[lcbarreto]

Hi guys,

I am simulating the flow over an aircraft using the S-A turbulence model and I came across a doubt about the Y+. I have seen in a lot of places saying that for tubulence models without wall treatment, my y plus should be approximately 1 on the wall, but is there any problem for it to be less than 1? My wall plus is ranging from 4e-2 to 7e-1, would that be ok?

Thanks in advance

[sbaffini]

Lower than 1 is even better. You should also have a certain number of points (5-10 or more) within the viscous layer.

Note that below y+ = 5 all fields have a linear behaviour, so going too much below y+=1 is actually an overkill (unless you are doing some grid convergence check).

In general, however, SA also allows y+ values higher than that, as its main variable has a linear behaviour also in log region. However, for the other variables you would still be using a wf.

[lcbarreto]

Originally Posted by sbaffini

Lower than 1 is even better. You should also have a certain number of points (5-10 or more) within the viscous layer.

Note that below y+ = 5 all fields have a linear behaviour, so going too much below y+=1 is actually an overkill (unless you are doing some grid convergence check).

In general, however, SA also allows y+ values higher than that, as its main variable has a linear behaviour also in log region. However, for the other variables you would still be using a wf.

Thank you so much for the reply.

I am doing a mesh convergence study and I am using CD and CL of the aircraft to test it. Both my CD due to viscous and pressure effects are converged, would that mean that my Boundary layer is well represented? I am using a first cell hight of 5e-6 m and 45 points with a 1.2 growth rate. I have used 45 because it gave a smooth transition from the prisms to my tet mesh.

[sbaffini]

Originally Posted by lcbarreto

Thank you so much for the reply.

I am doing a mesh convergence study and I am using CD and CL of the aircraft to test it. Both my CD due to viscous and pressure effects are converged, would that mean that my Boundary layer is well represented? I am using a first cell hight of 5e-6 m and 45 points with a 1.2 growth rate. I have used 45 because it gave a smooth transition from the prisms to my tet mesh.

Yes, if convergence is achieved in those values and your y+ is clearly within the viscous layer (as for your case) than it means that you have a well resolved boundary layer.

The minimum suitable grid for your case is just the coarsest one that gave you the final CD value.

Obviously, in theory, this does not means that there might not be other quantities which would still vary if you refine the grid. But if your interest is in CD, than you are ok.

[lcbarreto]

Originally Posted by sbaffini

Yes, if convergence is achieved in those values and your y+ is clearly within the viscous layer (as for your case) than it means that you have a well resolved boundary layer.

The minimum suitable grid for your case is just the coarsest one that gave you the final CD value.

Obviously, in theory, this does not means that there might not be other quantities which would still vary if you refine the grid. But if your interest is in CD, than you are ok.

Again, thanks for the reply.

I am also checking for continuity, momentum and turbulence (nut) convergence, the first 2 are below 1e-8 and the latter below 1e-6 and my CL has converged in the simulation as well as with respect to the mesh.

Hopefully I am getting the right results.