Learning Forum

[4sur]

Hi all,

I have to simulate the appearance and movement of o2 bubbles in water due to its heating. I have done simulations in 2D and 3D with the same model and geometry, but the results I get are not the same. The eulerian model I have used is as follows:

• Primary phase: liquid (water + o2) / Secondary phase: bubbles (o2 + o2_copy)
• Forces: surface tension and wall adhesion activated
• Spieces mass transfer: from liquid (o2) to bubbles (o2). Following Henry's law and vant hoffs correlation
• Bottom wall's temperature is set to 338.15K and ambient temperature is 298.15K.

In both 2D and 3D cases, I have initializated the model the same way. In 3D, it takes around 30s to form all the bubbles and move until the highest part of my geometry. However, in the 2D model (the centre slice of the 3D geometry) bubbles are formed but do not move to the highest part of the geometry. Why can it be? Do I have to change anything from 3D to 2D? Is my model right?

Thank you in advance

[Karthik R]

Hello:
Technically, 2D and 3D simulations are not always the same, especially when it comes to bubble dynamics. 2D run is just modeling a slice and is a simplification. Therefore, I'm not sure if you should expect the same behavior in both your runs.
Karthik

[4sur]

Ok thank you for your ansewr. And is it normal that my 2D model is closer to what happens in reality than my 3D model?

[Karthik R]

Hello:
Broadly, yes! But, if we are talking about near-spherical bubbles, I would not expect the behavior to be the exactly same between 2D and 3D.
Karthik

[4sur]

Thank you very much! One last question, which model do you think that is more appropiated to simulate these bubbles, mixture model or eulerian?

[Karthik R]

Hello:
You can use both - depending on your modeling needs. If you are modeling something similar to a bubble column reactor (where the flow of the two phases are generally in the same direction), you can use the Mixture model. Please remember that both models are trying to model the macroscopic effect of these gas-liquid interactions.
Karthik

[4sur]

Thank you very much. And what should I do if I need to simulate sub-millimeter bubbles? I mean, the height of my dispositive is no more than 0.2 mm, so bubbles are going to be much smaller. Can I use any of these models, or would it be better to use a UDF?