vibration case histories presentation, 091203.pdf · 9/19/2006 3 case history#1, excessive...

9/19/20061

Vibration Case HistoriesVibration Case HistoriesBarry T. CeaseMeadWestvaco

9/19/20062

IntroductionIntroduction

Case History#1 – ExcessiveVibration Of Motor

Case History#2 – Strange Motor &Gearbox Vibration

Case History#3 – Repeat FanBearing Failures

Case History#4 – Pump BearingCage Failure

9/19/20063

Case History#1,Case History#1,Excessive Vibration Of MotorExcessive Vibration Of Motor

EQUIPMENT: Nash 9000 seriesvacuum pump driven by a singlereduction, parallel gearbox and a700 HP synchronous motor.

The vacuum pump is one of manyused to pull water out of the paper(dry it) as it moves thru theprocess.

9/19/20064

CH#1 – Equipment LayoutCH#1 – Equipment Layout

9/19/20065

CH#1 - ProblemCH#1 - Problem

Excessive & noisy vibration atmotor.

This problem could be heard & feltclearly at the motor.

It sounded like something wasrubbing and/or loose.

9/19/20066

CH#1 – Test Data & ObservationsCH#1 – Test Data & Observations(Motor)(Motor)

Motor speed was exactly 400 rpm. Pump speed was 267 rpm. Pump vibration data showed no

abnormal patterns or levels. Motor spectra showed many

harmonics of running speed and of133 cpm (1/3 x rpm).

Motor waveform showed impactingat 400 & 133 cpm.

9/19/20067

CH#1 - Motor SpectraCH#1 - Motor Spectra

9/19/20068

CH#1 – Motor Spectra ZoomCH#1 – Motor Spectra Zoom

9/19/20069

CH#1 – Motor WaveformCH#1 – Motor Waveform

9/19/200610

CH#1 – Test Data & ObservationsCH#1 – Test Data & Observations(Gearbox)(Gearbox) Gearbox input speed was 400 rpm. Gearbox output speed was 267 rpm

(1.5:1 ratio and 56T/84T). Gearbox spectra showed many

harmonics of both input & output speedand 133 cpm (1/2 x output).

Gearbox spectra didn’t show highgearmesh frequency at 22,400 cpm.

Gearbox waveform showed heavyimpacting at 400 cpm.

Gearbox Peakvue data showed manyharmonics of 400 cpm.

9/19/200611

CH#1 – Gearbox SpectraCH#1 – Gearbox Spectra

9/19/200612

CH#1 – Gearbox Spectra ZoomCH#1 – Gearbox Spectra Zoom

9/19/200613

CH#1 – Gearbox WaveformCH#1 – Gearbox Waveform

9/19/200614

CH#1 –CH#1 – PeakvuePeakvue SpectraSpectra

9/19/200615

CH#1 –CH#1 – PeakvuePeakvue WaveformWaveform

9/19/200616

CH#1 – Maintenance HistoryCH#1 – Maintenance History

The motor was changed out in7/02 and the gearbox was changedin 8/02.

After motor change in 7/02, motorvibration levels initially dropped,but eventually resumed previouslevels.

The gearbox installed in 8/02 hadbeen used before, but was the onlyspare available at the time.

9/19/200617

CH#1 – Conclusions &CH#1 – Conclusions &RecommendationsRecommendations

It was recommended that the gearboxbe changed out when possible due to aprobable gear fault.

The gearbox was changed in 5/03. An inspection of the gearbox showed

many gear teeth broken off on thepinion gear with significant wear onboth gears.

Follow-up data on both machines afterthe gearbox change showed much lowervibration levels – the problem vanished.

9/19/200618

CH#1 – Gearbox InspectionCH#1 – Gearbox Inspection

9/19/200619

CH#1 – Motor Data, B & ACH#1 – Motor Data, B & A

9/19/200620

CH#1 – Gearbox Spectra, B & ACH#1 – Gearbox Spectra, B & A

9/19/200621

CH#1 – Gearbox Wave, B & ACH#1 – Gearbox Wave, B & A

9/19/200622

Case History#2, Strange Motor &Case History#2, Strange Motor &Gearbox VibrationGearbox Vibration

EQUIPMENT: Agitator driven thrusingle reduction, parallel gearboxby an induction motor.

This agitator helps maintain theconsistency of the stock in ourhydrapulper tank.

9/19/200623

CH#2 - ProblemCH#2 - Problem

Strange pulsing noise comingfrom motor & gearbox.

9/19/200624

CH#2 – Test Data & ObservationsCH#2 – Test Data & Observations(Motor)(Motor)

Motor speed was exactly 1192 rpm. Agitator speed was 236 rpm (5:1 ratio). Motor spectra showed many harmonics

of running speed & closer inspectionshowed 48 cpm sidebands around eachharmonic.

Motor waveform showed pulsations ormodulation at a period of approx. 1.25sec or 48 cpm.

9/19/200625

CH#2 – Motor DataCH#2 – Motor Data

9/19/200626

CH#2 – Motor ZoomCH#2 – Motor Zoom

9/19/200627

CH#2 – Test Data & ObservationsCH#2 – Test Data & Observations(Gearbox), Part 1(Gearbox), Part 1

Gearbox speed was 1192 rpm input &236 rpm output (single reduction @ 5:1ratio).

Gearbox spectra showed high gearmeshfrequency (26,200 cpm) with sidebandsat 945 cpm.

Gearmesh sidebands usually relate toone of the gear speeds, but 945 cpmdidn’t correlate to either speed (1192 or236 rpm).

9/19/200628

CH#2 – Test Data & ObservationsCH#2 – Test Data & Observations(Gearbox), Part 2(Gearbox), Part 2

Gearbox waveform showedimpacting or modulation at 945cpm.

Agitator data showed nothingabnormal.

9/19/200629

CH#2 – Gearbox DataCH#2 – Gearbox Data

9/19/200630

CH#2 – Gearbox ZoomCH#2 – Gearbox Zoom

9/19/200631

CH#2 – Conclusions &CH#2 – Conclusions &Recommendations (Motor), P1Recommendations (Motor), P1 It was concluded that the strange

vibration data on the motor was theresult of broken rotor bars.

The 48 cpm sidebands around themotor harmonics related directly to itspole pass frequency.

The classic spectral pattern of brokenrotor bars is running speed harmonicswith sidebands at pole pass frequency.

The classic waveform pattern of brokenrotor bars is pulsations at pole passfrequency.

9/19/200632

CH#2 – Conclusions &CH#2 – Conclusions &Recommendations (Motor), P2Recommendations (Motor), P2

Pole Pass Frequency =(Theoretical RPM – True RPM) * #Poles. PPF = (1200 – 1192) * 6 = 48 cpm. The recommendation was made to

changeout the motor at the next outage. Later inspection by a motor repair shop

showed many broken rotor bars.

9/19/200633

CH#2 – Conclusions &CH#2 – Conclusions &Recommendations (Gearbox)Recommendations (Gearbox) After consultation with the gearbox

vendor, it was concluded the thestrange data from the gearbox waslikely due to the 4-yoke design ofthe bull gear.

The 4-yokes in the gear hub resultin minor deviations from the gearpitch circle causing modulationeach time these teeth move in andout of the mesh.

4 * 236 rpm = 945 cpm.

9/19/200634

CH#2 - 4-Yoke GearCH#2 - 4-Yoke Gear

9/19/200635

CH#2 – Conclusions &CH#2 – Conclusions &Recommendations (Gearbox),Recommendations (Gearbox),Part 2Part 2

The vendor indicated it shouldn’tbe a problem, but recommendedan annual gear inspection &continued vibration monitoringlooking for any change incondition.

9/19/200636

Case History #3, Repeat FanCase History #3, Repeat FanBearing FailuresBearing Failures

EQUIPMENT: Overhung,centrifugal fan belt-driven by a 60HP induction motor.

This is a critical fan necessary tothe process of winding the paperinto customer-specified sizes.

9/19/200637

CH#3 – Equipment LayoutCH#3 – Equipment Layout

9/19/200638

CH#3 - ProblemCH#3 - Problem

Repeat fan bearing failures. In one instance, vibration detected

bearing faults on this fan less than amonth after changeout.

Predictive maintenance was able todetect these failures early enough toschedule repairs during outages, butafter three fan bearing changeouts in 12months, we knew something had to bedone differently.

9/19/200639

CH#3 – Test Data & Observations,CH#3 – Test Data & Observations,Part 1Part 1

Motor speed was 1786 rpm Fan speed was 1985 rpm Motor spectra showed running speed &

harmonics, fan speed & harmonics, beltfrequencies & little else.

Fan spectra was similar to motor data,but also showed fan bearing defectfrequencies (BPFO & harmonics).

9/19/200640

CH#3 – Fan SpectraCH#3 – Fan Spectra

9/19/200641

CH#3 – Test Data & Observations,CH#3 – Test Data & Observations,Part 2Part 2

Fan trend data showed initial dropwhen bearings were changed, butsoon jumped up to previous highlevels days or weeks afterchangeout.

One of the mechanics involved inthe bearing change told us, “it tookus over an hour to get the bearingsaligned to where the shaft wouldeven turn”.

9/19/200642

CH#3 – Fan Trend DataCH#3 – Fan Trend Data

9/19/200643

CH#3 – Test Data & Observations,CH#3 – Test Data & Observations,Part 3Part 3

The fan bearings were standardpillow block style housings withtapered roller bearings inside.

We assumed these were self-aligning bearings as most pillowblocks are, but this assumptionturned out to be false.

9/19/200644

CH#3 – Conclusion &CH#3 – Conclusion &Recommendations, Part 1Recommendations, Part 1 The relatively tight alignment tolerances

of the existing tapered roller bearingscombined with the poor condition of thefan base made for short bearing life.

We asked our bearing supplier for areplacement bearing type that wouldcarry the same load, but be moreforgiving for misalignment.

We also had our machine shop fabricatea new fan base that was machined flat& line bored to perfectly fit the newpillow block bolt pattern.

9/19/200645

CH#3 – Conclusions &CH#3 – Conclusions &Recommendations, Part 2Recommendations, Part 2

After installation of the new typebearings & new fan base we have nothad another bearing failure.

Our bearing life has went from anaverage of 4-months to 26 months andcounting.

Further inspection of maintenancehistory showed a fan speed increasewhich corresponded quite well to ourincreased rate of failure.

9/19/200646

CH#3 – Conclusions &CH#3 – Conclusions &Recommendations, Part 3Recommendations, Part 3

This case history is a good example ofboth predictive & proactivemaintenance.

Predictive maintenance allowed us toavoid catastrophic fan bearing failuresand perform repairs during scheduledoutages.

Proactive maintenance extended the life& reliability of the fan bearings.

9/19/200647

Case History#4, Pump OutboardCase History#4, Pump OutboardBearing Cage FailureBearing Cage Failure

Equipment: Double-suction, centrifugalpump driven by a 1250 HP synchronousmotor turning at 514 rpm.

This is the most critical pump in theprocess directly providing the product(stock) used to make paper on the papermachine.

9/19/200648

CH#4 - ProblemCH#4 - Problem

Increasing HFD vibration levels at pumpoutboard. Current levels had exceededmaximum of long term trend.

9/19/200649

CH#4 – Test Data & Observations,CH#4 – Test Data & Observations,Part 1Part 1

Since the bearing was oil lubricated, ananalysis of the outboard bearing oil wasrequested.

Oil analysis results showed very highcopper levels indicating cage wear.

Vibration data showed very high HFDlevels on outboard bearing.

9/19/200650

CH#4 – HFD TrendCH#4 – HFD Trend

9/19/200651

CH#4 – Test Data & Observations,CH#4 – Test Data & Observations,Part 2Part 2

Peakvue spectra showed harmonics offundamental train frequency (cage).

An 10/01 paper by J. Robinson & J. Berryrecommends a Peakvue fault level of 4.0g’s pk-to-pk for a 500 rpm machine.

Peakvue waveform on outboard bearingshowed levels at 11.3 g’s pk-to-pk!!!

Maintenance history showed a continuingproblem of a leaking pump outboard seal.

9/19/200652

CH#4 –CH#4 – PeakvuePeakvue Spectra, P1Spectra, P1

9/19/200653

CH#4 –CH#4 – PeakvuePeakvue Spectra, P2Spectra, P2

9/19/200654

CH#4 –CH#4 – PeakvuePeakvue WaveformWaveform

9/19/200655

CH#4 – Conclusion &CH#4 – Conclusion &Recommendations, Part 1Recommendations, Part 1

The recommendation was made tochangeout the pump outboard bearing.

Later inspection of the bearing showed thecage worn badly particularly in the areaseparating the rolling elements from oneanother.

During the bearing change, shims werefound under the bearing making up the fitwith the housing.

9/19/200656

CH#4 – Conclusion &CH#4 – Conclusion &Recommendations, Part 2Recommendations, Part 2 Failure analysis of the bearing found lube

contamination and skidding as the primarycauses of failure. Corrective actions were asfollows:

1) Replace the outboard bearing housing, 2) Repair or replace the pump outboard packing

gland, 3) Replace the pump rotor which was causing

thrust loading, and 4) Upgrade the lube from an ISO 68 to an ISO

150 weight oil (temporarily use grease untilpacking gland repaired).