Carbon Fiber Composite Does Not Tap

I should probably have investigated this before buying carbide drills, reamers, and taps, and 6-32 screws.  (The screws are cheap and likely useable for other projects. The drills, reamer, and taps were not cheap but work very well on steel.)

The tapping process was immediately suspect; while turning in the tap, it seemed as though it was just cutting through the material.  Indeed, it was.  The 6-32 screws just slide through.  I also verified that the 6-32 tap had a plausible major diameter: .132" which is just slightly smaller than the major diameter of 6-32 screws.

The right way to do this is with epoxy.  This means, of course, doing it right first time; there is no opportunity to remove screw and tap a new hole since even the first one will not be useful.

The good news is that epoxy sticks to carbon fiber composite and steel.

I also managed to write gCode to cut an exact circle for the 5/8" needle bearings.

And yes it is a press fit.  I suspect the epoxy will sort of slide out as I push the bearing into the hole.  Then apply a bit more epoxy on either side of the bearing/composite interface.  (Probably a toothpick application to avoid getting any on the bearing races themselves.

The bearing blocks will need to be glued into the housings. The gaps between bearing holders and housing are hundredths of an inch. Epoxy should work fine.

I had thought of machining the setting circles of carbon fiber composite but attaching them to the axes would be a problem.  The declination circle has a fixed position.  That can be glued in position.  The RA circle needs to be adjustable, usually by a set screw holding a low friction cylinder (usually Teflon) against the polar axis.  Set screw obviously not practical.  The setting circles will need to be aluminum or acetal.  Both of these tap well and likely have little or no friction to carbon fiber.  A set screw should hold them down just fine.

One of you pointed me to a company that makes precision adhesive scales which I might attach to an aluminum circle which would held in place on the axes by a set screw on Teflon.  This sounds more and more attractive if they make scales in degrees and RA.  (Not likely, but if not, I can upgrade my rotating table to the CNC version and write a program to cut lines and numbers.)