2016
DOI: 10.1137/151006135
|View full text |Cite
|
Sign up to set email alerts
|

Monolithic Multigrid Methods for Two-Dimensional Resistive Magnetohydrodynamics

Abstract: The magnetohydrodynamics (MHD) equations model a wide range of plasma physics applications and are characterized by a nonlinear system of partial differential equations that strongly couples a charged fluid with the evolution of electromagnetic fields. After discretization and linearization, the resulting system of equations is generally difficult to solve due to the coupling between variables, and the heterogeneous coefficients induced by the linearization process. In this paper, we investigate multigrid prec… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

4
67
0

Year Published

2016
2016
2020
2020

Publication Types

Select...
7

Relationship

4
3

Authors

Journals

citations
Cited by 47 publications
(71 citation statements)
references
References 60 publications
(87 reference statements)
4
67
0
Order By: Relevance
“…Different control volumes are defined depending on which variable is considered. In Figure 2, we represent in different colors the control volumes corresponding to the primary variables u, v, and p. 1 The pressure unknowns p are defined at the centers of the blocks (marked by ×-points in Figure 2), and the components of the velocity unknowns, u and v, are located at the centers of the block faces (denoted by the •-and •-points in the same figure). For the description of the discrete scheme, we need to fix an adequate indexing for the unknowns, which can be seen in Figure 3, where each unknown is depicted together with the corresponding control volume and the different variables around it.…”
Section: Discretizationmentioning
confidence: 99%
See 1 more Smart Citation
“…Different control volumes are defined depending on which variable is considered. In Figure 2, we represent in different colors the control volumes corresponding to the primary variables u, v, and p. 1 The pressure unknowns p are defined at the centers of the blocks (marked by ×-points in Figure 2), and the components of the velocity unknowns, u and v, are located at the centers of the block faces (denoted by the •-and •-points in the same figure). For the description of the discrete scheme, we need to fix an adequate indexing for the unknowns, which can be seen in Figure 3, where each unknown is depicted together with the corresponding control volume and the different variables around it.…”
Section: Discretizationmentioning
confidence: 99%
“…The proposed smoother in our work is based on Uzawa relaxation, but different relaxation schemes can be considered. The Braess-Sarazin method [1,5] is an example of another relaxation method. More concretely, the Braess-Sarazin method is based on the matrix system…”
Section: Multiblock Multigrid Algorithmmentioning
confidence: 99%
“…An alternative choice that yields less memory use and less time‐per‐iteration is the ‘diagonal’ Vanka submatrix : Mℓℓdiag=diag(Fℓℓ)BℓℓTBℓℓ0, where F ℓ ℓ , B ℓ ℓ , and BℓℓT are defined as previously. The methods using this submatrix will be called ‘diagonal Vanka’ if we are using the element‐wise blocks and ‘diagonal extended Vanka’ if we are using the extended blocks.…”
Section: Monolithic Multigridmentioning
confidence: 99%
“…One strategy is to consider the diagonal of F , C diag =diag( F ) . However, in order to improve robustness and convergence, we also consider a block‐diagonal preconditioner , denoted C blkDiag =blkDiag( F ), in which the blocks correspond to the two velocity degrees of freedom on an edge in the mesh. Note that this choice of blocks is dependent upon the discretization considered.…”
Section: Monolithic Multigridmentioning
confidence: 99%
“…This methodology leads to improved convergence for elliptic problems in a straightforward way [2,3], while partial differential equations with a dominant hyperbolic character are often handled with additional artificial dissipation [4,5]. Multigrid methods are used in various branches of applied mathematics and engineering, such as electromagnetics [6], magnetohydrodynamics [7] and fluid dynamics [8].…”
Section: Introductionmentioning
confidence: 99%