WindMiL: Equivariant Graph Learning for Wind Loading Prediction

Themistoklis Vargiemezis, Charilaos I. Kanatsoulis, Catherine Gorle

Keywords: Geometric deep learning, Graph Neural Networks, Equivariance, Wind loading, Computational Fluid Dynamics, Large-eddy Simulations

TL;DR: We introduce WindMiL, a reflection-equivariant graph neural network surrogate trained on a systematically generated LES dataset to predict wind pressures on buildings, reducing 24-hour CFD simulations to fast and accurate predictions.

Abstract: Accurate prediction of wind loading on buildings is crucial for structural safety and sustainable design, yet conventional approaches such as wind tunnel testing and large-eddy simulation (LES) are prohibitively expensive for large-scale exploration. Each LES case typically requires at least 24 hours of computation, making comprehensive parametric studies infeasible. We introduce \textsc{WindMiL}, a new machine learning framework that combines systematic dataset generation with symmetry-aware graph neural networks (GNNs). First, we introduce a large-scale dataset of wind loads on low-rise buildings by applying signed distance function interpolation to roof geometries and simulating 462 cases with LES across varying shapes and wind directions. Second, we develop a reflection-equivariant GNN that guarantees physically consistent predictions under mirrored geometries. Across interpolation and extrapolation evaluations, \textsc{WindMiL} achieves high accuracy for both the mean and the standard deviation of surface pressure coefficients (e.g., RMSE $\leq 0.02$ for mean $C_p$) and remains accurate under reflected-test evaluation, maintaining hit rates above $96\%$ where the non-equivariant baseline model drops by more than $10\%$. By pairing a systematic dataset with an equivariant surrogate, \textsc{WindMiL} enables efficient, scalable, and accurate predictions of wind loads on buildings.

Submission Number: 121