Microstructural-based physics of failure models to predict fatigue reliability

This paper presents a method for predicting fatigue failure using a virtual prototyping software tool that allows the simulation of real material behavior. Computer models are used to simulate the three dimensional microstructure in which fatigue evolves. Because grain structure properties are randomly distributed through any macro-sized structure, Monte Carlo simulation is used to give a probabilistic distribution of fatigue failure outcomes over the operating life of the structure. The reliability of structural elements with complex stress distributions is predicted by integrating the fatigue simulation model within traditional structural finite element models (FEM). The reliability analyses results for all the individual elements are combined using system reliability modeling techniques to determine the fatigue reliability of the entire component. This results in prediction of fatigue reliability as cycles to failure probability or the probability of exceeding a fatigue threshold. In making the fatigue reliability predictions, only the material micrographs, stress-strain curves and a single fatigue test data point (at a single stress) were provided. The results show excellent correlation between the predicted fatigue life and experimental fatigue life