Full-field polycrystal plasticity simulations of neutron-irradiated austenitic stainless steel: A comparison between FE and FFT-based approaches

Abstract

We compare two full-field approaches – a crystal plasticity finite element method (CP-FEM) and crystal plasticity fast Fourier transform-based (CP-FFT) method – for a specific crystal plasticity law introduced for neutron-irradiated austenitic stainless steel SA304L currently used in nuclear reactor vessel internals. This particular law is employed to identify and quantify possible advantages and drawbacks of the two approaches when used in the large-scale simulations to predict the effect of irradiation damage (e.g., crack initiation) in stainless steel microstructures. A comparison is performed in a polycrystalline context for different periodic Voronoi microstructures deformed under tension. Special emphasis is put on studying the performance of the two approaches in terms of mesh convergence analysis using aggregate models with different spatial discretizations. In the CP-FEM approach, the performance of linear as well as quadratic tetrahedral meshes is investigated. A similar performance between the CP-FEM and CP-FFT methods is demonstrated in a smaller 2-grain aggregate. However, a slower mesh convergence is observed for the CP-FEM method when comparing tensile responses of a larger 100-grain polycrystal. A drop in the convergence rate is much more pronounced on linear than on quadratic tetrahedral meshes. As a consequence, largest (average) grain boundary stresses are shown to be overestimated with the linear mesh CP-FEM approach, thus raising a concern of possible over-conservatism employed in the CP-FEM prediction of crack initiation in irradiated stainless steels. On the contrary, the mesh convergence of the CP-FFT approach is found to be practically independent of the applied macroscopic strain and also aggregate size. Therefore, for such steels, the CP-FFT approach seems to be better justified.