A three-dimensional Fourier analysis of fine mesh rebalance acceleration of linear discontinuous sub-cell balance method for diffusion equation on tetrahedral meshes. (November 2019)

• Record Type: Journal Article
• Authors: Muhammad, Habib; Hong, Ser Gi
• Is Part Of: Annals of nuclear energy. Volume 133(2019)
• Page Start: 145
• Page End: 153
• Publication Date: 2019-11
• Subjects: Fourier convergence analysis -- FMR acceleration -- Tetrahedral meshes -- Linear discontinous sub-cell balance method; Nuclear energy -- Periodicals; Nuclear engineering -- Periodicals; 621.4805
• DOI: 10.1016/j.anucene.2019.05.020
• Languages: English
• ISSNs: 0306-4549

Abstract

Highlights: Three-dimensional Fourier analysis for FMR acceleration of LDEM-SCB(1) method for neutron diffusion equation. The Fourier analysis was done both for homogeneous and heterogeneous problems with the translation and reflective boundary conditions. The results show that FMR significantly accelerates the Gauss-Seidel (GS)-like iteration. Abstract: This paper presents a three-dimensional Fourier analysis of the fine mesh rebalance (FMR) acceleration of the Linear Discontinuous Expansion Method with Subcell Balance (LDEM-SCB) discretization method for solving the neutron diffusion equation on three-dimensional unstructured tetrahedral meshes in order to theoretically understand the convergence characteristics. The three-dimensional Fourier analysis is performed on one and two cubes called the basic elements in which each element is divided into six tetrahedral meshes for translation and reflective boundary conditions on the external boundaries. The Fourier analysis shows that the FMR method significantly accelerates the Gauss-Seidel (GS)-like iteration to solve the LDEM-SCB discretized diffusion equation. The analytical results of the homogeneous test problem taking different aspect ratios show that the convergence of FMR with the translation boundary conditions is less sensitive on the aspect ratio than the one with the reflective boundary conditions. In addition, we identified the convergence characteristics of the GS-like iteration and its FMR acceleration of the heterogeneous problem.