Originally Posted by
oneclick
If the crankarm slides on, shim or no shim, there is no interference; hence a loose fit. On a straight spindle in a (too large) straight crank hole, tightening the bolt will not alter this.
I doubt it would be easy to make a shim strong and stiff enough at the correct thickness such that one could force it between the spindle and the bore; consider the pressure generated on the faces of a tapered connection. The end of the shim, on introduction, would be subject to the same pressure, while the free end behind it has nothing.
The only way it might work would be via differential thermal expansion; you'd be limited in what you could get from the crankarm (too high a temp for aluminium is quite low) so you probably would also want some co2 or liquid nitrogen for the spindle and shim.
I realize now that you were thinking that the shim(s) need to be forced in, as opposed to using an over-length shim(s) and leaving extra length hanging out of the hole from the side that the spindle gets pushed in (as I described in an earlier post, quoted below):
"An extra-long strip of shim stock along each of the internal flats of the crankarm could effect a compression-fit upon installation of the crankarm.
The extra length should be toward the inside face of the arm, so that the shim can't move away from the targeted areas of contact.
The shims can be trimmed away after installation, difficult as that might be(?)."
As described above, the shim is getting wedged in by the installation of the crankarm (assuming that the special Whitworth-threaded bolt is perhaps long enough to get things started).
Hopefully though, nobody here has to actually deal with any of these dreadful non-tapered crank interfaces on a bike that will be ridden much!