[sosat1012]

I'm trying to simulate a simple -only- diffusion problem (steady state). There is a box that its lower face produces CO2. The top face of the box has an opening and the box is surrounded by air (at 25C). Finally there should be due only to diffusion a mix of air and CO2 inside the box (is it not?). There are no velocites at all (as i said only diffusion and no convection).

What boundaries should i use supposing that initially there is no CO2 in the box (initialisation -> V=0 and CO2 mass fraction=0.0 in the domain)?


LOWER FACE: WALL type and define a source of C02 (flux).
OR
INLET type with velocity equal to 0 and mass fraction of CO2 equal to 1.0

TOP FACE: OPENING type with pressure equal to 0 and mass fraction of CO2 equal to 0. (air is defined as constraint so its mass fraction there would be 1.0)
OR
OUTLET type with Pres=0
OR

SOMETHiNG ELSE?????????

I get no good results with everything i try? Maybe my setup has a flaw buit i have checked everything many times.....Anyway is only diffusion and no convection problems solvable in CFX?

[ghorrocks]

You could set this up with mass fractions and use the expert parameters to not solve the fluids, heat transfer and turbulence equations. Then the only equation being solved would be the mass fraction and you would get a diffusion model.

A simpler approach may be to map it across to a thermal simulation. CFX supports a thermal only simulation for solids. Map mass fractions across to temperature and diffusivity to thermal conductivity.

[sosat1012]

Thanks a lot ghorrocks......just did it and it seems correct for the first time