ENGINEERING RESEARCH INSTITUTE UNIVERSITY OF MICHIGAN ANN ARBOR BIMONTHLY PROGRESS. REPORT NO VIII THERMAL-SHOCK INVESTIGATION By A. R. BOBROWSKY =L. L. THOMAS Project M949 WRIGHT AIR DEVELOPMENT CENTER, U. S. AIR FORCE CONTRACT AF 33(038)-21254; E. O. NO. 605-227 SR-3a March, 1953

42Q r\t 2 CK \ 'm 'I -

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN BIMONTHLY PROGRESS REPORT NO. VIII THERMAL SHOCK INVESTIGATION.OBJECT The object of this research is to evaluate optimum design of-test specimens and criteria which will permit correlation of thermal-shock data with performance of the material in the form of turbine buckets. SUMMARY I Data are presented on the themal-shock resistance of 21 specimens of N-155 alloy. The best specimens of this alloy appear to be as resistant or more resistant to cracking by thermal shock than those of Inconel and S-816 alloy previously reported. As would be expected, resistance to cracking by thermal shock decreased as the specimen temperature was increased from 1600~F through 17000~F to 18000~F The marked effect of difference in thermal-shock resistance between specimens from two different bars is definitely shown by results on two lots of this material. This effect may have been due in part to minor differences in supposedly identical heat treatments. INTRODUCTION Previous research has been concerned primarily with stainless steels, Inconel, and S-816 alloy. This research, conducted primarily during the period August 11, 1952, through November 11, 1952, was concerned with extending results to another material, namely N-155 alloy.

I i ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN APPARATUS The test rig employed was substantially the same as that used in tests reported in the last two progress reports, with modifications to the specimen-holding assembly and crack-detection procedure. The specimen holder (Fig. 1) was altered to hold the specimen horizontally, with the air blast coming from below to the cooled edge of the specimen. The right end of the specimen was held in a combination electrodesupport that was free to translate horizontally under the thermal expansion of the specimen during the test cycle. It was constrained, however, to prevent the specimen end from rotating, in order not to interfere with the positioning of the specimen relative to the air nozzle. The air nozzle was rigidly connected to the specimen-hold'ing assembly. A movable gage attached to the specimen holder was used to position the back of the specimen normal to the line of sight of the radiation pyrometero An end stop *on the specimen holder positioned the specimen along its axis at the left end. In this way all specimens were located in substantially the same position. Crack detection was facilitated by the use of a measuring telescope (Fig. 2). This telescope was mounted rigidly on the frame of the test rig, free only to rotate in a horizontal plane so that the telescope could be used to scan the length of the cooled edge. This setup enabled cracks to be detected almost at their inception, and the crack growth to be followed during the test. Only cracks which were obscured by oxide formation could not be definitely identified until the end of the test, when they were opened by bending the specimen; such cracks were present infrequently. A magnification of five diameters was found to be suitable for this study. The cracks were not equally visible during all portions of the test cycle, nor was the most suitable portion of the cycle the same for all specimens. In general, however, cracks were most easily discerned immediately after the air blast had started or immediately after the air blast had ceased, The specimens used contained no central holes, since the radiation pyrometer had been found to be reliable provided the surface of the specimen was adequately oxidized prior to test. All specimens were preoxidized prior to these tests. L 2

L; IL ' I. a m -'- - i "Z; Fig. 1. Front View of Specimen Holder, Specimen, Air Nozzle, and Radiation Pyrometer. Fig. 2. View of Specimen Holder With Measuring Telescope in Position To View Specimen 3

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - Two lots of specimens, identified by Roman numerals I and -I in the sixth column of the log, were heat-treated for 20 minutes at 220Q0F, waterquenched, and soaked for 50 hours at 1400~F. These specimens were obtained from different bars of the same heat, and were heat-treated at different times. Although no differences in microstructure were apparent on viewing the cross sections of the bars (Figs. 3 and 4), longitudinal sections (Figs. 5 and 6) -revealed substantially greater twinning in lot II. This twinning is best seen in Fig. 6. These two lots of specimens manifested quite different behaviors in the thermal-shock tests. RESULTS The results from the tests on the two lots are different, although results within eaqh lot are reasonably consistent. A table of results is shown below in order of the number of cycles to failure. 1600oF 1700 F 1800~F Cycles Lot Cycles Lot cles Lot No crack at 10,124 I 3764 I 2052 I No crack at 3,886 I 3248 I 1818 I 5,153 II 3211 I 15o8 I 3,530 I- 3195 I 1228 I1 3105 I 1130 II 2888 I 1095 Ii 2320 II 1042 II 2229 II 990 IT 1995 II The N-5-..- aliy:O from Lot I outpefrorme d the S-816 alloy'. - t is about -the same as -the best Ilconel specimens in resistance to cracking by thermal shock. This behavior is not evident in Lot II, which shows approximately the same behavior as the S-816 alloy previously tested. It is possible that the twinning which is present from large amounts of original cold-working appreciably reduces the thermal-shock resistance of N-155 alloy. It is known that small amounts of cold work (tensile strains up to 10q) had no major effect on specimens of Inconel. (Progress Report No. VII). If cold work is the reason for the difference inl behavior of the two lots of N-155 alloy, then the alloy is either more sensitive than Inconel to cold work or else a tensile strain of 10% was not sufficient to reduce appreciably the 4

OOTX 'XG Jo UOTGSS ssoD3 'XOTTV 55T-N '* '*1 II ~oq (q r ~~~~~~~~~ ~ '...,~ --., f '~\., '~ ~: ',"/.;' 'i, ~;h ':.'~.,~,~:.~ (5,.: ~.., ~,r ~_;>~ I 1%oa (I

I I Hi...J-~~~~~~~~~~~ A~ ~~~~ ~. $ " '''...,eA ~0* i 4 k~~~flIi o *- ', P'o -".'A ~,?.q,..~.' J~,.D H. A '.4. ~ c.!.', ON - J#.J r.... ~~~~~~ro ~;, -.~ '.~, ~ ct~... -. "... H-J o o. A H) - L.~. k.,,~..: ~o. $ r ~ r- 1Si74i~ C~\h~t ~~~J~~.,:. ~ '., ' ~el ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.. CC 1 1~~~~... ~ tr~ r...~.;,,...."':j O ~" ' rl "./.,\ P~~~~~~~ "' IIW" ilk,

L 001:x C'.t~' $0 urorq o@9S 'l:~pn q. T2~o ~I ~'~oT~rv ~'-~'~ II ~o? (q., '' r: ~..,~~~~~~~~~Y ~~~~~~~~LL~~~~~~~~ ~~~ '~ ~~~~~~~~~~~~~~~~~... ~I r~~~-'~~~~~~~~~~~~~""'! ',i -'L. ~ LC~SF~ I~' Y''U'i C~cC~sL~PYrf~y ItLrC!I!:1~~~"~ ~~~~~.~~:s~r7ry~TrL~c ~I q'o (~.~ ~~: ~.... ~~' ~~*c;-\r '~ J.-V,.n~r~~l~';,c'iChacCLc:~.,t~r-.~ j -~.X: x~ "2'zcG.:--nrcCI ~~r r;t -t;.J --- ~, d ~"b I/ I ~~ ~r 1 u '"' P:.r.1. :f * i 1 r.lcqr '1 lZ~?.. r.di ~,I ~Y I s~;r '-i i /"c~;:.r' ' - -7*(~, \~ " ~~ r. -~-C i-.. I, i ~.ir.- ~.;irkici~ -.C dC.xrqrr~ L.,.J 3~';I t ne -. ~, ;rL/ J L1 ~)C ~/~,-'~r

,~....'. l~~~~d' * K-"?~~~r, i'.:.;.,-."4 ~~". Ne ~ ~~,, *.). H'r - - f", ~~~~~~~~~~~~oq C'! " ~ ~ 2 "t r ~C' \:Ii o H~~~~~j ' r~~~~~~~~~~~~~~~w 0 0 U1 iCI r~~~i' ~ ~ ~1 C+~~~~~~~~~~~~~~' OD C~. ~r r bd~~~~~,,fc P ~ f,~. 0~~~~~~;'9t' 0.,(

F- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - the'thermal-shock resistance of Inconel, or both. In any case, N-155 appears to be the best or nearly the best alloy, from the point of view of thermal shock, of any of the materials studied to date. CONCLUSION Wrought Nl155 alloy, heat-treated as described above, shows thermal-shock resistance equal to or better than wrought S-816 alloy. One lot of N-155 alloy, free from twinning, was better in resistance to thermal shock than another lot, which showed large numbers of twins. The best Inconel and the better lot of N-155 alloy were about the same in resistance to cracking by thermal shock. 9

ENGINEERING RESEARCH INSTITUTE UNIVERSITY QF MICHIGAN -- Column (1) Column (2) W.045 P.F. x KERY TO LOG (1) Relative position on bar stock 1 Specimen number Arrow indicates direction and location of cooling jet; cooling medium is air unless otherwise stated Cooling medium is water Width of cooled edge, inches Previously subjected to rot&ting beam fatigue as shown in column (6) Failed during pre-fatigue Colu 4 mn (3) M Thermal shock cycle manually controlled 1500/5 Automatic cycle control.; maximum temperature, -~F, and length of cooling period, seconds P1800 Dead load, 1800 lbs -10/100 Starting with stated maximum temperature, maximum temperature was increased 100F after each 100 cycles 40o5K Reversed-bending (rotating-beam) fatigue tests; maximum stress, 40o,500 psi to 1800 Maximum temperature held constant after 18000F was reached Column ( no symI Column ( Column ( A 0. T300/1l Gl. A4) A Air cooling for stated number of cycles W Water cooling for stated number of cycles )ol Air cooling for stated number of cycles 0 No failure visible F Fracture C Cracks G Grooves FC Face crack PC Possible crack I6) B Specimen warped due to thermal strains.14 Area of cross section, square inch 500 Heat treated before testing 300 br at 1600~F 500 Grooves first appeared at 1500 cycles OH Stated maximum temperature was exceeded due to malfunction of control unit BT Broke through to thermocouple hole i I A-1 - - -- --

L ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN —. 40.5/f 8200o 'R N +100/5108 Check TII P PT1 C20/1700 Previously subjected to 82000 cycles at 40,500 psi Reproducibility test Specimen formed a neck due to tensile strain. Maximum temperature was increased 100~F at 5108 cycles. Second test to determninb the effect of" alteration' of testing procedure. Study of crack propagation Previously subjected to tensile strain of 1% at room temperature Long-time test at reduced severity, Test No. I Heat treated as shown in braces ( 3. Lot No. I Heat treated for 20 hours by heating to 1700~F and allowing to cool for 5 seconds by natural convection. i A-2

EEST LOG Specimen Cross Number Type of Number Section Cycle of Cycles Failure Remarks (1) (2) (3) (4). (5) (6) Type 304 Stainless Steel 1 aJ M 0 B 045 2 7 1600/10 4400 A C B 300 w 3 \7 1600/4 1783 C 4a Fatigue 40.5K 3300 F 4b Specimens 40.5K 2600 F 5 T 1700/4 1100 o __ ~7 1800/4 675 C 6 \O/ 1600/4 6240 0 G6500 190oo/4 1240 C 7 9 1500/4 4130 F A 0.16 9v ~P600 -A-3

TEST ILOG (cont) Secimen Cross Number Type of Number Section Cycle of Cycles Failure Remarks (1) (2) (3) (4) (5) (6) 8 ( 1600/5 3082 o T300/1600 1800/4 517 C 9 u 1500/3 5753 0 ~7-57 1600/4 1000 0 10 1700/4 1000 0 1800/4 80 C 11 ) 1500/5 1000 F A 0.132 V P1800 1500/5 P600 5000 0 A 0.133 12 P900 1200 0 P1800 203 F 13 1600/4 128.1 C G 115 14 1500/4 1000 F OH 15 _7,7 1600/5 1900 C T300/1600 16 1600/5 409 C......... -.....~~~~~~~~~~~~. A-4

LTEST LOG (cont) Specimen Cross NuTrr.r' Type of Number Section Cycle of Cycles Failure Remarks (1) ~(2) (3) (4) (5) (6) 17 ( 1500/5 300 F A 0.140 P1800 18 1800/4 1950 C G 1500 19 1700/3 C V w 20 1o500/3 1000 0 BT Type 347 Stainless Steel 1 \7 1600/4 866 C * 0is,, +10/100 2 97 '1600/4 1147 C H ozo +o10/100oo 3 9 1500/4 575 C BT +10/100 4a Fatigue 54K 5200 F 4b Specimens 54K 10400 F 40.,K 82000 A-5

TEST LOG (cont) Specimen Cross Number Type of Number Section Cycle of Cycles Failure Remarks (1) (2) (3) (4) (5) (6) +10/100 6;7 1500/4 1990 C +10/100 r6 1600/3.54 1990 7 +10/100 2700 G to 1800 8 (Defective) 9 1600/4 2863 C R 10 K7020 1600/4 3787 C Check II 11 zo o 160oo/4 2580 C 12 1600/4 3162 C G 736 13 1600/4 2204 C G 2072 14 7 0 1600/4 2707 C G 2604 -oZ I,,,.. 0 I...,.l,..I A-6

TEST LOG (cant) Specimen Cross Number Type of Number Section Cycle of Cycles Failure Remarks (1) (2) (3) (4) (5) (6) 15 \ 1600/4 3003 C G2820 R.2.O.. 16 1600/4 2518 C R 17 16oo00/4 4850 0 Check I 18 / Fatigue F 54K 64K 103300 19 1600/4 1825 C R,035 37K/217100 42K/llooo 20 Fatigue 4300 F 48K/35600 64K 54K/10000 59K/10400 21 1600oo/4 4430 C 22 (Defective) 23 1600/5 2962 C 24 Fatigue 3V,59K 52900 F 59Kc A-7

TEST LOG (cont) Specimen Cross Number Type of Number Section Cycle of Cycles Failure Remarks (1) (2) (3) (4) (5) (6) 25_ 1600/5 1562 C 54K/50000ooo Dor P.F. 53K/52000 2j67 59K/12000 26 1600/5 1960 C 64K/1000,v 10. o/70K/1000 75K/500 2X 53K/52000 V 0/0 P.F. F 59K/11300 53K/52000 59K/12000 28 1600/5 1594 C 64K/1000 __,O/O P.F. 70K/1000 75K/500 53K/52000 59K/12000 29 X C 64K/1000 V P.F. 70K/1000 75K/300 30 1600/5 1973 C 31 RE1V R 1600/5 2764 C 32 \Xo IO 1600/5 1500 C A-8

TEST LGy (coat) Specimen Cross Number Type of Number Section Cycle of Cycle. Failure Remarks (1) (2) (3) (2) (5P (6) (4) o P.F. F 59K/32600 (14.) V.0o P..F. (3) 4it36 P.iF 1811 C 60K/39000 35 (2) (Used f or calibration of 'Heat-'Eye), 36 \ 7/ 1600/5 1859 IC 58K/30000 (I) -. Votfo P.F. (5) Ofo 1600/5 4635 C 58 Ta/2000 39 G 2440 (7) 7 1600/5 2440,iG Rigid Support (__+~~~~~~~ - ~Nozzle No. 3 (8)~ \@/ 1600/5 31-43,G:izle No. 4 41 1600/5 2710 C R6id 000uppor Nozzle No. 5 ~ A*9

TEST LOG (cont) (1) (2) (3) (4); (5) (6).Specimen Cross Number Type of Number Section Cycle of Cycles Failure Remarks 42 43 P (11) 4.600/5 10708 C Rigid Support.)25 Nozzle INo. 4 44 45. H.S. 21 (vitallium) Cast I \ 1500/3.5 1000 C BT Inconel r1 \0KXi o3 '1450. C. ~ V.0155003 7 7.o ~. +10/0.03'55~T7 tf- T,/Is400 30 5. rLliCllrl11112730a 3 ( 150./0,/3 428 C.67., 5 \0/?1700t/5 18159 G!2/500 _ _s _T s.o 3'0,.7 _ ' A-10

TEST LOG (cont) (1) (2 —) -(S-3#)- (4) (5) (6) Specimen Cross. Number Type of Number.Section Cycle of Cycles Failure BRemarks _+,0:6 1600/4 7449 C.035 T2/500 ~7~ ~' 700/5 4706 C T1/35/1400.~35 Tl/3/1400.05 V 1 8/_1700/5 2090 C P* —T: 10. ~~~~~~~\ 0/ -T1/3/14oo00 10 \0 1700/5 3680 C T10,035V 0 - T1/3/1400 11 \0/ i700oo/5 286o0 c PT5.028 V 12 Vi\ o/ T1/3/1400 13 1700/5 2500 C PT1.02 14 ~\ o~~~~/. ' TSI3/1400 14 \/1700/5 2527 C PT5 ~\ Z/J T1/3/1400 15 1700/5 2804 C PT10 \o16 757/1700 T1/3/1400 1700/5 3590 C PTO.025 V C \ o / T1/3/1400 17.\/ 170~/5 2270 C PTI 17~~ 220T/310 A-11

TEST LOG (cont) () (2) (:3) (4) (5) (6) Specimen Cross Number Type of Number Section Cycle of Cycles Failure Remarks 2576 PC T1/3/1400 18 V 1700/5 3015 C PT 19 - 1700/5 80 C PT10 Tl/3/1400 20 1700/5 2898 C PTO. -03 21...... *; * - Tl/3/1400...\./ 4339 FC? flex. pipe 22 < 1700/5 6866 C to nozzle,035 V P 23 1700/5 2250 C T1/3/1400 24 3538 FC T1/3/1400 25 0 /1700/5 4229 C S-816 Alloy (wrought) 1500/4 A 0 x8 '.;7 +100/10000 3 1700/4 2256 C 700o o 18395 C +15108 3~:o 70425 A-12

TEST W.OG (cont) (:L) (2)- (3) (4) Specimen Cross Number Type of Number Section Cycle of Cycles Failure Remarks 4.j 1700/4.2250 C 1600/4.3870 C 6,V1500/4 2.630 C 7 15 100/4 13280 C 8,V1600/4- 749 C > 97~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Ai.3

TEST LOG (cont) (1) -- (2) (3) —.(4) (5) (6) Specimen Cross Number -'rType of Number Section Cycle of Cycles Failure Remarks N-155 Alloy (Wrought) \018C~fB- /3878 C 50/1400 V.0 ' \/ < 4949 2C 5 2.o40V7.o 0 1700/5 3211 C 50/1400~} 3 7 1700/5 3248 C W l/3/20 I 4 1800/5 1508 C T 0//21400 5 771600/5 3886 0 T11/ /22001 W.o36 V Removed for check; Nb crack 6 \1700/5 105 C Tl/3/2200 W 7 1800/5 1818 C Ttl/3,2200 WI 8 \77 1700/5 3195 C Tf 1/3/2200 Wi 039 o 1., 50/1400 *9 '77 1700/5 2888 C Tfl/312200 w 037 50/1400 10 \7 1600/5 10124 0 Tf1/3/2200 W.041 V 50/1400 J 11I. I _ 1800/5 2052 C -Il1/3/-. I J2200,W1 A-14

TEST LOG (cont) ( '1. (2 t3) (4 (5) (6) Specimen Cross Number Type of Number Section Cycle of Cycles Failure Remarks 12 180/5 1228 T1/3/2200 W II 13 1800/5 1095 C T /3/2200 WII.o48 1800/5 50/1400 '- --.. -. ~ "'..... 14 100/5 220429 C T 1/3/2200: WI ', o V4:....... 5 15 7; 1800/5 990 C T l/3/2200 WII 19.0385.... 1 /5/2200W}I 0. 1800/5 1130 C Tl/3/2200o Wi 17 1700/5 '2295 C T//2200 WII \040 / 50/1400:9:oo433:3 A-15

UNIVERSITY OF MICHIGAN 3 1015 02514 81911111 3 0502514 8191