← Datasets
Dataset

DrivAerNet++

Multimodal road-car dataset with CFD fields and deep-learning benchmarks

Elrefaie et al. (MIT) · Aero & CFD · Public: code, data

Data Mohamedelrefaie/DrivAerNet⁠

8,150 car designs (fastback, notchback, estateback) with high-fidelity CFD, surface pressure and wall shear stress, volume fields and geometry, released with benchmark code.

Licence
CC BY-NC 4.0
Access
Harvard Dataverse; splits in the repository
Cases
8,150 car designs
Fields
Surface Cp and WSS, volume fields
Size
39 TB

Used by

  1. Paper · TransformersTransolver without attention — Does Transolver really need a Transformer?
  2. Paper · TransformersTransolver++ — An Accurate Neural Solver for PDEs on Million-Scale Geometries
  3. Paper · PretrainingCross-domain pretraining — Cross-domain pretraining for steady neural CFD surrogates
  4. PaperCarBench — A common benchmark for neural surrogates of 3D car aerodynamics
  5. Paper · OperatorsGAOT — A geometry-aware operator transformer for PDEs on arbitrary domains

Share on LinkedIn, X or Bluesky, or or .

Know related work that belongs here?

Something wrong on this page?

More datasets

All datasets
  1. Transonic wing aerostructural dataset — Coupled aero and structural data for transonic wings
  2. SHIFT-Truck — A high-fidelity aerodynamics dataset for pickup trucks
  3. HiLiftAeroML — A high-fidelity CFD dataset for high-lift aircraft