Hello, 

I am currently studying hydrodynamics of a bubble column and I have an issue when I want to simulate it with 3 phases. 

Firstly, I studied the bubble column with only 2 phases (the column is filled with liquid tin, liquid height = domain height = 120mm, and methane is injected from the bottom part) and I did not have any problem with my simulation.

Now, I want to study the impact of liquid height and how the interface moves. So I have to add another phase (air) to fill the domain. Bubble column's height is now 320mm, and liquid height is 120mm, so air height is 200mm. During the simulation, there is no problem when methane is injected in the liquid tin phase, but when methane arrives to the interface between liquid tin and air, it never goes through the interface. Methane is stacking at the interface and then just push out the air phase (I joined a picture of the situation to illustrate. Red phase is methane, green phase is liquid tin and blue phase is air).

Here is a description of my set up :

• 2D axisymmetric, transient
• multiphase model : VOF, 3 phases : air, liquid tin, methane (I tried with and without surface tension)
• energy on
• turbulence model : k-epsilon RNG

Boundary conditions :

• Velocity inlet, turbulence : intensity and hydraulic diameter, methane volume fraction = 1 (0 for air)
• Pressure outlet, turbulence : intensity and hydraulic diameter, air backflow volume fraction = 1 (0 for methane)
• I have 2 walls with a fixed temperature

As I thougth that the problem occured because of wrong backflow volume fraction, I tried every combinations of backflow VF, I tried every phases combinations too but I still have the problem.

I have no more ideas on which parameter I can change to solve the problem. Maybe it is just because this kind of set up is not available with VOF model ?

Thanks for the future answers

Nicolas.