dolfin team mailing list archive

[Johan Hake <hake.dev@xxxxxxxxx>]

On 03/16/2012 10:28 AM, Garth N. Wells wrote:
> On 16 March 2012 09:17, Johan Hake<hake.dev@xxxxxxxxx>  wrote:
> > Hello!
> >
> > I have implemented a new version of UnitSphere based on subdivision of an
> > initial octahedral surface mesh and tetrahedralization using the CGAL
> > framework.
> >
> > This is all fine, but CGAL uses alot of time tetrahedralizing the initial
> > surface mesh. CGAL is also doing some funny things with the surface mesh.
> >
> >   1) It subdivide the initial surface mesh, which I want to prevent. Is
> >      that possible.
> >   2) 1) results in small triangles on the surface, but large triangles
> >      in the interior. I set a tentative volume constraint based on the
> >      size of a triangle of the input mesh.
> >
> > You can try it out here:
> >
> >   bzr branch lp:~dolfin-core/dolfin/hake-trunk
> >
> >   from dolfin import *
> >   mesh = UnitSphereNew(4)
> >   # Do not try larger subdivisions than 4, at least for now.
>
> Did you look at the CGAL demo for generating a mesh or a sphere:
>
>     http://www.cgal.org/Manual/latest/doc_html/cgal_manual/Mesh_3/Chapter_main.html#Subsection_50.3.1
>
> ? It uses an implicit surface description. I has planning to add
> support for meshing geometries from implicit descriptions.

No I didn't. Are we using the implicit interface in dolfin?

Using a subdivision algorithm is probably the fastest and cleanest way 
to generate the surface of a Sphere. Using an implicit function might 
create surfaces which does not follow the input mesh, which is what I want.

johan

> Garth
>
> > Johan
> >
> > _______________________________________________
> > Mailing list: https://launchpad.net/~dolfin
> > Post to     : dolfin@xxxxxxxxxxxxxxxxxxx
> > Unsubscribe : https://launchpad.net/~dolfin
> > More help   : https://help.launchpad.net/ListHelp