Today or tomorrow I will upload some changes to the trunk. These require a few changes to the input files. Next week (roughly) I will upload larger changes to the trunk, but these will NOT require input file changes.
Changes occur in the .mmp and .num files and are as follows:
$CAPILLARY_PRESSURE : must be made independent of relative permeability
$PERMEABILITY_SATURATION: Reverting to the use of sgr and slr for each phase.
$NON_LINEAR_SOLVER: Must allow the entry of multiple degree of freedom tolerances.
The coupled process entries are completely different, to allow correct process couplings.
For details, see the attached files (you may only need the new input files, but I also attach a description of the changes). As changes must be made to benchmark input files to prevent them all from failing, I will make these changes. And upload trunk source changes only once the benchmark input files are modified to the new system.
Please take a look and provide me with any objections.
Once I update the trunk, I will explain the new capabilities.
Should Kueper prob be solved with these example files revised
according to your 5.2.09 release? You know Kueper prob is very
important to me. Another question is that should it solve with
PARDISO?
Regards,
Chan-Hee
···
On Mar 20, 11:58 pm, Joshua <j.ta...@gmail.com> wrote:
Hello all,
Today or tomorrow I will upload some changes to the trunk. These require a
few changes to the input files. Next week (roughly) I will upload larger
changes to the trunk, but these will NOT require input file changes.
Changes occur in the .mmp and .num files and are as follows:
1) $CAPILLARY_PRESSURE : must be made independent of relative
permeability
2) $PERMEABILITY_SATURATION: Reverting to the use of sgr and slr for each
phase.
3) $NON_LINEAR_SOLVER: Must allow the entry of multiple degree of
freedom tolerances.
4) The coupled process entries are completely different, to allow correct
process couplings.
For details, see the attached files (you may only need the new input files,
but I also attach a description of the changes). As changes must be made
to benchmark input files to prevent them all from failing, I will make
these changes. And upload trunk source changes only once the benchmark
input files are modified to the new system.
Please take a look and provide me with any objections.
Once I update the trunk, I will explain the new capabilities.
I couldn’t look at your file changes (attachments didn’t load). You should be able to run the problem as before without changes (let me know if this doesn’t work). But the changes might be helpful. I attach the .mmp and .num files in their old format (from benchmarks/MULTIPHASE/KeuperProblem-PS) and the versions as I would revise them (this may be exactly what you did already).
The benchmark is set to BICGSTAB, but I suspect (and you would know best) that PARDISO is the best option for Keuper. Let me know if any of this doesn’t work.