Computational Flow Physics Group

We develop advanced numerical simulation and data-driven analysis tools to understand, model, and predict turbulent and multiphysics flows in engineering and nature. Our work combines high-fidelity computation with modal decomposition and feature extraction to reveal coherent flow structures and translate them into predictive reduced-order models for forecasting and optimization. We focus on aerospace problems, including jet noise control, unsteady aerodynamics and aeroacoustics, transition, dynamic stall, aero-optics, and hypersonics. While our work is grounded in physics-based modeling, we also curiously explore where modern machine learning can complement first-principles simulation, statistical methods, and classical closure techniques. Beyond aerospace, our methods extend to complex geophysical flows, with applications in atmospheric science and physical oceanography.