I'm trying to run an anlysis on some shrink fit parts and I'm getting odd, seemingly conflicting results from SW.  I've got a shaft with a sleeve that starts at with .001" clearance off the shaft across the diameter.  Then I'm shrink fitting a much larger part onto the sleeve with a .0087" interference with the sleeve across the diameter.  I run the analysis and SW shows a large compression on the face of the shaft (X/radial stress in the -4000 psi range), but when I check the X/radial displacement, it says the bore of the sleeve has only collapsed .0005" across the diameter, which means it should not have made contact with the shaft yet.

So how are those compressive stresses created if not for contact between the sleeve and the shaft?  Very confused about this and can't exactly go into a design review saying "well half our results showed a failure to make contact with the shaft, so we're just going to ignore those results..."