[Jayotpaul]

Hello users, i am trying to simulate something similar to a semi-permeable mebrane where both water and gas comes in and only air is allowed to flow through, while water accumultes on the inlet side and drains due to gravity.
I expected it to be a simple case, but the simulation is crashing mid way and i cannot also see any draining effect till then. Spent a lot of time playing with BC and time steps but nothing helps, and i am very confused .

I am using a euler-euler multiphase simulation (mixture model) with buoyancy inside (gas density as reference density), with opening as outlet B.C with specified pressure and a specified velocity and volume fraction as inlet B.C. Tried to replicate membrane using a high resistance to water flow in the flow direction only on a thing region of space.

Any help is appreciated.

[ghorrocks]

What form is the air/gas in? Bubbles, droplets, air dissolved in water, or something else?

[Jayotpaul]

Originally Posted by ghorrocks

What form is the air/gas in? Bubbles, droplets, air dissolved in water, or something else?

Dear Ghorrocks, my actual problem is an interface retension of oil when oil and air flows trough a porous media. The oil is retained on the porous side due to surface effects and the gas passes through with a high pressure drop. I wanted to first do a easy simulation and try to replicate this surface phenomenon using a 'semi permeable membrane' analogy, to see if it is possible.
So in my case both gas and fluid is a interpenetrating homogenous mixture, thats why i used both phases as continous fluid and a mixture model (beacuse both phases have separate velocity fields). I hope what i am trying to do makes sense. Thank you for your time.

[ghorrocks]

The starting point is understand what condition your fluid mixture is in. So what form is the oil/air in? Bubbles, droplets, air dissolved in oil, or something else?

[Jayotpaul]

Originally Posted by ghorrocks

The starting point is understand what condition your fluid mixture is in. So what form is the oil/air in? Bubbles, droplets, air dissolved in oil, or something else?

In my actual problem, the oil flows as thin films on the porous medium (more specifically fibrous filter elements) which flows along with air due to drag forces. Since i am using volume averaging, so both are assumed to be continuous fluids flowing through the medium, with momentum exchange due to drag forces. Does this help a bit?

[ghorrocks]

Yes, thanks. So you are using a eularian approach with oil and air phases.

Have a look at source terms, especially the phase-specific source terms. Hopefully you can do a source term which stops the oil phase at the entry to the porous material but lets the air through.

[Jayotpaul]

Originally Posted by ghorrocks

Yes, thanks. So you are using a eularian approach with oil and air phases.

Have a look at source terms, especially the phase-specific source terms. Hopefully you can do a source term which stops the oil phase at the entry to the porous material but lets the air through.

Thanks for the reply, thats exactly what i am trying to do. Create a momentum source term for water in the flow direction. I hoped to stop the flow of water in the flow direction, and expected drainage in the transverse direction. Unfortunately the solver diverges after water hits that area of momentum sorce . Is it possible the solver diverges due to high momentum source? Because i played with the values and the solver runs with low momentum sources. Since i need to completely stop the flow i need a very high source term. Any ideas for a workaround, if this is the problem?

[ghorrocks]

Is it possible the solver diverges due to high momentum source? Because i played with the values and the solver runs with low momentum sources.

Your second sentence is pretty strong evidence the first sentence is correct, isn't it?

The simple approach is to tune the strength of it to be low enough to converge but strong enough that it behaves as expected. So have a look at the results and do a sensitivity study and see if a value which does converge is close enough to not significantly affect results.

If that does not work you are going to have to look at more sophisticated convergence controls but let's not go there unless you need it.

[Pumba]

How did you define your momentum source?

Btw: maybe you should try FLUENT for that. There you can fix the velocity in the zone of the membrane to zero.

[Jayotpaul]

Originally Posted by Pumba

How did you define your momentum source?

Btw: maybe you should try FLUENT for that. There you can fix the velocity in the zone of the membrane to zero.

I use a phase specfic momentum source (linear resistance in cfx) along with a step function which activates only in a thin volume (my interface region)

This specific problem is only a part of a project, which i have been working on for a couple of month, so cant shift mid way . So have to make it work here.

[borandoruk]

Hello friends,
I model the two-phase flow through the membrane. While one phase is being held in the inner flow, the other phase must pass through the membrane. However, depending on the accumulation of the other phase, the mass flow rate of the phase passing through the membrane must decrease at the outlet. However, the mass flow rate of both phase 1 and phase 2 is not as expected at the outlet. I have read the clues you have given above. But I did not get results. I tried all models like mixture, DPM, VOF. Please come out with someone who can help or give tips.