Skip to content

Latest commit

 

History

History
87 lines (60 loc) · 5.19 KB

README.md

File metadata and controls

87 lines (60 loc) · 5.19 KB

ParaStell

Open-source Python package featuring a parametric 3-D CAD modeling toolset for stellarator fusion devices with additional neutronics support. ParaStell uses plasma equilibrium VMEC data and a user-defined radial build to model in-vessel components of varying thickness in low-fidelity. Furthermore, coil filament point-locus data and a user-defined cross-section are used to model magnet coils. Additional neutronics support includes the use of VMEC data and MOAB to generate tetrahedral neutron source definitions and Coreform Cubit to generate DAGMC geometries for use in Monte Carlo radiation transport software. In addition, an option is included to generate a tetrahedral mesh of the magnets using Coreform Cubit for use in Monte Carlo mesh tallies. A neutron wall-loading utility is included that uses OpenMC to fire rays from a ParaStell neutron source mesh onto a ParaStell first wall CAD geometry.

Dependencies

ParaStell depends on:

Install ParaStell

Download and extract the ParaStell repository:

git clone [email protected]:svalinn/parastell.git

or download the ZIP file from the repository home page. Once extracted, add the repository directory to your PYTHONPATH.

Install Python Dependencies

This guide will use the conda package manager to install Python dependencies. Conda provides straight-forward installation of Python packages and switching between different collections of Python packages through the use of environments.

If you have not already installed conda, you can use one of the following installers to do so:

A working conda environment with all ParaStell Python dependencies can be found in this repository's environment.yml file. To create the corresponding parastell_env conda environment on your machine, create the environment from the environment.yml file and activate the new environment:

conda env create -f environment.yml
conda activate parastell_env

Alternatively, view INSTALL.md for instructions on manually installing these Python dependencies using mamba.

Install Coreform Cubit

Download and install version 2023.11 from Coreform's Website, then add the /Coreform-Cubit-2023.11/bin/ directory to your PYTHONPATH by adding a line similar to the following to your .bashrc file:

export PYTHONPATH=$PYTHONPATH:$HOME/Coreform-Cubit-2023.11/bin/

Replace $HOME with the path to the Coreform Cubit directory on your system. Additional information about adding modules to your PYTHONPATH can be found here. While it is possible to use ParaStell with older versions of Cubit, additional steps not in this guide may be required.

If you do not have a Coreform Cubit license, you may be able to get one through Cubit Learn at no cost.

Executing ParaStell Scripts with YAML Input

While ParaStell can imported as a module to make use of its Python API, ParaStell also has an executable to alternatively call functionality via command line. This executable uses a YAML configuration file as a command-line argument to define input parameters.

To make use of this feature, add the ParaStell executables directory to your PATH by adding a line similar to the following to your .bashrc file:

export PATH = $PATH:$HOME/parastell/executables/

Replace $HOME with the path to the ParaStell repository directory on your system. Information about adding directories to your PATH can be found here.

Next, give any files in the executables directory file execution permission:

chmod -R u+x $HOME/parastell/executables/

Now, the executable can be run from command line with a corresponding YAML file argument. For example:

parastell config.yaml

See the executable's help message for more details.

Citing

If referencing ParaStell in a document or presentation, please cite the following publication:

  • Connor A. Moreno, Aaron Bader, and Paul P.H. Wilson, "ParaStell: parametric modeling and neutronics support for stellarator fusion power plants," Frontiers in Nuclear Engineering, 3:1384788 (2024). DOI: 10.3389/fnuen.2024.1384788