Root cause for motor failure 2

[paramathma]

Hello experts,

We are receiving failed electric motors from field. The motor is connected to a gear pump.
Winding open is observed after checking the connectivity.

Important observations after disassembly of motor
Contact marks on drive end side of rotor and stator see attached figure
Outer ring of Drive end bearing rotating in the housing
Dark color on rotor and stator only on the drive end

With the above information, please help me in identifying the root cause

 

[ScottyUK]

If a rub has really taken place - and isn't a result of poor assembly or dismantling - then one or both bearings are likely candidates. The picture is very poor quality and it is impossible to resolve detail.

 

[Sloss]

There's the culprit......

 

[jraef]

More questions.
1. Is this being run from an inverter?

2. Have you disassembled and inspected the bearings yet?

The rotor crash may be a secondary effect of bearing failure. EDM damage to motor bearings is a well known phenomenon in motors run from VFDs, but even if not, there are other causes of bearing EDM damage that can be considered. It can be mitigated, but the proper steps must be taken, which means identifying the problem correctly.


"You measure the size of the accomplishment by the obstacles you had to overcome to reach your goals" -- Booker T. Washington

 

[theode]

Misalignment at the gearbox can force the shaft to run out.
This results in bearing housing wear,
Eventually the rotor catches the stator.

 

[electricpete]

I agree "Outer ring of Drive end bearing rotating in the housing" could be an important element.

To validate whether it is enough to cause contact, have you done measurement of the as-found bearing housing bore? How loose is it?

How many motors are we talking about... one motor or several motors with ideantical symptoms?


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

 

[paramathma]

Friends,

Only in one case we have seen EDM damage in bearing. Please see attached image. The image is from SKF catalog, mentions current passage when the bearing is not running. A similar observation is seen in Drive end inner ring, outer ring and balls of ball bearing.
How will the current pass when the bearing is not running and later cause the rotor to touch stator.


Thanks,

 

[jraef]

You did not answer the first question; Is this being run from a VFD or not? If it is, we can elaborate. If not, it would be pointless to continue down that path.

But to address this new question, EDM damage on a stationary bearing, as shown in that photo, is typically the result of something like welding taking place on a machine with inadequate grounding, or the grounding has too much resistance, so the current travels through the bearing to ground.

EDM damage in a ROTATING motor bearing is called "fluting" and usually looks like this:

Obviously once your bearing races look like this, it becomes much more easy for the rotor to rub on the stator.


"You measure the size of the accomplishment by the obstacles you had to overcome to reach your goals" -- Booker T. Washington

 

[flexoprinting]

jraef
Thanks for that post Jeff, often times bearings are only replaced and not taken apart and examined for this type of failure. Nice heads up!

Chuck

 

[Skogsgurra]

Yes, Jeff is right.

It is usually a welding-like current that causes a damage like that.

But it is not very often caused by welding as such. There is a presentation that can be found if you google "ge bearing currents" and in that presentation frame 16 shows an example of such a failure and the failure mechanism behind such a damage is shown in frame 9.

An insulated coupling is one way to prevent such damages. Two insulated bearings is another way to avoid it. Yes, you need two. If you insulate one bearing, the current goes through the uninsulated one.



Gunnar Englund

http://www.gke.org

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

 

[flexoprinting]

Hi Gunner

Found your link, it is very informative.

http://www.gke.org/presentationer/files/GE Bearings and their electric environment w O and E.pdf

I don't understand your two bearing theory though, wouldn't the second non-insulated bearing self destruct essentially presenting another problem?

Thanks, Chuck

 

[Skogsgurra]

Hi flxprt, which frame are you referring to?

Gunnar Englund

http://www.gke.org

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

 

[ScottyUK]

Bearing failure guide from Timken. If you can split the bearing to see the working surfaces then it might be informative.

http://www.timken.co.uk/en-us/Knowl...s/Bearing-Damage-Analysis-Reference-Guide.pdf

 

[paramathma]

Thank you all for your valuable inputs

jraef Question:
Is this being run from a VFD or not? I do not know. The connection is from a control unit behind generator. We do not have access and information to this.

EDM failure mode is seen in one motor. The EDM damage is predominantly seen in DE bearing. Non drive end bearings have EDM damage but to less extent.

My question is when the rotor touches the stator windings does the current flow from the stator windings through rotor shaft to ball bearings?

 

[Skogsgurra]

"when the rotor touches the stator windings does the current flow from the stator windings through rotor shaft to ball bearings?"

The stator iron is solidly connected to the motor frame. And the end-bells/bearing housings are also solidly connected to the motor frame. So, there is not much that can happen electrically when rotor contacts stator iron. I would answer you question with a NO, it cannot.

Gunnar Englund

http://www.gke.org

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

 

[electricpete]

To talk in general about factors that may cause contact between rotor and stator:

1 - Poor initial centering.
2 - Wear at housing ID.
3 - Wear internal to the bearing.
4 – Rotor bending (result of unbalanced magnetic pull or operating near flexible rotor critical). I tend to discount this cause for your case because the rub was near the end of the machine rather than the center.
5 – Small airgap by design can make machine more susceptible to small variations above.

It doesn’t seem we have full info available. You’ve mentioned “motors’ many times. It’s not clear how many motors have rubbed and which symptoms are seen on those that rub.

A methodical approach would be to take a motor that rubbed and do some checks:

1 – Before even disassembling the motor, push pull on the rotor shaft extention and measure shaft movement radially with dial indicator. This is just extra info

2 - Remove rotor from stator but don’t remove bearings from rotor.

3 - Measure bearing internal clearance in radial direction. i.e. with bearings still mounted on rotor, push/pull on the bearing and measure movement in both directions with dial indicator (rotor solidly mounted on test fixture to prevent movement).

4 – Measure bearing OD. *

5 – Measure housing ID. *

* to be thorough, stand procedure would be to explore variability with 6 measurements: 3 angular locations (60 degrees apart) and 2 axial locations (1/4 distance in from each end).

6 – Measure rotor ID in vicninity of the rub. 3 locations 90 degrees apart. Measure at other locations for comparison.

7 – measure stator bore in the vicinity fof the rub. Measure at other locations for comparison.

8 – Attempt to determine deviation from concentricity of housing bore and stator bore. This is tough for a number of reasons and certainly complicated by rough geometry of stator bore. I’d talk to my machine shop about what they can do. Some options:
8A – with drive end endbell mouted to stator: Put on mill capable of rotating both. Center the fixture using one surface and indicate off the other surface.
with drive end endbell mouted to stator: with drive end endbell mouted to stator:
8B – with drive end endbell mouted to stator: use 3Dcoordinate measuring device.
8C – if cannot do either of above, might measure stator and endbell separately, comparing each one to the rabbet fit of endbell to stator.


With above info, you should have a good basis for making a determination.

If the cause does not jump out at you, you can take similar measurements at the ODE for comparison to help highlight measurements that are out of whack on the DE.


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

 

[electricpete]

I tried to be comprehensive but it may be a lot more than required. Leave the hardest stuff for last and start with the easy / likely checks.

Assuming your motor is already disassembled, I’d start with:
housing ID checks compared to bearing od to see how much clearance is there.
Compare to airgap measurements based on od of rotor and id of stator in area of rub.


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

 

[electricpete]

Another fairly easy check that should be added to the list of 8 items:
9 – check for clearance at rabbet fit between stator and endbell: Specifically, with endbell inserted onto stator, but not bolted tight (bolt loosely if needed), attempt to push endbell radially and check for movement with dial indicator.


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

 

[flexoprinting]

Hi Gunner

My reference was to your comment in the post just above mine, third paragraph

" An insulated coupling is one way to prevent such damages. Two insulated bearings is another way to avoid it. Yes, you need two. If you insulate one bearing, the current goes through the uninsulated one."

Are you suggesting this from field experience and found this technique to bleed off current without damage to either bearing?

By the way, could using the wrong grease contribute to the problem the OP has?

Thank's Chuck

 

[Skogsgurra]

OK, flexo - I see what you mean.

No. This is not about bleeding off any current. It is about stopping a welding current or, more likely, a ault current caused by an unwanted potential difference between the PE system and the bulk of a plant. Picture 9 in the presentation shows such a case.

There is a fault current flowing from failure to the earthing bar. From the earthing bar, most of the fault current flows to next ground point (or to main ground) while a lesser part (arrow width is proportional to current) flows to the motor's frame and then via the DE bearing to the grounded machine.

There are two ways to avoid damage in such a situation: The first is to use an insulated coupling and the second method is to use two insulated bearings. One insulated bearing will not work because the current then will flow through the non-insulated bearing.

Remember that the presentation shows all four fundamental failure mechanisms. The one in the example is the so-called "Frame Voltage" mechanism. It has nothing to do with inductive or capacitive coupling.

Gunnar Englund

http://www.gke.org

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