Updating legacy fenics code

dolfinx
1

Hello Fenicsx community,

I am attempting to upgrade some existing code depending on the legacy fenics package to fenicsx 0.72. The problem involves a variation of the poisson-boltzmann equations and a tetrahedral mesh with multiple subdomains.

I believe that I have successfully created the tetrahedral mesh in the new doflinx xdmi format (at least it looks right in paraview). However, my doflinx implementation is not returning the correct results. If anyone is able to provide some suggestions at how to correct the implementation I would be very grateful!

Below I will try to provide the files necessary to reproduce the original legacy fenics simulation, and my attempt at porting to dolfinx.

fenicsx files

fenics files

each contains mesh files, a python script, and some additional parameter data.

2

The files for this is missing.
In general, I would strongly recommend you to upload code directly to Discourse using

```python
# add python code here
```

as links to drives come and go and makes the reproducibility unreliable.

Secondly, I would suggest trying to reproduce the issue on a built in mesh first (say a unit cube), as it would then be easier to narrow down where the issue of your actual code is.