Yplus and wall functions for an airfoil

fede_freefly · December 3, 2014, 18:56

Hi guys!
That's my first time here, and I'm pretty new on CFD, thus I'll probably ask you something really basic.
I'm actually dealing with RANS simulation of a flow over an airfoil, with a Reynolds number of 300'000. I need to analyze the behaviour of lift and drag coefficients at changings of the angle of attack, from 0° till about 13°.

Now my problem is about the Yplus, the heigth of 1st cell and the wall functions. I know that i have to choose a certain value of Yplus to obtain the 1st cell spacing. People told me to choose Y+ = 1 but, with my data, the spacing ends up to be really small (about 0.03 mm), creating a too dense mesh. I've red that using wall funtions could avoid this problem, allowing you to use a higher value of Y+ (i guess i understood that these functions assume true the logaritmic law of the near-wall velocity)

Now my questions are:
-Which is the wall function i should use for this problem?
-therefore, what Y+ do i have to take?
-if i can use higher y+, will i miss information about the pressure coefficients?

Hope you can help !!

LuckyTran · December 5, 2014, 00:10

You don't really have much of a selection to choose any wall functions.

Only thing I'll say is that if you're using kwSST or k-omega models then your y+ had better be close to 1. For k-epsilon based models, enhanced wall treatment is wayy better than standard wall functions and is probably a good place to start. Most if not all of the wall functions used in Fluent are blended wall functions so you don't have to worry about getting a large enough y+ to make sure the log law is being used. Just make the mesh as fine as you can afford and let it crank.

fede_freefly · December 5, 2014, 05:53

Thanks for the answer !

Do you think k-epsilon is a good model to predict the pressure coefficients on an airfoil with a mesh C-type?
I have to calculate them till the stall, and i don't know if k-epsilon works good when the boundary separation happens.

Moreover is it possible to see the flow ricirculation with the k-epsilon? cause I've done some tries with high angles of attack and it doesn't seem to appear on the graphic animations (With an y+<10 on the mesh).

Thanks !

December 3, 2014, 18:56	Yplus and wall functions for an airfoil	#1
fede_freefly New Member Federico Join Date: Dec 2014 Location: Rome Posts: 7 Rep Power: 12	Hi guys! That's my first time here, and I'm pretty new on CFD, thus I'll probably ask you something really basic. I'm actually dealing with RANS simulation of a flow over an airfoil, with a Reynolds number of 300'000. I need to analyze the behaviour of lift and drag coefficients at changings of the angle of attack, from 0° till about 13°. Now my problem is about the Yplus, the heigth of 1st cell and the wall functions. I know that i have to choose a certain value of Yplus to obtain the 1st cell spacing. People told me to choose Y+ = 1 but, with my data, the spacing ends up to be really small (about 0.03 mm), creating a too dense mesh. I've red that using wall funtions could avoid this problem, allowing you to use a higher value of Y+ (i guess i understood that these functions assume true the logaritmic law of the near-wall velocity) Now my questions are: -Which is the wall function i should use for this problem? -therefore, what Y+ do i have to take? -if i can use higher y+, will i miss information about the pressure coefficients? Hope you can help !! aapinillosr likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[Other] Contribution a new utility: refine wall layer mesh based on yPlus field	lakeat	OpenFOAM Community Contributions	58	December 23, 2021 03:36
Wall functions - questions about implementation	AlmostSurelyRob	OpenFOAM Programming & Development	44	May 6, 2020 11:11
K-kl-transition model y+	Shamoon Jamshed	FLUENT	13	June 17, 2016 15:38
Solver Yplus of Automatic wall treatment	justjhy	CFX	2	May 4, 2013 08:19
Water subcooled boiling	Attesz	CFX	7	January 5, 2013 04:32

December 5, 2014, 00:10		#2
LuckyTran Senior Member Lucky Join Date: Apr 2011 Location: Orlando, FL USA Posts: 5,761 Rep Power: 66	You don't really have much of a selection to choose any wall functions. Only thing I'll say is that if you're using kwSST or k-omega models then your y+ had better be close to 1. For k-epsilon based models, enhanced wall treatment is wayy better than standard wall functions and is probably a good place to start. Most if not all of the wall functions used in Fluent are blended wall functions so you don't have to worry about getting a large enough y+ to make sure the log law is being used. Just make the mesh as fine as you can afford and let it crank.

December 5, 2014, 05:53		#3
fede_freefly New Member Federico Join Date: Dec 2014 Location: Rome Posts: 7 Rep Power: 12	Thanks for the answer ! Do you think k-epsilon is a good model to predict the pressure coefficients on an airfoil with a mesh C-type? I have to calculate them till the stall, and i don't know if k-epsilon works good when the boundary separation happens. Moreover is it possible to see the flow ricirculation with the k-epsilon? cause I've done some tries with high angles of attack and it doesn't seem to appear on the graphic animations (With an y+<10 on the mesh). Thanks !