UJNIVERI!TY OF >: ICHTIGAN,AitN2 ARBOR Preliminary Report on EFFECT OF METHOD OF LOADING o n THES R:S3UJTING CREEP CIKARACT;RIS fICS of PLAIIN CARBON STt~LIJ AT 850F. by:L. Clark A. E. hite Project XIumnber 491-48 Report Number 2 for The Detroit Edison Company July, 1932

j "II "I, \s I

EFFECT OF 4ETHOD OF LOADIN71G on THEI RESULTIING CREzP CHIARACTERISTICS of PLAIN CALRBON STEILS AT 8500F. In the present- report the results are given of an investigation to determine the effect of testing procedure on the observed resulting creep characteristics. As the usual creep test is now conducted, there are three possible ways in which the load may be applied. That is, either the single-step, the up-step, or the down-step method:nay be used. In the single-step method, a given fixed load is applied and this is maintained constant for tne entire duration of the test. In the up-step method, the original load is only maintained constant until the creep has either stopped, at least within the sensitivity of the measuring apparatus employed, or until it',as progressed at a constant rate for a given period of time. Then the load is increased and the process repeated. The down-step method differs from the up-step in that the original load is relatively large, and after the rate of creep has been determined, the load is decreased rather than increased. The down-step method of loading may be conducted according to either one. of two procedures. Either the inititlly applied, relatively large load may be maintained eonstant for

2. a sufficient period of time for the rate of creep to be determined, or else it may be manintained constant for only a few hours and the first rate of creep determined only after the load has been reduced. Although various claims have been made as to the relative merits of these three procedures, very little if any F:ork has been done to actually determine the effect of the method of loading on the resulting creep characteristics. Accordingly, the work herein reported was undertaken to throw light on this question. Also, attempts are to be made to determine the influence of the method of loading on the "Time-Yield" value of Dr. Hatfield. In a previous report entitled "Comparative Physical Characteristics of Grades A and B Seamless Steel Pipe at Normal Tevperatures and at 850~F,," it was shown that the values obtained for Hatfield's "Time-Yield" valueS, when the up-step method of loading was used, were considerably below those reported by Dr. Hatfield for sTillar steels. It was later learned that Dr. Hatfield employed the single-step, rather then the up-step method of loading in his tests, and,, accordingly, the results herein reported should show what infl.unce the method of loading has on the "Time-Yield"' values.

The results given in this report are not corm lete as only two steels have been considered, namely, two plain carbon steels containing 0.18 and 0.41 per cent carbon respectively, and only-one temperature has been considered, which was 8S500F. It is entirely possible that the findings obtained for these two steels at this temperature may not show the same relationships as would be obtained for steels of other types at this same temperature, nor is it entirely possible that the same relationships would be found to exist for these two same steels at other temperatures. Results will have to be available from other tests which are now in progress before the findings herein given may be extended. SUL7.rtaRY OF COiJCLU SIOP;S Creep tests undertaken on Grades A and B carbon steel at 8500F. by three different methods of loading., that is, the up-step, the single-step, and the tdown-step methods, show the higher carbon steel (Grade B) to possess the rmaximum creep characteristics in all cases. It was likewise found that the resulting observed creep characteristics varied appreciably depending upon the testing procedure employed. ith both steels the lowest

4. values for a rate of creep of n.01 per cent per 1000 hours (1.0 ner cent per 100,000 hours) were obtained with the upstep method of testinc, wthile the maximrum values were obtained with the dozn-step method. The expl natiion for this is believed to be as follows. In the up-step method sufficient ti-me is not allowed for the mraterial to assume its minimarl rate of flow and the rate reported is, tl-erefore somevzhat high. In the down-step method, the material has underg-one considerable stra in hardening under the relatively l1-rge initial load, and, therefore, assumies its minimut rcate of creep more rapidly w:hen the load is reduced to any 1'iven value. The observed creep characteristics obtained by these three different metthods of loading- are i Pren in Table I, Because of the greater speed with which results can be obtained by the up-step method, and because the results obtained by this method are always oh the safe or lower side, it is felt that this procedure does have a prasctical i,.nportance. It is believed, however, that the.maximum actual creep characteristics can best be obtained by the singele-step method, or by the up-step method in which the time under each load is increased to at least 650 hours, and Dreferably 1000 hours~

Tab le I Effect of Method of Loading on the Observed Creep Characteristics of Grades. i and B3 -teel at 850cF. Stress for Designated Rate of Creep Rate = Per Cent per 1000 Hours MethIod of Loadin 0.01 0.10 1.00 -Grade ih'teel Ij ip, Z'itep 6, 900 10 500 16,000 "i n gle- tep 2,800 12, 900 16,900 Dovwn-"-,teo (1) 11,500 14, 000 16,900 Dorn-:-tei (2) 11,700 13,500 15,830 irade 7B teel:Jp-Step 12, 000 1T, 900 2x, 500 - ingle-'3tep 13,000 18,3O00 27,000 Down-Step (1) 14,500 19, 000 24, 500 Down-,.;-tep (2) 14,500 19,000 24,500 (1) Original load mainttained constant for at least 500 hours. (2) Original load maintained constant for only 24 hours.

6. Likewise, it was found that the -method of testing influenced the total amount of plastic deformation which was obtained under any given load. In the majority of cases, the largest amount was obtained with the single-step method of loading, and this method also required the greater time period before creep would proceed at a given constant rate. If, however, in the other two methods or testing, that is, the up-step and down-step methods, allowance was r.ade for the plastic deformation which had occurred under the preceding stresses, the total amount of plastic deformation in the three cases would be more nearly equal. The effect of the method of loading on the Hatfield "Time-Yield"' value was also considered. Again it was found that in practically every case this value would be lower when the single-step method of loading was used than with the upstep or down-step methods. This is again due to the greater amount of plastic deformation obtained, especially during the early stages of the test, with the single-step method of testi ng,

7 PROCEDURE The steels used in this investigation were the same as those considered in the previous report on the "Comparative Physical Characteristics of Grades A and B Seamless $Steel Pipe at Normal Temperatures and at 8500F." The steels w"ere furnished in the form of seamless steel pipe and their compositions are given in Table II. Table II Chemical Compositions of Grades A and B Seamless Steel Pipe Chemical Composition, Per Cent Designation Carbon nanganese Silicon Sulphur hosrus Grade A 0.180 0.493 0.01 0.017 0.009 Grade B 0.408 0.917 0.21 0.031 0.014 The two steels differ primarily in their c irbon, silicon and manganese contents, with the Grade B material containing the higher amount of these three elements in each case. On the basis of the silicon content, the Grade A rmaterial is of the rimmed or open type, while the Grade B steel is of the killed type. The only tests which rere considered in this investigation were the creep tests. The apparatus used in these

tests has been fully described in the literature.1 An optical system which is sensitive to 2.8 millionths (0.0000023) of an inch is employed for measuring the elongation of the specimen and the tem.peratiare of the specimens is controlled to within ~2~F. Even with this degree of temperature control, however, the measuring system is sufficiently accurate to record the elongation and contraction of the specimen produced by the temperature variations. The elongation readings recorded on the. time-elongation curves given in this report are the average of the elongation and contraction ranges. Three different procedures were used in the creep tests. One consisted of the so-called up-step method of loading. This consists of applying an initial load of rather small magnitude and maintaining this load constant until flow has either stopped, at least within the sensitivity of the measuring apparatus employed, or until it has proceeded at a definite fixed rate for a given period of time, this time period gen rally being about 200 hours. Then the load is increased and the process repeated. The second method used was the down-step method, and this differs from the up-step only in that the initial..-'..hite, C.L.Clark, and L.Thomassen, "An Apparatus for the Determination of Creep at Eilevated Terperatures," A. S2c-P4. i, Fuels and Steamn Power, Vol. 52, No. 27, n. 347, 1930.

9. load is relatively large, and the load is decreased rather than increased, after the creep characteristics have been determined for any given stress. Two procedures were used in this method of testing. In one case the initially applied load was maintained constant for a sufficient period of time to allow the rate of creep produced by this load to be determined. In the second case the initially applied load was held constant for only 24 hours and then decreased. The first rate of creep obtained by this second method was, therefore, that which occurred after the first reduction of stress. The third, and last, method was that known as the single-step method. This differs fro-,m the preceding two in that a dfferent specimen is used for each of the loads. RESULTS' The creep results obtained by the different methods of loading on the two carbon steels at 8500F. are given in Figures 1 to 10 inclusive and in Table III through VI. Cr eeT ests on Grade A Steel As stated previously the three types of load application which were employed in the creep test on this material

10. were the up-step, the single-step and the down-step method. The down-step method was conducted according to two procedures. In one case the original load was maintained constant for 600 hours before it was decreased, while in the other it was maintained constant for only 24 hours. The results obtained on the Grade A material are given in Figures 1 through 5 and in Tables III and IV. Figures 1 through 4 give the time elongation curves which were obtained, while in Figure 5, the stress and the corresponding rate of creep are plotted on logarithmic coordinates. Table III summarizes the results which were obtained. It is seen that four stresses were employed in the up-ste.) method of loadirg, five in the single-step method, and three in each of the down-step methods. The stresses emn loyed ranged from 7,500 to 16,795 pounds per square inch, and the time of test employed for each load varied fron 340 to 890 hours. The results obtained on this material indicate that there is a considerable difference in the observed creep characteristics depending upon the method of loading which is employed. It should be emphasized that since a given heat of steel should possess a definite set of creep characteristics

Table III Deformations Obtained Durin- Creep Tests at F500F. on Grade A Steel Plastic Deformation -Total Time During Decreasing Total Plastic Rate of Method of Elastic of rest Stage of Creep Deforrnation Creep Loading Stress Deformation Hours In./In. Time, Hrs. I/100 Hrs. Up-Step 7,500 -0.000140 340 0.000150 100 0.000240. 0016 Up-Step 9,225 0.000115 475 0.000220 70 0.000430 0.054 Up-Step 11,700 0.000180 675 0.001520 500 0.001?00 0.160 Up-Step 14,250 0.000330 429 0.003519 225 0.304594 0.525 Single-Step 7,500 0.000140 340 0.000150 100. 0.000240 0.016 Single-Step 9,225 0.000565 605 0.000435 375 0.300485 0.006 Single-Step 11,700 0.000827 890 0.001403 425 0.301693 0.362 Single-Step 14,250 0.001620 865 0.003380 575 0.005960 0.213 Single-Step 16, 795 0.002320 600 0.006890 200 0.0109.20 1.05 Down-Step(1) 16,795 0.002320 600 0.006880 200 0.010920 1.05 Down-Step(1) 14,250 -0.000088 340 0.000000 0 0.000345. 119 Down-Ctep(l1) 11,700 -0.000080 380 0.003000 0 0.000075 0.016 Down —tep (2) 14,250 -0.000100 600 0.001110 225 0.002040 3.256 Down-Step(2) 12,525 -0.000065 466 -0.000010 30 0.300180 0.037 Down-Step(2) 11,700 -0.000025 385 -0.000010 25 0.000070 0.009 (1) Original stress of 16,795 pounds maintained constant for 600 hours. (2) Original stress of 16,795 pounds maintained constant for 24 hours,

at any given temperature that the differences obtained are not actual but only observed differences and are entirely due to the different creep testing procedures which were employed. It is also felt that many of the observed differences may be accounted for either on the basis of the time factor or on the fact that in certain of the testing procedures used the material was previously placel in a strained condition. The effect of the time factor is well shown by the results which were obtained from the two lowest stresses employed in the single-step method of loading. The test under a strpas of 7500 pounds per square inch was continued for 340 hours and the rate of creep during this timr.e was 0.016 per cent per thousand hours. with the stress at 9,225 pounds per square inch, the test was continued for 605 hours. and the rate of creep obtained was 0.006 per cent per thousand hours. It is impossible for a higher stress to produce a lower rate of creep than does a lower stress and the difference obtained is, without doubt, due to the variation in time.In other words, it is believed that if the test under the lower stress had been continued for a period of 600 hours or longer, the rate of creep would have continued to decrease

13. and would have become less than 0.006 per cent per thousand hours. Since, however, 340 hours is a tine period commonly used in the up-step method of testing, it is believed that the value of 0.016 per cent per thousand hours thus obtained should be used as representative of results obtainable with this type of loading. From Table III it is evident that the down-step method of loading yields the highest observed creep characteristics, while the up-step method yields the lowest. The magnitude of the relative creep characteristics yielded by the three methods is better shown in Figure 5. In this figure logarithmic coordinates are used and the rate of creep is plotted against the stress which produces it. This method of plotting is used because under many conditions of temperature and stress a straight-line relationship is obtained between stress and rate of creep.'Results taken from this figure are given in Table IV. Figure 5 and Table IV show that the differences in the observed creep characteristics obtained by the different methods of loading are greater at the lower rates of creep than what they are at the larger. For example, the stresses required to produce creep at the rate of.01 per cent per thousand hours ranged from 6,900 pounds for the up-step method

KCUrFtL L, LSSR 60., N. V. 10. 3(191111 L(rKnI-ithmfF, LYJ(, ii L) CScttt~d.;-i - ~~ ~ ( 1 -''`'-t~l~;". t - I r -! ~~~I ~?.r e ~~ -- -J —----— i-i —i.J r I - —-- r -u — i -- — I- — —--- "" -- t —~7c~ — r I r-.t " --- -Cli —r., i i F f I r I,, ~, r I 1 t I t I' t r I I i,: t 17 I t I I i(;"'' i.,, ; i: -. i i' T c; -$- r? —t —-.-TI —' --— ~ --— \ ----— ,, 1i t_~.l; I " r? ~ —-* f " D i. t ~I? lit:: t ... L i -~11 r 1.c.~ —, il.-1. 1 ~- 1.~ ~ ~ r t 1' :~" ~) ~z ~~~ I 1:- i;~ i 1:: t - t ~ 1'1:, i fl-~~- lr .." ~~ -f * ~~~ (~-~ r.. I .., I I 1, I. I L ~ i 1 -~ - f ~ ~ ~ I ~ ~ I ~ I ~ ~ 1~. I~', + iC;t~ 1~- ~- ~ - i —1 ^4: I. c;-.n:J - r-: —~ -- ~f I~t 1 I + -i — ~ C~~ c i i.,;. - ~.'i f"' -r ~~ 1 a I r -~r i: i "";, -~t — i f -t — —~ i ~~~ ~I.; ..' i'1't i i ~t~~i r ~~ ~ —— i~~~~: ~~ i f?-.1. 1.., J....... i t. r ~- tr i -r ~ I i ~-~,,~e -~-; ~ -~!~ t ~-~ ~'::~-:: L-'C r i; c.i r ..i: r —:c — I::: i.~: "'~;' 4" ~" i..'_i:-~ ~j_' ——' ~, *~ I I I i r i ~t~: ~:;~ I - —~ —t —cc* — ~~r i t I.ji'';I1;::. ~ -- t —- ~'' -F-3 —~ t- i -c- ~t j 1,~ 4L C _:*;4`q r..i.....j.:.. i i 2,;; i:'- - I ~r+' i' i ~.; ~ -r,t,...~ 7 I rc i i i C r, sl~ 1 1: -!-.- i ~. -t ~r: ~r r —— r -r — 1 r I;~; i i 3 i L',, t,, ~ i i ~~ — $ 6:i;~irL:.~C~.~ r r ~ r i i~ I r. , ~~- ( ~ ~~ - ~--~CC:~- t i i ~~ i~ I ~I~;-I - I ~ it ~.(-.~1 — ~- 1 f ~;l;j_ I) \ ( ~ r' i i' ~1: -i I ~~;(-~I ~: ~ t ~ ~ ~r ~i: f.. c~)-~:-C~- -- ~~~-~` ~ ) -~ 1'' r i- ~ -~.: I. r 1- ~111 r 1. ~1 - -1. — I~ ~ Lr-,,,-:1_ i II:I' 1 I i ~ ~ 1 I i ~ -'1. __1 1. 1., ~ I i j i e9t I~ ( i -4 i *:. i t;.I I i; *:-; I i_ I —t-_ —i f: I I...,`" ~'~ R:~_ ;t-l i h:~ t t I c ~: i ~.. I~- —-4~-~ —~~-~~ i; II::" i:~ t~,i;e1-; t;,-~:t;I:":t ~~' i- I i-:i- t.. ~ J:? r:"i 3i: ~~ - ~.. i ~- i 1 t 1 ~i~ 1 t t ~ t t r~ t -~ -1,~-.. ~-~ J 1,3 -— ~ -II i I -. i t 1~ I 1: - i... iif t":I ~.I ~:"~;-ieieS 8 d.41 P; t: -i ilr e: ~~,", ~1 -I i t j.! ..1..... 1' I I Ii.. i. t .t L r ~: i. L. ~ ~ I 1 I t -~ : L 1 r.r. f OI.ginaL 1Si cT ~ria~f-n 6 rCC ~j;ttj/tt.:11901 i.r.r r,L.:n-lr:i(i-i I ~ i L'~~~ i~;~ f 1 1 i "" Itoa a 2 3 4 5 G 7 8 ABO 2 3 4 5 6 7 8 9 1.00 - 2 3 d a:i

14. Table IV Effect of Mtethod of Loading, on the Observed Creep Characteristics of Grade A Steel at 8500F. Stress for Designated Rate of Creep Rate = Per Cent per 1000 Hours Method of Loadl ng 0.01 0.10 1 00 Up-Step 6,900 10,500 16,000 Single-Step 9,800 12,900 16,900 Down-Step (1l) 11,500 14,000 16,900 Down-Step (2) 11,700 13,500 15,800 (1) Original load of 16, 725 pounds held constant for 600 hours. (2) Original load of 16,725 pounds held constant for 24 hours. to 11,700 pounds for the down step method, However, the stresses required to produce creep at the rate of 1.0 per cent per thousand hours only ranged from 15,800 pounds to 16,900 pounds. The fact that the up-step method of loading yields the lowest observed creep characteristics is believed to be due to the time factor. It is felt that in this method of testing sufficient time is not allowed under any given stress for the metal to assume its minimum creep value. The fact that the down-step method produced maximum values may be explained on the assumption that sufficient strain hardening occurred under the original large load to enable the mnetal

15. to assume its minimum creep rate in a shorter period of time. Certain data in Table III substantiate this claim. For example, it is seen that with the doxwn-step method of loading the time required for the material to assume a constant rate of creep only ranged from 0 to 30 hours. 1?ith the other methods of loading the corresponding time ranged from 70 to 575& hours. It is also to be observed that the down-step method in which the original load was held constant for 600 hours gives a slightly greater resistance to creep after the first reduction of stress, that is, with a load of 14,250 pounds, than does the method whereby the original stress is maintained constant for only 24 hours, while at the lowest stress employed, 11,700 pounds per square inch, the converse is true. The explanation for the first of the above observations is believed to be as follows. rhen the specimen is subjected to the relatively high stress for the longer period of time, a greater resistance to continuous deformation is built up and thus when the stress is somewhat reduced it deforms at a smaller rate. This statement is supported by the shape of the time-elongation curve which shows that when the original stress is only maintained for

16. 24 hours, considerable plastic deformation occurs at a decreasing rate when the stress is first reduced before creep oroceeds at a constant rate. Thlen the original stress is maintained for G00 hours, creep occurred at a constant'rate immediately upon the reduction of stress. The second of the above observations may be explained on the assumption that for the lower rates of creep, the rate of recrystallization assumed greater importance, and that since the specimen which had been held under the original stress for 600 hours has undergone greater deformation, it will also recrystallize at a more rapid rate. Creep Tests on Grade B Steel Tests similar to those above were also conducted on Grade B steel at 850~F. The results obtained are shown in Figures 6 through 10 and in Tables V and VI. Figures 6 to 9 inclusive give the time-elongation curves which were obtained, while Figure 10 shows the results plotted to logarithmic coordinates. The results are summarized in Table V. Six stresses were employed in the up-step method of loading, six in the single-step method, two inaone down-step method, and three in the other dovrn-step method. The stresses em

Table V Deformatiorts ObteinedDuring Creep Tests at &50o' on Grade B Steel Plastic Deformnation Total Time During Decreasing Total Plastic Pate of Method of Elastic of Test Stage of Creep Deformation Cree Loading Stress Deformation Hours In_/In. Time, Hrs. Up-Step 7,500 0.000265 340 0.000313 240 0.003360 0.00 Up-Step 9,225 0.000159 305 0.000125 215 0 000l30 0.00 Up-Step 11,700 0.000160 600 0.000230 455 0.000240 0.00 Up-Step 14,400 0.000160 - 490 0.000320 320 0.000375 0.035 Tp-Step 16,795 0.000160 355 0.000200 75 0.000450 40.8 Up-Step 19,055 0.000160 245 0.000235 50 0.000725 0.275 single-Step 7,500 0.000265 340 0.000313 240 0.000360 0.00 Single-Step 9,225 0.000561 738 0.000439 550 0.000449 0.00 Single-Step 11,700 0.000650 785 0.000510 785 0.000510 0.00 Single-Step 14,250 0.000800 640 0.000350 50 0.000320 0.00 Single-Step 16,795 0000945 670 0.000685 100 0.001000 0.049 Single-Stp 19,055 0.001060 680 0.301040 175 0001570 0.113 Down -Step() 19,055 0.001060 680 0.001040 175 0.001570 0.113 Down-Step(l) 16,795 0.000090 775 0.000000 0 0.000245,.03l Down-Step(2) 16,795 -0.000090 600 0.000290 250 0.000510 0.036 Down-Step 2) 15,400 -0.000080 453 -0.000010 25 0.100000 0.016 Down-Step 2) 14,250 -0.000045 383 000000O 0 0.000070 0.009 (1) Original stress of 19,055 pounds maintained constant for 630 hours. (2) Original stress of 19,055 pounds maintained constant for 24 hours0 H

prloyed ranged from 7500 to 19,055 pounds per square inch and the time periods for each test varied fro-m 305 to 785 hours. The results indicate Grade B steel to be considerably more resistant to creep at this temperature than Grade A steel. In fact, in the up-step method of loading, this material was able to withstand a stress of 11,700 pounds without continuous creep. In the single-step method a stress of 14,250 pounds did not produce continuous creep. The method of loading exerts the same general effect on the observed creep characteristics of this steel as were found with the Grade A material. That is, the up-step method of loading produced the lowest observed creep value, Awhile the down-step method yielded the maximum value. The differences in the observed creep characteristics as obtained by the different methods of loading can best.be seen from Figure 10 in which the stress and the corresponding rate of creep are plotted on logarithnmic coordinates. Values taken from this figure are given in Table VI. From Figure 10 and Table VI it is evident that as in the case of Grade A steel there is an appreciable difference in the observed creep characteristics depending upon the testing procedure employed. In this case, however, the differences

"~t'~~~~~~~' ~ [ i -7 r~~ ~ ~~~ ~~~~~~~~~~~~~~~~~ T I -. I-.........., $...... - -~C~: — i H _.i —~-__ l ~_ I t — ~ —-----—;.......,~~~ r ~ -!~ - - ~-t —'~' t~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~'~- -"' t: ~ t~ 7' i. _.......... - - ) - ~;' "1'') ~~~~~~~~~~~~~~~ —... 1~ ~ ~ ~ ~ ~ ~ ~ ~~" i,,, -+~~ ~~~~~~~~~~~~~~~~~,-,-,,,:!!i::_ ~~~~~~t~ I::,,!,l,~L -.....-i —,-!~! ~,....~ iI~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ r~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ i i'i ~-t t~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~' ~~~.~ "m.....'r. i1 I 2 I~~~~~~~~,, 1'~~ ~ ~ ~~~~~~~~~~~~~~~~~~~~~'~ ". 1C~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i'.!r I..~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~! 1 I ~!: -I ~;I t 1 -- t~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 8 t~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~,~'..... i~t-i tk -~~~~~~~~~~~~~''...' ——'-11........ -— t —--...................~*..... -....- - I, I i31;~~~~~~ (R) Org 1 loa ma —t —ib i$~~~~~~~~~~~~~~~~~~~~~~~~~o j~ ~ ~ ~~].........~~ ~

19. Tab le VI Effect of MLethod of Loading on the Observed Creep Characteristics of Grade B Steel at 850~F. Stress for Designated Rate of Creep Rate e Per Cent er 1000 Hours Me thod of Loading 0*01 0.10 1.00 Up-Step 12,000 16,900 23,500 Single-Step 13,000 18,800 27,000 Down-Step (1) 14,500 19,000 24,500 Down-Step (2) 14,500 19,000 24,500 (1) Original load of 19,055 pounds maintained constant for 650 hours. (2) Original load of 19,055 pounds maintained constant for 24 hours. are not as great at the lower rates of creep, and they are more uniform over the entire creep range. The stress required for a rate of creep of 0.01 per cent per thousand hours ranged from 12,000 pounds for the up-step method to 14,500 pounds for the down-step method. For a rate of creep of i per cent per thousand hours, the creep stresses range from 23,500 to 27,000 pounds. In this latter case, however, it is the single-step method rather than the down-step which yields the highest results. The explanation for the observed differences are believed to be the same as those discussed under Grade A steel.

20.!ffect of Method of Loadinq on Tr1atfieldt's Time Yield Value As Dr. Hatfield's"Time-Yieldv' value is now defined, it is that stress which, during the 48-hour period immediately following the first 24 hours of the test, produces a total plastic deformation of 48 millionths (0.000048) of an inch per inch. In the following table are given values which show the deformation occurring during the 48-hour period imediately following the first 24-hour pe.riod for Steels A and B when subjected to creep tests by the single-step, the upstep and the dovmn-step methods of testing. In this table are also given the computed values for the "Tirme-Yield" value. The values in Table VI clearly indicfte that.in practically every case a larger plastic deforration is obtained in the period between the 24th and 72nd hour when the single-step method is employed than with the up-3tep or downstep me'thods. According to these values the timrre-yield value for Grade A steel at 8500F. would be 6,850 pounds per square inch when the single-step method of loading is employed, 7,800 pounds for the up-step, and 13,500 or 13,300 for the down-step method of loading. The corresponding stress for 1 per cent creep per 100,000 hours varies from 6,900 to 11,700 pounds per square inch, depending on the rmethod of loading employed.

21. Table VI Deformation During 48-1Hour Period Immediately Following First 24-'Tour Period Deformation During Period from 24th Hatfield's TimeAt ethod of Stress to 72nd Hour Yield Value Creep Stres, Loading Lb./. In. Inches per Inch..b./q.In. 1O,00 r Grade A Steel Single-Step 7,500 0.003060 Single-Step 9,225 0.000070 Single- tep 11,700 0.300210 6,850 9,5 00 Up-Step 9,225 0.000085 Up-Step 11,700 0.000150 7,800 6,90" Down-Step(l) 14,250 0.000060 Down-Step(l) 11,700 0.000020 13,500 11,500 Down-Step(2) 12,525 0.000035 Down-Step(2) 11,700 0.000020 13,300 11,700 Grade B Steel Single-Step 7,500 0.000043 Single-Step 9,225 0.000090 Single-Step 11,700 0~000075 %Single- Step 14, 250. 0.000030 Single-Step 16,795 0.000200 10,600 13, 000 Up-Step 9,225 0. 000053 Up-Step 11, 700 0.000054 U-o-Step 14,250 0.000080 Up-Step 16,795 0.000115 11,500 12,000 Down-Step(1) -19,055 0.000100 Down-Step(l) 16,795 0.000015 17,700 14,500 Down-Sten (2) 16,795 0.000090 Down-Jtep( 2 ) 15,400 0. 000025 Down-Utep(2) 14,250 0.000020 15,900 14,500 (1) Original load maintained constant for at -least 600 hlours. (2) Original load maintained constant for only 24 hours.

22. Wi th Grade B steel at F350~F. the "Ti-Me-Yield" value for the single-step method of loading is 10,600 pounds per square inch, with the up-step "method 11,500 pounds, and with the down-step method either 17,700 or 15,900 pounds. The corresponding stress for 1 per cent creep per 100,000 hours varies from 12,000 to 14,500 pounds per square inch depending on the rmiethod of loading employed. COPOC LUSIOITS Creep tests conducted on Grades A and B plain carbon steel at 8500F. by three different testing procedures allow the following general conclusions to be drawn regarding the effect of method of loading on the resulting creep characteristics. The higher carbon steel, that is, the Grade B material, was found to possess the maximum creep characteristics at 8500F. regardless of whether these values were determnined by the up-step, the single-step, or the downstep mre thods. For a rate of flow of 0.01 per cent per 1000 hours (1.0 per cent per 100,000 hours), the difference between the two steels for any given method ofloading varied frco 2,800 to 5,100 pounds per square inch.

The observed creep values were also found to vary considerably depending upon the rmthod of testing employed. With both steels, the minimuLm stress values for a rate of flow of 0.01 per cent per 1000 hours (1.0,per cent per 100,000 hours) were obtained with the up-step method of testing, and the maximum values were obtained with the down-step method. In the case of the lower carbon steel, Grade A, these values ranged from 6,900 pounds for the up-step method to 11,700 pounds for the down-step method. The corresponding values for the higher carbon steel, Grade B, varied from 12,000 to 14,500 pounds. The explanation of these apparent discrepancies in the observed, creep characteristics is believed to be as follows. At this temperature, creep, especially during the early stages of the test, is due to a combination of strain-hardening and recrystallization. The recrystallization Drocess is necessarily very slow and the rate of creep will become constant only after an appreciable ramount of strain hardening has occurred. In the up-step method of testing, the initial loads employed are relatively small and the rate of strain hardening is also slight. A considerable period of time is, therefore, required before the rate of creep assumes its minimum value. Evidently, the test was

24. not continued a sufficient length of time in order for the rate of creep to be at its minimum value. The results reported are, therefore, low. With the down-step method the conditions are different in thnt the initial stress is relatively large and a large degree of strain hardening occurs in a relatively short period of time. When the load is, therefore, reduced, creep proceeds at a uniform rate almost from the start of the test. The total amount of plastic deformation obtained also varies depending upon the testing procedure which is used. In practically every case, the greatest amount was obtained with the single-step method. If allowance is made in the other two methods, however, for the plastic deformation which occurred under the previous loads, the difference between the three methods would probably not be very marked. The effect of the method of loading on the resulting "Time-Yield" value was also determined. Again, it was found that, because of the relatively large amount of deformation which occurs during the first hundred hours or so of the single-step tests, the "Time-Yield" values obtained by this method were below those obtained by the up-step or down-step methods.

UNIVERSITY OF MICHIGAN III3 9015 02827 2766 LW 9015 02827 2766