[Getdp] Just a litle help ..

[Christophe Geuzaine]

Nacho Andres de la Fuente wrote:

> Dear Cristophe,
> 
> Many thanks for your response, I read carefully the manual and all the
> mailing list (it was quite helpful your discussions with Samuel
> Kvanisca) and I had already include SurfIll and SurfDark in the support
> of my basis functions. That is the reason for which I sent you all the
> *.pro and *.geo files, because I consistently checked everything up
> before bothering you.
> 
> Do you have any other suggestion?

Did you try to solve a "toy" problem first (2D, with constant
coefficients, etc.), in which you gradually incorporate all the features
you need in the real one (one material with Dirichlet BC first, then 2
materials, then exterior Neumann BCs, then interior BCs, etc.)? That's
usually the best way to proceed.

I don't have time to examine your .pro/.geo files in detail, but here
are some potential problems:

- your conducting domain contains both volume and line elements (for
which your integration method is not defined: getdp gave me an error
during the assembly!);

- the volume elements discretizing the region between the spheres is not
used anywhere;

- the Dirichlet condition on the "center" will not work: the point is
not connected to any volume element (see
http://www.geuz.org/gmsh/doc/texinfo/gmsh_fot.html#FOOT1).

Christophe

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