RUS  ENG
Full version
JOURNALS // Numerical methods and programming // Archive

Num. Meth. Prog., 2022 Volume 23, Issue 1, Pages 46–59 (Mi vmp1049)

Methods and algorithms of computational mathematics and their applications

Multigrid methods with skew-Hermitian based smoothers for the convection–diffusion problem with dominant convection

T. S. Martynova, G. V. Muratova, I. N. Shabas, V. V. Bavin

Southern Federal University, Vorovich Institute for Mathematics, Mechanics and Computer Science, Rostov-on-Don, Russia

Abstract: The convection–diffusion equation with dominant convection is considered on a uniform grid of central difference scheme. The multigrid method is used for solving the strongly nonsymmetric systems of linear algebraic equations with positive definite coefficient matrices. Two-step skew-Hermitian iterative methods are utilized for the first time as a smoothing procedure. It is demonstrated that using the proper smoothers enables to improve the convergence of the multigrid method. The robustness of the smoothers with respect to variation of the Peclet number is shown by local Fourier analysis and numerical experiments.

Keywords: convection-diffusion equation, multigrid methods, smoothing procedure, product-type skew-Hermitian triangular splitting, local Fourier analysis, convergence.

Received: 19.11.2021
Accepted: 01.02.2022

Language: English

DOI: 10.26089/NumMet.v23r104



© Steklov Math. Inst. of RAS, 2024