Hugging Face's logo

Datasets:
ArielLubonja
/
johns-hopkins-turbulence-database

Tasks:
Other
Size:
1B<n<10B
License:
Dataset card Files
xet
Community
Dataset Viewer

The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.

Johns Hopkins Turbulence Database

👉 Please visit: https://turbulence.idies.jhu.edu

There you will find:

  • Homogeneous Turbulence: Isotropic 1024³, 2048³, 4096³, 8192³ and 32768³
  • Wall-Bounded Turbulence: Channel Flow & Transitional Boundary Layer
  • Geophysical
  • Wind Farm Simulations

You will also be able to access these datasets in their full size, unlike the subset presented in this HuggingFace dataset

What is in this HuggingFace Dataset?

This repository hosts a small 3-D snapshot (256³ grid points) extracted from the publicly available Johns Hopkins Turbulence Database (JHTDB), meant for quick experimentation, tutorials, benchmarking, and as a lightweight entry point before connecting to the full database.

  • isotropic_1024_coarse_pressure.h5

    • shape: (256, 256, 256, 1, 10) - 256³ across 10 timesteps
    • fields: p - pressure
    • dtype: float32
    • data format: hdf5
    • size: 688 MB

  • isotropic_1024_coarse_velocity.h5

    • shape: (256, 256, 256, 3, 10) - 256³ across 10 timesteps, 3 fields
    • fields: u,v,w - three velocity direction components
    • dtype: float32
    • data format: hdf5
    • size: 2.02 GB

Dataset Description

Direct numerical simulation (DNS) using 1,024³ nodes. Only 256³ is present in the HuggingFace files.

✪ Navier-Stokes is solved using pseudo-spectral method.

✪ Energy is injected by keeping constant the total energy in shells such that |k| is less or equal to 2.

✪ After the simulation has reached a statistical stationary state, 5,028 frames of data with 3 velocity components and pressure are stored in the database. Extra time frames at the beginning and at the end have been added to be used for temporal-interpolations.

✪ The Taylor-scale Reynolds number fluctuates around Rλ~ 433.

✪ There is one dataset ("coarse") with 5028 timesteps available, for time t between 0 and 10.056 (the frames are stored at every 10 time-steps of the DNS). Intermediate times can be queried using temporal-interpolation.

✪ There is another dataset ("fine") that stores every single time-step of the DNS, for testing purposes. Times available are for t between 0.0002 and 0.0198.

✪ A table with the time history of the total kinetic energy and Taylor-scale Reynolds number as function of time can be downloaded from this text file.

✪ Radial spectrum E(k) averaged over time can be downloaded from this text file.

Downloads last month
102