[Gmsh] Help with 3D mesh

Christophe Geuzaine c.geuzaine at ulg.ac.be
Sat Jun 12 20:42:11 CEST 2004


Carrasco, Cesar J. wrote:
> Hello,
> 
>  
> 
> I am trying to develop a 3D mesh of a pavement section with several
> layers of materials.  The pavement section will be loaded on a small
> patch of about 6in (tire pressure). 
> 
> I successfully generated a 2D mesh that behaves exactly the way I want
> it to.  For the 2D mesh we would use an axisymmetric analysis.
> 
> I am attaching the geometry file for the 2D mesh.
> 
> What I want to do for the 3D mesh is to have something similar to the
> 2D.  What I have done is duplicate the 2D geometry rotated an angle of
> pi/2 radians.  This would constitute just one of the quadrants of the 3D
> mesh.  What I was trying to do next was to create the volumes between
> these two rotated planes.  I connected the two planes with the
> corresponding lines for just one of the volumes and created the
> necessary surfaces.  Then I created the volume.  When I try to mesh the
> volume, I get a series of warnings (I just copied two of them):
> 
> Warning : Coplanar points in circum sphere computation
> 
> Warning : (0,-155.24,0) (0,-148.353,0) (0,-155.24,0)
> (3.61401,-151.799,-3.61401)
> 
>             …………………………….
> 
> Needless to say, the 3D mesh does not get generated.  I am attaching the
> 3D geometry file.
> 

Surfaces 17 and 19 are incorrect (the loops don't close, due to
duplicate begin/end points).

You can fix this by adding "Coherence;" at the end of your file (or, of
course, by actually fixing the geo file).

>  
> 
> I am generating the geometry files using a program in Matlab, since I
> need to be able to change the number of pavement material layers and
> their corresponding dimensions.
> 
>  
> 
> I am sure there is a way to do this a lot easier.
> 

You can probably use loops/includes in the .geo file to achieve similar
results.

Best,

Christophe

-- 
Christophe Geuzaine
Applied and Computational Mathematics, Caltech
geuzaine at acm.caltech.edu - http://geuz.org