Improving source convergence diagnostics sensitivity using the complex components of Fourier fundamental mode

Abstract

Previous works have demonstrated that the Fourier fundamental mode coefficient (FFMC) is a potential fission source convergence indicator for Monte Carlo k-eigenvalue calculations. Unlike the conventional Shannon entropy indicator, it requires no spatial meshing, and its computation is computationally inexpensive. However, we found that FFMC is less sensitive to the convergence involving near symmetrical fission distributions. To rectify the problem, we focus on the analysis of the FFMC’s components (the real and imaginary parts) on a simple geometry problem and a full reactor core problem. The usage of the FFMC’s components is more conservative compared to the original FFMC method as supported by the improved results of the fission source convergence detection.