Calculating the Friction Velocity

cbarry · August 19, 2009, 03:18

Hi everyone,

I'm trying to implement new wall functions for the kOmegaSST model in OpenFOAM 1.5.

There is just one major hurlde I keep coming across. Does anyone know how I can calculate the friciton velcoity in OpenFOAM?

Best regards,

Chris

ngj · August 19, 2009, 04:03

Hi Chris

It depends on whether or not you want to use low- or high Reynolds version of SST. If you are using low RE, then (nu() + nut()) U.snGrad() computed at the wall yields satisfactory results.
If on the other hand you are having high RE, then an assumption of a log-profile would give reasonable results.

Best regards,

Niels

cbarry · August 19, 2009, 05:34

Thank you very much Niels.
I am trying to implement a universal, non-linear wall function. For y+ > 30, i.e. the log layer, I want the wall function to take the following form:

omega = u*/ (sqrt(beta2) * kappa * y)

where u* = friction velocity

It is thus in effect a high Reynolds approach. How would you recommend I calculate the friction velocity in that case?

Best regards,

Chris

ngj · August 19, 2009, 05:38

Hi Chris

Then the current implementation of SST is what you want. As I understand it, it is doing exactly that.

Best regards,

Niels

cbarry · August 19, 2009, 06:00

Hi Niels,

as far as I can see the current implementation of SST calculates omega in the wall region as follows:

omega= sqrt (k) / (Cmu25 * kappa * y)

When I looked earlier on I couldn't see how the two expressions are equivalent. Are they actually equivalent then? If so, how?

In that case I'll try leaving that expression as it is and adding different functions for the buffer layer and the viscous sublayer.

Thank you for your quick help.

Best regards,

Chris

ngj · August 19, 2009, 07:58

The following is from [1]:

k = u_f^2 / sqrt(C_mu)

and if you combine this with your equation for omega, you obtain the implemented result.

Best regards,

Niels

[1]: Wall Function Boundary Condition Including Heat Transfer and Compressibility, RH Nichols and CC Nelson, AIAA, vol. 42, no. 6., june 2004.

cbarry · August 19, 2009, 08:34

Thank you

. It all makes sense now

madad2005 · November 17, 2009, 11:24

Chris,

Did you get any further with your blended wall function implementation?

Regards,

Adriano

August 19, 2009, 03:18	Calculating the Friction Velocity	#1
cbarry New Member Christopher Barry Join Date: Aug 2009 Location: Switzerland Posts: 18 Rep Power: 16	Hi everyone, I'm trying to implement new wall functions for the kOmegaSST model in OpenFOAM 1.5. There is just one major hurlde I keep coming across. Does anyone know how I can calculate the friciton velcoity in OpenFOAM? Best regards, Chris

August 19, 2009, 07:58		#6
ngj Senior Member Niels Gjoel Jacobsen Join Date: Mar 2009 Location: Copenhagen, Denmark Posts: 1,900 Rep Power: 37	The following is from [1]: k = u_f^2 / sqrt(C_mu) and if you combine this with your equation for omega, you obtain the implemented result. Best regards, Niels [1]: Wall Function Boundary Condition Including Heat Transfer and Compressibility, RH Nichols and CC Nelson, AIAA, vol. 42, no. 6., june 2004. albet likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Velocity in Porous medium : HELP! HELP! HELP!	Kali Sanjay	Phoenics	0	November 6, 2006 06:10
How to update polyPatchbs localPoints	liu	OpenFOAM Running, Solving & CFD	6	December 30, 2005 17:27
Variables Definition in CFX Solver 5.6	R P	CFX	2	October 26, 2004 02:13
friction velocity	kei-tee	Main CFD Forum	10	May 25, 2003 19:31
Terrible Mistake In Fluid Dynamics History	Abhi	Main CFD Forum	12	July 8, 2002 09:11

August 19, 2009, 04:03		#2
ngj Senior Member Niels Gjoel Jacobsen Join Date: Mar 2009 Location: Copenhagen, Denmark Posts: 1,900 Rep Power: 37	Hi Chris It depends on whether or not you want to use low- or high Reynolds version of SST. If you are using low RE, then (nu() + nut()) U.snGrad() computed at the wall yields satisfactory results. If on the other hand you are having high RE, then an assumption of a log-profile would give reasonable results. Best regards, Niels

August 19, 2009, 05:34		#3
cbarry New Member Christopher Barry Join Date: Aug 2009 Location: Switzerland Posts: 18 Rep Power: 16	Thank you very much Niels. I am trying to implement a universal, non-linear wall function. For y+ > 30, i.e. the log layer, I want the wall function to take the following form: omega = u/ (sqrt(beta2) kappa * y) where u* = friction velocity It is thus in effect a high Reynolds approach. How would you recommend I calculate the friction velocity in that case? Best regards, Chris

August 19, 2009, 05:38		#4
ngj Senior Member Niels Gjoel Jacobsen Join Date: Mar 2009 Location: Copenhagen, Denmark Posts: 1,900 Rep Power: 37	Hi Chris Then the current implementation of SST is what you want. As I understand it, it is doing exactly that. Best regards, Niels

August 19, 2009, 06:00		#5
cbarry New Member Christopher Barry Join Date: Aug 2009 Location: Switzerland Posts: 18 Rep Power: 16	Hi Niels, as far as I can see the current implementation of SST calculates omega in the wall region as follows: omega= sqrt (k) / (Cmu25 * kappa * y) When I looked earlier on I couldn't see how the two expressions are equivalent. Are they actually equivalent then? If so, how? In that case I'll try leaving that expression as it is and adding different functions for the buffer layer and the viscous sublayer. Thank you for your quick help. Best regards, Chris

August 19, 2009, 08:34		#7
cbarry New Member Christopher Barry Join Date: Aug 2009 Location: Switzerland Posts: 18 Rep Power: 16	Thank you . It all makes sense now

November 17, 2009, 11:24		#8
madad2005 Senior Member Anonymous Join Date: Mar 2009 Posts: 110 Rep Power: 17	Chris, Did you get any further with your blended wall function implementation? Regards, Adriano