Methods of Antirotating a Miniature Ball Bearing?

[sdk_imported]

I am designing a high speed (20,000 rpm) machine with 1/2" bore grease packed ball bearings. The bearings are installed with an interference fit onto the shaft. I am looking for a clever way to anti-rotate the outer bearing race and still provide axial motion relative to the stationary housing. The appoach must be low cost and very simple to assemble. We have experimented with using orings - we've had limited success. We have tried notching the bearing and using a key - but this makes assembly a bear. Is there any other common way that this is handle for small bearing like this? Thanks.

SDK

 

[Tmoose]

If the load is one direction, AND the OD fit is good, AND the shaft is straight, the outer race should not creep.

If the load varies in direction there will be creeping, even with a good fit. Some claim a wavt preload washer can help, but I think the rotating load direction will still try hard to drive the ring around, just like a 1000-1 ratio friction drive.

 

[sdk_imported]

I agree with your comments. Unfortunately, I'm convinced that we need some form of antirotation in our application. I'm looking for a very simple way of providing this antirotation.

 

[EnglishMuffin]

You could consider mounting the outer race in an axially flexible metal diaphragm spring, if you have enough radial space.

 

[sdk_imported]

This is an interesting idea. Can you point me to supplier or a website?

 

[tlee123]

How about this. Rigidly mount the outer race to a housing. Then mount the housing with a parallel blade flexure. It's cheap to make, but will take up some space. It will allow 1 degree of freedom movement that be very close to, but not exactly, along the axis of the bearing.

Tom

 

[EnglishMuffin]

sdk: I think you would have to design and/or make the diaphragm yourself. I do know of one or two instances where this has been done successfully. The main thing is not to make it too stiff, to keep the preload low and allow the outer race to remain square to the bearing axis. I would envisage a one piece component incorporating both the diaphragm and the housing for the outer race, since the components are so small. It could be circular, but a lower stiffness is more easily achieved with a rectangular blade as tlee123 suggests. I don't think it has to be a parallel blade design however, assuming it only needs to be applied to one of the bearings, with the other fixed. There is a book by Mario Di Giovanni on diaphragm design, which I recommend:

http://www.amazon.com/exec/obidos/t...f=sr_1_1/104-9358829-3804713?v=glance&s=books

The bad news is they only have one left in stock at Amazon.