Artificial Viscosity

Artificial viscosity is a technique used to stabilize numerical solutions of fluid dynamics equations, particularly near discontinuities like shocks, where standard methods can fail. Current research focuses on developing adaptive and data-driven methods for determining optimal artificial viscosity, employing techniques like Bayesian inference, physics-informed machine learning (including neural networks and reinforcement learning), and unsupervised learning algorithms such as Gaussian Mixture Models. These advancements aim to improve the accuracy and efficiency of computational fluid dynamics simulations across a range of applications, from medical device design to high-speed aerodynamics, by automating the selection and application of artificial viscosity models. The ultimate goal is to create more robust and reliable simulations without requiring extensive manual tuning.