[zhubohong]

As shown in the attached figure, does anyone know how to make the injected gas phase disappear, when the injected gas floats out water and meets the top air, or injected gas meets the top gas when the volume fraction of air reach to 1. Thanks for your advices.

[Opaque]

Search for "Degassing" boundary condition in the documentation

2.5.1.9. Degassing Condition (Multiphase only)
Degassing boundary conditions are used to model a free surface from which dispersed bubbles are permitted to escape, but the liquid phase is not. They are useful for modeling flow in bubble columns.

[zhubohong]

Originally Posted by Opaque

Search for "Degassing" boundary condition in the documentation

2.5.1.9. Degassing Condition (Multiphase only)
Degassing boundary conditions are used to model a free surface from which dispersed bubbles are permitted to escape, but the liquid phase is not. They are useful for modeling flow in bubble columns.

Thanks for your reply. However, I have known the particle model with degassing condition. I am not simulating the bubble column. I want to achieve the purpose as I mentioned above.

[ghorrocks]

How are you modelling this?

If you are modelling it using Eularian bubble model you use a degassing boundary as Opaque stated.

If you are using a Lagrangian particle tracking model you could just make the top surface a wall with coefficient of restitution =0. Then the bubbles will hit the wall and be stopped.

If you are using a free surface model then you will need to explicitly model somewhere for the air to go, so an opening boundary is a common choice.

[zhubohong]

Originally Posted by ghorrocks

How are you modelling this?

If you are modelling it using Eularian bubble model you use a degassing boundary as Opaque stated.

If you are using a Lagrangian particle tracking model you could just make the top surface a wall with coefficient of restitution =0. Then the bubbles will hit the wall and be stopped.

If you are using a free surface model then you will need to explicitly model somewhere for the air to go, so an opening boundary is a common choice.

Thanks for your reply. let me describe my model firstly. There are three phases in my model, injected gas, water and top air. The injected gas was treated as the dispersed phase, and both water and top air are continuous. I used the particle model to predict the bubble column in the water, and also applied the free surface model to reflect the variety of water free surface between water phase and top air phase. Now, the problem is that if I simultaneously apply these two models, the injected gas will affact the flow field of top air. So, it will indirectly affact the shape of free surface. However, the dominated effect on the free surface is the bubble column. Therefore, I want to do some adjustments, as I mentioned above. I want to let the injected gas disappear when it floats out the water/or when it meets the top air phase with the certain volume fraction. I think this method should be acheived by modifying the Source term.

[ghorrocks]

Can I ask how were we meant to guess what you were doing based on your first post? Why didn't you post this information right from the start?

Anyway, to answer your question: If you are using Lagrangian particle tracking you will need to define a track termination criteria for when it reaches the surface. I suspect this could be looking at the volume fraction and if it is lots of air or bubbles over a defined volume fraction you terminate the track.

But then, as you say, you will need to use a source term to generate as much air as you terminate in the particle track.