Basic CFD simulation of thermally driven fluid flow and solid-liquid phase change problems
The CFD simulation framework presented here builds upon learning ressources established by others. Lorena Barba published a wonderful course on how to write a basic solver for the Navier-Stokes equations written in Python: 12 steps to Navier Stokes. The course ends with solving a popular 2d test case, the lid-driven cavity. Another brilliant course explaining linear algebra and CFD methods with examples written in Matlab is Computational Science and Engineering I by Gilbert Strang. Many resources and examples can also be found on the web page of the book Computational Science and Engineering. There is also a nice example code of a Navier-Stokes solver for the lid-driven cavity. All this learning material helped me to develop my own solver for solid-liquid phase change problems with natural convection. In the following, this solver is constructed in three steps.
The CFD method presented here starts with the lid-driven cavity flow problem in a first step. In the second step, the code is extended to thermally driven fluid flow (natural convection). Finally, in the third step a model for solid-liquid phase change is added. The final code is able to solve solid-liquid phase change problems with natural convection.
The Python code is found in 1-NS-LDC.py and the documentation is given in the Jupyter Notebook 1-NS-LDC.ipynb.
Pressure contours and velocity vectors of the lid-driven cavity with Re = 20 at steady state after time t = 5 s. |
work in progress ...
work in progress ...