Extreme value statistical analysis to determine the endurance limit of a 1045 induction hardened steel alloy

Surface hardened components are used in fatigue critical applications such as axles and gears. Inclusions are critical microstructural features where fatigue cracks have been observed to nucleate in these parts. In this investigation, the effect of inclusion populations on fatigue performance of induction hardened 1045 steel was examined. The steel was heat treated to have a tempered martensite starting microstructure and was induction hardened to two different case depths. Utilizing extreme value statistical analysis, the largest inclusion as well as the largest inclusion in each of five categories, MnS, MnCaS, MnAlS, Al2O3, and Al2O3-MgO, was estimated for a critically stressed area in a fully reversed cantilever bending fatigue sample. The inclusion size estimates were used to predict the endurance limit of the sample with a fracture mechanics-based model. This methodology has been traditionally used for homogeneous materials but has been modified here for bending fatigue and inhomogeneous case hardened material. The predicted endurance limits are closely correlated to experimentally measured endurance limits.