[Bug report] wrong computation of a section with BRepAlgoAPI_Section

Stephane Routelous
Tue, 05/01/2007 - 20:35

Forums:

Hi,

I found a bug with BRepAlgoAPI_Section.
I'm computing the section of a shape with a plane.
Using BRepAlgoAPI_Section, the section is wrongly computed (missing parts)
If I use BRepAlgo_Section (as in Draw), it works fine.

If you want, I can provide you the shape to cut and the coordinates of the plane.

thanks in advance,

Stephane Routelous

P G
Wed, 05/09/2007 - 08:22

Hi,

I also found these results are inconsistent between
BRepAlgoAPI* and BRepAlso* classes like in boolean
operations too.

Please provide the shape and the coordinates of the plane.
- pg

Stephane Routelous
Wed, 05/09/2007 - 16:08

the shape is here : http://www.exotk.org/model.brep.zip (372KB)
the plane is x=132.863 y=110.119 z=20.1127 nx=0 ny=0 nz=1

Stephane
http://www.exotk.org

Forum supervisor
Wed, 05/16/2007 - 09:52

Hello Stephane,

We were able to reproduce the problem using the command 'bsection' in OCCT 6.2 Draw as follows:

pload ALL
brestore model.brep b1
plane p 132.863 110.119 20.1127 0 0 1
mkface b2 p
bsection t b1 b2

In this case, the section 't' contains 46 edges - 1 edge is missing because the intersection algorithm generates a set of points not suitable for direct construction of a b-spline curve. To overcome the problem, the section command must be called with parameter '-a' that forces the intersection algorithm to approximate the resulting set of points with a b-spline curve:

bsection t b1 b2 -a

In general, it is recommended to set this parameter to approximate the intersection curves, except some simple cases.

Hope this advice can be of some help.
Best regards,
Forum Supervisor