I checked a motor today that had several harmonics of turning speed. This is a 3575 RPM motor with a high 120 Hz peak. There were distinct slip frequency times two sidebands around 6 harmonics of turning speed. The motor was making a loud, rough pulsing sound that did not sound so much like ohter motors I have heard when two times running speed comes in and out of phase with 120 Hz. I suspected possible rotor bar problems but current analysis did not confirm this. Has anyone seen this and if so, what was the problem. Could it be loose rotor bars in the slots or a loose rotor on the shaft? Amperage is steady at 48 amps rated at 57 amps full load. The motor was just installed two days ago and had been in a motor repair facility for storage. I talked to one of the guys at the motor shop and found out it actually came from another motor shop that use to have the contract for repairs and storage. So I have no idea what the repairs or history is on this motor. Any ideas would be greatly appreciated.

Thanks,
Ronnie

Could be dynamnic eccentricity. Possibly from excess bearing clearance or lack of downthrust on a vertical motor with rolling element thrust bearings. Or from bowed rotor (in which case expect high 1x).

Pete,
Thanks. The looseness I am seeing could be due to excessive clearance in the bearing fits. This is a horizontal motor. I spoke with one of the mechanics that installed the motor and he said you could move the shaft axially easily and he even asked me if it was sleeve bearings. It is definitely ball bearings. So there appears to be loose fits and my guess is the housing fits are loose.
I am attaching a word documnet with a route spectrum and some expanded high resolution spectrums to show the sidebands. I also put one waveform to show the modulation.

Wire_Shower_Pump_Motor.doc (72 Kb, 78 downloads)

when I expanded the spectrums to illustrate the sidebands, I cut of the difference frequency. It is .83 Hz on everything.

Ronnie:

First, the axial movement suggests that whoever assembled the motor forgot the expansion washers. These are usually wavy-washers (there are other terms, as well) that fit in the opposite drive end bearing housing that apply some level of pressure on the outer race of the bearing to hold the rotor in place. It is supposed to allow for the axial expansion of the motor shaft as it warms up. If this (or these) are missing, I would be concerned about other things, as well.

If the motor was stored for a long length of time, and it is large enough (what is the horsepower?), there is the chance that the rotor and shaft sagged, which would explain some of the issue that you are dealing with (dynamic eccentricity).

A loose rotor or loose bearing housings can also cause some of the signature that you are seeing, especially in a 2-pole motor.

Howard

Howard,
Thanks for your input. I am familiar with the washers you speak of. I worked in a motor shop for about 15 years as a mechanic. The motor is a 250 HP and I am not sure how long it was in storage. I can asked someone at work Monday. I was told there is only one spare so however long the other motor was in service is probably how long the motor in question was stored.
I was not aware that dynamic eccentricity could cause pole pass sidebands but when Pete suggested it, I trust his knowledge. With you confirming it surely sounds like something I should look further into. Unfortunately this type motor does not allow you to check air gap with the motor assembled so it will require disassembly and a runout check of the rotor iron. I think with the amount of looseness and end play, a wavy washer and some machine work might be in order.

Thanks,
Ronnie

In case we are off (EPete and I) and it actually turns out to be static eccentricity, if you take apart the motor and check the rotor eccentricity, if the bearing housings look good, you might want to check the rabbet fit to bearing housing.

Something else struck me, before you remove it, was there a significant amount of shim required to align it? If this has gone through a repair process where they forgot the washers, if it was rewound, could they have used too high a temperature?

In a published study just after I left Dreisilker Electric Motors, the the owner of Dreisilker and I did a mechanical study on the effects of temperature on the stator frame. The primary reason was because the Hydro Quebec/BC Hydro study we were involved in on the temperature impact of temperature on EE/PE iron did not cover the mechanical condition of the frame (ie: what happens to green steel).

This was published at the 1997 Electrical Insulation Conference.

Depending upon the way the frame 'twists' through high temperatures, if soft foot is angled versus flat, then the frame can be 'tuned' as the bolts are tightened down. This can cause some level of static eccentricity.

A way of quickly checking for this is to either use the dial indicator method, or to have your analyzer set up to watch the time waveform and loosen (then tighten) each foot. I would expect, to get the type of signature that you are seeing, would require a fair amount of soft foot - motor or base.

Howard

Mechanical_Effects_IEEE_EIC_1997.pdf (119 Kb, 41 downloads)

You definitely have mecahnical looseness. You could verify by lifting the motor shaft up and down...should be less than .003".
The sidebands are being created from the modulation you are hearing and it could also be loose rotor bars.
The data you supplied also indcates two more possible problems, loose rotor lamination stack to the motor shaft, or inner race defect. The time waveform data does not show significant impacting from a bearing defect..
Hope this helps
PS a motor with a bad rotor will loose performance, trip out, get hot, and with an analog meter the current will rythmically move back and forth. You could also do a single phase test on the motor prior to removal, this too can find a bad rotor...

Howard,
Friday, before leaving, I suggested we do a soft foot check Monday morning. I know from past experience that two pole motors are very sensitive to soft foot and will create the 120 Hz frequency by causing static eccentricity. I want to monitor the vibration and loosen one foot bolt at a time. This method has proven to be very effective at finding a soft foot. Of course a dial indicator will be very useful to determine the amount the foot lifts up. But it will not tell you if the foot is flat or angled. I suppose feeler gauges could be used for that and step shimming for correction.
A few years ago we were instructed to lower the temperature on our ovens to avoid warping the stator. But how widely known or practiced is that? The shop that had the motor was where I use to work and I know they are aware of this. But the shop that originally had the motor, I know very little about. Not to mention there is no telling how many times this motor has been to shops for repairs. Even reconditoned motor usually get dipped or VPIed and baked.
As for you suggestion of checking the rabbit fit to bearing housing, I think that is a good idea. Because of the contract, I am sure the motor will go to the shop I use to work for and I will recommend a thorough inspection of end frames and stator for eccentricity.
One thing I did not mention before, The motor shop did an MCE test and found some inductive imbalance, I think it was 9%. The technician said that was a little high but was still acceptable. The RIC test showed it to be OK but I did the current analysis with my analyzer to verify it was OK under load and it was. What would the inductive imbalance suggest to you?

Thanks,
Ronnie

Dave,
You must have posted while I was replying to Howard. Thanks for the input. I went to the MCC and looked at the amperage. It was very steady. I do not think it has lost any performance yet. The amps were 48 rated 57 FLA, seem to be in order and the slip is 25 RPMs which does not sound excessive. I had wondered about the loose rotor bars but not the loose laminations. I totally agree with you that there is mechanical looseness. I think the spare, which is the motor that was just pulled, will be given some priority on getting fixed and hopefully this motor will not need to be in service for too long. Thanks for the feedback.

Ronnie

Ronnie:

The static inductive unbalance will not tell you much, unless the unbalance remains similar in that phase if the shaft is moved and the motor retested. However, if there is an inductive shift with the RIC test, then that can indicate some issues. Also, with the RIC test, if there is eccentricity you may see the whole RIC waveform 'arc' from one side to the other. The attached is a RIC I performed on a nitrogen cooled annealing motor that I had designed (using an ALL-TEST IV Pro - would be similar results with an MCE). The arch in the waveform turned out to be the bearing housing off center from the rabbet, the impact at the valley of the sine-wave turned out to be a VERY large casting void that interrupted several bars.

If you have performed a current signature test with any current signature system, static and dynamic eccentricity have very specific FFT signatures, regardless of load (in dB). Static eccentricity would be the (RPM x Rotor Bars) with sidebands of 1 and 2 times line frequency. Dynamic eccentricity would have RPM sidebands around the 1 and 2 times peaks.

If you do not have the number of rotor bars, please provide the manufacturer name and model number and I will see if I have it in my database.

Howard

Rotor_Test.ppt (64 Kb, 49 downloads)

Howard,
That is an interesting RIC test. If you look back at the second spectrum in the document I posted, there are several peaks at the higher end of the spectrum spaced at 1X and they are synchronous to turning speed. My first thought was that one of them is the number of rotor bars, possibly 42, and the others are sidebands of turning speed. If these are indeed rotor bar and sidebands, would this be what you would expect to see with dynamic eccentricity? I have the mark set at 42 orders and the sidebands either side of that. But I could move the mark to the right or left and get the same effect. I would think 42 give or take a couple could be the number of rotor bars. I will check Monday for the name plate information.

Ronnie

Yes, it very well could be.

I will be onsite at a steel plant Monday working on a synchronous motor problem. I will be able to get back to you with rotor bars and how to confirm them Monday evening.

Howard

Howard,
Here is the Name Plate information:
GE 250 HP
3775 RPM 2300 Volt 58.9 Amps
Model 5K509EN1309H
Frame 509LS
6315 Ball Bearings in both ends.
We played around with the feet some this morning and was able to lower the 120 Hz peak slightly but the motor is still making a loud pulsing sound. For now I think the plan is to run it until they get the spare back.

Thanks,
Ronnie

Ronnie:

The closest I have is a 300 hp, 2-pole, GE in a similar frame size and 2300 Volts.

Rotor Bars = 39
Stator Slots = 48

You can confirm the number of each by multiplying the RPM (in Hz) times the rotor bars or stator slots, then adding 60 (LF) and subtracting 120(2LF). When viewing the FFT in a Linear format, you should see peaks at about those points. They may be very small.

So, if the RPM is 59 Hz (3540 RPM), then the confirming peaks will be (Example):

Rotor Bar Signature: (39 * 59) +/- 60Hz = 2361Hz and 2241Hz (within the resolution of the data collected).

Stator Slot Signature: (48 * 59) +/- 60Hz = 2892Hz and 2772Hz.

If there are distinct peaks at those points, then the rotor bars and stator slots are correct.

Howard

Howard,
Thanks for the help. If I use that formula, I could feasibly use anything from 36 to 46 rotor bars. 39 as well as 42 fall into that range. I will request the motor shop to provide me with the correct number when they pull this motor.

Thanks,
Ronnie

In the programs that we have initiated at General Motors and US Steel, we have included the count of rotor bars, stator slots and bearing size and manufacturer as being required information from the repair shop. The data tends to be difficult to obtain from the manufacturer.

You may want to consider adding that information to your repair specification reporting from the repair shop.

Let me know if I can provide any more help.

Howard