Probabilistic fatigue life prognostic using SNPL method and crack measurement data

Abstract

Fatigue life prognostic is a major concern in many structural infrastructures including bridges, which is exposed to various sources of uncertainty. The Paris law is widely used to predict the remaining life of a structure after detecting a crack during inspection. However, the Paris law alone is not sufficient to address the entire fatigue process, because there are several stages for fatigue process. The S-N Paris law (SNPL) method was recently proposed to quantify the uncertainties lying in the Paris law parameters, by finding the best estimates of their statistical parameters from the SN curve data. Through a series of steps, the SNPL method helps determine the statistical parameters (e.g., mean and standard deviation) of the Paris law parameters that will maximize the likelihood of observing the given S-N data. In this research, the SNPL method is introduced to probabilistically predict the remaining fatigue life based on crack measurement data. Although it is assumed that only deterministic S-N curve data is available, the proposed method enables the probabilistic fatigue life prognostic of a steel bridge based on various measured crack lengths.