Astm ds58 1975

The rupture and secondary creep rate data have been evaluated by both direct isothermal inter- polation or extrapolation, and by time-temperature parameter, to establish the temperature

EVALUATION OF THE ELEVATED

TEMPERATURE TENSILE AND

1916 Race Street, Philadelphia, Pa 19103

© by American Society for Testing and Materials

Library of Congress Catalog Card Number: 75-18417

NOTE The Society is not responsible, as a body, for the statements and opinions advanced in this publication

Foreword

This book represents a part of a continuing effort by The Metal Properties Council on behalf of the engineering community Individuals and organizations generating additional information concerning the materials evaluated in this report, or in others in this series, are urged to make these data available to the Council for incorporation in future revisions Address the Council at: The United Engineering Center, 345 East 47th St., New York, N.Y 10017

Related

ASTM Publications

Supplemental Report on the Elevated-Temperature Properties of Chromium-

Molybdenum Steels, DS 6-S2 (1971), $7.00, 05-006002-40 Evaluation of the Elevated Temperature Tensile and Creep-Rupture Properties

of C-Mo, Mn-Mo, and Mn-Mo-Ni Steels, DS 47 (1971), $6.25 05-04700-02

A Note of Appreciation

to Reviewers

This publication is made possible by the authors and, also, the unheralded efforts of the reviewers This body of technical experts whose dedication, sacrifice of time and effort, and collective wisdom in reviewing the papers must be acknowledged The quality level of ASTM publications is a direct function of their respected opinions On behalf of ASTM we acknowledge with appreciation their contribution

ASTM Committee on Publications

Editorial Staff

Jane B Wheeler, Managing Editor Helen M Hoersch, Associate Editor Charlotte E.Wilson, Senior Assistant Editor Ellen J McGlinchey, Assistant Editor

Figures 1 thru 6 Introduction Yield Strength, Tensile Strength, and Ductility Part 1 3Cr-lMo

Part 2 5Cr-^Mo Part 3 5Cr-HMo-Si Part 4 5Cr-V£Mo-Ti Part 5 7Cr-1/4Mo Part 6 9Cr-lMo Creep and Rupture Properties Part 1 3Cr-lMo

Part 3 5Cr-&Mo-Si Part 4 5Cr-'^Mo-Ti

Part 5 ICi-VMo

Part 6 9Cr-lMo

REFERENCE: Smith, G V., Evaluation of the

Elevated Temperature Tensile and Creep Rupture

Properties of 3-9 percent Chromium-Molybdenum

Steels; ASTM Data Series, American Society for

Testing and Materials, 1975

ABSTRACT: The evaluations of this report cover 6

grades of chromium-molybdenum steel of interest

for applications in boilers and pressure vessels:

ously published data and hitherto unpublished data

gathered by The Metal Properties Council from con-

tributing laboratories The properties that have

been evaluated include yield strength, tensile

strength, creep strength and rupture strength

In evaluating the yield and tensile strength

data, a normalizing procedure has been employed

that involves ratioing the elevated temperature

strength of a particular lot to the room temper-

ature strength of that same lot The method of

least squares is then employed to define a trend

curve for the ratio values representing a par-

ticular material grade

The rupture and secondary creep rate data have been evaluated by both direct isothermal inter- polation or extrapolation, and by time-temperature parameter, to establish the temperature dependence

of the average and minimum stresses to cause a secondary creep rate of 0.1 and 0.01 percent per

1000 hours, and of the average and minimum stresses

to produce rupture in 1000, 10,000 and 100,000 hours

Elongation and reduction of area data at frac- ture are included for both the short time elevated temperature tensile tests and for the rupture tests Summary figures, Figs 1-6, immediately fol- lowing this abstract show the temperature depen- dence of strength properties for the 6 grades of steel evaluated in this report In these illus- trations, the yield and tensile strength trend curves have been adjusted so that they corres- pond at room temperature to the specified minimum values of common ASTM product specifications The creep and rupture strengths represent the average values for a secondary creep rate of 0.01 percent per 1000 hours and rupture in 100,000 hours, respectively

Tabular comparisons of the yield strength ratio and tensile strength ratio trend curves for the six grades of steel are provided in Table V, and graphical comparisons are offered in Figs 57-

58

Tabular comparisons of the creep and rupture strengths are provided in Tables VIII through XI, with graphical comparisons of average 100,000- hour rupture strengths and average 0.01 percent per 1000 hour creep strengths in Figs 59 and 60 KEY WORDS: elevated temperature, mechanical pro- perties, tensile strength, yield strength, creep strength, rupture strength, elongation, reduction

of area, chromium-molybdenum steels, time-temper- ature parameters, data evaluation

DS58-EB/Oct 1975

Fig 1 Effect of temperature on yield strength, tensile strength, rupture strength (100,000 hours), and creep strength (0.01J* per

1000 hours) of 3 Cr - 1 Mo steel Yield strength and tensile

Pig 2 Effect of temperature on yield strength, tensile strength, rupture strength (100,000 hours), and creep strength (0.01# per

M

: -tf •HUJL1UUIIU1IJ-IL nnii i.4r

Pig 3 Effect of temperature on yield strength, tensile strength,

k Effect of temperature on yield strength, tensile strength,

:.': :::: :r: r ~~ :rt .: t-:-4;r;.; ^;.:r|r;irt:-rrt ;-.i :l^Uln4:]:::|;;-r:[4ri +^fV _J^:—;[•:: |_ W 1-t- \ -.i\::: :

::: :::: :::: ;; -~- ^^ 77^H: -itjrrfrtri.:- ^iy^HifiH ^ Hr ' t ixt+ iP+jSttt- +>Hr-H 3: Hitiritt^t -^P MtT'n-T- T i.;r 1 TM^'MT

fe"f -H-fj+H+l- TTmlTI

~r :: ": rH: H i'-rrffif:: ::.:: ::f.7.r;::: r-4-lu : 'Ai'.:: t' >3 ixHtt whw ±m&mm-# r:: : : 1-

+P*:-:~JEJ;:44fpffi:=:4^ \^\\\{ ! flj^i-ij-— -| i "S "V" S:::::::Etlc:!«i,:;::::: :±::::::::: Ji i^::

INTRODUCTION

The evaluations presented herein represent another

in a series of reports prepared under the sponsor-

ship of The Metal Properties Council (MPC) in the

interests of providing engineering design infor-

mation Included in the present evaluations are

6 grades of chromium-molybdenum steel ranging from

3 percent to 9 percent chromium, and Jj to 1 per-

cent molybdenum All are recognized for boiler

and pressure vessel usage

The tabular data, charts and the results of

the evaluations for the six grades have been

grouped separately as follows:

previously published in ASTM's DS Data Series Pub-

gathered by MPC from cooperating industrial organ-

izations Data representing different product

forms, plate, bar, pipe, tube, forging and casting

are included; however, for some grades, the number

of data representing certain product forms was

limited or completely lacking Data for weld metal

have been included, but are very limited in number;

data for weldments have not been included, owing

to the dependence of their behavior upon unstand-

ardized specific test details, e.g., the relative

fractions of base metal, weld metal and heat-

affected- zone encompassed

All of the data, including those from refer-

ences 1 and 2, are identified in Tables 1 and 2 as

to specification number, deoxidation practice,

heat treatment, product form and size, grain size

and chemical composition, in so far as these are

known

The properties that have been evaluated in-

clude yield strength, tensile strength, creep

strength and rupture strength Creep strength has

been evaluated at two levels, as the stress caus-

ing a secondary creep rate of 0.1% or 0.01% per

1000 hours; rupture strength has been evaluated at

1000, 10,000 and 100,000 hours Since the indi-

vidual strength properties employed in setting al-

lowable design stress intensities are each're-

quired in the temperature range for which they may

govern, the evaluations have been directed towards

developing trend curves that define the variation

of strength with temperature

Elongation and reduction of area at fracture

in the tensile and rupture tests are included in

the report, where available, and plotted in sum-

mary figures to reveal trends of behavior

average of duplicate tests The yield strength values represent either 0.2% offset, or the lower yield point, which is considered its equivalent Elongation at fracture was measured over a 2-inch gage length, unless otherwise noted Plate samples were taken at the quarter thickness position parallel to the rolling direction, unless other- wise indicated

The yield and tensile strength evaluations were made employing a normalizing procedure which has proved useful in prior evaluations (e.g., ref- erences 13-14) This procedure involves ratioing the elevated temperature yield and tensile strengths of individual heats of a particular grade

of material to the room temperature yield and tensile strengths of the same lots Then, by the method of least squares, the best fit curve is established for each set of such ratios to provide trend curves in ratio form, defining the variation

of strength with temperature These character- istic strength ratio trend curves may then be em- ployed to compute strength-temperature trend curves for specific room temperature strength levels of interest within the limits represented

by the original data Of frequent interest are curves anchored to the minimum strength specified

in the purchase specification (sometimes identi- fied as minimum position curves) Experience has indicated that such a curve may be expected to de- fine, approximately, a lower boundary for 95 per- cent of the data when the room temperature data population spans uniformly the permitted (or ex- pected) range of strengths

The yield strength, tensile strength and ductility data for the different grades are plotted

as dependent upon temperature in Figs 7-12

Part 1: 3 Cr - 1 Mo steels; Figs 7a, b, c The yield strength and strength ratio results are plotted in Fig 7a, the tensile strength and strength ratio results in Fig 7b, and the elonga- tion and reduction of area results in Fig 7c The trend curves for yield and tensile strength ratios, developed by the least squares procedure, are shown in the figures and tabulated in Table V The quantity of data for this steel at temperatures between room temperature and 800 F is minimal, and consequently there exists an uncertainty in the trend curves for this range Furthermore, some of the yield strength data represent 0.1 per- cent offset; the ratioed values for these data were nevertheless included in the least squares evaluation, on the assumption that the ratios for 0.1% offset might be expected to approximate those for 0.2% offset

Most of the data represented plate, in either the normalized and tempered or quenched and temp- ered conditions There were too few data repre- senting other product forms and heat treatments to warrant endeavoring to distinguish possible effects

of these variables, and therefore all of the data (except those for weld metal) and encompassing

and b, and tabulated in Table V, do approximate an

steel

Although excluded from the least squares

analysis (on general grounds), the limited yield

and tensile strength data for weld metal fall with-

in the scatter bands for the several wrought pro-

duct forms

The elongation and reduction of area data are

plotted in Fig 7c Owing to the paucity of data

between room temperature and 800 F, the trend of

ductility is poorly defined; above 800 F, ductility

increases with increasing temperature In asses-

sing the ductility data, note should be taken that

the strength at room temperature spans from 70 to

more than 130 ksi The one set of weld-metal re-

duction of area values falling below the general

scatter band, had a tensile strength at room tem-

perature of 127 ksi Also, a gage length of 1

inch had been employed for the plate material

Part 2: 5 Cr -^Mo steels; Figs 8a, b, c

The yield strength and strength ratio results are

shown in Fig 8a, the tensile strength ratio re-

sults in Fig 8b, and the ductility data in Fig

8c Trend curves are superimposed upon the

strength ratio data in Figs 8a and 8b, and tab-

ulated in Table V Weld metal data were again ex-

cluded from the least squares analysis, although

the ratio values fall not unreasonably in relation

to the scatter bands for the various product forms;

the latter were treated as though belonging to a

single population Although a wide range of

strengths is represented in the room temperature

data, inspection of the ratio scatter bands does

not suggest any significant effect of strength

level upon the trend curves

Fig 8c suggests some slight fall off of

elongation but not of reduction of area at inter-

mediate temperatures, before the usually observed

increase at higher temperatures

Part 3: 5 Cr - ^ Mo-Si steels; Figs 9a, b, c

Tensile test results were avilable for 4 heats of

bar and 1 of cast material; inspection of the

strength ratio scatter bands suggested that the

data should be treated as from a common population,

and this was assumed It is, of course, possible

that further data would prove the assumption to be

unwarranted It is of interest to note that the

steel is substantially identical with that for

falls somewhat more rapidly with increasing temp-

erature The yield strength ratio curves cross

one another at 550 F, and differ by less than 10

percent at temperatures up to 1000 F However,

with so few data for the 5 Cr-% Mo-Si steel, it

seems doubtful that the differences are signi-

ficant

In so far as can be judged, the limited

ductility data exhibit the same general trend with

increasing temperature as did the 3 Cr - 1 Mo and

the temperature range for which data are lacking have been sketched in, guided by the results of the least squares analyses for the higher temper- ature range Consequently, the trend curves can only be viewed as very rough approximations be- tween room temperature and 700 F Even at higher temperatures, the trend curves must be viewed as approximate, since they are based upon few test results Comparison with 5 Cr - ^ Mo steel, Table

V, reveals that although the derived trend curves differ importantly, the strength ratio scatter bands overlap sufficiently that it is possible that the data for the 5 Cr - !j Mo-Ti steel could

be viewed as belonging to the same population as that for 5 Cr - % Mo

The gap in data between room temperature and

700 F also exists for elongation and reduction of area Within this limitation, the data exhibit trends with increasing temperature similar to

Elevated temperature tensile strength results were

steel; yield strength results were available for only one heat, with a gap between room temperature and 1000 F The derived trend curves, particularly that for yield strength below 1000 F must be view-

ed as very approximate When compared with the

9 Cr - 1 Mo steel, to be considered next, and for which a somewhat greater number of data were avail- able, the differences are not large The data points for weld metal fell within the scatter bands for the wrought data, all of which repre- sented bar

Elongation and reduction of area values were available at elevated temperatures for only one heat of wrought material, with no values between room temperature and 1000 F

Part 6: 9 Cr - 1 Mo steel; Figs 12a, b, c The yield and tensile strength ratio trend curves for 9 Cr - 1 Mo steel appear fairly well defined, although there is not an abundance of data For this material, some of the weld metal strength ratios fall significantly above the scatter bands representing bar and tube stock

The ductility data exhibit the same general trend with increasing temperature as did the other steels of the present evaluation The low

ductility values for weld metal represent one lot whose tensile strength at room temperature was very high, 130 ksi, tested in the as-welded con- dition; commonly, a post weld heat treatment would

be imposed

CREEP AND RUPTURE PROPERTIES The available creep and rupture test results are tabulated in Table IV, separated into 6 parts ac- cording to nominal grade composition

The rupture data were evaluated to provide rupture strengths corresponding to 1000, 10,000

plotted in either instance, on double logarithmic

coordinates The indirect evaluations employed the

Larson-Miller time-temperature parameter The

direct evaluations were performed on individual

lots The indirect or parameter evaluations were

carried out only on a "universalized" basis assum-

ing a universal value for the parameter constant,

even though it is recognized that the constant may

vary from lot to lot As noted in earlier reports

in this series, the available data are of such a

character as to preclude individual lot parameter

evaluations (as would be preferable) on other than

a fraction of the available data

When the direct evaluations required extra-

polation, this was performed visually, with greater

weight given to the longer time or slower-rate

data Rupture extrapolations were made only when

it seemed reasonable to assume linearity for the

longer time data However, it is recognized that,

especially at the higher temperatures, a trend to

bilinearity or curvilinearity, downward in either

instance, may develop and, if not recognized, lead

to a non-conservative extrapolation The log

stress-log secondary creep rate plots frequently

exhibited curvilinearity at slower rates approach-

ing 0.01 percent per 1000 hours, and where extra-

polations were required, these were restricted

generally to not more than one log cycle, and per-

formed with a conservative assessment of the

scatter and the curvilinearity

The well-known Larson-Miller parameter was em-

ployed in evaluating the rupture data, using, for

all grades excepting 9 Cr - 1 Mo, the generally

assumed value of 20 for the constant c:

where T is the temperature in degrees Rankin, t

is the time in hours and F, (s) signifies that the

parameter is a function of the applied stress s

Leyda and Rowe^ had reported optimal values of c

for individual lots of various grades of steel

On the basis of these results, a value of c = 20

seems suitable for all except 9 Cr - 1 Mo For

this grade a higher value was indicated, and this

has been confirmed in unpublished work at ORNL,^^

from which the value of C = 25.1 used here was

taken

The Larson-Miller parameter was also employed

for evaluating the secondary creep rate data in

the form:

where r is the secondary creep rate in percent

per hour

Part 1: 3 Cr - 1 Mo steel; Figs 13-21

To show graphically the quantity of available data,

as well as their scatter, all of the time to rup-

ture data have been plotted in Figs 13a-c; all of

the secondary creep rate data have been plotted in

Figs 14a-c; and all of the elongation and reduc-

tion of area data at rupture have been plotted in

Figs 15a-g Data were available for bar, plate,

kept in mind in assessing the degree of scatter of the data in Figs 13-15

Rupture Strength The results of the individual lot interpolations

or extrapolations are plotted in Figs 17a-c as dependent upon temperature and tabulated in Table

VI The universalized Larson-Miller parameter scatter band (c = 20), representing all test times greater than 5 hours, is shown in Fig 16 In both Figs 16 and 17, a distinction is made as to heat treatment, and thus, indirectly, as to strength at room temperature Inspection of the plots re- veals that the data representing quenched and tem- pered plate material having a tensile strength at room temperature exceeding 110 ksi fall in a dis- tinctly different region of the plot than do the data for annealed or normalized and tempered material, for which the tensile strength at room temperature is less than 100 ksi The data for a single lot of quenched and tempered material having

a tensile strength of 91 ksi also fall within this second region

To explore further the relationship between rupture strength and room temperature tensile strength, the results of the individual-lot deter- minations of 10,000 hour rupture strength (Table VI) have been plotted (open symbols) in Fig 18 as dependent upon tensile strength Also shown are the few data for rupture in 100,000 hours (filled synbols) The 10,000 hour data have been assessed

by the method of least squares; and the resulting trend lines are shown in Fig 18 Estimated trend curves have been drawn visually for the few 100,000 hour data, paralleling those for 10,000 hours Although the quantity of data are too few, even for rupture in 10,000 hours, to provide more than an approximate assessment of the average de- pendence upon room temperature tensile strength, it

is clear that these properties are interrelated,

increases with increasing room temperature strength, the rate of increase decreasing with increasing temperature, and still faintly evident at 1100 F For quenched and tempered 2^g Cr - 1 Mo steel, the dependence had pretty well washed out by 1000 F The effect of room temperature strength upon the creep rupture properties has not heretofore been considered for purposes of establishing ASME Code allowable stresses, and in fact, the avail- able data for 3 Cr-Mo steel are inadequate to do

so Analyses have therefore been made of all data representing other than quenched-and-tempered material (the quenched and tempered condition having not yet been recognized by the ASME Code for service in the creep range), without regard to room temperature strength, to provide a current best assessment of creep and rupture strengths for purposes of setting allowable stresses, on an interim basis In one of these analyses, the uni- versalized Larson-Miller data for annealed and nor- malized and tempered wrought material have been evaluated by the least squares method to define the mean curve of best fit; this is shown on Fig

16 A minimum curve representing a lower bound

analysis, the rupture strength evaluations of indi-

vidual lots (excluding quenched and tempered mater-

ial), plotted in Figs 17a, b and c, have been

evaluated by the least squares method; the result-

ing mean and minimum trend curves are shown in the

figures and tabulated in Tables VIII and IX

The average rupture strengths derived by the

parameter analysis are superimposed upon Figs 17a-

c for comparison with the results by individual lot

analysis Inspection of the plots reveals gener-

ally good agreement, with the greatest differences

not exceeding about 10% The differences in min-

inum rupture strength, Table IX, are of the same

order

With the master parameter curve, it is also

possible to compute the isothermal log stress-log

rupture time curves to permit a visual test of how

well the computed curves represent the test data

Such computed isothermal trend curves are super-

imposed upon the plots of Figs 13a-c However,

there are too few data to form a judgment

Creep Strength

The results of the individual lot interpolations or

extrapolations to define creep strengths corres-

ponding to secondary creep rates of 0.1 and 0.01%

per 100 hours are included in Table VII, and are

plotted in Fig 20 The universalized Larson-

Miller secondary creep-rate parameter scatter band

is shown in Fig.19 Again, it is clear that, as

with rupture strength, creep strength depends upon

tensile strength at room temperature Accordingly,

a mean trend curve was developed for other-than-

quenched-and-tempered material; this is shown in

Fig 19, together with a minimum trend curve der-

ived, as previously, from the mean curve

From the individually evaluated data of Fig

20, mean trend curves were developed, as shown

However, there were too few data to warrant devel-

oping minimum values, nor, in fact for placing much

confidence in the mean trend curves Even so, the

mean curves agreed reasonably well with trend

curves computed from the mean master parameter

curve of Fig 19 and also shown in Fig 20 Be-

cause of the relatively few individual lot data,

greater confidence should probably be placed in

the parameter result The trend curves are pro-

vided in tabular form in Table X

From the mean master parameter curve (exclud-

ing quenched and tempered material), the isothermal

log stress versus log secondary creep curves have

been computed, and are superimposed upon the data

plots of Figs 14a-c The computed curves are in

conformity with the data

The individual lot estimates of creep, strength

(0.1 percent per 1000 hours) have been plotted

versus tensile strength at room temperature in Fig

21 The trends are similar to those evident in

Fig 18 for rupture strength There were too few

data for creep strength (0.01 percent) to warrant

plotting

Rupture Ductility

Relatively few data were available for other than

plate material and hence the trends of behavior

All of the time for rupture data are plotted in Figs 22a-c; all of the secondary creep rate data have been plotted in Figs 23a-c; and all of the rupture ductility data have been plotted in Figs 24a-f

Rupture Strength The universalized Larson-Miller parameter scatter band (c = 20), representing all rupture times ex- ceeding 5 hours, is plotted as Fig 25; the results

of the individual lot interpolations or extra- polations are tabulated in Table VI and plotted in Figs 26a-c In either instance, the data for lot 2-9, having a tensile strength at room temperature

of 126 ksi is clearly separated from data for the remaining wrought materials; consequently, data for this lot have been excluded from the various regression analyses related to Figs 25 and 26 The few data for cast material fell within the scatter band for wrought material and were in- cluded in the regression analyses

Examination of the wrought data, other than lot 2-9, reveals no effect of room temperature tensile strength upon elevated temperature rupture strength (10,000 or 100,000 hours) However, there are too few test results at temperatures below

1200 F to warrant drawing a general conclusion to that effect At 1200 F, no effect of room tempera- ture strength level is evident Accordingly evalu- ations for rupture strength have been performed for all data other than lot 2-9 and weld metal on the assumption of a common population

The temperature dependencies of the individual lot interpolations or extrapolations have been examined by the method of least squares, and the resulting lines of best fit are superimposed upon the plots of Figs 26a-c Rupture strengths cor- responding to 1000, 10,000 and 100,000 hours taken from these trend curves are included in Table VIII Minimum rupture strengths derived, as previously described, from the mean curves are also shown in the plots, and included in Table IX

Superimposed upon the parameter scatter band (Fig 25) is the mean curve of best fit for the data as determined by the method of least squares Also shown is a minimum curve derived from the mean by substracting 1.65 multiples of the stand- ard deviation From the mean curve, rupture strengths corresponding to 1000, 10,000 and 100,000 hours have been computed, and the resulting trend curves have been superimposed upon the plots of Figs 26a-c, and included in Table VIII The trend curves agree well with those derived from the individual lot evaluations Minimum rupture strength values have also been derived from the minimum curve of Fig 25 These values are in- cluded in the summary comparison of Table IX, and

on the whole, agree well with those developed by individual lot analysis

A final comparison may be made by computing the isothermal log stress versus log time for rupture curves from the mean curve of Fig 25, and superimposing the results upon Figs 22a-c The computed curves agree satisfactorily with the

or extrapolations assessing average creep strengths

corresponding to 0.1 and 0.01 percent per 1000

hours are included in Table VII and plotted in

Figs 28a and b Excluding the data for lot 2-9,

as discussed previously, the temperature variations

of creep strength were developed by the method of

least squares and are shown in Figs 28a and b and

included in Table X (mean values)

The universalized Larson-Miller secondary

creep rate parameter scatter band is shown in Fig

27, with superimposed average and minimum trend

curves From the average curve, creep strengths

corresponding to secondary creep rates of 0.1 and

0.01 percent per 1000 hours have been computed and

these are superimposed on Figs 28a and b, and in-

cluded in Table X Close agreement is evident be-

tween the creep strengths developed by the indi-

vidual lot and parameter evaluation procedures,

particularly for 0.01 percent per 1000 hours

From the average trend curve for the parameter

scatter band, isothermal log stress-log secondary

creep curves have been computed, and these are

shown superimposed on the isothermal scatter bands,

Figs 23a, b and c Reasonably good agreement is

evident

Minimum creep strengths have been computed

for both the individual lot data and the parameter-

ized data, by the method described previously, and

are included in Table XI

Rupture Ductility

Except at 1200 F, there were relatively few data,

especially for reduction of area At 1200 F, a

slight tendency for lower ductility at longer rup-

ture time may be detected The ductility of the

high strength lot 2-9 falls on the low side rela-

tive to other wrought material

The time for rupture data are plotted in Figs

29a-c, the secondary creep rate data in Figs 30a-

c, and the rupture ductility data in Figs 31a-f

Rupture Strength

The universalized Larson-Miller rupture parameter

scatter band (c = 20) is shown as Fig 32 Inspec-

tion reveals that the few data for castings fell

outside the scatter band of data for wrought metal

Within this latter category, there is no evidence

for an effect of room temperature strength level

upon rupture strength (or creep strength), but it

should be noted, Fig 9b, that the limited data

did not encompass a significant range in room temp-

erature strength Average and minimum trend

curves for wrought metal, by the method of least

squares, have been superimposed upon the data

scatter band

The individual lot rupture strength evalua-

tions are plotted against temperature in Figs 33a-

c with superimposed best fit average and minimum

trend curves Tabulated values for the individual

lot evaluations are included in Table VI Rupture

strengths derived from the trend curves are in-

by the two evaluation procedures This is es- pecially noteworthy in view of the relatively small number of available data

The average parameter trend curve of Fig 32 has also provided a basis for computing the iso- thermal log stress-log time for rupture curves that are superimposed in Figs 29a-c upon the available data Good agreement is evident

Creep Strength The individual lot evaluations for creep strength (0.1 and 0.01 percent per 1000 hours) are included

in Table VII and are plotted in Fig 35 Average trend curves developed by the method of least squares are also shown in Fig 35 and included in Table X

The universalized Larson-Miller secondary

steel is shown in Fig 34 Mean and minimum trend curves, excluding data for castings, have been superimposed upon the plot Creep strengths cor- responding to secondary creep rates of 0.1 and 0.01 percent per 1000 hours have been computed also from the mean master parameter curve and these are plotted in Figs 35a and b and included

in Table X The creep strength vs temperature trend curves by individual lot and by parameter evaluations are in fair (0.1%) to good (0.01%) agreement

From the master parameter curve, isothermal log stress vs secondary creep rate curves have been computed, and these are superimposed upon the scatter bands, Figs 30a-c Reasonably good con- formity with the few data is evident

Minimum creep strengths have also been de- veloped by the two evaluation procedures and are included in Table XI As expected on the basis

of the few data available, larger differences are observed than is evident for the average values Rupture Ductility

The relatively few ductility data show good rupture ductility within the restricted time and tempera- ture limits that they represent

The time-for-rupture data are plotted in Figs 36a-c, the secondary creep rate data in Figs

37a-b and the rupture ductility data in Figs

38a-b The number of data are extremely sparse except for temperatures of 1000 and 1200 F

Rupture Strength The universalized Larson-Miller rupture parameter scatter band (c = 20) is shown in Fig 39 All of the data represent wrought material However, in- spection of Fig 39 reveals that the data are poorly distributed, and it seems possible that the data population might be mixed, or perhaps that there is important microstructural instability The limited tensile strength data at room tempera- ture, Fig 10b, show only a very limited range in

curve derived from the average is also shown

The individual lot rupture strength evalua-

tions are plotted in Figs 40a-c and are included

in Table VI With the individual lot values

bunched at the two temperatures 1000 and 1200 F,

the least squares evaluation not surprisingly in-

dicated a nonlinear interrelation between the var-

iables, which, on inspection, seemed unreasonable

It therefore seemed best to force a linear relation,

and accordingly the average and minimum trend

curves for the individual lots shown in Figs 40a-c

and included in Tables VIII and IX reflect this

arbitrary decision Also superimposed upon Figs

40a-c are average trend curves computed from the

master parameter trend curve of Fig 39 These

values, as well as minimum values not shown in

Figs 40a-c, are included in Tables VIII and IX

In view of the poor character and quantity of the

data and the arbitrary judgments concerning the

least squares analyses, the agreement between the

results by the two evaluation procedures is good

From the average parameter trend curve of

Fig 39, isothermal log stress vs log time for

rupture curves have been computed and are super-

imposed upon the data plots of Figs 36a-c At

1000 F, agreement is poor; this is also evident in

Fig 40a, where the individual lot evaluations

should be accurate since they involve principally

interpolation on the isothermal plot On this

argument, the strength-temperature regression for

the individual lots in Fig 3 has also given a too-

conservative result

Creep Strength

The universalized Larson-Miller secondary creep

rate parameter scatter band is shown in Fig 41,

from which it is evident that the number of data

is quite limited Average and minimum trend curves

are superimposed upon the data

The individual lot creep strength evaluations

are tabulated in Table VII and plotted in Fig 42

Average and minimum trend curves, by the least

squares procedure, are superimposed upon the data,

and tabulated in Tables X and XI Little confi-

dence can be attached to the trend curves since

there are so few data (For this same reason, the

trend curves were arbitrarily forced to a linear

variation.) Also superimposed upon the plots of

Fig 42 are computed average trend curves by the

parameter evaluation procedure, which probably

warrant more confidence than the individual lot

trend curves These average creep strengths by

the parameter procedure as well as minimum values,

not plotted, are included in Tables X and XI

Isothermal log stress vs log secondary creep

rate curves, computed from the master parameter

curve, Fig 41, are superimposed upon Figs 37a-b

Rupture Ductility

Only scattered data are available at temperatures

other than 1000 and 1200 F At these two tempera-

tures, elongation remained good at the longest

test times, none of which, however, exceeded 6000

hours

Rupture Strength The universalized Larson-Miller rupture parameter scatter band (c = 20) is plotted in Fig 46 The data appear reasonably distributed in contrast with

minimum trend lines have been superimposed upon the data of Fig 46

The individual lot rupture strength evalua- tions are tabulated in Table VI and plotted in Figs 47a-c Average and minimum rupture strength

vs temperature trend lines are superimposed upon the data and included in Tables VIII and IX Also plotted in Figs 47a-c are average trend curves computed from the master parameter trend curve of Fig 46 These values, as well as minimum values not plotted in Fig 46, are included in Tables VIII and IX Comparison of the average trend curves of Figs 47a-c shows very good agreement between the results developed by the two evalua- tion procedures

Isothermal log stress vs log time for rupture curves have also been computed from the average parameter trend curve of Fig 46, and these are superimposed upon the plots of Figs 43a-c Good conformity is evident

Creep Strength The universalized Larson-Miller secondary creep rate scatter band is plotted in Fig 48 with superimposed average and minimum trend curves The individual lot creep strength evaluations are tabulated in Table VII and plotted in Fig 49 Average and minimum trend curves for the limited data are superimposed on the data of Fig 49 and included in Tables X and XI Average and minimum trend curves were also computed from the master parameter trend curves of Fig 48 The average and minimum values are included in Tables X and

XI The average values have also been plotted in Fig 49 for comparison with the individual lot trend curves Agreement is fairly good for 0.1% per 1000 hours but poor for 0.01% per 1000 hours

at temperatures above about 1050 F Finally, the master parameter curve has permitted computing isothermal log stress vs secondary creep rate curves, and these have been superimposed on Figs 44a-b They are in reasonable conformity with the data

Rupture Ductility The relatively few data exhibit no evidence for im- paired ductility within the restricted limits represented

Part 6: 9 Cr - 1 Mo steel; Figs 50-56 The time for rupture data are plotted in Figs 50a-c, the secondary creep rate data in Figs 51a-c, and the rupture data in Figs 52a-d

Rupture Strength The universalized Larson-Miller rupture scatter

trend lines have been computed from the master

parameter curve of Fig S3, and are also super-

imposed upon the plots of Figs 54a-c Tabular

values for these average curves and also for min-

imum trend curves, not plotted in Figs S4a-c, are

included in Tables VIII and IX The agreement be-

tween the average trend curves developed by the

individual lot and parameter evaluation procedures

is increasingly poor below about 1100 F as the ref-

erence time is extended beyond 1000 hours Since

the discrepancy is such that the parameter pro-

cedure produces the more conservative result, one

possible explanation for the discrepancy is that a

strengthening reaction occurs at lower temperature

that is not influential at the higher temperatures,

an inherent and well-recognized possible complica-

tion of the parameter evaluation procedure In

this connection, it is of interest that the secon-

dary creep rate parameter scatter band to be dis-

cussed later, Fig 55, does exhibit an unusual

upward concavity Perhaps a greater volume of

rupture data representing lower parameter values

would have resulted in an appropriately modified

master parameter curve for rupture

Isothermal log stress vs log time for rupture

curves have been computed from the average para-

meter trend curve of Fig 53, and these have been

superimposed upon the plots of Figs 50a-c With

reference to the discussion of the preceding para-

graph, it is of interest that the computed curves

fall below the data at 1000 and 900 F, but, of

course, the data are few in number

Creep Strength

The universalized Larson-Miller secondary creep

rate parameter scatter band is shown in Fig 55,

with superimposed average trend curve As cited

above, this trend curve has an uncommon upward

concavity

The individual lot creep strength evaluations

aTe tabulated in Table VII, and plotted in Fig 56

Average and minimum trend curves, which exhibited

least variance for the first order interdependence

of the variables, are shown superimposed upon the

data, Fig 56, and are included in tabular form in

Tables X and XI Also superimposed upon the plots

of Fig 56 are average trend curves computed from

the master parameter curve These average values

as well as minimum values, not plotted in Fig 56,

are tabulated in Tables X and XI The agreement

between the average values in Fig 56 are fair for

the 0.1% per 1000 hour rate, good for the 0.01%

per 1000 hour rate at 1000 and 1100 F, but diver-

gent at higheT temperatures

A comparison of computed isothermal curves

with the test data is afforded by Figs 5la-c

Reasonable conformity is evident

Rupture Ductility

Within the limits of the data, extending to beyond

10,000 hours at several temperatures, ductility is

maintained at greater than 20 percent elongation

100,000 hours, of interest for setting allowable stresses under the ASME Code, the average values for other grades were sometimes more conservative

by the one procedure and sometimes by the other, and similarly for minimum rupture strength Thus, the choice between the results by the two pro- cedures is not, on this basis, readily apparent However, the parameter procedure does suffer from the disadvantage that it cannot provide 100,000 hour values at the higher test temperatures, un- less a hazardous extrapolation of the master para- meter curve is made This disadvantage derives from the tradeoff between short time at higher temperature and longer time at low temperature that is inherent in the time-temperature parameter

As a consequence, 100,000 hour strengths towards the top of the temperature range of practical interest cannot be developed unless tests are made

at even higher temperature Such tests have seldom been included in the data gathered by MPC Prin- cipally for this reason, then, it seems that the choice between evaluation procedures for 100,000 hour rupture strength should favor the individual lot procedure

Creep Strength

In general, there was good agreement between the results by the two evaluation procedures, and no readily apparent basis for choosing between them However, in contrast with 100,000 rupture strength for which extrapolation in time is always involved, evaluation of the secondary creep rate data for the stress for 0.01% per 1000 hours may involve data interpolation, and parameter methods serve an important purpose in correlating data Also, there were often relatively few creep rate data, such that, for example, creep strength (0.01 percent per 1000 hours) could not be appraised at tempera- tures below 1000 F for several of the grades For these reasons, it has seemed best, on the whole, to favor the parameter result for the stress

to cause a secondary creep rate of 0.01 percent per 1000 hours

COMPARISON! OF GRADES The yield and tensile strength ratio trend curves

of the six grades of steel are compared in Figs

57 and 58 In either instance, there is no orderly trend evident for the dependence upon alloy con- tent Differences are sometimes small, as for examples amongst grades 21, 5, 5b and 9 for yield strength ratios at temperatures between 75 and

700 F, and may not be statistically meaningful The average rupture strengths (100,000 hours)

of the 6 grades are compared in Fig 59 and the average creep strengths (0.01% per 1000 hours) in Fig 60 Here, there is a measure of orderliness, with the 9 Cr - 1 Mo and 3 Cr - 1 Mo grades ex-

grades (except at the highest temperatures in the case of creep strength) Again, the extent to which the differences are significant is uncertain

References 17 Oak Ridge National Laboratory; unpublished

work of P Rittenhouse

1 W F Simmons and H C Cross: Report of the

Elevated-Temperature Properties of Chromium-

Molybdenum Steels; ASTM STP No 151 (1953)

2 J.A Van Echo and W F Simmons: Supplemental

Report on the Elevated-Temperature Properties

of Chromium-Molybdenum Steels; ASTM Data

Series No DS 6S1 (1966)

3 Resume of High Temperature Investigations

Conducted During 1955-56; The Timken Roller

Bearing Co., Steel and Tube Div., Canton,

Ohio (1956)

4 Lukens Steel Co

5 H R Voorhees and J W Freeman: The Elevated-

Temperature Properties of Weld-Deposited Metal

and Weldments; ASTM STP No 226 (1958)

6 Compilation of Available High-Temperature

Creep Characteristics of Metals and Alloys;

Joint ASTM-ASME Committee on Effect of Temper-

ature on the Properties of Metals; March 1938,

published by ASTM and ASME, Phil, and New York

7 Timken Roller Bearing Co., Steel and Tube Div.,

Canton, Ohio; Resumes of High Temperature

Investigations 1940 et seq.; also Digest of

Steels; also unpublished data

8 G N Emmanuel and W E Leyda: Long-Time

High Temperature Properties of Cr-Mo Weld

Metal; in "Properties of Weldments at Ele-

vated Temperatures; ASME (1968)

9 U S Steel Corp

11 Combustion Engineering, Inc

12 I Finnie and A E Bayce: Creep-Rupture

Tests on 9% Cr - 1% Mo Furnace Tube Material;

Proc of Joint Int Conf on Creep, New York

(Aug 1963), London (Oct 1963)

13 G V Smith: Supplemental Report on the

Elevated Temperature Properties of Chromium-

1 Mo Steel); ASTM Data Series Publication

DS 6S2, March 1971

14 G V Smith: Evaluation of the Elevated

Temperature Tensile and Creep-Rupture Pro-

Publication DS 50, September 1973

15 G V Smith: Quenched and Tempered Steels for

Pressure Vessel Service at Elevated Tempera-

tures; Second International Conference on

Pressure Vessel Technology, San Antonio,

Oct 1973; Part 3, page 273

Table I Identification of Steels

ASTM

Code Spec Deoxid

Product

Ref Code

No Part 1 - 3 Cr - 1 Mo steels

,T1190 ,T1320 ,A1550

,T1425 T1425 ,T1425 ,T1425 T1425

Tube, 5"x5/8"w Tube, 5"x5/8"w

A-Annealed; N-normalized; HR-hot rolled; T-tempered or stress-relieved; Q-quenched

Actual grain size, except when identified as M for McQuaid-Ehn

1" slice from quarterline; treated to simulate W.O of 8" plate

f41 '1" slice treated to simulate W.Q at centerline of 11" plate

Table I - page 2

Part 3 - 5 Cr - % Mo-Si steels

Categorized as 1% Si steel in Ref 1

Categorized as 1.5% Si steel in Ref 1

Table I - page 3

Metal arc Met a1 arc

N1750.A1500 Annealed A1575 A15S0 Annealed per A-213 Annealed per A-213 A1575

As welded A13S0

Bar, 1"

Wrought Bar, 1"

Bar, 1"

Bar, 3/4"

Tube Wrought Weld metal Weld metal

Table I - page 4

Part 6 - 9 Cr - 1 Mo steels

1200 min

Table II Chemical Composition of Steels

.010 014 007 018 014 012

.011 015 003 024 013 007

.92 80 88 80 72 97 44

.11 20 19

.59 53 52 47 50 58 46

.09 14 03

.018N

Table II - page 3

Other

Ti Part 6 - 9 Cr - 1 Mo steels

Table III Short-Time Tensile Properties

Code No

Test

Temp °F

1000 psi Yield Strength* Tensile Strength

70.2 60.4 58.8 57.0 51.4 43.5 29.4 92.3 75.4 71.4 69.3 62.3 52.6 34.6 102.8 86.3 81.5 75.2 73.4 53.3 41.1

* 0.2% offset, unless noted otherwise

• Elongation in 2 inches, unless noted otherwise,

a 0.1% offset

Table III - page 2

Table III - page 3

Percent Code No

Table III - page 4

Table III - page 5

Table III - page 6

Test

Part 3 - 5 Cr - % Mo-Si steels

Table III - page 7

Table III - page 8