Master thesis and internship[BR]- Master's thesis : Arbitrary Lagrangian-Eulerian and Fluid-Structure Interaction formulations to simulate bird impacts[BR]- Integration internship

Abstract

This thesis investigates the theoretical foundations, implementation details, and performance of the Arbitrary Lagrangian-Eulerian (ALE) and Fluid-Structure Interaction (FSI) formulations in LS-DYNA, and compares them to the Smoothed Particle Hydrodynamics (SPH) method. A two-part approach was adopted. First, a simple case study consisting of a plane-strain bar impact on a rigid wall was considered. The different ALE formulations in LS-DYNA were introduced, and their results were compared to those of the Lagrangian formulation. Generally more computationally expensive, they yielded comparable outcomes—sometimes of slightly lower accuracy— with their mesh quality remaining acceptable despite large strains. The FSI formulation was then introduced to model contact interactions between bodies and was compared to conventional Lagrangian contact. Also more computationally expensive, it produced similar results to the Lagrangian case, aside from slightly larger energy losses. This initial study of the ALE and FSI behaviors revealed the influence of various parameters and highlighted the strengths and limitations of each formulation. Second, a large-scale model—originally developed for SPH simulations—was used to evaluate the performance of the ALE and FSI formulations in terms of computational cost and result fidelity compared to experimental data. While the ALE and FSI formulations produced results similar to the SPH simulation, mesh-convergence issues limited the reliability of the comparison. The source of these issues could not be clearly identified, partly due to the inherited complexity and uncertainties of the provided model. Despite these challenges, the study demonstrated the potential of ALE and FSI formulations to compete with SPH, especially when extended features such as mesh motion are utilized to mitigate computational costs.