Sidebands vibration 1

[AlbertFM]

Hello all,

I have a problem with an electrical motor.

Two sidebands are apperaing next to syncronous peak and its harmonics (700Hz) while I am rumping up my electrical motor (1500 Hz nominal speed). A strange noise can be heard when sidebands appears.

I've been looking for some information of the problem:

Motor; with loose or otherwise defective rotor bars
Primary frequency is 1 x rpm. (With one or more sidebands on each side.) The difference frequency between a sideband and the primary frequency is equal to the rotor slip times the number of poles. Normally, its amplitude and amplitudes of sidebands are proportional to load.
The symptom for this defect at times produces excessive vibration at relatively high frequency. Instead of a sharp peak at rpm x number of rotor bars, there are many sideband peaks, producing what is usually referred to as a "haystack." The difference frequency between sidebands is usually equal to the electrical hum frequency.

I would appreciate if someone of you could suggest me other possible reasons of sidebands vibrations as well as an small explanation of what means the description I founded in internet.


You can find attached a spectrum. Speed is 700 Hz.

 

[electricpete]

"1500 Hz nominal speed"
"Speed is 700 Hz."
Both of these sound very high. Perhaps 70hz?

Since you ramped it up, I assume this is a vfd?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[AlbertFM]

Thanks electricpete but it seems that I've solved the problem by adding an o-ring between two components.

Yes, I was talking about 1500 Hz or 90.000 rpm.

 

[FeX32]

Your problem solution makes sense. I believe sidebands are commonly caused by mechanical compliance gaps (or looseness). You could be sure by measuring the response in more then one plane.


Fe

 

[AlbertFM]

It's strange because in all literature says that loosness and gaps appears at 0.5x, 1x, 1.5x, 2x...but anyway it has worked. However when I go down I still can heard a small sound when I pass through the frequency where sidebands appeared.

Refering to the measures in different planes I can manage easily to do at 90º but I'm not sure about what I will see. The contact with the two surfaces that made sidebands appear was cylindrical so I could read it in any radial direction.

 

[GregLocock]

I agree, sidebands are not typical symptoms of clearances and lash.

Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[FeX32]

Yes I lost my bearings. Actually I was not totally correct. In fact the solution does not totally make sense as sidebands usually are characterized by an imbalance modulated bearing defect within the machine.
Also, briefly looking at your spectrum, the amplitudes of the 2xfr are higher then that of the fundamental amplitudes. This could be in indication of an imbalance.
Measuring in the axial direction vibration could rule this out.
Although, knowing nothing about your machine makes it more difficult to analyze.
But you seem to have it under control.



Fe

 

[AlbertFM]

FeX32,

I know that the solution doesn not make sense but in fact what I have done is put some damping between parts. I have not killed the problem I have just smooth its effects. Ramping up further (1kHz) bands have appeared again.

Now I am working to reduce imbalance of the rotor since I also consider that vibrations are too high all along the coast up. I don't know if it will also solve sidebands definetly. The problem is that I am balancing at 4.000 rpm and it has nothing compared to 90.000 rpm.

Refered to 2xRPM, when I look in mm/s it does not seem as high as it seems in g's. Frequency is really high. However I agree that it could also confirm imbalance.

Please can you explain me what do you mean with

"imbalance modulated bearing defect"

May be a "mix" between 1xRPM and bearing frequencies?

Bearings were new, I do not expect any defect...

Thanks a lot

 

[FeX32]

Albert,
From my experience I know that it is common for an imbalance to "amplitude modulate" the bearing frequencies, which can produce sidebands (because of the bearing defect).
However, you mention that the bearings are new so it may or may not be the case.
Good luck,

Fe

 

[electricpete]

A typical rolling bearing defect pattern would be running speed sidebands around a non-synchronous series such as harmonics of BPFI. So expressed in orders (multiples of running speed), the pattern might be something like:
4.81 +/- 1
9.62 +/- 1
etc

This does not fit that type pattern at all. The running speed harmonics are center frequencies and the non-synch frequencies are the sidebands. Somethig like:
1 +/- 0.13
2 +/- 0.13
etc

I feel very confident to say this spectrum does not represent a rolling bearing defect pattern.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

I've been looking for some information of the problem:

Motor; with loose or otherwise defective rotor bars
Primary frequency is 1 x rpm. (With one or more sidebands on each side.) The difference frequency between a sideband and the primary frequency is equal to the rotor slip times the number of poles. Normally, its amplitude and amplitudes of sidebands are proportional to load.
The symptom for this defect at times produces excessive vibration at relatively high frequency. Instead of a sharp peak at rpm x number of rotor bars, there are many sideband peaks, producing what is usually referred to as a "haystack." The difference frequency between sidebands is usually equal to the electrical hum frequency.

I missed that. The sideband spacing you mentioned is called pole pass frequency. If present (*), is a very big clue that the source is related to the motor. It represents a rotor assymetry which rotates at 1x. It could be:
1 - Dynamic eccentricity. Rotor is moving at 1x within the airgap. This could simple be the case that your rotor is assuming a bow at the speed you operate it and the bow is an appreciable fraction of the airgap (let's say more than 10%).
2 - Rotor conductor assymetry such as casting porosity in cast aluminum rotor or bar-to-endring joint degradation in fabricated rotor.
For normal (lower speed motors), current signature analysis is often used to help distinguish the two. Look for sidebands of pole pass frequency around line frequency. The higher the magnitude relative to center peak, the more likely #2. If the sidebands are less than 1/400 of the main peak, it is not a severe case of 2. If sidebands are higher then likely 2 is the case. These thumbrule levels again are based for much smaller motors.

I tend to think you have a small motor and rotor degradation is not a concern. At these high speeds I think you are probably seeing the rotor assume a bow as discussed in 1. As long as you don't have a rub, it is not a really big problem. Since the magntidue of sidebands is small compared to the magnitude of running speed harmonics, it seems worthwhile to focus on reducing running speed harmonics.

How to address those running speed harmonics: Balancing you mentioned is a good start. You might also look for looseness in stationary or rotating components.

* "If present" - this assumes you confirm the spacing of the sidebands corresponds to pole pass frequency. It's not clear whether you have done this or not.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[AlbertFM]

Thanks a lot for your response electricpete.

I started by balancing the rotor and left a small residual imbalance. Now it works much more better. Harmonics are smaller (almost have disappeared). Referring to sidebands they're small and appears only in a few revolutions.

I will try to check your tips to optimize the system.

Thanks for the information.