Boundary Layer Mesh Conundrum (CFX/General Query?)
 

Hello all,

I'm working in the area of centrifugal pumps and apply and preach to the best of my ability all I learned at university, including the importance of mesh independence studies, appropriate Y+ values for the turbulence model, etc

I have however, a colleague who is self taught & doesn't generally use any boundary layer mesh at all, less calculate Y+ values. He feels I put too much emphasis on these points (we're not very well resourced with processing power so leaving out the boundary layer mesh allows a convenient speed up), and uses the example of a radial impeller which he solved with & without boundary layer mesh, finding ~3% difference in torque and even smaller difference in generated head.

I have no idea how the solution changes when no BL mesh is used, will the solver (CFX in this case) apply the wall function to the gap whatever the gap size is?

Could anyone (PLEASE!) arm me with a good response, or should I accept that sometimes it's "horses for courses"?

Thanks!

Ianto.