Raise mass conservation limits
Summary
Apparently, !14 (merged) changed the results of the mass conservation test. This might be due to the different time step behavior, as the tests run much faster now. This indicates that global mass conservation also depends on time step size.
What is the current behaviour?
Mass conservation test for static simplex grid fails. Mass conservation is three orders of magnitude worse than expected.
Proposal
- Raise mass conservation limits to appropriate value
- Add tests on rectangular grids again, because execution time is much lower now.