Steam turbine (driving a forced draft fan through a gear box) with normal running speed of 4500 rpm. Both i/b and o/b bearings experienced increasing vibration amplitudes while the turbine is running with constant operating conditions (from 1.5 to 4 mils). Mainly1X. Phase is not constant. Sometimes with time, amplitudes reduce (3.5 to 1.5) while phase is consistent, sometimes the phase is changing (140 to 10 deg) while the amplitude is steady. Orbit is semi circular. Vibration is collected using eddy current probes and confirmed actual using casing readings.

Thinking of: rotor dynamics changes, loose/creaked components, thermal sensitivity.
Any ideas. Running out of time!
Thanks.

Ali M. Al-Shurafa

What's the coupling type and arrangement? Has this type phenomenon abruptly entered into the picture or has it come with time? Do you have known resonant frequency data?

And, what is your time frame?

Ali,
More information is required.

Pressurized oil system?
If so, do casing measurements indicate any 1/3X?
Was this vibration gradual or sudden?
If sudden, cracks or rotor looseness, maybe?
If gradual, check bearing clearances or build-up.
Also, does FD Fan have damper controls set correct?
As Sam indicated, more info on coupling and resonant data would be useful.
Do you have any data you can post?

Thanks for your response.

no 1/3 x.
bearing clearance is 6 thou in both sides.
simple spacer coupling.
amplitude increases from 1.5 to 4 withing 1-3 min with exponenetial shape.
Critical speed based on the book is 6045 rpm.
Oil is pressurized by auxil then by main pump mounted on the gear.
I'll try to post some data asap.

Thanks again. Any more ideas?

As stated, you don't give much info but, the symptoms you describe fit those of:
a)rub
b) wet steam - condensate slugging the rotor

Attached is a doc with more details including a few plots.

Today machine had an increase in the vibration amplitude from 2 to 4 with minor phase change (about 35 deg) while increasing the speed from 4200-4500 rpm. Speed was increased very slowly.

Regarding rub: no indecations of 1/3, 1/2, 1 1/3, 1 1/2 ...x and no backward precs, waveform is not trancated from the top.

Thanks.

R81-KT-108Sat.doc (356 Kb, 52 downloads)

quote:

simple spacer coupling.

Exactly how is this coupling designed? Gear type with a spacer between each hub or what?

Did you say whether or not this was a sudden chnage or a gradual change or has this turbine been overhauled recently and this is what it is doing?

Have you any harmonics of 1X in the spectrum? I notice you are running CSI RBMware in the background. Have you any data from using it or is all this from the "ADRE"? Looks like in your waveform plots, there might be some harmonics. Just a guess though.

This looks alot like a rub, the type that casues the shaft to bow - not uncommon.

If you had coasted all the way down, I suspect you would have a different slowroll caused the the shaft bow. This would have rolled out over time. ... but you kept it at speed.

There is little doubt that you are looking at a rub - the difficulity now is determining the cause

rebuild with tight clearances?
Misalignment - excessive pipe strain/foundation problems


It would be interesting to see an orbit at the end of the run up and another at the same speed when the amplitude is at a maximum

BTW - you are interpreting the 'rules of rubs' far too literally. For example, you will usually only see exact 1/2X when you are running at above twicw the critical speed. Truncated waveforms are not always present.

The best indicators of rubs are changes that occur under steady state operating conditions and completely different behaviour during startups and shutdown (best seen on bode plots)

As already mentioned, inspection of runout data during startup and shutdown is also useful

The response is non linear, maybe can be related to one of the next two :
1.- Crack shaft : be carefull. Is there any body here experienced with this kind of failure?
2.- Crack foundation : several month ago somebody shown in this forum a similar problem and after was detected crack foundation If I find, I will recall in a next post.

The machine was tested solo (no fan , no gear). Similar trends were found.

Regarding foundation, yes it's old and went into fixing during the past weeks. Regarding misalignment, the coupling is dropped now and still no change in the behaivior. Max reading is in the i/b.

Regarding rub, orbit as mentioned and showed in the above attachment, there are small external loops. Spectrum shows 1x (main), small 2x and 3x and a bit bigger 4x. But when the vibration increases, they become very minor and 1x contain most of the orbit and spectrum.

Coast down data, did not show a clear increase in the amplitude. Seams normal but i'll check.

Important.. amplitude and phase change while speed fluctuates. Phase is not increasing; it is decreasing!!

Thanks again for all.

Kovac,
are u referring to? :

http://maintenanceforums.com/eve/forums/a/tpc/f/375...561026641#9561026641

We could run the turbine solo and collected the data. Now the data shows stronger indications of rub. Circular polar plot while machine is at constant (almost constant) speed. An interesting thing is that the phase is not increasing (not going down in the Bode Plot), it is decreasing (going up in the Bode Plot) and crossed the 12 O'clock three times.

How much could the piping stresses effect the turbine casing distortion resulting in a rub?
Provided that maintenance checked the clearances (bearing and seals) and they were as per manufacturer recommendations, what are root causes?

Regards-AMS

Check the overall casing expansion and make sure the rotor should not go faster the casing and that they free to move axially. This can be measured by a dial or check the indicator normally located on the front standard. Sometimes grease becomes solid after long period and bind the casing from moving. Check also all turbine drains to make sure no condensate is being trap. Look for possible changes in the process parameters and see any change from normal as this unit perhaps is exhausting to main condenser or exhaust is being used for process.

quote:

Kovac,
are u referring to? :

Yes, it is I told.
-----------------
Making relation between rub and thermal grown, What is the inlet overheated steam temperature and what would have to be ?