OpenFOAM post-processing

Python utilities for reading OpenFOAM fields and resampling them onto regular NumPy grids. Originally published as pyFoamPP, the project uses fluidfoam for field access and local radial-basis interpolation for unstructured meshes.

Quick start

Python 3.10 or newer is recommended.

git clone https://github.com/EngFlavioMartins/openfoam-postprocessing.git
cd openfoam-postprocessing
python -m venv .venv
source .venv/bin/activate
python -m pip install -r requirements.txt
python examples/plot_sample.py --output outputs/sample.svg

On Windows, activate with .venv\Scripts\activate. The example reads the bundled OpenFOAM snapshot directly and saves a velocity-sample plot. An OpenFOAM solver installation is not required to read it.

Notebooks

python -m pip install jupyterlab
python -m jupyter lab

Open Creating_Backups.ipynb to resample a case, or Examples.ipynb to inspect processed fields. Run notebooks from the repository root.

For a new unstructured case, place it under OFBackups/<case-name>/ and set its time, bounds and grid spacing:

from Libs.Subroutines import preProcess

preProcess(
    "429", "OFData",
    structured=False,
    domain_bounds=(8, 14, 0.2, 2.5, 8, 12),
    grid_spacing=0.15,
)

This call reads velocity U and pressure p, then writes a binary mesh backup to Data/OFData. Choose bounds inside the useful part of your own case.

Repository guide

Location	Purpose
Libs/Subroutines.py	Loading, interpolation and mesh container
examples/plot_sample.py	Direct-from-OpenFOAM plotting example
Creating_Backups.ipynb	Case-to-array workflow
Examples.ipynb	Analysis examples
OFBackups/OFData/	Bundled OpenFOAM snapshot
Data/	Historical processed example
docs/assets/header.svg	Conceptual vector overview

Data and numerical scope

The resampler uses five neighbours and a linear RBF kernel. It is not conservative remapping and can extrapolate beyond the sampled domain. Check resolution, domain bounds and interpolation error before using the arrays quantitatively.

The original structured=True branch is incomplete; use the documented unstructured path. Its backup writer appends objects to an existing file, so use a fresh case name or move an old backup aside first. Only load pickle backups from trusted sources; the quick-start example avoids them.

Contributing and attribution

See CONTRIBUTING.md. Maintained by Flavio Martins.

Built on fluidfoam and SciPy. The older README referred to MIT, but no licence file is present; no licence has been added or changed here.