The kinematic equations show this simply. Otherwise think of it as this, in one shaft revolution a point on the shaft passes some of the rolling elements, and what this point doesn't pass in 1 revolution (outer race fixed) these rolling elements have passed the starting point on the stationary race -- otherwise the inner race would have passed these balls.

--> sum of BPFI + BPF0 = number of rolling elements.

Thanks forr the comments and suggestions. Still thinking about some of it. A few responses.

quote:

David G
Do you set such a high Fmax=100,000cpm for such a slow machine?

Fmax was selected high for investigating the abnormal noise and suspected bearing defects. We also have routine route data with Fmax I think 10,000cpm where we can get better resolution on the low frequency stuff.

quote:

Danny Harvey
Was the bearing recently lubricated?

It was greased around last week of November. Shortly before the unusual noise was first heard.

quote:

IMO there is a possibility that (based particularly on 12/11/07 TWF: single sharp impact every revolution ) there is a crack in the inner race.

You can't tell from the powerpoint, but if you zoom in on the TWF, what you see is not impacts spaced at 1x, but an envelope peaking at 1x. Within that 1x envelope are BPFI spaced impacts. It corresponds to the spectral pattern of harmonics of BPFI with 1X sidebans.

quote:

Do you have enough historical data to calculate some statistical information? Instead of relying on a 100% increase (doubling), you could use the average plus 3 standard deviations. You could consider using data from the sister motors as well

We don't have a lot of history on the high Fmax acceleration measurement. I tend to think we are beyond 3 standard deviations compared to sister motors and there is no question it is much different.

quote:

In the meantime my advice is to lubricate the bearing to soften the spectrum,

We have thought about that. But we are also mindful that lubricating can sometimes make vibration worsen drramatically from one previous experience.


NOW – NEW INFORMATION:

You recall that on Monday and Tuesday morning, we had seen bearing temperature at approx 115F on housing and 130F on shaft (hotter than all the others). This afternoon (Wed) , we recorded bearing temperautre at 150F of the housing and 170F on the shaft ! Vibration level was stable at previous levels. Later in the evening, back down closer to 115/130 again.

What do you think would cause that?

Up until now I have been linking the temperature symptom and vib symptom under the assumption that something like misalignment increased the loading which caused both 1- incr in temperature 2 - bearing degradation. Now I'm wondering whether the temperature might be a result of the bearing degradation (which only occurs in final stages of bearing failure?)

We're not sure whether this might be a daily cycle or random variation associated with redistribution of the grease, or how it bodes for the future. We will be monitoring very very closely for the next several days.

Pete,

I have had two cases similar to what you are describing. One of my experiences was a motor bearing with irratic temperature spikes after I had diagnosed a bearing outer race defect. Vibration levels, as best as I can remember, remained pretty much constant.

The other case sounds very similar to your sitiuation. The motor is a 2000 hp, 1800 nominal rpm, ac motor driving a large ID fan. I had diagnosed early stages of a bearing defect. This motor has the internal RTD's so we store constant bearing temps and we can trend them. I had noticed bearing temps fluctuating during the hottest part of the day, this occured during July and August so ambient temps were at times into triple digits. But on two or three occasions the bearing temps would increase dramatically higher than usual for no apparent reason and later in the evening would return to near previous levels. During these temperature increases I could see little or no increase in vibration. Several weeks later vibration levels started to increase and more difinitive bearing defects started to show up in the vibration. The bearing temperatures never did increase the way they had earlier in the degradation of the bearing.
Like you, I wondered if this was caused by ambient increases, or lubrication redistribution.
In my case the bearing was replaced 6 to 8 weeks after the temperature spikes. Failure analysis showed a spalled outer race which originated from electrical eddy currents.

I hope this will help with your bearing. I would monitor the bearing very, very closely but from my experience you "may" be OK for the scheduled replacement.

Travis

Thanks Thud. That is some good info. Now that I think about it, I am coming to the general conclusion that grease lubrication can be a very transient phenomenon and changes in temperature when lub is poor are to be expected. This fits in nicely with another observation I have on some other motors with large bearings (6313+) running at 3600rpm. I have often found the grease discolored and races stained appearing to be overheated, but the bearing housings don't appear hot. I think the grease gets hot, then redistributes (grease location and oil partioning) and "cures" itself for a little bit. I have read somewhere others that concluded grease lubrication can be very dynamic, but I can't find where I read that.

UPDATE:

We took some shaft riders measurements:
On the motor shaft just below the motor:
11 mils N/S (all are pk/pk)
13 mils E/W

On the pump shaft just above the stuffing box
8 mils N/S
16 mils E/W

We monitored this for a few days. Saw the temperatures bounce up and down somewhat randomly without any correlation to ambient temperature.

We have pulled the motor. During removal, checked as-found alignment and found:
5 mils TIR N/S (2.5 mils offset)
15 mils TIR E/W (7.5 mils offset)

Checked hold-down bolts - tight. Checked pump lift - normal.

I'll be watching the motor disassembly at repair shop. Planned checks are:
Inspect for proper as-found grease quantity, grease location, grease consistency, evidence of mixing or contamination
Gather Grease samples from grease input area, from just above bearing, and just below bearing
Check as-found endplay
Check as-found shaft and housing fits
Check as-found rotor-stator airgap
Check upper bearing insulation resistance
Inspect bearing for evidence of upthrust, pre-existing defects, pitting from current, and other degradation mechanism
Check shaft runout during as-left uncoupled run

Does that sound like a reasonable set of checks? We had considered as-found uncoupled run, but believe it would not contribute much. We had considered checking upper radial tilting pad bearing clearances, but that requires extra disassembly and does not appear related.

Any comments: Could the relatively small misalignment be related to the bearing fault? What do you think of the shaft rider data? Any other inspections we should do?