Study of methods for evaluating the probability of multiple site damage occurrences

Abstract

Two methods currently available for evaluating the probability of Multiple Site Damage (MSD) occurrence were studied in this paper. One of the methods is a probabilistic analysis approach based on the statistical theory and fatigue characteristics of each structural detail, and the other is an approach which defines the initial damage scenario by means of Monte-Carlo simulation, and multiple initial crack scenarios are randomly generated. A modified method based on the Monte-Carlo simulation was proposed in this paper, in which the random fluctuation of the stress was considered to give more accurate evaluation results. In the presented method, the probability of MSD occurrence in a structural element containing multiple details was calculated based on the Monte-Carlo simulation and the p-S-N curve of a single structural detail. Fatigue tests were accomplished using specimens containing 21-similar-details to obtain the fatigue life corresponding to MSD occurrence. Tests on single-detail specimens and static calibration tests were also conducted to get the basic fatigue properties of the material and the degree of stress fluctuation. The aforementioned three methods were compared and validated via the test results. The influence of the stress random fluctuation degree on the probability of MSD occurrence and influence of the distribution types on evaluating the MSD occurrence probability were discussed. Multiple Site Damage (MSD), fatigue crack initiation, probability, stress fluctuation