This a 4-pole 2500 hp horizontal sleeve bearing motor. It drives a 6-blade multi-stage centrifugal pump through a Thomas Shim Pak coupling.

In Oct 2005 we replaced the rotor and stator (re-used the same stator frame). . The stator was rewound from a sister motor. The rotor was purchased new from the OEM. It is an evolution we are doing with all of the sister machines as a proactive measure (there were no specific problems with this motor). It's unusual, I know, but that's what we do on these machines for a variety of reasons.

The vibration pattern was roughly unchanged at that time as shown in waterfalls 9-13, except it appears that the RBPF sideband pattern shifted. I am pretty sure new rotor has same number of bars as old but I can't confirm. On slides 9-13 the 1x harmonics are labeled with dashed red line, the possible 2*LF (corresponding to frequency ~ 4x) and it's harmonics are labeled with solid red line, and the possible pump blade pass frequency and it's harmonics are labeled in black.

Within the past week, we have noticed a whining noise coming from out of the air openings. High resolution analysis shows multiple pole pass sidebands around the 1x, 2x, 3x etc running speed harmonics (even though those harmonics are small). See slides 1-3

Current signature analysis shows pole pass sidebands are very low 60+ db (slide 4).

Time waveform shows expected pulsing at pole pass frequency (slide 5). Slide 6 shows the zoom-in of the TWF over small intervals appears periodic at 1x but as you shift over longer intervals the pattern changes (and after a pole pass frequency interval it is again the same).

I am looking back over old data to see if vibration shows any changes. I can't detect many changes (other than that change in RBPF pattern associated with rotor change). I can't tell to what extent those pole pass sidebands existed in the past because we don't have any previous high-res spectra.

What do you think?
Is there any condition needing to be addressed? (or just trend).
Any additional on-line checks suggested?
Any off-line checks suggested?

FWBP23_NoisePostsmall.ppt (1,207 Kb, 63 downloads)

Hi Pete,
Hard to put its feet down on something like that but I see the reason of your concern. I am not familiar enough with current spectra or with electrical motor signature to give you advice but the offset waveform with the degree of RBPF modulation is a serious concern. On the safe side I would strongly recommend to perform a single-phase rotor bars test. You energize (AC) one phase of the rotor (10% rated current) than turn the shaft by hand. You record the current variation between each rotor bar (±3%). A pump of that size should turn by hand fairly easy or with a little help.
Hope it help, Marcel

epete,
You don't mention the load on this motor. Is it over 75%? If so, what happens when you drop load, do the sidebands drop?
The only other thing I would check is the number of slots and bars. If these are new and they changed the combination, there are faults referred to as 'cusping' and 'cogging' that can occur due to a bad slot/bar combination. I believe you can find these combinations spelled out in any IEEE reference manual.

Pete,
One of the reasons why I started off
http://maintenanceforums.com/eve/forums/a/tpc/f/7161085912/m/2981009643
is that I have found out that not always ppf in vibration is accompanied by low db in mcsa. Although theory would predict so. So I also feel there is a missing link. To be continued!
Arie Mol, www.ariemol.nl

There is a good chance that you are seeing rotor eccentricity. I have seen such data many times, it is generally accompanied by a characteristic hum. Mostly seen in 2-pole motors. 2FL too is often quite strong.

Often, such data is characteristic of the motor. If it has suddenly developed, look at shaft run-out & bearing housings at a given opportunity.

Regards,

Aditya

El'Pete,

I worked on a 2-pole boiler feed pump motor that developed a very loud modulated growling sound. My initial thought was that there was a rotor bar fault. The amp-meter on the motor controller was steady; I was expecting a modulated level. Motor current spectrum analysis and motor case flux spectrum analysis did not indicate a motor rotor fault.
The bearing housing vibration data showed pole pass sidebands around 1xSS and harmonics and 1xSS amplitude had increased considerably. Ultrasound measurements with a contact sensor indicated a considerable increase in dB level from previous survey and highly modulated sound level comparable to the the audible sound. The audio from the ultrasound meter also indicated a bearing fault at both bearings, but the inboard bearing had the higher sound level. My conclusion was that both bearings were bad and the inboard bearing was also loose.
The repairs confirmed this diagnosis. I was in the shop when a mechanic said that he had the new bearings installed. They were heated for a shrink-fit on the shaft. I took the new inboard bearing and slid it off the shaft by hand. The shaft required a collar that was machined to restore dimensions for a real shrink-fit.
An open mind and attention to detail is always required for both machine diagnostics and repairs.

Walt

The apparent discrepancy between low and high magnitude SB activity observed respectfully in el. current and vibration spectrum could be explained by dynamic eccentricity as oppose to a rotor bar fault as mentioned above. This could be caused possibly by a poor mounting/looseness. It looks like there are too many 1x orders which warrants a rotor lift check.

David

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