[tongx051]

Hello guys,

I have several questions regarding mesh resolution for turbulence modeling.

1. What’s the mesh resolution required to solve a turbulence boundary layer with RANS?
2. What’s the mesh resolution required to solve a turbulence boundary layer with LES?
3. How does result look like if the mesh resolution is not enough(in Fluent)?
4. If wall function is applied, what the required resolution then? How many grids do I need inside the viscous sublayer and buffer layer? What grid growth rate is recommended?
5. Is the wall function for LES different from the standard wall function for RANS?
6. In LES simulation, what happens if no wall function is applied and near-wall mesh is not fine enough?
7. For modeling flow over several bluff body, there are typically flow separation, adverse pressure gradients, and wake interaction. Is wall function still valid in this case?
8. What’s recommended the mesh resolution for the free shear region (far from the wall) with LES and RANS?

I appreciate your help!

Jeremy

[jdrch]

1) That depends on the physical scale of your model, but 5 prismatic layers is a good starting point
3) You might not even get a solution If you do, it may be physically wrong (obvious from inspection) or inaccurate
8) That depends on which region of the flow you're most interested in. If you're interested in the boundary layer, than it's OK to have a coarse mesh far from the wall