Prediction of microscale plastic strain rate fields in two-phase composites subjected to an arbitrary macroscale strain rate using the materials knowledge system framework

In this work, a data-driven reduced-order model is presented to predict the microscale spatial distribution of the plastic strain rate tensor in an isotropic two-phase composite subjected to an arbitrary macroscopically imposed strain rate tensor. This model was built using the framework of localization linkages called Material Knowledge Systems (MKS), which has been demonstrated to exhibit a remarkable combination of accuracy and low computational cost. In prior work, the MKS framework was successfully used to predict the local strain rate fields in multiphase composites subjected to a selected macroscale strain rate tensor. In this work, the MKS framework is extended to include the complete set of all macroscale strain rate tensors that could be applied. This is accomplished by developing novel representations that allow a parametrization of the localization kernel over the complete space of unit symmetric traceless second-rank tensors and implementing them with the required fast computational strategies. The MKS localization linkage produced in this work was calibrated and validated to results from microscale finite element models.