Data Science-coupled Multiscale Mechanics for Understanding Fatigue Behaviour of 3D Printed Ti-6Al-4V Alloy

Fatigue is a multiscale phenomenon, with the fingerprints of crack initiation and growth span over microscale to macroscale. A comprehensive understanding of fatigue behaviour requires bridging analyses across these length scales. This study aims to establish such a linkage by employing a rate-dependent crystal plasticity model for mesoscale analysis and the Hartman-Schijve equation for crack growth. Experimental data from low- and high-cycle fatigue tests are used for validation. Extreme value statistics and Bayesian inference are applied to correlate mesoscale parameters with fatigue life and crack growth rate. The results show good agreement between simulation outputs and experimental trends, enhancing fatigue initiation and life prediction for additively manufactured Ti-6Al-4V alloys.