Stator Slot induced Vibration

[mbailey]

I have a 250 hp vertical 12 pole motor that has vibration and noise spikes at 2760 and 2640 hz. One thought is this is RBPF (69 rotor bars) and RBPF - 120. I noticed on the vibration specrum that I have small peaks at 900, 1800, 2700 and 3600 hz. The spikes at 2760 and 2740 could be side bands of the 2700 peak +/- 60 hz. The stator has 90 slots. That's the background. My question is: Could this vibration induced by the rotor/stator slot combination be exciting the natural frequency of the stator and if so, can I calculate the natural frequency of the stator? I've read a little about mode shapes but am not up to speed on this subject. I think that is what I may have in this case. Your thoughts are greatly appreciated.

 

[electricpete]

I can't help you with natural frequency of the stator. But you are right it can have a big impact on the vibration.

How far below 10hz is actual running speed?

If you provide exact frequencies it may be easier to determine the origin.

My guess (as I suspect yours) is that 900, 1800, 2700, 3600 are a little less than those frequencies and multiples of stator slot pass frequency (90 times running speed).

My guess is that 2760 is 4x RPBF (4*69*1X) and 2640 is 4XRBPF-2*LF (4*69*1X - 2*60)

The rotor bar pass pattern as you know is pretty common and usually not a cause for concern. If it changes after working on a machine, it may be related to soft foot.

The stator slot pass frequency is less common. In my opinion MIGHT be an indication of a rotor bar problem or dynamic eccentricity. Have you done a high-resolution vibration spectrum to check for pole-pass frequency (12*<10hz-1x>) and pole pass sidebands around 1x, 2x etc? Current spectrum? Those might help start to confirm or disprove this).

 

[jbartos]

Suggestion: Visit

http://www.shef.ac.uk/eee/emd/research/stephenlongmotor/stephenlongmotor.html

(for stator natural frequency at switched reluctance motors to see the mechanism and any similarities)

http://www.shef.ac.uk/~am/research/nonlinear.html

for:
Nonlinear oscillations of synchronous motors
Email: Professor W D Collins
etc. for more info

 

[vanstoja]

Are the six peaks you have cited the only ones prominent in your frequency amplitude spectrum or are they only the ones you can't identify? I find it difficult to believe that a 4th Rotor Slot Passing (RSP) peak (with normal +/- 2E sidebands)would show up without the 1st, 2nd and 3rd RSP peaks and their 2E sidebands also being prominent. Also the possibility of 1st, 2nd, 3rd and 4th harmonics of Stator Slot Loading (SSL), as Electricpete suggests, is I believe, unusual for an air-cooled motor with rolling element bearings. Is this perhaps a water cooled motor with fluid film bearings? For water-cooled, fluid film bearing induction motors, SSL peaks are generally an hydraulic rather than an electromagnetic source due to watergap pressure perturbations over the slots which may have
recessed stator slot wedges whose top surface is below the
inner diameter of the stator punching stack. In fluid film
bearings, rotor eccentricity in the watergap under unbalance and electromagnetic loading tends to distort both
electromagnetic and gap pressurefield waveforms bringing
up multiple harmonics of both RSP and SSL sources.

Stator core vibration frequencies are difficult to calculate reliably. A simple formula by Alger,P.(1965), &quot;The Nature of Induction Machines&quot;, Gordon & Breach Science Publishers is f=36,700m(m^2-1)h/D^2*SQRT(K(m^2+1) for steel with E=30E06 and rho=0.29 in English units. K is ratio of gross weight of core and teeth to core alone. h is radial depth of stator core behind the slots (inches). D is mean diameter of the stator core (inches). m is half the number of nodes of core flexure. Better estimates of stator core resonances obtain from the experimental studies reported by Verma and Girgis mostly in IEEE transaction
papers.

 

[jbartos]

Suggestion: Reference:
Determination of radial-forces in relation to noise and vibration problems of squirrel-cage induction motors
Verma, S.P. Balan, A.
Dept. of Electr. Eng., Saskatchewan Univ., Saskatoon, Sask., Canada
This paper appears in: Energy Conversion, IEEE Transaction on
On page(s): 404 - 412
June 1994
Volume: 9 Issue: 2
ISSN: 0885-8969
References Cited: 10
CODEN: ITCNE4
INSPEC Accession Number: 4738364 Abstract:
The radial electromagnetic forces in induction motors play an important role in the production of audible noise and vibrations. The magnetic flux pulsations at the iron surfaces produce these radial forces, which act on the stator and rotor structures. An analysis for the calculation of the various field harmonics and radial forces in squirrel cage induction motors is presented in this paper. To verify the validity of the analysis, a squirrel cage induction motor is analyzed. Theoretical and experimental results are presented with a view to determine the actual role played by the air-gap harmonic fields on the radial forces. Also, the effects of loading on the radial forces and the ensuing vibrations are closely examined.
Index Terms:
squirrel cage motors; acoustic noise; vibrations; machine theory; machine testing; stators; rotors; harmonics; electromagnetic fields; magnetic flux; squirrel cage induction motors; radial electromagnetic forces; audible noise; vibrations; magnetic flux pulsations; stator; rotor; loading; air-gap harmonic fields

 

[electricpete]

John (vanstoja)
Good to see you on the motor board. Those are good comments.

I have seen one motor data where 3*rbpf and it's sidebands stood out as the only peak without any significant lower peaks on the lower harmonics of rbpf visible on linear scale. I suspect that if he looks on log scale he will see the lower harmonics of rbpf with similar sidebands.

The 120hz spacing between peaks seems usually a dead giveaway that those particular peaks are part of a rbpf pattern. But perhaps not always.

 

[1x1x1x]

In case like that the first step is to find if the vibration is mechanical or electrical and identify the source of it.
I would recommend that you make a vibration measurment when the motor is disconnected from the electric supply. It is done this way : run the motor at full speed and stop it with an ESTOP (this stop is usually remove the electric source from the drive). If the vibration recording shows vibration it is mechanical and not related to the electric supply. If the vibration does not show up it is electrically generated. In that case you should check the AC current of the motor when the motor is running fully loaded and check the AC current for harmonic distortion.
(this test can be done using an oscilloscope with fft functions.) Based on the harmonic spectrum of the AC current it is possible to narrow the investigation.
If harmonic components like the 3rd or 4th harmonic shows up in the current check with the drive manufacture.

 

[vanstoja]

Here are the Verma-Girgis references on stator resonant frequencies:
Verma,S.P. & Girgis,R.S.(1973), IEEE/PAS-92 #5 Sept.-Oct,pp.1577-1593, Resonant Frequencies of Electrical Machine Stators Having Encased Construction (2 Parts)

Verma,S.P & Girgis,R.S.(1975), IEEE/PAS-94, #6, Nov.-Dec., pp.2151-2159, Considerations in the Choice of Main Dimensions of Stators of Electrical Machines in Relation to Their Vibration Characteristics

Girgis,R.S, & Verma,S.P. (1979), IEEE/PAS-98, pp. 1446-1455, Resonant Frequencies and Vibration Behavior of Stators of Electrical Machines as Affected by Teeth, Windings, Frame and Laminations

Watanabe,S. etal (1983), IEEE/PAS-102, #4, April, pp. 949-956, Natural Frequencies and Vibration Behavior of Motor Stators

 

[electricpete]

Your motor is being discussed at:

http://www.reliability-magazine.com/ubb2000/ubb/Forum2/HTML/002009.html

 

[electricpete]

Hmmm, no matter what I do, eng-tips tries to parse that link.

You'll have to cut/paste it into your browser address bar.

 

[mbailey]

Thanks for the insight. As you found out from the other post, this is the same motor. We have plans to run it on the dyno later this week to get load data. I plan to do some high resolution spectrums to confirm rbpf. This is a vertical motor with vibration data collected on resilient pads and noise spectrums taken in a sound room. We will also take geometery readings on the stator/frame/brackets and verify air gap. At this point we still do not have a good &quot;fix&quot; to eliminate the noise. Is this noise level common for large slow speed motors?
Thanks Mike