Acoustics replacing vibration analysis???...

[sasuan]

I am assigned to evaluate a new program at our plant. Use of Acoustics for detecting vibration, over/under lubrication and arcing in transformers. We have a sound vibration analysis program for all rotating equipments already set in. Does the acoustics /ultrasound technology help any better than the vibration analysis?? Also, how useful is this technology in detecting lubrication problems and arcing in transformers. Thanks in advance.

 

[micjk]

Ultrasonics is a very useful "ADDITION" to your current program. I use it for leak detection also. As far as replacing vibe analysis I'd say no. It is also good for detecting arcing in transformers, and detecting corona using a parabolic dish. Lube issues can be detected with vibe just as well as with ultrasonics. You can also use untrasonics for steam traps. I would suggest using ultrasonics to complement your PDM program.

Roy Gariepy
Maintenance and Reliability Dept.
Dorlastan Fibers LLC.
Goose Creek, South Carolina USA

 

[GregLocock]

That's funny. In the acoustics world we are quite excited by using acceleration measurements instead of microphones for some work!

Cheers

Greg Locock

 

[svanels]

I agree with micjk that airborne ultrasound detection is rather an addiction then a replacement.
We use it for steam trap surveys and leak detection mostly. The problem with ultrasound is that it depends of the "listening experience" of the user. In most cases it is a portable tool where you make use of a handheld contact probe. For data aquisition you need a pick-up point that is not influenced by human errors, especially if you want to trend and compare data. It is a great tool to localize a hard to find leak in radiator, scanning compressed air connections or detect trap mal-functioning, but I would not use it to replace vibration based conditioning monitoring.

Steven van Els
SAvanEls@cq-link.sr

 

[Razor423]

I agree with the PDM program being a solid foundation for plant reliabilty. As far as acoustics, it is a great addition to our program also. It is used to help with inleakage detection in our condensor, check steam traps, and even mechanical deficiencies through several pumps have been located with the use of trend data from acoustics. As a result of good acoustic data, we were able to locate and track a leak in our boiler feed pump casing. There are also several products on the market that use acoustics by means of attaching it to your grease gun and one can monitor the bearing sound and determine the need for grease. These products only enhance a good PDM program and as a result your facility becomes a world class enviroment.

 

[electricpete]

On the topic of using ultrasound to determine greasing practices...

Using an ultrasound thing to determine how much grease grease your bearings is a somewhat new concept, heavily pushed by the ultrasound equipment vendors, but to my knowledge not embraced by any rotating equipment vendors, bearing vendors, or large industrial organizations. I was at a recent meeting including approximately 60 engineers from perhaps 30 nuclear plants. We had a presentation from an enthusiastic ultrasound vendor. He finished by asking how many of us used it. Zero hands raised. Everyone has heard of it but no-one among that group yet convinced it is the right way to go.

The basic reason is that controlled quantity at controlled interval is a simple time-proven method. If you follow that method from the beginning of the equipment life you know how much grease is in the cavity. If you use the ultrasound gun you will spend a lot more time on each lub task and you will never again know how much grease is in your cavity. In this respect use of the condition-based greasing might even present a bigger potential for overgreasing... which is the thing we are supposed to spend all this extra time to prevent. That may or may not be true. My biggest reason for not buying one is that as I mentioned no bearing vendors or rotating equipment vendors have recommended it to us (they recommend time-based controlled quantity) and no large industrail users have been using it to my knowedge.

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

 

[vanstoja]

A useful practice in machinery diagnostics is to combine accelerometer-acquired vibration signatures on a noise analyzer screen with their ear-audible sounds from an audio amplifier picking up the same signal. To do this you need to magnetically tape the accelerometer signals and then play them back through the noise analyzer. This used to be a rather routine noise diagnostic practice, but since the development of portable, digital noise analyzers, is now probably a very rare practice. The ear-audible sounds are actually structureborne, not airborne, signals that are transmitted through the machine and which may be totally inaudible to the ear situated close to the machine. The audio amplification of structureborne sounds is akin to the ancient practice of putting your ear to a metal rod pushed against the machine's outer wall. If impacting is occurring inside the machine, the audible sounds can be correlated visually with spectal frequency-amplitude "bursts" to determine the frequencies being excited by the transient impact, possibly identifying the location of the parts involved. For special cases of daunting transient noise problems(squeaks,squeals,whistles,chirps,etc.), it may still be worth the extra effort to tape the signals from the accelerometer and listen/look at them concurrently for a better understanding of the problem cause and/or location.

 

[micjk]

Although I don't record the sounds, I use a headphone set with my data collector and accel. So Greg I guess I do use a combination of vibe and acoustics .

Roy Gariepy
Maintenance and Reliability Dept.
Dorlastan Fibers LLC.
Goose Creek, South Carolina USA

 

[GregLocock]

There's a posh word for listening to vibration signals, but I can't remember it.

It is an important technique in my opinion, very good for deciding if the recording is a good quality one, and whether any transients are spurious or not.



Cheers

Greg Locock

 

[McVibe]

http://www.dliengineering.com/downloads/sound+vibe.pdf

http://www.dliengineering.com

The first link is to an application note that describes how to use Ultrasound along with vibration to improve your lubrication regimen. I apologize that it is not "generic" - but the concepts are. The secod one is to our website where you'll find alot of free and interesting technical info.

We have quite a few people using ultrasound along with vibration to increase bearing life (as described in the document). Most seem fairly pleased with it.

my take on the original question: Ultrasound complements vibration.

 

[ronfrend]

If I may add an additional comment.

I agree with the majority of what has been said here and ultrasonics is a useful addition to a vibration based system. However these units can be bought very cheaply (sometimes less than $400) and are an excellent trouble shooting tool for the craftsmen.

I was also involved in a system where we had to monitor the vibration of fan bearings for over 100 large fans in a car plant paint facility. We couldn't open the door to enter the fan room as this would dislodge dirt and damage the coating below. The bearings were fed grease from remote (time-based) lubricators outside the fan rooms through copper tubes about 3 metres long. We used the ultrasonic probe on the grease pipe next to the lube unit and hooked up the ultrasound gun to a standard FFT data collector. After a a lot of tests we found that we got a pretty good correlation between the vibration and the ulrasound SO LONG AS WE USED A DEMODULATION TECHNIQUE.

The technique was not quite as good as regular vibration but it worked very well and is still in use today, 9 years later, with an excellent track record.

Ron Frend

http://www.predicon.net

 

[sasuan]

Firstly, Thanks everybody for the suggestions.
I started working with each of the equipments regarding the bearing vibration and the lube issues.
As i said, I am trying to set a new program with this technology in our plant. Can anybody suggest how to go ahead with it. I have a long list of rotating equipments. I am trying to do comparative process between similar bearings and get historical data. Is there any better way of setting up this program...???

 

[sasuan]

RAzor 423 has mentioned up here that they have used the acoustics technology for detecting condenser leaks. Did you mean the tube leaks or the vacuum leaks on the valves flanges, and the fwhtr penetrations around the condenser. We have a huge condenser for which we do Helium leak test to check air leaks. It's pretty good, but time consuming. (It takes almost 3-4 days and 2 guys to test all the way from turbine deck down to vacuum pumps.) Does this ultrasound help in that way at all. Please let me know.

 

[goodvibes]

Hello,

My experience with equipment vibration/health monitoring using acoustics is that it is not effective if the equipment you are monitoring is in a high noise environment. We had a company try to predict bearing failures on a high speed aircraft gearbox using what they call "stress wave analysis" and it failed miserably. I understand that the FAA uses their equipment for bearing fault detection/PM on their radar dish gearboxes (very low speed low noise application), and it works well.

Good luck.

 

[struan11]

Hi,

like the majority of your replys I would suggest that Ultrasound should be used in conjunction with vibration and other types of CM techniques.

The best way to monitor the condition of Rolling element bearings is without doubt using vibration analysis, however if budget is a problem them Ultrasound would be a good way to go. If you are looking at going down this route seriuosly I would suggest that you contact UE Systems,

http://www.uesystems.com

, as they offer free downloads of their monitoring software which can be used with their entire range of instruments, and allows you to track overall levels of noise, and assign alert and alarm bands just like overall vibration levels. They also offer worldwide training so getting them to you site would be no problem.

Ultrasound however is the best way to detect partial discharge from any electrical machine running above 1000V, and I and my collegues have had success in detecting this at early stages in the UK sector of the North Sea. Again this can be used in conjunction with Thermography to great advantage in an electrical context, so I'd also consider looking at that too if the budget exists.

Ultrasound is also a well proven method of detecting both pressure and Vacuum leaks, and can be done whilst a unit is on load as the sound is highly directional and can be homed in on much faster than using a helium leak detector. Your 4 day survey could be quickly and easily carried out simply by boxing up the unit to be tested and turning on the vacuum system then walking around it and listening for leaks. No doubt the cost of a decent Ultraprobe Unit would be saved in the manning and helium costs for this. I have used this on a 660MW steam turbine condenser both inside and outside, it only took a day to find two leaks outside at the end of an outage. It is capable of finding very small leaks so it just depends how far down you want to go. I have also used it to detect condenser tube leaks. This was carried out with a condenser bank isolated with the unit running on reduced load, draining the water out of the pass climbing inside and quickly scanning the ends of the tube plate. This finds the leaking tube within minutes, a check at the other end of the tube shows if you are correct.

So yep go ahead and set up the system.

Cheers
John

John Wilson
Condition Monitoring Consultant
Clyde Platform
UK Sector North Sea.

 

[lilliput1]

Some mechanics use a screwdriver to put against a bearing & the other end to the ear.