[Gmsh] Anisotropy and refinement

Christophe Geuzaine cgeuzaine at ulg.ac.be
Thu Aug 5 16:32:12 CEST 2010


On 21/06/10 13:07, Laurent Michel wrote:
> Hi,
>
> I've been using gmsh for some time now (but very basically), and there
> is a couple of things I'd like to have a few hints for, if possible.
>
> First, I want to mesh a glacier. For this, I mesh its bed by hand. I
> don't care about its surface, because I am solving a free surface
> problem, hence I mesh a bit more than the actual glacier. The shape of
> the top of the mesh (I am referring to the part at the top of the
> glacier surface) can be whatever, I don't care. However, I'd like to get
> an anistropic mesh. How can I do that?

For structured grids, you can easily generate highly anisotropic grids 
with the "Transfinite" command.

For unstructured grids we currently only generate isotropic grids, but 
we are working on interfacing bamg, which will allow nice unstructured 
(2D) meshes.

>
> Second, I am doing parallel simulations, and I'd like to refine my mesh
> in a certain way. For instance, I'd like to exactly double the amount of
> nodes of my mesh. How do you do that with gmsh? There is a command
> refine, but it doesn't do that, as far as I have understood.

No problem for structured grids. Exactly doubling the number of nodes 
for unstructured grids is not easy...

>
> Thanks,
>
> L.
>


-- 
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine