From a leaky nose to a bent nose

Just coming to terms with the oil leak and concluded the nose is bent on the L5. Had to bore a shaft with a 1.3" id over 6 inches long for the Corlis. No matter what I tried I ended up with one end ten thou smaller than the other. Put the face plate on and what do you know - well out. I managed to face the nose shoulder and rebore but it was still six thou out. Most frusrating! Time for a better lathe me thinks.

Which spindle fitting does your L5 have? If you are producing a taper, it would mean that the headstock is not in line with the bed.

Edited By old mart on 27/12/2020 20:08:19

Its taper roller bearings from (I think) the late 40s. I would tend to agree with you except when I put a dti on the face plate it was not good! Everything clean so no swarf behind it. I managed to take a thou or so off the face of the spindle which reduced the offset but its stil out. For everyday use It does what I need but no good for boring to within a thou.

A bent spindle shouldn't produce a taper. Once you've done a clean-up pass the part is turning on the spindle axis so every pass after that should be consistantly true regardless of the state of the spindle nose or the chuck etc, so long as the chuck isn't rocking on the seat or the part rocking in the jaws but that should be evident in the finish.

Turning an unwanted taper is a geometry issue. Mis-aligned headstock, twisted bed, severe wear, but not a bent spindle.

not a bent spindle - what do you percieve a bent spindle would do? non of the proposed issues would show as a face plate out by 10 thou per half revolution...parts are rocking well and good...bent spindle. The other factors wouldnt cause the face plate to wobble.

Slow down there's no need to get defensive. You never mentioned rocking parts or face plate wobbling - you just said it was 'well out' which given the taper issue I took to mean you were running a DTI across the face and the reading was changing. Your main concern was the taper you're turning and that IS the one thing that wouldn't be caused by a bent spindle.

Gotta be more concise with your info if you're hoping for good advice.

Sounds like the faceplate just needs a skim. The problem is either the faceplate or the owner of the lathe.

Question: Should we be facing face-plates - being as most lathes cut concave to some degree? Obviously the only route - if the lathe is all one has at hand - and any flatting after machining on the lathe would need to carefully maintain the correct geometry.

Discussed in the summer

I can't imagine what you could do to an L5 to bend the spindle (at least without gross damage being visible on spindle nose, headstock casting, etc.).

Has it been taken apart and the taper roller bearings incorrectly re-installed?

Right, as a Member of the L5 Owners' Association...

Old Mart -

The Harrison L5 was offered with either screwed spindle (as is mine) or L00 fitting.

'

Andy -

I agree. It would take one hell of an accident or utter ham-fistedness to bend a Harrison lathe spindle by five thou; and damage like that would be noticeable in all sort of ways. Rather oddly, the official manual basically says the lathe is rugged enough to stand up to students and apprentices!

So Fizzy -

- let's not jump to conclusions without considering all other alternatives, but by all means carefully test the spindle. Even more certainly don't try to correct an assumed fault by machining the spindle, because that is likely only to create yet more problems. What you did there may have been compensation rather than true repair.

'

You say you were boring a shaft and it came out with a taper of 0.01 " on 1.3" dia over 6 inches length; but don't tell us which is the smaller end, nor describe the set-up. Did you bolt the cylinder to the face-plate or was the face-plate for the subsequent test?.

You don't mention using the faceplate for it, but a faceplate not truly perpendicular to the lathe axis would create non-cylindrical work - though I can't quite visualise what shape it would be.

Anyway, we have a significant error spoiling the work.

Six inches of work-length mainly unsupported just attracts problems, so we can assume you also used a steady or some other method to increase rigidity. We can also assume that the tool was sufficiently rigid, right height etc. etc.

So ....

If we want to blame the machine....

Are there wear-hollow in the bed-ways, perhaps due to the machine having spent years on small-part batch-production? Or on the saddle itself? Test the saddle for side-play and vertical play, and its travel: both but more especially side-play, would create tapers. A saddle out of adjustment or on a wavy bed can " walk " (proceed with a yawing motion), and it's feasible the cylinder's length is near the walking wave-length; with the geometry magnifying the yaw to that 0.005" radius error. Or more simply, wear has given the bed an unwanted effect resembling a taper-turning attachment.

Did you turn the shaft with the gap-filler in? I think the faceplate will just clear it. If so, did the saddle have to climb a little step up onto it, from a wear-ramp in the bed (it does on mine!). That may have at least contributed to the problem.

Now to the face-plate. Is this wobble radial or axial - the latter revolving it like a swash-plate? How certain are you that the face-plate's own register is not badly worn, or that someone in the past has not tried to re-face it and not quite got it right?

'

Finally, the headstock and spindle... Don't go cutting lumps out of it, certainly without being absolutely sure where the fault really lies, and that you will correct it by turning it!

(I wish I could have words with a past owner who cut my L5 nose back for no known reason, badly scarring its end face and significantly shortening the taper.)

First ensure all the bearings are in good condition and correctly adjusted for radial play and end-float. It's possible that the fault is due to a perfectly straight spindle sagging in what you say could after all be the original bearings. It could make the nose point either up or down, giving an effect akin to setting-over the tailstock for taper-turning. The cutting forces might cause the spindle to deflect sideways, with more marked tapering, but I'm guess they will have been too small in this instance to have much effect.

'

The only way you can assess the spindle's straightness is by sets of measurements, but you do need first assess and if necessary adjust the bearings correctly.

You need test the spindle's concentricity of rotation, by DTI , at both ends and if possible near its centre, but I am not sure if the main gear cluster has a suitably-machined surface there. I think it's cast or forged.

If after all that the spindle genuinely is bent (unlikely) you'll probably have to live with it.

If though it is simply being deflected by badly-worn or out-of-adjustment bearings, that might give enough swing to produce errors magnified by distance from the bearing.

Edited By Nigel Graham 2 on 30/12/2020 21:41:15