A recently developed high Mn austenitic steel damper was found to show an excellent low-cycle fatigue resistance, irrespective of specimen size. To explain the robustness against specimen size, this study proposes a probabilistic prediction method for the effect of specimen size on fatigue life. In the present study, a Fe-30Mn-4Si-2Al model austenitic alloy that shows a superior low-cycle fatigue property was used as a case study. The prediction method is based on the measurement results of the rate of fatigue subcrack propagations, which are separated into two parts: small and large fatigue crack propagations. An important aspect of the prediction is the scatter characteristics of the small and large fatigue crack propagations; the former showed a large scatter in the propagation rate data, while the latter did not show large scatter data. Based on the scatter, two characteristic crack lengths were introduced in the probabilistic analyses: a) a boundary, l0, for small and large fatigue crack propagations, and b) a critical fatigue crack length for fracture, lf. Moreover, the effect of specimen size on the small fatigue crack propagation scatter was estimated using a Poisson distribution. The effect of specimen size on the large fatigue crack propagation part was estimated by an exponential approximation of the relationship between fatigue crack length and number of cycles after the l0. In addition, the boundary condition between the fatigue crack propagation and the fracture was determined using the critical fatigue crack length for fracture obtained from a fractograph. According to our prediction model, the small fatigue crack propagation lives decreased with an increase in the specimen size, and the large crack propagation lives increased with an increase in the specimen size.
All Science Journal Classification (ASJC) codes
- Materials Science(all)