metadata
license: apache-2.0
task_categories:
- tabular-regression
tags:
- physics
- computational-fluid-dynamics
- cfd
- thermofluid
- physics-informed-machine-learning
- neural-operator
- pinn
pretty_name: ThermoFluid Demo Dataset
size_categories:
- n<1K
ThermoFluid Demo Dataset
Overview
A demonstration dataset for thermofluid physics AI models.
This dataset provides simulation-based field data for testing physics-informed machine learning, neural operators, and physics foundation models.
Currently, the dataset contains simulation-based CFD field variables including:
- velocity
- pressure
- temperature
Data Format
Each CSV file contains spatial field information:
| Variable | Description |
|---|---|
| x | Spatial x-coordinate |
| y | Spatial y-coordinate |
| u | Velocity x-component |
| v | Velocity y-component |
| pressure | Pressure field |
| temperature | Temperature field |
Dataset Structure
thermofluid-demo/
├── README.md
├── metadata.csv
└── data/
├── case_0001.csv
├── case_0002.csv
└── case_0003.csv
Parameters
Each simulation case includes operating conditions:
- Reynolds number (Re)
- Prandtl number (Pr)
- Inlet velocity
- Inlet temperature
Applications
This dataset can be used for:
- Physics-informed neural networks (PINNs)
- Neural operators (FNO, DeepONet)
- Physics transformers
- Thermofluid foundation models
- CFD surrogate modeling
Version
v0.1 - Initial demonstration dataset
Contents:
- 3 simulation cases
- Velocity, pressure, and temperature fields
- Tabular CFD representation
License
Apache License 2.0