Setup question: Is there a better way to do this?

[Torch]

A co-worker asked me to make him a new axle for his exercise machine. Basically nothing more than a round bar with circlip grooves at either end of the bearings and squares for the pedal cranks at either end. Somehow, he had managed to twist off one of the square ends.

The square ends were actually tapered, but I think I figured out the best way to deal with that. I mounted my RT on my tilt table like so:

I set the angle by chucking up the unbroken half of the old one and adjusting the angle of the table so it would indicate horizontal:

Taper was perfect, my square ended up square and it fit the crank quite nicely.

Then I ran into the bit I didn't fully consider when I agreed to attempt this project: The square on the other end has to align pretty much perfectly or the pedals won't be opposite each other. And the shaft is too short to access the other end through the chuck. Yes, it seems pretty obvious now...

What I did was to coat the end with some ink, set the milled square on a parallel on the surface plate and use the height gauge to lightly rub the ink off the top of the end to be milled (x4, just for spares and checks!) Then with the axle chucked up in the RT, I used the cross-slide to set the DTI at the peak height, then rotated the RT using the end of the DTI as a pointer, like so:

It seems to have worked:

I guess you can't argue with success but I still feel like my solution was a bit of a kludge, and there is probably a much better way of doing this that I have overlooked.

Suggestions?

[Conrad_R_Hoffman]

I'm not sure of a better way, but that looks darn good. I assume it's the standard taper used on most bikes. Just be glad he didn't want you to make the tapered hole to fit the axle!

[ken572]

Dwayne,

It looks like it turned out great to me.
You seem to have the setup rigid enough from
looking at your cut, so I would say Congrat's on
a job well done.
If your friend breaks the new one off, I think I would
pull a torsion bar out of an old chrysler product and
make the next one out of that. That stuff is machinable
and he won't break it no matter how he abuses it..

Ken.

[PixMan]

If there was enough space between the circlip groove and the end of the tapered square, it might have been stronger to just mill the tapered square by side milling. With the indexer flat on the table of the machine, but skewed to the angled required, you could mill with a nice radius in the corners for strength (no stress-rising sharp corners.)

[Rosco-P]

You could have milled a small flat across the length of the shaft to use as a datum mark for aligning the ends. You're using the tilt table under the dividing head to establish the taper angle? Then cutting the flats with the end of the endmill? Loss of rigidity right there. Mount the dividing head directly to the table, angle it slightly to establish the taper, side mill the flat. Or even simpler, use a 5C square collet block held at the correct angle in you vise. Same basic process.

[Bill_Cook]

Assuming that I'd thought ahead and caught it (not necessarily a good assumption), and building on PixMan's reply:

The shaft could have been clamped to the table and a flat reference area milled on each end to indicate off of. The ends could have been left a little longer and faced off later, or the reference flats could have been over but not into the tapered squares.

BC

[Torch]

Side milling would have necessitated a very small diameter end-mill, because of the required land between the circlip groove and the flats. I don't know that the tilt table made things worse than the flexibility of a small diameter tool.

A small flat across the full length of the shaft might have worked OK -- I didn't take this machine apart, but I believe the shaft rode in sealed bearings from what he said. Bill_Cook's modification would have worked great for this. Wish I'd thought of that at the time.

Thanks for the comments guys.