[Getdp] Face resistance, h-phi

Kubicek Bernhard Bernhard.Kubicek at arsenal.ac.at
Tue Oct 31 17:02:31 CET 2006


Hello!

With the help of Bernhard Laurent (Thank you very much for you help!),
there now exists another wiki-example demonstrating face conduction:
http://www.geuz.org/getdp/wiki/FaceConduct2D

Nice greetings,
 Bernhard Kubicek


-----Initial Message-----
Von: getdp-bounces at geuz.org [mailto:getdp-bounces at geuz.org] Im Auftrag
von Kubicek Bernhard
Gesendet: Montag, 23. Oktober 2006 12:48
An: getdp at geuz.org
Betreff: [Getdp] Face resistance, h-phi


Hello!
We are facing a probably tricky problem here, and would like to know if
there is a way to solve it with GetDP:
One has two brick-shaped electrodes which conduct quite well and who
have respectively one face set to a defined potential (e.g. 100 and 0 V)
as boundary condition.
Between the two electrodes, there is hot (and therefore conducting) gas
with a given conductivity as a function of the local temperature field.
However, at the boundary between gas and electrodes, there are a very
thin region, in which due to various effects a voltage drop occurs of a
couple of volt. This region is so thin that it is impossible to mesh it.

Hence, we want to simulate it by a very simple model, defining that at
the physical faces between gas and metal there should be a specific
voltage drop.

Is there a way in GetDP by which one could define a finite resistance of
a 2D layer in a 3D geometry?
Or is the only way to accomplish this by creating an extremely thin
additional region in which the conductivity is defined prop.  1/j ?

As a second point, it would really help us a lot to have an example of a
pro file, where from a source _current_ distribution a magnetic field is
calculated via the h-phi formulation. We have been looking at the
examples in the manual and the wiki, but still don't understand how to
calculate the source magnetic field from an arbitrary current
distribution within the formulation (e.g. InvRot we could only get to
work in the postprocessing). So, it would be very very kind if somebody
could share such a file. 

Thank you for your help,
and nice greetings,
Bernhard




More information about the getdp mailing list