AN AMERICAN NATIONAL STANDARD

ASME B89.4.1a-1998

ADDENDA
to

ASME B89.4.1-I997
METHODS FOR
PERFORMANCE EVALUATION OF
COORDINATE MEASURING MACHINES

THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS

Three Park Avenue New York, NY 10016-5990

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 Date of Issuance: June 8, 1998

ASME is the registered trademark of The American Society of Mechanical

Engineers.

This codeor standard was developed under procedures accredited as meeting the criteria for
American National Standards. The Standards Committee that approved the code or standard
was balanced t o assure that individuals from competent andconcerned interests have had an
opportunity to participate. The proposed codeor standard was made available for public review
and comment which provides an opportunity for additional public industry, input fromacademia,
regulatory agencies, and the public-at-large.
ASME does not "approve," "rate," or "endorse" any item, construction, proprietary device,
or activity.
ASME does not take any position with respectt o the validity of any patent rightsasserted in
connection with any items mentioned in this document, does not
andundertake to insureanyone
of any applicable LettersPatent, nor assume
utilizing a standard against liability for infringement
any such liability. Users of a code or standard are expressly advised thatdetermination of the
validity of anysuch patent rights, andthe risk of the infringement of such rights, is entirely their
own responsibility.
Participation by federal agency representative(s1or personts) affiliated with industry is not to
be interpreted as government or industry endorsement of this code or standard.
ASME accepts responsibility for only those interpretations issued in accordancewith governing
ASME procedures and policies which preclude the issuance of interpretations by individual
volunteers.

No part of this document may be reproduced in any form,

in an electronic retrieval system or otherwise,
without the prior written permission of the
publisher.

The American Society of MechanicalEngineers

Three Park Avenue, New York, NY 10016-5990

Copyright O 1998 by
THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS
All Rights Reserved
Printed in U.S.A.

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 ASME B89.4.1a-I998

Following approval by the ASME B89 Committee and ASME, and after public review, ASME B89.4.la-
1998 was approved by the American National Standards Institute on March 3, 1998.

Addenda to the 1997 edition of ASME B89.4.1 are issued in the form of replacement pages. Revisions,
additions, and deletions are incorporated directly into the affected pages. It is advisable, however, that this
page, the Addenda title and copyright pages, and all replaced pages be retained for reference.

SUMMARY OF CHANGES
This is the first addenda to be published to ASME B89.4.1-1997

Replace or insert the pages listed. Changes listed below are identified on the pages by a margin note, (a),
placed next to the affected area. The pages not listed are thereverse sides of the affected pages and contain
no changes.
Page Location Change
ix Contents Revised
46 5.5.7.1(g) Revised
5.5.7.1(h) Subparagraphs (l), (2), and (3) revised
5.5.7.3(h) Revised
49,50 6.2.1 Third paragraph revised
57-58.3 Appendix C Revised in its entirety
78-78.2 Appendix I Sections 16, 17, 18, and I9 added

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 35 Diagram Schematically RepresentingtheMeanings of the Radial. Tangential.
andAxial Working Tolerances for theRotaryAxis Performance Test ............ 44
36 TypicalResults of aVolumetricPerformance Test for aDCC Machine With a
Rotary
Axis ................................................................... 45
37 SchematicDiagram Showing the Locations of Probingin the Point-to-Point
Probing Test .................................................................. 48
Tables
I Location of the Reference Sphere on theRotary Table ............................ 40
2 DefaultNominal Angular Positions and Sample Data Sheet for Obtaining
Volumetric Performance WithaRotaryAxis .................................... 42
Appendices
AUser's
Guide to ASME B89.4.1 .................................................. 53
Thermal
B Environment Testing ..................................................... 55
C
CMM Site Vibration
Measurement ................................................ 57
Electrical
D Power Analysis ........................................................ 59
Utility
E Air ....................................................................... 61
F Hysteresis Test Design Recommendations ......................................... 63
G Ball Bar Test Equipment DesignRecommendations ................................ 65
H Straightedge Tests for RamAxisRoll ............................................. 69
I Interim Testing of CMM Systems ................................................ 73

ix

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 S T D - A S M E B B 9 * q - L A - E N G L L778 m 0759b70 Ob0399b T O O

METHODS FOR PERFORMANCE EVALUATION

OF COORDINATE MEASURING MACHINES ASME 889.4.1a-1998

3D/alpha radial error

15, 1

10

\
E
E , 5
ò
6 0

-5

-1 o
""
I I 1 nnm-rnnn
28 24 20 O16 12
4 8
Position

3D/alpha axial error

35 , 3D/alpha tangential error
1
20
1
3oL """"" """"",-I

.
E .
E
5
ì
5
e
L
i
2
w L
W

-15 J 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 l 1 1 1 1 1 1 1 1 1 J -20 1 I I I I I I I I I 1 I 1 I l I I I I I I I l I I I 1 1 1 I
O 4 16
8 12 20 24 i8 O 4 8 12 16 28
20 24
Position Position

FIG. 36 TYPICAL RESULTS OF A VOLUMETRIC PERFORMANCE TEST FORA DCC MACHINE WITH
A ROTARY AXIS
(The 3D/alpha radial, BD/alpha tangential, and BD/alpha axial working tolerances are clearly
labeled on the graphs.)

(d) The load at anyspecific contact pointwillbe (u) Placethetestweight on the machine.
no greater than twice the load of anyother contact point. (6) Perform the repeatability test as described in this
( e ) The center of gravity of the machine load must Standard (para. 5.3), withthe exception of location.
lie within the CG locationzone. Location isoptional in thistest.
The specific test load must fall within acceptable ( c ) Performsixball bar measurements, as physical
machine load limits, as defined by the Load Concentra- constraints allow, selectedfromthe following eleven
tion Chart (Fig. 3). user-selectable positions:
The following steps shouldbetaken for thetest (1) (four) 3D diagonals (as available);
procedure. (2) planar diagonal (front);

45

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
ASME 889.4.1a-1998 OF COORDINATE MEASURING MACHINES

(3) planar diagonal rear (opposite orientation); 5.5.8 Volumetric Performance Requirements.
(4) planar diagonal (top); VolumetricPerformance. as calculated in paras. 5.5.2.
(5) planar diagonal (left side); 5.5.3, 5.5.4 (if applicablc). 5.5.5 (if applicable). 5.5.6
(6) planar diagonal (right side - opposite orienta- ( i f applicable), and 5.5.7 (if‘ applicable) shall not exceed
tion); and the supplier’s specifications.derated as specified in
(7) twoorthogonal linear axes. paras. 4.2 and 4.3, if applicablc.

WARNING: Omission of 3D diagonalsmaypreventseeing the

full effect of loading.
5.6 BidirectionalLength Measurement
Capzbility
(d) Removeweight.
( e ) Repeat (b) above (repeatability test). 5.6.1 General. The precedingtestshaveproduced
cfl Repeat (c) above (ball bar measurements). a meaningfulpictureofthemeasurement system per-
(a) (8) Perform a repeatability analysis: resultsoftests formance; however, some errors, such as undue machine
(b) and (e) shall not exceed the stated repeatability or probehysteresis and improperprobe compensation,
specification. have not been fully analyzed since no two-sided length
(a) (h) Perform volumetric
analysis: measurementhasyet been performed. The following
( I ) range of readings of test (c) shallnotexceed tests removethisdeficiency by requiring the measure-
stated machinevolumetric performance specification; ment of a gageblock of a convenient length, in four
(2) rangeof readings oftest ( f ) shall not exceed positions in themachineworkzone.. Three of these
stated machine volumetric performance specification; positions areroughlyalignedwiththe machine axes,
( 3 ) the difference between a measuredlength in andthefourthposition is user-selectable. It isrecom-
mendedthatthisfourthpositionnotbe aligned with
test (c) and the measured length from the same position
in test ( f ) shall not exceed 50% of the machine volumet- anymachineaxis. The lengthoftheblockshallbe
ric performance specification. within the range ofatleast 25 mm (approx. I in.) to
100 mm (approx. 4 in.), withthe default value being
5.5.7.2 Optional Procedure(LaserorGage 25 mm (approx. 1 in.).The gage block shall be calibrated
Block).Followthe procedure described above using in accordance withtherequirementsof para. 7.3. I.
a laser interferometer, gage block, or other equivalent Before performing these tests, themachine probe shall
device as themeasured artifact. Analyze all data per be calibratedandqualified according to the supplier’s
para. 5.5.7.1. recommendations for normal operation of the machine
5.5.7.3 Rotary Table Machine Procedure. when measuring parts. Qualification on the gage block
For a rotary table machine, the procedure is as follows. to beused for thistestisspecifically excluded. The
(a) Calibrate therotarytable in anunloadedmode. measurementsfor this test are also to be performed
(6) Place weight on the machine in compliance with usingtheprobingparameters,probe approach rate,
the guidelines ofpara. 5.5.7.1 above. probe approach distance, and settling time specified for
(c) Performthe repeatability test as described in normaloperation in Fig. IA.
para. 5.3, withthe exception of location. Location is
optional in thistest.
5.6.2 Measurement Procedure -
Bidirec-
tional Length Measurement. The gage block con-
(d) Perform the volumetric performance test for DCC
forming to the requirements of para. 5.6.1 above shall
machines with a rotary axis (para. 5.5.6) using positions
be rigidlymounted in thework zone of the machine
listed in column A l of Table 2.
on a fixture that allows probing access to the faces of
(e) Remove weight. the gage block for thefour measurement positions in
cfl Repeat (c) above (repeatability test). turn. The mean temperature of the gage block and the
(g) Repeat (d) above (volumetric performancetest). appropriate machine scale(s) may be measured during
(a) (h) Analysis: results oftests (c), (d), (0, and
(g) this gage block measurement process for each position,
shall not exceed the stated machine performance speci- using a thermometer conforming to the requirements
fications for repeatability, radial, tangential, and axial of Section 7. The exact location of the gage block in
(3D/alpha) error. thework zone isnot critical; however, it isrecom-
NOTE It is recommended that a weight with simple geometric form mendedthat this positionbe near the location in the
be used for testing purposes to reduce potential difficulties in calculat- work zonewhere parts will most commonly be mea-
ing the CG location. sured. Aftermountingand alignment, which maybe

, 46

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
OF COORDINATE MEASURING MACHINES ASME B89.4.1a-1998

center is approximately 100 deg. from the pole of the -

6.1.3 Probe Approach Tests Optional. Many
testball in a directionparallel totheshank attached machinedprobe systems exhibit vastly different charac-
tothestylusball,12 equally spaced on the equator teristics depending on the probe approach distance and
with the pattern rotated about the stylus shank 10 deg., the probe approach rate. For the machine user desiring
12 equally spaced around the ball with the stylus center tousemorethan one value of these parameters, this
approximately 60 deg. from the pole and rotated about test of the machine performance is recommended. The
thestylusshank an additional 10 deg. relative to the procedure isthesamegiveninparas.6.1.1and6.1.2,
previous pattern, 12equallyspacedwiththe stylus except thatthistest is performedfortwodifferent
center approximately 30 deg.fromthepolewiththe probeapproach distances andprobe approach rates.
pattern again rotated an additional 10 deg., and finally, The working tolerance for point-to-point probing is
one on the pole of the test ball. This situationis depicted specified for each of these options.
in Fig. 37,- inwhichthedifferent probe positions are
shownwith dashed linesandlabeledpositions 1 to 5. 6.1.4 Point-to-Point ProbingPerformanceRe-
n e default test for manualmachines is the measurement quirements. Point-to-Point probingperformance- as
of 49 points distributed as uniformly as practical Over calculated inparas. 6.1.2 and 6.1.3 (if applicable),shall
the measurable portionofthetestball. not exceedthesupplier’sspecifications, derated as
On direct computer-controlled machines,theprobe ’Pecified in para*4.3y if applicable.
shall be vector-driven towardthetestball centerfor
each touch, providedthis is normal for the machine 6.2 Probing Analysis - Multiple-Tip Probing
when measuring parts. Ondrivenmanualand free- In addition to the probing errors highlighted in para.
floatingmanual machines, wherepossible, one axis 6.1.1, CMMs that use multiple stylus tip positions can
should be locked andtheremainingaxesmoved to haveadditionalerrors. These errors can be due to a
contact the ball in order to accurately hit the test ball. number of sources including the uncertainty in location
Inall cases, thesupplier’sprobeapproach distance, ofeachofthetips caused by tip calibration errors or
probe approach rate, and settling time, as given in Fig. bytheerrors associated withtheuseofanorienting
lA, shallbeused. head or probe changer. This is true for allmultiple-
tipsystemconfigurations,including:
6.1.2 Data Analysis - Point-to-Point Probing. (a) systems using multiple styli connected to the
From each set of 49 readings for each stylus, a sphere CMM probe, such as star clusters;
center is computed usingthesupplier’srecommended (b) systems using orienting heads;
algorithms. From this center a radius is then determined (c) systems using probe or stylus changers; and
for each measurement point.Theminimum radius is ( d ) systems using heads with multiple probes.
subtracted fromthemaximumradiustoproducethe The commonelement of these systems is that different
point-to-pointprobingperformance for each of the tips or tiplocations are usedto inspect a workpiece
styluslengths. Iftheresultobtainedfor a particular without any recalibration of the tips. As a result, it is
stylus is less thantheworking tolerance for the test, important to understand any additionalerrors which
then the testing is discontinued for that stylus and the might be contributed by these systems.
resultreported. If theresultforanystylus is greater
thanthe working tolerance,then the testshall be 6.2.1 Method of Test - Multiple-Tip Probing. (a)
repeated. If the new results agree to within the working The calibration ball diameter and all system configura-
tolerance for repeatability(para. 5.3), thenthe second tion dimensions in this Section are default values. Other
set of data is discarded andthefirstsetused for the dimensions may be substituted and it is recommended
analysis. If they donotagree,then a third set shall thatthisbedone if there is any concern thatthe
be taken. If this agrees with either of the two previous configurations required to measure actual workpieces
sets, then the first of theagreeingsetsshallbeused aresubstantiallydifferent from the default values.
in the analysis. Ifno agreement to within the working A precision reference ball conforming to the require-
tolerance for repeatability is obtained after three mea- ments of para. 7.3.3 shallberigidlymountedonthe
surement sequencesfor anygiven stylus, thetest is workpiece supporting surface intheworkzoneofthe
discontinued andthefaultdeterminedandcorrected. machine on a fixture that allows access by the machine
After correction, all of the tests describedin this section, probing system. The 6 mm (approx. 0.25 in.) diameter
even those for styluslengthsthatwere previously in test sphere used in the point-to-point probing test ( S e c -
tolerance, shall be repeated. tion 6.1) maybeused for this test. Any position may

49

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 S T D * A S f l E B 8 7 - q - L A - E N G L L998 H 0759b70Ob03779 7LT D

METHODS FOR PERFORMANCE EVALUATION

ASME B89.4.1a-1998 OF COORDINATE\MEASURING MACHINES

be chosen for thismountingwith the default position 6.2.3 Multiple-Tip ProbingPerformanceRe-

being the TVE positionasspecifiedinFig. 1. quirements. Multiple-tip probing performance, as cal-
Five different probing tip positions shall beused to culated in .para. 6.2.2, shall not exceed the supplier’s
perform thistest. These positionscanbecreatedby specifications, derated as specified in para. 4.3, if appli-
using a stylus configuration with five tips, five different cable.
orientations ofan orientinghead, or through the use
of a probe or styluschanging system using five different
tip positions. Two of theprobetippositionsshallbe
on a line perpendicular to theram axis.Two more 7 TEST EQUIPMENT
shall be onanothersuchline displaced 90 deg. The
fifth position shall beon a lineparalleltotheram 7.1 Temperature
axis through the intersection of the first two lines. The The time constant of thermometers shall be no more
default stylus length, including any extension members, than one-tenth the cycle time of the highest frequency
shall be 50 mm (approx. 2 in.) when using any of the component ofthe temperature variationofinterestin
above systems or combination of the above. a test. The time constant is the time required for the
The user is allowed to specify anytestpatternthat thermometer to indicate 63.2% of its final change due
contains 25 points. These 25 points shall be probed to a step change in temperature.
on the test ball as equally spaced as possible and cover The resolutionof thermometers needbenogreater
as much of thesphere surface aspractical.The 25 than one-tenth the amplitude ofthe lowest-amplitude
points shall betaken using five different tipsortip component of temperature variation of interest in a test.
locationsand each set offive points probed by each Thermometers shall be calibrated by suitable means
tip shall also be as widespread as possible. As an to an accuracy of +0.loC over the temperature range
example, these five points could be four points around of use.
the equator ofthe sphere (assuming the pole position
is directlyinlinewith the stylusshaft supporting the
tip) plus a point directly in linewiththestylusshaft. 7.2 Vibration
For the purposes of this Standard, relativemotion
shall be measured using a high-resolution,undamped
displacement indicator. Resolution of 0.1 Fm (approx.
6.2.2 Data Analysis - Multiple-Tip Probing. 0.000004 in.) or better is recommended.
From the set of 25 readings, a sphere center is computed
using the supplier’s recommended algorithm. From this 7.3 Displacement
center aradius is then determined for each measurement
point. The minimum radius is subtracted from the
maximum radius toproducethemultiple-tip probing 7.3.1 Gages. Step gages andgage blocks shall be
performance. If theresult obtained is less than the calibrated to within one-fifth the working tolerance for
working tolerance for thetest, then the result is reported. the repeatabilityspecified for theCMM.Indicating
If the result is greater than the working tolerance, then gages shallhave a resolution of nomorethanone-
the test shall be repeated. If the new result agrees with fifth the working tolerance for repeatability. All gages
the result of the first test within the working tolerance shall be calibrated following the supplier’s recommenda-
for the repeatability (para. 5.3), then the second set of tions.
datais discarded and the firstset is used forthe
evaluation. If they do not agree, then a third set should 7.3.2 Laser Interferometer. A laser interferome-
be taken. If this agrees with either of the two previous ter conforming to the requirements of this Standard
sets, then the firstof the agreeing sets shall be used shall have a frequency stabilitysuchthat this long-
in the evaluation. If no agreement to within the working term stability represents an error of less than one-fifth
tolerance for repeatability is obtained after three mea- the working tolerance for repeatability of the machine
surement sequences, thistest is discontinued and the (in meters), dividedby the length of the longest machine
fault determined andcorrected. After correction,the axis (inmeters). The resolution of such a system
repeatabilitytest(para. 5.3.3) and all ofthe tests shall be better than one-fifth the working tolerance for
described in thissectionshall be repeated. repeatability.

50

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 APPENDIX C
CMM SITE VIBRATION MEASUREMENT
(This Appendix is not part of ASME 889.4.1-1997 and is included for informationpurposes only.)

C l SCOPE the type of criteria, the amplitude ordinate canbe

definedin either thetime domain or thefrequency
The purpose of thisAppendixis to recommend
domain.
vibration measurement instrumentation and procedures
for measuring vibration at CMM installation sites. Vibra- C3.1.1Ordinate Units. Since theCMM is a di-
tion levels should be measured at the proposed CMM mensional measurement tool, units of displacement are
site(s) tocompareto allowable site vibration limits most useful in relation to CMM performance. However,
established by the CMM supplier. This Appendix also velocity and acceleration are alsoappropriate parameters
defines theinstrumentationand suggested procedures for measuringCMM site vibration.
to establish vibration on the CMM for additional analy-
sis. This Appendixdoesnot,however, address the (3.1.2 AbscissaUnits. The useof time or fre-
determination of vibration sources or the reduction of quency for the abscissa will depend on the acceptance
vibration levels. Such determination is usually involved criteria format of the CMM supplier. Time based criteria
and requires theknowledge of vibrationspecialists. are referredto as a Time History,whichprovides
measurement of transient or very low frequency vibra-
tory events such as beat signals. The frequency domain
C2DEFINITIONS allows measurement over a veryshort time range,
This Appendix is intended to be self-defining and is which provides an ability to diagnose many dynamic
written for individuals with an engineering background. events.
Definitions for specificvibrationterminology may be
found inIES-RP-CC024.1,Measuringand Reporting C3.2 Format
Vibration in Microelectronics Facilities,published by As defined in B89.4.1, the supplier provides, as part
the Institute of Environmental Sciences. of the machine specification, a statement of acceptable
vibration. This criterion should be provided by the
supplier, or listed as part of the CMMspecification
C3VIBRATION ACCEPTANCECRITERIA
form, if used. At least two criteria format options are
The CMM supplier is to provide site vibration levels presented: Frequency Function and TimeHistory.
of acceptability. Below these levelstheCMMcan The supplied acceptance criteria will definethe format
operate successfully,andabove these levels problems inwhichto present thevibrationdata for ease of
may occur. Each CMM manufacturer hasdifferent comparison.
formats and levels of acceptance. The type of vibration
measurements to be takenwilldependon format and C3.2.1 FrequencyResponseFunction.This
vibration unitsused by theCMM supplier.Basedon type of information is specified as a vibration amplitude
the type ofcriteria,thevibrationspecialist should as a function at specificfrequencies.The criteria are
determine the necessary measurement units, frequency usually presented as allowable vibration amplitude ver-
range, measurement locations,andinstrumentation. sus frequency in Hertz. The frequency range may vary
from supplier to supplier. In general, seismic vibrations
are applicable over a rangeof O (DC)to 100 Hz.
C3.1 Units
Vibrationlevelshavelarge dynamic range, and it is
Vibration is characterized by amplitude versustime sometimes helpful to present amplitude data in logarith-
or frequency. The amplitude can usually be defined in mic scale. If decibels are used,thestandard reference
displacement, velocity, or acceleration. Depending upon values must also beused.

57

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 C3.2.2 Time History. These measurements repre- ing, depending upon the sensitivity and signal-to-noise
sent thevibration during thetimeperiodofinterest. ratio. It is the responsibility of the vibration specialist
The supplier should specify a maximum peak-to-peak tousethe proper signalconditioners.
acceptable vibration level and a time period over which
it applies. The vibrationlevekcan be
in unitsof C4.3 Signal Recording/Analysis Instruments
displacement, velocity, or acceleration.
The type of instrumentation to usewill depend on
(3.2.3 As an alternative, theCMM supplier may the type of criteria and format that have been provided
choose to evaluate vibration related CMM performance by the CMM supplier. The frequency function criteria
degradation on the actual installation site, and compare requires a Fast Fourier Transform (FFI'), a Dynamic
it with an acceptable levelas a basis. Signal Analyzer, and, in some cases, a Digital Recorder.
Time History data can be acquired with an oscilloscope,
C4 INSTRUMENTATION a digitalrecorder, a FFT analyzer, ora frequency
analyzer.
This Section describes various instruments required
to perform on-site vibration measurements. Various C4.3.1 FFT Signal Analyzers. This typeof ana-
types of sensors, signal conditioners, recorders, com- lyzer offers the most sophisticated means of measuring
puter programs, and signal analyzers are available for vibration, by providing the greatest amount of informa-
use in acquiring this data.It is notthe intent of tion about thevibrationsignal. Inmost cases, this
this Section tosingle out any particular equipment additional information is necessary to understand the
manufacturer, but to recommend types of equipment vibration environment. Many types of FFT analyzers
whichmeet the requirements of this Appendix. exist, from many different manufacturers. One and two
channelunits,hand-held, and PC-based areformats
C4.1Transducers readily available. It should be noted that using a data
recorder as specified below will require the use of an
Many types of transducers existfor various types FFT analyzerafter thedata are acquired. It is the
of vibration measurements. The measurements specified user's responsibilitytounderstand their instrument's
in this Appendix require a specific accelerometer or a capabilitiesandlimitations.
specifictype of velocitytransducer. The following section offers guidelines for FFI:
C4.1.1SeismicAccelerometers. The two most analysis configuration andspecifications.
important requirements for the accelerometer are fre- (a) Noise floor. -100 dBV/rootHz.
quency response and sensitivity. Site vibration measure- (6) A/D resolution. The resolution of the analog-to-
ments generally require low frequency and high sensitiv- digital converter should be at least 12 bits. The better
ity. The minimum frequency response linearity should analyzers willhave 16 bit A/D resolution.
be lessthan 1 Hz, preferably 0.5 Hz. The maximum (c) Llynamic range. The dynamic range should be
frequency response should be greater than 100 Hz. The at least 70 dB. Better spectrum analyzers willhave
sensitivityof the accelerometer should be 10 Voltdg higher dynamic range.
or greater, where g is equal to 9.8 m/sec2 (386 in./sec2). (d) Frequency resolution. This parameter as it applies
C4.1.2 Velocity Transducers. These sensors are tothe analyzer is denoted in number of lines over
also referred to as geophones. The sensitivity ofthe which the analysis range is divided.Most analyzers
geophone should be 0.4 V/mm/sec (10 Volts/in./sec) can have selectable resolutionfrom 100 lines to 1600
or greater. The frequency response linearity requirement lines. The resolution in Hertz is calculated by dividing
of the velocity transducer is the same as the accelerome- the frequency range by the numberof lines. For example,
ter, 0.5 Hz to 100 Hz. a O to 100 Hz frequency range acquired with a 400
lineanalysiswill have 0.25 Hz (100/400) 'resolution.
C4.2 Amplifiers and Signal Conditioners The frequency resolution used must be compatible with
the resolution of the frequency response" criteria. If the
The transducers require amplifiersandsignal condi- criteria are defined at every 1 Hz,the data must be
tioners.Most seismic accelerometers require an ampli- acquired with a 1 Hzresolution. The overall frequency
fier,
but some models may have
built-in electronics resolution will also be dependent on the transducer
that do not require signal conditioning. Velocitytrans-frequency response. This information should be com-
ducers may require amplification and signal condition- pliedwithandmodifiedonlywhentheCMMmanufac-

58

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 turer'sspecification requests otherwise. determining the peak-to-peak voltage amplitude, and
(e) Anti-aliasing jlrer. This filter prevents incorrect using the transducer sensitivity for converting to appro-
reporting of frequency components due to under sam- priate amplitude units.
plingof higher frequencysignals. This filter is found
on most (if not all) FFT analyzers. It should always C5 TESTPROCEDURE
be used.
f'JAveraging. Most analyzers havethisfeature. The procedures for makingvibration measurements
When used, itreducesthe effects of transientevents are fairlysimple once the appropriateanalysis equipment
such as personnel or vehicular activity.It is recom- isselectedandconfiguredasrequired.
mended that 10 averages be taken for all measurements.
Some spectrum analyzers have various types of averag- C5.1Calibration
ing functions such as linear,rms, peak hold, or exponen- At a minimum, the vibrationmeasurement equipment
tial. Linear or summation averaging should beused. should have been calibrated by a qualified laboratory,
(g) Window functions. This feature is used to force traceable to NIST, in the past 12 months. Site calibration
a generalized vibration signal into discrete time domain comparison testing ofthe transducers atthe start of
periods.Whenwindow functions are not used,the the testing is required.
frequency response of the vibratory signal is incorrectly
distributed throughout the frequency range. There are C5.2 Transducer Mounting
many types ofwindowfunctions. The mostpopular
For all measurements,the transducers should be
are Hanning, flat top, anduniform. The Hanning window
mounted directly and firmly to the floor or a common
provides the best compromise in amplitude andfre-
interface for measuring three mutually orthogonal axes.
quency accuracy. Other windows provide excellent am-
Such mounting arrangements are referred to as triaxial.
plitude accuracy and poor frequency accuracy, and vice
Some transducers incorporate three mutually orthogonal
versa. The Hanning should be used for all measurements
axesinonedevice.Whenthismounting mangement
specifiedinthis Appendix.
is used, all three channels should be acquired simultane-
C4.3.2 Data Recorders. For ease of gathering ously. Time independent triaxial measurements should
vibration data inthefield,theuseof a multi-channel not be performed because simultaneous orthogonal re-
data recorder is foundto be useful and convenient. sponses willnot be achieved.
Such an instrument allows for three or more channels In case of measurements of floor tilt motions (rock-
of data to be recorded simultaneously while providing a ing), two sets of sensors are mounted at a designated
permanent record for archives and verbal data annotation distance for simultaneous measurements intwo orthogo-
during specific events. Additionally, the recorder allows nal verticalplanes.
a record of the real time response which can be most
useful. The data canthenbe processed at a later date C5.3 Measurement Location
using in-house data reduction techniques such as FFT In general, the transducers should be mounted in the
analyzers or signal analyzers specified in this Appendix. general area where the CMM will rest. This area should
The recorder formatmustbedigital and useDigital encompass the outer envelopeofthe CMM plus 3 m
Audio Tape (DAT) because of the excellent signal (10 ft) beyond this foot print on a uniform floor surface,
tonoiseratio and dynamic range as compared to or at the CMM support positions.
analog tape.
C5.4Acquiring/Recording Data
"3.3 Oscilloscopes. This piece of general labo-
ratory equipment may be easily obtained to make an Vibration measurements should be made during nor-
initial set of TimeHistoryreadings.Mostfacilities mal operations of the facility.Nearby equipment that
have an oscilloscope and personnel whocan operate willbe operating whenthe CMM is expected to be
the equipment, allowing users to take baseline readings inuse should berunningduringthevibrationtesting.
for themselves. The oscilloscope isalso usefulfor A written test log or voice channel on a data recorder
viewing beat signals,transient events, andhourlyand should be maintained by the individual performing the
daily vibratory changes. The oscilloscope should be set test so thatanyabnormalevents,such as temporary
to AC coupled and free run triggering. The vibration conditions resulting from construction, repair work, and
amplitude is determined by viewing thesignal and the like, may berecorded during thetest. A test

58.1

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 should be repeated if an abnormal event occurs. Normal work is required.It is the sole responsibilityofthe
vehicular trafficshouldnot be excluded.Whenthe supplier to maintain the performance of the CMM in
environmentalconditionsaresatisfactory,thedata order to meet specifications.
should be recorded on tape, saved to memory, printed,
or manually recorded.
C6.2 Measured Vibration Above Criteria
C5.5Comparing VibrationData
If, on theother hand,thevibrationlevels exceed
After the data acquisition and analysis are complete, the supplier’sspecifications,it is theresponsibilityof
the data must be compared to the vibration acceptance the user toisolate thevibration in orderto conform
criteria. to thespecification orelse accept a performance de-
rating as described in ASME B89.4.1. Again, this
C5.5.1 Time History. For Time History criteria, Appendix doesnot provide information onhow to
this simply involves comparing the measured peak-to- reduce excessive vibration levels, but vibration isolation
peakvibration levels tothepermissiblelevel. The will reduce the levels. Before the vibration levels
CMM supplier may provide horizontal, vertical, linear, can be reduced, thesource of the vibration must be
and. angular criteria. It is important to compare the determined. Itmay be easy to do thiswith the above
acquired data to the criteria in the appropriate direction. equipment. Shock and vibration isolator suppliers spe-
C5.5.2 Frequency Response Function. Compar- cializing in low frequency vibration attenuation should
ison of Frequency Functioncriteria to frequency domain be contacted if vibration isolation or a vibration survey
vibration data can be more effort than taking the data. is required.
If thecriteriahavethe same level at all frequencies
(straight line) or little changes in amplitude, it will be
easy enoughto draw the criteria over the printed C7 REPORT
vibrationlevels. Ifthe criteria are not constant or
uniform, itmay be easier to compare data and criteria A report should be issued by the vibration specialist
with various software programs. This involves digitizing within approximately three (3) weeks. The report should
thecriteria,whichinsome cases requires entering include all backup information and analyzed data with
levels at 1 Hz increments. The vibration data stored a comparison tothe CMMspecification, and include
on the FFT analyzer must be down-loaded into a PC. the following as a minimum: Title, Dates (issued and
This requires different steps depending on the analyzer when data was taken), Contract Number, Revision/
manufacturer. Using a spreadsheet, math, graphing, or Revision Date, Purpose or Scope, Instrumentation Used,
specialprogram,thevibration data and criteria are Calibration Information, Description and Diagram of
combined into asingle graph. Once the data is in a TestSetup, Procedure, Analysis, and Summary. It is
software format, it can be manipulated, graphed,and important to note that the report should serve to archive
analyzed in a usable format. the baseline vibration data for later review, if problems
arise after CMM installation.
C6SUGGESTEDCRITERIA ASSESSMENT

C6.1 Measure Vibration Below Criteria C8 FIELD INSTRUMENTATIONDIAGRAM

If thevibrationlevelsmeasuredbythe procedure A diagram of typical instrumentation is shown in
above are withinthesupplier’s criteria, no additional Fig. C-l.

58.2

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 X input

Y input Digital Signal

recorder analysis Printer

Z input

Option A

X input -J I

Option B

Z axis

X axis

1
conditioned output

“W To option

I
Yconditioned
axis output
A or B

Lb Zconditioned
axis output

Sensor Diagram

FIG. C-1 FIELD INSTRUMENTATION

58.3

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 f5
P .-
'81 Upper threshold I

IOI
8
6- n Upper threshold
4 - 0 0 0 0 ° 0 0 0 0 0 v

2- O 0 O v
O
o = o -" o o o , 0 0 :
v
O 0 O 0 0 0 o

FIG. 15 SUMMARY PLOTS OF SEVERAL INTERIM TESTRESULTS

Test number

FIG. 16 SUMMARY PLOT OF COMBINED INTERIM TESTINGRESULTS

balllocationrelativeto others withdifferent probe Figure I5 shows one possible method of data analysis
head orientations. Thefinalbody diagonal position for the interim test. For each interim test, all four center-
checks for any defective probes present inthe probe to-center length deviations, all eight ball diameters, and
rackandtherack'sprobe changing ability. The first the eight measured sphere form errors are plotted. The
ball of theballbar in thisposition is measured using test is passed if all these measurements are within the
the second probe obtained from the probe-changing threshold value limits. Some users may prefer a single
rack,andthe second ball of theballbar is measured plot representing thetestresults(instead of the three
withthefinal (#3) probefromtheproberack. The shown inFigs. I4 and 15). Such a plotcaneasilybe
form error and diameter, reported for each ball of the constructed,asshown in Fig. 16, by combining the
ball bar, test each ofthetwoprobes for probe lobing largest length deviation, the largest diameter deviation,
effects and stylussize calibration,
respectively.
(If and one half the largest form deviation, in a root sum
additional probes are available, these could be checked of squares (RSS) manner. (One-half thelargestform
by measuring each ball of the ballbar, in each ball deviation is used so each of the three contributions is
barposition,with a different probe.) appropriately weighted). This method has the advantage

77

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 of displaying only asingle graphbut provides less be difficult to fulfill with large CMMs, as the artifacts
information as to the sources of error. (If a CMM may become unwieldy, expensive, and difficult to Cali-
problem does develop, plotssuchas those in Fig.I5 brate.Furthermore, large interim testing artifacts may
could be constructed using data from the previous test require special fixturing to avoid distortions caused by
results.) gravity or the probing force of the CMMs. Since these
There are many different methods a user can choose distortions often increase as thecubeofthe artifact’s
to establish testing thresholds. These include using the length, acceptably small distortions on short artifacts
supplier’sstated CMM performance values for the canrapidlybecomesignificant error sources as the
particular CMM underconsideration,whichmightin- length of the artifact increases. Consequently, fixturing
volve specificationsfromthe ASME B89.4.1 or other which minimizes these effects is highly recommended.
appropriate national or international Standards. Other For example, when using a ball bar as the test artifact,
methods to determine the thresholds include examining a fixturingsystem such as the one showninFig. G4
the tightest tolerance of a feature found on the user’s is preferred to the free standing design shown inFig.
workpiece and reducingthis by an appropriate ratio. G1.Finally, thermal effectsare especially important
To avoid false alarms,the threshold levels should on largeartifacts. The magnitude of these errors can
exceed allvariations arising from normal operations. be estimated by the Nominal Differential Expansion
This may include suchfactorsas different operators (NDE), and the uncertainty in the NDE (Le., the UNDE,
and differentthermalconditions, e.g., time of day see para. 4.2).
or week. The following recommendations provide alternative
waysof overcoming the testingdifficulties of large
15 TESTING FREQUENCY CMMs.

The frequency of interim testing is highly user- 16.1 Subwork Zones

dependent. A CMM being operated three shifts a day
Since some large CMMs use a significantfraction
with multiple operators in a harsh environment is likely
of their work zone for part mounting, a smaller work
to experience manymore problems than the same
zone (or series of smaller work zones) might be used
machine being used one shift a day by a single operator
for the actual measurements (see work zone in glossary).
in an excellent environment. The frequency of testing
In these cases, thetesting artifact may comply with
is also strongly affected by balancing the cost of interim
the recommendation of using the length equal to 75%
testing against the consequences of accepting a bad
of the shortest axis of the subwork zone. An example
workpiece or rejecting a goodone.Itmay be useful
ofsuch a situationwouldbe a CMM whichinspects
to consider the interim testing interval as a percentage
physically largeparts that needto fit intothe work
oftotal CMM operating hours.Some users withhigh
zone butwiththeactual measurement region on the
value and/orsafetycritical workpieces may elect to
part being a small subvolume of the part’s physical
perform dailytests; other users might testweekly
size. Accordingly, a 0.9 m ball bar can easily be used
or monthly.Additionally, interim testingshouldbe
to test a measurement work zone having a 1.2 m length
conducted after anysortofsignificant event such as
side.Similarly, artifacts of length 1.5 m can beused
a C” collision, replacement of a subsystem compo-
to test measurement work zones having a shortest axis
nent, or the occurrence of abnormal temperature varia-
of upto 2 m. Artifacts greater than 1.5 m become
tions or gradients.
increasingly problematic, afact which represents the
limit of practically implementing this approach.
(a) 16 LARGE CMMS
16.2 Artifact Staging
CMMs with large work zones that are approximately
cubical (1 x l x 1)should follow Appendix I withthe For very large CMMs, withthe shortest work zone
following supplementary information. (For largeCMMs, axisgreater than 2 m, large physical artifacts may
approximately cubical work zones can include all cases become impractical. In this situation a reasonably large
where the ratio of the work zone’s longest to shortest artifact (e-g., 0.9 to 1.5 m) canbe staged in the work
axis is lessthan 2.) Appendix I recommends that a zone. The staging should cover a distance of at least
general purpose interim testing artifact should have its 75% of theshortest axis oftheworkzone. It is not
length at least 75% oftheshortest axis of a CMM recommended tostagethe artifact more than three
with a nearlycubicalworkzone. This condition may times since the artifact’s length relative tothe work

78

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 zone size is small in thissituation;hence,itloses plitter. The problem is easily avoided by using a good
sensitivity to angular errors (as explained in Appendix quality optic thandoesnot bend thetransmittedlight.
I) in addition to becoming very time consuming. Using Additionally, the correction for environmental effects
this strategywith a 1.5 m artifact allows testing of a on the wavelength of light over the measured distance
cubical work zone CMM with anaxis of up to 6 m. should be considered a potential error source (see para.
5.4.3.3).Similarly, the use of optical coordinate systems
(e.g.,lasertrackers)must have a sufficientlysmall
16.3 Testing With Optical Systems
system uncertainty relativeto the CMMundertest.
For CMM work zones with a shortest axis of more Since mostoptical systems used for interimtesting
than 2 m, the use of an optical displacement measuring do not involve the CMM probe or related subsystems,
system (e.g., a laser interferometer) maybe desirable. additional tests are needed to check these systems. A
If optical measurements are taken in nonstandard envi- testsphere,calibrated for form and diameter, can be
ronmental conditions,then the wavelengthcorrections employedto check theCMM probe, indexableprobe
of para. 5.4.3.3arerecommended. Additionally, long head,andCMM probe/stylus changing systems. For
beam paths may havespatialgradientspresent; this example, if all of the above subsystems are available,
effect should be assessed and reduced(e.g.,by air then asimple testwould be to measure a calibrated
mixing with fans if necessary). The use ofan optical sphere with a set of points taken using a combination
system canemploythe same procedurerecommended of different probedstyli (accessed thorough probe/stylus
for physical artifacts(i.e., the measurement ofbody changing). anddifferent probe headindexpositions.
diagonals) with at least one length being recorded for This collectionof points is (least squares) fittedto a
every 2 m of displacement traveled.For example, a sphere and theresulting form and diameter errors
CMMwith a 4 m x 5 m x 6 m work zone could be examined. The sphere’s diameter error is abidirectional
tested along thebody diagonal withatleast 3 m of length testand checks the probe’s calibration for features
distance checked (75% of 4 m), and with at least one of size (see para. 5.6), whereas the form error checks
intermediate pointrecorded. Since formostoptical the probe lobing of the different probes (see para. 6.1),
systemsthe measurement time is a smallfraction of and the index positions relative to each other (see para.
the setup time,addingadditionalmeasurementpoints 6.2). Additionally, if the CMM has a part temperature
is advisable (e.g., in the above situation a measurement compensation system, also known asanAutomated
ofthebody diagonal linesof 7 m withthe points Nominal Differential Expansion (ANDE) compensation
spaced at 1 m intervals would be desirable). For large system, this will not be tested during the optical mea-
CMMs that are not vector driven (i.e., cannot operate surement and should be checked independently; for
all 3 axes simultaneously), it may be impossible to example, by measuring a reasonably long calibrated
maintain thenecessaryoptical alignment requiredby artifact having nonzero expansion coefficient. During
thelaserinterferometer. For these CMMs,anoptical this measurement, the temperature of the artifact should
tracking system (e.g.,lasertracker)canmaintainthe bemeasuredwiththeCMM part sensor and used for
optical alignment asthebody diagonals are traversed the ANDE correction. Deviations between the thermally
and maybeused. compensated measured value and thecalibratedvalue
Care must be exercised to ensure thattheoptical for the artifact length may indicate problems with the
measurement system has a sufficiently low uncertainty compensation system.
relative to the CMM under test. If it becomes necessary
to movethe beamsplitterhemote interferometer rather
17 CMMS USED IN THE
DUPLEX MODE (a)
than theretroreflectorwhenmakinglengthmeasure-
ments, problems can arise if the beamsplitter is imper- For CMMs used in the duplex mode the procedures
fectly made and bends the transmitted light slightly. described inAppendix I can be usedwith atleast
Under these circumstances it is never possible to obtain some of theartifact measurements taken under the
good alignment of the beam with thedirection of duplex condition. This is achieved by measuring oppo-
motion; it thelaserbeam exiting the beamsplitter is site ends of the test artifact (ball bar, step gauge, gauge
well aligned withthedirection of motion, then the block, etc.) with different arms of the CMM. Similarly,
incoming beam will be misaligned and will walk across if a ball plate (or hole plate) artifact is employed, then
the face ofthe beamsplitter asthe beamsplitter is approximately half of the balls (holes) may be measured
translated. Thus a potential for bothsignalloss and with each arm. If theCMM is rarely usedin the
misalignment errors exists when translating the beams- duplexmode,then each a n n maybeinterimtested

78.1

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 independently and a fewadditional duplex measure- straightness of a long narrowpart.Inthis case,a
ments included.For very largeCMMsused in the special purpose test designed around the measurement
duplex mode,theuse of alaser interferometer (or requirement may be appropriate. In the above example,
similar optical system) is recommended.Inthiscase, theuseof a straightness interferometer together with
the retroreflectorismounted in the ramof one arm subsystem (e.g., probe) tests may be sufficient for the
and the interferometer is mounted intheramof the measurement application. In other situations the use of
second arm.(See the precautions above regarding testing twoball bars maybe sufficientto check theCMM.
with optical systems.) The distance betweenthetwo For example, one long bar could be oriented along some
CMM arms is varied along a common direction deter- combination of body diagonals, long face diagonals, and
mined by thelaser beam path. If such anoptical the long axis of the CMM. A second shorter bar could
procedure is used,thenthetesting of the subsystems be oriented along some combination of the short face
(e.g.,probehead)is also needed,as described in diagonals and the short axes of the CMM.
Section 16.
1TABLE
9 ROTARY CMMS (al
(a) 18 HIGH ASPECTRATIO CMMS
CMMs having a rotarytablecan be tested by an
CMMs having workzoneswiththe ratio of the abbreviated form of the 3D/alpha test described in para.
longest to shortestaxis(the aspect ratio) greater than 5.5.6. In cases wherethemeasurement volume of
2 may require modified testing procedures. For CMMs interest is approximately that of the rotarytable,the
with aspect ratios of 5 3,and having body diagonals two ballsetup of Fig. 33, with a minimumof four
lessthan 4 m long,interimtestingcanbe performed angular positions selectedfrom Table 2, is sufficient
using an artifact at least one thirdthe length of the to check the CMM. In situations wherethe measurement
body diagonal. For example, a CMM with axes of 0.5 volume is substantially larger thanthat accessible to
m x 1 m x 1.5 m has a body diagonal 2 m long, thus the rotarytable, additional measurementsusing a method
a minimal length testingartifact would be 0.7 m. CMMs previously described (e.g.,measuring a fixed length
with aspect ratiosgreaterthan 3 are usually designed artifact) are recommended. Note that part loading effects
fora special purpose; for example, measuring the can significantly affect the results of the 3D/alpha test.

78.2

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 ASME Services
ASME is committed to developing and delivering technical information. At ASME’s Information Central, we make
every effort to answer your questions and expedite your orders. Our representatives are ready to assist you in the
following areas:

ASME Press Member Services & Benefits Public Information

Codes & Standards Other ASME Programs Self-Study Courses
Credit Card Orders Payment Inquiries Shipping Information
IMechE Publications Professional Development SubscriptionslJournaldMagazines
Meetings & Conferences Short Courses Symposia Volumes
Member Dues Status Publications Technical Papers

How can you reach us? It’s easier than ever!

There are four options for making inquiries* or placing orders. Simply mail, phone, fax, or E-mail us and an Information
Central representative will handle your request.

Call Mail Fax-24Toll Free hours hours E-Mail-24

ASME US i? Canada: 800-THE-ASME 973-882-1717 Infocentral
00BoxDrive,
Law 22 (800-843-2763) 973-882-5155 @asme.org
Fairfield, New Jersey Mexico: 95-800-THE-ASME
07007-2900 (95-800-843-2763)
Universal: 973-882-1 167

* Information Central staff are not permitted to answer inquiries about the technical content of this code or standard.
Information as to whether or not technical inquiries are issued to this code or standard is shown on the copyright
page. All technical inquiries must be submitted in writing to the staff secretary. Additionalprocedures for inquiries
may be listed within.

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
COPYRIGHT American Society of Mechanical Engineers
Licensed by Information Handling Services
 STD.ASME B87.9-3-ENGL 3777 B 0757b70 0583028 893 II

The American Society of

Mechanical Engineers

A N
A M E R I C A N A T I O N A S
L T A N D A R D

METHODS FOR PERFORMANCE

EVALUAT
O
INOF COORD NIATE
MEASURNIE
MACHNIES

ASME 889.4.1 -1997

(Revision of ASME B8N.1.12M-1990)

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 S T D - A S M E B B S - L I * L - E N G L L777 0 7 5 9 b 7 0 0583027 72T m

Date of Issuance: June 16, 1997

The 1997 edition of thisStandard is beingissued with an automatic addenda

subscription service. The use of an addenda allows revisions made in
response to public review comments or committeeactions to be published
as necessary. The next edition of this Standard is scheduled for publication
in 2000.

ASME is the registered trademark of The American Society of Mechanical Engineers.

This code or standard was developed under procedures accredited as meeting the criteria for
American National Standards. The Consensus Committee that approved the code or standard
was balanced to assure that individuals from competent and concerned interests have had an
opportunity to participate. The proposed code or standard was made available for public review
and commentwhich provides an opportunityfor additional publicinput fromindustry, academia,
regulatory agencies, and the public-at-large.
ASME does not "approve," "rate," or "endorse" any item, construction, proprietary device,
or activity.
ASME does not take any position with respect to the validity of any patent rights asserted in
connection with any items mentionedin thisdocument, and does not undertake to insure anyone
utilizing a standard against liability forinfringement of any applicable Letters Patent, nor assume
any such liability. Users of a code or standard are expressly advised that determination of the
validity of any such patent rights, and therisk of theinfringement of such rights, is entirely their
own responsibility.
Participation by federal agency representative(s1or person(s) affiliated with industry is notto
be interpreted as government or industry endorsement of this code or standard.
ASME acceptsresponsibilityfor onlythoseinterpretations issued in accordance with governing
ASME procedures and policies which preclude the issuance of interpretations by individual
volunteers.

No part of this document may be reproduced in any form,

in an electronic retrieval system or otherwise,
without the prior writtenpermission of the publisher.

The American Society of Mechanical Engineers

345 East 47th Street, New York, NY 10017

Copyright Q 1997 by
THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS
All Rights Reserved
Printed in U.S.A.

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 FOREWORD
(This Foreword is not part of ASME 889.4.1-1997.)

The ambiguity of coordinate measuringmachine (CMM) manufacturers’specifications in

1978 made comparative evaluation of performance extremely difficult. Considering this fact
andtheincreasinguseof CMMs, theASMEMetrology Standards Committee B89formed
WG B89.1.12to develop an Americanindustrystandardthatwouldestablish equitable
means of determining machineperformance.
OnMarch17, 1983, the Standard, havingprogressedtotheapplicationlevel, was given
interim status andreleased for a one-yeartrialperiod. Thereafter, desirable modifications
wereimplemented in theofficial Standard, whichwasreleasedin 1985 as ASMWANSI
B89.1.12-1985.
Followingthisrelease,theapplicationand useofthedocumentweremonitoredand
opportunities for improvement were noted for inclusion in 1990 as ASME B89.1.12M, Rev. 1.
Subsequent new information is included in this current revision,such as impactof ,
workpiece weight effects, interim testing, large machines, and significant modifications with
respecttoprobing,ballbartests,andtherotaryaxistest.Dueto changes in the status of
theASME Committee, thisStandardisdesignatedASMEB89.4.1-1997.
The AmericanNational Standards InstituteapprovedthisStandardonJanuary 30, 1997.

...
111

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 ASME STANDARDS COMMITTEE B89
Dimensional Metrology
(The following is the roster of the Committee at the time of approval of this Standard.)

OFFICERS
R. B. Hook, Chair
F. G. Parsons, Vice Chair
B.P. Biddinger, Secretary

COMMITTEE PERSONNEL
K. L. Blaedel, University of California/Livermore Lab, Livermore, California
J. B. Bryan, Bryan Associates, Pleasanton, California
T. Charlton, Jr., Brown and Sharpe Manufacturing Co., North Kingstown, Rhode Island
W. T. Estler, NIST, Gaithersburg, Maryland
R. J. Hocken, University of North Carolina, Charlotte, North Carolina
B. P a w , Boeing Co., Seattle, Washington
B. R. Taylor, Renishaw PLC, Gloucestershire, England
R. C. Veale, NIST, Gaithersburg, Maryland

SUBCOMMITTEE 4 - COORDINATE MEASURING TECHNOLOGY

R. B. Hook, Chair, Metcon, Coventry, Rhode Island
D. Beutel, Caterpillar, Inc., Peoria, Illinois
K. L. Blaedel, University of California/Livermore Lab, Livermore, California
T. Carpenter, US. Airforce, Newark, Ohio
T. Charlton, Jr., Brown & Sharpe Manufacturing Co., North Kingstown, Rhode Island
T. D. Doiron, NET, Gaithersburg, Maryland
R. D. Donaldson, Giddings & Lewis, Dayton, Ohio
B. Edwards, Automation Software, Farmington Hills, Michigan
W. S.Gehner, Deere & Co., Moline, Illinois
A. J. Griggs, Brown & Sharpe Manufacturing Co., North Kingstown, Rhode Island
J. L. Henry, Sheffield Measurement, Dayton, Ohio
B. Parry, Boeing Co., Seattle, Washington
S. D. Phillips, NIST, Gaithersburg, Maryland
G. K. Stevenson, G.K.S. Inspection Services, Inc., Livonia, Michigan
B. R. Taylor, Renishaw PLC, Gloucestershire, England
R. C. Veale, NIST, Gaithersburg, Maryland

ADDITIONAL SUBCOMMITTEE 4 MEMBERS

M. Arenal, RAM Optical Instrumentation, Huntington Beach, California
M. A. Bahtiarian, Noise Control Engineering, Inc., Billerica, Massachusetts
J. Baldwin, Geomet Systems, Inc., Redwood City, California
W. L. Beckwith, Jr., Brown & Sharpe Manufacturing Co., North Kingstown, Rhode Island
F. K. Bell, Giddings & Lewis, Dayton, Ohio
E. Blackwood, Boeing Co., Seattle, Washington
B. Borchardt. NIST, Gaithersburg, Maryland
J. B. Bryan, Bryan Associates, Pleasanton, California
R. P. Callaghan, Jr., Independent Quality Labs, Inc., Wyoming, Rhode Island

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 STD-ASME B87-q-L-ENGLL777 m 0757b70 0583032 23q m

J. Caprio, Carl Zeiss, Inc., Thornwood, New York

G. W. Caskey, NIST, Gaithersburg, Maryland
R. Casto, The Gates Rubber Co., Denver, Colorado
J. D. Cogdell, The Timken Co., Canton, Ohio
J. Dankowski, Fabreeka International, Inc., Sandimes, California
M. E. Denomme, Independent Quality Labs, Inc., Wyoming, Rhode Island
F. Farzan, Intelligent Measurement, Inc., Encino, California
M. L. Fink, Beoing Defence 81Space Group, Seattle, Washington
M. T. Gale, Giddings 81 Lewis, Dayton, Ohio
G. E. Hafely, Lockheed Missles & Space, Sunnyvale, California
S. Harrison, SAIC, Las Vegas, Nevada
G. P. Hegarty, Fabreeka International, Inc., Stoughton, Massachusetts
E. Helmel, Helmel Engineering Products, Inc., Niagra Falls, New York
V. Hetem, Bradley University, Peoria, Illinois
R. J. Hocken, University of North Carolina, Charlotte, North Carolina
T. H. Hopp, NIST, gaithersburg, Maryland
R. W. Homing, Barry Controls, Brighton, Massachusetts
J. Hurt, SDRC, Milford, Ohio
J. A. Jalkio, Cyberoptics Corp., Golden Valley. Minnesota
L. R. Kamholz, Helmel Engineering Products, Inc., Niagra Falls, New York
W. Kilpatrick, United Technologies, Pratt & Whitney, Hartford, Connecticut
J. J. L'Eplattenier, Bobst Group, Inc., Roseland, New Jersey
J. S. Lifson, General Electric, Lynn, Massachusetts
T. Mclean, Brunson Instrument Co., Kansas City, Missouri
W. McLendon, Lockheed Aeronautical, Marietta, Georgia
D. R. McMurtry, Renishaw PLC, Gloucestershire, England
J. 1. Miles, Sr., Martin Marietta, Orlando, Florida
K. J. Moritz, Texas Instruments, Sherman, Texas
D. Moyer, Rank Taylor Hobson, Arlington Heights, Illinois
V. A. Mysore, Cyberoptics Corp., Minneapolis, Minnesota
R. W. Nickey, DoD, Naval Warfare Assesment Center, Corona, California
H. S. Nielsen, Cummins Engine Co., Columbus, Indiana
J. T. Nilsson, NCMS, Ann Arbor, Michigan
M. OLaughlin, Romer, Inc., Carlsbad, California
F. G. Parsons, Federal Products Co., Providence, Rhode Island
T. Posterick, Carl Zeiss, Inc., Minneapolis, Minnesota
W. H. Rasnick, Martin Marietta Energy Systems, Oak Ridge, Tennessee
R. M. Roterdam, Jr., MT1 Corp./CT Lab, Industry, California
R. J. Russell, Allied-Signal Aerospace, Kansas City, Missouri
S. C. Sandwith, Boeing Co., Seattle, Washington
J. R. Schmidl, Optical Gaging Products, Inc., Rochester, New York
R. Shelton, Electronic Measuring Devices, Inc., Flanders, New Jersey
J. N. Shry, Cubic Precision K & E, Tullahoma, Tennessee
D. Slocum, L. S. Starrett Co., Mount Airy, North Carolina
K. B. Smith, The Ohio State University, Columbus, Ohio
B. Tandler, Multi Metrics, Inc., Menlo Park, California
A. Traylor, Renishaw, Inc., Schaumburg, Illinois
K. Ulbrich, Electronic Measuring Devices, Inc., Flanders, New Jersey
G. L. Vander Sande, U.S. Army, Picatinny Arsenal, New Jersey
R. K. Walker, Westinghouse Marine, Sunnyvale, California
D. J. Warren, Industrial Measurement Systems, Norcross, Georgia
W. A. Watts, Glastonbury Gage, Glastonbury, Connecticut
F. J. Weingard, ACTCO Metrology Services, Meadville, Pennsylvania

vi

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 CONTENTS

Foreword ........................................................................... iii

Committee Roster ................................................................... V

1 Scope .......................................................................... i
1.1 Contents andSpecificationForm .............................................. 2
I .2 Alternatives ................................................................. 3

2 Definitions ..................................................................... 3
2.1 Glossary .................................................................... 3
2.2 MachineClassifications ...................................................... 13

3 Environmental Specifications ................................................. 17

3.1 General ..................................................................... 17
3.2 Temperature ................................................................. 17
3.3 Vibration .................................................................... 18
3.4 Electrical .................................................................... 18
3.5 UtilityAir .................................................................. 19

4 Environmental Tests .......................................................... 19

4.1 General ..................................................................... 19
4.2 Thermal Test ................................................................ 19
4.3 RelativeMotion Tests for Vibration .......................................... 24
4.4 ElectricalTests .............................................................. 25
4.5 UtilityAir Tests ............................................................. 25

5 Machine Performance ......................................................... 25

5.1 General ..................................................................... 25
5.2 Hysteresis ................................................................... 25
5.3 Repeatability ................................................................ 26
5.4 Linear DisplacementAccuracy ................................................ 27
5.5 VolumetricPerformance ...................................................... 31
5.6 BidirectionalLengthMeasurementCapability .................................. 46

6 SubsystemPerformanceTests ............................................... 47
6.1 ProbingAnalysis - Point-to-PointProbing ................................... 47
6.2 ProbingAnalysis - Multiple-TipProbing .................................... 49

7 Test Equipment ............................................................... 50

7.1 Temperature ................................................................. 50
7.2 Vibration .................................................................... 50
7.3 Displacement ................................................................ 50

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 7.4 Pressure ..................................................................... 51
7.5 Humidity .................................................................... 51
7.6
Utility Air .................................................................. 51

Figures
1A B89.4.1 Coordinate MeasuringMachineSpecificationForm ......................... 4
1B B89.4.1 EnvironmentalSpecification
Form ......................................... 5
IC B89.4.1
PerformanceSpecification
Form ........................................... 7
2 Schematic Illustration of Abbe Offset andAbbe Error ............................. 9
3 Load Concentration Chart ........................................................ 10
4 Fixed Table Cantilever Coordinate Measuring Machine ............................. 13
5 Moving Bridge Coordinate MeasuringMachine .................................... 14
6 FixedBridge Coordinate MeasuringMachine ...................................... 14
7 Column Coordinate Measuring Machine ........................................... 14
8 MovingRam Horizontal Arm Coordinate MeasuringMachine ...................... 15
9 Two Moving Ram HorizontalArm Coordinate Measuring Machines Witha
CommonBaseUsed in the Duplex Mode ....................................... 15
10 Moving Table HorizontalArm Coordinate Measuring Machine ..................... 16
11 Gantry Coordinate Measuring Machine ............................................ 16
12 L-ShapedBridge Coordinate Measuring Machine .................................. 16
13 Fixed Table Horizontal Arm Coordinate MeasuringMachine ....................... 16
14 Fixed Table Horizontal Arm Coordinate MeasuringMachine With aRotary
Table ......................................................................... 17
15 Moving Table Cantilever Arm Coordinate MeasuringMachine ...................... 17
16 Typical Setup for Performingthe TVE Test on Direct Computer-Controlled
Machines Withan Active Probe ................................................ 20
17 Typical DataFromthe TVE Test ................................................. 21
18 Typical Setup for theMeasurement of TVEonaFree-Floating Machine Using
Passive
Probes ................................................................ 21
19 Typical Setup for Measuring Displacement Errors AlongaBody Diagonal .......... 22
20 Typical Plot of Data for a TVE Test PerformedonaLargeMachine by
Measuring Displacement Errors onaBodyDiagonal .............................. 23
21 Typical ResultsofaRepeatability Test WiththeAxisRepeatabilityClearly
Labeled ....................................................................... 26
22 Typical Setup for RepeatabilityMeasurementUsinga Trihedral Probe .............. 27
23 Typical Example of Linear Displacement AccuracyDeterminedUsing Step
Gages ......................................................................... 29
24 Typical Setup for the Laser Test for Linear Displacement Accuracy ................ 30
25 Typical Resultsofa Linear Displacement Accuracy Test Usingthe Laser With
the Linear Displacement AccuracyClearlyLabeled .............................. 31
26 RecommendedBallBar Positions for MachinesWithNearly Cubic Work
Zones ......................................................................... 33
27 RecommendedBallBar Positions for MachinesWitha Single LongAxis .......... 34
28 RecommendedBallBar Patterns for aMachineWith Two Long Axesand
One Short Axis ............................................................... 35
29 Sample Fixture for HoldingaBallBarWithBothEndsFree ...................... 36
30 Ball Bar Test Results ............................................................ 37
31 Typical Setup for Offset ProbePerformance Testing ............................... 38
32 DefaultBallBar Positions for the Offset ProbePerformance Test ona
Vertical Ram Machine ......................................................... 39
33 Diagram of Test Ball Positions for the Performance Test on aRotaryAxis ......... 41
34 Default Positions -for Sphere .Locations on theRotaryAxisPerformance Test ....... 43

...
VI11

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 35 Diagram Schematically RepresentingtheMeanings oftheRadial.Tangential.
and AxialWorking Tolerances for the Rotary AxisPerformance Test ............ 44
36 TypicalResults of a VolumetricPerformance Test for a K C MachineWith a
Rotary
Axis ................................................................... 45
37 Schematic Diagram Showing theLocations of Probing in the Point-to-Point
Probing Test .................................................................. 48
Tables
1 Location of theReference Sphere ontheRotary Table ............................ 40
2 DefaultNominal Angular Positionsand Sample DataSheet for Obtaining
VolumetricPerformanceWith a RotaryAxis .................................... 42
Appendices
A User's
Guideto
ASME B89.4.1 .................................................. 53
B Thermal Environment Testing ..................................................... 55
C Vibration Analysis ............................................................... 57
D Electrical
Power
Analysis ........................................................ 59
E Utility
Air ....................................................................... 61
F Hysteresis Test DesignRecommendations ......................................... 63
G BallBar Test EquipmentDesignRecommendations ................................ 65
H Straightedge Tests for Ram AxisRoll ............................................. 69
I Interim Testing of CMM Systems ................................................ 73

ix

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 ASME 889.4.1-1997

METHODS FOR PERFORMANCE EVALUATION OF

COORDINATE MEASURING MACHINES

1 SCOPE surement capability. Subsystem performance consists

of proceduresto evaluate probingperformanceduring
This Standard establishes requirements and methods
point-to-point coordinate acquisitionwith single and
for specifying and testing the performance of coordinate
multiple tips.
measuringmachines (CMMs) havingthreelinearaxes
One of the most significant features of this Standard
perpendicular to each otherand up to one rotaryaxis
positioned arbitrarily with respect to these linear axes. is its treatment of environmental specification and test-
ing. The machineuser is assigned clear responsibility
In addition to clarifying the performance evaluation of
CMMs,this Standard seeks to facilitate performance for providing a suitable performance test environment,
comparisons among machines by unifying terminology, either by meetingthesupplier’sparameters or by ac-
generalmachine classification, andthetreatment of ceptingreducedperformance.Particular emphasis is
environmental effects. placed on the performance degradation caused by tem-
This Standard attempts to define the simplest testing perature variation and vibration. The treatment of ther-
methods capable ofyielding adequate results for the mal effects in this Standard is in conceptual conformance
majority of CMMsand is notintended to replace to the provisions of ASME B89.6.2. The key feature of
more complete tests that may be suitable for special this treatment is the relaxation of machine performance
applications. In particular, this Standard is most applica- requirements if the thermal environment causes exces-
ble to machines used in the point-to-point mode rather sive uncertainty or variation in the CMM performance
thanthe contour measurementmode.Althoughthis and does not meet thesupplier’srecommendations
Standardprovides checks for mostoftheparameters regardingthermal parameters.
relevantto coordinate measuringmachinesused in a Actualmachine performance testing is divided into
contouringmode,the checks do not actuallytest con- five major areas: repeatability, linear displacement accu-
touring accuracy, per se. Additions to this Standard to racy, streamlined artifact testing with a ball bar, rotary
include contouring performance are in process. axis testing, and bidirectional length measurement capa-
This Standard provides definitions of terms applicable bility. Supplements to the ball bar testing are provided
to CMMs. These definitions are separated intotwo for large machines and for machines used in the duplex
parts: first, a glossary covering technicaltermsused mode. (Note that the supplemental laser interferometer
throughoutthis Standard, and second, anexplanation diagonal displacement measurements will give numbers
oftwelvecommonmachine classifications. thatmaybe different from thoseobtainedwithlong
The actual specification of CMMs is subdivided into ballbars.However, these numbers also adequately
four sections: generalmachine classification, machine reflectthe performance of the machine.) Performance
environmental requirements and responses, machine per- tests for machines under loaded conditions are also
formance, andmachinesubsystemperformance.Ma- included. An important feature of these performance
chine classification includes machine type, measurement tests is the attempt to use normal operating procedures
ranges, position resolution, operating mode, and probing during the tests. This emphasizes theimportance of
method.Environmentalspecification includes thermal measurement procedure details, such as mode ofma-
response, electrical requirements, vibration sensitivity, chine operation and probetype. In addition,theuse
and utility air requirements. Machine performance speci- of normal operating procedures during the tests serves
fication includes repeatability, linear displacement accu- to emphasize the overall approach of this Standard in
racy,ballbarmeasurementperformance, offset probe considering measurement data as the results of the
performance, diagonal displacement performance (large complete measuring system, not just theCMM.
machines), duplex performance (machines usedin the Subsystem performance, at this time, provides a series
duplex mode), rotary axis performance,performance of tests for systematic point-to-point probing errors,
under loaded conditions, and bidirectional length mea- such as lobing. Tests are also providedformachines

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 STDOASME BB9.q.L-ENGL 3977 m 0 7 5 9 6 7 0 0583037 8Tb W

METHODS FOR PERFORMANCE EVALUATION

ASME 889.4.1-1997 OF COORDINATE MEASURING MACHINES

with multiple-tip probing. This includes the useof In order to clarify the use of this Standard, a short
probe changers and probeindexing capabilities. Tests guide is included as AppendixA. To assist theuser
for other subsystems, such as software, are of importance in tracing possible environmental problems, appendices
but arenotincluded in thisStandard. are also provided for thermal environment testing (Ap-
Throughout this Standard, the concept ofrange - pepdix B), vibration analysis (Appendix C), electrical
that is, the spread between the maximum and minimum power analysis (Appendix D), andutility air analysis
values in a set of data - isused as themeasure of (Appendix E). Appendices on hysteresis testing (Appen-
machine performance. This choice wasmadein favor dix F), ball bar test equipment (Appendix G), straight-
of more common statistical measures, such as standard edge tests for ram axis roll (Appendix H), and interim
deviation, and because the dominanterrors in coordinate testing of CMM systems (Appendix I), also provide
measuring machines are systematic as opposed to being theuserwithimportant subsidiary information.
random. In such cases, no generally accepted statistical Productivityisan important consideration in the
procedurescurrentlyexist. selection of a coordinate measuring machine. There are
Repeatability is defined as the “ability of a measuring numerous factors that affect relative productivity of
instrument to provide closely similar indications for measuring systems, including variables inherent to both
repeated applications of the same measurand under the the system and the workpiece. This Standard does not
same conditions of measurement.” The specified testing address methods to specifyand evaluate productivity;
of repeatabilityrequires a series of measurements of rather, productivityshould be evaluated withrespect
the center coordinates of a precisionball,usingthe totheexpecteduse of the system.
same testingprocedureasthe tests tomeasurethe
effect of thethermal environment.
The linear displacement accuracy of the machine is
1.1 Contents and Specification Form
measuredalongthreemutually perpendicular lines in
the work zone. The tests may be performed using either Any specification described as complying withthis
a step gage ora laser interferometer. This Standard Standard shall include at least the following items.
carefully details the treatment of these data if any mean (a) Machine classification (see para. 2.2). If no ma-
temperature in the tests departs from20°C (6S°F), at chine classification isapplicable, the actual configuration
whichmateriallength standards aredefined. shall be described in equivalent detail.
The overall measuringperformance of themachine ( b ) Principal mode of operation (free-floating manual,
is evaluated with a ballbar,providinglimitedbut driven manual, or direct computer control). If desired,
valuable testing of the machine. This method has been repeatability, linear displacement accuracy, volumetric
chosen due tothespeedandsimplicitywithwhich a performance, bidirectional length measurement capabil-
machine can be evaluated using a ball bar to simulate ity, point-to-point probing performance, andmultiple-
a real measurement procedure. For very large machines, tipprobing performance maybe specified for more
diagonal displacement measurements are used tosupple- than one mode of operation.
menttheballbarresults.Formachinesusedinthe ( c ) Principal probe type (passive, switching, propor-
duplexmode,measurements of a fixedball in various tional, or nulling). If desired, repeatability, linear dis-
positions are performed by both machines as a supple- placement accuracy, volumetricperformance,bidirec-
ment to ball bar measurementsby each machine. Further, tional length measurement capability, point-to-point
theball bar is measured in four positionswith offset probing performance, and multiple-tip probing perform-
probesto obtain theoffsetprobing performance. ance maybe specified for morethan one probe type.
The performance ofthemachine’srotary axis, if (d) Probe approach rate, probe approach distance,
applicable, is tested by measuring the locations of two and settling time(s) for theprincipalprobe type(s)
precision balls mounted at specified positions onthe specified.
rotary table. Again, this test is functional and is intended (e) Nominalvoltage,frequency,andpowerre-
to reflect the values that would be obtained from actual quirement.
measurements. The user of this specification is warned cf3 Utility air pressure, pressure variation, flow, tem-
thatrotary axes are particularly sensitive to theload perature,dew point, and particulate content.
distribution and the moment of inertia of the part being (8) Permissible environment vibration amplitude as
measured. A separate sectionisincludedthat allows a function of frequency. The amplitude must be specified
for performance testing of coordinate measuring systems at the interface between the equipment supplied by the
underloaded conditions. userandthat supplied by theCMM supplier.

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 ~~

STD.ASME B 8 7 m q . L - E N G L 3777 W 0757b70 0583036 732 W

METHODS FOR PERFORMANCE EVALUATION

OF COORDINATE MEASURING MACHINES ASME 889.4.1-1997

(h) Statement of availability of data required for 2 DEFINITIONS

foundationdesignandmachine installation.
( i ) Statement of the significant mean temperature 2.1 Glossary
change, if available, safe operating temperaturerange, This glossary contains brief definitionsof the majority
nominallocationforthe temperature variation error of technicaltermsused in this Standard. Omissions
test, and the availability of other thermal response data should be reportedtoASME (see Foreword).
for themachine.
( j ) Statement of nominalcoefficients of thermal Abbeerror: the
measurement error resulting
from
expansion of themachine scales, by axis. angular motion of a movable component and an Abbe
(k) Parameters describing a recommendedmachine offset betweenthe scale measuringthemotion of that
component andthemeasurementline (see Fig. 2).
thermal environment.
(1) Repeatability. Abbe offset: the instantaneous value of the perpendicu-
(m)Linear displacement accuracydefined by mea- lar distance between the displacement measuring system
surement with a laser interferometer or a mechanical (e.g., scale) of a measuring instrument and the measure-
master. The choice shallbeclearlyspecified. ment line where the displacement in that coordinate is
( n ) Volumetric performance includingballbarper- being measured. A schematic illustration of this concept
formance, offset probeperformance,volumetrictests isgiven in Fig. 2.
for machineswithlargeworkzones,tests for duplex acceptable machine load: themachineloadthat can
machines,rotaryaxis testing, and tests for machines be applied throughthespannedregion of contact as
underloaded conditions. defined in the load concentration chart (see Fig. 3). All
(o) Bidirectionallengthmeasurementcapability. standard machine specifications will remain unchanged
( p ) Point-to-pointprobingperformance. under “acceptable machineloading.” (Note: refer to
(4)Multiple-tip probingperformance. para. 5.5.7 for a detailed testing procedure thatdescribes
( r ) A sample machine specification form. This form acceptable machineloadingtest conditions.)
isillustrated inFig. 1 for a typicalmachine.Itis accuracy: a quantitative measure ofthedegree of
divided into three sections: General (Fig. IA), Environ- conformance to recognizednational or international
mental (Fig. IB), and Performance (Fig. IC). The standards of measurement.
General section is intended to characterize the machine
by configuration, size, operation mode, and probe type. axisdirection: the direction ofany line parallel to
The Environmental section is intended to describe envi- themotiondirectionof a linearlymoving component.
ronmental requirements for the machine. The Perform- ballbar: a gage consisting of twohighlyspherical
ance section illustrates theparametersused to specify tooling balls of the same diameter connected by a rigid
performance within the context of this Standard. In the bar. A ballbar, as used in this Standard, mustbe
casethatmorethan one operating mode/probetype sufficiently mechanically rigid that its length is constant
combination is specified, performance shall be specified duringthe course of a set of measurementsbut does
for each combination. This form cannot be effectively not haveto be calibrated (see Appendix G).
used outside the context of this Standard as the Environ-
mental and Performance sections are closely connected CG locationzone: a supplier-specifiedzonewithin
through working tolerancederating procedures described theloading area in whichthemachineload center of
in Sections 4 and 5. gravity, CG,must lie.
cosineerror: the measurement error in themotion
direction caused by angular misalignmentbetween a
linear displacement measuring system and the gage (or
1.2 Alternatives part) beingmeasured.Equationsfor computing cosine
error aregiven in para. 5.4.2.3.
This Standard allows parts of theenvironmental
tests sectionto be deferred or bypassedandonlythe deadpath: in laser interferometry, that distance be-
performance tests to be carried out. This alternative is tween the remote interferometer andtheretroreflector
acceptable only if it is acceptable to both the user and atclosestapproachwhich is not compensatedfor
the supplier and if deferred as specified in Section 4.1. changes in the indexofrefraction of air.

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR.PERFORMANCE EVALUATION
ASME 089.4.1-1997 OF COORDINATE MEASURING MACHINES

GENERAL

Machine Classification (para. 2.2): (A figure with axis designation and direction of positive machine motion shall be
supplied.) For duplex applications, relative positions and common elements of the two machines shall be shown.

Measuring Ranges (full travel):

X m m (in.)
Y m m (in.)
Z m m (in.)
D m m (in.) - diameter of rotary axis, if supplied (see Glossary)

Readout Resolution (least count):

X mm (in.)
Y m m (in.)
Z m m (in.)
a deg. (arc sec) - resolution of rotary table, if supplied

Principal Mode of Operation (more than one mode may be specified):

Free-floating manual -
Driven manual
Direct computer-controlled -
Principal Probe Type (more than one type may be specified):

Passive
Switching
Proportional
Nulling
Displacement-measuring
Proximity

Operating Parameters:

Probe approach rate - mm/sec (in./sec)

Probe approach distance ___ m m (in.)
Settling time
Passive (solid or hard) probes ___ sec
Proportional probes -sec
Probe configuration (describe):

Describe location of machine coordinate system origin:

Maximum acceptable machine load: -kg [Notes (11, (2)l

Safe machine load: kg

NOTES:
(1) The correct IS0 unit of weight is the Newton (N), but the kilogram (kg) is used customarily. One kg has a weight
approximately equal to 10 N in the earth's gravitational field.
(2) 1 kg weighs approx. 2.2 lb.

FIG. I A B89.4.1 COORDINATE MEASURING MACHINE SPECIFICATION FORM

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
OF COORDINATE MEASURING MACHINES ASME B89.4.1-1997

ENVIRONMENTAL Page 1 of 2
specified. If more
For the following parameters, the principal mode ofmachine operation and the probe type must be
:han one operatingmode, probe type, or mode type combination isdesired, a separate performance specification sheet
shall be used for each combination.

Operating Mode Probe Type

Significant Mean Temperature Change

(para. 4.2.3) "C ("F)

Safe Operating Temperature Range (para. 4.2.3)

Min. "C ("F) Max. "C ("F)

Nominal Location for TVE Test (Machine Coordinates)

X m m (in.)
Y mm (in.)
Z mm (in.)

Nominal Coefficient of Thermal Expansion ofMachine Scales [Note (1)1

X ppm per "C ("F)

Y ppm per "C ("F)
z ppm per "C ("F)

Electrical (para. 3.4 and Appendix D):

Frequency
Voltage V Hz
Amperage A SurgelSag V

Allowable transient voltages (0.5 to 800 ks): Magnitude V

Environmental Vibration (para. 3.3 and Appendix C)

Option 1: Response function data [Note (2)1

Option 2: Broad band data

Peak-to-peak
amplitude vibration km
Frequency range Hz

Utility Air [if applicable (para. 3.5 and Appendix E)]

Pressure 2 MPa (psi)

Flow rate IN/min (SCFM)
Dew Point "C
Particle removal requirements:
size Particle km % removal

Availability of Foundation/lnstallation Data:

Yes No

FIG. 1B B89.4.1 ENVIRONMENTALSPECIFICATION FORM

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
ASME 889.4.1-1997 OF COORDINATE MEASURING MACHINES

ENVIRONMENTAL (CONT'D) Page 2 of 2

Temperature:

Mean ambient temperature "C ("F)

Daily cycling
amplitude (24 hr) 5 "C ("F)
Superimposed cycle(s)

Amplitude "C ("F)

Frequency cycles/hr

Gradients

Vertical "Um ("F/ft)

Horizontal "C/m ("F/ft)

Mean Air Speed Surrounding

the Machine [Note (3)l m/min (ft/min)

(Additional parameters on machine component placement and special flow requirements aret o be attached, if
appropriate.)

GENERAL NOTE:
The parameters listed here are based on assumptions regarding normal air conditioned rooms. Another set, if
provided as part of the machine specification and agreed on between supplier and user, shall be acceptable for the
purposes of this Standard. In some cases, other fluids (rather than air) are used to provide thermal stability. In those
cases, separate parameters should also be provided, if possible.

NOTES:
(1) 1 ppm = parts per miIIion =
(2) Detailed vector vibration spectra shall be attached as part of this specification.
(3)Maximum air speed should not exceed 6 m/min (20 ft/min) at 20°C (68°F). See ASME B89.6.2 for a full discussion
of parameters affecting operator comfort.

FIG. 16 B89.4.1 ENVIRONMENTAL SPECIFICATION FORM (CONT'D)

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
OF COORDINATE MEASURING MACHINES ASME 889.4.1-1997

PERFORMANCE - BASIC MACHINES

For the following parameters, the principal mode of machine operation and the probe type must bespecified. If more
than one operating mode, probe type, or mode type combination is desired, a separate performance specification sheet
shall be used for each combination.

Operating Mode Probe Type

(For all tests below, the reported value is the maximum range of error.)

Repeatability [Note ( l ) ]-All Linear Axes (para. 5.3)

X mm (in.)
Y m m (in.)
Z m m (in.)

Linear Displacement Accuracy (para. 5.4)

Step gage (para. 5.4.2) - or laser interferometer (para. 5.4.3) -

X mm (in.)
Y m m (in.)
Z m m (in.)

Volumetric Performance (para. 5.5)[Note (2)1

Ball bar - or gage block __

Length m m (in.)
Working tolerance mm (in.)
Offset probe performance (para. 5.5.3) m m (in.)

Bidirectional Length Measurement Capability (para. 5.6)

Gage block length m m (in.)

Working tolerance m m (in.)

Point-to-Point Probing Performance (para. 6.1) [Note (311

Working tolerance - 10 m m stylus length m m (in.)

Working tolerance - 50 m m stylus length m m (in.)
Working tolerance - 50 m m stylus with a 20 mm offset m m (in.)

Multiple-Tip Probing (para. 6.2) [Note (411

Working tolerance - 50 m m stylus m m (in.)

NOTES:
( 1 ) For large machines, the supplier shall specify the second probe approach rate, probe approach distance, and tra-
verse speed to be used (see para. 5.3.3).
(2) The user may supply measuring positions and lengths. If not supplied, then default values are used. Optional
lengths and positions shall be attached, if required.
(3) Probe approach rate, probe approach distance, and settling time are the default values unless otherwise specified.
(4) The user may supply measuring positions and lengths. If not supplied, then default values are used. Optional
lengths and positions shall be attached, if required.

FIG. I C B89.4.1PERFORMANCESPECIFICATION FORM

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
ASME B89.4.1-1997 OF COORDINATE MEASURING MACHINES

PERFORMANCE - ADDITIONAL SPECIAL TESTS

For the following parameters, the principal mode of machine operation and the probe type must be specified. If more
than one operating mode, probe type, or mode type combination is desired, a separate performance specification sheet
shall be used for each combination.

Operating Probe Type

(For all tests below, the reported value is the maximum range of error.)

Ball Bar and Diagonal Displacement (large machines, para. 5.5.4)

Length of ball bar

Working tolerance
Length of longest diagonal
Working tolerance

Duplex Performance (machines used in the duplex mode, para. 5.5.5)

Description of test plane:

X m m (in.)
Y m m (in.)
Z mm (in.)

Rotary Axis Performance (para. 5.5.6)

Rotary table position(s) and orientation(s) [Note (111:

Radial separation, Rs m m (in.)

Height, Hs m m (in.)

Working tolerances:

3D/alpha radial p m (pin.)

3D/alpha tangential p m (pin.)
30/alpha axial p m (pin.)

Testing Under Loaded Conditions (para. 5.5.7) [Note (2)l

Machine Load kg (lb)

NOTES:
(1) More than one position and orientation may be specified. If unspecified, default values are provided in thisStandard.
(2) Working tolerances for testing machines under loaded conditions are the same as those in the unloaded condition.
Tests for ball bar performance and rotary axis performance are conducted.

FIG. I C B89.4.1 PERFORMANCE SPECIFICATION FORM - ADDITIONAL SPECIAL TESTS

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
OF COORDINATE MEASURING MACHINES ASME 689.4.1-1997

Object being measured

Displacement measuring
system (scale)

FIG. 2 SCHEMATICILLUSTRATIONOF ABBEOFFSET AND ABBEERROR

diagonaldisplacement: thedisplacement of theprobe driven manual mode: a modeof CMM operation in

on a machinealong either a face or bodydiagonalof whichtheprobeof a machine is movedfrompoint
itswork zone. topoint in itsworkzoneusing drive mechanisms
(gears, lead screws, etc.) that are manually controlled.
diameter o l a rotary axis: the maximum diameter of a
rotary table (outside diameter) supplied with a measuring duplex mode: an operating mode in which two coordi-
machine.Thisisthemaximumdiameteralongwhich nate measuring machines having a defined relationship
a part can be fixtured. If it is intended that parts larger betweentheircoordinatesystems are usedtomeasure
thanthe face oftherotarytable or an extension plate coordinates of points on a commonworkpiece.
beplacedontherotary table, this diameteris either
duplex pedormance: for twomachinesused in the
themaximumdiameter of thepart or themaximum
duplexmode,the difference in themeasuredposition
diameter of the extension plate (see para. 5.5.6). If the
ofan artifact reported by the two machines relative to
table is square, then this diameter is the diameter of the
a single coordinatesystem.
maximum inscribed circle that will fit the square table.
free-Joating manual mode: a mode of CMM operation
drift test(thermal): a type of testusedtomeasure
in whichtheprobeof a machine is movedfrom point
temperaturevariation error [see temperaturevariation
to point by direct operator manipulation of the machine
error ( W E ) ] on a machine.Oneform of thistest
ram or probewithoutuse of a motor drive.
consists of continuouslyrecordingtheoutput of dis-
placement sensors placed in theposition of a probe gage(gauge): a mechanical artifact used either for
onthemachinereading against a samplepartover a checking a part or forcheckingtheaccuracy ofa
period of time. Detailedprocedures for conducting machine; or, a measuringdevicewith a proportional
drift tests onmachinesof different types are given in rangeandsomeformof indicator, either analog or
para. 4.2.2. digital.

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
ASME 689.4.1-1997 OF COORDINATE MEASURING MACHINES

gage block: a length standard with rectangular, round,

or square cross sections having
flat,
parallel,
opposing rd
-
E 2
gaging surfaces.
(D

-
L

hysteresis: as applied to a measuring system, the S

property of that system whereby its response to a given
stimulus depends onthesequenceofpreceding stimuli. .-òm 1
-
Hysteresis is oftencaused by drive train clearance, 2 Acceptable load
U
guideway clearance, mechanical deformations, friction, C
c
m
andloose joints (seeAppendix F). For thepurposes n
v)
of this Standard, three types of hysteresis are defined: 1000 2000
( a ) machine hysteresis: the hysteresis of the machine
Load. kg
systems when subjectedto loads;
( b ) probe hysteresis: the hysteresis of the mechanical
or electrical elements of a probe;and FIG. 3 LOADCONCENTRATIONCHART
( c ) setuphysteresis: thehysteresisof the various
elements in a test setup, normally due to loose mechani- machine load effects: the changes in machine perform-
cal connections. anceduetothemachineload.
largework zone: a workzonehavingany one or maximum acceptable machine load: the maximum
moreofthe following characteristics: acceptable machine load identified by the load concen-
( a ) the least measuringrange (full travelalong an tration chart (see Fig. 3).
axis) of the work zone exceeds 1.0 m (approx. 40 in.);
maximum traverse speed: see traverse speed.
(b) the greatest measuringrange (full travelalong
an axis) of the work zone exceeds 3.0 m (approx. 120 mean ambient temperature: themeantemperatureof
in.); and theambient environment surrounding a machine as
( c ) thevolume of theworkzone exceeds IO m3 computed from at least two readings taken at the center
(approx. 350 ft3). of the machine’s work zone during the interval required
for a test. The timebetweenthetworeadingsshould
laser interferometer: in thisStandard,aninterferome- be at least two-thirds ofthetest interval.
ter for displacementmeasurementthatuses a laser as
a light source. mean gage temperature: themeantemperatureof a
gage used formachinetesting as computedfrom at
lineardisplacementaccuracy: the difference between least two readings taken on the gage during the interval
a true displacement along a straight line and that required for a test. The time between the two readings
indicated by a measuring system. For thepurposes of should be atleasttwo-thirds ofthetestinterval.
this Standard, this difference is understoodto bethe
maximum systematic error from any point to any other mean scaletemperature: themeantemperature of a
pointalongthemeasurementline (see para. 5.4). machine scale as computed from at least two temperature
readings taken on that scale during the interval required
loadconcentrationchart: the relationship ofthema- for a test. The timebetweenthetworeadingsshould
chine loadtothespannedregionof contact, shown be atleasttwo-thirdsof the test interval.
graphically in Fig. 3.
mean temperature: the average temperature computed
loading area: a specifiedarea (mm2 or in.2) of the from a stated number of temperature measurements at
CMM’s workpiecemounting surface that isused for equallyspacedtime intervals at a specifiedlocation.
supporting themachineload. measurand: the quantity beingmeasured.
lobing: a systematic error in themeasuringaccuracy measurement line: a line in the work
zone of a
of probing systems such that a measured value depends machinealongwhichmeasurementsare taken.
onthe displacement direction oftheprobetip.
measurement point: a point in theworkzoneof a
machine load: the load, in kilograms (kg), placedon machine at which machine coordinates are recorded as
the workpiecemounting surface. partof a measurement.

IO

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
OF COORDINATE MEASURING MACHINES ASME 889.4.1-1997

movable component: a major structural component proportional to a distance between a reference point
thatismovable relative tothemachinebaseduring on the machine ram andthe workpiece. Such probes may
measurement. be displacement-measuring probes, proximityprobes, or
nullingprobes.
nominal coeflcient of thermalexpansion: an estimate
( e ) proximity probe: a probethatgives a signal
of the coefficient of thermal expansion of a body. [Note
proportionaltothe distance fromtheprobe tip to the
that since the true coefficient for a bodyisunknown,
workpiece.
anuncertaintymustbe applied whenmakingnominal
cf) switching probe: a probe that gives a binary
differential expansion corrections (see Section 4).] For
signal as a resultofmaking contact with, or being in
thepurposesofthisStandardand in reference to the
proximityto, a workpiece.
nominal coefficient of expansion of machine scales, it
shallmeanthe effective coefficient of the scale and probeapproachdistunce: the distance of approachto
itsmountingto the machine as measured in linewith the part at which the machine traverse speed is reduced
the scale for typicalmachinesofthegiven design. totheprobeapproachrate for measurement.
nominaldifSerentialexpansion: the difference between probeapproachrate: thenominalspeed of approach
the nominal expansion of the part and the master. (Note oftheprobetowardthepartduringthe acquisition of
that in thisStandardthemachine’s scales shallbe data (usedprimarily for switchingprobes).
considered the master.)
radial separation: the perpendicular distance from the
performancetest: anyof a numberoftestprocedures axis ofrotationof a rotary axis to either of the two
that are usedtomeasuremachineperformance. test spheres used to assess the volumetric performance
periodicerror: an error in thelineardisplacement for a rotary axis (see para. 5.5.6).
accuracyof a machinethatis cyclic overaninterval ram: the moving component of a machine that cames
whichnormally coincides withthenatural periodicity theprobe.
ofthemachine scales. For example, in a leadscrew
driven machine with rotary encoders, the periodic error range: the difference between the maximum and mini-
is usually synchronous with the pitch of the lead screw. mumvaluesof a set of measurements ofnominally
thesame quantity.
pitch: theangularmotionof a carriage, designed for
linear motion, about an axis perpendicular to the motion repeatability (of a measuring instrument): the ability
direction andperpendiculartothe yaw axis. of a measuringinstrument toprovideclosely similar
indications for repeated applications of the same measur-
probe: in this Standard, a device that establishes loca- and under the same conditions of measurement. These
tion of the movable components ofa coordinate measur- conditions include:
ing machine relative to a measurement point. Six types ( a ) reductionto a minimumofthe variations due
ofprobesarediscussed in this Standard: to the observer;
(u) displacement-measuring probe: a probe that gives
( b ) thesamemeasurementprocedure;
a signalproportionalto a displacement oftheprobe
(c) thesame observer;
fromits free position. ( d ) thesamemeasuringequipment,usedunder the
( 6 ) nulling probe: a probe that, when referencedto
same conditions;
a workpiece, gives a signalwhichcausesthemachine ( e ) thesame location; and
tobedrivento a positionthat will null theprobe
cf) repetitionover a short period of time.
reading. Repeatability maybe expressed quantitatively in terms
( c ) passive(solid or hard)probe: a probethat of the dispersion characteristics ofthe indications.
mechanically fixes the movable components relative to
the workpiece. Two types are discussed in this Standard: repeatability (of results of measurements): the close-
seating probes, which are hard probes that are positively ness of the agreement between the results of successive
constrained tomaintain their locationwith respect to measurements of the same measurand carried out under
a measurementpointwithout operator contact; and the same conditions of measurement. These conditions
nonseatingprobes,which are hardprobesthat require are called repeatability conditions andincludethe fol-
force applied by a machine operator tomaintain their lowing:
positionwith respect to a measurement point. ( a ) thesamemeasurementprocedure;
( d ) proportionalprobe: a probethat gives a signal (6) thesame observer;

II

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
ASME B89.4.1-1997 OF COORDINATE MEASURING MACHINES

( c ) the same measuringinstrument,usedunder the step gage (uniandbidirectional): a gage comprising

same conditions; a rigid bar with calibrated features used for determining
( d ) the same location; and the accuracy of distance measurements in the direction
( e ) repetition over a shortperiod of time. of linearmotion. (Note thatsomegagescanonly be
Repeatability maybe expressed quantitatively in terms probed from one direction, while others are constructed
of the dispersion characteristics oftheresults. so that probing can be performed from both directions.)
resolution (of a displaying device): the smallest differ- supplier: a party who contracts, or indicates readiness
ence betweenindications of a displaying device that to contract, to supply a CMM to a user.
can be meaningfullydistinguished. (Note that for a
systematic error: theportion of a machine error that
digital displaying device,this is the change inthe
remains even after computing the mean of a very large
indication whentheleastsignificant digit changes by
number of similar measurements.
one step. This concept applies also to a recording
device.) temperature variation error (TVE): an estimate ofthe
maximum possible measurement error inducedsolely
resolution (of an indicating device): a quantitative by deviation of the environment from average thermal
expression of theabilityofan indicating device to conditions (see para.4.2.2 for complete specification).
distinguish meaningfully between immediately adjacent
valuesofthequantityindicated. thermal error index (TEI): thesummation,without
regard to sign, of the estimates of all thermally induced
roll: theangularmotion of a carriage, designed for measurement errors expressed as a percentage ofthe
linear motion, aboutthelinearmotionaxis. workingtolerance(see para. 4.2 for complete specifi-
safe machine load: themaximummachineloadthat
cation).
can be applied to the CMM’s work area without causing traverse speed: thespeedobtained by the tipofthe
damage, tipping, or other unsafe conditions. ramof a measuringmachine,measuredwithrespect
to thepartmounting surface, whenthemachineis
safe operating temperature range: the temperature
movedbetweennominal locations withoutmeasuring.
range in which a measuring instrument may be expected
For the purposes of this Standard, the maximum traverse
to operate withoutphysicaldamage to theinstrument
speed is themaximumspeed along anygiven ma-
or its support systems (Le., computers, probes, etc.).
chine axis.
settling time: thetimerequiredbetween contact of a uncertainty of nominal differential expansion (UN-
hard or proportional probe with a measurementpoint DE): the estimated possible difference betweenthe
and thetimeatwhichvalid data may betaken. actual differential expansion and the nominal differential
signijìcant mean temperature change: the change in expansion due to uncertainties in the accepted (nominal)
mean ambienttemperaturesurrounding a machine, coefficients of thermalexpansion(seepara.4.2.1).
which, in the supplier’s judgment, will cause sufficient user: a party who contracts to accept a coordinate
degradation in machine performance such that perform- measuringmachinefrom a supplier.
ance evaluation(Section 5 ) should be repeated.
vibration amplitude: the peak-to-peakamplitudeof a
spanned region of contact: the area bounded by all givenfrequency component of a vibrationspectrum.
pointsof contact betweenthemachineloadandthe
workpiecemounting surface. workingtolerance (WT): the maximum acceptable
range in themeasurements for anyperformancetest
specializedmachineload: a specialloadingcase in this Standard. In particular, this applies to repeatabil-
wherein the maximum acceptable machine load specifi- ity, lineardisplacementaccuracy,volumetricperform-
cation is modified; the resultantof a specially distributed ance,duplexmodeperformance, rotary axisperform-
or locatedmachineload. ance, performanceunderload conditions, bidirectional
lengthmeasurement capability, point-to-pointprobing
staging: themoving of a gagefrom one position to performance, and multiple-tip probing performance mea-
another suchthat a series of measurementsstarted surementresults.
in one position maybe continued in the subsequent
position. workpiece: anobject to be measured.

12

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
OF COORDINATE MEASURING MACHINES ASME 089.4.1-1997

workzone: themeasurementvolume of a machine as

specified by the supplier. Morethan one workzone
maybe specified for a givenmachine,andworking
tolerances may be specifiedseparatelyforeachwork
zone.
work zone aspecr ratio: the ratio of the greatest axial
measuring range (full travel) to the smallest measuring
range (full travel) for a workzone.
yaw: the angular motionof a carriage, designed for
linearmotion,about a specifiedaxisperpendicular to
themotion direction. In the case of a carriage with
horizontalmotion, the specified axis shallbevertical
unless explicitly specified. For a carriage that does not
havehorizontalmotion,theaxismust be explicitly
specified.

2.2 MachineClassifications
Thefollowingclassifications of different types of
CMMs are provided for ease of machine specification.
A place is provided in the standard machine specification
form, Fig. 1 , that shall be used to designate the machine
classification as described below. As part of the specifi-
cation, a drawing equivalent to Figs. 4 through 15 with
the axis designationanddirection of positivetravel, FIG. 4 FIXED TABLECANTILEVER
shall be provided. In the case whererotary axes are COORDINATEMEASURINGMACHINE
supplied, they shall be added to each machine classifica-
tion in theposition of their expected normaluse (if
movable). Figure 14 shows one example of a machine on thebase. A typicalmachine of thisclassification
with a rotary axis. Inthecasewheretwomachines isshown in Fig. 4.
are used in theduplexmode, a drawingshowingthe
positional relationship ofthetwomachinesandany 2.2.2 Moving Bridge. A machine employing three
elements common to thetwomachinesshallbepro- movable components moving along mutually perpendic-
vided. Figure 9 shows an example of two moving ram ular guideways. The probeisattachedtothefirst
horizontal ann machineshaving a commonbaseand component which moves vertically relative to the sec-
used in the duplex mode. If a machine is to be supplied ond. The second component moves horizontally relative
that does not conform to one of the described machines, to the third. The third component is supported on two
then a drawing similar in content to thoseshown in legs that reach down to opposite sides of the machine
thisclassification section, with axis designations and base, and moves horizontally relative to the base. The
directions of positivetravel,shall be provided as part workpiece is supported on the base. A typical machine
of themachinespecification. of thisclassificationisshown in Fig. 5.

2.2.1 FixedTableCantilever. A machine em- 2.2.3 FixedBridge. A machine employing three

ploying three movable components moving along mutu- movable components moving along mutually perpendic-
ally perpendicular guideways. The probe is attached to ular guideways. The probe is attached to thefirst
the first component whichmovesvertically relative to component whichmovesverticallyrelativetothesec-
the second. The second component moves horizontally ond. The secondcomponentmoveshorizontally along
relative to the third. The third component is supported at a bridge structure above it that is rigidly attached at
one end only, cantilever fashion, and moves horizontally eachendtothemachinebase. The third component
relative to the machine base.The workpiece is supported moveshorizontallyrelativetothemachinebase. The

13

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
ASME 689.4.1-1997 OF COORDINATE MEASURING MACHINES

FIG. 6 FIXEDBRIDGECOORDINATE
FIG. 5 MOVING BRIDGECOORDINATE MEASURING MACHINE
MEASURING MACHINE

workpiece is mounted on the third component. A typical

machine of thisclassificationisshown in Fig. 6.
2.2.4 Column. A machine employing two movable
componentsmovingalongmutuallyperpendicular
guideways. The probe is attached to the first component
whichmovesverticallyrelative to themachinebase.
The second component moveshorizontallyrelativeto
themachine base in two directions defining a plane
perpendicular to the first component motion. The work-
piece is supported on the second component. A typical
machine of thisclassification is shown in Fig. 7.
2.2.5 Moving Ram Horizontal Arm. A machine
employingthreemovable components movingalong
mutually perpendicular guideways.The probe isattached
to the first component which moves horizontallyrelative
to the second component. The second component moves
vertically relative to the third component. The third
component moves horizontally relative to the machine
base. The workpiece is mounted on the machine base.
A typicalmachine of thisclassificationisshown in
Fig. 8.
2.2.6 Duplex Mode Machine. Machinesused in
the duplex mode, a mode in which two machines have
a defined relationship between their coordinate systems,
are used to measure coordinates of points on a common FIG. 7 COLUMN COORDINATE MEASURING
workpiece. Two moving ram horizontal ami machines MACHINE

14

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 S T D - A S M E B B Y * q * l - E N G L 3797 0757b70 0583050 Z I T

METHODS FOR PERFORMANCE EVALUATION

OF COORDINATE MEASURING MACHINES ASME 889.4.1-1997

FIG. 9 TWO MOVING R A M HORIZONTAL

A R M COORDINATE MEASURING MACHINES
WITH A COMMON BASEUSED IN THE
FIG. 8 MOVINGRAM HORIZONTAL ARM DUPLEX MODE
COORDINATE MEASURING MACHINE

having a commonbaseandused in theduplexmode 2.2.9 L-ShapedBridge. A machineemploying

are shownin Fig. 9. Many other machinegeometries three movable components moving along mutually per-
shown in para. 2.2 can also beused in theduplex pendicularguideways.Theprobe is attached tothe
mode. firstcomponentwhichmovesvertically relative tothe
second. The second component moves horizontally rela-
2.2.7 Moving Table Horizontal Arm. A machine tive to the third. The third component moves horizontally
employingthreemovablecomponentsmovingalong on two guideways, one at the base level or below, the
mutually perpendicular guideways. The probe is attached other raised above the base. The workpiece is supported
tothefirstcomponentwhichissupportedhorizontally onthe base. A typicalmachineofthis classification
at one end only, cantilever fashion, and moves vertically isshownin Fig. 12.
relative to the second. The second and third components
movehorizontally relative to the machine base. The 2.2.10 Fixed Table Horizontal Arm. A machine
workpiece is mounted on the third component. A typical employingthreemovablecomponentsmovingalong
machineofthis classification is shown in Fig. 10. mutually perpendicular guideways. The probe attached
is
to the first component which is supported horizontally
2.2.8 Gantry. A machine employing three movable at one end only, cantilever fashion, and moves vertically
components moving alongmutuallyperpendicular relative tothesecondcomponent.Thesecondcompo-
guideways. The probe is attached to the first component nent moves horizontally relative to the third component.
whichmovesvertically relative tothesecond.The The third component moves horizontally relative to the
secondcomponentmoveshorizontally relative tothe machine base. The workpiece is supported on the base.
third. The third component moves horizontally on two A typicalmachine of this classification isshown in
guide rails raisedabove the machinebaseon either Fig. 13. An alternate machine configuration is shown
side. The workpiece is supported on the base. A typical in Fig. 14, where a rotarytableismountedtothe
machine of this classification is shown in Fig. 1 I . machinebasewithits axis vertical. In this case, the

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 ~~

STD. ASME B87=9-1-ENGL17770757b70 0583053 1 7 b m

METHODS FOR PERFORMANCE EVALUATION

ASME 889.4.1-1997 OF COORDINATE MEASURING MACHINES

FIG. 12 L-SHAPEDBRIDGECOORDINATE
MEASURING MACHINE

FIG. 10 MOVING TABLE HORIZONTAL ARM

COORDINATEMEASURINGMACHINE

FIG. 13 FIXEDTABLEHORIZONTAL ARM

FIG. 11 GANTRYCOORDINATE MEASURING COORDINATE MEASURING MACHINE
MACHINE

16

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 S T D - A S M E B A 7 - q - L - E N G L L797 m 0 7 5 7 b 7 0 0583052 O02 m

METHODS FOR PERFORMANCE EVALUATION

OF COORDINATE MEASURING MACHINES ASME B89.4.1-1997

FIG. 15 MOVING TABLECANTILEVER ARM

COORDINATE MEASURING MACHINE
FIG.14FIXED TABLE HORIZONTAL ARM
COORDINATE MEASURING MACHINE WITH A Sectionand Section 4 aremet. The usershall be
ROTARY TABLE
responsiblefor conducting allenvironmentaltestsat
theinstallation site. The supplier shallhavetheright
workpiece is mounted to the rotary table. This example towitnessalltests. The supplier shall, on request,
is intended to illustrate how rotary tables can be config- supply test equipment as specified in Section 7, as well
ured on measuring machines. All of the other machines as support for equipment and tests, at a price to be
showncould also be equipped with rotarytables. negotiatedbetweenthe supplier anduser. The useris
cautionedthatfailure to conform to the supplier's
2.2.11 Moving Table Cantilever Arm. A ma- recommendationson cleanliness and cleaning proce-
chine employingthreemovablecomponentsmoving dures can lead to performancedegradation. For example,
along mutually perpendicular guideways. Theprobeis particulates, oils, andwatercansignificantlydegrade
attached to the first component which moves vertically machine performance, increase friction, andacceler-
relative tothesecond component. The second compo- atewear.
nentissupported at oneend only, cantilever fashion,
andmoveshorizontallyrelativetothemachinebase. 3.2 Temperature
The third component moves horizontally relative to the
machinebase.Theworkpiece ismountedtothethird 3.2.1 General. Temperature has a significantand
component. A typicalmachine of thisclassification is often misunderstood influenceon the accuracy of dimen-
shown in Fig.15. sional measurements. The provisions of ASME B89.6.2
form a part of this Standard, but interpretation is needed
forapplicationtocoordinatemeasuringmachines.
3 ENVIRONMENTAL SPECIFICATIONS
ASME B89.6.2 defines two alternative conditions under
which a test environment is thermally acceptable. The
3.1 General first,thatallpertinent components of themeasuring
It shall betheresponsibilityoftheusertoprovide system be at exactly 20°C (68"F), is generally unobtain-
an acceptable environment for performancetesting of able. This Standard isprimarilyconcernedwiththe
the CMM at the installation site. The environment shall second: thatthethermal error index (see para. 4.2) be
be considered acceptable if therequirements of this a reasonable percentage of the working tolerance. It is

17

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
ASME B89.4.1-1997 OF COORDINATE MEASURING MACHINES

theopinion of the B89.4 Subcommittee and implied 3.3.2 Responsibilities. The user shall be responsi-
in ASME B89.6.2 that it is not yet possible to specify ble for site selection, environmental shock and vibration
parameters for a thermal environment thatwill assure analysis, andadditional special isolators required to
a specific valuefor the thermal error index. Acceptability ensure compliance with the maximum permissible vibra-
ofan environment that does notcomplywiththe tionlevelsspecified bythe supplier. All questions of
supplier’s thermalparametersis therefore specified in compliance shall be determined at the interface between
terms of its effect on themachine. the support system providedby the user and the machine
systemprovided by the supplier.
3.2.2 Thermal EnvironmentParameters. The
supplier shall provide, as part of the machine specifica- 3.3.3Environmental Vibrational Parameters.
tion, a statement of the acceptable thermal environment The supplier shallprovide, as partof the machine
parameters. Such parameters shallcontain a specification specification,astatementof the acceptable seismic
on meanroomtemperature,maximum amplitude and vibration spectra at theuser-supplier interface. (This
frequency range of deviations from this mean tempera- interface may be very different, depending upon details
ture,environmentalthermal gradients, and air speed ofthe contractual arrangementbetweenthe supplier
surroundingthemachine. The user shall be informed and ‘user. For example, if the machine is supplied with
that conformanceto these parameters does not guarantee isolators, the interface shall be between the foundation
anacceptablemachinethermal environment, but does and those isolators. However, if theuser provides an
constitute due care ontheuser’spartandthus shifts isolation system from another source, the interface shall
responsibility for performance degradation due to envi- beattheconnectionbetweenthose isolators andthe
ronmental sensitivity from user to supplier. If the user machine.) This statement can contain a complete de-
chooses not to conform to the supplied parameters, the scription oftheallowablevibration amplitude as a
tests of environmental sensitivity (see Section 4) may function offrequency for each vector component of
lead to an increase in the acceptable working tolerance thevibrationspectrum; or, cansimply be a limit on
for a given performancetest; in which case, the degrada- the total vibrationalamplitude over a specified frequency
tion in performanceshall be solely theresponsibility range. The sample specification form, Fig. 1, allows
oftheuser. for either option. The statement of acceptable vibration
3.2.2.1 Thermal Radiant Energy. The machine spectra applies withthemachine in place.
shall not be exposed to directsunlight or other powerful 3.3.3.1AirborneVibrations.Although not
radiantenergysources. Other direct radiantenergy specified in this Standard, measuringmachines are
sources (such as fluorescent lighting) shall notbe, susceptible to airborne vibrations in the form of pressure
wheneverpossible, closer to anypart of themachine waves, ¡.e., acoustic noise. Wherever possible, the ma-
thanthelengthofthe longest machineaxis.Where chine should not be exposed to large levels of acoustic
this distance requirement is impractical, indirect lighting radiation, but if such acoustics are present, the necessity
designed for diffuse reflection and increased path length for sound-deadening is theresponsibility of the machine
shall be used. user. Excessive vibration due to acoustic coupling will
be evidenced intherelativemotiontest described in
3.3Vibration para. 4.3.

3.3.1 General. The support surface (floor, founda-

tion,isolationpad, etc.) uponwhichthemachinewill 3.4Electrical
bemountedcanhavemotioninduced as a result of
external forces in the surrounding area (due to other 3.4.1General.The electrical power supplied to a
machines, lift trucks, compressors, etc.). This motion machine can haveastrong effect on its ability to
can be continuous vibration, interrupted shock, or both. perform accurate and repeatable measurements. This is
Such motion, .if transmitted to themachine,hasa particularlytrue when amachineusessomeformof
degrading effect on the overall accuracy and repeatabil- computer for anycontrol or readout function.
ityofa CMM by causing relative motionsbetween
theprobe,themachineaxespositionmeasuringtrans- 3.4.2 Responsibilities. It shall be the responsibil-
ducers, and the workpiece.In addition, certain excessive ityoftheusertoprovide electrical powermeeting
motionamplitudescan cause damage to themachine. requirementsspecified by the supplier.

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
OF COORDINATE MEASURING MACHINES ASME 689.4.1-1997

3.4.3Electrical Parameters. The supplier shall formance fails, the environmental tests may be per-
provide, as part of the machine specification,a statement formed as partofthe diagnostic process. However, in
of the steady state voltage(s) requirements of the ma- such cases, the computations oftheuncertaintyof
chine, allowable deviations fromthis voltage(s), fre- nominal differential expansion (UNDE) and the thermal
quency requirements, and amperagerequirements.These error index (TEI) resulting from this UNDE calculation,
parameters are listed in thesamplespecificationform, withthetemperature variation error (TVE) set equal
Fig.1. to zero, shall beperformed (see paras. 4.2 and 4.2.1).

3.5 Utility Air 4.2 Thermal Test

3.5.1 General.Air supplies tomachinescan sig- The thermal test shall be performed under conditions
nificantly degrade their accuracy and useful working life. equivalent to those pertaining during performance tests
Temperature variations can generate thermal gradients in (Section 5 ) . Thetestenvironmentshallbeconsidered
themachine; particulates, oils, andwatercandegrade acceptable if the thermal error index, as defined below
bearingperformance, increase friction, and accelerate for each test, does not exceed 50% for that performance
wear. test. If thethermal error indexexceeds 50% andthe
machine environment does not conformto the supplier’s
3.5.2 Responsibilities. For all machines requiring guidelines, either the user shall correct the environment
utility air, it shallbe the responsibility ofthe user to or permissible working tolerance limits for thattest
supply utility air meeting requirements specified by the shallbe automatically increased by anamountsuch
supplier. thatthe greatest thermal error index is 50% of the
3.5.3Specification.For utility air, the supplier working tolerance for the specified test. If the thermal
shall provide specificationfor all air parameters required error index exceeds 50% and the machine environment
for the proper operation and maintenance of the machine. conforms to the supplier’s parameters, no thermal derat-
For air bearing machines, these shall at least include the ingofthe permissible working tolerance limits for
mean air temperature, permissible temperature variation, anyperformancetest is allowed. Methods for testing
pressure, and pressure variations. Furthermore, on some compliance of the thermal environment to the supplier’s
machines the acceptable dew point and the particulate environmentalparameters are given in Appendix B.
content shall be specified. Theseparameters are listed Thethermal error index shall be calculated for each
in thesamplespecificationform,Fig. 1. Airquality performancetestfrom the equation:
parameters, such as particulate, oil, and water content,
are the sole responsibility oftheuser,although the TE1 = [(UNDE + WE)/m 1 0 0
X

supplier shall offer guidelines.

where
TEI = thermal error index
4 ENVIRONMENTAL TESTS
W E = temperature variation error
UNDE = uncertainty of nominaldifferential
4.1 General
expansion
As stated previously, it is the philosophyof this WT = working tolerance for thattest
Standard that the environment is the responsibility of
themachineuser. If the environmentcomplieswith The nominal differential expansion term of the ther-
theparametersspecified by themachine supplier, the mal error indexexpression(ASMEB89.6.2)hasbeen
responsibility for meetingperformance specifications deleted because it is a requirementofthisStandard
rests solely with the machine supplier. If because of that nominal differential expansion corrections be made
economic or other considerations themachineuser as indicated in para. 5.4.2.6. All values in the equation
chooses not to conform to the supplier’s environmental are absolute valuesand are considered positive. The
specifications, this Standard provides a derating proce- correct TE1 value tobe used for derating linear displace-
dure on the machine performance. Derating procedures mentaccuracy shall be the full TE1value, calculated
and the tests onwhichthey are based are specified in from the equation above, for the measurement direction.
paras. 4.2 and 4.3. The correct TE1 value to be usedfor volumetric perform-
The supplier and user may agree to defer the environ- ance derating and 3D/alpha derating shall be the vector
mentaltestuntil after performance testing. If theper- sum of the TEIs for the three machine linear axes with

19

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
ASME 889.4.1-1997 OF COORDINATE MEASURING MACHINES

theUNDE set equal to zero. The correct T E 1 value

to be used for derating duplexperformance (two ma-

d
chines used in theduplexmode)shall be the TE1 Ra m

calculated from a TVE determined as in para. 4.2.2.5. Switching,

The correct TE1 value to beused for derating ofthe proportional
offset probe test performance specification shall be the or nulling
probe
same as that used for derating the volumetric perform- Test ball
ance. The correct TE1 valueto be usedfor derating
Rigid
thebidirectionallengthmeasurement capability shall
be the TEI, with the TVE set equal to zero. The TVE
maybe set equal to zero due totheshort duration of
themeasurement sequences. Theworkingtolerances
for repeatabilityandprobingperformance maynotbe
thermally derated.

4.2.1Uncertainty of NominalDifferential
Expansion (UNDE). Uncertainty of nominal differen- Machine
table
tialexpansion(UNDE) isbasedonanuncertainty of
1 ppm (I Fm/m)/"C for the scale', and an equal
uncertaintyfor the step gage, gage block,etc. It shall FIG. 16 TYPICALSETUP FOR PERFORMING
be calculated as THE N E TEST ON DIRECTCOMPUTER-
CONTROLLED MACHINES WITH AN ACTIVE
UNDE = (0.000002)(L) I (T, - 20) I PROBE

where the TVE on direct computer-controlled machines, driven

T, = mean ambienttemperature, "C manualmachines,andfree-floatingpassiveprobema-
L = thenominallengthtobemeasured in chines. TVE testprocedures for DCC machineswith
eachgiventest large workzonesarespecified in para. 4.2.2.3. TVE
test procedures for machines used in the duplex mode
For example, for lineardisplacement accuracy, the are specified in para. 4.2.2.5. Since some of thetests
nominallength L would be themeasuringrangefor can beusedonmorethanonetypeofmachineand
the axis under test; for gage block measurement during therearetrade-offsbetween ease andtime of testing
thebidirectionallengthmeasurementtest,thenominal for thethreeprocedures, it is optional which test(s)
length L wouldbethelengthofthegageblock; etc. should be performed. However, the test(s) chosen shall
The UNDEabove applies even if a laseris used be clearly stated inthemachinespecifications.
for machine checking. (See ASMEB89.6.2 for a further Itshould benotedthaton some machinesthere
discussionandhistoryoftheUNDE.) could be variations in the mean value of the supplied
air pressure, which can be misinterpreted as TVE. This
4.2.2 Temperature Variation Error ( N E ) . Tem- is due to changes in machine squareness and positional
peraturevariation error (TVE)shall be determined by drifts. During this test, care should be takento ensure
a drift test. The drift testsspecified in thisStandard properairpressureregulation.
are tobeconducted for a periodoftimeequal to the
4.2.2.1TVETestforDirectComputer-
duration of thelongestperformancetest. This short
Controlled Machines and Driven Manual Ma-
period is a compromise, and users are strongly advised
chines.Immediatelyprior to initiatingthistest,the
to run this test for a time period of at least 24 hr, as
machine shall have been parked at a position geometri-
manytemperature effects exhibit daily periodicities.
cally opposite tothetestpositionselected by the
The following procedures are tobe used for determining
machine supplier, for a time period equal to the duration
ofthe TVE test. A testballshallbemounted to the
' This approximation does not apply for scales with a nominal machineworkpiece supporting surface at theposition
thermal expansion coefficient equal to zero, or to scales with selected by the supplier for the TVE test. A switching,
thermal expansion coefficients thatareaccurately known. In these
cases, the user and supplier shall negotiate a correct value for the proportional, or nulling probe shall be mounted in the
scale uncertainty to be used for calculation of the UNDE. probeholder. This setup isillustrated in Fig. 16. For

20

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
OF COORDINATE MEASURING MACHINES ASME 889.4.1-1997

Temperature Variation Error

4 Ram

+5 L
""""""""
V*" -
3-
5
.-c i -
U
-
D
**o**
**..
**
*m*
*.** ** *m
**9**

TVE
Clamp
I l Tapered probe

.-
i l lime, min

FIG. 17 TYPICAL DATA FROM THE TVE

TEST
(The TVE here is approx. 5 Pm.)

direct computer-controlled machines, an automaticcycle FIG. 18 TYPICALSETUP FOR THE

shallbe established torepeatedly take readingsonthe MEASUREMENT OF TVE ON A FREE-
ball and establish the ball center coordinates. For driven FLOATING MACHINE USING PASSIVE
manual machines, similar measurements shall be taken PROBES
as quickly as possible. Movement for each probe mea-
surementshallbe the minimumnecessarytoachieve
reliable readings. The test shall be conducted for a the counterbalance. Theprobe shall besecured in the
time period at least as long as the longest performance positiondetermined by themachine supplier for the
testofthemachine.The test period shall be divided TVEtest.Asample setup for such a test is shown in
into intervals of approximately one minute or the time Fig. 18. Thetest shall beconducted for a timeperiod
required to take a minimum of three readings. In order at least as long as the longest performance test, and
to minimize the effect of repeatability, the mean value normal activity shall be continued around the machine.
of each coordinate for each interval shall be determined. The test period shall be dividedinto intervals of approxi-
The range of variation of ball center readings for each mately one minute. The mean value of each coordinate
coordinateshallbetheTVE.Notethatdatafrom this for each interval shall bedetermined.Therange of
test may be used to determine repeatability as specified variations of readings for each axis shall betheTVE
in para. 5.3, in the case where the location specified forthat axis. Typical data for such a testwith the
for repeatabilityandtheTVEtestisthesame.Data TVElabeled are shownin Fig. 17.
shall beanalyzed as illustrated in Fig. 17. 4.2.2.3 TVETest for DCC Machines With
4.2.2.2 TVE Test for Machines Used in the Large Work Zones. Due to the large volume occupied
Free-Floating Mode With PassiveProbes. (This by the machine structure and the difficulty in achieving
test is invalid unless the machine has passed a hysteresis a uniformthermalenvironment, the results of TVE
check as described in Appendix F.) The active portion testing on a large machine will usually be more depen-
of this testisperformedwiththemachineprobeat a dentonthe location of thetestthanthey are on a
positiondetermined by themachine supplier. Immedi- smaller machine. Further, it is more difficult to predict
atelypriortoconducting this test, the machineshall in advancethe appropriate location for the TVEtest.
havebeenparkedat a positiongeometrically opposite Therefore, onlargemachines it is essential tosample
tothetest position selected by themachine supplier several widely spaced locations to determine the TVE.
for the TVE test, for a time period equal to the duration Thetestspecifiedhere requires measuring a quantity
of the TVE test. Afterthisparking cycle, theaxes similar tothe linear displacementaccuracy (para. 5.4)
shall beunlocked.A passive probeshallbemounted along a body diagonal through the machine work zone.
intheprobeholderandsecured to theworkpiece Currently, theonly practical instrumentation for this
supporting surface by clamping, or by engaging the testis the laser interferometer. Forworking in large
probe with some feature of the surface and unbalancing work zones, careful attention must be given tocorrecting

21

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
ASME 689.4.1-1997 OF COORDINATE MEASURING MACHINES

for wavelength changes due to air temperature and

pressurevariations during themeasurementtime, as
described in para. 5.4.3.3. Additionally, the setup shall
allowonly a minimum of dead path. The deadpath
that exists shall be measuredandcompensatedfor
according to the laser interferometer supplier’s recom-
mendations.
The supplier shall select one of thebody diagonals
of the work zoneas the TVE measurementline. Immedi-
atelypriorto initiating thistest,themachineshall be
parked at a position that is the maximum perpendicular
distance from this measurement line for a period equal
totheduration of theTVE test. Displacement errors
shall then be measured along this line for a time period
at least as longasthelongest performance test,but Remote
not less than one hour. The laser shall not be rezeroed interferometer

during thistest.Normal activity shall be continued

aroundthemachine. A typical setup for measuring
displacement errors along a bodydiagonalisshown
in Fig. 19. The position of the interferometer with
respect to thelaserhead and steering mirror may be
extremely critical and should not be altered except
by thoseintimately familiar withthe principles of
(metallic)
commerciallaser interferometry.
In order to minimize theeffects of machine repeatabil-
FIG. 19 TYPICALSETUPFOR MEASURING
ity, a group of sequential measurements maybetaken
DISPLACEMENTERRORSALONG A BODY
at each measurementpointalongthediagonaland
DIAGONAL
averaged,butthetimeintervalbetweenthefirstand
(In an actual setup, the remote
lastreadingofsuch a group shall not exceedone
interferometer should be as close to the
minute. Separate plots of the displacement error versus
steering mirror as possible to reduce dead
time shall be made using data from at least the midpoint
path. This figure shows a large separation
and the two ends of the measurement line. It is strongly
for illustration purposes only.)
recommendedthat data from at least one additional
position be measuredandplotted for each 20 m3 of
work zone volume. A typical plot for three measurement machine shall choose the three locations for these tests.
positionsisshown in Fig. 20. The greatest range of The data fromthis TVE testshall be reportedand
any of these plots is the TVE. (If the data plots appear analyzed as shown in Fig. 20.
to reveal a systematic relaxationthat could be due to 4.2.2.5 TVE Test for Machines Used in the
the machine structure approaching thermal equilibrium, Duplex Mode. For derating performance specifications
theuser may elect to repeatthistest to get a better of individualmachinesused inthe duplex mode,the
measure of the TVE.) methodof para. 4.2.2.1,4.2.2.2,4.2.2.3, or 4.2.2.4
shall be used, as applicable.For derating theduplex
4.2.2.4 TVE Testfor
DrivenManual performance specification for two machines used in the
Machines With Large Work Zones. The TVE test duplexmode, an auxiliarytestisrequired.
for drivenmanualmachineswithlargeworkzonesis For theauxiliarytest, a referenceballshall be
identical to the TVE test described in para. 4.2.2.1 for mounted in a positionspecified by the supplier. The
direct computer-controlled machines and driven manual means of mounting shall be the means chosen for the
machines; however, in the case of large machines, this duplex performance test, para. 5.5.5.3. The position of
TVE testshall be conducted at three locations within the ball shall be determined repeatedly by both machines
thework zone andthemaximumTVE for any axis by themethodspecified in para. 5.5.5.3. Differences
at any of these three locations shall be reported as the ofpositionreported bythetwomachinesshall be
TVE for that axis of the machine. The supplier of the determined in the three axial directions. The maximum

22

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 S T D - A S M E B B Y - q * L - E N G L 1797 0759b70 0581058 520 m

METHODS FOR PERFORMANCE EVALUATION

OF COORDINATE MEASURING MACHINES ASME 889.4.1-1997

PosRlon 1

Position 2
5 -
4 -

TVE4 microns

.-ul -1 -
I lime
3-2 -
-3 -
-4 -
-5 -
-6 I

Position 3

FIG. 20 TYPICALPLOTOFDATAFORATVETESTPERFORMED O N ALARGEMACHINEBY

MEASURINGDISPLACEMENTERRORS ON ABODYDIAGONAL

23

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 S T D - A S M E 8 8 9 - 4 * L - E N G L L997 0759b70 0583059 4b7 D

METHODS FOR PERFORMANCE EVALUATION

ASME 689.4.1-1997 OF COORDINATE MEASURING MACHINES

range of these differences, onan axis-by-axis basis, but rather, some complicated function that only relates
shall be theduplex TVE for the machine. All other in a verygeneral way to vibration amplitude. If the
preliminaryrequirements,test requirements, and data machine does notpassthe functional test,Appendix
evaluationrequirementsshall be thoseused for de- C provides recommended procedures for accurate mea-
terminingTVE of the individual machines. surement of the seismic vibration spectra at theuser-
For each axial direction, the TVE determined by this supplier interface for the purposes of determining con-
auxiliary duplex TVE test shall be compared withthe formance tothe supplier’s specifications.
sumof absolute values of TVE determined for the Should therelativemotion amplitude (as measured
individualmachines. The larger of theseshall bethe in this test) exceed the requirements andbetraced to
TVE usedto derate the duplex performance specifi- sources that are the user’s responsibility (see Appendix
cation. C), andiftheuser does not desire to upgradethe
machine interface, then the machine specification shall
4.2.3 Other Temperature Effects. The Commit- bederated so thattherequiredrepeatabilitywill be
tee recognizesthat when using CMMs, errors caused equal to the measured repeatability on an axis-by-axis
by differential expansion, scale hysteresis,and other basis.Notethattherepeatabilitytest (para. 5.3) must
effects can be induced in machines whentheyare be performed before this derating can be accomplished.
operated at mean temperatures significantly different The absolute value of the worst-case difference between
fromthetemperatureatwhichtheywere aligned and the measured repeatability and the specified repeatability
calibrated. Unfortunately, it is notwithinthe current shall be used to derate the repeatability, the volumetric
state-of-the-art to develop simple tests for these effects. performance,theduplexperformance (machines used
It is therefore the Committee’s recommendation that if in the duplex mode), the offset probe performance, the
a machineisto be accepted at a mean temperature four-axis performance,the 3D/alpha performance,the
significantly different from the one used during align- bidirectional length measurement performance, the prob-
ment and calibration, the linear displacement accuracy, ing performance, and the multiple-tip probing perform-
volumetricperformance,andbidirectionallengthmea- ance. This derating is performedby adding thedifference
surement capability tests described in Section 5 shall to thespecifiedworkingtolerance. If themachine
be repeated for each temperature. It is the requirement working tolerance is already subject to derating due to
of this Standard that the supplier specify the significant thermal environment, the derating due to relative motion
mean temperature change for a given machine of given shall be arithmeticallyadded to thethermal derating.
working tolerance. Furthermore, the supplier shall spec-
ify a safe operating temperature rangewithinwhich 4.3.1MethodologyforRelativeMotion
the machine should be kept to prevent physical damage Tests - Direct Computer-Controlled Machines
to the machine (see Glossary). In addition, temperature and Driven Manual Machines. A single-axis, high-
sensors used for compensation need to be periodically resolution displacement indicator havinglow damping
verified, as the sensors are subject to damage and drift. and conforming to the requirements of Section 7 shall
be used; and, withthemachine set at a positionnear
the middle of its work zone, set to read relative motion
4.3Relative Motion Testsfor Vibration
betweenthe ram andthemachinetable or suitable
The relativemotion tests shall be performedunder fixture attached to the table. The direction of displace-
the same conditions as those pertaining during the mentindicationshall be aligned with each machine
performancetests(Section 5). The test environment linear axis in succession, and the maximum spread of
shall be considered acceptable if the relative motion the indicator reading will be judged to be the machine
amplitude measured between the machine ramand the vibrationamplitude for that axis.
worktableis less than 50% ofthemachineworking
tolerance for repeatability. For thepurposes of this 4.3.2MethodologyforRelativeMotion
Standard, this amplitude is to be assessed by the Tests - Free-Floating Passive Probe Machines.
following simple functional tests. The test duration shall On free-floating passive probe machines, the probe shall
be at least 10 min.Both steady-state vibrations and be engaged with the table using the minimum amount
any transients that might occur during normal use shall of counter weightforcenecessary to holdtheprobe
beincludedwithinthetestperiod.Thesetestsare in position [30 g (approx. 1 oz.) is recommended].
specified with the understanding that theydo not consti- Any clamping mechanisms for the axes shall be disen-
tute a well-defined measurement of vibration amplitude, gaged. The range of flicker of the machine readout in

24

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
OF COORDINATE MEASURING MACHINES ASME B89.4.1-1997

allthreelinear axes shall be observed. The maximumaccording to the supplier’s recommendations inany
spread of thereadoutflickershall be judged tobethe environment meetingtherequirements of Section 3; to
machine relative
motion
amplitude for that
axis.
include, if required, derating of the
acceptable
working
tolerances as described in Section 4. A machine meeting
4.4 Electrical Tests performance
specifications and other conditions agreed
uponbetweenthe supplier and usershall be accepted
Well-defined procedures and highly developed instru-
by theuser. The criterion for meetingperformance
ments exist that enable the measurement of the parame-
specificationsshallbethesatisfactorycompletion of
ters characterizing the electrical powersupplied to a
all tests specified in this Section, except thatanytest
machine. It is, however, the opinion of this Committee
or tests may be omitted by mutual agreement between
that such tests are, in the general case, an unwarranted
the supplier and user. It should be emphasized tHat the
expense and shall be undertaken only in the event that
performance tests for repeatability, linear displacement
themachine does notmeetperformancespecifications
accuracy, volumetric performance, and bidirectional
andthereisreasontosuspect the electrical power.
length measurement capability described in this Section
Failures due to electrical powerusuallyshowup as
containmany options, andthatthese options willnot
intermittent control or readout failures whichare difficult
necessarily give the same results on any given machine.
to link to mechanical causes. In the case of power
This is due to minor differences between what is really
being suspect, this Standard provides AppendixD which
being measured when different options are selected. It
describes the recommended procedure for determining
istheopinionofthisCommitteethattheseminor
the conformance of the electrical environment to the
differences are not significant. The choice ofanyfull
supplier’s guidelines.
set - thatis, one repeatability option, one linear
displacement option, appropriatevolumetricperform-
4.5 Utility Air Tests
ance tests, andthebidirectionallengthmeasurement
As withthe electrical powertests,there also exist test - will give a fairand complete picture of the
many complicated procedures for determining the qual- machine’s capabilities forthatmodeofoperationand
ity ofthe utility air suppliedto a machine.Itisthe that probe type. It may be desirable to use this specifica-
opinionofthisCommitteethatsuch exhaustive tests tion for evaluating machines having large aspect ratios
should notbe requiredfor checking conformanceto (generally greater than 4:l). The user should be aware
specification unlessa problem traceable tothe air supply that if straightness of the axis with the greatest measur-
is evident. As stated previously, variations in the mean ing range(full travel) is critical, then a separate measure-
value ofthesupplied air pressurecan cause changes ment of this parameter should be performed. Where two
in machine squareness andpositional drifts, so that if machines are used in the duplex mode, determination of
such changes do occur, then air pressure is a possible performance of each machine does not ensure a known
suspect. It shalltherefore be theresponsibility ofthe relationship betweentheiraxis systems. Therefore, a
supplier to examine, using the gages and filters supplied test for duplex performance is specified in para. 5.5.5.
with the machine, the mean pressure, pressure variations, The usershallberesponsible for conducting all
and cleanliness ofthe utility air at the input tothe performance tests at his installation site and the supplier
machine. If, in the supplier’s judgment, the air supply shallhavetheright to witnessall tests. The supplier
is inadequate, then further tests are described in Appen- shall, upon request, supply test equipment as specified
dix E for determining conformance of theutility air in Section 7 including support for equipment and tests,
to the supplier’s specifications. If, however, the supplier at a price to be negotiated between supplier and user.
judges the air supply to be adequate, then the utility
air shall be judged as conforming tospecification
without further testing.
5.2 Hysteresis
5 MACHINE PERFORMANCE
It is strongly recommended that a mechanical hystere-
sistest be performedon the machineandonanytest
5.1 General
setup before time is spent on other testing. Any problems
The supplier shall be responsible for providing a suggested by thehysteresistestsshould be corrected
machine that meets all performance specifications agreed before proceeding with other tests. Hysteresis tests are
upon betweenthe supplier anduserwheninstalled described in Appendix F.

25

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
ASME 689.4.1-1997 OF COORDINATE MEASURING MACHINES

5.3 Repeatability

5.3.1 General. The concept of repeatability testing

incorporated in this Standard isthatthetestmust
5 +5 f"_
L
"""_
o

evaluate a complete system, which may include effects o Axis

o e o repeatability
due to machine characteristics, human operators, and , : : : : ' : : : : 1 1
computer algorithms. Hence, the test must be performed Measurement 10
in a manner closely representing the way in which the .
- number
-""""""""""
machine will be used after acceptance. Implementation
of this concept requires that different tests be used for
different modesofoperation. The testchosen for a
machineshall be thetest for theprincipalmodeof -5 L
operation, probetype,probeapproachrate, probe ap-
proach distance, andprobeconfiguration, as specified FIG. 21 TYPICALRESULTSOF A
in Fig. lA, by the supplier. In general, the stylus of REPEATABILITYTEST WITH THE AXIS
the probeshouldbeparalleltothe ram unless agreed REPEATABILITYCLEARLYLABELED
upon by the user and supplier. Where alternative princi- (For this test, the repeatability was approx.
palmodes are specified,morethan one repeatability 4.5 km.)
test may be required. Specific modifications to the test
procedures are provided for machines with large work 5.3.3 Standard Tests for Repeatability
zones.Forthesemachines,thetraversespeedmust
also bespecified. There maybe cases wherenoneof 5.3.3.1
Repeatability Tests
the
in
Computer-Controlled Mode or the Manual
the specified test alternatives comply with the concept
Mode Witha Switching, Proportional, or Nulling
of this Section. Forsuch cases, the supplier anduser
Probe. A switching, proportional, or nulling probe
shall agree onan alternative testbefore entering into
shall be mounted in the probe holder. Ten sets of four
contract.
contacts each shall be made on the ball (for computer-
controlled machines these contacts shall be made under
computer control). Contact points for each set shall be
5.3.2 Common Features. The requirement in the spaced as widely as possible andnotallin the same
definition of repeatability to measure the same measur- plane. The center coordinates for each set of four
andshall be satisfied by measuringthe center coordi- readings shall be calculated, andrepeatabilityshall be
nates of a precision reference ball rigidly mounted on determined as in para. 5.3.2.
the workpiece supporting surface at a position where the
5.3.3.2 Repeatability Tests With Passive
machine linear axesare approximately at the midpointof
Probes. There are twoimportant classifications of
their travel, unless otherwise agreeduponbytheuser
passive probes: seating and nonseating (see Section 2).
and supplier. Ten determinations of the reference ball Since themachinereacts differently to probing de-
center shall bemade as rapidly as is practical. For pending on the probe type, two tests are required and
each axis, therangeoftheball center coordinate two repeatabilities must be determined. Inboth cases,
shall be determined as a maximum minus a minimum. themachineandprobe shall be insulatedfrom the
Machinerepeatabilityshall be reported as either the operator's hand by some means, such as by a cotton
largest range in coordinate valuesmeasured or the glove.
range in coordinate valuesona per axis basis. The The first test is for nonseating probes. A ball probe
rangeofa set of data is defined as themaximum shallbemounted in theprobeholder.Rigidityofthe
spread ofthe data. In theeventthata data point reference ball mounting shall be confirmed by engaging
obtained during arepeatabilitytest appears to be an theballprobewiththereferenceballand applying
outlier, thenthispointmaynotsimply be discarded pressure. Contact pressureshall be estimated at twice
when defining the range; rather, thecomplete repeatabil- that normally used for a ball probe. The check shall be
ity test must be repeated and the range for a complete performed in each of the three machine axis directions.
test evaluated. Figure 21 illustrates typical results of Change in the machine readout, as pressure is applied,
arepeatability test. shall not exceed the working tolerance for repeatability.

26

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
OF COORDINATE MEASURING MACHINES ASME 889.4.1-1997

than120% of the supplier’s recommendedprobing

speed, and greater than 90% of the maximum traverse
W‘ To machine ram
speedspecified for themachine. Insofar as possible,
all machine axes shall be significantly exercised during
Contacts 120 deg. apart this test. The supplier shallspecifyprobeapproach
Ball
rate, probe approach distance, settling times (if applica-
ble), and traverse speeds to be used for this test. These
parametersshall be representative of thoseused for
normalmeasurement onthatmachineandshallbe
explicitly made part of the test procedure. Repeatability
shall be determined, as in para. 5.3.2, foreach of the
test traverse speeds. Also, for large machines, itis
stronglyrecommendedthattherepeatabilitytest be
repeated at one additional work zone location for every
20 m3 ofworkzonevolume. If thetestisrepeated,
the repeatability shall be determined as above for each
of thepositionsspecified.
Table
5.3.4 Repeatability Requirements.Repeatabil-
ity as calculated in para. 5.3.2 shall not exceedthe
FIG. 22 TVPICALSETUP FOR supplier’s specification,derated as specified in para.
REPEATABILITY MEASUREMENT USING A 4.3, if applicable.
TRIHEDRALPROBE
5.4 Linear Displacement Accuracy
Tensets of four contacts each shall be made onthe
referenceball,the contact patternbeingthe same as 5.4.1 General.Completeverification of measuring
in para. 5.3.3.1. Coordinates oftheballcentershall machine accuracy isa difficult and time-consuming task.
be calculated for each set. Repeatability shall be deter- All practical tests, therefore, represent some compromise
mined as in para. 5.3.2. between the cost of testing and the cost of inaccuracy.
The secondtest is for seating probes. An inverted The tests described in thisStandardaremeant to
probe ofthetrihedraltypeshall be mounted in the represent a minimumrequirement to ensure confor-
probeholder,and ten measurements ofthereference mance to specificationand are not to be considered
ball locations shall be taken approaching the ball from comprehensive. If morethoroughtestingisrequired
different directions. (Trihedral sockets are preferred for the intendeduse ofthemachine,thensuch tests
over conical sockets, as conical sockets do not provide shall be negotiated between the user and supplier. For
uniqueprobe seating.) This setup is illustratedinFig. the purposes of this Standard, only one accuracy test
22. Repeatability shall be determined as in para. 5.3.2. is specified. This test is the measurement of the linear
5.3.3.3 Repeatability Tests for Machines displacement accuracy for allthree axes, using either
With Large Work Zones. For large machines, modi- a step gage or a laser interferometer. This test is meant
ficationstotherepeatability tests arerequired.Dueto to assess the conformance ofthemachine scales to
thelargemassesbeingacceleratedanddecelerated the international standards of length. In later sections,
when positioning largemoveable components, several the performance of themachineanditsgeometryis
dynamic effects may influence repeatability results. To assessed, independent of conformance tointernational
assess theimportance of such effects, allrepeatability lengthstandards.
testingshall be performedwithtwodifferentvalues
5.4.2 Step GageTest for Linear Displacement
for the test traverse speed, where the test traverse speed
Accuracy
is themaximumspeedachieved (not commanded)
during themachinemovementtopre-positionthe ma- 5.4.2.1 General.Using a step gage to check
chine immediatelypriorto contacting the precision measuringmachines is a time-honoredprocess.How-
reference ball for each touch. (This meansthatthe ever, the step gage is also usedbymany suppliers
machineshallreachthetest traverse speedbetween to calibrate (adjust) themachine scales, oftenatthe
probings.) These two values shall be, respectively, less installation site. Forthepurposes of checking confor-

21

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
ASME 889.4.1-1997 OF COORDINATE MEASURING MACHINES

mance to specification, it is advisablethata different Depending upon the details ofthe setup andthe
step gage be used for accuracy checking than the gage desired measurement,the cosine error can be either
used for machinesetup. Systematic gage calibration positive or negative. It is therefore importantto correctly
errors andthermal expansion coefficient uncertainties align the measurement apparatus in order to make this
are not thoroughly assessed unless a different gage is error negligible.
used. However,for a very large number of the coordinate 5.4.2.4 Measuring Interval. The measuring in-
measuringmachinescurrently supplied, the step gage, terval shall be nomorethan25 mm (approx. 1 in.)
when properly used, is many times more accurate than for axes of 250 mm (approx. 10 in.) length or less.
thebasicmachine. Therefore, thisrecommendationis For longer axes, less than 1,OOO mm (approx. 40 in.),
relaxed in those cases andshouldonlybefollowed theintervalshould benot less than25 mm (approx.
whenthehighestaccuracyisrequired.Inany event, 1 in.) normorethan 1/10 of axis length.Foraxes of
it should be clearly stated, as part ofthemachine morethan 1,OOO mm (approx. 40 in.) in length, the
specifications, if a different step gagethan the gage measuringintervalshall be nomorethan 1 0 0 mm
provided by the machine supplier is to be used for the (approx. 4 in.). For all axes, the entire travelalong
machine acceptance. It should be noted that step gages the axis shall be measured.
are not particularly useful for the evaluation of periodic
error. If periodic error is suspected and a step gage is 5.4.2.5 Measurements. Measurementsshallbe
to be used for these measurements, it isadvisedthat made with the primary type of probe specified for the
ametric step gage beusedonamachinewithinch principal mode of operation. The machine readout shall
scales andviceversa.Furthermore,sucha step gage be zeroed at the first step of the step gage. Three sets
shall meet the requirements of accuracy and calibration of measurements shall be made for each axis.Each
as specified in Section 7. setofmeasurementsshallbesequenced in thesame
direction of machinemotion,and each measurement
5.4.2.2Measurement Lines.Measurement shall be madebetweengage steps facing in thesame
lines for step gage tests shall be along three orthogonal direction. Since the data are to be averaged, these
lines throughthe center oftheworkzoneparallel to measurements may be taken with or without establishing
thethreeaxisdirections. a new zero at the start of each set. The value obtained
for the linear displacement accuracy will be the same
5.4.2.3 Mounting. The gage shall be mounted
in either case. The nominal mean of machine readouts
ontheworkpiece supporting surface in accordance
for each step ofthe gage shallbedetermined.
withthe step gage supplier’s recommendations.Itis
extremely important that the mounting be done properly, 5.4.2.6 Nominal Differential Expansion Cor-
as the accuracy of some types of step gages is strongly rection. The mean temperature of the step gageand
dependent upon proper mounting. Care must be taken the appropriate machine scale shall be measured during
to ensure the gage is properly supported and restrained the step gage measurement process for each axis. The
without distortion. The gage shall be aligned with the machine readings shall be corrected for the mean scale
machine axis (measurement line) withsufficient accu- temperature. The machine readings shall be the compen-
racy that cosine emf does not exceed 10% of machine sated values on a compensated machine. These values
workingtolerance for linear displacement accuracy. may or may not be whatisshownonthe display,
(Mathematical correction for misalignment is an accept- andthe supplier’s recommendations shall beused to
able alternative to mechanical alignment.) Cosine error determine which values apply. Similarly, the step gage
is caused by the angular misalignmentbetweenthe length must be corrected for the mean gage temperature.
measurement line and the displacement to be measured. This shall be done using the following expression:
The magnitude ofthe cosine error is given by the
followingformula: CMR = MR[I + K,(T, - 20) - KR(Tg- 20)]
where
CE = G212D
CMR = corrected machinereading
Kg = thermal expansion coefficientof calibra-
where tion gage
CE = cosine error K, = effective thermalexpansioncoefficient
D = measured displacement of machine scales
G = misalignment of gage with machine axis MR = machinereading

28

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FORPERFORMANCE EVALUATION
OF COORDINATE MEASURING MACHINES ASME 889.4.1-1997

Tg = gagetemperatureduringmeasurements,
5r
"C
T, = scaletemperatureduringmeasure-
ments, "C
5
5
41
3 ---""-- "_" ""

This equation was derived based on the assumption

that the machinewas adjusted tomeasurelength cor-
rectlyon the international scale if the machinewere Length, mm
at a uniform temperature of 20°C (68°F). If the machine -"
scales are adjusted at a temperature other than20°C
(68°F) and appropriate nominal differential expansion
corrections are notmade,thisequationmaynotbe
applicable. It should be noted that the preceding equation
assumes that the mean temperatures of the step gage(s)
and scale(s) remain constant duringthemeasurement
-31
-5
-4

FIG. 23TYPICALEXAMPLE OF LINEAR

process.Thisisnotalwaysthe case in a changing DISPLACEMENT ACCURACY DETERMINED
environment where both machine and gages havediffer- USING STEPGAGES
ent time constants. Proper attention shouldbegiven (In this case, the linear displacement
to keeping the environment as stable as possible for accuracy was approx. 4.8 km.)
the duration ofthetest.

5.4.2.7LinearDisplacementAccuracyfor 5.4.3
Linear
Displacement Accuracy
an Axis.Lineardisplacementaccuracy for a given Measurements UsingaLaser Interferometer.
axis at a step position shall be the difference between The laser interferometer is an extremely useful tool for
step gage calibration andthemean corrected machine measuring displacement accuracy. However, there exist
reading (MCMR) for that position. Displacement accu- certain machines today that are difficult to check using
racy is determined by takingthe difference between a laser because they are corrected for systematic errors
the step gage calibration and the meancorrected machine in their computer systems and the display readouts do
reading at each step, and then determining the maximum notreflect these corrections. Theuser of a machine
displacement error fromanypoint to any other point should confer with the supplier to ascertain the suitabil-
in the full travel. This is equivalent to determining the ity of these tests before making them part of a machine
maximumrange of themean differences. Evaluation specification.
of linear displacementaccuracy is illustrated in Fig. 5.4.3.1 Lines of Measurements. Lines of mea-
23, wherethe linear displacementaccuracy is clearly surement for laser interferometer tests shallbethose
labeled. The measurement of the zero point of the gage specified in para. 5.4.2.2 for the step gage.
shallalwaysbe included.
5.4.3.2Alignment.Thelaserinterferometer
5.4.2.8 Staging. Where the step gage is shorter shall be mountedinsuch a fashion as to measurethe
thanan axis, thegage shall be staged. (If staging is relativemotionbetweentheramandtheworkpiece
required, it is the recommendation of this Standard that supporting surface. Particular attention shouldbepaid
the linear displacement accuracy be measured with the to cosine error, and alignment shall be such that cosine
laser interferometer (para. 5.4.3)rather than with a step error is less than 10% oftheworking tolerance ofthe
gage). In thestaged position, a step ofthegageshall axis under test. Deadpathshould also beminimized.
be set attheapproximateposition of the final step of A typical laser setup for linear displacementaccuracy
the original position, andthemachine shall bezeroed isshown in Fig. 24.
at that step. Corrected machine readings shall be deter- 5.4.3.3
Wavelength Correctionand
mined as before. The error of the last step ofthe gage Nominal Differential ExpansionCorrection. In
in thepreviousposition shall be algebraically added order to obtain proper results, interferometers must be
tothe error for eachsubsequentstaged position as a corrected for air temperature, air pressure, and air
further correction. The step gage shall be staged in a humidity.The correction shall be computedaccording
similar manner as many times as required to cover the tothefollowingequation for a laser measurement
completetravel of that axis. system set to read correctly at 20°C (68"F), 760 mm

29

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
ASME B89.4.1-1997 OF COORDINATE MEASURING MACHINES

n V = partial pressure of water vapor in mm Hg3

T, = mean air pressure, "C

The preceding equation is a linearization of the Edlén

equationandis accurate toapproximately 0.1 ppm.

W Other forms of thisequationare equally accurate and

are considered suitable for the purposes of this Standard.

W In order to compare the corrected laser reading to the

machinereadings,themachinereadingsmust also be
Retro1-eflecior . Remote
interferometer
corrected for temperature. The corrected machine read-
ings are given by

CMR = MR[ 1 + &(T, - 20)]

in the case wherethemachinewas set up at 20°C

(68°F). Variablesaredefined in para.5.4.2.6.
v beam If thelaser interferometer used has environmental
compensation features, the supplier's recommendations
FIG. 24 TYPICALSETUPFORTHELASER regarding the use of these accessories shall be followed,
TEST FORLINEAR DISPLACEMENT with the air temperature sensor nearthe laser beam
ACCURACY path and the material sensor placed on the appropriate
machine scale. It is a requirement of this Standard that
independent calibration of the temperature and pressure
Hg air pressure, and 10 mm Hg partialpressure of sensors of such compensation devices be performed on
watervapor.* a regularbasis(see Section 7).
5.4.3.4 Measuring Intervals.Measuringinter-
CLR = LDR[ 1 + K,(T, - 20) - K J P , - 760) + Kh( V - IO)] vals shall beno larger thanthosespecified in para.
5.4.2.4 for step gage measurements;however, due to
where ease of measurement using laser systems, smaller inter-
CLR = corrected laser reading valsarestronglyrecommendedwiththose intervals
Kh = coefficient of refractive index change due being chosen such that they are not even multiples of
to atmospheric humidity. The current best the machine scale spacing. With a laser interferometer
value is0.05 ppdmm Hg partial pressure it isparticularlyeasy to check for periodic error by
of water vapor. Because of the low value measuring a largenumberofcloselyspaced displace-
of this coefficient, the atmospheric hu- ments over an intervalequaltotheperiodicityofthe
midity can be neglected for most applica- machine scale. Althoughthismeasurement isnot a
tions. requirement of this Standard, it can yield useful infor-
Kp = coefficient of refractive index change due mation.
to atmospheric pressure. The current best
valueis 0.36 ppm/mm Hg pressure. 5.4.3.5 Sets of Measurements. Three sets of
Kt = coefficient of refractive index change due measurementsshall bemadealong each measurement
to atmospheric temperature. The current line, all in the same direction. The sets of measurements
best value is 0.93 ppm/"C. maybe takenwith or withoutrezeroingthemachine
LDR = laserdisplay reading andthe laser. For eachmeasurementpoint,the mean
P,,, = air pressure, mmHg of differences betweencorrectedmachineandlaser

The partialpressure of watervaporcan be calculated fromthe

' This form of equation also assumes atmospheric air with
the relative humidity by multiplying the saturated
vapor
pressure
at
a
normalmixture of gasses. Atmospheres that deviate significantly, particular
temperature by the relative humidity expressed as a
particularlyinregardto COZandaromatichydrocarbon concentra- fraction. The saturatedvaporpressureat 20°C (68°F) is 17.6 mm
tion, have been observed and can lead to measurableerrors.Ifthis Hg. Thus, for example, 50% relative humidityat 20°C would yield
situation is suspected, appropriate correction should be applied. partial
a pressure of 0.5 x 17.6 = 8.8 mm Hg.

30

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
OF COORDINATE MEASURING MACHINES ASME 889.4.1-1997

20
--
16
--
12
m
-""""""""""-""""" X""
8 t
4
accuracy
E,
2 0
e
L
"
""
""
-L
""
,
w

-4 -
L

-
-8-
-
--
-12
--
-16
-
-20 I I I I I I I I I I I I I I I I
O 100 200 300 400 500 600 700 800 900
Position, prn

FIG.25TYPICALRESULTSOFALINEARDISPLACEMENTACCURACYTEST USING THE

LASER WITH THELINEARDISPLACEMENTACCURACYCLEARLYLABELED
(For this example, the lineardisplacementaccuracy is approx. 12 p.m.)

readouts shall be calculated. Linear displacement accu- such as gage blocks, for the ball bar will be considered
racy shall be the maximum spread of the mean differ- in conformance with this Standard if these artifacts are
ences ofthe individual points. Thisisillustrated in of equivalent length and are measured in the positions
Fig. 25. specified for theballbar test (para. 5.5.2.1). Theuse
of such calibrated artifacts for these tests doesgive
5.4.4 Linear Displacement Requirements. Lin- additional information, but also incurs additional ex-
ear displacement accuracy, as calculated in para. 5.4.2.7 pense.
or 5.4.3.5, shall not exceed the supplier's specification,
Ball bars provide a rapid and easily understood check
deratedasspecified in para. 4.2, if applicable.
of machine volumetric performance. Properly conducted
ballbar tests allow precise comparisons ofthelength
5.5 VolumetricPerformance
scales on the various machineaxesand clearly point
5.5.1 General.Complete testing ofthevolumetric out deviations of machinegeometryfrom perfection.
performance of coordinatemeasuringmachinesis a They are also extremelyuseful for quicklyrechecking
difficult andtime-consuming process. ThisStandard a machineon a periodic basis. Inno case should
hasattempted to reducethetimeand cost associated theballbar tests alone be regarded as providing a
with testing by providing, wherever possible, simple measurement of machine accuracy. In this Standard,
self-checking procedures using measurements of uncali- accuracy is assessed in the linear displacement accuracy
brated artifacts. The primary uncalibrated artifact is the section (para. 5.4). In theballbar tests, as in the
ball bar. Specificationsthat substitute calibrated artifacts, repeatability and linear displacement accuracy tests, one

31

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCEEVALUATION
ASME 889.4.1-1997 OF COORDINATE MEASURING MACHINES

should expect that the precise value of error obtained at several positions along a long measurement line, in
is dependent on the particular mode chosen for that test. order to require nearly full travel of the machine along
Due to thepracticaldifficulty in transporting and thatmeasurement line. Patterns for machineswith a
usingverylongballbarsand in subdividing very single long axis (axis ratio 2: 1:l) are shown in Fig.
large workzones into many subvolumes,significant 27. These patterns require measurement of the ball bar
modifications to thenormalball bar procedures are in 30 locations. Sample patterns for machines with two
provided in para. 5.5.4 for machineswithlargework long axes andone short axis (axis ratio 2 2 : 1) are shown
zones.Herethe laser interferometer is introduced be- in Fig.28.Thesemachinesrequire 35 measurement
cause of its ability to measure over very long lengths. positions. The patterns were chosento provide maximum
The laser interferometer may give a different range of sensitivity to most angular and squareness errors. They
valuesthanwould a longballbar;however,these do not completelycheck angular motionsoftheram
numbersshould be representative of the machine’s axis, thus a separate testisprovided in para. 5.5.3 to
volumetricperformance. assess ram axis angular error effects when using offset
Users of this specification should also be aware that probes. Additionally, articulation of the probe head and
as the work zone aspect ratio increases on a machine, length of the stylus during thistestcansignificantly
the sensitivity of thesetests to the straightness of impactthe results of thetest.It is therefore recom-
the longest axis on thatmachineisreduced.Where mendedthatsucharticulationandlength changes be
straightness is critical, a separate check of this parameter minimized. (Articulating probe systems are tested in
should be performed whenthe aspect ratio of the para.6.2.)
machine axes exceeds 4: 1. The figuresshown are idealized and, onany given
This section on volumetric performance also contains machine, it ispossiblethattheballbar positions
performance tests for machineswith a rotary axis. shown will overlap. It is recommended that if positional
These testsfollowthe same philosophy in thatno overlaps betweenballbarsetups exceed 60%of the
calibrated artifact isused. ballbar length, then one ofthe overlapping setups
maybe eliminated. Most existing cases can be readily
5.5.2 Volumetric Performance Procedures obtained by rotations of the configurations in the figures.
UsingBall Bars No detailed recommendations are made regarding ball
5.5.2.1 General Patterns.The ball bar perform- bar fixturing; however, a limited discussion of fixturing
ance testsrecommended by thisStandard may be alternatives isgiven in Appendix G, and a sample
accomplished using a single ball bar of length slightly fixture used for holding a ball bar with both ends free
shorter (approx. 1 0 0 mm)thanthe least dimension of isshown in Fig.29.
thework zone! For nearly cubic machines,this ball Care should betaken in handlingballbars so that
bar is measured in 20 positions. The general approach heatfromthehandisnottransferredtotheballbar.
is to position the bar along 10 of the 12 edges of the The useof a plastic insulating sleeve ishelpful. The
work zone, along at least six work zone face diagonals time constant for thermal equilibration of a hollow
to require simultaneous motionof pairs of machine steelball bar is approximately 20 min (see ASME
axes, and along the four work zone body diagonals to B89.6.2, Temperature andHumidity Environment for
require simultaneous motion of all three machine axes. Dimensional Measurement, for an explanation of ther-
Recommended patterns for nearly cubic machinesare mal time constants). Typical ballbarswill stabilize
given in Fig.26. (The figures showing patternsare within about one hour after beingbrought into a
oriented for vertical ram machines.Theyshould be temperature-controlled environment.
rotated for horizontal ram machines.) For machines
havingworkzoneswith different aspectratios,the 5.5.2.2 Setup and Measurement Pro-
procedure still usesthe shorter.ballbarbutplaces it cedure - Ball BarTests. The ballbar shall be
suitably fixtured in the positions indicated for measure-
ment so that probing access to both balls is available.
As withother sections in thisStandard,the user isallowed to A fixturebasedon a knuckle joint is shown in Fig.
specify measurements different thanthe default option. To be in
compliance with this Standard, the user may specify measurements 29. Such a fixtureshould be portable so that it can
of ballbars in up to 40 different locations and is alsoallowed be easily moved around the table and sufficiently rigid
to specify up to 3 ball bar lengths. These positions and lengths must so thattheballbarwillnotsignificantlydeflect or
be clearly stated as part of the machine specification. Furthermore, if
more than one length is specified,each length ballbarmust be vibratewhilethe locations of the balls are being
measured in atleast 10 different positions. measured.Foreachofthepositionsspecified in the

32

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
OF COORDINATE MEASURING MACHINES ASME B89.4.1-1997

Location 1 Location 2 Location 3 Location 4

Location 5 Location 6 Location 7 8 Location

Location 9 Location 10 Location11Location 12

Location 13 Location 14 16Location15Location

Location 17 Location 18 Location 19 Location 20

33

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
ASME 689.4.1-1997 OF COORDINATE MEASURING MACHINES

Location 1
@ Location 2
Loution 3 Location 4

Louation 5 Location 6 Location 7 Location 8

Location 9 Location 10 Location 11 Location 12

Location 13 Location 14 Location 15 Location 16

Location 17 Location 18 Location 19 Location 20

Location 21 Location 22 Location 23 Location 24

Loution 26
@ €2§7
Location 26 Loution 27 Location 28

Location 29 Loution 30

FIG. 27 RECOMMENDED BALLBAR POSITIONS FOR MACHINES WITH ASINGLELONGAXIS

(The axis ratio here is 2:l:l. On machines with slightly different ratios, overlapping patterns
are recommended.)

34

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 STD.ASME B87.Y.L-ENGL 1777 m 0759b70 0583070 O28 W

METHODS FOR PERFORMANCE EVALUATION

OF COORDINATE MEASURING MACHINES ASME 689.4.1-1997

Location 1 Location 2 Lncation 3 Location 4 Location S

Location 6 Location 7 Location 8 Location S Location 10

-
Location 11 Location 12 Location 13 Location 14 Location 15

Location 16 Location 17 Location 18 Location 19 Location 20

Location 21 Location 22 Location 23 Location 24 Location 25

Location 27 Location 2 8 Location 29 Location 30

Location 31 Location 32 Location 33 Location 34 Location 35

FIG. 28 RECOMMENDED BALLBAR PATTERNS FOR A MACHINE WITH TWO LONG AXESAND
ON€ SHORT AXIS
(The axis ratio here is 2:2:1. Again, for machines that do not quite correspond, overlapping
positions are recommended.)

35

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
ASME B89.4.1-1997 OF COORDINATE MEASURING MACHINES

thefigure, or simplythetotalrange in thevalues in

the table. In cases wherethere appears to be a single
(or several) outlying point(s) that does not conform to
the general trend, it is recommended that this measure-
ment be repeated.
The recommended procedure for checking the repeat-
ability of a ballbarmeasurementis as follows. The
ballbar shall be measured twice in the suspected
position. If themeasurements agree within twice the
repeatability (para. 5.3), thenthefirstmeasurement
shall beused andthesecondmeasurement discarded.
If the measurements do not agree withintwicethe
repeatability, both are discardedandtheprocedureis
repeated. This procedure may be repeated three times;
at the end of which time, if repeatability has not been
obtained as defined above, the test shall be discontinued
and the fault determined and corrected. After correction
of theproblem,therepeatabilitytestandtheballbar
testmust be rerun in their entirety.
5.5.3 Offset Probe Performance Test. The an-
gular motion of the ram axiswasnottested by the
preceding procedure (see para. 5.5.2). This motion is
of particular importance when probeswith different
offset lengths are used. The following test is designed
FIG. 29 SAMPLEFIXTURE FOR HOLDING A to evaluate the machine performancewhen offset probes
BALLBAR WITH BOTH ENDS FREE are used. Although theillustrations show vertical spindle
machines, thistest applies equally to horizontal arm
patterns,both ends of the ballbarshallbemeasured. machines.
At least four probe contacts must be made on both of 5.5.3.1 Ball Bar Tests for Offset Probe Per-
theballs in order tomeasuretheballbar length. To formance. The ball bar can be used to place tolerances
achieve better accuracy, eight or morepointsshould onthemagnitudeof offset probing errors by using a
beusedto determine the center of each ball. These probe with a large offset. A typical test setup is shown
pointsshould be dispersed aroundtheball as far as in Fig. 31. The probe offset lengthshall be set at a
the probing system allows. (To check the repeatability reasonableamount [approximately 150 mm (approx.
of a setup, itis advisable tomeasuretheballbar 6 in.)isrecommended],theprobeshallbe oriented
lengthseveraltimes,butthisisnot a requirementof perpendicular to boththeballbar axis and the ram
this Standard.) From these probings, center coordinates axis, andmeasurementsshall be madeoftheballbar
of the ballsandthelength of theball bar shall be lengthwiththe offset probe,firstwiththeprobe in
calculated for eachballbarposition. The total spread one positionandthenrotated 180 deg. aboutthe ram
of calculated ball bar lengths shallbe assessed following axis with respect to that position. In performing these
theprocedure described in para. 5.5.2.3. The range of measurements, the ball bar may be moved to a second
these lengths shall not exceed the supplier's specifica- position, withnominallythe same angle withrespect
tions,derated as specified in paras. 4.2 and 4.3, if to the ram and probe offset axes, rather than reposition-
applicable. ing the cross-slide of the machine, as is shown in Fig.
5.5.2.3 Ball Bar Data Analysis. The data from 30. These twoprocedures may give different results.
ballbarmeasurements are analyzed by preparing a Whenthe cross-slide ismoved, this movement may
simple plot or a simple tableofthe deviations in the tilt the ram axis andlead to different results. For the
ball bar length without regard to measurement location. purposes of this Standard, either procedure is allowed.
An example of a scatter plot is given in Fig. 30. The Note that when offset probes are used, it is extremely
workingtoleranceof the machine is defined as the importantthatthey be properlybalanced so as notto
range of data in such a plot, as is clearly indicated in place undue moments on the ram. The default ball bar

36

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
OF COORDINATE MEASURING MACHINES ASME 889.4.1-1997

4l
3 -
o
""""""

o
"
6

o O
o
7
""

- o
o
- D
Working tolerance
U o
U o
-
o
- o o o

1
I
o o
"" & """"_""" "-
-4
-3
I I I I
a 5 10 15 20
Position number

FIG. 30 BALLBAR TEST RESULTS

(In this example, working tolerance is approx. 7 Fm.)

length for thesetestsshallbetheballbarlengthused 5.5.4 Volumetric Tests for Machines With

for thevolumetricperformancetest (para. 5.5.2). Large Work Zones. The goal in testing large machines
The ballbar shall bemeasuredin four locations. remains the realistic estimation of the expected accuracy
Theuser is free to chooseany four positions within of machinemeasurementsunderreal operating condi-
themachinevolume for theballbarmeasurements tions. To accomplish this on large machines,limited
with offset probes; however,the default positions, which ball bar testing is supplemented with specific additional
are most sensitive toram axis angularmotion, are linear displacementaccuracy tests. Thetotalnumber
shown in Fig. 32 for a verticalrammachine. In each ofballbarmeasurementsis IO plusone additional
position, theballbaris at 45 deg.totheram axis. position for each LO m3 of additional work zone volume
Two locations are sensitive toramaxisrollandyaw over 20 m3.The length of the ballbarisfixed,and
andtwo locations are sensitive to ram axis roll and shallbe 0.9 m (approx. 35 in.) for allmeasurements.
pitch. (Theusershouldbecautionedthatballbar The locations oftheballbarmeasurements shall be
positions, where the ballbarlengthisnearlyparallel specified in advance by mutual agreement between the
or perpendicular to the ram axis, are insensitive to ram supplier anduser.
axis roll.) Differencesbetween lengths measuredwith Additional linear displacementaccuracydata shall
thetwoprobe offsets shall becomputed.The results be collected, as specified in para. 5.4, along six supple-
are calculated as the ratio of these differences to twice mentary measurement lines. Four of the supplementary
theprobe offset length. The absolute valueof the measurement lines shall be the bodydiagonals of the
largest calculated ratio is reported as the offset probe work zone. A typical setup for these measurements is
performance andshall not exceed thesupplier's perform- shown in Fig. 19. Theremainingtwomeasurement
ance specifications, derated as specified in paras. 4.2 lines shall be one line parallel to each ofthenon-ram
and 4.3, if applicable. (It should be noted that this test axes. The location of each of these latter two lines in
is not a parametric test; thoseuserswishing to assess theworkzone shall bechosen to maximizethe offset
ram axis roll, specifically, are referred to Appendix H.) distance from the line to the position measuring system

31

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
ASME 889.4.1-1997 OF COORDINATE MEASURING MACHINES

U
Position 2

' FIG. 31 TYPICALSETUPFOROFFSETPROBEPERFORMANCETESTING

of the axis parallel to that line. The measurement points Methodsspecifiedin these sectionsare applicable
foreachlinechosenshall be atthepointspacing to a widevariety of duplexinstallations.Examples
specifiedin accordance withpara. 5.4.2.4. Therange arelarge and smallmachines,all applicable machine
of ball bar measurements and the greatest range of the configurations,machineswithindividual rotary tables
six supplementary linear displacement measurements or a shared rotarytable,machineson opposite sides
shallboth be determined. Neither shallexceedthe of a shared fixedtable,andmachineswith a shared
supplier'sspecifications, derated asspecified in paras. primaryaxis.The general principlefortesting duplex
4.2 and 4.3, if applicable. installations is thateachindividualmachineshall be
testedas a separatemachine,thentherelationship
5.5.5 PerformanceTests,Machines Used in between the two machines shall be tested by measure-
the Duplex Mode ment of duplexperformance.
5.5.5.1 General. Paras. 5.5.5.2 and 5.5.5.3 spec-
ify methods for measuring the performanceof machines 5.5.5.2 Tests of Individual Machines. The
operating in the duplex mode.Applicability ofpara. testsforindividualmachinesshall be the appropriate
5.5.5.3 shall be limited to installations where the work tests of this Standard.
zones of the two machines overlap. For other installa- When one of two machines used in the duplex mode
tions, it isrecommendedthatspecialtestprocedures isindividuallytested,the other machineshall be in
be developedand agreed uponbetweenthesupplier motionin a manner recommended by the supplier for
anduser. duplex measurement of a workpiece. Thus, if the sup-

38

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
OF COORDINATE MEASURING MACHINES ASME 889.4.1-1997

FIG. 32 DEFAULTBALLBAR POSITIONS FORTHEOFFSETPROBE PERFORMANCETEST

ON A VERTICAL RAM MACHINE

plierrecommendsno limitations onmotionofone coordinates for thefirstmachine shall beposition

machinewhilethe other ismeasuring, there are no coordinates of the probe tip reported by the firstma-
limitations onmotionduring testing. If the supplier chine. All ball center coordinatemeasurements shall
recommends one machine be at rest when the other is be performed by the methods specified in para. 5.5.2.2.
measuring, then it must be at rest during testing. Within The rammountoptionis generally preferred where
this limitation theusermaychoosethemanner of it can be used. Because there is only one set of probing
motion. errors perpairofball center determinations, it gives
5.5.5.3 DuplexPerformance Test. There are a better indication of the relationship between the axis
two options for conductingtheduplexperformance systems ofthetwomachines.Furthermore, it is faster
test: table mountandrammount.The supplier shall and requires less hardware. It cannot, however, be used
specify whichoptionisto be usedunlessotherwise onsomemachineswithbuilt-in proportional probes
contractually agreed upon between the supplier and user. becausethere is no suitable method for mountingthe
For the table mount option, a precision reference reference ball; additionally, itcannotbeusedonma-
ball shall be supported in successive specified positions chines that do not have a triggering means for reading
fromthetable or other workpiecesupport surface. thefirstmachine position.
Optionally, anarray of balls may beused. In each Forboth options, thetwomachinesshall either be
position, center coordinates ofthe reference ball or operated in the same coordinate system, or ball center
balls shall bemeasuredbyboth machines. coordinates measured by onemachine shall betrans-
For the ram mount option, a precision reference ball formed into the coordinate system of the other machine.
shallbemountedtotheramof a firstmachine in the Ball positions shall be in a planedefined by two
positionnormallyoccupied by theprobe tip. Thefirst perpendicularmachineaxes in theoverlapportion of
machineshallbemovedtosuccessivespecifiedposi- the twowork zones. A planehavingmaximum area
tions, and center coordinates oftheball shall bemea- shall be chosen. In the interest of clarity, the machine
sured by thesecondmachine.Measuredball center specificationshallcontain a description of theplane

39

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 STDDASME B 8 7 . q - L - E N G L 1777 W 0759b70 0583075 bOT m

METHODS FOR PERFORMANCE EVALUATION

ASME ~89.4.1-1997 OF COORDINATE MEASURING MACHINES

position. For planes of 1 m* or less,thereshall be 9 TABLE 1

ball positions. Forplanes of 1 to 4 m2,thereshall be LOCATION OF THE REFERENCE SPHERE
9 ballpositionsper m2. Forplanes over 4 m2,there ON THE ROTARY TABLE
shall be at least 36 ball positions. The user may specify Heights Radii
the ball positions. The default pattern specification shall H, mm R, mm
bethe intersections of a grid of squarescovering
substantially the entire plane. 200 200
400 200
For each reference ball position, center coordinates 400 400
measured by the first machine shall be subtracted from ao0 400
center coordinates measured by the second machine.The 800 800
range of such differences shall be duplex performance.
5.5.5.4DuplexPerformanceRequirements.
Duplexperformance, as calculated in para. 5.5.5.3, referred to ASME B89.3.4M-1985 and ASME B5.54"
shall not exceed the supplier's specification, derated as 1992, Methods for Performance Evaluationof Computer
specified in paras. 4.2 and 4.3, if applicable. NumericallyControlledMachine Centers.
5.5.6.2 Setup. At each of the locations chosen
5.5.6 Volumetric PerformanceTestfor DCC for theperformanceofthistest,therotarytableshall
Machines With a Rotary Axis bealignedfollowingtheproceduresspecified bythe
5.5.6.1 General. The performance test described supplier for table alignment duringnormal four-axis
in this section is applicable only to four-axis machines measurements on the machine.This alignment is critical.
withthreelinear axes plus a rotary table. This test is Next,twoprecisionspheresaremountedtotherotary
performed in addition to the other tests in Section 5, table in the position shown in Fig. 33. H, and R, shall
which are conductedwithoutthe useoftherotary be thelargest applicable correspondingvaluesfrom
table. The usershould be aware thatthistestonly Table 1. The value of H, shall be such thatthehigh
sphere can be measured by the machine. The value of
indirectlyassessesperiodic error in therotary scales.
Users particularly concerned with this error are advised R, shall be less than thetableradius R, andshallbe
toperformthe appropriate parametric calibration. such that the low sphere can be mounted on the table
Some coordinate measuring machines allow alterna- nearthetablesurface. R, ishalfthe diameter of the
tive locations and/or orientations of therotarytable. rotary axis asdefinedinthe Glossary. Inthe case
For such machines, the specified working tolerance for where axis motion limitations preventuse of the H,
this test must be metwhenthetestisconducted at value from Table I , H , shall be the maximum distance
anypermittedtablelocation and/or orientation ofthe allowed by the machine geometry.The precision spheres
rotary table within the machine work zone. The default shall meet the requirements for precisionreference
option for this test is that the user may select any one balls, para. 7.3.3. If an extensible ball bar conforming
position for theperformance of thistestfromthe tothe requirements of Appendix G wasused in the
previously specified permitted locations and/or orienta- preceding tests, the balls and extensions in this set can
tions.Performance may also be tested for morethan be readilyadaptedtoperformthis function.
one position and for a preferredrotarytablelocation 5.5.6.3 Measurement Procedure. The general
and orientation, if agreed. procedureis as follows. The rotarytableisrotated
The workingtolerances for a rotaryaxisarecalled through a series of angularpositionsandtheposition
the 3D/alpha working tolerances. They are determined of one of the two spheres is measured. Recommended
by usingall four axes of themachine to measurethe sets of angular positions, inwhichthe rotary tableis
centers of the two spheres on the rotary table, and by completely accessible to theCMMandwhere it is
analyzing the ranges of the measured center displace- partially accessible, are given in Table 2. (Certain
ments for each sphere. Since thistestdoes not use a measuringmachinesusetherotarytable as a means
calibrated artifact, it does not directly check accuracy; of extending the measuring range, so that only a portion
rather, it checks a complexcombination of rotary of therotarytableis accessible to a probe.Onsuch
table geometry, rotarytablealignment,probing,linear machines,the starting angularpositionof the rotary
accuracy, andmeasuringmachine coordinate transfor- table must be such that the positions of both balls can
mation algorithms. Usersinterested in analyzingthe be measured.) The angular positions given are default
geometry of the rotary table as a separate element are values. The usermay choose these or anyotherset

40

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
OF COORDINATE MEASURING MACHINES ASME 689.4.1-1997

FIG. 33 DIAGRAM OF TESTBALL POSITIONS FOR THEPERFORMANCETEST

ON A ROTARY AXIS
(The X, V, and Z designations in this figure are used to illustrate directions with respect
to the rotary table base and do not necessarily represent machine axis designations.)

of values, provided the chosen set is made part of the B set as the zero datum in this system, the X , Y plane
specification and contains thesamenumber of points. set normal to the rotary table axis, and the Z , X plane
All results ofthe measurementsarereported in a part set through the measured sphere centers. The coordinates
coordinate system. Usershavingmachineswithoutthe of the center of sphere A are calculated in this system.
software to enable this transformation can still perform The rotary table is then rotated 13 times to different
this test; however,the analysis iscomplexand isnot nominal angular positions (for example, those given in
included in thisStandard. Table 2). At each angular position,theposition of
In the following discussion, X, Y, and 2 are directions sphere A is measuredusingtheappropriateprobing
relative to therotarytableillustratedinFig. 33. They sequenceand these positions are recorded as in Table
do not necessarily correspond to the supplier’s labeling
2.Whenthe rotarytable is returnedtoitsstarting
of themachineaxes.
angularposition (point 14 in Table 2), thelocations
The following procedure shall be followed. The rotary
table isput in the starting position,positionzero (O) of both spheres are again measured, retaining the origi-
in Fig. 34, andthe positions of sphere A (the low nal datum on sphere B. The apparent X , Y, 2 positions
sphere) andsphere B (the high sphere) are measured of the centers of spheres A and B are calculated and
usingtheproceduresdescribed in para. 5.5.2.2 for recorded. The rotarytable is thenrotated 13 more
measuringthe center coordinates of a sphere during times to thepositions as shown in Table 2 andFig.
the ballbartest. The machine software is set so that 34. At each angle the apparent X, Y, Z position of the
themeasurement results willbereportedinthepart center of sphere B is recorded. (Note that since the
coordinate systemwiththemeasured center of sphere part coordinate system is being used with sphere B as

41

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
ASME 889.4.1-1997 OF COORDINATE MEASURING MACH.lNES

TABLE 2
DEFAULT NOMINAL ANGULAR POSITIONS AND SAMPLE DATA SHEET FOR OBTAINING
VOLUMETRIC PERFORMANCE WITH A ROTARY AXIS
Al A2 Sphere A Sphere B
eg. Point
No. [Note (111 [Note (211 XA YA ZA XB YB ZB

O O O -=O O O
1 75 135 D
l ... ...
2 125 225 a 2 ... ...
3 175 315 a 3 ... ...
4 385 405 a 4 ... ...
5 410 540 ZA 5 ... ...
6 510 630 ZA 6 ... ...
7 820 810 ZA 7 ... ...
8 510 630 "A 8 ... ...
9 410 540 ZA 9 ... ...
10 385 405 ZA 10 ... ...
11 175 315 -= 1 1 ... ...
12 125 225 2% 12 ... ...
13 75 135 a 13 ... ...
14 O O 14 14 YB 14
15 -75 -135 ... XB 15 YB 15
16 -125 -225 ... XB 16 YB 16
17 -175 -31 5 ... XB 17 YB 17
18 -385 -405 ... XB 18 YB 18
19 -41 O -540 ... XB 19 YB 19
20 -510 -630 ... XB 20 YB 20
21 -820 -810 ... XB 21 YB 21
22 -510 -630 ... XB 22 YB 22
23 -41 O -540 ... XB 23 YB 23
24 -385 -405 ... XE24 YB 24
25 -175 -315 ... x825 YB 25
26 -125 -225 ... XE26 YB 26
27 -75 -135 ... XB27 YB 27
28 O O ZA 28 XB 28 YB 28
GENERAL NOTES:
(a) In this table, an ellipsis (. . .) means that no measurement is made of the location of that sphere in that angular position

the datum, its measured center wouldremain at zero values obtained for spheres A and B are plotted on
on aperfectmachine.) similar scatter plots. The range of the X values for
The rotarytable isthenreturned to itsoriginal spheres A and B are compared, andthelargestrange
position and spheres A and B are remeasured, retaining reported as the working tolerance for 3D/alpha radial
the original datum on sphere B. The apparent X , Y,2 performance. Similarly, the greater oftheranges of
positions of thecenters of spheres A and B are calculated the Y values for spheres A and B is reported as the
andrecorded. working tolerance for 3D/alpha tangential performance.
Users are free to select their own angular positions Finally, the greater of the 2 ranges for spheres A and
for the performance ofthistest, as long as thesame B isreported as theworking tolerance for 3D/alpha
number (14) of angular positions is included and rota- axialperformance. The geometric meaningofthese
tions exceeding 360 deg. are allowed. parameters is conceptuallyillustrated inFig. 35.
5.5.6.4 Rotary Table Performance Data
Analysis. The data set isanalyzedtoobtainthe 3D/ 5.5.6.5 Rotary Table Performance Require-
alpha working tolerance. Here theX , Y and Z coordinate ments. 3D/alphaperformanceshallnotexceedthe
42

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 ~

METHODS FOR PERFORMANCE EVALUATION

OF COORDINATE MEASURING MACHINES ASME 689.4.1-1997

7 1.13

Point Nos.
4@10 /-

1-125 1-75 4-385 I 3-175

Al - CMM has access to '/2 of rotary table

7,20,22

4-05 15 3-3
1-135 2-225

A2 - CMM has access to full rotary table

FIG. 34 DEFAULT POSITIONS FOR SPHERE LOCATIONS ON THEROTARY AXIS

PERFORMANCE TEST

43

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
ASME 689.4.1-1997 OF COORDINATE MEASURING MACHINES

Tangential
Radial w o r k ï 9
tolerance x working
tolerance

I working Axial
I
I ! I tolerance

/ I .
T-
Rotary table

GENERAL NOTES:
(a) R, is the radius of Sphere B from the rotary table center.
(b) H, is the height of Sphere B from the rotary table face.

FIG. 35 DIAGRAMSCHEMATICALLYREPRESENTINGTHE MEANINGS OFTHE

RADIAL, TANGENTIAL, AND AXIAL WORKING TOLERANCES FORTHEROTARY AXIS
PERFORMANCE TEST

supplier’s specification,derated as specified in paras. their intended application. In cases where a specialized
4.2 and 4.3, if applicable. machineload is required,the supplier anduser shall
agree upon alternate specifications.
5.5.7 Performance Testing Coordinate
Measuring Systems Under Loaded Conditions. 5.5.7.1TestingProcedure -
Acceptable
The following procedures for evaluating machine load Machine Loading. The general requirements for .this
effects (anddefinitions supplied in the Glossary) are
testincludethe following:
intended to be used primarilyfor informational purposes.
(a) Weight used to perform testing shall not exceed
They will allow the purchaser of a coordinate measuring
machine to better understandtheresultofutilizing themaximum acceptable machineloadspecification.
incorrectloadingmethods or overloadingtheCMM. (b) The CMM supplier shall perform tests that com-
In general, this procedure is not intended to beused ply with all procedural requirements and shall meet or
as an “acceptance” test at the time of machinepurchase. exceedspecifiedperformancelevels.
It should be usedto differentiate machine models and (c) The physical volume of the weight supplied for
their relative “robustness” underload,and to ensure testing must lie within the measuring cube of the CMM
thatmachineshave appropriate weight capacities for andtheweightmustbe free-standing.

44

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
OF COORDINATE MEASURING MACHINES ASME 889.4.1-1997

3D/alpha radial error

10

.
s 5
E

i
e
h o
0 ABall
B Ball

-10
O 4 8 12 16 20 24 28
Position

BD/alpha
error tangential error
3D/alpha axial
35 20
1
ml""- """""""" .J 1 15

10

-1 o

-1 5

-15 I I 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 I -20 1 1 1 1 1 1 l l l l l l l l l l l l l l l I I I 1 1 1 1 1 J
O 4 B 12 16 20 24 28 O 4 8 12 16 20 24 28
Position Position

FIG. 36 TYPICALRESULTS OF AVOLUMETRIC PERFORMANCETESTFOR A DCC MACHINE

WITH A ROTARY AXIS
(The 3D/alpha radial, 3D/alpha tangential, and 3D/alpha axial working tolerances are clearly
labeled on the graphs.)

( d ) Theloadatanyspecific contact point will be ( a ) Place the test weightonthemachine.

no greater than twice the load of any other contact point. ( b ) Perform the repeatability test as described in this
( e ) The center of gravity of the machine load must Standard (para. 5.3), with theexception of location.
liewithinthe CG location zone. Locationis optional in this test.
(0 The specific test load must fall within acceptable (c) Performsixballbarmeasurements, as physical
machine load limits, as defined by the Load Concentra- constraints allow, selected fromthefollowingeleven
tionChart (Fig. 3). user-selectable positions:
Thefollowing steps should betaken for thetest (1) (four) 3D diagonals (as available);
procedure. (2) planar diagonal (front);

45

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
ASME B89.4.1-1997 OF COORDINATE MEASURING MACHINES

( 3 ) planar diagonal rear (opposite orientation); 5.5.8 Volumetric Performance Requirements.

( 4 ) planar diagonal (top); Volumetricperformance, as calculated in paras. 5.5.2,
( 5 ) planar diagonal (left side); 5.5.3, 5.5.4 (if applicable), 5.5.5 (if applicable), 5.5.6
(6) planar diagonal (right side - opposite orienta- (if applicable), and 5.5.7 (if applicable) shall not exceed
tion); and the supplier’s specifications, derated as specified in
(7) two orthogonal linear axes. paras. 4.2 and 4.3, if applicable.

WARNING: Omission of 3D diagonals may preventseeing the

full effect of loading.
5.6 Bidirectional Length Measurement
Capability
(d) Removeweight.
( e ) Repeat (6) above (repeatability test). 5.6.1 General. The preceding tests haveproduced
I f ) Repeat (c) above (ball barmeasurements).
a meaningfulpicture of themeasurementsystemper-
(8) Perform a repeatabilityanalysis:results of tests
formance; however, someerrors, such as undue machine
2 and 5 shall not exceed the stated repeatability specifi- or probe hysteresis and improper probe compensation,
cation. have not been fully analyzed since no two-sided length
(h) Performvolumetricanalysis: measurementhasyetbeenperformed. The following
( I ) range of readings oftest 3 shall not exceed testsremovethisdeficiencyby requiring the measure-
statedmachinevolumetricperformance specification; mentof a gage block of a convenient length, in four
(2) range of readings oftest 6 shall not exceed positions in themachineworkzone. Three ofthese
statedmachinevolumetricperformance specification; positionsareroughly aligned with the machine axes,
(3) the difference between a measuredlength in andthefourthpositionisuser-selectable.Itisrecom-
test 3 and the measured length from the same position mendedthatthisfourthpositionnotbealignedwith
in test 6 shall not exceed 50% of the machine volumetric anymachine axis. The length of theblockshall be
performancespecification. withintherangeof at least 25 mm (approx. 1 in.) to
100 mm (approx. 4 in.),withthe default valuebeing
5.5.7.2 Optional Procedure (Laser or Gage 25 mm (approx. 1 in.). The gage block shall becalibrated
Block).Followtheprocedure described above using in accordancewiththe requirements ofpara.7.3.1.
a laser interferometer, gage block, or other equivalent Before performing thesetests, the machine probe shall
device as themeasured artifact. Analyzeall data per be calibrated andqualified according to thesupplier’s .
para. 5.5.7. l. recommendations for normal operation of the machine
5.5.7.3 Rotary Table Machine Procedure. when measuring parts. Qualification on the gage block
For a rotary table machine, the procedure is as follows. tobeused for this testisspecifically excluded. The
(a) Calibrate therotarytableinanunloadedmode. measurements for thistest are also to be performed
(b) Place weight on the machine in compliance with usingtheprobingparameters,probeapproachrate,
the guidelines ofpara.5.5.7.1 above. probe approach distance, and settling time specified for
(c) Perform the repeatabilitytest as described in normaloperation inFig. IA.
para. 5.3, withthe exception oflocation.Locationis
optional in thistest.
5.6.2 Measurement Procedure - Bidirec-
tional Length Measurement. The gage block con-
(d) Perform the volumetric performance test for DCC
forming to the requirements of para. 5.6.1 above shall
machines with a rotary axis (para. 5.5.6) using positions
be rigidlymounted in thework zone ofthemachine
listed in column Al of Table 2.
on a fixture that allows probing access to the faces of
(e) Removeweight.
thegageblock for the four measurementpositions in
Repeat ( c ) above (repeatability test).
turn. The mean temperature of the gage block and the
(g) Repeat (d) above (volumetric performance test).
appropriate machine scale(s) may be measuredduring
(h) Analysis: results of tests 3, 4, 6, and 7 shall not
this gage block measurement process for each position,
exceed the stated machine performance specifications for
using a thermometer conforming to therequirements
repeatability,radial, tangential, andaxial (3D/alpha)
of Section 7. The exact location of the gage block in
error.
the workzone is not critical; however, it is recom-
NOTE: It is recommended thata weight with simple geometric form mendedthatthisposition be nearthelocation in the
be used for testing purposes to reduce potential difficulties in calculat- workzonewherepartswillmost commonly bemea-
ingthe CG location. sured. Aftermountingand alignment, which may be

46

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
OF COORDINATE MEASURING MACHINES ASME 889.4.1-1997

mechanical or performed by using the appropriate com- which includes the probe, the probe stylus, the machine
puter algorithm, the length ofthegageblockshallbe dynamics, and other variable parameters. The following
measured in eachpositionusingthemethodrecom- testshavebeendevisedto establish themagnitude of
mended by the supplier for measurement of distance the possible errors contributed by the probing sequences
betweentwoparallelplanes. for probes used in the point-to-point measuring mode.
For the purposesof this Standard, this includesswitching
5.6.3 Bidirectional Length Measurement Data probes, proportional probes, and nulling probes capable
Analysis. The length of thegageblockshall be of performing these measurements as they are used to
calculated using the coordinate measuring machine soft- acquire coordinate data onepointat a time(i.e.,not
ware in each ofthepositionsandmaybecorrected in a scanning mode). In all cases in thesetests, data
for temperature as described in para. 5.4.2.6. The worst are acquired by withdrawing the probe from the specified
case (largest) deviation, without regard to sign, between previously measured point and directing it to the new
the calibrated andthemeasuredvaluesofthelength positiontoacquirethenextpoint. The measurements
of the gage block,alongwiththenominallength of for thistestareto be performedusingtheprobing
the gage block,isreported as thebidirectionallength parameters,probeapproachrate,probeapproach dis-
measurement capability of the machine. In cases where tance, and settling time specified for normal operation
there appears to be a single (or several) outlying in Fig. 1A.
point(s) that does not conform to specification, it is
recommendedthatthismeasurement be repeated in 6.1.1 Method of Test - Point-to-Point Prob-
order to ascertainwhetherthelargedeviationactually ing. A precisionreferenceballconforming to the
reflects a systematic error. The procedure for checking requirements of para. 7.3.3 shall be rigidlymounted
repeatabilityisgiven in para. 5.5.2.3, which states a ontheworkpiecesupporting surface in theworkzone
gage block shall be measuredtwice in the suspected of the machine on a fixturethatallowsaccess by the
position. If themeasurementsagreewithintwicethe machineprobing system. The illustration (Fig. 37)
repeatability (para. 5.3), the first measurement shall be shows a calibration ballwiththe default diameter of
used and the second one discarded. If the measurements 6 mm (approx. 0.25 in.). Any position maybe chosen
do not agree withintwicetherepeatability,thenboth for thismounting,with the default positionbeingthe
are discarded and the procedure is repeated. This proce- TVE position as specified in Fig. l. Three probing tests
dure may be repeated three times; at the end of which shall be performed on this ball, using styli withdifferent
time, if repeatability has not been obtained as defined configurations. The three default styli are as follows:
above,thetestshall be discontinuedandthe fault a 10 mm (approx. 0.4 in.) longstraight stylus, a 50
determinedand corrected. mm (approx. 2 in.) long straight stylus, and a 50 mm
(approx. 2 in.) longstraightstylus with a 20 mm
5.6.4 BidirectionalLength Measurement Ca-
(approx. 0.8 in.) offsetperpendiculartothe ram axis.
pability Requirement. Bidirectional length measure-
The stylus tips can beofany diameter that allows the
ment capability, as calculated in para. 5.6.3, shall not
measurement to be made;however, a 6 mm (approx.
exceed the supplier’s specification, derated as specified
0.25 in.) diameter balltip[sphericity of 0.25 microns
in paras. 4.2 and 4.3, if applicable.
(approx. 10 kin.)or less] isthe default for each of
these three styli used to probe the test ball. Note that,
6 SUBSYSTEM PERFORMANCETESTS in order to allow measurement of the test sphere with
the offset stylus, the support holding the test ball must
The precedingsectionshaveprovided a reasonable
be rotated 90 deg. from the positionshown in Fig.
test of the coordinate measuring machine as a system.
37. Furthermore, sometypesofprobes maynotbe
Many errors have, however, either been hidden as part
able to perform this test with an offset stylus, thus the
of another measurement or not fully assessed. The
machine (probe) supplier shallbe consulted before
purpose of this section is to address the errors caused
performing this test.
by the mostimportant subsystems of themachine.
The userisallowedtospecifyanytestpattern
desired containing 49 points. The default test for direct
6.1 ProbingAnalysis - Point-to-Point computer-controlled machines is as follows. With each
Probing
of these styli, 49 points are probed onthetestballat
A major factor contributing to the total system mea- five different heights on thatball: 12 equallyspaced
suring error is the performance of the probing system, on a circle aroundthe test ballwhenthe stylus ball

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCE EVALUATION
ASME 889.4.1-1997 OF COORDINATE MEASURING MACHINES

Table

FIG. 37 SCHEMATIC DIAGRAM SHOWING THELOCATIONS OF PROBING IN THE

POINT-TO-POINT PROBING TEST
(Here the test ball is shown having a 6 mm (approx. 0.25 in.) diameter, the same as the stylus
ball. Also, the support for the test ball must be rotated 90 deg. when performing this test
with an offset stylus.)

48

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 STD-ASME B89*4=L-ENGL

METHODS FOR PERFORMANCE EVALUATION

OF COORDINATE MEASURING MACHINES ASME 589.4.1-1997

center is approximately 100 deg. from the pole ofthe 6.1.3 Probe Aproach Tests - Optional. Many
testball in a directionparallel to theshankattached machines/probe systems exhibit vastly different charac-
tothe stylus ball, 12 equally spaced on the equator teristics depending on the probe approach distance and
with the pattern rotated about the stylus shank 10 deg., the probe approach rate. For the machine user desiring
12 equally spaced around the ball with the stylus center to usemorethanonevalue of theseparameters,this
approximately 60 deg. from the pole and rotated about test of the machine performance is recommended. The
the stylus shank an additional IO deg. relativetothe procedureisthe same given in paras.6.1.1and6.1.2,
previouspattern, 12 equally spacedwiththe stylus except thatthistest is performed for twodifferent
center approximately 30 deg. fromthepole with the probeapproach distances andprobeapproachrates.
pattern again rotated an additional 10 deg., and finally, The workingtoleranceforpoint-to-pointprobing is
one on the pole of the test ball. This situation is depicted specified for each of these options.
in Fig. 37, in which the differentprobepositionsare
shownwithdashedlinesandlabeledpositions 1 to 5. 6.1.4 Point-to-Point Probing Performance Re-
The default test for manual machinesis the measurement quirements.Point-to-Pointprobingperformance, as
of 49 points distributed as uniformly as practical over calculated in paras. 6.1.2 and 6.1.3 (if applicable), shall
themeasurableportion of thetestball. not exceedthesupplier’sspecifications,derated as
On directcomputer-controlledmachines,theprobe specified in para. 4.3, if applicable.
shallbevector-driventowardthetestball center for
each touch, providedthisisnormal for themachine 6.2 Probing Analysis - Multiple-lip Probing
when measuring parts. On drivenmanualand free- In addition to the probing errors highlighted in para.
floatingmanualmachines,wherepossible, one axis 6.1.1, CMMs that use multiple stylus tip positions can
should be lockedandtheremaining axes movedto haveadditional errors. These errors canbe due to a
contact the ball in order to accurately hit the test ball. number of sources including the uncertainty in location
In all cases, the supplier’s probeapproachdistance, of each of the tips caused by tip calibration errors or
probe approach rate, and settling time, as given in Fig. by the errors associated withtheuseofan orienting
1A, shall be used. head or probe changer. This istrue for allmultiple-
tip system configurations,including:
-
6.1.2 Data Analysis Point-to-Point Probing. (a) systemsusingmultiplestyliconnectedtothe
From each set of 49 readings for each stylus, a sphere CMM probe, such as star clusters;
center is computedusingthe supplier’s recommended (b) systems using orienting heads;
algorithms. From this center a radius is then determined (c) systems usingprobe or stylus changers; and
for each measurement point. The minimumradius is (ci) systemsusingheadswithmultipleprobes.
subtractedfrom themaximumradiustoproducethe The common element of these systems is that different
point-to-pointprobingperformanceforeach ofthe tips or tip locations are usedtoinspect a workpiece
styluslengths. If the resultobtained for a particular without any recalibration of the tips. As a result, it is
stylus is less thantheworkingtoleranceforthetest, importanttounderstandanyadditional errors which
then the testing is discontinued for that stylus andthe mightbe contributed by these systems.
resultreported. If theresult for any stylus is greater
than
the
workingtolerance, then thetestshall be 6.2.1 Method of Test - Multiple-Tip Probing.
repeated. If the new results agree to within the working The calibration ball diameter and all system configura-
tolerance for repeatability (para. 5.3), thenthe second tion dimensions in this Section are default values. Other
set of data isdiscardedandthefirstset used for the dimensions maybe substituted and it is recommended
analysis. If they do not agree, then a thirdsetshall thatthis be done if there isanyconcernthatthe
be taken. If this agrees with either of the two previous configurations requiredtomeasureactualworkpieces
sets, thenthefirst of the agreeing sets shallbeused are substantially different fromthe default values.
in the analysis. If no agreement to within the working A precision reference ball conforming to the require-
tolerance for repeatabilityisobtained after threemea- ments of para. 7.3.3 shall be rigidlymounted on the
surement sequences for anygiven stylus, thetestis workpiece supporting surface in theworkzoneofthe
discontinued andthe fault determined and corrected. machine on a fixture that allows access by the machine
After correction, all of the tests described in this section, probing system. The 6 mm (approx. 0.25 in.) diameter
eventhose for stylus lengthsthatwerepreviously in test sphere used in the point-to-point probing test (Sec-
tolerance, shall be repeated. tion 6.1) may beused for this test. Any position may

49

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FOR PERFORMANCEEVALUATION
ASME 889.4.1-1997 OF COORDINATE MEASURING MACHINES

be chosen for thismountingwith the default position 6.2.3 Multiple-Tip ProbingPerformance Re-
beingthe TVE position as specified inFig. 1. quirements. Multiple-tip probing performance, as cal-
Five different probing tip positions shall be used to culated in para.6.2.2,shall not exceed the supplier's
performthis test. These positions can be created by specifications, derated as specified in para. 4.3, if appli-
using a stylus configuration with five tips, five different cable.
orientations ofan orienting head, or throughtheuse
of a probe or stylus changing system using fivedifferent
tippositions. Two oftheprobetippositionsshall be 7 TEST EQUIPMENT
on a line perpendiculartotheram axis. Two more
shall be on another such line displaced 90 deg. The 7.1 Temperature
fifthpositionshallbe on a line parallel to the ram
The time constant of thermometers shall be no more
axis through the intersection of the first two lines. The
than one-tenth the cycle time of the highest frequency
default stylus length, including any extension members,
component of thetemperaturevariationof interest in
measured from the intersection of the above lines, shall
a test. The time constant isthetimerequired for the
be 50 mm (approx. 2 in.) when using any of the above
thermometer to indicate 63.2% of its final change due
systems or combination of the above.
to a step change in temperature.
The user is allowedtospecify any test pattern that The resolution of thermometersneedbe no greater
contains 25 points. These 25 points shall be probed than one-tenth theamplitude ofthe lowest-amplitude
on the test ball as equally spaced as possible and cover component of temperature variation of interest in a test.
as muchof thesphere surface as practical. The 25 Thermometers shall be calibrated by suitable means
pointsshall betakenusingfive different tips or tip to an accuracy of ?O. 1°C over the temperature range
locations and each set offive points probed by each of use.
tipshall also beaswidespread as possible. As an
example, these five points could be four points around
the equator of thesphere (assuming thepoleposition 7.2 Vibration
is directly in line withthe stylus shaft supporting the
tip) plus a pointdirectly in line withthe stylus shaft. For thepurposes of this Standard, relative motion
shall be measuredusing a high-resolution, undamped
displacement indicator. Resolution of 0.1 p m (approx.
6.2.2 Data Analysis - Multiple-Tip Probing. 0.000004 in.) or better is recommended.
From the set of 25*readings,a sphere center is computed
using the supplier's recommended algorithm. From this
center a radius is then determined for each measurement 7.3 Displacement
point. The minimumradiusis subtracted fromthe
maximumradiustoproducethe multiple-tip probing 7.3.1 Gages. Step gages and gage blocksshall be
performance. If theresult obtained is less thanthe calibrated to within one-fifth the working tolerance for
working tolerance for the test, then the result is reported. therepeatabilityspecified for the CMM. Indicating
If the result is greater than the working tolerance, then gagesshallhave a resolution ofno morethanone-
the test shall be repeated. If the new result agrees with fifththeworkingtolerance for repeatability. All gages
the result of the first test within the working tolerance shall be calibrated following the supplier's recommenda-
for the repeatability (para. 5.3), then the second set of tions.
data is discarded andthefirst set isused for the
evaluation. If they do not agree, then a third set should 7.3.2 Laser Interferometer. A laser interferome-
be taken. If this agrees with either of the two previous ter conforming to therequirements of this Standard
sets, thenthefirstofthe agreeing sets shall beused shallhave a frequencystabilitysuchthatthislong-
in the evaluation. If no agreement to within the working term stability represents an error of less than one-fifth
tolerance for repeatability is obtained after three mea- the working tolerance for repeatability of the machine
surement sequences, thistestis discontinued andthe (in meters), dividedby the length of the longest machine
fault determined and corrected. After correction, the axis (inmeters). The resolution of such a system
repeatabilitytest(para.5.3.3)andallofthe tests shall be better than one-fifth the working tolerance for
described in thissectionshall be repeated. repeatability.

50

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 METHODS FORPERFORMANCE EVALUATION
OF COORDINATE MEASURING MACHINES ASME 689.4.1-1997

7.3.3 Precision Reference Ball(s). The precision 7.5 Humidity

reference ball(s) for the repeatability, TVE, and probing
Humiditymeasurement for correctionofthelaser
performance tests shall be spherical to within one-fifth
theworkingtolerance for repeatability of the CMM. interferometer wavelength shall be sufficiently accurate
that it contributes no morethanone-fifththe CMM
7.3.4BallBar.The ends oftheballbarshallbe workingtolerance for repeatabilitytolasermeasure-
spherical to withinone-fifththeworkingtolerancefor menterror.
repeatability of the CMM. Forinformationregarding
ball bars, see Appendix G .

7.4Pressure 7.6 Utility Air

The uncertaintyofthepressure sensor used for For the purposesofthisStandard,theutility air
correction of the laser interferometer shall be no greater pressureshall be measuredusing the gagessupplied
than T I mm Hg. withthemachine.

51

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 S T D - A S M E B87-9-3-ENGL3777 m 0757b70 0 5 8 3 0 8 73 2 3 m

APPENDIX A
USER'S GUIDE TO ASME B89.4.1
(This Appendix is not part of ASME 889.4.1-1997 and is included for information purposes only.)

A l PURPOSE (6) ballbar or gageblock length(s) for volumetric

performance;
This user's guide is intended to provide a framework
(c) ball bar or gage block placement for volumetric
for applyingthis Standard. Theguide is written in
performance;
checklist form to make it easier for first-time users to
( d ) probe offset for offset probeperformancetest;
beginusingthis Standard. Also, cross-references are
( e ) positionof the laser diagonalsand additional
provided from the guide text to the main body of this
linear displacements (large machines);
Standard for further detail on eachtestprocedure.
f'jj table mount or ram mount for the reference ball
(machinesused in theduplexmode);
A2 SUMMARY OF USAGE (g) radial separation and ball height (machines with
a rotary axis);
Theuse of this Standardmaybedivided into two
(h) bidirectionallengthmeasurementgageblock
distinct parts.First, this Standardisusedtoprovide
length;
a clear, commonmethod for specifying coordinate
(i) load,loading technique, location of load, and
measuring machines during negotiations between users
ball bar positions for testing performance under loaded
and suppliers. Second, this Standardprovidesuniform
conditions;
test proceduresto beused duringmachineacceptance
( j ) testballdiameterand stylus lengths and offsets
to establish conformancetothe specification.
for probing performance and multiple-tip probingper-
formance.
A2.1MachineSpecification
Stepgage tests are discussed in para. 5.4.2, and
Use para. 1.1 ofthisStandard to establish a clear laser interferometer tests are covered in para. 5.4.3.
understanding between supplier and user of the charac- Recommendationsonballbarlengthandplacement
teristics of the coordinate measuring machine. Detailed are provided in para. 5.5.2. The default lines for volu-
specifications of the machinecanbeitemizedusing metric tests onmachineswith large workzones are
thethree-part specification formprovided as Fig. 1. given in para. 5.5.4, the choice oftablemount or ram
Thefirst-timeuseris strongly urgedtorefertothe mountofthe reference ball for duplexperformance
technical glossary provided in para. 2.1 for clarification tests is covered in para. 5.5.5, the rotary axis parameters
oftheterminologyused on the specification form. in para. 5.5.6, guidelines for loading performance tests
In reaching agreement on the general machine speci- in para. 5.5.7, the bidirectional lengthmeasurement
fication, the principal mode of operation and the princi- capability defaults in para. 5.6, the probingtestball
palprobetypemustbe selected. If morethan one diametersand stylus lengths in para. 6.1, andthe
combination is specified, the supplier may chooseto multiple-tipprobingoffsetsandstyluslengths in
provide separate performance data for each combination. para. 6.2.
In specifying theperformance of a machine, several
choices' mustbemade: A2.2 MachineAcceptance
(a) accuracy test method (step gage or laser interfer-
For installation acceptance, the machinemustpass
ometer);
performance in an accepted environment. Acceptability
oftheenvironmentmaybedemonstrated in anyof
' Default values andor positional suggestions are offered forall four ways: by passingperformance tests, by showing
tests described inthisStandard. If theuser chooses to use the
default values and the position recommendations, only the accuracy compliancewith the supplier's environmentalparame-
testmethod needs tobe specitied. ters, by showinganacceptablelevel of performance

53

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 degradation due to environment, or by derating machine - Accept utility air, subject to later discovery
performance specifications to suit the environment. The of utility air inducedperformanceproblems
recommended procedure is to perform all of the environ- (para. 4.5).
mental tests before proceeding withthe performance Performrepeatability test, as appropriate for
tests. However, many installations maynot require the machine and probe configuration (para. 5.3.3).
full environmental testing to assure conformance to the If the vibration test failed, use these repeatabil-
performance specification. Therefore, as detailed below, ity datatoderatetheworkingtolerances
partsofthe environmental tests maybe deferred by (para. 4.3).
mutual agreement betweenthe supplier andtheuser. Perform linear displacement accuracy tests by
If the machine passes the subsequent performance tests, themethodpreviouslychosen(para. 5.4.2 or
performance of the deferred parts of the environmental 5.4.3).
tests isnot required.
Determinetherange of the deviations, either
arithmetically or graphically, and enter the
value on the acceptance testrecord (para.
A3ACCEPTANCE TESTING CHECKLIST 5.4.2.7 or 5.4.3.5).
Beforeproceeding to the following list, we assume Performballbartest, as appropriatefor ma-
thatthe supplier anduser are in generalagreement chine configuration, using the previously cho-
thatthemachineisproperlyinstalledandtheutilities senbar length(s) andpositions (para. 5.5.2).
are working satisfactorily. The numbers in parentheses Itis strongly recommendedthat a sketch or
following each itemarereferencestothemainbody description of numbered ball bar positions be
of this Standard. attached to the testresults.
- Perform or omit hysteresis test(para. 5.2 and Determinetherange of ballbarlengths for
Appendix F). each nominalbarlengthand enter onthe
- Determine if temperature environment meets acceptance testrecord(para.5.5.2).
supplier parameters; or, defer subject tolater Perform the offset probe performance test (para.
testing or later elimination. 5.5.3). Again, it is strongly recommended that
- Compute the uncertainty of nominal differential a sketch or description ofnumberedballbar
expansion (UNDE) (para.4.2.1). positions be attached tothetest results.
- Measurethetemperaturevariation error (TVE), Compute the range of the differences between
as appropriate for machine and probe configu- positive- offsetandnegativeoffsetmeasure-
ration (para. 4.2.2); or, defer TVE testing by ments at the same nominal locations, divide
setting TVE = O, subjectto later testing or this by twice the nominal length of the probe
later elimination. offset, and enter thevalueonthe acceptance
- Compute TE1 for all tests and enter teston testrecord (para. 5.5.3).
record (para. 4.2). IfTE1 exceeds 50% for For largemachines,supplementtheballbar
any specified working tolerance and the thermal testswith displacement measurements (para.
environment parameters are not met, compute 5.5.4).
deratedworkingtolerancesand enter onthe
For machines usedin the duplex mode, measure
acceptance testrecord.
duplex performance (para. 5.5.5).
- Measurerelativevibration as appropriate for
Performrotaryaxisperformancetests, ifre-
machine configuration, or accept vibration sub-
quired (para. 5.5.6), andrecordresults as in
ject to laterdiscovery of vibrationinduced
Table 2.
performance problems (para. 4.3). If the vibra-
tion effects are unacceptable due to excessive Ensure machinesupplier has proper documenta-
environmental sources and the userchooses not tion supporting loading effect specifications
to improve theenvironment, working tolerances (para. 5.5.7).
will be derated after performing the repeatabil- Performbidirectionallengthmeasurement ca-
itytests. pability tests (para. 5.6), and enter the results
- Accept electrical utility, subject to later discov- onthe acceptance testrecord.
ery of electrical utilityinduced performance Performprobetests(paras. 6.1 and6.2).
problems(para. 4.4).

54

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 APPENDIX B
THERMAL ENVIRONMENT TESTING

B I PURPOSE of the user-supplied environment. With these constraints

the following tests shouldbeperformed.'
The performance of coordinatemeasuringmachines
is strongly affected by the detailed characteristics of the
thermal environment that surrounds them. Parametersof B2.1Velocity
importance include thecoolingmedium (usually, but Since air isthemostwidelyusedcoolingmedium
notalways, air), thevelocityofthecoolingmedium, in dimensionalmetrology laboratories, thefollowing
the frequency and amplitude of temperature variations tests are structured for measuring air velocity. If some
ofthecoolingmedium, the meantemperature of that other medium is tobeused for heat transfer, then
medium, and the temperature gradients within that methods for testing its properties should be part of the
medium. The effects of these parametersand others machine specifications.
are discussed in detail in ASME B89.6.2. It is the For themeasurementofvelocity, several types of
thesis ofASME B89.6.2 and ofthisStandardthat instruments couldbe suitable. Theseinstrumentsand
currently it is notpossibleto specify parameters of a a discussion of the associated measurementproblems
thermalenvironmentthatwillensure a specificvalue are given in Tables BI and B2 ofASME B89.6.2. It
for thethermal error index (TEI). For a thorough isrecommendedthattheinstrumentsusedbeproperly
discussion of the technical situation, the reader is calibrated and the test personnel be'aware ofboththe
referred to ASME B89.6.2. The purpose of this Appen- limitations of their instrumentsand their operation.
dix is, however, to specify procedures and responsibili- Using appropriate instruments, the velocityofthe
ties for testingthethermalenvironment in theevent cooling medium around the machine shall be measured.
theTEI, as measuredinSection 4, exceedsthe 50% Measurements shall be made at the comers of a cubic
required for the machine, and the machine user contends volume that completely encompasses the machine, and
thathisenvironmentmeets the supplier's parameters. the velocity shall be computed as the average of these
For thepurposes of this Standard, these parameters eight measurements.
include cooling medium velocity, nominal mean temper-
ature, frequencyandamplituderange of temperature
B2.2 Mean Ambient Temperature
variation, and horizontal and vertical temperature gradi-
ents. The following tests are designed to measure these Themeanambienttemperature shall bemeasured
parameters for thepurposesof assuring conformance using a thermometer with characteristics as specified
tothe supplier's parameters. in Section 7 of this Standard. The mean ambient
temperatureshall be the timeaveragetemperature of
five readings takenat the center of the machinework
zoneover a periodoftimespanning the longest test
B2 METHOD OF TEST
(the use of five readings, ratherthantwo, for the
In ordertoensurethat the environmentitselfis measurement of the mean ambient temperature is justi-
tested rather thanany characteristic ofthecoordinate fiedherefor diagnostic purposes).
measuring machine supplied, these tests are to be
conducted with theCMM,supportcomputers (if sup-
plied), andany other auxiliary equipmentrelated to
' This appendix discusses only measurements of airtemperature,
andtheuser is warnedthat sometimes thermal effects are caused
the CMM, turned off for a periodof 24 hr preceding by coupling of infraredand visible radiation to the machine. If
thetesttoallowadequatesoak-outofCMM-induced the environment appears to conform to the supplier's parameters
afterperformingthe tests in these appendices, andthe TVEof
thermal gradients. Normal activity, however, should be the machine is still not within specification, then radiation coupling
continuedaroundthemachine as this constitutes part should be seriously examined.

55

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 B 2 3 Frequencyand Amplitude of locations ofthemachine.Thesetemperaturesshall be
Temperature Variation defined as theaveragevalueofno less thanfive
readings over an interval of 10 min. The vertical
The range of frequencies oftemperaturevariation
gradient shall be determined to be thedifference between
and the amplitude of those variations shall be determined
the maximum and the minimum temperatures anywhere
by measuring and continuously recording the tempera-
along a vertical line throughthemachinedivided by
ture at the center oftheworkzone over a periodof
the distance betweenthemeasurement point of these
timethat should, at a minimum, be representative of
extreme temperatures. The horizontal gradient is defined
a daily cycle (i.e., 24 hr). The maximum peak-to-valley
to be thedifference between the maximum and minimum
temperaturevariationshall be determinedfromthe
temperatures alonganyhorizontal line through the
recorded data. The data shall be analyzed to determine
machine divided by the distance between the measure-
therangeof temperature variation for a daily cycle
ment points of these extreme temperatures. These read-
andanhourly cycle, subject to theconditionthat
ings shall be taken over a period of time at least as
isolated disturbances thatare shorter in duration than
long.as the longest acceptance test (or 24 hr) andthe
theminimumperiod(maximum frequency) specified
greatest valueofthe gradient reported.
by the supplier shall be ignored. The dailyvariation
shall be defined as the maximum range of temperature
readings in 24 hr, subject to the condition on transients B3 ANALYSIS
mentionedabove. The amplitude of thesuperimposed
If any of the parameters measured in B2 exceed the
hourly cycle shall be defined as themaximumrange
supplier’s specifiedparameters, it isthe responsibility
of temperature variationin any one-hour interval, subject
of the user to correct the problem in order to conform
to the same condition.
withthosespecifiedparameters, or else to be willing
to accept the performance derating described in Section
B 2 4 Thermal Gradients
4 of this Standard. If the parameters so measured meet
Thermal gradients shall be determined by measuring the supplier’s specifiedparameters, it is the supplier’s
the temperature at the extreme comers of the machine responsibility to correct the performance of the measur-
in a horizontal plane and also at the highest and lowest ingmachine to meetthespecifiedworking tolerances.

56

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 APPENDIX C
VIBRATION ANALYSIS
(This Appendix is not part of ASME 889.4.1-1997 and is included for information purposes only.)

C l PURPOSE C = geophone calibration factor, V/mm/sec

(V/in./sec)
The purpose of these tests is to establish the vertical
D = peak-to-peakdisplacement, mm (in.)
andhorizontalvibrationenvironment at the interface
E = output transducer peak-to-peakvoltage,
betweentheCMMand the supportsystemprovided
V
by the user, in order to divide responsibility between
F = frequency, Hz
the userand supplier.
V = peak-to-peak velocity, mm/sec (inhec)

C2 METHOD OF TEST Vibrationbetweentheprobeandtheworkpiececan

result from forces external totheCMM, or from
The survey shallbe made using directionally sensitive, forces generatedwithintheCMM itself. Therefore, it
lowfrequency (0.5 to 100 Hz) transducers. Useof is generally advisable tofirsttest to see if the CMM
accelerometers having typical natural frequencies above isthesourceofvibration prior toconducting a full
25 Hzshouldbediscouraged, as these devices cannot vibrationmeasurementand analysis. IftheCMMis
measurethelowfrequency disturbances that are the motorized, the power to all motors should be removed
main cause of measuring machine accuracy degradation. and the CMMtestedtodetermineif the vibration is
The transducers shouldbecapableof discriminating still present. If theCMMuses air bearings, theair
motion amplitudes of 0.25 pm at 0.5 Hz (for example, pressure tothe bearings should bevaried sufficiently
Hall-Sears HS-10-1geophones). A band pass
filter to establish whetherairbearing instability is causing
of correspondingfrequencyrangeshouldbeusedto vibration.
discriminate betweenthe different signals sensed by If the CMM is not the source of the vibration, then
thetransducers. A signal generator. ofknown voltage a complete vibration analysis at the machine-to-support
shouldbeusedtodeterminethe attenuation ofthe interface shouldbeperformedandthefrequency spec-
bandpassfiltersatthefrequencybeing studied. This trumanalyzedtodetermine if frequencies andampli-
istoensurethat accurate data are beingrecorded on tudes are presentthatexceedtheCMM supplier’s
the strip chart recorder or the oscilloscope (or other defined limits. Measurementsshouldbemade in the
recordingmedium). vertical axis direction and the two mutually perpendicu-
Outputs ofvelocity or acceleration transducers must lar horizontal directions that approximately correspond
beconverted to displacement. The followingformulas tothehorizontal axis directions oftheCMM.
maybeused: Caremustbetakenthat the survey is run at a
suitable timeand for a longenoughperiodtoensure
D = W(6.28 X F) that conditions represent machine operating conditions.
Attention should also be paid to differentiating between
D = AJ(39.44 X F*) nominallysteady state conditions and transient condi-
tions.
V = (E/C)/A,

A = 6.28 X V X F C3 ANALYSIS
If any of the vibration parametersmeasured by
where the proceduredescribedaboveexceed the supplier’s
A = acceleration, mm/sec2 (in./sec2) specification,it is the responsibility of the user to correct
A, = bandpass filter attenuation factor the problem in order to conform to the specification or

57

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 S T D = A S M E B87.4.L-ENGL L977 m 0 7 5 7 b 7 0 0583072 777 m

else accept a performancederating as described in

Section 4. If, on the other hand, the vibration parameters
are within the supplier’s guidelines, then it is the sole
responsibility of the supplier to correct the performance
of the CMM in ordertomeet specification.

58

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 S T D - A S M E B A ? I - q = L - E N G L 3777 m 0 7 5 7 b 7 00 5 8 3 0 9 3 b25 m

APPENDIX D
ELECTRICAL POWER ANALYSIS
(This Appendix is not part of ASME 889.4.1-1997 and is included for information purposes only.)

D l PURPOSE D3 METHOD
The purpose of thisAppendix is to specifytest In order to ensure propermonitoring,thepower
procedures for analyzing the electrical power supplied supply to the machine should be monitored for a period
to a CMM and its supportequipment in the event that the thatincludesthenormal cycle ofCMM operation. In
electrical power is suspected to be causing inadequate the one-shift plant, this should include a complete shift.
machineperformance. Inthethree-shiftplant, complete 24 hrmonitoring is
required.Additionally, care should be takenthatthe
D2 TEST EQUIPMENT power line monitoring occurs over a representative
periodwhichincludesallnormal or even intermittent
Althoughtheparameters describing the electrical
electrical activity within the plant that could affect the
powersuppliedto a machinecan be measured by a
varietyofinstruments(voltmeters, oscilloscopes, etc.), machine. (As an obvious example, consider the case
it is the recommendation of this Standard that a power when arcweldingis done only a few days a week at
line disturbance analyzer be used for these tests because a locationthatusesthe same feeder as the CMM. In
of the excessive labor required when individual instru- this case, thepower line monitoring should include a
ments are used (an acceptable example ofsuchan typical arc weldingsequence.)
instrument would be the BMI 2400 series or the Dranetz Formakingthesemeasurements,an approved, cali-
Model626-PA-600X). These units are designedto bratedpower line monitor ofthetype discussed pre-
monitor a widerange of power line disturbances and viouslyshould be used.Appropriate thresholds (sag,
are capable of continuous, unattended operation. Typical surge, and impulse) should be set at the values corres-
measuredparameters include sags, surges, impulses, ponding to those levels set by the supplier in the CMM
andlinefrequency. specification. Monitoring shouldcontinue for a sufficient
Sags aresuddenvoltage drops that are detected by periodto ensure thatall of the effects mentioned are
analyzing each cycleand comparing its root-mean- included.
square level toa long timeconstant averaged steady state
voltage value. When the cycle-to-cycle level deviates by
morethanthepreselected threshold, a sag is detected.
Surges are sudden voltage increases that are normally
detected with the same techniques used to detect sags.
Again, a standardpower line monitorwillnoteand D4 ANALYSIS
recordboththevaluesofthe surge andthetimeat
Typicalpowerlinemonitorsprovide printouts of
which it occurred.
both the levels and times at which deviations from the
Impulses, inthe technicallanguage of power line
monitoring,refertoshort duration (approx. 1 to 1,OOO accepted thresholds occur. If themonitorissetwith
psec) spikes superimposed upontheac sine wave. the thresholdsdescribedabove,anysuch deviations
Typically, such impulses are measured as the amplitude recorded shall constitute nonconformance with the sup-
of the spike alonewithrespecttothevoltage level at plier’s specifications, and it shallbetheresponsibility
the time thespike occurred, Le., no subtraction or oftheuserto correct suchpowerline defects. If no
addition ismadeforthe sinusoidal component. deviations fromspecifications occur, then it is the
Frequencychanges in the line are also normally responsibility of themachine supplier to correct the
recorded by suchinstruments. Changes in frequency performance ofthemeasuringmachine in order that
areself-explanatory. machinespecifications are met.

59

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 APPENDIX E
UTILITY AIR
(This Appendix is not part of ASME 889.4.1-1997 and is included for information purposes only.)

E l PURPOSE inadequate, an air flow gageshall be mounted in the

supply line upstream of themachine air filter. Flow
The purpose of this Appendix is to specify procedures
shall be observed under the condition(s) which resulted
andresponsibilities with respectto utility air in the
in evidence of inadequate pressure.
event thatexcessivepressurefluctuation,inadequate
Measured flow rateshall be convertedto standard
supplypressureattherequired flow rate, excessive
(normal) flowrateusingthe relationship
machinethermal drift, or bearing contamination is
evidenced.
It should be noted that high frequency pressure spikes
in the air supply caused by external factors can cause
significant performance degradation, and thus attention
should be giventomeasuringthesepulses. where
VN = air flow in standard (normal), mN3/min
or IN/min (ft3/min)
E2TEST EQUIPMENT V = air ROW atlinepressure,m3/min or I/
min (ft3/min)
For thefollowingtests, an air pressure gage, air
P , = absolutelinepressure = gagepressure
flow gage, and a temperaturemeasuringsystemwill
be required. The air pressure gage should be calibrated
+ 0.981 MPa (14.7 psi)
PS = standard (normal) pressure,MPa (psi)
to cover the range between the minimum and maximum
T, = line temperature, "C
utility airpressurespecified by themachine supplier.
The resolution ofthe air pressure gage shall be20%
of thepermissible utility air pressurefluctuation as
specified bythe supplier. The air flow gage shall have E3.3 Machine Thermal Drift
anaccuracybetterthan 20% ofthemaximum air
A temperaturemeasuringsystempickup shall be
flow ratespecifiedbythe supplier. The temperature
mounted in the air line upstream ofthemachine air
measuringsystemshall be calibrated totheaccuracy
filter; or, if thisisimpractical, on a metallicpart of
specified in Section 7.
the supply line. If thepickupismounted ontheline,
the line and pickup must be insulatedfrom ambient
air to ensure thatthetemperature of the utility air is
E3 METHODS
beingmeasured. The temperatureshall be measured
underthe condition(s) whichresulted in evidence of
E3.1 Pressure Fluctuation excessive machinethermal drift.
An air pressure gage shall be mounted in the supply
lineupstream of themachine air filter. The pressure
shall be observed under the condition(s) which resulted E3.4 Contamination
in evidence of excessive pressurefluctuation.
Surfaces near air exhaust points shall be examined
for water, oil, or solid particulates. Machine filters shall
E3.2SupplyPressure and Flow
also be checked for excessive water, oil, and other
Withthesamegagesetup,thepressureshall be contaminants. If itissuspectedthatthefilters are
observedunderthecondition(s) whichresulted in inadequate, the machine supplier shall make appropriate
evidence of inadequatepressure. If thepressureis recommendations as to correct filteringparameters.

61

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 E4TEST ANALYSIS pressure (for example, by removing other devices using
air fromtheline).
E4.1 Pressure Fluctuation
E4.3’ Machine Thermal Drift
If fluctuation exceeds the supplier’s specification, it
is the responsibility of the user to correct the problem If thetemperature of the air supply line does not
(for example, by installing an accumulator). Of interest meet the supplier’s specification, it is the responsibility
hereare not just the short-termfluctuationsthat one of theuser to correct the temperature (for example,
would be able to observe using a normal pressure gage, by installing a heat exchanger) and retest the machine.
If the temperature meets the supplier’s specification, it
but also high-frequency pressure spikes in the air supply
is the responsibilityofthe supplier to correct the
which can cause performance degradation. If such spikes
performance of the measuring machine to meet specifi-
are suspected, further corrective action may be required.
cation.

E4.4 Contamination
E4.2SupplyPressure and Flow If contamination is present, it istheresponsibility
of the user to change the air filter cartridge, clean the
If the flow rate exceeds the supplier’s specification, machine air systemusingproceduresrecommended
it is theresponsibility of the supplier to reducethe by the supplier, and correct thesupply contamination
flow required by the CMM. If the flow rate meets the problem. Two methods of correction are available:
supplier’s specificationbutthe line pressure does not, reducesupply contamination or decrease theinterval
it is the responsibility of the user to increase the supply betweenfilter servicings.

62

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 APPENDIX F
HYSTERESIS TEST DESIGN RECOMMENDATIONS
(This Appendix is not part of ASME B89.4.1-1997 and is included for information purposes only.)

F1
PURPOSE F 2 1 Machine Hysteresis -
Machines Used in
the Free-Floating Mode
This Standard strongly recommendsthat a machine
hysteresis test be performed before starting performance For machines used in thefree-floatingmode, it is
testsandthateachtest setup be subjectedto a setup recommended that a ball probe be inserted in the probe
hysteresischeckbefore each test. This is to prevent holderand carefully tightened,thatthe rambe biased
wasted time and work. If these hysteresis tests are not toward the workpiece supporting surface, and that the
performed,any excessive hysteresisislikelyto be probe be inserted in a socket in either a rigid workpiece
revealed as a lack of repeatability in later testing. The or intheworkpiecesupporting surface. The machine
purpose of this Appendix is to provide general guidelines is then gently pushed and released in various directions.
for performingsuchtestswithoutunduly constraining Forceshould be abouttwicethatrequired to hold a
the user, especially since, due to the variety of machine passiveprobe in contact with a workpiece. Hysteresis
typesand setups, any single test maynotbe suitable. isrevealed by differences in machinereadout after
release.

F2.2 Machine Hysteresis -

Driven Manual
and Direct Computer-Controlled
F2 GENERAL Machines
A basiccaution for allmachinehysteresistestsis For drivenmanualand direct computer-controlled
to ensure thathysteresis ofthemachine,not of the machines, it isrecommendedthat an indicator,such
testsetup, is beingmeasured. All testhardwaremust as a flexure-type indicator or LVDT,bemountedon
berigidandtightly secured. A test setup ischecked theworkpiece supporting surface tomeasuremotion
for hysteresis by applying light forces, withthehand, of theprobeholder. The machineis thenpushedand
to thetest setup, checking for anychange in the released as in theprevious case, withthe force not
appropriatereadout. Sometimes a light tap,using a exceeding the amount specifiedby the supplier. Hystere-
pencil or light hammer, can alsobe effective for showing sis is revealed by differences between indicator readout
lackof suitability of a test setup. andmachinereadout after release.

63

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 APPENDIX G
BALL BAR TEST EQUIPMENT DESIGN RECOMMENDATIONS
(This Appendix is not part of ASME 689.4.1-1997 but is included for information purposes.)

G1 PURPOSE
Thisappendix contains informationregardingball
barsandballbarmounting platforms.

G2 BALLBAR DESIGNRECOMMENDATIONS
For ballbarsconforming to this Standard, thefixed
lengthbarmustbeadequatelyrigidand stable to
maintain a constant distance between the balls while
positioning theballbar in different orientations, and
to not deflect during probing. The bar is usually made
of tubingto increase itsnaturalfrequencyandreduce
its weight.
Thestandmust also be stable andrigid to holdthe
ball bar in its positions and to not deflect while probing
the balls. It must not obstruct access to the balls when
the bar is oriented in the various positions. At present,
the mostcommonlyusedfixture is the free-standing
ball bar system shown in Fig. GI. For low and moder-
ately accurate CMMs, most stands willbe adequate
for testing volumetricperformance.Formore accurate
CMMs (less than 10 pm over 1 m3)the stiffness of
the ball bar system must follow proper design guidelines.
Table G1 listsrecommended cross-sections for ball
barsand stands when used in the free-standing mode
FIG. G1 FREE-STANDING BALLBAR
to checkhigh-accuracy CMMs. These sizes were se-
SYSTEM
lected to maintain a l p m error in the ball bar system.
Other ball bars of equivalent geometrical precision and in one direction and reversing it to the opposite direction.
stiffness are equally suitable. Measure the lost motion of the system to return to the
null position. The applied force should betwicethe
probing force. Thetestshouldthenberepeated,but
G3 ERRORSOURCES
thistime by applyingthe force perpendiculartothe
There are various sources of error that can contribute ballbar axis. The hysteresis shouldbe less than 20%
to the total error of a ball bar system. Brief descriptions of the CMM repeatability.
of these errors are listedbelowandtypicalvalues are Deflections dueto gravity cause a foreshorting of
given in Table G2. The ball sphericity asdefinedin theballbardueto the sag of the cantilevered bar.
the Standardshouldbe less than 20% of the CMM Thissagismostpronouncedwhen the baris in the
repeatability. Balls 25 mm or less indiametercanbe horizontal position. The foreshortening is shown as
purchased to a sphericity of 0.2 pm or less which will ( A t ) in Fig. G2. This is a cosine error ofdeflection
meet this requirement. andusually a small error.
Hysteresis errors (see Glossary and Appendix F) can Probing forces may cause deflections in theballbar
be checked by applying a force parallel to the ball bar and stand. Lateral deflections have a direct influence

65

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 TABLE G I
BALL BAR CROSS-SECTIONS FOR VARIOUS LENGTH SYSTEMS FOR FREE-STANDING
BALL BAR SVSTEMS MADE OF STEEL
~~~ _ _ _ ~
~~~~ _ _ _ _ _ _ ~ ~ ~ ~ ~ ~ ~

Bar Bar Post Post

Ball Shortest Bar InsideOutside Post InsideOutside
Ball Bar CMM Axis, Length, Diameter, Diameter, Height, Diameter, Diameter,
Design mm (in.) mm (in.) mm (in.) mm (in.) mm (in.) mm (in.) mm (in.)

1 1,000 (40) 900 (36) 25 (1) 19 (0.75) 900 (36) 63 (2.5) 38 (1.5)
600 2 (24) 500 (20) 19 (0.75) 13 (0.50) 500 (20) 50 (2) 32 (1.25)
3 400 (16) 300 (12) 19 (0.75) 13 (0.50) 300 (12) 38 (1.5) 25 (1)

TABLE G2
ERRORS FOR A HIGH-ACCURACY FREE-STANDING BALL BAR SYSTEM
Ball Bar Ball System Vertical Horizontal Post
Design, Sphercity, Hysteresis, Distortion, Vibration,
Bending,
sag,
pm (pin.) pm (pin.) pm (pin.) +m (pin.) pm bin.) pm 1pin.t pm (pin.)

1 (5.5) 0.14 0.08 (3.0) 0.02 (0.8) 0.01(0.2) 0.52 (20.5) 0.35 (13.8)
(5.5) 0.14 2 0.05 (2.0) 0.01(0.5) 0.0 (0.0) 0.22 (8.8) 0.27 (8.8)
3 0.14 (5.5) 0.05 (2.0) 0.01(0.3) 0.00 (0.0) 0.16 (6.4) 0.20 (7.9)

GENERAL NOTE:
Values were calculated in microinches and rounded when converted to micrometers.

TABLE G3 "

UNCORRECTED THERMAL ERRORS(Pm) WHEN THE ENVIRONMENT AND BALL BAR ARE AT
DIFFERENT TEMPERATURES
Uncorrected Thermal Errors, hm

A T, "C Steel Invar

Lengthnime, min [Note (111 900 mm 500 mm 300 mm 900 mm 500 mm 300 mm

O 10 1O0 58 35 11 6.0 3.6

15 38 21 13 4.0 2.2 1.3
30 14 8.1 4.8 1.5 0.8 0.5
8 60 1.9 1.o D.6 0.2 o. 1 o. 1
0.02 90 0.26 0.14 0.09 0.03 0.01 0.01
003 120 0.03 0.02 0.01 0.003 0.002 0.001

GENERAL NOTES:
(a) Errors shown are for an initial temperature difference of 10°C when time = O min.
(b) Example: After 30 min the temperature of a 500 mm steel ball bar would be 1.4"C different from the'CMM andthe length
error would be 8.1 Fm. After 60 min the error is 1.0 Fm.

NOTE:
(1) A T is the temperature difference behnreen the environment and ball bar.

on determining the distance between ball centers because taken,due to its system inertia. If a nulling or scanning
deflectionscause the balls to appear smaller to the probe isused, theprobing forces may deflect theball
CMM andthus a longerball barlength (center-to- positionasignificantamountunlessarigidsystemis
centerdistance) is calculated.Usingswitchingprobes used. Errors arepresented in Table G2 for typical
withprobingspeeds of 5-10 mdsec does not permit probing forces. For high-accuracymeasurements,to
the
ball
bar system to deflect before
the
point
is avoidbending in the free-standingballbar,onecan

66

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 I I I

FIG. G2 STATICDEFLECTIONOFA BALL

BAR
usethemethodoffirstfindingtheballbar in space
and then only touching theends. This does not eliminate
FIG. G3 GAGE BLOCK STAND FOR SHORT
bending of the stand.
BLOCKS
Free-standing ball bar systems can be verysusceptible
(Less than 300 mm)
tovibrations.Boththepostandtheballbaract as
cantilever beamswithno restraints todampen the shown in Table G3. Temperaturecompensationcan
motion.Thevibrationcanbecaused by external floor greatly reducethesevalues.
vibrations or internally produced by the CMM.The Thermal errors are also caused by handling the ball
CMM will induce vibration when accelerating or decel- bar. Thiscanoccurduringassembly oftheballbar
erating duringservo drive moves.Manytimesthe toadd or remove an extension or by grabbinghold
CMM is resting on an isolation system to prevent floor of the barto reorient it. Usinggloves when handling
vibrations from being transferred to the CMM structures. thebarwillreducetheheat transfer fromoneshand
With these soft isolation systems, whentheCMM to the bar. Also, a plastic sleeve on the bar will reduce
moves it induces a rockingmotion ofthemachine. If the thermal growth when handling the bar. Experiments
theballbarsystemisnotrigidenough or securely have determined that a typical steel ball bar will return
fastened to thetable, a vibration is producedthatcan to its original lengthtowithin 1 Fm in 30 min after
causemeasuring errors. MovingtableCMMs are the minor handling. Invarballbarsdidnotexceed 1 Fm
worst for inducingthese vibrations. change in lengthduring a similar test.
Temperaturechangeshave a detrimental effect on A morethorough treatment of measurement errors
accurately measuring a ball bar. One event that causes in free-standing ballbarsispresented in a technical
temperature errors ismoving a ballbarsystemfrom paper, “Properties of Free Standing Ball Bar Systems,”
one environment to another at a different temperature. published in the Journal of Precision Engineering, Janu-
This is particularly truewhenbringingtheballbar ary 1993, Vol. 15, No. 1.
from the outside into a laboratory. Temperature changes
ofIO-20°C are common.Table G3 shows calculated
G4 BALLBAR MOUNTING PLATFORMS FOR
values of temperatureandlength error for ballbars
USE WITH SWITCHING, PROPORTIONAL,
with a thermaltime constant of 15 min.Thistime
AND NULLING PROBES
constant wouldbetypicalofsteelballbars in still air
sitting on a CMM.Thevaluesshowthechange in In thisStandard,severaloptions are provided for
ballbarlength for various temperaturechangesand the measurement of ball bars or length standards using
howlongoneshouldwaitbeforetheballbarisused switching, proportional, andnulling probes. Thefirst
for makingmeasurements. To maintain a stability of option, whichismostcommonlyused today, isthat
less than 1 km, oneshouldwait 30 to 90 minbefore the ball bar be supportedso that both balls are accessible
using a steelballbar. Steel and Invar ballbars are to probing. When using a setup of this type, the stand
considered. Invar greatly reduces the thermal errors, as mustprovide stable supporttotheballbar in defined

67

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 orientations in themachine’sworkzone. The stand
should be of a height slightly shorter than the CMM’s
verticaltravel. The standshouldhavetheabilityto
raise and lower the ball bar heightto a specific position.
A goodstandshouldalsoallowtheballbar to be
rotatedtotheproper angle. Stands of this typeare
shown in Figs. GI, G3, and G4. Notethatthestand
shown in Fig. G3 is holding a gage block,but,with
modification, it could be suitable for a ball bar. Stands
of thetypeshown in Figs. G3 and G4 can support
ballbarsupto one meter in length, butthestand
height should not exceed onemeter unless it is appropri-
atelybraced.

FIG. G4 COLUMN-TYPEBALLBAR STAND

68

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 APPENDIX H
STRAIGHTEDGE TESTS FOR RAM AXIS ROLL
(This Appendix is not part of ASME 689.4.1-1997 and is included for information purposes only.)

H l GENERAL the location of these points may be easily marked with

maskingtapeand a pencil.For the purpose of this
A straightedge maybeused for measuring ram axis
Appendix, “reasonably smooth” is defined as being flat
roll. In classical measuringmachinemetrology,this is
withintheworkingtoleranceforlineardisplacement
accomplishedusing an indicator in themachineram, accuracy over a distance of IO mm. Thepositions of
but it can also beaccomplishedusingthemachine’s these pointsaremeasuredusing an offsetprobe of
built-inprobing system. When the machine’sprobing reasonablelength, for example, 150 mm (approx. 6
system is used,other effects, such as machine dynamics, in.). Measurements are made at eachpointwiththe
probingperformance, etc., alter thetestresultssome- probe in the first position, shown in Fig. Hl, and then
what;however,the main parametermeasured is still in the second position. It is necessary to movethe
ram axis roll. The test in thisAppendixusesthe cross-slide in order to allow probe access to the straight-
machine’sprobe. edge surface, whichpossiblyintroduces a small error
due to roll in this axis as the machine is moved. Should
a machine appear to haveexcessive ram axisroll,the
H2 PROCEDURE roll caused by the cross-slide motion shouldbe measured
and subtracted fromtheresults in order to determine
A straightedge of reasonablequality is placed on the cause of this condition. The ram axis roll, in
thetablewiththe straight surface parallelto theram radians, is the ratio of the differences in the coordinate
axis. On a vertical ram machinethiswould beinan perpendicular to the face of the straightedge to twice
approximatelyvertical direction, asshown in Fig. H I . the probe offset length. It is possible for wide straight-
The setup is mostconvenient if theperpendicularto edges thatprobeinterferencecan occur in this test,
the straight surface is roughly aligned with the machine unless the probe is slightly angled with respect tothe
axis (X or Y). A number of points on the straightedge, lineparallel to the straightedge surface. This situation
encompassing therange of the ram axis, arechosen is depicted in Fig. H2, along with a diagram indicating
for sampling. If the straightedge is reasonably smooth, the correct offset to use when the probe must be angled.

69

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 .- Ram

Probe
(second position)

- Probe
/ I I (first position)

FIG. H I SETUP FOR STRAIGHTEDGEMEASUREMENTOF RAM AXIS ROLL

ON AVERTICAL RAM MEASURINGMACHINE

70

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 S T D = A S M E B 8 9 * 4 * L - E N G L 3777 m 0 7 5 7 b 7 0 0583303 3 T q m

Tilted to avoid interference

GENERAL NOTE:
In either case, the distance Bshould
be used to compute probeaxis roll.

FIG. H2 ILLUSTRATION OFTHE USE OF A PROBE ATASLIGHTANGLE IN ORDER TO AVOID

INTERFERENCE WHEN MEASURING RAM AXIS ROLL ON A MACHINE WITH A WIDE
STRAIGHTEDGE

71

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 APPENDIX I
INTERIM TESTING OF CMM SYSTEMS
(This Appendix is not part of ASME 689.4.1-1997 and is included for information purposes only.)

II INTRODUCTION length determination should be small compared to the

ThegoalofCMMinterim testing istoidentifyand threshold level at which the interim test fails. Similarly,
rapidlyremovefrom service defective CMMsbefore theformand surface finishofthe artifact shouldnot
significant numbersof good parts are rejected or bad significantly affect the measurement. (These conditions
parts are accepted. The frequent application of interim are similar to those stated in Section 7 of the Standard
testingwill increase confidence in CMMperformance which typically require the uncertainty ofan artifact
betweenCMM calibrations. Interim testing isnot a tobe ,20% oftheCMM stated performance.)
substitute or replacement for CMM calibration, and is Thermal properties of the artifact are also important
not normally diagnostic in nature. Rather, it checks the for workpiecesmeasured at a temperature other than
validity of the calibration by detecting common CMM 20°C. In .general, the user should select an artifact that
performance failures. It is recommendedthatusers has a thermalexpansion coefficient similar to that of
regularlyapply interim testing to their CMMs. An theworkpiecescommonlymeasuredwith the CMM.
effective interim test checkstheCMMmeasurement The uncertainty in the thermal expansion coefficient of
system as well as subsystem components that are used the artifact must also be considered, as discussed in
in the normal operation of the CMM. This may include para. 4.2 of the Standard. If the user commonly applies
suchcomponents as probes, probeheads,temperature
a correction for the thermal expansion of the workpiece,
compensation systems, and rotary tables. This document
then a thermalcompensationshould be applied tothe
assists CMM users by providing information on efficient
interimCMM testing. interim artifact. This will allow testing of the thermal
compensationsystem as a part of the interim test
procedure. Note that the temperature sensors are a part
12 GENERAL INTERIM TESTINGGUIDELINES of the thermal compensation system and are subject to
Limitedtimeis available for performinginterim damageand drift. Since environmental conditions may
testing, hencean efficient testmust concentrate on affect theperformance of a CMM, it is advisable
sources of performancedegradation that commonly torecordthetemperatureand other environmental
occur. The goal is to test for as many errors as possible parameters during an interim test, particularly if unusual
with a minimumnumberofmeasurements. If thetest conditions are present.
fails, additional actions are needed. These might involve It is important that the artifact be dimensionally
further diagnostic testing or involveCMM servicing stable between interim tests, so that the measurements
and calibration. CMMsubsystemcomponentsneedto obtainedduring an interim testcanbecomparedto
be included in the interim test to broaden its scope and thosefromprevious interim tests, and, if available,
insure that the entire measurement system is operating tothe artifact’s known length. Certain materials are
correctly. Each user has special needs, so interim testing
dimensionally unstable and may change in length by
proceduresand artifacts mayvaryfromusertouser;
many parts permillion(micrometerspermeter)over
however, the following guidelines may providesome
guidance in the matter. oneyear.It is importantthatthedimensional stability
CMM errors, whethersystematic or random,reveal (including any possible damage) of the artifact be
themselves as deviations fromknown lengths or as substantially less than the smallest CMM error of
variations of several measurements of a fixed (perhaps significance to the user. The interim artifact should be
unknown) length. Theuse of a knownlength artifact securely located on the CMMtabletopreventany
supplies additional usefulinformationfromthetest. If possiblerocking or slippage duringthemeasurement
a knownlength artifact isused,theuncertainty in its procedure. To compare interim test results to one

73

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 another, it is advisable to locate the artifact in approxi- features should be set comparable to those of the tightest
matelythe same positionand orientation for all tests. tolerances found in the actual production workpieces. In
Similarly, theinspection plan, such as thenumberof general, the interim artifact should be treated, fixtured,
probingpointstaken on the artifact, shouldbekept andmeasured in a manner similar to that of actual
constant foralltests.Widely distributing theprobing workpieces to reflect the actual measurement situation.
points over the gaging surface will aid in producing Althoughthe useof a testworkpiece as aninterim
consistent interimtesting results. artifact has merit, it is important to note that the testing
results are valid only for workpieces of a similar design
and may not indicate the errors present when measuring
13 INTERIM TESTINGSTRATEGIES
a workpiece significantly different from the test artifact.
There areseveral different strategies for choosing An artifact specificallydesigned for interimtesting
an interim artifact depending uponthe application of should be sensitive to common CMM errors. CMM
the user. For discussion purposes, we will consider two angular geometry errors typically increase in magnitude
categories: those strategies whichemployan artifact in direct proportion to thelengthofthe artifact. For
representing a typicalworkpiece (the artifact may be example, a squareness error of 10 arc secondscan
an actual workpiece from the production line) and those produce an error of 5 pm over a distance of 0.1 m,
strategies which employ an artifact specifically designed but it becomesan error of 50 pm over 1.0 m. This
for CMM testing. For all strategies, it is recommended illustrates a useful principle: to increase the sensitivity
that ten consecutive interim testing runs be conducted to angular errors, measurelong artifacts. Ideally,the
immediately after the CMM is calibrated. The mean artifact should be as long as practical, typically between
ofthese ten measurementscanbeused to establish a 75%-125%of the shortest axison a CMM with a
baselinevalue for theinterim artifact, andtherange nearly cubical workzone. On artifacts thatproduce
ofvalues indicates thetypicalvariationthat may be several lengths upon measurement, e.g., ball plates, the
expected under these conditions. Additional factors such longest length present will provide the greatest sensitiv-
as thermal conditions or different operators, may further ityto angular errors. (A short artifact positioned in
expand therange of interimtesting results. If upon several locations in the CMM work zone is not equiva-
recalibration of a CMM the new interim baseline mea- lentandwill not have the same sensitivity to angular
surements differ significantly fromthe previous baseline, errors as a long artifact.)
then the interimartifact or the CMMcalibration (or both) The orientation andpositionof the artifact is also
maybe suspect and further investigation is warranted. important.Certain artifact orientations canmaximize
Some CMMsarededicatedtomeasurementsof a the effect of geometry errors and hence allow them to
single type of workpiece or a family of similar work- be detected. Asan example, consider the squareness
pieces. In thissituationanactualworkpiece maybe error shown in Fig. I l . Inthefigure, L, isthetrue
used as the interim testing artifact. This type of artifact length of the ball bar as measured in a square coordinate
will be sensitive to errors that are important to actual systemandisthe (apparently foreshortened) length
workpiecemeasurements. An additionalbenefit is that as measured in an out-of-square coordinate system.
the user is familiar with the required workpiece measure- Note that L , isnot equal to h.It is apparent that the
mentsandconsequently may have a CMMprogram measured length of the artifact in the square coordinate
available that can be used for the interim testing. The system (X,,Y,)is longer than that of the out-of-square
selected workpieceandthemeasured features on that system (X2, Y2). If the artifact is a known length, then
workpiece should span the largest volume of the CMM this discrepancy appears as a measurement error. Even
work zone that is encountered during actual workpiece if the artifact length is unknown, this property can be
measurements, which will ensure that the relevant vol- exploited by measuringthesameartifact in two
ume of the CMM is tested. For users measuring many "crossed" orientations as shown inFig.12.By this
small workpieces located all over the CMM work zone, technique, theangular deviation from squareness (shown
it is suggested thatthesmallinterimtest artifact be as (Y in Fig. 12) can be determined in the absence of
measured at several different locations toinsurethat other errors. In Fig. 12. the XY squareness (in the
an adequate regionoftheworkzoneistested.Itis absence of other errors) can be estimated usingthe
notnecessarytomeasure every feature onthetest same artifact measured in twocrossed positions at
workpiece; rather, a representative group should be approximately 45 and 135 deg. In three dimensions,
selected (both for feature typeand location) for the the analogous situation is carried out by reorienting
interim testing procedure.The tolerance of these selected the artifact along all four of the body diagonals of the

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 2
+ Y2

FIG. I I AN EXAMPLEOFSQUARENESS
ERROR
FIG. 13 BALLBAR INDEXEDTHROUGH THE
BODY DIAGONALS OF THE CMM WORK
ZONE
(These positions are sensitive to the
squarenesserrors of all three axes.)

The CMMprobeshould be checkedto ensure it is

in goodworking order. This may involve an explicit
probe test that checks the directional sensitivity of the
probe, ¡.e.,probelobing (such as described in para.
6.1 of the Standard); or may be incorporated into part
of the general CMM geometry test, such as measuring
a long gageblockthatis oriented in several different
and X2
directions. Similarly, to test the probe calibration, which
involvestheaccuracy of theCMMprobecalibration
FIG. 12 ANGULARDEVIATIONFROM artifact (typically a sphere with a calibrated diameter),
SQUARENESS,REPRESENTEDAS OL a truebidirectionalmeasurement of knownlength is
((x may be computed from the length required. This might be the length of a gage block (as
measurements LA and Ln.) described in para. 5.6.2 of the Standard) or the diameter
workzone as seen in Fig. 13. This procedurecan be of a ring gage or of a precision sphere. It is important
conducted withmany different artifacts such as a ball tonotethatthemeasurement of a unidirectional step
bar, a step gage, or a longgage block. Alternatively, gage or the center-to-center distance between spheres
theuse of calibrated ball or holeplates may allow of a ballbar does notchecktheprobe calibration. If
morethanonesuchbody diagonal to be measured in multiplestyli are used (either with a stylus cluster,
each orientation. e.g., “starprobe,” or with an indexable probe head),
An artifact specificallydesigned for interimtesting then a testshould be includedthat checks theability
shouldprovide assurance thatthe entire measurement tolocate one stylus ballrelative to another. Such a
system is performing correctly. If only four body diago- test would include multiplestyliused in a single
nal positions are tested, the artifact should be of known measurement (e.g., see para. 6.2 of the Standard, “Prob-
length to test the accuracy ofthe scale on each CMM ingAnalysis - Multiple-Tip Probing”). If a probe-
axis. (If the artifact is of unknown length, then measure- changing rack is available, thenthissubsystemshould
ments in additional positions can identify relative errors betested by swapping probes in andout of therack.
betweenthe scales, butatleastoneknownlength is This not onlychecks the repeatability of probe changing,
requiredto establish thetrueaccuracy of the scales.) but defective probes in the rack may also be discovered.

75

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 For CMMs that include a rotary table, an appropriate Form deviations
test(such as the DCC rotary axis test,para. 5.5.5 of
the Standard) should be included as part of the interim
testing procedure. In summary, an effective interim test
I I I I
artifact should examine the complete CMMmeasure-
ment system to assure confidence in the entire measure-
mentprocess.
-
”

d
2
O - 1-
I I I
I Diameter deviations 1
4-
14 INTERIM TESTINGEXAMPLE I I I
5 2- I
I -
I
I I
I
I
The details of an interim test are highly user depen- ”
u -
dent since users have different types of CMMs, different dI
O:.
-2 -
m 1
I
I
I
. I
I
’

accuracy requirements, measuredifferent types of work- Q - I I I

pieces made of different materials, and their operators
-4 c I I I
I Length deviations I
and facilities aredifferent. Given theshorttimeto I I I
carry out the interim test, different users will optimize I I I
the interimtest in differentwaystosuit their needs. I I I
The following example is for a vertical ram CMM
with a nearly cubical (lxlx 1) workzonehavingan
indexable probe head, a probe-changing rack containing
two additional probes, and a temperature compensation
system. This specific example is givento provide
general guidance to CMMusers;however,theactual
CMMsystem may require different or further tests.
Additionally, some users may desire more extensive
testing or employ alternative strategies from the proce-
dures listedbelow. Probe X1 Probe X1 Probe Il Multiple
probes
Asan example of a specificinterimtest,theuser Straight
Offset
Multiple
dawn
ProbeX2 ( B a l l 1)
probe
Probe X3 (Ball 2)
chooses a ball bar calibrated for ball roundness, ball size, head
and center-to-centerdistance. The ball bar temperatureis positions

measured in each positionusingtheparttemperature

sensor and the appropriate thermal correction is applied FIG. 14 RESULTS OF AN INTERIM TEST
to all test results. A basictest involves measuring the USING ONE BALLBAR IN FOUR(BODY
four body diagonals of the CMM. In each position, DIAGONAL) POSITIONS
theuser decides to take eight pointson each ball of (The test includes checking temperature
theballbar. The user determines theapparentform compensation system, the indexable probe
error of theball,the difference betweenthebest fit head, and testing three different probes
sphere diameterandthe calibrated diameter and the available in a probe-changing rack.)
differences betweenthemeasuredballbar (center-to-
center) lengthsandthe calibrated value.These results
areplotted as shown in Fig. 14. For thesecondbody diagonal a similar measurement
In the first body diagonal position, the user employs is conducted, butwiththe probe headindexed so
a single probe(orientedalongtheram axis). In this that the probeis perpendicular to the ram axis. This
position, themeasuredform error oftheballs shows measurementwillproduce similar information to that
therepeatability oftheCMMandtheprobe,andany of thefirstbody diagonal position, butwill include
probelobing effects. The ball diameter measurements any Z axis roll error in theCMMgeometry. In the
check theprobe calibration (i.e., stylus ball size) and third body diagonal position, each ball of the ball bar
the short-range scale errors. The bar (center-to-center) ismeasuredwiththeprobeheadindexed in several
length measurementchecks for long-range (CMM geom- positions; thiswill supply information on probehead
etry or thermal expansion) errors inthatorientation. repeatability and the ability to accurately find a stylus

76

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 10
8
6
4
2
O

4
2
O
-2
-4

6- Upper threshold
4 - 0 0 0 0 ° 0 0 0 0 0 0 o o o o o o o o o o o o o O o ,
O O O
2- O 0 O O 0
0 - o " o o o _ 0 0 0
v
0 0 O 0 0 0 0 O 0
O' O O O
-2- O 0 O
O o o o o o O 0 0 0 o o o o o o o o o
-4-0 o o o o o' o 0 0 0 O
0 0 o o o o o o O
-6- 0 0 o o o o o o
- Lower threshold

FIG. 15 SUMMARY PLOTS OF SEVERAL INTERIM TESTRESULTS

5 12 -
6 10
.-E 8- Threshold
.-Tii 6 -
U
L 4-

balllocation relative to others with different probe Figure I5 shows one possible method of data analysis
head orientations. Thefinalbodydiagonalposition for the interim test. For each interim test, all four center-
checks for any defective probespresent in theprobe to-center length deviations, all eight ball diameters, and
rackandthe rack's probechanging ability. Thefirst the eight measured sphere form errors are plotted. The
balloftheballbar in this position is measuredusing testispassed if all these measurements are within the
thesecondprobeobtainedfrom the probe-changing threshold value limits. Someusersmayprefer a single
rack, andthesecondballof the ballbarismeasured plot representing thetestresults (instead of the three
withthefinal (#3) probefrom the proberack.The shown in Figs. I4 and 15). Such a plotcan easily be
form error and diameter, reported for eachballofthe constructed, as shown in Fig. 16, by combiningthe
ballbar, test each of the twoprobes for probe lobing largest length deviation, the largest diameter deviation,
effects and stylus size calibration, respectively. (If and one half the largest form deviation, in a root sum
additional probes are available, these could be checked ofsquares (RSS) manner.(One-halfthe largest form
by measuringeachball oftheballbar, in eachball deviation is used so each of the three contributions is
bar position, with a different probe.) appropriately weighted). This method has the advantage

77

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services
 of displaying only a single graphbutprovides less 15 TESTING FREQUENCY
information as to thesourcesof error. (If a CMM
The frequency of interimtesting is highlyuser-
problem does develop, plots such as those in Fig. I5
dependent. A CMM being operated three shifts a day
could be constructed using data from the previous test
with multiple operators in a harsh environment is likely
results.) to experience manymoreproblemsthanthe same
There are many different methods a user can choose machine being used one shift a day by a single operator
to establish testing thresholds. These include using the in an excellent environment. The frequency of testing
supplier’s stated CMM performancevalues for the is also strongly affected by balancing the cost of interim
particular CMM under consideration, whichmightin- testing against the consequences of accepting a bad
volve specifications fromtheASME B89.4.1 or other workpiece or rejecting a good one. It may be useful
appropriate national or international Standards. Other to consider the interim testing interval as a percentage
methods to determine the thresholds include examining of total CMM operating hours. Some userswithhigh
the tightest tolerance of a feature found on the user’s value and/or safety critical workpieces may elect to
workpiece andreducingthis byan appropriateratio. performdaily tests; other usersmighttestweekly
To avoid false alarms, thethreshold levels should or monthly. Additionally, interimtestingshould be
exceed allvariations arising from normal operations. conducted after any sort ofsignificant event such as
This may include such factors as different operators a CMM collision, replacement of a subsystem compo-
and different thermal conditions, e g , time of day nent, or the Occurrence of abnormal temperature varia-
or week. tions or gradients.

78

COPYRIGHT American Society of Mechanical Engineers

Licensed by Information Handling Services