Role of maximum grain size in fatigue life and its incorporation into strain-life modeling

Traditional fatigue life models rely on average grain size, which cannot adequately capture the effects of heterogeneous grain size distributions. In this study, 304 austenitic stainless steel specimens with different grain size distributions were produced via rolling. Despite comparable average grain sizes, significant differences in fatigue life are observed, which are associated with variations in the maximum grain size within the distribution tail. To address this limitation, a grain size heterogeneity correction factor based on the maximum grain size is incorporated into the conventional strain-life model. The proposed model improves prediction accuracy with minimal additional parameters. The results indicate that extreme grain size distribution features are associated with fatigue life variability, providing a microstructure-informed approach for fatigue life prediction.