I am modelling a very simple converging/diverging nozzle - just a tube with a restriction in the middle.  I've specified the inlet static pressure (and ambient pressure) as 1 Torr, and I have a volumetric flow at the outlet of 70 L/s.  I should be achieving about Mach 3 at the exit. 

I ran the simulation first without the 'high Mach flow' option.  Solidworks gave me warnings and came up with a Mach number just over 1.  But it did converge.  When I selected the 'high Mach flow' option I received this warning: Maximum Mach number exceeded dV/V=267.036.  It also did not converge, but simply stopped when the number of iterations hit 3200. 

When I graphed the results, a contour plot of the Mach number over the cross section gave a maximum of 2.3 (pretty close to what I expected).  But the default scale went all the way up to Mach 15.  When I did a surface plot of Mach number on the inside face of the exit lid, the entire thing was red.  So that's where my high mach number is.

I'm wondering what the problem is with my model.  The lids were generated automatically by Solidworks, so I'm assuming they are ok.  Is it the way I defined the outlet boundary condition?  Is it that my ambient pressure is 1 Torr but the pressure inside the tube at the exit is in the mTorr range (and Solidworks is trying to expand the flow as it exits the tube?).  I already submitted a question about how ambient pressure interacts with the inlet and outlet boundaries, because that's confusing me also.  Any help would be greatly appreciated.  Thanks!

Michael