[CIG-LONG] How to specify melt supply rate?

[Walter Landry]

Garrett Ito <gito at hawaii.edu> wrote:
> Walter, from what I recall I dont think we got this issue (as well
> as my weird "bat ears") sorted out.  Perhaps its related to the
> linear pressure-elements, which do not do as good a allowing an
> abrupt change in pressure?

I sort of figured it out.  Sorry for not letting you know earlier.  We
thought there might be a parallel bug, but I was able to duplicate the
problem in serial.  Parallel just changed the location of the problem.

I think that there is an interaction during the non-linear iterations.
The pressure goes up in a region, which causes the compressibility to
go up, which somehow causes the pressure to go up.  The result is a
runaway instability in a region where there is yielding.  Since there
is so much extra compressibility, that could account for the
inconsistent results you got.

Unfortunately, I do not have a fix at this time.

Cheers,
Walter Landry
walter at geodynamics.org