Mick,
I think I can explain how the type of clamped joint illustrated by John might fail to live up to expectations:
For the moment, ignore the little dowel-pin and consider just the clamping action.
Assume that:
- The crankpin is perfectly smooth, straight, and cylindrical.
- The hole in the con-rod has been bored to a close sliding fit on the pin.
You then tighten the cross-bolt, to apply clamping force; and it immediately goes very slightly oval … with the result that there is [at the microscopic level] only two-line contact.
Solution: Relieve those two areas of the bore very slightly [just a few thou'], and the contact lines will be redistributed. If you happen to get it right [which involves more luck than judgement, I fear] then you should get lines of contact at about 12, 4 and 8 o'clock, with the slot at 6. This is a much more secure and stable arrangement.
Note: In addition to it's obvious function; that little dowel-pin that John shows may well serve to "force" a contact line at 12 o'clock.
… I hope that all makes sense.
If it's any consolation; the problem described is, I believe, a fundamental failing of this design of clamp.
MichaelG.