-
Notifications
You must be signed in to change notification settings - Fork 239
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Fix surfaces lying in x-y-plane #1318
Conversation
Jenkins: OGS-6/Gui/Mac-PRs failed: https://svn.ufz.de:8443/job/OGS-6/job/Gui/job/Mac-PRs/2215/ |
rot_mat(2,1) = 0.0; | ||
rot_mat(2,2) = -1.0; | ||
return; | ||
} |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Why not modify the first and third diagonal entries via: rot_mat(i,i) = n[2]>0 ? 1.0 : -1.0;
(i = 0,2) and not duplicate the setting of the other matrix entries? ✅
👍 |
@@ -1 +1 @@ | |||
Subproject commit 3087421f7ad138c72be7ee85f6872925da48ba7b | |||
Subproject commit 28d74b31d74aa5f03a418cc5a11f33cd1e2d7662 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
In the Tests/Data could the Elliptic/cube_1x1x1_NeumannBC/ could you please cleanup the vtu files? There are unnecessary data arrays in the input vtu and output vtus. (Alternatively one could refer to the original input vtu (without copy) which is one level up, and also to the gml).) ✅
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks good. Later when merging the Tests/Data branch into Tests/Data/master please squash the changes, because the data will take space otherwise.
Jenkins: OGS-6/Mac-PRs failed: https://svn.ufz.de:8443/job/OGS-6/job/Mac-PRs/2238/ |
Jenkins: OGS-6/Mac-PRs failed: https://svn.ufz.de:8443/job/OGS-6/job/Mac-PRs/2239/ |
Jenkins: OGS-6/Gui/Gui-Linux-PRs failed: https://svn.ufz.de:8443/job/OGS-6/job/Gui/job/Gui-Linux-PRs/2320/ |
Jenkins: OGS-6/Gui/Gui-Linux-PRs failed: https://svn.ufz.de:8443/job/OGS-6/job/Gui/job/Gui-Linux-PRs/2321/ |
Jenkins: OGS-6/Linux-PRs failed: https://svn.ufz.de:8443/job/OGS-6/job/Linux-PRs/2373/ |
Jenkins: OGS-6/Linux-PRs-dynamic failed: https://svn.ufz.de:8443/job/OGS-6/job/Linux-PRs-dynamic/597/ |
Jenkins: OGS-6/Linux-PRs failed: https://svn.ufz.de:8443/job/OGS-6/job/Linux-PRs/2374/ |
Jenkins: OGS-6/Linux-PRs-dynamic failed: https://svn.ufz.de:8443/job/OGS-6/job/Linux-PRs-dynamic/598/ |
Jenkins, test this please. |
1 similar comment
Jenkins, test this please. |
WRAPPER time | ||
TESTER vtkdiff | ||
ABSTOL 1e-16 RELTOL 1e-16 | ||
DIFF_DATA cube_1e3_top_neumann_pcs_0_ts_1_t_1.000000.vtu pressure pressure |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Is it already fixed as in PR "Fix AddTest"?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
After the fix AddTest PR cmake will complain about this change. So there will be no silent failure.
After the test is named GroundWaterFlow/neumann... ⏩ |
4956ae8
to
419d719
Compare
419d719
to
b9cd0ae
Compare
Jenkins: OGS-6/Mac-PRs failed: https://svn.ufz.de:8443/job/OGS-6/job/Mac-PRs/2285/ |
OpenGeoSys development has been moved to GitLab. |
In our shape matrix computation routines there was some code detecting the special case when a 2D element lies in the x-y plane. This code led to Neumann BCs being added sign-flipped if the surface at which it was applied was in the x-y-plane with surface normal (0, 0, -1).
This PR fixes the special cases in the code. A small end-to-end test example has been added. Cf. image: Groundwaterflow process; upper row before the PR, lower row after the PR; one surface is kept at Dirichlet 1, the other has Neumann 1, i.e., the point at which the Neumann BC is applied must be the maximum over the entire z-axis; that did not always hold before this PR (upper right case).
Furthermore the det J > 0 check has been made strict in this PR. I.e., elements with zero volume or wrong orientation are disallowed now.