Turbulence models and wall boundary conditions

eugene · August 1, 2008, 07:15

G ~= tauw*(dU/dy)p, G != tauw*Up/yp, i.e. (dU/dy)p != Up/yp

Log-law
U = utau/kappa*ln(y*utau/nu * E);

or

dU/dy = utau /(kappa*y);

utau = Cmu25*sqrt(k);

Clear?

sradl · August 2, 2008, 02:50

Thanks Eugene, this is clear to me.

BUT: My literature (lecture notes from a AVL guy) tells me G=tau_wall*UP/yP. Furthermore, something like

G=tauw*UP/yP-rho*Cmu75*kP^1.5*u^+/yP

can be found in Versteeg and Malalasekera (Computational Fluid Dynamics, 1995), which was a little confusing to me.

However, OF'S approach sounds more physical.

By the way: can you recommend some recent literature for this kind of aspects (except your PhD thesis)?

cheers
Stefan

eugene · August 4, 2008, 06:45

This one:

G=tauw*UP/yP-rho*Cmu75*kP^1.5*u^+/yP

:
utau^3 = Cmu75*kP^1.5
utau^2 = tauw/rho
and
u+ = UP/utau

thus

G=tauw*UP/yp - tauw*UP/yP

Doesn't sound very correct to me unless some of my assumptions were wrong.

But don't take my word for it, you can work it out yourself. The definition of G is:

G = 1/yN * int(tau_turb * dU/dy.dy)[0-yN]

Normally people make various assumptions to simplify things, like the laminar region is negligible, tau_turb is constant and equal to tauw , etc.

For recent literature, do a google on "UMIST wall functions".

August 2, 2008, 02:50	Thanks Eugene, this is clear t	#22
sradl Member Stefan Radl Join Date: Mar 2009 Location: Graz, Austria Posts: 82 Rep Power: 18	Thanks Eugene, this is clear to me. BUT: My literature (lecture notes from a AVL guy) tells me G=tau_wallUP/yP. Furthermore, something like G=tauwUP/yP-rhoCmu75kP^1.5*u^+/yP can be found in Versteeg and Malalasekera (Computational Fluid Dynamics, 1995), which was a little confusing to me. However, OF'S approach sounds more physical. By the way: can you recommend some recent literature for this kind of aspects (except your PhD thesis)? cheers Stefan

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Wall functions and turbulence models implementations	jposunz	OpenFOAM Running, Solving & CFD	2	October 23, 2009 04:02
wall boundary conditions	sivaramakrishnaiah	CFX	2	August 21, 2008 10:25
Turbulence models and boundary layer	Stanislav Kraev	FLUENT	1	March 14, 2006 05:55
Turbulence models integrable upto the wall	vishwas	FLUENT	1	January 31, 2006 16:59
Turbulence : k-epsilon boundary conditions	Mohamed Sofiane KHELLADI	Main CFD Forum	1	April 25, 2000 20:10

August 1, 2008, 07:15	G ~= tauw(dU/dy)p, G != tauw	#21
eugene Senior Member Eugene de Villiers Join Date: Mar 2009 Posts: 725 Rep Power: 21	G ~= tauw(dU/dy)p, G != tauwUp/yp, i.e. (dU/dy)p != Up/yp Log-law U = utau/kappaln(yutau/nu * E); or dU/dy = utau /(kappay); utau = Cmu25sqrt(k); Clear?

August 4, 2008, 06:45	This one: G=tauwUP/yP-rho	#23
eugene Senior Member Eugene de Villiers Join Date: Mar 2009 Posts: 725 Rep Power: 21	This one: G=tauwUP/yP-rhoCmu75kP^1.5u^+/yP : utau^3 = Cmu75kP^1.5 utau^2 = tauw/rho and u+ = UP/utau thus G=tauwUP/yp - tauwUP/yP Doesn't sound very correct to me unless some of my assumptions were wrong. But don't take my word for it, you can work it out yourself. The definition of G is: G = 1/yN int(tau_turb * dU/dy.dy)[0-yN] Normally people make various assumptions to simplify things, like the laminar region is negligible, tau_turb is constant and equal to tauw , etc. For recent literature, do a google on "UMIST wall functions".