Using image-based computational modeling to study microstructure–yield correlations in metals

Abstract The potential of image-based computational modeling for determining the microstructural features that cause initial plastic flow (yield) in metals is investigated using a β-Ti alloy as the focus material. A high-fidelity three-dimensional (3-D) morphological and crystallographic description of approximately 100 metallic grains is embedded into crystal plasticity finite-element models to analyze the microscale heterogeneity of material response under a set of loading conditions applied at the mesoscale. Preliminary results show that regardless of the globally applied loading condition most of the plastic flow initially occurs generally within only one of the two grains at a grain boundary. Additionally, it was determined that initial local plastic slip is suppressed in a compliant grain neighbored by relatively stiff grains under multiple loading conditions. Overall, the analysis demonstrates that the image-based modeling framework should be used in conjunction with sophisticated data-mining tools for identification of microstructure–property correlations in polycrystalline metals provided a statistically relevant number of grains are accommodated in the models. However, smaller volumes may still be used to investigate the local effect of microstructural features on mechanical behavior.