An Experimental Investigation on Fatigue Damage by Means of Multiple Repeated Loads

Abstract

Fatigue tests with repeated loads of two steps and sequential multiple steps were made on 0.35% carbon steel specimens by means of a rotary bending fatigue tester of Ono's type. The specimens were normalized in blank before machining at the temperature of 870°C in 1 hour, with subsequent air-cooling. The mechanical properties of the material are given as follows: tensile strength σ_B=56.6kg/mm², lower yield point σ_S=32.2kg/mm², elongation φ=27.3% and the fatigue limit σ_wb=28.1kg/mm².
In the test of two steps loading, four kinds of loading, namely σ₁_??_σ₂=30_??_37.7kg/mm² (beyond the yield point) and 28.9_??_31.8kg/mm² (beneath the yield point) were applied in various cycle ratios, to investigate the aspects of fatigue damage. In the test of multiple steps loading, sequentially ascending and descending loadings of equal 2, 4 and 8 steps in equal cycle ratios were applied in the same range of stresses as in the case of the two steps loading, and the variations of the endurance life were measured.
The test results can be briefly summarized as follows. When the sequential loading is an ascending type, the total sum of cycle ratios up to fatigue fracture is larger than unity, ascending up to 1.3 in average, while the sequential loading is a descending type, the total sum of cycle ratios is smaller than unity, descending down to 0.7 in average. The value of the sum of cycle ratios is almost independent of the number of steps in sequential loadings, whether the stress amplitudes exceed the yield point or not. While scattering of the test results was more remarkable in ascending loadings than in descending loadings. The test results were compared to those of the theoretical prediction by D. L. Henry, and the reason for the scattering of the test results in ascending loadings and some other considerations are discussed.