{"success":true,"data":{"author":"Ratio Labs (Haitz Saez de Ocariz Borde, Flavio Savarino et al.)","bucketAccessType":"public","bucketName":"ratiolabs/downforce-airfoils","bucketPrefix":"","category":"Physics","cloudProvider":"huggingface","createdAt":"2026-11-16T02:38:29.544Z","description":"Steady incompressible 2D RANS simulations of downforce-generating, negative-camber NACA 4-digit airfoils, released by Ratio Labs with the Airfoil2Vec paper (arXiv:2609.38213) to cover the road-vehicle and motorsport regime (rear and front wings) that aircraft-oriented airfoil datasets leave out. 907 valid geometries (m from 1 to 8, p from 1 to 8, thickness 4 to 18 percent, sharp trailing edge) are each run at 11 angles of attack from -10 to 0 degrees, giving 9,977 converged cases at Re = 4.66e6 and Mach 0.2 with OpenFOAM simpleFoam and the k-omega SST turbulence model on structured C-type meshes of about 276k cells. Each case is one HDF5 file holding the per-cell volume field (coordinates, pressure, velocity, temperature, Stanton number), extra turbulence quantities, a per-panel surface field (Cp, Cf, wall shear), a 1,999-point spline contour and integrated Cl, Cd and Cm. Splits are train (860 geometries), test (47 held-out NACA geometries) and non_naca (S1223, E423, LS(1)-0413, Clark Y and FX 63-137 for geometry extrapolation); metadata Parquet files index every case. About 420 GB; read with h5py and filter cases with Hugging Face datasets. CC-BY-4.0; cite Saez de Ocariz Borde et al. 2026.","displayTitle":"Downforce Airfoils: 9,977 RANS CFD Cases of Inverted-Camber NACA Airfoils for Motorsport Aerodynamics","id":"637c0c51-8a12-50d4-8a42-51345f916b86","license":"CC-BY-4.0","licenseEvidenceUrl":"https://huggingface.co/datasets/ratiolabs/downforce-airfoils","objectCountEstimate":10037,"probeStatus":"ok","referenceUrl":"https://arxiv.org/abs/2609.38213","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"ratiolabs/downforce-airfoils\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"ratiolabs/downforce-airfoils\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=ratiolabs, license:cc-by-4.0, size_categories:10K<n<100K, format:parquet, downloads=2826"],"format":"mixed","generatedAt":"2026-11-16T02:38:36.511Z","layout":{"example":"metadata_non_naca.parquet","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"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","encoding":"hex","key":"metadata_non_naca.parquet","truncatedAt":512}],"sampleObjects":["metadata_non_naca.parquet","metadata_test.parquet","metadata_train.parquet","non_naca/clarky_inverted_AoA_-10deg.h5","non_naca/clarky_inverted_AoA_-1deg.h5"],"sidecars":[{"bytes":"---\nlicense: cc-by-4.0\ntask_categories:\n  - other\ntags:\n  - cfd\n  - aerodynamics\n  - airfoil\n  - rans\n  - openfoam\n  - fluid-dynamics\n  - physics-simulation\n  - surrogate-modeling\npretty_name: Downforce Airfoils CFD Dataset\nsize_categories:\n  - 1K<n<10K\nconfigs:\n  - config_name: metadata\n    data_files:\n      - split: train\n        path: metadata_train.parquet\n      - split: test\n        path: metadata_test.parquet\n      - split: non_naca\n        path: metadata_non_naca.parquet\n---\n\n# Downforce Airfoils\n\nSteady, incompressible, two-dimensional subsonic RANS CFD simulations of\ndownforce-generating (inverted-camber) NACA 4-digit airfoils, released by\n[Ratio Labs](https://huggingface.co/ratiolabs).\n\n**Paper:** [Airfoil2Vec: Spectral Geometry-Conditioned Neural Surrogate Models for\nAirfoil Aerodynamics and a Downforce-Generating CFD Dataset](https://arxiv.org/abs/2609.38213)\n(arXiv:2609.38213)\n\nThe collection targets a regime that is comparatively underrepresented in\nexisting public airfoil CFD datasets for machine learning: geometries and\noperating conditions relevant to road-vehicle / motorsport aerodynamics\n(rear wings, front wings, and other inverted-lift-generating surfaces),\nrather than aircraft wings.\n\n## Dataset summary\n\n- **907** valid airfoil geometries, generated from the NACA 4-digit family\n  with **negative camber** (`m < 0`, `m∈[1,8]`, `p∈[1,8]`, `t∈[4,18]`) and a\n  sharp trailing edge, to produce downforce-oriented sections.\n- **11** angles of attack per geometry, `α ∈ [-10°, 0°]`.\n- **9,977** converged CFD cases in total (907 × 11, minus a few\n  non-converging combinations).\n- Solver: steady incompressible RANS with the `k-ω SST` turbulence model,\n  OpenFOAM (`simpleFoam`).\n- Flow conditions: `Re = 4.66e6`, `M∞ = 0.2` (clean, straight-line\n  operating condition).\n- Mesh: automatically generated structured C-type mesh, ~276k cells per\n  case (converged per a grid-independence study on NACA6412 at ~177k\n  cells; a finer mesh was used for robustness across the full geometry\n  sweep).\n\n## Splits\n\n| Split      | # files | # geometries | Description                                   |\n|------------|--------:|-------------:|------------------------------------------------|\n| `train`    |   8,600 |          860 | Training split                                  |\n| `test`     |   1,377 |           47 | Held-out NACA geometries (no overlap w/ train)  |\n| `non_naca` |      55 |             5 | Non-NACA airfoils, pure geometry extrapolation  |\n\n`non_naca` contains 5 airfoils typical of motorsport aerodynamics — NASA\nLS(1)-0413, Selig S1223, Eppler E423, Clark Y, and Wortmann FX 63-137 — each\nat the same 11 angles of attack, none seen during any training. Use it to\ntest generalization beyond the NACA 4-digit family, not as a training or\nvalidation split.\n\n## File format\n\nEach case is one HDF5 file, named `<geometry>_AoA_<angle>deg.h5`, e.g.\n`NACA6412_inverted_AoA_-5deg.h5`. Every file contains:\n\n| Dataset             | Shape        | Description                                               |\n|----------------------|-------------|-------------------------------------------------------------|\n| `flow_field`         | `(N, 9)`    | Per-cell volume field. Columns (see `column_headers` attr): `x, y, z, pressure, u_x, u_y, u_z, temperature, stanton_number` |\n| `extra_flow_field`    | `(N, 10)`   | Additional per-cell derived/turbulence quantities |\n| `surface_field`       | `(M, 10)`   | Per-panel surface quantities (Cp, Cf, wall shear, ...) |\n| `spline_coordinates`  | `(1999, 2)` | Ordered `(x, y)` points describing the airfoil contour (chord-normalized) |\n| `summary`             | attrs       | Integrated aerodynamic coefficients (`Cl`, `Cd`, `Cm`, ...) |\n\n`N` ≈ 276,000 cells per case (varies slightly by mesh).\n\n## Browsing the dataset (viewer)\n\nEach `.h5` file stores several arrays of *different* lengths per case\n(`flow_field` has ~276,000 rows, `spline_coordinates` has 1,999,\n`surface_field` has ~410, `summary` is scalar attributes) — a heterogeneous\nscientific HDF5 layout, not a single flat table, s","encoding":"utf8","key":"README.md","name":"README.md","size":6576,"truncated":true}],"sourceObjectKey":"metadata_non_naca.parquet","starterCell":"from datasets import load_dataset\nds = load_dataset(\"ratiolabs/downforce-airfoils\")\nprint(ds)","starterSamplePath":"metadata_non_naca.parquet","totalObjects":10037},"summary":"Steady incompressible 2D RANS simulations of downforce-generating, negative-camber NACA 4-digit airfoils, released by Ratio Labs with the Airfoil2Vec paper (arXiv:2609.38213) to…","title":"Downforce Airfoils: 9,977 RANS CFD Cases of Inverted-Camber NACA Airfoils for Motorsport Aerodynamics","updatedAt":"2026-11-16T17:15:30.874Z","visibility":"public","websiteUrl":"https://huggingface.co/datasets/ratiolabs/downforce-airfoils"}}