epsilonWallFunction and omegaWallFunction G for k omega and k epsilon

TomWaits · May 18, 2024, 12:45

In both the epsilon and omega wall functions, the production term for the models is specified as

$G=w(\nu_{tw}+\nu_w)|\mathbf{n}\cdot(\nabla \mathbf{u})_f | C_\mu^{1/4}\frac{\sqrt{k}}{\kappa y}$

if y+ >= y+lam.

What do the vertical bars in "|n" and "grad u_f|" mean?

What are the corner weights w?

And does this mean the production term is turned off for the governing equations for k and epsilon if y+ < y+lam? Why is this done? I can't find any mention of this in the original papers.

Many thanks for any help,
Tom waits

alexeym · May 19, 2024, 09:03

Hi,

1. Vertical lines are magnitude operator.
2. Corner weights are calculated for the corner cells, where contribution of several faces should be accounted. Basically it is 1/number of cell boundary faces.
3. If I unwinded if condition correctly:

Code:

        if (!lowReCorrection_ || (yPlus(facei) > yPlusLam))
        {
            G0[faceCells[facei]] +=
                cornerWeights[facei]
               *(nutw[facei] + nuw[facei])
               *magGradUw[facei]
               *Cmu25*sqrt(k[faceCells[facei]])
               /(kappa*y[facei]);
        }

G == 0 for y+ < yLaminar only if we use low Reynolds correction. Otherwise it is calculated.

There is A Thorough Description Of How Wall Functions Are Implemented In OpenFOAM (https://www.tfd.chalmers.se/~hani/ku...nfoamFinal.pdf) for version 4.0, but implementation of the wall functions did not change much since then.

May 18, 2024, 12:45	epsilonWallFunction and omegaWallFunction G for k omega and k epsilon	#1
TomWaits Member Tom Waits Join Date: Aug 2018 Posts: 42 Rep Power: 8	In both the epsilon and omega wall functions, the production term for the models is specified as $G=w(\nu_{tw}+\nu_w)\|\mathbf{n}\cdot(\nabla \mathbf{u})_f \| C_\mu^{1/4}\frac{\sqrt{k}}{\kappa y}$ if y+ >= y+lam. What do the vertical bars in "\|n" and "grad u_f\|" mean? What are the corner weights w? And does this mean the production term is turned off for the governing equations for k and epsilon if y+ < y+lam? Why is this done? I can't find any mention of this in the original papers. Many thanks for any help, Tom waits

May 19, 2024, 09:03		#2
alexeym Senior Member Alexey Matveichev Join Date: Aug 2011 Location: Nancy, France Posts: 1,938 Rep Power: 39	Hi, 1. Vertical lines are magnitude operator. 2. Corner weights are calculated for the corner cells, where contribution of several faces should be accounted. Basically it is 1/number of cell boundary faces. 3. If I unwinded if condition correctly: Code: if (!lowReCorrection_ \|\| (yPlus(facei) > yPlusLam)) { G0[faceCells[facei]] += cornerWeights[facei] (nutw[facei] + nuw[facei]) magGradUw[facei] Cmu25sqrt(k[faceCells[facei]]) /(kappa*y[facei]); } G == 0 for y+ < yLaminar only if we use low Reynolds correction. Otherwise it is calculated. There is A Thorough Description Of How Wall Functions Are Implemented In OpenFOAM (https://www.tfd.chalmers.se/~hani/ku...nfoamFinal.pdf) for version 4.0, but implementation of the wall functions did not change much since then.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Low-Re Modelling	doerschlott	OpenFOAM Running, Solving & CFD	4	June 23, 2022 07:53
Openfoam Turbulent Car Simulation	meganrw	OpenFOAM Running, Solving & CFD	0	July 5, 2021 16:50
Turbulent flow around a cylinder with pimpleFoam	Nazim	OpenFOAM Running, Solving & CFD	2	May 19, 2020 07:58
K-Omega-Epsilon BCs for suction inlet	CFDBro	OpenFOAM Running, Solving & CFD	0	March 27, 2018 19:28
y+ selection with epsilonWallFunction and omegaWallFunction	wc34071209	OpenFOAM Running, Solving & CFD	3	March 13, 2018 10:20