NONDESTRUCTIVE

TESTING HANDBOOK
Second Edition

VOLUME 6
MAGNETIC PARTICLE TESTING

J. Thomas Schmidt
Kermit Skeie
Technical Editors
Paul McIntire
Editor

AMERICAN SOCIETY FOR

NONDESTRUCTIVE TESTING
 Copyright © 1989
AMERICAN SOCIETY FOR NONDESTRUCTIVE TESTING, INC.
All Rights Reserved.

No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any
form or by any means, electronic, mechanical, photocopying, recording, or othelwise, without
the prior written permission of the publisher.

Nothing contained in this book is to be construed as a grant of any right of manufacture, sale,
or use in connection with any method, process, apparatus, product or composition, whether or
not covered by letters patent or registered trademark, nor as a defense against liability for the
infringement of letters patent or registered trademark.

The American SOciety for Nondestructive Testing, its employees, and the contributors to this
volume assume no responsibility for the safety of persons using the information in this book.

Library of Congress Cataloging-in-Publication Data

Magnetic particle testing / J. Thomas Schmidt, Kermit Skeie, technical editors; Paul McIntire,
editor.
(Nondestructive testing handbook; v. 6)
Bibliographies.
Includes index.
ISBN 0-931403-03-0
1. Magnetic particle testing. 2. Nondestructive testing.. I. Schmidt, J. Thomas.
II. Skeie, Kermit. III. McIntire, Paul. IV. Series: Nondestructive testing handbook
(2nd ed.); v. 6.
TA417.3.M33 1989
620.1' 1278-dc20 89-6624
C.IP
Pl~blished by the American Society for Nondestructive Testing

PRINTED IN THE UNITED STATES OF AMERICA

PRESIDENT'S FOREWORD

'M any different means exist to disseminate technical computeIized data analysis. However, such automation will
information: conferences, technical papers, personal con- not eliminate the need for well trained and expetienced
tact, the classroom and handbooks. As the NDT industry's magnetic palticle inspectors.
technical society, ASNT is directly involved in all of these. A handbook on the magnetic pmticle test method should
This volume of the Nondestmctive Testing Handbook is include information on the physics and theory of the
the sixth in ASNT's ten-volume seties. \Vebster defines technique, written in a style that is accessible to both the
ell cyclopedia as "a work that contains information on all engineer and the technician. At the other end of the
hranches of knowledge or treats comprehensively a paltic- spectrum, it should also contain instmctions on how to do it
ular branch of knowledge." A handbook can be considered and, just as importantly, instructions on how not to do it.
an encyclopedia on a single topic, in this case the technology \Vhere and when the method should be applied, including
of magnetic particle testing. This book represents the the materials that can be tested, and the structural config-
colleeted and organized efforts of many authorities who urations where its use would be an advantage, should be
have generously made this contribution to the literature of 'discussed. Capabilities and limitations must be clearly
nondestructive testing. Distribution of such information is presented along with essential vatiables of the method and
clitically important to the advancement of nondestructive vatious techniques. Descliptions of equipment and its use,
testing throughout industry. in general and specific testing situations, are reqUired to aid
The magnetic particle method is much more than the inspectors in establishing their test procedures. Calibration
high school physics horseshoe magnet and iron filings guidelines are equally important. The qualifications of
demonstration. It is a simple yet highly effective tool for the ,personnel to develop techniques, to prepare procedures, to
detection of surface and near surface' discontinuities in perform the examinations and to interpret and evaluate test
ferromagnetic materials. In the general application, a ~11ag­ results should. be clearly delineated. Pitfalls and potential
netic field is applied to the test object. Discontinuities in the problems should be presented so that the reader might
matetial force a pOltion of the field to bridge the anomaly, anticipate and address them before they are encountered in
forming a n)agnetic leakage field. A powder of magnetic the performance of a test.
pmticles, either dry or in liquid suspension, is applied to the In short, the handbook must proVide the reader with most
component and the leakage field attracts and holds the of the information needed (1) to determine that the mag-
pmticles, forming a test indication. netic particle method should be used; (2) to establish the
Equipment used in the method ranges from a simple and techniques to apply it; and (3) to evaluate the test results.
inexpensive held-held yoke to large, wet hOlizontal systems These goals have been accomplished in this book by the
capable of deliveting thousands of amperes of electtic hard work and unselfish contributions of many volunteers
current thraugh the component or through a conducting who participated in preparing the text.
coil surrounding the part. Test objects range in size from a Several individuals merit special recognition. Technical
sevving needle to the very largest casting, forging or editors Tom Schmidt and Kermit Skeie, volume coordinator
weldment. Discontinuities onlv a few thousandths of an inch Rod Stanley, and mehic conversion reviewer Jan van den
long can he detected. . Andel have worked on this volume on their own time for
This highly sensitive, versatile and inexpensive technique more than four years and are directly responsible for its
is very dependent on operator capabilities and even more completion. Publication of this book has been accomplished
dependent on the operator's attention and persistence. An with the direction and editorial coordination of Paul
inspector's \·isual survey of the test surface is still the best McIntire of the ASNT headquarters staff.
and most effective way to detect and intelpret the indica- My personal thanks to all individuals, named and
tions formed by magnetic paItic1es. unnamed, who brought this project to its outstanding
\\'ith rapidly advancing technology and the need for conclusion.
inlTeased testing reliability, it is likely that automated
s~'stems will become increaSingly accurate for specialized Robert Baker
applications using advanced optical scanning systems \vith ASNT President (1988-1989)

iii
PREFACE

All of the Nondestructice Testing Handbook volumes are Alagnaflux Inspection (1940). The Amelican Society for
written as practical, instructional mateIials for NDT field Nondestmctive Testing gratefully acknowledges this debt
technicians. To fully explain the techniques and to encour- and extends thanks to Magnaflux Corporation, Chicago,
age accurate test results, the physics and mathematics of the Illinois, in pcuticular for access to its photographic archives.
technology are provided along with useful applications In an effoli to document nondestructive testing applica-
information. Each volume is wIitten and technically edited tions for an international audience and to acknowledge the
for ASNT by volunteers who provide their expeliise and differences in magnetic pmticle testing outside the United
donate their time to the project. States, this book was written and revie\ved by authorities
VVithin the NDT Handbook selies, this hook is the from across the US and from Germany, Unit~d Kingdom,
companion volume to ElectronUlgnetic Testing, the text Japan, Holland and Canada. Much effoli was made to
that covers eddy current, divelted flux and microwave tests. reference international speCifications, procedures and
Both books were produced with the guidance of the equipment whenever they differed from their American
ElectIical and Magnetic ~lethods Committee of ASNT's countelpmts: In addition, as \vith all the second edition
Technical Council, and the volumes share many authors and volumes, the International System of Units (S1) has been
reviewers. used throughout this text.
Within AS NT, the Handbook Development Committee Providing multiple units of measure is a time-consuming
deserves special recognition for its work on the seIies. task that affects all levels of book production, clerical and
Organized under the Methods Division of AS NT's Techni- . technical. Special recognition goes to Jan van den Andel of
cal Council, the Handbook Development Committee is VVestinghou?e Canada for again undertaking this difficult
directly responsibl~ for production of the Nondestntctive job. He has supplied metIic conversions for all of the books
Testing Handbook's second editio!l. As Handbook Develop- in the second edition, while at the same time serving as one
ment Director, RodeIic Stanley has chaired this committee of our most tlUstwOltlw and valued technical reviewers.
dming t~ publication qf two volllII\es and he;.along \vith all From among the publishing profeSSionals who contribut-
of the committee members, made special ·effoli to help 'ed to this project, ASNT extends 'its thanks to Hollis
complete this latest book. Humphries Black, who keyed all the text in this volume;
Even though this material was prepared by the authors Turner \Vainwlight Design, our technical drafting source;
especially for Magnetic Particle Testing, some of the first and Michael McGinn, ali director .for Lawhead Press. The
drafts had an unexpected consistency, if not a kind of quality of the finished volume is representative of their skills
familimity. It soon became clear thatthis occurred because and the dedication of all the book's contributors.
of the high quality and availability of several early publica-
tions in the field - Carl Betz's Principles of Magnetic Paul McIntire
Particle Testing (1967) and F.B. Doane's Principles of Handbook Editor

iv
 ACKNOWLEDGMENTS

Magnetic Particle Testing, Volume 6 of the Norulestruc- netic paIticle volume coordinator. In this role, Stanley
tiv£' Testing Handhook's second edition, is an effOlt by many motivated the book's contIibuting authors, served as the
volunteers to present a comprehensive gUide to this impor- priImuy reviewer and was also a reliable conhibutor of
tallt inspection technology. As Oliginally outlined and in extensive manuscript.
keeping with the objectives established by AS NT' s Elechi-
cal and I\1agnetic Committee, the volume presents both Volunteer Contributors
theory (lnd applications.
The three commentaries that follow were written to The organizers of this volume could not have accom-
highlight the volume's content and iI~tentions, and to sug- plished their task alone. The actual text of this book was
gest the book's place within the extensive literature of written by authors who are longtime expelts in their fields,
magnetic pmticle testing. More impOltantly, they also ac- and who have voluntaIily contributed their knowledge and
kllowledge the industry's debt to the volunteers who pro- time to this comprehensive effOlt. The editors wish to
duced all of this text. acknowledge and thank all those who contributed.

Future of the Magnetic Particle Method

Purpose of this Nondestructive The magnetic paIticle method is one of the oldest
Testing Handbook nondestructive tests. Some people think of it as low tech and
on the way to obsolescence. As some of the chapters of this
M'~IlY years ago, ,ASNT began what is pro~)ably its most volume amply demonstrate, magnetic particle testing is a,
sigllificant contribution to the science of nondestructive' complex procedure and its proper use demands as much
testing, namely publication of the Nondestnlctivp Testing skill and scientific foundation as any other method.
Halldbook. First published in 19.59, the NDT Handbook And the method is far from obsolete. It provides a fast,
aspired to he the encyclopedia which would include all the overall test of the surface and near surface integIity of
basic kI'lOwledge necessary for successful practice of nonde- magnetizable materials unmatched by other techniques, and
structive testing procedures. The first edition contained two it visibly indicates discontinuities directly on the test surfac-
volumes, edited bv Dr. Robelt C. McMaster. es, at their actual locations, and in nearly actual size.
Now, thirty ye<~rs later, NDT has grown and matured There is, in fact, a kind of mutual dependence among all
tremelldously. The sheer volume of knowledge has in- the nondestructive testing techniques. Common practice
creased to the pOint where a separate book is necessary for often includes a survey of new materials by the magnetic
eaeh tf-stillg teehnique and a group effort is required to particle method to find general discontinuity locations for
properly cover eve~) o~le inspection method. Such a group later detail testing by other methods. This author vividly
has been h~~)()]"ing for five years to organize this volume and remembers attending a meeting in 19.54 where the predic-
now thf~ eHOIt has come to fnlition. tion was made that the "obsolete magnetic particle method
[would] soon be replaced by sophisticated electronic meth-
Organization of the Volume ods." As it turns out, that predictor was long ago replaced,
but the test method has remained, increasing in use and
The organizational group for this book consisted of Paul value.
t\feIntire, Tom Schmidt, Kermit Skeie and Rod Stanley.
As all ASNT staff member and editor of the NDT
Use of the Nondestructive Testing Handbook
H(fI1.dh(~ok selies, I\ldntire is the one who produced all
drafts from allthor~' copy, coordinated the peer review This volume details the magnetic palticle test method in
process, made tht-' final t-'dits and sa\\' to all the details of all its facets. All levels of technical ability are addressed.
book production. PIinciples are covered simply for the begi~ner, but subject
Kermit Skeie and Tom Schmidt were the technical matter is also covered in great depth for the more accom-
editors who de\ised the book's basic format, outlined its plished practitioner. Thus, the material presented here will
eontent, enlisted most of the authors reviewed the manu- serve as a basic source of information on the magnetic
scripts f()l" technical accmac\' and th~mselves contributed particle method for many years to come.
mannscripts. -'
In addition to his duties as chair of the Handbook J. Thomas Schmidt
Deyelopment Committee, Rod Stanley acted as the mag- Volume technical editor

v
 reference materials by acknowledging and quoting from a
Role of this Nondestructive variety of intemational sources.
Testing Handbook A major application for magnetic paIticle methods is the
testing of piping and tubing, both new and used, by the
petrochemical industries. Methods, theories and techniques
Development of Magnetic Particle Tests for this application are included in this volume.
The magnetic particle method has histOlically been based Finally, for those with specific needs (electromagnetic
on empirical guidelines that are perpetuated because they theory, safety or existing specifications, for example) much
appear to work. Usually, no attention is given to the adage helpful and authoritative information is provided by this
that "we never know what we missed unless failure occurs." text.
With equipment, stmctures and products of all kinds This collection of data from dozens of authors could not
being deSigned with a relatively low safety factor, more be compiled or disseminated through a Single channel other
attention must be given to refining the magnetic particle than a professional society such as ASNT. This book is a
method to make it more reliable - application of fracture direct product of the Society's volunteer commitment and
mechanics demands predictable performance. effort.

Need for Standardization of the Technique Kermit Skeie

Volume technical editor
Why would a practitioner of NDT, NDI or NDE need a
book on magnetic particle testing? The method seems so
simple in concept and has served so well for fifty years that
nearly everyone knows all that is needed about the subject, Content of this Nondestructive
right? If so, then why in the 1980s would a controlled study Testing Handbook
on the magnetic particle method result in a probability of
detection less than 50 percent?
In the 1960s, a supervisor of NDT development in quality
Magnetic Testing.
engineering stated that no funding was available for studies Magnetic testing has been with us for very many years.
of magnetic particle testing because the test would soon be Magnetic methods have been used, and continue to be
replaced by other methods. High strength materials ~ould used, throughout industry to test ferromagnetic components
replace virtm~lly all the ferromagnetic materials in use 'and as divel'Se as bars, billets, railroad rails, pip~ 'lines, crmies,
the specifications on this old-line testing method were auto parts, landing gear, ball bearings and even nuts and
adequate and in place. Why then, nearly 10 years later, bolts. Objects are tested magnetically when new for con-
could qualified and certified inspectors find less than 60 formance to specifications and are also tested during and
percent of the known discontinuities in a monitored study? after use to determine the object's ability to continue in
Why, more than 20 years later, was their process standard service.
on magnetic paIticle testing rewritten? This volume of the The actual techniques employed are often determined by
Nondestmctive Testing Handbook was designed to help the size and shape of the test object, and the suspected
answer such questions. location of possible discontinuities. There is, however, a
common thread that holds magnetic test methods together.
Goals of this Volume This thread stalts with the fact that the test object must be
magnetized (in one or more directions, often at the same
For a long time, magnetic particle testing lacked the time), continues through the selection of the best possihle
reference standards so common and essential to other NDT sensor to detect the relevant test pararneters and ends "vitll
methods. Several such standards are described ill detail in the acceptance or rejection of the material in accordance
this book. with certain predetermined criteria.
This volume also reveals some misconceptions about the
magnetic particle method. One example is the often stated
Electromagnetic Testing
procedure that an effective magnetic field on a cylinder,
produced \vith a current canying conductor, is four times tvlethods for magnetizing test objects include permanent
the diameter of the conductor. This can be neither theoret- magnets and the use of electric current. Permanent magnet
ically nor practically supported. The fallacy was, in fact, first methods have been given a new lease on life \vith the
revealed by artificial discontinuity standards. introduction of rare emth magnets (neodymium iron and
Data and techniques established elsewhere in the world samarium cobalt). Electric current methods include the
have been largely ignored in the United States. This book creation of an electromagnet (commonly termed a yoke
has tlied to expand the scope of previously published which, Shipped to its essentials, is a coil with a high
 pcrlll('aiJility core), the use of air cored coils and threader Lead Authors and Reviewers
hars. and tl](' passage of current through the test object
itw/r .\11 of tlH'sC methods create magnetic flux in the tested Every Section of this Volume was assigned its own
IIlaft'rial. coordinator who in most cases served as the primmy author.
.'\t this point, however, the specialized topic of magnetic All of these individuals and the contlibutors they organized
particle testillg breaks away from the general field of deselve special recognition. Each of those named below
(·I(·drollla<Tl)ctic testing (covered in Volume 4 of the Non- produced their part of this book as a volunteer - many on
d('sfrll(,/ir;~ Testillg Ham/hook series). Electromagnetic tcst- their own time with their O\Vll resources, in an effOlt
ill~ is a lerlll that designates the entire subject of magnetic to make the technology available to the widest possible '
alld (·icd ric Helds and their interactions with materials, and audience.
so illl'Il1(I('s eddy current methods, magnetic flux leakage, Equally important is the large group of individuals who
lllino\\,lH' and thermoelectric methods. Magnetic particle reviewed all the text in this volume for technical content.
t('sting turns Ollt to he a branch of magnetic flux leakage The peer reviewers are also listed below and they deseIVe
iIlSPt'di()ll. much credit, for donating the time needed to refine these
documents for presentation. Because of the reviewers, many
changes and improvements were made pOSSible, to extend
the book\ coverage and to maintain its continuity.
Coverage of this Volume
Rodedc Stanley
II ist! lJ'ically, from the testing viewpoint, magnetic pmticle Handbook Development Director
illspediol1 has heen considered a stand-alone method. In
reality. what makes it so is its sensor: that collection of oddly
shaped and sized highly permeable matelials that we cause
to pass throllgh magnetic leakage fields from tight Handbook Development Committee
dis('oll'illllities ill the'hope that some of them might be held
h~ t 1)(' leakage field. Mike Allgaier, GPU Nuclear
This lH)ok attempts to proVide both the theory behind and Al Birks, Battelle MemOlial Institute
the kt'IIJ)()lot,TY of this bscinating topic. It covers the Bruce Bolis, Titanium Metals Corporation
dis('olltillilities detected bv the technique· methods for Jacques Blignac, Combustion Engineeling
Illagn<'tizillg rnatelials; metllOds for testing ~:vhether mate- Al Brown, Lawrence Livermore LaboratOlY
rials might he suffiCiently magnetized for the detection of Lawrence Blyant, Los Alamos National Laboratory
c('liaill t.''Pl's of discontinuities; and some of the codes and Chlistian Burger, Texas A&M University
specifications that have evolved around the technique. J olm Cavender, Duke Power Company
Sal(·t~ II<lzards are also addressed (try to take a black light \Villiam Chedister, Circle Chemical Company
Ollto all oil rig in the NOlth Sea!) and a group of contlibutors Peter Collins, Exxon Research Llnd Engineeling
ha\(' shan·d \'~lJiOllS industrial applications. Gene Curbow, Center for Applied Welding Research
H.~·<ld(TS are cautioned to keep an open mind when Mark Davis, ].A. Jones Applied Research Center
P(·r{o).'llJing magnetic tests, especially in relation to the many Boro Djordjevic, Martin Mmietta Laboratolies
sp,('cilic<ltions that have sprung up around the technology. Robert Green, Jr., The Johns Hopkins University
\\ hat Ilia.'" \\"Ork in a pmticular situation may not be Patricia Bouta Hearney, General Elecbic Company
addr('~s('d in specification documents or it may not be Edmund Henneke, Virginia Polytechnic Institute
:lddn'ss('d (·oITcctly. As inspectors, our job is to detect and Frank Iddings, Southwest Research Institute
Illterpret discontinuity indications not to be hide bound bv Ron Miller, MQS Incorporated
d(.H.·1JIIl(·~ltatioll. \\'e 'hope that tl~is text provides enoug!l Scott ~liller, Aptech Engineering Services
stlllllllatmg 1Ilat(Jlial that its readers will be able to pelform \Villiam Mooz, The Met-L-Chek Company
IIlagll:·ti(' paliiclt:' tests with a better knowledge of the Hamid Naveb-Hashemi, NOltheastem University
!I'VIIlIl(l'll' alld to try Ilew versions of older methods with a Bruce Pelligdno, EMCO Division of Intertest -
til'ep!'r 1Illderstandi;lg of their foundations.
J. Thomas Schmidt, ].T. Schmidt and Associates
'1'1 li~ Illat('rial was gathered from a wide vmietv of sources Charles Sherlock, CBI
alld \\(. Ilan' tried to include some of the le~s common Amos Sherwin, Sherwin IncOlporated
t('Clllliqllt's and some techniques that may be used in
Kermit Skeie, Kermit Skeie Associates
Europe hut Ilot in tilt:' United States. \Ve h~ve covered as
Rodelic Stanley, International Pipe Inspectors Association
lI1.all~· appli(,~lti()lIS as possible and regret if we have not Henry Stephens, J.A. Jones Applied Research Center
plO\']dl'd lIdormatioJ) that is specific to some inspection
prollJc.l1ls. Ming-kai Tse, Massachusetts Institute of Technology
Frank Vicki, Pratt and \\7hitney Aircraft

vii
 Michael Urzendo\Vski, Babcock and vVilcox
Volume 6 Lead Authors Jack Veno, General Electric Company
George Watson, BP Petroleum Development Limited
David Atkins, Packer Engineering Laurence Wong, ~lagnaflux Corporation
Bernie Boisvert, B&B Technical Services
John Bronk, Sandra T. Bnmk and Associates
Charles Exton, Ardrox Limited Volume 6 Reviewers
John Flaherty, Flare Technology
Bruce Graham, Magnaflux Corporation Hank Bogart
Donald Hagemaier, McDonnell Douglas Aircraft James Borocki, Ardrox Corporation
Company
Jim Chase, Magnaflux Corporation
Larry Haller
Dave Cornall, Inspectech Limited
Nathan Ida, University of Akron
David Dier, Magnaflux Corporation
Thomas Jones, Industrial Quality Incorporated Charles Exton, Ardrox Limited
Ken Kremer, McDonnell Aircraft Company
Richard Gaydos, Lucas Aerospace
Arthur Lindgren, Magnaflux Corporation Donald I-Iagemaier, McDonnell Douglas Aircraft
Calvin McKee
Company
Henry Ridder, Professional Engineering Systems Chuck Harpster, Advance Test Equipment
J. Thomas Schmidt, J. T. Schmidt Associates Dean Hanis, Universal NDT
Kenneth Schroeder, Schroeder and Associates William Holden, Spectronics Corporation
Roderic Stanley, International Pipe Inspectors Association Thomas Jones, Industrial Quality Incorporated
Don Lorenzi
George Luciw, American SOCiety for Testing and
Volume 6 Contributors Materials
Brian McCracken, Pratt and Whitney Aircraft
John Brunk, Sandra T. Bronk and Associates Patrick McEleney, US Army Ma~erials Technology
William Burkle, Precision Tubular Inspection Laboratory .
William Chedister, Circle Chemical Company Calvin McKee
Karl Deutsch, K. Deutsch GinbH and Company Will Ptomey, Econospect COlporation .
Brandon Fraser ~aul oRistuccia, Boeing Comm~rcial Ailphu'l.e Company
Lawrence Goldberg, Sea Test Services Sam Robinson, Sherwin Incorporated
Daniel Hafley, General Testing Laboratories James Rudins " ""
Donald Hagemaier, McDonnell Douglas Aircraft Hussein Sadek, Law Engineering Industlial Services
Company J. Thomas Schmidt, J.T. Schmidt Associates
Gunther Hanle, Ridge Incorporated Kenneth Schroeder, Schroeder and Associates
Grover Hardy, Wright Patterson Air Force Base Kermit Skeie, Kermit Skeie Associates
Dieter Kaiser, Mannesmann Forschungsinstitiit M. Stadthaus, Bundesanstalt fUrMaterialprufung
Art Lindgren, Magnaflux CorporatioI) Roderic Stanley, International Pipe Inspectors Association
David Lovejoy, Burmah Castrol Limited Meg Steigerwald, Society of Automotive Engineers
Brian McCracken, Pratt and Whitney Aircraft Lydon Swartzendrober, National Institute of Standards
Patrick McEleney, US Army Materials Technology and Technology
Laboratory Ron Sweet, Ultraviolet Products Corporation
John Mittleman, fvlaJine Inspection Technology vVillys Thomas
Stanley Ness Jan van den Andel, vVestinghouse Canada
Henry Ridder, Professional Engineeling Systems Jack Veno, General Electric Company
Kenneth Schroeder, Schroeder and Associates Frank Vicki, Pratt and Whitney Aircraft
Leo Shelwin, General Electric C0l1)oration
Kermit Skeie, Kermit Skeie Associates
Rodelic Stanley, International Pipe Inspectors Association Major contributors to individual chapters are listed
Lydon Swartzendruber, National Institute of Standards alphabetically on the Section title page under the name of
and Technology the primary author.

viii
INTERNATIONAL SYSTEM OF UNITS IN MAGNETIC PARTICLE TESTING

asmicrowatts per square centimeter (p:vVocm - 2). The units

Origin and Use of the SI System of measure llsed for the visible light range cannot be Ilsed
for ultraviolet encr!-,1)'.
Tht' S1 system was designed so that all branches of Visihle light intensity: expressed as lux (Ix). "Vas for-
\t 'i( '11('(:
('ould lise a Single set of interrelated measurement merly measured in footcandle (ftc). One lux is equal to
IIllitS. TIJ('se Ilnits are modifled in speCified ways to make 0.1 footcandle.
tll(,111 adaptable to the needs of individual disciplines.
"'itlum! the SJ system, this Nondestructive Testing Hand-
hlllik \'(JllIll\(' could have contained a confusing mix of TABLE 1. Base SI Units
11111)('rialll1lits, old cgs metlic units and the units preferred Quantity Unit Name Unit Symbol
h: certaill localities, industries or scientific specialties.
l ~~(' of the SI system also provides a mathematical length meter m
;l(kaIlLtgc. III equations, SI units and their powers balance mass kilogram kg
Oil ('adl side of the equal sign. This provides a double-check
time .. second s
of ;I('('III'a('y: all equation error will reveal itself not only electric current ampere A
tl,rolJ,~ll. all imhalance but through the different units
thermodynamic temperature kelvin. K
amount of substance' mole mol
("watdl Il\ ·the imhalallce~ .
luminous intensity candela cd
plane angle radian rad
solid a~gle steradian sr
SI Units for Magnetic
Particle Testing .
TABLE 2. Derived SI Units
Specific 1Inits used in this vorume are mainly those for
Relation to
1Ilaglll'liSIIl, \'isihle light and ultraviolet radiation: Originally,
Quantity Name Symbol Other SI Units
tlws{' IIIl its wc re developed by scientists using the cgs
(('('Iililllt'ter gram second) metric system. With the intro- frequency hertz Hz 1-s- 1
dll('!ioll of SL these quantities were rearranged into a selies force newton N kg x m-s- 2
oi hasl' (mits (Tallie 1) and delived units (Table 2). pressure (stress) pascal Pa N-m- 2
energy (work) joule j Nxm
For maglletic theOJies, this meant the removal of inter-
power watt W jos- J
Ilwdiat{· Il1lits (sl1ch as the unit pole) and made possible a
electric charge coulomb C A x s
din·('t C()JlwrSlon from flux cut per second to voltage. The 51
electric potential volt V WoA-'
1I1lits illc\wl(' the weher (\Vh), the tesla (T) aIId several
capacitance farad F CoV-1
d('ri\('(llIllits. TesIa is a large nnit and is often used with the
SI lI11dtiplins (Table 3). Listed below are the four basic
electric resistance ohm n VoA-J
conductance siemens S AoV-1
lInits l(lll1ld ill this text.
magnetic flux weber Wb V x s
;\Iag~wti~ fi~ld strength: expressed in ampere per lneter magnetiC flux density tesla T Wb-m- 2
1.\~1I1 !. \ \ as ionnerly measured in oen·ted (Oe), a nonex- inductance henry H WboA-1
I:\tlllll; pll:~il'aJ agl~nt f'nabling analysis of complex magnetic temperature degrees °C K - 273.15
hl,~d prol ,ll~Il's. Olle ampere per meter is equal to about Celsius
I.,) ).., 10 :.. 0('. luminous flux lumen 1m cd x sr
;\tagm:tie flux density: expressed in weber per square illuminance lux Ix Imom- 2
1I1~'kl: '\\ hOIll 2) or t(,8[0 (T) to indicate flux per unit area. radioactivity becquerel Bq los - I
\\ as i()nllcrh- measllred in gauss (C). One tesla is equal to radiation absorbed gray Gy jokg- J
J() I ll;allss dose
radiation dose sievert Sv jokg- J
t'lh· .. ,:iolet irradiance: expressed in watts per square
equivalent
11l!'tIT
-.. \ \ 'as r
1\\ °11)- .0'1 lormerI d in cgs mehic units
y measure

ix
 SI Multipliers
TASLE 3. 51 Multipliers
Very large or very small units are expressed using the SI Prefix
multipliers, prefixes that are usually of 10 3 intervals. The Symbol Multiplier
range covered in this text is shown in Table 3. These tera T 10 12
multipliers become a property of the SI unit. For example, giga G 109
a centimeter (em) is 11100 of a meter, and the volume unit, mega M 106
cubic centimeter (cm3 ), is (11100P or 10 -6 m 3 . Units such kilo k 103
as the centimeter, decimeter, decameter and hectometer hecto h
deca
102
are avoided in technical uses of the SI system because of da 10
their variance from the 10.3 interval. deci d 10- 1
centi c
Every effort has been made to include all necessary SI 10- 2
milli m
and conversion data in the text of this book. If questions micro
10- 3
remain, the reader is referred to the information available J-L 10- 6
nano n
through national standards organizations and to the special- 10- 9
pica p
ized information compiled by technical societies (see ASTM 10- 12
E-380, Standard Metric Practice GUide, for example).

Jan van den Andel

Westinghouse Canada

x
 ·t,t"""""""""""""""""""'"
""","""'"11""'""".

~r
CONTENTS
SECTION 1: FUNDAMENTALS OF PART 8: TYPES OF MAGNETIZING
MAGNETIC PARTICLE TESTING ........... . 1 CURRENT ................................................. . 27
Alternating Current ................................. . 27
P;\HT 1: INTHODUCTION TO MAGNETIC Half-Wave Direct Current ........................ . 27
P:\1\TICL£ TESTING ..... '" ......................... . 2 Full-vVave Direct Current ........................ . 28
Capahilities and Limitations of MagnetiC Three-Phase, Full-Wave Direct Current ..... . 28
Particle Techniques ............................. . 2 PART 9: DEMAGNETIZATION
Prindples of Magnetic Particle Testing ...... . 2 PROCEDURES .......................................... . 30
PABT 2: FABHICATION PROCESSES AND Justification for Demagnetizing ................. . 30
i\IACNETIC PARTICLE TEST Methods of Demagnetization .................... . 30
APPLICATIONS ........................................ .. 3 Demagnetization Practices ........................ . 32
Basic Ferromagnetic MateIials Production .. . 3 PART 10: MEDIA AND PROCESSES IN
Inherent Discoiltinuities .......................... .. 3 MAGNETIC PARTICLE TESTING ............ .. 34
Pri mmy Processing Discontinuities ............ . 3 MagnetiC Particle Properties ..................... . 34
Forging Discontinuities ............................ . 6 Effects of Particle Size ............................ .. 35
Casting Discontinuities ............................. . 7 Effect of Particle Shape .......................... .. 35
\Veldment Discontinuities ........................ . 7 Visibility and Contrast .............................. . 36
ManuhlCtming and Fablication Particle Mobility ...................................... . 36
Discontinuities .................................... . 8 Media Selection ...................................... . 37
Service Discontinuities ............................. . 9 Magnetic Particle Testing Processes .......... .. '37
PAnT :3: I\1AGNETIC FIELD THEORy .......... .. ·12 Conclusion ............................................. . 38
Magnetic "Domains .................................. . 12
Magnetic Poles ....................................... . 12 SECTION 2: GLOSSARY FOR MAGNETIC
'l\,pes of Magnetic Materials ..................... . 13 PARTICLE TESTING .............................. .. 41
Sources of Magnetism .............................. . 13
PAHT 4: MAGNETIC FLUX AND FLUX SECTION 3: A HISTORY OF MAGNETIC
LE/\KA.C;E ................................................. . PARTICLE TESTING ............................... . 49
15
Circular Magnetic Fields .......................... . 15 PART 1: ORIGINS OF THE MAGNETIC
l,ongit"dinal Magnetization ...................... . 15 PARTICLE INDUSTRy .............................. . 50
I\1aglletic Field Strength .......................... .. 15 The 1920s: Era of DiscovelY .................... .. 50
SllbsllI'face Discontinuities ........................ . 16 The 1930s: Years of Development ............ .. 50
Eff(x:t of Discontinuity Orientation ............ . 16 PART 2: EXPANSION OF THE MAGNETIC
Formation of Indications ................... " ..... . 17 PARTICLE INDUSTRy .............................. . 56
PAHT 5: ELECTRICALLY INDUCED The 1940s: Organization and Growth ........ .. 56
I\1:\CNETISM ............................................ . 18 The 1950s: Developing Markets ................ . 59
(:ircu!aJ' Magnetization ............................. . 18 The 1960s: Industry Challenges ................ . 66
~Iagn('ti<: Field Direction ........................ .. 19 The 1970s; Advanced Organization and
L()ngitudinal Magnetization ...................... . 19 Systems ............................................. . 72
!\ I IIltidirectiollal r-"iagnetization .................. . 20 The 1980s: IndustIy Improvements ........... . 73
PAHT fi: T\tAGNETIC PARTICLE TEST Conclusion ............................................. . 73
SYSTEI\IS ..... ............................................. .
~ 21
Stationary Magnetic Particle Test Systems ... 21 SECTION 4: DISCONTINUITIES IN
p( lWer Packs .... 21 FERROMAGNETIC ALLOyS .................... . 75
"lo!lik' and PO~~tbi~"T~~'~i~g' u'~i~~' ::: ::: ::::: :: 21 PART 1: INHERENT DISCONTINUITIES ...... . 77
Prods and Yokes 22 Cold Shut ............................................... . 77
PAHT 7: FEHHOMAG·NETic·~1AT'ERiAL ........ · Pipe ....................................................... . 77
Cll:\H.-\CTERISTICS .................................. . 24 Hot Tears ............................................... . 77
\Iagllctic Flux and Units of Measure ........ .. 24 Blowholes and Porosity ............................ . 77
\lagnetic Hysteresis ................................. . 24 Nonmetallic Inclusion~ ............................. . 78
\iaglletie Permeability ............................ .. 26 Segregation ............................................. . 78

xi
 PART 2: PRIMARY PROCESSING PART 5: MAGNETIC CIRCUITS AND
DISCONTINUITIES ................................... . 81 HYSTERESIS .................... : ........................ . 117
Seams .................................................... . 81 Magnetic Circuits .................................... . 117
Lalninations ............................................ . 81 Hysteresis ............................................... . 118
Stringers ................................................ . 81 Minor Hysteresis Loops ........................... . 121
Cupping ............................................... '" 82 Hysteresis Curve as a Classifier ................. . 121
Cooling Cracks ........................................ . 82 Energy Lost in a Hysteresis Cycle ............ .. 122
Forged and Rolled Laps .......................... .. 82 Eddy Current Losses .............................. .. 12.3
Internal and External Bursts ..................... . 83 PART 6: CHARACTEHISTICS OF
Hydrogen Flakes ..................................... . 83 ELECTROMAGNETIC FIELDS ................. . 124
\Velding Discontinuities ........................... . 84 Energy in the Electromagnetic Field ......... . 124
PART 3: SECONDARY PROCESSING Force in the Magnetic Field ..................... . 125
DISCONTINUITIES ................................... . 91 Torque in the Magnetic Field .................. .. 125
Grinding Cracks ...................................... . 91 Inductance ............................................. . 126
Heat Treating and Quench Cracks ............. . 91 PART 7: MODELING ELECTROMAGNETIC
Pickling Cracks ....................................... . 92 FIELDS ..................................................... . 127
Machining Tears ..................................... . 92 Modeling of Leakage Fields .................... .. 127
Plating Cracks ......................................... . 92 Experimental or Empirical Modeling ......... . 129
PART 4: SERVICE INDUCED Analytical Modeling ............................... '" 129
DISCONTINUITIES ................................... . Numerical Modeling ................................ . 130
94 PART 8: MATHEMATICS OF
Fatigue Cracking ..................................... . 94
Creep Cracking ...................................... .. ELECTROMAGNETIC FIELD STUDIES ... . 133
95
Stress Corrosion Cracking ........................ . Vector Algebra arid Calculus ..................... . 133
97
Hydrogen Cracking .................................. . Vector Identities ...................................... . 134·
98
Vector Theorems: .................................... . 134

SECTION 5: BASIC SECTION 6: MAGNETIZATION METHODS 147

ELECTROMAGNETISM .......; ..............., ... . 10.1,
PART 1: DESCRIPTION OF MAGNETIC
PART 1: FUNDAMENTALS OF FIELDS ...................................................... . 148
ELECTROMAGNETISM .......................... n. 104 Magnetized Ring ..................................... . 148
Coulomb's Law ....................................... . 104 Bar Magnet ........................................ : ... . 148
Use of Maxwell's Equations ..................... .. 104 PART 2: MAGNETIZATION WITH ELECTRIC
Units of Measure and Terminology ........... ,. 104 CURRENT ................................................ .. 1.50
PART 2: FIELD RELATIONS AND Circular Magnetization ............................. . 150
MAXWELL'S EQUATIONS ......................... . 106 Circular Magnetization of Solid Test
Linearity, Homogeneity and Isotropy ......... . 106 Objects .............................................. . 150
Static Fields ............................................ . 107 Circular Magnetization of Hollow Test
The Magnetic Vector Potential .................. . 107 Objects .............................................. . 152
Biot-Savmt Law ...................................... . 108 Longitudinal Magnetization ...................... . 152
Coil Magnetization .................................. . 1.52
PART 3: ELECTHOMAGNETIC FIELD TYPES
Field Flow Magnetization ......................... . 153
AND BOUNDAHY CONDITIONS .............. .. 109 PART 3: FACTORS CONTROLLING
Steady State Alternating Current Fields ..... . 109 MAGNETIZATION ..................................... . 154
Time Dependent Fields ........................... . 110 Material Propmties .................................. . 1.54
Electromagnetic Boundary Conditions ....... . 111 Types of Magnetizing Current. .................. . 154
PART 4: EFFECT OF MATEIUALS ON PART 4: DIRECTION OF THE MAGNETIC
ELECTROMAGNETIC FIELDS ................. . 114 .FIELD ....................................................... . 158
Matelial Properties and Constitutive Circular Magnetization ............................. . 158
Helatiolls ............................................ . 114 Limitations of Parallel Magnetization ......... . 159
Diamagnetic Materials ............................. . 114 Longitudinal Magnetization ...................... . 159
Paramagnetic Materials ............................ . 115 Combined Circular and Longitudinal
Ferromagnetic Materials .......................... . 115 Magnetization ................... " ................ . 160

xii
"""""",.... ", ••••• tt'II"'III'IIIIII'IIII'II'IIIIIIIIIIIII'II"I"''''1I11II

PART 5: MULTIDIRECTIONAL PART 2: WET METHOD TESTING

MAGNETIZATION ..................................... . 161 MATERIALS ............................................... 20.5
Combined Direct Current Fields .............. . 161 \Vet Method Particle Characteristics........... 20.5
Combined Direct Current and Alternating Oil Vehicles for \Vet Method Particles. ....... 206
Current Fields .................................... . 161 vVater Vehicles for \Vet Method Particles.... 207
Combined Altcmating Current Fields .... , ... . 162 Bath Contamination.................................. 208
PART 6: CIRCUMFERENTIAL Bath Preparation...................................... 209
MAGNETIZATION OF PIPE ...................... . 16.5 Bath Maintenance.................................... 209
Specifications for Testing Oil Field The Settling Test.... ...... ..... ..... ...... ..... ....... 210
'fubulars ............................................ . 16.5 Applications of \Vet Magnetic Pmiicles ....... 211
Magnetization Methods for Oil Field Viewing \Vet Method Indications............... 213
Applications ....................................... . 16.5
Use of B-H Curve ill Setting SpeciBcatiol1s . 167 PART 3: STUDIES OF MAGNETIC PAHTICLE
Typical Requiren:lents for Direct Current SENSITIVITY .................................... ,. ........ 214
Magnetization ................................... , .. 167 Studies to Determine Wet Method PaJiicle
Pulsed Current Magnetization ................... . 168 Sensitivity........................................... 214
Practical Testing Situations ....................... . 169 Dry Powder Sensitivity.............................. 225
Analysis of Pulse Current Magnetization .... . 169 Conclusion ...... ........................................ 22.5
Typical Values for L, C and R. .................. . 170
Design Considerations ............................. . 172
SECTION 9: DETECTION AND
Magnetization Recommendations .............. . 172
EVALUATION OF MAGNETIC
Current Pulse Effectiveness ...................... . 173
PARTICLE TEST..INDICATIONS ......... ...... 227
Use of Inductive Ammeters ...................... . 174
PART 7: MAGNETIC FLUX IN TEST
PART 1: CONTRAST AND CONTRAST RATIO .229
OBJECTS \VITH ?01~PLEX SHAPES ......... . 176
Brightne~s Contrast .............. : .................. : 229
Color Contrast......................................... 229
SECTION 7: MAGNETIC LEAKAGE FIELD
MEASUREMENTS .................................... . 179 PART 2: DETECTION DEVICES ........ :............ 230
PART 1: FUNDAMENTALS OF MAGNETIC. The Human Eye as a Detector .................. 230
FLUX LEAKAGE FIELDS ................ '" ....... . 180 Scanning Detectors................................... 231
Inducing Magnetic Flux Leakage .............. . 180 Television Detectors................................. 232
Effect of Flux Leakage on False
Indications ........ , .. , .................... " ....... . 181 PART 3: INTERPRETATION OF
Why Particle Indications Form ................. , 181 DISCONTINUITY INDICATIONS................ 233
PART 2: FLUX SENSITIVE DEVICES ............. . 183 Testing for Subsurface Discontinuities........ 233
Voltage Developed between the Ends of Ii Choosing the Magnetic Particle Technique.. 233
Straight "'ire ...................................... . 183 Determining the Nature of an Indication .... 233
Simple Pickup Coils ................................. . 185 Nonrelevant Test Indications..................... 234
Hall Element Sensors .............................. . 187 False Test Indications............................... 235
Magnetodiodes ........................................ . 190
PART 4: PROCESS CONTROL OF
Ferroprobes ............................................ . 191
MAGNETIC PARTICLE TESTS ................. ,. 236
Large Volume Magnetic Field Indicators .... . 192
Technique Sheets..................................... 236
Control of Wet Method Pmiicle
SECTION 8: MAGNETIC PARTICLES AND
Concentration .................................... ,. 236
PARTICLE APPLICATION ....................... . 199
Paliicle Control at the Indication Site......... 239
PART 1: DRY METHOD PARTICLE Brilliance and Contamination Tests............. 239
CHARACTERISTICS .................................. . 201 Viscosity Test........................................... 239
Drv Method Pmiicle Charactedstics .......... . 201 Steel Ring Test .... .... ....... .................. .... ... 240
D~' Paliicle Uses .................................... . 202 Verifying Illumination............................... 240
Application of Dry Magnetic PaJiicles ........ . 202 Visible Light Requirements....................... 241
Viewing and Inteq>reting Dry Particle Test Calibration for Current Output.................. 241
Indications ............................. , ........... . 203 Verifying the Magnetic Field ... ..... ...... ....... 241
Dry Magnetic Particle Specifications .......... . 204 Quick Break Test ..................................... 242

xlii
 SECTION 10: PROCESS AUTOMATION OF PART 2: PRESSURE SENSITIVE TAPE
MAGNETIC PARTICLE TESTING. ...... ..... 245 TRANSFERS............................................... 27:3
Dry Method Tape Transfers...................... 273
PART 1: TEST OBJECT HANDLING FOR \Vet Method Tape Transfers...................... 273
MAGNETIC PARTICLE TESTS ........ ........ ... 246 Archival Quality of Tape Transfer Records.. 274
Manual Handling with Automated Testing... 246 PART 3: FIXING COATINGS FOR TEST
Fully Automated Handling........................ 246 INDICATIONS............................................ 27,5
Designing Fully Automated Magnetic Application of Fixing Coatings................... 27,5
Particle Testing Systems....................... 246 Archival Quality of Fixing Coatings............. 275
PART 2: MONITORING AUTOMATED PART 4: ALGINATE IMPRESSION RECORDS 276
TESTING EQUIPMENT.............................. 248 Application of Alginate Impression
Automated Magnetizing Techniques........... 248 Techniqiles ......................................... 276
Computer Components of Automated Archival Quality of Alginate Impression
Testing Systems................................... 248 Records.............................................. 276
Monitoring Automated Test Systems for PART 5: MAGNETIC RUBBER METHODS...... 277
Malfunction. . . .. . . . . .. . . ... ... . . .. . . .. . . . . . . .. ... . . . 249 Application of Magnetic Rubber
PART 3: AUTOMATING THE MAGNETIC Techniques......................................... 278
PARTICLE TESTING PROCEDURE ........... 250 Preparation and Use of Magnetic Rubber.... 280
History of Automatic Scanning................... 250 Archival Quality of Magnetic Rubber
Production Automation of the Magnetic Records. . .. . . .. . . . . . .... . . . . . . . . .... . . .... . . . .. . . . .. . 282
Particle Method................................... 252 PART 6: PHOTOGRAPHY OF MAGNETIC
PARTICLE TEST INDICATIONS................. 283
PART 4: MONITORING THE PROCESSING Cameras for Indication. Recording .............. 284
STAGE OF MAGNETIC PARTICLE TESTS 254 Accessories for Indication Recording.......... 284
Verification of Cleaning Operations ............ 254 Choosing Film for Indication Recording...... 284
Verification of Magnetization..................... 254 Filters for Indication Photography.............. 285
Monitoring the Magnetic Particle Bath....... .255 Exposure Estimation and Metering............. 286
PART 5: AUTOMATING THE OBSERVATION . Bracketing Exposures...................... ....... ... 288
STAGE OF MAGNETIC PARTICLE TESTS 257 Lighting for Test Indication Recording....... 288
Methods of Scanning .................. ~............. '257 Developing· Procedures for Test Indication .
Television Scanning Systems ....... :.............. 257 Photographs .......................... '" . .. .... . . .. 289
Flying Spot Scanning Systems~................... ,257 Archival Quality of Indication Photographs.. 290
Application of the Flying Spot System ........ 258 Electronic Recording Techniques............... 290
PART 6: AUTOMATING THE Conclusion .............................................. 290
INTERPRETATION STAGE OF SECTION 12: DEMAGNETIZATION OF
MAGNETIC PARTICLE TESTS................... 260 TEST OBJECTS ..... ...................... .............. 291
Optical Pattern Recognition....................... 260 PART 1: DEMAGNETIZATION AND
Pattern Recognition with Dedicated Digital RESIDUAL MAGNETISM ...... ...... ............... 292
Processors. . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
Ferromagnetic Materials ............. .............. 292
Simple Microprocessor AlgOrithm............... 262
Requirement for Demagnetization.............. 292
The SRI Algorithms.................................. 264
Common Sources of Residual Magnetic
Neighborhood Processing .......................... 266
F'ields .. , ................ " . . . . . . . . . .. . . . . ... . . . . . . . . . . 293
PART 7: EXAMPLES OF PARTIAL AND Types of Residual Magnetic Fields............. 293
FULLY AUTOMATED MAGNETIC PART 2: PRINCIPLES OF
PARTICLE TESTING SYSTEMS .................. 267 DEMAGNETIZATION ......... .... ....... ..... ..... ... 295
Magnetic Hysteresis ............................... '" 295
Retentivity and Coercive Force.................. 295
SECTION 11: RECORDING OF MAGNETIC Basic Principle of Demagnetization.. . . .. . . . . .. 29,5
PARTICLE TEST INDICATIONS...... ........ 271 PART 3: SUMMARY OF DEMAGNETIZATION
PROCEDURES .... ....... ....................... ......... 297
PART 1: BASIC RECORDING OF MAGNETIC Alternating Current Demagnetization ......... 297
PARTICLE TEST RESULTS .... ...... ........ ...... 272 Direct Current Demagnetization................ 298
Drawing and Written Descriptions............. 272 Specialized Demagnetization Procedures..... 299

xiv
 PART 4: SELECTING A DEMAGNETIZATION
PH.OCEDURE ............................................ . PART 5: SAMPLE SAFE1Y HEQUIREMENTS ..
Limitations of Alternating Current 300 328
Purchaser's Utility Services and Conditions ..
328
~fethods " ............................. '" .......... . 300 EqUipment DeSign and Construction .......... 329
Maximum Effective Field Strength ............ . EqUipment Safety, Health and
300
.5 Reversing Direct Current ......................... . Environmental Criteria .........................
PART 5: MEASURING EXTERNAL FIELD 300 331
5 Other Safety Requirements ....................... 333
5 STRENGTH ............................................... .
302
Field Indicators ..................................... '"
302 SECTION 14: REFERENCE STANDARDS
LaboratOIY Instruments ............................ .
303 AND ARTIFICIAL DISCONTINUITY
Measllling Techniques .............................. .
PABT 6: TYPICAL DEMAGNETIZATION 303 INDICATIONS .......................................... . 337
PROBLE!vlS ............................................... .
304 PART 1: FUNDAMENTALS OF REFERENCE
Demagnetizing Field Strength Too Weak .. ..
304 STANDARDS FOR MAGNETIC PARTICLE
Alternating Current Demagnetization and
Direct Current Magnetization .............. . TESTING ....................................................
304 338
Test Object Orientation ............................ . EmpiIical Rules for Using Reference
304 Standards .. '" ...................... '" ..............
Poor Length-to-Diameter Ratio ................. . 338
304 Needs for a Known Indicator .....................
Magnetic Shielding .................................. . 339
304
Calibration of Field Indicators .................. . PART 2: REFERENCE STANDARDS FOR
304
Demagnetization Specifications ................. . SYSTEM EVALUATION ...............................
PART 7: DEMAGNETIZING EQUIPMENT .... .. 305 341
306 Tool 'Steel Ring Standard ..........................
Summary of Alternating .Current . 341
Reference Staudartl Test Blocks ................. 343
Demagnetization Equipment ................ .
Summary of Reversing Direct Current 306 PART 3: MAGNETIC DISCONTINUITY
Demagnetization Equipment ................ . STANDARDS .... : .........................................
306 345
PART 8: DEMAGNETIZATION OF Pie Gages and Raised Cross Indicators ....... 345
ELONGATED TEST OBJECTS .................. .. Shim Discontinuity Standards .................... 346
307
Circumferen tial Remagnetization ............... . PART 4: ELECTRONIC REFERENCE
307
Alternating Current Coil Demagnetization .. . STANDARDS ..............................................
309 347
Direct Current Coil Demagnetization ........ . Hall Effect Meters ...................................
310 347
Flux Sensed Demagnetization ................... . Eddy Current Devices ..............................
Problems Associated with Paliial 310 347
Conclusion .............................................. 347
Demagnetization ................................. .
311
SECTION 13: SAFETY IN MAGNETIC SECTION 15: EQUIPMENT FOR
PARTICLE TESTING ............................... . MAGNETIC PARTICLE TESTS .................
313 349
PART 1: INTRODUCTION TO MAGNETIC
PARTICLE TEST SAFETy ......................... .. PART 1: BASIC EQUIPMENT
314
Safety in Practice ................................... .. CONSIDERATIONS ....................................
PART 2: MATERIAL SAFETY DATA SHEETS .. 314 3.50
316 Effect of Testing Parameters on EqUipment
Potential Matelial Hazards ....................... . Choice ................................................
316 350
Use of 1-.1aterial Safety Data Sheets ........... .
PART ,3: SAFETY \VITH ULTRAVIOLET 316 PART 2: \VET HORIZONTAL EQUIPMENT .....
352
RADIATION SOURCES .............................. . Positioning the Test Object. .......................
352
320 Magnetization Procedures .........................
Ultrmiolet Energy ................................... . 352
320 Alternating and Direct Current Systems ......
Projection Ultraviolet Sources ................... . 353
Ult~'a\iolet Radiation Hazards ................... . 322 320 Wet HoIizontal System Components ...........
3.53
Conclusion ............................................. . Multidirectional Test Systems .................... 3.54
PART 4: ELECTRICAL EQUIP!\IENT 324 PART 3: STATIONARY MAGNETIC
SJ-\FETy ..................................................... . PARTICLE EQUIPMENT .. '" .......................
325 355
ElectIical Safety Hazards ......................... .. Direct Current Magnetizing Equipment. .....
325 355
Determination of Hazardous Areas ............ . Automated Magnetic Particle Systems .........
326 35.5
Designing for Hazard ............................... . Pulsed or Capacitor Discharge Units ..........
327 3.55
Quick Break Magnetization .......................
3.55
xv
PART 4: .MOBILE MAGNETIC PARTICLE PAHT .3: MECHANICAL PAINT REMOVAL
EQUIPMENT. ............................................ . 3.57 AND ITS EFFECT ON CRACK
Current and Voltage Parameters .............. '" 3.57 DETECTABILITY ...................................... . 388
Operation of Mobile Testing Units ............ . :358 Test Object Preparation .......................... .. 388
Mobile Capacitor Discharge Systems ......... . 358 Visual Test Results ................................. .. 388
Maintenance of Mobile Testing Systems ..... . 359 Magnetic Particle Test Hesults .................. . 389
PART.5: PORTABLE MAGNETIC PARTICLE Conclusions .............................. " ............ . 389
EQUIPJ\;IENT ............................................. . 360 PART 4: MAGNETIC PAHTICLE TESTING
Handheld Equipment .............................. . 360 USING REMOTE VISUAL EQUIPMENT ... .. 390
Portable System Configurations ................. . 360 Types of Instmments .............................. .. 390
Accessodes and Components .................... . 360 Compadson of Remote Vie\ving
PART 6: MAGNETIC PARTICLE TESTING Instnllnents ........................................ . 392
POvVER PACK SYSTE~'lS ........................... . 362 Measurement of Test Indications .............. . 392
Power Pack Applications .......................... . 362 Remote Viewing Applications .................... . 393
Multidirectional Field Power Packs ........... . 363 PART .5: TYPICAL FALSE INDICATIONS IN
PART 7: DEMAGNETIZATION EQUIPMENT .. 364 MAGNETIC PARTICLE TESTS .................. . 394
Theory of Demagnetization ............. " ....... . 364 False Indications in a Bellcrank Assembly .. . 394
Need for Demagnetization ....................... . 36.5 False Indications in Threaded Fasteners .... . 396
Coil Demagnetization .............................. . 36.5 PART 6: MAGNETIC PAHTICLE TESTS OF
Direct Current Demagnetization ............... . 366 j\lAINTENANCE INDUCED CHACKING .. .. 398
PART 8: LIGHT SOURCES AND Grinding Cracks ...................................... . 398
ACCESSORIES FOR ]\;lAGNETIC Repair of Hydroplane Blades .................... . 398
PARTICLE TESTING ................................. . 367 PAHT 7: CONTROL OF "VET MAGNETIC
Sources of Visible Light for Magnetic PARTICLES FOR YOKE
Particle Testing ................................... . 367 MAGNETIZATION ..................................... . 400
Use of Ultraviolet Lamps for Fluorescent Suspension Quality .................................. . 400
Magnetic Particle Tests ................. '" .... . 368 Use of Prism Block Standards ............. ·...... . 400
Response of the Human Eye to Light Conclusion ................ " ................'.... : ...... . 401
Sources ................. ·................. ·............ . 369 PART 8: MAGNETIC PARTICLE FIELD
Eyeglasses and Fluorescent Particle Tests ... . 371 TESTING OF .STR UCTURAL WELDS, ........ .. '402
Mercury Arc Ultraviqlet ~amps ..... : ........... . 372 Effect of Weld Surface ........................... .. 402
Fixtures for Ultraviolet Sources ................. . 373 Field Strength Adjustments for Tests of
Output Valieties for Ultraviolet Sources ..... . 375 Welds .......................................... :·..... . 402
Fluorescent Tubular Cold Discharge Effect of Coating on Tests of "Velds .......... . 402
Ultraviolet Sources.: ............................ . 37.5 Effect of Other Tests Parameters .............. . 403
Care of Ultraviolet Sources ....................... . 376 PART 9: OIL FIELD APPLICATIONS OF
Ultraviolet Measurement Instmmentation .. . 377 MAGNETIC PARTICLE TESTING ............ .. 405
Using Ultraviolet Radiometers .................. . 379 Longitudinal Magnetization ...................... . 40.5
Service Life of Ultraviolet Meters ............. . 379 Circumferential Magnetization .................. . 408
Pulsed Current Magnetization ................... . 410
SECTION 16: SPECIAL APPLICATIONS OF Evaluating Current Pulse Effectiveness ...... . 41.5
MAGNETIC PARTICLE TESTING .... ....... . 381 Use of Field Indicators and Simulated
PART 1: UNDERWATER MAGNETIC Discontinuities .................................... .
PARTICLE TESTS ..................................... . 382 Documents for Magnetic Particle Testing of
Magnetic Pmticle Testing through Coatings 382 rfubes ................................................ . 419
Single Leg Electromagnet Teclmique ........ . 384
Discontinuity Data Bases .......................... . 385 SECTION 17: CODES, STANDARDS AND
PART 2: DRY POWDEH MAGNETIC SPECIFICATIONS .................................... . 423
PARTICLE TESTS OF PAINTED WELDS .. , 386 PART 1: INTHODUCTION TO MAGNETIC
Yoke Break Test ...................................... . 386 PARTICLE SPECIFICATIONS .................... . 424
Weld Bead Crack Helerence Standards ...... . 386 Definition of Terms ............................... , .. 424
Reference Standards for 'Veld Cracking ..... . 386 The Need for Specifications ..................... .. 424
Effect of Coating Thickness on Test Need for Heview and Revision of
llesults .............................................. . 387 Specifications ..................................... . 424

xvi
 PART 2: SAMPLE MAGNETIC PARTICLE
:380 SPECIFICATION '" '" .............. '" ........... ...... 427
:388 Arrangement of the Specification .............. .
380 427
Sample Specification Test ......................... .
389 427
389 INDEX .......................................................... .
440

92
-)2
)3

14
4
6
 SECTION 1
FUNDAMENTALS OF MAGNETIC
PARTICLE TESTING
Bernie Boisvert, NOT consultant, Dayton, Ohio

PORTIONS OF THIS TEXT ADAPTED FROM NONDESTRUCTIVE INSPECTION METHODS I US AIR FORCE TO-33B-!-11
2 I MAGNETIC PARTICLE TESTING

PART 1
INTRODUCTION TO MAGNETIC
PARTICLE TESTING

may also be helpful for introductory studies by individuals

Capabilities and Limitations of already using magnetic particle testing or those preparing
for advanced training in the technique.
Magnetic Particle Techniques
Magnetic particle testing can reveal surface discontinui-
ties, including those too small or too tight to be seen \vith
the unaided eye. Magnetic particle indications form on an
object's surface above a discontinuity and show the location
Principles of Magnetic
and approximate size of the discontinuity. Magnetic particle Particle Testing
tests can also reveal discontinuities that are slightly below
the surf~l.Ce. Magnetic particle testing is a nondestructive method of
There are limits on this ability to locate subsurface revealing surface and slightly subsurface discontinuities in
discontinuities. These are determin~d by the discontinuity's magnetizable matetials. It may be applied to raw materials
depth, its size, type and shape, and the strength of the such as billets, bars and shapes; dUling processes such as
applied field. In some cases, special techniques or equip- forming, machining, heat treating and electroplating; and in
ment can improve the test's ability to detect subsurface testing for service related discontinuities. tvIagnetic paliicle
discontinuities. procedures cannot be used \vith non magnetizable matedals
Magnetic particle testing cannot be used on nonmagnetic such as aluminum or copper.
mateIials, including glass, ceramics, plasticS or such com- The testing method is based on the principle that mag-
mon metals as aluminum, magnesium, copper and austenitic netic flux i~ a magnetized object is locally distorted by the
stainless steel alloys. In addition, there are certain positional presence of a discontinuity. This distortion cau.sys some of
limitations: a magnetic field is directional and for best the magnetic field to exif and reentei· the test object at the
results must be Olie~lted perpendicular to the discontinuity. discontinuity. This phenomenon is calledm.agnetic flux
This generally requires two complete magnetizing opera- leakage. Flux leakage is capable of attracting finely divided
tions to detect discontinuities parallel and perpendicular to pmiic1es of magnetic mateIials which in turn form an
the test object's axis. Objects with large cross sections outline or indication of the discontinuity.
require a very high current to generate a magnetic field One of the objectives of magnetic pmiicle testing is to
adequate for magnetic particle tests. A final limitation is that detect discontinuities as early as possible in the processing
a demagnetization procedure is usually required follOwing sequence, thus avoiding the expenditure of effOli on mate-
the magnetic particle process. rials that will later be rejected. Practically evelY process,
This text provides an overview of the magnetic particle from the odginal production of metal from its ore to the last
testing process and is introductOlY to the detailed treat- finishing operation, may introduce discontinuities. Magnetic
ments in subsequent sections. Topics covered here include: pmiicle testin~ can reveal many of these, preventing flawed
(1) basic steel and component production and some of the components from enteting service. Even though magnetic
discontinuities produced; (2) the fundamental theOlY of pmticle testing may be applied durillg and hetween process-
magnetism, magnetic flux and types of magnetic fields; ing operations, a final test is usually performed to ensure
(3) plinciples of electrically induced magnetism and mag- that all detrimental discontinuities have been detected.
netizing current; (4) testing media and processes; and The test itself consists of three basic operations: (1) estab-
(5) basic principles and methods of demagnetization. lish a suitable magnetic flux in the test object; (2) apply
Such data can be helpful to managers, supervisors and magnetic particles in a dly powder or a liquid suspension;
personnel outside nondestructive testing who require gen- and (3) examine the test object under suitable lighting
eral information on the magnetic particle testing process. It conditions, inte111reting and evaluating the test indications.
 FUNDAMENTALS OF MAGNETIC PARTICLE TESTING / 3

PART 2
FABRICATION PROCESSES AND
MAGNETIC PARTICLE TEST APPLICATIONS

tis '1'1 )('r(' arc several ways to classify magnetic particle

it'slillg applications. Most approaches involve the discontin- Basic Ferromagnetic Materials
or
llili('S illlerest, hut these classifications prove to be too Production
sp(,cific or too geneml ic)r practical application.
(hw hroad classification method, which is also used in
In the production of ferrous alloys, iron ore is converted
liqllid }1t'lletrallt testing, is based on discontinuity location: to steel in one or more furnaces where it is melted, refined
SlIrhcl' or suhsurLlCe. This classification procedure is useful and allOying elements are added. While in the liquid state,
Ii II' magnetic pmticle tests since the ability to find discon- the metal is poured into a mold and allowed to solidify into
t i 1111 ities of each t~'l)e varies slimply. Beyondthis use though, a shape typically called an ingot.
I hl' local ion dassification system is too broad for common
Ingots are quite large and must be formed into more
IIS(, 1I:' magnetic palticle te'st inspectors.
manageable shapes by hot working through a seIies of rolls
AIl( )llwr method of classif)ring test applications is by using
II or mills. These semifinished shapes are called' blooms, billets
I hl' process in which they .. are applied: forging, casting,
or slabs, depending on their size and shape .. A bloom is an
m,ldillg, heat treating, glinding and so on. These processes intermediate product, rectangular in shape with a width not
a re also IIsed to define discontinuity types: forging cracks,
more than twice its thickness and a crass-sectional area
lH'al tn'at cracks m grinding·crucks, for example. \Vhile this ~'Pically larger than 0.02 m 2 (36·in.2). A billet can be round
sysll'm is \lsed extenSively for desclibing discontinuities, it is or square with a cross-sectional area from 1,600 mm 2 to
too specific for all magnetic particle applications. ' 0.02 m 2 (2.5 to 36 in. 2). A slab is an intermediate shape
The most Widely used classification system considers the between an ingot and a plate with a width at least mrice its
origin of discontinuities in the stages of fablication and thickness.
S('Jyice. These classes may be broadly categOlized as follows.

Inherent Discontinuities
1. Prilllary production and processing tests: this testing
This group of discontinuities occurs duling the initial
category is llsed to inspect the stages of processing
melting and refining processes and dUling solidification
from poming alld solidification of the ingot to produc-
from the molten state. Such discontinuities are present
tion of hasic shapes, including sheet, bar, pipe, tubing,
before rolling or forging is performed to produce interme-
r()r~illgs and castings. These tests are ~pically used to
diate shapes.
locate two discontinuity suhgroups: (a) those formed
during solidification are called inherent discontinuities;
and (Ill those formed dming mill reduction are called Pipe
jirill/(/ry processillg discontinuities. \Vhen molten metal is poured into an ingot mold,
,.., Secolldary processing or lIwl111facturing and fabrica- solidification progresses gradually, stmting at the bottom
tion t('sts: this testing category is used to inspect the and sides of the mold and progressing upward and inward,
res\) Its of processes that convelt rmv stock into finished The solidified metal is slightly smaller in volume than the
compon('nts. Forming, machining, \velding and heat liqUid and there is progressive sl1linkage dUling solidifica-
treatillg discontinuities are detected v,rith this kind of tion. The last metal to solidify is at the top and center of the
llIaglH'tk IXllticle test. mold. Because of the shlinkage, there is typically insuffi-
cient liquid metal remaining to fill the mold and a depres-
:3. Serde(' tests: these test procedures are widely used for sion or cavity is formed.
detecting over-stress and fatigue cracking. Ivlagnetic In additi~n, impurities such as oxides and entrapped
palticle tests are not used to detect corrosion, defor- gases tend to migrate to the center and top of a mold and
matiull or wear, three of the most common service may become embedded in the last portions to solidify. After
illduced problems. solidifIcation, the upper pOltion is cut off or cropped and
4 I MAGNETIC PARTICLE TESTING

steel cleanliness define the amount of inclusions or stringers

FIGURE 1. Cross section of an ingot showing that may be accepted.
shrinkage cavity at top center The addition of lead or sulfur to molten steel is a common
practice for the alloys known as free machining steels. These
alloys contain a large number of nonmetallic inclusions that
break or chip during machining operations. ~'1agnetic par-
ticle tests of free machining alloys often indicate an alarming
number of discontinuities that are not considered dehimen-
tal in service.

Blowholes
As molten steel is poured into an ingot and solidification
commences, there is an evolution of gases. These gases rise
through the liquid in the form of bubbles and many escape
or migrate to the cropped pOltion of the ingot.
However, some gases can be trapped in the ingot,
forming the discontinuities known as blOlcholes. Most
blowholes are clean and will weld or fuse shut during
primary and secondary rolling. Those near the surface may
have an oxidized skin and will not fuse, appearing as seams
in the rolled, forged or extruded product. Oxidized blow-
holes in the interior of slabs appear as laminations in plate
products.

Ingot Cracks
Contraction of the metal during solidification and cooling
of the ingot gene'rates significant surface stresses and
internal stresses which can result in cracking. If the cracks
are internal and no air reaches them, they are usually
discat:ded, removing most of the shrinkage cavity and welded shut dming rolling and do not result in discontinu- .
impurities. However, if the cavity is deeper than normal or ities. If they are open to the air or otherwise become
if the cropping is. short, some of tpe unsound metal will oxidized, they will 'not seal but remain in the finished·
show up in the intermediate shape as a void called pipe. product.'
Figure 1 shows an ingot cross section illustrating the During the rolling of an ingot into a billet, oxidized cracks
shrinkage cavity and impurities in the top center. Pipe is form long seams. It is common practice to use magnetic
almost always centered in the semifinished shape and is particle tests of billets before additional processing. Such
undesirable for most purposes. preprocessing tests permit the removal of seams by gIind-
ing, chipping or flame scarfing. If not removed before
Nonmetallic Inclusions rolling or working, seams are further elongated in finished
All steel contains nonmetallic matter that mainly origi- shapes and this may make the final product unsuitable for
nates in deoxidizing materials added to the molten metal many applications.
during the refining operation. These additives are easily
oxidized metals s~ch as aluminum, silicon, manganese and
others. The oxides and sulphides of the metals make up the Primary Processing Discontinuities
majority of nonmetallic inclusions. vVhen finely divided and
well distributed, these discontinuities are not ohjectionable. \\Then steel ingots are worked down to shapes such as
However, sometimes the additives collect during solidifi- billets, slabs and forging blanks, some inherent discontinu-
cation and form large clumps in an ingot. DUling primary ities may remain in the finished product. In addition, rolling
processing these large clumps are rolled out into long or forming operations may themselves introduce other
discontinuities called stringers. In highly stressed compo- discontinuities. The plimary processes considered here
nents, stringers can act as nucleation points for fatigue include the hot working and cold working methods of
cracking. In celiain test objects, stringers are acceptable in producing shapes such as plate, bars, rod, wire, tubing and
limited amount. Government and indushy specifications on pipe.
 FUNDAMENTALS OF MAGNETIC PARTICLE TESTING / 5

Forging and casting are also included in this categOlY

Iwc;lnst' they typically require additional machining or other FIGURE 2. Formation of seams and laps:
subsequent processing. All the pIimary processes have the (a) overfill produces excess metal squeezed out
potential fCJI" introducing discontinuities into clean metal. of rolls; (bl a lap results when the projection is
folded over and forced back Into the bar's
surface during a SUbsequent pass; (c) underfill
Seams results when there is not enough metal to fill
the rolls; and (d) a seam in the finished bar
Scams in hal's, rod, pipe, wire and tubing are usually
occurs when underfill is squeezed tight on a
ohjectionable and make the product unsuitable for many
subsequent rolling pass
apj)lic<1hons. Seams can oIiginate from ingot cracks and
despite preprocessing tests, some cracks can be overlooked fa)
or illcompletely removed.
Holling and drawing operations can also produce seams in
the finished product. If the reduction on any of the rolling
passes is too great, an overfill may then produce a projection
from the billet (see Fig. 2). This projection can be folded or
lapped on subsequent passes, producing a long deep sea~~
The reverse also occurs if the shape does not fill the rolls,
resulting in a depression or surface groove. On subsequent
rolling passes, this underfill produces a seam running the
fllll length of the shape. Seams oIiginating from overfilled
fbJ
rolls usually emerge at an acute angle to the surface. Seams
e<lllscd by imderfilled rolls are likely to be normal to or
p('lvcndicular to the surface.
Se.u11s or die marks can be introduced by defective or
dilty dies duIing drawing operations. Such seams are often
birl~- shallo:v and may not be objectionable, especially when
sui>s('(Iuent machining removes the seam. Seams are always
objectionable in components that expeIience repeated or feJ
cyclic- stresses in selvice. These seams can nucleate fatigue
cracks.

laminations
Laminations in plate, sheet and stIip are formed when
hlowholes or internal cracks are not welded shut duIing
rolling, hut are flattened and enlarged. Laminations are fdJ
large and potentially troublesome areas of hoIizontal dis-
continuity.
l\lagn~tic pmiicle testing detects lamination only when it
H'aches and hreaks the edges of a plate. Laminations that
aI"(' cOlllpletely internal to the test object typically lie parallel
to its slIrfaee and cannot be detected by magnetic particle
proccdures.

Cupping
Clipping occurs dming drawing or extruding operations Cooling Cracks
\\"h(~n the intelior of the shape does not flO\v as rapidly as the Bar stock is hot rolled and then placed on a bed or cooling
snr~a<:(4. The result is a seIies of internal ruptures that are table and allowed to reach room temperature. During
st'n~)llS :\'henever they occur (see Fig. 3).
cooling, thermal stresses may be set up by uneven rates of
(-uppmg can be detected by the magnetic patiicle meth- temperature change within the mateIial. These stresses can
od oIlly when it is severe and approaches the surface. be sufficient for generating cracks (see Fig. 4).
6 / MAGNETIC PARTICLE TESTING

FIGURE 3. Cupping formed during drawing or FIGURE 4. Cooling cracks indicated with
extruding fluorescent magnetic particles

Cooling cracks are generally longitudinal but because FIGURE 5. Forging laps in piston rods
they tend to curve around the object shape, they are not
necessarily straight. Such cracks maybe long and often vary
in depth "along their length. Magnetic particle indications of
cooling cracks therefore can vary in intensity (heavier where
the crack is deepest) ..

Forging Discontinuities
Forgings are produced from an ingot, a billet or forging
blank that is heated to the plastic flow temperature and then
pressed or hammered between dies into the desired shape.
This hot working process can produce a number of discon-
tinuities, some of which are described below.

Flakes positioning, flakes are not detectable by magnetic paliicle

techniques unless machining brings the discontinuity close
Flakes are internal mptures that some believe are caused to the surhlCe.
by cooling too rapidly. Another theOlY is that flakes are
caused by the release of hydrogen gas dming cooling.
Forging Bursts
Flakes usually occur in fairly heavy sections and some
alloys are more susceptible than others. These mptures are When steel is worked at improper temperatures, it can
usually well below the surface, typically more than half way crack or mpture. Reducing a cross section too rapidly can
between the surface and the center. Because of their also cause forging bursts or severe cracking.
 FUNDAMENTALS OF MAGNETIC PARTICLE TESTING / 7

Forging bursts may be internal or surface anomalies .. FIGURE 6, Flash lines and laps in forgings
"'hen at or near the surf~lce, they can be detected by the
magnetic palticle method, Internal bursts are not generally
detected with magnetic particles unless machining brings
them near the surface.

Forging Laps
During the forging operation, there are several factors
tllat can cause the surface of the object to fold or lap.
Because this is a surface phenomenon exposed to air, laps
are oxidized and do not weld when squeezed into the object
(see Fig. 5).
Forging laps are difficult to detect by any nondestructive
testing method. They lie at only slight angles to the surface
alld may be fairly shallow. Forging laps are almost always
ohjectionable since they selve as fatigue crack nucleation
poillts.

Flash Line Tears

Hot Tears and Shrinkage Cracks
As the dies close in the final stage of the forging process,
a small amount of metal is extruded between the dies. This Hot tears are surface cracks that occur during cooling
extruded metal is called flash and must, be .removed by after the metal has solidified. They are caused by thermal
tlimming. . stresses generated during uneven cooling. ·Hbt tears usually
If the trimming is not done or not done properly, cracks originate at abrupt changes in cross section where thin
or tears can occur along the flash line (see Fig. 6). Flash line sections cool more rapidly than adjacent heavier masses.
tears are reliably detected by magnetic paIticle testing. Shrinkage cracks are also surface cracks that occur after
the' metal cools. They are caused by the contraction or
reduction in volume that the casting experiences during
solidification.
Casting Discontinuities
Castings are produced by pouring molten metal into Weldment Discontinuities
molds. The combination of high temperatures, complex
shapes, liquid metal flow and problematic mold materials Welding can be considered a localized casting process
can cause a number of discontinuities peculiar to castings. that involves the melting of both base and filler metal.
Some of these are described below. Welds are subject to the same type of discontinuities as
castings but on a slightly different scale. In addition, other
discontinuities may be formed as a result of improper
Cold Shuts welding practices. Some of the discontinuities peculiar to
Cold shuts originate during pouring of the metal when a weldments are described below.
pOItion of the molten liquid solidifies prior to joining with
the remaining liquid. The presence of an oxidized surface,
Lack of Fusion and Lack of Penetration
eW'1l though it is liquid or near liquid, prevents fusion when
tv.'O surfaces meet. This condition can result from splashing, Failure to melt the base metal results in a void between
interrupted pouring or the meeting of two streams of metal the base and filler materials. This lack of fusion can be
coming from different directions. detected by magnetic particle methods if it is close enough
Cold shuts can be shallow skin effects or can extend quite to the weld surface.
<leeply into the casting. Shallow cold shuts called scabs \\lith lack of penetration, only a thin layer of the base
('an be removed by grinding. Deep cold shuts cannot be metal has been melted. Magnetic particle testing does not
repaired. generally detect lack of penetration.
8 / MAGNETIC PARTICLE TESTING

Heat Affected Zone Cracks and Crater Cracks by the body of the object. Comer cracking can also occur
during quenching because of the thermal stresses of uneven
Cracks in the base metal adjacent to the weld bead can be
cooling.
caused by the thermal stresses of both melting and cooling.
Such cracks are usually parallel to the weld bead. Heat
affected zone cracking is eaSily detected by magnetic parti- FIGURE 7. Grinding cracks indicated with:
cle testing. (a) visible magnetic particles; and (b) fluorescent
Cracks in the weld bead caused bv stresses from solidifi- magnetic particles
cation or uneven cooling are called crater cracks. Cracks
caused by solidification usually occur in the final weld fa)
puddle. Cracks caused by uneven cooling occur in the thin
portion at the junction of two beads. Magnetic particle
testing is widely used to detect crater cracks.

Manufacturing and Fabrication

Discontinuities
This classification includes discontinuities associated with
various finishing operations after the material has been
rough formed by rolling, forging, casting or welding. Pro-
cesses such as machining, heat treating or grinding may
introduce discontinuities, some of which are described
below.

Machining Tears
Machining tears occur if a tool bit drags metal from the
surface rather than cutting it. The primary cause of this is
improperly shaped or dull cutting edges on the bit.
Soft or ductile metals such as low carbon steel are more
susceptible to machining tears than harder medium carbon fb)
and high carbon steels. Machining tears are surface discon-
tinuities and are reliably detected by the magnetic particle
method.

Heat Treating CraCKS

When steels are heated and quenched (or otherwise heat
treated) to produce properties for strength or wear, cracking
may occur if the operation is not suited to the material or
the shape of the object. The most common sort of such
cracking is quench cracking, which occurs when the metal is
heated above the clitical transition point and is then rapidly
cooled by iimnersing it in a cold medium such as water, oil
or air.
Cracks are likely to occur at locations where the object
changes shape from a thin to a thick cross section, at fillets
or notches. The edges of keyways and roots of splines or
threads are also susceptible to quench cracking.
Cracks can also originate if the metal is heated too
rapidly, causing uneven expansion at changes of cross
section. In addition, rapidly increasing heat can cause
cracking at comers, where heat is absorbed from three
surfaces and is therefore absorbed much more rapidly than
1I1I1111111I111111111111111111111111111111111111111111111111'111'11111111111111111111111111111111

FUNDAMENTALS OF MAGNETIC PARTICLE TESTING I 9

I('Clir Straightening and Grinding Cracks Acid pickling can weaken surface fibers of the metal,
<-,veil allowing internal stresses from the quenching operation to
The lIl1('\'(~Jl stresses caused by heat treating frequently be relieved by crack formation (see Fig. 8). Another ~rack­
'es1Ilt in distortion or warping and the metal forms must be
:traightellccl into their intended shape. If th~ distortion is
ing mechanism is the interstitial absOlption .o.f hydrogen
released by the acid etching or electrodeposihon process
too great or thc objects are velY hard, crackmg can occur (Fig. 9). AbsOlption of nascent hydrogen adds to the inter-
It duri;w the straightening operation. nal stresses of the object and subsequently may cause
Slll~'ac(' cracks can also occur in hardened objects during cracking. This mechanism, called hydrogcn clnbrittlement,
improper grinding operatiOl~s. Such the.rmal cracks are can result in cracking dming the etching or plating opera-
created bv stresses from localIzed overheatmg of the surface tion or at some later time when additional service stresses
und('r th;' grinding wheel. Overheating can be caused by are applied.
using the wrong grinding wheel,. a dull or .glazed wh~el,
insuff'icicllt or poor coolant, feedmg too rapIdly or cuttmg
too he,wilv. Crinding cracks are especially detrimental since
the,' arc l~el11endicular to the object surface and have sharp
"Service Discontinuities
edges that propagate under repeated or cyclic loading (see
The remaining category of discontinuities are those
Fig. 7). .
Another t!ve of discontinuity that may occur dunng formed after fablication processes are complete and the
grinding is cracking caused by residual.stresses. Hardened object has been placed in selvice. The objective of m.ag?etic
particle testing during processing is to detect and ehmmate
ohjects may retain stresses that are not hIgl~ enough to cause
harmful discontinuities and to place into service objects that
cracking. Dming grinding, localized heatmg added to ~n­
are free of discontinuities. Even when this is fully accom-
trapped stresses can cause surface ruptures. The result~ng
cracks are lIsnallv more severe and extensive than typIcal plished,diseontinuities can occur frOl~ service conditions.
gJincling cracks. ' Some discontinuities such as deformatIon and wear are not
detected by the magnetic pmticle test, but the technique is
..useful for indicating tbe discontinui~ies listed below.
Plating, Pic'kling and Etching Cracks
Hardened surfaces are susceptible to cracking from Overstress ·Cracking
electroplating, acid pickling or etching processes. There are All matelials have load iimits (called ultinUlte strength)
several mechanisms that cause this cracking. and when service stressing exceeds this limit, cracking
occurs. Usually the failure is completed by surface fracture
of the object. In this case, the crack is easy to detect and
magnetic particle testing is not required.
FIGURE 8. Treating with acid weakens the
surface metal, allowing the release of a spring's
internal stresses through surface cracks
FIGURE 9. Hydrogen or pickling cracks on steel
spring

~~y . ,
./ . ~ ~..... '\ t

. ~, ~::y:",~' )
, \

.i~#~;.!;:( "
10 I MAGNETIC PARTICLE TESTING

There are instances where surfaces do not visibly separate

FIGURE 10. Fatigue cracking in manufactured and magnetic palticle testing is needed to detect and locate
components: fa) gear tooth roots; fb) automobile the cracking.
crankshaft; and fc) aircraft component

la' Fatigue Cracking

Objects subjected to repeated alternating or fluctuating
stresses above a specific level eventually develop a crack (see
Fig. 10). The crack continues to grow until the object
fi-actures. The stress level at ,",vhich fatigue cracks develop is
called the fatigue strength of the material and is \vell belO\,v
the ultimate strength of the matetial. There is an inverse
relationship between the number of stress applications
(cycles) and the stress level necessary to initiate cracking:
low cycles and high stress produce the same results as high
cycles and low stress .
. Another factor contributing to fatigue cracking is the
presence of surface anomalies such as copper penetration
(see Fig. 11) sharp radii, nicks and tool marks. These act as
stress risers and lower both the number of cycles and the
stress level needed to initiate cracking. Fatigue cracking
tylJically occurs at the sUlface and is reliahly detected by
magnetic particle testing.

Corrosion
Magnetic particle procedures are not used to detect
surface corrosion or pitting. However, there are secondary
discontinuities that can be revealed by the magnetic pmtide
method .. vVhen objocts are Ullder sustained stress, either

FIGURE 11. Fatigue crack from copper

penetration on a journal
 FUNDAMENTALS OF MAGNETIC PARTICLE TESTING / 11

illternal or external, and are at the same time exposed to a itself does not usually produce magnetic particle indications
('nrr()~i\'(> atmosphere, a pmticular kind of cracking results. (in some applications, sharp edged pits can hold patiicles).
KIlO\\1l as stress corrosion cracking, this discontinuity is Pitting can serve as a stress liser and often nucleates fatigue
l~asih detected by magnetic pmiicle testing. cracks. Fatigue cracks originating at corrosion pits are
:\;)( )tller occurrence related to corrosion is pitting. Pitting reliably detected by the magnetic particle method.

,>c(

) is
:)\V
"S(-,

'IlS

Iw
)Il

as
1('
12 / MAGNETIC PARTICLE TESTING

PART 3
MAGNETIC FIELD THEORY
the Earth's magnetic poles and therefore are called north
Magnetic Domains and south poles.
Figure 13 can be duplicated by placing a sheet of paper
Materials that can be magnetized possess atoms that over a bar magnet and sprinkling iron particles on the paper.
group into submicroscopic regions called magnetic dOlrUlins. It shows the magnetic field leaVing and entering the ends or
These domains have a positive and negative polarity at poles of the magnet. This characteristic pattem illustrates
opposite ends. If the material is not magnetized, the the term lines offorce used to describe a magnetic flux field.
domains are randomly oriented, usually parallel with the There are a number of important properties associated with
crystalline axes of the material. lines of force.
When the material is subjected to a magnetic field, the
domains orient or align themselves parallel with the extemal 1. They form continuous loops which are never broken
magnetic field. The material then acts as a magnet. Figure but must complete themselves through some path.
12 illustrates the domain alignment in nonmagnetized and 2. They do not cross one another.
magnetized material. :3. They are considered to have direction, leaving from
the north pole, traveling through air to the south pole
where they reenter the magnet and retum through the
Magnetic Poles magnet to the north pole.
4. Their denSity decreases with increasing distance from
A magnet has the property of attracting ferromagnetic the poles.
materials. The ability to attract or (repel) is not uniform over . 5. They seek the path of least magnetic resistance or
the surface of a magnet but. is concentrated at localized reluctance in completing their loop.
areas called poles. In every magnet, there are two or more
poles with opposite polarities. These poles are attracted to \-Vhen a bar magnet is broken into two or more pieces,
new magnetic poles are formed. The opPQsing poles att.t;act
'FIGURE 12. Orientation of magnetic domains: orie another as shown in Fig. 14. .
fa) in a nonmagnetized material; and fbJ in a
magnetized material
FIGURE 13. Magnetic field surrounding a bar
magnet

FIGURE 14. Broken bar magnet illustrating

locations of newly formed magnetiC poles
 FUNDAMENTALS OF MAGNETIC PARTICLE TESTING I 13

II t II(' center piece in Fig. 14 is reversed so that similar

\1\ )l.·~ ;1 n' adjacent, the lines of force repel one magnet from Sources of Magnetism
t lit- ()llwr. If one of the bars is small enough, the lines of
1;11""" (':lll cause it to rotate so that unlike poles are again
;hli:t(t'llt. This illustrates the most basic rule of magnetism:
Permanent Magnets
III i li L(, poles attract and like poles repel. Permanent magnets are produced by heat treating spe-
cially formulated alloys in a strong magnetic field. During
the heat treating process, the magnetic domains become
aligned and remain aligned after removal of the external
Types of Magnetic Materials field. Permanent magnets are essential to modern technol-
ogy, including applications such as magnetos, direct cur-
,\lI materials are affected to some degree by magnetic
rent motors, telephones, loud speakers and many electlic
li!·ld, \ latter is made up of atoms with a positively charged
instruments.
IIl1!'l. 'i 1\ surrounded by a field or cloud of negatively charged
Common examples of permanent magnetic materials
(,J. 'It l!>IlS. The electron field is in continual motion, spinning
include alloys of aluminum, nickel and cobalt (alnico);
anHllld the: nucleus. \Vhen the matelial is subjected to a
copper, nickel and cobalt (cunico); copper, nickel and iron
1I1.1~lldi(' field, the electron orbits are distorted to some
(cunife); and cobalt and molybdenum (como}).
(It'~n ',', The amount of this dist01tion (or the corresponding
chall~.' ill magnetic charactetistics) when subjected to an
('\It'rn;t! magnetic field p'rovides a means of classifying
Magnetic Field of the Earth
1~1:\1( 'rials into three groups: diamagnetic, paramagnetic or
("IT11lllapwtk, The planet Emth is itself a huge magnet, with north and
south poles slightly displaced from the Emth's axis. This
displacement results in a slight deviation between geograph-
Diamagnetic Materials ic n01th and magnetic n01t11.
"I'Ji(' krm diamagnetic refers to a substance whose mag- As a magnet, the Ealth ~s ·surrounded by magnetic lines of
lldi(' permeability is slightly less than that of a vacuum. force as shown in Fig. 15. These lines of force make up what
\\'!Jell phu:ed in' a strong magnetic field ,their induced mag- is sometimes called the earth field and they can cause
111'1 iSli I is in ,a direction opposite to that of iron. Diamagnetic problems in both magnetizing and demagnetizing of ferro-
ll1al('rials include mercury, gold, bismuth and zinc. magnetic test objects. The emth field is weak, on the order
of 0.03 mT (0.3 G).
Movement of ferromagnetic objects through the earth
Paramagnetic Materials field can induce slight magnetization. This is a problem in
aircraft where magnetized components can affect the com-
,Paramagnetic denotes a substance whose permeability is
shghtlv greater than that of air or unity. \Vhen such ma- passes used in navigation. Similarly, demagnetizing can be
difficult if certain objects, usually long shafts, are not
tv,rials' are pl~lced in a strong magnetic fi~ld, there is a slight
oriented in an east-west direction dming the demagnetiza-
aht.':.lllll<'lIt of the electron spin in the direction of the mag-
tion process. .
lld 1(' nllx flow. This alignment exists only as long as the
par~lIllaglletic material is in the external magnetic field.
~\hllllinllm, platinum, copper and wood are paramagnetic
llt;lkrials. FIGURE 15. Magnetic field of the Earth

/--~
Ferromagnetic Materials / '\"
I \1';"'-'-\-"
F(Trumagnetic substances have a permeability that is
1111 wit greater than that of air. "Vhen placed in an extemal
,
I \
\
\
\
Illa~lldil' ,fIeld, the magnetic domains align parallel with the I
l'~t('l'llaJ field and remain aligned for some period of time ,I
altt'1' r('111oval fi'om the field. I
" This c()ntinued alignment after removal from the extemal I
I
h(·ld is called retentivity and can be an important propelty
llI:o llle magnetic pcuticle testing procedures.
SOlll,e examples of ferromagnetiC materials are iron, co-
hall. lllckel and gadolinium.
14 / MAGNETIC PARTICLE TESTING

Mechanically Induced Magnetism It can be difficult to remove mechanically induced mag-

netization resulting from cold working. Disassembly is
Cold working of some ferromagnetic materials, either by usually impractical and demagnetization must be accom-
forming operations or during service, can magnetize the plished using portable yokes or cable coils. The operation is
objects. When mechanically induced magnetization occurs complicated when other ferromagnetic components are
as a result of forming operations, it can be removed by near the magnetized object: the demagnetizing operation
subjecting the magnetized object to a routine demagnetiza- can magnetize adjacent objects and a sequence of demag-
tion process.
netizing operations must then be performed.
 FUNDAMENTALS OF MAGNETIC PARTICLE TESTING / 15

PART 4
MAGNETIC FLUX AND FLUX LEAKAGE
crack, it can still create magnetic poles (see Fig. l7c) that
Circular Magnetic Fields hold magnetic matetials (the magnetic flux leakage field is
still finite),
TIl(' most familiar type of magnet is the horseshoe shape
sl](I\\1i in Fig. 16a. It contains both a n01th and south pole
willI the lines of magnetic flux leaving the north pole and Magnetic Field Strength
tL\\l'ling through air to reenter the magnet at the south
pol(' Ferromagnetic materials are only attracted and held at The amount of dist01tion or the strength of a flux leakage
or I wt\\'('en the poles of a horseshoe magnet. field from a discontinuity depends on several factors: (1) the
If t he ends of such a magnet are bent so that they are
closer together (see Fig. 16b), the poles still exist and the
Illaglldic flllx still leaves and reenters at the poles. However, FIGURE 16. Horseshoe magnet illustrating the
tIl!' lillt's of force are closer together and more dense. The fundamental properties of magnetism:.
llI11llll('r of lines of flux per unit area is called nwgl1ctic flux (a) direction of magnetic flux; (b) magnetiC flux
del/sitlj. ,in air around poles; moving poles close together
I r tfl(' magnetic flux density is high enough, f~rromagnetic raises the magnetic flux density; (c) fusing the
pari iell'S are strongly attracted and can even bridge the poles forms a circularly magnetized Object; and
ph:si( 'al gap between poles that are close enough together. (d) a discontinuity in a circularly magnetized
T\H' area where the flux lilies leave the pole, travel through Object and its resulting flux leakage field
air alld reenter the magnet is called a magnetic flux leakage
field fa'
When the ends of a magnet are bent together and the
pol(>s are fused to form a ring (see Fig. 16c), the magnet no
IOllg(,r attracts or holds ferromagnetic materials (there are
110 Illagnetic poles and no flux leakage field). The magnetic
/lIL\ lillt's still exist but they are completely contained within
tltt> 1I1agnet. In this condition, the magnet is said to contain
a cin'lIlar magnetic field or to be circularly nwgnetized.
If ;t crack crosses the magnetic flux lines in a circularly
llIaglldized object, north and south poles are immediately
cn'ated on either side of the discontinuity. This forces a
portio1l of the magnetic flux into the surrounding air,
cn'atillg a flux leakage field that attracts magnetic particles
(SIT Fig. 16d) and forms a crack indication.

Longitudinal Magnetization
. I r a horseshoe magnet is straightened, a bar magnet is
f()nlled with n01th and south poles (see Fig. 17a) . .rvlagnetic
flll\ fl()ws through the magnet and exits or enters at the
poll's. FerromagnetiC matelials are attracted only to the
poll's and snch an object is said to have a longitudinal field
I)r to IH:' longitudinally nwgnetiz.:ed.
. 1ft Il('se magnetic flux lines are intenupted by a discon-
tlllllih, additional n01th and south poles are formed on
eit\H'l side of the interruption (see Fig. 17b). Such second-
an P( )Ies and their associated flux leakage fields can attract
1lIa~ll{ 'tic patticles. Even if the discontinuity is a very narrow
16 / MAGNETIC PARTICLE TESTING

FIGURE 17. Bar magnet illustrating longitudinal FIGURE 19. Internal or midwall discontinuity in a
magnetization: (a) horseshoe magnet magnetized test object; there mayor may not be
straightened into a bar magnet with north and magnetic flux leakage, depending on the value
south poles; (b) bar magnet containing a of the flux in the object
machined slot and corresponding flux leakage
field; and (c) a crack in a longitudinally
magnetized object produces poles that attract
and hold magnetic particles
===- -==~~-~--====------==----==------===-
==::::---===-- - - - --- --
-----
---- --~.::-- -=---------=--=---
~ -==-==-- - - -=------.:.:---
(a)
IN "... '. sl

~
,'I' ""\\~
,H
I I
\ ,,
I I
,..., number of magnetic flux lines; (2) the depth of the discon-
,\ " tinuity; and (3) the width of the discontinuity's air gap at the
~\ /;J surface (the distance between the magnetic poles).
The strength of the leakage field directly determines the
(b) . number of magnetic particles that can be attracted to form
~GN~TI~ PARTICLES a test indication. The greater the leakage field strength, the
denser the indication, so long as the magnetic flux leakage
.. ':~ field is highly curved.
-----"-
N_-_ _ ~ ---- s
----
r-------":I""~S : :·: ....N-~'-----~
. .
Subsurface Discontinuities
A slot such as a keyway on the backside of an object
creates new magnetic poles that distOli the internal flux
(e) flow. If the slot is close enough to the surface, s<?me
magnetic flux lines'may be forced to exit and reenter the
magnetized object at the smface. The resulting leakage field
can form a magnetic particle test indication.
The size and strength of the indication depends on:
(1) the proximity of the slot to the top smface; (2) the size
and olientation of the slot; and (3) the intensity and distri-
bution of the magnetic flux field. A similar eff~ct occurs if
the discontinuity is completely internal to the object. Figure
18 is an illustration of a keyway on the far side of a bar and
Fig. 19 illustrates a midwall discontinuity.

FIGURE 18. Slot or keyway on the reverse side of

a magnetized bar
Effect of Discontinuity Orientation
The olientation of a discontinuity in a magnetized object
is a major hlctor in the strength of the leakage field that is
formed. This applies to both surface and internal discontin-
uities. The strongest leakage field is formed when the
discontinuity is perpendicular to the magnetic flux flow. If
the discontinuity is not pe1l)endicular, the strength of the
leakage field is reduced and disappears entirely when the
discontinuity is parallel to the magnetic flux flow.
Figure 20 illustrates the effect of discontinuity olientation
on the strength of the maglletic leakage field.
1""""'111111111111111111111111111111111111111111111'11111'1'1'111111111111111"111111'111111

FUNDAMENTALS OF MAGNETIC PARTICLE TESTING / '7

FIGURE 20. Flux leakage fields from Formation of Indications

discontinuities with different orientations:
faJ perpendicular to the magnetic flux; When magnetic paIiicles collect at a flux leakage site,
fbJ at 45 degrees angle to the magnetic flux; and they produce an indication that is visible to the unaided eye
fcJ parallel to the magnetic flux under the proper lighting conditions. The formation of
(a) ~ indications depends the magnetic flux lines' characteIistics,
including that they (1) form continuous loops; (2) do not
cross each other; (3) decrease in intensity with distance
from the surface; and (4) repel each other.
When a ferromagnetic particle is placed in a magnetic
field, it is drawn toward the magnetic source. As it gets
closer to a pole, more flux lines flow through it. This
J
concentrates the flux lines through the low reluctance
ferromagnetic path rather than the high reluctance air path.
(bJ It is this preferential action which causes palticles to begin
collecting at leakage fields and to subsequently form dis-

i
continuity indications.
18 I MAGNETIC PARTICLE TESTING
- --- ------ - - - ---~---

-- - - - --------------- - ---- - --

PART 5
ELECTRICALLY INDUCED MAGNETISM
a magnetic field in the same manner as the copper conduc-
Circular Magnetization tor. This magnetic field is known as circumferential rrwgne-
tization because the magnetic flux lines form complete
When an electric current flows through a conductor such loops within the object (see Fig. 21b).
as a copper bar or wire, a magnetic field is formed around A characteristic of circumferential magnetic fields is that
the conductor (see Fig. 21a). The direction of the magnetic the magnetic flux lines form complete loops without mag-
lines of force is always 90 degrees from the direction of netic poles. Because magnetic particles are only attracted to
current flow. When the conductor has a uniform shape, the and held where the flux lines exit and enter the object
flux density or number of lines of force per unit area, is surface, indications do not occur unless a discontinuity
uniform along the length of the conductor and it uniformly crosses the flux lines. The resulting accumulation of mag-
decreases as the distance from the conductor increases. netic particles forms an outline of the discontinuity over its
Because a ferromagnetic object is, in effect, a large exact location.
conductor, electric current flowing through the object forms

Inducing Circular Magnetization in a Test Object

FIGURE 21. Magnetic field generated around: Figure 22 illustrates a method for inducing a circular field
fa) a conductor carrying an electric current; and using a magnetic particle testing unit. The test object is
fb) a ferromagnetic test object used as a clamped between the contact plates so that electric current
conductor passes through it.
When tubes are tested by passing a current through
(aJ
MAGNETIC FIELD
them, the magnetic flux rises toward the outside sl:lrface,

r~ ~r
~(c=>, ~ l~ ,4> , ~Oc=>
~
L v
MAGNETIZING CURRENT
oJ'/-.J
L CONDUCTOR .
FIGURE 23. Inducing a circumferential magnetic
field using an internal conductor for: (a) a tube
with inside and outside surface discontinuities;
.and fb) multiple ring shapes with cracking on the
inside and outside surfaces

FIGURE 22. Inducing a circumferential magnetic

field in an object used as a conductor

(bJ

MAGNETIC FIELD
MAGNETIZING
ELECTRIC CURRENT CURRENT
 FUNDAMENTALS OF MAGNETIC PARTICLE TESTING / 19

\\ith negligible flux on the inside surface. The inside surface

i.c, often c(l'Ially important when testing for discontinuities. Longitudinal Magnetization
Sillcf' a magnetic field surrounds a conductor, it is possible
t() induee a satisLtctOlY field in the tube by inseliing a Electric current can be used to induce longitudinal fields
('oPlwr har Of some other conductor through the compo- in ferromagnetic materials. The magnetic Held around a
nellt and passing the current through the har. conductor is Oliented in a lengthwise direction by forming
This method is calJed internal conductor nwgnetization the conductor into a coil (see Fig. 25). Application of the
's('c Fig. 2:3). right hand rule shows that the magnetic field at any point
within the coil is in a lengthwise direction.
\\Then a ferromagnetic object is placed inside a coil
carrying an electric current (see Fig. 26), the magnetic flux
Magnetic Field Direction lines concentrate themselves in a longitudinal direction. An
object that has he en longitudinally magnetized is character-
The di feet ion of the magnetic lines of £()rce is always at ized by poles close to each end.
ri~ht angles to the direction of the magnetizing current. When a longitudinally magnetized object contains a
All easy way to determine the direction of the magnetic transverse discontinuity, a leakage field is produced that
flllx is to imagine the eonductor held in your light hand with attracts magnetic pmiicles and forms an indication. Figure
til(' thumh extended in the direction of the electric current 27 illustrates a typical coil found on magnetic particle test
Ilo\\,. Your clenched fingers then pOint in the direction of the systems used to locate transverse discontinuities.
It laglldic flux flow. This is 1..'11o\\'n as the right hand rule and
is illllstratt,d in Fig. 24.

FIGURE 26. Test Object containing a longitudinal

magnetic field induced by a ·coil
FIGURE 24. The right hand rule indicates the
direction of magnetic flux flow based an the
direct'ion the magnetizing current

FORTY-FIVE DEGREE CRACK

(DETECTED) TRANSVERSE CRACK
(DETECTED)

FIGURE 25. Formation of a longitudinal magnetic FIGURE 27. Formation of a transverse

field using a coiled conductor discontinuity indication during longitudinal
magnetization
 20 / MAGNETIC PARTICLE TESTING

Multidirectional Magnetization fields in different directions can be imposed on an object

sequentially and in rapid succession.
\Vhen testing for discontinuities in different directions, it When this is done, magnetic particle indications are
is standard practice to perform two tests, one with circular formed when discontinuities are favorably oliented to the
magnetization and the other with longitudinal. Two or more direction of any field. Such indications pe;sist as long as the
rapid alternations of current continue.
 FUNDAMENTALS OF MAGNETIC PARTICLE TESTING I 21

PART 6
MAGNETIC PARTICLE TEST SYSTEMS

Stationary Magnetic Particle Power Packs

Test Systems . Po\ver packs are tl~e electrical sources needed to produce
\Vet method hOlizontal magnetic palticle test systems
lllgh amperage, low voltage magnetizing current. They are
I:Vically consist of (1) a high current, low voltage magnetiz- used to magnetize test objects such as castings and forgings
ing source; (2) head stock and tail stock for holding test that are too large to be placed in a stationary testing unit.
ohjects and providing electrical contact for circumferential The size and weight of power packs prevent moving them
magnetization; (3) a movable coil for longitudinal magneti- and test objects are accordingly transpOlted to the test site.
zation; and (4) a paIticle suspension tank with an agitation The rating or current output of conlInercial power packs
svstem .
varies widely but is typically from 6 to 20 kA of magnetizing
. The basic components, along with magnetizing control current.
illdicators and ampere meters, are enclosed within a table The current is applied by ancillmy cable wraps, formed
coils, clamps and prods. Most'power pack units incOlporate
top structUl:al frame. Systems. are available in a large
Dumber of SIzes from a 25 mm (1 in.) contact plate opening an adjustable current control, one or two ammeters and an
lip t~) systems that are 6 m ,(20 ft) long. Manufacturer's .
automatic ...
shot duration timer.
prO\'lcle alternating current, direct current or a combination
of the. two. Maximum magnetizing current output is avail-
ahk: from 1,000 Ato 10 kA. Figure 28 shows a typical Mobile and Portab'e Testing Units
statlOn~uy or wet horizontal unit.
There are many applications where it is not possible to
bring the test object to the magnetic pmticle system. Mobile
FIGURE 28. Typical wet horizontal magnetic
particle test system
FIGURE 29. Typical mobile magnetic particle test
system
22 / MAGNETIC PARTICLE TESTING

units are one type of equipment that can be transported to The term portable equipment refers to compact, light
the test site and still provide relatively high magnetizing weight units that can be hand carried to the test site (see
currents (see Fig. 29). Traditional mobile units may be Fig. 30). Some portable units are mounted on wheeled carts
considered small versions of the power pack systems. While to facilitate portability. Like stational), and mobile equip-
some mobile units have a magnetizing current output of ment, portable units come in a variety of sizes, shapes,
6 kA, most are limited by size considerations to between 3 weights and amperage outputs. The most common method
and 4 kA. TransportabilitY is also improved by restricting the of applying current with a pOltable unit is with prods or
types of magnetizing current to alternating current and clamps. However, cable wraps and formed coils are also
half-wave direct current. Magnetizing current is applied to used in many applications. Reduced weight and size are
the test object by cable wraps, formed coils, prods and achieved by omitting the step down transformer needed for
clamps. Oil field portable magnetizing units can reach 15 kA demagnetization.
by capacitor discharge through internal conductors or cable
wraps.
Prods and Yokes
FIGURE 30. Portable magnetic particle testing Prods are magnetization accessories that can be used with
system with fixed distance prod assembly stationary, power pack, mobile and portable units. They
typically consist of a pair of copper bars 12 to 20 mm (0.5 to

FIGURE 32. Hand probe or yoke with a

preformed coil assembly

FIGURE 31. Circular magnetic field generated

around magnetizing prods
FIGURE 33. Longitudinal magnetic field
generated by a yoke
 FUNDAMENTALS OF MAGNETIC PARTICLE TESTING / 23

O,iS in,) in diameter with handles and connecting cables.

()JlC of the prods has a trigger to remotely activate the usually contain a magnetizing coil with a core of laminated
magnetizing current hom the unit's mainframe. Prods set transformer iron. Attached to the core are legs which may
IIp it circular magnetic field that diminishes in intenSity as either be fixed or <'lIticulated. \Vhen magnetizing current is
the distance between prods increases (see Fig. 31). applied to the coil, a longitudinal magnetic field is created in
Yokes are often cable connected to a mobile or portable the core and transmitted to the legs. When coupled to a test
IlIIi[ that proVides the magnetizing current. A yoke designed
object, a longitudinal magnetic field is generated between
\\'ith a self-contained magnetizing source is often called a
the poles as shown in Fig. 33.
if(/nd probe. Hand prohes contain small transformers that Yokes are often specified by their lifting ability or the
e:ellerate low voltage and high current (see Fig. 32). Yokes surface field they create midway between their poles, as
measured with a gaussmeter.
-241~ICPARTftL.E TESTING

PART 7
FERROMAGNETIC MATERIAL
CHARACTERISTICS

strength and zero magnetic flux) and increasing H in small

Magnetic Flux and Units of Measure increments, the flux denSity in the material increases quite
rapidly at first, then gradually slows until point A is reached.
A magnetic field is made up of flux lines within and At point A, the matelial becomes magnetically saturated.
surrounding a magnetized object or a conductor carrying an Beyond the ~<;aturation point, increases in magnetic field
electric current. The term nUlgnetic flux is used when strength do not increase the flux denSity in the material. In
referring to all of the lines of flux in a given area. Flux per diagrams of full hysteresis loops, the curve OA is often
unit area is called rrUlgnetic flux density (the number of lines drawn as a dashed line since it occurs only during the initial
of flux passing transversely through a unit area). magnetization of an unmagnetized material. It is referred to
There can be some confusion about the units of measure as the virgin curve of the material.
used to define these magnetic quantities. The unit of When the magnetic field strength is reduced to zero
magnetic flux was OIiginally called a rrUlxwell with one (point B in Fig. 34b), the flux denSity slowly decreases. It
maxwell being one line of flux. The unit of flux density was . lags the field strength and does not reach zero. The amount
the gauss with one gauss equal to one maxwell per square of flux density remaining in the material (line OB) is called
centimeter. In 1930, the International Electrotechnical residual magnetism or remanence. The ability of ferromag-
Commission redefined and renamed the gauss as an oersted, netic materials to retain a certain amount of magnetism is
or the intensity of a magnetic field in which a unit magnetic . called retentivity.
pole experiences a force of one dyne. 1 Removal of residual magnetism" requires the application
In 1960, the International Organization for Standardiza- . of a magnetic field strength in an opposite or negative
tion released ISO 1000:" The International System of Units direction (see Fig. 34c). When the magnetic field strength is
(S1). 'This document standardizes the metric units for first reversed and only a small amount is applied, the flux
magnetic flux. Flux Intensity is measured using the weber density slowly decreases. As additional reverse fieldstrength
(Wb) with one weber, equal to ·tQIS lines of flux. The flux is applied, the rate of ~'eduction in flux density (line Be)
denSity unit is the tesla (T) or one weber per square meter. increases until it is almost a straight line (point e) where B
One weber per square millimeter (or one tesla) is equal to equals zero.
1O,00Q gauss. The amount of magnetic field strength necessary to
reduce the flux density to zero is called coercive force.
Coercive force is a factor in demagnetization and is also
very important in eddy current testing of ferromagnetic
Magnetic Hysteresis materials.
As the reversed magnetic field strength is increased
All ferromagnetic materials have celtain magnetic prop- beyond point e, the magnetic flux changes its polarity and
erties that are specific to that material. Most of these initially increases quite rapidly. It then gradually slows until
properties are described by a magnetic hysteresis CUlve. The point D is reached (Fig. 34d). This is the reverse polarity
data for the hysteresis curve are collected by placing a bar of saturation point and additional magnetic field strength will
ferromagnetic mateIial in a coil and applying an alternating not produce an increase in flux density.
current. By increasing the magnetizing field strength Ii in \Vhen the reversed magnetic field strength is reduced to
small increments and measuring the flux denSity B at each zero (point E in Fig. 34e) the Hux denSity again lags the
increment, the relationship between magnetic field strength magnetic field strength, leaVing residual magnetism in the
and flux density can be plotted. material (line OE). The flux densities of the residual
The relationship between magnetic field strength and flux magnetism from the straight and reversed polarities are
denSity is not linear for ferromagnetic materials. A specific equal (line DB is equal to line OE).
change in H may produce a smaller or larger change in B as Removal of the reversed polmity residual magnetism
shown in Fig. 34, the initial CUlve for an unmagnetized requires application of magnetic field strength in the Oligi-
piece of steel. Statting at point 0 (zero magnetic field nal direction. Flux density drops to zero at point F in
 FUNDAMENTALS OF MAGNETIC PARTICLE TESTING I 2S

FIGURE 34. Hysteresis data for unmagnetized steel: (a) virgin curve of a hysteresis loop; (b) hysteresis
loop showing residual magnetism; (c) hysteresis loop showing coercive force; (d) hysteresis loop showing
reverse saturation point; (e) hysteresis loop showing reverse residual magnetism; and (f) complete
hysteresis loop

fa) fd)
B+ B+
ZERO FLUX DENSITY
.,, __ oA
A
ZERO MAGNETIC FIELD STRENGTH ,/
SATURATION POINT SATURATION POINT
/

H- \ 0
I
I
I
/

H+ H- H+

0
~REVERSE
B- MAGNETIZATION
SATURATION
B-
POINT

fbI (e,
B+ B+
A
RESIDUAL M~GNETISM \

• B
H- 0 H+ H-

REVERSE
MAGNETIZATION
B- POINT B-·

(e) (f'
B+ B+
A A

H- e H+ H- H+

COERCIVE FORCE /

REVERSE
B- RESIDUAL B-
POINT

LEGEND
8 MAGNETIC FLUX DENSITY
H + POSITIVE MAGNETIC FIELD STRENGTH
H - NEGATIVE MAGNETIC FIELD STRENGTh
 26 I MAGNETIC PARTICLE TESTING

Fig. 34f with the application of coercive force OF. Continu-

the magnetic field strength (B divided by H). Figure 3.5a is
ing to increase the field strength results in the magnetic
the virgin curve of a high permeability material and Fig. 3.5b
polarity changing back to its original direction. This com-
is the curve of a low permeability material.
pletes the hysteresis loop ABCDEF (note that the curve
CDEF is a mirror image of curve ABCF). The reciprocal of permeability is reluctance, defined as
the resistance of a material to changes in magnetic field
strength.
Magnetic properties and hysteresis loops vary widely
Magnetic Permeability between materials and material conditions. They are affect-
ed by chemical composition, microstructure and grain size.
One of the most important properties of magnetic mate- Figure 36a is a hysteresis loop for hardened steel and the
rials is perrneability. Permeability can be thought of as the loop is typical of a material with low permeability, high
ease with which materials can be magnetized. More specif~ reluctance, high retentivity and high residual magnetism
ically, permeability is the ratio between the flux density and that requires high coercive force for removal. Figure 36b is
the hysteresis loop for an annealed low carbon steel. It is
typical of a material with high permeability, low reluctance,
FIGURE 35. Magnetic permeability curves: low retentivity and low residual magnetism that requires a
low coercive force for removal.
fa) high permeability virgin curve; and fb) low
permeability virgin curve

fa)
FIGURE 36. Hysteresis loops: fa) hardened steel
hysteresis loop; and fb) annealled low carbon

/
//. . .l SATURATION POINT
steel hysteresis loop

>- /
I-
Vi / fa)
Z I
LU
o----------~-------- _____ >-
3
--.I
l-
v;
ll.. Z
LU
o
3
--.I
ll..

POSITIVE MAGNETIC
FIELD STRENGTH
fb) POSITIVE MAGNETIC
FIELD STRENGTH
~NPOINT
fb)
>-
>-
l-
l-
v;
......
v; ", Z
LU
z 1./' o
~ ----------~~------------
3 3
--.I
--.I ll..
ll..

POSITIVE MAGNETIC
POSITIVE MAGNETIC FIELD STRENGTH
FIELD STRENGTH
 FUNDAMENTALS OF MAGNETIC PARTICLE TESTING 127

PART 8
TYPES OF MAGNETIZING CURRENT

! Il the vel)' early days of magnetic paIticle testing, it was Use of Alternating Current in
'"ii, ·\(,d that the most desirable current for magnetization Magnetic Particle Tests
\\ ;t<, direct current provided by storage batteries.

\s knowledge of the magnetic particle process expanded There are three primal)' advantages to using alternating
ill td electrical drcuih)' continued to advance, many types of current as a magnetizing source. First, the current reversal
Ii uglldizing currents became available: alternating current,
causes an inductive effect that concentrates the magnetizing
ld!\\,;lve direct current and full-wave direct current. The flux at· the object surbce (called the skin effect) and it
(, 'Ill IS half-\~'ave rectified direct current and full-wave rec- provides enhanced indications of surface discontinuities.
tilH,d direct current are used for altemating current recti- Magnetic fields produced by alternating current are also
fl," I to produce half-wave and full-wave direct current. much easier to remove during demagnetization. A third
advantage is that the pulSing effect of the flux caused by the
current reversals agitates the particles applied to the test
object surface. This agitation increases particle mobility,
allOwing more particles to collect at flux leakage points, and
Alternating Current increases the size and visibility of discontinuity indications.
Concentration of the flux at the test object surface also
\lternating current is useful in many applications because
can be a disadvantage because most subsurface discontinu-
it is commercially available in voltages ranging from 110 to
ities are not detected. Another disadvantage is that some
~IO \ '. Elechical circuitry to produce altemating magnetiz-
specifications do not allow the use of altemating current on ...
iH~ (,~IITent is sim~e and relatively inexpensive because it
plated components when the coating thickness exceeds
(J! I h requires transforming commercial power into low
0.08 mm (0.003 in.). The flux in a test object may not be at
\ (lltage, high amperage magnetizing current.
peak value, depending on where within the magnetizing
III the United States and some other countries, altemat-
cycle the current is tumed off..
ill~ t'llJ"rent altemates sixty times in a second. Many other
t'lllilltries have standardized fifty altemations per second ..
'I'll(' alternations are called cycles. One hertz (Hz) equals
00,' c:'de per second and 60 Hz is sixty cycles per second.
Fi.!2;lIJ"c :37 shows the waveform of altemating current. In one Half-Wave Direct Current
c\(·1(', the current flows from zero to a maximum positive
\ ,t! II(' and then drops back to zero. At zero, it reverses When Single-phase alternating current is passed through
din'diol1 and goes to a maximum negative peak and retums a simple rectifier, the reversed'flow of current is blocked or
to 1.('l'O. The curve is symmetrical with the positive and clipped. This produces a series of current pulses that start at
1H'.~ati\'e lohes being mirror images. zero, reach a maximum point, drop back to zero and then
pause until the next positive cycle begins. The result is a
varying current that flows only in one direction. Figure 38
shows the waveform for half-wave direct current.
Half-wave direct current has penetrating power compa-
rable to single-phase full-wave direct current. Half-wave
FIGURE 37. Waveform of alternating current current has a flux denSity of zero at the center of a test
object and the denSity increases until it reaches a maximum
at the object surface. The pulSing effect of the rectified
.- wave produces maximum mobility for the magnetic parti-
Z cles; dry method tests are enhanced by this effect. Another
.J..i
3
::y: 0 ~----+-----~-----+----~-----4-- distinct' advantage of half-wave direct ~urrent is the simplic-
ity of its electrical components. It can be eaSily combined
with portable and mobile alternating current equipment for
weld, construction and casting tests.
TIME- One of the disadvantages of half-wave magnetization is
the problem in demagnetization: the current does not
 - --~- ------------
28 I MAGNETIC PARTICLE TESTING

reverse so it cannot be used for demagnetizing. Alternating Single-phase full-wave direct current has essentially the
current can be used to remove some residual magnetism but same penetrating ability as three-phase full-wave direct
the skin effect of alternating current and the deeper current. The current fluctuation causes a skin effect that is
penetration of half-wave direct current cause incomplete not Significant. It is also possible to incOlporate SWitching
demagnetization. devices in the circuitlY that reverse the current How. This
permits bUilt-in reversing direct current demagnetization.
Because of its Simpler components, the initial cost of
FUll-Wave Direct Current Single-phase full-wave direct current equipment is much
less than that of three-phase full-wave equipment.
It is possible for electrical circuitry to not only block (or One disadvantage of single-phase units is the input power
rectifY) the negative flOwing current, but to invert it so that requirement. Single-phase equipment requires 1.73 times
the number of positive pulses is doubled. Figure 39 shows more input current than three-phase units. This becomes
the waveform of single-phase full-wave rectified alternating very Significant at higher magnetizing currents where input
current. The resulting current is usually called single-phase values can exceed 600 A.
full-wave direct current

FIGURE 38. Half-wave direct current waveform Three-Phase FUll-Wave

Direct Current
Commercial elechic power, especially at 220 and 440 V,
is provided as three-phase alternating current with each
phase providing part of the total current. Figure 40 shows
the waveform of three-phase alternating current. Three-
ALTERNATING CURRENT INPUT _
phase full-wave magnetic particle equipment rectifies all
three alternating current phases and inverts the negative
flow to a' positive direction, producing a nearly flat line
direct current magnetizing current. Figure 41 shows the
waveform of three-phase full-wave direct current.

FIGURE 40. Waveform of three-phase alternating

current
. HALF-WAVE RECTIFIER _
+

n: ~lo~ - 1/60 2/60

HALF-WAVE DIRECT CURRENT OUTPUT
SECONDS

FIGURE 39. Single-phase full-wave direct current FIGURE 41. Three-phase full-wave direct current
waveform waveform

1+V\fVV\
u 0
TIME---
~ 1J;o\xxXXxxx TIME~
 FUNDAMENTALS OF MAGNETIC PARTICLE TESTING / 29

Tltree-phase full-wave direct current has all of the advan-

«lges of single-phase full-wave direct current plus some the current from one leg, resulting in an unbalanced line).
additional benefits. The current draw on the power line is lvfany power companies charge a higher rate to customers
"pread over three phases, reducing the demand by nearly with unbalanced, high current requirements. The three-
half. The demand on the line is also balanced, with each leg phase design also permits incorporating a qUick break
providing a pOltion of the current (single-phase pulls all of circuit that improves the formation of indications at the ends
of longitudinally magnetized test objects.
 30 I MAGNETIC PARTICLE TESTING

PART 9
DEMAGNETIZATION PROCEDURES

Ob' cts that have been magnetic particle tested retain

~e 'lgnetism. The amount of residual magnetism de-
some m,n the materiaI an d'Its con d't' L ow carb
cause the arc to deflect or wander. Arc deflection
PeneIs 1 IOn. on t lm
see . (caIIed arc blenc) is a palticular problem in automated
h
o . I' I
aled condition retams Itt e or no magnetIsm w 1 e . h'l welding systems that do not compensate for a shift in
t e anne . . [": Id [" I arc.
hardened alloy steels retam strong magnetIc He s lOr ong
periods of time. . '. . 6. Demagnetization may he required when remagnetizing
R ding or removing resIdual magnetIsm IS not a dIrect in another direction, if the second magnetizing field
Ii et' uc of the intensity of the retained magnetic field but intenSity is less than the original. If the second mag-
unc IOn f' h . [" fl' I netic field strength does not equal or exceed the ini-
. d' ct function 0 t e coerCIve IOrce 0 t 1e matena .
IS a Ire h h' h . . d I .
T h ere are Ina terials wit Ig . retentIVity
. d0'
an ow coerCIve tial field strength, the initial magnetic field remains
[" The ease of demagnetIzatIon epends on the mag- dominant.
lorce. . f tl t· I
netIc. properties or hysteresIs curve 0 Ie rna ena .
Reasons for Not Demagnetizing
Although demagnetization is generaIIy required, there are
Justification for Demagnetizing . occasions when it is not necesscuy Demagnetization is Bot
required when the test objects have very low retentivity
There are several ways that an ob~ect can be magnetize?: (such materials are demagnetized when the magnetic field
. d d magnetization from earth fIelds; use of a magnetIc strength is removed). On some occasions, the residual
III uce
chuck or plate during m~c h'IllI~g;
. mec.h ' IIy m
amca . d u~e d
magnetic field is such that it does not affect the function of
t ·
magne Iza, tion' and magnetIc
II partIcle testmg.
d DemagnetIza-
. I the object nor its service life. OccasionalIy, the test object
. IS
hon . req tll'red for the fa owing reasons, esplte t 1e source . is magnetic particle tested a second time, with equal or
of th~ magnetization. greater magnetic field strength in another direction.
Demagnetization is not necessmy when test objects are
1. A magnetized' object can affect the accuracy and subjected to extemaI magnetic fields such as clamping with
function of some, inst~l1men.t~ and meters. A vommon a magnetic chuck 4ming machining or hoisting with an
occu rrence in aircraft is the induced magnetization
, . electromagnetic crane. FinalIy, there is no need for demag-
resulting from traversing through the Earth s magnetIc netization if the test 9bject is exposed to a subsequent

~:~~val of the magnetic particle media following

heating above the Curie pOint (the temperature where
magnetic domains become random and the material be-
2. testing is necessary because resIdual pmtlCles can comes unmagnetized).
e problems during subsequent operations such as
caus . R . d . I
machining and surface coatmg. e~ame partIc es c~n
also cause excessive ,:ea~ on. movmg components m
assemblies. Demagneh~ahon IS n~cessary .because ~ux Methods of Demagnetization
leakage can retain partIcles despIte a typIcal cIeanmg
process. Curie Point Heating
Machining of magnet.ized .objects is objectionable
3. becaus . 'e Clll'pS.'mel
, .shavmgs may . adhere to the surface, All ferromagnetic materials containing magnetic flux call
el·ISrup ti' I19 the surface finish and duIIing the cutting be demagnetized by heating to a specific temperature and
allOwing the material to cool in the absence of an external
tool.
4. Magnetized objectslatltrac~ an d ~'etam. II' d I .
111eta IC . e :ns magnetic flux. The temperature at which the mateIiaI
during handling am c eanmg pnor to t1Ie app IIcc~tIOn changes from ferromagnetic to paramagnetic is called the
of surface coatings. The. entn~pped. metal pattlcIes Curie pOint. This temperature varies widely depending on
create selious imperfectIOns 111 pam ted Or plated alloy composition. For example, the Cmie pOint for nickel
surfaces. . . containing 1 percent silicon is 320°C (608 OF) while the
5. DemagnetizatiO 11 iS' reA,qUlre?1 wh en obJect~s cfl~'eldto be CUlie point for nickel containing .5 percent silicon is 4.5 0C
electric arc we l(e.
1d reSI( ua l magne IC Ie can (1l3 OF). The Curie pOint f()}' ferrous alloys ranges from
about 650 to 870°C (1,200 to 1,600 OF).
 FUNDAMENTALS OF MAGNETIC PARTICLE TESTING I 31

The transition from ferromagnetic to paramagnetic at the loops. The magnetizing current and flux intensity curves are.
( ~nrie point reverses on cooling and the matelial becomes plotted against time. When the current reaches zero, the
f(~rromagnetic in an unmagnctized condition. Some X-ray residual magnetism approaches zero.
diffraction studies show that this transition is not a clystal- Successful demagnetization depends on several require-
liJle structure transformation but a rearrangement of mag- ments. First, the magnetic field strength at the start of the t
Iletic domains. Demagnetization by heating through the demagnetizing cycle must be high enough to overcome the
( ~lllie pOint is the most thorough demagnetization possible coercive force and to reverse the direction of the residual
llllt because of its expense it is not commonly used. field. This is accomplished by demagnetizing at a slightly
higher current than that used in the magnetizing cycle. The
Electromagnetic Demagnetization second requirement is that, in each successive cycle, the
reduction of magnetic Held strength must be small enough
There are several techniques for demagnetizing an object that the reverse magnetic field strength exceeds the coercive
Ilsing electromagnetic energy. All of these techniques sub- force and reverses the flux direction from the previous
ject a magnetized object to a magnetic force that is contin- reversal. :rhis requires a number of cycles, depending on the
"ally reversing its direction and gradually decreasing in permeability of the material. Ten to thirty reversals are often
illtensity. required.
In Fig. 42, the top CUIve illustrates the magnetic field
strength used to generate the flux intensity CUlve below. As
Alternating Current Demagnetization
l he current diminishes in value with each reversal, the
hysteresis curve traces an increasingly smaller path. The A common method of demagnetizing small objects is by
h( lttom CUlve illustrates the decreasing residual flux inten- passing them through a coil carrying alternating current (see
,ity in the object, indicated by the shrinking hysteresis Fig. 43). The objects are moved into the coil while the

FIGURE 42. De,:"agnetization hystereSis loops with current and flux intensity curves
B+ +

>-
I-

~
H- ----------~~~~~~--~~------H+ ~~-~~~-r~~~~--------­
><
TIME_
~

- J-l - ----
+ -:T-
,,, I
FLUX CURVE

H- -~--~-~.......,----.......---,--;------- H +

CURRENT CURVE

j
8-
 32/ MAGN E
TIC PARTICLE TESTING

Alternating current coil Direct Current Demagnetization

FIGURE 4~.,.ng unit with rail assembly
demagnetlz The plinciple of demagnetizing \vith direct current is
identical to that of alternating current demagnetization. The
magnetic field strength or current must be sequentially
reversed and gradually reduced. One of the aJvantages of
reverSing direct current demagnetization is the deep pene-
tration that is possible.
Because reversing the direction of direct current is done
through elechical circuits, it is possible to control the rate of
reversal. The most commonlv used reversal rate is one cycle
per second or a frequency of 1 Hz. This produces' the
optimum depth of penetration, permitting the demagneti-
zation of large test objects. Direct current demagnetization
often reduces the residual field to a value lower than is
possible \Vith alternating current. In practice, the test object
is placed within the coil where it remains until the demag-
netization cycle is complete.

Yoke Demagnetization
Yokes or probes are often used for demagnetization when
portability is required. Either alternating current or revers-
ing direct current can be used, depending Qn the available
power supply. Pulsating half-wave direct current found in
. . for exposure to the maximum magnetic
·current IS fl?Wln~re then slowly and aXially withdrawn some
many self-contained power yokes cannot be used unless the
flux. The ?bJe~~~eacoil. Because flux intensity decreases ,vith
unit also contains a current reverSing circuit: Demagnetiza-
tion is accomplished by passing objects through the poles of
distance from its source, this procedure serves to reduce the
distanc~ fr?~~ strength. To ensure that the flux is reduced to
the yoke and withdrawing them while the current is flOwing.

I Ishoulc~ be. withdrawn to a

Yokes can also be used to demagnetize local areas on large
magnetIc
. ., 1um fIe Ieve,I the obiects
J objects. The poles are placed OR the surface to be demag-
a mmlll . t twice the coi c iameter. . netized, moved in a circular pattern, and then slowly
distance at leas ent demagnetization can also be accom- withdrawn while the yoke is energized. \Vhen demagnetiz-
Alt 'lting curr . 1 '1 d . d II
ern, I.' g the object m t le COl an gra ua Y ing small areas on a large object, care must be exercised to
plished by p aCl~ nt to zero. Some coils and some magnetic avoid magnetizing adjacent areas.
red~cing the curd:signs have built-in circuitry for current
partIcI~ syste::;n decaying alternating current is available on
reductIon. W I 'ts the current can be applied directly to
wet ho~zonta un~ 'the headstock and tailstock instead of
the ?bJect th~~U~ through the coil. This is more effective Demagnetization Practices
passmg the ~l ~;:chnique for long, circularly magnetized
than the COl There are practical limits to the demagnetization process.
objects. , limitations to alternating current demag- These limits ar~ controlled by the equipment, the size and
~hel.·e are ~f~~~e important is the f~lct that. alternating material of the object, and the Earth's magnetic field.
nehzatIOn. tv t. the magnetic flux at the object surface. Generally, the practical limit of demagnetization occurs at a
current concbe~trta ~s'e not effectively demagnetized by the point where a residual field remains but at a level that does
t st 0 ,lec s ell f . k' rr interfere or complicate the intended function of the object
Large e J. . t method because 0 ItS s III eHect.
alter~ating cfIlen tration also prohibits demagnetization of in service.
f
Tlus lack p~~eobjects piled in a basket (the alternating Longitudinal, residual magnetic fields are usually mea-
a number.o s;t 1
t demagnetizes only the outside surf~1Ce of
current skm1e eC t . hyer) Quantities of small objects can
sured with a field meter. Some meters read relative units
and are useful for comparison pUlposes only; other meters
b · t n t le ou el, . I' I read directly in tesla or gauss. The greatest flux leakage in a
o Jec s o . , d with altcmating current tec uuques on Y
be demagnetIze. '1 well separated Single layer with their longitudinally magnetized ohject is at the ends or comers of
bY PIacm. . g them 111. f'11le!
, f I "I
to the axis 0 t le coi . the object. These are the best places to check for the
long dimenSIOns pel , efTectiveness of demagnetization. Note that when the read-
 FUNDAMENTALS OF MAGNETIC PARTICLE TESTING I 33

ill!!," fire ill relative units there may be differences bctuxen

1//1' r{'(Jdill{!,s oj d{fJerent nzalllljactllrer~s' .field metel~S'. operation or to remagnetize in a longitudinal direction be-
The Ilmglletic fielu of a circularly magnetizeu object is fore demagnetization. This procedure allows the use of a
('(llnpldely contained within the object and there are no flux field meter to check the effectiveness of demagnetization.
II ·akag/' points, except at discontinuities. Therefore, field The E<uth's magnetic field is in a north-south direction
I( ' -
sf r(,llgth meters cannot indicate resiuual magnetism of a eir- and can cause problems when demagnetizing objects with a
('II lady magnetized object. A common practice is to perform high length-to-uiameter ratio. \Vhen low residual fields are
IOllgillldinal magnetization as the last step in a two-step required, these problems can be reduced by plaCing the
demagnetizing unit's coil axis in an east-west direction.

>n
is
d

Il
 34 / MAGNETIC PARTICLE TESTING

PART 10
MEDIA AND PROCESSES IN MAGNETIC
PARTICLE TESTING

The formation of reliably visible discontinuity indications powders have a higher permeability than the oxides used in
is essential to the magnetic particle testing method. An wet method suspensions. Yet a typical dry powder does not
important factor in the formation and visibility of indications produce indications of extremely fine surface fatigue cracks
is the use of the proper magnetic particles to obtain the best that are easily detected with wet method suspensions. High
indication from a particular discontinuity under the given permeability is desirable but is no more important than size,
conditions. Selection of the wrong particles can result in shape or the other critical properties. All of these charac-
(1) failure to form indications; (2) the formation of indica- teristics are interrelated and must occur in appropriate
tions too faint for detection; or (3) a distorted pattern over ranges in order for high permeability to be of value.
the discontinuity and the resulting misinterpretations.
In magnetic pmticle tests, there are two classes of media
that define the method: dry and wet. Dry method particles Coercive Force
are applied without the addition of a carrier vehicle. Wet
method pmticles are used as a suspension in a liquid vehicle. Materials used in dry method powders and wet method
The liquid vehicle may be water or a light petroleum dis- suspensions should have a low coercive force and low
tillate such as kerosene. retentivity. If these properties are high in dry powder.s, the
Magnetic particles are also categorized by the type of particle can become magnetized during manufacture or
pigment bonded to them to improve viSibility. Visible dming their first use, making them small, strong permanent
particles are colored to produce a good contrast with the test magnets. Such particles have an increased tendency to
surface under white or visible light. Fluorescent particles magnetically adhere to the test surface where they first
are coated with pigments that fluoresce when exposed to touch it. This reduces their mobility and produces a high
ultraviolet light. A third pigment category includes particles background, reducing contrast and masking relevant discon-
coated with a material that is both color contrasting under tipuity indications. .
visible light' and fluorescent when exposed to ultraviolet When wet method particles have a high coercive force,
fight. . they are also easily magnetized, prodUCing the same high
level of background. Magnetized particles are attracted to
any ferromagnetic material in the testing system (bath tank,
plumbing system or rails) and this causes an extensive loss of
Magnetic Particle Properties particles from the suspension. Particle depletion creates
process control problems and requires frequent additions of
The media used in magnetic particle testing consist of new particles to the bath.
finely divided ferromagnetic oxides. The pmticles can be Another disadvantage of magnetically retentive wet meth-
irregularly shaped, spheroidal, flakes, or acicular (elon- od particles is their tendency to clump, qUickly forming
gated). The properties of different materials, shapes and large clusters on the test object surface. These clumps have
types vary widely and some are discussed below. low mobjlity and do not migrate to leakage fields. This
causes distorted or obscured indications in heav~!, coarse
grained backgrounds. . '
vVhile the emphasis is on low coercive force and low
Magnetic Permeability
retentivity, there is an advantage to celtain levels of coercive
Magnetic p,uticles should have the highest possible per- force and retentivity. Low residual magnetization in dry
meability and the lowest possible retentivity. This allows particles appears to increase their sensitivity to diffuse, lo\~
their attraction only to low level leakage fields emanating level leakage fields formed by discontinuities lying below
from discontinuities. As the pmticles become magnetized, the surface. It is suspected that the small amount of polmity
they then attract additional pmticles to bridge and outline established in elongated pmticles assists in lining them into
the discontinuity, thus forming a visible indication. strings when attracted by weak leakage fields. This is similar
Magnetic permeability alone does not produce a highly to the effect of a compass needle swinging in the Emth's
sensitive particle material. For example, iron based dry comparatively weak magnetic field.
 FUNDAMENTALS OF MAGNETIC PARTICLE TESTING / 35

\ \ 'ct method particles are also enhanced by some coercive Wet Method Visible Particles
Ii )n'(' and residual magnetism. VelY fine wet method pmti-
Particles used in a liquid suspension are usually much
C)('<; hegin to collect at discontinuities as soon as they are
smaller than those used in dlY powders. Typically, the
appl ied to the test object surface. If the particles remain fine
particles for wet method testing range from 1 to 2.5 J.Ull
,11](1 s(,parate. their migration through the vehicle is velY while (hy powder ranges from 100 to 1,000 j.Ull. The upper
sl,)\\ I'l'c,lllse of their small magnetic field strength, small
limit of padicle size in most wet method visible materials is
sil'(' and tJj(~ir small mass in a high viscosity vehicle.
20 to 25 j..tI11 (0.0008 to 0.001 in.).
J III li<-atiol1s build lip very slowly under these conditions,
Larger pmticles are difficult to hold in suspension. Even
fakillg as long as ten seconds.
20 /-tm palticles tend to settle out of suspension rapidly and
Strollgly magnetized particles form clusters and adhere to
are stranded as the suspension drains off the test object.
fll<' fl'sl ohject surhtee as soon as bath agitation stops.
Stranded p;utides often line lip in drainage lines that could
Parfi,·I('s with low magnetic field strength cluster more
be confused vvith discontinuity indications.
sl()\\h while indications are forming. The leakage field at the
dis(,()'llfillllity draws the particles toward it anel the clusters
an' ('(lIlstantly enlarging due to agglomeration. 'At the same Wet Method Fluorescent Particles
tillH', lil(' dusters sweep up nearby fine pmticles as they PaIticles treated with a fluorescent dye or some of the
III()\(' Inward the discontinuity. visible pigments differ in size and behavior from black or
red (uncoated) visible partides. Fluorescent paIticles must
be compounded and stmctured to prevent separation of the
dye and magnetic material during use. A mixture of loose
Effects of Particle Size pigment and undyed magnetic material produces a dense
background and poorly formed indications. In addition, the
'I'll<' size and shape of magnetic particles play an impor- undyed magnetic pmticles may be attracted and held at
tallt f( )1(· ill how they behave when subjected to a weak leakage fi~lds but their lack of contrasting color makes them
lllaglldic field such as that from a discontinuity. Large, difficult to see.
I)('a\'\' particles are not likely to be attracted and held by a PrQducing fluorescent magnetic particles involves bond-'
\\'t'ak I('ah:.lge field as they' ni.ove over the Object surf~ce. ing pigment around each magnetic particle. The bonding
Ilo\\,l'\(,J'. \'(~ry small palticles may adhere by fliction to the. must resist the solvent action of petroleum vehicles and the
slIrLlc(' \\'lH're there is no leakage field and thus form an abrasive aotion occurring in pumping and agitation systems.
01 '.it' (·t i( )lIable background. Some manufacturers encapsulate the bonded dye particle in
a layer of resin. Particles built up synthetically are larger
than the minimum sized visible paIticles and such powders
Dry Powder Particles
have much fewer very fine or small particles. As a result of
\\'ithill limits, s~nsitivity to very fine discontinuities typi- their processing, fluorescent particles have a definite size
('alk illcreases as particle size decreases. Extremely small range that is maintained throughout the suspension's selVice
partid(·s. on the order of a few micrometers, behave like cycle.
diISt. TIH'y settle. and adhere to the object surface even
tl101Igh it lila), he very smooth. Extremely fine particles are
'.(']':' s(,lIsil iw to low level leakage fields but are not desirable
for prodtll'tioll tests hecause of intense backgrounds that Effect of Particle Shape
O\JSl'1 JI'(' or lIlask relevant indications.

Large paliicles are not as sensitive to fine discontinuities. The magnetiC materials used in magnetic particle testing
1l()\\(,\('I', ill applications where it is desired to detect large are available in a variety of shapes: spheres, elongated
dio.;(,()lItillllitit's. powders containing only large particles may needles or rods and flakes. The shape of the particles has a
1)(' IIwd. bearing on how they form indications. When exposed to an
\iost ('(:II11llercial dry powders are a carefully controlled extemal magnetic field, all pmtides tend to align along the
Iln\tlJr(' ()f p;lIticles containing a range of sizes. The smaller flux lines. This tendency is much stronger with elongated
Sl/.('<I particles pnl\ide sensitivity and mobility while larger particles such as the needle or rod shapes. Elongated shapes
\1/.(:<1 p;lrt id('o.; servE' two pmvoses. They assist in bUilding up develop internal north-south poles more reliably than spher-
Ill(h(';!! iOlls at larger discontinuities and help reduce back- oid or globe shaped particles, because they have a smaller
~r()lIll(l h.' a SOIt of sweeping action, brushing finer particles internal demagnetization field.
frulIl tli(' I('st ohject surface. A balanced mixture containing Because of the attraction of opposite poles, the nOlth and
a r;lllt;C of ~il',('s prO\ides sensitivity for both fine and large south poles of these small magnets arrange the particles into
ChS('()lItllllliti('s. \\ithout disruptive backgrounds. shings. The result is the formation of stronger patterns in
 36 / MAGNETIC PARTICLE TESTING

weaker flux leakage fields, as these magnetically formed

strings of particles bridge the discontinuity. The superior Visibility and Contrast
effectiveness of elongated shapes over globular shapes is
particularly noticeable in the detection of wide, shallow Visibility and contrast are properties that must be consid-
discontinuities and subsurface discontinuities. The leakage
ered .when selecting a magnetic pmticle material for a
fields at such discontinuities are weaker and more diffuse. specific testing application. Magnetic properties, size and
The formation of pmticle stIings based on internal poles shape may all be f~lvorable for producing the best indication,
makes stronger indications in such cases.
but if an indication is formed and the inspector cannot see
The disadvantage of elongated particles is their tendency it, then the test procedure has failed.
to mat and form clusters that mask indications. Particle Visibility and contrast are enhanced by chOOSing a p,utide
mobility is greatly enhanced by a spheroid or globular color that is easy to see against the test object surface. The
shape.
natural color of metallic powders is silver-gray. The colors
of iron oxides commonly used in wet method powders
are black or red. Manufacturers bond pigments to the
Dry Powder Shapes particles to produce a wide selection of other colors: white,
black, red, blue and yellow, all with comparable magnetic
The supeliority of elongated p<uticles in diffuse magnetic propelties.
fields holds true for dl)' powder testing. However, there is The white or yellow colors provide good contrast against
another effect that must be considered. Dry powders are mill surface ohjects. They are not effective against the
often applied to object surfaces by releasing them out of silver-gray of grit blasted or chemically etched surfaces or
mechanical or manual blowers. It is essential that the against blight, polished machine ground surfaces. For those
pmticles b~ dispersed as a uniform cloud that settles evenly applications, black, red or blue is used. The choice of color
over the object surface. Magnetic powder containing only depends on the surfaces of the test objects and the prevail-
elongated particles tends to become mechanically linked in ing test site lighting.
its contain~r and is then expelled in uneven clumps. The ability to bond fluorescent dyes to magnetic powders
When a powder behaves in this manner, testing becomes has produced a particle material that provides the best
very slow and it is difficult to obtain a smooth application possible visibility and contrast under proper lighting condi-
over the test object surface. Powders made of spherical tions. ,vVhen test objects are examined in ultraviolet light, it
particles flow evenly and smoothly under the same concli- is difficult not to see "the light ·emitted by a few particles
tions. A dly powder must have free flOwing properties for collected at a discontinuity. .
ease of application, yet must also have an optimum shape for Fluorescent particles are magneti9ally less sel]sitive thal1 .
the greatest sensitivity in forming strong indications over visible particles but the reduction in magnetic rete"ntivity is
weak leakage fields. These two conflicting reqllirements can more than offset by the increase invisibility and contn{st.
he met by selectively blending pmticles of different shapes. Visibility and contrast of fluorescent palticles are directly
A specific proportion of elongated pmticles must be related to the darkness of the testing site. In a totally
present for sensitivity and enough globular pmticles must be darkened area, even a small amount of ultraviolet enerb'Y
added to permit smooth and uniform application. activates fluorescent dye to emit a noticeable amount of
visible light. When the test site is pmtially darkened, the
amount of reqUired ultraviolet energy increases dnimatically
Wet Method Particle Shapes yet the emitted visible light is only barely noticeable.
Most militmy and commercial specifications require the
The performance of pmticles suspended in a liquid test site to be darkened to 20 lux (Ix) or 2 footcandles (ftc)
vehicle is not as shape dependent as dry pmticles. The or less, with a minimum ultraviolet intensity of 1,000
suspending liquid is much denser and more viscous than air microwatts per square centimeter (,u\V-CI1l - 2).' at the test
and this slows the movement of particles through the liquid object surface.
so that they accumulate more reliably at discontinuities.
Because of this slower movement, wet method pmticles
form minute elongated aggregates. Even unf~lVorable shapes
align magnetically into elongated aggregates under the Particle Mobility
influence of local, low level leakage fields. In suspension,
the pmticles are kept dispersed by mechanical agitation vVhen magnetic pmtides are applied to the surface of a
until they flow over the surface of the magnetized object. magnetized object, the pmticles must move and collect at
There is no need to add certain shapes to improve the the leakage field of a discontinuity in order to form a visihle
dispersion of the pmticles. indication. Any interference \vith this movement has all
 FUNDAMENTALS OF MAGNETIC PARTICLE TESTING I 37

(.n~ '(', Oll the sensitivity of the test. Conditions promoting or 3. Convenience: dry powder with portable half-wave
illlt'rfcrillg ''I'ith particle mobility are different for (hy and equipment is ensy to use for tests on site or in the field.
\ \ d IIwthod particles, \Vet method particles packaged in aerosol spray cans
are also effective for field spot tests.
Dry Powder Mobility
The dry powder technique is superior for locating sub-
I)n particles should be applied in a way that permits surface discontinuities, mainly because of the high perme-
lilt'llllo reach the magnetized object surface in a uniform ability and favorable elongated shape of the particles.
(·jnIHj ""ith minimum motion. When this is properly done, Alternating current with dl)' powder is excellent for smface
tlH' particles come under the influence of leakage fields cracks that are not too fine but this combination is of little
"llil(' suspended in air and are then said to possess three- value for cracks lying wholly beneath the surface.
dilll!'lIsionaJ mobility. This condition can be approximated \Vhen the requirement is to find extremely fine surface
011 sllrbl'cs that are veliical or overhead. cracks, the wet method is snperior, regardless of the
\\"ltt'll particles are applied to horizontal surbces, they magnetizing current in use. In some cases, direct current is
sdlk directly onto the surface and do not have mobility in considered advantageous hecause it also prOvides some
thre(, dilllel;sions. Some extension of mobility can be indications of subsurface discontinuities. The wet method
ad li('\(·r! hy tapping or vibrating the test object, agitating the also offers the advantage of complete coverage of the object
p~ui iell's and allowing them to move toward leakage fields. surface and good coverage of test objects with irregular
.\ltCrIldtillg current and half-wave rectified alternating cur- shapes.
relit plllsed direct current) can give palticles excellent
lll{ll,ilih when compared to direct current magnetization.
Visible or Fluorescent Particles
Wet Method Particle Mobility The decision between visible paliicles and fluorescent
particles depends on convenience and equipment. Testing
Tilt, \'Ispcmion of particles in a liqUid vehicle allows with visible particles can be accomplished under common
lltnlJilil\ {()f the paliicles in two dimensions when the
shop lighting while fluQrescent particles require a darkened
SIISI)('lbiotl Hows over the test object sUlface an?.in three , area and' an ultraviolet light source .
. diJlJ('lt~i()ns when the test object'is immersed in a magnetic Both wet method visible and wet method fluorescent tests
paliiei(' Lath. . have about th.e same sensitivity, but under proper lighting
Wt'I IIldhod pmiicles have a tendency to settle out of the conditions fluorescent indications are much easier to see.
S((Sp('II~i(Jn either in the tank of the test system or on the test'
ohJct'! sllrbce ShOli of the discontinuity. To be effective, wet
lildhod pcllticles must move \vith the vehicle and must
I"(-<lcll ("'\/'1':' surface that the vehicle contacts. The settling Magnetic Particle Testing Processes
ralt' of particles is directly propOltional: (1) to their dimen-
SiUllS; and (2) to the difference between their density and A test object may be magnetized first and particles
the l()\wr denSity of the liquid vehicle. Their settling ;ate is applied after the magnetizing current has been stopped
illYlTs('h propoltionaI to the liquid's viscosity. As a result, (called the residual method) or the object may be covered
the tl)obility of wet m'ethod pmiicles is never ideal and must v.rith patticles while the magnetizing current is present
he halalH'('d against the other factors important to wet (known as the continuous rnethod). \Vith test objects that
trwl hod k'it results. have high magnetic retenthrity, a combination of the resid-
ual and continuous methods is sometimes used.

Media Selection Residual Test Method

In the residual method, the test object is magnetized, the
'I'll(' <'il(li('(' hetween drv method and wet method magnetizing current is stopped and then the magnetic
t<'Clllli(jlh" is illf1uen<.:ed principally by the following con- pcuticles are applied. This method can only be used on
sid( 'r~tt inll." materials ha\iing sufficient magnetic remanence. The resid-
ual magnetic field must be strong enough to produce
T\ P' (lj' dis('ontinuitv (surface or subsurface): for discontinuity leakage fields sufficient for producing \risible
\Ill I~ II rLH.'(, discontinuities, dry powder is usually more test indications. As a mle, the residual method is most
s( 'lisit i\",'" reliable for detection of sUlface discontinuities.
;) Si /(' ()f ,II rface discontinuity: wet method paliicles are Hard materials with high remanence are usually low in
11'\11<llh h(~st for fine or broad, shallow discontinuities. permeability, so higher than usual magnetizing currents may
 38 / MAGNETIC PARTICLE TESTING

be necessary to obtain an adequate level of residual mag-

netism. This difference between hard steels and soft steels the magnetizing current are simultaneously stopped. TIl('
is usually not critical if only sUlface discontinuities are to be magnetic field strength continues to affect particles in tIl(>
detected. bath as it drains.
Either dly or wet method particle application can be used The wet continuous method requires more operator
in the residual method. vVith the wet method, the magne- attention than the residual method. If bath applicatiOll
tized test object may be immersed in an agitated bath of continues, even moment:uily, after the current is stopped.
suspended magnetic particles or it may be flooded with palticles held by a discontinuity leakage field can be washed
particle suspension in a curtain spray. away. If there is a pause between stopping the bath
In the immersion technique, the strength of discontinuity application and applying the magnetizing current, the SllS-
indications is directly affected by the object's dwell time in pension can drain off the test object leaving insufficiellt
the bath. By leaving the object in the bath for extended pmticles for producing discontinuity indications. Careless
periods, leakage fIelds have time enough to attract and hold handling of the bath and current sequence can seriollsly
the m,Ddmum number of particles, even at fine discontinu- hinder the production of reliable test results.
ities. If the test object has high retentivity, longer dwell time The highest possible sensitivity for very fine discontinui-
increases the sensitivity over that of the wet continuous ties is typically achieved by the foIlmving sequence: (1) im-
method. Note that the location of the discontinuity on the merse the test object in the bath; (2) pass magnetizing
object during immersion affects the accumulation 'of parti- current through the object for a short time during immer-
cles. Indications are strongest on upper horizontal surfaces sion; (3) maintain the current dUling removal from the bath;
and weaker on veltical or lower horizontal surfaces. (4) maintain the current during drainage of the suspension
Care must be exercised when remOving the test object from the test object; and (5) stop the magnetizing current.
from the bath or particle spray. Rapid movement can
literally wash off indications held by weak discontinuity
leakage fields.
Conclusion
Continuous Test Method
Magnetic palticle tests 'are effective nondestructive pro-
When a magnetizing current is applied to a ferromagnetic cedures for locating matelial discontinuities in ferromagnetic
test object, the magnetic field rises to a maximum. Its value objects of all. sizes and configurations. It is a flexihle
is derived from the magnetic field strength and the mag- technique that can be performed under a variety of condi-
netic permea~ility of the·.tyst object. When,the magnetizing tions, using a broad range 9f supplementary components.
current is removed, the residual magnetic field in the object Application of the magnetic particle method is deceptive-
is always less than the field produced while the magnetizing ly simple - good test results can sometimes be produced
current was applied. The amount of difference depends on with little more than practical expelience. In fact, the
the B-H curve of the materiaI. For these reasons, the development of the technique was almost entirely empirical
continuous method, for any specific value of magnetizing rather than theoreticaI.
current, is always more sensitive than the residual method. However, the method is founded on the complex plinci-
Continuous magnetization is the only method possible for pIes of electromagnetics and the magnetic interactions of at
use on low carbon steels or iron having little retentivity. It is least three materials simultaneously. In addition, there is the
frequently used with alternating current on these materials clitical consideration of the operator's ability to qualitatively
because of the excellent mobility produced by alternating and quantitatively evaluate the results of the inspection.
current. This volume of the Nondestructive Testing Handbook pro-
With the wet method,. the surface of the test object is vides many details on practical applications of the magnetic
flooded with particle suspension. The bath application and particle techniqne, while at the same time discussing its
important theoretical foundations.
 FUNDAMENTALS OF MAGNETIC PARTICLE TESTING I 39

-a/o)" REFERENCES
t iO/l
)('(/,

hee/
lath J. I Iallsman, Erich and Edgar Slack PhYSiCS, hventy-
;11.<;-
third edition_ PJinceton, NJ: Van Nostrand Publishing
l'lIt ( ;olllpany.
(·ss
sh-

'n
t.
 SECTION 2
GLOSSARY FOR MAGNETIC
PARTICLE TESTING
Calvin McKee, NOT consultant, Wayne, Pennsylvania
42 I MAGNETIC PARTICLE TESTING

For the purpose of consistency within the magnetic black light filter: A filter that transmits near ultraviold
particle testing industIy, the definitions here were taken radiation from while absorbing other wavelengths. s
from existing documents whenever ever possible. Almost all
the definitions have been slightly altered from their source
documents in order to reflect current terminology and to C
maintain the style of the Nondestructive Testing Handbook capacitor discharge method: A Single-shot magnetizatioll
series. This glossaq was prepared for educational purposes method using discharge from a bank of capacitors. A
and no other intention should be assumed. means by which electrical current is built up and stored
until a sufficient level is achieved to provide a pre-
A determined magnetic field in a test object, usual1~
acceptance standard: A specimen test object similar to saturation ..
the product to be tested, containing natural or artificial carrier fluid: The liquid vehicle in which fluorescent or
discontinuities that are well defined and similar in size nonfluorescent magnetic particles are suspended for
to or extent of the maximum acceptable in the product. 2 ease of application. See vehicle.:5
alternating current: An electric current that reverses the central conductor: An electric conductor passed through
direction of its flow at regular intervals. the opening in a part with an aperture, or through it
alternating current field: The active magnetic field pro- hole in a test object, for the purpose of creating a
duced around a conductor by an alternating current circular magnetic field in the object. 3
flowing in the conductor. circular magnetic field: The magnetic field surroundIng
alternating current magnetization: Magnetization by a an electrical conductor (test object) when a current is
magnetic field that is generated when alternating passed longitudinally through the conductor.·5
current is flowing.-t circular magnetization: The magnetization in an object
ampere: A unit of electIic current. Abbreviated A or amp. resulting from current passed longitudinally throllgh
ampere per meter: The magnetic field strength in air at the object itself or through an inserted central
the center of a single-turn circular coil having a conductor. 5
diameter of one meter, through which a current of one circumferential magnetization: See circular nwg-
ampere is flowing. Abbreviated A-m- 1 or Alm.4. netization.
ampere turns: The product of the number of turns of a coil coercive for~e: The reverse magnetic field strength need-
and the current in amperes flowing through the coil.·5 . ed to redqce bulk magnetism to zero ..
ar~: A luminous high temperature discharge· produced coil method: A method· of magnetization in which all or a
when an electric current flows across a gaseous gap.4 portion of the object is encircled by a current-carl)ing
~rc strikes: Localized burn damage to an object from the coiP
arc caused by breaking an energized electric circuit. coil shot: A technique of producing longitudinal magneti-
Also called arc burns. 5 zation by passing electric current through a coil encir-
arcing: Current flow through a gap, often accompanied by cling the test object. 3
intense heat and light. l coil technique: See coil method.
articulated pole pieces: On a magnetizing yoke, indepen- conditioning agent: An additive to water suspensions that
dently adjustable magnetic elements enabling the mag- imparts specific properties such as proper wetting,
netization of irregular test object profiles. particle dispersion or corrosion resistance.-s
artificial discontinuity: A manufactured material anomaly. contact head: Electrode assembly used to clamp and
See acceptance standard and reference standard. support an object to facilitate passage of electric
artificial flaw standard: See acceptance standard. current through the object fc)r circular magnetization ..')
contact method: See current flow technique.
8 contact pad: Replaceable metal pad, usually made of lead
background: In magnetic pmiicle testing, the appearance or copper braid, placed OIl electrodes to give good
of the surface against which test indications are viewed.;; electrical contact, thereby preventing damage such as
bath: See suspension. arc strikes to the test object.·s
bearding: See furring. continuous technique: A sequence where magnetic par-
Berthold penetrameter: A magnetic flux indicator con- ticles are applied to the test object while the magne-
taining an mtificial discontinuity in the shape of a cross, tizing force is present.-t
mounted below an adjustable cover plate.-t Curie point: The temperature at which ferromagnetic
black light: Electromagnetic radiation in the near ultravi- materials can no lOll gel' be magnetized by outside
olet range with wavelengths from 320 to 400 nm (3,200 forces and at which they lose residual magnetism
to 4,000 A).5 (between 650 and 870°C for most metals).,5
 GLOSSARY FOR MAGNETIC PARTICLE TESTING I 43

('UITcnt flow technique: A means of m~lgnetizing by examination: The process of testing materials, interpreting
passing current through an object USl11~ procl~ or and evaluating test indications to determine if the test
COli tact heads. The cnrrent may be alternatmg cm rent object meets specified acceptance cJitel~a. .
or reclified alternating current. 5 examination medium: A powder or suspenSIOn of magnet-
('u .... cut induction technique: A means of mag~}etizat~on ic particles aoplied to a magnetized test sur[ace to
ill which it circulating current is induced III a nng determine the presence or absence of surface or
component hy the influence of a fluctuating magnetic slightly subsurface discontinuities ..5
lIeld.'>
F
false indication: An indication that may be interpreted as
o being caused by a discontinuity but located where no
dark adaptation: The adjustmel~t of: the eye ove1: time to discontinuity exists.
reduced illumination, includmg mcreased retmal sen-
ferromagnetic material: A material that exhib.its th~
sitivitv, dilation of the pupil and other reHex physical . phenomena of magnetic hysteresis and magnetIc satu-
changes." . . ration and whose magnetic permeability is dependent
dcff.'d: A discontinuity whose size, shape, onentatIOn or on the magnetizing field strength.
location make it detrimental to the useful service of the field flow technique: See l1Ulgnetic flow technique.
test ohject or which exceeds the accept/reject cIiteria
fill-factor: In the coil method of magnetization, the ratio of
or all <{ppJicahle specification. 1 . .
the cross-sectional area of the object within the coil to
demagnetization: The reduction of reSIdual magnetIsm to
the cross-sectional area of the coil. 4
. an acceptable level:5 . . .
flash magnetization: Magnetization by a current flow of
dcmagnetizing coil: A coil of conductIve n:at~naI4carrymg brief duration. s See capacitor discharge method.
alternating current used for demagnetIzatIOn.
flash 'point: The lowest temperature at ,:",hi~h ~apo:s above
(liamagnetic material: A material with magnetic penne-
a volatile, combustible substance Igmte III aIr when
ahility less than l.
exposed to flame:5
direct ('o~ltact magnetization: See current flow technique.
flaw: See defect. ..
dircct eurrent: An electric current flOwing continually 'in
1 fluorescence: The emission by a substance of v1SIble
• (HI(, directio;) through a conductor.
radiation as a result of and only during the absorption
cJi"cel ('unent field: A residual magnetic field or an active
~f ultraviolet energy.
magnetic field produced by direct current flOwing in a
condllctor. I fluorescent magnetic particle testing: The process using
finely divided ferromagnetic particles that fluoresce
discontinuity: A change in the phYSical stmcture or con-
when exposed to ultraviolet light (320 to 400 nm).l
figuration of an object. May be intentional or
IIIlintelltiona1. 5 flux density: See magnetic flux density. .
flux indicator: A small device, generally a metal stnp or
domain: A saturated macroscopic substructure in ferro-
disk, containing artificial discontinuities. Used to de-
lIlag;netic materials where the elementmy pmticles
({'l(~ctroll spins) are aligned in one direction by inter- termine when correct mflgnetizing conditions or mag-
atomic f()rces. A domain would be a saturated penna-
netic field direction have been achieved.
n!'llt lllagnet.:1
flux leakage field: The magnetic field that leaves or enters
dry method: A magnetic pmticle testing method in which the surface of an object. s
thl' /(·rroll1agnetic p<llticles are applied in a dIy powder
flux leakage method: A method for the detection and
/()JTIl ..') analvsis of a discontinuity using the flux that leaves a
dry powder: Finely divided ferromagnetic particles select- magnetically saturated, or nearly saturated, test object
('d and prepared for magnetic paIticle testing.·3 at a discontinuity.2
flux lines: See lines of force.
fluxmeter: An electronic device for measuring magnetic
E flux.
d('drode: A conductor hy means of which a current passes
full-wave direct current: A Single-phase or three-phase
into or out of a test ~)hiect.·4
alternating current rectified to produce direct current
dedrolllagnet: A soft iro~1 core surrounded hy a coil of
characteristics of penetration and flow.
\\ire that temporc.llily becomes a magnet when an
furring: Build up or bristling of magnetic pmticles resulting
(·j(·dric (,IIITent flovvs through the wire ..5
from excessive magnetization of the test object:5
encin'ling ('oil: See coilm.cthod.
entluatiou: The proc('ss of determining the magnitude and G
sigllificallce of a discontinuity causing a test indication, gauss: A unit of magnetic flux densi~ or magneti~ induc-
after it has heen interpreted as heing relevant. tion. Magnetic field strength B IS measured m gauss
 44 I MAGNETIC PARTICLE TESTING

abbreviated (G). One gauss is one line of magnetic flux

interpretation: The determination of a magnetic pc.llticle
per square centimeter of are~.:3 See tesla.. '. indication's source and relevancy.·5
gaussmeter: A magnetometer usmg gauss to regIster held
strength.
K
keeper: Ferromagnetic material placed across the poles of
H a permanent magnet to complete the magnetic circuit
half-wave current: A unidirectional rectified single-phase and prevent loss of magnetism. 4
alternating current that produces a pulsating unidirec-
tional field.'" l
Hall effect: A potential difference developed across a laminated pole pieces: See articulated pole pieces.
conductor at dght angles to the direction of both the leakage field: See flux leakage field.
magnetic field and the electdc current. Produced. when leeches: Permanent or electromagnets attached to elec-
a current flows along a rectangular conductor subjected trodes canying magnetizing current, to proVide strong
to a transverse magnetic field. 4 electrode contact.'"
heads: The clamping contacts on stationary magnetic par- lifting power: The ability of a magnet to lift a piece of
ticle systems.:3 ferdtic steel by magnetic attraction alone,4
head shot: A short pulse of magnetizing current passed lines of force: A conceptual representation of magnetic
through an object or a central conductor whi~e.clamp~d flux based on the line pattern produced when iron
between the head contacts of a magnehzmg umt, filings are sprinkled on paper laid over a permanent
generating circular magnetization of the object. Dura- magnet. 5
tion of the current is usually less than one second.·3 longitudinal magnetic field: A magnetic field wherein the
horseshoe magnet: A bar magnet bent into the shape of a flux lines traverse the component in a direction that is
horseshoe so that the two poles are adjacent. Usually essentially parallel with its longitudinal axis,'''
the term applies to a permanent magnet.:3 longitudinal magnetization: Magnetization in which the
hysteresis: (1) The lagging. of the magnetic effec~ when th.e flux lines traverse the component in a direction essen-
magnetizing force actmg on a ferromagnehc body IS tially parallel to its longitudinal axis. 4
changed. (2) The phenom~n?n exhibited b~ a magn.etic
system wherein its state IS mfluenced by ItS .pr~VJ.ous . M .
histOIY· magnetic circuit: The closed path followed by any group of
hysteresis loop: ~ curve ~l~owing flux density B plott~~ as magnetic flux lines. 4 .

a function of magnehzmg force H as the magn~hzmg magnetic field: Within and surrounding a magnetized
force is increased to the saturation point in both the object, the space in which the magnetic force is
negative and positive directions sequentially. The curve exerted ..3
forms a charactedstic S shaped loop. Intercepts of the magnetic field indicator: A device used to locate or
loop with the B-H a;<is and the ~oints of minimum and determine the relative intenSity of a flux leakage field
maximum magnetizing force define important magnet- emanating from an object .."
ic charactedstics of a materiat3 magnetic field leakage: See flux leakage field.
magnetic field strength: The measured intenSity of a
magnetic field at a specific point. Expressed in amperes
per meter or oersted.
. indication: A magnetic pmticle accumuh:tion ywt serve~ as
magnetic flow tcchnique: \Vhen a test ohject or a pOltioll
evidence of a fielel leakage and reqlllres mteIvretatlOn
of it closes the magnetic circuit of an electromagnet.
to determine its significance. 5 .
The resulting field is longitudinal in direction.
induced magnetization: A magnetic field ge.nerated ~n an
magnetic flux: The total number of lines of f()rce existing
object when no direct .electrical contact IS m:lde... in a magnetic circuit. 4
induced current techmque: See current mductwn
magnetic flux density: The normal magnetic flux per unit
technique. . ' . area. Expressed in tesla or gauss. 4
inductance: The magnetism produced 1Il a ferromagnetIc magnetic flux leakage: See flux leakage field.
body by some outside magnetizing force:·3 . . . magnetic hysteresis: See hysteresis.
inherent fluorescence: Fluorescence that IS an mtnnslC magnetic leakage field: See flux leakage field.
characteristic of a materiaP
magnetic particlc test: A nondestructive test method
inspection: See examination. . . . . utilizing magnetic leakage fields and suitable indicating
inspection medium: See exammatlOll rnedlllm.
materials to disclose surface and near sllrf;lce
internal conductor: See central conductor. discontillui ties ..5
 GLOSSARY FOR MAGNETIC PARTICLE TESTING I 45

magnetic particle test system: Equipment providing the

electric current and magnetic flux necessiU)' for mag-
o
oersted: The cgs unit of magnetic field strength. Replaced
netic particle discontinuity detection. Provides facilities by the SI system's ampere per meter. 4
le)r holding components of v~u)'ing dimensions and for
overal1 magnetization: Magnetizing a complete object
adjusting and reading the magnetizing current.'!
with a Single energizing cycle:5
magnetic particles: Finely divided ferromagnetic material
capable of being individually magnetized and attracted
to flux leakage fields:5 p
magnetic permeability: See permeability. paral1el magnetization: A magnetic field induced in
magnetic pole: One of nvo sites on a magnet that gener- magnetizable matelial placed parallel to a conductor
ates magnetic fields. Flux leakage sites on an object.5 canying an electric current. 3 Not a recommended
magnetic powder: Magnetic particles in dl)' or powder practice for magnetic particle testing.
lemn with size and shape suitable for discontinuity paramagnetic matelial: A material with magnetic perme-
detectioll.~1
ability slightly greater than 1.
magnetic rubber: A specially formulated testing medium permanent magnet: An object possessing the ability to
containing magnetic pmticles. Used to obtain replica retain an applied magnetic field for a long period of
castings of component surfaces with discontinuities time after the active power of the field has been
being reproduced within the replica. A suitable mag- removed. 3
uetizing technique causes the migration of magnetic permeability: (1) The ease with which a material can
pmticles \vithin the medium to the position of the become magnetized. (2) The ratio of flux density to
discontinuity. -1
magnetizing force (BIH).l
~agnetic saturation: In a specific material, the degree of pole: See magnetic pole.
magnetization where an increase in H produces no powder: See dry powder. .'
iillther increase in magnetization. powder blower: A compressed air device used to apply dl)'
magnetic wdting: A nonrelevant indication sometimes magnetic palticles over the surface of a test object:5
caused when the surface of a magnetized object comes prod' magnetization': 'See currelit flow tecJl1iique.
in contact ,vith another piece of ferromagnetic material prods: Handheld electrodes for transmitting magnetizing
or a current carrying cable:5
current from a generating source to a test object. 4 .
magnetism: The ability of a magnet to attract or repel pulse magnetization: Direct or indirect application of a
an~ther l,nagnet or to attract a ferromagnetic material. high field intenSity, usually by the capacitor discharge
A force field surrounding conductors carrying electric method.
current. 1
magnetization: The process by which elemental), magnet-
ic domains of a matelial are aligned predominantly in Q
one direction. quick break: A sudden intermption of magnetizing
magnetizing current: The electric current passed through current:5 Used in magnetic palticle tests for materials
or adjacent to an object that gives rise to a designated with high residual longitudinal magnetism and limited
magnEtic field. to three-phase full-wave rectified altemating current.
magn~tizing force: The magnetizing field strength ap-
plIed to ferromagnetic matelial to produce magnetism.5
magnetometer: A device for measuring the strength of R
magnets or magnetic fields.l rectified altenlating current: A unidirectional electric
mult~.dire~tion~~ magnetization: Two or more magnetic current obtained by rectifying altemating current \vith-
fwlds m diHerent directions imposed on a test object out the deliberate addition of smoothing to remove the
inherent ripples. 4
sVfjucntiall)· and in rapid succession.4
reference standard: A specimen containing controlled
mtificial or natural discontinuities. Used for verif)':ing
N
near surface discontinuity: A discontinuity not open to the accuracy of discontinuity detection processes or
equipment. 4
IlIlt locc:ted near the surface of a test object. Produces
Ilroad, fuzzy, lightly held dry particle indications.s relevant indication: An indication caused by a condition
non~elevant ind~cation: A test indication produced by an or a type of discontinuity that requires e~aluation.5
(lC'cepta~)I(-' dIscontinuity or by spurious effects such as remanent magnetism: See residual m.agnetic field. 3
magnetIc "'Iiting, changes in section or the boundary residual magnetic field: The field remaining in a
hehveen materials of different magnetic propelties. -4. ferromagnetic material after the magnetizing force is
reduced to zero. 3
 46 I MAGNETIC PARTICLE TESTING

residual technique: Ferromagnetic particles are applied

to a test object after the magnetizing force has been through-coil method: See coil method.
discontinued. true continuous method: Test technique in which mag-
retentivity: A material's property of retaining to a greater netizing current is applied before application of mag-
or lesser degree some residual magnetism. 3 netic particles and is maintained \vithout intermption
ring standard: See test ring. throughout the examination:5
toroidal field: An induced magnetic field occurring in a
S ring test object when current is induced. See current
induction technique.
saturation level: See magnetic saturation.
sensitivity: The degree of capability of a magnetic particle
test to indicate surface or near surface discontinuities U
in ferromagnetic materials:5 ultraviolet light: Electromagnetic radiation \\lith \vave-
settling test: A procedure used to determine the concen- lengths between 200 and 400 nm (2,000 and 4,000 AL
tration of particles in a magnetic particle bath. The range of wavelengths used for fluorescent nonde-
shot: A shOlt energizing cycle in a magnetic particle test..5 structive testing is typically between 320 and 400 nm.
skin effect: The phenomenon that causes the magnetiza- Sholter wavelengths are vel)' hazardous. See black
light.
tion produced by alternating current to be contained
near the surface of a ferromagnetic object..5
slurry: A free-flbwing pumpable suspension of a fine solid V
in a liqUid. vehicle: A liqUid medium for the suspension of magnetic
subsurface discontinuity: See near surface discontinuity. palticles, often a light petroleum distillate or condi-
suspension: A two-phase system comprising finely divided tioned water. See carrier jluid:5
magnetic pc'trticles dispersed in a vehicle, often a liqUid visible light: Radiant energyo generated in the 400
. petroleum distillate:5 See vehicle. to 700 nm (4,000 to 7,000 A) wavelength range:5
swinging field:. See multidirectional magnetization.
W
T wet method: A testing technique in which the magnetic
tesla: The Systeme Internationale unit of measqre for particles are applied as a suspension in a liqUid veh~cle ..5
,magnetic flux denSity. One teshr is .e'quivalent to 10 4 wet. slurry technique: .A magnetic partic'le test in which
gauss. the palticles are suspended in high viscosity vehicle ..5
test piece: See reference standard.
test ring: A ring specimen typically made of tool steel, y
containing artificial subsurface discontinuities used to yoke: A V shaped magnet that induces a field in the area of
evaluate and compare the performance and sensitivity the test object that lies between its poles. Yokes may he
of magnetic particles:5 permanent magnets, altemating current electromag-
nets or direct current electromagnets. .
 GLOSSARY FOR MAGNETIC PARTICLE TESTING /47

REFERENCES
Ji.ccommended Practice for Field Inspection of New 3. \Veismantel, E. Materials Evaluation. Vol. 33, No.4.
Casing, Tubing and Plain End Drill Pipe, third edition. "Glossary of Terms Frequently Used in Nondestructive
API RP5A5. \Vashington, DC: American Petroleum Testing." Columbus, OH: The American Society for
Illstitute (1987). Nondestructive Testing (1975): p 23.
) Flrctromagnetic Testing: Eddy Current, Flux Leakage 4. Glossary of Tenns Used in Nondestructive Testing. Part
(/lid j\1icrowave Nondestructive Testing. The Nonde- 2 (November 1984). London, England: British Stan-
structive Testing Handbook, second edition. Vol. 4. R. darcls Institute.
;\lcMaster, P. McIntire and M. Mester, eds. Columbus, 5. Annual Book of ASTAl' Standards. E-269. Vol. 03.03.
OH: The Anlerican Society for Nondestructive Testing Philadelphia, PA: American Society for Testing and
\ 1986): p 654-659. Materials.
 SECTION 3
A HISTORY OF MAGNETIC
PARTICLE TESTING
Arthur Lindgren, Magnaflux Corporation, Chicago, Illinois
SO/MAGNETIC PARTICLE TESTING

PART 1
ORIGINS OF THE MAGNETIC PARTICLE
INDUSTRY

The Nondestmctive Testing Handbook is edited to avoid pipe. His work on this project resulted in the magnetic
using commercial wording - product names and trade- particle testing method in use today.
marks are deleted as a matter of policy. Because the history Alfred de Forest recognized the possibilities of the
of magnetic particle testing in the United States is so closely method if it could be perfected to detect cracks in any
linked with the inventors' corporation, generic text was not direction. This meant that the direction of the magnetic
always possible in this history. field in the object could not be left to chance, neither could
The chapter focuses on the magnetic particle industry's it be only longitudinal. The only means of magnetizing
origins in America, but important research and applications previously known were external magnets or coils carrying
were occurring at the same time in other countries, partic- current. Alfred de Forest used a system that passed mag-
ularly in Germany and Japan. netizing current directly through the test object.
This was the first recorded use of circular magnetization,
the method now so widely used. He also conceived of using
The 19205: Era of Discovery magnetic powders with controlled size, shape and magnetic
properties, essential for consistent and reliable results.
Magnetic flux leakage testing had its beginnings in the Conflicts with Major Hoke's original patent were worked
1800s. In 1868, a British engineering publication reported out the year de Forest met F.B. Doane (Fig. 1b), of
that discontinuities were being located in gun barrels using Pittsburgh Testing Laboratories. That year, 1929, the patt-
a magnetic compass to register the flux. 1 In 1876, A. Hering nership of AV. de Forest Associates was formed. In 19:34,
obtained US patent 185,647 for a test method using a the company became Magnaflux Corporation.
compass needle to detect discontinuities in rails. In 1967, Carl Betz (Fig. Ie) .wrote that de Forest and
Magnetic particle testing began in the ~enti.eth century Doane were individuals "who had the vision to see the value
and its development is the focus of this discussion. of a new idea and the courage and faith to devote their lives
to making this vision become a reality."2 kY. de Forest was
Hoke's Patent involved with magnetic palticle testing until his death in
1945. Doane remained active in the indus tty until his death
After World War I in 1918, Major vVilliam E. Hoke (US in 1963.
Army, on assignment to the Bureau of Standards) was
working on development of precision gage blocks as mea-
surement standards. Hoke observed that metallic particles
from hard steel parts being ground on a magnetic chuck The 19305: Years of Development
sometimes formed patterns on the face of the part, patterns
that frequently corresponded to cracks in the part's surface. Test System Development
This alert observation marked the birth of magnetic
particle testing (essentially, Hoke had recognized the basis In the early 1930s, de Forest and Doane showed great
for longitudinal magnetization). He applied for a patent enthusiasm for their practical testing method. They set up a
which was issued in 1922 but made no attempt to commer- small laboratory and shop in Doane's hasement and began
cialize the idea. making powder (dust, as it was then called). The pair
produced a variety of particles, making it possible to inspect
Contribution of Alfred de Forest both smooth and rough surfaces, machined pmts, castings or
welds. The particles could be used for detecting discon-
In the 1920s, Alfred Victor de Forest (Fig. 1a) was a tinuities of various 'widths and depths, and for detecting
research engineer, a consultant and a teacher at the both surface and some subsurface discontinuities. These
Massachusetts Institute of Technology. The study of metals first patticles were dry powders.
and their performance dominated his professional life. In The wet method (liquid suspension) technique was added
1928, de Forest was asked by Spang Chalfant Company to in 1935. At the \Vright Aeronautical Company in PatersoIl,
investigate the cause of failure in some of its oil well drill New Jersey, black magnetic oxide was suspended in a light
""""""'""""""""""""""""""""""~'II"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~ A HISTORY OF MAGNETIC PARTICLE TESTING I 51

petroleum product similar to kerosene. At the same time,

FIGURE 1. Pioneers of the magnetic particle the General Electric Company in Schenectady, New York
testing technique: (a) A.V. de Forest, one of the began using finely ground mill scale suspended in light oil.
method's inventors; (b) F.B. Doane, coinventor of In 1936, a German patent was issued to F. Unger and
the magnetic particle method; and (e) Carl Betz, R.S. HilpeIt who suggested that magnetic particles could be
magnetic particle authority and author suspended in water with wetting agents and rust inhibitors
added. About two years later, there were impoItant German
fa) developments in other areas of research. Based on a desire
to establish standardized magnetic paIticle test sensitivities,
a magnetic test gage was developed at the Reichs Roent-
genstelle at Berlin-Dahlem. \Vith few modifications, the
gage is the same type now known as the Berthold field
gage. 3
In the United States,' magnetic piuiicle testing was
introduced by Doane and de Forest to the Army Air Corps
at Vhight Field in Dayton, Ohio and to the Navy at the
Naval Aircraft Factory in Philadelphia, Pennsylvania. The
test method was soon being used in suppliers' plants as well
as aircraft repair centers. The original cracked san1ple used
for demonstrating the effectiveness of magnetic palticle
testing is shown in. Fig. 2. The reaction of those who saw the
particles attracted to the invisible cracks in his permanently
magnetized sample continually intrigued de Forest. 4
fbI In the early 1930s, an experimental magnetizing fixture
was first used to demonstrate the ~n<1gnetic particle tech-
nique (see Fig. 3). Figure ·4··shows an assembly of electrical
components like those built for early. applications. This
system employed alternating current for magnetizing 'and
was used to test tool steel bars. Figtire 5 shows a unit used
in the aircraft industry in the 1930s; it proVided circular and
longitudinal magnetization from storage battery power.

Early Inspection Applications

Hamilton Standard Company was among the first to use
the magnetic particle technique to inspect aircraft propel-
lers. Pratt and \Vhitney, producer of aircraft engines, soon
followed. The first airlines to use the method were American
Ie) Airlines and United Airlines. Both the Army and the Nav)'

FIGURE 2. Test object used by A.V. de Forest to

demonstrate the magnetic particle technique .
52 / MAGNETIC PARTICLE TESTING

FIGURE 3. Experimental magnetic particle testing FIGURE 5. Storage battery unit used by the
equipment used by F.B. Doane in t 930 aircraft industry from t 932 to 1940

FIGURE 4. Alternating current magnetizing

assembly (t 933)

FIGURE 6. Magnetic particle testing system first

used at the Indianapolis Motor Speedway (1936)
 A HISTORY OF MAGNETIC PARTICLE TESTING I 53

recognized the ,'alne of the technique for locating fatigue In Germany as well, the magnetic pcuticle test method
cfelch ilJ ('ngine parts, propellers and other highly stressed was developing parallel with welding techniques and their
pmis dllring periodic overhaul. .. use on steel structures, Magnetic particle techniques were
The r,tilroacls were also early users of magnetIc partIcle used to locate cracks and to detect misalignments of plate
techniql1f's, mainly fiJr the location of fatigue cracks in axles edges. Alternating current prods and direct current yokes
and IlwtiOIl parts of steam locomotives. The automotive were reported to produce the best results.
indllsln' hecame interested and by the end of the 1930s, Another early user was the Indianapolis Motor Speedway.
magn('iic p<uiicle systems were being used in metallurgical In 1936, magnetic particle testing was made mandatory
LtboratOlics and in some receiving inspection departments there for all steeling paIts. In that same year, more than 50
011 all experimental basis. The Greyhound Bus Company percent of the PaIts presented for testing were rejected. In
began its lise of the method for the location of btigue cracks 1948, Wilbur Shaw, president of the Speedway, stated that
in engine parts at overhaul. Early users also included steam the magnetic particle testing 111<~thod had contributed more
powl'r plants who began scheduling the testing method to their safety record than any other single factor. He
dllling maintcllance of turbines, boilers and piping welds. confirmed that not one accident at the track had been
Durint!; this same time, the merchant steel mills were caused by a defective steeling PaIt since the testing method
restriclil;g spending because of the Great Depression and was made compulsory.
had not yd hegun using magnetic palticle tests. However, Figure 6 shows the first magnetic paltiele unit sent to the
the method was accepted by the specialty steel producers, Speedway in 1936 to eliminate serious spindle failures,
those who made tool steel. This steel was more expensive which had been occurring during the two previous years. By
and top Cjllality was demanded. Seams in tool steel bars 1962, although not mandatory, owners were submitting
gencrall;' rcslIlted in cracked tools, punches or dies. The many engine parts for testing prior to pr~ctice runs. Figure
labor ('o.st of machining tools was high and specialty mills 7 shows a few of the many crankshafts rejected before'
were oit(,11 held responsible. Tool steel' and alloy steel failure in seIVice and a few magnesium wheels found
prodllcers were among the early advocates of magnetic defective by the fluorescent pon~trant met·hod. These . two
pmiic!(' tcchniqnes. methods continue to be widely used by the racing industry.

Improvements in Magnetizing Equipment

FIGURE 7. Racing driver Parnelli Jones and
inspector supervisor Ed Oelon reviewing parts Until the mid 1930s, most magnetic paIticle inspectors
with discontinuities at Indianapolis Motor made their own rnagnetizing equipment. About 1934, Doane
Speedway introduced the low voltage, 60 hertz alternating current unit
for steel mill bar stock testing. Up to this time, most
magnetizing equipment used direct current from storage

FIGURE 8. Mobile magnetic particle unit In

railroad shop, 1937)
 S4 / MAGNETIC PARTICLE TESTING

FIGURE 9. Standard design wet horizontal unit FIGURE 10. Magnetic particle test system for
119361 with motorized tailstock; used by the railroad axle units (1938J; one of the first
US Navy and US Army Air Corp systems designed to inspect a single kind of test
Object

batteries. The unit had a transformer with two taps in the

primary coil so that, by knife sV\itches, the output from the
secondmy could be adjusted. Several machines of this type
were designed to test heavy equipment in the railroads and
other industries. One of these (shown in Fig. 8 in a railroad
maintenance shop) is the prototype of today's mobile power
packs. It operated from a 220/440 V, 60 Hz power supply alternating current at diminishing amperages. Larger ob-
and delivered up to 3,000 A of low voltage magnetizing jects could be demagnetized while still in place on the
current through 4/0 flexible cables. The unit was made magnetizing unit.
mobile (wheeled) because most railroad parts were velY The 1930s also saw the introduction of the altemating
heavy and it was easier t.o move the testing system. current electromagnetic yoke. The yoke was limited then as
Around 19.38, the Navy and Army Air Corps agreed on a it is today to the location of surface discontinuities and
standard magnetic particle system design to be used in their found its'first uses as a preventive maintenance tooL The
overhaul sl~opSo (Fig. 9). It \vas a horizohtal, wet Ii1ethod, yoke was also repOliedly used in the metallurgical laboratory
direct current machine, with storage batteries and battelY of a steel mill as a means of locating discontinuities in
charger. The amperage :was controlled by a carbon pile sample disks 50 mm (2 in.) thick, cut from each end of
rheostat. The unit bore the designation AN for Army and forging quality billets.
Navy. Finally, this decade saw the first power plant applications
Another first came in 19:38 - a series of systems designed of the magnetic particle method, including the testing of
to inspect only one type of object (Fig. 10). This special steam turbine blades during overhaul.
design was built for the Denver and Rio Grande \Vestern
Railroad for the testing of railroad car axles. Many severe
Education and the Magnetic Particle Method
road failures with costly traffic delays and loss of equipment
had been occurring natio1l\\ide. During the first few months During this era, growth in many industries was hindered
of extensive magnetic patiicle testing, 45 percen~ of all by organizations that wanted to keep new developments
mO\ing locomotive parts were found to be defective. After proprietcuy. In industrial radiography for example, there
two years of planned overhaul programs, road failures due to was little published information available, even though
fatigue cracking were vitiually eliminated on the Denver I-LIl. Lester of \Vatertown Arsenal had demonstrated (as
and Rio Grande Western Hail road. early as 1922) that penetration of thick sections of steel \vas
Before 19.39, demagnetizers had been both crude and practical. He shovvecl that discontinuities such as cavities,
cumbersome. The use of alternating current for magnetiza- cracks, porosity anel nonmetallic inclusions could be re-
tion required special design efforts to attain closer current vealed where the discontinuitv thickness was as small as two
control. Research eventually produced a thirty-point motor- percent of the metal thickne~s.
ized tapswitcb that provided close control of the magnetiz- In 19:38 in Germany, Rudolph Berthold published The
ing current. This was a significant development that also Atlas of Nondestructive Test lVfctllOds which included the
provided the capability of rapid and automatic demagneti- phYSical fundamentals anel technical aids for magnetic
zation of objects through a succession of short shots of pmiicle testing. In 19.39, Doane proposed a textbook on
 A HISTORY OF MAGNETIC PARTICLE TESTING I 55

magnetic p<llticle testing, to include all of the design, Doane prevailed. The first edition of Principles of Magnaflux
application and method information then available - much Inspection was published in Febmary 1940 with Doane as
as this Nondestructive Testing Handbook does. Some indi- the author. He coauthored a second textbook, the Magnaflux
viduals in his own firm disagreed with this proposal but Aircraft Inspection Manual, with Willys E. Thomas in 1941.
56 / MAGNETIC PARTICLE TESTING

PART 2
EXPANSION OF THE MAGNETIC PARTICLE
INDUSTRY

The 19405: Organization FIGURE 11. Automatic magnetic particle testing

system for inspection of bolts and similarly
and Growth shaped Objects (early 1940sJ

Increased Wartime Production

The 1940s began with grO\\ing emphasis on military
procurement. Purchasing groups came to the United States
from Europe to buy militaIY hardware. The major manu-
facturers expanded production of weapons, vehicles and
aircraft. As a result, foundries, forging plants, machine tool
and gear manufacturers, steel mills and landing gear pro-
ducers were all affected.
As production increased, the need for magnetic particle
testing increased. Makers of tractors and diesel engines also
became users of .the technique. Suppliers of castings and
forgings showed genuine interest in the testing method.
Many machine tool and turbine manufacturers purchased
testing equipment for their metallurgical laboratories or
testing departments. Purchase orders requiring US militaly
specifications reqllired the use of magnetic particle testing.
\;Yorld War II not only created a need for more magnetic
particle testing, it also created a need for testing equipment
that could rapidly handle mass produced parts (see Figs. 11
and 12). Special purpose handling systems were designed
for larger and heavier objects (steel propeller blades, pro- FIGURE 12. Conveyorized 20 mm (O.B in.J
peller hubs, engine cylinders and engine mounts) and some magnetic particle testing unit (1942 J
of these systems were partially automated.
Up to this time, magnetic particle testing had been used
on many types of welded structures, paliicularly \velded
aircraft assemblies. However, on heavier welds, radiography
was the accepted nondestructive testing method and was
called for in the American SOciety of Mechanical Engineer's
Boiler and Pressure Vessel Code.
Because of the war, American shipbuilding soared and so
did the future of magnetic particle testing methods, for the
ships themselves and for their weapomy. The most common
naval weapon was the 127 mm (5 in.) gun for destroyers and
aircraft carriers. The gun's mounts were made of heavy steel
plate welded together to form a fixed and a rotating
platform; the welds were up to 150 mm (6 in.) thick.
Magnetic patiicle testing found an immediate application
for locating inclusions and laminations that interfered with
 A HISTORY OF MAGNETIC PARTICLE TESTING / 57

the making of crack-free root weld passes (early detection FIGURE 13. Mobile direct current (rectifier)
was critical hef()re depositing more weld metal). Direct power pack and powder blower ( 1942)
currcllt testing \\lith prods (Fig. 13) and the use of an
automatic powder blower also Oliginated at this time.

Effect of Military Specifications

The p,lce of production and the haste of all involved,
including the milital)" caused a great deal of over inspection
during the war. The first military specification on AN bolts
required 100 percent testing in both directions, even though
circlllllfi:rential discontinuities were the only ones of impor-
tance. ~I any usahle bolts were scrapped because of minor
seams or inclusions:5
III tIl(' "cars after the war, proposed military specifications
were 1I5,;all), reviewed by the potential users before being
isslled. This reduced but did not eliminate the problem
of o\'(,r inspection. There were still important differ-
ences between drawing specifications and floor inspection
. requin'lllents. Fluorescent Magnetic Particle Testing
In 1941, the' magnetic particle testing method took a
significant step fOlward with the introduction of fluorescent
~irect Current Testing and Qu!ck Break Des;9~s
magnetic palticles. The first fluorescent powders consisted
The sllperiOlity of direct current magnetization for iocat- of loosely bound agglomerates of magnetic powders and
iug sllhsnrface discontinuities had long been recognized. separate fluorescent powders. The two tended to be trapped
However, battel), maintenance represented a growing prob- together to form' fluorescent indications. The agglomerated
lem because of the need for larger and heavier duty units. In mixture was an improvement over visible powder but was
1941, the AN battel), powered series of wet horizontal test still not completely satisfactory.
machines was replaced on the assembly line by the ANQ Experimental work with the mixture of magnetic particles
(rectifier) sedes. and fluorescent powders had been canied out by a number
Wht'll a coil is used to impart a longitudinal field in a bar of individuals and groups. However, Robelt and Joseph
shaped object, special circuihy is required to ensure suffi- Switzer (Fig. 14a and 14b) are credited with first applying
cient fidd near the ends of the test object for detection of the idea of combining magnetic and fluorescent particles.
circumji'rential discontinuities. The condition or effect re- The brothers developed a method of coating magnetic
quired Jus been called qUick break or fast break. Battery particles with fluorescent material and this greatly enhanced
powered units had this feature built in. the particles' visibility. Pure black against pure white offers
Sp~'ciaJ quick break design considerations are required on a 25: 1 contrast ratio. Fluorescence in darkness offers con-
rectifier po\\'ered machines. The lack of this special design trast ratios as high as 1,000:1. Fluorescent particles can
can be catastrophic - one automobile producer dropped therefore be fewer in number while still offering a sizable
their sixty C,lr per hour production rate to zero when steel increase in sensitivity for detection of fine discontinuities.
conveyor pins hegan breaking in one of their main lines. The The bonding process had another advantage: it allowed the
m.agnetiC' p,lIticle test of these pins had been performed use of water instead of oil or kerosene as a· suspension
\\llth war SU!lJ1us systems built before the quick break vehicle.
phenomenon was recognized. Years later it was revealed that magnetic particle testing
This Olle occurrence was followed bv others in the years and fluorescent penetrant testing were very much a part
foll?\\1ng the war. Most of the trol~ble was caused by of building the first atomic bomb under the Manhattan
mamt<:>l1allce personnel who innocently removed what ap- Project. 6 These testing methods were also vital factors in
peared to be an extra breaker in the circuit of their the building of the first atomic reactor, erected beneath
machines. Even with today's modern circuitry, it is impor- the stands of Stagg Field at the University of Chicago.
tan~ tll.at all units using a coil for magnetizing be checked A.V. de Forest used both testing methods on various pieces
penodl<:all.\ to ensure that quick break is operating properly. of the reactor and the containers for the uranium fuel. Since
In. the 19(-j()s, a device for qUickly verifying the presence of that time, both methods have been intimately involved in all
tIllS feature was introduced. facets of the military and peaceful use of atomic energy.
 L-~'8!-w~m'lClpjiJ~IiJRrnTlcLE TESTINC.

FIGURE J4. Pioneers of the magnetic particle

testing technique: taJ RObert Switzer and CO!poration offered lecture courses, generally three dm's
tbJ Joseph Switzer, developers of the fluorescent
particle test long. DUring the war, more than fifty-five of their COurses
were presented to about five thousand people. After tIlt'
war, permanent schools were established by many nondE'-
structive testing firms, each specializing in the methods they
knew best. Most nondestructive testing innovators, includ-
ing pioneers like Phil Johnson, had a common philosoph;'
every defect has a cause - educated and constructivE'
inspectors can suggest a cure.

Peacetime PrOduction and Developments

Alfred de Forest said many times, "the closer the test can
be brought to the hot steel, the more economical it is."7 In
other words, catch the discontinuity early and save money
DUring the war, nondestl11ctive testing activities had been
directed toward supplying the needs of war material Con-
tractors and military agencies.
After t1le war, most metalworking firms continued to
emphasize production, not testing Or intensive quality con-
trol and all the nondestructive testing techniques, including
the magnetic particle method, had to be reintroduced. This
was done by selling the realities and advantages of accurate
testing. It was a slow educational process but by 1950 a
consumer oriented economy had helped shift the emphasis
toward quality. .
Peacetime manufactUring for profit in the late 1940s saw
an increasing acceptance of nondestructive testing. Many
metalworking firms were specifying magnetic particle test-
ing on purchased parts. Some finns (Cummins Engine
Company and the Electro Motive DiVision of General.
Motors COlporation, for example) established detailed pro-
cedures for magnetic particle inspeCtion of parts, including
testing procedures and accept or reject Criteria.
Developments in the hardware of magnetic particle
testing also helped improve the accuracy of the technique .
Technical Societies and Education and, in turn, the quality of products. In the 1930s, the first
mobile power pack for weld testing had used direct current
The 1940s also witnessed the beginning of nondestructive
testing symposiums and schools for industry. EvelY magnet- for magnetization. This provided a penetrating magnetic
ic particle testing system put into operation required oper- field (designed to supplement radiography for detecting
ators and inspectors, so people were hired by the thousands subsurface discontinuities) but it restricted the mobility of
and assigned to inspection work. Many of them knew little the magnetic pmticles. A method of magnetizjng c,{jled
about the equipment, the inspection method or what an surge magnetization followed in the 1940s. It was possible,
operator was expected to do. Military documents required by suitable Current control and S\vitching devices, to provide
contractors to train inspection personnel in botll theory and a velY high current for a short period (less than a second)
applications. Government inspectors then administered writ- and to then reduce the CUrrent Without interrupting it to a
mudl lower steady value.
ten and practical examinations before assigning government
certification. Compounding the training problem was a This surge magnetizing technique was used for many
shortage of supervisors and instructors. years to allow deeper penetration, but still did nothing for
Schools were soon established in impOltant industrial powder mobility. Half-wave magnetizing current was intro-
centers. For the magnetic pcuticle technique, Magnaflux duced after the war and provided increased penetration and
increased powder mobility. Despite the refinement of ra-
diographic techniques and the introduction of ultrasound.
 A HISTORY OF MAGNETIC PARTICLE TESTING / 59

!\"S half-wave magnetic palticle testing is widely used today for

FIGURE 15. Special purpose magnetic particle
the detection of both surface and near surface discontinuities
testing unit for raifroad car axles "ate 1940s)
in welds.
c- With the means available for magnetizing heavy cross
'\ sections in welds, castings and forgings, the rather simple
1- technique of passing test objects through an alternating
\; current coil for demagnetization was no longer always
(' adequate. A demagnetizer that could penetrate deeply was
necessary. One method utilized a direct current yoke and a
resonant carrier-to-interference ratio (CIR) circuit. Proper-
ly deSigned for a specific job, this method provided the
deepest penetration for demagnetizing. However, it was not
practical in those applications where several kinds of test
objects were involved.
The demagnetizer that did find wide acceptance during
this period was one that could be built into any direct
current power pack. A thirty-point switch and a set of
breakers provided reversing and decreasing direct current at
one reversal per second. This procedure was necessary for
FIGURE 16. Magnetic particle testing of steam applications such as demagnetizing heavy crankshafts used
turbine rotor blades in buck and off-road equipment.
The postwar years were most dramatic for the railroads.
DUling this period, many inspectors learned what the term
copper penetration meant: molten alloys from the bearings
of railroad axle journal~ penetrated the grain boundaries of
the heated journal. This prOvided the starting point for
fatigue cracks and eventual axle failure. The condition could
not be found visually nor with dly magnetic particles but
was readily seen with fluorescent magnetic particle testing'
(Fig. 15). The first mandatory use of magnetic particle
testing as a safety measure in railroads was introduced in.
1949 when the American Association of Railroads (Wheel
and Axle Committee) adopted the requirement for testing
of journals.
DUling the 1940s, developments also began in the utility.
indushies (see Fig. 16) and the petrochemical industries.
Firms laying pipelines, for example, began using magnetic
particle testing on their field-made welds, particularly on
the tie-in welds made after dropping the pipe sections into
place below ground level.

The 19505: Developing Markets

The early 1950s provided many new applications for
magnetic particle testing. In aerospace, it was the jet age.
Todav, airline accidents are seldom traced to structural or
engil{e component failure but this reputation had to be
earned and magnetic particle testing played a major part in
its achievement. Figure 17a shows the magnetizing of a jet
engine compressor disk; Fig. 17b illustrates the examination
of a compressor rotor assembly for transverse discontinuities
in the blades. Figure 18 shows a mobile unit being used on
a helicopter part in the field.
60 / MAGNETIC PARTICLE TESTING

FIGURE 17. Special magnetizing units employing Magnetic Particle Testing in the Steel Industry
the induced current method for: fa)jet engine By the mid 19.50s, many steel mills were consideIing
compressor disks; and fbI aircraft gas turbine magnetic particle testing as a means of increasing yield
compressor blading fassembled) rather than just as a crack finding method. By 1957, the first
fluorescent magnetic paIiicle testing system was installed
faJ for the production testing of tube rounds (matelial that was
subsequently pierced and rolled into oil well casing, pipe
and tubing).
The installation of this system was Significant because it
demonstrated that high quality steel product could he
obtained by spot conditioning of the semifinished product,

FIGURE 19. Automatic system for magnetic

particle testing of seamless tUbing: fa} loading
side; fb} inspection station; and fc} example of
fluorescent particle indication of spiral seams
(bl
(aJ

fbJ
FIGURE 18. Inspection of landing gear
components on a helicopter in the field using
a portable magnetic particle testing system

(e)
 A HISTORY OF MAGNETIC PARTICLE TESTING I 61

usiJl~ 111,l(~lldic particle testing; to disclose detrimental decade, thirty more major systems were installed in the
seaI;ls. '1'1;(' JlWdlOd proved far superior to unaided visual United States, Canada, Sweden, Gre,lt Britain, Italy and
testill~ pickliJlg and visual inspection, or skinning of the Australia. All of these systems incOllJorated design c~nsid­
entire ";lIrLl(,(' (ill spite of its great waste, skinning was erations not encountered before. The size and weight of the
whIch il,,('d at that tillle, especially for high quality steels). product, the effect of loose scale on the equipment, and the
In Hi!).'), aI/other fluorescent magnetic particle system was ambient temperature extremes in steel mills required the
install(d I()I" production testing of wire rod billets from development of unique testing hardware.
,50 1l11i1 ,2 ill.) to 64 mm (2.5 in,) square and nominally 9 1ll Figure 19 shows the JOO percent production testing of
(:30 fU j,. )II~. seamless tubes ranging from 100 to 250 111m (11 to 10 in.) in
Th(' S! I( Tt'S~ of these first f:"INO installations had a dramatic diameter and up to 12 m (40 ft) in length. This automatic
eHed Pi I t he magnetic particle industry. Over the next unit magnetized the tube by pas~ing direct current through
it while f1uorescent magnetic particles in a water suspension
were applied. Inspection under ultraviolet light revealed
surface seams, tears and other damaging discontinuities.
FIGURE 20. A combination ultrasonic and This ,,'as a finished test, coming at the end of the production
magnetic particle system for testing billets in process but before end trimming, threading or cutting iMo
the rolling mill shorter lengths.
Figure 20 shows the ultrasound portion of a combination
system for handling billets lip to 100 mm (4 in.) square and
12 m (40 ft) long at rates of three per minute. Billets with
discontinuities were segregated and divelted for proper
conditioning. The system used ultrasound f<:)r detecting
subsurface disc<.>ntinuities and n;mgneti(' paliicle testing·for
sUlface discontinuities. .

Magnetic Particle Testing in the

Automotive Industry
Quring the 1950s, the automotive industry began to
realize the potential of nondestructive testing. At that time,
almost all steel foundries used prod testing on their castings

FIGURE 21. Mobile magnetic particle testing unit

using prods; system employed half-wave current FIGURE 22. Production line yoke inspection of
on a 3,400 kg (7,500 IbJ steel casting gray iron castings
(early 1950s)
 62 / MAGNETIC PARTICLE TESTING

(Fig. 21). Gray iron foundries discovered the handheld

FIGURE 23. Wet horizontal and mobile magnetic magnetic yoke for detection of handling cracks in materials.
particle unit In an automotive overhaul shop Such tests were often included on the production line
(Fig. 22).
Automotive overhaul locations for trucks, buses and fleets
made use of the wet horizontal system, often with a mobile
unit as a power pack (Fig. 23). The grinding and heat
treating departments of manufacturing plants followed suit,
using small wet horizontal units for in-process testing
(Fig. 24).
In automotive manufacture, parts considered clitical and
subject to high loading were referred to as safety items and

FIGURE 25. Automated magnetic particle system

for automotive safety parts

fal
FIGURE 24. Small wet magnetic particle unit for
in-process testing after grinding or heat treating

fbJ

I'll"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~",""""""""'1""""""""""'-
 A HISTORY OF MAGNETIC PARTICLE TESTING / 63

slIbjcded to 100 percent magnetic particle testing. Ford areas were being checked. Figure 26 shows one of these
Mo'tor Company hegan 100 percent testing of steeling units in operation.
knuckles and arms as early as 1948 and other auto producers For testing objects easily handled on a wet testing table,
were to follow suit in the coming decade (Fig. 25). a fixtured machine was used in conjunction with the large
power pack (Fig. 27). A special combination system was
Magnetic Particle Testing for Forgings and Castings used when the test objects were very small and high
production was required, as in jet engine compressor blades
Dming this period, malleable iron and steel foundries (Fig. 28). The type of magnetizing current provided on
were dedicated to capturing business held exclusively by multidirectional units was direct current, alternating cur-
forge shops. Forgings were less prone to contain discon- rent, half-wave or a combination. The selection was made to
tinuities and when they did they were less likely to fail. suit the application.
HO\\'('\cr, castings were much less costly and magnetic
paJiiC'1c testing could be added to the production sequence
with c()llsiderable savings still remaining. Foundries began
offcrillg gllaranteed quality and were able to do so by
checking the pilot runs 100 percent. They seldom had to FIGURE 27. Multidirectional magnetizing of
buy hack any castings. 540 kg (1,200 Ibl steel missile forging
Large castings and forgings, too large for processing on
standard wet hOlizontal machines, were magnetized using
prods and mobile power packs. Unfortunately, the proce-
dure \\<lS tediolls because large surface areas had to be
painstakingly tested, a time consuming project when discon-
tinuities in both directions were possible: Retesting after
grinding sometimes rew~aled discontinu.ities missed.duling
the initial test. In 1956, the first high amperage power pack
was introduced, providing up to 8 kA of magnetizing· cur-
rent. This unit, installed at American Steel Foundry, was the
first lise of the overall magnetizing method, where the
entire brge casting ,vas magnetized in one operation.
A short time later, Birdsboro Corporation began using an
even larger'power pack. It not only had a high output but
was OIlC of the first multidirectional magnetizing deSigns,
one that could produce magnetic fields in a casting in more
than Olle direction during the same operation. The system
reduced testing time but still ensured that all eli tical surface

FIGURE 28. Multidirectional magnetizing system

for jet engine blades

FIGURE 26. High amperage direct current power

pack for overall multidirectional magnetization
of steel castings
 64 / MAGNETIC PARTICLE TESTING

Preventive Maintenance Uses of

Magnetic Particle Testing become a standard practice today. The year 19.57 w:itnessed
the first magnetic particle testing speCification for welded
During the 19.50s, magnetic particle testing equipment blidges used by the California State Highway' Depaltment.
purchased for. production \vas used more. and more f?r This specification \vas subsequentl.v adopted by many other
preventive mamtenance. Crane hooks, cham, forks on lIft technical societies and government agencies. Figure 311)
tnIcks, repair welds, fly\vheels, shafts, gears and hanger shows the test of another clitical \veld: the base of a
bolts gradually became part of planned preventive mainte- television mast on a bUilding in do\\l1town Chicago.
nance programs (Fig. 29). For expiosion-prooflocations, the In 1952, Taber de Forest, the son of Alfred de Forest,
permanent magnet yoke was used (Fig. 30) to prevent developed a testing method using a rubber sheet or balloon
shutdowns. inserted inside a pipe. \Vhen held in intimate contact with
In this period, shipyards began using the test ~m tailshafts, the test sUlface during magnetization, it retained the leak-
propellers, rudde~s and stem posts. In the oil fields.' pO.rta- age field from the test piece. The sheet was filled ",itl)
ble magnetic pmtlcle power packs were used .on dnIl pIpe, permanently magnetizable particles and the induced leak-,
casinO', tubing, drill collars, sucker rods and gnth welds. age field could be located on the sheet by means of
Str~lctural \veld testing with magnetic particle received its magnetic particles. This was espeCially useful for inspecting
first impetus in California \vhere the State Department of the inside of pipe or other cavities where indications could
Education began to use welded, single story school build- not be directly observed.
ings as a safeguard against earthquake damage (Fig. 31a).
The practice of all \velded f~lbrication spread slowly but has
Automated Visible Particle Scanning and
Induced Current Magnetic Particle Tests
FIGURE 29. Magnetic particle testing of a
fork lift The first experimental scanner for detecting and record-
ing magnetic particle indications was. installed dming the
1950s on a magnetic particle testing machine for handling
'large diameter resistance welded steel pipe (Fig. 32). The
scanner saw black particles on a white background. It made
use of ~. rotating mirror, a. small elongated aperture and a.
.
, FIGURE 30. Permanent magnet yoke used in an
explosion-proof room
 A HISTORY OF MAGNETIC PARTICLE TESTING / 65

FIGURE 31. Examples of magnetic particle FIGURE 32. Dry powder magnetic particle testing
methods on steel structures: (a) testing welds of of resistance welded pipe using a scanner for
an I·beam in a single story welded structure; and detecting and recording indications
(bJ inspecting a weld at the base of a 85 m
1285 ft) television tower

fa)

Ib,

. FIGURE 33. Noncontact magnetic particle system

using the induced current method for testing of
bearing races

photomultiplier tube. The refinement of fluorescent paIti-

cles alld a rotating prism with an ultraviolet light source
considerahly increased the system's sensitivity.
One of the first magnetic 'p,-uticle systems 'using induced
current is shown in Fig. 33. The induced current method
locates circumf<"rcntial discontinuities without making con-
tact to the bearing race. The unit provides magnetization by
central conductor. inspection, magnetization by induced
c.urrent. inspection and demagnetization. Another produc-
tIon. ullit using induced flux was later designed for the
testlll~ ()f hardened steel bearing balls, also eliminating the
need for electrical contact with the test object.
66 / MAGNETIC PARTICLE TESTING

The Nondestructive Testing Handbook bec,~me the focus of research for automating magnetic
The Society for Nondestmctive Testing (SNT) became partIcle tests.
very active during the 1950s. One of their important 'What was sought was a means of detecting and evaluating
contributions at that time was the publication of the first fluorescent magnetic particle indications more rapidly. mon'
edition Nondestructive Testing Handbook. accurately and at lower cost. The first practical applicatioll
Volume 2 of that edition includes chapters on electrified came in the 1970s when a fully automatic magnetic pmtide
particle and magnetic palticle tests, test indications, princi- scan unit was deSigned, built and successfully installed ina
ples and equipment. \-Vork on the Nondestructiue Testing bearing plant (see Figs. 35 and 36). Bearing~ were magne-
Handbook began in 19.50 and was continued by Robert tized and a bath applied. Bearings were then transferre~l t()
Mc~laster until its publication in 1959. Carl Betz, Taber de
a turntable and rotated under a vibrating laser beam,
Forest, Robert Eichen, Franklin Catlin, G.B. Baumeister Indications were seen by a phototube and an electric signal
and Kenneth Schroeder were among the magnetic particle was produced to divert defective bearings into a reject bill ,
L

authorities who contributed to the Nondestructive Testing ~utom~tic tes~ing ?f torsion bars was another early applica-
Handbook's first edition. tIon usmg a vlbratmg laser beam (Fig. 36c).

Advances in Measuring Magnetic Field Strength

The 19605: Industry Challenges Wo~k on de:eloping a reliable magnetic field gage began
early m the 11lstory of magnetic particle tests. In 193H, a
Advances in Test System Portability prototype of the Berthold magnetic field gage was cleve!-
Before developments in the 1960s, field testing was oped at the Reichs Roentgenstelle in Germany. In the eark
difficult with magnetic particle techniques. A typical testing 1940s, Lamellar test gages and wedge gages' were used i;)
unit measured 0.9 X 0.9 X 1 m (36 X 36 X 40 in.) and Germany for checking test sensitivity and penetratioll
weighed 180 to 270 kg (400 to 600 lbs). Magnetizing current depth.8
control was accomplished by adding long output cables. , In the United States through the 1950s, there was llO
Often, two 27 m (90 ft) lengths of 410 cable and two 4.5 m commonly us~d technique that permitted the exact mea-
(15 ft) lengths of 219 cable were needed to reduce the surement of field strength at a given point "vithin a piece of
current to the level required for pro.d inspection at 150 mm magnetized iron or steel. Inspectors depended on mles of
(6 in.) spacing. A large crane was used to move the test thumb and experience to determine what amount of mag-
system to the site and tbe inspector llsually had to drag netizing force should be used and in what direction it Sh01l1d
cables weighing 90 kg (200 lbs) through the stmcture. be applied. '
In 1961, work was begun at the Mare Island Naval
Shipyard to design and build a portable magnetic particle
testing system with a full-time use duty cycle. In 1965, a unit FIGURE 34. Portable magnetic particle testing
weighing 19 kg (42 lbs) was approved for use by the US system in use on an offshore oil drilling platform
Navy. Over the follOwing decade, improvements were made
to the basic design, mostly in the area of current control.
Such portable systems (see Fig. 34) helped lower the cost of
testing and dramatically improved accessibility for magnetic
palticle testing in shipyards, high rise buildings, off-shore
stmctures, bridges and pipelines.
Portability combined with the introduction of-articulated
legs for magnetizing yokes were the two developments that
have had the greatest impact on contemporary magnetic
patticle system designs.

Advances in Fluorescent Indication Detection

Even though fluorescent magnetic particles were easy to
see and provided for more rapid testing, the human eye was
still necessary for most magnetic particle testing. Each test
object had to be looked at and an evaluation had to be made.
In the 1960s, work began in earnest to automate many forms FROM ECONOSPECT CORPORATION. REPRINTED WITH PERMISSION.
of nondestructive testing. Laser beam technology soon
 A HISTORY OF MAGNETIC PARTICLE TESTING / 67-

During the 1960s, devices were developed and manufac- The proper level of sensitivity for detecting various sizes
tured witI) the intended purpose of ensUling that the field of discontinuities was achieved by vmying the artificial
distribution was of the proper magnitude and direction. In discontinuity's width and its depth below the surface. 'When
effect, these were adhesive altificial discontinuities. They used properly, these were a valuable testing component.
attempted to shunt some of the field out through the surface They did, however, more truly indicate the magnetic field
of the object into an extemal test piece and then back into strength H through the indicator rather than the flux denSity
the surfa~'e of the part again. One of these devices was the B through the surface layers of the magnetized test object.
BeIthhold field gage. Another was called a rruzgnetization Other versions of the original flux shunting devices
indicator or quantitative shim. included a reusable model. It could be handheld by the
These indicators were made of low carbon rolled steel. operator while magnetizing. Another semi reusable device
Into each had been machined an mtificial discontinuity in was developed in Japan in 1962. It featured more precise
the form of a straight line. In use, the indicators were placed artificial discontinuities and provided the discontinuities in a
on the test object so that the artificial discontinuity lay in the circular pattern that could be used to verify magnetizing
direction of cracks that were expected to occur. The test force in more than one direction.
object was then magnetized and magnetic palticles applied
Development of Test System Components
in' the usual manner. If the' artifIcial discontinuity in the
indicator was shown, then magnetization was considered In the early 1960s, some preventive maintenance and
acceptable. safety testing required more sensitivity than dly powder

FIGURE 35. Diagram of a laser scanning automatic discontinuity detection system based on the
fluorescent magnetic particle technique

PATTERN
RECOGNITION
ELECTRONICS

SCANNING MIRROR

·1
LIGHT COLLECTING
MIRROR

THRESHOLD
GATE

STATIONARY MIRROR

\\ FLUORESCENT
INDICATION
68 / MAGNETIC PARTICLE TESTING

could provide. Press1.llized cans of fluorescent magnetic

FIGURE 36. Magnetic particle systems employing
paIiicle suspension were introduced in a kit that also
laser technology: I a) system modified for use of
included an ultraviolet light source. The setup shown in
lasers in tests of bearing rollers; Ib) magnetic
Fig. 37 continued to be \videly used into the 19805 by
particle laser unit for rollers; and Ie) laser scan
meeting the safety testing re(lllirement of a six month check
testing of torsion bars
on all forks on lift trucks and crane hooks, a standard
fa) practice in many metalworking plants.
In 1962, copper oxide rectifiers. long the standard in
direct current wet holizontal units, were replaced by recti-
fier stacks using more efficient silicon diodes.
The demands of the flolllishing aerospace industry also
encouraged advances in magnetic padicle systems. Missile
motor case components required special magnetizing t~ch­
niques and fixturing in order to avoid arc bums from
elechical contact and to provide sufficient magnetizing
force in all directions. The Polaris, Titan III and Minute
Man missiles typically used such motor cases. Some motor
cases were as large as 3 m (10 ft) in diameter and 3.3 m
(11 ft) or more in length, with metal thicknesses less than
6 mm (0.25 in.). .
Figure 38 illustrates a yoke device that provided all
induced longitudinal field used in conjunction with a special

fbI
FIGURE 37. Portable magnetic particle testing
with aerosol and dry powder for preventive
maintenance and safety inspection

fe)
 A HISTORY OF MAGNETIC PARTICLE TESTING / 69

FIGURE 38. Induced mag~etic fiel~ har~ware: FIGURE 39. Special induced alternating current
(al special direct current Induced fIeld f,xture fixture for locating circumferential discontinuities
using a yoke to ensure detection of all around the nozzle area of a solid fuel motor case
circumferential discontinuities; and (b) diagram of
an induced field fixture

.small ball bearings and roller beaIings. Introduced in. +966,

. the method's advantage was that it allowed through-cutrent
(bJ without over-heating or burning the finished surface.

Expanding Applications for

Magnetic Particle Testing
In 1968, a lock was built in Sault Ste. Malie, Michigan. Its
C gates were 18 m (60 ft) high, 19 m (63 ft) wide and up to
140 mm (5.4 in.) thick.. Both magnetic pmticle and ultra-
sonic testing were specified by the COlpS of Engineers to
ENERGIZING COIL detect cracks and lack of penetration. This is one of many
applications now dependent on successful magnetic particle
tests.
Many skyscrapers also specified magnetic particle and
ultrasonic testing to verify weld integrity. The John Hancock
Building and the Sears Tower in Chicago are typical. The
Hancock BUilding's 100 stories total 332 m (1,107 ft) in
height and the Sears Tower is 435 m (1,450 ft) tall. The
central conductor arrangement (not shown) for testing thin Sears Tower required 775,000 magnetic particle tests, in-
walled motor cases with a circumferential field. Figure 39 cluding recordkeeping and repOlting. All was completed in
shows a ~pecial fixture for inducing alternating current in compliance with the American Institute of Steel Constmc-
the cases nuzzle area without making electlical contact \\lith tion, the American Welding Society and the American
the test ohject. This technique was used for detecting Society for Nondestmctive Testing. Figure 40 represents a
critical circllmierential discontinuities in the welded area. typical magnetic particle test of welds on a high lise tower.
In other indllstries, the method of magnetization called Production applications in the 1960s included quality
short shot lIUi,!!..IICtizil1g was being refined for the testing of steel railroad wheels shown in Fig. 41a (the unit handled
 70 / MAGNETIC PARTICLE TESTING

FIGURE 40. Magnetic particle testing of welds on

a thirty-two story building in Detroit, Michigan FIGURE 4'. Production line applications of
magnetic particle testing: faJ combination
multidirectional magnetic particle and ultrasonic
testing of railroad wheels; and fbJ residual
magnetic particle testing of nuts at 80 per
minute

fifty-five to sixty-five wheels per hour). Inspection of nuts at

5,000 per hour was performed using the magnetic particle
testing residual method (Fig. 41b). Many versions of semi-
automatic magnetic pmiicle systems were made for the
testing ofIarge crankshafts (Fig. 42). Multidirectional mag-
netizing of complex steel castings was introduced (Fig. 43). Advances in the Testing Profession
The c'onventional approach was used f()!' VelY large gas
turbine shafts and spindles used by utilities (Fig. 44). The first magnetic particle inspectors at nuclear and fossil
During the 1960s, production billet testing systems were fuel plants were a different breed of technician, working 011
produced at the rate of two and sometimes three per year, an unprecedented scale. The individuals were expert ill
handling larger and larger products at rates of two or more many nondestructive testing methods and they bore seriOllS
pieces per minute. Rotation of billets for the inspectors was responsibilities. The 2.2 million kilowatt plant built for the
provided by a Ferris wheel (Fig. 4.5) or by the chain sling Indiana and Michigan Electric Company near St. Joseph.
turner (Fig. 46). ObsClved indications were marked by the Michigan can serve as an example of their effOlis in the
1960s.
inspector (Fig. 47) for conditioning. Before the end of the
decade, the automatic indication detector was introduced The plant's construction began in 1969 and the unit went
(Fig. 48). on-line in 1976. Construction required on-site, around tIlt'
clock magnetic paliicle testing \vith crews operating out or a
"""'111111111111111'1111111111111111'111111'1"""'1"'1111111111111111111'111""""1""11
A HISTORY OF MAGNETIC PARTICLE TESTING / 71

FIGURE 42. Semiautomatic magnetic particle FIGURE 44. Magnetic particle testing system for
testing of large crankshafts for transverse and very large gas turbine shafts and spindles
longitudinal discontinuities

FIGURE 45. Magnetic particle billet testing using

the Ferris wheel

FIGURE 43. Three-way multidirectional

magnetizing of complex steel castings in one
operation

pen~allent testing laboratory. The twenty-four technicians

on SIte were ,til qualified and certified in magnetic particle,
penetran.t, ultrasonic and radiographic testing.
. Matenals were monitored at manufacture and during
InCorporation into the system at the job site. Nondestructive
72/ MAGNETIC PARTICLE TESTING

FIGURE 46. Magnetic particle billet testing using FIGURE 48. Automatic discontinuity indication
a chain sling turner detector for magnetic particle tests

testing was used to verify the integrity of valves, tltting:s

pipe joints, bolts, all types of welds and, most important, tIlt·
final structure·. The safety of people in and around the plallL
and the economy of the unit's operation, were both directl:
linked to the performance of the nondestructive testing
personnel on-site during constmction.
FIG~RE 17. Fluorescent magnetic particle..
indications of seams at various depths In a billet

The 19705: Advanced Organization

and Systems
During the 1970s, industries began recognizing the valli('
of Recommended Practice SNT-TC-1A introduced 11\
ASNT. The document began to be referenced in purchasing
contracts and the military began including its main featu res
in their own specifications. The era of Level I, II and [I r
magnetic particle inspectors spread across the conntr::
responsibility for celtification of oper~tors was in the hallds
of the firm making the product.

Advances in Systems and Components

In 1974, a magnetic particle slurry for discontintlil:
detection was introuuced. Unlike dly pO\vder or otlier
magnetic pmticle suspensions, the slurry was most eifedi\('
outside dUling rain and \vind. It also allowed maglletil'
pmticle testing under water. The material was first lIsed oil
pipeline welds in the field, on overhead stmctural welds and
for undenvater applications.
 A HISTORY OF MAGNETIC PARTICLE TESTING I 73

Later in the decade, special daylight fluorescent particles units. By 1975, wet horizontal units were using silicon
were introduced. They were designed to be dispensed thyristors with solid state power packs. The use of solid state
under water from plastic containers and were used with components allowed infinitely variable, fine current control
special electromagnetic yokes for offshore testing. between zero and maximum, replacing the tap switch's
In 1978, the first programmable component was intro- limited number of settings.
duced, replacing relay logic for controlling the sequencing By 1981, silicon diode rectifier stacks in commercial
of an automat ic magnetic particle system. The unit was direct current wet horizontal units were almost completely
designed for testing of 155 mm (6 in.) artillelY shells. In the replaced by silicon thyristors. This change allowed control-
same year, an in-line system was developed, using reversing ling the magnitude of magnetizing current while at the same
direct' current for the demagnetization of long tubular time converting alternating current to direct current. It also
product traveling at speeds of 20 to 30 m (80 to 100 ft) per prOvided the additional advantage of eliminating the need to
minute. reverse the polarity of the direct current while demagne-
An electromagnetic pOliable magnetization indicator was tizing.
introduced to verify reliable magnetization levels on test In the mid 1980s, a wet horizontal magnetic pUliicle
objects sllch as aircraft landing gears, complex castings and system was developed with microprocessor control. The
farge welded fal)l;cations. Unlike the earlier flux shunting system produced detailed test data and supplied hardcopy
devices that responded to a tangential magnetizing force for each test object, documenting such information as part
through the indicator, this instrument responded to changes number; type of magnetization used (head, coil or multidi-
in permeability detected by a cross coil probe near the rectional); type and amount of current; duration of current;
surface of the test object. Indicated magnetization levels concentration of bath at the time of processing; inspector's
correlated vel)' well with results on parts containing natural name; date of testing; and disposition of the part.
discontinuities.
In 1979, a newly developed process allowed fluorescent
magnetic pmticles to be manufactured directly to the
desired ri~llTOW size range. :Crinding of particles was no Conclusion
longer necessary. Inspection sensitivity could be controlled
•~y both t]~e type of pmiicles and the type and strength of the Magnetic particle testing has been in use since the 1920s.
applied magnetic field. The simplicity of its use is now understood to depend on
electromagnetic phenomena that are defined by complex
field theories. Yet the technique is still used as it was in the
The J 980s: Industry Improvements beginning - .to locate discontinuities in ferromagnetic
materials. Magnetic particle methods have been adapted to
Early in 1980 came the introduction of fluorescent include computerized data collection systems a'nd laser
magnetic particles that were three to five times brighter scanning modules - but magnetization, application of
than earlier products. The enhanced brightness made indi- particles, interpretation and demagnetization are still the
cations difficult to miss in dark testing booths. Out of the basic steps of the technique. Magnetic particle tests are
same research, daylight fluorescent pmiicles were devel- specified as complementary to more recent techniques
oped. They could be used wet or dly and were visible in (ultrasonic testing, for example). Yet magnetic particle
da~light, under blue-white mercUlY arc illumination, under inspection is still the only required test for many structural
whIte fluorescent lights, and also under bright white incan- and materials applications.
?escent light. They were not excited by yellow sodium vapor It is this combination of characteristics - a strong,
Illumination. original technique and an adaptability to contemporary
. As early (~s 1973, silicon thyristors were replacing silicon testing requirements - that has given magnetic particle
dIOde rectifier stacks in mobile magnetic pmticle testing testing its many successful applications.
 74 / MAGNETIC PARTICLE TESTING

REFERENCES
1. Saxby, S.M. Engineering. Vol. 5. London, England (1868): 5. Ibid: p 10.
p297.
2. Betz, Carl E. Principles of Jt/agnetic Particle Testing. 6. Ibid: p 14.
Chicago, IL: Magnaflux Corporation (1967).
3. Berthold, Rudolph. "Technical Aids of the Magnetic 7. Bogart, Henry G. "Cost Effectiveness in Nondestructive
Particle Method." Atlas of the Nondestructive Test Meth- Testing." lV1aterials Evaluation. Vol. 26, No.3. Columbus,
ods. Leipzig, Germany: Verlag von Johann Ambrosius OH: The American Society for Nondestructive Testing
Barth (1938): p 20/2. (March 1968).
4. Thomas, Willys E., et al. Magnaflux Corporation: A
History. Gary Plice and John McCall, eds. Chicago, IL: 8. Vaupel, O. Picture Atlas for Nondestructive Materials
Peabody International Corporation (1979): p 1. Testing. Ernst Hoeppner, ed. Berlin, Germany: Verlag
Bild und Forschung (19,5.5): p 721.
 SECTION 4
DISCONTINUITIES IN
FERROMAGNETIC ALLOYS
David R. Atkins, Packer Engineering, Naperville, Illinois
Michael A. Urzendowski, Babcock and Wilcox, Alliance, Ohio
Robert W. Warke, Packer Engineering, Naperville, Illinois
761 MAGNETIC PARTICLE TESTING

INTRODUCTION

Discontinuities can be defined as changes in the geome- In order to better detect and interpret magnetic particle
try or composition of an object. Such changes inherently discontinuity indications, it is necessary to know the basic
affect the physical properties of the object and may in tum material characteristics of the test object. Furthermore, it is
have an effect on the object's ability to fulfill its intended also important to consider how the material is produced,
service life. Not all discontinuities are defects. The defini- what manufacturing processes are used to form the finished
tion of defect changes with the type of component, its product and what discontinuities are typically initiated by
constmction, its materials and the specifications or codes in the processing operations.
force. It should be well understood that a harmless discon- During the various stages of material processing, certain
tinuity in one object may be a critical flaw in another kind of discontinuities can be expected. Typically, a discontinuity is
object. categorized by the stage of manufactUring or use in which it
Detection of discontinuities is a process that is largely initiates: inherent, primary processing, secondary process-
dependent on the discontinuity's physical characteristics - ing and service related discontinuities. The text that follows
in the case of cracks, a critical parameter is the ratio of is a description of discontinuities that may originate from
surface opening to crack depth. However, crack depth and the processing operations in each of the four categories. The
width are not the only factors affecting detectability; length listing is provided only for educational purposes and may
and orientation to the surface are also important. not apply to all magnetic particle test objects.
 DISCONTINUITIES IN FERROMAGNETIC ALLOYS I 77

PART 1
INHERENT DISCONTINUITIES
Ie
ic
is
:1,
When ferromagnetic materials are produced, molten appears as voids called pipe in the finished product. Pipe
d
metal solidifies i!lto ingot form, producing what is known as can also result from extrusion, caused by the oxidized
'y inherent discontinuities. :Many of these are removed by surface of a billet flOwing inward toward the center of a bar
cropping but a !lumher of them can remain in the ingot. at the back end. The presence of pipe is usually character-
n
Such discontinuities then can be rolled, forged and sec- ized as a small round cavity located in the center of an end
tioned along with the mateIial in its subsequent processing surface.
operations.
The follo\\-ing text is a brief description of common
inherent discontinuities that may occur in ferromagnetic Hot Tears
matelials (see Tahle 1).
At the elevated temperatures associated with solidifica-
tion, cast materials are susceptible to hot tears. Segregation
Cold Shut of low melting point impuritie~ results in localized loss of
ductility and strength. Lacking these, the cooling metal can
A cold shut is initiated dming 'the metal casting process. tear and crack in the mold because of restraint from the
It occurs because of imperfect fl.lsion between two streams mold. In addition, uneven cooling in thin sections or comers·
of metal that have .converged. Cold shuts may also be that adjoin heavier masses of metal can result in higher
attributed to surging, sluggish molten metal, an intermption metal surface stresses that in tum produce hot tears.
in pouring or any factor that prevents fusion where two Hot tears appear on the s~rface as a ragged line of
molten surfaces meet. variable width and numerous branches. In some instances
This discontinuity produces magnetic particle indications the cracks are not detectable until after machining because
similar to those of cracks or seams with smooth or rounded the tearing can be subsurface.
edges similar to those of Fig. 1.

Blowholes and Porosity

Pipe
Gas porosity or blowholes are rounded cavities (flattened,
DUring solidification of molten metal, a progressive re- elongated or spherical) caused by the accumulation of gas
duction in volume occurs. In the case of a casting, there bubbles in molten metal as it solidifies. A small percentage
e.v~ntually can be insufficient molten metal for completely of these bubbles rise through the molten metal and escape.
~llhng the top of the mold. As a result, a cavity forms, usually However, most are trapped at or near the surface of the
III the ~hape. of an invelted cone or cylinder (see Fig. 2). ingot when solidification is complete (see Fig. 2). During
I~ tIns shnnkage cavity is not completely removed before rolling or forging of the ingot, some of these gas pockets are
rollmg or forging into final shape, it becomes elongated and fused shut.

TABLE 1. Inherent discontinuities in ferromagnetic materials

Discontinuity Location Cause
Cold shut surface or subsurface the meeting of two streams of liquid metal that do not fuse together
Pipe subsurface an absence of molten metal during the final solidification process
Hot tears surface restraint from the core or mold during the cooling process
Porosity surface or subsurface entrapped gases during solidification of metal
Inclusions surface or subsurface contaminants introduced during the casting process
Segregation surface or subsurface localized differences in material composition
78 I MAGNETIC PARTICLE TESTING

The remaining pockets may appear as seams in the rolled elongated shapes and to appear in longitudinal sections as
product. Deep blowholes that are not rolled shut may stringers or streaks. In transverse cross sections, the inclu-
appear as laminations after becoming elongated in the sion's shape is more globular or flat (see Figs. 3 through 6).
rolling operation.

Segregation
Nonmetallic Inclusions
Segregation is localized differences in a material's chem-
Inclusions in ferrous alloys are usually oxides, sulfides or ical composition. During solidification of molten metal,
silicates introduced in the original ingot. During the melting
operation, the use of dirty remelt, crucibles or rods, or poor
linings may introduce nonmetallic inclusions into the mol- FIGURE 2. Longitudinal section of two types of
ten metal. Other contributing factors are poor pouring ingots showing typical pipe and porosity
practice and inadequate gating design that can produce
PIPE
turbulence within the mold.
POROSITY
Nonmetallic inclusions can become stress risers because
of their shape, discontinuous nature and incompatibility
with the surrounding material. In many applications, it is the
presence of these inclusions that lowers the ability of a metal
to withstand high impact, static or fatigue stresses. More-
over, the effect of inclusions depends on their size and
shape, their resistance to deformation, their orientation
relative to applied stress. and the tensile strength of the
material. Many inclusions can be of a more complex
intermediate composition than their host materials and each
grade and type of metal has its own characteristic inclusions.
Typically, inclusions are mechanically worked (from roll-
ing or forming), causing them to deform plastically into

FIGURE ,. Magnetic particle indication of a cold

shut in a casting .

BAR ROLLED FROM INGOT ABOVE

LEGEND
~INDICATES SECTION OF INGOTS USED
FOR ROLLING BARS BELOW

FROM MAGNA FLUX CORPORATION. REPRINTED WITH PERMISSION. FROM AMERICAN SOCIETY FOR METALS. REPRINTED WITH PERMISSION.
 DISCONTINUITIES IN FERROMAGNETIC ALLOYS I 79

iI:
celtain clements may concentrate limited areas, resulting
in an uneven distribution of some of the alloying elements of FIGURE 4. Cross section through raif sample,
the steel. Equalization of the compositional differences can showing magnetic particfe indications at center
be achieved by hot working (f(xging or rolling). However,
segregation is sometimes carJied into the wrought product.
When not detected, segregation can affect corrosion
resistance, f()rging and welding characteIistics, mechanical
properties, fracture toughness and fatigue resistance. Fur-
thermore, quench cracks, hardness variations and other
discontinuities are likely to result dUIing heat treating of
materials that exhibit segregation of alloying elements.

fiGURE 3. Inelusions present in wrought product

were elongated through the rOiling process and
discovered at a weld upset that jOined two rails
together; magnetic particle indications shown in
the web adjacent to the weld

FIGURE 5. Microstructure of transverse

(perpendicular to rOiling direction) section
through rail sample, away from weld
80 / MAGNETIC PARTICLE TESTING

FIGURE 6. Microstructure of longitudinal section

through rail sample; inclusion runs the length of
the specimen
 DISCONTINUITIES IN FERROMAGNETIC AllOYS I 81

PART 2
PRIMARY PROCESSING DISCONTINUITIES

Discontinuities that originate during hot or cold forming rolling process. They can be surface or subsurface, are
are said to be primary processing discontinuities. The generally flat and extremely thin (see Fig. 10).
processing of a w.nJught produc~ ~y f?lIing,. fOl:~ing,. casting Laminations can be detected by magnetic particle testing
or drawing may lIltroduce specIfIc chscontmmtIes mto the at an end or at a transverse cross section taken through the
product and inherent di~continuities that were at one time rolled plate.
undetectable or insignificant may propagate and become
detrimen tal.
The following is a hrief descliption of common pIimary Stringers
processing discontinuities that may occur in ferromagnetic
matelials (see Tahle 2). Shingers are predominantly found in bar stock. They
Oliginate by. the flattening and lengthening of nonmetallic
inclusions during the rolling operation.
Seams·
As an ingot is processed, inherent smface discontinuities FIGURE 7,· Formation of a seam: fa) underfill
such as gas pockets, blowholes and cracks are rolled and results when there is not enough metal to fill
drawn longitudinally. '''hen these discontinuities exist, an . the rolls;.and fb) a seam in the finished bar
underfill of mateJial occurs dUling the rolling operation. occurs when underfill is squeezed tight on a
Seams m:.-ty also he irritiated in the semifinishing and subsequent rolling pass
finishing mills because of faulty, poorly lubIicated or over-
6ized dies. fal

0:
As a result of m llltiple passes dUling rolling operations,
undelfilled areas are rolled together to form a seam (see
Fig. 7). The surfaces are typically oxidized and may be
intermittently welded together to form very tight, usually
straight cracks that vmy in depth from the sUlface (see
Figs. 8 and 9). '---. UNDERFILL

fbI
laminations
Laminations are separations that are t)pically aligned
parallel ~o the worked slllface of a matelial. They may be the
result of blowholes, internal fissures, pipe, inclusions, seams
or segregations that are elongated and flattened dUling the
o "'-SEAM

TABLE 2. Primary processing discontinuities in ferromagnetic materials

Discontinuity location Cause
Seams surface elongation of unfused surface discontinuities in rolled products
Laminations subsurface elongation and compression of inherent discontinuities during the rolling process
Stringers subsurface elongation and compression of inherent discontinuities during the rolling process
Cupping subsurface internal stresses during cold drawing
Cooling cracks surface uneven cooling of cold drawn products
Laps surface material folded over and compressed
Bursts surface or subsurface forming processes at excessive temperatures
Hydrogen flakes subsurface an abundance of hydrogen during the forming process
82 I MAGNETIC PARTICLE TESTING

FIGURE 8. Wet fluorescent magnetic particle FIGURE 10. Metallographlc cross section showing
indication of a seam in a steel billet laminations found In a resistance welded tube
made from SA-178 material

FROM MAGNA FLUX CORPORATION. REPRINTED WITH PERMISSION.

FIGURE 9. Photograph showing seams In bars,

from left: as-received condition, sand blasted
FIGURE .11. Cross section showing severe cupping
surface, pickled surface and wet fluorescent
in a 35 mm (1.4 in.) bar
magnetic particle indication

FROM MAGNAFLUX CORPORATION. REPRINTED WITH PERMISSION.

Cooling Cracks
FROM MAGNAFLUX CORPORATION. REPRINTED WITH PERMISSION.
After the rolling operation of cold drawn bars, woling
cracks may develop due to internal stresses caused hy
uneven cooling of the material. Such cracks are t)l)ically
Stringers are typically subsurface, semicontinuous straight
longitudinal and usually VaiY in depth and length. Although
lines parallel to the length of the bar stock.
often confused with seams, cooling cracks do not exhibit
surface oxidation.

Cupping
Typically occuning during the extrusion operation or as a Forged and Rolled Laps
result of severe cold drawing, cupping is a series of internal
mptures (chevrons) in bar or wire as shown in Fig. II. Forging laps are the result of metal being folded over,
Because the interior of a metal cannot flow as rapidly as the forming an area that is squeezed tight but not welded
surface, internal stresses build, causing transverse subsur- together (see Figs. 12 and 13). They are caused by faulty
face cupping cracks. dies, oversized blanks or improper handling of the metal in
 DISCONTINUITIES IN FERROMAGNETIC AllOYS / 83

FIGURE 12. Wet fluorescent magnetic particle FIGURE 14. Formation of a lap: faJ an overfill
indication of a forging lap in a connecting rod produces excess metal squeezed out of the rolls
causing a fin; fbJ a lap results when the
projection is folded over and forced back into
the bar's surface during a subsequent pass

fa)

o "FIN

o
fb)

FROM MAGNAFLUX CORPORATION. REPRINTED WITH PERMISSION.

"'-LAP

FIGURE 13. Micrograph of a forging lap with

included oxide in the lap . .
Internal and External Bursts
Internal bursts are found in bars and forgings and result
from excessive hot working temperatures. Discontinuities·
that exist plior to forming (porosity, pipe, inclusions or
segregation) are pulled apart because of the high tensile
stresses developed during the forming operation.
Rolled and forged metals may also develop internal bursts
when there is insufficient equipment capacity for working
the metal throughout its cross section (see Fig. 15) .
.External bursts typically occur when the forming section
is too severe or where sections are thin. External bursts may
also be formed when the capabilities of the equipment are
not great enough: the outer layers of the metal are de-
formed more than the internal metal and the resulting stress
FROM AMERICAN SOCIETY FOR METALS. REPRINTED WITH PERMISSION.
causes an external burst. Forming during improper temper-
atures may also cause extemal bursts.

the die. Forging laps are usually open to the surface and are
either parallel or at a small angle to the sUlface. Hydrogen Flakes
Rolled laps are a condition similar to a seam. Excessive
material is squeezed out dming a rolling pass, causing a Flakes are formed while cooling after the forging or
sharp overfill or fln. vVhen rotated for the follO\ving pass, the rolling operations. They are internal fissures attributed to
material is rolled back into the bar. Because of its heavily (1) stresses produced by localized metallurgical transforma-
oxidized surface, the overflll cannot be welded together by tions and (2) decreased solubility of hydrogen (hydrogen
the rolling operation. Rolling laps are usually straight or embrittlement) resulting from excessively rapid cooling.
slightly curved from the longitudinal axis and are either Hydrogen is available in abundance during all manufac-
parallel or at a small angle to the object surface (see turing operations. \Vhen permitted, hydrogen dissipates
Fig. 14). freely at temperatures above 200°C (390 OF), so that the
84 / MAGNETIC PARTICLE TESTING

solubility of hydrogen in material proportionally increases

with increasing time and temperature. Hydrogen flakes are Welding Discontinuities
usually found deep in heavy steel forgings, are extremely
thin and are aligned parallel with the grain. The following discontinuities are related primarily to the
fusion welding process, although a few may also apply to
resistance and solid state processes. This compilation covers
FIGURE 15. Cross section of a bar showing a only those discontinuities that I:Jpically lend themselves to
forging burst near the centerline; arrow detection by magnetic particle testing (see Table ,'3).
Indicates the direction of working Acceptance or rejection of a weldment, based on the
detection of a pmiicular discontinuity, is determined by the
requirements of the deSigner and the applicable code.

Cold Cracking
Cold cracking is also knovvn as underbead or delayed
cracking. It is a form of hydrogen induced cracking that
appears in the heat affected zone or weld metal of 19w alloy
and harden able carbon steels. Cracking of this type may
occur immediately on cooling or after a period of hours or
even days. The principal factors contributing to cold crack-
ing are (1) the presence of atomic hydrogen; (2) a hard
martensitic microstructure in the heat affected zone; and
(3) high residual tensile stresses resulting from restraint.
Sources of atomic hydrogen include moisture in the
electrode covering, shielding gas or base metal surface
(including hydrated rust), as well as contamination of the
filler or base meh~l by a hydrocarbon (oil or grease).
. Dissociati(m of water vapor or a hydrocarbon in the welding
arc results in the rapid diffusion of atomic hydrogen into the
molten weld pool and subsequently into the base metal's
heat affected zone. If the zone's cooling rate is high enough
and the steel is hardenable enough (a function of carbon
and alloy content), a mmiensitic microstructure may form
FROM AMERICAN SOCIETY FOR METALS. REPRINTED WITH PERMISSION.
and the hydrogen atoms may then collect at internal
discontinuities. Residual stresses caused by weld shrinkage,

TABLE 3. Primary processing discontinuities in welds

Discontinuity Location Cause

Cold cracking surface or subsurface a combination of atomic hydrogen, hardenable material and high residual stresses
Hot cracking
Solidi fication surface or subsurface dendritic segregation of low melting point constituents opening up during solidification
Liquation surface or subsurface HAl segregation of material in the liqUid state during solidification.
Lamellar tearing surface delamination of the base material during solidification and cooling of weld metal
Lack of fusion subsurface failure of the filler metal to coalesce with the base metal
Lack of penetration surface or subsurface inadequate penetration of the weld joint root by the weld metal
Porosity surface or subsurface vaporized constituents in the molten weld metal are entrapped during solidification
Inclusions
Slag subsurface improper cleaning of a previous weld pass
Tungsten subsurface molten weld pool or filler metal comes in contact with the tip of a tungsten electrode
Oxide subsurface mixing oxides on the base metal surface into the weld pool
Undercut surface oversized weld pool (related to excessive amperage. travel speed and electrode size)
Overlap surface insufficient amperage or travel speed
 DISCONTINUITIES IN FERROMAGNETIC ALLOYS / 85

or externally applied tensile stresses, result in hydrogen centerline of the deposited weld bead (see Fig. 19). DUling
induced cracks initiating at the hydrogen rich discontinuities. weld deposition, solidification of the progressing weld pool
Cold cracks produce sharply defined, heavy magnetic occurs from the outside in, beginning at both toes and
pmticle indications if ,they are open to the test object meeting at the center. The low melting point impurities are
sllrfilct\ as ill the case of underhead cracks that extend to the pushed ahead of these two joining solidification fronts
weld toe (Fig. 1(i). \Veld metal cracks may be oriented in where they are concentrated at the centerline and open up
'wv directioll ,mel are often associated with nonmetallic as a longitudinal hot crack under transverse solidification
;n~ltlsions (Fig. 17). Subsurface indications are less pro- shrinkage stresses. The likelihood of this occurrence is
nounced or may he undetectable, depending on depth. increased by high travel speed, high depth-to-width ratio of
the weld bead and a small \veld bead, particularly in the root
Hot Cracking pass.
Hot cracking is a term applied to several varieties of weld Another frequently observed type of solidification crack-
metal amI heat affected zone cracking, all of which occur at ing is called crater cracking, which occurs in the crater
elevated temperatures. The follo\ving t,ypes are two of the formed at the termination of a weld pass (Fig. 20). Crater
most common hot cracks. cracks are typically star-shaped on the surface and are the
Solidification cracking occurs near the solidification tem- result of three-dimensional shrinkage stresses brought about
perature of the \veld metal and is caused by the presence of by crater solidification. Sudden termination of the welding
low melting pOint constituents, typically iron sulfides, that are, rather than pausing at the end of a weld pass to fill in
segregate to the weld metal dendrite smfacesdming the the crater, is a common contributor to crater cracking.
liquid-to-solid transformation process. The shrinkage stress- Liquation cracking or hot tearing occurs in the heat
es induced by cooling cause cracks to open between the affected zone of a weld when the temperature in that region
dendrite surfaces (Fig. 18). results in the liquation of low melting point constituents
One common form of solidification cracking is called (inclusions or segregated allOying elements). These form a
centerline hot cracking, because it follows the longitudinal liquid grain boundary film that is unable to SUppOlt the

FIGURE 16. Cross section of a weld joint . FIGURE 17. Cross section of a weld joint
exhibiting hydrogen induced cold cracking in the exhibiting hydrogen induced weld metal cold
heat affected zone (underbead): this crack is cracking; this crack is oriented longitudinally but
detectable by magnetic particle testing because weld metal cracks may be oriented in other
it extends to the outside surface directions depending on joint restraint

FROM THE BRITISH WELDING INSTITUTE. REPRINTED WITH PERMISSION. FROM THE BRITISH WELDING INSTITUTE. REPRINTED WITH PERMISSION.
86 / MAGNETIC PARTICLE TESTING

FIGURE 18. Cross section of a weld joint FIGURE 19. Section through a weld joint
exhibiting solidification cracking; the weld on exhibiting centerline solidification cracking, a
the right side contains an interdendritic crack form of hot cracking
associated with a slag inclusion which acted as a
nucleation site; note that the crack curves as it
approaches the weld centerline, following the
dendritic solidification pattern

FROM THE BRITISH WELDING INSTITUTE. REPRINTED WITH PERMISSION.

FIGURE 20. location and typical appearance of' .

crater cracks

CRATER CRACKS
shrinkage stresses of the welding process. Such cracks are
often microscopic in size but may link up under applied
stresses to form a continuous surf~1Ce or subsurface crack.
In general, hot cracking is associated with steels having
high sulfur content and the effect is accentuated as carbon
content increases. The detectability of hot cracks by mag-
netic palticle methods is similar to that of cold cracks and
FROM THE AMERICAN WELDING SOCIETY. REPRINTED WITH
depends on their proximity to the surface. PERMISSION.

Lamellar Tearing
A lamellar tear is a base metal crack that occurs in plates and ductility in the through-thickness direction. 'When the
and shapes of rolled steel exhibiting a high nonmetallic shIinkage stresses induced by weld solidification are im-.
inclusion content. These inclusions are rolled flat in the posed in that direction on the hase metal plate, separation 01
steel plate manufacturing process, severely redUCing strength the base metal at the flattened inclusions might occur, as
 DISCONTINUITIES IN FERROMAGNETIC ALLOYS / 87

may sheming hetween those lamellar planes, resulting in a

terraced fracture (Fig. 21). One welding process that is palticularly susceptible to
this discontinuity is gas metal arc welding (G MAW) in the
Lamellar tearing is readily detectable by magnetic parti-
short-circuiting arc mode, because of its inherently low heat
cle techniques and is most often seen in base metal on the
input. Another frequent" cause of lack of fusion is attempting
edge of a skel plate or structural shape, adjacent to a
deposited \veld head (Fig. 22). to weld on top of a previously deposited weld pass that has
been inadequately cleaned of slag or welding on a dirty base
Lack of Fusion
Lack of fusion occurs when some portion of the weld FIGURE 22. Weld jOint designs in steel plate that
filler metal fails to coalesce with the adjacent base metal or are prone to lamellar tearing; typical locations
the weld metal from a previous pass. In welding processes of tears are shown
that use no filler metal, lack of fusion refers to incomplete
coalescence hetween the hvo base metal components being
jOined.
This conditiull is caused when the base metal smface fails
to reach melting temperature after application of the weld
metal. This t}JJically occurs when welding a large compo-
nent that call transfer heat rapidly because of its thermal
mass, pmticularly when it is at a relatively low temperature
prior to welding, thereby absorbing the heat applied to its
surface by the welding process. Lack of fusion is often seen
at the beginning of the first weld pass, where the base metal.
is at its lowest temperature during weld deposition; this is
commonly calkd a cold start (s.e~ Fig. 23) ..

FIGURE 21. Typical location and appearance of

lamellar tearing; this view is parallel to the
rOiling direction of the steel plate base metal

FROM THE BRITISH WELDING INSTITUTE. REPRINTED WITH PERMISSION.

FROM THE AMERICAN WELDING SOCIETY. REPRINTED WITH PERMISSION.
88 I MAGNETIC PARTICLE TESTING

metal surface, so that the heat of the arc is unable to reach usually oriented parallel to the direction of welding and the
the underlying metal. test indication often appears at or near the toe of the weld.
Lack of fusion mayor may not occur near the outside Lack of fusion in autogenous welds (without filler metal)
surface of the weld joint. The closer it is to the surf~lce, the may be the result of large inclusions in the base metal or
sharper the magnetic particle indication. Lack of fusion is impurities that become trapped between the faying surfaces
of the joint prior to welding. Susceptible processes are those
that produce a relatively shallow melted zone at the faying
FIGURE 23. Cross section of a weld joint surfaces and then expel most of that zone by a subsequent
exhibiting lack of fusion resulting from a cold upsetting force (high frequency resistance welding, projec-
start in a submerged arc weld tion welding, flash welding, friction welding). Other causes
of lack of fusion in autogenous welds include inadequate
heating and insufficient upsetting force. Figures 24 and 2,5
show a typical discontinuity of this type.

lack of Penetration
Lack of penetration is sometimes confused with lack of
fusion. Lack of penetration is inadequate (less than speci-
fied) penetration of the weld joint root by the weld metal
(Fig. 26). The condition can result from a number of incor-
rect parameters, most of them related to welding technique.
These include low amperage, use of an oversized electrode,
excessive travel speed, improper electrode angle, improper
arc manipulation and inadequate preweld cleaning.
Often, the joint design does not facilitate good penetra-
tion because of too large a root land, too narrow a root gap
or too small a bevel angle. Ma~y procedures for 'double
groove welds specify backgouging of the first p,ass on the
first side ptior to deposition of the first pass on the second
side. If this is neglected. or performeq inadequatcly duting '..
the joining operatioh, lack of penetration will likely occur.
The magnetic particle indication produced by lack of pen-
etration has an appearance similar to a subsurface longitu-
dinal crack and usually follows the centerline of the weld.
Porosity
Porosity is composed of cavities or pores that form when
some constituent within the molten weld metal vapotizes
and forms a small pocket of gas that is entrapped when the
weld metal solidifies. The pores can take a vaIiety of shapes
and sizes although they are usually spherical. One type of
FIGURE 24. Magnetic particle indication of lack
of fusion in a high frequency resistance welded FIGURE 25. Metal/agraphic cross section of the
tube tube shown in Figure 24, showing the depth of
lack of fusion from the outside surface inward
 DISCONTINUITIES IN FERROMAGNETIC ALLOYS I 89

elongated pore is often called a wormhole (Fig. 27). The Inclusions

distribution of porosity within the weld metal may be
clustered (llsually results [rom improper initiation or termi- Many weld processes use flux shielding, including shield-
nation of the welding arc) or linear (indicates gas evolution ed metal arc welding (SMAW), submerged arc welding
by welding over a contaminant confined to a linear junction (SAW) and flux cored arc welding (FCAW). Welds pro-
such as a COllIer or crevice). d~ced .by. t.hese methods are patticularly susceptible to
In general, porosity is often the result of dilt, rust or diSCOntI~U1tIes known as slag inclu.sions. Slag can be en-
moisture on the hase or fIller metal surface before welding trapped m the weld metal during solidification if it is unable
and can be prevented by maintaining cleanliness and dIY- to float out while the pool is still liquid. The factors· that
ness. Other contributing factors include base metal compo- p~·omote slag entrapment include high solidification rate,
sition (such as high sulfur content), high solidification rate hIgh weld pool viscosity, lIse of an oversized electrode and
and improper wC'ldillg technique (such as excessive arc. improper joint geometIy.
length). Sla~ allowed to remain on the surface of a deposited weld
Often the surf~lCe discontinuities called blowholes are bead IS rarely completely dissolved by subsequent passes.
found where gas pockets have reached the surface of the TI~erefor~, it is essential to remove all slag from each pass.
weld pool hut do not fully escape before solidifIcation takes Jomt deSIgns that exhibit a high depth-to-width ratio and
place. Blowholes should be removed before any subsequent weld beads with an excessively convex profile are promoters.
weld passes are deposited because they are likely places for of slag entrapment (Fig. 28). A magnetic particle indication
slag entrapment. p~·oduced by a slag inclusion is weak and poorly defined and
~ magnetic pmticle indication of subsurface porosity is lugh magnetizing field strength is required for detection.
typIcally weak and not clearly defined. All but the smallest . Tungsten inclusions are found in the weld metal depos-
. surface pores should be visible to the unaided eye. Ited by the gas tungsten arc welding (GTAW) process and·
~re usually the result of allOwing the molten weld pool or the
.fIller metal t~. come in contact With the tip of the tungsten
a
FIGUR~ 26. Cross section of weld joint electrode. Tlus type of inclusion is virtual1y undeteCtable by
magnetic patticle methods.
exhibiting lack of penetration

FIGURE 27. Longitudinal section through weld

metal containing wormhole porosity

~~~~,is75:'MERICAN SOCIETY FOR METALS. REPRINTED WITH

90 / MAGNETIC PARTICLE TESTING

Oxide inclusions are particles of high melting point oxides Undercut is caused by an oversized molten weld pool, which
present on the base metal surface. During welding, these is in turn related to excessive amperage, travel speed and
oxides are then mixed into the weld pool. The magnetic electrode diameter.
particle indications produced by oxide inclusions of signifi- A magnetic particle indication produced by undercut
cant size and quantity are similar to those produced by appears less pronounced than that produced by lack of
subsurface porosity. Small and isolated oxides are extremely fusion. Undercut is easily detected by visual examination.
difficult to detect by magnetic particle methods.
Overlap
Undercut
Overlap is the protmsion of weld metal over the weld toe,
Undercut occurs at the toe of a weld when the base metal producing a form of lack of fusion that creates a sharp
thickness is reduced. Essentiallv, a narrow crevice is formed mechanical notch or stress concentration (see Fig. 29b).
in the base metal, paralleling the weld toe and immediately The condition is caused by insufficient amperage or travel
adjacent to it (Fig. 29a). Undercut lessens joint strength in speed.
the static sense by reducing the base metal section thick- Overlap produces a magnetic particle indication at the
ness. It also creates a stress concentration that reduces the weld toe similar to that produced by lack of fusion. It is
impact, fatigue and low temperature properties of the joint. often detectable by visual examination.

FIGURE 2B. Cross section of a weld joint FIGURE 29. Diagram of weld discontinuities:
containing slag inclusions; note that the high (a) undercut (at arrow); and (b) overlap
depth-to-width ratio of the weld on the left side ('at arrow)
contributed to slag entrapment
fa)

fb)
 DISCONTINUITIES IN FERROMAGNETIC ALLOYS I 91

PART 3
SECONDARY PROCESSING
DISCONTINUITIES

Discolltinuities that originate from grinding, heat treat- However, in some cases, this process produces stresses that
ing, machining, plating and related finishing operations are exceed the matetial's tensile strength and cause it to crack
categorized as sccond,uy processing discontinuities. Such (Fig. :31). Similarly, when an object is heated to a vCly high
discontilluities may be the most costIy because all previous temperature then rapidly cooled (in air,· oil or water),
processing costs are lost when the component is diverted quench cracks may develop.
from senice. DUling the transformation from austenite (a face cen-
The f{)I1owing text bliefly desclibes the most common tered cubic stmcture) to ferrite (body centered cubic) and
secondmy processing discontinuities (see Table 4). martensite (body centered tetragonal) on cooling, a volu-
metric expansion occurs. "
When an object is quenched following heat treating, the
Grinding Cracks initial transformation occurs at the object's surface. Imme-
diately after the quenching process begins, a layer of body
GJinding cracks can be attributed to the use of giazed centered tetragonal or body centered cubic material is
wheels, inadequate coolant, excessive feed rate or attempt- formed at the smface. V\Then the interior cools and trans-
ing to remove too much material in one pass. Clipding forms, volumetric expansion takes place but the interior
Cl~lCks. develop where there is localized overheating of the expansion is restrained by the solidified lawr. If the !;olid
base material. They are typically at light angles to the layer does not expand enough or if the internal expansion is
gJinding direction and are velY shallow. Often, griQ.ding great enough, cracking through the outer layer results. .
cr~lCks are forked and sharp at the root (Fig. 30). A tempering process normally follows the quenching
"When located in high stress areas, such cracks may result operation. Beeause of this exposure to a high temperature,
in fatigue failures. Matetials that have been hardened or the surface of quench cracks become oxidized. IdentifYing
heat treated can be more susceptible to gtinding cracks oxidation is one method of determining if a crack was
because high residual stresses are retained duting the caused by queilching.
quenching operation. Quench cracks serve as stress concentration sites for
fatigue crack initiation and propagation. This may also selve
as the initiation sites for overload failures. Some quenching
Heat Treating and Quench Cracks operations are so severe that objects break up during the
process.
To obtain a specific hardness and microstmcture, mate- The amount of volumetric expansion is governed primar-
rials are cllstomatily heat treated. DUling this operation, the ily by the chemishy of the metal, pmticularly carbon. As the
metal is heated and cooled under controlled conditions. carbon content increases, so does the amount of expansion.

TABLE 4. Secondary processing discontinuities in ferromagnetic materials

Discontinuity Location Cause

Grinding cracks surface localized overheating of the material due to improper grinding proc"edures
Heat treating cracks surface uneven heating or cooling that produces stresses exceeding the tensile strength of the material
Quench cracks surface sudden cooling from elevated temperatures
Pickling cracks surface residual stresses being relieved
MaChining tears surface improper machining practices
Plating cracks surface residual stresses being relieved
.92 I MAGNETIC PARTICLE TESTING

The severity of the quench can be lessened by using a

lower carbon content material or by quenching in a less Machining Tears
harsh media such as oil or an elevated temperature bath.
Heat treating and quench cracks usually emanate from A dull machining tool shears metal off in a manner that
locations of thin cross section, corners, fillets, notches or produces rough, torn smfaces. As a result, the surface is
material thickness changes because these areas cool quicker work hardened to a degree that depends largely on the
and therefore transform first. Reshicted movement of the depth of cut, the type and shape of the tool and the material
material also influences the location of cracks during the properties (Fig. 32).
heat treating or quenching operations. Heat treating or Heavy cuts and residual tool marks from rough machining
quench cracks are typically forked, surface indications that act as stress risers and can contribute to premature failure in
are randomly placed in any direction on the test object. a component. Stress risers may also occur at a change in
section, such as in small fillet radii behveen hvo shaft
sections of different diameters or the poor blending of fillets
with shaft surfaces. Although difficult to detect, machining
Pickling Cracks tears must be thoroughly and meticulously located.

A pickling operation is used to remove unwanted scale for

the purpose of a more thorough test of the base material. It
can also be used to prepare the smface for finishing Plating Cracks
operations such as plating. Pickling cracks are predominate-
ly found in materials that have high residual stresses Plating is used for decoration, corrosion protection, wear
resistance and to correct undersized dimensions for a \vide
(hardened or cold worked metals) and in materials with
voids or similar discontinuities. variety of steel components. However, specific plating
During pickling, hydrogen is generated at the surface of·
the material. The diffusion of hydrogen into the metal
causes a breakdown of the molecular structure and a FIGURE 31. Magnetic particle indications of
subsequent propagation of cracks. When high internal quench cracks
stresses are present with preexisting cracks or other
discontinuities, hydrogen accelerates propagation of the
crack to relieve the stresses in the'material.'

FIGURE 30. Wet fluorescent magnetic particle

indication of grinding cracks in diesel engine
connecting pin

FROM MAGNA FLUX CORPORATION. REPRINTED WITH PERMISSION. FROM MAGNAFLUX CORPORATION. REPRINTED WITH PERMISSION.
 DISCONTINUITIES IN FERROMAGNETIC ALLOYS I 93

nlateJials produce residual stresses that can be either tensile (Fig. 33). This action produces crack propagation or initia-
or compressive. Plating mateIials that develop residual tion. Materials high in hardness or residual stresses are
tensile stresses (chromium, copper and nickel) can reduce more susceptible to damage from hydrogen absorption
the fatigue strength of a component. during plating or pickling operations. Fmthermore, cracks
Plating cracks may develop when there is penetration of that initiate exclusively in the plating mateIial may act as
either hydrogen or hot plating mateIial into the base metal stress Iisers and cause cracking in the base material.

FIGURE 32. Magnetic particle indications of FIGURE 33. Wet fluorescent magnetic particle
cracks resulting from cold working during indications of plating cracks due to hydrogen
machining embrittlement

FROM MAGNA FLUX C<?RPORATION. REPRINTED WITH PERMISSION.

FROM MAGNA FLUX CORPORATION. REPRINTED WITH PERMISSION.

94 / MAGNETIC PARTICLE TESTING

PART 4
SERVICE INDUCED DISCONTINUITIES

The life expectancy of a component is dependent on its Fatigue Crack Structure

service environment (both mechanical and chemical), the
quality of its maintenance and the appropdateness of its From an external surface, a fatigue crack resembles any
deSign. It is essential for testing personnel to know the other crack, but internally a fatigue crack has certain unique
service conditions of a component in order to accurately characteristics. Macroscopically, features called beach nwrks
perform a magnetic particle test. Although service induced or clarashell marks can be found. These distinct markings
discontinuities appear similar, the mechanisms that cause are the result of vaIiations in cycliC loading, either in
them are quite different in each case. frequency, environment or stress. Such marks are actually
The following text bdefly descdbes common service small ridges that develop on the fracture surface and they
induced discontinuities (see Table 5) found in ferromagnetic indicate the position of the advancing crack at a given time.
materials. The geometry and orientation of beach marks can help
establish the location of the crack origin and the direction of
propagation (Fig. 36).
Microscopically, the fatigue fracture mechanism is char-
Fatigue Cracking acterized bv features known as striations. Each striation
represents ~ne applied stress cycle. The distance between
Fatigue is a fracture mechanism induced by a cyclically striations can be equated to the crack growth rate.
applied stress that is lower in magnitude than the ultimate Striations a~d beach marks ai'e not always observed on
tensile strength of the material but high enough to initiate a the fracture surface. Many times, loading is such that
crack or to propagate a preexisting crack. striations formed during the tensile or positive stress cycle
Fatigue cracks can develop from stress risers such as are o.bliterated during compressive or negative stress. Stri-'
machining or t.oolil!g marks, nonmetallic inclusions present ations appear more often in softer materials such as alumi-
at or near the material surface, pores, holes or notches, num or low carbon steel.
keyways and may even develop on a smooth surface (see Fatigue cracks normally originate on the surface but can
Figs. 34 and 35). begin below the surface at discontinuities if the applied and
As a fatigue crack begins to propagate, the stress intensity residual stresses exceed the subsurface fatigue strength of
at the tip of the crack starts to increase. \Vith every the material. When this occurs, a circular pattern of beach
incremental growth period of the crack, there is a propor- marks may form around the origin, producing a bull's-eye
tional, incremental increase in the stress intensity. This appearance.
process continues until the stress intensity K reaches the The probability of fatigue cracking can be dramatically
critical value K[c where failure occurs. reduced if the designer is aware of the material's fatigue
This K[c factor, also known as the fracture toughness, is properties and deSigns the component accordingly. Proper
unique for each material. The variance in fracture toughness care in machining is necessmy to ensure tl1<tt no unantici-
partially explains the behavior of fatigue cracks: why there is pated stress risers are introduced. Additional fatigue resis-
such a range of fabgue crack sizes; why some cracks may tance can be gained by stress-relieving a component or hy
only propagate a small amount; and why others propagate shot peening to introduce a compressive stress on the
through nearly all the material before final fracture. object's surface.

TABLE 5. Service induced discontinuities in ferromagnetic materials

Discontinuity Location Cause

Fatigue surface cyclically applied stress below the ultimate tensile strength
Creep surface or subsurface material subjected to elevated temperatures and stress below the yield strength
Stress corrosion cracking surface combined efFects of a static tensile load and a corrosive environment
Hydrogen cracking surface or subsurface combined effects of applied tensile or residual stress and hydrogen enriched
environment
 DISCONTINUITIES IN FERROMAGNETIC ALLOYS I 95

FIGURE 34. Helicopter rotor component: FIGURE 35. Magnetic particle indications of
la) no discontinuities revealed by visual fatigue cracks in a gear
examination; and fb) fatigue cracks revealed by
wet fluorescent magnetic particle tests

la)

.s
~s
n
y
y

fbJ
FROM CIRCLE CHEMICAL COMPANY. REPRINTED WITH PERMISSION.

FIGURE 36. Photograph of a.fracture surface

·ty.pical of fatigue; note initiation area in the
upper left corner

Creep Cracking .
At temperatures greater than half the melting point and
at stresses below the yield strength of the material, defor-
~ation can occur by the action of grains gradually separat-
mg Over an extended period of time. This can eventually
le~d to cracking and finally to failure. This deformation or
faIlure mechanism is called creep.
Figure 37 shows a schematic representation of creep or The third pOltion is called secondary or steady state
deformation \vith a constant load. The curve can be broken creep. This peIiod accounts for the majoIity of a compo-
down into four regions. The Hrst is the material's initial nent's life and the rate of creep is nearly constant. During
response to loading. This is usually elastic in nature and is this stage, small voids begin to form and grow at the triple
applied very quickly, accounting for the vertical portion of points of the grain boundaries. Because the void formation
the curve. The next pOltion of the curve is where the is nearly constant, the creep rate can be predicted and the
ma~e~al' s rate of straining or creep is decreasing with time. remaining service life of the component can be estimated,
ThIs IS called prinwry or transient creep. based on the steady state creep.
 96 I MAGNETIC PARTICLE TESTING

Once the material moves into the region of tertiary creep, Nickel additions are beneficial if sufficient quantity is
the useful life of the material is over. In the tertimy stage, added to produce an austenitic structure that is mon-
the creep voids have become so large that they begin to link, resistant to creep. Austenitic stainless steels (pmiiculnrl:
forming a crack network that quickly leads to failure. 18Cr-8Ni tJl)es) have much better creep properties thall
Creep can be detected and controlled. Periodic tests, carbon steels.
paliicularly those involving field metallography and circum- Aside from alloying additions, heat treatment has <Ill
ferential measurement can be used to monitor the creep effect on creep properties. Heat treatment generally COll-
process (Figs. 38 and 39). By slightly decreasing operating troIs grain size and it has been found that a coarser grain at
temperature or stress, a substantial decrease in the creep elevated temperatures has higher creep strength than a
rate yields greater service life. Figure 40 shows the effect finer grain.
that various temperatures have on creep. Since materials can be subjected to such a variety of loads
It is generally recognized, that the most direct way to and temperatures for a pmiicular application, the t)Ve of
improve the creep properties of a metal is by adding alloying
elements. Carbide forming elements, such as molybdenum,
tungsten and to a lesser degree, chromium and vanadium, FIGURE 39. Photomicrograph of linked creep
effectively enhance the creep resistance of steels. voids in weld zone

FIGURE 37. Typical curve showing the three

stages of creep: (a) constant load test; and
Ib) constant stress test
PRIMARY CREEP SECONDARY CREEP

TIME
t

FIGURE 40. Curve showing the effect of

temperature on creep over time
FIGURE 38. Photomicrograph of fracture and
620 <os 605 "c
creep in various stages in heat affected zone (1.150 F) (1,120 'F)
near fusion zone interface
z 565 'c
<
~
(1,050 <OF)
Vl
Cl..
llJ 540 C
llJ (1,010 'F)
0:::
U

520 C

~~~==============:.
TIME
(970 F)

""""111111'1"""""""""""""""""""""""""""""""",,""""""'11
 DISCONTINUITIES IN FERROMAGNETIC AllOYS I 97

heat treatment should be based on the degree of stability exposed to ammonia (NH.3 ); and mild steel exposed to
that it imparts to the component initially and throughout its sodium hydrOXide (NaOH).
selvice life. The stress intensity versus time dependence for a typical
stress corrosion cracking situation is shown in Fig. 41. The
basic stress-time curve can be expressed in terms of the
initial level of K r, which is based on the tensile load and a
Stress Corrosion Cracking known crack length. The threshold value of stress intensity
is designated K rscc . Crack growth does not occur if the
Stress corrosion cracking is a fracture mechanism that stress intensity is below this value. If the initial stress
results from the combined effects of a static tensile load and intensity is above K[scc, a crack propagates. The higher the
a corrosive environment. The stress involved can either be initial K[ or the closer the value gets to the Clitical stress
from actual applicd loads or from residual stresses. One of intensity factor K IC , the shorter the life of the component.
the most common causes of this residual stress is the The initiation site of a stress corrosion crack may be a
shrinkage that occurs dming .cooling of weld metal. preexisting discontinuity or it may be a small pit acting as a
The corrosive emironment varies from material to mate- stress riser and produced by corrosive attack on the surface
rial. Some common examples of materials and their corro- (Fig. 42). Once a crack is formed, the corrosive environ-
sive environments include: aluminum and austenitic stain- ment penetrates the surface of the material. The tip of an
less steels exposed to saltwater; copper and its alloys advancing crack has a small radius and the attendant stress

FIGURE 41. Stress intensity versus time FIGURE 43. Photomicrograph showing a typical
dependence for a typical stress corrosion stress corrosion crack; note small pit produced by
cracking situation . corrosive attack acting 'as a stress riser
K,c

10
TIME
(minutes)

FIGURE 42. Stress corrosion cracking found in a

stainless steel tube (dye penetrant was used to
Illustrate the random crack orientation and
branching); note that similar cracking could exist
In ferromagnetic alloys
98 / MAGNETIC PARTICLE TESTING

concentration is great. This stress at the crack tip mptures

the nonnally protective corrosion film and aids in the FIGUR~ 44. Photograph of hydrogen cracking
corrosion process (Fig. 43). found In the heat affected zone adjacent to a
In addition to this, the formation of corrosion products by weld
local attack in confined areas produces high stress levels in
materials if the corrosion products occupy a larger volume
than the metal from which they are formed. This wedging
action of corrosion products in cracks has been measured to
produce stresses over 34 MPa (5 ksi) which aid in the
propagation of the crack.
Stress corrosion cracking produces brittle failure, either.
intergranular or transgranular, depending on the type of
alloy or the corrosive environment. In mo~t cases, while fine
cracks penetrate into a the cross section of a component, the
surface shows little evidence of corrosion.
In order to keep the stress intensity to a minimum, care
must be taken to avoid stress concentrations, such as tooling
marks, notches, arc shikes and large inclusions near the
surface.

Hydrogen Cracking If a crack is already present, it is quite common to see

~lydrogen induced cracking initiated at the tips of preexist-
Hydrogen cracking or hydrogen embrittlement is a ftac- mg cracks.
ture mechanism that results from the corrosive environment . In many instance~, l~ydrogen is. already present internally
produced by a hydrogen media and usually occurs in m a. metal befor~ It IS- placed mto service. Hydrogen is
covjunction ~th. an appI.~e.d tensile stress or residual stress. readIly absorbed mto molten metal during the initial solicl-
Hydrogen is introduced.into a material by processes such as ificati?~ of the ~l:ilterial and during welding processes. The
electroplating, pickling, welding in a moist atmosphere or solu~Ihty of hydrogen is quite high at elevated temperatures
the melting process itself. Hydrogen may also come from and m some cases, metals can become supersaturated with
corrosion or the presence of hydrogen sulfides, hydrogen hydrogen during cooling.
gas, water, methane or ammonia. Hydrogen cracking follows grain boundaries and rarely
If no crack or stress riser is present on a material surface, shows any signs ?fbl:anching (Fig. 44). vVhen such cracking
hydrogen can diffuse into the metal and often initiates re~u.lts from bhstenng or fro~ a static load, it always
cracks at subsurface sites, where triaxial stress conditions ongmates below the component s sUlface. Hydroaen crack-
are at maximum levels. In low strength alloys, this condition ing from other causes can begin below the objec~' s snr[lCe
can lead to what is known as hydrogen blistering. or at a stress riser.
 DISCONTINUITIES IN FERROMAGNETIC ALLOYS I 99

BIBLIOGRAPHY

1. "Nondestructive Inspection and Quality Control." Met- 6. "Metals and Their Weldability." Welding Handbook,
als Handbook, eighth edition. Vol. 11. Metals Park, OH: seventh edition. Vol. 4. Miami, FL: AmeIican Welding
Amelican Society for Metals (1976): p 56-58, 287-379. Society (1982).
2. "Failure Analysis and Pre~ention." Metals Handbook, 7. "Welding Technology." Welding Handbook, eighth edi-
eighth edition. Vol. 10. Metals Park, OH: American tion. Vol. 1. Miami, FL: American Welding Society
Society for Metals (1964): p 111-116,291-292,307-327. (1987).
3. "Failure Analysis and Prevention." Metals Handbook, 8. Betz, C. Principles of Magnetic Particle Testing. Chica-
ninth edition. Vol. 11. Metals Park, OH: American go, IL: Magnaflux Corporation (1967): p 70-113.
Society for Metals (1986): p 120-127~ 245-248, 309,..338. 9. Magnetic Particle Testing. New York, NY: General
Dynamics (1967): p 6-3, 6-13, 7-6, and 7-61.
4. "Welding, Brazing, and Soldering." Metals Handbook, 10. McGannon, H. The Making, Shaping and Treating of
ninth edition. Vol. 6. Metals Park, OH: American Soci- Steel, ninth edition: p 806-807, 850.
ety for Metals (1983). 11. "Quality Control and Assembly." Tool and Manufactu.r-
5. "Fundamentals of Welding." \Velding Handbook, seV7 ing Engineers Handbook, fourth edition. Dearborn, MI:
enth edition. Vol. 1. Miami, FL: American Welding Society of Manufacturing Engineers (1987): p 2!l-26, .
Society (l.981). 33-36. .
 SECTION 5
BASIC ELECTROMAGNETISM
Nathan Ida, University of Akron, Akron, Ohio

t
 102 I MAGNETIC PARTICLE TESTING

INTRODUCTION
Table 1 lists the symbols used in this section and their Tesia, for example, is a unit that replaces gauss for the
corresponding units of measure from the International measurement of magnetic fields. In vacuum (or for practical
System of Units (SI). Table 2 lists important physical con- purposes in air), a magnetic flux density (B) of 1 T corres-
stants and their values in SI units. Table 3 provides factors ponds to a magnetic field strength (H) of 107/47T A·rn- I .
for conversions to SI units.

TABLE 1. Symbols and units used in electromagnetics

Common
Symbol Quantity 51 Unit Abbreviation

A magnetic vector potential weber per meter Wb-m- I

8 magnetic flux density tesla T
weber per meter squared Wb-m-2
o electric flux density coulomb per meter squared C·m-2
E electric field intensity volt per meter Vom- I
F force newton N
f frequency cycles per second Hz
H magnetic field strength ampere per meter A·m- I
I current ampere A
J current density ampere per meter squared Aom-2
L inductance henry H
m,M magnetic moment ampere per meter Aom- I
V voltage volt V
v velocity meter per second m·s- I
W work joule J
w energy density joule per meter cubed Jom-3
q . • electric charge coulomb C
S area meter squared m2
C length meter m
ds differential area (vector) meter squareci m2
dl differential length (vector) meter m
R· position vector meter m
X, y, z, r, n, ;p, e, Q, R unit vectors meter m
¥m electromotive force ampere-turns Aot
n normal (component of vector)
Js surface current density ampere per meter squared Aom- 2
t time second s
P power watt per second W·s- I
Pd dissipated power watt per second Wos-I
15 Poynting vector watt per meter cubed Wom- 3
We electric stored energy joule J
Wm magnetic stored energy joule J
{) skin depth meter m
E, Eo, Er permittivity farad per meter Fom- I
/L, /La, /Lr permeability henry per meter Hom-I
(J' conductivity siemens per meter Som- I
W angular frequency radian per second rad-s- I
P charge density coulomb per meter cubed Com-3
Ax, A y ' A z' Ao- A cP , A,. AR components of a vector
A flux linkage
X magnetic susceptibility
T tangential component (vector)
Ps surface charge density coulomb per meter squared
 BASIC ELECTROMAGNETISM I 103

Although they are dimensionless quan~~ties, magnetic sus- TABLE 2. Physical constants used in
ceptibility an~l demagnetizing factors differ by a factor of 47T electromagnetism
in the GaUSSIaJl and SI systems.
Symbol Quantity Value (UnitsJ
In the SI system, relative pemteability J..Lr (fL = J..L,p.,J and
relative lJennittivity Er (E = ErEo) are used. Relative penne- permittivity of free space 8.8542 X 10- 12 (Fom- I )
ability and relative permittivity (dielectIic constant) are permeability of free space 47T x 10-7 (Hom-I)
dime~lsionless in the SI system. charge on an electron -1.602 x 10- 19 (e)
The mathematics of electromagnetic field studies are speed of light (vacuum) 2.9979 x 108 (mos- 1)
briefly discussed in PelIt 8 of this section.

I
I
TABLE 3. Conversion factors for SI units in study of electromagnetism

Multiply Non-SI Quantity by To Obtain SI Q~antity

Flux density: gauss 10- 4

flux density: tesla
Field strength: oersted 103 /4 field strength: ampere per meter
Susceptibility: emu/cc (dimensionless) 4 susceptibility (dimensionless)
Flux: lines 10-8 flux: weber
Flux density: lines per square inch 0.155 X 10-5 flux density: tesla
Demagnetizing factor (dimensionless) 1/4 demagnetizing factor (dimensionless)
Magnetic dipole moment: erg/G 10-3 magnetic dipole moment: ampere per meter squared
Magnetization: emu/cc 103 magnetization: ampere per meter
Magnetization: gauss 103/4 magnetization: ampere per meter
104 / MAGNETIC PARTICLE TESTING

PART 1
FUNDAMENTALS OF
ELECTROMAGNETISM

Coulomb's Law Use of Maxwell's Equations

Maxwell's equations do not take motion into account and
The nature of electromagnetism can be 'summarized by therefore do not include the induction of currents due to
four vector quantities, their interaction, their relations with motion. To do so, it is necessary to add the Lorenz force
each other and with matter. These four t.ime dependent equation and the so-called constitutive relations. It may also
vector quantities are referred to as...!lectromagnetic fields be useful to note that, by proper interpretation, relati\-istic
and include: electric field intensity (E), electric flux density effects can also be handled (this application is found in the
CD), magnetic field strength (H) and magnetic flux density literature). 1-5
(8),. . '. ' . d . The approach in this section is to start with MaxWell's
The study of electromagnetic fields begins with the stu y equations and derive from them all the necessary relation~.
of basic laws of electricity and magnetism and with the use In . doing so, the equations are accepted as the bas~c
of some basic postulates. In particular, it is customary to postulates. In particular, at low frequencies, Maxwell s
start with Coulomb's law. This law states that the force equations are identical to those of Coulomb, Faraday, Gauss
between two stationary charges is directly proportional to and Ampere. The results derived here are general and,
the size of the charges and is inversely proportional to the within the assumptions made in their derivation, apply to a
square of the distance between them. wider range of applications. Only those electric and m~g­
Adding Gauss' and Ampere's laws pro"ides a complete set netic phenomena most directly related to nondestructIve
of relations describing all electrostatic, magnetostatic and testing and, in particular, to magnetic particle testing are
induction phenomena, but not the propagation of waves. To considered in detail here. Other electromagnetic phenom-
include wave propagation in electromagnetic field equa- ena are mentioned briefly for completeness of treatment.
tions, the displacement current (continuity equation) is Mcu'Well's equations are a set of nonlinear, coupled,
added to Ampere's law. By doing so, Maxwell's equations second order, time dependent pmtial differential equations
are obtained. whose general solution is difficult to obtain. Some ~1ethods
Alternatively, Maxwell's equations may serve as the basic for the solution of the electromagnetic field equatIOns arc
postulates and, because they form a complete set describing discussed below, including numerical methods that have
all electromagnetic phenomena, the required relations may become prominent in recent years.
be deduced. By choosing Maxwell's equations as the starting
point, an assumption of the equations' accuracy is i~plicitly
made. This is not more troublesome than assummg that Units of Measure and Terminology
Coulomb's law applies or that displacement currents exist.
In either case, the proof of correctness is experimental. This One of the major sources for confusion when applying
is an important consideration: it specifically states that electromagnetic field theory is the system of units used ~or
Maxwell's equations and therefore the electromagnetic field measurements. The cgs (centimeter gram second \lIlltS,
relations cannot be proven mathematically. including the electromagnetic systetn of llnits or emu) and
 BASIC ELECTROMAGNETISM / 105

the MKSA (meter kilogram second mnpere units) are among involved) but the benefit gained in consistency far out-
the most familiar, but other systems such as the absolute weighs any inconvenience.
magnetic, the absolute electric and the so-called nonnalized Another problem encountered in practice is the confu-
system are also used. What is more disturbing is the fact that sion between rnagnetic field strength (H), sometimes called
mi.xed units are often used. For example, it is common for field intensity, and magnetic fl!:!::."C de:!!:sity (B). The term
practitioners to use SI (or MKSA) units such as the ampere magnetic field is often used for H or B or both, depenc!ing
for electrical quantities, and Gaussian emu units such as the OIl the situation. To avoid such confusion, the quan~ty B is
gauss for magnetic quantities. Only SI units are used used consistently for the magnetic flux density \vhile H is the
throughout this section. This may at times appear inconve- magnetic fielcL strength. Similarly, If is the electric field
nient (hecause of the Vel)' large or very small quantities intensity and D is the electric flux density.
 '06 / MAGNETIC PARTICLE TESTING

PART 2

FIELD RELATIONS AND MAXWELL'S

EQUATIONS

Maxwell's equations are summarized in Eq. 1 through

Eq. 4 in differential form and in Eq. 5 through Eq. 8 in At this pOint, the equations are neither linear nor
integral form. Equation 9 is the Lorenz force equation nonlinear. This important behavior is detem1ined through
which describes the interaction of electric and magnetic material properties and is not inherent in the equations. The
fields with electric charge.
material properties follow the constitutive relations:

VXE as s = JLR
(Eq. 10)
at (Eq. 1)
75 = €oE (Eq.11)
v x R =] a75
+ (Eq. 2) ] = uE
at (Eq. 12)
v ·75 p (Eq.3) If any of these relations is nonlinear, the field relations
are nonlinear. In particular, the permeability is known to be
v.os 0 highly nonlinear for ferromagnetic materials. In some cases,
(Eq. 4)
the conductivity and permittivity may also be nonlinear.
T~e electric conductivity (1', magnetic permeability JL and
i
c
E· ae dcfl
dt (Eq.5) the electric permittivity €o are generally tensor quantities.
While for many pl,'actical purposes it cail ·be assumed that
they" are scalar quantities, materials are encountered in
<fHoae ~ I +ja15 o(Ts (Eq. 6)
practice that behave differently. These exceptions are de-
c sat fined through. linearity, homogeneity and isotropy of the
materials. Only for linear, isotropic, homogeneous materials
are the material properties Single-valued scalar quantities.
is 75 .a.;= Q (Eq. 7)
Linearity, Homogeneity and Isotropy
iSea.;=o A medium is said to be homogeneolls if its properties do
s (Eq. 8)
not vary from pOint to pOint within the material. A medium
of
is linear in a property when that property remains constant
q(E + v x B) while other changes may occur. The permeability of most
(Eq. 9)
nonmagnetic materials is considered to be independent of
Equations 1 and 5 are a statement of Faraday's law of the field. These materials are usually considered to be linear
induction. Equations 2 and 6 are a modified form of in permeability. A material is isotropic if its properties are
Ampere's law. The addition of the displacement current independent of direction. This means that the permeability
term (a D/a t) was Maxwell's conhibution to the original laws of an isotropic material must be the same in all three spatial
of electricity. The displacement current, although often directions.
taken as an assumption, is nothing more than an expression A good example of an anisotropic material is a permanent
that can be derived from the continuity equation. Equations magnet. Most materials, including iron, are anisotropic on
4 and 8 are Gauss' law for magnetic sources and they state the clystal level. Because these cI)/stals are randomly oIi-
the nonexistence of isolated magnetic charges or poles. ented, it may often be assumed that a macroscopic material
Equations 3 and 7 are Gauss' law for electric charges. is isotropic. This is not necessarily the case for hard steels or
for steels with large, preferred orientations.
 BASIC ELECTROMAGNETISM / 107

Static Fields the normal to the surface enclosed by the contour C is in the
direction of the current, then the field is described by the
By setting to zero all time derivatives in Maxwell's light hand rule: if the current is in the direction of the
equations, the equati?ns for the stat.ic electric and magnetic thumb, the field is in the direction of the fingers (see Fig. 1).
fields are obtained. 1 he four equatIons become: The total current I in Eq. 18 is the current enclosed within
the contour.
v X E= 0 (Eq. 13)
v X H = .r CEq. 14) The Magnetic Vector Potential
v - I5 P (Eq. 15) Eq uati9n 14 is a cross product and results in a vector
quantity J. Because the cross pro~lCt of a vector is also a
v-n 0 CEq. 16) vector, the magnetic flux denSity B m~ be written as the
curl (cross product) of another vector A.

I i
iE-de
H - de = I
= 0 (Eq.17)

(Eq. 18)
B=VxX
Thus, Ampere's law (Eq. 14) can be written as:
(Eq.23)

c 1 __
V x - (V x A) = ]
f. l5 - ds = Q (Eq. 19)
/.L
(Eq. 24)

Th~ vector A is called a magnetic vector potential. For an

isotropic linear medium, the follOwing vector identity can be
fB-as=o (Eq:20) us.ed to simplify this expression: .
s

V x V X A = v(V· X) - V 2X (Eq. 25)

Equations 1.3 and 15 are the goveming equations for
electrostatic fields (Faraday's and Gauss' law). Equations 14 A vector is only defined when both its divergence and curl
and 16 are Ampere's and Gauss' (magnetic) laws for are speCified. The divergerice may be specified in different
magnetostatic applications. Note that Eqs. 13 and 15 do 'not ways. The Simplest is to set it equal to zero in Eq. 25.
contain the magnetic field while Eqs. 14 and 16 do not Ampere's law then becomes:
depend on the electric field. Thus, the equations for
electrostatics and magnetostatics are completely decoupled V2X = -J..t]
and electrical quantities can be calculated without resorting (Eq.26)
to the magnetic field and vice versa. which is the vector Poisson equation.
As with the general system of equations, the Lorenz
equation has to supplement these equations. Because of the
decoupling of the two sets, Lorenz' force equation should be FIGURE 1. The right hand rule: if the thumb of
used in its electrostatic form for electrostatic fields: the right hand is in the direction of the current,
the fingers show the direction of the field
(Eq. 21)
For magnetostatic fields, only the magnetic force exists.
MAGNETIC FLUX LINES

F = qv X B (Eq. 22)
The elechic current or, more conveniently, the elechic
(CSee
Urrent density] is the source of the magnetic field strength
Eq. 14).
calEqua~ion 18 is particularly useful because it allows the
~UlatlOn of the magnetic field strength and the determi-
nation of its direction. If the integration is taken such that
 108 I MAGNETIC PARTICLE TESTING

Equation 23 defines the magnetic vector potential. This

definition of the magnetic vector potential A allows the use FIGURE 2. A straight current carrying wire and
of a simpler Poisson equation instead of the original field the relation of the current and the field at
equations. point P

Biot-Savart law
The purpose of field relations is to solve field problems.
Any of the relations obtained previously may be used for this
purpose. In particular, Eq. 14 can be used for general field dl
problems while Eq. 18 is useful for solution of highly
symmetric problems. In problems where no such symmetry
can be found but which are simple enough not to require
the solution of the general Eq. 26, another method can be
used. For these types of problems, the magnetic vector
potential is used in still another form. Considering the
current in a straight wire in Fig. 2, the follOwing can be
written at point P:

dA = ~ dl (Eq.27)
If the magnetic vector potential in Eq. 28 is substituted
41T R back into its definition, the follOwing expression is obtained
for the flux denSity:
In Eq. 27, the definition of the magnetic vector potential in
Eq. 23 has been used. In order to find the total magnetic
vector potential, this differential is in~egrated over a closed (Eq.29)
contour to obtain:

A = J-tI 1· (fl. (Eq. 28) 'where; denotes a unit vector in the direction ofR in Fig. 2.
41T 1: R This expression, known as the Biot-Savart law, allows the
c<;l,lculation of the flux denSity directly. It is particularly
In order to find the field strength or the flux denSity, it is useful for situations where the current paths are clear and
necessary to find the curl of this expression by performing the required contour integration is in the direction of the
the operations in Eq. 23. currents.
 BASIC ELECTROMAGNETISM / 109

PART 3
ELECTROMAGNETIC FIELD TYPES AND
BOUNDARY CONDITIONS

1 - - -
Steady State Alternating V x - (V x A) = I + jawA (Eq.36)
p.,
Current Fields
This equation, often called the curl-curl equation is a
Low frequency alternating current fields are unique in diffusion equation and is the basis of many analytical and
that a simpler form of Maxwell's equation can be used. The numerical methods. By using the vector relation in Eq. 25,
displacenient current term (aD/at) in Eq. 2 or Eq. 6 Eq. 36 can be wlittenas:
~epends on frequency and is vel)' small for low frequencies.
In fact, this term may be neglected within conducting V 2 A - V(V • A) = - p.,J + jwp.,aA (Eq. 37)
materials ~mcl for all frequencies below about 10E + 13 Hz.
If this assumption is introduced and the term neglected, the Here, the divergence of the magnetic vector potential may
pre-Maxll)ellian set of equations is obtained. be chosen in any convenient, physically sound method. By
By introducing a phasor notation for all vectors, the time choosing a zero divergence and rewtiting.the Laplacian V2A
dependency' is not explicitly used and the' transformed . 'in its diffel'ential form; a partial differential equation can be
system is both simpler in presentation and in solution. A written for two dimensional (Eq. 38), axisymmetlic (Eq. 39)
general vector can be expressed as a phasor by the following and three-dimensional ge~metries (Eq. 40). Equations 37
definition: through .40 assume linear permeability.

A(x,y,z,t) = real [A(x,y,z)e jwt ] . (Eq. 30)

Maxwell's equations (in differe~tial form) can now be

- Is + jwaA;::. (Eq. 38)
written as:

vx E ~jwf3 (Eq.31)
(Eq.39)
v x fI = J + j wD (Eq.32)

V•D = p (Eq. 33) - Is + jwaAcfJ

V • f3 = 0 (Eq.34)

In this form, all time derivatives were w]itten as: (Eq.40)

aA _
-at = jwA. (Eq.35)
These equations were obtained by substituting the mag-
netic vector potential in Maxwell's equations. Because other
TI~is v~rsion of ~hmvell's equations is often called the potentials (vectors or scalars) can be defined, the field
quaslstatu: form and is vel'\' convenient for manv alternating equations may be obtained in terms of these_ functions or in
?urrent calculations at 10\~' frequencies including alternat- terms of the otiginal field quantities Band H. A Poissonian
. eu].rent Ieakage fields. The folloWing
Ing ' equation is ob- (or Laplacian) form as in Eq. 40 are paliicularly useful
tamed by a method similar to that used to obtain Eq. 26. because of the standard methods available for their solution.
 110 I MAGNETIC PARTICLE TESTING

Time Dependent Fields found in terms of the electric scalar potential (Eq. 45), the
magnetic field strength (Eq. 46) or the electric field inten-
Instead of neglecting terms in Maxwell's equations, if the sity (Eq. 47).
complete sets in Eqs. 1 through 4 or Eqs. 5 through 8 are
used, then the general time dependent form of Maxwell's
V' 2v - a2v _E
equations is obtained. As was mentioned above, this form is J.tE at 2 (Eq. 4.5)
E
completely general and, when combined with the appropri-
ate boundary conditions, can be solved to obtain all electro-
V'2H - a2H
magnetic phenomena, including those related to static JLE at 2 = 0 (Eq. 46)
fields. In practice, exact solutions are rarely obtained be-
cause of the complexity involved. Approximations or numer- a2lf
ical methods are often required for the solutions of this type V'2lf -
of problem. JLE at 2 = 0 (Eq. 47)

~ave Propagation The last two equations were obtained under Source-free
conditions and are therefore homogeneous wave equations
\rVhen describing wave propagation, the time dependent for the magnetic field strength and electric field intensities
respectively.
form must be used. A wave equation may be obtained by
using the definition of the magnetic vector potential in If phasors are used in Eqs. 46 and 47, instead of time
Eq. 23. By substituting this into Eqs. 1 and 2 and using the dependent vectors, similar forms are obtained for wave
constitutive relations in Eqs. 10 and 11, the following equations where the time derivative is replaced by jw. Thus,
~quation is obtained. the wave equations in Eq. 46 and 47 can be written as:

V' x V' x A
- = J.t]-'+ J.tE ata (.- V'V - at
ax) (Eq. 41) and (Eq.48)

o . '(Eq. 49)
Using the vector identity in'Eq. 2.5, Eq. 41 can be written as:
The constant k is defined as:
2

V'2jf - I.U aatjf2 =. - J.t] + V' ( V'. jf (Eq. 42) k = wVji€ (Eq. 50)

+ av)
J.tE- These equations are known as the hOmOgeneous Helmhol
at equations and describe the harmonic form of the electro-z
magnetic waves.
Because the magnetic vector potential requires the defi-
nition of the curl and the divergence, it is possible to define Skin Depth
its divergence in whatever way the situation requires. The
follo\ving form may be chosen: In a linear isotropic material, after substitutioh of the
constitutive relations in Eqs. 10 through 12 and using the
-
V' • A + J.tE-
av vector identity in Eq. 2.5, Eqs. 1 and 2 become:
at o (Eq.43)
V'2lf _ alf a 2It
Equation 42 then becomes: aJL at - EJL at 2 0 (Eq.51)

V'2H - aH a 2H
(Eq.44) aJ.ta-t - EJL at 2 = 0 (Eq.52)

This is a nonhomogeneous wave equation for the mag- Thus, It and B satisfy identical wave equations \vith a
netic vector potential. It is considered a wave equation damping (dissipative) term proportional to the conductivity
because its solution represents waves traveling at a velocity and magnetic permeability of the material. In a good
(J.tE)-1I2. conductor such as most metals, the second order derivative
This pmticular form was found by choosing to use the may be neglected for 1m\' frequenCies since it is due to the
magnetic vector potential. Similar wave equations may be displacement current in Maxwell's second equation. For
example, Eg ..52 becomes:
 BASIC ELECTROMAGNETISM / J J J

(Eq. 53) FIGURE 3. Boundary conditions between two

materials

These are simple diHllsion equations. If an alternating

magnetic field strength Ho exp (jwt) or Eo(jwt) is ~pp~ied
arallel to the surface of the conductor, tl~ electnc fIeld
lntensity E or the magnetic field strength H is attenuated
exponentially with distance below the sUlface. The attenu-
ation is exp (-xl 8) where x is the distallce below the surface
and 8 is the skin depth given by:

(Eq. 54)
ILl

This is an important factor to consider for alternating

current magnetic patticle testing because, even at 60 ~z,
the skin depth can be quite small. For example, for a t)pIcal
ferromagnetic material with a conductiVity of a = 0.5 X
7 The boundary conditions are the same for the magneto-
10 , a relative permeability J.Lr = 100 and a frequency of static and time vmying fields. These conditions are summa-
60 Hz, the skin depth 8 is 3 mm. This is proba~ly an rized as follows:
overestimate because a linear material was assumed III the
derivation. .' .
For this reason, alternating current magnetic particle 1. The tangential component of the el~ctric field ~nten- .
methods such as' the so-called swin{!.~ng field methods', sity E and the normal COI]lpOnent of the' magnetIc flux
'generaII; detect 'oBly disc~ntinuities which. ~re open to the density B' are continuous across the ~oundary. .
surface. These methods are. more senSItIve to suiface 2. The normal component of the electnc flux denSIty D
breaking discontinuities because the applied field is concen- and the tangential component of the magnetic field
trated at the surface. strength H are discontinuous across the boundmy- The
discontinuity depends on the existence of surface
charges and currents. For situations where no such
Electromagnetic Boundary charges or currents exist, either co~nponent ma~ be
Conditions continuous, depending on the matenals and the fIelds
involved.

Electromagnetic fields behave differently in different

materials. The constitutive relations in Eqs. 10 through 12 The four conditions presented in Egs. 55 through 58 can
are a statement of this behavior. \Vhen different materials be used in order to describe the fields in different materials
are present, the fields across the boundaries between these and across their boundaries.
materials must undergo some changes to conform to both The four conditions are not entirely independent and
materials. In such cases, the field may experience a discon- should be specified \\;ith care. For example, in time varying
tinuity at the boundalY. In order to find the necessary fields, specification of the tangential component of E
c~nditions that apply at material boundmies, assu,me two (Eq.55) is eqUivalent to the. s~ecification. ~f t!le normal
different matelials as in Fig. 3 and apply Maxwell s equa- component of B (Eq.58). Sllllllarly, speCIfICatIon of the
tions at the boundary. For convenience, the integral form is tangential component of H is equivalent to that of the
used. By doing so, th~ follovdng four conditions are obtained: normal component of D. Only hvo of the four may be
specified independently (the tangential component of E and
EIT = E 2r (Eq.55) the tangential component of H or any other acce~t~ble
combination). Over-specification of boundary condItIons
x (HIT - H )
1/
2r = Is (Eq. 56) may result in contradiction of conditions and may therefore
be in error.
1~ • (DIll - D 211 ) = Ps (Eq. 57) The boundar), conditions in Eqs. 55 through 58 were
obtained by using Maxwell's equations directly. In order to
BIll = B211 (Eq. 58) render th~se relations more useful, it is convenient to
introduce the constitutive relations in these conditions and
 112 / MAGNETIC PARTICLE TESTING

find the interface conditions for some special classes of

common materials. Two such groups of materials often vVhen D21l 0, then:
found in practice are:
(Eq. 69)
1. boundary conditions between two lossless media (a
lossless medium is one that has zero conductivity with BIn = B21l = 0 (Eq.70)
arbitrary permittivity and permeability; two perfectly
insulating materials are considered here); and Note that the boundary conditions in Eqs. 67 through 70
2. boundary conditions between a loss less material and a only apply for perfect conductors. This rarely arises except
good conductor. for Simplified problems and for superconductors. In the
case of a superconductor, these boundary conditions are
At the boundary between two good insulators, no current also correct for the static field.
densities and free charges are normally present. Thus, all Proper application of the field equations and imposition
four components in Eqs. 55 through 58 are continuous. of the correct boundary conditions result in a correct
These then can be rewritten using the constitutive relations solution to the field equations.
in Eqs. 10 and 12 as:

DIT The Continuity Equation

=~ (Eq·.59)
D 2r €2 Since charge cannot be destroyed or created, the only
possible way to charge a body is through flow of charge from
BIT one pOint to another. This is stated mathematically by the
=& (Eq.60) continuity equation:
B2r J.1.2

€IE ln €2 E 211 (Eq.. 61) _ ap

at (Eq. 71)
J.LIH ln J.L2 H 211 (Eq. 62)
This often neglected relation is fundamental to under-
At the interface between a good conductor and an
standing fiel.d be~avior and is responsible for two important
insulator, both surfacecu'rrent densities and free charges
aspects of electromagnetic fields: Kirchoffs current law and
may exist. The elechic field is zero inside a perfect conduc-
the linking of electric and magnetic quantities. Observing
tor and both the tangential component of the electric field
that the continuity equation is in fact a statement of the
and the normal component of the electric flux denSity must
divergence of the current denSity, the displacement current
be zero inside the conductor. The boundary conditions then
become: term in Maxwell's equations is shown to be a statement of
the continuity equation or preservation of charge. Ampere's

Elr = ° (Eq. 63)

law, before Maxwell's modification is:

n X (HIT - H 2r ) (Eq.72)
= Js (Eq. 64)
n··D IIl = Ps (Eq. 65)
The divergence of the curl of a vector is zero (identically).

BIn = B21l (Eq.66)

v • (V X H) = 0 (Eq. 73)

Note that while Eqs. 63 through 66 are ~rrect .!or the Thus, this equation can be written as:
static field, for the time vmying field, both Band H must
also be zero inside a perfect conductor. Thus, Eqs. 63
through 66 must be modified for the time varying case to: V • (V X H)
o (Eq. 74)

(Eq. 67)
From Gauss' law (MaX\vell's third equation), this can be
When H2r = 0, then: written as:

aD)
(]- +_
nXHIT=Js. (Eq. 68) V • (V X H) V·
at (Eq. 75)
 BASIC ELECTROMAGNETISM I 113

And the following is the correct form of Ampere's law:

tion need not be taken into account explicitly for purely
electrostatic solutions because this implies static charges
aD and therefore Eq. 71 is always satisfied implicitly. For direct
VxH=J+- (Eq. 76)
at current magnetic applications, Eq. 71 is again satisfied
because any flow of charges is constant. The only time the
equation needs to be introduced directly is when the
Note that this form is only necessmy when both electIic
and magnetic quantities are present. The continuity equa- displacement currents are large compared to conduction
currents (at very high frequencies).
114 / MAGNETIC PARTICLE TESTING

PART 4
EFFECT OF MATERIALS ON
ELECTROMAGNETIC FIELDS

direction, a total magnetization is generated. The magneti-

Material Properties and Constitutive zation M is then given by:
Relations
(Eq.78)
Magnetic properties are important because of their effect
on the behavior of materials under an extemal field (under The magnetic flux denSity of the material is then given by:
active excitation) or when the external field is removed
(residual magnetism). The magnetic properties are ofter (Eq.79)
discussed using the magnetic permeability of materials. This
important quantity is defined through the constitutive rela- m
The terms H, and M are vectors. This implies that a net
tion in Eq. 10. magnetic field or flux denSity can only exist if these vectors
Permeability govems two important features of the mag- are aligned in such a way that a total net vector M exists. If
netic field and therefore affects any application that uses the the independent vectors m are randomly oriented, as is
magnetic field. Flux density B is often the quantity of often the case, the net magnetization is zero.
interest and has ,higher values for high values of the For the purposes of this chapter, three types of magnetic
'permeability for a given source field strength H. Secondly, materials are important: diamagnetic, paramagnetic and
the permeability also defines whether the field equation is ferromagnetic.
linear or nonlinear. .
Tne permeabllity of free'space is /-to = 47T' X 10-7 Hem-I .
.Other materials may have larger or smaller permeabilities.
FIGURE 4. Representation of material properties:
Table 4 lists the relative permeabilities of some important (a) the field due to a current loop; and
materials.
(b) current loops created by spinning electrons
The magnetic properties of materials are defined through
the interaction of external magnetic fields and moving
charges in the atoms of the material (static charges are not fa)
influenced by the magnetic field since no magnetic forces
are produced in Lorenz' law). Atomic scale magnetic fields

cU&
are produced inside the material through orbiting electrons.
11
These orbiting electrons produce an eqUivalent current loop , I I

that has a magnetic mom<:nt. I

(Eq.77)

Where:

7Ta 2 = the area of the loop;

! = the equivalent current (see Fig. 4a); and
z = a unit vector normal to the plane of current flow.
Many such atomic scale loops or magnetic moments exist
and the matelial volume contains a certain magnetic mo-
ment denSity. If N magnetic moments per unit volume are
present, and if these moments are aligned in the same
 BASIC ELECTROMAGNETISM I 115

the internal structure of their magnetic moments. In these

Diamagnetic Materials materials, many atomic moments are aligned in a certain
direction within a very small region called a magnetic
In these materials, the internal magnetic field due to domain. NeighbOring domains have a similar structure, with
electrons is zero under normal conditions. In an external the net magnetic domain in one direction. In the demag-
magnetic fJdd, an imb~lanc~ occurs and a net inte.rnal field netized state, the magnetic domains tend to be aligned
opposing the external fIeld IS produ~ed. :rhus, MIll Eq. 7~ randomly, exhibiting a net intemal field that is either very
is negative \\ith respect to the apph:d fIeld. The magne~l­ small or zero.
zation is proportional to the extemal held through a quantlty This domain model is depicted in Fig. 5. \\Then an
called the magnetic susceptibility of the matmial x 1ll • extemal magnetic field is applied, those domains that have

M = xI/IBex (Eq. 80)

FIGURE 5. Domains in a ferromagnetic material:
dQmain-8 is aligned with the field and will grow
In terms of the applied flux density, this becomes:
as H is increased; domaln-3 is aligned against the
field and will shrink as H is increased
B= J.L,,(1 + x"JHex (Eq. 81)

The magnetic permeahility of any material can be written

as:
(Eq.82)
R
In diamagnetic materials, the magnetic susceptibility is
very small and negative. Its magnitude is usually on the
order of 10-5 . The net effect is that the relative permeabilities
exhibited by dhur~agnetic materials are slightly smaller than
1.0. This gro'up of materials includes many familiar metals
including pure copper and lead.
Under special conditions such as temperatures less than
- 150°C, some materials may become ~uperconducting. An TABLE 4, Relative permeabllities for some materials;
ideal superconductor has a magnetic susceptibility of - 1 values given for ferromagnetic materials represent
and· a permeability of O. A superconductor expels magnetic approximate maximum relative permeabilities
flux (the Meissner effect) from its interior.
Relative
Type of Magnetic Material Permeability
Paramagnetic Materials Diamagnetic materials
Gold 0.999964
This group of materials exhibits properties similar to Silver 0.99998
diamagnetics except that the magnetic susceptibility is Copper 0.999991
positive. In the presence of an applied magnetic field Lead 0.999983
strength, the atomic magnetic dipole moments can align to Water 0.999991
form a net magnetic dipole density. The effect is still Mercury 0.999968
relatively small, prodUcing observed relative permeabilities Bismuth 0.99983
slightly larger than 1.0. ParamagnetiC materials
The permeability of paramagnetic materials remains con- Vacuum 1.0
stant over a large range of applied magnetic field strengths. Air 1.00000036
Examples of materials in this group are air, aluminum and Aluminum 1.000021
some stainless steels. Ferromagnetic materials
Cobalt (99 percent annealed) 250
Nickel (99 percent annealed) 600
Iron (99.8 percent annealed) 6.000
Ferromagnetic Materials Iron (99.95 percent annealed in hydrogen) 2.0 X J05
SupermalloyTM (annealed, controlled cooling) 1.0 X 106
Ferromagnetic mateIials vary from diamagnetic and para-
Steel (0.9 percent carbon) JOO
~~gnetic matelials in two critical ways: (1) their suscepti-
Iron (98.5 percent cold rolled) 2,000
bIlIty is very large; and (2) there is a pronounced variation in
116 / MAGNETIC PARTICLE TESTING

a net field aligned in the direction of the applied field grow beyond a certain field, all the magnetic domains are aligned
in size while the other domains shrink. The internal field with the field and an increase in the magnetic field strength
and the external field H are aligned in the same direction cannot increase the net magnetization. Materials typical of
producing a larger total flux density B. The above argument this group are iron, steels, nickel and some stainless steels.
is related to the hysteresis curve of a ferromagnetic material Table 4 summarizes some of the more important ferromag-
and explains why any such curve has a saturation region: netic materials and their permeabilities. Table 5 lists con-
ductivities of various materials and Table 6 is a listing of
dielecbic constants.
TABLE 5. Conductivities of some materials As is evident from any hysteresis curve, the permeability
of ferromagnetic materials is not constant but varies with
Conductivity the field. This is exhibited through the slope of the initial
Material (siemens per meter)
magnetization curve to which the permeability is tangent.
Silver 6. I X 10 7 Thus, most ferromagnetic materials are highly nonlinear
Copper (100 percent lACS) 5.8x 10 7 materials.
Gold 4. I X 10 7
Aluminum 3.5 x 10 7
Tungsten 1.8 x 10 7
Brass 1.1 x 10 7
Iron (pure) 1.0 x 10 7 TABLE 6. Dielectric constants (relative
Soft steel 0.8x 10 7 permittivitiesJ for some materials
Carbon steel (I percent carbon) 0.5 x 10 7
18-8 stainless steel 1.4 x 106 Relative
Material Permittivity
Nichrome 0.9 x 106
Mercury 1.0 x 106 Vacuum I
Graphite 1.0 X 105 Air 1.0006
Carbon 3.0 X 104 Rubber 3
Germanium 2.3 Paper 3
Sea water 4.0 Bakelite™ 5
Silicon 3!9 x 10-4 Quartz 5
Bakelite ™ 1.0 X 10-9 Glass 6
Glass 1.0 X 10- 12 Mica 6
Rubber 1.0 X 10- 13 Water 81
Mica 1.0 X 10- 15 Barium titanate 1,200
Quartz 1.0 X 10- 17 Barium strontium titanate 10,000
 BASIC ELECTROMAGNETISM I 117

PART 5
MAGNETIC CIRCUITS AND HYSTERESIS

To develop this concept, it is convenient to use a toroid

Magnetic Circuits (Fig. 6). The gap is assumed to be small and the flux
densities inside the toroid and the gap are assumed to be the
The two cCluations that define the static magnetic field same. This in effect neglects any fringing effects in the gap.
are Eq. 14 and Eg. 16. These are writteE: below both in If the field strength is denoted in the gap as Hg and in the
differential and integral form in terms of B: toroid as He, then the fields can be calculated in terms of
the flux density Be in the toroid and the permeabilities of
v X B = J-tJ (Eg. 83) the gap and the toroid (/-Lo and J-t) as:

V· B = 0 (Eq. 84)
H e -- Be (Eg.88)
/-L
(Eq.85)
and

H
g
= Be
• . (Eg. 89)
{Beds = 0 (Eg.86) /-Lo

By substituting thes'e in Eq. 87, th~ magnetic flux density

These are Ampere's and Gauss' laws for the static field. is found to be related to the lengths of the gap (C g) and the
They can also be viewed as defining a veCtor quantity B length of the material in the toroid (27Tr - Cg ) where r is
through. its curl and divergence. . the mean radius of the toroid:
The line integral of the magnetic field str.ength around a
closed path is defined as a magnetomotive force (or (Eg. 90)
magnetomotance) :

v 111
= fH
C
e ([l = NI . (Eq. 87) FIGURE 6. A toroid with an air gap used to
define the magnetic circuit concept

The units of the magnetomotive force are customarily

expressed as ampere-turns although the correct unit is the
~mpere. The modification from I to NI simply states that,
If the total current inside the closed contour is divided into
N wires, then the number of turns may be used for
convenience.

The Use of Circuit Theory

A magnetomotive force VIII = NI causes a magnetic flux
i to exist ,\ithin the closed contour mentioned in Eg. 87. If
bor any reason this flux is contained within a material, it may
e assumed that a flux flows \x.1thin the material. This
concept allows flux to be treated much the same way as
~urrent and therefore circuit theory concepts may be used
or the solution of some specific fi~ld problems. '
 , '8 I MAGNETIC PARTICLE TESTING

If it is assumed that the magnetic flux density is uniform

within a material (it is uniform inside a toroid but rarely in For simplicity, an analogous magnetic circuit can be
other shapes), the flux can be calculated as: defined as in Fig. 7. Because of its simplicity, this approach
has found considerable use in many areas, especially in
devices with closed paths (transformers and machines). The
<I> = Bs (Eq. 91) approach is quite limited in scope because of the approxi-
mations used to derive the concept. First, the fringing
The total flux through the toroid or the gap is therefore: effects cannot be neglected for large air gaps. Second, there
are always some leakage fields that cannot be taken into
account. Finally, the permeability has been assumed to be
<I> NI
(Eq. 92) constant. In most cases of practical impOltance, the perme-
ability of a material is field dependent (Eq. 98).

Written in terms of the magnetomotive force V m , the

equation for the flux can be written as: Hysteresis
The constitutive relation between the magnetic field
. Vm strength and the magnetic flux denSity is shO\\1n in Eq. 10 .
<I> (Eq. 93)
Re + Rg The behavior of the field \vithin different materials has been
described above. However, these do not describe all phe-
where nomena that exist within materials.
Inspecting Eq. 10 shows that by increaSing the magnetic
= if. field strength H, the flux denSity B increases by a factor of
Re (Eq. 94) /-to However, for ferromagnetic materials, Eq. 10 must be
JLS
written as a nonlinear equation:
and

f-L(H)H
Rg =~ ·(Eg. '95)
(Eq., 98)
f-Lo S
An alternative way to look at this phenomenon is to
inspect the domain behavior of a ferromagnetic material.
The forms of Eq. 94 and 9.5 are analogous to that ~f the Initially, the domains are randomly oriented. As the applied
direct current resistance (R = pi fa) and are therefore
called magnetic resistances or reluctances. The reluctance of
the gap is Rg and Re is the reluctance of the material in the
FIGURE 7. Equivalent magnetic circuit
toroid. The units for reluctance are 1 per henry (l.H-I). representation
Similarly, if magnetomotive force is considered analogous to
voltage and flux analogous to current, Eq. 93 is analogous to
Ohm's law.
. For any closed magnetic path, the equation can be
</>
written as:

2: NJi = 2: I~<I>j j
(Eq. 96)
+

Similarly, by using the divergence of the magnetic flux

density V • B = 0, the law for a junction is:

(Eq.97)
 BASIC ELECTROMAGNETISM I 119

field increases, domains begin to grow by displacing other

FIGURE 8. HystereSiS curve: (a) the initial
domains and eventually occu~ying most of the material
volume. Any further increase of the field has little effect on magnetization curve and (b) the hysteresis curve
the domains and therefore has little effect on the flux
density in the material; thus the permeability depends
strongly on the applied field. la)

Magnetization Curves
A plot of the relati~n in Eq. 98 describing the flux density
as a function of the field strength is a useful way to look at
magnetic materials. For linear materials (materials for
which the permeability is constant at any field value), this
CUlve is a straight line whose slope is equal to the perme-
ability. Ferromagnetic materials behave differently. The
curves in Fig. 8 desclibe the behavior of iron. Initially, the
applied field strength is zero and so is the flux denSity.
As the field is increased, the flux denSity also increases
but, unlike linear materials, the curve is not linear. At some
field value HI, the curve starts bending and the slope of the
curve is reduced Significantly. Any increase beyond the field
H3 increases the flux density but not at the same rate as at
lower points on the curve. In-fact, the slope in this section
of the curve approaches unity, meaning that the relative fb' B
permeability approaches 1. This region is called saturatio,!!
and is dependent oh the mateIial tested. The whole curve
described in Fig. 8a is called a magnetization curve. Since it
starts with zero applied field it may also be callec;l an initial
magnetization curve.
H

Hysteresis Curves
RedUcing the applied field moves the curve to the left,
rather than retracing the initial magnetization curve (Fig.
8b). The flux density is reduced up to the point Br where the
applied field is zero. This residual flux is called renumance
or retentivity and is typical of all ferromagnetic materials.
Applying a reverse magnetic field fmther reduces the flux
de.nsity up to the point He where an applied field strength
eXists \\ithout an associated flux denSity. The field strength
at this point is called the coercivity or the coercive force of
the material. FU1ther increase in the negative field strength different materials, including their coercive forces and
traces the magnetization curve through point P2 where a remanance, are markedly different).
saturation point has again been reached, except that in this The slope of this cmve at any point is the magnetic
case the field strength and the flux denSity are negative. permeability. The slope is relatively high in the lower
. If the applied field is decreased to zero, a point symmet- portions of the initial magnetization curve and is gradually
nc to Br is reached. Similarly, by increasing the applied field reduced to unity. At this pOint, the material has reached
strength to a value equal (but positive) to He> the flux magnetic saturation. A curve describing the slope of the
~e.nsity ~s again zero. Further increase in the field strength initial magnetization curve of Fig. 8a is shown in Fig. 9.
dnng~ It back to the point Pl' Repeating the process Figure 9 shows that for this material (iron), the initial
escnbed above results in a retracing of the outer curve but relative permeability is low, increases gradually and then, as
not that of the initial magnetization curve. This unique the field approaches saturation, decreases and approaches 1.
ma~netization curve is called the hysteresis curve and is The hysteresis curve in Fig. 8b has four distinct sections
t)plcal of all ferromagnetic materials (hysteresis curves of described by the four quadrants of the coordinate system.
120 I MAGNETIC PARTICLE TESTING

Particularly important are the first and second. The CUIVe in in the second quadrant. Secondly, this is the quadrant in
the first quadrant is created by an applied field or source which permanent magnets operate. The coercivity and
and is therefore called a magnetization curve. In particular, remanance of ferromagnetic materials are very different
the initial magnetization CUIVe can only be described by from each other and define to a large extent the classifica-
starting with an unmagnetized sample of the material and tion of magnetic materials. The coercivity and remanance of
then increasing the field within the material. This section of some important materials are shown in Table 7.
the CliIVe is referred to as the active part of the curve. All The area under the hysteresis curve represents energy.
direct current applications of magnetic particle testing that This is understood by referring to the Poynting theorem. In
depend on active magnetization are governed by this section de"ices such as transformers, this is a detrimental property
of the CUIVe. because the energy is dissipated, primarily in the magnetic
The second quadrant (with the limits at Br and He) is core of the de\lice. In other cases, including permanent
called the demagnetization curve. It is important for two magnets or switching magnetic de\lices, this property is
reasons. First, any magnetic material, after being magne- useful.
tized, relaxes to the point B r or more commonly to a point
Magnetization
FIGURE 9. Initial permeability curve for iron In order to magnetize a sample, it is necessary to apply a
magnetic field to the sample. The form in which this field is
applied may vary depending on practical considerations but
the same basic effect must be obtained: the field in the
sample must be increased to a required value.
In general, if a sample is initially demagnetized, the field
is gradually in~reased through the initial magnetization
CUIVe to a required pOint. If a residual method is being used,
the field is reduced to zero and the material relaxes to a
point in the second quadrant of the hysteresis loop. For
pre\liously magnetized samples, it is. usu~Ily better- to de-
magnetize the sample first and then to magnetize it to the
required point.

Demagnetization
The hysteresis CUIVe indicates that when the source of a
field is reduced to zero, there is a remanant flux denSity in
~--------------------------------~~H
H, the material. This remanant or residual field is sometimes
used for testing purposes but in many cases it is desirable to

TABLE 7. Coercivityand remanence of some important materials; figures for He and Br are approximate and
are strongly dependent on thermoelectrical history
Coercive Force Remnant Flux Density Saturation Flux Density
He Br Bs
(amperes per meterJ (weber per square meterJ (weber per square meterJ

Soft magnetic materials

Iron (pure annealed) 100 1.2 2.16
Supermalloy ™ 0.2 10-4 0.8
Steel (0.9 percent carbon, hot rolled) 4,000 1.0 2.0
Silicon iron (4 percent silicon) 20 0.5 1.95
Ni-Zn ferrite 0.34
Hard magnetic materials
16
°
Carbon steel (0.9 percent carbon) 4,000
Alnico V 44,000 1.2
Alnico VIII 126,000 1.04
Samarium cobalt 560,000 0.84
 BASIC ELECTROMAGNETISM I 121

demagnetize a test object before a controlled field is applied a large direct current corresponding to a point on the
. in
r to demagnetize it after a test. hysteresis curve and a small alternating current superim-
md a Demagnetization cannot be achieved simply by creating a posed on it.
~nt
field opposing the original source field. The demagnetiza- Alternatively, if the magnetizing current is suddenly
;a-
tion process is complicated by shape effects that usually decreased and then increased again, the same effect is
of created. This situation results in a change in the hysteresis
cause different operating points to exist in different sections
of the material (see the curve in Fig. 8). curve as shown in Fig. 11. Thus, a small oval curve similar to
Effective demagnetization of materials can be achieved the hysteresis curve is described at the initial point. These
by heating the material beyond the Curie temperature and loops are called minor hysteresis loops to distinguish them
then cooling it in a zero field environment. Under most from the normal (or major) h~teres~ loop. Since perme-
circumstances, this method is impractical because of the ability is defined as the ratio of Band H, the permeability of
metallurgical effects associated with it. a minor loop may be defined as !lBI!lH:
A practical demagnetization approach is to cycle the
matedal through the hysteresis curve while gradually reduc- AB (Eq.99)
ing the magnetic field strength to zero. The effect is shown JLinc = AH
in Fig. 10. If started with a high enough field strength and
reduced slowly, this procedure results in a properly demag- Also called an incremental penneabilitfj, this quantity
netized sample. In practice, demagnetization is performed depends on the location of the minor curve on the hysteresis
by applying an alternating current field and gradually and decreases as the normal magnetization increases. The
. reducing its amplitude to zero. Complete demagnetization slope of minor loops is always smaller than that of the major
is usually a velY time consuming process. In practical loop at a given pOint. Thus, the incremental permeability is
situations, it is usually limited to reducing the flux density to lower than the normal permeability at any poin.t on the
some acceptable level. . hystereSiS curve. As the material approaches saturation, the
relative incremental permeability approaches unity.

Minor Hysteresis Loops

It often happens while a sample is at some operating
Hysteresis Curve as a Classifier
.point on the hysteresis curve (either on the initial mag-
When applying electromagnetic fields, it is necessary to
netization curve or on the outer loop) that a relatively
distinguish between applications, specialties and frequency
small change in magnetization occurs. An .example of this is
ranges. For example, electromagnetic nondestructive test-
ing is classified as a discipline separate from palaeomag-
FIGURE 10. Demagnetization of ferromagnetic netism (terrestrial magnetism), even though exactly the
materials same principles are involved and, often, the same methods
are used. Moreover, within each discipline different appli-
cations are distinguished.
B

FIGURE 11. Major and minor hysteresis loops

B

--------~~~~--------~H

------------~--------~------------~H
122 / MAGNETIC PARTICLE TESTING

In nondestmctive testing, active leakage field, residual

leakage field, eddy current and other electromagnetic phe- FIGURE 12. Classification of testing methods:
nomena are used. This distinction helps focus the treatment (a) active leakage fields (direct current);
of different problems. Often, the distinction parallels that of (b) residual leakage fields; and (c) alternating
the various areas of electromagnetic fields: active leakage current operation
fields are associated with magnetostatics; residual leakage
fields with source-free magnetostatics; and eddy currents
with steady state alternating current fields. fa} 8
It is far more practical to distinguish between the various
applications based on the point of operation on the hyster-
esis curve. This offers a visual description as well as some
physical insight into the application.
Active leakage field methods are those that employ the
H
initial magnetization curve (Fig. 12a). The point on the
----------~--------~--
initial magnetization curve is obtained by increasing the
current from zero to some predetermined value. It is
possible to apply a field to an initially magnetized sample
but this is usually not done because of the difficulty in
determining 'the exact operating point.
Residual leakage fields are obtained when an active fb}
excitation is removed and the operating points of the 8
material are allowed to relax into the second quadrant
(Fig. 12b). Similarly, alternating current leakage methods
may be defined as those that employ a normal hysteresis
curve. The operating point is on the major loop (Fig. 12c).
Eddy current methods require alternating current exci- H.
tation but this is usually very low. hi terms of the hysteresis
curve, it may be said that the operating point is at the origin
~lthough small hysteresis loops are described around the
origin as in Fig. 12c. .

fe}
Energy Lost in a Hysteresis Cycle 8

The energy stored in the magnetic field is given as a

volume integral of an energy density w:

w= f
v
wdv (Eq. 100) H

After integrating over the hysteresis curve or over any

pali of it, the area under the curve may be written as:

w ~ r HdB (Eq. 101)

The units of this integral are those of a volume energy

density and, under linearity assumptions (dB = dlI), the
energy density becomes tV = J,LH2/2. If this is then integrat-
ed over the volume of material in which the magnetic field
w ~J
v
(f
()
lidB )dv (Eq. 102)

exists, the total work done by external sources can be written The fact that energy is transformed in the process
as: becomes apparent by consideIing that work needs to be
 BASIC ELECTROMAGNETISM / 123

tformed in order to change the magnetic Held in the

Plume of a material. The expression in Eq. 102 is the work eddy current generated for any particular situation but the
dO ne for a complete cycle over the hysteresis loop. If the relative quantity involved is easy to obtain.
fi~ldchanges at a certain frequency, the energy per cycle in For any conductor, the electric field due to the induced
Eq. 102 must be multiplied by the frequency to obtain: electromagnetic force is directly proportional to the mag-
netic Held as:

(Eq. 103) - dB _
E oc - oc fB
(Eq. 105)
elt
This equation is exact but of limited use because it
requires integration over the hysteresis loop. Being a com- The dissipated energy due to heating losses (I2R) is
plex function and in many cases only known expelimentally, related to the square of the electric field. In terms of
the hysteresis loop is difficult to integrate .. For practical Eq. 105 and the magnetic Held, this becomes:
purposes, an approximate expression in terms of the maxi-
mum induced flux denSity Bmax is often used. The dissipated (Eq. 106)
power is:
This relation clearly indicates that the losses due to eddy
currents can be very large, especially for large flux densities
(Eq. 104) and higher frequencies. Eddy current losses may be re;.
duced by: (1) redUcing the conductivity of the materials
This expression is credited to Charles Steinmetz and involved (ferrites); (2) by special allOying to produce very
assumes that the constant TJ is known. It ranges from 0.001 narrow hysteresis curves (silicon steels); and (3) by breaking
for silicon steels to about 0.03 for hard steels. It is an the eddy current paths (laminated cores). ..
experimental value and the equation is only correct for The total losses in magnetic materials due to hysteresis
relatively large saturation fields (above 0.1. T). For low and eddy current losses can be summarized in terms of th~
saturation flux denSities, the equation cannot be' used. actual field as: . .

Eddy Current Losses (Eq. 107)

In addition to hysteresis losses, the change in flux denSity In terms of the saturation flux denSity Bmillo total losses may
be written as:
inside conducting materials generates induced electromag-
netic forces in those materials. The existence of this elec-
tromagnetic force, and the relatively large conductivity of Pd = k 1 fBl.6 + k2 f2 B~1ilX
~~st metals, results in a relatively large current flOwing max p (Eq. 108)
mSlde the material in a path that is a mirror image of the
SOurce generating the field. It is difficult to calculate the The constants k, kl and k2 depend on geometry as well as on
material properties.
124 / MAGNETIC PARTICLE TESTING

PART 6
CHARACTERISTICS OF
ELECTROMAGNETIC FIELDS

p ExT! (Eq. l12)

Energy in the Electromagnetic Field
In order to examine the energy in a magnetic field it is The advantage of such an expression is that it also
convenient to look first at the general time dependent indicates the direction of the energy flow, information that
expression for energy. This expression includes stored mag- is important for wave propagation calculations.
netic energy, stored electric energy and dissipated energy. The first term on the right side of Eq. III represents the
The following vector identity is used: time rate of increase in the potential or stored energy in the
system. It has two components: the stored electric energy
v e (f X H) = H e V X E - EeV X H (Eq. 109) and the stored magnetic energy. These 'energy densities
reduce to simpler expressions for the static electric and
Into this expressi9n, a ~ubstitution is made: the expres- magnetic fields.
sion for the curl of E and H from Maxwell's. first, and second
equations. EE2
We (Eq.1l3)
2
V e (E X H)
-aB
H- (Eq. l10)
at 2
Wm
= p,H (Eq. 114)
- aD -- 2
Ee--EeJ
at
Assuming the energy flow in a volume v bounded by an The second term at the right side of Eq. III is the power
area s, it is then possible to integrate this expression over the diSSipated and the power due to sources that may exist in the
volume v. Before this, transform the left side from a volume volume v. If no such sources exist, this term represents
integral to an area integral using the diverge~1Ce theorem ohmic losses.
(Eq. 168). The Poynting theorem describes all of a system's energy
relations: electrostatic, magnetostatic or time dependent.

fs
(f X H) - dS _~f(H' If
at v 2
(Eq. lll)
Because the cross product between the electric field and the
magnetic field is taken, these two quantities must be related,
otherwise the results have no meaning.

+E
-2e
- 0) dv - f--
vEe Jdv
The expression in Eq. III is an instantaneous quantity.
For practical purposes, a time averaged quantity is more
useful. This can be done by averaging over a time T (usually
a cycle of thc alternating current fIeld).
The left side of the expression represents the total flow of
energy thr~gh the area bounding the volume. The expres-
sion E X H is a power density with units of W-m- 2 . The Pav = T1 iTP(t)dt
0 (Eq. 115)
power density is called a Poynting vector.
 BASIC ELECTROMAGNETISM I 125

Forces in the magnetic field may also be expressed i?

· Force in the Magnetic Field terms of the energy stored in the magnetic field. A system s
mechanical work is done at the expense of its potential
The force in the magnetic field is governed by Lorenz' energy, so that:
force equation given in, Eq. 9=-For tl~'pUl"p?ses here, the
I ctJic force (Coulomb s law) Fe = qE IS not Important and F • de = - (V w) • M (Eq. 122)
e e removed from Lorenz'
was .
equation. Th·· f'£ t
IS III ~ tee assumes The force due to this reduction in the stored energy is
that the charge q only experiences a magnetIc force: therefore:
q(v x Ii) (Eq. 116) F = -Vw (Eq. 123)
This expression for the force is pmticularly convenient
v
Here, is the velocity of the charge. when the actual current distributions are not known or are
While forces on charges are impOltant in themselves, the
too complicated to permit calculation of the flux densities of
force on current canying conductors is more important in each current separately.
conjunction with magnetic fields. If it is assumed that an Because the magnetic field energy may be expressed in
element of conductor dC, with a cross sectional area s carries
terms of inductances (see Eq. 131), the force in the
N charge particles per unit volume moving with an average
magnetic field may also be expressed in terms of inductanc-
velocity v, then the magnetic force that this conductor es. Thus, for example, the force between two conductors
experiences is:
carrying currents II and 12, having inductances L 1 , L2 and
mutual inductance L12 can be written as:
dF = NqslvlM x B (Eq. 117)
F = 11 12 (VL 12 ) (Eq. 124)
Since Nqsu is the total current in the conductor, the
magnetic force becomes: The stored energy is calculat~~ as:

dF = llI1 x B (Eq.118) W = "21 LI1I2 + L121112 +"21 L 2 122· (Eq. 125)

In order to obtain the force due the complete conductor,

integration is taken over the length of the conductor.
Torque in the Magnetic Field
(Eq. 119) "The torque on a current carrying system may be calcu-
lated by using the definition of torque: the product of force
Another impOItant consideration is that of the force and the moment arm length. For simplicity, consider the
exerted on a current carrying conductor due to the field of
a second conductor. This is treated by assuming that the
field B is due to one conductor. If there are two conductors,
FIGURE 13. A rectangular loop and the direction
the force on conductor-l due to the field of conductor-2 is:
of the forces and torque on the rotating loop

(Eq. 120)

Similarly, the force on conductor-2 due to conductor-l is: B

m _

(Eq. 121)

In these equations, the integration is assumed to be over

the entire (closed) path of the currents. This is not merely a
COnvenience but is required to ensure that the forces F12
~nd F 21 are equal and opposite in direction. In other words,
Ih~egration over paIt of the contours will violate N ev.rton' s
t ud law (action and reaction forces).
126 I MAGNETIC PARTICLE TESTING

square loop in Fig. 13. If this definition is used, the torque conductors or turns in the system. For a simple solenoid,
is equal to; this is defined as;
T= s1sB sin () (Eq. 126) A = N<I> = NIB. ds (Eq. 129)
s
where s is the area of the loop. The product of current and
area is defined as a magnetic moment m. In effect, this includes only the flux that passes through
the center of the solenoid. The integration is over the cross
m = t; (Eq. 127) sectional area of the solenoid. A more complicated defini-
tion can be used, one that includes flux linkages that do not
The magnetic moment has a direction normal to the area link all the conductors, but this has little practical use
s and in the direction described by the right hand rule. because of the difficulties in calculation.
Thus, torque is a vector quantity and can be written as; If the system under consideration is linear, the field is
directly proportional to the current and the inductance of a
T=mxB (Eq. 128) system (coil) may be defined as the ratio of the flux linkage
and the current. The unit of measure for inductance is the
henry (H).

L = NfsB. as (Eq. 130)

Inductance 1
The inductance in Eq. 130 can be calculated, provided
Inductance is a property of the arrangement of conduc- that the flux linkages can be obtained. In many practical
tors in a system. It is a measure of the flux linked within the situations, it is more convenient to use an energy relation.
a measure of the magnetic energy stored in the system of
conductors. Flux linkage is defined as the flux that links the 12
whole system of conductors, multiplied by the number of. W = L- (Eq. 131)
2
"""1"'1""1111"'1"1"""1'111"1"""'1'1"1'""1""""'1,,""""""'1""""1'"
BASIC ELECTROMAGNETISM / J27

PART 7
MODELING ELECTROMAGNETIC FIELDS

Modeling of Leakage Fields FIGURE 14. Magnetic leakage field over distance
for faJ the tangential and normal (perpendicular)
Modeling of leakage fields can take many forms. I~ldeed, components of the leakage field from a
any attempt to explain a phenomenon may constItute a cylindrically shaped discontinuity (the
model. It is custommy to refer to a model as a process by numerical quantities shown are for P » Po'
which a particular problem or set of problems may be solved a = 1 mm, h = 1 mm and Ho = 6.4 kA·m- 1 J: and
and, with the solution, gain some insight into the processes fbJ the geometry of the test arrangement
and interactions involved. The more general the model, the
more useful it becomes and the wider its application is likely
to be. fa)
Because of the impoltance of leakage fields in a variety of
applications, including magnetic particle testing, their mod-
eling has rec,eived considerable attention. It is important to 20~
remember that an air gap, regardless of application, gives
rise to leakage fields. Modeling of leakage fields comprises
three categories: (1) experimental or empirical,modeling;
(2) analytical n'lodeling; and (3) numerical modeling.
When a process is classified, there is a temptation to
assign values,to different methods. No such claims are made
here. The various methods of modeling are presented and
only their usefulness is stressed. .
-10
Nature of the Leakage Field
-4 -2
As a first approximation, leakage fields around discontin- DISTANCE
uities may be considered to be superpositions of dipolar millimeters

fields. It is therefore of interest to p'resent a simple model

that illustrates the nature of such a field. Consider a Ib)
discontinuity in the shape of a long cylinder with its axis
parallel to the surface, its center a distance h below the
surface and with the applied magnetic field strength H0
pa~ll~l to the surface and perpendicular to the cylinder axis.
ThIS sItuation is illustrated in Fig. 14.
The discontinuitv is assumed to be in a material with a
perm~ability of J.ll ~ The leakage field at the surface above
the discontinuity is the point of interest. It has been shown 6
~hat the leakage field is closely approximated by a dipole
leld. The tangential and normal (perpendicular) compo-
nents of the field at the surface can be written as:

(Eq. 132)

H.L (Eq. 133)

128 I MAGNETIC PARTICLE TESTING

Where: A small magnetic particle of approximately spherical

shape can be cha~cterized by diameter d, density p and a
x = the distance along the surface in the direction of dipole moment m. The energy "V of this dipole in a
the applied field; and magnetic field strength H is given by:
m = is a magnetic dipole moment (per unit length).
(Eq. 135)
Assuming that p. « /-to, then m is given by:
The magnetic force on this dipole is the negative gradient
(Eq.134) of the energy. Consider two cases: (1) a permanently
magnetized particle with a moment that does not vary with
The shape and magnitude of these field components are position; and (2) a permanently magnetiz~ particle with a
illustrated in Fig. 14. Note that the tangential component of moment proportional to the field strength H. The latter case
the field is a maximum just above the discontinuity, whereas is more characteristic of magnetic particles and the force
the perpendicular component is zero at the same location. can be written as:
Although the leakage field illustrated in Fig. 14 is for an
idealized case, a linear magnetic material and a simple
geometrical shape for the discontinuity, real leakage fields
F = rxp.oV(H· V)R (Eq. 136)
have shapes that look like a distortion of the leakage field
shown in Fig. 14. Here, V is the volume of the particle and rx is a proportion-
The magnitude of the leakage field is nearly independent ality constant ~latiI~ the p~rticle magnetization and the
of permeability, provided that this is much greater than p.o. field strength m = Mv = rxHv.
While this is the case for nearly all ferromagnetic materials, If the particle has a low remanance and a nearly spherical
such materials are not linear. The effect of a nonlinear shape, then the field magnitudes of interest (its magnetiza-
magnetic material on the leakage field can be expl.ored tion) depend mainly on the demagnetizing factor N associ-
through experimentation or through numerical modeling. ated with its shape and are given by:
Large but slowly varying bias fields may be added to the
leakage fields. For example, the tangential field has the - 1-
added value Ho because. of the boundary conditions on the
M--H
. N
(Eq.137)
tahgential Hela component. For a nonparallel (to the sur-
face) field strength, the magnetizing field strength may also For a sphere, N = 1/3 and hence M = 3H. Using Eqs. 13.5
have a normal component resulting in a bias field in that and 137, the force on such a spherical particle directly above
direction. the cylindrical discontinuity shown in Fig. 14 has the
magnitude:

The leakage Field's Role in Formation of

Magnetic Particle Testing Indications (Eq. 138)

In magnetic particle testing, discontinuities are detected

by the indications formed when magnetic particles are The force calculated for a spherical particle is only a small
attracted to leakage fields caused by discontinuities. The fraction of the gravitational force (for particles with densi-
formation of such an indication is a complex process and has ties similar to iron) and isolated particles are not captured
thus far been difficult to model realistically. by the leakage field. A cooperative effect is reqUired, in
PaIiicles of the size used in magnetic particle testing are which the particles chain together to reduce the demagne-
affected in their motion by the following: (1) gravitational tization factor. If the particles chain together before they are
forces; (2) viscous forces; (3) leakage field forces; (4) in the vicinity of the leakage field, they may be too large to
exciting field forces; (5) image forces; and (6) interactions be affected by the leakage field, or they may be too strongly
between particles. attracted to their own image and may stick to the surface,
Viscous forces include those due to the motion of the forming a confusing background.
medium in which the particle is suspended. Image forces Because of these complexities, magnetic measurcments
are the attraction of the magnetized particle to its electrical on bulk powders by themselves give very little information
image in the material being tested. Particle interactions are regarding their suitability for magnetic palticle testing.
very important and depend on (1) particle shape; (2) mag- Many other factors must be considered and currently the
netization; (3) surface tension; and (4) the number of best test of such powders is to compare their ability to form
particles per unit volume. indications under actual test conditions.
 BASIC ELECTROMAGNETISM I 129

be solved analytically except for the most trivial situations.

Experimental or Empirical Modeling For this reason, simplified models have been sought which,
while simplifying the problem, still provide a reasonable
Experimenta~ modeling n:ethods are tho~e derived fro.m description of the actual behavior.
easurem ents 1I1 tests - eIther actual or sImulated. In Its Perhaps the most successful methods for representing
~lre form, this method of modeling constitutes a series of leakage fields are those that employ analogy: one situation
~xperiments designed to prove a point or to produce a has a simple solution and another situation is analogous to
product. In practice, the technique is nearly always paIt of that solution. The basis for this approach is in the analogy
a larger process and is used in conjunction with other between the electrostatic field in Eq. 17 and Eq. 19 and the
methods, p<uticularly analytical modeling. magnetostatic field in Eqs. 18 and 20. If a solution to an
Considering the difficulty of solving Maxwell's equations electrostatic problem is found, the solution to an analogous
for a nonlinear material, it is not surprising that experimen- agnetostatic problem can also be found. For the solution of
tal modeling is featured so prominently in analysis of leakage field problems (direct or alternating current), vari-
leakage fields and magnetic particle testing. Yet from a ations on the analogy between the electrostatic dipole
practical point of. vie,:, empirical metl,lOds have limitat~~ns . distribution and the actual magnetostatic field were sought
. As their name llnphes, the methods results are empmcal in the early period of field modeling.
and therefore applicable to the problem tested and perhaps Perhaps the earliest effoIts at using this equivalent
to a limited number of similar problems. Their extension to solution was an attempt to solve the field distribution of
other applications is not always possible or advisable. While recording heads. 7 Because the geometry included an air gap
a series of tests can be devised to model surface breaking over which a leakage field is generated, the same solution
cracks, the same model" cannot be used for subsurface applies to leakage fields from discontinuities in either active
. discontinuities or inclusions. or residual excitation. Th~ gap of the recording head
The value of such models is limited because of its (Fig. 15) is modeled by assuming a constant magnetic
empirical nature and the extension of such models to other potential difference across the gap. Assuming that this
geometries"is more or less speculative. This does not mean potential produces a constant flux density Bo in the gap, the
that accurate, controlled experimental data are not valuable magnitude of Bo is given oy IlVlb where 2b is the gap width
in modeling. Both analyti~al and especially immerical mod- and p., is the permeability in the gap. The flux denSity above
~ling rely on such data for confirmation. the gap can also be calculated as:
The limitatio~ of experimental methods are many and,
although very useful at times, it is not possible to rely solely 1 [ Ib+x (Eq. 139)
on experimental methods for modeling. Bx. = - 1T Bo tan - --y-

+ tan- 1 --y-
b - x]
Analytical Modeling and
Anal)tical models are those derived from elementary field
and circuit theory relations. At the very base of this By =
1
21T Eo In
[t/l ++ (b +_(b X)2]
X)2 (Eq. 140)
approach is the fact that some simplifying assumptions can
be m~de about the testing environment. These assumptions
may mclude those that are satisfied \-vith few or no errors Similarly, the field in a magnetic layer above the gap can
!ike linemity in calculations. There are also assumptions that be calculated based on the assumptions outlined above.
Imply. large errors or even a modification of the test geom- A similar approach was introduced in an attempt to find
a direct equivalent magnetic dipole to an electrostatic
~try, mcluding symmetry considerations, boundalY condi-
tIons an.d discontinuity shape approximations. ' dipole. 8 ,9 This was done by assuming that opposite magnetic
In spite of extensive simplification, analytical models are charges are uniformly distributed on the opposing walls of a
extremely complicated and their results tend to be limited rectangular slot. Assuming a charge denSity s, the field at a
to a singl.e
- geometry or class of problems. point p (Fig. 15) is:
An~IYhcal models are attempts to solve the governing
71(x + b)
equa.hons directly. Solution of Maxwell's equations can Hx = 28 tan -1 ----;:------- (Eq. 141)
proVIde all the data and understanding necessary for mean- (x + b)2 + y(y + 71)
l~gful application of electromagnetic methods and for de-
b gn of instrumentation and tests. This ideal situation is
. adly hindered by the fact that Ma\.well's equations cannot
- 28 tan
-1 h(x - b)
----=-<J~--.-:.---
(x - b-) + y(y + h)
 130 / MAGNETIC PARTICLE TESTING

and

Hy = s In {[(x + b)2 + (y + h)2] 2s [ tan- 1 -X + b - tan- l _x - b]

(Eq. 142) y y (Eq. 143)
[(x - b)2 + y2]}
and
-s In {[(x + b)2 + y2] [(x _ b)2 + (y + h)2]}

In these equations, h is the depth of the discontinuity. As

_
H" - s In
J
[(X - b)2 +
(.t + lJ )"- + y_"
.1/] (Eq. 144)
the depth of the discontinuity approaches infinity, the
expressions in Egs. 141 and 142 become:
These relationships are only as good as the assumptions
behind them. If infinitely long planes are assumed \vith
charge distributed over them, then the solution is not
FIGURE' 5. Leakage field models: (al Karlqv;st capable of modeling short, tight discontinuities. In addition,
model of a recording head; and (bl dipole model these models exemplifY the basic limitation of analytical
of a slot
models: it is often necessary to assume an oversimplified
geometry (the infinite parallel planes) or even a nonphysical
assumption (the magnetic charge idea). This oversimplifica_
laJ y tion is at the root of two problems: (1) the inability to
adequately solve the field equations analytically; and in turn
(2) the inability to obtain an adequate model.
t Many other methods for the solution of magnetostatic
field equations exist and have found application. Because
• PIxy)
the magnetostatic field equation (Eq. 26) is either a Laplace
I equation for zero excitation or a Poisson equation for
nonzero eXcitation, the classical methods for 'the solution of
j!
~
-x Laplace and Poisson's equations can be used. It is also
reasonable to consider such well known solutions as the
separation of vaIiables, the method of images and conformal
transformations .. None of these .methods is satisfactory in
thege~eral sense. Separation of variables can only be t;sed
with simple geometries and boundaries,1O the method of
images is only applicable to highly symmehic problems, 10
while conformal transformations are strictly two-dimensional
methods that can be used for some simple boundaries.

IbJ y Numerical MOdeling

Numerical modeling is conceptually different from ana-
lytical modeling. For the purpose of modeling nondestmc-
f tive tests, the most important advantage in numerical
modeling is the' fact that lIone of the simplif)ling assump-
tions made in the analytical approach are necessary in order
to reach a satisfactOlY solution.
Stmting with Mcu:well's equations, it is possible to pro-
ceed in many different ways, each of which requires the
formulation of these equations in Some p,uticular form. 11,12
Assumptions are made for the sake of simplicity and
economics of the solution and not to render the equations
solvable. Thus, a hvo-dimensional or axisymmetric solution
may be assumed if tIle geometry is approximately two-
dimensional or has axisymmetIy. A two-dimensional solu-
tion is a special case of the more general three-dimensional
solution which, if needed, can be used. Similarly, an
axisymmetric solution is the solution of a three-dimensional
 BASIC ELECTROMAGNETISM I 131

problem in c:,jillllrical coordinates, Other <:ssumptions may technical: large computer resources, non closed form solu-
be employed in order to shorten the solutIOn process or to tion, and so on.
obtain insight into the problem before a more general Numerical modeling in nondestl11ctive testing is an out-
solution is attempted. A linear formulation may be used as growth of the failure of analytical models to reliably predict
a first approximation or in cases where the problem is the necessary field discontinuity interactions with any de-
gree of generality. A numerical model uses a digital com-
indeed linear.
As \\rith clJ1<llytical methods, there are many approaches puter to solve the governing equations directly without
and numerical methods available. Some are more accurate, simplifying assumptions and this is enough to explain the
more convenient, more economical or more suitable for the significance of such models for the solution of field prob-
solution of a particular problem. lems, including those related to active and residual leakage
The successful application of a numerical process for fields. It not only allows the solution of very complex
obtaining a model is determined by three preliminar), problems but at the same time does not require the user
decisions: (1) choosing the appropriate numerical method; to know the intricacies of electromagnetic theOl), or dif-
(2) making correct assumptions about the geometry and the ferential calculus. All that the user is required to do is input
nature of the solution; and perhaps most important the problem variables and, if necessary, verify the results
(3) resolving the question of economics. Accuracy can be experimentally.
obtained regardless of the problem's geometI)" linearity,
nonlinearity or dimensionality, pro\rided the necessary ex-
penses associated with such a general solution are justifiable
The Finite Difference Method
and that the computer resources are available.
A significant point and one of considerable practical The finite difference method has been traced back to
importance is the fact that the numerical solution to many Gauss and has mO~'e recently found Significant use as a
significant problems .does not require large computers or general method for the solution of partial differential
highly trained personnel, despite the prevalent notion. equations. There .are many reasons for its widespread
Desktop computers may be used for the solution of many application, espetially in the early days of computer mod-
two-dimensional' or a.xisymmehic problems. Some large eling. The method is not only general but relatively easy to
scale models or specialized three-dimensional problems still apply. It is equally applicable to direct current fields,
require supercomputers but it is reasonable to assume that quasistatic or transient fields and to linear and nonlinear
tomorrow's miJlicomputers will be more powerful than problems. In its simplest form, the formulation of the field
today's mainframes. The developments in computer tech- equations consists of simply replacing the partial derivatives
nology favor numerical modeling approaches. by appropliate difference formulas. A solutJon can then be
Numerical methods are generally far more powerful than obtained for the dependent variable at discrete points
analytical methods. At the same time, the solution is within the solution region either by an iterative process or
obtained as numerical data rather than as a closed form by the solution of a system of algebraiC equations, depend-
solution. As such, the result obtained for a particular ing on the finite difference formula choset:t.
p.ro?lem is not usable for the analysis of a different, perhaps The application of the finite difference method is com-
s~mllar, problem. Parameter study requires repetitive solu- plicated by problems of convergence and stability of the
tIon and it is not ah. .'avs possible to deduce it from the solution as well as by restrictions on the discretization
solution, as is the case ~'ith analytical solutions. This is not process. While regular sets of discretization points (grids)
a significant clis(l(h'antage becau~e the same repetitive pro- are easy to handle, irregular grids are not. Discretization of
cess allows study of parameters for which the analytical complex geometries into regular grids is not practical and
approach cannot he applied, including changes in the irregular grids may in some cases render the solution
geometry.of an arhitrclIily shaped discontinuity, nonconvergent. In field problems, the inability to properly
Numencal modeling can be \riewed as a special subset of discretize small areas (such as air gaps or discontinuities) is
the analytical methods because the goals are to solve the critical and is detrimental to the use of the method. In
same system of equations; the methods do differ in ap- addition, the finite difference method is in effect a nodal
proach. In anal)iical methods, there are attempts to simplify method and cannot take into account distributed parame-
tl~e ge~me.try or the equations. Numerical methods rely on ters such as current densities, conductivities and permea-
dIscretIzatIon of the solution space into subspaces, for which bilities. These have to be described as equivalent nodal
k go~d ap'pr~ximation to the solution can be obtained. At quantities with all the associated errors. The solution is only
I' a~t I~ pnnclple. the numerical methods suffer none of the valid at the nodal points.
ImI~atIOns of the anal)iical methods and are capable of Despite its limitations, the finite difference method has
s?lvmg the goveming equations Vvith few, if any, approxima- been applied successfully to a variety of electromagnetic
tions. The main difficulties with numerical methods are field problems. The problems related to leakage fields are
 '32 I MAGNETIC PARTICLE TESTING

those applications that deal with electrical machines. 13-16 In Being a discrete method, it requires discretization of the
those applications, the actual geometry, machine parame- solution region but no reshictions are imposed on the
ters, current densities and realistic material properties were shape, size or number of finite elements. 28 The solution
modeled. process and the formulation are identical, regardless of the
Nonlinearity of ferromagnetic materials is taken into size and shape of the elements used. \Nhile the finite
account by iteration methods 14 to produce results that no difference method assumes linear relations between the
analytical model can match either in accuracy or flexibility. unknowns, the finite element method can handle higher
While two-dimensional and axisymmetric models are suffi- order relations as wel1. 28 Problems with convergence have
cient in many applications, there are also three-dimensional no meaning in the context of finite elements.
models using the finite difference method. These models These factors are of particular importance for the simu-
can be used for alternating current applications as well. lation of electromagnetic nondesttuctive testing techniques
and accordingly the method has received considerable
The Finite Element Method attention. Numerical models based on the finite element
method have been developed for two-dimensionaP9-32 and
The finite element method has a briefer history than the three-dimensionaI27 ,33-36 eddy current applications.
finite difference method. It evolved in the late 1950s as a The advantages of the method come with a price. vVhen
numerical method in structural analysis 17 but has spread compared to finite difference methods, problems solved by
qUickly to become a major analysis tool in diverse areas of the finite element method generally require larger comput-
engineering,18-22 in the physical sciences and in medical er resources, especially for nonlinear and time dependent
research. 23 problems. In addition, the method does not lend itself well
Because of its success in modeling intricate geometries to the solution of transient problems because it cannot
efficiently and accurately, the potential for its application to efficiently handle time discretization.
electric and magnetic fields was recognized in the early These two methods are complementary, each being
1970s and has been applied with great success to the study suited to the solution of different problems. For example,
of direct current and low frequency electromagnetic fields time integration in finite element computer codes is usually
in electrical machines,24 large magnet structures 25 and handled by various forms of finite difference methods. 37
permanent magnet design. 26 The finite element method has The flexibility of finite element models has been demon-
considerable advantage over the finite diffe~ence. method, . strated by diverse applications to th~ solution of a variety of
including the ease of handling boundary conditions and the leakage field problems. Both active38 and residuaP9 appli-
ability to follow awkwardly shaped boundaries. ~7 cations have been handled successfully in linear or nonlinear
The method is by definition a volumetric method: various environments. The extension to three-dimensional prob-
parameters are defined by their association with the volume lems is relatively simple,27 although extensive computer
or, in the case of two-dimensional and axisymmetric formu- resources are required, especially for nonlinear problems.
lations, with the surface. The finite element method is Alternating current applications, either eddy current or
therefore naturally suited to the modeling of continuum general problems, have also been treated extensively in
problems and is quite flexible in terms of the discretization two-dimensional and axisymmetric geometries and in three-
process. dimensional geometries.
 BASIC ELECTROMAGNETISM I 133

PART 8
-MATHEMATICS OF ELECTROMAGNETIC
FIELD STUDIES

This can be interpreted as the total flux through an

Vector Algebra and Calculus elemental volume. It is therefore a measure of the outward
flow of flux per unit volume. The divergence can be written
The presentation and use of electromagnetic field rela- in terms of partial derivatives in Cartesian coordinates as:
tions requires the manipulation of vector and scalar quan-
tities. The following text attempts to define the necessalY
quantities, relations an? theorems. These should be viewed
" • A-
v =aax
Ax
- +a-
ay
+ adZ
Ay Az (Eq. 149)
'as a listing rather than as a mathematical treatment of any
depth. Fmther reference to sources on fields or vector
algebra and calculus <l;re encouraged. .
In cylindrical coordinates, the divergence is written as:
. . . .
V• A = 1:. ~ (rA: ) + aAp + aA z '(Eq.150)
r ar r ra 4> . az
The gradient of a scalar function U is defined as the
steepest slope of the function at a given point and is denoted
by VU where V is called the del operator. In Cartesian. In sphelical coordinate systems, divergence is written as:
coordinates'this is:
- 1 d (2 ) (Eq. 151)
V•A = R2 dR R AR
(Eq. 145)
1 a. 1 dAp
In cylindrical coordinates, the gradient is written as: + R sin (J ali (A 0 sm (J) + R sin (J a~
Curl
VU = r"au
- + ~"au
- + z"au
- (Eq. 146)
ar rd ~ az The gradient and the divergence define important quan-
tities relating to fields. There are however fields with special
In spherical coordinates, the gradient is written as: properties that cannot be defined in terms of these two
properties alone. The circulation of a turbulent flow field,
Vu _ RdU A dU " 1 au for example, has rotation properties that must be described.
- dR + (J Rd(J + ~ R sin 8 ~ (Eq. 147) Rotation of a vector field can be described by the circulation
of the vector.
Note that the gradient is only defined as operating on a
scalar function, The gradient of a vector is not defined.
c = fA. de
c
(Eq. 152)
Divergence
The divergence of a vector function is defined as: This definition relates the field to a contour around a
specified area. It is helpful to envision a small area around
which this contour integral can be evaluated. If a differential
(Eq. 148) form of the circulation is needed, a point circulation vector
134 I MAGNETIC PARTICLE TESTING

should be defined. This is done by taking a limiting pro£ess A· 13 x C C· A x If (Eq. 157)

as the area s tends to zero. The curl of the vector A is
defined as th~ vector whose magnitude is the maximum A x (8 x C) == 13 (A • C) - C(A • B) (Eq. 158)
circulation of A per unit area:
V(c/>V) = c/>'VV + VVc/> (Eq. 159)
v x A == lim~s-+o ;s [n £ A • de ] ma.x (Eq. 153)
V • (c/>A) = </>'V. A + A· 'Vc/> (Eq. 160)

In terms of partial derivatives in Cartesian coordinates, this V x (c/>A) = </>( V x A) + V </> x A (Eq. 161)
is written as:
V• (A x B) = 13· (V x A) - A· (V x B) (Eq. 162)
T7
v
X -
A=x (d-
A:;
A

- dAy)
- oy oZ (Eq. 154)
(Eq. 163)

+y. . (d- A:::) +Z. . (a-

Ax- d- Ay- a-
Ax) V X V x if = V(V • A) - V 2X (Eq. 164)
az ax ax ay
V x V</> = 0 (Eq. 165)
In cylindrical coordinates, it is written as:

V • ('V x A) = 0 (Eq. 166)

(Eq. 155)
These identities are used to put an expression in a
;;.(aA
+ 0/ -
r - aA~) + z. . -l[a- (rA.p ) _ aaAc/>r]
--
particular form or to simplify the expression. For example,
a vector triple product may be expressed as the sum of the
. dZ dr . r ar
gradient of the divergence of a vector K minus its vector
In sphe.rical coordina~e systems, tt is written itS: Laplacian (see Eq. 164). .

v x A::: R __l _ (Eq. 156)

R sin 0
Vector Theorems
d - _ dAoJ
[ ac/> (A.p sin 0) ac/> Like any other mathematical discipline, many important
theorems can be derived from the basic properties of
+ {) ~[_1_ aA R _ ~ (RA )]
vectors and scalars. There are three theorems that are
R sin 0 a </> aR .p particularly important in field problems.
Stokes' theorem relates the line integral of a vector with
the surface integral of its curl. It is an impOltant tool that
+ 4>.l[~ (RA ) - OAR] allows the use of simpler line integrals whenever the
R aR 0 ao theorem can be applied.
In these relations ~, y, z, ;,
R, {) , ¢ , are the unit vectors in
the x, y, z, r, R, 8, c/> directions respectively.

J(V
s
x A) • ds = fc
X • de (Eq.167)

Vector Identities
The divergence theorem relates the volume integral of
In the following identities, all symbols without a bar are the divergence of a vector to the closed surface integral (or
scalars while those with a bar are vectors. The operator is a flux) of the vector.
vector. The identities listed here are those that are most Its impOltance lies in its abilities for Simplification and
often used. derivation of field relations.
1"""""""""""""""""""""""""'""~"~'ll"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~'"~ BASIC ELECTROMAGNETISM I 135

specified. Thus, the divergence specifies the solenoidal

(Eq. 168) properties of the field while the curl specifies its rotational
properties. A field is said to be solenoidal if its divergence is
nonzero and rotational if its curl is nonzero. Similarly,
Helmholtz' theorem states that a vector field is defined nonsolenoidal and rotational fields can be defined, or fields
(within a constant) if both its divergence and its curl are with any combination of the two properties.
136 I MAGNETIC PARTICLE TESTING

REFERENCES

1. Morse, P.M. and H. Feshbach. Methods of Theoretical 14. Erdelyi, E.A. and E.F. Fuchs. "Nonlinear Magnetic
Physics. New York, NY: McGraw-Hill Book Company Field Analysis of Direct Current Machines: Applica-
(1953). tion of the Improved Treatment." Transactions on
2. Jackson, J.D. Classical Electrodynamics, second edi- Power Apparatus and Systems. Vol. PAS-90. New
tion. New York, NY: John \-Viley and Sons (1975). York, NY: Institute of Electrical and Electronics En-
3. Abraham, M. and R. Becker. The Classical Theory of gineers (1970): p 1,555-1,564.
Electricity and Magnetism. New York, NY: Hafner 15. Demerdash, N.A., H.B. Hamilton and C.W. Brown.
Publishing Company (1951). "Simulation for Design Purposes of Magnetic Fields in
4. Clemnow, P.C. An Introduction to Electromagnetic Turbogenerators with Symmetrical and Asymmetrical
Theory. New York, NY: Cambridge University Press Rotors: Model Development and Solution Technique."
(1973). Transactions on Power Apparatus and Systems. Vol.
5. Elliot, R.S. Electromagnetics. New York, NY: McGraw- PAS-91. New York, NY: Institute of Electrical and
Hill Book Company (1966). Electronics Engineers (1972): p 1,985-1,992.
6. Shcherbinin, V.E. and M.L. Shur. "Calculating the 16. Demerdash, N.A., H.B. Hamilton and G. W. BroWn.
Effect of the Boundaries of a Product on the Field of "Simulation for Design Purposes of Magnetic Fields in
a Cylindrical Defect." Soviet Journal of NDT. Vol. 12 Turbogenerators with Symmetrical and Asymmetrical
(1976): p 606.
Rotors: Model Calibration and Applications." Trans-
7. Karlqvist, O. "Calculation of the Magnetic Field in the
, actions on' Power Apparatus and Systems. Vol. PAS-
. ' Ferromagnetic Layer of a Magnetic Drum.'" Transac- 91. New York, NY: Institute of Electrical and Elec-
tions of the Royal Institute of Technology. No. 80. tronics EngiJ:l.eers (1972): p 1,992-1,999.
Stockholm, Sweden (1954): p 1-6.
8. Zatsepin, N.N. and Y.E. Shcherbinin. "Calculation of 17. Zienkiewicz, O.c. The Finite Element Method in
Engineering, third edition. London, England:
the Magnetostatic Field of Surface Defects I: Field
McGraw-Hill Book Company (1977).
Topography of Defect Models." Defectoskopiya. Vol. 5
(September-October 1966): p 50-59. 18. Desai, C.S. and J.F. Abel. Introduction to the Finite
9. Zatsepin, N.N. and V.E. Shcherbinin. "Calculation of Element Method. Van Nostrand Reinhold Publishing
the Magnetostatic Field of Surface Defects II: Exper- (1972).
imental Verification of the Principal Theoretical Rela- 19. Sabir, A.B. "Finite Element Analysis of Arch Bridges."
tionships." Defectoskopiya. Vol. 5 (September-October Finite Elements for Thin Shells and Curved Members.
1966): p 59-65. D.G. Ashwell and ·R.H. Gallagher, eds. London,
10. Binns, K.J. and P.J. Larenson. Analysis and Compu- England: John \Viley and Sons (1974): p 223-242.
tation of Electric and Magnetic Field Problems. New 20. Bruch, J.e. and G. Zyroloski. "Transient Tvvo-
York, NY: Macmillan (1963). Dimensional Heat Conduction Problems Solved by
11. Brown, W.F. Magnetostatic Principles in Ferromag- the Finite Element Method." International Journal
netism. New York, NY: Interscience (1962): p 12-29. for Nwnerical Methods in Engineering. Vol. 8, No.3
12. Guancial, E. and S. DasGupta. "Three-Dimensional (1974): p 481-494.
Finit~, Element Program for Magnetic Field Prob- 21. Ie Provost, C. and A. Poucet. "Finite Element Method
lems. Transactions on MagnetiCS. Vol. MAG-13, No.3. for Spectral Modeling of Tides." International Journal
New York, NY: Institute of Electrical and Electronics for Numerical Methods in Engineering. Vol. 12, No.5
Engineers (May 1977): p 1,012-1,015. (1978): p 853-872.
13. Erdelyi, E.A. and E.F. Fuchs. "Nonlinear Magnetic 22. Arlet, P.L., et al. "Application of Finite Elements to
Field Analysis of Direct Current Machines: Theoreti- the Solution of Helmholtz' Equation." IEEE Proceed-
cal Fundamentals." Transactions on Power Apparatus ings. Vol. 115, No. 12. New York, NY: Institute of
and System'). Vol. PAS-89. New York, NY: Institute of Electrical and Electronics Engineers (December
Electrical and Electronics Engineers: p 1,546-1,554. 1968).
 BASIC ELECTROMAGNETISM I 137

23. McNeice, C.M. and H.C. Amstutz. "Finite Element 32. Palanisamy, R. and W. Lord. "Finite Element Model-
Studies in Hip Reconstructions." Biomechanics. ing of Electromagnetic NDT Phenomena." Transac-
Baltimore, MD: University Park Press (1976): p 394- tions on Magnetics. Vol. MAG-IS, No.6. New York,
405. NY: Institute of Electrical and Electronics Engineers
24. Demerdash, N.A., H.B. Hamilton and C.W. Brown. (November 1979): p 1,479-1,481.
"Simulation for Design Purposes of Magnetic Fields in 33. Hodgkins, W.R. and J.F. Waddington. "The Solution
Turbogenerators with Symmetrical and Asymmetrical of Three-Dimensional Induction Heating Problems
Rotors: Model Development and Solution Technique." Using an Integral Equation Method." Transactions on
Transactions on Power Apparatus and Systems. Vol. Magnetics. Vol. MAG-18. New York, NY: Institute of
PAS-91. New York, NY: Institute of Electrical and Electlical and Electronics Engineers (1982): p 476-
Electronics Engineers (1972): p 1,985-1,992. 480.
25. Simkin, J. and C.W. Trowbridge. Three-Dimensional
Computer Program (TOSCA) for Nonlinear Static 34. Rodger, D. and J.F. Eastham. "A Formulation for LoV(
Electromagnetic Fields'. HL 7,9/097. Hutherford Labo- Frequency Eddy Current Solutions." Transactions on
ratOlY (1979). Magnetics. Vol. MAG-19, No.6. New York, NY:
26. Rafinejad, P., J.L. Coulomb and C. Meunier. "Penna- Institute of Electrical and Electronics Engineers
nent Magnet Three-Dimensional Field Computation (1983): p 2,443-2,446.
by Curvilinear Finite Elements." Proceedings of 35. Pillsbury, R.D. "A Three-Dimensional Eddy Current
COMPUMAG Conference. Grenoble, France (1978). Formulation Using Two Potentials: the Magnetic Vec-
27. Ida, N. Three Dimensional Finite Element Modeling of tor Potential and Total MagnetiC Scalar Potential."
Electromagnetic Nondestructive Testing phenomena. Transactions on Magnetics. Vol. MAG-19, No.6. New
Ph.D. thesis. Fort Collins, CO: Colorado State Uni- York, NY: Institute of Electrical and Electronics En-'
verSity (Spling 1983). gineers (1983): p 2,284-2,287.
28. Huebner, K.H. The Finite Element Method for Engi- 36. Ida, N. and W. Lord. "A Finite Element Model fOJ:
neers. New York, NY: Wiley Interscience, John Wiley Three.-Dimensional Eddy Current NDT Calculation~."
and Sons (1975). •. . . . Transactions on Magnetics. New York, NY: Institute of
29. Palanisamy, R. and W. Lord. "Finite Element Model- Electrical and Electronics Engineers (November.
ing of Electromagnetic NDT Phenomena." Transac- 1985).
tions on Magnetics. Vol. MAG-IS, No.6. New York,
37. Wood, W.L. "A Further Look at Newmark, Houbolt,
NY: Institute of Electrical and Electronics Engineers
. (November 1979): p 1,479-1,481. etc., Time Stepping Formulae." International Journal
30. Palanisamy, R. and W. Lord. "Finite Element Analysis of Numerical Methods in Engineering. Vol. 20, No.6
. of A..'{isymmetric Geometries in NDE." Proceedings of (1984): p 1,009-1,018 .
the ARPNANFL Review of Progress in Quantitative 38. Lord, W. and D.J. Oswald. "Leakage Field Methods in
NDE. La Jolla, CA (July 1979): p 25-32. Defect Detection." International Journal of NDT.
31. Palanisamy, R. Finite Element Modeling of Eddy No.4 (1972): p 249-274.
. Current Nondestructive Testing Phenomena. Ph.D. 39. Satish, S.R. and W. Lord. Finite Element Modeling of
thesis. Fort Collins, CO: Colorado State University Residual Magnetic Phenomena. Boston, MA: Intermag
(1980). Conference (1981).
t 38 I MAGNETIC PARTICLE TESTING

BIBLIOGRAPHY

See Electromagnetic Testing: Eddy Current, Flux Leakage 15. Bradshaw, M.D. and W.J. Byatt. Introductory Engi-
and Microwave NDT (Volume 4 of the Nondestructive neering Field Theory. Englewood Cliffs, NJ: Prentice-
Testing Handbook, second edition) for additional informa- Hall (1967).
tion on electromagnetic testing techniques and supplemen- 16. Cmter, C.W. The Electromagnetic Field in Its Engi-
tal coverage of electromagnetic phenomena. neering Aspects. New York, NY: Amelican Elsevier
Publishing Company (1967).
Entry level 17. Cheston, W.B. Elementary Theory of Electric and
Magnetic Fields. New York, NY: John vViley and Sons
1. Plonus, M.A. Applied Electromagnetics. New York, (1964).
NY: McGraw-Hill Book Company (1978). 18. Holt, C.A. Introduction to Electromagnetic Fields and
2. Cheng, D.K..Field and Wave Electromagnetics. Read- Waves. New York, NY: John Wiley and Sons (1966).
ing, MA: Addison-Wesley Publishing Company (1985). 19. Paris, D.T. and F.K. Hurd. Basic Electromagnetic
3. Shen, L.c. and J.A. Kong. Applied Electromagnetism. Theory. New York, NY: McGraw-Hill Book Company
Monterey, CA: Brooks/Cole Engineering Division (1969). .
(1983). 20. Silvester, P. Modem Electromagnetic Fields. Engle-
4. Neff, H.P., Jr. Basic Electronwgnetic Fields, second wood Cliffs, NJ: Prentice-Hall (1968).. .
edition. New York, NY: Harper and Row Publi~hing 21. Tamboulian, D.H. Electiic and Magnetic Fields. New
Company (1987). ' . York, NY: Harcourt, Brace and World (1965) .
.S. Lorain, P. and D.R Corson. Electromagnetism. San 22. Walsh, J.B. Electromagnetic Theory and Engineering
Francisco, CA: W.H. Freeman (1978). Applications. New York, NY: The Ronald Press Com-
6. Hyatt, W.H. Engineering Electromagnetics, third edi- pany (1960).
tion. New York, NY: McGraw-Hill Book Company 23. Whitmer, R.M. Electro magnetics, second edition.
(1974). Englewood Cliffs, NJ: Prentice-Hall (1962).
7. Kraus, J.D. and K.R Carver. Electromagnetics, second
edition. New York, NY: McGraw-Hill Book Company
Intermediate level
(1973).
8. Magid, L.M. Electromagnetic Fields, Energy and 24. Plosney, Rand R.E. Collin. Principles and Applica-
Waves. New York, NY: John Wiley and Sons (1972). tions of Electromagnetic Fields. New York, NY:
9. Feynman, RP., R.O. Leighton and M. Sands. Lectures McGraw-Hill Book Company (1961).
on Physics. Vol. 2. Reading, MA: Addison-vVesley 25. Lorain, P. and D.R. Corson. Electromagnetic Fields
Publishing Company (1964). and "Vaves, second edition. San Francisco, CA: W.H.
10. King, H.W.P. Fundamental Electromagnetic Theory. Freeman and Company (1970).
New York, NY: Dover Press (1963). 26. Shelkunoff, S.A. Electronwgnetic Fields. New York,
11. Paris, D.T. and F.K. Hurd. Basic Electrornagnetic NY: Blaisdell Publishing Company (1963).
Theory. New York, NY: McGraw-Hill Book Company 27. Shadowitz, A. The Electromagnetic Field. New York,
(1969). NY: McGraw-Hill Book Company (1975).
12. PopOvic, B.D. Introductory Engineering Electromag- 28. Bolm, E.Y. Introduction to Electromagnetic Fields and
netics. Reading, MA: Addison-Wesley Publishing Com- Waves. Reading MA: Addison-\Vesley Publishing
pany (1971). Company (1968).
13. Purcell, E.M. Electricity and Magnetism. New York, 29. Della Torre, E. and C.Y. Longo. The Electromagnetic
NY: McCraw-Hill Book Company (1963). Field. Boston, MA: Allyn and Bacon Publishing (1969).
14. Seshadli, S.H. Fundamentals of Transmission Lines 30. Durney, C.H. and C.C. Johnson. Introduction to
and Electromagnetic Fields. Reading, MA: Addison- Modem Electromagnetlcs. New York, NY: McGraw-
Wesley Publishing Company (1971). Hill Book Company (1973).
 BASIC ELECTROMAGNETISM I 139

31. Fano, R..tvL, L.J. Chu and RE. Adler. Electromagnetic 51. Hartington, RF. Time-Harmonic Electromagnetic
Fields, Energy and Forces. New York, NY: John \Viley Fields. New York, NY: McGraw-Hill Book Company
and Sons (1960). (1961).
32. John, c.T. Engineering Electromagnetic Fields and 52. Johnson, e.C. Field and Wave Electrodynam.ics. New
\Vaves. New York, NY: John Wiley and Sons (1975). York, NY: McGraw-Hill Book Company (1965).
33. Moon, P. and D.E. Spencer. Foundations of Electro- 53. Langmuir, RV. Electromagnetic Fields and Waves.
dynmnics. Ptinceton, NJ: D. Van Nostrand Publishing New York, NY: McGraw-Hill Book Company (1961).
(1960). 54. Bladel, J. Van. Electromagnetic Fields. New York, NY:
34. Nussbaum, A. Electromagnetic Theon) for Engineers McGraw-Hill Book Company (1964).
and Scientists. Englewood Cliffs, NJ: Prentice-Hall 55. Weeks, W.L. ElectronUlgnetic Theory for Engineering
(1965). Applications. New York, NY: J olm Wiley and Sons
35. Maxwell, J.C. A Treatise on Electricity and Magnetism. (1964).
Vols. 1 and 2. New York, NY: Dover Press (1954).
36. O'vveD, G.E. Introduction to Electromagnetic Theory. Waves
Boston, :MA: Allyn and Bacon Publishing (1963). 56. Crawford, F.S., Jr. Waves. New York, NY: McGraw-
37. Popovic, B.D. Introductory Engineering Electromag- Hill Book Company (1965).
netics. Reading, MA: Addison-Wesley Publishing Com- 57. Kong, J.A. Theory of Electromagnetic Waves. New
pany (1971). York, NY: John Wiley and Sons (1975).
38. Ramo, S., J.R. Whinnery and T. Van Duzer. Fields and 58. Towne, D.H. Wave Phenomena. Reading, MA:
Waves in Communication Electronics. New York, NY: Addison-Wesley Publishing Company (1967).
John Wiley and Sons (1965). 59. Skilling, H.H. Fundamentals of Electric Waves. New
39. Rao, N.N. Elem.ents in Engineering Electromagnetics. York, NY: John Wiley and Sons (1964).
Englewood Cliffs, NJ: Prentice-Hall (1977). 60. Dearholt, D.W. and W.R McSpadden. Electromag-
40. Reitz, Rand F.J. Milford. Foundations of Electro mag- netic Wave Propag(l~ion. New York, NY: McGraw-Hill
netic Theory. Reading, MA: Addison-Wesley Publish- Book Co.mpany (lS73): .
. ing Company (1960). • 61. Jordan, E.C. and KG. Balmain. Electromagnetic
. 41. Scott; \\1. T. The PhtJsics of Electricity and Magnetism, Waves and Radiating Systems. Englewood Cliffs, NJ:
. second edition. New York, NY: John Wiley and Sons Prentice-Hall (1968).
(1966). .
62. Kong, J.A. Theory of Electrornagnetic Waves. New
York, NY:. John Wiley and Sons (1975).
Advanced Level
42. Stratton, J.A. Electromagnetic Theory. New York, NY: Vector Calculus
McGraw-Hill Book Company (1952). 63. Boast, W.B. Vector Fields. New York, NY: Harper and
43. Panofsky, W.KH. and M. Phillips. Classical Electricity Row Publishing (1964).
and Magnetism. Reading, MA: Addison-Wesley Pub- 64. Bowen, RM. and C.C. Wang. Introduction to Vectors
lishing Company (1956). and Tensors. New York, NY: Plenum Press (1976).
44. Smythe, W.R. Static and Dynamic Electricity. New 65. Wolstenholme, E. Elementary Vectors, second edition.
York, NY: McGraw-Hill Book Company (1950). Oxford, England: Pergamon Press (1971).
45. Harrington, RF. Field Computation by Mom.ent 66. McDougle, P.E.V Vector Algebra. Belmont, CA:
Methods. New York, NY: Macmillan Publishing (1968). Wadsworth Publishing Company (1971).
46. Morse, P.M. and H. Feshbach. Methods of Theoretical 67. Merder, L. Vector Analysis. New York, NY: American
Physics. New York, NY: McGraw-Hill Book Company Elsevier Company (1970).
(1953).
47. Jackson, J.D. Classical Electrodynamics, second edi- Field Analysis
tion. New York, NY: John Wiley and Sons (1975).
48. Abraham, M. and R Becker. The Classical Theory of 68. Binns, KJ. and P.J. Larenson. AnalysiS and Compu-
Electricity and Magnetism. New York, NY: Hafner tation of Electric and MagnetiC Field Problems. New
Publishing Company (1951). York, NY: Macmillan Publishing (1963).
49. Clem now, P.e. An Introduction to Electromagnetic 69. Courant, Rand D. Hilbert. Methods of Computation-
Theory. New York, NY: Cambridge University Press al Physics. Vol. 2. New York, NY: Interscience (1961).
(1973). 70. Javid, M. and P.M. Brown. Field Analysis and Elec-
50. Elliot, R.S. Electromagnetics. New York, NY: McGraw- tromagnetics. New York, NY: McGraw-Hill Book Com-
Hill Book Company (1966). pany (1963).
140 I MAGNETIC PARTICLE TESTING

Analytical Modeling of Electromagnetic Fields 82. Root, R.R., et al. "On the Solution of Large Three-
71. Karlqvist, O. "Calculation of the Magnetic Field in the Dimensional Magnetostatic Problems Utilizing MSC/
Ferromagnetic Layer of a Magnetic Drum." Transac- NASTRAN." Proceedings of the First Conference on
tions of the Royal Institute of Technology. No. 80. Finite Element Modeling (September 1980).
Stockholm, Sweden (1954): p 1-6. 83. Simkin, J. and C.W. Trowbridge. Three-Dimensional
72. Zatsepin, N.N. and v.E. Shcherbinin. "Calculation of Computer Program (TOSCA) for Nonlinear Static
the Magnetostatic Field of Surface Defects I: Field Electromagnetic Fieldc;. RL 79/097. Rutherford Labo-
Topography of Defect Models." Defectoskopiya. Vol. 5 ratory (1979).
(September-October 1966): p 50-59. 84. Sarma, M.S. "Magnetostatic Field Computation by
73. Zatsepin, N.N. and V.E. Shcherbinin. "Calculation of Finite Element Formulation." Transactions on Mag-
the Magnetostatic Field of Surface Defects II: Exper- netics. Vol. MAG-12, No.6. New York, NY: Institute'
imental Verification of the Principal Theoretical Rela- of Electrical and Electronics Engineers (1976):
tionships." Defectoskopiya. Vol. 5 (September-October p 1,050-1,052.
1966): p 59-65. 85. Sarma, M.S. and J.c. Wilson. "Accelerating the Mag-
74. Shcherbinin, v.E. and M.L. Shur. "Calculating the netic Field Iterative Solutions." Transactions on Mag-
Effect of the Boundaries of a Product on the Field of netics. Vol. MAG-12, No.6. New York, NY: Institute
a Cylindrical Defect." Soviet Journal of NDT. Vol. 12 of Electrical and Electronics Engineers (1976):
(1976): p 606. p 1,042-1,044.
86. Sarma, M.S., J.C. Wilson, P.J. Lawrenson and A.L.
Joki. "End Winding Leakage of High Speed Alterna-
Numerical Modeling of Electromagnetic Fields
tors by Three-Dimensional Field Determination."
75. Brown, W.F. Magnetostatic Principles in Ferromag- Transactions on Power Apparatus and Systemc;. Vol.
netism. New York, NY: Interscience (1962): p 12-29. PAS-90, No.2. New York, NY: .Institute for Electrical
76. EI Markabi, M.H.S. and E.M. Freeman. "Electromag- and Electronics Engineers (1971): p 465-477.
netic Shielding Effect of a Set of Concentric Spheres 87. Armstrong, A.G.A.M., J.C. Collie and J. Simkin. "The
in an Alternating Magnetic Field." IEEE Proceedings: Solution of Three-Dimensional Magnetostatic Prob-
Vol. 126, No. 12. New York, NY: Institute of Electrical 1~I?s Using Scalar Potentials." Proceedings of
and Electronics Engineers (I979): p 1,338-1,343. COMPUMAG Conference. Grenoble, France (1978).
77. Hammond, P. and T.D. Tsiboukis. "Dual Finite Ele- 88. Rafinejad, P., J.L. Coulomb and G. Meunier. "Perma-
ment Calculations for Static Elechic Magnetic Fields." nent Magnet Three-Dimensional Field Computation
IEEE Proceedings. Vol. 130, No.3. New York, NY by Curvilinear Finite Elements." Proceedings of .
Institute of Electrical and Electronics Engineers: COMPUMAG Conference. Grenoble, France (1978).
p 105-111. 89. Brauer, J.R. "Saturated Magnetic Energy Function for
78. Guancial, E. and S. DasGupta. "Three-Dimensional Finite Element Analysis of Electrical Machines."
Finite Element Program for Magnetic Field Prob- Proceedings of the PES Winter Meeting. Paper C75,
lems." Transactions on Magnetics. Vol. MAG-13, No.3. Part 2. New York, NY: Institute of Electlical and
New York, NY: Institute of Electrical and Electronics Electronics Engineers (1976): p 151-156.
Engineers (May 1977): p 1,012-1,015. 90. Zienkiewicz, O.C., A.K. Bahrani and P.L. Arlett.
79. Weiss, J. and C.M. Stephens. "Finite Elements for "Solution of Three-Dimensional Field Problems by
Three-Dimensional Magnetostatic Fields and Its Ap- the Finite Element Method." The Engineer (October
plication to Turbine Generator End Regions." Trans- 1967): p 547-550.
actions on Power Apparatus and Systems. Vol. PAS- 91. Pissanetzky, S. "Solution of Three-Dimensional Ani-
100, No.4. New York, NY: Institute of Electrical and sotropic Nonlinear Problems of Magnetostatics Using
Electronics Engineers (April 1981): p 1,591-1,596. Two Scalar Potentials, Finite and Infinite Multipolar
80. Demerdash, N.A., T.W. Nehl and F.A. Fouad. "Finite Elements and Automatic Mesh Generation." Transac-
Element Formulation and Analysis of Three- tions on Magnetics. Vol. MAG-18, No.2. New York,
Dimensional Magnetic Field Problems." Transactions NY: Institute for Electrical and Electronics Engineers.
on Magnetics. Vol. MAG-16, No.5. New York, NY: (March 1982): p 346-350.
Institute of Electrical and Electronics Engineers 92. Lord, W. "A Survey of Electromagnetic Methods of
(September 1980): p 1,092-1,094. Nondestructive Testing." Mechanics ofNondestmctive
81. Polak, S.J. and A. Watchers. "MAGGY 2 and PADY Testing. W. Stinchcomb, ed. Plenum Press (1980):
Program Packages for Two and Three-Dimensional p 77-100.
Magnetostatic Problems." Proceedings of the 93. Demerdash, N.A., T.VV. Nehl, F.A. Fouacl and O.A.
COMPUMAG Conference. Grenoble, France (1978). Mohammed. "Three-Dimensional Finite Element
 BASIC ELECTROMAGNETISM I 141

Vector Potential Formulation of Magnetic Fields in Electromagnetic Nondestructive Testing Phenomena.

Electrical Apparatus." Transactions on Power Appara- Ph.D. thesis. Fort Collins, CO: Colorado State Uni-
tus and Systems. Vo1. PAS-lOO, No.4. New York, NY: versity (Spring 1983).
Institute of Elechical and Electronics Engineers (ApIil 105. Hodgkins, W.R. and J.F. Waddington. "The Solution
1981): p 4,104-4,111. of Three-Dimensional Induction Heating Problems
94. C.S. Holzinger. "Computation of Magnetic Fields Using an Integral Equation Method." Transactions on
'Vithin Three-Dimensional Highly Nonlinear Media." Magnetics. Vol. MAG-18. New York, NY: Institute of
Transactions on Magnetics. Vol. MAG-6, No. 1. New Electrical and Electronics Engineers (1982): p 476-
York, NY: Institute for Electrical and Electronics 480.
Engineers (March 1970): p 60-65. 106. Rodger, D. and J.F. Eastham. "A Formulation for Low
95. Silvester, P.P. and M.V.K. Chari. "Finite Element Frequency Eddy Current Solutions." Transactions on
Solution of Saturable Magnetic Field Problems." Magnetics. Vol. MAG-19, No.6. New York, NY:
Transactions on Power Apparatus and Systems. Vol. Institute of Electrical and Electronics Engineers
89. New York, NY: Institute of Electrical and Elec- (1983): p 2,443-2,446.
tronics Engineers (1970): p 1,642-1,651. 107. Pillsbury, R.D. "A Three-Dimensional Eddy Current
96. Aoki, S. "Three-Dimensional Magnetic Field Calcula- Formulation Using Two Potentials: the MagnetiC Vec-
tion of the Levitation Magnet for HSST by the Finite tor Potential and Total MagnetiC Scalar Potential."
Element Method." Transactions on Magnetics. Vol. Transactions on MagnetiCS. Vol. MAG-19, No.6. New
.MAG-16. New York, NY: Institute of Electrical and York, NY: Institute of Electrical and Electronics En-
Electronics Engineers (September 1980): p 725-727. gineers (1983): p 2,284-2,287.
97. Kaminga, W. "Finite Element Solutions for Devices 108. Emson, C.R.I. a~d J. Simkin. "An Optimal Method for
with Permanent Magnets." Journal of Physics (D: Three-Dimensional Eddy Currents." Transactions on
Applied PhysiCS). Vol. 8, No.7 (May 1975): p 841-855. MagnetiCS. Vol: MAG-19, No.6. New York, NY:
98. Hammond, P. "Use of Potentials in Calculation of Institute of E,lectrical and Electronics Engineers
Electromagnetic Fields." IEEE Proceedings. Vol. 129, (19&3): p 2,450-2,452 .
.No.2, Part A, (March 1982); p 106-112.
109. Trowbridge, C.W. "Applications of Integral Equation
99. Sarma, M.S. "Potential Functions in Electromagnetic
Methods for the Numerical Solution of Magnetostatic
Fiel~ Problems." Transact~ons on Magnetics. Vol.
and Eddy Current Problems." Proceedings of the
MAG-6, No.3. New York, NY: Institute for Electrical
International Conference on Numericai Methods in
and Electronics Engineers (September 1970): p 513-
Electrical and MagnetiC Field Problems. Santa
518.
Marghe rita, Italy (1976).
100. Trowbridge, C.W. "Applications of Integral Equations
Methods for the Numerical Solution of Magnetostatic no. Webb, J.p., G.L. Maile and R.L. Ferrari. "Finite
and Eddy Current Problems." Proceedings of the Element Solution of Three-Dimensional Electromag-
International Conference on Numerical Methods in netic Problems." IEEE Proceedings. Vol. 130, No.2,
Electrical and Magnetic Field Problems. Santa Part H (1983): p 153-159.
Margherita, Italy (1976). 111. Anderson, O. W. "Transformer Leakage Flux Programs
101. Webb, J.P., C.L. Maile and R.L. Ferrari. "Finite- Based on Finite Element Method." Transactions on
Element Solution of Three-Dimensional Electromag- Power Apparatus and Systems. Vol. PAS-92. New
netic Problems." IEEE Proceedings. Vol. 130, No.2, York, NY: Institute of Electrical and Electronics En-
Part H (1983): p 153-159. gineers (1973): p 682-689.
102. Penman, J. and J.R. Fraser. "Dual and Complemen- 112. Palanisamy, R. and W. Lord. "Finite Element Model-
tary Energy Methods in Electromagnetism." Transac- ing of Electromagnetic NDT Phenomena." Transac-
tions on Magnetics. Vol. MAG-19. New York, NY: tions on MagnetiCS. Vol. MAG-I5, No.6. New York,
Institute of Electlical and Electronics Engineers NY: Institute of Electrical and Electronics Engineers
(1983): p 2,311-2,316. (November 1979): p 1,479-1,481.
103. Ida, N. Three Dimensional Finite Element Modeling of 113. Palanisamy, R. and W. Lord. "Finite Element Analysis
Electromagnetic Nondestructive Testing Phenomena. of Axisymmetric Geometries in NDE." Proceedings of
Ph.D. thesis. FOlt Collins, CO: Colorado State Uni- the ARPNANFL Review of Progress in Quantitative
versity (SpIing 1983). NDE. La Jolla, CA (July 1979): p 25-32.
114. Palanisamy, R. Finite Element Modeling of Eddy
Eddy Currents Current Nondestructive Testing Phenomena. Ph.D.
thesis. Fort Collins, CO: Colorado State University
104. Ida, N. Three Dimensional Finite Element Modeling of (1980).
t 42 I MAGNETIC PARTICLE TESTING

115. Palanisamy, Rand vv. Lord. "Finite Element Model- 128. Silvester, P.P. and RL. FerraIi. Finite Elements for
ing of Electromagnetic NDT Phenomena." Transac- Electrical Engineers. Cambridge University Press
tions on Magnetics. Vol. MAG-IS, No.6. New York, (1983).
NY: Institute of Electrical and Electronics Engineers 129. Silvester, P.P. and M.V.K. Chari. "Finite Element
(November 1979): p 1,479-1,481. Solution of Saturable Magnetic Field Problems."
116. Ida, N. and vV. Lord. "A Finite Element Model for Transactions on Power Apparatus and Systems. Vol.
Three-Dimensional Eddy Current NDT Calculations." 89. New York, NY: Institute of Electrical and Elec-
Transactions on Magnetics. New York, NY: Institute of tronics Engineers (1970): p 1,642-1,651.
Electrical and Electronics Engineers (November 130. Kaminga, \v. "Finite Element Solutions for Devices
1985). \\'ith Permanent Magnets." Journal of Physics (D:
117. Brown, M.L. "Calculation of Three-Dimensional Eddv Applied PhysiCS). Vol. 8, No.7 (May 1975): p 841-855.
Currents at Power Frequencies." IEEE Proceeding.;. 13l. Demerdash, N.A. and T.W. Neh!. "An Evaluation of
Vol. 129, No.1, Part A (January 1982): p 46-53. the Methods of Finite Elements and Finite Differenc-
118. Hammond, P. "Use of Potentials in Calculation of es in the Solution of Nonlinear Electromagnetic Fields
Electromagnetic Fields." IEEE Proceedings. Vol. 129, in Electrical Machines." Transactions on Power Appa-
No.2, Part A (March 1982): p 106-112. ratus and Systems. Vol. PAS-98, No. l. New York, NY:
119. Anderson, O.W. "Transformer Leakage Flux Program Institute of Electrical and Electronics Engineers
Based on the Finite Element Method." Transactions (January-February 1979): p 74-87.
on Power Apparatus and Systems. Vol. 92. New York, 132. Oden, J.T. "A General Theory of Finite Elements I:
NY: Institute of Elechical and Electronics Engineers Topological Considerations." International Journal for
(March-April 1973): p 682-689. Num.erical Methods in Engineering. Vol. 1, No. 2
120. Chari, M.V.K. "Finite Element Solution of the Eddy (1969): p 205-221.
Current Problem in Magnetic Structures." Transac- 133. 04en, J.T. "A General Theory of Finite Elements II:
tions on Power Apparatus and Systems. Vol. PAS-93, Applications." International Journal for Numerical
No. 62. New York, NY: Institute of Electrical and Method,; in Engineering. Vol.. 1, No.3 (1969): p 247-
Electronics Engineers (1974). 259.
121. Sa~ma, ,M.S. "Potential Functions in ElectromagnetiC The Finite Difference Method
Field· Problems." Transactions on MagnetiCS. Vol.
MAG-6, No.3. New York, NY: Institute of Electrical 134. Erdelyi, E.A. and E.F. Fuchs. "Nonlinear Magnetic
and Electronics Engineers (September 1970); p 513- Field Analysis of Direct Current Machines: Theoreti-
518 .. cal Fundamentals." Transactions on Power Apparatus
and Systems. Vol. PAS-89. New York, NY; Institute of
ElectIical and Electronics Engineers (1970): p 1,546-
1,554.
General Numerical Methods
135. Erdelyi, E.A. and E.F. Fuchs. "Nonlinear Magnetic
122. James, M.L., G.M. Smith and J.C. Walford. Applied Field Analysis of Direct Current Machines: Applica-
Numerical Methods for Digital Computation. New tion of the Improved Treatment." Transactions on
York, NY; Halper and Row (1977). Power Apparatus and Systclns. Vol. PAS-90. New
123. Ida, N. Three Dimensional Finite Element Modeling of York, NY; Institute of ElectIical and Electrollics En-
ElectromagnetiC Nondestructive Testing Phenomena. gineers (1970); p 1,555-1,564.
Ph.D. thesis. Fort Collins, CO: Colorado State Uni- 136. Demerdash, N.A., H.B. Hamilton and G:W. Brown.
versity (Spring 1983). "Simulation for DeSign Purposes of MagnetiC Fields in
124. Donea, J., S. Giuliani and A. Philippe. "Finite Ele- Turbogenerators with Symmehical and AsyrmnetIic~!
ments in the Solution of Electromagnetic Induction Rotors: Model Development and Solution Technique.
Problems." International Journal for Num.erical Meth- Transactions on Power Apparatus and Systems', Vol.
ods in Engineering. Vol. 8 (1974): p 359-367. PAS-9l. New York, NY: Institute of ElectIical and
125. Trowbridge, G\V. "Three-Dimensional Field Compu- Electronics Engineers (1972): p 1,985-1,992.
tations." Transactions on MagnetiCS. Vol. MAG-18, 137. De merdash , N .A., I-LB. Hamilton and G.\iV. Brown.
No. l. New York, NY: Institute of ElectIical and "Simulation for Design Purposes of MagnetiC Field~ in
Electronics Engineers (Janumy 1982): p 293-297. Turbogenerators with Symmetlical and Asym,~11etncal
126. Courant, Rand D. Hilbelt. Methods of Computation- Hotors: Model Calibration and ApplicatiOns. Trans-
al Physics. Vol. 2. New York, NY: Interscience (1961). actions on Power ApparatHs and Systems. Vol. PAS-
127. Harrington, RF. Field Computation by Moment 91. New York, NY: Institute of Electrical and Elec-
Methods. New York, NY: Macmillan Publishing (1968). tronics Engineers (1972): p 1,992-1,999.
 BASIC ELECTROMAGNETISM I 143

: 38 Sieminieniuch, J.L. and I. Gladwell. "Analysis of Numerical Methods in Engineering. Vol. 1, No. 2
1... . Explicit Differ~nce Meth.ads for a Diffusion-Con~ec­ (1969): p 205-221.
tiOl1 Equation. Infernatwnal Journal for Numerical 150. Oden, J.T. "A General TheOlY of Finite Elements II:
Methods in Engineering. Vol. 12, No.6 (1978): p 899- Applications." International Journal for Numerical
ent Method~ in Engineering. Vol. 1, No.3 (1969): p 247-
ts."
916.
139. DufOlt, E.C. and ~.P. Frankel. "Stabili~ COI~ditions .in 259.
'01. the Numerical 1reatment of ParabolIc DIfferential 151. Zienkie\vicz, O.C. The Finite Element Method in
~c-
Equations." Mathematical Tables and Aids to Compu- Engineering, third edition. London, England:
tation. Vol. 7 (1953): p 1,35-152. McGraw-Hill Book Company (1977).
es 140. RichtI11)'er, R.D. and K.\V. Molton. Differential A1eth- 152. Oden, J.T. and G.F. Carey. Finite Elements, Special
J: ods for Initial Value Problems. John \-Viley and Sons Problems in Solid Mechanics. Vol. V. Englewood
5. (1967). Cliffs, NJ: Prentice-Hall (1984).
)f MaeNad, R.H. "An Asymmehical Finite Difference 153. Sabir, A.B. "Finite Element Analysis of Arch Bridges."
Network." Q: Applied A1athematics. Vol. 11 (1953): Finite Elements for Thin Shells and Curved Members.
p 295-310. D.G. Ashwell and R.H. Gallagher, eds. London,
142. Forsythe, G.E. and \-V.R. Was ow. Finite Difference England. John Wiley and Sons (1974): p 223-242.
MetflOds of Partial Differential Equations. John Wiley 154. Bruch, J.C. and G. Zyroloski. "Transient Two-
and Sons (1964). Dimensional Heat Conduction Problems Solved by
143. Wood, W.L. "A Further Look at Newmark, Houbolt, the Finite Element Method." International Journal
etc., Time Stepping FOffi1Ulae." International Journal for Numerical Methods in Engineering. Vol. 8, No.3
for Numerical Methods in Engineering. Vol. 20, No.6 (1974): p 481-494.
(1984): p 1,009-1,018. 155. Ie Provost, e. and A. Poucet. "Finite Element Method
144. Demerdash, N.A. and T.W. NehI. "An Evaluation of for Spectral Modeling of Tides." International Journal
the Methods of Finite Elements and Finite Differenc- for Numerical Methods in Engineering. Vol. 12, No.5
es in the Solution of Nonlinear Electromagnetic Fields (1978): p 853-872. . '. .. .
in Elechical Machines." Tran.~actions on 'Power Appa- 156. Arlet, P.L., et al. "Application of Finite Elements to
ratus and Systems. VoL PAS-98, No. 1. New York, NY: the Solution of Helmholtz' Equation." IEEE Proceed-
Institute of Electdcal and Electronics Engineers ings. Vol. 115, No. 12 (December 1968).
(January-Februaty 1979): p 74-87. 157. McNeice, G.M. and H.C. Amstutz. "Finite Element
Studies in Hip Reconstructions." Biomechanics.
Baltimore, MD: University Park Press (1976): p 394-
The Finite Element Method 4Q5.
145. Aneelle, B. and J.e. Sabonnadiere. "Numelical Solu- 158. Demerdash, N.A. and T.W. Nehl. "An Evaluation of
tion of Three-Dimensional Magnetic Field Problems the Methods of Finite Elements and Finite Differenc-
Using Boundmy Integral Equations." Transactions on es in the Solution of Nonlinear Electromagnetic Fields
MagnetiCS. Vol. MAG-16, No.5. New York, NY: in. Electrical Machines. Transactions on Power Appa-
Institute of Elechical and Electronics Engineers ratus and Systems. Vol. PAS-98, No. 1. New York, NY:
(1980): p 1,089-1,091. Institute of Electrical and Electronics Engineers
146. DeSai, C.S. and J.F. Abel. Introduction to the Finite (January-February 1979): p 74-87.
Element Method. Van Nostrand Reinhold Publishing 159. Huebner, K.H. The Finite Element Method for Engi-
(1972). neers. New York, NY: Wiley Interscience, John Wiley
147. Demerdash, N.A., H.B. Hamilton and G.W. Brown. and Sons (1975).
"Simulation for Design Purposes of MagnetiC Fields in 160. Wood, W.L. "A Further Look at Newmark, Houbolt,
Turbogenerators with Symmetrical and Asymmetrical etc., Time Stepping Formulae." International Journal
Rotors: Model Development and Solution Technique." for Numerical Methods in Engineering. Vol. 20, No.6
Transactions on Power Apparatus and Systems. Vol. (1984): p 1,009-1,018.
PAS-91. New York, NY: Institute of Electrical and 161. Oden, J.T. "A General Theory of Finite Elements I:
Electronics Engineers (1972): p 1,985-1,992. Topological Considerations." International Journal for
148. Wood, \VL. "A Further Look at Newmark, Houbolt, Numerical Methods in Engineering. Vol. 1, No. 2
etc., Time Stepping Formulae." International Journal (1969): p 205-221.
of Nwnerical Methods in Engineering. Vol. 20, No.6 162. Oden, J.T. "A General Theory of Finite Elements II:
1 (1984): p 1,009-1,018. Applications." International Journal for Numerical
49. Oden, J.T. "A General Theory of Finite Elements I: Methods in Engineering. Vol. 1, No.3 (1969): p 247-
Topological Considerations." international Journal for 259.
144 I MAGNETIC PARTICLE TESTING

163. Penman, J. and J.R Fraser. "Dual and Complemen- 176. Oswald, D.J. A New Nondestmctive Testing Tech-
tary Energy Methods in Electromagnetism." Transac- nique. Master of Science thesis. Fort Collins, CO:
tions on Magnetics. Vol. MAG-19. New York, NY: Colorado State University, Colorado (1969).
Institute of Electrical and Electronics Engineers 177. Lord, \V. and D.J. Oswald. "The Generated Reaction
(1983): p 2,311-2,316. Field Method of Detecting Defects in Steel Bars."
164. Silvester, P.P. and RL. Ferrari. Finite Elements for Materials Evaluation. Vol. 29, No.2. Columbus, OR:
Electrical Engineers. Cambridge University Press The American Society for Nondestructive Testing
(1983). (February 1971): p 21-27.
165. Djurovic, M. and C.J. Carpenter. "Three-Dimensional 178. Brown, W.F. Magnetostatic Principles in Ferromag-
Computation of Transformer Leakage Fields and As- netism. New York, NY: Interscience (1972): p 12-29.
sociated Losses." Transactions on Magnetics. Vol. 179. Djurovic, M. and C.J. Carpenter. "Three-Dimensional
MAG-6. New York, NY: Institute of Electrical and Computation of Transformer Leakage Fields and As-
Electronics Engineers (1970): p 828. sociated Losses." Transactions on MagnetiCS.
166. Forsythe, G.E. and W.R Wasow. Finite Difference Vol. MAG-6. New York, NY: Institute of Electrical and
Methods of Partial Differential Equations. John Wiley Electronics Engineers (1970): p 828.
and Sons (1964). 180. Sarma, M.S., J.C. Wilson, PI Lawrenson and A.L.
167. Desai, C.S. and J.F. Abel. Introduction to the Finite Joki. "End Winding Leakage of High Speed Alterna-
Element Method. Van Nostrand Reinhold Publishing tors by Three Dimensional Field Determination."
(1972). Transactions on Power Apparatus and Systems. Vol.
PAS-90, No.2. New York, NY; Institute of Electrical
and Electronics Engineers (1971): p 465-477.

Numerical Modeling of leakage Fields

Materials and Material Properties
168. Lord, W. and D.J. Oswald. "Leakage Field Methods
in Defect Detection." International Journal of NDT. 181. The Nondestructive Testing Handhook, second edi-
No.4 (1972): p 249-274. tion. Vols. 1-6. Columbus, OH: Th~ AINerican Society
169., Lord, W. and rH. Hwang. "Finite Element Modeling' . for Nondestructive Testing (1983-1989).
182. Metals Handbook, eighth edition. Vol. 1. Metals Park,
of Magnetic Field/Defect Interactions." Journal of
Testing and Evaluation. Vol. 3, No. 1. Philadelphia, OH; American Society for Metals (1976).
PA: American Society for Testing and Materials 183. Brown, W.F. Magnetostatic Principles in Magnetism.
(January 1975): p 21-25. New York, NY: Interscience (1962).
170. Hwang, J.H. and "v. Lord. "Magnetic Leakage Field 184. Tebble, RS. and D.J. Craik. MagnetiC Materials.
Signature of Material Discontinuities." Proceedings of London, England: Wiley Interscience, John vViley and
the Tenth Symposium on Nondestnlctive Evaluation. Sons (1962).
San Antonio, TX (April 1975); P 63-65. 185. Cullity, RD. Introduction to MagnetiC Materials.
171. Hwang, J.H. Defect Characterization by MagnetiC Reading, MA: Addison-Wesley Publishing Company
Leakage Fields. Ph.D. thesis. Fort Collins, CO: (1964).
Colorado State University (1975). 186. Snoek, J.L. New Developments in Ferromagnetic lrfa-
172: Lord, W. and J.B. Hwang. "Defect Charactelizatioi1 terials. New York, NY: Elsevier Publishing Company
from MagnetiC Leakage Fields." British Journal of (1947).
Nondestmctive Testing. Vol. 19, No.1 (January 1977): 187. Chen, C.W. Magnetism and "Metallurgy of Soft Mag-
p 14-18. netic Materials. Amsterdam, Netherlands: North
173. Satish, S.R and W. Lord. Finite Element Modeling of Holland Publishing Company (1977).
Residual MagnetiC Phenomena. Boston, MA: Intermag 188. Dielectric Materials and Applications. A.R. von Hippe!,
Conference (1981). ed. New York, NY: John Wiley and Sons (1954).
174. Lord, W., J.M. Bridges, W. Yen and R Palanisamy. 189. Bozorth, RM. Ferromagnetism. New York, NY: D.
"Residual and Active Leakage Fields Around Defects Van Nostrand Publishing (1951).
in Ferromagnetic Materials." Materials Evaluation.
Vol. 36, No.8. Columbus, OB: The American Society
Units of Measure
for Nondestmctive Testing (1978): p 47-54.
175. Satish, S.R Finite Element Modeling of Residual 190. Rasmussen, E. The Fundamental Units of PhysicS and
MagnetiC Phenomena. Master of Science thesis. Fort the Logic of Theoretical Physics. Philadelphia, PA:
Collins, CO: Colorado State University (1980). Dorrace Publishing (1970).
l"""""""""""'"''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
BASIC ELECTROMAGNETISM I 145

Cheremisinoff, N.P. Unit Conversion and Formulas 193. "A Supplement to Information for IEEE Authors and
Manual. Ann Arbor, MI: Ann Arbor Science Publish- IEEE ReGommended Practice for Units in Published
Scientific and Technical Work." IEEE Spectrum. New
ers (1980). York, NY: Institute of Electrical and Electronics En-
n "Infomlation for IEEE Authors." IEEE Spectrum.
Vol. 2. New York, NY: Institute of Electrical and gineers (May 1966).
Electronics Engineers (August 1967).
g
 SECTION
6
MAGNETIZATION METHODS
Donald Hagemaier, McDonnell Douglas Aircraft Company, Long Beach, Ca,lifornia
Volker Deutsch, Karl Deutsch and Company, Wuppertal-Elberfeld, Federal Republic ofGermany
Roderic Stanley, 'nterna~i.on.al Pipe Inspectors Association, Houston. Texas.
148 I MAGNETIC PARTICLE TESTING

PART 1
DESCRIPTION OF MAGNETIC FIELDS

The magnetic particle testing method uses magnetic Magnets only attract materials where the lines of flux
fields to reveal material discontinuities in ferromagnetic leave or enter the magnet. \Vhen magnetic matellal is
materials. The common horseshoe magnet attracts ferritic placed across the poles of a horseshoe magnet, the lines of
materials to its ends or poles. Magnetic lines of flux flow flux flow from the north pole of the magnet through the
from the south pole through the magnet to the north pole as material to the south pole. MagnetiC lines of flux flow
illustrated in Fig. 1a. preferentially through magnetic material l"ather than non-
magnetic material or air.

FIGURE 1. Magnetism in various shapes: (a) a

horseshoe magnet; (b) a ring magnet with air Magnetized Ring
gap; Ic) a closed magnetized ring; and
Id J magnetic particles attracted to a radial crack If a horseshoe magnet is bent so that its poles are close
In a circularly magnetized object together (see Fig. 1b), the' poles still attract magnetic
materials. Iron filings or other magnetic materials cling to

Q.
the poles and bridge the gap between them. In the absence
of a slot, the magnetic flux lines are enclosed within the ring
(Fig. 1c). No external poles exist, and magnetic paiticles
lal ... dusted over the ring are not attracted to the ring even ·1
though there are magnetic flux lines flowing thr0ug!1' it.
N S
'Magnetized matetials attract externally only when poles
.exist. A ring magnetized in this manner is said to contain a
circular magnetic field which is wholly within the ohject.
Ib)
Small changes in the cross section of the ring or in the
permeability of its material may cause external flux and the
attraction of magnetic particles.

Effect of Cracks in a Magnetized Ring

A radial crack in a circularly magnetized object creates
north and south magnetic poles at the edges of the crack.
This forces some of the magnetic lines of force out of the
metal path. These disrupted lines of force are called
magnetic leakage flux. Magnetic particles are attracted to,
the poles created by such a crack, forming an indication of
the discontinuity in the metal test object (Fig. Id), I

Bar Magnet
\Vhen a horseshoe magnet is straightened, it becomes a
bar magnet with poles close to each end. Magnetic flux lines
flow through the bar from the south pole to the north pole
but the flux denSity is not uniform along the bar (Fig, 2a).
MagnetiC patiicles are attracted to any location where flux
emerges and particularly to the ends of the magnet where
the concentration of extemal flux lines is greatest. Since the
 MAGNETIZATION METHODS I 149

magnetic flux within a bar magnet may run the length of the
FIGURE 2. Straightening a horseshoe magnet bar, it is said to be longitudinally magnetized or to contain
results in a bar magnet: (a) lines of magnetic flux a longitudinal field.
pass through the magnet from its south to its
north pole; and (b) a crack in a bar magnet Effect of Cracks in a Magnetized Bar
creates magnetic poles that attract magnetic
A crack in a bar magnet (Fig. 2b) distorts the magnetic
particles outside the bar; magnetic lines of flux
lines of force and creates poles on either side of the crack.
flow from north pole to south pole These poles attract magnetic particles to form an indication
of the crack. The strengths of poles formed at a crack
depends on the number of magnetic flux lines interrupted.
laJ [5 ~: A crack at right angles to the magnetic lines of force
interrupts more flux lines and creates stronger poles than a
crack more nearly parallel to the flux lines. Test indications
of maximum size are formed when discontinuities are at
right angles to the magnetic lines of flux.

figN)\
Ir'--- ---:-;,
li___ _ ----~ ..
1SO I MAGNETIC PARTICLE TESTING

PART 2
MAGNETIZATION WITH ELECTRIC
CURRENT

Electric currents are used to create or induce magnetic The passage of current induces a magnetic field strength
fields in electrically conducting materials. Since it is possible in the conductor as well as in surrounding space. An object
to alter the directions of magnetic fields by controlling the magnetized in this manner is said to have a circular field or
direction of the electrical magnetizing current, the arrange- to be circularly magnetized (Fig. 3b).
ment of current paths is used to induce magnetic flu.x lines
at right angles to expected discontinuities in the test object.
Circular Magnetization of
Circular Magnetization Solid Test Objects
To induce a circular magnetic field in a solid test object,
Electric current passing through a straight conductor (a current may be passed through the object. This creates
wire or bar, for example) creates a circumferential magnetic poles on both sides of discontinuities that are parallel to the
field around that conductor (see Fig. 3a). The magnetic length of the test object. These poles attract fine magnetic
lines of force are always at right angles to the direction of particles and form an indication of the discontinuity (see
the current that induces the magnetic field. Fig.4a). It is also possible to generate a circular field in
To determine the direction tak~n by magnetic lines of
force around a conductor, imagine that the conductor is
grasped with the right hand so that the thumb points in the
direction of the electric cJ1rrent. The fiugers then point in FIGURE 4. Circular magnetization of typical tf7,t
the direction taken by the magnetic field lines, surrounding Objects: fa) circular magnetization caused by
the conductor. This is known as the Flel1iing right hand rule. passing electric current from contact plates .
through the test object; and (bJ production Qf a
localized circular field by passing electric current
FIGURE 3. Fields in circular magnetization: between contact prods
fa) circumferential magnetic field surrounding a
straight conductor carrying an electric current~ la.
and fb) circular magnetization of a test object
through which a magnetizing electric current
passes

fa) MAGNETIC FIELD

+ - 9v
L
'hi'~~~
~CL-
oJ ,.j/V
LCONDUCTOR fb) SWITCH
MAGNETIZING
CURRENT
MAGNETIZING CURRENT

(b) MAGNETIC FIELD

~~
+ -tt'-It=) P-l='rG--
L ~\et '. . . ' ~
L
MAGNETIZING CURRENT
TEST OBJECT
 MAGNETIZATION METHODS 1151

localized areas of the object using prods to pass current Circular Magnetization with Induced Current
through the area being tested (Fig. 4b). ~ current flowing circllmferentially around the ring can
The prod electrodes (generally solid copper or braided be Induced by making the ring a single-turn, short circuited
copper tips) are first pressed firmly against the test object. secondary transformer (Fig.6b illustrates this effect). To
The magnetizin~ current is passed through the prods and accomplish this effect, a standard magnetizing coil can be
into the area of the object in contact with the prods. This
used.
establishes a circular magnetic field in the test object The ring is placed inside the coil with its axis parallel to
around and between each prod electrode. that of the coil. When the coil is energized with alternating
The use of alternating current limits the prod technique
current, the arrangement constitutes an air core transform-
to the detection of surface discontinuities. Half-wave recti- er; the magnetizing coil is the primary and the ring is the
fied direct current is more desirable here because its greater ~ingle-turn ~econdaIy. The total current induced in the' ring
particle mobility helps detect surface and near surface
IS greatly mcreased by inserting a laminated core of
discontinuities with greater particle mobility. ferromagnetic material through the ring.
The prod technique generally is used with dry magnetic For materials with high magnetic retentivity, direct cur-
particle matelials because of increased particle mobility on rent can be applied in the method known as qUick break and
rough surfaces and better penetration. In the United States, the objects may then be tested by the residual method.
wet magnetic particles are not normally used with the prod Quick break is when a direct current field is caused to
technique because of electrical and fire hazards. In Europe, collapse suddenly due to an abruptly interrupted magnetiz-
wet r~~icles are regularly used with prods to achieve higher ing current. The circular field generated by the induc;ed
senSItivIty. Care should be taken to maintain clean prod tips, current leaves the test object with a strong residual induc-
to minimize heating at the point of contact and to prevent
tion. A bearing race is a good example of the type of object
arc strikes and local heating of the test surface. Aluminum
that can be tested advantageously by this method. .
or copper braided tip prods or pads (rather than solid
For t{(st objects made of soft material, with little teten':'
copper tips) are recommended because of the possibility of
tivit5', the continuous method must be used and the collaps-
copper penetration if arcing occ\lrs. .
ing dire~t current field method is not applicable. By using
A remote con~rol switch should be built into the prod
altematmg current (or half-wave direct current) in the
handles to permIt control of the current after positioning
magnetizing coil, the current may be left on and an
and before remOving, to minimize arcing.
alternating current (or half-wave direct current) of the same
frequency as the magnetizing current is induced in the ring.
Circular Magnetization with Prods This current should be allowed to flow long enough to
In ci~cular magnetization with prods, the field strength is produce indications by the continuous method.
prop.ortlOnal to the current used but varies with prod
spacmg and the thickness of the section being tested. A
magnetizing current of 90 to llO A for each 25 mm (1 in.) FIGURE 5. Direct contact method of magnetizing
of prod spacing is recommended for material under 20 mm ring shaped test objects to locate circumferential
(0.75 in.) thick. A magnetizing current of 100 to 125 A for discontinuities
each ~5 mm (1 in.) of prod spacing is recommended for
:ate~a~ over 20 mm (0.75 in.) in thickness. Prolonged
ner~IZll1g cycles may cause localized overheating. Prod CIRCULAR MAGNETIC FIELD
spacmg should not exceed 200 mm (8 in.).
Pr~d spacing less than 75 mm (3 in.) is usually not
practIcal because the particles tend to band around the
prods, making interpretation difficult.

CirCular Magnetization with Direct Contact

. Figure 5 shows the direct contact method for prodUcing
CIrcular
. field·S In
. a nng
. t0 ·In d·lcate CIrcum
. £erentiaI cracks. To
achrfleve a reI·1abie exal1unahon
.. 0 f t h e entire cy1indrical
su T~~e, .two magnetizations are required.
IS IS done because the points of contact (where the
:rrent. enters and leaves the ring) are not adequately
th:~~~bzed for discontinuity indication. The ring must
ore be turned 90 degrees and then retested.
1S2 / MAGNETIC PARTICLE TESTING

FIGURE 6. Magnetizing with the induced current FIGURE 7. Circular magnetization of cylindrical
method to locate circular discontinuities in ring objects using an internal current carrying
shaped objects conductor: fa) internal bar conductor and
fbI internal cable conductor

(a)
faJ

~~~0D1~~ LAMINATED IRON CORE

ALTERNATING CURRENT

(b) DISCONTINUITY CABLE

INDUCED CURRENT PATH

CIRCUMFERENTIAL
DISCONTINUITY
~~'H"~~

LAMINATED CORE

\ CURRENT

Circular Magnetization bf Hollow

Test Objects
With hollow objects or tubes, the inside" surfaces may be
Longitudinal Magnetization
as important for testing as the outside surfaces. When such
an object is circularly magnetized (by passing the magne- Electrical current can be used to create a longitudinal
tizing current through it), no magnetic flux is produced on magnetic field in magnetic materials. When electric current
the inside surface. is passed through a coil, a magnetic field is established
Since a magnetic field surrounds a current carrying lengthwise or 10ngitudinally within the coil (Fig. 8). "
conductor, it is possible to induce a satisfactOlY magnetic The nature and direction of this field are the result of the
field by sliding the test object onto an internal conducting field around the conductor which forms the tums of the coil.
bar (Fig. 7). Passing current through the bar induces a Application of the right hand nIle to the conductor at any
circular magnetic field throughout the volume of the test point in the coil (Fig. 8a) shows that the field within the coil
object. is longitudinal.
When a conducting bar is not available, an electrical cable
may be passed through the test object and connected to
receptacles in the magnetic particle unit (Fig. 7b). For large Coil Magnetization
diameter cylinders, the cable can be brought back on the
outside of the test object, then threaded through again - When magnetic material is placed within a coil, most of
each pass through increases the effective field by a factor of the magnetic lines of force created by the electric cUl-re~t
two. For long finished tubes, unill5ulated conductors are not concentrate themselves in the test object and induce longI-
permitted because of arc bums. tudinal magnetization (Fig. 8b). "
 MAGNETIZATION METHODS I 153

Testing of a long cylindrical object with longitudinal

FIGURE 8. longitudinal or coil magnetization:
magnetization is shown in Fig. 8c. With a transverse discon-
(a) longitudinal magnetic flux within a current
tinuity in the test object, magnetic poles are formed on both
carrying magnetizing coil; (bJ longitudinal
sides "O!' the crack. These poles attract magnetic particles to
magnetization with a coil; and fc) typical
form an indication of the discontinuity. Figure 8c shows that
arrangement of coil and test object for
a magnetic field has been induced at light angles to the
longitudinal magnetization
discontinuity,
Test ohjects too large to fit in a fixed coil can be
magnetized longitudinally by making a coil fro~n several (a)
turns of flexible cable (see Fig. 8b). The use of portable
magnetizing equipment with cables and prods or clamps has
broadened the use of magnetic particle testing - there is no
theoretical limit to the size of the object that can be tested
in this manner.

Field Flow Magnetization fbI

Another means of producing a longitudinal field in a test
object is the field flow method. Here, the field is produced
by electromagnets and passed through objects as if they
were the keepers in a yoke. The field is almost wholly
contained \vithin the test object.
While there is theoretically no limit to the length of a test
object th~t can be magnetized this way, as a practical matter
with an alternating current source, power requirements fe)
limit the effective leng~h to about 1.3 m (4 ft). However,
special techniques using direct current have accomplished
longitudinal magnetization in one step for lengths over 3 m
(10 ft).
The field flow (or yoke) method may have some advan-
tages over the current flow (coil) method in some produc-
tion applications because: (1) moving a magnetizing coil
several times may be impractical and time consuming;
(2) test objects with length-to-diameter ratios less than 3:1
require no special handling; (3) a consistent field wholly
contained within a test object may be required. Frequently
the field flow method is indicated in multidirectional mag-
netization.
A reference standard containing known artificial discon-
tinuities should be located in the center of the test object
dUring setup to ensure adequate field strength along its
entire length.
154 I MAGNETIC PARTICLE TESTING

PART 3
FACTORS CONTROLLING
MAGNETIZATION

Factors that should be considered when selecting a proportion amounts to about 30 percent of the peak value.
method of magnetization include: (1) alloy, shape and con- Figure 9c shows full-wave rectified alternating current
dition of the test object; (2) type of magnetizing current; with a Single-phase bridge circuit. The direct current pro-
(3) direction of the magnetic field; (4) sequence of opera- portion amounts to about 64 percent of peak value.
tions; (5) value of the flux density; (6) desired throughput; Alternating current with a frequency of 50 to 60 Hz
and (7) type of discontinuities anticipated. Material proper- (Fig. 9d) shows excellent uniform magnetization of the
ties and current type are considered below. Direction of the surface even with large changes of cross section. Penetration
magnetic field is covered in the next part of this Section. depth is frequency dependent and equals about 2 mm at
50 Hz (skin effect). There is rapid reduction in indication
sensitivity with increasing depth when using alternating
Material Properties current magnetization.
Pulsed current is illustrated in Fig. ge. Because of the
The alloy, its heat treatment, cold working and other pulse train at predetermined intervals, the danger of heating
conditioning treatments determine the magnetic permeabil- from current flow at the contact points is minimized. Thin
ity of a test object. It is necessary to consider these when walled test objects can therefore be tested using higher
selecting the sequence of operations, the value of flux currents.
density or the magnetic field strength. They, in tum: affect .. Impulse current is normally used with the residual
selection of the magnetization method. method (Fig. 9f). The magnetization effect results from a
The size and shape of the test object determine the most high intensity single-current pulse of sh01t duration (milli-
practical method of magnetization with the available equip- second range).
ment. The surface condition of the test object influences the
selection of particles as well as the magnetization method. Direct Current Magnetization
Surface coatings such as paint, chemical conversion or Direct current obtained from storage batteries was first
lacquer coatings are poor electrical conductors and affect believed to be the most desirable current to use for
testing because it is difficult or impossible to pass magne- magnetic particle testing because direct current penetrates
tizing current through such coatings. "Vhenever a test object more deeply into test specimens than alternating current.
can be properly magnetized with an induced method, The big disadvantage of storage batteries as a source of
coating thickness is the main concern for the inspector. current is that there is a definite limit to the magnitude and
duration of current that can be drawn from a battelv before
recharging. BattelY maintenance is costly and c~n be a
Types of Magnetizing Current source of trouble.
Direct current obtained through chy-plate rectifiers from
Many types of magnetizing current can be used for each alternating current power lines is similar to battelY current
type of testing (see Fig. 9)1 Full-wave rectified alternating and has the advantage of permitting an almost unlimited
current with three-phase blidge circuitry is shown in Fig. 9a. number of magnetizing shots. Today, nearly all direct
With a ripple of 5 percent, nearly the entire cross section current is produced with silicon diode rectification. Current
can be magnetically saturated. This means that the proba- obtained by passing three-phase alternating current through
bility of detecting subsurface discontinuities may be im- special rectifiers is called three-phase rectified alternating
proved over other forms of current. Because only ohmic current.
resistance is involved, large lengths can be tested at high Single-phase full-wave rectjfied altcmating current has
current values. the advantage of greater paIticle mobility when compared to
Half-wave rectified alternating current ~ith a single- the three-phase bridge circuit. There is no detectable
phase circuit is shown in Fig. 9b. The direct current difference in discontinuity detection capability between the
 MAGNETIZATION METHODS / 155

methods. However, a three-phase circuit demands low-

es of current used for magnetic
FIGU RE 9' Typ 'f' d er power consumption to achieve an equivalent current
article testing: (a) full-wave recti Ie , density.
P
alterna I t'ng current with a ,
three-phase
I '
bndge
, , fbI half-wave rectified a ternattng current
circuit, f ., t'f' d Half-Wave Rectified Magnetizing Current
with single-phase circuit; Ie) u -wave r~c I Ie
alternating current and single-phase bridge Half-wave rectified alternating current is the most effec-
't' (d) alternating current at a frequency of tive current to use for the detection of subsurface and
circUi , d tf) , I
50 to 60 Hz; fe) pulsed current; an .mpu se surface discontinuities when dry magnetic particles are
current used. This type of current is produced by putting single-
phase alternating cmrent through a rectifier. The rectifier is
lle. fa) WAVEFORMS a nonlinear electronic component that permits unimpeded
f'nt current flow in one direction and thus simulates direct
1'0-
current characteristics for purposes of testing. Half-wave
rectified current imparts a very noticeable pulse to the
lz particles. This gives them mobility, aids in the formation of
le indications and helps prevent the formation of nonrelevant
III indications.
FIVE PERCENT RIPPLE
It
II
Alternating Current Magnetization
fb'

h -..:.,... t
Alternating current at line frequency is the most effective
current to use for the detection of surface discontinuities,
particularly fatigue cracks. It. is, important that alternating
current testing equipment be built to include proper cur-
rent controls. .
An advantage of alternating current testing is' the ease
fe) with which test objects can be ·demagnetized.

Choosing Alternating or Direct Current 2 •3

A direct current is distributed uniformly across the entire
cross section of the test object. The field and flux density it
creates are maximum at the outer surface and zero along the

IdJ~
object's central axis, as required by Ampere's law. An
alternating current field is forced to the surface because of
the skin effect. This fact has technical consequences for the
current flow in a billet of square cross section, for example.
Figure lO shows the difference in field strength distribution
along the test object surface in cases of alternating and
fe, I direct current.

~ n~t
For direct current flow, the magnetic field near the billet
comers is considerably lower than in the middle of each
side; it is nearly constant in the case of an alternating
current field. For corner cracks to be accurately indicated
by direct current fields, an overmagnetization in the center

,f, I
area of each side may be unavoidable. With the alternating
current method, the' same field strength is obtained across

~t-t
the sUlface of the billet at half the current density.
Figure 11 shows the distIibution of field strength on the
surface of a stepped sample. The difference in magnetic
Held strength for the largest and smallest cross section is
much greater in the case of direct current magnetization.
FROM TIEDE GmbH AND COMPANY. REPRINTED WITH PERMISSION. For the same value of direct current, the field strength is
inversely proportional to the square of the diameter; an
t 56 I MAGNETIC PARTICLE TESTING

FIGURE 10. Magnetic field strength in FIGURE 11. Distribution of field strength in
kiloamperes per meter on the surface of a kiloamperes per meter on the surface of a
current flow in square billets If = 50 Hz); current stepped sample: raj alternating current field and
flow is through the length of the billet (b) direct current field

HDC = 27 5.2 7.4

lal~
HAC"" 5.2 5.2 5.2

/~------74 4.8
,_ 5.3

T
~ 1'---:;
1 = 2.040 A
H'heo = 1/ 4a
1
= 6.4 kA'm- fb)

~.
E
E 1.8 •
a
00 2.8 55
'-
-.L.L-___~
FROM KARL DEUTSCH GmbH AND COMPANY. REPRINTED WITH FROM KARL DEUTSCH GmbH AND COMPANY. REPRINTED WITH
PERMISSION. PERMISSION.

alternating current field is inversely P!oportional only to the Direct current field magnetization cannot guarantee the
diameter. When maximum and minimum field strengths for ability to indicate all subsurface discontinuities, particularly
a specimen are given in test specifications, a complicated .as. depth ·increases. The application of magnetic particle
test object can often be inspected ·onlY in sections, if the test testing methods should be for detecting surface or near
is carried out with a direct current magnetic Held. surface discontinuities. Wet method indications of discon-
More obviously, this effect can be seen on objects with tinuities below 0.25 mm should be relied on only when
sudden enlargements in cross section. In camshafts, for other nondestructive testing methods cannot be used.
example, the alternating current field (in accordance with
the skin effect) follows the contour; the direct current field
follows a more complex path. As a consequence, alternating
Depth of Penetration
current systems are often found at airline and automotive .Figure 12 is a plot of threshold values for the magnetizing
overhaul stations, where surface fatigue cracks can occur at currents necessary to produce a readable indication of holes
various locations on complex shapes such as landing gear in a tool steel ring standard. Holes parallel to the cylindrical
and actuating mechanism components. surface are drilled 1.8 mm (0.07 in.) in diameter at increas-
Furthermore, the magnetic pole areas at the ends of a test ing depths' below the surface. The depths vmy from 1.8 to
object are much smaller with an alternating current field. 21 mm (0.07 to 0.84 in.), in 1.8 mIll (0.07 in.) increments,
The indications of cracks near the contacting area, are much from hole-1 to hole-12.
better with an alternating current field. The results plotted in Fig. 12 were obtained using the d,y
An advantage of the direct current field is its increased continuous testing method with an internal conductor using
depth of penetration. This provides the ability to test for 60 Hz alternating current, and three forms of direct current:
subsurface cracks with magnetic particle techniques. Such (1) direct current from batteries; (2) three-phase rectified
tests are generally used to find cracks under chromium alternating current with surge; and (3) half-wave rectified
plating; subsurface cracks in flash welds; lack of root single-phase alternating current.
penetration or lack of fusion in weldments. . The alternating current test required about 475 A to
Extensive research has shown that the depth effect not indicate hole-l and over 1,000 A to indicate hole-2. Hole-3
only depends on the distance of the crack from the surface could not be shown at any available altemating current
but also on the crack's shape, size and its relations to the test value. Hole-2 was indicated with 475 A straight direct
object dimensions. The indications from subsurface cracks current, with 275 A full-wave rectified alternating current
are relatively blurred and indefinite and can be recognized preceded by a surge of double this amount, and by ~ 75 A
safely only on a sufficiently smooth surface. using half-wave rectified alternating current. Half-wave
 MAGNETIZATION METHODS I 157

rectified alternating current of 750 A indicated hole-12,

FIGURE 12. Comparison of the sensitivity of
while 975 A of direct current from batteties was needed to
alternating current, direct current, direct current
with surge, and half-wave rectified current for indicate hole-lO.
. These comp.arisons velify the importance of choosing the
locating discontinuities wholly below the surface
nght current for producing the best indications. The com-
of a test object; threshold indications produced
parisons also show how current should vmy, depending on
using dry particles and continuous magnetization
the nature and location of the discontinuities.
on an unhardened tool steel ring with an
artificial discontinuity diameter of 1.B mm
(0.01 in.)

21008) --
41016)
>-
lJ...t: 61024)
O:::J-
a::::~Q} 8 (032)
wl-..c
I-Zu
rE°~
UU Vl
10 lOAD)
Vl~

OO~ 12 (0.48)
I---JE
I~c:::: 14 (056)
I-U:-=
friff E 16 (0.64)
al-a::::
« 18 (072)

20 (0..80)

2210.88)
0 250 500 750 1.000

MAGNETIZING CURRENT
(amperes)
158 / MAGNETIC PARTICLE TESTING

PART 4
DIRECTION OF THE MAGNETIC FIELD

The proper orientation of the magnetic flux in relation to cables so that magnetizing current can be passed through
the direction of a discontinuity is the most important factor the test object or through an area of it. Portable equipment
affecting discontinuity detection, even more than the mag- should have a remote control switch on the prods, enabling
nitude of the magnetizing current. the operator to control the current while moving the prods
For reliable testing, the magnetic flux should be at right or viewing the test indications.
angles to the discontinuity. If the magnetic flux is parallel to Another means of local magnetization is to use contact
the discontinuity, there is little magnetic leakage; if an clamps with cables, particularly when the test objects are
indication is formed at all, it is likely to be extremely small relatively small in diameter. Tubular struct~lres can be
or indefinite. To ensure that proper field direction exists at tested this way by positioning the clamps so that the current
the desired magnitude, reference standards containing ar- passes through the area of interest and along the line of
tificial discontinuities are needed. suspected discontinuities (Fig. 13).

Internal Conductor Magnetization

Ci.rcular Magnetization
In a tubular test object, a circular magnetic .field may be
Fundamentals of Circular Magnetization set up by passing current through the tube itself; no field is
induced on the inside surface of the tube. If the test object
The magnetizing method that is easiest to control is is hollow or has holes through which an internal conductor
circular magnetization. This is the metho~ in which th,e can be passed, it is best to induce.a circular magnetic field
magnetizing current is passed directly through the test in the object by passing the magnetizing current through the
object, setting up circular magnetic field lines at right ~mgles conductor. Circular magnetization with an internal conduc-
to the direction of current flow (see Fig. 4a). tor has the following advantages over passing current through
A good way to circularly magnetize the outer regions of a the test object itself:
test object is to place the object between the contact plates
of a stationary magnetic particle testing system. Care should
be taken to clamp the test object firmly between the soft
lead contact plates. Enough of the object's surface area FIGURE 13. Current carrying clamp electrodes
must contact the plates to permit passage of the magnetizing used for testing tubular Objects with small
current without burning. As the area of the surface decreas- diameters
es, the probability of burning increases.
On irregular test objects, it may be helpful to use copper
braid contact pads between the objects and plates to prevent
overheating. When testing objects with irregular cross sec-
tion, it may be necessalY to circularly magnetize with a low
current to inspect the thin areas. A method specific to the
application must be devised to pass a higher current through
the heavier sections for testing of those areas.

Prod Magnetization of Large Test Objects

When an object is too big to fit into available test
equipment, the test object (or areas of it) can be circularly
magnetized by either of two methods.
One method is to use prod contacts with cables to
transmit the magnetizing current from the source to the test
object (see Fig. 4b). Prods are attached to the ends of the
 MAGNETIZATION METHODS I 159

1. it induces nux at the inside diameter of the test object,

permitting testing of inner as well as outer surfaces; Longitudinal Magnetization
2. direct electrical contact is not made with the test
object, thereby eliminating the likelihood of burning; A longitudinal field can be induced in a test object by
anCl placing the object in a fixed current canying coil, mounted
3. several small test objects (washers or nuts, for exam- on the rails of a stationary unit or attached by cables to a
ple) can be suspended on the same conductor and portable unit (Fig. 8). The effective magnetic field induced
tested in groups (see Fig. 14). by a coil extends from 150 to 220 mm (6 to 9 in.) beyond
either end of the coil. Depending on fill-factor, if the test
Capacitor Discharge Circular Magnetization object is long, it is necessary to magnetize and test it in
sections along its lcngth.
Capacitor discharge methods are used for the circumfer- Longitudinal magnetization with pOltab]c equipment is
ential magnetization of oil field pipes. The technique is accomplished by wrapping current canying cable ill a coil
discussed in detail under Circumferential Magnetization of around the test object.
Pipe later in this Section.

Important Considerations in Coil Magnetization

limitations of Parallel Magnetization To induce an adequate longitudinal magnetic field with a
coil, the long dimension of the test object should be at least
A circular magnetic field surrounds any electrical conduc- twice as great as its short dimension or end pieces should be
tor and this is the magnetic principle underlying circular added and the long axis of the test object should be parallel
magnetization with an internal conductor. Knowing this, to the coil axis. This is especially true in the case of
some operators have assumed that they can induce a circular irregularly shaped test objects, because the shape of the
field in a test object by placing it next to instead of around object affects the direction of the induced flux.
a conductor. This is not true. When a wheel, smaller in diameter than a coil, is placed
Some field is induc~d· in the test ,object by such a i'n the coil (as 'shown in Fig. 15), a field is induced in the
procedure, but since a portion of the magnetic flux path is white areas of the test object in such a direction that radial
in air, the field in the object is greatly reduced, distorted discontinuities ·create indications. However, radial cracks in
and unevenly distributed. This procedure is sometimes the snaded areas of the test object are parallel (or nearly so)
called parallel magnetization. It is not dependable and to the induced magnetic field, so that few or no indications
should not oe used. are formed. Furthermore, magnetic poles and attractive
forces occur in these areas. To indicate radial discontinuities
in the shaded areas, it is necessary to rotate the test object
FIGURE 14. Internal conductor method used to through 90 degrees and remagnetize it, although this
produce circular magnetization: (a) several ring technique is not recommended.
shaped test objects magnetized simultaneously; The detection of radial cracks in a test object of this shape
and fbJ close-up of a ring with cracks in several is more accurately and rapidly done using an internal
locations and orientations conductor (see Fig. 14). Better methods for finding circum-
ferential discontinuities in ring shaped test objects are
shown in Figs. 5 and 6. Ring shaped objects, disks, wheels or
fa) races are best checked for circumferential cracks using the
induced method of Fig. 6. An iron core, for example, is used
with a coil surrounding it to produce a toroidal field. This
method has an advantage over the direct contact method
(Fig. 5) in that no danger of arcing or burning exists, and the
field is constant throughout the test object.
fb)
Yoke Magnetization
A longitudinal magnetic field can be induced in a test
object or in a limited area of an object by using a handheld
yoke. A yoke is a U shaped piece of soft magnetic mateIial,
either solid or laminated, around which is wound a coil
carrying the magnetizing current (Fig. 16).
160 / MAGNETIC PARTICLE TESTING

FIGURE 15. Coil magnetization of a circular FIGURE 16. Longitudinal lines of force induced by
shape; radial discontinuities will be indicated a yoke magnet: fa) electrically energized yoke
only in white areas (to reveal radial magnet and fbI permanent magnet yoke
discontinuities in the dark areas, the test object
must be rotated 90 degrees and remagnetized);
circular discontinuities will be indicated in the
shaded areas (to reveal circular discontinuities in
the white areas, the part must be rotated
90 degrees and remagnetized)

fb, MAGNETIC PARTICLES

,COLLECT AT CRACK

Combined Circular and Longitudinal

Magnetization
Complete testing for discontinuities in different direc-
When a test object is placed across the opening of the tions requires that two or more magnetizations and tests be
U shape and the coil is energized, the object completes the performed. The test object should first be circularly mag-
path of the magnetic lines of force. This sets up a longitu- netized and examined for indications, then longitudinally
dinal field in the test object between the ends of the yoke. magnetized and inspected. Demagnetization is the final
Permanent magnetic yokes can also be used to create a step.
longitudinal magnetic field (Fig. 16a). Such yokes are often It is critical to remember that discontinuities are best
specified by their lifting power or by the tangential field detected when they are at right angles to the magnetic lines
strength midway between the legs. of force. -
 MAGNETIZATION METHODS 1161

PART 5
MULTIDIRECTIONAL MAGNETIZATION 1,2,3

With all magnetizing methods, discontinuities pelvendic- alternating frequency). The direction change again occurs in
ular to the magnetic flux are optimally indicated. However, such a way that, for at least a short time, some field
discontinuity detection depends heavily on mateIial perrne- component' is perpendicular to any existing crack direction.
ability, nux density and the properties of the testing medium This in turn causes a magnetic paIticle accumulation and
(see Table; 1). subsequent detection.
It is true that magnetic excitation also permits the In the case of a combination of direct current yoke (or
detection of disC()}Jtinuities that ate not exactly pelvendic- coil) with an alternating current flow (Fig. 18), the resulting
ular to the flux direction. In this case, the line of flux can be field swings around the axis of the test object (Fig. 19). This
decomposed into two components, one of them parallel and combination of static and dynamiC fields results in a vector
the other pe1l)endicular to the direction of the crack. The swinging over an angle. The magnetic vector swings around
perpendicular component contributes to the indication of the position of the direct current field and, in any given
the discontilluity. In some cases, even cracks appearing to position, a sufficient component of it is at right angles to a
be parallel to the flux direction may be weakly indicated. possible discontinuity. If both fields have the same strength,
The reason is that most cracks are ragged in outline a total angle range of ± 45 degrees (totally 90 degrees) is
(intercrystalline cracks) so that some sections may be covered.
properly oriented for detection. However, at best cracks can The disadvantages of direct current yokes are: (1) stray
only he detected when the angle behveen them and the fields may form at the rounded ends of the test objects or. at
.. ..
direction of magnetization is more than 30 degrees.
.
FIGURE 17. Superimposition of direct current
Combined Direct Current Fields magnetic fields: (a) addition of field vectors and
(b) relationship of field directions
When a direct current magnetic field of a certain direc-
tion and strength is superimposed on one of a different
direction and strength, both fields can be combined to form fa)
another field as shown in Fig. 17. PhYSically, the resulting
field is [urnit'd by the addition of the hvo magnetic field
vectors (something like the combination of forces in a
paral.Ielogram). The resulting field has a direction and STRENGTH AND DIRECTION
OF RESULTING FIELD
strength different from either of the plimary fields, and is
therefore very djfficult to predict, especially when induced
in complex shapes.
VECTOR OF LONGITUDINAL FIELD
Two or three Held directions may be superimposed by

r r
sequencing. If the fields vary in strength with time, a
~winging vector field is created. It is essential that multiple
unposed fields he balanced.
fb) ULTlNG
fiELD TEST OBJECT

Combined Direct Current and ,t FORCE LINES OF

LONGITUDINAL
MAGNETIZATION
Alternating Current Fields FORCE LINES OF CIRCULAR
MAGNETIZATION

F?r combined direct and alternating current magnetic

partIcle testing. hvo perpendicular magnetic fields are FROM KARL DEUTSCH GmbH AND COMPANY. REPRINTED WITH
~uperimposed in such a way that the resulting field changes PERMISSION.
Its direction \\ith time (generally in rhythm with the
162 I MAGNETIC PARTICLE TESTING

FIGURE 18. Complete discontinuity detection by FIGURE 19. Combined crack detection by a
the traditional swinging field method circular alternating current and a longitudinal
(combination of direct current yoke and direct current field
alternating current head current flow)

CLAMPING DEVICE WITH

CURRENT CONTACTS
CRACK INDICATIONS TEST OBJECT

I I
I I
: I
I I I I
I

DIRECT CURRENT FOR YOKE ABC D

A C D
ALTERNATING CURRENT FOR CURRENT FLOW

FROM KARL DEUTSCH GmbH AND COMPANY. REPRINTED WITH

PERMISSION.

cross-sectional differences between the yoke and the object; LEGEND

and (2) considerable field reduction occurs with large cross- R = VECTOR OF CIRCULAR MAGNETIZATION
L = VECTOR OF LONGITUDINAL MAG~ETIZATION
sectional changes of the test object. SiI:nilar. situations can G = VECTOR OF RESULTANT FIELb (AT TIMES ABCDI
also occur with coil magnetization.

Combined Alternating Current FROM KARL DEUTSCH GmbH AND COMPANY. REPRINTED WITH
PERMISSION.
Fields
It can be advantageous to magnetize in two directions
with alternating current fields. An effective combination for
must be assembled from laminated transformer sheet to
periodic alteration of the resulting field vector cannot be
prevent eddy current losses. These constructions are more
realized if the hvo fields are in phase or in counter phase.
expensive than direct current yokes.
At a phase shift between 50 and 130 degrees, a rotating
Alternating current yokes are usually built to operate over
magnetizing vector of sufficient uniformity can be obtained.
limited clamping lengths, because a longitudinally Olientcd
At a phase shift of 90 degrees and for equal field strengths,
alternating current field is reduced with increasing clamping
a circularly rotating vector is generated, as illustrated in
lengths.
Fig. 20. When three-phase current from the mains of a
The advantages of an alternating current yoke are: 1
phase shift of 120 degrees is used, the behavior is similar
(elliptical rotation).
1. uniform field distJibution, even over test objects with
geometrically complicated shapes and over changes of
cross section;
Yoke for Combined Alternating Current Fields
2. the possibility of induced field flow;
For structural reasons, a direct current yoke of solid steel 3. simple and rapid demagnetization of the test object;
has been used for many years to indicate transverse cracks. and
Alternating current yokes have recently been used but they 4. short testing times.
 MAGNETIZATION METHODS I 163

FIGURE 20. Alteration of the magnetic field by FIGURE 21. Combined auxiliary magnetization by
combination of two phase shifted alternating a combination of alternating current trow and
current fields (rotating vector) current induction methods

CIRCULAR
MAGNETIC
FIELD

/ .... L "\
( """II
\ Rt.~~
.... _-G ALTERNATING ~CU~R:R~E~N~T----_ _ I
LEGEND
R = VECTOR OF CIRCULAR FIELD
L = VECTOR OF LONGITUDINAL FIELD FROM KARL DEUTSCH GmbH AND COMPANY. REPRINTED WITH
G = VECTOR OF RESULTANT FIELD PERMISSION.

magnetization method. The technique can detect very small

FROM KARL DEUTSCH GmbH AND COMPANY. REPRINTED WITH discontinuities because at some period in the magnetization
PERMISSION.
cycle the field vector is normal to the discontinuity direc-
tion. Also, in comparison with the single magnetization
procedures (head and coil magnetizations), the combined
\Vith clamping lengths from 900 mm to 1,200 mm (36 to current method presents considerable economic advantages
48 in.), longitudinal magnetization may better be achieved since it requires only one processing step instead of hvo or
using a movable alternating current coil. more.

Testing Procedures with Multidirectional

Magnetization 2 Combined Auxiliary Magnetization
With multidirectional magnetization methods, the result- An exceptional variation of the multidirectional magneti-
ing magnetic field changes its direction in the rhythm of the zation method is called combined auxiliary 'nuzgnetization,
altemate frequency of the current application, becoming used for cylindrical test objects. Here, the auxiliary alter-
alternately perpendicular to ceIiain cracks. However, the nating current flow method is combined with the alternating
direction of such a vector is also parallel to a particular crack current induction method (Fig. 21). A magnetizing bar is
for a very sh01i time. For this reason, when using the put through the hole of the test object and this laminated
multidirectional method, the application of the testing steel, copper coated bar serves as both a current conductor
medium must always take place during magnetization. An and a phase shifted magnetic field conductor.
~ndication previousl)' established cannot be held by magnet- This reliable technique is non contacting, pole free and
IC force and could theoretically be flushed away by the can indicate cracks of any direction on inside surfaces,
testing fluid. The procedure does possess a certain inertia; outside surfaces and face areas of cylindlical test objects.
the magnetic vector has already left the parallel direction Combined aUXiliary magnetization can be carried out only
when the pmiicles begin to move away. by systems equipped with an alternating current yoke -
Experience in Germany and more recentlv in the United contributing still further to the trend toward alternating
States indicates hvo ad~'antages for the 1~1Ultidirectional current yoke techniques (see Table 1).
 '64 I MAGNETIC PARTICLE TESTING

TASLE ,. Techniques for multidirectional

magnetization
Circular magnetization:
• current flow method (for solid and tubular test objects)
• internal conductor method (recommended for tubular test
objects)

Longitudinal magnetization method:

• yoke or coil magnetization

Current induction method:

• circulating current induced in a ring using a laminated core
and the influence of the Fluctuating longitudinal alternating
current yoke Field /Fig. 6)

Combinations of methods For overall coverage in one test mode:

• yoke or coil magnetization with current Flow or internal
conductor method
• internal conductor method with current induction method
 MAGNETIZATION METHODS I 165

PART 6
CIRCUMFERENTIAL MAGNETIZATION
OF PIPE

The dJilling and production of natural hydrocarbons

generally require that the tub~lar p'ro~u.ct (casing, ~ubing Magnetization Methods for Oil Field
and dIill pipe) be tested for dlsconhnmhes. Magnehc flux Applications
leakage methods are the most commonly used tests for
detection of outer and inner sllrface discontinuities. Ultra- Two distinct methods 7 are used for the circumferential
sonic methods are used for testing regions difficult to magnetization of tubes up to 14 m (45 ft) long (see Fig. 22).
inspect vvith magnetic flux leakage techniques. Both methods use an insulated rod (generally made from
A common form of testing for longitudinally oriented, aluminum although this is not required) w-hich passes
surface breaking, tight discontinuities (seams, laps and through the bore of the tube. In Fig. 22a, the rod is
cracks) involves magnetization of the tube circumferentially reasonably well centered in the bore and fed ~th some
by the internal conductor method, followed by testing with . form of direct current. In mill installations, this might be
some form of magnetic flux .leakage sensor. The use of full-wave or half-wave rectified alternating current with the
ferroprobes, coils and solid state sensors for this application subsequent test being done using wet fluorescent magnetic
is summaIized in Volume 4 of the Nondestructive Testing particles.
Handbook series: Electrorrwgnetic Testing. The fext below In field operations, banks of batteries have been used for
deals with the magnetization of oil field tubes and treats current. When the magnetizing current is pure direct
magnetic particles as the sensor. current, the magnetic field strength H at the outer surface
of the tube, when the conductor is centered within the bore,
is given in amperes per meter by:
Specifications for Testing Oil Field
Tubu/ars (Eq.l)

Specifications and recommended practices for the testing

of oil field mateIials are wIitten by both the Amelican Where:
Petroleum Institute (API)5 and by oil companies. 6 Table 2
summarizes the API documents that pertain to the testing of I the current (amperes); and
oil field tubular product by magnetic palticle techniques. A R" the outer radius of the tube (meters).
more detailed descIiption of these documents is given in the
Section titled Codes, Standards and Specificationsfor Mag- Figure 23 illustrates some of the values used in this
netic Particle Testing. discussion. In Eq. 1, the field strength is given in amperes

TABLE 2. American Petroleum Institute specifications and recommended practices for magnetic particle
testing of oil field tubular products
Specification 5CT Specification for Casing and Tubing (first edition, 1988)
Specification 5D Specification for Drill Pipe (first edition, 1988)
RP 5A5 Recommended Practice for Field Evaluation of New Casing, Tubing and Plain-End Drill Pipe
RP 7G Recommended Practice for Drill Stem Design and Operating Limits
 166 I MAGNETIC PARTICLE TESTING

FIGURE 22. Two methods for establishing FIGURE 23. The eddy current Ie created in a steel
circumferential magnetization in elongated tube at the beginning of a pulse I in a rod
tubes: (a) the central conductor method with a centered within the bore of the tube; the
battery pack to provide high current; and direction of Ie on the inner surface opposes that
(b) the internal conductor method with a of I; the outer surface forms the return path
capacitor discharge system (a peak and duration
meter may be used to measure pulse amplitude
and duration) \ RETURN CURRENT TOGETHER WITH Ie
\ FORMS EDDY CURRENTS IN TUBE WALL

fa}
H =-L..
~ +
bra

+
fbJ
~Ie-

per meter since Ro is expressed in meters. However, this field strength at some radius r at some instant while the rod
field strength is sometimes measured with a Hall element and eddy current fields are finite, is given in amperes per
gaussmeter, and since one gauss is numerically equal to one meter by:
oersted in air, conversion to gaussian units yields:
(Eq ..3)
(Eq.2)
The He term is the field strength created by the eddy
where I is in amperes and Ro is in centimeters. current itself.
The second magnetization method is shown in Fig. 22b.
The motive force is proVided by a capacitor discharge unite
and the necessity for rod centralization is eliminated.
The Eddy Current Effect
Magnetization by this method obeys no scientifically simple In consideling Fig. 2:3, Ampere's law indicates that the
rules because the rapid rise of rod current during magneti- field at the radius r is caused by the currents inside that
zation causes the induction of an eddy current in the tube radius (rod current and inner wall eddy current). The outer
and this detrimentally affects penetration of the magnetiz- wall eddy current is the return loop for the inner wall eddy
ing field strength into the material. current, and plays no role in the theory so hlr outlined.
The direction of the induced eddy current Ie with respect However, since it does represent an unwanted current
to the rod current I is shown in Fig. 2:3 for a centered rod. flowing in the tube, its presence does lead to two very
By Lenz's law, the eddy current induced on the inner practical considerations.
surface of the tube must create a field within the matelial First, pipes being magnetized before testing should he
which opposes the field caused by the rod current I. The insulated from each other by an air gap. If this does not
 MAGNETIZATION METHODS 1167

occur, then the outer surf~lce eddy current can jump from
protrusions in the pipe being magnetized to the next pipe in FIGURE 24. Magnetizing force versus flux density
the string. The resulting arc can cause burns on both tubes. (B-H) curves: (a) high strength (OGS-90) tubular;
This in tum can cause hard spots on the materials at which material is a sour gas grade of special chemistry
corrosion might preferentially occur. This is paiticularly to and heat treatment; and fb) a lower strength
be avoided with corrosion resistant matelials, some of which fK-55) oil field casing; the dashed lines indicate
require a hardness less than 22 on the Rockwell C scale for that the materials are magnetized almost to
1
saturation by application of 3.2 to 4.0 kA-m-
longevity in sour environments. .
Secondly, the matClial should be msulated from the metal (40 to 50 Oe)
racks that carry it. If pipe racks are not insulated with a layer
of nonconductive JnateJial (ru!)ber or wood, for example), laJ B B,
then the outside diameter surface eddy current can flow to
ground through the !'ack anJ there is a real possihility of arc
burn at tll(~ point ot contact.

Use of 8-H Curve in Setting

Specifications
There is an important fact about the magnetization of test
objects by the capacitor discharge, internal conductor meth-
od: the ling sample B-H curve govems the flux density value MAGNETIZING FORCE
in the mateliaJ. In effect, knowing the B-H propelties of the kiloamperes per meter (oersted)
material from a ling sample investigation allows field strength
levels to be set. Figure 24 shows the B-H properties of two
typical oil Hel~. tubular. n~ateIials: a Ej20 MPa (90,000 psi) fbJ B
proprietmy material (DSG-90) and a 390 MPa (55,000 psi)
K-55 casing mate1ial.. . )-Vi
The impOltant point that can be made from these curves 1-V1
-:J
Vlru
is that after application bf about 3,200 A-m -1 (40 Oe), the ZQ)
9 UJO
mateJials are effectively saturated. It is generally true of oil o~
field tubular matelials that 3,200 to 4,000 A-m -1 (40 to 3:;-
-JV1
500e) are required 'A-ithin the matelial to magnetize to a LL~
level sufTicient for subsequent residual induction testing. It
is required that this fIeld strenbrth level be reached at each
pOint in the tube wall, despite the demagnetizing effect of
the eddy current.
This requirement does not lead to a simple current MAGNETIZING FORCE
equation that can be used by a typical operator in the field kiloamperes per meter (oersted)
(experimental specifications found effective in saturating
AFTER LORD. REPRINTED WITH PERMISSION.
tubes are presented later in this Section).

Typical Requirements for Direct electrical properties of the matelial - do not affect the field
strength. The actual value used is often determined by
Current Magnetization specifications, as agreed between the manufacturer of the
material and the user. A typical specification° is given by:
. If. the central conductor method is used for the magne-
tIzatlOll of tubes, then the values given in Table 3 reflect the
magnetizing field at the outside diameter of either 3,200 I = 12,000 D (Eq.4)
A-m- I (40 Oe) for I I or 4,800 A-m -1 (60 Oe) for 12 for
typical pipe sizes. Since the magnetization method is direct where D is the tube diameter in meters, and:
Current, then the wall thickness, mass per meter (weight per
foot) and tube grade _ which affect the magnetic and I = 300 D
168 / MAGNETIC PARTICLE TESTING

TABLE 3. Current requirements for direct current

magnetization of oil field tubes; direct current or FIGURE 25. Plots of the capacitor discharge
long pulse (> 0.5 sJ only; not valid for capacitor internal conductor current (I versus tJ and the
discharge magnetization average flux density induced (8 versus t) in a
tube; Imax and T are measured with a peak and
Tube Diameter duration meter; note that the flux density peaks
millimeters finches) well after the current
60 ( 2.4) 600 910
73 ( 2.9) 730 UOO
89 ( 3.5) 890 1,340
102 ( 4.0) 1,020 1,530
114 ( 4.5) 1,150 1,720
127 ( 5.0) 1,280 1.910
140 ( 5.5) 1.400 2,100
168 ( 6.6) 1,690 2,530
178 ( 7.0) 1,790 2,680 8,
194 ( 7.6) 1,940 2,920
219( 8.6) 2200 3,300
244 ( 9.6) . 2,450 3,680
273 (10.8) 2,740 4.110
298 (I 1.8) 3,000 4.500
340 (13.4) 3,410 5.120 o~ __________________________
• 3.2 kA'm- 1 /40 Oel AT OUTSIDE DIAMETER
TIME
•• 4.8 kA'm- 1 /60 Oel AT OUTSIDE DIAMETER (milliseconds)

where D is the tube diameter in inches. This amperes per

diameter unit specification is equivalent to 3,760 A-m-} variations that are measurable for single pulses, such as are
(4.7 Oe) at the tubt; surface. It can be seen from Fig ..24 tl}at provided by capacitor discharge units. ThE! flrst vaIiation'is
such a field strength raises the value of the flux density in that of the magnetizing current (I versus t). In this variation,
the tube to a high level, so that after the field has fallen to a relatively rapid rise of current to its maximum vahle Imax
zero, the flux density in the material is at a value close to is followed by a much slower fall toward zero, the entire
remanence (Br). pulse length being on the order of 200 ms. This time
variation is the response to the discharge of a capacitor C,
initially charged to Vo volts through a resistor R in a circuit
FUll-Wave and Half-Wave Rectified Alternating containing inductance L. A simple mathematical analysis is
Current provided later.
For the central conductor method outlined above, some The second variation is that of the average bulk flm:
form of rectified alternating current is often used. It should denSity within the material (B versus t). This quantity rises
be noted that such current waveforms induce eddy currents at a much slower rate than I(t) due to the shielding effect of
in the test object. The field strength waveform at the outer Ie. A high level of magnetization is reached ",:hen the flux
surface can be seen by positioning a Hall element to detect denSity at the point F is close to the material's saturation
the field and feeding the output of the gaussmeter to an value Bs. The consequences of conditions shown in Fig. 25
oScilloscope. are that deep magnetization of the tube only occurs when
the detrimental effect of the eddy current is overcome by
elongating the electrical pulse in time so that the magnetiz-
ing current is still effective as the eddy current is dying
Pulsed Current Magnetization away.
The fall in induction from F to B r is that which is normally
Internal conductor magnetization (using Single pulses of expected when the magnetizing field strength falls to zero,
current) differs from direct current or continuous magneti- as it does after the passage of a pulse. This is determined by
zation by the central conductor method because account the B-H curve for the material undergOing magnetization.
must be made for the fact that the induced eddy current Should the point F not represent saturation (BJ, then the
may not have time to die away before the field strength from material reaches some average bulk flux density lower than
the conductor current dies away. Figure 2.5 shows two time B r . This is often not a problem: while the surbces are
 MAGNETIZATION METHODS I 169

magnetized sufficiently for longitudinally oriented discon- A material might also arrive with an induction between
tinuities to hold magnetic particles, no information of the zero and - Br and it is desired to perform testing at + B r.
intelior condition is required. However, in the case where During magnetization, the material should take the path
relativel), thin elongated tubnlars can be tested from the - BrHr;PBsB,-. That is, through saturation B~ to remanence
outside diameter surface only, saturation of the material is Br .
necessaJY for inside diameter discontinuities to pro<;luce In the case of material initially at zero induction, the tube
magnetic nux leakage at the outside diameter. is at 0 on Fig. 26 and during magnetization takes the path
OPBsBr'
In cases where the pulse is insufficiently strong, the
Practical Testing Situations mateIial may follow a magnetization path such as - BrHcPQ
or OPQ. It is then essential to pulse more than once. A
Commonly encountered testing situations for the magne- possible magnetization path during a second pulse is QBsB r.
tization or remagnetization of tubulars are discllssed below, The net final induction is raised as shown.
including: (1) matelial at unknown induction; (2) ll1aterial at
zero induction; and (3) mateIial not saturated by pulse.
The magnetic condition in which a sample arrives is often
not known to the inspector, who must assume the worst Analysis of Pulse Current
possible case: the matelial is at saturation in a direction Magnetization
directly opposed to that caused by the magnetic particle test
equipment. This is resolved by taking the material from an In the text below, an analysis of the pulse current internal
unknown value of remanence in one direction to remanence conductor method for magnetizing elongated tubulars is
in the other direction as is shown in the schematic B-H presented. Simplified equations are given for the types of
curve for the mateIial (Fig. 26). current pulses available for magnetization. From the theo-
retical viewpoint, the cun;ent pulse time dependence (I
versus t ·of Fig. 25) is discussed and then formulas are
FIGURE 26. Possible paths taken by presented for the inductance experienced by the magnetiz-
. ing circuit.
clrcumferentlaJJy magnetized material from
various initial magnetization conditions to These formulas illustrate the dependence of such induc-
saturation Bs and ~hen remanence Br in a known
tances (1) on the average value of the differential perme-
direction: fa) material at remanence in the ability (dB/dH) of the object under magnetization; and
opposite direction; (bJ material at zero induction;
(2) between the field strength and flux density limits im-
point P indicates weak pulse followed by a posed by the exciting current and B-H propelties of the
second pulse material.

B Current Pulse Time Dependence

For LCR circuits, the time variation of the current pulse
obeys the equation:

d(LI)
dt
+ IR + II dt
C
= 0 (Eq.5)

o ...-_f--_____
H The three terms on the left of Eq.5 represent the
instantaneous voltages across the inductance, the resistance
and the capacitance in the circuit (Fig. 22b). The induc-
tance in the circuit is mainly that of the rod tube system,
since by careful design the presence of additional induc-
tance between cables and ground can be minimized. Be-
fa) cause inductance is time dependent, it is included in the
derivative term. The resistance is the combined resistance
of the rod, cables and their connection, and any internal
resistance in the capacitor discharge box. Resistance in the
discharge box may be due to the forward resistance of a
170 I MAGNETIC PARTICLE TESTING

silicon controlled rectifier included to eliminate the possi-

bility of current oscillation. The capacitance of the system is FIGURE 27. Typical pulses from capacitor
generally in the region of 2 to 8 F. discharge systems; the longer pulse is more
Equation 5 has three solutions if the time dependence of effective in magnetizing a tube sample (line
L is ignored. These solutions depend on the relative values pipe): (a) short pulse and fb) long pulse
of L, C and R.

2Vo
I exp ( - f3t) (Eq. 6)
)4~ _ R2

sin ~4L
C - R"-

I ~p(f32t) exp ( - f3t) (Eq.7)

TIME
2~)
(milliseconds)
I exp ( - f3t) (Eq. 8)
~R2 - -
4L
C
Definition of Puls.e Length
2 4L
sin h JR - - It has become commonplace to define the length of such
C
pulses as the time taken for the pulse to reach 0.5 IIlHLx
during decay T (Fig. 25). Both Imax and T are measur~ble
Where: with an inductive ammeter or peak and <;Juration meter
(Fig. 22b). Such pulses are effective in magnetizing tubular
Vo = the voltage to which the capacitor bank is charged; test objects because the field strength from the rod current
and is still high as the eddy current in the test object dies away,
f3 = RJ2L. so that penetration of the field into the bulk of the material
occurs.
Since the inductance is a function of time, a full solution
The solution to Eq. 6 is oscillatory, but the presence of for the variation of the pulse current I(t) can only be
the SCR limits the pulse to only the first· positive-going obtained by modeling the effect that the induced eddy
peak. This is shown in Fig. 27. In this example, the pulse has current has on the instantaneous value of L. Experimental
a length of 17 ms and reaches 10,500 A. Such pulses are evidence indicates that, at least for elongated tubulars, the
ideal for magnetizing objects of low electrical conductivity, physics of the magnetization process can be illustrated by a
such as ferrite magnets. However, with highly conducting discussion of the constant Lease.
materials such as steel tubes, the initial rapid current rise
(up to millions of amperes per second) induces a shield of
eddy currents that does not permit field penetration into the Typical Values for L, C and R
bulk of the material. The net effect of this is a magnetized
outer layer only. In the design of a capacitor discharge pulsing system, it is
The exponential solution (Eq. 7) is known in its mechan- essential to aim at a pulse that has sufficient length to deeply
ical analog as critical damping. It is difficult to achieve in magnetize the material. There are two reasons for this.
this situation because it depends on the value of L which in First, the material to be tested may arrive at the test site in
turn is dependent on the physical and magnetic parameters a longitudinally magnetized condition and it may be neces-
of the test object. The formula for the inductance of a tube sary to remagnetize it circumferentially prior to testing.
is given below. Second, some specifications call for relatively low emergent
The sin h solution (Eq. 8) leads to the longest pulses longitudinal field strengths at the ends of such elongated
because there is no oscillation. Pulses of full length up to test objects. Rotation of the bulk flux density into the
160 ms are commonly used in the oil tube testing industry circumferential direction may be simplest way to achieve
(Fig. 27b). this.
 MAGNETIZATION METHODS / J 7,

An additional consideration unrelated to the physics of

the magnetization of the test object is the safety of the
system in both installed and field situations. The National Using the International System of Units (51), Eg. 11
becomes:
Electric Code should be consulted for details. However, it
appears essential [()J' field use to limit the charging voltage of
tIle capacitor hank to SO V. The tendency of this restriction
is to add capacitance to the system. L ~ 0.16
r
(iI)(dB)
dH (Eq. 12)
The resistance of the magnetization system is a factor in
permitting high currents to flow. It is minimized for field
where r is O.S(R" + RJ All lengths are in meters and dB/dH
lise by employing parallel strands of AWG 0000 copper is dimenSionless (see Fig. 29).
welding cahle for the connections between the rod and the
capacitance discha~'ge box. The rod is made of aluminum The inductance of thin walled tubes is seen from Eq. 11
(mainly becallse of the continued need to make and break to be proportional to the length (e) and wall thickness (T) of
the tube, and inversely proportional to its radius or diame-
the rod) hut any highly conductive material would work
ter. Neither of these physical parameters nor the value of
equally well. The reqUirement of elongating the pulse
lengtl1 to ensure the presence of its, field after eddy currents dBldIi can he controlled by the designer of the magnetizing
eqUipment.
have died away f~lr outweighs the requirement of minimiz-
ing the overall resistance of the magnetizing circuit. Typical For much of the tubular product used in oil fields, the
values, which might include that 0£.5 m (16 ft) of cable and value of T/R does not vary a great deal, perhaps only by a
15 m (49 ft) of rod, are 1 to ,15 milliohms. factor of two. The average value of dB/dH encountered
The capacitance within the capacitance disc11arge supply during magnetization can he seen from Fig. 26 to vary
is generally within the range of 2 to 8 F, which is compar- widely, depending.o n (1) the point (P) reached on the B-H
atively large. This OCCurs because of the need to maintain Cl.llve by the material dUring magnetization; and (2) the
relatively low voltages around the circuit and to elongate the starting point for Ihagnetization (anywhere from _ B r to Q
pulse. • . on the B axis). EXi'lmples of typical inductances follow ..
While tIle val lies Rand C can be controlled by the
manufacturer, the value of L. cannot, mainly because it Example 1: pipe.. magnetized to saturation follOwing the
depends on the test object undergoing magnetization. In path - B,HcPQ
the. case of tubulars, the inductance is given by:
B" :::: 1.2 T (12 kG)
e P = 1.2 T (12 kG)
L
27T' (:!) (Eq. 9) e = 10m (33 ft)
Where: T = 12.6 111m (0.,15 in.)
r = 136.,15 mm (SA in.)
dH = 2,400 A·m -1 (30 Oe)
e :::: the length of the tube;
dBldH :::: the differential permeability;
R() :::: the outer radius; and Here dB is 24,000 G, so dB/dH = 800, and
j
R :::: the inner radius (see Fig. 23).
7
L :::: (2 X 10- )(10 m)(12.6 mm)(800)l(136.S mm)
It often Occurs that the wall thickness T is much smaller :::: 148 /-LH
than the avemge radius of the tube. Under these circum-
stances, Eg. 9 may be converted to:

Example 2, same tube as example 1, taken from Q

T tlll'ougl1 P and B\ to Br by a second pulse.
r +
L e 'dB) 2
27T' (dH In T (Eq. 10) Q = 1.0 T (10 kG)
r - B,\ = 1.,15 T (1,15 kG)
2 dH :::: 4,000 A.m -1 (SO Oe)
Which reduces to:

The average value of dBldIi is now only 100. Dividing the

value obtained in example 1 by 8, the ratio of the two
(Eq. 11) values of dB/dH exhibited by the steel yields:
L :::: 18.S;..tH
172 / MAGNETIC PARTICLE TESTING

Example 3: pipe initially unmagnetized follows path high level of residual induction. However, broad gUidelines
OPB~Br with: based on research "vith a variety of tubes indicate that the
values given in Table 3 provide'adequate magnetization.
Bs = 1.5 T (15 kG)
= 9.09 m (30 ft)
T = 4.83 mm (0.2 in.) Pulse Duration
R = 27.7 mm (1.1 in.)
dH = 3,200 A-m -1 (400e) In Table 4, pulses are classified by duration. Long
duration pulses are those in excess of 100 ms. For such
So that: pulses, the induced eddy current can be assumed to have
died away while the magnetizing field strength is still high
L = (2 X 10- 7 )(9.09 m)(4.83 mm)(15,000140)1 enough to cause saturation. Moderate duration pulses are
(27.7 mm) those between 40 and 100 ms. For magnetization, the
= 119 JLH longevity of the induced eddy current is acknowledged by its
effect on the tube (shown through the use of the lineal mass
The relatively large change in inductance exhibited by the of the tube rather than the outer diameter). Short pulses are
tube in examples 1 and 3 affects the shape of the pulse those below 40 ms. The maximum current requirement for
waveform, notably the easily measurable parameters of peak the Single short pulse compared to that for the Single
current Im'Lx and pulse duration 7'. These parameters are moderate pulse is higher for the same lineal mass of tube. In
shown in Fig. 25. effect, the higher current causes a larger magnetizing field
strength in an attempt to overcome the eddy current.
Should it·be necessmy to use two such pulses, the peak
current requirement falls because the material is partially
Design Considerations magnetized. If the peak current can only reach Imax = 180
(W), then tWo such pulses are required. Should the pulse be
Good equipment design must include user input about of insufficient magnitude to magnetize the tube with two
the mate'rial being magnetized.' The worst case for the pulses, then a third pulse is necessary so that the three
internal and external resistances of the magnetizing system pulses meet the requirement of Imax = 145 (\y).
should bt:: known to the manufacturer and worst values of . These requirements are deSigned .td ensure that the bulk
inductance should be investigated. Under no circumstances induction follOwing the pulses is at least 90 percent of the
should peak currerits be stated for th~ purpose of magneti- remanence value. In most cases it is higher.
zation withe _,( an electrical and magnetic load being used
for the system evaluation.
Depending on the use of the equipment, the relevant TABLE 4. Pulse classification by duration with
regulations should be <;onsulted with regard to insulation, current requirement
isolation, explosion proofing, intrinsic safety and purging.
Such regulations are found in a variety of places, depending Current
on the use of the product. Notable among these are the Magnetization Duration Requirement
System (milliseconds J Equation
Occupational Safety and Health Administration (OSHA),
the National Institute of Occupational Safety and Health
(NIOSH), Code of Federal Regulations (CFR) and a variety Long pulse > 100 1= 11.8 (Od
of foreign specifications, many of which are a great deal 1=300 (0 21
more stringent than those in the United States. Moderate pulse 40 to 100 1= 74 (Wd
Equipment designers should particularly note the re- 1=110 (W21
quirements of the CSA when designing for Canada; the Single short pulse o to 40 1=161 (Wd
1=240 (W21
requirements of the United Kingdom, Norway and West
Germany are applicable when designing for the North Sea.
Double short pulse o to 40 1=121 (Wd
1= j80 (W21
Triple short pulse o to 40 1= 97 (Wd
1=145 (W21
Magnetization Recommendations
I CURRENT IN AMPERES
D. OUTER DIAMETER IN MIlliMETERS
Tubular product has such wide limits of diameter and wall 02 OUTER DIAMETER IN INCHES
thickness that it is difficult to provide a universal specifica- W, TUBE WEIGHT IN KILOGRAMS PER METER
W2 TUBE WEIGHT IN POUNDS PER FOOT
tion for the measurable parameters of current pulses for a
 MAGNETIZATION METHODS I 173

Current Pulse Effectiveness FIGURE 28. Measuring the flux density induced in
the circumferential direction by the methods
There are two methods used to evaluate the effectiveness shown in Fig. 22
of a current pulse. A third technique, which detects surhl.Ce
fields only, is also outlined here. The first method is a
variation on the Rowland ring technique for the evaluation
of the magnetic parameters of magnetizable materials, and
involves the me,lsurement of magnetic flux. The second
method is an indirect technique using an inductive ammeter
(peak and duration meter). The third method uses sinmlat-
ed contact discontinuities. .
H

The Fluxmeter Method

Magnetic fluxmeters measure the total magnetic flux
threading an area defined by a search coil. In the case of
circumferential magnetization of a hollow product by the
intemal conductor method, the search coil can be a single- than one turn around the test object, the resulting error in
turn coil through the test object. Flux c.hanges are given by: final bulk flux can be reduced. However, for the pUlpose of
establishing the presence of a residual induction in the test
~<p = A~B (Eq.13) object to excite magnetic flux leakage from discontinuities,
this procedure is not neces,s4l)'.
Where:

~<I> = changes in flux;

The Inductive Ammeter Method
A = the area of the test object perpendicular to the
search coil (A = tC); and Figure 29 shows an inductive ammeter or peak and
~B = the change in flux density of the test object induced duration meterY As shown in Fig. 22b, the pickup coil of
dUJing magnetization. . the device is threaded onto any convenient part of the
magnetizing circuit and when the pulse is fired the meter
Commercially available fluxmeters can generally be com- reads the peak current (Im,lX of Fig. 25) and the duration of
pensated for the test object area so that the device can be the pulse (T of Fig. 2.5).
made to read the average flux density directly. The problem Saturation of the material occurs when successive read-
with this approach to the measurement of the final flux ings on the ammeter are identical. This can be explained as
density is that the initial flux density in the test object, \\lith follows. When the first pulse is fired, the material exhibits its
respect to the vector direction of the search coil, must be highest value of dBldH because of the inclusion of the steep
zero. This problem occurs because, when flux changes are part of the B-H curve into the value of L (Fig. 27). The
to be measured, the initial value must be known. However, average value of dBldH is effective in determining the value
if the tube shown in Fig. 28 is initially unmagnetized or the of the inductance in Eqs. 6, 7 and 8. This value is relatively
prior magnetization is longitudinal, then the fluxmeter reads large compared to what the material might exhibit during a
t~le average density of induced circumferential magnetiza- second pulse. High values of dBldH for a first pulse (0.001
hon. T-m-A -1 in the SI system or 800 in the gaussian system)
~hould the output of the fluxmeter be presented on an correspond to lower values during the second pulse (1/8,000
oscIlloscope, it should be noted that flux values (caused by T-m-A -1 in the SI system or 100 in the gaussian system). In
the passage of a pulse) between the beginning and the end effect, the second pulse expeliences a lower inductance
o.fthe magnetization process represent the flux linked by the than the first pulse.
smgle turn coil and contain the effect of the flux in the air The effect of the lowered inductance experienced by a
between the terminals of the fluxmeter. second pulse is to permit the peak current I IlHlX to reach a
DUring the pulse, the air field caused by the current I in higher value than it reached on the first pulse - in effect,
~he rod and eddy currents Ie in the test object affects the the material is different - but the system response is also to
mstantaneollS fluxmeter reading. \\fhen these currents have lower the duration T. By monitoring I m,lX and T, it is possible
~ied away, only flux perpendicular to the Single-turn coil for inspectors to determine the relative degree of magneti-
affects the final result. If the operator has time to wind more zation of the test object.
174 / MAGNETIC PARTICLE TESTING

FIGURE 29. Inductive ammeter or peak and FIGURE 30. Plots of tube wall thickness versus:
duration meter; measures the peak current Imax (a) magnetizing force; and (b) resultant flux
in kiloamperes and pulse duration 'T in density; the lower lines represent field and
milliseconds for magnetizing current pulses induction at the beginning of the pulse (time
proceeds up the figures); the central regions are
the last to be magnetized

Cl:::
UJ
f-
UJ
fa)
~
UJ
Vl
0
Cl::: UJ
UJ f-
a.. Vl
Cl:::
2 UJ
« 0 ID OD

8.000 100

FROM INSPECTION RESEARCH. REPRINTED WITH PERMISSION.

The Simulated Test Discontinuity

Strips of high permeability matelial are commercially
available and these may be placed in intimate contact with
the test object after magnetization. Such stlips contain three
test discontinuities encapsulated in brass so that the lift-off
between the test object and the stlip is minimized to that of 20 40 60 80 100
the brass encapsulation. PERCENT OF THICKNESS
Under such circumstances, the magnetized material
shares flux with the strip and, if the test discontinuities give fb)
an indication with magnetic particles, then so will a similarly
sized discontinuity in the test object.

1.4 14
Use of Inductive Ammeters
1.2 12

When magnetizing with p~Ilse techniques, the value of

the material's field strength H and flux denSity B both
S
-l
10
U.
change with time. In Fig. 30, the horizontal a,"{es show the U 0.8 8
percent distance from the inside diameter surface to the i=
UJ
Z 0.6 6
outside diameter surface. The vertical axes show either the lJ
fraction of H required to saturate the material or the flux «
2 0.4
density B. The lowest lines indicate time ii'om the very start
of the pulse. The uppermost lines indicate the field strength 0.2

and flux density levels at later time increments.

It can be seen from Fig. 30 that three phenomena occur 20 40 60 80 100
during pulse magnetization: (1) both inner and outer sur- PERCENT OF THICKNESS
f~lCes are rapidly magnetized; (2) the midwall region is the
last part of the mateIial to be magnetized; and (3) the
midwall region can be left with a low state of magnetization This last phenomenon contlibutes to magnetic fields from
if the pulse field strength is insufficient to saturate the discoutinuities at one surface, prodUcing no leakage field <It
matetial. the other surf~lce, when the material is not saturated. The
 MAGNETIZATION METHODS 1175

leakage field into the midwall section of the material merely Operation of Inductive Ammeters
raises the local magnetization level to a higher degree.
The inductive ammeter is a microprocessor based instru-
During magnetization, if parts of the matelial do not
ment that employs an inductive pickup coil. This coil
reach a field strength level that ensures saturation (the pOint
contains a large number of turns wound onto a noncon-
P in Fig. 26), then the ensuing bulk residual flux denSity is ducting nonmagnetic ring shaped core. It is threaded onto
low and the matelial requires additional pulses to saturate it.
the cables from the capacitor discharge system or onto the
The magnetization process calls for the highest values of the
rod itself. When a pulse is fired, the flux caused by the
inductance L in Eqs . .5 through 12 during the first pulse and
current surge links with the ling and the voltage induced in
lower values dUling subsequent pulses. The general effect of
the coil (see Fig. 31) is given by:
a high value of inductance is to lower the value of Imax and
elongate the value of T. E = (2 X 10 -7)Nd ell In ~ (Eq. 17)
In order to show that this is the case, and to limit the elt a
necessary mathematical computation, Eg. 7 is selected and
Where:
from it the closed form results for Imax and T are found.
First, the time t (in seconds) at which 1DlllX occurs is found N = the number of turns in the ring;
by differentiation of Eg. 7 to be: d = the axial length of the ring;
b = the outer radius of the ring;
2L a = the inner radius of the ring; and
t =- (Eq.14) dIldt = the rate of change of the current.
R
This equation is derived from Faraday's law of induction. In
Where: order to provide a signal related to the current itself, Eq. 17
must be integrated. The result is:
L = self inductance (henrys); and
[(2 x 1{)-7)Nd In· ~ JI
R = resistance (ohms).
.e = f E • dt i"' .. (Eq. 18)
When this value is used in Eq. 7, the result for 1n1<lX is:
Here, e is the output voltage of the integration circuit. Since
all the terms in the brackets are known, the output of the
(Eq. 15) integration of the induced voltage is proportional to the
instantaneous current, and the instrument can be calibrated
Where: to read current. Electronic circuits are used to measure the
peak current I m,lX and the pulse duration T.
Vo = voltage; and
C = capacitance (farad).
FIGURE 31. Diagram of an inductive pickup coil
This result indicates that the value of Imax is inversely (dimensions used in Eq. 17)
proportional to that of L (the greater is L, the lower is Im,lX)'
In order to find T, l(t) must be set at 0 ..5 Imax. The result in
seconds is:

L
T = 5.36 Ii (Eq. 16)

I~ this case, the pulse duration as defined by T is propor-

tIonal to the value of L. Larger values of L, such as are found
for the initial pulse, lead to the longest pulse durations. READOUT
CALIBRATED
The B-H curve indicates that the lowest value of induc- IN AMPERES
tance that can occur under these magnetization conditions is
that exhibited by saturated matelial, when the value of
dBldH is at its lo~est (see Eq. 11). If two identical readings
are obtained from an inductive ammeter as shown in
Fig. 22, the material must be exhibiting its lowest induc-
tance to the magnetizing circuit and must therefore be at
remanence B r'
176 I MAGNETIC PARTICLE TESTING

PART 7
MAGNETIC FLUX IN TEST OBJECTS
WITH COMPLEX SHAPES
When a discontinuity lies perpendicular to the magnetic
field and is at or near the surface, a leakage field occurs FIGURE 33. Typical geometry problem in
which attracts and holds magnetic particles applied to the magnetic particle testing
test object surface. The capturing and holding power of the
leakage field is determined by both the size of the discon-
tinuity and the magnitude of the magnetic flux in the test
object.
There are rules used to define the current needed and
how it is applied to produce the desired direction and flux
density in the test object. Those rules, along with recom-
mended particle concentrations, are specified in military
and commercial specifications. Unfortunately, many com-
plex ferromagnetic aircraft components have varying cross
sections, large cutouts and protruding extremities. With
these, it is difficult to apply empirical rules that guarantee
effective testing of the object in its entirety.
The testi.ng of the object ,shown in Fig. 32 has been field· direction and magnitude does not reveal the field level
detailed in the literature. ~l The direction and intensity of required for crack detection but does uncover problem
the external flux fields are measured using circular and areas caused by part geometry. Locations having cross-
lo~gitudinal magnetization by, means ·of a transverse Hall sectional areas significantly larger thaT). those where. the
probe gaussmeter. The magnitude of the current pulse' is current enters the test object' exhibit fields whose magni-
determined, froin empirical rules in order to ascertain the tudes are extremely low compared to those at the entry area
validity of the rule in each case. The determination of the (pOints 5, 6 and 7 in Fig. 32).
Magnetization was performed using full-wave direct cur-
rent techniques. Low fields were also observed in deep
cutouts (between points 6 and 7 of Fig. 32) and at the
FIGURE 32. Complex shape of steel forging for extremity areas not directly in the path of the current flow
aircraft '
(points 4, 6 and 7 of Fig. 32). Any area producing a field of
less than 10- 3 T (10 G) was deemed inadequately magne-
tized. It was also determined that when the current branched
into different directions at an intersection, the field at this
intersection was zero (see Fig. 33 and points 1, 2, 3 and 5 of
Fig. 32). These particular ai-eas required additional tests
using portable electromagnetic yokes.
When dealing with complex test objects, initial investiga-
tions conceming the direction of the flux field and adequacy
of field strength should be determined using calibrated
artificial discontinuity shims placed at a sufllcient number of
locations on the test object. Special techniques are required
to establish adequate field strength and direction in sonie
areas of the test object. The techniques should be docu-
mented for future reference when identical objects are to be
tested. In cases of high production rates, special magnetiz-
ing tools may need to be fabricated to achieve reliable
testing of the entire test object. This is impOltant because
areas containing changes in shape or thickness are likely
3 locations for development of cracks dllling fabrication or
smvice.
 MAGNETIZATION METHODS / 177

REFERENCES

1. "~\'1agnetic Crack Detection Techniques for the Detec- 7. Stan lev, R. "Circumferential l\lagnetization of Tubes
tiOll of Surface Cracks." Nondestructive Materials Test- and tl{e Measurement of Flux De~lsity in Such MateIi-
ing. Essingen, Federal Republic of Germany: Tiede als." MateJials Evaluation. Vol. 44, No.8. Columbus,
GmbH and Company. OB: The AmeIican Society for Nondestmctive Testing
(1986): p 966-970.
2. Deutsch, V., A. Becker and M. Vogt. Crack Detection 8. Moake, G. and R. Stanley. "Inspecting OCTG Using
by Magnetic Particle Examination. K. Deutsch, ed. Capacitive Discharge Systems." Materials Evaluation.
'¥uppclial-Elberfeld, Federal Republic of Germany: Vol. 41, No.7. Columbus, OH: The American Society
Karl Deutsch GmbH and Company (1979). for Nondestructive Testing (1983): p 779-782.
3. Deutsch, V. and M. Vogt. "A Comparison of AC and 9. Stanley, R. "Basic Principles on Magnetic Flux Leakage
DC Fields for Magnetic Particle Methods." The British Inspection Systems" and "Capacitor Discharge Magne-
Journal of Non-Destructive Testing. No.4 (July 1982). tization of Oil Countly Tubular Goods." Electromagnet-
4. Betz, C.E. Principles of Magnetic Particle Testing. ic Methods of Nondestructive Testing. W. Lord, ed.
Chicag.o, IL: Magnafl.ux COlporation (1967): p 234. . Vol. 3. Go,rdon and Breach Publ1shing (1985): p97-160.
10. Schindler, John. Current Pulse Monitor. US Patent
5. American Petroleum Institute Specifications 5CT and 4,502,004 (June 1980).
SD. Dallas, TX: Amelican Petroleum Institute. See also 11. Gregory, C., et a1. "Approaches to Velification and
API Recommended Practices RP 7G and RP 5A5. Solution of Magnetic Particle Inspection Problems."
6. Specifications for the Nondestructive Evaluation of API Materials Evaluation. Vol. .30, No. 10. Columbus, OH:
Oilfield Tubular Goods. Rev. l. Exxon Company (May The American Society for Nondestructive Testing
1984). (1972): p 219.
 SECTION

MAGNETIC LEAKAGE FIELD

7
MEASUREMENTS
Roderic K. Stanley, International Pipe Inspectors Association, Houston, Texas
Laurence C. Wong, Magnaf/ux Corporation, Chicago, Illinois
180 I MAGNETIC PARTICLE TESTING

PART 1
FUNDAMENTALS OF MAGNETIC FLUX
LEAKAGE FIELDS

Magnetic palticle testing is not an isolated technical displaced flux is forced out of the object's surface into the
discipline. It is a combination of two distinct nondestructive surrounding environment (air or water), where it can be
testing techniques: magnetic flux leakage testing and visual detected (see Fig. 1).
testing. The basic principle of the magnetic particle tech-
nique is to magnetize an object to a flux density that causes
magnetic flux leakage from a discontinuity. Powdered Magnetizing Current
ferromagnetic material is then passed through the leakage To induce flux leakage, magnetizing current can he
field and those held over the discontinuity are visually passed through the test object by direct contact. This is
interpreted by the operator. commonly do~e but because of the danger of arc bums, it is
From a theoretical point of view, the only difference not always recommended (see Fig. 2). Insulated current
between magnetic flux leakage testing and magnetic particle carrying rods or cables may be used, by passing them·
testing is the use of iron or iron oxides as a sensor. In effect, through holes in the test object. Other alternatives are the
magnetic particles may be considered a commonly used use of coils to carry the current around the test object and .
form of sensor for the detection of magnetic flux leakage the use of electromagnets or permanent magnets applied to
(sometimes called stray fields). the· test object. 3
The key to ideal magnetic particle testing is to provide the When current· is present, there is an associated field
highest sensitivity to the smallest discontinuities by a careful strength R(t) that raises localized areas to various flux
corpbinrtion of: (1) applied magnetic field strength. R(t); density B(t) values, based on the B-H properties' of the test
(2) flux denSity B(t) in the test object; (3) particle size and material. Figure 1 shows a computer simulation of field
appliqation method; and (4) optimal viewing conditions. In lines in and above a material at some value below magnetic
order to do this, experiments are necessary with all of the saturation, as can be seen from the bending of the field lines
parameters. The best combination is then chosen for a under the discontinuity. 4
particular application.
Writers of specifications have often over-generalized this
empirical process in order to provide the magnetic particle
test operator with a set of rules that govern all situations. FIGURE 1. Field lines around, through and above
This generalization can lead to inappropriate specifications a discontinuity (an oblique slot), as computed by
for certain magnetic particle tests. a finite element computer model; note the
There are many forms of magnetic field sensors, includ- asymmetry of the magne~ic flux leakage field
ing the Hall element, the magnetodiode, the ferroprobe and
the sensor coil.l ,2 Tape recorder heads are magnetic sen-
sors, as are the triaxial flux gate magnetometers that are
orbited above the Emth to detect very small changes in
magnetic fields, The purpose of this chapter is to provide
details about the use of sensors in measuring and detecting
fields for magnetic nondestructive tests,

Inducing Magnetic Flux Leakage

The essence of all magnetiC flux leakage testing is to
induce a magnetic flux denSity B(t) around a discontinuity.
The flux denSity mayor may not be time dependent but
should be at such a level that some of the flux is displaced
AFTER HWANG. REPRINTED WITH PERMISSION.
by the discontinuity's higher magnetic reluctance. The
 MAGNETIC LEAKAGE FIELD MEASUREMENTS I 181

FIGURE 2. Typical arc burn caused by direct FIGURE 4. Highly curved magnetic field from a
contact magnetization narrow surface breaking discontinuity

FROM BAKER-HUGHES TUBULAR SERVICES. REPRINTED WITH

PERMISSION.

be possible to distinguish the cause of the leakage without

the use of additional nondestructive testing technology.
FIGURE 3. Minor surface flux leakage from
One way around this problem of excessive magnetization
variations in local magnetic permeability may be
is to localize magnetizing fields at the object surface. This
the source of false test indications
can be done using alternating current fields and the corre-
sponding skin effect. As a rule, skin depth .(also called
standard depth of penetration) fOl: 60 Hz alternating current
fields is 1 mm (0.04 in.). The field strength falls.to lie of its
surface value, or 37 percent, at this. depth. At two skin
depths, field strength falls to (lie) x (lie) or 13 percent of
its surface value. Magnetic flux leakage from discontinuities
depends on the value of H(t) and, in tum, on how large a
B(t) value the field strength causes around the discontinuity.

Why Particle Indications Form

SUlface-breaking discontinuities best detected by mag-
Effect of Flux Leakage on netic particle tests are those that expel the optimal magnetic
flux leakage for the technique. In order to gain a clearer
False Indications inSight of this, it is necessary to understand three sets of
variables:
In a magnetic p,uticle test, it is impOltant to raise the field
strength and f1ux density in the object to a level that 1. how discontinuity parameters affect the external mag-
produces magnetic f1ux leakage sufficient for holding parti- netic flux leakage; . _
cles in place over discontinuities. On the other hand, 2. how magnetic field parameters affect the external flux
excessive magnetization causes particles to stick to minor leakage field; and
surface leakages not caused by discontinuities. 3. how the sensor reacts to passing through such fields.
If such surface leakage occ~lrs \ Fig. 3) and attracts large
numbers of particles, the result is a f~llse indication and the
test object is said to be over magnetized for this inspection.
Discontinuity Parameters
It may then be necessary to verif\· the test results -with shear The discontinuitv characteristics that are critical to the
Wave ultrasound or another nondestructive testing method. formation of magl{etic particle indications include depth,
Such false indications may result from local permeability \vidth, and angle to the object surface. The effects of
changes which are caused lw local stresses in the test object. discontinuity v,7jdth on the topography of the magnetic flux
In some cases, the magne'tic flux leakage field might be leakage field has been desClibed in what might be termed
caused by a suhsurface matelial discontinuity and it may not classical approaches S - 7 where the discontinuity may be
182 I MAGNETIC PARTICLE TESTING

replaced by arrays of poles. Higher ambient field strengths

FIGURE 5. Effects of induction on flux lines in the
or flux densities are included within such models by increas-
presence of a discontinuity: I a J compression of
ing the pole densities that give rise to the magnetic flux
flux lines at low levels of induction around a
leakage fields. More recently, computer models have been
discontinuity, so that no surface flux leakage
developed4 ,8 to explain how magnetic flux leakage fields are
occurs; IbJ lack of compression at high induction,
related to discontinuity parameters (Fig. 1 is an example of
showing some broad surface magnetiC flux
such work).
leakage
In cases where the discontinuity is narrow and surface
breaking (seams, laps, quench cracks and grind tears), the
magnetic flux leakage field near the mouth of the disconti- (a) ....
nuity is highly curved (Fig. 4). The activating field strength
may be quite small (a few amperes per meter) or, after
saturating the test object, inspection can be performed with
the resulting residual induction.
·.. n
=====:.:.-=,:~-=--=--=--=--=------"~-------=--=--=--=--=--=-:,:,:,=====
In the case of subsurface discontinuities (inclusions and
laminations), the magnetic flux leakage field at the inspect-
ed surface is much less curved (Fig. 5). Relatively high FLUX LEAKAGE

ill~I~1111
values of field strength and flux density within the object
are required for testing. This lack of leakage field curva-
ture greatly reduces the particles' ability to stick to such

-C5
indications. (b, ~

@?
Magnetic Field Parameters
The properties of the magnetic field that most affec~ flux
leakage include the field strength, local B- H properties and
the angle to the discontinuity opening. The leakage field's
ability to attract magnetic particles is determined by several Where:
additiqnal factors. These include: 9
K = a mathematical constant [(Nem3)eA -2];
!l. = the ambient leakage field strength (A-m- 1
); and
1. the magnetic forces between the magnetic flux leakage VH = the vector gradient of the field (A-m-2 ).
field and the particle;
2. image forces between a magnetized particle and its It can be seen from Eq. 1 that Fm is dependent on the
magnetic image in the surface plane of the test object; local field stre~h H and how it changes over th~ength of
3. gravitational forces that may act to pull the particle the particle VH. For surface discontinuities, VH is large
into or out of the magnetic flux leakage field; and (because the field is highly curved), while IT it_self need not
4. surface tension forces between the palticle vehicle and be large. For subsurface discontinuities, vIT is relatively
the object surhtce (wet method tcsts). small and IT itself must be raised to compensate for the
small change. Unfortunately, raising IT will also raise surbce
noise.
Some of these forces may in turn VaIY with: discontinuity In other forms of magnetic flux leakage testing, the flux
olientation; the Earth's gravitational field; paIticle shape density is raised to a higher level than is common with
and size (in effect, with paIticle effective permeability); and magnetic particle testing and nonrelevant indications (noise)
with the particles' containing medium. are in some way recognized. For example, the Signals that
The magnetic force FlII (newtons) that holds a single noise induces in flux sensitive dctectors may be filtered out.
paIticle to a magnetic flux leakage field is determined by the Magnetic flux leakage testing is therefore not limited by a
vector relation: human inability to distinguish real from apparent disconti-
nuities. It is limited by an electronic inability to perform the
Fill K(H e VB) (Eq. 1) same function.
 MAGNETIC LEAKAGE FIELD MEASUREMENTS I 183

PART 2
-
FLUX SENSITIVE DEVICES

Desclihed helow are flux sensitive devices used in mag- L the length of the wire behveen PQ and RS (meters);
netic nondestructive testing. The sensors detailed here and
measure either magnetic fields or their gradients. Llx the distance between position AA' and ee'
The inability to measure magnetic fields has seriously (meters).
hampered progress in magnetic particle testing. Hecent
research 10 indicates that the lack of discontinuity detection The magnetic f1ux interrupted by the wire is:
can be blamed on the magnetizing method, the particles
used, and the capability of the inspector. The important <P = 13· ~dA (Eq.3)
question that must be answered before beginning any
magnetic pmtic1e test is: what is the best possible combina- Where:
tion of 1nagllcti;:.illg method, particle shape, type and size,
and operator training that will reliably detect a discontinu- ~ = the interrupted magnetic flux (weber);
ity of a specific size every time? Commonly accepted ~ = the magnetic flux density (tesla); and
magnetization methods may not always be the best. Flux n = the unit vector for the area dA.
measurement de\ices can help prOvide more accurate in-
formation about the test procedure. The two equations tog~ther give:
Commonly used magnetic flux sensitive devices include:
(1) a long straight \'lire passing th~'ough a magnetic field, <P = LB· ~LlX (Eq.4)
(2) the-search coil, (3) search coil de11vatives such a C and E
cores, (4) the Hall element, (5) the magnetodiode, (6) the Faraday's law of induction states that an electromotive
ferroprobe Hnd (7) the nux gate magnetometer. For sensors force e will be induced in the wire and its magnitude is given
in categories 1 through 3, the output signal depends on by the relation:
some form of time vmiation for the amhient field strength.
Sensors in categOJies 4 through 7 are not time dependent.
A long straight wire passing through a magnetic field is
not llsed for nondestructive testing, but it is a crucial FIGURE 6. Wire cutting magnetic flux between
concept for understanding the Signal developed in coil AA' and CC'
sensors as they pass through magnetic flux leakage patterns.

Voltage Developed between the

Ends of a Straight Wire
p
As shO\Vll ill Fig. 6, two conducting \\ires PQ and RS are
placed at light angles to a magnetic Held (shaded area) of
constant flux densih' (B) directed towards the reader. Let R
another free wire A.1' be moved to positiun ec', a distance
Ax away. The area swept out by the \\ire is then:

dA = L~x (Eq.2)

Where:

dA ::::: the area swept out bv the moving wire (square

meters);
 184 I MAGNETIC PARTICLE TESTING

d<f> so that the value of v is known and Eg. 8 can then be wlit-
e = (Eg. 5) ten as:
dt

This is the rate at which the magnetic flux is cut. Eliminat- e = - v JB J.dl (Eg.9)
ing the flux between Egs. 4 and 5, and taking the compo-
nent of B perpendicular to ndA gives: where BJ. is the perpendicular component of the magnetic
field, such as the magnetic fllL'( leakage field shown in Fig. l.
The tangential flux density B t plays no role in the
e = - BL dx (Eg.6) development of the electromotive force in the conductor
dt
since sin 8 is zero for this field component.
Finally, since dxldt is actually the velocity v of the wire, The electromotive force developed between A and A I

the induced electromotive force becomes: appears across PR (Fig. 6) and can be measured with a
sensitive voltmeter. No current flows if P and R are not
e = - BLv (Eg. 7) connected. Furthermore, in the general case of conductor
motion through magnetic fields, the variation of B along the
As an example calculation, consider a truck traveling conductor must be known so that the integral of Eg. 9 can
north at 100 km-h -1. If the length of the truck's axle is be computed.
2 m (6.6 ft) and the vertical' component of the Earth's
magnetic field strength is 3 X 10-.5 Wb-m -2 (0.3 G), then Example of a Straight Wire Signal
from Eg. 7 the electromotive force between the ends of the
The electromotive force generated in the leading edge of
axle is:
the coil shown in Fig. 7 is deduced from the perpendicular
field component of a tight crack. The simplest approxima-
e = (3 X10- 5 Wb-m- 2 ) (2 m) ( 100,000 m-s- I ) tion for this magnetic flux leakage field is: 5 ,6
. 60 X 60
e = 1. 7 X 10 - 3 volt
(Eq. 10)
This example indicates the magnitude of voltages induced
'when metal objects move in relatively small magnetic fields .
. As another example, compute the electromotive force 'gen-
erated between the ends of a 10 mm long wire when moving FIGURE 7. Parallel and perpendicular coils
at 500 mm·s - 1 through a field of 1.6 X 10 - 3 Wb-m - 2 cutting magnetic flux leakage fields from a
(16,000 G). discontinuity at a speed v; for discontinuity
fields longer than the coil, the output of the coifs
is given by the formulas shown
e = (1. 6 X 10 - 3 Wb - m - 2) (0.01 m) (0.5 m - s ~ 1 )
e = 8 X 10 - 6 volt y
~
I
I
. It is unusual for B to be at right angles to v and under
such circumstances a more general form ofEq. 7 is required:

e = - J (v X B)dl (Eg.8)

where v X B is the vector cross product of the wire velocity

and the flux denSity through which it passes. Numerically,
this is (vB) sin 8, where 8 is the angle between v and B. The
integral is taken along the length of the wire because the Ell :< [B" - By,ILv
local value of B through which each segment of wire is E.l x [BYI - By21Lv
L = COIL LENGTH INTO PAGE
passing may vary.
In magnetic nondestructive testing, wires in the form of
coils are moved in a controlled fashion over a test surface
 MAGNETIC LEAKAGE FIELD MEASUREMENTS I 185

1. the value of BgLg (the magnetomotive force of the

and
discontinuity);
2. L (the length of the wire, provided that the aforemen-
(Eq. 11) tioned conditions are met); and
3. v (the relative velocity between the object and the
conductor).
Where:
Bg == the flux density deep 'within the discontinuity The dependence on liftoff can be seen by differentiating
(weher per square meter); Eq. 12 for the turning points (x o = ±h) and using these
Lg == the \vidth of the discontinuity (meter). values to compute the swing ~e in e. The result is:

The Oligin of coordinates is the mouth of the discontinu- (Eq. 13)

ity. If the length of the wire L is parallel to and shOlier than
the discontinuity opening, then the electromotive force
developed between the ends of the wire is taken from In this field approximation, the swing in voltage as the
Eqs. 10 and J 1 as: conductor passes through the magnetic flux leakage field is
inversely proportional to the liftoff.
c = (Eq. 12)
Simple Pickup Coils
In traditional magnetic flux leakage testing equipment
(see Nondestructive Testing Handbook Volume 4 on elec- Figure 7 shows tWo commonly used pickup coils: parallel
tromagnetic testing), the value of y is maintained at some and perpendicular. In some cases, the turns of th~se coils
constant value h (liftoff of the sensor). The form of the are wound onto small blocks of forrite to increase the value
electromotive f(:m.'e is shown in Fig. 8 for increasing value of of B.L above its air ~al~e (B.L is the flux leakage component
liftoff h. From Eq. '12, the magnitude of the electromotive perpendicular to the test surface). Air core coils are dis-
force is shown to depend linearly on: cussed below. .

Perpendicular Coil
FIGURE 8, Electromotive force develop'ed
betweer'! the ends of a conductor passing at With a one-tum coil passing at speed v through the same
constant speed through a leakage field such as magnetic flux leakage field as above, the signal electromo-
Equations 11 and 12 tive force is the difference between the two electromotive
forces developed in the upper and lower branches:
e
e = _ BgLgLv (x) [ 1 (Eq. 14)

I hi
2
11" x +

,J.. - x2 ~ h~ ]
t------l" LARGE LIFTOFF
where hI and hz are the liftoffs of the two branches.
If the coil has N turns and a width of 2b, then hI = H +
band hz = H - b (the liftoff is measured to the center of
the coil). The electromotive force then becomes:

- - - - - - - - - - - -____L-~__________ x
-Xo Xo
The results of varying hand b are shown in Fig. 9 where
the electromotive force is similar in form to that of the
186 I MAGNETIC PARTICLE TESTING

FIGURE 9. Form of the voltage signal developed FIGURE 10. Voltage induced in a parallel coil by
in a perpendicular coil when passing at constant passing it through a magnetic flux leakage field
speed through a leakage field of the form given such as in Equation 11; the coil voltage is the
by the simple Foerster relation difference between leading and trailing edge
signals
SIGNAL
VOLTAGE IN COIL (e)

, / RAW SIGNAL FROM LEADING

, , ,- \ EDGE OF COIL
It \
,\ \
I \ \ RAW SIGNAL FROM TRAILING
, ~ EDG£; OF COIL
\ \
I
\ \
1 \ \
,I
..
\ \

I
\
\
\
,
\ \
I
I
" \
'" ), COIL MOTION
1 , ,
I. ' .... ' .......

\ ,
\
\
\
\
\,1
\1
\ V
" , I1\ \JI
AFTER BRAY AND STANLEY. REPRINTED WITH PERMISSION.

straight wire. The turning points in the electromotive force

are given by the solution to Eq. 16. . the leading and trailing edge distances from the center of
the coil:

For example, when b = H/2, then the turning points are

Xo = ± O.74H and 2xo (the distance between the turning
points) is 1.49H. The swing in signal e is difficult to compute
in closed algebraic form.
The forrn of Eq. 18 is shown in Fig. 10. The dashec1lines
are voltages induced in the leading and trailing edges. The
solid line is their difference or the form of the electro-
Parallel Coil motive force. The signal consists of a major peak at X = 0
\Vhen the coil is oIiented so that one set of wires f()llows and two smaller side peaks. The roots of Eq. 18 occur at
another, then the output signal is the difference between Xo = ± (h 2 + b 2 ) liZ so that the distance between the points
the signals developed in the leading and trailing branches: at ,>vhich e = 0 is given by:

(Eq.17) (Eq.19)

Using Eq. 11 for the leading and trailing edge fields The maximum value of the coil signal occurs at x = 0 and
(B .LL and B .LT), substitute XL = X - b and Xl' = X + b for is propmtional to b/(h 2 + b2 ). Differentiation \\lith respect
 MAGNETIC LEAKAGE FIELD MEASUREMENTS I 187

the fact that fenites have velY low electlical conductivities,

FIGURE 11. Ferrite cored magnetic flux leakage minimizing detrimental eddy' current effects in them.
detector coil systems
Electronic Considerations for Coil Voltages
fa) It is essential that pickup coils are used to generate

,
voltages and not currents. Once a current is allowed to flow
in a coil, it creates its own magnetic field, one that can
interfere with the field under investigation. The output of
such coils is therefore generally fed to a high resistance
operational amplifier.
____T
.. 2fo
Coil Applications and Derivatives
Examples of the use of coils as detectors for magnetic flux
leakage are presented in the Nondestructive Testing Hand-
book on electromagnetic testing. Such coils can be connect-
ed in series adding, seIies opposing (a figure eight), over-
MAGNETIZED MATERIAL
lapping and many other configurations.
Search coils are often wound on ferIite cores to increase
the flux through them (Fig. 11 shows two common config-
urations). A detailed discussion of such sensors is given in
the electromagnetic testing volume.

fb)
Hall Element Sensors
1 3 to
Hall elements are crystals of semiconductor mateIial.
vVhen a current is passed through them and they are placed
~-J__~____--------------:::15mm
in a magnetic field, then a voltage develops across two of the
faces of the crystal. The voltage is.propOltional to the
t...=l=+=----iJ strength of the magnetic field.
A solid state gaussmeter is made up of the electronic
~__------2{ ------~.~I components needed to supply current to a Hall element, to
detect and measure the resulting yoltage and to then
convert it to the measured field value (see Fig. 15).

to b indicates that the coil signal is ma.ximized when b = h. Theory of Hall Element Operation
Thus when the half-width of the coil is equal to the liftoff,
the coil output voltage is ma.ximized with respect to the Electrically conducting solids are almost transparent to
magnetic flux leakage from the discontinuity. This argument the flow of conduction electrons, since the ions in the cIYstal
also indicates that this type of coil discriminates against lattice do not deflect conduction electrons as might be
relatively long range material sUlface noise such as might be expected from a typical billiard ball model. As current is fed
caused by local permeability variations. into one end of a crystal (see Fig, 12), electrons are
deflected toward one another or toward the other side of the
crystal, in accordance \\lith the Lorentz force F:
Ferrite Cores in Coils
Ferrites are useful in pickup coils because they not only F = - e (E + vx B) (Eq.20)
provide support for the wire turns but they also amplify the
flux density through the coil Windings by a value equal to the \Vhere:
effective permeability of the ferrite.
For small pieces of fenite (Fig. 11) where the dimension- e = electronic charge (coulombs);
al rati,o is smalL the effective permeability of the ferrite may E = electric field strength (volts per meter) on particle;
vary from the low teens to the thousands. The advantage of v = velOcity of particle (meters per second); and
Using fenite occurs not only in this amplification but also in B = applied flux density (tesla).
'88 / MAGNETIC PARTICLE TESTING

The term v X B is a vector cross product and is itself a

vector at right angles to both v and B. Its direction FIGURE 13. Typical Hall element probes: fa) flat,
determines which side of the crystal the electrons are (b) high linearity, (c) miniature, (d) subminiature
deflected toward. The theory of solid state physics provides and (e) axial (see Table 1)
a voltage V" across the crystaP of:
fa, 2 ,x 5 mm (0.08 x 0,2 in,)

(Eq.21) -----------]
____________ a

Where:

I = the applied current (amperes); Ib,

Bo; = the component of the applied field
at right angles 2,4 x 6,4 mm (0 I x 0,25 in,)

to the current (webers per square meter);

b = the thickness of the crystal in the direction of the
magnetic field (meter); and
~·-eI~~~~t~trj[-_~-]
Rh = the Hall coefficient (A - 1.S - 1). ~ BRASS HOLDER

Ie) 0,75 x 1.5 mm (0.Q3 x 0,06 in,)

FIGURE 12. Magnetic flux, drive current and Hall
effect voltage relationship ¥T"-~~-=-'~'-'---~,:7·:T~·7°--"-:.][:::~[:~::~~:~:~~
~ EPOXY COATED
(WITH SLIDE PROTEC TOR)

Id) 0,75 x 1.5 mm (003 x 0,06 in,)

~~~~~~~~§'~-~'jJ1::::::::::j .
. ~ EPOXY COATED '
(WITH SLIDE PROTECTOR)

Ie) 0,64 x 2 mm (0,025 x 0,08 in)

~ BRASS TUBE (NONMETALLIC OPTIONAL)

FROM LDJ ELECTRONICS. REPRINTED WITH PERMISSION. FROM LDJ ELECTRONICS. REPRINTED WITH PERMISSION.

TABLE 1. Specifications of typical Hall element probes (see Figure 13)

Hall Nominal
Output Current Temperature Operating
Voltage Control Coefficient Temperature
Probe Type (millivoltsJ (milliamperesJ (0C) (0C)

Flat or transverse 340 200 -0,1 -65 to 85

High linearity 350 350 -0.1 -65 to 85
Miniature 200 25 -0.25 -65 to 85
Subminiature 200 25 -0.25 -65 to 85
Axial 100 100 -0,1 -65 to 85
FROM LDJ ELECTRONICS. REPRINTED WITH PERMISSION.
 MAGNETIC LEAKAGE FIeLD MEASUREMENTS I 189

In general. if the Clystal is placed at an angle to the field FIGURE 14. Hall element configurations:
B, stich that 13 = B cos (), then the cosine of the angle must
0
(aJ sensor at a fixed distance from wire;
be found. Normally, the clystal is rotated until the maxi-
(bJ ferrite core; (cJ free-standing flux
mum gaussmcter reading is found. At that point, () is 0 since
concentrator; (dJ symmetrically positioned
cos () is l. contacting concentrator; and (e) asymmetric
The vallie t he Hall coefficient H" is determined by the
contacting concentrator
interaction of charge carriers 'with the clystal lattice. In
metal crystals, it is given by:
fa)
1
H" = (Eq.22)
ne

Where:

11 = the electron concentration; and

e = charge on the electron (-1.6 X 10- 19 coulombs).

Metals do not make the best Hall sensors because their

Hall coefficients are often relatively low. As can be seen in
Eq. 21, the larger the Hall coefficient, the larger the Hall
fb) I
voltage. Investigations of Hall coefficients for many sub-
stances have shown that combinations of elements from t
groups III and V of the periodic table give the highest Hall
voltages and have the least sensitivity.to changes in temper-
ature. Also, the charge carrier for these groups is more likely
to be a hole rather than an electron:

Excitation of Hall Elements

Where contacts occur between two dissimilar metals such
as the current and voltage attachments on the Hall crystal,
thermoelectric electromotive forces are generated. fe)
If direct current is used to excite the crystal, the voltage
read by circuih)' following the voltage contacts is the sum of
the Hall voltage and the thermoelechic voltage. For this
reason, Hall element clystal excitation is usually performed
with 25 to 350 mA altcmating current with frequencies in
kiloheltz.

Manufacture of Hall Elements rd)

Bulk Hall elements are generally bismuth doped semi-
conductors such as indium antimonide (InSb). These are
----_. "
produced by solid state crystal growth technology, cut into
small rectangular blocks and have current and voltage leads
attached before being encapsulated. Typical sizes are as
small as 0.8 mm (0.03 in.) long by 0.4 mm (0.015 in.) wide fe)
by 0.5 mm (0'()2 in.) thick. 2
Vapor deposited I-Ial1 elements have been reported for
use iI:
the testing of ball bearings by the magnetic flux
techmque. 11 In this application, bismuth was evaporated
onto an alumina substrate. A newer development is to ~ HALL ELEMENT
combine the Hall sensor, its power supply and an amplifier
on ?ne chip. Figure 13 and Table 1 show configurations of
typICal Hall sensors and their specifications.
'90 / MAGNETIC PARTICLE TESTING

FIGURE 15. Checking external field level with a FIGURE 16. Multidimensional arrays of Hall
meter after partial demagnetization elements used to measure directional
components of magnetic field intensity:
(a) two-dimensional array of Hall detectors;
components of magnetic field in XV plane are
sensed individually; (b) three-dimensional array
of Hall magnetic field detectors; each detector
senses the magnetic field component
perpendicular to the face of the semiconductor
fa)

Applications of Hall Elements

Hall elements are used in conjunction with gaussmeters
or other devices to detect or measure magnetic fields.
Typical configurations are shown in Fig. 14. In Fig. 14a, the
Hall sensor is held a fixed distance from a current carrying fbI
wire and the gaussmeter measures the field strength created

---~.
~
by the current. In the case of pulsed currents, the peak
current can be measured with a peak reading gaussmeter .. -
I EHZ - RHlcB z
.1ri Fig. 14b, a ferri'te ring is added to measure small Helcls
or currents. The high permeability of the ferrite aids. in Bz
creating a high B value in the vicinity of the sensor's active
area. Figure 14c, 14d and 14e show combinations of Hall
elements and ferrite flux concentrator configurations used
in magnetic flux leakage testing.
The level of external field just outside a partially demag-
netized material may best be measured with a Hall element
meter. Figure 15 shows an inspector checking the external
field level with a gauss meter after partial demagnetization of
the test object, a 270 mm (10.75 in.) diameter steel tube.
Crossed Hall elements can also be used, as shown in
Fig: 16. Such configurations are used to check welds or to
reconstruct the total field from the measured components. l2 Applications of Magnetodiodes
Figure 20 shows the use of magnetodiodes for detecting
magnetic flux leakage from discontinuities in tubes. 2 The
Magnetodiodes magnetic Hux leakage is excited by alternating current
electromagnets arranged to detect either internal or exter-
The magnetodiode is a solid state device whose resistance nal surface breaking discontinuities. The system illustrates
changes with field strength. l .3 The device consists of 17 and the general principles of magnetic flux leakage testing.
11 zones within a semiconductor, separated by a region of Sensors are connected differentially to eliminate Signals
material that has been modified to create a recombination from the applied field and from relatively long range
zone (Fig. 17). Its frequency response (Figs. 18 and 19) is vaIiations in surface field strength. This system and mag-
flat from direct current to 3 kHz and the device is stable netic flux leakage systems like it are used to rapidly evaluate
without temperature dependence from - 10 to 50°C (15 to the surface condition of tubes and can detect tight discon-
120 OF). tinuities with a depth of only 0.1 mm (0.004 in.).
 MAGNETIC LEAKAGE FIELD MEASUREMENTS I 191

FIGURE 17. Diagram of the magnetodiode FIGURE 19. Characteristics of the magnetodiode:
showing p-zones and n-zones in semiconductor raj frequency response and (bl temperature
material along with intrinsic and recombination dependence
la) T.. = 25°C
INTRINSIC ZONE

~~
, ~
~

0.02 0.04 0.1 0.2 0.4 I 2 4 10 20 40

MAGNETIC FIELD FREOUENCY
RECOMBINATION ZONE
(kilohertz)

H- H+
Ib) E = 6V
H = ±80kA'm- I (± 103 Oe)
LU
\J
::;- :s
:! 1.5 : - ~
V"' 30 g
t? ~
FIGURE 18. Typical characteristic curve showing
that the response of a magnetodiode is linear to
about a 40 kA·m - 1 {500 Oe J field
.~
~
I-
2
I-
1.0

05
10- ~ aV f- lo- - :-..
r--....
2.0

1.0
Z
w

2
C>:
LU
:::l I-
0 Z
LU
U
- 50 - 30 - 10 0 10 20 30 40 50 60 70 80
AMBIENT TEMPERATURE (OC)

Ferroprobes
F erroprobes (also called Foerster 11licroprobes) take many
forms but for the purposes of nondestructive testing they
generally consist of cylindrical or rectangular ferrite upon
which one or two coils are wound (Fig. ll).
Flux gate magnetometers are used to detect small chang-
es in the Earth's magnetic field. As might be used by
geophysical prospectors, these devices consist of ferrite
rings carrying many coil configurations.
Both of these devices are based on the same physical laws
as a tape recorder head or any other ferrite cored magnetic
field pickup. The difference between the two is that
-200 -160 ~ 120 -80 -40 0 40 80 120 160 200
ferroprobes are activated at high frequency.
MAGNETIC FIELD STRENGTH Typically, one coil is excited \vith alternating current at a
(kiloamperes per meter)
frequency f. The voltage induced in a second coil at
frequency 2f is then detected. This secondary signal carries
information about the scanned magnetic flux leakage field.
Figure 21 is an example of the tangential magnetic flux
Magnetic particle testing is often used to inspect such leakage field taken with such a probe over an angle slot in
tubes but while it is extremely sensitive to outer surface, residual induction at a liftoff of 1 mm (0.04 in.).
tight discontinuities, its use fOl: inner surface discontinuities Ferrite cores might be solid or hollow, to reduce eddy
requires the use of a vie",ring de\rice. currents in the ferrite.
 192 I MAGNETIC PARTICLE TESTING

FIGURE 20. A magnetodiode testing system for tubes: (a) alternating current magnetizing method and
(b) electrical block diagram

fa) fb) PEN

RECORDER

MAGNETIZATION
ELECTRIC SOURCE

DETECTION DETECTION DETECTION r~"···-·~

STAND-I STAND-2 STAND-3 '-,.-.-:.
A, A2A) C, C2 C3 E, Ez E3
B, B2 B3 D, DlD, F, F2 F3 MARKING SElECTING
DRIVER CRADLE
1:::u4 1=:::u4 1==···4

HOT ROLLED STEEL BAR

MARKER CONTROLLER

SUMITOMO METAL INDUSTRIES. REPRINTED WITH PERMISSION.

Beeause the relative magnetic permeability of air is 1, this

Large Volume' Magnetic reading is also the numerieal equivalent· of the' magnetic
Field Indicators field strength of air. .
The magnetic field indicator is also used to determine the
existence of a magnetic field external to a ferromagnetic
Bulk Field Indicators object. To do this, the indicator is O1iented against the
The field measurement systems discussed above are object's surfaee and moved until it registers the maximum
deSigned and used for the assessment of magnetic leakage external field reading.
fields from material discontinuities. In all cases, the active
sensing area of such a device is very small. In the case of the
Hall element, which is rectangular in shape, it is possible to
Bulk Field Indicator Construction
integrate the field over the active area of the Hall crystal,
and so compensate for it. Then, by taking measurements at Many magnetic field indicators are round (see Fig. 22),
controlled distances above a magnetized surface, it is pos- about 64 mm (2.5 in.) in diameter with thicknesses of 13 to
sible to extrapolate the field values to that at the smface .. 25 mm (0.5 to 1 in.). The indieators eommonly ·used for
Once this is done, the electromagnetic boundmy conditions cheeking external fIeld levels after magnetic particle tests
indicate the magnetic field strength just inside the surface. have a range of about 1 to 2 mT (10 to 20 G) in divisions of
The following text relates to the detection or measure- 0.05 to 0.1 mT (0.5 to 1 G). Positive readings are n01th and
ment of magnetic fields over much larger areas, since the negative readings are south.
active area of the sensor is larger than that used for leakage A key component of the magnetic field indicator is a small
field testing. The instruments are handheld, moving magnet movable field sensing magnet. The magnet is mounted so it
sensors used to measure bulk external fields at a relatively is free to rotate, Its angular deflection is shown by the
high liftoff from the test object. They are often used as a movement of a pointer.
practical check on the external demagnetization state of an A second key eomponent is a fixed permanent magnet. Its
object. Their use to detect magnetic field strength within magnetic field strength limits the useful range of the unit by
coils should be discouraged, since the coil field may proViding a restraining force to prevent the sensing magnet
remagnetize the moving magneties. Note that these devices from rotating freely. With no external magnetic field, these
do not measure leakage fields from discontinuities. two magnets stay antiparallel to each other and the pointer
A bulk magnetic field indicator can be used to measure remains in a neutral pOSition, registeling a zero reading
the value of a uniform magnetie induction field in air. (Fig. 22).

"~"~'l "~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~'"~"~"~'l "~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~"~

 MAGNETIC LEAKAGE FJELD MEASUREMENTS I 193

FIGURE 21. Tangential magnetic flux leakage FIGURE 22. Diagram of a typical magnetic field
fields in saturated residual induction over a indicator; distance from the pivoting point of
40 degree slot; curve 1 is from experimental data the sensing magnet to the closest point on the
at 1 rom (0.04 in.) liftoff; curve 2 is a model with edge of the casing is about 18 mm (0.75 in.); the
increased charge on acute face of slot size, shape, material, magnetic field strength and
relative positions of the sensing magnet and the
reference magnet vary with manufacturer

~REES
0.4 (4)

0.2 12)

o~~~----~~-+----;-----
PLAN VIEW

-0.2 {-2)

I
iii ~

~
-0.8 {- 8) '--__-'-_ _ _'--_--..L._ _......._ _
-4 8 SIDE VIEW

x
(millimeters) I
Sj I IN
AFTER ZATSEPIN AND SHCHERBININ. 7

The field indicator is designed so that the net magnetic

field from these two magnets is weak outside the device.
~lacing the indicator on an unmagnetized object does not
mduce poles on the object sufficient for causing inaccurate magnetic field in the same area, the field sensing magnet
meter readings. rotates into a direction parallel to the resulting, combined
In principle, magnetic field indicators could use a coiled magnetic field.
spring instead of a calibrated magnet to return the pointer The instrument's pointer is attached to the sensing
to zero once the external field is removed. However, slight magnet and correspondingly rotates into a direction perpen-
ch~nges in the sensing magnet's strength would then re- dicular to the resulting field. If the external field changes
qUIre recalibration of the unit. Also, a strongly poled sensing polarity, the pointer rotates in the opposite direction and the
magne.t, when placed very close to an unmagnetized object, reading's algebraic sign changes. If the magnetic field
could mduce localized poles and cause inaccurate readings. indicator is rotated through 180 degrees about its painter's
zero direction, there is no algebraic sign change in the
Principles of Field Indicator Operation reading (the scale is also rotated 180 degrees).
However, in practice a reverse of polarity or rotation of
A fixed magnet inside the device's housing sets up a the indicator often produces a change in a reading's mag-
reference magnetic field. A small movable sensing magnet is nitude unless the external field is perpendicular to the
mounted inside this field. If a nearby object also sets up a reference field.
 '94 / MAGNETIC PARTICLE TESTING

FIGURE 23. Pointer deflection 8 in first FIGURE 24. Pointer deflection a and fJ when B is
calibration type; B is perpendicular to 8* and not perpendicular to B* (a < 8 < PJ
tan (J = B/B*
POINTER'S
INITIAL/ZERO
POSITION
!
, I POINTER'S NEW
" ,: i 8 /OSITION DUE TO 8
-- ------...-,--- -
~. --/-~---...--- --- ... '~ /
I'
I
I
I
I SENSING MAGNET'S
__ c....._____ -L-~:::.I,~ INITIAL POSITION
8* ,I!'': - - ---
/ I '
/ I "
I / I ~~~~~~~'g"JE TO 8
I /

Calibration and Use of Field Indicators

Generally, there are two ways to calibrate magnetic field ,FIGURE 25. Pointer d~flection l/> In second
indicators. These calibration methods provide two distinct calibration type; B and the pointer are parallel
ways of using the devices. .and tP is at ~ maximum.(Q* > B cUld q, > 8)
The most common calibration method correlates the POINTER'S
angular deflection e of the indicator's pointer with the INITIAL IZERO
POSITION
magnitude B of a uniform external field whose direction is
parallel to the zero direction of the pointer. To measure a
I
"" i POINTER'S ~X~ti~C'~'g~
uniform field, the field indicator is positioned so that the ~ . /
zero direction of the pointer is parallel to the field. --- T:::::'' = ~,-:::\:-- --- - -- :;::=-- -!-- --", /
Used in this manner, the field B from an external object I ' ........ ' "'). ,I . . " /.
is perpendicular to the reference field B* inside the mag- ' . . ,\ ----------·1--------- 8
netic field indicator. Or as depicted in Fig. 23: /
~ ....
.... '\ .
/~~ ", 1

B = B* tan (J (Eq. 23) ~

~

"
I" "
For a small deflection, B* may be considered uniform and - -O'"'~___- - - - - ' - -......-_r:.r.:
a large (J indicates a relatively strong B. To keep within a 8* ~
practical calibrated scale when (J is between + 45 and - 45 "-
~
degrees, the measured field must be weaker than the ~
~
reference field (B less than B*). SENSING MAGNET'S
In many applications, B may not be perpendicular to B*. NEW POSITION

For example, the direction of B may be unknown or the

field could be nonuniform. In a case like this, the magnetic
field indicator is positioned against the object's surface and
oriented in such a way that the directional marking on the The alternate way of calibrating correlates the uniform
device's casing is near to and perpendicular to the object's field value B with the maximum deflection ¢ of the
surface. Because the field is not necessarily normal to the magnetic field indicator's pointer, .The value of 4> is obtained
object's smface, the reading can be less or greater than the by orienting the instrument inside the field B. As shown in
actual value, depending on whether the field makes an acute Fig. 25, the field vector B changes its direction relative to
or obtuse angle with the reference field (see ll' and f3 in the field vector B* and their resulting vector traces out a
Fig. 24). circular path of radius B centered at the tip of B*.
 MAGNETIC LEAKAGE FJELD MEASUREMENTS I 195

For B less than B* and for a maximum deflection 4>, the not attract or deflect magnetic materials dming normal use
resulting held vector is tangential to this circular path. It unless surface discontinuities occur, producing strong exter-
follows that the field B is actually parallel to the pointer nal poles.
(Fig. 25). When B < B*, Eq. 24 is valid. Consider a cylinder with flat ends that has been longitu-
dinally magnetized. A magnetic field indicator is placed
B = B* sin 4> (Eq.24) against the cylinder's end surface and the directional mark-
ing on the indicator's casing is lined up vvith the length of
Note the differences between Eqs. 23 and 24. When the test object. The indicator is aligned normal to the
measUling very vveak fields, these two calibration methods cylinder's surface and a reading is taken.
are about the same and the pointer's angular deflection is .If a demagnetization procedure has been properly per-
approximately linear with the uniform field's magnitude formed, the indicator reading will be 0.1 mT (1 G) or less.
(0 = B/B* = cP for a small BIB*). In general, for the same Similar readings obtained on the side of the cylinder (with
uniform field B: the directional marking perpendicular to the side surface)
should be about zero. Sometimes, if residual magnetism is
4»(} (Eq.25) high, a nonmagnetic spacer is placed between the object
and the magnetic field indicator and relative readings are
As a consequence of this inequality, the scale of the obtained.
second type of calibration is generally wider if marked on For a cylinder that is large compared to the indicator's
the same arc inside the same magnetic field indicator. When size, measurements made at the center of the end surface
using an inshument that is calibrated in the second way, the are close to the actual values immediately beneath the
.unit. is often rotated to verify that the readings are maxi- sui-face of the object. The reason for this accuracy is that the
mized. Sometimes this is inconvenient for objects with magnetic fluxes immediately inside and outside the cylin-
. complicated g~ometry because a rotation of the device' may aer's' end are perpendicular to the end surface and' the
move its sensing magnet away from the area of interest. perpendicular component of the magnetic induction field
If magnetic .field indicators of the first type are used' as if across the boundary surface is continuous, according to
they· had the second type of calibration (maximizing their electromagnetic field theory.
readings by rotation) then the resulting maximum values are However, the same accuracy is not possible for geome-
actually greater than the true field values. Sometimes, in tries with sharp comers or for objects that are small
this way, an estimate can be made for the size of a uniform compared to the size of the indicator. In these instances, the
or nea;'ly uniform field, even though its direction is un- measured field may not be uniform (the direction and
known. density of the flux lines vary across a small distance) and a
When measming a nonuniform field, the reading of a tangential component exists. It is known from electromag-
magnetic field indicator is at best the average field value netic field theory that the tangential component of magnetic
over the area covered by the sensing magnet. For example, induction may not be continuous when crossing the bound-
assume that the flux lines from an external magnetic source my surface between two media of different permeabilities.
are almost parallel with the special directional marking on Therefore, the magnetic induction inside and outside the
the magnetic field indicator. The two ends of the magnetic object may not be the same.
field indicator's field sensing magnet may still expelience When the test object has an irregular shape or the
deflection forces of different magnitudes because of differ- residual field readings are large, one way to test extemal
ent field values and the resulting pointer deflection is an magnetism is to scan the entire surface of the object with a
average of the two field values. magnetic field indicator. Maximum readings occur at loca-
tions where significant external poles exist.
At such maximum reading locations, the field indicator
Measuring Residual FieldS can be used to determine if the field is normal to the
The primary function of a field indicator is to measure the object's surface. The field indicator is positioned against the
external magnetic field strength close to an object, but not object and oriented with its directional marking normal to
~very kind of residual magnetism can be detected by these the surface. The device is rotated through 180 degrees
lllstruments. about the directional marking (normal to the object's sur-
. In a circularly magnetized object, where residual flux face). During rotation, variations of the indicator reading
lmes are circumferential and form closed loops inside the are noted.
matelial, the induction is Significantly different from zero If the rotation does not affect the indicator reading, then
but .the field may not produce poles outside the' object. Such the reading is the tme field value and the true field is
a clfcular field does not produce Significant readings in normal to the object's surface at this location. If the field is
magnetic field indicators. Circularly magnetized objects do not perpendicular to the object's surface, it is likely that, at
196 / MAGNETIC PARTICLE TESTING

a ceItain time during rotation, the actual field vector will TABLE 2. Nearly uniform magnetic field values for a
have no projection along the direction of the indicator's five-turn coil carrying direct current, compared to
reference field, and the reading at that time will be exactly the linear distance from the coil center
the component of the field normal to the object's surface. In
Distance from Coil Center Measured Value
other words, the normal component of the field at this meters (feet) millitesla (gauss)
location will be no greater than the largest value observed in
the rotation. More precisely, it is in between the readings o 31 (309)
obtained at the beginning and at the end of the rotation and 0.45 (1.5) 1.0 (9.8)
is no greater than the average of these two values. If the 0.9 (3.0) 0.14 (1.4)
same value appears twice during rotation, then it must be 1.0 (3.3) 0.1 (1.0)
the normal component of the field.
In most applications, the purpose of external residual
field measurement is to ensure that the objects are free of and connected in series-aiding mode. In about 30 percent of
magnetic poles that detract from serviceability. The exact the volume between the two coils, there is a very uniform
value of the external residual magnetism is not critical, so magnetic field parallel to their axes. The field value can
long as it is lower than a limit predetermined by the user's either be measured with an appropriate meter or calculated
empirical data. from the coils' dimensions and the value of applied direct
current (in Eq. 26, xlR is 0.5 and Bo is replaced with 2Bo).
Checking Indicator Reading Accuracy In addition to the Helmholtz coil or commercial calibra-
If inconsistent results occur in different sets of magnetic tion fixtures, an approximately uniform magnetic field may
field indicators, it is likely that some of the devices are be established using a large direct current coil. Over a small
malfunctioning. Certain indicator malfunctions are easy to distance along the coil's axis, a. magnetic field can be
detect, such as an imbalanced or damaged pointer, or considered nearly uniform. As examples, Table 2 shows a set
mechanical failures at the support of the sensing magnet of magnetic field values for· a five-tum coil of 300 mm
and pointer assembly. High mechanical impact or sudden (12 in.) diameter carrying 1,500 A direct current. The table
exposure to a ~tron,g magnetic field are among the less values'were calculated using the following·e'quations.
obvious causes for 'erratic readings.
A magnetic fielq. indicator can also become inaccurate if
B
its magnetization is changed by exposure to a strong direct
current or a decaying alternating current field. If the fixed (Eq. 26)
reference magnets become partially demagnetized, the unit
can give readings much larger than a good unit's results
(smaller B* in Eq,23). If an indicator's magnetic compo- Where:
nents are totally demagnetized, its pointer may not return to
the zero position, remaining virtually anywhere on the scale. Bo = magnetic field at the center of the coil (millitesla);
Two different field indicators may give different results at x = distance from the center of coil along the axis
the same location on the test object. However, these (meters); and
differences alone do not indicate that one of the field R = coil radius (meters).
indicators is malfu~ctioning. The sensing magnets of differ-
ent devices may have different sizes and their location inside
the units may be different. They may therefore not be
measuring the field at exactly the same location. In addition, (Eq.27)
reference fields inside the units may also differ.
In a highly nonuniform field, readings may not vary in the I

same ratio as varying measurement locations. As a result, it Where:

may not be possible to verify the accuracy of a magnetic
field indicator by comparing its readings to a known refer- = the permeability constant (417" X 10 -7);
f..L(J
ence unit (other field measurement devices may be equally N = number of turns in the coil; and
inaccurate in the nonuniform field). The best way to test the I = applied direct current (amperes).
accuracy of a particular device is to perform reference
comparisons in uniform or nearly uniform magnetic fields. Sometimes, the Earth's magnetic field can indicate a
To set up a uniform magnetic field for calibration, a meter's accuracy: the Emth's field is about 0.0.5 mT (0 ..5 G).
Helmholtz coil may be used. This device contains two If the indicator's accuracy is within ± 0.03 mT (0.3 G), then
parallel coils separated at a distance equal to their radius with proper north/south and hOlizontal orientations, the
 MAGNETIC LEAKAGE FIELD MEASUREMENTS / 197

device should be able to register an approximate reading of

the Ealth's field, provided there are no other magnetic In cases where the field direction is uncertain, the
objects nearby. indicator may be rotated about its directional marking,
The best magnetic field indicators are precision calibrat- which is in turn positioned normal to the object's surface.
ed. Their accuracy may also be less susceptible to the The rotation moves through 180 degrees to get a maximum
influence of a strong magnetic field. In some applications, reading. The component of the magnetic field normal to the
less costly magnetic field indicators may be used to do object's surface is no greater than the maximum value
measurements. Precision calibrated units are then used as registered during rotation and no greater than the average
reference standards, verif)ring the readings of the less costly of the readings at the beginning and the end of rotation. If
devices. Periodically, the reference devices are returned to the rotation does not affect the reading, then the field is
the manufacturers for calibration. perpendicular to the object's surface and the reading is the
field's true value .
Conclusion
. For a nonuniform field, the reading of the magnetic field
A magnetic field indicator is a convenient low cost tool for indicator is an average value (at the spot where the indica-
measuring the residual external field strength of ferro- tor's field senSing magnet is located).
magnetic objects. To measure a uniform field, these units Good magnetic field indicators have sound mechanical
are often calibrated in a way that requires the operator to supports for their reading pOinters and these supports
align a special directional marking (line or arrow on the cannot be' easily damaged. Their magnetic components
device's casing) with the field's known direction. cannot be'. eaSily demagnetized by strong external fields.
For an external flux measurement, the field indicator is Also, they do not induce Significant magnetic poles on the
positioned against the object with its directional marking objects they test.
near to and perpendicular to the object's surface. This For an accurate calibration ·of a field indicator, a uniform
positioning i~ hased on the fact that flux lines are eJ\:pected magnetic field may be provided by a Helmholtz coil. For a
to be pelpeI1dicular to the object's surface at the location of quick check of calibration, various aEprmdmate uniform
~ignificant l)oles.
field values' along the axi~ of a large direct current coil may
be used.
198 I MAGNETIC PARTICLE TESTING

REFERENCES
1. Bray, D.E. and RK. Stanley. Nondestructive Evalua- 7. Zatsepin, N. and V. Shcherbinin. "Calculation of the
tion - A Toolfor Design, Manufacturing and Service. Magnetostatic Field of Surface Defects: Part 1, Field
New York, NY: McGraw-Hill Publishing (1989). Topography of Defect Models" and "Part 2, Experi-
2. Electromagnetic Testing: Eddy Current, Flux Leakage mental Verification of the Principal Theoretical Rela-
and Microwave Nondestructive Testing. Nondestruc- tionships." Defektoskopiya. No.5 (1966): p 50-65.
tive Testing Handbook, second edition. Vol. 4. R
McMaster, P. McIntire, M. Mester, eds. Columbus, 8. Heath, Scott. Master of Science thesis (unpublished).
OH: The American Society for Nondestructive Testing Fort Collins, CO: University of Colorado (1983).
(1986) .
9. Swartzendruber, L. "Magnetic Leakage and Force
3. Stanley, R "Basic Principles of Magnetic Flux Leak-
Fields for Artificial Defects in Magnetic Particle Test
age Inspection Systems." Electromagnetic Methods of
Rings." Proceedings of the Twelfth Symposium on
Nondestmctive Testing. W. Lord, e~. New York, NY:
NDE. San Antonio, TX: Southwest Research Institute
Gordon and Breach (1985). (1970).
4. Hwang, Jackson. Defect Characterization by Magnetic
Leakage Fields (unpublished). PhD.thesis. Fort Col- 10. Skeie, K. and D. Hagemaier. "Quantifying Magnetic
lins, CO: Colorado State University (1975). Particle Inspection." Materials Evaluation. V?l. 46,
5. Foerster, Friedrich. "Nondestructive Inspection by No.6. Columbus, OH: The American Society for
the Method of Magnetic Leakage Fields: Theoretical Nondestructive Testing (May (1988): p 779.
and Experimental FouQdations of the Detection of
Surface Cracks of Finite and Infinite Depth." Defek- 11·.. Beissner, R, G. Matzkanin and C. Teller. NDE Ap-
toskopiya. Volume 11 (1982): p 3-25. plications of Magnetic Leakage Field Methods; a State
6:' Foerster, Friedrich. "On the Way from 'Know How' to of the Art Survey. NTIAC-80-1. San Antonio, TX:
'Know Why' in the Magnetic Leakage Field Method of Southwest Research Institute (1980).
Nondestructive Testing" (Part 1). Materials Evalua- 12. "Hall Effect Transducers: How to Apply Them as
tion. Vol. 43, No. 10. Columbus, OH: The American Sensors." Freeport, IL: MicroSwitch Company (1982).
Society for Nondestructive Testing (September 1985):
p 1,154. See also Part 2, Volume 43, No. 11 (October 13. "What is the Sony Magnetodiode?" New York, NY:
1985): p 1,398. Sony Corporation of America.
 SECTION 8
MAGNETIC PARTICLES AND
PARTICLE APPLICATION
Bruce Graham,· Magnaflux Corporation, Chicago, Illinois

.. . .
200 I MAGNETIC PARTICLE TESTING

INTRODUCTION

The magnetic particle technique provides a test indica- magnetic particle characteristics, the magnetization meth-
tion that is located very near actual material discontinuities, od, the level of magnetization, test environment lighting
describing their size and shape on the test object surface. In intensity and inspector training.
this important way, the magnetic particle method differs This chapter focuses on: the importance of the magnetic
from most other nondestructive tests. In other techniques, particle's characteristics, test indication contrast, mainte-
test indications are typically produced in a medium separate nance and handling, and specification requirements. There
from the test object, as an oscilloscope trace, through an are two types of magnetic particles in general commercial
acoustic transducer or on radiographic film, for example. By use: dry particles and wet particles. Dry particles are applied
outlining and precisely locating the discontinuity, magnetic to the test surface as a solid suspension or as a cloud in air.
particle indications are comparatively simple to interpret. Wet particles are applied as a suspension of particles in a
The magnetic particle method is best used for locating liquid vehicle;either oil or water. Each method is discussed
small surface discontinuities on ferromagnetic test objects. here, with details of their contrast, color type, fluorescence
The technique may also indicate the presence of certain and some general information about the appropriate indus-
linear discontinuities slightly below an object's surface. The try standards.
absolute sensitivity of the magnetic particle method has not Note that the Nondestructive Testing Handbook uses Sf
been firmly established, although this text discusses several units of measure followed by American standard units.
applications that demonstrate how fine a crack may be When quoting from us testing specifications, the Sf units
reliably detected with careful testing procedures. In practi- are often proVided for consistency in the text but are not
cal situatio.ns, absolute sensitivity i~ not as important as the necessarily part of the original document's language. Con-
probability of detection. This probability is a product of the versions are typically rounded.
 MAGNETIC PARTICLES AND PARTICLE APPLICATION / 201

PART 1
DRY METHOD TESTING MATERIALS

Fine dry particles that can pass through a US Standard 270

Dry Method Particle Characteristics mesh sieve of 50 J-tm (0.002 in.) show much greater sensi-
tivity to small discontinuities and their small leakage fields
Particles used in dry method magnetic particle testing are when compared to 100 mesh particles of 150 J-tm (0.006 in.).
manufactured to emphasize three sets of phYSical proper- The coarser particles, nearly three times the diameter of the
ties: (1) magnetic characteristics, (2) size and shape and 270 mesh particles, are more than twenty times heavier and
(3) visibility. are too large to be held by weak leakage fields.
However, for two reasons, a dry testing powder cannot be
Magnetic Properties made exclusively of small magnetic particles: (1) small
Nearly aU dl)' magnetic powders are finely divided iron particles adhere to all surface anomalies (traces of oil,
particles coated "'ith pigments. These ferrous materials are dampness, fingerprints or roughness\ producing dense
chosen to prOvide the characteristics critical to magnetic particle backgrounds; and (2) the testing environment be-
particle test procedures. Primary arilong these are low comes extremely dusty and unsafe for the. inspectors. As a
coercive forc6, .low magnetic retentivity and high mugnetic result, a practical dry magnetic powder contains a range of
permeability. . particle sizes. Small particles are needed for sensitivity to
Low retentivity increases the dry powder's ability to fine discontinuities. Coarser particles are needed to bridge
clearly indicate disconti~uities. If the particles retain mag- large discontinuities and to diminish the powder's dusty
netism (high magnetic retentivity), they adhere to each nature. In addition, large particles can help reduce masking
other and cannot be properly applied. When such particles by dislodging the background often formed by fine particles.
reach the test object, they adhere to its surface and cause an Particle shape also has important effects on background,
intense background, masking discontinuity indications. particle application and test results. Elongated particles
High magnetic permeability also increases the powder's (large length-to-diameter ratios) are easily attracted to
ability to indicate discontinuities. Particles with high per- leakage fields. When compared to particles of equal perme-
meability are easily attracted to the small magnetic leakage ability but compact shape, long particles more readily form
fields from discontinuities, where they are trapped and linear discontinuity indications, possibly because of their
retained for intelpretation. ability to eaSily achieve magnetic polarity. However, elon-
The concentration of the magnetic material in dry pow- gated particles cannot be used alone because of their
ders is a fourth cdtical consideration. Increasing the amount tendency to mat and form clusters that cannot be easily
of magnetically inert pigment in a dry powder composition applied.
naturally lowers its magnetic sensitivity. Dry magnetic pow- Particles with a compact shape flow easily, have high
ders are therefore manufactured as a compromise between mobility and can be simply dispersed into clouds for proper
tI~e pdmal), need for sensitivity and the secondary need for application. As a result, the most sensitive dry powders
hIgh viSibility. contain both shapes, in a ratio determined by the testing
application and empirical data.
Size and Shape In AMS 3040, the upper size limit for sensitive dry
Particle size and shape are to some extent more critical particles is about 80 mesh or 180 J-tm (0.007 in.). Larger
than magnetic permeability for achieving sensitivity and particles can plug powder applicators and do not add
e.ase of use in a dry powder. Magnetic powders are not sensitivity.
SImply an aggregate of metallic filings. The particles are Not all dry powder test procedures require high sensitiv-
~ade from carefully selected magnetic matelials of specific ity. In some applications, high sensitivity is actually detri-
SIze: shape, magnetic permeability and retentivity. They are mental to the testing procedure. In these circumstances,
des~gned to be used in air, not to be mixed with a liqUid size and shape are less critical and the most important
velucle. Reclaiming dry powders is not recommended. particle characteristic is low residual magnetism.
 202 I MAGNETIC PARTICLE TESTING

Visibility and Contrast continuously activated while the inspector applies the powl
der and removes the excess).
Dry method magnetic testing powders are commercially Automatic processing has been used for testing lineai
available with three types of colors: (1) visible colors for
welds on large diameter pipe. However, most .weldedl
viewing under white or visible light; (2) fluorescent colors
stmctures are shaped in ways that make them difficult to
for viewing under ultraviolet light; and (3) daylight fluores-
handle automatically. In addition, much weld testing is done I
cent colors. The daylight fluorescent powders fluoresce on site and only portable testing systems can be used.
brightly in visible light, greatly amplifying their visible color. Direct current magnetization is usually preferred for weld I
The visible light powders are normally available in gray, testing because it penetrates more deeply and allows the
red, black yellow, blue and metallic pigments. This range of indication of slightly subsurface linear discontinuities. Half- I
colors allows the user to choose the one that contrasts most wave direct current also has the' advantage of providing
strongly with the test object surface. increased particle mobility and increased potential for
Fluorescent colors are seldom used in dry powder appli- forming accurate discontinuity indications while redUcing
cations, partly because dry powder tests are typically per- background.
formed on site or on large stmctures, making it difficult or
impossible to enclose and darken the testing area. In
addition, fluorescent dry powders often produce back- Tests of Castings
ground that is bright enough to be objectionable. Dry particles are particularly useful for magnetic pmiicle
Daylight fluorescent colors are occasionally used when a testing of large castings. Cast objects are normally tested
test indication's high visibility or contrast is more important using prods or yokes, with the test covering small overlap-
than absolute sensitivity (fluorescent color augments the ping areas. Large portable power supplies may be used. In
visible color and indications are extremely bright and these applications, the magnetization equipment is set 'up in
visible). With these powders, there is no need for an advance, with connectors for through current shots firmly
enclosed, darkened test environment. In fact, the brighter clamped in place and coils or looped cables wound where
the ambient light, the brighter the indications. Daylight, needed. The current is applied throughout powder applica-
even· in deep shadow, excites their fluorescence, as does tion making this a continuous method. Because the test
blue mercury vapor lighting or the light from white fluores- procedure may take several minutes .on a large castinga all
cent tubes. Ordinary incandescent lights work less effective- test system cabling must be'properly filted to avoid excessive
ly than white or blue light sources. However, the yellow light Iteating.
from sodium vapor lights does not excite fluorescence in The choi~e of magnetizing current depends on the type of
daylight powders and cannot be used for this application. discontinuities being sought. ,. As a rule, direct current is
In some applications, particle contrast may be enhanced recommended for weld testing because the direct current
by coating the' test object surface with a thin white lacquer. magnetic field penetrates deeper into the test object,
The characteristics of the lacquer then become important providing some subsurface detection capability.
considerations for the magnetic particle inspector. The
white surface coating must dry immediately, so as not to
slow the testing procedure and it must be eaSily removed
after the test to avoid delays and additional costs. Application of Dry Magnetic
Particles
Dry Particle Uses Particle Applicators
DIy magnetic particles are often used to test welds or Dry magnetic powders must be applied in a gentle stream
castings for surface discontinuities. Dly powder is not or cloud. If applied too rapidly, the stream may dislodge
recommended for the detection of fine discontinuities such indications already formed and v-.rill certainly build up too
as fatigue cracks or grinding checks. Wet method testing is much background for easy removal or interpretation. If
much more sensitive to small surface discontinuities. applied with too much velocity, typical particles gain too
much momentum to be reliably trapped by discontinuity
leakage fields. Manual and mechanized powder applicators
Tests of Welds
can help provide proper density and speed of particle
In weld testing, the typical magnetic particle technique application.
uses prods or yokes, with the inspector magnetizing and The Simplest and most common padicle applicators are
testing short overlapping lengths of the weld. The continu- rubber bulbs or shaker bottles. For most magnetic particle
ous magnetization method is used (the magnetic field is applications, these simple devices provide the kind of
 MAGNETIC PARTICLES AND PARTICLE APPLICATION I 203

application required for accur~te ~esting .. Mechanical pow- their magnetic properties remain intact and they can still
der blowers are another apphcahon optIOn. They are de- indicate discontinuities. Beyond 370°C (700 OF), magnetic
signed to float a cloud. of pcuticles onto the. test object powders can ignite and burn.
surface and to then proVIde a gentle stream of aIr to remove Fluorescent and daylight fluorescent powders lose their
the lightly held background pmticles. For best results, visible contrast at 150 °C (300 OF) and sometimes at lower
magnetizing current should be present throughout the temperatures. This occurs because the pigments are organic
application of particles and the removal of background. compounds that decompose or lose their ability to fluoresce
at particular temperatures.
Monitoring Particle Application
Other Magnetic Particle Considerations
Regardless of the applicator that is used, the inspector
IDust carefully monitor the test object while paiticles are Dry method powders must not be used in wet method
being applied. It is critically impOliant that powder appli- applications. DIy particles are designed and manufactured
cation be done properly in order to ellSure the reliability of at densities that work well in air but cause them to settle
the test. . quickly out of liqUid suspensions. In water, typical dry
This is especially true if subsurface discontinuities are powders settle at rates around 150 mm (6 in.) per second
being sought. Their particle indications are weakly held and and accordingly cannot be kept suspended. In addition, dry
not well delineated, so that they are very susceptible to powders are very susceptible to oxidation when exposed to
damage from patticles applied later. water.
Note also that ferrous powders in general and particularly
Particle Reuse numy of their pigments are classified (!s nuisance dusts by
the Occupational Safety and Health Administration (OSHA)
It. is recommended that dry magnetic palticles be used and must be handled accordingly. .
only once. There are occasions when reuse is permitted but
inspectors should understand the effects 'of frequent reuse.
Ferrous magnetic powders are dense (specific gravity of
7.68). \\Then, ~gitated in .bulk, as in'a powder blower or a Viewing and Interpreting Dry
bulb, a lot of sheating and abrasion occurs and this wears off Particle Test Indications
some of the pigment. Each reuse, with its additional
a
rehandling, wears off more pigment. As result, color and Producing a discontinuity indication is the first stage of a
contrast continually dinlinish to the point that discontinuity magnetic particle test. Viewing the indication and interpret-
indications are not visible. ing its meaning is the second important step, and both
Reuse of magnetic powders also encourages fractionation, procedures are critically dependent on the characteristics of
the separation of different sized padicles from the compos- the magnetic particles and light intensities.
ite. Padicles sprayed over a test object are designed to have
different sizes and to fall in different trajectOlies. Generally,
Visible Particle Light Intensities
finer paiticles float fudher away from the point of introduc-
tion. Unless great care is taken to collect all the particles, Military standard MIL-STD-1949A calls for minimum
reused powders become gradually coarser and correspond- light intensities of 1,000 lux (100 footcandles) for magnetic
ingly less sensitive. padicle testing with nonfluorescent powders. Other light
level recommendations range from 800 to 2,000 Ix (80 to
Particle Storage 200 ftc).
The optimal light level is often a compromise between
The storage condition for dry method powders is critical operator fatigue and visibility. On bright or reflective sur-
t? their subsequent use. The primaly environmental con- faces, high light intensities can cause glare that interferes

I
Sl?eration is moisture. If magnetic paiticles are exposed to \vith interpretation. On darker surfaces or those covered
hl~h levels of moisture, they immediately begin to form \vith thin scale, rust or other staining, the 1,000 Ix (100 ftc)
oXides. Rusting alters the color, but the major problem is level may be barely adequate.
that the padicles adhere to each other, forming lumps or
large masses that are useless for magnetic particle testing.
Fluorescent Particle Light Intensities
Though not severe, there are also limitations on the
t~~peratures at which dry powders can be stored and used. Because of their limited industrial application, there is
\ VISIble light powders \vork on surfaces as hot as 370°C
(700 OF). Near this temperature some palticle materials
little empiIical data for light levels in dry fluorescent testing.
In addition, there is presently no standard or specification
become stid,'),. Others lose much of their color, although for dry fluorescent intensity ranges. The few dry fluorescent
 204 / MAGNETIC PARTICLE TESTING

powders that have had marginal applications were much

brighter than typical wet method powders. Their viewing Automotive and Aerospace Engineers. This specification is
conditions were correspondingly less demanding than the referenced in Magnetic Particle Inspection (MIL-STD-
conditions for wet fluorescent particle viewing (see below). 1949A) issued by the US Department of Defense. Specifi-
Daylight fluorescent dry powders also have no specifica- cation AMS 3040 places an upper limit on particle size:
tions for required light intensities. Sunlight, shaded or 98 percent must be finer than an 80 mesh in the American
direct, as well as artificial blue or white light all excite Standard Sieve Series or 180 J.Lm (0.007 in.) in diameter.
daylight fluorescent powders to a high level of brightness The specification also defines the limiting amount of loose
and visibility. Incandescent light is somewhat less effective pigment (in a test called magnetic properties) and specifies
and yellow sodium vapor light is totally ineffective. Daylight a minimum sensitivity, six indications on the tool steel ring
standard.
fluorescent powders also fluoresce brightly under ultraviolet
light but cannot be recommended for this method of Other methods have been developed for dry magnetic
viewing. They exhibit a high level of fluorescent background particle testing of pipe, including requirements for the dry
when viewed in the dark, high enough to nearly obliterate particles to be used. 1 Powder is required to have a range of
fine discontinuity indications. Testing under full visible light particle sizes: at least 75 weight percent must be finer than
hides this background and keeps contrast hig~. 120 mesh ASTM sieve size (125 J.Lm or 0.005 in.) and at least
15 weight percent must be finer than 325 mesh (44 J.Lm or
0.0017 in.).
Dry Magnetic Particle Specifications This specification also outlines a magnetic permeability
test that uses a cylindrical powder sample as the core of a
While the manufacturers of dry magnetic powders have simple transformer. For an 11 kA-m -1 (1400e) input in
their own quality control tests, additional testing is often the primary, at least 2.5 V shall show up across the second-
needed to prove conformance to user specifications. One ary, with the prescribed cirCUitry. Such tests are sensitive to
important specification is Magnetic Particle Inspection Ma- the amount of powder packed into the cylinder and effec-
terial, Dry Method (AMS 3040) issued by the Society of tively put a ceiling on the amount of magnetically inert
pigment that the powder can contain.
 MAGNETIC PARTICLES AND PARTICLE APPLICATION I 205

PART 2
fi-
WET METHOD TESTING MATERIALS

g
occurs because the methods of forming low retentive
c
y
Wet Method Particle Characteristics powders (grinding or precipitating from a highly agitated
f solution) preferentially break down long, narrow particles.
In wet method magnetic particle testing, particles over Exceptions to this are the high retentivity, high coercive
500 mesh (25 p,m or 0.001 in.) in diameter, are considered force oxides and ferrites used in magnetic tapes.
coarse. These large particles exhibit diminished sensitivity Fluorescent wet particles have a definite and measurable
to fine surface cracks and indicate no more than four size, as do particles based on finely divided metallic iron.
indications on the tool steel ring standard. These coarse Synthetic iron oxides are more difficult to measure, with
particles settle out of suspension in five to fifteen minutes diameters around 0.1 p,m. They are almost too fine to settle
and are difficult to keep in suspension. out of suspension.
Sensitive wet method patticles 'range from 5 to 15 p,m in Because of their slight residual magnetism, oxide particles,
diameter (0.0002 to 0.0006 in.) and unpigmented ferromag- collect to form loose clusters that settle out of suspension
netic oxide particles are an order of magnitude finer. much faster than individual particles. The degree of clus-
The small size and generally compact shape of wet tering depends on the intensity Qf agitation. At hig~ agita-'
method particles have a dominating effect on their behavior... ·tion rates, the dusters are small. At low agitation rates, they
Their size rel),ders permeability measurements highly inex- become larger. Clusters can often be seen on smooth shiny
act and of limited utili tv. In addition, the size influences the test surfaces after bath application. In this quiescent state,
brightness of fluoresce1;t powders made from such palticles. the clusters grow large enough to be seen with the unaided
eye, often tens of micrometers in diameter.

, Wet Particle CompOSition

Commercial wet method particles are made from finely Magnetic Properties
divided iron, black iron oxide, brown iron oxide, and Because they can be shaped into ideal toroidal samples of
experimentally from ferrites, nickel and nickel alloys. one hundred percent density, metals and ferrites are subject
Black iron oxide or magnetite (Fe304) is available as to accurate measurements of their magnetic properties.
ground ore or preferably as a fine synthetic powder. Brown Powders cannot be accurately measured because it is not
iron oxide (gamma Fe203) is chemically identical to possible to make samples with densities higher than about
nonmagnetic red iron oxide (alpha Fe203) but has the same 50 percent. As an example, pure iron has an initial perme-
ferromagnetic cubic crystalline structure as magnetite. ability of 1,000 at 8 kA-m -1 (100 Oe) field strength. Pure
Fenites are hard ceramic materials that are difficult to iron has a maximum permeability around 5,000. In labora-
make into fine powders. Some nickel alloy powders show tory tests, 10 p,m (0.0004 in.) iron powder showed a perme-
good magnetic test sensitivity, if fine enough, but are slightly ability of about 10 (5 for iron oxides) at 8 kA-m -1 (100 Oe).
denser than iron and even harder to keep in suspension. With the lower field strengths associated with small leakage
Fluorescent powders also contain fluorescent pigments as fields at discontinuities, permeabilities may well be lower
well as a binding resin to attach the fluorescent pigment to still (the data above cannot be used to predict a ferromagnetic
the ferromagnetic core. powder's behavior).
Because of their compact shape, the true material per-
meability of fine wet method particles may not be of
practical importance. However high the core material's
permeability might be, the apparent permeability of the
1 At sizes of 10 p,m (0.0004 in.) and under, practically all individual particles does not exceed a value of about 2.5.
ow retentivity ferromagnetic powders have a compact This is because of the large demagnetization factor 2 associ-
shape, with length to diameter (LID) ratios around one. This ated with an LID ratio of one.
 206 I MAGNETIC PARTICLE TESTING

1 N the particle strongly influences its fluorescent brightness as

(Eq.l)
J-L 47T the following data illustrate.
In 1 kg (2.2 Ib) of 12,5 J-Lm (120 mesh or 0.00,5 in.) iron
or in SI units: particles, the pigmented surface area is about 6 m 2 (6,5 ft 2 )
and can be made brightly fluorescent with very little
1 pigment. A finer iron powder, about 40 J.Lm (32,5 mesh or
0.0017 in.) has a surface area around 18 m 2 (190 ft 2 ). In 1 kg
J-L
(2.2 lb) of a 6 J-Lm (0.00024 in.) oxide based powder, the
Where: pigmented surface area is about 420 m 2 (4,500 ft2). Because
the inspector's eyes register the fluorescent pigment on the
J-L = the tme permeability of the parent ferromagnetic particle's surface, ultrafine particles require 30 to 60 times
material; as much pigment as the coarser particles in order to achieve
J-L' = the apparent permeability of the sample particle; the same relative brightness. Such ratios make manufactur-
N = the demagnetizing factor; ing of bright ultrafine particles impossible (the pmticle
Nd = the shape demagnetization factor. would contain virtually all pigment with a trace of
ferromagnetic core and no electromagnetic sensitivity).
An important consideration in fluorescent particle con-
At lower material permeabilities, the apparent permeabil- trast is its durability. In an agitation system where the bath
ity decreases, becoming 2.0 at a material permeability of 10. constantly passes through a centrifugal pump, the particles
The apparent permeability decreases rapidly at still lower are subjected to constant high speed impact and shearing
permeabilities. As a result, magnetic particles made from action from the pump's impeller. This slowly breaks the
very high permeability material are little more effective than particle down to fragmt:nts. In the extreme case, two kinds
those of moderate permeability values (about 20 to 100). of particles are formed: (1) nonfluorescent magnetic frag-
The most important consideration is avoiding the' use of ments which form indications that cannot be seen; and
particles with' very low permeabilities. (2) nonmagnetic fluorescent fragments which do not indi-
Tme material permeability becomes important when two cate discontinuities but which do cause background. In
or more particles touch and align in a leakage field. The LID practice, baths never reqch this stage of total deterioration,
.ratio of the jOined pamcles begins to effectively exceed a but as breakd6wn progresses, indications become dimmer
. value of one and the demagnetization factor of the string of while background fluorescence increases (indication-to-
particles shrinks. This allows the string to become more background'contrast diminishes).
effectively magnetized and more firmly attached to the sites The onset of particle breakdown can be detected in an
of discontinuity leakage fields. already operating bath by an extension of the settling test.
Fluorescent fragments are both less dense and finer than
the intact particles and settle out of suspension much more
Visibility and Contrast
slowly, requiring 10 to 1,5 hours. After a settling test is
Wet method particles are commercially available as fluo- performed, allow the settling tube and sample to sit
rescent and nonfluorescent materials. Most nonfluorescent unagitated overnight. Loose fluorescent pigment will pro-
particles are simply ferromagnetic iron oxides, either black duce a thin, brightly fluorescent layer on top of the
or brown. These are used in their natural color with no sediment (see Settling Test below).
added pigment. As a result, these pmticles are very slightly No finn correlation has becn made between the extent of
more sensitive than the pigmented fluorescent particles. pmticle breakdown, the relative amount of free pigment and
For this reason, the unpigmented iron oxides are preferred the reliability of the bath. This settling test simply detects
for some applications where sensitivity is more important the occurrence of breakdown. The individual user can best
than easy visibility. Bearing testing is an impOltant applica- relate the evident breakdmvn to the quality of the fluores-
tion of these oxides - sensitivity is critical and the smooth cent magnetic padicle bath.
reflective surface of the test objects give the best possible
contrast to the dark oxide pmticles.
On darker surfaces, indications from brown and black
particles are very difficult to see and locate, though a thin
Oil Vehicles for Wet
white lacquer painted on the test surface can improve Method Particles
contrast, much as it does with dly powders.
Fluorescent magnetic pmticles are composites, contain- There are two kinds of vehicles used for wet method
ing a ferromagnetic core, a fluorescent pigment and pref- testing: water and oil. Oil vehicles are preferred in certai~
erably a binder to hold the composite together. The size of applications: (1) \vhere lack of corrosivity to ferrous alloys IS
 MAGNETIC PARTICLES AND PARTICLE APPLICATION I 207

'tal (finished bearings and bearing races, for instance); having a flash point of 93°C (200 OF) or more, and burdens
~) where water could pose an electrical hazard; (3) on some the inspector's breathing air with many times more solvent
high strength ;.llloys, where exposure to w<~ter may cause vapor.
hydrogen eml)littlement (hydrogen atoms from water can The MIL-STD-1949A standard no longer provides re-
diffuse into the crystal stmcture of celtain alloys thereby quirements for magnetic particle oil vehicles, but instead
causing embrittlement). '. references DOD-F-87935 and AMS 3126.
Water vehicles are preferred for the followmg reasons: The requirements of DOD-F-87935 are listed below.
(1) lower cost, (2) little fire hazard, (3) no petrochemical
fumes, (4) quicker indication formation and (5) little clean 1. Viscosity: maximum of 5.0 mmz·s - 1 (5.0 cs) at bath
up required on site. temperature (ASTM D445).
2. Flash point: 93°C (200 OF) minimum (ASTM D93).
3. Fluorescence of vehicle: no more than reference stan-
Specifications for Oil Vehicles
dm-d.
In the past, a vClJiety of petroleum solvents and oils were 4. Odor: none.
wIitten into magnetic particle testing specifications but 5. Particulate matter: 0.5 mg-L-l' maximum (ASTM
most of these early oil vehicles were designed for other D2276).
purposes. For example, federal specification P-D-680 (dry 6. Total acid number: 0.0015 mg KOHegm- 1 maximum
cleaning solvent) and federal specification W-K-220 (kero- (ASTM D3242).
sene, deoclolized) were referenced in MIL-I-6868E (1976), 7. Color: 1.0 maximum (ASTM DlOOO).
Magnetic Pmticle Inspection Process. .
The P-D-680 document was originally applied to nonde- The requirements of AMS 2641 (January 1988) are as
structive testing before high flash pOint fluids were re- follows.
quired. This specification called for: (1) a min~mum flash
point of 59°C (138 OF); (2) a low distillation range (no 1. Flash point: 93°C (200 OF) mInImUm for Type I
contempormy oil vehicle would meet this specification); and vehicles; 60 to 93 °C (140 to 200 OF) for Type II
(3) two inappropriate purely chemical tests (th~ doctor test vehicle (ASTM D-93). . . . . .
and sulfuric acid absorption). . 2. Viscosity: no more than 3.0 mmz·s - 1 (3.0 cs) at 38°C
The VV-K-220 document allowed a Similarly low flash (100 OF); no more than 5.0 mm 2·s- 1 (5.0 CS) at bath
point and low distillation range and required a different temperature (ASTM D-445).
sulfuric acid reactivity test. 3. Fll~orescence: same as specified in DOD-F-87935.
Militaty specification MIL-STD-1949 (August 1985)
called for magnetic patticle oil veh~cles that met the re- The AMS 2641 specification prescribes the same limits as
quirements of AMS 3161 and DOD-F-87935. DOD-F~87935 on the amount of particulate matter, acidity,
The American National Standard Institute's AMS 3161 odor (inoffensive) and visible color.
(1972), Odorless Heavy Solvent Inspection Oil, addresses
one of the specific needs of magnetic particle testing, a
viscosity limit of 32 to 34 Savbolt Universal Seconds.
Viscosity is measured in the SI system as square meters per Wate'r Vehicles for Wet
second (m 2es - 1). It may also be measured in mm 2.s - 1 or Method Particles
its eqUivalent in the cgs system, centistokes (cs). The 34
SUS level is a markedly low viscosity range equivalent to
~.O mmzes -1 (2.0 cs) to 2.3 mm 2es -1 (2.3 cs). Such viscosity Water Conditioning
IS typical of light, volatile, low flash point petroleum sol- \Vater cannot by itself be used as a magnetic particle
vents. The flash pOint minimum given in AMS 3161 is only testing vehicle. It rusts ferrous alloys (including the testing
65°C (150 OF). <
equipment), it wets and covers test surfaces poorly and does
!lash pOint is impOltant for two reasons. Fire safety is the not reliably disperse fluorescent magnetic particles. Water
pnm~~ consideration: the Occupational Safety and Health conditioners or wetting agents must be added to remedy
AdmIl1lstration has placed costly restlictions on the use of these shortcomings.
lower flash point solvents in open tanks. 3 These restrictions \Vater conditioners are not well covered in existing
may ~e avoided if the vehicle's flash point is over 93°C specifications. MIL-STD-1949A requires only that a condi-
(200 F). Health considerations are less obvious but also tioned water vehicle (1) wets a test surface without the film
h~portant. Low flash point vehicles are more volatile than of water breaking; and (2) has an alkalinity not exceeding
fllgh flash solvents. A typical petroleum solvent having a pH 10.0. Common sense indicates that a conditioned bath
ash point of 65°C (150 OF) is much more volatile than one should not rust test objects and should not foam so as to
208 / MAGNETIC PARTICLE TESTING

interfere with the formation of indications. In magnetic agent concentrations are critical: if too much is present, it
particle tests, a water conditioner needs to perform the four acts like an oily contaminant, coagulating the magnetic
follOwing functions. paliicles and destroying the bath's wetting ability.

1. Reliably wet and cover all test surfaces.

2. Encourage wetting and dispersion of fluorescent par- Corrosion Inhibition
ticles, with their water repellent organic pigments and Corrosion was at one time controlled by including small
binders. amounts of sodium nitrite or traces of sodium chromate
3. Minimize foaming caused by the necessary presence in the magnetic particle bath. These chemicals have
of wetting agents in the bath. high levels of toxicity and, when present, must be listed
4. Retard rusting of test object surfaces. by suppliers on Materials Safety Data Sheets as ordered by
the Occupational Safety and Health Administration
Where the bath does not at first cover the test object (29 CFR 1910.1200). Sodium nitrite and sodium chromate
surface, there can be no particles and no indication forma- are also among the waste water contaminants regulated by
tion. Beyond this, the bath film must cover the test object the Environmental Protection Agency. These safety restric-
surface (without breaking) throughout the magnetic particle tions limit the use of the chemicals and alternative corrosion
test procedure. If the water film does break or peel from the inhibitors are being studied.
surface to form separate drops, it also peels off most of the \Vater baths have never been expected to provide long
particles in affected indications. The result is poor and term corrosion protection for the test object after testing is
unreliable inspection. complete. Accomplishing this protection requires separate
post testing treatment.
Wetting Abilities
Different surfaces require different degrees of wetting.
Steel billets, with porous, oxidized surfaces are eaSily wetted
Bath Contamination
by untreated water. At the other extreme, very smooth
'. J30th water and Qil vehicles can be contaminated by s?lid
surfaces covered with a trace of oil reqtoJ.ire very strong
materials (see Settling Test below). Introduced oil may also
wetting ability. In such' cases, surface tensions as low as be a contaminant of water and oil baths. In addition; oil
0.025 N-m -1 (25 dynes-cm - 1) may be required. ..
vehicles can be contaminated by introduced water. .
Magnetic iron oxides used in nonfluorescent wet method
The most common form of contamination is oil in a water
baths are easily wetted by untreated water. However,
bath. This occurs in recirculating systems where bath runofI
fluorescent particles typically contain organic pigments and returns to the bath reservoir and oil on the test surface is
binding resins that tend to be water repellent. Fluorescent
washed into the water. When enough oil is present, it caus~s
baths must therefore include treated water to achieve the magnetic paliicles to congeal and also destroys the bath S
wetting ability needed to cover common oily test surfaces wetting ability. Oil contamination can be avoided by effec-
and to adequately wet and disperse the fluorescent particles. tively precleaning test objects. If this is not possible,
When fine fluorescent magnetic particles are not wetted by
stringent control of the bath is required and should inclt~de
the vehicle, they float on the bath surface like dust and no regular visual inspections and monitoIing of productIOn
amount of agitation will disperse them. levels. Contaminated bath is discarded and replaced.
Highly fluorescent oil or grease on test object surfaces
can eaSily dissolve into oil-based magnetic pmticle baths.
Foaming Solutions The accumulation of introduced oil produces a blue fluo-
Because they contain powerful wetting agents, magnetic rescent background that can hinder discontinuity ind~cat~on
particle water baths easily generate stable foams when detection. Frequent inspection of the bath and mOl1Jt?nn.g
agitated at the surface. Masses of foam that reach the test of production levels are again the solutions, if precleanmg IS
surface during bath application slide to the lowest edge of not possible. .
the surface, erasing indications in their path. It is therefore Particle coagulation also occurs when water contammates
important that water baths (1) do ?ot foam exc~ssively a?d an oil bath. The condition is accompanied by a build-up of
(2) generate unstable foam that disa~~ears qUIckly.. ~Vh.Ile a sticky mass on the bath tank walls. Adding small ~~lOl1n~S
careful formulation of a water condItioner can mmimize (0.1 percent or less) of a suitable oil-soluble emulSIfier WII!
foaming, full foam control occaSionally requires the use. of restore the bath. The addition of emulsifiers must be don,e
antifoaming agents. Such agents cover the bath surface WIth carefully - too much will markedly increase the bath s
a microscopically thin layer of an oily substance. Antifoaming viscosity and fluorescence.
 MAGNETIC PARTICLES AND PARTICLE APPLICATION I 209

concentrations. Most industry standards specify only the

Bath Preparation acceptable settling volume.

Wet magnetic particle baths may be mixed by the supplier Visible Particle Concentrations
r may b~ sold chy for mixing by the user. When the The Society of Automotive Engineers specifications
~agn;tic particle system contains a recirculating bath, AMS 3042 and AMS 3043 for visible wet method particles
mixing is relatively easy. call for settling volumes from 10 to 24 mUL or (using a
100 mL centrifuge tube) 1 to 2.4 mL (1 to 2.4 cm 3 ). The
Recirculating System Bath Preparation MIL-STD-1949A specification calls for a range of 1.2 to
2.4 mL after thirty minutes settling in a vibration free
After the liquid vehicle is added to the bath tank, the location.
manufacturer's required amount of powder is added dire~tly All three standards specify the use of the particular
above the sump. The hath is allowed to recirculate for five pear-shaped tube illustrated in ASTM D-96, 'Vater and
minutes \vhile the pump disperses the powder. . Sedi1nents in Crude Oils (Fig. 1 shows a similar settling
When a conditioning agent is needed for a water vehicle, tube). The D-96 tube has a 1.5 mL (1.5 cm3 ) stem
the agent is added to the bath and dissolv~~ b~fore the graduated in 0.1 mL (0.1 cm3 ) increments. The next
magnetic powder is added .. Because the C?ndI~lO.nmg agent graduation is at 2.0 mL (2.0 cm3 ) in the conical portion of
must be dissolved before It becomes actIve, It IS faster to the tube. If the expected settling volume is in an inade-
begin the process with warm water. quately graduated portion of the settling tube, a permanent
mark should be made at the correct location so that a
Nonrecirculating System Bath Preparation constant bath strength can be accurately maintained without
estimation.
Ifthe magnetic particle system has an air agitated bath, or
if the bath is applied from spray guns, then bath preparation Fluorescent Particle Concentrations
is more difficult. Magnetic particles often stick together
during manufacture, storage or shipment ,and simple stirriQg' The AMS 3044, AMS 3045°and' MIL-STD-1949A stan-
d6es riot separate them - manually mixed baths are' dards also specify concentrations for fluorescent particle
therefore susceptible to uneven distribution of particles and
are unreliable for testing.
The best way to manually prepare a finely dispersed bath FiGURE 10 Diagram of a typical centrifuge tube
is to add pre measured powder to a small food blender along used for magnetic particle settling tests
with enough oil or conditioned water to nearly fill the
container. At low speed this takes about thirty seconds.
An altemative method is to add pre measured powder to a
small container along with enough liqUid to form an easily
worked paste. Mixing by hand precedes adding the paste to
the bath. .

Bath Maintenance
The effectiveness and reproducibility of a magnetic par-
ticle bath depend on its concentration. If the concentration
is too low, indications will be weak and difficult to locate. If
the concentration is too high, the background will be intense
enough to camouflage indications. Correct particle concen-
trations fall between these extremes.
Keeping the concentration at a constant level eliminates
one variable in the test: the indication-to-background con-
t~ast. It is important to regularly monitor bath concentra-
tIons throughout the testing cycle, not only after bath
preparation. Suppliers of magnetic powders specify bath
Concentration both in weight of powder per volume of bath
and also in settling volume ranges that result from these
210 / MAGNETIC PARTICLE TESTING

baths but again the requirements differ slightly. The AMS

specifications call for a range of 0.2 to 0.5 mL (0.2 to The Settli ng Test
0.5 cm 3 ), while MIL-STD-1949A calls for 0.1 to 0.4 mL
(0.1 to 0.4 cm3 ) settling volumes. The ASTM D-96 tube is Since the 1940s, a settling test has been used to measure
required for the test. magnetic paIiicle bath concentrations. 5 It is a convenient
Fluorescent particle baths are much more dilute than method that requires little equipment, a simple procedure
visible baths, so that all likely settling volumes fall within the and only thirty to sixty minutes to perform. Its accuracy is
graduated stem of the tube. More reproducibility for small sometimes less than 80 percent but the levels of precision
settling volumes can be achieved if the tube has a 1 mL stem are appropIiate for most applications.
graduated in 50 ,uL intervals. The military standard MIL-STD-1949A specifies that the
filled settling tube be demagnetized before the settling test
begins. Because the magnetic condition of a bath can affect
the speed of settling and the final settling volume, the
Use of Settling Ranges demagnetization procedure is an effort to standardize the
It is not productive to assume that all concentrations magnetic level of the bath regardless of its use.
'within broad specified ranges are equally effective or desir-
Settling Parameters
able. For example, it has been sho\\11 that fluorescent
particle concentrations around 2 mUL give the best results It is essential that the settling test take place in a location
on a tool steel ring standard. 4 Higher concentrations pro- that is free from vibration. The settling tube must be
duce backgrounds that obscure faint discontinuity indica- positioned in an area that is proven to be free of strong
tions. Lower concentrations produce indications too faint to magnetic fields. .
be easily detected. Freshly magnetized bath settles very rapidly, often in
In another special case, it was found that increaSing fifteen minutes or less. Magnetization causes the particles to
particle concentration from 1 to 4 mUL gives increaSingly clump together quickly and form large, fast-Sinking clusters
brighter indications with velY little increase in background. (agglomerated settling). However, these clusters form a
In this application, clusters of velY fine cracks were being mud1 larger settling volume than if the individual pmiicles
located under a chromium plate. The cracks were very close were unmagnetized (the stmcture of the clusters cannot be
tog~ther (less than 0.5 mm apart) so thqt expanding indica- compacted by gra~ty). .
tions totally depleted the narrow strips of background Speed of settling and settling volume depend on the
between the cracks, producing dense bright indications. particle' $. magnetization level and this is the basis for
Some users prefer a substantial fluorescent background requiIing demagnetization of the settling tube sample.
during testing because in their applications this lessens the Vibration during settling does not affect the speed of
need for frequent bath concentration measurements. Such settling but it can compact the sediment to give falsely low
modifications of recommended procedure are allowed only settling volumes.
after firmly establishing and verifYing the results of such
The Settling Tube
tests for the individual application.
The broad concentration ranges outlined in most speci- Settling test equipment is Simple: (1) a 100 mL pear-
flcations cover the limits that are allowed and that may be shaped graduated glass centrifuge tube (see Fig. 1); (2) a
required. Testing beyond either specified extreme may not stand for supporting the tube vertically; and (3) a timer to
produce the best results and higher or lower concentrations Signal the end of the specified settling peIiod.
should be considered. The tube referenced in most specifications has a 1.5 mL
There is a strong tendency in some industries to test at (1.5 cm3 ) stem graduated in 0.1 mL intervals. According to
the lowest possible concentration. This is valid for economic ASTM D-96, the maximum reading error of this tube is
reasons and it additionally ensures that excessive back- 30 ,uL. Another common tube has a 1.0 mL stem graduated
ground does not become a problem. There is one precau- in 0.05 mL intervals. This configuration is easier to read for
tion necessary for this approach: if particles are applied baths with small settling volumes, including most fluOloes-
from a recirculating system where the excess bath returns to cent particle baths.
the reservoir, particle depletion is a likely result. Most For vel)' dilute baths, even 0.05 mL intervals are too large
magnetic particles adhere to the test object while most of and a tube with a 0.2 mL stem in 0.01 mL graduations gives
the vehicle returns to the reservoir. The particle concentra- the most reprodUCible results. This stem is VCIY nearly the
tion decreases steadily vvith usage over time and may not be size of a capillmy tube and is extremely difficult to clean
noticed by inspectors. For reliable, reproducible magnetic after a settling test.
pcuiicle tests 'with low bath concentrations, the frequency of The typically specified tube is probably a compromise. It
concentration tests must be increased. can be used to measure dilute fluorescent particle baths
 MAGNETIC PARTICLES AND PARTICLE APPLICATION I 211

(under 0.2 mL vvith diminished accuracy) as well as more In addition, the upward How of the displaced liquid now
. concentrated visible pmticle baths (also with diminished becomes rapid enough to fmther retard settling. The final
accuracy if the settling volume is over 1.5 mL). The three stage is compact settling, where all the particles are in
ure tubes do not show the same settling volumes for the same contact and apparently settled but the sediment slowly
ent bath. Tubes with narrower stems show higher settling shrinks in volume as more liquid is displaced.
ure volumes. Hindered settling behavior mentioned below acts
1 is to retard settling in the more constricted stems. Condition of the Vehicle after Settling
on In its early versions, MIL-STD-1949 required that the
Effect of Contaminants supernatant liquid (the vehicle after settling) be essentially
he
Recirculating magnetic particle baths can pick up solid nonfluorescent. Naturally, fluorescence in the liquid vehicle
~st
contaminants from three main sources. The first source can detract from the contrast of fluorescent indications. It
ct
includes solids washed into the bath from test object can also signal the breakdown of Huorescent magnetic
Ie
surfaces. Vapor phase degreasing, a common procedure for pmtides into their components. A Huorescent magnetic
e
cleaning test ohjects before magnetic particle testing, does particle bath composed of unattached fluorescent pigment
. not· remove non-oily contaminants such as sand, dust, lint along with nonHuorescent particles, whose indications can-
or grit. not be seen under ultraviolet light, is not a usable bath and
A second source is palticulate matter in the testing should be discarded.
atmosphere that settles into the bath. The quantity of these Three different conditions can lead to fluorescence in the
solids is determined by the geographic location and the supernatant liquid after settling is complete but not all of
,nature of the manufacturing facility, but some airborne them indicate a substandard bath. Fluorescent oils or grease
particulates can be in quantities sufficient for increasing the can be swept into the bath from test surfaces, making the
settling test volume. whole liquid fluorescent. :particles can break apart, leaving
A third source, unique to water baths, is a highly dilute tiny slow settling fragments of fluorescent but nonmagnetic
but bulky· and gelatinous precipitate. The. contaminant is pigment behind. Finally; sQme very small but complete
caused by increasing water hardness in th~ bath as water is particles can escape agglomeration and, settling at individ-
added to fep,lace evaporation loss. Water hardness is deter- ually slow rates, may require another hour or more to finally
'mined by the concentration of certain salts, including settle out.
calcium and m.agnesium. The resulting precipitate does not A fluorescent supernatant liquid is a warning to further
affect settling tests of baths with high particle concentra- monitor the condition of the bath. Is the blue fluorescence
tions because too little of it is present and the weight of the due to the presence of oils bright enough to interfere with
settled ferrous particles compacts it. In dilute baths (settling indication contrast? If not, then it is not excessive. Does the
volumes of 0.1 mL or less in the 100 mL centrifuge tube), sediment after overnight settling show a bright fluorescent
the precipitate adds bulk to the sediment and can give layer on top? If not, the particles have not been broken
falsely high settling volume readings. apart. Does the green fluorescent supernatant liquid give
Contaminating patiicles do not necessarily settle out first feeble indications because of the extremely low concentra-
or last to form obvious layers in the sediment. The effects of tion of particles? If so, then the bath contains fine, slow
contamination contribute strongly to the low absolute accu- settling but useful particles. The converse to any of the
racy of the settling test. Contamination of oil baths by water above answers indicates that the bath should be replaced.
and contamination of water baths by oil tend to produce Various instruments are available for automatically mon-
bul~ sediments and substantial amounts of the magnetic itoring bath concentrations. Low consumer acceptance of
partIcles adhering to the sloping walls of the settling tube. these instruments is based on two disadvantages: (1) the
much lower cost of simple settling test equipment; and
Stages of Settling (2) the accuracy of continuous monitoring equipment can
. The sedimentation of magnetic particles during the set- be compromised by certain kinds of bath contaminants .
thng test consists of four separate but overlapping stages. 6
. ~~e first stage is simple unhindered settling of the
mdIvidual particles. This changes to agglomerated settling Applications of Wet Magnetic
where their slight residual magnetism causes the pmticles to Particles
collect into larger and faster settling clumps. When most of
the p~lticles have settled into the narrowing portion of the
centnfuge tube, hindered settling begins. In this stage, the
Magnetization Methods
loc~ palticle concentration is high enough that the falling Wet method materials can be used in practically any kind
partICles get in each others way and restrict fmther settling. of magnetic particle testing, although they are most useful
212 / MAGNETIC PARTICLE TESTING

for indicating the fine surface cracks that dry powders present on the test object. Often, the continuous magneti-
cannot reliably locate. zation method is employed in this test, with the magnetizing
Wet particles can be used with the residual and the current deliberately kept on throughout bath application.
continuous magnetization methods. In the residual method, Several means of spraying are available: aerosol cans,
the test object must have high magnetic retentivity. It is first prepressurized spray guns and spray guns supplied from
magnetized, then the magnetic field is removed and the a separate container (a dmm of bath or a pressurized
object is dipped into an agitated magnetic particle bath. spray pot).
Agitation must be gentle or sometimes stopped during Each of these devices has its own advantages and disad-
immersion to avoid dislodging indications. Longer immer- vantages. Aerosol cans free the testing process from the
sion or exposure times yield larger indications, an advantage need for pressurized air, but the cans must be frequently
of the residual method. shaken to keep the bath suspended. Prepressurized spray
The continuous magnetization method is more widely guns cost less than bath in aerosol cans, but they require
used, with slight variations in procedure that depend on the pressurized air or a supply of carbon dioxide cartridges and
specific application. In the typical wet horizontal testing must still be shaken and constantly agitated. Prepressurized
unit, the inspector flows agitated bath onto the test object guns are also much heavier than aerosol cans and handling
from a hose. After the hose is shut off, the magnetic field is can be strenuous. Spray guns supplied from separate con-
activated. At this point, the bath is draining off test surfaces tainers often have recirculating systems immersed in the
slowly enough to avoid dislodging indications. If these bath. These automatically keep the bath agitated and are
actions are not properly sequenced and bath is still being quite effective where electrical power is available. Pressur-
applied after the magnetizing current ceases, the continuous ized spray pots can be equipped to supply constant agitation
method is no longer . being used but rather the residual and are handy to use where pressurized air is available.
method. This degrades the inspection and can actual~y
nullify it if the test objects are of low magnetic retentivity.
Sample test objects ·are available to teach, as well as test,
application technique. They feature a very low retentivity
Wet Magnetic Particles for Special Applications
iron test object containing a discontinuity. If the sequencing
of bath application and magnetization is not correct, the vVet method fluorescent par.ticl~s can be used 'for nonde-
discontinuity is not indicated; .. stmctive testing of undelWater stmctures such as drilling or
Multidirectional magnetization allows the use of automat- production platforms. Test procedures parallel those used
ic bath application. Precise timing and sequenCing are also on dry land with a number of added complications.
important in this application, where the test object is First there are the personnel hazards: (1) exposing in-
exposed to various magnetic orientations in order to locate spectors to deep undelwater conditions for long periods;
discontinuities positioned in various ways throughout the and (2) the electrically conductive environment demands
object. perfect insulation to avoid the danger of electrocution.
Because an object can retain only one magnetic field at a Beyond the safety aspects, the magnetic pmticle test proce-
time, the fields are rapidly switched. The discontinuities dure is itself complicated. Often the tests are performed in
indicated by the final field orientation are effectively located murky surroundings with poor visibility where currents can
by a short lived continuous method. Discontinuities magne- cany particles away from the test surface. Before testing can
tized by earlier fields in the rapid sequence are subject to begin, the structure must be cleared of all sediment and
erasure if the bath flow is too rapid. This also e'xplains \vhy other marine fouling.
test objects with some surface roughness (13 RMS or The magnetic pmticle materials are usually special fluo-
rougher) are better candidates for multidirectional testing rescent powders that mix equally wen with fresh water or
than those \vith velY smooth surfaces. Not only do rougher sea water. The particle suspension is taken to the test site in
surfaces retard the rate of bath flow and drainage, but they a plastic bottle where it is applied near the magnetizing yoke
also retain indications by increased friction. commonly used for underwater tests. The particles are in a
concentrated aqueous sluny that is immediately diluted to
normal particle concentrations by the surrounding \\'ater.
In another special application, thick, concentrated slur-
ries of paramagnetic flakes in a viscous liquid are occasion-
Means of Applying Wet Magnetic Particles ally used to locate sU&lce cracks. These reflective paIticles
In on-site testing using portable or movable power sup- do not migrate to leakage fields but rotate in place to line up
plies, wet magnetic particles are typically applied in a spray, with the fields.
with no provision for collecting or recoveIing excess bath. The inspector brushes the sluny onto the surface, pro-
This method avoids bath contamination from particulates ducing a shiny film that stays in place and does not nm or
 MAGNETIC PARTICLES AND PARTICLE APPLICATION / 213

drip off. vVhere a leakage field exits, the particles present the 1 percent level is 100 times more than OSHA's permis-
their dull edges to the surface, delineating cracks with dark sible exposure limit (PEL) of 100 ppm. This level can be
lines OIl the otherwise blight surface. Post testing clean-up considered hazardous. For health reasons, an oil bath
is tedious fix this technique. should be used at maximum temperatures far below the
flash point. Excellent ventilation is required for any use of
an oil bath above 43°C (110 OF).

Temperature limits for Water Vehicles

Wet method testing is susceptible to temperature limits
that do not affect dry particle testing. Water baths change
little in viscosity between the freezing and bOiling points of Viewing Wet Method Indications
water but they do freeze on test surfaces colder than 0 °C
(32 OF). At near subfreezing temperatures, water does not Visible lighting Specifications
freeze instantly and magnetic particle indications have time
Visible wet method magnetic particle testing is subject to
to form before the water bath solidifies.
the same lighting requirements as dry particle testing in
At lower temperatures, antifreeze must be added to keep
MIL-STD-1949A. This is a minimum of 1,000 Ix (loa ftc) at
the bath liquid. Antifreeze is useful only in those applica-
the testing surface.
tions where high sensitivity is not desired, such as the
Because the nonfluorescent particles are often used on
testing of steel billets. Solutions of water and either ethylene
fairly bright, reflective surfaces (contrast on dark surfaces
glycol or methyl alcohol rapidly become far too viscous
can be poor), glare can easily become a pro~lem. Lowering
below freezing and drastically retard the formation of
discontinuity indications. the light intensity, when not prohibited by specifications,
can decrease glare but this should not be done without
The upper temperature limit of the test 9bject has not
been established for water baths . .At 100°C- (212 OF), water careful evaluation. of the test result,s . Where sp~cular reflec-
tion of the light source is not a problem to inspectors, the
evaporates too quickly for indications to form and this can
happen at somewhat lower temperatures as well. specification requirements should be followed ill all cases.

Fluorescent lighting Specifications

Temperature limits for Oil Vehicles
For' fluorescent wet method testing, MIL-STD-1949A
Oil baths, though still liquid below -18°C (0 OF), in- requires a minimum ultraviolet light intensity of 1,000
crease noticeably in viscosity as the test piece or bath JLWemm -2 at the test surface. Further, the maximum
temperature decreases. Type I baths (defined in AMS 2041) allowable visible light intensity at the surface is 20 Ix (2 ftc).
have a 'viscosity around 2.4 mm 2es -1 (2.4 cs) at 38°C This is because the presence of visible light reflected from
(100 OF) and reach the 5 mm 2es -1 (5 cs) limit at about the test surface lowers the contrast of the fluorescent
10 0c. (50 OF). Type II baths (AMS 2641) have a viscosity indications. Visible light intensity must be measured with
around 2.7 mm 2es -1 (2.7 cs) at 38°C (100 OF) and reach the the ultraviolet light source on since they themselves emit
5 mm 2es- 1 (5 cs) limit at approximately 13°C (55 OF). small amounts of visible light, primarily in the violet
These figures may vary with the source or manufacturer of portions of the visible light spectrum. This small amount of
the oil vehicle.
visible light, added to the ambient visible light already in the
The upper limit of practical temperature for oil baths is testing booth, can help lower a fluorescent indication's
influenced more by health considerations than by fire contrast.
hazards. \Vhen an oil bath is heated to its flash point (either The reliability of testing increases dramatically as the
in bulk or by contacting a test surface at this temperature), ultraviolet light intensity increases. 4 The probability of
air in the immediate vicinity contains nearly 1 percent oil detection of 97 discontinuities on 37 test pieces has been
vapor (10,000 ppm). The vapor condenses to a fine oily mist shown to rise from 34 percent at 1,000 JLW-mm -2 to 100
as the vapor cloud cools. A vapor or mist concentration at percent at 4,000 JLWemm - 2.
2141 MAGNETIC PARTICLE TESTING

PART 3
STUDIES OF MAGNETIC PARTICLE
SENSITIVITY
The data below were compiled from application studies of direct current, a central conductor and a head shot, \vith
magnetic palucle process sensitivities. Discontinuity sizes current passing lengthWise through the projectiles. 1l
were usually not known by the investigators. In five instanc- Though not a common application, the two test reports
es, the test objects were closely examined, sometimes with present documented correlations of discontinuity size, loca-
destmctive techniques, to determine exactly what the mag- tion, magnetizing current and indication quality. The re-
netic particle method was detecting. ports' original drawings are reproduced in Fig. 2 through
Fig.H.
Studies to Determine Wet Method
FIGURE 2. Tracings of photomicrographs showing
Particle Sensitivity fluorescent magnetic particle indications on shell
Discontinuity Size limits for Carburized Steel casings

Case hardened (carburized) steel pins of unspecified size la)

were tested with wet method magnetic particles, both
visible and fluorescent. 7 A direct current coil was used and
crack depths .were not measured. Some indications of fine
transverse cracks were found.
Micros~opic examination of the smooth test object sur- -----'----..--OUTSIDE DIAMETER
,face showed Grack widths ranging from·0.25 to 2.3 p,m (10 0.84 mm (O.0334 in.)

to 90 p,in.). Such crack widths are typical of a surface under

. high compressive stress.
~m 100039 in.)

Ol.I4mm (0.045 in.)

This range 'of widths could be considered the narrowest - - - - - - - ' , ' - - - INSIDE DIAMETER
detectable discontinuity for a direct current test, which HEAD CENTRAL CONDUCTOR
produces a magnetic field largely contained within the IT = 1.300 A 'T ~ 1,000 A
IG = 1.700 A IG = 1,500 A
higher permeability core of this test object. An alternating I" I 000 = NONE 'R 1.000 = NONE
current technique concentrates the magnetic field at the
surface, possibly altering the detectable discontinuity size
limit. ,bJ
Discontinuity Size Limits for Nitrided Steel
NitIided steel bushings of unspecified sizc were mag-
netic particle tested at 1 kA. for grinding cracks. 7 The
bushings had been nitriclecl to an average depth of 0.15 mm
- - - - - - - - I - O U T S I D E DIAMETER
(0.006 in.).
The magnetic particle discontinuity sites were metallur-
gically examined and showed grinding cracks from 13 to
'" ~:~'[~OO:'l9;:"[[
65 p,m (0.5 to 2.5 milli-in.) in depth. Some quench cracks I -J
0.42 mm (0.0166 In.) ___
0.67 mm (0.0266In.j
-------...Jt~INSIDE DIAMETER
with depths from 0.05 to 0.13 mm (0.002 to 0.005 in.) were
HEAD CENTRAL CONDUCTOR
also detected with magnetic particle techniques. All crack- IT = 500 A 'T = 200 A
ing was normal to the surface. IG = 800 A I" ~, 400 A
'R 1.000 = NONE 'R 1.000 = NONE

Correlation of Discontinuity Size with LEGEND

Other Parameters 'T = CURRENT AT WHICH INDICATIONS ARE FIRST DETECTED
IG = CURRENT AT WHICH INDICATIONS ARE EASilY SEEN
Several 20 mm (0.8 in.) high explosive projectiles were ' R.x = RESIDUAL CURr<ENT

tested with fluorescent magnetic particles. The tests used

 MAGNETIC PARTICLES AND PARTICLE APPLICATION I 215

FIGURE 3. Tracings of photomicrographs showing FIGURE 4. Tracings of photomicrographs showing

fluorescent magnetic particle indications on shell fluorescent magnetic particle indications on shell
casings casings

faJ

[ill-3 OUTSIDE DIAMETER

fa)

[ill
----'4;""·
J'
INDICATIONS RATED HEf<E
----.-t--
,:a OUTSIDE DIAMETER

t
0.79 mrn 10.0314 Inl
t
I mm 1004 in.)
075 mm 100296 in)
-L 0.95 mm (0.0377 In)

+- J ""--L
...i,/

1-1
0.27mm (O.Ollln)
0.02 mm (0.00 rfn)./ 042mm 10.01671nl
! t INSIDE DIAMETER
t f---j
0.38 mm (0015 inl
INSIDE DIAMETER

028mm (00111 in)

HEAD CENTRAL CONDUCTOR

~0~.1-:'8-m-m- 10-: 0-.!-;7~I~in-.:-)---+r---

IT = 300 A IT = 300 A
I" = 500 A IG = 500 A OUTSIDE DIAMETER
IR 1.000 = NONE IR '.000 = NONE
t -::;.- ~ 100154 '"I
INSIDE DIAMETER
fbJ I~

[ill-----'"J-'\ \
3
INDICATIONS RATED HERE

- - - . - - - - - . . . - - - OUTSIDE DIAMETER
HEAD
IT = 400 A
IG = 600 A
0.45 mm 10.018In)

IR., 000 = NONE

CENTRAL CONDUCTOR
IT = 400 A
IG = 600 A
IR '000 = NONE

--+1-'"<..
0.71 mm 100283 in 101.1 mm

7
--
10.0434in) fbI SURFACE ON OUTSIDE
I
/
005 mm (0002 In) 142 mm (0056 inl DIAMETER

t " t ~~ -
t H ~005
INSIDE DIAMETER

~
mm 10002 in)
____ RATED SU8SURFACE
0.32 mm 1001261n) :J INDICATIONS
ON INSIDE
~ INDICATIONS RATED HERE
DIAMETER HERE
0.12 mn !O005 I in J. OUTSIDE DIAM ETER

208mm
10.082111.)
t -' -f- ~mm
",
L-
0.39 mm 10.0157 In )

1001131n)
t
0154 mm (0061 in)
OUTSIDE DIAMETER

--L
t ~ 0.05 mm (0002 WI.)
INSIDE DIAMETER
~m10032inl
J
HEAD CENTRAL CONDUCTOR INSIDE DIAMETER

I, = 400 A IT = 400/\ HE/\D CENTRAL CONDUCTOR

Ie, = 600 A Ie, = 600 A
NO RATING IT = 400 A
IR ,oon = \A/EAK IR '. 00(' = WEAK
Ie, = 600 A
IR '000 = NONE

LEGEND
LEGEND
IT = CURRENT AT WHICH INDICATIONS ARE FIRST DETECTED
IT = CURRENT AT WHICH INDICATIONS ARE FIRST DETECTED
1(. = CURRENT AT WHICH INDICATIONS ARE EASILY SEEN
IG = CURRENT AT WHICH INDICATIONS ARE EASILY SEEN
I", = RESIDUAL CURRENT
IR> = RESIDUAL CURRENT
 216 / MAGNETIC PARTICLE TESTING

FIGURE 5. Tracings of photomicrographs showing FIGURE 6. Tracings of photomicrographs showing

fluorescent magnetic particle indications on shell fluorescent magnetic particle indications on shell
casings casings

J~ faJ

-------...--OUTS
DE DIAMETER
~m (0.022 in.)

J.. o I mm (0.004 in.)

--.- 0.27 mm (00 I I in.)
I- ·1
_____ --Lt_.46_m_m_(01~;,~~b,AMETER SURFACE AGGLOM~032In.)
. . :::L
~.""~ IUL
HEAD CENTRAL CONDUCTOR
IT = 800 A
IG = 1.000 A
IT = 600 A
IG = 1.000 A
f--J
005 mm (0.002 In.)
,R,.ooo = NONE ' R·, 000 = NONE
IT = 400 A
IG = 6QO A
LEGEND 1"000 = WEAK
Ir= CURRENT AT WHICH INDICATIONS ARE FIRST DETECTED I" I 000 = WEAK
IG = CURRENT AT WHICH INDICATIONS ARE EASILY SEEN
I" x = RESIDUAL CURRENT

fbJ

Correlation of Discontinuity Indication Brightness

A study of indication brightness versus seam depth was
made on two 60 mm (2.4 in.) square steel billets. Ultraviolet
photographs were taken at the ends of six billets at two
continuous current settings and one residual current level. 0.008mm
- . / ) \ '\" 0.02 mm (0 0008 In.)
Figures 12 through 18 show the brightness of magnetic (00003 in.) • t t
particle indications for the various discontinuity depths. ~~ 0.06 mm (0.0027 In.)
Although some clarity is lost in the reproduction process,
the original photographs indicate a direct relationship
STRINGER I (em l
between increased crack depth and increased indication
brightness. 9 -I
0.05 rnm (0.002 in)

I] = 400!\
Ie, = 600 A
Discontinuity Size Limits in Steel Billets ",'(,00 = NONE
I" '000 = WEAK
DUling processing, certain steel billets are heated to a
temperature where the outer 0.4 mm (0.01.5 in.) oxidizes to
form scale. Magnetic particle tests of such billets are
concemed only with cracks likely to remain after scale
LEGEND
removal. The goal of the study below was to lower the
'T= CURRENT AT WHICH INDICATIONS /"\RE FlflST DETECTED
sensitivity of the testing technique to avoid detection of fine IG = CUr/RENT AT WHICH INDICATIONS /\RE EASILY SEEN
tempormy cracks. I" x = RESIDW\L CLJRRENT
Test indications were established using a special low
sensitivity fluorescent wet method pmticle. These coarse
pmticles had diameters from 25 to .50 p.,m (0.001 to
0.002 in.), Test indications were compared to those for
 MAGNETIC PARTICLES AND PARTICLE APPLICATION / 217

FIGURE 7. Tracings of photomicrographs showing fluorescent magnetic particle indications on shell

casings

fa)

7f
EAT TREATING CRACK

. ~mmlO025,"1

IT < 200 A
IG = 200 A
'. 200 = GOOD
'. , .000 = GOOD

fb)

3
0.05 mm 10.0022 in.)
0.001 mm 1000005 in.) I y==o.12mm 10005 in.)
____~t___ < t
'-T"--'
l 100005in)J~- t ~\.--r
'Y, ~
J;mm 10.001 in.).

~0002mm 100001 In)

0.01 mm
'-
/..----1-1-
0.4 mm 10.016 in.)
0.07 mm 100028 in.)
0001 mm

10 00005 'n ~
•

11- TI
mm 10003, ,n I

0.4 mm 10.016 in.)

'T= 1.000 A 'T = 400 A IT = 600 A

IG = 1.200 A IG = 600 A IG = 1.200 A
'. , .200= NONE ' R600 = WEAK '. , .200 = NONE
IR 1.000 = NONE 'R 1.000 = WEAK 'R 1.000 = NONE

LEGEND
'T= CURRENT AT WHICH INDICATIONS ARE FIRST DETECTED
Ie, = CURRENT AT WHICH INDICATIONS ARE EASILY SEEN
I. x = RESIDUAL CURRENT
21 B I MAGNETIC PARTICLE TESTING

FIGURE 8. Tracings of photomicrographs showing fluorescent magnetic particle indications on shell

casings

taJ

0.06 mm (0.0025 If,.) 004 mm (0.0017 in)

==II- -I h
0.002 mm (0000 I in.)

'"I t 0.01 mm 1"...:::I!03 mm 10.0014 in.)

j 0.25 mm (001 in)

(0.0004n)
L~
::__1_-
IT = 400 A IT = 600 A
IG = 1,000 A IG = 1.000 A
I" 1.000 = NONE I" 1 000 = NONE

,bJ

0.005 mm (0.0002 In.) 0.08 mm (0.0032 in.)

\:+
-I I-
r ~, 0~8
Urn
mm (00032 in.)

10028"01 0,15 mm (0.006 In.) t 10003 '"I

1 J HE~i TREAT CRACK

I-l
0.05 mm (00023 in )
HEAT mEAT cr~ACK

IT = 200 A 'T = 200 A IT = 400 A

IG = 400 A IG = 200 A IG = ROO A
'''400 = GOOD '''200 = GOOD I"t'()() = NONE
'II 1.000 = GOOD IR 1.000 = GOOD

lEGEND
'T = CUI(RENT AT WHICH INDICf\T10NS ARE Flr<ST DETECTED
le;= CURRENT AT WHICH INDICATIONS ARE E/\SILY SEEN
Iflx = RESIDUAL CURRENT
 MAGNETIC PARTICLES AND PARTICLE APPLICATION I 219

FIGURE 9. Tracings of photomicrographs showing fluorescent magnetic particle indications on

shell casings

J
(a)

mJ]--+-J
0.0 I mm 100005 In)

0.03 mm 10001 1n)4~m (00007 In)

£,~ 1000041n)

IT = 400 A
IG = 600 A
IR600 = NONE
I. 1.000 = NONE

fb)

[ill'3 ~ 4

-Lo.002mm (0.0001 In)

1
If
Jf
0.06 mm 100024 In ) 0.05 mm 10.002 In)

y ; .~mm (00016 In)

) 1mmlOoo27'01
T
CRACK
...i..
1
043mm 10.0171n)
0 12 mm 10005 In )
CRACK
mm 10 00 IS '" I

I, = 400 A IT = 200 A IT = 400 A IT = 200 A

Ie, = 800 A IG = 200 A IG = 800 A IG = 400 A
1"800 = NONE 1. 200 = GOOD IR800 = NONE IR800 = NONE
I, 1000 = NONE I" I 000 = GOOD IR 1.000 = NONE IR' 000 = NONE

LEGEND
Ir = CURRENT /\T WHICH INDICATIONS ARE FIRST DETECTED
Ie = CURRENT AT WHICH INDIC'\TlONS ,'\RE EASILY SEEN
If<" = RESIDUAL CURRENT
220 I MAGNETIC PARTICLE TESTING

FIGURE 10. Tracings of photomicrographs FIGURE 11. Tracings of photomicrographs

showing fluorescent magnetic particle indications showing fluorescent magnetic particle indications
on shell casings on shell casings

la)
(a)

I\±
O.IBmm 10.0071 in.)
!---t , t 0.03 IYlm (000 14 If).)

~IOOIJ410I IOffiM 101 1 003 mm (0.0013 Ifl.)

" ~
~CRACK ~ 0.05 mm (0002 In.)
W-
0.2 mm (OOOB in.)
IT = 200 A
IG = 200 A IT =200 A IT = 200 A
Ie; = 400 A IG = 400 A
1. 200 = GOOD
'. 400 = NONE '''BOO = NONE
I. 1.000 = GOOD
'.1.000 = NONE I" 1.000 = NONE

(b)
1 T005.~~ (0.0002 In)

,0.04 mm (0.00 17 In)

H-L
0.02 mm (0.001 in.)

'T = 400 A
'G = BOO A
'''800 = NONE

1~±",IOOOlinJ '. , .000 = NONE

0.05 mm (0.002 in.) (b)

IT = 400 A
IG = BOO A
' RBOO = NONE
'. , .000 = NONE

,
'T = 400 A
Ie; = 1.500 A
'. .500
_ _ 0.03 mm 10.0014 in.)

ot 2 mm 10005 in)

= NONE
-rr:
'T = < 200 A
10.002"0·1

'
'. , .000 = NONE
Ie; = 200 A
' RlOO = WEAK
I" I 000 = GOOD

LEGEND
IT= CURRENT AT WHICH INDICATIONS ARE FIRST DETECTED LEGEND
IG = CURRENT AT WHICH INDICATIONS ARE EASILY SEEN IT = CURRENT AT WHICH INDICATIONS AT?E FIRST. DETECTED
I. x = RESIDUAL CURRENT I" = CURRENT AT WHICH INDICATIONS ARE EASILY SEEN
I" x = RESIDUAL CURRENT
 MAGNETIC PARTICLES AND PARTICLE APPLICATION I 221

FIGURE 12. Ultraviolet photographs taken at the FIGURE 13. Ultraviolet photographs taken at the
ends of square steel billets for comparison of ends of square steel billets for comparison of
magnetic particle indication brightness versus magnetic particle indication brightness versus
crack depth: fa) continuous direct current at crack depth: fa) continuous direct current at
200 A; Ib} continuous direct current at 800 A; 200 A; fbJ continuous direct current at 800 A;
and Ic} residual direct current at 200 A and fc) residual direct current at 200 A

la) (a)

Ib, Ib)

Ie, Ie)
222/ MAGNETIC PARTICLE TESTING

FIGURE 14. Ultraviolet photographs taken at the FIGURE 15. Ultraviolet photographs taken at the
ends of square steel billets for comparison of ends of square steel billets for comparison of
magnetic particle indication brightness versus magnetic particle indication brightness versus
crack depth: fa) continuous direct current at crack depth: faJ continuous direct current at
200 A; (b) continuous direct current at 800 A; 200 A; fbI continuous direct current at 800 A;
and fcJ residual direct current at 200 A and (c) residual direct current at 200 A

faJ faJ

fb' fb'

fe) fe)
 MAGNETIC PARTICLES AND PARTICLE APPLICATION I 223

FIGURE 16. Ultraviolet photographs taken at the FIGURE 17. Ultraviolet photographs taken at the
ends of square steel billets for comparison of ends square steel billets for comparison of
magnetic particle indication brightness versus magnetic particle indication brightness versus
crack depth: (a) continuous direct current at crack depth: (a) continuous direct current at
200 A; (b) continuous direct current at 800 A; 200 A; (b) continuous direct current at 800 A;
and (c) residual direct current at 200 A and te) residual direct current at 200 A

fa'
fa)

0.27 mm 10.011 Ill.)

fbI Ib)

fel

0.27 mm (0.011 ill)

224 / MAGNETIC PARTICLE TESTING

FIGURE 1B. Ultraviolet photographs taken at the FIGURE 19. Standard sized fluorescent magnetic
ends of square steel billets for comparison of particles with bath concentrations at 1.25 gol -1
magnetic particle indication brightness versus used to test a 64 mm (2.5 in.) square steel billet
crack depth: (a) continuous direct current at at: (a) 300 A full-wave direct current;
200 A; (bJ continuous direct current at BOO A; (b) 600 A full-wave direct current; and
and (c) residual direct current at 200 A (c) 1,050 A full-wave direct current

fa)
'al

fb)

fe)

fbI

FIGURE 20. Coarse, large sized fluorescent'

magnetic particles with bath concentrations at
1.25 gel -1 used to test a 64 mm (2.5 in.) square
steel billet at: faJ 300 A full-wave direct current;
fbI 600 A full-wave direct current; and
(c) 1,050 A full-wave direct current

fa)

leI ~-----t~ --

,~ , '
:~
fb)

fe)
- -,-. " '- -:~~~
~ ,
 MAGNETIC PARTICLES AND PARTICLE APPLICATION / 225

COI1llll crci <ll powders "'ith 3 to 5 }-tIn (0.1 to 0.2 milli-in.)

TABLE 1. Tangential magnetic field component at
diameters. Figure 19 shows the test indications from stan- centerline of milled reference slots
dard sized particles. Figure 20 shows the results of tests v.1th Magnetic Field
the coarse particles. The larger pmticles did exhibit depth Depth Width Lift-Off Strength (H t )
discrimination, scarcely indicating fine cracks. Crack depth millimeters millimeters millimeters amperes per meter
(inches) (inches) (inches) (oersteds)
was verifIed by grinding.
The lo\\' sensitivity palticles indicated discontinuities with 0.18 (0.007) 0 190 (2.4)
0.38 (0.015)
depths of 0.37 to 0.8 mm (0.015 to 0.033 in.), but did not 0.38 (0.015) 0.18 (0.007) 0.5 (0.02) 111 (1.4)
consistentlv detect shallower cracks. By this cIiterion, sur- 0.38 (0.015) 0.18 (0.007) 1.0 (0.04) 80 (1.0)
face discOl;tinuities less than 0.37 mm (0.015 in.) in depth 0.38 (0.015) 0.18 (0.007) 1.5 (0.06) 48 (0.6)
Illay be called .flnc cracks. Those deeper than 0.8 mm 0.75 (0.03) 0.18 (0.007) 0 549 (6.9)
(0.0:33 in.) may be considered coarse cracks. 0.75 (0.03) 0.18 (0.007) 0.5 (0.02) 334 (4.2)
0.75 (0.03) 0.18 (0.007) 1.0 (0.04) 223 (2.8)
0.75 (0.03) 0.4 (0.016) 0 573 (7.2)
374 (4.7)
Dry Powder Sensitivity 0.75 (0.03)
0.75 (0.03)
0.4 (0.016)
0.4 (0.016)
0.5 (0.02)
1.0 (0.04) 255 (3.2)
0.75 (0.03) 0.5 (0.021) 0 653 (8.2)
Little information exists on the surface discontinuity 0.5 (0.02) 390 (4.9)
0.75 (0.03) 0.5 (0.021)
sensitivity limits for dty magnetic particles. However, a 0.5 (0.021) 1.0 (0.04) 255 (3.2)
0.75 (0.03)
typically sensitive dq powder contains by weight about 35 1.40 (0.057) 0.·18 (0.007) 0 1,274 (16.0)
percent coarse particles, in the 25 to 50 }-tm (0.001 to lAO (0.057) 0.18 (0.007) 0.5 (0.02) 892 (11.2)
0.002 in.) diameter range. It is likely that this portion of the 1.40 (0.057) 0.18 (0.007) 1.0 (0.04) 573 (7.2)
dty powder conglomerate has roughly the same sensitivity as
the similarly sized coarse wet particles cited above.

Effect of Discontinuity Depth on the cracks but they do indicate a downward trend of H t for
.Tangential Field Component decreasing slot' size. At distarices beyond 0 ..5 mm (0.02 in.)
from the slot centerline, there is no reading for the
In a pmticle sensitivity study, measurements were record-
ed for the leakage fields emanating from a seIies of five tangential compone~t.
milled slots. The slots were milled lengthwise down the
center of one face of a 82 'mm (3.25 in.) square steel billet.
Three of the slots were a constant 0.12 mm (0.07 in.) Conclusion
Width, with depths of 0.38, 0.75 and 1.25 mm (0.015, 0.03
and 0.05 in.). The two other slots were a constant 0.75 mm The magnetic particle testing technique is extenSively
(0.0:3 in.) depth \\1th \vidths of 0.4 and 0.5 mm (0.016 and used in virtually all of the world's major industries - its
0.021 in.). All measurements were made at a magnetizing successful application on any ferromagnetic matelial has
current of 500 A using full-wave rectified direct current. ensured its position among the most valuable nondestruc-
The tangential component of the leakage field H twas tive testing methods. Because it is so widely used, so simple
measured with a meter operating in the differential mode. to apply and so familiar, the complexity of magnetic particle
Errors in prohe position were no greater than 0 ..5 mm (0.02 testing is often underestimated.
in.). Field strength readings were estimated to be accurate The technique is, in fact, founded on the complicated
to 40 A-m - 1 (0..5 Oe). The results of this stud" are summa- interactions of several electromagnetic fields and at least
l;zed in Table 1. .. two separate materials for evety test. The particles serve as
Note that HI increases with the slot depth. The tangential magnetic field sensors and, as with any clitical testing
component also increases \vith slot width, though possibly component, need to be fully understood for efficient and
not in direct propOltion. These slots are not particularly fine accurate use.
 226/ MAGNETIC PARTICLE TESTING

REFERENCES

1. Moyer, M. and B. Dale. "Methods for Evaluating the

Quality of Oilfield Tubular Inspections." Journal of 5. Doane, F.B. and c.E. Betz. PrinCiples of Afagnajlll
Petroleum Technology (January 1986). third edition. Chicago, IL: Magnaflux COIporatio
2. Bozorth, R.M. Ferromagnetism. New York, NY: D.Van (1948): p 178.
Nostrand Publishing (1951).
6. Allen, T. Particle Size Measurement. Chapman and Ha
3. US Department of Labor. Occupational Safety and Publishing (1965).
Health Administration. Flammable and Combustible
Liquids. 29 CFR 1910.106 through 29 CFR 1910.108 7. Grutzmacher, R. Internal applications laboratory re
(1987). port. Chicago, IL: Magnaflux Corporation (1955).
4. Skeie, K. and D. Hagemaier. "Quantifying Magnetic
8. Schroeder, K. Internal applications laboratOlY repOlt.
Particle Inspection." Materials Evaluation. Vol. 46,
Chicago, IL: Magnaflux COIporation (195.5, 19.59).
No.6. Columbus, OH: The American Society for Non-
destructive Testing (May 1988): p 779-785. 9. LorenZi, D. Internal applications laboratory repOlt.
Chicago, IL: Magnaflux Corporation (198.5).
·,
1
SECTION 9
DETECTION AND EVALUATION
OF MAGNETIC PARTICLE TEST
INDICATIONS
Henry Ridder, Professional Engineering Services, West Hills, California
. J. Thomas Schmidt, NOT. consultant, Ar,iington Heights, Hlinois
228/ MAGNETIC PARTICLE TESTING

INTRODUCTION

This section of Magnetic Particle Testing covers the two Part 2 covers detection devices, plimarily the human eye.
procedures that occur after the formation of discontinuity Part :3 outlines the categOlies used for evaluating test
indications: (1) detection of the test indications and (2) in- indications. Part 4 covers the process controls used to verif~'
teIpretation of their meaning. that the magnetic particle test has been properly performed,
With information on contrast ratios, Part 1 discusses the providing discontinuity indications for accurate detection
detectability of visible and fluorescent test indications. and interpretation.
 DETECTION AND EVALUATION OF MAGNETIC PARTICLE TEST INDICATIONS I 229

PART 1
CONTRAST AND CONTRAST RATIO
"tvlagnetic p::lIiicle discontinuity indications on the surface detectors. Contrast ratios as low as 1.25: 1 can be detected,
of a test object serve no purpose until they are detected and although the ratio should be at least 2: 1 to achieve reason-
intelpreted. Until the early 1950s, al1 indication detection able probability of detection.
was accomplished by the eye of the human inspector. In In visible test systems, there is sometimes a correlation
some cases, detected indications were then recorded by between bIightness and color of the particles: low bright-
photography, bllt no other mechanical or electronic detec- ness patticles are often dark colored and high brightness
tion systems existed. patticles are light colored.
. Today, dcctro-optkal devices can be used to detect In fluorescent systems, the indication is bright, both
discontinnity indications and computers are then used to because it emits large amounts of visible light and because
intelpret the detected signals. Because these devices are ultraviolet light is not visible to the eye. This makes a
costly and illflexible, their use is limited to high volume nonfluorescent background appear low in brightness and
applications where speed and reproducibility are cost justi- the contrast ratio is accordingly high. In a fully darkened
fied. Nearly all magnetic pmticle tests still base their . viewing area, the effective contrast ratio can be 200: 1 or
procedures on the human eye as a detector. To obtain the higher.
highly sensitive results within the capability of the method,
it is therefore necessmy to understand the operation and
limitations of the eye. Color Contrast
Since all test r~sults do not automatically indicate a
'rejectable discontinuity, it is also necessalY to understand The visible portion of the electromagnetic spectrum in-
. test indication interpretation. Some magnetic particle indi- ~ludes wavetengths betwe(!n 400 and 700 nanometers (nm).
o cations are relevant, others are not relevant and stilloo~hers If all wavelengths are present in equal amounts, the light is
are false indications of discontinuities. white. If no wavelengths '~re present, the eye sees black.
Test indications are visually detectable because they Light from 450 to 480 nm is pure or saturated blue. Light
contrast with the background they are formed on. All de- from 550 to 600 nm is pure or saturated red. If both
tection devices, includiIlg the eye, work by detecting this wavelengths are present, the color is perceived as purple.
contrast. In electronics, contrast is. called signal-to-noise Few visible light sources produce pure light of any color.
ratio. Provided there is e,nough energy present to operate a Generally, the light is predominantly of one color but also
detector, then the greater the contrast, the more reliable is contains some of all the other visible wavelengths. As the
its detection. Two types of contrast are important to mag- proportion of white light (all wavelengths) to colored light
netic paltic1e tests: hrightness contrast and color contrast. (a specific range of wavelengths) increases, the color be-
comes less saturated or more impure.
Brightness Contrast Color contrast can be defined as the difference between
two colors of equal brightness and it therefore includes both
B.rightness contrast is the amount of light reflected by one color (wavelength) and saturation (percentage of a certain
surface compared to that reflected by another adjacent wavelength). At the same brightness, two impure colors
surface. High blightness materials transmit or reflect most have velY little contrast, while one pure color and one
of the light shiking them, a characteIistic related to the impure color may contrast strongly.
~Ul~ace texture of the mateIial. Magnetic particle test Color contrast contributes a small percentage to total
mdlcations can he high or low brightness, so IOJig as the visibility or total detectability, with bIightness contrast
background is the opposite blightness. making up the primary contribution. A pure red indication
!n the visible test systems, indications are typically low is not highly visible against a pure blue background of equal
bnghtness and the backgrounds are high brightness. The pUlity and bIightness. In fact, such a situation can decrease
backwound may reflect 15 to 75 percent of incident light. detectability because of eye strain (red and blue wave-
T~s~ l?dicatiolls may reflect 3 to 8 percent of incident light. lengths focus at different points in the eye).
?lv1dmg an average background reflection of 4,5 percent by In nondestructive testing, the magnetic particle material
an average indication reflection of 5 percent produces a usually has a relatively pure color, while the background is
COntrast ratio of 9: 1. typically as nearly white or nearly black as pOSSible. The
I[ this indication is large enough to be resolved, a 9: 1 main importance of color to detectability is its difference
Contrast ratio is eaSily seen by the human eye and most other from the background.
230 / MAGNETIC PARTICLE TESTING

PART 2
DETECTION DEVICES

There are three types of detection devices for magnetic

particle testing: the human eye, imaging detectors (cameras) FIGURE 1. Average human eye response to
and nonimaging detectors (scanners). various wavelengths at different light levels

VISIBLE LIGHT

The Human Eye as a Detector

For magnetic particle tests, there are two important
characteristics of visible light. Illumination is measured in
lux (Ix) or footcandles (ftc). Luminance or photometric
brightness is measured in candela per square meter
(cd-m -2) or footlamberts (ftL). For the practical purposes.
of visible magnetic particle tests, these two characteristics
may be considered equivalent.. . lOa I:-.----+----+---+--+--+----\---------l
The eye is the first and still the most common detector
used in the magnetic particle testing industry. When con-
sidered simply as a component of the test system, the eye is
widely available, highly sensitive, very flexible and inter-
faced to a sophisticated computing device that can produce
instantaneous interpretations of test results.
The eye is sensitive to both color and brightness. How-
ever, it is not equally sensitive to brightness at all intensity
levels and it is not equally sensitive to all colors. The human
eye responds to all wavelengths between 400 and 700 nm
but it responds most strongly to the yellow green wave-
lengths in the center of the spectmm. The response curve
is bell shaped, falling off to no response at about 400 and
700 nm (see Fig. 1).
The eye changes its absolute and color sensitivity at
different ambient light levels. Figure 1 shows this effect for
three typical light levels: 340, 3.4 and 0.03 cd-m -2 (100,1.0 0.1 f------jf------jf----I--

and 0.01 ftL).

Bright light Conditions

The 340 cd-m - 2 (100 ftL) level is the amount of light
0.01 L - -_ _' - - -_ _, _ _.L-...L--L_---.J._ _--'-_ _~
found in a brightly lighted indoor room or outdoors on a 600
100 200 300 400 500
bright day in deep shade. It is a light level often used for
visible magnetic pmticle tests. It is also the level where the WAVELENGTH
(nanometers)
eye is considered fully adapted to bright light and where
increased brightness does not cause further sensitivity
 DETECTION AND EVALUATION OF MAGNETIC PARTICLE TEST INDICATIONS I 231

change. Under these conditions, the peak sensitivity is at conditions. Color is barely discernible and most test indica-
555 nm and the eye is only 10 percent as sensitive to blue tions are perceived as blue white.
(470 nm) and red (650 nm) light. This means that blue and The eye does not adapt instantly to changes in light level.
red appear only one tenth as bright as yellow green or that Moving from full bright adaptation to full dark adaptation
ten times as much blue and red light is needed to appear as can take as long as twenty minutes. Going from full bright
blight as yellow green. adaptation to the 3.4 cd-m -2 (l ftL) level can take two
Figure 1 also indicates why fluorescent palticles are often minutes. If critical tests are being performed, the inspector
green, yellow or orange: Since the peak response of the eye should wait at least two minutes after entering the booth
is to yellow, less light of this color is required for minimum before beginning the test. Less critical tests in lighter booths
visibility. Less exciting ultraviolet radiation (or less fluores- can often be done after 30 seconds of adaptation time.
cent dye surface area) is necessalY for the indication to be With visible magnetic pmticles, testing should not be
visible. Green (S20 nm) or orange (590 mn) produce 75 attempted with less than 100 Ix (10 ftc). Between 300 and
percent as much eye response as yellow. If a testing 1,000 Ix (30 and 100 ftc) are best for most visible testing
application requires one of these for color differentiation, applications. Critical tests of small discontinuities may
the drop in lnightness is often acceptable in exchange for require 2,000 to 5,000 Ix (200 to 500 ftc). Extended testing
the increased color contrast. at levels over 2,000 Ix (200 ftc) may produce eyestrain.
Fluorescent tests should be carried out in the darkest
conditions possible. Specifications typically call for less than
Low Light Conditions
20 Ix (2 ftc) of visible light in the booth and in some cases
At low light levels, the operating mode of the eye changes. the measurement must be made at the test site on the test
The retina or detecting portion of the eye is coated with two object surface.
types of receptors known as rods and cones. Cones are used Rough magnetic particle tests can be made with up to
in high brightness conditions and respond to color as well as 100 Ix (10 ftc) but contrast levels and detectabilitv are low.
brightness. Rods have much more sensitivity to light and are A light level of soo lx.(SO ftc) may be acceptable if (1) large
used in low brightness conditions, but they are. do not detect amounts of ultraViolet are available; (2) bright fluorescent
color. palticles are being used; and (3) high sensitivity is not
Cone vision (340 cd-m -2 or 100 ftL) is called photopic required. In no case should fluorescent tests be attempted
vision. Rod vision (0.03 cd-m -2 or 0.01 ftL) is called when the visible light level is higher than 1,000 Ix (100 ftc).
scotopic vision ..The region in between, where both rods and Fluorescent tests should be done outdoors in daylight only
cones are partially operative, is called rnesopic vision. Full if the test site can be partially darkened.
rod vision is about 800 times' as sensitive to brightness as full
cone vision but it is totally insensitive to color. The eye
cannot see color in low light. Personnel Vision Requirements
Although the dark adapted eye cannot distinguish or dif- In order to properly perform magnetic particle tests, it is
ferentiate colors, it can detect certain wavelengths and necessary for the inspector to possess a ceIiain level of visual
is more sensitive to some colors than others. As the acuity. Vision tests are typically done with near vision and all
0.03 cd-m - 2 or 0.01 ftL curve shows, the maximum re- inspectors should be tested to ensure proper visual acuity.
sponse wavelength shifts to 512 nm (blue green) and most The American Society for Nondestructive Testing's Recom-
sensitivity to red light is lost. However, sensith,ity to blue mended Practice SNT-TC-1A recommends minimum type
and ultraviolet increases. It is possible to see wavelengths as size reading ability at specified distances. The document
1m\' as 350 nm in the absence of all other light. also recommends that the inspector demonstrate the ability
to differentiate between colors used in the magnetic particle
test method.
Lighting in Typical Testing Conditions Good close vision is required to properly perform mag-
In a magnetic paIiicle testing booth, it is not possible to netic particle tests. Good color vision may be helpful in
attain full darkness, Ultraviolet lamps emit some blue visible visible light tests, but it is not as critically impOltant for
light and stray fluorescence or light leaks are almost un- fluorescent tests because evervone loses some color vision in
avoidable. The light level in most testing booths ranges the low light levels of a testiI;g booth.
between 1.7 and 14 cd-m -2 (0.5 and 4 ftL).
A good booth has about 3 cd-m - 2 (l ftL) of visible light
on the test object surface. Figure 1 shows that peak eye Scanning Detectors
response at this level is at 530 nm (green) and that light of
380 nm and above is visible. The figure also indicates that Scanners are nonimaging detection systems. In their
the eye is about 30 times as sensitive as in full bright Simplest form, they consist of a light source, a means of
 232 / MAGNETIC PARTICLE TESTING

moving the test object, a detector and signal processing unwanted light and undesirable background can often he
equipment. In use, the light source illuminates the test eliminated.
object surface and the detector measures the light reflected Scanner systems are relatively simple, compared to tele-
or emitted from it. The test object is moved past the vision systems. Scanners are optically and electronically less
detector in a uniform manner. If the measured amount of complicated and are potentially less costly. They are often
light changes significantly, a discontinuity indication is deSigned with very sensitive detectors. A major advantage is
detected. the scanners' great depth of field. They usually have no
Usually, the light source is focused or collimated into a lenses in the detector system and do not need to be focllsed.
narrow beam that moves over the test surface. Since only a
small area is illuminated, it is possible to exactly locate any
indication that may be detected. It is also possible to focus Television Detectors
the detector on a small area, but the mechanical alignment
becomes much harder. Television is often used as a test indication detector ill
Scanners may be built for use with either the visible or automatic magnetic palticle systems. Television cameras
fluorescent method. With the visible method, the light produce images that may include the test object as well as
source can emit any wavelength the detector can see. Lasers the test indications. This property allows more sophisticated
are often used as light sources because they produce very interpretation, a task usually accomplished through ~t dedi-
narrow, intense beams. "Vith visible systems, the reflection cated computer.
from the background is typically high and reflection from Television camera detector systems use broad field illu-
the indication is low. A sudden decrease in signal results mination sources that cover a large portion of the test
from the presence of a particle an indication. object. Television is essentially sensitive to visible Jight so
With the fluorescent systems, the detector is covered by standard visible or ultraviolet light sources may be used.
a filter that absorbs the wavelength of the light source but These sources may in turn be specially filtered or filters may
transmits' the wavelengths produced by fluorescence in the also be placed over the camera lens. As with scanners,
particles. It is not necessary that either light be in the visible. television does not have the same color sensitivity as the
range, as long as the detector is sensitive to the emitted human ey.e, so care must be taken to guarantee the com-
light. patibility of light sources, camera and test materials.
The photoelectric detectors used in scanners sel?om have . Because .of the higher contrast ratios, most television
the same color or brightness sensitivity as the human eye. It camera detector systems use fluorescent testing techniqlles.
. is necessary that the detector be sensitive to some of the. In such tests, the light source and camera are filtered to
. emitted or reflected wavelengths. Filters are used to remove transmit and receive the necessary wavelengths.
 DETECTION AND EVALUATION OF MAGNETIC PARTICLE TEST INDICATIONS I 233

PART 3
INTERPRETATION OF DISCONTINUITY
INDICATIONS

Intelpretatioll is the culmination of the magnetic particle Sensors can and should be used to indicate the pres-
.test procedure but certain interpretation decisions must be ence and direction of magnetic flux leakage, particularly
made early in the testing process. Two of the most critical with test objects having changes in cross section or unusual
involve the appropriateness of the magnetic pmticle meth- configurations. Sensors can also provide a relative value for
od: (1) is it the correct testing technique for the anticipated the number of flux lines needed to obtain specified test
discontinuities; and (2) is it appropriate for the type of test reliability.
object and its characteristics. Once these decisions are
made, magnetic particle indications can be formed. These
are then detected and interpreted to determine if they Choosing the Magnetic
represent discontinuities and, if so, to determine severity
and the effect on test obje¢t service. Particle Technique
The fact that an object is ferromagnetic does not mean
that magnetic particle testing is the best surface inspection
. Testing for Subsurface method. The exceptions .depend mainly on the size, shape
Discontinuities and finish of the test object. For example, under some
specifications, over-magnetization of finish machined ob-
MagnetiC palticle tests can be used to locate specific jects can create high background, making the magnetic
types of subsurface discontinuities in ferromagnetic materi- particle test ineffective. In forgings, flow lines and particle
als, depending on: (1) the t)1)e of magnetic particle equip- buildup at comers or grooves produce non relevant indica-
ment and the kind of magnetizing current; (2) the nature tions that mask possible discontinuities and reduce the
and characteristics of the discontinuity (its orientation and effectiveness of the magnetic particle method.
depth under the surface); and (3) the dimensions and shape In addition, if magnetizing current is applied. to the test
of the test object. This ability to detect subsurface discon- object by direct contact, there is a high probability of arcing
tinuities should not be overemphasized. Other nondestruc- and the resultant damage to fine surfaces. In aerospace
tive test methods have better capabilities for subsurface components, for example, arcing inevitably causes minute
discontinuity detection. cracks into which copper from contact pads can penetrate.
For the dI)' particle method, only the detection of major Copper penetration is cause for rejection in virtually all
linear discontinuities (sluinkage cracks or incomplete pen- aerospace components.
etration) is possible and then only if the discontinuity's
depth below the object surface is less than 6 mm (0.25 in.).
\Vhen the wet method is used, the maximum depth below Determining the Nature of
the surface is 0.2 mm (0.008 in.). These depth limits are
b~sed on empiIical data and vary with type and size of the an Indication
dIscontinuity, the dimensions and shape of the test object,
and with the magnetic properties of the test object material. The first step in interpretation is to decide the character
of a magnetic particle indication: is it relevant, nonrelevant
or a false indication?
Effect of Test Object Shape
If it is relevant (representative of an actual material
The shape of a test object can cause magnetic flux lines to discontinuity), the indication is then evaluated, based on the
bypass some areas of the object surface. No flux leakage data given in the acceptance criteria for that type of test
Occurs in those areas and no particle indications are formed, object. If the test object is within the acceptance criteria, it
despite the presence of possible discontinuities. can be accepted, even though there is a discontinuity
234 / MAGNETIC PARTICLE TESTING

present. However, such acceptable discontinuities do not strength of the test object but it can cause nonrelevant
affect the quality, future use or service life of the test object. particle indications. In some cases, high internal and exter-
If the acceptance criteria are not met, the test object is nal stresses cause variations in magnetic properties and this
rejected and a full report goes for further evaluation to a can also cause nonrelevant indications.
materials review board. Members of that board typically
include quality assurance managers, engineers and in most
cases a customer representative. Contact Indications and Stamping Marks
The review board decides future action for the test
object: scrap, use as is, rework or repair. After rework or Objects that touch each other during magnetization may
repair the object is retested. Scrapped components are set up local polaIities at their surfaces and can cause leakage
destroyed to prevent accidental use. fields. This happens most often when a number of objects
N on relevant indications are caused by magnetic leakage on a central conductor are magnetized at the same time.
fields resulting from the shape of the test object: sharp Contact indications are sometimes called magnetic writing.
comers, splines, thread roots or magnetic writing, for Confusing particle patterns can be caused and, as with all
instance. Nonrelevant indications are not cause for rejection non relevant indications, these patterns can mask relevant
but they can mask actual discontinuities critical to the discontinuity indications.
object's service life. Proper testing techniques reduce the Numbers" or letters sh~mped onto a component are
occurrence of non relevant indications. sometimes ground out before magnetic particle testing. The
False indications are not caused by magnetic flux fields changes to the structure caused by stamping are sometimes
but by material obstmctions and improper processing: dirt, sufficient to cause changes in the metal's magnetic proper-
fingerprints, gravity, scale or drain lines are examples. ties. Magnetic particles occasionally can indicate these
Proper housekeeping can prevent false indications. They are magnetic changes and Otitline or indicate the previously
not cause for rejection but often require retesting before stamped characters. This is considered a nonrelevant indi-
acceptance. cation but it may also be used to benefit by verifYing the test
object's identification.

Nonrelevant Test Indications Structural Indications

Nonrelevant indications are caused by magnetic leakage At abrupt changes in section of a magnetized object,
fields but they do not represent accidental discontinuities. there is an increase in internal flux density which in tum
Flux leakage may be caused by shmp comers, holes drilled creates local external polarity and can produce magnetic
close to the surface, threads, changes in the structure of the particle indications. Sharp corners, keyways, intemal splines
material, shrink fits or dissimilar metals. or holes close and parallel to the object surface are the kinds
The main problems with nonrelevant indications are: of component design features that produce nonrelevant
(1) they can mask actual discontinuities or (2) actual discon- indications. Such indications are characterized by their
tinuities can be interpreted as nonrelevant. width and lack of clarity. Their relationship to the object's
design is usually apparent.
Again, the primary concern over structural indications
is their ability to mask relevant discontinuity indications.
Metallurgical Properties \Vhen structural indications occur, the magnetic "lines of
Alloy and hardness directly influence the magnetic prop- force exit the mateIial normal to the surface, an llnfavor-
eities of metals such as steel. Variations in hardness from ahle orientation for the detection of discontinuities. Quick
cold working create localized variations in magnetic prop- break magnetization can partially eliminate structural
erties. Depending on the alloy, such vaIiation can cause discontinuities.
magnetic particles to form sharp, distinct patterns. A shIink fit may also be considered a structural indication.
Heat affected zones near welds can give similar indica- The interface between two objects gives a distinct magnetic
tions. Some of the tool steels with high retentivity and high particle indication that is apparently not related to the
coercive force are well known to produce sharp, well pressure used for the shlink fit operation.
defined nonrelevant patterns. If such a condition occurs, Pmticle indications seen in the roots of threads are often
magnetic particle test results should be veIified by a test caused by gravity rather than by a magnetic leakage field.
method such as a liquid penetrant or ultrasonic testing. Very careful examination of the threaded area is required to
When a coupon is taken from such an area and metallur- distinguish relevant hom nonrelevant indications. In this
gically examined, a minor change in the grain structure is case, relevant indications turn in a slightly transverse direc-
often visible. This change does not typically influence the tion vvith a hook at the end of a crack.
 DETECTION AND EVALUATION OF MAGNETIC PARTICLE TEST INDICATIONS I 235

Dissimilar Metals and Welded Joints forming patterns that resemble discontinuity indications.
These indications do not reappear after the object is cleaned
\Vhen two metals with different magnetic properties are
and retested, thereby establishing their false character.
fused together, the interface produces a very sharp indica-
Scale on the test object sUlface often produces false test
tion during magnetic particle testing. An automotive valve is
indications but the source of scale indications is simply
typical of such a component: the valve body is fusion welded
determined and accounted for. If scale is forced into the
to a stem fabricated from a different metal. Under test, the
surface of the test object during forging operations, a
fusion line shows up very clearly, making it impossible to
significant discontinuity is formed and, depending on the
inspect the weld for lack of bond. A supplementaty ultra-
stage of manufacture, such a discontinuity can be very
sonic test is required for complete inspection of an object
relevant to future service life. .
such as a valve body, containing dissimilar metals.
In addition to the indications that can occur in the heat
Scratches and Burrs
affected zone of welds, nonrelevant patticle indications can
also appear in the weld reinforcement. Indications caused Surface scratches and burrs trap magnetic particles and
by abrupt changes in the crack's cross section can also occur form patterns that may have the characteristics of a discon-
at that location. A very thorough visual test with magnifica- tinuity indication. For example, these false indications can
tion or gIinding of the weld cap is often necessary to mimic a crack with an orientation transverse to the magnetic
supplement magnetic particle procedures. particle flow. Often, such indications can be distinguished
from relevant indications by the lack of particle buildup.
Verification at 10 x magnification in visible light may
sometimes prove necessary.
False Test Indications Scratches and burrs are classified as false indications,
unless the scratches occur in notch sensitive and highly
Dirt and Scale
stressed materials,. on polished surfaces or if burrs are fDund
If a test object is inlproperly cleaned and foreign material in threads and splines.
remains on the surface, magRetic particles may become A good visual examination before the magnetic particle
trapped. Improp~r cleaning contributes to contamination of test generally locates these conditions. Since the indications
a wet ,Patiicle suspension and it causes drainage lines, are linear, they must be reported.
236 / MAGNETIC PARTICLE TESTING

PART 4
PROCESS CONTROL OF MAGNETIC
PARTICLE TESTS

Note that the particles in suspension are an insignificant

Technique Sheets part of a testing operation's total cost. Frequently changing
the magnetic particle suspension does not materially inflate
One of the basic ways for controlling the outcome of a the cost of the testing operation. Even if the bath is changed
magnetic particle test is to properly use technique sheets each week, it is still a very small fraction of the total cost,
(see Figs. 2 and 3). These forms comprise a valuable record when compared to equipment depreciation and labor.
of the test procedure, including all pertinent magnetizing
data: (1) the direction of the magnetic field; (2) how the
field is generated (coil, central conductor, yoke); and (3) the Uses of the Settling Test
magnetizing current values. Certain positional data can also . Low signal-to-noise ratio is the principal reason for f~~ilure
be found on technique sheets (the test object's location . 'to detect fluorescent discontinuity indications - it is nearly
within the coil, the yoke contact points .or the placement of impossible to detect fluorescent indications in high fJuores-
shim standards, for example). .. cent backgrounds, for either automated tests or the human
Under military standards, such records are mandatory eye. The principal causes of low signal-to-noise ratio are:
and must be Signed and approved by a certified Level III (1) excessive current denSity, (2) excessive magnetic parti-
magnetic particle inspector. In most cases, technique sheets cles in suspension, (3) excessive fluorescent background in
are retained for future use. Filing systems are often based the vehicle or (4) excessive particle contamination. The set-
on part number and technique information can be trans- tling test can detect tlle last three causes.
ferred to computerized data banks. In a typical setting test done with fresh pmticles, a
As a source of empirical data, technique sheets perform a concentration of 0.15 to 0.25 mL particles should be found
vital system analysis function. Their information can help in a 100 mL centrifuge tube. A lower concentration, on the
determine the cause of problems that occur throughout order of 0.1 mL is usually satisfactory. Anything above
magnetic particle test procedures. If properly completed 0.3 mL is excessive and should be avoided, even though
and retained, technique sheets can also prOvide valuable many specifications allow up to 0.5 mL.
information on successful inspection setups used for similar The settling test may also be used to determine two of the
test objects. primaly kinds of bath contamination: (1) the loose fluOl'es-
cent material from the particles themselves; and (2) extra-
neous oils (such as cutting oils) that remain on the test
Control of Wet Method object after cleaning. The degree of such contaminatio.n ~~n
be monitored with a centrifuge tube that retains an lI1ltml
Particle Concentration sample of the vehicle for reference purposes. This is then
compared to a concentration test after at least one hour
Cost of Suspension Control settling.
The majotity of wet magnetic particle installations de- Another source of contamination is sand from prior sand
pend on the settling test for control of paIticle concentra- blasting operations, residue from grinding or shot dilt.
tions in suspension. The settling test has been in use since These contamination sources are the result of inadequate
the 1940s and it is now possible to use this simple technique precleaning and can be determined by a settling
to detect other attributes in a particle bath. comparison.
 DETECTION AND EVALUATION OF MAGNETIC PARTICLE TEST INDICATIONS 1237

FIGURE 2. Technique sheet for wet method magnetic particle testing

238 / MAGNETIC PARTICLE TESTING

FIGURE 3. Technique sheet for dry method magnetic particle tests of welds and castings

TECHNIOUE SHEET
Magnetic Particle Testing
Dry Method
Welds and Castings
IDENTIFICATION
Job Site _______ _
Construc!lon

MAGNETIZATION MOliel /1tl1per3iJ e SpCfCtng

Prods
COil
Yoke

MATERIALS
o Dry powder color
:J Black oXICie suspension sprdy can

APPLICABLE SPECIFICATIONS: ----~---- __ ~--~------------------ - ------ -~----"~~7ffS%:t3jift~C-~;:J

o Single V groove o Single J groove

o Doul'll.;> V groove o Double J glOOVI'
o Corner, edge flange o Stngle U groove
o Square groove G Double U groove
o Single tJevel groove o Flare bevel qroQve
o Double bevel groov<: o Flare V qroove
o Full penetration o Pamal penepafton
JOINTS
CBut!JOlrlt • o LilP.lolnt
o Corner Joint o E(Jgejolnt
[I Tee JOint

CASTINGS
Surface conditio'! ------~-- - -":'- ----------~------------- ----:--c,-----"-;.;,. ::<C:"i;~;::;::~<;;,:d

SYMBOLS
Conteler pOrnt) +. f + +
Iflrernal conductor
F,cld dlleCliofl

The ttI5PC({(lf" ')hal! t'OS!/ft=: cornpIJ.:1[)Cf.' ~·1/i(h c:1/1leqUirerrJenn" of lilt:" (]lJ.:1!Ity

c7S"SlI{rincr pro( t\1Uft'5 to! rn.7t.]n:l{IC par ticle le_\[f09

INSPECTOR: FOR APPROVAL:

Signaturr' Siynatl1le
SNT-TC-IA LevellJ MT SNT-TCIA Lc'vrllJl MT
Delre ___ _ Dare
 DETECTION AND EVALUATION OF MAGNETIC PARTICLE TEST INDICATIONS I 239

Brilliance and Contamination Tests

Each week that magnetic particle equipment is used, the
brilliance of fluorescent suspension should be checked by
Control of particles in suspension at the pOint of the comparing it to a fresh, unused sample. Whenever suspen-
discontinuity is of vital importance. Fluorescent magnetic sion is mixed, a minimum of 200 mL should be retained in
particles are used in the overwhelming number of applica- a dark glass bottle for reference purposes.
tions (about 10 percent of the magnetic particle suspensions Used suspension is compared with the unused sample
are nonfluorescent), so that fluorescent particle control is under ultraviolet light. If the brilliance is noticeably differ-
more often needed than control of visible pmticles. Ideally, ent, the test system's tank should be drained, cleaned and
a discontinuity should have a high contrast ratio under refilled \vith fresh sllspension. The results of these proce-
ultraviolet ligl~t. The indication should be shalp with virtu- dures are recorded in the log book.
ally no background either from the particles themselves or Contamination of the suspension by foreign matter should
from the vehicle. also be checked at least once a week. If the suspension is
The relative number of particles in suspension per unit contaminated, the tank should be drained, cleaned and
.. volume should he controlled within close limits. This is refilled with fresh suspension. The following procedure is
important because the basis for evaluation of an apparent used to verify contamination.
discontinuity is typically the Visual detection of particle
accumulation. Mass fluorescence in the background of an 1. Run the circulating pump for thirty minutes.
indication can greatly diminish detectability by Significantly 2. Fill a graduated 100 mL centrifuge tube With suspen-
lowering the contrast between .background and indication. sion and allow it to stand for sixty minutes.
3. Examine the liquid above the precipitate with an
ultraviolet light. If the oil or ~ater fluoresces. green or
yellow green, the tank should be drained, cleaned and
Application of a Wet Method Concentration Test
refilled with fresh suspension.
After a maximum of eight hours; the concentration of a. 4. Examine the precipitate. If two distind layers are
magnetiC particle suspension s40uld be measured· (concen- visible and the top layer of contaminate exceeds 50
tration must be at the proper level before testing can percent of the bottom layer's magnetic particles, then
continue). If there are insufficient particles in the suspen- the tank is drained, cleaned and refilled with fresh
sion no indications can form. Too many particles give a high suspension.
background that can mask small indications. 5. Record the results in the log book.
The concentration is usually measured with a technique
called a settling test that is pelformed as follows. Specifications may vary in the allowable ratio between
particles and contaminant in the precipitate. Results of
these tests are always logged in writing.
1. Run the circulating pump for a minimum of thirty
Since both of these tests take at least sixty minutes, it is
minutes. normal to begin magnetic palticle testing before the bril-
2. Flow enollgh suspension through an applicator nozzle
liance and contamination procedures are completed. If the
to ensure uniformity of the suspension.
suspension does not meet specification, all objects tested
3. Fill a thoroughly cleaned, 100 mL centrifuge tube to during the thirty minutes must be retested with new
the 100 mL line with the suspension sample.
suspension.
4. Allow the tube to stand away from magnetic fields and
Vibration for sixty minutes when oil is the vehicle,
thirty minutes when water is the vehicle.
5. Veri!)! the volume of settled particles on the tube's Viscosity Test
graduated scale. The precipitate volume should be
0.15 to 0.3 mL per 100 mL for fluorescent parti- Viscosity is a measure of a fluid's resistance to flow. It is
cles and 1.0 to 2.4 mL per 100 mL for black oxide an important property of the oil used as a vehicle for
particles. magnetic particle testing. Viscosity is measured in square
6. Adjust the suspension concentration by adding parti- millimeters per second (these units are sometimes called
cles or vehicle and repeat the settling test. centistokes, after George Stokes, the British physicist who
7. Repeat step 6 until the correct concentration is ob- developed the theory of motion through viscous fluids).
tained. For many years, petroleum distillates or kerosene was
8. Record each concentration reading in the log book. used as a vehicle for magnetic particle tests. The flash point
240 / MAGNETIC PARTICLE TESTING

of these vehicles was typically around 60°C (140 OF). The

FIGURE 4. Schematic diagram of tool steel ring
Occupational Safety and Health Act of 1972 raised the
used as a reference standard
requirement for open tank usage of such vehicles to 93°C
(200 OF) for a tank \\-lth surface exposure of 0.93 m 2 (10 ft:2)
or more. The viscosity of kerosene was about 40 percent
lower than those vehicles meeting OSHA minimum require-
ments. For equivalence, the settling time for particles had to SIDE TOP
be increased from 30 to 60 minutes.
In general, the measurement of viscosity is done in
specialized commercial laboratories. Common viscosity val- x
ues for the vehicle used in magnetic particle suspensions
are: (1) must not exceed 3 mm 2 ·s- 1 at 38°C (100 OF); and
(2) must not exceed 5 mm 2·s -1 at the temperature of use.
Contaminants from the surface of test objects tend to
build up in the particle suspension and this increases the
viscosity. Precleaning test objects to remove oil and grease
helps resolve but does not eliminate this problem. Depend-
ing on the specifications in force, viscosity measurements
are usually performed monthly. A 90 mL (3 oz) specimen
from the tank is sufficient for accurate appraisal.
If the viscosity exceeds specified values, the suspension is
discarded and the tank is drained, cleaned and refilled with reannealed with the follOwing procedure: (1) heat to 790°C
fresh suspension. The results of viscosity tests are reported (1,450 OF); (2) furnace cool in 10 °C (50 OF) increments to
in the log book (third party reports are also kept on file). 480°C (900 OF); and (3) air cool.
Suspensions, solvents and similar materials must be dis- With the follOwing test parameters, 8 or 9 indications are
carded i~ compliance with fede~al, state and local laws. detected: (1) a fresh suspension of suitable paiiicles having
a concentration of 0.2 mL or 1.6 mL visible particles;
. (2) current at 1,400 A induced willi central conductor;
Steel 'Ring Test (3) light levels at a minimum 3,000 JLvV·cm -.2 fluorescent
light. Increasing the amperage does not increase detectabil-
In the early 1940s, a steel ring was first used to demon- ity. Former indications of holes were set as listed in Table x.
strate that magnetic particle techniques would detect sub- If used with threshold values, the ring standard becomes
surface discontinuities (see Fig. 4). It was difficult then to a measure of the efficacy of particles themselves. It is thus
delineate linear discontinuities that were very tight and used as an acceptance test in many specifications. It cannot
slightly angular, including those cracks found with radiog- indicate overall system effectiveness.
raphy in nominally full penetration welds (ultrasonics was
not yet being used).
The first application of the ring was by Robert Roehrs at Verifying Illumination
McDonnell Aircraft for a measure of system effectiveness.
At that time, there was no other reference standard for
Ultraviolet Light Requirements·
magnetic pmiicle testing.
The ling was made of AISI-01 oil hardening, cold work Ultraviolet light measurements are made daily. The com-
tool steel and was generally machined from a-nnealed bar mon minimum intensity of ultraviolet light for magnetiC
stock. The hardness of the ring was specified between 90 pmiicle testing is 1,000 J-LW·cm -2 at a distance of 380 mill
and 95 on the Rockwell-B scale. (15 in.) from the face of the source (filter or bulb).
In 1985, the ring standard's accuracy was seriously re- Intensity may be measured with a commercial ultraviol~t
futed. I With all parameters remaining equivalent, the num- light meter. Such instmments often measure infrared ra?l-
ber of holes detected in a series of sample tests ranged from ation with the ultraviolet, but filter out visible light. With
3 to 11. In addition, it was determined that the hardness of this kind of meter, measurements of ultraviolet intensity are
the ring did not VaIY from the annealed or unannealed done in three steps: (1) determine the combined intens~.tyof
condition. The only property that was measurable and ultraviolet light and infrared; (2) by means of a hlter,
indicative was initial permeability. It was shown that a ring measure only the infrared radiation; and (3) mathematically
standard that could be repeatedly depended on to show 5 or determine the ultraviolet intensity by taking the difference
6 indications should show 9 indications when annealed or between the two measurements.
 DETECTION AND EVALUATION OF MAGNETIC PARTICLE TEST INDICATIONS I 24'

The foIlov,ing procedure is used. First, switch the meter ampere meter, it is essential that the inspector obtain the
to the least sensitive scale and place the sensor 380 mm exact current value that the instrument indicates. For that
(15 in.) from the center of the light source. Move the sensor reason the equipment must be calibrated at three month
until a il1cL'dmum reading is obtained and record the intervals. This procedure is entered into the log book and a
measured intensity value. Switch the meter to the high calibration sticker is placed on the calibration adjustment of
sensitivity scale and position the infrared filter without the meter.
moving the sensor. Deduct the first reading from the second
to determine the ultraviolet light intensity. Yoke Calibration
If the minimum intensity is not met, the follOWing actions
must be taken. Calibration requirements for magnetic yokes are relative-
ly Simple: the yoke must be able to lift a specific dead
1. Clean thc lamp filter, bulb or reflector. It is manda- weight. For alternating current yokes, a 4.5 kg (10 lb) weight
tory to clean the ultraviolet source on' a daily basis. is lifted with the pole pieces spaced at the testing distance.
2. Replace the ultraviolet light bulb if cleaning does not For direct current yokes: (1) a 13.5 kg (30 lb) weight is
correct the situation. Dispose of used ultraviolet bulbs lifted with the pole pieces 50 to 100 mm (2 to 4 in.) apart;
as hazardous waste. and (2) an 18 kg (40 lb) weight is lifted with the pole pieces
100 to 150 mm (4 to 6 in.) apart.
As a final test for the ultraviolet light source, inspect the The plates for these tests must be certified for the correct
filter for cracks or visible light leaks and replace cracked or weight and are then considered and identified as standards.
broken filters. Do not touch ultraviolet bulbs with the bare Two months is the interval between weight lifting tests for
hands: the acidity of the fingers can cause the glass to fail. yokes.

Visible Light Requ;re.ments.. Verifying the Magneti.c Field

The magnetic particle testing area must be darkened. Pie Gages
Usually, a hood is mounted over the test equipment to keep
out ambient light. Ambient light may not exceed 20 Ix (2 ftc) T11e magnetic field indicator or pie gage was developed in
for fluorescent magnetic particle ~ests. Germany as an aid to determine the direction of magnetic
Ambient light may be measured with a commercial light fields for the detection of discontinuities. It is not a
meter. Such meters should be filtered to exclude ultraviolet quantitative device and cannot determine the adequacy of
radiation. When visible particle suspension is used, the the magnetic field.
intensity of the visible light should be 1,000 Ix (100 ftc) at The indicator is a disk made of ferrous material with very
the testing surface. This may also be measured with a low retentivity. It is cut into four or eight slices, similar to
commercial light meter and a reduction filter. Visible light the slices of a pie. The slices are bonded together with
intensity measurements are performed every week and the nonmagnetic material and, depending on the design, are
results reported in the log book. covered with a thin copper shim or 0.25 mm (0.01 in.)
Light meters must be calibrated every six months. This copper chrome plating.
procedure is recorded in the log book and a calibration The disk is mounted in a holder so that the inspector can
sticker is placed on the meter, covering the calibration place it on the test object during the application of magne-
control. tizing current. Depending on the direction of the magnetic
field, the position of the indicator and the adequacy of the
magnetic field, one or more of the bond lines between the
sections have a sufficient leakage field to attract magnetic
Calibration for Current Output particles. This in turn shows the direction and to some
extent the adequacy of the field. It is recommended that
Under military specifications, the current output of all indicators be demagnetized before each use.
magnetic particle equipment must be calibrated. This is The magnetic particle indicator has limitations and should
u~ually performed by an independent service organization be used accordingly. When placed on a nonmagnetic mate-
\\llth instruments calibrated to National Institute of Stan- rial through which a current is passed, the pie gage shows
dards and Technology (NIST) standards. The NIST stan- indications that could be misleading to an inexperienced
dards should be documented in the calibration report. inspector. With circular magnetization, magnetic flux lines
Since the magnetizing field strength depends on the exit the test object to pass through the deyice (only a strong
CUrrent output of the system and on the accuracy of the magnetizing current causes indications). This is good, since
242 / MAGNETIC PARTICLE TESTING

it gives a margin of safety, but with longitudinal magnetiza- recognized that field strength (H) not flux denSity (B) is
tion, the case is different. The magnetic field generated by being measured. \Vhere the field is essentially contained
the coil passes by preference through the object but it also within the test object, such as in circular magnetization of a
passes through the magnetic particle indicator. For longitu- regularly shaped object, the readings obtained have long
dinal magnetization, such indications do not represent an been disputed as inconsequential.
accurate estimate for the adequacy of the magnetizing However, it has been shown that a suitable instrument
current. can be both quantitative and repetitive, when properly used.
In the absence of a suitable reference standard, an actual The limitations beyond that point are that a meter shows
part representative of the test object has been used. This direction and approximate magnitude but does not show a
so-called flaw standard has the following limitations. gradient such as can be provided with a reference standard.
The Hall effect instrument can be used to establish bases for
1. The discontinuity is usually preferentially in a single quantifying other reference standards.
direction. The Hall effect gaussmeter is relatively useless for guar-
2. The discontlnuity is rarely at threshold but is typically anteeing that magnetic fields are existent and balanced in all
gross. directions in a multidirectional magnetizing operation. The
3. The material is usually retentive and can be used in only practical and inexpensive means of ensuring that this
either a residual or continuous application. balance exists is with the shim standard.
4. Most discontinuities used in this way are so gross that
excessive background becomes a factor.
5. It is frequently possible to use more than one direction Magnetoinductive Instruments
of magnetization to produce a satisfactory indication. It is possible tb electronically obtain a relative value for
the adequacy of a magnetic field. This method is accurate
Consider this illustration of component used as a flaw and simple to apply, but it is costly and may only be used
standard: a small forging with 25 mm ,(1 in.) sections with continuous direct current magnetization.
throughout. A continucyus magnetization, cycle of 60 A in The instrument has two sensor coils to which alternating
either of two directions (circular magnetization) produces a current is applied at a frequency of 100 kHz. The sensor
satisfactory test indication. The usual requirements of coils ~re mounted at right angles to eacp ot~er and connect-
present proce~ standards or'specifications would b'e'300 to ed in series opposing. When the probe with the two sensors
1,000 A instead of 60. How then could this specimen be is placed on the material, eddy currents ax:e generated in the
indicative of a malfunction in the system?" test object, inducing a secondary current in the two sensor
coils.
Shim Gages Since the coils are on the material at the same time and
Thin metal shims with artificial subsurface discontinuities since the induced current is the same for both coils, the
have been deSigned and widely used in Japan. Such low output for this series opposing circuit is, zero. During the
retentivity shims have circles or lines etched close to one application of magnetizing current, magnetic lines of flux
smface. When placed face down on the magnetized object, increase the current induced in the sensor coil, at right
the circles or lines are made visible by magnetic particles, angles to the direction of the magnetic field, thereby
indicating the direction of the field and some quantitative unbalancing the circuit.
value for the adequacy of the magnetization. The unbalance is registered on a digital or analog meter
By accurately controHing the depth of etching, different built into the instrument. The higher the' applied current,
sensitivities can be obtained. When the circular etching is the more the circuit is unbalanced and the greater the
used, the direction of the field can be determined by noting indication on the meter. Because magnetic properties vary
the sections of the circle that are indicated. with aHoy and hardness, the instrument only indicates the
Similar devices in the United States have artificial level of magnetization. The meter is generally divided into
discontinuities such as circles, lines and crosses precision ranges corresponding to insufficient, questionable and suf-
etched to an exact depth below the testing surface. Since ficient magnetization.
these shims are flexible, they can be placed in critical areas
such as fillets, near protrusions and other complex areas.

Measuring the Hall Effect

Quick Break Test
A meter using Hall effect probes is particularly useful in The dIect of a quick break was first observed in the
setting up testing procedures. Readings can be related 1940s. The only major design change since then involves t1~e
directly to artificial discontinuity standards. It must be introduction of a break in the secondary magnetizing circUlt.
 DETECTION AND EVALUATION OF MAGNETIC PARTICLE TEST INDICATIONS I 243

Utilizing only thp primmy break results in some self- updated, quick break systems, it was discovered that grind-
demagnetization and failure to observe circumferential ing cracks underneath a flash chrome was the cause of the
discontinuities. . problem. These were not detected with the equipment
An early result of this self-demagnetization was repetitive having only the primmy breaker.
failure of an automotive plant in 1946. At that time, all It was thought at the time that no failures of this sort
automotive manufacturers could sell every vehicle that was could occur in properly designed magnetic particle systems.
manufactured and a failed conveyor line at any time directly However, with the advent of electronic controls and trig-
reduced income. In the case in question, conveyor pins were gering, it was discovered that a malfunction in the firing
shearing at frequent intervals. These had been tested in the circuit could result in a loss of the quick break. It is essential
manufacturer's plant with magnetic particle techniques and that a periodic measurement with a quick break tester be
older equipment. By testing returned conveyor pins with required.
244 / MAGNETIC PARTICLE TESTING

REFERENCES

1. Skeie, K., D.J. Hagemaier. "Quantify Magnetic Particle Vol. 11. Metals Park, OH: American Society for Metals
Inspection." Materials Evaluation. Vol. 46, No.6. Co- (1976): p 44-74.
lumbus, OH: The AmeIican Society for Nondestructive 5. "Nondestmctive Testing: Magnetic Particle." Pro-
Testing (May 1988): p 779-785. grammed Instruction Handbook, fourth edition. Vol. PI-
4-3. San Diego, CA: General Dynamics Convair Divi-
sion (1977).
6. Ridder, Henry. Classroom Instruction Manual. "Vest
Hills, CA: Profe;ssional Engineering Services.
BIBLIOGRAPHY 7. Boiler and Pressure Vessel Code. Section V, Article 7.
New York, NY: American Society of Mechanical Engi-
1. Liquid Penetrant Tests. The Nondestructive Testing neers.
Handbook, second edition. Robert McMaster, ed. 8. Recommended fractice for Magnetic Particle Examina-
Vol. 2, Sec. 30-34. Columbus, OH: The AmeIican tion. ASTM E-709. Philadelphia, PA: American Soc~ety
Society for Nondestmctive Testing (1982). forTesting and Materials (1985).
2. Loeb, L. Fundamentals of Electricity and Magnetism. 9. Magnetic Particle Inspection. MIL-STD-1949. Wash-
New York, NY: Dover Publications. ington, DC: Department of Defense (1985).
3. Betz, C.E. Principles' of Magnetic Particle Testing. 10. Magnetic Particle Inspection Process. MIL-I-6868.
Chicago, IL: Magnaflux Corporation (1967). Washington, DC: Department of Defense. .
4. "Nondestructive Inspection and Quality,Control." M/lt- 11. Nondestructive Testing Requirements for. Metals. MIL-
als Handbook, eighth edition. Howard Boyer; ed. STD-271E. Washington, DC: Department of Defense.
 SECTION 10
PROCESS AUTOMATION OF
MAGNETIC PARTICLE TESTING
John Flaherty, Flare Technology, Elk Grove, Illinois
Charles Exton, Ar.drox Limited, Buckinghamshire, United Kingdom .'
 246 / MAGNETIC PARTICLE TESTING

PART J
TEST OBJECT HANDLING FOR MAGNETIC
PARTICLE TESTS

1. Safety: statistics indicate that a very large portion of

Manual Handling with industrial accidents happen during the movement of
Automated Testing matelials. Many of these accidents can be prevented
by the use of mechanical handling systems.
The simplest automation of the magnetic paIticle test 2. Space utilization: it is often possible to increase output
procedure occurs when test objects are manually positioned from the same amount of space by a mechanical
in the testing system and the magnetizing process then handling system. Such automation ensures the flow of
proceeds without operator intervention. objects from process to process and eliminates the
Consistency and reproducibility are achieved with this need for temporary storage spaces before and after
level of automation by: (1) providing accurate levels of each manufactUring process.
magnetizing current or field; (2) producing accurate timing 3. Quality control: accurate control of htmdling system
of magnetizing field or current flow; and (3) applying speed ensures that each test object receives the same
consistent volumes of magnetic bath in the same manner treatment. Counters can be installed to monitor sys-
and the same direction for each test object. This consistency tem speeds and this information may be stored on-site
of processing conSiderably improves both the quality and or delivered to a remote controlling center.
efficiency of the testing. To include this kind of automation
in a standard magnetic particle testing system is relatively
·simple and inexpensive, and can be achieved with electronic
cu~rent leveling devices and electronic timers. Designing Fu"y Automated
Magnetic Particle Testing Systems
Fully automated magnetic particle testing systems must
be specifically made for a particular product (or a limited
range of products) and are usually deSigned for a specific
Fu"y Automated Handling manufactming site. This is done so that the testing system is
compatible with existing manufactUring facilities. For exam-
Fully automated magnetic particle testing systems can be ple, the system for moving objects through manufacturing
cost justified only when large volumes of Similarly sized and operations may be an overhead monorail conveyor or a
Similarly shaped test objects are inspected. To move from powered roller track. The testing system must be deSigned
siml)le to complete automation requires the installation of as an integral pad of the site's existing handling facility. It is
mechanical handling systems that feed test objects into and equally important for the testing system to accommodate
out of the testing machine. Automatic scanning after mag- palticular production rates, depending on whether the plant
netization, classification and distribution after the testing uses a batch system or continuous flow line production.
process are very complex and costly procedures. Another impOltant design consideration is the size of the
The cost of handling in a manufacturing operation forms test object. If large, they may be attached to or loaded onto
a large part of the typical production cost: in the United the conveyor Singly and are presented in that way to the
States, the cost of handling varies from 15 to 85 percent. 1 testing system. Small objects are often loaded into baskets
More than any other single factor, materials handling offers or fixtures and conveyed through the plant in these contain-
a greater oppOltunity for cutting production costs and ers. On anival at the magnetic palticle testing system
increasing productivity. (which may have its own system for presenting test objects
When conSidering a fully automated testing system, the into magnetizing positions), the fixtures must be unloaded
reduction of labor costs is not necessarily the most impor- Singly and this requires a simple form of robot arm. Vel)'
tant factor. Other considerations are detailed below. small test objects (fasteners, for instance) may be singly
 PROCESS AUTOMATION OF MAGNETIC PARTICLE TESTING I 247

presented into the testing system by means of a vibrating process with a single program installed in a program-
... bowl feeder . mable logic controller.
The most impOltant considerations for successful design
of a fully automated magnetic particle testing system are: It is also important for the programmable logic controller
to have sufficient capacity for future expansion and the
1. reducing test object movement to a minimum; capability for linking with other on-site computers. This
2. mechanizing handling in a manner compatible with permits communication of test data to other systems and
the rest of the manufactUling process; and enables the magnetic particle testing system to be con-
3. controlling test object movement and the testing trolled by the plant's manufactming mainframe.
 248 I MAGNETIC PARTICLE TESTING

PART 2
MONITORING AUTOMATED TESTING
EQUIPMENT
Magnetic particle testing systems are designed to meet current is applied to the headstock and tailstock simulta-
specific testing requirements for detectability of discontinu- neously, with a second alternating current 120 degrees out
ity location, direction and size. Also included in design of phase with the first, applied to the coil. These currents
considerations for automated systems are such parameters are two phases of a three-phase electrical supply.
as test object symmetry, size, weight, configuration, mag- A multidirectional, contact and noncontact magnetization
netic properties and throughput requirements. technique provides a significant increase in the speed with
All of these factors influence the final design criteria for which magnetization can be established in a given test
the load station, conveyor system, test object positioning object. The implementation of electronic firing circuitry
fixtures, magnetizing method, magnetizil1g apparatus, bath enables phase switching' rise times on the order of 4 ms in
application, test object manipulation, visual or automatic a 0.5 s shot. Controlling the current level in each firing
scanning testing stations, demagnetization apparatus, off- circuit is another valuable characteristic. \Vith proper
loading and tested part segregation. multidirectional magnetization, time savings up to 70 per-
cent have been reported in the literature.
A test system design incorporating multidirectional mag-
Automated Magnetizing Techniques netization techniques must account for the follOwing con-
siderations: (1) the strength and direction of the magnetiz-
Establishing the proper magnetizing technique for a ing fields; (2) the sequencing of the magnetizing' shots;
specific test object is a fundamental requirement. Selection (3) the duration of the overall shot; and (4) the application
. of the type and level of magnetizing current may be dictated of the mag~etic bath and, in tum, the availability of
by the contrplling process sp,ecifications. Alternating cur- m.agnetic particles on the test object and their freedom to
rent, half-wave direct current and full-wave direct current, move and form indications.
Single-phase and three-phase, and Single-shot all have their , It takes considerable time and experimentation to ensure
place in automated test systems. that all these parameters are correct. If the magnetic field
In most manufacturing applications of magnetic particle strengths in different directions are not correctly adjusted
testing, test objects are sequentially processed with circular relative to each other, test indications may not be distinct
and longitudinal magnetization using coil and contact tech- and significant discontinuities can be missed. Field strength
niques. This has long been standard practice for manually adjustments may not necessarily be accurate if a pie gage or
operated equipment and semiautomated systems as well as other artificial discontinuity devices are used. A multidirec-
for automated tests. tional magnetic particle system should be set up using a
This is a reliable but slow magnetization method when gaussmeter and a reference standard (a typical test object
conSidering the requirements of automated testing line containing known discontinuities),
throughput .. For example, test objects magnetized in both
the circular and longitudinal methods require: (1) magne-
tizing in a circular direction, (2) testing and interpretation,
(3) demagnetizing, (4) magnetizing in a longitudinal direc- Computer Components of
tion and (5) a second testing and inteI1Jretation. Automated Testing Systems
Multidirectional Magnetization
An automated magnetic particle testing system may be
When automatically processing large numbers of similar managed by a programmable logic controller consisting of
test objects (automotive forgings or castings, for example), the following components: (1) central processing unit
there is a critical need for performing magnetic particle tests (CPU), (2) programmable memory, (3) input cards, (4) out-
at the highest reliable speed. This is done by using a system put cards and (5) power supplies for the internal electronics
that indicates discontinuities in all directions with one and the outputs to be controlled. ,
magnetizing operation. The CPU processes instmctions stored in the computer s
Several systems for this kind of testing are commercially memory. Among other fimctions, the central processing unit
available, including alternating current swinging field sys- uses the system software for the follOWing pUlposes: (1) to
tems. In a typical alternating current swinging field system, interrogate the status of the inputs; (2) to set or reset the
 PROCESS AUTOMATION OF MAGNETIC PARTICLE TESTING / 249

outputs; (3) to perform counting and timing functions; and current leveling capability to ensure that, when the current
(4) to continuously monitor the system for malfunction. output control has been set, the precise amount of current
This kind of controller provides much flexibility for many is supplied at each shot.
of the magnetic pmticle testing or manufactming proce- vVithout such a device, the variations in output from one
dures. Magnetizing current levels, shot duration and mag- shot to another can be enough to set off a soft alarm. In
netic bath application parameters can be easily changed and addition, when the output current is kept consistent, a drop
monitored. Test object identity and programming require- in current may be attIibuted to conditions external to the
ments can be fed in from an external computer and data on power unit (bad contact with the test object or loose cable
inspected parts may be filed as the test objects progress connections, for example).
through the system.

Magnetizing Current Monitors

Monitoring Automated Test Systems Current passing through the test object or the coil must
for Malfunction be monitored. Current levels at 90 to 100 percent of a
preselected value should be acceptable in most testing
\\lith a totally automated magnetic particle testing system, situations. Current values at 80 to 90 percent should cause
it is clitical to verifY correct performance of the system. For a soft alarm. Below 80 percent, a hard alarm should sound
instance, if the magnetizing current has not reached its full and the system is shut down. These threshold pOints may be
value because of bad contact, or if the magnetic bath adjusted to suit particular applications.
concentration is low or high, test objects move through the MonitoIing current levels for multidirectional magnetiza-
system but :discontinuities are not reliably indicated. Subse- tion is more complex because the current is only flowing in
quent intelpretation, manual or automatic, will consider this anyone direction for very sholt periods of time. It is
lack o{ indications to. me'an freedom from discontinuities. probaBly more efficient to integrate the current levels in
It is essential therefore to monitor all the parameters of each direction and average them over the duration of the
the magnetizing pro,cess and to monitor with equivalent current shot. Monitoring of the duration of the shot itself is
care the transducers and other devices introduced into the also necessary but this can be a straightfonvard function of
testing system to detect malfunction. the programmable controller.

Test System Malfunction Alarms

Magnetic Particle Bath Monitoring
Malfunction systems are normally connected to alarms
and two levels of alarm should be provided. A soft alarm The concentration of magnetic particles in the bath is a
sounds an audible tone, warning of a malfunction requiring very important system characteristic yet few devices have
operator intervention. The operator interrogates the pro- been deSigned for this monitoring function. One commer-
grammable controller to determine the nature of the mal- cially available model continuously monitors the opacity of a
function. In a typical soft alarm malfunction, the concentra- magnetic bath. As the bath flows through the monitor, a
tion of the magnetic bath may have moved beyond preset light source and photocell translate the concentration of
limits or the magnetizing current may have dropped within suspended particles into a signal which in tum activates
ten percent of a preset value. Based on established proce- alarms when the concentration is outside preset tolerances.
dure, the operator then decides: (1) to accept the test The system also sounds an alarm if the bath becomes
objects in question; (2) to reprocess the test objects; or contaminated. These are all soft alarms and in each case the
(3) to stop the system. operator takes a sample of the bath and performs a settling
A hard alarm sounds a different audible tone and imme- test.
diately shuts down the system. A hard alarm is sounded if a The programmable controller should also monitor the
test object is jammed in the clamping mechanism, if the magnetic bath application time, the magnetic bath pump
Lath pump fails, or if any part of the hardware or software pressure and the flow rate. This ensures conSistency in the
fails. application of the bath and helps provide consistent, repro-
ducible test results.
Current Leveling Monitor All of these monitoring steps - current level of the
system, current level in the test object, bath concentration
It is necessary to monitor a test system's current leveling and dwell time - are important steps for verifying the
device to ensure that it is operating correctly. The magne- operation of an automated testing system and, in tum, for
tizing power unit in an automated testing system includes ensuring the reliability of the system's test results.
250 / MAGNETIC PARTICLE TESTING

PART 3
AUTOMATING THE MAGNETIC PARTICLE
TESTING PROCEDURE
In today's industrial environment, interest in automation is used to describe a system with some combination of item I
for production operations, from basic fabrication to final 1,2 and 3, but which still requires human intervention (sud
assembly, is commonplace. The interest has generated a system is not fully automatic). The term (lutonuzti( I
extensive research efforts and as a result of these efforts scanning refers to the processes in items 4 and 5 above.
there are many claims, promises and hopes for lower costs
and better quality through automation.
The claims for lower cost are based largely on the
reduction of direct labor due to anticipated yield improve- History of Automatic Scanning
ments and better quality. Better quality comes about be-
cause of the uniformity of products being produced. This Visible Light Scanning
uniformity is based on the consistency that is possible when
fully automatic machines are used in manufacturing and One of the first recorded attempts at the full automation
because of the removal of human judgment from the testing of magneti? particle testing was in 1956. 2 The product being
function. While there is ample proof that lower cost and tested was a welded steel tube commonly c~lled electric
better quality do occur when automation is used in produc- resistance welded (ERW) pipe. Prior to this time, a number
tion, some questions persist about the benefits of automat- of partially. automated magnetic particle testing systems
ing the testing function. , were operational, and these performed as desclibed below.
One important. comparison is the cost of equipment The pipe was formed from a continuous sheet of steel
needed to perform automatic tests versus the anticipated bent into a tubular shape. The seam was continuously
savings from reduced labor, reduced scrap ·.a,nd reduced ,welded. The product passed through' a large yoke electro-
testing e~penses. Perhaps the greatest savings occur by magnet with the weld facing upward and dry black magnetic
autonlating only a portion of the testing process. A feasibility particles were dusted onto the weld zone (Fig. 1). As the
study should compare savings with e~penses for each stage dusted area moved into the magnetic region between the
of the testing process. Such a break-down can help deter- poles of the yoke, the magnetic particle material was
mine which individual activities could best be automated. attracted to leakage fields developed around weld zone
Considered below are some of the. factors that apply to discontinuities. A weak air flow removed excess particles,
the full automation of the magnetic particle nondestructive leaving only the magnetic particles over leakage fields in the
test method. Details are given for the history of automatic weld zone. At this point in the test, an operator visually
scanning; various factors that enter into automation of inspected the test object and determined if discontinuities
magnetic particle systems; specific methods of process were present.
verification; methods of scanning; image processing; and To automate the system, an optical scanner was devel-
some examples of systems using various forms. of automa- oped to detect the presence of black magnetic powder.
tion. Detection of an indication by the phototuhe higgered an
Automation can be used in magnetic particle testing by automatic marking device to identify the discontinuoHs area
celtain combinations of the follOWing processes: (Fig. 2).
For reliable discontinuity detection by the phototllhe, it
1. automatic handling and positioning of the test object; proved necesscuy to paint the top smface of the pipe white
2. automatic test object magnetization; to increase the contrast of the black powder indications.
3. automatic timing of the bath application relative to the This additional operation made the cost of automation
timing of the magnetization cycle; prohibitive.
4. automatic detection and location of magnetic particle
indications on the test object; and Ultraviolet Light Scanner
5. automatic inte11netation (deciding if indications are
caused by discontinuities). Before 1960, an improvement was made to the visihle
light scanner, eliminating the need for painting the top
If all of these processes are used, then the process is said smface of the pipe. 2 A wet fluorescent method was used fc)r
to exhibit full automation. The term automatic system often the test instead of the dly method. The surface was
 PROCESS AUTOMATION OF MAGNETIC PARTICLE TESTING I 25 J

FIGURE 1. Magnetic particle system for tests of illuminated with ultraviolet light and the resulting fluores-
electric resistance weld pipe
cent indications were scanned by a phototube. As in the
earlier application, an automatic marking device painted the
area of interest when biggered by a discontinuity.
A new problem occurred with this system: it was neces-
smy to use direct current in the ultraviolet lights. This kept
the power line fi-equency noise to a minimum and aBowed
reliable detection of indications. The use of direct current
shortened the life of the high intensity ultraviolet lamps to a
point that again made the system impractical and expensive.

Dye and Paint System

In 1959, another approach was developed to aid the
detection of magnetic particle indications on steel billets
and other products that required surface conditioning to
generate a high quality product. 2
In this approach, a special chemical was added to the
pigment coveIing the magnetic particles_ Normally the
pigment is fluorescent and is the medium that absorbs
ultraviolet energy and emits visible light. In the new form,
the pigment also contained a bright visible red dye that was
released by applying a suitable solvent. The solvent \\Us. in
'tum' mixed with a white paint. When the test object surface
containing retained magnetic particle material was sprayed
with this mixture, the solvent released the dye from the
pigment and the dye bled into the white paint, highlighting
discontinuities (Fig. 3). .
The intention of this approach was to aid the inspectors
and also to have a permanent indication on the billet
surface, allOwing further testing at a convenient location and
time.

FIGURE 2. Automatic scanner for magnetic

particle tests of electric resistance weld pipe FIGURE 3. Results of paint/dye magnetic particle
testing
252 I MAGNETIC PARTICLE TESTING

The cost of the operation was high, mainly because of the same results would occur by automating testing and quality
large amount of paint used. A few years later, this system control functions.
was improved by coupling to a television scanner that Machines can usuallv be calibrated more accurately than
located and detected fluorescent indications. 3 Detection of human senses and si~ple go/no-go interpretations ~an be
the fluorescence activated a system to spray the white paint uniformly made by automated systems. Machines are HSU-
and solvent mixture only over suspect areas. If the discon- ally more consistent, obtaining reliable and reproducible
tinuity was present, the red dye bled into the white paint test results. In addition to these advantages, automated
verifYing its presence and location so that the grinder testing systems can offer testing speeds for certain applica-
operators could remove it. This operation could be econom- tions that no human inspector can match.
ically justified. Beyond this point of comparison, care is needed! As
automation is applied and human involvement is removed
Laser Scanning from any operation, the requirements for consistency at all
The first laser scanning system for magnetic particle levels of processing become significant.
testing4 was installed in 1972 (see Figs. 4 and 5). This This can be illustrated by examining in more detail those
application was for testing torsion bars in automobiles. The routine procedures (cleaning, magnetization, particle bath-
discontinuities were seam inclusions and the system sepa- ing, observation and interpretation of the indications) used
rated those bars that contained no seams from those that in the manual magnetic particle testing method ..5 In the
contained seams. Beam shaping was used for pattern rec- manual method an inspeCtor performs these functions each
ognition (described later). time a test object is handled. For study, the functions may
be divided into three general groups.

Production Automation of the Processing Stage

Magnetic Particle Method In a typical manual system, the testing process usually
proceeds after the test object is rinsed or cleaned in some
Improved quality and lower cost have occurred- by auto- way, so that the surface is clear and free of debris, oil or
mating the production-sequence and it seems likely that the foreign matter.

FIGURE 4. Laser scanning in a magnetic particle FIGURE S. Laser scanning test system loading
testing system bars
"""""""""."",."",,"'1111111"""""'"""""""""""""",,"""""""1
PROCESS AUTOMATION OF MAGNETIC PARTICLE TESTING / 253

The test object is then magnetized. The operator can use require less magnetization time. They also have a less
a number of methods to determine if the magnetizing level critical bath stad time because of fewer problems with bath
is correct. These include: the monitoring of the magnetiza- washing of developed discontinuity indications.
tion current level with an ammeter;5 the use of Hall effect
devices to monitor surface field levels;6 flux shunting de- Observation Stage
vices;7 use of a reference standard containing known
The operator rotates, twists Of othenvise manipulates the
discontinuities; visual observation of the background level;
test object through three dimensions in order to observe all
or the use of specialized eddy current equipment to mea-
the cIitical surface areas. Magnetic pcuticle indications
sure internal magnetization levels. 8
formed duIing the processing stages are located at this
Magnetic pcuticle bath is applied after the operator
stage.
determines that the magnetizing level is appropriate. While
applying the hath, the operator carefully observes the
proceclul"e, veIifying that all of the surfaces of interest are
. Interpretation Stage
covered with magnetic material. The required skill level for The operator studies the indications and determines if
this process vaIies with the shape of the test object, its discontinuities are present. This is done with the benefit of
magnetic characteristics and the material it is made of. training and expeIience, making judgments based on the
If the wet testing method is used, the application effort is shape and bIightness of the retained magnetic padicle
less critical than with the dry method, providing that the indications.
bath is applied during the correct magnetization time Each of these three fundamental stages may be automat-
window. The time \\lindow is determined by the magnetic ed. Details of those procedures are covered in the text that
retenti\lity of the test object. Higher retentivity materials follows.
 254 I MAGNETIC PARTICLE TESTING

PART 4
MONITORING THE PROCESSING STAGE
OF MAGNETIC PARTICLE TESTS

Full automation comprises more than automatic scan- The cleaning solution should be also monitored to deter-
ning. The term automatic scanning usually refers only to the mine if the bath is viable or contaminated and at proper
automation of those actions performed in the observation concentration.
and interpretation stages discussed earlier. When the mag- The cleaning of magnetic particle test specimens is
netic particle testing process is fully automated, all three usually not a very critical matter. Unlike penetrant testi.ng,
stages are performed automatically; there is no operator or magnetic particle tests are generally very forgiving of
inspector. Processing, observation and interpretation are nonmagnetic dirt. For example, it is possible to have
performed and monitored by automatic apparatus. nonfenitic foreign material contained in surface cracks and
Partial automation is achieved when certain operations still do an excellent job of detecting these discontinuities
are automated. For example, it is common to use automatic with magnetic particle techniques. Good judgment and cost
cleaning and bath application methods tied to an automatic considerations determine the degree of cleaning process
magnetization cycle (processing stage) and still use inspec- monitoring that is justified.
tors to observe and interpret the magnetic particle indica-
tions. 9
. It is most important to remember that a good inspector
verifies that each stage of the testing procedure is properly Verification of Magnetization
performed. If full automation is used, automatic verifitation
. is required and maintenance of such components becomes A number of approaches can be used for this procedure.
a critical procedure. At production speeds, inoperative veri- . Current flOwing during the correct magnetization window is
fication equipment can invalidate large numbers of mag- a good indication that the test object is being magnetized
netic particle tests. properly:5 Further assurance occurs when the monitored
There is a variety of ways to verify proper performance of current value falls within predetermined threshold levels.
processing stages. The text below is a summary of these Monitoring of the current time and value is straightfolVll:ml.
important procedures. 10 Flux shunting devices are commercially available for deter-
The first decision in this and all verification processes is mining proper magnetization levels 7 and such devices can
determining the need for verification. For example, if be used in a manner similar to reference standards contain-
cleaning is critical to the accurate testing of a particular ing minimum rejectable discontinuities.
product, then the cleaning procedure must be verified. It is also possible to measure the normal and tangential
The next decision is determining the characteristics of components of the surface magnetic flux density with a Hall
interest (in this case, those that affect magnetic particle test effect device. 5 This measurement is then used as a relative
results) and how they can be quantified. indicator of intemal flux density for test objects 'With simple
shapes. It is best to measure this value dming the magne-
tization time window and use it to activate predetermined
Verification of Cleaning Operations threshold limits.
Another approach is to use an eddy current device to
If a spray cleaning system is used, a flowmeter may be measure the change in relative permeability of the material
lIsed to monitor the cleaning bath as it moves through the being magnetized (Fig. 6). The optimum value of perme-
pipes leading to the spray head. The flowmeter should be ability is that value at the knee of the initial magnetization
placed as close to the spray head as possible and the signal curve. 8 It is possible to monitor this point with eddy current
generated by the flowmeter should be compared to the techniques. As with the other methods, predetermined
control signal used to activate the flow valve. Some tim- values should be used, proper threshold levels should be set
ing differences occur but this is a good way to verify that and subsequent measurements should be taken during the
valves are working and that the flow is at proper magnitude. magnetization time window.
 PROCESS AUTOMATION OF MAGNETIC PARTICLE TESTING 1255

FIGURE 6. Eddy current response versus FIGURE 7. Typical m~gnetic particle bath f'ow
magnetization curves for two sample test bars versus magnetization time
13 x 25 x 150 mm (0.5 x 1 x 6 in.J with
electrodischarge machined notches at
0.6 pm 10.015 in.J depth BATH FLOW

UJ

tz
o
t ~
EDDY CUf(RENT
INSTRUMENT
0...
800 to 80
a,:
I(EADING

>- a,:
~ o
g
I-
w I

«
UJ
400 :?
I-
40
~ Z
a,: UJ
UJ a,:
3
I
0...
o -----r----t-----t--.L----f'>-l----
2
o U 0L--1---+--;---~~--~--;--==
TIME
o>- 1'600
1201
3.200
1401
4.900
160 1 (seconds)
o
UJ
GOOD INDICATION
FAINT INDICATION

FIELD STRENGTH
ampere per meter (oersted)

LEGEND
.0 ANNEALED Bath Application for High Retentivity Test Objects
64 RH ,

A 6 SPHEROIDIZED
If the test object has high magnetic retentivity or if the
91 RB surface is rough, bath application is not as critical. If proper
magnetization is used in a high retentivity test object, tl~e
magnetic leakage field around the discontinuity continues to
exist and attract particles after the magnetizing current is
Monitoring the Magnetic shut off. In this case, discontinuity indications develop even
if bath flow continues after the end of the magnetization
Particle Bath cycle .
. Multidirectional applied fields are not retained in test
Bath Application for Low Retentivity Parts
objects.5 (only one of the directions remains after the
Bath application is a vel)' clUcial parameter for test current is stopped). The usual compromise is to end
objects ,vith low magnetic retentivity. If such a test object magnetization with a field orientation in the direction
has a smooth surface and bath continues to flow over covering the most critical discontinuity Olientation.
developed indications after the end of the magnetization If alternating current magnetization is used, the current
cycle, it is probable that the developed magnetic particle should b~ cut off at the same point In the cycle for evel)' test
indications "ill he washed off. object to ensure consistency of magnetization. For maxi-
In this situation, it is clUcial that the timing of the mum retained field, the current should be cut off at zero or
magnetization cycle be carefully considered when determin- when it is decreasing from a maximum value (see Fig. 8).
ing the bath procedure:s Flow conditions should be checked vVith modern electronic components, it is most convenient
and cycles should be carefully adjusted to ensure that all to cut the current at zero. If coils or yokes are used for
bath flow has stopped in clitical areas of the test object alternating current magnetization, it is vel)' easy to demag-
before the magnetization cycle ends. Once these parameters netize the test object as it moves through the tapering field
have been determined, then both the magnetization and at the exit end of the coil or yoke:s As in the case of low
bath cycles must be monitored in time and compared with retentivity materials, little or no retained field then exists
the predetermined values. Appropriate threshold levels are around the discontinuity and washing of the indications
set to detect deviations. must be considered. However, if the test object's surface is
Flow can be monitored with a flowmeter (Fig. 7) located rough, embedded particles may resist the washing effect,
as closely as possible to the spray heads. Photocells and even if the test object is demagnetized. A sufficient number
acoustic path de"ices have also been used to measure flow of particles might remain in place to indicate the presence
rates. of discontinuities.
256 / MAGNETIC PARTICLE TESTING

Bath Coverage and Concentration

FIGURE 8. Flux density versus magnetizing
current; to ensure maximum retained field, It is vital for critical areas to be covered with magnetic
current should be shut off during the two parts particle bath. Positioning of the test object relative to the
of the waveform illustrated as solid lines; shutoff bath applicator is therefore a very important parameter.
must be at the same point for all test Objects to When an automatic loading system is used, it is possible for
ensure consistency the test object to shift and critical areas are not then exposed
to the bath. For this reason, test object positioning should
be monitored by means of limit switches, photocells or
+8 max acoustic path monitors.
Another critical bath parameter is the magnetic particle
concentration. In manual systems, this is monitored using a
centrifuge tube as part of a settling test. 6 It is possible to
monitor particle levels automatically with electromagnetic
methods, optical methods and particle counters.
H
A number of bath parameters may be checked at the
same time by using a standard test object containing a
known discontinuity. The reference standard is passed
through the entire test system in front of each part being
tested or in some frequent, repetitive manner. The standard
is cleaned, magnetized, bathed and tested automatically.
When the system is working correctly, the standard's
discontinuities are accurately detected and this verifies that
follOwing test objects are also being processed and inspected
prop~rly.
 PROCESS AUTOMATION OF MAGNETIC PARTICLE TESTING / 257

PART 5
AUTOMATING THE OBSERVATION STAGE
OF MAGNETIC PARTICLE TESTS

Methods of Scanning Flying Spot Scanning Systems

Scanning of fluorescent magnetic partiCle indications is System Operation and Arrangement
done in order to automate the viewing of test results. Auto-
matic scanners usually comprise the following components: The flying spot scanner uses a deep blue, violet or an
ultraviolet beam to illuminate the test object (Fig. 10). The
source of the beam is usually a laser. For example, a
1. a source of ultraviolet radiation (may be a commercial helium-cadmium laser operates in the visible mode at a
ultraviolet floodlight or a laser with a movable beam); deep blue wavelength of 441.6 nm (4,416 A). This wave-
2. a photodetector sensitive to visible light but not length will excite the pigments used in fluorescent magnetic
sensitive not to ultraviolet radiation (may be a photo- particle materials. Such a light source produces a vel)'
tube, a photomultiplier or a vidicon camera); narrow and intense beam.
3. an imaging apparatus (may be a television photode- The scanning occurs as the beam passes over the test
tector or a flying spot system); object and illuminates a vel)' small area at one time. the
4. amplification and discrimination equi.pment (electronic position of the beam is controlle6 ~y mirrors; by moving the
processing· networks witn the ability to determine light source, or by moving the test object in front of a
whether intensity and patterns of detected indications stational)' beam. The level of fluorescence gerrerated in the
are caused by discontinuities); and area covered by the beam is low unless the laser strikes a
5. material handling accessories. retained indication. When this occurs, a larger amount of
fluorescent (visible) radiation is emitted (reflected). The
fluorescence is detected by a single phototube equipped to
Matelial handling accessories are needed to move test
filte·r out blue or ultraviolet light from the illumination
object to and from the scanner; to index or move the test
source.
objects under the scanner so that all desired areas are
in~pected uniformly; and to mark or separate rejected test
The detection signal takes the form of pulses that are then
processed through a computer or other electronic equip-
objects.
These components are integral to the two basic types of ment. Signal processing is used to discriminate rejectable
scanning systems: the television scanner and the flying spot
scanner. II
FIGURE 9. Television camera with an image
Intensifier for automatic scanning of magnetic
particle test Indications
Television Scanning Systems
A television scanner first illuminates the test object with
filtered ultraviolet light. The camera is used as a detector of
visible indications (Fig. 9). It is equipped with an image
intensifier and a filter to remove superfluous ultraviolet
light. The output video signal is processed through a
computer or other electronic equipment designed to distin-
gUish noise from discontinuities.
After receiving a discontinuity signal, the computer is
often programmed to initiate some rejection procedure
such as marking the area with paint for later disposal, or
sorting appropriate test objects into a reject bin. Some
pattern recognition and classification of discontinuities are
possible with a television scanning system.
258 I MAGNETIC PARTICLE TESTING

FIGURE 10. Flying spot scanner for automated magnetic particle tests

PATTERN
RECOGNITION
ELECTRONICS

,----------
/
/
/ /
/ /
/ / THRESHOLD
GATE
/
/
/
/ FLUORESCENT
INDICATION

discontinuities from background noise and to initiate rejec- useful accessories for television scanning systems are inex-
tion action. pensive, eaSily available and versatile.

Comparison of Flying Spot to Television Systems

Application of the Flying
The flying spot scanner has several advantages over
televIsion systems. The laser provides much more intense Spot System
illumination than all ordinal)' ultraviolet light. The resulting A technique using a flying spot laser for the detection of
fluorescence is correspondingly intense and easily detected. fluorescent indications has been reported in the literature. 12
The detail discIimination is much better in a flying spot As shown in Fig. 10, the system consists of several functional
system. A phototube can handle a larger contrast ratio than components, including a scanning laser, a photodetector
a normal vidicon or the conventional television camera. and electronic pattern recognition circuits.
Problems with depth of field are minimal in flying spot The scanning laser causes a beam of intense light to move
methods because the system is not limited by optical across the test object. vVhen this beam stIikes fluorescent
imaging criteria. The depth of field is determined by the material held ove~' discontinuities on the test object's sur-
beam divergence not the {-number of the imaging lens. In bee, a pulse of light is generated. The emitted ligl{t does not
addition, flying spot systems require electronic circuits that have the same wavelength as the incident light. The
are much simpler than a television system. photodetector converts this fluorescent pulse into an eIec-
The plincipal advantage of television systems is their wide tticaJ signal that is processed to determine if it represents
dishibution and the resulting economies of volume. Many disconti nuities.
 PROCESS AUTOMATION OF MAGNETIC PARTICLE TESTING I 259

Pattern recognition techniques and signal intensity mea-

FIGURE 11. Line scanning procedure for magnetic
surements are used in the signal processing. The optical
pattern recognition system recognizes shape and size for particle tests
discontinuity determination and generally ignores back- STAIRCASE SYSTEM
ground fluorescence effects. GENERATOR CLOCK
The system has extensive depth of field. System resolu-
tion is determined from the scanning beam cross section
and not by the effective aperture of an imaging lens. High
densities of excitation energy are possible with a laser and
extremely sensitive photo detectors allow detection of weak
fluorescent indications.

System Characteristics 13
The scanner PQrtion of the system contains a helium-
cadmium (Heed) laser operating at a deep blue wavelength
of 441.6 nm (4,416 A). The pigments used in magnetic
particle testing materials absorb blue and ultraviolet excita-
tion and emit visible yellow light. A blue excitation source
may therefore be used instead of the familiar ultraviolet
light. The blue is filtered out of the viewed image for visual
interpretation and the ultraviolet light is invisible.
Photosensitive devices detect both wavelengths, so filters motion of the test object so that the entire surface is covered
are required when either wavelength is used for automatic by the scanning beam.
testing. The laser beam diameter is about 1 n1m (0.04 in.)
and the output power is nominally 15 mW. The laser berm .
Signal Detection and Processing
d\vergence is vely.small so that no optical components are
required. It is possible to use a lens or a combination of As the beam strikes retained fl~orescent material, it emits
lenses to reduce the beam diameter to a verv small value a pulse of yellow light. Some of this light strikes the face of
and thereby increase system. resolution. .I • the photocell and is converted int<? an electrical pulse Signal.
Figure 11 shows the arrangement of the system's compo- The amplitude of this pulse is directly related to the
nents. The laser beam is directed to a scanning mirror that intensity of the pulse of yellow light. The phototube pulses
causes the light to move back and forth across the test are amplified, filtered and, if they are to be used in a digital
object. This procedure effectively produces a line scan. The circuit, converted to a binalY value representing amplitude.
waveform used to drive the scanning mirror motor is a These pulses may be used to generate a raster similar to
staircase function derived from the system clock. Thus each a television image. The beam position information and the
position of the scanning mirror can be directly related to a phototube analog output are required to produce such an
specific clock pulse. The scan motion is at right angles to the image.
260 / MAGNETIC PARTICLE TESTING

PART 6
AUTOMATING THE INTERPRETATION
STAGE OF MAGNETIC PARTICLE TESTS

Recognition of significant test indications and distinguish-

ing them from background indications are generally the two FIGURE 12. Optical pattern recognition of
most complex problems for an automated testing system. straight, linear fluorescent test indications
Background often appears as a foggy glow containing a
number of isolated bright spots. Frequently, there are also
some false indications that share the characteristics of
discontinuities. These include tool marks, scratches, thread
crests, valleys, hole edges and other anomalies. PHOTOTUBE

A trained operator ignores nonrelevant indications and so

should an automated system. This discrimination capability c;9
dramatically increases system complexity, yet most auto-
mated testing applications do require some sort of pat-
tern recognition. This may range from a simple coincident
light technique to a digital computer programmed with
pattern recognition algOrithms. The most prevalent pattern
recognition procedures include: (1) optical techniques,
(2) dedicated digital processors, (3) microprocessor algo-
rithms, (4) SRI algOrithms and (5) neighborhood processing
algOrithms.

illumination could conc'eivaoly permit optical pattern rec-

Optical Pattern Recognition ognition for straight line indications in other orientations.

In optical pattern recognition, a linear laser beam is

positioned in the direction of a discontinuity.13 Figure 12 Pattern Recognition with Dedicated
illustrates the formation of this line of light using an Digital Processors
anamOlphic lens. An anamorphic lens is curved so that it
focuses the collimated laser light into a line parallel to the The system described below is a parallel processor de-
lens axis. The beams' cross sectional width in the orthogonal signed for use at high speeds. 14 It consists of multiple circuit
direction is maintained because there is no focUSing in that boards, each containing a number of logic functions
axis. (Fig. 13). The basic component of each board is a large
A phototube detects and integrates the fluorescent flash memory array with multiple bit elements. Each scan of the
that occurs when the exciting beam 'is coincident with the testing system is adjusted fcn zone width and the disconti-
indicated discontinuity. Good parallel alignment of the nuity input Signal is gated and fed into thc array. System
heam and the discontinuity are neceSSaIY and this require- clock pulses llsed to determine the scanning waveform are
ment in some ways limits system flexibility. also used as shift pulses. Discontinuity Signals shift in
A similar result can be obtained using a scanning laser synchronism with the scanning beam. Auxiliary registers
system and integrating the output of the photodetector arranged at the output of each array are grouped together to
across the scan. This has the advantage of a larger depth of form a two-dimensional array.
field (no optics are used for focusing) but requires more If a discontinuity indicati~n is straight, long and bright
complex scanning and electronics. enough to activate the input threshold, the input pulse \vill
Far less sensitivity results if the beam and the disconti- be exiting the last register in a 2,400 bit line after twelve
nuity are not parallel. Using a parallel line of exciting scans. Simultaneously, signals will be exiting all the other
radiation is considered a form of pattem recognition be- registers in 200 bit increments, beginning with the 200
cause optical integration of the fluorescent light occurs from· register. These Signals are coupled into the matIix array and
the entire length of the indication. Slits of excitation appear as a row of ones on the top. After eight clock pulses,
 PROCESS AUTOMATION OF MAGNETIC PARTICLE TESTING 1261

FIGURE 13. Block diagram of digital processor used for optical enhancement of magnetic particle test
indications

SHIFT PULSES

REMOVE 200 BIT

REVERSE MOS
INPUT SCAN SHIFT
SIGNAL REGISTER

'----~I '----~ I.i....-_ _...J I'~- MAIN MEMORY

I
DIRECTION OF
SCANNING
BEAM
I~
co
00

1 .J
-----,
PARALLEL TO TIME PRESETTABLE I
I SEOUENTIAL
CONVERTER CONTROL I OUTPUT

L ____________ J
STRAIGHT LINE DECODER

this row appears at the bottom and exits the system

(Fig. 13). Background fluorescence shows up as a random FIGURE 14. Scanner memory matrix signal
distribution of ones in the array. When a row is filled, the storage for a straight discontinuity (at bottomJ
probability is high that a seam or some other linear discon-
tinuity has caused it.
MOTION OF TEST OBJECT
Linear discontinuity conditions are detected with a num-
ber of decoders wired to specific registers in the matrix
array. It is shown in Fig. 13 for straight line decoding (the
decoding is \vired to detect straight lines only). These
parallel channel pulses are converted into a time sequence
I ',,~
I-
1 1 1 1 1 t-- ANGLED SEAM
ENTERING SYSTEM

and then registered on a counter. Any number of pulses may m ~~~+-~~~~-+~ BACKGROUND
00 ~~~+-~~~~-+~
be preset on the system so that an intermittent linear
discontinuity with four to eleven counts in a twelve count
\\lindow CaI~ activate the output gate. A decoder can be
wired to detect an angled linear discontinuity indication in a
similar manner (Fig. 14).
1:""""" I I-- STRAIGHT LINE
DISCONTINUITY
ABOUT TO LEAVE
SYSTEM
 262 I MAGNETIC PARTICLE TESTING

FIGURE 15. Scanner memory matrix signal FIGURE 16. Scanner memory matrix after an
storage for a discontinuity at an angle to the angled discontinuity has moved out of the array
line of test object motion (at bottom) and a new discontinuity parallel to
test object motion is in storage (near center)
z MOTION OF TEST OBJECT
o;::
u MOTION OF TEST OBJECT
:st

I
(5
z ~-+-+-+--+-+~4-~r+--1, V ANGLED SEAM
~ _ B/\CKGROUND
~ I' ,!-j
00 ~-+~~F4~4-~r+-1 ~-+~~~--+-+-+--~~
co ' '-1""" I ./' BACKGROUND ~-+-+-~+--i-+-+-+~ ,-+, __ NEW STRAIGHT SEAM

1~~ -+--~--t-+--f--+-~/
BACKGROUND

TOP OF ANGLED SEAM

Figures 14, 15 and 16 show memo!), matrix signal states

dming scanning of sequential linear discontinuity indica-
tions. In Fig. 14, the matrix signal pattern shows: (1) a FIGURE 17. Conversion of analog test signals to
horizontal indication about to exit at the bottom, (2) an digital input data
angled linear discontinuity entering above and (3) back-
ground. The direction of laser beam scanning across the
discontinuity is shown vertically. The direction of the test
object motion is shown hOrizontally.
figure 15 shows the matrix signal state after the horizon-
tal discontinuity has left and the angled discontinuity is in
. storage. Because all discontinuities eventually pass down
through the matrix, a given. decoder 'wired to detect this
lingled seam 'could be connected to the registers shown (or
one above or below) so long as the pattern is the same.
In Fig. 16, the angled discontinuity has nearly exited the
array and a new horizontal discontinuity is present in the
middle of the display system. Random background signals
are indicated by ones scattered throughout the matrix.

Simple Microprocessor Algorithm

spot scanner as well. It is required that the scanning format
A variety of ways are available for using a digital computer be -similar to the television format and that the video Signal
to detect ce}tain patterns common to discontinuities. The be delivered in binary or digital format. These are powerful
important idea is to differentiate discontinuities from back- approaches and are used in machine vision applications to
ground. This is usually done based on the fact that most aid in pattern recognition by reducing noise and clutter on
discontinuities are linear indications. the images.
The ability of humans to visually distinguish lillear incli- Figure 17 shows how analog Signals move from the optical
cations in high noise clutter is well known. A large variety of scanner and are processed in analog-to-digital convClters
algorithms are available for Simulating this human charac- and then coupled to a digital computer for analysis.
telistic. The follOwing example shows a simple linking
algorithm used to detect linear discontinuities in high
Adjacent Cell Linking
background levels.
Celtain computer based optical processing techniques Figure 18 shows the simplified logic diagram of a pattel11
may be used to enhance television images and it is possible recognition program that will find hOlizontal and vertical
to use these approaches with the image data fi'om a flying lines. 13 The computer program uses a procedure called
 PROCESS AUTOMATION OF MAGNETIC PARTICLE TESTING I 263

FIGURE 18. Logic diagram of a computer program for finding horizontal and vertical test indications

DISCONTINUITY
IS PRESENT
OUTPUT ....----.-.1'--------1 N LONG
AT LOCATION X

adjacent cell linking. In this instance, if test indications be reached to store a binarY value in memory. All values
(st~red numbers in memOl), locations that are an analog of below this level are recorded as zero. .
~he~r locations on the test object) can be linked to adjacent The scanning format is 12 X 200. Assuming a starting
mdIcations, then the probability is high that these are location of 1, the program first looks for the presence of data
caused by discontinuities. The greater the number of links, in location 1. If data are there, the program next looks in
the w~ater the discontinuity probability. location 1 + 1. If data are there, the program checks loca-
Il11tlally, the scanner data are stored in memOl), with tion 1 + 1 + 1 (or 3) and so forth until a preset number N is
ad~ressing referenced to the position of discontinuity indi- reached. Reaching N indicates a vertica1linear discontinuity
catIons on the test object. To keep matters simple, the data indication with length proportional to N and a starting point
are reduced to a single value. Thus, a certain threshold must at location 1.
264 I MAGNETIC PARTICLE TESTING

To detect horizontal linear discontinuity indications the

number 200 is added to X and, if proper discontinuity'data FIGURE 20. Use of neighborhood processing to
are present in this location, the counter moves to X + 400 enhance the image of a cylinder: (a) top view
and so forth. After N progressions with discontinuity data showing mirror held by gray block placed in the
recognized in each mahix cell, the computer reco~izes a cylinder to image the inner wall; (b) video
horizontal linear indication N units long starting at loca- image of inner wall and crack indication;
tion 1. After this, X is changed to 2 and the process repeats. (c) image before removing background; and
Analogous programs to recognize angled linear discon- (dJ image after processing
tinuity indications can be incorporated with additional
statements.
lal

The SRI Algorithms

A popular set of algOrithms used in pattern recognition
systems are those developed at the Standard Research
Institute (SRI).I.5 These algorithms consist of a series of
procedures that control about fifty different features ex-
tracted from a binm)' image. Some ~f the extracted features ,bl
are: (1) the size of a blank area, (2) the size of solid objects,
(3) the centroids of holes, (4) the centroids of blank areas,
(5) the centroids of solid objects and (6) the number of
these details registered. .

FIGURE 19. Typical arrangement and labeling of

neigh!:>oring pixels

leI

U Dl [J /
""
[]~ -0 [ PIXEL [

/ ~ --l ,dl
0
HORIZONTA~ 8 D 8
i=
a::
LlJ
>
 PROCESS AUTOMATION OF MAGNETIC PARTICLE TESTING I 265

FIGURE 21. Use of neighborhood processing to enhance the image of a tube section, starting with the
image before processing, progressing through several enhancement techniques to remove background,
and ending with the final image of the discontinuity

fa) fb,

. Ie, 'Id,

fe'
266 / MAGNETIC PARTICLE TESTING

An image is analyzed based on these features and a Cellular automata theory was first developed by John von
description of the object or image is stored and used for Neuman who is primarily known for developing the algo-
later comparison with images of unknown objects. The rithms that drive the modem digital computer. Neuman's
purpose of the analysis is to reduce complex images to interest was in Simulating in a machine some of the aspects
essential values (a kind of categorizing), so that images may of self-reproduction in biological cells.
be selected and compared based on the details of interest. As in all electronically processed images, the stalting
There are three significant points to remember about the point of neighborhood processing is a large rectangular area
SRI algorithms. First, algorithms usually work with a binary of picture cells or pixels, in which the rows and columns are
image. Therefore all data below the threshold value are arranged in a standard television format. Typically each pixel
defined as zero and all values above the threshold are has eight neighbors as shown in Fig. 19.
defined as one. Second, the resulting binalY image is In the normal analog television image, each pixel has a
analyzed in terms of certain predetermined characteristics brightness level that corresponds to some shade of gray
(number, size, centroids) of certain predetermined features (typically 256 levels or 8 bits). In digital imaging, the pixel
(holes, lines and so on). Finally, the description is stored for has a binary number that corresponds to the specific gray
future reference for comparison with unknown images. level. If the computer program word representing a pixel
The chief limitation of SRI algorithms is the significant has enough memory capability, additional memory can be
amount of data discarded in the generation of the binary used to store other information about that pixel's portion of
image. Also, the severity of the thresholding makes such the image. For example, the most significant bit can be
systems difficult to set up and not very repeatable. It is not made to indicate that its pixel is part of a discontinuity. A
easy to digitally manipulate the large amount of video data field of all these significant bits forms wliat is called the bit
present in a typical television image. One way to accomplish plane of a discontinuity and can be used to indicate the
this is to reduce the number of gray levels from a typical image's discontinuity condition. .
value of 256 to a value of 8. Even eight is too large a number ~n cellular automata, the data in all the pixels are changed
to effectively accommodate the SRI algorithms .. according to a transition rule and according to the state of
Because of these limitations, other techniques have been . the pixel neighbors (all data change sequentially).
developed to expand the number of the gray levels involved There are four key stages in neighborhood processing.
in the data processing. First, the. pixel word acquires gray levels in the least
significant bits and other information in more significant
bits. Second, a series of successive transformations occurs in
Neighborhood Processing a c~refully determined sequence. Third, when the final
transformation has occurred, the important information is
Neighborhood processing is also known as cellular autom- located in the most significant bit; irrelevant data are
ata and rrwrphological image processing. 16 The concept is eliminated. Finally, the image is scanned and specific
particularly suitable for industrial nondestmctive testing features are studied by examining certain significant bits.
purposes. The advantage of the neighborhood processing technique
In this image processing technique, each pixel in the is that it removes background so well (Figs. 20 and 21).
image is transformed or converted in a particular way that After such improvement, it is relatively easy to use spatial
primarily depends on the state of the adjacent or nearest gating of a video image system to trigger alarms or other
neighbor pixel. warning devices.
 PROCESS AUTOMATION OF MAGNETIC PARTICLE TESTING I 267

PART 7
EXAMPLES OF PARTIAL AND FULLY
AUTOMATED MAGNETIC PARTICLE
TESTING SYSTEMS
The magnetic particle system illustrated in Fig. 22 is for Figure 24 illustrates a system using multidirectional
billet testing. The system uses a specially developed mag- magnetization for processing. Test objects are loaded onto
netic. paIticle material that is permanently bonded to the holders and moved into the magnetization area where a
surface of a billet when heated. This allows manual testing
to be performed at a later, more convenient time. The
processing stages of the test are automated, including FIGURE 23. Piston dome processing system for
multidirectional magnetization, test object bathing and ap- magnetic particle tests
plication of heat to fix the magnetic particle indications.
Figure 23 illustrates a system designed to automatically
process piston domes using longitudinal coil magnetization
at speeds exceeding 200 test objects per hour. The test
objects are automatically moved to a pickup point, clamped
securely by jaws to a special turnover mechanism, then
lifted in an arc and positioned inside the coil. Bath is
auto~natical1y applied to criti{:al surface areas, coil current is
turned' on, the bath is shut off and the coil current is then
shut off.
Next the test object is returned, dome side down, to the
pickup point, with the escapement cycling to push the
processed test object onto the exit conveyer, at the same
time bringing a new test object into the system for process-
ing. Inspected parts are moved to a testing area where they
are visually examined.

FIGURE 22. Fixed indication billet testing system

FIGURE 24. Magnetic particle testing unit used

for bolt inspection
268 I MAGNETIC PARTICLE TESTING

FIGURE 25. Coupling testing unit using a central FIGURE 26. Magnetic particle unit for testing
conductor for circular magnetization bars using residual magnetization (see also
Figures 4 and 5)

sequential coil (longitudinal magnetization) and head (cir-

cular magnetization) are applied. The bath flows on to the -Residual magnetization is used in the 'system shown in
te!rl: object during -the magnetization cycle and is cut· ~ff Fig. 26. This is a fully automatic system. Parts are circubrly'
before the magnetization cycle ends. The test object is ~hen magnetized two at a time then moved by means of a walking
moved to the testing area where it is visually inspected. beam conveyer into the testing chamber. There they are
Figure 25 shows a system that uses a central conductor to rotated two at a time and moved laterally under laser
circularly magnetize test objects. Bath is applied dming the scanning heads. The beams are shaped optically into lines
magnetization cycle and test objects are viewed immediately and after scanning, good product moves to the next manu-
after bathing. The operator can manually rotate the test facturing station. Material with discontinuities is automati-
object for better observation if necessary. cally removed from the production cycle.
 PROCESS AUTOMATION OF MAGNETIC PARTICLE TESTING I 269

REFERENCES
1. llJateriaZ Handling in Industry. The Anglo American 10. Foley, Eugene. "NDT: A Real Profit Center." Ma-
Productivity Team. London and New York: The Anglo chine and Tool Blue Book. Chicago, IL: Hitchcock
American Council on Productivity (1950) Publishing Company (December 1984): p 58.
2. Shroeder, Kenneth. Private correspondence. Chicago, 11. "Mechanized Scanning of Fluorescent Penetrant In-
IL (May 1987). dications." The Nondestructive Testing Handbook: Liq-
3. Van Kirk, Keith, et a1. Method of Detecting Inhomo- uid Penetrant Tests, se~ond edition. Vol. 2. Columbus,
geneities by the Use of Mixtures of Fluorescent and OH: The American Society for Nondestructive Testing
Visible Dye Colored Magnetic Particles. US Patent (1982): p 211.
3,609,532 (September 1971). 12. "New Laser System for Scanning Metal Defects Read-
4. O'Connor, D., et al. Defect Detecting and Indicating ied for Market." Metalworking News. New York, NY:
Means for Non-Destructive Testing. US Patent Fairchild Publications (May 1973): p 18.
3,774,030 (November 1973). 13. Flaherty, John, et a1. Laser Scan Testing System
5. Betz, C.E. Principles of Magnetic Particle Testing. Having Pattern Recognition Means. US Patent
Chicago, IL: Magnaflux Corporation (1967). 3,774,162 (1973).
6. Dry Powder Magnetic Particle Inspection. E-709-80. 14. Flaherty, J. and E. Strauts. "Alltomatic Scanning of
Philadelphia, PA: American Society for Testing and Fluorescent Indications." Proceedings of the ASM
Materials (1980). Metals Show. Metals Park, OH: American Society for
7. Proceedings of the Third Intematjonal Conference of Metal~ (October 1971)....
Nondestructive Testing. Tokyo, Japan: Japanese Soci- 15. Negin, M. and N. Zuech. "Review of Vision System
ety for Nondestructive Testing (March 1960). Techniques for Inspection of Electronic Components."
8. Lorenzi, D. Establishing Reliable Magnetization Lev- Proceedings of EIA Electron .Device Conference.
els for NDT Inspection. Columbus, OR: The American Washington, DC: Electronics Industries Association
Society for Nondestructive Testing (Spling 1981). (May 1985).
9. Flahelty, J. "Automation and NDT." ManufactUring 16. Wilson, S., Q. Holmes and T. Limperis. "New Ma-
Systems. Chicago, IL: Hitchcock Publishing Company chine Vision System Called PIXIE." Proceedings of the
(August 1985). Control Exposition (May 1980).
 SECTION 11
RECORDING OF MAGNETIC
PARTICLE TEST INDICATIONS·
J. Thomas Schmidt, NOT consultant, Arlington Heights, Illinois
 272 / MAGNETIC PARTICLE TESTING

PART J
BASIC RECORDING OF MAGNETIC
PARTICLE TEST RESULTS
Discontinuity indications fonned by the magnetic particle
test method may be very visible on the surface of the test Drawings and Written Descriptions
object at the time of the test but they are seldom permanent
or even durable. In many cases, it is desirable to make a The simplest method of recording test indications is the
permanent record of these indications. Sometimes, the written deSCription. To be practical and meaningful, this
record is used to document and justifY the return of a written record must be detailed and descriptive. At the very
rejected palt and in other cases the recorded indication is least, it should include the sizes and locations of test
used to prove that a discontinuity is small enough to be indications. Additional information should be supplied as
insignificant. warranted by the application.
Indications may be recorded in two basic ways: (1) on the Drawings are often used in place of or as complements to
test object itself (often called fixing); or (2) on other media written deSCriptions. Drawings can provide more informa-
for storage remote from the test object. The text that follows tion than written descriptions or they can be confusing,
includes data on fixing methods, plus details on recording depending on the skill and dedication of the illustrator. If
with media separate from the test ·object, including: draw- large numbers of similar test objects must be recorded by
ings and written descriptions, pressure sensitive tape trans- drawing, a master diagram can be prepared and copied.
fers, alginate impressions, magne~ic rubber replicas and Indication records are then added to the copies by the
photography. magnetic particle inspector.
 RECORDING OF MAGNETIC PARTICLE TEST INDICATIONS 1273

PART 2
- PRESSURE SENSITIVE TAPE TRANSFERS
One of the oldest and Simplest mechanical methods of making the transfer. Wet surfaces prevent the tape from
recording magnetic particle indications is the tape transfer. sticking properly and may dissolve the tape or adhesive.
This technique Hses a piece of pressure sensitive tape
pressed over the indication then lifted off and pressed onto
Water Vehicle Tape Transfers
s a piece of paper that is kept as the test record. It is usually
necessmy to include some additional descriptive matter With water vehicle tests, the drying procedure can be
the coveling the location and orientation of the discontinuity on fairly Simple. \Vater dries quickly in a typical testing envi-
this the test ohject. . ronment, leaving the test object surface in proper condition
ery In order to make a good tape transfer, the test surface for a tape' transfer.
est must he dean hefore the tape is pOSitioned. Dust and dirt If there is not sufficient time for the water vehicle to
as deactivate the adhesive if present in excessive amounts, as evaporate, or if the particle background is too dense, the test
do water and oil. It is often necessmy to clean the surface object can be gently rinsed with acetone. Acetone dissolves
to before the magnetic pmticle test is performed and again water and dries very quickly. It also linses away some
a- hefore the indication transfer is made. surface palticles, performing the function of the dry method
g, air flow. However, caution must be exercised to' avoid
If washing away indications.
'y
I. Dry Method Tape Transfers Oil Vehicle Tape Transfers
• With oil suspensions, the drying problem is more diffi-
The tape transfer recordh~g method' is most easily used
cult. Oil drains slower than water and dries much slower.
with dry magnetic palticle tests. Large indications are
With some velY volatile oils, it may be possible to wait until
usually produced with dly techniques' and there is. no
the vehicle evaporates, but it is usually necessary to remove
moisture to resist the tape adhesive. .
excess oil bath with a volatile solvent (petroleum ether,
Excess magnetic powder is removed, usually by a gentle
hexane, naphtha or 1,1,1 trichloroethane).
air flow, before the transfer is made. Once the indication is
Rinsing with solvents must be done very carefully to avoid
fon~1ed and the excess powder removed, a piece of tape
loss of the indications and the solvent must be completely
.. sufficiently long to cover the desired area is pressed firmly
dry before the tape is placed.
over the indication. The tape is then lifted off the test object
Caution must be observed when using these solvents
and pressed onto the paper backing that is retained as the
record. since some are very flammable. If possible, it is best to
transport the processed test objects to a fume hood to
If fluorescent particles are used, dark colored paper
conduct linsing operations. The area where rinsing is
(paliicularly black photographic paper) is the best backing.
conducted must be well ventilated and free of sparks or
Most commercially available pressure sensitive tapes are
other sources of ignition. Strict storage and handling re-
~u~rescent and this greatly reduces contrast between the
strictions apply to all flammable 1JUlterials.
ll1dlcation and the backing. If possible, nonfluorescent tape
should be obtained and used to make transfers with fluo-
rescent pmticles. Wet Fluorescent Tape Transfers
Since wet method particles are often fluorescent, the
procedures and precautions mentioned for dry fluorescent
particles must be observed. The use of black paper as a
Wet Method Tape Transfers permanent backing for the retained tapes is of particular
importance.
Tape transfers can also be made with wet method test Fluorescent indications may sometimes peIform as visible
indications, but this is not as simple as transfers with the dry light indications but their size and vi~ibility are not ideal
method. The first requirement is that the object surface and (particle concentrations are normally much lower than
the test indication be dry before the transfer is taken. The needed for white light visibility). Therefore, normal tape
procedure is to form the indication normally, allow the transfer techniques can provide low quality but usable
excess bath to drain off and to then dry the surface before transfers of fluorescent indications.
 274/ MAGNETIC PARTICLE TESTING

Archival Quality of Tape Transfer paper substrate and may be lost.

Records To minimize this problem, high quality translucent tape~
should be used. Tape records should be stored in a cool and
Tape transfers are fairly durable records of magnetic moist environment. vVrapping records in aluminum fOil
particle tests, but they do deteriorate with time. The retards degradation of the tapes. Good tapes properly stored
adhesive dries out and becomes brittle, and in some cases, last at least five years and often longer. Serious deterioration
so does the plastic tape base. The transfers then falloff the may be expected from the best tape records in about ten
years.
 l\\\\\\\\\\\\\'\\\'\\\\\"\\\\\\\\\\\\\'\\\\\\\\\\\\\'' ' \\'\\'\\\''\\\\\\\,,\\\\\\\'"
RECORDING OF MAGNETIC PARTICLE TEST INDICATIONS 1275

PART 3
('lit tapes
("quI <lIld
FIXING COATINGS FOR TEST
Will 1<>i1
k ston'd
INDICATIONS
'joration
)Ollt {pn

A helpful but seldom used method of recording magnetic Fixing for Removal from the Test Object
padicle test indications is the fixing coating. These coatings The other type of fixing coating is deSigned to be stIipped
are usually bulk matelials that are gently sprayed onto the off the test object so that the actual indication is preserved
test object surface. The coatings can fix a discontinuity elsewhere as a record of the test. These coatings are very
indication in one of two ways: (1) bind the indication firmly special mateIials containing a release agent that allows the
in place on the test object; or (2) lift: the indication for dried coating to be lifted off the object surface. Usually a
preservation elsewher~, much like a tape transfer. . . specially prepared pressure sensitive strip is pressed over
Only celtain matenals can be used as magnetlc partI- the area and used to pick up the coating with its entrapped
cle fixing coatings. Because test surfaces typically contain indication. The stIip provides support for the coating and
bath or other materials (scale, oil or water), a coating must . allows easier removal and storage.
adhere to the object surface in spite of contami- vVith fluorescent indications, the usual precautions apply.
nants. Specially formulated materials are required for this Strips and coatings may be fluorescent and can compromise
charactelistic. indication viSibility.

Application of Fixing Coatings

It is best to have the test object surface as clean and dry
, as possible before applying the coating. In production
Archival Quality of Fixing Coatings
. operations, this normally requires special dl)ing equipment
. Test indications fixed. to the test object retain their
to complete t~le recording in a reasonable amount of time.
validity indefinitely while the test object is inactive. The
Water baths are probably the easiest to fix because they have
characteristics of a normal service environment can quickly
shorter drying times than other vehicles and because water
erode or mask typical test results fixed onto the object
does not emit noxious or flammable fumes.
surface. Fixed indications that are removed from the test
Fixing on the Test Object object must be stored with the same precautions as tape
transfer records. Humidity and temperature controls are
One use for fixing coatings is to protect the test indica- needed to extend the life 'of fixed records.
tions so that abrasion from mechanical handling does not The pressure sensitive strips used for removing a fixed
damage the test result. Coatings designed for this purpose indication are special grades of expensive materials. Their
must he durable after dlying. life expectancy is limited but longer than that of the low
Such coatings are typically applied by spray and may be grade tapes often used for simple tape transfers. The
water based or organic mateIials. The coating must dry removed test indications may be mounted on paper backing,
hefore it can perform its protective function and in the wrapped in foil and stored in a cool, moist environment for
interest of speed, drying equipment must often be provided. archival purposes.
276/ MAGNETIC PARTICLE TESTING

PART 4
ALGINATE IMPRESSION RECORDS
Alginate impression compounds are based on mixtures of to the inspected part. The test object does not have to he
potassium alginate, calcium sulfate, sequestering agents dried before application of the recording media.
such as sodium phosphate, and fillers such as silica, If the area of interest is a cavity, it should immediatelv be
diatomaceous earth or calcium carbonate. \Vhen the com- fllled with the compound paste~ If the area of interest is
pound is mixed with a prescribed amount of water, a soft large or has a complex geometry, the recording compound
paste is formed. In three or four minutes, the paste becomes can be applied to a thin polyethylene sheet and the sheet is
a rubbery solid with the characteristic of accurately con- then smoothed over the full test area. Movement of the sheet
forming to and replicating, in negative, the surface to which or excessively u;orking the compound will smear the test
it is applied. This physical record in 1: 1 scale has proved indications.
valuable in many areas, including dental applications. After the paste sets into a solid, the magnetic particles are
The impression solid also absorbs and lifts particulate gently separated from the object by slowly peeling the
matter from the object surface and it is this absorption polyethylene sheet away from the surface or gently dislodg-
property that makes alginate compounds useful for record- ing the solid from its cavity. In a successful impression
ing magnetic particle indications. The primaty benefit of recording, the discontinuity indication is transferred to the
these compounds is their ability to lift patiicle indications compound material in its Oliginal size and shape. Test
from test object locations that cannot be directly viewed by results may then be examined and interpreted at a conve-
the inspector. nient location remote from the test object surface.

Archival Quality of
Application of Alginate Alginate Impression Records
Impression. Techniques
Because of their size, impression records are difficult to
To make alginate impressions, the test object is first retain for archival purposes. A common procedure is to use
processed through a typical magnetic particle test proce- the alginate impression method to retrieve a discontinuity
dure. Discontinuity indications are allowed to form nor- indication from an inaccessible location and to then photo-
mally. graph the indication for filing purposes. Supplementary
After the test is completed, the alginate impression documentation provides the magnetic particle test method,
compound is mixed according to the instructions of the the impression recording procedure and the photographic
manufacturer. The mixed compound is immediately applied data required for complete archival recordkeeping.
 RECORDING OF MAGNETIC PARTICLE TEST INDICATIONS 1277

PART 5
MAGNETIC RUBBER METHODS
Magnetic rubber recording uses a dispersion of magnetic process. Figures 4 and 5 show the preparation for magnetic
palticles in a rubber base that cures at room temperature. rubber tests of horizontal and veItical fastener holes.
The test object is magnetized with the uncured mbber Magnetic lUbber inspection is useful for inaccessible
covedng the area of interest. The magnetic paIticles then areas such as cavities or bolt holes where test indications
migrate to the leakage fIeld caused by a discontinuity. As the cannot be seen. The magnetic rubber method also prOvides
rubber cures, discontinuity indications remain in place on a permanent replica of the test area and indications.
the rubber (and over test object) surface. The process is time consuming because special holding
After curilig, the rubber is removed from the test object dams must often be constructed to keep the rubber in the
and the test results are interpreted in a convenient location. area of interest. Another disadvantage is that the recording
Figures 1, 2 and 3 depict the magnetic rubber recording medium must be mixed only at the time of the test and must

FIGURE 1. Typical magnetic rubber inspection FIGURE 2. A magnetic leakage field at a materia'
using electromagnetism: (a) diagram of rubber discontinuity causes the particles in magnetic
application;. and (b) cross section of an rubber to .mlgrate and concentrate over the
inspected hole discontinuity: (a) diagram of yoke positioning;
and (bJ cr~s~ section of an inspected ho.e
(a)
faJ

(bJ
(b)
DAM \ DUCT SEALER
PUnY'
ELECTROMAGNET
~
278/ MAGNETIC PARTICLE TESTING

FIGURE 3. After curing, the magnetic rubber replica is removed from the test object, revealing the
location of a crack as a vivid dark line; other surface discontinuities and conditions are also shown

'~
~.,
(/~....
f' ~_-.='~ \
\
., I I '._",-I'\. '
',.
---.. ........ ... ~\_/'-
"\ . J

"
'---' ~
RUBBER REPLICA

TABLE 1. Materials and accessories for magnetic

ru,bber indiqlti.on rect:?rping
Applications of Magnetic
Basic Materials
Rubber'TeChniqLies
Magnetic rubber compound
Cure and stabilizer
Magnetic rubber techniques require special supplies and
Dibutyltin dilaurate catalyst equipment (see Table 1). NecessalY cleaning materials in-
Duct sealer putty clude organic solvents such as acetone, naphtha, xylene or
Duct tape 1,1,1 trichloroethane. These must be able to dissolve oil and
Aluminum foil grease and must be stored with appropriate precautions.
Water and strong detergent may be used for cleaning.
Material Preparation Accessories
Balance (1,000 g capacity capable of reading 0.1 g)
Magnetizing Equipment
Disposable syringes (30 to 50 mL capacity)
Small disposable cups (paper or plastic) Because of the long magnetizing times, magnetic rubber
Disposable stirrers methods require the same kind of magnctizing systems llsed
for dlY method magnetic particle tests. The system may
Optional Special Equipment
Geometric pole pieces (special design to fit test object)
include yokes, prods, clamps, coils or central conductors
Central conductor and dam (assorted sizes) with alternating or direct current. The direct current yoke is
Microscope (7 x to lax for viewing very small indications) the preferred magnetization method. Permanent magnets
Field meter (electronic or dial probe) may also be used in certain test applications.
Heat lamp
Portable vacuum pump Test Object Preparation
Desiccator chamber
Using the chosen cleaning solvent, oil, grease and soil are
removed from the test object surface and the area to he
tested. Also remove metal burrs, smears and sharp edges.
be used immediately to avoid premature cure. Further-
more, the rubber medium is velY viscous and extremely long Be sure the surface is dry before proceeding.
magnetization shots are rcquired before indications can
If needed, a central conductor and dam assembly (Fig. 6)
are then installed and connected. Using tape, foil and duct
form.
 RECORDING OF MAGNETIC PARTICLE TEST INDICATIONS 1279

FIGURE 4. Horizontal hole preparation for magnetic rubber inspection

SLOTS FOR PAPER

IDENTIFICATION TAG

MAGNETIC
RUBBER

AIR VENT

TAPE

sealer putty, a reservoir is formed for the magnetic rubber. Permanent magnets should be used only if specific proce-
Virtually any test object configuration can be prepared for dures have been developed using reference standards with
magnetic rubber inspection using a dam assembly. Upside known discontinuities.
down surfaces may be tested by building a reservoir beneath Central conductors are best suited for fastener and
the test area and pressure filling it with magnetic rubber. A attachment holes, particularly when there are layers of
vent hole must be provided in the reservoir to prevent material requiring replication.
entrapment of air. Good magnetic contact is critical. Magnetic field strength
is greatly reduced \-vith poor contact between the test object
and the magnetizing yoke. To increase contact, auxiliary
Magnetization and Current Parameters
pole pieces may be machined from soft iron and attached to
Magnetization may be applied with: (1) portable electro- the poles of electromagnets. Pole pieces should be deSigned
magnets or yokes, (2) central conductors, (3) conventional to have the smallest possible reduction of cross section
magnetic paIticle units with clamps, prods or coils and consistent with space requirements. Figures 7 through 10
(4) pe.rmanent magnets. Direct current yokes are preferred demonstrate good and bad magnetic contact arrangements
~d, In areas of limited access, steel extensions or pole and the use of pole pieces.
pIeces may be used to transfer magnetism to the test object. Magnetic field requirements include both field strength
. Pe~manent magnets may be useful in certain specialized and duration. These are clitical to the success of the test
SItuatIOns where test object shape makes magnetization 'with because of the high viscosity of the rubber and the long
yokes difficult. The magnetic fields produced by permanent times necessary for migration of the particles to the discon-
magnets are often quite low and sometimes unreliable. tinuity. These requirements are detailed in Table 2.
280 / MAGNETIC PARTICLE TESTING

FIGURE 5. Vertical hole preparation for magnetic FIGURE 6. Typical central conductor setup for
rubber inspection magnetic rubber inspection of horizontal or
vertical fastener holes: (a) cutaway view of
fastener; and (b) cross section through the
fastener hole

SLOTS FOR PAPER

IDENTIFICATION TAG

FASTENER HOLE THROUGH

STRUCTURE

TYPICAL IDENTIFICATION TAG

TO POWER SUPPl.Y

SLOTS FOR PAPER Ib) .

IDENTIFICATION TAG

FROM MAGNA FLUX CORPORATION. REPRINTED WITH PERMISSION.

TABLE 2. Magnetic strengths and durations for

magnetic rubber tests Preparation and Use of
Field Strength Time Magnetic Rubber
Inspection Area Type millitesla (gauss) (seconds)
Magnetic rubber is prepared according to manufacturer
Uncoated holes 5 to 7.5 (50 to 75) 30
specifications and is only mixed immediately plior to its use.
2.5 to 5 (25 to 50) 60
First, the base rubber is thoroughly mixed to resuspend
1.5 to 2.5 (15 to 25) 240
the magnetic particles (a paint shaker is helpful). Mixing
Bare surfaces 15 to 25 (150 to 200) 30 to 60 should continue until there are no streaks or color variations
10 to 15 (100 to 150) 60 to 180 in the matelial.
5 to 10 (50 to 100) 180 to 600 Weigh or measure the specified amount of magnetic
40 to 50 (40 to 50) 600 to 900 rubber into a paper cup or other disposable container.
20 to 40 (20 to 40) 900 to 1,800 About 1 g of base matelial fills 1 mL. Do not measure out
more matelial than can be used within the allowable
VARIATIONS TO THIS TABLE MAY BE NEEDED FOR SPECIFIC APPLICATIONS
magnetization peliod.
 RECORDING OF MAGNETIC PARTICLE TEST INDICATIONS 1281

FIGURE 7. Good and bad magnetic contact FIGURE 8. Improved contact and improved
situations for magnetic rubber inspection: magnetic field using pole pieces: (a) poor contact
(a) very little magnetic flux gets through the (field partially lost); (b) improved field with the
aluminum to the steel; (b) a better magnetic field addition of pole pieces; and (c) magnetic field
Is obtained by leaving alternate steel bolts in and contact are improved with correct pole
place; and (c) the best magnetic field is obtained piece configuration
by magnetizing from the steel side
fa)

fa)

~
::
I,

: I
::

;;
I I
::

::
I I
ALUMINUM

STEEL
,I I I I I
c:&::9 ;;;;;;;;;;:i I&JlJ
fbI

fbI rjJJJl
~I
I
I
I
I'
!..
I
I!
I

I I I I i '1·
I I '1'.1 I I
" I' I I
<:&JJ C&;J \0&&1

fe) q f? A
i:
,. :t.I i ,.
: Ie)

~~
. Deaerate the material if air bubbles are a problem. This When longer magnetization times are necessary, the cure
ISdone by placing the mixed material in a vacuum chamber time may be extended by reducing the amount of catalyst.
and reducing the pressure to between 85 and 100 kPa Five drops of catalyst per 10 mL of base material provides
(25 to 30 in. Hg) for 60 to 120 s. about twenty minutes of usable magnetization time. Final
cure is about three hours. One drop of catalyst per 10 mL of
Catalyst and Stabilizer base provides about one hour of usable magnetization time
and final cure in about eight hours.
The catalyst and stabilizer are then added as follows. A
standard fo;mula is used whenever the magnetization peri-
Applying the Rubber to the Test Object
od does not exceed 480 s: 15 drops of dibutyltin dilaurate
catalyst and 2 drops of cure stabilizer for each 10 mL of base \Vhen testing holes with scored surfaces, small diameters
material. !\·1ix thoroughly, avoiding reintroduction of air or unusual configurations, the test area may be coated with
bubbles. After about eight minutes, this mixture becomes a thin layer of petrolatum to aid removal of the replica.
~oo thick for magnetic particle migration. Cure is completed Rubber may be applied by pouring from the disposable
1U about an hour. container if sufficient access is available. Inaccessible areas
282 / MAGNETIC PARTICLE TESTING

FIGURE 9. Improved contact and improved FIGURE 10. Correct positioning of pole pieces:
magnetic field using pole pieces: (a) very poor (a) cross section unnecessarily reduced by pole
contact (field partially lost); (bJ magnetism in the pieces; and (b) greater cross section maintains
radius is much improved with the addition of a less loss of field
pole piece
fa)
fa'

fb)

fb'

The replica is then visually examined for its condition and

discontinuity indications. A low powered microscope may
aid in this process. After inspection, test objects may be
demagnetized and cleaned with solvent or vapor degreased
may be filled with a disposable syringe. Such areas must be
if necessary.
equipped with vent holes to allow air to escape and rubber
to completely fill the cavity. These vent holes must often be
closed with tape or putty after filling to prevent escape of
the rubber compound. Archival Quality of Magnetic
Magnetization commences and must be completed in the Rubber Records
allotted time. Allow the magnetic rubber to cure for the
time specified by the chosen formula. Avoid movement of The magnetic rubber technique is not a means of record-
the test object and contamination of the magnetic rubber ing indications from a typical magnetic palticle test. I t is a
during the cure period. To determine that the magnetic different test procedure, requiting its own test parameters
rubber is completely cured (tack free), lightly touch the and its own specialized equipment.
replica or the mixture remaining in the mixing container. One of the advantages of the method is that it provides a
To remove the replica, first remove tape, aluminum dam, permanent replica of the test object surface as well as the
duct sealer putty and central conductor dam assembly, then particle discontinuity indications. This record may be re-
carefully extract the replica from the test area. The replica tained intact for as long as space is available. Magnetic
must be identified by a tag or label or by placing it in a rubber replicas may also be photographed for archival
labeled container. purposes.
 RECORDING OF MAGNETIC PARTICLE TEST INDICATIONS I 283

PART 6
PHOTOGRAPHY OF MAGNETIC PARTICLE
TEST INDICATIONS
Photography is the recommended method for making Typical test indications can occupy small areas on compar-
high quality recordings of magnetic particle test indications, atively large test objects and it may be possible to accurately
but it is by no means the simplest or least expensive method. record either the entire object configuration or the test
Test indication photography is not a pOint-and-shoot proce- indication, but not both in the same frame. It may be
dure. Specialized equipment is required and the photogra- necessary to photograph the test object and the test indica-
pher must have a good working knowledge of test object tions separately, with guide marks on the test object.
characteIistics and magnetic paIticle testing procedures. Another solution is to photograph the test object in part, but
EmpiIical data are critical to successful indication pho- with sufficient visual detail to show the location and orien-
tography. Considerable experimentation and reshooting is tation of the indications. Separate close-up photographs are
often necessary to obtain quality results, even with appro- then made to accurately record the test results. .
priate equipment and operator knowledge. Another difficulty comes-with fluorescent test indications.
For these reasons, indication photography is normally a Many magnetic particle tests use paIticles that are excited
labor intensive and expensive recording method. The tech- by near ultraviolet light and emit fluorescent visible wave:
nique is cost justified because it produces very high qu~lity lengths. The inspector's eye is sensitive only to the paIticles
recordings. Often it is the only method capable of providing emitted wavelengths, not the ultraviolet light. However,
t ' the results needed for publication, instruction or perm.ane~t many camera films and photodetectors an:; sensitive to
records_ Figure 11 is a typical photograph· of visible mag-
netic particle indications. Figure 12 is a photograph of
fluorescent magnetic particle indications. .
One of the difficulties in indication photography is the FIGURE 12. Ultraviolet photograph of fluorescent
difference between test result size and test object size. wet method Indications. on a gear

FIGURE 11. Visible light photograph of a dry

magnetic particle indication on a crane hook

FROM MAGNAFLUX CORPORATION. REPRINTED WITH PERMISSION. FROM MAGNAFLUX CORPORATION. REPRINTED WITH PERMISSION.
284 I MAGNETIC PARTICLE TESTING

ultraviolet wavelengths emitted by the light source. If a Even when a flash unit is used, a camera support allows
camera is to record only what the eye sees, then special control of composition and focus while releasing the shutter.
filters must be used. Producing ultraviolet photographs is a In addition to the SUppOlt, a cable shutter release is
very specialized procedure. Once the technique is learned, recommended for indication photography because it will
it is possible to produce accurate and impressive photo- virtually eliminate camera movement when tripping the
graphs of fluorescent particle indications. shutter.

Cameras for Indication Recording Choosing Film for

The most important feature of a camera used for indica-
Indication Recording
tion photography is the ability to take close-up photographs. Indication photographs can be made on nearly any kind
This capability is a product of the camera's physical config- of film, regular or instant, black and white or color. Special
uration and its lens. Close-up photographs may be taken ultraviolet and infrared films are not necessary and are often
with a macro lens, bellows or extension rings. Close-up inappropriate. Even photographs of fluorescent indications
photography also requires that the photographer know rely on visible light and the film must respond to visible light
exactly what is in the frame and what is in focus. Reflex and very much as the eye responds.
view cameras provide this ability while other types do not. For highest definition, slower fine grained films are
Built-in light meters are another valuable characteristic of chosen unless exposures become excessively long and a
cameras used for indication recording. Such meters are compromise is struck. If there is a choice, exposure indices
standard on many 35 mm systems, but completely automat- between 100 and 500 ASA are used. The most common
ed cameras should be avoided. Because indication exposure exception is 3000 speed black and white instant film. This is
parameters are far from normal, most automated cameras commonly used for quick record photographs not intended
must be used in the override mode, allOwing manual control for publication.
and compensation for the extreme conditions. When using color film, daylight types are normally chosen
unless visible indications are being photographed under
incandescent ~photoflood) illumination. Flash units emit
Accessories for' Indication 'Recording light of approximate daylight balance and natural light
requires daylight film. Fluorescent indication photographs
lenses are also better with daylight balance film because it has a
lower sensitivity to the visible blue and violet light emitted
The best lenses for indication photography are those by most ultraviolet sources. Incandescent balance films are
classified as normal or mild telephoto (50 to 135 mm focal very blue sensitive to compensate for the yellowness of
length in the 35 mm format). Wide angle lenses produce the typical incandescent source. Ultraviolet photographs
distorted images at the short camera-to-subject distances made with incandescent films tend to have an overall
required for small indications. Furthermore, it may be blue-green cast.
necessary to place a wide angle lens so close to the subject
that it prevents proper lighting.
Long telephoto lense~ are not often used because their Instant Films·
small depth of field sometimes prevents clear focus on the
entire area of interest. A long lens is necessary only if it is Instant films may be effectively used for archival indica-
impossible to closely approach the test object or the particle tion photographs. Instant film can be chosen hom a valiet)'
indication. of speeds, color or black and white, depending on the
Fast lenses are of no pmticular advantage - most requirements of the testing application.
indication photography is done at [5.6, [8 or even smaller In some cases, instant films are used to set up the final
lens openings in order to get acceptable depth of field. regular film exposure. Subject, camera and lighting are
arranged for the record photograph and a tlial exposure is
made on instant film. The instant film may be in a diHcrent
Equipment Supports
camera or it may be paIt of a preview or proof exposure pack
Some sort of camera support is usually necessmy for test mounted on the back of the camera body.
indication photography. This could be a standard tripod, The photographer examines the proo( and makes IIeces-
clamp or copy stand. SUppOlt is needed because exposures sary corrections before taking the archival exposures. !-.l uch
are long in this application and camera movement can be a time and expensive film is saved by this procedure. It is very
problem. difficult to predict exactly what an indication photograph
 RECORDING OF MAGNETIC PARTICLE TeST INDICATIONS I 285

wiII register without some soli of trial or extensive exposure Gelatin filters are sometimes available from commercial
bracketing. photography stores but more often they must be ordered
from a manufacturer.
Gelatin filters can be mounted in a filter frame that
Filters for Indication Photography attaches to the camera lens. For occasional use, they may
also be taped over the lens. Gelatin filters are not durable
Filters are often used in indication photography because accessories. They are easily scratched or contaminated and
the available light is seldom what is needed. The pUll)ose of are not easily cleaned. In addition, they have little heat
fllters is to remove unwanted light or to correct the color resistance and some types fade after extensive exposure to
balance of the light in use. vVhen using filters, there is a loss light. Gelatin filters must be inspected and replaced regu-
of light transmitted to the film and this must be compen- larly, but they are excellent for their purpose if their
sated hy longer exposure. . limitations are not exceeded.

Filters for Visible Light Photography Types of Photographic Filters

Filters are used with visible light photography to alter The filters listed below are typical of those produced by
color balance, improve contrast and produce special effects. many manufacturers and are representative of the types of
This is a voluminous subject well documented in the photographic filters useful for recording magnetic particle
literature. test indications (see Table 3 and Fig. 13).
One visible light filtering technique is especially impor- A valuable filter for this kind of photography is one
tant for test indication recording: the use of p~larizing filters deSigned to cut off wavelengths at 405 nm. Such a filter
to reduce highlights and reflections. Reflections are often a transmits no ultraviolet and almost no violet light. It shows
problem with magnetic particle test methods because the test indications in their normal color but does not show the
test objects are often metallic and highly re!Jective. test object very. well unless suff~eient ambient light ·is
The effect of the polarizing filter is vari~d by its angle of present. The 405 nm filter is excelleht when test indications
rotation. If the camera is a reflex or view model, the are small and much ultraviolet light is used. It is.also a good
polalizing effect is seen through the viewer . Otherwise, a filter when highly reflective surfaces are present or when
manual orientation of the filter is reqUired. Remove the the test object has a slightly fluorescent coating. The 405 nm
filter from the camera and, staying as close as possible to the filter 'can be used with fluorescent tube sources that emit
expected camera position, view reflections off the test object considerable amounts of visible blue and violet light.
through the filter while rotating the filter in front of the eye. The best all around filter for fluorescent indication
When the proper filter orientation is determined, the filter photography cuts off at 390 nm and passes the violet 405 nm
is reinstalled on the camera in that position and the band of a mercury light source. This band is visible to the
exposure is made. eye and the filter uses it to photograph the test object and
make. it visible along with the indications. The result is a
picture closely resembling the image seen by the eye in a
Filters for Fluorescent Indication Photography
Photographic films are responsive to ultraviolet light and
this causes serious photographic problems unless the ultra-
violet wavelengths are filtered out. Fluorescent indication TABLE 3. Filters for fluorescent magnetic particle
photography is only possible if a ultraviolet absorbing, indication photography
visible transmitting filter is placed over the camera lens.
A further requirement for such a filter is that it not be Cut-Off Manufacturers
fluorescent itself. A fluorescent filter used in an ultraviolet Wavelength
(nanometers) Kodak B+W Tiffen Rolyn
environment causes fogging of the film image. Since the
light emitted by fluorescent indications is visible light, the 390 28 UV-15 65-1290
filter should not change the apparent color of this light, 405 2A 420 Color Haze 2
unless the change is done intentionally for special effect. 415 2E 021 65-1300
The filters normally used for indication photographs are 420 3 Color 3 65-1305
called ultraviolet filters or haze filters. Some of the best are 450 4 65-1310
gelatin filters. These are sheets of plastic or acetate contain- 460 8 (K-2) 022 Color 8
ing carefully controlled amounts of dye to produce the NOTE: FILTERS NOT LISTED ABOVE. INCLUDING KODAK lA AND B+11I409
desired light transmission. They are produced in sizes from FILTERS. MAY TRANSMIT TOO MUCH ULTRAVIOLET LIGHT FOR FLUORESCENT
MAGNETIC PARTICLE INDICATION PHOTOGRAPHY
.50 X 50 mm (2 x 2 in.) to .350 X 450 mm (14 x 18 in.).
286 I MAGNETIC PARTICLE TESTING

A sharp cut yellow filter limits exposures at 450 mn. No

FIGURE 13. Transmission curves of filters used for blue is included and indications appear unnaturally yellow.
fluorescent indication photography This filter is best restricted to black and vvhite exposures
where its background limiting properties may be useful. A
separate white light exposure is necessary if the test object
is to be photographed.
./ Another yellow filter has been deSigned to cut otT at
460 nm and to reduce blue-green light at the same time.
Indication colors are heavilv altered and the test object is
not visible at all without a separate white light exposure. The
460nm filter may be used for black and white work if only
the indications must be recorded or if additional visible light
,"', "'.
exposures are acceptable for photographing the test object.

Other Haze Filters

\/
o. I jJ...:....--+----'-+JI.--4I---+-----I---I---____-
Many commercial glass covered and ring mounted filters
are sold as haze filters. Most of these .are a glass sandwich
300 350 400 450 500 550 600 650 700
that may be threaded into a filter mount on the camera lens.
WAVELENGTH Some such filters (usually called lA filters) cannot always be
(nanometer) used for recording magnetic particle test indications be-
cause they transmit too much ultraviolet light.
However, some commercial haze filters can be used if
LEGEND they are deSigned to limit the proper wavelengths. It is
HAZE FILTER recommended that test exposures be' made to verify the
------------- 390 nm FILTER
'results of a new filter in its intended application. Glass haze
415 nm FILTER
.----.-- - 460 nm FILTER filters can have several practical advantages. They are more
MERCURY ARC FILTER convenient to use, easier to keep clean, more heat resistant
and they· k1st much longer- than gelatin filters. Glass filters .
FROM EASTMAN KODAK COMPANY. REPRINTED WITH PERMISSION do not often require markedly increased exposures and they
do not noticeably change the color balance of the final
photographic image.
darkened testing booth. This realistic image is produced
without a supplementary visible light exposure to show the Polarizing Filters
test object. The high ultraviolet light levels needed to bring In most instances, polarizing filters do not transmit
out small indications can result in overexposures with this l.lltraviolet light and can be used in fluorescent indication
filter. Also, if the background is slightly fluorescent, the photography. However, they only transmit 10 to 2,5 percent
390 nm filter does not provide sufficient contrast. of the visible light and greatly extend the lengthy exposure
A pale yellow filter is available to cut off exposures at times necessary for fluorescence photography. Polarizing
415 nm. It does not transmit the 405 nm violet that often filters can be used if an ultraviolet filter is not available or to
produces the image of the test object. Unless the test object reduce blue renections from a shiny test object.
is slightly fluorescent or the source contains some blue or Figure 13 shows the light transmission of several filters
green light, only the test indications are shown with the often used for fluorescent indication photography.
415 nm filter. If the test object should be photographed, a
separate white light exposure is necessary. The 415 nm filter
does not alter the color of the indication so it is the correct
filter if fluorescent tube sources are used or if background Exposure Estimation and Metering
fluorescence is present.
A light yellow filter is designed to cut off exposures at The light meter built into a tyl)ical 35 mm camera helps
420 nm. It is similar in use and effect to the 415 nm filter determine proper exposure. Some meters set exposure
except that it also partially reduces blue wavelengths and speed at a chosen lens opening, while other meters set lens
makes test indications appear more green. The 420 nm filter opening for a chosen shutter speed. This is a reCiprocal
may be used where this effect is desired or with black and relationship: more light from the apelture requires less
white film records. speed from the shutter. The light meter establishes this
 RECORDING OF MAGNETIC PARTICLE TEST INDICATIONS I 287

· No feiat iO/lship by averaging the light striking the whole frame. Long exposures must be expected for fluorescent indica-
lIow. The light values from the center of the scene are often tions, particularly since the lens opening should never be
!Ires prci'<-rt'lltially weighted by the meter. larger than [.5.6. Limiting the [-stop provides adequate
Il.A sharpness (depth of field) over the pictured area of interest.
'.iect Even with ASA 400 film, exposures are seldom less than 1/8
Visible Indication Exposures
second. In many cases exposures are two minutes or more.
f at \'isihle indication photography requires normal exposures For this reason, fast films are often chosen at the expense of
me. with ccrtain adjustments. Exposures can be estimated with high grain in the image.
t is a t:Vi('al exposure meter but often the meter must be Proper exposure is less difficult for black and white
I'he O\'('lTidden to obtain the desired result. This occurs because recording, because there is no concern for color accuracy. In
Illy Ih(' kst ohject (typically filling the frame) is a monotone and color photography, underexposure and overexposure can
~ht is oliell at a bIightness level the meter is not designed to produce images of an indication, but the color will be
ct. read. \lost photographic light meters are designed to give illaccurate. Underexposed indications usually appear bluish,
prnpn ('xposure of a 17 percent gray scene. This compro- regardless of their actual color. Overexposed indications are
mise ~r;ly value is a medium tone. In recreational photog- pale or even white.
rapll\. light subjects such as snow or water scenes are
'rs lllltlr'f('\posed if photographed at the meter settings; dark
~h Changing Exposure Time
SC('lj('~ klve overexposed light areas.
s. Changes are commonly made to exposure time (shutter
SilW(' test objects are mostly monotones and test indica-
'e tions are a very 'small part of the total area, it is necessary to speed) rather than to the lens opening. For magnetic
adjllst the exposure to make the subject appear natural. A particle test indications, the aperture is usually limited to [8
gra.'·islr test object reflecting 15 to 20 percent of the light and is no larger than f5.6, for proper depth of field.
if Because of the reciprocal nature of photographic exposures,
striking it should be exposed as the meter indicates. A velY
s the remaining option is to change the exposure time. Start at
light colored test object reflecting 70 to 90 percent of the
light st rikillg it should be exposed two [-stops over the one-half second, make another exposure at one second and
mdcJ'(·d value. For example, changing a recommep.ded [11 • possibly another at tWo seconds.
to all r.S.6 setting increases the aperture opening, admits Most cameras do not automatically time exposures longer
more lig:ht and properly exposes a dark test indication in a than one second. Longer exposures are made by using the B
bligllt field. A very dark test object reflecting 5 to 10 setting and timing the exposure manually. Unfortunately, it
perc('])t of the ligh~ should be exposed one or two [-stops is difficult to accurately time exposures less than four
lIllde] the metered value. seconds. Therefore, exposures in the one to four second
E\pl 'stires must be set to compensate for the filters being range should be avoided by using a smaller lens opening and
IIsed. This is done by the meter when using a reflex camera an exposure of four seconds or more.
",it h t hrough-the-lens metering. Otherwise, position the
salll(' I1ltn over the meter or use recommended compensa- Reciprocity Failure
tion Ltdors to determine exposure settings. Documentation Reciprocity failure is a problem caused by the long
frollJ thl' filter manufacturer can also recommend exposure
exposures needed for indication photography. The reciproc-
('han~('s for the filters in question.
ity law states that the optical density of a developed
photograph is directly proportional to the exposure time and
Fluorescent Exposures
the illumination. Decreases in illumination require corre-
Fhlurt'scent indication photography requires more exper- sponding increases in exposure time, and vice versa. Films
j] III 'llLltiOll to get proper exposures from a test indication. are balanced for exposure times between 1110 and 1/1,000
Pn )of ('\jlosures are required because of broad variations in second because most exposures are made in this range.
t}}(, photographic parameters: (1) incident ultraviolet inten- Longer or shorter e'-posures result in changes of effective
si/\. i.:2 \ test object surface reflection, (3) fluorescent back- exposure index, color balance or both.
~n "lIlel Jlld (4) indication brightness. Long exposures typically result in a lower exposure index.
Li!.!;llt meters register the visible light emitted by fluores- However, the color balance changes that occur mayor may
('('lit p;lrti('le discontinuity indications. The best estimation not be compensated by longer exposure. In some cases, only
of nposmC' is obtained from a typical light meter equipped a filter can correct color balance. The only way to ensure
\\ it h t I j(' same type of ultraviolet filter used on the lens. The proper technique is to run a proof test, making careful notes
pr()!'!t'1I1 with this is that the measured light intensity may of all conditions. Once the results are seen, corrections can
!Jr, "() 10\\' that the meter is not sensitive enough to read it. be made for the archival exposures. This is a tedious and
Ti,; ultra\iolet intensity can then be increased or estimated expensive process that should be used only if high quality
!l Lit I II 'lllatil'allv. . photographs are required.
288 / MAGNETIC PARTICLE TESTING

visible test indication photography is daylight. It is readily

Bracketing Exposures available, of sufficient intensity, and evenly distributed. Its
character does change depending on whether it is direct
Indication photography is more difficult than scene pho- sunlight, cloud obscured or in shade.
tography and is typically much less predictable. Even
experienced photographers find it necessary to bracket Incandescent Light Sources
exposures to ensure accurate representations of a test
indication. Bracketing is the exposure of additional frames Incandescent light is usually produced by tungsten fila-
with increased or decreased [-stops to compensate for ment lamps. It is readily available and often of sufficient
unforeseen conditions. intensity. Its color varies according to the temperature of
Most photographers routinely bracket critical shots, time the filament but it is always yellower than daylight or
and conditions permitting, since the cost of the additional fluorescent light. When using incandescent light with color
film is much less than the cost of reshooting the session. film, special color balance film is reqUired or special blue
Bracketing can be safely omitted when the exposure is one filters are needed over the lamp or over the camera lens.
of a series done under identical conditions with well known Figure 14 shows the spectral distribution of daylight and
results. incandescent sources used for magnetic particle test indica-
tion photography.
Visible Indication .Bracketing
Visible indication photography generally needs little Fluorescent Light Sources
bracketing because meter readings are more trustworthy in Fluorescent lamps produce light by bombarding gases
this application. A normal visible bracketing series includes with electrons from a cathode and producing ultraviolet
a shot with one [-stop increased exposure and another with energy. This energy excites a fluorescent coating inside the
one [-stop decreased exposure in addition to the shot at" the tube which in turn emits several wavelen~ths of visihle and
calculated or metered exposure.
The calculated exposure is the exposure including all
necessary compensations. Note that moving one [-stop FIGURE 14. Typical spectral distribution of noon
doubles or halves the exposure. Photographic films are daylight and incandescent nght .
relatIvely tolerant of errors smaller than. th~s. and smaner
increments are necessary only in the most critical cases.
110 I

Fluorescent Indication Bracketing 100 ~ ./ ~

io-'
Fluorescent indication photography is much less predict- 90
V
/ ro...J'
able than visible indication photography and more bracket- I I
I

ing is required to ensure a good result. Experience can 80

,
reduce the amount of bracketing required but at least one or
~~
70
I I

two bracket exposures are typically needed for each subject.

If the test indications are large and bright, one additional
we:
Z::J 60 1 ,
,
I

shot at one [-stop decreased exposure may be made.

1--(1)
I.~
\.:J~ 50 1 ,
I

Average size or small indications can often dispense with -~

Cl::::(I)
CO'- 40
~ I

this decreased exposure. vVith fluorescent test indicatiolls, I

always make one additional exposure at one [-stop increased 30 J

exposure. It is recommended that a second bracketed 1/ I
I
I

20
exposure be made at two [-stops increased exposure. For , I

10
added assurance, still another exposure can be made at a
three [-stop increase.
400 500 600 700

Lighting for Test WAVELENGTH

(nanometers)
Indication Recording
LEGEND
Daylight Sources NOON DAYLIGHT
- - _. 2.848 K INCANDESCENT
Lighting techniques and light sources are extremely
important in photography. The preferred light source for
 RECORDING OF MAGNETIC PARTICLE TEST INDICATIONS / 289

dilv invisible light. Fluorescent lamps have adequate inten.sity

1t~'i FIGURE 15. Spectral distribution of mercury arc
and very even distribution but their color balance is differ-
'ect and fluorescent light sources
ent fro111 all other sources.
Fluorescent light appears white to the eye but it has a 100
/
very green tint on film. Filters are available to correct this / \
light to approximate daylight. The filters are a purple color I \
I \
and should always be used when making color photographs I \
Ia- I \
nt under fluorescent light. I \
I
I \
of I \
)r \ I
)r
10~--7:---4--~--~H+------+---­
Sources for Fluorescent Indication Photography 32' I
e LUCIJ
ru I
Fluorescent indication photography is normally done VlQ.. I
z~ I
under illumination provided by mercury vapor ultraviolet 0u... ...,
0 I
I
lamps. These lamps are often used for inspection of mag- VlC
LUCIJ I
Q::~ I
netic particle test objects and may be spot lights or flood CIJ
.9: I
lights. For fluorescent photography, the lamps are equipped I
with a black glass filter that cuts off all visible light except a ,,
small amount of violet. Such lamps are often used ~r:ithout a
filter for space illumination in factory areas and it may
.sometimes be necessary to photograph visible indications
using this source. Such light is extremely blue and a
yellowish filter is reqUired to obtain proper color balance
with color films.
Mercury arc. ultraviolet sources are often highly direc- 300 350 400
fiOlial and spotty when photographed. They illuminate the
area of an indication well but may not produce enough WAVELENGTH
distributed light to photograph the rest of a large test object. (nanometers)
In such cases, it may be necessary to use a visible light
double exposure to show the test object. If this is done, the LEGEND
test object must be underexposed in visible light to retain FLUORESCENT
good images of the indications. Such visible light exposures FILTERED MERCURY ARC

are made at least two [-stops less than normal exposures.

Tubular Black light Sources

Another light source used for fluorescence photography is Developing Procedures for Test
sometimes called the tubular black light. These are standard Indication Photographs
fluorescent fixtures equipped with special bulbs containing
an ultraviolet emitting phosphor enveloped in ultraviolet Films containing indication photographs can be devel-
transmitting black glass. Such lamps produce a majority of oped with normal procedures. Special techniques are not
their light in the near ultraviolet spectrum but they also have necessary although they can be used for some special
a considerahle output of blue and violet visible light. Their effects. Color prints produced by commercial processors
output is not focused so that they illuminate a large area may be a problem because their processing procedures are
~ith()ut high intensity in anyone spot. often color compensated and based on an easily recognized
Tubular sources can produce an acceptable photograph of feature of at least one shot in the series. Few indication
a large area containing large, bright test indications, partic- photographs have any feature easily recognized by a
ularly if the camera is equipped v,r:ith a filter that cuts off photoprocessor and color compensation may not be accu-
violet light. Such sources are not good for small or dim rately performed.
indications because: (1) they have low spot intensity; and Instant color films are used for indication photography
(2) their large visible light output cuts contrast. even though they are subject to serious reciprocity failure at
Figure 15 shows the spectral emission of several ultravi- the long exposures needed for fluorescence photography
olet light sources often used in magnetic particle testing. N onnal developing of these films produces a very blue
290 / MAGNETIC PARTICLE TESTING

picture. This can be corrected by using one-half to three- These programs are designed to enhance contrast, isolate
quarters of the recommended developing time. indications, differentiate signal from noise or use indication
characteristics (image density, size or shape) to help deter-
mine if an indication is a valid representation of a disconti-
Archival Quality of Indication nuity. Automated magnetic particle testing systems use
digitized image enhancement as the first step in interpreting
Photographs test results and determining the actual condition of the test
object.
Photographs of magnetic particle test results have two
distinct advantages. \\-'hen properly exposed, photographs
have the highest quality reproduction of any of the record-
ing methods. Conclusion
In addition, photographs can show much more than just
the indication's particles. The test object can be completely The magnetic particle test method produces visible indi-
pictured, with the particle indication in position. The cations of surface and slightly subsurface discontinuities in
indication's size, shape and intensity can be seen and it is ferrous test objects. These indications are visible to the eve
this abundance of information that most recommends the in size andlocatioIl, but they are not durable. If records 'of
photographic recording method. If stored with a minimum magnetic particle tests are necesscllY, additional procedures
of care, photographs have an indefinite shelf life. for recording the indications are required .
. This text summarizes some of the techniques commonly
used to produce test records. The inspector must choose a
Electronic Recording Techniques method appropriate for the particular application and must
then become proficient in its use. In some cases, new
Electronic imaging is regularly used as a test indication recording procedures must be developed.
recording medium. Black and white and color television In one case, a combination photographic record was
cameras have the sensitivity needed to record visible or r.equired to show the visible outline of the test object and
fluorescent test indications. Sufficient lighting and filtering the fluorescent test indications. Because double exposures
of unwanted ultraviolet wavelengths are requirements are difficult with modern cameras, a Single exposure was
shared with still photography. It is often necessary to adjust made. Using a polarizing filter and the camera's B or bulb
illumination, levels to o~tain the n~<;essary quality f~r video setting, the lens was .0peI}ed in the presence of ambient
recordings of test indications. visible light and a hvo second exposure was produced. The
Video tape exposure time is basically fixed so that the visible light was then removed and a handheld ultraviolet
camera's aperture size and gain setting are the controls to source introduced to expose the particle indications on the
adjust for improved exposure. Video tape has a somewhat same frame.
different color sensitivity than the eye or still films and This kind of recording may be a unique application, but it
filtering is adjusted for the individual camera system and the is indicative of the flexibility available to the magnetic
recording conditions. Ultraviolet filtering can be achieved particle inspector. There are many recording techniques,
with the same filters used on still cameras. from drawings and tape transfers to alginate impressions
One distinct advantage of recording discontinuity indica- and photography. Each should be considered for certain
tions on video tape is the availability of subsequent image indication recording situations, and each can be adapted in
processing. Television images may be digitized and pro- its own way for producing successful magnetic particle
cessed by computerized image data analysis procedures. testing archives.
 SECTION

DEMAGNETIZATION OF
TEST OBJECTS
Kenneth Schroeder, Schroeder and Associates, Arlington Heights, illinois
Roderic Stanley, International Pipe Inspectors Association, Houston, Texas
Lawrence Wong, Magnaflux Corpo!atiC?n, Chicago, Illinois
292 / MAGNETIC PARTICLE TESTING

PART J
DEMAGNETIZATION AND RESIDUAL
MAGNETISM

Demagnetization is the process of removing magnetism called leakage field'} and are the phenomena responSible for
from a ferromagnetic material. This is accomplished on the attracting nearby magnetic particles.
molecular level by establishing or reestablishing random
orientations in the material's magnetic domains.
The demagnetizing process is often required in conjunc-
tion with magnetic particle testing. However, demagnetiza-
Requirement for Demagnetization
tion may also be required for reasons other than magnetiC
Components that retain a relatively strong residual flux
particle tests.
density can be a source of problems ~uring subsequent
manufactUring processes or during service. Some typical
problems are detailed below.
Ferroma.gnetic Materials When subsequent machining is performed on a test
object, the presence of a strong residual flux denSity can
Ferromagnetic materials are characterized by a relative attract and hold chips or particles. This ca~ adversely affect
eas~ a
of magnetization when exposed to magnetizing force. surface finish or tool life. . .
. This can be attributed to the relatively high magnetic In an arc welding operation, the pI:esence of strong
permeabilityJ.L exhibited by these materials. Once magne- leakage fields can deflect the arc away from its intended
tized, ferromagnetic material retains a level of magnetic flux location. This phenomenon is sometimes called arc blow.
denSity B after the magnetizing field strength H has been Electron beam welding is also adversely affected by residual
removed. This remanent flux density is commonly referred magnetism. Even leakage fields of moderate strength can
to as residual magnetism or the residual field. The magni- deflect the electron beam away from its intended target.
tude of residual flux denSity is. a function of the following Plating quality also can be affected by residual magne-
factors: (1) the magnetic characteristics of the material; tism. In an operation such as chrome plating on a steel
(2) the immediate history of the material's magnetization; surface, the presence of a strong residual magnetic field can
(3) the strength of the applied magnetizing field; (4) the divert the plating current away from its intended location.
direction of magnetization (longitudinal or circular); and Such mishaps are costly, incorporating the expenses of the
(5) the test object's geometry. first faulty plating, possible repair and a second plating
Residually magnetized materials contain magnetic flux <l> operation.
measured in webers. In the International System of Units In-selvice operation of an object may be impaired if it is
(SI), 1 weber (\Vb) passing through 1 m 2 gives a flux denSity the source of excessive leakage fields, attracting metallic
of 1 tcsla (T). In the centimeter-gram-second (cgs) system, chips or tramp paIticles. This condition can cause malfunc-
flux is expressed in maxwells or lines of flux. Flux denSity tions in rotating assemblies such as bealings or excessive
is expressed as lines per square centimeter or gauss (G). wear on bealing surfaces. Strong residual flux densities also
One weber is equivalent to 10 8 lines and one tesla equals may cause a substantial magnetic attraction between adja-
104 gauss. cent moving parts. This can produce increased fIiction or.
Depending on the direction of the flux with respect to the may ill other ways interfere with the intended function of
surface of the test object, these lines may leave the object at the component.
one point and continue their path through air before Cleaning operations for the removal of metallic chips and
reentering at another point. The points where magnetic particles can be hampered by the presence of residual
lines of flux leave and reenter a ferromagnetic material are fields. This can he of speCial concern when cleaning ohjects
called magnetic poles. The north pole is where flux lines that have internal openings such as oil passages.
leave the material. Flnx lines reenter the material at the Residual leakage flelds can have a direct adverse effect on
south pole. Such fields emanating hom a test object are certain types of instnunentation. The magnetic compass
 DEMAGNETIZATION OF TEST OBJECTS I 293

aboard aircraft is an example, as are many electronic Residual magnetization can occur with improper opera-
components. tion of alternating current through a coil demagnetizer. This
Rotating shafts containing residual magnetism may act as occurs when the coil current is terminated and the test
electric generators in conjunction with the Earth's magnetic object is still within the coil or within its influence. Pene-
field or other stray fields. The generated electricity can heat tration depth is determined by the frequency of the alter-
the shafts, causing operating losses and other service prob- nating current.
lems. Finally, under certain conditions the Earth's magnetic
Finally, objects sometimes need to be demagnetized field can impart a longitudinal residual field. This occurs
before using established magnetizing current levels for only with low coercivity steels, usually when the object is
magnetic p,-uticle testing. The presence of strong residual shocked or vibrated while its long axis is parallel to the
induction can lead to erroneous results and faulty test Earth's magnetic field. Such a residual field may become
procedures. quite significant for long parts subjected to severe vibrations
Demagnetization is not required when the test material in service or transport. The intensity (horizontal compo-
exhibits very low magnetic remanence. Low carbon or nent) of the Earth's field in the United States is about
nodular steels are included in this category. If the next 16 A-m -1 (0.2 Oe). This is equivalent to the magnetic field
manuf~lCtUling process calls for the object to be heated strength at the center of a five ampere-tum coil 300 mm
above the Curie point, the matelial becomes nonmagnetic (12 in.) in diameter.
and demagnetization is accomplished by the heating pro-
cess. At the Curie point, steel temporarily transforms from
a ferromagnetic to a paramagnetic state and subsequently Effect of Magnetic Field Origin on
cools with zero net induction. Demagnetization
If the pmt does not require additional machining and its The type of magnetizing source becomes significant when
intended function is not compromised by the presence ofa conSidering the nature of residual induction and its subse~
residual field, then demagnetization becomes unnecessary. quent demagnetization.· .
Alternating currents tend to flow near the surface of a
conductor and this so-called skin effect produces residual .
Common Sources of Residual fields that are sUlface oriented. Such fields respond well to
Magnetic Fields alternating current demagnetization techniques ..
There is practically no skin effect associated with a direct
Ferromagnetic .material can become magnetized in a current magnetizing source and consequently the entire
numl~er of ways, intentionally and unintentionally. cross section of an object can be residually magnetized.
Parts may be purposely magnetized to perform magnetic Deep seated residual fields in larger objects may not be
particle tests or to facilitate a magnetic flux leakage test. affected by alternating current demagnetiZing techniques
Magnetization may also come from biasing magnetic fields because the skin depth is only about 1 mm for steel of
such as those used to negate the effects of permeability relative permeability 100 at 60 Hz.
c.hanges in eddy current tests.
Magnetic chucks are a common source of residual fields.
Pronounced fields can be left in an object if a chuck's
built-in demagnetizing cycle is faulty or inadequate. Prema-
turely removing test objects before completion of the de- Types of Residual Magnetic Fields
magnetizing cycle can also be a source of residual magnetism.
Lift magnets are convenient material handling devices,
Longitudinal Magnetic Fields
but because they operate with direct current, they can leave
strong residual fields in fabricated objects or in raw mate- MateIials magnetized by a coil or solenoid sometimes can
rial. Physical contact with any permanent magnet or highly be left with a longitudinal residual induction. The field is
magnetized object (a machine table or fixture) can also oriented lengthwise in the test object and there is a high
establish a residual field. concentration of emergent fields at each end. These fields
Low hequenc), induction heating that is abntptly termi- constitute poles, lines of magnetic flux entering or leaving
nated can induce very strong residual flux density. The the material.
lower the frequency, the deeper the magnetizing field Longitudinal magnetization is easily detected by field
strength penetrates into the matelial. measming devices such as the gaussmeter or by the attrac-
\Vith certain oIientations of high amperage cables, an tion of other ferromagnetic matelials. \Vhile this type of
electIiC' arc welding operation can magnetize ferromagnetic field can adversely affect subsequent machining, it is usually
mateIial. vel)' responsive to demagnetiZing techniques.
 294 / MAGNETIC PARTICLE TESTING

Circular Magnetic Fields

keyway. All holes or slots cutting through a circular field
Unlike longitudinal residual induction, circular residual produce magnetic poles that can attract materials such as
induction can exhibit little or no external evidence of its chips or dust from subsequent machining.
presence. The flux may be entirely confined within the Without special equipment, demagnetization of a circu-
material, depending to some extent on part geometry and larly magnetized object can be very difficult. Confirmation
the magnetizing procedure. of an adequate demagnetization level is an additional prob-
For example, if magnetizing current is passed through a lem. Leakage field measuring devices are ineffective since
homogeneous length of ferromagnetic bar stock having a there may not be an external leakage field to monitor. In this
circular cross section, the resulting circular residual flux case, reorientation of the circular field into a longitudinal
density is for all practical pUlposes undetectable without field prior to demagnetization may be advantageous in some
altering the bar in some manner. Virtually no leakage fields instances when such a procedure is compatible with part
emanate from its surface because the magnetic flux path is geometry and size.
closed within the object. Because of this, the internal
residual flux density B may be much stronger than if the test
object had been magnetized longitudinally in a coil or Multiple Magnetic Poles
solenoid having a comparable magnetic field strength H.
The circular residual flux density becomes an apparent Multiple magnetic poles can be induced ip and retained
problem when the geometry of the object is altered by by ferromagnetic material that has been exposed to direct
subsequent machining. For example, if a keyway is cut in a current magnetization, as in a magnetic chuck or lift mag-
piece of shafting that is circularly magnetized, the circular net. These fields can be pronounced and take the form of
field becomes quite evident. Strong leakage fields (north alternate north and south poles at relatively'dose intervals.
and south magnetic poles) occur on either side of the Demagnetization of multiple poles may be difficult using
conventional alternating current through-coil procedures.
 DEMAGNETIZATION OF TEST OBJECTS / 295

PART 2
PRINCIPLES OF DEMAGNETIZATION

Ferromagnetic material differs from other material in As H is reduced to zero, the level of B falls to + Br rather
that it contains magnetic domains, localized regions in than zero. This level of retained magnetization ( + B r) is the
which the atomic or molecular magnetic moments are remanent or residual field remaining in the material. As
aligned in parallel. vVhen a matclial is not magnetized, the negative values of H (opposite polarity) are applied, the
domains are randomly O1iented.and their respective mag- curve passes through - Hc and on to - HTIl and reaches a
netic inductions sum to zero. \Vhen the material is exposed magnetization level of - Brn' Reversing the applied field to
to. a magnetizing field strength H, the domains tend to align a value of + Hm completes the magnetic hysteresis loop.
Wlth the applied magnetic field and add to the applied field.
\Vhen tIle magnetizing source H is removed, some of the
magnetic domains remain in their new orientation rather
than ~etuming to the original random orientation and the Retentivity and Coercive Force
matenal retains a residual magnetic field.
The value Hc is the coercive force and is an indicator of .
the difficulty involved in demagnetizing a material. The .
Magnetic Hysteresis : value of Br is the material's retentivity and indicates the
residual induction within a section of the material. As a mIa, .
. Magnetic hy~teresis is a lag in the change.of magnetiza- high co@rcive forces are associated with harde'r materials
~lOn values after a change in the magnetizing force. Figure 1 and low coercive forces with softer materials. Therefore,
Illustrates the relationship between the magnetic field harder materials usually offer more resistance to demagne-
stren.gth H (magnetizing source) and th~ magnetic flux . tization and require a higher demagnetizing field than softer
denSIty ~ (level of magnetization). When an unmagnetized materials.
marenal IS magnetized by gradually increasing H from zero No definition has been made for the dividing line between
t~ -: Hm, the level of magnetization B increases along the hard and soft materials. However, if Hc ;::: 8,000 A-m- 1
~rgl,n magnetization curve to a maximum value correspond- (100 Oe), then the material is typically considered hard. If
mg to + Bm. Hc ~ 400 A-m -1 (5 Oe), then the material is considered
soft. Coercive forces as large as 8 x 105 A-m -1 (10 4 Oe)
and as small as 0.08 A-m -1 (10- 3 Oe) have been observed.
F'~URE J. Magnetic hysteresis curve The fact that an object retains a strong residual field B r is
not necessarily indicative of a high coercive force H c' Some
MAGNETIC FLUX DENSITY materials retain an appreciable residual induction and yet
(LEVEL OF MAGNETIZATION I
are eaSily demagnetized (transformer steels are an exam-
+8
pIe). On the other hand, some materials that retain relative-
ly weak residual fields can be extremely difficult to demag-
1:
GLiJ netize because of a high coercive force.
z~ i VIRGIN
l.U :J r-..;---:-: MAGNETIC

§~ .,:1 CURVE
Basic Principle of Demagnetization
-J Z -H H
~~ _ _~-~no_ _~~_ _-¥~~ ______+-___
~tiJ Practically all demagnetizing methods are based on a
tu~ common procedure. A magnetizing field strength H, suffi-
z«
~~
ciently high to overcome the initial coercive force He, is
!/ -8,
alternately reversed in polarity and gradually reduced to
~--. zero. The diminishing hysteresis curve shown in Fig. 2
illustrates this principle.
-8 The value of the total coercive force He is often unknown
and varies with proximity to the test object's ends. However,
296 / MAGNETIC PARTICLE TESTING

it is always less than the maximum magnetizing force H 11/

FIGURE 2. Diminishing hysteresis curve value. This may serve as a guide in cases where the
magnetizing field strength Hm is known, as is sometimes the
FLUX DENSITY case in magnetic particle testing. Successful demagneti-
+8 zation procedures often start with an initial demagnetiz-
ing field which equals or exceeds that of the magnetizing
force Hm.
An object may also be demagnetized by raising its
temperature above the Curie point. Although this method
provides thorough demagnetization, it is often impractical.
-H FIELD
---------+++~~~------+---H STRENGTH

-8
 DEMAGNETIZATION OF TEST OBJECTS I 297

PART 3
SUMMARY OF DEMAGNETIZATION
PROCEDURES

influence of the coil's field. The coil should not be

Alternating Current deenergized until the test object has reached this pOint,
Demagnetization three or four coil diameters away or about 1 m (3 ft). This
mle applies to most demagnetizing coils in industry, where
peak values of the field at the center of the coil are about
1.2 X 105 to 2.4 X 10,5 A-m - I. Small test objects can be
Through-Coil Method
hand held, placed within the coil and withdrawn. Field
The simplicity of the alternating current through-coil penetration may be several skin depths.
method makes it .one of the most prevalent demagnetization The magnitude of the field can also be reduced by
techniques. The method uses a coil powered from a current withdrawing a coil such as a cable coil away from a stationary
source altemating at line frequency (usually 60 Hz in the test object. This method is advantageous for high produc-
United States). Operating at a fixed amplitude, the coil tion rates since a P!operly designed coil can be continuously
prqduces a continuously reversing magnetic field because of energized while a steady stream of test objects is conveyed
the cyclic nature of tbe current. As a test object is conveyed through the cC?il opening. Typical alternating current
through the coil, it is subjected to the most intense magnetic through-coil demagnetizers are shown in Figs. 3 and 4.
field.while within the confines 'of the coil. Demagnetization of test shims may be performed by
The strength of the field is gradually reduced to zero as removing them slowly from one pole of an alternating
the. object exits the :oil and teaches a point beyond the current yoke.

FIGURE 3. Manually operated alternating current FIGURE 4. Production alternating current coil
coil demagnetizer with roller carriage and demagnetizer incorporated into conveyorized
automatic timer magnetic particle testing unit

FROM MAGNAFLUX CORPORATION. REPRINTED WITH PERMISSION FROM MAGNAFLUX CORPORATION. REPRINTED WITH PERMISSION.
298 / MAGNETIC PARTICLE TESTING

Cable Wrap Method object should remain stationary \\rithin the coil until tIlt'
demagnetizing cycle has been completed. vVhen the contact
There are applications (such as large or immobile test
method is used, contact should be maintained until the cvcle
objects) where a cable wrap, in conjunction with an appro-
is completed. The direct current method provides deep
priate high amperage alternating current power source, can
penetration and is usually very successful on objects that are
provide a convenient means of demagnetization. The high
otherwise difficult to demagnetize.
amperage power source must also incorporate a suitable
Because of economic considerations, the reversing direct
current control to facilitate the gradual reduction of current
current demagnetiZing feature is usually incorporated into
from maximum to zero.
the design of direct current horizontal wet magnetic particle
Portable or mobile alternating current power packs are
testing units or relatively large stationary power packs.
available for such applications. Equipment of this type
usually features stepless solid state current control. Some
units incorporate special current decay circuitry that auto-
matically reduces the current from maximum to zero in a Reversing Cable Wrap Method
matter of three or four seconds. This method is used for demagnetizing objects too large
or heavy to process on a horizontal wet testing unit. The
object to be demagnetized is wrapped with multiple turns of
Through-Current Method high amperage flexible cable (such as 4/0) connected to a
stationary direct current power pack.
Through-current demagnetization is a contact method. The current is alternately reversed in direction and
High amperage alternating current is caused to flow directly reduced in amplitude through multiple steps until the
through a test object starting at contact electrodes. The current reaches zero. This is usually accomplished using
current control system must prOvide gradual reduction of built-in automatic circuitry similar to that deSigned into
current as r.equired for demagnetization. A portable alter- some wet testing units. Power packs that lack an automatic
nating current power pack used for cable wrap demagneti- demagnetizing cycle can sometimes be used to accomplish
zation may also be used for the through-current method. demagnetization by manually interchanging cable connec-
Through-current demagnetization is used on horizontal tions (current reversal) and ma~ipulating manual·current
wet magnetic particle testing units that have comparable controls (current reduction). However, this is til.ue consum-
current control systems. Because of part configuration and ing and requires an appropriately fine current control to he
the accessibility of alternating current, the through-current successful.
'method can have real advantages for ~ertain ap·plications.

Pulsating Reversing Method

Direct Current Demagnetization A high amperage direct current coil demagnetizer has
been designed to produce alternate pulses of positive and
negative current. The pulses are generated at a fixed
Reversing Direct Current Contact Coil Method
amplitude and a repetition rate of five to ten cycles per
Reversing direct current contact coil demagnetization is second. This permits relatively small objects to be demag-
usually associated with relatively large test objects that have netized by the through-COil method.
been magnetized using a direct current magnetic field. It is The object is subjected to a constantly reversing magnetic
also applicable in certain instances where alternating cur- field as it passes through the coil and the magnitude of the.
rent demagnetization proves ineffective. effective field is reduced to zero as the object is gradually
The method requires high amperage direct current or withdrawn from the coil. This mode of operation is identical
full-wave rectified alternating current that can be directed to the altemating current through-COil method described
to a coil or contact plate. There must also be provisions for above, except for the reduced repetition rate (five to ten
reversing the polarity of the current and suitable means for cycles per second compared to sixty cycles per second).
gradually reducing its amplitude to zero. The lower repetition rate substantially reduces the ~kill
The direct current is alternately reversed in polarity effect with a corresponding increase in magnetic field
(direction) and reduced in amplitude to zero. Although penetration. Therefore, the method provides additional
fewer steps may provide satisfactory results, greater reliabil- demagnetizing capabilities on some relatively small test
ity is achieved by using about thirty reversals and current objects with thick cross sections that are difficult to handle
reductions to approach zero asymptotically. with line frequency coil demagnetizers. Another standard
The cycle is usually controlled automatically and requires mode of operation allows the object to remain stationary
about thirty seconds to complete. When using a coil, the test within the coil while the current gradually decays to zero.
 DEMAGNETIZATION OF TEST OBJECTS I 299

when facilities for overall (full length) demagnetization are

Specialized Demagnetization not available.
Procedures One such procedure applies a bucking (opposite polarity)
field at the ends of the object using a direct current coil
positioned near the ends. By trial and error, magnetizing
There is a \\lde valiety of miscellaneous demagnetizing current to the coil is increased until external field readings
techniques that have specific applications designed around a are reduced to an acceptable level. The result is a redistri-
specific test object. bution of the field at each end of the tube. Unfortunately,
mechanical agitation subsequent to this procedure, from
handling or transpOlt, can result in the redistlibution of the
Demagnetization with Yokes
longitudinal field and can reestablish strong leakage fields at
A yoke confIguration is sometimes used to demagnetize the ends of the test object.
small objects having high coercive force. Yokes are also used Another specialized demagnetization procedure uses a
for demagnetization of objects with smalllength-to-diameter full length internal conductor in conjunction with a unidi-
(LID) ratios, including spheres with near-perfect LID ratios rectional current pulse to reorient the longitudinal field and
of 1. Yokes are usually designed for a specific type or range form a circular field. The magnitude of the circularly
of test objects. magnetizing current is increased until acceptable external
Some alternating current yokes are similar in operation to field readings are achieved. While this is not demagnetiza-
the altemating current coil method, where the test object is tion, it does reduce leakage fields to a level suitable for butt
passed between or across the surfaces of the pole faces welding operations. It also creates the undesirable condi-
(maximum field strength) and then withdrawn. Direct tions associated with circular magnetization and hinders
" current yokes a~e sometimes used with the reversing direct subsequent operations, including weld preparation or
current method on specific objects requiring deep penetra- threading ...
. !ion. Howeyer, this is an inhetently slow process because of . Overall demagnetization is the best solution for long
the associated high inductance of the circuit. tubes and this is usually carried out using some form of
. Some coil or yoke arrangements use damped oscillation to reversing direCt current. For static conditions, an internal
obtain the required reversing and diminishing field. The conductor is used in conjunction with a conventional revers-
oscillation is delived from a special circuit design based on ing direct current cycle. A cable, wrapped to give several
specific values of capacitance, inductance and resistance. internal conductors, could be used in place of the central
conductor. However, when object' s1ength is taken into
Demagnetization of Oil Field Tubes consideration, the central conductor approach is much more
time efficient.
Valious testing processes involving direct current magne- The advent of pulsating direct current equipment has
tization leave the test object with a significant longitudinal lead to dynamic demagnetizing systems for overall demag-
residual field. Demagnetization is usually required to elim- netization of long test objects. In such systems, the object is
inate any possibility of the field having an adverse effect on conveyed through a special coil arrangement at speeds up to
subsequent machining or welding operations. Some tubular 1.5 m per second (300 ft per minute). The coil current
products have lengths of 9 to 15 m (30 to 50 ft) and this consists of constant amplitude pulses of direct current
presents a formidable challenge to standard demagnetiza- altemating relative to polarity (positive and negative) at a
tion procedures. Some unusual methods have been used rate of about five cycles per second.
300 I MAGNETIC PARTICLE TESTING

PART 4
SELECTING A DEMAGNETIZATION
PROCEDURE

Selecting a suitable demagnetization procedure is a mat-

ter of matching capabilities with the specific application. Maximum Effective Field Strength
Each of the procedures discussed above has advantages and
limitations based on test object size, hardness, production The maximum effective field strength Hill of a pmticular
rate and source of magnetization (see Table 1). demagnetization procedure is an important indicator of the
procedure's capability. BaSically, the Hill value must be
sufficient to overcome the coercive force He associated \-vith
limitations of Alternating Current the material to be demagnetized. Generally, He increases
with hardness.
Methods Test object size and geometIy (including the LID ratio)
are also factors that influence field strength requirements.
The alternating current through-coil method is probably The demagnetization potential of a procedure inherently
the most widely used demagnetization technique because of increases with field strength capabilities. \Vhen related t~
its simplicity and its adaptability to high production rates. various alternating current coil sizes with equal ampere-tum
However, the skin effect associated with alternating current ratings, the field strength increases as coil size decreases.
magnetic fields results in limited penetration capabilities.
Consequently, alternating current demagnetization meth- .
ods may be ineffective for removing deep residual fields Reversing Direct Current
such ~s th9se associated with direct current magnetization .. . . . .
.
This becomes a significant factor on large test objects.· As Reversing direct current procedures are usually required
a rule, alternating current methods can effectively demag- when alternating current techniques prove inadequate. For
netize most objects with a cross section less than 50 mm those difficult applications, reversing direct current has
(2 in.), regardless of the original magnetizing source. Larger unexcelled demagnetization capabilities. Almost any object
test objects with deep residual fields usually require some can be demagnetized to an acceptable level with reversing
form of reversing direct current demagnetization. If alter- direct current. While production rate capabilities are limit-
nating current is the only source of magnetization, alternat- ed, this is usually not a significant hindrance; since larger
ing current demagnetization is effective regardless of test objects are often involved, the associated production rates
object size. are relatively low.
 DEMAGNETIZATION OF TEST OBJECTS 1301

TABLE 1. Demagnetizing method selection guide based on test object size, hardness and production rate for
residual fields from direct current magnetizing source
Test Object Size Material Hardness Production Rate

Small' Medium 2
Large 3
Soft Medium Hard Low Medium High

Alternating Current (50/60 Hz)

Through-coil A 0 N A A N A A A
Cable wrap N 0 Q A A N A N N
Through-current (30 point step down) N A N A A N A Q N
Through-current (current decay) N A N A A N A 0 N
Yoke A N N A A 0 A N N

Reversing Direct Current

Through-coil (pulsating) A A A A A A A A A
Coil (30 point step down) A A 0 A A A A A N
Cable wrap (30 point step down) N 0 A A A A A N N
Through-current (30 point step down) N A A A A A A Q N
I. HAND HELD
2. LESS THAN 50 mm 12 In., DIAMETER
·3. MORE THAN 50 mm /2 In.1 DIAMETER

LEGEND
o = OUESTIONABLE
A = APPLICABLE
N = NOT APPLICABLE
302 I MAGNETIC PARTICLE TESTING

PART 5
MEASURING EXTERNAL FIELD STRENGTH

In SI units, magnetic flux density B is measured in tesla north pole of the magnet. In the absence of external fields,
or webers per square meter. In cgs units, magnetic flux the pointer indicates zero on the graduated scale.
density is measured in gauss or as magnetic lines of force 'When the magnetic north pole of a residually magnetized
per square centimeter. One tesla is equivalent to 10 4 gauss. object is moved close to the pivot end of the pointer, the
In SI, the magnetic field strength H is measured in amperes south pole of the vane is attracted toward the object and the
per meter (A-m -1); in cgs, the unit is oersted (Oe). One n01th pole of the vane is repelled. The resulting torque
ampere per meter is equivalent to 0.013 oersted. causes the pointer to rotate in the positive or plus direction.
Flux density and field strength are related to each other The restraining torque is prOvided by the tendency of the
and to magnetic permeability by the following equation: vane to remain aligned with the leakage field of the
permanent magnet.
(Eq. 1) Sensitivity of the field indicator is primarily a function of
the permanent magnet's characteristics and the spacing
Where: between the magnet and the vane. The device can be
calibrated in relative units, milliteslas or gauss. Calibration is
f.L(J = the permeability of free space; and only valid when the device is subjected to a uniform
f.L = the relative permeability.
The value of permeability is not affected by the 'choice of
units. Permeability values in customary units are the same FIGURE 5. Typical magnetic field indicator
magnitude as those qf relative permeabilities in Si units. construction .
Because gauss and oersted have by definition the same
magnitude in air, the f.Lo factor is not used in the equation
for customary u n i t s . '
By itself, the permeability of a material has no meaning.
Its value at speCific values of H or B does have meaning.

Field Indicators
Theory of Operation
The field indicator is a hand held instrument used to
measure the relative strength of magnetic leakage fields.
The construction of a typical field indicator is shown in
Fig. 5 and its the01Y of operation is relatively Simple. As
illustrated, an elliptically shaped soft iron vane is attached to PERMANENT MAGNET
a pointer that is free to pivot. A rectangular permanent
magnet is mounted in a fixed position directly above the soft
iron vane.
Because the vane is under the influence of the magnet, it
tends to align its long axis in the direction of the leakage
field emanating from the magnet. In so dOing, the vane
becomes magnetized and has a magnetic pole induced at
each end of its long axis.
On the end of the vane below the south pole of the
magnet, a north pole is induced. Correspondingly, a south
pole is induced at the opposite end of the vane below the
 DEMAGNETIZATION OF TEST OBJECTS / 303

magnetic field snch as that fonnd at the center of a large coil by a flexible multiconductor elechical cable. This permits
or Helmholtz coil arrangement. leakage field measurements light at the surface of a test
The pattern of external fields emanating from a residually object, and consequently, the readings are substantially
magnetized object is anything but uniform and varies with greater compared to those obtained with a field indicator.
changes in pmt geometry. In view of this, readings obtained In the past, use of the gauss meter was confined to
with ~ field indicator are effectively in relative nnits but still laboratory work because of instmment cost, complicated
usefill for comparative purposes. operating procedures and the delicate nature of the probe.
However, gaussmeters are now widely used in many ways,
Use of the Field Indicator such as qualifying demagnetization procedures for small or
unusual test objects; establishing maximum permissible
The relative strength of an extemal field is measured by field indicator readings; and verifYing field indicator calibra-
bIingiIlg the field indicator near the object and noting the tion procedures.
deflection of the pOinter. The edge of the field indicator at
the pivot end of the pointer should be closest to the object
under investigation.
The required degree of demagnetization is usually spec- Measuring Techniques
ified as a maximum field indicator reading (in terms of
. divisions, tesla or gauss) for a specific device. Leakage Leakage field measurements are undertaken to ascertain
field patterns of very small objects may be too limited for the level of residual magnetic fields emanating from an
detection by a field indicator - the sensing element (soft object. To increase the accuracy and repeatability of such
iron vane) is located approximately 12 mm (0.5 in.) from the measurements, it is good practice to isolate the object
edge of the device's casing. magnetically to eliminate the influence of extraneous mag-
The field indicator is the most common device used in netic fields or other ferromagnetic material. This can. be
. industry for mon~tOling the effectiveness of demagn~tizing . accomplished by moving the test object to an appropIjate
. processes. It is convenient, easy to operate and inexpensive. area and supporting it manually or by placing it on a suitable
nonferromagnetic platform.
Caution should be exercised to keep certain types of field
Laboratory 'nstruments indicators at a safe distance from demagnetizing fields. Such
exposure can partially demagnetize the internal permanent
The Hall effect gaussmeter is the most common labora- magnet and effectively change the calibration of the device.
tory instmment used to measure the strength of leakage If this occurs, the sensitivity of the device increases substan-
fields quantitatively. The sensing element is located in a tially. In strong fields, remagnetization of the device's in-
remote hand held probe connected to the basic instrument ternal magnet may occur, rendering it unfit for future use.
304 I MAGNETIC PARTICLE TESTING

PART 6
TYPICAL DEMAGNETIZATION PROBLEMS

Demagnetizing Field Strength Poor Length-to-Diameter Ratio

Too Weak
The ratio of an object's length to its diameter (the LID
Demagnetizing capability is directly related to the tech- ratio) is as important to coil demagnetization as it is to coil
nique's magnetic field strength. Higher field intensities magnetization. Values below 3:1 can be a problem, but the
increase the variety of objects that can be successfully situation may be corrected by effectively increasing the LID
demagnetized. ratio.
Conventional alternating current coils are usually rated in This correction is done by adding ferromagnetic pole
terms of ampere-turns. However, for the same applied pieces at both ends of the object. Pole pieces should be
ampere-turns, the strength of the field increases as coil about 1.50 mm (6 in.) long and nearly the same diameter as
diameter decreases. It should be noted that the field the object. \Vith pole pieces in place, the assembly can be
strength within a coil is greater near the inside wall than at passed through the demagnetizing coil in the usual manner.
the center. For large lots, it is sometimes convenient to pass the objects
through the demagnetizing coil end to end in a chain.

Alternating Current
Demagnetization and Direct MaQnetic Shi~lding
Current Magnetization Whenever possible, it is desirable to demagnetize an
object before its assembly with other components. Once
Objects magnetized by a direct current source can be assembled, the test object may be adjacent to or surrounded
difficult if not impo"ssible to adequately demagnetize with an by other ferromagnetic material. In such cases, the demag-
alternating current process. This is more prevalent in test netizing field may be shunted through the adjacent mateliaJ
objects with diameters greater than 50 mm (2 in.) because rather than through. the object itself and demagnetization
of the pronounced skin effect associated with alternating will be ineffective.
current fields. Small objects should not be passed through a demagne-
Deep interior residual magnetism remains unaffected by tizing coil in bundles or layered in handling baskets. Test
surface oriented alternating current fields. On such objects, objects in the center of a stack are shielded from the
some form of reversing direct current demagnetization must demagnetizing field by the outer layer of ~laterial. Ferro-
be used. magnetic baskets or trays are objectionable for the same
reason.

Test Object Orientation

Improper test object orientation relative to an alternating Calibration of Field Indicators
current demagnetizing coil can have an adverse effect on
the demagnetization process. For the best results, a test Field indicators are convenient devices for monitOling the
object should he passed through the coil lengthwise or with effectiveness of a demagnetization process but ohs~tlre
its longest axis parallel to the coil's central axis. changcs in calibration can cause problems vvben verifying
Ring shaped test objects such as bearing races can be the demagnetization procedure.
rolled through an alternating current coil to obtain the Inadveliently subjecting the device to a demagnetizing
desired results. In a similar manner, objects with a complex field can pmiially demagnetize the field indicator's intern,~1
configuration may require some form of rotation as they are permanent magnet and significantly increase the de\ice s
passed through a demagnetizing coil. sensitivity. Subsequent measurements would then indicate
 DEMAGNETIZATION OF TEST OBJECTS I 305

the erroneous conclusion that an established demagnetiza-

example, objects exhibiting a field indicator reading around
tion process is inadequate. In such cases, the results of
0.3 mT (3 G) usually will not have an adverse effect on
demagnetization should be verified with another device
known to be accurate. subsequent machining or welding, instrumentation or end
use.
Strictly specified limitations should be placed only on a
Demagnetization Specifications limited number of demagnetization applications (particular
test objects manufactured for specific service) as required
Stringent demagnetization specifications can be a source by experience or laboratory data.
of selious production problems. Specifications are often
To eliminate ambigUity, a speCification should define how
wlitten without accommodation to the practical limitations
a reading is made - on the surface with a Hall effect meter
of the demagnetization process and its verification. For or with a field indicator of a specific deSign.
 306 I MAGNETIC PARTICLE TESTING

PART 7
DEMAGNETIZING EQUIPMENT

offered as an option on direct current magnetization equip-

Summary of Alternating Current ment but separate units are also available. Some typical
Demagnetization Equipment units are listed below.

Alternating current coil denwgnetizers function only as

Coil denwgnetizers are designed for demagnetization
demagnetizers. They are designed for operation at line
only. A constant amplitude, high amperage direct current is
voltage and line frequency. Most of the units are rated for
directed through a fixed coil. The polarity of the current is
intermittent duty but some are available for co'ntinuous
continuously reversed at a fixed rate. Relatively small test
duty. A variety of coil openings are commercially available.
objects may be demagnetized dynamically \vith the through-
Portable alternating current half-wave units are light-
coil method or statically with modified current decay tech-
weight magnetization and demagnetization units used to
niques. Coil sizes may be limited.
power high amperage cables. Solid state steples's current
control permits demagnetization by coil, cable wrap or
contact methods. Limited alternating current output usually Horizontal wet nwgnetization and denwgnetization units
restricts application of these systems to relatively small or permit objects to be magnetized, tested and demagnetized
mediuI1) sized objects. in place ·on the system. Demagnetization is based on
Mobile alternating current half-wave units are heavy automatic polarity reversal and a step down cycle using coil
. duty, dual output s.ystems used in cpnjunction with high or contact methods. These systems are suitable for relatively
amperage cables. Newer designs have solid state current large test objects.
control and built-in current decay circuitry. Relatively large
test objects can be demagnetized by the coil, cable wrap or
contact methods. Stationary nwgnetization and denwgnetization pou:er
The horizontal wet alternating current units \vith solid packs produce high amperage directed to the test object
state current control feature current decay circuitry that can through flexible high amperage cables. Either the contact or
be used for demagnetizing either by the coil method or the cable wrap method can be used to demagnetize large
contact method. objects. The reversing direct current demagnetization pro-
cess is automatically controlled through a fixed cycle.

Summary of Reversing Direct Specially designed systems are available for applica-
Current Demagnetization Equipment tions that require high rates of produdivity. SpeCial power
pack and coil configurations are available to demagnetize
Equipment for reversing direct current demagnetization 9 to 15 m (30 to 50 ft) lengths of tubing on a continuous
is often expensive. The reversing feature is commonly basis at rates up to 1.5 m per second (300 ft per minute).
 DEMAGNETIZATION OF TEST OBJECTS / 307

PART 8
DEMAGNETIZATION OF ELONGATED
TEST OBJECTS

:M lIch flnished or semifinished tubular product is tested In some testing specifications, it is also reqUired that the
for surface breaking tight discontinuities by the magnetic external longitudinal magnetic fields close to the ends of the
flux leakage me.thod. In the case of ferromagnetic oil fIeld tubular product (especially oil field transmission line pipe)
tubes, for example, the ends of the tube might first be tested be reduced to a value below a specified minimum follOWing
for transversely oriented discontinuities, with the magnetic magnetic flux leakage testing. 1 This is generally done so that
flux being induced by placing the end of the tube in a coil. such materials can more eaSily be field welded.
The tube is then tested for longitudinal discontinuities using Typically a specification might require that not more than
circumferential induction resulting frpm a current pulse 800 A-m- 1 (10 Oe) be measured with a gaussmeter at the
along an internally placed conductor connected to a capac- ends of the material. After magnetic flux leakage testing for
itor discharge system. Both of these methods are shown in transversely oriented discontinuities, the magnetic field
Fig. 5. intensities emerging from the ends of tubes might easily be
40 kA-m -1 (500 Oe) and it is essential. that some form of
deI!1agnetization be applied.
FIGURE 6. Magnetization of elongated products In the text below, commorily used fOr'ms.of demagnetiza-
as commor;tly practiced for oil field tubular tests': tion or remagnetization are ·outlined.
fa) circumfetential magnetization of a tube using
the capacitor diSCharge internal conductor .
method; and fb) longitudinal magnetization of Circumferential Remagnetization
an end region using an encircling coil
In some cases, after longitudinal magnetization has been
established and transverse discontinuity testing completed,
la) CAPACITOR DISCHARGE UNIT
it is necessary to remagnetize only in the circumferential
direction. The resulting magnetization then causes little or
no external field. This is easy to accomplish in tubes by using
C ~ the internal conductor method 2 shown in Fig. 6a. The

FIGURE 7. Hall element gaussmeter measuring

longitudinal magnetic field in air close to the
end of a tube

TUBE

Ib)

GAUSSMETER

tI\j~9J}1
308 / MAGNETIC PARTICLE TESTING

remaining longitudinal field strength can be measured easily tube is not circumferentially magnetized.·3 It merely incli-
with a gaussmeter as shown in Fig. 7. cates that the circumferential and axial flux densities in-
In this test, a calibrated gaussmeter is used to measure volved obey a vector relationship such as that shown in
the maximum value of the magnetic field strength at the end Fig.8.
of the material. If it exceeds a specified value, additional In this particular example, the material just inside the end
shots from the capacitor discharge internal conductor are of the tube is saturated with magnetic flux that is virtually
applied to reduce this emergent field. For this application, circumferential at a value very close to the remanence
a gaussmeter capable of reading up to 100 mT (1,000 G) is Br for the steel. Suppose that an external flux denSity
preferred over other types of field indicator. of 1 mT (10 G) occurs for a material with a remanence of
1,400 mT (14 kG). This flux denSity measurement was made
Angled Nature of the Resulting Field with a gaussmeter after attempted circumferential mag-
netization by the method outlined above. Its value incli-
The presence of some small axial field strength at the end cates that the direction of the flux inside the material is
of a tube (or other elongated part) does not indicate that the only tan -1 (111,400) or 0.04 degrees from being perfectly
circumferential.

FIGURE 8. Possible magnetic flux density Drill Pipe End Area Tests
configuration at the end of a tube: (a) the
emergent flux density is measurable with a A commonly practiced magnetic fllLX leakage test is that
gaussmeter but the circumferential component is of longitudinally magnetizing the ends of drill pipe (Fig. 6b)
not; and (b) the resulting field within the tube In and looking for thread root cracks with wet fluorescent
this case Is almost circumferential magnetic particle techniques. Figure 9 illustrates an exam-
ple of pin thread stretching, the roots of these threads heing
(a) particularly prone to cracking. 4 Following such a test, the
rotation of the flux into the circumferential direction might
require several pulses from a capacitor discharge internal
conductor system. Unde.r such circumstances, longer pulses
(25 Ins) are more effective than shorter ones (3 ms). .
In the resulting circumferential magnetization, the drill
stem can be tested for longitudinally oriented discontinuities'
and the tool joints can be tested for heat-check cracks \\lith·
(b) LONGITUDINAL FIELD magnetic particle tests. The threads are also relatively easy
1 mT 110 G) to clean of particles.

FIGURE 9. Pin stretch on used drill pipe threads;

wet magnetic particle tests are often used to
o detect fatigue cracking on these thread roots
8
<T'

f-
E
o
~.

FROM EXXON CORPORATION. REPRINTED WITH PERMISSION.

"""""""11""""""1"'11'11'1'1'1'1"11"'""""1"""""""","""""""1'1"
DEMAGNETIZATION OF TEST OBJECTS I 309

upset Area Problems the surface of the test object; (2) the thickness of the
material; and (3) the matelial's B-H properties.
It is generally necessary. to apply only ?2 ~o ~ kA-m- 1
In traditional demagnetization with alternating current, it
(40 to .50 Oe) circumferentlally to m~gnetIze.Ol~ fIeld tt~bu­ is required that the test object be slowly removed axially
Jar materials to a point near saturatIOn. ThIs IS .especIal!y from the demagnetizing coil and that the strength of the
tme of the region between the ends of the matenal (QR III field be sufficient to produce saturation in the direction
Fig. 10). Some matelials may require highe.r field streng~h opposed to the existing magnetization direction.
values to cause saturation (the B-H curve for the matenal Because of the skin effect, the field strength at any point
should be consulted). within the matelial is lower than that at the surface. If
However, plior to threading the ends of such tubes, it is suffiCiently high fields are to penetrate the material to cause
sometimes necessaIY to bell or upset the ends (J'Q and RS) demagnetization at that given point, the surface field am-
of the material. This process may lock in existing magnetic plitude must be several times higher than that reqUired to
fluxes from earlier tests, the result being that the flux cause saturation. At five skin depths (about 5 mm) one
density at the ends can require a higher demagneti~ing percent of the surface field remains. It is clear that the
circumferential field than is required for the central regIOns alternating current technique is confined to relatively thin
of the tube. This is due to ineffective or incomplete stress or thin walled test objects and that some other form of
relief at the ends after upsetting, which also causes the B-H demagnetization is needed for thicker materials.
properties to differ from those of the central regions.
When this situation is encountered, a gaussmeter can
indicate the number of pulses required to rotate the flux
within the upset areas. FIGURE 11. The axial bulk flux density In a
10 m (30 ft) long tube: (a) after longitudinal
magnetization to saturation; and (bJ after direct
current coil demagnetization using a constant
Alternating Current Coi~ current; 'note the remanent flux density (1.14 T),
the opposed dipole effect caused by the
Demagnetization . demagnetization process, and the values of the
flux densities in the tube after this form of
After tubes have been longitudinally magnetized to satu- demagnetization
ration, ,lowe ling the remaining induction is usually done
with alternating current or direct current through a central-
ized, surrounding coil. With the alternating current tech-
nique, it is clitical to know that, for a frequency of 60 Hz, (a)
the skin depth being demagnetized is only about 1 mm
(0.04 in.). The effect of this form of demagnetization is
dependent on: (1) the strength of the demagnetizing field at

FIGURE 10. Different circumferential magnetic

field 8-H curves: (a) at the belled or upset ends
of tubes; and (b) the unworked center section of
a tube; curve (a) shows lower Br and higher He
Ib)

~I~T
than curve (b)

P Q R S
>-
f-

~
UJ
I
N J S
0
'- ./S + N
3
fa)
LL.

0 3
JST 1
6
I
10
(10) (20) (30)

DISTANCE ALONG PIPE

meters (feet)

AFTER LORD (SEE REFERENCE 51. REPRINTED WITH PERMISSION.

310 I MAGNETIC PARTICLE TESTING

provided to show that lower bulk fluxes can be obtained by

Direct Current Coil Demagnetization sensing the remaining flux and adjusting the demagnetiz<l-
tion current to make the flux as small as pOSSible.
Partial demagnetization of tubes is often accomplished by A commonly used method for sensing the flux density
passing the tube through a coil in which the field strength H level in an elongated object is to jerk a loosely fitting coil
opposes the direction of the residual induction in the from a position of flux linkage with the pmt to a position of
material. Demagnetization current settings are obtained by no flux linkage. This is shown in Fig. 12. "Vith the coil in
trial and error since only the external fields at the ends of position-I, the flux linkage with the magnetized object is
the tube are available to the inspector. Currently accepted given by:
practice is to adjust the demagnetizing field strength to
provide the minimum fields at either end of the test object. Net> = N x J 13 -n dA (Eq.2)
As shown in Fig. 11, this process leaves the flux density Where:
within the material at a relatively high level.
The actual value of the axial flux density in the material is Net> the flux linkage (weber-tums);
=
governed by two factors. The first is the local magnetic N =
the number of tums on the coil;
property or, in effect, just how high a value of flux density B 13 = the flux density of the test object in the coil
can be sustained. The second is the proximity of the ends of (webers per square meter or tesla)~ and
the material (this affects the flux density through the ndA = the area of the coil perpendicular to B
demagnetizing field). The latter factor is very important (square meters).
within about 1 m of the tube ends and has a considerable
effect on the local B-H properties. The general effect is to With the coil in position-2, the_flu~ linkage is zero and the
lower the local permeability, due to the internal demagne- net change in flux linkage is NJB - ndA. When the voltage
tizing field inside the test object. induced in the coil during this process is integrated with a
As an example, the effect of direct current coil demag- fluxmeter, the output (suitably compensated for the value of
netization on the axial flux density in a typical 10·m (30 ft) A and N) is the average value of the flux density in the test
tube is shown in Fig. 11. As illustrated in Fig. lla, the axial object directly inside the coil at position-I.
flux density of the tube after magneti~ation is relatjvely Applying this theory to the demagnetization process, the
constant at 1.14 T (11.4 kG) except within about 1 m of test object is passed through a s'ensing coil and the output of
either end. This value is close to the remanence for the the fluxmeter is used to electronically compensate the
material, which varies with localized stress, chemical com- curr~nt through the demagnetizing coil. This compensation
position and other factors. produces a net flux density in the sensing coil, and therefore
After demagnetization, in which only the external field is in the object, as close to zero as pOSSible. This is shmvll
sampled at both ends of the tube, the remaining flux density schematically in Fig. 13.
within the material is as shown in Fig. 11b. The flux density
for the majority of the material is relatively low and in the FIGURE 12. Measurement of the flux at a specific
same direction as the saturated state. However, at one end pOSition close to the end of an elongated test
of the tube the direction of the flux has been reversed by the object; a search coil is taken from position-1 to
coil field for roughly 1.4 m (4.4 ft). The demagnetized position-2 and the resulting voltage is integrated
material then exists in an opposing dipole configuration, within a fluxmeter; the fluxmeter output is
with poles as shown in the illustration. proportional to the flux at position-l
If this form of demagnetization is used, the inspector
must consider each application individually. The direct

_~~I
current required depends on the wall thickness of the tube.
Passage of the material through an altemating current coil, lfJ=O
perhaps of low frequency, further lowers the contained flux.

"
L~-:.:::::::. ===:=:::.:::::. -:.:::.
Flux Sensed Demagnetization
Traditional altemating current and direct current demag-
netization techniques do not totally demagnetize an object. FLUXMETER
They merely reduce the bulk flux density to the value at
which emergent fields are low enough not to hinder subse-
quent metalworking processes. The follOwing information is
 DEMAGNETIZATION OF TEST OBJECTS / 311

FIGURE 13. Diagram for optimizing the direct of the skin effect, surface demagnetization by the alternat-
current demagnetization of an elongated test ing current method is at best a temporal), measure so that
object; the sensing coil develops voltage as a the bulk magnetization of the test object may eventually
magnetized materral passes through it and the lead to large extemal fields.
voltage is fed to a fluxmeter; the meter's signal The alternative practice is that of local application of a
is proportional to the flux density within the direct current field which, while it may remove the flux from
test object and is used to control the a small region of a test object, does not leave the entire part
programmable power supply of the demagnetized. Unless the entire object can be demagne-
demagnetizing coil tized totally, presently accepted practice will continue: the
f1ux is reduced in certain regions of the object so that
DIRECT externally applied field" indicators show a low field. The end
CURRENT user will continue to accept material that is highly magne-
MAGNETIZING
COil tized even though it appears to be demagnetized. No
specifications appear to exist for demagnetization proce-
dures that lead to bulk flux reduction in previously magne-
tized objects. Below are examples of situations where the
lack of demagnetization can cause problems at a later date.
A common practice in the testing of oil field tubes is to
reduce the bulk flux to. the condition shown in Fig. II b.
Subsequent transportation and other mechanical vibrations
FWXMETER can cause the external field to rise to levels unacceptable to
the end user. Such lev~ls cannot be caused by the Earth's .
magnetic fi~ld (only f?02 tnT or 0.2 G) because the field
strength required to magnetize these materials is much
larger than 0.02 mT. .
A second common practice is to pass a longitudinally
magnetized tube or rod through an alternating current coil
Problems Associated with Partial carrying 50 or 60 Hz current. Skin depth considerations
Demagnetization reveal the volume of material actually demagnetized. The
material is left in a highly magnetized state, apart from a
Partial demagnetization can lead to field problems when surface layer. Subsequent handling causes the reappearance
the material later appears to be highly magnetized. Because of high external fields.
312 I MAGNETIC PARTICLE TESTING

REFERENCES

1. Electromagnetic Testing: Eddy Current, Flux Leakage 3. Oilfield Magnetism and the Mythology 'Which Sur-
and Microwave Nondestructive Testing. The Nonde- rounds It. Chapter 6. Houston, TX: International Pipe
structive Testing Handbook, second edition. Vol. 4. R. Inspectors Association.
McMaster, P. McIntire and M. Mester, eds. Columbus, 4. Moyer, M. and B. Dale. "A Test Program for the
OH: The American Society for Nondestructive Testing Evaluation of Oil Thread Protection." Journal of Petro-
(1986). leum Technology. Richardson, TX: Society of Petroleum
2. Stanley, R. and G. Moake. "Inspecting Oil Country Engineers (February 1985): p 306.
Tubular Goods Using Capacitor Discharge Systems."
Materials Evaluation. Vol. 41, No.7. Columbus, OH: 5. Electrornagnetic Methods of Nondestructive .Testing. \V.
The American Society for Nondestructive Testing Lord, ed. New York, NY: Gordon and Breach (198.5):
(1983): p 779-782. p 97-160.
 SECTION

DEMAGNETIZATION OF
TEST OBJECTS
Kenneth Schroeder, Schroeder and Associates, Arlington Heights, illinois
Roderic Stanley, International Pipe Inspectors Association, Houston, Texas
Lawrence Wong, Magnaflux Corpo!atiC?n, Chicago, Illinois
292 / MAGNETIC PARTICLE TESTING

PART J
DEMAGNETIZATION AND RESIDUAL
MAGNETISM

Demagnetization is the process of removing magnetism called leakage field'} and are the phenomena responSible for
from a ferromagnetic material. This is accomplished on the attracting nearby magnetic particles.
molecular level by establishing or reestablishing random
orientations in the material's magnetic domains.
The demagnetizing process is often required in conjunc-
tion with magnetic particle testing. However, demagnetiza-
Requirement for Demagnetization
tion may also be required for reasons other than magnetiC
Components that retain a relatively strong residual flux
particle tests.
density can be a source of problems ~uring subsequent
manufactUring processes or during service. Some typical
problems are detailed below.
Ferroma.gnetic Materials When subsequent machining is performed on a test
object, the presence of a strong residual flux denSity can
Ferromagnetic materials are characterized by a relative attract and hold chips or particles. This ca~ adversely affect
eas~ a
of magnetization when exposed to magnetizing force. surface finish or tool life. . .
. This can be attributed to the relatively high magnetic In an arc welding operation, the pI:esence of strong
permeabilityJ.L exhibited by these materials. Once magne- leakage fields can deflect the arc away from its intended
tized, ferromagnetic material retains a level of magnetic flux location. This phenomenon is sometimes called arc blow.
denSity B after the magnetizing field strength H has been Electron beam welding is also adversely affected by residual
removed. This remanent flux density is commonly referred magnetism. Even leakage fields of moderate strength can
to as residual magnetism or the residual field. The magni- deflect the electron beam away from its intended target.
tude of residual flux denSity is. a function of the following Plating quality also can be affected by residual magne-
factors: (1) the magnetic characteristics of the material; tism. In an operation such as chrome plating on a steel
(2) the immediate history of the material's magnetization; surface, the presence of a strong residual magnetic field can
(3) the strength of the applied magnetizing field; (4) the divert the plating current away from its intended location.
direction of magnetization (longitudinal or circular); and Such mishaps are costly, incorporating the expenses of the
(5) the test object's geometry. first faulty plating, possible repair and a second plating
Residually magnetized materials contain magnetic flux <l> operation.
measured in webers. In the International System of Units In-selvice operation of an object may be impaired if it is
(SI), 1 weber (\Vb) passing through 1 m 2 gives a flux denSity the source of excessive leakage fields, attracting metallic
of 1 tcsla (T). In the centimeter-gram-second (cgs) system, chips or tramp paIticles. This condition can cause malfunc-
flux is expressed in maxwells or lines of flux. Flux denSity tions in rotating assemblies such as bealings or excessive
is expressed as lines per square centimeter or gauss (G). wear on bealing surfaces. Strong residual flux densities also
One weber is equivalent to 10 8 lines and one tesla equals may cause a substantial magnetic attraction between adja-
104 gauss. cent moving parts. This can produce increased fIiction or.
Depending on the direction of the flux with respect to the may ill other ways interfere with the intended function of
surface of the test object, these lines may leave the object at the component.
one point and continue their path through air before Cleaning operations for the removal of metallic chips and
reentering at another point. The points where magnetic particles can be hampered by the presence of residual
lines of flux leave and reenter a ferromagnetic material are fields. This can he of speCial concern when cleaning ohjects
called magnetic poles. The north pole is where flux lines that have internal openings such as oil passages.
leave the material. Flnx lines reenter the material at the Residual leakage flelds can have a direct adverse effect on
south pole. Such fields emanating hom a test object are certain types of instnunentation. The magnetic compass
 DEMAGNETIZATION OF TEST OBJECTS I 293

aboard aircraft is an example, as are many electronic Residual magnetization can occur with improper opera-
components. tion of alternating current through a coil demagnetizer. This
Rotating shafts containing residual magnetism may act as occurs when the coil current is terminated and the test
electric generators in conjunction with the Earth's magnetic object is still within the coil or within its influence. Pene-
field or other stray fields. The generated electricity can heat tration depth is determined by the frequency of the alter-
the shafts, causing operating losses and other service prob- nating current.
lems. Finally, under certain conditions the Earth's magnetic
Finally, objects sometimes need to be demagnetized field can impart a longitudinal residual field. This occurs
before using established magnetizing current levels for only with low coercivity steels, usually when the object is
magnetic p,-uticle testing. The presence of strong residual shocked or vibrated while its long axis is parallel to the
induction can lead to erroneous results and faulty test Earth's magnetic field. Such a residual field may become
procedures. quite significant for long parts subjected to severe vibrations
Demagnetization is not required when the test material in service or transport. The intensity (horizontal compo-
exhibits very low magnetic remanence. Low carbon or nent) of the Earth's field in the United States is about
nodular steels are included in this category. If the next 16 A-m -1 (0.2 Oe). This is equivalent to the magnetic field
manuf~lCtUling process calls for the object to be heated strength at the center of a five ampere-tum coil 300 mm
above the Curie point, the matelial becomes nonmagnetic (12 in.) in diameter.
and demagnetization is accomplished by the heating pro-
cess. At the Curie point, steel temporarily transforms from
a ferromagnetic to a paramagnetic state and subsequently Effect of Magnetic Field Origin on
cools with zero net induction. Demagnetization
If the pmt does not require additional machining and its The type of magnetizing source becomes significant when
intended function is not compromised by the presence ofa conSidering the nature of residual induction and its subse~
residual field, then demagnetization becomes unnecessary. quent demagnetization.· .
Alternating currents tend to flow near the surface of a
conductor and this so-called skin effect produces residual .
Common Sources of Residual fields that are sUlface oriented. Such fields respond well to
Magnetic Fields alternating current demagnetization techniques ..
There is practically no skin effect associated with a direct
Ferromagnetic .material can become magnetized in a current magnetizing source and consequently the entire
numl~er of ways, intentionally and unintentionally. cross section of an object can be residually magnetized.
Parts may be purposely magnetized to perform magnetic Deep seated residual fields in larger objects may not be
particle tests or to facilitate a magnetic flux leakage test. affected by alternating current demagnetiZing techniques
Magnetization may also come from biasing magnetic fields because the skin depth is only about 1 mm for steel of
such as those used to negate the effects of permeability relative permeability 100 at 60 Hz.
c.hanges in eddy current tests.
Magnetic chucks are a common source of residual fields.
Pronounced fields can be left in an object if a chuck's
built-in demagnetizing cycle is faulty or inadequate. Prema-
turely removing test objects before completion of the de- Types of Residual Magnetic Fields
magnetizing cycle can also be a source of residual magnetism.
Lift magnets are convenient material handling devices,
Longitudinal Magnetic Fields
but because they operate with direct current, they can leave
strong residual fields in fabricated objects or in raw mate- MateIials magnetized by a coil or solenoid sometimes can
rial. Physical contact with any permanent magnet or highly be left with a longitudinal residual induction. The field is
magnetized object (a machine table or fixture) can also oriented lengthwise in the test object and there is a high
establish a residual field. concentration of emergent fields at each end. These fields
Low hequenc), induction heating that is abntptly termi- constitute poles, lines of magnetic flux entering or leaving
nated can induce very strong residual flux density. The the material.
lower the frequency, the deeper the magnetizing field Longitudinal magnetization is easily detected by field
strength penetrates into the matelial. measming devices such as the gaussmeter or by the attrac-
\Vith certain oIientations of high amperage cables, an tion of other ferromagnetic matelials. \Vhile this type of
electIiC' arc welding operation can magnetize ferromagnetic field can adversely affect subsequent machining, it is usually
mateIial. vel)' responsive to demagnetiZing techniques.
 294 / MAGNETIC PARTICLE TESTING

Circular Magnetic Fields

keyway. All holes or slots cutting through a circular field
Unlike longitudinal residual induction, circular residual produce magnetic poles that can attract materials such as
induction can exhibit little or no external evidence of its chips or dust from subsequent machining.
presence. The flux may be entirely confined within the Without special equipment, demagnetization of a circu-
material, depending to some extent on part geometry and larly magnetized object can be very difficult. Confirmation
the magnetizing procedure. of an adequate demagnetization level is an additional prob-
For example, if magnetizing current is passed through a lem. Leakage field measuring devices are ineffective since
homogeneous length of ferromagnetic bar stock having a there may not be an external leakage field to monitor. In this
circular cross section, the resulting circular residual flux case, reorientation of the circular field into a longitudinal
density is for all practical pUlposes undetectable without field prior to demagnetization may be advantageous in some
altering the bar in some manner. Virtually no leakage fields instances when such a procedure is compatible with part
emanate from its surface because the magnetic flux path is geometry and size.
closed within the object. Because of this, the internal
residual flux density B may be much stronger than if the test
object had been magnetized longitudinally in a coil or Multiple Magnetic Poles
solenoid having a comparable magnetic field strength H.
The circular residual flux density becomes an apparent Multiple magnetic poles can be induced ip and retained
problem when the geometry of the object is altered by by ferromagnetic material that has been exposed to direct
subsequent machining. For example, if a keyway is cut in a current magnetization, as in a magnetic chuck or lift mag-
piece of shafting that is circularly magnetized, the circular net. These fields can be pronounced and take the form of
field becomes quite evident. Strong leakage fields (north alternate north and south poles at relatively'dose intervals.
and south magnetic poles) occur on either side of the Demagnetization of multiple poles may be difficult using
conventional alternating current through-coil procedures.
 DEMAGNETIZATION OF TEST OBJECTS / 295

PART 2
PRINCIPLES OF DEMAGNETIZATION

Ferromagnetic material differs from other material in As H is reduced to zero, the level of B falls to + Br rather
that it contains magnetic domains, localized regions in than zero. This level of retained magnetization ( + B r) is the
which the atomic or molecular magnetic moments are remanent or residual field remaining in the material. As
aligned in parallel. vVhen a matclial is not magnetized, the negative values of H (opposite polarity) are applied, the
domains are randomly O1iented.and their respective mag- curve passes through - Hc and on to - HTIl and reaches a
netic inductions sum to zero. \Vhen the material is exposed magnetization level of - Brn' Reversing the applied field to
to. a magnetizing field strength H, the domains tend to align a value of + Hm completes the magnetic hysteresis loop.
Wlth the applied magnetic field and add to the applied field.
\Vhen tIle magnetizing source H is removed, some of the
magnetic domains remain in their new orientation rather
than ~etuming to the original random orientation and the Retentivity and Coercive Force
matenal retains a residual magnetic field.
The value Hc is the coercive force and is an indicator of .
the difficulty involved in demagnetizing a material. The .
Magnetic Hysteresis : value of Br is the material's retentivity and indicates the
residual induction within a section of the material. As a mIa, .
. Magnetic hy~teresis is a lag in the change.of magnetiza- high co@rcive forces are associated with harde'r materials
~lOn values after a change in the magnetizing force. Figure 1 and low coercive forces with softer materials. Therefore,
Illustrates the relationship between the magnetic field harder materials usually offer more resistance to demagne-
stren.gth H (magnetizing source) and th~ magnetic flux . tization and require a higher demagnetizing field than softer
denSIty ~ (level of magnetization). When an unmagnetized materials.
marenal IS magnetized by gradually increasing H from zero No definition has been made for the dividing line between
t~ -: Hm, the level of magnetization B increases along the hard and soft materials. However, if Hc ;::: 8,000 A-m- 1
~rgl,n magnetization curve to a maximum value correspond- (100 Oe), then the material is typically considered hard. If
mg to + Bm. Hc ~ 400 A-m -1 (5 Oe), then the material is considered
soft. Coercive forces as large as 8 x 105 A-m -1 (10 4 Oe)
and as small as 0.08 A-m -1 (10- 3 Oe) have been observed.
F'~URE J. Magnetic hysteresis curve The fact that an object retains a strong residual field B r is
not necessarily indicative of a high coercive force H c' Some
MAGNETIC FLUX DENSITY materials retain an appreciable residual induction and yet
(LEVEL OF MAGNETIZATION I
are eaSily demagnetized (transformer steels are an exam-
+8
pIe). On the other hand, some materials that retain relative-
ly weak residual fields can be extremely difficult to demag-
1:
GLiJ netize because of a high coercive force.
z~ i VIRGIN
l.U :J r-..;---:-: MAGNETIC

§~ .,:1 CURVE
Basic Principle of Demagnetization
-J Z -H H
~~ _ _~-~no_ _~~_ _-¥~~ ______+-___
~tiJ Practically all demagnetizing methods are based on a
tu~ common procedure. A magnetizing field strength H, suffi-
z«
~~
ciently high to overcome the initial coercive force He, is
!/ -8,
alternately reversed in polarity and gradually reduced to
~--. zero. The diminishing hysteresis curve shown in Fig. 2
illustrates this principle.
-8 The value of the total coercive force He is often unknown
and varies with proximity to the test object's ends. However,
296 / MAGNETIC PARTICLE TESTING

it is always less than the maximum magnetizing force H 11/

FIGURE 2. Diminishing hysteresis curve value. This may serve as a guide in cases where the
magnetizing field strength Hm is known, as is sometimes the
FLUX DENSITY case in magnetic particle testing. Successful demagneti-
+8 zation procedures often start with an initial demagnetiz-
ing field which equals or exceeds that of the magnetizing
force Hm.
An object may also be demagnetized by raising its
temperature above the Curie point. Although this method
provides thorough demagnetization, it is often impractical.
-H FIELD
---------+++~~~------+---H STRENGTH

-8
 DEMAGNETIZATION OF TEST OBJECTS I 297

PART 3
SUMMARY OF DEMAGNETIZATION
PROCEDURES

influence of the coil's field. The coil should not be

Alternating Current deenergized until the test object has reached this pOint,
Demagnetization three or four coil diameters away or about 1 m (3 ft). This
mle applies to most demagnetizing coils in industry, where
peak values of the field at the center of the coil are about
1.2 X 105 to 2.4 X 10,5 A-m - I. Small test objects can be
Through-Coil Method
hand held, placed within the coil and withdrawn. Field
The simplicity of the alternating current through-coil penetration may be several skin depths.
method makes it .one of the most prevalent demagnetization The magnitude of the field can also be reduced by
techniques. The method uses a coil powered from a current withdrawing a coil such as a cable coil away from a stationary
source altemating at line frequency (usually 60 Hz in the test object. This method is advantageous for high produc-
United States). Operating at a fixed amplitude, the coil tion rates since a P!operly designed coil can be continuously
prqduces a continuously reversing magnetic field because of energized while a steady stream of test objects is conveyed
the cyclic nature of tbe current. As a test object is conveyed through the cC?il opening. Typical alternating current
through the coil, it is subjected to the most intense magnetic through-coil demagnetizers are shown in Figs. 3 and 4.
field.while within the confines 'of the coil. Demagnetization of test shims may be performed by
The strength of the field is gradually reduced to zero as removing them slowly from one pole of an alternating
the. object exits the :oil and teaches a point beyond the current yoke.

FIGURE 3. Manually operated alternating current FIGURE 4. Production alternating current coil
coil demagnetizer with roller carriage and demagnetizer incorporated into conveyorized
automatic timer magnetic particle testing unit

FROM MAGNAFLUX CORPORATION. REPRINTED WITH PERMISSION FROM MAGNAFLUX CORPORATION. REPRINTED WITH PERMISSION.
298 / MAGNETIC PARTICLE TESTING

Cable Wrap Method object should remain stationary \\rithin the coil until tIlt'
demagnetizing cycle has been completed. vVhen the contact
There are applications (such as large or immobile test
method is used, contact should be maintained until the cvcle
objects) where a cable wrap, in conjunction with an appro-
is completed. The direct current method provides deep
priate high amperage alternating current power source, can
penetration and is usually very successful on objects that are
provide a convenient means of demagnetization. The high
otherwise difficult to demagnetize.
amperage power source must also incorporate a suitable
Because of economic considerations, the reversing direct
current control to facilitate the gradual reduction of current
current demagnetiZing feature is usually incorporated into
from maximum to zero.
the design of direct current horizontal wet magnetic particle
Portable or mobile alternating current power packs are
testing units or relatively large stationary power packs.
available for such applications. Equipment of this type
usually features stepless solid state current control. Some
units incorporate special current decay circuitry that auto-
matically reduces the current from maximum to zero in a Reversing Cable Wrap Method
matter of three or four seconds. This method is used for demagnetizing objects too large
or heavy to process on a horizontal wet testing unit. The
object to be demagnetized is wrapped with multiple turns of
Through-Current Method high amperage flexible cable (such as 4/0) connected to a
stationary direct current power pack.
Through-current demagnetization is a contact method. The current is alternately reversed in direction and
High amperage alternating current is caused to flow directly reduced in amplitude through multiple steps until the
through a test object starting at contact electrodes. The current reaches zero. This is usually accomplished using
current control system must prOvide gradual reduction of built-in automatic circuitry similar to that deSigned into
current as r.equired for demagnetization. A portable alter- some wet testing units. Power packs that lack an automatic
nating current power pack used for cable wrap demagneti- demagnetizing cycle can sometimes be used to accomplish
zation may also be used for the through-current method. demagnetization by manually interchanging cable connec-
Through-current demagnetization is used on horizontal tions (current reversal) and ma~ipulating manual·current
wet magnetic particle testing units that have comparable controls (current reduction). However, this is til.ue consum-
current control systems. Because of part configuration and ing and requires an appropriately fine current control to he
the accessibility of alternating current, the through-current successful.
'method can have real advantages for ~ertain ap·plications.

Pulsating Reversing Method

Direct Current Demagnetization A high amperage direct current coil demagnetizer has
been designed to produce alternate pulses of positive and
negative current. The pulses are generated at a fixed
Reversing Direct Current Contact Coil Method
amplitude and a repetition rate of five to ten cycles per
Reversing direct current contact coil demagnetization is second. This permits relatively small objects to be demag-
usually associated with relatively large test objects that have netized by the through-COil method.
been magnetized using a direct current magnetic field. It is The object is subjected to a constantly reversing magnetic
also applicable in certain instances where alternating cur- field as it passes through the coil and the magnitude of the.
rent demagnetization proves ineffective. effective field is reduced to zero as the object is gradually
The method requires high amperage direct current or withdrawn from the coil. This mode of operation is identical
full-wave rectified alternating current that can be directed to the altemating current through-COil method described
to a coil or contact plate. There must also be provisions for above, except for the reduced repetition rate (five to ten
reversing the polarity of the current and suitable means for cycles per second compared to sixty cycles per second).
gradually reducing its amplitude to zero. The lower repetition rate substantially reduces the ~kill
The direct current is alternately reversed in polarity effect with a corresponding increase in magnetic field
(direction) and reduced in amplitude to zero. Although penetration. Therefore, the method provides additional
fewer steps may provide satisfactory results, greater reliabil- demagnetizing capabilities on some relatively small test
ity is achieved by using about thirty reversals and current objects with thick cross sections that are difficult to handle
reductions to approach zero asymptotically. with line frequency coil demagnetizers. Another standard
The cycle is usually controlled automatically and requires mode of operation allows the object to remain stationary
about thirty seconds to complete. When using a coil, the test within the coil while the current gradually decays to zero.
 DEMAGNETIZATION OF TEST OBJECTS I 299

when facilities for overall (full length) demagnetization are

Specialized Demagnetization not available.
Procedures One such procedure applies a bucking (opposite polarity)
field at the ends of the object using a direct current coil
positioned near the ends. By trial and error, magnetizing
There is a \\lde valiety of miscellaneous demagnetizing current to the coil is increased until external field readings
techniques that have specific applications designed around a are reduced to an acceptable level. The result is a redistri-
specific test object. bution of the field at each end of the tube. Unfortunately,
mechanical agitation subsequent to this procedure, from
handling or transpOlt, can result in the redistlibution of the
Demagnetization with Yokes
longitudinal field and can reestablish strong leakage fields at
A yoke confIguration is sometimes used to demagnetize the ends of the test object.
small objects having high coercive force. Yokes are also used Another specialized demagnetization procedure uses a
for demagnetization of objects with smalllength-to-diameter full length internal conductor in conjunction with a unidi-
(LID) ratios, including spheres with near-perfect LID ratios rectional current pulse to reorient the longitudinal field and
of 1. Yokes are usually designed for a specific type or range form a circular field. The magnitude of the circularly
of test objects. magnetizing current is increased until acceptable external
Some alternating current yokes are similar in operation to field readings are achieved. While this is not demagnetiza-
the altemating current coil method, where the test object is tion, it does reduce leakage fields to a level suitable for butt
passed between or across the surfaces of the pole faces welding operations. It also creates the undesirable condi-
(maximum field strength) and then withdrawn. Direct tions associated with circular magnetization and hinders
" current yokes a~e sometimes used with the reversing direct subsequent operations, including weld preparation or
current method on specific objects requiring deep penetra- threading ...
. !ion. Howeyer, this is an inhetently slow process because of . Overall demagnetization is the best solution for long
the associated high inductance of the circuit. tubes and this is usually carried out using some form of
. Some coil or yoke arrangements use damped oscillation to reversing direCt current. For static conditions, an internal
obtain the required reversing and diminishing field. The conductor is used in conjunction with a conventional revers-
oscillation is delived from a special circuit design based on ing direct current cycle. A cable, wrapped to give several
specific values of capacitance, inductance and resistance. internal conductors, could be used in place of the central
conductor. However, when object' s1ength is taken into
Demagnetization of Oil Field Tubes consideration, the central conductor approach is much more
time efficient.
Valious testing processes involving direct current magne- The advent of pulsating direct current equipment has
tization leave the test object with a significant longitudinal lead to dynamic demagnetizing systems for overall demag-
residual field. Demagnetization is usually required to elim- netization of long test objects. In such systems, the object is
inate any possibility of the field having an adverse effect on conveyed through a special coil arrangement at speeds up to
subsequent machining or welding operations. Some tubular 1.5 m per second (300 ft per minute). The coil current
products have lengths of 9 to 15 m (30 to 50 ft) and this consists of constant amplitude pulses of direct current
presents a formidable challenge to standard demagnetiza- altemating relative to polarity (positive and negative) at a
tion procedures. Some unusual methods have been used rate of about five cycles per second.
300 I MAGNETIC PARTICLE TESTING

PART 4
SELECTING A DEMAGNETIZATION
PROCEDURE

Selecting a suitable demagnetization procedure is a mat-

ter of matching capabilities with the specific application. Maximum Effective Field Strength
Each of the procedures discussed above has advantages and
limitations based on test object size, hardness, production The maximum effective field strength Hill of a pmticular
rate and source of magnetization (see Table 1). demagnetization procedure is an important indicator of the
procedure's capability. BaSically, the Hill value must be
sufficient to overcome the coercive force He associated \-vith
limitations of Alternating Current the material to be demagnetized. Generally, He increases
with hardness.
Methods Test object size and geometIy (including the LID ratio)
are also factors that influence field strength requirements.
The alternating current through-coil method is probably The demagnetization potential of a procedure inherently
the most widely used demagnetization technique because of increases with field strength capabilities. \Vhen related t~
its simplicity and its adaptability to high production rates. various alternating current coil sizes with equal ampere-tum
However, the skin effect associated with alternating current ratings, the field strength increases as coil size decreases.
magnetic fields results in limited penetration capabilities.
Consequently, alternating current demagnetization meth- .
ods may be ineffective for removing deep residual fields Reversing Direct Current
such ~s th9se associated with direct current magnetization .. . . . .
.
This becomes a significant factor on large test objects.· As Reversing direct current procedures are usually required
a rule, alternating current methods can effectively demag- when alternating current techniques prove inadequate. For
netize most objects with a cross section less than 50 mm those difficult applications, reversing direct current has
(2 in.), regardless of the original magnetizing source. Larger unexcelled demagnetization capabilities. Almost any object
test objects with deep residual fields usually require some can be demagnetized to an acceptable level with reversing
form of reversing direct current demagnetization. If alter- direct current. While production rate capabilities are limit-
nating current is the only source of magnetization, alternat- ed, this is usually not a significant hindrance; since larger
ing current demagnetization is effective regardless of test objects are often involved, the associated production rates
object size. are relatively low.
 DEMAGNETIZATION OF TEST OBJECTS 1301

TABLE 1. Demagnetizing method selection guide based on test object size, hardness and production rate for
residual fields from direct current magnetizing source
Test Object Size Material Hardness Production Rate

Small' Medium 2
Large 3
Soft Medium Hard Low Medium High

Alternating Current (50/60 Hz)

Through-coil A 0 N A A N A A A
Cable wrap N 0 Q A A N A N N
Through-current (30 point step down) N A N A A N A Q N
Through-current (current decay) N A N A A N A 0 N
Yoke A N N A A 0 A N N

Reversing Direct Current

Through-coil (pulsating) A A A A A A A A A
Coil (30 point step down) A A 0 A A A A A N
Cable wrap (30 point step down) N 0 A A A A A N N
Through-current (30 point step down) N A A A A A A Q N
I. HAND HELD
2. LESS THAN 50 mm 12 In., DIAMETER
·3. MORE THAN 50 mm /2 In.1 DIAMETER

LEGEND
o = OUESTIONABLE
A = APPLICABLE
N = NOT APPLICABLE
302 I MAGNETIC PARTICLE TESTING

PART 5
MEASURING EXTERNAL FIELD STRENGTH

In SI units, magnetic flux density B is measured in tesla north pole of the magnet. In the absence of external fields,
or webers per square meter. In cgs units, magnetic flux the pointer indicates zero on the graduated scale.
density is measured in gauss or as magnetic lines of force 'When the magnetic north pole of a residually magnetized
per square centimeter. One tesla is equivalent to 10 4 gauss. object is moved close to the pivot end of the pointer, the
In SI, the magnetic field strength H is measured in amperes south pole of the vane is attracted toward the object and the
per meter (A-m -1); in cgs, the unit is oersted (Oe). One n01th pole of the vane is repelled. The resulting torque
ampere per meter is equivalent to 0.013 oersted. causes the pointer to rotate in the positive or plus direction.
Flux density and field strength are related to each other The restraining torque is prOvided by the tendency of the
and to magnetic permeability by the following equation: vane to remain aligned with the leakage field of the
permanent magnet.
(Eq. 1) Sensitivity of the field indicator is primarily a function of
the permanent magnet's characteristics and the spacing
Where: between the magnet and the vane. The device can be
calibrated in relative units, milliteslas or gauss. Calibration is
f.L(J = the permeability of free space; and only valid when the device is subjected to a uniform
f.L = the relative permeability.
The value of permeability is not affected by the 'choice of
units. Permeability values in customary units are the same FIGURE 5. Typical magnetic field indicator
magnitude as those qf relative permeabilities in Si units. construction .
Because gauss and oersted have by definition the same
magnitude in air, the f.Lo factor is not used in the equation
for customary u n i t s . '
By itself, the permeability of a material has no meaning.
Its value at speCific values of H or B does have meaning.

Field Indicators
Theory of Operation
The field indicator is a hand held instrument used to
measure the relative strength of magnetic leakage fields.
The construction of a typical field indicator is shown in
Fig. 5 and its the01Y of operation is relatively Simple. As
illustrated, an elliptically shaped soft iron vane is attached to PERMANENT MAGNET
a pointer that is free to pivot. A rectangular permanent
magnet is mounted in a fixed position directly above the soft
iron vane.
Because the vane is under the influence of the magnet, it
tends to align its long axis in the direction of the leakage
field emanating from the magnet. In so dOing, the vane
becomes magnetized and has a magnetic pole induced at
each end of its long axis.
On the end of the vane below the south pole of the
magnet, a north pole is induced. Correspondingly, a south
pole is induced at the opposite end of the vane below the
 DEMAGNETIZATION OF TEST OBJECTS / 303

magnetic field snch as that fonnd at the center of a large coil by a flexible multiconductor elechical cable. This permits
or Helmholtz coil arrangement. leakage field measurements light at the surface of a test
The pattern of external fields emanating from a residually object, and consequently, the readings are substantially
magnetized object is anything but uniform and varies with greater compared to those obtained with a field indicator.
changes in pmt geometry. In view of this, readings obtained In the past, use of the gauss meter was confined to
with ~ field indicator are effectively in relative nnits but still laboratory work because of instmment cost, complicated
usefill for comparative purposes. operating procedures and the delicate nature of the probe.
However, gaussmeters are now widely used in many ways,
Use of the Field Indicator such as qualifying demagnetization procedures for small or
unusual test objects; establishing maximum permissible
The relative strength of an extemal field is measured by field indicator readings; and verifYing field indicator calibra-
bIingiIlg the field indicator near the object and noting the tion procedures.
deflection of the pOinter. The edge of the field indicator at
the pivot end of the pointer should be closest to the object
under investigation.
The required degree of demagnetization is usually spec- Measuring Techniques
ified as a maximum field indicator reading (in terms of
. divisions, tesla or gauss) for a specific device. Leakage Leakage field measurements are undertaken to ascertain
field patterns of very small objects may be too limited for the level of residual magnetic fields emanating from an
detection by a field indicator - the sensing element (soft object. To increase the accuracy and repeatability of such
iron vane) is located approximately 12 mm (0.5 in.) from the measurements, it is good practice to isolate the object
edge of the device's casing. magnetically to eliminate the influence of extraneous mag-
The field indicator is the most common device used in netic fields or other ferromagnetic material. This can. be
. industry for mon~tOling the effectiveness of demagn~tizing . accomplished by moving the test object to an appropIjate
. processes. It is convenient, easy to operate and inexpensive. area and supporting it manually or by placing it on a suitable
nonferromagnetic platform.
Caution should be exercised to keep certain types of field
Laboratory 'nstruments indicators at a safe distance from demagnetizing fields. Such
exposure can partially demagnetize the internal permanent
The Hall effect gaussmeter is the most common labora- magnet and effectively change the calibration of the device.
tory instmment used to measure the strength of leakage If this occurs, the sensitivity of the device increases substan-
fields quantitatively. The sensing element is located in a tially. In strong fields, remagnetization of the device's in-
remote hand held probe connected to the basic instrument ternal magnet may occur, rendering it unfit for future use.
304 I MAGNETIC PARTICLE TESTING

PART 6
TYPICAL DEMAGNETIZATION PROBLEMS

Demagnetizing Field Strength Poor Length-to-Diameter Ratio

Too Weak
The ratio of an object's length to its diameter (the LID
Demagnetizing capability is directly related to the tech- ratio) is as important to coil demagnetization as it is to coil
nique's magnetic field strength. Higher field intensities magnetization. Values below 3:1 can be a problem, but the
increase the variety of objects that can be successfully situation may be corrected by effectively increasing the LID
demagnetized. ratio.
Conventional alternating current coils are usually rated in This correction is done by adding ferromagnetic pole
terms of ampere-turns. However, for the same applied pieces at both ends of the object. Pole pieces should be
ampere-turns, the strength of the field increases as coil about 1.50 mm (6 in.) long and nearly the same diameter as
diameter decreases. It should be noted that the field the object. \Vith pole pieces in place, the assembly can be
strength within a coil is greater near the inside wall than at passed through the demagnetizing coil in the usual manner.
the center. For large lots, it is sometimes convenient to pass the objects
through the demagnetizing coil end to end in a chain.

Alternating Current
Demagnetization and Direct MaQnetic Shi~lding
Current Magnetization Whenever possible, it is desirable to demagnetize an
object before its assembly with other components. Once
Objects magnetized by a direct current source can be assembled, the test object may be adjacent to or surrounded
difficult if not impo"ssible to adequately demagnetize with an by other ferromagnetic material. In such cases, the demag-
alternating current process. This is more prevalent in test netizing field may be shunted through the adjacent mateliaJ
objects with diameters greater than 50 mm (2 in.) because rather than through. the object itself and demagnetization
of the pronounced skin effect associated with alternating will be ineffective.
current fields. Small objects should not be passed through a demagne-
Deep interior residual magnetism remains unaffected by tizing coil in bundles or layered in handling baskets. Test
surface oriented alternating current fields. On such objects, objects in the center of a stack are shielded from the
some form of reversing direct current demagnetization must demagnetizing field by the outer layer of ~laterial. Ferro-
be used. magnetic baskets or trays are objectionable for the same
reason.

Test Object Orientation

Improper test object orientation relative to an alternating Calibration of Field Indicators
current demagnetizing coil can have an adverse effect on
the demagnetization process. For the best results, a test Field indicators are convenient devices for monitOling the
object should he passed through the coil lengthwise or with effectiveness of a demagnetization process but ohs~tlre
its longest axis parallel to the coil's central axis. changcs in calibration can cause problems vvben verifying
Ring shaped test objects such as bearing races can be the demagnetization procedure.
rolled through an alternating current coil to obtain the Inadveliently subjecting the device to a demagnetizing
desired results. In a similar manner, objects with a complex field can pmiially demagnetize the field indicator's intern,~1
configuration may require some form of rotation as they are permanent magnet and significantly increase the de\ice s
passed through a demagnetizing coil. sensitivity. Subsequent measurements would then indicate
 DEMAGNETIZATION OF TEST OBJECTS I 305

the erroneous conclusion that an established demagnetiza-

example, objects exhibiting a field indicator reading around
tion process is inadequate. In such cases, the results of
0.3 mT (3 G) usually will not have an adverse effect on
demagnetization should be verified with another device
known to be accurate. subsequent machining or welding, instrumentation or end
use.
Strictly specified limitations should be placed only on a
Demagnetization Specifications limited number of demagnetization applications (particular
test objects manufactured for specific service) as required
Stringent demagnetization specifications can be a source by experience or laboratory data.
of selious production problems. Specifications are often
To eliminate ambigUity, a speCification should define how
wlitten without accommodation to the practical limitations
a reading is made - on the surface with a Hall effect meter
of the demagnetization process and its verification. For or with a field indicator of a specific deSign.
 306 I MAGNETIC PARTICLE TESTING

PART 7
DEMAGNETIZING EQUIPMENT

offered as an option on direct current magnetization equip-

Summary of Alternating Current ment but separate units are also available. Some typical
Demagnetization Equipment units are listed below.

Alternating current coil denwgnetizers function only as

Coil denwgnetizers are designed for demagnetization
demagnetizers. They are designed for operation at line
only. A constant amplitude, high amperage direct current is
voltage and line frequency. Most of the units are rated for
directed through a fixed coil. The polarity of the current is
intermittent duty but some are available for co'ntinuous
continuously reversed at a fixed rate. Relatively small test
duty. A variety of coil openings are commercially available.
objects may be demagnetized dynamically \vith the through-
Portable alternating current half-wave units are light-
coil method or statically with modified current decay tech-
weight magnetization and demagnetization units used to
niques. Coil sizes may be limited.
power high amperage cables. Solid state steples's current
control permits demagnetization by coil, cable wrap or
contact methods. Limited alternating current output usually Horizontal wet nwgnetization and denwgnetization units
restricts application of these systems to relatively small or permit objects to be magnetized, tested and demagnetized
mediuI1) sized objects. in place ·on the system. Demagnetization is based on
Mobile alternating current half-wave units are heavy automatic polarity reversal and a step down cycle using coil
. duty, dual output s.ystems used in cpnjunction with high or contact methods. These systems are suitable for relatively
amperage cables. Newer designs have solid state current large test objects.
control and built-in current decay circuitry. Relatively large
test objects can be demagnetized by the coil, cable wrap or
contact methods. Stationary nwgnetization and denwgnetization pou:er
The horizontal wet alternating current units \vith solid packs produce high amperage directed to the test object
state current control feature current decay circuitry that can through flexible high amperage cables. Either the contact or
be used for demagnetizing either by the coil method or the cable wrap method can be used to demagnetize large
contact method. objects. The reversing direct current demagnetization pro-
cess is automatically controlled through a fixed cycle.

Summary of Reversing Direct Specially designed systems are available for applica-
Current Demagnetization Equipment tions that require high rates of produdivity. SpeCial power
pack and coil configurations are available to demagnetize
Equipment for reversing direct current demagnetization 9 to 15 m (30 to 50 ft) lengths of tubing on a continuous
is often expensive. The reversing feature is commonly basis at rates up to 1.5 m per second (300 ft per minute).
 DEMAGNETIZATION OF TEST OBJECTS / 307

PART 8
DEMAGNETIZATION OF ELONGATED
TEST OBJECTS

:M lIch flnished or semifinished tubular product is tested In some testing specifications, it is also reqUired that the
for surface breaking tight discontinuities by the magnetic external longitudinal magnetic fields close to the ends of the
flux leakage me.thod. In the case of ferromagnetic oil fIeld tubular product (especially oil field transmission line pipe)
tubes, for example, the ends of the tube might first be tested be reduced to a value below a specified minimum follOWing
for transversely oriented discontinuities, with the magnetic magnetic flux leakage testing. 1 This is generally done so that
flux being induced by placing the end of the tube in a coil. such materials can more eaSily be field welded.
The tube is then tested for longitudinal discontinuities using Typically a specification might require that not more than
circumferential induction resulting frpm a current pulse 800 A-m- 1 (10 Oe) be measured with a gaussmeter at the
along an internally placed conductor connected to a capac- ends of the material. After magnetic flux leakage testing for
itor discharge system. Both of these methods are shown in transversely oriented discontinuities, the magnetic field
Fig. 5. intensities emerging from the ends of tubes might easily be
40 kA-m -1 (500 Oe) and it is essential. that some form of
deI!1agnetization be applied.
FIGURE 6. Magnetization of elongated products In the text below, commorily used fOr'ms.of demagnetiza-
as commor;tly practiced for oil field tubular tests': tion or remagnetization are ·outlined.
fa) circumfetential magnetization of a tube using
the capacitor diSCharge internal conductor .
method; and fb) longitudinal magnetization of Circumferential Remagnetization
an end region using an encircling coil
In some cases, after longitudinal magnetization has been
established and transverse discontinuity testing completed,
la) CAPACITOR DISCHARGE UNIT
it is necessary to remagnetize only in the circumferential
direction. The resulting magnetization then causes little or
no external field. This is easy to accomplish in tubes by using
C ~ the internal conductor method 2 shown in Fig. 6a. The

FIGURE 7. Hall element gaussmeter measuring

longitudinal magnetic field in air close to the
end of a tube

TUBE

Ib)

GAUSSMETER

tI\j~9J}1
308 / MAGNETIC PARTICLE TESTING

remaining longitudinal field strength can be measured easily tube is not circumferentially magnetized.·3 It merely incli-
with a gaussmeter as shown in Fig. 7. cates that the circumferential and axial flux densities in-
In this test, a calibrated gaussmeter is used to measure volved obey a vector relationship such as that shown in
the maximum value of the magnetic field strength at the end Fig.8.
of the material. If it exceeds a specified value, additional In this particular example, the material just inside the end
shots from the capacitor discharge internal conductor are of the tube is saturated with magnetic flux that is virtually
applied to reduce this emergent field. For this application, circumferential at a value very close to the remanence
a gaussmeter capable of reading up to 100 mT (1,000 G) is Br for the steel. Suppose that an external flux denSity
preferred over other types of field indicator. of 1 mT (10 G) occurs for a material with a remanence of
1,400 mT (14 kG). This flux denSity measurement was made
Angled Nature of the Resulting Field with a gaussmeter after attempted circumferential mag-
netization by the method outlined above. Its value incli-
The presence of some small axial field strength at the end cates that the direction of the flux inside the material is
of a tube (or other elongated part) does not indicate that the only tan -1 (111,400) or 0.04 degrees from being perfectly
circumferential.

FIGURE 8. Possible magnetic flux density Drill Pipe End Area Tests
configuration at the end of a tube: (a) the
emergent flux density is measurable with a A commonly practiced magnetic fllLX leakage test is that
gaussmeter but the circumferential component is of longitudinally magnetizing the ends of drill pipe (Fig. 6b)
not; and (b) the resulting field within the tube In and looking for thread root cracks with wet fluorescent
this case Is almost circumferential magnetic particle techniques. Figure 9 illustrates an exam-
ple of pin thread stretching, the roots of these threads heing
(a) particularly prone to cracking. 4 Following such a test, the
rotation of the flux into the circumferential direction might
require several pulses from a capacitor discharge internal
conductor system. Unde.r such circumstances, longer pulses
(25 Ins) are more effective than shorter ones (3 ms). .
In the resulting circumferential magnetization, the drill
stem can be tested for longitudinally oriented discontinuities'
and the tool joints can be tested for heat-check cracks \\lith·
(b) LONGITUDINAL FIELD magnetic particle tests. The threads are also relatively easy
1 mT 110 G) to clean of particles.

FIGURE 9. Pin stretch on used drill pipe threads;

wet magnetic particle tests are often used to
o detect fatigue cracking on these thread roots
8
<T'

f-
E
o
~.

FROM EXXON CORPORATION. REPRINTED WITH PERMISSION.

"""""""11""""""1"'11'11'1'1'1'1"11"'""""1"""""""","""""""1'1"
DEMAGNETIZATION OF TEST OBJECTS I 309

upset Area Problems the surface of the test object; (2) the thickness of the
material; and (3) the matelial's B-H properties.
It is generally necessary. to apply only ?2 ~o ~ kA-m- 1
In traditional demagnetization with alternating current, it
(40 to .50 Oe) circumferentlally to m~gnetIze.Ol~ fIeld tt~bu­ is required that the test object be slowly removed axially
Jar materials to a point near saturatIOn. ThIs IS .especIal!y from the demagnetizing coil and that the strength of the
tme of the region between the ends of the matenal (QR III field be sufficient to produce saturation in the direction
Fig. 10). Some matelials may require highe.r field streng~h opposed to the existing magnetization direction.
values to cause saturation (the B-H curve for the matenal Because of the skin effect, the field strength at any point
should be consulted). within the matelial is lower than that at the surface. If
However, plior to threading the ends of such tubes, it is suffiCiently high fields are to penetrate the material to cause
sometimes necessaIY to bell or upset the ends (J'Q and RS) demagnetization at that given point, the surface field am-
of the material. This process may lock in existing magnetic plitude must be several times higher than that reqUired to
fluxes from earlier tests, the result being that the flux cause saturation. At five skin depths (about 5 mm) one
density at the ends can require a higher demagneti~ing percent of the surface field remains. It is clear that the
circumferential field than is required for the central regIOns alternating current technique is confined to relatively thin
of the tube. This is due to ineffective or incomplete stress or thin walled test objects and that some other form of
relief at the ends after upsetting, which also causes the B-H demagnetization is needed for thicker materials.
properties to differ from those of the central regions.
When this situation is encountered, a gaussmeter can
indicate the number of pulses required to rotate the flux
within the upset areas. FIGURE 11. The axial bulk flux density In a
10 m (30 ft) long tube: (a) after longitudinal
magnetization to saturation; and (bJ after direct
current coil demagnetization using a constant
Alternating Current Coi~ current; 'note the remanent flux density (1.14 T),
the opposed dipole effect caused by the
Demagnetization . demagnetization process, and the values of the
flux densities in the tube after this form of
After tubes have been longitudinally magnetized to satu- demagnetization
ration, ,lowe ling the remaining induction is usually done
with alternating current or direct current through a central-
ized, surrounding coil. With the alternating current tech-
nique, it is clitical to know that, for a frequency of 60 Hz, (a)
the skin depth being demagnetized is only about 1 mm
(0.04 in.). The effect of this form of demagnetization is
dependent on: (1) the strength of the demagnetizing field at

FIGURE 10. Different circumferential magnetic

field 8-H curves: (a) at the belled or upset ends
of tubes; and (b) the unworked center section of
a tube; curve (a) shows lower Br and higher He
Ib)

~I~T
than curve (b)

P Q R S
>-
f-

~
UJ
I
N J S
0
'- ./S + N
3
fa)
LL.

0 3
JST 1
6
I
10
(10) (20) (30)

DISTANCE ALONG PIPE

meters (feet)

AFTER LORD (SEE REFERENCE 51. REPRINTED WITH PERMISSION.

310 I MAGNETIC PARTICLE TESTING

provided to show that lower bulk fluxes can be obtained by

Direct Current Coil Demagnetization sensing the remaining flux and adjusting the demagnetiz<l-
tion current to make the flux as small as pOSSible.
Partial demagnetization of tubes is often accomplished by A commonly used method for sensing the flux density
passing the tube through a coil in which the field strength H level in an elongated object is to jerk a loosely fitting coil
opposes the direction of the residual induction in the from a position of flux linkage with the pmt to a position of
material. Demagnetization current settings are obtained by no flux linkage. This is shown in Fig. 12. "Vith the coil in
trial and error since only the external fields at the ends of position-I, the flux linkage with the magnetized object is
the tube are available to the inspector. Currently accepted given by:
practice is to adjust the demagnetizing field strength to
provide the minimum fields at either end of the test object. Net> = N x J 13 -n dA (Eq.2)
As shown in Fig. 11, this process leaves the flux density Where:
within the material at a relatively high level.
The actual value of the axial flux density in the material is Net> the flux linkage (weber-tums);
=
governed by two factors. The first is the local magnetic N =
the number of tums on the coil;
property or, in effect, just how high a value of flux density B 13 = the flux density of the test object in the coil
can be sustained. The second is the proximity of the ends of (webers per square meter or tesla)~ and
the material (this affects the flux density through the ndA = the area of the coil perpendicular to B
demagnetizing field). The latter factor is very important (square meters).
within about 1 m of the tube ends and has a considerable
effect on the local B-H properties. The general effect is to With the coil in position-2, the_flu~ linkage is zero and the
lower the local permeability, due to the internal demagne- net change in flux linkage is NJB - ndA. When the voltage
tizing field inside the test object. induced in the coil during this process is integrated with a
As an example, the effect of direct current coil demag- fluxmeter, the output (suitably compensated for the value of
netization on the axial flux density in a typical 10·m (30 ft) A and N) is the average value of the flux density in the test
tube is shown in Fig. 11. As illustrated in Fig. lla, the axial object directly inside the coil at position-I.
flux density of the tube after magneti~ation is relatjvely Applying this theory to the demagnetization process, the
constant at 1.14 T (11.4 kG) except within about 1 m of test object is passed through a s'ensing coil and the output of
either end. This value is close to the remanence for the the fluxmeter is used to electronically compensate the
material, which varies with localized stress, chemical com- curr~nt through the demagnetizing coil. This compensation
position and other factors. produces a net flux density in the sensing coil, and therefore
After demagnetization, in which only the external field is in the object, as close to zero as pOSSible. This is shmvll
sampled at both ends of the tube, the remaining flux density schematically in Fig. 13.
within the material is as shown in Fig. 11b. The flux density
for the majority of the material is relatively low and in the FIGURE 12. Measurement of the flux at a specific
same direction as the saturated state. However, at one end pOSition close to the end of an elongated test
of the tube the direction of the flux has been reversed by the object; a search coil is taken from position-1 to
coil field for roughly 1.4 m (4.4 ft). The demagnetized position-2 and the resulting voltage is integrated
material then exists in an opposing dipole configuration, within a fluxmeter; the fluxmeter output is
with poles as shown in the illustration. proportional to the flux at position-l
If this form of demagnetization is used, the inspector
must consider each application individually. The direct

_~~I
current required depends on the wall thickness of the tube.
Passage of the material through an altemating current coil, lfJ=O
perhaps of low frequency, further lowers the contained flux.

"
L~-:.:::::::. ===:=:::.:::::. -:.:::.
Flux Sensed Demagnetization
Traditional altemating current and direct current demag-
netization techniques do not totally demagnetize an object. FLUXMETER
They merely reduce the bulk flux density to the value at
which emergent fields are low enough not to hinder subse-
quent metalworking processes. The follOwing information is
 DEMAGNETIZATION OF TEST OBJECTS / 311

FIGURE 13. Diagram for optimizing the direct of the skin effect, surface demagnetization by the alternat-
current demagnetization of an elongated test ing current method is at best a temporal), measure so that
object; the sensing coil develops voltage as a the bulk magnetization of the test object may eventually
magnetized materral passes through it and the lead to large extemal fields.
voltage is fed to a fluxmeter; the meter's signal The alternative practice is that of local application of a
is proportional to the flux density within the direct current field which, while it may remove the flux from
test object and is used to control the a small region of a test object, does not leave the entire part
programmable power supply of the demagnetized. Unless the entire object can be demagne-
demagnetizing coil tized totally, presently accepted practice will continue: the
f1ux is reduced in certain regions of the object so that
DIRECT externally applied field" indicators show a low field. The end
CURRENT user will continue to accept material that is highly magne-
MAGNETIZING
COil tized even though it appears to be demagnetized. No
specifications appear to exist for demagnetization proce-
dures that lead to bulk flux reduction in previously magne-
tized objects. Below are examples of situations where the
lack of demagnetization can cause problems at a later date.
A common practice in the testing of oil field tubes is to
reduce the bulk flux to. the condition shown in Fig. II b.
Subsequent transportation and other mechanical vibrations
FWXMETER can cause the external field to rise to levels unacceptable to
the end user. Such lev~ls cannot be caused by the Earth's .
magnetic fi~ld (only f?02 tnT or 0.2 G) because the field
strength required to magnetize these materials is much
larger than 0.02 mT. .
A second common practice is to pass a longitudinally
magnetized tube or rod through an alternating current coil
Problems Associated with Partial carrying 50 or 60 Hz current. Skin depth considerations
Demagnetization reveal the volume of material actually demagnetized. The
material is left in a highly magnetized state, apart from a
Partial demagnetization can lead to field problems when surface layer. Subsequent handling causes the reappearance
the material later appears to be highly magnetized. Because of high external fields.
312 I MAGNETIC PARTICLE TESTING

REFERENCES

1. Electromagnetic Testing: Eddy Current, Flux Leakage 3. Oilfield Magnetism and the Mythology 'Which Sur-
and Microwave Nondestructive Testing. The Nonde- rounds It. Chapter 6. Houston, TX: International Pipe
structive Testing Handbook, second edition. Vol. 4. R. Inspectors Association.
McMaster, P. McIntire and M. Mester, eds. Columbus, 4. Moyer, M. and B. Dale. "A Test Program for the
OH: The American Society for Nondestructive Testing Evaluation of Oil Thread Protection." Journal of Petro-
(1986). leum Technology. Richardson, TX: Society of Petroleum
2. Stanley, R. and G. Moake. "Inspecting Oil Country Engineers (February 1985): p 306.
Tubular Goods Using Capacitor Discharge Systems."
Materials Evaluation. Vol. 41, No.7. Columbus, OH: 5. Electrornagnetic Methods of Nondestructive .Testing. \V.
The American Society for Nondestructive Testing Lord, ed. New York, NY: Gordon and Breach (198.5):
(1983): p 779-782. p 97-160.
 SECTION

REFERENCE STANDARDS AND

14
ARTIFICIAL DISCONTINUITY
INDICATIONS
Thomas Jones, Industrial Quality IncorpOrated, Gaithersburg, Maryland
Donald Hagemaier, Douglas Aircraft Company, Long Beach, California
Kermit Skeie, Kermit Skeie Associates, Hacienda Heights, California
Roderic Stanley, International Pipe Inspectors Association, Houston, Texas
Lydon SWartzendruber, National Institute of Standards and Technology, Gaithersburg, Maryland
338 I MAGNETIC PARTICLE TESTING

PART 1
FUNDAMENTALS OF REFERENCE
STANDARDS FOR MAGNETIC
PARTICLE TESTING

Nondestmctive tests are typically designed to reveal the the early days of the method. These rules have persisted
presence of discontinuities or to measure specific properties through the years in various standards and specifications.
in a stmcture or material. The discontinuity may be an The reliance on empirical data occurred because of the
anomaly in a homogeneous material or a change in one of enormous complexity of magnetic fields and their interac-
the material's properties (thickness, hardness or density, for tions with ferromagnetic components. Unfortunately, niles
example). of thumb have sometimes been used exclusively for deter-
Before testing, some form of artificial discontinuity or mining the adequacy of certain test setups. As with most
reference standard is commonly used to verifY the operation empirical data, the rules developed for magnetic p,lltide
of a magnetic particle testing system. This verification is testing should be used with caution and with an understand-
performed in order to (1) provide a sensitivity check of the ing of their limits. Caution dictates that regular system
testing procedure; and (2) establish a known correlation monitoring be used to verify acceptable test sensitivity -
between the response of the test system, the magnitude of most existing formulas ensure over-magnetization in some
the material property or the severity of a discontinuity. test objects.
Magnetic particle testing uses magnetic fields to inspect It is easy to demonstrate the connection between misu~ed
ferromagnetic materials. Discontinuities in the material rules of magnetization and inaccurate testing. Figure 1
cause disturbances in the magnetic field and this in tum shows ~here a null field is produced at the for~ in a simple
produces a leakage fluJ!:. It is this leakage· flux that permits. . test object configuration. Failure to use soine form of field
the formation of particle indications. Both the direction and strength indicator to verifY the presence of a valid magnetic
intensity of the magnetic field are critical in determining the flux could lead to inadequate testing of this critical area.
sensitivity of the test procedure. Both of these factors are in Measurements with the means described here have shown a
turn affected by the nature of the material, the test object variation from 0.3 to 8.5 mT (3 to 85 G) within the same test
geometry and the way in which the magnetic field is object, because of its geometric differences.
induced.
All of these parameters are interrelated to determine the
direction and intensity of the magnetic field in a particular FIGURE 1. Magnetic particle testing problem
location in a test object. The mathematics of these interre- area in a simple part geometry
lationships does not lend itself to straightforward, closed
form solutions even for relatively simple geometries. A
generalized solution for more complex test objects has not
been obtained and would probably not completely solve
the problem of establishing magnetic particle testing
procedures.

Empirical Rules for Using Reference

Standards
Perhaps more than any other nondestructive technique,
magnetic paIiicle testing has based its procedures on em-
piIical data (rules of thumb) developed by trial and error in -
 REFERENCE STANDARDS AND ARTIFICIAL DISCONTINUITY INDICATIONS I 339

Reliability Studies compound; and (3) verifying the adequacy of a test pro-
cedure for detecting discontinuities of a predetermined
The absence of verification has led to widely varying magnitude.
effectiveness. Two often cited reliability studies are excel-
lellt examples of the problems associated with the lack of a
sensitivity check. The first of the studies l was the result of System Evaluation
a round-robin among four major aerospace contractors, two Unlike other nondestructive testing methods, magnetic
jet engine manufacturers, three landing gear manufacturers, particle testing systems give little evidence of malfunction.
one major forging supplier and a commercial test lab. Each The absence of a test indication could mean that (1) tests
company processed and tested twenty-four pedigreed spec- were properly performed on samples without discontinuities;
imens with known discontinuities. All of the discontinuities or (2) the testing system was not working and therefore not
were considered detectable at that time in the method's locating existing discontinuities.
development. Figure 2 shows the results of that study. As a result, some form of reference standard is needed to
The companies varied from less than 20 percent detec- determine proper system performance and adequate sensi-
tion to over 90 percent. Only one of the eleven companies tivity. Such a system evaluation tool should ch~ck ~~r
scored better than 60 percent of the discontinuities. At the contamination of the magnetic particle bath, matenal VISl-
time of its completion, the study drew few conclusions about bility (loss of fluorescence on fluorescent oxides), particle
t be cause for this variability. concentration (for wet methods), adequate particle mobility,
In another study,2 it was concluded that the maximum and the ability to generate an appropriate magnetic field.
probability of detection was 55 percent on jet turbine blades
made of magnetic alloys (Fig. 3). The reasons for the poor
results were not individually assessed. System Standardization
When multiple variables C'ln affect the. outcome of. a test,
a means should be used to normalize or standardize the test.
Needs for a Known Indicator This ensures that consistent, repeatable results are achieved,
independent of the machine, the operat.or or the time of the
The goal of system monitoring is to verify that the system test.
is performing in the desired way and at the desired . The most direct way to achieve consistent results is to
sensitivity. The most direct 'Yay to achieve this goal is to regularly use a reference standard to compar~ system
1'<'lify the system's ability to detect one or more known sensitivity to preestablished tolerances. If the deslred sen-
discontinuities. Ideally this would be done with a disconti- sitivity is not achieved, testing should be stopped to allow
IlIdt}' of the smallest critical size in an exact duplicate of the required system adjustments.
test piece. This ideal is rarely if ever practical.
More often, some form of artificial discontinuity indicator
is IIsed. This so-called reference standard is designed to help
('\'aluate several aspects of a magnetic particle system's FIGURE 3. Magnetic particle testing discontinuity
performance, including: (1) testing the magnetizing equip- detection probability
Illent; (2) checking the sensitivity of the magnetic particle
Z 80
0
i= 70
FIGURE 2. Magnetic particle discontinuity U
UJ
60
detection survey results I--
UJ
O~ 50
U.
c
Qi
~
iOO~

90~
_ _ _ _ _ _ _ _ _ _ _ _ __
_ _ _ _ _ _ _ _ _ _ _ _ _ ___ o>- Qi
~
40

~ BO~ __________________ 1--..90 30

o85~ ~------------------­
:::i
70 co 20
~~
wI-- ~
60
50
~--------------­
-l------------
«
Cl) 10
twoZ~ 40 0Cl<:
0u- 30 C!.. 0.25 05 0.75 10 1.25 1.5 1.75
~ 20
o Ig (O.I) 10.2) (0.3) (0.4) (0.5) (0.6) 10,7)

CRACK LENGTH
6 B 9 10 II
millimeters (inches)
COMPANY
AFTER RUMMEL. REPRINTED WITH PERMISSION.
340 I MAGNETIC PARTICLE TESTING

Parametric Evaluations The selection of an appropriate test technique may be the

single most important factor in the success of a magnetic
It is at times useful to examine a system's sensitivity to particle test. The use of reference standards and artificial
changes in one or more variables. For example, to evaluate indications can Significantly improve system pelformance
the effectiveness of magnetic particle testing on chromium and may also reduce the cost of testing by eliminating
plated components, it would be appropriate to investigate: unnecessary configurations or scrappage caused by exces-
(1) the effect of various plating thicknesses; (2) the sensitiv- sive current. The use of reference standards during tech-
ity of the test to changes in current levels or field strength; nique development can quickly verify the completeness of
and (3) the effect of changes in the particle type or bath coverage, the direction of magnetizing fields and the level of
concentration. field strengths.
Reference standards are used to study these changing In many cases, it can be demonstrated that common rules
parameters. Indications of the known discontinuities help of thumb produce field strengths far in excess of those
determine the effect of the individual parameters on test necessary for detecting particular discontinuities. Excessive
sensitivity. The results of such studies are used to generate field strength might appear to provide a margin of safety for
or modifY testing procedures for the material and geometry unknown effects of test object material and geometry.
of interest. However, in many cases, this excess produces a significant
field component normal to the test object surface. This in
turn reduces particle mobility, increases particle back-
ground and actually reduces rather than enhances the
Technique Development sensitivity of the test. Reference standards are often used to
regulate field strength to avoid excess flux while achieving
In the past, it was common for some operators to rely accurate indications.
solely on empirical rules for establishing magnetic particle Two kinds of artificial discontinuities are used for mag-
testing procedures. This practice frequently lead to over- netic particle test systems: (1) those deSigned to indicate the
magnetization, poor coverage, inappropriate selection of adequacy of the field in an' unknown test object; and
test geometries or some combination of all three disadvan- (2) those deSigned to measure the effeetiveness of the
tages. tes.tingo system independent of tlie test objecOt.
 REFERENCE STANDARDS AND ARTIFICIAL DISCONTINUITY INDICATIONS /341

PART 2
REFERENCE STANDARDS FOR SYSTEM
EVALUATION

Heference standards may be llsed to evaluate the fimc- Using the Ring Standard
tionality or performance of a magnetic particle testing
The ring standard is used by passing a specified direct
s~'stem. On a periodic basis, reference standards are used as
current through a conductor which in tum passes through
test objects in order to monitor the system for changes in
the ring's center. The magnetic particle testing procedure
Illagnetic field production, particle concentration, visibility
(or system) is evaluated based on the number of holes
or contamination. It is helpful to have graduated discontin-
detected at various current levels. The number of holes that
lIities. so that a numerical indicator of the system perfor-
should be detected at a particular current level is provided
lIlallCe can be recorded and monitored.
by written specifications (see Table 2).4
Standard test objec~s such as the ring have proven to be
a valuable aid in controlling magnetic particle test system
Toq' Steel Ring. Standard
The tool steel ring is a commonly used and universally
TABLE 1.· Comparative dimensions for a tool steel
n'('ogni~ed reference standard fop magnetic particle testing
ring standard (see Figure 4 J
s)sterns (see Fig. 4 and Table 1), but it essentially indicates
only particle efficacy. It appears in.viItually all US codes and Distance from Edge
'ipccifications as the means for checking magnetic particle Hole Diameter to C~nter of Hole
perfol:mance. The ring was first used in 1941 3 . Number millimeters ,inches) millimeters (Inches)
Since that time, its use has expanded for both wet and dry I 1.78/0.07) ..1.8 (0.07)
Illl'thods, to the point that the ring is widely used for 2 1.78/0.07) 3.6 (0.14)
IlwasUling system performance. It is important, however, to 3 1.78 (O.07) 5.3/0.21)
recognize the ring's limits. For example, a current density 4 1.78 (O.07) 7.1 /0.28)
level less than 20 percent of that usually applied (Table 2) is 5 1.78 (0.07) 9.0/0.35)
all that is needed to indicate a surface discontinuity 0.25 mm 6 1.78 (0.07) to.7 (0.42)
i.O.01 in.) in length. 7 1.78 (0.07) 12.5/0.49)
8 1.78/0.07) 14.2 (0.56)
9 1.78/0.07) 16.0 (0.63)
10 1.78 (0.O7) 17.8/0.70)
FIGURE 4. Tool steel ring standard (see Table 1J 11 1.78 (0.07) 19.6/0.77)
12 1.78/0.07) 21.4/0.84)

TABLE 2. Test indications required when using the

tool steel ring standard
Type of Magnetic Current· Minimum Number of
Particles Used lamperes) Holes Indicated

Wet suspension** 1.400 6

'8 2.500 7
10 '9 3.400 7
Dry powder 1.400 7
-..j f.-
22 mm (087 In.' 2.500 9
3.400 9
*FULL-WAVE DIRECT CURRENT AT CENTRAL CONDUCTOR
"VISIBLE OR FLUORESCENT
342 / MAGNETIC PARTICLE TESTING

parameters. 5 However, in addition to magnetizing current negative X direction (since J.Ll > J.LJ, or that which would he
level, other factors influence test results, including the obtained from a current dipole with a current separation
properties of the particles, operator skill, magnetization product of 21Tm pointing in the negative Y direction.
level, direction of the magnetic fields produced, and particle
concentration. An evaluation of all the contributing factors
requires the development of mathematical models 6 to de- Influence of a Boundary
scribe their effect on the formation of test indications. 7 The field from a cylindrical discontinuity next to a plane
boundary can be obtained using the method of images. The
Ring Standard Magnetic Fields field above the plane surface is the sum of the fields from a
dipole at the center of a discontinuity with strength as given
All magnetic leakage fields are a superposition of dipolar
by Eg. 3 and an infinite series of images as illustrated in
fields. This dipole character is usually evident when the field
from the discontinuity is measured. 8 ,9 The field arising from Fig. 6. To a good approximation, when h > 2a, all image
dipoles are assumed to be at the center of the cylinder. 11
a long cylindrical discontinuity in a linear isotropic medium
The field above the surface can then be described by Eq~. 1
can be exactly calculated and has a pure dipolar character.
and 2 (rather than by Eg. 3), with m given by:
Referring to the coordinate system defined in Fig. 5, the
applied field has a value of Ho and is in the X direction.
This is a classic problem in magnetostatics 10 whose m ( 2J.Ll ) [1 + (J.Ll fLo)2 (Eq.-4)
solution in the region of permeability J.Ll consists of the J.Ll + J.L o J.Ll + fLo
vector sum of the applied field Ho with a dipole field
centered at the origin. The dipole field can be written in
polar coordinates as:

cos ()
Hr -m-- (Eq. 1)
r2 Magnetic particle test materials generally have J.L t » J.Lo·
The major effect of the surface is an approximate douhling
sin () of the effective dipolar field from the first term on the right
Ho = - m - r2
- (Eq. 2) of Eg. 4 and a small increase from the second tenn (ahout
.. '7 percent when'h = 2a).
Here, m is a dipole moment per unit· of length along the
cylinder axis and is given by: . Sample Leakage Field Calculation
The calculated field from a subsurface cylindrical discon-
(Eq.3) tinuity (as obtained by Eqs. 1,2 and 4) has the familiar form
illustrated in Fig. 7. The illustration shows the components
tangential and perpendicular to the surface (the most
The field from the cylindrical discontinuity is exactly that commonly measured components).
which would be obtained from a magnetic dipole of strength
m per unit length centered at the origin and pointing in the
FIGURE 6. Dipole images for the field from a
cylindrical discontinuity next to a plane surface
FIGURE 5. Coordinate system for cylindrical
discontinuities y

• HH
~

~;r
,Hr tI
/~
/ - '-, -, ./~ (j

.(lLn,:;~) __ 1_ _.- _X
\,,, ___ ,_CYLINDRICAL
DISCONTINUITY
OF RADIUS a
 REFERENCE STANDARDS AND ARTIFICIAL DISCONTINUITY INDICATIONS / 343

The numerical values are based on the hole depth A ring that showed 5 holes at a given current denSity
(II = 3.56 mm or 0.14 in.) and hole radius (a = 0.89 mm or produced 9 holes in a repeat of the sensitivity test af-
0.03 in.) of hole 2 in the ring, with a central conductor ter annealing. Standardization efforts are being pursued to
current of 1,000 A. This corresponds to an Ho value of control this problem.
2.5 kA-1ll -1 (31.5 Oe). This calculation applies to the ideal
case of a linear isotropic magnetic medium with a high
permeability, a two-dimensional geometry (a cylindrical
discontinuity of infinite length) and a plane parallel bound- Reference Standard Test Blocks
ary. Although the test ring of Fig. 4 deviates considerably
from this ideal condition, compalison of measured leakage
fields with the form shown in Fig. 7 reveals that the leakage Split Prism Test Block
fields frolll the ring can be closely approximated by Eqs. 1
and 2. The prism block shown in Fig. 8 is another reference
standard containing an artificial discontinuity.13 Truncated
half-prisms are built with one face at an angle and when two
limitations of the Ring Standard such components are bolted together, an artificial crack is
formed. The sloped surface of the block can be positioned
A 1986 study revealed a lack of uniformity among ring at variable distances from the conductor.
standards used around the United States. 12 When tested in When current is passed through the conductor~ the
a prescribed manner, the rings were found to produce leakage field from the crack gradually weakens along the
indications anywhere from 4 to 11 holes, while indicating no prism face. A specified amperage is applied through the
difference in hardness or spectrographic analysis. The data conductor and the length of the magnetic particle indication
appear to cluster around two sensitivities, one at 5 to 6 holes is used to measure the test sensitivity.
(40 percent of the rings) 'and the other at 9 to 10 holes (46
percent of the rings).
Magnetized Test Blocks
Another version of the block standard consists of two
FIGURE 7. Calculated components of the ground steel blocks forming an artificial crack at their
magnetic leakage field at the. surface of a
subsurface cylindrical discontinuity in a linear
isotropic magnetic medium with high FIGURE 8. Prism sensitivity indicator
permeability and with the applied field in the
positive X direction

DISTANCE
finches) /

o
-0.3 -0.2 -0.1 0.1 0.2 0.3

400

o~
tJ/
-l
~Vl 2
200 uj E
l..L.u w:: ....
UJQ; UJQ;
~~ .----~-+-+_-~:_'_--___I
CJQ
~~
0
<:-
UJ
-l -2
~~
-lE
o
H" -200 ~

-~LI0------_·5------~0------~----~10
I S I
DISTANCE L· ·._/I~--.J?
(millimeters)
0"
./~
344 I MAGNETIC PARTICLE TESTING

contact surfaces (see Fig. 9), similar to the discontinuity different crack widths on its surface (see Fig. 10). A typical
formation in the split prism test block. On one of the face standard of this type is 50 mm (2 in.) in diameter and
ends, a small permanent magnet is fixed below a brass cover, 10 mm (0.4 in.) thick. VelY fine cracks are situated between
causing magnetic flux leakage from the artificial discontinu- grosser discontinuities across the block standard's face. As
ity. This leakage field decreases with greater distance from an example of this standard's use, the loss of indications for
the magnet, so that longer discontinuity indications reveal the fine cracks (or their appearance as points rather than
higher test sensitivity. lines) indicates that the magnetic particle bath is no longer
This same task can be fulfilled by another block standard usable.
using a slightly different test principle: a residually magne-
tized block is manufactured to contain a network of many
FIGURE 10. Residually magnetized block
standard
FIGURE 9. Block standard containing a
permanent magnet

FROM TIEDE GmbH. REPRINTED WITH PERMISSION.

FROM KARL DEUTSCH GmbH. REPRINTED WITH PERMISSION.

 REFERENCE STANDARDS AND ARTIFICIAL DISCONTINUITY INDICATIONS / 345

PART 3
MAGNETIC DISCONTINUITY STANDARDS

Altificial discontinuity standards and magnetic field incli- Both of these devices are used to determine the approx-
cators of the shared flux type all have some common imate orientation and, to a limited extent, indicate the
characteristics. The basic shared flux gage was developed by adequacy of field strength. However, they do not measure
Dr. B. Berthold and was first used in Germany in the late the internal field strength of the test object. They merely
1930s to indicate magnetic field direction. Refinements detect external fields in the vicinity of the test object. The
have resulted in the pie gage, widely used in weld testing presence of multiple gaps at different orientations helps
applications (specified by NavShips 271), and in the so- reveal the approximate orientation of the magnetic flux.
called magnetic penetrameter or raised cross indicator. Slots perpendicular to the flux lines produce the most
Both of these devices respond to applied magnetic fields, distinct indications. Slots parallel to the flux lines produce
regardless of the test object or lack thereof. little or nothing.
On flat ferromagnetic surfaces, the shared flux devices
respond well and verify magnetic field strengths capable of Limitations of Pie Gages and Raised Cross
.detecting sUlface discontinuities about 0.025 mm in length Indicators
and 0.01 mm deep. On complex test objects with convex
surfaces, the va,Iue of shared flux devices is limited and The raised cross indicator· is designed to ~ave a large
invariably resuits in over-magnetization when so applied. liftoff. (.~he discontinuity distance from the testing surface).
Shared flux discontinuity standards should only be used in Yet because of this, there is some question about what the
conjunction with. the continuous method of particle appli- raised cross device' actually detects during' residual induc-
cation. When the continuous method is not used, commer- tion tests. The pie gage sits closer to the testing surface and
cial pie gages may indicate a magnetic field, but this is generally performs better than a raised cross indicator at the
eVidently residual magnetism in some gages or possibly even same point. This occurs for two reasons.
physical entrapment of particles. . Both devices attract and hold magnetic particles, as
All mtificial discontinuity standards can produce results determined by the leakage fields from the test object. The
comparable to a Hall effect gaussmeter. A shim with 30 gap in the raised cross is farther from the testing surface
percent discontinuity depth and pie gages can indicate than the slot in the pie gage and it is exposed to less ambient
discontinuities 0.25 x 0.05 mm (0.01 X 0.002 in.) at 0.9 to field than the slot in the pie gage. Also, being thinner, the
1.5 mT (9 to 15 G), depending on the permeability of the raised cross gap causes less flux leakage at the slot, mainly
test object material. because such slot fields are width dependent.

FIGURE 1'. Gap indicators for magnetic field

Pie Gages and Raised Cross measurements: (a) pie gage and (b) raised cross
indicator
Indicators
fa) i~ 25mm ~i
Pie gages are disks of high permeability material divided ri========~~========~:_j~12~~.
into triangular segments separated by known gaps (see 1- I I ~l~~-~~-
Fig. lIa). The gaps are typically filled with a nonmagnetic
material, to protect the integrity of the gap and to strength-
en the disk. The testing surface is coated with a nonmagnetic
/1/111 /! ! illlill \ \

'\. . GAP 0.8 mm

Ia)'er. The pie gage contains eight segments separated by
gaps up to 0.75 mm (0.03 in.). The gaps run the full depth
of the gage.
Raised cross indicators contain four gaps (in the shape of
fb)
I ~ I
a cross) approximately 0.13 mm (0.5 in.) in width. The
segments are cut away so that the known gap is raised a fixed
distance off the test object surface (see Fig. lIb).
111/111/111111
346 / MAGNETIC PARTICLE TESTING

direction of maximum field strength and the angular toler-

Shim Discontinuity Standards ance of sensitivity. Such shim standards can be used when
setting up multidirectional magnetization, to ensure that
Shim indicators are thin foils of high permeability mate- balanced fields exist in all directions during the magnetizing
rial containing well-controlled notch discontinuities (see cycle at any given point in the test object.
Fig. 12). Frequently, multiple shims are used at different
locations and different orientations on the test object to
examine the magnetic field distribution.
Application of Shims
One popular version of the shim indicator is a strip of
high permeability magnetic material containing three slots Shims are sometimes called paste-on discontinuities be-
of different widths. The strip is placed in contact with the cause they must be attached to the test object with pressure
testing surface and shares flux with the test object. For the sensitive tape. Shims are used most often during the
purposes of producing test indications, the slots in the strip development of test procedures, where they help indicate
act as if they were cracks in the test object. A principle the relative strength and direction of a magnetic field for a
limitation of this standard is that a 50 mm (2 in.) gage length particular test configuration.
is needed. Because shims are often made of high permeability foils,
Another type of shim (sometimes called a block) has been they are generally small and flexible enough to fit into fairly
used in Japan since the 1960s (see Fig. 13). As described in complex test object geometries to help determine the
Japanese Industrial Standard G 05665, these indicators are adequacy of field strength in these critical areas.
used for examining the "performance of the apparatus, Once the field distribution is found adequate, the testing
magnetic particles and suspension, and the strength and procedure is recorded and the components are tested with
direction of effective magnetic field on the surface of the the parameters established by the shims.
test article."
The blocks are available in a variety of thicknesses and
slot depths. Linear and circular slots are available. Circular
slots are particularly effective when the direction of the FIGURE 13. Japanese type A standard test shims:
magnetic flux is not known. The shims are taped to the test (a) circular and fb) linear
object (sl9tted side in close contact with the part), in areas
where the strength and direction of the magnetic field are in
(a)
question. The slots share magnetic flux with the test object
ilnd simulate slightly subsurface discontinuities. .
An American version of the Japanese shim is manufac-
tured in 0.05 mm (0.002 in.) and 0.1 mm (0.004 in.) thick-
nesses, with discontinuity depths of 15, 30 and 60 percent of
the shim thickness. The controlled discontinuities are avail-
able as either linear grooves or circular grooves. The circular
discontinuities have the added advantage of indicating the

FIGURE 12. Shim indicators for magnetic field

1•
20mm 108 in)
-\
verification: (a) simple shim and (b) slotted strip (b) - -

fa)

11111111//111/ - --i r-
-
s:: 6mm
10.15 in)
<Xl
2-
E
NONMAGNETIC E
fb) I 17 MATERIAL
0

'"

/-1/~/~II~/~/~II~/~/~/~I/~
I- 20 mm 10.8 in)
-I
 REFERENCE STANDARDS AND ARTIFICIAL DISCONTINUITY INDICATIONS I 347

PART 4
ELECTRONIC REFERENCE STANDARDS

degree of magnetization. Because of poor repeatability, few

Hall Effect Meters of these procedures are widely used. The repeatability
problem stems from the large number of variables that can
Hall effect devices are commonly used to measure the affect eddy current response in a ferromagnetic material.
strength of the magnetizing force tangential to the surface A magnetization level indicator has been developed to
of a test object. Though often called a gaussmeter, the detect imbalance in the permeability along the lines of flux
device does not actually measure the magnetic flux B within compared with the permeability transverse to the lines of
the component. Rather, it measures the magnetic field flux as the material approaches magnetic saturation.
strength H adjacent to the test object. The Hall effect meter
is a relatively effective indicator and is in widespread use for
establishing magnetic particle testing procedures. They Conclusion
effectively measure residual fields and indicate the direction
of the remanence. MagnetiC particle testing historicaHy relied on empirical
Vari~)Us specifications. call for the use of different tesla. '.. guidelines for the' development of test procedures. This
(gau9S) or ampere per meter (oersted) values in particular practice led to widely varying dis<;ontinuity detection capa-
applications. In air, the non metric gauss and oersted units bilities. Various codes a~d specifications have perpetuated
are numerically equal in value. Required values commonly the problem by citing nIles of thumb for establishing
range from 1.6 kA-m -1 (20 Oe) to 4.8 kA-m -1 (600e) procedures.
when the residual method is used. A residual field less than Several forms of reference standards are available for
240 A-m -1 (3 Oe) usually does not attract conventional verifying procedures and for evaluating the performance of
magnetic particles. a magnetic particle testing system (as a whole or by
components). Reference standards for this purpose are
usually made of high permeability materials containing
controlled graduations of artificial discontinuities. Refer-
Eddy Current Devices ence standards are also available for evaluating the effec-
tiveness of a magnetic particle test for a particular test
The ability of a material to store electromagnetic energy object. These devices include a variety of shim configura-
in the form of eddy currents is a function of both the tions containing known discontinuities.
conductivity and permeability of the material. Because the Electronic reference standards such as Hall effect meters
permeability of a ferromagnetic material changes as the and eddy current devices can also be used to evaluate the
material is magnetized (from a relatively low initial perme- adequacy of a test procedure.
ahility through a higher maximum value), the eddy current Widespread use of reference standards and test discon-
coil impedance also changes. tinuities is needed to improve the consistency of magnetic
Several eddy current procedures have been developed to particle tests and to increase the detection reliability of the
detect this change in permeability and thereby indicate the method.
348 / MAGNETIC PARTICLE TESTING

REFERENCES

1. Lewis, W.H., W.H. Sproat, B.D. Dodd and J.~'1.

Rings. Proceedings of the Twelfth Symposium on NOl/-
Hamilton. Reliability of Nondestructive Inspections _
destructive Evaluation, ·W.VV. Bradshaw, ed. San Anton-
Final Report. Report No. SA-ALC/MEE 76-6-38-1.
io, TX: Southwest Research Institute (1979).
Kelly Air Force Base, TX: San Antonio Air Logistics
Center (1978). 8. Lord, W., J.M. Bridges, vv. Yen and R. Palanisam:i,
"Residual and Active Leakage Fields Around Defects in
2. Rummel, W.D., D.H. Todd, Jr., S.A. Frescka and RA. Ferromagnetic Materials." Materials Evaluation.
Rathke. The Detection of Fatigue Cracks by Nonde- Vol. 36, No.7 (July 1978): p 47.
structive Testing Methods. NASA CR-2369. National 9. Foerster, F. Developments in the Magnetograp~y of
Aeronautics and Space Administration (1974). Tubes and Tube Welds. Non-Destmctive Testing. Vol. H
(1975): p 304.
3. Betz, C.E. Principles of Magnetic Particle Testing.
Chicago,IL: Magnaflux Corporation (1967). 10. Stratton, J.A. Electromagnetic Theory. New York, NY:
McGraw-Hill Book Company (1941): p 258. .
4. Magnetic Particle Inspection. MIL-STD-1949. Wash- 11. Shcherbinin, VE., and M.L. Shur. Calculating the Ef-
ington, DC: Department of Defense (1985). fect of the Boundaries of a Product on the Field of a
5. Gregory, C.A., VL. Holmes and RJ. Roehrs. "Ap- Cylindrical Defect." Soviet Journal of NondestructitY'
proaches to Verification and Solution of Magnetic Testing. Vol. 12 (1976): p 606.
Particle Inspection Problems." Materials Evdluation. 12. Hagempjer, D.J. Magnetic Particle Ketos Ring Standard
Vol. 30, No. 10 (October 1972). Evaluation. Report LR-11711. Long Beach, CA: Doug-
6. Beissner, RE. An Investigation of Flux Density Deter- las Aircraft Company. (April 1986). .
minations. Report AFML-TR-76-236. San Antonio, TX: 13. Sheli~ov, G.S. and A.G. Aleksandrov. "Coagulation of
Southwest Research Institute (1976). the Particles in a Magnetic Suspension and its Influence
on the Sensitivity of the Magnetic Powder Method of
7. Swartzendruber, L. "Magnetic Leakage and Force
Testing." Soviet Journal of Nondestructive Testing,
Fields for Artificial Defects in Magnetic Particle Test Vol. 13 (1977): p 26.
 SECTION

EQUIPMENT FOR MAGNETIC

PARTICLE TESTS
Larry Haller, NOT consultant, La Habra, California
Stanley Ness, consultant; Mission Viejo, California
K~r.mit Skei~, NOT cC!nsultant, H.acienda Heights, California
 350 / MAGNETIC PARTICLE TESTING

PART 1
BASIC EQUIPMENT CONSIDERATIONS

Magnetic particle testing equipment can be as small as a Test Sensitivity

handheld yoke or as large as the billet testing units found in
In some high production applications, the magnetic
steel mills. Magnetic particle systems have evolved dramat-
particle testing system is used as a product screening device:
ically from the relatively small and simple units first pro-
duced. The primary improvements parallel developments in any indication of a discontinuity becomes cause for rejec-
tion. Once removed from the production line, the rejected
other technologies and include faster electronic Switching,
automatic current control and improved materials for par- object may then be reevaluated or reworked as n~edecl.
In applications such as aerospace or plant maintenance,
ticles, coatings and vehicles.
magnetic particle tests need to be very sensitive. Extremely
In its fundamentals, the magnetic particle testing tech-
small discontinuities must be detected and correspondingly
nique has not changed much since its conception in the
small test indications must be produced for evaluation and
1930s. However, Significant changes have occurred in three
areas: (1) the magnetic particle materials; (2) the configura- interpretation.
tion of testing systems; and (3) the automated components For some applica.tions, high sensitivity is actually a prob-
. lem: excess leakage .fields produce false indications and
. designed to meet contemporary manufacturing and produc-
dense backgrounds. The magnetic particle test sensitivity
tion needs.
must be established in order to indicate discontinuities
Magnetic particle test systems must fulfill two basic
requirements: (1) to accurately perform a nondestructive within a range of severity appropriate to the application.
test based on amperage requirements, test object size,
magnetic field levels and suitable testing area; and (2) to Configuration of the Test System
perform the test with or without operator intervention at a
The follOwing list is a summary of the choices and
rate required by the particular production facility.
considerations that determine the configuration of a mag-
In turn, these two requirements determine the size,
netic particle testing system for a specific application:
shape, speed and configuration of the magnetic particle
testing system.
1. particle type (wet or dry);
2. magnetization requirements of the test object;
Effect of Testing Parameters on 3. degree of automation required;
4. demagnetization requirements;
Equipment Choice 5. current requirements;
6. test object size and corresponding test system size;
Production Speed 7. electrical power availability;
8. air supply requirements;
Some magnetic particle testing systems operate at slow
9. accessories needed for the application; and
production speeds, often as low as a few large components
10. test specifications requiring verification.
per day. In other instances, the testing equipment is an
integral component of a production line, processing and
testing hundreds of objects per hour. Each of these ten considerations is affected bv lIIany
If only because of handling considerations, these two other testing or manufachuing parameters. MagnetizatiOI~'
applications require completely different types of testing for instance, may need to be achieved with alternating
systems, designed and built to accommodate their specific current, half-wave rectified alternating current, full-wave
uses. rectified alternating current or Single-phase rectified alter-
 EaUIPMENT FOR MAGNETIC PARTICLE TESTS I 351

nating curreht, depending on the nature of the test object

and the purpose of the test. As another example, demagne- understanding is then applied to the specific requirements
tization requirements are determined by the test object's of the expected test objects and their anticipated discontin-
subsequent use and the magnetization method. In addition, uities. The test object's magnetic characteristics, its geom-
demagnetization procedures may need to be incorporated etry and its intended se'rvice are all important factors that
into the main testing system or performed at a separate affect the choice of the magnetic particle system. Finally,
station. the test system must be configured and installed so that it is
an integral part of the existing production facility.
Choosing the correct magnetic particle testing equipment
is the last of several preliminary decisions. First, there must The text that follows is a generalized view of magnet-
ic particle testing systems and can serve as a guide for mak-
be a thorough understanding of the magnetic particle
technique, especially its capabilities and limitations. This ing the more specific decisions required by each testing
application.
 352 / MAGNETIC PARTICLE TESTING

PART 2
WET HORIZONTAL EQUIPMENT

The magnetic palticle equipment most commonly used

FIGURE 1. Typical horizontal magnetic particle
for production testing is the wet horizontal unit (see Fig. 1).
bench testing units: faJ direct current unit with
The name is derived from the unit's use of the wet method
split magnetizing and control functions for two
test techniques and its horizontal bed for positioning the
operators; and fbJ small alternating current unit
test object.
with decay magnetizing
The nominal length of such a unit is determined by the
size of the object that can be fixed inside its clamping
subsystem. Lengths of 1 to 4 m (3 to 12 ft) are used for most
applications. Many other system lengths have been de- (a)
signed, some for objects as sn1a11 as aerospace bolts a few
millimeters in length. On the other end of the scale, very
long systems have been built for testing steel billets, gun
tubes, oil field pipe or railroa? engine crankshafts.

Positioning the Test Object

Before performing magnetic particle testing, the object is
·.c!amped between a headstock and an adjustable tail stock
. that moves horizontally along rails in the unit's bed. .'
The headstock holds the test object by means of a
compressed air cylinder. An electrically operated sWitch,
often a foot switch, is the common method for controlling
the throw of the headstock air cylinder.
The tailstock's position may be controlled by a gear screw
and hand crank or, in some smaller units, the tailstock is
simply pushed into position along the bed. Some systems (b)
use motor driven tailstocks but these are slow and are
normally used only in special applications.

Magnetization Procedures
A wet horizontal testing system is usually capable of
producing circular magnetization by the direct contact
method and longitudinal magnetization with an encircling
coil or a yoke configuration.
Once the test object is clamped into position, electrical
current is passed through it or through a conductor for
setting up an induced field. For longitudinal magnetization,
most horizontal systems have a rail-mounted magnetizing
coil that may be moved along the horizontal axis of the
machine. \.vhile the test object is being loaded, the coil is
moved clear of the headstock and tail stock opening. After
circular magnetization, the coil is moved to encircle the
object. On test objects over 450 mm (18 in.) in length, the
 EQUIPMENT FOR MAGNETIC PARTICLE TESTS I 353

encircling coil must be repositioned in 350 to 425 mm (14 to vVith direct current magnetization, the current and the
17 in.) increments. \Vith yoke procedures, longitudinal resulting field follow the path of least resistance, which is
fields are set up along the entire length of the test object, not necessmily the location of discontinuities. Therefore,
vvith no need for repositioning. A suitable reference stan- alternating current equipment is often chosen because of an
dard with known discontinuities should be used to ensure improved probability of detection. In addition, an equiva-
adequate field strength at the midpoint of the test object. lent relative field strength in an alternating current system
Following each magnetization procedure, the test object requires only about 50 percent of the current used by a
is examined using wet method magnetic palticle techniques. direct current system.
Large test objects remain in position on the bed of the unit.
Ivlany wet horizontal testing systems are also equipped for
demagnetization procedures.
Wet Horizontal System Components
The rated magnetizing output and duty cycle of wet
Alternating and Direct Current horizontal systems vary with the model and manufacturer.
Alternating current equipment usually has a maximum
Systems output of 1.5 to 3 kA with some rated as high as 6 kA.
Full-wave rectified alternating current equipment usually
\'Tet horizontal units used for in-service and in-process
has a 2, 4 or 6 kA rating. For larger test objects, long systems
testing are usually Single-phase alternating current systems.
can be manufactured with 10 kA or more direct current.
Alternating current equipment is less expensive to manu-
,Table 1 lists the components of a typical wet horizontal
fac~ure because costly rectifying circuitry and associated
magnetic particle testing system.
cooling systems are not required. In addition, a separate
d,emagnetizing unit is not necessary wit~ alternating current
units and test objects are demagnetized in position on 'the Test Systems tJsed Outsid~ t6e United States
unit. Magnetic particle testing systems used outside the US
Aerospace codes and many manufacturing standards spec- differ in several ways: First, because alternating current is in
ifY the use of direct current magnetic particle testing more widespread application, nearly all systems are de-
equipment. These are usually three-phase full-wave recti- signed for use with Single-phase input. Where applicable,
fied alternating current systems. In the early days of the rectified circuitry is generally half-wave rectified alternating
magnetic particle method, direct current was supplied by a current. Single-phase equipment is appreciably less costly to
hank of acid storage batteries that were often inefficient and build than the three-phase rectified systems used in the US.
unreliable. In 1941, the rectifier circuit was developed, Single-phase equipment must have primary current double
producing unidirectional current from alternating current that of a three-phase system for the same secondary or
and replacing the troublesome battery bank. magnetizing current output.
The main advantage of three-phase rectified alternating
Comparison of Alternating and Direct Current current is that the demanded theoretical power usage is
lower. However, because most testing applications have a
Equipment
10 percent duty cycle (0.5 seconds on and 5.0 seconds off),
There are many differences between the basic circuitry in the saving in power is comparatively inSignificant. Single-
direct and altemating current magnetic particle testing phase magnetizing current provides far greater particle
svstems. mobility for dry and wet particles.
- For example, direct current equipment often has a The second important difference, especially in European
current timer that is preset for a half second duration so that systems, is the nearly universal use of yokes rather than coils
a low duty cycle can be used. The timer may be bypassed for developing a longitudinal field in test objects. Many
manually and in contemporary units it is adjustable. Test European systems have no magnetizing coil at all mounted
systems using fully rectified alternating current require on the unit.
augmented cooling. Fans help fulfill the cooling require- A third difference is that most of the electrical or
ments of the rectifier circuitry. Such systems mav offer electronic components are separated from the mechanical
demagnetization using reversi~g direct ~urrent with step or handling pOltions of a system. In many countries, this
dmvn procedures or altern~ting current demagnetizing cir- separation is a code requirement with safety to personnel as
cuilI)' with rapid decay. justification. Note that since 1950, no instance of electro-
Because alternating current produces a skin effect (with cution has been traced to a magnetic particle apparatus,
penetration depth dependent on line frequency), the result- mainly because secondary (magnetizing) circuits produce
ing magnetic field follows the contour of the test object. relatively low voltage.
354 / MAGNETIC PARTICLE TESTING

TABLE 1. Components of a typical wet horizontal

magnetic particle testing unit Multidirectional Test Systems
Multidirectional testing systems should provide magne-
• Hood: encloses the unit for ultraviolet light tests tizing current in two or more directions. Fluorescent mag-
• Instrument pedestal: holds meters and other controls netic particle suspensions are predominantly used ""ith
• Ammeter: measures amperage of direct current multidirectional magnetization.
• Adjustable timer: controls elapsed time of magnetizing Conventional magnetization in one direction is accom-
current plished in a multidirectional unit by energizing a specially
• Visible light source designed circuit. In addition, one, two or three circuits can
• Switches: fan and visible light controls
be individually energized in rapid succession. These quickly
• Coil: for longitudinal magnetization
changing magnetizing currents produce overall magnetiza-
• Contact plate: for circuit connection at tailstock
tion of the test object, allOwing complete coverage of
• Tailstock: adjustable to test object length
• Crank: adjusts tailstock position
potential discontinuity areas.
• Curtain: encloses hood to restrict visible light Multidirectional test systems may be designed for very
• Actuator bar: triggers actuator to operate magnetizing specific applications: the steel blades in jet engines, t()[
current example. For this particular test, the third phase of three-
• Transfer switch: selects coil or head shot magnetization phase alternating current is used to provide a circular field
• Push button: magnetizing current start control directed through the dovetail of the blade. The circular field
• Current control: selects current levels adds to a longitudinal field supplied by a coil, giving
• Foot switch: controls air pressure valve for headstock increased magnetic field strength in the blade's critical
contact plate dovetail area.
• Actuator: connected to actuator bar to operate magnetizing Theoretically, multidirectional magnetization can be ap-
current plied in a majority of production applications with improve-
• Power indicator: .indicates when power supply to unit is ments . in resolution and cost. 'Suitable reference standards
active
with known discontinuities must be used for set lip of
• Pump switch: activates pump for magnetic particle
suspension· .
multid'irectional tests. When properly applied with assur-
• Shelves: headstock and tailstock contact plates to support ance of adequate field and directional balance, improve-
test objects ments in discontinuity detection can be as high as :30
• Hose: delivers magnetic particle suspension percent over typical unidirectional applications. Reduction
• Nozzle: applies magnetic particle suspension to test object in labor costs can exceed 60 percent when compared to
• Contact plate: for circuit connection at headstock conventional test procedures on wet horizontal equipment.
• Headstock: supports air cylinder that operates contact plate Time studies have indicated that forgings weighing 2.2.5
• Swing arm: allows pOSitioning of ultraviolet light to 9 kg (5 to 20 lbs) require about 35 percent less time for
• Ultraviolet source multidirectional tests because of reductions in handling amI
testing times. Test objects with multiple apeltures are also
likely candidates for significant reductions in labor costs
when tested with multidirectional methods. Savings in labor
up to 90 percent have been reported on tests of plates in
shipbuilding through the use of heavy duty multidirectional
magnetizing power packs.
 EOUIPMENT FOR MAGNETIC PARTICLE TESTS I 355

PART 3
STATIONARY MAGNETIC PARTICLE
EQUIPMENT
Many different stationary bench units are available for by the line frequency. In this way, the test object is both
specific magnetic particle testing applications. The size of circularly and longitudinally magnetized. Switching trig-
the test system is determined by the size of anticipated test gered by the line frequency allows the field direction to vary
objects. Larger models are commonly used for heavy com- at vel)' high speed.
ponents such as diesel engine crankshafts, landing gear
sections or gun ·barrels.
Stationary magnetic pmiicle units are generally designed Pulsed or Capacitor Discharge Units
to operate from a 440 V three-phase alternating current
source and to deliver alternating or rectified magnetizing Very high magnetizing currents can be obtained over
current. Current control is often infinitely variable. In older short periods of time by discharging a capacitor. The main
units, the current was manually controlled with a step advantage of this procedure is that it can be done with a
switch. Because of expanding requirements for close and relatively small unit.
repetitive control over current density, filtering of the power When the capacitor is discharged through the test object
source may be necessary. or through a coil, a strong magnetic field is set up. The exact
magnetizing current may be determined with a peak and
duration meter and, with th~ advances in solid state tech-
nology, it is possible to accl;lrately cOlilt'rol. pulse amplitude
Direct Cl:lrrent ·Magnetizing and duration. The duration of the magnetizing current is
Equipment less than 500 ms.
For longitudinal magnetization, the pulse can be applied
For magnetic particle' testing of large, complex castings, to a yoke with the test object clamped between a headstock
welded structures or plate, overall magnetization with high and tail stock. This is an important advantage because it
magnetizing current is usually employed because of cost eliminates the possibility of arcing.
savings. The rated maximum output values for such appli-
cations is usually around 12 kA.
Multidirectional magnetization if often used through two Quick Break Magnetization
or three magnetizing circuits, making it possible to detect
discontinuities in all directions in a setup. This multidirec- A quick break magnetization feature is needed in three-
tional magnetization is usually done by electromechanical phase full-wave rectified systems using coils for direct
s\vitching. Better results can be obtained when the switch- current magnetization. When an object is placed in a coil
ing is done electronically, rather than electromechanically. and magnetized, the field lines leave the test object in the
Electronic Switching prOvides faster rise time and the ability vicinity of the north pole and reenter near the south pole.
to program Switching cycles for more reliable performance. The lines of force may be normal to the surface at the ends
of the test object. A discontinuity near the end of the object
may therefore be in a very unfavorable position for detec-
tion by magnetic particle techniques.
Automated Magnetic Particle In order to overcome this, the direct current applied to
Systems the coil is quickly turned off. The rapid collapse of the
magnetic field creates low frequency eddy currents \vithin
Automatic or semiautomatic magnetiC particle testing in the object in a direction favorable for the detection of
most cases requires magnetization in two directions to transverse discontinuities at the ends of the object.
detect randomly oriented discontinuities. Since two mag- The use of a yoke or the field flow method makes qUick
netic fields cannot exist simultaneously in one test object, it break magnetization unnecessaJ)' since the test object is part
is necessary to switch from one direction to the other. of the magnetic circuit (see Fig. 2). Periodic checking of the
Electro~ic Switching provides advantages in this applica- break is critical - on electronically higgered equipment, a
tion, as it does in other system configurations. Current can malfunctioning firing module could result in evidence of
be switched several times per second and is often triggered quick break failure.
356 / MAGNETIC PARTICLE TESTING

FIGURE 2. Adjustable electromagnetic yokes for

magnetic particle testing
 EQUIPMENT FOR MAGNETIC PARTICLE TESTS / 357

PART 4
MOBILE MAGNETIC PARTICLE
EQUIPMENT

Mobile magnetic particle testing systems have outputs up Mobile equipment is widely used for many types of
to 20 kA and may be designed to deliver alternating current, magnetic particle testing. The advantage of these system
direct current, half-wave rectified or pulse current. Table 2 configurations is that they can be moved to the test site,
lists the components of a typical mobile magnetic particle whether that is a flight line, a refining tank, large plant
test unit. In older systems, the current was adjusted with a machinery or a structural steel weldment. The advantage of
thirty-point step switch. Newer units are equipped with, a mobile system's high amperage is its ability to inspect large
infinitely variable current controls. Push button demagne- castings, forgings, welds or any other test object requiring
tization is often incorporated into contemporary designs. strong magnetizing field strengths.
In Europe, mobile systems are used as power packs for
stationary testing units. This is a safety require,P1ent (install-
ing heavy duty electrical equipment under a kerosene or Current and Voltage Parameters
water suspension tank is prohibited).
For cenain ·applications, prods or clamps are ~sed with
TABLE 2. Components of a typical mobile magnetic mobile magnetic particle equipment. A solenoid (or a cable
particle testing unit wrapped into a coil) may also be used when longitudinal
magnetization or demagnetization is desired. A length of
• Lifting hooks: For positioning the unit with a crane cable can serve as an internal conductor for magn~tization
• Current control: for <;!djusting alternating or direct current procedures or, with clamps, a ferromagnetic bar can be
strength used.
• Line switch: for power to unit Mobile power packs operate on 230 or 460 V single-phase
• Power indicator: indicates power to unit is on alternating current. Most units manufactured through the
• Remote switch: allows use of remote controls 1970s had a maximum output of 2 or 3 kA of half-wave
• Common connector: either-end connector for one side of alternating current or Single-phase alternating current. Sys-
output current tems are now available with 4 and 6 kA half-wave alternating
• Power outlet 120 V for accessories (powder blower,
current and single-phase alternating current outputs or a
grinder, light)
• Half-wave connector: with common connector, supplies range of pulse outputs.
half-wave alternating current to cables In a typical system, selection of alternating or direct
• Alternating current connector: with common connector, current is at the option of the operator using a common
supplies alternating current to cables terminal on the front panel.
• Cables: 4/0 cables for supplying current to prods*
• Connector: either-end slip joint connector for additional Cabling Parameters
cable or accessories
• Remote control: for stopping or starting magnetiZing and Cable length can be varied and many applications require
demagnetizing current lengths exceeding 30 m (100 ft). Mobile system cables are
• Casters: two swivel and two fixed casters usually fitted with slip joint, either-end connectors. Cables
• Remote outlet: for remote control are connected to the test system, prods, clamps or to each
• Current light: indicates current is on for output circuit other (for lengthening the magnetizing circuit). Some high
• Ammeter: indicates amperage in alternating or half-wave amperage systems use bolted terminals to permit the
current circuits passage of high currents without overheating.
• Louvers: ventilation for electrical equipment A 4/0 (12 mm diameter) flexible rubber coated cable is
• Recess: for cable and accessory storage most commonly used for mobile magnetic particle applica-
"4/0 cable Is 12 mm diameter copper equivalent
tions. For easier handling, 210 (9 mm diameter) cabling is
sometimes used for connections to prods. With short cables
3581 MAGNETIC PARTICLE TESTING

5 to 10 m (15 to 30 ft), mobile equipment delivers close to procedures. A second 120 V outlet is sometimes provided
its rated amperage. As the cables (magnetizing circuits) are for remote power to trouble lights, powder blowers, grinders
extended, the amperage available at the test point is and other accessOlies.
considerably decreased. For example, a 4 kA unit with two
30 m (100 ft) cables in the magnetizing circuit delivers
about 1 kA at the test point, because of internal resistance of Demagnetization Procedures
the secondary circuit. Demagnetizing can be accomplished in several ways with
The input power cable to a mobile unit can be almost any a mobile system. The choice of technique depends on the
convenient length with little or no loss in output. Often the object being tested and the means of magnetization. Small
input cable is fitted with a heavy duty contact plug for objects can be demagnetized by selecting alternating cur-
connection to 230 or 460 V outlets. These outlets are rent and forming a coil with the cables. The object is drawn
commonly available for welding equipment and other heavy through the coil while the coil is energized. The test object
machinery. Mobile magnetic particle equipment may also must be moved sufficiently far from the coil to outdistance
be operated from alternating current generators. the coil's magnetic field, usually about 1 m (3 ft).
As the cables and connectors become worn or subject to For larger test objects, demagnetization is achieved by
overheating, their electrical resistance increases. Note that touching the smface with the energized coil and then
this ,:"ear also decreases the available amperage. moving the coil away from the test object.
An alternate method of demagnetization uses demagne-
Input and Output Current Requirements tizing current flowing through the test object. Cables are
connected to the object with clamps. With solid state
Mobile magnetic particle testing units can be operated on equipment, the decay current or ramp current reducing
230 or 460 V single-phase alternating current. Changing the system is used. With older equipment or tap switch units,
operating voltage is easily done through an access port that demagnetization is done with the step down system. An
exposes the terminals. Jumpers are alternately positioned to internal conductor can also be used for demagnetization if
change the transformer connections and vary the input the test object geometry lends itself to such procedures.
requirements. With either voltage input, the output rating is
the same.
Mobile test systems with 6 ,kA output are not available
with 230 V input and, because of high primary current
requirements, must be operated on 440 to 460 V alternating Mobile Capacitor Discharge Systems
current. Many systems have tap switches to regulate the
current by selecting one of eight or one of thirty taps to the Capacitor discharge systems are used for circular magne-
transformer. This selection determines the current output tization of long hollow test objects and the ends of massive
for either alternating current or half-wave alternating cur- steel components, such as the thread connections of drill
rent. pipe and collars. These systems consist of 2 to 4 F boxes
For demagnetization, mobile magnetic particle testing charged to a voltage of 50 to 75 V. This low voltage is used
units use a decaying alternating current or a ramp demag- because of safety regulations for unshielded conductors.
netizing system. The capacitor bank, once charged, is discharged either
through a 410 cahle coil or an insulated aluminum rod. The
insulation prevents contact with the inside sUl'f~lce of the test
object to avoid arc bums. Peak amplitude and pulse dura-
Operation of Mobile Testing Units tion are recorded for each shot and are then compared with
published specifications to ensure that the objects are
magnetized to 90 percent of remanence (0.9 Br)' Second
Remote Operation Capabilities
and third shots may often be needed to produce this high
A remote control cable may be connected to a four-point level of magnetization
outlet. This allows the operator to control the mobile unit Following magnetization, the objects are inspected for
from the actual testing location at the ends of the cables. discontinuities either by wet or dry particle methods. A
vVhen using prods, a microswitch for control1ing current typical capacitor discharge system includes the following
flow may be located in the prod handle. With clamps, coils components: (1) a 2 to 4 F bank charged to 50 to 75 V;
or an internal conductor, a special remote switch station is (2) current pulse filing control; (3) a 15 m (5 ft) insulated
available or the prod switch may be used. aluminum rod; «1) 4/0 (12 mm diameter) cabling for con-
Solid state units have a remote control station that also nection to the rod and for coil wrap; and (5) meters [or peak
allows current selection, on-off option and demagnetization current and pulse duration.
 EQUIPMENT FOR MAGNETIC PARTICLE TESTS I 359

Prod tips are another source of concern. They may

Maintenance of Mobile Testing become corroded or burned, hindering good contact with
Systems the test object. Defective prod tips can produce arc bums
and these are proven sources of cracking. Similarly, clamps
should have good copper mesh gripping points.
Some critical maintenance procedures are required for Frayed cables and continuous overheating decrease the
mohile magnetic particle testing systems. For instance, the conductivity of the cable. This produces heat, increases
cooling intake of the unit must be kept clean to permit the resistance and reduces the amperage available for testing.
free How of air, especially if the unit is moved into a dilty Connectors and cable joints should be tight and solid to help
testing environment. prevent overheating.
360 / MAGNETIC PARTICLE TESTING

PART 5
PORTABLE MAGNETIC PARTICLE
EQUIPMENT

Current and Voltage Parameters

Handheld Equipment
Portable systems vary greatly in electrical parameters.
The simplest and perhaps most common magnetic pmti- Lighter weight units are usually deSigned to operate on
ele test system is a handheld magnetic yoke (see Fig. 3). For 120 V single-phase alternating current. Output amperage of
small test objects (automotive parts, for example) surface these units can vmy from 400 to 900 A, depending 011 the
discontinuities can be reliably detected at 10\,,' production model and manufacturer. Because the usual limit on U,S V
rates \\lith a portable magnetic yoke. Yokes are also used for is :30 A, outputs over 1.0 kA require a 230 or 460 V IInit.
mclgnetic particle tests of welds, especially when arc strikes Most portable magnetic particle testing systems contain
cannot be tolerated. It is difficult to prevent arc strikes when half-wave rectified altemating current and altemating cur-
current is applied directly with prods. rent output.
Yokes operate on altemating current at standard line Either two or three connectors are provided for dual
voltage, either 120 V or 220 V. For areas where a shock output systems: a common connector, one connector for
hazard exists, yokes can be made to. operate at 42 V. The half-wave rectified and a third connector for alternating
typical electromagnetic yoke has articulated legs that assist current. Output is controlled with a tap switch or a poten-
positioning on complex shapes. tiometer. A Single ammeter selves to measure both tht'
Yokes 'With four poles have been developed in Japan and alternating cqrrent and the half-wave. output.
are in use in several pcuts of the world, although they are not
often used in the US. By switching the. poles in pairs at right
angles to each other, discootlnuity detection in all directions Accessories and Components
can be performed in a single setup.
Table 3 and Fig. 3 illustrate the compOllents of a portable
magnetic pmticle test system.
Accessories such as prods, clamps and cables for portahle
systems are similar to their counterpaIts for mobile testing
units. Cable lengths for portable systems are usually limited
to 10 m (30 ft) or two 5 m (1.5 ft) cables.
Portable System Configurations \Vith the pOltable systems, demagnetization is aCCOIll-
plished using (1) manual altemating current step dowll or
Larger equipment is needed when a higher magnetizing (2) a coil formed from cables using alternating current.
current or a higher duty cycle is required. r..lagnetic pmticle vVhen step down demagnetization is used, connection to the
test systems often require heavy transformers and, exclud- test ohject is made using clamps or an internal condllctor.
ing the magnetizing cables, can eaSily weigh 34 kg (75 Ib), A 120 V coil is a useful accessOlY f()l' prodUCing longitu-
For that reason, portable units are sometimes mounted on dinal magnetization. The coil consists of many turns of fine
carts and become, by definition, mobile units. wire; a switch is used to close the circuit when the coil is
POltable systems may operate from a 11,5, 230, 460 or connected to 120 V alternating current. The coil's plilllary
380 V single-phase source. Magnetizing output currents use is f<)f testing elongated objects such as spindles or ,edt'S.
range from 400 to 2,000 A for alternating current or The coil can also be used for demagnetizing many' kinds of
half-wave rectified alternating current applications. Dual test objects.
prods or clamps are used for direct COli tact magnetization. A caiJ(lcitance discharge pOltable system is also availahle.
Most cables have either-end connectors, aUmving the oper- Though it operates much like a standard portable unit, it
ator to manually form a magnetizing coil from a standard weighs considerably less and can use lighter, smaller mag-
cable. netizing cables. The input voltage is loyv in capacitance
Portable magnetic p;uticle equipment is most often used discharge units, providing the additioJlal henefit of in-
for testing welds, hut it is not restricted to that application. creased operator safety.
 EOUIPMENT FOR MAGNETIC PARTICLE TESTS / 361

FIGURE 3. Portable magnetic particle testing TABLE 3. Components of a typical portable

equipment: laJ direct or alternating current magnetic particle testing unit
model with infinite current control;
(bJ lightweight pulse portable operating from • Current control: adjustable or tap switch
120 V outlet; and IcJ adjustable yoke, 120 V coil • Cables: extra flexibility required for connection to prods
(for magnetizing or demagnetizingJ and assorted • Prods: in-line components for making electrical contact
accessories with test object
• Handle: for positioning and transporting unit
• Half-wave connector: with common connector, supplies
fal half-wave alternating current to cables
• Remote receptacle: for remote cabling hook-up
• Common connector: either-end connector for one side of
output current
• Power indicator: indicates when power supply to unit is on
• Alternating current connector: with common connector,
supplies alternating current to cables
• Control cable: connects to prod remote switch
• Ammeter: measures alternating and half-wave current
output

Yokes
Magnetic particle yokes have a muItiturn coil wrapped '.
around an assembly of soft iron laminations. A power cord
and switch are patt of the typical yoke design. In service, the
yoke is placed on the test object and energized. Magnetic
fbI particles are applied between. and adjacent to the ends of
the yoke. A magnetic field is set up between the ends (poles)
of the yoke. To find discontinuities in both directions, the
yoke is rotated 90 degrees and the operation is repeated.
The Oliginal alternating current yokes had a fixed U shape.
A more recent and useful design has the ends of the
U jOinted so that wider or narrower objects can be tested
more effiCiently. A rectifier is sometimes used in conjunc-
tion with the yoke, providing a direct current field for some
penetration below the test object surface.

leJ
362 / MAGNETIC PARTICLE TESTING

PART 6
MAGNETIC PARTICLE TESTING POWER
PACK SYSTEMS

A power pack is used to set up magnetizing fields when an Automotive Industry Applications
object is too big for the available statiomuy testing unit.
In the automotive industl)" important components sitch
~Iost magnetic particle systems used in contemponuy pro-
as spindles, connecting rods and crankshafts are routinely
duction environments have a power pack as part of their
tested with magnetic particle techniques. System configu-
design. The electrical components of the magnetizing and
ration is based on the use of a power pack to supply the
demagnetizing circuitry are generally part of the power
proper magnetizing current. Timing circuits are a \,ita! p,ui
pack. Timing circuits adapted to the particular productibn
of this power pack application: magnetizing current must be
line are also included in the power pack.
applied at a specific time dUting the test object's conveyance
Power pack timing circuits are designed for specific
through the system.
applications. This might be hundreds of ohjects per hour in
Magnetization may be circular or longitudinal and is
the automotive industry or an object per minute in the steel
routinely followed by demagnetization. For small test ob-
industry. In other applications, the power pack could have
jects, demagnetization is achieved using an altt'mating
timers that control the length of the cnrrent application. in
current coil encircling the conveyor. On some systems, the
order to meet product specifications. Power pack design is
operator drops the test object through a coil after inspec-
·usually dictated by the production requirements of the
tion. For large test objects, special contact demagnetizing
products being tested.
circuitry is deSigned as an integral p,ui of the power pack.

Power Pack Applications

FIGURE 4. MagnetiC particle system for steel mill
Steel Industry Uses product testing: fluorescent test unit showing
In the steel industry, magnetic particle testing is applied square billet turner and inspection station
to products as diverse as billets, blooms, bars and tubing
(see Fig. 4). Magnetic particle technology is used to test
semifinished products and is sometimes applied at an
intermediate stage of manufacture for process control.
A steel mill testing system typically uses a power pack that
supplies magnetizing currt;nt to contacting fixtures. De-
pending on the test ohject's size, amperage as high as 20 kA
full-wave direct current may be used. The test system is like
a large bench unit, featuring a meallS of magnetically
contacting the test object after positioning. Steel mill test
equipment includes handling systems for positioning steel
objects measUling from 12 to 20 m (40 to 60 ft) in length.
After magnetization, the object is rotated and positioned
for testing. MagnetiC particle test results determine where
the test object goes next: a grinding station, a second,uy
rolling operation or to a shipping point.
For the majority of their tests, steel mills oftell use wet
fluorescent magnetic particle techniques. In addition to the
power pack, circulating pumps, a filteling system and
applieation nozzles are all patt of the standard steel mill test
system.
 EOUIPMENT FOR MAGNETIC PARTICLE TESTS I 363

Applications for Large Castings full-wave power pack is also useful when large weldments
are tested. Instead of prods, the test object is contacted with
An impOltant application of the swinging Held (see
cl~mps and magnetizing. current is passed along the length
MllltidirectionalPowcl- Packs below) is the nondestructive
of the weld. After this first test, the clamps are repositioned,
testing of large, heavy castings. The power pack used in this
contacting the object so that current is passed perpendicular
method can supply up to three separate magnetizing cir-
to the weldline.
cuits. Contact is made by cables clamped to the test object
and the circuits are arranged so that magnetizing current
crosses the test object in three different directions. The
powe~' pack supplies timing circuitry for shifting current Multidirectional Field Power Packs
flow from circuit to circuit. In some applications, one of the
three magnctizing circuits is used to torm a coil, providing In som~ systems, a power pack is used to supply a circular
longitudinal magnetization in the test object. Demagnetiz- field to the test object then within milliseconds the field is
ing circuitry, usually reversing direct current, can be a changed to a longitudinal field. This sequence is continuous
feature of this power pack unit. and extremely fast since it can be based on the reversals of
This technique is an overall testing method that uses wet the 60 Hz power supplied to the unit.
flllo.rescent magnetic palticle suspensions to test the entire For example, an elapsed time of 0.5 seconds allows fifteen
surface of large cast objects. In a foundry, for example, this circular field shots and fifteen longitudinal field shots. The
technique can save as much as 90 percent of the labor and resulting magnetic field is a swinging field that sweeps
time required to inspect an object with prods. In addition, across the test object and crosses the plane of a discontinuity
areas can be tested that are inaccessible to the prod method. regardless of its orientation.
Often, wet fluorescent magnetic particles are used with
this magnetization method. The suspension is applied to the
Other·Power ~ack.Applications
test object as the .magnetic field swings across. the' material.
Full-wave direct current power packs are used for ener- The rapidly changing field requires high mobility from the
gizing bench units. These systems are usually designed for particles and that in tum recommends th~ free, small
testing large objects such as landing gear forgings. A particles used in wet fluorescent suspensions.
364/ MAGNETIC PARTICLE TESTING

PART 7
DEMAGNETIZATION EQUIPMENT

fields and poles needed for typical detection methods.

Theory of Demagnetization Complete demagnetization is virtually impossible to oh-
tain. Consequently, the demagnetization process is act-llall.,
a reduction of the residual field to a level that is appropriate
Ferrous materials usually retain some residual magnetism for the test object's subsequent service or manufacture.
after the magnetizing current is removed. The amount of The demagnetization process exposes the test object to a
residual magnetism depends on (1) the retentivity of the reversing magnetizing field that gradually diminishes ill
material; (2) the coercive force of the material; and (3) the strength. This causes a corresponding reversal and reduc-
strength and direction of the magnetic field. A longitudinal tion in the magnetic field strength in the object. Figure .5
residual field is often simple to detect with a tag wire or a shows a typical demagnetization hysteresis curve and a flu\:
field meter. A circular magnetic field is completely con- curve during reversal and reduction (for more information,
tained in the test object and may not exhibit the leakage see the Section titled Demagnetiz,ation of Test Objects).

FIGURE 5. Typical flux curve for the demagnetizatipn procedure

B+ + .•

>-
f-

~
H- -----+-+HJ'++-+--.......- - - H + ~~--4--4--4-~-+-4~-------­
x
TIME_
2

-~-----
FLUX CURVE
+ -:;- I
I
I
I

H - _ _ _ _- - I L - - _ . L - - r -_ _ _-t--_.,.....,,-_ _ H +

CURRENT CURVE

I 8-
 EQUIPMENT FOR MAGNETIC PARTICLE TESTS / 365

Need for Demagnetization Coil Demagnetization

I t is sometimes necessalY to demagnetize an object Alternating current demagnetization 'with a specially built
before magnetic particle testing. This is particularly tme if demagnetizing coil is the most convenient and widely used
tIl(' test object has a strong residual field from previous demagnetization method (see Fig. 6). Single-phase alternat-
operations such as magnetic crane handling or from contact ing current is used to power a multiturn coi1. VaJ;ous
with a magnetic chuck. voltages (120, 240 or 480 V) are used, depending on
I t is not always necesscuy to demagnetize aH:er magnetic availability and the application.
P;l rtide testing. Below are some of the factors that deter- The t}1)ica] demagnetization procedure comprises three
. Illine whether demagnetization is needed to reduce the re- main steps: (1) the test object is positioned within the coil;
sidual magnetic field. These factors focus on the test ob- (2) the coil is energized to a predetermined level; and
ject's subsequent manufacture and its seIVice life. (3) the test object is removed from the coil and placed
When an object is magnetized for magnetic particle outside the coil's magnetic field.
ksting, the residual field might adversely affect later stages As the Single-phase alternating current reverses in the
ill its pr()(luction. A residual field interferes with machining coil, the magnetic field in the test object also reverses. As
operations by causing chips to adhere to the test object and the object is moved out of the coil's magnetic field, the
this is detrimental to both tool life and the object's finish. A object's field is weakened as it is reversed. Moving the test
magnetic field can also interfere with welding operations by object completely away from the coil (usually 1 m away)
actually shifting the arc and prodUcing potentially defective effectively demagnetizes -the part. The current through the
welds. An important reason for demagnetizing is to clean coil is maintained until the test object is completely out of
the test object before it moves to any o~her surface related the coil' s fiel~. If the cl1rrent is stopped before this, or if the
.. production step. object's movement is stopped within the coil's field, residual
. Hesidual fields can also interfere with the function of the magnetism could remain. in the test material.
test object after it is placed in service. Magnetized milling
('litters, for instance, hold ferromagnetic chips and this not
I >Illy intelferes with the cutter's function but damages the

snrfaces of cut objects. Residual fields can interfere with FIGURE 6. Typical alternating current
c(:,ltain magnetic instruments (compasses, for example) or demagnetizing unit: ra) structure of a complete
\\;th any number of sensitive electronic components. A system with track and carriage; and (bl detail of
residual magnetic field is pmticularly destmctive to closely the coil assembly
fitted moving components such as crankshafts, connecting
rods and bealing smfaces. fa)
CARRIAGE

When Demagnetization Is Not Needed

When test objects have low magnetic retentivity (an
alltomotive block, for example, or low carbon steel used for
I
\\'dded tanks), the residua] field diSSipates as soon as the
III,tgnetizing current is removed. Sometimes, this is also true
in large structures and no demagnetization is required after
Illagnetic pcutic1e testing.
Demagnetization occurs as a side effect of heat treatment
if the test object temperature is taken to the Curie point or
:Iho\'e (about 750°C or 1,400 OF for steel). At the Curie
(('mperature, magnetic domains return to their random
orientations and the mateIial is demagnetized \vhen it cools, fbI
Ilwking fmther demagnetization unnecessary.
Finally, when an object is remagnetized in a different
direction to a level equal to or below its previous magneti-
I.ation, then demagnetization is not required. Some early
~pecifications required demagnetization between circular
and longitudinal magnetization steps of magnetic particle
tests. This was unnecessary for the accuracy of the test.
 366 I MAGNETIC PARTICLE TESTING

To facilitate handling, a small rolling carriage and track

are often furnished with coil demagnetizing systems. The Direct Current Demagnetization
test object is positioned on the carriage next to the coil and
pushed through the coil and beyond it to a neutral area. Alternating current does not penetrate ferromagnetil'
materials very deeply. For this reason, large test objects and
those magnetized with direct current cannot be demagne-
tized with alternating current.
Circular Field Demagnetization
Direct current demagnetization requires equipment \-vith
Alternating current demagnetization may also be achieved (1) a means of reversing the direction of current flow; and
with a circular magnetic field. The test object is placed (2) a means of gradually reducing the current level. A
between the contact points (headstock and tail stock) of a motorized thilty-point tap switch connected to a tapped
magnetizing system. An applied alternating current is grad- transformer is the most common current reduction method.
ually reduced to zero and this demagnetizes the object. The The reducing voltage is fed into rectifiers with each tap
procedure is widely used with alternating and direct current setting.
systems designed for testing heavy objects. The coil of a magnetic particle testing unit can be usE,d for
Reducing alternating current was formerly done with a demagnetization but a more effective method is to put the
step down switch connected to a tapped power transformer. demagnetizing current directly through the test object with
The step down was either a motorized switch or the manual what is called a head shot. For heavy objects, this fOI'm of
current control tap switch of the magnetizing unit. Similar demagnetization is the most efficient because it bn'aks
circuitl)' and procedures were used for mobile equipment, down longitudinal and circular fields. Head shot demaglle-
where contact with the object was made with cables and tization is usually done with large power packs.
clamps. . . On production equipment, demagnetization circuitry can
A faster system of demagnetizing heavy objects with be a part of the test system design. For example, th.e unit's
alternating current is the solid state. decaying method. The conveyor can pass through a demtlgnetizing coil beyond tIlt'
amperage flOwing through the object is rapidly and smooth- test station or a circular magnetization shot can be incolVo-
. . .
ly reduced by the demagnetizing system's elect~onics. rated into the conveyor. operation.
 EQUIPMENT FOR MAGNETIC PARTICLE TESTS / 367

PART 8
LIGHT SOURCES AND ACCESSORIES FOR
MAGNETIC PARTICLE TESTING

Table 4 lists typical accessoIies used for magnetic palticle grays absorb fairly uniformly throughout the visible spec-
testing when the tests are palt of a quality control program. tnllll and the spectral dishibution of the illuminating light is
Tal )Ie 5 lists accessOlies used with a typical production virtually unaffected.
Illagnetic pmticle test system. By knowing the color of the magnetic particles and the
Visible light and ultraviolet sources are often considered color of the background, it is theoretically possible to
accessories, even though they are a vital part of the magnetic increase the contrast of a colored particle indication by
p<lIticle testing method. An ultraviolet light source is usually selecting a particular colored light source.
i IIc0l1)()rated into a hood that darkens the testing area. The Because a greater area is made visible, floodlights are
hood can be pmt of the magnetic pmticle test system or may often an advantage for inspecting large and relatively flat
I)f' located in a separate area designed specifically for the surfaces. On intricate test objects where areas are accessible
illtelvretation of test results. with difficulty, handheld spotlights may be more effective.
. ..' . .
Illumination levels for Magnetic ~article Testing
Sources of Visible Light for The proper intensity of visible light illumination is' deter-
mined by the parameters of the test and the phYSical
Magnetic Particle Testing characteristics of the test object and the particle materials.
For large discontinuity indications, a brightness level of 700
~lagnetic particle testing techniques often rely on an
to 1,000 Ix (70 to 100 ftc) at the surface of the test object is
inspector to locate and interpret test indications. The
generally sufficient.
lighting provided for this visual examination is extremely
illlpOJtant. It not only affects the sensitivity of the test but is
also an impOItant conhibuting factor to inspector fatigue ..
TABLE 4. Accessories for magnetic particle tests in
Visible light sources for magnetic pmticle tests are not
quality control
signiflcantly different from those used for other visual
• Ultraviolet light meter: measures ultraviolet light intensity
t('sting applications. Sunlight, incandescent lamps, fluores-
with digital or analog readout; may also include a sensor
('('Ilt tubes or metal vapor arc lamps are generally quite
for visible light
satisfactory. • Test meter: measures amperage output of the magnetic
particle unit; selection must be compatible with the
unit's current capabilities
Spectral Characteristics of Visible Light Sources • Centrifuge tube: used to test the magnetic particle
Spectral characteIistics are usually not important when suspension concentration
lIsing dsible light sources. However, it could be better to • Tool steel ring standard: used to verify the accuracy of a
lise a light source that is defIcient or rich in paIticular
test set up
• Field indicator: provides relative measurement of residual
\\;LH>]engths, depending on the color of the particles and the
magnetism
('())or of the background.
• Calibrated field indicator: provides actual measurement of
B:' measUling the spectral quality of light reflected and residual magnetism
"cattered from typical magnetic particles, it can be seen that • Magnetic field indicator: indicates direction of magnetiC
red materials reflect highly in the longer wavelengths of the field
\isihle spectrum, fall off at about 580 11m and remain • Hall effect meter: measures magnetic fields in either
('omparatively low through the remainder of the spectrum. dynamic or static modes
)'(·)10\,- powders absorb heavily at the shorter ,vavelength or • Reference standard: indicates the direction. penetration and
hltw end of the spectrum up to about 480 11m, absorb less to strength of a field
;11)( )ut .5.50 nm, above which they reflect strongly. Blacks and
 368 / MAGNETIC PARTICLE TESTING

For extremelv critical tests or verv small indications, TABLE 5. Accessories for operating a production
higher visible light intensities are need~d. These light levels magnetic particle testing unit
should be in the 1,500 Ix (150 ftc) range. Note that excessive
- Hand spray gun: for application of wet method suspension
light levels increase inspector f~ltigue.
- Powder spray bulb: for application of dry powder materials
• Contact block: for uSing cables with a bench unit
• Cables: for portable or bench units; available in 2/0 and 4/0
sizes With lug terminals or slip jOint (either-end)
connectors*
Use of Ultraviolet Lamps for • Contact clamps: used with cables to make electrical contact
Fluorescent Magnetic Particle Tests with test object
• Prods: used with cables to make electrical contact with test
object; commonly used for tests of welds or large test
Fluorescent magnetic pmticle indications should be in- Objects
spected in a darkened area. The lower the level of ambient • Split coil: used with mobile units for longitudinal
visible light, the more brilliant fluorescent indications will magnetization
appear. This is important paIiicularly when testing for very • Central conductor set: central conductors in several sizes for
small discontinuities that mav attract onlv a small number of use with bench units
fluorescent paIiicles.'; - - Test Object supports: headstock, tailstock and rail mounted
supports; various sizes useful for tubing, bars and shaft
It is also impOltant that the test station be free of random
shapes .
fluorescent materials, whether or not they are directly' • Transport truck: used to carry portable testing units and
related to the testing procedure. accessories
- Contact pads: copper mesh pads used to ensure contact
with test objects; some models have rubber cushion
backing and V blocks
Visible Light Interference with Ultrav,iolet Testing
• Connectors: slip joint (either-er.1d) CQnnectors for quick
connections to prods, clamps, cables and contact block
The visible light intensity in a test area has a dramatic • Small part adapters: permits use Of standard wet hOrizontal
effect on pelfotmance and reliability. More visible light unit for testing of small objects
makes fluorescent indications harder to see, requiring high- *2/0 cable Is 9 mm diameter copper eqUivalent
er ultraviolet irradiances to permit detection of indications. 410 cable Is 12 mm diameter copper equivalent
Table 6 lists several ambient visible light level equivalents.
The results of laboratOlY tests on light levels for fine and
coarse cracked panels and a dark adapted inspector are
sho\\11 in Table 7. TABLE 6. Ambient visible light level equivalents
Ultraviolet testing booths cannot typically achieve visible Visible Light Intensity
light levels less than 10 to 20 Lx: (1 to 2 ftc) because lux (footcandles) Light Level Equivalent
ultraviolet lamps (with filters) have some Voisible light out-
put. In addition, the induced fluorescence from test objeds, 10 (I) Best ultraviolet testing booth
inspectors' clothing and spills of fluorescent material add to 100 (10) Dim interior ambient lighting
the visible light level 1,000 (100) Bright interior; deep shade under sun
The ultraviolet irraJiance at the test surface can be
altered by adjusting the distance between the ultraviolet
source and the test object. A new, appropriately filtered
100 \V ultraviolet bulb can produce up to about 30 vV-m - 2
(3,000 .u\V-cm -2) peak intensity at 450 mm (18 in.) from TABLE 7. Empirically determined minimum light
the test smface. BIinging the light source within 50 mm levels for discontinuity location with fluorescent
(2 in.) of the surfwe increases the intensity to over magnetic particle techniques
300 \V-m -·2. Flood bulbs and those lamps pro~!ided with Light Intensities for Light Intensities for
fluted filters generally provide substantially lower irradiance Small Discontinuities Large Discontinuities
levels. W·m- 2 p,W·cm- 2 W-m- 2 p,W-cm- 2
Because test surfaces can he blocked by lamp hOUSings,
the shoItest practical source-to-object distance is about 0.3 30 0.1 10
50 mm (2 in.). A t)11ical ultraviolet light tcst call meet most 5 500 0.5 50
specifications with a distance of 380 mm (1.5 in.) and a 50 5,000 5 500
minimum intensity not less than 10 W-m - 2.
 EQUIPMENT FOR MAGNETIC PARTIClE TESTS I 369

FIGURE 7. Electromagnetic spectrum showing the narrow range of ultraviolet radiation used for
fluorescent magnetic particle testing
VISIBLE LIGHT
(400 to 700 nm)

~
X RAYS
(IOpmto JOnm)

I I~~FRARED
(700 nm to J mm)
MICROWAVES
(I mm [0 I m)
RADIO WAVES
(10 (0 100,000 m)

ULTRAVIOLET
(410400 nm)
[ UHF VHF
I

100
I I 1 1 I I I

10 100 10
I

100
I
'.
10 100
10 I Ia I00 )1 Ia I00
PICOMETERS NANOMETERS MICROMETERS MILLIMETERS METERS KILOMETERS

RADIATION USED FOR FLUORESCENT

MAGNETIC PARTICLE TESTS
(320 to 400 nm)

WAVELENGTH

wavelengths down to 350 nm (this causes the cornea and

Response of the Human Eye to ' lens of the eve to fluoresce). The eve also detects the
Light Sources presence of 100nger wavelength visible 'light more easily at
low background levels.'
Response to Ultraviolet Light
Comparative Eye Response to Visible Light and
The human eye is relatively unresponsive to radiation at Colored L!ght
\\avelengths shOlier than 400 nm (see Fig. 7). However, in
the absence of longer wavelengths, the sensitivity of the eye Visual acuity is the eye's relative ability to resolve detail.
to shOlier 'wavelengths greatly increases. As shO\vn in Fig. 9, visual acuity drops as the illumination
Figure 8 shows the response of the average human eye (brightness level) decreases. Light levels are detected in the
ullder various lighting levels. The highest luminance level retina by structures known as rods. Color is detected by the
sl,owll is 340 cd-m -2 (l00 ftL), a normal, bright light structures called cones. As bIightness diminishes, the iIis of
\'i('wing condition. the eye opens wider and the rods become preferential
The second light level at 3.4 cd-m -2 (1 ftL) is about the detectors oecause of their low light sensitivity. At low levels
average found in a darkened testing booth. Total darkness is of illumination, the eve is color blind because rods are not
llt'\'er achieved in a typical testing booth, for the following color sensitive. .
/('asons: (1) near ultraviolet sources also produce some blue The blightness of the area surrounding the target of
;lIId 'iolet visible light; and (2) most testing booths contain vision also affects visual acuity. Reducing the contrast of the
SOll1e sources of fluorescence, often the inspector's clothing. background area reduces visual acuity (see Fig. 10). Nor-
\1 the 3.4 cdem -2 ambient light level, the eye becomes mally, visual acuity is determined for visible light illumina-
V'lIsitive to radiation in the 380 to 400 nm range. This is tion. Visual acuity for monochromatic light is higher for the
almost thirty times higher than the sensithity in Inight light yellow and yellovv' green wavelengths.
cOllditions. The 380 to 400 nm radiation causes a deep blue
\isnaJ sensation in the eye, plus greatly increased sensith-ity
ill tIle hlue region to the 405 nm spectral line of mercury
Mechanisms of Dark Adaptation
\ apor lamps. This does have the advantage of allO\ving the The eye adjusts to changing light intensity by varying the
(lark adapted inspector to move around the booth safely, size of the pupil and by changing the retinal sensitivity. This
,!(,l'mately locating objects in the testing area. is an autonomic or reflex action in normal vision but full
The top response curve in Fig. 8 is the 0.03 cdem - Z adjustment of 10\\' level light does not occur instantaneously.
(),()1 ftL) level. This is almost total darkness and is seldom The change from bright, visible light to the darkened
"llL'olmtered in magnetic padicle testing applications. Note environment needed for reliable fluorescent testing normal-
t hat ",hen compared to normal levels, the eye is over 800 ly requires at least five minutes for an average, healthy
timcs as sensitive under these conditions and can respond to inspector (see Fig. 9).
370 I MAGNETIC PARTICLE TESTING

FIGURE 8. Relative average response of the FIGURE 9. Human vision threshold: upper and
human eye to various wavelengths in the visible lower dashed curves show the effect of high and
range of the electromagnetic spectrum at three low illumination levels before dark adaptation
different levels of ambient light: (a) 340 cd-m- 2 begins; for areas subtending more than five
(100 footlamberts); the maximum response for degrees, the threshold is almost constant but
this curve is set arbitrarily at 1.0 on the vertical rises rapidly as target size decreases; curves are
scale for relative eye response, corresponding to for a target subtending about two degrees
the photopic eye at maximum sensitivity;
(bJ 3.4 cd-m - 2 (1 footlambertJ; and /0- 2
V)
(c) 0.03 cd-m - 2 (0.01 footlambert) V)
10- 2
V)
V1
~ W
Z Z-
r- r-Qj
I /0-3 I'-'
\:)QJ
\:J
02- /0-3 _E
CQ~ g§~
rv
~~ /0-4
W::J
VISIBLE LIGHT en E /0-4
-10-
envo
t~ f=
Q... Qj
:?;
~E
f- Z wQ
w
Q::
0 :3
w w
w
Q:: ~ 0 0.::-
W
/0-5 Urv
0.:: -
wQJ
1,000 ~ '> CD lJ
llJ
>-
W
Q:: Q... /0-5 Q...D
r-
V) ~~
«
W
/0-6 «.:::.
~
-I -I

30 ,40

TIME
lOa I - - - - - t - - - - j - - - + - I - - f - - - - \ - - - - l
(minutes)

FIGURE 10. Contrast sensitivity of the eye as a

function of field brightness;, the dashed curve
indicates contrast sensitivity for a dark
surrounded field '
BRIGHTNESS
(candela per square meter)
10
1.0 I I I I

>- _\.
~ 05
f=
i7i
Z
W
V1
"[\ I
O. I f-------II__-~I__---j-_+--t---+-_+_-+-I
I-
<en
~-
Zen
0.1
[\
0::1
U

oo
-I 0.05
'\ I

0.0 I L--_---l'--_----.J'--_.L-L~_----'_ _---.J._ _-Y

0:::
Q...
[\ /
/1

lOa 200 300 400 500 600 7 a o

W
0:::
r\. //
WAVELENGTH !
(nanometers) STAR~'GHT MOO~UGHT INTE1RIORS FXTEr'ORS
001
/0-5 /0-4 /0-3 10 2 /0-1 10 10" la' 10"

BRIGHTNESS
(millilamrJert)
 EOUIPMENT FOR MAGNETIC PARTICLE TESTS I 371

The time required for dark adaptation generally increases

FIGURE 11. Dark adaptation curves measured with iu(;reasing age of the inspector. Dark adaptation is not
with a 25 mm (1 in.) test stimulus: (a) subject easilv retained - there is a decided physiological safety
preadapted to 6,760 trolands fa unit of retinal adva;ltage to having pupils that quickly adjust to bright light.
illuminance) for five minutes; and (b) subject Even velY hrief visible light e:q)osure requires complete
preadapted to 389 trolands for five minutes suhsequent dark adaptation.
\Vhen light levels are reduced, the pupil of the eye
expands in diameter to allow more light to enter and the
retina of the eye becomes more sensitive. Fmther clark
adaptation occ~rs below 2 X 10 - 3 cd-m - 2 as the eye
switches from cone vision to rod visioll. Figure 11 shows the
~~,-"-,,,~,...o(aJ variation in brjg~]tness discrimination that occurs as a
PHOTOPIC function of dark adaptation time.
Photopic vision refers to vision under high level lighting
conditions (73 (;d-m - 2 ). ScotopiC vision is a term used to
describe vision under conditions of total dark adaptation
... fbJ (3 X 10-''5 cd-m -2). The change in perception by the
human eye fc}r vat),ing conditions of illumination is shown
schematically in Fig. 12. The dm* adaptation gap is that
-2oL-----------1~0------------20~--------~3~0
petiod of time during which the eye is changing and is not
TIME IN THE DARK capable of pelforming at. maximum sensitivity in eit~er
(minutes) vision condition. The time required for dark adaptatIOn
before testing vades \\1th the individual and depends on the
FROM HATIWICK-ACADEMIC PRES:;. REPRINTED WITH PERMISSION.
overall health alld.age of the inspector .. A dark adaptatiOll
time of f]ve minules is typically required for fluorescent
magnetic particle testing ~ith .ultrmiolet irradiation .. Com-
FIGURE 12. Visual acuity as a function of object plete dark adaptation may. take as long as tvventy mmutes.
brightness: the dashed and dotted curves show Note that some specifications (M1L-STD-1949, for exam-
the effect of increased and decreased . pIe), require only one minute of dark adaptation before
background brightness; the open circles indicate performing fluorescent magnetic palticle tests.
the diameter of the pupil The llHLXimul11 sensithity of the eye shifts in wavelength
dming dark adaptation. Figure 13 ;how~ the sensitivity of
BRIGHTNESS the eve as a function of light wavelength for normal levels of
(candela per square meter)
illumination and also for the dark adapted eye. As dark
10- 3 10- 1 10 10 3
2.0 ,----,-'i,r-r--.--.--
,r--'-'Ir-T--'-'I-'---' adaptation progresses, the peak sensitivity of the eye shifts
./ toward the blue end of the visible spectrum \\ith reduced
1.8 f---+-+--l---t---i--t-- l7"'--t-j-----J
v sensithitv in the red. This so-called Purkinje shift is a result
1.6 t----t--;-----t---+----+--t-6"'-t--t----t
~I--- of the different chromatic sensithities of tl;e rods and cones
J
V of the retina.
Vl 1.4 ~-+-+--f____+-.+_-I...I-t~_:::::=+--+-__i
>- Q; 1--:':
t:::5
::JC
SdE
1.2 f---+-+---i----t----t-Vr--t---t--t--'
Eyeglasses and Fluorescent
~ ru 1.0f---+--+-+--t---f'r j--t-,--t- -t-~ Particle Tests
~~
I

0.8 f---+--+----1I---++-+--t---t---i------I6.0 ffi At present, there are no official federal regulations

So
~ 0.6 0 0
V 5.0
~~
2 i'::
coveling the permissible amount of ultraviolet exposure an
inspector can receive dming a work day. However, both the
_-+-__+-~~oJ01~0..:.oo>l..o9.o;;-::+.-t__t___j4.0 ~~
OAr
1
V 000
oPOOOQ,
I 1i! f-
National Institute for Occupational Safety and Health
(N10SH) and the American Conference of Governmental
02
1L---+--'1
STAFUGHt;?: EXTERIORS 30 2 1ndustlial Hvgienists (ACGIH) have recommended the
follO\~ing lil;1its: (1) for the ultraviolet spectral region
MOONUPHT ,INTERiORS SKY ,

from 315 to 400 nm, total irradiance incident on unprotect-

BRIGHTNESS ed skin or eves, based on either measurement or output
(mi//ilamberts) data, shall not exceed 1.0 m\V·cm -;2 (1,000 .u\V-cm -2) for
periods greater than 1,000 S; or (2) for exposure times of
 372 / MAGNETIC PARTICLE TESTING

FIGURE 13. Relative luminous efficiency curves FIGURE 14. Mercury arc lamp construction
for the human eye showing response as a
function of incident light wavelength:
fa) photopic vision with adequate illumination
levels; and fbI scotopic vision with dark adapted
eye and low levels of illumination ELECTRODE

iO STARTING
LU ELECTRODE
V1
Z MERCURY
0 0.8 OUTER BULB VAPOR
0..
V1 CARTRIDGE
LU
a:::
.....J 0.6 ELECTROIJE
«-
::>-<
RESISTOR

V1~
:> 0.4
LU
>
~
.....J
0.2
ultraviolet radiation. They are also helpful when vi('will~
LU
a::: clitical small particle indications. Such visors or ~lasSl'S
510 should block all ultraviolet, most violet and blue li~ht,
WAVELENGTH without diminishing the yellow green test indication.
(nanometers)

Mercury Arc ,Ultraviolet Lamps

1,000 s or less, the total radiant energy shall no~ exceed
1,000 mW s-cm- 2 (l.0 J-cm- 2 ). The most common light source for fluorescent ma~llcti('
Photosensitive (photochromic) eyeglasses darken in the particle testing is a mercury arc bulb rated at 100 \V.
presence of ultraviolet radiation. The darkening is propor- Ultraviolet lamps require the use of an autotransformer to
. tional to th,e amount 9( incident radiation. While this type of regulate current and step up voltage. The ultraviolet li~l~t
lens has advantages under sunlight conditions, such glasses source ·for. fluorescent magnetic particle tests emi!s light
decrease the ability to petform fluorescent magnetic parti- with an average wavelength around 365 nm (3,650 A). The
cle tests and are not permitted for use in the test area. output was formerly and incorrectly measured in footcal Icllt's
\tVearing red lensed eyeglasses when exposed to visible with a visible light meter. Ultraviolet light meters are 1I0W
light may aid subsequent dark adaptation. These glasses available to measure the ultraviolet intensity directly at I he
must be removed after entering the darkened testing booth testing point as watts per square meter or a~ micro~atts pCI'
before viewing fluorescent test indications. square centimeter.
Some commercial eyeglass frames fluoresce and can A filter is placed over an ultraviolet light source h))" two
cause glare or unnecesscu), fluorescent background illumi- reasons: (1) to block visible light, exposing the test oiljed
nation. As with any object in the testing booth, glasses only to ultraviolet wavelengths; and (2) to resttict the
should be examined and care should be taken to eliminate emission of ultraviolet wavelengths harmful to the hllln<ln
extraneous sources of light. body. No magnetic pmticle test should be attempted \vith-
It is possible that an inspector may expeliencc a tempo- out the filter in position. Damaged filters Ulust be replaccd
rary clouding of vision if black light is permitted to shine immediately. Dirty filters seriously impede the emissioll or
into the eye or if it is reflected into the eye from test object ultraviolet radiation - cleanliness is essential for a('cII rate
surfaces. This clouding occurs because the comea, lens and magnetic pmticle test results.
the liquid in the eye (the vitreous humor) are also fluores- The 100 \V mercUlY arc lamp requires about ten millutes
cent matelials. Under no circumstances should shorter to reach full intensity. To lengthen service life, the lalllp
wavelength ultraviolet light be allowed to shine or reflect should remain energized after warm up. It is good practic('
into the eye. to keep ultraviolet lamps energized for the entire working
Because of the eye's own fluorescence, ultraviolet sources day.
in the testing area should be positioned so that neither Ultraviolet lamps are sensitive to low voltage and volta~c
direct nor reflected light shines into the operator's eyes. fluctuations. For this reason, they must be connected to
Ultraviolet absorbing eyeglasses should be wom if this voltage sources that have little fluctuation (within 5 percent
fluorescence becomes a SeIiOliS problem and are recom- of the voltage rated for the transformer). A tap switch jll the
mended to prevent unnecessary exposure of the eye to transformer allows adjustment to several voltages.
 EQUIPMENT FOR MAGNETIC PARTICLE TESTS / 373

construction of a Mercury Arc lamp

FIGURE 15. Spectral emission of mercury arc and
i\1crcUlY arc lalTlps are gaseous discharge devices in tubular ultraviolet sources
\\'hich an electlic arc takes place in a controlled atmosphere
and ('mits light whose charactelistics depend on the nature 100 .------~.,.--y----,-----_r_-__,

oj' t1w atmosphere. I

The construction of a typical mercUlY arc bulb is shown in /
Fig, 14. The mercury is confined in a qucutz or hard glass /
cartridge and hvo main electrodes cany current to the arc I
stream (along the length of the cartridge). An auxilial)' I
,o.;tarting electrode and a current limiting resistor are also I
included in the electrical design. The entire assembly is
I
scaled in an outer protective bulb that may either be
n(lclIated, filled with air or filled with an inert gas, depend- :2
iJIg on the design of the bulb. The lamp is fed from a current rv
LU~
regulating ballast reactance or transformer. This is required
IWC<lllSe the arc tube shows negative resistance characteris-
20
0<--
Q..C
til'S and would quickly destroy itself if not throttled by an VlQJ
LUU
('\tcrnal device. Cl<:Qj
3
Energizing a Mercury Arc lamp
\Vhen the lamp is first turned on, the mercury l'n the
l';utridge is not in vaporous form but is condensed in
droplets on the inside of the tube. Under this conditiop, it
would be difficult or impossible to shike' the arc. To
hcilitate stmting, a small amount of neon gas is incorporat-
ed into the caJtridge and a stmting electrode is sealed
t hrollgh an end of the tube near one of the main electrodes. o~----~------~-------~~~
300 350 400 450
\Vhen voltage is first applied, a discharge from the
..;talting electrode moves through the neon. This so-called WAVELENGTH
;:J( lIe discharglJ carries a small current limited by the (nanometers)
protective resistor but is sufficient to vaporize and ionize the LEGEND
JIH'rt'Ul), and eventually cause an arc to shike between the
- - - - TUBULAR FLUORESCENT
lilain electrodes. The heating and ionization process takes - - - FILTERED MERCURY ARC
five to fifteen 'minutes after the lamp is first energized.

Spectral Characteristics of Mercury Arc Sources must be filtered. A commonly used and effective filter is
One of the advantages of the mercury arc lamp is that its made of heat resistant purple glass. A typical transmission
()lItput can be controlled by design and manufacture. By CUlve for an ultraviolet filter peaks rather sharply near
proper choice of vapor pressure, spectral output can be 360 nm and starts to lise again in the neighborhood of
\ <tried from a few intense but widely scattered lines (when 700 nm.
! he pressures are near 1 mPa) to - an almost continuous
"pectrum at about 10 MPa (100 atm).
At medium pressures from 100 to 1,000 kPa (1 to 10 atm), Fixtures for Ultraviolet Sources
t 11(:' light output is about evenly distributed between the
\isihle, black light and hard ultraviolet ranges. These medi- For several reasons, ultraviolet sources require a housing
nlll pressure lamps have been used for magnetic patticle
and fixturing: (1) to support the filter; (2) to prevent leakage
t(,sting. Figure 15 shows the spectral emission of two kinds of unwanted visible light; and (3) to permit positioning of
,)/' ultrmiolet light sources. the beam onto the test surface.
Various ultraviolet fixtures are commerciallv available.
Some are small and portable. Others are mOUl;ted penna-
Transmission Characteristics of Ultraviolet Filters
nently inside a testing booth or on a production testing
Because a limited portion of the ultraviolet spectmm is system. Figures 16 and 17 show portable 100 \V mercury arc
needed for testing, the output radiation from a light source ultraviolet sources and tubular fluorescent ultraviolet lamps.
374 / MAGNETIC PARTICLE TESTING

FIGURE 16. Portable ultraviolet sources FIGURE 17. Low intensity ultraviolet lamps:
(aJ portable, battery powered source; and
(bJ tubular source powered by alternating
current
fa)

fbi

FROM ULTRAVIOLET PRODUCTS AND THE UNITED STATES AIR FORCE.

FROM ULTRAVIOLET PRODUCTS. REPRINTED WITH PERMISSION. REPRINTED WITH PERMISSION.
 EQUIPMENT FOR MAGNETIC PARTICLE TESTS 1375

The 100 \\1 lamp is small enough to be portable, although of large or medium sized indications. This source docs not
it is often mounted more or less permanently in a testing produce as high a maximum irradiance in anyone area as
unit. These lamps come in a v<uiety of configurations fi'om the 100 VV bulb.
various manufacturers, in spot and flood lamp !yves. Spot A 125 \V BlercUlY vapor lamp is also regularly used in
lamps are used almost exclusively to attain high intensities nondestuctive testing, however several aerospace companies
for localized use. As shown in Fig. 18, spot lamps produce a do not permit use of the 12.5 W source. The bulh has a filter
vel)' intense but narrow beam. built into its extelior glass shell. Because this is a pear
shaped bulb, it has an extelior reflector. The bulb socket
and handle are mounted on the reflector. Though the
Output Varieties for Ultraviolet 125 \V bulb is bulkier that the 100 VV model, it has a number
Sources of advantages. The 125 \V lamps have less of a warm up
peliocl and come up to brilliance much Elster. In adcliti~n,
the 125 \V bulb is much less sl~sceptible to voltage vana-
There are many mercury arc ultraviolet sources ranging
tions.
dmvn to a 2 \V size. These have found celtain vel)'
There is still another ultraviolet source available, but it is
spedalized uses in magnetic palticle testing. They usually
not suitable for magnetic pmticle testing applications. It is
do not have built in reflectors so their lower power is widely
dispersed. an incandescent ultraviolet bulb in 75 and 1.50 \V sizes.
These are standard incandescent bulbs with a filter glass
A 400 \V ultraviolet source is also available. It is large and
envelope. Measured on a standard ultraviolet meter, the.se
t)1)ically limited to a stational), mounting. The 400 \V source
sources produce ultnl\,iolet irradiance similar to that of a
produces a large amount of ultraviolet over a large area and
4 \V fluorescent tubular source. In addition, the incandes-
IS we]] adapted to illuminating a large area for quick location
cent bulb produces nearly 30 times as much visible light as
the fluorescent tubular sources. Experiii1ents shO\v that
small discontinuity indications detected \vith other sources
FIGURE 1B. Intensity variations as a function of
cannot be detected with incandescent ultraviolet sources.
distance from the beam center for several
ultraviolet sources Large indications might be dete'ctable but the high risk.of
missing indications makes the use of incandes<;ent ultravIO-
DISTANCE FROM BEAM CENTER let bulbs unwise.
(inches)

J 0.000 J0 9 8 7 6 5 4 3 2 a I "2 3 4 5 6 7 8 9 JO 100

J
Fluorescent Tubular Cold Discharge
100 W SPOT
PLAIN FILTER -
Ultraviolet Sources
Another t)'Pe of ultraviolet light source f~r magnetic
~ pal:ticle testing is the fluorescent tubular ultraVIOlet source.
<!.J
Elechically and mechanically, these are standard fluores-
S 1.000 ['------1'-1-+---\-\-----_ I a ~
C <!.J
cent bulbs' that come in sizes from 2 to 60 input watts. They
<!.J

>-<!.J
u zE are cold discharge tubular lamps containing low pressure
1- ..... O<!.J
-ru i=~ mercmy vapor glow discharges. The plimary radiation
Vl::::J
Zu <::::J generated \\rithin the glass envelope is hard ultraviolet of
i.J.JVl zu
1- .....
Z<!.J ~~ 254 nm (2,537 A) wavelength. The primary radiation is used
-Q ::J~
-.J to excite a special celium activated calcium phosphate
~~
phosphor coated on the inside of the tube. This phosp,hor,
I when activated bv ultraviolet radiation, emits ultravIolet
light \\rith a range -'of 320 to 440 nm, peaking at 360 nm.
40 W ../~
TU8ULAR Because a Significant amount of \isible light is eBlitt~d
FLUORESCENT
along \\ith the ultraviolet, these bulbs are often made \\lth
a purple filter glass similar to that used over high pressure
I arc lamps. This greatly reduces the emitted ultraviolet but
lO i
28-'-::O---'--::2-'-OO:--'--1-L20--l.80------'40---.JOL-J4-0-aLa-12LO---1...-2...LOO----.l--.J28~ I often still leaves an excessive amount of visible blue light,
conSidering the relatively low intensity of its ultraviolet
DISTANCE FROM BEAM CENTER
(millimeters) output. Figure 15 shows the spectral emission of the
fluorescent tubular source compared to the high pressure
mercury arc.
376/ MAGNETIC PARTICLE TESTING

Advantages and Limitations of Fluorescent Tubular

FIGURE 19. Variation in ultraviolet output of
Ultraviolet Sources
100 W mercury arc lamp (normal power factor
Fluorescent tubular sources produce sufficient amounts ballast and 117 V tap) as line voltage varies
of ultraviolet light but because of their configuration they
cannot be easilv focused. The irradiance is much lower than
II C1
that provided 'by the high pressure mercmy arc lamps. I
Accordingly, fluorescent tubular sources are not often con-
sidered adequate for critical fluorescent magnetic particle 100
I /V
/V
tests. 1 I
Fluorescent tubular sources offer the significant advan-
tages of instant stmting, cool operation and low cost. Used in 90
I I

~
typical four-lamp fixtures, the 40 VV tubular sources add up I
\
to 160 VV and produce near ultraviolet radiation intensities .;;-

f
of 1 to 5 \V-m -2 at normal operating distances. Although 80
,,"l.-'t'
fluorescent tubular sources may not meet certain specifica- 0'
tions (MIL-STD-1949, for example), they are commonly
used in industrial applications where stringent performance 70
}7
limits are not required. II
Fluorescent tubular sources, especially in the smaller 60 I
sizes, are the most practical for battelY powered portable 80 90 100 110 120 130

ultraviolet sources. They are more efficient in their use of LINE VOLTAGE
electticity when compared to high pressure mercury arc
·sources. Even more important, they start and reach full
output in a few seconds rather than the fifteen minutes
required by the high pressure arc. Aging and life Expectancies of Mercury Arc
Ul,traviolet BUlbs
Care of Ultraviolet Sources The ultra~iolet output level of any bulb decreases \vith
age. As a bulb nears the end of its service life, output may
Note: care should be taken to avoid breaking· mercury dropas much as 7.5 percent. The selvice life of these bulhs
vapor arc lamps. Mercury constitutes a Significant health may vary widely, depending on their care and the original
hazard. It may also lead to cracking of aluminum and other manufacturer. N ominallife expectancy is typically proVided
metallic components that it contacts. Broken filters can by the manufacturer (1,000 hours for a 100 W spot). For
pennit exposure to dangerous hard radiation. various reasons, the actual selvice life is less for a bulb used
The output level of an ultraviolet source depends on the in magnetic patiicle testing. The manufacturer's service life
cleanliness of the filter, the applied voltage and the age of is an estimate based on a standard operating cycle in a fixed
the bulb. To keep the output high, the filter should be and ventilated position. Ultraviolet lamps used in magnetic
periodically removed and cleaned, the voltage should be particle testing are subject to numerous stmts and shllt offs
held constant and the bulb should be replaced when its and to rough handling. In addition, because of filtering and
output drops. pOliable hOUSings, ultraviolet lights for fluorescent tests
Low voltage can extinguish a UlercUlY arc and over time often operate at higher than optimum temperatures.
decrease the life of the bulb. \Vhere line voltage is subject
to wide fluctuations, \vith low points at 90 V or less,
Extending the Service Life of Ultraviolet Sources
ultraviolet sources cannot be expected to operate properly
(see Fig. 19). High voltage surges also decrease bulb life. The magnetic pcuiicIe inspector can contribute to lamp
Line voltages above 1:30 V can cause velY early burnouts. life in nvo ways. One of these is to avoid operating
On fluctuating p~:)\,yer lines, specially designed constant ultraviolet sources above their rated voltage. Slight increases
voltage transformers are recommended to control the ef- in voltage decrease lamp life substantially. In some tests,
fects of high and low voltage alld extend the source's selvice increasing supply voltage to behveen 12.5 ancl130 V resulted
life. An additional advantage is obtained by using such a in bumecl out lamps in as little as 48 hours.
transformer. Certain power lines are subject to shal1) drops A second way to prolong the life of an ultraviolet sOt~rce
when heavy machinclY is stmted and the regulating trans- is to keep the number of starts as low as possihle. Each tlIn.e
former helps eliminate the nuisance of extinguished ultra- a lamp is started, a small amount of active material 1S
violet sources that must cool before re-igllitioll. removed from the electrodes. A Single staIt is equivalent to
 EQUIPMENT FOR MAGNETIC PARTICLE TESTS /377

several hours of continuous hurning. It is generally more watts but is more often stated in milliwatts (mW) or
C('Ollomical to leave lamps burning over rest peIiods and microwatts (j-t\V). Although ultraviolet sources are common-
Illoch hours that to turn them off and on again. ly rated by their wattage, these figures are actually electrical
energy input rather than optically radiated output. Because
of conversion losses, the radiated output is much less than
Ultraviolet Measurement the input.
Instrumentation Measurement of ultraviolet irradiance requires equip-
ment sensitive in that spectral region calibrated in watts
I t is often diffkult to determine the predse amount of per square meter (W-m -2), milliwatts per square centi-
lIltraviolet radiation required to cany out a particular test. A meter (m\V·cm -.2) or microwatts per square centime-
snitable level can generally be determined by trials on ter (j-tW-Cl11 '-2). Such meters are typically filtered so that
rdc:rcnce standards with known discontinuities. . they rcspond only to the appropIiate ultraviolet wavelengths
Fluorescent magnetic particle test interpretation may be (see Fig. 20).
considered a visual test and most of the lUles applicable to The ideal responsivity function for such a measurement
\'isible light testing apply. Since different amounts of ultra- in magnetic pmticle testing is a constant sensitivity from 320
violet are necessmy for different types of testing, some to 400 nm (the so-called black light range) and zero sensi-
method of evaluating and speci~ying irradiance level is tivity elsewhere. vVith such a responsivity, meaSllrements
Ilceded. Experience indicates that ultraviolet irradiance could be made accurately and Simply. Unfortunately, large
levels of lO W-m - 2 are generally adequate. departures from this ideal responsivity are commonplace in
ultraviolet meters. Detectors or filters needed for realizing
Selenium Cell Measurement of Ultraviolet Li9.ht the ideal are not available.
Intensity When the spectral distribution of a calibration 'iamp and
the lamp to be measured are significantly different, the
• Beginning around 1942, selenium. cell phdtoelectric measured irradiance may be substantially in error. Some
IIleters were used to measure ultraviolet levels. The most manufacturers sell calibration equipment (based on black
,common, easily used portable footcandle meters were de- light) to calibrate radiometers pIimarily used with the
signed for illumination engineers. It was discovered that if 100 W parabolic ultraviolet source. Without a correction
the filters were removed from these meters, they became factor, such calibrated meters provide erroneous metering
sensitive to ultraviolet radiation. values.
The lux orfootcandle is a unit of visible light illmninance. All common indushial ultraviolet radiometers use a
I t is deflned in terms of the human eye's response under bandpass fiTter. Such filters have vmiable high transmission
Inip;ht light (photopic) conditions. There is no such thing as from 320 to 400 nm and much lower transmission else-
a footcandle of ultraviolet radiation. Ultraviolet radiation where. An ultraviolet meter's ability to measure with mini-
wa~ incorrectly measured in footcandles for many years with mal interference from other radiations is based on two
lInfiltered footcandle meters. Even though these readings characteIistics: (1) its spectral flatness, the upper limit and
(lid not make lOgical sense, they did give reproducible constancy of its transmission in the ultraviolet region of
relative quantities. interest; and (2) its blocking ability, the lower transmission
In addition, such metering was always dependent on an limit that can be maintained at all other wavelengths.
ltllcontrolled portion of the spectral sensitivity of selenium. Interfeting radiations, usually UV-B (280 to 320 nm), visible
This part of the selenium response was of no ~oncern to the (400 to 760 nm) and infrared (760 to 10 6 nm), can be
manufacturer of the visible light meter and could have been palticularly troublesome in magnetic particle testing appli-
('hanged at any time \-\lithout notice and 'without effect on cations.
t he visible light measurements. The effect on ultraviolet Unfortunately, filters with shalp cut-on and shatp cut-off
Ille(lsurement could have been Significant. Despite these slopes are velY costly and they have never been induded in
disadvantages, no portable meter was then available to give magnetic particle testing meters. In addition, errors and
(Ille measurements in the near ultraviolet wavelength range. confusion can be caused by using radiometers \,v1th the
wrong spectral response for the application. Military stan-
Ultraviolet Meters dards and other industry specifications clearly state that the
spectral region of interest is 320 to 400 nm for magnetic
Ultraviolet energy is invisible electromagnetic radiation particle tests. Typical spectra of lamps used in nondestruc-
similar in nature to radio and infrared waves. These radia- tive testing are shown in Fig. 15.
l.ions are measured in energy per unit time or watts and The excitation spectrum is such that some fluorescent
frequently, as in the case of broadcast radio waves, in magnetic particles increase in fluorescence efficiency be-
kilowatts. Ultraviolet power output can also be measured in yond 380 nm. Also, the output of commonly used ultra~rjolet
378/ MAGNETIC PARTICLE TESTING

sources does not fall to zero at this wavelength. For these

FIGURE 20. Meters used to measure ultraviolet reasons, a meter whose sensitivity falls to zero at 380 mn
irradiance: (a) combination radiometer and may fail to explain why one type of lamp causes greater
photometer; and (b) long wave ultraviolet meter fluorescence than another type. The meter may also difler
in relative readings from a radiometer with a spectral
bandwidth extending to 400 nm. Using broader bandwidth
(a) radiometers generally gives a better indication of the quan-
tity of usable ultraviolet.

Units of Measure for Ultraviolet Sources

The US Military Standard MIL-STD-1949, Afagnetic
Particle Inspection, calls for a minimum ultraviolet light
intensity of 1 ,000 microwatts per square centimeter (10 watts
per square meter) at the test surface, with a maximum of
20 Ix (2 ftc) of visible light in the testing area. Some inc1ushy
specifications vary from these limits.
As shown in Fig. 21, ultraviolet irradiances vaI)' greatly
with the distance between the light source and the radiom-
eter and with the type and wattage of the source. Note that,
under normal conditions, 'the irradiance does not vmy as the
inverse square of the distance. Ultraviolet sour~es are
typically broad beam and the inverse square law shictly
applies only to point sources. .
Irradiance is an average over the area of the sensor and
most radiometer-s used in' nondestmctive testing have dif-
ferent sized apertures. For this reason, there may be some

FIGURE 21.' Variations in ultraviolet irradiance

with distance from the source face for 400 Wand
100 W tubular lamps
fb,
DISTANCE FROM SOURCE TO METER
(inches)

o 4 12 16 20 24
200 I 0,000

, I'
--~ KIOO W BULB'
~'2
100 ' - - -
80
...
--~-

--=
Z ~ 60
'\
<:( QJ 50
i3 E 40 \ f-
400 W BULB I'\.
~ ~ 30
~~
t:u Sf 20
~ ,~
O~ 40W;~
TUBULAR ~
~~
:> V) 10
~OM:~
,I
<:( ±::: 8 - 8W'm 2 :::", ::::::" ,
a:::ro --
f-S .,~ -I:::.;..
~- ~
~
2 1 - - 1 00
o 100 200 300 400 500 600 700

DISTANCE FROM SOURCE TO METER

(millimeters)
FROM SPECTRONICS AND ULTRAVIOLET PRODUCTS. REPRINTED WITH
PERMISSION.
 EOUIPMENT FOR MAGNETIC PARTICLE TESTS / 379

reading variations even if all sensors were similarly calibrclt- commonly used. The user of this patticular meter must
(·d. TI~e 100 vV spot lamp used as a portable ~tltraviolet determine the multiplication factor of the metal attenuator
source produces a narrow beam of high intensity radiation under normal conditions.
alld some manufacturers' 12.5 \V lamps can be even narrow- One manufacturer's radiometer indicates that a 50 per-
(·r. Snch beams may not cover the entire area of meters with cent correction factor is needed when using its meter to
rt'lativdy large sensors and this can produce measurements measure fluorescent ultraviolet sources.
differing considerably from meters \vith smaller sensing Not all meters are deSigned for the same range of
ar('as. conditions or for the pmticular conditions found in magnetic
palticle testing applications. Some meters are used in
applications such as ultraviolet cUlillg, photolithography or
medical phototherapy where the spectral output of the
Using Ultraviolet Radiometers source is considerably different from the sources llsed in
magnetic particle testing.
I deaJ/y, the irradiance of a surbce could be measured
directly 'by placing a calibrated detector at that surface at
all:' time, hut this is usually not possible \vith radiometers.
To avoid additional sources of error, it is essential that the
~'()lllplete area of the calibrated detector be irradiated. If at Service Life of Ultraviolet Meters
.tli possible, the detector should be oriented normally to the
incident radiation to avoid modification of the detector's Petiodic recalibn~tion is absolutely essential if the reliabil-
projected area and to minimize reflections hom the filters in ity and accuracy of the ultraviolet meter is to be maintained.
front of the detector. Apmt from catastrophic events such as thermal or mechan-
1r it is not possible to have most of the incident radiation ical irppact, the components of most radiometric sensors are
striking the detector pelpendicularly, then an angular sen- subject to deterioration., even with careful use. Ultraviolet
-;itivity plot for the detector system must be used to avoid filters are subject to' aging effects, pmticularly those provid-
large-errors. Sensors with interference filters are especially. ed \vitb the intelference filters.' H timidity and heat that
prone to large errors if the irradiating source is off to one builds up \vithin the multilayer filter can have irreverSibly
side of pel1Jendicular from the sensor face or if measure- damaging effects on its transmission characteristics.
Illents ~ire made close to extended sources. A plastic wavelength convelter is used in most sensors to
One commercial meter sensor is provided \vith an change the ultraviolet radiation to visible light that is closer
IIl}calihrated, perforated metal plate transmitting 20 to to the peak sensitivity of the meter's photodiode. These
:30 percent of the ultraviolet radiation stIikirig it, depending cOllveIiers undergo irreversible photochemical damage with
ill) the geometry of the measurement. This allows an use and must be replaced at regular intervals. Most manu-
eHt'dive scale extension of three to five times the normal facturers recommend a recalibration peliod of six months. If
htll scale value of 6,000 J.t\V·cm -2 and about the same the meter is heavily used and under extreme environmental
()\'C'rall range (20,000 J.t\V·cm -2) as the digital radiometers conditions, the recalibration petiod should be shortened.
I"""""""""""""""""""""""""'"""","""""1"""""""""""'"

SECTION
16
SPECIAL APPLICATIONS OF
MAGNETIC PARTICLE TESTING
John Brunk, Sandra T. Brunk and Associates, Overland Par~, Kansas (Parts 4, 5, 6, 7 and 8)
William Burkle, Precision Tubular Inspection, Lone Star, Texas (Parts 2 and 3)
William Chedister,.Circle Chemical Company, Hinckley, Illinois (Part 6)
, Brando'n Fraser, William Burkle Associates, Bartlesville, Oklahoma (parts 2 and 3)
. Lawrence Goldberg, Sea Test Services, Merritt Island, Florida (Part 1)
Daniel Haney, General Testing laboratories, Kansas City, Missouri (Part 8)
Donald Hagemaier, Douglas Aircraft Company, long Beach, California (Part 5)
John Mittleman, Marine Inspection Technology, Panama City, Florida (Part 1)
Roderic Stanley, International Pipe Inspectors Association, Houston, Texas (Part 9)
382 / MAGNETIC PARTICLE TESTING

PART 1
UNDERWATER MAGNETIC PARTICLE TESTS

Innovative technology bas been tieveloped in the Amer-

ican offshore underwater nondestructive testing indushy. Magnetic Particle Testing
These developments provide signifIcant cost reductions, through Coatings
without reducing sensitivity, for standard magnetic particle
techniques as well as undelwater tests. Pelforming any type of underwater testing is costlv
Underwater magnetic particle weld testing is different hecause of the required peripheral diving snppOlt. The co;t
from f:)pical dry and wet magnetic pmticle testing. In is increased by the need to remove marine grO\vth prior to
practice, it resembles a hyl)lid form of both the dry and wet visual and magnetic particle testing. Traditionally, most oil
methods, more closely resembling (hy hllnication and in- and gas companies have required welded joints to he
service testing. The main differences are that the particles cleaned to bare metal, especially for magnetic pmticle
are delivered in a wet slurry and the inspector is a diver. testing. ~Jost industrial codes have limited magnetic particle
Figure 1 shows underwater magnetic pmticle testing being testing to surf~lCes having less than .50 /-Lm (2 milli-in.) of
performed as part of a damage slIIVey. thin nonconductive coating. However, research has S110\\l11

FIGURE 1. Diver per.forming underwater magnetic particle test during a damage survey; remote operated
vehicle provides additional surveillance
 SPECIAL APPLICATIONS OF MAGNETIC PARTICLE TESTING I 383

that it is possible to test through thin coatings with reliable Magnetic particle testing through coatings is performed
rcsults. 1,2
using an alternating current yoke. The sensitivity of the
This is impOltant because high pressure water guns system is based 011 the detection of flne in-service fatigue
operating at 138 MPa (20 ksi) can efficiently clean under- cracks with minimum dimensions of 13 mm (0.5 in.) length,
water stmctures to a thin, tightly adheling layer of black 0.025 mm (0.001 in.) width and 0.75 mm (0.03 in.) depth.
oxide having a coating thickness in the range of 75 to This sensitivity is within the range of fatigue crack sizes most
12,5 J.tm (3 to 5 milli-in.). Compared to black oxide surface oil and gas companies require for detection. Figure 2 shows
jJ IIi shes, the cost of cleaning to a bare metal surface finish a magnetic pmticle indication formed on a black oxide
cOllselvatively doubles the testing cost. It has been deter- sU&1ce finish.
ruined that magnetic pmticle testing on black oxide coatings
has the same required sensitivity as testing performed on
Field Tests through Coatings
hare metal for detection of incipient fatigue cracks at joint
illtersections on welded offshore structures. After' magnetic pmticle techniques were performed suc-
cessfully underwater for several documented case studies,
Underwater Applications of Testing research was initiated to look at similar tests through
through Coatings coatings in air. 4 ,.5 Research focused on defining threshold
coating limits for certain discontinuity sizes. Reliable results
A magnetic palticle system sensitivity test was demon- were achieved using an alternating current yoke with dry
stratedto the Amelican Bureau of Shipping (ABS) to obtain powder on hairline indications through as much as 300 to
cC'ltification for using magnetic particle techniques in the 400 J.tm (12 to 16 milli-in.) of paint.
f1eld ..'3 The demonstration, performed at the Paramus, New There.were two significant findings of these studies. The
Jersey laboratory of ABS, used painted objects having most important variable for performing magnetic particle
discontinui.ties of the type that must be located offshore. testing is not the ability to get sufficient flux denSity to the
The technique was certified and subsequently found to discontinuity site, but rather to get sufficient flux leakage
produce the required se'nsitivity without costly cleaning to . out from it discontinuity to form a detectable magnetic
I >are metal. The benefits to the customer are Significant particle indication.
reduction in cost (typically fifty percent) without saCrificing Secondly, when using the alternating current yoke (indi-
reliability.. rect magnetization), it is not always necessary to clean in the
area of yoke leg contact. The altemating current provides
high surface flux denSity because of the skin effect. The
FIGURE 2. Magnetic particle indication produced same test results can be achieved when the area of interest
on black oxide using an alternating current is cleaned, with or without cleaning the base metal at the
electromagnetic yoke points of yoke contact.
For example, a crack-like indication detected on a coated
weld can be verified by magneti~ particle testing through as
much as 750 J.tm (30 milli-in.) of coatings because the break
or crack in the coating provides a flux leakage path.
Conversely, cracks that occur only in the paint do not
produce magnetic particle indications.
Cracks occurring in offshore stmctural welds are almost
always found in the toe of the weld. If the paint coating
thickness is 750 J.tm (20 milli-in.), one way to reduce cost is
to clean the weld and 13 mm (0.5 in.) on each side of the
weld and then perform the magnetic particle test. Figure 3
shows an example of a tight fatigue crack detected with only
the suspect area cleaned to bare metal.
As a result of these studies,4,.'5 the American Society for
Mechanical Engineers amended its Section V, Nondestruc-
tive Examination Code (T-724d, Surface Preparation) as
follows:

If coatings are left on the part in the area being examined,

it must be denwnstrated that indications can be detected
through the maximum coating thickness applied. 6
384 I MAGNETIC PARTICLE TESTING

In some cases, the Single leg technique can provide better

Single Leg Electromagnet Technique sensitivity than small conventional yokes. The main factors
affecting test sensitivity for both yokes and Single leg
\Nhen testing T, X, K and Y tubular connections in areas electromagnets are: (1) the size of the coil producing the
of acute angle, it is often impossible to access the geometry leakage field; and (2) the flux denSity at the area of interest.
with a typical yoke configuration. A technique that can be Although no industrial specification references the Single
used successfully under these circumstances is the single leg leg technique, the method can meet all the necessary
technique. The single leg electromagnet produces a magne- specification requirements by (1) picking up a 4.5 kg (10 lb)
tization pattern called a radial field that is essentially one weight; (2) producing clearly defined indications on mag-
half of the longitudinal field produced by the yoke. Because netic field indicators; (3) producing greater than 240 A-m -1
structural steel has a higher magnetic permeability than the (30 Oe) in the tested area; and (4) detecting discontinuities
surrounding water,7 the magnetic field can be localized in production applications or in reference standards with
where it is needed. Figure 4 shows testing with the flux known discontinuities.
leakage field produced by a Single leg electromagnet. Alternating current is always used with the Single leg
technique because fatigue cracks typically initiate as exter-
nal surface discontinuities and then fracture through the
FIGURE 3. Magnetic particle indication produced thickness of the test object. Direct current magnetization
when only the suspect area is cleaned to bare and permanent magnets are not recommended for detec-
metal and the yoke legs contact 0.5 mm (0.02 in.) tion of in-service weld cracks. Figure 4 shows an example of
the single leg technique used in an area of tight access.
Figure 5 shows a magnetic particle indication produced on
black oxide with Single leg magnetization.
Single leg technology also offers some unique advantages
for robotic .test systems. Its geometry is universal and it is a
low profile package. Since robotic arms can maneuver heavy
components, coil size is not a restriction. Layering alternat-
ing current with pulsed direct current by stacking or
winding one coil over the other ~an provide the skin effect
. and particle mobility of alternating current with the fIeld
penetration of direct current. 8

FIGURE 4. Diver performs a magnetic particle

test using a single leg electromagnet to access FIGURE 5. Magnetic particle indication produced
tight areas of Interest on black oxide surface finish using single leg
alternating current electromagnet
 SPECIAL APPLICATIONS OF MAGNETIC PARTICLE TESTING I 385

Discontinuity Data Bases FIGURE 6. Optimum orientation of an

electromagnetic yoke for the detection of weld
Underwater magnetic p~uticle testing can also be used to toe cracks on offshore structures
define the types of discontinuities that occur, their frequen-
cy and the probability of detection. A discontinuity data base
fl.)!" fixed offshore platforms 2 found that, of all the discon-
tinuities detected, the majority was confined to the toe of
the weld. This trend applies to fixed offshore platforms (T, K
and Y connections, built primarily of mild strength steels)
and may not apply to other marine structures, such as
mobile offshore drilling units made of high strength steels.
The significance of the trend is that both fracture me-
chanics models and documented empirical data for fixed
offshore structures indicate a low probability of occurrence
for transverse discontinuities. Yoke manipulation and scan-
ning efficiency can be maximized by restricting testing
procedures to those that detect longitudinal discontinuities.
The yoke should straddle the welded joint so that the
lines of flux cross discontinuities at the most nearly perpen-
dicular angle. Techniques that use criss-cross (45 degree)
pole placement are less sensitive for detecting longitudinal
cracks and should not be used. Figure 6 shows the proper
yoke setup for testing in the longitudinal direction.
386 / MAGNETIC PARTICLE TESTING

PART 2
DRY POWDER MAGNETIC PARTICLE TESTS
OF PAINTED WELDS

A literature search produces little information about the bead in which copper ferrite dilution cracking had been
technical feasibility of alternating current yoke magnetic induced. Copper ferrite cracks are very fine and barely
particle tests of painted welds. Empirical data and practical visible to the unaided eye. Each plate was tested '.'\lith yok~
opinion are somewhat mixed concerning this procedure and magnetization and a record of the patiicle indications was
there are reports of both successful and unsuccessful mag- made for each direction of magnetization.
netic particle tests of welds through paint. Each plate was then coated with inorganic zinc, zinc
Such a wide variation in reported feasibility is not entirely chromate, enamel or phenolic epoxy. Coating thicknesses
surprising. Thousands of different paints exist and are were measured and recorded and the plates were then
applied in a wide range of thicknesses and in a nearly retested with a yoke technique. Particle indications were
infinite number of primer and topcoat combinations. Each recorded and the indications were compared to those
paint or coating system has a characteristic magnetic per- obtained during magnetic particle tests of the bare rnetal.
meabilIty that influences the degree to which magnetic flux This process was repeated until each plate reached the point .
may be introduced to a ferromagnetic substrate. at which coating thickness diminished the.indication detect-
ability. Table 1 summarizes the results of this study and hsts
the critical thickness of each coating.
Yoke Break Test
A yoke break test was devised to help evaluate the effect Reference Standards for
of coating type and thickness on the introduction of flux to Weld Cracking
a ferromagnetic substrate. This simple test relies on the
contention that the flux strength produced in a ferromagnetic To confirm that bead crack test data would be valid when
material may be gaged by the amount of pull produced by applied to actual welds, three additional plates were pre- .
the yoke on that material. The test uses a dynamometer to pared to contain multipass groove butt welds with multiple
pull an energized yoke from a bare test plate. The plate is bead weld caps. Each weld cap was apprOXimately 25 mm
then coated with progressively increasing thicknesses of (1 in.) wide. These welds contained undercuts of varying
paint. After the application of each paint layer, the force severity as well as copper ferrite dilution cracks. Except for
required to pull (or break) the yoke from the plate is the cracking, the welds were typical of those commonly
measured along with the coating thickness. During this test, encountered dming industrial magnetic particle tests.
the poles of the yoke are maintained at a coilstant spacing. An initial test was performed and records of pmticle
Such a test was performed using ASTM A-36 steel plates indications were made. Each plate was then coated with
as a substrate. Four separate coating types were evaluated,
including inorganic zinc, zinc chromate, enamel and phe-
nolic epoxy. Yoke break test data obtained for each coating
TABLE 1. Copper ferrite dilution cracking
are presented graphically by Fig. 7, where pull is plotted
detectability for various applied coating thicknesses
against applied coating thickness.
Critical Thickness·
Coating Type micrometer (millHnch)
Weld Bead Crack Enamel 300 II 2)
Reference Standards Inorganic zinc 225 (9)
Zinc chromate 200 (8)
Phenolic epoxy 175 (7)
In an attempt to quantifY the thickness of the coating at
which sensitivity is diminished, four ASTM A-36 plates were *MAXIMUM THICKNESS OF APPLIED COATING AT WHICH NO lOSS OF
INDICATION DETECTABILITY WAS OBSERVED
prepared, each containing a Single shielded metal arc weld
 SPECIAL APPLICATIONS OF MAGNETIC PARTICLE TESTING I 381

illorganic zinc, zinc chromate or enamel. The paints were that this difference may be the result of concurrent leakage
allowed to dry, coating thicknesses were measured and fields (Fig. 8) that existed in the multiple bead weld cap but
recorded and each plate was retested with yoke magneti- not in the single bead weld. This could have produced
zation. decreased flux denSity in leakage fields at the crack sites and
Hecords of the particle indications were made and test in tum could have reduced the ability to attract and hold
results were compared to those obtained from the unpainted particles through the coatings.
plates. Even though each of the applied coating thicknesses
was less than 80 /-Lm (3 milli-in.), all of the test objects Concurrent Leakage
revealed a dramatic loss of indication detect ability.
The concept of concurrent leakage was investigated
fmther. A Single shielded metal arc weld bead containing
copper fenite dilution cracking was deposited on a carbon
Effect of Coating Thickness on steel plate. The bead was tested with magnetic particle yoke
Te'st Results techniques and the discontinuity indications were recorded.
An additional bead was then deposited on each side of the
The data in Fig. 7 show that the introduction of magnetic Original weld and the new weld was retested. After seven
J1IIXto a ferromagnetic substrate with an alternating current beads had been deposited (three on each side of the original
~!oke is affected by interposing thicknesses of paint. The weld bead), a distinct decrease in indication detectability
extent of this effect may be significant, depending on the was observed.
type of paint and its thickness. This indicates that magnetic particle testing of painted
There is a surprisingly large difference in .the data welds is dependent not only on coating type and thickness,
obtained from single bead test welds and from multiple but also on the weld profile.
head. test welds, especially as it relates to the coating
thicknesses that permit yoke magnetization. It is believed.
FiGURE 8. Concurrent leakage fields produced in
a mul.tibead weld cap as the result of abrupt
FIGURE 7. Decrease in pull with increase in cross-sectional change at the weld toes;
coating thickness; yoke produces a pull of redistribution of the magnetic flux caused by
8.7 kg (19.3 Ib) on a bar~ test plate; yoke these fields reduces the leakage field in an
center-to-center pole spacing maintained at adjacent discontinuity
110 mm (4.25 in.) during all tests
DISCONTINUITY
LEAKAGE FiElD
8.0 (/8)
ENAMEL
Vi' 7.6 (/7)
U
c 72 (16)
::J
0 6.8 (15)
-l Q INORGANIC ZINC
-l
::J
~

Vl 6.3 (14) - Z/NCC HROMATE

Q..
E 5.9 (13)
~ PHENOLIC EPOXY
CJ)
54 (12)
.2
.Q 5.0 (II)
4.5 (10)

25 50 75 100 125 150 175 200

(I) (2) (3) (4) (50) (6) (7) (8)

COATING THICKNESS
micrometer (mil/i-inch)
388 I MAGNETIC PARTICLE TESTING

PART 3
MECHANICAL PAINT REMOVAL AND ITS
EFFECT ON CRACK DETECTABllITY

To assess the effect that mechanical paint removal might

have on weld discontinuity detection, a comparison was Visual Test Results
made of the detectability of copper ferrite dilution cracking
in welded test plates before and after a paint removal Copper ferrite dilution cracks tend to form fine, branched
procedure. Copper ferrite dilution cracking is characteristi- indications that vary Significantly in length and width.
cally very fine and difficult to detect visually. Its small size Depending on orientation and their reflective characteris-
makes it representative of the sort of weld discontinuity that tics, many such cracks are not conSistently detectable dudng
could be affected by surface working during paint removal. visual testing. To ensure that crack detection compmisons
were based on a reliable and consistent frarpe of reference,
an effort: was made to determine which of the sample weld
Test Obje.ct Preparation crack indications could be repeatedly and conSistently
detected.
Fiv~ ASTM A-36 steel plates were prepared, each about Fifteen welding inspectors ceItified by the Amedcan
300 m.m (12 in.) 10I)-g by 150 n;ml: (6 in.)". \}'ide by 13 mm . Welding Society visually inspected each of the five test
(0.5 in.) thick. A Single V groove butt weld was produced in plates and produced a detailed skdch of crack locations and
the center of e~ch plate by the shielded metal arc welding orientations. Each inspector's sketch was compared to the
(SMAW) process, using E-7018 electrodes. Each weld was previously obtained magnetic particle tape transfer record
divided into segments 50 mm (2 in.) long and two of these for the appropriate sample segment.
segments in each weld were selected for crack implantation. Those cracks that were most consistently reported by :lll
Copper ferrite dilution cracking was induced in each of fifteen inspectors were selected for use as reference cracks
the two selected segments. Each plate was assigned a in comparing detectability before and after paint removal.
sample identification number and each was tested using an
alternating current yoke and dry magnetic particles.
A transparent tape transfer record was made of all
Coating Application
discontinuity indications for each direction of magnetiza- The five plates were sandblasted in preparation for
tion. Because liquid dye penetrant materials are difficult to coating application. Five commonly used paint systems were
remove from capillmy voids such as the cracks produced in applied by brushing or spraying in strict accordance with the
the sample plates and because these materials could inter- manubcturers' recommendations. The applied paint sys-
fere with subsequent coating application, an initial liquid tems and tlle average applied coating thicknesses are de-
dye penetrant examination was not performed. tailed in Table 2.

TABLE 2. Coatings and thicknesses for paint removal study

Thickness Thickness
Sample Primer micrometer (milli-inch) Topcoat micrometer (milli-inch)

81 Zinc chromate 35 (1.4) enamel 98 (3.9)

82 TNEMEC'M 48 (1.9) enamel 88 (3.5)
83 TNEMEC'M 45 (1.8) Carboline ™ 305 130 (5.2)
84 Carbozine ™ II 110 (4.4) Carboline ™ 305 275 (11.0)
85 Carboline ™ II 85 (3.4) none
 SPECIAL APPLICATIONS OF MAGNETIC PARTICLE TESTING / 389

Each plate was given fourteen days to dry and cure. After Cross Sections
cming, the welds were visually inspected to determine if
The welds were then cross sectioned through vmious
cracking was visible through the paint and none was
crack locations. Cross sectioning was used to determine
detected.
phYSical characteristics such as crack width; to determine if
smearing or working of the metal surrounding the crack
Visual Retesting openings had occurred; and to determine the extent to
which the coating may have entered and occluded the
The coatings were removed by applying surface condi- reference cracks.
tioners (solvents) in combination with manual and power On the basis of laboratory reports and a subsequent test
wire brushing. Five certified welding inspectors then per- of the samples, crack widths were determined to be in the
formed visual examinations of the plates. Each inspector range of 25 /-Lm to 0.25 mm (0.001 to 0.01 in.) at the plate
produced a detailed sketch of the location and oIientation of surface. After paint removal, a layer of disturbed metal was
detected cracks. observed to partially seal the crack openings. Laboratory
A redetection frequency was determined for each of the tests indicate that this layer was on the order of one or two
reference cracks by comparing the sketches to the original metal grains in thickness. No paint was detected in any of
magnetic particle transfer record and to the sketches of the crack voids.
crack locations before coating. In all cases, the reference
cracks were redetected and recorded without Significant
change. A reduction in the cracks' reflective characteristics
was noted.
Conclusions
Magnetic particle tests of welds produce reliable results
after coating removal operations. No Significant degradation
in crack detectability was noted (all reference cracks detect-
Magnetic P~rticle Test Results ed by magnetic particle methods were also detected by
visual examinations).
An alternating current yoke, dry powder magnetic parti- Test results suggest that, compared to magnetic particle
cle test was performed on each sample segment. Transpar- techniques, the liquid. dye penetrant testing method is not
ent tape transfers of all discontinuity indications were reliable for detect!ng discontinuities ·after paint removal. It
obtained for each direction of magnetizatiQIl. Comparisons is suspected that this is primarily the result of a partial
were made of tape transfers taken before and after coating smearing of the weld surface metal across discontinuity
removal and the discontinuity indications were found to be openings.
identical. . In a further investigation, a test weld containing copper
A liquid dye penetrant test was then applied to each test ferrite dilution cracking was liquid penetrant tested both
plate using a solvent removable visible dye penetrant and before and after sandblasting. Before sandblasting, the
nonaqueous wet developer. A dwell time of twenty minutes cracks were readily detected. After sandblasting, crack
was observed before penetrant removal" and developer detectability was significantly reduced. This suggests that
application. Eighty percent of the reference cracks were not surface preparation also contributes to smearing and again
revealed by this method. points to the advantages of magnetic particle test methods.
 390 / MAGNETIC PARTICLE TESTING

PART 4
MAGNETIC PARTICLE TESTING USING
REMOTE VISUAL EQUIPMENT

Magnetic particle methods are sometimes required for

testing surfaces that are not accessible for direct viewing. FIGURE 9. Borescope viewing heads for various
The problem may be inherent in the test object (the inside applications: (a) bends the cone of view at
surfaces of hollow forgings or castings, for example) or in right angles to the borescope axis, providing a
the design of the testing facility. lateral view; (b) circumference or panoramic
Television systems are sometimes used to solve accessi- head projects forward to view at slanting
bility problems, such as the need to inspect both sides of an incidence a 360 degree band; (c) bottoming
object without actually moving it. Television systems can head provides cone of view directly forward
also be considered if it is necessary to increase the speed or with uniform circumference illumination;
to automate a testing process. High production quantities (d) bends cone of view at a retrospective angle
for cost justification, adequate lighting and sufficient space to borescope axis, providing a view of the area'
are all necessary to support a television system. just passed by an advancing borescope; and
When it is impossible or impractical to use standard (e) bends cone of view to a forward oblique
television cameras, other types of remote visual testing angle to borescope axis
instruments may be effective.

fa) 55 6EGR~ 7.
Types of Instruments
-~--
. When access to the area of interest is restricted, remote
visual equipment may be used to achieve inspectability. The
following text details such equipment for use with visible
(color contrast) and fluorescent magnetic particle testing
techniques. Four primary types of remote visual testing
instruments should be considered.

Lamp Illuminated Rigid Borescope

. Rigid borescopes are available ,vith inseltion tube cliam- (e)
eters from 9 to 70 mm (0.4 to 2.8 in.). Extendable borescopes
may have lengths up to 30 m (100 ft) or more. There are
several types of viewing heads available to give the desired
direction or angle of view (see Fig. 9).
The rigid borescope can provide very bright visible light fd) 155 DEGREES

illumination. For ultraviolet applications, large borescopes

can be equipped with the same 125 W ultraviolet sources
used in portable lamps, although bulb size can then restrict
-~
access. Much smaller ultraviolet tubes (Fig. 10) may also be
fe) 55 DEGREES-'\
used to increase accessibility but these typically do not meet
certain intensity requirements, such as those specifications
that call for measurement at a distance of 380 mm (15 in.)
from the source surface. This ultraviolet illumination
~
problem is common to all four types of remote visual FROM LENOX INSTRUMENT COMPANY. REPRINTED WITH PERMISSION.
instruments.
 SPECIAL APPUCATIONS OF MAGNETIC PARTICLE TESTING / 391

The ligid borescope has the advantage of supelior image scope tip is limited by absorption of ultraviolet in the
quality. Resolution in excess of 75 line pairs per millimeter fiberoptic illumination guide. Subject illumination can be
is available with some instruments when viewing directly improved by llsing high ptuity quartz illumination fibers
through the eyepiece. When a video adapter is added, some in the scope itself and a quartz or liquid filled guide to
loss of detail results, and the amount of image degradation conduct light from the lamp to the borescope. Borescopes
depends on the video components. This can be a clitical of this type can produce ultraviolet light intensities above
consideration, depending on the resolution needed to de- 1,000 p,W·cm -2 at typical scope-to-surface distances of
tect the smallest significant magnetic particle indication. 50 mm (2 in.).
The fiberoptic illuminated borescope typically uses the
Fiberoptic Illuminated Rigid Borescopes same type of ligid optical lens imaging system as the lamp
illuminated borescopes and gives comparable detail resolu-
The fiberoptic borescope uses a separate light source and tion. The effects of adapting video efluipment are similar for
fiberoptic filaments to transmit light through the insertion both types of viewing apparatus.
tube to illuminate the area of interest. Such borescopes are
available with insertion tube diameters from 1 mm (0.04 in.)
to 20 mm (0.8 in.) or more. A few have been produced with Fiberscopes
diameters as small as 0.5 mm (0.02 in.). Like the lamp
Fiberscopes are used when a flexible insertion tube is
illuminated borescopes, fiberoptic instruments are available
needed to reach the test surface. Some fiberscopes are
with several kinds of directional viewing heads.
available with two-way or four-way articulating tips for
In addition to increased accessibility because of their
changing the direction of view. Fiberoptics are used to
small size, fiberoptic borescopes may also be used in places
conduct light to the probe tip and to conduct the image back
where electlical wiling or heat from a lamp is hazardous.
to the eyepiece.
The intensity of fiberoptic illumination is determined by
Compared to a ligid borescope of the. same length, a
the type of light source used and the size, number and
fiberscope has much greater loss of initi~l light intensity
length of the illumination fibers. Visible light sources are
reaching the eyepiece. For lengths beyond about 1 m (3 ft),
available up to 300 W. In 'g~neral, meeting visible light
there is often insufficient visible light intensity for color
intensity requirements is not a problem with insertion tube
contrast magnetic particle detection and interpretation.
diameters of 3 mm (0.12 in.) or more and lengths at le~st
Light transmission depends on the size and number of
1 m (3 ft).
both the illumination and imaging fibers .. All else being
. Very high ultraviolet intensity is produced by sources with
equal, larger diameter instruments can be made in longer
200 to 300 W xenon lamps. The intensity available at the
lengths to achieve the same image brightness. For fluores-
cent magnetic particle testing, it is generally necessary to
use a high pulity quartz fiberscope. A 10 m (30 ft) quartz
FIGURE 10. Borescope head with a miniature fiberscope has about the same ultraviolet light transmission
ultraviolet source efficiency as an otherwise similar 1 m (3 ft) conventional
fiberscope.
Compared to conventional fiberscopes of the same diam-
eter and length, quartz instruments are more expensive and
often less flexible (larger minimum safe bending radius).
Their advantage is that they can be used for ultraviolet light
tests, an application not possible with conventional
fiberscopes.
Some quartz fiberscopes produce light intensities up to
1,000 J.LW·cm - 2 at tip-to-surface working distances of 25
to 50 mm (1 to 2 in.). They are available with insertion tube
diameters as small as 0.5 mm (0.02 in.). The minimum
diameter needed for ultraviolet light testing is about 3 mm
(0.12 in.).
Image resolution with fiberscopes is not as good as with
ligid borescopes. Thirty-four line pairs per millimeter is an
expected value for an 8 mm (0.3 in.) diameter fiberscope.
This resolution depends on the number and size of imaging
fibers and the size of the field of view. In general, smaller
FROM LENOX INSTRUMENT COMPANY. REPRINTED WITH PERMISSION.
diameters and more flexibility necessitate fewer fibers and
 392 / MAGNETIC PARTICLE TESTING

less resolution. With adaptation to video systems, fiberscopes quartz fiberscope gives a satisfactory image with less light
tend to suffer more image degradation than rigid borescopes. intensity on the test surface than a comparable conventioilal
fiberscope. For rigid borescopes, the important factor is the
Flexible Video Borescopes light transmission efficiency of the optical system. For
flexible video borescopes, it is the sensitivity of the camera
. Instead of imaging fibers or a rigid lens system, a flexible chip.
Video borescope uses a very small charge-coupled television The recommended approach to demonstrating the ade-
camera in the tip of a flexible insertion tube. There is no quacy of a testing technique with any type of remote vie'Ning
eyepiece and the image is viewed on a video monitor. instrument is to prove the ability to image an appropliate
Optical fibers are used to transmit light from an external discontinuity indication. This image might also be compared
SOurce to the probe tip. Transmission loss problems occur with the same indication viewed directly under acceptable
with ultraviolet sources. As with the other types of remote conventional lighting conditions. The demonstration of the
viewing instruments, there is enough light intensity on the adequacy, documentation and control of a nonstandard test-
test surface at normal working distances but not at 380 mm
ing technique is allowed by many codes and specifications.
(15 in.) from the probe tip.
These probes generally give better detail resolution than Image Resolution
fiberscopes used with television but not as good as the
~esolution from rigid borescopes. Unlike other flexible Differences in resolution capabilities of remote viewing
mstruments, video borescbpes do not develop black spots in instruments may be very important when it is necessary to
the image from fiber breakage and they can safely negotiate detect fine cracks. It may be of little consequence for
tighter bends than fibersGopes of the same diameter. locating larger discontinuities. The adequacy of a speCific
Flexible video borescopes are available with insertion technique can only be determined by practical application
tube diameters from 6 mm (0.25 in.) to 13 mm (0.5 in.) and of the test.
with lengths up to 31 m (100 ft). For ultraviolet light It is advisable to use a rigid bores cope when flexibility is
applications, I?iniature camera probes are limit~d tO,lengths not needed because they are les~ expensive, easier to lise
less than 2 m (6 ft). Because of the high sensitivity of the and less susceptible to damage than typical fiberscopes.
i~aging chip, much longer lengths can be used for visible
lIght testing. Like fiberscopes, video borescopes are avail-
able with two-way and four-way articulating tips and remov- Measurement of Test Indications
able adapters to change the direction of view.
It is often necessary to measure the size of a discontinuity
indication to determine whether it is acceptable. When
Comparison of Remote using a remote viewing device, it is not possible to measure
Viewing Instruments in the usual ways. Most test setups produce some image
magnification, depending on several factors: (1) the wider
Light Intensity the field of view, the lower the magnification at a given
distance (instruments are available with fields of view from
Except where there is access for an instrument with a about 10 to 90 degrees); (2) the greater the distance from
standard ultraviolet source, it is not now possible to obtain the lens tip to the test object, the lower the magnification.
800 p.W·cm -2 at a distance of 380 mm (15 in.) from the MagnifYing eyepieces, zoom lenses or video monitor screen
distal tip of the insertion tube. It may well be possible to size can also affect magnification.
obtain 1,000 p."V.cm -2 or more at the test surface under For an individual system, the relationship of magnifica-
normal tip-to-surface operating distances. The subject of tion to distance can be determined experimentally and it
ultraviolet light intensity requirements and appropriate may then be possible to maintain a constant distance \vith
measurements techniques is regularly under review. Mili- fixturing, such as when testing the inside of a straight
tary and industrial specifications do not currently address cylinder \vith a rigid scope. Some borescopes are also
the use of remote viewing instruments for magnetic particle available with measming eyepieces. In addition, accessory
tests. Present specifications typically cover the light needed video components are available to provide a direct digital
~or direct viewing of discontinuity indications by an readout of the indication size when adjustable lines OIl the
mspector. screen are positioned at the edges of the indication.
Consideration must be given to the loss of intensity In some situations, object distance may be vadable and
between the distal tip of the instrument and the eyepiece or uncontrollable. For example, when flexible bores copes arc
the video monitor. These losses vmy between different used, it may be difficult to precisely control the distance
types of instruments. For ultraviolet or visible light tests, a from the lens tip to the test object. One type of instrument
 SPECIAL APPLICATIONS OF MAGNETIC PARTICLE TESTING / 393

solves this problem by using a small depth of field - the image has adequate resolution (home-type recorders may
discontinuity indication appears velY fuzzy unless the object not be good enough).
distance is within a specific, narrow range. When the image Video hard copy machines show considerably more fine
is in focus, then the nominal distance is known. detail. Thilty-five millimeter cameras are a good choice for
Some flexible video borescopes and fiberscopes have use with eyepiece viewing systems.
working channels that allow \vires to be passed through the
insertion tubes. This permits attachment of devices that are
visible through the scope and can serve as reference objects
for comparative measurements. Small scales, wires of known Remote Viewing Applications
length or diameter, or spheres can be compared with the
ncarhy images of discontinuity indications. The probe is It is possible to use a remote viewing instrument to
manipulated so that the comparator is resting on the test inspect any area where it is possible to spray a magnetic
sUlface beside the indication. particle suspension. The corners of inside diameter surface
The reference device can also be an extension of known grooves, the far sides of inside diameter weld reinforce-
length beyond the probe tip, brought into contact with the ments and blind holes or cavities may often be inspected
test surface to define object distance and thereby magnifi- more completely and reliably with a borescope than with a
cation. This technique is best used with a video monitor. It mirror.
is not reliable when the probe is not perpendicular to the There are also exposed surface situations, such as weld
test surface. Lasers have also been used to project a ref- root passes and discontinuity grind-outs, where the inspec-
erence spot of known size on the test surface. tor does not have the space to properly position the light
source for. direct viewing.
Recording Test Results Figure 11 shows two magnetic particle indications on
. the inside su.rface of a 75 mm (3 in.) diameter pipe weld.
VideO' tape is an ob'(riou~ choice when television systems Figure 12 shows fluorescent magnetic particle indications of
are used, but it should first be determined that the recorded cracks in a steel forging. Both figures were made with a
video hard print machine.

FIGURE 11. Black magnetic particle indications

on the inside. diameter surface of a pipe weld, FIGURE 12. Fluorescent magnetic particle
viewed with a video borescope system and a indications on a forging, viewed with a video
rigid borescope adapter borescope system and a rigid borescope adapter

FROM SANDRA T. BRUNK AND ASSOCIATES. REPRINTED WITH FROM SANDRA T. BRUNK AND ASSOCIATES. REPRINTED WITH
PERMISSION. PERMISSION.
394 I MAGNETIC PARTICLE TESTING

PART 5
TYPICAL FALSE INDICATIONS IN
MAGNETIC PARTICLE TESTS

Whenever a testing technique produces a localized, char- the hvo bellcranks exhibited undercuts, overlaps, splatter
acteristic nonrelevant indication, there is some danger of and evidence of arcing.
inspectors getting a false sense that any indication in this Fatigue cracking of the failed bellcrank (Figs. 13 and 15)
area is nonrelevant. The following application is typical of appeared to emanate from a prior crack that was typical of
many examples of in-service failures due to unfortunate a weld or heat treat quench crack as evidenced by a
combinations of manufacturing processes and testing tech- featureless topography, high temperature oxidation and an
niques that led inspectors to disregard indications of real elliptical crack geometry. The quench crack extended ap-
discontinuities. 9 proximately 115 degrees around the forging circumference
and had occurred through 75 percent of the thickness at its
maximum depth. This crack grew by fatigue and final failure
False Indications in a occurred by ductile overload (Fig. 15).
Bellcrank Assembly
Magnetic Particle Test Results
An initial, single case of failure occurred in a slat drive
bellcrank, resulting in complete fracture (Fig. 13). Subse- Examination of the fracture face on the unfailecl hell-
quently, inspectors found other cracks in or near weld areas crank revealed that the two cracks were also typical of weld
of the bellcranks. Complete failure of one bellcrank had or quench cracks. Both fractures had occurred through
occurred circumferentially through· the forging adjacent to 50 percent of the forging cross section and there was no
weld 4 (Figs. 13 and 14). evidence of in-service propagatiO!l (F)g. 15). Laboratory
The secol1d (unfailed) component ~ontained tw'o circum- magnetic particle tests of the failed bellcrank revealed
ferential cracks in the forging between welds 2 and 3 numerous indications in the remaini~g three weld beads.
(Figs. 14, 15 and 16). The two crackswere diametrically Metallographic examination through a number of sections
opposite each other. One crack had occurred through revealed these indications to be false because no cracks
approximately 180 degrees of the forging circumfer-
ence, while the other encompassed 70 degrees of the
circumference. FIGURE 14. Slat drive actuator bellcrank weld
identifications
Discontinuity Characteristics
The weld bead geometry and quality in the two bellcranks
did not meet the dimensional requirements of the engineer-
ing drawing or the weld quality specification. All welds in

FIGURE 13. Failed beJlcrank assemblies

CRACK CRACK
 SPECIAL APPLICATIONS OF MAGNETIC PARTICLE TESTING / 395

\vere observed. Therefore, a low concentration oxide solu- welding was necess~uy to build up and repair weld discrep-
tion in conjunction with residual magnetization was devel- ancies and selialization was added for tracing the compo-
oped to retest the bell crank. This procedure eliminated nents through the manufactUling cycle.
false indications and indicated numerous cracks along the The magnetic pmticle testing specifications were revised
toes of welds 2 and .3 (Figs. 14 and 16) that were subse- to clarify in-process testing requirements and the detailed
quently verified by sectioning. The topography of these methods to be followed. These methods were worked out in
cracks was typical of quench cracks. the laboratory on cracked components, prior to release. The
A laboratorv search revealed three cracked bellcranks laboratory investigation revealed that thermal cracks, in-
previously rej~cted due to weld discrepancies. These three duced by welding or quenching, were located at the weld-
and the previous hvo cracked bellcranks prompted a review to-parent metal interface, in the forged parent metal adja-
of the manufactmillg and quality assurance standards appli- cent to the weld and in the weld itself.
cable to the components. The investigation disclosed that
there were hvo basic problems: (1) the weld bead was of Characterizing Nonrelevant Indications
poor quality; and (2) the magnetic particle testing proce-
dure did not accurately represent the discontinuities. In attempting to determine why the cracks were missed,
The engineering dra\ving was changed to speci(Y compo- it was found that false indications of background magnetic
nent selialization and to allow tungsten inert gas (TIG) particles were intermittently generated at the toe of the
welding repairs of the initial arc welds. These changes were weld (intersection of the weld and parent metal). These
necessary because it was physically impossible to arc weld indications appeared when the test objects were inspected
the bellcrank to drawing requirements. Tungsten inert gas using the wet continuous method and standard concentra-
tions of black or fluorescent oxides.
A cracked belle rank was taken to the five testing labora-
tolies responSible for the testing of these components and
FIGURE 15. Cross section through cracked the inspectors were asked to evaluate them. In all cases, the
beJlcranks: (a) weld 2 and weld 3 of the unfailed inspectors claimed that the indications were typical and
beJlcrank; and (b) weld 4 of the failed false due to weld geometry or permeability changes in the
component· .. heat'affected zone. ..
Other experience indicated that permeability changes are
faJ not intermittent and produce strong indications of magnetic
palticles on both sides of the weld for its full length.

FIGURE 16. Enlargement of magnetic particle

crack indication in unfailed bellcrank

fbJ
396 / MAGNETIC PARTICLE TESTING

Metallurgical evaluation revealed that the magnetic particle accurate or consistent magnetization from one size screw to
indications were caused by a combination of unacceptable another.
conditions - cracks, cold laps and sharp geomehical
transitions between the weld and parent metal. Magnetic Particle Test Parameters
The assembly is fabricated from a 4140 steel forging to To produce reference standards with known discontinu-
which are welded two links also fabricated from 4140 hand
ities, a variety of screws were pulled with a tensile testing
forgings. The links are heliarc welded to the forging using machine to initiate cracks. Previously accumulated rejects
mild steel. The assembly is subsequently heat treated to the were also used. Some cracks were produced in unthreaded
1,240 to 1,380 MPa (180 to 200 ksi) strength range, then
lengths of screw stock, allOwing the use of a magnetic field
cadmium plated and primer coated. In the course of meter in intimate contact with the test object surface.
fabrication, the bellcrank is magnetic particle tested four A field strength of about 1.2 kA-m -1 (15 Oe) was regis-
times - after machining, after welding, after heat treating tered with the meter probe resting on the thread crests and
and after cadmium plating. The cracks initiated during ensured detection of cracks open to the surface. Depending
welding or heat treatment should be detected by the on the thread size, nonrelevant indications in the roots
magnetic particle procedures. caused Significant interference with proper evaluation of
cracks when the field strength reached 3.2 to 4.8 kA-m - 1
Resolution of the Misinterpretation (40 to 60 Oe).
All bellcranks in stock and in manufacturing were then These two readings provided a practical range for suitable
tested using a low concentration of oxides and residual magnetizing current levels. When the test objects were
magnetization. Numerous test objects were reworked by magnetized in accordance' with MIL-STD-1949, field
tungsten inert gas welding or were scrapped. All testing strength readings ranged from 2.7 kA·m -1 to 17.8 MA-m- 1
personnel were instructed in the use of the new testing
technique. .
FIGURE 17. Fluorescent magnetic particle
Because of the problems associated with manufacturing
indication of a longitudinal seam extending from
the welded bellcrank, a one-piece forging was designed for
a bolt head across several thread crests
future production. Th~se 'Components show a marked im-
provement in quality. Finally, an in-service magnetic particle
test procedure was developed for use by airline personnel.

False Indications in
Threaded Fasteners
Threaded fasteners, especially those with thread diame-
ters less than 8 mm (0.3 in.), are often difficult to test
properly. When they are inspected for circumferential
discontinuities in accordance with coil magnetization cur-
rent formulas such as in MIL-STD-1949, nonrelevant incli-
cations in thread roots, shatp fillets and minor surface
irregularities due to magnetic saturation are commonly
observed.
Simply reducing the magnetizing current creates an
uncontrolled risk that real discontinuities will not be detect-
ed. Sometimes, low power magnification is used to distin-
guish discontinuities from nonrelevant indications. With a
binocular microscope and a standard 12.5 W ultraviolet light
source, it is extremely difficult to position the lamp so that
thread roots can be viewed with sufHcient illumination.
One effort to solve these problems led to the use of
reduced magnetizing current levels and a black oxide water
bath, replacing the Oliginal fluorescent oil suspension. 10
Small screws and bolts provided no appropriate location for FROM KERMIT SKEIE ASSOCIATES. REPRINTED WITH PERMISSION.
magnetic field meters that could quantitatively indicate
 SPECIAL APPLICATIONS OF MAGNETIC PARTICLE TESTING 1397

(34 to 223 Oe). For a given length-to-diameter ratio, the

I t is very awkward to use the headstock and tailstock of a
reading varied with material composition (permeability) and
overall size. horizontal magnetic particle system to test for longitudinal
discontinuities in small fasteners. It is also extremely slow, a
It was found that reducing the magnetizing current to
critical consideration if large numbers of components are
eliminate nonrelevant indications greatly reduced the need tested.
for a microscope. However, changing to a black oxide
One semiautomatic system tests screws and bolts up to
suspension and visible light illumination made it much
9.5 mm (0.38 in.) in dia~eter. This allows detection of both
easier to confirm findings with 5 X or 10 X magnification.
circumferential and longitudinal discontinuities in one op-
With either fluorescent or black oxide magnetic particles, a
eration. The test object is sequentially magnetized by three
water suspension was found to give less background inter-
mutually perpendicular coils wrapped around cores.
ference from particles held by gravity in the thread roots,
The component is held in a vertical position in one of four
possibly because of faster nmoff with the water suspension.
stations around an indexing turntable. The test sequence is:
Although circumferential or transverse discontinuities are
load the test object, apply suspension, magnetize, magne-
the most serious in threaded fasteners, some specifications
tize, magnetize, remove the test object, inspect. The current
require inspection for discontinuities of any direction. Seams
through each of the coils is individually adjustable. Proper
in the wire used to produce screws and bolts often become
current values must be determined for each configuration
longitudinal discontinuities in finished test objects (see
Fig. 17). and material to. allow for the residual effects of one
magnetization on the next magnetization stage.
398 I MAGNETIC PARTICLE TESTING

PART 6
MAGNETIC PARTICLE TESTS OF
MAINTENANCE INDUCED CRACKING

These illustrations represent an important application (If

Grinding Cracks magnetic particle testing. The technique may be lIsed [t)
locate discontinuities in an object, but it must also he llsed
There are many operations designed or intended to repair after operations intended to repair the original prohlelll.
material discontinuities in manufactured components. Mag- Such retesting is espeCially important subsequent to control
netic particle tests often reveal the need for such operations operations such as sharpening cutting wheels and straight-
by providing visible indications of the discontinuities. Be- ening drill bits.
cause repair procedures typically introduce thermal or
mechanical stresses, attempts to restore a component or
remove discontinuities can often create new problems. Repair of Hydroplane Blades
Grinding is an operation us~d to remove surface or near
surface discontinuities or to provide a specified surface Unlimited class hydroplanes use aircraft engines deH-I()p-
finish. Grinding operations can also produce thermal cracks ing 3,000 or mot:e horsepower and'may reach speeds O\t'l'
from localized overheating. These cracks are typically per- 90 m·s --;-1 (200 mph). Highly. stressed components that
pendicular to the grinding direction. Figure 18 shows the require magnetic particle tests include propellers, propdll'l"
grinding cracks present in a se.ction removed from a steel sha~ts, gears and turbocharger components:· .
landing gear. Als'o vis'ible in the photograph are the legs of
an electromagnetic yoke used to magnetize the component
for magnetic particle tests. Figure 19 shows grinding cracks FIGURE 19. Fluorescent magnetic particle
around the flanges, of two steel fittings. indications of grinding cracks
Figure 20 shows cracks in a twist drill bit caused by
straightening the bit with a peening hammer. Figure 21
shows cracks in a steel test fixture caused by using a rivet
gun to loosen fasteners.

FIGURE 18. Fluorescent magnetic particle

indications of cracks in a section removed from a
steel landing gear component

FIGURE 20. Fluorescent magnetic particle

indications of cracks caused by straightening a
drill bit with a peening hammer

" '". -. ~,.,.",/~",,~ (/ -' .

~",,~=_J.4""'"~"'~~),i:Wi~~
. . 'I'
. -" - \ ' d-__
~
,.. t ~","~'.j':"P~- ~
~, ~
~~
.
;z:" ~ ",", iN 0. ."f. t 40:i /!' 8 1" '" 'i!" ~" ;J+~? < ~ l'
 SPECIAL APPLICATIONS OF MAGNETIC PARTICLE TESTING I 399

FIGURE 21. Fluorescent magnetic particle FIGURE 22. Magnetic particle indications of
indications of cracks caused by using a rivet gun cracks in a racing hydroplane propeller blade
to loosen fasteners from a steel fixture

fb,

Hydroplane propellers rotate at more than 10,000 rpm.

In service, such rotation rates alone can cause stress
·cracking. In addition, under racing conditions, propellers
can leave and return to the water very quickly. This causes
rapid and dramatic changes in the density of the blades'
operating medium, which in turn produces sudden flexing
. and very high stresses.
A propeller blade is originally shaped for thrust by
machining, grinding and polishing. For purposes of operat-
ing efficiency, the blade tip is very thin, sometimes tapering
to a shalp edge that serves as a stress liser during use.
After magnetic particle or visual tests locate damaged
areas, propeller blades are often repair welded. This may
initiate new problems in the heat affected zone and can
introduce any of the discontinuities associated with welding
procedures. Retesting after weld repair is a critical magnetic
particle application.
Figure 22 shows three magnetic particle indications of
cracks in a propeller blade. These cracks were found using
coil magnetization and a fluorescent magnetic particle water
sllspension. FROM CIRCLE CHEMICAL COMPANY. REPRINTED WITH PERMISSION.
400 / MAGNETIC PARTICLE TESTING

PART 7
CONTROL OF WET MAGNETIC PARTICLES
FOR YOKE MAGNETIZATION

overly sensitive suspension on a rough surface or from

Suspension Quality excessive loose fluorescent material in the suspension.

One of the important factors that determine the effec- Use of the Settling Test
tiveness and reliability of wet m~thod magnetic particle
testing is the efficacy of the magnetic particle suspension. The settling test is intended to evaluate these character-
The commonly required settling test has some impOltant istics based on the assumption that they are directly related
limitations that become very evident during field tests using to the concentration of magnetic particles in the suspension.
the wet method with an electrom~gnetic yoke. The settling The availability of magnetic materials with different particle
test is appropriate for a suspension that is mechanically size distributions - to produce suspensions of low, medium
agitated in a large stationary magnetic particle system, but or high efficacy - is another reason for adopting alternative
the test is less appropriate for suspensions applied manually means of evaluating the suspension. Tests to measure the
from pump spray bottles or for suspensions purchased in overall sensitivity of a testing procedure (for example, ·the
sealed pressurized cans. When spray bottles are used, the tool steel ring standard for wet horizontal machines) are not
inspector is responsible for mixing the suspension properly specified for yoke techniques. Reference standards can help
and a~itating it sufficiently during use. Pressurized cans! on . evaluate the end product of several independent valiables .
the other hand, are labeled to show the particle concentra: and how they work in the overall test system, but they not do .
tion when filled, but this concentration level may not be specifically address the quality of the magnetic particle
maintained from the first use to the last. suspension.
The time and amount of material required for settling One way to evaluate the quality of a suspension is to use
tests to monitor each container (and each inspector's agita- a block of permanently magnetized material with a network
tion technique) make this method much slower and more of many natural cracks of various sizes. This block standard
expensive than inspecting with dry magnetic powder tech- is similar to the aluminum crack blocks used to compare·
niques. For example, if inspection continues while waiting penetrant procedures. Because the crack pattern is very
for the results of a settling test, then a reading outside complex, it is difficult to rely on visual observation and
acceptable limits make it necessary to retest every area me~ory for evaluation. A better way to compare magnetic
inspected since the last acceptable settling test. This is partIcle suspensions with this block is to compare photo-
avoided by stopping all inspections while waiting for the graphs made under identical lighting and exposure cO!1(li-
results of a settling test. Neither alternative is efficient tions. The block standard described below is a better choice
enough for production applications. for routine use because it gives a numelical reading of
relative sensitivity.
Important Suspension Properties
There are three properties of a magnetic particle suspen-
sion that are important to the inspector. Use of Prism Block Standards
Relative sensitivity: all else being equal, how does the
smallest discontinuity indicated with one suspension com- A prism block standard consists of a steel block with an
pare to that indicated with another? artificial linear discontinuity and a calibrated scale. The zero
ConSistency: are detectable discontinuity size minimums end of the scale is adjacent to a permanent magnet. The
the same every time a suspension is used? strength of the leakage field across the crack is inverscl:'
Signal-to-noise ratio: does background fluorescence in- propOltional to the distance from the magnet. The better
tmfere with detecting and evaluating discontinuity indica- the quality of the suspension, the lower the magnetic ~leld
tions r Excessive background could result from using an strength required for it to produce an indication. Therefore,
 SPECIAL APPLICATIONS OF MAGNETIC PARTICLE TESTING 1401

the length of an indication measured from the scale is TABLE 3. Comparative discontinuity indications on a
directly proportional to suspension sensitivity. block standard
To make a measurement, the block is held with the Length of Indication
Particle Block Standard
artificial discontinuity on top, tilted enough to allow excess Material Reading millimeters (Inches)
suspension to lUn off. Suspension is applied (sprayed, in this
case) and the length of the indication is read from the scale A 9 22 (0.9)
under ultraviolet light. After making the reading, the indi- A 12 28 (1.1)
cation can be removed by wiping the block with a lint free A 15 29 (1.2)
cloth. Magnetic particles remaining in the discontinuity can
be removed with a soft bIistle blUsh. B 15 30 (1.2)
B 17 33 (1.3)
To develop a working procedure for using a block
B 21 33 (1.3)
standard, freshly prepared suspensions were evaluated by
the settling test and the block standard. These tests gave a
range of acceptable values for periodic field checks of
suspensions applied from a pump bottle. Because the block One example of the significance of the numerical read-
standard field checks are almost instantaneous and use only ings was obtained by inspecting a natural crack visible to the
a small amount of suspension, it is practical to repeat them unaided eye at one end and tapered to a sharp edge at the
frequently. other. Two particle materials and three suspension concen-
For pressuIized spray cans, the pIism block standard may trations of each material were compared (see Table 3).
also be used to veIiry concentration levels as the can is used . When the suspension concentration is repeatedly in-
or to check variations between different cans of particles. .creased, a point is reached where block standard readings
no longer increase proportionally and background fluores-
Significance o.f the Block Standard Measurements .. ce!1ce does increase. Although the ~lock does give a quali-
tative indication of background fluorescence, this is' better
It is imp~rtant to understand what the block standard test
does and does not do. It evaluates the magnetic particle judged on the actual test sUlface. It appears that an
optimum combination of high sensitivity and low back-
su.spension as an independent vaIiable. It gives no informa-
ground is typically reached at concentrations well below the
tion about the functioning of the yoke (or any other
magnetizing apparatus that might be used for the test) or maximum allowable settling volume reading.
about the effectiveness of any specific test setup for detect-
ing discontinuities. Its use can replace some settling tests
but not any required test of field strength, field direction or Conclusion
equipment function for overall test sensitivity.
It does not provide a traceable calibration procedure The use of a block standard with a permanent magnet and
because there are slight variations between blocks. For a scaled artificial discOli.tinuity is a simple and practical way
example, a test of the same suspension with several different to monitor two of the three impOltant characteIistics (sen-
blocks might give a range of readings from 17 mm (0.7 in.) sitivity and repeatability) of wet magnetic particle suspen-
to 19 mm (0.8 in.). This vaIiation is not critical to an sions used in the field with yokes.
inspector using a Single block. However, it must be consid- The third important characteIistic, background fluores-
ered in any applications where different blocks are used. cence (signal-to-noise ratio) is a function of the test surface
There is no predetermined minimum reading for a good as well as the suspension. The block standard can be used to
suspension. Appropriate suspension concentrations must be find the concentration above which sensitivity shows little or
determined for a specific application (v.rith reference stan- no increase and background fluorescence' is expected to
dards) and these concentrations may then be verified with increase.
the block standard. Tests made by preparing suspensions of The use of the block standard with understanding of its
low, medium and high sensitivity magnetic powders accord- functions and limitations can greatly improve the effective-
ing to their manufacturers' recommendations gave block ness and control of wet magnetic particle tests performed in
readings from 6 mm (0.25 in.) to 27 mm (1.1 in.). the field v.rith yoke magnetization.
402 / MAGNETIC PARTICLE TESTING

PART 8
MAGNETIC PARTICLE FIELD TESTING OF
STRUCTURAL WELDS

Magnetic particle testing is pelformed on a valiety of increased to the point where these things occur and is then
welds used in bridges, buildings and other structures. The reduced slightly for the actual test. A magnetic field pie gage
American Welding Society Structural -Welding Code with its can be used conveniently if the test is being performed on
nondestructive testing requirements is often incorporated a horizontal surface with sufficient space for positioning the
into constmction contracts. gage.
Inspectors typically must work under less than ideal Alternating current and switchable yokes do not have a
conditions, particularly when existing structures are tested. means for adjusting the magnetizing current. However,
Field welds made to connect fablicated subassemblies can some alternating current yokes do have removable pole
·also present significant problems. extensions to compensate for variations in pole spacing on
the surface. These are available as more than one length and
are sometimes articulating. Depending on their length and
Effect of Weld Surface design, the ~ddition of pole pieces c~n reduce the tangential
magnetic field strength midway between the poles hy a
Most weld surfaces are left in the as-welded condition for factor of 25 to 50 percen~. It is poss\ble to deteonine the
service or for testing. It is up to the inspector to decide reduction of field strength caused by incomplete contact of
whether a particular 'sur£ice is good enough for magnetic the poles with rough surfaces using a Hall element
particle testing. Dry magnetic particles, half-wave rectified gaussmeter.
current or alternating current are typically used with elec- Partial penetration in structural welds is common, as are
tromagnetic yokes. welds joining sections with substantial differences in thick-
The mobility of dry powder with the pulsed magnetic ness. These conditions are common causes of nonrelevant
fields produced by yokes is essential for effective testing of indications with half-wave rectified magnetization. This
irregular surfaces. Testing with yokes and fluorescent sus- problem may be eliminated by changing to alternating
pensions is increasingly required for welds that are ground current - because of its skin effect, variations in magnetic
smooth, such as those in vessels fabricated to Amelican field strength with section thickness are minimized.
Society of Mechanical Engineers Boiler and Pressure Vessel Although rectified current is sometimes specified because
Code Section VIII. of its deeper penetration and its increased ability to detect
On rough surfaces, particle suspension can be trapped in subsurface discontinuities, alternating current is becoming
depressions, creating unwanted background f1uorescence. more Widely accepted in the United States. Empirical data.
The geometry of the test surface, particularly where fillet have shown that altemating current yokes are capal )Ie of
welds are used, often requires a yoke with articulating legs. detecting discontinuities that are not open to the snri~tce, to
The distance between the yoke poles may have to be a depth of at least 1 or 2 mm (0.04 to 0.08 in.). With any
changed frequently to fit the available contact areas. Some type of yoke, it is difficult to predict the ma.ximum depth at
yokes have adjustable magnetizing current when used in the which a subsurface discontinuity will be detected.
half-wave rectified mode.

Field Strength Adjustments for Effect of Coating on Tests of Welds

Tests of Welds When welds in existing structures are being tested, they
The magnetic field strength is adjusted to correspond to are often coated with paint or rust. One study had sho\~1l
the reqUired pole spacing and the test material. This is done that the maximum thickness of paint that does Hot sigllifi-
by observing the test surface for nonrelevant indications and cantly reduce the detectahility of a particular size discollti-
excessive buildup of pmticles around the poles. With recti- nuity depends on the types of coating matelials and also Oil.
fied current electromagnetic yokes, the current is usually the weld surface profile. 11 Because the type and thickness of
 SPECIAL APPLICATIONS OF MAGNETIC PARTICLE TESTING 1403

coating on an existing structure are generally unknown, the intensity lamp. Instruments such as borescopes and fiber-
safe thing to do is to remove it in the areas of interest. scopes can be used to inspect areas where direct visual
Although it is possible to make measurements of paint access is limited.
thickness with a pOltable magnetic induction coating gage, \Vhen dl)' fluorescent powders are used outdoors, it is
this does not provide information about the relative mag- important to have shielding for minimizing visible light. This
netic effects of different coatings. is at least as critical as providing sufficient ultraviolet light
Studies have also sho'wn that mechanical removal of paint, intensity.
such as by power wire brushing in conjunction with applying
solvents, does not Significantly reduce the reliability of
magnetic particle testing. 12 Relationship with the Welder
The magnetic particle inspector often works closely with
the welder who is called on to clean surfaces that are too
rough and to chase discontinuity indications. The welder is
Effect of Other Tests Parameters often the best source of information about what types of
discontinuities are probable and where they are likely to be.
Lighting for Weld Inspection Developing good communication with the welding con-
tractor can result in more effective and reliable testing. A
It is important to have adequate illumination of the test good start is to explain the magnetic particle procedure and
surface. Even when working outdoors in bright daylight, the need for particular surface conditions. It is equally
some welds are covered by shadows. Some specifications helpful if the inspector is informed of the welding process
leave determination of adequate visible light intenSity to the used. .
discretion of the inspector,13 while others specify a mini- Figure 23 shows a fillet weld presented as ready for
mum intensity (for example 2,200 Ix) at the test surface. 14 testing. The inspector has .marked three deep crater pits to
The common practice of using a standard flashligJlt is velY . be rewarked before n1agnetic particle testing. In Fig. 24, a
often inadequate. short overhead fillet weld is tested for transverse discontin-
Electric power is required at the test site for the electro- uities. Figure 25 shows an indication of a longitudinal
magnetic yoke and it can also be used with a pOltable high discontinuity' going around the comer of a weld.

FIGURE 23. A fU,et weld with three deep crater FIGURE 24. A short, overhead fillet weld tested
pits to be reworked before magnetic particle for transverse discontinuities
testing

FROM GENERAL TESTING LABORATORIES. REPRINTED WITH

PERMISSION. FROM GENERAL TESTING LABORATORIES. REPRINTED WITH
PERMISSION.
404 I MAGNETIC PARTICLE TESTING

FIGURE 25. Magnetic particle Indication of a

longitudinal weld discontinuity

FROM GENERAL TESTING LABORATORIES. REPRINTED WITH

PERMISSION. •
 SPECIAL APPLICATIONS OF MAGNETIC PARTICLE TESTING / 405

PART 9
OIL FJELD APPLICATIONS OF MAGNETIC
PARTICLE TESTING

In the Unite.d States, there are more than 660,000 oil and
FIGURE 26. Axial component of flux density in
gas wells containing an enormous amount of steel tubular
an 11 m (36 ft) steel bar; the value in the central
material down the wells, in wellhead fittings and in cross
country flowlines. The text below outlines some of the region is close to the ring sample value
(remanence or Br); emergent fields at the ends
magnetic particle testing applications routinely performed
in oil fields, including: (1) aspects of magnetization from the are about 0.03 T (300 G)
oil field inspector's viewpoint; (2) specifications used for oil
field tests; and (3) some of the misconceptions about oil
field applications of magnetic particle tests. -J "Bf 1,000 (10)
MagnetiC particle tes,ting is applicable only to ferromag- «>- ::J
zt: ~ 800 (8)
netic materials and should be studied, as a specialized
branch of electromagnetic testing. It is impossible to tes.t all
§~~ 600 (6)
!::o-;;-
tubes for imp~r.fections; both inside and out, using magnetic l'Jx Vi 400 (4)

particle tests alone. Other magnetic flux leakage techniques ~3~ 200 (2)

are more cost effective and have been used on oil field tubes
-lu..'E
for many years. These tests often make use of Hall element,
magnetodiode and coil pickup methods for sensing the DISTANCE ALONG BAR
, meters (feet)
magnetic flux leakage from material discontinuities.

Longitudinal Magnetization
Because of bearding effects, wet particles are preferred,
When a tube 10 to 15 m (33 to 50 ft) in length is just as they are when any short object is being tested for
longitudinally magnetized, discontinuities transverse to the transversely oriented discontinuities. Wet particles do ~ot
tube axis can be detected by their magnetic flux leakage. In fur along external field lines because of the surface tenSIOn
this circumstance, two distinct test object geometries must of the liquid vehicle (the surface tension effect greatly
be considered: (1) the ends of the tube; and (2) the larger predominates over the magnetic field strength). .
region away from the ends. The axial component of the Using wet particles at the ends of tubes and bars, despIte
magnetic flux density (after the material has passed through the emergent normal fields, also allows the applied coil field
a magnetizing coil) is shown in Fig. 26. In the central region to be raised to levels in excess of those normally used for
and for much of the tube, the steel is magnetically saturated body waIl testing.
and holds a flux density close to the remanent value for the
material. At the ends, the flux lines begin to emerge and the
Coil Field Criteria
axial magnetic component falls as the normal component
rises. When magnetizing the end of an elongated object, it is
''''hen testing tubes for longitudinally and transversely critical to ensure that the magnetization levels produce flux
Oliented discontinuities in the end regions, magnetic parti- leakage from surface discontinuities sufficient to hold par-
cle testing is the technique preferred from among the flux ticles and give a discontinuity indication. With solid bars,
leakage methods. This is mainly hecause the high normal this may be one magnetization level and with tubes it may
component of the emergent magnetic field strength can be another. Sometimes the appropriate magnetization level
obscure discontinuity indications when using other flux is determined by the ease with which indications can be
leakage methods. seen.
406 I MAGNETIC PARTICLE TESTING

For the magnetization of tube ends, two very different than 200 mm (8 in.), For tube diameter Dl measured in
types of coil are used. One is the traditional wound wire millimeters and field strength measured in millitesla:
magnetization coil, excited by alternating current or direct
current. The other is a coil made of a few turns of 0000 cable B = 20 + 0.12Dl
(over 100 mm 2 of copper) pulsed from a capacitor discharge or
system. B = 200 + 30D 2

Direct Current Coil (Residual Induction Tests) for outer diameter D2, in inches and field strength in gauss,
The field strength is over 40 mT (400 G) when the tuhe's
One method for magnetizing the ends of long test objects outside diameter is greater than 200 mm (8 in.).
is to apply the magnetizing field from a direct current coil. This equation covers the worst situations and explains the
In view of the high field strengths used, such coils should be general increase in tuhe wall thickness with increased
supported either from the floor or the roof of the testing outside diameter found in Ametican Petroleum Institute
facility so that the test object can be centered in the bore of (API) tubular matetials. It is a relatively simple matter to
the coil. Under such conditions, a simple test may be determine the field strength at the center of such a coil
pelformed to ensure the required magnetization of the using a gauss meter. This eliminates the misuse of older
ends. equations that state only the number of ampere-turns on
The test object is saturated in one direction then the coil coils. '
is turned and placed with roughly 300 mm (12 in.) protrud-
ing beyond the coil. A Hall element gaussmeter is posi-
tioned to detect the field strength that emerges from the Alternating Current Coil
end of the matetial. 1.5 The current in the coil is raised and
\Vhen testing with alternati~g current coils, two points
turned off in increments until the gaussmeter reading are important to remember. The first is that the magnetic
saturates. Figure 27 illustrates typical data for just such a field roughly obeys the typical eddy current skin depth
test, although it must be realized th~t coil sizes, material relation, so that at 50/60 Hz, the skin depth in steels is on
diameters and matetial wall thicknesses all conttibute to the the order of 1 mm (0,04 in,). A peak surface field of
applied field strength at which the matetial saturate~. 2,400 A-m -1 (30 Oe) gives good discontinuity indications
Using this technique, the ~nds of tubes ilre shown to be from outside diameter surface breaking discontinuities. .
adequately magnetized if the following equation is applied. The second point is that there is little penetration throilgh
The equation is wtitten for tubes with Qutside diameters less to the inside diameter for tube wall thicknesses in excess of
4 mm unless a direct current field is also applied to lower
the effective relative permeability of the mateIial. S11ch
FIGURE 27. Hall element data taken at the end fields can eaSily be measured with a gaussmeter.
of a tube to ensure saturation prior to particle
application; coil positioned 300 mm(12 in.) from
the end of a pipe with 90 mm (3.5 in.) outside Capacitor Discharge Coil Field
diameter A common method for the magnetization of tube ends is
to wrap a 0000 (over 100 mm 2 ) welding cable around the
30013) 1.200112) end of the test object and to apply several shots from a
cklpacitor discharge system, '
25012.5) 1,(00110)
This is typically a bank of capacitors charged to a voltage
ACTIVE
<W~ 20012) 80018) I-WVl limited by safety requirements (limit may diHer depending
«Q..
V1
-::3
oe::::3
-.JQ..rv oQ..rv on local regulations). The bank is deSigned to produce a
WLLO"l
u::o~
15011.5) • _._ -e- -.-
60016)
LdoF single spike of unidirectional current. Ideally, the current
u:::o~ pulse has a long decay time so that its field strength call
-.Jo- 100 II) 40014) WZrv
«Zrv >WVi penetrate the matetial despite eddy current effects created
BW~ 5W~ dUling the initial rapid lise of the current.
[TI~~ 5010.5) 20012)
«I=
I-E The theOlY of capacitor discharge magnetization includes
a<:I-E
0 considerations of the inductance L, capacitance C and
-501-0.5) -
," - 200 1- 2) resistance R of the entire system, including the ohject heiug
magnetized, Because many different types of capacitor
o 2,000 4,000 6,000 8,000 10,000 12,000
discharge systems exist, it is not possible to make sufficien~l)'
AM PERE-TURNS
general statements about the measurable parameters of a
pulse and the resulting flux denSity in the test object.
 SPECIAL APPLICATIONS OF MAGNETIC PARTICLE TESTING I 407

Magnetization of the material can be ascettained by any Magnetization Requirements

one of three methods. The first is simply to hold a
!n many cases: the number of ampere-turns supplied by
?aussmeter near the end of the material and require that it
COlIs can be specIfied. This in turn prOvides a flux denSity in
IS close to saturation. The second method is to use a
the material that produces residual induction indications
pottable surface discontinuity. The third is to use a flux-
from surface breaking discontinuities. When a coil field is
meter. 16 The shape of the cur:ent pulse is shown in Fig. 28.
applied as shown in Fig. 29 the magnetization field at the
Sho~er pulses th~t are not effective for magnetizing deeper
outer layers of the material is caused by coil turns very close
pottlOns o~ an object are shown in Fig. 28a. More elongated
to the test object. This is different from the situation
and effectIve pulses 17 are shown in Fig. 28b.
described in residual induction with direct current. When
As the number of coil turns increases around the test
the tube's outer diameter is less than 200 mm (8 in.) and
object, the inductance in the LCR circuit increases as the
l:neasured in millimeters:
square ~f the number of turns. This elongates the pulse and
ca.n e~slly cause a capacitor discharge box (normally oper-
atmg m the. curve of Fig. 28a) to begin operating in the
NIm = (50 + 0.33D 1 )D 1 (Eq.2)
or
curve of FIg. 28b. The problem is that pulse systems NIm = (1,300 + 200D 2 )D 2
generally operate with a rectifier that shuts down the
curr~nt at .the instant the pulse reaches its negative-going when measured in inches. When the diameter is greater
pOlilOn (FIg ..2.8a). Should the pulse be similar to Fig. 28b, than 200mm (8 in.) and measured in millimeters:
then the rectIH~r never closes and circuitry must be includ-
ed to clos.e off the pulse from the capacitor charging NIm = 115D 1 (Eq.3)
circuitry. or
NIm == 2,900D2

. when mea~ured in inches. For drill 'coll~rs and tool joints

FIGURE 28. Typical pulses from capacitor
measured in millimeters:
discharge systems: (a) short pulse effective for
deeper magnetization; and (b) long pulse more
NIm ~ 230D I (Eq.4)
effective for magnetizing a t~be or
NIm = 5,800D2
fa'
Where:
66 PERCENT SATURATION
J-~
Zv
wQ
N = number of coil turns;
~E DI outside diameter of the coil (millimeters);
~ tl 5 D2 outside diameter of the coil (inches); and
u~
1m maximum current from the capacitor discharge
system (amperes).
o~~ __ ~ ____ ~~ ____ ~ ____ ~

~ 100 I~

TIME
(milliseconds) FIGURE 29. Longitudinal magnetization of drill
pipe tool joint with a capacitor discharge system
fb' and several turns of 0000 (100 mm2) cable
10

50 100 150
TIME
(milliseconds) --========= TO CAPACITOR
DISCHARGE SYSTEM
408 I MAGNETIC PARTICLE TESTING

For magnetizing the end of a 200 mm (8 in.) outside inspected with a borescope. The presence of one fatigue
diameter drill collar, the diameter of the coil is about crack is sufficient for rejecting the tube. The field strength
225 mm (9 in.). The number of ampere-turns is 230 x 225 used for the test in Fig. 30 is around 40 mT (400 G). The
or 5,800 x 9 or about 52,000. Should the capacitor coil is passed back and forth over the suspect region to
discharge system produce a peak current of9,000 A, then six obtain a crack indication in the presence of f'llrring. \Vhen a
turns are required. Alternatively, two or three pulses should borescope is used, appropriate crack indications can gener-
be fired and the resulting residual induction checked with a ally be produced with longitudinal residual induction
field indicator. techniques.

Oil Field Applications for

Longitudinal Magnetization Circumferential Magnetization
The ends of drill pipe and drill collars are magnetized
longitudinally so that cracks may be detected in their Two distinct methods 16 are used for circumferential
threaded regions. If the connections are not made tig4t magnetization of tubes up to 14 m (45 ft) long (see Fig. 31).
enough (torque values are listed in API Recommended Both methods employ an insulated rod typically made from
Practice RP 7G), HI the threaded regions may first elongate aluminum and designed to pass through the bore of the
and then cracks may form at the roots of the threads. The tube. In Fig. 31a, the rod is reasonably well centered in the
most common place for such cracking is in the last engagyd bore and carries some form of direct current. In mill
thread. installations, this might be full-wave or half-wave rectified
In order to perform good magnetic particle testing, alternating current for wet. fluorescent magnetic particle
threads must be cleaned and burrs removed. The particle testing. In field operations, banks of batteries have been
~uspension should be more dilute than normal (1 to 3 mL extensively used to provide magnetizing current.
per liter of solution), to reduce the number of false
indications. Surface ultraviolet illumination should be
around 20 JLW-mm.- 2 •
For internal fatigue cracking in drill pipe, the leakage FIGURE 31.' Two methods for establishing
field must be strong enough at the outside diameter of the circumferential magnetization in elongated
tube to hold dry particles (Fig. 30) or the inside must be tubes: fa) central conductor method with battery
pack to provide high current; and fb) internal
conductor method with capacitor discharge
system; a peak and duration meter is often used
FIGURE 30. Direct current magnetization of drill to measure pulse amplitude and time
pipe for internal fatigue crack detection; the coil
field must produce sufficient flux leakage at the faJ
outer surface for dry particles to bridge the H =.-..!..-.
21TR
discontinuity

~COll

-------:/ - -
 SPECIAL APPLICATIONS OF MAGNETIC PARTICLE TESTING /409

Ie
Central Conductor Magnetization surface of the tube must create within the material a
-11
magnetic field that opposes the field produced by the rod
Ie When the magnetizing current is pure direct current and current I. The field strength at radius r (in meters), at some
o the conductor is centered within the bore, the magnetic
instant while the rod and eddy current fields are finite, is
a field strength H at the outer surface of the tube is given by:
given by:

(Eq.5) I
H= 27Tr +H e (Eq.7)

\Vhere: where He is the magnetic field strength created by the edJy

current itself.
I magnetizing current (amperes); and
Ro the outer radius of the tube (meters).
Eddy Current Effect
In Eq. 5, the field strength is given in amperes per meter Ampere's law indicates that the magnetic field at the
since Ro is expressed in meters (see Fig. 32). Field strength radius r is caused by the currents inside that radius, the rod
can also be measured with a Hall element gauss meter and, current and the inner wall eddy current (see Fig. 32). The
. because 1 G is equal to 1 Oe in air, conversion to gaussian outer wall eddy current is the return loop for the inner wall
units gives: eddy current and plays no role in the theory as outlined so
far.
However, since the outer wall eddy current does repre-
H (Eq. 6)
sent an unwanted current flOwing in the tube, its presence
leads to the very practical cQnsideration that pipes being
Where: magnetized- before testing should be insulated from each
other by an air gap. If this insulation does not occur, then
I = magnetizing current (amperes); and the outer surface eddy current can jump from protrusions in
Ro = the outer radius of the tube (centimeters). the pipe being magnetized to the next pipe in the string.
The resulting arc can cause bums on both tubes and this can
This equation should b~ used with Hall element meters that in tum cause hardening and locations where corrosion
have scales or digital readouts in gauss. preferentially occurs. It is particularly important to avoid arc

FUll-Wave and Half-Wave Rectified

Alternating Current
FIGURE 32. The eddy current IE created in a steel
Rectified alternating current is often used with central tube at the beginning of a pulse I in a rod
conductor magnetization. It is important to remember that centered within the bore of a tube; the direction
such current waveforms induce eddy currents in the test of IE on the inner surface opposes that of I; the
object. The field strength waveform at the outer surface can outer surface forms the return path
be seen by positioning a Hall element to detect the field and
then feeding the output of the meter to an OScilloscope.

Capacitor Discharge Magnetization

In another circumferential magnetization method (shown
in Fig. 31b), the motive force is provided by a capacitor
discharge unit. 17 There.is no need for precise rod centering
with this magnetization method, providing a distinct advan-
tage for field testing. Unfortunately, magnetization by ca-
pacitor discharge obeys no simple rules. The rapid rise of
the rod current causes the induction of an eddy current in
the tube and this detrimentally affects penetration of the
magnetizing field strength into the material.
The direction of the induced eddy current Ie with respect
to the rod current I is shown in Fig. 31 for a centered rod.
By Lenz's law, the eddy current induced on the inner
 410 I MAGNETIC PARTICLE TESTING

burns on API 5eT Group 2 materials,19 some of which

require a hardness less than 22 on the Rockwell C scale for FIGURE 33. Magnetization curves for oil field
longevity in sour environments. tubular materials; the dashed vertical lines
Magnetized material should also be insulated from the indicate that the materials are magnetized
metal racks that carry it. If pipe racks are not insulated with almost to saturation by application of 3,200 to
a layer of electrically nonconductive material (rubber or 4,000 A-m- I (40 to 50 Oe) magnetic field
wood), then the outer wall eddy current can flow to ground strength: (a) high strength tube material; and
through the rack and there is a finite possibility of arc bum (b) low strength casing material
at the point of contact with the rack.
faJ
Use of 8-H Curves for Setting Specifications -v,-

When test objects are magnetized with the capacitor

~~
ViOl
discharge internal conductor method, a tool steel ring B-H ~;[ 1,000 (10)

curve governs the flux denSity value in the material. In

effect, knowing the B-H properties of the material from a
3~ 800 (8)

ring standard investigation allows field strength levels to be LL=E 200 (2)
set. Figure 33 shows the B-H properties of two typical oil
-3.200 3.200 6.400 9,600
field tubular materials: a 620 MPa (90,000 psi) proprietary (-40) (40) (80) (120)
material and a 380 MPa (55,000 psi) casing material. 17
After application of about 3,200 A-m -1 (40 De)., these MAGNETIC FIELD STRENGTH
amperes per meter (oersteds)
materials are effectively saturated. It is generally true of oil
field tubular materials th~t 3,200 to 4,000 A-m -1 (40 to
50 De) are required within the material to provide magne- fbJ
. tization sufficient for residual induction testing. V1

This magnetic field strength level is required at each ~ ~ 1.400 (14)

• point in the tube wall, despite the ·d'emagnetizing effect of ViOl

the eddy current. Unfortunately, this requirement does not
iE ~
0-
1.000 (10)

lead to a: current equation t~at can be simply executed in the x~

~ QJ 800 (8) I
field. Experimental specifications discussed later have been --l

found effective for saturating tubes.

LL='E ~

200 (2)
: . - 3,200 to 4,000 A'm
I (40 to 50 Oe)

-3,200 3,200 6.400 9,600

(-40) (40) (80) (120)
Typical Requirements for
Direct Current Magnetization MAGNETIC FIELD STRENGTH
amperes per meter (oersteds)
If the central conductor method is used for magnetizing
tubes, then the values given in Table 4 reflect the magne-
tizing field at the outside diameter for typical pipe sizes. The
wall thickness, the mass per meter (weight per foot) and the
tube grade all affect the magnetic and electrical properties
such a field strength raises the value of the flux density in
of the material, but because the magnetization method is
direct current, these parameters do not affect the magnetic the tube to a high level so that, after the field has fallen to
field strength. The actual field strength value is often stated zero, the flux denSity in the mateIial is at a value close to
by specifications agreed on by the manufacturer of the remanence B r.
material and the user. A typical specification is given by
Eq. 8, where DI is the diameter of the tube in millimeters:
Pulsed Current Magnetization
I = 12DI (Eq.8)
or Internal conductor magnetization using single pulses of
I = 300D 2 current differs from direct current (and from continuo liS
magnetization by the central conductor method) because
where D2 is the diameter of the tube in inches. I6 The the induced eddy current may not have time to dissipate
specified equations give the equivalent of 3,760 A-m -·1 before the field strength from the conductor current dies
(47 De) at the tube surface. It can be seen from Fig. 3:3 that away.
 SPECIAL APPLICATIONS OF MAGNETIC PARTICLE TESTING I 411

TABLE 4. Current requirements for magnetization

of tubes: direct current or long pulse (more than FIGURE 34. Plots of the capacitor discharge
0.5 seconds) only; not valid for capacitor discharge internal conductor current (I versus t) and the
magnetization average flux density (B versus t) induced in a
tube; Imax and T are measured with a peak and
Diameter duration meter; note that the flux density peaks
millimeters (inchesl I, * 12**
well after the current
60 (2.375) 600 910
73(2.875) 730 1,100
89(3.50) 890 1,340
102(4.0) 1,020 1,530
J 14 (4.50) 1,150 1,720
127.(5.0) 1,280 1,910
140 (5.5) 1,400 2,100
168 (6.625) 1,690 2,530
B,
178 (7.0) 1,790 2,680
194 (7.625) 1,940 2,920
219 (8.625) 2~200 3,300
245 (9.625) 2,450 3,680
273 (10.75) 2,740 4,110
299 (11.75) 3,000 4,500
340 (13.375) 3,410 5,120
TIME
*3,200 A'm - I 140 Oel AT OUTSIDE DIAMETER . (milliseconds) •.
**4,800 A'm- I 160 Oel AT OUTSIDE DIAMETER

Time Variations While the smfaces are magnetized sufficiently to form

Figure 28 shows two variations that ~re measurable for indications of longitudinal discontinuities, no information
single current pulses such as those provided by capacitor on the interior condition is required. However, when
discharge units. The first variation is that of the magnetizing relatively thin elongated tubes can be inspected from the
current versus time (I versus t): a relatively rapid rise of outside diameter surface only, saturation of the material is
current to its maximum value I max is followed by a much necessary for inside diameter discontinuities to produce
slower fall to zero. The entire pulse length is about 200 ms. magnetic flux leakage at the outside diameter.
This time variation is a response to the discharge of a
capacitor C initially charged to Vo volts through a resistor R Common Pulsed Current Application
in a circuit that contains inductance L.
A new material's magnetic condition is often unknown to
the inspector who must generally assume the worst possi-
Flux Density Variations
bility: magnetic saturation in a direction directly opposite
A second variation is that of the average bulk flux density that of the operator's equipment. Stated another way, the
\vithin the material (B versus t). This quantity rises at a worst possible case is that of taking a material from an
much lower rate than I(t) due to the shielding effect of the unknown value of remanence in one direction to remanence
eddy current Ie. A high level of magnetization is reached in the other direction. This is shown in the B-H curve in
when the flux density at the point F is close to the saturation Fig. 35.
value Bs for the material. As shown in Fig. 34, deep The material might arrive with an induction between zero
magnetization of the tube only occurs when the detrimental and - Bn with testing to be performed at + B r . During
effect of the eddy current is overcome by elongating the magnetization, the material should take the path
pulse in time. In this way, the magnetizing current is still - BrHcPBsBn through saturation Bs to remanence Br.
effective as the eddy current is dissipating. For materials at zero induction, the tube is at 0 on the
The fall in induction from F to B r is expected when the B-H curve in Fig. 35 and takes the path OPBsBr during
magnetizing field strength falls to zero, as it does after the magnetization. Materials not saturated by a pulse may
passage of a pulse. This is determined by the B-H curve for follow a magnetization path such as - BrHcPQ or OPQ. It is
the material undergoing magnetization. Should the point F essential to then fire additional pulses. A possible magneti-
not represent saturation B s , then the material reaches some zation path during a second pulse is QBsB r. The final net
average bulk flux density lower than B r' induction is raised as shown.
412 I MAGNETIC PARTICLE TESTING

additional inductance between cables and ground can be

FIGURE 35. Possible paths taken by minimized. Because the inductance is time dependent, it is
circumferentially magnetized material from included in the derivative term.
various initial magnetization conditions to The resistance is the combined resistance of the rod,
saturation Bs and then remanence Br in a known cables and their connection and resistance within the
direction: (a) material at remanence in the capacitor discharge box. The internal resistance of the
opposite direction; (b) material at zero capacitor discharge box could be from the forward resis-
induction; point P indicates a weak pulse tance of a silicon controlled rectifier included to eliminate
followed by a second pulse the possibility of current oscillation. The capacitance of the
system is generally in the region of 2 to 8 F.
There are three solutions for Eq.9, if the time depen-
dence of L is ignored. These solutions depend on the
relative values of L, C and R.

I (Eq. 10)

o !'--+_ _ _ _.-;.;.H

-8, I (Eq. 11)

I (Eq. 12)
Analysis of Pulse Current Magnetizatic:m
Generalized analysis of the pulse current internal conduc-
tor method for magnetizing elongated tubes is presented 4L
below. Simplified equations are given for the types of
C
current pulses used for magnetization. The theory of cur-
rent pulse time dependence (r versus t of Fig. 34) is
discussed and equations are presented for the inductance Where:
experienced by the magnetizing Circuit. Vo the charged voltage of the capacitor bank; and
The equations illustrate: (1) the dependence of induc-
tances on the average value of the differential permeability f3 Rl2L.
(dBldH) of the test object; and (2) the dependence of the Equation 10 is an oscillatory solution, but the presence of
field strength and flux density limits (imposed by the the silicon controlled rectifier limits the pulse to the first
exciting current) and the B-H properties of the test object.
positive-going peak (see Fig. 28). In this particular example,
the pulse has a length of 17 ms and reaches 10,500 A. Such
Current Pulse Time Dependence pulses are ideal for magnetizing objects of low electrical
For LCR circuits, the time variation of the current pulse conductivity (ferrite magnets). However, with highly con-
obeys the equation: ducting materials such as steel tubes, the initial rapid
current rise (may be millions of amperes per second)
d(LI) + IR + I dct
cIt
I = 0 (Eq.9)
induces shielding eddy currents Ie that do not permit field
penetration into the bulk of the material. The net efIect of
this is magnetization at the outer layer only.
The exponential Eq. 11 is known in its mechanical ana-
The three terms on the left of Eq.9 represent the
logue as critical damping. It is difficult to achieve in this
instantaneous voltages across the inductance, the resistance
magnetic patticle application because it depends on a
and the capacitance in the circuit (Fig. 31b).
kno",m value of L that is in turn dependent on the physical
The inductance in the circuit is mainly that of the rod and
tube system, since by careful design, the presence of and magnetic parameters of the test object.
 SPECIAL APPLICATIONS OF MAGNETIC PARTICLE TESTING 1413

The sin h solution in Eq. 12 leads to the longest pulses overall resistance of the magnetizing circuit. Typical resis-
because there is no oscillation. Pulses up to 160 ms are tance values for 50 m (165 ft) of cable and 15 m (50 it) of
commonly used in oil tube testing. rod are 1 to 5 mD.
It has become common to define the length of such
pulses as the time taken for the pulse to reach 0.5 I max
during decay T. Both Imax and T are measurable with an Capacitance Values
inductive ammeter or a peak and duration meter. These
pulses are effective for magnetizing tubes because the field The capacitance in a capacitor discharge supply is gener-
strength ii·om the rod current is still high as the eddy ally between 2 and 8 F. This relatively large value is
current in the test object dissipates (there is penetration of provided for two reasons: (1) because of the need to
the field into the bulk of the test material). maintain relatively low voltages around the circuit; and
Because the inductance is a function of time, a full (2) to elongate the pulse.
solution for the variation of the pulse current I(t) can only While the values of resistance R and capacitance C can be
be obtained by modeling the effect that the induced eddy controlled by the capacitor manufacturer, the value of
current has on the instantaneous value of L. Experimental inductance L cannot.
evidence indicates that, for elongated tubes, the physics of
the magnetization process can be illustrated by a study of
the constant-Lease. Inductance Values
Inductance is dependent o~ characteristics of the test
object. In the case of a tube, inductance is given by:
Typical Values for Inductance, Capacitance
and Resistance
\VrH~m desig~ing a capacito~ discharge puising system, it is
L=~ r dB") In R"Ri
27T \dB
(Eq. 13)
essential to provid.e a pulse length suffic:ient for deeply
magnetizing the material. There .are two reasons for this. Where:
First, the tested material may arrive in a longitudinally
magnetized condition and it may be necessary to remagnetize the length of the tube (meters);
it circumferentially before magnetic particle testing. Sec- the differential permeability;
ondly, some specifications call for relatively low emergent the outer radius of the tube (meters); and
longitudinal field strengths at the ends of elongated test the inner radius of the tube (meters).
objects and rotation of the bulk flux density into the
circumferential direction may be the simplest way to meet See Fig. 32 for a diagram of these dimensions. Tube. wall
the specification. thickness T is often much smaller than the average radius of
An additional consideration, unrelated to the physics of the tube. Under such conditions, Eq. 13 may be converted
test object magnetization, is the safety of the system in to:
permanent and field testing situations (the National Electric
Code should be consulted for details). For field use, it is r + !..
essential to limit the charging voltage of the capacitor bank
to 50 V. The tendency of this limit is to add capacitance to L ~ (dB)
27T dB
In 2 (Eq. 14)
r - T
the system.
2

or
Resistance Values
The resistance of the magnetization system is an impor- L= T)
(f27Tr
-- dB (Eq.15)
tant factor in permitting high currents to flow. Resistance is - dB
minimized for field tests by using parallel strands of 0000
(100 mm 2 ) copper welding cable for the connections be- In gaussian units, Eq. 15 becomes:
tween the rod and the capacitor discharge box.
The rod is made of aluminum, mainly for its handling
advantages, but any highly conductive material works equal-
L == 2 X 10- 7 (e -T) -dB
r
dB (Eq. 16)
ly well. The requirement of elongating the pulse length to
ensure the presence of its field after eddy currents have where r is 1/2(R" + Ri). All lengths are in meters and dB/dB
dissipated far outweighs the requirement of minimizing the is dimensionless.
 414 / MAGNETIC PARTICLE TESTING

The inductance of thin-walled tubes is seen from Eq. 15 Bs 1.5 T (15 kG)
=
to be proportional to the tube length e and wall thickness T e= 9.1 m (30 ft)
and inversely proportional to its radius or diameter. Neither T = 4.8 mm (0.2 in.)
of these physical parameters nor the value of dB/dH can be 28 mm (1.1 in.) and
controlled by the designer of the magnetizing equipment. dH = 3,200 A-m -1 (40 Oe)
However, for much of the tubular products used in oil fields,
the value of T/R does not val)' a great deal (perhaps by a The inductance value is calculated in the SI system as:
factor of two).
The average magnitude of dB/dH encountered during
L 9.1 m ( 4.8 mm ) ~
magnetization can be seen from Fig. 33 to val)' widely. The 271' - 28 mm 3,200
value is dependent on: (1) the point reached by the material L 116 J.LH
on the B-H CUIve during magnetization; and (2) the starting
point for magnetization (anywhere from - Br to Q on the or in the gaussian system as:
vertical axis of Fig. 35).
As an example of a typical inductance calculation, con- L (2 x 10- 7 )(9.1 m)(4.8 mm) 15,000/40
sider a pipe magnetized to saturation following the path 28.mm
- BrHcPQ and having: L 119J.LH

Br = 1.2 T (12 kG) The relatively large change in inductance exhibited by the
P= 1.2 T (12 kG) tube in the first two examples affects the shape· of the pulse
e= 10 m (30 ft) waveform, notably the easily measurable parameters of peak
T = 12.6 mm (0.5 in.) current Imax and pulse duration T (see Fig. 28):
R = 136.5 mm (5.25 in.) and
H = 2,400 A-m -1 (30 Oe) Design Considerations
Here dB is 2.3 T (23 kG), so that dB/dH is 0.001 in the SI Good equipmynt design must consider the material being
system (800 in gaussian units) and L in 81 units is: . magn~tized. The' worst internal and external resistances of
the magnetizing system should be known to the manufac-
12.6 mm turer and corresponding inductance values should be inves-
L 10 m 27T _ 136.5 mm 0.001 tigated. "tJnder no circumstances should peak currents be
stated for the purpose of magnetization without an elechical
L 141 J.LH and magnetic load being used for the system evaluation.
Depending on the use of the equipment, regulations
or, in gaussian units: should be consulted with regard to insulation, isolation,
explosion proofing, intrinsic safety and purging. Such regu-
L (2 X 10- 7 )(10 m)(l2.6 mm) 800 lations are available from many sources, depending on the
136 mm use of the product. Notable among these are the Occupa-
L 148 J.LH tional Safety and Health Administration and the National
Institute of Safety and Hygiene. Various foreign standards
Using the same tube as above, a path is taken from Q are considerably more stringent than those in the United
through P and Bs to Br by a second pulse, with: States. Equipment deSigners should pmticularly note the
requirements of the United Kingdom, NOlway and West
Q = 1.0 T (10 kG) Germany when designing for North Sea applications. The
Bs = 1.5 T (15 kG) and Canadian Standards Administration should be consulted
dH = 4,000 A-m -1 (50 Oe) when designing for applications in Canada.
The average value of dB/dH is then only 118,000 (1.25 x
Magnetization Recommendations
10- 4 ) in the SI system or 100 in the gaussian system.
Dividing the original inductance value by 8 (the ratio of the Tubular products v,:uy between such wide limits of
two dB/dH values exhibited by the steel) yields: diameter and wall thickness that it is difficult to provide a
universal specification for the measurable parameters of
L = 18 ..5 J.LH current pulses for high residual induction. However, the
values in Table .5 are based on research with a vmiely of
As another example, consider an unmagnetized pipe that tubes and can serve as broad guidelines to ensure adequate
follows a path OPBsBr with: magnetization. 1.'5
 SPECIAL APPLICATIONS OF MAGNETIC PARTICLE TESTING / 415

As tabulated, the pulses are classified into long, moderate TABLE 5. Generalized current duration requirements
and short duration. Long duration pulses are those in excess for adequate magnetization of tubes
of 100 ms (see Fig. 34). For such pulses, the induced eddy Magnetization Duration Current Requirement
current is assumed to have dissipated while the magnetizing System (millisecond J Equation
field strength is still high enough to cause saturation.
Moderate pulses are those with durations between 40 and Long pulse >100 I=II.8(Od
100 ms (Fig. 34). For magnetization, the longevity of the 1= 300 (0 2 )
induced eddy current is acknowledged by its effect on the Moderate pulse 40 to 100 1= 74 (Wd
tuhe, shown in calculations by the use of the tube's lineal 1= I 10 (W2 )
mass rather than its outer diameter. Single short pulse o to 40 1=161 (Wd
1=240 (W2 )
Short pulses are defined as those with durations below
40 ms. By comparing the requirement for the single short
Double short pulse ° to 40 1=121 (Wd
1= 180 (W2 )
pulse with that for the single moderate pulse, it can be seen Triple short pulse Ot040 1= 97(Wd
that the maximum current requirement is higher for the 1= 145 (W2 )
same lineal tube mass. In effect, higher current causes a
I = CURRENT IN AMPERES
larger magnetizing field strength, in an attempt to overcome 01 = OUTER DIAMETER IN MILLIMETERS
the eddy current. 02 = OUTER DIAMETER IN INCHES
WI = TUBE WEIGHT IN KILOGRAMS PER METER
Should it be necessary to use two such pulses, the peak W2 = TUBE WEIGHT IN POUNDS PER FOOT
current requirement falls because the material is partially
magnetized. If the peak current can only reach an 1max value
of 180(W), where W is tube weight in Ibeft -1, then two such
pulses are required. Finally, should the pulse be t~o small to FIGURE 36. A technique for measuring flux
magnetize the tube with two pulses,. then a third pulse is density induced in the circumfer.e~tial direction
• necessary. The three pulse sequence should be such that
1max is equal to 145(W). .
The requirements summmized in Table 5 are deSigned to
ensure that the bulk induction follOwing the magnetization
pulse is at least 90 percent of the remanence value. In many
cases, it is higher.

Evaluating Current Pulse

Effectiveness
There are two methods for evaluating the effectiveness of
a magnetizing current pulse. The first method is a variation Where:
on the ring standard technique for the evaluation of mag-
netic flux. The second method is an indirect technique that = changes in magnetic flux density (tesla);
~<I>
employs an inductive ammeter (peak and duration meter). A A = the area of the test object (A = TC) perpendicular
third method, using simulated contact discontinuities (field to the search coil (square meter); and
~B = the change in the test object's flux density induced
indicators), is a valuable technique for evaluating smface
fields only. during magnetization.

COml)1ercially available flu.xmeters often compensate for

the area A so that the device reads the average flux density
Fluxmeter Techniques
directly. The problem with this approach for measuring the
Fluxmeters measure the total magnetic flux threading an final flux density is that the initial flux density (with respect
area defined by a search coil. \iVhen circumferentially to the vector direction of the search coil) must be zero.
magnetizing a hollow object by the internal conductor "Then flux changes are measured, the initial value must be
method, the search coil can be considered a one-tum coil known. However, if the tube shown in Fig. 36 is unmagne-
through the test object. Flux changes are given by: tized or if prior magnetization is longitudinal, then the
fluxmeter reads the average density of induced circumfer-
~<I> = A~B (Eq. 17) ential magnetization.
416/ MAGNETIC PARTICLE TESTING

The output of the fluxmeter can be presented on an Figure 37 indicates that three phenomena occur (!tiling
oscilloscope. Flux values intermediately between the begin- pulse magnetization: (1) inner and outer surfaces are rapidlY
ning and the end of the magnetization process represent the magnetized; (2) the midwall region is the last part of th~'
flux linked bv the one-turn coil. These intermediate values material to be magnetized; and (3) the midwall regioll call
contain the' effect of the flux in the air between the be left with a low state of magnetization if the pulse field
terminals of the fluxmeter. strength is insufficient to saturate the materiaL
During the pulse, the air field (caused by current [ in the This last phenomenon contributes to magnetic fIelds from
rod) and eddy currents (Ie in the test object) affect the discontinuities at one surface but produces no leakage field
instantaneous fluxmeter reading. When these currents have at the other surface when the material is not saturated, The
died away, only the test object flux perpendicular to the leakage field into the midwall section of the material merely
one-turn coil affects the final reading. If the operator has raises the local magnetization level to a higher degree.
time to wind more than one turn around the test object, the
error in final bulk flux density can be reduced. However,
this procedure is not necessary for establishing the presence FIGURE 37, Plots against elapsed time and wall
of residual induction in the test object for magnetic flux thickness of a tube for fa J magnetiC field
leakage f~om discontinuities. strength; and fbJ magnetic flux density; lower
curves represent magnetic field and induction at
Inductive Ammeter Techniques the beginning of the pulse; time proceeds from
bottom to top; central regions are the last to be
As shown in Fig. 31b, the pickup coil of the device is
magnetized
threaded onto a convenient part of the magnetizing circuit.
When the pulse is fired, the meter reads the peak current fa) 10 00
(I max of Fig. 34) and the duration of the pulse (T of
Fig. 34).20 2: u 8,000 (100) ~----;------------i
1J~
Saturation of the material has- occurred when successive Z~
LU ClJ
readings on the ammeter are identical. This can be ex- 0<: 0
I- -
Vl '--
plained as follows. When the first pulse is fired, the material o~
exhibits its highest dBldH valu'e (th'e steep slope of the B-H ~E
curve is included in the value of L). The average value of u::: '--
u~
dBldH is effective for determining the value of the induc- i=Vl
LUClJ
tance in Eq. 9 through Eq. 13. This value is relatively large zv
in comparison with the value exhibited during a second ~E-
pulse. Figure 35 indicates high values of dBldH for a first 2 1iJ

pulse (about 0.0008 in SI and 800 in gaussian) and much

lower values during the second pulse (around 1.25 X 10- 4 20 40 60 80 100
in SI or 100 in gaussian). In effect, the second pulse DISTANCE FROM INNER TO
experiences a, lower inductance than the first pulse. OUTER SURFACE
The second pulse's lowered inductance permits peak (percent)
current [max to reach a higher value than it reached on the fbI 10 OD
1,600 (16)
first pulse. In effect, the material is different. The system
>-
response is also to lower the duration T. By monitoring [max I-
Vi_
1,400 (14)

and T, it is possible for inspectors to determine the test ZVl

Vl
1.200 (12)
LU
object's relative magnetization. o~
X01 1,000 (10)
:J,g
-.J -"L
Using Inductive Ammeters LL - 800 (8)
~~
I-ClJ
vVhen materials are magnetized with pulse techniques, LU·:=
600 (6)

the values of field strength H and flux density B change with Z==
1J E 400 (4)
time as shown in Fig. 37. The horizontal axes show the «
2 200 (21
percent distance from the inner surface to the outer surface.
The vertical axes show either the fraction of H required to 20 40 60 80 100
saturate the material or the flux density B. The lowest lines DISTANCE FROM INNER TO
show time from the shut of the pulse. The uppermost lines OUTER SURFACE
show the field strength and flux density levels at later time (percent)
increments.
 SPECIAL APPLICATIONS OF MAGNETIC PARTICLE TESTING I 4 t 7

When sections of the test object are not maintained at vVhere:

field strength levels sufficient for saturation (point P in
Fig. 3,5), then the ensuing bulk residual flux density is low N = the number ~f turns in the ring;
and the material requires additional pulses to saturate it. d = the axial length of the ring;
The magnetization process calls for the highest value of b = the outer radius of the ring;
inductance L in Eq. 9 through Eq. 15 dming the first pulse a = the inner radius of the ring; and
and lower values during subsequent pulses. The general dIldt = the rate of change of current.
effect of a high inductance value is to lower the value of [max
and elongate the value of T. This equation is derived from Faraday's law of induction. In
In order to verify this and to limit the necessary mathe- order to prOvide a signal related to the current itself, Eq. 21
matical computation, Eq. 11 was selected and from it the must be integrated. The result is:
closed form results for [max and Tare found below. First, the
time'T at which [m.lX occurs is found by differentiation of.
Eq. 11 to be: e f
= E • dt = [(2 X 10 -7)Nd In ~ J1 (Eq.22)

2L
t =- (Eq. 18) where e is the output of the integration circuit.
R In effect, since all the terms in the brackets are known,
the output of the· integration of induced voltage is propor-
When this value is used in Eq. 11, the result for [max is: tional to the instantaneous current and the instrument can
be calibrated to read current directly. Electronic circuits are
(Eq. 19) used to measure the peak current 1max and the pulse
duration T.
ThiS. result indicates that the value of 1m,lX is inversely
pmportional to that of L (greater L gives lower 1mruJ In . .
order to find T, l(t) is set at 0.5 [max' The result is: . Use of Field Indicators and
Simulated Discontinuities
(Eq.20)
Typical Configurations of Field Indicators
In this case, the pulse duration defined by T is proportional
OccaSionally, a magnetic particle test operator needs to
to the value of L. That is, latger values of L, such as those
verify the orientation of induction after tubes have been
for the initial pulse, lead to larger T values (the longest pulse
durations). magnetized by the capacitor discharge internal conductor
technique. In the case of perfectly concentric tubes with
The B-H curve indicates that the lowest value of induc-
constant metallurgical properties throughout, there should
tance that can occur under these magnetization conditions is
be no external leakage field. This fact eliminates the use of
exhibited by saturated material, when the value of dBldH is
Berthold reference standards (elevated discontinuities sur-
at its lowest (see Eq. 15). If two identical readings are
rounded by high permeability material) in this application.
obtained from an inductive ammeter, the material must be
Figure 38 illustrates three common field indicators (er-
exhibiting its lowest inductance to the magnetizing circuit
roneously called penetrameters) widely used in magnetic
and must therefore be at remanence B r .
particle tests of tubes.
Operating Principles of Inductive Ammeters
The inductive ammeter is a microprocessor with an Pie Gage and Raised Cross Devices
inductive pickup coil. The toroid coil contains a large Compared to the raised cross device, the pie gage has
number of turns wound onto a nonconducting nonmagnetic somewhat wider acceptability for tests of tubes, mainly
ring core. The ring is threaded onto cables from a capacitor because it can sit closer to the test object surface than the
discharge system or onto a central rod. When a pulse is raised cross configuration.
fired, the flux caused by the current surge links with the The raised cross device contains perpendicular slog with
windings on the ring and the voltage induced in the coil is a 0.13 mm (0.005 in.) gap. A nonmagnetic shield is some-
given by: times used to cap the device ..The pie gage comprises six 60
degree sections of high permeability steel with no cap and
E 2 X 10 - 7
1
N d d In ~ (Eq.21) gaps up to 0.75 mm (0.03 in.). In both cases, the width of
dt a the air gaps determines the sensitivity of the device.
418 I MAGNETIC PARTICLE TESTING

FIGURE 38. Typical portable magnetic field FIGURE 39. The magnetic force on an isolated
Indicators: (aJ raised cross with nonmagnetic cap; particle in the flux leakage field from a slot;
(bJ pie gage; and fc) strip device; widths of the the semicylindrical magnetic flux leakage
air gaps or slots made from nonmagnetic approximation is according to Foerster
materials differ between the types (see Equation 26)
y

fa)
/"
/' / - - -...............

cx::J I I " / / /'

/ I I/',....
( I I / I /'"
I I I ( I /,
--
x

fbJ
H~Jl~

fe) Principles of Operation

The principle of operation for aU field indicators relies on
the ability of the device to tum a relatively constant surface
• • • field into a highly curved field (it is well known that
magnetic particles are attracted to highly curved fields). The
surface field of the test object is converted by a air gap or
slot into a curved field. The magnetic force on a particle in
The device's sensitivity may be defined as the ability to such a field is given by:22
produce a particle indication in an ambient magnetic sur-
face field. In simple Foerster leakage field theoI)',13 the (Eq.23)
particle holding ability depends on the product (H L g )2,
where Hg is the magnetic field strength in the gap anJ Lg is Where:
the gap width.
a = a demagnetization factor of the particle (dependent
Strip Indicators on particle shape);
The field indicator preferred for tube tests are strip }-to = the permeability of hee space (4 'Tf X 10 - 7 H -m - l );

indicators (Fig. 38). Such strips typically contain three V = the particle volume; and
discontinuities encapsulated for support in high permeabil- V = the vector calculus gradient operator.
ity material (often brass). The strip is placed in intimate
contact with the test object after magnetization so that the For a two-dimensional discontinuity (Fig. 39) placed at right
lift-off is limited to the thickness of the brass carrier. Under angles to the surface field, the vector operation reduces to:
such circumstances, the magnetized test object shares flux
with the strip. (Eq.24)
The encapsulation is necessary because ships contain
discontinuities that are very small in comparison to the other
types of indicators. In effect, the gaps in the high perme- where fIx and Htl are the leakage Held components in the
ability base material are activated by the same surface fields region above the slot. Calculations beyond this point require
that activate magnetic flux leakage from similarly sized equations for the leakage field components. The simplest
discontinuities in the test object. relation for a crack leakage field is given by:
 SPECIAL APPLICATIONS OF MAGNETIC PARTICLE TESTING I 419

HgLgY Where:
Hx (Eq.25)
7T(X2 + y2)
AA = the cross section area at A;
and Ac = the cross section area at C; and
HgL~
Hy 7T(X2 + y2)
B,.. = the remanence value for the test material.

If the leakage field falls off in an inverse manner with

distance, then the strongest surface leakage flux B" is
These equations work reasonably well as long as the particle
is located at a distance more than Lg from the mouth of the approximated by:
slot, as is the case if the slot has a thin nonmagnetic cover. dR
Bo = B"'R (Eq. 28)
The result of combining Eq. 25 with Eq. 24' is:

J.t oV(HgLg)2
(Eq.26) where dR is the amount of eccentricity present in a tube of
Fm IX 7T 2r.3 radius R. Equation 28 indicates that:

where r is the distance of the magnetic particle from the 1. the higher the remanence B,.., the stronger the leakage
mouth of the slot. In this simple approximation, Eq.26 field;
indicates that the magnetic force holding a particle is: 2. larger eccentricities give larger leakage fields for the
(1) proportional to the gap parameter (HgLg)2 and various same radius and remanence; and
particle dimensions; and (2) proportional to the iiwerse 3. for the same B,.. and dR, the leakage field falls off as
cube of its distance from the mouth of the slot. It is the tube diameter is raised (leakage fields are stronger
necessary then to determine the cause of .fig.. • on the smaller tubes).
Fortunately, when seamless tubes are 'made by the
mandrel piercing process, there is generally some eccentric- At low inducti.on values, the flux at A and C can be equal
ity present (see Fig. 40). If the material is magnetized to without the presence of a leakage field. Figure 40 shows that
remanence by' the capacitor discharge internal con·ductor the flux density at C (the thin wall) must be the remanence
method, then the conservation of magnetic flux indicates value for the material if there is to be any leakage field.
that the flux present in the material at point A must equal When this occurs, the flux denSity BA at A obeys the
the total of that at C plus the leakage flux. Consideration of relation:
the tube's cross section areas at A and C yields:
(Eq.29)
<l>leakage = (AA - AdB,.. (Eq.27)
Before this occurs, it may be possible to obtain a magnetic
particle indication with a strip indicator. Because there may
be no external fields, it is not possible to obtain an indication
with the pie gage or other field indicator. It must then be
FIGURE 40. Broad magnetiC flux leakage field recognized that metallurgical variations within the tube
from an eccentric circumferentially magnetized affect B,.. and cause some external leakage flux.
tube with points A and C at remanence Br; such
leakage fields can activate magnetic field
Indicators (portable discontinuities)
Documents for Magnetic Particle
Testing of Tubes
Table 6 lists American Petroleum Institute specifications

" \
\
and recommended practices that refer to magnetic particle
testing for tubular materials. Of these, RP 5A5 addresses
the most commonly used magnetization methods, plus wet
\ \ \
and dry magnetic particle testing procedures. Many oil and
I
I I
I I gas companies now produce their own specifications, in-
/ I I cluding those for magnetic particle tests.
'" / ;¢lEAKAGE In most documents, particular attention is given to the
,. /
follo\\ing parameters: (1) wet particle concentration; (2) dry
particle bulk permeability; (3) dry particle visual contrast
with the inspected surface; (4) dry particle filler content;
420 I MAGNETIC PARTICLE TESTING

(5) verification of saturation for circular magnetism; (6) in- TASLE 6. Specifications and recommended practices
tensity of ultraviolet light at the test surface; (7) time referring to magnetic particle testing of tubular
intervals between ultraviolet intensity checks; (8) relation products; except as noted, all documents were
between field strength at the center of a coil and the current written for tests of tube body, internal upsets,
through the coil; and (9) time intervals between magnetiz- external upsets and pipe couplings
ing coil current calibration.
Though not always specified, it must be recognized that Document Use
the test object surface condition plays a critical role in
determining which kinds of magnetic particles are accept- Specification 5CT Casing, tubing and drill pipe
able for specific applications. Recommended Practice 5A5 Field evaluation of casing, tutJ-
High contrast is required for visibility and accurate ing and plain-end drill pipe
interpretation. High permeability is required so that the Recommended Practice 7G Drill stem design and operatinq
limits -
particles are magnetized by relatively low leakage fields
Specification 5L* Line pipe
from tight cracks. High bulk permeability can be verified by
a test performed on bulk powders placed in a chemical test
*FOR VERIFICATION OF WELD REPAIRS
tube.
 SPECIAL APPLICATIONS OF MAGNETIC PARTICLE TESTING 1421

REFERENCES

1. Goldberg, L. Reliability of Magnetic Particle Inspection

13. Standard Recommended Practice for Magnetic Particle
Performed Through Coatings for Offshore Structures.
Inspection. ASTM E-709, Section 5.3.1. Philadelphia,
Houston, TX: Exxon Produ~tion Research Company
(1985). FA: American Society for Testing and Materials
2. Goldberg, L. Philosophy for Undenvater Weld Inspec- 14. Magnetic Particle Inspection. MIL-STD-1949, Section
tion. Houston, TX: Exxon Production Research Com- 4.9.1. Washington, DC: Department of Defense (1985).
pany (1986) 15. Moyer, M. "Magnetic Requirements for Oilfield Tub-
3. Certificate of Conformance. 86-NY-2902-X. New York, ulars." Materials Evaluation. Vol. 44, No.6. Columbus,
NY: Amelican Bureau of Shipping (1986). . OH: The American Society for Nondestructive Testing
4. Goldberg, L. Reliability of Magnetic Particle Inspection (1986): p 616. See also Specificatiom for the Nonde-
Pelformed Through Coatings. Palo Alto, CA: Electlic structive Evaluation of API Oilfield Tubular Goods;
Power Research Institute (1988). Revision 1. Houston, TX: Exxon Production Research
5. Cook, J. Nondestructive Examination afWelds Through Company (May 1984).
Painted Suifaces. Palo Alto, CA: Electlic Power Re- 16. Stanley, R. "Circumferential Magnetization of Tubes
search Institute (1988). •. . aqd the Measurement of Flux Density in ·Such Mateli- .
6. Nondestructive Examination . .Section V, 1986 edition, als." Materials Evaluation. Vol. 44, No.7. Columbus,
1987 addenda. New York, NY: Amelican Society of OH: The American Society for Nondestructive Testing
Mechanical Engineers (1987). (1986): p 966-970.
7. Stanley, R. "Magnetic Field Measurement: The Gauss 17. Stanley, R. "Basic Principles of Magnetic Flux Leakage
'Meter in Magnetic Particle Testing." Materials Evalu- Inspection Systems" and "Capacitor Discharge Magne-
ation. Vol. 46, No. 12. Columbus, OH: The Amelican tization of Oil Country Tubular Goods." Electromagnet-
.Society for Nondestructive Testing (1983): p 1,509- ic Methods of Nondestructive Testing (Nondestructive
1,512.
Testing Monographs and Tracts). W. Lord, ed. Vol. 3.
8. Goldberg, L. Magnetic Particle Impection Using Re- New York, NY: Gordon and Breach Publishing (1985):
mote Operated Vehicles. Houston, TX: Shell Oil Com- p 96-160.
. pany (1989).
18. "API Recommended Practice for Drill Stem DeSign
9. Hagemaier, D. "A Critical Commentary on Magnetic
and Operating Limits." Recommended Practice API RP
Particle Inspection." Materials Evaluation. Vol. 41,
7G, twelfth edition. Dallas, TX: American Petroleum
No.9. Columbus, OH: The American Society for Non- Institute (May 1987).
destructive Testing (1983): p 1,064-1,065.
10. Brunk, J. Improved Magnetic Particle Inspection of 19. "API Specification for Casing and Tubing." Specifica-
Small Threaded Fasteners. BDX-613-3283 (DE-AC04- tion SCT, first edition. Dallas, TX: American Petroleum
76-DP00613). Washington, DC: Department of Energy Institute (March 1988).
(1986). 20. Schindler, J. Current Pulse Monitor. US Patent
11. Burkle, \V. and B. Fraser. "Dry Magnetic Particle 4,502,004 (1985).
Examination of Painted \Velds Using an AC Yoke." 21. Foerster, F. "Nondestructive Inspection by the Method
Materials Evaluation. Vol. 44, No. 10. Columbus, OH: of Magnetic Leakage Fields: Theoretical and Experi-
The Amelican Society for Nondestructive Testing mental Foundations of the Detection of Surface Cracks
(1986): p 1,156-1,161. of Finite and Infinite Depth." Defektoskopiya (Novem-
12. Burkle, W. and B. Fraser. "The Effect of Mechanical ber 1982): p 3-25.
Paint Removal on the Detectability of Cracks by Visual,
22. Schwartzendruber, L. "Magnetic Leakage and Force
Magnetic Particle, and Liquid Dye Penetrant Testing."
fields for Artificial Defects in Magnetic Particle Test
Materials Evaluation. Vol. 45, No.8. Columbus, OH:
Rings." Proceedings of the Twelfth Symposium on
The Amelican Society for Nondestructive Testing
(1987): p 874-875. Nondestructive Evaluation. San Antonio, TX: South-
west Research Institute (1979).
IIII1I11111111111111111111111111111111111111111I11111I111111111111111111111111111111I111111111111

SECTION

CODES, STANDARDS AND

17
SPECIFICATIONS
Ken Kremer, McDonnell Aircraft Company, St. Louis, Missouri
424 I MAGNETIC PARTICLE TESTING

PART 1
INTRODUCTION TO MAGNETIC PARTICLE
SPECIFICATIONS

Defi nition of Terms The Need for Specifications

In general usage, a standard is a document written by Specifications are written to eliminate the variable cllar-
recognized authorities to recommend actions for achieving acteristics of human operators and system deSigns, t()
certain goals. Theory and a broad educational foundation produce an accurate result regardless of who is performillg
are sometimes included to justify the recommendations. A the magnetic particle test. Specifications must be wn'ttC'1I
standard can present minimum parameters or it can suggest with a full knowledge of (1) magnetic particle test tech-
certain courses of action based not on minimums but on niques; (2) a technique's individual sensitivities; (3) the ,test
efficiencies such as cost, labor or quality. A standard is object's design; (4) its material characteristics; and (5) tIle
typically enforced or given authority by an agent separate discontinuities critical to the test object's service life. In
from its author. When a standard becomes law, it could then most manufacturing applications, nondestructive tests art'
be called a code. • considered during design.apd such specific<~tions are part ()f
In nondestructive testing, a specification is often written the' test object's original drawing.
by a commercial organization, usually one of the primary Magnetic particle specifications are produced to stan-
parties in a purchasing agreement. A specification is product dardize test results, not to eliminate the initiative (1' tlti'
specific and may be considered a tailored form of standard. technician. There is no substitute for an experienced oper-
A specification can require more stringent limits than a ator who assumes personal responsibility for the quality and
related standard's limits. In practice, a specification pro- accuracy of the test.
vides a clearly organized list of testing parameters (a specially Table 1 is a summary of specifications commonly llsed fur
written procedure) that describes the technique for locating magnetic particle testing and material selection.
and categorizing discontinuities in a specific test object. A
typical specification includes acceptance criteria and is
required by the designer, buyer or manufacturer of the Need for Review and Revision
article it covers. of Specifications
.For magnetiC patticle tests, the term procedure refers to
a set of brief generalized guidelines that show the technician Advances in the sciences have a three-fold effect Oil
how to perform an accurate test for a given contract. A nondestructive tests and their specifications: (1) new ways
procedure often includes details about the in-house testing can be f()Und to produce or interpret test results; (2) tests
setup. Other details, based on the organization's past expe- must be developed to accurately examine newly developed
rience, provide the practical information needed in a pro- materials; and (3) test procedures must be designed as
cedure. Occasionally, procedures are written for broad integral parts of the modern manufactming environnwlll.
categories of material or for specific classes of test objects. All of these developments affect magnetic paltide tests and
These procedures are not required or reshictive but edu- must be reflected in their specifications. The guidelines
cational in intent. used in countries such as Japan and \Vest Germany differ in
Unfortunately, a standard is also a reference test object many respects from those in the United States, and the:
\vith known discontinuities used to verify the accuracy of a may give better test results in some applications, In the
testing procedure. In addition, the words standard, specifi- United States, many specifications are being updated to
cation (spec) and code are wrongly used as synonyms in make better use of artificial discontinuities and multidin'('-
some applications and they are often used differently tional magnetic fields.
depending on the discipline in question. The definitions Testing specificatiolls are working documents that tell
above are prOvided only for the educational purposes of this how to locate discontinuities in a specific test object. Even
volume. for well established and successful methods like magnetic
 CODES, STANDARDS AND SPECIFICATIONS I 425

TABLE 1. Summary of magnetic particle testing specifications; because latest revisions always apply, contact
the issuing organization for current documents
Issuing Organization Document Title

American National Standards Institute

ANSI/ASME B31. 1 Power Piping
ANSI/ASTM A275 Magnetic Particle Examination of Steel Forgings
ANSI/ASTM A456 Magnetic Particle Inspection of Large Crankshaft Forgings
ANSI/ASTM E 125 Reference Photographs for Magnetic Particle Indications on Ferrous Castings
ANSIIASTM E269 Definition of Terms Relating to Magnetic Particle Inspection

American Petroleum Institute

API5CT Specification for Casing and Tubing
API 5D Specification for Drill Pipe
API 5L Specification for Line Pipe
API5LU Specification for Ultra High-Test Heat Treated Line Pipe
API RP 5A5 Recommended Practice for the Field Inspection of New Casing, Tubing and Plain End Drill Pipe

American Society for Testing and Materials

A275 Magnetic Particle Examination of Steel Forgings
A456 Magnetic Particle Inspection of Large Crankshaft Forgings
E45 Standard Practice for Determining the Inclusion Content of Steel
E1:25 Standard Reference Photographs for ·Magnetic Particle Indications on Ferrous Castings
E269 Definition of Terms Relating to Magnetic Particle Inspection
E709 Standard Recommended Practice for Magnetic Particle Examination
D93 Flash Point by Pensky-Martens Closed Tester
D96 Water and Sediment in Crude Oils
D445 Kinematic Viscosity of Tra~sparent and Opaque L[quids (and the' Calculation of Dynamic Viscosity)

American Society of Mechanical Engineers IBoiler and Pressure Vessel Code)

Sec I Power Boilers
Sec" Material Specifications
Sec III Nuclear Power Plant Components
Sec V Nondestructive Examination
Sec VIII (Div I) Unfired Pressure Vessels
Sec VIII (Div 2) Alternative Rules for Pressure Vessels
Sec XI Rules for In-Service Inspection of Nuclear Power Plant Components

American Welding Society

D1.I Structural Welding Code
D14.6 Welding of Rotating Elements of Equipment

Society of Automotive Engineers

AMS 2300F Premium Aircraft-Quality Steel Cleanliness, Magnetic Particle Inspection Procedure (also MAM 2300)
AMS 2301 G Aircraft Quality Steel Cleanliness, Magnetic Particle Inspection Procedure
AMS 2303A Aircraft Quality Steel Cleanliness, Martensitic Corrosion Resistant Steels, Magnetic Particle Inspection Procedure
AMS 2640J Magnetic Particle Inspection
AMS 3040 Magnetic Particle Inspection, Material Dry Method
AMS 3041 A Magnetic Particles, Wet Method, Oil Vehicle
AMS 3042A Magnetic Particles, Wet Method, Dry Powder
AMS 3043A Magnetic Particles, Wet Method, Oil Vehicle, Aerosol Canned
AMS 3044B Magnetic Particles, Fluorescent Wet MethOd, Dry Powder
AMS 3045B Magnetic Particles, Fluorescent Wet Method, Oil Vehicle, Ready to Use
AMS 30468 Magnetic Particles, Fluorescent Wet Method, Oil Vehicle, Aerosol Packaged
AMS 3161 Inspection Vehicle, High Flash, Odorless
CONTINUED NEXT PAGE
426 / MAGNETIC PARTICLE TESTING

TABLE 1 continued

Issuing Organization Document Title

United States Department of Defense

MIL-STD-27 J Nondestructive Testing Requirements for Metals (ACN- J )
MIL-STD- J 949 Magnetic Particle Inspection
MIL-I-6867 Magnetic Inspection Units
MIL-M-23527 MagnetiC Particle Inspection Unit. Lightweight
MIL+83387 Magnetic Rubber. Inspection Process
DOD-F-87935 Fluid. Magnetic Particle Inspection. Suspension (Metric)
MIL-M-47230 Magnetic Particle Inspection Soundness Requirements for Materials. Parts and Weldments
MIL-STD-4 J 0 Nondestructive Testing Personnel Qualification and Certification
MIL-STD-2175 Castings. Classification and Inspection of
FED-STD-595 Colors

United States Department of Energy

E J5-2NB-T Supplementary Requirements for Sec /II ASME Code
EJ 5-2NC-T Supplementary Requirements for Sec "' ASME Code
E J 5-2ND-T Supplementary Requirements for Sec /II ASME Code
E J5-2NE-T Supplementary Requirements for Sec III ASME Code
RDT-F3-6T Nondestructive Examination (Supplement to ASME Boiler and Pressure Vessel Code Sec V)

British Standards Institution

BS 6072 Method for Magn"etic Particle Flaw Detection

Deutsche Gesellschaft fur Zerstorungsfreie Prufung

Guidelines for Magnetic Particle Flaw Detection

Japanese Institute of Standards . •

JIS G 0565 Methods for Magnetic Particle Testing of Ferromagnetic Materials and Classification of
Magnetic Particle Indications

particle testing, specifications are in regular need of review critical areas. In other cases, the fields were very low and not
and revision. As an illustration of this, consider one piece of strong enough to hold magnetic particles. Adja'cent areas of
peripheral equipment used in establishing test procedures: simple geometry were often found to be inagnetized to
the magnetic field indicator. levels satisfactory for testing. These empirical data indicated
The increased reliance on artificial discontinuity stan- that existing current level equations could be applied only to
dards, pie gage field indicators, tool steel ring standards and the Simplest test object geometry and were unreliable for
tangential field indicators is driven by incidents of cata- complex geometries. It is very important that this kind of
strophiC failure after magnetic particle tests failed to detect knowledge be incorporated as quickly as possible into
critical discontinuities. These unreliable tests were per- industty specifications.
formed using practices found in early specifications. A As another example of the need for review and revision,
major problem was that current level equations suitable for consider the increased use of alternating current tests
simple geometries were applied to more complex geome- outside the United States. This has occurred because of
tries. Discontinuities in comers, radii, bearing races, grooves alternating current's ability to produce a nearly uniform
and other areas are not detected when these formulas are magnetic field over the entire surface of a test object.
inappropriately applied to complex test objects. However, alternating current, because of its low penetrating
Subsequent field strength measurements were performed ability, further enhances the need for field indicators and
in areas of known failure on test objects of similar design. In surface field measurements, and increases the demand for
some cases, there was no magnetic leakage field in the their inclusion in test specifications.
 CODES, STANDARDS AND SPECIFICATIONS I 427

PART 2
SAMPLE MAGNETIC PARTICLE
SPECIFICATION

The following text is provided as a simple introduction to 2.0 Reference Documents

magnetic particle specifications and is not a sample working
document. It contains much useful information about per- List of all documents addressed or referenced in the
forming a typical magnetic particle test, including items specification.
found in most North AmeIican specifications, plus details
unique to Japanese and West German specifications. Typi- 3.0 General Requirements
cal parameters are given for the performance of magnetic 3.1 Magnetizing and Demagnetizing Equipment
particle tests using any of the electIic current techniques.
By following the specified instructions for artificial dis- Magnetization can be accomplished either by passing
continuity standards and magnetic field indicators, it is likely an electIic current directly through the mateIial or by
that all surface discontinuities can be identified or the placing the material within the magnetic flux of an
.process's tn~bility to find discontinuities will be determined. . '. external ~ource such' as a coil. The types of equipment
available include yokes, portable or mobile units, station-
ary units arid multidirectional units'.
Arrangement of the Specification The .types of currents used for magnetization are
full-wave rectified unidirectional current (FWDC), half-
. This sample specification does not include. reference wave rectified single-phase direct current (HWDC) and
documents. Typically, all such docur.nents referenced in the alternating current (AC). The equipment used shall
body of the specification should also be detailed in Sec- adequately fulfill the magnetizing and demagnetizing
tion 2.0. requirements as outlined herein, without damage to the
Section 5.0 includes all definitions of terms needed in the test object and shall include the necessary features
specification, including general ter~ns for discontinuity de- required for safe operation.
scIiption and reporting. This section may list the terms and Permanent magnets shall not be used.
their definitions or may provide source documents where
appropIiate definitions are found. 3.1.1 EqUipment Calibration
If acceptance criteria are included in such a document,
they are listed in Section 5.0 and the glossary moves to Magnetic particle testing equipment shall be
Section 6.0. checked for performance and accuracy at the time
of purchase and at intervals thereafter as given in
Table 2; whenever malfunction is suspected; and
Sample Specification Text whenever electrical maintenance which might affect
equipment accuracy is performed.
1.0 Scope
3.1.2 Ammeter Accuracy
This specification establishes minimum requirements for
magnetic particle testing. \\Then specified, the magnetic To check the ammeter on a testing unit, a calibrated
particle testing method shall be used to detect cracks, laps, ammeter shall be connected in seIies with the output
seams, inclusions and other discontinuities at or near the circuit. Comparative readings shall be taken at three
surface of ferromagnetic materials. Magnetic particle test- output levels encompassing the usable range of the
ing may be applied to raw material, billets, finished and equipment. The equipment meter reading shall not
semifinished mateIials, welds and components in-service. deviate more than ± 10 percent of full scale from the
Magnetic particle testing is not applicable to nonferromag- current value shown by the calibrated ammeter. When
netic metals and alloys such as austenitic stainless steels. measuIing half-wave direct current, the current values
428 / MAGNETIC PARTICLE TESTING

TABLE 2. Typical verification intervals; maximum If the current is applied in shots and if alternating
time between verifications may be extended when current or half-wave direct current is used, the
substantiated by technical stability or reliability gaussmeter should have a frequency response frolll
data direct current greater than 300 Hz. The direction and
Maximum Time magnitude of the field can be determined by t\VO
Test Parameter Between Verifications measurements made at right angles with this meter.
Lighting
Ultraviolet light intensity J day 3.2 Lighting
Visible light intensity J week
Background visible light intensity J week 3.2.1 Ultraviolet Type and Intensity

System performance using ring standard J day Unless otheT"Wise specified, the ultraviolet light
intensity at the examination surface shall be
Wet particle concentration 8 hours 10 J.LW·mm -2 (1,000 J.LW-cm -2) or greater when
measured with a suitable ultraviolet light meter and
Water break test J week shall have a wavelength in the range of 320 to 400 Hill.
Portable or handheld ultraviolet light sources shall
Wet particle contamination I week
produce an intensity greater than 10 J.LW-mm - 2 (1,000
Equipment calibration J.LW-cm -2) when meastired at 380 mm (15 in.) from
Ammeter accuracy 6 months
the ultraviolet light source. The intensity of ultra,;io-
Timer control 6 months let light incident on unprotected skin or eyes shall
Dead weight check 2 months not exceed the levels established by local safety
regulations.

shown by the calibrat.ed direct current ammetec read-

ing shall be douDled. The ammeter shall be checked 3.2.2 Visible Light Intensities
according to the schedule in Table 2.
Visible light shall be used when testing with no"uI1ll-
3.1.3 Timer Control Check orescent particles or when assessing fluorescent. imli-
cations. The intensity of the visible light at the surhlce
On equipment using a timer to control current of the test object shall be maintained at a minimum of
duration, the timer should be checked to within ± 0.1 2,000 Ix (200 ftc). Unless otherwise specifled, Huo-
second using a suitable. electronic timer. . rescent magnetic particle testing shall be performed ill
a darkened area with a maximum visible light level of
20 Ix (2 ftc).
3.1.4 Dead ·Weight Check

Yokes shall be dead weight tested at intervals stated 3.2.3 Viewing Aids
in Table 2. Alternating current yokes shall have a lifting
force of at least 4.5 kg (10 lb) \vith a 50 to 100 mm (2 to When using fluorescent materials, inspectors shall
4 in.) spacing between legs. Direct current yokes and not wear eye glasses equipped with photochrolllic
permanent magnets shall have a lifting force of at least lenses (lenses that darken when exposed to ultraviolet
13.5 kg (30 lb) with a 50 to 100 mm (2 to 4 in.) spacing light or sunlight). Magnifiers may be used.
between legs; or 22.5 kg (50 lb) with a 100 to 150 mm
(4 to 6 in.) spacing.
3.3 Reference Standards and Field Indicators
3.1.5 Tangential Field Strength Meters A standard shall be available for procedure develop-
ment and system checks. The standard may be an actual
The active area of the Hall effect probe should be test object with known discontinuities; a replica with
no larger than 5 by 5 mm (0.2 by 0.2 in.) and should be discontinuities; a magnetic field indicator (see Fig. I); it
within 2 mm (0.08 in.) of the test object surface. The tool steel ling (see Fig. 2, Tables .3 and 4); an <.utiflcial
plane of the probe must be approximately perpendic- discontinuity or shim standard (see Fig. 3 and Tahle .5);
ular to the surface of the test object at the location of or a tangential field meter with an associated reference
measurement. standard.
 CODES, STANDARDS AND SPECIFICATIONS I 429

TABLE 3. Ring standard magnetic particle test

FIGURE 1. Pie gage magnetic field indicator
indications for full-wave rectified alternating
comprising eight low carbon steel triangles,
current (see Figure 2)
furnace brazed together and copper plated
Magnetizing Minimum
Magnetic Current Number of
Particle Type (amperes) Holes Indicated

Black suspension 1,400 3

2,500 5
3.400 6

Dry powder 1,400 4

2,500 6
3,400 7

Fluorescent suspension 1,400 3

2,500 5
E~ J.. r COPPER PLATE
0.25 mm (0.01 in.) ± 0.05 mm (0.002 in.)
3,400 6

~~
rvi~T·
[r=JO~'~~:~:~:~I==========~
..
TABLE 4. Typical distances from· outer edge for holes
in a tool steel ring standard; distance tolerances are
0.13 mm (0.005 in.); 110le numbers 8 through 12 are
FIGURE 1. Ring standard for magnetic particle optional depending on manufacturer (see Figure 2)
test system verificat40n (see Table 3 and Table 4); Hole Number Distance from Edge
material is AISI 01 tool steel from annealed
I 1.8 mm (0.07 in.)
round stock; hardness is 90 to 95 on the
2 3.6 mm (0.14 in.)
Rockwell B scale; dimension tolerances are
3 5.3 mm (0.21 in.)
± 0.8 mm (± 0.03 in.) 7.1 mm (0.28 in.)
4
5 8.9 mm (0.35 in.)

f_
22 mm (0.875 in.) 6 10.7 mm (0.42 in.)
7 12.4 mm (0.49 in.)

r 8
9
14.2 mm
16.0 mm
(0.56 in.)
(0.63 in.)
125mm 10 17.8 mm (0.70 in.)
11 19.6 mm (0.77 in.)

] 1.75 mm 10.07 in.)

± 0.125 mm 10.005 In.)
HOLES TYPICAL
12 21.3 mm (0.84 in.)

0.1 mm (0.004 in.) thicknesses. Artificial discontinuities

are 0.008 mm (300 /-Lin.) \vide at depths of 15, 30 and
60 percent of the material thickness. These three
3.3.1 Configuration of Artificial Discontinuities and depths represent three sensitivity levels. Simulated
their Designations discontinuities are machined air gaps filled \vith inert
nonmagnetic material.
The configuration of the artificial discontinuities Type C (circular) standards can be used to ascertain
(shim standards) is shown in Fig. 3. The artificial field direction and ensure complete coverage on two
discontinuities are labeled. The designation gives the or more magnetizing steps or for developing multidi-
thickness, discontinuity type and percent depth of rectional tests. Type B (bar) standards are used when
discontinuity. For example B215 is a bar type discon- discontinuities in a particular direction are most criti-
tinuity 2 milli-inches (50 /-Lm) thick \vith a discontinu- cal. Type R (radial) standards are most useful for test
ity depth at 15 percent of the bar's thickness. objects \vith narrow spaces and small radii.
The material is rolled, low carbon, high permeabil- These artificial discontinuities should be discarded
ity, annealed steel \vith 0.05 mm (0.002 in.) and if excessively worn or misshaped.
430 I MAGNETIC PARTICLE TESTING

TABLE 5. Typical dimensions for artificial

FIGURE 3. Artificial discontinuity standards: discontinuity standards (see Figure 3)
(a) bar or type B standard verifies fields for
discontinuities in particular directions; (b) circular Thickness Discontinuity Depth
or type C standard for determining field Designation micrometers (inches) (percent)
direction; and (c) radial or type R standard for 8215 or (215 50 (0.002) 15
small test objects (see Table 5) 8230 or (230 50 (0.002) 30
8260 or (260 50 (0.002) 60

8415 or (415 100 (0.004) 15

8430 or (430 100 (0.004) 30
la, 8460 or (460 100 (0.004) 60

~ R230 50 (0.002) 30
6mm
(0.25 in.)
The artificial discontinuity should be firmly attached
to the test object with its back surface in intimate
contact with the object's surface in locations where
discontinuities are being sought. The true continuous
magnetizing method must be used: applications of
Ib,

I0 I
magnetizing force continue after application of paIti-
cles ceases. In dry powder tests, blowing ofT the excess
05 powder must occur while magnetizing current is flow-
13 mm 1: in) : [ 0 7 5 in)
ing. DIy or wet particles may be used.
These standards should be used in conjunction with
a tangential field meter. The magnetic field level shall
be translated to current densities and shall be included
in the wlitten procedure.
Ie) r 50mm(2 in.) I 3.3.3 Care of Standards
Following use, artificial discontinuity standards shall
lOmm 10EI :: : : : : : : : I be thoroughly cleaned and checked under ultraviolet
or visible light, as appropriate to the testing process, to
--.II..-
5 mm (0.2 in.)
ensure that residual indications do not remain.

3.4 Qualification and Certification of Testing Personnel

All personnel performing magnetic particle tests
shall be qualified and certified in accordance with
MIL-STD-41O. Personnel making accept/reject decisions
in accordance with the process described by this standard
shall be qualified to at least Level II in accordance with
3.3.2 Use of Artificial Discontinuities
MIL-STD-41O.
Personnel performing processes described in this spec-
Altificial discontinuity standards should be placed in
ification other than accept/reject decision making shall
an area of high stress concentration and in changes of
be qualified to at least Level I in accordance with MIL-
section to determine levels of current density and
STD-410.
directions of magnetizing fields needed to produce test
indications. 3.5 Acceptance Criteria
When used for attaching artificial standards, tape
should have the properties of good adhesion to steel; Acceptance criteria for test objects shall be incorporat-
nonfluorescence; imperviousness to oil or solvents; ed as part of the written procedure whether specifically
and shall not cover the surface of the standard above or by reference to other documents containing the
the discontinuities. Standards shall be used with care necessary information. Methods for establishing :lCcep-
to avoid cold working, bending or climping. tance requirements for large crankshaft forgings are
Itllllllllllllllllllllllllllllllllllllllllllllllll'lll1I11111111111111111111111111111111111111I11

CODES, STANDARDS AND SPECIFICATIONS I 431

discussed in ASTM A456. Methods for establish- 3.7 Record of Testing

ing requirements for steel forgings are discussed in
th
ASTM A275. Methods for classif)ring castings are given in The results of all magnetic particle tests shall be
MIL-STD-2175. MIL-M-47230 provides a classification recorded. Records shall be identified, filed and made
scheme for ferromagnetic forgings, castings, extrusions available to the procuring agency on request. Records
and weldments. The testing of aircraft steel for cleanli- shall provide traceability to the specific test object or the
ness using magnetic particle testing is discussed in lot inspected and shall identify the testing facility and the
AMS 2~OO, AMS 230101' AMS 2303 as appropriate to the procedure used in the testing.
type of steel being tested.

3.6 Written Procedure 3.8 Marking Test Objects

Magnetic particle testing shall be performed in accor- Test obj~cts that have passed magnetic particle testing
dance with a written procedure applicable to the test shall be marked in accordance with the applicable draw-
objects. The procedure shall be in accordance with the ing, purchase order, contract or as specified herein.
requirements and gUidelines of this specification. The Marking shall be applied in such a manner and location as
procedure shall be capable of detecting the smallest to be harmless to the test object. The identification shall
rejectable discontinuities specified in the acceptance not be obliterated or smeared by subsequent handling
requirements. and when practical shall be placed in a location visible
The written procedure may be a general one which after assembly. When subsequent processing removes the
clearly applies to the specified objects being tested. All identification, the applicable marking shall be affixed to
written procedures shall be approved by an individual the record accompanying the finished assembly.
qualified and certified to MIL-STD-410 Level III for
magnetic particle testing, and .shall be subjeet to the
approval of the procuring agency. The written procedure 3.8.1 Impression Stamping
shall include at least the following elements, either
directly or by reference to the applicable documents. Impression stamping or vibroengraving may be used
when permitted or required by written procedure,
A. Procedure identification number and date it was detail specification or drawing and shall be located
written. o!J.ly in the area provided adjacent to the part number
B. Identification of the test objects to which the proce- or inspector's stamp.
dure applies, including material and alloy.
e. Areas of the test object to be examined.
D. Directions of magnetization to be used; the order in ~.8.2 Etching
which they are applied; and any demagnetization
procedures used between shots. Test objects may be etched using etching fluid or
E. Method of establishing magnetization (prods, yoke, other means provided that the etching process and
cable wrap, multidirectional system). location do not adversely affect the object's function.
F. The type and level of magnetizing current (alternat-
ing current, half-wave direct current, full-wave direct
current) and the equipment used. 3.8.3 Dyeing
C. The current level or the number of ampere-turns
used and the duration of its application. Identification may be accomplished by the applica-
H. Test object preparation before testing. tion of dye, if the test object surface finish and
I. Type of magnetic particle material (dry or wet, visible subsequent handling permit.
or fluorescent) and the method and equipment used
for its application.
J. Type of records and method of marking objects after 3.8.4 Other Identification Methods
testing.
K. Acceptance criteria and disposition of objects after Other means of identification such as tagging may
testing.
be used for objects that have construction or function
L. Post-test demagnetization and cleaning requirements. precluding the use of stamping, vibroengraving or
M. Sequence of magnetic particle testing as related to etching, as in the case of completely ground or
manufacturing process operations. polished balls, rollers, pins or bushings.
432 / MAGNETIC PARTICLE TESTING

3.9 Test Materials

or DOD-F-8793.5. The viscosity of the suspension shall
3.9.1 Dry Particles not exceed 5 mm 2 -s - I (,5 centistokes) at the temper-
ature of use when tested in accordance with ASTM
Dry particles shall meet the requirements of AMS D445.
3040. AMS 3040 particles shall be capable of produc- The flash point of the suspension should be a
ing indications (listed in Table 3) on the test ring minimum of 93°C (200 OF) when tested in accordance
specimen of Fig. 2 using the follOwing procedure. with ASTM D93.
Place a conductor with a diameter 25 to 32 mm (I to The background fluorescent of the suspension
1.25 in.) and a length greater than 400 mm (16 in.) vehicle shall be less than the limit specified in
through the center of the ring and circularly magnetize DOD-F-87935.
the ring by passing the current specified in Table 3
through the conductor; 3.9.5 DetemLination of 'Vet Particle Concentration
Apply the particle powder to the ring using the and Contamination
continuous method. Examine the ring within one
minute after current application. Examination of Agitate the particle suspension a ffilIllmum of 30
nonfluorescent baths shall be under a visible light of minutes in order to ensure uniform distribution in the
not less than 2,100 Ix (200 ftc). Examination of bath. Place a 100 mL sample of the agitated suspen-
fluorescent baths shall be under ultraviolet light with a sion in a parachute shaped centrifuge tube (specified
minimum intensity of 10 JLW-mm -2 (1,000 JL\V-cm -2) in ASTM D96). Demagnetize and allow the tube to
at the surface. The minimum number of hole indica- stand undisturbed for at least 60 minutes. Read the
tions shall meet or exceed those specified in Table 3. volume of settled particles.
Hole indications exceeding 4.75 mm (0.2 in.) in width If a fluorescent suspension is used, the liquid above
shall not be counted. the precipitate is examined with ultraviolet light. The
liquid shall be nonfluorescent. Comparison may be
3.9.2 .Wet Particles made with fresh vehicle. If the liquid suspension with
the particle settled out is seriously contaminated with
Wet pqrticles shall .meet fhe requirements. of cutting oil or with fluorescerit pigment froin the
AMS 3041, AMS 3042, AMS 3043, AMS 3044, particles, the suspension will fluoresce brightly and
AMS 3045 or AMS 3046, as applicable. In applying shall be discarded.
these specifications, the 'particle shall show indications Particle concentration and contamination shall be
as listed in Table 3 on the test ring specimen of Fig. 2, determined at start' up; whenever the bath is changed
using the procedure described in 3.9.1. or adjusted; and at regular inteIVals according to
The concentration of patticles in the test bath shall Table 2.
be in the range of 0.1 to 0.5 mL in a 100 mL bath
sample for fluorescent particles and 1.2 to 2.4 mL for 3.9.6 Caution
nonfluorescent particles. Fluorescent particles and
nonfluorescent particles shall not be used together. Dry cleaning solvents are not to be used as the
vehicle for wet particles. Proper precautions must be
3.9.3 'Vater Vehicles taken to prevent the ignition of hydrocarbon suspen-
sion baths by overhe~lting or eleetrical arcing. Precau-
Water used as a suspension vehicle shall be sUitably tions shall also be taken to prevent inhaling of dl)'
conditioned to provide proper wetting, patticle disper- particle materials.
sion and corrosion protection. Proper wetting shall be
determined by a water break test. Whenever possible,
4.0 Procedures
a nonionic wetting agent should be used. However, the
amount of wetting agent should be limited so as not to 4.1 Testing Sequence and Coverage
raise the alkalinity of the suspension above 10 pH.
Use of conditioned water vehicle on cadmium plat- Magnetic pmticle testing shall be performed after the
ed steels "vith tensile strength of 1,240 MPa (180 ksi) completion of operations that could cause surface or near
or above is prohibited. surface discontinuities such as forging, heat treating,
plating, cold forming, welding, grinding, straightening,
3.9.4 Oil Vehicles machining and proof loading.
Magnetic particle testing shall not be pcrformed "vith
The suspension vehicle for the wet method shall be nonferromagnetic or ferromagnetic coatings in place that
a light petroleum distillate conforming to AMS 2641 could prevcnt the detection of surface discontinuities in a
 CODES, STANDARDS AND SPECIFICATIONS I 433

on shall ferromagnetic substrate. Such coatings include paint or (0.4 in.). For the same flux, magnetization at the
emper- chrome plate greater than 0.075 mm (0.003 in.) in thick- surface using alternating current yokes is higher than
ASTM ness or ferromagnetic coatings such as electroplated that for direct current magnetization.
nickel greater than 0.025 mm (0.001 in.) in thickness. When alternating current flows through conduc-
he a When such coatings are nonconductive, they must be tors of circular cross section, the fIeld strength at the
-dance removed where electlical contact is to be made. Unless surface is exactly the same as for a direct cur-
otherwise specified, production parts shall be magnetic rent flow of the same amperage. For conductors
'nsion particle tested plior to application of the coating. When having noncircular cross sections (pmiicularly note
~d in plating operations could result in hydrogen emblittlement conductors with sharp edges), the field strength at
cracking, testing will be performed before and after the the surface becomes more uniform with higher
plating operation. The post plating magnetic particle test frequency.
l{ion may be substituted with a penetrant test without a C. 'When electIical contact is made with the test object
pickling operation.
using prods, clamps, magnetic leeches or other means,
c·30 precaution shall be taken to ensure that the electrical
A. When it is not possible to test the whole test surface current is not flowing while contacts are applied or
the in one operation, divide the test surface into suitable removed and that excessive heating does not occur in
en- sections and perform separate tests on each section, the contact area. Verify that contact surfaces are
led clean ..
with a ten percent overlap at the section boundaries.
to
B. \"'here the direction of discontinuities cannot be
he presumed or detection of discontinuities in various 4.2 Preparation of Test Objects
directions is required, magnetic fields in at least two
ve
directions shall be applied to the test object. The 4.2.1 Pretesting Demagnetization
Ie
directions shall vary 90 degrees from one another and
the fields may be imposed successively or simulta- The test object shaH be 'demagnetized before testing
neously in a multidirectional system. . if prior operations or conditions have produced a
C. When the residual method is used, other ferromag- residual magnetic fIeld which will interfere with the
netic material shall not be brought in contact with the test.
test surface, fro:r:n the finish of the magnetizing
operation to the completion of magnetic particle 4.2.2 Precleaning
pattern inspection. .
D. An object carrying a direct current will conduct the For tests performed with dry powders, the cleaning
current throughout the whole of its cross section. solvents or cleaning process must be capable of pro-
Alternating current flows predominantly along the viding a dry surface free of oil, scale or grease. For
surface because of the skin effect. Regard full-wave tests performed with wet particle suspensions, the
rectified alternating current as direct current because cleaning process must be capable of providing a test
of its high direct current proportion (63 percent). surface that passes a water break test.
Half-wave rectified direct current and the other
current types may be regarded as a superimposition A. The area of pretreatment shall be wider than that
of a direct current proportion with an alternating under test. In the case of a weld, the pretreatment
current proportion.
area shall be enlarged about 20 mm (0.75 in.) from
E. With direct current flow, the flux denSity is inversely the area under test toward the parent material.
proportional to the cross sectional area of the test B. When dry magnetic particles are used or when a
object. In the case of alternating current, the flux wet pmiicle suspension different from the cleaning
denSity at the surface is inversely proportional to the liquid is used, the test surface shall have been
circumference of the test object. The attenuation of dried.
the flux denSity with increasing circumference or C. To prevent burning and to improve current trans-
cross section is less for an alternating current field mitting efficiency, the surface of the test object
than for a direct current field. and the electrodes that contact one another shall
F. \"'hen using direct current yokes, magnetization at be clean.
the surface is dependent on the thickness of the
material.
4.2.3 Water Break Test
\Vllen magnetizing 'with alternating current yokes,
the field strength at the surface is virtually indepen- An object to be tested is flooded with conditioned
dent of plate thickness for thicknesses above 10 mm water and the appearance of the surface is noted after
434 I MAGNETIC PARTICLE TESTING

flooding is stopped. If a continuous even film forms 4.3.2 Head Shot

over the entire object surface, a suspension has suffi-
cient wetting ability and the surface is sufficiently \Vhen magnetizing by passing current directly
clean for magnetic palticle testing. If the film of through the test object, the current shall be from 12 to
suspension breaks, exposing bare surface, insufficient 32 A per millimeter (300 to 800 A per inch) of test
wetting agent is present or the test object has not been object diameter. The diameter shall be taken as the
adequately cleaned. largest distance between any two points on the outside
circumference of the test object less any void areas
4.2.4 Plugging and Masking (holes or passageways) across the line between the
points. Avoid overheating and arcing. Use one of the
Small openings and oil holes leading to obscure devices listed in 4.3 to verify that sufficient field
passages or cavities shall be plugged with a suitable strength is obtained.
nonabrasive material which is readily removed and, in
the case of engine parts, is soluble in oil. Effective 4.3.3 Central Conductor
masking shall be used to protect components which
may be damaged by contact with the suspension. \Vhen magnetizing by passing current through a
conductor inside a hollow test object, alternating
4.3 Magnetization current is to be used only for smface discontinuities on
the inside surface of the test object. If only the inside
The current levels given in the following subparagraphs of the object is to be tested, the diameter shall be the
are to be used as a guide in establishing the test largest distance between two points, 180 degrees apmt
procedure. Actual current levels must be established by on the inside circumference.
verification of adequate field strength by using one or \-Vhen the axis of the central conductor is located
. more of the following: near the central axis of the test object, the same
current levels established for a head shot shall apply,
A. a tangential field meter; but ensure that sufficient magnetic field strength is
B. an artificial discontinuity standard; . obtained by using one of the devices listed in 4.3.
C. a pie gage field indicator.

A field strength value measured at one point on a test 4.3.3.1 Offset Central Conductor
object is not indicative of the field strength els~where,
\-Vhen the central conductor is placed against an
particularly if the test object has a complex shape.
inside wall of the test object, head shot current
Therefore, field strength shall be measured at several
levels shall apply except that the diameter shall be
locations on the test object and particularly in comers
considered the sum of the diameter of the central
and grooves. . conductor and twice the wall thickness.
When the tangential field meter is used, the field
The distance along the test object circumference
strength should be in the range of 3 to 6 mT (30 to 60 G).
(interior or exterior) which is effectively magnetized
If an artificial discontinuity standard or pie gage field
shall be taken as four times the diameter of the
indicator is used, the artificial discontinuities must be
central conductor (see Fig. 4). The entire circum-
detectable in all tested areas at anyone shot. It is typical
ference shall be tested by rotating the test object on
to use the minimum cnrrent required to prodnce 3 to
the conductor, allowing for about 10 percent mag-
6 mT (30 to 60 G) readings or to produce indications at
netic Held overlap between shots. Use one of the
the adificial discontinuities to avoid excess magnetization
devices listed in 4.3 to verify that sufficient field
or overheating.
strength is obtained.

4.3.1 prod
4.3.4 Coil Shots
When using prods, 3.5 to 5 A per millimeter (90 to
125 A per inch) of prod spacing shall be used. Prod Passing current through a coil encircling the test
spacing shall not be less than 50 mm (2 in.) nor greater object or encircling a section of the test object pro-
than 200 mm (8 in.). Verify that contact surfaces are duces a magnetic field parallel to the axis of the coil.
clean and in contact before current application. Also, This magnetizing method must be used when an
use one of the devices listed in 4.3 to verify that object's width or diameter is Significantly greater than
sufficient field strength is obtained. its length (bearing races, disks, rotating pads).
"111111111111111111111111111111111111111111111""111III1IIIIIII11111111111111111111111111111111

CODES, STANDARDS AND SPECIFICATIONS / 435

For low fill-factor coils, the effective field extends a

distance on either side of the coil center approximately FIGURE 4. Magnetization with an offset central
equal to the radius of the coil (see Fig. 5). For cable conductor; region effectively magnetized for
wrap or high flll-f~lCtor coils, the effective distance of particle testing is equal to four times the
magnetization is 225 mm (9 in.) on either side of the diameter of the conductor
coil center (see Fig. 6).
For test objects longer than these effective distanc-
EFFECTIVELY
es, the entire length shall be tested by repositioning MAGNETIZED
REG
the object within the coil, allovving for about 10 per-
cent effective magnetic field overlap. For coil magne-
tization, it is essential that artificial discontinuity stan-
. danis are placed to determine the extent of the
effective field on either side of the coils when prepar-
ing technique documentation.

4.3.4.1 Effect of Coil Size

When the cross sectional area of the coil is ten or

more times the cross sectional area of the test
object, then in amperes through the coil, the cur-
rent I shall be as follows.

A. F~r test object~ positioned t.o the side of the

co&
FIGURE 5. Magnetization with a low fill-factor
coil; region effectively magnetized for particle
_4,s-...:..,0_0_0_D testing is equal to two times the radius of the
I =. NL (Eq. 1)
coil
Where:

L = length of the test object;

D = diameter of the test object (measured in
the same units as the length); and
N = number of coil turns.

B. For test objects positioned in the center of the

coil:
11
1= R43,000
(Eq.2)
N(~) -05 FIGURE 6. Magnetization with a high fill-factor
Where: coil

R = the radius of the coil.

If the cross sectional area of the coil is less than

twice the cross sectional area (including hollow
portions) of the test object:

1= 305,000
(Eq. 3)
N(~) + 2
436 / MAGNETIC PARTICLE TESTING

These equations hold only if the LID ratio is B. Superimposition of Alternating Field on a Con-
greater than 3 and less than 15. If LID is less than stant Field
3, ferromagnetic material with the same diameter as The resulting field direction oscillates around tlw
the test object shall be placed on each end of the direction of the uniform field. If both fields arc
test object to effectively increase LID to 3 or equal in size, then the angle between the fields will
greater. If LID is greater than 15, use 15. approach 90 degrees. If this range is to be en-
For a hollow or cylindrical test object, D shall be larged, set the alternating field to a higher strength
replaced with an effective diameter Dejf calculated than the constant field.
using:
When two alternating fields are superimposed with
(Eq.4) the same or opposite phases, the two altemating
fields interact to generate one field having a constant
Where: direction.

At = the cross sectional area of the test object; 4.3.5.1 Multidirectional Magnetization Current
and
Ali the' cross sectional area of the hollow por- Alternating current, half-wave direct current and
tions of the test object. full-wave direct current may all be used separately
or in combination depending on the testing appli-
Effective fields diminish according to the inverse cation and the restrictions regarding coating thick-
square law from one pole to the other. Artificial ness. The continuous method of particle application
discontinuity standards shall be used to ensure must always be used when emplOying multidirec-
proper cu~rent levels along the entire axis of the test tional magnetization techniques.
object. Use alternating current for magnetization of
complex shapes. 4.3.6 Yokes

~.3.5 Multidirectional Systems Yokes meeting the re.quirements o( 3.l.4 may. he

used, prOvided discontinuity standards are used to
With suitable circuitry, a multidirectional field in a verify field strengths.
test object may be established. For example, this may
be done by selectively switching· the magnetizing 4.4 Particle Application
current between contact pairs positioned about
90 degrees apart. The electrical phase of the fields 4.4.1 Continuous Method
must be different so that the resulting field within the
object changes its direction periodically. The magnetic In the dry continuous method, magnetic particles
fields must be attuned to one another with respect to are applied to the test object by blOwing, brushing or
their size, angle and phases. To illustrate such com- dusting while the magnetizing force is present. In the
bined methods, two examples are given below. wet continuous method, the magnetizing force is
present or the current is on while the suspension is
A. Superimposition (~f Two Altemating Fields with (J heing applied to the test object.
Phase Shift
By shifting the phase of the magnetization, a field 4.4.2 Residual Method
is produced of changing size and changing spatial
characteristics. Such a field can be produced using In the residual magnetization method, magnetic
three-phase current with a phase difference of 120 particles are applied to the test object immediately
degrees. after the magnetizing force has been discontinued.
Phase shifts between 50 degrees and 130 de- The residual method shall be used only when specitl-
grees can be used. Phase shifts exceeding this cally approved by the procUling agency or as all
range produce snch great changes in field strength interpretation aid.
that the detection of discontinuities in random
direction is reduced. With a phase shift of 90 4.4.3 Dry Method
degrees and fields of the same size peq)Cndicular
to one another, the field strength will be equal in VVhen using (hy pmticles, the flovv of magnetizing
all directions. current shall be initiated before the applicatioll of
 CODES, STANDARDS AND SPECIFICATIONS / 437

magnetic particles to the test surface and shall be 4.4.6 MagnetiC Polymer
terminated after powder application has been com-
pleted and any excess blown off. Polymerized material (low temperature vulcanizing
The duration of the magnetizing current shall be at rubber) containing magnetic particles shall be held in
least 0.5 second and short enough to prevent damage contact with the test object dming the period of its
to the test object from overheating or other causes. cure. Before curing takes place and while the magnetic
DIy powder shall be applied such that a light, uniform pmticles are still mobile, the test object shall be
particle dust settles on the surface of the test object magnetized to the specified level. This requires pro-
while it is being magnetized. Specially deSigned pow- longed or repeated peliods of magnetization. This
der blowers or shakers using compressed air or hand method is for special applications such as bolt holes
power sha11 be used. The applicators shall introduce that cannot be readily tested by the wet or dly method.
the particles into the air in a manner such that they The method shall be used only when specifically
reach the test object smface in a uniform cloud with a approved by the pr.ocllling agency.
minimum of force.
Excess powder shall be removed by means of a dry 4.5 Interpretation and Recording of Indications
air current with sufficient force to remove excess
palticles but not enough force to disturb particles held Operators must wait at least 60 seconds after entering
by leakage fields indicative of discontinuities. Observe a darkened testing booth before examining particle
carefully during powder proper application and during indications.
removal of the excess powder. Sufficient time for
formation and examination of indications shall be 4.5.1 Recording of Indications
allowed during the testing process.
All indications shall be evaluated according to the
applicable acceptanee criteria specified in the written
4.4.4 Wet Method procedure. Relevant indications which have been eval-
uated a!ld determi'ned to be unacceptable shall be
Fluorescent or nonfluorescent particles suspended described in detail on the appropriate rejection form
in a liqUid vehicle at the required concentration shall and submitted for disposition. The location of all
be applied either by gently spraying or flOwing the indications shall be' marked on the test object and
susp~nsion over the area to be tested. Proper sequenc- permanent records of the location, direction and
ing and timing of magnetization and proper applica- frequency of indications may be made by one or more
tion of particle suspension are required to obtain of the follOwing methods.
formation and retention of indications. This generally
requires that the stream of suspension be applied to 4.5.2 Recording Sketch
the test object slightly before and simultaneously with
energizing the magnetic circuit. Record the location, length, direction and number
The magnetizing current shall be applied for a of indications by a sketch or on a tabular form.
duration of at least 0.5 second for each application
with a minimum of two shots being used. Care shall be 4.5.3 Tape Transfer
exercised to prevent damage to the test object due to
overheating or other causes. Weakly held indications Apply special transparent pressure sensitive tape to
on highly finished parts are readily washed away and the dry particle indication and then peel it off the test
care must be exercised to prevent high velocity of the object. Place the tape transfer on an approved form
bath flow over critical surfaces. with information giving its former location on the test
object.

4.4.5 Magnetic Slun'y or Paint 4.5.4 Indication Fixing

Magnetic paints or slurries are applied to the test Cover the indication with commercially available
object with a brush, squeeze bottle or aerosol can strippable film to fix the indication in place. Strip off
before or during the magnetization operation. This the film and indication after cure (use transparent
method is for special applications such as overhead or pressure sensitive backing, if included with the fixing
underwater magnetic particle tests. This method shall product). Place the indication on an approved form
be used only \\Then specifically approved by the pro- with information giving its previous location on the test
curing agency. object.
438/ MAGNETIC PARTICLE TESTING

4.5.5 Photography 4.6.1.3 Yoke

Photograph the magnetic particle indications, the
\Vhen demagnetizing with a yoke, use alternating
tape transfer or the strippable fixed indication. File the
current and energize the yoke when it is in contact
photograph with information giving its location on the
with the test object. ·With current still applied,
test object, if such information is not visible in the
separate the test object from the yoke for a distance
photograph (grid lines may be used).
of 1 m (3 ft) while rotating the test object or yokE'
relative to one another.
4.6 Post-Test Demagnetization and Cleaning
4.6.1.4 Field Strength Levels
Unless directedotheIWise by the procuring agency, all
magnetic particle test objects shall be demagnetized,
After demagnetization, a magnetic field strength
cleaned and corrosion protected after testing.
meter shall not detect fields greater than ± 0.3 mT
( ± 3 G) anywhere on the test object.
4.6.1 Demagnetization
4.6.2 Cleaning and Care after Testing
4.6.1.1 Alternating Current
Remove all traces of magnetic particles and vehicle
When using alternating current demagnetization, from the test object. Cleaning shall be done with a
the test object shall be subjected to a field with a suitable solvent, air blower or by other means. All
peak value equal to and in nearly the same direction traces of the magnetic particles and magnetic particle
as the field used during testing. This alternating test materials which might interfere with subsequent
current field is then gradually decreased to zero. use of the test object shall be removed. Chlorinated
When using an alternating current demagnetizing solvents shall not be used on test objects containing
coil, hold the. test object about 0.3 m (1 ft) in front crevices or on 400 series stainless steels.
of the coil and move it at least 1 m (3 ft) beyond the Thoroughly remove all plugs in cavities and all
enq of the cOiJ. ~epeat this process if ne,cessary. masking. Test objects shall b~ protected from possible
Rotate and tumble test objects of complex configu- corrosion or dam~ge during the cleaning process and
. ration while passing them through the field of the shall be treated to prevent the occurrence of corrosion
. coil. before continued· processing.

4.6.1.2 Direct Current

5.0 Definitions
When using direct current demagnetization, the If acceptance criteria are included in a specification, they
initial field shall be equal to and in nearly the same
are listed in Section 5.0 and the definitions move to Sec-
direction as the field reached during testing. The tion 6.0.
field shall then be reversed, decreased in magni-
The speCification may reference other documents that
tude, and the process repeated (cycled) until an
contain appropriate definitions for particular terms. Alter-
acceptably low value of r.esidual magnetic field is nately, the specification may provide its own alphabetical
achieved. Typically, a thirty step cycle shall be used. glossary of words and their definitions.
4401 INDEX

INDEX

A
Absolute electric system . . . . . . .. . . . . . . . . . . . . .. . . . . . .. . . . .. . . . . . . . . . . .. .. ... 10.'5 American Welding Society..... . .... 69,388, -t!.5
Absolute magnetic system .......... " .............. , . . . . . . . . . . . . . . . . . . . . .. lOS Stmctllfal Welding Code...... 40~
Acceptance criteria specifications .................................. 430-431 Ammeter...................................... .. ...................... 42
Acceptance standard. See also Artificial discontilluity; Flaw accuracy specifications ... . . . . . .. .. ..... .... . 427 -42<~
standard, artificial.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 42 and automated test systems .. . . . . . . . . . . . . . . . . . . . . 25:3
ACGIH TLV. See also Threshold limit value .............................. 316 Ampere's equation .. :.... ................ .. ................... 104
Actinic ultraviolet. See also UV-C ..........................................320 Ampere's law....... 104,106-107,112-113,117,1.5.5,166. 409
Active leakage field .......................... " .............. " .. " . . .. . . . .. 122 Ampere meter kilogram second. .. .. .. . .. . .. . .. .. . .. . . .. . . . .. . .. .. .. .. . 10.')
Adjacent cell linking ................................................. " 26.3-264 Ampere-tunt .................... .. .. . 42. 117
Agglomerated settling ....................................... " ............. 211 Analytical modeling .............................................. 127. 129-l:30
Alginate impression record.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 276 compared with numerical modeling.... . .. . .. . . ... . .. .. .. .. ... 131
Alloy ............................................................................ 2.34 AN battery... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ........................ . . " :54
Alnico ............................................................................ 13 AN bolt....... ....... .... ...... ....... ... ......... .. .... .... .... .... .5,
Altentating current ................................................. 27, 42, 1.54 Anisotropy................................ .............. 106
advantages of . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 27 ANS (rectifier) series.................. ............................ :5,
and high retentivity... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 2S.5 Arc blow. . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
and hysteresis curve ......................................................... 24 and demagnetization.. . .. .. .. .. .. .. .. . .. . . .. . .. .. .. . . .. . .. . . . .. . . :30
and magnetodiodes ......................................................... 190 Arc bunt .................... , ............... '" ..... ....... .... .... ......... 42
and mobility ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 37 Arc deflection................................................................. :30
Arc strike .. .. .. .. .. .. .. .. .. . .. .. .. . .. .. . .. .. .. . • .. .. .. . .. .. .. .. . .. . . .. . .. .. . . 4:2
:~~ :~~~~~1~~;1:e~~:~:~::::::::::::::::::::::::::::: :.::::: :::::::: :::::::: ;~~ . Arc welding. . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
compared with direct current ......... .'.............................. 1.55.156 ~~~~;a.l. ~~.~~i.~ ............................................................................................ ~:. ~~~.~~: .2~~: ;j~~
early use of ................................................................... Sl
anp safety ............................................... , . . . . . . . . . . . . . . :31·~
equipment, compared with direct curr.ent equipment. .................. 353
field definition ..................... : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 42 Articulated pole piece .. ..... .. .. .. .. . .. .. .. .. .. .. .. .. .. .. .. .. .. .. . .. .. .. 42
for circular magnetization ................................................... 151 Artificial discontinuity ................................ " .. .. .. . .. .. .. . . .'. 41 •
for field flow magnetization ......................... : ..................... 15:3 and prism block standard. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400-401
for multidirectional magnetization ..................... , . . . . . . . . . . . .. 162-164" and reference standard.... . . . . . . . . . . . . . . . . . . . . . .. ......................... :340
generators ................................................................... 358 specifications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 429-4:30
in wet method testing ....................................................... 21 standards. See also Acceptance standard;
mobile test systems ............................................... 22, 357-358 Reference standard; Simulated discontinuity .... 34.5-:]·Hi
portable test systems .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 360 Austenite. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . \:) 1
skin effect.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 32 Autogenous weld...................... .. . . .. .. .. .. .. .. .. kS
swinging field systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 248 Automated handling.. .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2..10
to field test welds ......................................................... " 402 Automated test system ... .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. :3.5.')
Altentating current yoke ........................................ 162-163,383 computer components ............................................ '" 248-24~)
Altentating current coil design ...... '" ...... " .................................................. 246-247
demagnetization of tubular product ...................................... 309 Automated test system malfunction. See Bath monitor; Current
demagnetizers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .................. " 306 level rrwnitor; Magnetizing current monitor; Malfunction alarms
testing tubular product. . . .. ................... 406 AutOmated visible particle scanning .....
Altentating current demagnetization :31-32,297-298.304 Automatic powder blower ..
equipment. .. .. . .... .. .. .. . .. .. .. .. . . . .. .. . . . .. .. .. '" 306 Automatic scanning ... 250,25 ..1
limitations of.... .. .. . .. .. .. .. .. .. .. .. .. .. ..... 32, 300 Automatic verification ... . 2.5-1
specifications. .. .. .. .. .. . . . .. .. .. .. . . . . .............. ' .......... 438 Automation, partial. .... .
Altentating current magnetization .................................. 42, 1.55 Automotive industry applications ..... . :3fi:2
Average leakage field .... . p.)
by wet horizontal systems.... .. .. . .. .. .. .. .. .. .. .. . .. .. .. . .. . . .. .. .. . .. . ... 35:3
Ambient light ............................................................... " 368 Axial flux density ................. .. :310
American Association of Railroads. . .. .. . . .. . .. . .. . . . .. .. .. . .. . . .. .. .. .... 59
American Bureau of Shipping. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 383
American Conference of Goventmental B
Industrial Hygienists.................. . .... :316. 322, 324, 371 Background 42
American Institute of Steel Construction.. . ........................... 69 Background fluorescence. Sec also Sigll(J!-lo-rwisc ratio ... -!OJ
American National Standard Institute ..... 207,314,324,328, .332, 425 Bar magnet ............ . 15, 14b-119
American Petroleum Institute ..................... 165, 406, 408, 419, 425 Bar standard ....... . UH
American Society for Nondestructive Testing ................... 69, 231 Bath. See also SlIspension; Vchicle
American Society for Testing and Materials ......................... " 425 and automated testing ..
American Society of Mechanical Engineers ........... 56, 331, 383, 425 application. " ..
Hoiler and Pressure Vessel Code.................. 331, 402, 425 concentration ........ .
 INDEX / 441

contamination ......................................................... 208, 236 Capacitor discharge. See also Flash maglletization ..................... 159
coverage ..................................................................... " 2,56 Capacitor discharge coil field testing ............................. , 406-407
maintenance ....................................... " .................. , 209-210 Capacitor discharge magnetization ........................... 406-407,409
monitor ..... " .................................................... 249, 255-256 Capacitor discharge method ............................................... 42
preparation .... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 209 Capacitor discharge pulse system ....................................... 170
settling range ............... , ................................................ 210
Capacitor discharge test system .............................. 355, .358, 361
settling test ................ '" ......... " .................................... 236
Carbon steel. . . . . .. .. . . . . . . . .. . . . . . . . . . . . . . . . . . . .. . . . . .. . . . .. . .. . . . . . .. . . . . .. . ... 8
:::~~i~~~;~eS=~oa;~ ,~,~~l~~. ~:.n~~~................................................................................ '4~: Carburized steel. ............................................................ 214
Bellcrank . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 394-396 Carrier fluid. See also Vehicle; Suspension .......... ....................... 42
,42.5 Carrier-to-interference ratio............................................... S9
402 Berthold field gage ................................................. 51,67,345
Berthold penetrameter ................. " ., ......... " . . . . . . . . . . . .. . . . . . . . .. 42 Casting ............................................ , ................... 7, 202, 363
42
Berthold referen('e standard......... ............... . ..... 417 Cellular automata ........... " ........................................ " .... 265
--121)
2.5:3 Bctz, Carl.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 50, 66 Centerline hot crack. See also Solidificatio1l cmck....................... 8,5
104 B-H curve Centimeter gram second ................................................. , 104
409 and flux density variations ................................................. 411 Centistoke .................................................................... , 239
10.5 and inductance value ....................................................... 414 Central conductor. See also Internal cOllductor . .. , ................. 42, 279
117 and inductive ammeter............... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 17.5 Central conductor magnetization ...... " .......................... 409, 434
1:30 and pulse current magnetization ..................................... 168-169 Circuit theory ........................................................... 117-118
131 and testing tubular product ................................................ 410 Circular field demagnetization... . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . .. 366
.54 Bias field. See also Leakage field.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ... 128 Circular magnetic field ............................................ 15, 42, 150
.57 Billet ................ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 3, 267 from prods ............................................................. , . . . . .. 23
111\) Binary imaging. See also Patter1l recognition....................... 263, 26.5 residual ...................................................................... 294
.57 Biot-Savart law .......................... : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 108 Circular magnetization. See also Circumferential
292 Bit plane. : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 265 lIwgnetizatioll...... .. .. .. .. ......... .. .. .. .. .. .. .. .. .. .... 15, 18-19, 42, 1.50
30 Black light. See also Long wave ultraviolet; Ultraviolet and alltomated magnetic palticle testing .................. '" ............ 268
42 light; (JV-A ......................................... , .... " ........... 42, 320 and direction of magnetic field ..... , . ... . .. . . . . . . . . . .. . . . .. .. . . . . ... 158-1.59
:30 filter ........................................................................... 42 and pie gage ....................................... : .................... 241-242
-t2 humap eye response to ........ ; ....................... " .. : .......... , ..... 372 and wet horizontal systems ................................................ 3S2
:92 regulatory requirements .............................................. 322-323 demagnetization of....... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 33
DO Black magnetic particle .... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 3,5 early use of .................................................................50-S1
:3:3 Black oxid~ coating................................ . . . . . . . . . . . . . . . . . . .. 383-384 fundam'entals of ................•............................................ 1.58
14 Black oxide suspension.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 397 inducing ................................................................... 18-19
42 Black oxide water bath. . .. .. .. .. .. .. .. . .... .. .. .. .. .. .. .. .. .. .. . .. .. .. .. .... 396 of hollow test objects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 1S2
t2 Block standard ....... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 400, 401 of solid test objects .................................................... 1,50-1SI
J1 Bloom ............................................................................. 3 with direct contact. .. . . . . . .. . . . . . . . . . . . .. . . . . . .. .. . .. . . .. . . .. . . . . . . . .... . . .. 1.51
!O Blowhole. See also Porosity .... ........... : ....................... 4,77-78,89 with induced current ........... , .. .. .. .. .. . .. .. .. .. .. .... .. .. .. .. .. ........ lSI
10 Boiler and Pressure Vessel Code ................................... 402,42,5 with prods...... .. .. . .. . . .. . . . . . . . . . ... .. .. . . .. .. ..... . . .. . . ... .. . . . .. . . .. . .. lSI
Boiling point .................................................................. 317 Circular standard. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 429
() Borescope ............... " ................................... 390-393, 403, 408 Circumferential discontinuity ....................................... 396-397
1 Boundary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 342 Circumferential magnetic field. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 1SO
S Boundary conditions .............................................. 109-113, 195 Circumferential magnetization. See also Circular lIUlgnetization ..... 18
fi Bracketing .................................................................. " 288 and demagnetization of tubular product. . . . . . . . . . . . . . . . . . . . . . . . . . .. 307-309
Breakdown potential. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 326 and testing tubular product .......................................... 408-410
Bright light conditions ... " ................ '" .............. , ......... 230-231 in wet method testing ....................................................... 21
Brightness of tubular product ..................................................... 16.5-17.5
adaptation ................................................................... 231 Clamshell mark See also Fatigue crack. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 94
contrast. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 229 Cleaning
contrast ratio... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 229 and magnetism......................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 292
sensitivity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 230 and underwater testing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 382
Brilliance' test .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 239 post-test specifications ..................................................... 438
British Standards Institute ............... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 426 pretest specifications ...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 433
Bucking field..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 299 verification of.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 254
Bulk field indicator.......... .. .............................. 192-197 Closed magnetic field equation. .. .. .. .. .. .. . .. .. . . . . . .. .. .. . .. .. . .... ... 118
Burst ............................................................................. 83 Coatings ..................................................................... " 383
effect on weld field testing. . . . . . . . .. ................................ 402-403
reference standards for ............................................... 386-387
c removal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 388-389
Code of Federal Regulations ....... , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 172
Cable parameters ........ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 3,57-3,58 Coercive force............. .. ...................... 24, 34-3.5, 42, 119
Cable wrap demagnetization. . . ...................... 298 and demagnetization ............................................ , 30, 29S, 300
Calibration and nonrelevant indications ................................................ 234
of field indkator .. . ............ . 303-30.5 and yoke demagnetization . . .. . .. .. .. .. . . .. .. . . .. . .. .. .. .. .... ............ 299
of ultmviolet meters ... . 379 Coil demagnetization. See also Alternating curre1lt coil
specifications .......... . ............ 427 dellwgnetization ................................................... 365-366
Candela ............. '" ............................................ 230 Coil field criteria. . . . . . .. .............................................. 405-406
Capacitance .. .... 171,406 Coil magnetization ..................................................... 1S2-1S3
Capacitor values .... .. 413 specifications. . . . . . . . . . . . .. ....................... .................... 434-436
4421 INDEX

Coil magnetizer...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 306 Cunico.................................................................... I,';

Coil method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 42, 159 Cunife ............... ...................... ..................... I:)
Coil pickup method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 405 Cupping .................... , ............ , ... '" ................. , .............5. ,..,:2
Coil shot ...................................................................... " 42 Curie point....................................... ...................... 30.4:2
Coil technique. See Coil method Curie point heating... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1:2 I
Cold crack. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 84 and demagnetization ............................................ 293. 296. :311:;
Cold lap ................ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 396 Curl ..................................................................... 1:3:3-1:3·(
Cold shut .................... '" .......... , .. " .............................. 7, 77 Curl-curl equation. . . .. . . .. . . . . . . . .. . . . . . . . .. . . . .. .. . . . . . . . . . . . . . . . . . . IO\)
Current flow method. See also Direct contact TlUlgnetization;
Cold start. See Lack of fusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 87
Prod magnetization.................................................. ·4:,
Color contrast .............................................................. " 229
Current induction technique. See also Induced Ctlrrent technique;
Color sensitivity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 230 Toroidal field ............................................................ , . . t3
Combination current...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 21 Current level monitor .................................. '" ................. 24~1
Combined auxiliary magnetization ................................. 163-164 Current output ............................................................... 2A I
Combined current ..................................................... 161-162 Current parameters ............................................. , 3.57-3.'58. :360
Combined magnetization .................................................. 160 Current pulse ................................................ 412-413.414.41.'1
Comol ........................................................................... 13 Current reduction method .......... , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .366
Compact settling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 211 Cylindrical discontinuity .......... " ...................... , . . . . . .. . . 342
Compliance. . . .. . . . .. . . . . .. . . . . .. . . . . . . .. . . . .. . .. . .. . . . . . .. . . .. .. . . . .. . .. . . ... 333
Components. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 353-354
Composition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 202
Concentratiml control............... . . . .. . . .. . . . . . . . . . .. . . .. . . . . .. . .. . . . . .. 236
o
Concentration specifications ............................................ " 432 Dark adaptation ............. , .......................................... 43. 3fiH
Concurrent leakage field. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 387 Dark adaptation gap ....................................................... :371
Conditioning agent ................................ , . . . . . . . . . . . . . . . . . . . . . . . . .. 42 Daylight fluorescent particle..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 7:3
Conductivity .................................................................. 110 Dead weight specifications............... . . . . . . . . . . . . . . . . .... . . . . . . . . . . . . 42H
Cone ...................................................................... 231,369 Decay current. .. ... . . . .. . . . .. . . .. . .. . . . .. . .. . .. . . .. . .. . . . . .. .. . . . . . . . . . . .. . . 3.'5H
Configuration ................................................................ 360 Defect. See Discontinuity.................. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ., 4:j
Constant voltage transformers ........................................... 376 de Forest, Alfred Victor ..................................... : '" 50, .'57..'58, fi(i
Contact head ................................................................... 42 Delayed crack. See also Cold crack. . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . .. 84
Contact indication ...................................................... , . . .. 234 Del operator. .. . . .. .. . . . .. . . . .. .. . . . .. .. . . .. . . . .. . . . . .. . . . . . .. . . . .. . . . . . . .. . .. J:3:3
Contact magnetization .................................. , . . . . . . . . . . . . . . . . .. 360 Demagnetization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 30-33. ·,n
Contact method. See Current flow technique and high retentivity.......... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 255
Contact pad ................................................................... , 42 and,hysteresis .. : .. e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120-121
Contamination . ' and magnetic shielding .. , .... " ., ., ..... , .............................. , ... :304
ai~borne . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 332 and residual magnetism ............................................... 292-294
and false indications......................................................... 235 by alternating current ............................................. 27,297-298
and reference standards.... . . . . . .. .. . . . . .. . . . .. . . . . . . .. . . . . . . . .. . . . . .. . . ... 341 by direCt current. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 2~)I)
and settling test.. . .. . .. . . . . . . . .. . . . . . . .. . .. .. .. . . . . . . . .. . .. . . .. . . . . . .. . . .... 211 by single-phase full-wave direct current. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 21>
of wet method particles.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 208 cable wrap method ........................................ " . . . . . . . . . . . . . .. 298
specifications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 432 coercive force in ........................................................... '" 24
test of fluorescent suspension.................. . . . . . . . . . . . . . . . . . . . . . . . . . .. 239 Curie point testing ..................................................... , 30-31
Continuity equation. See also Displacement Ctlrrent . . . . . . . . . . . . .. 106, 113 effect of magnetic field origin on .. . .. .. .. .. .. .. .. .. . .. .. .. .. .. .. .. .. .. ... 293
and electromagnetic boundary conditions .......................... 112-113 electro'magnetic.. .. . . .. . .. . . . . . .. . . .. . .. . .. . . . . . . . . .. .. . . . . .. .. . . . . . . . . . .. . .. :31
Continuous direct current magnetization .............................. 242 equipment .......... , ............... , .... '" ... '" .......... 306, 3.'50, .364-:366
Continuous test method ............................................ 37,38,42 field indicators.................................................. . . . . . . . . . . .. 303
and circular magnetization ................................................. 151 flux curve .................................................................... :364
and wet method testing .......................... " ...................... " 212 half-wave direct current.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 27
indications in ................................................................. 38 hysteresis curve .. ·...... .. ... . .. .... .. . .. ... . . . ... .. . .. .. .. .. .. :36·!
suitability of................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 38 in circularly magnetized objects ........................................... ' :Q
to test castings .............................................................. 202 leakage field.................................. .............................. 128
to test welds............ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 202 limitations of. ....................................................... , ...... , . :32
specifications ............................................................. '" 436 method selection.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300-30 I
Contrast .. , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 229 methods of... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 30-:32
Contrast ratio ................................................................ 239 mobile test systems .............................................. 298, 3.57-:3.51>
Conversion factor ...... .. .. . .. .. .. .. . .. .. .. .. .. .. . . . .. .. .. .. .. .. . .. .. .. . . ... 103 of circular fields .................................................... , ........ 294
Cooling crack. ........................................................... , .'5-6, 82 of longitudinal fields....................................................... 29:3
Copper oxide rectifier. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 68 of multiple poles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
Copper penetration .................................................... .'59, 233 of pie gage................................................................ 241
Corona discharge.......... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 326 of tubular product........................... . ...... :307-:311
Corrosion. . .. .. .. .. .. .. .. .. . .. .. .. .. .. .. .. . . . . .. . .. . .. . .. .. . .. .. . .. ... 1O-11, 208 partial .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 311
Coulomb's law .......................................................... 104,125 particle shape and. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5
Crater crack. See also Solidification crack ...... ........................ 8, 85 portable test systems .................................................. :360-361
Creep .......... , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 9.'5-97 post-test specifications .................................................. _. ·t:38
Criterion for a Recommended Standard Occupational power pack test systems .............................................. 298. 36Z
Exposure to Ultraviolet Radiation.. . .. .. . . .. .. .. .. .. .. . .. .. .. .. . ... 322 pretest specifications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 4:33
Critical damping ........................................................ 170-412 principles of.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . .. 29.5-296
Crystalline stmcture .... " .. " ............................ , .................. 31 problems .......................................................... 304-:30.5, 311
 INDEX 1443

1:3 pulsating reversing method ................................................ 298 indications ................................................................ 17. 27

1:3 orientation, effects of ..................... , . . . .. . .. . . . . . . . . . . . . .. . . . . . ... . . .. 16
reasons against........................ .............. :30
;,82

~;1;~~:~~~'~ :~f. :. ::".:::::::: :.': :.:. . :.:.:. :.: :.:.::.:::::.:::.:.::. :.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.: ':i:~-~~~
reasons for.................................................................... 30
',42
121
remote ..................................................................... " 3.58
.36.'5 reversing cable wrap method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 298
reversing direct current contact coil method...................... .. . . .. 298 sources of... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 2
J:3.f

~~(~;¥~::~+te< 51::31~
Og specialized procedures... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 299
specifications ......................................................... , 30,5, 4:38
43 theOlY ....................................................................... 364
through-coil mcthod.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 297
·1:3 through-current method .................................................. 298 Documentation ......................................................... 419-420
:49 wet hOlizontal systems .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 3053 Domain. See Magnetic dOlll(Jill . ............. " ... , ... , ........ , ............. , 43

........................~
41 when Ilcedcd ................................................................ 365 Drill pipe. See Tubular product
60

g~ ::'i:~~~;;;~
Demagnetizing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 13
]5 Demagnetizing coil. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 43
')6 Demagnetizing current .................................................. " .310
12 Depth of penetration. See Skin depth ............................... 156-157 and subsurface discontinuity indications ........... , ..... , .. , ... , ........ 233

;f~yf~~:\i(i.ii;ii~2:~~
Detection device.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 230-232
human eye ................................................................. " 230
seanning'detector ....................... , ................................... 231
television detector ....................... " ...... , . . . . . . . . . . . . . . . . . . . . . . . . .. 232
9
Deutsche Gesellschaft fur Zerstorungsfreie Pnafung ...... . . . . . . . .. 426
1
Diamagnetic material .............................................. 13, 4:3, 115
3
Dielectric ·constant. See Relative permittivity ............................ 116
Differential permeability .................................................. 171
Dipolar field ................... , ............................ '" . . . . . . . .. 127, 342

E:~~f~:;~.· •. •. •. •. •.~• . •. •. •. •. •. •. •. •. •.~. . . . . . . . . . . . . . . . . . . . ·2~·

Dipole moment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 128
Direct ~bta«;t magnetization.· See CurreTlt flow techntque ........ '" 352
Direct contact method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 233
Direct current. . . . . . . . . .. .. . . . .. .. . . . . .. . . . . .. . . . . .. . . . .. . .. . . . . . . . .. . . . .. .. . .. 43 to test c~ated welds ......................... : . . . . . . . . . . . . . . . . . .. . . . . .. 386-387
and circular magnetization ................................................. 1051
and field flow magnetization .............................................. 153 ~:~~!!~e'.'.'.'.'.'.'.'.'.'.' . . . . . . . . . . . . . .:. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2~~
and mobile test systems.... . .. .. .. .. .. .. .... .. .. .. .. .. .. . .. .. .. .. .. .. .. .. ... 357 Dye and paint system ................................................. 251-252
and multidirectional magnetization .......... " ....................... , . .. 161
and residual magnetization ............... , ................. ; ............ " 293
and stationary test systems ................................................... 35.5
and ultraviolet light scanning .............................................. 251
E
coil ........................................................................... 196 Earth's magnetic field ............................................ 13, 196-197
coil testing .. , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 406 Eddy current .......................................................... , 122, 123
compared with alternating current .................................. 155-1.56 and altemating current coil method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 406
equipment compared "'ith alternating c;urrent equipment ............. 353 and capacitor discharge magnetization..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 409
field definition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 43 and finite element method ................................................ 132

!:s~~~ ~~.t~~~ .t.~~~i.~~. : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : :: ;;

and hystereSis loop ......................................................... 122
and rectified alternating current.. .. . .. .. . .. . .. .. .. .. .. .. . .. .. . .. .. .. .. ... 409

~~~1~1~1~.~~.~1~~~ ~!~
to test welds. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 202
yokes .............................................................. 161-162, 279 :: : : : : : : : : : : : ::: :: : : : : :: : : ::: ::: : ::: : :::: : ::: : :: : : ::: :: :: : : : ::

~?I~i~:;:~7~' ~:~ri'~~i~J~~' ;~. :::: ::::: :::.: :::::: :::::::::: :::: :::::::: ::: :::::: !.~~
Direct current demagnetization .............................. , 32, 298, 366
specifications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 438
Direct current magnetization ................................58-59, 1054-155 testing.. . .... .. . .. . .. .. . .. .. . .... .. . .. ... . .. .. .. .. .. .. .. .. . .. . .. ...... .... 24, 293
and demagnetization ....................................................... 304 to test tubular product .......................................... 166-167, 409
and wet horizontal systems ................................................ 353 Electric arc welding ........... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 293
of tubular product ............................................... 1605, 167-168 Electric conductivity ...................................................... " 106
requirements ................................................................ 410 Electric constant ............................................................. 116
Discontinuity. See also Artificial discontinuity; Defect; Horizontal Electric current density........ . . . . .. . . . .. . .. . . .. . . .. .. .. .. . .. .. .. .. .. . 107
discontinuity; Inherent discontinuity; Linear discontinuity; Electric energy.... .... ............. .. .. .. .. .. . .. .. .. .. .. .. .. .. .... 124
Lougitlldinal discoutinllity, Mauufactl/ring discontinuity; Midwall Electric equipment design ................. , ... , .... , ............ , ... 329-331
discontinuity; Primary processing discontiullity; Radial discontinllity; safety ......................................................... 314-31.5, 3205-327
Secondary processing discontinuity; Service induced discontinuity; Electric field ............................................................. 106-107

EI::;~~ih~·d~~~i~:::::::::::::::::: :::::::::::::::::::::::. ~.~~-~g~~l~~~-~~~

Shim; Simulated discontinuity; Subsmface discontiTluity; Transverse
discontinuity; Welding disconti1luity ................................. 43,76

:~~ ~:~e~~~~i.~~. ~.~~~~~~ : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : :: ~~

Electric hazards. See Shock hazards; Sparking hazards ........... 3205-327

and dry powder sensithity .............. " ................................. 218

and magnetic field....... . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . .. . . . . . ... 160
: :~!~o~d: ~;.~ ;.~ i.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.. ~!~
characteristics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 3 Electromagnet. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 43
data bases........ .. .. .. . .. . .. .. .. .. . .. .. .. .. .. .. .. .. . .. . 385 Electromagnetic boundary conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . .. 111-112
depth ........................................................................ 218 Electromagnetic demagnetization ........ " .............................. 31
 444/ INDEX

Electromagnetic field. See Steady state alternating current field; Field indicator
Time dependent field .............................. 104. 109-113, 114, 124 and testing tubular product . . . . . . . . . . . . .. . 417-4 I (i
Electromagnetic spectrum ................................................ 229 specifications. .... .. .. .. .. .. . . .... .. . .. .. .. ... .. .. . .. . . .... 4:2,"
Electromagnetic symbols ............................................. 102-103 theorv................................................................ 302·:31);1
Electromagnetic units ........................................... 102-103, 105 Field i~tensity. See iHaglletic field strength.......................... W.'i
Electromagnetism. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 102 Field penetration ......................................... " .. " . . .. . :li-,I
Electromechanical switching .............................................. 355 Field strength. See External field strength; Magnetic field strength
Electron beam welding.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 292 and Hall effect probes ........................ ;2·L2
Electron spin .................................................................. 13 and permeability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ;2b
Electronic imaging. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 290 calculation of. .. .. .. .. .. .. .. .. .. . .. .. .. . .. .. .. .. . .. .. . .. .. .. . . .. . . .. .. .. .. . ·HJ9
Electronic Industries Association...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 328 demagnetizing ............................................... _.. _. . . . . .. . . ... :30·+
Electronic switching.. . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 355 end area. calculation of ................ _.. .. .. .. .. . . ... .. . . . .. .. .. . .. .. .... ·Wfi
Electro-optical device ..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 229 maximum effective, and demagnetization .......... " ...... " ....... :300-:30 I
Electrostatic field ............................................................ 107 related to flux denSity and permeability.... .. .. .. .. .. .... .... .. .. .. ..... ;jll:?,
Electrostatic phenomena ................................................ " 104 Field testing _...... __ ......................................... _......... . ....m;;
Empirical modeling. See Experimental TTWdeling ....................... 127 Fill factor ..................................... _. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. -n
Empirical rule ................................................................ 340 Filters ... _. _.... _.................................................... 285-286. :37:2
Enamel coating ................................. " .................... , 386, 387 lA filters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2Hfi
Encircling coil. See Coil method Finite difference method... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 131-1 :32
End area ................................................................. 405-406 Finite element method .... , ........................................... .' . . .. 1:32
Energy ......................................................................... 124 Fire and explosion hazards list.. ....................................... , ;J Ii-,
Energy loss ............................................................. , 122-123 Fixing. See Recording test indications .... .. " .... .. . . .. . . . . . . .. 272, 27.'5, 437
Flake ............................................................................ Ii
Environmental protection ............................................ 331-333
Flammability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . .. 317<31 i)
Equipment
Flash line tear . ........................................................ :: . . . 'I
for magnetic rubber records ............................................... 278 Flash magnetization. See Capacitor discharge . ....... , .. , . . ... . .. . . .. . .. . 4;)
for photographic records ................................................... 284 Flash point ........................................................ " 4.3, 207, :31 i)
safety features ............................................................... 331 Flaw. See Defect; Discontinuity
design ........................................................................ 414 Flaw standard .. '. . . . . . . ... . . . .. . . . .. .. . . . .. . . . . .. . . . . ... . .. . .. .. . 2·12
specifications ...... '" ..... , ................................................. 427· mtificial.......................... ................................... ......... 42
Erythemal ultraviolet. See [N-B; iHidwave ultraviolet ........ : .. 320,322 Fleming right hand rule. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. V50
Etching crack. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 9 Fluid containment ........................................................... 3:31

::;~~!;:::. :~:I~~··.::::·.:·.::::::·.:·.·.::::::·.·.·.·.::·.·.·.·.·.·.·. .:. . . . . '~'.'.' 'i27, ~~~

fluorescence .... , .......... :,. ... .. .. . .. . ... . . .... ... ... . . .. .... ..... 4.3, 44, :37:2
Fluorescent indication
. External field strength .. .. .. . .. .. .. .. .. . .. .. .. .. .. .. .. .. .. .. . .. .. .. ... :302-303 advances in..... . . . . . . . . . . . . . . . . . . .. . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 5f)
External reflector lamp. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 321 . and fixing coating records ................................................. 275
Eyeglasses ................................................................ 371-372 and photographic records .................................................. 4:3
Fluorescent oil bath ....................................................... :396
Fluorescent particle
choice of ............... , .............................................. " ...... 37
F concentration i~ bath .................................................. 209-210
Fabrication discontinuity... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 8-9 control of .................. '. . .. . . . . .. . . . . .. . . . .. . . . . . .. . . . . . . .. . . .. . . . .. . .. .. 2,30
False indication ......................................... 43, 181, 235, 394-397 for wet method testing........ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .. 20(i
causes of............... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 234 light requirements ............................................... 203-204.21:3
Faraday's equation .......................................................... 104 Fluorescent penetrant testing ............................................. , .'57
Faraday's law of induction .......................... 106, 107, 175, 183-184 Fluorescent testing ............................................................57
Fast break. See QUick break and eyeglasses... . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . :371
Fatigue crack ................................................................. , 94 and fiberscopes ............ ~.. .. . . .. .. .. .. .. . . . .. .. .. .. ;391
causes of ........... . .. . . .. .. .. .. . .. .. .. .. .. .. . . .. . .. . .. . . .. .. . . .. .. . . .. . ... 5, 10 ancl fIying spot scanJling .. ... ...................... 2.'57
in bell cranks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 394 and multidirectional systems ..... , . .. .............. :3.54
origins of ...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 4 and ultraviolet lamps.. .. .. . . .. .. .. . .. . . .. .. . . .. .. .... .................. :3(1)
prevention of . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 94 of welds. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402
structure of . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 safety. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . :32-1
Fatigue strength ............................................................. , 10 Fluorescent test systems
Federal Communications Commission .................................. 332 and scanning detectors .. , ................... 2;32
Ferrite ........................................................................... 91 brightness contrast... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22~J
core of pickup coil. .... . .. .... .. . .. .. .. .. .. ... .. . . .. .. . .. .. .. . .. .. .. .. .. .... 187 light requirements... . . .. .. . . . .. . . .... . .. .. .. .. ... .. .. .. . . .. . 2:31
Ferromagnetic material. ..................................... 13,43, 115-1]6 Fluorescent tube .................. . .. . .. .. .. .. . .. .. :37.'5
characteristics of. ............................. , ......................... " 24-26 Fluorescent ultraviolet lamps, tubular. .................. ............. :37:3
production of................................................................... 3 Flux. See Magnetic flux ....... .. .. .. .. .. .... ... .... ...... 11 S, 417
Ferroprobe .................................................................. 191 Flux changes................................ ............... "11.5
Fiberoptics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391 Flux cored arc welding ........................ " ..... " .. ". . . . . . . !lU
Fiberscopes .............................................................. 391-392 Flux curve demagnetization................... .................. :3(),1
and indication measurement .. .. .. . .. . .. . . . .. .. .. . . . .. . . . . .. . .. . .. .. .. . ... 393 Flux density. See Afagnctic flux density......................... It>. 2,1
and weld field testing... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 403 and half-wave direct current. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . 27
Field flow magnetization ................................ ................. 153 and hysteresis.... .. .. . . . . .. . . . .. . . . .. .. . . . .... . . .. ...... '" . .. 12:3
Field flow technique. See Magnetic flow technique and permeability. . .. .. . . .. . . .. .. . . . .. .. .. . . . .. . . . .. .. .. . .. . . .. .... :W
 INDEX 1445

related to fleld strength and permeability... . . . . . . . . . . . . . . . . . . . . . . . . . . .. 302 Half-wave direct current ................................................ 27-28
saturation. . . . . .. . . . . . . . . .. . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . .. . . .. 123 and circular magnetization ............................................... 151
variations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 411 in mobile test units......... .. . .. .. .. . .. . .. .... .. .... .. .. .. .. .. .. ... .. .. .. ... 22
lclux gate magnetometer..................... . ...................... 191 to test welds.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 202
Flux indicator. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 4.3 Half-wave rectified alternating current .................... .' 1.54-155, 353
Flux leakage ............. " . . . . . . . . . .. . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... 27, 126 and mobility .... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 37
and demagnetization.. .. ............... . . . . . . . . . . . . . . . . . . . . . . . .. 30 and pOltabl~ test systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 360
induced. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 180 to test tubular product ..................................................... 409
throllgh coatings............................ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 383 Half-wave rectified current ......................................... 3.57, 402
Flux leakage field. See Magnetic flux leakage . .................. 36, 43, :384 Half-wave rectified direct current ....................................... 151
Flux le:lkage method. . . . ................................... 43 Hall coefficient......... .. .. .. .... .. .. .... .. .... .. .. .... 188-189
Flu"'( lines. See Line's Off01(;('
Hall crystal. . . . . . . . . . . . . . . . . . . .. . . . . .. . . . . . . .. . . . . . . .. . . . . . . . .. . . . . . .. 189
Flux linkage... . . . . . . . . . . . . . . . . . .. . . . . . . . . . . .. . . . . . .. :310-:311 Hall effect... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ................................. 44
Flux sensed demagnetization ........................................ 310-311 devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 253
Flux sensitive de\ice. " .. . . . . . . . . . .. . . . . . . . . . . . .. . . ... . . . . . . . . . . . . . . . . ... . .. 183 gauss meter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 242, 303, 345
Flux shunting de\ice. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 253 measurement ............................................................... 242
Fluxmeter ........... " . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 43, 407 meter ......................................................................... 347
and current pulse magnetization ..................................... 415-416 probe ......................................................................... 242
and pulse current magnetization... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 173 Hall element ................................................. 168, 180, 187-190
Flying spot scanning ..................... : ....................... 257-258,262 gaussmeter . . . . . . . . . . . . . .. . . . . .. . . . . . . . . . . . . . . . . . . . . .. . . . . . . . .. . .. 166, 402, 406
applications .......... ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 258-2.59 methods. .. . . . .. .. . . . . . . . . .. . .. . . . .. . .. .. .. . . . . . . . . .. . . . .. .. . .. .. .. .. . . . . . .... 40.5
characteristics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 259 element sensor ......................................................... 187-190
compared with television scanning ....................................... 258 Hall probe gaussmeter ..................... : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 176
Foaming ....................................................................... 208 Hall sensor ................................................................... 189
Foerster leakage field theory .............................................. 418 Hall voltage....................... . . ... . . . . . . ... .. .. . . .. . .. . . . .. . . . . .. .. . . . .... 189
Foerster microprobe. See also Fen-oprobes ............................. 191 Hand probe ................. : . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 23
;~~:!:: ~';:~.'.'.'.'.'.'.'.'.:.'.'.'.:.'.'.'.'.'.'.'.'.'.'.'.' . . . . .;. . . . . . . . . . . . . . . . . . . . . ~ . . . . . ~ . . . . . . . . . ~-~
Footcandle .................................................................... 230
Handheld equipment .....................' ...... .'. . . . . . . . . . . . . . . . . . . . . . . . . . .. 360
Handling.. .. . . ... . . . . . .. ..... . .. . . . .. .. . . . .. . . . .... .. .... . .. .. . . . . . . . . ... 246-247
Hard ultraviolet light....................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 375
Footlambert .................................................... _............. 230
Hardness ...................................... : . .. .. . . . .. . .. .. .. . . .. . . .. . . . ... 234
Force ............................................................................ 125
Hazard Awareness Communications Program ........................ 319
Forged lap ............ : ..................................................... 82-83
Fracture toughness ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 94, 97 Haze filter .................................................................... 286
Full automation... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 2.50 Head shot .................... : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 44, 366, 434
Full-wave direct current.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 43, 362 Heat ............................................................................ 332
Full-wave direct current magnetization .... , . . . . . . . . . . . . . . . . . . . . . . . . . .. 176 Heat affected zone.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ........ 2:3"4
Full-wave rectified alternating current. . . . . . . . . . . . . . . . . . . . . . . . . .. 298, 409 Heat affected zone crack. See also Hot crack ......................... 8, 85
Furring .......................................................................... 43 Heat treating crack .................................................... 8, 91-92
Hazards. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 327
Hazardous areas..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 326-3~7
G Hazardous ingredients list ................................................. 317
Hazardous installation .................................................... " 332
Gas metal arc welding ................................................... 87-88
Gas tungsten arc welding. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 89 Health ..................................................................... 331-333
Health hazards list .......................................................... 318

g:~::; ~~·~~ti~~·"""""""""""".·"""""""""""""""""""""""""""""""""""""""" ..~~: .~~' ~~!

Gauss' law ................................................... 104, 107, 112, 117
Gauss' law for electric charges......... .. .. .. .. .. .. . .. .. .. .. .. .. .. . .. .... 106
Heliarc welding ............................................................ " 396
Helmholtz coil .......................................................... 196,303
Helmholtz' theorem ......................................... , . . . . . . . . . . . . . .. 13.5
Gauss' law for magnetic sources ......................................... 106 High intensity discharge lamp ....................................... 320-321
Gaussmeter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 44, 407 Hindered settling ............................................................ 211
and circumferential remagnetization .................................... 308 Hoke, Major William E. ...................................................... .50
and residual induction measurement ..................................... 347 Hoke's patent .................................................................. .50
Hall effect... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 242 Homogeneity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 10fi
Gennicidal ultraviolet. See also UV-C ................................... 320 Homogeneous wave equation. . .. .. .. . . . .. .. .. .. .. . .. .. .. .. .. .. .. . . . .. . ... 11 0
Glow discharge ............................................................... 373 Horizontal discontinuity ............. , ....................................... .5
Gradient. . . . . . . .. . . . . . . . . . .. . . . . . . . . . . . . . .. . . . . . . . .. . . . . . . . . . . . . . . . . . . . ... 133 Horizontal wet alternating current magnetizer .. . . . . . . . . . . . . . . . .. 306
Gravitational forces... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 128 Horseshoe magnet.................................................. 44, 148
Grinding crack............ .................. .. ............. 9, 91, 398
Hot crack..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 8.5-86
Grounding. . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 329-330
Hot tear................. ..................... .. .............. 7, 77, 85-86
Human eye................ .. ........ 230-2:31, 369-371
Hydrogen blister... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ................... 98
H Hydrogen crack.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 98
Half-wave alternating current. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 358 Hydrogen embrittlement................................ . ..... 9,98
Half·wave current. See also Single-phase rectified Hydrogen flake ............................................................ 83-84
alte'nwtillg clIn-ent.. . .. .. . .. .. . .. . . ............................ 44, 59 Hydrogen induced crack. See also Cold crack; Pickling crack .. 84
Half-wave demagnetizer.... . ...................................... 306 Hysteresis.... .............................. .. ....... 24-26, 44, 118-121
446/ INDEX

Hysteresis curve. ...... 24. 119 Internal stress .. I

and demagnetization .................. . " .. , 31. :364 International Association for Research on Cancer .. :3J\
first quadrant .............. . 120 International Committee on Illumination ... :320
of ferromagnetic material ...... . ........ ...... . .... 116 International Labour Office .. :32:3
Hysteresis loop .......... . ...... 24, 26, 44, 121-122 International System of Units (SI) 24. 102.171
Interpretation ... . -u
Intrinsic safety ... . :327
Irradiance ......... . :3:20
Isotropy. IO(-j
Illuminance .. .... ..... ...... . 320
Illumination. . . . . . . . . . . . ... 2:30, 240-241. 368
Image forces........................................ . .... 128 J
Image resolution.................................... . ......... 391-.392
Japanese Industrial Standard ....
Immersion technique. .. . . .. . . . . . .. . . . . .. . . . . . . . . . . . . .. . . . . . . . . . . . .. . . . . .. 38
Japanese Institute of Standards.
Impact load......................... ............... . ...............331
Joint Industrial Council
Impulse current ..... , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . ... 1.54
Impurity ...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 3-4
Incandescent ultraviolet bulb ............................................ :37.5
Incremental permeability........... . ........................ 121 K,L
Inclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .................... 89-90 Keeper ........ ..... ....... ............. .. ... 4-l
nonmetallic. .. .. . ... . .. .. .. .. .. . . . .. .. .. .. .. . .. . . .. .. . .. .. . .. .. . . 78 Kircholrs current law..... 112
oxide ............... : ........................................................... 90 Lack of fusion....................... 7. K7-8k
tungsten. .. .. . .. . . .. .. . .. .. . . . . . . .. . .. .. .. .. .. . . .. . .. . . .. .. . .. .. . . . 89 Lack of penetration. . . . . . . . . . . . . 7, 80
Indianapolis Motor Speedway .......... . . . . . . . . . . . . . . . . . .. .............. .53 Lamellar tear..................... ................ ,'-;6-S7
Indication. See also False indication; NOT/relevant indication; Lam~lIa~ test gage. . . . . 6(-)
Relevant indication. .. . . . .. . .. . . .. . .. . .. .. .. .. .. ... 44 Lammahon............... ........... .... 4-S. ill
and coil magnetization.. .. .. .. . . . .. .. .. .. . .. .. .. . . .. . .. .. .. ... 1.59 Laplacian equation .............................................. 109, 1:30. l:3-t
and continuous test method. . .. .. .. . . .. . .. .. .. . .. .. .. .. .. . . .. .. .... 38 Laser ................................................................. 66,2.57. :393
and leakage field ....................................... " ............ : ..... , 128 Laser scanning. .. . . .. .. .. .. . .. . .. . . .. . . . .. . .. . .. .. .. . .. .. .. . 2,52
and residual test method...... .. . .. .. .. .. .. .. . .. .. .. . .. .. . .. . .. .. . .. .. . .. ... 38 LID ratio.................................... .. ........ 299-300, 3(H. :397
brightness .................................................................... 214 Leakage field................ .. .. .. .. . .. .. 292
failure to form .......................................................... 34, 212 and Hall effect gaussmeter ...... : . . . . :30:1
formation of.. .......................................................... 181-182 and magnetic particle size ......... , . . . . . . . . . . :3,'5
inteq>retation ........................................................... 233-2.3.5 calculation of.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . ... ,342-343, 418 c'! 19
mt:asurement of....................................................... .' 392-393 concurrent. . . . . . . . . . . . . . :387
of cold crack .......................................... : ............... : . . . . . .. 8.5 discontinuity ............................. . 127
of cold shut.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .. 77 forces ...................................... . 12.8
of lack of penetration ............................... : . . . . . . . . . . . . . . . . . . . . . . .. 88 magnitude ..................................... . 128
of porosity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ................................ 89 modeling ............................................. . 127-128
particle size and shape.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 3.5-36 Leech ...................................... . .. .. 41
viewing dry method .. . . . . . . . . . . . . . . . . . .. ......................... 203 Lenz' law ............................... . 166, . W\:)
wet method ................................................................. 213 Lester, H.H ............................ . .')4
Induced current technique. See also Current induction Lifting power ............... . ..l4
technique ............................................................... 64-6.5 Light conditions ............ .. 2:)[
Induced magnetization... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 44 Light intensity ............... . :3D2
Inductance ............... .. .. . .. .. .. . .. .. .. . . .. .. .. .. .. . .. .. .. .. . .. .... ....... 44 Lighting ........................ . 288-289, 4():3
and capacitor discharge magnetization.. . .. . .. . . .. ... ...... .. . .. .. .. ... 406 specifications .................. . ·128
in magnetic fields...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 126 Lines of force ................. . 12.44
in tubular product ......................................................... 171 Line frequency coil magnetizer .. 29S
values .. , 41:1-414 Linear discontinuity .... . 2m
Induction L~near. permeability .... . IOU
and residual field .. ... 19.5 Lmeanty ...... . lOG
heating. . . . . .. . . . . . . .. 293 Liquid dye penetrant test .. :iSH
maximum, calculation of... .. 417 Liquid suspension technique. SeC' Wet method
phenomena....................... .. ................... 104 Liquidation crack. See also Hot crack ....... . 85-S(i
Load limit.................... . ................... . (.)
Inductive ammeter................... 173-17.5,416
Ingot crack ....... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .......... 3, 4 Long wave ultraviolet. See also Black light; (TV-A :320, :32:3-:32·1
Inherent discontinuity. See also Blowhole; Cold shut; Hot tear; Longitudinal coil magnetization ..... . . .. 267
Inclusion; Pipe; Segregation ..... .............................. 3-4, 77-80 Longitudinal discontinuity ..... . 3K.5. :3\)7
Initial magnetization curve. . . . . . . . . ..... ..... .... 122 Longitudinal magnetic field ...... . 2:3.4-1
Inorganic zinc coating.. . .. .. .. .. .. .. ... 386, 3137 residual ................. . . ... 32-:33.29:3
Insulator. See Lossless medium Longitudinal magnetization ...... . 1,5, 19, 44, 149, 1,52. 150- [(j()
Internal conductor. Sec also Central conductor.... 1.58-1.59 and pie gage ........... .. 2-1:2
Internal conductor magnetization. See also Pulse cllrrent and residual field measurement .... 1\:)5
1Iwg1letization .................................... . 19, 166 and testing tuhular product .... ·iO.'5--108
Internal conductor method ., . . . . . . . . . . . . . . . . . . . .. 307-308 and wet method testing . 21
Internal flux flow....... . .. .. .. .. .. .... 16 by pOltable test system. .'361
by three-phase full-wave direct current ..... . 2\)
Internal magnetic field................... .............. 115
 INDEX 1447

:l 1I') by wet horizontal system .................................................. :352

demagnetization of....................................................... 32-33 education in ........................... .. ......................54-5.5, .58
120
early use of.. .............................................................. 50-.51 fluorescent paJiicles ... '" .......... '" ........... " .................. 34
of t~lbular end areas ........................................................ 40S
32:3
171 forging and casting applications ......................................... '" 6.3
Lorenz' equation ............................................ 104, 106-107, 125 fundamentals of. ......................................................... 2, 180
·~4
127 Lorenz' force .............................. " ... " .................... " .... " lS7 intelpretation automation ............................................ 260-26.5
120
parameters of.. .............................................................. 182 media used in. . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 34-38
O() themy ................................................................... '" 12-14 lIlilitmy application ............................................................56
uniform ................................................................... '" 195
observation automation ............................................... 2.57-259
verifIcation of .......................................................... 241-242
oxides used in ................................................................ 34
pigment used in .............................................................. 34
M preventive maintenance applications. . . . .. ................................ 64
plillciples of.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 2, 180
Magnetic field indicator. See also Field indicator .. , .......... '14, :H5-346 process control.... .. .. .. .. .. . .. .. .. .. .. .. . .. .. .. 2.36-24:3
Magnetic field intensity. See Magnetic field strength process monitoring ....................... _ . . . . . . . . . . . . . . . . . . . . . . . . . . .. 2.54-256
Magnetic field production ................................................. 341 production automation ................................. " ............. 2.52-2053
Magnetic field strength.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .... 15-16, 44, 104, 10.5 safety in ................................................................. 314-333
and demagnetization .................................................... 30, 31 specifications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 424-438
and electromagnetic boundal), conditions ............................... III speed of. ..................................................................... 248
and weld field testing... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 402 steel industry applications ................................................ 60-61
effects of . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 305 through coatings.............. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 382-383
7 in complex shapes ....... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 177 ultraViolet light sources for ........................... , . . . . . . . . . . . . . .. 320-322
measurement advances in .. .. . . . . . . . . .. . . . .. . . . . . . . .. . . . . . .. . . . . . . . . . . .. 66-67 verification of................................................................ 181
wave equations .............................................................. 110
visible particle... . . . . . . . . . . . . .. . .. ........................................... 34
source of. ..................................................................... 107
wet method.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 34
Magnetic flow technique. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 44
Magnetic flux. See also Flux .. ........................... : .. .. .. ... 24, 44, 292 pattern recognition ................................... ::. . . . . . . . . . . . . . . . . . .. 260
and demagnetization .................................................... 30, 32 Magnetic penetrameter .....................................................345
field ....................................... : ............... : .................... 12 Magnetic permeability. See also Permeability. '" .. . . . .. . . .. .. . . .. 106, 292
in complex shapes...................... .' ...................... : .. .. .. .. .. .... 176 and underwater testing ... " .......... " •. _................•................. ' 384
in hollow test objects ..................................................... " 1.52 in time dependent fields .................. : ................................ 110
of superconductors ....... .. .. .. . .. .. .. . . .. .. .. .. .. .. .. .. . .. .. .. .. .. .. . .. ... 11.5 of air ......................................................................... 192
Magnetic flux density. See also Flux density .... 105,24,44, 104, lOS, 117 of diamagnetic materials ............................... : .. .. . .. .. .. . .. .. .... 13
and circuit theory ...... " . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 118 Magnetic polarity. See Polarity
and electromagnetic boundary conditions ............................... III Magnetic pole. See also Pole. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 292
surface, measlIIing .......................................................... 2.54 Magnetic polymer specifications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 437
Magnetic flux flow and paramagnetic materials ..................... " 13 Magnetic powder. ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 45, .50
Magnetic flux leakage. See also Flux leakage .................... 2, .50, 1605 Magnetic properties ......................................... 36,201,205-206
Magnetic flux leakage field. See also Flux leakage field ...... 1.5, 180-182 Magnetic reluctance .................................................... 12, 118
Magnetic force ............................................................... 418 Magnetic remanence ........... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 293
Magnetic hysteresis. See also Hysteresis .. ............................... 295 Magnetic resistance .................................................... 12, 118
Magnetic leakage field. See Flux leakage field; Leakage field Magnetic rubber ................................................... 4.5, 280-281
Magnetic leakage flux .................................................... " 148 records of indications ................................................. 277-282
Magnetic moment ................................................ 114-115, 126 Magnetic saturation. See also Sat1l7,,(ltion ............................ 45, 396
Magnetic particle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 405 Magnetic shielding .................... " ................................... 304
black ........................................................................... 36 Magnetic slurry... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 437
contrast. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 36 Magnetic susceptibility. . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. lIS
properties of . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 34-35 Magnetic vector potential. ................................ 107-108, 109, 110
red ......................................... '" ................................. 36 Magnetic writing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 45, 234
shape, and indications ..................................................... " 36 Magnetism. . . . .. . . . . . .. . .. . . . . .. . . . . .. . . . . . . . . . . . . .. . . . . . . . . . .. . . .. .. 13, 4.5, 181
slurry, early use of. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 72 Magnetization ................................................................. 4.5
systems, automated.................................................... 267-268 and hvsteresis .................. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 120
visihility .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 36 and ~et horizontal equipment ....................................... 3.52-.3.53
Magnetic particle testing... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 44, 45 factors in ................................................................ 1054-157
advances in. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 70-73 level indicator ............................................................... 347
applications ................................... 2, 051-05.5, 056, 60-63, 64, 69-70 of tubular product, calculation of. ................................... 407-408
automation of. ........................................... " 2, .50-2.53, 2.54-256 overall ......................................................................... 4.5
automotive application ................................................. " 61-63 recommendations for tubular product.... . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . .. 172
blightness contrast in ..................................................... " 229 recluirements for testing tubular product ........................... 407-408
capabilities and limitations of ............ " ............................ 2, 182
choice of method ........................................................ '" 23.3 ~~~~~~~~~~~~f: :::::::::::::::::::::::::... :::::::::::::::::::::::::::: ::::::: :~:
coated particles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 with electric current............ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 1.50-105.3
color contrast in................. .. ................................... 229 Magnetized test blocks...... .. .. .. .. .. .. .. .. .. .. ... 343-344
COmponents advances in. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 72-73
detection devices ...................................................... 230-232 M~~:i~~i~?~. ~~~~~~ i.5.5~ 1:~
................................................................................................... '.' .
development of. .. . . ....... 50, .58-59 levels and threaded fastener testing. . . . . . . . . . . . . . . . .. 396
dry method .... . . . . . . . . . . . . . . . . . . . . .. ........................... 34 monitor . .. .. . . . .. .. .. 249
t}pes of . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
448/ INDEX

Magnetizing equipment ........... . . . . . . . . . .. 53-54 Multidirectional test systems. :3.5-1

design of...................... . ... 172 Multiple magnetic poles 294
Magnetizing field strength " . . . . . . . . . . . . .. 24. 45, 292
Magnetizing techniques. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 248
Magnetodiode............................ . ....... 190-191,405
N
Magnetoinductive instruments... . . . . . . . . . . . . . . . . . . . . .. 242 National Electric Code ...................... 171,325, :327.328,3:30.41:3
National Fire Protection Association. ................. :32S
Magnetometer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
National Institute for Occupational Safety
Magnetomotance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
and Health ................................. . . ..... 172, 371, 322. :32~l
Magnetomotive force................... . .............. 117, 118 National Institute of Safety and Hygiene ... . 414
Magnetostatic field equations ......... . . . . . . . . . . . . . . . . . . . . . . . .. 130 National Institute of Standards and Technology ............. . 2~H
Magnetostatic phenomena.... . ................. 104, :342 National Machine Tool Builders Association ......... " ..... .
Magnetization curve.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 119-120 National Safety Council ................................. .
Magnetization indication.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 67 National Toxicology Program .............. .
Maintenance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 359 Naval Aircraft Factory .................. . 51
Maintenance crack testing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 398-:399 NavShips .................................... . 271. :3~l5
Malfunction alarms ......................................................... 249 Neighborhood processing ........... . 26:3
Mandrel piercing process .................................................. 419 Nitrided steel ........................... , ......................... . 214

~::~:~~:n~~~~;~~ . ·.·.·.·.·.· . . . :. . . . . . . . . . . . . . . . . ....... .. . . . . . . . . . . . . . . . . . 2~~

Manufacturing discontinuity ..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 8-9
Noise ......................................

Nonhomogeneous Helmholtz equation.

. .................... .
Nonfluorescent particle. . . . . . . . . . . . . . . . . ......................... .
. ................ .
:3:3:2
206
lIO
Martensite. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 91 Nonhomogeneous wave equation. . . . . . . . .................. . Ito
Massachusetts Institute of Technology ................................... 50 Nonmetallic inclusion ............................................... . -I
Nonrelevant indication. See also Indications..... 4.5, 23:3-234. :395-:39(j
Material handling and storage ...................................... 318-319
Normalized system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . lO.5
Material hazards ............................................................. 316
Numerical modeling.. . . . . . . .. . . . . . .. . . . . . . . . . .. . . . . . . .. . . . . . . . .. 127, 1:30-132
.Material identification specifications .................................... 431
Material properties ........ _. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 154
Material safety data sheet ............................................ 316-319 o
Maxwell's equations .................. 24, 104, 106-108, 109, 110, 129, 130 Occupational Safety and Health Act .... _.. .. .. .. . 24.0
and continuity equation ........................................ " . . . . . . . . .. 112 Occupational Safety and Health Administration ... 172, 20:3, 207. 2(1),
and electromagnetic boundary conditions ............................... III 213,314,316,318, 324,32.5, 328,3:12, .!l.!
in steady state alternating current fields.. . . . . . . . . . . . . . ... . . . .•. . . . . . . . .. 109 Ocular fluorescence ........................................................ :32:3'
in time dependent fields ..................................... .'. . . . . . . . . . . .. 110 Ohm's law................ ............. ......... lIS
quasistatic form ............................................................. 109 Ohmic loss .............. . l2,4
Maxwell's first equation ..................................................... 124 Ohmic resistance. . .. .. .. .. .. .. .. . . .. . . . .. .. . . . .. .. . I S·1
Maxwell's second equation ................................. : ............ 1.24 Oil field applications .......... ................ ................... 40.5-420
Maxwell's third equation ................................................... 112 Oil field tube. See Tubular product
Media selection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 37 Oil vehicle. See also Suspension
Meissner effect. . . . . . . . . . . . . . . . . . . . . .. ................................. ..... 115 flash pOint. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... .... .. ..... .. .. .... 20.
Mercury arc lamp. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 372-373, 376 in wet method testing .......................................... _. . 20G-207
and photographic records ................................. , ............... 289 tape transfer record.. . .. . .. .. .. . .. .. . .. .. .. .. .. . . .. .. . . .. .. . . . .. 27:3
high pressure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 375 specifications.......................... ........................... 207, ~1:32
Mercury vapor arc lamp ....................................... 289, 322, 375 temperature limit................................................... 21:\
Mesopic vision .......................... , .................................... 231
Metallurgical properties.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 234 O;;;~~~~o'~ ~ '. ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ : : ~ ~ : ~ : ~ : ~ ~ ~ : ~ ~ ~ . . . . . . . . . .... .... . ... ... :30t, ;~~
Metals, dissimilar............................................................. 235 OSHA Hazard Communication Rule.... . . .. . . . .... . . ......... :31 (j
Microprocessor algorithm. . . . . . . . . . . . . . . .. . . . . .. . . . . . . . . 262-264 OSHA PEL. See Permissible eXpOSltlT limit
Midwall discontinuity. . . . . . . .. ...................... Hi Overlap ......... .. ......... .. DO
Midwave ultraviolet. See also Frytherrwl ultraviolet. UV-B .... ....... 320 Overstress crack .. 9-10
Mobile alternating current ................................................ :306 Oxidized crack .... ~I

Mobile power pack. . . . . .. . . . . . . . . . . .. .................. . ....58-59, 298

Mobile test equipment ...................................... 21-22, 357-359
Mobility. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........ , 36, 38
p
Modeling ................................................................. 127-128 Parallel magnetization ~1.5, 15D
Monitoring magnetic particle test systems ............ 248-249, 256, 339 Parallel pickup coil IS(-j-1K'
Morphological image processing. See also Neighhorhvocl Parallel processors ............ . 2(iO
processing. . . . . . . . . . . . . . . . . .. ..... .. .... . . . . . .. 265 Paramagnetic matedal .. n. -is. II.')
Multidirectional magnetization. 20, 45, 70, 161-164 Parametric evaluation. :310
and automated magnetic palticle testing.. 267 Particle
and field flow magnetization .. ... .... .. ..... ... . . . . . . . . . . . . . . 15.'3 application .................. . ·t3()-cU'
and power pack test systcms .................... . .... .... ..... '" 36:3 bath .... :1:m
and wet method testing. . ................. . 212 concentration. T39, :J·11
automated testing systems for ................ . 248 consistency .. . ·lOO
current monitor for ......... . 249 control ... '... . 2:39
specifications ................. . 436 mohility ..... . 27, :36-:37, :3:m, :3.'):3
techniquc .................. . 248 sensitivity IOU
testing procedures with ...... . shape ..... :3.S-:S(-i
 INDEX /449

·-4 size ............................................................................ 35 Primary creep ............................................................... " 9.5

·1 suspension tank. ... " ............................................. '" ......... 21 Primary processing discontinuity. Sec also Burst; Cooling crack;
type .......................................................................... 3.50 Clipping; Hydmgen flake; Lamillatioll; Rolled lap; Seam;
vehicle safety.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 314 Sf1inger; 'Welding discontinuity .............................. 3, 4-6, 81-90
Paste-on discontinuity. See Shim Primary production and processing steel. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .3
Pattern recognition............ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 260-26.5 Prism block standard..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 400-401
Penetrameter. See also Field indicator . .................................. , 417 Probability of detection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 200
Penetration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. '106 Probe. See l;'erroprobe
Probe demagnetization....... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 32
Permanent magnets.. .. . . . .. . . . . .. . . . . . . . . . . . . .. . . .. . . . . . . . . . .. . . .. 1:3, 4.5, 279
Process control .......................................................... 236-243
Permeability. Sec also Magnetic penneahility .. ............. 26, .34, 4.5, 116 Prod ..................................................................... Z2-23,45
alternating current coil method...... . ................................... 406 and circuial' magnetization. .. .. .. .. .. . .. .. .. .. .. .. . . .. .. .. .. .. .. .. ... 151, 158
continuous test method ............. ,' ........................................ 38 early use of. .............................................................. 53, 57
eddy current coil impedance .............................................. :347 Prod magnetization. See Current flow techniqlle
electromagnetic fields.. .. .. . . . .. . . . . .. .. . . . . . .. . . .. . . . . .. . . .. .. . . . . . . .. . . .. 114 specification ................ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 434
field relations..... .. . . .. . . . .. . . . .. .. . . . . .. . . . . . .. .. . . .. . . . .. . . .. . . . . . . . .. . ... 106 Product identification ...................................................... 316
incremental. . . .. . . .... .. . .. .. . . . ... . . . . . .. . . . . . .. . . .. . . . . .. . . . .. . . . . . .. .. . ... 121 Production speed..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 3.50
nonrelevant indications ..................................................... 39.5 Programmable logic controller ................................ 247-248,249
related to flux density and field strength ...... : .................... '.' .... :302 Pulsating half-wave direct current ...................................... ·. 32
residual test method ..................................................... 37-38 Pulsating reversing demagnetization .................................... 298
threaded fastener testing ........................ : ............ , ............. 397 Pulse current ......................................................... ·. 1,54, 3.57
of diamagnetic particles ....................... " . . . . . . . . . . . . . . . . . . . . . . . . . .. 11.5 Pulse current magnetization. See also Half-wave rectified
of dry method particles .................................................... 201 altenUlting current
of ferromagnetic pmticle.... . . . .. . . . . . . .. . . . . .. . . . .. . . .. .. . . .. . .. . . .. .. . ... 11.5 analysis of............................................................... 169-170
of free space .......................................: .......................... :114 and testing tubular product .......................................... 410-415
of paramagnetic particle. . . . . . . .... . . . . . . . .. . ... .. .. . . .. ... . ... . .. .. . .... ... 115 evaluation of.. ..........................................................,' '173~174
of wet method particles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 205
Pennissible e-xposure limit .......•..'................ ! . . . . . . . . . . . . . . . . .' 213, 316 .~o~~~ll:~pir~d~l~t· :: ::: : ::: : : : : :: :: : :: :: : : ::::::::: ::: :'::: : : :: :: ::: ::: : . i'68-i~!
Perpendicular pickup coil .......... .'. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 185-186
Phenolic epoxy coatings ............................... '.' . . . . . . . . . . . . . . . . . .. 386 p!J~:~;:~:~: :::::: ::::: :::: ::::: :::: ::::: : : ::::::: :::: : ::: :::: ::: ::::: ~;~
Photochromic eyeglasses ................................................... ;372 Pulse magnetization................. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 4.5
Photoelectric detector.... . .. . .. . . .. . . . . . .. . . . . . .... . .. . . . . . . . . . .. . . . . . . .. . .. 232 Pulse length .................................................................. 170
. Photographic record ................................................... 283-290 Purkinje shift ................................................................. 371
specifications .............................................................. " 438
Photometric brightness. See also Luminance ................. '" ........ 230

.~~~!~;!~~~~:r:;~~gi~.~~~~. :. :::::::: . :: . :. :. :. :. :. :. :: . ::: . :. . ::: . :::: ..~~.1.' ~~~ a

Quality control ....................... , .......................... , 246, 2.50, 367
Physical properties list ................................................ 317-318
Quantitative shim............. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 67
Pickling crack. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 9, 92
Quartz fiberscope ..................................................... 391,392
Pickup coil ............................................................... J85-187
Quench crack ..................................................... 8,91-92,394
Pie gage
Quick break.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 4,5
and residual induction measurement ............................. , ....... 34.5
and circular magnetization .............. , .. .. .. .. . .. . . . . . . . . . . . . .. . . . .. . . .. 1,51
and testing tubular product .......................................... 417-418
circuit .......................................................................... 29
and weld field testing.......................... .. .. .. .. .. .. .. .. .. .. . .. .. ... 402
design ...........................................................................57
as magnetic field indicator ............................................ 241-242
magnetization ....................................... , ........ , ........ , 234, 356
Pipe. See also Tubular produd .................................... 3-4, 11, 165
test ...................................................................... 242-243
Piping system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 331
Pitting ........................................................................... 11
Plating. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 292
Plating crack ............................................................ 9,92-93 R
Poison Control Center ..................................................... 318 Radial discontinuity.................. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 1.59
Poisson equation ................................................. 107, 109, 130 Radial field .................................. . ............................. , 384
Polarity. See also Bucking field .. ................................. '" 12, 24, 34 Radial standard ............................................................. , 429
Polarizing filters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 286 Radiation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332
Pole. See also Magnetic pole .............................................. 12-13 Radiography ..........................................................56, .59. 240
Pole pieces, laminated. See Articulated pole piece Raised cross indicator ........................................... 345,417-418
Porosity ............................. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 77-78, 88-89 Ramp current.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 358
Portability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 66 Reactive materials list ...................................................... 318
Portable alternating current half-wave demagnetizer . . . . .. 306 Recirculatory system ....................................................... , 209
Portable power pack.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 298 Reciprocity failure ............................................... 287, 289-290
Portable surface discontinuity. See Shim Recording test indications. See Alginate impressioll record; Archival
Portable test equipment.. .................................... 21-22,360-361 quality; Fixing coating; MagnetiC mbber; PhotographiC record;
Potentiometer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .............. 360 Tape transfer record
Powder. See Dry powder Records. specifications ............................................ 431, 437-438
Powder blower ....... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ............ " 4.5 Rectified alternating current. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 4,5
Power pack test system .......................................... 21,362-363 Red magnetic particle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 3,5
Poynting theorem................................ . ............... 120, 124 Reference standard. See also Art!ficiaZ discontilluity; Ring standard;
Pressure system ...............................................................331 Sensitivity testing; Test block; Verification. . . . . . . . . . . . . . . . . .. 45, 338-347
450/ INDEX

and yoke testing ........... . 400 of dry method particles .............. .. :201
for coated welds .......... . 386 of ferromagnetic matelials ...... . I :~
for magnetic field production............... 341 reverse polaJity ........... . 24-:26
for multidirectional svstems .. 3,54 Reverse magnetic field strength .. . .. 2·[
fix particle concentr;tion . . . . . . . . 341 Reverse polarity ............................ . 24-2(i
for threaded fasteners. . . . . . . . . . . . . . . . 396-:397 Reversing cable wrap demagnetization. . . ........... . .. ... 29K
for weld bead cracks ............................ ..................... 386-387 Reversing direct current contact coil demagnetization ..... . 29K
for wet horizontal systems.... .. .. . . . .. . . . .. .. .. . . .. .. . .. . . ... 35:3 Reversing direct current demagnetization .......... . 301. :106
Right hand rule.. .. .... " ........... .. 19, 107. 1.'52
specifIcations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 428
Ring magnet. . . . . . . . . . . . . . . . . . . . . . .................... . 14.';
Regulatory requirements. See also Safety. . . . . . . . . . . . . . . . . . .. :328-33:3
Ring standard. See also Reference standard; Test ring " 46, :341·:3 c [:,

Reichs Roentgenstelle ................................... . ........... 51 Robotics .............................................. . :1K4

Relative permeability .. .. .. .. .. .. .. .. .. .. .. .. .. . .. .... 102, 254 Rockwell-B scale .................................. . 240
Relative permittivity. See also Diclectlic constant... ..". 102. ll6 Rockwell-C scale... . . . . ............................ . Wi
Relevant indication. See also Indication ....... ................. 45. 233-234 Rod ................... . ............................ . 231, ;JG9
Reliability studies ................................................ .......... 339 Rolled lap....................... . ............ . sa
Reluctance ...................................................................... 26 Rowland ring technique' ............................. . .. .... 17:)
Remagnetization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. :36.5
Remanence. . .. . .. .. . ... . ................................................ 24, ll9
and current pulse magnetization........................ .. . . .. .. .. .. .. .... 41.5
and demagnetiza.tion ....................................................... 310
s
and mobile capacitor discharge systems .................................. 358 Safety .. .. ..... , ..................................... 63, 246, 314-:3:3:3
and residual test method .................................................... 37 regulatory requirements..................................... 316-:319
Remanent magnetism. See Residual magnetic field Safety testing.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ..' :327
Remote operation ................................................... , ...... 358 Saturation level. See also MagnetiC saturation...................... 229
Remote viewing equipment ................................ : . . . . . . . .. 390-393 Saturation point... . . . . . . . . . . . . . . . . . . . . . . . . . .. .................... 24
Residual demagnetism ........... : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 30 Saturation region.. .. .. . .. .. . .. .. .. . .. .. . .. .. .. .. .. .. .. .. .. .. . .. 1 19
Residual field ........ "................................... : ......... 19,5-196,292 Scab .................................. , ........ ............... ....... i
Residual flux density.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 292 Scalar factor ..................... '" ............................... ... l:.n
Residual flux intensity ....................................... , .......... .'.. e' 31 Scalar quantity ................................................................ 106
Residual induction. See Residllal l7U1gnetism Scanning: See also Dye and paint system; Laser sca~ltIillg: Ultraviolet
Residual induction testing.. . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 406 light scanning; Visible light scanning
Residual leakage field. . . . . . . . . . . ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 122 automatic, history of.. .................................... : ............ 2,50-2,52
Residual magnetic field ........................................ 4,5, :37, 293-294 detectors ........ .' ...................................................... 231-2:32
Residual magnetism ............................................. 292-294, 36,5 fluorescent testing ...................................................... " ." 2.57
and half-wave direct current... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 28 Scotopic vision ......................................................... 231, :171
in reference standards ..................................................... , 344 Seam.. .. .. .. .. . .. ... ................................ ................ 4. 5, K 1
in settling test ....................................... " ................ 210-211 Secondary creep.. . . .. .. .. .. . .. .. .. .. .. . .. .. .. .. .. . . 9:')
in wet method .......................... : .. .. .. .. .. . .. . .. . .. .. . .. .. .. ... :3,5 Secondary processing discontinuity. See also Grinding crack;
of dry method particles ..................................................... 201 Heat treating crack; Quench crack; Machining tear; Pickling crack;
polarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 34 Plating crack ......................... .. .. ....... ... . . ... 91-9:3
problems ................................................................ 292-293 Secondary processing test. .. .. . .. .. . . . .. . ... .. .. . . .. . .. .. .. .. .. .. .. . . :3
sources of.... .. . .. .. .. .. .. .. . . . .. . .. .. . .. .. .. . .. .. .. . . .. .. .. .. .. .. . .. .. .. . ... 293 Segregation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7K-I';o
Residual magnetization.......... ",,,,,,,,,,,,,,,,,,,,,,,,,, 31 Selenium cell photoelectric meters ....... '" .. .. .. .. .. . .. .. .. .. .. .. . :377
and automated testing. . . 268 Self-demagnetization. '. . . .. . .. . . .. . . . . . . . . . . . . . . . . . . . . . . .. . .... 24:3
and false illdications . . . . . .. .39,5 Service induced discontinuity. See also Creep crack; Fatigue cra('k;
charactelistics.... 34 Hydrogen crack; Stress corrosion cmck. 9·11, ~)'!-9fl
Residual method. . . . . . . 37 -:38, 46 Service test. . . . . . :)
and circular magnetization. . . . . .. 151 Settling range .. . . . . . . . . . . . . . . 210
and wet method testillg. . . . . . . . . " ..... , 212 Settling test. . . . . .. . . . . . 46, 20G, 201->. 400
indications in ................... . . . . . . . . . . .. 38 and bath contamination. . . . 2:3(i
speciHcations . . . . . . . . . . . . . . . .. 436 and slIspension control ;2:36
Residual stress. .. . ... '''''''''''' ..... "..... . .. ... 9 directions for ............... . 2:39
Resistance effect of contaminants. 211
and capacitor discharge magnetization ...... 406 parameters ............................. . 210
in current pulse magnetization... . ... 412 supernatant condition ........... . 211
minimizing.............................. 171 tube ........... . ................... . 210-211
of human body............ 325 Shared tlm.: discontinuity standard .. J·l5
values...................................... ............. 4]3 Shielded metal arc welding ... K9, :3i'jS
Resonant carrier-to-interface ratio circuit. . . ,59 Shim ........................ . ... 242, J4(),W,
Retentivity .. .. .. .. .. .. .. .. .. .. .. . .. .. .. ... 24, 34, 46, 119 Shock hazard ................ . :3:25·:3:2fi
and hath application.. . . . . 2,5,5 Short shot magnetizing... . ........... . fig
and circular magnctization........... . ....... 151 Short wave ultraviolet. See also UV-C ;)20
ami continuous tcst method... . 38 Shot ..................................... . <\(i
and demagnetization. . . . . . . . . . . . 30, 36,5 Shrink fit . 2:3·[
and residual magnetism... . .............. 29,5 Shrinkage cavity. . ................ . 4
and residual test method.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 38 Shrinkage crack ..... ... ..... . ................ .
nonrelevant indications of . .. .. . .. .. . .. . .. . .. .. . . .. .. . . .. . . .. . . . .. . . .. .. ... 234 Signal-to-noise ratio. See also Background fluorescence. .. 229, 236. ·toO
 INDEX 1451

:20]
1:3
Silicon diode . 0 ••• 0 0 •• 0 • 0 • 0 0 73
0 0 0 0 0 •••••• 0 0 0 •• 0 0 •••••• 0 • 0 • , 0 ••• 0 00 •• 0 ••• 0 0 •• 0 0 • ••

Stray field. See also Magnetic flux leakage................ . ....... 180

:?A-26 Silicon thyristor .. 0 0 0 0 0 0 ••• 0 73
• 0 ••• , • - ••• 0 0 0 0 0 0 • • •• • • • • • • • • • • • • • • • • • • • • • • • • • • • • ••
Stress corrosion crack ....... : ................................. 11, 97-98, :399
Simulated discontinuity. See also Reference standard ....... 174, 417-419

~~]~~~;il~ii~:':::.t··.···.·.·.·. ·. ·.· .· . ·.· .· .· . •.4:.81~

M
;2·4-:?Ji Single-leg technique ... 0 :384
•••••• 0 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • ••

29H Single-phase alternating current ..... 27, 353 0 ••••••••••••••••••••••••••••••

29H Single-phase fuD-wave direct current ............. 27, 28 o •••••• " ••••••••••

,:]O(i Single-phase full-wave rectified alternating current. . . . . . . . . . .. 28, 154 ··.··.··.··.i··.··.··.··.··.··.··.·••••

~~~~i~~~i~::~::~:i~:·:·:·:·:·:·:·:·:·:·:·:·:·:·:·:·:·:':':':':':':':':':':':':':':':':':':':':0:':':':':':'.·~.:·.i6'2~!
, 1.'52 Single-phase rectified alternating current. See Half-waV(' current
lotH Size and shape
1-:343 of (hy method particles....... ..................................... 201
:31).1 of wet method pmticles... . . .. ..................................... 205 and ~;Iternating current.. .. . . . .. . .. .. . .. .. . .. .. .. .. . .. .. . .... . .. .. .. . .. 27
240 Skin dcpth Superconductor. . . . . . . . . . . . . . . . . ........................ , . . . 11,5

!~~a~~~#/):~
W7 and alternating current coil method....................... . . . . . . . . . . . . . .. 406
. :3()9
and through-coil demagnetization .............. 297 0 •• 0 ••••••••••••••• 0 • • • • ••

H:3
equation ........... o .... 0 ...... 0 .............................................. III
173
in time dependent fields ............................... 110 o
.......
o
...........

Skin effect. .................................... 27, 46, 1.54, 1.5.5, 156, 293, 3.53
Surface tension ........................................................ ..... 208
and altemating current demagnetization ................................. 300
Surface treatment....... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 333
and demagnetization .............. 0 ••••••••••••••• 32 _ •••• ' •••••••••••••••• 0 • • • ••

Surge magnetization...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . .. ,59

-J:n and flux leakage fields .......................... 181 0 ...........................

Suspension. See also Bath; Vehicle .......................................... 46

-:31\:) and pulsating reversing demagnetization ........... 298 0 ••••••••••••••• 0 • • • ••

concentration and settling test. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 400

:327 by Single-leg technique ........... 0 • '" ••• 0 384 •••••••••••••••••••••• 0 ••••••••••

control ............................................. .......................... 236

229 by single-phase full-wave direct current 28

£:~S~ ~~~.~ ~~~~~~ ~~~~~~.~:::::::::::::::::::::: ~6~-:~~

0 ••••••••••••••••••••••••••••••• 0 ••

24 testing through coatings .................................................... 383

J 19
weld field testing ............................................ :.............. 402
Slag inclusion ............................................ : . . . . . . . . . . . . . . . . . . . .. 89

i2.::':Yf;i:..~~~~~rt...., •. •. •. •. •. •. •. •. •. •. •. •. .~g
7 Swinging held. See also MultidirectiOluil nlOgnetiza!ion .. :............. III
IT3 Slurry ..................... : .......... : .................... ;' . . . . . . . . . . . . . . . . . . .•. 46'
lOG s'ociety for Nondestructive Testing ....................................... 66
Society of Aerospace Engineers....... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 343
Society of Automotive and Aerospace Engineers ......... 204, 209; 42.5
~.'52
Solid state physics ............................................. '. . . . . ... .. .. .. 188
132 Solidification crack. See also Hut' crack .... ................................ 8.5 T
2.57 Space utilization.. .. . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 246
Tangential field strength meter ....... :................ . . . . . . . . . . . . . . . . .. 428
371 Sparking hazard.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 326
Tap switch ..................................................................... 360
HI Specification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .. 424
Tape transfer record ....................................... 273-274,388,389
and B-H curve .............................................................. 167
Te:;~~!:~:ti~~:~~. :~~
9.5
for magnetic particle testing .......................................... 424-438
:: :: :: : ::: : : : :: :: : : : : :::::: :: : : : : :: : : : : :::: : : : ::: :: : : :: :
sample ........... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 427-438

• .•.•.•.•. • .• . . .• . • •.• . .•. . . .• •. ~5~.257-~~

Television detector........ . . . . . . . . . . . . . . . . . . . . . . . . . . .. ..................... 232
steel cleanliness. . .. .. .. .. . .. . .. .. .. .. .. .. . .. .. .. . .. .. .. .. .. .. .. .. .. .. .. .. .. . ... 4

~:~=~~S:g
9:3
:3 testing ............................................................. 16.5, 431-438
Spectral characteristics
HO
77 of mercury arc lamp.... . . .. . .. .. . .. . . .. .. .. . .. .. . .. .. . .. .. .. .. .. .. . .. . .. ... 373
Tensor quantity. See Electric conductivity; Electlic permittiDity;
of visible light .......................................................... 367-368 MagnetiC peaIWability
of ultraviolet light... . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . .. 377
Tertiary creep............................... .. ......................... 96
f}H Split prism test block. . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . .. 343
Tesla ................................................................... 24,46, 102
:3 SRI algorithms .............. " ..................... " ................... 264-26.5
Test block See also Reference stoTldalli . ............................ 343-344
10 Stamping marks.. . . .. . . . . . .. .. . . . . . . . . . .. . . . . .. . . .. . . . . . . . . . . .. . . . . . . . . . .. ... 234
Test piece. See Reference standard
10 Standard. See Reference Standfm/ or Specification ...................... 424
Test ring. See also Ring standard. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 46
30 Standard depth of penetration. See also Skin depth; Skin effect ..... 181
36
)8
Static electric field.. . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 124
Static field.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 107
~::: :7t:~~~~~:.::::::::::::::::::::::::::::::::::::::::::::::::::::::::::' 3'i4-~~~
1
o
Static magnetic field ...................................................... " 117
Stationary magnetic particle test system ................................ 21
~:::i:;:;;a::~;;~~.ts. : : : : : : : : : : : : : : : : : : : :: : : :: : : : : : : :: : : : :: : : : : : . ~7;.:o-~~~
Testing procedure specifications .................................... 431-43~
Stationary power pack ....................... _.............................. 306

~~~=l~;:~::;~~~!~h:: ::::::: ::: . ::.:.::: :: :::.::: :: ~~i

1
1 Stationary test equipment.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 35.5-3.56
,5 Steady state alternating current field. _................................. 109
H Steady state creep .................................... _. . . . . . . . . . . . . . . . . . . . . .. 9.5
Threaded fasteners.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 396-397
7 Steel industry application. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 362
Three-phase alternating current. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 28
(i Steel billet .................. " ................................................ 216
Three-phase alternating system. . . . . . . . . . . . . . . . . . . . . . . . . . . .. ............ 3.55
9 Steel. machining... . . . . . . . . . . .. . . . . . . . . . . . . . ............................. " 4
Three-phase full-wave direct current. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 28-29
Steel ring test. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 240
Three-phase fuD-wave rectified alternating current. .. ....... 154, 3.53
Steinmetz, Charles. -. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 123
Three-phase full-wave rectified magnetization ...................... , ~,56
Stokes. George.............................................. . . . . . . . . . . . . . . . . .. 239
Three-phase rectified alternating current . . . . . . . . . . . . . . . . . . . . . .. 1,54, .353
Stokes' theorem. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 134
Threshold limit value.. .. . .. .. .. . .. .. . .. .. .. .. .. ... .................... 316
Storage battery ...........................................................51, 1.54
Through-coil demagnetization ............................................ 197
Straight wire signal................................. . ................ 183-18.5
Through-coil method. See Cui/method
Straightening crack. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 9
Through-current demagnetization ....................................... 298
 452 I INDEX

Time dependence ........................................... 110-lll, 169-170 US Department of Labor. . . . . . .. . ............................ 316, 328. 332
Time variations ................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 411 US Environmental Protection Agency ..................... , . . . . . . . . . . .. 201'
Timer control specifications ., ........................................... , 428 US Navy .....................................................................51, 54
Tool steel ring standard. See Ring standard. . . . . . . . . . . . . .. . . . . . .. . . . . .. 117 Utilities ., . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328-329
Toroid coil.. . . . . . . . . .. . .. . . .. . . .. . . . . . . . . . . . . . .. .. . . . . . . . . . . . . . . . . . . . .. . . . . . .. 417 UV-A. See also Black light; Long wave ultraviolet . .......... , ...... 320, 323
Toroidal field. See also Curr'eTlt induction technique .. .................. , 46 UV-B. See also Enjthelllalltltraviolet ....................................... 320
Torque .................................................................... 125-126 UV-C. See also Actinic ultraviolet. . . . . . . . . . . . ... . .. . . . .. . . . . .. . . . . .. . .. . . ... :320
~;::::~~:::t~~~~t~rl'~~i~ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ·.3~~
Transverse discontinuity
and coil magnetization ..................................................... 153 v
and longitudinal magnetization ............................................. 19 Vapor density ................................................................ 317
and underwater testing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 38.5 Vector. See also "\-Jagnetic vector potential ........................... 110, 411'
in threaded fasteners..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 397 Vector factors ........................................................... 133-134
True continuous method...... .. .. .. .. .. . .. .. . .. . . .. . . .. . .. .. .. .. .. .. .. .. ... 46 Vector theorems ........................................................ 134-13.5
Tubular black light. See also Black light .... ............................ , 289 Vehicle. See also Bath; Carrier Fluid; Suspension ......................... 4(1
Tubular product. See also Pipe Verification. See also Reference standard; Sensitivity testing ..... 338, 350
and alternating current coil demagnetization. . . . . . . . . . . . . . . . . . . . . . . . . . .. 309 Vibration .................................................................... " :331
circumferential remagnetization ..................................... 307-311 Virgin hysteresis curve ..................................................... 24
demagnetization of .............................................. 299,307-311 Viscosity ..................................................................... " 239
end testing .............................................................. 308-309 and wet method particles ................................................... :37
testing documentation ................................................. 419-420 effects of... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .35
testing situations ........................................................... , 169 of oil vehicles ................................................ , .. .. .. .. ... .... 207
Tungsten inert gas welding .......................................... 395-396 test ...................................................................... 239-240
Type B standard. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 429 Viscous force ............................................................... " 12H
Type C standard....... . . .. .. . . . . . .. . . . . .. . . .. . . . .. . .. . .. . . . . . . . . . . .. . . . ... . .. 429 Visibility ............................................................. " . 339, 341
Type R standard.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 429 and contrast .................................... ,........ ............. 202, 20(-i
Visible indications .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . .. .. .. .. .... 287, 288
Visible light.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 4(-i
and borescopes ......................................................... 390-:392
and fiberscopes .............................................................. :391
• Ultimate stren~h ............................ :................................. 9 filters ......................................................................... 285
Ultrasonic test method ...... .'... : ........... : ......................59, 16.5, 235 intensity .......................................... : ......................... " 392
Ultraviolet tlXtures .................................................... , 373-375 requirements .............. : ... '.' . ~ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 241
Ultraviolet light. See also Black light; VV ....................... 46, 320-324 scanning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... . . . . . . . . . . . . . . . . . . . . . . . . . . .. 2.'50
and borescopes ......................................................... :390-392 sources ................................................................... 367-368
and fiherscopes ............................................................. 391 specifications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 421-1
and television scanning .................................................... , 257 to test threaded fasteners. ..... . .. . . .. . . . .. . .. .. .. .. . .. .. .. .. .. . .. .. . .. .. ... :397
filters for photographic records. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 285 Visible test systems ............................................... 229,231-2:32
human eye response ............ , .. .. .. .. .. . . . .. . .. .. .. . .. .. .. .. .. .. .... . ... 369 Visible particle
intensity .... :. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 392 choice of . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . :37
measurement of. ............... : ..................................... " 377-379 concentration in bath.............. . . .. .. .. .. .. .. . . . .. . .. . .. . . .. .. .. . .. .. . .. 20D
meters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 372, 377-379 light requirements .................................................... 203, 21:)
radiometers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . .. 379 Visual acuity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 231. :3hD
regulatory requirements .............................................. 377-378 Voltage parameters .............................................. 357-35S, :3(-j()
safety ......................................................................... 315 Voltage sources ...................... , ............... , ................. , ..... :37()
scanning ................................................................ 250-251
sources ...................................... :. . . . . . . . . . . . . . . . . . . . . . . . . .. 320-324
specifications ............................................................... , 428
testing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368 w
to test threaded fasteners ..... , .............................................396 Water bath ............................................................. . ... ,. 27:)
to Held test welds. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 403 ,"Vater break test.. ..................................................... . 43:3-431
tubes .......... , .............................................................. :322 Watcr conditioning. . . . . . . . . . . .. ....................................... 207-2(1)
verification procedure ................................................. 240-241 Water vehiclc. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 2.07 -20S
visible light interference... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 368 and tape transfer records.. .. .. . . .. . .. .. .. .. .. . . .. . . . .. .. .. . .. .. .. . .. .. ... 27:)
Underbead crack. See also Cold crack ... ................................. , 84 specifications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1:32
Undercut ........................................................................ 90 temperature limits" .. .. .. . . .. ... . ... .. .. .... .. .. ... . .. .. .. . .. .. .. 213
Underwater testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 3R2-385 \Vatertown Arscnal ... .. .. .. .. . .. .. .. .. .. . . .. .. . . .. .. . . .. .. . .. . .. .. . '}-(
Underwriters Laboratories............................ .............. 27,328 Wave equation ..................................................... . 110
Unhindered settling ......................................................... 211 \Vavc propagation ....................................... . lO4, 110
Units of measurement ................................................. 104-105 Weber ........................... .
Upper explosive limit ....................................................... 318 Wedge gage ........................ .
US Army ........................................................................ 50 Weld
US Army Air Corps ...................................................... 51,54 and nonrelevant indications. . ... ... . ..... . . 2.:3:')
US Bureau of Standards ..................................................... 50 bead................... ., ........................................ }oj
US Corps of Engineers ..................................................... , 69 (hy method testing of ...... , ....................... . ..... 202, :306<)·,,7
US Department of Defense. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 204, 426 field testing........................ ... 402··1(14
US Depal1ment of Energy ............................................... , 426 geometIy.. .. .. .. . .. .. .. .. .. . .. . .. . . .. .. .. . . . ...................... :39.')
 INDEX 1453

in hellcranks ................................................................. 394

profile ........................................................................ 387
surface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 402-403
::E:;7~~. :'~~~~~~i.~~ .~~. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .~~.~:2~~
Welding discontinuity. See also Cold crack; Hot crack; InclusioTl; testing 'tubular product.. .. . .. . ...... . .. .. .... . .. .. .... . .. .... .... .. .. .... 405
Lack offusioll; Lack of penetratiOll; Lamellar tear; Overlap; oil vehicles. . . . . . . . . . . .. .......................................... 206

§:::?>\((((((((<:~;~
Porosity; Ulldercut ............................................... 7-8,84-90
Wet continuous method .................................................... 395
Wet fluorescent method ..................... '" ................ 273, 362, 363
Wet horizontal system ........................ " .................. 57, 352-354
and demagnetization.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 32, 298
Wet method .................................................................... 46

~~~~~~~~i;;;;(i:i~
and subsurface discontinuity indications ..... " .......................... 233
hath .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 208-209
concentration test ........................................................... 239
disadvantages of ............................................................... :34
early use of . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 50
hOJizontal test system ............... " ........ '" ............................ 21
sensitivity .. " ................ " ...................... " ................ 214, 216
specifications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 437 X,Y,Z
sUitability .................................................................... " 37
tape transfer records ....................................................... 273
Wet method particle ~~?~~~~~~~~. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . '2'2~23, .~!
~~b~~~?~ ~:;:~:~~~"e~t . ••••••••••..•. . ••.•.••..••••.••.••.• m
applications. .. . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 212-213
bath .......................................................................... 208
coagulation ........... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 208
composition .... " ................... " ....................................... 205
concentration control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 236
::;i~a~I;:t~r~~~ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ~~'5;~~4~~~
contrast ...................................................................... 206 for longitudinal magnetization ....................................... 159-160

yo:~r::::!.· • · •· •· .•. • • •. .•.•.• •.•.• .•.•.•.•.•. • ,•. ~-~~~: ~

control ......................................................... , ........ 400-401
fluorescent ................................................................... ~. '35
how to apply ............ " ............. : ...................... : .. '" ......... 212
Zinc chromate coating ........................................... : .... 386,387