My Fellow Colleague,

Been battling with a water injection pump for weeks now, the pump is a centrifugal multistage pump with 6 stages driven by a siemens electric motor through a lufkin gearbox . The pump had failed early this year due to process issue, pump was run for a long time without load, signal was sent to the breaker to trip pump, but it did not trip, luckily the pump was hand tripped by an operator, when we started, the pump refuse to turn, so we got another bundle which has been laying horizontally on the workshop for close to 9 years. After rebundling, we started the pump and had high radial vibration on the X-Y axis of the pump drive end of about 145 microns and 120 microns pk-pk respectively, spectrum analysis at first indicated some level of misalignment and rub, we opened the bearing and found some level of rubs on the DE bearings, changed it, corrected the alignment, changed the coupling and then started, this time, the vibration increased further to 160 microns and 128 microns respectively on same points. But the interesting thing is that at first 4X was dominating, now we have corrected the alignment 1X is now dominating, when we checked the phase between DE and NDE, X-X 108 degrees, Y-Y 138 degrees, X-Y DE bearing; 127 degrees X-Y NDE bearing; 97 degrees. Attached is a copy of the data

Note: Run out has been carried out on the shaft and was found to be 1 thou, which is still okay

Attachments

Original Post

Just one information, the speed of the motor is 3000 rpm, pump is 5216 rpm. Attached is the configuration of the probe, it is advisable to open the plots on paint so you can see the headings, the probe number the last two numbers after 68409-(44)

Oli, by knock test did u mean a bump test or a modal test to check for resonance frequencies?,..The balance details were actually extracted from the polar plot, spectrum and orbit....We have seen some forcing frequencies at phase coinciding with unbalance but some are still out of play...so I want to be double sure it's imbalance before I give them a go to pull out the rotor

Yes, just a bump. I have "normally" trim balance on site, weir feed pumps on voith for example or dresser charging pumps on mech seal mount screws....
If the plots are the latest after adjustment it still looks a bit like touching, looseness problems anyway so yes there may be more than one thing to look into.

 

I quite agree with you on bumping the equipment...coz the problem is more or looking like looseness exciting resonance in the pump..but it's funny enough we don't have the capacity to do a bump test i.e I mean no vibration analyzer...the only thing we can rely on is the data from system 1 monitoring hmi...on unbalance I have been suspicious of it, but the phase shift doesn't really set the rythm but thanks anyway

Where's @John P   

If the system 1 can provide a FFT it's not that fancy if it is undamped, just hit it by a 2"x2" or 2"x4" and see if you get a peak, when  not running that is and pending what sensors there are...... If you do get a peak, repeat H and V if you can and possible to get at the rotor that is with a reasonable oil supply or lift oil, jack oil or what available.

"Trying and getting a peak is better than never striking the rotor." I rarely break them that way :-).

Oli, if I understand you well, you want me to impact the pump at the casing or bearing housing and watch the spectrum live to see if I would get a peak which obviously would be a resonating frequency? Actually the probe there are proximity probes and u know they aren't that good with impact that's y I needed an analyzer so I can use a seismic probe (accelerometer)

 

Your freq. range of suspected resonances are within a few 100Hz well within eddy probe range. No I don't see any reason for eddy probes not reacting on impacting if you get relative movement btw. bearing and shaft. If you try to excite the shaft that is one of the normal suspects there are a reasonable chance that you get a peak in your FFT if you try it with the tools you have IMHO. You can't try it with tools you don't have so the choice is yours. Slightly out of the box maybe but what options do y have? Good luck.

No guarantee but worth a try it may bring some information anyway, do you have some kind of circ oil? It may be not possible if shaft is not floating on some oil, if it is stuck in the bearings it may not give any movement across the eddy probes but it is worth a try anyway.

Add Reply

Post

HACKATHON in association with IMC-2019

×
×
×
×