ad banner
  Reliability Magazine« Message Boards
  Alignment/Balancing/Vibration Forum
  High Axial Vibration on Outer Race Freq. for Electric Motors

Post New Topic  Post A Reply
profile | register | preferences | faq | search

UBBFriend: Email This Page to Someone! next newest topic | next oldest topic
Author Topic:   High Axial Vibration on Outer Race Freq. for Electric Motors
kseyler
Member
posted 07-04-2001 12:19 PM     Click Here to See the Profile for kseyler   Click Here to Email kseyler     Edit/Delete Message   Reply w/Quote
Hello,
We have been seeing a lot of a/f bearing motors (900 to 1200 RPM and 2000 to 6000 HP) with outer race frequencies. The outer race frequency has been showing up at 1x BPFO, 2x BPFO and 4x BPFO. Usually the 1x or 2x is the highest. This is mainly seen in the axial direction with much, much lower levels on the Hor. and Vert. The levels can be in the range of up to 0.07 in/s at a specific frequency. We do not have an accelerometer to check the acceleration. We are seeing this on the initial run-up under no-load conditions. These are brand new bearings. We have had one of the bearings removed and sent out for analysis. On the particular bearing we sent out there was one very tiny mark on one of the balls. That doesn't seem to correspond with the outer race frequency that we have been seeing. We do follow very good practices for installing the bearing and the cartridge.

1) What do you recommended in regards to runout for the outer race when the indicator is rotating with the shaft? What is the recommended runout on the cartridge?
2) Is it possible that this outer race frequency is being seen due to the fact that the motor is not being loaded? 1/2 of the rotor weight is more then triple the recommended bearing minimum load.
3) The bearings being used are 6332M-C3. Is it possible that under this cool no-load run (Brg. Temp. Rise approx. 10 deg.C) in which the rotor and everything else is very cool, the clearance is too large in the bearing and we are seeing this outer race frequency due to excessive clearance?
4) The bearings use an outer ring/bearing cartridge fit that is on the larger clearance size due to the expansion of the bearing/shaft since its a large electric motor. Is is possible that during the no-load runs this clearance is a little too much?
5) Is it possible that poor bearing alignment could cause a large axial outer race frequency?
6) I do not believe that the bearings are being damaged during installation or prior to installation. Is it likely that these bearing frequencies are being seen even though there is no damage? If so, please list any potential causes that may show outer race frequencies without actual damage.

Any other suggestions or comments would be appreciated. If you need any further information just ask and I will let you know.

Thank you,
Kevin

IP: Logged

Arne.Lindholm
Member
posted 07-04-2001 07:07 PM     Click Here to See the Profile for Arne.Lindholm   Click Here to Email Arne.Lindholm     Edit/Delete Message   Reply w/Quote
Dear Kevin, recent experience from root cause research on this issue has convinced me that the mounting is the main problem when you see x1 BPFO and multiples already from start.
1) Runout should be nothing readable. (shaft holds the gauge, the tip measures radially.
But danger is that when you later will close up, tighten bolts, the actual deformation takes place.
2) I do not believe that this comes from lack of load.
3) With just a 10C raise it is still almost C3, but that does not explain 1xBPFO, etcetera. A perfectly round bearing that sits 90 degrees to the shaft does not show any BPFO at all.
4) I agree. Are you sure that you must take a larger clearance (outer ring to housing)? The outer ring will expand almost the same as the housing, or is that of aluminium? 5) Yes. 6) I agree that damage during mounting does not seem likely. Apart from the fact that I believe that the outer ring is not any more round and/or is skewed to the shaft. Can you arrange to run to full temp and arrange a way to measure without removing bolts? - Stop and check axially to the outer ring and radially at the rings edge? I would recommend to look for a reason why the bearing deforms to get out of round.
A scetch or sectional drawing would be a great help to understand the situation. Any format. Best regards Arne

IP: Logged

kseyler
Member
posted 07-05-2001 09:07 AM     Click Here to See the Profile for kseyler   Click Here to Email kseyler     Edit/Delete Message   Reply w/Quote
Hi Arne,

1) Recently we had a motor with a 0.006" runout on the outer ring. We adjusted it down to 0.002" but there was no significant decrease in vibration. The type of assemblies that we have always have the cartridge bolted to the shield so the only thing that isn't on is the outer cap during the runout check. I would assume there should be no deformation when the outer cap is later bolted on.
3) In regards to a perfectly round bearing, one thing that I didn't mention that I was wondering is if people have seen outer race due to oval bearing housings? I believe I had read that somewhere before but I hadn't found the e-mail. With an out of round housing is it likely to see multiples of outer race as well?
4) With regards to the outer ring/housing fit I believe this is the fit that was determined with bearing manufacturers and our engineers a number of years ago. I will be reviewing these fits with the bearing manufacturers again. With a larger clearance could an outer race frequency show up though?

We should be doing a full-load run on a motor early this fall.

Thanks for the help,
Kevin

quote:
Originally posted by Arne.Lindholm:
Dear Kevin, recent experience from root cause research on this issue has convinced me that the mounting is the main problem when you see x1 BPFO and multiples already from start.
1) Runout should be nothing readable. (shaft holds the gauge, the tip measures radially.
But danger is that when you later will close up, tighten bolts, the actual deformation takes place.
2) I do not believe that this comes from lack of load.
3) With just a 10C raise it is still almost C3, but that does not explain 1xBPFO, etcetera. A perfectly round bearing that sits 90 degrees to the shaft does not show any BPFO at all.
4) I agree. Are you sure that you must take a larger clearance (outer ring to housing)? The outer ring will expand almost the same as the housing, or is that of aluminium? 5) Yes. 6) I agree that damage during mounting does not seem likely. Apart from the fact that I believe that the outer ring is not any more round and/or is skewed to the shaft. Can you arrange to run to full temp and arrange a way to measure without removing bolts? - Stop and check axially to the outer ring and radially at the rings edge? I would recommend to look for a reason why the bearing deforms to get out of round.
A scetch or sectional drawing would be a great help to understand the situation. Any format. Best regards Arne

IP: Logged

davmar48
Member
posted 07-05-2001 04:03 PM     Click Here to See the Profile for davmar48   Click Here to Email davmar48     Edit/Delete Message   Reply w/Quote
quote:
Originally posted by kseyler:
If so, please list any potential causes that may show outer race frequencies without actual damage.


Kevin and Arne,

One of the explanations can be found in Vibrations, September 1999, Volume 15, Number 3. Article titled: " Is it really an outer race problem?" In authors' opinion an impact takes place between the cage and the ball even in a perfectly good bearing when each of the balls is passing the load zone. That is occurring at Nballs) x (FTF) rate, which is equal to BPFO according to the well known equation.

A question for Kevin. Correct me if i am wrong, but I've got the impression that you measure the outer race runout by having the base of a dial indicator attached to the shaft and indicating the race by rotating the shaft? Isn't this true that the trajectory of the base (a circle in a perfect world ) due to the static load is eccenric to the outer race? If this is correct, it explains why you are seeing "out of round shape" of the outer race.
Arne, even if the outer race is trully out of roundness, I have hard time picturing why would BPFO appear. Is it similar to an impact in case if race is really damaged?
Isn't this true that BPFO would rather modulate 1xRPM or/and 2xRPM then appear itself?

David

IP: Logged

kseyler
Member
posted 07-05-2001 04:34 PM     Click Here to See the Profile for kseyler   Click Here to Email kseyler     Edit/Delete Message   Reply w/Quote
Hi David,

The runout check is taken with a dial indicator base sitting on the shaft and the indicator on the side of the outer race. This runout check records the axial runout of the bearing in respect to the shaft. The deflection from the rotor weight in this case would be approximately 0.0005" or less for a 5000 lb. rotor. I would assume this may have a small impact on the axial runout check but not too large.

As I stated in the last message with regards to improving the runout from 0.006" to 0.002": I believe that we currently have some instances in which the face of our frame is not completely perpendicular to the bore as well that the frame end rings are not parallel. On the case stated above, I added a 0.006" shim between the shield and frame (at the bottom of the shield.) That is when the runout improved from 0.006" to 0.002". There was no significant decrease in vibration from this improvement though.

If anyone has any further information on why BPFO can be seen without any actual damage please let me know.

Thanks again,
Kevin

IP: Logged

Arne.Lindholm
Member
posted 07-05-2001 07:14 PM     Click Here to See the Profile for Arne.Lindholm   Click Here to Email Arne.Lindholm     Edit/Delete Message   Reply w/Quote
Hello David, with due respect for the background for that article, I wonder why on our swedish railway frequency converters at 220 ton rotor mass sitting in double row spherical roller bearings at 110 ton load each, there is not even a hint of BPFO. C3 play, running very cold or very hot pending situation. Does not matter. And similarly, why do I not see the same on large yankee cylinders at past 100 ton load.

But when by mistake or deliberately I arrange an ovality, the BPFO pops up. From large diameter thin bearings to heavy duty bearings, roller and ball.

Reason for 1xBPFO is that the motion is nearly sinusoidal. A pending damage with an inch or so spalling has a family of BPFO peaks at x5 to x12. The more the spalling grows around, the lower the hump moves. The final, when most of the outer ring has the spall throughout the load zone, the x1 and x2 starts to dominate. I have a case here where the cage is broken and rolls where tumbling and only massive greasing prevented a seizure. The spectrum is full of x1, x2 and x3 BPFO and a lot of noise carpet and above 10 g broadband noise. Best regards Arne

IP: Logged

Jan Krepela
Member
posted 07-06-2001 10:23 AM     Click Here to See the Profile for Jan Krepela   Click Here to Email Jan Krepela     Edit/Delete Message   Reply w/Quote
Dear Kevin,
We had just recently had a motor in the shop for repairs. Final vibration signatures contained clearly BPFO on both inboard and outboard bearings in all directions, but it was the largest in the axial direction. The motor:900HP,3600RPM, bearings 6216 C3. Your description of the problem fits our motor almost perfectly. The only diffrence is, that we could see 1xBPFO plus 11 harmonics-on both bearings.
I believe I had an advantage compared to you, seeing this problem before - in 1999. Again it was an electric motor, DC this time, with large NU bearings (identical on both ends). We had spend lots of time trying to fix the problem. After taking the machine apart several times, replacing both bearings with new ones flown directly from the manufacturer, we had decided hat there is no problem. At that time I wrote the article mentioned by David. ("Is it really an outer race problem?", Vibrations, Sept.1999, Volume 15). Since than I have seen the same phenomenon many times.
In discussions with fellow vibration analysts the most frequent possible explanation was the out-of-roundness. But the large number of harmonics always points to an impact. The outer race out of round can never generate an impact. If for example the outer race is deformed into an ellipsoid with horizontal axis larger than the vertical, then the rotating part of the bearing sees an outer race of a slightly larger diameter. No reason for impact, no reason for BPFO. But the impacting of the cage by the rolling element in the loading zone will always be present. As far as I know, this impacting is more pronounced when the motor runs loose on the test bed than when bolted to the base and coupled to the driven machinery. Is it due to the increased stiffness and /or different loading in the load zone? Probably.

jank

IP: Logged

kseyler
Member
posted 07-06-2001 10:55 AM     Click Here to See the Profile for kseyler   Click Here to Email kseyler     Edit/Delete Message   Reply w/Quote
Hi Jan,

In my case we usually either see 1x and/or 2x BPFO and sometimes 4x at low amplitudes. Have there been any published papers or has anyone else actually seen any of these frequencies show up due to out of round?

In situations like this when this BPFO is seen is it sufficient to explain that this is due to no-load and being un-coupled? Any further explanation would be appreciated. Any things that can be done during a no-load test to help ensure these amplitudes are as low as possible?

The other thing in regards to this BPFO that we see: Why is it that it mainly shows up in the axial direction? The Hor. & Vert. sometimes have the BPFO but at very low amplitudes. Could this be caused due to the fact that on A/F motors they are not shimmed completely for magnetic center and the axial magnetic pull helps add to these forces?

Thanks,
Kevin

IP: Logged

Jan Krepela
Member
posted 07-09-2001 11:05 AM     Click Here to See the Profile for Jan Krepela   Click Here to Email Jan Krepela     Edit/Delete Message   Reply w/Quote
Hi Kevin,
When looking for harmonics, do you use the program to locate them? I am using the CSI equipment, and the Harmonic Analysis option located lots of harmonics I did non know existed.
I am not avare of any paper on this subject.
Customers have always been reluctant to accept a motor that looks substandard. You will have lots of convincing to do. But I believe the best way is to make them avare that problem like this does exist (if we should call it "a problem"). If it is possible to go back to your motors with "BPFO" in the field and analyze them would be a big help. I have a great difficulty accepting the out of roundness as a cause for the cases I have seen. Especially on motors with identical bearings on both ends and there were several of those.
Yes I have tried to rig up a belt load. No use (it is hard to do a belt drive on something like 1500Hp motor). I tend to believe that the education will have better success in the long run than trying to minimize the "BPFO" during the test runs.
The same as you, also our experience shows the highest amplitudes in axial direction. Looking back on those cases, the axial stiffness was always only a fraction of the horizontal or the vertical stiffness.
The axial magnetic pull is virtually negligible for reasonable out of mag. center positions.

jank

IP: Logged

kseyler
Member
posted 07-17-2001 11:14 AM     Click Here to See the Profile for kseyler   Click Here to Email kseyler     Edit/Delete Message   Reply w/Quote
Hi Jan,

With regards to the harmonics: I currently just calculate the harmonics of BPFO, BPFI, FTF, etc. and then just locate them on the spectrum.

Part of the changes we will make will include an improvement of the shield stiffness. I have some other ideas but nothing right now that I think will work for sure.

It seems odd to me still that if this is a very common problem that there aren't more people that discuss it and write papers on it.

Thanks for the help

IP: Logged

Jan Krepela
Member
posted 07-19-2001 11:55 AM     Click Here to See the Profile for Jan Krepela   Click Here to Email Jan Krepela     Edit/Delete Message   Reply w/Quote
Hi Kevin,
I think that the main problem with those "BPFO" frequencies is that they are not a big problem. I do not believe they are causing any failures (just my guess). And if they do not cause failures, why would anybody spent too much time on them. At the same time it might be useful to know that something like that does exist.

Jan

------------------
jank

IP: Logged

kseyler
Member
posted 07-19-2001 12:59 PM     Click Here to See the Profile for kseyler   Click Here to Email kseyler     Edit/Delete Message   Reply w/Quote
Hi Jan,

I haven't heard of any of our customers having bearing failures for a long time. The only one that I can remember hearing about was a lubrication problem. Hopefully I can look at reducing these levels somewhat and then besides that I will just need to be able to explain it to our customers if they have concerns.

Thanks,
Kevin

IP: Logged

Chris Olsen
Member
posted 07-19-2001 04:41 PM     Click Here to See the Profile for Chris Olsen   Click Here to Email Chris Olsen     Edit/Delete Message   Reply w/Quote
The problem you describe sounds like a case I had in 1991 on a 400 HP grinder motor. The vibration was a major concern to the customer and the final fix was an adjustment to the internal clearances on the taper lock bearing. That fix came after replaced they bearings (twice) and housing (once).

------------------
Chris Olsen
GLViB.com
tel(513)755-6673
pager(800)398-8593

IP: Logged

All times are ET (US)

next newest topic | next oldest topic

Administrative Options: Close Topic | Archive/Move | Delete Topic
Post New Topic  Post A Reply
Hop to:

Contact Us | Reliability Magazine« Home

Except for message content, this material is Copyrightę 2000 by Industrial Communications, Inc. All rights reserved


Ultimate Bulletin Board 5.45b