Toward an understanding of local variability of fatigue strength with microstructures

A probabilistic approach is taken to provide a means to predict variability in fatigue properties given a measured or predicted distribution of a critical microstructural feature. Under the assumption that local fatigue life is a monotonically decreasing function of the size of the local microstructural feature responsible for fatigue crack initiation we obtain analytically the cumulative fatigue strength distribution, cumulative life distribution and their corresponding distribution density functions. The theory has been used to study the variability of cast aluminum alloys, which are widely used in the automotive industry for engine components. In this case, the mean fatigue strength and its variation are based on the distribution of micropores that are ubiquitous in cast aluminum. The impact of the size of the highly stressed area, represented by local surface area S and the local density of the microstructural feature is found to have a critical role on the fatigue strength distribution. The influence of other parameters, such as mean and standard deviation of the distribution of the microstructural feature, on fatigue distributions is discussed.