Sorry about the lack of info As i don't have a very good camera and it's awkward trying to get the pictures off it, but i will put some on here a bit later to show you what i was on about.

It's a spindle for a horiz-mill machine and part of it is the arbor for the 7mm bore cycloid pinion cutter, it needs to be concentric in order to make a decent cut with no chattering.

I could've bought some more material and made another but i'm not a fan of wasted effort so i like to save things if i think it can be rescued. So the 4.5" length of silver steel is completely turned ok, this length is what the machine itself is holding and the last little bit is the cutter arbor.

I managed to turn everything to the right dia. on the 1/2 inch bar, however, i messed up on the last length which was critical because it needs to sit on the center line of the milling axis, if i dont get the length right, it'll be cutting pinions askew to one side. So thats why i cut the end off basically.

Soo, i carried out what i said i'd do, the two dia. screwed and locked together and turned the end in the 3 jaw, held by the main length and tested it with a dial indicator and it barely moves to 0.01mm, i'm happy with this, and because i cut a relief at the end of the thread, the jointing piece fits right up to the shoulder, i turned the two together and now, you'd barely realise that they were once separated at all. I will call this a success!

hooray! Once it's all pieced together i'll put pictures on here.

Michael W

Thanks ray, it's going ok so far,

This is a picture of the machine at work:

It might not look it but to my surprise that 3 jaw is very accurate, it only has an inherent run out of 0.01-0.02 (when a bar of known size is used within 50mm or so) i remember using an old 3 jaw at work that was 0.2mm out of round

And this is a picture of the fused bar in the lathe, with loctite on the internal M6 thread (i personally prefer a high strength anaerobic compound called bondloc, the specific grade is for fusing pulleys to shafts, much thicker and gloopier, i'd definitely recommend it),

the maximum runout of the fused bar is shown on the plunger dial indicator when held in it's mating spindle cartridge, i can't get more than 0.02 out of it. As you can see, no mating line is visible after turning, it's like it was always 1 piece!

sorry i couldn't get the picture the right way up.

If all else fails for runout you can always superglue cigarette papers to the offending O/D, it's a black art though and certainly not kosher, i now prefer to call them "diameter shims" to legitimize it's proper engineering usage. If you could hear me i'd be telling you in hushed tones 

Michael W

Edited By Michael Walters on 30/04/2016 21:14:04