Electrical discharge to earth of motor bearings

[peter53]

I'm currently undertaking ultrasound inspections of some 6600 volt soft start motors and appear to be getting indications of fluting whereby i might be getting a discharge to earth through the motor bearings.
Iknow VFD controlled motors can cause an induced voltage on the motor shaft but not aware if soft start motors would do the same. I'd be grateful for some advice on the subject please.

 

[Skogsgurra]

Softstarters are usually employed for a short while, a minute or so. So, they do not cause any appreciable bearing problems.

I can imagine that you get problems if there's no by-pass contactor/breaker that closes after start is complete.

Running with SS at 100 % can bring problems, especially if it isn't an SCR softstarter.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[peter53]

Thank you for your kind reply although i have just been told that these motors may be DOL so i assume the problem of discharge through the bearings must be comming from some other source perhaps. I assume i would now need to use a voltmeter and test the motor casing to earth.

 

[LiteYear]

VFDs and other external sources of high frequency currents are one cause of bearing current, but internal sources have also been known to exist for 80 years or more. The causes are things like eccentricity and non-symmetrical rotors, which create circulating currents from stator, through bearings, down the rotor and back through the bearings. Since they're internally sourced, they're very hard to detect.

Measuring earth current might be a way to understand whether your damaging currents are internally or externally sourced, but I think you'll have a hard time making a call just by "use a voltmeter and test motor casing to earth". Unfortunately I'm out of my depth here. Hopefully someone with some actual field experience will chime in.

 

[Skogsgurra]

No SS? Never, ever?

Then you should look not only at motor frame voltage to solid ground (if you can find it) but also check voltage across bearings. Measure from shaft to frame. If you have more than 500 mV to 1 V RMS, you may have a problem that needs to be eliminated.

These days, DOL motors are known to have problem with PWM residues coming from other drives with PWM inverters in the vincinity. To check that, you need an oscilloscope with at least 20 MHz BW to measure shaft to frame voltage. If you can see HF components with a couple of volts or more, peak, then there's the cause for the fluting.

To find the remedy needs deeper investigation.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[peter53]

Thanks guys you have been very helpful

 

[purplepete]

This experience is on a 415 volt 3 phase motor.
We had about 150 metres of 300 sq mm singles Cable feeding a 400 amp FLC motor.
The earth conductor ran neatly alongside the same phase conductor almost the whole length.
The motor suffered a bearing failure after about a year (on Christmas Eve, when else?).
The repairer said we've got earth currents.
We tong tested the earth conductor: 90 amps! After we relocated it away from the triad, down to next to nothing.
This may be of use to someone, not necessarily with the 6600 volt problem..

 

[Tmoose]

Do you make conventional vibration measurements on those motors?

What are the symptoms of fluting using Ultrasound?

thanks

Dan T

 

[peter53]

Dan another company do the vibration analysis what i find using ultrasound is a buzzing sound which is backed up by a spectralyzer showing dB and hz, basically i find peaks every 100hz with a small peak every 50hz which indicates fluting. As we are 50hz supply thats where i'm looking for electrical faults compared to mechanical.

 

[Skogsgurra]

peter53
The 100 Hz may also come from the mains voltage. At 50 Hz and force being current squared, you will see lots of 100 Hz and multiples thereof.

Fluting at exactly 100 Hz is not very common. BPF is all over the place, but not very often right on double mains frequency. And given that the fluting period and BPF interact, I would say that a bearing in a machine where you have 100 Hz and its multiples is a healthy machine.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[peter53]

Thats very interesting as i have found the same symptom on another motor at another location but on the same site. If this is a supply problem what would be the best way to prove this issue compared to a bearing issue.

 

[Skogsgurra]

For a perfectly built motor and a perfectly balanced grid, there should no be much 100 Hz mechanical vibration.

The first thing to look for is if your grid has any imbalance. Already at one or two percent difference between phase voltages, you will get noticeable 100 Hz vibrations. And more at higher imbalance levels.

Harmonics do not usually cause this kind of problem, so do not pay too much attention to them.

Are the two motors fed from the same transformer?

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[peter53]

I'm not sure if they are fed from the same transformer but i'll find out today and post the answer, thanks for your help i appreciate it.

 

[peter53]

Skogsgurra, On site there 3 different locations for these 3 phase 6600v blower motors. Each location has 4 motors and named as phase 1, phase 2, phase 3 [just a location name] on 1 motor on phase 1 the NDE has the 100hz issue as does 1 motor on phase 3.
Would i be correct in assuming that if this was a supply problem then i would see this on all the other motors. 1 of the motors i have now been told has indicated a vibration alert so any advice you have i'd be grateful as this is more indepth than i normally would be expected to deal with.

 

[peter53]

Another bit of info i just found out is that the motor shaft is ceramic coated as to protect the motor NDE from this type of problem?

 

[ScusaMe]

Peter53

- Are these 2-pole (i.e. ~3000 rpm) motors ?

- What is measured value of mains voltage ?

 

[peter53]

ScusaMe, all the info i can get is that the motor is 3 phase, 6600volts, rpm 2990 and that it is an ex rated motor as it is located in a gas storage area.

 

[ScusaMe]

Peter53

Excellent!

2900 rpm is indeed a 2-pole motor ....
i.e.,3000 rpm sync. speed (-) 100 rpm (nominal)slip at full load.

2-pole motors are notorious for their vibration problems.
One only needs to do a search— i.e., Google — on ...
Vibration in 2-pole motors .....

You'll find several articles / papers on the subject, some in PDF format. There are also some IEEE papers which, unfortunately are not free, even for those of us that are members thereof.

As a Prologue to my comments below, I offer excerpts from Neil Mulji's excellent article ... entitled— Vibration in Electric Motors

Vibration in Electric Motors
Neil Mulji | Saturday, August 06, 2011

.... A basic electric motor has a magnetized stator that surrounds a rotor; the magnetic interaction between the two causes the rotor to spin.

Faults in motors can develop due to resonances, imbalance, misalignment and foundation problems or improper mounting. These faults can lead to bearing failures and cause vibration.

Some of the electrical and mechanical faults unique to motors include:

Eccentric or loose rotors or stators
Broken rotor bars
Bowed rotors
Uneven air gap between the rotor and stator
Loose laminations
Electrical discharges between the various electrical components

When analyzing vibration data for motors, it is typical to observe a vibration component at twice the line frequency (at 100/120Hz). The line frequency (50/60Hz) is the frequency at which AC power is supplied to the motor; this causes changes in magnetic attraction between the rotor and stator at twice the line frequency. The varying magnetic forces cause small dimensional changes in the iron material, leading to vibration.

In 2-pole motors, it may be difficult to discern differences between the 2X frequency and twice the line frequency(100/120Hz) and it is important to use a high enough resolution when collecting data. One method to verify the existence of a peak in the vibration at the 2X frequency is to take measurements while the motor is running, and then cutting off power; any peak at twice the line frequency should disappear once the power is turned off, leaving the 2X frequency if it exists.

Frequencies that may show up on the spectral data include:

- Slip frequency - the difference between the motor’s running speed and the synchronous speed; it is typically small in value and will be present as a sub-synchronous frequency.
- Pole-pass frequency - it is equal to the number of poles times the slip frequency.
- Sidebands spaced at intervals equivalent to the to the pole-pass frequency.
- Running speed:

3600 or 3000 RPM for a 2-pole motor
1800 or 1500 RPM for a 4-pole motor
1200 or 1000 RPM for a 6-pole motor

Eccentrically positioned rotors cause a variable air gab between the rotor and stator which leads to a pulsing vibration. Once again, a vibration component at twice the line frequency will be present, however it will have sidebands spaced at intervals equivalent to the pole-pass frequency; the sidebands will also surround the 1X frequency.

Comments:

In re-reading your post of 8 Jun 12 ... (quoted below) —

...another company do the vibration analysis what i find using ultrasound is a buzzing sound which is backed up by a spectralyzer showing dB and hz, basically i find peaks every 100hz with a small peak every 50hz which indicates fluting. As we are 50hz supply thats where i'm looking for electrical faults compared to mechanical.

— I would take umbrage with the conclusions drawn as well as the methodology by which the testing was performed.

The 50-Hz and 100-Hz components from which you draw conclusions are most likely related to— 1X & 2X— Mains Supply Frequency, consequent to testing while mains excitation was applied to the MUT (Motor Under Test). The better way to conduct these tests is while the motors are coasting-down immediately upon removing excitation therefrom. In this mode, you eliminate any and all influences from the magnetics— (electromagnetic & electromechanical)—consequent to motor excitation.

see:
Application Notes: Vibration Diagnostics for Industrial Electric Motor Drives

An Analytical Approach to Solving Motor Vibration Problems
​

-------
Bearing fluting resulting from dV/dt of the source voltage is not as likely on medium-voltage 50-Hz supplies as 60-Hz supplies because 50-Hz produces a lower value of dV/dt. Somewhere in my literature collection there is a reference to dV/dt causing bearing fluting on medium voltage 60-Hz supply back in the 1920 era. Also it is more likely on ball type bearings than on roller bearings. (Check with UR MTR MFGR. to see what type bearings are installed in affected machines of your concern.)

With induced electromechanical vibration, one is more likely to see bearing problems from brinelling as opposed to fluting. Most positive way to be sure is to remove the bearing exhibiting vibration during coast-down ... cutting it open and inspecting the races. Expensive... but informative; particularly where more than one motor is affected.

Other causes of low frequency vibration can be attributed to installation practices. Here are but a few for consideration:

1) Misalignment
2) Soft Footing
3) Weak Foundation
4) Shaft Coupling Eccentricities
5) Loose bolts of pump, gear box etc.
6) Bearing loose in housing (The bearing housing will be heated up)

Take a look at this article, which offers diagnostic advice for pursuing the above installation problems leading to vibration issues.
Basic Motor Vibration Troubleshooting Tips

Hope this helps..
\'scusaMe/

 

[electricpete]

Like Tmoose, I would really like to see some traditional vibration data.

I don't know a lot about ultrasound, but the frequency band is generally high (audible range is up to 20khz, so ultrasound is generally higher?). I don't think twice line frequency vib would show on ultrasound.

There are however a variety of higher frequency electromagnetic slot-related vibrations. For example rotor bar pass frequency +/- 2*LF. The signal processing that you are using might output to the user some frequency which is different than the raw data. For example envelope detection of RBPF+/- 2*LF vibration (similar to Entek Spike Energy or CSI Peakvue) yields 2*LF output. I seem to recall that some ultrasound instruments use something like heterodyne, which is different than envelope detection but I'm guessing it might somehow have a similar effect? Just at thought. In any case the data presented is very sketchy...need more dat.

=====================================
(2B)+(2B)' ?

 

[peter53]

ScusaMe, thanks for your input i'll certainly follow the links that you have posted as i clearly need to understand more about vibration and motors in general.
electricpete, I'm hooking up later with a trainer from UESystems to help me analyse some of my issues and will post further information later next week. You are correct in that ultrasound 'heterodynes' the sounds so they can be displayed in a visual form in their software package.
I attended a conference recently regarding ultrasound and one of the speakers mentioned that if you get multiples of 50hz frequency readings basically meaning high peaks at 100hz then a small peak inbetween at 50hz this would indicate an electrical problem. He went on to say that it would indicate an electrical discharge through the motor bearings.
The issue i now face is that the shaft of the motor has a ceramic coating precisly designed to stop any discharge through the NDE bearings of the motor so i'm not convinced that it is actual discharge causing the read outs.