End Of Line Vibration Testing 3

[das712]

I am interested in end of line vibration analyis? I work for a manufacturer of electric ac motors. I would like to deploy this kind of testing to eliminate noisy motors from the production line and also determine root cause of defect. Can anyone help?

 

[GregLocock]

In the mechanical world lots of people do this for lots of systems. I think some alternator manufacturers do it as well.

Does the customer have a problem with noise or vibration? If you can use a microphone then it might simplify your setup. Can you measure your typical problems via the base plate, or do you need an accelerometer actually on the motor itself? If so your repeatability might suffer.

What do you need help with?

Cheers

Greg Locock

 

[electricpete]

what does "end-of-line" mean? is it simply referring to the fact that motors are tested when they they reach the end of the production line?

 

[dgallagher]

Feel free to contact me directly as we have made many test stands and acceptance testing systems.

http://www.stiweb.com

 

[das712]

All,
Thank you for your quick responses. I really appreciate the time you took to answer my question.

GregLocock,
The customer is having a problem with noisy motors. The application the motors are used in is noise sensitive. I agree that a microphone would be easier to use than a accelerometer, however the environment really does not permit this. The test booth that the motors are tested in is "semi" enclosed and designed to eliminate some of the noise from the production floor. This has also been tried with little success.

Unfortunately, I think an accelerometer will have to touch the motor itself. This could be difficult, considering the number of motor models that are produced on this line. That means the fixturing will be more complicated, or each model will need its own. What do you know about laser vibrometers? I was thinking maybe a generic fixture that could be adjusted for a variety of motors, and then a laser that does not have to actually touch the motor, but may be positioned away from the unit instead.

electricpete,
When I used the term "end-of-line" I was just referring to testing at the end of the production line - 100% of all units.

dgallagher,
Thanks, I will look at the website.

 

[electricpete]

Here is one comment that doesn't address your question, but may need to be considered:

NEMA allows vibration testing of motors either ridigly bolted down or on a resilient flexible base (rubber pads).

Most small motors are tested on the rubber pads since it's much easier.

But vibration can be substantially higher when bolted down (as installed) due to frame distortion affecting the airgap and/or frame resonsnce. In particular the 2*LF vib and k*rotor bar pass frequency +/-n*2*LF pattern of electromagnetic vibrations are worse when the machine is bolted down. I suspect that the high frequency rotor bar pass frequency pattern vib may contribute to audible noise. (although internal fans more likely are the dominant source of audible noise)

 

[das712]

electricpete,

Thank you, I am looking for all considerations. I am new to this feild and have a huge learning curve to overcome.

 

[electricpete]

Back to the microphone issue... if noise (vs vibration) is what you're interested in, and background noise poses a measurement obstacle, then there are directional microphones which will reduce the background noise significantly.

 

[das712]

We have tried directional microphones before with limited results. I was not part of that project, but have recently seen some impressive results with vibration analysis. For instance, a signature was developed for a model using vibration analysis. This was developed with known good and bad units. Each unit was then run, tested, and compared to the signature. Part of the signature also showed common faults at certain frequencies. This was in a lab environment, and we are now interested in deploying this same analysis in production.

I suppose the microphone is still an option, but everyone is excited about giving the vibration analysis a try.

 

[GregLocock]

Laser vibration sensors can work - but I suggest you develop a method using conventional accelerometers, which are relatively cheap, and then get the guy selling you the laser vibrometer to come in and demo it and lend you it for a week.

Things you need to worry about - what frequency range do you need? Laser vibrometers have fixed frequency and displacement capabilities. for instance the first B&K laser vibrometer only worked up to 400 Hz. Also they are buggers to mount - obviously any floor vibration gets picked up by the laser.

If your motors have steel casings you may be able to use a magnetic base on the accelerometer. Alternatively get your machine shop to make up a quick release clamp for each motor type - this will test your patience and their ingenuity.

Roughly how many motors do you need to test per day?
Cheers

Greg Locock

 

[vibegeek]

take a look here, there are some very cost effective automated solutions available

http://www.processvibration.com/