High vibration trip on BN probes

Hello experts! 

On one of our critical blower, wWe are facing frequent trips due to high vibration on Bentley Nevada probes on gearbox high-speed shaft (12100 RPM).  Vibration readings (over the bearing) are high as well on gearbox but we are unable to find the root cause, we have conducted MCSA, perform alignment/ soft foot check but with no success. Oil analysis results came out clean.

Particularly two things we found in spectrum are puzzling us- see spectrum attachment

  1. Peak at 2LF ( 100 Hz) appears to as a twin peak and not precisely at 100 Hz
  2. There is a weird peak at 1620 Hz, which is not matching to any of the fault frequency and has been fluctuating/trending up. The same peak is present on all points with differing magnitudes.

If you could please go through the attachments and share your thoughts as to what could possibly be the cause of the problem and what other test (s), we should perform to find the root cause and any thoughts on counter measures.

Please let me know if any other info/ specs are required for the diagnosis of the issue.

Attachments

Original Post

Dear Yousuf

Roughly calculating the mentioned peak at 1620Hz it corresponds to 8.03xRPM(12100) hence it could be the vane pass for your blower kindly confirm the number of vanes on the blower. 

Secondly considering the casing vibrations maximum vibration is around 4.08mm/s rms which does not seems to be much high. However Proximity probes are giving high vibration and to analyze that plots related to proximity probes are required. Kindly provide plots to further analyze that.

Regards

Rohit

Thanks for the responses . Following are some clarifications in response to your comments

  • I initially suspect vane pass as well but there are 13 vanes in the blower. So this is (1620 Hz) not vane pass 
  • It does not occur at start ups. Machine just trips during normal operation and every time its a high vibration trip. 
  • Waveform ( acceleration ) is not showing any signs which may be attributed to high bearing stress. 

For a short term , I have implemented following rules to control frequent trips

  • we have two similar machines and our process requirement is to run one machine at full load and other on varying load ( 30 %- 70%). We were using this machine as main machine ie., on 100 % load . So now we switch this machine on varying load but even on that , it tripped once within 12 hours , however restarted again and doing in between 60-70 micron. 
  • Trip level setting on BN probes was set to 80 micron ( zone B ISO standard 7919). I suggested to increase it to 90 micron ( slightly in zone C) to avoid frequent trips and startups. 

 

Your thoughts and suggestions required 

Yousuf Masood posted:

Thanks for the responses . Following are some clarifications in response to your comments

  • I initially suspect vane pass as well but there are 13 vanes in the blower. So this is (1620 Hz) not vane pass 
  • It does not occur at start ups. Machine just trips during normal operation and every time its a high vibration trip. 
  • Waveform ( acceleration ) is not showing any signs which may be attributed to high bearing stress. 

For a short term , I have implemented following rules to control frequent trips

  • we have two similar machines and our process requirement is to run one machine at full load and other on varying load ( 30 %- 70%). We were using this machine as main machine ie., on 100 % load . So now we switch this machine on varying load but even on that , it tripped once within 12 hours , however restarted again and doing in between 60-70 micron. 
  • Trip level setting on BN probes was set to 80 microns ( zone B ISO standard 7919). I suggested to increase it to 90 micron ( slightly in zone C) to avoid frequent trips and startups. 

 

Your thoughts and suggestions required 

Is your proxy data real? Try change the sensor with similar machine 

John from PA posted:

In light of your title "High vibration trip on BN probes" is there any reason data from the BN probes hasn't been added to the thread?  All I see is casing measurements.  Assuming you have orthogonal probes, are both probes high at the time of the trip or just one?

 values of probes are there as well and YES there are orthogonal probes . Both probe values are high but usually it always trips on high value ( more than 80 micron) on one probe 

Yousuf Masood posted:

values of probes are there as well v and YES there are orthogonal probes . Both probe values are high but usually it always trips on high value ( more than 80 micron) on one probe 

Every plot I see has "v rms[mm/s]" in the top left corner.  And others as well have asked about proximity probe data.

John from PA posted:
Yousuf Masood posted:

values of probes are there as well v and YES there are orthogonal probes . Both probe values are high but usually it always trips on high value ( more than 80 micron) on one probe 

Every plot I see has "v rms[mm/s]" in the top left corner.  And others as well have asked about proximity probe data.

Please see the data attached. 

Attachments

Yousuf Masood posted:
John from PA posted:
Yousuf Masood posted:

values of probes are there as well v and YES there are orthogonal probes . Both probe values are high but usually it always trips on high value ( more than 80 micron) on one probe 

Every plot I see has "v rms[mm/s]" in the top left corner.  And others as well have asked about proximity probe data.

Please see the data attached. 

As an fit attachment, 200hz have high amplitude and all harmonics have been low amplitude. What kind of blower? Does it lobe? Maybe 200hz related to lobe mesh frequency?

aziz58 posted
an 80 micron) on one probe  

As an fit attachment, 200hz have high amplitude and all harmonics have been low amplitude. What kind of blower? Does it lobe? Maybe 200hz related to lobe mesh frequency?

From data already provided, do the math.

Motor speed= 2970 RPM and tooth combination is 159/39.

2970 * 159/39 =12108 pinion speed = 202 Hz which is likely to be the frequency seen at approximately 200 Hz.  The proximity probe data clearly shows that as the dominant frequency but I have to ask is there any proximity probe data when the machine is acting up?

To the OP, you might want to double check your tooth combination (159/39) since it represents a non-hunting tooth situation.

 

Yousuf Masood posted:

Thanks for the responses 

@aziz58  Its not a lobe type blower, its a centrifugal one 

@John from PA : Unfortunately we do not have the proximity probe data when the machine is acting up. 

You reckon its an issue on pinion only or may be a hunting tooth problem? 

It's unfortunate that you don't have proximity probe data when the machine is acting up.  Seeing what the spectral content is would likely be very beneficial.

As to hunting tooth problem; I only brought that up because you don't have a hunting tooth, which is desired.  In your case, with a GCD of 3, a hunting tooth problem would most likely create 1/3 of mesh.

By the way, with respect to your machine sketch, the gearing was shown as single-helical.  Is that correct?  You do not want to mount a blower wheel directly to a double helical pinion although I have seen it done (with issues).

@John from PA Thanks a lot for your response.  you are right that gearing is single helical.

I have added the real time graphical data of the probes. It may give you an idea how the values are fluctuating despite the fact that the faulty machine is on low load and drawing low current ( 138 amps where as other is drawing 160 amps) 

Kindly share what are your thoughts about the corrective action(s)? 

Attachments

Yousuf Masood posted:

@John from PA Thanks a lot for your response.  you are right that gearing is single helical.

I have added the real time graphical data of the probes. It may give you an idea how the values are fluctuating despite the fact that the faulty machine is on low load and drawing low current ( 138 amps where as other is drawing 160 amps) 

Kindly share what are your thoughts about the corrective action(s)? 

According to the real time graph I think you have pulse vibration by 200hz frequency .as John mentioned it can be hunting frequency 

aziz58 posted:
Yousuf Masood posted:

@John from PA Thanks a lot for your response.  you are right that gearing is single helical.

I have added the real time graphical data of the probes. It may give you an idea how the values are fluctuating despite the fact that the faulty machine is on low load and drawing low current ( 138 amps where as other is drawing 160 amps) 

Kindly share what are your thoughts about the corrective action(s)? 

According to the real time graph I think you have pulse vibration by 200hz frequency .as John mentioned it can be hunting frequency 

I didn't say it was hunting frequency, I did say lack of a hunting tooth combination would yield 1/3 of gear mesh which would be around 2623 Hz.

FYI, anytime someone gives me a tooth combination that is a non-hunting tooth combination, I suggest they verify it.   

Add Reply

×
×
×
×