Predicting plastic anisotropy using crystal plasticity and Bayesian neural network surrogate models

This work presents an efficient data-driven protocol to accurately predict plastic anisotropy from initial crystallographic texture. In this work, we integrated feed forward neural networks with Variational Bayesian Inference techniques establish accurate low-computational cost surrogate model capable of predicting the anisotropic constants based on texture polycrystalline material quantifiable uncertainty. The developed was trained results 54,480 crystal plasticity simulations. performed simulations parametrized Hill’s yield for single crystals and textures, which were robustly represented using generalized spherical harmonics (GSH). Subsequently, GSH-based representation different textures linked its corresponding coefficients a variational network. efficacy accuracy critically validated 20,000 new textures. predictions network showed excellent agreement obtained experiments high-fidelity finite element