Stochastic approach for remaining fatigue life prediction considering parametric distributions

Abstract

AbstractIn the present work, a stochastic framework, which combines the well‐recognized linear damage rule (LDR) and the uncertainty of material properties under constant amplitude loading, is proposed to improve the accuracy in prediction of remaining fatigue life. The uncertainty quantification approach based on parametric statistics is employed together with the Monte‐Carlo simulation method for modeling the randomness of remaining fatigue life. The impact of several commonly‐used parametric statistics is investigated, and the pertinent one is identified utilizing the Kullback‐Leibler divergence. The proposed framework is validated through the use of experimental fatigue life data under two‐step loading test.