Hello to all

The gas bubbles induced liquid flow field in a water tank is to be calculated in my model. I got a problem in my calculation using DPM, that is i have to define the face(3d) or edge(2d) adjacent to the injections(randomly distributed on the face or edge)as a wall zone. But in this case, i don't know how to define its boundary conditions, since the injection is the only flow inlet, with a face or edge of the tank open to the air, without the flow outlet. the injiection particle is assumed to be bubbles of air with a fixed diameter.As the bubbles rise along the face or edge because of buoyancy, the continuous phase(water here)will be dragged to form a flow field in the tank. how to define the boundary conditions? how to initialize the calculation with the gas bubbles?

Any help will be appreciated!

Hi

I was thinking in you case.

In my opinion, the only way to estimate the flow field is applying multiphase models, because the Discrete Phase Model was developed under the assumption of low volume fraction (less than 12% DP/CP).

I guess you require a volume fraction higher than 12% to provide at least a gentle mixing. Even if you were able to use the discrete model available in fluent, you need a model for the continuous phase and none of them would give you a satisfactory answer.

The best model to simmulate mixing in tanks is the LES, however this model is incompatible with the DPM. Therefore, you should talk to someone in this forum about the potential to use LES and multiphase models for your case. Unfortunately, I have never worked with multiphase model, so I cannot give you a better insight.

Best regards

Alex Munoz

Hi, Alex Munoz

Thanks for your reply!

In my model the volume fraction of the dispersed phase is about 10%, so i want to use DPM. The problem you mentioned above do exist, it's not easy to find a model for the continuous phase.

I am not so familiar with the LES and Multphase model either. i will start to learn them. thanks again for your advice!

best regards

X.P.Lee

Hi, my friend

I think maybe you can define the face(3d) or edge(2d) adjacent to the injections(randomly distributed on the face or edge)as velocity inlet condition and set the velocity equal to zero. Then you can initialize the field using inlet zone, then you give the initial field with zero velocity everywhere.

Since I am not familiar with your problem, I don't know whether my advice is useful.

Good luck!

winnie

Hi, Winnie

Thanks a lot!

That's a good idea, i will try it!

Thanks!

X.P.Lee