Rolling Resistance of Dry Ball Bearing 2

[zipcurse]

I am evaluating a medical product design that includes a number of stainless steel ball bearing that have no lubrication. To be clear, the manufacturers normal lubrication has been intentionally cleaned out. I do not fully understand the logic of this, other than to avoid any biocompatability issue.

The ball bearings are 10mm in diameter, ABEC-3, running at <5rpm, very low radial loading (<2% of static rating), no thrust load, and very short life requirements. My only real issue with the bearings is rolling resistance. I have found a number of tools for calculating the lubricated torsional resistance, but nothing for dry. What kind of difference can I expect?

Thank you in advance.

 

[tr1ntx]

At a very light load, the rolling resistance will be infintesimal. It will be a function of the surface hardness' of the ball and race.

That being said, there will be friction if 2 balls rub against each other, or friction between the balls and the cage if there is a cage to separate the balls. That will be a function of load (normal to the surfaces in question) and relative speed between the 2 surfaces.

What kind of difference can you expect? That depends on how tight the bearing is put together and on what kind of lubricant it had in the first place. If it was a heavy grease, unlubed will move more freely.

 

[zipcurse]

Thanks for responding. Under the light loads, I get about 2X what SKF predicts for a similar lubricated bearing. This turns out to be a number that I do care about. I plan on trying the bearing with lubrication, but I was hoping to find a calculation of what I should expect from the unlubricated bearing.

 

[MikeHalloran]

I suppose you could start from first principles, and calculate the torque required to slide the cage on the inner race given the cage's weight, then the torque required to rub a ball or two against the cage to rotate it, etc.

Then maybe from railway engineering you could find an equation relating the normal force at a Hertzian rolling contact to the lateral force required to make the object roll, and relate that to a torque.

OR

Since by removing even the light preservative oil, you're working well outside the recommended application envelope, SKF may not be of much help, but I suppose it's possible that some grad student somewhere took a look at completely dry bearings, maybe for use in a vacuum.



Mike Halloran
Pembroke Pines, FL, USA

 

[zekeman]

.................., no thrust load, and very short life requirements. My only real issue with the bearings is rolling resistance.

Well, since you have very low life requirements, why don't you get some empirical data on your own by speeding up the testing, by ,say , doubling or tripling the speed and use a correlation of actual life to speedup life data.

I think the data you get from this would be as good as you would get from some "analytical" approach.

 

[tbuelna]

zipcurse,

With a ball bearing operating under the conditions you describe, your losses will be due to cage/ball sliding frictions, microslip, and hysteresis as the ball and race contact surfaces get strained passing through the loaded sector of the bearing.

In "Advanced Concepts of Bearing Technology", ch. 5, Harris gives a detailed description of calculating these losses.

One other thing to consider is the nature of your bearing materials. Clean, dry stainless steels in contact with each other tend to gall and cold weld. This may produce higher frictions as a result.

Hope that helps.
Terry

 

[Warpspeed]

Biocompatibility sounds like a sterile environment, which usually means mechanical washing followed by autoclaving after each use.

No lubrication also suggests no lip seals for bugs to hide behind.
That also to me implies a fully open exposed rolling bearing.

I would worry more about foreign objects being trapped and jamming the bearing, especially if the available torque is minimal.

I don't think the actual bearing itself will be the problem, but very fine debris finding its way into the bearing might be.

 

[unclesyd]

I would contact the Dyanroll people for information or have them test your bearing.

Have you considered a full ceramic bearing for this service?
They can run without lube better than metal bearings.

http://www.dynaroll.com/torque-analysis.asp

 

[asimpson]

In theory un-lubricated bearings would have less viscous drag than lubricated.

If friction is a problem it may be the bearings binding due to miss-alignment.

To simplify the system less tight bearings such as ceramic ball bearings would reduce the influence of miss-alignment at the expense of less accuracy. Friction would be much more predictable.

 

[tr1ntx]

"Under the light loads, I get about 2X what SKF predicts for a similar lubricated bearing. This turns out to be a number that I do care about."

It's possible you have a more highly engineered, higher-duty bearing design than you need. A much simpler design probaby wouldn't have that much difference between dry & lubed.