Development of a physics-informed neural network to enhance wind tunnel data for aerospace design

Abstract

In recent years, physics-informed neural networks (PINNs) have emerged as a novel approach to solving PDEs in many applications, including the Navier-Stokes equations in fluid mechanics. We seek to develop a tool based on PINNs that can bridge gaps in the experimental and computational methods currently utilized in aerospace design and analysis. The feasibility of such a product is demonstrated in the context of 2D steady flows over airfoil geometries. Reconstruction of full flow fields from spatially sparse measurements of pressure and flow direction is presented, motivated by prevalent practices in subsonic wind tunnel testing campaigns. Areas of algorithm improvement, directions for future research, and broader visions for the collaboration are discussed.

Additional details