mem4py - A membrane finite element solver based on kinetic dynamic relaxation

This tool provides a solution method to solve the deformations of pressurized membrane structures. The steady-state is found with the method of kinetic dynamic relaxation which proved its effectiveness for a variety of membrane structures.

Installation:

mem4py is written in cython which is a C-compliable python script. In benefits from C like speedup and can be called using a simple python script.

1. Get eigen3 either with

sudo apt-get install libeigen3-dev

or download under http://eigen.tuxfamily.org.

2. Clone mem4py to folder of choice

git clone https://github.com/pthedens/mem4py.git

3. If eigen3 is not located at /usr/include/eigen3 or C:\Program Files\eigen3 change EIGEN_PATH to eigen3 in setup.py

4. Compile in mem4py/ with

"python setup.py build install"

or using pip

"pip install -e ."

Windows

Compiler

Install a cython compatible compiler (Visual Studio/Windows SDK C/C++) if you don't already have a C/C++ compiler.

Setup environment and dependencies with Anaconda/Miniconda

# Assuming you are running a recent python 3.x version
conda create --name mem4py cython numpy matplotlib scipy
# Run the build script
python setup.py build install
cd tutorials && python main.py

Test

In ./tutorials run main.py to test example cases.

Setting up your own problem

You can set up your own problem using gmsh and a python script. In gmsh you define the geometry and boundary conditions, and the python script is used for material and solver properties.

Mesh:

A surface mesh in gmsh format has to be provided in the /msh folder.

• Only 3 node triangular shells are implemented as membrane elements.

• Boundary conditions are defined on physical surfaces, lines or points

Material and solver properties:

In a python script the properties are defined in a dict which is fed to the solver. Closely follow the example case in the tutorials folder.

License:

The software is licensed under MIT.

Acknowledgments

mem4py extensively uses a cython version of eigen3 which is originally based on https://github.com/strohel/Ceygen. Many thanks to the developers to make eigen3 cython compatible.

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No 642682 (AWESCO).