[madad2005]

I'm investigating the wall function approaches used in OpenFOAM. Epsilon and omega wall functions were fairly straight-forward in the end. However, I am looking at nutWallFunction and I'm at a loss as to why nut at the wall is defined as so:

nutw[faceI] = nuw[faceI]*(yPlus*kappa_/log(E_*yPlus) - 1.0); //for y+>11

Why not use the values the computed values of k, epsilon, or omega?

[olesen]

Originally Posted by madad2005

nutw[faceI] = nuw[faceI]*(yPlus*kappa_/log(E_*yPlus) - 1.0); //for y+>11

Why not use the values the computed values of k, epsilon, or omega?

I'm confused - do you mean not using any wall functions at all?

[madad2005]

Sorry to have confused you; I'll elaborate more. In the k-epsilon and k-omega models, the very last thing they do is compute nut:

//from RealizableKE
// Re-calculate viscosity
nut_ = rCmu(gradU, S2, magS)*sqr(k_)/epsilon_; // (1) nut = Cmu*k/omega
nut_.correctBoundaryConditions();

nut is computed as I'd expected here. Why not use the same equation in nutWallFunction to compute the value at the boundary, as you have k, omega, and epsilon already computed at the boundary face? Is it because you require a general expression for application to all turbulence models?

The reason I've asked is because I'm not particularly conversant in wall functions and I wasn't expecting a different equation for nutWallFunction.

[chb]

• At the walls
  nutw[faceI] = nuw[faceI]*(yPlus*kappa_/log(E_*yPlus) - 1.0);
  is used to apply the wall shear stress
  tau_wall = rho * (nu + nut) * du/dy
  according to the log law.

• For the rest of the flow field
  nut_ = Cmu*sqr(k_)/epsilon_;
  is used to model turbulence effects.

[madad2005]

Thanks, Christof.

"At the walls
nutw[faceI] = nuw[faceI]*(yPlus*kappa_/log(E_*yPlus) - 1.0);
is used to apply the wall shear stress
tau_wall = rho * (nu + nut) * du/dy
according to the log law. "

Ok, if that's the case, then this is a different form from what I've read other people doing in the past. Thank you, anyway, for helping clear that up.

[83_Ale_83]

Hi to everybody,
i was looking for the "origin" of this nut equation, It's from:

nut=nu*[(Y+/U+)-1] and U+=1/k*ln(E*Y+) ???

I don't know the origin of the first equation, I have read Pope-Turbulent Flows but in that book nut=k*y*u_t

Thank in advance
Alessandro

[83_Ale_83]

Originally Posted by 83_Ale_83

Hi to everybody,
i was looking for the "origin" of this nut equation, It's from:

nut=nu*[(Y+/U+)-1] and U+=1/k*ln(E*Y+) ???

I don't know the origin of the first equation, I have read Pope-Turbulent Flows but in that book nut=k*y*u_t

Thank in advance
Alessandro

Anyone ?
thanks

[83_Ale_83]

Originally Posted by 83_Ale_83

Anyone ?
thanks

Up again please

[Anne Lincke]

Hey Alessandro,

in this thread you might find help:

http://www.cfd-online.com/Forums/ope...-function.html

[Anne Lincke]

Originally Posted by madad2005

Thanks, Christof.

"At the walls
nutw[faceI] = nuw[faceI]*(yPlus*kappa_/log(E_*yPlus) - 1.0);
is used to apply the wall shear stress
tau_wall = rho * (nu + nut) * du/dy
according to the log law. "

Ok, if that's the case, then this is a different form from what I've read other people doing in the past. Thank you, anyway, for helping clear that up.

Dear madad,

can you please explain how the nut computed by the nutWallFunction applies the wall hear stress according to the log law?
Am I right that there is no explicit near wall treatment for the velocity and the wall function is only enabled via omega, k and nut?

Thank you for an answer in advance.
Anne