The MHM method for the second-order elastodynamic model - Inria - Institut national de recherche en sciences et technologies du numérique Accéder directement au contenu
Pré-Publication, Document De Travail Année : 2018

The MHM method for the second-order elastodynamic model

La méthode MHM pour le modèle élastodynamique de second ordre

Résumé

The wave propagation phenomena in heterogeneous media is an actual challenging problem to numerical simulation. In today's scenario of increasing computing power and advent of exascale computing, machines provide good environments to obtain solutions to tough problems. However, it is necessary to construct good algorithms that can enjoy this massively parallel computing. The recent Multiscale Hybrid-Mixed (MHM) method is an alternative to be used in such a scenario since it constructs solutions with high-order precision by localizing computations in independent local problems. The final solution is obtained by a post-processing, using solutions of the local problems and the one from a coarse global problem. This work presents the mathematical background of the MHM method for a second-order elastodynamic model in time domain considering highly heterogeneous coefficients. We show several numerical simulations to verify the optimal error convergence of the MHM method and its capacity to deal with heterogeneous three-dimensional media on non-aligned coarse meshes. All the simulations are performed in a hybrid parallel environment and we are able to conclude that the MHM method is a competitive option to handle multiscale wave propagation problems.
Fichier principal
Vignette du fichier
slides.pdf (2.3 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01840081 , version 1 (31-07-2018)

Identifiants

  • HAL Id : hal-01840081 , version 1

Citer

Weslley da Silva Pereira, Claire Scheid, Frédéric Valentin. The MHM method for the second-order elastodynamic model. 2018. ⟨hal-01840081⟩
184 Consultations
46 Téléchargements

Partager

Gmail Facebook X LinkedIn More