Lump in Lathe MT3 Spindle

[SillyOldDufferModerator @sillyoldduffer]

Posted by not done it yet on 26/10/2016 19:12:00:

…

This thread is surely one of how not to attempt a repair.

…

This has thread has turned into a nightmare and my sympathies are with Iain.!

When I needed to replace the gears on my mini-lathe the advice was to tap the spindle out with a large raw-hide mallet. (Judging by the picture the mallet head was 6" diameter and about 10" long.)

Not having such a thing I used a lump-hammer and a block of wood. The bearing was tighter than I expected and needed more force to shift than I was comfortable with, both coming out and going back.

I'm pleased to report that I got my lathe back together without doing any damage, but that was more luck than judgement. It seems that Iain's spindle was even tighter than mine.

Noting that Iain unsuccessfully tried to shift the spindle with a home-made puller what's the best way of dealing of with this kind of 'jammed solid' problem?

Thanks,

Dave

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Michael Gilligan on 26/10/2016 15:53:00:

This is mere idle speculation on my part, but : Is it plausible that the spindle-wall around the MT3 is rather thin, and that the beating it received has 'concertina'd' the spindle

I think I would carefully measure the outside diameter before proceeding much further.

MichaelG.

.

Reading back over the recent posts … I think Hopper beat me to this thought too.

MichaelG.

[Neil WyattModerator @neilwyatt]

For the record, here are some pics from when I changed my bearings that may help those who come after.

I used some of these tools when I changed my X2 to angular contact bearings as well. For the record although I did use bit of rubber mallet & wood to get one bearing free, it was very light taps and I think all four of the removed bearings are still in good nick.

Honest.

Neil

[Iain DownsParticipant @iaindowns78295]

The puller in your last picture but one is what I made. I rather think that was advice you gave at the time (the same picture included). My angle iron was a little thicker than in your picture and it deformed in a big way without the spindle moving a micron. As well as bowing, I couldn't get any more force on the spanner. More gym, perhaps.

As SillyOldDuffer says, the manufacturer's advice is to bash it with a hammer. I too ended up with a lump hammer and some serious blows to get it moving.

So. I wouldn't do that again.

The spindle has pretty certainly had some compression artefacts. most of the bluing was round the throat at first, particularly right at the edge. once I'd scraped for a bit I quite quickly got coverage half plus the way down the taper.

What I see now on the blued taper is a set of small cleared blue areas mainly in a line pretty close to the axis which is the high point of the run out and a few more high spots here and there near the bottom. I can't really see the blue on the spindle though.

So I think there is evidence that The throat collapsed on the re-assembly bashes and possibly an off-centre hit pushed a ripple down one side.

One of the things I tried briefly was to wrap some 1500 grit paper round my taper and lightly lap. this actually appeared to clear off some of the worst of the high spots. However, it's had little or no effect on the run out and I then reverted back to the forum for words of wisdom (not mentioned this before as it's either a bit clever or really dumb…).

It seems to me that I probalby have two options. Replace the spindle and bearings or leave it alone and make the best of it.

Anything that's smallish a thou or so out isn't going to make a lot of difference. Any thing I need long and accurate I would turn between centres.

I'll see what the spindle costs.

Iain

[HopperParticipant @hopper]

Posted by SillyOldDuffer on 26/10/2016 19:35:13:

…

Noting that Iain unsuccessfully tried to shift the spindle with a home-made puller what's the best way of dealing of with this kind of 'jammed solid' problem?

Thanks,

Dave

Heat. Depending on the set up, you can heat the headstock housing to loosen the grip on the outer race of the bearing, or you can heat the inner race of the bearing to loosen the grip on the spindle.

You can use a propane torch to heat the headstock/race. Or you can use an electric heat gun available from any hardware store. Even a hair dryer will work, just takes a bit longer to get up to temp. Recommended temp is about 120C (250F). Roughly, hot enough that if you spit on the heated job, the spit sizzles and evaporates instantly.

With a real sticky job, you can use a combination of methods. Put tension on the bearing/shaft with the puller as shown in the pics posted above, then if the bearing will not move, tap judiciously on the end of the spindle with a soft hammer or hammer and soft drift. (Be aware though that this destroys teh bearings and they should be replaced). Or, better option, put tension on the shaft/bearings with the puller and then heat the housing until the bearings move. If desperate, do all three, puller plus heat plus hammer persuasion.

On assembly, small bearings like these should be able to be installed on the shaft with light pressure from an arbor press, or from a piece of tube slid over teh spindle and tapped with a hammer so that all the impact is on the INNER RACE ONLY.

If the fit is too tight to do this, the spindle diameter needs measuring and emerying down slightly and carefully until the fit is right. Too tight a fit can expand the race — or indeed crush a thin walled hollow spindle — and adversely affect bearing clearance and life.

Likewise, bearings should tap easily into the housing on the OD. If not, burrs and maybe some metal need removing to get the right fit. Again emery tape or paper should do the job. We are usually only talking half a thou or so. And you can warm the housing, and even freeze the spindle and bearings overnight, to ease assembly, providing sizes are right.

[HopperParticipant @hopper]

Posted by Iain Downs on 26/10/2016 21:32:22:

It seems to me that I probalby have two options. Replace the spindle and bearings or leave it alone and make the best of it.

Anything that's smallish a thou or so out isn't going to make a lot of difference. Any thing I need long and accurate I would turn between centres.

I'll see what the spindle costs.

 

Iain

In view of the trouble you had getting the spindle out last time you might be best to try to make do with what you have in place already as first option.

Just use a soft centre when turning between centres and take a clean-up cut over it before each time you use it. This will make the centre run true regardless of whether the taper is running true or not.

And if the lathe is turning a thou or two tapered,  with no  tailstock centre in place, you can adjust the headstock alignment regardless of the condition of the spindle's internal taper.

I prefer turning a test piece over Bob's Dad's method because the test piece gives you a real world result of what you will get, taking into account any headstock bearing play etc under real-world working conditions, which may be different from static conditions when checking bars with dial indicators etc.

Edited By Hopper on 27/10/2016 05:41:06

[Iain DownsParticipant @iaindowns78295]

Thanks Hopper.

With the soft centre, I'v marked position on the chuck plate and the centre so that i can re-assemble in the same orientation each time (after an initial skimming) which means (I hope!) that I don't have to re-skim for each job.

The headstock adjustment is basically tightening up the securing bolts more or less so it's a bit of a blunt instrument. I will most likely use Rollie's Dad's as a starter because it's quick. Depending on results I may end up tuning with a test bar.

Most likely I will leave any spindle / bearing replacements for now, though I will have a try at aligning the headstock. In particular if it's too far vertically from the tailstock it makes centre turning pointless.

If I'm going to replace the bearings / spindle then I most likely don't care how hard it is to get out. I did have in mind to put the headstock in the oven (cook for one hour at 120…) to warm it up, rather than a blowtorch as 1) it should give more even heat and less stress and 2) I don't have a blowtorch.

Iain

[Martin KyteParticipant @martinkyte99762]

George Thomas says in one of his books that inaccuracy in machine tools becomes very much less important when you know what they are. You can compensate for many of them by the way you work. Many people have produced good work on very old worn out lathes.

Don't heat the headstock, put the spindle in the freezer and do the job on a warm day.

regards Martin

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Martin Kyte on 27/10/2016 08:49:06:

Don't heat the headstock, put the spindle in the freezer and do the job on a warm day.

.

Tricky … when you're removing the spindle

MichaelG.

[Martin KyteParticipant @martinkyte99762]

I meant when fitting the new one obviously. You can cheerfully hit the old one with a sledge. Sounds like you can't do it much more damage.

You could of course put a bung in the end, tip it on it's back and fill it with liquid nitrogen. It should drop out on it's own then.

:0)

Martin

[Michael GilliganParticipant @michaelgilligan61133]

I did realise, Martin … Sorry; just couldn't resist.

Seriously, though: If [as it appears] the parts are not toleranced properly, who knows what the best approach will be?

As you say … for getting the old one out [again], brute force should suffice.

The new spindle can [and should] be 'fitted' to suit the bearings [*] but things may get more intersting if the bores in the headstock casting are undersize for the O.D. of the bearings.

MichaelG.

.

[*] … on a Record Wood Lathe spindle, did this by hand, using a very fine diamond hone to remove the peaks of the grinding marks !!

**LINK**

http://www.ezelap.co.uk/tools/Diamond-Hone—Stone.html

[Martin KyteParticipant @martinkyte99762]

Well personally I would fit as many bits together as I could individually. Inner races to spindle, outer races to headstock etc. That was I know that everything that should go together will go together and I would get a feel for 'fits' and any high spots on the way in. I realize that nothing is ever in perfect alignment so it's going to be a little tougher doing everything at once.

Having said all that it occurs to me that if the bearing outer races are not seated well in the headstock and they are at all on the skew it's going to be really hard to get the spindle in so maybe that is what has happened in this case.?

regards Martin

[Michael GilliganParticipant @michaelgilligan61133]

[HopperParticipant @hopper]

Posted by Iain Downs on 27/10/2016 06:34:27:

If I'm going to replace the bearings / spindle then I most likely don't care how hard it is to get out. I did have in mind to put the headstock in the oven (cook for one hour at 120…) to warm it up, rather than a blowtorch as 1) it should give more even heat and less stress and 2) I don't have a blowtorch.

Iain

You should care how hard they are to get out. Headstock castings are not indestructible.

Best to use a torch or heat gun (cheap to buy) so you heat up the housing from the outside but not the spindle and bearings so much. The housing will expand away from the bearings. A wet rag slapped on the bearings periodically helps keep them cooler than the surrounding housing. Soaking both housing and bearings/spindle in an oven will result in them both expanding about the same amount and so the interference fit will remain about the same.

I wouldn't rely solely on freezing in the domestic fridge to contract your spindle a great amount. Freezers are usually only 10C below 0. So only 30 degrees or so below nominal room temp of say 25C. Heat with a torch to 120C gives you 95 degrees above room temp. About three times the expansion. In extreme cases , housings can be heated to about 200C for another three times the expansion. "Freezing" of bearings in industry is done with liquid nitrogen to a temp of about -200C.

Edited By Hopper on 27/10/2016 12:28:08

[HopperParticipant @hopper]

Posted by Martin Kyte on 27/10/2016 10:47:27:

…

Having said all that it occurs to me that if the bearing outer races are not seated well in the headstock and they are at all on the skew it's going to be really hard to get the spindle in so maybe that is what has happened in this case.?

regards Martin

Ideally, if the spindle comes in and out from the righthand end of the headstock, the main bearing (righthand end) should be mounted on the spindle before fitting into the housing. The lefthand bearing should be mounted in the housing before the spindle is installed..
Then when you slide the spindle in, the lefthand bearing guides it into the correct alignment before the righthand bearing is tapped home in its recess. So all should line up within the amount of clearance between the balls and races.

Edited By Hopper on 27/10/2016 12:38:16

[Martin KyteParticipant @martinkyte99762]

Liquid nitrogen is for shrink fits mostly. What we are talking about here is gentle fitting. You should not have to resort to millwright methods on a little lathe so the fits need to be sorted out first. I am, as I said talking about reassembly not getting it out.

Martin

[Iain DownsParticipant @iaindowns78295]

Hopper. I had in mind the oven for re-assembly.

warm the front bearing and put it on the spindle. Let it cool.

head the headstock. Assemble the spindle from the front. then press the rear bearing in from the rear using something turned to extend across the full width of the bearing (less the hole in the middle).. I actually managed this on the rear bearing in re-assembly and this worked reasonably well.

With a bit of luck that's got all my bearings on without resorting to a hammer.

I imagine the next step is to let it cool and then preload the bearing.

Blowtorching the headstock may be a useful way of making the extraction easier. If I ever get round to it!

As and when and if I might measure the spindle diameter and headstock retainer (if that's the word) and see how for out of fit they are. I might have done that when I had the issue, but have no recollection of the result.

Iain

[not done it yetParticipant @notdoneityet]

The simple principle of never force fitting any bearing by transmitting force through the bearing contact surfaces is a basic fundamental. A simple mechahanical rule. Ignore it at your peril.

 

For ball races, the outer is bedded into position by force applied to the outer race only, and the sindle is fitted by force between the inner race and spindle only. Removal often means the bearing should be replaced, if the above simple rules are not able to be adhered to, while stripping the spindle from the housing. One very good reason why roller bearings are the better option in many cases (providing preload adjustments are carried out to specification).

Edited By not done it yet on 27/10/2016 19:47:33

[Iain DownsParticipant @iaindowns78295]

My supplier (SPG Tools) tells me that they don't supply spindles and that the run out I'm seeing is normal for this type of lathe.

ARC tell me that they don't supply a compatible spindle.

So it seems I'm left with making the best of what I have. which is fine.

Apart from the run out, the headstock is not in line with the bed at all (~0.5mm out at 200mm, accounting for run out) so I shall have a go at tweaking that out, perhaps skim the faces of the chuck plate and reskim the end of my soft centre.

Thanks for the advice and help.

Iain

[Neil WyattModerator @neilwyatt]

Posted by Iain Downs on 28/10/2016 21:52:47:

My supplier (SPG Tools) tells me that they don't supply spindles and that the run out I'm seeing is normal for this type of lathe.

ARC tell me that they don't supply a compatible spindle.

So it seems I'm left with making the best of what I have. which is fine.

Apart from the run out, the headstock is not in line with the bed at all (~0.5mm out at 200mm, accounting for run out) so I shall have a go at tweaking that out, perhaps skim the faces of the chuck plate and reskim the end of my soft centre.

I stuck a 1" brass bar in my Zither 3-jaw on my mini lathe earlier today. At 90mm from the front of the chuck and 160mm from the end of the spindle runout was exactly 0.001" on a 0.0005" dial gauge.

I think that's about as good as any lathe/3-jaw can give, in fact I wouldn't believe it if anyone else claimed a result like that and I'm sure many folks won't believe me! Maybe I was lucky with errors cancelling out but I've had similar readings before (and only when the chuck is spotless)

I very much doubt the headstock is misaligned the design is fundamentally accurate, as you can bore it out with the headstock on a jig that mimics the inverted-v bed. Much more likely the bearings aren't seated properly.

You can get a new spindle from the USA:

littlemachineshop.com/products/product_view.php?ProductID=1341

You coudl also put up a wanted ad to see if anyone who has fitted a 4" flange spindle has their old spindle going begging.

Neil

[Author]

Posts