How an added mass matrix estimation may dramatically improve FSI calculations for moving foils - Université de technologie de Compiègne
Article Dans Une Revue Applied Mathematical Modelling Année : 2017

How an added mass matrix estimation may dramatically improve FSI calculations for moving foils

Emmanuel Lefrançois
  • Fonction : Auteur
  • PersonId : 867462

Résumé

This paper presents a corrected partitioned scheme for investigating fluid–structure interaction (FSI) that may be encountered by lifting devices immersed in heavy fluid such as liquids. The purpose of this model is to counteract the penalizing impact of the added mass effect on the classical partitioned FSI coupling scheme. This work is based on an added mass corrected version of the classical strongly coupled partitioned scheme presented in Song et al. (2013). Results show that this corrected version systematically allows convergence to the coupled solution. The fluid flow model considered here uses a non-stationary potential approach, commonly termed the Panel Method. The advantage of this kind of approach is twofold: first, in restricting itself to a boundary method and, second, in allowing an added mass matrix to be estimated as a post-processing phase. Whereas the classical scheme encounters an acceptable (no numerical oscillation) convergence limit for fluid densities higher than 8 kg/m3 for the considered case, our corrected scheme is not dependent on fluid density and converges with only 6 iterations. This makes it possible to investigate the dynamic behavior of a 2D foil immersed in heavy fluids such as water. For example, it recognizes that frequency shifting may occur as the consequence of a strong added mass effect.

Dates et versions

hal-01993640 , version 1 (25-01-2019)

Identifiants

Citer

Emmanuel Lefrançois. How an added mass matrix estimation may dramatically improve FSI calculations for moving foils. Applied Mathematical Modelling, 2017, 51, pp.655-668. ⟨10.1016/j.apm.2017.07.022⟩. ⟨hal-01993640⟩
16 Consultations
0 Téléchargements

Altmetric

Partager

More