DrivAerStar: An Industrial-Grade CFD Dataset for Vehicle Aerodynamic Optimization

2025-10-19

AI Summary

High-precision aerodynamic optimization is critical for extending electric vehicle (EV) driving range, yet conventional CFD simulations are prohibitively time-consuming (weeks per case), and existing machine learning approaches are hampered by low-resolution datasets, incomplete component representation (e.g., missing engine bays and cooling systems), and validation errors exceeding 5%. Method: We introduce the first industrial-grade automotive CFD dataset, comprising 12,000 high-fidelity STAR-CCM+ simulations across three vehicle classes. Geometric variations are systematically generated via Free-Form Deformation (FFD) over 20 CAD parameters, with full modeling of underhood flow and thermal management systems. Simulations employ refined boundary-layer meshing and strict y⁺ control. Contribution/Results: Validation error is reduced to 1.04%—a fivefold improvement over prior state-of-the-art. Our trained AI models achieve production-grade accuracy, slashing prediction time from weeks to minutes. This work establishes a new paradigm for data-driven aerodynamic design, bridging academic AI research and industrial CFD practice.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Machine Learning: Deep Neural Architectures and Foundation Models
• Computer Vision: Diffusion Models for Vision

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Web Mining and Content Analysis: Web data generation and simulation

Abstract

Vehicle aerodynamics optimization has become critical for automotive electrification, where drag reduction directly determines electric vehicle range and energy efficiency. Traditional approaches face an intractable trade-off: computationally expensive Computational Fluid Dynamics (CFD) simulations requiring weeks per design iteration, or simplified models that sacrifice production-grade accuracy. While machine learning offers transformative potential, existing datasets exhibit fundamental limitations -- inadequate mesh resolution, missing vehicle components, and validation errors exceeding 5% -- preventing deployment in industrial workflows. We present DrivAerStar, comprising 12,000 industrial-grade automotive CFD simulations generated using $ ext{STAR-CCM+}^unicode{xAE}$ software. The dataset systematically explores three vehicle configurations through 20 Computer Aided Design (CAD) parameters via Free Form Deformation (FFD) algorithms, including complete engine compartments and cooling systems with realistic internal airflow. DrivAerStar achieves wind tunnel validation accuracy below 1.04% -- a five-fold improvement over existing datasets -- through refined mesh strategies with strict wall $y^+$ control. Benchmarks demonstrate that models trained on this data achieve production-ready accuracy while reducing computational costs from weeks to minutes. This represents the first dataset bridging academic machine learning research and industrial CFD practice, establishing a new standard for data-driven aerodynamic optimization in automotive development. Beyond automotive applications, DrivAerStar demonstrates a paradigm for integrating high-fidelity physics simulations with Artificial Intelligence (AI) across engineering disciplines where computational constraints currently limit innovation.

Problem

• Industrial CFD datasets lack accuracy for vehicle aerodynamic optimization
• Traditional CFD simulations are computationally expensive and time-consuming
• Existing datasets have validation errors preventing industrial deployment

Innovation

• Industrial-grade CFD dataset with 12,000 simulations
• Free Form Deformation algorithms explore CAD parameters
• Refined mesh strategies achieve 1.04% validation accuracy