Fracture closure

[Skonki]

Fracture closure in the case of re-meshing and/or the presence of fluid compressibility seems to have a bug.

• [x] #22

[Skonki]

Edit: The issue is also there if we do not have re-meshing and no fluid compressibility.

This means that between the simulations run for the recession paper and this update there seems to be an issue (as we do not use iterative solve here).

[Skonki]

See how the opening goes up shortly before closure. Oddly, the efficiency does not.

I will now investigate volume.

Note that this is without re-meshing..the result in case of re-meshing is identical.

[Skonki]

We observe similar issues with different (order one results) parameter sets.

However, this might be a known problem of leak-off continuing in cells with activated width constraint.

I add the task to check this.

See figure of the results

[Skonki]

For the data of Mohsen (used in the recession paper), this was actually the bug, and the results now show the expected behavior of the opening.

The test on the example (with re-meshing and all, rather coarse) is now running; let's see.

[Skonki]

The problem is solved, pull request follows.

See here the new graphics as a result.