Sign up Calendar Latest Topics Donate
 
 
 


Reply
  Author   Comment  
fburgos

Sr. Member
Registered:
Posts: 670
Reply with quote  #1 
I already call for an inspection/replacement of this bearing, we have 3 fans but only two operate at the time.

Induced draft fan between bearings (SKF22340)
891rpm and 800kw 

For fan and pumps im using the wizard and triax with CSI-2140, and trust the peakvue setting.

I want to understand why the signal is "weak" in horizontal and peakvue (primary axis) but very clear and strong on vertical and axial, what could be the cause of this, is my peakvue setting "bad".

This is my vibration "history" from 5/24/2017 to 02/15/2018 only overalls are available lost the database and was with a different collector it was really hard to analize and report on that.

From motor NDE to Fan out board, bearings 1, 2, 3 and 4.
Capture.png 





 
Attached Files
docx Bearing 3, coupling side.docx (425.19 KB, 33 views)

Danny Harvey

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,402
Reply with quote  #2 
Ferdy,

Keeping in mind that only the horizontal axis is measuring radially, I would think that the vibrations you are measuring are not being emitted in that direction.

Whenever I see vibrations at bearing defect frequencies that appear in the velocity spectrum with 1 x bearing defect frequency predominant, I suspect some sort of impingement.  Most common is probably a high point on the outer race but a cocked bearing will also do it and it can happen with the inner race, too.  A cocked inner race would generate axial vibration and might excite the vertical, too since it is tangentially mounted instead of radially.

It certainly does not look like normal wear, though.
Ralph Stewart

Avatar / Picture

Sr. Member
Registered:
Posts: 235
Reply with quote  #3 
I see where you have editted your original posting. This may answer my questions below. [smile]

Danny said "It certainly does not look like normal wear, though."

Am I missing seeing something? All I see in the attached file is a picture of the bearing itself.

Are there spectrums showing somewhere that, for some reason, I can not see on my computer? I see where you said you lost your database, and does this mean there are no spectrums available?

Quote:
fburgos quote:
I already call for an inspection/replacement of this bearing, we have 3 fans but only two operate at the time.

Induced draft fan between bearings (SKF22340)

891rpm and 800kw For fan and pumps im using the wizard and triax with CSI-2140, and trust the peakvue setting.

I want to understand why the signal is "weak" in horizontal and peakvue (primary axis) but very clear and strong on vertical and axial, what could be the cause of this, is my peakvue setting "bad".

This is my vibration "history" from 5/24 to 15/2 only overalls are available lost the database and was with a different collector it was really hard to analize and report on that.


I may not be understanding all you are asking, but where you say "I want to understand why the signal is "weak" in horizontal and peakvue (primary axis) but very clear and strong on vertical and axial, what could be the cause of this, is my peakvue setting "bad".", what is this "weak" referring to? Is this the only question you are asking?

Sorry for my confusion.

Thanks and Have a Great Day,
Ralph

__________________
Ralph Stewart
AlertAnalytical.com
fburgos

Sr. Member
Registered:
Posts: 670
Reply with quote  #4 

I'm measuring triax, primary axis in radial directly on the bearing-metal interface.

Opened bearing, and we can see a lot of mud and water in the oil chamber, if I take a look closer I can see some shiny small pieces In the mud and oil, I believe is metal from bearing.

could not see any scratches on the rollers or outer race, inner race view is covered by metal cage.

if this indeed is metal from bearing inner race, i think my peakvue should have higer amplitues, or my measurement point is it too far from load zone?

Attached Images
jpeg FJIMG_20180613_084436.jpg (184.20 KB, 15 views)
jpeg FJIMG_20180613_084615.jpg (237.90 KB, 16 views)
jpeg FJIMG_20180613_101403.jpg (291.93 KB, 15 views)

fburgos

Sr. Member
Registered:
Posts: 670
Reply with quote  #5 
Quote:
Originally Posted by Ralph Stewart
Are there spectrums showing somewhere that, for some reason, I can not see on my computer? I see where you said you lost your database, and does this mean there are no spectrums available?


lost my old database during a power failure or computer unexpeted shutdown, database opens but all spectrums are crazy bad, this was from another data collector, got my CSI240 on may, I have just one set of spectrums from this survey.


yes my question is why peakvue is not clear on the failure pattern, my location or the peakvue setting is bad?

Attached Images
png img00012.png (50.61 KB, 13 views)
png img00013.png (53.02 KB, 13 views)
png img00014.png (38.19 KB, 13 views)
png img00015.png (35.97 KB, 14 views)

RustyCas

Avatar / Picture

Admin
Registered:
Posts: 1,810
Reply with quote  #6 
I don't think PeakVue will show "inner race" problems too well because there is no direct path to your sensor.  PeakVue is looking at "stress waves" (or so they say) and the path from the inner race to the outer race will not pass those stress waves, imo.  Perhaps on large, heavy, slow turning units it would, where there is heavy impacting.  But on a higher speed unit with lighter loading on the bearings, I think not.

Or I could be wrong.

__________________
"The trend is your friend"
OLi

Sr. Member
Registered:
Posts: 1,916
Reply with quote  #7 
It could also be that this is an older damage that flatten out and the bulk of the signal moved down under the filter setting of the PV... I do see that sometimes as I get a higher freq. data accel HP 2kHz and compare to velocity HP 600Hz and you can see the new damage getting higher in accel and if you wait a couple of years it will lower and get higher in the lower freq. and in the final stage also increase the 10-1000Hz range data but then it is final. Not always, rarely the same behavior but in general, it happens.
__________________
Good Vibrations since early 1950's, first patented vibrometer 1956 in the US.
http://www.vtab.se
Danny Harvey

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,402
Reply with quote  #8 
As I understand it, a portion of the high frequency "stress wave" is reflected back every time they pass through an element. So in the case of an inner race defect, it may show ONLY in the load zone because out of the zone, there is clearance and very little signal will transmit through the gap. But even then, it must pass from the inner race, through the oil film to the roller, then through the oil film and into the outer race, then through the outer race and into the housing and then into you sensor if you can mount directly there. That's a lot of losses.  So I look for the most direct path to the bearing which might not necessarily be the closest.  But quite often the load zone is in the one place that you cannot mount your sensor.  (That's why on spherical roller bearings in pillow block housings, the best signal often comes in the axial direction in the load zone. My belief is that if you could get to it, radial in the load zone would be best but we take what we can get in most cases.)

So I almost always take BPFI signals in PeakVue or Demod much more seriously than BPFO.


Danny Harvey

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,402
Reply with quote  #9 
Ferdy,

If you still have the oil you can put it in a sample jar and turn the jar upside down and leave it on a magnet for a couple of hours.  It's kind of a poor man's wear particle analysis.

And I wonder about the security of having such a fine resolution image of your fingerprint on the internet.  I can't imagine how it could be used for evil, but I'll bet somebody can.
Ralph Stewart

Avatar / Picture

Sr. Member
Registered:
Posts: 235
Reply with quote  #10 
fburgos,

Thanks for the replies. [smile]



Quote:
"I want to understand why the signal is "weak" in horizontal and peakvue (primary axis) but very clear and strong on vertical and axial, what could be the cause of this, is my peakvue setting "bad".




Why does the "Inner Race Signal" appear to "die" after the 2x BPFI position in the Velocity data? There looks like there is no 3x BPFI in the spectrum at all in the horizontal position, or any positions.

My guess in your situation is, after you saw what was possibly an Inner Race pattern in the vertical and axial directions, the Peakvue data should have been taken in the vertical or axial directions, in my opinion.

I am not sure if you noticed this condition on the meter display (BPFI in the Vertical and Axial) while in the field taking the original data, but from my experience, I personally would have acquired "extra" Peakvue data (with the "Acquire" function of the meter) and placed my  transducer in the true Vertical and true Axial positions. The use of a "triaxial" transducer and the Horizontal Position assigned to the Peakvue data to be collected, might have had some bearing or result that is related to the question you asked "I want to understand why the signal is "weak" in horizontal and peakvue (primary axis) but very clear and strong on vertical and axial, what could be the cause of this, is my peakvue setting "bad"."

Just my opinion and I could be totally wrong.

Thanks and Have a Great and Safe Day,
Ralph




__________________
Ralph Stewart
AlertAnalytical.com
fburgos

Sr. Member
Registered:
Posts: 670
Reply with quote  #11 
Did not measure any axial peakvue or velocity because I didnt notice on the moment, I was under heavy rain and I could go back if i need to, in office I considered that what I have was good to call for a shut down.

Yesterday after cleaning bearing #3 vibration is lower, I think a pice of "debris" could squeeze the rollers but not impact to generate stress waves (once I had oval bearing housing/outer race no impacts on demodulation but velocity spectrum had strong signal), some of this days will check and update.



Capture.png 


Attached Images
png img00000.png (29.63 KB, 9 views)
png img00001.png (55.83 KB, 9 views)
png img00002.png (26.63 KB, 9 views)
png img00003.png (28.00 KB, 7 views)
png img00005.png (29.46 KB, 7 views)

Ralph Stewart

Avatar / Picture

Sr. Member
Registered:
Posts: 235
Reply with quote  #12 
Thanks for the good information return.

Ralph

__________________
Ralph Stewart
AlertAnalytical.com
Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.