Tailstock Offset Issue

All,

I did put this in my "Tests" thread in Workshp Techniques, but I think that thread ran its course. Apologies if folks picked it up there, but there's actually no advice left!

Anyway, this is regarding my ML7 that I've just done a light re-furb on...I'm getting frustrated with this whole thing now. Yesterday, after hours of work, I got the tailstock offset such that I couldn't detect any difference in turned diameter on two c. 1" bosses spaced 5" apart - turned between centres. Measured with a 0.0001" micrometer.

I also re-checked with the Myford test bar between centres - fine.

Just now, I decided to do a quick check: I removed the test bar, and both centres.

Then replaced the centres and turning test bar in random rotational orientations and took a 0.001" cut. Result: a difference of 0.002" in diameter.

I checked the centre in the headstock, and it's true. Then substituted the tailstock centre, and there's about 0.0005" - 0.001" runout.

What are you supposed to do about a brand new dead centre that's out of true? Are you supposed to mark it such that it goes in the same orientation in the tailstock?

It just sems at the moment that it'll be impossible, day-to-day, to get consistent results out of this lathe when turning between centres. If it's not a perfect dead-centre that puts the offset out, then no doubt it would be tailstock ram position, or tailstock on bed position, or probably some other small detail.

It's so frustrating.

Any adive as ever is more than welcome! Thanks.

Is the centre with the error a hard or a soft centre? ( No I'm not talking about choccies!). I ask because soft centres, which look very similar to their hard equivalent, are intended to be trued up after mounting in the headstock. They're not intended to be used in the tailstock.

Just a thought.

Edited By Clive Brown 1 on 18/11/2020 10:54:07

Dunno about others, but I get to a point where I find it unprofitably time-consuming to pursue each variation, although I'd probably think .002" over 5" worth investigating if I was expecting soon to do a job where it might matter. But I'd most likely never reach the root cause, and settle for whatever pragmatic solution resolved it for the job in hand.

I can imagine that you would have cleaned both inside tapers as clinically as you could before you did the setup where you're now finding discrepancies, but in the past I've found myself 'gaslighting' my own previous work. Bits of grit or swarf in or on any of the tapers could cause this kind of change, or any or all of the causes you mention.

The standard of your work we've seen on here makes me think you worry too much!

I made a dumbbell setting piece. The two bosses are machined to the same diameter as far as I can measure in a workshop with no climate control. This is used to set the tailstock with whatever tooling is going to be used so it does not matter if the centre is of unknown accuracy or if wear anywhere connected with the tailstock changes things. The extension of the tailstock barrel and how tight the clamp is can also make a difference. If the tailstock offset is adjusted is there any rotation of the tailstock about a vertical axis for example? I've never checked for it but I would expect it to be there. There are so many possible sources of variability in the tailstock it can result in you chasing your tail trying for repeatability.

Martin C

Don’t know how accurate my lathes are, never wasted time checking them, my opinion for what it’s worth is that you are wasting your time fussing around over a thou or two. Yes your centre may be a thou or so out, this is the problem nowadays with all this stuff being bashed out in China and India, mark it and always put it in the same way is the only realistic solution, also be aware that it is a lathe, not normally used in industry for chasing thous, or getting long lengths parallel, that is why most decent factories had a grinding shop, need a bit of small diameter bar about three or four inches long parallel, rough turn it to plus ten thou and bung it through the centreless grinder, job done. If the job has to be turned out will come the turners friend, a bit of emery tape, he will turn it as near as possible and polish the rest to within limits, note the within limits, not dead parallel.

If you consider the practical application of turning between centres the repeatability you need is removing the turned part from the centres and replacing it again accurately, sometimes end for end. The centre distance will be the same in each case, the centre will not have been removed from the tailstock or or the headstock, the tailstock will not have needed to be unclamped from the bed and the extension of the tailstock quill will be the same and it will be clamped so play in the quill is non existant.

First a comment. Between centres turning does is not required very often, neither is this level of accuracy so the occasional effort in setting up is a small overhead. Few items require turning parallel for the entire length but short bearing surfaces are often required on either end to be concentric and of equal diameter. This means any taper is over a short distance.

When settting up for between centres work the tailstock should be set by turning a test bar with centres the same distance apart as the workpiece(s). Once you are satisfied that the lathe turns parallel get on with the job. You can remove the work but not the centres and don't unclamp the tailstock. Everything should be very repeatable.

Don't lean on the lathe.

regards Martin

You can drive yourself crazy trying to make an old hobby grade machine tool accurate to the nth degree, or you can just get on with it and make stuff.

Many steam engines work better with a little extra clearance rather than ultra fine fits, as an example. Through lapping and honing and other hand processes very fine fits can be achieved in IC engines for example.

My own lathe and mill are very old and quite worn out but I still have a lot of fun with them and produce models and machines and tooling accessories that work fine.

If it is perfection you want in machine tools, you can have it, with brand new industrial machines. The cost of perfection is very high though. Have a look at Mazak or DMG Mori machine tools, or a gigantic old Loewe jig borer - they are as close to perfect as I have ever come across, in industry.

If the Tailstock centre is soft, and all the Morse tapers were clean and undamaged, did you trim the centre before using it? Since it runs out, possibly not.

Also was the Tailstock barrel locked or free?

Any slight clearance between barrel and Tailstock body might account for the the run out on the second test piece.

If conditions are not the same for each check, the results are likely to differ, even if both tests were carried out with the barrel unlocked, and with the same amount of extension,.since it could move between checks.

As said, in many instances, the mode of operation is not turning between centres. In 17 years of owning my lathe, I don't think that I have ever turned between centres.. Nor can I remember doing so with the previous ML7!

Mostly, the Tailstock is only used for support of longer work, or when drilling, tapping, or cutting external threads .with a Die.

If you are worried about keeping long work on centre, you can always use a steady!

A different Tailstock sees more use with the Rotary Table to provide support when gear cutting.

Howard

Thanks all.

I take the points about accuracy, but when the Myford test sheet states 0.0004" over a 12" test piece, that equates to 0.00017" over 5", it just made me think hmmm...

To answer the other points:

The headstock centre: Soft.

Tailstock centre hard: Hard.

Tapers were both cleaned up, and cleaned before installing the centres (yes, there's a thread bout that too!).

I probably do worry too much, but a double-check on something that's so tightly specified seemed the logical thing to do.

Thanks all.

What can be achieved with a brand new machine just off the assembly line may well differ from the same machine after 30 years of use, abuse, and wear.

Don't get too hung up on tenths of a thou, unless you are totally confident of your measuring equipment being consistently accurate and repeatable..

Howard

You can get that sort of accuracy with the Myford but you have to set it up specifically for your workpiece beforehand. You cannot just lob a centre in, clamp the tailstock any old how and expect it to be spot on.

Try this test. Mount a DTI in the tool post, extend the tailstock quill about an inch, rest the indicator on the side of the tailstock quill. Now unclamp the tailstock, twist and reclamp. Note the reading. Unclamp, twist the opposite way and reclamp and note second reading. This will give you the max error to expect when moving the tailstock.

That is one of tthe reasons I said you should set everything up and then not disturb the setup whilst working.

regards Martin

Edited By Martin Kyte on 18/11/2020 13:23:04

Edited By Martin Kyte on 18/11/2020 13:23:38

The test sheets are like MOT certificates, it was within limits when tested. After Myford test the lathe it goes to a retailer, not all machines were supplied direct from factory to user, the machine could be dumped on a concrete floor for a couple of months before being purchased and then possibly bolted onto a stand because that’s what the customer ordered and then in the back of a van to be delivered to your shed, being man handled into place. Will that machine still match the factory inspection certificate? I doubt it.

Dr

601 posts on chasing Myford accuracy is beginning to look a little obsessive

Just as well over half of that 601 was about making a very good job of a model

Dave, All - I understand what you guys must think - stop faffing just get on with building something!

I have done that earlier this year:

https://www.model-engineer.co.uk/forums/postings.asp?th=161327

...and to be fair a lot of my posts were in that thread, not just on lathe fettling and the basics of model engineering.

This is really just an exercise in improving my machine tools, before starting the next one - mainly based on what I found during my first build.

It's not like I'll bin the lathe; I know it's worn. It's just the way it is.

Ha ha thanks Jason - I missed that one.

If there's a resource available to ask a question (ie this forum) why not ask? If folks set up for between centres for every job, fair enough - I didn't know that.

Apparently some folks do, not one of them myself and I do turn between ctrs quite a bit for things like crankshafts, conrods, long tapers. etc.

Take a measurement after the first cut and if it's a little further out than you want give the tailstock a nudge one way or the other, expect there is a bit of wear between the inner bed faces allowing some twisting of the tailstock as well as barrel wear.

We are dealing with hobby machines, often at least secondhand, rather than tooroom machines.

ABSOLUTE accuracy is probably not needed. In real life, bearing clearances in high power, high speed engines have a tolerance band of 0.002" to 0.004" .

So, for us, if the shaft runs happily in the housing without excessive wobble, it is most likely to be fit for purpose.

A few years ago, after a rebuild, KIng George V ran a hot box on a special out of Paddington. The clearances had been made too small (And remember that GWR worked to much finer tolerances than most other railways )

Lapped fits are more appropriate for making very small displacement IC engines. We are not making fuel injection equipment for diesel engines.

Great precision can become a hobby in itself, but takes a lot of time. Depends on what is seen as the hobby, rebuilding a machine to as great, or greater precision than when it left the factory, or just making it able to produce the end product that you want .

Very often close attention to detail can change a bit of a sow's ear into a near silk purse.

You make your choice. (I make tools because of lacking both the skill and the patience to make good models )

Howard

I'd say that after it was levelled, and the other setup process gone through, then it would match the certificate. Unless it had been grossly mistreated, I can't see why it wouldn't.

Thanks Martin.

So I did as you suggested, and I also moved the tailstock to different positions on the bed, and extended the quill to various degrees, then checked the DTI at the same point on the quill. The maximum deviation from an ititial zero is - rather surprisingly - about 0.0005".

I have adjusted the gibs in the tailstock previously to give a firm but smooth sliding action, and it's prretty consistent along the working area of the bed.

I also tried twisting the very lightly clamped tailstock with the DTI against the quill, and could only get movement with fairly heavy load, and even then it wasn't much - like spring inthe assembly rather than a true clearance movement.

I don't know for sure, but I think it's probably the hard centre in the tailstock that's the biggest issue. Might try it with a rotating centre, but perhaps replacing it always in the same orientation is the answer. I'll try a few more tests then leave it the best I can and move on.