Dear All,

I am a first timer here and will appreciate suggestion from this forum.This Fan & my whole plant is down since last night.

This is a overhung fan with antifriction bearlings(DE 6224 & NDE NU 1030),oil cooled,3000 rpm(induction motor with sleeve bearings).The coupling is distance type with shims at both end(dont know exactly what it is called).
Vib in both fan & motor was maxm 1.8 mm/s on 03/07 .But at 02:30 vib suddenly reached 15 mm/s in Fan NDE Hor(only probe in Fan) in our on line accleleration probe.Data recorded with FFT analyzer(maxm 35 mm/s at Fan DE Axial)and Fan was shut down.There was no change in flw/pr etc and jump was sudden.I am attatching readings & some spectrum.

1)The Fan spectrum shows harmonics of rpm(with sidebands of app 400 cpm).Inspected bearing housing(integral type housing,cvers entire bearing unit) not much wear.Clearance in NU1030 s ok 0.15 mm.
2)Some skirts at 3 x rpm.Bump test in fan housing steel frame natural fr is above 50000cpm.Fan bearing housing natural fr show a peak at 10500 cpm(close to 3 x rpm).Hard to understand resonance will act suddenly.
3)Coupling appears ok.Could not take cross phase & costdown,will take when restarts

Pl suuggest urgently

Regards
csp

rab040707.doc (307 Kb, 101 downloads) ms word

Looking at the velocity data suggests that your fan has developed a huge and sudden increase in 1X vibration.

I would suggest the following:
a) Look for bits missing from the fan (possibly due to cover/blade resonance leading to fatigue failure) - use a borescope if you can

b) Look for broken holding down bolts (Check by measuring on the bolt and on the structure immediately next to the bolt to ensure no relative movement between the 2 locations) - or you could try using a strobe to identify areas where relative motion is taking place

c) Try a lift check on any exposed parts of the fan shaft

That level of increase is quite extreme - the pressure to get the plant back up will be intense, but do make sure that any further investigation is carried out safely

This message has been edited. Last edited by: Buzz LightYear,

Dear Buzz,
Thanks for reply.
-----------------------------------------------
'Looking at the velocity data suggests that your fan has developed a huge and sudden increase in 1X vibration.'
----------------------------------------------

The peaks are at 1x, 2x,3x,45 etc(major peak at 3x & not 1x)as shown in attatched file.Vib is relatively higher at DE end.So I do not suspect unbalance in this overhung fan.Any way we have inspected & cleaned the impeller.

From spectrum the problem appeared to be looseness.But all fasteners are ok & bearing housing contact area with bearing outer ring is ok(no visible wear,outer ring is not rotating in housing).There are unusal sidebands of 400 cpm in rpm harmonics.

Resonance may be an issue.But resonance acting up suddenly is very remote

-----------------------------------------------

c) Try a lift check on any exposed parts of the fan shaft
------------------------------------------------
Can you pl detail the lift check & how it is done & why.

We will start the fan & see how it behaves ,will take cross phase readings.

Still waiting for further suggestions;Fast

Regards

csp

CSP
Could there have been something go through and stick to the fan while you where running and then fall off when you stopped this unit. Also the lift check is to see how much play or looseness you have in the bearings using somthing to lift the shaft and record the movement with a dial indicator. This will give you an idea of how loose this is.

Sorry - took a quick look at the spectra and read cpm instead of orders

Hmm 3X on a fan?

What did the bearing cage look like?

Possible rub (exciting the resonance?) But what would cause the rub - something stuck between the rotating and stationary parts?

or coupling problem? Can you put a DTI on any exposed portion of the coupling (fan side) and carry out axial and radial runout checks?

The lift check involves putting a DTI on the shaft and seeing how high you can lift the shaft - for roling element bearings in good condition you should not have any movement.

If you do start the fan, it would be useful to see if the vibration readings remain constant with time or are changing - can you take a spectrum reading say every 15 seconds at the worst point and see how the vibration trends for the duration of the run?


Do you have any details on the fan construction - no of vanes? Coupling assembly details?

What are the chances that this is not a mechanical problem but a process related issue?

Could this be a result of a severe flow disturbance? Could you vent the fan discharge to atmosphere to check for this?

This message has been edited. Last edited by: Buzz LightYear,

quote:

Originally posted by csp:
Fan bearing housing natural fr show a peak at 10500 cpm(close to 3 x rpm).Hard to understand resonance will act suddenly.

It is obviously a resonance, likely in the pedestal or housing, and coast down test in this case is not going to reveal it. As far as the cause is concerned, it could be, for instance, a crack or mounting problem (loose fastening bolts, grout, etc).

The source of 400cpm sidebands is puzzling. Does not look it is motor's Pole Pass freq. It may provide a clue if source of PPF is found but I do not think this is the major problem. My suggestion is to bump as many components of the fan as possible (including vanes) in all directions.

Dave

Looking at the acceleration time waveform. If I count the positive peaks I see 16 peaks in 0.1 seconds. Approx 0.006 seconds interval corresponds to 3x.

But then look at the negative peaks in the range around 0.3 seconds 0.4 seconds. Every 3rd one is much larger. (corresponds to 1x). In fact these peaks spaced at 1x are the higest in the whole TWF. This might possibly be an indication of impacting/looseness creating 1x harmonics and exciting a 3x resonance as others said. But I think the very large magnitude of the peak (14 g’s true pk/0) suggests to me the impacting/looseness is the problem to attack, not the resonance excited by it. Even if you managed to take away that apparent resonant 3x peak somehow, you still have 8 mm/sec at 2x and 6 mm/sec at 4x which is still a problem. Once again this suggests to me that resonance is not the main problem.

What exactly causes 400 cpm (~0.12 orders) sidebands around harmonics of running speed?... beats me. The best I can guess is severe looseness, most likely in the bearings. Check all the stuff listed above seems like a pretty good list to me.

I tend to agree with most of the other guys here. for this to happen all of a sudden, seems like something has come loose, maybe a crack in the pedestal around the bearing mounts. I have seen these "strange" sidebands around harmonics of 1x and they are related mostly to a "weak" mounting surface. So like most have said, look for something loose, internal and external.

Wow, do we have lives or what? Answering vibration questions on the Fourth of July!!!!
I have seen sidebanding like this and it is usally cage related and involves axial bounce in ball bearings. The 3x may be resonance, but where did the sudden excitation come from? The coupling, as you describe it sounds like a shim pack type. I would check the coupling for broken or loose bolts. That would then explain the issues with the 400 cpm sidebanding, since with sleeve bearings in the motor, the fan bearings are carrying all of the thrust load.

Thanks for all the input.Sorry i couldnot give the feedback as i was busy

What follows now is really strange and this KKK make Fan is turning out to be very unusal case study.

My understanding from the spectrums was basically looseness or rub(in bearing housing) and possible crack in fan unit/baseframe(as suggested by most).As I mentioned earlier no fastener looseness,bearing rotating in housing were found.I searched for cracks in base frame/foundation & housing .Its not possible to check every nook & cornor but i could no find anything.Coupling was ok.Bearing cage,clearance etc are ok.Only real work was impeller cleaning as 4-5 mm/s 1 x rpm was there in fan NDE side(fines gets stuck in impeller)

Started the fan after all this checks.Surprisingly the vib was down to app 2 mm/s in all points.Everyone felt happy(except me) and 'deposits in impeller' was declared the culprit, as there was good rain yesterday.I asked the maint guy that vib is down but I am not so sure about the equipment so do closely monitor the online Fan NDE H (only on line probe)readings in control room.


Operation started in evening and again vib increase was reported in midnight and this used to come down to 2 mm/s on its own even though fan is operating in its performace curve.Vib readings (no FFT spectrum could be measured)up to 20 mm/s in Fan DE & NDE Axial was measured ,also 12-14 in radial direction.This vib shooting up and coming down is very puzzling.

I will try to get some spectrum at high vib when i 'catch' them.The present readings are app 2 mm/s with minor 1x,harmonics and minor bearing fault fr(This has been the trend since last overhaul8 months back).No subharmonics to suggest flow instability,Little Vane Pass Frq(6 ).

Sorry this is a bit lengthy,but i want to give you a clear picture.Now please hurry up with all your experience on this issue.

1)I dont know exactly how safe to run this equi with erratic 20 mm/s
2)What is happening here.Is an undetected crack responsible for this .
Flow is in the performace curve.I cannot adjust the inlet damper to observe more(process parameter wont permit),Couplings are ok.The fan in ground floor sucks air from atmosphere above 50 m pipe and suction filter is ok.Is some stuck material in suction pipe is getting into the fan.

This vib shooting up and coming down is very puzzling

This problem has RUB written all over it.

Since the machine can run intermittently at low vibration levels, you have effectively eliminated problems such as unbalance, bearing and coupling damage.

The difficulty here is that rubs are a symptom of another problem i.e. a rub will occur due to loss of clearance, so you will need to identify where the rub is occurring and eliminate its cause.

A starting point would be to study a sectional drawing of the fan and identify any likely rub locations, such as seals or baffles and then check these items carefully. It doesn’t take much physical contact to excite a resonance and generate high vibration – you might want to consider clearance checks using plastigage.

Are there any locations where a build up of hard deposits could cause a rub?

Another thing to consider is the timing of this problem – does the high vibration only occur at night? If so, this suggests the possibility of a temperature related cause. Can you correlate ambient air temperatures and process temperatures with the appearance of the high vibration and possibly tie this problem into an alignment issue? This is always going to be a tricky one to resolve since cold alignment might be OK. It could be some internal component that undergoes a dimensional change with temperature, causing a rub.

I guess the possibility of a crack (possibly under a sleeve or collar or somewhere on the structure) still remains, but I would have thought that once the any crack had opened up (due to thermal growth?) and caused higher vibration then it would remain high and not return to ‘normal’ levels. I recall some Bently training course material showing that a growing crack can reduce shaft stiffness which in turn causes the critical speed to pass through the machine, eventually giving a reduction in vibration as complete failure approaches.

Apart from inspection you are going to need some more detailed data to get much further.
I don’t know if you have the equipment available, but fitting a once-per-turn tacho pulse and using something like an Adre or Zonic book allowing a comparison of startup and shutdown data (bode plots in particular) will go a long way in helping to resolve this problem.

Is the machine safe to run? – I would not be comfortable with it – high and steady I can live with for a short time but high and unsteady worries me, since you never know where it will go. Whilst 20/30 mm/sec is OK for very short term running, these levels will not do much for the bearing life. The other difficulty is that if you get comfortable running the machine with these high levels, you might well miss any other developing problems?

If this is a rub it will either clear itself or go 'bang'.


I suggest you try to get more data (at least spectra and phase angle) just before, during and after the vibration increase.

(1) A machine running close to this equipment can also excite the natural frequency.

(2) Also perform the Motor Current Signature Analysis if possible.

Finding the excitation source, of course, is right strategy since it could be the only problem even if the machine has "built-in" resonance at 3x. Rub definitely is one of the posibilities and Bode plot will help to detect it although it requires installation of a tach.

Another (simpler) way to detect a rub will be using PeakVue or possibly the transient program. Compare TWF during low vs. high vibration.

But possibility of structural faults (pedestal or shaft crack) and resonance which has moved to 3x range as result, still has to be ruled out.

Dave

Is it possible that one of the bearings is rotating inside the housing? Maybe when it is spinning it is acting like a rub with the sidebands being the frequency that the bearing is turning in the housing. Then the housing heats up and the clearance goes away and the bearing stops turning, resulting in vibration getting back to more normal levels. Then things cool off and the bearing starts to spin again and everything cycles?
Just a SWAG!

I'm with vibeguy 2004. Turning bearing 400cpm, causes change in stiffness.
Strange fan set up with a cyclindrical roller - how do you get the angular bearing to shaft tolerance correct? Do you bother to check 'em?

I have some more info & data regarding this Fan and vib goes up suddenly (upto 27 mm/s in both DE & NDE axial) and comes down to 2 mm/s .
As most of us fill Rub / possible unidentified crack seems to be the problem and to be sorted out.
Flow disturbance can be another reason(as already suggested in forum) I was not entirly right when i said earlier that Fan is operating in its performance curve( 60000 to 95000 m3 /Hr)I spend some time in DCS and observed
1) vibration shoots up even when there is not much flow change i.e. minor up down in performace curve(flow is regulted by process dept as per furnace condition)
2)Vib shoots up from 2 mm/s and sometimes drops from high value also when there is flow change as exhibited in DCS.When fan was running with high vibration we intentionaly asked process guy to drop flow to 50000 m3/hr and then go to normal flow.Vibration droped to normal level.But I must add that we tried the same thing today twice but vib didnt go down.I can understand that fan can have high vib when fan operates at left side of performance curve but not this.

I have taken spectrums in high vib today which i am attatching.Vib reduces to 2 mm/s

1)Signifacant vib at Fan DE & NDE axial927 mm/s) as ealier with predominant 3 x rpm(close to bearing casing natural frequecy of 10500).RPM Harmonics in radial direction with Bearing fault frequecy
2)Impacting in acc TW at 1 x rpm more in Fan NDE
3) 392 cpm sidebands are lower at harmonics .One thing observed was this 392 cpm minor peak in SKF envoleped acceleration spectrum.I have seen this type of sidebands when bearing outer ring rotates in housing

It appears to me that flow change/instability creates some kind of stiffness change in the fan and thus resonance .I understand its very difficult for you guys to give suggestion regarding this fan without 'feeling' the fan and reviewing the spectrum in software,but all inputs are welcome.I will try to get bode plot in my FFT analyzer when fan stops.In the meanwhile we are trying for suggestions from KKK Germany which will take time.

Pl give some thought on this flow disturbance aspect and any experience like this.Also rubbing/ crack issues are also to be invesigated but dodnt exactly what to do.THe mechanical seal appeared ok.
Jim
------------------------------------------------
Strange fan set up with a cyclindrical roller - how do you get the angular bearing to shaft tolerance correct? Do you bother to check 'em?
----------------------------------------------
This NU 1030 cylindrcal bearing is floating type as per desgin.we checked the clearnce.ok.Do you mean tolerance of shaft

Regards
CSP

rab070707.doc (455 Kb, 36 downloads)

Hi VibeGuy2004
Bearing turning in housing is a possibility,But during inspection it appeared normal.The prblem is i have a lot of possible causes. and I have to get the bode plot aswell.

Hi Buzz,

This doenot happen in night only,Seals ,baffale are ok.When in high vibration,abnormal noise is there but no metallic sound.Waiting to hear your opinion as well

CSP
Expectin

Started the fan after all this checks.Surprisingly the vib was down to app 2 mm/s in all points

This vib shooting up and coming down is very puzzling.

When fan was running with high vibration we intentionaly asked process guy to drop flow to 50000 m3/hr and then go to normal flow.Vibration droped to normal level

Changing the flow will presumably change the thrust on the fan. How is the thrust set up on this unit? Can you measure the axial float/allowable thrust clearance and compare with OEM recommendations?

I am wondering if you have excessive axial float, through either installation or wear, that allows the rotor to make contact with a stationary component under certain flow/temperature conditions.

I suggest you try to measure the axial float when cold, then inspect the drawings and identify a location where you could add a shim to reposition the rotor so that changes in axial load cannot facilitate a rub.

You will have to figure out the best shim thickness, it might be no more than 5 or 10 thou - this can be achieved by machining an existing collar shim.

I don't believe that you are looking for large signs of damage or excessive clearance, just something towards the end of its limits. It doesn't take much of a rub to generate high vibration when a resonance is involved

This message has been edited. Last edited by: Buzz LightYear,

Hi Buzz,

------------------------------------------------
I am wondering if you have excessive axial float, through either installation or wear, that allows the rotor to make contact with a stationary component under certain flow/temperature
conditions.
----------------------------------------------
I think you are right,and you have explained that also.This flow varition gives axial thrust which leads to this rubbing exhibitted in TW.I shuuld not have missed this logic.
Last time we inspected the eqipment in coupled condition and axial float was there but we presumes that to be from motor.Now I will check this float after decoupling.There is no thrust setup in this fan unit with antifriction bearing(like other sleeve bearing units)

Down here people feel this erratic vibration is due to some random material from suction line getting into the fan.I wiil check the axial float with a Dial indicator in cold condition.I presume this fan with antifriction bearing should have negligiable play/float.

Thanks a lot for this inputs sitting miles away.By the way where is Niue


csp

Is the bearing mounted directly to the shaft or is there an adapter sleeve between the bearing and the shaft?
Could you attach a picture of the fan?