[Gmsh] Gmsh: 3D Volume with a hole (not through-going)

Oliver Willekens oliver.willekens at elis.ugent.be
Fri Aug 29 00:34:48 CEST 2014


You could delete the volume of the cylinder you’ve created by means of
extrusion and then delete one of the cylinder’s flat surfaces using the
Delete command. However, it won’t give you what you want, because the box
“holding the cylinder” will be unaware of the presence of the cylindrical
hole.

I tried this:

Plane Surface(1) = {1, 2};  // square minus the circle
out1[] = Extrude {0.0, 76.2, 0.0}{Surface{1};};
Plane Surface(2) = {1};
out2[] = Extrude {0.0, -76.2, 0.0} {Surface{2};};

Which results in the 2D mesh you find in the attachment. Perhaps this is
what you’re looking for?

I also tried combining those two volumes using Compound Volume(4) =
{out1[1], out2[1]};, but like you, I got GRegion Compound errors. Probably
some basic restriction from triangulations I’m overlooking. This only means
your 3D meshes will be clearly stopped at a flat interface about half-way
through those two boxes, but you’ll still have a 3D mesh. I was hoping that
the Compound Volume would make some tetras across this boundary though… But
with the above lines, you’ll have a working mesh.
​


2014-08-28 17:08 GMT+02:00 Omid Mahabadi <omid.mahabadi at geomechanica.com>:

>  Hi Christophe and Gmsh team,
> I've been trying to model a simple cube with a cylindrical hole that is
> NOT through-going (i.e., its depth is smaller than the depth of the cube),
> as shown in the attached picture. The Extrude command doesn't seem to work
> since it will extrude both surfaces (exterior and interior -- hole) at the
> same time. I know I can define all the surfaces, surface loops, and volumes
> manually, but is there a better way of defining something like this in
> Gmsh?
>
> I also tried to use the Compound Volume command by first defining two
> volumes from Extrude and then trying to combine them but I'm getting errors
> for the Compound Volume visualization (Error: Cannot evaluate bounds on
> GRegion Compound) and when I mesh the geometry, the actual shared
> interfaces are still existing, although by the notion of compound from the
> documentation, the internal interfaces should be neglected. Here is the
> geometry file:
>
> // Characteristic length (==> element size)
> cl_external   = 25;
> cl_excavation = 5;
>
> // External boundaries
> Point(1)      = {-127, 0.0, -127, cl_external};
> Point(2)      = {+127, 0.0, -127, cl_external};
> Point(3)      = {+127, 0.0, +127, cl_external};
> Point(4)      = {-127, 0.0, +127, cl_external};
> Line(1)       = {1, 2};
> Line(2)       = {2, 3};
> Line(3)       = {3, 4};
> Line(4)       = {4, 1};
> Line Loop(1)  = {1, 2, 3, 4};
>
> // Excavation boundaries
> Point(5)      = {0.0, 0.0, 0.0, cl_excavation};
> Point(6)      = {19.05, 0.0, 0.0, cl_excavation};
> Point(7)      = {0.0, 0.0, 19.05, cl_excavation};
> Point(8)      = {-19.05, 0.0, 0.0, cl_excavation};
> Point(9)      = {0.0, 0.0, -19.05, cl_excavation};
> Circle(5)     = {6, 5, 7};
> Circle(6)     = {7, 5, 8};
> Circle(7)     = {8, 5, 9};
> Circle(8)     = {9, 5, 6};
> Line Loop(2)  = {5, 6, 7, 8};
>
> //Using compound volumes
> Plane Surface(1) = {1};
> out1[] = Extrude {0.0, 76.2, 0.0}{Surface{1};};
>
> Plane Surface(2) = {1};
> out2[] = Extrude {0.0, -76.2, 0.0}{Surface{2};};
>
> Compound Volume(3) = {1, 2};
>
>
> I also tried to create the geometry in CAD software and imported it as
> iges, brep or step formats but for reason the hole is not there completely.
> See for instances the iges file attached.
>
> Can you kindly shed some light here? Am I doing something wrong? Or are
> there Gmsh tricks/commands that I can use to achieve my goal?
>
> Thank you,
> Omid
>
> --
> Omid Mahabadi, Ph.D.
> Geomechanica, Inc.http://www.geomechanica.com/
> Tel : +1-647-478-9767
> Cell: +1-416-824-2408
>
>
> _______________________________________________
> gmsh mailing list
> gmsh at geuz.org
> http://www.geuz.org/mailman/listinfo/gmsh
>
>


-- 
Oliver Willekens
PhD Student
[image: LCP group logo] <https://lcp.elis.ugent.be/>

Liquid Crystals & Photonics Group
Sint- Pietersnieuwstraat 41
9000 Gent
Phone: +32 9 264.89.51
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.geuz.org/pipermail/gmsh/attachments/20140829/c646c50a/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cyl_in_box.png
Type: image/png
Size: 307120 bytes
Desc: not available
URL: <http://www.geuz.org/pipermail/gmsh/attachments/20140829/c646c50a/attachment.png>