CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   How to add a expression to Turbulent Prandtl Number (Prt) (http://www.cfd-online.com/Forums/cfx/68349-how-add-expression-turbulent-prandtl-number-prt.html)

Redpoko September 16, 2009 07:11

How to add a expression to Turbulent Prandtl Number (Prt)
 
hi, Guys,

I would like to use a expression in Turbulent Prandtl Number rather than a constant in CFX 12.0, the form maybe like this: Prt = f (x,y). For some certain purposes, I only want to change it in the Energy Equation.

I've been browsing through some posts here. But I didn't get the answer yet. There is one suggestion in 2006:

------------------------
Add the CCL below to the HEAT TRANSFER MODEL section (for each fluid domain)

HEAT TRANSFER MODEL:
Option = Thermal Energy
TURBULENT FLUX CLOSURE:
Option = Eddy Diffusivity
Turbulent Prandtl Number = 0.9
END
END

-----------------------

I don't know if I could do the same to replace 0.9 by f(x,y) in CFX 12.0

Is there anyone here could give me some ideas? Thanks in advance



Red

ghorrocks September 16, 2009 18:55

I don't know whether you can make the Prandtl number a function of anything, it might need to be a constant... But I am not sure. Try it and find out!

Redpoko September 17, 2009 04:30

thanks Glenn, I will give a try and later post the result here.

one more question: is the CCL above correct or not? if there is any other approach to change Prt in CFX 12.0?

ghorrocks September 17, 2009 07:35

You can easily check you CCL by either selecting the options you want in CFX-Pre then right click on the feature tree and select edit in command editor. Alternately you can enter the text in the command editor in CFX-Pre and check the right options are activated.

Redpoko September 23, 2009 10:56

it didn't work......

I give a simple expression to Turbulent Prandtl Number. The Warning Message is :

Parameter 'Turbulent Prandtl Number' in object '/FLOW:Flow Analysis 1/DOMAIN:Default Domain/FLUID MODELS/HEAT TRANSFER MODEL/TURBULENT FLUX
CLOSURE' is not allowed to be assigned an expression value that depends on variables. It must be assigned a numeric value, or an expression that resolves to a constant value.

now I have no idea how i could solve the problem.....

ghorrocks September 23, 2009 18:51

Well, the error message just about sums it up - if you want to define Turb Prandtl number as a function of anything you have to either develop your own turbulence model or drop CFX and use other software.

It may also be worthwhile talking to CFX support as they may have a hidden way of doing what you want.

Redpoko September 24, 2009 08:10

ya, I also think so.

I'll give Ansys a call to ask if they have some 'secret path' to solve it.

Thank you again for your patience and those answers !

Redpoko October 1, 2009 04:16

I got the answer from ANSYS. They said CFX 13.0 will probably solve the problem.

well, that's it.....

rockzh February 11, 2012 01:41

Hi,I have a question about the Turbulent Prandtl Number in CFX 12.0.
After importing the up CCL,where can i check the Prandtl Nb value in CFX12.0?

hustxinxin July 1, 2013 02:50

I have tested that, in CFX14.0, the Prandtl Number is not allowed to be assigned an expression value that depends on variables. Redpoko, have you solved your problem?

Quote:

Originally Posted by Redpoko (Post 231118)
I got the answer from ANSYS. They said CFX 13.0 will probably solve the problem.

well, that's it.....


ghorrocks July 1, 2013 03:07

I think this thread has thoroughly explained what is going on. Making Pr a variable introduces many complex high order terms and they are not implemented in CFX so this will not work. Alternately Pr might define parameters which are calculated once, whereas is you make it a variable then the variation of Pr is not properly accounted for.

It looks like V13 does not do the check for constant Pr but I bet it does not account for the terms and/or changes in other variable so use it at your own risk - no guarantees it is correct.

hustxinxin July 1, 2013 03:29

So, do you think it is a good way to assign the Pr a constant A, and treat the remained term (Pr-A) as a source term attaching to the turbulence equation?

Quote:

Originally Posted by ghorrocks (Post 436989)
I think this thread has thoroughly explained what is going on. Making Pr a variable introduces many complex high order terms and they are not implemented in CFX so this will not work. Alternately Pr might define parameters which are calculated once, whereas is you make it a variable then the variation of Pr is not properly accounted for.

It looks like V13 does not do the check for constant Pr but I bet it does not account for the terms and/or changes in other variable so use it at your own risk - no guarantees it is correct.


ghorrocks July 1, 2013 06:03

You have to treat the Pr as a constant, so if you can include the effect of non-constant Pr as a source term then that will be the only way of implementing it in CFX.


All times are GMT -4. The time now is 17:55.