LAUSR

Abstract Polycrystals heterogeneities generate stress concentration which can be responsible for premature failure of mechanical parts and thereby need to be considered in mechanical models. A grain’s close neighborhood has shown to significantly impact its mechanical behavior. The more the crystal is anisotropic, the more this impact can be, which in some specific configuration can double a grain’s stress level than it would normally have in an “average” random environment. The impact of such stress concentration on the elastoplastic behavior of polycrystalline material has been studied by means of a cellular automaton (CA) and finite element (FE) models. The studied aggregates were single-phase with grains of identical size and spherical shape. The grains’ anisotropy and crystallographic orientations were the only sources of heterogeneities studied. The CA model, originally developed for the study of elastic loadings, was adapted for the study of elastoplastic loadings in high cycle fatigue regime and compared to the FE model. Using the CA model, a statistical study was carried out to determine the true elastic limit probability distribution due to the random character of the neighborhood effect. The obtained probability distribution could be linked to the fatigue test scatter observed experimentally.