has anyone got experience of regular bearing failures in machines with VSD? the motors in question range from approx 22kw - 250kw and are double star-wound to facilitate a wide speed range. We have been told that the motors and VSDs are matched (not sure who manufactures the motors but inverters are from various sources including Emotron, Bosch, Yasakawa to name a few).
The bearings are not showing the usual ESD damage indications but we are experiencing short life (less than 8000 hours in some cases). The bearings on some units are insulated (80kw plus). I want to know if it is the bearings that re undersized or if there may be another issue

Since you have eliminated ESD, some possible issues could be heat, speed, load, lack of load and vibration.

Heat can be an issue since some "inverter duty" motors are rated at a higher temperature than the bearing seal material.

Speed is going to reduce L10.

Load can reduce L10.

Lack of load can cause the balls to skip and chatter. This can sometimes be prevented with a belleville/wave spring to add a small axial load.

Vibration, of course.

What are the bearings rated for and what conditions are they being operated in?

Bearing faailures with VSD is very common. To rule out ESD I would make sure I had a shaft grounding assembly on the motor. It is inexpensive way to rule that out and make sure the bearings are protected. Helwig Carbon offer the best solution - check out www.helwigcarbon.com.

From a bearing manufacturer I once heard that one of the most common reasons for bearing failures is wrong type of lubricant. Selection of the right lubricant depends on: temperature, position of motor (vertical, horizontal), are there corrosive gases / liquids in connection with bearing, operational speed range of the motor ... and so on..

Common mode currents are bearing killers. 1) use as low a carrier frequency as the VFD will allow. 2) make sure that one bearing is grounded and the other is insulated 3) use a dv/dt filter 4) make sure that the bearings installed are proper for the load, roller or ball bearings 5) Use an Aegis or Helwig shaft grounding device

These failures usually will occur faster in a high torgue situation at low to medium speeds like an extruder.

Also. if you are using a gearbox directly coupled to the motor, take the same precautions with it. Common mode can be transferred through it and cause similar issues.

Concerns others have raised about lubricants are valid. Make sure the lube is NON conductive.

Good Luck!

Have you had the bearings looked at for a root cause analysis? If it is due to shaft current there should be some tell tale wear patterns to point to this. You could also have a mixture of things going on but in the end a root cause analysis of the bearings will help determine what is going on. It would be beneficial to have your local bearing expert come in and take a look.

You first need to find where the bearings are manufactured. The quality of the 52100 steel, heat treatment etc play an important part in the life rating of the bearing...if the bearings are being manufactured in a less than adequate facility; premature failures are more than guaranteed...if need be I can put you in touch with one of the best bearing failure analyst around....

Andrew, all are good points and should be looked at. On these large horsepower motors typical design for VFD power would be to insulate ODE, opposite drive end bearing and provide a shaft current device to mitigate any bearing currents(Aegis). Also maintain per the motor manufactures guidelines amount and frequency of greasing, dependent on temperature, running hours, atmospheric conditions/application. The recommended grease for electric motors has been PolyRex EM by Baldor, Reliance for a number of years as typical industrial applications.
The other thing not mentioned is lead lengths and switching frequency. It is not uncommon with various manufactures and drive designs to see common mode noise which will become an issue quickly with lead lengths in excess of 100 feet. There are numerous ways to help reduce this using reactors on the drive output, reducing switching frequency, dv/dt filters, etc. Performing a simple google search on common mode noise will help immensely.

If you reach a point of frustration try:

Lew Sibley
Tribology Systems, Inc.
610-466-7547

He has over 50 years of bearing failure analysis experience.

Good luck!

Thanks for the comments and the information. We have recently changed the lubricating grease for these machines and had discussions with the machinery manufacturer regarding grease type, quantity and interval to make sure we are maintaining everything correctly. The first failure we experienced saw the bearing balls "eroded"away to look like pebbles on a beach. The second failure on the same motor was a seized bearing on the DE (motor is directly coupled to a compressor air-end with a flexible rubber insert). This unit was an 80kw motor which since we replaced the motor has run for two years without any problems. (the customer insisted on a new motor after the seconf failure took place only 4 weeks after the first repair). Our re-winder told us there was no need for insulated bearings as the motor end cap and retaining bolt holes were already insulated. I am still sceptical about that. Unfortunaately I havent been able to find the bearings we removed from the failed unit to have them examined. We have had a similar problem on a 37kw and a 55kw version.
Unfortunately the smaller one cannot be greased as it uses "sealed for life" bearings. We are now scheduling in the re-bearing of this motor at about 6000 hour intervals so as to prevent an out-of-warranty failure. The lead length from the inverter is quite short on all units - max 2 metres. The frequecny range varies depending on the kw rating of the machine - motor rpm is approx 1500-7000, once again depending on kw rating.

Reducing the carrier frequency might help, but as Kenneth says it is all about the dV/dT sine filter to surpress the voltage spikes that is causing your bearings to desintegrate like ECM milling.

A few comments:

Yaskawa has a G7, three level inverter line (NPC) that resolves this issue. They have a considerable body of information on the subject.

Output inductors on the VSD will help, but it will not resolve the common mode voltage applied to the motor windings. It may not completely resolve the problem.

There was a good paper on this subject in PCIC about 10 years ago, where a semi conductive additive was used in the oil of hydrodynamic bearings. This was effective and cheap

I would take shaft readings with a scope meter to rule out ANY kind of EDM, then you know what you have coming through your bearings. Then I would run some simple Pdm tests, IR, vibration and ultrasonic on the bearings to rule other factors out.