Thought I would post some info on a motor bearing defect that I've just found. This is a motor mounted in a vertical position driving a centrifuge. The bearing in question is a NU220...not sure what manufacturer, but it really doesn't matter anyway. I threw this presentation together rather hastily and I know it's a bit long, but I thought it might be useful for some folks.....oh yeah, I've recommended to have the bearing lubricated for now.....I'll keep an eye on it for the next few months.

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

Bearing_Defect.pps (390 Kb, 269 downloads) Bearing Defect Presentation

Cool stuff, Billy.

I guess there is a thrust bearing at the top of the motor?

Billy,

This is very cool!

I'm just thinking about how to interpret all this.

The way I read it is that BPFO is modulated by FTF. Trying to picture what kind of defect in the cage would modulate magnitude of impact on the outer race. In particular when this is a vertical machine. . Clearly, there is a huge defect in the outer race.

David

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

Danny, I'm not exactly sure about the thrust bearing....I'm told by maintenance that there is something there to handle the thrust, but the motor tag just shows a NU220 and 6318 bearing for inboard (top) and outboard (bottom)respectively....so I'm not sure.

David, I'm not sure how to interpret the BPFO sidebands around cage frequency....normally, I'll see ball spin instead of outer race, but that doesn't seem to be the case here. I have heard of this before though. Maybe someone can shed some insight into this.....

I had one case with a horizontal motor with the same bearing where the tab that engages the lock washer with the shaft keyway sheared off (it had been reused ) and that allowed the nut to slowly back off until it became pinched between the bearing and the end cap. Rapid heating resulting in a total lockup and rotor to stator contact and a total wreck of a vacuum pump motor. The only clue I ever saw was an elevated floor.

IMO there's a defect on the outer race (a spall or crack). Also the cage might be distorted since and a section of the rollers impacts the defect in a stronger manner thus the cage modulation. The attachment illustrates this theory.

distorted_cage.ppt (27 Kb, 83 downloads)

I find the patterns in the Autocorrelation for the PeakVue very interesting. Since the Peak g-level in the PeakVue waveform on the inboard is near twice that on the outboard, I would postulate the problem is on the inboard bearing.

I have seen the sinuosidal pattern in the autocorrelation data similar to what you have in cases where the bearing was slightly 'cocked' placing undue stress on the cage (postulated to lead to rub between rollers and cage), but the amplitude of the autocorrelation factor at cage period was much less than what you are seeing (this would be expected if rub was occuring since there would be 'randomness' in the wave form). Your peak correlation factor is 0.65 leading to the conclusion that greater than 80% (square root of 0.65) of the waveform is from a very periodic source (very little randomness). Thus 'rub' is not very probable.

For this bearing (Fag NU 220), the BSF fault frequency is very close to 8 times the FTF, and the BPFO is 16 (# of balls) times the FTF. Of course a defect in the rollers would lead to activity at BSF side banded with cage. There are activity at BSF as well as BPFO. I would suspect you have both a BSF and a BPFO fault in this bearing. I would be suprised if lubricating would lead to a long term reduction in the impacting.

Thanks for the insight Mr. Robinson....I agree that there is probably a ball spin defect present also....I'll try to be sure to update on the progression and post back with details.

Billy, you have way too much time on your hands. But that was a good job. Autocorrelation is nice. Kind of makes me yearn for RBMWare.

Believe it or not, it didn't really take that much time to put together....but I do find some extra time sometimes Hey David, did you get the stuff I sent ya....and good luck...give me a call and let me know how it went?

Billy,

I have 47 vertical centrifuges that I've been monitoring very closely for 3+ years. The motor nameplate is probably correct as all the motors on the machines that I monitor in this plant and all that I've seen in other plants have roller bearings on the drive end and ball bearings on the opposite drive end. The roller bearing handles the radial load from the belts and the ball bearing carries the axial (thrust) load. Of all the motors that we've removed due to mechanical problems over the past few years the primary damage on all of them has been on the inner ring of the ball bearing. Whether this is due to thrust load remains to be seen.

Are these by chance Merco Centrifuges? Of all the machines that I've ever monitored these vertical centrifuges are by far my favorite as we see and learn something new all the time. It never gets boring!

Hey Michael, thanks for the info. I'll keep that in mind while looking at these. I have not seen an outboard bearing fail yet, it has been my experience, in my plant only, with these centrifuges that the drive end cylindrical roller bearing usually fails first. Matter of fact, we just changed one last week that I called out back in August. I've got a request into the motor shop to save the bearing and have it sent back to me because it was doing some humming near the scheduled outage. If it actually returns, I'll try to post some pictures of the faults. As far as our centrifuges, they're Dorr Olivers.....

In my experience, it is not at all unusual to see cage sidebands around BPFO. I wouldn't begin to try to explain it, but I see it fairly often.

Just following up...ran this route again today and levels were not affected by the added lubricant. Amplitudes are all about the same. My 1k to 20kHz band is creeping up a bit more that other parameter which is to be expected. I'll continue to update as time passes...

Billy / Michel
We also use "Westfalia" centifuges and one in particular(test machine) the NDE bearing(ball) keeps on failing in the motor. All signs lead to lack of lubrication although we constantly keep adding grease. Anny problem like this on your vertical motors?

Haven't seen anything on the NDE bearing for our centrifuges. Could it be in your case that you're adding too much grease?

Also, as an update to this cage defect problem that I've found.....wouldn't you know it. A major process change has been inacted that will take these centrifuges out of the loop. So I won't be seeing any more data collected for these machines....oh well that's the way it goes sometime.

Billy,
We add the required amount as sugested by the vendor and still had 4 bearing failures due to dry running in a 12 month span. We are tracking a dB reading and ad more grease as soon as the reading reach about 90dB and prolong the MTBF. The last bearing taken out had a cracked ball, spalding on the thrust side of the outer race and was bone dry. It seems to me that the grease is flowing into the motor and and not through the bearing.The grease path, when adding, is from inside the motor to the outside. If taking in consideration that the motor is mounted vertical, would it not be better to change the arrangement or redesign the NDE to run in an oil bath? See attached picture
Riaan