dolfin team mailing list archive

[Bartosz Sawicki <sawickib@xxxxxxxxxxxxx>]

cosby@xxxxxxxxx:
> Dear Dolfin developers
>
> I have been looking around for a relative easy-to-use poisson
> solver for a 2D problem and stubled upon the Fenics/Dolfin
> package.
> I have managed to compile the package (including PyDolfin which
> is what I hope to be able to use) and have a look at the Poisson
> demo.
> However, from the demo and the manual it is not clear to me whether
> Dolfin can deal with a dielectric subdomain. The boundary condition
> on a vacuum-dieletric interface is that the normal derivative has
> a discontinuity.
> Could anyone help clarifying this issue ?

I am not a dolfin developer, so my knowledge is limited, but your
problem is rather classic one. I don't predict any problems.

When you use Lagrange element potential has to be continous, but
derivative doesn't has such limitation.

Moreover Fenics has Discontinous Langrange elements and possibility of
using boundary integrals inside the form definition. So you are
associating with powerful tool.

regrd.
BArtosz