Angular contact bearing preload

[JeanMicheling]

I have a spindle running at 3600 rpm with a 550 lbs, 12'' diameter tool, mounted with a pair of 7020 15deg ang. cont. at each end. Two companies did bearing analysis for me. One (bearing manuf.)said there was too much preload, one (spindle repair company) said the opposite. It's a high impact, high vibration application. The actual bearings are light preloaded. I'd like to have some experimental feedback on ang. contact. bearing preload.

Thanks,

 

[Tmoose]

Load, speed, lubrication, preload method, shaft and housing design, and manufacturing tolerances all combine to influence the "right" preload.

Too little preload for a given radial+axial force allows "unloading" and excessive ball excursion and poor spin/roll ratios. Adding (enough) preload reduces those bad things. Spring preloading does not offer nearly the control that "rigid" preloading does.

At some point higher preload generates more heat than the system can handle, and thermal runaway results (squeak).

Sometimes somewhere in the middle heat generation will make the spindle run warm enough to cause tolerance and dimensional issues, which can be a pretty big problem too.

We made a pretty successful motorized milling spindle running a pair of rigidly preloaded HC 7020 at 15,000 rpm with air-oil lube. There were SEVERAL tricks required to make it live. It was real easy to turn the bearings (Actually one of the pair) into junk in a few minutes.

 

[carburized]

Preloaded bearing do work to eliminate excessive vibration, clearances and increaases the stiffness of the system. But on the other hand it shortens the life of a bearing and also depending upon how much is preload, it could create lot of heat which of course is detrimental for the system. so, really how much preload you need, depends upon the application and accuracies you need.

http://www.gearsandbearings.com

 

[Tmoose]

http://www.timken.com/products/bearings/techtips/PDFs/Vol6No3.pdf

"slight preload can be beneficial to the success of long wheel end life. Slight preload can improve bearing, seal and tire life, but only if the entire process is in control."

 

[JeanMicheling]

What would be for you a control process?

 

[Tmoose]

"What would be for you a control process?"

That wording was from the PDF Timken link.
I believe the "process" is whatever controls the assembled bearing preload. The "control" would somehow cause the bearings to end up installed with Goldilocks (just right) preload.

The graphic image in the right hand column on page one shows the "bearing life" reaching a maximum with some amount of preload. Greater or lesser than optimum preload both result in less "bearing life."

I last fought with Timken bearings a few years ago. I think the preferred preload was 1 0r 2 0.0001s of an inch tight than zero axial play.

As I recall the mechanism explaining the counter-intuitive notion of greater bearing life with preload was NOT smaller stress reversal, like in a bolted joint. Instead it had something to do with more of the rolling elements sharing the load.