Hello, I am using the mixture multiphase model to solve gas-liquid flow in a pipe. The volume fraction of the liquid phase is very small (less than 3%).

The two phases move at different velocities, and I want to define both. Therefore I want to enter a slip velocity between the two phases. However when I turn on the slip velocity option, the secondary phase becomes discrete (ie. bubbles, particles, droplets). The secondary phase is a continous liquid phase.

Can anyone help me with this problem?

Regards, John

For a liquid volume fraction of 3%, I assume you are trying to say that the gas volume fraction is 97%. Is this really correct? Please check again. Normally, for the bubbly regime in fluent, fluid-fluid flows are defined using a comtinuous (primary) phase which usually has a large volume fraction (say more than 50%) and a dispersed (secondary) phase which is also a fluid (liquid or gas) but has a lower volume fraction (typically 1 - 25%). These are only typical ranges, give or take 10 to 15 % for practical purposes. Secondly, the slip velocity is usually estimated when using either model (eulerian or mixture), not specified. In the case of the mixture model, you have the flexibility to write your own user-defined drag function which will affect the slip to some extent. All this changes of course if you are trying to model stratified flow. Is this the case?

Thanks for your reply,

Yes the liquid volume fraction is very small (less than 3%). I am modeling it as a thin film around the walls of the pipe. It closely resembles annular flow regime. Since the liquid fraciton is very small, it is assumed the entrained fraction does not exist.

The model is being used to simulate a natural gas transmission pipeline, where only very small liquid deposits are known to exist.

So ideally I want the liquid to flow as a thin film along the walls, without excessive mixing with the gas. The mixing should only occur at the gas-liquid interface (where a sharp velocity gradient exists).

You are correct with the slip velocity, I can't specify it.

So in my mind the best way to model this flow is to use the mixture model, with or without the slip velocity option. Is this what you would suggest?

Also, would it be wise to use the implicit body force option?

Thankyou for your time,

Regards,

John

I think VOF is your best bet.

Aside: If you activate implicit body force, Fluent support says convergence improves sometimes. However, I do not think that the mixture model is appropriate for your problem. VOF seems more suitable.

Yes you are probably right. However when i use the VOF model, the film seems to disappear after moving a small distance along the pipe. Can this be because I havn't used a high enough liquid inlet velocity? Or could it be something to do with the VOF model limitations (i.e. can't handle differing velocities too well)?

Thanks again,

Regards,

John

Sorry can't help you there. I've never worked with the VOF model. Wait and see if someone else can help you out. Good Luck!

No worries. You've been a great help.