MantisBT - Community
View Issue Details
0027798Community[OCCT] OCCT:Meshpublic2016-08-22 07:322019-02-11 10:37
Vico Liang 
closedno change required 
WindowsVC++ 201364 bit
[OCCT] 7.4.0* 
0027798: Bad or degenerated triangles should be avoided (e.g for Cone meshing)
OCCT meshing algorithm generates good mesh for most cases. But there are some speical cases which will generate bad triangle with angle near 180 degree or degenerated triangle with two coincident nodes. It can be reproduced with Cone Shape. It will be a big issue if these bad or degenerated triangles serve as input of remesher algorithm or simulation solver. The root cause might be about edge meshing of cone generatrix. The straight edge of cone generatrix will never be divided whatever how long it is. If the straight long edge can be divided as necessary, the bad triange might be avoided.
No tags attached.
duplicate of 0025042verified bugmaster Community Cone Meshing: Missing two connecting edges. 
Issue History
2016-08-22 07:32Vico LiangNew Issue
2016-08-22 07:32Vico LiangAssigned To => oan
2016-08-22 07:34Vico LiangDescription Updatedbug_revision_view_page.php?rev_id=14458#r14458
2016-08-22 11:07oanRelationship addedduplicate of 0025042
2016-11-03 17:38abvTarget Version7.1.0 => 7.2.0
2017-07-20 12:43oanTarget Version7.2.0 => 7.3.0
2018-02-25 21:09abvTarget Version7.3.0 => 7.4.0*
2018-11-07 16:30oanRelationship addedrelated to 0028379
2018-11-07 17:00oanNote Added: 0080853
2018-11-07 17:00oanAssigned Tooan => msv
2018-11-07 17:00oanStatusnew => feedback
2018-11-07 17:18msvNote Added: 0080859
2018-11-07 17:19msvNote Added: 0080863
2018-11-07 17:19msvAssigned Tomsv => Vico Liang
2018-12-24 11:20oanNote Added: 0081603
2018-12-24 11:21oanNote Edited: 0081603bug_revision_view_page.php?bugnote_id=81603#r20504
2018-12-24 12:24Vico LiangNote Added: 0081604
2018-12-24 12:26Vico LiangAssigned ToVico Liang => oan
2018-12-24 12:26Vico LiangStatusfeedback => assigned
2018-12-25 12:25oanNote Added: 0081617
2018-12-25 12:25oanAssigned Tooan => bugmaster
2018-12-25 12:25oanStatusassigned => feedback
2018-12-28 14:10oanStatusfeedback => assigned
2019-02-11 10:37bugmasterStatusassigned => closed
2019-02-11 10:37bugmasterResolutionopen => no change required

2018-11-07 17:00   
Implemented in context of #28379.

Test case not needed.
2018-11-07 17:18   
The fix for #28379 splits the seam edge of the conical face. It does not deel with degenerated triangles.
2018-11-07 17:19   
Dear Vico, please check the new behavior and tell if it is OK for you now.
2018-12-24 11:20   
(edited on: 2018-12-24 11:21)
Please note that BRepMesh algorithm is not supposed for use in solvers, but only for visualization needs.

Another point is that, it will always provide at least one degenerative triangle near the apex of the cone and e.g. pole of the sphere as far as it tries to produce connected mesh in 3d/2d space and without such triangles mesh will contain a hole.

In case of necessity to avoid degenerative triangles, please skip them manually before feeding triangulation to any solver.

As for general issue, division of seam edge has already been implemented in context of #28379 and integrated to master branch.

I propose to close this issue.

Vico Liang   
2018-12-24 12:24   
sorry that I don't get time to check this. please go ahead to close this issue.
2018-12-25 12:25   
Dear bugmaster,

please manage the issue.