CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Possible to account for surface roughness with SST model? (http://www.cfd-online.com/Forums/cfx/65302-possible-account-surface-roughness-sst-model.html)

LSC June 11, 2009 00:51

Possible to account for surface roughness with SST model?
 
I need to account for surface roughness for my aerofoil simulations but found it is possible to input roughness height for the k-epsilon models and not SST. I am wondering is it possible to work around and account for surface roughness while using SST model with transition? Please advise

ckleanth June 11, 2009 01:47

I just tried it and the option was avaiable with the sst model in v12 (however sometimes the gui does play tricks) if its not avaiable try to add the option on your wall boundary using the same ccl as in the k-e case

maybe you already know this but i thought to mention it; you need to be careful with the wall roughness as the sand roughness height option as used in cfx is not the standard roughness factor found commonly in textbooks

LSC June 11, 2009 02:04

I am currently using V11. Would it still work?

ckleanth June 11, 2009 02:23

:( it appears its a new feature in v12

Timon June 11, 2009 03:00

As ckleanth mentioned you have to add the ccl manually in V11 (this indeed works).

Just add this section to your no slip wall boundary (with RH to be specified of course):

WALL ROUGHNESS:
Option = Rough Wall
Roughness Height = RH
END

ckleanth June 11, 2009 04:37

Quote:

Originally Posted by Timon (Post 218910)
As ckleanth mentioned you have to add the ccl manually in V11 (this indeed works).

Just add this section to your no slip wall boundary (with RH to be specified of course):

WALL ROUGHNESS:
Option = Rough Wall
Roughness Height = RH
END

Well there are loads of hidden features in CFX - its quite interesting when you do find them out ;)

ghorrocks June 11, 2009 07:01

Hi,

Surface roughness in SST was a beta feature in V11. I believe it is now fully released in V12 so it now appears on the GUI. I don't know if it was developed before release - best go to the current version (V12) to be sure you have the most current version.

Glenn Horrocks

LSC June 11, 2009 22:54

Hi,

is the editing of CCL to account roughness works in V11(not beta). Anyway just another question, I am now working on forced transition for my aerofoil simulations. I have divided the mesh into Laminar (up to the explicitly determined transition point) and Turbulent Zones in Gambit. I have created a subdomain in CFX for the Laminar Zone. I am wondering which source term I should turn off for the Laminar Zone. I have read that turbulent eddy viscosity should be turn off but I don't see this option in the source tab. Can anyone shed some light?

ghorrocks June 11, 2009 23:50

Hi,

No you have missed the point of the turbulence transition model. When you turn the turbulence transition model on what it does is modify the SST model so it can handle laminar and turbulent flow. It then needs a sub-model to get the turbulence transition point. The default model for the transition point is the Langtry-Menter model, but if you know where the transition point is up-front you can choose your own model and you can explicitly set the transition point. But you are still working within the turbulence transition model.

There is no need for source terms or fiddling with the turbulent viscosity relations.

Glenn Horrocks

LSC June 12, 2009 00:11

Hi Glenn,

I do have some predicted transition point data from journals for my aerofoil profile and they leads to reasonably good results. For my case, I started off with default values using the transition model and although the CL results is good the CD error is quite large. I did a CF plot and found that the transition points were predicted too early which resulted in large drag deficit. How do I go about explicitly input the transition locations( using the transition point data I have) in CFX?

ghorrocks June 12, 2009 01:46

Hi,

Look in the documentation under "Gamma Theta Transition Model". The Gamma Theta model is the default two equation model. A one equation and a zero equation are also available. The zero equation model is what you are looking for where you specify the transition point by defining the intermittency variable explicitly, usually through a CEL expression.

Glenn Horrocks

LSC June 12, 2009 02:03

Hi Glenn,

how do I relate the intermittency variable to the transition point value I have? I read through the help files but could not find any clues. For example if I want to specify transition location at x/c = 0.36, I would have to define a relationship using the CEL expression and input the expression varaible name in the intermittency? Sorry for being a noob as I am completely new to CFD..

ghorrocks June 12, 2009 08:45

Hi,

You'll have to read the documentation for the details. I have never used it, I just know that's how you do it.

The expression to trigger the transition can be a CEL function (sin, cos, +,-,*,/ etc) or an interpolation function (either 1D or 3D). Use whichever approach will describe your transition best.

Glenn Horrocks

Tom June 18, 2009 22:24

Would like to learn how to add roughness to the SST model too.
Beta features are enable.

I tried adding creating a CCL file using a notepad and i added the line

BOUNDARY CONDITIONS:
WALL INFLUENCE ON FLOW:
Option = No Slip
WALL ROUGHNESS:
Option = Rough Wall
Roughness Height = 0.0001 [m]
END

I got the message

ERROR
CCL validation failed with message:
Error: Sub-object 'WALL ROUGHNESS' is not allowed in /BOUNDARY CONDITIONS/WALL INFLUENCE ON FLOW

I edited it to

BOUNDARY CONDITIONS:
WALL INFLUENCE ON FLOW:
Option = No Slip
END
WALL ROUGHNESS:
Option = Rough Wall
Roughness Height = 0.0001 [m]
END

There is no problem importing it but when i run the solver the command is not added.
I tried editing the the wall to Scalable. Same thing there are no changes. Anyone could write a small guide and pm me. Thanks in advance.

ghorrocks June 18, 2009 23:08

Hi,

Look in the RULES file. It is located in (CFX ROOT)/etc/RULES. It specifies the full CCL syntax of all commands. You will see the syntax there.

Glenn Horrocks

Tom June 18, 2009 23:33

Ok i get what you mean i placed it in the wrong tree of the command line.

I have corrected it. How do i activate it. I tried importing the the CCL file with those few lines but when i run the solver it does not appear there. How do i start it or will it be shown in the GUI.

LSC June 19, 2009 00:48

Hi,

does this work in V11( not beta version)? My school only has V11(not beta)..Anway Hi Tom! looks like we are looking into the same problem.

Tom June 19, 2009 01:07

Some feature are in Beta not V11 itself. V11 is a full working software. Im working on a cylinder with roughness. Just want to compare transitions and wanted to try add roughness on to it.

LSC June 19, 2009 01:16

Hi Tom,

understand that some of the new features are not available in V11. i am actually studying the effect of surface roughness on airfoil. So far, SST model with transition predicted relatively well for my case but not for the K-epsilon family models. With V11, I can only account for surface roughness using the k-epsilon models and not SST. It would be good if I can account for roughness using SST model with V11. Can anyone shed some light as to whether editing the CCL to account for roughness using SST model actually works for V11?

Timon June 19, 2009 03:08

As I stated earlier, roughness height for SST can indeed be enabled as a beta feature in v11 by manually adjusting the ccl. If you properly edit the ccl, it works...

However, in v11 it is beta and therefore not thoroughly validated. I don't now if there were any big changes going to the full release in v12, but in our experience the model worked fine in v11 as well.

I always used the cfx5cmds command to edit the ccl:

extracting ccl:

"cfx5cmds -read -def yourdeffile.def -ccl yourcclfile.ccl"

then edit the yourcclfile.ccl and write it back to the yourdeffile.def with:

"cfx5cmds -write -def yourdeffile.def -ccl yourcclfile.ccl"


All times are GMT -4. The time now is 23:34.