Thread: Roller walking out of lathe chuck

A few weeks ago we had to fix a hollow roller that had the shaft bores worn out. We welded up the id and had to rebore both ends. The shaft mount is welded in about 1 1/2 from the edge and only about 1/4 tube thickness. 6 foot long and O.D. is worn a little. The chuck jaws wouldn't reach the solid part and with the steady rest it kept walking out.
I remembered some work holding tips from Forest (pretty sure). One of them was using wire in the chuck jaw grooves to increase the grip. Worked slick. Thanks Forest!
I have trouble with names but something like this I retain.

Dave

A few years ago I had the joy/frustration of manufacturing large hydraulic cylinders, 10" diameter, 96"long- bored from each end close to size- then finish honed. This involved extensive use of a steady rest. A constant problem was the tube walking out of the jaws- I remembered a solution an old gentleman told me about several years prior-- I bent a piece of rod 5/16" diameter in a 90 degree bend (that fits the clamp out of a Starrett 196 indicator set) - the rod was clamped to the OD of the cylinder using a chain vice-grip, a 196 Starrett indicator is set to the face of the chuck and axial runout is then adjusted to zero using the steady rest. There is a hidden element - sag - from the weight of the indicator- that has to be compensated to reach true zero axial runout. When I took the time to use this method, the walking stopped completely and distortion of the cylinder from clamping too tight went away - that was a problem also...

Chuck jaws are slippery. Unless you have them mashed down enough to flow metal into a serration you have nothing but friction to prevent axial mpovement. If you drill from the solid it's likely the work will push in until it contacts the chuck body. If your steady isn't adjusted to hold the work axis in strict alignment with the spindle axis the work will drift in the direction of the splay and/or the axial component of the feed force.

You need a raised rib to engage a jaw serration or a wire to catch a groove in the work, or a work shoulder to prevent creep. Or a death grip. Shutinlead mentions an indicator stop to display the instant of movement. Works good if hazard to the unstrument is nil and RPM is slow enough to sight the dial as it whizzes past.

Seen the "walk out" from two other main reasons. Chuck not gripping as tightly as you think, because of lube neglect. The other being (of course) one roller in too far pushing the part off center.

Learn to set up the steady correctly and your problems will end.
That is the classic result of a steady not on centre.

you do not mention rpm.
.
if steady rest is too far out of alignment AND you are going fast you will drift fast, for many things you might want to run more like 100 rpm and not a 1000 rpm. even if you have carbide tooling which you probably should not bother with anyway. HSS still works for most things.
.
don't skip the steady rest setup alignment. saving a few minutes in alignment will cause 10x the walking out of chuck problem

I agree with Mark 100%! If you don't have a set-tru 3 jaw use a 4 jaw and get it dead nuts. Have your steady mounted and dial in the supports until you get a zero reading both ways. It won't walk out. Here's a simple device you can make for truing in the bar. Simply slide it from the chuck to the steady 'til dials read zero. Steady rest not shown for clarity.

The centering of the steady is very important also the direction that the steady rest land is cut.

The land is afterall a thread just real fine.

An indicater between the chuck and the workpiece will give you the true chucking deflection.

I use slips of paper that does NOT contain clay to get a death grip on slippery metal. You could use brown paper bag for example. Put a slip over each chuck jaw. This works perfectly in the milling machine's smooth jaw chucks on materials like brass,which can move even if clamped tightly.

I don't doubt the steady wasn't correct, but it was as close as I could get it considering the roller was well worn and not even straight.
Normally I would have turned a set up plug to set the rest up Didn't have a chunk even close to the rolller size. Part of the problem was also the hollow section compressing. Time was also an issue with the machine down.
Speed wasn't the problem, I was running slow anticipating this and chatter problems.
Ray, I like your indicator mount.


Thanks for the info guys
Dave

I agree with several above that true tube likely was not "true with the world" here. If you consider how a U-joint works at say a 10 degree angle you will see why a chuck has no chance here. In fact things can get bad enough that even if the part doesn't walk out - a thread cut which measures "good" won't accept a gage (due to velocity change every 90 or 120 degrees depending on the chuck type). The part changes rotation velocity while the tool travel does not... It can also produce "lobed" surfaces also (you won't see any problems with a mic, but the mating part won't fit well).

There are a number of solutions out there, the one I use most is demonstrated in the link below, you do have to be aware of indicator sag and adjust for it.

Indicating_Lathe_Steadyrest.wmv - YouTube


If you are brave, here is the part being centered in the steady with just a dog & center at the faceplate in the link below;

Lathe_Steady_Rest_Center_Cutting.wmv - YouTube

Good luck
Matt

Matt, I spent some time watching some of those videos and they are full of knowledge. Thanks for the link.

Great video.