Graduate researcher working at the intersection of computational fluid dynamics, deep learning, and nuclear thermal-hydraulics. I build surrogate models and hybrid AI–CFD pipelines that aim to be faster than classical solvers without giving up on the physics.
- Neural surrogates and operator learning for accelerating fluid simulations
- Hybrid coupling of deep learning with traditional CFD solvers (OpenFOAM)
- Applications to nuclear thermal-hydraulics: passive cooling, steam-generator flows, reactor systems
- ML–CFD Computation: a cross-computation framework implementing the RePIT strategy for hybrid AI–CFD prediction
- MARL: Multi-Agent Reinforcement Learning for active flow control of 2D Rayleigh–Bénard convection
- LinkedIn — shilaj-baral
- Email — shilajbaral@gmail.com