what is the relationship between the radial & axial play in a metal-to-metal bearing? 1

[rollingcloud]

The axial play is defined as the distance by which the inner ring/ball can be moved in an axial direction relative to the outer race from one extreme position to the opposite extreme position.

Radial play is the total displacement of the ball in the radial direction.

Sometimes I see both radial and axial clearance specified, sometimes it's only one and sometimes it's mixed. For example, in this case, NHBB listed max axial play where SKF provided axial play.
So, in an ideal world where both profiles of the race and the ball are perfectly spherical, axial play would equal radial play? In real world, the differences are due to manufacturing variation such as roundness tolerance? Or is the race inner profile not exactly intended to be spherical?

 

[3DDave]

In that example you can calculate from the geometry the relation between axial play and radial play. Since radial play is diametral and axial play is along a chord they aren't the same.

If there was only infinitesimal width the radial play could be finite and the radial play infinite as the ball would pull out of the housing.

Typically a maker will include the resistance to radial and thrust loads. If a bearing is not suitable either axial loads or radial loads the play in that direction are of not much interest. Sleeve bushings, for example, have no resistance to axial loads, and have no specified axial play because that is infinite.

 

[rollingcloud]

You nailed it again! I get it now, I did not have the correct picture in my mind when thinking about the axial play.
For the same bearing, RBC provided no axial nor radial play, but they did provide no-load-rotational breakaway torque, is this another way of indicating internal clearances?

 

[3DDave]

That is indicating the bearing is preloaded and has no play.

 

[rollingcloud]

For an ultra-low speed application where positional accuracy is non-critical, I believe the difference between loose play (0.002'' max) and no play (0.25 - 5 lbs of no-load rotational breakaway torque) would be insignificant,correct? My understanding is that the no-play case would offer better positional control, with the only downside being an increase in heat during rotation. However, since full rotations are rare, this difference would be undetectable?