Public data to train and test surrogates on.
Each with its licence as written on the data card: read it there before you train on one. How entries are chosen is on About.
- Koratikere et al.Aero & CFD · Structures & crashTransonic wing aerostructural dataset — Coupled aero and structural data for transonic wings2,471 static aerostructural analyses of transonic wings, varied over planform, airfoil sections, Mach number, angle of attack, altitude and load factor, plus three out-of-distribution samples.
huggingface.co/…/transonic-wing-aerostruct-datasetCC BY 4.0arXiv 2610.36442
- Raut et al.Aero & CFDSHIFT-Truck — A high-fidelity aerodynamics dataset for pickup trucks1,000 delayed detached-eddy simulations of a pickup truck varied over 17 shape parameters, on about 100 million cells each, with surface and volume fields and wind-tunnel comparison.
huggingface.co/…/TruckCC BY-NC 4.0arXiv 2609.38638
- Emmi AIGeneral PDEsNeuralDEM — Discrete element and CFD-DEM simulations of hopper discharge and fluidized-bed reactorsSimulation data for training and evaluating surrogates of particulate flows: hopper discharge of about 250,000 particles across hopper angles and friction regimes, and a fluidized-bed reactor of about 500,000 particles with air injection at varying inlet velocities.
huggingface.co/…/NeuralDEMCC-BY-NC 4.0arXiv 2411.09678Public: data
- Luminary CloudThermalSHIFT-Battery — Conjugate heat transfer simulations of battery cold-plate cooling channelsOver 1,300 conjugate heat transfer simulations of serpentine cold-plate cooling channels under a battery module, varied over six channel parameters and the operating conditions, with surface and coolant volume fields and the pressure drop and peak module temperature of each case.
huggingface.co/…/BatteryCC BY-NC 4.0
- Elrefaie et al.Structures & crashCarCrashNet — A large-scale dataset and neural solver for crash simulationAn open crash benchmark: more than 14,000 bumper-beam pole impacts with varied geometry, materials and boundary conditions, and 825 full-vehicle crashes of three public car models, run in OpenRadioss and validated against crash tests and LS-DYNA, with CrashSolver as a full-vehicle baseline.
github.com/…/CarCrashNetLicence not statedarXiv 2605.07098
- Ashton et al.Aero & CFDHiLiftAeroML — A high-fidelity CFD dataset for high-lift aircraft1,800 wall-modelled LES runs of 180 variants of the NASA Common Research Model high-lift configuration, at ten angles of attack from 4° to 22°, on 300–500 million-cell grids, released under CC BY 4.0.
- Luminary CloudStructures & crashSHIFT-Crash — Explicit finite-element frontal crash simulations of a morphed Toyota YarisAbout 5,200 OpenRadioss simulations of a 56 km/h frontal impact into a rigid wall by a 2010 Toyota Yaris model morphed over four body dimensions and front-rail thickness and yield strength, with displacement, stress, plastic strain and energy at 13 instants over 120 ms.
huggingface.co/…/SHIFT-CrashCC BY-NC 4.0
- Luminary CloudAero & CFDSHIFT-Pump — Rotating-frame CFD simulations of centrifugal pump impellers and volutesOver 5,600 parametric centrifugal pump geometries, impeller and volute, simulated with water in a rotating reference frame at and away from the design flow rate, with surface and volume fields and torque and outlet total pressure for each case.
huggingface.co/…/PumpCC BY-NC 4.0
- Luminary Cloud with nTopAero & CFDSHIFT-CCA — A CFD dataset of Group 5 UAV planform variantsSimulations of planform variants of a parameterised Group 5 UAV reference configuration, of the class the card calls Collaborative Combat Aircraft, generated in nTop and run on the Luminary Cloud platform at cruise conditions, with surface and volume fields and integrated forces.
huggingface.co/…/CCAProprietary (Luminary Cloud, all rights reserved)
- Luminary CloudAero & CFDSHIFT-Submarine — A CFD dataset of Myring-type submarine hullsSimulations of geometric variants of a parameterised Myring-type submarine, six of its more than 20 shape parameters varied, run on the Luminary Cloud platform with adaptive meshing, with surface and volume fields and integrated forces.
huggingface.co/…/SubmarineProprietary (Luminary Cloud, all rights reserved)
- Luminary Cloud with Otto Aviation and NVIDIAAero & CFDSHIFT-Wing — Adapted-mesh simulations of parametric NASA CRM wing-body variants for transonic aerodynamicsSimulations of thousands of parametric variants of the NASA Common Research Model, with the wing re-lofted from six spanwise sections and the fuselage tube varied in diameter and length, across a range of flow conditions, with surface and volume fields.
huggingface.co/…/WINGCC BY-NC 4.0
- Luminary Cloud with Honda and NVIDIAAero & CFDSHIFT-SUV — Scale-resolving simulations of morphed AeroSUV variants, at quarter and full scaleTransient DDES runs of thousands of variants of the FKFS AeroSUV, morphed with a deformation cage over design parameters informed by Honda, in estate and fastback form at quarter and full scale, with surface and volume fields.
huggingface.co/…/SUVCC BY-NC 4.0
- Polymathic AIAero & CFD · Energy & fusionThe Well — A Large-Scale Collection of Diverse Physics Simulations for Machine LearningSimulations of biological systems, fluid dynamics, acoustic scattering, magnetohydrodynamics and supernovae, in one format.
huggingface.coCC BY 4.0 (Hugging Face data cards)NeurIPS D&B 2412.00568Public: data
- Ashton et al.Aero & CFDDrivAerML — High-Fidelity Computational Fluid Dynamics Dataset for Road-Car External AerodynamicsHundreds of scale-resolving simulations of parametric car variants, surface and volume.
huggingface.co/…/drivaermlCC BY-SAarXiv 2408.11969Public: data
- Elrefaie et al. (MIT)Aero & CFDDrivAerNet++ — Multimodal road-car dataset with CFD fields and deep-learning benchmarks8,150 car designs (fastback, notchback, estateback) with high-fidelity CFD, surface pressure and wall shear stress, volume fields and geometry, released with benchmark code.
github.com/…/DrivAerNetCC BY-NC 4.0Public: code, data
- Bonnet et al.Aero & CFDAirfRANS — High-Fidelity CFD Dataset for Approximating RANS SolutionsOver a thousand RANS airfoil simulations with full volume fields.
github.com/…/AirfRANSODbL-1.0NeurIPS D&B 2212.07564Public: data
- Takamoto et al. (NEC Labs and University of Stuttgart)General PDEsPDEBench — Benchmark datasets of generic PDEs for scientific machine learning1D, 2D and 3D generic PDEs (advection, Burgers, Darcy, compressible and incompressible Navier–Stokes, shallow water) with varied initial and boundary conditions and parameters, plus code and baselines.
Missing a public dataset or benchmark? Suggest a link
New surrogate models and papers, in your inbox.One email a week, on Saturday morning in your time zone: the week’s news, papers and datasets with their sources, funding kept apart at the bottom.