Ball/roller temperature

[Hup]

I have read properly older threads concerning bearing temperature and changes in clearance, fits, preload, etc. It seems to me that one thing was not mentioned yet and this is the rolling element (RE) temperature.

Tmoose in thread821-120829 mentions typical ring temps during steady state conditions. Then, I suppose, could be stated such a typical temperature of RE for specified type of lubrication (which probably determines the heat losses on RE). I understand, that it is just the result of heat transfer in the bearing.

In some article I've read an assumption, that at each contact (RE/IR or RE/OR) half of the heat enters the ring and half enters the RE. When heat losses from lubrication are small and thermal resistances of RE contacts are large, the RE temp could be high enough to affect internal clearance together with IR and OR temperatures.

Does any of you have any experience, theory, opinion on the "typical" RE temp assuming known temps of rings and type of lubrication? I saw some pictures from WZL Aachen, where guys were using some kind of "IR endoscope" to examine spindles.

Thanks, H-up

 

[Tmoose]

On all the stuff I work on the final answer is the bearing "speedability." If either the balls or the races get hot and swell up to the point of run-away thermal preload the conclusion is mutual total destruction. The temp I mentioned was a number often mentioned in bearing literature, and was of a general nature, and may well have originated in Aschen.

 

[Hup]

I'm just trying to make sense out many discussions on the topic of clearance without considering the ball temperature. When I think about the thermal equilibrium in a bearing, it seems to me, that the ball temperature could be (espec. for grease lubricated applications) much higher than the temp of either of the rings.

Although the ball diameter is much smaller than the ring radius, its contribution to thermal problems could be also signaficant due to higher temperature. This way I explain to myself the superiority of oil-air lube in high-speed applications.

 

[rnd2]

"........Although the ball diameter is much smaller than the ring radius, its contribution to thermal problems could be also signaficant due to higher temperature. This way I explain to myself the superiority of oil-air lube in high-speed applications."

Assuming other factors to be equal and there is no thermal transfer, it is conceivable the ball temperature will increase more than the rings. However, in a >normal< system this does not exist because there is efficient thermal transfer between the metal parts. As to oil -V- grease, Oil-mist provides an auto-replacing super clean lubricative surface film, reduces overall drag and overall assists with cooling. Grease provides only a lubricative film.

 

[Hup]

What a pity that it's so complicated to estimate all the thermal issues (resistances of contacts, discussed in thread821-99946, heat).

Concerning heat flow generation, I performed several "coast tests" on my test spindle with spring-preloaded ACBBs. Observing spindle's deceleration in time I figure out frictional moments and heat flow produced by bearings.

In Harris and also in many catalogues, formulas for the speed-dependent part of total heat (mostly viscous and some spin) contains speed with 2/3 exponent. Nevertheless from my measurements it looks like the exponent must be greater than 1 in any case (making function H=f convex instead of concave). A also saw some papers from other people having similar results.

 

[Tmoose]

Small point.
In modern times air-oil and oil mist are 2 distinctly different lubrication methods. Even some semi-modern literature fails to differentiate between them, but the spindle innards and the lubrication equipment are not the same, or interchangeable. Any body who has worked around machines with each system can see (and breathe) the difference.

 

[Hup]

Sorry for my beeing undeterminate. By mentioning oil-air I meant any means of lube that removes heat from the bearing by some medium (air).

Anyway the spindle I mentioned is grease-lubricated and well run-in. So I would expect the well-reputed semi-modern literature to work.

 

[Hup]

Ehh, ..., there should have been mentioned "frictional torque" instead of "heat flow" and the function M=f in my last but one post. The heat flow is convex at any rate because there is one speed more in the equation of course. I'm sorry for confusion...
Nevertheless the issue remains.

 

[bonerigo]

What external temperature should I expect to measure as a maximum at the outter surface of a new pump drive atnbthe lower motor bearing, a vertical 100 HP motor, This vertical inline pump has solid plate bearing guards on both sides with only 1/2 to 1" clearance at the top for air to flow. The bearing are double row ball angular contact and are grease lubricated. There doesn't appear to be too much grease.
Bonerigo