CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FLUENT (
-   -   Multiphase Flow Problem (

Robi July 12, 2005 14:31

Multiphase Flow Problem
Problem: I am trying to model a 2-D two phases flow using FLUENT multiphase model. I am considering water and air as the two phases. I am defining the two zones separately in GAMBIT itself. For the interface edge, I am defining the interface boundary. In Fluent when I choose the Multiphase model, it automatically chooses the mixture of air and water (ofcourse I choose the water in the material panel also). But I don't want it to be mixture. I just want both phases to be present separately and interact through the interface. For that I want one face to be full of water and another face to be full of air. I might have to define the interface boundary conditions by the UDF. Is there anyone who could help? I would really appreciate your help.


samim July 12, 2005 16:07

Re: Multiphase Flow Problem
You need the VOF model. -sami

matt July 12, 2005 18:44

Re: Multiphase Flow Problem
You may want to consider "patching" in water (or air) as the secondary phase in FLUENT itself rather than pre-defining zones in Gambit.

Adapt --> Region --> define secondary phase zone Solve --> Initialize --> patch in secondary phase

This is covered in FLUENT 6.2 Tutorial Guide #16 "Using the VOF Model"

Good luck !

Robi July 13, 2005 10:31

Re: Multiphase Flow Problem
Matt I understand what you are saying i.e. like in the fluent tutorial. But when I went for adept->region, I can not define the secondary phase zone ( because it is very irregular). In fluent it only has option that you can either adept rectangular or circle or you can give the minimum and maximum coordinate of two points only. Could you please explain how could I adept irregular shape also?

Thanks a lot, Robi

matt July 13, 2005 10:48

Re: Multiphase Flow Problem
Hello Robi, actually I've only ever had to use very ordinary patches but is it possible to describe your irregular interface with a set of adapted regions which can then be patched one by one ?

The other thing that may work is using the two zones you've made already define a radiator boundary condition between them (in Gambit). This is an interior BC which can be changed in FLUENT to "default-interior" BC - pretty innoxious.

But what is your simulation? I'm not sure what you mean by wanting the two phases to interact but not to mix - stratified flow?

Robi July 13, 2005 12:58

Re: Multiphase Flow Problem
Matt, I was trying to adept the region by using very ordinary patches (one rectangle and one triangle). But they do not have that option for triangle(for rectangle it is easy). This is so ridiculus that the user can not adept irregular shapes. Only if I could adept the triangle I could simulate the problem. Let me know if you know some other way to adept.

Thanks a lot. Regards, Robi

matt July 13, 2005 13:23

Re: Multiphase Flow Problem
Agreed, this is a clear limitation of FLUENT (and one that would seem easy to eliminate)

I can't think of how you would patch a triangle so I'd go the interior BC route between the two zones you've already created in Gambit - something to try anyway. I've had luck getting this technique to distinguish between two zones (seperated by the BC) for some particle tracking experiments...

Good Luck !


samim July 15, 2005 16:23

Re: Multiphase Flow Problem
You may want to try the Grid-Rotate feature. Since FLUENT is using the global coordinates, you may trick it by rotating your model about an axis such that the new reactangle actually covers half the area. This would show up as a triangle when you diaply the register.

You may also try a custom field function defining the triangle. In the VOF tutorial, it is explained in step 7, item 5 and 6. good luck

khalid July 23, 2005 05:02

dear sir: I want some tutorials in stratified two phase flow modeling so i was thankful if you send me any tutorial in this subject

All times are GMT -4. The time now is 11:53.