[ROY4]

Hello,
I wanted to ask about the value of y+ in k epsilon model. I read that for k epsilon model it most be more than 11. In some places I read that it should be between 30 to 300. I am simulating a bubble column which has two phase flow. In most of the papers I have read till now, no one discussed about the value of y+. I don't know if it is not sensitive to y+ value or if they just did not mention it. My advisor told me that he thinks the y+ value must be less than 10. My supervisor who had worked on bubble column before says that they did not care about the value of y+.
Now I would like to ask:
- do you think that y+ matters in a bubble column? And if it matters, is it important on the wall around the column or the bottom wall that the air source points are defined on it?
- what should be the value of y+ in my case (two phase flow)? I get different values for different mesh but they are all more than 11.
Thanks for your answers in advance.

[Gert-Jan]

1. You could opt for SST and bother less about Y+.
2. Y+ is important if you want to solve the flow along a wall accurately, which will affect pressure drop, wall shear force, heat transfer, wall lubrication force, etc.

So, the main question is:
- what question do you want to solve using CFD? Is it the spatial distribution of your bubbles in a large volume, I think Y+ is not the most important parameter. Your phase interaction models are.
- how important is the wall in your question? Are the bubbles of the same size as your column, then Y+ can be important.
- Do you need to include heat transfer? then it is of prime importance.

So you tell........

[AtoHM]

For k-e-Model, there are actually two y-Plus values: one being the original Yplus, the other being SolverYplus. The original one refers to the actual Yplus of the first cell, the other is an artifical value used by the solver. It is not going below 11.06 (I think), so if you have a finer mesh, ANSYS will still use this value. In principal, both k-e and SST model are using mesh-resolution independend wall functions. Take look into the Modelling and Theory guide, the information is a bit split up there, but its all written down.

The k-e-Model does neglect the viscous sublayer in momentum and mass balance, you have to decide if its useable for you case - as Gert-Jan wrote.

[ROY4]

Originally Posted by Gert-Jan

1. You could opt for SST and bother less about Y+.
2. Y+ is important if you want to solve the flow along a wall accurately, which will affect pressure drop, wall shear force, heat transfer, wall lubrication force, etc.

So, the main question is:
- what question do you want to solve using CFD? Is it the spatial distribution of your bubbles in a large volume, I think Y+ is not the most important parameter. Your phase interaction models are.
- how important is the wall in your question? Are the bubbles of the same size as your column, then Y+ can be important.
- Do you need to include heat transfer? then it is of prime importance.

So you tell........

Dear friend
In my case I have a bubble column with quiescent water which air is injected through source points. I have no heat transfer or mass transfer now. There is only intraction between bubble and water and forces such as lift, drag, virtual mass, turbulent dispersion. And the parameters that are mostly reported are bubble size distribution, gas hold up, liquid axial velocity, k and epsilon values. These are time averaged. As I said before most of papers I have checked , which are plenty of them, did not report or talked about y+.
I don't know if I should care about it or in my case I can just ignore it. I don't even know what other researchers did?
My problem is that no matter how Much I make the mesh finer adjacent to the wall, the y+ is still more than 11.
I read in this site that even if the grid is some how that the y+ is less than 11, the k epsilon model which uses scalable wall function will use the y+ more than 11 and ignores the user's grid. Now what should I do? Should I just ignore it?
If not, what is the value of y+ for k epsilon model? Is it less than 1??less than 10 ?? Among 30 to 300?? Or what?
Thanks