Nbstechnologicpapert362 PDF
Nbstechnologicpapert362 PDF
BUREAU OF STANDARDS
                          George K. Burgess, Director
Bureau of Standards
                                   PRICE,   15    CENTS
                      $1.25     Per Volume on subscription
    Sold only by the Superintendent of Documents, U.     S.   Government Printing   Office
                                   Washington, D. C.
                                   UNITED STATES
                       GOVERNMENT PRINTING OFFICE
                              WASHINGTON
                                           1928
                                                                                   T   362
ABSTRACT
   This report describes so-called creep tests in which the elongation of metal
specimens is observed with time under a fixed load. Tests were made on a low-
carbon structural steel, a high-chromium steel, a chromium-molybdenum struc-
tural steel, high-speed steel, and a high chromium-high nickel austenitic steel and
correlated with short-time tension tests at corresponding temperatures within the
range 70 to 1,350° F. (20 and 730° C.).
   So-called creep charts are given in which the relations are shown between stress,
temperature, elongation, and time for each of the steels.       These charts enable
the approximate determination of the stress permitting life of different durations
with different total elongations. The application of these charts is discussed in
some   detail.
  Correlation of creep tests with short-time tension tests show that, when using
accurate equipment, the proportional limit was in the range of stresses which could
be sustained for long periods with small amounts of deformation.
  The best resistance to oxidation in air was shown by the high chromium-high
nickel steel and the high-chromium steel.    These appear to be superior to the
high-speed steel, which, however, with the high chromium-high nickel steel
showed the best load-carrying ability in the range 1,100 to 1,350° F. (595 to
730° C).
  The chromium-molybdenum           steel was not structurally stable at 1,200° F.
(650° C).        Oxidation was accompanied by decarburization and grain growth.
                                          CONTENTS
                                                                                       Page
  I.   Introduction                                                                    236
 II.   Previous investigations                                                         236
            1.   Creep   tests   compared to short-time tension   tests                236
            2.   Effect of temperature    on creep in steels                           237
III. Steels tested                                                                     238
 IV. Test methods and equipment employed                                               239
            Creep tests
            1.                                                                         239
            Tension tests
            2.                                                                         240
  V. Experimental results                                                              242
         1. Time-elongation curves                                                     242
         2. Creep charts                                                               242
            3.    Correlation of creep tests with short-time tension tests             250
             Comparison of data from different sources
            4.                                                                         256
          5. Comparisons of the different steels                                       260
 VI. Summary and conclusions                                                           263
VII. Selected bibliography on flow in metals                                           266
                                                                             235
236                  Technologic Papers of the Bureau of Standards                                       i
                                                                                                             vol.   22
                                       I.   INTRODUCTION
  The high temperatures and    pressures associated with recent devel-
opments                   equipment, oil-cracking processes, and the
                in power-plant
chemical industries have shown the need for additional information
on the behavior of metals at high temperatures. Attention has
recently been focused in particular upon so-called " creep tests" or
the flow in metals subjected to stresses for long periods at elevated
temperatures.
  Within the past five years many investigations have been made in
this field, but the information obtained has apparently not satisfied
the requirements of designing engineers who insist upon having a
more nearly quantitative evaluation of the relations between stress,
temperature, elongation, and time for different metals.
   This report is prepared to permit selection of working stresses for
five steels at temperatures between atmospheric and 1,350° F.
(732° C).     Preliminary results for three of these were previously
reported (8, 24).    Comparisons are given of the results of the creep
                            l
French, Cross,
Peterson              Creep in Five Steels at Different Temperatures                                                    237
   Creep       tests, unless carefully                made and                   the results reasonably inter-
preted,     may        create a false sense of security on the part of engineers,
for even with accurate determinations on a single bar, melt, or lot of
steel the question  remains what variations from observed numerical
values may ordinarily be expected if tests are repeated on other bars,
melts, or lots of the same type of steel.
   Where maximum and minimum values                                                  of tensile    strength              (or
elastic properties) are included in specifications for engineering ma-
terials for use at atmospheric temperatures, the range is seldom less
                                  2                                                                                        2
than 10,000            lbs./in.       ,   and a variation                 of only 3,000 to 5,000 lbs./in.
                                                                          22-"
                                                           ^ vo£
                                                           \>>
                                                                                                                    I
j£ o
                                              c                  =\fc__
                                                                                           i
                                                                                           Q               7
                                                                                           Id
                                                                  =14                                      /'
=lt
                                                                                           1               /
                                              r                   =10
                                                  <                                                    J
                                              I                                                        /
                                                                                          /        y           =t
               TEMPERATURE
                                                                                          L                    =i
   Fig.   1.        Diagram showing the character of the relations between stress, tem-
                     perature, time, and elongation in ordinary structural steel
  A, Creep chart.
  B, Time elongation curves where strain hardening is observed.
  C, Time elongation curves where strain hardening is not observed.
high temperatures, only half this total variation is observed, the per
cent variation may be enormous, as the " strength" or load-carrying
ability may only be a few hundred or thousand pounds per square
inch.
      Under such conditions an approximate determination would offer
 many     advantages, provided it could be secured quickly and could
likewise be depended upon.                            Therefore, attention will be given in the
 latter portion of this report to                      methods               for a quick        and approximate
 determination of working stresses.
 re         Cross
pe terfon           ']       Creep in Five Steels at Different Temperatures                                239
                                                               U.5.5TD.      THREAD
                                                                     14   PER INCH~~~~
4-R.
                                                 TEMPERATURE
                                              DEGREES FAHRENHEIT
                                        T   70    Z8Z 5Z5 755 885
                                        C   70    300 550 800 1000
                                        B   70    Z91 541 779 943
                                        F   70    Z87 516 151 921
                                 U.S. ST D. THREAD
                                    10 PER    INCH
       DEPTHS OF HOLES FOR THERMOCOUPLES
          MARK.  DEPTH
                a                  z
                    b              i                                U.S. STD.   THREAD
                                   4"                                  14 PER   INCH
                    c
       DIAMETER OF ALL HOLES =0.045'
Optical micrometer used in measuring creep is shown in the foreground at the right.
                                      2.   TENSION TESTS
  A   description has already been published of the equipment used
in the tension tests (24),                 which were made with a hydraulic                    test-
                                 —
French, Cross,
Peterson         Creep in Five Steels at Different Temperatures                         241
  The limited space between the movable head and base of the test-
ing machine made it necessary to use a relatively short furnace.  The
temperature variations observed under such conditions are shown in
Figure 2. Care was taken to reach thermal equilibrium before
         63552°— 28              2
242          Technologic Papers of the Bureau of Standards               [   vol. 22
applying the stress and measuring the strain, and with the practice
employed smooth       stress-strain curves were secured.        The need for
reaching thermal equilibrium is shown by the fact that steels, with
an expansion coefficient of about 0.000008 parts per unit of length per
degree Fahrenheit, will elongate four-millionths of 1 inch with a tem-
perature rise of one-half of 1° F. Temperature control within such
limits is very difficult to obtain in tension tests if temperature fluctu-
ations must be avoided for any appreciable time.        It is fortunate that
under ordinary conditions determination of the stress-strain relations
requires only from 5 to 15 minutes, and the electrical circuits can
frequently be balanced manually so that during such an interval
fluctuations in temperature of the control thermocouple may be kept
within 2 or 3° F. (about 1 or 2° C). Where variations in tempera-
ture take place slowly in one direction, there will be a progressive
change in the observed elastic modulus but no marked irregularities
in the stress-strain curves.     However, such changes may have an
important effect upon the observed proportional limits. Hence,
improvements in furnace construction and methods for close tem-
perature control are greatly to be desired.
   The need for an accurate extensometer for high-temperature tension
tests was discussed in a previous report (24).        With a proportional
limit at high temperatures of the order of 1,000 lbs. /in. 2 and an  ,
                   V.     EXPERIMENTAL RESULTS
                     1.   TIME-ELONGATION CURVES
  The time-elongation curves obtained for each of the five steels
when subjected to stresses at different temperatures are summarized
in Figures 5 to 9, inclusive. These are similar to the curves shown
in Figures  1 B and C and may be divided into two groups, depending
                             2.    CREEP CHARTS
  The time-elongation curves may be summarized in the manner
                   1 A to show the stresses which can be sustained
illustrated in Figure
for long periods at different temperatures with different degrees of
deformation.    Such terms        as "long life," "   freedom from appreciable
                 —
French, Cross,
Peterson               Creep in Five Steels at Different Temperatures                                                                                              243
                      S4O00 TrrJT
                     -54000 ('76) J
                                                                            Room     Td mk ercture
020 3
                             SSC JO-f'SS)
                                         7QL '-C74J -(jnbrai &n.
                                                                                                            S3400-t   M ^3^3                    .
43000 - ^£9/
                                         .   3QOOV-C4')- Jnbnjken
                                 100                 ZOO                 300      400      500               600          700             800       300
         02 °
    I                38* 100- *u)
                  L-3 %T0                        c
                                                       ^
                                                   J ?re
                                                 rbr
                                                           )
brof e
^ <
o.io s / t r
                                                                                                        9    bn )HS
           0.30                          «2
                                             ^
                             1
                                                                                                    |                 9    t   roki
         020
                                             /
                                                                                                <*4 ?0C£
                                                                                                                ^v
           0.10                      /
                                     /
                                 /                                                                                                     —     4C
                                                                                                                                                     u nbrol -en
                                                                                                                1+ —o—
                                 50                  100                 150
                                                                            ^=5   ^200
                                                                                    oj£
                                                                                          -»
                                                                                           250
                                                                                                •
                                                                                                        ro£a_
                                                                                                        _•
                                                                                                            300
                                                                                                                -•
                                                                                                                      350
                                                                                                                               t
                                                                                                                                      *
                                                                                                                                      400
                                                                                                                                           300- ('32, -unbr iken
                                                                                                                                                    450
                                                                                  Time         in   hours
    Fig.    5.       Time-elongation curves for hot-rolled 0.24 per cent carbon                                                                            steel
                               under stress at different temperatures
       The numerical values given above the curves refer to the applied load expressed in pounds
     per square inch of original cross section. Test specimen numbers are in brackets.
       See Table 1 for the chemical composition of this steel.
                            —                                                          .
deformation," etc., which are used in Figure 1 are not definite but
may    be justified by the fact that differences in mechanical properties
may    ordinarily be expected in different melts of the same type of
steel.   They may equal or exceed the differences in working stresses
which are based on various conceptions of "long life " that is, whether                                                                                 ;
0.20 0.40
        0.15                                                                                 0.30
               1
               §
        0.10                                                                                 120
               *
1 No 2^ 948/0 bn*i[
  ?     0X
                        fjo   4 3.BIZ 00
                                                                                             010
                                                                                                    f                       No.
                                Not   .617.
                                                             No-I   7.   SOl   OO                                                     No.i   2,*Z 100
  ^K           trtS s=*=                                                                       n         • •            •   •     1                         •   •«
                              60           120      180           240               300                     100                        ZOO          300                   400          SOO   ,       For
                            120            240     360            480               600 fa                  800                       1600         2400                  3200          4000\Nos.i2
                                                                                                                                                                                             I   and 51
  I
  I
  l<j   0.50
                                              800 T(4 251 j
                                                             \                               0.50       A
                                                                                                                                         /   too °F{6 50'C)
                                                                                                    1
                                                                                                        w
                                                                                                        M
        0.40                                                                                 0.40       i\\i
                                                                                                    s   $   s
                                                                                                                -
                   1                                                                                    jlK
                                                                                                                                                                     I
        0.30 r<                                                                              0.30
                   8
                                                                                                        IE
                                                                                                                                                            (
                                                                                                                                                             i
        0.20                                                                                 0.20       H-
                   *
                                                                                                        rr                                              J
        Q/0
               -
                                                                                             0.10
                                                                                                                                             \f
                                                                                                                                                /
                                                                                                                                      ,-*                                      i' 00
                                                                                                                                                   ^ W
                                                                                                                                                                          it
                        No. 3   .SOO OO
                                                  N0.3& ,30Ot O                                                                              f£m
                                                                  800           mo                       400                                        1200                 1600          2000,
                                                                                                         800                          1600         2400                  3200          4W0l»°>.*.»
          The numerical values given above the curves refer to the applied load expressed in pounds
        per square inch of original cross section.
          See Table 1 for details of composition and treatment of this steel.
French, Cross,']
JPeterson          J
                       Creep in Five Steels at Different Temperatures                                 245
       I                                         6000- (*2Z)-Unbroken
                               400           800           1200         1600         2000, ,
                               800          1600           2400        3200 4000              Wj
    I
    jg         0.30
           The numerical values given above the curves   refer to the applied load expressed in   pounds
     per square inch of original cross section. Test specimen numbers are given in brackets.
       See Table 1 for details of composition and treatment of this steel.
246                     Technologic Papers of the                                                 Bureau of Standards                                                   \   vol.   n
10, may be similar to stress-cycle graphs of fatigue tests at atmospheric
temperatures and show a more or less sharp bend or "knee," indicating
a limit of " static endurance," or they may appear to take the form
of a hyperbola.     Curves of the latter type do not always approach
the horizontal coordinate within the range of life included in the
tests and are found mostly in tests at the higher temperatures.
                                                     0.10                                         >-/«<tf-
                                                                                         fZ^-*-*inbro<cen
                                                                                                             jnbri    Ken
                                                            —*-                                              —• unbroren
                                                                                       100                              200
0.10 0.10
Bo.4L 0.40 I,
  %
                                                                                                                 i.
                                                                                                                                           S
                                                                                                                 %
              f
                              I                                                                  0.30
  B>0.30
  Q                           v       Jken
                                                      T>i roke
                                                                                                                                           J
                                                                                                                                           1
020 020 i
              5   8
                                                     j/
                                                       f
                                                                                                                             J   VJ
       0.10
                                          Sy                             jnbri>Aen
                                                                                                 0.10
                                                                                                                        S    /
                                                            WL^&Cf
                                                                     t
f •*
              hk&
                400
                                                ~3Sfe                                                    \   It *«M <W>*J                  *—<m< "JO) i      nbra ten
    The numerical values given above the curves refer to the applied load expressed in pounds per
  square inch of original cross section. Test specimen numbers are given in brackets.
    See Table 1 for details of composition and treatment of this steel.
Similar effects are  shown by the relation of stress and the time to
produce    per cent elongation, likewise summarized in Figure 10.
                  1
   However, the point has been made that allowable working stresses
in practice are frequently restricted by the permissible deformation
and that a more useful summary for many engineers would include
evaluation of life in hours and the per cent deformation during this
life.  Creep charts giving such information are probably much to be
desired, but little data are available upon which such summaries may
be based. The tests described in this report, which have been con-
French, Cross,
Peterson             Creep in Five Steels at Different Temperatures                247
 tinued generally for longer periods than similar tests made by earlier
 investigators  and which have likewise included determinations at
 more temperatures under a greater range of stresses, do not give
 sufficient data upon which to base such detailed comparisons without
   <=s    3      a
                             ljoui   jdcf ui 'uoyo6uo/j
  estimation of some of the values sought. The reasons for this can be ex-
  plained by considering what has been called initial and secondary flow.
    As already pointed out, creep is not continuous at temperatures at
  which strain hardening takes place, but                 ceases, for   most practical
  purposes, after a variable initial period. At such temperatures the
  determination of the working stress is a function of the permissible
  initial creep.
                                                          —        —
                '4wa
                   d   U">°H   — P B *<?S   H B !H
"8 s
                                                                                                       /    /°
                                                                                                       /   If
J6
                                  $1                      A                                    6j
                                                          sv            •V       /*
                                                                       $     /
                                                     tl/               t>/   /
*T $
                                                                   7
                                                                                                   I
                                                                   1
                                                               -I
                                                               —
                                                               /
                                                                                               r
                                                                                              /£
                                            I                  ;                          A>
                   „
                                                                                          k
                                                                                          V
                                       I                                              /
                                                                                      (
French, Cross,
Peterson             Creep in Five Steels at Different Temperatures                   251
                             Average flow rate in 2nd stage -Inch per Inch per hour
     Fig. 12.        Relation between applied load and secondary flow rate in different
                                  steels az different temperatures
  See text for discussion.
                     zk°t
          u-
 4.0
                                                            V
                      1%                                   -^
                                                                 \\\\
                                                                                      ELONG.               HR5
                                                                     \\                        IS 1000
                                                                     \v
                                                                         \\
                                                                         \
                                                                                  v
                                                                                  V
                                                                                  vV
                                                                                  T3
                                                                                               ^.10%
  Z0
                     0.1%
                                                                ^\
                                                                     s
                                                                          \
                                                                            vX /l%
                                                                         \\ -V
                                                                                           ^   ^4%
                                                                                               ,0.1%
                                                           "—
                              -^__
                              /
                                       "~~-~
                                               -   —   _
                                                                            IX
                                                                            S£
                         /
                                                           *
                                                               "-X- -•
                                                                         — -5^ \               \
                    PROF ORTI )NAL LIMI1                                       x\
                                                                               ^^
                      Z00                   AOO                  600                  800
                                                                                                   ^   ^^r
                                                                                                     1000                12100
           ^ x 5%                 INIT AL E .ONG.
                     \
  80
           °s       Z±%
                     \
  60       Qv
                    ^1%   v
  40
                                                                                                       1
t"i=r-
                                             TEMPERATURE. DEGREE5 F.
                Fig. 13a.                   Creep charts of three of the                             steels tested
       These give the relations between                             stress,   temperature, elongation, and time.
       See text for the               manner       of application of              data and other details. Proportional
  limits   were obtained in short-time tension                                    tests.
252
                                         —
French, Cross,
Peterson            Creep in Five Steels at Different Temperatures                                             253
            100
                   IN ITIAL    EL0N G.
                         5%                              \
            60          _£zA
                                                            \
                                                         \\ V \^
                                                    ~-o
                  °T"   ^1%
                        \^                                   \\ \             EL 0NG. N              HR'
             60
                                \                                \
                                                                     \s
                                                                                              100
—\-< \ )
                        _OI%                 \                                           A%
       z     40
                                                                \\
                                                                                  y      >l
                                                                                         ,0.1%
                                                                                              %
                                                     / /N N ^
       d
                                                                              X
       &.   zo
                         PR< )P0R1 I0NA      L   LIM T                    c       \\
                                                                              \          \V
       CO                                                                         \^          V^\\
       o
       o                                                                                            \NV
        )                 zoo            400              600        800               1000           IZ00
     33
1
Vi
     2+
<0
K 3'
£~
                                                                                         71 be
                                                                                         fee
"to o
S3
                                                                                   Cs 53
                                                                                               o
Pi 2
SI. 03 >
                                                                              f
                                                                              CO
                                                                              03
                                                                                         -g t»=3
                                                                              CO         ©-Q.2
                                                                             >
                                                                              £          IS'g
                                                                              O
                                                                             *^>
                                                                              Q          S' Hs
                                                                                         W ol2
                                                                              <»         m gfl
                                                                              5-
                                                                                         as
                                                                              1
                                                                                         !•?!
                                                                             00          gOQo
                                                                                           ."O
                                                                                         k^ ^      03
                                                                             o           ^£        03
                          'NI'&S   a3J'Cai O00I-SS3S19                       f=)          oS
   Fig. 19.           Differences between the proportional limits in tension tests and the
             stress giving    a selected elongation in 1,000 hours in the creep tests
      The    tests   made
                      in this investigation are summarized in the column at the left; those made by-
   Pomp and Dahmen       (33) are given in the columns at the center and right side of the chart.
      "Creep limit"        refers to 0.1 per cent elongation in 1,000
                                                              hours in the authors' tests and less than
   0.0001 per cent elongation per hour in the third to sixth hours in tests of Pomp and Dahmen.
   "Proportional limit" is the stress giving the smallest detectable deviation from the assumed
         modulus in the authors' tests and 0.02 per cent deformation in tests of Pomp and Dahmen.
   elastic
French, Cross,
Peterson         Creep in Five Steels at Different Temperatures                                                                259
                                                              LOW     C     5TEEL5
                       v\
                                   \
t\
                                                                                                      .1%
                                                                                                   \/0.l7.
                                                                                              \ X,O.OI%
                      »ai7.
                                                                                                    w
                                                                                                    ""0.1%   '^\
10%.
                                                                                                     17.   v^
                       -   —   -
                                        '
                                       o.e%c,
                                                •
                                                3
                                                    3.9%Cr, I3.6%W, 1.9%   V,   as receivea
TEMPERATURE, DEGREES F.
                                                                    l*Cr-Mo    steel   ,
                                                                                           quenched and tempered
                                                                   Z=rlioh   N'\   -High Cr    steel   ,   annealed
                                                                   3" High Cr steel, quenched and tempered
           50
                                                                   4=High speed        steel
      ^40
      of
                                      400          600
                                        TEMPERATURE -DEGREES f
      Fig. 21.           Comparison     of the five steels   on   the basis of the stress                    producing
                            O.l per cent or less elongation in 1,000 hours
800° F. (315 and 425° C.) marks the beginning of a rapid drop in
load-carrying ability, and at 1,000° F. (540° C.) none of the steels
                                                 2
tested can withstand a stress of 10,000 lbs. /in. with 0.1 per cent or
less elongation in 1,000 hours.
   At somewhat higher temperatures, between 1,100 and 1,350° F.
(595 and 730° C), the high-speed steel and the high chromium-high
nickel steel have better load-carrying ability than the three other
alloys.  The high-chromium steel appears to be about the same as
the low-carbon structural steel in its ability to withstand stress pro-
ducing 0.1 per cent elongation in 1,000 hours, while the chromium-
molybdenum steel gave the poorest results. This latter steel scaled
                           —
French, Cross,'
Peterson             Creep in Five Steels at Different Temperatures                                          261
very badly at 1,200° F. (650° C), as will be evident in Figure 22, and
showed measurable creep     at this temperature under very low loads,
as is illustrated in Figure 6. This poor load-carrying ability is prob-
ably due in part to poor resistance to oxidation in air. Coating the
chromium-molybdenum steel samples with electrodeposited gold or
chromium or even sprayed metallic coatings increased the life and
decreased the rate of creep. Some of these results are included in
Figure 6 (sample Nos. 20, 7, and 45 tested at 1,200° F. (650° C).
  The chromium-molybdenum steel was not stable structurally at
1,200° F. (650° C). Oxidation was accompanied by what appeared
to be decarburization and likewise by grain growth throughout the
entire mass, although the latter effect was most marked in the decar-
burized outer areas, as is shown in Figure 23.
 Fig. 23.       Structure of the   chromium molybdenum      steel before   and   after test at
                                     1,200° F. {650° C.)
  a, After 2,800 hours under 1,500 lbs./in.s X 12.
  b, Microstructure of steel before test. X 250.
  c, Microstructure of specimen "a" at center. X 250.
  d, Microstructure of specimen "a" near edge.     X 250.
  Samples etched with 2 per cent nitric acid in alcohol.
                     —
  re         Cross
pe terfo'n           ']    Creep in Five Steels at Different Temperatures                                            263
       Fig. 24.           Microstructure of the low-carbon structural                  steel before   and    after test
                                           at 1.100° F. (595° C.)
        a,   Before test. X100.
        b,   After 125 hours under 4,000 lbs./m.*           X100.
        Specimens etched       for   from   five to six   hours in cold 5 per cent picric acid in alcohol.
Fig. 25.         Structures of the high chromium-high nickel steel, the high                    chromium
                     steel and the high-speed steel before and after test
  Allspecimens X 100. Etchants for a and b were 15 per cent sulphuric acid in water; for c and d
were 15 per cent sulphuric acid plus a few drops of nitric acid in water; for e, 2 per cent nitric acid
in alcohol.
       264
                                                                     —
French, Cross,
Peterson         Creep in Five Steels at Different Temperatures    265
numerical values was not possible even with the extended tests
described.
   There has, however, been a demand for the type of chart reproduced
in the report  and these are the best that can be prepared at this time.
They should be useful if employed with an understanding of the man-
ner of their preparation. Careful comparisons with the work of other
investigators are included to show the magnitude of variations in
results from different sources for similar metals.
   In creep tests, as in short-time tension tests at different tempera-
tures, the results are dependent upon the equipment and procedure
employed in the tests and the manner of interpreting the results. A
deviation of 0.1 per cent from an assumed proportionality between
stress and strain in the tension tests was generally observed at a
stress which was higher than that producing a similar deformation
in  1,000 hours in the creep tests at corresponding temperatures.
In other words, for a given accuracy of strain measurement the
short-time test gave somewhat higher values than the long-time
tests, but the proportional limit determined by accurate equipment
                                                              —
such, for example, as was used in the described tests was in the
range of stresses which could be sustained for long periods with small
amounts of deformation. However, there appeared to be a tendency
for the proportional limit to become higher than the allowable creep
stresses as the temperature of the test was raised.
   The best resistance to oxidation in air was shown by the high
chromium-high nickel steel and the high-chromium steel.           They
appeared to be superior in this respect to the high-speed steel, which,
however, with the high chromium-high nickel steel showed the best
load-carrying ability in the range 1,100 to 1,350° F. (595 to 730° C).
  The chromium-molybdenum steel was not structurally stable at
1,200° F. (650° C); it showed decarburization and grain growth.
266          Technologic Papers of the Bureau of Standards                      [   vol.
 25     1926      Safe stresses at high temperatures,           The   Metallurgist, July 30, 1926,
                       p. 104.
 26     1926      H. Shoji and Y. Mashiyama,     On the plasticity of metals at high
                    temperatures, Science Reports, Tohoku Imp. Univ., 15, p. 442.
 27     1926      P. G. McVetty and N. L. Mochel, The tensile properties of stainless
                    iron and other alloys at elevated temperatures, Trans. Am.
                    Soc. Steel Treating, 11, No. 1, p. 73.
 28     1926      A. E. White and C. L. Clark, Properties of boiler tubing at ele-
                    vated temperatures determined by expansion tests, Preprint for
                    Annual meeting Am. Soc. of Mech. Eng., New York, December,
                       1926.
 29     1926      S. H. Inberg and P. D. Sale, Compressive strength and deformation
                    of structural steel and cast-iron shapes at temperatures up to
                    950° C. (1,742° F.), Proc, Am. Soc. Test. Matls., 26, pt. 2,
                    p. 33.
 30     1927      L. W. Spring, H. W. Maack, and I. Kanter, Testing flow in metals
                    at various temperatures, Power, 65, p. 205; also, Valve World
                     (Crane Co., Chicago, 111.), June, 1927, p. 191.
 31     1927      V. T. Malcolm and J. Juppenlatz, Investigation of bolt steels,
                    Trans., Am. Soc. Steel Treating, 11, No. 2, p. 177.
 32     1927      A. Michel and M. Matte, Variations des proprietes mechaniques
                    des aciers et alliages avec la temperature, Rev. Met. Mem., 24
                    year, No. 4, April, 1927, p. 200.
 33     1927      A. Pomp and A. Dahmen, Entwicklung eines abgekurtzten Prufver-
                    fahrens zur Ermittlung der_ Dauerstandfestigkeit von Stahl bei
                    erhohten Temperaturen. Mitt. a. d. Kaiser- Wilhelm-Inst. f.
                    Eisenforschung zu Dusseldorf, IX, No. 3; also, Stahl u. Eisen,
                    Mar. 10, p. 414.
 34     1927      D. Hanson, Some observations on creep of metals, The Metal-
                    lurgist, Apr. 29, 1927, p. 54.
 35     1927      Creep stresses, Engineering, May 6, 1927, p. 551.
 36     1927      F. Schleicher, Tension conditions at the flow limits, Zeit. Ange-
                    wandte Mathematik U. Mechanik, 6, No. 3, p. 199; Trans.
                    A. S. S. T., January, 1927, p. 140.
 37     1927      H. J. Tapsell and W. J. Clenshaw, Properties of metals at high
                    temperatures: I. Mechanical properties of Armco iron, 0.17 per
                    cent carbon steel and 0.24 per cent carbon steel, with special
                    reference to creep.     Department of scientific and industrial
                    research, Engineering Research, special report No. 1, published
                    under authority of His Majesty's Stationery Office.