Lathe Adjustments for Better Runout

What is the wall thickness?

Are you holding them in jaws? if so, what type if jaws? (hard jaws, soft jaws, pi jaws?)

You could try releasing jaw presssure before the last cut and taking a minimum amount of stock.

What are the feeds and speeds you are running?

What type of insert?

I would suggest c3,c4 carbide, positive rake angle (7-11 deg.)

Start by using soft jaws that you have cut in situ, for the job! after that, need more info!

If you still have problems with using jaws, turn a mandrel so you locate on the inside dia and pressure drive it off the tailstock.

All these are good ideas. Let me see if I can sum it up for you so your not just doing a trial and error.

If you have a thin wall, you need to use pie jaws to keep from distorting the part as much as possible.

I know it shouldn't need to be said, but I'll say it, make sure you use the machine to re-cut the jaws every time you set the job up. You cannot trust the chuck to repeat that closely.

That means you cannot expect hard jaws to work.

This is independent of whether your cutting and ID or an OD, you need to follow this 'guidlines'...

good luck and stop back and let us know of your success.

What is the indicated runout of your chuck or workholding on the spindle?

What is the length / depth of the ring out from the chuck or faceplate?

What is there deflection of the tool post?

milo

Your question goes much farther than workholding.... there are many other factors to consider.... what kind /brand / style of lathe are you using? how old is it? is it a manual ? cnc ? slant bed ? VTL ? if a manual are you running your tools from a compound slide? what kind of chuck ,what size and how old? is it a manual chuck or hyd? is the runout consistent across the face and length or is it a taper? if so which way?

I run a lathe nearly every day, cutting rings in the 6" diameter range. I do not need nearly the precision that you are calling out for, however, my experience is that in the range of 6", the variation in diameter will exceed .001" + or -, simply due to the temperature of the ring. Perhaps this accuracy is needed because you are press-fitting parts together? In any case, temperature variation in aluminum causes a considerable amount of dimension change, and must be considered. If the density of the ring varies, then expansion will vary throughout the piece. As with aluminum pistons in an automobile, the diameter is "cam ground". That is, the final dimension at operational temperatures is projected at the thickest mass, and is cut thinner at that point to allow for expansion.I run a tool-post grinder when needing that precision.

Oh Ken... for the first time I can remember, I have to disagree with you.

Your talking about a lot of different things. If the spindle bearings are bad, then that would effect runout, I'll have to agree.

But now, the test bar and the bad ways would show up in taper, not runout. On an old manual lathe, you would need to have the ways checked and corrected, but with new CNC lathes, you can go into the control and place in 'fudge factors' in the control to adjust for ways that are out of position. (to a point.)

Your comments on the chucking problem... you are right on!

Have a nice day

Actually, I don't think you're disagreeing with what I thought I was writing. My post rambled a bit.

I jumped from talking about runout to the idea of just generally getting the lathe set up right before looking for the particular runout cause. So the point re the test bar is to make one, and get the lathe spot on first, while looking for runout in the test bar. Then with the lathe able to make a near-perfect test bar, look for problems unique to making the 6" ring (if they remain).

Significant runout is pretty hard to achieve on a lathe, even if you try. My lathe is as poorly maintained as anyone else's, and I typically indicate with a metric indicator, so one space between lines is .01mm (.0004"). Ordinarily, the needle looks likes it's stuck -- I often have to check to see that it not fully extended or something. The width of a line (not the space between lines) is probably equivalent to .00004", and I can usually expect the needle to stay within a line's width.

Of course, if the piece gets removed from the chuck, then all this is thrown out. (But if it's a ring, then getting TIR on it outside of a chuck would be tricky, unless there is a test mandrel to mount it upon.)

There are cases where runout can be significant while the piece is still in the chuck. For example, if the piece is thin enough to be distorted by jaw pressure, TIR will normally be near zero immediately after machining. But as the part cools (still in the chuck) jaw pressure changes as the piece shrinks, and runout appears in the cooled part. If that is the case, and if runout is out-of-tolerance on the cooled piece chucked with light jaw pressure, then using a mandrel, or a six jaw chuck, or even clamping with end pressure against a face plate (instead of using a 3 or 4 jaw chuck) can help.

More rambling: I recently wanted to press fit four automotive ball joints into holes in steel plate. The ball joints were intended for this type of installation, and were actually made in Germany (not just German branded). Imagine my surprise to find no two the same ("diameters" -- which I put in quotes because they weren't round -- varying by 008", and roundness varying by .003") These were a little less than 2" diameter, so .002" can make the difference between a tight sliding fit and a very heavy press fit. To get even remotely close to the right fit, I had to make each hole different -- and even then was not sure the things would fit right (perish the thought I should ever have to replace a ball joint!) I wondered how these pieces could be so different given even the slightest thought to reasonable order of operations. For these to work in subsequent assembly, they'd have to rely on actually shearing off some of the metal of a significantly undersized hole. (Or maybe they made the holes random too (!!!) and selectively fit them -- setting back the idea of interchangeable parts by about a century.)

I cannot agree more... you do Ramble

Well put!

I try to operate on the principal that if I simply type anything, and make it long enough, eventually something that makes sense will come out. It's the "Give a monkey a typewriter and enough time, and he'll type the complete works of Shakespeare" principal.

As Mark Twain wrote, "Sorry this letter is so long. If I had more time, I could have made it short."

Bore. Turn OD. Part to length. Bore. Turn OD. Part to length.Bore. Turn OD. Part to length. Bore. Turn OD. Part to length.Bore. Turn OD. Part to length. Bore. Turn OD. Part to length.

Or if you like.

Turn OD. Bore. Part to length. Turn OD. Bore. Part to length. Turn OD. Bore. Part to length. Turn OD. Bore. Part to length.

High speed tool steel will work just fine if it's all you got.

I'm with you on that to a degree! But if you start with a blank of oversized Ally then things like work holding are fundamental! Most important, Repeatability. If the machine cannot keep within a said tolerance to start with then maybe you need to check the machine itself. If normally this machine can hold tighter tolerances, then it is most likely a work holding problem! Narrow down your options, use SPC, make a couple of histograms and SD curves of other parts made by the machine to give you an idea of the health of your machine!

Yes. Potential causes of discrepant or undesirable results need be eliminated from front to back (metaphoric front to back). The fact is that this is an exceedingly simple operation that should pose little frustration.

We used to make titanium rings @ 3/16 ht x 085" th x 8' OD with opposing angles to a theoretical point. after much customer dialouge it was agreed that we would first article parts on machine prior to final op (parting to length) as the deformation was excessive once removed. There are numerous other examples as this is a normal issue on thin-wall parts.

If indicator shows acceptable run out at chuck then there is a chance of a similar situation for his parts.

But again yes I agree, and much good advice has been offered.

CR3