Hi all
Any phase analysis gurus on board I need a bit of help with this.
I was asked to confirm that high vibration of approx 8-9 mm/sec was present in the axial direction on a motor driving a compressor
I took a simple measurement and was able to confirm 8+ mm/sec
While I was at it I managed to take some phase readings.
The customer then decided to uncouple the compressor just to see if it made a difference, Axial vibration was still there at about 8 mm/sec

Regardless of the vibration levels a decision had been made to fit new shims to the motor check for soft foot and re-align the drive motor to the compressor

I went back to site and took another set of measurements and there was a decrease in axial vibration to approx 5.5 mm/sec rms, I could not take any phase measurements this time

My knowledge is limited regarding phase analysis but the data suggests (there seems to be a 180 degree phase shift everywhere) that it may be both couple unbalance as well as a resonance problem.

I wanted to take some coast down/star up measurements’ and an impact test but the coupling was guarded, with the data shown can anyone shed any light or offer any suggestions.
Thanks
Mcdm

Phase_Compressor.xls (24 Kb, 59 downloads)

measurements on the motor

Phase_Motor_only.xls (24 Kb, 54 downloads)

It looks as if there could be some internal misalignment in the motor possibly due to a cocked bearing or the frame being distorted from the hold down bolts. Check your phases on each side of the shaft in the axial direction to see if you get a 180 degree phase shift to determine if the bearing is cocked and also loosen one motor foot at a time while monitoring vibration to see if the internal alignment is being changed due distortion of the frame from the hold down bolts.

MCDM,

Vertical readings on the motor are about 6 and 15 times higher then that in Hor. Most of the time the picture is in reverse and with less ratio. Resonance in Vertical?

Mcdm,
How were the phase measurements calculated, Cross channel(two accels) or keyphaser(accel and a reference point on shaft)? This is important because it will establish a reference point depending on the method.
For example, If you used the cross channel method did you leave the reference accel on the motor in one spot and move the roaming accel including the compressor. This method establishes phase relative to 2 accels using absolute vibration, which means vibration is not relative to the shaft nor does it have a reference point on the shaft. The keyphaser method uses the reference mark on the shaft.

Without knowing the method used, I can't say for sure but you do have axial vibration that is much higher compared to the rest of the points for both machines. Radial vibrations are low and typicaly the axial measurements are low compared to radial. You either have a coupling issue or a structural resonance in the axial plane and of course this all depends on how the vibration was colected. We also don't now what type of machine, speed, HP, bearings, etc. Once we know more about the machine a severity can be assessed.

Thanks for the help and replies
This is a 2 pole motor fitted with rolling element bearings can't day exactly how big but its about 250 Kw
last summer a new bigger drive motor was fitted new support struts were made and fitted a new coupling was used and a new compressor. Female from male
high axial vibration was reported then but this was the first time that this comp could be topped.

I used 1 accelerometer and a laser tacho to take the measurements

Some more information the integrity of the drive motor and coupling assembly is now in doubt as I have noticed that the coupling key is full on the motor side (it sticks out approx 3 inches from the coupling)
Apparently the motor, coupling, and comp were in situ and were fitted as is just an alignment check was made, I doubt that the coupling was balanced with the motor.

Once again thanks for the help and advice I will post an image of the support struts as soon as possible; I am not in work at the moment its sunday and i'm trying to type this and dodge my wife so she dosen't catch me working on my day off.

Thanks Mcdm,
Now that we know the phase readings come from a keyphaser it looks like the coupling is forcing the machines away from each other 180 deg. You mention in your drawing that the axial phase readings are not corrected for direction. After correction it shows machines are 180 deg out from each other. What type of coupling is it? Radial vibration amplitudes are not bad but the axials are a little high for axial reading.

Please attach spectrum and waveform readings

Hi Eric
Attached is a picture of the coupling assembly

coupling.doc (96 Kb, 44 downloads)

I have attached the spectra page 5 show high res data taken on the motor 2x line frequency? how close to 100 Hz does it have to be?
This is not running on a VFD
I have seen this 2xLF occasionally when the air gap is uneven, misalignment between end shields and or bearing cartridges

As David L has said "internal"

Spectra.doc (118 Kb, 30 downloads)

Mcdm,

Now as it appears to me there is a whole new game. A profound 2xLF is present. This might be due to frame twisting causing internal bearing misalignment, and as result high axial vibration. First thing to do is to get rid of 2xLF and then go from there.

High V/H ratio still is weird and a bump test is warranted IMO.

David

This message has been edited. Last edited by: David_G,

Thank you for all the replies
David G, I completely agree I have recommended a bump test and inspect the drive motor.

This is a very interesting case; the information is invaluable and IMO can be used as an example of a real life vibration problem.

I just hope I can get some feedback from the client so I can post results and prove it.

Can you view the coupling during operation with a tunable strobe? If you can, look for a wave in the shimpack. Are you certain that the axial spacing between the coupling hubs is correct?

Thanks for the advice Danny
If I get invited back to site I will try the strobe
I have been assured that the axial spacing and alignment is correct, I am yet to find out what the tolerances or spec figures were