CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Yplus values for walls in CFX4.4 (http://www.cfd-online.com/Forums/cfx/20205-yplus-values-walls-cfx4-4-a.html)

Forrest March 1, 2004 15:48

Yplus values for walls in CFX4.4
 
Hello, How to work out the Yplus values for wall boundary conditions? For example, a structured mesh from CFX4.4, I know the first cell height of the mesh but not sure how to estimate Yplus values. What are the general criteria for the Yplus values using the wall function when two equations turbulence models applied (say k-epsilon model)? Thank you for your help. Regards Forrest

Glenn Horrocks March 1, 2004 16:57

Re: Yplus values for walls in CFX4.4
 
Hi Forrest,

It has been a long time since I used CFX4.4. I think you can get the y+ values put into the dump file by selecting it as an additional variable or something like that. I know it is discussed in the documentation so have a look there.

CFX4.4 uses standard wall functions, meaning for the k-eps turbulence model it is only applicable for y+>11. That means there is actually a limit to the grid refinement you can do at the wall boundaries. There is no real upper limit to allowable y+ values (especially for high Reynolds number flows), but grid independance generally starts to fall apart at 100>y+>500.

Regards, Glenn

Forrest March 2, 2004 05:12

Re: Yplus values for walls in CFX4.4
 
Glenn,

Thank you for your reply. One more thing, if I import a CFX4.4 structured mesh into CFX5.6 and run it to get a results file, I can view the Yplus values using CFX5.6 Post. Can this be a way to examine Yplus for a structured mesh?

Thank you again

Regards, Forrest

Glenn Horrocks March 2, 2004 17:28

Re: Yplus values for walls in CFX4.4
 
Hi Forrest,

Assuming the results are the same between CFX4.4 and 5.6 then the y+ should be the same. Keep in mind that CFX5.6 by default uses different wall functions then 4.4, so if you want the results to be properly comparable you should run 5.6 with the same wall functions.

Glenn

Jeff March 2, 2004 19:53

Re: Yplus values for walls in CFX4.4
 
Actually, you'll not get the correct values in CFX-5, because CFX-5 is vertex based and CFX-4 is cell centered. This means your y+ values in CFX-4 will be roughly half the values in CFX-5 (half the distance to the wall).

To estimate the y+ in CFX-4, use the formula:

y+ = density * Vmean * wall dist / viscosity

This is good for a first cut at approximating what cell size you need at the walls given that the criteria of 30 < y+ < 100.

To see the actual solver values, add a user scalar and name it 'YPLUS'.

>>CFX-4
:>OPTIONS

.

.

USER SCALAR EQUATIONS 1

.
:>VARIABLE NAMES

USER SCALAR1 'YPLUS'

You can then plot this scalar on the walls to see where the min and max are and where you might need to modify the grid.

Hope this helps, Jeff

Forrest March 3, 2004 05:13

Re: Yplus values for walls in CFX4.4
 
Hi, Glenn and Jeff

Thank you both very much for your comments. No further questions for Yplus now. Thanks again

Regards, Forrest


All times are GMT -4. The time now is 16:37.