An Investigation Into the Robustness of a Crystal Plasticity Finite Element Model
Abstract A crystal plasticity model has been developed for describing the plastic and viscoplastic behaviour of 316H stainless steel. The model has been used successfully to predict the macroscopic response of the material for monotonic and cyclic loading, however the robustness of the model is now being scrutinised to ensure that it captures the underlying mechanisms and local meso-scale deformation characteristics correctly. In order to look at this in more detail, the model has been scripted to allow simulation of diffraction studies on the grains. This is being used to compare the simulation output with neutron and synchrotron experiments. A theoretical study has been completed which adjusts the values of each material parameter within the crystal plasticity finite element (CPFE) framework in isolation to analyse the effect each has upon the shape of the hysteresis loop and how this relates to predictions. A further study had been conducted to investigate the amount of scatter that is produced by altering the initial microstructure of a relatively small volume. The results show that changing the initial microstructure has a negligible effect on the subsequent stress-strain response. This indicates the influence that the grain morphology will have upon the diffraction measurements and that there is no need to consider this further when testing small specimens. To conclude this investigation two further aspects of the model have being scrutinised; the effects of constraining boundary conditions and altering the local environment of a single grain, again to explore the influences which these may have on diffraction studies. The plane boundaries of the volume are currently fixed at zero displacement which will influence the local grains on these boundaries as they are over constrained, but should not alter the overall stress-strain response. To investigate this, grains within the model that have boundary conditions applied have be discarded from the final results to leave the response from the central volume that should be free of the boundary condition effects. The results show that the macroscopic response of the bulk volume and the central volume are very similar but when the individual grain family responses are analysed, it can be seen that the boundary condition do alter the response of the representative volume element (RVE). To analyse how the local environment surrounding a single grain affects its stress-strain response, a centrally located grain will be selected and kept constant while the orientation of the surrounding grains are randomly altered. It has been found that changing the local environment surrounding a single grain will impact the stress-strain response seen by that grain.