Abstract:
Background High-fidelity neutronics simulation of nuclear reactor cores, particularly those with complex geometries such as the AP1000, remains computationally challenging. Efficient deterministic methods that can achieve Monte Carlo-level accuracy are highly desirable for design and analysis.
Purpose This study aims to develop, apply, and validate the FLASH code, which implements an advanced Fission Response Function (FRF) algorithm, for performing efficient and accurate full-core, pin-wise neutronics calculations of the AP1000 reactor core.
Methods The FRF database was generated through reference-state simulations using the Serpent Monte Carlo code. To enhance accuracy in complex geometries, the methodology incorporated a local inter-assembly environmental correction factor to address fuel assembly heterogeneity and a predictor-corrector scheme to precisely simulate reflector environmental effects. The performance of the FLASH code was validated against reference Monte Carlo solutions under Hot Zero Power (HZP) conditions.
Results The validation results demonstrated high accuracy. Deviations in the effective multiplication factor (keff) were within +220 pcm for all 2D axial slices and +209 pcm for the full 3D core calculation. The root-mean-square error (RMSE) was below 1.1% for the 2D pin power distribution, while the 3D pin power RMSE was 1.05% and the 3D assembly power RMSE was 0.67%. In terms of efficiency, the FLASH code completed the pin-wise full-core 3D calculation for the AP1000 in 106 seconds using 64 CPU cores.
Conclusions The developed FLASH code, based on the FRF algorithm with integrated correction schemes, successfully bridges the gap between efficiency and high fidelity. It provides a rapid and accurate computational tool for AP1000 core analysis, confirming the practicality and effectiveness of the proposed methodology for detailed reactor physics calculations.