Hi,

i got a call from the client that one of his motor NDE bearing is getting failed on the 2nd day after replacement. this motor is connected to the blower/fan assembly. i am visiting the site tomorrow morning. is it a mechanical problem or an electrical problem. motor DE bearing seems to be ok. from where so much force is coming to destroy the bearing within 2 days.

need help. i will update with more data after coming back from site.

Cheers,

Madhu

This message has been edited. Last edited by: Madhu,

When it goes that fast a minor sledge hammer adjustment or several or to tight fit may have some connection with the fail or no grease at all maybe. I have not seen VFD fluting or similar do that kind of quick damage unless it´s welding.... Olov

Bearing manufacturers cite installation practices as a major contributor to bearing failures, especially short term where an "original" survived for a resonable period of time.

Look carefully at all aspects of how it was installed.

John from PA

Loss of clearance caused by thermal expansion is a common cause of rapid bearing failure. It's possible to diagnose loss of clearance by short term trending of average level of acceleration and envelope spectra during the time from startup until bearing temperatures are stabilized. If loss of clearance is not an issue, signals generated by the bearings should be stable. If the signal levels are not stable, then clearance loss is likely. Trending the average level(0x), 1x, and 2x levels of the ebearings envelope spectra can give very usefful information about what is happening within the bearing.

The NDE bearing is generally the thrust bearing. Attention needs to be paid to the coupling gap. Even the ones with rubber flex elements can put undue axial stresses on bearings. Monitoring the bearing will let you know you are destroying another one, but identifying the root cause and correcting it will eliminate the issue. Has an autopsy been performed on one of the failed bearings? Have the fits been measured to verify they are in tolerance? Was the lube inspected in the failed bearing? All this information has a story to tell and by knowing that story, you can eliminate the cause of the failure.

Have a good one,

Gary B

Dear all,

Here is the data what I had collected at site and my comments based on the vib data taken. require your help. please go through the attachment and give me your thoughts. had anyone come across this kind of fault earlier.

Cheers,
Madhu

AHU_case_study_1.pdf (325 Kb, 183 downloads)

quote:

i got a call from the client that one of his motor NDE bearing is getting failed on the 2nd day after replacement.

Has this thing failed more than once, or just this once. Judging from what your quote says, it fails every time it is replaced, but how many times has it been replaced, say in the last 2 weeks? Is this a one time failure before they called you? Had or did it show any signs of failing when you were balancing it?

I personally have never seen this type bearing (Cylindrical Roller Bearing:::straight roller NU type) bearing on the Non Drive End of a motor. This being a direct coupled motor to the fan, why even have this type bearing in the motor at all?

From the picture it does appear to have some type of electrical discharge scarring the inner race, but again, how many times has this happened, once or more? If only once, were they doing any welding around the motor before startup?

Just some thoughts.

50Hz, 3000RPM running speed. This is a 2 pole Synchronous motor? (if so, there must be a good reason?)

Like Ralph, I've never seen this type of bearing on the NDE of a motor before.

When it was noted in the report that the outer race showed no damage, was the outer ring disassembled to allow inspection of the actual raceway?

I'd be a bit suspicious of false brinelling. A close inspection with a magnifying glass or similar should reveal the mechanism of the damage.

Ian

Mahdu,
Is Fmax=30000 cpm or it is just a zoomed scale? Although it won't explain the cause, BPIR should be present in the data. Hard to read the image of the damaged ring, but it appears that it has smear marks. That leads to believe what Duncan and OLI have mentioned - loss of clearance. Possibly initial clearance is set wrong.
What kind of mounting has been used?
Are there signs of inner race rotating on the shaft?
Was noise the only symptom making them to shut the fan down?
Are there any signs if races' internal misalignment?

Dave

Hi,
Thanks for the inputs. this bearing has failed more than 5-6 times in the last 3 weeks. within 24 hours they will hear the abnormal sound from the Motor NDE side. if they allow the bearing to run for the 2nd or 3rd day the line marks on the inner race will go deeper. alignment is ok. foundation is intact. motor DE bearing is fine. bearing clearences are good. some of you had mentioned that this is not a suitable bearing for the motor NDE side. then which is the bearing which is suitable. will doing a motor current analysis on the unit help..???

Cheers,
Madhu

quote:

some of you had mentioned that this is not a suitable bearing for the motor NDE side. then which is the bearing which is suitable. will doing a motor current analysis on the unit help..???

Not to say this is not a "suitable" bearing for this motor bearing position, I have never seen this type bearing on the NDE and not to say it is wrong or right. This type bearing is usually on the DE of a heavy loaded belt driven machine.

The misalignment question might have been referring to the alignment of the bearing to the shaft/housing or maybe I misread the statement and not motor to fan alignment.

How is the bearing being installed on the shaft as far as heating it in order to get in on the shaft? Oven heat, magnetic bearing heater, torch,????? Is the cage assembly being left on the inner race while being heated? Is the bearing being allowed to cool before finishing the installation? Has the shaft been check for "how tight" of a fit there is between the inner ring and the shaft? There may be an electrical discharge coming from a magnetic bearing heater if one is used.

Personally I would disassemle the bearing and install the inner ring on the shaft seperate from the cage assembly and the outer ring if possible. But even more so, I would determine why the motor has this design bearing on the non drive end and then make the decision as whether or not to install a 63XX or 62XX series bearing instead of this NU type. Does the motor tag say this what goes on the NDE, if so, check with the manufacturer as to why, etc.

Has there been any shaft current test been done before starting the motor?

Only my opinion and theories and I could be totally wrong.

I have never seen a Cylindrical roller bearing on a direct coupled application. If it were me I believe I would try a ball bearing in its place. Do we know what bearing was in place before these failures began to happen 3 weeks ago ?

I have seen these bearings used in pumps where there is movement needed to set impellar or seal clearences, but never on motors. I believe this bearing could be wrong for this application. Just my 2 pennys worth.
Please let us all know what the outcome is.

if brg is lightly loaded the rollers maybe skating when out of load zone due to over lubrication by well intentioned oiler, may also explain the noise and roller marks that are produced as the roller regains traction and re enters the load zone, this will usually be accompanied by higher temps

I do not see anything wrong in principal in using a cylindrical roller bearing on NDE of a motor. It will handle thermal expansion away from the coupling without the outer ring having to move inside the bearing housing.

Bearing loading is another consideration and since cylindrical bearing can carry more load, its minimum loading requirement is also higher then that for other types. Normally loading on NDE bearing is expected to be less then that on DE. So, this might be a design problem in the discussed case as also mentioned by captnb. Unfortunately, there is no information on whether or not the bearing was overheating though.

I also have never seen a cylindrical roller bearing applied to the NDE of an electric motor. But as was pointed out here earlier, we have to assume that there was a good reason for this non-standard bearing configuration design demanded by the driven machine. It would be interesting to know the name of the motor manufacturer; is it a reputable company, established globally like GE, Teco-Westinghouse, ABB, Siemens, Reliance, US-Motors, WEG, Ansaldo etc. ?

My guess is that the bearing failures were caused by poor installation practices during motor overhaul. Let me give you an example from my own experience :

Prevention of Scuff-Marks when assembling Cylindrical Roller Bearings.

Before assembly, the inner race is located on its journal on the shaft.
The outer race, containing the roller elements, is inserted in the bearing housing which is part of the end shield ( in case of a motor or generator ).
The end shield is suspended from a crane. When slipping the outer race over the inner race the end shield must be positioned perfectly perpendicular to the shaft which is in general difficult to achieve. If it is not perfectly perpendicular the leading edges of the rolling elements can ( and often will ) produce scuff marks and scratches in the exposed surface of the inner race.
To avoid this kind of damage, which will remain undetected until the machine is put into operation, we machined a sleeve from a solid block of white nylon. Using a lathe, a hole is drilled to match the shaft diameter for a snug fit. The outer dimension of the sleeve is machined to match the outer diameter of the bearing's inner race ( perhaps a few thousands of an inch larger ). The finished sleeve is pushed over the shaft extension until it touches the inner race. Now, when assembling the two halves of the bearing, the rolling elements are gently guided by the sleeve into a perfectly perpendicular position, and when the leading edges of the rollers touch the inner race they cannot cause any scuffs or scratches.


Another possible reason ( as pointed out by David_G above) isw under-loading. NU - type bearings require a minimum amount of loading during operation. If this load is not sufficiently large and constant, the rolling elements tend to " skid ", thus precipitating early bearing failure.

Siemens 1LA8 series motors are available with a roller and ball (2 bearings) on the de and a roller on the nde. The ball takes the thrust and the two rollers take the load. Actually the housing for the ball bearing has about .020 clearance on the OD, it's just for thrust. A customer of mine has 12 of them direct coupled to fans (over hung loads). And another 6 that drive screw compressors.

MAN!!!

I am glad some have had dealings with this type bearing on the NDE of a motor. Some of us have just never seen one, but were not saying they did not exist.

I am sure it will work in the correct application so long as there is a thrust bearing somewhere in the motor. Duh!

If there is no load involved seems like the "skid" problem would be seen on the rollers also, but seems like there was non showing on the rollers in this case. In fact it looked like there was no damage anywhere except on the inner race, unless I overlooked it in the available pictures.

Looking at the outer ring and the cage, they may not seperate as we suggested. I am still thinking maybe a preheating problem at installation. The inner ring looks burnt in the picture.

ungereng's explanation makes pretty good sense to me and seems to agree with the pictures. If there are marks spaced at roller spacing, the possibilities are reduced. Skidding is not a candidate to produce that pattern. Electrical current would not likely produce marks spaced at roller spacing. False brinneling would look like that but we can rule it out since it occurred very fast. Installation damage is the only way I can think of to produce that pattern in these circumstances.

Hi,

Thanks for your feedback. i had written to my client regarding the installation procedures followed during assembly and disassembly of the bearing. i hope some of you are doubting regarding the installation procedure followed. however i had few questions,

1. This motor belongs to one of the reputed manufacturer here in India with the company being around 40 years old.
2. simillar 9 machines in the same floor is running fine.
3. however i was told that the frequency of failures of other machines are very less.
4. but whenever they open the motor NDE bearing of other units they see these line marks.
5. the client maintenance dept are also well experienced in assembling the bearings.
6. simillar motors of the same specs are running well in another steel plant just opposite to my client's site.
7. bearing temperatures are normal when they start hearing the abnormal noise from the motor NDE bearing side
8. do you mean to say that everytime they open the motor NDE bearing for replacement the wrong installation procedure is followed..???
9. I was told that they preheat the inner race while removing bearing from the casing. but i will comeback with more details.

hope i had given some more inputs for your reference.

Regards,