Speaker
Description
Living organisms such as fish propel themselves through periodic, self-generated deformations of their body shape. Modeling this motion within a fluid-structure interaction (FSI) framework requires a careful separation of the active deformation, which is prescribed and drives the motion, from the passive, resulting motion of the body, which follows the laws of rigid body dynamics. In this talk, we present a modeling and discretization framework for such problems, building on classical monolithic ALE-based FSI approaches for fluid-rigid body and fluid-structure interaction problems.
We describe how the body's motion is decomposed into a prescribed active deformation, modeling the fish-like undulation, and an emergent rigid body motion, determined self-consistently from the interaction with the surrounding fluid. This splitting leads to a coupled system in which the rigid body position and velocity become additional unknowns, coupled monolithically to the incompressible Navier-Stokes equations through kinematic and dynamic transmission conditions on the moving boundary.
We discuss the resulting numerical discretization, based on a finite element ALE formulation, and address the solution of the strongly coupled nonlinear system arising at each time step. Since the propulsion mechanism is periodic in time, we further consider the direct computation of the time-periodic limit cycle, avoiding long transient simulations by formulating and solving a periodicity constraint as part of the coupled system.
Finally, we present first results on the optimization of the active deformation with respect to propulsion efficiency, laying the groundwork for a systematic design and control framework for self-propelled, fish-like locomotion in viscous fluids.