Research on fatigue life evaluation method of shafts based on small sample P–S–N

Abstract

In view of the fact that there are few fatigue data in practical engineering, Weibull distribution is introduced to propose a fatigue life prediction method considering failure probability based on small samples. Firstly, a modified bootstrap sampling method is given by introducing random numbers between (0–1), and the small sample fatigue life data is expanded to obtain the correlation factor of Weibull distribution. Secondly, the relationship between equivalent stress and characteristic life is given by exploring the law between them. Meanwhile, combining with the Osgood equation and using the Weibull distribution to describe the fatigue life, an iterative procedure is established to obtain the predicted life, and a first-order reliability factor is introduced to improve the reliability of the iterative procedure to modify the iterative procedure. Finally, the P–S–N curves of different failure probabilities are given, and Cast iron and TC4 alloy are selected to evaluate the correctness and reliability of the proposed method in this study. The comparison results show that the predicted life obtained by the proposed method is within the double dispersion band, and possesses high accuracy. At the same time, the P–S–N given by the proposed method in this paper under different failure probabilities conforms to the actual law, which can provide a theoretical basis for solving practical engineering problems.