F- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN ENGINEERING RESEARCH INSTITUTE UNIVERSITY OF MICHIGAN ANN ARBOR PROGRESS REPORT NO. 7 THE INFLUENCE OF SURFACE TREATMENT ON THE FATIGUE PROPERTIES OF TITANIUM AND TITANIUM ALLOYS By L. THOMASSEN M. J. SINNOTT A. W. DEMMLER, JR. Project 2034 WRIGHT AIR DEVELOPMENT CENTER, U. S. AIR FORCE CONTRACT NO. AF 33(616)-26 SUPPLEMENT S-2(53-582) E. 0. NO. R615-11-SR 3c, PROJECT NO. 53-670A-61 February, 1954

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN SUMMARY The effect of surface rolling annealed V-notched specimens of Ti-75A, RC-130B, and A-110AT titanium alloys has been evaluated. In all alloys a very marked increase in fatigue life, as compared to specimens given a machined notch or a ground notch, was obtained. The effect of grinding a V-notch and a square notch in Ti-75A and V-notches in RC-130B and A-11OAT alloys has been evaluated. The fatigue properties of the ground notches were inferior to the properties obtained on the same type of notches that were machined or rolled. J ii

F- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN INTRODUCTION Previous progress reports and the summary report have presented the results of fatigue testing titanium alloys Ti-75A, RC-130B, and A-11OAT to determine the influence of various types of surface treatments and notches on the resulting fatigue life characteristics. The present report deals with the fatigue life characteristics of ground notched specimens and rolled notch specimens of these same alloys. EXPERIMENTAL PROCEDURES Two of the three alloys used had been previously used in work already reported. The analyses and tensile properties, as determined after a one-half hour anneal in argon at 1450~F, are given in Table I. Also included are the properties of a new heat of RC-130B that was used in this work. Since the previous work on the influence of surface conditions on the fatigue properties of these alloys had shown that the method used to prepare the surface had, in some cases, a pronounced effect on the resulting fatigue life the work on the notch fatigue strength properties was planned to evaluate the effect of the method of notch preparation. The previous progress report (No. 6) gave the fatigue properties of machined notches. The present report presents results of notches prepared by grinding and surface rolling. The ground notches were prepared by machining the specimens to size except for the tapers and the notch proper. These regions were left 0. 007 inch oversize. The specimens were then annealed in argon for one-half hour at 14500F to remove the effects of the prior machining. The tapers were then ground to finish size and the unfinished notches 1

F- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - were forwarded to Dawson Carbide Industries of Detroit, Michigan, for the diamond grinding of the notches to the finish dimensions. This diamond grinding procedure was found to be necessary since it was impossible to grind the notches to a controlled root radius of 0. 010 inch, using the conventional alundum or silicon carbide types of grinding wheels. The wear and fracture, particularly the fracture, of these wheels was excessive and gave poor root radii and surface finish. The diamond grinding gave excellent results. The root radius was held exactly to dimension and an excellent surface was obtained. Diamond grinding from the oversize diameter to the finished size was done in steps of 0. 002 to 0. 003 inch. Since shot peening the surface of specimens improved the fatigue life properties it was believed that a thread rolling type of operation, if used on the V-notch specimens, might improve the notch fatigue properties by cold working the surface layers of the notch in somewhat the same fashion that the shot peening works the surface of a metal. Notch rolling was accomplished by machining the specimens to size except for the notch which was made 0. 005 inch oversize. The specimens were then annealed for one-half hour in argon at 14500F to remove the effects of the machining. The notches of the specimens were then rolled by adapting a pipe cutting tool in such a fashion that the cutting wheel which had been remade, hardened, ground, and polished, rolled the material to the finished notch size with a root radius of 0. 010 inc It was found that 0 005 inch could be reduced on Ti-75A and yield a good surface in the root radius. The surface on the RC-130B and A-11OAT alloy was not as good as that obtained on the Ti-75, in that it tended to be rough, although the root radii were on dimension. 2

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN The first tests on Ti-75A specimens of the rolled notch type showed that such a treatment very markedly increased the fatigue strength. It raised the strength to such a level that the capacity of the testing machines was exceeded. This necessitated the redesign of the test specimen. The notch diameter was decreased to 0. 250 inch in place of the previous 0. 300 inch. The outside diameter was decreased from 0. 480 inch to 0. 400 inch. The theoretical notch sensitivity is not particularly sensitive to these changes and is more dependent on the root radius which was held at the original value of 0. 010 inch. The notch sensitivity of these new type specimens is on the order of 2. 7 to 2. 9 versus the value of 3. 2 of the previously used notches. The actual fatigue data on the various alloys are presented in Tables II and III. The lack of sufficient stock prevented the use of extensive replicate testing; generally duplicate or triplicate specimens were tested at each stress level for each alloy. SUMMARY AND DISCUSSION The stress-cycles of stress plots constructed from the data of Tables II and III are given in Figures 1 and 2. Table IV is the estimated mean fatigue life data taken from these plots. Figure 3 is a plot of the fatigue strength reduction factor, the ratio of the fatigue life of the hand finished unnotched specimens to the fatigue life of the notched specimens, as a function of fatigue life. It is apparent from these data that a rolling operation very markedly improves the fatigue life properties. The notch effect is completely obliterated and the fatigue properties are superior to those chosen as the standard, the annealed hand finished surfaces. The amount of 3

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - reduction given during rolling was quite small, on the order of 4. 0%, but this was sufficient to give the improvement noted. It is questionable whether still further reduction would give a correspondingly greater increase. This evidence seems to confirm the conclusions drawn from the data on the shot-peened surfaces, i. ee, the improvement is due to residual compressive stresses and surface and sub-surface working of the metal. The large scatter noticeable in the data on the RC-130B and A-11OAT alloy may be traceable to the non-homogeneity of the working operation. An interesting development occurred in testing these rolled notches. The specimens of all the alloys showed no tendency to heat internally to any appreciable extent but the notch radius of the Ti-75A was noticeably oxidized at the conclusion of testing and some tendency for oxidation was observed in the notches of the RC-130B and A-11OAT alloy. It is believed that the rolling operation changed the internal friction characteristics of the material to such an extent in these areas that localized overheating of the surfaces occurred. In spite of the overheating, however, the fatigue life properties still showed an improvement over the unnotched surface properties. The diamond grinding was not as severe a grinding operation as has been performed on the specimens that had been surface ground. It was anticipated that such grinding would not adversely affect the properties. The results indicate that generally this is true but that the ground notches tend to have slightly higher fatigue reduction factors indicating a slightly greater notch sensitivity. In view of the fact that the grinding gives such similar results to the machined notches, further work on ground notches is being discontinued. 4

-- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - FUTURE WORK A heat of an all beta alloy (30% Mo) has been received and work is proceeding on evaluating this alloy under the conditions used in testing the previous alloys. An all alpha alloy (6% Al) ordered at the same time as the all beta alloy has not yet been delivered. Unless receipt is obtained shortly, the work on this alloy will not be completed at the termination of the project. The effect of annealing at various temperatures a ground surface of Ti-75A is being investigated to determine if such a procedure will produce a recovery in fatigue life. 5

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN TABLE I Analyses and Mechanical Properties* ~~~~~~~~~ r.,.,. Ti-75A RC- 130B A-11OAT Heat M-270 C-0. 055%; N-0. 10%; Fe-O. 20%; W-0. 01% Heat B-5331 C-0. 1%; N-0. 04%; Mn-4. 4%; A1-3. 9% Heat R-31314 C-0. 1%; N-O. 05%; A1-4. 13%; Sn-2. 58% Ti-75A RC-130B A-110AT Prop. Limit psi 47, 900 138, 000 124, 000 Tens. Str. psi 84,300 148, 000 138, 100 % Elong. 2" 27. 3 18. 7 16. 0 % Red. in Area 42. 0 46. 0 40. 0 Duplicate Specimens 6

-- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN TABLE II R. R. Moore Rotating Beam Fatigue Data Rolled 600 V-Notch Specimens 6000 Cycles per Minute 1000 Cycles of Stres Ti-75A 70,000 psi 38 19 81 65,000 psi 188 203 RC- 130B 120,000 psi 59 15 432 110, 000 psi 496 555 169 A-1 OAT 115,000 psi 58 263 110, 000 psi 69 87 2, 852 844 62,500 psi 250 678 57,500 psi 794 1,314 50, 000 psi 11,454 * 10,052 * 11,543 * 105,000 psi 1,496 704 686 100, 000 psi 782 1,078 95, 000 psi 1, 113 7,904 105, 000 psi 557 4, 143 100,000 psi 5,470 5,617 13,013 * 95, 000 psi 10, 195 * 11,655 * 90, 000 psi 5,541 2, 853 Did not fail Did not fail. 7

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN TABLE III R. R. Moore Rotating Beam Fatigue Data Diamond Ground Notches 6000 Cycles per Minute 1000 Cycles-of Stress ---- 60~ V-Notch Square Notch Ti-75A RC- 130B A-11 OAT Ti-75A 30, 000 psi 73 41 56 25, 000 psi 132 284 564 55, 000 psi 11 12 50,000 psi 15 19 40, 000 psi 54 42 35, 000 psi 92 125 90 35,000 psi 40 52 43 30, 000 psi 137 181 128 20, 000 psi 1, 003 1,079 648 17, 500 psi 2, 773 10,043 * 10, 162 * 45, 000 psi 28 26 40, 000 psi 61 52 30, 000 psi 250 244 121 25, 000 psi 908 945 660 25, 000 psi 1, 116 251 361 22,600 psi 2, 187 1, 285 10, 000 * 35, 000 psi 127 172 20, 000 psi 11,043 * 11,012 * 10, 595 * 20, 000 psi 10, 000 * 11, 104 * 2, 357 30,000 psi 690 370 25, 000 psi 10, 000 * 10, 000 * Did not fail Did not fail. 8

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN TABLE IV Estimated Stress-Mean Fatigu.e Life Data Cycles 4 5 6 7 10 10 10 10 Stress, 1000 psi Ti-75A, Ground V-Notch 44. 0 27, 5 19. 0 16. 5 RC-130B, Ground V-Notch 56. 0 37.0 28. 5 25.0 A-110AT, Ground V-Notch 51.0 35.0 24.0 20.5 Ti-75A, Ground Square Notch 48. 0 32. 0 23. 0 20. 0 Ti-75A, Rolled V-Notch 78.5 67.0 56.0 50.0 RC-130B, Rolled V-Notch 136.0 119. 0 103. 0 87.0 A-110AT, Rolled V-Notch 128.0 116.0. 105.0 96.0 - 9

'HLON-A.09 aasrToi*'SLO"Id SS3EaLS JO S'33D-SSHLS'1 nflDI 931 0AO o01 901 g01 01 rn U) () II ) I

o 40 0 _,, IC wo 30 A-i \TToS 10. 104 I 05 106 IC CYCLES FIGURE 2. STRESS-CYCLES OF STRESS PLOTS. DIAMOND GROUND NOTCHES.

3.5 2.5 xm I 0.5 05 106 CYCLES FIGURE 3. FATIGUE STRENGTH REDUCTION FACTOR VERSUS FATIGUE LIFE.