Spindle bearing issue

[N8PDX]

I’m wrapping up an overhaul of a Sheldon EM-56-P lathe and hit a snag with the spindle. I was planning on not replacing the tapered roller spindle bearings that it came with as they still feel and sound good to me. But I just discovered that the bearings appear to have been installed incorrectly in that the hand engraved markings on the outer races are not aligned (see pics - arrows point to engravings). In fact they are 180-degrees off, which I understand would maximize the spindle runout. I’ve had the lathe running in the current configuration and runout at the spindle is around half a thou, but I suspect it should be better.

My questions:
- Am I assessing the situation correctly?
- Is there a recommended method to pull the outer bearing race without damaging it? I only need to pull one and reinstall it correctly.
- Should I just live with it for now and not risk damaging the bearings until I’m ready to replace them?
Thanks!

Front bearing:


Rear bearing:

 

[ltlvt]

Personally, I think .0005 is pretty good for worn bearings, to get that other .0003 it is gona be costly. Did you check the spindle for TIR without the bearings? I would not just assume it is the bearings without data.

 

[projectnut]

I'm not sure you're going to get any less runout regardless of which way the bearings are installed. The last page of the parts manual shows a copy of the inspection form used when the machine came from the factory. The tolerance for run out at 12" from the spindle nose is .0000" to .0006". You're within the original specification, and I doubt realigning or replacing the bearings would do much but drain your pocketbook.

Here's a copy of the inspection sheet that came with the parts manual of my MW-56P.

 

[Cadillac]

To know if the races are orientated correctly you need to know the spindle runout. Orientating the outer races 180* would give you maximum runout with each revolution of inner race.
The inner races should be 180* out hopefully cancelling each other. The outer should be aligned.
You need to know spindle runout so you can offset the runouts from bearings to spindle. For removing race best would be a expandable arbor on a slide hammer.

 

[Nutfarmer]

I wouldn’t mess with them unless you are prepared to buy new bearings. To remove a bearing race from what looks like a blind bore has a high chance of damage.

 

[Ischgl99]

I would suggest checking to make sure the outer rings are not loose in the housing. These should be snug, but it’s possible they were installed correctly, but rotated over time if the fit is too loose. See if there is any looseness in the fit, and if not, leave it as is.

As other have said, don’t mess with them until you have to. Precision tapered roller bearings are expensive and have limited availability. 0.0005” runout may not perfect, but it’s far better than spending hundreds of dollars per bearing to marginally improve it.

 

[N8PDX]

Thanks all for the input and to projectnut for the inspection doc. I’ll use that to see how close I can get to original specs when I get it back together.

To check the spindle runout with no bearings, here’s what I did… I laid the spindle in v-blocks on my surface plate, rotated it, and looked for variance at/near the bearing seats with a 5 tenths indicator on a surface gauge. I’m really not seeing any appreciable movement, like maybe a tenth. Is there a better way to measure the spindle TIR? The two bearings have different IDs so I can’t directly compare the seats with this setup.

I’m fairly certain the outer races will not come out easily. They seem to be a tight press fit, and I don’t believe they have rotated on their own. They are in blind recesses in the head casting.

It sounds like the consensus so far is to leave well enough alone until the bearings need replacing, and that’s the way I’m leaning. Given that the outer races are already installed 180 deg apart, how should I orient the inner races on the spindle for the least overall runout? Would a different test of the spindle TIR affect the answer? Thanks!

Nate

 

[ltlvt]

Given the data that has been provided by your inspection and the Data sheet provided by Projectnut you may want to consider that you are over thinking something that may be taking valuable sleeping time away from you. I do it all the time. You may want to consider counting sheep instead.

 

[RJSakowski]

I can't see how orientation of the outer races would affect runout. They are fixed in place so their position doesn't change. At most, an eccentricity between the o.f. of the race and the bearing surface would shift the spindle axis cause an effective headstock misalignment. The inner race, on the other hand, rotates with the spindle and an eccentricity in that bearing race would show up as runout. Ineven wear in the rollers themselves would also show up as spindle runout, as would runout in the seat for the inner race or loose bearing adjustment.

 

[N8PDX]

I think when I put it back together I‘ll take some runout measurements with the inner races in different orientations to find the best option.