Compressible Flow

Compressible flow research focuses on accurately and efficiently modeling fluid dynamics where density changes significantly, crucial for applications like aerospace and weather prediction. Current efforts leverage machine learning, employing neural networks (including DeepONets, physics-informed neural networks, and generative models like diffusion models) to create fast and accurate surrogate models for computationally expensive simulations, often focusing on improving the handling of discontinuities and turbulence. These advancements promise to accelerate scientific discovery and enable real-time simulations for engineering design and optimization, particularly in scenarios involving complex geometries and high-speed flows.