Meshing 3D geometries defined on lattice

On 11/20/2012 03:11 PM, Abbas Shirinifard wrote:
> New question #214742 on FEniCS Project:
> https://answers.launchpad.net/fenics/+question/214742
>
> Do we have to mesh multi-domain 3D geometries that are defined on a
> rectangular lattice to be able to use them with FEniCS? or FEnics is
> able to solve PDEs (e.g. reaction diffusion) on lattices.

How are these lattices given? DOLFIN only works with tetrahedrons, for
now, so you need to figure out a good way to transform you lattices mesh
into a tetrahedon mesh.

> The domains represent individual biological cells in a tissue (in
> CompuCell3D lattice-based simulation framework). Boundary of these
> domains move (slowly) so we have to re-mesh which I think will be
> expensive if we do it very often.

If only the boundary is moving you should be able to move vertices in
the tetrahedron mesh.

> If we have to mesh, what are the options for meshing lattices that
> works with FEniCS?

I do not know.

Johan

> Suggestions are much appreciated.
>
> Abbas
>

Lattice data is in cubic pixel format.

Would GAMer (I found it in your webpage) mesh lattice data?

On 11/20/2012 03:50 PM, Abbas Shirinifard wrote:
> Question #214742 on FEniCS Project changed:
> https://answers.launchpad.net/fenics/+question/214742
>
> Status: Answered => Open
>
> Abbas Shirinifard is still having a problem:
> Lattice data is in cubic pixel format.

Is it a binary data format?

> Would GAMer (I found it in your webpage) mesh lattice data?

It does read _a_ lattice data format, which is a binary data format
resulting from a segmentation of some image data. If it is the same data
it should work. There is however not a 1-1 correspondence between the
lattice data and the generated mesh.

Johan