FEniCSx examples at CISM 2024

FEniCSx examples at CISM 2024#

Introduction#

This repository contains source code for the computational examples presented at the 9th CISM-ECCOMAS Advanced Course on Variational Fracture Mechanics and Phase-Field Models taking place between July 1st 2014 and July 5th 2024.

The built book can be read at https://jhale.github.io/cism-2024-varfrac-code/. The git repository can be found at jhale/cism-2024-varfrac-code.

Updates#

  • 12/2024; Bug fixes in reduced space construction in stability solver.

  • 11/2024: Update for DOLFINx v0.9.0.

Installation instructions#

The recommended approach for installing the software is using Docker. By using Docker we ensure a consistent software environment across all participants and minimal use of the wireless network. As alternatives, we also provide instructions for Google Colab and Anaconda (conda).

Docker#

  1. Install Docker Desktop on your system following the instructions here.

  2. Pull the DOLFINx laboratory image:

    docker pull dolfinx/lab:v0.9.0r1

  3. (macOS, Linux). Start a DOLFINx laboratory container using a Unix-like shell:

    mkdir ~/cism-varfrac-course
cd ~/cism-varfrac-course
docker run -ti -v "$(pwd)":/shared -p 8888:8888 -w /shared dolfinx/lab:v0.9.0r1

  4. (Windows Powershell). Start a DOLFINx laboratory container using Powershell:

    mkdir ~/cism-varfrac-course
cd ~/cism-varfrac-course
docker run -ti -v "${pwd}:/shared" -p 8888:8888 -w /shared dolfinx/lab:v0.9.0r1

  5. (Windows cmd). Start a DOLFINx laboratory container using Windows cmd.

    mkdir %HOMEPATH%\cism-varfrac-course
cd %HOMEPATH%\cism-varfrac-course
docker run -ti -v "%cd%":/shared -p 8888:8888 -w /shared dolfinx/lab:v0.9.0r1

  6. A URL e.g. http://127.0.0.1:8888/lab?token=544f7380ab06eb1d175d8c2b35a362e7fd7a29471b56818c will be printed to the terminal. Copy this to a web browser. You should see the Jupyter notebook environment open.

  7. Click File > New > Notebook in the top menu. Use the Python 3 (ipykernel) kernel by pressing Select. In the first notebook cell type:

    import dolfinx
import pyvista
import gmsh

    and then press Shift + Enter to execute the cell. You should receive no errors (e.g. ModuleNotFoundError).

Conda (macOS and Linux)#

  1. Download the Anaconda Python distribution.

  2. Create and activate a new conda enviroment:

    mkdir ~/cism-varfrac-course
cd ~/cism-varfrac-course
curl -o environment.yml https://raw.githubusercontent.com/jhale/cism-varfrac-code/main/environment.yml
conda env create -f environment.yml
conda activate fenicsx-cism-2024

  3. Launch a Jupyter lab environment in your browser:

    jupyter-lab .

  4. Click File > New > Notebook in the top menu. Use the Python 3 (ipykernel) kernel by pressing Select. In the first notebook cell type:

    import dolfinx
import pyvista
import gmsh

    and then press Shift + Enter to execute the cell. You should receive no errors (e.g. ModuleNotFoundError). If macOS asks about accepting incoming network connections, click Allow.

Google Colab#

  1. Go to Google Colab and create a new notebook. We will use the FEM on Colab project to install FEniCSx. Copy and paste into a new notebook cell:

    try:
    import dolfinx
except ImportError:
    !wget "https://fem-on-colab.github.io/releases/fenicsx-install-real.sh" -O "/tmp/fenicsx-install.sh" && bash "/tmp/fenicsx-install.sh"
import dolfinx

    and press Shift+Enter. You should see output from the install process and no errors.

Getting up and running#

We recommend cloning the repository to run the notebooks using git.

From a terminal inside Jupyterlab:

cd /shared
git clone https://github.com/jhale/cism-2024-varfrac-code.git
cd cism-2024-varfrac-code

If git does not work you can also download the current version using:

cd /shared
curl -L -o main.tar.gz https://github.com/jhale/cism-2024-varfrac-code/archive/refs/heads/main.tar.gz
tar -xvf main.tar.gz 
cd cism-2024-varfrac-code-main

Users using Google Colab should paste the following content into a cell:

from google.colab import drive
drive.mount('/content/drive')
%cd "/content/drive/MyDrive/Colab Notebooks" 
!git clone https://github.com/jhale/cism-2024-varfrac-code.git
%cd cism-2024-varfrac-code

To open the notebook in Colab navigate using Google Drive to Colab Notebooks/ and double click on it.

Authors#

  • Jack S. Hale, University of Luxembourg.

  • Laura de Lorenzis, ETH Zurich.

  • Corrado Maurini, Sorbonne Université.

Credits#

This work includes elements from Computational fracture mechanics examples with FEniCSx under the terms of the BSD license.

This work includes elements from NewFrac FEniCSx training under the terms of the MIT license.

Developer instructions#

Building#

To build this Jupyter book run:

docker run -v $(pwd):/shared -w /shared -ti --entrypoint /bin/bash dolfinx/lab:v0.9.0r1 
pip install -r requirements-docs.txt
export PYVISTA_OFF_SCREEN=true
export PYVISTA_JUPYTER_BACKEND=html
jupyter-book build -W .

and then on the host:

open _build/html/index.html

On ARM it is necessary to install pyvista from a custom binary wheel, see below.

Linting and formatting#

To lint and format using ruff:

pip install ruff
ruff check .
ruff format .

Converting .ipynb to .py#

This repository stores stores demos in percent-formatted .py files that can then be converted to e.g. notebooks on demand. To convert legacy notebooks e.g. notebook.ipynb to this format automatically use:

pip install jupytext
jupytext --to py notebook.ipynb

which will create a file notebook.py which you can then edit with any text editor. For more information see here.

pyvista on ARM#

To install pyvista in ARM docker:

python3 -m pip install "https://github.com/finsberg/vtk-aarch64/releases/download/vtk-9.2.6-cp310/vtk-9.2.6.dev0-cp310-cp310-linux_aarch64.whl"
python3 -m pip install pyvista