Physics-Informed Surrogates for Design
Neural-network surrogates that respect conservation laws, for fast aerodynamic design sweeps.
Design engineers often need hundreds of flow solutions to explore a design space. This completed project trained physics-informed neural networks that predict surface pressure and skin friction in milliseconds, while penalising violations of the governing equations.
Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.