Fatigue Reliability Analysis Of Structural Components Based On SFEM

Fatigue failure is one of the main failure modes of actual engineering structures. It is very important in engineering to analyze the fatigue reliability under service loading. Usually, mean fatigue life is obtained from traditional nominal stress approach and local stress-strain approach, but the fatigue obtained doesn't satisfy fatigue life reliability analysis; On the other hand, it is very time-consuming and cost-wasting to determine the fatigue life depending on experiment, and the fatigue life obtained is uncertain because it is subject to many random factors. So, it is very important and more valuable to analyze the stochastic life and reliability of structural component suffered from stochastic loading and inherent dispersion.First of all, this paper reviews roundly the research status of stochastic finite element method (SFEM) in home and abroad, Its application in the reliability analysis is also presented. And then, the methods are summarized, classified and discussed in the thesis.For the fatigue life issue of medium and high cycles, the randomicity of fatigue response and fatigue analysis method is considered. The stress stochastic distribution in fatigue zone is computed by The Monte Carlo stochastic finite element method (MCSFEM). A probabilistic nominal stress approach is proposed in this paper based on Miner's fatigue damage cumulative rule, the p-S-N curves of material and stress stochastic distribution. The fatigue life distribution of structural elements under the constant amplitude loading is computed by the probabilistic nominal stress approach. Then the stochastic fatigue life is computed combined with the stochastic damage cumulative rule and reliability is calculated resorting to the residual life model under the constant and variable amplitude loading.For the fatigue life issue of low cycles, he randomicity of fatigue response, loading, cyclic stress-strain curves and strain-life curves is considered. The stochastic stress distribution is given by MCSFEM and the fatigue life obtained by probabilistic local stress-strain approach. In the end, stochastic fatigue life and reliability under the constant serial loadings and variable serial loadings is given.