- PhD,
PhD Defense - Catherine Ramirez Villalba - LHEEA/ED SPI
Catherine Ramirez Villalba will defend her PhD on Friday, November 20, 2020. The PhD is entitled "Towards an efficient modeling of fluid-structure interaction". This PhD defense is confidential.
On November 20, 2020 from 10:00 To 13:00
Committee
Chairman:Yannick HOARAU - Professeur, Université de Strasbourg
Examiners:
Elisabeth LONGATTE-LACAZEDIEU - Enseignant-chercheur, EPF Ecole d'Ingénieurs
Yann ROUX - CEO, K-Epsilon
Wulf DETTMER - Professor, Swansea University (reviewer)
Joris DEGROOTE - Professor, Ghent University (reviewer)
PhD supervisor:
Michel VISONNEAU - Research Director CNRS, LHEEA, Ecole Centrale de Nantes
Co-supervisor:
Alban LEROYER - Assistant-professor, LHEEA, Ecole Centrale de Nantes
Guest:
Corentin LOTHODE - Research Engineer CNRS, LMRS, Université de Rouen
Abstract
TOWARDS AN EFFICIENT MODELING OF FLUID-STRUCTURE INTERACTIONFSI industrial applications are often described by complex geometries and materials. In order to accurately predict their behavior, high computational costs are associated, both in time and in computational resources. To improve the quality of the prediction without penalizing the computational time, and to reduce the computational time without impacting the accuracy that is available today, two main axes are explored in this work. The first one is the study of an asynchronous algorithm that could allow the use of complex structural models. The second axis consists of the study of the strip method while combining the use of a RANS model and a non-linear FEM model. On one hand, the study of asynchronicity in the FSI domain revealed different aspects of interest that must be addressed before the approach can be used industrially. However, a first treatment of the limitations found showed signs of an improvement that could lead to a promising algorithm, one that naturally lies between the implicit external algorithm and the implicit internal algorithm. On the other hand, it was shown that the strip method developed in this work achieves a significant reduction in calculation time while maintaining excellent accuracy.