Hi,

I am trying to model the diffusion of air through a venturi inside a tank which is filled with water in contact with the atmosphere (free surface). I want to model the mixture air-water with the non-homogeneous particle model, and at the same time to model the free surface. Does anybody know if this is possible? If so, I have the following doubts:

1) which boundary condition should be used at the top of the tank? Opening with constant pressure? Degassing outlet? (there is only air at that boundary because the free surface is below the boundary).

2) How should I model the free surface? Is it possible to do it with the non-homogeneous model?

3) the density reference for buoyancy should be that of the water or that of the air? (in free surface flow with homogeneous model the density of air is used)

In anybody can give me a clue with any of these doubts or any other advise, It would be of great help.

Thanks,

Luis

1) Opening with constant pressure.

2) use free surface sharpening, factor 2.

3) use density of water.

Define an initial guess rho*g*h for pressure. Otherwise you will never get started. See the help on how to do this.. There is a clear example.

Gert-Jan

www.bunova.nl

Thanks,

I was missing the initial guess for the pressure. Hope that will make it work,

Luis

Look into the help. Search for "Modelling Advice for Free Surface Flow". There it is explained how to set this up.

Gert-Jan

Wait... if he uses water density as reference, won't the initial guess of rho*g*h be wrong, since pressure in water will be zero?

That is just for buoyancy computation.

Well, yes and no. The variable 'Pressure' excludes the "hydrostatic" pressure based on the reference density. That means that a fluid with a density equal to the reference density will have no buoyancy. Therefore if you have a free surface you probably should set your reference density as the lighter fluid, which in your case would mean pressure does not vary in air (which in real cases is a good assumption), while in water you would have your hydrostatic pressure field.

That is the problem. For free surface flow it is recommended to use the density of the lighter fluid as the reference density, but for multiphase dispersed flow (bubbles in water) it is recommended to use the density of the heaviest fluid as the reference density. In my problem I have a free surface as well as many dispersed bubbles of air in the water which move up towards the free surface. So, which density is better? Gert-Jan has recommended to use the density of water, as in multiphase disperse flow. I am trying this week to see if I get any trouble with the free surface definition...

Actually I think you should only use the density of heavier fluid if you have a very dispersed fluid, which is not your case since you have a free surface. My guess is that this is solely for numerical reasons, since you would then cancel the buoyancy source term on most of your domain.