Factorized Implicit Global Convolution for Automotive Computational Fluid Dynamics Prediction

TL;DR

This paper introduces FIGConv, a novel neural network architecture that efficiently predicts automotive CFD outcomes on large 3D meshes, significantly reducing computational complexity and improving accuracy over previous models.

Contribution

The paper presents a new architecture, FIGConv, that achieves quadratic complexity for CFD prediction on large 3D meshes, enabling more efficient and accurate automotive design simulations.

Findings

Achieves $R^2$ of 0.95 on DrivAerNet dataset.

Outperforms previous state-of-the-art by 40% in relative MSE.

Reduces computational complexity from $O(N^3)$ to $O(N^2)$.

Abstract

Computational Fluid Dynamics (CFD) is crucial for automotive design, requiring the analysis of large 3D point clouds to study how vehicle geometry affects pressure fields and drag forces. However, existing deep learning approaches for CFD struggle with the computational complexity of processing high-resolution 3D data. We propose Factorized Implicit Global Convolution (FIGConv), a novel architecture that efficiently solves CFD problems for very large 3D meshes with arbitrary input and output geometries. FIGConv achieves quadratic complexity $O(N^2)$, a significant improvement over existing 3D neural CFD models that require cubic complexity $O(N^3)$. Our approach combines Factorized Implicit Grids to approximate high-resolution domains, efficient global convolutions through 2D reparameterization, and a U-shaped architecture for effective information gathering and integration. We validate our approach on the industry-standard Ahmed body dataset and the large-scale DrivAerNet dataset. In DrivAerNet, our model achieves an $R^2$ value of 0.95 for drag prediction, outperforming the previous state-of-the-art by a significant margin. This represents a 40% improvement in relative mean squared error and a 70% improvement in absolute mean squared error over previous methods.