Converge problem of Mixture Model
 

I am using k-epsilon turbulent model to calculate two phase flow in 2D axisymmetric space (6*20 in). Water inlet (R=0.01in) is in the center of the chamber and air inlet(Rin=0.05in,Rout=0.103in) is an annular region around the water inlet. The velocity of air is 70m/s, that of water is 1.97m/s. I set the turbulent intensity of air is 4.957, hydrolic diameter is 0.00254. The values of k and epsilon for water are both 0.

I meshed this domain using Quad/Tri Pave, the mesh quality is pretty good.

---------------wall---------------------- | | | | | |

|pressure air inlet |outlet | | water inlet | ----------------axis---------------------

What I want to know is the slip velocity between these two phases. To use an Eulerian Model, I think the solution of Mixture model is a good initial condition to start with. When I calculate the mixture model, I first temporarily disable slip velocity equation in Solve/Control, and change the under-relaxation factor for Volume Fraction to 0.1. When iterating, the vf-water is never converge.

What is the problem with my settings? Is there anything missing?

Thanks a lot!

Jen

Re: Converge problem of Mixture Model
 

I don't know why it displays wrong of the geometry. :( It is like an rectangle, and one hole (water inlet) in the lower left conner, and one hole above(air inlet). The lower edge of this rectangle is Axis,upper edge is wall, right edge is pressure outlet.

Re: Converge problem of Mixture Model
 

Make sure that what you draw is within "" and "" tags. Like this dispersed flow in a pipe for example:



------------------------
o o o
O o
o o o o
------------------------

Re: Converge problem of Mixture Model
 

oops the tags are not there anymore. They are pre and backslash pre enclosed in lesser and greater than symbols.

Re: Converge problem of Mixture Model
 

Did I get that right -> 70 metres per second of Air? Is this some kind of high speed jet?

Re: Converge problem of Mixture Model
 

Yes. The air speed is 70m/s, and the water speed is 1.97m/s. Their ratio is pretty large. Do you have some methods to deal with this kind of problems?

Thanks,

Jen

Re: Converge problem of Mixture Model
 

I do not think the mixture model is appropriate. Please check with the guidelines in the manual. Maybe you need to do a VOF analysis or something?

Re: Converge problem of Mixture Model
 

I think Eulerian model is the best to know the slip velocity between the two phases, is that right? VOF can not provide this information.

Re: Converge problem of Mixture Model
 

I cannot give you a definite answer on that one. But for sure all three models can give you the slip (at least theoretically).