Fatigue Analysis for Structures under Stochastic Loading

Predicting the peak distribution of stochastic stress responses is critical for various fatigue analysis problems with stochastic uncertainties. In this paper, a method, based on the first-order reliability method (FORM), is proposed for estimating the peak distribution of stochastic response. The method linearizes the stochastic process at the associated most probable point. After linearization, the non-Gaussian, nonlinear stochastic response is transformed into an equivalent standard Gaussian process. The peak distribution is then approximated using the rice’s formula by considering the correlation of the stochastic process at time instants. The method is capable of dealing with problems with both random variables and stochastic processes. A fatigue analysis example of the hydrokinetic turbine blades is adopted to demonstrate the proposed method. The results show that the solutions obtained from the proposed method are close to their counterparts from Monte Carlo simulations (MCS).