Archived issue #0026928
Edges are meshed too coarsly
Description
In the appended case, the edges seem to be discretized too coarsly by BRepMesh_IncrementalMesh. It is especially visible because the corresponding surface, which is created from the edges by ThruSections is meshed much finer.
If you meassure the distance between the curve and its tessellation, it seems to be bigger than the linear deflection.
If you meassure the distance between the curve and its tessellation, it seems to be bigger than the linear deflection.
Steps to reproduce
bugs moddata_3 bug27356
#manual steps
restore Attachment 1 (BREP) c
vinit
vdisplay c
vfit
vsetdispmode c 1
# These are the Parameters the shape was originally meshed with => no Change
incmesh c 0.715 -a 11.459
vdisplay -redisplay c
incmesh c 0.5 -a 11.459
vdisplay -redisplay c
incmesh c 0.3 -a 11.459
vdisplay -redisplay c
incmesh c 0.1 -a 11.459
vdisplay -redisplay c
# only here the curve looks better
#manual steps
restore Attachment 1 (BREP) c
vinit
vdisplay c
vfit
vsetdispmode c 1
# These are the Parameters the shape was originally meshed with => no Change
incmesh c 0.715 -a 11.459
vdisplay -redisplay c
incmesh c 0.5 -a 11.459
vdisplay -redisplay c
incmesh c 0.3 -a 11.459
vdisplay -redisplay c
incmesh c 0.1 -a 11.459
vdisplay -redisplay c
# only here the curve looks better
Public activity
18 archived notes
Participants are labeled by their role within this record.
This problem is similar to #0026532 because in both cases edges are meshed too coarsly. But the current Problem has nothing to do with the minSize parameter.
Initial shape contains triangulation for face and 3D polygon for free edges. 3D polygons has deflection 0.184044204640587 that is less than values specified in "steps to reproduce" - 0.715, 0.5, 0.3. BRepMesh checks deflection of existing polygonal data and generates new one only if it is greater than specified value. That is why much more appropriate polygons are generated when value 0.1 is used. Like this, it is necessary to clean up all existing data before call to incmesh in order to produce correct result.
Problem described in issue is reproduced on current state of OCCT.
Please check on current master.
The problem is still reproduced. See http://occt-tests/IR-2018-11-02-OCCT/Windows-64-VC14/bugs/mesh/bug26928.html
Please note that results are different comparing to scenario given in Steps To Reproduce.
Dear Author,
In the current Dev version, you should use "tclean" command to vanish existing mesh (already stored in the file). After that, the shape will look OK.
In the current Dev version, you should use "tclean" command to vanish existing mesh (already stored in the file). After that, the shape will look OK.
Branch [archived branch] has been created by Participant.
[revision removed]
Detailed log of new commits:
Author: nbv
Date: Wed Nov 7 16:56:57 2018 +0300
0026928: Edges are meshed too coarsly
Adding test case for the problem.
The issue has been fixed by a patch for #0026106.
[revision removed]
Detailed log of new commits:
Author: nbv
Date: Wed Nov 7 16:56:57 2018 +0300
0026928: Edges are meshed too coarsly
Adding test case for the problem.
The issue has been fixed by a patch for #0026106.
Branch [archived branch] has been updated forcibly by Participant.
[revision removed]
[revision removed]
Branch [archived branch] has been updated forcibly by Participant.
[revision removed]
[revision removed]
Branch [archived branch] has been updated forcibly by Participant.
[revision removed]
[revision removed]
Dear Mikhail,
Please review the branch CR26928.
Test results are here:
Windows: http://jenkins-test-12.nnov.opencascade.com/view/MeshTest-master_NBV/view/TESTING/job/MeshTest-master_NBV-OCCT-Windows-64-VC14-opt-test-restart/HTML_20Report/
Linux: http://jenkins-test-12.nnov.opencascade.com/view/MeshTest-master_NBV/view/TESTING/job/MeshTest-master_NBV-OCCT-Debian80-64-opt-test-restart/HTML_20Report/
Please review the branch CR26928.
Test results are here:
Windows: http://jenkins-test-12.nnov.opencascade.com/view/MeshTest-master_NBV/view/TESTING/job/MeshTest-master_NBV-OCCT-Windows-64-VC14-opt-test-restart/HTML_20Report/
Linux: http://jenkins-test-12.nnov.opencascade.com/view/MeshTest-master_NBV/view/TESTING/job/MeshTest-master_NBV-OCCT-Debian80-64-opt-test-restart/HTML_20Report/
Please redesign the test:
- remove checkcolor command
- make view showing more detailed picture.
- remove checkcolor command
- make view showing more detailed picture.
Branch [archived branch] has been updated forcibly by Participant.
[revision removed]
[revision removed]
Done. Please review the branch CR26928 again.
Reviewed.
Update testing case
Branch [archived branch] has been deleted by Participant.
[revision removed]
[revision removed]
Related records