Another issue. I tried to do this with some different geometry. I used the same settings, except the analysis is no longer axisymmetric but rather 2D cartesian with multiple bodies. Instead of the high velocity flow at the wall I am now seeing very low velocity flow at the wall, almost as if there were a boundary forming. I have attached my share file, if you have a second to look it over I would appreciate it.
Here are the XY coordinates for data extraction. FYI: the cp's I gave in the spreadsheet were calculated from the velocity magnitude (rather than the calculated cp). cp= 1- (v/V)^2 and V=1 m/s so cp = 1 -Vmag^2.
Turn off the slip boundary conditions for the potential flow simulation. That is degrading the results. Go back and try your other problem again after removing the slip boundary conditions and see if that helps with the comarison to your benchmark.
I would also consider changing the edge mesh around the profiles to 3600 elements with no biasing. Since we have a 2D model that solves in 1 iteration, we might as well increase the mesh around the profiles.