Closed agzimmerman closed 6 years ago
Actually for the benchmarks in Monolithic simulation of convection-coupled phase-change - verification and reproducibility, it should be straightforward to make a single file which isn't too scary and shows a minimal FEniCS implementation which runs the benchmarks with our model.
Actually let's make a minimal Jupyter notebook for each of the benchmarks.
Complete as of PR #278 .
For the development of Phaseflow and implementing new models, the object oriented approaching using the phaseflow.Simulation class as an API is working very well.
On the other hand, this is probably daunting for anyone who wants to use Phaseflow as a reference implementation for convection-coupled phase-change. I think it would be nice, and not too much effort, to add minimal FEniCS scripts to our docs folder which solve some benchmarks. I wouldn't want to maintain these necessarily, so there is some chance that they will break as FEniCS is updated; but it's fine as long as we write down which version of FEniCS they are originally tested with.