Learning Forum

[zukki]

Dear All,

I want to analyze the microbubble generator using ANSYS fluent or CFX. The purpose is to predict the number density and bubble diameter distribution of generated microbubbles. In the experiment, the diameter of the microbubbles is about 10 μm to 100 μm, has a peak value at 50 μm, and has a number density of several thousand number / ml.

In the microbubble generator, the microbubble can be generate by deformation or split the gas phase by at the high shear regions in swirling liquid. In the analysis, I think that the model of bubble splitting and coalescence is important.I have studied basice tutorials of two phase flow.

Fluent has a VOF and a hybrid model of the dpm model, and checked the spray can tutorial. In the case of microbubbles, is it possible to analyze if the liquid phase is the gas phase and the gas phase is the liquid phase, as opposed to the configuration of spray can tutorial?

At CFX, I checked the tutorial with the MUSIG model, but I could not know if it can be analyzed with this model. And I don't know how to set the each bubble group.

But, I am wondering whether to work with fluent or CFX software. Also, I don't know which model to start with.

Can you give me some advice?

Thank you.

[Rob]

I'd avoid VOF as that requires you to track the bubble free surface. It'll work but the cell count will be excessive. DPM has a built in collision & break up model but is designed for low volume fractions and the collision & break up models have their limitations. Population balance may be a better solution, and is available in Fluent and CFX: can better advise on the pros & cons of the model in the two solvers.

[DrAmine]

I won't use VOFtoDPM for Gas Bubbles: you will a mass deficit and this will scale with the size of the bubbles.
VOF alone should be enough if you can afford the spatial and temporal requirements.
If you cannot pay for VOF especially if you start having large numbers of bubbles, then the most efficient engineering solution is to rely either on the the Eulerian Population Balance Methods: MUSIG, IMUSIG or Discrete or QMOM) or rely Lagrangian Methods (especially oif the volume fraction of bubbles remains small).

[zukki]

Dear Rob
Dear DrAmine

Thank you for your useful advice.
I would like to challenge the prediction using MUSIG for microbubbles.
The results which I want in this analysis is to obtain the graph which has bubble diameter distribution on the horizontal axis and the number of bubbles on the vertical axis. Is it correct to assume that this can be obtained with MUSIG?
And, I have no experience with the MUSIG model, so I should study first. If you know any useful tutorials or videos for this problem, it would be greatly appreciated if you could let me know.
Thank you.