Excessive Spindle Play Fixed (newbie Error — I Didn't Need New Bearings)

[Redlineman]

And...



With all the pieces. Gets a little confusing, but it shows much more clearly where/how everything sits... once you get used to what you are looking at!

 

[wa5cab]

Rex,

I'm also an EE, first Aerospace and then for much longer, NDT in the Oil Patch.

Although I still don't see what advantage there was to complicating the design, I agree that the Belleville washer is installed correctly. And that assuming that the left bearing is free to slip in the hole bored through the headstock, it doesn't put an axial load on the right bearing. However, if you were able to turn LA-259 Nut a full turn, that is 0.050", almost a sixteenth. So the several components on the left end of the spindle were as loose as geese.

I also agree that the illustrated parts drawing of the 11" (and the 10", which is the same except for a couple of parts) was poorly done. The two rows of parts are not shown in the order that they actually exist on the spindle. The draftsman who drew it must have been drunk and the engineer who approved it high as a kite.

 

[Rex Walters]

And... With all the pieces.

Beauty! (As Bob and Doug McKenzie would say.) Many, many thanks! (Just a tiny bit more useful than my chicken scratches. ;-)

That colorized diagram is wonderful. It clarifies the assembly tremendously, and now I can see precisely whats bearing against what.

When I torqued the red take-up nut, it pressed against the green spindle gear, which bears on the left spacer (dark blue), which bears on the yellow rear bearing, which bears against the right dark blue spacer, which bears against the step to the 1.225" diameter portion of the spindle. Whew.

The surprising thing is that this means the right spacer must locate the left edge of the rear bearing such that it slightly overhangs (or at worst lines up with) the step from 1.1233" diameter to 1.1806" diameter (otherwise the left spacer can't bear against the inner race of the bearing).

I think Robert is right, things were seriously loose on the rear of the spindle. In fact, since the rear bearing must be close to that leftmost groove (1.098" diameter) I think my explanation of the rear bearing canting is almost certainly correct. Once I snugged it up it was better registered on the shaft and (presumably) in the bore of the headstock.

This weird thread taught me way more about my lathe than I expected! Thanks to all.
--
Rex

 

[Redlineman]

Interestingly;

The Belleville washer was not used on the early versions, and as such was not depicted in the drawing as you can see. Neither my lathe (early #15235) nor the later headstock I bought for parts had one. My guess it that it was put there simply to take up space. The end cover at the main bearing does not load the spindle at all, but has a relief inside, matching one in the headstock, that the washer easily fits into. Perhaps that gap allowed the spindle to pull to Starboard under repeated heavy cutting? We'll never know, but it is interesting to speculate.

For a company to get into the lathe business that never made that sort of equipment might be reflected in some of the things they did that do not seem normal, or might be deemed sub standard. Yet, all things considered, they did a pretty good job and made a nice little machine.

 

[Rex Walters]

The Belleville washer was not used on the early versions, and as such was not depicted in the drawing as you can see. Neither my lathe (early #15235) nor the later headstock I bought for parts had one. My guess it that it was put there simply to take up space. The end cover at the main bearing does not load the spindle at all, but has a relief inside, matching one in the headstock, that the washer easily fits into. Perhaps that gap allowed the spindle to pull to Starboard under repeated heavy cutting? We'll never know, but it is interesting to speculate.

Actually, I think my lathe pre-dates yours slightly (mine is a Wards 2130, s/n 3124A, circa 1941) and definitely had the belleville washer. The diagram from my manual (posted above) still shows the washer (part LA-247).

Logan's serial number table says they started making lathes in May of 1940. You must be more familiar with their history than me, I just found their history page on the wayback machine (the page on the Logan Actuator site seems to have gone missing).

Interesting that they'd get rid of the washer but leave the relief in the end cap. I can see the relief in the colored diagram — I guess since there was enough bearing surface there was no reason to re-design the part, but I'm surprised they didn't bother to eliminate a machining process step.

I don't know enough about mechanical engineering to have an opinion on the quality of Logan's design, but I definitely prefer my lathe to the Atlas lathes that it was competing with back in the day!

Also, I just now realized I've been incorrectly referring to my lathe as an 11" model. It's actually a 10" model! I just measured and it's just a bit over 5 1/4" from center to the flat part of the ways.

Lastly, I think I'm going to laminate that colored diagram and stick it into the inside of my the belt cover on the headstock! Love that diagram.
--
Rex

 

[Redlineman]

Indeed;

I AM speculating, which is all that is left to us in many cases. The early section view drawing does not show/list the LA-247 Belleville washer. Neither my '42 (what I call a "coffin" horizontal emblem early model, still using a lot of MW parts), nor the later headstock I bought (what I call a "tombstone" vertical emblem "series production version) had it. The later exploded diagram shows it, and your earlier MW has it. Did mine disappear? Was yours added? Who knows, eh? Always fun to play sleuth though.

Given that the drive end bearing is not blocked against the headstock in any axial sense, I'm not sure what LA-247 is supposed to do?

Guesswork.

 

[Rex Walters]

Given that the drive end bearing is not blocked against the headstock in any axial sense, I'm not sure what LA-247 is supposed to do?

I'm not sure what you mean about the bearing not being blocked against the headstock. The snap-ring and groove in the bearing blocks mine from moving farther to the left (axially) into the headstock bore.

[Whoops. Now I see what you wrote -- the drive end bearing in the rear. It's not located axially. Apologies.]

I think the belleville washer was there to provide constant pressure. It pushes the front-bearing to the left with relatively constant force (pushing the snap-ring into the headstock casting) while still allowing some expansion to the right if things heat up. Without the belleville washer (spring) there is no give. On a hobbyist machine I doubt it makes any difference, but that's the only explanation that makes any sense to me.

Regards,
--
Rex

 

[wa5cab]

Well, now that we know that you have a 10". I understand why the illustrated parts list .looked just like the later one in the 200 Series manual (which is the only manual that we have in Downloads). The earlier 200 does not have the Belleville washer. As I said earlier, I'm an EE by profession, not an ME. But given that the left spindle bearing is free to move axially relative to the headstock, I can't see any valid reason for the Belleville washer to have been added. It neither added nor eliminated any machining step But it's there on later production.

I don't want to start any discussion that often degenerates into a fight over the pros and cons of Atlas versus Logan or any other US badges. But I have to say that there can be no doubt that tapered roller bearings of adequate spec are superior to ball bearings for lathe and milling machine spindles. And flat belts were already quite obsolete by 1940. Those are two of the reasons that 35 odd years ago I bought a new Atlas instead of a Logan or a South Bend. Even that long ago, I already had a decade of design experience with small and large machines using both types of bearings and both types of belts.

 

[Rex Walters]

Shoot, I still can't believe that Sears (the pre-Kmart Sears) isn't around any longer. I suspect I'd have gone new as well if they were.

I'm still not 100% confident that I understand what pre-loading is or does, though. Please validate if my understanding jibes with yours.

If I understand correctly what you and the engineerlive.com article that Henry provided are saying (a very big "if") then the normal way to spring pre-load this type of rotating spindle would be to have shoulders on the spindle or some other mechanical way to prevent the two bearings from moving toward each other axially. The shoulders would bear on the inside faces of the inner races, and springs (or Belleville washers) would bear on the outer faces of the outside races. This would cause the inside and outside races of each bearing to offset minutely, apparently taking up internal clearance between the balls and the races. This is what the nmbtc article refers to as "duplex face-to-face" preload.

I'm basing this understanding on your comment about not seeing a valid reason for the Belleville washer in my lathe since the rear bearing floats axially, and from the following two sentences from the engineeringlive.com article:

For mounting, the spring is normally applied to the non-rotating part of the bearing, typically the outer ring.

and

Axial adjustment calls for the mounting of at least two bearings in opposition, so that inner and outer rings of each bearing are offset axially.

Here's what I think I was missing:

I'd always visualized bearings as having the balls running on flat races, but if I understand correctly (that "if" again) the balls actually ride in a curved groove in the races. So when the inner and outer races are offset axially, the balls are more tightly squeezed in the races (both axially and radially!).

If I've got it right, then "pre-loading" bearings just means minutely offsetting the inner and outer races of a bearing axially to remove internal clearances. This works because the ball bearings ride in groove, not on flat races.

Visually, here is what I think pre-loading means (a cross-section looking at a bearing on edge). Please let me know if I've got it right (finally!).



If this is correct, then I understand your comment that the Belleville washer in my lathe must be there for some reason other than pre-loading. Because both the solid stop (the clip-ring) and the spring (Belleville) bear on the outer race, the inner and outer races aren't being offset.

Sorry for prolonging this thread so long, but I figure if I'm this confused by the term "pre-loading" then I'm probably not alone. The conversation has improved my understanding of my lathe immensely. I hope it's useful for others.

On a sadder note, while the play at the spindle itself was greatly reduced, I'm still getting a few thou of play at the chuck jaws. I'm still getting a lot of chatter when parting a chucked piece (but parting with collets works a treat). It's more than possible I'm still doing something wrong, but several thou of play at the chuck jaws can't be helping.

Regards,
--
Rex

 

[Redlineman]

Gad no;

Play at the chuck will never come out in the wash. The trick is identifying it. I've got an old plain bearing lathe that moves many 10ths of an inch axially. Not 10ths of a thou, 10ths of an INCH. Now... THAT machine is an adventure! The only way I can get a decent surface finish - or most of the time a close tolerance - is to use a FILE for the last few 10ths.

I think you have the preload thing pretty well figgerd. Look at MBFrontier's rehab thread and reference his preload mod on his spindle. This mod coming from another enterprising machinist named "JST" who ran into these issues and has some real brain skillz.