An Effect of Load Variation on High-cycle Fatigue Behavior and Fatigue Life

点击翻译按钮获取中文摘要

In recent years several fatigue studies have shown that fatigue failures can occur even at the stress amplitude below the conventional fatigue limit in the ultrahigh cycle ( N > 10^7) or gigacycle fatigue regime, and S-N curves indicate a multi stage shape depending on the failure mechanisms, i.e. failure originated from specimen surface or internal defect.S-N curves in a high carbon chromium steel obtained by rotating bending test revealed that the fatigue data for the internal defect induced fracture (fish eye type fracture) were different from those for surface slip induced fracture. It was found that the S-N curve consisted of two straight lines, one corresponding to shorter life due to fracture from surface, which shows the fatigue limit, and another to ultra-long life due to fracture from internal inclusions. Repeated two step loading tests were conducted on the condition that the high level stress is above the conventional fatigue limit and the low level stress below the fatigue … More limit obtained form the data for fracture from surface. Cumulative fatigue damage values based on the Miner rule, D, where the low level stress does not result in fatigue damage, represent the values larger than unity for fish-eye type fracture indicating the conventional S-N curve give an over estimation of fatigue damage, while that for surface fracture gives a smaller value .than unity indicating that the stress below fatigue limit may give a fatigue damage. Fatigue damages based on the modified Miner rule, D, which was obtained by extrapolating the S-N curve below the fatigue limit for surface fracture or by using the data for fish-eye type fracture are discussed, D for surface fracture shows a conservative value, but D for fish-eye type fracture shows fairly small values, suggesting that the effect of the high level stress on the crack initiation from internal inclusion becomes more detrimental by the load variation.S-N curves obtained by a push-pull type fatigue test in Cr-Mo steel, which were conducted with the cycle rate at 1kHz. In this case no fish-eye type fracture was observed and S-N curve showed the fatigue limit in the region over 10^8 cycles. Repeated two step tests were conducted by changing the low stress level. Fatigue damage due to the lower level stress was clearly recognized and cumulative damage values based on the Miner rule become less than unity. Modified S-N curve was found to lie below the S-N curve for constant amplitude test and to be extended to the stress level far below the fatigue limit. Less