Forced Magnetohydrodynamic Turbulence DNS

Image 1

Description

These snapshots are derived from a Direct Numerical Simulation (DNS) of forced, statistically stationary incompressible magnetohydrodynamic (MHD) turbulence, which is publicly available through the Johns Hopkins Turbulence Database (JHTDB). The original simulation was performed on a triply periodic cubic domain of size $2\pi \times 2\pi \times 2\pi$, discretized on a grid of $1024 \times 1024 \times 1024$ nodes, with periodic boundary conditions applied in all three directions. The incompressible MHD equations — the Navier–Stokes equations coupled to the magnetic induction equation through the Lorentz force — are solved using a pseudo-spectral method. Both the velocity field and the magnetic field are forced at large scales to sustain a statistically stationary turbulent state, so that energy injected at large scales cascades to and is dissipated at the smallest resolved scales.

For this machine-learning-oriented release, we subsampled the original JHTDB forced MHD turbulence fields and re-saved them in a new, ML-friendly format. Each snapshot retains the complete three-dimensional flow state with all three velocity components (u, v, w), pressure and all three magnetic magnetic vector potential components (Ax, Ay, Az) . Two versions of the dataset are provided:

  • Sequential version — 20 consecutive snapshots sampled at the stored time interval,
  • Random version — 10 snapshots drawn at random (non-consecutive) times from the statistically stationary regime. This version is intended for tasks that benefit from diverse, decorrelated samples, such as single-frame reconstruction, statistical modeling, or generative learning.

Quick Info

  • Contributors: Charles Meneveau
  • Nɸ = 4 + 3
  • Nx = 1024, Ny = 1024, Nz = 1024
  • DOI
  • .bib
  • Download.sh
ID Description Size (GB) Links
0 Sequential Snapshots 561
1 Collection of snapshots at different time 281

Updated: