DrivAerML: High-Fidelity CFD Dataset for Road-Car External Aerodynamics
- Provider
- Neil Ashton et al. (caemldatasets.org)
- License
- CC-BY-SA-4.0 (source)
- Size
- 100,000 files
- Format
- CSV (in sampled files)
- Stored at
- Hugging Face (https://huggingface.co/datasets/neashton/drivaerml)
- Access
- Free and public. No account with the data provider is needed.
- Category
- Physics
- Source
- https://arxiv.org/abs/2408.11969
About DrivAerML: High-Fidelity CFD Dataset for Road-Car External Aerodynamics
DrivAerML is a high-fidelity computational fluid dynamics dataset for machine learning in automotive aerodynamics, published by Neil Ashton and colleagues, with the CFD workflow by UpstreamCFD. It contains 500 parametrically morphed variants of the DrivAer notchback generic road car, each simulated with scale-resolving hybrid RANS-LES in OpenFOAM v2212 on industrial-scale unstructured meshes. Measured 2026-11-14, the repository root holds run_N folders plus openfoam_meshes/; each run folder contains the STL and STEP geometry, geometry parameters and reference values (CSV), a boundary VTP surface file with native polygon areas (NPY), volume fields, slice data, time-averaged forces and moments, and images. The publisher lists about 31 TB stored, and 16 run ids are documented as non-public. It is a reference benchmark for surrogate models of surface pressure, wall shear stress, drag and lift (for example in NVIDIA PhysicsNeMo). Open VTP/VTU files with PyVista, VTK or ParaView and CSVs with pandas. Licensed CC-BY-SA-4.0; cite Ashton et al., arXiv:2408.11969.
How to load this dataset in Python
import pandas as pd
df = pd.read_csv("hf://datasets/neashton/drivaerml/force_mom_all.csv")
df.head()Data structure
File types in sampled files: CSV
Files
Example file: force_mom_all.csv
Analyze this dataset without downloading it
On Scigantic, DrivAerML: High-Fidelity CFD Dataset for Road-Car External Aerodynamics opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.
This page is also available as Markdown and as JSON.
Related Physics datasets
- Quantum Physics 0.5 Corpus
- Kolmogorov Flow Exact Coherent States: 8,952 Equilibria and Traveling Waves of 2D Navier-Stokes (nu = 1/40)
- AhmedML: High-Fidelity CFD Dataset for Low-Speed Bluff-Body Aerodynamics
- BOOSTR Fermilab Booster Gradient Magnet Power Supply (GMPS) Regulation Time Series, Dec 2019 to Feb 2020 (reduced copy)
- CMS Open Data HH to 4b: Run 2 Collision Data, Hemisphere-Mixed Pseudo-Data and Simulation for Multijet Background Modelling
- The Well: Acoustic Scattering Through Mazes
- NeuralConvection 2D: Boussinesq-to-Compressible Natural Convection Tensors for Fourier Neural Operators (Menglik, Shao, Hyde)
- Tensile2d: 2D Quasi-Static Non-Linear Structural Mechanics under Geometric Variations (PLAID)