Struggles with holding size on shaft work.

I've been in this business for 26 yrs. Seven years manual and last 19 years on CNC lathes.
I always seem to have a problem holding size when using a centre.
The job I'm currently doing is 390mm long, 40 mm diameter and have to turn back 295mm
There is a journal at the tailstock end with a +/- .01 mm tolerance.

I just can't get any consistency on the diameter of this journal.

When doing chucking work the machine will hold this tolerance all day long but when working with the centre It can vary +/- .02
I'm having to come slowly down to size on each one to hit the tolerance. I've got 70 of these to do.

This is a regular job, so I'd like to bottom the problem.

I've had lots of theories about this, the current one is that the stock is stress relieving as it is machined down and is putting a side load on the centre causing the effective centre line of the machine to move slightly?

The bearings on the machine are built into the quill and were recently changed. This made no difference to the accuracy. Still had the same problem.

I've not had much experience with using centres to be honest, most of our work is chucking style.
I'll also say that everyone else has the same problem but I'm lumbered with it this time.

Be grateful if anyone can share their experience.

Cheers

Dave

Think of all the different places that an error can be introduced when using a tail stock,
and start there.

If you don't move the tailstock back, in other words, only use the quill in between
parts is it more consistent? That answer will give you a direction to head.

On the sneaking up on it thing. You just need to get yourself a stable process.
One test cut, measure, offset, and let her rip. None of this "slowly coming down
to size" stuff. One test cut, one measure, one offset, and off she goes.

When I say test cut, I mean a cut that mimics your finish cut almost exactly. If
your finish is going to take .030", then take .030 on your test cut. keep it consistent.

I do it a lot on tight tolerance stuff,
especially when the tail stock is involved.

+1 to don't move the tailstock. I made an OD holder that holds a centerdrill so I can sneak between the shaft and center to drill.

If you need to move the tailstock to load and unload I don't have any ideas.
Maybe stick more of the shaft out so you can load/unload?

Thanks guys, all good advice.
I'm currently centring with the a modified od tool holder as Booze suggested.
I don't move the tailstock at all once set. Just the quill.
What I'm doing at the moment is
roughing to plus .2mm
Semi finish to plus 0.1mm
M0
measure.
My theory was that I could use this dimension to decide if I need to alter the offset for the finish cut.
But it's just not working like that.
eg semi finish = 0.1 big then I ought to be able to just let the machine continue.
Sometimes it's ok, but then the next time it's either bigger or smaller than it should be.
This is without even retracting the quill.
Maybe I've got too cautious and the finish cut is too small.
Problem is if it's too small it's scrap.
Ahh the joys of machining.
Anyway, I'm off this job for a couple of days. (Ripping down some big lumps of duplex tol +/- 0.04", no stress)
I'll get back to it next week. I'll have a word with the boss, see if we can have a bit of time to go through this methodically.
Thanks again.

Dave

Maybe I've got too cautious and the finish cut is too small.

Click to expand...

Small finish cuts does not equal more accurate. I've fallen into that
trap more than once myself.

Depending on your insert, its possible that your tiny finish cut is your
source of error.

You don't say what matl but that's an awful light cut. It's probably sometimes rubbing, sometimes cutting. You have to get that nose radius into the matl. (what radius are you using?)

I'd leave at least .38mm for finish cut. Also, don't let the machine sit. My experience with CNC lathes is they move when they cool off or warm up. Usually about .02

For close grinding, I would figure to get double a live center error on the part.

Deads are often the best but with overuse, they can wear a part center.

Honing a center drilled center was often a must-do for close work.

The quill may be loose in the tailstock.Stick a tenths indicator to the tailstock and check the quill for movement when pushing and pulling on the part (gently).

Ed.

I'd be looking at the finished center to see if its really a true center and has a deep enough pilot hole.
if the pilot hole isnt deep enough the post on the center will hit first before it even touchs he 60 angle part of the center.

When I load a part when using a center I chuck the part in the machine against the stop. pop in the live center reopen the chuck and close it quickly then reseat the live center. this way your not cocked in the machine.
also after you put the live center in stick a indicator on the center and seat what it runs out. try moving the part with you hand to see if theres any play.

Sorry for getting back so late on this guys. The job got put on the back burner as some urgent stuff came in.
So I managed to get some time troubleshooting and I've got rid of 90% of the errors.
Checked the quill for stability, fine. Left the T/S in position once set.
Basically I was doing just about everything wrong.
1. After the bar was centered, 1st op. The center hole was running out! Could actually see the bar move when the center went in. Too aggresive on centering. Fixed that by slowing the feed down.
2. After roughing the center was out again. The material stress relieving I guess. So carefully re-centered.
Feel like a complete novice but very little experience with tailstock work :-P But at least the job is a lot more straightforward now.
Thanks for all your help!!

Check tailstock pressure. In many cases, I've seen the hydraulic pressure dialed up to the moon and the shaft is actually "bowing out" because of it. I keep mine around 100 - 120 PSI.

One other thing to try-put a slight dwell on the center drill to make sure you get an accurate surface.

+- .0004” tolerance over 15” length. Yeah, you’re going to have to slowly get it to size.
That tight of tolerance is never easy to hit every time for 70 parts.

I think the issue is caused by runout of the part due to inaccurate chucking, and/or stock with a bend in it. You'd literally have to drill the center and never move the part to get a true running center. This is why they used to turn stuff between centers (which I don't do).

What I've done for this is to customize a collet for my chuck: since I was working in the range of 5C, I took a soft collet and machined relief behind the front lip, so the collet basically grabs on a surface about 1/4 inch long. Then the bar is put in, basically use the tailcenter to get it close, then clamp the collet. Then indicate the center hole and push and pull the bar until it runs true (in mid air), Install the tailstock center and machine the part.

If the part is beyond collet capacity, you can make a collet like substitute to clamp in your 3 jaw power chuck. Again, relieve it so that only the front 1/4" does the gripping. You can thread this surface to improve grip if you like. But you have to dial in the center hole in free air before inserting the tail center. I know, it's crazy that the tail center can't pull that shaft over, but the part will precess around and this changes the effective cut diameter by that pesky fraction that drives you nuts.

I never clamp the tailstock quill as a matter of habit. I don't think it helps fix any issue. I find that it is more important to be able to test the position and pressure of the tail center in the part, in case the part moves slightly, or heats or cools. I have the habit of always having the handwheel handle just past the 12 o'clock position so it will not fall back the wrong way. I basically monitor the tailstock handwheel for the duration of the cut. I find it is pretty common (with a smooth hard jaw) for the bar to push back slightly during a rough cut, so that needs monitoring. Or, the tailstock may slip back a bit too, until the clamping effect between the clamp, the bed, and the tension bolt gets really set.

DLodge said:

1. After the bar was centered, 1st op. The center hole was running out! Could actually see the bar move when the center went in. Too aggresive on centering. Fixed that by slowing the feed down.

Click to expand...

Easy for this to happen either by drill wandering or material moving slightly after roughing. I did some shafts last year that were off somehow and needed to run very straight. Had to bore the center taper back on center with micro boring bar. Little tricky to get in there but worked great.

When I do tailstock work, I put the work in, clamp every thing down, rotate the part with power to the spindle, then stop and loosen both the chuck and tailstock just enough to relax the gripping points, then retighten.

Tom

Remember he is on a CNC lathe- no quill clamp, n handwheel etc. And then again, sometimes its just easier to admit you need to grind the darn things as a 2nd OP.