Hi,

some time ago I posted a message asking if somebody knows where the rule of thumb of having one cell below a y+1 for two-equation TMs comes from. I haven't had any reply and I am surprised nobody knows. So I am asking the same question again to see if this time somebody who knows the answer (I don't!) will reply.

Thanks again.

Valerio

Hi,

I don't know about the rule that y+ has to be below 1. The rules I know are that for the High Reynolds k-e model with the wall function y+ should be around 30 for the near wall cell. The reason for this is that the near wall cell is supposed to be in the log layer (usually in the region of 30 < y+ < 300 - 500) for the wall function to work. The other y+ rule is for the Low Reynolds k-e model. Here the equations for k and e are solved through the boundary layer. The eddy dissipation at the near wall cell is determined with an algebraic function. For this one the y+ value at the near wall cell should be around 1. Again this is necessary for the near wall function to be applicable. I hope that helps a bit.

It is just a guideline rather than a strict relationship, and it may vary for different flows. It definitely varies between different turbulence models.

I'd say it was formulated based on grid convergence studies.

hi how are you

If You are using models that integratyes to the walls (Low Re k-Eps, k_Omega, Menter etc.) you then want small y+ values of the order of 1 or below. In models with wall functions, this is not the case because yuou approximate the flow profile at the walls!

The y+ of one value comes from an understanding of the velocity sub-layer in a viscous flow. The sub-layer corresponds to the region of the flow where the velocity is linear. Hence, having a y+ of roughly one puts your first cell in the linear sub-layer and your solution will be accurate.