Attached is an image of a failed cage. It is a spherical roller double row bearing, supposed to be floating, circulating oil lubricated, and water cooled. Hot gases of about 1000F are passing through the rotor.

I'm trying to visualize as to what caused the steel cage to rupture. Apparently, after being ruptured the rollers lost their orientation and were dragged instead of rolling, that is why flats are present.

What forces in general are applyed to a cage causing it to strain (bearing misalignment, axial load, excessive temperature ....) rather then normal wear and why?

Thanks,
David

In a deep groove ball bearing, misalignment of inner ring and outer ring would definitely stress the cage but I don't think that is as much a factor for spherical roller bearings which are very tolerant of this type of misalignment.

Some excerpts from SKF Publication PI 401 E 1994 "Bearing Failures and their causes", Page 43, section titled "Cage damage"

"Excessive speed - If the bearing is run at speeds in excess of that for which the cage is
designed,the cage is subjected to heavy forces of inertia that may lead to fractures.Frequently,where very high speeds are involved,it is possible to select bearings with cages of special design." [I think this may include gyroscopic force in a spherical roller bearing the axis of rotation of the roller is continuously being forced to change - gyroscopic force resists this change and I think some of the counterforces come form cage to overcome this]

"Blockage- Fragments of flaked material or other
hard particles may become wedged between the cage and a rolling element,preventing the latter from rotating round its own axis.This leads to cage failure." [It's easy to imagine that when roller gets stuck to cage and stops spinning it creates friction which puts continuous stress on cage]

"Vibration - When a bearing is exposed to vibra-
tion,the forces of inertia may be so great as to cause fatigue cracks to form in the cage material after a time. Sooner or later these cracks lead to cage fracture." [I don't understand this one]


"Cages in bearings subject to severe acceleration and retardation,in conjunction with fluctuations in speed,are affected by forces of inertia.These give rise to considerable pressure between the contacting surfaces,with consequent heavy wear." [It sounds like wear was not likely the root issue in your case. If it was, that opens up the issue of lubrication and contamination as well]

I see various references which I think imply that sudden variation in speed causes load on the cage. Perhaps reciprocating machinery may cause this?

Also if there is skidding I think the cage has to reaccelerate the roller as it comes back into the load zone.

This message has been edited. Last edited by: electricpete,

David
Can you tell from the tracking bands on the raceways that there may have been a predominant axial load. If so the float may not have been set correctly for all temperature conditions. You may have ended up with a row of rolling elements basically very lightly loaded and not tracking correctly. This would causes uneven loading in the cage pockets and ultimate fatigue failure of cage.
Lubrication issues can also lead to premature bearing cage failure. Contamination ( dirt ) being the usual suspsect in most cases.
Cage failure is difficult to detect using standard velocity readings but there has been some success using demod techniques.

Cage_failure.PDF (88 Kb, 43 downloads) Cage Failure pdf

quote:

Originally posted by electricpete:

"Vibration - When a bearing is exposed to vibra-
tion,the forces of inertia may be so great as to cause fatigue cracks to form in the cage material after a time. Sooner or later these cracks lead to cage fracture." [I don't understand this one]


"Cages in bearings subject to severe acceleration and retardation,in conjunction with fluctuations in speed,are affected by forces of inertia.These give rise to considerable pressure between the contacting surfaces,with consequent heavy wear."

Also if there is skidding I think the cage has to reaccelerate the roller as it comes back into the load zone.

Concerning bearing vibration allegedly causing cage fracture. Perhaps, SKF considers axial vibration and excessive stress cage is experiencing due to its inertia mass and impacting coming from the rollers. Inertia forces in this case are trying to rupture the cage in axial direction.

Transition from possible skidding to rolling as roller enters load zone would be another plausable cause of cage stress. This time the accelerating roller generates a force pulse applied to cage in tangential direction.

Petri,
It was hard to interpret marks on the races. The outside of the outer ring of the bearing and bearing housing had clear marks of bearing rotation in the housing. No discoloration though. So I don't think temperature was high.

Assuming proper lubrication and listed previously types of strain applied to the cage, I am not sure there is a way to detect those conditions in vibration data (including demodulation technique) in advance before wear to the cage has been done and predict this type of cage failure.

Any opinions on that?

This message has been edited. Last edited by: David_G,

David,

This particular failure gave you no warning?

I agree most likely no warning from the cage itself weakening.

The only chance would be if this unknown stressor shows in vibration, lube analysis or temperature - or much less likely if cage partially fails such that cage makes contact with stationary part or abnormal rolling element movement cause abnormal vibration - you might catch it before it completely lets go.

quote:

Originally posted by electricpete:
Also if there is skidding I think the cage has to reaccelerate the roller as it comes back into the load zone.

I think I was wrong on this part. Accelerating the roller back to it's normal rolling spped as it reenters the load zone is done by sliding friction against the races.

But if there is not good preload then the races will not do as good a job of keeping the rollers in position against forces such as centrifugal and gryoscopic forces, so the cage will have to do more work to keep them in position.