Accelerating Multi-scale Simulations of Nuclear Components via PCYS Interpolation Tables
Zirconium alloy core components in nuclear reactors, such as spacer grids and fuel cladding, undergo anisotropic dimensional changes driven by coupled irradiation creep and growth. While micromechanical crystal plasticity frameworks like the Viscoplastic Self-Consistent (VPSC) formulation capture these microstructurally driven phenomena, their integration into macroscopic Finite Element Method (FEM) solvers is computationally prohibitive for engineering-scale components. To bridge this gap, this work presents a multi-scale framework implemented within the open-source FEM solver Code_Aster. The developed interface uses a 5D Interpolation Table (IT) as a static material surrogate to govern instantaneous viscoplastic responses, coupled with a periodic recalibration and first-order Taylor series linearization scheme to track microstructural drift due to radiation damage without on-the-fly database updates. The predictive accuracy, numerical stability, and performance of this Polycrystal Yield Surface (PCYS) interpolation approach are benchmarked against VPSC-FEM simulations under continuous high-dose irradiation scenarios. Material-level assessments demonstrate that the linearization scheme bounds relative errors below 1% for representative deformation paths, maintaining structural compatibility. Furthermore, structural simulations of a spacer grid domain revealed meaningful computational savings, overcoming the multi-scale computational penalty while preserving microstructural fidelity. The proposed framework shows potential for multiphysics structural assessments and safety margin evaluations of core internals over operational lifespans.