Boundary Layer roughness/ low reynolds wall treatment
 

Is my topic title an oxymoron? Does anyone have experience with wall roughness while looking at the boundary layer behavior?

I'm modeling the boundary layer along a wall of a wind tunnel. The boundary layer is what I'm interested in, so I'm using the two layer all y+ wall treatment that is an option with Reynolds Stress models and my mesh is resolved below y+=1.0 In the lab, my wind tunnel entrance isn't long enough to get the thick boundary layer that I want, so I use a boundary layer trip of sand particles upstream from where I'm looking at the boundary layer behavior. Can I model this with a roughness height, or will it not mean anything since I'm resolving my first wall cell and the rougness parameter is greater than the y+ of this first cell?

If just specifying wall roughness for the region doesn't work, any feedback on my ideas?
1. Use a different mesh where the trip is located. Being an all y+ wall treatment, if I use a first wall cell thickness that corresponds to a high y+ treatment, shouldn't the roughness factor be compatible with this?

2. For the fluid around the boundary layer trip surface, separate it into a different region, then I could assign a different physics continua to it (I guess if I still have a small mesh there, it still wouldn't be legit).

3. Model the boundary layer trip explicitly.

Any feedback would be great. Just seeing if there is anyone out there that has modeling something similarly. Thanks!

One solution I have seen for dealing with trip strips (I think this is what you're suggested in option 2), is to use two regions:

- A region upstream of the trip strip which uses a laminar formulation

- A region downstream of the trip strip using a turbulent formulation

see the following paper for an example

Kim, S.-E., S. H. Rhee, et al. (2003). Application of Modern Turbulence Models to Vortical Flow Around a 6:1 Prolate Spheroid at Incidence. 41st Aerospace Sciences Meeting and Exhibit. Reno, NV, AIAA.