Sign up Calendar Latest Topics Donate
 
 
 


Reply
  Author   Comment  
MAA

Member
Registered:
Posts: 4
Reply with quote  #1 
Hi guys,

Is it possible for 1X vibration of a particular machine to get amplified by background 1X vibration from adjacent machine which is running at same speed? One of our compressor's main lube oil pump is currently experiencing high 1X vibration. The LO pump is driven by a main motor which also drives the compressor on the other end. There are three identical compressor packages installed in our plant with the one with problematic lube oil pump located in between the others. We have carried out bump test on the pump and its associated structure to check for any possibility of resonance. A thin peak was observed in the high resolution spectrum which coincide with running frequency of the machine. However, we believe this peak is coming from background vibration as natural frequency peak would typically show-up as a mountain with gradual slope while background machinery vibration would show up on high-resolution spectrum as a super-thin vertical line. We proceeded to measure the pump vibration as it is (without running) and we can see the thin 1X peak is still there, prompting conclusion that it is coming from the running adjacent machines. We could not shutdown the other two machines to see its effect due to restriction on plant demand.

Seeking for expert view on this. Many thanks. 
Beatnik

Sr. Member / Supporter
Registered:
Posts: 183
Reply with quote  #2 
Yes it's a common problem when doing bump test. Usually we'll take a measurement without bumping and measurements with bumping and compare.

You could try negative averaging also but I never did it in this situation. Look in the documentation of your collector/software to see how to do that.
Shurafa

Sr. Member / Supporter
Registered:
Posts: 310
Reply with quote  #3 
Based on the information provided in the post, what you saw as a peak in the impact test could be a background vibration (not the real natural frequency).

To evaluate the data quality of an impact test, I usually review the waveform. A quick screening is this: if the impact could not generate vibration more 3 times the background vibration, the data is rejected. If the structure is relatively small or flexible, my limit is 5 times.

It would be easier for the forum members to help is you can post the plots.

Regards-Ali M. Al-Shurafa
MAA

Member
Registered:
Posts: 4
Reply with quote  #4 
Many thanks for the replies. Noted that the thin spectrum line represents background vibration and not the actual resonant peak. I've attached the bump test result for your reference and further evaluation. Since the thin spectrum line is not valid, none of the natural frequency peaks in the spectrum seems to coincide with the machine running speed (~49.8-50 Hz) to cause state of resonance. However, we are still observing high 1X vibration on the pump. FYI, we have already carried out trim balancing on the driver motor previously and manage to reduce the vibration at the motor drive end by 25% but with no significant changes on the pump vibration. Regarding my previous question, by any chance that the 1X vibration of this machine has been amplified by interference with background vibration from adjacent units?

 
Attached Files
pdf Bump Test Result.pdf (2.54 MB, 17 views)

MarkL

Avatar / Picture

Sr. Member
Registered:
Posts: 1,120
Reply with quote  #5 
Not being able to look at the graph and not sure if you mentioned run speed are you on 50hz supply? Sounds like a electrical thing to me. I’ve seen electrical peaks are most times unnaturally ‘perfect’.......if you get me? Thin and straight, and if I’m getnting you right in this context always present? A electrical shut down test of machine would be my first call to rule out.
Also have you considered beat frequency if a few similar run speed machines are in close proximity this is a possibility?
I once Had a 1 x fan speed peak from a 2.2kw motor on a huge extruder show all the way across the machine train through a 2 ton gearbox and visible in the gearbox readings.
MAA

Member
Registered:
Posts: 4
Reply with quote  #6 
MarkL,

The pump's driver is a 2 pole induction motor with supply of 50 Hz. The motor also runs a compressor at its other end. I have checked the vibration with high resolution setting on the analyzer and the spectrum peak is observed at 49.8 Hz. In my view, given that the motor has some slip, the peak seems to be associated with the mechanical speed instead of electrical line frequency (correct me if I'm wrong). Since there are another two identical machines running adjacent, beat frequency due to background vibration is also a possibility. The background vibration from adjacent units is indeed felt at the pump (~0.2 mm/s RMS) when we took the measurement while this particular machine is stopped. But the pump vibration seems to be stable while running without any fluctuation, which is what one normally expect when dealing with beat vibration. Also, how severe could it be if beat vibration is actually the case? Adjacent unit's lube oil pump are having highest vibration at 4 mm/s RMS while this particular unit is vibrating at 12 mm/s RMS. Could background vibration from the adjacent units really amplify the vibration to such extent? Appreciate for your advise.
MarkL

Avatar / Picture

Sr. Member
Registered:
Posts: 1,120
Reply with quote  #7 
The speed and line frequenciy are very close, have you taken a low resoloution 400 line 200 hz fmax reading to see if ther is any separation of peaks?

I can't say with any conviction I've experience of a frequency being amplified by another except in a resonance situation.
Curran919

Avatar / Picture

Sr. Member / Supporter
Registered:
Posts: 599
Reply with quote  #8 
Quote:
Originally Posted by MAA
Also, how severe could it be if beat vibration is actually the case? Adjacent unit's lube oil pump are having highest vibration at 4 mm/s RMS while this particular unit is vibrating at 12 mm/s RMS. Could background vibration from the adjacent units really amplify the vibration to such extent? Appreciate for your advise.


About beat frequency:

Run pump A, shutdown pump B: Vibe on A 4.0mm/s, on B 0.2 mm/s. The cross talk from A to B is then 0.2 mm/s. Run pump B and shutdown pump A. Pump B vibrates at 12.0 mm/s. Assuming only a mechanical coupling between the pumps (discounting electrical and hydraulic interference), running both pumps should result in pump B beating between 11.8 mm/s and 12.2 mm/s.

Its not too complicated. Beating itself isn't bad. Its only if the crosstalk from other pumps is high (like 30%, not 5%), then it can lead to large swings in vibration levels which means intermittent high vibration.

If you see a much higher beat amplitude than the cross-talk from A to B, there is something more complicated happening.
Danny Harvey

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,402
Reply with quote  #9 
If you are equipped, you could do a cross channel phase and coherence test to determine if the sharp peak in you spectrum comes from one of the other pumps.
John from PA

Sr. Member
Registered:
Posts: 963
Reply with quote  #10 
Quote:
Originally Posted by MarkL
The speed and line frequenciy are very close, have you taken a low resoloution 400 line 200 hz fmax reading to see if ther is any separation of peaks?


OP already stated:

Quote:
I have checked the vibration with high resolution setting on the analyzer and the spectrum peak is observed at 49.8 Hz.


That works out to about 10 cpm so likely is the motor slip frequency.  A good tach or strobe would answer that question.
Ralph Stewart

Avatar / Picture

Sr. Member
Registered:
Posts: 235
Reply with quote  #11 
Is the main concern with this LO Pump the "high" vibration it has when it is not running or is the concern with it being worse when it is running?
__________________
Ralph Stewart
AlertAnalytical.com
MAA

Member
Registered:
Posts: 4
Reply with quote  #12 
Ralph,

The LO pump vibrates heavily when running (1X peak). Vibration readings while running and layout of the machine are as follows:

Pump DE Vertical: 6.5412 mm/s RMS
Pump DE Horizontal: 6.7813 mm/s RMS
Pump DE Axial: 2.6463 mm/s RMS
Pump NDE Vertical: 12.5862 mm/s RMS
Pump NDE Horizontal: 13.2908 mm/s RMS
Pump NDE Axial: 2.2350 mm/s RMS

air comp.jpg 
FYI, the LO pump is of screw type. It is mounted to the motor frame and it is kind of cantilevered. As mentioned earlier, there are 3 identical machines located side by side. Background vibration from adjacent machines is felt when the unit is not running albeit very small. Appreciate your view on this.


Ralph Stewart

Avatar / Picture

Sr. Member
Registered:
Posts: 235
Reply with quote  #13 
MAA,

Does the waveform show any modulation, if so, what is the distance between them?

When the motor is running does the 1x or any frequency in the spectrum increase and decrease in amplitude in a repeatable pattern?

When he motor is running, are there any frequencies with sidebands, if so, what are their spacing off the carrier?

I have attached a file with some direction questions. [smile]

Attached Images
jpeg Questions on motor and LO pump.jpg (50.45 KB, 6 views)


__________________
Ralph Stewart
AlertAnalytical.com

Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.