Parametric 3D wing dataset created to improve AI aerodynamic predictions
Ground-Wing: A Parametric 3D Wing-in-Ground-Effect Dataset for Benchmarking Neural Surrogate Models
Machine Learning
Summary
Predicting how air flows around airplane wings is important but complicated, especially when wings are close to the ground. The paper introduces a new collection of detailed computer simulations of 3D wings flying near the ground, which is common in motorsport. The data helps test different kinds of AI models that try to predict aerodynamic forces quickly without full physics simulations. The authors compare six AI architectures to see which predict pressures and forces best on these tricky wing shapes.
What this means in practice
- •For motorsport engineers: Use the dataset to develop AI tools that predict race car wing aerodynamics near the ground more efficiently and accurately.
- •For drone designers: Improve aerodynamic predictions for low-flying drones operating close to surfaces using benchmarked neural surrogate models.
Authors
Andrea Cenzato, Lukas Karch, Andre Luckow, Yoran Pichon, Francesco Pasqua, Abouzar Ghasemi, Rishikesh Ranade, Ira Shokar, Julia Kowalski
Abstract
This paper presents a CFD dataset of a highly parametrized 3D wing operating in ground effect, designed for benchmarking neural surrogate models for aerodynamic prediction, using a wing shape and flow conditions typically encountered in motorsport applications. The dataset comprises 770 RANS simulations of 194 unique wing configurations, sampled across 27 airfoil profile variants and 4 distances from the ground, spanning variations in wing sweep, chord, taper, angle of attack, and tip twist. Six neural network architectures representing distinct paradigms --- convolutional (DoMINO), graph-based (X-MeshGraphNet), transformer-based (AB-UPT, GeoTransolver, GAOT), and JEPA-based (AeroJEPA) --- are evaluated on their ability to predict surface pressure coefficient and wall shear stress distributions, from which integral drag and lift coefficients are recovered. The dataset is intended to be used as a challenging benchmark for surrogate CFD models, due to the uniqueness of the ground effect being absent from other wing and airfoil datasets.