Which is the best method to diagnostic bearing faults?

VASTOCEAN,

 I wouldn't  really call myself an 'expert' but have been doing this for a little while and have had a lot of success in finding bearing faults before they failed catastrophically and have saved my company hundreds of thousands of dollars.  And I have received a lot of help from people on this forum.

I guess the first question you need to ask is how early is early detection for you and your org?  What is your industry?  We can find faults in bearings using vibration analysis and enveloped acceleration (SKF version of Shock Pulse) that truthfully could operate worry free for a year or more.  We use SKF CMXA 70's(with @ptitude Analyst software), oil analysis, and thermal imaging together and have had incredible success.  There are detection methods out there that can detect the most minute fault in a bearing, electrical issues, lubrication issues.... you name it.  I am in the mining industry so many of our bearings are on a 'run to failure' strategy, because we have many redundant systems. Not all of them..... but a good portion.  The main goal is to have the information ahead of time to have parts on site when the failure occurs to minimize the scenario where the back-up equipment is running while the primary is being repaired.

Long answer, I know... But something to keep in mind.  For my money the SKF micrologs and software is an excellent tool, not to mention SKF is a legit company that has great customer support, industry respect, and a TON of educational materials available to the user.  

Hope that helps a little.

You really need to define what type of bearing and the speed before I could give any meaningful answer to your question.  If you are asking about Conrad style roller bearings running at 1800 rpm, equipment that uses the vibration spectrum for identifying bearing faults cant be beat. 

As answered above by our colleagues, depends of which operational parameters the bearing shall withstand for, temperature, load, speed, etc.

Wise advice is that despite which it is the bearing work environment, once you detect a warning, an indicator, the bearing is done. 

Following this approach, you can monitor the bearing health by PeakVue technology or cheaper solutions as 4-20mA sensors, IMI bearing fault detector sensor and carefully set the alarms limits. Once that trend reaches the limits, a FFT analysis shall be performed with CSI/Emerson or SPM, etc (there are many choices in the market) to know what is exactly happening in the bearing and for how long can run without any catastrophic failure. This analysis will help you to know which could be the causes of the failure.

The key to monitoring bearings is trending data. Shock pulse, gSpike, Enveloping spectrum all see the same stuff. Time wave sees the same stuff. Vibration specta software will work with the right Fmax and alarm levels and almost all vendors have some type of high frequency enveloping. Vibration with a permanent mounted accelerometer in the location of interest is the best method for collecting good quality data, the next best would be a magnet mount at the same location every time. The direction horizontal, vertical or axial to the bearing is another consideration. Structural stiffness, belt drives and coupling types all figure into the analysis. The key is to collect a baseline and repeat same parameters, same location, same method of collection and trend that information. Ultrasonic has its place in bearing diagnoses but is hard to trend, use ultrasonic for lubrication and acceptance, remember lubrication can change the dynamics of the bearing. A stethoscope can even be useful for bearing evaluation, if you can hear click get ready for replacement. Good trending starts with a good baseline data, requires repeat-able data collection and results in accurate trending of bearing life.

 

Hi

this is first time I share my opinion in the AMS website, and I hope it helps you

To know what the best method to diagnostic bearing faults we need to identify several information and parameters as follows:

  • What is the used technology, if any (vibration analysis, oil analysis, ultrasound etc), based on this we can choose the technique(s) of diagnosis and detection of faults.
  • Knowing what is the type of the Brg (journal, or anti-friction,) and its lubrication method to determine the type of sensor(s) to be used, the settings of the Brg information in the software & instruments, and the parameters or frequencies to be looking at, followed up.
  • knowing some information and parameters which can help for directing Brg faults, such as Brg temperature, machine speed, the type of the machine and its other components,. This will help to choose the best CM technology, set up the used tools and instrumentation, and eliminate confusion or misdiagnosis during data analysis.

The more accurate and correct information are available, the best method to diagnostic bearing fault can be chosen.

Well, probably your choice about SKF Microlog is a matter of personal experience. Hardly you can compare it with CSI-2140 or with SPM (see http://www.spminstrument.com/M...ues/HD-technologies/) or many others... As well as for you seems the better choice, for me would be the latest for a long list of instruments, at same position that Fluke. Don't make me wrong, all those instruments can give you useful results but once the bearing is damaged, your priority shall be when it will fail and why happened. Early warning would be easy (and cheap) as a 4-20mA sensor

Well, probably your choice about SKF Microlog is a matter of personal experience. Hardly you can compare it with CSI-2140 or with SPM (see http://www.spminstrument.com/M...ues/HD-technologies/) or many others... As well as for you seems the better choice, for me would be the latest for a long list of instruments, at same position that Fluke. Don't make me wrong, all those instruments can give you useful results but once the bearing is damaged, your priority shall be when it will fail and why happened. Early warning would be easy (and cheap) as a 4-20mA sensor

The original poster requested "best methods" for "diagnosis of bearing faults". I didn't see any part of his question that asked "what is your favourite technology brand". Yet the replies are full of SKF this and SPM that or CSI here. Kudos to everyone who resisted the urge to pump or plug their favourite vendor.

OP asked about Diagnosing Bearing Faults. What about Monitoring and observing Conditional Changes? These are two distinctly different outcomes which are attainable with a host of technologies.

Ultrasound - Fantastic for condition monitoring (detecting changes in rotating assets that are related to increases or decreases in friction or the presence or emergence of impacts). Strongly advocate using ultrasound to help optimize lubrication issues.

Oil Analysis - Finding wear debris and contamination - is a change in condition, but not necessarily a specific diagnosis. Still good tech for identifying failures and even causes of failures.

Vibration - in the hands of an experienced operator, probably the best tech for diagnosing bearing faults and identifying specific causes of failure.

I/R - Too late in many cases because the heat is already there. Plus only measuring surface heat so difficult to for specific diagnosis. Still, a good trending tool that should be part of every asset condition management program.

There may not be any one good answer to your question, OP, but I would wager that most would agree that a complimentary mix of the four technologies I alluded to above will solve many reliability issues.

My closing question to the OP? Are you sure you want to diagnose bearing faults? Isn't it better to "anticipate bearing faults" instead of "diagnose" them? Then apply corrective action to avoid them altogether?

Khairul HN posted:

Vibration analysis is the best to 'diagnose' or 'anticipate' bearing fault for REB type of bearing. The brand? - depending on each analyst experience...

You don't get away that easy, Kairul HN.

Why? Its not enough to say, "Vibration is the best". You have to support your statement. 

"A car is the best means of transportation" is a true statement; until I arrive at the end of the road and need a boat, a train, or a plane to continue; or a traffic jam and wish I'd brought a bicycle or just plain continue on foot. 

I maintain there are no absolutes when it comes to condition monitoring, predictive maintenance, or asset condition management, or whatever other label you need to describe it.

Personally I agree with Allan as far as using ultrasound, vibration and oil analysis to determine if you have bearing issues as well as to make your best attempt to control or eliminate future issues.  Infrared is great and I have about 15 years of IR experience.  However (again agreeing with Allan) by the time bearing issues show up with IR, it might be to late.

You might want to look into a simple lube monitoring system such as the UE 201 or possibly a frequency adjustable ultrasound instrument that can be utilized for other inspections (electrical, steam trap and compressed air/gas leaks).  This link will get you headed in a good direction. http://cstsolutionsllc.com/wp-...versaltechnology.pdf

Good luck

Steve

http://cstsolutionsllc.com/

Add Reply

×
×
×
×