Development of a Multi-Scale Computational Framework for Radiation Damage Prediction in Structural Materials for Nigerian Nuclear Infrastructure

Radiation damage in structural materials of nuclear reactors and radioactive waste storage facilities is a multi-scale physics problem spanning electronic stopping power interactions at the femtosecond timescale through defect cluster evolution at the year timescale. Predictive models for structural integrity of materials in Nigeria's proposed nuclear infrastructure must capture this multi-scale behavior to ensure reliable performance lifetime estimates. This research develops an original multi-scale computational framework that links quantum mechanical density functional theory calculations of displacement threshold energies and electronic stopping, through molecular dynamics simulation of displacement cascades, to rate theory and phase field models of microstructural defect evolution under neutron irradiation. New interatomic potentials for reactor-relevant austenitic stainless steel and zircaloy alloys are developed by fitting to density functional theory energy surfaces using machine learning force field methodology implemented in the TensorFlow framework. The molecular dynamics cascade simulations are performed on the NSCC supercomputing facility through a National Research Foundation Nigeria collaboration, covering primary knock-on atom energies from 1 keV to 150 keV. Original contributions include a novel defect recombination rate theory that accounts for the spatial heterogeneity of damage microstructure produced by cascade clustering, which reduces over-prediction of steady-state vacancy concentration by 47 percent relative to standard homogeneous rate theory. The framework predicts swelling, embrittlement, and irradiation creep rates for candidate materials of the Nigerian Atomic Energy Commission's proposed research reactor pressure vessel over a 20-year design life. Keywords: radiation damage, multi-scale modeling, nuclear materials, molecular dynamics, Nigeria