NASA-TN-D-2743-1965 Studies of fatigue crack growth in alloys suitable for elevated-temperature applications《适用于高温应用的合金疲劳裂纹扩展研究》.pdf

1、NASA TECHNICAL NOTE STUDIES OF FATIGUE CRACK GROWTH IN ALLOYS SUITABLE FOR ELEVATED-TEMPERATURE APPLICATIONS by C. Michael Hudson LangZey Research Center Langley Station, Hampton, Va. NATIONAL AERONAUTICS AND SPACE ADMINISTRATION WASHINGTON, D. C. APRIL 1965 4 Provided by IHSNot for ResaleNo reprodu

2、ction or networking permitted without license from IHS-,-,-I NASA TN D-2743 STUDIES OF FATIGUE CRACK GROWTH IN ALLOYS SUITABLE FOR E LEVATED-TEMPERATURE APPLICATIONS By C. Michael Hudson Langley Research Center Langley Station, Hampton, Va. NATIONAL AERONAUTICS AND SPACE ADMINISTRATION For sale by t

3、he Office of Technical Services, Department of Commerce, Washington, D.C. 20230 - Price $1.00 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-STUDIES OF FATIGUE CRACK GROWTH IN ALLOYS SUITABLE FOR ELEVATED-TENPEBATWE APPLICATIONS By C. Michael Hudson

4、 Langley Research Center Constant-amplitude axial-load fatigue-crack-propagation tests were con- ducted on 8-inch (20.3-cm) wide sheet specimens made of AM 350 (CRT) and AM 367 stainless steels, two thicknesses of Ti-8Al-lMo-lV (duplex annealed) titanium alloy, 2020-6, 2024-T81 ( clad), and RR-58 (c

5、lad) aluminum alloys, and Inconel 718 superalloy. Tests were conducted at room, elevated, and cryogenic temperatures to determine the effect of temperature on crack propagation in each material . The fatigue-crack-growth resistance of the materials was determined and compared with materials tested s

6、imilarly in a previous investigation. At ele- vated temperature, the O.O-inch (1.27-mm) thick titanium alloy, Ti-8Al-lMo-lV, in either the duplex- or triplex-annealed condition showed the greatest resist- ance to crack growth. At the room and cryogenic temperatures, the superalloy Inconel 718 appear

7、ed to be the most resistant. showed good resistance to crack growth at all temperatures but only a limited number of tests were conducted on this material. The AM 367 stainless steel INTRODUCTION A study of the fatigue-crack-growth characteristics of nine materials having potential use in supersonic

8、 aircraft is reported in reference 1 which is extended herein to include seven additional materials. Axial-load fatigue tests were conducted at positive mean stresses on 8-inch (20.3-cm) wide sheet specimens. Identical tests were conducted at elevated, room, and cryogenic temperatures to determine t

9、he effect of temperature on fatigue crack growth. The experimental results of this study are presented in this paper. The effects of temperature on crack propagation in each material were determined. In addition, the crack-growth characteristics of the seven materials tested are compared with the ch

10、aracteristics of the most resistant materials tested in the previous investigation (ref. 1) to provide a comprehensive ranking of each material with respect to resistance to fatigue crack propagation. Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-S

11、YMBOLS The units used for the physical quantities defined in this paper are given both in the U.S. Customary Units and in the International System of Units (SI). Factors relating the two systems are given in reference 2. a one-half of the total length of a central symmetrical crack, inches or centim

12、eters ( cm) N number of cycles Sa s, t specimen thickness, inch or millimeters (mm) alternating stress amplitude, ksi or meganewton/meter2 (MN/m2) mean stress, ksi or meganewtons/meter2 (MN/m2) TESTS Specimens The materials tested in this investigation are listed in the following table : Thickness M

13、aterial Stainless steel Stainless steel Aluminum alloy Aluminum alloy Aluminum alloy Titanium alloy Titanium alloy Super alloy Condition AM 350 (CRT) AM 367 2020-6 RR-58 (clad) 2024-T81 (clad) Ti-8Al-lMo-1V (duplex annealed) Ti-8Al-lMo-lV (duplex annealed) Inconel 718 in. 0.050 -050 .050 .063 063 .0

14、50 .250 -050 mm 1.27 1.27 1.27 1.61 1.61 1.27 6.35 1.27 All the specimens for each alloy were obtained from the same mill heat. The tensile properties of each material tested are listed in table I and the nomi- nal chemical compositions, in table 11. The general configuration of the specimens may be

15、 seen in figure 1. The All spec- specimens were 24 inches (61 cm) long and 8 inches (20.3 cm) wide. imens were made with the longitudinal axis of the specimens parallel to the 2 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-grain of the sheet. A 0.

16、1-inch (0.254-cm) notch was cut into the cen- ter of each specimen by means of an electrical discharge process. Very localized heating occurs in making notches in this manner. Thus, virtu- ally all of the material through which the fatigue crack propagates is unal- tered by the cutting process. Prio

17、r to shearing the specimen blanks, the sheet materials were covered with tape to protect the sur- faces. Following shearing, all spec- imens were chemically cleaned. Those specimens requiring heat treatment were then heat treated according to the procedures outlined in table 111. A reference grid (f

18、ig. 2) was photographically printed on the spec- imen surfaces to define intervals along the crack path. This photo- graphic method produces no mechanical defects in the specimen surface, and, consequently, no stress concentrations are introduced. Metallographic exam- ination and tensile tests condu

19、cted on specimens bearing the grid indi- cate that the grid had no detrimental effects upon the materials tested. Detail of stress raiser Figure 1.- Specimen configuration for crack propagation studies. Testing Equipment Three axial-load fatigue testing machines were employed in this inves- tigation

20、. having an operating frequency of 1800 cpm (30 Hz) was used for tests expected to last more than 1 000 000 cycles. lic fatigue machine which applied loads at a rate of 1200 cpm (20 Hz) was employed in tests expected to last from 10 000 to 1 000 000 cycles. tion hydraulic and subresonant fatigue tes

21、ting machine (ref. 4) capable of applying loads up to 132 000 pounds (587 IrN) hydraulically or 110 000 pounds (489 m) subresonantly was used as the needs for testing dictated. The oper- ating frequencies were 40 to 60 cpm (0.7 to 1 Hz) for the hydraulic unit, and approximately 820 cpm (14 Hz) for t

22、he subresonant unit. A 20 000-pound (89-MY) capacity subresonant fatigue machine (ref. 3) A 100 000-pound (445-kN) capacity hydrau- A combina- In all tests, loads were monitored by measuring the output of a bridge circuit whose active elements were wire-resistance strain gages. These gages were fixe

23、d to weigh bars through which the load was transmitted to a specimen. Monitoring precision was approximately *1 percent. 3 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-Figure 2.- Grid used to mark intervals in crack path. (1.27 m). Notch L-63-4299

24、. 1 Grid spacing is 0.05 inch The apparatus used in the elevated-temperature tests (fig. 3) consisted of three heating units and a steel framework which held the heating units in con- tact with the specimen. The heating units were composed of a 1/2-inch (l.27-cm) thick graphite plate, a ceramic bloc

25、k containing wire resistance heaters, and an insulating pressure plate. A machine screw was jammed against the insulating pressure plate to hold the heating unit in contact with the specimen surface. The screws were carefully tightened to insure thermal contact without intro- ducing significant fric

26、tional forces. observation side of the specimen; one above and the other below the region of crack growth. view of the propagating crack. The third unit was located on the opposite sur- f ace immediately opposite the crack- growth region. Two heating units were placed on the A 1/2-inch (l.27-cm) gap

27、 was provided to insure an unobstructed A control thermocouple was fixed in the expected crack path near the edge of the specimen. By using an edge control point, the temperature was found to vary k5O F (k30 K) across the specimen width. was found to vary +2O F (ELo K) during the course of the test.

28、 control was maintained in the elevated-temperature tests by a controller- recorder which regulated current flow through a saturable reactor. The con- troller operated at 208 volts using 60-cycle single-phase ac power. The temperature at a given point Temperature The equipment used in the -1090 F (1

29、950 K) tests (fig. 4) consisted of three blocks of dry ice, the same steel framework used for the furnace, and an insulating cover box. The dry ice blocks were mounted in the steel framework 4 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-L-63-9528

30、.2 Figure 3. - Elevated-temperature-test apparatus. L-63-9529. 2 Figure 4.- Cryogenic-temperature-test apparatus. and held in contact with the specimen surface in the same manner as the heating units. Test temperature was governed by the sublimation temperature of the dry ice and was found to vary l

31、ess than 5O F (30 K). The entire cooling apparatus was isolated from circulating air drafts by the insulating cover box. sublimation rate of the dry ice satisfactorily. periodically sprayed with ethyl alcohol to prevent frost buildup in the crack- growth region. This isolation was necessary in order

32、 to control the The specimen surfaces were Specimens were clamped between 3/8-inch (0.gP-cm) thick aluminum guides (ref. 5) to prevent buckling and out-of-plane vibrations in all the room- temperature tests. Guides were also used in the elevated- and cryogenic- temperature tests in which compressive

33、 loadings were applied. tests, the heating or cooling units were placed directly against the guide plates and the specimen was heated or cooled by heat conduction through the guides. In these latter Good temperature control was maintained throughout these tests. 5 I Provided by IHSNot for ResaleNo r

34、eproduction or networking permitted without license from IHS-,-,-Specimen surfaces were lubricated with light oil in the room- and cryogenic- temperature tests and with dry molybdenum disulfide in the elevated-temperature tests. One side of the guide contained a 1/2-inch (l.27-cm) cutout across its

35、width to allow visual observation of the crack path. fitted into the guide cutout to prevent buckling of the specimen. A transparent plate was Test Procedure Constant-amplitude axial-load fatigue tests were conducted at positive mean stresses of 40 ksi (276 MN/m2) for AM 350, AM 367, and Inconel 718

36、; 25 ksi (173 and 202 -T 1 (clad). area of the specimen. k30 m/m2) for AM 350, AM 367, and Inconel 718; k25 to +2 ksi (S73 to k-14 m/m2) for Ti-8U-lMo-lV; and -5 to k2 ksi (k104 to k14 MN/m2) for 2020-6, RR-58 (clad), and 2024-T81 (clad). stant throughout each test. for Ti-8U-mo-lVj and 15 ksi (104

37、MN/m2) for 2020-6, RR-58 (clad), All stresses mentioned herein refer to the original net Alternating stresses ranged from k60 to +5 ksi (+414 to Mean and alternating loads were kept con- Tests were conducted at 800 F (300 K) and -lO9O F (1950 K) on all mate- rials, at 550 F (5610 K) on the stainless

38、 steels, titanium alloys, and the superalloy, and at 2500 F (394 K) on the aluminum alloys. tested at the same stress levels at all test temperatures in order to evaluate the effect of temperature on crack propagation. Specimens were The test data were obtained by observing the crack growth through

39、30 power microscopes while illuminating the specimen with stroboscopic light. ber of cycles required to propagate the crack to each grid line was recorded so that the rate of crack propagation could be determined. Tests were termi- nated when the cracks reached a predetermined crack length, and the

40、specimens were reserved for the subsequent residual static-strength investigation reported in reference 6. The num- RESULTS AND DISCUSSION The crack-propagation-test results are presented in table IV which gives the number of cycles required to propagate a crack from a half length 0 .l5 inch (0.38 c

41、m) . The number of cycles given in table IVY and in fig- ures 5 to 12, is the mean number of cycles required to grow cracks of equal length on both sides of the central starter notch. The numbers of cycles are referenced from a half crack length of 0.15 inch (0.38 cm) because at this length the fati

42、gue crack growth is no longer influenced by the starter notch (ref. 7). a of 6 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-550 F 1561KI SOo f WQn KI -Id F 1195 KI 0 1.0 - 0.5 - .I 81 $1 .I 103 Id io7 550 F l56loKI 80 F SWo KI -IWo F ll9So KI - 25

43、 - 20 - 1.5 I I -LO ; g - 0.5 I I I 0 1.5 LO 0.5 Sa. 10 ksi (69 MN,m2) 550 F 1561KI 80 F OW KI I 1 10 Id lo5 lo7 N. cydes N. cycler Sa .5 kII (15 MNim) 550 F 1561 KI SOo F OOOo KI -109F 1195KI ,I *I I I I I 1 10 id Id N. cycler 2.5 2.0 1.5 5 1.0 $ 2 - r a5 0 I Figure 5.- Fatigue-crack-propagation cu

44、rves for Inconel 718. S, = 40 ksi (276 MN/m2). 7 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-. . . 0 I -I I I I. .I. . L I - 1 10 12 Id N. cycles L1L ,I I .I ,I .I .I lo7 1 10 Id Id N. cycles I I I I I I 1 10 12 Id lo7 N. cycles Figure 6.- Fatigu

45、e-crack-propagation curves for AM 350 (CRT). S, = 40 ksi (276 MN/m*). 8 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-5 01 I 1 I I I I 1 Sa - 20 ksi (138 kwlm2) c 1 I I I I I I 10 1 10 Id lo5 10 1 LO lo3 Id 10 N. Cycler N. cycler 1 10 10 lo3 lo5 N,

46、 cycle5 Figure 7.- Fatigue-crack-propagation curves for AM 367. S, = 40 ksi (276 M.N/m2). 9 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-sa - 15 kri (104 MNh2) - - 250F 1394 KI 80 F 13W0 KI -109 F 1195 K) - - - 0 I I I I I 1 -* 250 F 1394 KI 80 F

47、UW KI -1W F 1195“ KI 2.5 2.0 5 1.5 - - z 1.0 g - I 0.5 0 I I I ,I 81 250F 1394 K1 0.5 -109F 1195KI I I ,I I1 #I 1 10 lo3 16 N, cycles I ,I ,I I I -0 10 1 10 lo3 16 N, cycler 2.5 2.0 0.5 lo7 Figure 8.- Fatigue-crack-propagation curves for 2024481 (clad). S, = 15 ksi (104 m/m2). 10 Provided by IHSNot

48、for ResaleNo reproduction or networking permitted without license from IHS-,-,-550 F (561 KI -1WO F 1195 KI -, 2.5 sa - 15 ksi (104 MN/m2) - LO 550 F 1561 KI Sao F OW KI -id F 1195 KI -15 g 2 r - LO - 5 - r - 0.5 Figure 9.- Fatigue-crack-propagation curves for Ti-8Al-lMo-lV (duplex annealed). t = 0.050 inch (1.270 mu); = 25 ksi (173 MN/m2). 11 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-0 I I I - 0.5 - 593