Sign up Calendar Latest Topics Donate
 
 
 


Reply
  Author   Comment   Page 2 of 2      Prev   1   2
electricpete

Sr. Member
Registered:
Posts: 651
Reply with quote  #16 

I think a given physical size size defect creates lower peak acceleration at slow speed, but degrades faster at high speed.

So knowing we have certain g’s on low speed machine, we envision larger physical defect than if it was on a fast machine, but we envision slower degradation than if it were the same physical size (which would also be higher g’s) on faster machine.

Which do we worry about more, considering there are some things working in different directions.

The CSI alarm limits for peakvue TWF true peak get lower as the speed gets lower.  So to me the reported g level more concerning knowing it is a slow speed machine.

 

I have case history of a 324rpm vertical motor lower bearing showing around 30g’s twf true peak-to-peak (not as severe as this one based soley on the peak-to-peak g value *)

When we pulled the bearing it was in some sad shape – inner race SEVERELY spalled around a large arc (slide 12) and cage broken (slides 9). 

 

More details discussed here:

http://www.maintenance.org/topic/monitoriong-a-degraded-bearing-was-fault-frequency-question

 

* Some things to consider to explain why this severe damage didn't create higher g's:

1 – it was an inner ring fault (outer ring would have given higher g’s)

2 – we weren’t using peakvue (peakvue gives higher true peak whereas traditional twf can miss the peaks if sample rate is too low).

3 - Since it was vertical motor guide bearing, there is no significant load on the bearing in our case (this defect likely would have made higher g's under radial load). 

4 – we didn’t have a trend of g’s (I agree with Rusty it can be a useful trend)

 
Attached Files
pdf cw11vibR1postSmall.pdf (1.17 MB, 11 views)

RustyCas

Avatar / Picture

Admin
Registered:
Posts: 1,810
Reply with quote  #17 
This has been mentioned recently somewhere else, but in terms of impact severity, I am not a fan of looking at acceleration levels as peak-peak. I think the maximum peak is a better representation of an impact.
__________________
"The trend is your friend"
MEllis

Member
Registered:
Posts: 10
Reply with quote  #18 
  Capture.png 

MEllis

Member
Registered:
Posts: 10
Reply with quote  #19 
Thanks for the comments and that is a good note on speed and how much wire is on the reel Rusty.  We have not been tracking that or attempting to keep it constant. 

Here are some updated graphs of the original ones I posted..  We are taking readings every week, next reading will be this Friday.  One graph is a lower frequency spectrum and the other is higher frequency.  Both in velocity as you can see.  I asked our contractor to give me a TWF with every reading from now on.  We are getting 1X side bands on the BPFO frequencies and notice the raised floor in the high frequencies. 

bcv.png 

nfgn.png 

 

electricpete

Sr. Member
Registered:
Posts: 651
Reply with quote  #20 

Quote:
We have some conflicting opinions on this across the company and I wanted to see what yall think

I hope everyone agrees the bearing needs to be replaced and it’s just a question of when. My opinion is that failure could easily be just around the corner for this bearing.

The motor lower bearing case study I linked above is my personal reference point to analyse this.    Roughly same speed (it was 324rpm).  Roughly same size bearing (it was MRC 148R  with 15” OD, although it’s an angle contact ball bearing rather than spherical).   It was similar pattern with respect to harmonics of a fault frequency and 1x sidebands around those harmonics (although our fault frequency was bpfi and yours is bpfo…. slightly different mechanisms for 1x sidebands).   Ours had not reached the point of creating an increased noise floor or other defect pattern.  Ours was less severe than yours by the true peak-to-peak acceleration metric, with some caveats mentioned above.  And as I said the damage found was SEVERE with intentional capital letters.  See slides 9,11 and especially 12.  The spall around the inner race extended most of the width of the race and more than half the circumference.  Metal flakes from that spalling spread throughout the grease and bearing, embedding in the cage.   The cage was completely broken in one place.  Maybe you can show the photos  to your coworkers for shock value if you are inclined to want to increase the sense of urgency.  

By the way, it's a guide bearing on vertical motor, so probably not much radial load. Axial load is low, just what is generated by the preload wavey washer.  If this had been a loaded bearing, failure probably would have progressed much faster.

If you look at my linked thread, you’ll see we were 
monitoring vibration over a month with little change. The bearing housing temperature was slightly hotter than the sisters by a few degrees initially.  At the end we were monitoring everything daily.  Bearing housing temperature started spiking erratically like from 130F to 150F in one day  and that temperature behavior is what finally convinced us to take the machine down for maintenance.

 
Attached Files
pdf cw11vibR1postSmall.pdf (1.17 MB, 11 views)

MEllis

Member
Registered:
Posts: 10
Reply with quote  #21 
electricpete, I did read your case study and it really helped me out.  Ours is not increasing temp much yet but we are monitoring temp every day and vibration once a week. 

Thank you very much for all of your feedback and to be honest, internally we have one guy saying it could be a year from now and me saying that we need to do it within the next 2 months.  We went ahead and ordered all parts needed which some have 4-6 week lead times.  It's almost gotten to be a pissing match at this point!  The plan from above right now is to monitor it monthly, even though im monitoring it weekly, and see what happens.  T
Barry

Avatar / Picture

Member / Supporter
Registered:
Posts: 132
Reply with quote  #22 
IMHO never apply the alarm criteria to vertical roller thrust bearings as one would for horizon loaded bearings. Vertical bearings have or should have a lode zone the full circumstance as opposed to a horizontal loaded bearing that has very narrow load zone.
__________________
Barry
John from PA

Sr. Member
Registered:
Posts: 963
Reply with quote  #23 
Quote:
Originally Posted by electricpete


I hope everyone agrees the bearing needs to be replaced and it’s just a question of when. My opinion is that failure could easily be just around the corner for this bearing.



Like e'pete I always feel that when you have well established evidence that a bearing is failing, it all too often occurs rapidly.  One common characteristic is in the later stages, the vibration will start to decrease.  That is particularly true of inner and outer race faults.  Unfortunately a walk around program might not detect the trend due to the route interval.

Good luck moving forward and keep us posted.

  

electricpete

Sr. Member
Registered:
Posts: 651
Reply with quote  #24 

Quote:
Setting my vertical alarm lower in a roundabout way increases the severity we predict for op’s bearing.


I agree the configuration makes a difference in a number of ways, not always so straightforward to compare.   Personally, I conclude vertical guide bearing true peak (or pk/pk) acceleration alarm should be lower.  Let me know if you disagree. 

In my mind any vertical guide bearing has lower load (in this specific case, it has no radial load other than misalignment and very small axial preload from wavewasher).

Lower load (like lower speed) means less impacting vibration (peak acceleration) for a given physical defect size, but more “life” left in the bearing for a given physical defect size (slower degradation for a given physical defect size).   These two factors act in opposite directions in terms where we might set our alarm.   But if I follow CSI’s approach to lower speed (which acts similar to lower load in these two respects), then I'd select lower alert limit. It means we apply more weight to the present assumed physical defect size than we would to future speed dependent/load dependent degradation rate of a given physical defect size defect.

There is also the question of how much the impacting force gets transmitted from the impact to the particular monitoring location.  If we are monitoring in load zone of a normal horizontal machine, we get a better transmission of vibration from the impact to the point of monitoring. Contrast to my case with somewhat axial loading or other vertical guide bearings with unknown loading, we get worse transmission to the point of monitoring and accordingly we would set alarm on the vertical guide lower… IN GENERAL. 

EDITED - DELETED THE FOLLOWING:

Quote:
 In contrast to “in general”, I have to say I just noticed the op probably has stationary inner ring and rotating outer ring (correct?). That’s not the general case I imagined.  For rotating outer ring, I guess you have to  monitor on a stationary piece associated with inner ring.  In that case his bpfo pattern with 1x sidebands pattern resembles my bpfi with 1x sidebands pattern (and you can ignore one of the caveats from my first post).  In both his case and my case we are monitoring the stationary piece while the defect is rotating.   Defect moves around with respect to the monitoriong point and comes near it once per revolution (one good reason for 1x sidebands).   Assuming the weight load is directed downward and he places his accelerometer on the bottom of the stationary piece associated with inner ring, he still has the benefit of monitoring in the load zone where I don’t (which again suggests my vertical alarm would be lower).


TLDR:  in this post, I generally concluded vertical machine guide bearing alarm would be lower. In a roundabout way, this mostly increases the urgency we assign  for op’s bearing when comparing to my case study.  

John from PA

Sr. Member
Registered:
Posts: 963
Reply with quote  #25 
Quote:
Originally Posted by electricpete
In contrast to “in general”, I have to say I just noticed the op probably has stationary inner ring and rotating outer ring (correct?). 


Are you referring to the machine by OP = "MEllis".  If so what made you draw that possible  conclusion?  
MEllis

Member
Registered:
Posts: 10
Reply with quote  #26 
The inner race is rotating, outer is stationary.  Typical horizontal shaft. 
electricpete

Sr. Member
Registered:
Posts: 651
Reply with quote  #27 
Sorry, I misunderstood about that piece.  So I just went back and highlighted in grey the piece of my response that should be ignored/deoeted
Thanks
Danny Harvey

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,403
Reply with quote  #28 
Sorry I missed this one.
Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.