Hi all,

as part of the whole problem, I´m trying to solve the electric field in a box. I set an additional variable and set a Diffusive transport equation and so on.

The problem: Imagine that there is a planar metallic grid in the box (parallel to any face) which is set to a given potential, as the rest of the boundaries (conductors and/or insulators). The problem here is that the grid is inside the domain so what I´m doing is to set a subdomain identical to the box and create a source valued zero everywhere except in the grid. Summarizing, I should solve: div(grad(V))+S=0 where S=rho/epsilon_0 in the grid and S=0 else everywhere

The charge density has to be translated to superficial density (the grid is conductor, sorry)and the mesh has to be fine there at the interface.

The results are like this implementation has not effect and the field is like there is no grid at all, but maybe it is due to a bad guess in the choice of S.

Anyone has solve a similar problem?, or knows if the set up of the problem is right?

thx pi

I believe that there was an article to present on this topic in Avignon in November 2004 (CFX User Meeting Avignon) by the ifremer, enitiia and cetim. Ask cfx

dd.nvs

thx. I´ll look.