Physics-constrained Gaussian Processes for Predicting Shockwave Hugoniot Curves
Researchers have developed a novel physics-constrained Gaussian Process regression framework to predict shockwave Hugoniot curves. This method uses a limited number of shockwave simulations to accurately estimate material behavior under extreme conditions and quantify prediction uncertainties. The approach is demonstrated on silicon carbide and can inform future experiments and simulations for material science. AI
IMPACT Introduces a novel AI-driven approach for material science simulations, potentially accelerating discovery and reducing computational costs.