If there is anyone who can help me with this I would be very, very grateful. I have been trying to do this for two weeks now with no luck.

I am trying to model the species mass fraction in an electrochemical system. The charged ions are drawn towards an oppositely charged electrode. I know the equation for the flux of the species. There is NO flow.

I have reduced the problem to a 2D, rectangular grid. At the top a flux of species in using a UDS flux. At the Bottom an opposite flux of species out. The net gain of species in the system should be zero.

I set my UDS_FLUX term so that is returns a constant * area of horizontal face. The constant is the same as the flux on the top and bottom wall, but that shouldn't matter. The UDS_FLUX takes the retuned value and multiplies it by the value of the UDS I have removed diffusion by setting the value to zero.

I cannot get the value of the UDS to settle on a fixed value, it just keeps going up.

I know that you shouldn't set two NEUMANN B.C's, but when you do it with diffusion, it settles nicely to a constant species mass fraction gradient. Why dose it not with my model?

Please, Please help me, anyone.

Andrew, Maybe you could post the main part of the your define uds flux udf and there may be a clue in there.

OK, I am using a massivly simplified 2d model to try to understand what is going on - a 5 x 5 2D grid. Fluxes on the top and bottom face, say 0.001.

The UDF is a little like this. I just want the mass flux to be the same through each cell (later I will link it as a function of another variable)

DEFINE_UDS_FLUX

{

F_AREA(A, f, thread);

if(f == internal face); return 0.001 * A[1]; else return 0;

}

This should mean that the flux of the UDS through each plane (where y = const) is the same, right?