Aircraft Aerodynamics
1 article · search the full text for this term
-
Physics-Informed Machine Learning for Aircraft Aerodynamics and Fluid-Flow Prediction: Methods, Validation, and Research Challenges
Abstract: Physics-informed machine learning (PIML) is increasingly studied as a means of combining governing fluid-flow equations with data-driven prediction of aerodynamic fields and integral loads. This review critically evaluates residual-based physics-informed neural networks, neural operators, conventional surrogate models and hybrid CFD–learning systems for aircraft aerodynamics. Attention is given to the incompressible and compressible governing equations, boundary enforcement, turbulence closure, distribution shift, uncertainty calibration, and computational-cost accounting. A structured benchmarking protocol is …
Published in Recent Trends in Fluid Mechanics Read article