Text

List of References Associated with ASME Code Case N-499 to Support NRC Review of Inservice Inspection (ISI) Program Relief Request No. 33 for Pressurizer Base Material Heating
	ML051790349
Person / Time
Site:	Palo Verde
Issue date:	06/21/2005
From:	Mauldin D; Arizona Public Service Co
To:	; Document Control Desk, Office of Nuclear Reactor Regulation
References
	102-05298-CDM/TNW/GAM
	Download: ML051790349 (360)
	v • d • e

ZAI FS 10 CFR 50.55a(a)(3)(i)

David Mauldin Vice President Mail Station 7605 Palo Verde Nuclear Nuclear Engineering Tel: 623-393-5553 PO Box 52034 Generating Station and Support Fax: 623-393-6077 Phoenix. Arizona 85072-2034 102-05298-CDMITNW/GAM June 21, 2005 U.S. Nuclear Regulatory Commission Attn: Document Control Desk Washington, DC 20555-0001

Dear Sirs:

Subject:

Palo Verde Nuclear Generating Station (PVNGS)

Unit 3 Docket No. STN 50-530 List of References Associated with ASME Code Case N-499 to Support NRC Review of Inservice Inspection (ISI) Program Relief Request No. 33 for Pressurizer Base Material Heating In letter 102-05296, dated June 19, 2005, Arizona Public Service Company (APS) requested NRC approval of proposed alternatives to 10 CFR 50.55a(c), "Reactor Coolant Pressure Boundary," for a portion of the PVNGS Unit 3 pressurizer base material surrounding the heater sleeves that was subjected to elevated temperatures (ISI Relief Request No. 33).

During telephone conferences on Monday, June 20 and Tuesday, June 21, 2005 regarding Relief Request No. 33, the following references were discussed:

1. Portions of ASTM DS 47, "Evaluations of the Elevated Temperature, Tensile and Creep - Rupture properties of C-Mo, Mn-Mo and Mn-Mo-Ni Steels," Prepared for MPC By G. V. Smith, 1971.

2. Portions of ASME Publication, "Symposium on Heat-Treated Steels for Elevated Temperature Service," September 1966; Article: "Characterization of Heat Treated Pressure Vessel Steels for Elevated Temperature Service," Edited by M. Semchysen, 1966.

A member of the STARS (Strategic Teaming and Resource Sharing) Alliance Callaway f Comanche Peak 0 Diablo Canyon O Palo Verde 0 South Texas Project 0 Wolf Creek

U. S. Nuclear Regulatory Commission ATTN: Document Control Desk List of References Associated with AMSE Case Case N-499 to Support NRC Review of Inservice Inspection (ISI) Program Relief Request No. 33 Page 2

3. Portions of ASME-MPC Publication, "Analysis of Data from the Symposium on Heat-Treated Steels for Elevated Temperature Service," By E. B. Norris and R.

D. Wylie, 1966.

4. DOE-HTGR-88383, 'Tensile and Creep Properties of SA533 Grade B Class 1 Steel," December 1989.

5. DOE-HTGR-90286, "Documentation of ASME Code Case for Elevated-Temperature Service of MHTGR Reactor Vessel Materials," September 1991.

6. Combustion Engineering Report MML-89-142, "Creep and Tensile Properties of SA508 Class 3 Forging Material," December 1989.

7. Oak Ridge National Laboratory Letter 0409-49-90 regarding submission of data package for SA-533 Grade B, Class 1 plates, SA-508 Class 3 forgings and their weldments, dated April 20,1990.

8. ASME Publication, "Criteria for Design of Elevated Temperature Class 1 Components in Section III, Division 1, of the ASME Boiler and Pressure Vessel Code," May 1976.

The applicable portions of the documents listed above are attached to this submittal.

No commitments are being made to the NRC by this letter. If you have any questions, please contact Thomas N.Weber at (623) 393-5764.

Sincerely, CDM/TNW/GAM Attachments: Applicable portions of the documents listed in this letter.

cc: All w/o attachments B. S. Mallett NRC Region IV Regional Administrator M. B. Fields NRC NRR Project Manager G. G.Warnick NRC Senior Resident Inspector for PVNGS

1. ASTM DS 47, "Evaluations of the Elevated Temperature, Tensile and Creep -Rupture properties of C-Mo, Mn-Mo and Mn-Mo-Ni Steels".

I aI1 :

-4 EVALUATIONS OF THE i'0 0

2.

T1 ELEVATED TEMPERATURE TENSILE AND CREEP-RUPTURE PROPERTIES OF C-Mo, MnMo and Mn-Mo-Ni STEELS*.

L Preparedfor the

METAL PROPERTIES COUNCIL

' I

'3

.by G.V.Smith

-4 DS 47 .,

I V

S,01AMERICAN SOCIETY FOR TESTING AND MATERIALS i

I

.A

SC--

EVALUATIONS OF THE ELEVATED TEMPERATURE TENSILE AND CREEP-RUPTURE PROPERTIES OF C-Mo, Mn-Mo and Mn-Mo-Ni STEELS Prepared by G. V. Smith ASTM DATA SERIES PUBLICATION DS 47 05-047000-02 List price $6.25

'I' Gol AMERICAN SOCIETY POR TESTING AND MATERIALS 1916 Race Street, Philadelphia, Pa. 19103 mummoo

REFERENCE:

Smith, G. V., Evaluations of the Several summary figures immediately following Elevated Temperature Tensile and Creep-Rupture this abstract, Figs. 1-6, show the temperature Properties of C-'Mo, Mn-Mo and Mn-Mo-Ni Steels; dependence of strength properties for the various ASTM Data Series DS 47. American Society for grades evaluated in this report. in these figures, Testing and Materials, 1971. the-yield and tensile strength curves have been computed from the respective ratio trend curves ABSTRACT: This report evaluates the elevated so that they corresnond at room temnerature to the temperature strength properties of carhon-moly specified minimum values of common ASD. specifica-steel, and various modified versions of that steel tions.

commonly identified as manRanese-molybdenum or The body of the report provides in tables, manganese-molybdenum-nickel steels. The data that text, and figures, details concerning the identifi-have been evaluated encompass test results pre- 1 cation of the individual lots of material, the viouslv included in ASTM Data Series DS 6 (lDS3)t evaluation procedures, and the results.

and DS6-SI (1966)'2), and previously unpublished test results gathered by The Metal Properties KEY WORDS: elevated temperature, tensile strength, Council. yield strength, creen strength, rupture strength, Employing the method of least squares, trend elongation, reduction of area, carbon molv and curves depicting in ratio form the characteristic manganese moly steels, time-temperature parameter, temperature variations of yield and tensile data evaluation, mechanical properties.

strengths have been developed. The rupture data have been evaluated by both direct isothermal interpolation or extrapolation, and by time-temperature parameters, to establish the tempera-ture dependences of the average and minimum stresses to cause rupture in 1000, 10,000 and 100,000 hours0 days 0 hours 0 weeks 0 months . The secondary creep rate data have been evaluated by direct interpolation or extra-polation to determine the temperature dependences of average and minimum stresses to cause secondary creep rates of 0.1 and 0.01 percent per 100n hours.

These latter trend curves could be developed onlv for C-M1o steel, there being too few data for the remaining grades to warrant such evaluation.

Elongation and reduction of area data are included for both the short time elevated temperature tensile tests and for the rupture tests.

1 I.-

A . gg}iM t 4wg MM MIMIMU-4 iIa NO 044+ WW mm MM# MU7M' fr a mm0WWWm 44 +

l l Ue5 WOM 74 RR a 9 Qi M Wt RRt W gggg gt gM tX m@ !g i i iMM 0 Fig 2 Effc of ite ertr on yield strengt7 andi tenile strHengtEh of b~-o(Iptee A 302 (o A53 Cls 1) adjuw tedWW Ws f B FE3 503an 80 ksHIa te EH:iF- HiFE H 75 4

INTRODUCTION The materials evaluated in this report include fracture have been included in the report for both C-No steel (0.5 percent molybdenum) which has been the tensile and rupture tests, when available.

widely used in elevated temperature service over many years, and the newer Mn-Mo, and Mn-kMo-Ni Yield Strength, Tensile Strength, Elongation and modifications of this basic grade. Evaluations Reduction of Area have been made for various heat-treated conditions, including quenching and tempering, and for various The original tensile test results, excepting product forms. The report is another in a con- those previously reported in DS 6 and DS 6SI, are tinuing series of evaluations sponsoTed by The tabulated in Table IIT. Many of the reported Metal Properties Council (%IPC) (6-1 '- values represent the average of replicate tests.

Included in the evaluations are data pre- As with the previous evaluations in this series, viously reported in ASTN's DS Data Series,(l, 2 ,l 9 ) the tests are presumed to have been conducted under sponsorship of the Joint ASTH-ASME Committee generally at strain rates within the limits per-on the Effect of Temperature on the Properties of mitted by AS1'f Recommended Practice E 21, and the Metals, as well as data recently gathered by MPC yield strengths to represent either 0.2% offset, from cooperating laboratories. Pertinent data or the lower yield point. Unless otherwise indi-reported many years ago in the 1938 Creep Data cated, elongation values represent a gage length Compilation( ) have also been included. All of of two inches, and in the case of plate material, the data are identified in Table I as to ASTM test specimens were taken from the quarter thick-specification, deoxidation practice, heat treat- ness position.

ment, product form and size, grain size, and data Employing a data normalizing procedure at source, to the extent that these were known, and has proved useful in previous evaluationstl i) in Table II with respect to chemical composition. the elevated temperature yield and tensile Some of the data sets from DS 6 and DS 6S1(' 2) strengths of individual lots have been ratioed to were excluded from the evaluation, owing to inade- the room temperature yield and tensile strengths quate identification or non-conformance with of the same lots. Then, each set of such ratios, specifications. The data gathered by MPC are tabu- representing individual data populations, e.g.

lated in this report; data from DS 6 and DS 6S1 C-1o steel, has been evaluated by the procedure of have not been copied into this report, but a coding least snuares to establish a "ratio trend curve" key to the DS data that have been integrated into of best fit through all the data. With the tempera-the evaluations is provided in Table I of the ture dependence of strength expressed in terms of present report. strength ratios, it becomes possible to compute The tabular data for each of the several types strength trend curves for any specific room-of materials, or heat treatment variations of a temperature strength level of interest, within the given type of material, included in the evaluations, limits encompassed by the original data.

have been grouped separately, as follows: The tensile test results for the different categories are plotted as dependent upon tempera-Part l: C-Mo steels (Specs. A204, A209, A335, ture in Figures 7-12, corresponding with the indivi-A369, A182, A217) dual Parts into which the data have been grouped.

In each figure, part (a) charts yield strength and Part 2: Mn-Mo and Mn-Mo-Ni Steel Plates yield strength ratio; part (b) charts tensile (A302) strength and tensile strength ratio; and part (c)

Part 3: Mn-Mo and Mn-Mo-Ni Steel Plates, charts elongation and reduction of area. No data Ouenched and Tempered (AS33) for either weld metal or weldments are included in the figures inasmuch as no elevated temperature Part 4: Mn-Nlo Steel Forgings (A372, Class IV) test results were received by MPC, nor were there Part S: Mn-Mo Steel Castings (A487, Classes data in anyglier report covering weld metal and 2N, 20) wcldments.(J Specific comments concerning the individual Part 6: Mn-Mo Steel Plate, Quenched and groups follow:

Temnered (A514, Type C).

Part 1: C-Mo steels, Figs. 7a, b and c In considering the data of Part 1, a dis-tinction has been preserved in the early stages of Data for the different product forms, plate.

the evaluation as to product form (bar, plate, pipe-tube, bar, and castings have been dis-tube or pipe, casting), but in the final stage of tinguished from one another. All of the plate trend curve evaluation, it has seemed appropriate materials fell within limits corresponding to to consider the data for different product forms Grade B of A204, with several lots also meeting as from one population. In the remaining Parts, the requirements of either Grade A or C. Con-the product form was unique to the individual siderable scatter is evident, especially for yield Part.

The properties that have been evaluated in strength,* and normalizing the strength data by this report include yield and tensile strengths, and creep and rupture strengths. Unfortunately, the latter two properties could he evaluated only As suggested in an earlier publication , a for Part l, the number of data being either too significant portion of the scatter in yield strength few or nonexistent for the remaining Parts. For probably reflects the difficulty of measuring Part 1, rupture strength has been evaluated for small strains at elevated temperatures, the pos-three rupture intervals, namely; 1000. 10,000 and sible presence of residual stresses from 100,000 hours0 days 0 hours 0 weeks 0 months , and creep strength for two secondary straightening or specimen preparation procedures, creep rates, namely 0.1 and O.0l percent per 1000 and possible differences in strain rate, factors hours. Elongation and reduction of area at which have a lesser effect upon tensile strength than upon yield ltrength.

9

- I

r ratioing has not proved particularly effective in The trend curve for tensile strength indicates reducing scatter. Inspection of the ratio plots by the rise at intermediate temperatures a slight suggests somewhat greater yield strength ratios tendency for dynamic strain aging. However, the for plate than for bar, but this difference is not strength ratio remains below 1.0, in contrast to evident for tensile strength ratio. Unfortunately, the C-Mlo steels of Part 1, and the peak occurs at the extent of overlap in temperature for these two a slightly higher temperature, SSO F as compared product forms is limited, there being no plate to 4f00F. Again, the scatter in strength ratio data above 1000lF, and only isolated data for bar at the peak temperature is probably to be associated below 7nO-F. For the other two product forms, the with differing susceptibilities to strain aging. l number of data are quite limited. All in all, the character of the data is such that it has seemed appropriate to treat all of the yield and tensile Perhaps reflecting the reduced strain-aging susceptibility indicated by the tensile strength results, there is little tendency for reduced I

strength ratios as belonging to the same individual ductility at intermediate temperatures, Fig. 8c.

populations for the purposes of the least squares The scatter in ductility for the plate represented analyses. The resulting regression lines of best in Fig. Sc, is less than that evident in Fig. 7c, fit, or trend curves, have been superimposed on which represented various product forms.

the ratio plots, and are included in the tabula-tions of Table V. Part 3. Mn-Mo and Mn-Mo-Ni steel plates, quenched The tensile strength ratios that resulted and tempered, Figs. 9a, b and c from the least squares analyses using a computer began increasing immediately above room tempera- ASPS Spec. A533 includes 4 grades A, B, C and ture, and it is apparent that the two values at D corresponding with different levels of nickel 200'F, having a ratio less than one, were "over- within the range 0 to 1% and fixed amounts of the whelmed" by the weighting of all the other data. remaining elements, and 3 classes of tensile Since carbon and low alloy steels typically exhi- requirements, Classes 1, 2 and 3. Because the bit(17 ) a decreasing ratio immediately above room strength classes overlap one another and because temperature, before dynamic strain aging can be experience has indicated an insensitivity of manifested, the trend curve shown in Fig. 7b and strength ratio to strength level, within a given included in Table V has been drawn visually for material category, the plots of Figs. 9a, b and c temperatures between room temperature and 700'F. distinguish only among the several grades. At It is of interest to note that the spread in least some data were available for each of the tensile strength ratio at intermediate temperatures grades A, B and D, but none for grade C. Although is probably in part associated with differences in it is possible that more adequate samples of data strain-aging susceptibility of the different lots, for the different grades might reveal differences the degree of susceptibility depending primarily amongst them, the character of the strength ratio upon nitrogen concentration, deoxidation practice data that are available are such that it has not and heat treatment. It is also of interest to note seemed appropriate to distinguish at this time that the maximum susceptibility evident in Fig. 7b amongst the different grades. Hence the ratioed is on the order of that exhibited by carbon steeL(1 7 ) data have been considered as belonging to a common Elongation and reduction of area of C-Mo steel population. The trend lines resulting from the also exhibit much scatter, Figure 7c, with some regression analyses are shown on the ratio plots tendency for reduced ductility in the range of and included in Table V.

temperature in which dynamic strain aging is indi- The tendency for dynamic strain-aging has been cated in the tensile strength data. At higher lessened still further relative to the materials of temperature, ductility trends to higher levels. Parts 1 and 2, with the tensile strength ratio Plate and castings tend to exhibit somewhat less trend curve essentially level at intermediate ductility than bar and pipe. temperatures.

The ductility data, Fig. 9c, showed only Part 2. Mn-Mo and Mn-Mo-Ni steel plates, Figs. Ba, limited scatter, with no evident differentiation b and c amongst the three grades for which data are available.

The data that were available were reported to represent either grade B or Grade C of ASDI speci- Part 4. Mn-Mo steel forgings, Figs. 10, b and c fication A302, and are so distinguished in Figs.

Ba, b and c. However, it will be recognized that Data were available for only one lot of Mn-Mo Grade B requirements overlap those of Grade A, and steel forgings, and thus the conversion of the hence a number of the Grade B data may equally well strength data into ratios serves only the purpole represent Grade A material. Inspection of the of making it possible to express the trend curves ratio plots, Figs. Ba and b does not indicate any in ratio form. With data for only one lot, there need to distinguish between grades B and C as to is, of course, no way of knowing how representative trend curves. Some data representing material the trend curves shown on the ratioed plots and in having room temperature tensile strengths greater Table V are. The requirements of ASTM Specification than permitted by spec. A302 have been incudt in A372, IV covering this material, overlap in many the evaluation. Work previously reported, as respects with those for the materials of Parts 2 well as analysis of the present data, has indicated and 3 of this report, and the strength ratios fall that for a specific grade of material the depen- within the scatter bands of Figs. 8 and 9. How-dence of strength upon temperature, when expressed ever, the specified molybdenum content for A372, in ratio form, is insensitive to absolute strength IV is only about one-half of that required of the level within wide limits. A common population has other materials, and it has not seemed appropriate been assumed for the least squares regression to include this material with the others. Clearly, analyses. The resulting trend curves have been further testing of this material is desirable.

superimposed upon the ratio plots, and included in Table V.

IF 10

Part S. Mn-Mo steel castings, Figs. Ila, b and c isothermal relation between stress and rupture time, commonly plotted on log-log coordinates, to Data were available for 3 lots of this 100,000 hours0 days 0 hours 0 weeks 0 months , whereas the latter involves one or material, but each lot had been tested in both the another time-temperature parameter. For reasons normalized-and-tempered and quenched-and-tempered described in the earlier evaluations, the character conditions, conforming to grades 2N and 20 of AS7M of the available data is such that the indirect or Specification A487. The data are distinguished as parameter extrapolations have not been performed to heat treatment in the plots. Ratiolng of the on an individual lot basis, but rather on a strength data has been effective in reducing "universalized" basis, assuming universal values scatter, especially for tensile strength. Yield for the parameter constants.

ratios for Grade 2N exhibit relatively high scatter, but except for this grade, the ratio trend curves shown in Figs. Ila and b and included in Part 1. C-Mo steels Table V seem reasonably well defined by the data from only three lots. To show both the quantity of data and their The ductility values, Fig. llc, exhibit only scatter, all of the data are shown in isothermal modest scatter. scatter band plots of log stress versus log time for rupture (Figs. 14a, b and c); of log stress Part 6. Mn-Mo steel plate, quenched and tempered, versus log secondary (or minimum) creep rate (Figs.

Figs. 12a, b and c ISa, b and c); and of percent elongation and reduction of area at rupture versus log time for Data were available for only two lots of rupture (Figs. 16 a-g). In each plot, data for material in this category, and although it is different product forms are differentiated. A few possible that the availability of further data at data available for weld metal have been plotted, some future date may reveal that this material can for purposes of visual comparison, but these have be grouped with one or another of the other not otherwise been included in the evaluations. A material groups evaluated in the present report, it few weld metal data for temperatures of 842, 932 has seemed desirable to treat it separately for the and 1022F have not been plotted, but inspection present. Interestingly, a relatively small differ- reveals that these are not inconsistent with the ence between the strengths of the two lots has been data for the several product forms, as true also even further reduced by ratioing, giving a measure of the plotted data. A few data for weldments of confidence in the trend curves that have been (as contrasted with weld metal) have not been developed, Figs. 12a and b and Table V. Caps in included in this report owing to the inhomogeneous test temperature between room temperature and nature of weldment test specimens, and the depen-300'F and between 300 and 600'F, do raise some dence of the results upon geometrical considera-uncertainty as to the true shapes in the range tions.

below about 600*F.

Rupture Strength Comparisons of the Trend Curves The rupture data have been extrapolated iso-Tabular comparisons of the yield and tensile thermally to 100,000 hours0 days 0 hours 0 weeks 0 months both visually on an strength ratio trend curves are afforded for all individual lot basis, giving weight to the longer of the material categories in Tables Va and Vb, time tests, and by extending the line of best fit respectively, and a graphical comparison is pro- resulting from least squares analysis of the scatter bands, assuming all of the data to have vided in Fig. 13 for the three categories for cone from a common population. No clearly identi-which there were significant volumes of data. It fiable effect of product form is evident in the may be seen in Fig. 13 that the trend curves for material categories corresponding to Specifica- scatter bands, nor in the results of the individual tions A302 and A533 are reasonably similar to one lot extrapolations assembled into plots of strength versus temnerature (see later). However, for some another. For this reason, trend curves for the combined populations have also been developed, and product forms, there are few or no data. In 16con-are included in Tables Va and Vb. The common trast to earlier evaluations in this series( 18 ),

trend curves have been used for summary Figures 2 these evaluations have indicated that the relation between log stress and log time to rupture may be and 3. bilinear or curvilinear. Thus, for a number of the isothermal scatter bands, the variance of the Creep and Rupture Properties data was observed to decrease as the order of the assumed relation between the variables was The original creep and rupture data not pre- increased beyond the first degree:. In the viously reported in DS 6 and DS 6S1 (References I individual-lot extrapolations some of the log-log and 2) have been tabulated in Table IV, separated plots (not included here) seemed to exhibit a into Parts according to specification or nominal break from one slope to another steeper one; such composition. Only a few data were available for breaks might well be reflected as curvilinearity Parts 2 and 3 (corresponding to specifications in the scatter bands. Unfortunately, departure A302 and AS33) and for Part 6 (AS14C), and none from linearity introduces an element of uncertainty at all were available for Parts 4 and 5 (A372IV into the direct extrapolations, beyond the usual and A487, 2N and 20). Only for Part 1 (C-Mo) were uncertainty associated with extending a linear the data adequate to warrant evaluation to the line for one or more log cycles.

extent of developing trend curves suitable for The results of the individual lot evaluations establishing allowable stresses. 7 of the stress to cause rupture in 1N0, 10,000 and As in earlier evaluations l l"). both direct 100,000 hours0 days 0 hours 0 weeks 0 months , by interpolation or extrapolation, and indirect procedures have been employed in as required, have been assembled in Table VI, and extrapolating the rupture data to 100,000 hours0 days 0 hours 0 weeks 0 months . plotted in Figs. 17a and b. Most of the data of The former procedure involves extending the Figs. 17a and b represent bar stock. The 11

- I -

relatively few data for pipes or tube and for cast- In either parameter, T is temperature in degrees ings fall rcasonablv within the scatter of the bar Rankin, t is the runture-timc in hours, and data. T.east squares evaluations were made for F1(s) and F2(s) denote that the parameters are each rupture time, Figs. 17a, b, assuming the data different functions of the stress s.

to represent common populations; the resulting regression lines, of third order in each instance, Scatter band plots showing the dependence of have been superimposed upon the data, and included log stress unon the two parameters are provided in tabular form in Table TX. It is of interest in Figures 18a and b. Parameter values correspond-that the regression curves for the three intervals ing to 100,OnO hours at specific temperatures have have similar shapes. MIinimum position curves, been superimposed upon the Figures. Also super-derived from the mea curves by a procedure imposed upon the plots are the results of least described previousy I ), are also shown in Figs. squares evaluations, taking parameter as the 17a and 17b, and tabulated in Table TX. independent variable. Mean curves together with In the least squares evaluations of the iso- minimum curves (90% confidence) derived from the thermal scatter bands, the longer time data were mean curves(2 1) are shown. Mean and minimum values weighted by the expedient of excluding rupture for the stresses to cause rupture in 1000; 10,000 times less than 50 hours5.787037e-4 days 0.0139 hours 8.267196e-5 weeks 1.9025e-5 months , and time was taken as and 100)000 hours0 days 0 hours 0 weeks 0 months are included in Table IX, for the indepenfent variable, for reasons given comparison with the results derived by the other earlier. 21 The results of the evaluations, in procedures employed.

terms of the stresses to cause rupture in 1000; Figure 19 shows a graphical comparison of the 10,000; and 100,000 hours0 days 0 hours 0 weeks 0 months , are assembled in Table results for rupture in 10,nOO and 1nD,000 hours0 days 0 hours 0 weeks 0 months , VIII. For each of the temperatures evaluated, derived by regression of the temperature dependence rupture strengths are shown corresponding to a of the individual lot extrapolations and by the first order relation between log stress and loz two universalized Parameter procedures. Both of time-for-rupture. In addition, for a number of the parameter evaluation procedures give a more temperatures, rupture strengths are given, corre- conservative result than the procedure involving sponding to a second or third order equation. individual lot visual extrapolation, but the exhibiting a reduced variance. Even when the differences may be viewed as rclativel modest.

variance is reduced only marginally (expressed as a (In past evali ons of carbon steel.(l) and percentage), the 100,000 hour0 days 0 hours 0 weeks 0 months rupture strength, 2 1/4 Cr-l 'tolJ , the individual lot procedure developed in all instances by extending the had given a more conservative result than the regression line, may be reduced significantly, and universalized Larson-Iiller procedure.) Of the for larger reductions in variance, the reduction twn n;:rnmeter procedures, the Manson compromise in strength is even greater. Note particularly, parameter gave the more conservative result, as the 100,000 hour0 days 0 hours 0 weeks 0 months strength at 900*F corresponding expected, with divergence increasing towards to a third order equation. The differences in higher temperature. Since the parameter procedures 10,000 hour0 days 0 hours 0 weeks 0 months and 1000 hour0.0116 days 0.278 hours 0.00165 weeks 3.805e-4 months strengths corresponding provide an estimate of long time strength (e.g.

to different orders of the regression equation arc I1D,000 hours0 days 0 hours 0 weeks 0 months ) from shorter time tests at higher correspondingly less than those for 100,000 hours0 days 0 hours 0 weeks 0 months temperatures, it follows that they are inherently as might be expected, but may still be significant mnable to provide estimates of 100,000 hour0 days 0 hours 0 weeks 0 months rupture (as for example at 900'F). It must be concluded strengths over the entire range of temperatures for from a study of Table VIII that extension of the which there are test data (except as questionable isothermal scatter regression line to 100,On hours extrapolations of the "master" parameter curve is an especially hazardous procedure, and might be employed). Thus 100,000 hour0 days 0 hours 0 weeks 0 months rupture increasingly so as the order of the assumed equa- strengths can he derived in the present instance tion increases (see also reference 21). For this only to a maximum temnerature of l000'F, and even reason values for 100,000 hours0 days 0 hours 0 weeks 0 months rupture strength at this temneraturo, only a small portion of all by this procedure have not been included in the the data actually define the corresponding value comparisons of Table IX. However, 1000 hour0.0116 days 0.278 hours 0.00165 weeks 3.805e-4 months and of parameter. To be sure, many engineers would not 10,000 hour0 days 0 hours 0 weeks 0 months rupture strengths have been included choose to use C-%Io steel above lnOnfF because of in Table IX (for the linear case only), as of the excessive scaling to be expected.

possible comparative interest. In choosing amongst the three sets of 100,000 Two "universalized", time-temperature para- hour rupture strengths in Table TX or Fig. 19, one meter evaluations were made, in which all of the obtained bv direct and two by indirect extrapola-rupture data, except those for rupture tine less tion, the advantages bv the direct individual lot than S hours, were "parameterized", and the result- extrapolation Procedure of visual weighting of the ing scatter bands of stress versus parameter longer time results seems to be outweighed by the evaluated by the method of least squares. Uni- uncertainties associated with possible bilinearity versal values were assumed for the constants, and or curvilinearitv. It follows then that greater all data were assumed to come from a common sta- weight should be given to the results of the tistical population. The parameters that were indirect or parameter procedures, in spite of employed were firstly, the well-known Larson-liller inherent reservations aov such procedures, parameter, with an assumed value of 20 for the previously expressed.( Also, as previously constant: noted, either parameter procedure leads to a more conservative estimate. The choice between the two T(20 + log t) = Fl(s); parameter results is more difficult; several con-siderations seem appropriate. Firstly, the maximum and the more recent compromise parameter proposed difference between the two results is at 10000F, at which the Hranson compromise parameter result is 13%

by Mlanson:( 2 2) less than that by the Larson-Miller procedure; the 2

difference diminishes progressively to zero at 800*F.

log t

I log t _ 40 0 - F2FC 40.T460 Cs) . Interestingly, the positions of the parameter master 12

curves are least well defined by data at 1000lF, are fairly well defined and which are in the range and best defined at 800'F. Secondly, at 1000 of practical interest, ductility at rupture at hours, at which the direct result might be 10,000 hours0 days 0 hours 0 weeks 0 months is on the order of only 10 percent.

reasonably viewed as superior to the result of any This reduced ductility may reflect a trend to indirect procedure, since it can be derived by intergranular mode fracture at longer time, but no interpolation, the Larson-Miller result more information concerning mode of fracture was made closely approximates the direct result. Similarly, available by the contributors of data.

at 10,000 hours0 days 0 hours 0 weeks 0 months , the need for extrapolation was minimal, with a few values derivable by interpo-lation, and again the Larson-Miller result more Parts 2 and 3. Mn-Mo and Mn-Mo-Ni steels (Speci-closely approximates the direct result. The fore- fications A302 and A533) going train of reasoning suggests that the Larson-Miller procedure has given the most reasonable Since the yield strength ratio and tensile estimates of 100,000 hour0 days 0 hours 0 weeks 0 months rupture strength; and it strength ratio curves for material corresponding is this result that has been integrated into the to Specifications A302 and A533 were relatively summary chart of Fig. 1. Also as a matter of similar, and since the number of creep and rupture possible interest, isothermal log stress vs. log test results available is limited, data from these time-for-rupture curves have been computed from latter tests have been plotted together, though the Larson-Miller master curve, and these have differentiated by symbol. Figs. 21 a-b shows the been superimposed upon the scatter bands, Figs. interdependence between time-for-rupture and 14a, b and c. stress, Fig. 22, that for secondary creep rate and stress, and Fig. 23a, b that for rupture ductility and time-for-rupture (except that, in the interest Creep Strength of saving space, a few relatively short time rupture ductility results have not been plotted).

The secondary creep-rate data shown in the The rupture and creep data, extending to a scatter bands of Figs. 15a, b and c were visually maximum test duration of only about 1000 hours0.0116 days 0.278 hours 0.00165 weeks 3.805e-4 months , are interpolated or extrapolated (by not more than so limited in number that it has not seemed worth-about 1 log cycle) on an individual lot basis, to while to attempt the development of trend curves, determine creep strengths corresponding to 0.1 and particularly, wshen, based upon the behavior of 0.01 percent per 1000 hours0.0116 days 0.278 hours 0.00165 weeks 3.805e-4 months . Curvilinearity in the normalized and tempe fit or quenched and tempered relation between log stress and log secondary 2 1/4 Cr-l Mo steel, it is to be expected that creep rate in the region 0.1 to 0.01% per 1000 the elevated temperature creep and rupture hours was evident for many of the individual lots. strengths will vary with the level of strength at The results are assembled in Table VII and plotted room temperature, and hence depend upon the temper-in Figs. 20a and b. There is no effect of product ing temperature. However, individual lot interpo-form evident in the data, but the number of data lations or extrapolations were made, and these for other than bar is severely limited. The results are summarized in Tables VI and VII.

scatter plots of strength vs. temperature have Evidence of dependence upon room temperature been evaluated by the method of least squares and strength is apparent when the level of the 900°F the resulting lines of best fit (trend curves) scatter band is observed to be higher than that superimposed upon the plots and tabulated in for 850*F, Fig. 21a. The 900*F data represent lots Table X. Minimum position trend curves, derived having room temperature tensile strengths in the from the mean curves, are also given. range 130-140 ksi (which, of course, exceeds the The trend curves for variation of creep level permitted by specification), whereas the strength with temperature, Fig. 20, exhibit a AS33 data at 850@F represent a lot having a tensile complex character, requiring a third or higher strength at room temperature of only 87 ksi.

order equation, if the entire range of temperature Certainly, under the circumstances there is no is to be represented by a common curve. As a possibility of differentiating between lots con-consequence, the result becomes sensitive to the forming to the two different specifications.

distribution of data and at the mercy of the The limited ductility data extending only to procedure; thus the indicated maximum in the 0.1% 1000 hours0.0116 days 0.278 hours 0.00165 weeks 3.805e-4 months test duration arc essentially indepcn-per 1000 hours0.0116 days 0.278 hours 0.00165 weeks 3.805e-4 months creep strength at 850-F is of dent of rupture time and at a satisfactory level at questionable validity. It is of interest to note 850*F, but trend downward at 950*F.

that creep strength falls off rapidly above about 950F, whereas at lower temperatures there seems to be a levelling tendency. (The variations of Part 6. Mn-Mo Steel Plate, Ouenched and Tempered 1000 hour0.0116 days 0.278 hours 0.00165 weeks 3.805e-4 months and 10,000 hour0 days 0 hours 0 weeks 0 months rupture strengths, Fig. (A514, Type C -

17a, exhibited a somewhat similar tendency.) This relative insensitivity of strength to temperature The time-for-rupture, secondary creep rate, at the lower temperatures may be related to the and rupture ductility for A514, Type C are plotted tendency to secondary hardening conferred by moly- in Figs. 24, 25, and 26 a-b, respectively, but in bdenum. The trend curves have been extended to view of the limited character of the data, and an the upper limit of temperature for which data were expectation that the results depend sensitively available, i.e. 1200F, even though as indicated upon level of strength at room temperature, no earlier, extensive scaling may be expected above effort has been made to develop trend curves. How-about 1000'F. ever, interpolations and extrapolations of the data for the individual lots have been made, and the results are included in Tables VI and VII.

Rupture Ductility The elongation and reduction of area at rupture of C-Mo steel tend, on the whole, to be relatively low at longer rupture times, Figs.

16 a-g. Thus, at 850 and 9500F, where the trends 13

Acknowledgment 19. Elevated-Temperature Properties of Weld-Deposited Metal and Weldnents, AS1M Special The evaluations of this report were developed Technical Pub. No. 226 (1958).

for The Metal Properties Council under the guidance of a subcommittee of which Dr. M. Semchyshen is 20. G. V. Smith: Strength of 2 1/4 Cr-l Mo Steel Chairman. Appreciation is expressed to the chair- at Elevated Temperatures; Symposium on 2 1/4 man and members of the subcommittee for helpful Chrome-l Molybdenum Steel in Pressure Vessels suggestions. and Piping, Sept. 1970, Denver, Colorado.

American Society Mechanical Engineers.

References

21. G. V. Smith: Evaluation of Elevated-Tempera-

1. Report on the Elevated-Temperature Properties ture Strength Data: Jnl. of Materials, 4, of Chromium-Molybdenum Steels; AS7M Data No. 4, Dec. 1969, pp. 878-908.

Series Pub. DS 6 (formerly STP No. 151); Oct.

1953. 22. S. S. Manson: Time-Temperature Parameters --

A Re-Evaluation and Some New Approaches;

2. Supplemental Report on the Elevated-Temperature published in "Time-Temperature Parameters for Properties of Chromium-Molybdenum Steels, ASTM Creep-Rupture Analysis"; Amer. Society for DS 6S1; June 1966. Metals, Pub. No. D8-100 (1968).

3. Babcock and Wilcox Co.

4. U.S. Steel Corp.

5. General Electric Co.

6. Compilation of Available High-Temperature Creep Characteristics of Metals and Alloys; Joint ASME-ASTM Committee on Effect of Temperature on the Properties of Metals; March 1938; published by ASTM and ASME, Phil. and New York.

7. A. 0. Schaefer: Review of Available Data on Steel in Heavy Sections for Pressure Vessel Applications; unpublished work, Pressure Vessel Research Committee, Welding Research Council, 1964; also 1968.

8. U.S. Steel Corp.

9. Lukens Steel Co.

10. Babcock and Wilcox Co.

11. Combustion Engineering, Inc.

12. Lehigh U1niv., Department of Metallurgy and Materials Science: The Creep Rupture Proper-ties of Pressure Vessel Steels - Data Summary, Sept. 1968.

13. S. S. Strunck, A. W. Pense, and R. D. Stout:

The Properties of Spray-Ouenched Thick-Section Steels; Welding Research Council Bulletin, No.

120, Feb. 1967.

14. Texas Electric Steel Casting Co.

15. Jones and Laughlin Steel Corp.

16. An Evaluation of the Yield, Tensile, Creep and Rupture Strengths of Wrought 304, 316, 321 and 347 Stainless Steels at Elevated Tempera-tures; ASTM Data Series DS SS2 (1969).

17. An Evaluation of the Elevated Temperature Tensile and Creep-Rupture Properties of Wrought Carbon Steel; ASTM Data Series DS lSt (1970).

18. An Evaluation of the Elevated Temperature Tensile and Creep Rupture Properties of 2 1/4 Cr-l Mo Steel; ASTP Data Series DS 6S2 (1971).

14

Table I Identification of Steels Code Snec. Deoxid. Heat (1) Product Grain(2) Ref.

No. No. Pract. Treatment Form-Sizc-Si- 2e R.

Cn.4 No.

Part 1 - Carbon-lolybdenum Ste els (Specs. A204, A209, A335, A369, All32, A217)

.. ^ ;

r rs 1-1 01-A& ti lb.) AISSO Bar, 1" 6-8^ 1 1-2 A182 IIR Plate IRXInXl.S" I 2 1-3 A206 N1650, T1250 Rod, I" 3 1-4 Si#A1 (.4 lb.) N16so Forged 23 2 4 1-5 Si+AI (1.6 lb.) N1650 6 3N511 1I 1-6 Si N1650 L- I 6 1-7 Al (.S lb.) N1700, T1310 Cast 1 7 1-8 CaMnSi (2 lbs.) N1700, T11S0 Cast I 8 1-9 Si.AI N1560, T1380 Wrought I 9

.. 89 1-10 N1650, T1300 I 10

.. 8-9 1-11 N1650, T1300 I 11 1-12 Si+Al N1650 8-9 12 1-13 Al (.5 lb.) A1850 I 13 1-14 A1740, T1330, T1200 Cast - I 14 1-15 Annealed Tube, _ I 16 14 O.D.ml 1/4"w.

1-16 Al (2 lb.) N1750, T1200 Cast - 1 is 1-17 CaMnSi+A1 (3 lbs.) 1N1700, T1300 Cast 8 1 20 1-18 Ca&tnSi*Al (2 Ibs.) 1N1700, T1300 Cast 7 1 21 1-19 CaSi+Al (.5 lb.) MM170 T1300 Cast 2-3 I 22 1-20 Al (2 tbs.) N1700, T1300 Cast 6-7 23 1-21 Carbotram (4 lbs.) N1700, T1200 Cast 7-8 1 24 1-22 Al (2 lbs.) N1700, T1300 Cast 6-7 1 25 I

1-23 Al (.5 lb.) N1700, T1300 Cast 3-4 I 26 1-24 None N1700, T1300 Cast 3-6 27 I

1-2S Al (.5 lb.) N1700, T1300 Cast 4 28 1

1-26 CaInSi-Al (3 lb.) N1700, T1300 Cast 7 1 29 1-27 CaSi.A1 (2.5 lb.) N1700, T1200 Cast 2-3 30 7-8 I 1-28 CaSi*AI (2.4 lb.) N1700, T1200 Cast 31 1-29a Al (I lb.) N1800 Bar 5 2 7-6a 1.

1-29b .. N1800, T1250 1-30a N19S0, T1250 o, 9,.

5 2.. 7-6b

.9 1-30b N2000, T1250 1-31l MR, T1300P Rar 3/4" 2,3 7-18 1-31b N1750, T1200 2,3 7-18 1-Slc N1650, T1200 2,3 7-18 1-32 T1400 (4 brs) Tube 21 O.D0.x.3" - 2.3 7-19 1-33 A209 Si+AI (1.6 lb.) AIS75 - 1200, AC 2.3 7-20 1-34 A209 Si of ..

Bar 3/4" - 2,3 7-21 1-35 A209 CD, T1300 Tube, 2 0.I3.x.344'" - 3 1-36 A1600 Tube - 3 (1) A-Annealed; N-Normalized; HR-Hot Rolled; Q-Quenched; T-Tempered or Stress Relieved; CD-Cold Drawn (2) Actual grain size except when identified as H for AicQuaid-Fin 15

Table I - page 2 Code Spec. Deoxid. HCat Product Grain Rer.

No. No. Pract. Treatment Form-SIze Sire Ref. Cnde No.

1-37 1S25 - 1200, AC Wrought 1-38' - Si N16SO Bar, I" 1-3 1-39* - Si+.AI (1.6 lb) N1650 Bar. I" 6-8 1-40 A204-RC - Plate, 5 3/16" 1-41' - Si., ;I (.4 lb) N16S0 Bar, 1" 1-3 1-42 T1270 Cast 1-43 N1650, T125I Bar, 1" 6-8ba la 1-44a N1650, QISSO, Tl 100 Bar, I 1/8" 714 lb 1-44b N1650, TIOS0 Ic 1-44c N1650, 01550, Tl110o 4 .S}I1 Id 1-45 N1650, T122S Bar, 1", 2 1-46"* - S1-4 LI A1550 Bar, I" 3a 1-47 - Si-ALI AISSO Bar, 1" 3b 1-48 _ Si-I Li A1550 Bar, I" 3c 1-49 _ Si- Ai AISSO Bar, I" 3d I-so Nl6SO, T1200 Tube 44 1-51 S. ,, 4h 1-52 4c I-S3 4d 1-54 4e I-55 N16S.

4f 1-56 led N41650, T1200 (I wk.) Bar, 1"1 Sa 1-57 -Kill led N416S0, T1200 (1 wk..) Bar, I" 5b I-58 ,ed N41650, 71400 (I wk.) Bar, I" Sc I-59 - Si-ALI Bar 6a 1-60 - KillAI Si-A 14165n, T1200 (5 hr) Bar 6b 1-61 - Si-ALi N1650, T1200 (1tLAhr) Bar 6c 1-62 - Si-ALI N1650, T1300 (S hr) Bar 6d 1-63 _Si-A Li N1650. 7130D (11 Bar 6e 1-64 - Si-AAI N1650. T1400 (S hr) Bar 6f 1-65 -Si-A LI N1650, T1400 (I1,Rhr)Bar 6R 1-66 N1650, Tl11O Plate, 1 114's 7 1-67 Normalized Bar, 1' 8a 1-68 _Si T7200 Plate, 1 1/2"1 2 4M Sb 1-69 _Si T120n Plate, 1 1/2"1 8c 1-70 T120D Weld metal 9b 1-71 T1200 Weld metal 9c 1-72 As Received Pipe 9a 1-73 A2l4-B - N1675 Plate, S 114' 4 -5b1 1-74 A204-B-C - N1700 Plate, 3'*

1-75 A204-B-C - N170n, T1100 Plate, 3" 1-76 A204-B - N1700, TlO Plate, 4 3/4I

Identical with Code 1-6

Identical with Code 1-5

Identical with Code 1-4

- Identical with Code 1-1 16

1 Table I - pagc 3 Code Spec. Deoxid. llent Product Grain Ref.

No. No. Pract. Treatment Form-Size Si2e Ref. Code No.

1-77 A204-B - S1700, TO00 Plate, 4 3/4" 9 1-78 A204-A-R - N1675 Plate, 6" 9 1-79 A204-B - N1650, TllSO Plate, 3 1/4" 9 I. 99 1-80 . - N16S0, T1200 9 1-51 N165n, T1250 9 1-82 T1200 lfcld netal 19 1 I, It 1-83 T1200 19 2 1-S4 T1200 19 3 I-8S T1275 1s 4

,. I.

1 T1112 1s S 1-87* T1202 19 S T1200 I... S 1-88 1-89 T1275 5 Part 2 - Mlangancse-molvbdenum and manganese-nolvbdenum-nickel steels (Spec. A302) 2-1 A302B Norm. & Temp. Plate, 6" 2 7-14ab,c 2-2 ., Plate, 8 1/2" 2 7-lSab 2-3 Plate, 3" 2 7-16a,b 2-4 A302B FG;A1 X1775F(l), T1125 Plate, 3 1/8" 7 2-5 A302B FG Spray Quenched 6 Plate, 1S" 7 Tempered 2-6 A302B N1650, T1200 Plate, 4" a 2-7 A302C N1650, T1225, T1125 Plate, 2 7/8" a 2-8 A302C X1650, T127S Plate, 5 9/16" 9 2-9 A302B N1600, T1100 Plate, 6" 9 2-10 A302C N1650 Plate, 2" q 2-11 .. N1650, T1290 Plate, 2" 9 2-12 9. N1650, T1300 Plate, 2" 9 2-13 ..

N1650, 71275 Plate 9 2-14 A302B FG Q157S, T1225, TllSO Plate, 10 1/2" 7 2-15 ., , II IS .. .. 7 2-16 ., .. .. *. .. I.

7 Part 3 - Mlanganese-nolybdenum and iianganese-molybdenum-nickel steels (Spec. AS33) 3-1 A533B,2,3 - 01700, 01625, T117S Plate, 6 3/1"8 10 3-2 *- ,, ,, 55 9, 5B -

10 5-3 10 3-4 A533B,2 FG QlS7S, T12225, T1150 Plate, 7 7/8" - I1I 3-S II II Is 1. i Plate - 11 3-6 A5330,2,3 FG Q16SO, T1240 Plate, 1 3/4" - 9 3-7 It Ouenched G Temp. Plate, 1 3/4" - 9 3-8 to .' Q1675, T1225, T1075 Plate, 1 3/4" - 9 3-9 A5330,2,3 - 01675, T1225 Plate, 1 3/4" - 9 (1) Spray quench to 600F nata Sdentical with Codes 1-70 and 1-71 17

- I

Table I - nage 4 Code Spec. flcoxid. Heat Product Grain Ref.

No. No. Pract. Treatment Form-Size Size Ref. Code No.

3-10 A533A,1,2 - 01600, TllOO, T1150 Plate, 6" 9 3.11 A53381 - Q16S0, T1240, TllS1 Plate, S 9/16" 9 3-12 A533B,1,2 - 01650, T1240, TlISO Plate 9 3-13 AS33B,1 FC-Al 01600, T1225. T11SO Plate, 9 5/8" 7 3-14 * . ..4. .... ..*. 7 3-1S " . 7 3-16 A533B,3 - Q1650, T1200 Plate, 2 3/8" 3.171 AS33A 016S0. T1150 Plate, 3/4"-1" 12 3-l8al A533B -

4, ..

12 3-18b AS33B,1 - Q1650, T1200, M1100 plate( 2 ) 13 3-18c " .. 49 of ." (2): 12 13 Part 4 - Carbon-Manganese-Polybdenum steel forgings (Spec. A372, Class IV) 4-1 A372,1V - Q16SO, 7900 Pipe 24"0.D.xl.08"v, - 8 Part S - Carbon-Manganese-Hoblybdenum steel castings (Spec. A487, Class 2)

S-la A487,2N - N1600, TllOO Casting 14

.. 14 S-lb A487,20 - 01600, T1225 _

.. 14 5-2a A487,2N - N1600, T71Oo -

S-2b A487,2Q - 01600, T1225 - 14 S-3a A487,2N - N160M , TllnO 14 5-3b A487,2Q - 01600, T1225 - 14 Part 6 - Ouenched and Tempered carbon-manganese-molvbdenum steel plate (Spec. AS14, Tyve C) 6-1 A514C - Ql6S0, T1125 Plate, 1/2" - is 6-2 AS14C - ,. .. . is (1) Yield and tensile strengths at room tenmerature exceed limits of current specification.

[2) Heat treated to simulate thickness indicated.

18

1-w-Table II Chemical Composition of Steels r2.At - r. M P S Si C-r K4 Mo Cu Al Part 1. - a M lbde . eel - -- .t Part I - Carbon-Molybdenumn steels 1 -3S .18 .52 .009 .022 .28 .12I 08 54 .08 1-36 .19 .50 .no8 .018 .26 .16 .12 .59 .10 1-37 .14 .54 .01S .015 .22 .06 .08 .52 .08 1-38 .16 .85 .020 .019 .24 .028 .015 .51 .03 .010 .005 1-39 .13 .52 .012 .021 .16 .058 .042 .S2 .044 .005 1-40 .23 .85 .26 .54 1-41 .22 .51 .012 .020 .17 .046 .048 .50 .09 .004 .005 ..001 1-42 .18 .58 .008 .008 .36 .12 .10 .53 1-43 .18 .68 .030 .019 .20 .10 .12 .51 .26 1-44 .16 .78 .032 .015 .25 .03 .04 .48 1-45 .17 .51 .32 - - .46 1-46 .13 .49 .011 .010 .25 - - .S2 1-47 .16 .49 .015 .018 .30 - - .49 1-48 .16 .47 .016 . 01S .23 _ _ .42 1-49 .11 .19 .010 .012 1.35 _ _ .50 I-so .17 .52 .16 .05 - .S4 1-51 .21 .48 .31 .05 .11 .53 1-52 .15 .47 .13 .02 - .55 1-53 .15 .41 .19 .07 .14 .S8 1-S4 .13 .50 .13 .06 - .52 1-55 .16 .45 .14 .06 - .58 1-56 .16 - _ _ .50 1-57 .10 - _ _ .50 I -S8 .10 - _ _ .50 1-59 .11 .47 .010 .014 .17 _ _ .S4 1-60 .11 .47 .010 .014 .17 _ _ .54 1-61 .11 .47 .010 .014 .17 _ _ .54 1-62 .11 .47 .010 .014 .17 _ _ .54 1-63 .11 .47 .010 .014 .17 .54 1-64 .11 .47 .010 .014 .17 .54 1-65 .11 .47 .0n1 .014 .17 .54 1-66 .17 .50 .21 - _ .54 1-67 .16 .66 .016 .012 _ _ .54 1-68 .17 .60 .01 .02 .22 _ _ .4S 1-69 .17 .60 .01 .02 .22 .45 1-70 .06 .48 .03 _ - *.56 1-71 .06 .48 .03 _ _ ,56 1-72 .12 - _ _ .50 1-73 .2S .75 .007 .024 .25 .05 .09 .45 .12 .010 1-74 .25 .75 .007 .024 .25 .05 .0s .4S .12 .010 1-75 .25 .75 .007 .024 .25 .05 .09 .45 .12 .010 1-76 .25 .75 .007 .024 .25 .05 .09 .45 .12 .010 19

- I

Table T1 - nate 2 Code No. C Mn P S Si Cr Ni -lo Ci1 Al N 1-77 .2S .75 .007 .024 .25 .05 .09 .45 .12 .010 1-78 .21 .62 .015 .024 .23 .08 .21 .47 .24 .012 1-79 .20 .70 .008 .012 .17 .10 .16 .47 .21 .02 1-80 .20 .70 .008 .012 .17 .10 .16 .47 .21 .02 1-81 .20 .70 .008 .012 .17 .10 .16 .47 .21 .02 1-82 .14 .44 .12 .59 1-83 .13 .66 .18 .46 1-84 .12 .52 .23 .48 1- S .08 2.32 .02n .021 .71 .48 I 1-86 .06 .. 8 .03 .56 I 1-87 .0f6 .48 .03 .56 1-88 .08 .26 .10 .55 i 1-89 .20 .65 .004 .003 .51 .52 iI I

Part 2 - %langanese-molybdenum and manganese-molybdenum-nickel steels (Suec. A302) 2-1 .25 1.36 .018 .036 .23 .49 2-2 .25 1.36 .018 .036 .23 .49 2-3 .25 1.36 .018 .n36 .23 .49 2-4 .20 1.27 .020 .028 .21 .48 2-S .22 1.45 .01.1 .015 .18 .29 .35 .s0 .20 .037 2-6 .18 1.26 .015 .012 .29 .09 .56 2-7 .19 1.37 - - .20 .62 .55 2-8 .21 1.27 .011 .016 .25 .10 .56 .60 .12 .014 2-9 .20 1.28 .009 .018 .24 .07 .ns .47 .11 .031 2-10 .22 1.30 .010 .017 .22 .07 .62 .52 .13 .029 2-11 .23 1.30 .010 .)15 .22 .06 .52 .42 .12 .039 2-12 .23 1.30 .010 .015 .22 .06 .5: .42 .12 .039 2-13 .24 1.32 .009 .016 .15 .11 .56 .55 .11 .019 2-14 .20 1.46 .013 .010 .22 .48 .12 2-1S .20 1.37 .010 .021 .19 .47 .12 2-16 .2n 1.42 .010 .nl4 .26 .47 .10 Part 3 - klanganese-molybdenum and-manganese-molybdenum-nickel steels (Spec. A533) 3-1 .26 1.35 .n12 .021 .26 .15 .61 .46 .28 3-2 .25 1.37 .n12 .024 .33 .13 .63 .46 .23 3-3 .25 1.37 .nl3 .024 .32 .13 .65 .46 .23 3-4 .22 1.56 .no5 .013 .38 .07 .57 .51 .19 .03 3-5 .22 1.31 .008 .025 .22 .15 .S1 .49 .25 .048 3-6 .19 1.48 .012 .022 .21 .29 .35 .50 .20 .037 3-7 .20 1.37 .009 .017 .19 .08 .25 .48 .15 .041 3-8 .18 1.18 .010 .016 .22 .31 .48 3-9 .22 1.30 .011 .025 .24 .14 .25 .4S .23 .03 3-10 .20 1.28 .009 .018 .24 .07 .n8 .47 .11 .031 3-11 .21 1.27 .011 .016 .25 .10 .56 .60 .12 .014 3-12 .24 1.32 .009 .015 .15 .08 .56 .55 .11 .019 20

Table II - page 3 Code No. C Mn P S Si Cr Ni Mo Cu Al N 3-13 .22 1.29 .011 .018 .25 .16 .57 .46 .16 .062 3-14 .22 1.30 .014 .020 .22 .10 .46 .50 .14 3-15 .20 1.28 .010 .019 .25 .15 .58 .46 .25 3-16 .25 1.34 .012 .023 .23 .10 .50 .53 .054 3-17 .19 1.28 .013 .010 .22 - - .45 3-18 .20 1.28 .019 .030 .21 .15 .53 .52 .27 .031 Part 4 - Carbon-manzancse-molybdenum steel (Snec. A372, Class IV) 4-1 .44 1.S6 .013 .A21 .20 .21 Part 5 - Carbon-nanganese-molyhdenum steel castings (Spec. A487, Class 2) 5-la,b .30 1.10 .015 .015 .43 .26 .18 .18 5-2a,b .30 .87 .022 .018 .41 .21 .18 .19 5-3a,b .28 1.03 .01(. .020 .38 .21 .16 .21 Part 6 - Quenched and temnered carbnn-maneanesc-molvydenum steel palte (A514C)

Ti V B 6-1 .18 1.17 .012 .023 .22 .03 .02 .24 .037 N.D. .ons .°02 6-2 .18 1.30 .012 .024 .25 .02 .02 .25 .04 N.D. .nns .no2 21

-I---

Table III - page 3 T 1000 psi Per Cent Test I Code No. Temp. _F Yield Stren. Tensile Stren. Elone. Rcd. Area 1-73 700 42.8 79.5 24. 48.

cont. 750 38.5 69.3 21. 52.

800 40.3 69.5 20. 52.

850 41.6 67.7 21. 49.

1-74 70 55.6 83.3 24. 51. )

so5 57.8 97.8 21. 35.

675 48.1 77.3 21. 54.

850 51.5 76.7 20. 56.

1000 46.0 62.7 20. 63.

1-75 70 51.7 78.8 28. 52.

500 40.6 78.6 17. 41.

675 43.5 79.6 23. 49.

8S0 47.6 71.2 25. 51.

1000 40.8 60.5 22. 54.

1-76 70 47.2 74.4 30. 51.

500 42.8 79.7 24. 40.

675 39.8 67.3 21. 49.

850 41.9 70.4 20. 45.

1000 36.3 58.2 23. 54.

1-77 70 46.9 73.6 32. 52.

So0 44.5 7S.5 21. 38.

675 38.5 65.5 23. 49.

8s5 36.8 66.7 23. 52.

1000 42.2 57.0 24. 55.

1-78 7n 46.6 73.2 31. 56.

500 36.0 64.6 26. 52.

675 33.5 65.3 25. 42.

850 30.4 60.5 26. 61.

1000 29.9 50.6 30. 68.

75 55.6 80.6 29. 61.

750 31.4 75.8 28. 64.

1-80 75 5S.1 78.6 .31 . 63.

750 31.1 75.4 28. 63.

1-81 75 52.3 76.7 29. 66.

750 31.4 69.0 31. 64.

Part 2 - Manganese-molvbdenum and manganese-molybdenum-nickel steels (Spec. A302) 2-4 75 77.0 98.5 23. 61.

100 73.7 99.2 26. 65.

150 70.2 96.4 21. 60.

200 69.5 94.5 22. 66.

300 68.1 91.0 22. 65.

400 65.9 89.9 20. 56.

4 Test samples from 1/2 T position.

24 Ii

Table T11 - paRe 4 1000 psi Per Cent Test Code No. Temp. eF Yield Stren. Tensile Stren. Elong. Red. Area 2-4 5SO 67.5 92.0 17.3 52.

I' cont. 700 65.9 92.0 24. 54.

I 2-S 7S 85.6 104.90) 20. 50.

4 1000 1200 1400 64.5 42.5 13.6 75.4 46.6 16.6 23.

27.

58.

66.

55.

51.

1600 7.2 10.3 51. 44.

1800 4.0 5.4 59. 57.

2-6 80 76.4 96.1 25. 69.

200 70.1 90.0 23. 68.

400 66.9 SS.8 23. 67.

600 66. S 94.0 27. 65.

80s 61.0 79.9 26. 73.

1oon 53.4 61.7 29. 83.

1200 32.0 38.7 31. 91.

1400 12.0 17.4 79. 92.

1600 6.6 13.4 50. 92.

1900 3.0 5.6 91.

2-7 80 71.6 8s8 1 27.

so0 55.0 82.4 21. 67.

700 58.5 84.5 28. 74.

900 49.6 68.6 21. 76.

son 26. 50.

2-8 80 54.2 100.8 750 53.2 90.5 33. 64.

850 S0.1 74.9 34. 68.

950 47.9 64.4 34. 72.

2-9 80 69.8 94.9 22. 56.

750 S8.9 84.8 23. 62.

850 56.3 79.8 22. 64.

950 53.4 71.7 25. 68.

2-10(2) 75 94.4 116.3(2) 22. 60.

400 90.7 111.0 14. *46.

600 84.4 111.9 26. 69.

800 82.2 103.5 22. 71.

1000 69.6 78.4 20. 69.

2-11 75 73.3 9S .0 26. 69.

400 67.2 93.4 24. 64.

600 70.3 96.5 25. 61.

son 67.3 81.D 28. 72.

1000 S4.4 63.3 26. 78.

2-12 75 64.2 90.5 28. 69.

400 60.1 84.S 23. 68.

600 61.1 9S.8 32. 64.

I (1) Slightlv higher than permitted by spec. A302.

(2) Exceeds limit of spec. A302C.

I1 25 iI ,:S I--

V. .

, I - 11

Table III - page S

-F 1000 psi Per Cent Test Code No. Temp. *F Yield Stren. Tensile Stren. Elong. Red. Area 2-12 800 56.2 70.7 22. 72.

cont. 1000 50.6 57.1 22. 80.

2-13 80 60.7 87.0 30. 65.

750 47.3 79.1 30. 63.

850 44.3 69.2 31. 69.

950 43.7 61.2 31. 72.

2-14 75 61.7 81.4 27. 72.

200 58.6 76.8 25. 67.

400 53.2 74..2 22. 69.

600 53.1 75.2 27. 65.

2-15 75 58.7 81.0 29. 70.

200 56.5 76.4 25. 68.

400 54.1 76.6 22. 69.

600 51.0 74.1 27. 65.

2-16 75 62.8 83.5 29. 69.

200 59.5 79.1 27. 67.

400 54.6 75.7 26. 66.

600 52.4 80.0 29. 62.

Part 3 - M1anganese-molybdenum and manpanese-molybdenum-nickel steels (SDec. A533) 3-1 75 87.2 108.1 24. 67.

-,8 200 84.5 101.9 23. 66.

400 78.5 97.5 22. 67.

600 76.5 100.9 19.S 51.

800 69.2 87.3 21. 63.

3-2 75 87.0 107.5 25. 66.

200 82.0 101.0 24. 66.

400 78.6 98.1 22. 63.

600 78.0 101.3 21. 55.

800 71.2 86.1 21. 69.

3-3 75 91.5 110.4 25. 67.

200 86.1 102.9 24. 67.

400 81.2 99.0 21. 63.

600 81.8 103.5 21. 54.

800 72.8 89.1 22. 66.

3-4 75 72.6 94.5 24. 65.

300 65.4 85.5 23. 65.

500 62.9 85.9 21. 64.

700 60.9 84.6 25. 67.

900 54.6 67.1 22. 74.

3-5 7S 69.4 91.3 25. 66.

550 61.8 88.0 23. 60.

650 60.1 84.9 25. 61.

3-6 75 88.3 105 .7 26.

200 91.8 107.7 23. 71.

400 88.4 105.3 22. 65.

26

--- I.

Table III - page 6 1000 psi Per Cent Test rneAr Mi Tm.- Or VYieA ctzrepn Tensqile Rtren. Flono - Red. Area uu:.-u *x1as .-- *w sw* -.

3-6 600 84.8 111.1 20. 62.

cont. 800 78.1 97.7 22. 69.

1000 66.7 74.4 24. 79.

3-7 75 82.8 100.5 25.

200 81.0 98.3 23. 70.

400 76.0 96.5 19. 70.

600 74.7 106.5 24. 56.

800 68.5 91.5 22. 64.

1000 61.6 72.9 22. 79.

3-8 75 89.1 107.1 23. 64. ,

700 76.4 100.6 18.7 60.

800 71.9 92.2 19.0 63.

900 68.5 83.2 18.5 67.

1000 62.9 74.5 20. 72.

3-9 75 86.7 107.6 21. 64.

300 77.3 97.0 22. 65.

500 73.1 102.5 28. 63.

700 71.7 100.0 27. 70.

800 69.1 90.4 25. 66.

3-10 80 62.1 83.1 30. (1) 69.

750 57.8 82.5 71.

850 54.5 71.4 0 .(1) 75.

950 50.7 64.1 29. (1) 80.

3-11) 80 64.5 87.0 29. 67.

73.1 29. 66.

SBBi 750 54.2 67.0 33. 71.

850 50.9 950 47.9 58.9 32. 79.

3-12 75 70.8 95.1- 27. 63.

750 59.5 79.7 31. 67.

850 56.0 71.4 43. 72.

200 53.0 62.7 31. 76.

3-1 675 65.3 87.3 26. 67.

200 67.7 89.0 22. 66.

400 56.7 78.3 22. 66.

58.1 85.2 25. 66.

3-14 75 64.6 87.1 27. 70.

600 200 63.4 84.1 25. 69.

400 52.2 7S.5 23. 68.

600 54.1 80.1 23. 66.

3-16 75 67.2 88.8 25. 67.

200 64.8 84.6 24. 68.

400 58.8 81.0 22. 66.

600 57.3 83.2 22. 64.-

3-16 so 100.0 117.0 21. 66.

200 99.4 114.0 20. 65.

400 91.4 111.0 20. 62.

(1) I" gage length.

27

WNP Table III - page 7 1000 psi Per Cent Test Code No. Temp. OF Yield Stren. Tensile Stren. Elong. Red. Area 3-16 600 90.0 117.0 25. 66.

cont. 800 79.0 98.8 23. 72.

1000 69.8 79.8 22. 82.

3-17 75 121.3 138.0 22. 70.

3-184 75 121.5 130.5 20. 63.

3 --&Bs-

75 70.0 88.7 28. 69.

200 64.8 82.5 25. 69.

600 58.1 83.6 l I .al. 900 51.7 64.6 21.

22.

60.

74.

1100 44.2 45.5 26. 81.

3-18c 75 64.1 84.6 29. 69.

I 200 56.7 77.5 28. 68.

600 48.5 78.9 23. 58.

900 41.4 58.4 28. 75.

1100 36.1 39.8 38. 82.

Part 4 - Carbon-manganese-molybdenum steel forgings (Spec. A372, Class IV) 4-1 75 95.9 125.9 19. 54.

200 91.8 120.3 18.5 53.

300 86.9 118.1 17.0 54.

400 85.5 119.3 18.0 55.

500 87.1 123.7 22.0 60.

600 85.9 127.4 29.0 71.

700 81.4 117.5 28.0 74.

800 77.6 101.4 24.0 78.

900 72.5 88.6 27.0 84.

1000 63.3 74.4 25.5 85.

Part 5 - Carbon-manganese-molybdenum steel castings (Spec. A487, Class 2) 5-la 75 65.0 98.5 23. 41.

300 62.5 90.0 28. 46.

500 56.5 89.5 20. 39.

700 55.0 91.2 22. 39.

5-lb 75 88.0 107.0 22. 48.

300 85.0 101.5 18.0 48.

500 78.0 101.0 18.0 44.

700 78.0 103.7 24.0 46.

5-2a 75 59.0 88,5 24.0 43.

300 50.0 80. 0 26.0 51.

son 41.0 80.0 23.0 41.

700 40.0 81.0 24.0 42.

5-2b 75 77.0 98.0 24.0 57.

300 71.0 92.0 22.0 56.

500 65.0 89.0 21.0 53.

700 65.0 97.0 27.0 59.

28

.r -- t I

Table TV - page 6 Test Stress, Duration- Min. creep At Runture Code No. Temp. *F ksi Hours Rate-%A/r. % Elong. % Red. Area 2-72 932 20.0 1030. c .0000194 932 20.0 1030. c .0000248 1-88 900 58.0 1.5 11.6 62.

52.0 33. 5.7 18.

46.0 136. 5.9 17.

42.0 871. 3.0 10.

38.0 343. 2.0 13.

38.0 1035. 2.3 11.

1-89 1150 7.0 648. 15.6 44.

1050 17.0 331. 14.5 33.

1000 27.0 132. 25.4 48.

900 27.0 10,046. 10.2 24.

800 38.0 38,307. 13.3 58.

Part 2 - Manganese-molyhdenum and manganese-molybdenum-nickcl steel (spec. A302) 2-8 830 S7-5 219. .01319 .^ 31. 70.

850 65.0 0.5 8.4 - 29. 64.

hstnc 62.5 1.9 6.2 29. 65

57. 5 48.6 .253 . 34. 66.

55.0 445. .0268 ::; 30. 64.

52.5 2667. c .00426 2 50.0 1438. c .002171.- -

45.0 SOOO. c .0004 TA.^-

900 32.5 171. c .00125 :P -

9S0 60.0 during loading - 30. 70.

47.5 10.3 1.76 42. 74.

45.0 26.9 1.12 41. 73.

40.0 101. .077 8 37. 57.

35.0 321. .0218 '-.7 17.8 35.

30.0 1281. .00651 /5y.- 19.5 25.

950 53.0 183. - 11.4 13.8 45.0 727. - 7.5 10.0 42.5 476. - 8.6 11.6 Part 3 - Manganese-molybdenum and manganese-molybdenum-nickel steels (spec. A533) 3-10 850 60.0 61. .249 31. 68.

ti 5A0 I 950

57. S 50.0 95.

19.4

.154

.870 34.

39.

62.

77.

47.5 630. .01165 22. 39.

40.0 288. .0337 35. 54.

37.5 1198. 22. 25.

35.0 1017. .0033 16.5 25.

32.5 1938. c .00IS9 35

-!I Table TV - page 7 I

Test Stress, Duration- Miin. creep At Rupture Code No. Tenp. *F ksi Hours Rate-%Ihr. % Elong. % Red. Area 3-11 850 60.0 2.25 3.21 30. 71.

55.0 18.6 .371 26. 69.

jjf7S36I 52.5 21. .423 29. 71.

50.0 253. .0326 31. 70.

47.5 160. .0358 30. 64.

4S.0 907. .0034 31. 71.

40.0 1000. c .000679 9S0 50.0 0.7 9.14 36. 74.

45.0 7.0 .864 37. 78.

40.0 S4. .1095 35. 76.

35.0 188. .0399 32. 61.

30.0 657. .0161 24. 38.

3-17 900 80.0 110. 33. 44.

70. 0 OARS1 60.0 265.

905. 7.

1000 57.0 is. 27.

50.0 0.6 16. 66.

45.0 10.7 21. 42.

40.0 39.0 24. 42.

33.5 53. 8.

30.0 108. 10.

25.0 271. 33. 21.

20.0 740. 29.

1100 20.0 23.3 41.

15.0 103.

3-18a 900 80.0 14. 72.

70.0 186. 15. 21.

A53 B 60.0 50.0 283.

535. 41. 11.

1000 57.0 12.1 21. 3S.

50.0 19.2 19. 24.

45.0 29.7 20. 22.

40.0 27.5 38. 53.

33.5 96. 23. 27.

30.0 171. 14. 13.

25.0 320. 27. 33.

20.0 712. 29.

1100 15.0 106. 70. 60.

Part 6 - Ouenched and tempered carbon-mangancse-molybdenum steel plate (AS14 C) 6-1 700 100.0 2.5 11.1 !

95. 0 650.

92.0 717. 10.2 90.0 526.

800 90.0 5.3 85.0 23.5 13.4 36

7- --- --

Table Va Ratio of Elevated Temperature Yield Strength to Room Temperature Yield Strength Combined Temp. °F C-1/2 'Ho A302 A533 A302-AS33 A372.IV A487.2N A487.20 ASl4C 75 1 no0 1.000 1.000 1.000 1.00 1.0 1.0 1.0 100 .988 .983 .990 .988 .990 .986 .990 .980 200 .940 .938 .951 .950 .947 .945 .959 .942 3no .902 .914 .923 .922 .917 .904 .925 .918 400 .870 .903 .903 .903 .902 .860 .890 .901 500 .845 .898 .887 .890 .896 .820 .860 .881 600 .818 .893 .870 .877 .886 .775 .859 .857 700 .788 .880 .848 .861 .862 .738 .858 .829 800 .749 .853 .817 .833 .815 .788 900 .698 .805 .770 .788 .746 - .732 1000 .631 .729 .705 .718 .663 .654 11O0 .545 .615 .617 .613 .S49 1200 .435 .466 - .463 1300 .297 1400 .140 Table Vb Ratio of Elevated Temperature Tensile Strength to Room Temperature Tensile Strength Combined Temn. DF C-1/2 Mo A302 A533 A302-As33 A372,IV A487,2N A487,20 As14C 75 1.000 1.000 1.000 1.000 1.00 1.0 1.00 1.00 100 .990 .981 .980 .983 .990 .980 .990 .990 200 .965 .928 .944 .937 .951 .930 .959 .970 300 1.035 .928 .935 .933 .939 .909 .935 .952 400 1.118 .948 .942 .946 .958 .902 .923 .942 so0 2.125 .967 .947 .958 .986 .90S .928 .938 600 1.075 .966 .938 .952 .989 .909 .941 .922 700 1.010 .934 .908 .920 .939 .918 .968 .878 R00 .935 .868 .850 .857 .828 .815 900 .840 .771 .764 .766 .686 .738 1000 .735 .653 .652 .653 .596 .650 1100 .590 .545 .518 .531 .545 1200 .440 .42S .417 1300 .295 1400 .175 38

T- Y-MEM ig! M:9;IHE4 'N --'I M iii) :E, gl: ..r Til U1 q:'

ial r E -tIT -:H7 111: iv. J:J:

fph Fl r-I" 1 -5 TR!"n .'a '241 O"ir UrE I av TE ITT- 11 =.M

'!I ITt:

L: Z. r jiv =1 Ij 51, MH I

i RM. T O, =TIM ST A 19i 4 -09 AM 7.-.t air f

'frp' m:

. gi ;i -Ir t m E' 9 O.' WI -011-1 a- RIM--li

-- jll

t:- fin N T' M 2.M Tr MR, v HE r.-

,=

I- 13 M

-R iR T Lti -AL IR T--ii Im.. U all:

i IWIR, M. Q'M im !ranksqM ffa$m VT Ira t'M4 N V 9 is! M M,O- 6-Y--

i+l: WR UP, Pig. Sa Variation of yield strength of Mn-Mo-(Ni)

A 302, with temperature. steel, 50

rtzs FzL- .r..l- s:T. - jI t ...... ..

i2 i-:.

t::;-.

'I =1=

Pi Z.'.1p- -! .. e, Airs A it's( .... .......................

i7 R

%Pz;-O=

r:. Ml-CZ, i9i CZ air E-z-

!LE -

jg SFimlm'. 1:--l.

K..*..- -..-.

I .;.'l.gX3=i Ii E: l- '~-,_ ;.- 1 t r7:I! I'-T1. IIj:T. 7 . ' 71I ..- 7::-'7 [4-.~I . F ,r In' -E 7r:i... 1i,-d;['~'..u i, ills.,- Hi

+/-:X, -- Zi- . _$4w*t! ~-V -r. 1~il rI. Aj-e e:lS .r*--4 3:ZA -, 11fooiE

:4"t 3

.-. 'i Ed -

I' r ig. 8b Variation of tensile strength of Mn-Mo- (NO) steel, A 302, with temperature.

51

.*srl.6 -

- l d-M11' .1 I--,-,.-. - *'.-:1.:!

__J_- ::: ..1 ..:. :.

.- __M, :M :!.. -. !: .. ..... ....

.... !i:

NA ... ...

I JV- I...

1. -M.-MI.T.-THMMM

=t:lz; :mpn-n R-T IE; 3 :t'. r qr: J-' :11. t;!;r ii': li 1:.J.

!!w-L Mc L1

1 IM Hir M.. n. tz. :I; II L:

1U.P: it rk m

-fr ni N T;,:,; =f LU!; M: F,h= T 1 7: ri: I Fit T!I1-'J:&!: .3 I.

r 'j:i:1 In: 6T. lui U U: I 5-: .:1.

Hi: a i--1:I-q

_EF Rn 'M M:

i;:1.

iLt, 1:1..

-.iKl IW1-11n: I: 3:1 .:Z MI TM "Pl.

an =z M ir-i

7 r 4 JET 91"ZI Eq Lit, tz T... --= U

M W! am Mi WurM'5 tl:_'_F R: =r. li' Ur.

tn gnti-m t4 -4 j MI,- 4Yi - o t--T' 4--"

Ell-Z-

!Alto .... .... .. -. ....

test n-M't._ 5 i =r TEW"M i=.4.= . e_

-... ..:..Mli

_ -Tral:F1Il:*t Ir-l

-I ' .. :N . 14-.0- gg:!-,

AA HPTF1Jb:TM.4fg~ ~_

Mfl 14" A I,-hmed .. 4' _-' I _"- z F,' =-

-Z L i-Fig. 9a Variation of yield strength of Mn-Mo-(Ni) steel, A 533, with temperature.

53

I ::n .1I i I

m*.:

i l

4 a

4 4

i 4 ;iL Pig. 9b Variation of tensile strength of Mn-Mo-(Ni) steel with temperature.

54

Mi.'I; -

- I:s- :~ FzLO

.... . W .E -

n * .nI.--

FE ... ...

I

=;it 9i i I!ff j I a.,

u d-. iq-q

= m z4F

=Lg liq La &- Z

- 1. t

-=--Z 1 ER I

U i

m I m m

-Fig,:. I OMY 1- n -r -. ..-- --r, --- ---

am1rao tensile strength ratios for O-Mo and Mn-Mo-(Ni), A 302 and A 533, steels.

65

-1. 2 A 2 I A 5 6 7tt 91 2 3 4 5 6 7891 2 3 4 5 6 7 9 1 2 3 4 5 6 7B89l1 j, . ,- : ,.`

I i X f..

_. . IIl'...1 :!.A:,

7q i-H*; z'1!sl-l - Ez Mr

,Y t1 7 " rn-.:: ::::1' a:: _ . __

t Cl)

U) tr cn1 5

J I *

.h TvTI

. I :..

t ...

,i .,_ _ [

t I

1+H1 I ,IlI

.i1_

-. 'ii~i.:1:4 .

t'. _.

4 7I 2

.5 I 1lI- F1- -I 1 1B-11 T.' J,- T .1 .alu I I1 1 'UU 7

__ R-4inqirii-mIn-Q-li1T;

. J I, Wu

- , I 14- 7 I.__}

JI U,

-A WuUn TIME FOR RUPTURE - hours Fig. 21a Stress vs time for rupture of Mn-Mo steels.

U. . -"

t;. III.

44-t ..t' I .....

W.4%A-4...... i;_ .;i,3142:

l.*_t !F

1 if 1"Lf_ t

.... 4 ...................

i: i. 11iil.r..

1 Z;:TI.: ORRPTR co -. i __

ST: r .ii1 _ ---- ii-- 'i----4----

J.-

10 ,0001,000 0,0 Fig, 21b Stress vs time for rupture of Mn Mo steels.

I I 2 A 4 IA 5 67 9 2 3 4 2 3 7a a 2 4 A & 7nG I 2 3 4 5 6 7899 4

3 2

Ill .R HHEIPI t-1 4 TiV0iM ml tZ E,,

CE AT 2 ..

F 2 S v seodr pe lkt w!fr rte of100 hours 2cent steels.

0.0 T {tmSg1 1d 0f i l 10mi i2 Eif rHShn.

.0 11 II lb .16 CREEP RATE -per cent per lOOO hours l Filg. 22 Stress vs secondary creep rate of Mn-Mo steels.

2. ASTM Publication, "Symposium on Heat-Treated Steels for Elevated Temperature Service," September 1966; Article:

"Characterization of Heat Treated Pressure Vessel Steels for Elevated Temperature Service."

- I

. '.K I-.-.

;.- Vf

, * -* E-

. '%{

."SeTEELRM UFOR; *..~.1

.- l -. I- ;

> / ~.

- , -. z.- ., ; .. ; , .v; , J>

^.~,S.,> * , ~.<

I-...

SYMPOSIUM ON HEAT TREATED s * * *-

STEELS FOR I.-* >

ELEVATED TEMPERATURE 4 SERVICE I I

- - \ B:BB SEPTEMBER 19-20, 1966

. .-. NEW ORLEANS, LOUISIANA Held as Part of the 21ST ANNUAL PETROLEUM-MECHANICAL ENGINEERING CONFERENCE Sponsored by ASTM-ASME JOINT COMMITTEE ON EFFECT OF TEMPERATURE ON THE PROPERTIES OF METALS and METAL PROPERTIES COUNCIL OF TIHE ENGINEERING FOUNDATION Edited by M. SEMCHYSHEN CLIMAX MOLYBDENUM COMPANY ANN ARBOR, MICHIGAN THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS United Engineering Center 345 East 47th Street New York , N. Y. 10017

rupture streng was recorded.

the coarse-gri of this phenot Characterization of Heat Compariso steels reveal steels, the qt Treated Pressure Vessel 10,000 hour0 days 0 hours 0 weeks 0 months I I Studies of Steels for Elevated 1000 F, 100, steels thant~

became limit

' ' Temperature Service always limit lower than e tested.

Exposure A. W. PENSE and R. D. STOUT cold work di Department of M1etallurgy and Materials Science but the allo Lehigh University temperature Bethlehem, Pennsylvania toughness I to offset mO Abstract The extension of service conditions of pressure vessels to higher Service c pressures and temperatures has stimulated consideration of steels capable of dustries hai developing more useful combinations of such properties as strength, toughness, practical li fatigue resistance, and creep-rupture strength. The Pressure Vessel Research anticipated Committee of the Welding Research Council has sponsored an extensive series of ingly heavy investigations at Lehigh University to assess the potential usefulness of becomes dii quenched and tempered carbon and alloy steels in pressure vessels for elevated comitant pri temperature service. The program included studies of stress-rupture properties, toughness, fatigue properties, tensile properties, and notch toughness. Eight pressure vessel Anticipa steels, A212 Grade B, A387 Grade B, A517 Grades E and F, A533 Grades A and vated tempt B, A 542 and A543 were included in the program and were studied in both the j Vessel Res welded and unwelded condition. series of in The investigations showed that the steels could be placed into three groups on low-alloy h I

the basis of stress for rupture in 10,000 hours0 days 0 hours 0 weeks 0 months at 1000 F. As expected, the low- the investil est strength was exhibited by the carbon steel, A212 Grade B. At an intermediate standing of with suffici strength level were A533 Grades A and B, A517 Grade F and A543. In the highest strength group were A38? Grade B, A517 Grade E and A542. Using the same basis tively heav of comparison, i.e., 10,000 hour0 days 0 hours 0 weeks 0 months life at 1000 F, the steels could also be grouped tions on qu on the basis of rupture ductility. The steel with the greatest ductility was the favorable c A212 Grade B steel. Materials with intermediate ductility were A542 and A517 resistance Grade E. The A533 Grades A and B, A517 Grade F, A387 Grade B and A543 all steels, ma1 had low rupture ductilities. This low ductility was associated with intergranular cases.

cracking during creep. The pre When welded, the rupture strengths of most of the steels were nearly compa- tions of th, rable to their unwelded counterparts, although for three steels a slight loss in ties at ele~

8

rupture strength was observed and for one steel, A517 Grade F, a substantial loss was recorded. All of the alloy steels were susceptible to intergranular cracking in the coarse-grained heat-affected zone of the weld, while the carbon steel was free of this phenomenon.

Heat Comparison of quenched and tempered and normalized and stress relieved steels revealed that below 800 F for the carbon steel and 950 F for the alloy steels, the quenched and tempered structure is superior to the normalized one for

?ssel 10,000 hour0 days 0 hours 0 weeks 0 months life.

Studies of fatigue at elevated temperatures indicated that between 800 F and cd 1000 F, 100,000 cycle fatigue stresses were more limiting to service for the alloy steels than the stress for rupture in 100,000 hours0 days 0 hours 0 weeks 0 months . Above 1000 F, rupture stress became limiting for the alloy steels. For the carbon steel, rupture stresses were ce always limiting above 800 F. Below 800 F, one-quarter of the tensile strength was lower than either of these fatigue or stress rupture criteria for all the steels tested.

Exposure to elevated temperatures in the 500 F to 1200 F range without prior cold work did not significantly influence the notch toughness of the carbon steel, but the alloy steels had substantial losses in toughness after exposure to this temperature range. When strained prior to aging all of the steels tested had some toughness loss. The inherently good notch toughness of the alloy steels served YI to offset most of this loss.

Introduction higher Service conditions for pressure vessels in the chemical and nuclear power in-

.s capable of dustries have extended the application of the more familiar grades of steel to the gth, toughness, practical limit of their capabilities. The high temperatures and operating pressures

sel Research anticipated and even currently utilized for such vessels have resulted in increas-tensive series7 of ingly heavy sections of the normalized grades of low-alloy steels. As a result, it ilness of becomes difficult to manufacture such heavy-walled vessels, and some of the con-Ols for elevated comitant properties normally considered essential to extended service, such as ure properties, toughness, cannot easily be maintained.

t pressure vessel Anticipating the increasing need for experimental data dealing with the elc-

3 Grades A and vated temperature properties of low-alloy high-strength steels, the Pressure d in both the Vessel Research Committee of the Welding Research Council in 1962 initiated a series of investigations at Lehigh University to explore the characteristics of

o three groups on low-alloy high-strength steels in the elevated temperature range. From the start of ected, the low- the investigation, it was determined that the greatest need was for a better under- I t an intermediate standing of the capabilities of quenched and tempered steels, particularly those

13. In the highest with sufficient hardenability to respond favorably to heat treatment even in rela-ig the same basis tively heavy sections. Previous Pressure Vessel Research Committee investiga-also be grouped tions on quenched and tempered low-alloy steels had already established that tility was the favorable combinations of strength, toughness, high cycle fatigue resistance, and

?

.542 and A517 resistance to aging phenomena in the 500 F to 700 F range were possible in these B and A543 all ; steels, making them more attractive than their normalized counterparts in many th intergranular cases.

The present PVRC studies, which are still continuing, have included investiga-nearly compa- tions of the tensile properties, the fatigue properties and the stress-rupture proper-slight loss in ties at elevated temperatures of a variety of representative quenched and tempered 9

lr;'

Er: , ."...:.' . -.. ... .

N %. - ' ' '

A.I II Iai' steels. The influence of section size and of fabrication operations such as welding were considered in the program, as well as properties such as ambient temperature strength and toughness that are complementary to properties at service tempera-tures. Special consideration has been given to alteration of low temperature proper-Table II. In most cas thicknesses listed in Figs. 16-20, are tak in Fig. 16, some dat ties due to elevated temperature exposure. As a part of this program, the tensile strength, fatigue strength and creep-rupture strength, particularly as they repre-ii sent limiting stresses for design, have been investigated in the 700 F to 900F range to determine over what temperature ranges each becomes limiting.

I While a number of materials were included, the purpose of the PVRC program I has not been to define the behavior of a given material but to develop the over- 0.

all pattern of behavior that is typical of these materials. It is the characterization Steel yin i

of the quenched and tempered steels that is the primary objective, with emphasis

.i on the grouping of materials and generalizations about the kinds of mechanical A212B properties at room and elevated temperature that are representative of this class A533B I of steels. In the presentation of the results of this work, therefore, emphasis will A387B I A542 be placed on the general behavior of the quenched and tempered low-alloy high-I strength steels and their suitability for elevated temperature service as a whole. A543 A517E A517F Review of Data Deta courtesy of the Arm Materials and Heat Treatment Stress Rupture Data Most of the materials included in the original investigation are now covered by recent ASTM designations for quenched and tempered steels. The materials are Dead load stress listed by ASTNI designation in Table 1. Two of these steels, A212B, and A387B, in the program in bot which were studied in the quenched and tempered condition, are not yet covered these tests are show in these compositions by ASTM quenched and tempered designations. The two while the welded tes steels in the A517 designation are more commonly known by their proprietary this latter specimen names "SSS-100" (Grade E) and "T-l" (Grade F). While the steels designated as deform together durinj A533, Grades A and B, fall within the chemistry limits for these grades, the room which are particularly temperature mechanical properties of the specimens exceed the maximum tensile selected to produce fe strength allowed by this specification. They are therefore A533 composition ma- between 100 and 5000 terial only and this is so indicated when appropriate. The thicknesses of the I

plates tested and the heat treatment temperatures for the various steels are listed in Table 1. The room temperature mechanical properties of the steels are listed in I I

Table I I Chemical Compositions, Heat Treatment Temperatures and Welding Electrodes t A

Temper-Weld.

Aust. Ing Ing C Mn P S 51 N1 Cr Mo V T1 Cu B Temp.Temp. Eleetr. I A212B 0.26 0.70 0.010 0.024 0.23 1650 1150 E7016 I A533A 0.19 1.28 0.013 0.010 0.22 -

A533B 0.20 1.28 0.019 0.030 0.21 0.53 0.45 0.52 1650 1650 1150 E10018 1150 E10018 i'I B A387B 0.17 0.59 0.012 0.024 0.21 - 0.91 0.51 - - - - 1675 1150 E11018 A542 0.11 0.37 0.010 0.010 0.26 - 2.20 0.96 - - - - 1700 1150 E11018 A543 0.15 0.26 0.010 0.022 0.18 2.84 1.52 0.45 - - - - 1700 1150 E11018 AS17E 0.15 0.65 - - 0.28 - 1.76 0.50 - 0.072 - 0.002 - - E11018 AS17F 0.18 0.85 0.008 0.017 0.25 0.85 0.48 0.50 0.04 0.003 0.27 0.004 1700 1260 E11018 Welding done with a travel speed of 10 in. per minute. 200A. approx. 22V. FIG 10

1_

ns such as welding Table II. In most cases, the data presented in this paper are for the steels and -

mbient temperature thicknesses listed in Table I. Some of the data, particularly those illustrated in ervice tempera- Figs. 16-20, are taken from other heats of the same steel tested at Lehigh, and temperature proper- in Fig. 16, some data from the literature are included.

,ram, the tensile I as they repre- Table 11 700 F.to 900 F Room Temperature Mechanical Properties imiting. of the Quenched and Tempered Steels PVRC program velop the over- 0.2% Offset Tensile Elongalion Reduction ie characterization Steel Yield Strength Strength In I in. of Areo e, with emphasis psi psi % %

of mechanical A212B 49,400 79,500 30.0 67.1 ve of this class A533B 121,500 130,500 20.0 63.2 re, emphasis will A387B 129,500 140,600 16.0 63.8 low-alloy high- A542 119,000 137,000 25.0 70.4

.vice as a whole. A543 126.000 135,000 21.5 68.4 A517E 107,500 117,800 19.5 68.9 A517F 119,000 131.000 19.5 63.2 fData courtesy of the Armeo Steel Compmny.

Stress Rupture Data e now covered by e materials are Dead load stress rupture tests were performed on the eight materials included

.2B, and A387B, in the program in both the unwelded and welded condition. The specimens used in not yet covered these tests are shown in Fig. 1. The base plate test specimen is specimen A, ions. The two while the welded tests were run on composite specimen B. It was the purpose of r proprietary this latter specimen to force the base plate, weld metal and heat-affected zone to

ls designated as deform together during testing, thereby revealing those regions of the composite grades, the room which are particularly sensitive to fracture initiation. Testing stresses were maximum tensile selected to produce failure in less than 10,000 hours0 days 0 hours 0 weeks 0 months with the majority of the data

omposition ma- between 100 and 5000 hours0.0579 days 1.389 hours 0.00827 weeks 0.0019 months [11, [2].

esses of the steels are listed eels are listed in i.

ding Electrodes Temper. Weld.

Aust. ing ing Temp.Temp. Electr.

1650 1150 E7016 1650 1150 E10018 1650 1150 E1018 1675 1150 E11018 1700 1150 E11018 1700 1150 E11018 I .

a - - E11018 1 1700 1260 E11018 I FIG. 1 - STRESS RUPTURE TEST SPECIMENS i 11 I

i

-As.

Wa 0TP A summary of these data for welded and unwelded specimens are found in Figs. above 40 per cent re 2-15. It should be noted that these figures include two kinds of information- Grade E and A542. I the stress for rupture in 10, 100, 1000, 5000 and 10,000 hours0 days 0 hours 0 weeks 0 months , and the rupture drops below 40 per c ductility for these same time periods. ing are A387 Grade I Base Plate Tests. The stress rupture data for the base plate specimens are be noted that the stt presented in Figs. 2-8. For case of comparison, selected rupture strength data those with the lowes taken from these curves are also listed in Table 111. If we consider the rupture level have the lowei strength for 10,000 hour0 days 0 hours 0 weeks 0 months life at 1000 F, however, it is possible to classify the Metallographic e.N steels into three groups. In the lowest strength (5000 psi) category is the A212 has revealed that thi Grade B steel. The next group, of intermediate strength (10,000- 12,000 psi), con- prior austenite grai tains four alloy steels including A533 Grades A and B, A517 Grade F and A543. the lowest ductiliti The highest strength (20,000-23,000 psi) group includes three steels, A387 fracture, while exte i Grade B, A517 Grade E and A542. It may be observed that the intermediate the test gage length strength group of steels consists of manganese-molybdenum (with or without eluding microprobe nickel) and nickel-chromium molybdenum steels, while the highest strength group cracking.

is primarily a chromium-molybdenum group. The material with the highest rupture strength of any tested in the program is the quenched and tempered 2% per cent Welded Composi chromium-1 per cent molybdenum steel (A542). of the same steels Although these groupings hold strictly only for 10,000 hour0 days 0 hours 0 weeks 0 months life at 1000 F, were welded withou Table III indicates that much the same relationship between the steels continues in Table 1. The sp to hold at lower temperatures and shorter times. The A542 and A517 Grade F are, in shallow grooves however, proportionately somewhat higher in strength at lower temperatures and mens transverse to shorter times than indicated above. When comparing noted that the same Table IlI the base plate test A533 Grades A and Rupture Stresses for the Quenched and Tempered Steels or slightly greater Grade B and A543, Stress to Rupture (psi) Stress to Rupture (psi) in 1000 hr. at, OF in 10,000 hr. at, 0F from welding and t 800 900 1000 1100 800 900 1000 1100 about the same ma A212B 30,000 17,000 8,500 - 23,000 11,000 5,000 - sensitive to weldin A533A - 60,000 19,000 10.000 - 43,000 11,000 - the 800 F to 1000 A533B - 55,000 20,000 10,000 - 40,000 11,000 - rupture times longe A387B - 77,000 35,000 13,000 - 62,000 20.000 7,500 As Figs. 9-15 A542 - 88,000 41,000 15,000 - 75,000 23,000 8,00(, of A212 Grade B, A543 85,000 55,000 23,000 11,000 - - 12,000 6,000 60,000 35,000 16,000 79,000 48,000 20,000 10,000 times were increas A517E* 82,000 A517F 89,000 74,000 32,000 8,500 85,000 t2,000 13,000 4,000 both in the weld m rupture cracking di

Data courtesy of the Armco Steel Company was the coarse gre

1 a common characte Although the rupture strength of the steels appears to be of primary importance strongest in those for service, rupture ductility may also be of interest. Reference to Figs. 2-7 re- .1 in the base plate t veals that the steels would be grouped in quite a different order on the basis of J~ steels the fracture reduction of area at failure. The carbon steel A212 Grade B, has the best rupture zone. The stress r ductility of any steel tested, always exhibiting reductions of area of 60 per cent toe of the weld an or greater, a characteristic not observed with any of the alloy steels. The alloy with final fracture steels can be divided into two groups on the basis of ductility. In the first group, affected zone regi ductility at failure decreases regularly as time for rupture increases, particularly appears that the t in the 900 F to 1100 F range. However, up to 10,000 hours0 days 0 hours 0 weeks 0 months , the ductility remains intensified in the 12

arc found in Figs. above 40 per cent reduction of area. The two steels in this category are A517 f information- Grade E and A542. In the second alloy steel group, the fracture reduction of area ind the rupture drops below 40 per cent, and in some cases well below 20 per cent. In this group-ing are A387 Grade B, A517 Grade F, A533 Grades A and B and A543. It should

specimens are be noted that the steels with the highest stress-rupture strength are not necessarily ire strength data those with the lowest rupture ductility, but that the steels of intermediate strength der the rupture level have the lowest ductilities.

to classify the Mletallographic examination of specimens failing with low rupture ductilities

ory is the A212 has revealed that this phenomena is associated with intergranular cracking along

-12,000 psi), con- prior austenite grains in the quenched and tempered structure. In specimens with nade F and A543. the lowest ductilities, little if any plastic flow appears to be associated with the steels, A387 fracture, while extensive grain boundary sliding and cracking is evident throughout nrtermediate the test gage length. Extensive studies using light and electron microscopy, in-th or without cluding microprobe analysis, did not reveal any specific cause for intergranular lst strength group cracking.

e highest rupture red 2% per cent Welded Composite Tests. The results of tests on welded composite specimens of the same steels shown in Figs. 2-8 are presented in Figs. 9-15. The steels ife at 1000 F, were welded without preheat using the electrodes and welding parameters indicated steels continues in Table I. The specimen used, (Fig. 113) was obtained by depositing weld metal k517 Grade F are, in shallow grooves on opposite sides of the base plate and cutting out the speci- fI emperatures and mens transverse to the weld beads.

When comparing the base plate and weld composite specimen results, it may be noted that the same general grouping of steels on the basis of rupture strength for the base plate tests is followed by the welded specimen tests. For four steels, A533 Grades A and B, A542 and A387 Grade B, the weld composite was equal to iteels or slightly greater in rupture strength than the base plate. For two steels, A212 Rupture (psi)

Grade B and A543, approximately 15 per cent reduction in rupture strength resulted 3 hr. at,0 F from welding and testing at 900 F; while for A517 Grade E, a loss in strength of 1000 1100 about the same magnitude occurred in the 1000 F to 1100 range. The material most 5,000 -

sensitive to welding in terms of decreased rupture strength was A517 Grade F. In 11,000 - the 800 F to 1000 F range, rupture strengths were reduced by 30 - 40 per cent for I I ,000 - rupture times longer than 100 hours0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months in the welded condition.

20,000 7,500 As Figs. 9-15 indicate, failures in the welded specimens, with the exception 23,000 8,000 12,000 6,000 of A212 Grade B, had an increasing tendency to occur in the weld zone as rupture 20,000 10,000 times were increased. The term "weld zone" in these figures includes failures l 13,000 4,000 both in the weld metal and the heat-affected zone. In some of the steels, stress

'! rupture cracking did occur in the weld metal, but the most common site of fracture was the coarse grained region of the heat-affected zone. While this appears to be a common characteristic of all the alloy steels tested, the tendency appears to be

)rimary importance strongest in those steels that were the most susceptible to low ductility fracture to Figs. 2-7 re- in the base plate tests, (for example, A543 and A517 Grade F). In many of these

, on the basis of steels the fracture was almost exclusively in the coarse grained heat-affected s the best rupture zone. The stress rupture cracks usually initiate at the specimen surface at the ea of 60 per cent toe of the weld and progress along the heat-affected zone under the weld metal, teels. The alloy with final fracture occurring through the base plate separating the two heat-In the first group, affected zone regions. The fracture was along prior austenite grain boundaries. It ases, particularly appears that the tendency for intergranular cracking observed in the base plate is ductility remains intensified in the microstructure in the coarse grained heat-affected zone.

.i 13 L

E5Z-,.*. -

, V., , ". "; ,

Il

2 i

700 8a) 900 IC10O 1COO 1200 Temperoture-OF FIG. 2 - STRESS RUPTURE

SUMMARY

CURVE FOR A212 GRADE B,OUENCHED AND TEMPERED 11PEA ZU. - . I I I A533 Grades ABB ItX) Crnpcsition Rupture in Rupture Reduction of Arec Grade f30 \

a08rs 1003 0** 60-90%

E Doa a 0 20-40%

a E3 -20%.

so In V) I

.~

.~' I P.

410 S-f

!0 -

C Unwelded

__ I I I_

O 700 800 900 1000 1100 130 Temperature-OF FIG. 3 - STRESS RUPTURE

SUMMARY

CURVE FOR A533 GRADES A AND B 14 I

El2l l. l

-- 7-I A387Grade B (Quenched 8Tempered) 100 upture in Rupture Reduction IOHI'I4 Of Area e?80 40 -0640%

.2 0.

IOHrs 0040 it

_ 60 0540 40Hr\

20[

Unwe Ided

.I I OL I. . . . . .

700 Boo 900 000 1000 1200 Temperature -IF l FIG. 4 - STRESS RUPTURE

SUMMARY

CURVE FOR A387 iR A212 GRADE B,QUENCHED AND TEMPERED .

.,7 I

C) 0-50%

0-60% 0.

0 -40% I 3 -ZOS i p2 CA, I

0 CO, i U) 800 900 1000 1100 Temperature -°F OR A533 FIG. 5 - STRESS RUPTURE

SUMMARY

CURVE FOR A542 15 k-,'-- ._ '. --

, .1 I.

vi.."." ... - -,.. .

-17nI 1201 A543 Rupture In 100 Rupture Reduction

'.N. Of Area

n ICO 6% 660-90%

A40-60%O 2 80 020-40% : 8C 0 0o-20% 1!I

5 0.. 40% \

,' 60 6(

v0 10 to In M 40 4C v/)

20S 0 20 . . . ..

Unwelded r 2 ) :

II _ C 700 800 900 IC00 1100 1200 Temperature-IF FIG FIG. 6 - STRESS RUPTURE

SUMMARY

CURVE FOR A543 l

12(

, i - - III A5J7 Grade E 11'cccelorNH'1 icx 100 - Rupture in ' .IIJ I inRupture low.* ReductiorI In

0 Of Areo Kxo10 s-

60soS90 I 8c - S A 40-60%

I

~-

w 60 I

va4 40

.8I LI-I 6

E40 I

-.1 I 2 20 I Unwelded A iI 6

I - lI I I II 700 800 900 1000 1100 1200 Temperature--F FIG. 7 -STRESS RUPTURE

SUMMARY

CURVE FOR A517 FIG. '

GRADE E 16 lii [

4, 0

G) 70( 800 900 KW00 1100 1200 Temperature-'F FIG. 8 - STRESS RUPTURE

SUMMARY

CURVE FOR A517 GRADE F .

i I

.4 t

0.

P-a.-1 4,

800 900 1000 1100 .i Temperature- 0F FIG. 9- STRESS RUPTURE

SUMMARY

CURVE FOR WELDED A212 GRADE B QUENCHED AND TEMPERED 17 J.. S ,. -.: . ..

/f11 low 10D n ~ . I-

A 533 Grodes A 8 'B Composition l

I I

ur Rupture in Fbilure Location I 80 IoH5- Grade I~A B R XCOHrs *

Base Plate -

CL. \ 0 Weld Zane

_60 lOOOHr1-0 \

I U, S4O i 201 Welded I IIf I 0

700 800 900 1000 1100 1200 Temperature -F FIG. 10 - STRESS RUPTURE

SUMMARY

CURVE FOR WELDED A533 GRADES A AND 8

.1 aI

.I 09 D1 r

i I, 800 900 1000 Temperature-OF I

FIG. 11 - STRESS RUPTURE

SUMMARY

CURVE FOR WELDED A387 GRADE B QUENCHED AND

'TEMPERED 18

jLr. - ,- , I I I A542 Rupture in 100 IOHKr -t Failure Locaft

- 0OH,-*\\

Base PIC1in v 80 A oWeld Zon ale J.

, 60 n 40 201 Welded 0/

U . . .

v 700 800 900 1000 1100 1200 Temperature -IF I

FIG. 12 - STRESS RUPTURE

SUMMARY

CURVE FOR WELDED A542 I

7-1201 I i a A543 100~

late Failure Location one

Base Plote I0OOHrs- O WeldZone 80

. 60 I- r!

J in An I.

_vj

20) Welded ri, 700 800 900 1000 1100 1200 Temperature-T i FOR AND FIG. 13 - STRESS RUPTURE

SUMMARY

CURVE FOR WELDED A543 19

.1 I.

I

.1

. ,- ; ,;w,, , ,

vw, - . , ': , - :

,;L;

i:

-ii I

. .II

.. II Although the r I

I have a bearing on

.Ii Lehigh Universit]

.I., weld zone cracki

'el thermal stress re]

0. current tests on r 1i I relief that are ide tests [2]. The st4 W 61 have the lowest r tendency for heat b-)

I' IV

.t: 4( Influence of Micr4 U,

Although quen normalized and st strength and notc I tempered microstn temperature range I

II normalized structu C0 1000 1100 elsewhere for cart I Temperolure-IF the temperature ra FIG. 14 - STRESS RUPTURE

SUMMARY

CURVE FOR depends on compo WELDED A517 GRADE E : Figs. 16 and 17. (

quenched and teml and stress relieve 11n I j

.1

.v II I I I I

- A517 Grade F 80 I

100i ( T-1')

P iure in Failure Locotion 70 x

i 1 80 a0Hrr--\

10HC& BasePlote 0 Weld Zone ' 60

-.R ,I

0) 50-

-P 60 I.'

C0 g-40 U 40 930

\0 \

20 a 20 Welded 10 n _ L - . - St j'II u 700 800 900 1000 1100 120 Temperature-IF 0L FIG. 15 - STRESS RUPTURE

SUMMARY

CURVE FOR WELDED A517 GRADE F Fl i

20 II,

II Although the results of these tests are of significance to service, they also have a bearing on the heat treatment of these steels. Tests presently underway at Lehigh University appear to indicate that the steels which have a tendency for weld zone cracking may also be susceptible to heat-affected zone cracking during thermal stress relief. Such cracking has been reported in the literature [31, and current tests on restrained weldments have produced cracks during thermal stress relief that are identical in appearance to those produced in the stress-rupture tests [2]. The steels most susceptible to this phenomena appear to be those that have the lowest rupture ductilities in the base plate tests and have the greatest tendency for heat-affected zone failure in the welded composite tests.

Influence of Mlcrosfructure Although quenched and tempered steels have a significant advantage over normalized and stress relieved steels from the standpoint of yield strength, tensile strength and notch toughness, it has been accepted that the quenched and tempered microstructure will be subject to spheroidization in the creep-rupture temperature range and therefore may be inferior in rupture strength to the coarser normalized structure. Comparison of stress-rupture data obtained at Lehigh and elsewhere for carbon and alloy steels has confirmed this general conclusion, but the temperature range over which the normalized microstructure is advantageous depends on composition. A comparison of this type for four steels is seen in Figs. 16 and 17. On the basis of 10,000 hour0 days 0 hours 0 weeks 0 months life, these figures show that the quenched and tempered microstructure does hold advantage over the normalized and stress relieved one up to 800 F for the A212 Grade B, up to 950 F for the IV II I

Temperature-OF FIG. 16 - THE INFLUENCE OF MICROSTRUCTURE ON STRESS FOR RUPTURE IN 10,000 HOURS ON A212 GRADE B AND A302 GRADE B 21 e---

.Q 14.:: .... . .- -

x I .. . "

". .f - -

I:

ait

' Vk *I

-60 strain in L shown in F design witl is a djstin~

fatigue co cyclic ope ture range.

I. V) 50 cycle fati

°2 In the tern 0 40 becomes Below 800 a:3 ing for deS considerat one quart d 20 stress rup been conc obtained t impaired l specimen.

900 1000 I100 900 0D) IKX)

Temperoture-OF FIG. 17 - THE INFLUENCE OF MICROSTRUCTURE ON 40 STRESS FOR RUPTURE IN 10,000 HOURS ON A387 GRADES B AND D 30 A302 Grade B (A533 Grade A composition) and for the A387 Grade B. The stress.

oh2 relief temperature (following normalizing treatment) in these cases is 1150 F. For U'

the A387 Grade D (A542 composition) the stress relief temperature is one specifi-2 E 20 cally selected to produce structural stability (1350 F), while the tempering temper- V7 ature for the quenched steel is more typical of that used to produce high room .1 temperature strength and satisfactory toughness (1150 F).Under these conditions 10 Al of heat treatment the quenched and tempered structure is superior up to 1050 F. ii Elevated Temperature Fatigue Data 01 While stress-rupturc properties are of primary consideration in pressure vessel service above some threshold temperature, presumably greater than 650 F, it is of interest to know what role fatigue plays in this threshold range, and where the threshold temperature for creep-rupture may be expected to fall. Fig. 18 is a compilation of data that delineates some of these relationships in the threshold range, i.e., between 700 F and 1100 F. The one-quarter tensile strength curves and extrapolated stress-rupture curves (extrapolated one cycle on the conventional Elevated log stress-log rupture time curve) are derived from the current PVRC program. The fatigue curve is derived from a companion PVRC study using fully reversed canti- All of lever bending specimens on the same quenched and tempered steels studied in rupture s the stress-rupture investigation [4]. These tests were conducted at 1100 cycles steels at per hour and were strain controlled. The safety factor of four applied to this temperat, curve is a result of full scale pressure vessel tests sponsored by PVRC at South- and oper west Research Institute. The fatigue strength reduction factor for 100,000 cycle for press life based on vessel membrane strain in these full scale vessels as compared to strength, il 22

I- B k

strain in Lehigh tests on the same material, was about four [5). The curves shown in Fig. 18 suggest that fatigue failure should not be a serious problem in 587 deD design with the carbon steel, but that for the two alloy steels this type of failure brPS0) is a distinct possibility in preference to either tensile or stress-rupture failure if fatigue conditions are encountered in service. Some vessel applications involve cyclic operations applying fatigue cycles at slow rates in the elevated tempera.

ture range. Hence, in service lives that extend for a number of years, 100,000 cycle fatigue life and 100,000 hour0 days 0 hours 0 weeks 0 months rupture strength may both be of significance.

In the temperature range between 800 F and 1000 F the stress for fatigue failure becomes more limiting from the design standpoint than the stress for rupture.

Below 800 F, the quarter-tensile-strength at temperature criterion becomes limit-ing for design, while above 1000 F, stress rupture becomes, the most important consideration. For the carbon steel, design below 850'F appears to be limited by one quarter of the tensile strength at temperature while above this temperature, stress rupture is limiting. Current studies on elevated temperature fatigue have been concerned with the influence of welding on this fatigue behavior. The results obtained thus far indicate that the fatigue resistance of the alloy steels is not impaired by the presence of iveld metal or heat-affected zone structures in the specimen.

I_ a I I-I I & & I I JRE ON 40 i A212B (OrT)

A387B I (OMT) A517F I ("T-1") -

URS ON I I Stress I Streom I Rupture 30 VYSlrength I (lOPre I Grade B. The stress

Stress RupureStrro A-_\I 0C,00DHW cases is 1150 F. For I- fioomirns) l-Tensile fFPI% it I I I 4

-2 Strenth A i ok ~Stwgth \4P" erature is one specifi. 0In e the tempering temper-IA 20 - A~ \ '-. .

9 Strengtn Stror I Fatigue \18\

produce high room I tloaocI II katigue 'i nder these conditions 10 perior up to 1050 F.

_t @ Z I . I I ^ I. I. I. . .

n W70D . . . . .

ion in pressure vessel 900 1100 700 900 1100 700 900 1100

er than 650 F, it is of Temperoture- 0 F nge, and where the FIG. 18 - THE RELATIONSHIP BETWEEN TENSILE, fall. Fig. 18 is a STRESS-RUPTURE AND FATIGUE CRITERIA FOR lips in the threshold QUENCHED AND TEMPERED A212 GRADE B, A387 GRADE B,AND A517 GRADE F sile strength curves cle on the conventional Elevated Temperature Exposure Data nt PVRC program. The g fully reversed canti- All of the mechanical properties discussed thus far-tensile strength, stress-d steels studied in rupture strength and fatigue strength-are directly related to the behavior of the ucted at 1100 cycles steels at the service temperature. One property not directly related to elevated ir applied to this temperature service but still of importance under certain conditions of fabrication red by PVRC at South- and operation is notch toughness. Quenched and tempered steels are advantageous tor for 100,000 cycle for pressure vessel service not only because of the improved yield and tensile ssels as compared to strength, but also because of the decided improvement in notch toughness that 23 I,"

7.r '...

Z.

,Z"-

ve_.

Z' 1::

.. , V

!ZZ i

..1.

...' I 4!.,.

they provide. Even in relatively heavy section sizes, it has been shown [6] that adequate toughness can be maintained in quenched and tempered steels. Exposure to elevated temperatures, either with or without prior cold work, can lead to marked reductions in toughness. Figs. 19 and 20 illustrate the losses of tough-ness that have been found to occur. In Fig. 19, the toughness loss due to long-time exposure is illustrated for both light and heavy sections of four quenched and tempered steels, while in Fig. 20, the effect of short-time exposure to ele-vated temperatures after 5 per cent cold forming is illustrated. In these figures the term "light section" refers to a quenched and tempered plate approximately 1 in. in thickness, while "heavy section" refers to a plate approximately 12 in.

in thickness. It should be observed that in the non prestrained condition, the carbon steel A212 Grade B was not sensitive to toughness losses by exposure to the 500 F to 1150 F range. On the other hand, the initial transition temperature of this material was high compared to the alloy steels. When strain aged the em-brittlement of the A212 Grade B is more substantial. In the unstrained condition, Fig. 19, the A533 Grade A material was also relatively insensitive to toughness losses up to 1150 F, where some loss is observed. The sensitivity of this ma-terial to aging phenomena after cold work is evident in Fig. 20. The most sensi-tivity to aging is found in the more complex alloy steels such as A543 and A517 Grade F. These materials show substantial toughness losses due to exposure either with or without prior cold work. The initial excellent toughness of these materials, however, serves to compensate to a large extent for these losses dur-ing exposure, rendering them superior to the carbon steel under most conditions of service.

+200 _

+160 v This d

.120 - fito TI T

1. A pered loA E- _

+~I80 _ for 10,00(

stress as a) Ru b) Ru ie > t s wg se ~fs~ue 8 v Gri 2CL __hn tLih Ti hk Ti hc c) Ru j -120 - i SsI an

2. As Z-160 - -zo u.e i reduction 10 -k ExaseurTang c ~posare ciated wi 6L>

t) -200 _

- nretfo s curki !jEx 5 !j.IrW

-5o.TranilflnToemp. -

10,000 hc ture redui

--4 Thin Thick I Thin Thick Thin Thick a) Re t~lt Roetim. Ror n _Re4n -qonn Ro~tr~fi b) Re Ii A212B A533A A517F A543 c) Re am FIG. 19- THE INFLUENCE OF ELEVATED TEMPERATURE 3. In EXPOSURE ON THE TOUGHNESS OF A212 GRADE the base B, A533 GRADE A, A517 GRADE F AND A543 Grades 24

I 1'§LI I'1 ias been shown [61 that empered steels. Exposure d work, can lead to te the losses of tough.

tness loss due to long- I.

AL a

E tp8>U.age wn L-

. km1mi- ~t

1 gE~

1 D=

I Mone

ions of four quenched

>-time exposure to ele-rated. In these figures rcd plate approximately hi I.

II I

>s 7 -

' 11, I

mte approximately 12 in.

rained condition, the Ii

. ' SL . "."

1-11.01

.1.1.000.'

I a3

.00 '. :

ss losses by exposure to

.T-1-Ir"-.' b 00 .11 .'

transition temperature of .5

-.1 I.,

n strain aged the em- ~

U

- *.~

§-Aolng Tear, h;%rin he unstrained condition, -160 gl-incraj OutingAgi j nsensitive to toughness J Baselondta~mp.

-200 sensitivity of this ma- IEDeevaDurisng gn_

l _ . _ .

'ig. 20. The most sensi- .

_eœn_

-iS4U Thin ThidI - Thin Thick Thin Thick such as A543 and A517 _e__l__ __

Cole-.

g I _

4 Aon c *ation Sudde[n

___ A5 ofntinfn sses due to exposure A212B I A533A IA517F IA543 ent toughness of these -

nt for these losses dur- FIG. 20- THE INFLUENCE OF 5 PER CENT STRAIN AND I under most conditions 1 HOUR EXPOSURE ON THE TOUGHNESS OF A212 GRADE B, A533 GRADE A, A517 GRADE F AND A543 Summary This data review may be summarized as follows:

1. A comparison of the stress-rupture properties of eight quenched and tem-fi * -

S

pered low-alloy high-strength steels in the 900 to 1100 F range has indicated that for 10,000 hour0 days 0 hours 0 weeks 0 months life at 1000 F, these steels may be grouped on the basis of rupture stress as follows:

a) Rupture stress approximately 5000 psi - A212 Grade B steel.

b) Rupture stress approximately 11,000 psi - A533 Grades A and B, A517 Grade F, and A543 steels.

c) Rupture stress approximately 20,000 psi - A387 Grade B, A517 Grade E, and A542.

2. As time to rupture increases in the 900 to 1100 F temperature range, the b-P reduction of area at failure decreases for the alloy steels. This decrease is asso-ciated with extensive intergranular cracking along prior austenite grains. For

~I'

. -- 'o 10,000 hour0 days 0 hours 0 weeks 0 months rupture life at 1000 F, the steels may be grouped on the basis of rup-ture reduction of area as follows:

rhin Thick a) Reduction of area greater than 60 per cent - A212 Grade B steel.

c7ion Section b) Reduction of area between 40 and 60 per cent - A517 Grade E and A542.

517F A543 c) Reduction of area below 40 per cent - A387 Grade B, A517 Grade F, A543, and A533 Grades A and B.

EMPERATURE

3. The welded composite specimens were comparable or nearly comparable to

A212 GRADE the base plate in stress rupture strength for most of the steels. Four steels, A533 kND A543 Grades A and B, A542 and A387 Grade B when welded displayed equal or greater 25

. . I..

stress rupture strength than the base plate. Three steels, A212 Grade B, A543 and A517 Grade E suffered approximately 15 per cent loss in strength when welded compared to the base plate. One steel, A517 Grade F, lost 30 - 40 per cent of its rupture strength when welded.

4. The welded composite specimens of the alloy steels showed a tendency for low ductility intergranular fracture in the coarse-grained heat-affected zone of the welds. The materials most susceptible to this type of failure were those failing The with low ductilities in the base plate tests. The carbon steel, A212 Grade B, was free of this type of fracture.

5. The quenched and tempered microstructure is superior in 10,000 hour0 days 0 hours 0 weeks 0 months stress rupture life to the normalized and stress relieved up to 800 F for the car-bon steel A212 Grade B and up to 950 F for the three alloy steels, A387 Grade B, of (

A533 Grade A and A542.

6. For applications where elevated temperature fatigue is a consideration, in the temperature range between 800 F and 1000 F the 100,000 cycle life stress is 6 lower than the stress for rupture in 100,000 hours0 days 0 hours 0 weeks 0 months for A387 Grade B and A517 Grade F. For the carbon steel A212 Grade B, the rupture stress is less than the fatigue stress over this range. Below 800 F, one-quarter of the tensile strength is lower than either of the other two strength criteria for these three steels.

7. Exposure to elevated temperatures either with or without prior cold work can cause losses in notch toughness in the quenched ana tempered low-alloy high-strength steels. -The carbon steel A212 Grade B was not so seriously affected.

Three alloy steels tested, A533 Grade A, A517 Grade F and A543 all had some substantial toughness losses due to exposure, but the inherently good notched toughness of these steels served to substantially offset the loss.

Acknowledgment Until 2'4Cr-1m~

The authors gratefully acknowledge the financial assistance and technical l veloped I cooperation of the Pressure Vessel Research Committee of the Welding Research , can be ol Council, and of many of the companies represented by the council. Special ac- :.1 evaluati knowledgment is made to the United States Steel Corporation, the Arnco Steel i! ized an Company and The American Oil Company for their contribution of data to the pro- showed gram. The authors also are appreciative of the contribution of Mr. J. J. deBarba- i ties can dillo of the Lehigh Staff to the program. However, References to those conduct

[1] V. S. Robinson, A. W.Pense and R. D. Stout, "The Creep Rupture Properties of uniformr Pressure Vessel Steels," Welding Journal, Vol. 43, Research Suppl., pp. 534-s to 540-S, and temp 1964.

[2] J. J. deBarbadillo, A. W.Pease and R. D. Stout, "The Creep Rupture Properties of Pressure Vessel Steels - Part II," IVelding Journal, to be published, 1966.

[3] W.D. Doty, and C. E. Crotke, "Some Observations on the Weldability of Quenched and Tempered High-Yield-Strength Alloy Steels," Conf. Proc., U. S. Navy Bureau of Ships, A nu Washington, D. C., pp. 34 - 67, March 21-22, 1960. grade 1

[4] R. A. DePaul, A. W.Pense and R. D. Stout, "The Elevated Temperature Fatigue presenc Properties of Pressure Vessel Steels," Welding Journal, Vol. 44, Research Suppl., reactio pp. 409-s to 416-s, 1965.

[5] J. It. Gross, "PRVC Interpretive Report of Pressure Vessel Research Section 2- crakin Materials Considerations," Welding Research Council Bull. No. 101. C

[61 R. D. Stout, "Higher-Strength Steels for Weldments," Welding journal, Vol. 39, stocks, Research Suppl., pp. 273-s - 2B3-s. future.

26

3. ASTM Publication, "Analysis of Data from the Symposium on Heat-Treated Steels for Elevated Temperature Service."

7 - . ---- --

ANALYSIS OF DATA FROM SYMPOSIUM ON HEAT- TREATED STEELS FOR ELEVATED TEMPERATURE SERVICE Prepared by:

E. B. NORRIS and R. D. WYLIE Southwest Research Institute Department of Materials Engineering Prepared for:

THE METAL PROPERTIES COUNCIL United Engineering Center 345 East 47th Street New York, New York 10017 THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS United Engineering Center 34S East 47th Street New York , N. Y. 10017 p..

- . 1! 1!,,,., ri -. *P: H,

SI. *JUPIWJ .- ,'*

Contents Page Tables......................................................................................... ii Illustrations.................................................................................. iv Introduction................................................................................... 1 Evaluation of Data Presented at Symposium. 1 Tensile and Creep Rupiure Properties. 1 Determination of Allowable Stresses. 4 Notch Toughness Considerations ....................... ,,................. 5 Notched Rupture Characteristics. 5 Long Time Exposure Effects. 5 Hydrogen Environmental Studies. 5 Properties of Welds. 5 Fatigue Studies .12 Recommendations for Future Work .12 General..12 Chemical Analysis Limits .................. , . ............ 12 Strength Determinations .13 Microstructural Stability ................ , ................................ 13 Fracture Toughness.................................................................... 13 Effect of Welding .14 Low Cycle Fatigue................................................................... 14 Environmental Effects .14 Summary - Priority .14 References .15 Appendix A - Illustrations of Experimental Data .18 Appendix B - Tables of Experimental Data.............................. ,.,.,.,.40

WN I Tables Table Page I Materials Evaluated in Symposium Papers ............................... 2 11 Three Classifications for 2-4 Cr-1 Mo Data............................. 2 III Summary of 2-Y Cr-1 Mo Stress Rupture Strength ....................... 3 IV Stress Rupture Strength of Several Quenched and Tempered Steels. 4 V Comparison of Derived Allowable Stresses Based on 25 per cent of Minimum Tensile Strength for the Four Groups of Steels Evaluated.............................................................. 10 VI Comparison of Derived Allowable Stresses Based on 33-%, per cent of Minimum Tensile Strength for the Four Groups of Steels Evaluated...................................................................... 11 B.I Tensile Properties of A542 Steel Tempered at 1075 to 1125 F.41 B.1l Tensile Properties of A542 Steel Tempered at 1150 to 1200 F.42 B.IH1 Tensile Properties of A387 Grade D Steel Tempered at 1250 to 1350 F.43 B.IV Tensile Properties of A212 Grade B Steel .44 B.V Tensile Properties of A387 Grade B Steel .44 B.VI Tensile Properties of A517 Grade F Steel Tempered at 1260 F ............................. 45 B.VII Tensile Properties of A517 Grade E Steel .............................. 45 B.VIII Tensile Properties of A387 Grade E Steel ............................. 46 B.IX Tensile Properties of A508 Steel (Tempering Temperature 1050 F')......................................................... 47 B.X Room Temperature Tensile Properties of Several Quenched and Tempered Steels .47 B.XI Stress Rupture Properties of A542 Steel Tempered at 1075 to 1125 F .48 EB.XII Stress Rupture Properties of A542 Steel Tempered at 1150 to 1200 F.49 B.XIII Stress Rupture Properties of A387 Grade D Steel .50 B.XIV Stress Rupture Properties of A212 Grade B Steel Tempered at 1150 F .51 iil1

I B.XV Strcss Rupture Properties of A387 Grade B Steel Tempered at 1250 to 1350 F........................................ 52 B.XVI Stress Rupture Properties of A387 Grade D Steel Tempered at 1150 F........................................ 53 B.XVII Stress Rupture Properties of A517 Grade F Steel Tempered at 1260 F ........................................ 54 B.XVIII Stress Rupture Properties of A517 Grade B Steel Tempered at 1260 F........................................ 55 B.XIX Stress Rupture Properties of A517 Grade E Steel ...................... 56 B.XX Stress Rupture Properties of A517 Grade. D Steel ...................... 57 B.XXI Stress Rupture Properties of A543 Steel .................................. 58 B.XXII Stress Rupture Properties of A533 Grade A Steel ...................... 58 B.XXIII Stress Rupture Properties of A533 Grade B Steel ...................... 59 B.XXIV Stress Rupture Properties of A387 Grade E Steel ...................... 59 B.XXV Stress Rupture Properties of 2-4 Cr-i Mo Weld Metal Quenched and Tempered ........................................ 60 B.XXVI Stress Rupture Properties of 2-/ Cr-I Mo Weld Metal Stress Relieved .............. , 60 B.XXVII Selected Creep Rate Data on A542 Steel ................................. 61 B.XXVM Summary of Code Strength Criteria for A542 Class 2 Steel ..................................................... 61 B.XXIX Summary of Code Strength Criteria for A542 Class 1 Steel .................................................... 62 B.XXX Summary of Code Strength Criteria for A387 Grade D Steel Tempered to 75,000-psi Minimum UTS ................................... 63 B.XXXI Summary of Code Strength Criteria for the Remaining Low Alloy Steels Studied, Quenched and Tempered to 115,000 psi Minimum UTS .................................................. 64 iii

.i~r Illustrations e l ~Figure H 1 Comparison of Allowable Stress Criteria for A542 Class 2 Steel .6 2 Comparison of Allowable Stress Criteria for A542 Class 1 Steel .7 3 Comparison of Allowable Stress Criteria for A387 Grade D Steel .8 4 Comparison of Allowable Stress Criteria Based on Tensile Strength Data for Various Quenched and Tempered Steels ......... 9 A.l Tensile Strength of A542 Class 2 Steel . .................................

19 A.2 Yield Strength of A542 Class 2 Steel . ....................................

20 A.3 Tensile Strength of A542 Class 1 Steel . .................................

21 A.4 Yield Strength of A542 Class I Steel . ....................................

22 A.5 Tensile Strength of A387 Grade D Steel . .................................

23 A.6 Yield Strength of A387 Grade D Steel . ....................................

24 A.7 800 F Stress Rupture Properties of A542 and A387 Steels ... 25 A.8 850 F Stress Rupture Properties of A542 and A387 Steels ........... 26 A.9 900 F Stress Rupture Properties of A542 and A387 Stecls ... 27 A.10 1000 F Stress Rupture Properties of A542 and A387 Steels ......... 28 A.ll Tensile Strength of Various Quenched and Tempered Steels ........ 29 A.12 Yield Strength of Various Quenched and Tempered Steels........... 30 A.13 800 F Stress Rupture Properties of Various Quenched and Tempered Steels ............................... 31 A.14 900 F Stress Rupture Properties of Various Quenched and Tempered Steels.. 32 A.15 1000 F Stress Rupture Properties of Various Quenched and Tempered Steels .33 A.16 1100 F Stress Rupture Properties of Various Quenched and Tempered Steels .............................................. 34 A.17 1200 F Stress Rupture Properties of Various Quenched and Tempered Steels .35 A.18 Stress Versus Larson-Miller Parameter for 2-% Cr-i Mo Steels . 36 A.19 Charpy V-Notch Impact Properties of 2-4 Cr-1 Mo Steel .37 A.20 Charpy V-Notch Impact Properties of 2-Y Cr-1 Mo Steel .38 iv

IP00

'e Analysis of Data from Symposium e-at I on Heated-Treated Steels for nElevated Temperature re Service E. B. NORRIS and R. D. WYLIE Southwest Research Institute San Antonio, Texas Introduction The use of quenched and tempered low alloy steels for special purpose petro-chemical vessels operating in the intermediate temperature range of 750 to 850 F has accelerated development of data on steels which are potentially useful for such vessels. As a result of this development, a conference was held September 19-20, 1966 at the annual meeting of the Petroleum Division of the ASME in New Orleans with the specific purpose of presenting the available data on these steels.

The conference was jointly sponsored by the ASTM-ASME Joint Committee on the Effect of Temperature on the Properties of Metal and The Metals PrQperties Coun-cil. The newly organized Metals Properties Council contracted with Southwest Research Institute to review the data presented at the conference, to prepare a detailed analysis of the data for submission to the ASME Boiler Code Subcom-mittee on Properties of Metals, and to outline specific programs which may be re-quired to provide needed data for this class of steels.

The following report presents the results of the study program.

Evaluation of Data Presented at Symposium Tensile and Creep Rupture Properties The mechanical property data presented at the Symposium on Heat Treated Steels for Elevated Temperature Service were obtained on materials representing ten chemistries covered by seven ASTNI specifications, as summarized in Table I.

The tensile and stress rupture data were analyzed in a manner as consistent as possible with the methods employed by the Subgroup on Strength Properties for Steel and High Temperature Alloys of the ASME Boiler and Pressure Vessel Code. For the "elastic" range, this involves determining the tensile and yield trend curves, as a function of temperature, by lowering the average curves to the 1

I'

. Table I i.I Materials Evaluated in Symposium Papers i e Alloy Class ASTM Specifications t

Carbon steel A212-B r

% M Mo b A533-A i Mu Mo.N! modified A533-B I Cr-4 Mo A387-B 2 % Cr-I Mo A387-D*, A542 a Cr-I Mo A387-E* I I 3 Cr-'A Mo A517-E

% Nib% Cr-4 MO A517-F Ni-Cr-hlo A543 I Ni-Cr-Mo-V A508

Theac specificativne are for annealed or normalized and tempered material.

I iI minimum specified values at room temperature. In the "creep" range, the average and minimum stresses to produce rupture in 100,000 hours0 days 0 hours 0 weeks 0 months and the average stress to produce a secondary creep rate of 0.01 per cent per 1000 hours0.0116 days 0.278 hours 0.00165 weeks 3.805e-4 months are also re- II quired. However, the creep data presented at the Symposium were too limited to permit the latter calculation. The creep data will be discussed in more detail in I a later section. I The bulk of the data reported wcre on the 2-Y4 Cr-i Mo steel composition. The I chemistries and tensile properties reported showed that they met either the ASTM A542 or the ASTM A387 specification. For purposes of analysis, the data were divided into three groups or classifications as given in Table 11. For each of I these groups, the minimum ultimate strength and minimum yield strength trend I

curves were established, as shown in Figures A.1 through A.6, Appendix A. The first step was to construct an average curve for the ultimate or yield strength as i a function of temperature. Secondly, a minimum trend curve was developed by re- II I

ducing the average curve by the ratio of the minimum specified room temperature strength to the average room temperature strength. Therewas noadjustmentmade to i this average curve for the purpose of establishing design stresses. The raw data I used to construct these curves are given in Tables B.I through B.JII, Appendix B.

The stress rupture data contained in Tables B.XI through B.XIII, Appendix B, i were platted on log-log coordinates in Figures A.7 through A.10, Appendix A. In- I dividual rupture curves were constructed for each lot of material in Figures A.7 Table 11 Three Clossificotions for 2-14 Cr-l Mo DATA i

Group Minimum UTS Heat Treatment Tempering Temperature II A542-2 115,000 Q &T 1075 0 to 11250 F I A542-1 105,000 Q &T 1150 0 to 1200"F i A387-D 75,000 N &T 12500 to 13000 F I

It 2

I i

II I

_9P"N~ _- . .

through A.10. A lot was considered to consist of a material with a specific com-bination of heat of steel, heat treatment, material thickness, specimen location, etc. The log-log plots were extrapolated to 100,000 hours0 days 0 hours 0 weeks 0 months life if any individual test within a lot exceeded 1000 hours0.0116 days 0.278 hours 0.00165 weeks 3.805e-4 months and if at least one lot in a group had a test result exceeding 3000 hours0.0347 days 0.833 hours 0.00496 weeks 0.00114 months . The 100,000-hour rupture strengths so obtained are summarized in Table III.

Table III Summary of 2-% Cr-1 Mo Stress Rupture Strengths I 05-Hour Rupture Strength, ks I Group Temnp. e°F)

Minimum Average A542-2 800 84.0 97.7 850 60.0 64.0 900 32.0 38.5 A542-1 800 73.0 76.0 850 54.0 57.0 900 36.0 42.5 A387-D 800 50.0 50.0 850 __ _

900 26.0 26.0 The tensile data on the remaining materials listed in Table I were grouped together for study because they were.insufficient to permit evaluation of each alloy individually. The tensile data on these alloys are presented in Tables B.IV through B.X, Appendix B, and in Figures A.11 and A.12, Appendix A. Except for A212-B1[2] and A508[7],the tensile data agreed well with the trend curves deter-mined previously for A542-2. Also shown in Figure A.11 is a curve representing four times the allowable stresses given in Code Cases 1204 and 1298. The shape of this curve also agrees well with the 115,000-psi trend curve.

The individual stress rupture data utilized but not specifically tabulated in one paper [2] was obtained by consulting two additional references [12, 13].

These data, along with other data presented at the Symposium, are given in Tables B.XIV through B.XXVI, Appendix B, and in Figures A.13 through A.17, Appendix A. Although there is a large amount of test data, less than 13 per cent were over 1000-hours duration, and over half of all tests failed in less than 100 hours0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months . For those materials (all single heat data) on which test results in excess of 3000 hours0.0347 days 0.833 hours 0.00496 weeks 0.00114 months were reported, extrapolation to 100,000 hours0 days 0 hours 0 weeks 0 months were made and are given in Table IV.

The creep rate data reported at the conference were quite limited. Only those obtained on tests in excess of 5000 hours0.0579 days 1.389 hours 0.00827 weeks 0.0019 months were considered, Table B.XXVII, Ap-pendix B. The number of variables in this tabulation prohibits any detailed anal-ysis, and the relatively high creep rates reported do not lend themselves to extra-polation to determine stresses for a creep rate of 0.01 per cent in 1000 hours0.0116 days 0.278 hours 0.00165 weeks 3.805e-4 months .

3

Table IV Stress Rupture Strength of Several Quenched and Tempered Steels t Group Estimated 105.Hour Rupture Strength, ksl I 8000F 900 0F 1000'F A387-B-QT I 52.0 13.3 A517-E 71.0 40.0 13.0 A517-F __ 46.0 8.9 The stress rupture properties of the 2-% Cr-i Mo class of materials were also studied with the aid of the Larson-Miller parameter technique, Figure A.18. A constant of 25 was employed instead of the normally used value of 20 because a better correlation between test temperatures was obtained. Two primary points can be noted in this figure. The first is the tendency for the two lower bands to merge. The second is the indication that the quenched and tempered groups not only merge but may cross, indicating that the material starting with the higher rupture strength may be the weaker after long periods of exposure to temperature and stress.

Determination of Allowable Stresses In the "elastic" range, one of the factors for determining allowable stresses is the minimum tensile strength. Certain sections of the Code utilize one-quarter of this value, while others employ a factor of one-third. In the "creep" range, 80 per cent of the minimum and 60 per cent of the average 100,000-hour rupture strength are used along with the average stress to produce a minimum creep ratc of 0.01 per cent per 1000 hours0.0116 days 0.278 hours 0.00165 weeks 3.805e-4 months .

At the temperature of transition between the tensile-governing and the creep-governing curves, a "faired value" was taken as being equal to 75 per cent of the difference between the creep-governing value at a temperature 50 F higher and the tensile-governing value at a temperature 50 F lower. An example of this calculation is given below:

Temperature Material Governing Property and Allowable Stress 850 F A542-2 25% UTS - 23,500 psi 900 F A542-2 (To be determined) 950 F A542-2 10'-SR Strength - 11,000 psi "Faired value" of Allowable stress at 900 F

- 11,000 psi + % (23,500 psi - 11,000 psi)

= 20,400 psi The 7 5-per cent factor was chosen because it provides a smooth transition between the two curves.

The allowable stresses based on each of the above criteria are tabulated in Tables B.XXVII through B.XXXI, Appendix B, and are presented graphically in 4

_ .j

- !I !"twomp"MUMM.. -- P. V.

I - I. -- - --- 1 . -, . . a_

. .- . .. . - - - 1 . .

i II Figures 1 through 4. For A542 Classes 1 and 2, Figures 1, 2 and 3 show that based I on the 25 per cent of minimum tensile strength criteria the "elastic" range ex-tends to 850 F and that based on the 33 '/Xper cent of minimum tensile strength criteria the "elastic" range is limited to 800 F. For the remaining quenched and i

tempered alloys (except A212-B carbon steel), the tensile data supported the use I of the same 115,000-trend curve used for A542 Class 2. The stress rupture data were insufficient to reliably establish allowable stresses in the "creep" range.

Based on this analysis, the suggested allowable stresses for the four groups of materials, using 25 per cent of the minimum tensile strength, are presented in Table V. A similar summary using 33 %/ per cent of the minimum tensile strength is contained in Table VI.

Notch Toughness Considerations Except that in one paper [9], the Charpy V data presented were generally insufficient to construct transition curves. These data, presented in Figures A.19 and A.20, Appendix A, exhibit a large amount of scatter and a significant directional effect. A directional effect is also indicated by a second author [4].

Another important characteristic is the large range (30 to 80 ft-lb) reported in the drop weight NDT Charpy V correlation energy level [4, 9]. The Naval Re-seach Laboratory has also reported large variations in correlation energy levels of over 100 ft-lb absorbed energy (161. This observation raises a question con-cerning the adequacy of the Charpy V test in defining the fracture safe operating temperature range of a structure employing these higher strength materials.

Notched Rupture Characteristics Combination smooth-notched bar rupture test results [3] indicate that strain aging may produce embrittlement at elevated temperatures, as evidenced by two failures in the notch at 1000 F.

Long Time Exposure Effects The effect of long time exposures on the tensile and impact properties was studied by several investigators [4, 7, 9]. On the basis of tensile properties, the data indicated that below 1000 F there is no significant loss in tensile strength when exposed approximately a year, with or without stress. However, there appears to be an indication of embrittlement, as defined by the Charpy V test. In general, the as-quenched notch toughness of the quenched and tempered alloys are good. However, data were presented [71 which indicated that the Ni-Cr-Mo-V alloy was embrittled by aging, particularly at 800 F.

Hydrogen Environmental Studies The effects of hydrogen on Cr-Mo steels was studied [6]. Although tests were limited to the lower strength grades (annealed, normalized and tempered), a re-duction in stress-rupture strength was noted. This is of concern because the ef-fect might be more pronounced on the high strength quenched and tempered grades.

Properties of Welds The results reported on weld metal and heat affected zone properties were quite limited. One paper [2] employed a composite specimen containing a shallow 5

. .t I

F II i

I i

I I9 I

ta AOu I 140 115,000-pot iron curve from F 120 I. J00 c

'I 4x allowable stress E jCade C sea IZ04 and 1298)

D 80 .... ............ MM-.

Code-60 OASI?-E AAS17-F C3AS43 VA387-B OAS 3-A 40 +A533-B

+A508 20 V 400 quo bug 0o0 logo Temperature, IF FIG. A.11 TENSILE STRENGTH OF VARIOUS QUENCHED AND A TEMPERED STEELS I,

I i

29 II

I I

a I.;

1.1 Ii 4:.

I 400 600 800 1000 0 200 Temperature. *F AND FIG. A.12 YIELD STRENGTH OF VARIOUS QUENCHED TEMPERED STEEL.S I i- I I .

PI .1; P

il..

I!

I, I 30 I1'

- i -I - 1 ----. ---

N i *.:1i:: .::4-- A:ls@ ' -.1I.: I * ;l@

..... l Bi.

-; I*:- 1-.:

1;;.

T-.: -:;:I

. :1.1. 1:i. .1::I

1 Ia id li ,_...,

11W-d T... , 1- - - -

I .ii H-iI.

. F... 1;t; w ;tIz I:. I I.

-t t:I: :l- . ~"-

l:4;::=.:.:~I .- r e__ _- - 1..l...

l.::I-~ -.I- .t: t .I1. - I -I- 1-I- .I I- I 1- I > --I:e ~I- II-t - .-I-. I-- l.-I .-I -I -.1-4 1-a ... ... - i tt . i

'.4:1' IPEiIE: I :::I r-1'. I.. F.':1:::: .

I;14'-: I;:

1 -'-* I:;1 10' f., . I m..

Read Stres : 0.

I F:: lI+/- IEl En.,1':::, - -

. *1"A::

-1 F rr: I144 - T.,;;

I.

Ca . W a.:.II W -. 1.1. .... ..I.

I; im I g ...1-1.:-.1111 I

1..1 L1. I . I. 1n.

iI.I s~ VIMI.t

. .: ffi ::::1

.r I.. I'. .1 I OA51-fASIT-T

A517Dl I

1to4 AA517-B O ASIS V A387D-A387D-L-.-ST - I Q A53 3-A

a. iA::i

I. , ziw '14i: ,l1

'TI II il..-

tI: :.i:

.1. W .1 I

_. .4.4.

5m. 1 I-;- L Ron d Stress Len !T."'1_ . 1,4,

. j _, . 1I

..1 --.;-.1 : : :1 1 .1.

4-H

-I 7puvl:

I......

tt I

.I.f-l,-

1. 1. 1 I I

. . l§(, hl ITUI1 U I L

1.

J-,., I 1 I V I; idI:3I. l : IlI7.'l Ii-4----I--4--4--eth+t4-I il I.L? I.i1i314W 4;--'1.-T.-J: ;1 L 1:1 'I 1'. 1...-.1 ::I i .i . - I-14 1'If II}J '-l-Fl-ii INJ- I IIt182' lt-Il'- t. *-

'- N- - H1' I I I -_&ll-*@44*l Ia jot 10b Tim. for Riapta-re. Sr FIC. A.14 900 F STRESS RUPTURE PROPERTIES OF VARIOUS QUENCHED AND TEMPERED STEELS

_ I .

I I

I

-I I

, I . .1

- .e..

83 .

AC

.idlibU, W i # itsl...

,gtgl .w:

,g1!iti.

tX.gsgg ig- I l ~

l X gg lfi. a tia it A IT g1 r i M I iiirlllLi

i. *.

.- 1; Iii1an reli 1 V1) 1 Rg 4t .QJi;I R Stress .e ffit 0110 Time fee Auptere. He FIG. AJ15 1000 F STRESS RUPTURE PROPERTIES OF VARIOUS QUENCHED AND TEMPFERED STEELS tI.44+1 4. ..... t.... . .. __ . , _._44 . 4.. ........ .

-I I

I(

Table B.IX Tensile Properties of A508 Steel (Tempering Temperature 1050°F)

Source Plate Test Test 0.2% YS UTS Elong R.A.

Gouge (in.) Loc'n Temp. (OF) (ksl) (ksl) (%) (%)

7 4-% XT RT 156.0 170.0 18.0 60.0 200 148.0 160.0 15.0 55.0 400 140.0 156.0 14.0 53.0 600 134.0 150.0 16.0 59.0 800 126.0 138.0 18.0 63.0 1000 108.0 114.0 19.0 72.0 1100 84.0 90.0 21.0 81.0 1200 48.0 54.0 39.0 91.0

'See Relerences for scurce.

Table B.X Room Temperature Tensile Properties of Several Quenched and Tempered Steels Source' Alloy Plate Test 0.2% YS UTS Elong R.A.

Source Ident. Gouge (in.) Loc'n (M) O(ksl) () (%)

2 AS43 126.0 135.0 21.5 68.4 A517F 113.0 121.0 26.0 59.7 A517E 112.0 120.0 22.0 67.9 A533A 121.3 138.9 22.5 70.5 A533B 121.5 130.5 20.0 63.2

'See Refcrences for source.

t.

f h:

47 I;

.11 r

"'I-- .

Ii

.. t * - l i

I I

Table B.XXI I Stress Rupture Properties of A543 Steel I Source* Test Stress Rupture Source Temp. (OF) (psi) Life (hr) 13 800 94,000 51.0 89.000 325.0 900 80.000 16.5 70,000 350.0 64,000 360.0 60,000 600.0 50,000 1490.0 1000 70,000 1.1 60,000 21.0 50,000 51.0 45,000 85.0 35,000 112.0 30,000 307.0 25,000 585.0 20,000 1690.0 1100 40,000 1.1 30,000 14.5 20,000 86.0 15.000 307.0 1200 15,000 6.1 10,000 72.0

'See Relerences for source.

Table B.XXII Stress Rupture Properties of A533 Grade A Steel Source* Test Stress Rupture Source Temp. (0F) (psi) Life (hr) 13 900 80,000 108.0 70,000 257.0 60,000 920.0 1000 57,000 14.5 50,000 <1.0 45,000 10.5 I 40,000 38.5 35,000 5.4 30,000 114.0 25,000 266.0 20,000 725.0 l 1100 20,000 24.0 15,000 105.0

'See References for source.

58

- I -.--.- -----. .-- -- -

I t}

Table B.XXIII Stress Ruvnire Prnin.rties of I

A533 Grade B Steel II Source

Test Temp. ( 0F)

Stress (psi)

Rupture Life (hr)

IIi 3 900 80,000 8.4 80,000 15.8 i

I 1000 70,000 60,000 58,000 184.0 585.0 11.4 0

50,000 I9.2

_ 1100 15,000 114.0

'See References for source.

I 1 i

Table B.XXIV i Stress Rupture Properties of A387 Grade E Steel i Source* Plte l TITest Test l Stress Rupture Guge (in.) Loc'n Temp. (°F) I (psi) Life (hr) i i Quenched and Tempered (1125°F) 4 7-l1M/ C T 850 90.0 49.6 85.0 114.0 80.0 364.1 I 76.0 1596.0 I 900 85.0 20.6 80.0 26.6 75.0 230.0 70.0 430.8 68.0 453.8 66.0 687.7 Normalized and Tempered (1225°F) 4 7-%. 6 T 850 56.0 37.7 55.0 28.1 52.0 138.5 i 900 50.0 55.0 50.0 305.5 6.1 42.6 0 45.0 181.3 I 'See Rclerencea for source.

45.0 43.0 187.9 358.0 i

I 59

4. DOE HTGR 88383, "Tensile and Creep Properties of SA533 Grade B Class I Steel," December 1989.

A V6-DOE-IHTGR-88383 ORNTM-1133813 TENSILE AND CREEP PROPERTIES OF SA533 GRADE B CLASS 1 STEEL APPLIED TECHNOLOGY Any hurber distribution by any bolder of this document or data therein to third prtkes rePreCent forgn Interests, foign Vvermmu, fortlgn copankes, and foign suWidiaties or fortlgn divio of US.

companies shall be approved by the ssocdate Deputy Adastnt Sectary (or Reactor Ssteaes, Development ad Technology, US. Depatment of EnerW. Further foreig pay telease may require DOE approal punsuant to Federal Regulation 10 CER Part 810, and/or may be sutbed to Section 127 of the Atomic Enagy Act.

AUTHORS/CONTRACTORS Oak Ridge National Laboratory H. E. McCoy Oak Ridge National laboratory Oak Ridge, Tennessee 37831 operated by MARTIN MARIETTA ENERGY SYSTEMS, INC for the UNITED STATES DEPARTMENT OF ENERGY December 1989

DOE-HTGR-88383 ORNL/TM-11338

2. Distribution Category UC-522T Metals and Ceramics Division TENSILE AND CREEP PROPERTIES OF SA533 GRADE B CLASS 1 STEEL H. E. McCoy Date Published: December 1989 NOTICE: This document contains information of a preliminary nature. It is subject to revision or correction and therefore does not represent a final report.

Prepared for the DOE Office of Advanced Reactor Programs AF 20 10 15 2 Prepared by the OAK RIDGE NATIONAL LABORATORY Oak Ridge, Tennessee 37831-6285 operated by MARTIN MARIETTA ENERGY SYSTEMS. INC.

for the U.S. DEPARTMENT OF ENERGY under Contract DE-AC05-840R21400

CONTENTS LIST OF TABLES.. . . . . . . . . . . . . . . . . . . . . . . . . . . . v LIST OF FIGURES .......................... . vii ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 INTRODUCTION .... . . . . . . . .. 1 EXPERIMENTAL DETAILS .... . . . . . . . . . . . . . . . . . .. 2 MATERIAL .... . . . . . . . . .. 2 TEST METHODS .... . .. . . . . .. 2 EXPERIMENTAL RESULTS .... . . . . . . . 2 EFFECTS OF HEAT TREATMENT ON TENSILE PROPERTIES OF BASE METAL 2 METALLOGRAPHIC EXAMINATION OF MATERIAL . . . . . . . . . . . . . 3 MEASURED TENSILE PROPERTIES ... . . . . . . . . . . . . .. 3 MEASURED CREEP PROPERTIES ... . . . . . . . . . . . . . .. 4 TEMPER EMBRITTLEMENT .... . . . . . . . . . . . . . . . .. 6 DISCUSSION .... . . . . . . . . . . . . . . . . . . . . . .. 7

SUMMARY

.... . . . . . . . . . . . . . . . . . . . . . . . .. 7 ACKNOWLEDGMENTS .... . . . . . . . . . . . . . . . . . . . . .. 7 REFERENCES .... . . . . . . . . .. . . . . . . . . . . . . . . . 8 iii

LIST OF TABLES Table Eag 1 Summary of test materials .... . . . . . . . . . . . . . . 9 2 Summary of product chemical analyses . . . . . . . . . . . . 10 3 Specimen lot designations (modified heat number used in report) . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 4 Tensile properties specified by code and measured after various heat treatments .... . . . . . . . . . . . . 12 5 Tensile test of A533 base metal heat treated by ORNL HT 1 (Lot 1) . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 6 Tensile test of A533 base metal heat treated by ORNL HT 2 (Lot 2) ..... . . . . . . . . . . . . . . . . . . . . . . 14 7 Tensile test of A533 base metal heat treated by CE (Lot 3) . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 8 Tensile test of A533 transverse weld (HT 9583A) heat treated by CE (Lot 4) . . . . . . . . . . . . . . . . . . . . 16 9 Tensile test of A533 weld metal (HT 9583A) heat treated by CE (Lot 5) . . . . . . . . . . . . . . . . . . . . . . . . 17 10 Tensile test of A533B base metal (HT 5795) heat treated by CE (Lot 6) .18 11 Tensile properties of A533B all weld metal (HT 5795) heat treated by CE (Lot 7) .19 12 Tensile properties of A533B transverse weld (HT 5795) heat treated by CE (Lot 8) . . . . . . . . . . . . . . . . . . . . 20 13 Tensile properties of A533B base metal (HT 9583B) heat treated by CE (Lot 9) .21 14 Tensile properties of A533B transverse weld (HT 9583B) heat treated by CE (Lot 10) .22 15 Tensile properties of A533B weld metal (HT 9583B) heat treated by CE (Lot 11) . . . . . . . . . . . . . . . . . . . 23 16 Tensile properties of A533B base metal (HT 64535) heat treated by CE (Lot 12) .24 17 Creep data for A533B steel-heat 9583A . . . . . . . . . . . . 25 v

Table . *Pg 18 Creep data for A533B steel-heat 5795 . .. .. .. . .. . . 26 19 Creep data for A533B steel-heat 9583B . . . . . . . . . . . . 27 20 Creep data for A533B steel-heat 64535 . . . . . . . . . . . . 27 21 Effect of creep exposure on tensile properties . .. . . 28 vi

LIST OF FIGURES Figure Paug 1 Yield strength of several lots of A533 base metal as a function of test temperature . . . . . . . . . . . . . . . . 29 2 Yield strength of three lots of weld metal as a function of temperature .... . . . . . . . . . . . . . . . . . . . . 29 3 Yield strength of three lots of transverse weld samples as a function of temperature .... . . ..... . . .. . . . 30 4 Yield strength for lots of base metal, transverse welds, and weld metal having the lowest strength . . . . . . . . . . 30 5 Yield strength of A533 base metal at a strain rate of 3.7 x 10's- 1 as a function of temperature . . . . . . . . . . 31 6 Yield strength of A533 weld metal at a strain rate of 3.7 x 10-5s'l as a function of temperature . . . . . . . . . . 31 7 Yield strength of transverse weld specimens of A533 at a strain rate of 3.7 x l0's l as a function of temperature . 32 8 Yield strength at a strain rate of 3.7 x 10-s-1 for lots of base metal, transverse weld samples, and weld metal as a function of temperature .32 9 Tensile strength of several lots of A533 base metal as a function of temperature .. 33 10 Tensile strength of several lots of A533 weld metal specimens as a function of temperature . . . . . . . . . . . . 33 11 Tensile strength of several lots of A533 transverse weld specimens as a function of temperature . . . . . . . . . . . . 34 12 Tensile strength for lots of base metal, transverse welds, and weld metal having the lowest strength . . . . . . . . . . 34 13 The ultimate tensile strength of several lots of base metal as a function of temperature . . . . . . . . . . . . . . 35 14 The ultimate tensile strength of several lots of weld metal as a function of temperature . . . . . . . . . . . . . . 35 15 The ultimate tensile strength of several lots of transverse weld specimens as a function of temperature . . . . . . . . . 36 vii

16 -Ultimate tensile strength for lots of base metal, weld metal, and transverse weld metal having the lowest strength. . . . . . . . . . . . . . . . . . . . . . . . 36 17 Yield strength of Lot 6 measured at strain rates of 1.3 x 10-4S' and 3.7 x 106s 2 . . . . . . . . . . . . . . . 37 18 Ultimate tensile strength of Lot 6 measured at strain rate of 1.3 x 10-4s-1 and 3.7 x 10-6s'l . . . . . . . . . . . . . 37 19 Larson-Miller correlation for Lots 1, 2, and 3 base metal crept to 1% strain . . . . . . . . . . . . . . . . . . . . . . 38 20 Larson-Miller correlation for Lots 3, 6, 9, and 12 base metal crept to 1% strain . . . . . . . . . . . . . . . . . . . 38 21 Larson-Miller correlation for 1% strain showing average and minimum strengths for Lots 3, 6, 9, and 12 base metal . . 39 22 Larson-Miller correlation for tertiary creep for Lots 3, 6, 9, and 12 base metal . . . . . . . . . . . . . . . . . . . . . 39 23 Larson-Miller correlation for rupture for Lots 3, 6, 9, and 12 base metal .40 24 Time to 1% strain as a function of time to rupture for Lots 3, 6, 9, and 12 base metal . . . . . . . . . . . . . . . 40 25 Time to tertiary.creep as a function of time to rupture for Lots 3, 6, 9, and 12 base metal . . . . . . . . . . . . . . . 41 26 Fracture strain as a function of rupture time for Lots 3, 6, 9, and 12 base metal .41 27 Allowable stresses as a function of temperature for base metal based on various design criteria . . . . . . . . . . . . 42 28 Larson-Miller correlation for 1% strain in Lots 5, 7, and ll weld metal . . . . . . . . . . . . . . . . . . . . . . 42 29 Larson-Miller correlation for 1% strain in Lots 5, 7, and 11 (combined) weld metal .43 30 Larson-Miller correlation for 1% strain in Lots 4, 8, and 10 transverse weld samples . . . . . . . . . . . . . . . . 43 31 Larson-Miller correlation for 1% strain in Lots 4, 8, and 10 (combined) transverse weld samples . . . . . . . . . . 44 32 Larson-Miller plot comparing the average properties of base metal, weld metal, and transverse weld samples . . . . . 44 viii

TENSILE AND CREEP PROPERTIES OF SA533 GRADE B CLASS 1 STEEL H. E. McCoy ABSTRACT Tensile and creep tests are being performed on several lots of base metal and weldments to determine the design stresses for 1% strain in 1000 h over the temperature range of 371 to 538'C.

Short-term tensile tests indicate that the strength is least for base metal, intermediate for transverse weld specimens, and greatest for weld metal. Creep tests show much less variation, with about equivalent creep strength for base metal and trans-verse weld samples and slightly greater creep strength for weld metal. This is an interim report on a continuing program.

INTRODUCTION The Modular High-Temperature Gas-Cooled Reactor (MHTGR) concept utilizes a pressure vessel constructed of SA533 Grade B Class 1 steel. The allowable stresses given for this material in Section III (Nuclear Construction) of the ASHE Boller and Pressure Vessel Code are for tempera-tures to 371'C (700'F).1 It is anticipated that, in the operation of MHTGR's, Level C and D events may occur that last for a total of less than 1000 h in which the vessel temperature will be in the range of 371 to 538'C (700 to 1000'F). The purpose of the current testing is to determine the mechanical properties of this steel over the temperature range 371 to 5931C (700 to 1100'F) needed to support code approval for use of this steel under the Level C and D conditions.

The approach taken to this problem is that of measuring the tensile and creep properties of this steel over the temperature range of 317 to 5939C (700 to llOO1F). The construction will involve two basic welding processes, namely, tandem electrode submerged arc (machine) and shielded metal arc (manual). Hence, samples of welds made by these processes are included. The test materials currently being evaluated include three heats of base metal, two submerged arc welds, and one shielded metal arc weld.

The current status of the testing is described in this report.

Research sponsored by the Office of Advanced Reactor Programs, Division of HTGRs, U.S. Department of Energy, under contract DE-AC05-84OR21400 with Martin Marietta Energy Systems. Inc.

1

2 EXPERIHENTAL DETAILS MATERIAL The materials were procured from Combustion Engineering (CE),

Chattanooga, Tenn., and information about the test materials is summarized in Table 1. All of the materials underwent the vendor chemical analysis and an analysis by CE; the results of these tests are shown in Table 2, The tests showed excellent agreement, and the material seemed to be of the composition specified by the ASHE Boiler and Pressure Vessel Code. 2 The type of specimen, the associated heat treatment, and vendor designation are given in Table 3. The lot numbers shown in the last column of Table 3 will be used throughout this report.

As will be discussed further, all three heat treatments resulted in mechanical properties that satisfied the code requirements: ultimate tensile strength in the range of 550-690 MPa (80-100 ksi), the minimum yield strength of 345 MPa (50 ksi), and the minimum elongation in 50.8 mm (2 in.) of 18.0%.

TEST METHODS The tensile tests were run in accordance with ASTM E8, and the creep tests were run according to ASTH E139.s The specimen had a gage section 6.35 mm (0.25 in.) in diameter by 31.75 mm (1.25 in.) long. Extensometers were attached by set screws into small grooves outside the gage section so that the set screws would not induce rupture.

EXPERIMENTAL RESULTS EFFECTS OF HEAT TREATMENT ON TENSILE PROPERTIES OF BASE METAL The range of properties allowed by the code for this material is quite broad. The first piece of base material obtained from CE did not have the desired heat treatment. Individual test specimens were given the first heat treatment listed in Table 4, referred to as ORNL HT 1. The ultimate tensile strength of the material given this heat treatment is at the top of the code-specified range, with duplicate tests having values of 705 and 667 MPa (102.2 and 96.8 ksi). The heat treatment used by CE is the last one listed in Table 4. The ultimate tensile strength of the base metal is near the middle of the allowable range, and the other mechanical properties are acceptable.

The next heat treatments investigated were based on the premise that the most conservative results would be obtained by heat treating to the lower side of the allowable strength band. The material was solution annealed at 871-C, water quenched, and tempered at 663*C. After tempering for 30 h, the ultimate tensile strength was barely above the code minimum value. Tempering for 50 h at 6630C reduced the ultimate tensile strength considerably below the code minimum value. Thus, the heat treatment of

3 solution annealing at 871-C, water quenching, and tempering 30 h at 663'C was selected as ORNL HT 2. The mechanical properties of heat 9583 were evaluated following three heat treatments: ORNL HT 1 (Lot 1), ORNL HT 2 (Lot 2), and CE HT (Lot 3).

HETAIIOGRAPHIC EXAMINATION OF MATERIAL Not all of the materials have been evaluated metallographically. The primary microconstituent is tempered bainite. The hardness of the base metal is 90 to 92 Rockwell B. The weld deposit is slightly harder with a hardness of 95 Rockwell B. There is a region in the base metal a short distance from the fusion line that is tempered during welding to a hardness of about 88 Rockwell B. Samples that transversed the fusion line (noted transverse weld samples) usually failed at this weaker location.

MEASURED TENSILE PROPERTIES Tensile properties of the following were measured: A533 base metal, the deposited weld metal, and transverse across the fusion line. In the transverse specimens, the fusion line was midway along the gage length; weld metal is on one side, and base metal on the other. The test matrix consisted of test temperatures of 25, 317, 427, 482, 538, and 593'C.

Specimens were tested at initial strain rates of 1.3 x 10-4 and 0.27 x 10-4s1. In general, duplicate tests were run at each condition, but the availability of material was a limitation in a few cases.

The tensile test results for Lots 1 through 12 (see Table 3) are given in Tables 5 through 16. The data from these tables were used to construct Figs. 1 through 18. The yield strengths of the various lots of base material are shown in Fig. 1. The data for each lot were fit with a second-order polynomial, and the lines developed in this way are shown in Fig. 1. The Lot 1 material is the strongest, the Lot 2 material is second weakest, and Lot 6 is the weakest material. Lots 3, 6, 9, and 12 were given the standard CE heat treatment. The yield strengths of these four lots agree well at temperatures above 300'C, but there is considerable scatter at 25*C. The spread in the yield strengths of weld metal are shown in Fig. 2. Lots 5 and 7 were submerged arc welds, but their properties fall at the extremes. Lot 11 was a shielded metal arc weld, and its strength is close to that of submerged arc Lot 7. There is no obvious explanation for this variation in properties. The yield strengths of the three lots of transverse weld specimens are shown in Fig. 3. There is some variation in strength between the three lots, and the strength order is not the same as that shown in Fig. 1 for the base metal.

The minimum yield strength curves are shown in Fig. 4 for the 12 lots of material. These data have not been treated in any special way but are the lowest curves selected from Figs. 1, 2, and 3. The weakest weld metal was Lot 5, and the weakest transverse weld samples were from Lot 4.

Comparison of the minimum curves in Fig. 4 shows that the order of yield strengths from least to greatest is base metal, transverse weld samples, and weld metal samples.

4 Curves are shown in Figs. 5, 6, and 7 for the yield strength deter-mined at the slower strain rate (0.27 x 10-4s'1 ). Curves are shown in Fig. 5 for the base metal, Fig. 6 for the weld metal, and Fig. 7 for the transverse weld specimens. The relative positions of the curves are similar to those noted in Figs. 1, 2, and 3 at the higher strain rate of 1.3 x 10 4s'1. The minimum curves from Figs. 5, 6, and 7 are shown in Fig. 8. The base metal (Lot 6) is the weakest, the transverse weld samples (Lot 8) are next highest, and the weld metal (Lot 5) is the strongest.

The ultimate tensile strengths of the various lots of base metal are shown in Fig. 9. Lot 1, which was heat treated to obtain the maximum allowable strength, is significantly stronger, but the other lots of material fall in a rather narrow band. The tensile strengths of the three lots of weld metal are shown in Fig. 10. The strengths of the two sub-merged arc welds fall at the extremes, and the shielded-metal arc weld has intermediate strength. The transverse weld sample properties are shown in Fig. 11. The spread in strengths between the three lots of material is very small.

The minimum ultimate tensile strength curves at the higher strain rate of 1.3 x 10-4s-1 are shown in Fig. 12. The spread in strength is rather small, but Lot 6 base metal has the lowest strength, Lot 5 weld metal has the highest strength, and Lot 8 transverse weld specimens have intermediate strength.

The ultimate strength curves at the lower strain rate of 0.27 x 104s-1 are shown in Fig. 13 for the base metal. Lot 1 has significantly higher properties, but the properties of the other base metals fall in a rather narrow band. The ultimate tensile strengths of the two submerged arc welds are shown in Fig. 14. There is a variation in strength of about 10% between the two lots of weld metal. The ultimate tensile strengths of the three lots of transverse weld specimens are shown in Fig. 15. The spread in strengths is quite small.

The minimum ultimate tensile strength curves at the lower strain rate of 0.27 x 10-4s-1 are shown in Fig. 16. The variation in strengths is small, but the order from weakest to strongest is base metal Lot 6, transverse weld samples Lot 4, and weld samples Lot 5.

The effect of strain rate on the yield and ultimate tensile strengths for Lot 6 base metal is shown in Figs. 17 and 18. The peak in strength predicted by the curves near 200'C is not real but is a result of the lack of data between 25 and 371'C. The yield and ultimate tensile strengths show an effect of strain rate at test temperatures of 538 and 593-C; there is no systematic influence of strain rate at lower temperatures.

The fracture elongation of all samples exceeded the 18% minimum required by the code. (See Tables 5 through 16 for exact values.) The reduction of area was about 75% for most samples.

MEASURED CREEP PROPERTIES The creep results obtained thus far are summarized in Tables 17 through 20. A "D" after the discontinued time in the `T-R (time to rupture)" column indicates that the test was discontinued prior to rupture.

Tests that do not have a reduction in area noted in the last column of each I

5 table are still in progress. The creep results indicate, as did the tensile properties, that the order of strengths from least to greatest is the base metal, the transverse weld specimens, and the weld metal.

The time to 1% strain for Lots 1, 2, and 3 is correlated by use of the Larson-Miller parameter in Fig. 19. Lot 1 was heat treated by ORNL to the maximum allowable tensile strength, Lot 2 was heat treated by ORNL to the lowest allowable tensile strength, and Lot 3 was given the standard heat treatment by CE (Table 3). Lots 1, 2, and 3 were from the same heat of material. The line in Fig. 19 was fit to the Lot 3 data with a.second-order polynomial. The Lot 1 data points, except one, fall above the line, indicating higher creep strength for Lot 1. The data for Lots 2 and 3 are interspersed, and there appears to be little, if any, difference in the creep strengths of the two lots.

The data for the base metal heat treated by CE (Lots 3, 6, 9, and 12) were analyzed as a group. There were 34 data points for the time to 1% strain. Similar analyses will be performed for the transverse weld and weld metal samples when more data become available; however, the design stresses probably will be determined by the properties of the base metal.

The data for the four lots of base metal (all receiving the CE standard heat treatment) are shown in Fig. 20. The data were fit with a second-order polynomial on a Larson-Miller plot to give the average properties.

These same data were treated statistically to obtain the minimum properties using the premise that the minimum properties were less than the average properties by 1.65 times the standard deviation. The data in Fig. 20 were fit on the basis of the stress being the independent variable and the Larson-Miller Parameter (K/1000) (20 + log t) being the dependent variable, where K is the absolute temperature in degrees Kelvin and t is the time in hours. The standard deviation was determined to be 0.7453, and this value was used to establish the curve for minimum properties in Fig. 21. The crossover at the left side of the figure does not have any physical significance.

Because the test program has emphasized good definition of the stress-temperature-time relationships to low strains, most of the tests have not been taken to rupture. However, the rupture and tertiary creep data for Lots 3, 6, 9, and 12 were used to obtain estimates of the average proper-ties. Eighteen data points were available for tertiary creep (intersection of line parallel and offset 0.2% from minimum creep rate with creep curve),

and these are shown in Fig. 22. There were nine data points for rupture, and these are shown in Fig. 23.

The time to 1% strain is shown in Fig. 24 as a function of the time to rupture. The data are not distributed well enough to obtain a very accurate correlation, but they are approximated by the line shown in Fig. 24, represented by the equation 0

(T-1) - 0.688(T-R)°874 A similar correlation is shown in Fig. 25 for the time to tertiary creep as a function of the time to rupture. The line shown in Fig. 25 is repre-sented by the equation (T-t) - 1.48(T-R)0 J3

6 The symbols all have units of time; specifically, T-1 is the time to 1% strain, T-t-is the time to the beginning of tertiary creep, and T-R is the time to rupture.

The fracture strain of the base metal samples is very dependent on test temperature, with higher temperatures favoring higher strains. The fracture strains from 10 base metal tests are shown as a function of rupture time in Fig. 26. The horizontal line is at 18% strain, and two tests fall at this minimum value; all other tests failed at higher strains.

Even the 18% value is quite high compared with the 1% limit imposed by design.

Correlations were developed for base metal on the basis of 1% strain (average strength), 1% strain (minimum strength), and tertiary creep and rupture strengths based on average properties. These correlations were used to obtain allowable stresses at various temperatures based on a design life of 1000 h. These data are summarized in Fig. 27 for the four lots of base metal (Lots 3, 6, 9, and 12). The order of design stress at a given temperature from highest to lowest stress is based on rupture (average),

tertiary creep (average), 1% strain (minimum), and 1% strain (average).

The three curves for the three lots of weld metal are shown in Fig. 28. The data points for the three lots fall rather close together, so they were fit as a single set of data (Fig. 29). The three individual curves for the three lots of transverse weld specimens are shown in Fig. 30. The spread between the three lots is quite small, so these data were combined as a single set to obtain the correlation shown in Fig. 31.

The comparative strengths of base, transverse weld, and weld metal specimens are shown in Fig. 32. The base metal and the transverse weld specimens have about the same creep strengths. The weld metal is slightly stronger.

TEMPER EHBRITTLEMENT Most of the temper embrittlement studies of this alloy show that embrittlement is associated with the enrichment of grain boundaries with impurities such as phosphorus during welding. The worst embrittlement was noted in the heat-affected zone.4'7 The heats of commercial material currently being evaluated are quite low in residual impurities such as phosphorus (Table 2), so it is not very likely that commercial heats of this material produced to the chemical specifications for nuclear use in Section II of the code will be susceptible to temper embrittlement.

However, a small program is being carried out to demonstrate this point.

In the meantime, it was felt that tensile tests at 25'C on creep samples discontinued after a few percent strain would reveal any significant embrittlement.

The results of tensile tests at 25'C on samples from discontinued creep tests are shown in Table 21. The creep tests involved exposures up to 7292 h to temperatures ranging from 371 to 593'C. The tensile results for samples that had only been heat treated before tensile testing are given in Table 21 for comparison. The ductilities (elongation and reduction of area) are higher for the samples creep tested before tensile testing than those tensile tested without the creep test history. Thus,

7 there is no indication of temper embrittlement in these materials based on these tests.

DISCUSSION The data available are sufficient for only part of the analytical analyses necessary to determine allowable stresses at various temperatures. Numerous tests are in progress, and the data base will increase markedly over the next few months.

The tensile data indicate significant differences in strength, with base metal having the lowest, weld metal having the highest, and transverse weld metal samples falling in between. Under creep conditions, the spread in strengths is much smaller than in short-term tensile tests. Base metal and transverse weld samples have equivalent strengths, but weld metal is slightly stronger.

The Larson-Killer parameter with a second-order polynomial fit of the data has been used extensively in analyzing the data. This method has appeared satisfactory, but the data at parameter values of about 20 do not fit well. The data in this region will determine the design stress at 538'C, so it may be necessary to alter the analytical methods being used.

SUCKARY Three lots of SA533 Grade B Class 1 plate were procured and are being evaluated. Two submerged arc welds and one shielded-metal arc weld were also procured for evaluation. Tensile tests were run on all materials at 25, 371, 427, 482, 538, and 593'C at two strain rates. These tests indicate that the short-term tensile strength varies appreciably, with base metal having the lowest strength and weld metal the highest strength.

Transverse weld samples have intermediate strength. Numerous creep tests are complete on these materials, others are in progress, and still others remain to be started. The indication thus far is that the spread in strengths is smaller than noted in short-term tensile tests. Base metal and transverse weld samples have equivalent creep strength, and weld metal is slightly stronger.

Creep samples tested to fracture have fracture strains in excess of 18%. Several discontinued creep samples were subjected to short-term tension tests, and there was no evidence of embrittlement as a result of the creep exposure. A program to more systematically evaluate whether temper embrittlement occurs in this material is in progress.

ACKNOWLEDGMENTS Research is sponsored by the Office of Advanced Reactor Programs, Division of HTGRs, U.S. Department of Energy, under contract DE-AC05-840R21400 with Martin Marietta Energy Systems, Inc.

8 The author is grateful to P. L. Rittenhouse for programmatic direction and to C. R. Brinkman for technical assistance. Nancy Cole, CE, Chattanooga, Tenn., was very helpful in providing the test materials.

D. F. Wilson and D. J. Alexander reviewed the manuscript and provided several helpful suggestions. Several technicians provided their expertise, including R. H. Baldwin, T. P. Kirkland, and H. D. Upton. The draft was typed by C. Erickson, and the final manuscript was prepared by B. Q.

Bridges.

REFERENCES

1. ASME Boiler and Pressure Vessel Code,Section III, July 1, 1986, American Society of Mechanical Engineers.

2. ASHE Boiler and Pressure Vessel Code,Section II, July 1, 1986, American Society of Mechanical Engineers.

3. "Metals - Mechanical Testing; Elevated and Low-Temperature Tests,"

in 1984 Annual Book of ASTlI Standards, Vol. 03.01, American Society for Testing Materials.

4. Masayoshi Haseguawa, Nobuya Nakajima, Nobuharw Kusunoki, and Kazuhiro Suzuki, "Effects of Copper and Phosphorus on Temper Embrittlement of Mn-Mo-Ni Low Alloy Steel (ASTM A533-B)," Trans. Jpn. Inst. Met. 16, 641-46 (1975).

5. S. C. Druce and B. C. Edwards, "On the Temper Embrittlement of Manganese-Molybdenum-Nickel Steels," Nucl. Technol. 55, 487-98 (November 1981).

6. W. A. Logsdon, "The Influence of Long-Time Stress Relief Treatments on the Dynamic Fracture Toughness Properties of ASME SA508 CQ 2a and ASME SA533 Cr B Cl 2 Pressure Vessel Steels," J. Hater. Energy Sys.

3(4), 39-50 (March 1982).

7. S. G. Druce, G. Gage, and G. Jordan, "Effect of Aging on Properties of Pressure Vessel Steels," Acta Recall. 34(4), 641-52 (1986).

I

9 Table 1. Summary of test materials Heat Vendor Plate thickness Evaluated as8 number cm in. BR SA weld SMA weld D9583 Lukens 8.9 3.5 X X X C5975 Lukens 24.9 9.625 X X 64535-1 Marrel Freres 24.5 9.625 X aBM - base metal, SA - submerged arc weldment, and SMA - shielded metal arc weldment.

Table 2. Summary of product chemical analyses Heat Element Code 8 specified D9583bb D95 8 3b D9583c 6 4 53 5 -lb 6 4 5 35 - 1 b c5795b c5 7 9 5b C5795c D9 5 83 d (Lukens) (CE) (CE) (M. Freres) (CE) (Lukens) (CE) (CE) (CE)

C 0.25 (max) 0.19 0.18 0.11 0.22 0.21 0.20 0.22 0.13 0.10.

Hn 1.07-1.62 1.31 1.27 1.41 1.45 1.43 1.45 1.44 1.53 1.27 P 0.015 (max) 0.012 0.010 0.008 0.009 0.006 0.01 0.008 0.009 0.008 S 0.018 (max) 0.003 0.002 0.006 0.006 0.004 0.015 0.021 0.015 0.014 Cu 0.12 (max) 0.08 0.04 0.04 0.04 0.07 0.06 0.04 V 0.06 0.004 0.003 <0.005 0.001 0.003 0.004 0.008 St 0.13-0.45 0.22 0.21 0.45 0.19 0.21 0.23 0.25 0.45 0.35 Ho 0.41-0.64 0.57 0.52 0.55 0.52 0.51 0.54 0.58 0.57 0.48 NI 0.37-0.73 0.69 0.76 0.13 0.62 0.62 0.65 0.66 0.12 0.03 Cr 0.10 0.06 0.04 0.08 0.14 0.03 Cb, Ti. U <0.01 <0.01 <0.01 <0.01 <0.01 <0.029 Co 0.01 .013 0.014 0.016 0.011 0.005 Al 0.026 0.01 0.019 0.022 0.011 0.005 B 0.001 <0.001 <0.001 <0.001 0.001 <0.001 '-a As. Sn 0.005 <0.01 <0.020 <0.007 <0.007 <0.005 0 Zr <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 N 0.006 0.01 0.006

'ASHE. Section 11. Standard Chemical Requirements plus Special Reactor Beltline Requirements for Cu. P. S. and V.

bBase metal.

cUeld metal deposited by submerged are process.

d4 eld metal deposited by the shielded-metal arc process.

11 Table 3. Specimen lot designations (modified heat number used in report)

Vendor Modified Heat Lot heat heat Forma treatmentb number number number D9583 9583A BM ORNL HT 1 1 D9583 9583A BM ORNL HT 2 2 D9583 9583A BM CE-STD 3 D9583 9583A TW-SA CE-STD 4 D9583 9583A WM-SA CE-STD 5 C5795 5795 BM CE-STD 6 C5795 5795 WM-SA CE-STD 7 C5795 5795 TW-SA CE-STD 8 D9583 9583B BM CE-STD 9 D9583 9583B TW-SMA CE-STD 10 D9583 9583B WM-SMA CE-STD 11 64535-1 64535 BM CE-STD 12 8BM - base metal specimen; TU - transverse weld specimen, fusion line in specimen center; WM - weld metal specimen; SA - weld made by the submerged arc process; and SMA - weld made by shielded metal arc process.

bORNL HT 1 - 1 h/871'C/WQ/4 h/663'C; ORNL HT 2 -

1 h/871-C/IQ/30 h/663-C; and CE-STD - 2.5 h/871VC/WQ/

2.5 h/663'C/20 h/607-C.

Table 4. Tensile properties specified by code and measured after various heat treatments Yield strength Ultimate tensile strength Elongation MPa ksi MPa ksi (%)

Code-specified 345 min 50 min 552-689 80-100 18 min Heat-treated 1 h/871-C/4 h/663-Ca 1641 93.0 [705 102.2 26.7 (1 h/1600'F/4 h/1225-F) 1599 86.9 1667 96.8 130.0 1 h/871-C/30 h/663-Cb 492 71.3 578 83.8 33.0 (I h/1600-F/30 h/1225F) 1 h/871VC/50 h/663-C 426 61.8 518 75.2c 33.7 (I h/1600'F/50 h/1225F) 2.5 h/871VC/2.5 h/663'C/ 492 71.3 621 90.1 35.0 20 h/607*Cd (2.5 h/1600'F/2.5 h/1225F/

20 h/1125-F) 8 0RNL HT 1.

bORNL HT 2.

CBelow code minimum.

dCE HT.

I : It Table 5. Tensile test of A533 base metal heat treated by ORNL HT 1 (Lot 1)

Test Strain Hodulus Yield Tensile Total Uniform Reduction Specimen temperature rate strength strength elongation elongation of area

.C *F 10's l0'min' GPa 106 psi MPa ksi HPa ksi (t) (t) ()

504 24 75 1.3 8.0 210 30.5 599 86.9 668 96.8 29.95 9.98 76.47 526 24 75 1.3 5.0 231 33.5 641 93.0 705 102.2 26.72 8.05 74.88 529 24 75 1.3 8.0 224 32.4 621 90.1 707 102.5 25.70 7.30 74.96 505 24 75 0.27 1.6 215 31.2 604 87.6 676 98.0 29.30 8.24 76.68 527 24 75 0.27 1.6 228 33.0 641 93.0 700 101.5 26.21 7.99 74.54 530 24 75 0.27 1.6 211 30.7 649 94.1 707 102.6 26.50 7.87 74.58 506 371 700 1.3 8.0 193 28.0 531 77.0 636 92.2 24.60 6.85 74.60 508 371 700 1.3 8.0 179 25.9 497 72.1 609 88.3 25.30 6.93 76.76 507 371 700 0.27 1.6 114 16.5 525 76.2 629 91.3 23.39 4.79 74.48 509 371 700 0.27 1.6 165 23.9 488 70.8 607 88.0 28.53 7.05 76.57 519 427 800 1.3 8.0 159 23.0 462 67.0 562 81.6 27.78 4.74 78.70 522 427 800 1.3 8.0 162 23.5 476 69.0 569 82.6 24.80 5.59 78.57 516 427 800 0.27 1.6 153 22.3 482 69.9 559 81.0 23.50 3.31 77.71 La 517 427 800 0.27 1.6 169 24.5 478 69.3 549 79.6 26.40 3.96 76.94 523 482 900 1.3 8.0 189 27.4 434 62.9 489 70.9 30.25 4.32 83.23 525 482 900 1.3 8.0 224 32.5 454 65.9 502 72.8 24.48 2.15 81.82 520 482 900 0.27 1.6 241 35.0 427 61.9 483 70.1 33.60 1.91 82.47 521 482 900 0.27 1.6 172 24.9 446 64.7 486 70.6 25.58 1.72 82.39 510 538 1000 1.3 8.0 151 21.9 376 54.6 411 59.6 33.51 1.40 85.58 512 538 1000 1.3 8.0 123 17.9 389 56.4 410 59.5 32.38 0.89 83.48 511 538 1000 0.27 1.6 107 15.6 331 48.1 368 53.4 46.51 1.16 84.20 513 538 1000 0.27 1.6 88 12.8 348 50.5 376 54.5 54.45 1.19 83.03 538 593 1100 1.3 8.0 100 14.6 286 41.5 307 44.5 57.15 1.62 82.30 539 593 1100 0.27 1.6 88 12.7 177 25.7 236 34.2 61.40 0.70 80.30

Table 6. Tensile test of A533 base metal heat treated by ORNL UT 2 (Lot 2)

Test Strain Modulus Yield Tensile Total Uniform Reduction Spetimen temperature rate strength strength elongation elongation of area

C *F 10-4s- 103min" CPa 106 psi MPa ksi HPa ksi (%) (t) ( )

905 24 75 6.7 40.0 185 26.9 491 71.3 578 83.8 33.00 11.54 78.56 910 24 75 6.7 40.0 210 30.5 478 69.3 597 86.6 32.10 12.38 77.09 911 24 75 6.7 40.0 218 31.6 470 68.2 593 86.0 32.25 11.47 77.67 920 24 75 0.27 1.6 214 31.0 485 70.3 582 84.4 31.10 11.16 77.84 912 371 700 6.7 40.0 189 27.3 393 57.0 551 80.0 35.72 11.30 78.61 921 371 700 0.27 1.6 216 31.3 370 53.7 539 78.1 30.11 9.60 74.96 913 427 800 6.7 40.0 189 27.4 371 53.8 495 71.8 32.55 8.84 79.02 922 427 800 0.27 1.6 144 20.9 357 51.8 479 69.5 32.50 7.66 80.14 914 482 900 6.7 40.0 193 27.9 346 50.2 419 60.8 32.28 6.31 82.13 923 482 900 0.27 1.6 167 24.2 329 47.7 392 56.8 36.25 4.15 84.21 915 538 1000 6.7 40.0 157 22.8 322 46.7 357 51.7 41.70 3.19 87.01 t-S 924 538 1000 0.27 1.6 125 18.1 283 41.0 314 45.5 47.70 1.39 77.55 916 593 1100 6.7 40.0 132 19.2 254 36.9 282 40.9 59.25 1.19 89.89 925 593 1100 0.27 1.6 84 12.2 174 25.2 220 31.9 62.00 1.23 88.15

I{

Table 7. Tensile test of A533 base metal heat treated by CE (Lot 3)

Yield Tensile Test Strain Yield Tensile Total Uniform Reduction Modulus rate strength strength elongation of area Specimen temperature elongation CPa 104 psi C" (C) (M)

C F 10"-Is lO0"min', HPa ksI HP& ksi 600 24 75 1.3 8.0 238 34.5 492 71.3 624 90.5 31.65 9.20 73.93 601 24 75 1.3 8.0 199 28.8 491 71.2 619 89.7 29.32 9.73 75.14 602 24 75 0.27 1.6 255 37.0 478 69.3 604 87.6 29.66 9.20 74.58 603 24 75 0.27 1.6 208 30.2 562 81.5 663 96.1 29.30 a.47 75.61 604 371 700 1.3 8.0 191 27.7 407 59.1 547 79.3 29.59 8.30 77.34 605 371 700 1.3 8.0 221 32.1 413 59.9 553 80.2 28.35 8.82 76.51 606 371 700 0.27 1.6 201 29.2 461 66.9 574 83.2 26.18 5.67 75.55 607 371 700 0.27 1.6 199 28.9 410 59.4 562 81.6 29.00 9.45 77.26 609 427 800 1.3 8.0 192 27.8 392 56.8 494 71.6 29.75 6.72 80.47 610 427 800 1.3 8.0 213 30.8 427 61.9 513 74.4 27.10 5.25 81.37 611 427 800 0.27 1.6 136 19.7 346 50.2 508 73.7 28.60 6.82 81.22 612 427 800 0.27 1.6 183 26.6 438 63.5 510 74.0 18.00 3.60 80.43 on 613 482 900 1.3 8.0 125 18.1 334 48.5 453 65.6 30.25 4.69 84.33 614 482 900 1.3 8.0 165 24.0 378 54.8 435 63.1 30.28 3.48 83.50 615 482 900 0.27 1.6 127 18.4 362 52.5 415 60.2 33.40 6.14 82.55 616 482 900 0.27 1.6 183 26.5 368 53.4 414 60.2 36.15 3.06 85.39 617 538 1000 1.3 8.0 146 21.2 334 48.5 359 52.1 36.58 2.12 86.01 618 538 1000 1.3 8.0 161 23.4 334 48.3 363 52.6 35.45 1.72 84.95 619 538 1000 1.27 1.6 82 12.0 317 46.0 331 48.0 41.90 1.60 84.06 620 538 1000 0.27 1.6 119 17.3 321 46.5 341 49.4 42.40 1.15 84.80 630 593 1100 1.3 8.0 80 11.5 264 38.3 281 40.8 37.65 0.86 78.92 631 593 1100 1.3 8.0 110 16.0 257 37.3 282 40.9 66.42 0.78 87.91 632 593 1100 0.27 1.6 61 8.9 212 30.8 240 34.8 55.65 1.13 73.50 633 593 1100 0.27 1.6 47 6.9 209 30.3 240 34.8 45.08 1.09 71.34

Table 8. Tensile test of A533 transverse 'weld (HT 9583A) heat treated by CE (Lot 4)

Test Strain Modulus Yield Tensile Total Uniform Reduction Specimen temperature rate strength strength elongation elongation of area

F 10 s' 10'tmin'I 1C CPa 106 psi MPa ksi HPa ksi (%) (%)

700T 24 75 1.3 8.0 213 30.9 519 75.3 621 90.0 26.50 8.98 77.93 701T 24 75 1.3 8.0 227 32.9 511 74.1 622 90.2 24.75 6.69 75.04 702T 24 75 0.27 1.6 227 32.9 505 73.3 609 88.4 20.02 7.55 66.56 703T 24 75 0.27 1.6 231 33.5 509 73.9 616 89.3 26.78 8.83 67.76 704T 371 700 1.3 8.0 190 27.5 430 62.3 572 83.0 22.35 8.60 65.75 705S 371 700 1.3 8.0 245 35.5 424 61.5 565 82.0 27.18 8.21 77.23 706T 371 700 0.27 1.6 201 29.2 422 61.2 561 81.4 25.00 6.30 78.17 707T 371 700 0.27 1.6 195 28.3 429 62.2 564 81.8 21.49 6.02 75.87 709T 427 800 1.3 8.0 173 25.1 422 61.3 516 74.8 23.09 4.72 78.31 710T 427 800 1.3 8.0 188 27.3 406 58.9 SOS 73.2 24.10 4.60 79.94 711T 427 800 0.27 1.6 196 28.4 416 60.3 509 73.9 18.30 4.06 78.66 712T 427 800 0.27 1.6 233 33.7 405 58.7 498 72.2 24.18 3.95 80.49 I.'

714T 482 900 1.3 8.0 180 26.1 376 54.5 445 64.6 21.50 2.62 81.52 715T 482 900 1.3 8.0 160 23.3 353 51.1 447 64.8 26.00 3.09 83.84 716T 482 900 0.27 1.6 159 23.0 383 55.6 428 62.0 22.12 2.01 80.68 717T 482 900 0.27 1.6 136 19.8 368 53.4 426 61.7 20.50 2.19 82.29 71ST 538 1000 1.3 8.0 176 25.6 334 48.5 376 54.5 24.55 1.46 83.52 719T 538 1000 1.3 8.0 141 20.4 341 49.4 377 54.6 21.32 1.38 82.60 720T 538 1000 0.27 1.6 179 25.9 335 48.6 351 50.9 25.05 0.87 81.23 721T 538 1000 0.27 1.6 149 21.5 328 47.6 349 50.6 25.90 1.15 84.16 724T 593 1100 1.3 8.0 134 19.5 258 37.5 293 42.5 26.90 0.28 84.61 725 593 1100 1.3 8.0 103 14.9 268 38.9 285 41.4 29.33 0.75 86.79 726T 593 1100 0.27 1.6 83 12.3 205 29.7 235 34.0 25.10 1.01 81.51 727T 593 1100 0.27 1.6 71 10.3 229 33.2 253 36.7 28.18 0.95 80.59 I .

t-Table 9. Tensile test of A533 weld metal (HT 9583A) heat treated by CE (Lot 5)

Test Strain Modulus Yield Tensile Total Uniform Reduction Specimen temperature rate strength strength elongation elongation of area

C *F 10-4 s-1 10 3min't CPa 106 psi HPa ksi MPa ksi a M M 800 24 75 1.3 8.0 213 30.9 517 74.9 615 89.2 28.31 9.57 66.42 801 24 75 1.3 8.0 211 30.6 517 74.9 621 90.0 28.99 10.15 66.41 802 24 75 0.27 1.6 221 32.0 500 72.5 607 88.0 28.80 9.55 66.96 803 24 75 0.27 1.6 231 33.6 500 72.5 608 88.2 28.99 9.59 67.47 804 371 700 1.3 8.0 218 31.7 432 62.6 571 82.8 25.61 8.70 67.04 805 371 700 1.3 8.0 203 29.5 459 66.5 579 84.0 25.80 8.03 66.74 806 371 700 0.27 1.6 215 31.2 438 63.6 573 83.1 26.35 7.41 66.61 807 371 700 0.27 1.6 221 32.1 428 62.0 570 82.7 25.00 8.33 67.62 808 427 800 1.3 8.0 207 30.1 423 61.4 530 76.9 26.40 6.54 69.43 809 427 800 1.3 8.0 176 25.5 411 59.7 520 75.5 25.88 7.61 68.24 810 427 800 0.27 1.6 221 32.1 418 60.6 525 76.2 24.51 5.94 65.74 811 427 800 0.27 1.6 211 30.6 425 61.6 532 77.1 26.40 6.38 65.79 w-

-.1 814 482 900 1.3 8.0 150 21.7 388 56.2 474 68.8 28.80 5.48 71.44 815 482 900 1.3 8.0 174 25.2 372 53.9 469 68.0 29.39 6.21 72.51 816 482 900 0.27 1.6 228 33.1 383 55.5 456 66.2 31.70 5.57 72.59 817 482 900 0.27 1.6 211 30.7 391 56.7 460 66.8 31.40 4.65 72.10 819 538 1000 1.3 8.0 215 31.2 372 53.9 408 59.2 34.49 2.53 77.79 820 538 1000 1.3 8.0 155 22.5 376 54.6 411 59.6 40.02 2.39 77.76 821 538 1000 0.27 1.6 164 23.8 339 49.1 374 54.2 30.35 2.15 73.82 822 538 1000 0.27 1.6 159 23.0 344 49.9 371 53.8 36.10 2.52 76.13 828 593 1100 1.3 8.0 102 14.8 308 44.7 319 46.2 56.12 0.99 81.21 829 593 1100 1.3 8.0 94 13.6 312 45.2 321 46.6 49.85 1.08 81.99 830 593 1100 0.27 1.6 151 21.9 254 36.8 281 40.8 56.35 1.07 63.61 831 593 1100 0.27 1.6 .121 17.5 264 38.2 276 40.0 47.45 1.09 75.38

Table 10. Tensile test of A533B base metal (HT 5795) beat treated by CE (Lot 6)

TYst Strain Modulus Yield Tensile Total Uniform Reduction Specimen temperature rate strength strength elongation elongation of area IC *F 1041s' 10 l1min GPa 108 psi HMa kst MPa ksi (t) ()) ('

8101 24 75 1.3 8.0 242 35.1 432 62.6 578 83.9 30.95 11.47 70.44 5102 24 75 1.3 8.0 202 29.3 451 65.4 596 86.4 30.95 11.33 68.09 8103 24 75 0.27 1.6 197 28.5 434 63.0 583 84.5 29.20 10.12 67.04 B104 24 75 0.27 1.6 226 32.8 429 62.2 569 82.5 30.10 10.86 67.83 B105 371 700 1.3 8.0 156 22.6 365 53.0 543 78.7 27.65 10.20 70.27 8106 371 700 1.3 8.0 118 17.1 366 53.1 539 78.1 30.15 9.86 72.56 B107 371 700 0.27 1.6 133 19.3 391 56.6 552 80.1 29.00 8.45 71.99 8108 371 700 0.27 1.6 156 22.7 409 59.4 560 81.3 28.35 8.25 71.85 B109 427 800 1.3 8.0 174 25.3 361 52.3 472 68.5 31.50 7.12 76.93 8110 427 800 1.3 8.0 108 15.6 398 57.7 494 71.7 29.22 7.98 73.88 B1ll 427 800 0.27 1.6 186 26.9 379 55.0 498 72.3 27.99 CHART 74.52 8112 427 800 0.27 1.6 123 17.9 340 49.4 479 69.4 32.60 7.70 75.14 B113 482 900 1.3 8.0 158 22.9 339 49.2 433 62.7 29.25 5.91 76.19 B114 .482 900 1.3 8.0 150 21.7 311 45.1 419 .60.8 34.51 6.72 78.08 B1S 482 900 0.27 1.6 155 22.4 336 48.8 419 60.8 34.00 4.42 77.61 5116 482 900 0.27 1.6 107 15.6 315 45.7 404 58.6 39.60 5.95 80.38 5117 538 1000 1.3 8.0 109 15.8 312 45.2 361 52.4 43.60 3.77 82.67 B118 538 1000 1.3 8.0 105 15.3 320 46.4 368 53.4 45.75 2.38 81.25 B119 538 1000 0.27 1.6 116 16.9 299 43.3 332 48.2 51.75 2.68 72.46 5120 538 1000 0.27 1.6 104 15.1 283 41.0 325 47.1 48.60 3.44 77.62 B121 593 1100 1.3 8.0 95 13.7 245 35.5 273 39.6 55.88 1.58 77.36 B122 593 1100 1.3 8.0 97 14.1 258 37.5 277 40.1 63.40 1.27 76.81 B123 593 1100 0.27 1.6 102 14.8 201 29.2 235 34.1 56.53 0.76 58.44 B124 593 1100 0.27 1.6 103 15.0 224 32.5 250 36.3 59.72 1.08 56.37 I .

l Table 11. Tensile properties of A533B all weld metal (ET 5795) heat treated by CE (Lot 7)

Test Strain Hodulus Yield Tensile Total Unifor Reduction Specieen temperature rate Modulus strength strength elongatlon elongation of area

C F 10'4s'1 10'min' CPa 10 s MPa ksl MPa ksi (t) (a) (I)

B307 24 75 1.3 8.0 216 31.3 608 88.1 699 101.4 28.70 9.02 65.73 B308 24 75 1.3 8.0 222 32.3 582 84.3 673 98.1 26.35 8.23 66.88 B309 24 75 0.27 1.6 254 36.8 588 85.3 683 99.0 27.00 7.72 66.78 3310 24 75 0.27 1.6 207 30.0 615 89.1 687 99.6 25.30 7.61 67.37 3311 371 700 1.3 8.0 201 29.2 S06 73.3 639 92.7 26.75 8.30 67.66 3312 371 700 1.3 8.0 200 29.1 506 73.4 638 92.6 23.90 7.07 68.61 B313 371 700 0.27 1.6 159 23.1 493 71.5 621 90.1 25.40 7.84 67.33 B3114 371 700 0.27 1.6 192 27.8 500 72.6 634 92.0 22.90 7.31 65.67 B315 427 B00 1.3 8.0 186 27.0 462 67.1 577 83.7 26.28 6.55 70.33 B316 427 800 1.3 8.0 168 24.4 456 66.1 563 81.7 23.20 6.46 56.68 B317 427 800 0.27 1.6 177 25.7 472 68.5 560 81.2 25.25 5.74 70.64 8318 427 800 0.27 1.6 153 22.2 463 67.2 566 82.0 25.02 6.20 68.14 F-h

%0 B319 482 900 1.3 8.0 156 22.6 449 65.1 509 73.8 30.35 6.26 73.00 B320 482 900 1.3 8.0 149 21.6 457 66.3 523 75.8 27.08 5.08 70.58 3321 482 900 0.27 1.6 148 21.4 451 65.4 507 73.5 33.10 4.66 74.70 B322 482 900 0.27 1.6 142 20.6 441 64.0 492 71.3 33.35 4.51 71.03 B323 538 1000 1.3 8.0 141 20.4 396 57.4 430 62.3 41.40 1.68 77.47 B324 538 1000 1.3 8.0 143 20.7 419 60.7 453 65.7 37.80 1.84 77.30 B325 538 1000 0.27 1.6 173 25.2 382 55.4 412 59.8 51.25 1.78 74.68 B326 538 1000 0.27 1.6 130 18.9 396 57.4 420 60.9 40.50 1.01 74.23 B331 593 1100 1.3 8.0 107 15.4 304 44.2 330 47.9 45.60 0.67 42.81 B332 593 1100 1.3 8.0 112 16.3 311 45.2 330 47.8 50.55 0.97 47.88 B333 593 1100 0.27 1.6 162 23.5 272 39.4 290 42.1 58.65 0.50 53.07 5334 593 1100 0.27 1.6 . 162 23.5 281 40.7 292 42.3 52.50 0.95 54.07

Table 12. Tensile properties of A533B transverse weld (11T 5795) heat treated by CE (Lot 8)

Test Strain Modulus Yield Tensile Total Uniform Reduction Specimen temperature rate strength strength elongation elongation of area 3

C *F 0losy 10- min'- CPa 10 psi HPa ksi HPa ksi (%)

B200T 24 75 1.3 8.0 212 30.8 483 70.0 647 93.8 22.03 6.79 57.21 B201T 24 75 1.3 8.0 194 28.2 472 68.5 630 91.4 21.75 6.38 62.55 B202T 24 75 0.27 1.6 207 30.1 463 67.1 631 91.5 19.80 5.89 62.57 B203T 24 75 0.27 1.6 202 29.2 480 69.6 643 93.3 21.75 6.30 61.48 B204T 371 700 1.3 8.0 162 23.5 411 59.7 567 82.2 21.25 5.66 72.09 B205T 371 100 1.3 8.0 190 27.5 417 60.4 572 83.0 21.70 5.48 68.46 B206T 371 700 0.27 1.6 176 25.5 434 62.9 576 83.5 18.00 5.13 65.17 B207T 371 700 0.27 1.6 177 25.7 423 61.3 578 83.8 19.30 5.54 67.61 8208T 427 800 1.3 8.0 189 27.5 413 60.0 515 74.7 20.65 3.79 73.29 B209T 427 800 1.3 8.0 209 30.2 399 57.9 508 73.6 22.10 2.96 73.74 B210T 427 800 0.27 1.6 198 27.3 402 58.3 497 72.1 21.00 3.22 78.32 3211T 627 800 0.27 1.6 257 37.2 470 68.2 497 72.1 22.50 3.62 77.45 N,

B212T 482 900 1.3 8.0 149 21.6 387 56.2 445 64.6 23.75 2.91 77.08 B213T 482 900 1.3 8.0 190 27.5 377 54.7 435 63.1 22.85 2.74 78.91 B214T 482 900 0.27 1.6 123 17.9 373 54.2 428 62.1 21.75 1.69 76.75 B215T 482 900 0.27 1.6 146 21.1 355 51.6 423 61.3 26.65 3.76 78.29 B216T 538 1000 1.3 8.0 119 17.3 351 51.0 377 54.7 25.12 1.76 80.35 B217T 538 1000 1.3 8.0 150 21.8 317 45.9 357 51.7 31.65 81.56 B218T 538 1000 0.27 1.6 177 25.7 322 46.7 354 51.3 26.43 1.48 77.69 B219T 538 1000 0.27 1.6 195 28.3 323 46.9 353 51.3 24.30 1.50 76.14 B220T 593 1100 1.3 8.0 78 11.3 277 40.2 304 44.0 25.12 1.04 72.92 B221T 593 1100 1.3 S.0 106 15.4 291 42.3 299 43.4 24.30 0.82 77.56 B222T 593 1100 0.27 1.6 77 11.1 214 31.0 239 34.7 30.20 1.34 65.56 3223T 593 1100 0.27 1.6 127 18.4 245 35.6 262 38.0 20.30 0.96 67.05

I0 k, .:

Table 13. Tensile properties of A533B base'metal (HT 9583B) heat treated by CE (tot 9)

Test Strain modulus Yield Tensile Total Uniform Reduction Spcimen temperature rate strength strength elongation elongation of area

.C *F 10-4s't 10-3min-' CPa 104 psi MPa ksL MPa ksi (%) (t)

C1O0 24 75 1.3 8.0 231 33.5 582 84.3 652 94.6 30.55 9.11 76.31 C101 371 700 1.3 8.0 196 28.4 406 58.9 543 78.8 30.25 8.49 79.12 C102 427 800 1.3 8.0 178 25.8 399 57.9 495 71.8 28.20 6.54 81.69 C103 482 900 1.3 8.0 214 31.1 384 55.7 438 63.3 27.99 3.78 84.75 C104 538 1000 1.3 8.0 140 20.3 327 47.4 352 51.1 40.90 1.86 86.60 C105 593 1100 1.3 8.0 100 14.5 240 34.8 258 37.4 20.000 1.02 82.94

'Specimen broke In gage marks.

Table 14. Tensile properties of A533B transverse weld (HT 9583B) heat treated by CE (Lot 10)

Test Strain Modulus Yield Tensile Total Uniform Reduction Specimen temperature rate strength strength elongation elongation of area-

C F 10'4s', 10"min' CPa 10 psi MPa k5t MPa ksi. (' )

C200T 24 75 1.3 8.0 210 30.4 569 82.6 643 93.2 23.70 7.47 69.02 C201T 24 75 1.3 8.0 192 27.9 568 82.4 647 93.9 23.35 7.19 70.25 C202T 24 75 0.27 1.6 211 30.6 554 80.4 633 91.8 25.58 8.52 72.29 C203T 24 75 0.27 1.6 194 28.1 487 70.6 646 93.7 21.50 7.09 70.96 C204T 371 700 1.3 8.0 148 21.4 474 68.8 586 85.0 21.00 6.44 68.57 C205T 371 700 1.3 8.0 184 26.7 471 68.3 582 84.3 23.20 6.67 70.82 C206T 371 700 0.27 1.6 173 25.1 474 68.8 583 84.5 20.80 6.72 69.69 C207T 371 700 0.27 1.6 183 26.5 472 68.4 579 84.0 23.22 6.33 72.30 C208T 427 800 1.3 8.0 184 26.7 451 65.4 538 78.0 20.05 5.68 72.31 C209T 427 800 1.3 8.0 160 23.2 459 66.6 538 78.0 22.01 4.88 73.87 C210T 427 800 0.27 1.6 176 25.5 451 65.4 528 76.6 24.40 5.21 74.74 C211T 427 800 0.27 1.6 177 25.7 443 64.2 504 73.1 22.80 1.17 78.39 76.69 r'3 C212T 482 900 1.3 8.0 148 21.4 422 61.2 486 70.4 23.25 4.29 C213T 482 900 1.3 8.0 172 25.0 427 61.9 475 68.9 22.25 4.13 76.19 C214T 482 900 0.27 1.6 158 22.9 420 60.9. 464 67.3 26.41 4.09 78.64 C21ST 482 900 0.27 1.6 138 20.0 431 62.5 475 68.9 24.10 3.45 71.58 C216T 538 1000 1.3 8.0 180 26.1 379 55.0 409 59.3 28.40 1.84 80.01 C217T* 538 1000 1.3 8.0 172 25.0 374 54.3 407 59.1 17.90 2.50 76.26 C218T 538 1000 0.27 1.6 148 21.4 355 51.4 383 55.5 34.10 1.69 80.07 C219T 538 1000 0.27 1.6 166 24.0 357 51.7 382 55.4 29.61 1.46 80.89 C220T 593 1100 1.3 8.0 124 18.0 294 42.6 316 45.9 35.93 1.09 80.21 C221TO 593 1100 1.3 8.0 149 21.7 294 42.6 316 45.8 12.68 0.93 67.21 C222T 593 1100 0.27 1.6 94 13.7 267 38.7 284 41.1 38.80 1.16 79.68 C223T 593 1100 0.27 1.6 121 17.6 263 38.2 282 41.0 38.90 0.82 82.98 0

Specimen broke outside of gage marks.

I .

Table 15. Tensile properties of A533B weld metal (HT 9583B) heat treated by CE (Lot 11)

Test Strain Kodulus Yield Tensile Total Uniform Reduction Specimen temperature rate strength strength elongation elongation of area -

C *F 10's-' 10'min-I CPA 10 psi MPa ksL HPa ksi, ( )

C300 24 75 1.3 8.0 214 31.1 606 87.9 670 97.2 26.75 9.81 60.82 C301 371 700 1.3 8.0 171 24.8 486 70.5 596 86.4 30.50 8.63 68.65 C302 427 800 1.3 8.0 144 20.9 464 67.3 544 78.8 23.15 6.42 74.45 C303 482 900 1.3 8.0 134 19.4 438 63.5 484 70.3 25.42 4.78 76.58 C304 538 1000 1.3 8.0 133 19.4 395 57.2 417 60.5 37.75 2.70 82.22 C305 593 1100 1.3 8.0 139 20.1 319 46.3 332 48.1 39.55 0.98 77.89

Table 16. Tensile properties of A533B base metal (HT 64535) heat treated by CE (Lot 12)

Test Strain Modulus Yield Tensile Total Uniform Reduction Specimen temperature rate strength strength elongation elongation of area

C *F 10'4s-' 10' 3 mIn'l CPa 10' psi ?Pa kai HPa ksi (M)

D100 26 75 1.3 8.0 244 35.3 478 69.3 607 88.0 30.50 12.45 71.51 D101 24 75 1.3 8.0 219 31.8 481 69.8 622 90.2 30.52 10.89 70.29 D102 24 75 0.27 1.6 212 30.8 485 70.3 621 90.1 29.40 10.20 69.99 D103 24 75 0.27 1.6 247 35.8 481 69.8 616 89.3 29.35 10.52 70.37 D104 371 700 1.3 8.0 148 21.5 400 58.0 551 80.0 27.95 10.34 72.00 D105 371 700 1.3 8.0 204 29.6 395 57.3 556 80.7 30.65 9.45 72.48 D106 371 700 0.27 1.6 229 33.3 402 58.3 552 80.1 30.65 8.64 74.60 D107 371 700 0.27 1.6 133 19.2 404 58.6 549 79.6 30.00 8.58 74.44 D108 427 800 1.3 8.0 161 23.4 387 56.1 491 71.2 32.28 7.94 77.98 D109 427 800 1.3 8.0 148 21.5 394 57.2 495 71.8 29.00 7.16 77.53 DU1O 427 800 0.27 1.6 212 30.7 385 55.8 486 70.5 32.61 7.43 78.35 Dill 427 800 0.27 1.6 173 25.2 377 54.7 472 68.5 29.10 6.66 78.58 41 D112 482 900 1.3 8.0 166 24.1 358 52.0 432 62.6 34.20 5.44 82.00 D113 482 900 1.3 8.0 173 25.0 359 52.1 437 63.4 33.99 4.69 79.65 D114 482 900 0.27 1.6 102 14.8 371 53.8 416 60.3 41.85 3.54 81.78 D115 482 900 0.27 1.6 143 20.8 347 50.4 415 60.2 35.25 4.28 80.59 D116 538 1000 1.3 8.0 197 28.6 342 49.6 350 50.8 45.42 2.29 84.46 D117 538 1000 1.3 8.0 194 28.2 333 48.3 355 51.5 41.85 2.20 86.30 D118 538 1000 0.27 1.6 141 20.4 308 44.6 340 49.3 41.00 1.70 86.06 D119 538 1000 0.27 1.6 116 16.8 310 45.0 332 48.1 48.50 1.92 76.39 D120 593 1100 1.3 8.0 94 13.6 249 36.2 271 39.3 68.35 1.04 88.00 D121 593 1100 1.3 8.0 88 12.8 278 40.4 280 40.6 35.30 1.06 85.46 D122 593 1100 0.27 1.6 114 16.6 194 28.1 237 34.4 44.62 1.59 82.22 0123 593 1100 0.27 1.6 65 9.4 215 31.1 244 35.3 63.10 1.39 86.26 I .

I a Table 17. Creep data for A533B steel-heat 9583A ApedLt mal

.20?

o.o01 °7.° oe 7490!

".a tn 2512 16.71 11240 41720 I20 40250

-1

o 2 *7 5 I I.*

'.- 43s9 a 02 76.5 D0 '.4 MIt o00029 eon 07 0an

- I -LuiX Ij noo252-0I'l "s§ rir.1r I.27 N4M0=1 =-11s.

-1 04_

-I..- a 03 2447 0.9 F _00 S

Table 18. Creep data for A533B steel-heat 5795 lESTNOt torra STRF5S.K51 TEMP. C SR. Wm T-0,1 T.1.% T.2% T-5% T-t TfR LOAING CS. % ED IN AREA.A 5972 £ 70 271 0.00062 0.025 242 1700 25150D 2.5 5.21 25968 6 65 371 0.00020 0.1 956 1.97 0.7 26194 6 6s 371 0 00016 4.4 1960 4291 0.e 1 2.4 2ss7e 6 427 0.003t 0.oS 45 240 679D 1.04 22 5.0 2s972 a so 427 0.000094 40 6480 0.01 0.23 25971 a 45 427 0.000025 e5 1t201D 0.29 0.2 1.42 6 i5 482 0.00036 e.4 2000 35150 0.28 2.8 25963 6 27 462 0.000026 170 13.S. o le 0.2 0 2ct0 6 20 "a O.0o0I9 9.s 440 810 126S 720 1990 0.13 19.6 t9.9 2970 6 t 53 0.0016 23 900 1780 1960 20260 0.07 2.5 3.94 25974 6 6 593 .0041 10 230 478 825 s60D 0.06 4.7 7.05 2668 6 4 sOS 0.0007 54.2 119s 1269 0.03 1.1 2596s 7 so 427 0.000076 22 8790 sos5035 _0.32 0.7 25961 7 35 482 0.0002 0o- 4450 33360 0.14 0.6 25662 7 is 826 0.00023 s5 2786 4370 2600 4799 0.14 24 2s9ee8 e 593 0.0026 0.1 350 630 1015 660 11710 0.os 7.4 13.80 26464 7 4 592 0.0003 62 1650 1 473 0.02 0.7

.25sso 6 60 427 0.05q 025 sso 11720 o s.. J.91 26702 a8 40 462 0.019 7.5 . 440 . 65es 10090 0.21 2.1 9.Si 0' 25s976 a 35 2 o0.00063 55 1540 2600 2400 2622 0.16

J.1 2.3t 2s577 e15 5e2 0.0009 1t.s 970 1640 1340 e18200 0.08 2. 54.3 25979 8 e9 0.eon 2.5 J07 202 370 170 362D 0.06 4.6 3jj 25220 6 4 593 0.00068 43 1300 1654 0.02 1.3 I

4.

Table 19. Creep data flcr A533B steel-heat 9583B TES'rta LWND STRESS l(SI 7E S.%*I 1_.V.% 1.1% 1.2% 'T.$% T-I T.R LODMS CREFPS% EOINAREA 23411 3 75 -371 j0.. 0.05 3 9.6 22.4 19.6 i 23.5 6.4 16.6 71.1 26193 9 el 427 0.016 O,1 32 96 256 212 2650 0.65 6.6 5 felo1 0 s0 482 --. 00026 12~9 2970 _ ___ ___ 4202 -0.17 1.4 26192 6 5 63 .01 50 90 2060 1160 22C80 0.08 S 3.26 258 92 0005 4.5_ l65 240 765 450 1362 0.06 J_1 _

762160 4 593 0.0006 36.9 1400 _ ___ 1774 0.01 1.21 _ ___

£9412 10 75 371 0.00024 0. 2S04 5 as_________ 1.76 0.6 ____

626 10 61 427 0.00024 0.1 490 4362 _____0_ _0__31 ______

o t602 __0 462 0.00014 as G10len__ __________ 3026 0.14 0.6 26I7 n is 536 0.0007 16.7 1160 ______ ____________ ______

245 10 6 593 0.0043 4.7 170 380 750 460 1241 0,0 20 681.61 2 00 10 4 502 0.00095 42 975 3& ,3 _L41.

26 90It6 427 0.00021 1.7 26 5 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 9 0 _ _ _ _ _ _ _ _ _ _

04482 0.0023 7. 70 160 10 L 0.1j. 37

-j T~sble 20. Creep data EA'A533B steel-heat 64535 I*

I Itp-Maf. .- L-Lt_ ..J I .. 1 mc.1. Ofa*

Ka. ,~ ..

-OfE IIC-I.U I -.-afl.-Thr I*

1 vI

-c...- . -_ . . _.

. _.. . - -1. - - _. _.-

I galas 1 i9 I at I j17 I AnE 'tv I 1ot [ -* s I 1713 1 740 1 1430 I 1964 1 1.09 I 75.1 I ,t-a I 1 26199 *I 12 15S 536 0.00016 I . 600 1 1270 .... 1ZZ __ __ _ 13450 . .

1 26201 1 1I 6 I 593 1 0.0092 1 0,36 90o 210 1 480 1 420 -1 1409 1 I

0.1

.9 .

1 5 It.

912

.- 1I7 I -- -- -- - _ __ __ -- -4 .I 19 45 a I61Q5 I I13 I A I Rai I n001 I ffm I 1020 I 1800 I 1 1740 1 19450 1 0-09 I I- 9.; 1 fin I

Table 21. Effect of creep exposure on tensile properties Reeults for exposed asaple. 0soults for unexposed mamples (tS9la HIAI K LOCATION HEATTRwIAl _

TEMP. C StRt.SKfSI l fill. l S.RSI U15.XSI l tc.N6.x l RAX i I U15 I (LONG tA OA t A477 4 1209 J04 67 . . _ AeL..* 6*

e ..

.. 69 3 739?I9 9 MII .I OAS 4_2 50 6ee 1e 9 9 20 i _ S 6 64 31 39 69 3 275" 5795 AS t do? 2? . 319 645 *j j5 o L.2 .o

. .2 . e.1 t.'i 23594a 5*95 o e H CI IN ..... 960 e +/- . es .. __-9'd e5.L S 693

.5570 9 xA _ BAs ORML . 311 75 sn9. 1g6 1M0 20 7I.7 _90 .. L 2.3 75 7 ns553 _jtexA OAS( ORK 1 _71 or i49- j2j l0. jj2 74 s o 00.5 _ lj 7I7 2st14 9563A ASL 0seO.2 III so SO70 nZ 3 es t 29 3 j 75 3I e oS S 7?e

?d" 925 exs es 2e 7 70.3 71 90.1 9 05 as-

_250s_ es3 BAS Ite tt 3i1 e 7 25799 tISIA AS( I ORM2 a? 50 1321 64 . 042 22.4 72.5s .. .esot 32.5 776 1340 s 7 106 26.4 725 - 90 109 3 2o ) ?I 2ST01 OU vKs 5~a_~T 47_ eQ I.s 3062 320. . 02 Ij 90.0 205 745 25A97 eeM _jt '127 a5 3193 25961 95A .A..ORIII. L2? 4. . 2015 oe 03. 33.4 749 7o 3 05.2 e2.. S 76 25179 OSO3 ASE 2 e f2eo

RM*

2 j5Je 7re 3S 7o 2 70,3 2 _e 3 2.5 7r0 o

956_ e2st SAI e M .. j9 707 7 . Jj s 1s 727 703 85.2 32.5 7 e "no 2l22 95GA C AsM O. . i 1360 37e 70, _ .7MI. 70.2 703 85.2 32f 175 25690 sex SW tt 311 e7 6o tl e J 707 e 2 74 7 90.9 25 e 163 2sW n 6e7 554) 03 2 2.3

1) *eV7 e Ct e 134 07 0X. 205 6. . 09 206 6b4 23524 257 5 2os552e 3MIg 9

_302 L C C1 oa ..ara...

7 2 .....

402....n q62 2j7279 7 .. Me +/-L. 66 1.... .+/-L.. .... L... 4 ..I.. .... 22.. 6 25942 n Ct 4& 50 ?IS At& o+s7 9eeb z ee b 290 Y Weo C1 s~~~~~~~~e 20 es os e 47 n255ss Xb7e eeeooo6e evQse24 co

29 ORNL-DWG 89-18608 700 600 0 500 w

cn

, 400 S.-

300 200 _

o *O0 z00 300 400 500 600 TEMPERATURE (C)

Fig. 1. Yield strength of several lots of A533 base metal as a function of test temperature.

ORNL-DWG 89-18609 Wo 700 In _ __ _ __

300 0 100 200 300 :OD Soo 600 TEMPERATURE (C)

Fig. 2. Yield strength of three lots of weld metal as a function of temperature.

30 ORNL-DWG 89-18610 600 500 IZ 0.

400 I.-

U, 300 200 0 100 200 300 400 500 600 TEMPERATURE (C)

Fig. 3. Yield strength of three lots of transverse weld samples as a function of temperature.

ORNL-DWG 89-18611 600 T

0.

tu I.-

V)

TEMPERATURE (C)

Fig. 4. Yield strength for lots of base metal, transverse welds, and weld metal having the lowest strength.

31 ORNL-DWG 89-18612 0.

C,)

un I-U, 100' I 0 100 200 300 400 500 600 TEMPERATURE (C)

Fig. 5. Yield strength of A533 base metal at a strain rate of 3.7 x 10's-l as a function of temperature.

I ORNL-DWG 89-18613 W,

(n U,

I-U, 200' 0 100 200 300 400 Soo 600 TEMPERATURE (C)

Fig. 6. Yield strength of A533 weld metal at a strain rate of 3.7 x 10-&s- as a function of temperature.

32 ORNL-DWG 89-18614 600 0

E-c M

n u

0 100 200 300 400 500 600 TEMPERATURE (C)

Fig. 7. Yield strength of transverse weld specimens of A533 at a strain rate of 3.7 x 10-5 s-1 as a function of temperature.

ORKL-DWG 89-18615 600 500 tu Cn SL cm 400 LU BI-0) 300 200 0

TEMPERATURE (C)

Fig. 8. Yield strength at a strain rate of 3.7 x 10'6s' 1 for lots of base metal, transverse weld samples, and weld metal as a function of temperature.

33 ORNL-DWG 89-18616 IZ M.

I-Ml 200 300 400 TEMPERATURE (C)

Fig. 9. Tensile strength of several lots of A533 base metal as a function of temperature.

ORNL-DWG 89-18617 800 700 v 600 CA, C 500 I.-

Cn 400 300 _

o 100 200 300 400 500 600 TEMPERATURE (C)

Fig. 10. Tensile strength of several lots of A533 weld metal specimens as a function of temperature.

34 ORNL-DWG 89-18618 0.

I-in on mu 200 0 o 100 200 300 400 Soo 600 TEMPERATURE (C)

Fig. 11. Tensile strength of several lots of A533 transverse weld specimens as a function of temperature.

ORKL-DWG 89-18619 700 600 2.0 .0 3 100 200 0 100 200 300 400 500 600 TEMPERATURE (C)

Fig. 12. Tensile strength for lots of base metal, transverse welds, and weld metal having the lowest strength.

35 ORNL-DWG 89-18620 Cn XU U, 400 0 100 200 300 400 S00 600 TEMPERATURE (C)

Fig. 13. The ultimate tensile strength of several lots of base metal as a function of temperature.

I ORNL-DWG 89-18621 IZ a.

=

w I.-

mu 20CoI 0 100 200 300 400 S00 600 TEMPERATURE (C)

Fig. 14. The ultimate tensile strength of several lots of weld metal as a function of temperature.

36 ORNL-DWG 89-18622 700 cn Lu i,

M 600 TEMPERATURE (C)

Fig. 15. The ultimate tensile strength of several lots of transverse weld specimens as a function of temperature.

ORNL-DWG 89-18623 700 6 4 600 -

500 U3 I-400 Ln 300 200 0 100 200 300 400 500 600 TEMPERATURE (C)

Fig. 16. Ultimate tensile strength for lots of base metal, weld metal, and transverse weld metal having the lowest strength.

37 ORNL-DWG 89-18624 Soo 400 m

w 200 0 100 200 300 400 500 600 TEMPERATURE (C)

Fig. 17. Yield strength of Lot 6 measured at strain rates of 1.3 x 10'4s-1 and 3.7 x 10'5s-1.

ORNL-DWG 89-18625 700. .

a HIGH STRAINRATE 600

LOW STRAIN RATE L 500 L

CI I

II Lr 400 300 200 .00 4 0 a 100 200 300 400 So0 600 TEMPERATURE IC)

Fig. 18. Ultimate tensile strength of Lot 6 measured at strain rates of 1.3 x 10-4s- and 3.7 x 10'6s-1.

38 ORNL-DWG 89-18626 1000 . . . v . . . .

I

a 0 100

a. C ORNLHT I *

ORNLHT2 .3

CE STDI4T a I-cn 10 am I I 12 14 16 I8 20 22 P.1 Fig. 19. Larson-Miller correlation for Lots 1, 2, and 3 base metal crept to 1% strain.

ORNL-DWG 89-18627 T

a.

0 100 Ia 0

w.

P.1 Fig. 20. Larson-Miller correlation for Lots 3, 6, 9, and 12 base metal crept to 1I strain.

39 ORNL-DWG 89-18628 1000 T

0.

l 100 I-Cn uz 10'.-

12 14 16 1I 20 22 P-I Fig. 21. Larson-Miller correlation for 1 strain showing average and minimum strengths for Lots 3, 6, 9, and 12 base metal.

ORNL-DWG 89-18629 1000 10 P.'

Fig. 22. Larson-Miller correlation for tertiary creep for Lots 3, 6, 9. and 12 base metal.

40 ORNL-DWG 89-18630 1000 I-0.

0 10 L_

12 13 14 15 16 17 18 I9 20 21 22 P.R Fig. 23. Larson-Miller correlation for rupture for Lots 3, 6, 9, and 12 base metal.

ORNL-DIG 89-18631 1000 , . . . .. . .... .... . . . .. . .... .... . . . . ....

a 100

/e a

- aaa a

10 10 100 1000 10000 T.R (h)

Fig. 24. Time to o 1% strain as a function of time to riupture for Lots 3, 6, 9, and 12 base metal.

41 ORNL-DWG 89-18632 4.,

10woo a

1000 a a

a 100 10 10 100 10oo 10000 T4R (h)

Fig. 25. Time to tertiary creep as a function of time to rupture for Lots 3, 6, 9, and 12 base metal.

ORNL-DWG 89-18633 A.

Eo . . ....... .............. , . . ...... ......... - . . .............. . . -

a 60 z a 40 I-a 20 a a 0 . . ...... .......... . . . ......................... . . . . ...

100 1000 10000 RUPTURE TIME (h)

Fig. 26. Fracture strain as a function of rupture time for Lots 3, 6, 9, and 12 base metal.

42 ORNL-DWG 89-18634 1000

( 100 0

w so 10 _.

300 600 TEMPERATURE (C)

Fig. 27. Allowable stresses as a function of temperature for base metal based on various design criteria.

ORNL-DWG 89-18635 1000 on 100 c

°I 00 10 P.'

Fig. 28. Larson-Miller correlation for 1% strain in Lots 5, 7, and 11 weld metal.

43 ORNL-DWG 89-18636 1000 _

I U

0.

e 100 Cd, 10 _

16 17 15 I6 20 21 P.1 Fig. 29. Larson-Miller correlation for 1% strain in Lots 5, 7, and 11 (combined) weld metal.

ORNL-DwG 89-18637 1000 , * *-

C LOT4 i* LOT N LOTIO

0. 0 a,100 as 16 17 I 1I9 20 21 P-1 Fig. 30. Larson-Miller correlation for 1% strain in Lots 4, 8, and 10 trans-verse weld samples.

44 ORNL-DWG 89-18638 1000 E-0 100 lIa l1 P-1 Fig. 31. Larson-Miller correlation for 1 strain in Lots 4, 8, and 10 (combined) transverse weld samples.

I ORxL-DWG 89-18639 1000 T

C.

0 100 w

V-0 0 L-12 1A 16 I8 20 22 P-I I

Fig. 32. Larson-Miller plot comparing the average properties of base metal, weld metal, and transverse weld samples.

.45 DOE-HTGR 88383 0RNL/TM-11338 l.

Distribution Category UC-522T V

INTERNAL DISTRIBUTION 1-2. Central Research Library 25. D. W. Heatherly

3. Document Reference Section 26. J. R. Hightower 4-5. Laboratory Records Department 27. F. J. Homan

6. Laboratory Records, ORNL RC 28. D. T.Ingersoll

7. ORNL Patent Section 29. U. Jones 8-10. M&C Records Office 30. M. J. Kania

11. C. A. Baldwin 31. P. R. Kasten (Consultant)

12. W. P. Barthold (Consultant) 32. T. S. Kress

13. E. E. Bloom 33. J. C. Mailen

14. J. A. Bucholz 34-38. H. E. McCoy

15. T. D. Burchell 39. F. R. Mynatt 16-17. E. E. Clemner 40. N. H. Packan

18. J. C. Cleveland 41. P. L. Rittenhouse

19. W. P. Eatherly (Consultant) 42. J. P. Sanders

20. L. C. Emerson 43. 0. M. Stansfield

21. E. C. Fox 44. J. 0. Stiegler

22. U. Gat 45-49. K. R. Thoms

23. R. K. Genung 50. D. B. Trauger

24. H. A. Glovier 51. A. W. Trivelpiece EXTERNAL DISTRIBUTION

52. COMBUSTION ENGINEERING, 1000 Prospect Hill Road, Windsor, CT 06095 S. A. Caspersson

53. EG&G IDAHO, INC., P.O. Box 1625, WCB-E3, Idaho Falls, ID 83415 E. Arbtin

54. GENERAL ATOMICS, P.O. Box 85608, San Diego, CA 92138-5608 A. J. Neylan 55-57. JAPAN ATOMIC ENERGY RESEARCH INSTITUTE, Tokai-Mura, Ibaraki-Ken 319-11, Japan I.I K. Fukuda T. Kondo K. Shiba

5. DOE-HTGR-90286, "Documentation of ASME Code Case for Elevated-Temperature Service of MHTGR Reactor Vessel Materials,"

September 1991.

QHRS HOFMANN DOE-HTGR-90286 REV. 0 JIFGR 11111 I 11111 DOCUMENTATION OF ASME CODE CASE FOR ELEVATED-TEMPERATURE SERVICE OF MHTGR REACTOR VESSEL MATERIALS APPLIED TECHNOLOGY Any Further Distribution by any Holder of this Document or Data Therein to Third Parties Representing Forlgn interests. Foreign Govemmertu. Fomign Companles, and Foreign Subsidiaries or Foreign Dlvimons of U.S. Companies Shall be Approved by the Associate Deputy Assistant Secretary for Reactor Systems. Development and TechnooW U.S. Department of Energ Further Foreign Party Raelse May Require DOE Approval Pursuant to Federal Regulation 10 CFR Pat 810, andkr May Be Subject to Section 127 of the Atomic Energy Act.

AUTHORS/CONTRACTORS COMBUSTION ENGINEERING,INC.

ISSUED BY: COMBUSTION ENGINEERING, INC.

FOR THE DEPARTMENT OF ENERGY CONTRACT DE-AC03-89SF17886 SEPTEMBER 1991

MHTGR TECHNOLOGY DEVELOPMENT PROGRAM TASK 1603.3 REACTOR VESSEL MATERIALS DOCUMENTATION OF ASME CODE CASE FOR ELEVATED-TEMPERATURE SERVICE OF MHTGR REACTOR VESSEL MATERIALS Prepared by:__ _ , -, ,.

Date: 2-w-!5/

t. L. Hoffmann, g6gnizant Engineer Approved by: t A Date: A W. R. Gah~171ler, Supervisor, Materials Technology Approved by: . -

Date: to/ / )

S. M. Schloss, Manager, Materials/Chemical Technology Approved by: -- :5, Date: 1__ _ _

S. A. Caspersson Project Manager COMBUSTION ENGINEERING, INC.

ABB COMBUSTION ENGINEERING NUCLEAR POWER MATERIALS/CHEMICAL TECHNOLOGY WINDSOR, CONNECTICUT DOE-HTGR-90286, Rev. 0

TABLE OF CONTENTS Section Title

1.0 INTRODUCTION

1 2.0

SUMMARY

I 3.0 DISCUSSION 2 3.1 MHTGR VESSEL ELEVATED-TEMPERATURE SERVICE 2 3.2 INQUIRY TO ASME CODE 3 3.3 ASHE COMMITTEE ORGANIZATION 3 3.4 APPROVAL PROCESS FOR THE MHTGR CODE INQUIRY 3 3.5 SUPPORTING MATERIALS TESTING PROGRAM 4 3.6 SUPPORTING DESIGN ANALYSIS 5

4.0 REFERENCES

5 APPENDIX I INQUIRY AND REPLY FOR NEW CODE CASE APPROVED BY THE ASME CODE MAIN COMMITTEE LIST OF FIGURES Figure No. Title Pap-e 3-1 Committee Procedure for Adopting Proposed 6 Changes to the Code 3-2 Flowchart for Code Committee Actions on Code 7 Inquiry for MHTGR Elevated Temperature Service 3-3 Chronology of MHTGR Reactor Vessel ASME Code Case 8 DOE-HTGR-90286, Rev. 0

MHTGR TECHNOLOGY DEVELOPMENT PROGRAM - TASK 1603.3 DOCUMENTATION OF ASME CODE CASE FOR ELEVATED-TEMPERATURE SERVICE OF MHTGR REACTOR VESSEL MATERIALS

1.0 INTRODUCTION

The HHTGR vessel system includes an uninsulated, steel, reactor pressure vessel to allow decay heat removal by conduction and radiation during a total loss of coolant and/or coolant flow event. Certain low-probability conduction-cooldown events can raise metal temperatures of the reactor vessel above 370'C (700'F), the maximum temperature allowed by Section III of the ASME Boiler and Pressure Vessel Code for the selected pressure vessel materials. An Inquiry was submitted to the ASME Code Committee requesting a special Code Case which would provide allowable stresses and design rules for the limited elevated-temperature service of the MHTGR reactor vessel.

A materials test program was performed to provide the material properties required to obtain the Code Case approval. The time/temperature dependent behavior of the MHTGR pressure vessel materials was characterized for the range of times and temperatures occurring during conduction cooldown events.

The test results were used as the basis for establishing time-independent and time-dependent stress allowables.

Analytical work was performed to determine the possible range of material response for the time-temperature conditions of the MHTGR duty cycle events.

The results of the analysis were used to define the design rules for the Code Case.

The following discussion reviews the HHTGR elevated-temperature service conditions, design approach, and the technical issues associated with obtaining approval for the new Code Case. Previous progress on this task was reported in Reference 1.

2.0 SUMl Y The design for limited elevated-temperature service of the MHTGR reactor vessel will be governed by the rules and allowable stresses of a new Code Case to the ASME Code. An Inquiry was submitted to the ASME Code Committee. The Inquiry and Proposed Reply received final approval of the ASME Code Main Committee on September 13, 1991. Following the Main Committee approval, the new Code Case is sent to the Board on Nuclear Codes and Standards (BNCS) for approval and is then published for public comment. The Code Case is available for use upon BNCS approval. C-E actively participated in the ASME Code committee meetings to obtain the approval of the new Code Case.

Materials data and analysis of elevated-temperature materials response were provided to demonstrate the capability of the HHTCR vessel materials.

Elevated-temperature time-dependent materials properties were generated in test programs at ORNL and C-E. Both of these test programs were completed in FY 1990. The final data package was assembled following completion of all the testing and presented to the Code Committees at the May 1990 ASME meeting.

I DOE-HTGR-90286, Rev. 0

The data package included elevated-temperature tensile and creep data necessary for establishing allowable stresses. In addition, results from the temper embrittlement study, cyclic stress-strain curves, and elevated-temperature fatigue tests were presented. The elevated-temperature materials properties were used to establish allowable stresses for MHTGR design during Service Level C and D events at temperatures above 370'C C700F) for limited times.

Analytical work to define the materials response to elevated-temperature service was performed by C-E and ORNL. The analyses demonstrated that a simplified set of design rules could not be developed to address all of the potential elevated-temperature design considerations. Based on the analytical results, the existing elevated-temperature design rules of ASME Section III Code Case N-47 were incorporated into the MHTGR Inquiry.

3.0 DISCUSSION 3.1 MHTGR VESSEL ELEVATED TEMPERATURE SERVICE Reference 1 contained a detailed discussion of the MHTGR reactor vessel elevated-temperature service. The MHTGR reactor vessel is cooled by passive heat transfer, with no active decay heat removal systems. The design duty cycle of the MHTGR contains several low-probability events (Service Level C and D conditions) in which all forced circulation is lost. The elevated-temperature events are characterized by the following conditions:

1) Pressurized conduction cooldown, with a maximum metal temperature of approximately 410'C (770'F), with a time duration above 370'C-(700'F) of approximately 150-200 hours, and

2) Depressurized conduction cooldown with a maximum metal temperature of approximately 470'C (880'F), with a time duration above 370'C (700'F) of approximately 400 hours0.00463 days 0.111 hours 6.613757e-4 weeks 1.522e-4 months .

Two events of each type are included in the HHTGR duty cycle. The number of events was restricted to reduce elevated-temperature creep-fatigue interaction concerns, while maintaining the flexibility within the duty cycle to return the MHTGR plant to service following one of these events. If two events of either type were to occur, the duty cycle for that event would be reduced to zero. Therefore, the maximum number of temperature cycles above 370-C (700'F) is effectively limited to a total of three. The maximum number of events permitting the return of the vessel system to operation is only two, since the vessel can only be returned to operation when the duty cycle provides for the possible reoccurrence of either event.

Allowable stresses for the SA 533 Grade B, Class 1 and SA 508 Class 3 pressure vessel steels were limited to 370'C (7006F) by Section III of the ASME Code.

Section III designs are based only on the time-independent strength properties of materials. Time-dependent material behavior must be considered at elevated-temperatures. There was no provision in Section III for deriving and using allowable stresses in the elevated-temperature regime.

2 DOE-HTGR-90286, Rev. 0

A request was made to the ASHE Code Committee for a special Code Case to allow use of pressure vessel steels at temperatures above 370'C (700'F) for limited times. The ASME Code Section III design procedures, supporting data and allowable stresses are applicable to virtually the entire duty cycle with the exception of the low-probability Service Level C and D events. Obtaining approval of the new Code Case required establishing the necessary allowable stresses and supplementary design rules to be used in conjunction with existing ASHE Code Section III data and rules for duty cycle events where temperatures are greater than 370'C (700'F).

3.2 INQUIRY TO ASME CODE An Inquiry was prepared and submitted to the ASME Code Committees for consideration. The scope of the Inquiry enveloped the conditions for the anticipated Service Level C and D events in the MHTGR duty cycle. The general form of the Inquiry was as follows:

May SA533 Grade B, Class 1 plates, SA508 Class 3 forgings and their weldments be used in Section III, Division 1, Class 1 construction at temperatures exceeding 700'F up to 1000'F during Service Level C or D events for limited times of exposure not to exceed 1000 hours0.0116 days 0.278 hours 0.00165 weeks 3.805e-4 months ?

The Inquiry remained essentially the same since it was originally submitted.

However, the original Proposed Reply, which the Inquirer is required to provide to the ASME Code committee, underwent numerous revisions. The wording of the Inquiry and Reply approved by the ASME Code Main Committee is provided in Appendix I.

Approval by the ASME Code Committee resulted in the establishment of a new Code Case for the materials and design of the MHTGR vessel. The Code Case provides the Code Committee's response to the Inquiry in terms of design rules which will govern the limited elevated-temperature service of the MHTGR vessel and the time-temperature related allowable stresses which govern the vessel stresses at temperatures above 370'C (700F).

3.3 ASME COMMITTEE ORGANIZATION The overall organization of the ASME Code Committee was described in Reference

1. Appendix XX to Section III of the ASME Code describes the requirements for preparation of technical inquiries to the Code committee. Technical inquiries to the Code are forwarded to the appropriate Subcommittee(s) for review and action. Figure 3-1 shows a generalized flow chart of the committee procedure for considering and approving changes to the Code. Proposed changes to the Code must pass through the approval sequence shown in Figure 3-1 in order to be adopted into the Code or as a Code Case.

3.4 APPROVAL PROCESS FOR THE MHTGR CODE INQUIRY The Inquiry was considered and approved by several Subgroups and/or Subcommittees of the ASME Code Committee and was evaluated in terms of both design and materials requirements. The MHTGR design is intended to adhere to the rules of Section III of the ASHE Code for all other design and operating conditions. The design for the limited elevated-temperature service will be based on the supplemental rules and requirements of the Code Case.

3 DOE-HTGR-90286, Rev. 0

Figure 3-2 shows the specific subgroups and subcommittees involved in addressing the C-E inquiry for MHTGR reactor vessel material elevated temperature service. Figure 3-3 shows a complete chronology of the ASME Code Case for the MHTGR reactor vessel, which was initiated in 1987.

Since the Inquiry was for a Section III Class 1 application, the overall responsibility for the Inquiry was with the Subcommittee on Nuclear Power (Section III). Due to the elevated-temperature requirements of the request, the Inquiry was assigned to the Subgroup on Elevated Temperature Construction.

The Inquiry was addressed by Subgroups of the service subcommittees on properties and design in order to establish stress allowables based on the material properties and to provide the design rules that will be incorporated in the Code Case.

The Inquiry and Proposed Reply were initially approved by the Subgroup on Elevated-Temperature Design at the May 1990 ASME Code meeting. Additional approvals were obtained from the other Subgroups and Subcommittees indicated in Figure 3-2 at subsequent committee meetings, with the Main Committee approval obtained at the September 1991 meeting.

Progress on developing the committee response to the Code Inquiry required material property data for the SA533 and SA508 pressure vessel steels and weldments and the results of design analysis to demonstrate the material response to the time and temperature range requested in the Inquiry. The status of the materials test program and design analysis performed to support the Inquiry are discussed below.

3.5 SUPPORTING MATERIALS TEST PROGRAM A test program was performed to provide the materials property information required to develop allowable stresses and design rules for the special Code Case. Testing was performed by Oak Ridge National Laboratories (ORNL) and Combustion Engineering's Metallurgical and Materials Laboratory. The test matrix included three heats of SA533 plate, one SA508 forging and three weldments (two submerged arc welds and one shielded metal arc weld). The testing performed on these materials was as follows:

1) Tensile testing from 700F to 1100'F,

2) Creep testing at several different stress levels from 800'F to 11000F for test times up to 2000 hours0.0231 days 0.556 hours 0.00331 weeks 7.61e-4 months , and

3) Evaluation of the potential for thermal embrittlement of these materials by aging base and weld metal test specimens at 850'F and 950'F for 2000 hours0.0231 days 0.556 hours 0.00331 weeks 7.61e-4 months and comparing Charpy impact test results for the aged and unaged materials.

The completed data package of the testing was provided to the ASME Code Committee in May 1990 (Ref. 1). The test data were used as the basis for establishing allowable stresses for the Code Case. The materials data submitted to the Code committee included room and elevated temperature tensile test data and trend curve analysis for establishing Sm values.

4 DOE-HTGR-90286, Rev. 0

Information in the materials property data package provided all of the necessary material property information for Code Case N-47. The material properties have been incorporated as part of the Reply of the new Code Case (See Appendix I Figures and Tables). This data is required for performing the elevated-temperature design analysis according to Code Case N-47 rules. The elevated-temperature data included S values based on 1% creep strain in 1000 hours0.0116 days 0.278 hours 0.00165 weeks 3.805e-4 months , onset of tertiary creep and stress rupture data. Creep and creep rupture equations were generated by analysis of the data. The package included information on creep strain to failure, isochronous stress-strain curves, cyclic stress-strain curves, elevated temperature fatigue curve and thermal aging.

3.6 SUPPORTING DESIGN ANALYSIS In response to the suggestion of the ASME Subgroup on Elevated Temperature Design, evaluations of simplified geometries with an enveloping duty cycle were performed. A simple, plane-strain, symmetrical, cylindrical model was analyzed. The analyses were intended to bound the MHTGR transients that can occur for Service Level C events and provide an envelope for load-controlled and strain-controlled stresses subject to the criteria of Code Case N-47. The results of these analyses demonstrated where creep strains and creep-fatigue damage became significant.

Based on the results of the preliminary design analysis, the approach of using a bounding design envelope on primary and secondary stresses was discontinued.

The analysis had shown that a bounding envelope would be overly restrictive, since it was attempting to generalize all possible stress conditions in the structure. The existing elevated-temperature design rules of Code Case N-47 were adopted for analyzing the MHTGR Level C and D events that exceeded 700F.

4.0 REFERENCES

1) TR-MCC-149, Rev. 02, "Progress Report for FY 1990 on ASHE Code Inquiry for Elevated-Temperature Service of Reactor Vessel Materials," Combustion Engineering, Inc., September 1990.

5 DOE-HTGR-90286, Rev. 0

Figure 3-1 COMMITTEE PROCEDURE FOR ADOPTING PROPOSED CHANGES TO THE CODE PROPOSED CHANGE - CORRESPONDENCE, TASK GROUP, WORKING GROUP I

SUBGROUP - (SECTION III HAS 6, SECTION XI HAS 7)

SUBCOMMITTEE - VOICE VOTE MAIN COMMITTEE - (FIRST CONSIDERATION - 1 WRITTEN NEGATIVE BALLOT STOPS ACTION)

MAIN COMMITTEE - (SECOND CONSIDERATION - 4 WRITTEN NEGATIVE BALLOTS DEFEAT ITEM, CANNOT BE BROUGHT BACK UNLESS IT IS CHANGED IN TECHNICAL CONTENT)

BOARD ON NUCLEAR CODES AND STANDARDS PUBLIC REVIEW - (ANNOUNCED IN MECHANICAL ENGINEERING, ASME MONTHLY MAGAZINE, 4 MONTHS AFTER MAIN COMMITTEE MEETING)

PUBLISHED - CODE CASE (4 TIMES A YEAR)

OR ADDENDA (ONCE A YEAR) 6 DOE-HTGR-90286, Rev. 0

3-2 FIGURE ASME COMMITTEE WORK FLOW DIAGRAM MHTGR CODE INQUIRY FOR REQUEST REACTOR FOR LIMITED PRESSURE ELEVATED VESSEL TEMPERATURE SERVICE MATERIALS INQUIRER lASME STAFF I SUBCOMMITTEE ON NUCLEAR POWER (SECTION III)

SUBGROUP ON STRENGTH, SUBGROUP ON ELEVATED FERROUS ALLOYS TEMPERATURE CONSTRUCTION (SECTION III)

I O O SUBGROUP ON STRENGTH, SUBGROUP ON ELEVATED WELDMENTS TEMPERATURE DESIGN ISUBGROUP ON TOUGHNESS SUBCOMMITTEE ON DESIGN SUGOP ON MATERIALS, FABRICATION & EXAMINATION (SECTION III)

SUBCOMMITTEE ON SUBGROUP ON DESIGN MATERIALS (SCII) (SECTION III)

SUBCOMMITTEE E ON NUCLEAR POWERl-I (SECTION III)I AS-ME CODE MAIN COMMITTEE1 IBOARD ON NUCLEAR CODES & STANDARDSl l PUBLIC REVIEW PUBLISHED CODE CASE 7 DOE-HTGR-90286, Rev. 0

FIGURE 3-3

CHRONOLOGY OF MHTGR REACTOR VESSEL ASME CODE CASE o CODE INQUIRY SUBMITTED TO ASME CODE COMMITTEE - 11/87 o INTERACTION WITH ASME CODE COMMITTEE AT QUARTERLY MEETINGS - 1987-1991 o MATERIALS TEST PROGRAMS COMPLETED BY ORNL AND C-E (MML) - 3/90 o REVISED INQUIRY/PROPOSED REPLY SUBMITTED TO CODE COMMITTEES - 5/90 o FINAL MATERIALS PROPERTIES DATA PACKAGE SUBMITTED - 5/90 o CODE CASE APPROVED.BY ASME CODE MAIN COMMITTEE - 9/13/91 8 DOE-HTGR-90286, Rev. 0

APPENDIX I INOUTRY AND REPLY FOR NEW CODE CASE APPROVED BY THE ASME CODE MAIN COMMITTEE DOE-HTGR-90286, Rev. 0

INOUIRY AND REPLY FOR NEW CODE CASE APPROVED BY THE ASHE CODE MAIN COMMITTEE INQUIRY May SA-533 Grade B, Class 1 plates, SA-508 Class 3 forgings and their weldments be used in Section III Division 1, Class 1 construction at temperatures exceeding 700'F up to 1000F during Service Level C or D events for limited time of exposure not to exceed 1000 hours0.0116 days 0.278 hours 0.00165 weeks 3.805e-4 months ?

REPLY It is the opinion of the committee that Class 1 nuclear components, fabricated from SA-533 Grade B, Class 1 plates, SA-508 Class 3 forgings and their weldments, may be used when metal temperatures exceed 700'F during Level C and D events in accordance with the following considerations:

(1) Metal temperatures shall not exceed 10000F.

(2) The component design shall be based on a maximum cumulative time of 1000 hours0.0116 days 0.278 hours 0.00165 weeks 3.805e-4 months when metal temperatures exceed 700F.

(3) The number of anticipated events where metal temperatures exceed 800F shall be limited to a total of 3.

(4) The rules for materials in Section III, Division 1, NB-2000 and Code Case N-47 for Class 1 Components in Elevated-Temperature Service shall apply to the materials of this case with the following additions:

(a) The material specifications permitted by this Code Case are SA-533 Grade B Class 1. SA-508 Class 3 and their weldments.

(b) The allowable stress intensities in Table I of this case shall be considered as extensions to the values of Tables 1-1.0 in Appendix I for the materials and conditions addressed by this Case.

(5) The rules for design are:

(a) The design rules of NB-3000 shall be satisfied for all Design and Operating Conditions for which metal temperatures do not exceed 700-F. The Design Conditions shall be as defined in NB-3000.

(b) Metal temperatures exceeding 7000F are permitted only for Service Level C and D events. The applicable rules of Code Case N-47 for Class 1 Components in Elevated-Temperature Service shall be satisfied for these events.

Appendix I Page 1 of 22 DOE-HTGR-90286, Rev. 0

The creep-fatigue interaction damage envelope shown in Figure 12 shall be used for the materials of this Case.

The mechanical and physical property values at elevated-temperatures are provided in Figures 1 thru 11 and Tables I thru 8 of this Case.

The properties include:

(1) Isochronous Stress-Strain Curves, (2) Yield Strengths, (3) Stress-to-Rupture Values, (4) Elevated-Temperature Fatigue Strength, (5) Moduli of Elasticity, and (6) Instantaneous and Mean Coefficients of Thermal Expansion.

The stress-rupture factors for welds shall have a value of 1.0 for the materials and conditions of this Case.

(c) In addition, the sum of the creep damage and fatigue damage, summed over the entire lifetime, shall not exceed the limit of Figure 12 anywhere in the structure. When performing the creep-fatigue interaction analysis, load history effects and residual stresses from prior low-temperature operation shall be considered in the evaluation. Since the fatigue curve at temperatures above 700F is more restrictive than that for temperatures below 700F, strain cycles which have one extremum at elevated-temperature and one extremum at low temperature shall be evaluated using the values in Figure 11 and Table 8.

(6) The following additional rules of the following Code Cases for elevated-temperature components shall apply:

(a) Code Case N-48, Fabrication and Installation of Elevated Temperature Components,Section III, Division I, (b) Code Case N-49, Examination of Elevated-Temperature Nuclear Components,Section III, Class 1, (c) Code Case N-50, Testing of Elevated-Temperature Components,Section III, Division 1, Class 1, and (d) Code Case N-51, Protection Against Overpressure of Elevated-Temperature Components,Section III, Division 1, Class 1.

(7) The component stamping and Data Report shall indicate this case number and the revision applied.

Appendix I Page 2 of 22 DOE-HTGR-90286, Rev. 0

Smt - ALLOWABLE STRESS INTENSITY VALUES SA-533B & SA-508 CL. 3 63 ... . . . . ... . . ... .. .. .... .. . .. ... . . .. . . . ... .. . ... ..

. S. CURV ES 55 50

R 45 40

-4 35 _.- .. . .. . .. .. . .. .. . ... .... . . ... .... ....

30 cn Ctl 25 20 . . . .*. . . . .. . . . .. ... ..... .. .. . ....... .. .. . .... .

Sm* . \ \'.\\,10 * -

15

_~~~~. .. ;,Si>..\_

i............

i>.. .. _.. .

10 5

0

600 650 700 750 80 850 900 95( 1000 1050 1100 TEMPERATURE, F FIGURE 1 - SMt VALUES FOR SA-533 GRADE B CLASS 1 AND SA-508 CLASS 3 Note: The S t values are the lower of the two stress intensity values, S (time-independent) and St (time-dependent).

Appendix I Page 3 of 22 DOE-HTGR-90286, Rev. 0

TABLE 1 SMt - ALLOWABLE STRESS INTENSITY VALUES FOR SA-533 GRADE B CLASS 1 AND SA-508 CLASS 3

- ALLOWABLE STRESS INTENSITY VALUES, ksi SMt TIME AT TEMPERATURE, HOURS TEMPERATURE I 10 30 100 300 1000

( F) 700 26.7 26.7 26.7 26.7 26.7 26.7 750 24.9 24.9 24.9 24.9 24.9 24.9 800 24.2 24.2 24.2 24.2 24.2 24.2 850 22.7 22.7 22.7 22.7 22.7 22.7 900 21.3 21.3 21.3 21.3 21.3 21.3 950 20.2 20.2 20.2 20.2 20.2 16.0 1000 18.0 18.0 18.0 18.0 14.0 9.5 Note: The S values are the lower of the two stress intensity values, Sm (time-independent) and St (time-dependent).

Appendix I Page 4 of-22 DOE-HTGR-90286, Rev. 0

S - ALLOWABLE STRESS INTENSITY VALUES t

SA-533B & SA508 CL. 3 60 55 700°F 50

--'--- C......... .....................

45 _.:,.................... ...........

40 _. .. N . ;...... .. .........................

35 - as;;;;;, oo ---- ;----.. ........................

750 30 cn (0

a-- 25 _. v._........... .............. .................... ,

co 20

. . Ni950 15 10 _ .............. ......... ... 8......<............. ........................ .........................S 5 _. .. .. . N3OO; 0

I 10 100 l1900 1CIOCI LOAD DURATION. t, Ill HOURS FIGURE 2 - St ALLOWABLE STRESS INTENSITY VALUES Appendix I Page 5 of 22 DOE-HTGR-90286, Rev. 0

TABLE 2 St ALLOWABLE STRESS INTENSITY VALUES, 1000 psi TIME AT TEMPERATURE, HOURS TEMPERATURE l 10 30 100 300 1000 O F) 700 54 54 54 53 53 52 750 54 54 54 52 49 47 800 54 53 50 48 44 40 850 53 49 46 41 37 32 900 49 43 39 34 29 24 950 45 36 31 26 22 16 1000 39 28 24 18 14 9.5 Appendix I Page 6 of 22 DOE-HTGR-90286, Rev. 0

TABLE 3 S - ALLOWABLE STRESS VALUES, YIELD STRENGTH AND TENSILE STRENGTH m VERSUS TEMPERATURE TEMPERATURE S YIELD TENSILE m STRENGTH STRENGTH

(-F) (ksi) (ksi) (ksi) 75 26.7 50.0 80.0 100 26.7 50.0 80.0 200 26.7 47.5 80.0 300 26.7 46.5 80.0 400 26.7 45.0 80.0 500 26.7 43.5 80.0 600 26.7 43.0 80.0 650 26.7 43.0 80.0 700 26.7 43.0 80.0 750 24.9 42.5 74.8 800 24.2 42.0 72.6 850 22.7 40.0 68.2 900 21.3 38.5 63.8 950 20.2 32.0 60.5 1000 18.0 34.5 53.9 Appendix I Page 7 of 22 DOE-HTGR-90286, Rev. 0

EXPECTED MlINIMlUMl STRESS-TO-RUPTURE VALUES, IN0O psi SA-533B t SA-M08 CL. 3 too . ................. .............. ............... ............... ............... ............... : 7 0

................. ........ e7

............................ ..... .. t1bu

-4

0. 900 Go .............. ...............

Q3 0e I10 ................

a: ............... ............... ............... .......... ....................

(n ............. .

7! ................................. ............... ............... ............... :..

(A)

................................. ............... ............... 1000 1 .............. .............. ............... ............

I to 100 (100 10000 I EAS IMIMTHUM TIME TO RUPTURE, HR FIGURE 3 - STRESS-TO-RUPTURE (MINFIRUM)

Page 8 of 22 DOE-HTCR-90286, Rev. 0 Appendix I

TABLE 4 EXPECTED MINIMUM STRESS-TO-RUPTURE VALUES (ksi).

TIME TO RUPTURE TEMPERATURE 1 10 30 100 300 1000 3000 10000 30000 100000 (C

F) 700 80 80 80 79 78 77 74 70 66 60 750 80 80 78 77 72 70 67 59 54 48 800 79 78 75 70 66 60 54 48 43 36 850 78 72 69 61 56 50 44 37 31 23 900 72 63 59 51 45 38 32 26 20 16 950 67 54 48 41 34 27 22 17 12 9 1000 58 44 37 29 23 18 14 9.5 7 4.5 Appendix I Page 9 of 22 DOE-HTGR-90286, Rev. 0

TABLE 5 MODULUS OF ELASTICITY VS. TEMPERATURE ELASTIC TEMPERATURE MODULUS (OF) (ksi) (xlO )

-325 31.1

-200 30.5

-100 29.9 70 29.2 200 28.5 300 28.0 400 27.4 500 27.0 600 26.4 700 25.3 800 23.9 900 22.2 1000 20.1 1100 17.8 Appendix I Page 10 of 22 DOE-HTGR-90286, Rev. 0

TABLE 6 INSTANTANEOUS COEFFICIENT OF THERMAL EXPANSION VS. TEMPERATURE Instantaneous Coefficient of Whermal Expansion in/in - 'F x 10 Temp, 'F SA-533B SA-508 Class 1 Class 3 (Mn - 1/2Mo - 1/2Ni) (3/4Ni - 1/2Mo - Cr - V) 70 7.02 6.41 100 7.13 6.53 150 7.29 6.73 200 7.45 6.93 250 7.60 7.12 300 7.74 7.30 350 7.88 7.49 400 8.01 7.66 450 8.13 7.84 500 8.25 8.03 550 8.36 8.21 600 8.46 8.35 650 8.55 8.51 700 8.63 8.64 750 8.71 8.78 800 8.78 8.90 850 8.84 9.04 900 8.90 9.13 950 8.95 9.22 1000 8.99 9.30 Appendix I Page 11 of 22 DOE-HTGR-90286, Rev. 0

TABLE 7 MEAN COEFFICIENT OF THERMAL EXPANSION VS. TEMPERATURE Mean Coefficient of Thereal Expansion in/in - 'F x 10 Temp, *F SA-533B SA-508 Class 1 Class 3 (Mn - 1/2Mo - l/2Ni) (3/4Ni - 1/2Mo - Cr - V) 70 100 7.06 6.50 150 7.16 6.57 200 7.25 6.67 250 7.34 6.77 300 7.43 6.87 350 7.50 6.98 400 7.58 7.07 450 7.63 7.15 500 7.70 7.25 550 7.77 7.34 600 7.83 7.42 650 7.90 7.52 700 7.94 7.59 750 8.00 7.68 800 8.05 7.76 850 8.10 7.85 900 8.14 7.89 950 8.19 7.98 1000 8.23 8.05 Appendix I Page 12 of 22 DOE-HTGR-90286, Rev. 0

stcrcLn, X FIGURE 4 - ISOCHRONOUS STRESS-STRAIN CURVES FOR 700'F 1.

Appendix I Page 13 of 22 DOE-HTGR-90286, Rev. 0

0)

_flx1O2

-J 02 6.4 02 02 100 tr) 4 4 4 p p= I .

0.2 0.4 0'.6 o.8 1.0 1.2 1.4 1'.6 1.8 2.0 2 sLratn, 7-FIGURE 5 - ISOCHRONOUS STRESS-STRAIN CURVES FOR 750OF Appendix I Page 14 of 22 DOE-HTGR-90286, Rev. 0

it straon, Z FIGURE 6 - ISOCHRONOUS STRESS-STRAIN CURVES FOR 800OF Appendix I Page 15 of 22 DOE-HTGR-90286, Rev. 0

CURVE OP 850 STRESS STA I -ISOCHOOVS Appendix I Page 16 of 22 E HTCR-90286 Rev. 0

I hr 3

to 3x102 3XIOZ lx10l 0.2 0.4 0.6 0.0 1.0 1.2 1.1 1.6 1.8 2.0 stratn. :

FIGURE 8 - ISOCHRONOUS STRESS-STRAIN CURVES FOR 900F Appendix I Page 17 of 22 DOE-HTGR-90286, Rev. 0

-4.

0)

CL CL stroan, t.

FIGURE 9 - ISOCHRONOUS STRESS-STRAIN CURVES FOR 950-F Appendix I Page 18 of 22 DOE-HTGR-90286, Rev. 0

he 1.0 1.2 str-cLn, X FIGURE 10 - ISOCHRONOUS STRESS-STRAIN CURVES FOR 1000*F Appendix I Page 19 of 22 DOE-HTGR-90286, Rev. 0

DESIGN FATIGUE STRAIN RANGE, £ FOR SA-533B & SAM08 CL. 3 UP TO 100e F 0.1 . . .................

.. . .. ....................- . r.................. .. .................

C C

-4 6.01 ... . . . .... .. ... . .. ... . ... . . . . .. . . . .

... ... . ... .. ... .. ... . . .. . .. ... ... e . . . .. .. .... .

(D

z cc Li) i-10000 F

CYCLIC STRAiN : RATE: x-l0f in./i:'/sec.

1E-3 .... ............ . ,.:_

10 too 10oCo I0COMO 1E6 NUtlEER OF ALLOWABLE CYCLES. N4 FIGURE 11 - DESIGN FATIGUE STRAIN RANGE FOR SA-533B CLASS 1 AND SA-508 CLASS 3 Page 20 of 22 DOE-HTGR-90286, Rev. 0 Appendix I

TABLE 8 DESIGN FATIGUE STRAIN RANGE FOR SA-533 GRADE B CLASS I & SA-508 CLASS 3 UP TO 1000*F ND, NUMBER OF Ct. STRAIN RANGE (in./irn.)

CYCLES [NOTE (1)1 AT TEMPERATURE 101 0.030 4 x 10I 0.011

°02 0.0071 2 102 0.0056 4 x 102 0.0048 0.0042 2 x 0.0037 4 0.0027 104 0.0021 2 x 104 0.00190 4 x 104 0.00170 105 0.00155 2 x 105 0.00145 4 x 105 0.00130 106 0.00120 NOTE: (1) Cycle Strain Rate: 1 x 10 3 in./in./sec.

Appendix I Page 21 of 22 DOE-ITGR-90286, Rev. 0

I4 I 4ir] I I I I I I 4 I IS 6 I 5l46I1 I 4 9 6 4 4 TXlz11111IT 1

0.9 ., ................ . . . . . . . . . .

0.8 . ................ . . . . . . . . . .

0.7 0.6 . . ............. . ,. . . . . . . .

.,..... ....................... Z . . . . . . . . . . .

0.5

......... ........... ....... ....... . .I . . . . . . .

%.e~

N 0.4 0.3 8.2

0. 1 0

e 0. 1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 I X -nNd)

FIGURE 12 - CREEP-FATIGUE DAMAGE ENVELOPE Appendix I Page 22 of 22 DOE-HTGR-90286, Rev. 0

6. Combustion Engineering Report MML-89-142, "Creep and Tensile Properties of SA508 Class 3 Forging Material," December 1989.

. I Creep and Tensile Properties of SA508 Class 3 Forging Material Project No. 900773 Job No. 98192177 MML-89-142 December 13, 1989 F. V. Ellis and J. E. Bynum Metallurgical and Materials Laboratory Combustion Engineering, Inc.

911 W. Main Street Chattanooga, TN 37402

DISCLAIMER NOTICE This report was prepared as an account of work sponsored by Oak Ridge National Laboratory. Neither Oak Ridge National Laboratory, nor Combustion Engineering, Inc., nor any person acting on their behalf:

a. makes any warranty or representation, express or implied, with respect to the accuracy, completeness, or usefulness of the information contained in this report, or that the use of any information, apparatus, method, or process disclosed in this report may not infringe privately 6wned rights; or

b. assumes any liabilities with respect to the use of, or for damages resulting from the use of any information, apparatus, method, or process disclosed in this report.

Creep and Tensile Properties of SA508 Class 3 Forging Material ABSTRACT Creep and tensile properties were measured for a single heat of SA508 Class 3 forging material. The test material was given a heat treatment of 40 hours4.62963e-4 days 0.0111 hours 6.613757e-5 weeks 1.522e-5 months at 607C to simulate an extended post weld heat treatIent. Duplicate tensile tests at strain rates of .008 min.- and .0016 min. were conducted at room temperature, 371C, 427C, 482C, 538C and 593C. The yield and tensile strengths for the SA508 Class 3 material were lower than those for SA533 Grade B Class 1 steels which had been given a shorter term PWHT. The creep tests were conducted in air out to times of 3000 hours0.0347 days 0.833 hours 0.00496 weeks 0.00114 months in the temperature range of 371C to 593C. The strain-time data were acquired and stored using a computer controlled data acquisition system. In addition to the rupture data, the minimum creep rate, time to end of secondary creep, and 0.2 percent offset tertiary time and strain data were obtained. All creep curves had the classical shape. For the primary plus secondary stage of creep, both power law and rational polynominal creep constitutive equations were good descriptors. For the tertiary creep stage, a creep equation was written based on a linear relationship between log creep strain rate and linear creep strain. Based on the tertiary creep equation, a rupture parameter was defined which correlates well with the observed rupture time. For the complete creep curve, a power law primary plus exponential tertiary model had better conformity to the observed creep curve than the Theta Projection.

Metallographic examination of creep specimens indicated no significant hardness or microstructural changes for temperatures up to 482C and out to times of 2000 hours0.0231 days 0.556 hours 0.00331 weeks 7.61e-4 months .

INTRODUCTION Currently, the ASME Boiler and Pressure Vessel Code allowable stresses in Section III for the two common nuclear pressure vessel steels, SA508 Class 3 and SA533 Grade B Class 1, are given only to a maximum temperature of 371C.

For a design application, both allowable stresses and design rules were required for limited duration elevated temperature service using these steels.

A critical review of the available elevated temperature property data for Mn-Mo-Ni steels indicates the need for further testing in order to develop allowable stresses. One of the earlier studies sponsored by PVRC (DeBarbadillo, Pense, Stout, 1966) on A-302B (Ni) steel is inappropriate since the steel meets the chemical specification of A533B but exceeds the tensile strength requirement. One important aspect of this study was the testing of weldment as well as base material . The recent results reported by Reddy and Ayres (1982) on SA533B and SA508 give only short term data (approximately 48 hours5.555556e-4 days 0.0133 hours 7.936508e-5 weeks 1.8264e-5 months maximum test duration) and emphasizes the creep strain-time properties with no rupture data provided. The American Society for Testing and Materials (ASTM) data series (Smith, 1971) gives rupture property data for six heats of Mn-Mo-Ni steels. Of these, only two heats would appear to meet the chemical and strength requirements for the SA533B Class I or SA508 Class 3 specifications. Although the data may be available from the original sources, other data needed for the allowable stress determination - time to 1% strain and time to tertiary creep - were not given in the ASTM data publication (Smith, 1971).

The objective of the current testing program was to measure the short term elevated temperature properties of SA508 Class 3 forging material required to support ASME Code approval for this steel under transient conditions in nuclear service. Only a limited testing program was conducted on the SA508 material, since similar properties to SA533 were anticipated for the two steels and a more comprehensive multiple heat testing program was being conducted by ORNL (McCoy, 1989) on SA533 Grade B Class 1 base materials, weldments, and compatible SA and SMA weld metals.

-Page 2-MATERIAL The test material was a large diameter (approximately 5.3 m) reactor vessel closure head with a nominal wall thickness of 480 mm. An initial study (Borden, 1989) of the homogeneity of the flange material included chemical analyses, metallography, Charpy and tensile testing. Macroetched sections revealed inhomogeneities in the form of dark grey streaks scattered throughout the forging which were identified as resulting from alloy element segregation.

Near the quenched surface the microstructure was 100 percent bainite. For the mid-thickness position, approximately 15 percent ferrite - 85 percent bainite was found. On a microscale, faint banding due to the varying proportions of ferrite and bainite in the microstructure was observed. The chemical composition for the SA508 Class 3 forging material is given in Table 1.

Prior to creep and tensile testing, the material was given a simulated post weld heat treatment of 607C i 15C for 40 hours4.62963e-4 days 0.0111 hours 6.613757e-5 weeks 1.522e-5 months . The Vicker's hardness (20 kg load) was 188 (equivalent Rockwell B 91) following this heat treatment.

RESULTS AND DISCUSSION The tensile and creep specimens had a 6.4 mm diameter cross section and a 31.8 mm gage length. Creep testing was performed on conventional lever-arm type constant load creep testing machines. These machines have automated beam leveling to accommodate specimen elongation. The tensile tests were conducted in a modified creep machine at two strain rates. All test data were acquired with a computerized data acquisition and process control system (Roberts and Cullen, 1973), (Bynum, 1989).

Tensile Properties Results of the tensile tests at room temperature, 371C, 427C, 482C, 538C and 593C are listed in Table 2 and shown in Figures 1-4. The nominal strain rates for these tests were .0080 min. 1 and .0016 min. 1. Duplicate tests were performed at each temperature and strain rate for a total of 24 tests in the basic program. Because of the 64 MPa difference in ultimate tensile

-Page 3-strength between the replicate tests at 371C and 0.0080 min* -1, a third test was performed at this test condition. This third test agreed with the higher strength of 523 MPa. The yield and ultimate tensile strength data shown in Figures 1 and 2 were fit to cubic spline curves shown as solid lines in the figures. These fits are non-physical and should not be used to infer expected material behavior in the temperature range where test data does not exist.

The yield and tensile strengths for the SA508 Class 3 material are consistently lower than those found for the SA533B base material tested at equivalent strain rates (McCoy, 1989). This difference may be due to the longer simulated PWHT time of 40 hours4.62963e-4 days 0.0111 hours 6.613757e-5 weeks 1.522e-5 months for SA508 material compared to the 20 hours2.314815e-4 days 0.00556 hours 3.306878e-5 weeks 7.61e-6 months for the SA533B material. However, it should be noted that the room-temperature strength values are considerably less than those obtained on the same material tested at more "normal" strain rates in a different machine in the Metallurgical and Materials Laboratory after heat treating a section of the forging (Borden, 1989). These nine tests were conducted at strain rates on the order of .050 min.-1 to verify that the heat treatment resulted in acceptable tensile properties. The yield strengths from the nine tests ranged from 416 MPa to 460 MPa and the ultimate tensile strengths ranged from 572 MPa to 611 MPa. An additional room-temperature test conducted at an intermediate strain rate of .024 min. 1 (test AKNZ in Table 2) confirmed the higher strengths of the .050 min.-1 strain rate tests. These higher strain rates are more representative of normal production type testing. Therefore, it appears that the tensile properties of the SA508 heat used in this investigation are highly strain-rate sensitive.

Tensile ductility of the SA508 material is shown in Figures 3 and 4 as a function of temperature at the two different strain rates. The total elongation (31.2 mm gage length) exceeds 20 percent over the entire temperature range. While both the reduction in area and the total elongation increase as the test temperature increases, the uniform elongation decreases with values approaching 2 percent at 593C.

Creep Properties The criteria for establishing the ASME code time-dependent allowable stresses, St, for a specified time, t, is defined as the lower of the

. I

-Page 4-following three time dependent values: (1) two-thirds of the minimum stress to produce rupture (2)80 percent of minimum stress to cause tertiary creep and (3) the minimum stress to produce one percent total strain. For the intended application, the 103 hour0.00119 days 0.0286 hours 1.703042e-4 weeks 3.91915e-5 months allowable stresses were required for temperatures from 371C to 538C. The approach taken to gather the necessary data for the code analysis was to perform short term (approximately 3000 hours0.0347 days 0.833 hours 0.00496 weeks 0.00114 months maximum duration) creep and stress rupture tests at five temperatures (371C, 427C, 454C, 482C, and 593C) with three stress levels at each temperature. The data analysis will be performed by ORNL and will use the creep and rupture properties obtained for both the SA533B and SA508 materials.

Strain versus time curves for each of the respective isothermal test temperatures of 371C, 427C, 482C, 538C and 593C are shown in Figures 5-9.

Over the full range of stresses and temperatures, these curves exhibited the classical shape having primary, secondary and tertiary regimes. Time to a given creep strain data for the SA508 base material are given in Table 3. The creep strain is found by subtracting the loading strain (Table 3) from the measured total strain. Other measures of the creep curve -- minimum creep rate, Monkman-Grant constant (minimum creep rate times rupture time), time to end of secondary creep (t2), and 0.2% offset tertiary time (to 2%) and strain -- are given in Table 4.

Uniaxial stress rupture properties and Larson-Miller Parameter values are given in Table 5. The Larson-Miller Parameter, P, is defined as P = T(C + Log tr) where T is absolute temperature, tr is the rupture time, and C is the Larson Miller constant. The universal constant value of 20 was used in Table 5. The rupture test results are shown in Figure 10 plotted as log rupture stress versus log time. Various correlations have been proposed for relating rupture time to creep parameters other than stress. Figure 11 shows the minimum creep rate versus rupture time relation of Monkman-Grant. The time to end of secondary creep and time to 0.2% offset tertiary versus rupture time relations are shown in Figures 12 and 13. There is an apparent temperature dependence for both t2 vs tr and the to0 2% vs tr relations which could be easily modeled using heat-centered techniques.

-Page 5-The time to one percent creep strain data is shown in Figure 14 on a log stress versus log time basis and in Figures 15-18 on a log stress versus Larson-Miller Parameter basis using Larson-Miller constants (C) of 20.0, 20.8, 23.1 and 24.9, respectively. The solid line drawn in Figure 15 represents the mean of the 533B Class I material calculated by the relation given by McCoy (1989). The SA508 material is weaker in creep at the high stress region and tends to merge in the lower stress (approximately 55 MPa) region. This crossover in creep strength is similar to that reported by Pense and Stout (1966) for A302B steel. They showed that the 10,000 hour0 days 0 hours 0 weeks 0 months rupture strength for Q&T steel was vastly superior than that for N&T steel up to temperatures of 510C while the N&T steel had slightly better strengths at 538C and 593C. This change in rupture strengths was associated with the accelerated rate of spheroidization at the higher test temperatures for the Q&T microstructure compared to the coarser N&T microstructure. A similar microstructural based explanation may apply in the current study with the predominantly bainite microstructure of SA533B steel-being less stable at the higher test temperatures compared to the ferrite-bainite (slower cooling rate) microstructure of the SA508 material. The solid line shown in Figures 16, 17 and 18 was drawn using a log stress plus stress master curve and a Larson Miller parameter fit to the data in Table 2 for different stress ranges.

Comparison of the data in these figures indicates that the lower constant values correlate the low stress-high temperature data while higher constant values correlate the higher stress-low temperature data. It should be noted that the Dorn parameter correlations had lower standard deviations in log time to one percent creep strain than the Larson-Miller parameter.

Creep Constitutive Equations The measured engineering strain, e, was converted to true strain, c, using the following relation e ln (1 + e)

Creep strain was found by subtracting the loading strain from the total strain. The resulting creep strain versus time curves were fit using several

-Page 6-relationships. Nonlinear-least squares fits were performed on the complete creep curves (through approximately 6% creep strain), the tertiary portion of the curves, and the portion from beginning through the end of secondary creep.

The primary plus secondary curve fit coefficients and models are given in Tables 6, 7 and 8. In general, the power law and the rational polynominal description were in good conformity to the creep curves while the exponential primary law tended to underpredict the initial creep and had higher standard deviations in creep strain than the other models.

The tertiary creep law is based on the observation that the plot of log creep strain rate versus creep strain is linear in the tertiary regime (Sandstrom and Kondyr, 1976). Using this relationship and integration, the following creep strain-time equation can be derived cc = -A In (C - Bt) where c is creep strain, t is time, and A, B and C are constants. A tertiary creep rupture parameter, P, can be defined by evaluating the tertiary equation at rupture and solving for the rupture time as follows P = tr = (C - e ,r/A)/B where er is the true creep rupture strain.

For values of er/A >3, the rupture parameter is given by the following approximate relation P = C/B The values for the regression coefficients A, B, and C for the tertiary creep model are given in Table 9. The exact and approximate rupture parameters are also given in Table 9. Figure 19 shows a plot of log rupture time versus log tertiary creep rupture parameter (C/B) for the SA508 for Class 3 forging

-Page 7-material. It is intended that some of the creep tests in progress will be continued in order to extend this correlation out to longer times.

The complete creep curve models and regression coefficients are given in Tables 10 to 13. In general, it should be noted that the Theta Projection tended to underpredict the observed strain-time response in the primary regime while the power law primary plus exponential tertiary model had excellent conformity to the entire creep curve.

Metallography Creep specimens exposed for the longest time at each temperature were metallographically prepared and examined to characterize the metallurgical structure. At 371C and 427C, no readily discernible microstructural changes or hardness decrease was observed as a result of the creep exposure for 3149 hours0.0364 days 0.875 hours 0.00521 weeks 0.0012 months and 2566 hours0.0297 days 0.713 hours 0.00424 weeks 9.76363e-4 months , respectively. The specimen that ruptured in 2035.6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months at 482C (LMP = 17600) also showed only very slight changes and is similar in both time and temperature to the maximum anticipated for the low probability transient condition. Both specimens exposed at the higher test temperatures had decreased hardness values and from slight to significant spheroidization of the microstructure. Figure 20 compares the microstructure of the creep specimens at 427C, 482C, 538C and 593C. The hardness values, converted to Rockwell B scale from measured Vicker's (20 kg load), and exposure duration are also given in Figure 20.

SUMMARY

Creep and tensile property data have been obtained for a single heat of SA508 Class 3 forging material in order to support a proposed Code Case to extend the allowable stress up to a maximum temperature of 538C and out to times of 103 hours0.00119 days 0.0286 hours 1.703042e-4 weeks 3.91915e-5 months .

-Page 8-REFERENCES Borden, M. P., 1989, "Properties of SA508-Class 3 Reactor Vessel Closure Head Flange," Metallurgical and Materials Laboratory, Final Report MML-89-111, Combustion Engineering, Chattanooga, TN.

Bynum, J. E., 1989, "Automating a Creep Testing Laboratory for Real-Time Data Acquisition and Control Using an HP 1000 Computer System,", Proceedings of the 1989 INTEREX HP Users Conference, San Francisco, Paper 1004.

DeBarbadillo, J. J., Pense, A. W., and Stout, R. D., 1966, "The Creep-Rupture Properties of Pressure Vessel Steels - Part II," Welding Research Supplement, Welding Journal, pp. 357s-367s.

McCoy, H. E., 1989, "Tensile and Creep Properties of SA533 Grade B Class 1 Steel," Oak Ridge National Laboratory, Oak Ridge, TN.

Pense, A. W., and Stout, R. D., 1966, "Characterization of Heat Treated Pressure Vessel Steels for Elevated Temperature Service," Heat-Treated Steels for Elevated Temnerature Service, ASME, NY, NY, pp. 8-26.

Reddy, G. B., and Ayres, D. J., 1982, "High Temperature Elastic-Plastic and Creep Properties for SA533 Grade B Class 1 and SA508 Materials," EPRI NP-2763, Electric Power Research Institute, Palo Alto, CA.

Roberts, B. W., and Cullen, T. M., 1973, "Computerized System Collects Creep Data," Metals Progress, pp. 49-64.

Sandstrom, R., and Kondyr, A., 1976, "Model for Tertiary-Creep in Mo-and Cr-Mo-Steels," Mechanical Behavior of Metals, Vol. 2, Pergamon Press, NY, NY, pp. 275-284.

Smith, G. V., 1971, Evaluations of the Elevated Temperature Tensile and Creeo-Rupture Properties of C-Mo. Mn-Mo and Mn-Mo-Ni Steels, ASTM Data Series Publication DS-47, American Society for Testing and Materials, Philadelphia, PA.

Tabl e 1 Chemical Composition of SA508 Class 3 Forging Material Composition SA508 Class 3 El ement (Weiaht Percent) Reaui renents C 0.19 0.25 max.

Mn 1.35 1.20 - 1.50 p 0.006 0.025i max.

S 0.003 0.025i max.

Si 0.31 0.15 - 0.40 Ni 0.78 0.40 - 1.00 Cr 0.14 0.25 max.

Mo 0.49 0.45 - 0.60 V 0.005 0.05 max.

Cb 0.001 Ti 0.001 Co 0.014 Cu 0.06 Al 0.012 B <0.001 w <0.001 As 0.007 Sn 0.006 Zr <0.001

Table 2 Tensile Properties of SA508 Class 3 Base Material Test Strain Yield Tensile Total Uniform Reduction Specimen Temperature Rate Strength Strength Elongation Elongation Of Area Code 0C *L 10-3/min. Mna ksi MDa ksi AJZZ 24 75 8.0 348 50.5 494 71.6 23.2 10.71 66.8 AKAZ 24 75 8.0 348 50.4 501 72.6 26.0 11.53 69.0 AKBZ 24 75 1.6 352 51.0 496 71.9 24.8 11.06 66.8 AKDZ 24 75 1.6 338 49.0 487 70.6 26.0 11.61 67.3 AKMZ 371 700 8.0 335 48.6 523 75.8 24.8 9.87 69.0 AKHZ 371 700 8.0 318 46.1 487 70.6 24.0 9.41 64.5 ALAZ 371 700 8.0 350 50.8 523 75.8 25.6 9.92 66.8 AKEZ 371 700 1.6 301 43.6 465 67.4 22.4 9.51 69.1 AKFZ 371 700 1.6 294 42.6 459 66.6 23.0 9.14 64.5 AKKZ 427 800 8.0 332 48.2 463 67.1 24.8 8.02 69.1 AKLZ 427 800 8.0 348 50.4 466 67.6 23.2 7.31 69.1 AKIZ 427 800 1.6 328 47.6 452 65.6 23.2 7.91 66.6 AKJZ 427 800 1.6 331 48.0 450 65.2 23.2 7.13 69.1 AKQZ 482 900 8.0 308 44.6 401 58.2 26.4 6.66 79.1 AKRZ 482 900 8.0 325 47.2 404 58.6 24.8 6.16 77.1 AKOZ 482 900 1.6 328 47.6 383 55.6 24.8 4.73 80.9 AKPZ 482 900 1.6 316 45.9 383 55.5 24.8 6.05 79.1 AKUZ 538 1000 8.0 292 42.3 330 47.8 33.2 3.14 82.6 AKVZ 538 1000 8.0 288 41.7 332 48.1 36.0 3.36 84.2 AKSZ 538 1000 1.6 270 39.2 303 43.9 41.6 2.73 85.0 AKTZ 538 1000 1.6 279 40.4 311 45.1 41.6 2.12 85.0 AKXZ 593 1100 8.0 223 32.4 254 36.8 44.8 2.72 88.6 AKZZ 593 1100 8.0 225 32.6 258 37.4 40.0 2.07 87.2 AKWZ 593 1100 1.6 194 28.1 220 31.9 41.6 2.18 85.7 AKYZ 593 1100 1.6 191 27.7 225 32.6 47.2 2.02 87.2 AKNZ 24 75 24.0 432 62.6 585 84.9 27.2 11.46 66.8

Table 3 Time to a Given Creep Strain Data for SA508 Class 3 Forging Material Code Stress Loading Time to a Given CreeD Strain - Hours (KSI) AMPa) Strain-% 0.2% 0.5% 1% 2% 5%

700F (371C)

AMH 67 462.0 2.355 - 12 102 495 AMA 65 448.2 2.357 - 10.6 121 506 1952 ALU 60 413.7 1.597 .4 117 877 800F (427C)

ALP 60 413.7 1.88( 1.6 8 AMI 53 365.4 .97' .3 5.2 24 85 299 AMB 50 344.8 0.44! - 7.8 74 328 1177 ALS 45 310.3 0.237 16.5 198 1036 850F (454C)

AMJ 45 310.3 0.33' 1.2 11.0 45 134 387 AMK 40 275.8 0.18C6 26.6 191 616 1436 900F (482C)

ALO 45 310.3 0.240 .2 1.8 5.8 15 35 AMD 40 275.8 0.238 1.6 8.6 26.9 67 166 ALY 35 241.3 0.160 12.3 71 207 475 1056 ALT 30 206.9 0.124 38.7 250 687 1519 AMF 27 186.2 0.165 141.6 619 1500 1000F (538C)

ALW 30 206.9 .119 0.9 3.5 8.1 17 39 AME 25 172.4 .043 3.6 13 7 35.0 AMO 25 172.4 .115 2.4 10.1 26.2 59 139 ALR* 20 137.9 .060 9.1 42 129 302 692 ALM 15 103.4 .098 36.7 194 531 1168 2500 AMC 12 82.7 .033 112.7 629 1492 llOOF (593C)

ALV 15 103.4 .038 1.0 5.1 14.5 34 82 AMG 12 82.7 .059 5.1 19.2 49 108 238 AMP 10 69.0 .033 8.7 38 94 ALN 8 55.2 .033 19.4 96 223 431 876 ALX 4 27.6 .018 165.3 650 1432 2726

Specimen overtemperature by 10C for approximately 1 hour1.157407e-5 days 2.777778e-4 hours 1.653439e-6 weeks 3.805e-7 months .

Table 4 Minimum Creep Rate, Monkman-Grant Constant, Time to End of Secondary Creep, 0.2% Offset Tertiary Time and Strain for SA508 Class 3 Forging Material Monkman-Code Stress Minimum Creep Grant t2 0.2% Offset Tertiary (KSI) (MPal Rate - %/Hour (VA (Hrs.) Time-Hrs. Rfrain-V.

700F (371C)

AMH 67 462.0 .0075 7.2 500 636 8.6 AMA 65 448.2 .002 1400 2311 8.2 800F (427C)

ALP 60 413.7 .38 5.1 9.5 7.4 AMI 53 365.4 .0138 9.5 220 354 6.9 AMB 50 344.8 .0034 8.7 800 1247 5.7 850F (454C)

AMJ 45 310.3 8.5 165 308 4.3 900F (482C)

ALO 45 310.3 .107 6.9 12 24 3.4 AMD 40 275.8 .025 9.1 60 130 4.0 ALY 35 241.3 .0035 7.1 250 653 2.9 ALT 30 206.9 .0012 1000 2374 3.3 1000F (538C)

ALW 30 206.9 .106 8.0 14 28 3.4 AMO 25 172.4 .030 9.2 50 102 3.6 ALR 20 137.9 .0056 7.3 200 466 3.2 ALM 15 103.4 .0015 900 1628 3.0 HOOF (593C)

ALV 15 103.4 .050 9.9 30 59 3.5 AMG 12 82.7 .016 9.5 70 143 2.7 AMP 10 69.0 .0089 90 ALN 8 55.2 .0037 7.6 125 387 1.8 ALX 4 27.6 .00063 1800 2769 2.1

Table 5 Uniaxial Stress Rupture Properties of SA508 Class 3 Forging Material Code Stress Rupture Time LMP* Elong. Reduction (KSI) (MPa) (Hours) (Y) In Area-%

700F (371C)

ALQ 75 517.1 .05 12040 23 64 AMH 67 462.0 959.8 14800 25 57 800F (427C)

ALP 60 413.7 13.5 14790 21 52 AMI 53 365.4 688.2 15980 24 69 AMB 50 344.8 2566.4 16390 32 73 850F (454C)

AMJ 45 310.3 853.7 16670 49 69 900F (482C)

ALO 45 310.3 64.1 16470 32 74 AMD 40 275.0 365.3 17040 49 75 ALY 35 241.3 2035.6 17600 30 37 IOOOF (538C)

ALW 30 206.9 75.9 17740 25 80 AMO 25 172.4 305.5 18240 40 65 ALR 20 137.9 1302.2 19130 36 75 IIOOF (593C)

ALV 15 103.4 197.7 19320 60 91 AMG 12 82.7 591.7 19730 80 87 ALN 8 55.2 2051.0 20190 60 92

LMP is the Larson-Miller Parameter, P, defined as P = T(K) (20 + Log1 o Time (Hours))

Table 6 Power Law Primary Plus Secondary Creep Curve Regression Coefficients Streus Code (KSI ) K n (%/Or.

700F (371C)

AMH 67 .587 .162 .00627 AMA 65 .300 .187 .00197 ALU 60 .221 .132 .000642 850F (454C)

AMJ 45 .18246 .333 .0078813 AMK 40 SOOF (427C)

AMI 53 .2663 .320 .009981 AMB 50 .3106 .217 .. .002683 ALS 45 .0795 .325 .000276 ALS 45 .0768 .340 .000180 900F (482C)

ALO 45 .28826 .301 .08730 AMD 40 .15295 .358 .01879 ALY 35 .06503 .376 .00246 ALT 30 .065354 .224 .0010527 1000F (538C)

ALW 30 .1079 .287 .09879 AME 25 .072465 .565 .013029 AMO 25 .09816 .401 .02420 ALR 20 .07092 .400 .003976 ALM 15 .053426 .291 .0012715 AMC 12 .0304 .371 .000226 IIOOF (593C)

ALY 15 .16338 .293 .044475 AMG 12 .059809 .559 .009615 ALN 8 .05734 .315 .00269 ALX 4 .010775 .378 .00056419 E = Ktn + i t

+ 1 KSI = 6.895 MPa

Table 7 Exponential Primary Plus Secondary Creep Curve Regression Coefficients Strews 6 0 Code (KSI ) (%i (1/Ar.) (%/Pr.I 700F (371C)

AMH 67 .998 .694 .00776 AMA 65 ALU 60 .352 1.058 .00093 800F (427C)

AMI 53 .720 .153 .01391 AMB 50 .669 .233 .00396 ALS 45 .310 .055 .00085 850F (454C)

AMJ 45 .482 .178 .011208 AMK 40 900F (482C)

ALO 45 .343 1.482 .11072 AMD 40 .308 .365 .02511 ALY 35 .238 .093 .00373 ALT 30 .172 .071 .00127 1000F (538C)

ALW 30 .124 1.871 .10640 AMO 25 .203 .388 .02994 AME 25 .197 .246 .02276 ALR 20 .268 .0912 .00563 ALM 15 .182 .0577 .00158 AMC 12 llOOF (593C)

ALV 15 .257 .758 .05063 AMG 12 .202 .158 .01615 ALN 8 .132 .260 .00383 AUX 4 .075 .0426 .000629

-8 2 t C = el (1

-e )+ mt

+ I KSI - 6.895 MPa

Table 8 Rational Polynominal Creep Curve Regression Coefficients For SA508 Class 3 Forging Material Strems C p Code (KSI ) (%) (1/Hr.)I (%/Pr.)

700F (371C)

AMH 67 1.0646 .83497 .00756 AMA 65 ALU 60 800F (427C)

AMI 53 .82860 .20909 .013376 AMB 50 .80623 .20148 .0034609 ALS 45 850F (454C)

ANJ 45 .56213 .23147 .010713 AMK 40 900F (482C)

ALO 45 .40919 1.9423 .10601 AMD 40 .36681 .47142 .024190 ALY 35 .27355 .11611 .0035508 ALT 30 .21307 .076690 .0011496 AMF 27 1000F (538C)

ANW 30 .14368 2.4885 .10514 AMO 25 .24667 .46827 .029147 AME 25 .26735 .23220 .021613 ALR 20 .31890 .11536 .0054166 ALM 15 .22626 .061773 .0014685 AMC 12 IIOOF (593C)

ALV 15 .29389 1.0678 .049690 AMG 12 .27137 .15721 .015407 ALN 8 .15130 .35085 .0036738 ALX 4 fc = Cpt/(1 + pt) + Emt

+ 1 KSI = 6.895 MPa

Table 9 Tertiary Creep Curve Regression Coefficients for SA508 Class 3 Forging Material Rupture Par.

Code Temp. Stress A to/%

B-1 C Monkman AB (Hours)

(F) (KSI) I/0e)IIII II Grant-% 1%/Hr.. I C/B P AMH 700 67 3.5472 .00044010 .46709 7.2 .00156 1061 1057 AMA 700 65 7.0278 .0001498 .79966 .00105 5336 AMI 800 53 8.1765 .001001 .85361 9.5 .00818 853 781 AMB 800 50 6.5431 .00028077 .80651 8.7 .00184 2872 2821 AMJ 850 45 7.9323 .0009288 .90134 8.5 .00737 970 963 ALO 900 45 5.6687 .013902 .91165 6.9 .07880 66 65 AMD 900 40 6.6791 .0026266 .91753 9.1 .01754 349 348 ALY 900 35 5.8241 .00047703 .93761 7.1 .00278 1966 1942 ALT 900 30 5.7847 .00014585 .93170 .000843 6388 ANW 1000 30 6.1578 .012809 .94751 8.0 .07888 74 72 AMO 1000 25 7.4320 .0031179 .95118 9.2 .02317 305 302 ALR 1000 20 6.1206 .00070041 .93519 7.3 .00428 1335 1326 ALM 1000 15 5.8514 .0002100 .95997 .00123 4571 ALV 1100 15 7.6095 .0050559 .94003 9.9 .03847 186 186 AMG 1100 12 7.4976 .0019004 .97388 9.5 .01425 512 512 ALN 1100 8 7.5924 .0005473 1.0062 7.6 .004155 1838 .1835 ALX 1100 4 2.648 .00018357 .97672 .000486 5320 EC - -A ln (C - Bt)

P = (C-e )/B; Define rupture parameter, P, as for er/A >3 P - C/B

Table 10 Modified Theta Projection Regression Coefficients For SA508 Class 3 Forging Material Code Stres A a-1 B (KSI I f VO1 (hr I (,L, 700F (371C)

AMH 67 4.5574 .0045705 .084342 800F (427C)

AMI 53 4.3012 .00634 .20527 850F (454C)

AMJ 45 3.2824 .0060798 .18871 900F (482C)

ALO 45 2.5378 .065284 .28846 AMD 40 2.7300 .013710 .27620 ALY 35 2.3486 .0019268 .42162 100OF (538C)

ALW 30 2.3082 .043787 .67508 ALR 20 2.4695 .0029282 .41212 ALM 15 2.1530 .000800 .4685 110OF (593C)

ALV 15 2.8000 .022679 .46211 AMG 12 2.3530 .0071674 .65130 ALN 8 1.6373 .001446 1.4454

- -= _(Vt. -. ,ft-

-1)

EC A (1-e -) + B (e -

Rupture Parameter, P, defined as P - 1/a In ((er - A)/B) where Er = true rupture creep strain

+ 1 KSI = 6.895 MPa

Table 11 Theta Projection Regression Coefficients For SA508 Class 3 Forging Material Code Stress 91 02 03 64 (KSI+) (%) (hrIl) (%) (hri) 700F (371C)

AMH 67 1.184 5 .23784 7.8177 .00076069 800F (427C)

AMI 53 .989108 .07042 8.1661 .0012956 850F (454C)

AMJ 45 .645(08 .08376 8.7066 .0010257 900F (482C)

ALO 45 .52630 .60865 3.5850 .022633 AMD 40 .43781 .16860 6.3077 .0032132 ALY 35 .36425 .031352 3.7160 .00075258 IOOOF (538C)

ANW 30 .42294 .22343 3.3333 .022001 ALR 20 .40627 .036399 3.9854 .0010893 ALM 15 .33558 .011152 3.1011 .00036104 IIOOF (593C)

ALV 15 .35939 .36571 6.7031 .0063152 AMG 12 .30535 .090205 4.2198 .0030879 ALN 8 .14050 .23369 3.6661 .0009461 cc = B(1 - e02 ) + O3 (eO4t-1)

+ 1 KSI - 6.895 MPa

Table 12 Power Law Primary Plus Exponential Tertiary Creep Curve Regression Coefficients For SA508 Class 3 Forging Material Code Streis K n 93 84 (KSI I W% (H/Hr.I 700F (371C)

AMH 67 .43571 .292 1.2320 .0019527 80OF (427C)

AMI 53 .22642 .426 1.6185 .0029152 850F (454C)

AMJ 45 .16702 .406 2.0696 .0023087 900F (482C)

ALO 45 .31873 .424 1.4753 .034231 AIMID 40 .14802 .446 2.1305 .0057862 ALY 35 .061437 .409 1.8866 .0010469 IIOOF (538C)

ALW 30 .15403 .683 1.0910 .034275 ALR 20 .067694 .445 1.8022 .0015968 ALM 15 .040160 .398 1.8292 .00046175 110OF (593C)

ALV 15 .16942 .426 3.1328 .0096733 AMG 12 .07616 .455 2.4914 .0039988 ALN 8 .09017 .104 3.5066 .00096906 cc = KtL+ 3 (eO4t.1)

+ 1 KSI = 6.895 MPa

Table 13 Power Law Primary Plus Tertiary Creep Curve Regression Coefficients For SA508 Class 3 Forging Material Code Strews K n A B C (I(-T I ( Y. (Hr. I 900F (482C)

AMD 40 .22765 .298 5.4308 .0031433 1.0153 ALY 35 .36886 .122 5.2892 .00056307 1.0640 IOOOF (538C ALW 30 .13257 .488 4.9427 .014575 1.0003 ALR 20 .22238 .201 5.2717 .0008210 1.0320 1000F (538C)

ALV 15 .38538 .185 7.0709 .0056853 1.0319 EC = Ktn Aln (C - Bt)

+ 1 KSI = 6.895 MPa

FIGURE CAPTIONS Figure 1. Tensile strength properties for SA508 Class 3 forginy material as a function of temperature at a nominal strain rate of .0080 min.- .

Figure 2. Tensile strength properties for SA508 Class 3 forgipg material as a function of temperature at a nominal strain rate of .0016 min.

Figure 3. Ductility of SA508 Class 3 forging _qaterial as a function of temperature at a nominal strain rate of .0080 min.

Figure 4. Ductility of SA508 Class 3 forging 9 aterial as a function of temperature at a nominal strain rate of .0016 min.- .

Figure 5. Strain time curves for SA508 material at 371C.

Figure 6. Strain time curves for SA508 material at 427C.

Figure 7. Strain time curves for SA508 material at 482C.

Figure 8. Strain time curves for SA508 material at 538C.

Figure 9. Strain time curves for SA508 material at 593C.

Figure 10. Stress rupture properties for SA508 base material.

Figure 11. Log minimum creep rate versus log rupture time data for SA508 base material.

Figure 12. Log time to end of secondary creep versus log rupture time for SA508 base material.

Figure 13. Log time to 0.2% offset tertiary creep versus log rupture time for SA508 base material.

Figure 14. Log stress versus log time to one percent creep strain data.

Figure 15. Log stress versus Larson-Miller parameter (C = 20) data for time to one percent creep strain.

Figure 16. Log stress versus Larson-Miller parameter (C - 21) data for time to one percent creep strain.

Figure 17. Log stress versus Larson-Miller parameter (C = 23) data for time to one percent creep strain.

Figure 18. Log stress versus Larson-Miller parameter (C = 25) data for time to one percent creep strain.

Figure 19. Log rupture time versus log tertiary creep rupture parameter.

Figure 20. Comparison of microstructure and hardness values for creep specimens tested at 427C, 482C, 538C and 593C.

600 I I I I I I, I C = 0.0080 min.71 500 0

0 CL 400 I

-C

.4-CD 300

-Ie-C/)

200 o Ultimate Tensile Strength O Yield Strength I I I

I

I

100 I I 0 100 200 300 400 500 600 Temperature - °C

.600 I I I -- i I I- I I I -- _ _ I__9_

e = 0.0016 min.-I 500 a-400 I

C-

-O.-

0) 3 C 00 L2 U) 200 _ 0 Ultimate Tensile Strength 0 Yield Strength I

100 I I I I I I

I I I 0 100 200 300 400 500 600 Temperature - °C

100 a Reduction in Area = 0.0080 mini' 0 Total Elongation 80 _ O Uniform ElongationH 600 1 60__

00 in 40 - 0 20_ , s , - .. 0 -

0 0 100 200 300 400 500 600 Tempera-lure - °C

100 I l I I r I , I I o Reduction in Area E = 0.0016 min.-i o Total Elongation 80 0 Uniform Elongation 2 1 60 0 03 8

1 60

-O0-4-

0 C._

40 - 0 0 U

in

§ 0 0 0 20 4

I I I I II 0 U I I I I I 0,

0 100 200 300 400 500 600 Temperature °C

E2X I 0 .000 ' ~ -

T2 3200. 000 SI El 0.00000 S E2 . i8000 I AMH T 2 AMA/

3 ALU/

R A

I N

IN IN El TII T2 ii TIME-HRS

s R

A N

IN IN I El'-

TI T2 TIME-HRS

S I A T 2 A 3 A R 4 AT

. 5 A F A

N XN IN I*

TI T2 TIME-HRSi

E2 T2 38oo . oso E: O .000001 S E2 . 2eooo IME-AME T

-9I AMO R 4 ALA

-5 ALMX A lS AMC/

a -,

I T

TIME-HRS

S T

R A

N IN IN Ti T2 TIME-HRS

2.8 L

0 .

G X a x

S 2.4 x T

R A E

S S

A M a P 2.0 TEMPERATURE - C a + 371 3

427 0 454 X 482 A 538 13 593 a 1.2. I 4

1. 2. 3. 4.

LOG TIME - HOURS

G MI TEMPERATURE - C N

I + 371

427 I

0 454 U X 482 U ,1 x ^A 538 R l E

E X P

13 R la A

a T -2.

E l A

H 0

U -3. 2 R 1. 2. 3. 4.

LOG TIME - HOURS

4.0 .

TEMPERATURE - C L + 371 0

427 G 0 454 X 482 3.0 A 538 T a 513

I M F E

x H

al 0

U 2.0 R ai S

13 i .0 I I 1.0 2.0 3.0 4.0 LOG RUPTURE TIME - HOURS

4.0 6 I TEMPERATURE - C L + 371 0

427 G 0 454 X 482 .4 3.0 A 538 T 593 I OF M

E Of a

/

H a 0

U 2.0 R

S a i1.0 I- --- -- ----- I 1.0 2.0 3.0 4.0 LOG RUPTURE TIME - HOURS

4- +

L X 0*

L X 0 .

G 2.4 0 x a x

A X S A A x TA R

E S 13 A S 2.0 03 A

/ 1 M a P TEMPERATURE - C a + 371 i.6 '* 427 0 454 X 482 A 538 1 13 593 1.2 - -

.5 1.0 1.5 2.0 2.5 3.0 3.5 LOG TIME -- HOURS

2.8 L

0 G 2.4 S

T R

E S

S 2.0 M

P a

1.6 1.2 12000. 14000. 16000. 18000. 20000. 22000.

PARAMETER C - 20 I

2.8 1* F

+

f.

L )o 94 0 X 0 G 2.4 x a X S X AA T

A E

S S 2.0 M

P TEMPERATURE - C a

1.6 r

+

0 X

371 427 454 482 n

A 538 o 593 4 0 L I 14000. +/-6000. 18000. 20000. 22000.

PARAMETER C - 21

2.8 I A I - a

+ IF

+

be K

L CK

0 x O G 2.4 x S

T R

E S

S 2.0 M

P TEMPERATURE - C a + 371 i.S

427 0 454 X 482 A 538 o 593 4 .

0 L 0

14000. 16000. 18000. 20000. 22000. 24000.

PARAMETER C - 23

2.8 i 1-

+ F F9 9.

L a

G 2.4A s

T E

S S 2.0 M

TEMPERATURE - C a + 371 1.8

427 0 454 X 482 Ai 538 El 593

.& . L I I I I 18000. 18000. 20000. 22000. 24000. 26000.

PARAMETER C - 25

4.0 L

0 G

3.0 T

I M

E H

0 U 2.0 S

1.0 i.0 2.0 3.0 4.0 LOG PARAMETER

427C 2566.4 Hours 482C 2035.6 Hours 92 RB 91 RB 538C 1302.2 Hours 593C 2051.0 Hours 88 RB 81 RB Figure 20. Comparison of microstructure and hardness values for creep specimens tested at 427C, 482C, 538C and 593C (1OOX).

APPENDIX A QA Documentation The following copies of the QA documents for creep testing machine 5 are typical of the records on file for all such machines in the Metallurgical and Materials Laboratory of Combustion Engineering, Inc. located in Chattanooga, Tennessee. Copies of documents for other machines are available on request.

I.

. I CREEP MACHINE CALIBRATION Creep Machine Morehouse Proving Ring Machine No. : S Serial No. : 4563 C-£ Asset No.: 8023 Capacity: 6000 pounds Machine Capacity: 12000. pounds NBS Reference No. : 737.229759 Lever Arm Ratio : 20:1 Calibration Date: 08/07/89 Proving Ring Temperature = 24.2 C Prov.

Prov. Prov. Ring Prov. Error Pan Applied Ring Ring Def 1. Ring of Weight Load Reading DefI. at 23C Load Mach Lbs. Lbs. Div. Div. Div. Lbs. Lbs. Error

0. 0. 45.4

40. 800. 1 07.3 61.9 61.88 798. 97 1. 03 .13 1 00 2000 199.8 154.4 154.35 2003.88 -3.88 -. 19

_ 160 3200 291.1 245.7 245.62 3204.81 -4.81 -. 15 220 4400 381. 6 336.2 336.09 4406.96 -6.96 -. 16 280 5600 471. 1 425.7 425. 57 5607.69 -7.69 -. 14 Date: 30 OCT., 1989 Calibrated by

______________ __EL_XA__

A-1

CREEP MACHINE LOAD CELL CALIBRATION Creep Machine No. : 5 Load Cell Manufacturer : Revere Model No. : USPI-S-A Serial No. : 245043 Capacity : 5000. pounds.

Excitation Voltage : 16.850 volts Pan Applied Load Cell .Load Cell Load Cell Weight Load Reading Reading Error %

Lbs. Lbs. mv Lbs. Lbs. Error

40. 800. 8. 057 805.67 -5. 67 -: 709 100 OOO. 20. 058 2005.82 -5.82 -. 291 160 3200. 32. 069 3206.87 -6.87 2O

_ 4400. 44. 053 4405.27 -5.27 -.120 280 5600. 56. 032 5603.24 -3.24 -. 058 Date : 17 NOV., 1989 Calibrated by A-2

ASL CAPACITANCE TRANSDUCER CALIBRATION ASL Transducer Calibration Source Model No. : 1083B Federal Proaucts Corp. Micrometer Serial No. : 1109 Klodel No. : 6888 Creep Machine : S C-E Asset No. : 8220 Transducer : A Resolution : 0.00001 inch Microieter Micrometer Transducer Transducer Error of Reading Change Reading Change Transducer inch Inch inch inch inch Error 0.00000 . 903886

. 10000 .10000 .803898 .095988 .00 01e .O12

.20000 .20000 .704067 . 199819 .000181 .091

.30000 .30000 .604097 .299789 .000211 . 070

.40000 .40000 .50407t .399815 .00 0185 .046

.50000 .50000 . 404023 . 49863 .000137 . 027

.60000 .60000 .303936 .559950 .000050 .008

.70000 .70000 .203897 .699989 .000011

.80000 .80000 .103831 .800055 -. 000055 -. 007

.85000 .85000 .053773 .850113 -. 000113 -. 013 Date : 17 NOV., 1989 Calibrated by

/1 A-3

ASL CAPACITANCE TRANSDUCER CALIBRATION ASL Transducer Calibration Source Model No. : 10833 Federal Products Corp. Micrometer Serial No. : 1068 Model No. : 6888 Creep Machine :5 C-E Asset No. : 8220 Transducer : B Resolut ion : 0.00001 inch Micrometer Micrometer Transducer Transducer Error of Reading Change Reading Change Transducer inch inch Inch inch inch Error

0. 00000 .905220

.1 0000 .10000 .805213 . 100007 -. 000007 -. 007

.20000 .20000 .705225 .1 99995 .000005 . 002

. 3000n , 30000 .605236 .299984 , 000016 .005

. 40QO0 .4000 0 .505263 .399957 .000043 .01 1

. 50000 .50000 . 405296 . 499924 .000076 .015

. 60000 .60000 .305325 .599895 .0001 05 .017

.70000 .70000 .205364 .699856 .000144 .021

. 800 . 00000 .1 05389 .799831 , 0001 69 .021

. 85000 . 85000 . 5541 0 . 84981 0 . 00 0 90 . 0Z2 Date : 17 NOV.,

1989 Calibrated by

.- E. Y--.

A-4

THERMOCOUPLE ANALOG INPUT CALIBRATION Creep Machine S Center Thermocouple Input Leeds & Northrup Millivolt Potentiometer Model Number 8686 Serial Number 18181921 Calibration Date 9-29-89 Kaye Instruments Ice Point Reference Model Number K140-8 Serial Number 1Z9S Input Input Min. Piax. Avg. Avg. Temp. x Channel Temp.-F mv mv mv Temp.-F Diff.-F Error 224 600 12.850 1Z.857 12.853 599.81 . 188 .031 224 700 15. 140 15.173 t S. 167 699.50 .496

071 224 800 1 7.510 t7 .5t6 1T.S13 799.58 .417 .052 224 900 19.865 19.870 19.867 899.3S .653 .073 224 1 000 22.237 22.242 22.239 999.58 .416 .042 224 11 0 0 24.603 24.607 24.604 1099.56 .437 .040 224 1200 26.953 26. 956 26.954 1199.25 .753 .063 224. 1300 2 9.292 29 .294 29.293 1299.14 .857 .066 224 1400 31 .594 31 .597 31.595 1398.42 1 .575 .113 224 1500 33.875 33.881 33.879 1498.25 1.752 .117 Date: 17 NOV., 1989 Calibrated by

______________ -a----

A-5

THERMOCOUPLE ANALOG INPUT CALIBRATION Creep Machine S Top Thermocouple Input Leeds & Northrup Millivolt Potentiometer Model Number 8686 Serial Number 18181921 Calibration Date 9-29-89 Kaye Instruments Ice Point Reference Model Number K140-8 Serial Number 129s Input Input Min. Max. Avg. Avg. Temp. X Channel Temp.-F mv mv mv Temp.-F Diff.-F Error 1___08_ _____

1 08 600 12.771 12.829 12.8Z2 598.46 1.544 .257 1 08 700 15.119 15.155 15. 149 698.72 1.284 .183 1 08 800 17.469 17.499 17.495 798.79 1.211 .151 1 08 900 19.848 19.858 19.856 898.89 1.1 14 .124 1 08 1 000 22.224 22.230 22.227 999.08 .919 .092 1 08 11 00 24.592 Z4.595 24.594 1099. 10 .899 .082 108 1200 26.950 26.953 26.951 11 99.11 .889 .074 1 08 1300 29.289 29.293 89.292 1899.08 .922 .07t 1 08 1400 31.574 31.578 31.576 1397.63 2.370 .169 I 08 1500 33.863 33.866 33.865 1497.63 2.368 .IS8 Date: 17 NOV., 1989 Calibrated by 711111--

A-6

7. Oak Ridge National Laboratory Letter 0409-49-90 regarding submission of data package for SA-533 Grade B, Class I plates, SA-508 Class 3 forgings and their weldments, dated April 20, 1990.

I I Letter 0409-49-90 OAK RIDGE NATIONAL LABORATORY POST OFFICE BOX 2008 OPERATED BY MARTIN MARIETTA ENERGY SYSTEMS. INC. April 20, 1990 Mr. A. W. Dalcher Chairman, SC-ETC General Electric Company 6835 Via Del Oro P. 0. Box 530954 San Jose, CA 95153-5354 Mr. Michael Gold Chairman, SG-SFA (SC II)

Babcock & Wilcox 20 S. Van Buren Ave.

Barberton, OH 44203 Mr. R. I. Jetter Chairman, SG-ETD Rockwell International Energy Technology Engineering Center P. 0. Box 1449 Canoga Park, CA 91304 Gentlemen:

Submission of Data Package for SA-533 Grade B, Class 1 Plates, SA-508 Class 3 Forgings and Their Weldments Attached find the data package and analysis for the above named materials.

We are requesting that discussion of this data package be placed on your agenda, for the May Code Meetings in Nashville.

Sincerely, C. R. Brinkman Group Leader Mechanical Properties CRB:las.

Attachment cc: J. M. Corum

'Hb; MfaiN H. E. McCoy H. Prager S. Roberts

1 1. III !

CE REQUEST FOR CODE CASE FOR SA533B CL1 AND SA508 CL 3 INQUIRY NUMBER N87-37 REVISION 1 APRIL 17, 1990

N I -

SUPPORTING DATA PACKAGE(ORNL)

TABLE OF CONTENTS 1.0 MATERIALS 1.1.a.BASE METALS - SA533 PLATES & SA 508 FORGING bIDENTIFICATION OF HATERIALS, HEAT NO.'S cHEAT TREATHENIS OF MATERIALS 4CHEMICAL ANALYSES COMPARED TO SPECIFICATIONS eVIECHANICAL PROPERTIES COMPARED TO SPECIFICATIONS

1.2 o.WELOMENTS bIOENTIFICATION OF WELOMENT & WELDING PROCESS c.WELDMEHT HEAT TREATMENT (PWHT) d1WELD METAV CHEMISTRY eQWELD METAL MECHANICAL PROPERTIES

. 1.

2.0 Smn DATA - TENSILE PROPERTIES 2.1 TENSILE PROPERTIES FOR EACH MATERIAL RT, 700, 800, 900, 1000, 1100 COMPARISON TO CODE VALUES IN SECTION lI-APPENDlCES TABLES t-2.1 YIELD STRENGTH VALUES (RT TO 1OOF) 1-3.1 TENSILE STRENGTH VALUES (RT TO QOOF)

RATIO TREND CURVE ANALYSIS FOR Sm VALUES see- COMPARISON TO CODE VALUES IN TABLES 1-2.1 & I-3.1 2.3 PLOTS OF TENSILE & YIELO STRENGTH TREND CURVES RATIO VALUES VERSUS TEMPERATURE (RT TO 1100F)

COMPARISON TO CODE VALUES IN TABLES 1-2.1 & 1-3.1 2.4 TABLE OF Sm ALLOWABLE STRESS VALUES (N-47 1-14.5)

TIME INDEPENDENT VALUES AS A FUNCTION OF TEMPERATURE INCLUDE YIELD AND TENSILE STRENGTHS IN TABLE 2.5 PHYSICAL PROPERTIES TO 100F (N-47 1-14.7 TO 1-14.9)

MODULUS OF ELASTICITY.& COEFFICIENT OF THERM.L EXPANSION

f .,

PAGE 2 OF 3 3.0 St DATA - CREEP PROPERTIES 3.1 TABLE OF CREEP DATA FOR EACH MATERIAL 3.2 LARSON-MILLER CORRELATIONS FOR 1%, TERTIARY & RUPTURE ALL MATERIALS SHOWING AVERAGE AND MINIMUM CURVES 3.3 STRESS VERSUS TEMPERATURE PLOTS FOR 1%, TERTIARY & RUPTURE 3.4 STRESS VERSUS TEMPERATURE PLOT MHINIHUM STRESS TO 1%, 80% TERTIARY, & 67% RUPTURE STRESS 3.5 STRESS VERSUS TIME PLOTS (N-47 1-14.4)

St ALLOWABLE STRESS VALUES ISOTHERMAL CURVES FOR 700, 750, 800, 850, 900, 950 AND IOOOF 3.6 TABULATED St ALLOWABLE STRESS VALUES (N-47 1-14.4)

COLUMNS FOR 1, 10, 30, 100, 300 AND 1000 HOURS-ROWS FOR 700, 750, 800, 850, 900, 950 AND 1000 DEGREES F 3.7 PLOT OF.MINIMUM STRESS TO RUPTURE VALUES (N-47 1-14.6)

STRESS VERSUS MINIMUM TIME TO RUPTURE ISOTHERMAL CURVES FOR 700, 750, 800, 850, 900, 950 AND QOOPF 3.8 TABULATED MINIMUM STRESS TO RUPTURE VALUES (N-47 1-14.6)

COLUMNS FOR 1, 10, 30, 100, 30D AND 1000 HOURS ROWS FOR 700, 750, 800, 850, 900, 9S0 AND 1000 DEGREES F 3.9 ISOCHRONOUS STRESS-STRAIN CURVES (N-47 T-1800)

ISOCHRONOUS CURVES FOR 1, 10, 30, 100, 300 AND 1000 HOURS CURVES PLOTTED FOR 700, 750, 800, 850, 900, 950 AND 1000 F 4.0 Smt ALLOWABLE STRESS VALUES 4.1 TABLE OF ALLOWABLE STRESS, Smt, VALUES (N-47 1-14.3)

COLUMNS FOR 1; 10, 30, 100, 300 AND 1000 HOURS ROWS FOR 700, 750, 600, 850, 900, 950 AND 1000 DEGREES F 4.2 PLOT OF ALLOWABLE STRESS, Smt, VALUES (N-47 1-14.3)

ALLOWA LE STRESSES VERSUS TEMPERATUR. Sro AND St CURVES ISOCHRONAL CURVES FOR 1, 10, 30, 100, 300, AND 1000 HOURS

PAGE 3 OF 3 5.0 FATIGUE AND CYCLIC STRESS-STRAIN DATA 5.1 FATIGUE CURVE AT 1000 F (N-47 T-1420-1)

STRAIN RANGE VERSUS CYCLES 6.0 TEMPER EMBRITTLEMENT 6.1

SUMMARY

OF LITERATURE FINDINGS 6.2 RESULTS OF TENSILE TESTS ON CREEP TESTED SPECIMENS COMPARISON OF INITIAL VERSUS POST-CREEP PROPERTIES 6.3 TENSILE TEST RESULIS ON THERMALLY AGED SPECIMENS COMPARISON OF INITIAL VERSUS AGED PROPERTIES (90OF AND 100OF FOR 1000 HOURS) 6.4 CHARPY IMPACT TEST RESULTS ON THERMALLY AGED SPECIMENS COMPARISON OF INITIAL VERSUS AGED PROPERTIES (AGED AT 900F AND 100OF FOR 1000 HOURS) 7.0 SUBJECTS. NOT COvERED ELSEWHERE 7.1 Comparison of SA533 base metal, and weldment properties.

7.2 Comparison ofrProperties of SA533 and SA508.

7.3 Guide used.in preparing this document.

7.4 Direction of questions.

I

1. 1.a,b IDENTIFICATION OF HEAT NUMBERS AND HEAT TREATMENTS FOR SA533B MATERIAL HEAT NUMBER VENDOR FORM HEATT] REATMENT(a)

SA 533 PLATE D 9583 LUKENS 3.5 INCH PLATE A SA533 PLATE C 5975 LUKENS 9.625 INCH PLATE A SA533 PLATE 64535-1 MARREL FRERES 9.625 INCH PLATE A (a) A=2.5 h at 871C, WQ, 2.5 h at 663C, air cool, 20 h at 607C, air cool.

I 1.l.c CHEMICAL ANALYSIS OF SA533B COMPARED WrI SPECIFICATION.

Heat number Code Source of analysis Element specified' D9583 D9583 64535 64535-1 C5795 C5795S

_ Lukens CE M. Freres CE Lukens CE C 0.25 max 0.19 0.18 0.22 0.21 0.20 0.22 Hn 1.07-1.62 1.31 1.27 1.45 1.43 1.45 1.44 P 0.015 max 0.012 0.010 0.009 0.006 0.010 0.008 S 0.018 max 0.003 0.002 0.006 0.004 0.015 0.021 Cu 0.12 max 0.08 0.04 0.04 0.07 0.06 V 0.06 . 0.004 <0.005 0.001 0.003 Si 0.13-0.45 0.22 0.21 0.19 0.21 0:23 0.25 Mo 0.41-0.64 0.57 0.52 *0.52 0.51 0.54 0.58 Ni 0.37-0.73 0.69 0.76 0.62 0.62 0.65 0.66 Cr 0.10 0.06 0.04 0.08 Cb, Ti, W <0.01 <0.01 *' <0.01 <0.01 Co 0.01 0.013 0.014 0.016 Al 0.026 0.010 0.019 0.022 B *0.001 <0.001 <0.001 <0.001 As, Sn 0.005 <0.010 <0.020 <0.007 Zr <0.001 <0.001 <0.001 <0.001 N.

0.006 . 0.010

ASME,;Section II, Standard Chemical Requirements plus Special Reactor Beltline Requir qements for Cu, P,. S,-.and V;

l.l.d COMPARISON OF SPECIFIED AND MEASURED TENSELE PROPERTIES AT ROOM TEMPERATURE ULTIMATE TENSILE MATERIAL CONDMTION(a) YIELD STRENGTH STRENGTH ELONG RED IN AREA MPa ksi MPa ksi % %

SA533B, CL 1 SPEC 345 min 50 min 552-689 80-100 18 min SA533B, CL 1 A 492 69.4 , 621 88.4 29.5 71 a) A= 2.5h/871Ct2.5h/663C/20h/607C

1.2.a IDENTIFICATION OF WELDMENTS AND WELDMENT PROCESSES FOR SA533.

Heat Vendor Plate thickness Evaluated asa number cm in. SA weld SMA weld D9583 Lukens 8.9 3.5 X X C5975 Lukens 24.9 9.625 X 8 - base metal, SA - submerged arc weldment, and SMA - shielded metal arc weldment.

1.2.b WELDMENT HEAT TREATMENTS(PWT THE POST WELD HEAT TREATMENT USED FOR SA533 WAS 20 HOURS AT 607C

l.1c

l. CHEMICAL COMPOSTONS OF DEPOSD WELD METAL IN SA 533. BASE METAL COMPOSTONS GIVEN FOR COMPARISON.

Ueat Element Code c D53 specifleda D9S83b D9583b D95B3c 6 4 5 3 5 -lb 6 4 53 5 -lb C5 79 5 b C5795b C5 795c D9583d (Lukens) (CE) (CE) (H. Freres) (Ca) (Lukens) (CE) (CE) (CE)

C 0.25 (max) 0.19 0.18 0.11 0.22 0.21 0:20 0.22 0.13 0.10 Hn 1.07-1.62 1.31 1.27 1.41 1.45 1.43 1.45 1.44 1.53 1.27 P 0.015 (max) 0.012 0.010 0.008 0.009 0.006 0.01 0.008 0.009 0.008 S 0.018 (max) 0.003 0.002 0.006 0.006 0.004 0.015 0.021 0.015 0.014 Cu 0.12 (max) 0.08 0.04 0.04 0.04 0.07 0.06 0.04 V 0.06 0.004 0.003 <0.005 0.001 0.003 0.004 0.008 SI 0.13-0.45 0.22 0.21 0.45 0.19 0.21 0.23 0.25 0.45 0.35 Ho 0.41-0.64 0.57 0.52 0.55 0.52 0.51 0.54 0.58 0.57 0.b8 Hi 0.37-0.73 0.69 0.76 0.13 0.62 0.62 0.65 0.66 0.12 0.03 Cr 0.10 0.06 0.04 0.08 0.14 0.03 tCb, tl. U <0.01 <0.01 <0.01 <0.01 <0.01 <0.029 Co 0.01 0.013 0.014 0.016 0.011 0.005 Al 0.026 0.01 0.019 0.022 0.011 0.005 B 0.001 <0.001 <0.001 <0.001 0.001 <0.001 As. 'n 0.005 <0.01 <0.020 <0.007 <0.007 <0.005 Zr . <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 N 0.006 0.01 0.006 -

aASME,Section II, Standard Chemical RequIrements plus Special Reactor Beltline Requirements ror Cu. P. S. and V.

b8 83 6 metal.

CUeld metal deposited by submerged arc process. AWS CLASSIFICATION NO. EH14.TYPEF-80FLUX(INDE124). EH14SPECIFIEDTOHAVELOWCu. LOWP.

nd~ld metal deposited by the shielded-netal are process. AWSCLASSIFICATIONNO.E8IOGsRW.

13 MATEIAL NUMERICAL IDENTIFCATION CODES USED IN DATA PROCESSING.

Vendor Modified Heat Lot heat heat Forma treatmentb number number number D9583 9583A BM CE-STD 3 D9583 9583A TW-SA CE-STD 4 D9583 9583A WH-SA CE-STD 5 C5795 5795 BM CE-STD 6 C5795 5795 WM-SA CE-STD 7 C5795 5795 TU-SA CE-STD 8 D9583 9583B BH CE-STD 9 D9583 9583B SW-SMA CE-STD 10 D9583 9583B IH-SHA CE-STD 11 64535-1 64535 BK CE-STD 12 8BK - base metal specimen; TW - transverse weld specimen, fusion line in specimen center; WH - weld metal specimen; SA - weld made by the submerged arc process; and SHA - weld made by shielded metal arc process.

- I

2.1 TENSIE PROPERTIES OF SA5338 CLIl.SA0.0016fmnn vn eln n c1GIs IcOoEuTSIb)i ALj L1F0lREOLFTKN B

-.- _ Do o r s I I l I l I _scn -

kal 9563A 29.66 9.20 74.58 29.30 - 8.47 - 75.61 29,20 10.12 67.04 30.10 - 10.86 67.83 29.40 10.20 - 69.99 29,35 10.52 70.37 26.0 10.23 67.04 23.38 1 7.60 1 64.00 5795 AA rn Ri i I *a .n nE

-2 - .4 LL 5795 I600.00 l 52.70 l 52.70 l 43.80 800.70 1 60.70 l 60.00 l 30.6 so 9583A 700.00 L66..0 59.87 l 43.tO 83.20 l 801s 80.00 26.18 - .1 7 7s5 9583A 700,00 59.40 _1 l 7 81.60 _ _ _ 29.00 9.45 77.26 5795 700.00 56.60 _ _ 80.10 29.00 6.45 _71,99 5795 700.00 59.40 61.30 _ _ 28.35 8.25 71.85 64535 700.00 58.30 _ _ I eo60.10 - _ 1 1 305 CS 664 - 7 60 64535 700.00 58.60 _ 79.60 _ _ _ 30.00 8.56 74.44 64535 750.00 56.00 56.00 42.30 73.60 73.60 60.00 30.90 7.71 76.12 9S83A 800.00 50.20 84.77 41.60 73.70 71.43 80.00 28.60 6.82 61.22 9s63A 800.00 63.50 _ 740028_

o- 26.00 3.60 80.43 5795 800.00 55.00 -7 2.30 227.99 74.52 879S 80o00 49.40 9 s40 .3260 7.70 75 14 64635 600 00 55.80 _ 70=_0 7 = 32 61 7.43 78.35 64535 600.00 54.70 s_____29.10 J 668 78 58 84565 850.00 52 10 _ 2.10 40.60 63.60 63.60 76.60 t<,9S 4.75 60.06 9583A 900.00 62.50 50.77 39.40 60.20 6o.0s 72.?7 33.40 6.14 52.55 9583A 900.00 53.40 _6_ _0_ S0.20 _ J6.15 3.08 85.39 5795 900.00 48.80 _ -60.80 - 34.00 4.42 77.61 5795 900.00 45.70 _.60 5_ 39.60 5.95 60.36 64836 soo.oo 53.80 - 60.30 41.6s JM -2.76$.

64s5s 900.00 50.40 _-eo.20 35.25 4.28 60.s9 6453s 950.00 46.70 46.70 37.80 5t.30 _5.30 67.30 50.75 2.12 86.75 9583A 2000.00 46.00 44.44 35.90 48.00 48.35 J 62.20 42.90 2.60 84.06 95a3 I 1000.00 46.50 9 40 _ 42.40 1.5 84.80 S795 2000.00 43.30 480 52.77_ 2.68 72.46

... . 1000.o0 41.00 47.10 4.60 3.44 77.62 64535 1000.00 44.60 - 49.30 - _ 42.00 1.70 66.06 64535 1000.00 45.00 -As 4 0 46.50 1.92 76.39 l

1i5so.0o0 1 3s5.: 35.2i 39. - .4-

.60 2.1 e gs 63 A I I 0A.

9583IA 1100.2 56.37 64535 35.30 a) Ta¢cI-2.1Appecai SccsuioUIL.DIvidonI SeenoteonnextIpae.

Tensile tests were performed at crosshead speeds of 0.01 and 0.002 inches/min. The specimen gage length was 1.25 inches, so the strain rate was either 0.008 or 0.0016 in/in/min. The values used in the data analysis were from the tests at the lower strain rate.

However, the yield stress and the UTS were not significantly different at the two rates at temperatures from room temperature through 1000F, and.either set of data could have been used with the same results.

2.3 RATIO ANALYSIS FOR SA533B TENSILE PROPERTIES RATIO ANALYSIS FOR YIELD STRENGTH 1.1

- -I----- -

1.0 - - - - - - - . 0 I . .n U IVASURED Os 0.8 C,,

0.7 0.6 0.5 OA To 200 400 600 800 1000 1200 TEMPERATURE(F)

Average results from duplicate tests on three heats of material (see Table 2.1).

2.3 RATIO ANALYSIS FOR SA533B TENSILE PROPERTIES YIELD STRENGTH OF SA533B MODIFIED BY RATIO ANALYSIS 60 50 E-By 40 Us W

30 20 0 200 400 600 800 1000 1200 TEMPERATURE(F)

2.3 RATIO ANALYSIS FOR SA533B TENSILE PROPERTIES RATIO ANALYSIS OF UTS FOR SA533B 1.2 1.0 0-%

E- 0.8 E-F-

0.6 UW 0.4 02 _

200 400 600 800 1000 1200 TEMPERATURE(F)

Average results from duplicate tests on three heats of material (see Table 2.1).

2.3 RATIO ANALYSIS FOR SA533B TENSILE PROPERTIES UTS FOR SA533B BASED ON RATIO ANALYSIS 90 - - - 1----

so S4X 70 - - - - -

m60=== ==<=

40 - - - - - -

30 - - - -

0 200 400 600 800 1000 1200 TEMPERATURE(F)

2.4 ALLOWABLE STRESSES(Sm) VS TEMPERATURE FOR SA533B (o) (C) (d)

TEMPERATURE DERIVED Sm YIELD STRESS UTS CODE Sda) RECOMMENDEDS OF ksl ksl ksl ksi ksl 75 26.4 50.0 80.0 26.7 26.7 100 26.4 48.0 80.0 26.7 26.7 200 26.4 44.0 80.0 26.7 26.7 300 26.4 42.0 80.0 26.7 26.7 400 26.4 41.5 80.0 26.7 26.7 500 26.4 41.5 80.0 26.7 26.7 600 26.4 41.5 80.0 26.7 26.7 650 26.1 41.5 79.2 26.7 26.7 700 25.0 41.0 75.9 26.7 26.7 750 24.3 40.5 73.7 244._3 e 800 23.1 39.5 69.9 .2.8- .2y-J' 850 21.8 38.0 66.0 _________A__

900 20.0 37.0 60.5 s o -2 /

950 18.2 33.5 55.0 16_, 2 1000 16.0 31.0 48.4 Id./

(a) Table 1-1.1, Appendix I, Section ImI, ASME Boiler and Pressure Vessel Code, Class I Components.

(b) Derived Srequals 1/3 x (UTS column).

(c) Yield Stress is taken directly from the yield trend curve based on the ratio analysis and a room temperature value of 50 ksi.

(d) UTS is the lesser of 1) 80 ksi(minimum specified room temperature UTS or (2) 1.1 times the elevated temperature UTS from the trend curve.

2.5 MODULUS OF ELASTICITY vs TEMPERATURE FOR SA533B TEMPERATURE ELAST MOD SCURCE F 1000000 psi_

-325 31.1 SECT. III, APP I, TABLE 1-60

-20 0 30.5 00 29.9 _

70 29.2 200 28.5 300 28.0 400 27.4 .

500 27.0 600 26.4 .

700 25.3 800 23.9 _

900 22.2 SECT VIII, DIV 1, TAB UF-27 1000 20.1 l _

I 1100 17.8 _

I .. I

2.5 THERMAL EXPANSION OF SA533B TEMPERATURE SA533B SA 533B F INST.COEF. MEAN COEF.

70.00 7.02 -

100.00 7.13 7.06 150.00 7.29 7.06 200.00 7.45 7.25 250.00 7.60 7.34 300.00 7.74 7.43 350.00 7.88 7.50 400.00 8.01 7.58 450.00 8.13 7.63 500.00 8.25 7.70 550.00 8.36 7.77 600.00 8.46 7.83 650.00 8.55 7.90 700.00 8.63 7.94 750.00 8.71 8.00 800.00 8.78 8.05 850.00 8.84 8.10 900.00 8.90 8.14 950.00 8.95 8.19 1000.00 8.99 8.23 Units on both parameters are in/in x°F x 10O Source of data is: "R. A. Moen, "Thermophysical Properties of Ferrous Structural Alloys," HEDL-TME 78-47, UC-79b,h, April 1978.

3.1 CnEEP DATA FOR SAS= 5E (a) (b) (b) (b) . (b). (b)- (b)

.lTESrQ Ha I OTNQL f STFtEsS-KSI I MAP,;- It SR. YWH I T-T0.I% . .. T..I%

. ... i Th2% . 6ICREEPas% MD MAFS 2s0 I I5S I II aIs I To:t .... -T-R 2S708 I O9583fA 371 0.006 O-as A5 195 I 660 1 660 1004 2.62 25694 67 371 0.00012 0.1 2050 9500 7490D 1.44 1.74 2.7 25527 61 427 0.0062 O.1 48 175 630 670 1261 0.87 32.74 .. ,s_

25707 55 427 0.00032 4 1200 4300 51150 0.32 2.2 3.24 25520 -40 a 482 0.0019 0.2 340 840 1840 t8so 2612 0.09 16.76 35.1 25987 40 482 0.0022 4.4 340 790 1124D 0.21 2.8 2.21 25698 35 482 0.0011 12 S0o 1600 3220 1900 41J720 0.04 9.3 13.82 25518 _ 20 538 0.021 t 35 as 230 270 620 0.1 52.6 45.0 25709 15 538 0.003 10 250 600 1350 900 2995 0.14 18.8 23.9 25724 _ 12 538 0.0004 83 1820 3100 t OOt1950 40250 0.02 2.8 7.3 26502 15 566 0.013 1 66 135 295 19s 615 0.09 4B.6 52.7 26497 10 566 0.0038 2.9 220 460 1040 620 3378 0.04 56.2 64, 26s17 3 566 0.00016 145 5375 14870 0.02 0.4 0 26219 8 593 0.013 1.4 63.5 140 320 200 1330 0.06 79 88 257235 593 0.0019 25 . 478 920 870 1171D 0.02 2.6 7.1 25989 4 593 0.00046 36 1660 25921D 0.07 1.4 7.3 26208 _ 3 593 0.00058 5.5 1540 3050 6350 3500 9399D 0.18 8 27.4 26215 2 593 0.00015 175 4150 a 5295D 0.02 1.2 12.4 26216 1 593 0.000053 1040

3810D 0.03 0.9 11.1 25972 5795 70 371 0.00062 0.025 242 1700 25t50 3 2.5 SA.t 25968s 65 371 0.00016 0.1 956 _ .97 0.7 26194 65 371 0.00063 4.4 1960 78950 0.81 1.8 4.9 25976 a60 427 0.0031 0.05 45 240 679s 1.34 3.2 .01 25973 _a 5 427 0.000094 40 8480 0.31 0.23 25971 __ 45 427 0.000028 65 1201D 0.29 0.2 1.42 25975 35 482 0.00036 8.4 2000 38150 0.28 1.8 25969 27 482 0.000026 170 1319D 0.16 0.2 0 26190 20 538 0.0019 9.8 440 810 1395 720 1990 0.13 19.8 19.P1 25970 tS 538 0.001 23 900 17s0 1960 2026D 0.07 2.5 26501 15_ 566 0.0061 4.1 160 310 600 420 1163 0.09 42 7 47.5 26503 10 566 0.00014 12 625 1150 2350 950 38280 0.09 0 5.54 25974 8 593 0.0041 10 230 478 825 9600 0.06 4.7 7,0 26488 4 593 0.0007 54.2 1195 2250 4575 1900 . 8400D 0.03 15.8 26.5 26499 2 593 0.00019 454 5000 49730 0.06 _ 14.71 26491 9563B 75 371 0.15 0.05 3 9.6 2,4 I 10.8 23.5 6.4 16.6 71.1 26509 _ 70 371 0.00057 0.19 675 2375 2976D 1.7 2.3 3.3 26193 61 427 0.016 0.1 32 96 256 212 285D 0.85 S.8 7.5 .

26821 35 482 0.00072 17.2 1200 669D 0.11 0.8 0 26191 30 482 0.00026 12.9 2970 6100 1 1200 6500 11572D _0.17 5.5 3.8 25192 t 5 538 0.0018 3,6 450 960 2060 1160 2208D 0.08 5.5 3.i6 26505 10 566 0.0026 5.9 325 700 1650 1250 3648D 0.09 16 9.8. .

26486 8 593 0.005 4.5 165 340 765 450 2241 0.08 62.8 *67.1 26218 4 593 0.0006 36.9 1400 3100 5550 3350 90460 0.01_ 10.4 2S5 26196 64535 61 427 0.0051 o0.1 45 170 740 1480 1964 1.09. 25.1 71t8 -

26195 _ 30 482 0.0003 27.2 2675 5400 5700 68710 0.15 2a8 0 26820 20 538 0.0074 2.3 100 225 . 520 290 34D 02 1 *. *;. -

16.t.

26199 15 I_ 538 0.00015 8.9 600 1270 13450 0.1 2.3 .1.4 * ':

26510 t10 566 0.0031 0,08 260 580 1260 800 27820 0.13 15.9. 107 to :^

26201 8 593 0.0092 0.35 90 210 480 420 1409 0 61.3 . 93.2 26195 4 593 0.001 100 1020 1800 _ 1740 1945D 0.09 2.25 12.8 *

(a) minimum creep rate-(b) Times given in hours to various creep strain percentages, t-tertiary, and r=,rupture..

I 3.2 LARSON-MILLER CORRELATION FOR 1% STRAIN, AVERAGE AND MINIMUNI PROPERTIES 100 I.i - - - - --- I

- _ AVERAGE I I 13 + MINIMUM

._4 I _ I 77 cn 10 t:

C4 V: _ -

_ _ l I Ita\. -

_-- -- ___ I _--- tI.,S I I I I II 1

13 14 1s 16 17 18 19 20 21

_ P.1% .

P-l1o is a modified Larson-Millerpararneter defined as equal to K/1000(20+ log t) where: K=temperature in degrees Kelvin, t= time to 1% strain in hours.

K -.---6-f NOTES:

(a) Minimum properties defined as the average value rninum 1.65 multiples of the standard deviation.

(b) Larson-Miller constant variation between 15 and 25 had no effect on R2

3.2 LARSON-MILLER CORRELATION FOR TERTIARY CREEP, AVERAGE AND MINIMUM 100 I 6 I - I a I I i a a . I i rz uA 10 C:

En I'

13 14 15 16 17 18 19 20 21 P-t P-t is a modified Larson-Miller parameter defined as equal to K/1000(20+ log t) where: K=temperature in degrees Kelvin, t= time to tertiary creep (based on 0.2% strain offset) in hours.

3.2 LARSON-MILLER CORRELATION FOR RUPTURE-AVERAGE AND MINIMUM 100 l___I -

_ -* MINIMUM

=I = = I -n I I-I._

U) vn 10

___ .'_ ___ i _--

UE Ed I__

I_ i_

_ _7 ___ ,'_ ,

=

,\

(n

___ _3 I_ __ ___ I 1L I3 14 15 16 17 18 19 20 21 P.R P-R is a modified Larson-Miller parameter defined as equal to K/1000(20+ log t) where: K=temperature in degrees Kelvin, s t= time to rupture in hours.

e

3.3 MINIMUM STRESS FOR NOTED EVENT IN 1000 HOURS AT TEMPERATURE 10010 _ _ _ __ _ _ __ _

.__ _ 01%_

_____E -

TRT

.RUPT up 10 tso700 800 900 1000 1100 TEMPERATURE (F)

3.4 MINIMUM STRESS vs TEMPERATURE FOR 1% CREEP, 80% TERTIARY, AND 67% RUPTURE 100

__ _+- I Ii I

_ __ __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

a 1%

_--_ -_ * .8 TERT -

a ._.-i_

67 RUPT i

Q:z rnz 10 _ I_ _ _ _ _

_ . -1. ;'\

E-wD

= I = . .I.

I 600 700 800 900

<a 1000

-- 1100 TEMPERATURE (F)

3.5 St vs TIME-ISOTHERMAL CURVES 60 --

50-110 7 D 40 U,4 30 so-20 10 11 0 100 1000 10000 TIME (h)

3.5 Si, vs TIME-ISOTHERMAL CURVES JAw 70boo:'

-- I - - 14

-I 1--I4 iz _j=r Ia t__wm_ -I r 4-C-, 10 I ai Tl I -I_ _:t___1 I I II .,_

(n

=1 ___t_ ".t t _ A__

I 1 10 100 1000 10000 TIME (h)

3.6 St-Allowable Stress Intensity Values, 1000 psi Temp. F Ih 1Oh 30h 1 OOh 300h I OOOh 700 54 54 54 53 53 52 750 54 54 54 52 49 47 800 54 53 50 48 44 40 850 53 49 46 41 37 32 900 .49 43 39 34 M 724 950 45 36 31 26 _ 16 1000 39 28 24 18 14 9.5

3.7 MINIMUM STRESS TO RUPTURE AS A FUNCTION OF TIME AND TEMPERATURE 100 CI,

< 40 E-U, 1 . 10 100 1000 10000 100000 TIME (h)

. 1. p.;.. *1/4

. . . 1. . . . .

3.7 MINIMUM STRESS TO RUPTURE AS A FUNCTION OF TIME AND TEMPERATURE 100 :1 I.

I

I LLi6

.1. 1 .1 -rn I I 11 .1 I

III If MPIT

-i-F~q

_ eS_

4-.T-i I -1I.LH T"m I;Poe' O'P 1-4.iIT -*I II -4 I I . 4-I Ifi1II

?1 p) lJC Y7 T--V- III 1 _-+-6

~s~o*z 1111 U2 Is

CQ2 10 = i I I fil I i I I:It --- ?5O9 E- I I- I I t11111 -

S4 -. l4.4& . .... 1 1 I I Ii 4.

. 4 .- 4--. .4 54-44----

- 44 -. 5444.- 5- . z4--5

/ n-4 .1i I I~t I111 I III IIlill_ Ill- III t I III 1

I I1111 II IIIIII I I

1 :lltl lll 1111 t1,1f 1 10 100 1000 10000 100000 TlME (h)

3.8 MINIMUM STRESS-TO-RUPTURE VALUES, ksl

.5 . . .-

TENIERATURE Ih 10h 30h 100h 300h I000h 3000h 1 0000h 30000h 100000h F ksl ksl ksl ksl ksl ksl ks_ ksl ksl ksl 700 80 80 80 79 78 77 74 70 66 60

750 80 80 78 77 72 70 67 59 54 48 800 79 78 75 70 66 60 54 48 43 36 8E0 78 72 69 61 56 50 44 37 31 23 900 72 63 59 51 45 38 32 26 20 16 950 67 54 48 41 34 27 22 17 1 9 1000 ' S8 44 37 29 . 23 18 14 9.5 ()4.5

3.9 These isochronus stress-strain curves were prepared by W. K. Sartory of ORNL. The modulus of elasticity values in 2.5 and an ORNL-derived creep equation were used in formulating these curves.

.1. . .

. I * . .

I .

3.9 Isochronous stress-strain curve for SA533B at 700 F.

co Not Tensile lOOhr 3

x100 CL C)

CD 8) tn.

V4 CO (1D L

strcLn, X

3.9 Isochronous stress-strain curve for SA533B at 750 F.

IOOhr 3x10 2 Ix103 I

02 (O

0 D,

straLn, X

. .1 3.9 Isochronous stress-strain curve for SA533B at 800 F.

Hot Tensile hr

102

102 co 0O CL, 0CD, L

.4.

C, 0

0 1.4 1.6 1.8 2.0 2.2 strain, *!

3.9 Isochronous stress-strain curve for SA533B at 850 F.

9 0'

c:>

C.

0 0-Do CD 0

D 1i l<

strain, X

3.9 Isochronous stress-strain curve for SA533B at 900 F.

1 hr 3

10 30 lx102 C) 3x10 2 CL (0

lxlO a) strain, %

3.9 Isochronous stress-strain curve for SA533B at 950 F.

3.9 Jsochronous stress-strain curve for SA5338 at 1000 F. i 0

Cs rN c

a- Hot Tensile u) co 0Q 0.~

CD C;

4.1 ALLOWABLE STRESS INTENSITY VALUES, Smt(ksi)

TEMPERATURE Ih 1Oh 30hb OOh 300h I OO0h F ksi ksi ksi ksi ksV ksi 700.0 26.7 26.77 26.7 26.7 26.7 26.7 750.0 24.3 . 24.3 24.3 24.3 24.3 24.3 800.0 23.1 23.1 23.1 23.1 23.1 23.1 850.0 21.8 21.8 21.8 21.8 21.8 21.8 900.0 20.0 20.0 20.0 20.0 20.0 20.0 950.0 . 18.2 18.2 18.2 18.2 1 8.2 '16.0 1000.0 16.0 16.0 16.0 16.0 14.0 9.5 i

60 so 40 I

42 PLOT OF ALLOWABLE STRESS VALUES (So,is TEMPERATURE

= I_ ___

_I I I__

Vq30 W-t 20

__ I I_ _

.-I ; I- 3s--4 10 I1' I ' =I .. i n

6o0 700 800 900 1000 110 TEMPERATURE (F)

5.1 FATIGUE DATA AT 1000F FOR SA533B 10

___ I 1 6 i -i I a ll

= orginal S_

t- I

_=1- laV r;

rn IL

.1 0 100 1000 10000 cycles

(. I 4.

TEKPER EHBRITrILE.DE Hose of the temper embrittlement studies of this alloy show that embrittlement is associated with the enrichment of grain boundiries-with

impurities such as. phosphorus during welding. The worst embrittlement was noted in the heat-affected-zone. 4'7: The heats-of commercial material

-currently being evaluated are quiteilow in restdual impurities such.as phosphorus (Tablse2), so it is n6ot.e-r. likely that'commercial heats of this material produced to. the chemidal-specifications for nuclearuse in Section II of the code-will be'susceptible to.temper embrittlement.

However, a small program is being Tcarried out to demonstrate this point.

In the meantime,.it waSd felt that.tensile.tests at 25'C on creep samples

discontinued after a few perc~entstiain *would reveal any significant

' embrittlement. , , . .

RCF~E t'F-CE'

.; '-. . *,;s

4. Hasayoshi Haseguawa, Nobuya Nakajima, Nobuharw Kusunoki, and t.

Iazuhiro Suzuki, "Effects of Copper and Phosphorus on Temper Embrittlement..f' of Mn-Ho-Ni Low Alloy Steel (ASTH A533-B)," Trans. Jpn. Inst. Met. I6...& .,

641-46 (1975). '::"'. .

5. S. G. Druce and B. C. Edwards, 'On the Temper Embrittlement.of: :

Manganese-Kolybdenum-Nickel Steels," Yucl. Technol. 55, 487-98 '(November *" '*

1981). .

6. W. A. Logsdon, "The Influence of Long-Time Stress Relief Treatments on the Dynamic Fracture Toughness Properties of ASHE SA508 Cl 2a' and ASHE SA533 Gr B Cl 2 Pressure Vessel Steels," J. Hater. Energy SYs. V .

3(4), 39-50 (Harch 1982). *:

7. S. G. Druce, G. Cage,'-and 0. Jordan,'"Eff Agi Properties of Pressure Vessel Steels," Acts Hetall. 34(4), 641-52 (1986j;*::....

~ .. ~ 1 -. *b -. '

6.2 Effect of creep exposure on tensile properties at 75F.

creep exposure before tensile tensile without exposure

!ESlrp). *-AT LCCAIl(tI IEAT I&AT I 1PC 6Ih£55. TitE. H YSt UlSt'5I tLON6.S j YSSI MJTS.SI (LOMt. RA.S S97t _.71^ AM_ CE 4.17 *tS 1201 68.4 67.1 27 63.5 64 05.2 3J 69.I 25973 _ *-75_ M% *- CE 402 50 648 71.8 89.7 26.8 65.6 64 05- Jt 69.3 215909 5795 -E5E- C£ 462 27 1319 64.5 05.6 30.2 65 64 85.2 31 69.3 S9474

!-%55:6 9.

57*

J0A LEAXE ASE C

ORTM I 593 371 a 960 52.4 _31 I5 5619 64 os01~a - 61 75 5494 89.6 110 29 71.7 90 100.5 .20.3 7

-- 513 ;4563A BAM _ cS. I 371 67 1491 92.9 IO3 24.9 74 90 75.7

$714 583A BA5E Oft2 371 60 5016 .. 75.J 85.2 29.3 75.8 70.3 85.2 32.5 77.8

St94 9 3C71 67 719 8OZ.5 83,7 28.7 70.3 71.3 901 30.5 74.5

-25719 %J.63A BASE ORSL 2 427 50 1321 64.2 84.2 32.4 72.5 _7Q.3 8502 32.5 77.8 S2I517 9.3A MSE- ORM-IL I 121 61 1346 91.7 106 28.4 72.5 90 I Q.5 26.. 75.7

5707 9M3A BA5E CE 427 55 51S5 74.5 94.2 32.1 70.2 71.3 90.1 30.5 74.5

_5<61 '43A E . 0L 182 30 201^5 83.4 314_ 74.9 70.3 J. 32.5 77s 25721 *..S3A USC *IUL 530 12 329': 55.8 77.0 35 76.2 70.3 85.2 32.5 77.0

(.6 - is 3 tA

% ORIX 2 530 9 707 53.9 71.5 2 72.7 70.3 .. 5.. 32.5 77.6 25722 -503A eASE ORIL2 593 5 1300 57.6 76.1 35.7 78.1 70.3 8S.2 32.5 77.6

_$ 8 Tw CE 371 67 666 016 91.3 28.7 64.2 74.7 90.1 25.6 76.5

5t47 5'.3J3A WELD CE 371 67 55.13 83.2 92.9 23.3 66.4 74.9 09.6 26.6 66.4 i5529 9;03A WELD E 427 61 6121 1.2 93.2 20. 65 74.9 89.6 2S.b _.4 255.5 9 83A WELD CE 462 410 7292 72.3 91 .6 27.9 60.2 74.9 89.6 25.6 66.4 62 %J03A WEL P C E 402 50 21 82.6 95.8 J 26 63 74.9 166 09.6 66.4 23572 9f.04A WELP CE 530 20 3185 50.1 78.4 24.7 62.1 74.9 09.6 20.6 66.4

5524 I WELD (E 5383 20 2349 667 82.7 .8 S 63 8 74.9 89.6 28.6 66.4 The results of tensile tests at 25C on samples from discontinued creep tests are shown above. The creep tests involved exposures up to 7292h to temperatures ranging from 371 to 593C. The tensile results for samples that had only been heat treated before tensile Ia testing are given for comparison. The ductilities ( elongation and reduction of area) are higher for the samples creep tested before tensile testing than those tensile tested without. the creep test history. Thus, there is no indication of temper embrittlement in these materials based on these data.

6.3 Tensile specimens were prepared of SA533B and these were aged for 1000h at 900 and 1000F. The specimens were tested and the results are given in the following table. Specimens were tested at room temperature and at the aging temperature. Test data for unaged specimens are also give for comparison. Aging did not have a detectable effect on the ductility parameters (elongation and reduction of area). Thus these results indicate that temper embrittlement is not a problem under these conditions.

TENSILE PROPERTIES OF A5338 TRANSVERSE WELD (HT#5795) AGED MATERIAL TEST STRAIN YIELD TENSILE TOTAL UNIFORM REDUCTION HEAT TEMPERATURE RATE MODULUS STRENGTH STRENGTh GflNGAT)ON RlNGATKON OF AREA TREATMENT 0 *F 10*4/sec 10-3/min GPa 10+6psi MPa ksl MPa ksi % e %

ASRECEIVED 24 75 1.3 8.0 210 30.4 451 65.5 602 87.3 22.40 6.48 66.81 ASRECEIED 24 75 1.3 8.0 220 31.9 470 68.2 632 91.7 20.51 6.50 62.03 AGED/i K14820C 24 75 1.3 8.0 202 29.2 480 69.6 641 93.0 20.50 6.48 64.41 AGED/IK/482*C 24 75 1.3 8.0 189 27.4 463 67.2 631 91.5 23.59 5.84 63.07 AGED/IK/5380C 24 75 1.3 8.0 206 29.8 428 62.1 567 82.2 23.95 6.44 66.95 AGED)1K/538*C 24 75 1.3 8.0 219 31.7 438 63.5 595 86.2 22.40 5.99 63.33 ASRECEIVED 482 900 1.3 8.0 176 25.5 360 52.1 430 62.4 25.35 3.03 77.92 ASRECEIVED 482 900 1.3 8.0 171 24.7 362 52.6 435 63.1 23.60 2.71 78.56 AGED/1K/4829C 482 900 1.3 8.0 .158 23.0 380 55.2 455 66.0 21.70 3.04 77.31 AGED11K/482 0C 482 900 1.3 8.0 128 18.6 354 51.3 435 63.1 25.20 3.33 79.59 ASRECEIVED 538 1000 1.3 8.0 135 19.6 338 49.0

358 51.9 26.20 1.61 81.03 ASRECEIVED 538 1000 1.3 8.0 120 17.4 330 47.8 365 53.0 25.00 1.61 80.30 AGED/1K/5380C 538 1000 1.3 8.0 134 19.5 305 44.2 334 48.5 27.00 1.56 83.58 AGED/IK/5380C 538 1000 1.3 8.0 148 21.4 304 44.2 330 47.9 25.60 1.66 83.10

S.

6.4 Impact specimens were prepared of SA533B and these were aged for 1000h at 900 and 1000F. These specimens were tested and the results are presented below. The results do not indicate that temper embrittlement occurs under these conditions.

EFFECT OF AGING ON HAZ IMPACT PROPERTIES 250 200 150 A:

CD z t00 W

so o L-

-200 -too 0 100 200 300 TEMPERATURE (OC)

Table 1. Charpy impact data

Transition temperature Upper-shelf Condition energy TO T4O.7 T67.6 ()

As-welded -21 -42 -24 143 900OF for 1006 h -22 -46 -28 154 1000'F for 1000 h -37 -64 -46 163

7.1 COMPARATIVE PROPERTIES OF BASE METALAND WELDMENTS FOR SA533B.

Numerous tests were performed on base metal, weld metal, and transverse weld samples. The data presented in this submission are almost entirely for the base metal. The next two figures compare the yield and ultimate tensile strengths of the three types of material. Base metal has the lowest strength, so design values based on base metal properties should be conservative.

The third plot shows a comparison of the strengths of these same materials in creep. The differences in strength are smaller in creep than in short-term tensile tests, but the base metal is still weaker. Thus the use of base metal properties should be conservative.

7.,

The numbers on these plots refer to the lot designations defined in table 1.3.

ORNL-DWG 89-18611 ORNL-DWG 89-186:

-c 600 Soo 4z (Iz

£00 U) w Ca I) 30 f 200 600 0 100 200 300 400 Soo 60 TEMPERATURE (C)

TEMPERATURE (C)

.Fig. 12. Tensile strength for lots c Fig. 4. Yield strength for lots of base metal, transverse welds, and weld base mecal, transverse welds, and weld metal having the lowest strength.

metal having the lowest strength.

RNL-DWG 89-.18639.

1000 o 0.

,, 100 10'L 12 14 1- 1i . 20 22 S P - I

i 7.2 . Comparison of Properties of SA533 and SA508..

SA533B AND SA508, CL3 HAVE SIMILAR CHEMICAL REQUIREMENTS(a)

ELEMENT CONTENT (%)

SA533B SA508. CL3 carbon 0.25 max 0.25 max manganese 1.07-1.62 1.20-1.50 phosphorus 0.015 max 0.015 max sulfur 0.018 max 0.018 max silicon 0.13-0.45 0.15-0.40 molybdenum 0.41-0.64 0.45-0.60 nickel 0.37-0.73 0.40-1.00 copper 0.12 max 0.10 max vanadium 0.06 max 0.05 max (a) ASME, Section 11, standard chemical requirements plus special reactor beltline requirements for Cu, P, S, and V.

I

7.2 Comparison of Properties of SA533 and SA508.

SA533B AND SA508, CL3 HAVE SIMILAR MECHANICAL PROPERTY REQUIREMENTS MATERIAL YIELD STRENGTH UTS ELONGATION REDUCTION OF AREA ksi ksi %

533 50 min 80-1 00 18 min 508 50 min 80-105 18 min 38 min

7.2 ComDarison of ProDerties of SA533,and SA505. t COMPARISON OF SA533 AND SA508 TENSILE PROPERTIES AT A STRAIN RATE OF 0.001 6/min i SA533-AVG 3 HEATS SA508-ONE HEAT:

TEMPERATURE YIELD SIRENGTH UTs ELONGATION RED OFAREA YIELD STRENGTH WS ELONGATION RED OF AREA F ksi ksi % % ksi ksi %

75 69.4 88.4 29.0 69.0 66.9 91.4 31.3 64.7 400 54.0 79.5 28.2 66.5 69.3 94.7 20.6 53.0 700 59.9 81.0 28.3 74.0 *59.6 90.1 24.4 67.7 800 54.8 71.4 30.1 77.2 56.3 81.2 31.0 71.6 900 50.8 60.1 38.4 80.3 56.0 70.9 34.0 78.8 1000 44.4 48.4 45.0 80.8 49.4 55.3 53.3 76.8 1100 30.3 35.0 55.0 75.2 34.9 38.2 54.0 58.5 The ultimatestensile strength controls Svtx from room temperature through:900F.hSince.the:UTS of SA508 equals or exceeds that of SA533,'.Sit'fori.SA533s' and SA508 should be the same at temperatures from 75 through 900F.

7.2 ComDarison of ProDerties of SA533 and SA508.

COMPARISON OF CREEP PROPERTIES OF SA533 AND SA508 100

&n 10 rn CJ2 cc~

E-1'-

13 14 15 16 17 18 P.1%

P-1% is a modified Larson-Miller parameter defined as equal to K/1000(20+ log t) t where: -K=temperature in degrees Kelvin, .

t- time to 1% strain in hours.

The creep strength of SA508 is slightly lower than that of SA533 at the left side of the above graph. However, at the right side(longer times and higher temperatures) the properties of the two materials converge. Spiis controlled by creep only at 950 and 1000F, and IOCh at these '.'Vw semier;Wuras -r-, imrara-d on the bOVe p!ot.

this range the properties are about. the same for the two materials, so the values of Sxtbased on'creep would not be significantly

.. different.;

I I 7.3 Guide used in preparing this document.

Unless stated otherwise, the correlation methods described in the following document were used in preparing this data package.

el' 6vv o-s /SY2 (A/t-zf7J_~ -'?7G _

-. ;APPENDIX A PROCEDURES EMPLOYED TO ESTABLISH THE BASIC TIME AND ..

TEMPERATURE STRESS INTENSITIES AND ISOCHRONOUS STRESS.STRAIN -

CURVES.

A.1 .MAXIMUM ALLOWABLE VALUE OF GENERAL PRIMARY-

' '....-. . MEMBRANE STRESS INTENSITY FOR DESIGN:CONDIlTIONS. ' : * .::-

.. ' The-smbol S. isused for this value. The SO values are identical to the values for S.

-*.- i * ;.** gien in Section VlO, Division 1 of the Code.

A.2 tIME-INDEPENDENT DESIGNSTRESS INTENSITY

*The symbol Sm isused for this value. These values are based on tensile and yield-

': .. *

strengths of the material. The criteria employed are defined as folloiws:

,..-: '. .- : .. ., (a) The allowable stress Intensity value, Sm, for ferritic steels and nonferrous '..;;

metals and alloys, except those covered in paragraph (b)below, is thc leist of.

.~~~~~~h 'val.........teowigoraues:...folwn fou -. - .*;,,

. (1) one-third of the specii7id minimurn lensile.strengh at room timpere Z.

the -. 'ature value '. .. -d.

(2) one-third 6f the shorte-iteim tensile.strength at temperature (as dc

- ,* . .

The fined in A.2.1 in this Appendix). *

The ratio of UTSTIUTSR 7. for austenitic stainless steel, Isfound In ASTM DSS-S2.

* .(3j two-third of the specified minimum yield strength at room temper-

-- (4).-thirds6f theishort-term 0.2% offset yieldstngtattemp rat i* suienthfa tempratue: :.

* > ; *--. § *(efnedasth ratio of yield strength at tcrnperature .divid'd by the

>.< <4;\,>-,,'c~l-\S... J y~ce strcnt. at room tempeiftuicrniultipldbtenim . -;-: ..-

N .- - - specified room temperature yield strenith)' * *.

sK. The rat to ofSTSR foraustenticstainessstel, isfound iASTM

.The h is-determined strengt ai rate ofr.O.OS *nn:.

at siatn

.stors ;st;*8 *s'l ;

1'.. ()- T'hceallowable stress intensityYaiucSm; for austcnitic steelsnickel chromiunk' ;i is the lowest orthe fotlowyng four;

.onndanilkl-iron-chromiumalloys

it-.rd of the specified mininum t tt r mtuertr

; .e.wo-thirdslof 1 the specified minimum yield strength at'roorn. :.

'~ o *-'--.'-.-

- ---* .temperature .*--,*;. ..

- . . *- :. : ; r(4) 90%of the yield strength it the operating temperature. -'.

.-.A.21 ...YIELDEANDTENSlLESTRENGTH

*'The proceduies for establishing yield and ultimate strengths are discussed beIow: . .

The yield and tensile strength data available for a particular material grade are normal-. :-.

. .ized by ratiotng the elevatcd temperature strength of individual lots to the room

tc peraturc stcnzth of c sam lots cand then all iets of such ratios representing a P.9 . ; . . 'Not appticable to boltingmitelgits.

79

ft 7.3

*particular grade are evaluated by the method of least squares to establish the curve of abest fit for the data. The rcsulting strengrh-ratlotrend curve is considered to rcprcscrit

. .

the typical or characteristic variation of yicid or tcnsile strength with icmperature. .

Using such a ratio trend curve, it becomes possible to compute strength trend qurves for any particular room temperature strength level of interest within the limits encorn-passed by the original data. . ..- . ._

~ ince the design stress intensiTly criteria include fractions of the Specifled yield and*:.**.

*-tensile strengths, it is necessary to factor the ratio trend curves against the specified-

; .'minimum yield and tensile strcngths to define what may be termed minirurn position yield and tensile strength curves. At temperatures above roomn temperature, the property ieldstrength at temperature is taken, for purposes of the criteria, to be this minimum position value. However, th'e property tensile strength at terpeioture is taken as the smaller of:

(a) specified minimum tensile strength at room temperature; ori (b) a value 10% greater than the minimum position value cited abovei

. .,. - -. ,,l.

A3 TIME-DEPENDENT DESIGN STRESS INTENSITY 2 *.*:

The symbol SI is used for the basic time and temperature dependent allowable stress *.

Intensity for load-controlled stresses. Sg "atues are the least of cach of the three * -

-.. -. q4uantities: * .... ' '

... ' . (a) two-thirds of the minimum stress to cause rupture in time t,

. (b). 80% of the minimum stress to cause the onset of 1ertiary creep in time t; and

. 1. . I.S Mthe minimum stress to produc cone percent T(c) total strain'in time t. .

^ -' '. A.3.1 "--'-MINiMUM STRESS-TO-CAUSE -RUPTURE IN TiME't **-. .

'The basic dita and a description of the'rumpture stength procduresemployed

I-,

. .':foreveral of the materials incl-uea in Cis&:.1592 (i.e.,Typess34 and 316 Stainless*'

Stel) may e found InMe8al Propert 6 cjI data.-ealitirs [Smith (17) and * - * :

Tw6principal procedur are in evaluating the depcndenc ofthestcss-t. .

.: *- ; ; 5.4 .. rupture upon time and temperature. A choike betwven the results, basedon engineer-. *

> >t udgmentis ther'madei *- * . ::*

. ihe first procedure, the i othermal relation between stress and time for rupture of'- ' .s * -

'-:kj; ;

Z- .individual lots is interpolated or extrapolated;ais equired, to ideaitify the strcss-to- * '.e ii .s.-

g.Causc-rupturein 100, !0bo, ip,00, etFc:,h. Plottingon log-togcoordinites tends to

.-Z linearize the variation, and the'teby. fadllitatc extrapolation, particularly.at lower '

mperateresa 'At highei temperaturcs; the variation tends to cunirinearity at Ionger-:

> %'... .. **. irnes, and extrapolation Involves gieater risk. .The results ofsuch interpolatiosnor . .

- - .capolations for individual lots, as they vary with temperature, are then etvaluated by.'.

-. *the methou of least squaris to dmfine an avoftge rupture sdingth-terimpcrature trtnd: ,.i..*

- ':'*- curve. A trend curve for minimum rupture itrcrigth Is derived from the mean trend *'

.:4.- :..- >.8 ." *.: turve by subtracting t1.65 multiples of the stanar d C tin of the sampfie; If ceruan b,.-'-a Implicit assumptions hold (e.g., that the data are normalty distributed,or that the *-.

.- average is without error), this minimum trend curve defines a lowei boundary for 95% .*-;:.

. of the data. Inspection of the Uata plots has alwayishown that thsis approximately ._

tric.

^ .-;* IBased on dala deternined from tests performed on anr Rfl

4a

'7.3~

The second evaluation procedure that has been employed is an indirect one, Involving

one or another timc termpcrature parameter. One of the simplest of these, widely used, is Larson and Millcr's

~ i.j. Pi T(C+log t) F(S) t . *. . * * .

, the time for rupture in ho'ufs; C is a *

. * . where T Is the temperature in degrees Rankin t is material constant, and F(S) denotes that the parameter depends upbn stress. Ordinar-ily, the available data are not adequate to permit evaluation of individual lots by the R _?..

. . *.:

parameter procedure as w~ould be desirable; instead, the total data population must be-.- *. -

-.. -' treated as a common population on a universalized basis, assuming the constant C to remain invariant from lot to lot within the data population. Accordingly, the scatter -

!. *t;_t*~ '- *~, band for the stress-parameter variation is evaluated by the metod of least squares to

determine an average curve of best fit, from which averageval ues of rupture strengih

-.: *St as dependent upon temperature can be derived. A minimum curvey s developed, as in .

l.65 multipes of the standard*

.-...x ' .. "-'thepreviously described procedure by subtracting

. ' -deviation of the sample from the average curve.

. ' . A.3.2. MINIMUM STRESS-TO-INITIATE-TERTIARY-CREEP IN TIME t **

. - F the ferrous alloys included in Case 1592, Leydaand Rowe (19) had repoitd that*

.-'h*. *- etime to initiate tertiary creep isa fixed fraction of the time for rupture for givn a .Yi. .

. '*

material at a specific temperature. Thus" for 24 Cr-1. Mo.indTipes 304 and 316, the *.:.

minimum stiess-to-initiate-tertiary-crep was computed from the minimurstress-to-...: -'

. ** '.. - .. 'rupture datL For the highnickel alloystertiary creep.lnfor T ration was ieveloped by

examination of available'creep curves; In all instances, the initiationof tertiary creep was assessed visually. Extrapolationstb'lower temperatures were made by time-

. temperature parmetermethodsexactly analogous to thosedescribed for rupture

.. strength, eicept that the time for tertia"y creep to. begtin became the measured quantity w: .y::.-..:.. % .:

"A3J MiNiIMUM STRESS TO CAUSE~I.@' . .T *IN.IM't-

-' ;s* *:- common the criterion cith for stressito initiate tertiarycreep,the criterion-on

- -.* . M \mninImum stress fo&i % totalstrain isnew to the ASME C6de.. For purposes of Sec.: ...

.. *. ... tiono landd*cinVli'iin-ce Sction YlllDivision 1;creep strength his for many years been evaluated

.;..-¶.

A. * - In terms thestirsi causingasecondarycreep rate of 0.01 % per I000 hr.- In the new. ;.

of

.*; : c terion, stress to cause 1% tptal.strain can be.evaluated in an exactly analogous

.-; . ;.-; .. mannerasthe stress to cause rupture or to Initiate tertiary creep-the dependence uporA.'. .'*

Ž~**:..-.; -5*'stress tends to parallel that of rupture time-or Itcan be derived directlj'froin the . .-

.; tr..l~-e$:S G' A i :..-

ba * :Isochronous stress-strin curves. Both approaches have been employed In developing:'** -

rnateri maal fr Case 1}592.- * '

. *- - . ** . *. -* .:A. .- *

. ;.~"-8~v,,;R e.,.i*.:

. .. * **

i

-, * .s.e . * .. - *.t ...

A4 -. ISOCHRONOUS STRESS-STRAIN CURVES

'. ',; For relatively

-- times, isbchronous stress-stratn curves may be derived bytkingcnstant

.- ;. At;-g;; * .~ i ...timesections hrough a family ofcreip.curves,.such as those.of Fig. 2. if owener; this approach..:

* .*.;....^ ..'.to their derivation is not generally practicable (unless fdrning only apartof a multi-approacli.

-.: procedure) since the time scale of inierestextends bejond the feasible limit f experknintl*

testing. Extrapolative procedures are required for the longer test times and for strains below 1 ..

. :..'* * - * . For generating the isochrbnous stress-strai curves n In ase 1592, two procedures hive been

ployed. In one of these, the curves are-developed by performing evaluations exactly anao-

. Z. that . *S ' dscribed for dc rn:!:F' t' o: r d' q 9t:::s.:c strain, except

': . .'-' .'a other speciric strains encompassing thle range of interest are also evaluated and the various

- parts assembled in plots of stress versus strain for different fixed times.

S. * * ** 81-.

*:*.-: A ~~fund if-Me Gc olion of isochto.-~~~

g-
..... ^
i- *offi s
~~s nousSiess-Strain Cilrves. ASME, 1972(4). -~~
~~@. * @ @ @ @ f~~
B @ w *
$:-- . j
'- - . a. .,;i . .-. ' * *, ........... ,, X . - : * ,
_ *
~~-I-~~
I@
~~. e S~~
' :.' . : .. . , a,: : ; . . h s $ _;; l t ...
r D b
,.-- * ....... . *. . * **;.!--;5t;'
~~; 2w. ; .* i.r-.~~
~~tS-.l*-*t.~~
* * , J
z * ., am > *
~~_ * ........................ > . . As ..... * ... * , *- . a, .......... , l,; . , g ............ * -~~
Ad Or ant ' . !
~~.- t- * .* t o . -t S . ;- _, ¢ . . . - , - .s~~
-JT ".. '......,, ,, -- 'S
'.A .. ':C
.H..:*. .::,. '. ;.* ,:,.
,::-...-;-2 .,.- ... t *,N ' . ;-4.....-;
7.4 Direction of questions.
QUESTIONS QUESTIONS ON THE DATA PACKAGE SHOULD BE DIREClED TO:
H. E. MCCOY, 615-574-5115.
OR CHRIS HOFFMAN, 203-285-4929 OR C.R BRINKMAN, 615-574-5106
~~8. ASME Publication, "Criteria for Design of Elevated Temperature Class I Components in Section III, Division 1, of the ASME Boiler and Pressure Vessel Code," May 1976.~~
Criteria for Design of Elevated Temperature Class 1 Components in Section 111, Division 1, of the ASME Boiler and Pressure Vessel Code May 1976 II
'I "The American Society of Mechanical IE3IES Engineers
- ..345 East 47 Street
~~NeNew York, I!J.Y..~~
. 10017*
.. I I-: ..
'.- 6 N.?I
. I Copyright @1976 The American Society of Mechanical Engineers
' All Rights Reserved No part of this document may be reproduced in any form, in an electronic retrieval system, or otherwise, without prior written permission from the publisher.
Library of Congress Catalog Card Number: 75-28980 Printed in the United States of America
9,
FOREWORD Experience in performing stress analysis of nuclear power plant components in accordance with the ASME Code Cases for elevated temperature service has indicated a need for a docu.
ment to provide the background criteria for the rules of these Code Cases. Accordingly, this document was prepared.
To assure technical accuracy of this document, numerous reviews of the document material were performed during the period of preparation. The design criteria described herein were developed over a period of several years. It isnot to be expected that this document will answer all the questions which will bc asked, but it ishoped that it will provide a starting point.
Con tents 1.0 Introduction .......... 7 2.0 Elevated Temperature Material Behavior. 8 3.0 Structural Failure Modes Considered .10 4.0 Stress and Strain Categories, and Controlled Quantities. 13 5.0 Design Rules for Load-Controlled Stresses in Structures other than Bolts. 15 5.1 Criteria for Design Conditions 5.2 Criteria for Operating Conditions-Introduction 5.3 General Primary Membrane Stress Intensity Limits for Operating Conditions 5.4 Allowable Stress Intensities for Loading Involving Bending 5.5 Use-Fraction Summations 6.0 Design Rules and Limits on Strain for Structural Integrity .. . 22 6.1 Reasons for and Objectives of Strain Umits 6.2 Limits on Cumulative Inelastic Strain for Structural Integrity 6.3 Phenomena of Plastic and Creep Ratcheting 6.4 Methods of Satisfying Strain Limits Using Elastic Analyses 7.0 Creep-Fatigue ........................... ,.,,. 35 7.1 Correlation and Evaluation Methods 7.2 Derivation of the Creep-Fatigue Interaction Rules 7.3 Creep and Fatigue Design Curves 7.4 Mean Stress and Multiaxial Effects 7.5 Rotating Principal Strains 7.6 Rules and Limits for Use with Inelastic Analysis 7.7 Rules and Limits for Use with Elastic Analysis 8.0 Design Rules and Limits for Buckling and Instability .58 8.1 Elevated Temperature Design Factors 8.2 Dependence on Post-Buckling Behavior 8.3 Time Design Factor 8.4 Effect of Initial Imperfections 8.5 Strain-Controlled Buckling 9.0 Special Limits and Consideration ......................... 62 9.1 Requirements for Welds 9.2 Rules and Limits for Bolts 9.3 Elastic Follow-up 9.4 Load Environment Histograms 10.0 References .73 Appendix A .79 List of Figures .83
I
~~1.0 INTRODUCTION~~
The purpose of this publication is to set forth the criteria, reasoning, and supporting data for existing ASME Boiler and Pressure Vessel Code Section 111, Division 1, rules for design of Class 1 pressure boundaries of components intended for elevated temperature service. The term elevated temperature refers to temperatures that exceed those for which allowable stress values are given in Section Ill.
The key consideration that sets these high temperature rules apart from the Section 111, Sub-section NB rules is creep effects. Unlike Subsection NB design rules which primarily guard against time-independent failure modes, the elevated temperature rules are applicable for service conditions where creep effects are significant.
These elevated temperature design rules for Section Ill have been under development for a number of years, but the last half-decade has been the period of most growth. The latest rules have been provided through Case 1331 (1)' and Its successor, Case 1592 (2). It is intended that this publication explain the basis of the rules in Case 1592 and aid in the application of the rules.
In nonnuclear applications, the design of pressure boundary components for elevated tempera-ture service dates back forty years. Over those years, vessels designed by the rules of Section I (Power Boilers) and Section VIII, Division 1 (Pressure Vessels) of the ASME Code have established a record of successful elevated temperature operation. This success is due to a combination of factors: (1) the ease of inservice inspection for such vessels has made it possible to detect incipient failure conditions before gross failures could occur, (2) the variety of operating and environ-mental conditions have been successfully addressed outside the Code rules by individual engineer-ing effort, and (3) the Code design rules have required extra wall thickness for service where creep phenomena are significant.
In Section I and Section VIII, Division 1, the design rules employ allowable stress criteria which utilize creep rate and stress-rupture properties in addition to short-term tensile strength properties. However, these Sections do not have mandatory requirements for a detailed stress analysis but set the wall thickness necessary to keep the basic hoop stress below the tabulated allowable stress, and they rely on the design rules for details and the design factor to hold secondary bending and high localized stresses at a safe level consistent with experience. Even though these Sections employ criteria based on the average stress to provide a creep rate of 0.01% per 1000hr and the average and minimum stress toproduce rupture in 105 hr, it isnot to be in-ferred that 10' hr (or any definite interval) is the intended design life for such construction. The fore-word to the ASME Code states that the objective of its rules "is to afford reasonably certain protec-tion of life and property and to provide a margin for deterioration in service so as to give a reasonably long safe period of usefulness." Neither Section I or Section Vill, Division 1,have mandatory require-ments for a cyclic fatigue analysis. The Power Piping Code (ANSI 83 1.1), which ex tends into elevated temperature application, gives allowable values for the thermal stresses which are produced by expansion of piping systems and varies the allowable stresses with the number of expected cycles. However, a complete evaluation of localized and combined peak stresses and and associated fatigue life assessment is not required.
As discussed in the ASME Publication, Criteriaof the ASME Boller and Pressure Vessel Code for Design by Analysis (3), the Code Committee recognized that additional design considerations t and rules were desired for nuclear components. These components would be exposed to severe service conditions (e.g., highly cyclic loadings, recurring severe th ermal shocks) where superior reliability was required to offset potentially serious consequences of failure. These needs led to the preparation of the design rules in Section 111, Division 1. The specialized service of nuclear 4 components also made it feasible to expand the Code rules to cover all service conditions, whereas the diversity of service for Section VIllI vessels made such a choice unwise.
' All references appearing in the bibliography are Indicated by numbers in parentheses.
7
Section III rules are intended to prevent three different types of failure which are as follows:
(1) Bursting, gross distortion, and elastic instability (buckling) from a single application of pressure are prevented by the limits placed on primary stresses.
(2) Progressive distortion is prevented by the limits placed on primary-plus-secondary stresses. These limits generally assure shakedown to elastic action after a few repeti-tions of the loading.
(3) Fatigue failure is prevented by the limits placed on the peak stresses.
Rules to account for creep, stress-rupture, or other time-independent failure modes were not included in Section IlIl.
After rules in Section III were initially developed for low-temperature design, there was an effort to extend the rules to elevated temperature design. By 1963, the first version of Case 1331 was approved. Subsequent versions of Case 1331 appeared, but it was not until 1971 and Case 1331-5 that extensive rules and limits were given to address the additional failure modes associated with elevated temperature operation. Additions and improvements led to the
-6, *7; and *8 versions of Case 1331. With the further preparation of rules for use in elevated temperature construction, the Code Case for elevated temperature design was changed to the (
present Case 1592.
This publication first provides a brief discussion of metal behavior when operated at elevated temperatures (i.e., in the creep range) under sustained and cyclic loadings. Then the relevant structural failure modes are described, and an explanation of the associated design rules and limits is provided. Special limits and considerations relative to the design procedure are also presented. Throughout this publication, comparisons of elevated temperature and low tempera-ture design rules are made.
2.0 ELEVATED TEMPERATURE MATERIAL BEHAVIOR When discussing elevated temperature material behavior, it is common to distinguish between elevated temperature and low temperature behavior by whether or not significant creep effects are present. As an introduction to the time-dependent effects associated with creep, results from experiments using uniaxial test specimens will be employed.
f Consider a uniaxial tensile specimen subjected to a load-induced stress level at a given test temperature. As shown in Fig.1 (a), if the temperature is low enough so that no significant creep occurs, the stress and strain both achieve their maximum value at a time to and remain constant at the maximum values for as long as the load is maintained. There are no time-depend-ent stress or strain changes after time to, and the stress and strain magnitudes are related to each other independent of time. However, if the test temperature is high enough for creep effects to be significant, the strain will increase with time after load application as depicted in Fig. 1(b),
and may ultimately lead to rupture of the specimen. If, on the other hand, the elevated temper.
ature uniaxial test is one where the specimen is initially strained a fixed amount and then held, the stress-strain history would be similar to that shown in Fig.1 (c). The reduction of stress under the constant imposed total strain of Fig.1 (c) is usually referred to as relaxationdue to creep effects. In both of the above elevated temperature tests, either stress or strain is time and temperature dependent. Also, the amount or creep strain or relaxation Isstress and strain level dependent Figure 2, sketches (a) and (b) depict the effects of initial stress level on typical elevated temperature creep and relaxation behavior. Elevated temperature material behavior is a function of time, temperature, and stress level, and each material has its own creep character-istics and stress-temperature-time domain where creep is significant Standard material specimen creep tests commonly used for characterizing basic creep properties are constant-temperature, constant-load tests of uniaxial tensile specimens. The terms commonly used to describe the stages of these creep tests are shown in Fig. 3.
8
t - TIME Fig. 1(a) HISTORIES FROM A LOADING AT LOW TEMPERATURE 4 STRESS 0 FRACTURE STRAI N I
TIME t
0 Fig. 1(b) HISTORIES FROM A LOAD-CONTROLLED LOADING AT ELEVATED TEMPERATURE Figure 4shows the typical manner of presenting stress-rupturc data. If one were to cross plot the value of total strain (elastic-plus-plastic-plus-creep) as a function of stress at agiven constant cnd-ofrtest time, a curve commonly referred to as an isochronousstress-strain curve would result. Typical isochronous strcss-strain curves arc given in Fig. S.Isochronous stress-strain curves are very convenient for presenting basic material creep data. The isochronous stress-strain curves for all pressure boundary materials allowed in Case 1592 are found in its Appendix T.
General descriptions of the generation of the isochronous stress-strain curves are given in Appendix A of this document. More complete descriptions of these procedures may be found in the ASME publication, The Generation of Isochronous Stress-Straln Curves (4).
So far, the material behavior in the creep regime has been discussed in terms of time-dependent deformation and fracture characteristics for sustained load conditions. When cyclic load condi-tions exist, fracture and incremental cumulative deformation (ratcheting) can be significantly 9
III rg --
[
to ~T MlE Fig 1(c) HISTORIESFROM ASTRAIN TRE AT ELEVATED TEMPERATURE accelerated by elevated typical reductions temperature effects.
in strain-controlled For example, Fig. 6
[from Weeks (5)1 shows variables associated fatigue life as a function with the elevated temperature of hold times. Moreover, variables include la) operation affect many the temperature; (b) rate, hold time, and the loading level (load the material cyclic life. These total exposure time); wave-form, frequency, (d) the environmentally (c) thermally activated strain induced factors such as irradiation, metallurgical transformations; and (e)the material form and its fabrication oxidation, process with the degree erosion, and corrosion; 3.0 STRUCTURAL FAILURE of cold work.
MODES CONSIDERED The structural failure modes considered 1 are the following:
(1) ductile rupture from short-term loadings (2) creep rupture from long-term loadings (3) creep-fatigue failure (4) gross distortion due to incremental
~~15) loss of function collapse and ratcheting IM8 buckling due to excessive deformation due to short-term~~
'7} cre buckling loadings due to lrig-zerm In determining what u"I ,t rules were appropriate Il-nde thee failure modes badigtohheScio failure that the rules shouldto preclude.
elevated temperature design, the first step by adding to the I ta-t mode The failure modes above was to
-' L.. ,sl(.ad()mds were developed Up to this point, discussion scctere onltheur loading. As in Section of failure modes has failur as relate lI, the elevated temperature centered on the failude 10th load applications and as related to the aplp applied loading sequences. Code Case provides guidance environmental effects However, the Owner for all possible stress-rupture and are a key area not and the Designer must specifically addressed recognize that tests performed creep-fatigue data used in developing by the Code Case rules.
in an air environment. the Code Case limits All the AtPeft were taken from Also, subtle changes hv elevated temp by synergistic effectsin mechanical properties brour,11.
related mechanisms resulting from the combined irtfuence are references (6) through not necessarily accounted for by (016).
ave". -,irimn of aging, recovery,yan the Code. The reader and e1ironmCn I
is referred to 10
I V)
LAJ I-j .07a
~~Cya5 a6 0 4' 3~~
.0I
- TIME (a) STRESS RELAXATION CURVES AT FIXED TEMPERATURE 2--
tn TIME (b) CREEP CURVES AT FIXED TEMPERATURE Fig.2 EFFECTOFSTRESSLEVELONELEVATEDTEMPERATUREBEHAVIOR 11
I rEMPE RATURE - Tc 0-4-RUPTURE STRES S - 00 I-c INTERC EPT - tREEP STRAIN
/ t l STRAIN
-PRIMARY I I- '4.-CREEP - - J -
STRAIN TIME 0
-+ ITo
-... ELASTIC-PLASTICI RUPTURE I
-STRAIN (TIME I R INDEPENDENT)
) I IIME 0
L PRIMARY.J. SECONDARY al- TERTIARY I l CREEP l CREEP I CREEP I STAGE STAGE STAGE STRAIN HISTORY AND NOMENCLATURE Fig. 3 DATA FROM THE STANDARD CREEP TEST w
a-.
(I)
-j TEP =CNS I LOG (TIME-TO-RUPTURE)
Fig. 4 STRESS-RUPTURE DATA FROM CONSTANT TEMPEhATURE CREEP TESTS 12
tI t2 tt4 v) LOAD DURAT1ION, HOUIRS t I< t 2< t3<t4 TEMP. = CONSTANT ACCUMULATED STRAIN Fig. 5 ISOCHRONOUS STRESS-STRAIN CURVES FROM CREEP TESTS AT CONSTANT TEMPERATURE Another design consideration area not controlled by Case 1592 is the area relating to functional performance. This area isleft to the Designer to assure adequate performance as required by the Owner.
4.0 STRESS AND STRAIN CATEGORIES, AND CONTROLLED QUANTITIES The elevated temperature rules follow the approach of Section 111, NB-3000 in categorizing different types of stress and strain and applying different limits to each category. The categories and symbols used are generally the same as in Section III with limits related to primary, second-ary, and peak stresses and strains.
As the structural behavior at elevated temperatures can be significantly different than at low temperatures, it was recognized that different grouping of the stress and strain categories for purpose of applying limits was needed. Thus, placing limits on quantities related to the type of expected structural behavior under loading was adopted. The two basic types of controlled quantities are:
(a) load-controlled quantities; (b) deformation-controlled quantities.
The load-controlled quantities are stress intensities which result from equilibrium with applied loads during plant operation. As in Section III, the stress intensity is defined as twice the maxi-mum shear stress and equal to the largest algebraic difference between any two of the three principal stresses. The load-controlled quantities are determined using linearly elastic material models. The familiar primary stress intensities are load-controlled quantities.
13
10 AISI TYPE 304 STAINLESS STEEL 1100 F e=4 x 10 3 SEC 1 5
TENSION HOLD TIMES:
D 1 MIN.
A 10 MIN.
2 V 60 MIN. , e=4 x 10 SEC (SOLID CURVE REPRESENTS FIT
< \ >vAs sTO ZERO HOLD TIME DATA)
~~1 v~~
o 0.5 0.2 0.1 , ,, I, ,,,1 01 02o2 , .,. I, ,,,1 ,
103 1 1 I1 11 104 105 CYCLES TO FAILURE Fig. 6 EFFECT OF TENSION HOLD TIME ON THE FATIGUE LIFE OF AISI TYPE 304 STAINLESS STEEL AT 1000 F IN AIR
(deflection Deformation-controlled quantities are stresses, strains, and deformations which result from load and strain compatibility. These quantities may vary with both time and the applied loads, and creep effects may be a major time-dependent influence. Thus, accurate analytical evaluation of deformation-controlled quantities usually requires inelastic stress analysis when creep effects are significant. The stress intensities usually categorized as secondary and peak in NB-3000 are usually considered to be deformation-controlled quantities. An exception, however, is made in regard to the expansion stress, Pe, defined in NB-3222.3. Such stress must be treated as either primary or secondary. Thermal expansion net axial and net shear force on the structural cross section are categorized as primary, Pm or PL, load-controlled quantities. Thermal expan-sion leading to net bending stresses in piping are treated as secondary, Q, stresses unless elastic follow-up effects preclude the reduction of the thermal expansion bending stress through small deformations. If elastic follow-up effects (which are discussed later In more detail) are significant, instead of performing a detailed inelastic stress analysis it is conservative to consider the thermal expansion bending stress as a primary bending, Pi, stress and then use the elastic stress analysis rules for assessing compliance with the design rules and limits.
The grouping of stress and strain categories according to the controlled quantities and the operating conditions is illustrated in the flow diagram of Fig. 7. Notice that the "Load-Con-trolled Stress Limits" column depicts a design evaluation procedure similar to NB-3000, but the various operating conditions are handled with different parameters. For example;whereas the Design Condition limits are also used for primary stress control for Operating Conditions in NB-3000, the elevated temperature procedure controls primary stress intensities by placing limits on both Design Conditions and Operating Conditions.
The Design Conditions still retain single-value stress limits, S,, based on extrapolated 105 hr properties-similar to Sections I and VIII, Division 1. Unlike Section 1I1,the Operating Condi- t tions have their own load-controlled stress limits in the Code Case. Time-of-loading becomes an additional variable so that elevated temperature stress limits, St, Sm, and Smr, are a family of curves such as those shown in Fig. 8. This separation of load-controlled stress limits allows short-time loads (e.g., earthquake and severe shock loads) to have stress limits based on short-term tensile properties. Short-time loads need not be considered in the analysis for the Design Conditions since these same loads will be considered under the limits for Operating Conditions.
As shown in Fig. 7, the "Strain and Deformation Limits" for elevated temperature design cover Normal, Upset, and Emergency Operating Conditions, but no limits are placed on Design Conditions or Faulted Operating Conditions. The inclusion of coverage for Emergency events with strain and deformation limits is a new feature of the elevated temperature design methods.
It was recognized that the low temperature Section III limits on primary-plus-secondary stress intensity range will not, due to creep effects, necessarily assure that an elevated temperature structure will shake down to elastic action in gross areas when fixed strain ranges are repetitively applied. Thus, the primary-plus-secondary stress intensity limits of Section III were replaced by strain limits, where inelastic analyses are performed, and by more conservative rules where elastic analyses are employed.
5.0 DESIGN RULES FOR LOAD-CONTROLLED STRESSES IN STRUCTURES OTHER THAN BOLTS The basic feature of a load-controlled stress is that it is necessary to satisfy equilibrium of the structure under externally applied loads. As a result, deformations will not generally relieve load-controlled stresses. In developing the rules for components at elevated temperature, the Committee was faced with the precedent established by Sections I and VIII and by the previous elevated temperature Case 1331-4 which used elevated temperature allowables based on I05 hr creep rupture properties and a creep rate of 0.01% per 1000 hr.
15
1.85j LIj -g
. O j r.
I CREEP-FATIGUE EVALUATION I <l_ Ix ERGENCY I +p _4 I - 'b F OR l n/d t/TId f41 I LI I I _t _l*O IODESIGN FACTORS I BUCKLING &
Se INT t INSTABtLtTY APPLIED F~t
.SNO LIMITS UNLESS SPECIFIED l K IN THE DESIGN SPECIFICATION OR CL R LEGEND 0UANTLT0 0 CONTROLLED
%>___/\___>/> s_ FOR ELASTIC ANALYSIS L r tr l lCONTROLLED QUANTITY AULTED ' P _ P P/ FOR INELASTIC ANALYSIS COMPUTED QUANTITY
-~I t (NON-MANDATORY)
Fig. 7 FLOW DIAGRAM FOR ELEVATED TEMPERATURE ANALYSES 16
22 l I l 18 F. SmA
- 14 -1 HOURS-CD1 CD
_`12
-0800 l 00 0 l20 0 0 1 O 8 30 6 30 4 3~
2 800 900 1 000 1100 1200 1300 1400 TEMP. F Fig. 8 S., ALLOWABLE STRESS, TYPE 316 STAINLESS STEEL The ASME Code wanted to explicitly recognize the actual service life for Section 1II, Class 1 components and still utilize the proven methods of Section I and VillI. The result is two sets of primary stress allowables, one for Design Conditions and the other for Operating Conditions.
5.1 CRITERIA FOR DESIGN CONDITIONS The Design Condition allowable stress values, SO, are the same asSection I and Section VIII, Division 1. For Design Conditions for Case 1592, the allowable primary membrane stress is 5O and the allowable primary membrane-plus-bending stress is 1.5 SO.
5.2 CRITERIA FOR OPERATING CONDITIONS - INTRODUCTION For Operating Conditions, the allowable stress criteria arc more complex. The first problem is to identify the numerical criteria to be used in setting the actual allowables. Then there is the problem of accounting for different loads for differing times and temperatures. It was decided to retain the basic Section III criteria for determining the time-independent allowable stress, SM.
17
To that isadded a time-dependent criteria, St. Another term, Smt, was identified as the tower of Sm or St for aparticular time or temperature. The St values are the least of three quantities:
(1) two-thirds of the minimum stress to cause creep rupture In time, t; (2) 809 of the minimum stress to cause the onset of tertiary creep in time, t; and (3) the minimum stress to produce 1%total strain in time, t.
The evaluation procedures employed in developing the St values for Case 1592 are described in Appendix A of this document.
5.3 GENERAL PRIMARY MEMBRANE STRESS INTENSITY LIMITS FOR OPERATING CONDITIONS The Normal, Upset, Emergency, and Faulted Conditions have general primary membrane stress intensity limits as shown in the following table.
Loading Condition Pm Limit is Lesser of Normal and Upset Sm or SI Emergency 1.2 Sm or S, Faulted Section III limits or 1.2 St The time, t, which corresponds to the total duration of the loading at temperature, T, during the entire design life of the component, isused for determining the St value. Note that there is no significant increase in the time-dependent (St related) limits for Emergency or Faulted loading conditions. Also, the time-independent (Sm related) limits are the same as those of Section 1II.
5.4 ALLOWABLE STRESS INTENSITIES FOR LOADING INVOLVING BENDING The limits for local membrane plus bending stress intensities in Case 1592 are tabulated below.
Loading Condition Pt + Pb Limit is Lesser of
'Normal and Upset 1.5 5 m or KtS, Emergency 1.8 Sm or KtS, Faulted Section III limits or 1.2 Kr St In Section III the 1.5 factor applied to Sm for Normal and Upset loading is the limit load shape factor for a solid rectangular section bar loaded in bending, assuming aperfectly plastic material.
Thus, collapse of the bar due to attainment of limit loads is precluded in design by limiting the elastically calculated PL + Pb. In elevated temperature design, where creep effects can exist, the K, factor is intended to recognize the additional load-carrying capacity in bending due to the effects of creep on the distribution of stress and strain throughout the section which results in a reduction in the actual maximum bending stress below the elastically calculated value.
In Case 1331, versions -S through -8,a factor K was used in the allowable stress intensities for primary membrane-plus-bending and in the strain limits for elastic analysis. The notation for this factor was changed in Case 1592 from K to Kt, 2 and it was made dependent on the actual cross section being considered.
The definition of the factor K, and of the more traditional section factor K isillustrated in Fig. 9, which depicts the stress-distribution in a beam under pure bending. The elastically calculated dis-tribution islinear and has a maximum value denoted by Pb. If Pb does not exceed the material yield stress, then the linear distribution isalso the actual distribution at zero time. As creep occurs, the stress-distribution becomes nonlinear, as shown in Fig. 9, and it continues to change with time. However, as the transient creep becomes depleted in the outer fiber portions of the cross section, the stress-distribution change diminishes and approaches a stable distribution 2
In the Winter 1972 Addenda to the 1971 Edition of Section Ill. Kdenotestthe more traditional shope factor or section factor based on an elastic-ideally plastic material response. Thus, there was a conflict In notation.
18
M M
-ACTUAL INELASTIC STRESS DISTRIBUTION ELASTICALLY CALCULATED STRESS DISTRIBUTION Fig. 9 ELASTICALLY CALCULATED PRIMARY BENDING STRESS DISTRIBUTION COMPARED TO ACTUAL TIME-DEPENDENT INELASTIC STRESS DISTRIBUTION
'.0
I K,
15 14 13 12 11 1.00 .
Z0.75 M STEADY-STATE CREEP STRESS
- DISTRIBUTION /
w0.50 BASED ON 2
o 2 U 0.25 6 I 0
0 2 4 6 8 10 12 14 16 18 20 STRESS (1OOpsi)
Fig. 10 STEADY-STATE CREEP STRESS DISTRIBUTION ACROSS A RECTANGULAR BEAM IN PURE BENDING AND HAVING A STEADY-STATE CREEP LAW OF THE FORM, f An throughout the section. Creep analysis of the pure bending of a rectangular section beam, neglect-ing primary creep and assuming the secondary creep rate is proportional to the n power of stress, reveals that the distribution of stress is independent of time for any given n value. The stress-distribution ranges from linear through the section when n is unity, to perfectly plastic when n is infinite (see Fig. 10). However, at any given time during the creep period, the actual maximum stress is less than Pb and is denoted by PblKt, and the factor Kt is a time-, temperature-, and material-dependent quantity, as well as being dependent on the cross-sectional shape of the beam.
To investigate the role of the Kt factor and the values which should be used in the Code Case, a number of elastic-creep and elastic-plastic-creip beam bending analyses using material properties for Types 304 and 316 Stainless Steel were performed. The resulting inelastic stress distributions and maximum stresses were examined relative to the K, factor. The objective of these compar-isons was to identify Kt values for Code Case use that would, for simplicity, be independent of time and would conservatively represent the long-term, steady-state creep stress. This work Is reported by J. M. Corum (54).
Results of typical beam bending analyses are shown in Figs. 10 and 11. These figures show the steady-state creep stress distributions in a solid rectangular beam and in a tubular beam or pipe 20
15 1.4 l3 1.2 1.1 I O 1.00 2 0.75 x
-j 0:
rrg 12 0.50 I'.
a 0.25 a1
-s 0 14 16 lB 20 0 2 4 6 8 10 12 STRESS (1000psi)
Fig. 11 STEADY-STATE CREEP STRESS DISTRIBUTIONS ACROSS A THIN, CIRCULAR TUBE IN PURE BENDING AND HAVING A STEADY-STATE CREEP LAW OF THE FORM, ic=An
[Note that for the particular thickness used, the value of Kl for n = a or for elastic-ideally-plastic response is approximately 1.31. The limiting value as the thickness goes to zero is 1.271 or circular cross section for various values of n in the assumed steady-state creep equation ic = Auv.For the rectangular section, K, varies from about 1.28 to 1.38 for n between 3 and 6, the usual range for the materials of interest at low stresses.
Additional calculations based on the steady-state creep portion of the two-exponential creep law recommended in L. D. Blackburn (20) were used to examine Types 304 and 316 Stainless Steel beams in the 900 to 1200 F temperature range. For elastically calculated stresses ranging from 5,000 to 15,000 psi, it was found that the ratios of the maximum steady-state stress to the elastic stress varied from 0.70 to 0.74, giving a K, value ranging from 1.35 to 1.43.
These results show that avalue of 1.25 for pure bending conditions is reasonable and conserva-tive for steady-state stress distributions in rectangular cross sections. On the same basis, however, Fig. 11 indicates that 1.25 is high for a thin circular tube. A comparable value to the 1.25 would be about 1.1 5 for the circular section. This illustrates the dependence that Kr must have on the cross-sectional shape. !
21
For pure bending, the formula for 14 chosen for Case 1592 is K, 1 + k5 where ks ca(K - 1)
The factor K is the section factor for the cross section being considered. Values of K for various sections are tabulated in Table A-9221(a)-1 of Appendix A of Section Ill. Presently, a is specified as 0.5. With this value, Kr for a rectangular beam (K<= 1.5) in pure bending is 1.25; for a thin tube (K = 1.27), 1.14; and for an l-beam with a very thin web (K = 1), 1.0.
5.5 USE-FRACTION SUMMATIONS The purpose of the use-fraction summations is to allow the Designer to take credit for the fact that a component may not be operating at a single temperature or stress level throughout its entire operating history. Thus, one is permitted to use higher allowables associated with shorter periods ot operation, provided that the creep damage associated with the entire operating history at elevated temperatures is taken into account.
To account for varying loads at variable times and temperatures, a modified linear damage use-fraction approach was employed. This damage rule takes the form Et;\4 13 where:
to= the total duration of time at a particular stress level and temperature during the service life of the component the= the allowable time of operation at the same stress level and temperature 2;tr= the total operating time at temperatures in the creep regime B = factor which is equal to unity or, alternatively, it can be specified to be less than unity in the Design Specifications to account for nonlinearities in the use-fraction rule The time used in the denominator of the use-fraction is the time to reach the time-dependent stress limit, St, for a primary stress P. at temperature T.
For primary membrane-plus-bending loads, the stress limits are increased by a factor K, which accounts for the stress redistribution in bending for various cross sections due to plastic flow.
Emergency Conditions, in addition to Normal and Upset Conditions, are included in the use-fraction summation.
6.0 DESIGN RULES AND LIMITS ON STRAIN FOR STRUCTURAL INTEGRITY The incremental growth of a component subjected to pressure loading with superimposed cyclic thermal stresses may lead to distortion or fracture unless the accumulated strain is kept within allowable limits. For low temperature service when the creep influence is negligible, the stress regimes for which freedom from ratcheting may be demonstrated are described by NB-3222.5 of Section IlIl. However, for operation where creep and relaxation cannot be neglected, the stress regime, E, as shown in the Bree (22, 23) diagram of Fig. 12, is the only regime where the response of the structure is ratchet-free. In many nuclear power plant components it is not feasible to keep all elastically calculated stresses less than the yield strength. In such compo-nents, plastic/creep cycling and ratcheting can be allowed, provided that both effects can be limited, calculated, and kept within safe allowable values.
22
Ut 7
6 at = THERMAL STRESS 5 op = PRESSURE STRESS ay = YIELD STRESS 4
R1' R2, PI Sit S2 & E ARE STRESS REGIMES 3
- a ) = C2 p y 2 ,..a =a2 P t y t y y
a y
E REGIME OF NO CREEP RATCHETING OR PLASTIC RATCHETING
[tjote however that creep due to -p can exist.)
Si S2I P & E: REGIMES OF BOUNDED CREEP/PLASTIC RATCHETING ANALYZED IN REFERENCE 28 RI & R2: REGIM4ES OF PROGRESSIVE RATCHETING Fig. 12 STRESS REGIMES
[Taken from Bree (22)] II 23
6.1 REASONS FOR AND OBJECTIVES OF STRAIN LIMITS Conceptually, the elevated temperature rules for primary stresses, creep-fatigue and buckling should provide an adequate basis for design. However, there are sufficient limitations to the current state of knowledge of elevated temperature materials behavior that additional rules for strain and deformation were deemed necessary to ensure structural integrity. These strain and deformation limits are not necessarily related quantitatively to specific failure modes.
However, they have been formulated to help ensure the applicability of the other rules of the Code Case,' provide additional comparative design data, and provide additional quantified assess-ment relative to the failure modes of:
(1) creep rupture from long-term loadings; (2) creep-fatigue failure; (3) gross distortion due to incremental collapse and ratcheting.
Materials test data related to the first of the above failure modes is presented in jakub and Moen (24). The data depicted in Fig. 13(a) and (b), demonstrate that prior creep deformation strains tend to reduce residual short-term tensile elongation and creep ductility (as measured by total elongation) as time-to-rupture increases.
For the creep-fatigue failure mode, much of the current data, even at elevated temperatures, has been obtained from tests which involve zero mean strain. Strain cycles between two fixed values as shown in Fig. 14(a) are typical in actual components.
For low temperatures the NB-3000 rules for primary-plus-secondary stress intensity limits rely on the simplified shakedown concept and 3Sm (or 2Sy) limits along with the simplified thermal ratchet rules to allow gross section yielding and ratcheting (increasing mean strain) to occur only during the first few load cycles.
At elevated temperatures, creep strains make it necessary to handle the more general strain history as shown in Fig. 14(b). Due to creep effects, the simple 3Sm limit does not assure shake-down to gross area elastic action [Townley and Poynor (25) and Penney and Marrietts (26)1.
Thus, strain concentrations higher than the stress concentrations may occur and the use of strain-controlled, zero-mean-strain fatigue data may not be sufficiently conservative.
It was recognized that a repetitive application of severe load cycles which induce large cumulative inelastic strains and ratcheting could cause exhaustion of the material ductility and lead to frac-ture in fewer cycles than would be indicated by normal creep-fatigue design rules. An approxi-mate method for evaluating the effect of cumulative permanent strain of fatigue is given by B. F.
Langer (27). It was felt that limiting the cumulative strains to a sufficiently low level in both gross and local areas would help ensure the applicability of the existing fatigue data. Thus, due to the lack of thorough experimental verification for creep-fatigue phenomena, the strain limits were also considered warranted for backing up the applicability of the creep-fatigue rules.
Finally, in the gross distortion failure mode, the limits on inelastic cumulative strain provide a direct control on incremental collapse and ratcheting.
In addition to the direct applications to failure modes, strain limits provide a vehicle to add quantitative rules for the design and location of welds. The present elevated temperature design rules require the use of material models that neglect the changes of mechanical properties at the welds-the entire structural model uscs base material properties. However, the rules also restrict the limits for the calculated strains in the weld regions. (See discussion on welds in the Special Limits and Considerations section.) The decision to reduce all of the allowable strain limits at welds by a significant factor was a positive step to keep weld locations outside of areas expecting large creep strains.
3 This function Issimilar to the role of the 3Sm limit of Section ill.
24
50 40
'4
.: ANNEALE D
~~30 o)~~
-J 0
. l1200 F
~207 010 O -
10 CREE KTAIh,.
1000 F 010 20 30 4)
CREEP STRAIIN, Fig. 13(a) EFFECT OF PRIOR CREEP DEFORMATION ON TENSILE ELONGATION OF TYPE 304 STAINLESS STEEL The strain limits are also related to the small deformation theory which has been the cornerstone for analyses of structures at low temperatures. The same small deformation or small strain as-sumptions are retained by the majority of current computer structural models being used for elevated temperature analysis. However, the possibility of unlimited creep or relaxation strains have requircd that some way be found to ensure that the theory still applies. If one overall rule cannot ensure compliance, then there must be some measure which tells when large deformation theory must be applied. At present the strain limits in Case 1592 are set low enough that they effectively ensure that small deformation theory is applicable for most structural analyses ol elevated temperature components. Obvious exceptions include some buckling instability analyses and thin-walled large deflection shell analyses.
For structural design of metal components, stress and strain are two obvious material state par-ameters which can be calculated and thus be considered as candidates for Code-allowable limits.
25
~~n av~~
' ' Iaa' I I I111111 -I 1 I lTIIII I I 11I 111 I 1
0 SUBMERGED-ARC WELDMENTS FLUX-APCOS S-4 L740 ENERGY INPUT-40,000-270,000 JOULES/INCH TEST TEMPERATURE-1000-1200 F R"'30 0 30 TYPE 304 BASE METAL 2°TEST TEMPERATURE-1000-1050 F o S 20
-JS 10 100 1000 10,000 RUPTURE LIFE, HOURS Fig. 13(b) VARIATION OF CREEP DUCTILITY WITH RUPTURE LIFE FOR SELECTED AUSTENITIC STAINLESS STEEL SUBMERGED-ARC WELDMENTS The two quantities are not independent of each other since a complete history of one should, theoretically, provide a good description of the history of the other parameter. The strain par-ameter, which is still the most common physical measurement in experiments on elevated temperature structures, is more closely related to structural deformations. Thus, the Committee concluded that strain was the more useful parameter for setting Code-allowable limits at elevated temperatures. Appendix T of Case 1592 reflects this decision..
6.2 LIMITS ON CUMULATIVE INELASTIC STRAIN FOR STRUCTURAL INTEGRITY In choosing what the controlled strain parameters and their limits should be, the state-of-the-art analysis methods were considered. For analysis economy it was and still is desirable to relate the limits to elastically calculated parameters. However, recognizing that the significant elevated temperature structural response is often inelastic, it was judged that inelastic stress analyses, rules, and limits were also necessary. Accordingly, for the less informative elastic analysis evalua.
tions, conservative but often restrictive design rules and limits were developed. The rules and limits for inelastic analysis, being less restrictive and considered to provide more informative and accurate control, are a means to demonstrate component design adequacy even if the elastic analysis screening rules cannot be satisfied.
The elevated temperature Code Case places the following limits on the maximum accumulated inelastic strain for parent material (see T-1310 of Case 1592):
(1) strains averaged through the thickness, 1%
(2) strains at the surface, due to an equivalent linear distribution of strain through the thickness, 2%
(3) maximum local strains, 5%
26
ttreA E mu 51KAI STRAIN HOL.D
- TIM/.EE
'to (a (b)
CONSTANT STRAIN RANGE Fig. 14 TYPICAL STRAIN-TIME RESPONSES The above limits apply to computed strains accumulated over the expected operating lifetime of the element under consideration and computed for some steady-state period at the end of this time during which significant transients are not occurring. These limits apply to the maxi-mum positive value of the three principal strains. A positive strain is defined as one for which the length of the element in the direction of the strain is increased. The principal strains are computed from the strain components (e,, e~,exzt, 4., 7xy, 7 y)- When the strain is computed at several locations through the thickness, the strains are first averaged and linearized on a component level and then combined to determine the principal strains for comparison to the limits on average and surface strains. The limits for discontinuity strains are based on the computed strains at the point of interest.
Inelastic strains accumulated in weld regions are computed using parent material properties and these calculated strains are limited to one-half the strain values permitted for the parent material.
(See Section 9.1 of this document.)
6.2.1 Membrane Strain Limits The membrane strain limit magnitude of 1%was selected to accomplish the goals outlined in Section 6.1 of this document.
6.2.2 Bending and Local Strain Limits High linearized bending strains may result in a large deflection which could compromise the integrity and reliability of the structure, particularly in the case of the thin-walled structures commonly found in elevated temperature systems. Classical small deformation theory, used for most high temperature design analyses, may also be compromised if linearized bending strains are not carefully limited. Thus, a 2% limit due to a linearized bending strain distribution was deemed appropriate.
The local creep fracture ductility in the creep regime was not considered sufficiently high so that local strain accumulations could be ignored. Monotonic creep tests have shown that the creep strain at fracture can be very low-well under 10%. It was desired to limit creep strains to a value under 5%. However, because of the practical difficulties that occur in trying to partition strains between plastic and creep strains, the limit of 5% was used for the total accumulated local strain.
6.2.3 Application of Strain Limits The strain limits are applied to the maximum accumulated inelastic strain. This is illustrated by Fig. 15. The steps in the strain cycle are as follows:
o-a initial elastic strain a-b subsequent creep strain due to sustained load 27
C f
b i_
d ACCUMULATED INELASTIC STRAIN a
0 0 TIME Fig. 15 STRAIN-TIME HISTORY SCHEMATIC b-e strain cycle resulting from superimposed thermal transient
[Note that there isa net decrease in the strainr e- f additional creep strain due to sustained load or residual stresses f- g recovery of elastic strain due to removal of primary membrane stress The strain limits apply to the accumulated inelastic strain, o - g.
The strain quantity limited by these criteria is the maximum principal tensile strain. The tensile strain was chosen since it was felt that this is the most appropriate index for the onset of tensile failure.
6.3 PHENOMENA OF PLASTIC AND CREEP RATCHETING Ratcheting is defined in Case 1592 as progressive cyclic inelastic deformation that can occur in a component that is subjected to cyclic variations of mechanical secondary stress, thermal secondary stress, or both, in the presence of asustained primary stress. The ratcheting phenom-enon Isof particular concern in elevated temperature reactor systems because of the cyclic stress conditions induced by frequent, and often rapid, temperature changes.
Several different mechanisms can contribute simultaneously, or singly, to the ratcheting phenomenon. The specific structural problem determines whether or not all of the various mechanisms come into play and the relative importance of each mechanism. However, an in-sight into the ratcheting phenomenon can be obtained by considering a simplified example in which the basic mechanisms are assumed to be present.
28
TIME 4HOL
_J W
-Jw I 41 1 TH 4M Wi= OUTSIDE z W 2 _
TL TIME UP (a) COMPONENT WALL
._ BAR I BARZ w ,-1. Go HOLD TIME ,
41 K -OUTSIDE I- TH a- I I
'J i C-
-12 l-BARS I AND 2 TL I
TIME (b) SIMPLE TWO-BAR REPRESENTATION Fig. 16 SIMPLE EXAMPLE OF CONDITIONS LEADING TO ELEVATED TEMPERATURE RATCHETING Such an example isdepicted in Fig. 16(a). Here, a component which carries a sustained primary membrane stress, op, isalso subjected to a periodic thermal down-shock on the inner surface, as represented by the histogram on the left in Fig. 16(a). Assume that the material creeps at the temperature TH and that the temperature drop from TH to TL issufficiently rapid for a relatively small inner portion of the component wall to drop almost to TL. while the outer portion remains near the original temperature, TH. Eventually the entire wall reaches TL. The temperature Is then slowly increased uniformly back to TH. This isfollowed by a period of steady operation before another thermal down-shock occurs.
f 29
Now consider the possible effects of the down-shock. During the temperature drop, high secondary stresses are induced, and if these stresses are sufficiently large, plastic yielding will occur on both the inner and outer surfaces of the wall. Yielding has two effects in this example:
it results in an Increment of plastic growth, aiid it leaves a residual stress pattern in the wall at the beginning of the subsequent hold period. The residual stress pattern adds to the existing primary membrane stress, ap, and, because the creep response of the material generally has a nonlinear dependence on stress, more net creep strain is accumulated during the hold period than would have occurred due to the membrane stress up alone. This effect is referred to as enhanced creep, and it adds to the progressive growth caused by the thermal down-shock.
Finally, since the residual stresses are self-equilibriating, they tend to relax during the hold period. Relaxation alters the stress pattern during the subsequent thermal down-shock in such a manner that the time-independent plastic increment of growth may be larger than it would have been had the down-shocks not been separated by a hold period In the creep regime.
To better understand each of the assumed mechanisms in this simplified example, consider the simple two-bar representation of the component wall that Isshown in Fig. 16(b). Bar 1, which has a smaller cross-sectional area than bar 2, represents the inner portion of the compo-nent wall In which the temperature drops rapidly during the down-shock. Bar 2 represents the larger outer portion of the wall which experiences a less rapid temperature drop. The bars are constrained to the same length, and they jointly carry the primary load, P, which initially produces a uniform tensile stress op in both bars. The temperature history for the bars is shown on the left in Fig. 16(b). The temperature of bar 1 drops from TH to TL and subse-quently the temperature of bar 2 drops. The slow heat-up and the hold period are the same as in the actual component wall.
With this simple representation, we can follow the ratcheting process step by step:
(1) As bar 1 is cooled, its stress increases from cp until it yields in tension. When bar 1 reaches TL, bar 2, which is still at Ti, carries a reduced load that may even be compressive.
(2) As the temperature of bar 2 subsequently drops from TH to TL, its stress increases again until it also yields in tension. At the same time the stress In bar 1 decreases and becomes compressive.
(3) If we assume, for simplicity, that neither the elastic modulus nor the yield stress vary with temperature, then no stress change occurs during the reheat phase of the cycle as both bars are heated together from TL back to TV.
(4) At the beginning of the hold period, both bars have yielded in tension. The assembly is longer than it was in its initial elastic state, and it can continue to grow longer with each succeeding cycle. This growth is time-independentplastic rotcheting.
(5) Also at the beginning of the hold period, the stress in bar 2 is tensile, while a compres-sive stress exists in bar 1. The averaged stress in the two bars is,of course, still up. If the first down-shock were followed immediately by another down-shock with no intervening creep period, the residual stresses in the bars resulting from the first down-shock would alter the stresses during the subsequent down-shock, and in this particular case reduce the subsequent increment of plastic ratcheting.
(6) Consider now the hold period. Bar 2 is subjected to tensile stress greater than up, and creeps more than it would if the stress were op. This increased time-dependent growth rate or enhanced creep, which isa direct consequence of the thermal down-shock, (referred to as enhanced creep) contributes to the ratcheting increment during each cycle.
(7) As creep occurs during the hold period, the compressive stress in bar 1 unloads and becomes tensile as the tensile stress in bar 2 is transferred to bar 1. Further, the beneficial effect of the residual stresses in reducing the increment of plastic ratcheting in the subsequent down-shock (step 5) is lost.
30
Although actual specified loading cycles are likely to be much more involved, this simple example serves to illustrate that ratcheting at elevated temperatures is a complex phenomenon consisting, generally, of both time-independent plastic ratcheting and time-dependent creep ratcheting. Depending on the particular thermal and mechanical loading histories, the tempera-tures, and the material behavior, plastic ratcheting and creep ratcheting may occur alone or together, and one type may or may not interact with the other. 4 Likewise, in actual service, events of varying types and severities are intermixed. The residual effects of one event may remain to combine with, and possibly reinforce, the effects of a subsequent event.
Thus, because of the complexity of the ratcheting phenomenon, a reasonable prediction of the actual incremental growth from cycle to cycle in a component subjected to a realistic operating history can generally come only from a detailed inelastic analysis. The elevated temperature Code Case provides screening rules that are based on elastic analysis results and which identify whether ratcheting is or is not a problem to be further evaluated. Also, the Code Case provides a procedure for obtaining an upper bound estimate of ratcheting strains using only the results of an elastic analysis. Since these rules and procedures ignore many of the complexities of the ratcheting phenomenon and the possible interactions involved, they are necessarily very conservative. Nonetheless, they often allow the Designer to evaluate complex inelastic struc-tural behavior without resorting to a costly and time consuming inelastic structural analysis.
6.4 METHODS OF SATISFYING STRAIN LIMITS USING ELASTIC ANALYSES 6.4.1 Rules to Preclude Plastic or Creep Ratcheting If the primary stress intensity and the range of cyclic secondary stress intensity are small enough that they define a point in the elastic E regime of Fig. 12, then no progressive ratcheting in a component will occur. Thus, the elevated temperature Code Case has some screening equa-tions intended for use with elastic analysis stress results that are based on the equations defining the E regime of Fig. 12. This equation is IpL + (PblKr)IMAX+ (QR)mx < Sy (1) where Sy = the average of the yield strength values at the maximum and minimum wall-averaged temperature during the operating conditions being evaluated (PL + (Pb /K)IM = the maximum value of the primary stress intensity, adjusted for bending via KS,during the operating conditions being evaluated (QRq)MAX = the maximum range of the secondary stress intensity during the operating condition being considered Note that the above Equation 1 is evaluated by adding primary and secondary stress Intensities.
This is different from the normal NB-3000 approach where stress levels of different categories are combined at the stress component level prior to determining the stress intensity level.
Equation I assures that there will be no plastic ratcheting. Further, it assures that there will be no creep ratcheting provided that the average wall temperature at one of the stress extremes defining each secondary stress range, QR, is below the creep regime. The latter temperature is defined in the Code such that creep effects do not control the primary stress limits for 105 hr of operation.
In cases where the average wall temperature at one of the stress extremes defining the secondary stress cycle is not below the creep regime, creep ratcheting can be avoided by reducing Sy in Equation 1 to a stress value which is low enough to avoid significant creep relaxation. This value was selected as 1.25S, taken at the highest average wall temperature during the cycle at 104 hours0.0012 days 0.0289 hours 1.719577e-4 weeks 3.9572e-5 months .
'Creep ratcheting can occur in the absence of any plastic now, for example. if a component Issubiccted to a sustained primary membrane stress and a cyclic radial temperature gradient.
31
The above screening limits are intended to preclude plastic and creep ratcheting. However, It is very difficult to design elevated temperature equipment within these limits because the thermal stress ranges are usually too high. Moreover, it Iscompletely safe to allow plastic and creep ratcheting provided that the resulting accumulated strains are kept within safe limits. Further elastic screening limits are, therefore, given In the Code Case which bound the total accumulated inelastic strains to the specified limiting values.
6.4.2 Rules to Limit Accumulated Strains with Creep Ratcheting A relatively new method, the O'Donnell-Porowski method (28), developed under the sponsorship of the Pressure Vessel Research Committee, gives a way of evaluating the total inelastic strains that could be accumulated under creep ratcheting conditions. Only the results of elastic stress analyses are needed in this method to obtain quantitative upper bounds using the creep proper-ties of the material. The creep acceleration due to the secondary thermal stresses is included by deriving an equivalent creep stress *c for the particular combination of primary and secondary stresses which exists. The creep properties can be obtained through use of the isochronous stress-strain curves which are included in the Code Case.
A cylindrical vessel subjected to Internal pressure and a cyclic temperature drop through the wall is solved in detail in O'Donnell-Porowski (28). The resulting upper bound strains are rigorously derived, based on an elastic-perfectly-plastic material model. The equivalent creep stress, accounting for the accelerated creep due to thermal cycling, is shown in Fig. 17.
The effective creep stress, oc, may then be obtained using Fig. 17,5 or the closed form solution given in O'Donnell-Porowski (28). These stresses are then used in isochronous stress-strain curves to obtain upper bounds for the total inelastic strain accumulation, including the strains due to creep ratcheting.
Since these results were derived for a cylindrical vessel, their use is restricted to axisymmetric structures subjected to axisymmetric loading away from local structural discontinuities. How-ever, nonaxisymmetric loads such as the bending of a pipe or vessel may often be conservatively included as axisymmetric loads, and the present rules may be applied.
Further, the average wall temperature at one of the stress extremes defining each secondary stress range QR must be below the creep regime, since no relaxation at the cold extreme of the cycle was considered in O'Donnell-Porowski (28). Minimum isochronous curves, assumed to be 25% lower than the "average" curves given in the Code Case, should be used. The total service life may be subdivided into temperature-time blocks, and the strain increment for each block may be evaluated separately. However, the times used in selecting the isochronous curves should sum to the total operating life. The strain increments for each time-temperature block are added to obtain the total strain. The resultingvalue is limited to 1%for base metal and 'A1% for weld regions.
6.4.3 Operating Cycle Definition for Ratcheting Evaluation In applying the ratcheting rule as just described, it is important to note the significance of cycle definition. The intent of the term [PL + (PbIKr)I MAX isto consider the maximum value of load-controlled stress throughout the operating life, and the intent of the expression Qrange is to consider the maximum secondary stress range throughout life. This is shown schematically on Fig. 18. The reason for this can be illustrated by a hypothetical example, as shown by Fig. 19.
Assume that one initially has a pure applied radial thermal gradient, and the resultant thermal stress is less than Sq, as defined in Case 1331 and Case 1592, and is maintained for a sufficiently long time that the true stress in the wall in the presence of the thermal gradient relaxes down to a very low value (O-A-B). Assume now that the radial thermal gradient Is removed. One will then have a superimposed residual stress which very nearly equals the original thermal stress in
'Note that the (QR)M AX= 3 limit on the abscissa in Fig. 17 was used for convenience In scaling the drawing and has no other significance.
32
3.0 2.8 2.6 2.4 2.2 : .
'~2.0 2: .) x + y /4u 1
>b1.6 Lui 1.4 Um, 1..
~~0. 0 06-0.4 - 1 E~~
~~0. 2 0 0.2 0.4 0.6 0.8 1.0 PRIMARY STRESS X - bMX. 5 S~~
Fig. 17 EFFECTIVE CREEP STRESS, a%, FOR UPPER BOUNDS ON TOTAL ACCUMULATED INELASTIC STRAIN WITH CREEP RATCHETING 33
[a X + {Pb/Kt) MAX IM .1 ,
LLI Am . l- / Q- RANGE TIME Fig. 18 LOAD HISTORY FOR AN EXAMPLE WITH RATCHETING the presence of a radial gradient (Point C). If one now applies a pressure cycle which approaches the maximum allowable primary stress, then an incremental deformation can result (D-E). If one now holds the pressure stress until the residual thermal stress relaxes out (E-F), then re-moves the pressure stress (G), the cycle can be repeated. Applying this to the special case where two separate cycles are defined, one a thermal cycle and the other a pressure cycle, then each of these cycles can be shown to independently satisfy the elastic analysis ratcheting rules and yet, taken together, they will result in incremental deformation.
The interpretation of the rules given in the above example is clearly conservative. If the applied radial thermal gradient is only present for a very short time and no stress relaxation occurs, then the removal of the radial gradient will leave no residual stress to interact with the applied primary stress in the pressure cycle. In extending the elastic analysis rules, the possibility of using dif-ferent bounding techniques on different cycles was explicitly recognized in the Code Case
[T-1 324(e)] . Thus it is possible to use inelastic analysis methods to calculate the strains for a selected number of maximum strain cycles provided that one adequately accounts for the inter-action effects with the remaining cycles being elastically evaluated.
6.4.4 Experimental and Rigorous Inelastic Analysis Verification of Elastic Ratcheting Rules An experimental and analytical study of ratcheting in a simple structural component is described in Corum and Sartory (30). A straight pipe from a well-characterized heat of Type 304 Stainless Steel was subjected to a series of thermal down-shocks followed by sustained periods of elevated temperature operation under internal pressure. The test was performed in a special sodium test facility built for the purpose. The inelastic analysis predictions were obtained using a one-dimensional finite-element procedure. Good agreement between the measured and predicted ratcheting behavior was found.
To further validate the upper bound method, an analysis [Pickel et al. (31)1 was accomplished which applied the upper bound method to over fifty thin-walled cylinders subjected to sustained 34
ELASTICALLY CALCULATED PRESSURE STRESS ELASTICALLY CALCULATED THERMAL STRESS
+0y ACTUAL STRESS V)
WD S u E P. TINE yENHANCED CREEP RATCHETING
(-)
Fig. 19 STRESS HISTORY FOR AN EXAMPLE WITH RATCHETING primary and cyclic secondary stresses that had been analyzed using arigorous inelastic finite element computer code. Comparison of results shiwed the upper bound results to always be conservative relative to the rigorous inelastic analysis results.
7.0 CREEP-FATIGUE Following a discussion of the correlations and evaluation methods that were considered by the Code Committee, the basis for the interaction rules is presented. Then, the text describes the creep and fatigue design curves, mean stress and multiaxial effects, and the rotating principal strains. Finally, the background and intent of the rules and limits for use with elastic and inelastic analysis are treated.
7.1 CORRELATION AND EVALUATION METHODS Many theoretical correlations and evaluation methods for creep-fatigue have been proposed during the last twenty years without conclusive evidence that a universal method exists (32)-(46).
Kitagawa and Weeks (47) compare five analytical methods (including the methods of Campbell (39) and Coffin (60)) for correlating the results of hold-time fatigue testing. Linear damage rules using time ratios for creep damage appear to be at least as good as, and perhaps better than, linear damage rules using strain ratios or the frequency-modified fatigue life equation of Coffin (60).
35
The Committee chose the creep-fatigue interaction approach wherein damage due to creep is accounted for on a time-fraction basis and damage due to fatigue Isaccounted for by using Miner's cumulative damage criteria. The allowable total damage is based upon observed material behavior and is a function of the calculated damage for both creep and fatigue.
7.2 DERIVATION OF THE CREEP-FATIGUE INTERACTION RULES Subarticle T-1 400 describes the general rules for the damage summation which is used to assess the adequacy of the component to withstand the specified cyclic thermal and mechanical loadings. These rules were originally based on the behavior of AlSl Type 304 Stainless Steel when subjected to hold times at peak tensile strain.
Material specimen creep-fatigue tests data covering the effect of hold time on the life of Type 304 Stainless Steel were described in references (48) through (51). Hold times were introduced individually both in the tension and compressive portions of the fatigue cycle and under several different strain rates. The strain rates varied from 6.4 x 10- In/in/sec to 4 x 10-6 in/in/sec.
Hold times Introduced into the fatigue cycles ranged from 0.1 to 600 minutes. Strain ranges varied from 0.25% to 4%; however, most of the strain rate and hold time data were generated on %2% and 2%total strain ranges.
Using data in references (32), (34), (48-51), the effect of hold times Introduced into a fixed strain cycle are shown in Fig. 21. Note that a marked reduction in fatigue life is observed when only tensile hold periods are introduced into the fatigue cycle at fixed strain. Fig. 22, taken from references (32), (33), (43), (52) and (53), compares hold time effects for various strain ranges for 21 Cr -I Mo, 1 Cr -I Mo, and 1 Cr -1 Mo -0.25 V steels.
It is observed from Figs. 21 and 22 that hold time has significant influence on cyclic life, but mainly in the low strain ranges which are the ranges of most interest to pressure vessel Designers.
Most of the available test data are for strain ranges greater than the range of interest to the Designer. It was thus necessary for the Committee to choose a method of correlation and, with this method and the available data, extrapolate to strain ranges and time of interest to the Designer.
The life-fraction rule (T-1411 of Case 1S92) was adopted based largely upon a life-fraction evaluation of available data by Campbell (39).
An interaction value was also determined from this evaluation. The life-fraction concept was not based solely on stainless steel behavior. It was also applied by Wundt (40) to very limited test data on 1 Cr-i Mo-0.25 V steel of Krempl and Walker (53), and the general behavior appeared the same as for austenitic steels. Other investigators [references (38), (41), (47)1 have applied the life-fraction rule to different alloys as well as austenitic stainless steel and observed similar behavior. Subsequent to providing creep fatigue rules and fatigue curves for austenitic stainless steel in Case 1331-5, an evaluation of Ni-Fe-Cr, Alloy 800H hold time data was made and fatigue curves and a D value for this alloy were included in Case 1331-7. The hold time rules for stainless steel were based on 1200 F behavior. Thus, at temperatures lower than 1200 F, the rules were anticipated to be conservative. Subsequent test data at 1000, 1050, and 1100 F [references (55), (56), and (57)1 have shown the data to fall within normal scatter, further supporting the Code rules.
7.3 CREEP AND FATIGUE DESIGN CURVES Equation 5 of T-1411 of Case 1592 or Equation 14 of Cases 1331-5, -6, -7, and -8 is the linear life-fraction relationship discussed previously. The terms are explained in the Case; however, the following additional description is offered.
The Fig. 15 fatigue curves of Case 1331-5, -6, -7, -8, and Fig. T-1420-1 of Case 1592 are design curves. The design curve is constructed by reducing the best-fit curve of continuous cycling 36
I
. { ..
I0.0. II I .
{. & . . . . .-
l
.l .l .l l zl e
'I I.- a I I I I,._
CT = TOTAL STRAIN RANGE PP = PUSH-PULL TEST BEND = REVERSE BENDING TEST T a HOLD TIME IN TENSION C a HOLD TIME IN COMPRESSION N a CYCLES TO FAILURE WITH HOLD TIME Nf = CYCLES TO FAILURE WITHOUT HOLD TL[ME 1000 LINE 1 0 304 SS - GE-NSP - 1200 1 CT 1.98% PP -T LINE 2 0 304 SS -GE-NSP -1200 1 CT 0.49% PP -T -
FI LINE 3 0 304 SS - GE-NSP - 1200 1 -CT 0.25% Pp -T -
IC LINE 4 0 316 SS - DAWSON - 1112 F CT 2.86% PP -T F
LINE 5 A 316 SS - DAWSON - 1112 1 It 2.86% PP -T
.CT LINE 6 T 316 SS - DAWSON -1112 F -ET 3.0% BEND -T _
100 LINE 7
~~316 SS - DAWSON -1112 F -cT " 1.6% - BEND -T V 304 SS - GE-NSP -1200 1 -CT 0.49% - PP - T&C-C.~~
LINE 8 V 304 SS - GE-NSP -1200 1 CT aa 1.98% - PP - T&C:
U 304 SS - GE-NSP - 1200 F -eT 1.98% - PP -c A 16 SS - WAIKER - q95 F -ET a 2.17% - PP -C 10 5 316 SS WALKER 950 FF CT Sa 3.69% - PP -T -
S C 316 SS WALKER 950 I CT 5.17% - Pp -T i
1.0 0.1 = \ 3 I I I I{ ,1ii , , I \1, I I I 1I. 1
, . . ...... .............. I .
I . .
II .
111 0.1 1.0 100 1000 HOLD TIME (min)
Fig. 21 EFFECT OF HOLD-TIME ON FATIGUE LIFE OF AUSTENITIC STAINLESS STEEL 37
10 w
0o 10 1 2
910 10-2 10-4 10 104 107 CYCLES TO FAILURE Fig. 22 EFFECT OF HOLD-TIME ON FATIGUE LIFE OF VARIOUS STEELS
fatigue data by a factor of 2 on total strain range or a factor of 20 on life, whichever results in a minimum value. The present curves are higher on the low cycle end than the Case 1331-4 curves.
Extra design margin was built into the original 1331-4 curves to account for hold times, slow strain rates and inaccuracy in calculating inelastic strain by elastic methods. The current curves do not contain design margins for the above conditions and these conditions must be evaluated by other steps of the fatigue analysis. Figure 15 and T-1420-1 curves are generally used only for inelastic analysis where strain-time relations are rigorously calculated using appropriate plasticity and creep solution methods. The rigorous inelastic analysis eliminates the need for extra design margin since the strain range can be computed more accurately than by elastic methods. Also, the effects of hold times and slow strain rates are calculated by the second half of Equation 14 and Case 1331.5, *6, *7, and -8 and Equation 5 of Case 1592 using the integral form for creep damage shown in 5.3(b) and T-1 420.
Figure 15 and T-1 420.1 curves may be used for elastic analysis only for cases of continuous cycling at strain rates equal to or greater than those shown on the fatigue curves. If there are hold times at elevated temperatures or if the cycling is slow, then creep damage is being intro-duced which is not accounted for in the elastic analysis rules of 5.3(c)(2) and T-I 433.
Figure 15 of Case 1331-5, *6,-7, and -8, and T-1420-1 fatigue curves for austenitic stainless steel do not show a difference between 1000 F and 1200 F. Test data from references (48)-(51),
and (65) were plotted for temperatures of 800 F,1200 F, and 1300 F. Data for 1000 F, Jaske et al. (59), appear very similar to the 1200 F data of references (48)-(51). In fact, some of the 1000 F data points fell below the average 1200 F curve. The 1200 F curve was then used as a representative curve from 1000 F to 1200 F. Curves between 800 F and 1000 F were inter-polated by linear scaling on the log-log scale since there were no data in-between. The 1300 F data, Brinkman, et al. (65), fell distinctively below the 1200 F data and a separate curve was constructed for 1300 F.
Subsequent to originating the Fig. 15 and T-1420-1 fatigue curves, test data ranging from 800 F i to 1200 F, references (55) and (63), indicate a definite difference between 1000 F and 1200 F.
Figure 10 of Weeks et al. (55) compares the temperature effect. Fatigue data for Ni-Fe-Cr, Alloy 800H were available from references (57) and (64) for 800 F, 1000 F, 1200 F and 1400 F.
Figure 15 and T-142D-1 curves for these temperatures were constructed in the same manner as for austenitic stainless steel. Intermediate temperature curves were interpolated linearly on a log-log scale. Figure 11 curves were constructed using the Fig. 15 and T-1420-1 curves as a base line and modifying them to account for the life reduction associated with slow strain rates and hold times. Figure 11 and T-1 430.1 curves are reduced below those in Fig. 15 and T-1420-1 to account for creep damage due to hold times and slow strain rates. The hold-time effect is more severe than the strain rate effect, and thus, the curve is constructed based upon the fatigue life reduction determined for hold times at fixed strains.
The hold-time tests at fixed strain result in pure relaxation, hence the curves in Figs. 11 and T-1430-1 are reduced to account for creep damage due to pure relaxation of peak residual stresses. The relaxation damage is based upon uniaxial test specimen relaxation curves. Creep damage due to primary stresses and the relaxation of secondary stresses is not contained in Fig. 11. This creep damage effect is calculated separately in 5.3(c)(2) and T-1433.
The stress-to-rupture curves, Figs. 16 and 1-14.6 (from which the values of Td are obtained), are minimum stress-to-rupture curves. The factor K' is used to adjust the minimum curve to some other percentage of stress to cause rupture. The K' factor of 0.9 adjusts the curve to about a 89% of the minimum stress-to-rupture curve. In comparison, the primary membrane allowable curve, St, is based principally on two-thirds of minimum stress-to-rupture. The primary bending life-fraction summation is based upon a maximum of 1.25 (213) or 83h/3o of minimum stress-to-rupture by using the factor K, from Equation 7, 3223(c) as a divisor for the calculated PL + P, stress.
39
7.4 MEAN STRESS AND MULTIAXIAL EFFECTS Very little data exists for mean stress effects at elevated temperature. Chowet al. (58) contains three data points for the effect of mean stress on high-cycle fatigue and indicates that the effect is small.
For strain-controlled, inelastic cycling, mean stresses tend to cyclically relax to zero. This tendency is more prominent at higher temperatures as shown by comparing Type 304 Stainless Steel data at 1000 F and 1200 F in Jaske et al. (59). In this work it was shown that significant mean stresses can be introduced by variable-amplitude straining. It is also generally known that mean stress effects become more important when the inelastic strains are small and the cyclic life is large.
Adjustment of the fatigue curves for stainless steels to account for mean stress effects was con-sidered using the method described in the ASME Criteria Document for Section 1I1,Class 1 Components (3). Using this modified Goodmon diagram approach, it was found that no adjust-ment was indicated. At the end of the fatigue curve (106 cycles) plastic straining is still occur-ring (neglecting cyclic hardening) and the adjusted mean stress iszero.
Weeks et al. (55) indicates that there is very little effect on creep-fatigue life of austenitic stainless steels when a mean strain of up to 1.5% is imposed during the tests. This Indicates that the effects of creep ratcheting do not significantly alter the creep-fatigue life when creep ratcheting strains are small. Thus, the Committee has not taken action to modify the fatigue curves for mean stress or strain as there are insufficient data to identify a definite need.
Compressive stresses are considered to be equally as damaging as tensile stresses in computing creep damage. The uniaxial test results of references (48), (49), and (55) would indicate that compressive stresses should be considered less damaging than tensile stresses since strain-controlled low cycle fatigue tests with hold times at fixed strain in compression indicated very little reduction in fatigue life, while tensile hold-time tests indicated a large reduction in fatigue life (Fig. 21). However, the uniaxial loading case does not generally appear in pressure-contain-ing components. The state of stress is most always multiaxial. Until there is additional multi-axial load fatigue data (61), and a method to separate out the compressive strain effects in a general multiaxial stress/strain condition, the Committee chose to use the conservative and simpler approach of treating compressive and tensile strain as equally damaging.
7.5 ROTATING PRINCIPAL STRAINS The rules for rotating principal strains are based on the applicability of the effective strain range for predicting fatigue behavior in multiaxial stress states. In Manson (62) this is presented in terms of the principal strains as Afeguiv = 3 [A(eI -C2)1 + [A(E2-63)2 + [A(C3-ea)J1 112 which can be rewritten as AIequiv 2 c2) +(AE 2 - AE)
A-he2) +( 3- Ae)2 ] :/2 where the algebraic strain (positive for tension) must be used in calculating the change in strain Aci. However, when the principal directions change during the cycle, the analysis must consider the six strain components. That is, the equivalent change in strain is calculated from the change in strain component from some reference time, tref.
Aet. ) = E. W -C,(tra
&ey (t = Y (t). - eY Orae) etc.
40
The equivalent change in strain at any time during the fatigue cycle is AEequivet) = ŽL_(&C - ( 2e) 3
+ (brxy2 +a_2 + AZ 2 )Ia/?
The fatigue damage is estimated based on the maximum equivalent change in strain. This may require the calculations at several times during the cycle when the external conditions are not obvious.
At the present time there are no test data for fatigue under changing principal directions to confirm the validity of the above approach. However, the above approach reduces to Acequiv = Ae, for uniaxial tests with large plastic-strains (e2 Ze 3 = 0.5e). It also provides improved agreement with the change in stress intensity approach at low temperature. This was considered desirable to reduce any discontinuity of the design criteria at the limits of applicability of Subsection NB of Section III.
7.6 RULES AND LIMITS FOR USE WITH INELASTIC ANALYSIS From the inelastic analysis, the stress-strain-time relationships are determined. The linear life-fraction summation of creep damage may then be represented by an integral form t dt f, Ted The original inelastic rules contained in Case 1331-5 were based on the behavior of austenitic stainless steel described in Conway (48) and the analysis of these data which was documented in Campbell (39). Only 1200 F tests data were available so it was assumed that the crcep-fatigue interaction for other temperatures would be the same as that at 1200 F. Test data from refer-ences (55) and (63) at 1100 F support this original assumption. Figure 23 reproduced from Campbell (39) plots the original 1200 F failure data. Figure 24 shows a bilinear trend curve generated by a statistical evaluation of the Conway (48) data and documented in Campbell (39).
From this data the D value for austenitic stainless steel was derived for Case 1331-5.
Figure 25 [reproduced from Brinkman et al. (63)) compares actual failure data at 1100 F with design allowable points of Case 1592. It can be seen that there is considerable scatter and the design factor on life varies from a minimum of approximately four to greater than 40. Figure 26
[reproduced from Weeks et al. (55)] shows the distribution of design factor in bar form for Type 304 Stainless Steel at 1100 F. The design factor ranges from about 7 to 40 for tests where hold times in tension only were introduced. With hold times in compression, the design factor on life increases significantly. This difference in life for tension versus compressive hold-time stresses was discussed previously.
When Ni-Fe-Cr, Alloy 800H was introduced into Case 1331-7, data at 1000, 1200 and 1400 F contained in Jaske et al. (64) were analyzed in Corum (54) in the same manner as was done in Campbell (39) for austenitic stainless steel. Figure 27 taken from Corum (54) shows considerably more scatter than for austenitic stainless steel. The creep.fatigue interaction curve, Fig. 27, shows that a D value of unity was reasonable.
Examination of Fig. 27 reveals that, when damage is primarily due to creep, the D value of unity is more conservative.
7.7 RULES AND LIMITS FOR USE WITH ELASTIC ANALYSIS Elastic analysis rules and limits for creep-fatigue evaluation were intended to be more conserva-tive than inelastic rules. To effect this, a number of assumptions, techniques, and rules were 41
I 1.6 1.5 1.4 1.3 1.2 1.1 1.0 0.9 4J 4
-J 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 N/N5 Fig. 23 CREEP vs. FATIGUE DAMAGE AT 5% LOAD DROP-OFF, TYPE 304 STAINLESS STEEL AT 1200 F (Campbell (39)]
42
I 1.0 0.8 0.6
-S 0.4 AVERAGE TREND 0.2 WMINIMUM TREND\(
0.0 0.0 0.2 0.4 0.6 0.8 1.0 N/Nf Fig. 24 AVERAGE AND MINIMUM TREND CURVES FOR CREEP-FATIGUE INTERACTION FOR TYPES 304 AND 316 STAINLESS STEELS employed. These are listed below.
(1) Adjusted fatigue curves were developed to account for creep damage due to peak stress relaxation during hold times and slow strain rates.
(2) A technique was developed for determining the maximum strain range in the presence of inelastic deformation for use with the fatigue curve.
(3) Rules were established for evaluating the creep damage due to secondary and primary stresses.
(4) The rule for determining combined creep-fatigue damage was established.
Most of the above assumptions, techniques, and rules are quite conservative.
To obtain adjusted fatigue curves that incorporate creep damage due to relaxation of peak stresses, the continuous-cycling fatigue curves for use with inelastic analysis were used for the 43
100.0 10.0 304 STAINLESSO3 *_ 6 316STANLES1 1.0 LI 0 -- LINEAR CREEP/FATU INTERACTION LI 0 n/Nd 0 0 O"D 1 ASSUMING Zn/Nd < 1 0.10 0.01 0.01 0.1 1.0 10.0 I FATIGUE DAMAGE Fig. 25 COMPARISON OF CODE CREEP-FATIGUE ALLOWABLES TO EXPERIMENTAL DATA unadjusted baseline data. This data Isgiven in Fig. 15 of Case 1331.5, -6,-7, -8 and Fig. T-1420-1 of Case 1592. The fatigue curves were constructed in the same manner as was done for the lower temperature Section Ill curves incorporating a design factor of 2 on strain range or 20 on cycles, whichever isless.
7.7.1 Adjustment of Fatigue Curves In order to determine the effects of slow strain rates and hold times, the fatigue-life-reduction factors were plotted from available data (references (48) through (51) for austenitic stainless steels, and references (57) and (64) for solution annealed Ni-Fe-Cr, Alloy 800H]. Such plots for 44
-304 SS 26 HOLD TIME TESTS AT 593 C (l10o F) iMMTENSION HOLD TIME M TEN-COMP. OR COMP. HOLD TIME ASME CODE CASE 1331 1.0 5 10 20 40 2.5 Fig. 26 DISTRIBUTION OF ACTUAL-TO-CALCULATED LIFE RATIOS FOR STRAIN-CONTROLLED CREEP-FATIGUE TEST SPECIMENS
(.
austenitic stainless steel are shown in Figs. 28 and 29. The fatigue design curves are based on a strain rate of 10-3 in/in/sec or 3.6 in/in perhr. It was assumed that at 10-3 in/in/sec strain rate, the damage mechanism was pure fatigue or that any creep damage included in the data would be already built into the fatigue curves. Therefore, the fatigue-life-reductlon factor was defined to be equal to 1.0 at a strain rate of 1O-3 in/in/sec. All the data points in Fig. 28, with the excep-tion of one for e = 4 x 1O-6 in/inlsec, span two orders of magnitude in strain rate.
The fatigue-life-reduction versus hold-time curves in Fig. 29 are derived from actual data in Conway (48) and Berlinz and Conway (49) for Type 304 Stainless Steel at 1200 F. Extension of the curves beyond data points was accomplished by analytically extending the relaxation curves of Conway (48) and computing creep damage by I dt.
The equation used in Campbell (39) and Conway (48) to fit the relaxation curve was used to extend the curves for longer hold times.
Figure 29 shows data points used to plot the fatigue-life-reduction curves. Open points are for actual test data and solid points are for computed values. Since strain ranges greater than 0.01 will rarely be encountered in actual service, the curve for CT = 0.01 was used for Er D0.01-The fatigue-life-reduction curve for er = 0.001 was constructed using part of the relaxation curve for ET = 0.0025. It was desirable to use relaxation curves for the actual strain range in question: however, there were no available data on ET = 0.001. Use of the lower portion of a 45
100 S AAl 300 A I a 10-
-c a + - 1
/ AVERAGE TREND OF DAMAGE 31 RESULTS FOR TYPE 304 SS 0o (CAMPBELL) 0.1 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 TOTAL FATIGUE DAMAGE, 2 n Fig. 27 CREEP-FATIGUE DAMAGE INTERACTION FOR TENSION HOLD-TIME TESTS OF SOLUTION ANNEALED Ni-Fe-Cr, ALLOY 800H 0.25% relaxation curve for the er = 0.001 relaxation curve should be, if anything, conservative since examination of relaxation curves for higher strain ranges shows that strain hardening slows down the relaxation process, thus increasing calculated creep damage for this case.
The maximum fatigue-life-reduction factors for hold-time effects given in Fig. 29 determined the actual curve used to modify Fig. T-1 420.1 to obtain Fig. T*1430-1 in the Code Case. All test data Indicate that the hold-time effects are more significant than slow strain rate effects. Hence, the fatigue-life-reduction curves generated for hold-time effects were used to construct Fig. T-1430-1 of the Code Case.
The low cycle end of Fig. T-143D-1 was constructed by reducing the number of cycles to failure for a given strain range in Fig. T-1 420-1 by the maximum life-reduction factors for hold-time effects. The high cycle end of Fig. T-1430-1 was constructed in a different manner. The fatigue curves of Fig. T-1420-1 of the Code Case were extended at a slope of -0.12 on a log-log scale.
46
10I3 00 Li g
AVERAGE SL01E e10 A--*4 X Io e
_4 lo13 1o02 lo-I 100 10 STRAIN RATf (in./in.-hr '
Fig. 28 STRAIN RATE EFFECTS ON FATIGUE LIFE REDUCTION
I-IM 0 1O1 HOLD TIME (hr)
Fig. 29 HOLD-TIME EFFECTS ON FATIGUE LIFE REDUCTION FOR TYPES 304 AND 316 STAINLESS STEEL
Then, the fatigue-life-reduction factors from Fig. 29 were used to establish the high cycle end of the Fig. 11 and T-1430-1 fatigue curves of the Cases 1331 and 1592, respectively. Fatigue curves for Ni-Fe-Cr, Alloy 80011 were generated in a similar manner.
7.7.2 Maximum Strain Range Prediction Equation for Use With Elastic Analysis Strain concentrations significantly larger than the elastically calculated stress concentrations can occur when gross yielding or creep strains are occurring in the material surrounding local con-centrations. In a low temperature Section III (NB-3000) design, this is prevented by the use of basic stress limits which assure shakedown to elastic action, or else a strain-concentration factor has to be applied to the elastically calculated stress-intensity range prior to entering the fatigue curve. This strain-concentration factor is based on the shape of the stress-strain curve. An extension of this method into the creep range using isochronous stress-strain curves was attempted. Unfortunately, thisSection III simplified elastic-plastic approach appeared to be unconservative in some cases. The Committee then undertook the evaluation and correlation of the Krempl (66) test data using the Stowell (67) and Neuber (68) methods of predicting strain concentrations.
Test data, elastic-plastic finite element analysis, and the methods of Neuber and Stowell were compared. It was concluded that use of either the Neuber or the Stowell relationship for predicting strain in a notch (when the gross section is inelastic) is conservative. Mowbray and McConnelee (69) and Topper and Gowda (70) give further justification for the applicability of the Neuber method. The use of Neuber or Stowell equations with isochronous strcss-strain curves was thus deemed to be conservative and represent methods by which elastic analysis may be used to compute peak strains in the creep range.
Application of either the Neuber or Stowell methods requires an iterative solution technique, a feature not desired for ASME Code rules. Therefore, an equation not requiring iteration was developed to simplify the calculation and yet maintain conservatism when the net section is undergoing inelastic strain due to creep or plasticity. The equation developed and given as (
Equation 16 of Case 1331-5, -6, -7, -8, and Equation 7 of T-1432 of Case 1592 is as follows:
eCT KCee + KE2 ep + KTCF where CT = the derived maximum strain for the loading condition c, = the elastic strain in the region under consideration, exclusive of strain concentrations Kfe = the theoretical elastic strain-concentration factor e = the inelastic strain in the region under consideration, exclusive of strain concentrations and peak thermal strains EF = peak thermal strain associated with the peak thermal stress intensity as defined in Section III Kr = strain-concentration factor applied to peak thermal strain component, eF The value, eC,is determined by subtracting the elastic strain component; ee from the calculated total nominal strain. The total nominal strain, en, is the sum of the load-controlled strain and deformation-controlled strain, exclusive of strain concentration and peak thermal strain. The load-controlled strain is determined by entering the appropriate isochronous stress-strain curve at a stress intensity equivalent to the load-c6ntrolled stress intensity in the region under con-sideration. The deformation-controlled strain is determined from the elastically calculated stress intensity due to the applied deformation.
en = eloadcontrolled + (Sstrain-controlled /E)
This maximum strain range equation becomes very conservative if there is significant inelastic behavior in the net section. However, it was judged that proper design would not produce 49
3 2 1, i
0 10 20 30 40 50
°n, KSI Fig. 30(a) COMPARISONS OF ANALYTICAL AND EXPERIMENTAL STRAIN RANGES FOR A NOTCHED BAR 50
3 FINITE ELEMENT - CYCLIC 1331-5 INITIAL o-e
-wl - 1331-5 CYCLIC a.-e o EXP. INITIAL CYCLE o EXP. SHAKEDOWN 2
(b) 304 Stainless Steel Kt 3.46 lI_!
1.
JII R /I!
00 0 10 20 30 40 50 on.KS!
Fig. 30(b) COMPARISONS OF ANALYTICAL AND EXPERIMENTAL STRAIN RANGES FOR A NOTCHED BAR 51
3 2 _ Ail -~ , -/ ,
Kt 1.97 _
Hw 8I I 6I
.6 _>
.4
.2 0
0 10 20 30 40 50 e KSI Fig. 30(c) COMPARISONS OF ANALYTICAL AND EXPERIMENTAL STRAIN RANGES FOR A NOTCHED BAR 52
3 2 Kt 3.46 nii 0 10 20 30 40 50 an KS]
Fig. 30(d) COMPARISONS OF ANALYTICAL AND EXPERIMENTAL STRAIN RANGES FOR A NOTCHED BAR 53
significant gross inelastic behavior except for a very few infrequent and severe events, hence the apparent extra conservatism should not be detrimental to the design of components. Figures 30(a) through (d) gives a typical comparison of the Code Case equation, the Neuber equation, experl.
mental data, and finite element analysis results.
When Case 1331-5 was published, the maximum strain range prediction equation contained only two terms:
Er = K, cc + K"1e This resulted in extra large and unwarranted design factors when a skin-type peak thermal strain occurs from a thermal transient and the skin effect is small in comparison to the size of the geometric discontinuity. With the release of Case 1331-7, the equation was expanded to separate out the peak thermal strain from the remainder of the inelastic strain (ep). There is no guidance in the Code Case for determining what value Kr should have; KT would normally be 1.0 for smooth surfaces, but in the case where there ire surface irregularities that are small in comparison to the skin effect depth of the peak thermal strain, there should be a value of KT> 1.0 applied. In the case of a weld which is not ground flush, an appropriate KT should be applied.
Case 1331-5 required that all welds meet stringent contour requirements analogous to the flush weld requirements for Section 1I1. Upon the release of Case 1331-7, the weld contour control was relaxed but a stress analysis of the finished joint was required using stress-concentration and strain-concentration factors appropriate for the worst surface geometry. The ASME Code does not tabulate stress-concentration and strain-concentration factors for welds or other discon-tinuities except for those Included in piping stress indices. The Designer must determine his own KY and KJ factors and justify them in the Stress Report.
7.7.3 Elastic Analysis Rules for Evaluating Creep Damage Due to Primary-plus-Secondary Stresses As discussed previously, the fatigue curves for use with elastic analysis are adjusted to account for creep damage during pure relaxation of peak residual stresses. Creep damage from primary and secondary stresses are evaluated separately. The rules established in the Code Case use the time-fraction approach of for determining creep damage.
Consider the following simplified loading histogram.
-SZ z
to It TIME 54
Pm = primary stress intensity SI = sustained primary-plus-secondary stress intensity for sustained operating condi-tions between cycles S2 = transient primary-plus-secondary stress intensity Case 1331-8 and Case 1592 specified that creep damage from primary-plus-secondary stresses resulting from sustained operating conditions between cycles be evaluated by considering that the stress intensity used for obtaining rd be the lesser of the minimum specified yield strength of the material under evaluation, orPm + 0.5 S,, where St = SI because only the sustained operating conditions between cycles are evaluated.
In Case 1331-5 the stress used for creep damage evaluation was different than that described above. Paragraph 5.3(c)(3) of Case 1331-5 required that creep damage from primary and secondary stresses shall be evaluated with no differentiation between sustained and transient conditions. For Case 1331-5 the stress intensity to be used for obtaining Td would be the minimum specified yield strength, or Pm + O.S S. = S2 -on the previous figure since all operating conditions during the cycle are considered.
The philosophy and assumptions that were used to arrive at the above rules are further discussed below.
Relaxation of residual stresses due to mechanical and thermal loading are a function of the temperature, primary stress, prior strain history, and elastic follow-up. Since the above effects cannot be assessed by elastic analysis, it was assumed in Case 1331-S that the secondary stresses would not relax during the cycle and that the creep damage must be assessed as though the stresses were primary. The maximum calculated elastic primary-plus-secondary stress intensity is not, however, a realistic stress intensity to use for creep damage since the material will yield and the stresses will redistribute. Hence, the stress to be used was assumed to be the lesser of or, or Pm + 0.5 S,. The assumption implies that if (PL + P, + Q) range is greater than 2a., the maximum practical stress that can be sustained during a hold period is a.. If (PL +Pb +Q) range is less than 2a., then after a few repeated cycles the stress will redistribute to a value of approximately Pm + 0.5 S, where S, is the maximum range of (PL + Pt, + Q).
With the issuance of Case 1331-7 and -8, the wording was changed so as to consider only the PL + PER+ Q for sustained operating conditions in order to remove some of the conservatism inherent in Case 1331-5. The logic behind the change was that residual secondary stresses caused by short-term thermal and mechanical transients would relax just the same as peak stresses relax, and creep damage due to the relaxation process would already be accounted for by use of the Fig. 9 (or Fig. T-1430-1 in Case 1592) fatigue curve. The only additional creep damage of concern would involve primary-plus-secondary stresses due to sustained mechanical or thermal loading. In this situation the secondary stresses are conservatively assumed to not relax.
7.7.4 Rules for Determining Allowable Creep-Fatigue Damage for Use With Elastic Analysis The combined creep-fatigue damage allowable was set at D = 1.0 in the cumulative damage equation of In The value of D was chosen to be unity for use with the elastic analysis creep-fatigue rules. It was felt that the conservatisms already inherent in the adjusted fatigue curves and the creep damage rules were sufficient to obviate the need for a reduced value of D as in the inelastic creep.fatigue rules.
The procedures used to evaluate the creep-fatigue damage are straightfoward for simple shapes and loading such as a notched bar subjected to alternating tension and compression, but be-55
P P
A Tf LOAD HISTOGRAM
.n nn P
to ti t2 t3 t4 t5 AXIAL LOAD VS TIME THERMAL TRANSIENT (T 2 - T1 )
Tf T1 T VCf f7\
to tI t2 t3 t4 t5 FLUID TEMPERATURE VS TIME 01 oX(Psi) PM Q02$L Q 0J AXIAL STRESS VS TIME AT SECTION A-A Fig. 31 SAMPLE PROBLEM IN PLANE STRESS 56
P *LOAD I FLUID TEMPERATURE
~~ECCENTRICITYOF LOAD PATH UT. THFRHAL MOMENT' 2 7y dy~~
~~STRE5fINTENS17T~~
~~J S~~
~~YIELO STRESS LC' LOAD-CONTROUED STRESS INTENSITY~~
~~PJA . MC C SOC-DISPLACEMENT CONTROLLED STRESS IKTEtSITY T T. C t,* THEORETICALELASTIC STRESS CONCENTRATION FACTORKg FORTHEFILLET~~
'LC *LOAO-CONTROLLED STRAIN FROM ISCCHRONOJS STRESSSTRAIN CURVE
'DC
~~DISPLACEMENT CONTROLLED STRAIN. SOC/E.~~
Fig. 32 ANALYSIS FLOW DIAGRAM 57
comes much more difficult when the component geometry and loading history arc complex.
Thus, a flow chart for an example problem, which exercises all the features of the elastic analysis procedures but does not contain the complexity of multiaxial geometry and loading, is presented in Figs. 31 and 32.
8.0 DESIGN RULES AND LIMITS FOR BUCKLING AND INSTABILITY The buckling and instability design limits given in Case 1592 differ fundamentally from those provided in Section III in two ways:
(1)Section III provides charts for determining allowable stresses in spherical shells and cylindrical shells with or without stiffening rings loaded by external pressure, and for determining allowable stresses in cylindrical shells under axial compression. Design limits are not provided for other configurations or other loading conditions. Case 1592 provides minimum design factors for calculated buckling loads or times for any case where instability due to compressive loads or strains may be a possible failure mode.
However, it does not provide methods or charts for calculating buckling loads for specific cases.
(2) The design charts of Section III include the effects of initial imperfections, where appropriate, and the effects of temperature upon the short-time stress-strain curve but do not account for the effects of creep. The design limits of Case 1592 are applicable for long-term loadings at elevated temperature where creep effects may become significant and require that the effects of geometric imperfections and tolerances be considered explicitly in all calculations.
8.1 ELEVATED TEMPERATURE DESIGN FACTORS The design factors of Table T-1 520-1 are specified to provide a margin of safety against two types of buckling failure, elastic or elastic-plastic buckling that may occur instantaneously at any time in life, and creep buckling which may cause geometrical instability as a result of creep enhancement of initial imperfections with time. The essential difference between elastic and elastic-plastic buckling and creep buckling is that elastic and elastic-plastic buckling occur with increasing load (or strain) independent of time, whereas creep buckling is time-dependent and may occur even when the loads are constant. Elastic and elastic-plastic buckling depend only on the geometrical configuration and short-time material response at the time of application.
Creep buckling occurs at leads or strains below the elastic or elastic-plastic buckling loads or strains as a result of accumulated creep strains over a period of time. Creep buckling is highly dependent upon the geometrical configuration as well as the time-dependent material response.
In most cases creep buckling occurs in two stages: an increase in initial geometrical imperfections with time due to creep, and instantaneous buckling when the critical deformation is reached for the particular loading condition. Therefore, separate design factors are specified (1) for load and strain to provide a margin of safety against instantaneous collapse at any time during life, and (2) for time to account for the uncertainty in the magnitude of initial imperfections.
8.2 DEPENDENCE ON POST-BUCKLING BEHAVIOR The margin of safety provided by the design factors of Table T-1 520-1 is dependent upon the post-buckling behavior of the component. Three distinct types of post-buckling behavior may be identified as illustrated by curves 1, 2, and 3 of Fig. 33. The solid lines represent the be-havior of theoretically perfect components; the dashed lines represent the behavior of real components with Initial imperfections; and Pr,. is the buckling load that would be calculated using small deformation theory for perfect components.
Components that behave according to curve 1 show considerable post-buckling strength; i.e., it takes an increase in load beyond the theoretical buckling load, Pc,, to get significant deforma-58
I I
- Behavior of theoretically perfect component
-- Behavior of actual 1.0 (imperfect) component
- IP = Critical load determined cr by small deformation 0
-- I theory.
Fig. 33 TYPICAL BUCKLING BEHAVIOR C: 6 0 cr tA 4.
ciw-,
a)-
100 iol 102 103 1 Time (log scale)
Fig. 34 CREEP BUCKLING DEFLECTION-TIME CHARACTERISTICS 59
tion. Rectangular plates compressed In the long direction with edges supported parallel to the load typify this type of behavior. Components that behave according to curve 2 show little or no post-buckling strength; i.e., as the load approaches the theoretical buckling load, the defor-mation Increases very rapidly toward catastrophic collapse. Columns loaded axially and cylinders under uxternal pressure typify this type of behavior. Components that behave accord-ing to curve 3 show post-buckling instability; i.e., equilibrium states exist at loads well below the theoretical buckling load, Pcr, and real components tend to jump to these equilibrium states by processes often described as snap-through or oilconning. It is well known that thin-walled cylinders loaded axially buckle Into a diamond pattern at loads on the order of one-half of the theoretical buckling load, P&,. Spheres under external pressure and cylinders loaded in torsion also exhibit post-buckling instability.
It is apparent that buckling loads calculated for theoretically perfect components using small deformation theory provide conservative estimates for components with post-buckling strength.
Magnification of initial deformation Issmall even at the theoretical buckling load. For compo-nents with little post-buckling strength, initial deformation is magnified as the load approaches Pc,. However, If initial deformation is sufficiently small, the load that causes significant magnifi-cation can approach Pc, very closely, so that Pc, provides an upper bound but close approxima-tion to the buckling load. Magnification of larger initial deformations may become unacceptably large at loads well below Pc,. In this case it is necessary to calculate the actual load deflection curve, taking account of initial deformation and perhaps using large deformation theory. For components that show post-buckling instability, Pcr grossly overestimates the buckling load.
It is intended that the load factor of 3.0 for elastic buckling be applied to the actual buckling load, taking into account the effects of geometric imperfections and tolerances whether initially present or induced by service.
Consideration of geometrical imperfections is required to provide an adequate design margin for components which are imperfection-sensitive (post-buckling unstable), typified by the be-havior shown in curve 3 of Fig. 33. For components that are imperfection-insensitive (post-buckling stable), typified by the behavior shown in curves 1 and 2 of Fig. 33, the load-controlled design factor of 3.0 may be shown to provide a sufficient margin of safety without Inclusion of the effects of geometrical imperfections so long as their magnitudes are modest. For these components, the effects of geometrical imperfections are relevant only to creep buckling calculations.
A reduced design factor of 2.5 is used for elastic-plastic buckling where the effect of plasticity is significant, in recognition of the fact that strain-hardening tends to make the post-buckling be-havior more stable and thereby reduces sensitivity to geometric imperfections. It is well known, for example, that when an axially loaded cylinder buckles in the bellows mode, typical of in-elastic buckling, predicted buckling loads coincide closely with those actually measured.
8.3 TIME DESIGN FACTOR The magnitude of the time design factor for creep buckling was selected, based on observed scatter in typical creep buckling test data, to account for uncertainty in initial imperfection and sensitivity to creep rate. The sensitivity of creep buckling time to initial imperfections Is illustrated by the deformation-time relations in Fig. 34. Although typical of the behavior of axially compressed columns and cylinders under external pressure, these curves generally represent the behavior of any component with at least some post-buckling strength, such as those represented by curves 1 and 2 of Fig. 33. In general, a structural component will deviate initially from the perfect geometrical shape (straight, round, cylindrical, etc.) by some small amount. Under a system of loads, below those that would cause elastic or inelastic instability, the initial deflection is magnified with time as a result of creep. The deflection increases until -
the geometrical configuration becomes unstable, as shown at point A In Fig. 34, and buckling 60
occurs. The final buckling load in this situation would be dependent upon the instantaneous elastic or inelastic properties of the material at the time of buckling.
As can be seen in Fig. 34, a small difference in initial deflection is associated with a large dif-ference in the time-to-buckling. Geometrical deviations within the range of manufacturing tolerances can easily result in variations of a decade or more in buckling time. It might be argued that the manufacturing tolerance would provide a safe upper limit on initial imperfec-tion. But that only applies to the material as supplied. Fabrication and heat treatment can alter the geometry so that the actual magnitudes of initial imperfections in the as-fabricated condition are not well known. Considering the extreme sensitivity of buckling time to initial imperfection and the sensitivity of creep rate to various material parameters, a time design factor of 10 is not overly conservative.
A time design factor was considered to be necessary to provide an adequate margin of safety based on the current state of knowledge. However, it was recognized that considerable difficulty will occur in predicting component behavior at times up to 10 times design life.
In addition to properties, the question arises as to what load history to assume beyond the end of life. It would be prohibitively expensive to reproduce a detailed inelastic analysis of a com-plex load history for 10 times the design life. The intent of the Code Case is to identify the dominant features of the load history and construct, for repeated application, a simplified load history which has the same net response. The simplified load history could be as simple as a steady state load condition.
It is apparent that, although the time design factor is philosophically desirable, there is consider-able computational difficulty with its application. Further Code development is presently underway to establish alternate limits which account for sensitivity to initial imperfections and variation in material creep rates without requiring calculations for times greater than the design life.
8.4 EFFECT OF INITIAL IMPERFECTIONS Section T-1 520(b) implies rather strongly the need for a complete and detailed inelastic analysis to account for the effect of geometry changes due to creep. In reality there are many situations in which a simplified analysis, which may assume an idealized component geometry, is perfectly adequate to assure structural stability. In this regard it should be noted that the design factor of 10 on time essentially assures that creep enhancement of initial geometric imperfections will be negligible for the design life. The geometrical configuration will not deviate significantly from the initial configuration and the instantaneous buckling load (load design factor) will remain essentially constant during life. Therefore, if the design factor of 10 on time is met and there is reasonable margin above the required load design factor at the beginning of life, calculation of the load design factor at every point in time is not needed.
8.5 STRAIN-CONTROLLED BUCKLING In the determination of design factors, distinction is made between load-controlled buckling and strain-controlled buckling (see T-1510(b)l. Load-controlled buckling is characterized by situations where the application of load continues after buckling; i.e., the load is not relieved by buckling. Strain-controlled buckling is characterized by loads which are strain limited, thus, in a sense, relieved by buckling. Examples of load-controlled buckling include spheres and cylinders under external pressure, columns and cylinders loaded axially by a dead weight, and pipingelementsloaded by a moment or torque due to dead weightor seismic motion of connecting components. Examples of strain-controlled buckling included heated plates and shells restrained from in-plane thermal expansion, buckling of a cylinder near the intersection with a hemispheri-cal head due to internal pressure, and buckling of plates and shells due to severe local variations in temperature which cause large compressive stresses. Thus, load-controlled buckling is 61
characterized by catastrophic collapse whereas strain-controlled buckling is a self-limiting process. Once the structure deforms, the strain is accommodated and the load is relieved. As a matter of fact, in the deformed state, the structure is able-to accommodate increasing amounts of strain mismatch with smaller changes in the deformed shape. Initial deformation, rather than Increasing additional deformation, tends to decrease additional deformation. Although strain-controlled buckling must be avoided or conservatively bounded to guard against failure by fatigue, excessive strain, and interaction with load-controlled Instability, the minimum strain design factor can be lower than the minimum load design factor because of the self-limiting nature of deformation in the post-buckled state and the relative insensitivity to initial imperfec-tions. Thus a design factor of 1.67 is used for strain-controlled buckling. A design factor on time is not required because strain-induced loads are reduced concurrently with resistance of the structure to buckling when creep is significant.
Note that for thermally induced, strain-controlled buckling, the material properties should be taken as those at the actual operating temperature in calculating strain design factors. Although, strictly speaking, this is nonconservative, the nonconservatism vanishes as the actual strain ap-proaches the calculated buckling strain. Furthermore, the strain design factors are not employed to guard against uncertainty in operating temperatures.
9.0 SPECIAL LIMITS AND CONSIDERATIONS 9.1 REQUIREMENTS FOR WELDS A significant portion of the rules for low temperature design in Subsection NB are concerned with weld materials, design at welds, fabrication of welds, and inspection of welds. However, the analysis requirements for weld regions are minimal, and the weld region is modeled as an extension of the base material with identical properties. The attitude of the Code rules toward weldments did not stem from a belief that weld regions are the same as base material. Instead the attitude reflects a belief that welds can be made as strong as base metal by means of proper control of weld fabrication, materials, and processes, and further, that bend tests will assure that the actual welds will have the necessary ductility in the weld region. With these controls on fabrication and materials, weld regions may be considered at least as strong as the base metal, -
and they can be conservatively modeled as extensions of base material.
For elevated temperature design, two new concerns caused the prior weld region assumptions to be re-examined. The first concern was that the long-term exposure to elevated temperatures could lead to changed mechanical properties by way of diffusion processes within the material. These are difficult to detect by tests on the deposited weld metal since the heat-affected zone between weld and base materials consists of a continuously changing chemistry of alloys over a region often a fraction of an inch thick. The second concern was that a weldment with strength greater than base metal may produce a poor structural joint since differences in material plastic-ity and creep properties can lead to discontinuity stress and strain patterns similar to those characteristic of a bimetallic junction with different coefficients of thermal expansion. This could lead to extra strain in the neighborhood of weld metal that in turn could lead to failure during elevated temperature service.
The Code personnel judged that weld regions needed good ductility more than an abundance of strength beyond that of the base metal. Unfortunately, the major source of data on weld region materials was generated only for austenitic stainless steel material and seldom covered anything but properties of the weld metal itself. Even this limited data showed that weld ductility, as expressed in elongation at fracture, was often far below values for base metal, especially in uni-axial creep-rupture testing [see Fig. 13(b)] . The Committee believed that more than words of general guidance were required for weld regions. Thus, the rules of Appendix T (see T-1710) not only warn of limited ductility but also restrict total calculatedinelastic strain values. Due 62
to the lack of data for weld region materials, the restriction is only an interim measure for in-fluencing component designs, and the degree of conservatism is not known for the method.
The method requires that weld regions be assumed to have properties identical to the surround-ing base metal, but the calculated total inelastic strains shall not exceed one-half the values allowed in base metal. This is not the same as saying that all parts of the weld regions have one-half the ductility of base metal. Even with these restrictions, some parts of the weld region may undergo greater elongation than in the nearby base metal.
9.2 RULES AND LIMITS FOR BOLTS 9.2.1 Design Conditions The intent of Design Condition rules isto keep the primary stresses below the lesser of one-third the expected minimum yield strength and the criteria for allowable stresses established in Sec-tion VII1, Division 1 of the ASME Code. The primary stresses are those required to resist the internal design pressure and to provide an adequate seal in terms of Section III requirements.
Section VIII criteria govern the allowable stresses at elevated temperatures where time-dependent properties predominate; however, for most of the temperature range, time-Independent proper-ties govern. The combination of the lesser of one-third yield strength and Section VI II allowable stresses provide a smooth transition of design allowables between Section 11bolting rules at low temperatures and the extension of Section III to elevated temperature.
9.2.2 Maximum Average Stress Through the Bolt Due to Pressure Loading The intent of this paragraph in Case 1592 is to limit the normal pressure stress sustained by the bolt to the lesser of one-third the yield strength at temperature or YaSt of a structural material.
The Sm1 values for bolting are one-half of those values given for structural materials in Case 1592. A design factor of approximately 2 is utilized in Section III for Sm values of bolts as compared to structural materials, and this philosophy was also used for the elevated temperature rules.
9.2.3 Maximum Membrane Stress in a Bolt Cross Section The intent herein is to specify a maximum membrane stress for bolt preload which allows the Designer to neglect creep-rupture damage in his structural evaluation. If the bolt preload mem-brane stress is kept to the lesser of two-thirds yield strength and St value for a structural material, then creep-rupture damage in the shank does not have to be evaluated as a function of the number of bolt-tightening applications.
If the Designer wishes to preload a bolt in excess of the stress limits established above for various reasons such as minimizing potential leakage, he may do so, but at the expense of evalu-ating the creep rupture damage for each bolt-tightening application.
9.2.4 Maximum Membrane-Plus-Bending Stress in the Bolt Periphery These rules and limits are intended to limit the bolt cross-sectional stress induced by a combina-tion of bending-plus-membrane loadings that may result from flange rotation. The maximum periphery stress value is limited to the lesser of the yield strength or 2KtSt [the St values are those taken for bolting given in curves of 1-14.13 of Case 15921. The stresses are allowed to exceed this value provided ihe possibility of creep-rupture due to bending is guarded against by computing the creep-rupture life fraction expended for each bolt torque application. This calculation assumes that the bolt isorientated in the same position for each torque application so that the maximum bending stress is always applied to the same bolt region. The allowable use fraction of 0.67 was obtained by multiplying the use fraction of 0.5 for membrane loading by Kr for a solid circular cross section.
63
9.2.5 Creep-Fatigue Because of the critical application of bolts at high temperature, fatigue exemptions are not permitted as in design for low temperature. For notches as exemplified by screw threads, fatigue-strength-reduction factors of 4.0 have been shown to be adequate. Unless it can be shown by tests that a lower value is justified, the use of lower stress-concentration factors is not permitted.
A strength-reduction factor of 1.8 is used for creep-rupture as well.
9.2.6 Emergency and Faulted Conditons The philosophy for Section III Isextended to elevated temperature in that Emergency and Faulted events must be considered if they exist; however, the Sm and Smt values for structural materials (Table 1-14.3) may be utilized for the evaluation.
9.2.7 Strain Limits The strain limits for structural materials are also applied to bolting. The previously discussed limits should preclude membrane and bending strains from exceeding 1 and 2% strains, respectively. Bending strains of 2% could exceed the functional requirements of most bolt applications; therefore, the membrane limits should be carefully studied in terms of the specific applications to ascertain whether or not they should be reduced.
9.3 ELASTIC FOLLOW-UP There are two areas of the Code Case which require consideration of elastic follow-up. One area is the classification of secondary stresses with a "large amount of elastic follow-up" as a load-controlled quantity in paragraph 3213(a).
The only definition of elastic follow-up currently in the Code Case is contained in paragraph 3138 which is a modified version of the discussion in the Power Piping Code, 831.1, on local overstrain. The definition and examples of paragraph 3138 relate to the classification of load-controlled quantities. In practice, piping is the most common application of the paragraph 3138 definition and the term "large amount of elastic follow-up" would usually apply to an analysis of a complete pipe line. It is difficult to quantitatively describe what is "significant elastic follow-up.
Robinson (71) in 1955 published a paper which explored the possibility of localized creep strain concentrations in elevated temperature piping systems. Robinson started with the example of a bolt In a rigid flange, which is a case of pure relaxation, and continued through to the case of four large bending loops in series with four smaller loops of half-size pipe in parallel.
Figure 35 is a summary of the behavior of a bolt in an unyielding flange. The stress and creep strain as a fraction of initial elastic extension are shown as a function of time. In this example, the creep extension in 10,000 hr is 0.75 times the initial elastic extension. Figure 36 isa similar plot for a simple beding loop. 1I this case, the creep extension is 0.97 times the initial elastic extension. The final stress is also somewhat higher. Finally, Fig. 37 plots the results for smaller diameter pipe loops in series with large diameter pipe. In this case, the 10,000 hr creep strain is 3.47 times the initial elastic extension and the final stress issignificantly higher than the pure relaxation case.
In the context of the stated rules in 321 3(a), which are largely based on the results of Robinson and others [references (71), (72), and (73)l, the intent is to consider restrained thermal expan-sion stress in systems with localized weak areas, such as Fig. 37, as primary. However, restrained thermal expansion in a well-balanced system, as represented by Fig. 36, should not be considered primary.
The other part of the Code Case which considers elastic follow-up is in Appendix T, and it relates to the application of elastic analysis to the satisfaction of strain limits. Further illustra-tion of elastic follow-up concepts is necessary in order to explain the intent of the rules in Appendix T.
64
15,000 0 CD
-LUJ VI.
10,000 L C 05 LU 0.5 5,000 0 5 coRESIDUAL .
______ i ZIL.1.0 0 1,000 5,000 10,000 TIME, HR.
Fig. 35 BEHAVIOR OF BOLT IN AN UNYIELDING FLANGE Consider two bars in series as shown in Figs. 38 through 40 subjected to a displacement, B. The stress and strains associated with the displacement, 6, would nominally be considered deforma-tion-controlled quantities. However, depending upon the relative stiffness of the two bars, the stresses in each will require separate interpretation of the Code Case rules. Consider the follow-ing cases:
In Fig. 38, the area of bar A is very much larger than B. In this case all the deformation will take place in B. This is analogous to a local thermal stress such as a small hot spot in a vessel wall. In essence, there is no elastic follow-up because the displacement-induced load in B causes a negligible change in the displacement of A. Therefore, the resultant load history in B can be considered solely on the basis of creep-fatigue.
In Fig. 39, A and B are of equal area and length and operate at the same temperature. When subjected to a displacement 6, the bars deflect an equal amount and this displacement does not change even if yielding or creep takes place. This case is analogous to the stresses produced in a vessel wall due to a radial thermal gradient. Radial thermal gradients remote from discontinu-ities produce stresses which are a function of the radius only, and the net strain is a function of 65
a
-I -TC 0
15,000 0 LAJ P-U.-
0.
rI- 10,000 0.sW 0.
U VI cn Ln 5,000 I .5 E -J 1'-
0 0 5,000 10,000 1 ,000 TIME, HR.
Fig. 36 BALANCED LOOP BEHAVIOR the superimposed primary stress. The case of acylinder under primary stress with a superimposed cyclic thermal stress Is the basis of the rules for the application of elastically calcufated stresses to the prediction of strain. Thus, stresses due to radial, through-the-wall, temperature variations are specifically exempted from the restriction that secondary stresses must be considered primary in applying the O'Donnell-Porowski (28) technique incorporated in T-1 323 of Case 1592. The intent, as will be subsequently described, isto consider displacement-induced stresses with elastic follow-up as primary in applying the O'Donnell-Porowski technique.
In Fig. 40, where there Iselastic follow-up, the area of B issmaller than A but still strong enough to cause an initial deflection in A. First, consider the mutual deflections of A and B if the elastically calculated stress in A is less than its yield strength. The elastically calculated deflections of A and B will be as shown. However, since the actual stress in B will not exceed its yield strength, the actual load and resultant deflection in A will be less than the elastically calculated load and deflection. Since the deflection in A and B must add up to the total applied deflection, the deflection of B will be greater than elastically calculated. This isan example of elastic follow-up. Next, consider the relative deflections of A and B assuming that creep is taking place; the area of B issmaller than A and A's Initial deflection issignificant with respect to B.
Since the stress in B isgreater than the stress in A, the stress In Bwill relax at a faster rate. As the stress in B relaxes, the stress and deformation of A will decrease (i.e., tend toward its un-loaded position) and the deformation of B will increase; again this iselastic follow-up. Further-more, if A isa relatively long bar, which stores a large amount of elastic energy, then it will tend to exert an almost constant load on B since the deformation of the relatively short bar, B,will 66
-I I
I F - a- -w- F 1.0 CLU 10,0002.0 Et Lii X 5,000 so 0 C-l<
3.0 5,000 04.00 50 1 cCc 5.0 0 - 6.0 0 1,000 5,000 10,000 TIME, HR Fig. 37 FOUR LARGE BENDING LOOPS IN SERIES WITH FOUR SMALLER LOOPS OF HALF-SIZE PIPE IN PARALLEL not relieve the load in the long bar, A. Thus, if there isa large amount of elastic follow-up, the load on B will be practically constant and can be considered primary even though the initial source of the load was a displacement. The presence of elastic follow-up results in amore constant load, slower stress relaxation, and more strain accumulation as compared to cases where there is no elastic follow-up. Ineffect, a purely elastic analysis will tend to underestimate the creep-fatigue damage (because of slower stress relaxation) and strain in the presence of elastic follow-up. This has important implications to the implementation of elastic methods for calcu-lating the strain accumulation in some common structural configurations.
Consider the case of the stresses produced by the temperature difference between a nozzle and the shell to which it isattached. This can be idealized as a built-In cylinder as shown in Fig. 41.
The initial elastically calculated deflection curve is shown as 0-A. If there were no moment resistance at the built-in joint, then it would behave structurally as a pinned joint as shown by curve O-D. If the initial stress at the built-in end exceeds yield, then a partial plastic hinge will 67
/
/aI
/ /
/ /
/ A UNDEFLECTED
/ /
/ /
/
/
I4
`I /
6 I
I0 I4
/
/
/ A fr DEFLECTED AREAA>> AREAB Fig. 38 TWO-BAR MODEL WITH STRAIN CONCENTRATION BUT NO ELASTIC FOLLOW-UP LA = LB UNDEFLECTED
+
DEFLECTED A =zB AREAA = AREA 8 Fig. 39 TWO-BAR MODEL WITH NO STRAIN CONCENTRATION OR ELASTIC FOLLOW-UP 68
OA ELASTICALLY CALCULATED DEFLECTION A B 6 B= 6 -° A I_
K H6 ACTUAL INITIAL DEFLECTION WHEN BAR B YIELDS A B SO°B= 6 - 6'A 8 A ACTUAL DEFLECTION AFTER CREEP AREAA >AREAB Fig. 40 TWO-BAR MODEL WITH ELASTIC FOLLOW-UP AND STRAIN CONCENTRATION result with an increase in strain at the joint as shown by O-B. If the joint creeps, there will be further strain redistribution as shown by curve O-C. In effect, the angular rotation of the relatively localized high stress area at the joint isbeing driven by the lower stressed, beam-on-elastic-foundation behavior of the cylinder. Thus, the built-in cylinder isan example of elastic follow-up. If one were to apply the results of the elastic analysis directly, as in the case of stresses generated by a through-the-wall gradient, then the surface strain at the builtin end would be underestimated. This is the reason for the restriction in T-1 324(d) which states that any secondary stress with elastic follow-up must be considered primary for purposes of that evaluation. This isconservative in that it assumes the built-in cylinder effect will have alarge amount of stored elastic energy. Unfortunately, there have not been sufficient inelastic analysis of actual joints as of the drafting of these rules to permit anything other than this conservative assumption. However, it isrecognized that specific geometries, materials, operating conditions, and analytical assumptions may be demonstrated to be sufficiently conservative that the noted restrictions on secondary stresses need not be invoked. If this Is the case, the justifica-tion for not considering secondary stresses with elastic follow-up as primary stresses must be included in the stress report.
69
S T1 > To 0>
0
=
Fig. 41 BUILT-IN CYLINDER WITH ELASTIC FOLLOW-UP Note that the restrictions of T-1 324(d) do not apply to T.1 322 and T-1323. This is because the stress limits for these cases are sufficiently conservative that the additional restrictions on elastic follow-up of secondary stresses are not warranted.
9.4 LOAD ENVIRONMENT HISTOGRAMS A histogram, which is required by the Design Specification for each component, is a graphical representation of how the design parameters (such as pressure, temperature, force or flow) change with time during a particular event, and a histogram can depict the sequential order of events during the lifetime of a plant. These are the events that must be considered in the design of the component. In this larger sense, the histogram can depict more than a single loading event; it may contain a set of operational cycles, giving the various loadings, the number of occurrences of each, the order in which they are anticipated, and the expected time interval between events. The term load histogram and expected loading history are often used synon-omously with histograms.
9.4.1 Event Order and Time Duration Effects When material behavior is nonlinear (such as creep at elevated temperature), the order and time duration of the loads applied to a structure influence the total deformation and length of life of the structure. Different order and time duration of loads will yield different strain ranges for loading, creep deformation, and stress distributions for stress-rupture evaluation. The rate at which an elevated temperature structure recovers and readjusts its stress distribution following a transient load depends on the structure configuration, level and distribution of the primary stresses, and the creep properties of the material. Low creep strength materials usually relax the high residual stresses in a short time period so that the Influence of the high residual stresses does not have a major effect on the component stress-rupture life, although the relaxation of the residual stress may be at the expense of some additional creep deformation. If the transient load leaves a residual stress distribution, these stresses will be added algebraically to the stress field of the next applied load and may thus increase or decrease the total stress. The total stresses, in turn, Influence the creep strain rates.and damage from stress-rupture for this applied load, especially during the early period of this time interval. The order of the transients will have its greatest influence when the time between the events is not sufficient to establish a stable stress distribu-70
tion in the part. Further, the order of transients will also influence the magnitude of the calcu-lated stresses and strains In a structure which is made of a material of high creep strength and subjected to small primary stresses and high secondary (including thermal) stresses. Such struc-tures have an ability to retain residual stresses.
9.4.2 Design Specification Considerations The Design Specifications require a histogram of the Normal and Upset operating conditions for the design life of the plant. Experience in the design of elevated temperature equipmenthasshown that an even spacing of the various transient events throughout the plant life is not a realistic histogram. As an illustration, consider the normal startup and shutdown events. At the begin-ning of operations there will be a number of these events at rather close spacing because of problems during the shakedown period. At a base loaded plant, there will be long intervals be-tween these events. As the plant becomes older, it will be used only during the normal work week and placed on standby operation for the weekend. In the future this may be three days or more. Also as the plant becomes older, the unplanned shutdown becomes more frequent be-cause of aging equipment. Finally the plant will be used only for peaking service because of its high cost of operation when compared with newer equipment in the system.
Electrical load patterns differ from one utility to another, and this part of the histogram is con-trolled by local conditions. The reaction of the various parameters such as temperature, pressure, and time to various transients depends on plant design and modes of operation. Thus, both the Owner of the plant, or his agent, and the system Designer have input to the histogram, much of which must be forecast with some degree of uncertainty.
The influence of the loading histogram may be factored into the design of the structural compo-nents in several ways. One procedure which would entail excessive cost would be to analyze each period of operation as it occurs on the histogram. Another procedure would be to find a recurring sequence of events and do the analysis on this section of the histogram and then consider the total damage as a multiple of the recurring events. However, there are several different stages in the plant life which depend on plant age and the ever changing social and economical pattern of the region in which the plant is located. At each stage in plant life, certain events will occur in approximate cyclic behavior in order that damage could be deter-mined for each stage and then summed for total damage.
Another less detailed procedure would analyze each cyclic transient from steady-state operating conditions where each cyclic event would cover a time span that included the transient and the time required to bring the plant back to Normal operating conditions with a stable stress dis-tribution. As an example, if a transient resulted in a plant shutdown, the total cyclic event would include the Upset transient, the shutdown, the startup, and sufficient time at Normal operating conditions to establish a stable stress distribution in the part being analyzed. Some modes of plant operation, for instance working in two shifts, do not allow sufficient time for a stable stress distribution to be established before the cyclic events are repeated. In this case two or more events should be analyzed back-to-back to establish a stable cyclic behavior. This procedure would handle Emergency events which are unpredictable in a similar manner. This method considers the influence of the residual stress system from the transient but neglects the order of the transients. However, the fatigue damage could be based on the maximum strain ranges over all transients regardless of order of events. Moreover, it is possible to order the events in a manner which will influence the magnitude of the calculated stresses and strains to some extent, especially if events occur at a frequency which does not allow a stable stress dis-tribution to be formed.
For all cases a certain minimum factor of safety is required because of the inability to forecast the actual operating history of the plant, the variations in material properties, the limits of mathematically modeling the structural behavior of the component, and the lack of a theoretical basis for evaluating the combined damage from creep and fatigue. Different procedures for 71
evaluating creep and fatigue damage may require different factors of safety because of the amount of details involved in the analysis. In many cases a high degree of precision in one facet of the analysis is not justified when other phases of problems are not amenable to a similar determination.
In conclusion, the Design Specifications require a histogram which must be factored into the stress analysis of the particular component. Varied procedures are available to estimate the creep and fatigue damage from the periodic operating history of the plant. There is no typical histogram for all the components, or all components of the system. Also, histograms will-vary for different types of systems and are dependent on the regional social and economical climate of the plant location because of the wide variation in electrical load demand. A realistic histogram requires input by the Owner to describe load demand throughout plant life and by the system Designer to describe Upset transients and maintenance replacement periods.
72
I.
~~10.0 REFERENCES~~
~~1. Code Case 1331.4, -5,-6, 7, *8,ASME Boiler and Pressure Vessel Code, Case Interpreta-tions, 1971 Edition, The American Society of Mechanical Engineers, New York.~~
~~2. Code Case 1592, ASME Boiler and Pressure Vessel Code, Code Cases, 1974 Edition, The American Society of Mechanical Engineers, New York.~~
~~3. Criteria of the ASME Boiler and Pressure Vessel Code for Design by Analysis, The~~
~~--American Society of Mechanical Engineers, New York, 1969.~~
~~4. The Generation of Isochronous Stress-Strain Curves, The American Society of Mechanical Engineers, New York, 1972.~~
~~5. R.W.Weeks, Mechanical Properties Test Data for Structural Materials Quarterly Progress Reports, October 1973, ORNL-4936, Oak Ridge National Laboratory.~~
6. G. V. Smith, W.B. Seens, H. S.Link, and R.R. Molenock, "Microstructural Instability of Steels for Elevated Temperature Service," Proceedings of The American Society for TestingandMaterlals, 51: 895-917 (1951).
~~7. A. J. Lovell, L. D.Blackburn, and J.J.Holmes, Quarterly Progress Report, Irradiation Effects Progress Report, Hanford Engineering Development Laboratory, WHAN-FR-40-1, (January 1970).~~
~~8. D.Fahr, Fuels and Materials Development Program Quarterly Progress Report, ORNL-TM-3703, Oak Ridge National Laboratory, (December 1971).~~
~~9. M.C.Murphy and G. D. Branch, "Metallurgical Changes in 2.25 Cr-Mo Steels During Creep-Rupture Test"Journol of Iron Steel Institute, 209 (Pt. 7): 546 (July 1971).~~
~~10. K. Natesan, T. F.Kassner, and Che-Yu Li, "Effect of Sodium on Mechanical Properties and Friction-Wear Behavior of LMFBR Materials," Reactor Technol. 15(4): 242 (Winter 1972-1973).~~
~~11. C R.Brinkman, G. E. Korth, and R. R.Hobbins, "Estimates of Creep-Fatigue Interaction in Irradiated and Unirradiated Austenitic Stainless Steels," Nuclear Technology 16(1): 297 (October 1972).~~
~~12. C. R.Brinkman, Mechanical Properties Test Data for Structural Materials Quarterly Progress Report April 30, 1974, ORNL-4963, Oak Ridge National Laboratory.~~
~~13. C.R. Brinkman and G. E. Korth, "Heat-to-Heat Variations in the Fatigue and Creep-Fatigue Behavior of AISI Type 304 Stainless Steel at 5930 C," Journal ofNuclearMaterial 48: 293-306 (1973).~~
14. H. E. McCoy and R.D.Waddell, "Mechanical Properties of Several Products from a Single Heat of Type 304 Stainless Steel," Presented at the Winter Annual Meeting of the ASME, Detroit, 1973, published
15. W.E.White and 1.Le May, "Some Effects of Microstructure on the Creep Behavior of AISI Type 316 Stainless Steel," p. 1,Symposium on Elevated Temperature Properties of Austenitic Stainless Steels, The American Society of Mechanical Engineers, New York, 1974.
~~16. H. E. McCoy, "Tensile and Creep Properties of Several Heats of Type 304 Stainless Steel,"~~
~~ORNL-TM4709, Oak Ridge National Laboratory, November 1974.~~
17. G.V. Smith, "An Evaluation of the Yield, Tensile, Creep, and Rupture Strengths of Wrought 304, 316, 321, and 347 Stainless Steels at Elevated Temperatures," ASTM Data Series DSSS2 American Society for Testing and Materials, Philadelphia, February 1969.
73
18. W.. F. Simmons and J. A. Van Echo, "The Elevated Temperature Properties of Stainless Steels," ASTM Data Series Publication DSSS1, American Society for Testing and Materials, Philadelphia, December 1965.
~~19. W. E. Leyda and 1. P. Rowe, "A Study of the Time for Departure from Secondary Creep for Eighteen Steels," ASM Technical Report No. P9-1 01, American Society for Metals, Metals Park, Ohio, 1969.~~
20. L. D. Blackburn, "Isochronous Stress-Strain.Curves for Austenitic Stainless Steels," in The Generationof Isochronous Stress-StrainCurves, p. 15-48, The American Society of Mechanical Engineers, New York, 1972.
~~21. Liquid Metal Fast BreederReactor MaterialsHandrbook, Hanford Engineering Develop-ment Laboratory, HEDL-TME 71-32.~~
~~22. J. Bree, "Elastic-Plastic Behavior of Thin Tubes Subjected to Internal Pressure and Inter-mittent High-Heat Fluxes with Application to Fast Nuclear-Reactor Fuel Elements,"~~
~~Journalof Strain Analysis, Vol. 2, No. 3, 1967.~~
~~23. J. Bree, "Incremental Growth Due to Creep and Plastic Yielding of Thin Tubes Sub-jected to Internal Pressure and Cyclic Thermal Stresses,"Journalof Strain Analysis, Vol. 3, No. 2, 1968.~~
24. M. T. jakub and R. A. Moen, "Translating Elevated Temperature Material Properties into Rules for Structural Design," Proceedingsof the First InternationalConference on Structural Mechanics In Reactor Technology (held in Berlin, Germany, September 1971)
~~T. A. Jaeger, Luxembourg, 1971.~~
25. C. H. A. Townley and J. F. Poynor, "Creep and the Combined Effects of Creep and Fatigue," Chapter 3 of Pressure Vessel Engineering Technology, Edited by R. W. Nichol, Elsevier Publishing Company, Ltd., Amsterdam, 1971.
~~26. R. K. Penny and D. L. Marriotts, Design for Creep, McGrawHill, New York 1971.~~
~~27. B. F. Langer, "Design of Vessels Involving Fatigue," Chapter 2 of Pressure Vessel Engineering Technology, Edited by R. W. Nichols, Elsevier Publishing Company, Ltd.,~~
~~Amsterdam, 1971.~~
28. W. J.O'Donnell and 1. Porowski, "Upper Bounds for Accumulated Strains Due to Creep Ratchetting," Welding Research Council Bulletin No. 195,-June 1974. Also Presented at the ASME Pressure Vessel, Piping and Materials Conference in Miami, June 1974, to be published in the ASME Transactions.
30.* J. M. Corum, and W. K. Sartory, "Elastic-Plastic-Creep Analysis of Thermal Ratchetting in Straight Pipe and Comparisons with Test Results," ASME Paper No. 73-WA/PVP-4.
Contributed by the Pressure Vessels and Piping Division of ASME for presentation at the Winter Annual Meeting, Detroit, November 1973.
31. T. W. Pickel, G. T. Yahr, W. K. Sartory, and J. M. Corum, Studies of Shakedown and Ratchetting of Structures, pp. 161-171 in "High-Temperature Structural Design Methods for LMFBR Components Quarterly Progress Report for Period Ending June 30, 1973,"
~~USAEC Report ORNL-TM-4356, Oak Ridge National Laboratory, December 1973.~~
32. C. D. Walker, "Strain Fatigue Properties of Some Steels at 9500 F (51 0C) -With a Hold in the Tension Part of the Cycle,"Paper No. 24, Joint InternationalConference on Creep, Institution of Mechanical Engineers, London, 1963.
~~Reference 29 omitted.~~
74
I
~~33. H. G. Edmunds, and D. J. White, "Observations of the Effect of Creep Relaxation on High Strain Fatigue;" journal of Mechanical Engineering Science, Vol. 8, No. 3, 1966, pp. 310-321.~~
34. R. A. T. Dawson, W. J. Elder, G. S. Hill, and A. T. Price. "High Strain Fatigue of Austenitic Steels," Thermal and High Strain Fatigue, Proceedings of International Conference organized by Institute of Metals In association with the Iron and Steel Institute, Church House, London, June, 1967, p. 239.
~~35. E. L. Robinson, "Effects of Temperature Variations on the Long-Term Rupture Strength of Steels," Transactions of The American Society of Mechanical Engineers, Vol. 74, 1952, pp. 777-780.~~
~~36. S. Taira, "Lifetime of Structures Subjected to Varying Load and Temperature," Creep In Structures, J. Nicholas Hoff, ed., Academic Press, New York, 1962, pp.96-124.~~
~~37. D. A. Spera, "The Calculation of Elevated Temperature Cyclic Life Considering Low-Cycle Fatigue and Creep," NASA TND-5317, NASA Lewis Research Center, Cleveland, Ohio.~~
38. S. S. Manson, G. R. Halford, and D. A. Spera, "The Role of Creep In High Temperature Low Cycle Fatigue," pp. 229-249. Ch. 12 in Advances In Creep Design, A. 1.Smith and A. M. Nicolson, eds., Applied Science Publications Ltd., London, 1971.
~~39. R. D. Campbell, "Creep/Fatigue Interaction Correlation for 304 Stainless Steel Sub-jected to Strain-Controlled Cycling with Hold Times at Peak Strain," ASME Paper No. 71-PVP-6.~~
40. B. M. Wundt, "Cumulative Damage Due to Combined Creep and Low Cycle Fatigue of 1%Cr.1% Mo-'4% V Steel at 1000 F," Consulting Memorandum No. 9, The Metals Property Council, New York, November 9, 1970.
~~41. R. Lagneberg, and R. Altermo, "The Effect of Combined Low Cycle Fatigue and Creep on the Life of Austenitic Stainless Steels,"Met. Transactions Vol. 2, 1971, p. 1821.~~
~~42. E. P. Esztergar, and J. R. Ellis, "Considerations of Creep-Fatigue Interaction in Design Analysis," Design for Elevated Temperature Environment, American Society of Mechanical Engineers, 1972.~~
43. S. S. Manson, G. R. Halford, and M. H. Hirshberg, "Creep Fatigue Analysis by Strain Range Partitioning," Design for Elevated Temperature Environment, American Society of Mechanical Engineers,1972, pp.12-28 (NASA TMX-67838,1971).
44. J. F. Polhemus, C. E. Spaeth, and W. H. Vogel, "A Ductility Exhaustion Model for Prediction of Thermal Fatigue and Creep Interaction," Fatigue at Elevated Temperatures, ASTM STP 520, American Society for Testing and Materials, 1973, pp. 625-636.
45. E. P.Esztergar and 1. R. Ellis, "Cumulative Damage Concepts in Creep-Fatigue Life Prediction," Central Electricity Generating Board, International Conference on Thermal Stresses and Thermal Fatigue, Berkeley, Glos., England, September, 1969, Paper No. 28.
~~46. M. M. Abo El Ata and lain Finnie, "A Study of Creep Damage Rules," ASME Paper 71-WA/Met-1.~~
47. M. Kitagawa and R. W.Weeks, "Analysis of Hold Time Fatigue Test Results of A1Sl 304 Stainless Steel by Five Existing Methods," Submitted to the 1974 Symposium on Mechanical Behavior of Materials, Kyoto, Japan, August 1974.
48. J. B. Conway, "An Analysis of the Relaxation Behavior of AISI 304 and 316 Stainless Steel at Elevated Temperatures," GEMP-730, General Electric Nuclear Systems Programs, Cincinnati, Ohio, December 1969.
75
49. J. T. Berling and J. B. Conway, "Effect of Hold Time on the Low Cycle Fatigue Resistance of 304 Stainless Steel at 1200'F," Part II, Materials and Fabrication, FirstInternational Conference on Pressure Vessel Technology (Delft, Netherlands, September 1969),
~~published by The American Society of Mechanical Engineers, 1970.~~
50. J. T. Berling, and T. Slot, "Effects of Temperature and Strain Rate on Low Cycle Fatigue Resistance of AISI 304, 316 and 348.Stainless Steels" ASTM STP 459, Fatigue at High Temieratures, pp. 3-30. (Also available as General Electric Report GEMP-642, General Electric Nuclear Systems Programs, Cincinnati, Ohio, June 1968).
~~51. J. B. Conway, "Evaluation of Plastic Fatigue Properties of Heat-Resistant Alloys,"~~
~~GEMP-740, General Electric Nuclear Systems Progrims, Cincinnati, Ohio, December 1969.~~
52. G. R. Halford, M. H. Hirshberg, and S. S. Manson, "Temperature Effects on the Strain Range Partitioning Approach for Creep/Fatigue Analysis," ASTM STP 520,1973, pp. 658-667, (NASA TMX-68023, 1972).
~~53. E. Krempl, and C. D. Walker, "Effect of Creep-Rupture Ductility and Hold Time on the 10000F Strain-Range Behavior of a 1 Cr - 1 Mo - 0.25V Steel," ASTM STP 459, pp. 75-98.~~
54. J. M. Corum, "Examination of K-Factor Used In Code Case 1331 and Recommended Modifications," pp. 172-179, High Temperature Structural Design Methods for LMFBR Components Quarterly Progress Report For Period Ending June 30, 1973, Oak Ridge National Laboratory Report, ORNL-TM-4356.
~~55. R. W. Weeks, D. R. Diercks, and C. F. Cheng, "ANL Low-Cycle Fatigue Studies-Program, Results and Analysis," U. S. AEC Report ANL-8009, March 1973.~~
~~56. C. R. Brinkman, G. E. Korth, and R. R. Hobbins, "Estimates of Creep-Fatigue Interaction in Irradiated and Unirradiated Austenitic Stainless Steels," Nuclear Technology, Vol. 16, 1972.~~
~~57. C. E. Jaske, H. Mindlin, and J. S. Perrin, "Combined Low-Cycle Fatigue and Stress Relaxation of Alloy 800 and Type 304 Stainless Steel at Elevated Temperatures,"~~
~~Fatigue at Elevated Temperatures, ASTM STP 520, American Society for Testing and Materials, 1973, pp. 365-376.~~
~~58. J. G. Y. Chow, J. R. Hare, R. H. Jones, and Tremel, "Quarterly Progress Report on High Cycle Fatigue of LMFBR Materials," BNL 50363, October 1972.~~
~~59. C. E. Jaske, H. Mindlin, and J. S. Perrin, "Low-Cycle-Fatigue Evaluation of Reactor Materials," in BMI-1914, July 1971, pp. A2-A44, Battelle Memorial Institute report.~~
60. L. F. Coffin, Jr., "Predictive Parameters and Their Application to High Temperature Low Cycle Fatigue,"SecondInternatlonalConference on Fracture (Brighton, England, 1969) published by Chapman and Hall, London, and Barnes & Noble, New York, 1969.
61. S. Y. Zamrik, Study of the Effects of Biaxality in Creep-Fatigue at Elevated Temperatures, pp. 177-189 in "High Temperature Structural Design Methods for LMFBR Components Quarterly Progress Report for Period Ending September 30, 1972," USAEC Report ORNL-TM-4058, Oak Ridge National Laboratory, December 1973.
~~62. S. S. Manson, Thermal Stresses and Low Cycle Fatigue, pp. 193-273, McGraw-Hill, 1966.~~
~~63. C. R. Brinkman, G. E. Korth, and J. M. Beeston, "Comparison of the Fatigue and Creep-Fatigue Properties of Unirradiated Type 304 and 316 Stainless Steel at 593 0 C (11 00 0 F),"~~
~~ANCR-1078, August 1972.~~
~~64. C. E. Jaske, H. Mindlin, and 1. S. Perrin, "Low Cycle Fatigue and Creep Fatigue of Incoloy 800," BMI-1921, February 1972.~~
76
I .
~~65. C. E. Brinkman, G. E. Korth, and J. M. Beeston, "InFluence of Irradiation on Creep-Fatigue Behavior of Several Austenitic Stainless Steels and Incoloy 800 at 700"C,"~~
~~Effects of Radiation on Substructuresand Mechanical Propertiesof Metals andAlloys, ASTM STP 529, 1973.~~
~~66. E. Krempl, "The Effect of Stress Concentration on the Low-Cycle Fatigue of Three Low-Strength Structural Steels at Room Temperature and 5500 F," GEAP-101 70, March 1970.~~
~~67. E. Z. Stowell, "Stress and Strain Concentration at a Circular Hole in an Infinite Plate,"~~
~~NACA-TN-2073, April 1950.~~
68. H. Neuber, "Theory of Stress Concentration for Shear-Strained Prismatical Bodies with Arbitrary Nonlinear Stress-Strain Law," Transactions of The American Society of Mechanical Engineers, Series E, Journalof Applied Mechanics, pp. 544-550.
69. D. F. Mowbray and McConnelee, "Application of Finite Element Elastic-Plastic Stress Analysis to Notched Fatigue Specimen Behavior," First InternationalConference on StructuralMechanics in Reactor Technology (West Berlin, Germany, September 1971),
~~published by T. A. Jaeger, Luxembourg.~~
70. T. H. Topper and C. V. Byre Gowda, "Local Stress-Strain Approach to Fatigue Analysis and Design," ASME Paper 70-DE-24. Presented at the Oesign EngineeringConference andShow, Chicago, Ill., May 11-14, 1970.
~~71. Ernest L. Robinson, "Steam-Piping Design to Minimize Creep Concentrations," ASME Transactions, 1955, pp. 1147-1162.~~
~~72. J.Spence, "An Analysis for Pipework Systems Under Creep Conditions," Proceedings of the First International Conference on Pressure Vessel Technology, Part 1 - Design and Analysis, ASME, 1969.~~
~~73. W. Gorczynski, "A Method of Assessing Structural Efficiency of Pipework Designed for High Temperature Service," Proc. Instn. Mechanical Engineers, 1963-1964, Vol. 178.~~
77
t.
I APPENDIX A PROCEDURES EMPLOYED TO ESTABLISH THE BASIC TIMEAN'b TEMPERATURE STRESS INTENSITIES AND ISOCHRONOUS STRESS-STRAIN CURVES' A.1 MAXIMUM ALLOWABLE VALUE OF GENERAL PRIMARY' MEMBRANE STRESS INTENSITY FOR DESIGN CONDITIONS The symbol So isused for this value. The S0 values are identical to the values for S given in Section VIII, Division 1 of the Code.
A.2 TIME-INDEPENDENT DESIGN STRESS INTENSITY The symbol Sm isused for this value. These values are based on tensile and yield strengths of the material. The-,criteria employed are defined as follows:
(a) The allowable stress intenslty.value, Sm, for ferritic steels and nonferrous metals and alloys, except those covered in paragraph (b) below, is the least of the following four varues:
(1) one-third of the specified minimum tensile strength at room temper-ature (2) one-third of the short-term tensile strength at temperature (as de-fined in A.2.1 in this Appendix).
The ratio of UTSTIUTSR J, for austenitic stainless steel, is found in ASTM DS5-S2.
(3) two-thirds of the specified minimum yield strength at room temper-ature (4) two-thirds of the short-term 0.2% offset yield strength at temperature (defined'as the ratio of yield strength at temperature divided by the yield strength at room temperature, multiplied by the minimum specified room temperature yield strength).
The ratio of YS1IYSRT, for austenitic stainless steel, is found in ASTM DS-S2.
The strength isdetermined at a strain rate of 0.0005 min.'
(b) The allowable stress intensity value, Sm, for austenitic steels, nickel-chromium-iron and nickel-iron-chromium alloys is the lowest of the following four values:
(1) one-third of the specified minimum tensile strength at room temper-ature (2) one-third of the tensile strength at the operating temperature (3) two-thirds of the specified minimum yield strength at room temperature (4) 90% of the yield strength at the operating temperature A.2.1 YIELD AND TENSILE.STRENGTH The procedures for establishingyield and ultimate strengths are discussed below:
The yield and tensile strength data available for a particular material grade are normal-ized by ratioing the elevated temperature strength of individual lots to the room temperature strength of the same lots, and then all sets of such ratios representing a 1Not applicable to bolting materials.
79
lo 0 particular grade are evaluated by the method of least squares to establish the curve of best fit for the data. The resulting strength-ratio trend curve isconsidered to represent the typical or charakcteristic variation of yield or tensile strength with temperature.
1 Using such a ratio trend curve, it becomes possible-to compute strength trend curves f6ir anyrpartiita'rroom temperature strength level of interest within the limits encom- i passed by the original data.
Since the design stress intensity criteria include fractions of the specified yield and tensile strengths, it is necessary to factor the ratio trend curves against the specified minimum yield and tensile strengths to define what may be termed minimum position yield ana tensile strength curves. At temperatures above room temperature, the property yieldstrength at temperature is taken, for purposes of the criteria, to be this minimum position value. However, the property tensile strength at temperature is taken as the smaller of:
(a) specified minimum tensile strength at room temperature; or (b) a value 10% greater than the minimum position value cited above.
A.3 TIME-DEPENDENT DESIGN STRESS INTENSITY 2 The symbol St is used for the basic time and temperature dependent allowable stress Intensity for load-controlled stresses. SI values are the least of each of the three quantities:
(a) two-thirds of the minimum stress to cause rupture in time t; (b) 80% of the minimum stress to cause theonset of tertiary creep in time t; and (c) the minimum stress to produce one percent total strain in time t.
A.3.1 MINIMUM STRESS-TO-CAUSE-RUPTURE IN TIME t The basic data and a description of the rupture strength evaluation procedures employed for several of the materials included in Case 1592 (i.e., Types 304 and 316 Stainless Steel) may be found in Metal Property Council data evaluations [Smith (17) and Simmons and Van Echo (18) 1.
Two principal procedures arc employed in evaluating the dependence of the stress-to-rupture upon time and temperature. A choice between the results, based on engineer-ing judgment, Isthen made.
In the first procedure, the isothermal relation between stress and time for rupture of individual lots is interpolated or extrapolated, as required, to identify the stress-to-cause-rupture in 100, 1000, 10,000, etc., hr. Plotting on log-log coordinates tends to linearize the variation, and thereby facilitate extrapolation, particularly at lower temperatures. At higher temperatures, the variation tends to curvilinearity at longer times, and extrapolation involves greater risk. The results of such interpolations or extrapolations for individual lots, as they vary with temperature, are then evaluated by the method of least squares to define an average rupture strength-temperature trend curve. A trend curve for minimum rupture strength is derived from the mean trend curve by subtracting 1.65 multiples of the standard -deviation of the sample. If certain implicit assumptions hold (e.g., that the data are normally distributed, or that the average is without error), this minimum trerid curve defines a lower boundary for 95%
of the data. Inspection of the data plots has always shown that this is approximately true.
XBased on data determined from tests performed on air.
80
The second evaluation procedure that has been employed is an Indirect one, involving one or another time-temperature parameter. One of the simplest of these;.widelyiused, is Larson and Miller's P =T(C + log t) F(S) where T is the temperature in degrees Rankin, t is the time for rupture Kn:hours, Cmis.a material constant, and f (S) denotes that the parameter depends upon stress. Ordinar-ily, the available data are not adequate to permit evaluation of individual lots by the parameter procedure, as ,Yould be desirable; instead, the total data population must be treated as a common population on a universalized basis, assuming the constant C to remain Invariant from lot to lot within the data population. Accordingly, the scatter band for the stress-parameter variation is evaluated by the method of least squares to determine an average curve of best fit, from which average values of rupture strength as dependent upon temperature can be derived. A minimum curve is developed, as in the previously described procedure, by subtracting 1.65 multiples of the standard deviation of the sample from the average curve.
A.3.2 MINIMUM STRESS-TO-INITIATE-TERTIARY-CREEP IN TIME t For the ferrous alloys included in Case 1592, Leyda and Rowe (19) had reported that the time to initiate tertiary creep is a fixed fraction of the time for rupture for a given material at a specific temperature. Thus, for 2Y4 Cr-1 Mo and Types 304 and 316, the minimum stress-to-initiate-tertiary-creep was computed from the minimum-stress-to-rupture data. For the high-nickel alloys, tertiary creep information was developed by examination of available creep curves. In all instances, the initiation of tertiary creep was assessed visually. Extrapolations to lower temperatures were made by time-temperature parameter methods exactly analogous to those described for rupture strength, except that the time for tertiary creep to begin became the measured quantity.
A.3.3 MINIMUM STRESS TO CAUSE 1%TOTAL STRAIN IN TIME t In common with the criterion for stress to initiate tertiary creep, the criterion on minimum stress for 1 %total strain is new to the ASME Code. For purposes of Sec-tion I and Section VIII, Division 1, creep strength has for many years been evaluated in terms of the stress causing a secondary creep rate of 0.01% per 1000 hr. In the new criterion, stress to cause 1%total strain can be evaluated in an exactly analogous manner as the stress to cause rupture or to initiate tertiary creep-the dependence upon stress tends to parallel that of rupture time-or it can be derived directly from the isochronous stress-strain curves. Both approaches have been employed in developing the material for Case 1592.
A.4 ISOCHRONOUS STRESS-STRAIN CURVES For relatively short times, isochronous stress-strain curves may be derived by taking constant time sections through a family of creep curves, such as those of Fig. 2. However, this approach to their derivation is not generally practicable (unless forming only a part of a multi-approach procedure) since the time scale of interest extends beyond the feasible limit of experimental testing. Extrapolative procedures are required for the longer test times and for strains below 1 %.
For generating the isochronous stress-strain curves in Case 1592, two procedures have been employed. In one of these, the curves are developed by performing evaluations exactly analo-gous to those described for developing information on the stress to reduce 1%strain, except that other specific strains encompassing the range of interest are also evaluated and the various parts assembled in plots of stress versus strain for different fixed times.
81
I In'a ieCond procedure, the strain hasbeen expressed as an analytical function of time according to the general form:
e =cading +etranslent esteady state where each 'of the parts is a function of stress and temperature..The Individual parameters of the equation are then evaluated from the experimental results generated in conventional creep' tests.
More complete descriptions of these procedures may be found in The Generation.of Isochro-nous Stresg-Strain Curves, ASME, 1972(4).
8 .
' I 82
LIST OF FIGURES Fig. 1(a) Histories from a Loading at Low Temperature ....... ................. 9 Fig. 1(b) Histories from aLoad-ControlledLoading at Elevated Temperature ................ 9 Fig. 1(c) Histories from aStralnzControlledLoading at Elevated Temperature................ 10 Fig. 2 Effect of Stress Level on Elevated Temperature Behavior .................... ...... II Fig. 3 Data from the StandardCreep Test ..................... ................. 12 Fig. 4 Stress-Rupture Data from Cbtstant Temperature Creep Tests .......:
.......... .. 12 Fig. 5 Isochronous Stress-StralnCurves from Creep Tests at Constant Temperature ............ 13 Fig. 6 Effect of Tension Hold Time on the Fatigue L1fe of AISI Type 304 Stainless Steel; .
at 1000 FinAir .14 Fig. 7 Flow Diagram for Elevated Temperature Analysis . . . 16 Fig. 8 Sm: Allowable Stress, Type 316 Stainless Steel ................................ 17 Fig. 9 Elastically Calculated Primary Bending Stress Distribution Compared to Actual Time-Dependent Inelastic Stress Distribution .................. ,.,,.-.-.-.-.-.-.-. 19 Fig. 10 Steady-State Creep Stress DistributlorlsAcross a RectangularBeam in Pure Bending and Having a Steady-State Creep Low of the Form, ec =A an ............................... 20 Fig. 11 Steady-State Creep Stress DistributlonAcross a Thin, CircularTube In Pure Bending and Having a Steody-Stote Creep Low of the Form ~e -Au' . ............................... 21 Fig. 12 Stress Regimes ....... ............................................. 23 Fig. 13(a) Effect of PriorCreep Deformation on Tensile Elongation of Type 304 Stainless Steel ... 25 Fig. 13(b) Varlation ofC/eep Ductility with Rupture Life for Selected Austenitic StainlessSteel Submerged-Arc Weldments ................... ,. . 26 Fig. 14 TypicalStraln-Time Responses ..... ..................... 27 Fig. 15 Strain-Time History Schematic . . .28 Fig. 16 Simple Example of Conditions Leading to Elevated Temperature Ratcheting .... ........ 29
Fig. 17 Effective Creep Stress, oc, for Upper Bounds on Total Accumulated Inelastic Stroan with Creep Ratcheting....................... 33 Fig. 18 Load History for an Example with Ratcheting ......... ........ ............ 34 Fig. 19 Stress History for on Example with RotchetIng ................... .............. 35 83
.a-
-Fg. 21 Effect of Hold-Time on Fatigue Llfe of Austenitic Stalnless Steel ..................... 37 Fig. 22 Effect of Hold-Time on Fatigue Life of Various Steels .......................... . 38 Fig. 23 Creep vs. Fatigue Damage at 5%Load Drop-off, Type 304 Stainless Steel at 1200 F. :42 Fig. 24 Average and Minimum Trend Curves for Creep-Fatigue interoction for Types 304 and 316 StaInless Steels ........................................................... ,.43 Fig. 25 Comparison of Code Creep-Fatigue Allowables to Experimental Dta .44 Fig. 26 Distribution of Actual-to-Calculoted Life Rotlos forStrOak7-Controlled Creep-FOtigue Test Specimens .............................. 45 Fig. 27 Creep-Fatigue Damage Interaction for Tension Hold-Time Tests ofSolution Annealed Ni-Fe-Cr, Alloy 800H ......................................... ; 46 Fig. 28 Strain Rate Effects on Fatigue Life Reduction .47 Fig. 29 Hold-Time Effects on Fatigue Life Reduction for Types 304 and 316 Staonlesi Steel ........ 48 Fig. 30 Comparison of Anolytical and Experimentol Strain Ranges for a Notched Bor .......................................
.. 5O-53 Fig. 31 Sample Problem In Plane Stress ......... -
................. 56 Fig. 32 AnalysIs Flow Diagram-............ ,:.5 Fi. 32 A..y t lw iga ............................... ~1............................. 57 Fig. 33 Typical Buckling Behavior 59 Fig. 34 Creep Buckling Deflection-rime Characteristics 59 Fig. 35 Behavior of a Bolt in an Unyielding Flange 65 Fig. 36" Balanced Loop Behavior - :.....66 Fig. 37 Four Large Bending Loops in Serles with Four Smaller Loops of Half-Size Pipe In Parallel 67 Fig. 38 rwo-BorModel with Strain Concentration butno Elostic Follow-up 68 Flg. 39 Two-Bar Model wlth no Strain Concentrotlon or Elostic Follow-up ..... ............... 68 Fig. 40 Two-Bor Model with Elostic Follow-up and Strain Concentration ..................... 69 Flg. 41 Built-in Cylinder with Elastic Follow-up ................................. 70 84

v t e Letter Sequence Request
TAC:MC7342, (Open)
Initiation Request, Request, Request Acceptance... Administration Questions Answers Results Approval... Other: ML052420712
MONTHYEAR2005-06-19 [Table View]

ML051790349

Text

Dear Sirs:

Subject:

REFERENCE:

7 r 4 JET 91"ZI Eq Lit, tz T... --= U

SUMMARY

SUMMARY

SUMMARY

SUMMARY

SUMMARY

SUMMARY

SUMMARY

SUMMARY

SUMMARY

SUMMARY

SUMMARY

SUMMARY

SUMMARY

SUMMARY

SUMMARY

1.0 INTRODUCTION

SUMMARY

4.0 REFERENCES

1.0 INTRODUCTION

4.0 REFERENCES

SUMMARY

SUMMARY

1.0 INTRODUCTION

10.0 REFERENCES

Navigation menu

Search