Odd bearing failure

Lacking much information, a good guess would be too much belt tension. The 6311 bearing does not have a very high radial load capacity. Over tensioned or misaligned belts/pulleys would cause overheat/overload failure. What did the grease look like?

Are these electrical arc marks on the rollers? If so, are there any on the inner race or outer race.If so, has there been any welding close to this motor recently?

If these are not welding grounding marks, it could be the electrica discharge through the bearing from the VFD might have been allowed to "run" too long.

There had not been any welding by the unit. As far as the grease these fans had just had PM's done recently so there was plenty of fresh grease. The belt tension was OK. We definitly need to get a handle on the fluting issue.

Robert,

What about the marks circled, are they as they appear to be in the photo, electrical arcs?

Ralph, yes they appear to be arcs. Here is a photo of the rolling element.

Robert,

I have never seen electrical flutting do this to a ball, maybe some one has, but not I.

This looks like pure electrical shorting, like from welding arcing through the ball or DC or AC current shorting out wnile the ball was not rolling.

Welding would not have to done on the motor or fan or even the fan frame in order to short through the motor. Depends on the ground passing through the motor.

I would look for something drastically wrong in the VFD and inquire about any maintentance near the motor in the recent days before this happened.

The mysterious symptom of craters on bearing balls has been discussed in many other threads:

http://maintenanceforums.com/e...061080213#6061080213

http://maintenanceforums.com/e...911023731#1911023731

http://maintenanceforums.com/e...411046412#7411046412

http://maintenanceforums.com/e...681084221#4681084221

http://maintenanceforums.com/e...871055632#5871055632

If you read through those threads you'll find a lot of good suggestions from knowledgeable folks about possible causes for this type of damage.

I don't claim to know any more about it than anyone else. I do have a little bit of a theory, but it's a little bit out there. In the very first one of the links above I summarized what I thought I knew of this symptom (craters) at the time. That includes our own experience where the sequence of events leading up the the failure very strongly suggested that the cause was a loss of oil but at that time had no logical explanation why loss of oil should create this pattern.

Now I think I have a possible explanation for how loss of lubrication can result in these creaters due to... Hydrogen embrittlement Here is an exceprt from "Advanced Concepts of Bearing Technology"

quote:

Another mode of rolling contact surface failure is caused by hydrogen ions, which attack the surface material, resulting in pitting or spalling of the surface. Figure 10.44 depicts the
spalled surface of a bearing ball caused by hydrogen embrittlement. This failure mode,
which is relatively rare, is generally associated with rolling contact surface operation at
steady-state temperature above that at which degradation of the mineral oil lubricant
commences. It is generally associated with significant differential or gross sliding between
rolling contact surfaces subjected to high Hertz stress, the surfaces incompletely or marginally
separated by a mineral oil lubricant film. The high temperature that results in the
contact causes the chemical breakdown of the lubricant, releasing hydrogen ions. Hydrogen
embrittlement is also associated with an environment surrounding the bearing, which does
not allow the hydrogen ions to easily dissipate from the vicinity of the bearing; for example,
in a well-sealed application.

The essentially circular shapes of the spalled areas of Figure 10.44 are probably associated
with the axisymmetric residual stress distribution existing in the bearing ball after heat
treatment and surface finishing. As illustrated in Figure 10.45, the hydrogen ions penetrate
the steel from the surface of the component, resulting in cracks. These in turn propagate
weakening the material until spalling occurs. Many researchers have investigated the occur-
rence of hydrogen embrittlement failure. In all these reported experiments, hydrogen was
introduced into the application in the presence of excessive contact stresses, and in most cases,
in the presence of elevated temperatures. In none of these cases did the production of
hydrogen ions result from lubricant chemical breakdown.

The figure 10.44 from the book is attached and looks similar to most of the craters in the posts. So the possible scenario for our failure could have been that the known loss of lube oil caused overheating which caused hydrogen embrittlement, which caused the known craters. There are a few questions that I still can't exactly answer: Question 1: wouldn't this overheating create more discoloration? Answer 1: I don't know what temperature is required for this hydrogen release and embrittlement to occur; Question 2 if loss of lube caused the high temperature... then wouldn't there be no lube around to create hydrogren? Answer 2: I dunno. In in the absence of any better theory I can come up with, this is the best theory I can come up with ;-) The bearing balls did not split open in our case.

And while everyone agrees the marks on the races are clearly related to vfd, I do agree with Ralph these crater marks on the bearing balls are not likely related to vfd. But I think 2 or 3 of the posts linked above were vfd applications (ours was not). Now yours also on a vfd seems to me like a high proportion of these occuring on vfd machines (I perceive that much less than half the machines monitored are vfd drive). Coincidence or relevant fact? Beats me.

One more thought: 3 or 4 of these cases of crater cases posted were accompanied by splitting of the bearing balls (not ours).

The splitting has the appearance of a "brittle" failure since no plastic deformation is evident. The craters provide a possible initiating flaw for brittle fracture.

Maybe this type of failure is something that should be classified under the broad category of "embrittlement" in the sense that brittle failure would not normally be expected in this type of application (based on material and loading pattern)?

Any opinions on whether the splitting of the balls appears as brittle? Whether a brittle failure mode is expected ? Whether this might be consistent with "embrittlement" of some type ?

=======
Am I rememberin right that brittle fracture always requires tensile load?

Robert,

The outer race has major electrical fluting. You may have bad iron in your stator or rotor. When I say "bad iron" I mean you may have laminations that are shorted together causing a hot spot and also an electrical imbalance in your motor. Has this motor ever been rewound? Sometimes when the stator is cleaned before rewind the laminations can be shorted out with the sandblasting. Do you have a grounding brush on the shaft?

Mark

Here is another picture of a rolling element from the same motor getting ready to split.

Sledder. The pictures of the outer race are fluting on the outboard bearing. The cracked elements are from the drive side bearing. This is a motor that had never been rewound. It did not have a grounding brush but does now.

What does the inner and outer races look like on the 6311?

Robert U.
Change the power cable between your inverter and your motor to a 'shielded' power cable. These cables have either a braided shield around all three conductors, or a copper film shield. It is very important connect this shield ground to a terminal lug inside the motor terminal box to a bolt that goes through the box into the frame, and also to the PE ground bus in the inverter cabinet. This shield will carry the high frequency current that is not only fluting your bearings, but it also has the ability to breakdown grease so you lose all lubrication. Also, you mentioned this motor is in a belted application and this bearing is the drive end, then it may have had a high belt tension. That, coupled with the severe fluting, minus lubrication would definitely destroy the bearing. As far as why the ball damage has manifested itself with these splits, I would be very curious to know if this type of failure isn't just due to a faulty manufacturing process. The most important issue here is to remove the cause of the fluting (VFD/Motor/Cable), and then see how the bearings behave.

What is very interesting to me is that not one response to you mentioned the above issues, yet they have been mentioned on this bulletin board numerous times.
The lack of knowledge in the field for this very simply fixed problem is almost laughable. The NEC ignores the issue (My guess is there are major marketing/sales issues at stake, and that should not be their focus!) The European market has eliminated these issues by simply not allowing a motor and inverter application without the proper power cable.
Please read the attachment and call if you have ANY questions.
Regards,
Ron Brook
215-365-1500

Ron,

I remember this subject well and thought I posted my standard recommendation that he search the site for your prior comments on this subject but it must have been in another topic.

Anyway, thanks for the paper. (Thanks to ABB, too.)

quote:

What is very interesting to me is that not one response to you mentioned the above issues, yet they have been mentioned on this bulletin board numerous times.
The lack of knowledge in the field for this very simply fixed problem is almost laughable. The NEC ignores the issue...

Ron - I believe your comments about omissions in others' responses are based on your mis-reading of the original question.

quote:

Outboard bearing 6210 had typical fluting issues (VFD). Drive bearing is 6311. Any ideas on the cause of drive bearing failure? No recent vib. data.

The fluting occurred on the outboard / ODE bearing. He did not ask about the fluting on the ODE bearing or how to prevent it (quite the contrary he called it typical). He asked about the odd damage on the inboard / DE bearing (craters and splitting of balls). By your own response , the crater/splitting failure on the inboard side is not related to vfd (bearing manufacturing defect). So.... there is no logical reason to question anyone's response because they did not mention VFD fixes in response to a question about a symptom (crater/splitting) that you yourself attribute to causes other than vfd.

Robert - I do think the physical evidence of craters and splitting suggests a brittle failure mode where we would not expect it for normal ball materials. Stepping back for a broader view of possibilities:
1 - Hydrogen embrittlement changed the material characteristics as mentioned in the article. Conditions leading to that as discussed in the article. Note it was characterized in the article as "relatively rare"
2 - Manufacturing defect caused improper material properties from the beginning as mentioned by Ron is certainly a possible cause. i.e. the material was not properly processed to give the right properties, which resulting in a brittle ball. This type thing is more likely if the brand of bearing is obscure or even worse if you had a counterfeit bearing. (What is the brand?)
3 - Again due to the fact that so many of the links mention vfd's (I think all of the splitting occurred on vfd motors), I wonder if there is some mechanism where bearing current over time results in channge in properties (embrittlement)?


If you had an opportunity to send the balls (split and non-split balls) for analysis I'll bet a metallurist could confirm brittleness and possibly provide insight on the cause.

I imagine you might be able to do a homemade test to confirm brittleness like putting standard and suspect non-split balls into a vice ...I'll bet the suspect balls will crack open and the standard won't. Of course this would be a test that would require a lot of planning regarding personal safety precautions related to possible shrapnel.

Robert,

Several years ago I sent a bearing in to NTN for analysis that had split balls with the small hollow space in the center (balls looked very much like the balls in your pictures). They called this "water quenching," where the bearing looses lubrication and heats up and then water, or some other liquid, get into the bearing and "quenches" the hot rolling elements causing them to crack. This could even be caused by greasing a very hot bearing. Don't know if this is possible in you case on not. The bearing I sent in was off of a spar mill where flood coolant got into the front of the motor because the front seal drain hole was plugged allowing the coolant into the bearing. We increased the size of the drain hole and I never saw this happen again.

Regards,
John J

EP,
Yes, but he also mentioned needing to 'get a hand' on the fluting issues.
Again, this type of equipment is being installed without the customers guarding themselves with a properly worded spec. I still believe that the root cause to his problem is in this order:
EDM due to incorrect wiring
Fluting and breakdown of lubrication as a result
Failure of the bearing due to lack of lube and excessive g loads due to fluting
I really like John J's response. The failure does look like something you would expect from fast cooling of very hot metal.

John very good response on perhaps fast cooling when very hot. As I said this had been recently PM'd so it could have had cold grease added when hot.

Lake man here is the 6311 brg pictures.

Ron thanks for the ABB article I will follow up

Pete thanks for the links very useful info.

Ron - You are correct Robert mentioned wanting to getting a handle on the fluting as one of four informational statements (not questions) in a followup post - I honestly didn't even see that. That you read and understood that particular comment out of many closely enough to understand he was asking for help on that issue is a good thing. That every other person read and responded to the original post, original question and original subject line without responding to one particular statement among many in a followup post should not be viewed as a bad thing or mentioned in the same breath as "laughable lack of knowledge" imo.

The possible connections between vfd and craters/splitting are certainly open for discussion but by no means conclusive. Imo embrittlement from any source including current flow (if current flow can cause embrittlement) fits the damage pattern of craters and splitting. I'm not sure I understand the link you are drawing between vfd and craters/splitting as in your comments below:

quote:

Fluting and breakdown of lubrication as a result
Failure of the bearing due to lack of lube and excessive g loads due to fluting

Seems that you are suggesting that fluting contributes to the craters/splitting (if I understood you correctly). It could be, but in my opinion fwiw not likely. Let's look at it closer:

• 1 - ** First and foremost we note that bearing with the severe fluting was not the one that had the craters/splits. We had multiple crater/split defects on the bearing with less severe fluting and no crater/split defects on the bearing with more severe fluting (did I describe that right Robert?)
  
• 2 - Maybe you're suggesting fluting causes g loading which causes damage to balls....could be but I would think we would get some fatigue on the races as well. Also my judgement (right or wrong - interested in other opinions) is that the brittle failure pattern we see would not be expected if the material has the correct properties. Nothing about the g loading changes the material properties of the ball. And again, item 1 does not support that g loading was the cause.
  
• 3 - Maybe you're suggesting fluting causes lubrication breakdown which causes damage to the balls. I agree with the lack of lubrcication part but don't see this as a consequence of fluting. The only mechanisms for fluting to breakdown lubrication that I can see is when the temperature gets very high in the very final stages of failure ...I think you will have a lot more evidence of problems on the races at that late stage. (by the way Robert, was there indication of overheating... races stained brown? What did grease look like?). If on the other hand you were to suggest that current flow can break down the lubrication... on that I would agree - it is well documented in available literature as I'm sure you know.

I certainly may not have understood exactly what link you were trying to draw.

I definitely agree JohnJ's scenario is worth adding to the list. I have to go back to review our own scenario to see if it matches. I do remember that we let the oil run low completely out of sight in the oiler, then found these craters a week later when we went in to repair the seals that had caused the leaks (hadn't seen anything on vibration the week before the oil ran low). I don't remember if we added oil while the machine was still running .. if we did that might have given the quenching effect.

EP,
You are correct in the statement that I was suggesting that the current caused the breakdown in the lubrication. Once that happened, the other variables are not known to us, for example, the belt tension, dynamic load distribution between the two bearings (maybe the load, due to sheave overhang puts a much bigger load on the NDE bearings), etc.
Robert also made the statement that there was plenty of new fresh grease in the bearing. John J's comments about 'cool' hitting 'hot' make sense. It is also not unheard of to see maintenance folks (not throwing stones here, just stating fact), when confronted with a motor that they see running hot, proceed to fill it with grease. There are so many reasons for auto lubers I can't begin to cover all of them.
You can imagine my frustration, EP, when I contacted the NEC personally and was not even given the courtesy of a return call to get EDM/grounding issues addressed.
I will never forget a customer asking me why my company's sales engineer didn't warn him about EDM and requiring specific power cable. I replied that he would have lost the sale because the competition might have told him it wasn't required. He replied, 'Yes, but I would have had more respect for him' and my reply was '... and his sales manager would have chewed him out.' The customer nodded, understanding and said 'It's all about the money, isn't it?'
That is what is so infuriating about the NEC. They are supposed to be above all that.
Sorry, I will step down off the soap box.
NEXT!