Closed matthewhoffman closed 10 months ago
A recent AIS run showed weird melt rates occasionally: Some investigating indicated that this occurs on a timestep after one in which the velocity solver failed to converge, which led to the fix in this PR.
After this fix, the same run looks correct:
When the velocity solver fails, the internal dissipation field it returns can have large artifacts in it that lead to unphysical melt rates being applied. This merge skips updating the heatDissipation field when the solver fails so that the previous timestep's value is retained.