SpectralFluid

An open-source Julia ecosystem

Spectral methods for fluid dynamics

Fast, readable tools for partial differential equations, spherical harmonic transforms, stability analysis, and geophysical flow.

Explore the software Browse the source
pseudo-spectral form

∂ₜu = 𝓛(u) + 𝓕⁻¹[𝓝(û)]

physical spacespectral space
Fourier · Chebyshev · Spherical harmonicsCPU · GPU · MPIOpen source · Julia

01

Software

A family of numerical tools

Each package is useful on its own. Together they cover the path from spectral transforms to complete fluid solvers.

Tarang.jl

PDE framework

A high-performance spectral and pseudo-spectral PDE solver with natural mathematical syntax across CPUs, GPUs, and distributed clusters.

  • Fourier
  • Chebyshev
  • GPU
  • MPI
Docs Source

SHTnsKit.jl

Transform library

Pure-Julia spherical harmonic transforms with scalar and vector analysis, multithreading, and distributed-memory parallelism.

  • Spherical harmonics
  • Pure Julia
  • MPI
Docs Source

GeoDynamo.jl

Geophysical solver

A Julia solver for rotating convection and self-consistent dynamos in spherical shells and full balls.

  • MHD
  • Rotating flows
  • Spherical domains
Docs Source

BiGSTARS.jl

Stability analysis

Bi-global linear stability analysis for geophysical flows using Chebyshev–Fourier spectral collocation.

  • Eigenproblems
  • Collocation
  • Geophysical flows
Docs Source

02

Science

Fields, flows, and modes

Spectral representations expose the structure of a field while retaining the accuracy needed for serious computation.

Ocean-current streamlines forming eddies south of Africa
Ocean circulation. Fine vortical structures embedded within a planetary-scale flow.NASA Scientific Visualization Studio · source
Rows of blue and gold lobes representing real spherical harmonic modes
Spherical harmonics. Global fields decomposed into an orthogonal basis of modes.Inigo Quilez / Wikimedia Commons · CC BY-SA 3.0

03

Approach

Research software you can inspect

SpectralFluid develops numerical software close to the mathematics: explicit algorithms, composable Julia interfaces, and parallel execution paths that scale from a laptop to a cluster.

  • Spectral accuracyfor smooth multiscale physics
  • Parallel by designacross CPUs, GPUs, and MPI
  • Reproducible methodswith open code and documentation

Open research software

Read the code. Reproduce the method. Build the next solver.

Visit GitHub