(feat) UTVD fluxlimiter #2374
(feat) UTVD fluxlimiter #2374
Conversation
Manangka
force-pushed
the
unstructured_fluxlimiter
branch
4 times, most recently
from
866076a to
1240d20
Compare
Manangka
force-pushed
the
unstructured_fluxlimiter
branch
from
1240d20 to
dfeb932
Compare
# Conflicts: # make/makefile # msvs/mf6core.vfproj # src/Model/TransportModel/InterpolationScheme/AdvSchemeEnum.f90 # src/Model/TransportModel/InterpolationScheme/CentralDifferenceScheme.f90 # src/Model/TransportModel/InterpolationScheme/TVDScheme.f90 # src/Model/TransportModel/tsp-adv.f90 # src/Utilities/LinearAlgebraUtils.f90 # src/meson.build
Manangka
force-pushed
the
unstructured_fluxlimiter
branch
from
42b2e65 to
1306626
Compare
aprovost-usgs
left a comment
aprovost-usgs
left a comment
There was a problem hiding this comment.
Elegant code. Just minor things for your consideration
…d of the G matrix
# Conflicts: # make/makefile
# Conflicts: # make/makefile # src/Model/TransportModel/tsp-adv.f90
mjr-deltares
left a comment
mjr-deltares
left a comment
There was a problem hiding this comment.
I really like how you did this. I do think we need to present the core design to the team so if they visit tsp-adv, they know what is going on there. You have increased the level of software engineering to come up with a better structured code, and succeeded. Let's make sure people can follow along.
|
|
||
| select case (this%limiter_id) | ||
| case (2) ! van Leer | ||
| theta = max(0.0_dp, min((r + dabs(r)) / (1.0_dp + dabs(r)), 2.0_dp)) |
There was a problem hiding this comment.
I know we have DZERO, DONE, etc. but I like it better how you have done it here, it's readable
There was a problem hiding this comment.
Nice to isolate generic functionality. There are similar units in the ModelUtilities folder. Should this go there as well?
aprovost-usgs
left a comment
aprovost-usgs
left a comment
There was a problem hiding this comment.
Nice job, @Manangka
# Conflicts: # make/makefile
4 participants
This commit adds the new TVD flux limiter based on the paper of Darwish.
For now i called the new limiter UTVD where U stands for unstructured. But as @christianlangevin and i discussed last week this doesn't cover the because it is also suited for structured grids.
A small description of the code layout.
The main starting point is the tsp-adv.f90 file. Here the cell based gradient is created and passed on to the UTVD class.
I would recommend reviewing tsp-adv.f90 -> IGradient.f90/LeastSquaredGradient.f90 -> UTVDScheme.f90 first to follow the logic of the implementation.
A helper class
DisUtilsis added which provided common functions e.g. getting a cell centroid locationI've updated existing advection tests to also test the new limiter.
It depends on #2367 and #2370
Checklist of items for pull request
ruffon new and modified python scripts in .doc, autotests, doc, distribution, pymake, and utils subdirectories.fprettifyFor additional information see instructions for contributing and instructions for developing.