All, I have attached some data collected on this motor. It currently sits on a test stand and the customer wants it back next week. (like always) At 1000 volts (60hz) the motor runs smooth, as the current is increased to the 4000 volts it runs at, the vibration increases. The speed remains constant and the rotor has been checked for balance. The spectrum shows mechanical looseness and we shimmed the bearing housings with little change. This is a plain bearing motor. Low frequency with high resolution shows what appears to be low amplitude pole pass freqs around running speed. This motor has no load and the slip is very minimal. Rotor problem or am I missing something?

Allis_Chalmers_AC_induction_motor.doc (354 Kb, 43 downloads)

How are you loading the motor on test stand? May be, that might be causing problem. But , the 1X and harmonics shown in the spectrum are not that significant to worry about. May be little looseness.
Regards
Irshad

The motor is not loaded on the test stand. The owner requires less then 1 mil and we aint there yet. Another identical motor runs smooth without any of this. I believe they're going to swap rotors today to see what happens. Would the unloaded condition explain the low amplitudes?

Looks like might have rotor problems. May be loose rotor bars becoming more "loose" with higher voltage due to increase in heat. May even be loose on the shaft.

Was the rotor check during the rebuild or just assumed to be ok based on the customer's information? What happened to the motor that required a rebuild?

What does the phase do as the voltage is increased and the amplitude increases, steady or unsteady or shifting? If steady, then may be that if there is a rotor bar problem, you can narrow the area down as to where it is located based on balancing theory (phase and amplitude).

Thanks Ralph, The phase remains normal horizontal, and vertical. However, axial we have 180 shift accross both plain bearings, horizontaly and verticaly. Very low amplitudes as you can see from the spectrum. The rotor was tested so I've been looking into the looseness and cocked bearings. The outboard bearing was off by a couple thousands and set straight with shims. We spent lots of time making sure the bearings were'nt loose in the housing. In fact, tightening these up actualy increased the amplitude and harmonics as seen in the last spectrum. Even though the rotor was tested, I'm back to thinking it has a problem. Have you ever seen this show up with such low amplitudes? We have no way of loading it because of its size. I performed a bump test on the rotor and stator to rule out resonance.
Pete

I tend to agree with Akhtar. At these low levels, I'm not sure you have any problem.

Every one of those graphs is in g's as far as I can tell. I am used to looking at ips, particularly for the lower frequency stuff. 0.01 g's is 0.03 ips at 1200cpm, 0.005 ips at 7200cpm, 0.0017ips at 22,500cpm.. All the peaks in your first slide (outboard horizontal) are less than 0.006g's = half that. The highest scale on any of your plots was 0.030 or 3 times that - still pretty small.

You had 1x and harmonics that varied proportionally to voltage. That suggests possible dynamic eccentricity. Rotor is traveling sync speed so the field pulling outwards is travelling at 1x. Could be slight bow in rotor (check TIR). Could be excess clearance in bearings. (check clearanace both shaft to bearing and bearing to housing). In a vertical motor once I saw this behavior when there was excess clearance between thrust hub and the shaft but there is nothing similar on horizontal motor.

You mentioned that the vib was sensitive to adjustment of bearing housing bolts. ** Is the vib sensitive to adjustment and shimming of the feet? I would check carefully at the feet for problems that might contribute to the vib.

Also check that the rotor is not running axially against the shaft shoulder. (perhaps scribe rotor position while running and then check after shutdown that the scribe mark is not at a limit of travel).

quote:

However, axial we have 180 shift accross both plain bearings, horizontaly and verticaly.

Can you clarify this statement?

Thanks Pete,
First the feet. I checked phase and amplitude from the feet to the plate. Low amplitude with about 20 deg phase shift on all 4. Adjusting the feet made no differance. The mentioned axial phase readings are on each end of the motor from the right to left of the bearing and above and below. (horz. and vert.)This is where I pick up 180 shift which would indicate cocked bearings right? Axialy from the inboard to outboard is in phase along with the same vertical and horizontal data. The magnetic center is marked and falls into place when the voltage is increased to 4000V. I have already ruled that possibility out. Using ultrasound, I listened for looseness on both ends at the bearing and housing. No noise. Also, I have already sugested to the shop to check the ecentricity of the rotor. They should give me some results later today. I will attach some data in IPS for you.
Pete

Hi,

whilst your conducting the various tests described, try collecting overall vibration Vs time on a coast down. If you loose overall vibration as soon as power goes then its a fairly good indicator that there is some kind of electrical force causing your mechanical vibration.

Ive come across similar faults during the testing of motors and more often or not it has been the motor bearing housings being off-set in the radial's, causing an uneven air gap.

Cheers

DJ

Thanks Spike,
Been there, done that. Yes, as soon as power is cut, vibration disappears. It is beeing caused by the electromagnetic field. Airgap has been checked and rechecked at around .045"
Pete, Attached is the data in velocity. I also have some waveform data. The low frequency data was collected in 6400 LOR. Do you want to see waveforms in autocorrelate?
Pete

SAME_DATA_IN_VELOCITY.doc (215 Kb, 20 downloads)

Here's some autocorelated waveforms that show some wierd patterns.

AUTOCORRELATE.doc (196 Kb, 20 downloads)

quote:

as soon as power is cut, vibration disappears.

I am still betting on rotor. Seems like if the airgap was the problem there would be a 2x LF somewhere in the data.

These low amplitudes are not "low" when displayed in velocity and mils. Why are you working in Gs at such a low cpm? You have over 2 mils amplitude and your customer wants under 1 mil, so you do have a problem somewhere, huh?

As has been pointed out, the vibration levels are very low. Yet, a lot of effort is being expended to obtain even lower vibration. The fact that the vibration disappears at coast down suggests an electrical issue. Perhaps the question here is this--does anyone believe there is evidence of a problem that might impact reliability on this machine? For example, could slight dynamic eccentricity cause failure down the line? If there is slight rotor bar looseness, will it deteriorate? Is this machine subject to numerous starts and stops, which would stress the rotor bars, or does it run continuously, which could mean that slight looseness may never become a reliability issue. It appears likely that this problem will go away when the rotor is changed, but, is it worth the effort?

At .16 ips on a test stand you definitely have a problem somewhere. What make me think is the large ratio of amplitude from vertical to horizontal? .16 and .03. Seems the possibility of a resonance created from something. Has the shaft been UT'd for cracks? Just something to consider. At the lower voltage this may not show up. Good Luck.
Jason

These are not low levels in the vertical, by any means, for an unloaded motor running on a test stand. Are you bolting the motor down on the test stand? Do a finger pinch test under each foot. Any looseness there may be causing the high verticals.

My best advice:
1. Check the feet on base for any looseness/soft foot putting the rotor in a bind.
2. Inspect bearings after running for any rubs or unusual wear patterns that might show bearing misalignment
3. try a PeakVue reading to see if their might be metal-metal contact (rub)
4. Conduct a running bump test to test for natural frequency in vertical direction on the housings. Make sure rotor is on the oil film whe you do this. The 4:1 ratio is very high.
5. Simple ODS (phase/amplitude) of motor/feet/base to see how it is moving vertically.

Contact me if you think I may be able to help you. I'll be happy to try.

thanks

This message has been edited. Last edited by: Bill Kilbey,

OK, I just returned from the shop. They swapped out rotors. I will post some data shortly after a bite to eat and a cold beer.
Looks like the rotor was the problem.
Pete

I have attached the vertical readings taken on the new rotor. One thing we all missed was the second harmonic of RBPF. F max is too low to capture this. 69 rotor bars X 1200 = 82800 x 2 = 165600. RBM wizard did this for me. I wish I would have changed that before I went up there. Don't always trust the wizard! I guess being focused on the low frequency and looseness along with being a weekend, made me temperarly stupid.
The good thing is, it's now exceptable.
Bill, thanks for joining. I did all the feet things and took peakvue. Peakvue shows a harmonic of looseness as well. It still shows this on the new rotor.
Testtech. In a saw mill I would probably agree and run it, but NuclearJason has tighter tollerances. With everything now below .5 mils on a test stand, and clamped to base on machine pads (I forgot to mention that)It should be fine. Right?
Ralph, Sorry, I'm so use to acceleration I sometimes forget. The sister motor will need to be checked as well when the repair the rotor. I'll have some better data setup for it if it's OK with the wizard.
Pete

Allis_Chalmers_Data.doc (102 Kb, 21 downloads)

I agree with the comments concerning the vibration level. I was evidently looking at the wrong plot-horizontal rather than vertical. Those polar plots of correlation are interesting. Can anyone explain the significance of the polar plot shapes? Glad your problem is resolved.

Thanks Testtech,
You know, that's what I'm realy interested in is those correlated plots. I'm trying to learn more about these and I think they provide valuable data if you get good at reading them. Where's Danny Harvey? He's getting good at these.
Those patterns are still there and I'm not sure if they represent normal, or do they indicate a different problem. I would love to learn more about this analysis. Anyone?
Pete