Is anyone actually using ultrasound to check/trend motor bearings and other rotating equipment?
If you are, I'd like to hear how you are doing it and what kind of successes you've had.

Thanks, Paul

Hi Paul,
there are lots of organisations using u/s to trend failing bearings - and particularly failing lubrication of bearings. A typical set of decisions might be 8dBuV above base = grease, 16dBuV above base = change soon and 32dBuV = change immediately.
I have personally had some excellent successes using this approach, some of which I documented and reported in a paper presented at PdM2006.
Best Regards,
Tom Murphy

Hi Tom,
Thanks for your reply. I'd like to read the paper you presented. Is there antwhere I can get a copy?

Thanks,
Paul

Paul,

I use ultrasound for machine monitoring along with vibration. I suggest you be very careful about "trending" numerical ultrasound levels. It is possible to get high/increase in ultrasound level and it can be caused by differnt things, including the background. "Trending" is one of many methods of fault detection. I rely on numerical value, range and type of level variation with time, and the audio characteristcs (headphones). I compare levels on same machine (different points and over time/trend) and compare identical/similar machines.

For motor-pump monitoring: 1 contact per bearing, airborne for coupling, and airborne scan of motor for electrical faults [add contact for mechanical seal]
For motor-belts-fan monitoring: 1 contact per bearing, airborne for Vee belts, and airborne scan of motor for electrical faults
For Gearbox monitoring: 1 contact per bearing and 1/more contact on case for gear faults
For Compressor monitoring: add airborne scan for leaks
For Chain Drives: add airborne scan of chain and sprockets
I have also done diesel engine fuel injector and fuel pumps.

I find that it is better to diagnose fault with ultrasound while at machine rather than relying on numbers transferred to a computer with little or no notes for qualification.

Walt

quote:

Originally posted by Tom Murphy:
I have personally had some excellent successes using this approach, some of which I documented and reported in a paper presented at PdM2006.

Tom,

I don't have access to PdM2006 but will be very interested in reading your paper. I am doing something in this regard but did not put together the material. Will be glad to share it later. I'll be very appreciative if you can email it to me (dgluzman at hotmail dot com ) or attach it here.
Regards,
David

If in conjunction to normal vibration parameters you collect either Shock Pulse, PeakVue or Spike Energy, you're essentially covering the u/s bearing task.

Stress Wave technology also does same.

See the paper (first in chronological order = bottom of page) in Articles & Papers on www.vibra-k.com

And DO cover sleeve bearings...

Xo,

I agree with you although I have not read yet the link. Ultrasound is just a more convenient tool for the mechanic while he is lubricating a bearing.

David

Lubricating a bearing has nothing to do with it. What did we have before vibration? Your "finger" like the fender of a car, you place your finger on it and feel the vibration...Ultrasound has been around for awhile, it is an excellent tool to use especially if you do not have vibration analysis, you don't outscource your vibration, slow-speed bearings, and if you do have the capabilities to actual use it in conjunction with your vibration analyzer or record a wave file using ultrasound and then run it through wave-form analysis, you then may find something worth knowing. But, to think it is just another tool to use when lubricating....come on...
Later

There are several factors to consider when using ultrasound to monitor motor bearings and other rotational equipment.
1. Change in intensity of ultrasound - RMS or dB (relative, logarithmic scale)
2. Spikes in otherwise quiet, smooth sound
3. Smooth vs. fluttery
etc.
As you can see, many of the characteristics detectable using ultrasound are not really quantifiable .

A good ultrasonic device, when used properly on a routine basis will provide good clarity of signal, good sensitivity, the ability to distinguish between different sounds (background or bearing), and ability to adjust sensitivity adequately to eliminate interfering ultrasound and white noise.
There is no red light/green light. Using a good ultrasonic sensor, however, provides the ability to easily identify subtle to gross changes in the ultrasound produced by the component under test. A little experience, signal analysis, and comparison of data gives weight to any call to action.

Hi all, i'm new to all this. I use vibration and shockpulse, but use ultrasound as a further tool. i use it to listen to a suspect bearing, a good bearing being like white noise, a good bearing needing lubrication like very loud white noise and a worn or damaged bearing sounds like a bag of spanners. What I am saying is that I use it as another method to proove to the non believers that there is a problem, easier than showing them a spectrum because they can hear it.

I thought somebody said we can ensure proper greasing of motor bearings using ultrasonic. Of course, there are many other applications.

Walt:

I too am new to Ultrasound and am excited to hear of your reference of use to drive couplings.

I realize that it is a broad question but could you focus on and discuss the use of US on drive couplings in more detail? For example, what sounds indicate problems vs what sound indicates OK? What other parameters are useful and how may one trend?

I realize that what I ask is difficult to quantify as it may be impossible to describe what a sound sounds like but please give it a shot!

Thanks,
Gary Forsythe

Hello,

The company I am working for is implementing a CBM program covering our conveyor motors and gearboxes. We have 1700 motors and gearboxes. We are doing a 100 motor and gearbox test using a SKF Marlin unit and a U/E Ultra sonic gun. Right now I am walking the route once weekly for the next month taking three readings 2 on the motor and one on the gearbox. We are trying to get trending data. So far it is going well. Much easier doing Ultra sonic than doing these smaller components with our vibration equipment. As for greasing motors with the Ultrasonic it is very effective.

Gary,

The basic principle is that coupling (flexible type) faults will cause excessive friction or impacting (or both) that can be detected by airborne ultrsound. Typical coupling faults include:
1) Shaft misalignment (offset and angular)
2) Shaft ends or element clearance insufficient
3) Internal worn or damaged components
4) Loose parts
5) Poor lubrication

Ultrasound is measured near coupling's assemby joint as close as reasonably and safely possible. The overall level (dB) is measured and audio characteristcs are noted. Friction has no real character and may sound like White Noise or Static or whoosh or (__??). Impacting has a distinctive sound like Clicking or Popping. Be aware that there are background ultrasounds that can mask the coupling sound, such as:
a) Machine bearing fault
b) Machine shaft seal rub or leak
c) Pump Cavitation
d) Compressor leak
e) Direct or reflected sound from another machine or source
f) Coupling's windage from spinning bolts or protruding surfaces or holes

A good consistent measurement is not necessarily a fixed measurement point, but one or more locations that may be needed to optimize receiving the sound from the coupling.

I have used ultrsound on several types of couplings including: elastomer element, metal grid, gear, and metal disk-pack.

Try it , you'll like it!

Walt