Three-dimensional numerical analysis of fluid-structure interaction with structural contact

Abstract

The study of Fluid-Structure Interaction (FSI) problems with structural contact generates relevant contributions to Engineering (e.g. problems with immersed solids collision, dam gates and valves) and also to biomechanics (e.g. heart valve simulation). One of the challenges for these problems is the case of flows with topological changes in the fluid domain, which can be caused by free surface flow effects, such as waves breaking of bridges or offshore structures, or by contact between immersed solids. A widely employed robust technique to deal with such problems consists in considering the structures or free surfaces as boundaries immersed in a fixed fluid mesh under Eulerian description. However, the method present the difficulty to represent the immersed boundaries with precision, once they are generally represented by a level-set function, and the impossibility to ensure a suitable adaptive mesh refinement close to the moving boundary in order to represent local effects, such as boundary layer. The Particle-Finite Element Method (PFEM) is one alternative to modeling these problems, which combines finite elements with particle concepts, where the fluid is discretizes as a cloud of particles in Lagrangian description, allowing the simulation of problems with severe topological changes. This proposal aims at the development of a computational model under space-time formulation and PFEM concepts, able to simulate 3D FSI problems with structural contact, allowing topological changes in the fluid domain due to both, structural contact and free-surface effects. To do so, we propose an extension of the 2D PFEM code, developed during the candidate's Masters, to 3D case, followed by the implementation of the space-time formulation. We seek a better precision, once instead of a continuous re-meshing, a 4D mesh is employed in which the spatial mesh in the past and in the current instants are allowed to be completely different. Aspects regarding FSI coupling technique (monolithic and partitioned), solid constitutive model (Neo-Hookean e Mooney-Rivlin) and contact model (node-to-surface and surface-to-surface) are also to be considered in the study. (AU)