What's new
What's new

Deflated, went to look at a Smart & Brown 1024 found the spindle bent.

Peter.

Titanium
Joined
Mar 28, 2007
Location
England UK
This machine caught my eye on eBay, it was only about 45 miles away. It looked pretty good in the pics and had all the necessary bits with it bar the steadies so I went to look it over. It was still powered so I ran through all the controls, everything worked and it ran at full speed without making any strange noises. I put a 14" long bar in the collet and tugged as hard as I could on it up and down and my 1um dial indicator didn't so much as twitch. Then I ran it slow and put the indicator against the camlock mounting faces, and this is where my heart sank.

http://peterrimmer.myby.co.uk/vids/1024spindle1.mp4

http://peterrimmer.myby.co.uk/vids/1024spindle2.mp4

It's showing 14um swing on the face of the camlock, that's just over 0.0005", and 8um on the camlock OD taper, about 0.0003" with a similar amount on the ID collet taper.

I guess it's had a pretty heavy crash to bend what is a fairly chunky spindle for a 2hp lathe.

I've already bid on it before going to view, and it's a fairly low bid so I expect it to be surpassed. If not I'm going to be faced with either parting the machine out for bits and scrapping the rest or looking at ways to repair the spindle. What are the options - grind in situ? It isn't much to grind a few tenths surely and the machine does have a taper attachment. What would your thoughts be on this?
 
I'd machine it to trueness right where it sits. The most important feature is to get the flat face running true, because angularity there becomes exaggerated in whatever chucking extension you add. The small bit of eccentricity of the taper only affects the radial runout of the chuck and a half thousandth isn't much to fret over.

If you machine the taper true, you need to calculate the proper (exact) amount to remove off the face, because contact is simultaneous with both features when a good backplate is mounted up. So, if you clean the minimum off the face, then you'd barely need to scratch the taper, and even then, you'd only be taking a wee bit off the high side.

I did all of this when I rebuilt an older cnc lathe that had worn bearing seats on the spindle. I first tried metallizing the journals, found that didn't stick, so then I arc welded them up, which resulted in about .003" face runout of the spindle. I simply machined it back into trueness. In my case, I did have the comfort of knowing that the bearing journals were both straight and concentric with each other (after remachining), so there was no fretting stress going on with each rotation. I'm not sure how bad that would be on your spindle in this case.

Maybe put a chuck on that machine (before you do any fixing), and machine a piece, and see if you can detect out of roundness, or gear noise, etc. Do some lift tests on the spindle too, to see if it is actually loose in the bearings. It might be a case of wear underneath a bearing, or a bad repair job done by someone else.
 
Hi Peter; I own a S&B 1024 and have had the spindle and just about ever thing else completely apart, I would suggest checking the spindle bearing on the left end (not chuck end) for clearance. On mine that bearing had measurable clearance. Contact me if you need any further info I have parts list user manual and the like.
 
Hi Peter; I own a S&B 1024 and have had the spindle and just about ever thing else completely apart, I would suggest checking the spindle bearing on the left end (not chuck end) for clearance. On mine that bearing had measurable clearance. Contact me if you need any further info I have parts list user manual and the like.

Thanks, if it ends up in my posession I'll be in touch. I would have thought that any play in the rear bearing would show itself when I was pulling on the bar in the chuck though.
 
Hi Peter. Not familiar with S&B, other than that they are very good and well respected machines. Was the owner there when you were doing the testing and did he make any comment? If a lathe has been crashed this hard I would have thought there would be other outward signs of damage, and I would doubt the spindle is bent. In a crash gears usually fail long before this happens. I would look at the bearings first, and certainly strip the spindle and check it, but machine nothing until you are absolutely certain you know what and where the problem is. TBH I never try the accuracy of my lathes, only the accuracy of the work I produce on it, if that leaves something to be desired, and I know I didnt screw it up, then I start to look at things, if not, and it is producing work as good as I need I leave it alone, life is too short..................perfectionism can drive you insane ;-) You did right to put in a low offer!
Phil
 
As you say, it takes a really heavy crash to bend a spindle as strong as that one. So much so that I would want to look for other side effects of the crash. It is very easy to remove the lid on top of the headstock and check for any other damage and runout.

It should be possible to have the spindle taper and face re-ground. In the event that new bearings are required they are available at a price. If the price of the lathe is not too high and everything else checks out, in your position I might still take a chance on it. These lathes are so good that they are worth saving if everything else is right.

The readings that you got in the two tests (14 and 8 microns) are not ideal, but neither are they very bad. I looked up the relevant Schlesinger test charts for a "Toolrom lathe (Highest degree of accuracy) up to 200mm height of centres". The Smart & Brown 1024 definitely falls into that category, although it is much smaller than the upper limit of size and so might be expected to be well within the test limits. Test 5 is similar to your first video and the permissible error for that are .005 mm (5 microns). Your second video is similar to test 6 and the permissible error for that is 0.005 mm. I think I have a test chart somewhere for a 1024; I will have a look for it and if I can find it will let you know what the typical test results would have been.

If you do buy it and then find you need to part it out, please let me know; there are parts I would be interested in.
 
I've watched and paused the vids many times, and have a feeling that one particular camlock has been strained - over tightened etc etc etc ..........probably by some Neanderthal that shouldn't have been let loose near a lathe.
 
So when you put a chuck on it and you have .002 run out do you contemplate suicide? Reality check! You are buying a USED machine. If you want to see no movement in that needle buy a NEW one. If that machine was crashed hard enough to bend the spindle the gear box would be trashed, does it sound like it's full of gravel? Those numbers are better than most used machines. After you check the bearings put it back together and use your tool post grinder to fix it. Job done! If it were me I would treat it like a sore dick....DON'T F&CK WITH IT. Anything you make on it will probably be machined on all sides anyway. Your part will never know the chuck mount face is .0005 out of true.
 
Think re-working the taper and face of the spindle while relatively easy is not a great choice to repair this machine....If the rest of the machine is a "keeper" then i would disassemble
the headstock and evaluate exactly the root of the problem.
Bent spindle is not only an issue for the output (mounting of work holding) but affects the spindle bearings as well.

Careful work may result is the ability to straighten the spindle to a more acceptable level....I once rebuilt a CNC lathe spindle with much worse runout than the one shown...(bent from a crash i believe)
Using purpose built cradles to support the beating seats machined from heavy aluminum plate i got that spindle nose to run round and flat at under .0001" , no machining required, just careful work.

Barring that approach, at the least the bearing seats should be under cut, hard chromed and ground back to standard so that they run true...
then the face and taper at the nose can be corrected....Of course new bearings should be part of any repair.

Cheers Ross
 
If it was nearer here i would be happy to out bit you, my customer would not notice 8 microns of error if it jumped up and kicked him in the gonads the full 14 microns :-)

You would do well to listen to sami, i have seen lathe spindles with error also caused by cam lock abuse, in the case i saw it was a welder trying to play machinist and swap a 4 jaw for a 3 jaw and trying to leaver it of one of the cam locks with a fork lift. The idea of lining up the cams with the marks was a bit to hard for that mig monkey to grasp. In that case a simple stoneing was all it took to put right. (no we did not bury mig monkey to his waste and stone him to death though it would have been fitting end for him!)
 
S&B 1024 ?? another bell just rang ...........lift the headstock top cover and make sure the back gears haven't been ''over crashed'' (the RH of the headstock levers) I've seen a good few where operators haven't let the spindle stop before engaging back gear, it's been a long time but I've a feeling the back gears are a helical cut - either way, unless you've a good set under the bench, walk away.

While the HS top's off, you might be able to get a clock reading from around the mid way point of the spindle - that'll show if it's bent.

FWIW The worst I ever saw, had over half the face width chewed away.

At the back of my mind there's something else to look out for on S&B 1024s, but it won't come - if it does I'll post.
 
Peter ,did you check for turning parallel ,my Model A which has ways that look pretty decent is actually quite worn ,I am having to live with it at the moment but eventually I am hoping to get the bed re ground ,I have never even checked the chuck mounting face therefore I don't worry about it.
 
Is that the (only) one that's on the Bay at the moment? The one that's in Surrey?

I'd tend to buy it and fix any problems after I'd got it. Current price is very competitive with what I've got in my Hardinge HLV, that's a similar spec after I spent several years rebuilding it from scratch.
 
Yeah it's that one Mark. The bed is ok though it is a little tighter at the tailstock end, no ridges or scores though so it's just normal wear. The only real blot is the spindle. I have the highest bid right now, I was willing to go to about a grand after I'd seen it but now I'll just let the dice roll.
 
I'd machine it to trueness right where it sits. The most important feature is to get the flat face running true, because angularity there becomes exaggerated in whatever chucking extension you add. The small bit of eccentricity of the taper only affects the radial runout of the chuck and a half thousandth isn't much to fret over.

If you machine the taper true, you need to calculate the proper (exact) amount to remove off the face, because contact is simultaneous with both features when a good backplate is mounted up. So, if you clean the minimum off the face, then you'd barely need to scratch the taper, and even then, you'd only be taking a wee bit off the high side. snip

I was not aware of what you've written (red) and would like to see documentation of the intent of "simultaneous contact". Can you provide any reference(s) supporting your statement.

In short, I don't believe it, as achieving it would require absolute perfection in machining both the spindle and the mate. I suspect the taper is actually the controlling feature and the face has nominal clearance to any proper mating part.
 
Before machining anything on the spindle I would pull it, check for collateral damage and see where the bend is (if indeed there's one there at all). Machining a face of a spindle that's bent may well be a quick fix, but the spindle is still bent. Typically that's right outboard of your front bearing, and that being the case you're not going to have your precision bearing for long! There's no way in heck I'd just grind the spindle in place, shrug my shoulders, and consider it as good as new.
 
Bent spindles are very rare, certainly in manually operated machines. I can recall repairing a handful. One was caused by careless rigging and another by a fool of an operator using an hydraulic jack to stretch ( straighten ) a large shaft. Nobody knew how the third one got bent.
Personally I'd live with the error.

Regards Tyrone.
 
I'll echo what the others have said. Lovely lathe (got one), certainly worth fixing. The spindle is quite beefy and I doubt that a crash could bend it. But be warned about the bearings; they are expensive and a craftsman at the factory was allowed one week for getting them perfect. And this was for someone who knows exactly what they were doing.
 








 
Back
Top