Rayleigh B\'enard Convection

Rayleigh-Bénard convection (RBC) studies the heat transfer in fluids heated from below, focusing on understanding the complex patterns and dynamics of convection rolls that arise. Current research emphasizes developing advanced control strategies, employing machine learning techniques like multi-agent reinforcement learning and Koopman-based models to manipulate and predict these flows, often using data from direct numerical simulations. These efforts aim to improve our understanding of turbulent heat transfer and enable more efficient control of industrial processes and improved modeling of geophysical flows, such as atmospheric and oceanic circulation. The development of accurate surrogate models for turbulent RBC is a key focus, with ongoing work exploring the limitations and potential of data-driven approaches for subgrid-scale modeling.