BCs for k and epsilon at walls with y+=1

vangelis · October 15, 2012, 15:18

Hello everybody,

I would like to know the following for OpenFOAM 2.1.1:

when setting up a turbulent flow case with a mesh that leads to y+ values of 1 I do not need to use wall functions.
What is the correct BC for k and epsilon in such a case
Are they fixedValue=0 or
zeroGradient?

Similar question for k-omega models also.

I have found this post
http://www.cfd-online.com/Forums/ope...m-1-5-dev.html
but it refers to previous OF releases

Thanks!

sylvester · October 16, 2012, 03:22

Hi Vangelis,

To my understanding you would (theoretically) set both at fixedValue = 0, as there is no turbulence possible in the viscous sublayer. And as there is no turbulence, you also can't have dissipation of turbulence.

I say theoretical because actually setting the BC's to 0 would result in a division by zero in the OpenFOAM implementation. In practice you would therefore set them to a very small value, for example fixedValue = 1e-12.

Best regards,
Sylvester

vangelis · October 16, 2012, 03:51

Hi Sylvester!

Thank you very much for your reply.

Best regards

Vangelis

vangelis · October 16, 2012, 08:57

Hi Sylvester

I forgot to also specify that
nut should also be set to fixedValue=0
like k, right?

Vangelis

sylvester · October 16, 2012, 09:14

Hi Vangelis,

Or use nutLowReWallFunction, which makes the patch accessible from yPlusRAS.
From nutLowReWallFunctionFvPatchScalarField.H:

Quote:

Description
Boundary condition for turbulent (kinematic) viscosity when for use with
low Reynolds number models.

Sets nut to zero, and provides an access function to calculate y+.

regards,
Sylvester

ps. Can we perhaps expect improved OpenFOAM pre-processing in ANSA v13.3?

vangelis · October 16, 2012, 10:25

We are trying to do this.
And in this sense thank you for your help!

so you mean that it is better to use
nutLowReWallFunction

so that afterwards I can get values for y+?

Otherwise it would be the same as setting fixedValue=0 for nut at walls?

sylvester · October 16, 2012, 10:52

Hi Vangelis,

Yes, from what I can gather it is better to use the nutLowReWallFunction. But I haven't done much work with low Re turbulence models, so I may be wrong here.

Best regards,
Sylvester

vangelis · October 16, 2012, 11:45

Thanks again Sylvester!

October 15, 2012, 15:18	BCs for k and epsilon at walls with y+=1	#1
vangelis Senior Member Vangelis Skaperdas Join Date: Mar 2009 Location: Thessaloniki, Greece Posts: 287 Rep Power: 21	Hello everybody, I would like to know the following for OpenFOAM 2.1.1: when setting up a turbulent flow case with a mesh that leads to y+ values of 1 I do not need to use wall functions. What is the correct BC for k and epsilon in such a case Are they fixedValue=0 or zeroGradient? Similar question for k-omega models also. I have found this post http://www.cfd-online.com/Forums/ope...m-1-5-dev.html but it refers to previous OF releases Thanks! Last edited by vangelis; October 15, 2012 at 16:44.

October 16, 2012, 03:22		#2
sylvester Member Roland Join Date: Mar 2009 Location: Netherlands Posts: 92 Rep Power: 17	Hi Vangelis, To my understanding you would (theoretically) set both at fixedValue = 0, as there is no turbulence possible in the viscous sublayer. And as there is no turbulence, you also can't have dissipation of turbulence. I say theoretical because actually setting the BC's to 0 would result in a division by zero in the OpenFOAM implementation. In practice you would therefore set them to a very small value, for example fixedValue = 1e-12. Best regards, Sylvester saeed jamshidi likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Epsilon boundary condition for walls in LamBremhorst LowRe model	maruthamuthu_venkatraman	OpenFOAM	1	April 21, 2019 06:25
SimpleFoam k and epsilon bounded	nedved	OpenFOAM Running, Solving & CFD	16	March 4, 2017 08:30
How to set up BCs for k epsilon model	Julian K.	OpenFOAM Pre-Processing	44	October 26, 2015 09:23
Dealing with BC's in OF 1.6	vkrastev	OpenFOAM Running, Solving & CFD	5	September 4, 2012 11:58
SimpleFoam k and epsilon bounded	nedved	OpenFOAM Running, Solving & CFD	1	November 25, 2008 20:21

October 16, 2012, 03:51		#3
vangelis Senior Member Vangelis Skaperdas Join Date: Mar 2009 Location: Thessaloniki, Greece Posts: 287 Rep Power: 21	Hi Sylvester! Thank you very much for your reply. Best regards Vangelis

October 16, 2012, 08:57		#4
vangelis Senior Member Vangelis Skaperdas Join Date: Mar 2009 Location: Thessaloniki, Greece Posts: 287 Rep Power: 21	Hi Sylvester I forgot to also specify that nut should also be set to fixedValue=0 like k, right? Vangelis

October 16, 2012, 10:25		#6
vangelis Senior Member Vangelis Skaperdas Join Date: Mar 2009 Location: Thessaloniki, Greece Posts: 287 Rep Power: 21	We are trying to do this. And in this sense thank you for your help! so you mean that it is better to use nutLowReWallFunction so that afterwards I can get values for y+? Otherwise it would be the same as setting fixedValue=0 for nut at walls?

October 16, 2012, 10:52		#7
sylvester Member Roland Join Date: Mar 2009 Location: Netherlands Posts: 92 Rep Power: 17	Hi Vangelis, Yes, from what I can gather it is better to use the nutLowReWallFunction. But I haven't done much work with low Re turbulence models, so I may be wrong here. Best regards, Sylvester

October 16, 2012, 11:45		#8
vangelis Senior Member Vangelis Skaperdas Join Date: Mar 2009 Location: Thessaloniki, Greece Posts: 287 Rep Power: 21	Thanks again Sylvester!