Centrifugal Compressor Vibration

[ad90]

Hi,

I have also posted this in the Acoustic & Vibration Engineering Forum.

We are trying to commission a new motor-gear-centrif, gas compression train. We have an issue with the vibration levels at the Compressor Drive End. As the machine runs up to speed (16,000 rpm driven) the shaft centreline plot (both DE & NDE) is a near vertical line rising ~130 microns. Once on its centre line the NDE vibration is a constant ~12 microns (primarily 1x & round orbit) as is the DE. However the DE amplitude continues to rise over time flattening out at ~39 microns (primarily 1x & round orbit)after 90 minutes.

The machinery OEM is blaming nozzle loads due to the piping growing, but my experience tells me that nozzle loads, if large enough to cause problems, will apply a pre-load to the bearing resulting in an elliptical orbit.

FYI. The machine alignment is within OEM spec and the piping certainly looks well enough designed that loads would not be transferred back on to the nozzles.

Any ideas? I am looking at Morton Effect.

Thanks

AD90

 

[nickelkid]

Has the OEM made a specific remark about your piping configuration?

 

[ad90]

Only the Service Hand who was onsite, made the comment that because we re-aligned the machine (moved it 0.1mm, as that is what they blamed originally for the vibration. NB: When out-of-alignment (cold) by 0.1mm the pipework was 'floating' at the nozzle) with the piping done up, he thought the pipework could be at fault. His FS office have backed him on this although we have had no feedback from their engineering dept.

I am trying to stay one step ahead.

Thanks

AD90

 

[TPL]

You need to take care when using shaft centreline information - monitoring a start from cold and then through to loading for any length of time will include effects due to thermal growth. For the purpose of assessing any alignment influences, it is better to observe shaft centre line changes during a shutdown following a period of loaded operation AND using the gap volts measured as soon as the unit comes to rest as the reference parameters.

What is the bearing clearance - have you added the bearing clearance to the shaft centreline to see where the shaft settles when on load?

If this is a new machine, what does the supply contract say about vibration - API 617 suggests a limit of 22 microns for a new machine running at 16000rpm (runout needs to be added to this, say 28 microns total)

 

[ad90]

Bearing clearance is (design) 0.15mm - 0.179mm and as the machine is new, I expect it to be at the lower end. The shaft centreline runs down in a similar manner to the run up.

AD90

 

[Rotator]

Are the thermal growth assumptions contained in the vendor alignment specs valid? They will be based upon an assumed cold condition and temp rise to operating temp which may not be valid. Can set up laser to measure actual relative thermal growth of bearing housings which will validate the cold offset assumptions and might give you a clue as to what's happening.

 

[ad90]

Thanks Rotator, Permalign checks were on the list. What we have done is trim balance the coupling (very small amounts 0.6g) and we now have a machine that is showing maximum 25 microns and shows little sensitivity to time/temp.

AD90