ML10 - ready for the boneyard?

Regardless of Rollie's great-uncle and precision ground test bars and the like, the ultimate test is exactly what you have been doing: Turn a test piece and measure it for parallel.

All the rest is theoretical. The turning test gives you the real result of what a worn lathe will do under working loads and stresses with accumulative error from a chain of worn components.

The other methods are best used when reconditioning a machine that is stripped down and unable to be operated at the time. But upon final assembly, a turning test is normally done and the final adjustments are made. As your lathe is useable, i would go ahead as you are, turning test pieces.

A 1" diameter piece sticking out the chuck by 4 to 6 inches is ideal. To save a bit of time, rough turn the mid section - three inches out of four, or five inches out of six - to .050" or so smaller than the ends. Then you can just take fine cuts over the two end sections at fine feed to get your readings.

If you can get it down to .001" per six inches you will be doing exceptionally well on an old machine. And that is plenty close enough for most home shop work. (Factory tolerance was max of .0008" per six inch test piece.) A bit of emery paper rubbed over the spinning job will remove that in seconds if the job is that critical.

Edited By Hopper on 01/08/2018 10:15:05

Martin, I will this further advice into consideration. Thank you.

Hopper,

Also good advice. Time for me to place an order for some clean turning stock to run some (better) tests.

Martin, chuck runout has no bearing on the test piece once a cut has been taken over it. Even if the chuck is off centre and cock-eyed, the turned surface of the test piece will be in line with the lathe spindle. And if the lathe is correctly aligned, that surface will be parallel end-to-end.

If you take a piece of bar and mount it .050" off centre in the four jaw chuck, then take a .055" cut along it, the resulting cylinder will be in line with the spindle. Likewise if the bar (or chuck) is mounted at a slight angle.

For what it is worth I'm measuring 0.03mm ( = 0.0011 inch ) of runout on the back of the 3-jaw chuck.

Other tests will have to wait. Need to order some turning stick and perhaps a length or two of silver steel.

That runout is well within tolerance for a chuck body. No problem there at all.

One common problem with those old Burnerd chucks is bellmouthing of the jaws, which can lead to some tricky results, including poor finish and tapered turning and uneven turning as the job moves about while the cut is taken.

If you gently grip a piece of silver steel, or a good quality socket wrench etc, that is a good cylindrical shape, in the jaws, you can often see a gap at the front of the jaws if they are bellmouthed. With a longer piece of bar you may even be able to grip it gently then grab the end and move it about quite a bit.

There was an article in MEW about three issues ago on diagnosing and regrinding these old chucks.

Hopper,

Gave the chuck and spindle another look.

Found that there was a tiny bit of lateral play in the spindle (along bed axis) of 0.02mm ( = 0.0008 inch ). Careful tightening of the screw-collar at the gear end of the spindle enabled me to cut that amount in half. The spindle still turns freely.

The chuck does have some bell-mouthing, oddly enough, only on one jaw. That is showing some daylight and I can just slip a 0.05mm shim ( = 0.00197 inch ) into the split at the very front. This would suggest I should dress the jaws with a grinder held in the toolpost with the chuck jaws tensioned and the lathe turning at low speed.

I'm going to check up on that bell-mouthing again once I have some ground silver steel bar in hand.

Am I correct in thinking that I should first address any play in the spindle journals (by lapping the shims) before doing anything to the chuck?

We have already established the lathe is turning tapers, the challenge now is to identify the reason by a staged process of investigation and rectify.

Agreed, if you only consider radial runout of the chuck. Radial runout doesn't matter that much. However, you also need to consider the axial runout, and 1+ thou runout behind the chuck is small and probably not the main cause of error in this case, but if it was 2 or 3 thou, you might expect 4 to 6 thou angular deviation 100mm out from the point of measurement if that error was consistent through the chuck body (because it might only be the back face of the chuck/backplate that is out).

Try this experiment, take a small piece of straight thick wire, put a slight bend in it, imagine one side is the work piece in the 'bent chuck' and the other side the spindle, twirl the 'spindle' side between thumb and forefinger as if in the headstock, observe the circle described by the workpiece end and then imagine a tool cutting on an axis parallel to the spindle - result is a taper cut (as you may get with workpiece set poorly in a bellmouthed chuck).

However, in the two cases of distorted or a bellmouthed chuck, the effect will be the same, i.e., the end of the work piece is most likely to precess as per the bent wire experiment, which would lead to work piece being cut narrower at the tailstock end than at the headstock. From the measurements provided, I think I am right in saying the test piece is actually narrower at headstock end following the cut, by some 3.5thou over 3.5 cm, (sorry to mix systems!) turning a longer piece say 10 cm would mean a taper of approx. 10 thou over the diameter, which may be significant (or not at all, depending upon the intended use of the piece)

Assuming the work piece is well mounted, there are no bellmouthing issues and Ignatz wishes to remove the sources of error, there aren't too many options left to consider now, most likely being headstock alignment or possibly a twist to the bed or concatenation of both. Easily addressed once you diagnose the problem

The original question was 'is my 10 a boneyard candidate', the answer is clearly no, it is not. It is just a 50 year old lathe that has recorded all of the bad habits if its previous owners. The simplicity of the 10 is it will only need some minor adjustment once the issues have been fully identified. It will then be capable of sustained and accurate work

0.4 of a thou axial play is fine, leave the spindle bearings until after you have confirmed spindle/headstock alignment (as you confirm the finish is acceptable for now)

Bellmouthed chuck unless really severe, or you are mounting the work right at the front of the jaws, is more of an irritation than a real problem.

Suggest you grind jaws(with jaws griping something inside) only after you are happy with the rest of the system, or simply buy a new chuck keeping the old one for non critical work

Martin, might be a bit wrong with this, but bear with me.

Currently, I can force some movement of the main spindle at the chuck end by virtue of inserting my fist underneath and exerting pressure. I am reading a total deviation of about 0.035mm ( = 0.00138 inch ). I am thinking that a cutter working against a rotating steel workpiece will generate even more force, so my readings may be 'conservative'.

Quoting Hopper from another post:

"Myford's factory inspection sheet specifies 0 to .0003" runout on the spindle, for a brand new lathe leaving the factory. "

... so it looks like my machine is showing wear of something between four and five times the amount of normal runout as compared to the lathe when new.

The best way to correct that would be with a brand new replacement spindle. We all know that Myford stopped supplying spares for this old girl a long time ago, so that leaves the alternate approach of tightening the casting journal.

But is it not so that if I (delicately) lap the headstock shim to allow the iron casting to close a bit more snugly around the main spindle to take up some of that wear that I am, in essence, forcing the spindle slightly downwards... thus changing its angle with respect to the bed way?

Would it then not make more sense to first take out the play in the main spindle and only after that check the spindle headstock alignment?

Currently, I can force some movement of the main spindle at the chuck end by virtue of inserting my fist underneath and exerting pressure. I am reading a total deviation of about 0.035mm ( = 0.00138 inch ).

Ignatz,

The play in the bearing is unfortunate, but recoverable. Is there significant side to side play, which is more of an issue?

Difficult to predict what will be worn and where, so I don't think you will 'push' the spindle down at the nose as you imply, even if you do, that will be less significant than any side to side play

The effect of squeezing the bush down will be to close that area of wear slightly, hopefully reducing play. In any case up and down play in the spindle of +- half a thou would probably not account for the amount of taper you have measured (side to side play might).

Spindle play generally manifests as poor finish rather than a tapered cut, but the forces at the tailstock end of the piece when cutting, could in theory deflect the spindle in the bearing sufficient to result in a shallower cut at the tailstock end ( and so can deflection of the work piece)

Proceed as you wish, but if it were my machine, I would do the simple checks first, headstock alignment and twist. It is easier and safer to deal with these before hacking the bearing!

If you still have a significant taper on a test piece after confirming the alignment is OK and there is no twist (and you have mounted the work deep in the chuck to avoid the effects of the jaw belling) then it is time to look at the bearing.

Check front and back bearings/spindle contact for play both lateral and vertical, The front one will have suffered the most impact, but they need to be considered together. The amount of wear is not that great an you have the right approach - proceed delicately! Lap off half a thou at a time from the shim and recheck the spindle movement. Personally, anything less than 1/2 a thou play static would do me. If you go tighter, make sure to run the lathe up at high speed for half an hour (with oil) and check to see the bearings are not warming up unduly.

If you are going to the trouble of adjusting the bearing shims, you may want to consider taking the spindle out to remove any gunk/swarf etc from the bearings giving the lot a good clean. It is unlikely you will damage the headstock with the level of adjustment proposed, just make sure you remember to reinsert the shim before cranking down on the bearing bolt and check spindle rotation by hand as you tighten up STOP as soon as the spindle starts to feel the least bit tighter

You will never reduce the errors to zero so don't sweat the small stuff!

Don't confuse play in the spindle with runout, they aren't the same thing. A spindle in plain bearings needs clearance and for a 1 1/2" spindle a thou and a half would be about right for a close running fit.

This won't necessarily translate to excessive play when the spindle is running with adequate lubrication and under load.

For comparison I googled to see what sort of play other people have:

"I can measure 0.010-0.015" of vertical movement when I pick up on the chuck. Seems like a lot to me."

Edited By Neil Wyatt on 01/08/2018 19:35:44

Neil, that is comforting to know. However...

I've just tested the front-to-back play on the main spindle bearing. In that direction when I attempt to force the chuck to move I measure about 0.02mm ( = 0.00079 inch ) which is only about half of what I measured vertically.

I am also assuming that after forty to fifty years of occasional on-and-off use the 0.02mm I measure is already more generous than when brand new.

But this does suggest that snugging down the casting to make the vertical play equal that of the horizontal would not be amiss. Am I wrong in this?

Spotted Neil,

I thought it looked a bit tight, obviously at eyes glazed over at runout and just assumed play was what was meant.

Having said that, I can hardly detect any play in in the spindle on my 10, sub 1 thou, but it doesn't seem to get warm when run, so that was my justification for 1/2 a thou on play.

I guess if you were really working the lathe in an industrial environment you might want a bigger clearance.

BTW, what in your view would wear more in steel to iron bearings, the spindle or the iron bearing itself? someone told me once that you get a sort of carbidey type glaze on the surface of the iron, and counter-intuitively it is the hardened steel spindle that wears more... is that possible?

No way can a tool cutting parallel to the lathe axis cut a taper as you describe. Does not happen. Go out in your shed and try it if you doubt it. Bend a bit of bar in the vice and then hold it in the chuck and take a cut over it.

Is the spindle even hardened? The original ML7 spindles were not and as the ML10 was the budget model it would be surprising if they were.

The .0007" is maybe a tiny bit more than you might like, but it's not too bad. About .0005 - .001" is pretty good on these old lathes. You might reduce the vertical movement by carefully tightening the clamping bolt on the split bearing as described in the post above. To tighten it up half a thousandth or so may not even require the shim to be removed. Just tightening the bolt could do the job.

Certainly, the amount of play you have now is not enough to cause the kind of taper you are turning.

And you can compensate to a degree for bearing wear by putting thicker oil in the bearings. I use 20/50 engine oil in mine (which will have the purists baying for blood) quite successfully.

Edited By Hopper on 01/08/2018 23:41:14

Placed an order for the ground silver steel stock and so forth. Should be arriving early next week. Then I can go further with testing spindle alignment.

Thanks to everyone up to this point for the really solid and helpful advice!

Just as a theoretical thought..

Bed wear on a hobby machine will usually be by the headstock end. So a large diameter tube unsupported by tailstock and a length skimmed well away from the headstock might be another worthy test? Unless one gets into a goldilocks zone on the area of bed twist?

pgk

Received my steel stocks from the supplier the other day. So, on with the testing...

Rollie's Dad's Method of Lathe Alignment

Almost immediately I realized that the method has its limitations. This deviation test would work very well if we assume that the bed of the lathe is flat and true. That is not the case with my old ML10 where there is definitely wear on the saddle. This means a hollowing at the chuck end, vertically... and also horizontally bacause of the wear on the dovetail and gib... I am assuming there is also some wear to the carriage slide (bellmouthing?) where it contacts the rear dovetail of the bed.

So rather than simply take a deviation reading from the chuck end as well as at one spot further out, I elected to take a series of stepped readings along the length of my test bar. I chucked up a length of 14mm silver steel and proceeded to test horizontal and vertical deviations in the manner suggested. I did this at five locations along the test bar, starting 1cm away from the jaws of the chuck and then at four other locations, each time 6.5cm further on down the length of the test bar towards the tailstock end of the lathe. At each test location I tightened up the carriage locking screw.

The five locations are thus: 0cm, 6.5cm, 13cm, 19.5cm and 26cm

The vertical readings suggest a condition of 'spindle high' towards the tailstock. But is this just a trick of the low readings due to saddle wear near the chuck end of the lathe?

The horizontal readings are also much more out at the chuck end... resolving to what would indicate correct alignment at the tailstock end (where there would be far less dovetail wear).

And, oh yes...

One other problem I found is a certain amount of what I will term 'Vertical Oil Film Droop'. This must be attributed to wear in the main spindle and journal. When turning the test bar the measurement would start at one reading, but rapidly change as the end of the bar sagged... doubtless due to the weight of the chuck and workpiece forcing the spindle down through the layer of oil between it and the cast iron journal.

I measured this 'droop' at 0.02mm ( = 0.0008 inch ) out 26cm away from the chuck.

I don't know if the deviations I've measured are necessarily so bad for this old machine. In any case, I wanted some feedback on the results before I start tinkering with the headstock mount and playing around with shimming.

After that I can try some test cuts for parallel on some free maching steel I've secured.