Level-Set Reinitialization Project

This repository contains an implementation of reinitialization methods for the Level-Set method, using FEniCSx and CutFEMx.

Project Structure

• main.py: Main entry point for the reinitialization process.
• data.py: Configuration file for simulation parameters.
• src/: Contains the main source code (library).
  • reinitialization.py: Main class for reinitialization.
  • reinitialization_mpi.py: MPI-compatible version.
  • mesh_generation.py: Mesh generation utilities.
  • geometry_ls.py: Geometry definitions and noise application.
  • compute_error.py: Error computation functions.
• scripts/: Execution and demonstration scripts.
  • demo_convergence_2d.py: Convergence study on a 2D case (circle).
  • demo_3d.py: 3D demonstration (torus/sphere).
  • demo_mpi.py: Demonstration using MPI.
• requirements.txt: List of Python dependencies.

Quick Start

To run the standard reinitialization process:

1. Configure your simulation parameters in data.py (mesh size, level set function, noise, etc.).
2. Run the main script:

python3 main.py

Results will be saved in the res/ directory.

Installation

It is recommended to use a Conda environment with FEniCSx installed.

# Example installation with Conda
conda create -n fenicsx-env
conda activate fenicsx-env
conda install -c conda-forge fenics-dolfinx mpich pyvista matplotlib
pip install -r requirements.txt

# Install cutfemx
# Please refer to the official repository for installation instructions:
# https://github.com/sclaus2/CutFEMx

Make sure the project root folder is in your PYTHONPATH or run the scripts from the root as follows.

Usage

To run the demonstrations, execute the scripts from the project root:

2D Convergence

python3 scripts/demo_convergence_2d.py

3D Demonstration

python3 scripts/demo_3d.py

MPI Demonstration

mpirun -n 4 python3 scripts/demo_mpi.py

Authors

Amina El Bachari 2025 ONERA and MINES Paris - PSL

Citation

If you use this code in your research, please cite the following paper:

A predictor-corrector scheme for approximating signed distances using finite element methods Amina El Bachari, Johann Rannou, Vladislav Yastrebov, Pierre Kerfriden, Susanne Claus Preprint, 2025 Read the paper