Sign up Calendar Latest Topics Donate
 
 
 


Reply
  Author   Comment  
Curran919

Avatar / Picture

Sr. Member / Supporter
Registered:
Posts: 599
Reply with quote  #1 
I have a machine (development) on experimental journal bearings that is showing wear. To detect oil-film failure, I would generally look at eddy current probes, but these are not possible to install on the machine. Does anyone know if it is possible to detect hydrodynamic bearing contact with accelerometers on the casing? Maybe something similar to asperity/Smith shock detection on gears?
arie mol

Avatar / Picture

Member
Registered:
Posts: 30
Reply with quote  #2 
Do it the electrical way: Why don't you try a silver-graphite brush shaft raider. Apply some dc voltage via high ohmic series resistor on brush. A voltage dip on brush indicates metal-to-metal contact between rotary and stationary components.
If there is already a voltage on the shaft (electrical machine, steam turbine) then a shaft voltage dip indicates metal-to-metal contact.
Curran919

Avatar / Picture

Sr. Member / Supporter
Registered:
Posts: 599
Reply with quote  #3 
Thanks Arie,

I was thinking about doing it electrically somehow. I'll mention this, but if they are not willing to install a proximity probe, I don't imagine they are willing to install a contact brush.

Come to think of it, they didn't mention it in the last meeting, but from a previous conversation, the bearing may even be non-metallic... I guess that would be important in any case.
OLi

Sr. Member
Registered:
Posts: 1,918
Reply with quote  #4 
If you have a material that will give hi frequency noise on contact so bearing is not completely Teflon or such absolute sensors can be used. I have seen it on a small steam turbine where a standard bearing condition parameter gave a increased value and trending and when opening the bearing it was blue in spots and after "adjusting" the value went back to a low value. Measurement was included by "doing the same everywhere" not by thinking. Measurement was basic RMS sum of everything above 600 Hz. Have fun. Olov
__________________
Good Vibrations since early 1950's, first patented vibrometer 1956 in the US.
http://www.vtab.se
Walt Strong

Sr. Member
Registered:
Posts: 889
Reply with quote  #5 
I would measure multiple vibration parameters during a controlled test so that the best parameter or combination of parameters can be utilized in the future. My first choice would be PeakVue or some form of demodulated spectrum, since it sensitive to friction noise caused by poor lubrication. You could add one or more high-frequency acceleration bands as suggested by Olov. I also would include measuring 1X shaft speed and harmonics and sub synchronous frequencies as indications of rubbing and changes in support stiffness. The loss of lubricating oil film should increase the bearing support stiffness so that the same dynamic forces of unbalance and misalignment could cause higher vibrations on the housing. I've also had good success detecting poor lubrication with ultrasound measurements by using a SDT 150 ultrasound meter with contact sensor. Any method of high-frequency detection requires good structural contact between the inner bearing housing and the outer bearing housing where the sensor is mounted. I think you can be successful, but you might have to think outside the box of normal measurements!

Walt
OLi

Sr. Member
Registered:
Posts: 1,918
Reply with quote  #6 
I happened naturally to use a moving coil velocity transducer [cool]. I don't think it made a difference but anyway. So in that case it was HF velocity... Olov
__________________
Good Vibrations since early 1950's, first patented vibrometer 1956 in the US.
http://www.vtab.se
Curran919

Avatar / Picture

Sr. Member / Supporter
Registered:
Posts: 599
Reply with quote  #7 
Quote:
Originally Posted by Walt Strong
I would measure multiple vibration parameters during a controlled test so that the best parameter or combination of parameters can be utilized in the future.


You mean... I can't tell everything about a machine's condition from RMS and crest factor? And here I thought that any signal above 1kHz was just noise artefacts from having super worn dual rail magnets.

I knew we had the machine learning matlab toolbox for a reason...

Quote:
Originally Posted by Walt Strong
...using a SDT 150 ultrasound meter with contact sensor.


The SDT webpage is all razzle dazzle and salesmen-speak. I am not familiar with ultrasonic meters in general, and I don't understand what you mean with contact sensor. Can you point me somewhere with some foundational info?

Quote:
Originally Posted by Olov
Measurement was included by "doing the same everywhere" not by thinking.


I'm not sure what you mean here. Did you just throw a bunch of parameters at the data and see what correlated? I guess I need some control data.



I'm a young guy self-teaching myself through this world of vibration. This forum has been a wave of new information so far...
OLi

Sr. Member
Registered:
Posts: 1,918
Reply with quote  #8 
No the guy that made the measurements used a ready made vibrometer with the magic bearing condition L-method as described inside it and he used it on all bearings regardless if it was a fan, pump or steam turbine and one day he got a elevated reading on the turbine.... (also). So normally you get a noise from metal-metal contact and most general bearing condition methods can pick that up and the type of bearing and level of deterioration is only influencing the levels and trending maybe but you may also get it from seals touching, couplings with low lube etc. etc. Special for shaft touching is that you may get a varying level, amplitude towards time if the touching get in to a cyclic mode by touching spot heating shaft, shaft bending and cooling and touching again in the simple description so you can get a sinusoidal trend towards time, sometimes and maybe in cooperation of being close to a shaft resonance as it is where I seen it. That is then in the regular vibration range RMS sum of 10-1000Hz. Have fun your machine may have some other tell tale. Olov  
__________________
Good Vibrations since early 1950's, first patented vibrometer 1956 in the US.
http://www.vtab.se
Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.