Lathe spindle bearings and pre-tensioner

[Dan_the_Chemist]

Last night I notice a bit of a rumble in my Rockwell-Delta 11x36 lathe. I narrowed it down to the front bearing area. I tightened the pre-tensioner a little, and the rumble went away. However the front end now runs more warm then I've ever felt it before. Not hot, but I can feel that it's warm when I put my hand on it. The manual tells how to tighten the pre-tensioner, but it doesn't give any specifications for torque, etc.

How do I tell if the tension is right?

[chipmaker4130]

On my Sheldon, the official tensioning practice is to run the lathe at high speed for 15 minutes, the bearing housing should be warm to the touch but not too hot to hold.

[old mart]

If the bearings are taper roller, then just warm is good, but do a frequent hand check during the next few hours of usage just to be safe. Check the rear bearing area at the same time. The old Smart & Brown model A at the museum has plain bearings made much like er collets, and individually adjustable. They respond also to the hand warm test. Both mills have taper roller bearing spindles and also get the hand warm checks.

[Richard P Wilson]

The manual for my Denford lathe suggests tightening the end float so that the bearing runs warm but not hot when run at high speed for some time. Sounds as if you've got it about right.

[nc5a]

The manual for my 13" Colchester outlined a procedure to adjust spindle bearings. First run the lathe at 1000 rpm for 20 minutes with an 8" chuck on the spindle. After the warmup (20 minutes) remove the the spindle drive belt so the only load on the bearings is the spindle itself. Wrap a string around the chuck several times and with a small loop knot in the string pull on the string with a scale that measures in pounds. I used a small fish weighing scale that measured to 0-10 pounds. The spec says the 8" chuck should start to rotate when 4 pounds pull is exerted on the string. If less than 4 lbs tighten the spindle bearing by moving the adjustment nut approximately 1/8" measured on the circumference of the nut. If more than 4 lbs loosen the same amount. The idea is to make very small adjustments until you get it right.

Hope this helps.

[DR]

The instructions for new bearings in my CNC lathe are a bit complicated. You install the bearings and adjust the pre-load nut. Run at a specified low rpm for a time period. Check for heat and spindle end play with an indicator to make sure the end play isn't excessive with a given end load. Keep repeating the process gradually increasing rpm until it can be run at the top rpm without overheating.

This process can take most of an eight hour day. Or, you could send the lathe headstock to the manufacturer and they would do the installation for $1800 + bearing cost (1988 costs).

[Dan_the_Chemist]

I appreciate everyone's help on this. I guess the "warm but not hot" seems to be the most common standard (at least among us hobbyists). I had expected more answers like this...

The manual for my 13" Colchester ...

[snip]

... spindle bearing by moving the adjustment nut approximately 1/8" measured on the circumference of the nut. If more than 4 lbs loosen the same amount. The idea is to make very small adjustments until you get it right.

[BCRider]

No reason why you could not apply the same idea on tension but adjusted to your own spindle and chuck size and used on concert with the temperature test. Plus it sure reads that nc5a's test uses a similar combination of warmth check and moving friction.

The value for the tension on the string will alter based on chuck size and seal and bearing size. So some math might need to intrude to scale his procedure to your lathe. The chuck size would clearly change things. But so would the bearing and seal diameters.

Some of that 4 lbs will be due to friction drag from the seals. I assume that the manual value would be assuming a new or near new machine. For older well loved seals with some wear the seals won't grip as tightly. So I think in that case I'd start with more like a 3lb pull.

Another option since you're doing this is to slack off the preload so there's no shake but the bearings don't give any significant drag either. Then check to see how much moving drag on the string and scale is from the seals alone. Then up that by some amount. If you marked where the nuts are now where it just gets warm you could then set things back to that mark and then you'd have a figure for that preload causing the present temperature rise. And perhaps make an educated guess at any change you might feel is warranted based on an educated guess? Without a more specific procedure for your particular machine that's all you've got after all.

I looked up the manual for a lathe very similar to mine (never have found the exact match). The description for setting the bearing preload for this lathe can be found starting at page 58 in this PDF link.

g9249_m.pdf (grizzly.com)

The issue that I see is that the correct movement after the initial contact depends on a few specific factors. First is the thread pitch of the locking nuts that set the preload from the first point of contact where there is contact but no significant preload.

And second is the type of bearings used. For tapered roller bearings it's one thing. But for preloaded angular contact ball bearings it would be quite different. So there's room for some variation.

[Dan_the_Chemist]

Thank you. I'm not sure how well I can adapt this to my old lathe, but it gives me some ideas.

I looked up the manual for a lathe very similar to mine (never have found the exact match). The description for setting the bearing preload for this lathe can be found starting at page 58 in this PDF link.

g9249_m.pdf (grizzly.com)

[old mart]

Dan, just in case you haven't seen this, there is a bit of info on R D lathes here:

http://www.lathes.co.uk/delta%20metal/page2.html

[The Metal Butcher]

Interesting, the 10" just uses a small 3200 series bearing. No adjustment. Rear is a floating ball bearing, same way. The 11" is a vastly better machine, and yet the 10" is still fantastic!

[J Tiers]

Interesting, the 10" just uses a small 3200 series bearing. No adjustment. Rear is a floating ball bearing, same way. The 11" is a vastly better machine, and yet the 10" is still fantastic!

Many of the Logan machines use the same system. A double row preloaded (internally) nose bearing, and a single row floating bearing at the rear. It works, although for a while Logan was supplying ordinary double row non-preloaded (positive clearance) bearings as replacements. Predictably, they did not work at all. They wised up eventually.

[BCRider]

Many of the Logan machines use the same system. A double row preloaded (internally) nose bearing, and a single row floating bearing at the rear. It works, although for a while Logan was supplying ordinary double row non-preloaded (positive clearance) bearings as replacements. Predictably, they did not work at all. They wised up eventually.

On occasion when I bother to think about them at all... like when I'm actually adjusting the lathe spindle pre-load..... I wonder why the dual bearings at the nose is not a far more common setup. Either preloaded paired bearings as per Logan or even back to back regular tapered roller bearings where they are manually preloaded to aid with lower costs. Or is it the norm on the bigger and heavier machines?

If we think about it there's far more sensitive control by grouping them closely together. And the preload won't change as the spindle bearings heat up from running by as big an amount as I'm sure occurs on a typical setup. Just seems to make a lot more sense.

[The Metal Butcher]

Heat is the reason they are close. Older lathes like my Sidney have them at both ends for the most part, but newer designs use them close at the front. Downside is that the headstock bearings must be adjusted with the cover open on oil ath machines.