Sign up Calendar Latest Topics Donate
 
 
 


Reply
  Author   Comment  
JohnJ

Member
Registered:
Posts: 66
Reply with quote  #1 
I have a pump that I am told has about 2.5 mils physical shaft run-out at the non-drive-end bearing.  Seems like I remember API has a limit that you are allowed to compensate run-out data.  If that is correct, can someone please tell me what that limit is?

Thanks,
John J
John from PA

Sr. Member
Registered:
Posts: 963
Reply with quote  #2 
I assume by the question that you are monitoring with proximity probes?  That being said, the 5th Edition of API 670 states

Quote:
6.1.2.2 The surface areas to be observed by the probes (probe areas):

d) shall be properly demagnetized or otherwise treated so that the combined total electrical and mechanical runout does not exceed 25 % of the maximum allowed peak-to-peak vibration amplitude or 6 ìm (0.25 mil), whichever is greater.
electricpete

Sr. Member
Registered:
Posts: 647
Reply with quote  #3 
0.00025" ? Wow, that's a small runout. 
John from PA

Sr. Member
Registered:
Posts: 963
Reply with quote  #4 
Quote:
Originally Posted by electricpete
0.00025" ? Wow, that's a small runout. 


Quote:
25 % of the maximum allowed peak-to-peak vibration amplitude or 6 ìm (0.25 mil), whichever is greater


The 1/4 mil is readily attainable on most ground journal/probe target areas.  There are some exotic steels that are always a problem, some stainless for instance.  In this case the OP has an issue with a pump and a common problem is a plating of the bearing journal with chrome as a repair procedure.  If the repair shop is not aware of the probe target area location, it too often gets plated driving proximity probe nuts.
JohnJ

Member
Registered:
Posts: 66
Reply with quote  #5 
John PA,

Yes, it is proximity probes.  Neither the shaft nor bearings have been out of this pump since installation about two years ago.  And there is very little run time on the motor/pump.  They start it up once a week just to rotate spares, but they only run the motor-driven pump for a short time because if electrical grid concerns (not sure what that means). The orbits and shaft centerline plots show the shaft running outside the bearing clearance at the top of the bearing during steady state operation and out the lower right quadrant during coast down.  See the attached file and you mayl understand my concerns about compensating the data.

Thanks,
John J

 
Attached Files
pdf Uncompensated orbits.pdf (48.33 KB, 24 views)

John from PA

Sr. Member
Registered:
Posts: 963
Reply with quote  #6 
Quote:
Originally Posted by JohnJ
And there is very little run time on the motor/pump.  They start it up once a week just to rotate spares, but they only run the motor-driven pump for a short time because of electrical grid concerns (not sure what that means). The orbits and shaft centerline plots show the shaft running outside the bearing clearance at the top of the bearing during steady state operation and out the lower right quadrant during coast down.  See the attached file and you may understand my concerns about compensating the data.

Thanks,
John J


Usually grid concerns occur with the starting of a relatively large motor which can impact the peak demand charge.  But I see in your shaft centerline (SCL) data about 4 mils clearance so the shaft size can't be that big, probably something like 2.5 inches.

I also note your comment about not being run too often.  You might pull a probe and look down the probe mounting hole and examine the surface while turning.  It would not be the first time that some corrosion has occurred on a shaft.

The reference data for compensation should be acquired on a shutdown.  In light of the mechanical or electrical runout you may have , don't use a DC gap voltage at a stopped condition, rather utilize a point at 100 to 150 RPM.

I like to think of shaft centerline plots as supporting information.  If I have other information pointing to a wiped bearing and the SCL plot supports it, then I might disassemble the machine.  But I would not likely disassemble a machine based solely on the SCL.  

 
JohnJ

Member
Registered:
Posts: 66
Reply with quote  #7 
John PA,

The pump is run every week but only for a short time. Corrosion seems unlikely.

The spectral data show high 2X (higher than the 1X component) along with several high level harmonics (0.8 mil at 1X and 1.0 mils at 2X).  The orbits show external loops along with truncation.  I know I didn't include that information in the previous posts. Would that be enough evidence of a wiped bearing?  If not, can you tell me what else would you expect to see with a wiped bearing?

Quote: The reference data for compensation should be acquired on a shutdown.  In light of the mechanical or electrical runout you may have , don't use a DC gap voltage at a stopped condition, rather utilize a point at 100 to 150 RPM.

I probably should have said this before.  The pump is an API centerline mount pump. 

The coastdown data were taken after about an hour and a half of operation.   I realize that is not enough time for complete temperature stabilization, but I don't understand using the gap voltages at 100 to 150 rpm instead of stop voltages. There is about 0.200 volt (1 mil) difference between the stop gap voltages and the 108 rpm gap voltages.  To me, that means the shaft is off the bottom of the bearing.  Can you please explain to me why I should use the DC gap voltages at 100 to 150 rpm.

Thanks,
John J
John from PA

Sr. Member
Registered:
Posts: 963
Reply with quote  #8 
Quote:
Originally Posted by JohnJ
John PA,

The pump is run every week but only for a short time. Corrosion seems unlikely.

The spectral data show high 2X (higher than the 1X component) along with several high level harmonics (0.8 mil at 1X and 1.0 mils at 2X).  The orbits show external loops along with truncation.  I know I didn't include that information in the previous posts. Would that be enough evidence of a wiped bearing?  If not, can you tell me what else would you expect to see with a wiped bearing?

Quote: The reference data for compensation should be acquired on a shutdown.  In light of the mechanical or electrical runout you may have , don't use a DC gap voltage at a stopped condition, rather utilize a point at 100 to 150 RPM.

I probably should have said this before.  The pump is an API centerline mount pump. 

The coastdown data were taken after about an hour and a half of operation.   I realize that is not enough time for complete temperature stabilization, but I don't understand using the gap voltages at 100 to 150 rpm instead of stop voltages. There is about 0.200 volt (1 mil) difference between the stop gap voltages and the 108 rpm gap voltages.  To me, that means the shaft is off the bottom of the bearing.  Can you please explain to me why I should use the DC gap voltages at 100 to 150 rpm.

Thanks,
John J


High 2X in the runout is a bit surprising.  Sometimes that means a magnetized shaft; sometimes it can be a coupling issue.  I would still be inclined to visually check the probe target area, unless of course it is a lot of grief.

Grabbing slow roll DC values at low RPM, as opposed to a stopped condition, provides some averaging of the voltage.  If you use a stopped condition just where have you stopped?  You say you have 2.5 mils of slow roll.  Are you at the peak or the minimum?  2.5 mils of slow roll is a lot in an SCL plot that has a 4 mil clearance circle.
JohnJ

Member
Registered:
Posts: 66
Reply with quote  #9 
John PA,


Quote:
High 2X in the runout is a bit surprising.  Sometimes that means a magnetized shaft; sometimes it can be a coupling issue.  I would still be inclined to visually check the probe target area, unless of course it is a lot of grief.

Grabbing slow roll DC values at low RPM, as opposed to a stopped condition, provides some averaging of the voltage.  If you use a stopped condition just where have you stopped?  You say you have 2.5 mils of slow roll.  Are you at the peak or the minimum?  2.5 mils of slow roll is a lot in an SCL plot that has a 4 mil clearance circle.


The measured runout is actually 2.26 mils overall at 100 rpm and the 1X and 2X from the spectrum is 0.901 and 1.18 mils respectively.  The time waveforms are pretty smooth, even at 100 rpm.  I would expect them to look like there were a lot of little scratches if there were corrosion.  I will try to get the probe area examined, but I doubt they will want to do that until they have a planned outage.

My understanding about using stopped gap voltages for the SCL, is that the shaft should be at bottom dead center (I know that doesn't have to be true in all cases).  I referenced the gaps at 100 rpm and plotted the SCL and compensated orbits with bearing clearance in the attached file.  I also have a plot of the coastdown waterfall with the 2X referenced with the dark line and the 1X referenced with the green line.  Note that the spectral data cannot be compensated in my software.  The SCL and orbits are still outside the bearing clearance. but changed positions to the left.

I might not worry too much about the shaft running outside the bearing clearance by 1 mil except it running out the top of the bearing.  The impellers are between bearings and the DE bearing runs within the clearance but just a little higher than expected.  I have a hard time believing that coupling misalignment (Thomas series 71) could cause the NDE of the shaft to run at the top with between bearings impellers (2 stage).  I am suspecting bearing misalignment.  What are your feelings?

Thanks,
John J


John from PA

Sr. Member
Registered:
Posts: 963
Reply with quote  #10 
Quote:
Originally Posted by JohnJ
Neither the shaft nor bearings have been out of this pump since installation about two years ago.  And there is very little run time on the motor/pump. 


Has this pump always had this issue or is this something relatively new?

Quote:
Originally Posted by JohnJ
John PA,

The measured runout is actually 2.26 mils overall at 100 rpm and the 1X and 2X from the spectrum is 0.901 and 1.18 mils respectively.  The time waveforms are pretty smooth, even at 100 rpm.  I would expect them to look like there were a lot of little scratches if there were corrosion.  I will try to get the probe area examined, but I doubt they will want to do that until they have a planned outage.

My understanding about using stopped gap voltages for the SCL, is that the shaft should be at bottom dead center (I know that doesn't have to be true in all cases).  I referenced the gaps at 100 rpm and plotted the SCL and compensated orbits with bearing clearance in the attached file.  I also have a plot of the coastdown waterfall with the 2X referenced with the dark line and the 1X referenced with the green line.  Note that the spectral data cannot be compensated in my software.  The SCL and orbits are still outside the bearing clearance. but changed positions to the left.

I might not worry too much about the shaft running outside the bearing clearance by 1 mil except it running out the top of the bearing.  The impellers are between bearings and the DE bearing runs within the clearance but just a little higher than expected.  I have a hard time believing that coupling misalignment (Thomas series 71) could cause the NDE of the shaft to run at the top with between bearings impellers (2 stage).  I am suspecting bearing misalignment.  What are your feelings?

Thanks,
John J


I don't particularly like "stopped" gap voltages.  What little error is introduced due to the shaft not being at the bottom dead center is minimal, especially when you have a shaft exhibiting 2+ mils of some form of runout.  If you want to use stopped voltages, but this procedure neglects thermal effects, then roll the shaft by hand taking note of the max and min DC voltages, then use the average.
JohnJ

Member
Registered:
Posts: 66
Reply with quote  #11 
John PA,

Thanks so much for your input.  Your expertise are much appreciated by me and I'm sure the rest of the board.

John J


hriveraq

Member
Registered:
Posts: 13
Reply with quote  #12 
hola hij , te adjunto informacion donde podras ver Transactions of the ASME

 
Attached Files
pdf Mechanical and Electrical Runv out.pdf (611.27 KB, 12 views)
pdf D04_Shaft_deflection_runout_vibration_and_axial_motion.pdf (1.36 MB, 11 views)

Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.