MAGIC-GPS global variance reduction method for large-scale shielding calculation

Abstract

We propose a Method of Automatic Generation Importance by Calculation with Growth Population Scheme (MAGIC-GPS) for efficient and accurate shielding calculation. The new method adopts a Growth Population Scheme (GPS) in the reciprocal iteration process of flux and importance for the Monte Carlo fixed-source calculation. Mathematical derivation is performed to prove that the GPS scheme is conducive to a rapid flux distribution convergence and can improve the Figure of Merit (FOM) of the Monte Carlo deep-penetration simulation. The effectivity of the MAGIC-GPS method is tested by a single concrete rod model, the H.B. Robinson-2 reactor benchmark (HBR2) and the China Fusion Engineering Test Reactor (CFETR). The results show that the MAGIC-GPS method can accelerate the Monte Carlo deep-penetration simulation, and the faster the population of particles increases, the better the acceleration. The MAGIC-GPS method is helpful for large-scale shielding calculation.