I think you are talking about the black bands on the brass cage?

It looks to me that they are equally-spaced at the same spacing as the rolling elements. That would seem to suggest the marks cccurred while the bearing was stationary. Maybe the balls facilitated heat conduction from a hot shaft (heated by the fluid) to that cage and the hot cage interacted with the oil or oil vapor?

I don't think I'd attach whole lot of significance to those marks. I'd be more interested to see the interior of the bearing (balls and races and cage pockets) if you can have someone cut it apart. Although if vibration was normal (and you are looking closely at the vibration signature, not just the velocity overall), then I really don't expect to see much there by inspection.

If I step back and think about bearing noise without vibration, two things come to my mind:
1 - skidding. Is it possible the bearings don't have enough axial preload? I'm not familiar with the configuration of your machine, or the speed etc.
2 - ball defect. Ball defect vibration pattern can come and go as the ball spin axis changes so it can easily be missed during vib surveys. Although if noise from a ball defect was present at the time of vib measurements then I'd expect it to show in vibration.

=====================================
(2B)+(2B)' ?

Electricpete:
Thanks for replying to my post. Pump speed is 3600 RPM and these two bearings are located at the non-drive end of between bearings pump. Maybe because they were installed in 2016 and were not used until recently, and that is causing noise??

ok, so bearings were installed and pump sat for a few years. I'm curious to know if it was in standby or in storage (mostly to test my theory of whether heat from shaft / fluid may have been responsible for those marks).

But either way I guess your question is whether it is plausible that some degree of corrosion might form giving a waviness of the surfaces that creates noise but not detectable vibration. Periodic rotation of an idle machine helps spread oil through the bearing to keep a thin protective coating. Bad environment (especially moisture) of course contributes to corrosion. Also high temperature. My first thought was that probably corrosion products are either hard enough to cause detectable defect patterns in vibration or soft enough to be chewed up without vibration/noise. But that's going a step further than I should, I don't really know. I can say that at our plant we've pulled apart probably 30 or 40 sets of bearings that showed bearing defect patterns in vibration, and 4 or 5 of them had corrosion of some type, often along with other damage. We don't pull apart bearings that don't show up in vibration (other than a few undetected catastrophic failures, long stories mostly without an answer). We have never pulled one apart where the only complaint was noise. So it's plausible to me for all I know. But the other scenario (about preload) is also plausible to me. If we were evaluating it our next step would definitely be taking apart the bearing for further inspection. We would also go back and look at the vibration signature again with a magnifying glass to see if there was anything we missed before. Maybe analyse the oil if a sample was saved.

Maybe others have more comments / insight.





=====================================
(2B)+(2B)' ?

https://www.youtube.com/watch?v=-F87xKCJen4

A bearing manufacturer's "noise" list is attached.

• https://files.engineering.com/getfile.aspx?folder=d165e022-39f5-4418-9436-8

Haha, yeah that gives a lot more to think about when bearings are "making a noise".
But I can't believe they left out "chirp" and "snarl" !

=====================================
(2B)+(2B)' ?

Making decisions or even having discussions based on Well intentioned Descriptions like "Noisy" bearings and "bad" bearings is danged dangerous in my opinion.

Assuming ANYthing is about like playing with one of these in my experience.
https://previews.123rf.com/images/tiero/tiero1503/...

I ain't never been smart enough or quick enough to get away unscathed - 1:50 here -
https://www.youtube.com/watch?v=YWUeAAaSPTI

Hello All,

Thanks for responding to this post. I realized later that there is a lot more involved in figuring out root cause for bearing noise including visual inspection is needed.

We are still working on this and I will share the findings when available. Thanks

Etching is a condition where bearing parts are subject to corrosion by moisture by the direct intrusion of water from outside. When used in high-temperature humid environments, bearings should be protected from these effects. a boiler feedwater pump may leak flashed steam from the pump seal directly down the shaft through the labyrinth seal on the bearing housing and contaminate the oil. Flashed steam may not be noticed but may exist. labyrinth seals are great for fluids but not vapors. Frequent changing bearing oil or lubricating oil that provides corrosion protection from water contamination might help.

OP is putting a lot of emphasis on the time spent stationary. Is the pump in a high vibration environment? False Brinelling can occured if the shaft is left stationary for extended periods in high vibration environments.

I sent the bearings to SKF (bearing OEM) and asked for deep analysis and detailed report. It is somewhat expensive but is necessary to find the root cause and for learning purposes. Will share the results.

> False Brinelling can occured if the shaft is left stationary for extended periods in high vibration environments.

false brinneling is indeed common for long-time idle equipment installed near running equipment. but false brinelling damage is typically easy to detect in vibration, which is why I had not mentioned it in my earlier comments.

but you never know. no point in my speculating any more... there is a bearing inspection report coming.

=====================================
(2B)+(2B)' ?