pystormflow

A Python interface to Stormflow, a CFD solver specialized for actuator line simulations of lifting surfaces with the Stormbird library.

Requirements

⚠️ Rust Toolchain Required

This package is distributed as source only and must be compiled during installation. You must have the Rust toolchain installed before installing this package.

Install Rust: Visit https://rust-lang.org/

Stormflow can run parts of the solver on the GPU through wgpu, which requires up-to-date graphics drivers with support for Vulkan, Metal or DirectX 12.

Installation

Once Rust is installed, install pystormflow with pip:

pip install pystormflow

Or install from source:

git clone https://github.com/NTNU-IMT/stormbird.git
cd stormbird/interfaces/pystormflow
pip install .

Use Cases

• Running actuator line simulations with Stormflow through Python scripting
• Post-processing of flow fields and sectional forces directly in Python

Implementation Details

The interface is made using PyO3. A simulation is set up from a JSON string, which can be created using Python's built-in json module for dictionaries, or using the stormbird_setup helper package. See the Stormflow README for a description of the solver and its settings.

from pystormflow import Simulation

with open("setup.json") as f:
    simulation = Simulation(f.read())

simulation.do_steps_until_end_time(end_time=10.0, courant_number=0.5)

simulation.export_fields_as_vtk("fields.vtk", binary=True)
result = simulation.get_stormbird_simulation_result()

Development

For a test build using cargo:

cargo build

To build a distributable source package:

maturin sdist

Publishing to PyPI (maintainer cheat-sheet)

This section is a reminder for the main developer. Only the package owner has upload rights to PyPI.

The package is published as a source distribution only (no pre-built wheels), so users compile it locally with their own Rust toolchain. Do not upload wheels from maturin build.

1. Bump version in Cargo.toml (the Python version is read from there) and update CHANGELOG.md.
2. Remove old build artifacts and build the source distribution:

   rm -rf target/wheels
   maturin sdist
   

3. Optionally check that the sdist installs: pip install target/wheels/pystormflow-*.tar.gz
4. Upload with an API token from PyPI (username __token__, password = the token):

   pip install twine
   twine upload target/wheels/pystormflow-*.tar.gz