Gauss-compatible Galerkin schemes for time-dependent Maxwell equations - Archive ouverte HAL Accéder directement au contenu
Pré-Publication, Document De Travail Année : 2014

Gauss-compatible Galerkin schemes for time-dependent Maxwell equations

Résumé

In this article we propose a unified analysis for conforming and non-conforming finite element methods that provides a partial answer to the problem of preserving discrete divergence constraints when computing numerical solutions to the time-dependent Maxwell system. In particular, we formulate a compatibility condition relative to the preservation of genuinely oscillating modes that takes the form of a generalized commuting diagram, and we show that compatible schemes satisfy convergence estimates leading to long-time stability near stationary solutions. We next apply these findings by specifying compatible formulations for several classes of Galerkin methods, such as the usual curl-conforming finite elements and the centered discontinuous Galerkin (DG) scheme. We also propose a new conforming/non-conforming Galerkin (Conga) method where fully discontinuous solutions are computed by embedding the general structure of curl-conforming finite elements into larger DG spaces. In addition to naturally preserving one of the Gauss laws in a strong sense, the Conga method is both spectrally correct and energy conserving, unlike existing DG discretizations where the introduction of a dissipative penalty term is necessary to avoid the presence of spurious modes.
Fichier principal
Vignette du fichier
mcomp-conga-6_-_submitted_version-v2.pdf (347.81 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-00969326 , version 1 (02-04-2014)
hal-00969326 , version 2 (24-06-2015)

Identifiants

  • HAL Id : hal-00969326 , version 1

Citer

Martin Campos Pinto, Eric Sonnendrücker. Gauss-compatible Galerkin schemes for time-dependent Maxwell equations. 2014. ⟨hal-00969326v1⟩
530 Consultations
405 Téléchargements

Partager

Gmail Facebook X LinkedIn More