Jim Wilkins wrote:
On Sep 7, 4:53 pm, Louis Ohland wrote:
...
I'm not going to machine the Thompson shaft at all. Concept- drop the
shaft between a dead center in the head and a live center on the
tailstock and use it as a test bar. I was wondering about how much
deflection would be caused by the tailstock being tightened. I err on
the side of cross section thickness.

The normal straightness on Thompson shaft is .001 per foot, with .0005
being available. Well within the tolerance of the lathe that I'm using.

I suspect the deflection will be far less than you can measure with a
dial indicator unless the tailstock or its spindle shifts under
pressure (my lathe, for example).

Lets see: 0.785in^2 * 29E^6 = 22.7 million Lbs to stretch it 100%,
22,700 Lbs for 0.1%, 227 Lbs should shorten it by 0.001%. [Insert
Poisson's Ratio calculations here] If it expands anything close to
uniformly you shouldn't see the difference.
Bending with centered pressure isn't much of an issue until length/
diameter approaches 50.

It's easy to test, though. Leave the carriage stationary while you
tighten up the handwheel and see if the needle moves.

The bigger problem is drilling centered holes in the rod. That's why
the suggestion is to make the alignment bar, finishing both ends to
final diameter at the tailstock end without touching the crossfeed.

I align the tailstock with a cylinder square that I got cheap at a
machine-shop auction since no one else realized it was a precision
measuring tool and not just some shiny cutoff.

jw


I was going to ask the folks at Danaher Motion about doing the centers
at each end. Way more precise than I could do. If they can hold .001 I/m
a happy camper.

http://www.danahermotion.com/product...?parent_id=353