Dear all. 

I need some help to analysis abnormal vibration trend. 

This machine is centrifugal compressor and has 5 impellers. And it is driven by steam turbine. This machine is compressing propylene from -35'C to 70'C.

At normal state, compressor drive side radial vibration was increased gradually from 20 micron to 30 micron. After load change this vibration have cyclic pattern from 20 micron to 30 micron. 

We did some action to do reduce vibration level and to find root cause. (Base bolt tightening, proximitor replacement) But We can not find root cause for this abnormal phenomenon. please review attachment and give some advice.



Photos (10)
Original Post

I know current vibration level is not high and is under this machine alarm set 66um. But vibration pattern is not normal. And I worry that amplitude could be higher than nor gradually. 

Attached 2 polar plots with amplitude and phase. DE side plot shows that unbalance is strong and no phase change. NDE side plot shows that phase is moving and returning but vibration level is very low. It could be negligible. But I think, NDE polar plot pattern is similar with light rubbing or Morton effect. And it effect to DE side badly making strong unbalance. 

Especially NDE side temperature is very low about -35'C with propylene. DE side is about 70'C. And NDE side primary seal vent is almost zero though sometimes it shows some level of leakage. So I think NDE gas seal have possibility of icing and plugged to vent hole. And it makes unnecessary rub or effect bearing and oil. 


Actually, In my experience, i have not met the morton effect yet. So, my thinking is just my sense. The speed could be not exactly in the case but i think that your case is not morton effect. Your phase change is not cyclic and i sure that your compressor is between bearing type which motorn effect is rarely occurred. 


My thinking could be total wrong!


I would suggest checking more operating parameters in order to have a better chance to understand the possible problem.


Sometimes,  compressor upstream or downstream equipment are unstable and affect the compressor. Sometimes, compressor auxiliaries such as lube oil temperature control system is not tuned perfectly and can result in compressor instability.


Spectrum comparison could help a little if you can post data for high and low times.


I tend to agree with the comment given about the low value of the amplitude. Fortunately or unfortunately some problems can be easier analyzed if the magnitude is high.

Regards- Ali M. Al-Shurafa 

Add Reply