It is a 4 pole 2.5 MW motor. The vibration of motor is not very high (some high frequency peaks were observed). In acceleration spectrum , BPFO harmoics are observed , however no change in bearing temperature or current. In spectrum also peaks present at BPFO. The images are attached. 

Any suggestions for the problem?



Photos (1)
Files (1)
Original Post

A 360KCPM FFT with 1600 or 3200lines resolution should clearly show you if there is an electrical etching/fluting condition present.  This can often times cause the condition that you are seeing within your data.  Typically you will see race wear and cage fault frequencies scattered out in the rotor bar pass area/frequency range of the FFT.

What do you mean you are unable to make a call as the bearing is only 4-5 months old?

It is your job to make a call!!!! New bearing means nothing, the readings show an issue and you aren't doing your job if you ignore it. You sound like you have an issue with maintenance trusting your judgement, the instrument doesn't have an agenda, it shows you what it see's the data looks good so can be trusted. I was once told by someone far smarter than me, that with gE if you have a clear peak over 1 gE then you will at least see marking on a bearing raceway when examines, I see a peak at 1.5 gE in that spectra, so I think you are safe enough to have some noticeable damage if you open the bearing.

It may piss off some motor maker/rebuilder but so what, sometimes we break a few eggs to make an omelette.

Sounds like some false brinnelling possibly. Monitor the bearing closely and if the gE levels start to go up again then I would be getting more forceful in your recommendations.

what does the fft look like in acceleration? When your on velocity hit ‘A’ on your keyboard to view acceleration and then hit ‘V’ to return to velocity.

Dharun Mari posted:

In this case, I suspect that the motor has been stored at a warehouse a long time(without running ). So the bearing balls mark may be the reason for the gE value high.


Initially after installation the values were not higher. Also defect frequencies were not present. After 2 months of running the defect frequencies appear in the spectrum.

False brinnelling may be the issue but it should be present from the beginning .

Further spectrum & twf is also attachedpectrum is attached above is in acceleration


Photos (3)

I don't mind making calls on brand new motors, on the contrary. Is it on VFD or VFD close by? It may be slight change in operation position that made a move in to a damaged operating zone or infinite other reasons. Do it sound like a disaster with stethoscope, screw driver or yellow handle analyzer? Call it then, maybe as planned as suitable. There are no worse than having a fail sitting on data indicating that is the case and not reporting, minimum issue a warning report to swap brg if stop of motor happens anyway. 2.5MW is no toy it is not 38MW that was treated the same in my world but still expensive IMHO.

Add Reply


HACKATHON in association with IMC-2019