[kaaja]

When I have a wall I typically use "noSlip" as boundary condition for the velocity U also when dealing with turbulent flow.
At the same time I use turbulence models with wallfunctions for the turbulence variables.
It does not seem like there are specific wallFunctions for the velocity.
Is the answer that in practice the velocity U is indirectly determined by wallfunctions through the turbulence variables (that is modelled with wallFunctions)?

[RobertHB]

To my knowledge, the typical expectation is that your velocity is Zero (0 0 0) at the wall. That is why you use a noSlip, or fixedValue (0 0 0) which does the same, boundary condition.

[kaaja]

Originally Posted by RobertHB

To my knowledge, the typical expectation is that your velocity is Zero (0 0 0) at the wall. That is why you use a noSlip, or fixedValue (0 0 0) which does the same, boundary condition.

Thanks for replying!
So one has no slip at the wall, but do you know what one have between the wall and the 1st cell for U when one use e.g. nutUSpalding for nut at the wall?

[piu58]

There is no wall function for U because the flow is the result of your calculation. If you use only p and U, then you don't have any additional model for turbulence. If you use such a model (e.g. k-epsilon) this model influences the flow. It does that in a different way in the near of walls.

The wall function is not a property of the flow but a property of the model of turbulence.

[kaaja]

Originally Posted by piu58

There is no wall function for U because the flow is the result of your calculation. If you use only p and U, then you don't have any additional model for turbulence. If you use such a model (e.g. k-epsilon) this model influences the flow. It does that in a different way in the near of walls.

The wall function is not a property of the flow but a property of the model of turbulence.

What you say seems to confirm my suggested answer, that U is indirectly affected by wall functions if one use turbulence models that apply wallfunctions.
Thanks!