FEniCSx & GMSH Tutorial Repository (v0.9.0)

FEniCSx & GMSH Tutorial Repository (v0.9.0)#

This repository contains all materials, scripts, and documentation for a Finite Element Modeling workshop, focusing on FEniCSx v0.9.0 and GMSH v4.11+. The project is directly related to Thomas Lavigne’s AFR-FNR research, a collaborative effort involving the University of Luxembourg, Bordeaux University (I2M), and ENSAM Paris (IBHGC); and to the workshop conducted at I2M Bordeaux of the 9th-11th September 2024.

The primary goal is to guide users through advanced finite element methods using FEniCSx and robust, complex mesh generation with GMSH.

Please give a look here https://th0maslavigne.github.io/FEniCSx_GMSH_tutorials/README.html

1. Repository Structure Overview#

The repository is logically divided into Installation methodologies and practical Tutorials.

Directory

Content Summary

Install/

Scripts and comprehensive documentation for setting up the FEniCSx environment on various platforms: Windows (WSL2, Docker) and High-Performance Computing (HPC) systems (Spack).

Tutorials/

Python scripts demonstrating a range of finite element simulations (Elasticity, Fluid Dynamics, Thermal Analysis) and advanced GMSH meshing techniques.

Core Installation Sub-Folders Breakdown#

Sub-Folder

Target Platform

Primary Files

WSL_Docker_installer

Windows (WSL2 + Docker)

install_wsl2_fenicsx_docker.bat, FEniCSx_interactive.bat

WSL_Local_installer

Windows (WSL2 Native)

install_local_fenicsx.bat, apt-installed-list.txt

Windows_docker_installer

Windows (Docker Desktop Direct)

FEniCSx_docker_0-9-0.bat, Interactive_FEniCSx_0-9-0.bat

Spack_Install_FEniCSx (MCIA or Cassiopee)

Linux / HPC

Install Procedures and batch files

2. Requirements & Environment Setup#

To successfully execute the provided examples, the following components are necessary:

  1. FEniCSx v0.9.0 runtime environment (Requires one of the installation methods outlined below).

  2. GMSH software (version >4.11) for mesh generation.

  3. ParaView for post-processing and visualization.

Installation Guidance#

Detailed, step-by-step instructions for all setup methods are located in the Install/Install_processes.md file.

For Windows users, the WSL2 Docker Installer (Install/WSL_Docker_installer/install_wsl2_fenicsx_docker.bat) is the most recommended approach, offering a self-contained, high-performance Linux environment.

Here-after are links to the official download pages:

3. Usage and Execution Launchers#

Environment Aliases (WSL2 Docker Method)#

After running the WSL2 Docker Installer, custom aliases are available inside the resulting WSL distribution (FEniCSxenv):

Alias

Function

Description

fenicsx

Interactive Shell

Launches an interactive Bash shell inside the FEniCSx container.

fenicsx-run

Run Script/Command

Executes a single command or Python script non-interactively.

fenicsx-jupyter

Jupyter Lab

Launches Jupyter Lab, exposing it for access via your web browser.

Note on Graphical Interfaces (Crucial)#

By default, Docker containers lack a graphical environment. Any script that attempts to open a GUI window (e.g., GMSH’s interactive viewer or certain Matplotlib backends) within the container may fail or crash.

To prevent this when running GMSH-related scripts, you must pass the close argument to suppress the GUI:

# Example of running a GMSH script non-interactively
fenicsx-run "python3 Tutoriels/Other_GMSH_Examples/chip5x5.geo close"

4. FEniCSx & GMSH Workshop Contents#

The workshop is designed to provide a comprehensive introduction to meshing with GMSH and finite element analysis with FEniCSx.

4.1. Finite Element Method (FEM) Overview#

  • A brief reminder about the Finite Element Method fundamentals.

4.2. GMSH Mesh Creation Techniques#

The workshop explores versatile methods for generating complex meshes, focusing on control and export compatibility:

  • Geometry Definition:

    • Creation from a sketch.

    • Creation from elementary geometries.

    • Utilizing symmetries and boolean operations.

    • Generating meshes from a pre-existing STL file.

  • Advanced Meshing:

    • Implementing local refinement procedures (referencing Lavigne et al., 2023).

    • Boundary and domain tagging for material and boundary condition mapping.

  • Export:

    • Export compatibility: FEniCSx vs. Fenics Legacy.

4.3. FEniCSx Computation and Problem Solving#

The computational portion of the workshop covers a range of physics and solution methods:

  • Problem Types:

    • Thermal Problems: Stationary and transient analysis.

    • Fluid Dynamics: Stokes Equation solving in 2D and 3D.

    • Poro-Mechanics: Terzaghi poromechanical model (referencing Lavigne et al., 2023).

    • Solid Continuous Mechanics:

      • Elastic and hyper-elastic problems.

      • Problems involving penalty contact.

      • Incremental resolution (for updated Lagrangian formulation).

      • Evaluation of a quantity (e.g., reaction force).

  • Numerical Methods:

    • Linear and Non-Linear resolutions.

    • Updated mesh resolution (mesh update).

For updated lagrangian please contact Carla Cornillon

4.4. Key Academic References and Resources#

The specialized topics (local refinement, poromechanics, and incremental resolution) are supported by the author’s published work, available in full on their GitHub repositories:

  • Porous Media and Local Refinement:

  • Incremental Resolution and Mesh Update:

    • Lavigne et al., 2024 (Work presented for an example of incremental resolution and mesh update.)

    • This includes the use of volume tags for material mapping and reaction force evaluation.

    • GitHub Repository: Skin Porous Modelling

5. Utility Code Snippet#

A simple Python function is provided here for exporting data lists to a CSV file:

def export_to_csv(data, filename, header=None):
    import csv
    try:
        with open(filename, 'w', newline='') as file:
            writer = csv.writer(file)
            if header:
                writer.writerow(header)
            writer.writerows(data)
        print(f"Data exported to {filename} successfully")
    except Exception as e:
        print(f"An error occurred while exporting data to {filename}: {e}")

6. Acknowledgments and Licensing#

Research and Funding#

This activity is a core component of Thomas Lavigne’s PhD work. This research was entirely or partially funded by the Luxembourg National Research Fund (FNR), under grant reference No. 17013182.

Open Access License#

For the purpose of open access, the author has applied a Creative Commons Attribution 4.0 International (CC BY 4.0) license to this work.

Inspiration and Support#

The author extends thanks to:

  • Jørgen S. Dokken and Christophe Geuzaine, whose exemplary work and tutorials served as inspiration for this repository.

  • F. Daghia and Ludovic Chamoin for the high-quality finite element courses delivered at the École Normale Supérieure Paris-Saclay.

  • Jack Hale and Stéphane Urcun for their invaluable assistance in code debugging throughout the PhD journey.

  • Giuseppe Sciumè for the invitation to co-organize the workshop at the Bordeaux University.

Contents