Bearing failure with no sign 1

[PAP]

Dear sir
I'd to discuss with anyone who experiences the failure of bearings of motor(2Z type) with no sign of vibration problem. What 's the possible root cause of this problem?
REGARDS
Prasit P.

 

[electricuwe]

Is your motor supplied by a variable speed drive (inverter)?

 

[stardelta]

There are many reasons for bearing failure. Can you provide us with some more info, such as....

Bearing type.
Motor type.
Motor duty.
Speed and power.
Driven application.
Coupling type.
Enviromental conditions.
IP rating.
Starting and speed control devices.
Shaft and motor loading.

Any attempt to diagnose a failure without being aware of these factors would be nothing more than a good guess!

 

[electricpete]

As mentioned above there are lots of possibilities. It may help to do a post-morten on the bearing and see what clues it holds.

When you say "no warning", I assume that means you are doing some monitoring of vibration and possibly temperature. Can you clarify.

One comment is that defects on balls can be hidden during vibrations because the ball spins in somewhat random fashion where defect will sometimes make contact and sometimes miss. Another possibility is cage failure... once the cage fails, the bearing goes very fast.

By the way how old is the bearing? I assume 2Z means double-shielded. That provides limited ability to relubricate. What is the speed and size of the bearing? Have you ever checked bearing housing temperatures in service. Vibration shows any indications of misalignment or other possible abnormal load mechanism? Does the unit vibrate when it is idle?

 

[EdDanzer]

One other thing to check is housing bore/roundness and the shaft diameter and roundness.

 

[Skogsgurra]

PAP,

You may have a PWM-related problem. I have seen a lot of these the last five - six years. The report on

http://gke.org/rapporter

and look for "El Lagerskador.pdf" The report has a Swedish title, but is written in English. It describes what is going on in an asynchronous motor with one insulated bearing. It turns out that one insulated bearing does not help, the other bearing needs a shorting brush. Or, you can use common-mode filters in the motor cable (a non-chrystalline toroid, "amourphous iron" that keeps fast voltage changes from the motor winding).

The report also contains an "autopsy" of the bearing. It was carried out in the central SKF laboratory and is a good introduction to bearing current damages from PWM inverters.

Gunnar Englund

 

[GusD]

Hi PAP

If it is important to know the exact cause of failure,it might be worthwhile to get someone to do a thorough analyses of the failure.Whenever we have a bearing failure that seems hard to explain,we send it to a Bearing Failure Analyses Eng. firm.
They send us a complete falure analyses report of the failed elements and kind of takes the guess work out of it.As some of the resoponses indicate,bearings can fail from a multitude of reasons.You won't be able to correct causes of failure,unless you have knowledge of what causes them to fail.

GusD

 

[jbartos]

Suggestion: Visit

http://www.ab.com/drives/techpapers/ieee/ieee.html

for several papers dealing with bearing current problems and remedies, e.g.

Doyle Busse, Jay Erdman, Russel J. Kerkman, Dave Schlegel, and Gary Skibinski, Bearing Currents and Their Relationship to PWM Drives, IEEE Industrial Electronics Conference (IECON), Orlando, FL, Nov. 6-10, 1995, pp. 698-705 and IEEE - Transactions on Power Electronics, Vol. 12, No. 2, March 1997, pp. 243-252.

Jay M. Erdman, Russel J. Kerkman, David W. Schlegel, and Gary L. Skibinski, Effect of PWM Inverters on AC Motor Bearing Currents and Shaft Voltages. IEEE - Applied Power Electronics Conference (APEC), Dallas, TX, March 6-9, 1995, pp. 24-33 and IEEE Transactions on Industry Applications, Vol. 32, No. 2, March/April 1996, pp. 250-259.

 

[AeroPD805]

Pap,
A couple of less complicated details to check first
are:
a) did the bearing have the correct preload?
This is a common cause of failure - too much and
the bearing runs hot, too little and the bearng
won't track correctly. The preload depends on the
internal clearance of the bearing and application.
Adjustment of preload is done by the use of a wave
washer(s) and/or shims.
b) lubrication; this is a science in itself. Again
this depends on application. Without knowing your
application, I can't recommend a specific grease
or oil.
Cheers
Pete
p.s. JBar, thank you for the link, interesting paper
on pwm and bearing life.

 

[electricpete]

I would like to ask the posters if they have any thoughts specifically on the subject of bearing failing without much vibration warning. (My only thought on the subject as identified above are that failures starting as cage faults and ball faults can often proceed rapidly to bearing failure.)

 

[GusD]

Hi electricpete

Cage failures are normally caused from poor lubrication.
The cage failure is normally visible as a sideband of other fundamental frequencies and at very low levels.They could easily be lost in the shuffle of spectral noise without showing large discrete peaks.

As for bearings that can fail without much notice,I would say that oil lubricated bearings when they are suddenly starved for oil and subject to preload they can fail very rapidly.
The vibration would probably be there, but unless you happen to have an Online vib system you won't know about it.(Input bearings on gearboxes which are normally spinning at motor speed,can fail in 1/2 hrs or less.

Some very large bearings can fail at very low levels of vibration (eg .002 in/sec).It could be missed.
Hi-frequency spectra analyses like Shock pulse/Spike Energy/can be very good to identify bearing defects,even at very low levels.


GusD

 

[electricpete]

Good comments gus.

I have seen cage frequency (FTF) show up frequently for NON-CAGE failures. In particular FTF shows up as sideband of 2*BSF in case of ball fault. That can be explained by the fact that the fact that the defective ball moves into the load zone at the same speed as the cage.

It is widely acknowledged that a true cage failure often proceeds from the point of indetectable degradation to bearing failure in very short time. I can tell you I have seen brass cages very badly worn (not failed) and no cage frequencies anywhere in the spectrum. If cage fault were detectable I believe it would show up at FTF+harmonics and possibly FTF sidebands around 1X, but not FTF sidebands around other fault frequencies. The pattern of FTF sidebands around other fault frequencies is associated with non-cage bearing faults.