ENHANCING THE PERFORMANCE OF FISSION SOURCE CONVERGENCE WITH FUNCTIONAL EXPANSION TALLIES IN SERPENT 2 MONTE CARLO CODE

Abstract

This paper presents an upgrade to the built-in response matrix based solver implemented in Serpent 2 Monte Carlo code aiming to improve the fission source convergence when obtaining the forward solution to the k-eigenvalue criticality source problems. The functional expansion tallies are introduced in an attempt to improve the accuracy of the cell-wise form factors that feed the response matrix solver, replacing the current mesh-based approach. The functional expansion tallies reconstruct the binning surface and collision tallies, by using high-order series expansion to represent the original and continuous spatial distributions. This new feature is implemented to Serpent 2 and tested by single-assembly and full-core PWR calculations (BEAVRS benchmark). The results show enhanced performance of the convergence acceleration methodology based on an improved initial guess of the fission source.