Hi vibration problem of AC Motor

[PAP]

Dear all
Anyone can suggest about the problem of high vibration at high frequency (at aout 911 Hz ,Motor 50 HZ , 1000rpm) ,what are the possible causes of this problem ?. In vibration theory ,the cause may come from the stator or rotor defects , so how can we check the internal stator or rotor defect?
Your sincerely
PAP.

 

[electricpete]

A very common pattern is RBPF with 2*LF sidebands. You can recognize that by looking for several equally-spaced peaks spaced at 2*LF. That is a very common vibration pattern, possibly associated with soft foot or misalignment, in my opinion not a problem unless it is increasing.

If you wanted to check for rotor bar problems (I see no reason to suspect them yet), look for pole pass sidebands around 1x and harmonics. Then confirm with a current signature test.

It's also possible that you have a bearing problem or some type of impacting which is exciting the natural frequency of the bearing.

To narrow it down, it will help if you can identify the larger pattern to which this peak belongs. Is there a series of harmonics of a non-sync peak? Are there sidebands spaced at 1x? etc.

Also time waveform can tell you a little. Bearing or impacting is obvious impacting spike and ringdown. Most electrical patterns do not resemble impacting at all, they are smooth modulated patterns.

 

[mdonner]

I think you just overspeed the motor considerably beyond its top speed limit. Consult motor data sheet. At 911 Hz the motor is running at about 18,000 RPM ?!!!
The rotor is not balanced for values beyond guaranteed top speed, and bearings may overheat (this will also deteriorate lube material).

 

[electricpete]

md - 911hz was the frequency of the measured vibration, not the speed of rotation.

 

[GusD]

HI PAP.
You didn't say if your HI -vibs are IN/SEC /Gs/or MILLS
I am assuming that it is high for whatever unit you used.
I would rule out a bearing Freq (BPFO/BPFI/BSF/or FTFs. The 911 Hz=45550 Cpms.Bearing defect frequencies are very seldom higher than 300 Hz=15000 cpms.
As for your rotor stator problems I would look for the 2*LF with pole pass side bands,or if you know the # of Rotor Bars you could look for a Freq around 1000 RPM * Rotor bars.Amplitudes of the sidebands in relation to the Fundamental Freq. gives you an indication of the severity of the problem.If you did not experience any trips or a lack of torque I would look for other causes.
900 HZ is about 18* Run speed.Could you have a FAn with 18 blades or something.A gearbox with a input pinion of 18 teeth.There are a lot of other reasons for the 911 Hz.
If you would like to do a Motor Current Signature Analyses
you need some special equipment.Good Luck
Let us know what you find GusD

 

[electricpete]

Good comments by Gus. I would suggest that bearing fault cannot be ruled out based on the frequency range. Typically bearing fault spectra occurs in harmonics of the fault frequency... the harmonics peak at the bearing resonance often in the range 30kcpm-150kcpm.

 

[electricpete]

Not to beat a dead horse... . I just got a new web server and I am eager to try it out.

Here is an example outer race defect on a 6308 bearing. The highest frequency content is ~ 100kcpm (g's scale).

http://electricpete.freeservers.com/OuterRaceFault.jpg

I cut open the bearing for inspection and found clear false brinneling damage on the outer race.

 

[electricpete]

Whoops. Don't bother clicking on the above, the link won't work. I'm pretty sure we have all seen similar spectra anyway. If anyone want the file, email me at the address in my profile.

 

[GusD]

Hi electricpete.
The BEARING HI-FREQ RINGING AT ITS OWN NATURAL FREQ it is indeed in the ranges as you mentioned.Theorectically, the
1st failure mode on any bearing would happen at this levels of vibration 30KCPm to 150Kcpm.These would be a sub-service defect that in all likelyhood would not be visible to the naked eye.The Hi vibration level at 911Hz as reported by PAP, just seems too high for a natural bearing resonant condition.Unless it is being excited by something else.Havng said that,in this field,nothing can be ruled out.Everything is possible.

Thanks Elecpete. GusD

 

[electricpete]

Good comments.

I will mention that 915hz is 55kcpm..... not even above the middle of the range 30kcpm-150cpm mentioned. The spectrum that I posted has highest content at 100kcpm and the defects are most definitely plainly visible. I have spectra and photos available. I have at least one other example of similar behavior. I do agree in general we expect the frequency content to move toward lower frequencies as the fault progresses.

 

[jbartos]

Suggestion: Is there a VFD? if yes, visit

http://www.transcoil.com/

for:
dV/dT and Reflective Wave unsuppressed causes Motor failures
Peak Voltages on a 460V system can reach 1200 to 1600 V, causing rapid breakdown of motor insulation, leading to motor failure. On 575 V systems, the peak voltages can easily reach 2100 V. If this is left uncontrolled, insulation failure may occur. The same peak voltages that damage the motor can also damage the cable. Since the KLC and KLCUL filters are designed to be placed at the output of the Drive, these units will also protect the cable runs.
The added inductance of a KLC or KLCUL filter will also help reduce motor heating, motor noise, and motor vibration by reducing the current harmonics in the system.

 

[PAP]

Hi all
Thank you very much for your useful comments: and I have any questions from your suggestion:
For Electricpete :what's the Motor Current Signature Analyses,How can I measure it ?
For GuSD :- I've calculated the WBPF for the stator slot (Stator slot=54 ,54*1000*60 is about 900 Hz ) ,Is this the problem may come from stator side?
- in your first comment "Amplitudes of the sidebands in relation to the Fundamental Freq. gives you an indication of the severity of the problem" :what's amplitude of the sidebands come from? and what does it mean?
Best regards
PAP

 

[electricpete]

Current signature analysis can usually be done by connecting output of a current clamp-on probe to your data collector. But I think you would be jumping the gun to do motor current signature analsysis. That is only required if you see evidence of pole pass sidebands in vibration spectrum.

If not can you describe in more detail. What is the exact running speed of your motor (49.7hz?). What other peaks do you have in the neighborhood of 911hz? Is there a regular spacing of peaks in this frequency range?

Do you have available horizontal/vertical/axial readings at inboard and outboard positions?

 

[sparky1976]

Electricpete, I cannot open your link, it says not finish yet.

 

[GusD]

Hi Pap

If you don't mind to E-mail some spectra regarding your vibration problem it might be easier to understand what your problem is.My E-mail is dalmeida.gus@syncrude.com

To elaborate further on your question regarding the sidebands.Most discrete peaks showing on any particular
spectrum are usually related to the Fundamental Freqs.That is usually the 1xRPM of the motor.If that motor has stator problems related to excentricity,shorts or others it normally shows a peak at around the 2XLf (120Hz or 100Hz for 50 cyc).If these larger peaks or fundamentals have sidebands around them,depending on their amplitude and # of them, they could indicate other elctric problems such as rotor bars,excent, etc,etc.The deggree of defect is usually related to the amplitude of the sideband as well as the # of sidebands.In a nutshell,the greater the SBpeaks and the greater their # the greater the defect.

Pap a word of caution.I too, do vibration anylises (or try) as well as electric motor cond.monitoring.
We use all the technologies that are available today.Vibration,electric motor testers,Thermography,MCSA
Motor current signature analyses.The # of electrical motor defects (rotor bars,exc,etc)identified by Vibration analyses is small when compaired to the ones identified by motor protection relays.At timestoo late.
If I have a motor, that is not drawing excessive Amps,
develops rated torque and runs at normal temperature,I would look for a problem other than electrical problem with that drive.

GusD

 

[jbartos]

Suggestion: It is also known that the 18-phase motor is causing mechanical vibrations. That could potentially be linked to 900Hz vibration frequency since 18phase x 50Hz is 900Hz.

 

[electricpete]

what 18-phase motor?

 

[jbartos]

Comment: There were some engineered, designed and built. However, preferable number of phases appears to be 15 since the mechanical vibration characteristic to 18 phase motors are not present. The some of these motors are now being implemComment: There were some engineered, designed and built. However, preferable number of phases appears to be 15 since the mechanical vibration characteristic to 18 phase motors are not present. The some of these motors are now being implemented in elevators and NAVY ships since they have very appealing good torque characteristics for elevator operations.
Visit

http://web-cat.cs.vt.edu/PEBB/UC-Hardware/tsld005.htm

http://web-cat.cs.vt.edu/PEBB/controller/Universal_Controller_Review_09-27-01.pdf

Handle complex algorithms and computations
• Field-Oriented Control
• 15-Phase Motor Control

http://www.ship.org.tw/Upload/Electric_Warship_VI.pdf

for:
ALSTOM Advanced Induction Motors have been designed and manufactured at a variety
of ratings and phase numbers since their development for industrial applications in the
early 90’s. The machines are fully compliant with Noise and Vibration, EMC/EMI and
Shock requirements in accordance with the appropriate NES or DEFSTAN. Typical
parameters for an H-bridge, PWM inverter fed, 15-phase motor for delivery in 2004 are
shown in Table 3 and the construction of a typical advanced induction motor is shown in
Figure 2.
etc. for more info

 

[jbartos]

Suggestion: Visit

http://www.chorusmotors.gi/technology/animations.shtml

http://www.chorusmotors.gi/technology/motortutorial1.pdf

http://www.chorusmotors.gi/technology/simulator2.shtml

http://www.chorusmotors.gi/technology/simulator.shtml

http://www.chorusmotors.gi/technology/chorus_exp.shtml

http://www.chorusmotors.gi/

http://www.chorusmotors.gi/technology/photos1.shtml

etc. for more info
The problematic 911hz vibration frequency could also be linked to 3rd harmonic x 3rd harmonic x 2nd harmonic = 18th harmonic (electrical)
18 harmonic x 50hz = 900hz electrical that impacts 911hz mechanical vibration. However, this is under an assumption that there is the 3rd harmonic, 9th harmonic and 2nd harmonic content in the motor power supply.

 

[electricpete]

I was e-mailed a copy of the spectrum. Unfortunately my ISP is switching and it's difficult for me to respond directly by email.

The spectrum shows a pattern of RBPF with multiple sidebands at spacing of 2*LF, 4*LF, 6*LF . The highest peak 915hz is probably noy RBPF, but one of the sidebands. That peak has mangnitude approx 1.5 mm/sec.

This RBPF pattern is very common and usually does not indicate any destructive/degrading condition within the motor. Sometimes it is inherent in the motor design, and sometimes it canbe lessened by adjusting soft foot condition or correcting eccentricity.. I believe the poster has disassembled/inspected and checked the gap.

I don't think any more action is required. It should operate reliably. Just continue to monitor.

The well-known Technical Associates by Jim Berry states that this pattern indicates rotor bar problems. I think 99% of vib professionals disagree with that. As long as you don't see pole-pass sidebands around 1x you probably don't have rotor bar problems. In your case you have inspected the rotor also which gives you higher confidence.