Aldo is completely correct on both points.
In particular, most bar's bends create twisting forces going into the clamp, and relatively long, slender glued or pressed joints do not fare well in torsion. These junctions fail starting at one end, and the failure progresses along the length to complete failure, due to the twisting that occurs along the length of tubing. I wouldn't count on the outer sleeve twisting along with the inner bar, since the stress on the adhesive will be localized at the outermost points of the bond (left and right in this case). Nor would I assume that any adhesive won't leave large areas un-bonded where air can become trapped, surrounded by glue.
David Snyder, BSME theorizing in Auburn, CA, usa
Aldo Ross wrote:
> I've tried drilling & pinning (one pin 1/8" dia) through the
> sleeve and bar, but the aluminum (non-heat-treated Cinelli
> bars) is so soft that the slippage continued, while the pin
> just gouged a slot through the bar and the sleeve.
>
> If you try glue or epoxy, remember that under load it could
> shear (break loose) at any time.