Hi everyone! I have a global problem with the diffusion in fluent. I have to simulate a very simple problem, namely there are two channels which are separate by a membrane. N2 flows in the one channel and in the other He. There are in principal different flow rates at the inlets. The pressure overall have to be constant, e.g. there is no viscous flow trough the membrane and the mass transport is defined only by molecular and Knudsen diffusion. Theoretically the mass flow at the each outlets should be different comparing to the inlet mass flow of each channel, because in the porous media region the diffusion coefficients are not equal for both component and He diffuses easier as N2. But the results in Fluent don't show this behavior, they show the same mass flow at the inlet and outlet of certain channel. I know, that the sum of diffusion fluxes over the control volume have to be 0, but because of different diffusive flows through the membrane it should be generated velocity trough the membrane. But this doesn't happen by the simulation, because the experimentally determined permeability constant in the porous region is so low, that the velocity is already 0. I will be glad, if someone could help me with this problem. Thank you

[kdep]

Very old thread, but I am desperate for an answer to my question. Any chance you could explain how you managed a membrane that only had diffusive transport through it and no convective transport?

Kurtis