CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   CFX (https://www.cfd-online.com/Forums/cfx/)
-   -   Multiphase flow modeling (https://www.cfd-online.com/Forums/cfx/19814-multiphase-flow-modeling.html)

Paul August 5, 2003 17:14

Multiphase flow modeling
 
Hi, I am asking a quite general question. I am using CFX4.4 and trying to simulate multiphase (air-water) pipe flow of different flow regimes: bubble flow, churn flow, annular flow, slug flow. My understanding is that CFX4 has two models available: homogenous and two-fluid model. Whatever model I use, I can get good predictions only for cases of bubbly flow with low air void fraction. I have a bad luck for any annular flow even in a straight pipe. I am wondering if there are some tricks for multiphase flow simulations? And also how the multiphase capabilities are promoted in CFX5?For anyone experienced with multiphase flow, your response is appreciated.

Thank you.

Aaron Latty August 11, 2003 05:32

Re: Multiphase flow modeling
 
Before you read this, understand that I have only used CFX5 so I don't know what is possible in CFX4:

As a general rule, you are probably better off using the inhomogeneous model if you are concerned with how the bubbles will behave. The homogeneous model assumes that the fluids share a common velocity field. This might be appropriate for some slug flows, but generally, you will get better results with an inhomogeneous model.

Sometimes it is worth adding a third phase or possibly even a fourth phase (of air) to account for the different types of flow. You then have to worry about how these phases interact with each other. For slug flow, for example, you could model water as a continuous fluid, air as a continuous fluid and then create another phase called 'dispersed air' as a dispersed fluid. You can specify the mean diameter of the fluid particle depending on whether you are expecting large or small bubbles.

You can also play around with the free surfance settings. As I understand it, this provides a 'surface sharpening' algorithm that might improve the resolution of the interface between the water and air phases.

Good luck.

Aaron

Paul August 11, 2003 09:41

Re: Multiphase flow modeling
 
Good point. Thank you, Aaron.


All times are GMT -4. The time now is 23:49.