A general method for determining probabilistic S-N curve based on static strength data of material

Abstract

The conventional group test method for obtaining probabilistic S- N curve is a time-consuming and expensive test program. This study aims to present an alternative method for determining the probabilistic S- N curve of material based on its static strength data. Firstly, the probabilistic mapping relationship between the static strength and fatigue life of material is elaborated. Subsequently, the Weibull distribution is utilized to model the static strength and fatigue life data of material, and the correlation between the distribution parameters of them is investigated. Finally, an alternative method for determining the probabilistic S- N curve of material is proposed, and the experimental data of carbon steels and composite laminates are applied to verify its validity. The results show that the proposed method is capable of determining the probabilistic S- N curve of material based on its static strength data, which can significantly save test time and reduce test cost. In engineering applications, when the parameters of static strength distribution and the median S- N curve are known, the probabilistic S- N curve with any given survival probability can be determined through a unified analytical expression.

Fatigue life prediction of eccentric springs for the automobiles considering residual stress

“…A general program was proposed to predict the high and low cycle fatigue life of steam turbine shaft 14 . And some scholars have also put forward some new methods to solve the fatigue problem 15–17 . A microscopic observation method of in situ damage was proposed 18 .…”

“…14 And some scholars have also put forward some new methods to solve the fatigue problem. [15][16][17] A microscopic observation method of in situ damage was proposed. 18 It had a high resolution in size and time and could reliably measure the propagation rate of small cracks.…”

Research on fatigue life evaluation method of shaft parts based on small sample

“…Buar et al (2006) established an appropriate cumulative damage model based on a multilayer perceptron to investigate the effects of stress amplitude, notch factor and temperature on fatigue life under constant loading conditions. Gao et al (2022) determined the P-S-N curve of the material based on its static strength data and established a relationship between the static strength and the probability of fatigue failure. Then, a new method is proposed to determine the P-S-N curve, which reduced the testing time and cost.…”

“…Researchers have proposed several methods to solve the inefficiency of simulation methods based on variance reduction, such as Markov chain Monte Carlo sampling, Sequential Importance Sampling and Bootstrap. But due to complex process for application of the Markov chain Monte Carlo sampling method, the practicality of specific method also should be considered (Zhu et al, 2022). Therefore, it is vitally important to develop a reliable data augmentation method that can greatly save costs and should be enhanced for engineering problems.…”