A growing number of computational material science and computational mechanics research is currently devoted to the explicit modeling of microstructures at various length and time scales. The finite element models of grains and grain boundaries in polycrystals include discretization of the grain interior. In addition, grain boundaries are explicitly discretized as cohesive zones with appropriate damage properties to facilitate the simulation of intergranular cracking. Such finite element models may easily involve hundreds of grains and millions of finite elements. They may also be combined with advanced lattice orientation dependent constitutive models, such as for example anisotropic elasticity and crystal plasticity. The complexity of the model, including the random lattice orientations, may therefore represent a serious difficulty in detecting possible issues in the finite element model and the interpretation of the results. A number of self-consistency model-checks are therefore needed to verify the model. Two tests are proposed and demonstrated in the paper. The first is aiming at the assessment of the finite element mesh quality within the grains in terms of the results. The second is primarily aiming at the verification of the consistent modeling of the cohesive layer at the grain boundaries. In addition, some useful information about the finite element mesh quality in terms of results is also given.

This content is only available via PDF.
You do not currently have access to this content.