Published February 27, 2024 | Version 2.1
Dataset Open

BLASTNet Simulation Dataset

Authors/Creators

  • 1. Stanford University
  • 2. Sandia National Labs
  • 3. University of Melbourne
  • 4. Polytechnique Montréal
  • 5. University of Connecticut
  • 6. ROR icon Google (United States)
  • 7. ROR icon Centre Européen de Recherche et de Formation Avancée en Calcul Scientifique

Description

Go to https://blastnet.github.io/ to access and download this reacting and non-reacting flow physics simulations.

Mission

BLASTNet 2.1 was developed to provide the researchers in  reacting and non-reacting flow physics communities with high-fidelity simulation datasets in a convenient format for ML applications. With ~5 TB, 765 full-domain samples, and 36 configurations, BLASTNet can effectively address these gaps and aid in fostering open/fair ML development within reacting and non-reacting flow physics communities.

Application

This data is useful for fluid flows in a wide range of ML applications tied to automotive, propulsion, energy, and the environment. Specifically, scientific engineering tasks related to these domains may include turbulent closure modeling, spatio-temporal modeling, and inverse modeling.
    

Notes

URL: https://blastnet.github.io/

Files

Files (5.1 kB)

Name Size Download all
md5:b7ea010a05426fb01cad2682b0da2135
5.1 kB Download

Additional details

Related works

Documents
Conference paper: 10.48550/arXiv.2309.13457 (DOI)
Is cited by
Journal article: 10.1016/j.jaecs.2022.100087 (DOI)
Conference paper: 10.48550/ARXIV.2207.12546 (DOI)
Is supplement to
Journal article: 10.1016/j.jaecs.2022.100087 (DOI)
Conference paper: 10.48550/ARXIV.2207.12546 (DOI)

Dates

Updated
2024-02-27
Version 2.1

References

  • Wai Tong Chung, Bassem Akoush, Pushan Sharma, Alex Tamkin, Ki Sung Jung, Jacqueline Chen, Jack Guo, Davy Brouzet, Mohsen Talei, Bruno Savard, Alexei Y Poludnenko, Matthias Ihme (2023). Turbulence in Focus: Benchmarking Scaling Behavior of 3D Volumetric Super-Resolution with BLASTNet 2.0 Data, Advances in Neural Information Processing Systems (NeurIPS) 36.