Addressing biases in spectral databases for increasing accuracy and computational efficiency of crystal plasticity computations

Spectral databases of crystal plasticity solutions stored as discrete Fourier transforms (DFTs) have now been used for over a decade to speed-up the simulations. These approaches demonstrated impressive computational advantages conventional numerical that explicitly solve governing constitutive equations. However, their implementation in finite element models (FEM) referred spectral method (SCPFEM) framework has encountered significant hurdles mainly related instabilities jacobian calculations leading convergence issues solving equilibrium In this work, these main are tracked certain biases introduced formulation databases. choices made selection independent diagonal components deviatoric stress tensor and ordering eigenvalues plastic stretching tensor. Specific strategies designed implemented address such as, generating separate DFT database all three tensor, standardizing assignment among other improvements. improvements shown improve markedly accuracy both stress-strain responses well jacobians needed computations. The benefits improved with selected case studies.