Data-driven reduced order modeling for flows with moving geometries using shifted POD

Philipp Krah; Miriam Goldack; Thomas Engels; Kai Schneider; Julius Reiss

Communication dans un congrès

Data-driven reduced order modeling for flows with moving geometries using shifted POD

Philipp Krah ¹ Miriam Goldack ^{2, 3} Thomas Engels ³ Kai Schneider ² Julius Reiss ³

1 Institut für Mathematik, TU Berlin

2 I2M - Institut de Mathématiques de Marseille

3 Institute of Fluid Mechanics and Engineering Acoustics, Berlin University of Technology, D-10623 Berlin, Germany

Abstract : A new approach, based on the shifted Proper Orthogonal Decomposition (sPOD), for modelling and optimizing flows with moving geometries is presented. This novel model order reduction technique is shown to be efficient for advection dominated systems and its approximation properties are assessed. Moreover its applicability for path optimization of moving geometries will be shown.

Keywords : Fluidics and Thermodynamics Numerical Simulation and Co-Simulation Model Simplification and Optimization Model Reduction

Type de document :

Communication dans un congrès

Domaine :

Mathématiques [math] / Analyse numérique [math.NA]

Physique [physics] / Mécanique [physics] / Mécanique des fluides [physics.class-ph]

Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-03396325
Contributeur : Kai Schneider Connectez-vous pour contacter le contributeur
Soumis le : vendredi 22 octobre 2021 - 16:32:56
Dernière modification le : mardi 26 octobre 2021 - 04:00:32
Archivage à long terme le : : lundi 24 janvier 2022 - 16:59:13

KrahGoldackEngelsSchneiderReis...

Fichiers produits par l'(les) auteur(s)

Data-driven reduced order modeling for flows with moving geometries using shifted POD

Fichier

Identifiants

Collections

Citation

Exporter

Métriques

186

70

Data-driven reduced order modeling for flows with moving geometries using shifted POD

Fichier

Identifiants

Collections

Citation

Exporter

Partager

Métriques

186

70