1、-Z_ were obtained intests conducted with the Hi-Lo sequence and the lowest values wereobtained in tests conducted with the Lo-Hi sequence. The Random,Lo-Hi-Lo,In with someand Hi-Lo-Hi sequences resulted in intermediate values oftendency for results of Lo-Hi-Lo and Hi-Lo-Hi tests to be higher than th
2、eresults of tests with the Random sequence. The summary of the statisticalanalysis presented in table VIII(a) indicates that the foregoing observa-tions were supported in the great majority of cases. For example, the_n for the tests with Lo-Hi sequence are shown to be sig-values ofnificantly lower t
3、han those for other sequences in all cases where com-In for tests withparisons were possible. Similarly, the values ofthe Hi-Lo sequence were significantly higher than others in all but two72ncases. Differences between _ for Random sequence and for Hi-Lo-Hior Lo-Hi-Lo were smaller and significant in
4、 only two of seven possiblecases.If the random sequence may be regarded as being most nearly repre-sentative of service experience, it would appear that considerable biaswas introduced in the tests with Lo-Hi and Hi-Lo sequences of loading.Additional work is needed to determine which of the sequence
5、s used inthis investigation reproduces service loadings most faithfully.One possible explanation for high values of life in Hi-Ix) sequencetests is that beneficial residual compressive stresses were produced byapplication of high tensile loadings early in the test.L798Effect of Mean StressValues of
6、_ were consistently higher for tests with high valuesof mean stress than for tests with lower values. This observation issupported strongly by the statistical analysis (Gable VIII(b) in allcases where comparisons were possible between tests with 0 and 17.4 ksifor 2024-T3 and between tests with 0 and
7、 20 ksi mean stress for 7075-T6.Differences between results of tests with mean stresses of 0 and lO ksiwere not statistically significant.The probable reason for this behavior is that beneficial residualstresses produced by tensile loadings are effective when the mean stressProvided by IHSNot for Re
8、saleNo reproduction or networking permitted without license from IHS-,-,-iiis positive, but are canceled by compression loads when the meanstressis zero (ref. 9). The tests with Smean= i0 ksi are probably marginalas regards their ability to show an effect.L798MaterialL n computed for results of test
9、s of 7075-T6 specimensValues ofwere consistently higher than those for corresponding tests of 2024-T3specimens; the statistical analysis indicated that these differenceswere significant in all but two cases. (See table VIII(c).) This con-clusion is probably less significant than is indicated by the
10、statistical_n are based upon S-N curves foranalysis because the calculations oftwo different materials.Number of Cycles Per BlockComparisons to evaluate the effect of block size on _ are possi-ble in only six cases (table VIII(d). The results of these comparisonsare quite inconsistent. In three of t
11、he six cases the statistical treat-ment indicated that smaller blocks produced significantly higher valuesof _ than did larger blocks; in the other three cases the effectsg.p._ -were not significantly different. It should be pointed out that blocksizes in the present investigation were varied within
12、 a limited range.It is possible that effects would be found if much larger blocks wereused. Considerably more data would be required to establish whethervariations in block size in a practical ranga_produce a significant effecton specimen life.Number of Stress StepsThe data are not sufficient to eva
13、luate the effect of the number ofstress steps used to simulate the stress spectrum. Table VIII(e) presentsconflicting data on this point.Shape of Stress Frequency SpectrumThe data are insufficient to allow critical evaluation of the shapeof the stress frequency spectrum. The one comparison shown in
14、table VIII(f)is inconclusive; this is not surprising since other data (ref. !0) indicateProvided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-12no consistent effect even for spectra which differ muchmore than thespectra used in the present investigation.S-
15、N Curve ReliabilityAs noted previously, the S-N curves were refaired as indicated infigure 9 after additional constant-amplitude test results were obtainedin the present investigation. Since the value of fatigue life N istaken from the S-N curve the cycle ratios, n/N, were greatly affectedby this re
16、vision in the S-N curves. In somecases individual stressband cycle ratios changedmore than !00 percent and the sumof the cycle_nratios _ for the load schedule as much as 35 percent. The resultspresented in tables VI and VII and figures 7 and 8 were obtained usingthe revised S-N curves, but the S-N c
17、urves taken from references 6 and 7were used in the numerical-integration process for determining the testschedules. It was found that by using the revised S-N curves the numeri-cal integration would have yielded different values of stresses to repre-sent the stress bands in some cases. The precedin
18、g discussion clearlyshows the necessity for establishing_,_a representative S-N curve, partic-ularly if the values of _ obtained from tests on similar specimenconfigurations are to be used for comparative purposes. Failure to estab-lish a reasonably reliable S-N curve may be responsible for some of
19、the_nlarge variations in _ or other life index used in discussing resultsof variable-amplitude fatigue tests in the literature.L798Stress at FailureFor 7075-T6 specimens the stress step in which specimen failureoccurred had a definite pattern. For Smean = 0 ksi most of the spec-imens failed at the h
20、igher stresses while for Smean = 20 ksi most ofthe specimens failed at the lower stresses. (See fig. I0.) The patternwas not so clear for 2024-T3 specimens_ however_ fewer tests with eightsteps were made. The sequence in which the loads were applied did notseem to affect the load at failure.CONCLUSI
21、ONSThe results of programmed variable-amplitude axial-load fatiguetests of 2024-T3 and 7075-T6 aluminum-alloy sheet specimens support thefollowing conclusions:Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-13L798Ini. The values of the summation of c
22、ycle ratios _ for 7075-T6specimens were consistently higher than the values obtained for 2024-T5specimens.2. The sequence in which the loads are applied has a marked effecton the life of the specimen, with sequences involving progressivelyincreasing stresses within each block giving the shortest lif
23、e andsequences involving progressively decreasing stresses in each blockgiving the longest life.3. The mean stress at which the specimen is tested has an effect on_, with the life increasing as the mean stress is increased.4. A reasonable change in the fairing of the S-N curve produces anappreciable
24、 change in the value of _.5. Additional data are needed to establish effects due to the numberof load cycles per block, the number of load steps used to approximatethe load spectrum, and changes in the load spectrum.Langley Research Center,National Aeronautics and Space Administration,Langley Field,
25、 Va., October 6, 1959.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-14REFERENCESI. Miner, Milton A.: Cumulative Damagein Fatigue. Jour. Appl. Mech.,vol. 12, no. 3, Sept. 1945, pP. A-159 - A-164.2. Grover, H. J., Bishop, S. M., and Jackson, L. R.: F
26、atigue Strengthsof Aircraft Materials. Axial-Load Fatigue Tests on Unnotched SheetSpecimensof 24S-T3 and 75S-T6AluminumAiloys and of SAE4130 Steel.NACATN 2324, 1951.3. Grover, H. J., Hyler, W. S., Kuhn, Paul, Landers, Charles B., andHowell, F. M.: Axial-Load Fatigue Properties of 24S-T and 75S-TAlum
27、inumAlloy as Determined in Several Laboratories. NACARep. 1190,1954. (Supersedes NACATN 2928.)4. Spaulding, E. H.: Design for Fatigue. SAETrans., vol. 62, 1954,pp. 104-116.5. Rhode, Richard V., and Donely, Philip: Frequency of Occurrence ofAtmospheric Gusts and of Related Loads on Airplane Structure
28、s.NACAWRL-121, 1944. (Formerly NACAARRL4121.)6. lllg, Walter: Fatigue Tests on Notched and Unnotched Sheet Specimensof 2024-T3 and 7075-T6 AluminumAlloys and of SAE4130 Steel WithSpecial Consideration of the Life RangeFrom 2 to i0,000 Cycles.NACATN 3866, 1956.7. Grover, H. J., Bishop, S. M., and Jac
29、kson, L. R.: Fatigue Strengthsof Aircraft Materials. Axial-Load Fatigue Tests on Notched SheetSpecimensof 24S-T3 and 75S-T6AluminumAlloys and of SAE4130 SteelWith Stress-Concentration Factors of 2.0 and 4.0. NACATN 2389,1951.8. Anon.: A Tentative Guide for Fatigue Testing and the StatisticalAnalysis
30、 of Fatigue Data. Special Tech. Pub. No. 91-A, ASTM,1998.9. Anon.: Discussion in NewYork by Herbert F. Hardrath (Langley Field,Va.). Proc. Int. Conf. on Fatigue of Metals (London and NewYork),Inst. Mech. Eng. and A.S.M.E., 19563p. 830.I0. Hardrath_ Herbert F. 3 Utley, Elmer C., and Guthri% David E.:
31、 Rotating-BeamFatigue Tests of Notched and Unnotched 7075-T6 Aluminum-AlloySpecimensUnder Stresses of Constant and Varying Amplitudes. NASATND-2103 1959.L798Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-15L798TABLE I.- TENSILE PROPERTIES OF ALUMINU
32、M-ALLOY MATERIALS _STEDAv. Min. Max.7075-T6 (152 tests):Yield stress (0.2 percent offset), ksi . 75.50 71.54 79.79Ultimate tensile strength, ksi . 82.94 79.84 84.54Total elongation (2-inch gage length),percent 12.3 7.0 15.02024-T3 (147 tests):Yield stress (0.2 percent offset), ksi 52.05 46.88Ultimat
33、e tensile strength, ksi . 72.14 70.27Total elongation (2-1nch gage length),percent . 21.6 15.0 25.0Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-16TABLE II.- VARIABLE-AMPLITUDE LOADING SCHEDULES FOR 2024-T5 ALUMINUM-ALLOY SPECIMENSStepSmeanSmax_ n
34、n/N,ksl per step= 17.4 ksi; gust frequency curve B1 18.1 62,0002 19.5 24,0003 20.9 9,4004 22. 3 3,4005 23.7 88O6 25.1 2207 26.5 608 27.9 269 29.3 7.8i0 30.7 3.2ii 32-1 1.812 33.5 .5i3 34.9 .34i4 36.3 .1615 37.7 .08i6 39.1 .05417 40.5 .02418 41.9 .012Smean_alO0,O0000.000235.002618.003260.002296.oo135
35、o.001000.000520.000348.000295.000119.00o117.000070.000052.000047.000025.o00o1_0.012367= 17.4 ksi; gust frequency curve A19.5 82,00022.5 15,00025.6 2,80028.7 35031.9 4635.1 7.438.4 1.641.5 .35ZblO0,2050.016660.038360.018940.oo7o8o.OO268O.001208.000466_0.085394Step Smaxksl n n/N,per stepSmean = 17.4 k
36、si; gust frequency curve Ai 18.12 19.53 20.94 22.35 23.76 25.17 26-58 27.99 29.3lO 30.7ii 32.112 33- 513 34.914 36.315 37.716 39.117 40.518 41-9Smea_ = 0;2.28.013.218.523.8e9.234.840.446,80027,20014,_06,8002,7501,12049O2008039187.63.01.3.6.3.25.09ZI00,OO000.0oo363.005230.010160.Oli68o.010650.o07700.
37、oo533o.004240.002918.001810.001035.000634.000387.000261.0OO153.00O113_0.062664gust frequency curve A41,0007,850980143233.73.ii_c50,0000.001869.o196oo.014300.012105.008824.004294.001896Z0.062888I.-.1_0CoaTests were also made at 5n.bTests were also made at n/2.CTests were also made at 2n.Provided by I
38、HSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-3Z17TABLE III.- VARIABLE-AMPLITUDE LOADING SCHEDULESFOR 7075-T6 ALUMINUM-AI/)Y SPECIMENSL798Step Smax n n/N,ksi per stepSmean = 20 ksi; gust frequency curve Ai234567821.525.328.732.636.34o.145.947.542,0007,500I,1901
39、75232.5.5.i_a50,9000.046875.071687.050172.007931.001678.000610.000208_o.159161Smean = i0 ksi; gust frequency curve Ai23456781317.121.927.131.736.842.547.024,4004,700i,00092141.8.3.0730,2000.oo9o38.o50ooo.02o9o9.008485.002687.000231.000121_0.091471lSmean = O; gust frequency curve Ai23456783.89.115.02
40、1.227.233.739.946.324,4004,80069098141.8.33.074_50,0000.002087.025091.023333.007778.005625.002538.001276_0.067728aTests were also made at n/5.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-18TABLE IV.- RESULTS OF CONSTANT-AMPLITUDE AXIAl-LOAD FATIGU
41、E TESTS OF2024-T3 ALL_4INUM-ALLOY SPECIMENS WITH KT = 4.0 EDGE NOTCHksi cycles ksi cyclesSmean = 0 ksiA36N2-2 iA26N2-5 34.5A29N2-1A29N2-2A29N2-8 18A29N2-3 1A29N2-4A26N2-1A29N 2-ioA2TN2-5 12A26N2-4A29N2-7A39SI-IOA43Sl-3A$2Sl-7A39SI-2A$3Sl-6A39SI-6A39SI-4A39SI-7A39SI-IA43SI-8A_4SI-6A42SI-IA44Sl-7A39Sl
42、-3A44SI-I0A39SI-5A43Sl-4A43SI-9A40SI-3A39SI-8A43Sl-10A40Sl-8A59SI-9A_3Sl-5A47SI-IA47SI-10A46SI-6A46SI-10A47Sl-717414615,00014,O0013,00012,0008,000233,000211,000178,000126,000113,000A29N2-5A29N2-9A29N2-6A26N2-7A2TN2-6A2TN2-7A27N2-8A26N2-6A36N2- iA27N2-10A36N2-47.25Smean = 17.4 ksi572557534529475825i,
43、1571,0597977937227621,4703,2522,4421,897i,691i,3803,56815,89312,4739,8336,6415,54414,00014,CO013,00012,000ii,000A42SI-2A42SI-6A38SI-6A42S1-4A42Sl-5A47Sl-6A47Sl-5A48SI-2A48Sl-8A48S1-10A41SI-7A41Sl-3A41SI-10A41SI-8A46Sl-7A46SI-5A41S1-5A47SI-9A41SI-9A38SI-IA38S1-2A41S1-4A37Sl-lOA38S1-9A48S1-6A57Sl-7AJT
44、Sl-9A37Sl-3A37Sl-8A45Sl-8A45Sl-6A37Sl-4A37SZ-1A45Sl-3A46SI-227.525I22.52221716,000300,000197,0006,284,COO4,609,0004,488,0003,663,0003,361,0005,518,00017,039,0004,148,00073,00071,00029,00026,00023,00064,00062,00058,00057,00049,0004,810,000346,000274,000231,000155,00047,106,00023,345,00016,479,00010,2
45、04,0009,817,0006,167,0005,496,000551,000255,000214,000169,709,000146,780,000141,040,000112,905,000i10,430,00070,568,00053,157,00052,381,00014,952,000605,000t_IMC(XProvided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-19TABLE V.- RESULTS OF CONSTANT-AMPLITU
46、DE AXIAL-LOAD FATIGUE TESTS OF7075-T6 ALK_4INL_-ALLOY SPECIMENS WITH KT = 4.0 EDGE NOTCHcO_!oc onI lls ec on Ifoksl cycles ksl cycles=_Smean = 0 kslB55NI-5B55NI-4 50B59NI-10 40B59NI-9B60NI-4 )B60NI-2 30BI29SI-8B59NI-8 ) 20B6ON1-5B6ONI-1B55NI-I _ 15_oONi-3 #4O36136iN)9178656546,0006,0o035,00030,00018
47、, ooo_meanB55N1-6B6Oni-7B88NI-IB59NI-6B59_1-2B59NI-5B59_I-4B59NI-3B88NI-IOB88NI- 7B88NI-912IO149,000130,00095,000i,292,000673,000532, 000456,000510,0003,874,0003,309,0002,290,000= i0 kslB88N1-2B29NI-4B88NI-4B27NI-6B27N1-7B27NI-2B27N1- lOB27NI-9B55NI-9B27N!-8B55Ni- 5B88N1-3)i135O 928440 44037438.2 45
48、535 95525 6,82557,82020 32, 99029,00022,520B60NI-9B59NI-7B60Ni-i0B60NI-6B60NI-8B5TNI-2B57NI-3B57NI-9B55Nl-IOB88N1-6B55N1- 2B27N1-5I 1815i_i06,000162,00052,00045,00042,0002,241, 0002,102,000i,093,00024,204,00013,877,0008,247,000I0,000,000Smean = 20 ksiB55NI-7B55NI-8BSTNI-IB57NI- I0B88N1- 5B129SI-6B56
49、NI-8B129SI-9B129Sl-10BI29Sl-5)5o3O2536350913,000lO, 8009,000674, 000335,000120,000112,00092,000B129S1-4B129S1-7Bi28Sl-1B56_1-9B58N1-9B58Nl-lOB57_-5B57N1-8B57NI7B57NI-4I 2524.581,00075,00063,00042, 0009,648, ooo5,875, o00176,00044,606,00018, 575,0008,355,000Provided by IHSNot for ResaleNo reproduction or networkin
