AI-Optimized Flexible-Particle Shapes in Turbulent Flows

3IA Côte d'Azur invites applications for a PhD thesis focused on optimizing particle shapes to maximize sampling in turbulent flows. The project blends physical modeling with AI-driven optimization, using elastic deformations projected onto a finite-dimensional space.

Bayesian optimization and reinforcement learning will solve the inverse problem under specified objectives. The candidate will study dynamics in Stokes flow, model flexible particles as rigid-sphere assemblies, and test approaches