MantisBT - Community
View Issue Details
0022882Community[OCCT] OCCT:Modeling Algorithmspublic2011-12-28 14:592017-12-05 17:09
szy 
jgv 
normalminor 
assignedopen 
ALL
[OCCT] 6.5.2 
[OCCT] 7.4.0* 
0022882: Bug in fillet operation
Post from the Forum (author: Sean) - http://www.opencascade.org/org/forum/thread_22488/?add_id=22488. [^]
"Hi
Attached, please find a zip file with 4 brep files.
I used BRepFilletAPI_MakeFillet to create the fillet on small.brep with radius=0.002. It results in small2.brep. The same step is applied to another part (big.brep) with radius=0.2. and it results in big2.brep. big.brep is 100 times bigger than small.brep. However, the results are different. small2.brep has 14 edges and big2.brep has only 8 edges.

Thanks.

Sean"
=================================

Fillet with radius r=0.002 gives the specified number = 14 of edges on both shapes (as on small.brep as on big.brep)

Starting from r = 0.02 it gives stable number of 8 edges in resulting shape.
When you try to check the filleted face itself you can find that in case of more small radius (less than 0.02) two vertexes are enough far from the curve itself (see the picture)and fillet gives the next warnings also:

Dump of SweepApproximation
Error 3d = 9.64427631904698e-006
Error 2d = 9.64427631549123e-006 ,
9.6120996504437e-006
23 Segment(s) of degree 11
length of the trajectory : 4.94398697453071
distance psp/papp : 0.732435912823124

COMPUTE: temps total 0.109375s� dont :
- Init + ExtentAnalyse 0s
- PerformSetOfSurf 0.109375s
- PerformFilletOnVertex 0s
- FilDS 0s
- Reconstruction 0s
- SetRegul 0s

For r <= 0.02 the filleted surface consists from 3 faces with degenerated edges.
pload ALL
restore small.brep s
explo s
renamevar s_1 s
don s
explo s E
## s_1 s_2 s_3 s_4 s_5 s_6 s_7
fillet r s 0.002 s_2
# Dump of SweepApproximation
# Error 3d = 2.0842025270003e-006
# Error 2d = 1.82208251552354e-006 ,
# 1.60944430249344e-006
# 9 Segment(s) of degree 8
# length of the trajectory : 0.0494398675219756
# distance psp/papp : 0.000530648888485438

# COMPUTE: temps total 0.0625s dont :
# - Init + ExtentAnalyse 0s
# - PerformSetOfSurf 0.0625s
# - PerformFilletOnVertex 0s
# - FilDS 0s
# - Reconstruction 0s
# - SetRegul 0s

don r
explo r F
don r_2 r_3 r_4
exolo r_4 V
dump r_4
## it has degenerated edges.
No tags attached.
zip fillet.zip (82,328) 2011-12-28 15:20
https://tracker.dev.opencascade.org/
png filleted_surface.PNG (21,483) 2011-12-28 15:20
https://tracker.dev.opencascade.org/
Issue History
2011-12-28 14:59szyNew Issue
2011-12-28 14:59szyAssigned To => jgv
2011-12-28 15:20szyFile Added: fillet.zip
2011-12-28 15:20szyFile Added: filleted_surface.PNG
2012-01-12 10:05abvStatusnew => assigned
2012-01-13 14:52szyNote Added: 0019114
2012-02-09 09:06abvTarget Version6.5.3 => 6.5.4
2012-10-23 18:41abvTarget Version6.5.4 => 6.6.0
2013-03-29 21:27abvTarget Version6.6.0 => 6.7.0
2013-12-21 10:21abvTarget Version6.7.0 => 6.7.1
2014-04-04 18:19abvTarget Version6.7.1 => 6.8.0
2014-09-26 09:55abvNote Added: 0032192
2014-09-26 09:55abvTarget Version6.8.0 => 7.1.0
2016-10-28 15:33msvTarget Version7.1.0 => 7.2.0
2017-07-21 11:34msvTarget Version7.2.0 => 7.3.0
2017-12-05 17:09msvTarget Version7.3.0 => 7.4.0*

Notes
(0019114)
szy   
2012-01-13 14:52   
The specified bug is not a regression.
It is presented in all available versions till OCCT6.3.0 (the last which was tested).
(0032192)
abv   
2014-09-26 09:55   
Still reproduced on current master