how to hold part on arbor?

I need to grind down the outside of an SK40/MT5 reduction shell to fit a short-MT5 taper socket. I've got a suitable SK40 taper mandrel to hold it, but am afraid that any "clamping force" to keep the shell on the mandrel will distort the mandrel away from true and result in the inside and outside no longer being concentric.

Is there a good way to glue it on? I could use a drop of cyanoacrylate ("super glue") in the SK40 socket but am not sure that's a good idea. If it lets go during grinding that's going to ruin the part and other bad things might happen. And if the grinding goes OK it's not clear that I can reverse the process to get the shell off the mandrel, except with heat which might then distort the part.

Loctite 680, it's strong stuff

Grinding on any taper even grinding a part that is set on a grinding taper mandrel the grind travel direction taking any stock is up-taper so travel is toward the direction that pushes tighter so heat expansion only makes a tighter fit. No infeed on the end that might loosen the fit to the ID taper.

I ran a part that I set a tube threaded on one end over the tapered mandrel, then set a nut on the tube that was turned tight to a stack of rubber washers and so then pushed tight to a locked collar to push/hold tight to the going tighter direction.

Yes, a between centers job with checking the set-on mandrel for near zero. the center can be restored to near zero using a center point hone.

1" Diam 80 Grit 60deg Included Angle 09193939 - MSC

michiganbuck said:

Grinding on any taper even grinding a part that is set on a grinding taper mandrel the grind travel direction taking any stock is up-taper so travel is toward the direction that pushes tighter so heat expansion only makes a tighter fit. No infeed on the end that might loosen the fit to the ID taper.

Click to expand...

Good point. In my case the larger taper diameter is towards the headstock. So I should infeed at the tailstock end.

I set a tube threaded on one end over the tapered mandrel, then set a nut on the tube that was turned tight to a stack of rubber washers and so then pushed tight to a locked collar to push/hold tight to the going tighter direction.

Click to expand...

Yeah I was thinking along similar lines if the glue is a bad idea. (Which I assume it is, since you didn't mention it).

The bit of mandrel poking out on the TS side will be about 20mm in diameter (a bit over 3/4"). This means that if I have any kind of hard clamping, that could distort the mandrel off true center. So the stack of rubber washers is important because it provides enough force to hold the part in place but not enough to bend the mandrel one way or the other, even if the part grows from heat. Is that the right idea?

Yes, a between centers job with checking the set-on mandrel for near zero. the center can be restored to near zero using a center point hone.

Click to expand...

I've got two, one 80 grit and one 150 grit, exactly for this!

To move a mandrel center just a tad you put the mandrel between a machine center with the center lap chucked at the other end, then with a heavy hand on top side of the mandrel rotate back and forth perhaps 10 -15 degrees.
I have never glued a part to a mandrel but am not saying someone may have a working method.
Very low temperature solder can be had and might be a method (??). Perhaps to solder just the end after the part is tight (??).

Low temperature alloys made by Indium Corporation

We used to use solder of near 100* F for weighting pine-wood derby cars, this so it would be safe for the boys.