Should bearings be warming up?

cdhknives

Active User
H-M Supporter Gold Member
Joined
Apr 12, 2013
Messages
560
I swear I read or was told (under directions from a career machinist to allow some warmup time before cutting anything high precision) that lathe bearings should warm a bit after long runs. I followed the instructions to preload the bearings on my QC54 when I changed them out a few years ago, and rechecked/re-tightened them by backing off a turn and re-tightening per spec several times as they 'ran in'...and did have a bit of loosening for the first couple of hours. Recently I spent a few days running it long hours on a project and idly checked the temperature...stone cold. Same temperature as the surrounding metal. No measurable rise using an IR temperature gun. I don't have any felt play in the spindle but still wonder if I should tighten the preload collar another tooth or two. I was lead to understand that properly tightened lathe spindle bearings should have 5-10 degrees of warming after 30 min+ of runtime. True or false???
 
It really depends on the spindle speed, the faster you turn the more they heat up. If you are not having chattering or other finish problems, then I would not adjust them.
 
The Atlas MOLO from 1937 through 1988 says to run the lathe for 30 minutes to an hour and then tighten the thrust nut on the left end of the spindle until there is zero endfloat and then another 2 teeth past the zero end-float point. The Atlas Technical Bulletins that came out around 1960 don't mention the warming run, so the implication would be that you do it with the headstock at ambient temperature. The difference between the two is that the adjustment is done at ambient temperature or the adjustment is done at an elevated temperature. The rest is the same.

At the moment, I can't say which way I think is the most likely to be the best. My suggestion until more information is available would be to do it without the pre-warm, and then do the pre-warm for an hour and see whether or not there is any end float. If there isn't, I would leave it alone. Then report what you found.

My 3996 is about due for a belt change after nearly 40 years. I intend when I have it back together to do some tests. But it might be a few months before I get a round tuit.
 
Can see on page 5 how to adjust the bearings. I would run it 75% of fastest speed. http://vintagemachinery.org/pubs/51/18293.pdf
Also buy a temp probe or infrared temp gun so you can accurately check the temp. A good rule of thump about bearing temps is if you can lay a hand on the casting and it's hot but not uncomfortable it is good. approx. 120 to 135 F. If the spindle cuts good and not getting hot. Let it alone. If you tighten the spindle nut and noting happens maybe the gears are rusted to the spindle and it doesn't slide.

If you want to zoom in on the book. Press CTRL and + keys on laptop and you can zoom in to see it better. If you find this good info please donate a few dollars to Vintage Machinery as it's a non profit.
 
Last edited:
Agreed. Thats about what my machine does. Gets warm to the touch but not hot.
 
All our high precision machines at work are run for 30-60 minutes prior to using them. They will feel slightly warmer than room temps as the oils and bearings etc warm up and the increased accuracy is achieved. This is especially of a concern in using the Landis Cylindrical grinder in grinding our rolls for the rolling mills. Need to get within 0.0002” in diameter and less than 0.0005” taper. When the machine is cold, this is not possible.
 
A hard push/pull on the chuck inline with the spindle with an indicator in the toolpost shows about .0005 endplay max...cold.

I'll speed it up and let it run to see what T rise I get at high speed.
 
About 5 degrees F rise after 45 minutes at around 800 RPM...same .0005" on the indicator. I guess I'll call that good enough.
 
Reading between the lines, one could make a case that the Atlas designers intended for there to be 0.002 of preload on the spindle bearings. As thermal expansion of steel it linear at least over relatively small temperature changes, and using the stated 0.002" length increase for 50 deg. F temperature rise given in the MOLO, 5 deg. would only be 0.0002", which could be a little difficult to discern with most of our equipment.
 
Yeah, my .0005 is a guesstimate, really about half a division of the standard .001" indicator scale. Within the limits of my dial indicator and eyeballs, it was the same half division cold and warm. Not exactly calibrated to NIST standards like some of the stuff I deal with at work, but good enough for hobby work!
 
Back
Top