Learning Forum

[zj X]

Hello, everyone!
I am simulating gas-liquid-solid three-phase mass transfer. Also I want to simulate the aggregation and breakage of bubbles, but I am distressed with the fact that after adding the PBM model I find that there is a bubble particle size distribution and liquid velocity above the liquid surface yet I can't monitor the liquid phase volume fraction.

Now, I'm not sure what is wrong. All I know is that the above problem occurs when I use PBM model. I have tried using udf to suppress the liquid velocity above the liquid level but it doesn't help. I have also tried modifying the mesh and using higher order discretization to reduce numerical diffusion but that didn't help either.

So the question is how to avoid having liquid velocity and bubble particle size distribution above the liquid surface when using PBM

Here is the specific setup for my simulation: an Eulerian multiphase flow model with a mixture of gas-liquid-solid phases, with a liquid phase (water + liquid toluene) in the primary phase, and gas (air + gaseous toluene) and solid phases (silicon + liquid toluene) in the secondary phase, with gas-liquid, gas-solid, and liquid-solid mass transfer.Before the simulation, I patch a part of the area above the model with a gas phase volume fraction of 1.

Thanks in advance for your patience.

[Rob]

What is the liquid volume fraction in the head space? 

[zj X]

There is no liquid volume fraction at the head space（at very small scales there is still no），I dont know why.

[Rob]

There will be a tiny amount, and that's why you're seeing a velocity result. 

[zj X]

Thanks for your replying.

However, I use “volumeInt” of the liquid volume fraction in head space results in 0，is that normal?
Also, I found some similarity in the velocity contour of the gas and liquid, and guess that the gas is carrying the liquid upward.Could I do anything to avoid this？

[Rob]

It's unlikely that you'll get zero liquid in the head space, and it's that tiny volume fraction that gives the velocity field you're seeing. One trick is to use a Custom Field Function (in the User Defined tab) of vol_frac_liquid x velocity as that'll mostly eliminate the odd result. Clip planes are another option. 

[zj X]

Thanks for your replying.I will try Custom Field Function.

And is there a solution to the bubble bin fraction in the head space? 
I write an udf to make the aggregation rate only below the liquid surface by limiting the height, but it didn't help.There are still bubble bin fraction above the liquid surface.
I can give more information if you need.

[Rob]

The bubble size will be determined by the PBM model, so what in the model will switch over to putting all of the gas into the largest bin size? 

[zj X]

I use default luo aggregation model and luo breakage model to simulate bubble aggregation and breakage.
When the gas does not reach the liquid surface, there is no bin size in the headspace, but when the gas breaks through the liquid surface, bin size appears in the headspace.
It looks like the gas is aggregating and breaking in the headspace, which doesn't make sense.
The bottom is the velocity-inlet for the gas. The gas rises from the bottom to the outlet.Pressure-outlet at the top，all bin fraction‘s boundary value is 0，is there mistakes in my settings?
Thanks for your patience.

[Rob]

Why doesn't it make sense? The PBM set up has a trigger to break & form bubbles, and the Eulerian model has a size associated with the second phases. 

[zj X]

So I get all normal results.
Thanks for all your replies Mr. Rob.