Field Application Note

7/28/2019 Field Application Note

http://slidepdf.com/reader/full/field-application-note 1/124

Field Application Note

Eccentricity TSI

Shaft Eccentricity plays a veryimportant role as part of a TurbineSupervisory Instrumentation (TSI)System on large steam turbines andshould be included in retro-fit planswhen at all possible.

Operators use eccentricitymeasurements to determine when acombination of slow roll and heatinghave reduced the rotor eccentricity to

the pointwheretheturbine

can safely be brought up to speedwithout damage from excessivevibration or rotor to stator contact.

Eccentricity is the measurement of Rotor Bow at rotor slow roll whichmay be caused by any or acombination of

1. Fixed mechanical bow2. Temporary thermal bow3. Gravity bow

In extreme cases of thermal/gravity bow, caused by a sudden trip of theunit and failure of the turning gear toengage, the rotor may be positionedand stopped 180o out of phase (bowup) to allow gravity to work entirelyon the bow and substantially shortenthe time required to reduce the bow.

Eccentricity is measured while the

Mechanical Runout Eddy Current transducers are alsosensitive to the shaft smoothness for Eccentricity. A smooth (64 micro-inch) area approximately 3 times thediameter of the probe tip must be provided for a viewing area.

Electrical Runout Since Eddy Probes are sensitive tothe permeability and resistivity of the target material and the field of the transducer extends into thesurface area of the shaft byapproximately 15 mils (0.015"), caremust be taken to avoid nonhomogeneous viewing area materialssuch as Chrome.

Another form of electrical runoutcan be caused by small magneticfields such as those left by Magna-fluxing without proper degaussing.

Perpendicular to shaft centerline Care must be exercised in allinstallations to insure that the EddyProbe is mounted perpendicular tothe shaft center-line. Deviation bymore than 1-2 degrees will effect theoutput sensitivity of the Probe.

Transducer (Probe) side

clearances The RF Field emitted from the probetip of the transducer isapproximately a 45 conical shape.Clearance must be provided on allsides of the Probe tip to prevent



turbine is on slow roll (1 to 240 RPM below the speed at which the rotor becomes dynamic and rises in the bearing on the oil wedge) andrequires special circuitry to detect the

peak- to-peak motion of the shaft.This is accomplished using circuitrywith long update times selectable between 20 seconds (> 3 RPM) and 2minutes (<3 RPM).

As the eccentricity measurement isnot required after a turbine is broughtto speed and under load provisionsare made to lock the measurement tozero. This can be accomplished

without external contacts through theuse of a speed measurement channelwith underspeed or overspeedalarms.

As it is impractical to mount EddyProbe Transducers (Non-ContactingPickups) midspan on the rotor wherethe eccentricity measurement would be the highest the transducer(s) aremounted outside the pressure case as

far from the bearing (Node Point) as practical.

The bearing should be avoided as amounting location because duringslow roll operation the rotor isturning in the bottom of the journal bearing and is not dynamic while theeccentricity measurements are beingmade. This effect forces the bearings

to

becomenodal points.

Assuming

uniform stiffness and weight, therotor midspan eccentricity may be

interference of the RF Field. Caremust also betaken toavoidcollars or

shoulderson the shaftthat may

thermally "grow" out from under theProbe tip as the shaft expands.

Eddy Probe tip to tip clearances Although Eddy Probe tip to tipclearances are not normally an issueon most machines, it should be notedthat the probes radiate an RF Field

larger than the probe tip itself. As anexample, SKF-CM CMSS65 and 68Eddy Probes should never beinstalled with less than one (1) inchof Probe tip to tip clearance. Larger probes require more clearance.Failure to follow this rule will allowthe driver to create a "beat"frequency which will be the sum anddifference of the two driver RFfrequencies.

System Cable Length and

Junction Boxes Eddy Probe Systems are a "tuned"length, and several system lengthsare available. System length ismeasured from the probe tip to theOscillator/Demodulator, and ismeasured electrically which can beslightly different than the physicallength. For example, the Model 403

is available in 9, 20, and 30 footsystem lengths. Care must be takento insure that the proper systemlength is ordered to reach therequired Junction Box.

Grounding and Noise Electrical noise is a very serious



expressed as the ratio of thetransducer span from the bearingover the transducer measuredeccentricity to 1/2 the bearing spanover the midspan eccentricity or

calculated using the followingformula, (Tecc x ½Bspan)/Tspan =MSecc.Where Tecc = Transducer measuredeccentricityBspan= Bearing SpanTspan= Transducer span from bearingMSecc= Midspan eccentricity

OEM's (Original EquipmentManufacturers) should be consulted

for actual calculations.

Turbineowners whoareretrofittingexistingeccentricitysystemssupplied bythe OEM or

others willmount the eccentricity transducer atthe same location as the originalinstallation. In many cases onlyminor modifications to the existing bracket are required. Using the samelocation has several advantages andsimplifies installation.

1. OEM's original installation asa rule included an eccentricity

collar or other good target for an Eddy Probe System.

2. Eddy Probe eccentricitymeasurements will agreeclosely with the originalOEM supplied system as themeasurements will be takenat the same location.

consideration when installing anyvibration transducer, and specialcare needs to be taken to preventunnecessary amounts of noise. Asmost plant electrical noise is at 60

HZ, and many machine runningspeeds are also 60 HZ, it is difficultto separate noise from actualvibration signal. Therefore, noisemust be kept to an absoluteminimum.

Instrument Wire A 3-wire twisted shielded instrumentwire (ie; Belden #8770) is used toconnect each

Oscillator/Demodulator to the SignalConditioner Card in the Monitor.Where possible, a single run of wirefrom the Oscillator/Demodulator (Junction Box) to the Monitor location should be used. Splicesshould be avoided.

The gauge of the selected wiredepends on the length of theinstrument wire run, and should be

as follows to prevent loss of highfrequency signals:

Up to 200 feet 22 AWG



The following wiring connectionconvention should be followed:

Red -24 VDC Power Black Common

White Signal

Common Point Grounding To prevent Ground Loops fromcreating system noise, system



3. Operators will need lesstraining on how to interpretthe new systemsmeasurements as they will be basically the same.

4. Eccentricity historical datawill be valid.

5. Existing brackets may bemodified.

6. Case or standard penetrationfor cable may be reused withminor modification.

Eccentricity isnormally

measured P/P(Peak to

Peak) to agree with previouslyestablished conventions. The actualexcursion from shaft centerlinecaused by bow would be one half that measurement or the 0/P (Zero toPeak) measurement. The TurbineSupervisory Instrumentation may becalibrated in either fashion to suite

the users requirements.

Theory of Operation

Eddy CurrentTransducerswork on the proximitytheory of operation. A

system consists of a matched

component system: a Probe, anExtension Cable and an Oscillator /Demodulator (driver). A highfrequency RF signal @2 mHZ isgenerated by theOscillator/Demodulator, sent throughthe extension cable and radiated fromthe Probe tip. Eddy currents are

common, ground and instrumentwire shield must be connected toground at one location only. In mostcases, the recommendation is toconnect commons, grounds and

shields at the Monitor location. Thismeans that all commons, groundsand shields must be floated (notconnected) at the machine.

Occasionally due to installationmethods instrument wire shields areconnected to ground at the machinecase and not at the monitor. In thiscase, all of the instrument wireshields must be floated (not

connected) at the monitor.

Conduit Dedicated conduit should be provided in all installations for bothmechanical and noise protection.Flexible metal conduit should beused from the Eddy Probe to theOscillator /Demodulator junction box, and rigid bonded metal conduitfrom the junction box to the monitor.

Calibration All Eddy Probe systems (Probe,Cable and Oscillator Demodulator)should be calibrated prior to beinginstalled. This can be done by usinga SKF-CM P/N CMSS601 StaticCalibrator, -24 VDC Power Supplyand a Digital Volt Meter. The EddyProbe is installed in the tester withthe target set against the Eddy Probe

tip. The spindle micrometer withtarget attached is then rotated awayfrom the Eddy Probe in 0.005" or 5mil increments. The voltage readingis recorded and graphed at eachincrement. The SKF-CM CMSS65and 68 systems will produce avoltage change of 1.0 VDC ±0.05



generated in the surface of the shaft.The driver demodulates the signaland provides a modulated DCVoltage where the DC portion isdirectly proportional to gap

(distance) and the AC portion isdirectly proportional to vibration. Inthis way, an Eddy CurrentTransducer can be used for bothRadial Vibration and distancemeasurements such as ThrustPosition and Shaft Position.

Special Considerations

Mounting Orientation

All vibration transducers measuremotion in their mounted plane. Inother words, motion either directlyaway from or towards the mountedEddy Probe will be measured aseccentricity.

For eccentricity measurements it isrecommended that the transducer bemounted vertically. As mosteccentricity sensors are internally

mounted and are not visible from theoutside of the machine whatever theangle of orientation is finally chosenit is very important that the mountinglocation be documented for futurereference.

Linear Range Several versions of Eddy ProbeTransducers are available with avariety of Linear Ranges and body

styles. In most cases, a sensor with alinear range of 90 mils (0.090") ismore than adequate for Eccentricitymeasurements.

Model Range Output Size

CMSS65 90 mils 200mV/mil

1/4"x28 UNF1" to 5"

VDC for each 5 mils of gap changewhile the target is within the NCPU's linear range.

Gap

When installed, Eddy Probes must be gapped properly. In mostEccentricity applications, gappingthe transducer to the center of thelinear range is adequate. For theModel 403 transducer gap should beset for -12.0 VDC using a DigitalVolt Meter (DVM), this correspondsto an approximate mechanical gap of 0.060" or 60 mils. The voltagemethod of gapping the Eddy Probe is

recommended over mechanicalgapping because it is more accurateand easier to accomplish. In allcases, final Eddy Probe gap voltageshould be documented and kept in asafe place.

Eccentricity Installation Checklist

1. Machine Slow Roll Speed2. Transducer Orientation

Documented3. Target Material, 4140 Other 4. Smooth Target Area5. Size of Target Area6. Junction Box Location(s)7. Metal Conduit (Junction Box

to Monitor)8. Flexible Conduit (Junction

Box to Probe)9. Correct Instrument Wire10. Shielding Convention,

Monitor or Machine11. Calibration12. Gap Set



Length


3/8"x24 UNF1" to 9"Length

Target Material/Target Area

Target Material EddyCurrenttransducersarecalibrated atthe factory

for 4140 Steel unless specifiedotherwise. As Eddy Probes are

sensitive to the permeability andresistivity of the shaft material, anyshaft material other than 4000 seriessteels must be specified at the time of order. In cases of exotic shaftmaterial a sample may need to besupplied to the factory.

RDI Homepage

Copyright © 1989-2000 by Sales Technology, Inc. ALL RIGHTS RESERVED

Table of Contents

Application Notes

General CAB Custom Cabinets

FS Field Service

Maint Maintenance Methods

MCM Machine Classification for Monitoring

USERS Monitoring Systems Users

MON Monitoring Classifications

REB Rolling Element Bearings

JB Journal Bearings

Basic Vibration F=MA System Response

BAL Balancing

CVR Comparing Vibration Readings

JBF Bearing Failure Modes

http://www.reliabilitydirect.com/default.htm

http://www.reliabilitydirect.com/appnotes/appnotes.html


http://www.reliabilitydirect.com/appnotes/cabapp.htm


http://www.reliabilitydirect.com/appnotes/fieldservice.html


http://www.reliabilitydirect.com/appnotes/maintain.html


http://www.reliabilitydirect.com/appnotes/classmac.html


http://www.reliabilitydirect.com/appnotes/systemusers.html


http://www.reliabilitydirect.com/appnotes/classmon.html


http://www.reliabilitydirect.com/appnotes/reb.html


http://www.reliabilitydirect.com/appnotes/jb.html


http://www.reliabilitydirect.com/appnotes/fma.html


http://www.reliabilitydirect.com/appnotes/balance.html


http://www.reliabilitydirect.com/appnotes/cvr.html


http://www.reliabilitydirect.com/appnotes/jbfail.html















http://www.reliabilitydirect.com/default.htm



Indoor Indoor___ Outdoor Nema4 ___ Nema4 ___

Environment:

Both indoor and especially outdoor cabinet environments need to be looked at for special environmental concerns and requirements.Temperature is a common issue for outdoor cabinets. Both the annual temperature extremes of the proposed cabinet location and the heatgenerating capacity of the equipment to be installed in the cabinet must be carefully looked at. Heating and/or Air Conditioning will beinstalled as required.

Min.Temp ______ C/F. Max. Temp_______C/F

Type and Manufacture of Cabinet:

Cabinets come in two (2) basic types, Solid and Modular. Solid Cabinets have welded seams where modular cabinets are bolted together.Solid cabinets will be generally be used outdoors and modular cabinets indoors. In some cases a customer may have a preference betweenmanufactures.

SolidSolid___ Modular___ Manufacture___________

Modular___ Manufacture___________

Access:

The location that the cabinet is to be installed in needs to be looked at carefully as to access for maintenance and repair of the installedequipment. In general most cabinets have access front and rear. In some cases where a cabinet is to be mounted against a wall only side or front access is available. We always recommend Front and Back Access as any other type is difficult and expensive unless absolutelyrequired. If side access is necessary, specify which side as if you are looking from the front.

Front & BackFront & Back ___ Front & L/R ___ Front Only ___Front & Back___ Front & L/R ___ Front Only ___

Dimensions or Size:

Size is in the end determined by the installed equipment and required options. In most cases a desired size is specified and then will bechanged, as engineering requires. Other existing cabinets and the desire to have the installed equipment close to eye height normallydetermine height. Height will normally be between 78" (2000 mm) and 82" (2100 mm). Width will be determined by available space and theinstalled equipment and will be 19" EIA, 24" (610 mm) or 36" (914 mm). Depth is determined by the depth of the installed equipment and willbe either 26" (650 mm) or 34" (850 mm).

Height_______Width________Depth_______

Power Distribution:

Power distribution is recommended and is usually provided. A power distribution panel with breakers and two (2) accessory power outlets willbe provided. Lighting will be provided in the top of the panel for ease of maintenance. Top or Bottom power and wire entry to the cabinetneeds to be specified along with input voltage.

TopTop___ Bottom ___ 110 VAC ___ 220 VAC ___

Intermediate Terminal Blocks:

Intermediate terminal blocks are a required feature on most cabinets. This allows for all the pre-wiring of the installed equipment to take

place at the cabinet shop. Only power and external wiring need to be attached to the terminal blocks. A listing of the equipment to beinstalled, installed transducers, recorder outputs and required relays along with a percentage of spare terminal blocks to be supplied isrequired. As an alternate, direct wiring to the installed equipment can be accomplished.

Intermediate TerminalsIntermediate Terminals___ Direct ___ 2-Wire Transducers _____ 3-Wire Transducers _____ Recorder Outputs _____ Relay's Required _____ % Spares Required _____



2-Wire Transducers _____ 3-Wire Transducers _____ Recorder Outputs _____ Relay's Required _____ % Spares Required _____

Paint and Color:

The specific exterior paint color of the cabinet may be optionally specified to meet existing cabinet specifications or preferences. Any color other than standard OEM colors must be specified in advance. Non stock paints will effect price and lead times.

Standard Standard Gray___ Specified ______________

Specified ______________ Standard Gray___ Specified ______________ Specified ______________

Doors:

Doors need to be specified as either left handed or right-handed.

Left HandedLeft Handed ___ Right Handed___

Right Handed___Left Handed ___ Right Handed___ Right Handed___

Accessories:

Smoke Detector___ Heat Sensor___ Transportation Eyebolts___ Door Locks___

Air Conditioning___ Heating___ Fan ___ w/Filter___ Glass Front Door ___

Air Purging___

Drawing Package:

Please see our standard drawing package as a starting point of drawing package options. Drawings can be provided in AutoCAD Format(Standard) as well as Intergraph. Drawings can be provided in prints, reproducible and digital formats.

AutoCADAutoCAD

Mechanical Drawing Package___ Electrical Drawing Package___ Field Transducer Interconnect___

Field Service

Proper installation is critical to the performance, integrity andcredibility of your VibrationMonitoring or Information System.A proper installation is dependenton, quality design and planning,

The basic steps to complete a turnkeyare:

1. Site Visit2. Preparation of Bid Documents3. Purchase Order



quality workmanship, training, and proper configuration/programmingof your vibration monitoringsystem.

Several levels of Field Service areavailable for your convenience and budget. First time users of vibrationmonitoring equipment andinformation systems usually requiremore assistance than those whohave completed several installations.In many cases even experiencedcustomers simply do not have themanpower available during crucialoverhaul periods to properly

complete system installation withoutassistance.

The following services are availablefrom our Field Service Group andyour requirements may of course becustomized depending on your circumstances and requirements.

Pre-Installation Survey Installation Assistance

Startup and Calibration Complete Turnkeys Training

Pre-Installation Engineering

The purpose of the pre-installationengineering is to finalize as manydetails as possible as to the systeminstallation. A site survey iscompleted and detailed drawings

and documentation are prepared toguide the customer in completingthe actual installation.

Site (Machine) Survey

Machine Survey Available Documentation

4. Preliminary DesignDocuments

5. Electrical Work (Conduit, J-Boxes and Wire)

6. Machine Shutdown

7. Mechanical Work (Machining/Drill & Tap)

8. Transducer Installation9. Monitor Installation10. Wiring Terminations11. Startup and Calibration12. Training13. Final Report and

Documentation

Training

Training is an integral part of asuccessful Vibration System Startup.Operators must be instructed in theuse of the system, InstrumentTechnicians must be instructed as tothe care and maintenance of thesystem. Training may be customizedfor each plants requirements andneeds.

Basic Vibration Training Transducer Theory and

Operation Operator Monitor Training Instrument Technician

Training

Customer Responsibility

The customer normally takesresponsibility for the following items

except in the case of a Turnkey project or if other arrangements have been made:

Machine preparationincluding:

o Bearing Drilling andTapping



Survey Transducer Application

Review Transducer Installation

Methods

Junction Box Locations Conduit Layout Monitor Location and

Mounting Alarm (Relay) Requirements

and Logic Startup Contact

Requirements Recorder Output

Requirements

Office Engineering

Machine Layout Drawing Transducer Installation

Detail Drawing Transducer Bracket Design Bracket Fabrication

Drawings Conduit and Instrument

Wire Drawings Purchase Order Review

Formal Report withDrawings

Installation Assistance

In many cases customers desire onsite supervision during systeminstallation. Our Field ServiceTechnician will supervise and assistin all phases of the actualinstallation to insure system

reliability and integrity.

Transducer Installation Monitor Installation Instrument wire terminations

Startup and Calibration

o Bearing Machining if required

Transducer Installation Junction Box Installation Conduit Installation

Instrument wire pulls Wire Terminations Panel Cutout and preparation AC Power for Monitor Installation of Monitor

Project Checklist

Pre-Installation:

Application Review

o Plant Objectiveso Transducer Call-outo System Call-out

Purchase Vib. System

System Design:

Transducer InstallationDetails

o Mounting Designo Bracket Design

Field Wiring Layouto Junction Box Layouto Conduit Layouto Transducer Field

Wiringo AC Power o Recorder Outputso Relay Outputso Digital

Communications Monitor/Mux. Installation

Details Custom Panel Design (if

required) Bill of Materials Purchase Non-IRD Supplied

Materials



The most often performed service,Startup and Calibration includes thefollowing after the preliminaryinstallation has been completed bythe customer:

Transducer calibration check Transducer mounting (gap)

check Wire termination Check Monitor Power Up Monitor Configuration and

Programming Monitor calibration Baseline machine

information

Formal Report

Turnkey

In some cases customers may electto have a Turnkey performed due tomanpower or time shortages.Turnkeys remove almost all planning and manpower requirements from the customer. Wewill take responsibility for all facets

of the installation from transducer installation, conduit and wire, tomonitor installation and programming. Extensivedocumentation packages areincluded for future reference.

Turnkeys require advanced planningand a prior site visit by one of our Field Service Representatives whereyour requirements, installation

details, electrical and mechanicalcontractors can be discussed.

Installation:

Install Monitor/Host/Mux. Install Conduit Pull Wire:

o Transducer InstrumentWire

o AC Power o Recorder Outputso Relay Outputso Digital

Communications Fabricate Brackets/Adapters Prepare Brgs/Caps (Drill/Tap) Prepare Notch/Proj. for

Speed/Phase

Install Transducers Install Flex. Conduit & Trans.

Wire Terminate Field Wiring

Start-up and Calibration:

Wiring Check Power up System Calibrate System Install/Program

Software/Firmware System Start-up Validate System Training System Acceptance\Sign-off

Maintenance Methods

Every facility that produces a This method has the highest cost for



consumer product has somerequirement for maintenance or upkeep of their machinery.Depending upon the product and, tosome extent, the size of the facility,

this maintenance activity may becontinuous in nature or periodic.Some maintenance activities mayconsume a significant portion of thefacility expenses and manpower.

Facility maintenance activitiesgenerally fall into three categories: breakdown, preventive, and predictive. Each category has particular costs associated and

specific benefits.

BREAKDOWN MAINTENANCE

This method has no continuousactivity associated with it.Essentially, no maintenance activityis performed on machinery until itfails or produces unacceptable product. At first impression thismethod seems the most cost

effective because the manpower andtheir associated costs are minimal.

But closer examination shows thatwhen the machinery fails,considerable expense is required toallocate manpower on an emergency basis, repair/replacement parts, andlost revenues due to non-productioncan mount rapidly depending uponthe manufacturing process or

product. Clearly, this method has thehighest associated cost andmaintenance is unpredictable at best.In addition, an unexpected failurecan be dangerous to personnel andthe facility.

PREVENTIVE MAINTENANCE

replacement parts because thefacility may have a separate programor department with the sole purposeof maintaining an inventory of spare parts and scheduling outage activity.

Maintenance costs are reduced because the "annual outage" or "turnaround" is usually scheduled for a period when the product demand islow. Additional cost savings arerealized because manpower and anyheavy equipment are scheduled.

PREDICTIVE MAINTENANCE

Throughout the decade of the 1980s

many facilities began to seek solutions to high maintenance costsand spare parts inventories. Byadopting a continuous approach tofacility maintenance these reductionscan be realized. Supporting thisapproach was the profusion of portable data collectors and databasesoftware. As an extension or enhancement to a portable datacollector system, which can have an

elevated associated manpower cost,is a permanently installed monitoringsystem. Many of these systems can be interfaced to advance softwaresystems that can assist with signalanalysis. The key to this enhancedsystem is having the sensors installedwhich are available for signalacquisition continuously.

Using these systems, and the

appropriate training necessary for signal interpretation, a facility canimplement a predictive maintenance program. This method relies on thedata collected, either on a continuous basis or on a routine, periodic basis,to dictate the required maintenance procedure and when to schedule the



An advancement on a breakdownmaintenance program is a preventive program. This periodic approach tomaintenance has little continuous

activity associated with it. It involvesscheduling a regular outage, usuallyon an annual basis, where the entiremachine train or plant is shutdown,or removed from production, for careful inspection and routinereplacement of specific parts.

maintenance activity. Granted, thescheduling is a subjective topiccontrolled by spare parts inventory,manpower availability, and productdemand. By evaluating all these

parameters a scheduled outage can be determined and all associatedcosts can be reduced.

Maintenance Method Checklist

1. Breakdown Maintenance2. Preventive Maintenance3. Predictive Maintenance

Machine Classifications

The type of vibration monitoringsystem to install on a machine trainwill depend upon how the machinetrain is used in the plant process. Avery important machine train willrequire more instrumentation tomonitor its health and operation

condition, while other machine trainshaving less important roles in plantoperation will have less monitoringrequirements. Using the techniques presented in this application note themachine trains may be classified ascritical, essential, or general purpose.Topics such as how the machinetrain relates to plant production, plant and personnel safety, andwhether they are spared are used as

classification criteria. Additionally,costs, such as lost production,replacement, insurance, andmaintenance expenses are factoredinto this classification.

CRITICAL

Vibration monitoringinstrumentation should providecontinuous, full time monitoringcapabilities. Some systems willdisplay every channelsimultaneously so that rapidassessment of the entire machine

train can be made. See STIApplication Note, MonitoringClassification, Classif-1 for additiondiscussion about monitoringinstrumentation.

ESSENTIAL

Essential machine trains may havethe same attributes as criticalmachine trains, but their importance

to the plant production process willnot be as important. They may haveinstalled spare units which can bestarted without significantinterruption of the plant process.They may be high horsepower or high speed, but will not have tooperate for extended periods or



Critical machine trains are requiredto maintain plant production andmany times are an integral part of the plant process. They are highcapital cost items and must operate

continuously, may not be spared, or have an installed backup unit, because interruption of the processto startup the backup machine trainwould have an undesirable effect onthe plant operation. Other machinetrains may be involved in anoperation which is important to plantor personnel safety.

Some plant designs that incorporate

several identical unists may appear to have installed spares, but all unitswill be reauired for 100% plantoutput. Other plants will have certainmachine trains which are required tooperate continuously duringemergency situations.

High horsepower and high speedmachines would be classified criticalif they are required to operate

continuously for extended periodswithout interruption of the plant process. All of these machine trainsshould be considered critical tocontinued plant operation and,therefore, qualify for higher expenditure on monitoringinstrumentation.

continuously. Maintenance budgetswill not be as costly when themachine fails, thus classifying thesemachines as essential and will nothave the same monitoring

instrumentation requirements ascritical machines.

Vibration monitoring systemsinstalled on essential machines can be of a scanning type, where thesystems switches from one sensor tothe next to display the sensor outputlevels. Many of these systems arecontrolled with solid statemultiplexers and switch channels

every second or so.

GENERAL PURPOSEGeneral purpose machine trains are all otherswhich are not classified as critical or essential. They are usually sparedand are not critical to plant production. They usually haveauxiliary roles or support other processes, may only operate ondemand, stocked replacement parts,

and maintenance costs are relativelylow when compared to critical or essential machine.

Due the machine classification, thesemachines do not qualify for permanently installedinstrumentation and a continuousmonitoring system. These machinesare usually monitored with a portableinstrument.

Machine Classification Checklist

1. Critical2. Essential3. General Purpose



Monitoring System Users

Smooth operation of a machine traininvolves the cooperation of severaldisciplines which interface with eachother and a vibration monitoringsystem in differing manners. Somedisciplines will have very frequentusage of the information supplied,while others will interfaceinfrequently or even indirectly.

OPERATIONS Operations personnel are usually thefront line users of vibrationmonitoring systems. They will beinterfacing on a frequent basis todetermine whether the production process requires adjustment.Operating vibration levels are onethe indications of how a machinetrain is behaving to whatever theoperators are doing. Shouldsomething happen to the machinetrain, the operators will require someform of warning of an impendingfailure. This warning may be audibleor visual in nature. Advancedwarning of an impending failure will provide the opportunity for acontrolled machine shutdown, whereauxiliary support or process supplysystems can be secured.

MAINTENANCE Maintenance personnel will mostlikely interface indirectly toinformation supplied via a vibrationmonitoring system. Their involvement will be to scheduleactivities economically, if at allfeasible. Scheduling activities allowsallocating manpower effectively,spare parts on an "as-needed" basis,and any required externalcontractors. All of these activitieswill reduce unplanned outages andoverall maintenance costs.

VIBRATION SPECIALIST The vibration specialist'sinvolvement with a vibrationmonitoring system will be infrequentunder normal conditions. As thelevels increase, the vibrationspecialist will be called upon todiagnose the cause for the higher levels. This type of analysis isusually at the component level,where interest is focused at one particular machine case or bearingwhere the levels are elevated. After any required maintenance has beencarried out the vibration specialistwill be called upon to collect detailedvibration information as part of a"new" machine acceptance test program. Successful completion of these tests will culminate in the

machine being commissioned, readyfor process operation.

Users Checklist

1. Operations2. Maintenance



3. Vibration Specialist

Monitoring Classification

A subset of the decision of purchasing a monitoring system isthe decision of what type of systemis required. Monitors are available inmany varieties; some simply displaythe overall signal levels, some haveelaborate interface systems, somecan automatically collect differenttypes of information. The end user

must decide what is really necessary.Will the monitor be required to provide some form of protection?Does the end user require that themonitor provide some type of information? Must the monitoringsystem provide diagnosticcapabilities?

PROTECTION

Protection is available in manyforms. Nearly all monitor systemsavailable today can providemachinery protection. This meansthat should a sensor signal exceed a predetermined set point the monitor can initiate a shutdown to preventinternal machinery damage. Thisform of protection is tangible andcan be quantified for accounting purposes. Additional intangible

protection provided by a basicmonitoring system are personnel and production protection. If a machinecan be shutdown prior tocatastrophic damage, which couldinvolve unexpected shrapnel fromthe machine, the personnel that arein the vicinity of the machine are

INFORMATION

An information system will providedata that is useful for planning andscheduling. This information can beused for a "Go No-Go" decisionwhether to continue operating themachine train or produce goods.Basic monitoring systems are

capable of providing this type of information by alerting personnel tocurrent conditions.

Maintenance planning and outagescheduling requires additionalinformation. Information systemswill provide data as trends whichgive advanced notice of elevatingoverall signals.

DIAGNOSTICS

Advanced monitoring systems will provide additional information aboutthe condition of the machine trainconnected to the monitor. Thisinformation can be collectedautomatically or manually, and uponalarm activation or on a regular basis.

This information has many benefitswhich when properly used can produce cost savings and downtime.By analyzing the collectedinformation the root cause of theelevated signals can lead to the causeof the machine problem. This type of information can lead to reducing



protected.

An orderly shutdown of a machinetrain can benefit the facility production and its product. Certain

production processes, such as paper and sheet steel, are sensitive toexcessive vibration. High vibrationlevels produce poor quality product.These facilities will benefit from amonitoring system that can alertoperation personnel whenunacceptable product is being produced.

machine train downtime. After themaintenance has been conducted,this type of monitor can be used for acceptance testing and machinecommissioning. Many end users

have reported correction of designflaws and incorrect operating procedures using advanceddiagnostic information.

Monitoring Classification Checklist

1. Protection2. Information3. Diagnostics

Rolling Element Bearings

The basic purpose of a machine bearing is to provide a near frictionless environment to supportand guide a rotating shaft. Twogeneral bearing styles are utilized atthis time: the journal bearing and

the rolling element bearing. For lower horsepower and lighter loaded machines, the rollingelement bearing is a popular choice.

Until the 1940's, the journal bearingwas the prevalent style used onmachines. As metallurgy andmachining techniques progressed,the rolling element bearing gainedgreater usage. Today many of the

smaller process support machineshave rolling element bearings.

FAILURE MONITORING

This style of bearing is typicallymonitored using a case mountedtransducer: an accelerometer or velocity pickup. A displacement

sensor observing the shaft relativevibration would show little, if any,vibration due to the vibration nodecreated by the bearing.

Using signal integration techniques,found in many industrial datacollectors, specific frequency rangesrelating to certain defects can beemphasized. Acceleration signals,obtained from case mounted sensors,

emphasize high frequency sources,while displacement signals emphasizelower frequency sources, withvelocity signals falling between theextremes. Recent innovations for determining bearing condition areAcceleration Enveloping, SpectralEmitted Energy (SEE), and Spike



BEARING DESIGNS

RollingElementBearings

have four components: an inner race, an

outer race, a rolling element, and acage to support, space, and guidethe rolling elements. The rollingelements found in today's rollingelement bearings include: balls,rollers, and tapered rollers. Allrolling element bearings have one

thing in common: all parts must bein physical metal to metal contact atall times. Installation instructionsspecify the amount of bearing pre-load to maintain the componentcontact.

Rolling element bearings havesome unique concerns not found in journal bearings. A rolling element bearing will always force a

vibration node at its location.Because of the metal to metalcontact, this bearing will providevery little vibration damping.Although these bearings are a very precisely machined part they have alimited lifetime. Each componentof the bearing will generate specificfrequencies as defects initiate and become more prevalent.

Spherical Ball Spherical ball bearings, as thename implies,utilize sphericallyshaped balls asthe rolling or loadcarrying element.

Energy. These measure highfrequency resonances generated by bearing defects. As a trended variable,in conjunction with other parameterssuch as displacement, velocity or

acceleration, they can give the earliestindication of bearing defects.

The figure depicts the overallamplitude levels obtained from a

bearing as it progresses throughcontinuing phases of failure. Thischart depicts overall vibration levelsonly. As time progresses the earliestindication of failure are obtained fromfiltered high frequency signals because these signals are generated bythe resonance of the bearing and by bearing component defects.

During the early stages of failure the

other three parameters may notgenerate enough signal to be detected because these parameters emphasize progressively lower frequency ranges.As failure continues and the damaged bearing generates the individual bearing defect frequencies, the other parameters register signals.



The number of balls used in a bearing varies depending on theapplication. This rolling element bearing type is designed to carry both radial and axial loads. By

modifying the design, this bearingcan also accommodate large axialloads.

Cylindrical/Spherical Roller This type of bearing utilizescylindricallyshaped rollers asthe load carryingelement. This

bearing type isdesigned to carry

large radial loads. This bearing, inits basic configuration, is not wellsuited to counter axial loads. Therollers may actually be slightly barrel shaped in certain designs.Barrel shaped rollers and their associated outer race allow for some self alignment of the bearing. Needle bearings are a special

adaptation of the cylindrical roller bearing.

Tapered

Roller/Land This bearingdesign is aspecialadaptation of the cylindricalroller bearing.

This bearing is designed to counter axial thrust loads along withcarrying radial loads. Due to thegeometrical summation of theradial and axial loads, this bearinghas a lower radial load limit than asimilarly sized cylindrical or spherical bearing.

Viewing the four monitoring parameters as spectra, additionalinformation can beobtained

about thefailuremodes.Thisfigureshowsthespectrumfrequency contentduring

four stages of bearing failure. A normal bearing or newly installed bearing willshow no frequencies except thoseassociated with shaft phenomenonsuch as balance or misalignment.

Stage I

Stage I has some very high frequencycontent in the Spike Energy region.This zone is in the ultrasonic regionwhich requires a sensor specificallydesigned to detect in this region.Special circuitry filters pass onlythose signals. Physical inspection of the bearing at this stage may not showany identifiable defects.

Stage II

Stage II begins to generate signals



Certain applications may employtapered rollers along with taperedraces, hence the name. Special bearings may have inner and outer races with differing angles.

VIBRATION MONITORING

APPLICATIONS

Rolling element bearings, by their design and installation, provide avery good signal transmission pathfrom the vibration source to theouter bearing housing. Also, these bearings require monitoring of theunique bearing frequencies

generated by the various parts of the bearing, in addition to the rotor fault frequencies.

Bearing Frequency Calculation Although modern rolling element bearings are very preciselymachined, they do have micro-defects which are potential sites for future damage. Due to the precisetolerances, improper installation

practices can reduce bearing life.Extensive information has beencompiled about bearing defectfrequencies.

The figure lists the bearing defectfrequency formulas for a defect onthe balls or rollers, outer race, inner race, and cage. The assumption for these formulas is that the outer raceis stationary while the inner racerotates.

associated with natural resonancefrequencies of the bearing parts as bearing defects begin to "ring" the bearing components. A notableincrease in Zones 3 and 4 region

signals is associated with this stage.Beginning signs of defects will befound upon inspection.

Stage III

Stage III condition has the

fundamental bearing defectfrequencies present. These frequenciesare those discussed previously in this paper. Harmonics of these frequenciesmay be present depending upon thequantity of defects and their dispersalaround the bearing races. Theharmonic frequencies will bemodulated, or side banded, by theshaft speed. Zone 4 signals continueto grow throughout this stage.

Stage IV

Stage IV is the last condition beforecatastrophic failure of the bearing.This stage is associated withnumerous modulated fundamental

frequencies and harmonics indicatingthat the defects are distributed aroundthe bearing races. Due to the increaseddegradation of the bearing the internalclearances are greater and allow theshaft to vibrate more freely withassociated increases in the shaftfrequencies associated with balance or



If the bearing dimensions areknown, the individual bearingdefect frequencies can be calculated precisely, or a general rule of thumb can be applied. Using the

generalized form the inner racefrequencies would be N x RPM x60% and the outer race frequencieswould be N x RPM X 40%. If the bearing manufacturer modelnumbers are known severalcomputer programs are available tocalculate the necessary frequencies.

mis-alignment. During later phases of stage IV, the bearing fundamentalfrequencies will decline and bereplaced with random noise floor or "hay stack" at higher frequencies.

Zone 4 signal levels will actuallydecrease with a significant increase just prior to failure.

Journal Bearings

Industrial machinery with highhorsepower and high loads, such assteam turbines, centrifugalcompressors, pumps and motors,utilize journal bearings as rotor supports.

One of the basic purposes of a bearing is to provide a frictionless

environment to support and guide arotating shaft. Properly installed andmaintained, journal bearings haveessentially infinite life.

BEARING DESIGN

A journal bearing, simplystated, is acylinder which

surrounds theshaft and isfilled withsome form of fluid lubricant. In this bearing a fluidis the medium that supports the shaft preventing metal to metal contact.The most common fluid used is oil,

Plain Bearing The plain bearing is the simplest andmost common design with a highload carrying capacity and the lowestcost. This bearing is a simplecylinder with a clearance of about 1-2 mils per inch of journal diameter.Due to its cylindrical configuration itis the most susceptible to oil whirl. It

is a fairly common practice duringinstallation to provide a slightamount of "crush" to force the bearing into a slightly ellipticalconfiguration.

Lemon Bore The lemon or elliptical bore bearing is avariation on the

plain bearingwhere the bearingclearance isreduced on one direction. Duringmanufacture this bearing has shimsinstalled at the split line and then bored cylindrical. When the shims



with special applications using water or a gas. This application note willconcentrate on oil lubricated journal bearings.

Hydrodynamic principles,which areactive as theshaft rotates,create an oilwedge thatsupports the

shaft and relocates it within the bearing clearances. In a horizontallysplit bearing the oil wedge will lift

and support the shaft, relocating thecenterline slightly up and to one sideinto a normal attitude position in alower quadrant of the bearing. Thenormal attitude angle will dependupon the shaft rotation direction witha clockwise rotation having anattitude angle in the lower leftquadrant. External influences, suchas hydraulic volute pressures in pumps or generator electrical load

can produce additional relocatingforces on the shaft attitude angle andcenterline position.

An additional characteristic of journal bearings is damping. Thistype of bearing provides much moredamping than a rolling element bearing because of the lubricant present. More viscous and thicker lubricant films provide higher

damping properties. As the availabledamping increases, the bearingstability also increases. A stable bearing design holds the rotor at afixed attitude angle during transient periods such as machinestartups/shutdowns or load changes.The damping properties of the

are removed the lemon bore patternis results. For horizontally split bearings, this design creates anincreased vertical pre-load onto theshaft.

This bearing has a lower loadcarrying capacity that plain bearings, but are still susceptible to oil whirl athigh speeds. Manufacturing andinstallation costs are considered low.

Pressure Dam A pressure dam bearing is basically a

plain bearingwhich has beenmodified toincorporate acentral relief groove or scallop along the top half of the bearing shell ending abruptlyat a step. As the lubricant is carriedaround the bearing it encounters thestep that causes an increased pressure at the top of the journal

inducing a stabilizing force onto the journal which forces the shaft intothe bottom half of the bearing.

This bearing has a high load capacityand is a common correction for machine designs susceptible to oilwhirl. Pressure dam bearings are aunidirectional configuration.

Another unidirectional bearing

configuration is the offset bearing. Itis similar to a plain bearing, but theupper half has been shiftedhorizontally. Offset bearings haveincreasing load capacities as theoffset is increased.

Tilting Pad



lubricant also provides an excellentmedium for limiting vibrationtransmission. Thus, a vibrationmeasurement taken at the bearingouter shell will not represent the

actual vibration experienced by therotor within its bearing clearances.

Journal bearings have many differingdesigns to compensate for differingload requirements, machine speeds,cost, or dynamic properties. Oneunique disadvantage whichconsumes much research andexperimentation is an instabilitywhich manifests itself as oil whirl

and oil whip. Left uncorrected, this phenomenon is catastrophic and candestroy the bearing and rotor veryquickly. Oil whip is so disastrous because the rotor cannot form astable oil wedge consequentlyallowing metal to metal contact between the rotor and the bearingsurface. Once surface contact existsthe rotor begins to precess, in areverse direction from rotor rotation

direction, using the entire bearingclearance. This condition leads tohigh friction levels which willoverheat the bearing babbit metalthat leads to rapid destruction of the bearing, rotor journal, and themachine seals.

Some common designs employed arelemon bore, pressure dam, and tilt pad bearings. These designs were

developed to interrupt and redirectthe oil flow path within the bearingto provide higher bearing stabilities.

GEOMETRIES

Journal bearings installed inindustrial machinery today generally

Tilting pad bearings is a partial arcdesign. Thisconfiguration

has individual bearing padswhich areallowed to pivot or tilt to conformwith the dynamic loads from thelubricant and shaft. This type of bearing is a unidirectional design andis available in several variationsincorporating differing numbers of pads with the generated load appliedon a pad or between the pads.

VIBRATION MONITORING

A shaft supported by journal bearings will move relative to the bearing housing as various forces areimposed onto the shaft. A vibrationtransducer is required which canmonitor the relative motion betweenthe shaft and the bearing. Higher vibration frequencies are not of

prime concern since they would not be transmitted through the oil filmreliably.

The only sensor available that canmeasure relative measurements of the shaft is the non-contacting pickup, sometimes called adisplacement, eddy current, or proximity pickup. This type of sensor measures the relative

vibration of the shaft and, also, therelative position of the shaft withrespect to the bearing clearances.High frequencies such as blade passage and cavitation would beattenuated by the lubricant. Casemounted sensors would not providean accurate indication of the



fall into two categories: full bearingsand partial arc bearings. Full bearings completely surround theshaft journal with many differinggeometries such as elliptical, lobed,

or pressure dam configurations andusually are two pieces, mated at asplit line. Partial arc bearings haveseveral individual load bearingsurfaces or pads and are made up of numerous adjustable components.

The bearing inner surface is coveredwith a softer material, commonlycalled babbit. Babbit, which is a tinor lead based alloy, has a thickness

that can vary from 1 to 100 milsdepending upon the bearingdiameter. A babbit lining provides asurface which will not mar or gougethe shaft if contact is made and toallow particles in the lubricant to beimbedded in the liner withoutdamaging the shaft.

vibration due to the inherentdamping offered by the lubricant between the shaft and the bearing.For more information aboutinstallation and theory of operation

of NCPUs, see the STI Application Notes: Eddy Current Transducer Installation, Part 1-Radial Vibration.

System Response

An understanding of how a springmass system responds to vibratoryinfluences is helpful inunderstanding, recognizing andsolving many problems encounteredin vibration measurements. In thisapplication note the combinedeffects of system mass, stiffness, anddamping properties are presented to

reveal the cause and characteristicsof resonance.

All machines have three fundamentaltraits which combine to determinehow the machine will react toexcitation forces. These traits arestiffness K, damping D, and mass M.

Below Resonance If each of theindividualterms arerepresented by a vector,and the influences of frequency areincluded, the result is a type of

graph, similar in shape to a triangle.The figure is a graphicalrepresentation of the relationships of the terms at low frequencies, i.e.slow rotor speeds. The total restraintvector is the summation of all threevector terms. Note that the dampingand mass terms do not have much



These traits, actually represent forcesinherent to every machine andstructure, tend to resist or opposevibration.

From an analysis standpoint, itshould be remembered thatmachines, along with their supporting structures, are complexsystems consisting of many spring-mass systems, each with its ownnatural frequency. Also, each of these systems may have differingdegrees of freedom with a differingnatural frequency. This collection of possible resonant frequencies, and

the many excitation frequencies, allcombine to make resonance a verycommon problem for the transientvibration analyst. Understanding the basics of how a system responds tovibratory forces is important toanyone involved in vibrationmeasurement, analysis, and balancing. From a measurementstandpoint, it is important toremember that every object has a

resonant frequency...machinery, pickups, brackets, etc. Resonance of a pickup mounting bracket, or the pickup itself, will introducesignificant errors to measurements.

RESTRAINING FORCE

The combined effects of therestraining forces of stiffness,damping, and mass determine how a

system will respond to a givenexciting force. Mathematically therelationship is represented by:M a + D v + K x = Me ² e sin( t - )

For simplification, the aboveequation can be written as:Mass term + Damping term +

influence on the total restraint at lowfrequencies, leaving the stiffnessterm as the dominant term. Thismeans that at frequencies below theresonance frequency the rotor

behaves as a pure spring, sometimescalled a stiff shaft rotor.

At Resonance As the rotor frequencyincreases,theinfluence of

the damping and mass terms becomegreater due to the influence of and ²

in the mass and velocity terms. At acertain frequency the stiffness andmass terms cancel each other due tothe 180 phase difference in theterms. The figure presents thevectorial relationships and theresultant vibration amplituderesponse at resonance condition.

When these terms cancel each other the only remaining restraint term is

the damping term to control thesystem vibration. As the stiffness andmass terms approach the point of canceling each other, the system'svibration amplitude will increase to amaximum, constrained only by theavailable damping from anylubricant present. At resonance thesystem has lost the restraining forcesof the stiffness and mass terms. Amachine supported by rolling

element bearings, which have littleor no damping capabilities, willexhibit a dramatic and sharp increasein vibration amplitude in this region.This phenomenon is referred to theresonance frequency or "critical"speed. Operation in this zone should be avoided since any change in the



Stiffness Term = Restraining Force

The restraining forces, represented by the various terms in the equation,are what determines how a rotor

behaves throughout its operatingrange. Any excitation force, such asunbalance, is always in equilibriumwith the restraining forces of mass,damping, and stiffness. The amountof measured vibration, as a result of these combined forces, will dependupon the combined effect of all threeterms in the equation. The phaseangle ( ) change as a rotor increasesspeed and surpasses a resonance

region is dependant upon on therelationship between the variousterms.

PHASE RELATIONSHIP

To understand the phaserelationships of the terms, consider that the mass term is proportional toacceleration, damping term is proportional to velocity, and the

stiffness term is proportional todisplacement. In equation form, theacceleration term = -x ² sin( t) andthe velocity term = x cos( t).Examining the relationship of theacceleration and velocity equations,a 90 phase difference exists as theterms are integrated. Another integration produces the stiffnessterm that is proportional todisplacement (x) only, and the

relationship between the stiffnessand damping terms have another 90 phase shift.

The effects of frequency ( ) shouldalso be considered along with the phase shifts noted. Stiffness being proportional to displacement only,

available damping can have adramatic effect upon the measuredvibration levels.

Above Resonance

As the rotor frequencycontinues toincrease, themass term,which is proportional to ², becomes the predominant portion of the totalrestraint force, growing faster thanthe other terms. The figure shows thevector representation of the forces

and the vibration amplitude at highrotor frequencies. Note that as speedincreases further the phase anglechange approaches another 90 shift.The rotor behaves as a pure masswith little impact from the constantstiffness term and the relativelyslowly changing damping term. Arotor operating in this region iscalled a flexible rotor since it rotatesaround its mass centerline, not its

geometric centerline.

Thus, as rotor frequencies increase,three regions are found where one of the component terms is dominantover the other two terms. Thesummation of the three terms isrepresented by the vector labeled:Total Restraint. The total restraintvector is what is measured asvibration amplitude and its

associated phase angle. As the rotor speed passes through each of theseregions the measured phase anglewill change by 90 and will exhibit anoverall phase shift of 180 as itsurpasses a critical "resonance"speed: Nc.



and not influenced by frequency,means that essentially the stiffnessterm is constant throughout allfrequency ranges. However, thedamping and mass terms are

influenced by and ², respectively.

Balancing

Unbalance is a very common source of highvibration that is identified by excessivelevels of vibration amplitudes at a frequencythat is synchronous with machine speed.Possible sources for this malfunction may

be uneven product deposits around a fan or pump impeller, damaged or missing bladesor vanes, improper shaft componentassembly, or any other uneven massdistribution around the rotor axis. Amachine that is operated within designconditions and is clean should notexperience severe unbalance vibrationlevels.

Correction of unbalance situations involves

characterizing the heavy spot. The heavyspot is the radial location at which theexcessive radial

massdistribution exists. This heavy spot is

always a location which is opposite thelocation where weight needs to be added.Unfortunately the location of the heavy spotcannot be identified directly. However, wecan identify the high spot location which isthe point of maximum displacement. Thehigh spot is a radial location where the shaftexperiences its maximum displacement, or

If the vibration levels are notacceptable after a balanceweight has be placed on therotor, a further weight calleda trim weight can be installed.

Sometimes this weight will be required to be located at aslightly different locationthan the previous weight.

Balance Weights Trial, balance, and trimweights may be made fromany material. Commonmaterials are bolts, washers,or C- clamps. Strong tape,

such as duct tape, issometimes used totemporarily attach a trialweight. Some machines willrequire an imagination todevise an attachment method.Other machines, such asturbo- machinery, have balance weight holes provided which may requiresome machine disassembly to

access them.

If the weight must be locatedat the same location as a previous weight both weights(previous balance weight andtrim balance weight) should be combined into a single



excursion, due to the unbalance force(s).

A definable relationship exists between theheavy spot and the high spot which dependsupon the rotor critical speed. If a rotor never

experiences a critical speed throughout itsoperating range the heavy spot will always be the exerting force resulting in the highspot coinciding with the heavy spot, just asa weight being swung on a string exerts adisplacing force on the rotation axis.

As the rotor surpasses the first critical speedthe heavy spot and high spot split apart untilthe they are separated by about 180 . This phenomenon occurs because the rotor now

rotates around its mass centerline instead of its geometric centerline, forcing the highspot to be the location of maximumdisplacement. If the rotor continues toincrease its speed and experiences another critical speed the high spot rotates another 180 until the high spot coincides with theheavy spot. This 180 shift in the high spot'srelationship to the heavy spot continues assubsequent critical speeds are surpassed.For further discussion about the relationship

between critical speeds and the associated phase shift, see STI Application Note,System Response (F=MA).

Correcting an unbalance condition involvesusing phase measurements to locate the highspot orientation, determining therelationship of the high spot to the heavyspot, and finding the magnitude of theunbalance by measuring the influence of correction weights. Careful observation of

the machine vibration level and phase angleas it progresses from stand still to fulloperating speed will identify whether thetrial weight should be installed at the highspot phase angle or opposite the high spot phase angle. Following this observation thevibration amplitude levels will almostalways be reduced during the first balance

weight by making a larger weight or using a material of greater density.

All weights, regardless of

material or attachmentmethod, must be held rigidlyin place. This may beaccomplished by staking,welding, etc. Relying on a bolt and nut with split washer should not be considered. If a bolt and nut configuration isused the bolt should have itsthreads staked to prevent thenut from shifting.

Balancing Speed Operating speed has a drasticeffect upon the measuredunbalance vibration levelsince the centrifugal forcedue to unbalance is proportional to speed squared( rpm²). During the balancing procedure the operating speedat which amplitude and phase

measurements are gatheredmust be held steady at a predetermined speed. Allsubsequent balance attemptsshould be made with themachine operating at thesame speed as previousattempts. Future balanceactivity should be made atthis same speed so that the balance response

coefficient(s) can be includeinto calculations to reduce thenumber of machine restarts.

Balance Response

Coefficient Once a rotor has been balanced, or has had a trial



run, which usually impresses the machineoperators.

Two balancing procedures prevail: the fixed protractor method where a mark on the rotor

is used for

phasemeasurements with a strobe light; therotating protractor method where a phasetransducer observing a once-per-turn event

is used for phase measurements. Bothmethods are a trial weight balancing procedure where an initial weight isinstalled to determine how the rotor responds to it.

Fixed Protractor Method In the fixed protractor method a strobe lightis usually used to measure the phase angle by observing a

singlemark, or unique item, such as a key or machined hole. Some procedures use a

optical tachometer, instead of a strobe light,to obtain phase measurements. Phase anglemarkings are made on the stationary part of the machine. This method counts increasing phase angles in the direction of rotor rotation. Phase measurements are taken byobserving a single mark on the rotating shaftusing the strobe light which is tuned to the

weight installed that producesa change in vibrationamplitude or phase, anunbalance constant can becalculated which can be used

in future balance attempts.Each rotor will have adifferent unbalance constant,given in units of weight per amplitude (lbs/mil, oz/mil,gm/in/sec, etc.). A single plane balance procedure will produce one balance responsecoefficient. Multiple plane balancing will produce anynumber of coefficients

depending upon the number of balance planes. A three plane balance procedure will produce nine balanceresponse coefficients. Thisnumber is because as the trialweight is moved to eachsubsequent balance plane twoadditional responsecoefficients are calculated.

Correction WeightCalculations Once the vibration levels and phase angles have beencollected with and without atrial weight, the requiredcorrection weight can becalculated. Basically, thecorrection weight must besized to counteract thecentrifugal force generated by

the existing unbalance mass.The method for calculatingthe weight differs dependingupon the whether the rotor experiences a critical speedand the number of balance planes.



machine speed or using the photo-tachwhich observes a piece of highly reflectivetape placed on the shaft.

This method is commonly employed on

machines that do not operate above their first critical speed because the high spot andthe heavy spot are coincident meaning thatthe final balance weight location will beopposite the measured high spot. Also,instrumentation setup is easy, usually usingsurface mounted vibration transducers.

Rotating Protractor Method The rotating protractor procedure utilizes a phase Eddy Probe or an optical tachometer

to determine the

phaseangle.

The trigger for this method comes from a pulse signal observing a once-per-turnevent, such as a keyway or a key. Phaseangle measurements are the relation between the keyway and the high spot, withincreasing phase angles measured againstshaft rotation.

Setup for this method requires that the phasetransducer be installed observing a once-

per-turn event. An existing key or keyway isideal, but if non-existent a temporary keymay be epoxied in place.

The rotor will require phase marks installed,not to directly obtain measurements, but toassist in weight placement. A simplemethod for marking the shaft is to use a

One plane balance, operation below 1st critical speedCW = (G X CF) / (R X SPD²)where CF = (V X K) / 2and K = WR² X M

Two plane balance, operation below 1st critical speedInboard CWi = (CFi X G) /(Ri X SPD²)where CFi = (Vi X Ki) / 2and Ki = (WR² X M) / 2

Outboard CWo = (CFo X G) /(Ro X SPD²)where CFo = (Vo X Ko) / 2

AND Ko = (WR² X M) / 2

One plane balance, operationabove 1st critical speedCW = (CF X G) / (R X SPD²)where CF = (WT X V XSPD²) / 2 G

Two plane balance, operationabove 1st critical speedCWi = (CFi X G) / (Ri X

SPD²)where CFi = (½ WT ) X Vi XSPD² / 2 G

CWo = (CFo X G) / (Ro XSPD²)and CFo = (½ WT) X Vo XSPD² / 2 G

CF = Centrifugalforce due to

unbalance (lbs) WR = First critical

speed (rpm) K = Rotor spring

constant (lb/in) M = Rotor mass (rotor

weight (lb/G) G = Acceleration of



permanent marker to label the rotor with thezero degree angle aligned with the keywaythat the phase transducer observes. Whenmarking the rotor remember that phaseincreases against the rotor rotation direction.

Trial Weights Regardless of which balancing procedureused a "trial weight" is must be installed tomeasure how the rotor will respond to thisweight. This weight will induce a different balance condition with an accompanyingchange in the vibration level and/or phaseangle. This change, once introduced into balance calculations, will dictate how muchweight is required to be added and the radial

location for weight placement.

gravity (386.4 in/s²) V = Peak-peak

vibration (in) SPD = Balance speed

(rpm x 0.1047 rad/s)

CW = Correctionweight (lbs)

R = Balance weightradius (in)

WT = Rotor weight(lbs)

CFi = Inboard centrif.force due tounbalance (lbs)

Ki = Inboard rotor spring constant (lb/in)

Vi = Inboard pk-pk vibration (in)

CWi = Inboardcorrection weight(lbs)

Ri = Inboard balanceweight radius (in)

CFo = Outboardcentrif. force due tounbalance (lbs)

Ko = Outboard rotor

spring constant (lb/in) Vo = Outboard pk-pk

vibration (in) CWo = Outboard

correction weight(lbs)

Ro = Outboard balance weight radius(in)

Balancing Checklist

1. Balancing Method2. Trial Weight(s)3. Weight Material4. Balancing Speed5. Response

Coefficient(s)6. Final Weight



PlacementDocumented

Comparing Vibration Readings

Comparing vibration level readings taken bydifferent types of instruments and transducers can be very confusing and can lead to mistrust of thesystems involved.

Knowledge of how to properly compare readings isrequired before comparing any readings isattempted.

This application note explains the variablesinvolved in some detail and will act as a guidelineas you compare vibration readings.

Transducer Type

Three (3) basic types of vibration transducers areavailable which correlate with the three (3) types of measured physical motion, Acceleration, Velocityand Displacement.

Accelerometer Accelerometers are a piezo-electronic (crystal)device. A pre- loaded crystal is charged withcurrent and as the crystal is compressed or de-compressed by vibration an output proportional tog's (gravity) is provided. A "g" is equal to 9.80meters/second2 or one (1) standard earth gravity.

Shaft Absolute Shaft Absolute is the measurement of the shaft's motion relative to free space(or absolute). Shaft Absolute can bemeasured two (2) ways, the first beingelectronically summing the signals

of both a Eddy Probe measuring shaftrelative and a accelerometer measuringcase absolute, the second being using ashaft rider which is a spring mounteddevice that physically rides on thesurface of the shaft, normally a velocitysensor integrated to displacement ismounted on top of the shaft rider. Shaft

Absolute is normally used where therotating assembly is five (5) or moretimes heavier than the case of themachine.

Engineering Units

0 to Peak (0-P) Both Velocity (in.sec, mm/sec) andAcceleration (g's) by definition aremeasured in 0 to Peak or one/half the

Peak to Peak signal as viewed on anoscilloscope.

Peak to Peak (P-P) Displacement by definition is measuredin Peak to Peak or the actual Peak toPeak Motion of the Shaft.



Accelerometers are normally used for high-frequency bearing cap vibration readings(Case/Bearing Cap Absolute on machines usingrolling element bearings. Usually the output isintegrated electronically to velocity (in/sec or

mm/sec). Other applications include monitoringGears and High Frequency Applications.

Velocity Pick-up Two (2) types of Velocity Sensors exist,mechanical and electronic. Mechanical types arethe most common and are made up of a springmounted coil mounted inside a magnet. Vibrationcauses the coil to move in relation to the magnetwhich produces a voltage output directly proportional to Velocity. Electronic Velocity

Sensors are Accelerometers with an electronicintegrator built in to the unit. Output of a VelocitySensor can be expressed in many different terms,inches/second (in/sec) or millimeters/second(mm/sec) being the standards.

Velocity Transducers are normally used for BearingCap Vibration Monitoring (Case/Bearing CapAbsolute) on machines with rolling element bearings. They have the advantage of high outputsand the signal is read directly in velocity (in/sec or mm/sec).

Eddy Probes (Proximity) Eddy or Proximity Probes are a displacementdevice that measure the relative motion between the probe mounting location and the target (shaft).Output is directly proportional to displacement andis usually measured in mils (.001") or millimeters(mm).

Root Mean Square (RMS) Root Mean Square (RMS) is a popular method of measuring Case or BearingCap Vibration as many vibrationengineers have found that RMS is more

indicative of actual rolling element bearing condition. Although rarelyfound in vibration wave-forms a puresine wave RMS would be .707 timesthe 0 to Peak Value.

Transducer Considerations

Frequency Response The frequency response of a vibrationtransducer is very important when

comparing readings. Transducers with awider or broader frequency responsewill typically see more vibration if it is present than a narrower bandwidthtransducer. How different vibrationfrequencies contribute to overall valuesis dependent on their phase relationshipto each other, some may add, some maysubtract from the overall value.

Eddy Probes Displacement200

mv/milVelocity(Mechanical)

Velocity500mv/in/sec

Velocity(Piezoelectric)

Velocity500-1000mv/in/sec

Accelerometer Acceleration 100 mv/g

Mounting

How a transducer is mounted is also

critical to comparing measurements.Accelerometers are extremely sensitiveto the method of attachment.Differences in bandwidth can bemeasured between hand-held, magnetattached, epoxy, and stud mountedinstallations.



Eddy Probes are used on machines with Journal(Sleeve) type bearings. Where the measurement of motion between the Bearing and Shaft is critical.

Bearing Type

Two primary types of bearings are in use today,Rolling Element Bearings and Journal or SleeveBearings.

Rolling Element Bearings are zero (0) clearancedevices. All vibration of the shaft is transmitteddirectly to the bearing cap.

Journal or Sleeve Bearings are designed so that theoil film provides damping. The shaft is free tovibrate within the bearing. Due to the damping provided by the oil film very little of the shaftvibration is transmitted to the bearing cap. The oilfilm damping provides even higher levels of attenuation to higher

frequencies.

Measurement Type

Installation instructions must befollowed precisely to obtain themanufactures transducer specifications.Accelerometers not mounted perfectly perpendicular to the surface or on a flat

surface will produce stress risers whichwill also produce false signals.

Measurement Location

When comparing readings it is essentialthat all readings are taken at the samelocation and the same plane. Even smalldifferences in location can effect theoverall readings. All vibrationtransducers are single plane devices and

only measure in the plane in which theyare held or are mounted.

Instrument Considerations

All Instruments handle signal isdifferent ways. Different instrumentshave their own frequency response andfiltering. Knowledge must be gained onthe instruments used before the outputscan be compared even when they use

the same transducer.

Conversion Formulas

Displacement, Velocity andAcceleration are mathematically relatedto each other as a function of frequency.Electronic integrators or differentiationare also used to change one term to theother. Once again it must be understoodthat the readings be of the same type or

they will not agree.

D = Displacement, P-P, Mils.

V = Velocity, 0-P, in/sec.

A = Acceleration, 0-P, g's.

D = 19.10 x 103 x (V/CPM)

D = 70.4 x 106 x (A/CPM2)



Only measurements of the same type can becompared. Bearing Cap or Case Vibration cannot be directly compared to Shaft Relative or ShaftAbsolute and visa versa.

Case Absolute Case or Bearing Cap Absolute is the measurementof the Case or Bearings Caps (Location of Transducer) motion relative to free space (or absolute motion). Case or Cap Absolute is usuallyused for monitoring Rolling Element

Bearings.

Shaft Relative Shaft Relative is the measurement of motion between the Shaft and whatever the measuringdevise is mounted to. This measurement isnormally taken with a NCPU or Proximity Sensor.Shaft Relative measurements are used for Journalor Sleeve Bearing

Applications.

V = 52.36 x 10-6 x D x CPM

V = 3.87 x 103 x (A/CPM)

A = 14.2 x 10-9 x D x CPM2

A = 0.27 x 10-3 x V x CPM

Summary

In General it is difficult to get any tworeadings to precisely agree with oneanother. Even when care is taken toinsure that transducers and locations arethe same and that the measurement typeis the same, agreement within +-30%depending on the instrument isconsidered good.

Even though overall values will notagree precisely spectrum Data or frequencies will be comparable withinthe limits of the bandwidth of thedifferent instruments.

Checklist

1. Is Transducer Type the same2. Bearing Type3. Is Measurement Type the Same

4. Engineering Units the same5. Frequency Response of

Transducer 6. Mounted Transducer Frequency

Response7. Where Readings Taken at the

same location8. Where Readings Taken in the

same Plane9. Instrument

Bearing Failure Modes

This application note presents specific failure modes associated withmachinery having journal bearings. Signal measurements presented here are



time waveforms that are collected with two orthogonally located EddyProbes, or proximity type sensors. Although certain faults can be analyzedwith spectrum analyzers, the majority of these faults can be diagnosed usingorbit analysis alone.

BALANCE

Diagnosis of a degrading balance condition is performed by concentratingon the synchronous amplitude which coincides withthe rotor speed. This can be accomplished by viewingthe spectra from any single Eddy Probe sensor. Asimilar diagnosis can be made by viewing the filteredsignals from two orthogonally mounted Eddy Probessensors as orbits. As the balance conditiondeteriorates the size, and sometimes the shape, of the orbit will grow larger until the peak-topeak amplitude exceeds acceptable limits.

CRACKED SHAFT

A crack in a rotor, or shaft, can generate several different effects on how themachine behaves: a change in the vibration level, a change in the operating phase angle, and/or a change in the resonance frequency as the machinestarts or stops. Spectral analysis can be used toidentify this fault, but observing filtered, synchronousorbits with the phase angle superimposed on the orbitallows rapid identification of this condition.

Changes in the filtered amplitude can be determinedusing orbits analysis. By superimposing the phaseangle input signal onto the orbit a shift in this parameter can be easilydetermined. By noting the operating speed at which the resonancefrequencies occur, a change in this frequency may indicate the "possibility"of a crack existing.

The "possibility" must be emphasized and carefully analyzed because manyother causes can produce these changes, such as, a damaged or loose bearing support, foundation problems, loose rotating parts...basicallyanything that can influence the "system" mass, damping, and/or stiffness.

LOOSE ROTATING PART

A loose rotating part can generate unusual vibration signals. They mayfluctuate in amplitude and the phase angle may shift, also. This fault isdiagnosed easiest using filtered, synchronous orbit analysis. Imagine amass, such as an impeller, which has come loose; it can rotate freely on theshaft independently.



As the loose part rotates it influences the balancecondition of the rotor which appears as a cyclicalincrease and decrease in the synchronous amplitude.This is observable using a spectrum analyzer, but thechanges may be too rapid for the sampling rate of the

instrument. An oscilloscope set up to observe afiltered orbit will sample continuously so that the

changes can be seen. The phase shifting can, also, be observed using anoscilloscope.

The inception of a loose part condition will produce a "nervous" filtered,synchronous orbit. The orbit will appear to vibrate slightly as this conditionis created; the part may be slipping and then sticking on the shaft just prior to becoming a full fledged loose rotating part.

OIL WHIRL

Oil whirl and oil whip are sometimes listed as a single machine fault, butcloser observation of the vibration signals and the machine conditionscausing these signals will produce different, distinct signal displays for eachcondition. This fault is caused by a condition which prevents the rotor fromcreating a stable oil wedge on which ride. An improperly designed bearingis the usual source for oil whirl conditions, but a change in the fluidviscosity or machine alignment state are other possibilities.

Generally, an oil whirl condition precedes an oil whip condition. Spectraland orbit analysis can be used to identify either condition. This

phenomenon creates an individual subsynchronous frequency which canoccur within a frequency range from 35% to 48% of rotor speed, dependingupon the machine/bearing design or construction. As the machineaccelerates the whirl frequency will increase asmachine speed increases.

Observing oil whirl as a filtered, synchronous orbit produces a distinctive display. The orbit will be moreor less round in shape with an amplitude that nearlyapproximates the bearing clearance, and when the phase angle is superimposed upon the display, the

orbit will appear to have two phase marks on it. This characteristic is due tofiltering at shaft speed and the fault being generated at a subsynchronousfrequency. The two phase marks will not be displayed symmetrically on theorbit because the whirl frequency is not at exactly ½ machine speed.

Acceleration Enveloping



Acceleration envelope measurementsare a new introduction to theinventory available to vibrationanalysts. Although initially it would

seem to be most useful for detectingrolling element bearing defects(BPFO, BPFI, BSF, FTF), much morecan be performed once the frequencyranges and applications of other moretraditional vibration parameters areconsidered. Once it is understood howacceleration envelope signals are processed, nearly all analysis could beaccomplished with these signals.Phenomena associated with motor

electrical and gear meshing problemsfall into the frequency range of acceleration envelope processing.Unbalance and other lower order problems do not initially fall intoenvelope analysis, but as these problems degrade they can bedetected.

Acceleration envelope signatures areessentially band passed signals where

lower order and higher order frequencies are removed. Theremaining frequency range is mostcommonly associated with bearingdefect frequencies. Normally, runningspeed related frequencies are removed by the filters and are not present inacceleration envelope spectra.However, when a running speed order frequency becomes severe enough itwill appear in the envelope signature.

This is due to the presence of numerous harmonics of the lower order frequency, which are notdetectable in velocity signals, fallinginto the envelope filter range.

Figures 1 and 2 illustrate that whenlower velocity signals are present the

Inadequate lubrication, which canencompass insufficient quantity,excessive quantity, and/or improper specifications, can be readily

identified using acceleration envelopemeasurements. This phenomenonmanifests itself as a shifting of thenoise floor. Properly installed andoperating bearings will not necessarilygenerate specific bearing defectfrequencies when an inadequatelubrication condition exists.

Figure 5

Figure 5 illustrates how a lubrication problem can be identified. The left portion of the spectra has beenelevated with overall amplitudes of

1.1 G's. Once the bearing was greasedthe elevated portion shifted back down, figure 6. Overall amplitudeswere reduced to 0.7 G's.

Figure 6

Sometimes the entire noise floor iselevated when a lubrication problemexist, as illustrated in figures 7 and 8.Figure 7 shows the elevated noisefloor. No identifiable bearing defect



running speed components will not be present in acceleration envelopesignatures. These spectra werecollected on the same day during aroutine PM data collection visit.

Figure 1

Acceleration envelope measurementswere taken on same day. Note that1Xrpm is missing in Figure 2 due tothe filtering used (30k-600k cpm).

Also, note the low noise floor and thelack of bearing frequencies in thespectrum. This bearing had recently been installed, no defects wereidentifiable, and the lubrication isadequate for continued operation.

Figure 2

As the running speed components andtheir harmonics become larger alongwith the presence of higher order harmonics the "ringing" associatedwith these harmonically relatedfrequency components fall into thefrequency range of the accelerationenvelope measurements. Runningspeed frequencies are now present inacceleration envelope measurements.Figures 3 and 4 illustrate this phenomenon.

Figure 3

frequencies are present, only anelevated noise floor. Overallamplitudes were at 8.2 G's.

Figure 7

Figure 8 was collected after the bearing was lubricated. Note that bearing defect frequencies are not present and the noise floor hasdropped. The overall amplitudedropped to 1.1 G's.

Figure 8

CONCLUSION

Acceleration Envelope measurements,due to the signal filtering, can be usedto identify more machinery defectsthan damaged rolling element bearings. This measurement parameter is commonly only used to identify thesmaller frequencies associated with bearing defects.

If machinery defects are classifiedinto lower order problems and higher order problems they can be effectivelydiagnosed with accelerationenveloping measurements. Lower order problems would be balance,misalignment, looseness due tofasteners, or looseness due to worn bearings. Higher order problemswould be bearing defect frequencies,motor bar pass frequencies, or



The velocity spectrum, figure 3, has amuch larger 1Xrpm frequencycomponent which now appears in theacceleration envelope spectrum, figure

4.

Figure 4

LUBRICATION

Lubrication problems, not normally

analyzed with velocity signatures, butcan sometimes be detected inacceleration signatures, can readily bedetected using acceleration envelopesignatures. Experience has shown thatan inadequate lubrication conditionwill cause a shift in the noise floor of the signature.

gearmesh frequencies.

The presence of lower order problemsin envelope spectra indicate a severe problem which most likely will need

immediate attention. The filtering willeliminate these frequencies from anotherwise healthy machine.

Tracking of higher order problems can be accomplished normally usingenvelope measurements

Piping Vibration

Piping vibration can be an annoying problemwhich can consume unnecessary maintenanceactivity and can affect pumping system performance and endurance. The systemincludes the pipe, all piping supports, hangers,snubbers, pipe to pipe interfaces, and machineryor devices attached to the pipe. All these itemscan influence the pipe vibration patterns.

This testing method will determine the pipingsystem vibration amplitudes, frequencies, nodal points, and the pipe modal shape. It can, also, beused to identify defective supports, incorrectly placed supports, and the locations of maximumdeflection requiring additional supports.

Analyzer/Data Collector

Corrective Actions

Generally, the pipe supports should be anodal point with little or no motion.Excessive motion at these locationsindicate that the support is faulty or improperly installed. Vibration amplitudesshould decrease as a complex joint, such asa tee connection, an elbow, or machine

connection, is approached.

Convert all the collected data todisplacement units using the formula:A = 14.2 x 10-9 x D x F²where:D = Displacement (mils pk-to-pk)A = Acceleration (G's pk)



Many data collectors have internal circuitry withlow frequency range limitations: output displaysin acceleration units are 2 Hz and outputdisplays in velocity and displacement are 5 Hz.

This circuitry is an internal high pass filter setfor a 2 Hz roll-off frequency for accelerationsignals and 5 Hz as the velocity anddisplacement roll-off frequency. The filter eliminates excessive noise from beingdisplayed.

This means that if an accelerometer is connectedto the data collector and the display is setup for acceleration units, the low frequency signals arecorrectly displayed down to 2 Hz. If the

accelerometer signal is integrated to velocity or displacement the low frequency limitation is 5Hz. Similarly, a velocity transducer has thevelocity low frequency limitation and theintegration limitation also applies.

Methodology

Piping vibration analysis involves describinghow much the pipe is moving and at whatfrequency the motion exists. The piping motion

can be further described by showing the motionas a modal plot. Pipes can vibration in threeorthogonal directions just like a machine.Vibration data should be collected in the X, Y,and Z axis. Since most data collectors do nothave the capability of calculating transfer functions collected from impact/response inputsignals, all data collection should be taken whilethe pumping system is

operating.

The vibration transducer may be attached to the

F = Frequency

(Hz).

Plot all the amplitude information which isat a common frequency on the graph todetermine the modal shape at which the pipe is vibrating. Compare the calculatedamplitudes and frequencies with theallowable piping vibration levels chart to

determine if corrective action is warranted.

Wachel, J. C. and Bates, C. L., Techniques for Controlling PipingVibration and Failures, ASME Paper 76-PET-18.

The listed ASME Paper includes a"severity chart" which could be used as astarting point in determining the piping

system acceptability. This chart wascompiled from 25 years of data and may beoverly conservative for long flexible pipingsystems commonly found in power stations.

Pipe vibration correction will involve re-tuning the pipe system to a different



pipe using a magnetic mount without affectingthe lower frequency response of the transducer.The overall pipe length should be separated intoequal spaced lengths 3-5 feet ( 1-2 meters ) for this test and plotted on a graph sheet. The pipe

hangers/restraints and their orientation to the pipe should be noted on the plot.

Setup the data collector for a frequency rangefor a 0-12,000 CPM (0-200 Hz) and displayunits of acceleration (G's). Collect spectra ateach measurement point. Evaluate the spectrafor the components at common frequenciesnoting their amplitude and frequency.

frequency. This may be accomplished byre-locating the pipe supports, installingdifferent supports, isolating the pipe fromits hangers or joints, or installing expansion joints in the pipe. Before any modification

is undertaken another pipe analysis should be carried out to determine that themodification does not violate other design parameters such as machine couplingmomentums or connection stresses.

Testing Checklist

1. Piping System Defined2. Proper Accelerometer 3. Graph Paper

4. Analyzer Set-up

Transient Analysis

Transient analysis is the study of how various signals change during amachine's process or load change.This analysis is particularly

important during a machine's startupor shutdown when critical speeds or natural resonances are encountered.Observing the overall amplitudes or individual spectra will not besufficient to document how theentire machine behaves duringtransient conditions. Data collectionmodes fall into two generalcategories: observing transducer output signals with respect to

machine speed and observingtransducer output signals withrespect to time. Sometimes, acombination of the two modes may be required.

BODE' PLOT

The center of the plot represents zerospeed and zero amplitude withsubsequent amplitude and phaseangle measurements plotted with

their associated machine speed.Phase angle measurements are plotted around the circumference of the chart, against machine rotationdirection. Polar plots are alwaysfiltered to machine running speed or some multiple of the machine speed,depending upon the fault beinginvestigated.

Machine critical speeds and natural

resonances are displayed as loops,with the critical speed situated 90from the start of the loop. Thischaracteristic makes identification of resonances and critical speeds easy.Other information available from the plot are: slow speed runoutassociated with NCPUs, system



A Bode' , pronounced bow-day, plotis a display of three variables:

signalamplitude, machine speed, and phaseangle. The three variables are usuallysynchronized to the machineoperating speed, but occasionallymay be synchronized to somemultiple of machine speeddepending upon the type of machinefault being investigated.

The Bode' plot is essentially two

displays: phase angle versus machinespeed and amplitude versus machinespeed. Analysis of the plot candetermine at what speed(s) a criticalspeed exists, the amount of runoutassociated with a Non-ContactPickUp (NCPU), the balancecondition, system damping, and theoperating phase angle and amplitudeat various machine speed.

POLAR PLOT

The polar, or nyquist, plot is also a plot containing the same three

variables as aBode' plot.The variablesare plotted ona singlecircular chartinstead of two

separate plots.

damping, balance condition, andoperating phase angle and amplitudeat various machine speeds.

CASCADE PLOT

Cascade or Waterfall plots, alsohave threedisplayedvariables:amplitude,

frequency, and machine speed. Thisdisplay differs from the other transient plots because it is

essentially a collection of spectra.Each spectrum is associated with aspecific machine operating speed.Setting the maximum frequency(Fmax) and the spectrum resolutionare important for accurate diagnosis.See the Application Note, DiagnosticTechniques-Part 2, FrequencyDomain for further discussion of these settings.

Transient Analysis Checklist

1. Bode' Plot2. Polar Plot3. Cascade Plot

Transducer Conventions



This application note provides a set of conventions to provide order andmeaning to how transducers relate totheir mounting and the machine.

MOTION TOWARDS

Vibration transducers measure allvibration they sense regardless of thedirection of the mechanical vibration.The displayed output of a transducer will resemble a sinusoidal wave pattern, with some portion of the signalabove (or on the positive side) the zeroreference and some portion below ( or on the negative side) the zero

reference.

A common convention employed inthe design and construction of alltransducers is that motion towards thetransducer will cause a positive signaloutput. This convention allows theanalyst to determine how the rotor or bearing case is moving in relationshipto the sensor. This information is vitalwhen determining phase relationships

between various sensors and betweenthe rotor and its case.

BEARING NUMBER

Distinguishing one bearing fromanother is an important considerationwhen installing or analyzing several bearings on a machine train. Acommon convention is to startnumbering the bearings starting from

A dual channel oscilloscope allowsviewing the output signals fromtwo different sensorssimultaneously on a cathode raytube (CRT). When the oscilloscope

is in the XY, or cartesian mode, thecenter of the screen represents zerooutput from both sensors. Thisoperational mode is sometimescalled orbit mode. As the signalfrom the horizontal sensor increases (in the positive direction)the dot, or trace on the displaymoves to the right. An increasingsignal in the vertical plane causesthe trace to move up on the

display. Thus, following standardcartesian notation, the positive X(horizontal) input is to the left of the positive Y (vertical) input.

Twosensorsmust beinstalled90 fromeach

other tomeasure

the entire amount of vibration intwo planes (vertical andhorizontal).

This requirement stems from thefact that each sensor is designed tomeasure vibration only along thesensor's axis. If the sensors are notinstalled at a 90 orientation, one of

the sensors will be measuring someof the vibration which is beingmeasured by the other, which canonly confuse any analysis.



the driver end. The choice of whether to use numbers or letters is immaterial,the sequence should increase asadditional bearings are added to thediagram.

Machine trains with gear increasersmay cause some confusion when thegear is considered. When numbering agear, continue numbering the bearingsalong the gear shaft instead of aroundthe gear casing.

X & Y

When two transducers are installed in

a single plane, such as two EddyCurrent sensors on one bearing, theyhave a relationship to each other whichmust have a standard specification.Defining their orientation andrelationship to each other requiresunderstanding how an oscilloscopeoperates. The oscilloscope was thefirst, and still is, a basic analysisinstrument.

Todistinguishonesensor

fromtheother they

are labeled X and Y.

Determining which is X and whichis Y requires using the bearingnumber convention discussedearlier (viewing the sensors fromthe driver end) and utilizing

standard oscilloscope orientation.Within the 90 orientation the Xsensor is located to the left of theY sensor when the sensors aremounted above the horizontal splitline of the bearing. Should thesensors be oriented at a differentlocation on the bearing, simplyimagine the sensors located abovethe horizontal and follow the statedorientation convention. This

convention applies regardless of shaft rotation direction.

Transducer Conventions Checklist

1. Motion Towards2. Bearing Number 3. X & Y

Frequency Domain

Frequency domain measurementsand analysis have becomeincreasingly popular to diagnose a particular machine fault. Thismeasurement mode relies on

A band is essentially a band passfilter allowing only the frequencieswith the selected range to bemeasured. All other frequencies areexcluded from analysis. Many



processing the transducer outputsignal using Fast Fourier Transform(FFT) algorithms to display thesignal amplitudes as a function of frequency. FFT processing

essentially separates complex signalsinto individual components having asingle frequency content. This typeof display is commonly termed aspectrum.

An enhancement of spectral analysisis to define specific frequency rangesto perform band analysis.Conceptually, band analysis issimilar to filtering a signal. The

"filter" searches for frequencies onlywithin its frequency range. Certain permanently installed machinemonitoring systems offer thiscapability. This feature is quiteeffective, once the particular spectralrange and resolution has beendetermined, to rapidly diagnosemachine faults.

SPECTRUM

A spectrum display is a display of signal amplitudes on the vertical axisand the signal frequencies on thehorizontal axis. The frequency axis

units may be in hertz (hz) or incycles per minute (CPM). Hertz or cycles per second may be convertedinto CPM by multiplying by 60. For example: 10 hz = 600 CPM. Thehorizontal axis is scaled from 0 tosome maximum frequency (Fmax).Individual signal frequencies will

modern machine monitor systemsare capable of monitoring specificfrequency ranges using bandanalysis.

RESOLUTION

Frequency resolution is an arearequiring considerable attention. If the resolution is inadequate theentire analysis process could bemeaningless or incorrect. Someinstrument specifications list thespectral resolution as lines. A highresolution would be 3200 lines per spectrum and a low resolution would

be 100 lines per spectrum.

Each lineof resolutioncan beviewed asa bucket or pail of aspecificsize. The

signal frequencies can be viewed asa tennis ball. If a tennis ball'sfrequency matches the frequencyrange of the bucket, it is placed inthe bucket. As the bucket fills withtennis balls the peaks on thespectrum display rise. Should thefrequency range of the buckets betoo large, the tennis balls will not beadequately separated to detectindividual frequencies. This would

lead to always using the highestresolution for spectral or bandanalysis. Higher resolutions requiregreater amounts of time to displaythe spectrum, thus a balance must bereached between the capture timeand the spectral resolution.



appear as peaks or spikes, eachhaving a specific amplitude. Properlysetting the Fmax will ensure that allof the input signal is being analyzed.This setting can be verified

mathematically by summing thesquare of each amplitude peak. Thesquare root of this summation shouldapproximate the overall amplitudelevel obtained directly from thetransducer's output.

BAND ANALYSIS

Bandanalysis

involvesselectingfrequencyranges of interest toallow

rapid determination of a machine'scondition. Generally, each machinefault will generate a specific, uniquefrequency as the conditiondeteriorates.

Frequency Domain Checklist

1. Overall Amplitude2. Time Base Waveform3. Orbits

Turbine Supervisory Instrumentation

Specification of a TurbineSupervisory Instrumentation (TSI)system can be an exhausting processwhen the individual parameters must be specified. This application note issupplied to provide a guide to be

used in selecting an appropriate TSIsystem. TSI systems not onlymeasure bearing vibration levels, butcan include shell expansion,differential expansion, valve position, turbine speed andacceleration, thrust position, phaseangle, and bearing temperatures.



When an existing TSI system is being retro-fitted the immediateindication is that a one-for-onereplacement of each original

parameter is sufficient. Thisapproach may be adequate if theoriginal system was a complete package.

Recent experience with retro-fittingTSI systems has brought to light thatmany of the existing systems could be enhanced with additional parameters. Also, certain parametersshould be considered for complete

replacement with a different typesensor.

General

The information required under thistopic will define and describe theturbine generator along with whowill perform and/or supply thevarious tasks and parts of the TSIinstallation. The time frame for the

system installation should getconsideration at the point.

Describing the turbine generator involves listing the number of bearings, type of bearings,turbine/generator manufacturer, thenumber and function of each rotor segment, etc. This information may be obtained from the OEM operationand maintenance manuals and is

required whether a retro-fit or anentirely new installation is beingspecified.

Documentation of the proposed TSIshould include who supplies theindividual components and service of the new system, along with the



number of operation and servicemanuals and/or drawings required.

For more information aboutinstallation services see the STI

Application Note, Field Service, FS.STI Application Note, Field WiringInstallation, FWI covers many topicsof particular concern prior to andduring the electrical systeminstallation.

Monitor

Selecting the monitor follows the process of detailing the turbine

generator layout. The monitor selection generally involves decidingwhat the monitor should do and howthe user will interface with it.

The monitor can be specified to be astand-alone output with user interface or to interface with ananother existing output device suchas PLC or DCS.

Radial Vibration

Radial vibration is usually the heartof the TSI system. It gets the mostattention and generally gives the firstindication of out of specificationconditions. Most OEM TSI systemsutilized a shaft rider transducer system to monitor vibration with ashaft absolute output signal. Anexact replacement transducer system

can be supplied, but most customersand OEMs are specifying a EddyProbe Systems. A complete vibrationsystem would install two sensor systems per bearing with the sensorslocated 90 from each other.

For more information about Eddy



Probe Vibration Sensors and their application see the STI Application Note, Eddy Probe Transducer Installation, Part 1-Radial Vibration.

Thrust Position

Thrust position indication includes oneor two Eddy Probe Systems to observethe position of the thrust collar withinits bearings. This system is an internalinstallation and need not replace theexisting system because many originalinstallations utilize a differential pressure system that interfaces with theturbine hydraulic control system.

For more information about thrust position sensors and their applicationsee the STI Application Note, EddyProbe Transducer Installation, Part 2-Thrust Position.

Shell Expansion

Shell expansion is the measure of aturbine case or shell moves in relation

to a fixed location usually measuredwith a Linear Variable DifferentialTransformer (LVDT). Some existingOEM systems still use spindlemicrometers or dial indicators that aresubject to mechanical damage andhuman error. Although many systemsinstalled with only one LVDT areadequate, a complete TSI systemspecification should consider twoLVDTs located at each corner of the

turbine shell. A second sensor willmonitor shell cocking or uneventhermal growth which is a fairlycommon occurrence during startupwhen the sliding feet may haveinadequate lubrication.

For more information about shell



expansion systems and applicationssee the STI Application Note, ShellExpansion, TSI Part-4.

Differential Expansion

Differential expansion measurementsare an important parameter receivingmuch attention during turbine startupand warming. This parameter measures how the turbine rotor expands in relation to the turbineshell, or casing.

A new differential expansion systemusing Eddy Probes can be retro-fitted

to any existing system. A EddyProbe is more reliable and robustthan OEM supplied induction coilsystems.

For more information aboutdifferential expansion systems andapplications see the STI Application Note, Differential Expansion, TSIPart-3. Valve Position

Correct valve positioning is requiredto efficiently operate a steam turbine.Some turbines may require severalthrottle valves be monitored andsome turbines will instrument themain stop valve(s) to determinewhen they crack from their seats.

Valve Position

Correct valve positioning is requiredto efficiently operate a steam turbine.

Some turbines may require severalthrottle valves be monitored andsome turbines will instrument themain stop valve(s) to determinewhen they crack from their seats.

Retro-fit valve positionmeasurements use DC LVDTs or DC



Rotary Potentiometers. All OEM TSIsystems include valve positionmeasurement(s) as a startup andoperation parameter. Some OEMsystems utilized AC LVDTs while

others use mechanical linkages andscales for indication.

A retro-fitted system can be installedin the same position or at relocated toa more accessible or protected position.

For more information about valve position systems and applications seeSTI Application Note, Valve

Position,TSI Part-2.

Eccentricity

A rotor which has been sitting idleduring overhaul or has beeninadvertently stopped duringcoastdown for an extended periodwill develop a bow or bend. Thiscondition must be corrected byturning gear operation and, possibly,

with auxiliary heating prior to highspeed operation to prevent internalclearance rubbing.

Eccentricity systems installed byOEMs monitor the turbine stub shaftor a shaft collar using inductioncoils. A retro-fit Eddy Probe systemwill monitor the same location andmany times use the same bracketry.

For more information abouteccentricity systems and applicationssee STI Application Note,Eccentricity, TSI Part-1.

Speed

Turbine speed indication supplied by



OEMs come in many forms: observinga gear wheel located inside the frontstandard, electrically converting thegenerator output frequency, or monitoring the turning gear. A retro-

fitted system using Eddy Probe's can be specified to observe any multi-toothed gear wheel. Applicationsmonitoring generator output frequencywithout an integral turning gear mayrequire installation of a custom gear wheel.

Speed indication may be specified asan analog display or as a digitaldisplay and can be interfaced to a zero

speed system for turning gear engagement.

Rate of Acceleration

The rate of acceleration parameter isusually monitored during startup to prevent over-torquing the rotors, as theturbine approaches critical speeds, andas the operating speed is reached prior to line synchronization. Once the

generator has been synchronized and is being controlled by load dispatchersthe acceleration rate is not monitored.

Acceleration rate measurements use aspeed input to derive its output display.Eddy Probe systems can be installed asa replacement or supplement anexisting application. See STIApplication Note, Eddy ProbeTransducer Installation, Part-1 Radial

Vibration for relevant informationabout this type of sensor.

Phase

Phase, or phase angle, is a measure of the relationship of how one vibrationsignal relates to another vibration



signal and is commonly used tocalculate the placement of a balanceweight. This parameter is not usuallydisplayed continuously but ismonitored periodically to determine

changes in the rotor balancecondition, deviations in systemstiffness such as a cracked shaft.

Phase angle measurements aresometimes not supplied by OEMs, but can be installed using a EddyProbe system. Installation involveslocating or installing a once-per-turnevent such as a key or notch that theEddy Probe will view. An Eddy

Probe viewing a notch is easier toinstall and adjust, but the installationof the notch requires special toolingto cut the notch. Keys are easier toapply using glues or epoxies and aresubject to coming off due tocentrifugal forces.

Temperature

Bearing temperature is a measure of

the how hot a bearing is operating. Itmay be due to overloading, mis-alignment, improper lubricant pressure and/or flow.

Nearly all turbine generator bearingswere originally installed or retro-fitted with bearing temperaturesensors. These sensors may bethermocouples or RTDs. This parameter is often overlooked

possibly due to the OEM outputdisplay located at some other panelnot within the vicinity of the retro-fitted TSI system. Any bearings thatwas not originally equipped withtemperature sensors can be retro-fitted to accept thermocouples or



RTDs.

Custom Cabinet

Congested control boards may

preclude installing the TSI rack requiring a stand-alone cabinet. Thiscabinet can house auxiliaryequipment associated with the newTSI system, such as power supplies,termination strips, external relays,etc.

The cabinet can be configured tomany differing designs dependingupon the user's requirements.

Cabinets should be sturdy enoughwithstand environmental conditions,such as moisture content, explosiveatmospheres, temperature,

etc. Eccentricity TSI

Eccentricity TSI

Shaft Eccentricity plays a veryimportant role as part of a TurbineSupervisory Instrumentation (TSI)System on large steam turbines andshould be included in retro-fit planswhen at all possible.

Operators use eccentricitymeasurements to determine when acombination of slow roll and heatinghave reduced the rotor eccentricity to

the pointwheretheturbine

can safely be brought up to speedwithout damage from excessive

Mechanical Runout Eddy Current transducers are alsosensitive to the shaft smoothness for Eccentricity. A smooth (64 micro-inch) area approximately 3 times thediameter of the probe tip must be provided for a viewing area.

Electrical Runout Since Eddy Probes are sensitive tothe permeability and resistivity of

the target material and the field of the transducer extends into thesurface area of the shaft byapproximately 15 mils (0.015"), caremust be taken to avoid nonhomogeneous viewing area materialssuch as Chrome.



vibration or rotor to stator contact.

Eccentricity is the measurement of Rotor Bow at rotor slow roll whichmay be caused by any or a

combination of

1. Fixed mechanical bow2. Temporary thermal bow3. Gravity bow

In extreme cases of thermal/gravity bow, caused by a sudden trip of theunit and failure of the turning gear toengage, the rotor may be positionedand stopped 180o out of phase (bow

up) to allow gravity to work entirelyon the bow and substantially shortenthe time required to reduce the bow.

Eccentricity is measured while theturbine is on slow roll (1 to 240 RPM below the speed at which the rotor becomes dynamic and rises in the bearing on the oil wedge) andrequires special circuitry to detect the peak- to-peak motion of the shaft.

This is accomplished using circuitrywith long update times selectable between 20 seconds (> 3 RPM) and 2minutes (<3 RPM).

As the eccentricity measurement isnot required after a turbine is broughtto speed and under load provisionsare made to lock the measurement tozero. This can be accomplishedwithout external contacts through the

use of a speed measurement channelwith underspeed or overspeedalarms.

As it is impractical to mount EddyProbe Transducers (Non-ContactingPickups) midspan on the rotor wherethe eccentricity measurement would

Another form of electrical runoutcan be caused by small magneticfields such as those left by Magna-fluxing without proper degaussing.

Perpendicular to shaft centerline Care must be exercised in allinstallations to insure that the EddyProbe is mounted perpendicular tothe shaft center-line. Deviation bymore than 1-2 degrees will effect theoutput sensitivity of the Probe.


clearances The RF Field emitted from the probe

tip of the transducer isapproximately a 45 conical shape.Clearance must be provided on allsides of the Probe tip to preventinterference of the RF Field. Care

must also betaken toavoidcollars or shoulderson the shaft

that maythermally "grow" out from under theProbe tip as the shaft expands.

Eddy Probe tip to tip clearances Although Eddy Probe tip to tipclearances are not normally an issueon most machines, it should be notedthat the probes radiate an RF Fieldlarger than the probe tip itself. As anexample, SKF-CM CMSS65 and 68

Eddy Probes should never beinstalled with less than one (1) inchof Probe tip to tip clearance. Larger probes require more clearance.Failure to follow this rule will allowthe driver to create a "beat"frequency which will be the sum anddifference of the two driver RF



be the highest the transducer(s) aremounted outside the pressure case asfar from the bearing (Node Point) as practical.

The bearing should be avoided as amounting location because duringslow roll operation the rotor isturning in the bottom of the journal bearing and is not dynamic while theeccentricity measurements are beingmade. This effect forces the bearingsto become nodal points.

Assuming

uniformstiffnessand

weight, the rotor mid- spaneccentricity may be expressed as theratio of the transducer span from the bearing over the transducer measuredeccentricity to 1/2 the bearing spanover the midspan eccentricity or calculated using the followingformula, (Tecc x ½Bspan)/Tspan =

MSecc.Where Tecc = Transducer measuredeccentricityBspan= Bearing SpanTspan= Transducer span from bearingMSecc= Midspan eccentricity

OEM's (Original EquipmentManufacturers) should be consulted

for actualcalculations.

Turbineowners whoareretrofittingexistingeccentricitysystems

frequencies.

System Cable Length and

Junction Boxes Eddy Probe Systems are a "tuned"

length, and several system lengthsare available. System length ismeasured from the probe tip to theOscillator/Demodulator, and ismeasured electrically which can beslightly different than the physicallength. For example, the Model 403is available in 9, 20, and 30 footsystem lengths. Care must be takento insure that the proper systemlength is ordered to reach the

required Junction Box.

Grounding and Noise Electrical noise is a very seriousconsideration when installing anyvibration transducer, and specialcare needs to be taken to preventunnecessary amounts of noise. Asmost plant electrical noise is at 60HZ, and many machine runningspeeds are also 60 HZ, it is difficult

to separate noise from actualvibration signal. Therefore, noisemust be kept to an absoluteminimum.

Instrument Wire A 3-wire twisted shielded instrumentwire (ie; Belden #8770) is used toconnect eachOscillator/Demodulator to the SignalConditioner Card in the Monitor.

Where possible, a single run of wirefrom the Oscillator/Demodulator (Junction Box) to the Monitor location should be used. Splicesshould be avoided.

The gauge of the selected wiredepends on the length of the



supplied by the OEM or others willmount the eccentricity transducer atthe same location as the originalinstallation. In many cases onlyminor modifications to the existing

bracket are required. Using the samelocation has several advantages andsimplifies installation.

1. OEM's original installation asa rule included an eccentricitycollar or other good target for an Eddy Probe System.

2. Eddy Probe eccentricitymeasurements will agreeclosely with the original

OEM supplied system as themeasurements will be takenat the same location.

3. Operators will need lesstraining on how to interpretthe new systemsmeasurements as they will be basically the same.

4. Eccentricity historical datawill be valid.

5. Existing brackets may be

modified.6. Case or standard penetration

for cable may be reused withminor mod

ification.

Eccentricity is normally measuredP/P (Peak to Peak) to agree with previously established conventions.The actual excursion from shaftcenterline caused by bow would beone half that measurement or the 0/P

instrument wire run, and should beas follows to prevent loss of highfrequency signals:





Red -24 VDC Power

Black Common

White Signal

Common Point Grounding To prevent Ground Loops fromcreating system noise, systemcommon, ground and instrumentwire shield must be connected toground at one location only. In mostcases, the recommendation is toconnect commons, grounds andshields at the Monitor location. Thismeans that all commons, groundsand shields must be floated (notconnected) at the machine.

Occasionally due to installationmethods instrument wire shields areconnected to ground at the machinecase and not at the monitor. In thiscase, all of the instrument wireshields must be floated (notconnected) at the monitor.

Conduit Dedicated conduit should be provided in all installations for bothmechanical and noise protection.Flexible metal conduit should beused from the Eddy Probe to theOscillator /Demodulator junction box, and rigid bonded metal conduit



(Zero to Peak) measurement. TheTurbine Supervisory Instrumentationmay be calibrated in either fashion tosuite the users requirements.

Theory of Operation

Eddy CurrentTransducerswork on the proximitytheory of operation. A

system consists of a matchedcomponent system: a Probe, anExtension Cable and an Oscillator

/Demodulator (driver). A highfrequency RF signal @2 mHZ isgenerated by theOscillator/Demodulator, sent throughthe extension cable and radiated fromthe Probe tip. Eddy currents aregenerated in the surface of the shaft.The driver demodulates the signaland provides a modulated DCVoltage where the DC portion isdirectly proportional to gap

(distance) and the AC portion isdirectly proportional to vibration. Inthis way, an Eddy CurrentTransducer can be used for bothRadial Vibration and distancemeasurements such as ThrustPosition and Shaft Position.


Mounting Orientation

All vibration transducers measuremotion in their mounted plane. Inother words, motion either directlyaway from or towards the mountedEddy Probe will be measured aseccentricity.

For eccentricity measurements it is

from the junction box to the monitor.

Calibration All Eddy Probe systems (Probe,Cable and Oscillator Demodulator)

should be calibrated prior to beinginstalled. This can be done by usinga SKF-CM P/N CMSS601 StaticCalibrator, -24 VDC Power Supplyand a Digital Volt Meter. The EddyProbe is installed in the tester withthe target set against the Eddy Probetip. The spindle micrometer withtarget attached is then rotated awayfrom the Eddy Probe in 0.005" or 5mil increments. The voltage reading

is recorded and graphed at eachincrement. The SKF-CM CMSS65and 68 systems will produce avoltage change of 1.0 VDC ±0.05VDC for each 5 mils of gap changewhile the target is within the NCPU's linear range.

Gap When installed, Eddy Probes must be gapped properly. In most

Eccentricity applications, gappingthe transducer to the center of thelinear range is adequate. For theModel 403 transducer gap should beset for -12.0 VDC using a DigitalVolt Meter (DVM), this correspondsto an approximate mechanical gap of 0.060" or 60 mils. The voltagemethod of gapping the Eddy Probe isrecommended over mechanicalgapping because it is more accurate

and easier to accomplish. In allcases, final Eddy Probe gap voltageshould be documented and kept in asafe place.

Eccentricity Installation Checklist

1. Machine Slow Roll Speed



recommended that the transducer bemounted vertically. As mosteccentricity sensors are internallymounted and are not visible from theoutside of the machine whatever the

angle of orientation is finally chosenit is very important that the mountinglocation be documented for futurereference.

Linear Range Several versions of Eddy ProbeTransducers are available with avariety of Linear Ranges and bodystyles. In most cases, a sensor with alinear range of 90 mils (0.090") is

more than adequate for Eccentricitymeasurements.







Target

Material EddyCurrenttransducersarecalibrated at

the factory for 4140 Steel unlessspecified otherwise. As Eddy Probes

are sensitive to the permeability andresistivity of the shaft material, anyshaft material other than 4000 seriessteels must be specified at the time of order. In cases of exotic shaftmaterial a sample may need to besupplied to the factory.

2. Transducer OrientationDocumented

3. Target Material, 4140 Other 4. Smooth Target Area5. Size of Target Area

6. Junction Box Location(s)7. Metal Conduit (Junction Box


Box to Probe)9. Correct Instrument Wire10. Shielding Convention,

Monitor or Machine11. Calibration12. Gap Set



Differential Expansion TSI

Differential Expansion (DE) and/or Rotor Expansion (RE) are veryimportant measurements usuallysupplied by the Original Equipment

Manufactur e (OEM) as part of a

Turbine SupervisoryInstrumentation System (TSI) onlarge steam turbine generator applications. Although theapplication of DifferentialExpansion and Rotor Expansionhave much in common, it isimportant to understand thedifference between the twomeasurements.

Differential Expansion on a turbineis the relative measurement of therotor's axial thermal growth withrespect to the case.

Rotor Expansion on a turbine is theabsolute measurement of the rotor'saxial thermal growth with respect tothe turbine's foundation.

A typical large steam turbine power generation unit will have a thick case, on the order of 12" to 18". Dueto the mass of this case, it willexpand and contract at a slower ratethan the relatively thin (hollow)rotor. During turbine startup, extra

care must be used to ensure that thecase has been properly heated andexpanded sufficiently to preventcontact between the rotor and thecase. Several DE or a combinationof DE and RE measurements may be employed on a single large

The signal sensor generates a highfrequency oscillating RF signal that issent through the extension cable to the pickup tip. The pickup tip, having awound coil of fine wire, radiates aelectromagnetic field. As the radiatedfield is bisected by the rotor surface,eddy currents are created on the rotor surface. As the rotor surface movescloser to the pickup tip, a greater amount of eddy currents are created proportional to the gap between thesurface and the pickup tip. The signalsensor contains a demodulator whichmeasures the increase in eddycurrents, and generates an equivalentDC voltage proportional to the gap.

Extra care must be exercised to ensurethat the pickup axis is perpendicular tothe viewed surface. An error of ±1ocan cause a significant error in themeasurement.

Calibration In all cases, it is extremely importantto complete a calibration curve or graph prior to installation. This graphallows the engineer to document andgraphically visualize how the requiredmechanical range and alarm setpointsfit the Probe's system's operatingrange.

First, prepare an Eddy Probe system

calibration curve using the suppliedcalibration data, or using actualempirical calibration data derived inthe field. Then overlay the requiredfull scale range and alarm setpointsfor the machine on the calibrationcurve. The two curves should be



turbine generator.

For many years, Power GenerationUtilities and other owners of largesteam turbines have been retrofitting

the original OEM supplied TSIsystems with modern TSI systemswhich are more accurate, and easier and less costly to maintain.

Eddy Probe Transducer Systemshave been very successful replacingthe older single or dual coilcapacitive and inductance typetransducer systems. Eddy ProbeTransducer systems are available

with up to 1000 mils (1") of linear range. When properly applied, thesetransducers can replace any existingDE or RE transducer system eventhose with ranges over severalinches.

Existing OEM Applications

In general, modernizing or replacingDE and RE transducers and monitor

systems is reserved for thosemachines that were originallyequipped with these measurements by the OEM. Usually, the user of the system wants to replace theexisting DE or RE system with amodern system. The new systemhas exactly the same range andalarm setpoints as the old system.This insures that the user followsthe original OEM operating

instructions and simplifies operator training and understanding.

Large steam turbine rotors wereoriginally manufactured with two(2) types of DE and REmeasurement surfaces:

centered so that there is an equalamount of unused linear range at eachend of the Probe curve.

With this graph completed, gapping

the Eddy Probe transducers is simplya matter of knowing the mechanical position of the rotor, finding thematching Probe DC Gap on the graphand gapping the Eddy Probetransducers appropriately.

Pickup Calibration - Single DE with

Collar Once the required DE range has beenestablished, the operating range of the

pickup must not be exceeded by theDE measurement. The pickups should be calibrated, or adjusted, when theturbine is in a known thermalcondition. This condition is almostalways when the turbine is in a coldstate. The OEM documentation maylist the cold setpoint as the "GreenMark" condition with correspondingDE measurements that theinstrumentation system should

indicate.

Next, the direction of thermal growthmust be determined. The indicationsmay be termed "rotor long" and "rotor short". The OEM may list the twoextreme alarm setpoints as "RedMark".

For example, consider a turbine withthe thrust bearing at the front

standard, the DE collar located between the LP turbine and thegenerator, and the single DE pickuplocated on the governor side of thecollar. The "rotor long" indication will be when the relative gap measurement between the collar and the pickup hasincreased. The "rotor "short"



Perpendicular Collar Many turbines have one or moreintegral collars machined on the

rotor. Thesecollars will

usually havesmoothlymachinedfaces of 2"to 3" and beseveralinches thick.

The existingOEM

transducer system will usually

employ a custom bracket with dual(complimentary) coils one on eachside of the collar.

Rotor Ramp Some turbines will have machined

conicalsurfaces or ramps onthe turbineshaft at one

or morelocations.

Ramps increase the operating rangeof the measurement transducer bythe sine of the ramp angle allowinglarger measurements or the use of shorter range transducers.

The existing OEM transducer system will usually employ acustom bracket with dual

(complimentary) coils mounted perpendicular to the ramp surface.

Ramp angles can usually bedetermined by referring to theoriginal OEM Turbine Manuals, or they can be measured if necessary.Common Ramp angles are 7o, 11o,

indication will be when the gap hasdecreased.

The pickup should be chosen so thatthe linear range of the pickup exceeds

the expected DE range including thealarm setpoints. The cold setpoint for the pickup adjustment should allowthe center of the DE range tocorrespond with the center of the pickup's operating range. A simplecalculation will provide the required pickup gap for the cold setpoint.

Pickup Calibration - Single DE with

Ramp

The same requirements as listed in thesection above apply for a single DE pickup system with a ramp target.

Additional requirements are that theramp angle must be included in thetransducer calibration curve with thegap and alarm setpoint calculations.If the ramp surface is small, the pickupdiameter sizing must be aconsideration. Also, the pickup

orientation with respect to the rampmust be considered to insure that theramp will not "grow" out from under the pickup at either alarm extreme.

Consider the above example, but thetarget is a collar with a 45o rampmachined on the outer edge. The DE pickup is located on the governor sideof the collar. When the turbine is inthe cold state, the pickup to ramp

orientation should have the pickuplocated near the bottom of the ramp.As the indication approaches "rotor long", the pickup will be oriented near the top of the ramp.

Pickup gap voltages (VDC) would becalculated to include the 14 ramp



and 14o as examples.

Occasionally a ramp may bemachined on a collar. This rampwill usually have a much larger

angle with a smaller viewingsurface (ie: 45o).

Retro-fit Applications

In almost all cases, retrofitting theexisting DE or RE transducers withEddy Probe's is simply a matter of determining the required operatingfull range with alarm setpoints,whether its a collar or ramp

application, angle of the ramp, andthe shaft viewing or target area.

Whenever the expected rangeexceeds a single transducer range acomplimentary system is required.A complimentary system utilizestwo (2) Eddy Probe transducersviewing the opposing faces of thecollar or ramp. The complimentarysystem extends the operating full

range of the system. This systemoperates such that as the collar or ramp moves out of the operatingrange of one transducer it movesinto the operating range of thesecond transducer.

As Eddy Probe systems operate onthe proximity theory of operation,they are not effected by oil or other non- conductive material that may

come between the target area andthe transducer. As the Eddy Probetip radiates a high frequency RFsignal roughly in a 45o conicalshape from the probe tip the targetarea must be three (3) times the tipdiameter. Care must be taken toinsure that the Eddy Probe

angle as follows:Cold Gap Volts = Cold Gap x sin(14 )Rotor Long Volts = Long Gap xsin(14 )Rotor Short Volts = Short Gap x

sin(14 )

Pickup Calibration - Dual DE with

Collar Dual DE pickup systems viewing acollar will have two separate Probesystems that view either side of theDE collar. This application is acomplimentary system where thecollar should grow out of the range of one pickup and into the range of the

other pickup.

One pickup will be installed on thegovernor side of the collar and theother installed on the generator side of the collar. Generally the pickups will be installed so that when the collar iscentered in its operating range, both pickups should be indicating the samegap voltage reading.

Cold pickup adjustment involvesdetermining the equivalent gapvoltage for the pickup on the governor side of the collar. This pickup may begapped electrically.

The other pickup may be out of itsoperating range when the turbine is inthe cold condition. This means the pickup on the generator side of thecollar will have to be gapped

mechanically.

Mechanically gapping the generator pickup involves figuring thedifference between the cold gap of thegovernor side pickup and theequivalent gap of the governor side pickup when the collar is centered,



transducer is mounted perpendicular to within a few degrees of the targetor viewing area. Excessivemounting error will change thecalibration factor of the installation.

Perpendicular Collar Perpendicular 90o Collars can bemeasured by either a single EddyProbe transducer or bycomplimentary transducers mountedon either side of the collar. The sizeof the collar face available for theProbe viewing surface and therequired measurement range willdetermine which method is best

suited for the application.

The following operating ranges areavailable. The operating full rangefor single transducer applicationsviewing a perpendicular collar is thesame as the transducer's linear range.

Single Complimentary

CMSS65/68

90

mils 180 mils

CMSS62240mils

480 mils

CMCP-1000

1000mils

2000 mils

Installation of the required EddyProbe transducer(s) can be made byeither modifying the existing DE or RE bracket, or by fabricating a new

bracket.

Rotor Ramp A Rotor Ramp installation can also be measured using either a singleEddy Probe transducer or bycomplimentary transducersconfiguration. The size of the Probe

and adding this value to the equivalentgap of the generator side pickup whenthe collar is centered. This calculationwill result in an equivalent mechanicalgap at which the generator side pickup

is to be gapped.

Pickup Calibration - Dual DE with

ramp Dual DE pickup systems follow thesame requirements discussed abovefor collar applications, but the gapvoltage or

mechanical gap values willincorporate the ramp angle. Certain

turbines may have the rampsmachined so that the pickups, wheninstalled, are facing away from eachother. Other turbines will have theramps machined so that the pickupsface towards each other. Regardless of the orientation, the pickup calibration procedure is the same.

For this application, one the governor side pickup will be calibrated

(adjusted or gapped) electrically,while the other will be calibratedmechanically while the turbine is inthe cold condition.

This application will have each pickupindicating the same gap voltage whenthe both ramps are centered on the



viewing surface and measurementrange required will determine whichmethod best suites the application.

The operation range of the

transducer is calculated by dividingits published linear range by thesine of angle of the ramp. Doublethe single transducer range for complimentary transducers.

For example, the followingoperating ranges are available usingthe following extended range EddyProbe systems viewing a 14o Ramp.

Single Complimentary

CMSS65/68372mils

744 mils

CMSS62992mils

1984 mils

For example, for a 500 milmeasurement range with a 14oramp, the complimentary P/NCMSS68 system should be used.

This will maximize resolution of themonitoring system, and yield themost economical installation.

Installation of the required EddyProbe transducer(s) can be made byeither modifying the existing DE or RE bracket, or by fabricating a new bracket.

Installation

DE and RE Eddy Probe systemsoperate on the proximity theory. AEddy Probe system consists of amatched component system: a pickup, an extension cable, and asignal sensor.

pickups. The governor side pickupcold setpoint is calculated for a 11ramp as follows:Cold Gap Volts = Cold Gap x sin(11 )

Calculating the required mechanicalgap for the generator pickup involvesfiguring the difference between thecold gap of the governor side pickupand the equivalent gap of the governor side pickup when the ramps arecentered, and adding this value to theequivalent gap of the generator side pickup when the ramps are centered.This calculation will result in anequivalent mechanical gap at which

the generator side pickup is to begapped.

DE Installation Checklist

1. Mounting Type, Collar Ramp2. Ramp Angle3. No. of Pickups, Single Dual4. Pickup Linear Range5. DE Cold Setpoint6. DE Rotor Long Setpoint

7. DE Rotor Short Setpoint8. Pickup Orientation, Gov Side

Gen Side9. Target Sizing10. Pickup Bracket(s)

Documented11. Pickup Calibration

Documented12. Pickup Cold Gap Documented




Part-2 Valve Position

Valve position measurements are animportant aspect of a completeTurbine Supervisory Instrumentationsystem. Typically, the Main SteamControl (or Throttle) Valve is alwaysincluded in the system with other valves added depending upon thecontrol system incorporated in theturbine design. Addition of a position transducer to a hand wheeloperated throttle valve, which was

equipped only with a graduated scalefor indication, will allow more precision in valve positioning.

Valve position indication is actuallyameasurement of theamount avalve isclosed or

open. Thismeasurement is usually made with a Linear Variable Differential Transformer (LVDT), but sometimes a RotaryPotentiometer is used for specialapplications. Nearly all applicationsrequire bracket made to attach thetransducer and another bracket in physical contact with the moveable portion of the valve (stem or

linkage).

Theory of

Operation

LVDT LVDTs areelectromagneti

Rotary potentiometers, by their design, require that they be installedat the end of the throttle valve camshaft or valve linkage axle. Thevalve axle shaft will requiremodification to allow the potentiometer shaft to be clampedrigidly, possibly with a coupling.The potentiometer body must beinstalled so that it will not rotate asthe valve shaft rotates.

Conduit Dedicated conduit should be provided in all installations for mechanical protection of theinstrument cable. Rigid conduit isrequired from the monitor location tothe LVDT or Potentiometer location.The final 2-3 feet of the conduitinstallation should be completedwith flexible conduit to allow

transducer removal.

Measurement Convention Measurement convention involvesdetermining which direction thevalve operates in relation to theselected transducer location. TheSKF-CM LVDT rod is spring loadedwith the rod forced in the extendeddirection. The standard installationhas the extended orientation

representing 0% indication or valveclosed and the fully compressedorientation representing 100%indication or valve fully open.

Many times the LVDT cannot beinstalled in the standardconfiguration and orientation due to



selected by using the center portionof the standard operating range.

SKF-CM's rotary potentiometer P/N# CMSS30503100 is a direct

replacement for GE Catalog No.9888323 or equivalent.

Transducer Installation The body of the LVDT is designedto be rigidly attached to animmoveable location or bracket withthe rod pressing against the valvestem or bracket. Proper installationof the LVDT involves selecting alocation where the operating range is

not exceeded and the LVDT rod hasfree travel.

For valves having longer strokes(greater than 2 inches), a suitablelocation along the valve operatinglinkage must be selected where theLVDT operating range is notexceeded. An alternative is to designa custom linkage to achieve therange required. A second alternative

is to install a circular cam to thethrottle valve cam assembly andinstall the LVDT so that the rodcontacts the cam. The latter alternative is not applicable onsmaller turbines which may not havea throttle valve manifold containingseveral steam valves intended toopen at differing loads.

until 0 VDC is obtained at thetransducer output. As the valve isstroked throughout its full range, theLVDT output voltage should benoted for monitor re-calibration.

Rotary potentiometers are installedso that the potentiometer shaft isattached to the cam shaft or other rotating shaft with the potentiometer body rigidly held so it does notrotate. The potentiometer should beadjusted so that when the rotationangle is 0 the signal output should be0 VDC. As the valve is strokedthroughout its full range, the output

voltage should be noted for monitor re-calibration.

Valve Position Checklist

1. Xdcr Type, LVDT RotaryPot.

2. Operating Range3. Transducer Location(s)4. Measurement Convention5. Xdcr Installation

Documented6. Correct Instrument Wire7. Flexible Conduit8. Calibration


Shell or Case Expansion

Shell or Case Expansion is a veryimportant measurement as part of a




Turbine SupervisoryInstrumentation (TSI) System for large steam turbines. Thismeasurement should be included inturbine retrofit plans when at all

possible.

The Shell Expansion measurementis utilized by operators to monitor

the proper thermalgrowth of theturbine's shellduring

startup, operation, and shutdown.The turbine's shell is anchored to

the foundation at one end of themachine and allowed to expand or grow by sliding towards theopposite end. The expansion or growth of the turbine's shell ShellExpansion is the measurement of how much the turbine's shellexpands or grows as it is heated. Aslarge turbine cases grow or expandthermally, in some case up toseveral inches, and was usually

supplied as part of the OriginalEquipment Manufactures TSIsystem supplied with the turbine.

Used in conjunction with aDifferential Expansion (DE)measurement (Case to Rotor) thethermal growth of both the case androtor can be monitored to preventcostly rubs between the rotatingand stationary parts of the turbine.

The recommended Shell Expansionmeasurement device is a SKF-CMLVDT (Linear VariableDifferential Transformer)engineered and manufactured to provide long measurement ranges,long life and simple installation.

Operating Range The standard SKF-CM LVDT has astandard operating range of 1.0, 2.0 or 4.0 inches. The most common used is0-2 inches (0-50 mm) with an output

of (-7.5)-0-(+7.5) VDC and anaccuracy of ±0.5 % full scale. Ashorter range may be selected byusing the center portion of the LVDTstandard operating range. Longer ranges are available on request.

Transducer Installation The body of the LVDT is designed to be rigidly attached to the turbinefoundation and the spring loaded

roller tipped plunger is to pressagainst a bracket that is attached to theFront Standard or Turbine Case.

The bracket must be designed not tointerfere with turbine operation andallow the roller tip of the plunger to

ride against freely it throughout theentire range.

Measurement ConventionThe SKF-CM LVDT operates on the standardinstrument convention that as the plunger or rod is compressed into theLVDT body (motion towards the



Both retrofit and new applicationsmay be accommodated easily withSKF-CM's LVDT design as itincorporates a protective epoxycoated aluminum housing with

mounting flanges and a springloaded plunger with an adjustableroller tip.

Occasionally due to improper turbine shell pre-heating,maintenance or the location of thesteam inlets being used to preheatthe turbine the turbine shell may become distorted which can causeinternal damage.

Turbine "Cocking" occurs when theturbine sliderhangs up or sticks onone side of the foundation andcontinues to grow or slide on the

other. Thisconditionsometimescorrects itself by breakingloose quite

dramatically.To monitor for distortionor cockingtwo (2)

LVDT's may be utilized and areinstalled on either side of the FrontStandard or turbine case. If theTurbine Case does not grow evenlythe case is allowed to cool and thenreheated with more even heat

distribution.

Theory of Operation

transducer) the signal output increasesor goes more positive. The LVDTmay be installed in either direction sothat thermal growth causes a more positive going signal or a negative

going signal. The monitoring systemcan be configured for either direction.

ConduitDedicated TSI Systemconduit should be provided in allinstallations for mechanical protectionof the instrument cable. Rigid IMCconduit is required from the monitor location to the LVDT location. Thefinal 2-3 feet of the conduitinstallation should be completed with

flexible conduit to facilitate transducer removal. The SKF-CM LVDT body isequipped with a 3/4" NPT conduitfitting.

Instrument WireFor LVDTapplications a 4-conductor, twisted,shielded, insulated, instrument wireshould be utilized between themonitor location and the LVDT. Thiswire should be a continuous run and

not be spliced. Alternately two (2)individual twisted, shielded pair wiresmay be used with extreme care takento properly tag the cables to preventimproper connection.


Red -18 VDC White Signal

Black Power Ground Green Signal Ground

The following Belden Part# have been provided for your convenience. Theymay be cross referenced to other wiremanufactures.

Belden Part Numbers



LVDT's are electromagneticdevicesthat havethree coilsof wire

wound ona hollowtube and ametal rodmovinginside thehollow

tube. The primary coil of wire isexcited by a supply voltage whichinduces a voltage in the other coilsas t&he rod or plunger travels

throughout its range. When the plunger is centered in its range theinduced voltage of the twosecondary coils is equal inmagnitude, but opposite polarity.As the plunger moves to either sideof the center position the voltage of one of the secondary coils increaseswhile the other secondary coilexperiences a decreased voltage.DC LVDT's differ from AC

LVDT's in that they aremanufactured with an internalcarrier generator/signalconditioning module and onlyrequire DC Power to operate.

Pair Nom.O.D.

4-Cond. Nom.O.D.

18AWG 8760 0.22" 9418 0.25"

20AWG

8762

0.20"

N/A

Calibration The SKF-CM 2" LVDT is designed to be installed so that when the plunger

is centered in itsoperating rangethe LVDToutput voltagewill be 0 VDC.This calibration

may be accomplished by temporarilyinserting a block whose thickness isequal to exactly one-half (1/2) thedesired range (for 2" range use a 1.0" block) under the plunger tip when theturbine case is in it's cold position andadjusting the roller tip, located at theend of the plunger, until 0 VDC isobtained at the transducer output.Alternatively, the entire body of theLVDT may be repositioned.

Shell Expansion Checklist

1. Number of LVDT,s, One Two2. Operating Range3. Transducer Location(s)4. Measurement Convention5. LVDT Installation

Documented6. Correct Instrument Wire7. Metal IMC Mainline Conduit8. Flexible Conduit9. Calibration

Eddy Current Transducer Installation

Part-1 Radial Vibration



Eddy Current Transducers(Proximity Probes) are the vibrationtransducer of choice when installingvibration monitoring on Journal

Bearing equipped rotatingmachinery. Eddy CurrentTransducers are the only transducersthat provide Shaft Relative (shaftrelative to the bearing) vibrationmeasurement.

Several methods are usuallyavailable for the installation of EddyCurrent Transducers, includinginternal, internal/external, and

external mounting.

Before selecting the appropriatemethod of mounting Eddy CurrentTransducers, special considerationneeds to be given to severalimportant installation considerationsthat will determine the success of your monitoring program.

Theory of Operation

Eddy Current Transducers work onthe proximity theory of operation. A

EddyCurrentSystemconsists of amatchedcomponentsystem: aProbe, an

Extension Cable and an Oscillator /Demodulator. A high frequency RFsignal @2 mHZ is generated by theOscillator/Demodulator, sent throughthe extension cable and radiatedfrom the Probe tip. Eddy currents aregenerated in the surface of the shaft.The Oscillator /Demodulator

The gauge of the selected wiredepends on the length of theinstrument wire run, and should beas follows to prevent loss of high

frequency signal:Up to 200 feet 22 AWG




Red -24 VDC

Black Common

White Signal

Common Point Grounding To prevent Ground Loops fromcreating system noise, systemcommon, ground and instrumentwire shield must be connected toground at one location only. In mostcases, the recommendation is toconnect commons, grounds andshields at the Monitor location. Thismeans that all commons, groundsand shields must be floated or notconnected at the machine.

Occasionally due to installationmethods instrument wire shields areconnected to ground at the machinecase and not at the monitor. In thiscase, all of the instrument wireshields must be floated or notconnected at the monitor.

Conduit Dedicated conduit should be provided in all installations for bothmechanical and noise protection.Flexible metal conduit should beused from the Eddy Probe to the



demodulates the signal and providesa modulated DC Voltage where theDC portion is directly proportional togap (distance) and the AC portion isdirectly proportional to vibration. In

this way, a Eddy Current Transducer can be used for both RadialVibration and distancemeasurements such as ThrustPosition and Shaft Position.


Number of Transducers All vibration transducers measuremotion in their mounted plane. In

other words, shaft motion either directly away from or towards themounted Eddy Current Probe will bemeasured as radial vibration.

On smaller less critical machines,one (1) Eddy Current Transducer system per bearing may be adequatefor machine protection.

The single Eddy Current Probe will

then measure the shaft's vibration inthat given plane. Therefore, the EddyCurrent Probe should be mounted inthe plane where the greatestvibration is expected.

On larger more critical machines,two (2) Eddy Current Transducer systems are normally recommended per bearing. The Probes for this typeof installation should be mounted

900 apart from each other. Since theProbes will measure the vibration intheir respective planes, the shaft'stotal vibration within the journal bearing is measured. An "Orbit" or cartesian product of the twovibration signals may be viewedwhen both Eddy Current Transducers

Oscillator /Demodulator junction box, and rigid bonded metal conduitfrom the junction box to the monitor.

Calibration

All Eddy Current Systems (Probe,Cable and Oscillator Demodulator)should be calibrated prior to beinginstalled. This can be done by usinga SKF-CM CMSS601 StaticCalibrator, -24 VDC Power Supplyand a Digital Volt Meter. The Probeis installed in the tester with thetarget set against the Probe tip. Themicrometer with target attached isthen rotated away from the Probe in

0.005" or 5 mil increments. Thevoltage reading is recorded andgraphed at each increment. TheCMSS601 Calibrator will produce avoltage change of 1.0 VDC +-0.05VDC for each 5 mils of gap changewhile the target is within theSystems linear range.

Gap When installed,Eddy Current Probes

must be gapped properly. In mostRadial Vibration applications,gapping the transducer to the center of the linear range is adequate. For the Model CMSS65 and 68 gapshould be set for -12.0 VDC using aDigital Volt Meter (DVM), thiscorresponds to an approximatemechanical gap of 0.060" or 60 mils.The voltage method of gapping theProbe is recommended over

mechanical gapping. In all cases,final Probe gap voltage should bedocumented and kept in a safe place.



are connected to an SKF-CMInformation System or anOscilloscope.

Linear Range

Several versions of Eddy CurrentTransducers are available with avariety of Linear Ranges and bodystyles. In most cases, SKF-CM'sCMSS68 with a linear range of 90mils (0.090") is more than adequatefor Radial Vibration measurements...


CMSS65 90mils

200mV/mil

1/4"x28 UNF1" to 5" Length

CMSS68 90mils

200mV/mil

3/8"x24 UNF1" to 9" Length

CMSS62 240mils

50mV/mil

1" x 12 UNF 1"to 5" Length


Target Material Eddy Current Transducers are

calibratedat the

factory for 4140 Steelunlessspecifiedotherwise.

As Eddy Currents are sensitive to the permeability and resistivity of theshaft material any shaft materialother than 4000 series steels must bespecified at the time of order. Incases of exotic shaft material a

sample may need to be supplied tothe factory.

Mechanical Runout Eddy Current Transducers are alsosensitive to the shaft smoothness for Radial Vibration. A smooth (64micro-inch) area approximately 3

Internal Mounting

InternalMounting isaccomplishe

d with theEddyCurrentProbesmounted

internally to the machine or bearinghousing with a SKF-CM CMSS903Bracket or with a custom designedand manufactured bracket. TheTransducer system must be installedand gapped properly prior to the

bearing cover being reinstalled.Provisions must be made for thetransducer's cable exiting the bearinghousing. This can be accomplished by using an existing plug or fitting,or by drilling and tapping a holeabove the oil line. The Transducer'scables must also be tied down withinthe bearing housing to prevent cablefailure from "windage".

For added safety and reliability, allfasteners inside the bearing housingshould be safety wired, or otherwise prevented from working loose insidethe machine.

Advantages of Internal Mounting

Most economical installation. Less machining required. True bearing relative

measurement. Usually good viewing

surface for Eddy Probe.

Disadvantages of Internal Mounting

No access to Probe whilemachine is running.



times the diameter of the Probe must be provided for a viewing area. The prepared journal area on most shaftsare wider than the bearing itself allowing for Probe installation

immediately adjacent to the bearing.

Electrical Runout Since Eddy Current Transducers aresensitive to the permeability andresistivity of the target material andthe field of the transducer extendsinto the surface area of the shaft byapproximately 15 mils (0.015"), caremust be taken to avoid nonhomogeneous viewing area materials

such as Chrome.

Another form of electrical runout can be caused by small magnetic fieldssuch as those left by Magna-fluxingwithout proper degaussing.

Perpendicular to shaft centerline Care must be exercised in allinstallations to insure that the EddyCurrent probes are mounted

perpendicular to the shaft center-line.Deviation by more than 1-2 degreeswill effect the output sensitivity of the system.

Orientation of Transducer(s) As most machine casings arehorizontally split, transducers arecommonly found mounted at 450 both sides of vertical 900 apart.

If possible transducer orientationshould be consistent along the lengthof the machine train for easier machine diagnostics. In all casesorientation should be welldocumented.


Cables must be tied downdue to "windage".

Transducer cable exits must be provided.

Care must be taken to avoid

oil leakage.

External/Internal Mounting

External/Internal mounting isaccomplished when the Eddy Probesare mounted with a MountingAdapter (SKF-CM CMSS911 or

904). Theseadoptersallow

externalaccess to theProbe yetallows theProbe tip to

be internal to the machine or bearinghousing. Care must be taken indrilling and tapping the bearinghousing or cover to insure that theEddy Probes will be perpendicular tothe shaft center line.

In some cases due to spacelimitations External/Internalmounting is accomplished bydrilling or making use of existingholes in the bearing itself, usually penetrating at a oil return groove.

Advantages of External/InternalMounting

Eddy Probe replacementwhile machine is running.

Usually good viewing areafor Eddy Probe.

Gap may be changed whilemachine is running.

Disadvantages of External/Internal



clearances The RF Field emitted from the Probetip of a Eddy Current Transducer inapproximately a 450 coned shape.

Clearance

must be provided onall sides of the Probe tipto preventinterference

with the RF Field. As an example, if a bearing is drilled to permitinstallation, the hole must be counter bored to prevent side clearanceinterference. Care must also be taken

to avoid collars or shoulders on theshaft that may thermally "grow"under the Probe tip as the shaftgrows from heat.

Eddy Current Probe tip to tip

clearances Although Probe tip to tip clearancesare not normally an issue on mostmachines, it should be noted thatEddy Current Probes radiate an RF

Field larger than the Probe tip itself.As an example, Model CMSS65 and68 probe should never be installedwith less than one (1) inch of Probetip to tip clearance. Larger Probesrequire more clearance. Failure tofollow this rule will allow theOscillator/Demodulator to create a"beat" frequency which will be thesum and difference of the twoOscillator/Demodulator RF

frequencies.

System Cable Length and Junction

Boxes Eddy Current Transducer Systemsare a "tuned" length, and severalsystem lengths are available. Lengthis measured from the Probe tip to the

Mounting

May not be true bearingrelative measurement.

More machining required.

Long Probe/Stinger length(Resonance).

External Mounting Pure external Eddy Probe mountingis usually a last resort installation.The only valid reason for using thismethod is inadequate space available

within the bearinghousing for

internalmounting.Special care

must be given to the Eddy Probeviewing area and mechanical protection of the transducer andcable.

Advantages of External Mounting

Most Inexpensive

Installation.

Disadvantages of External Mounting

May be subject to "Glitch" or Electrical/Mechanical runout.

Requires mechanical protection.

Installation Checklist

1. Mounting Type, InternalExternal/Internal External

2. Number of Transducers, X or X&Y

3. Target Material, 4140 Other 4. Smooth Target Area5. Size of Target Area6. Junction Box Location(s)



Oscillator/Demodulator, and ismeasured electrically which canslightly vary the physical length. For example, the Model CMSS65 and 68are available in 5 and 10 meter

system lengths. Care must be takento insure that the proper systemlength is ordered to reach therequired Junction Box.

Grounding and Noise Electrical noise is a very seriousconsideration when installing anyvibration transducer, and special careneeds to be taken to preventunnecessary amounts of noise. As

most plant electrical noise is 60 HZ,and many machines running speed isalso 60 HZ, it is difficult to separatenoise from actual vibration signal.Therefore, noise must be kept to anabsolute minimum.

Instrument Wire A 3-wire twisted shielded instrumentwire (ie; Belden #8770) is used toconnect each Oscillator/Demodulator

to the Signal Conditioner in theMonitor. Where possible, a singlerun of wire from theOscillator/Demodulator (JunctionBox) to the Monitor location should be used. Splices should be avoided.

7. Metal Conduit (Junction Boxto Monitor)

8. Flexible Conduit (JunctionBox to Probe)

9. Correct Instrument Wire

10. Shielding Convention,Monitor or Machine

11. Calibration12. Gap Set

Thrust Position



Thrust Position Monitoring was oneof the earliest applications for EddyCurrent Transducers. When utilizedfor this application, it is one of themost important components of a

complete machinery protectionsystem.

Thrust bearing failure produces oneof the most catastrophic failures of arotating machine. This could resultin very expensive repairs and the possibility of machine replacement.

Theory of Operation

Eddy Probe Transducers work on the proximitytheory of operation. ASystemconsists of amatchedcomponent

system: a Probe, an Extension Cableand an Oscillator /Demodulator. Ahigh frequency RF signal @2 mHZ

is generated by theOscillator/Demodulator, sent throughthe extension cable and radiatedfrom the Probe tip. Eddy currents aregenerated in the surface of the shaft.The Oscillator /Demodulator demodulates the signal and providesa modulated DC Voltage where theDC portion is directly proportional togap (distance) and the AC portion isdirectly proportional to vibration. In

this way a Eddy Current Transducer can be used for both RadialVibration and distancemeasurements such as ThrustPosition and Shaft Position.


Zero (0) Center of Float Zone When usingthe "Center of FloatZone"

method, theshaft ismechanically barred to

accurately measure the total floatzone with the Eddy Probe or a dialindicator. The Eddy CurrentTransducer and monitoring systemare then gapped and calibrated toread zero (0) when the thrust collar isin the center of the float zone. This

method provides equal range in bothactive and inactive directions, andwhen the machine is runningnormally with no wear themonitoring system will display onehalf the float value (ie: +5 mils).

Plus (+) Active (Normal) Direction All thrust position monitoringsystems are installed and calibratedso that wear on the active thrust

bearing or normal direction producesa plus (+) or up scale reading. Minus(-) or down scale readings indicatemotion towards the inactive thrust bearing.


Target Material Eddy Current Transducers are

calibrated at

the factoryfor 4140Steel unlessspecifiedotherwise.As Eddy

Probe's are sensitive to the permeability and resistivity of the



Single or Dual Voting When determining the number of transducers for monitoring thrust position on a machine, severalfactors should be considered. First,

will the system be required to trip themachine if thrust failure is detected,or secondly what other means areavailable to verify thrust failure.

One of the rotating machineryinstrumentation standards, API- 670(American Petroleum Institute),specifies Dual Voting ThrustPosition at each thrust bearing. Thisapproach to thrust position

measurement requires that twotransducers be mounted at eachthrust bearing. Their respectiveoutput signals are then compared toalarm and shutdown limits. Bothoutput signals must exceed theshutdown limit before the machine istripped. This method of thrustmeasurement increases the system'sreliability, and is recommended for shutdown operation.

A single Eddy Current transducer for thrust position measurement shouldonly be used when the monitoringsystem is not required to shutdownthe machine, and other means areavailable to verify thrust failure.

Circuit OK and Fault DetectionCircuits are not used whenmonitoring thrust position, as they

could effect the proper monitoring of this parameter. The reason for this isthat a rapid thrust failure could causeFault Detection Circuits to operateinhibiting a valid shutdown alarm.

Linear Range Several versions of Eddy Current

shaft material, any shaft materialother than 4000 series steels must bespecified at the time of order. Incases of exotic shaft material, asample may need to be supplied to

the factory.

Mechanical Runout Eddy Probe Transducers aresensitive to the shaft or targetsmoothness. For thrust position, a perfectly smooth finish is notrequired as the system will averagethe DC signal. An areaapproximately 3 times the diameter of the Eddy Probe must be provided

for a viewing area.


clearances The RF Field emitted from the Probetip of a Eddy Probe is approximatelya 450 coned shape. Clearance must be provided on all sides of the Probetip to prevent interference with theRF Field. As an example, if a bearing is drilled to permit

installation, the hole must be counter bored to prevent side clearanceinterference. Care must also be takento avoid collars or shoulders on theshaft that may thermally "grow"under the Probe tip as the shaftgrows from heat.

Eddy Probe tip to tip clearances Although Eddy Probe tip to tipclearances are not normally an issue

on most machines, it should be notedthat Eddy Probe's radiate an RF Fieldlarger than the Probe tip itself. As anexample, SKF and Bently Nevada probes should never be installed withless than one (1) inch of Probe tip totip clearance. Larger Probe's requiremore clearance. Failure to follow



Transducers are available with avariety of Linear Ranges and bodystyles. In most cases, SKF andBently Nevada models with a linear range of 90 mils (0.090") is more

than adequate for thrust positionmeasurements. Other Eddy Currenttransducer systems are available withranges of 250, 500 and 1000 mils.

The range required for each thrustapplication can be calculated per thefollowing example:

1. Active Direction

Allowable Wear 30

mils

2. Float Zone 10

mils

3. Inactive Direction

Allowable Wear 30

mils

------

-

Total Range Required 70

mils

Transducer Location Transducer location is very

important for a proper ThrustPosition monitoring system. Theobjective of the system is to measureactual thrust position. Thusly, careneeds to be taken that the system isnot observing items such as thermalgrowth of the shaft.

STI recommends that the EddyCurrent Transducers used to monitor thrust position be located within two

(2) shaft diameters of the thrust bearing. This assures that the eddyCurrent system is not adverselyaffected by shaft thermal growth. Insome cases this is not possible, andthe engineer needs to be aware of thethermal growth expected and planaccordingly.

this rule will allow theOscillator/Demodulator to create a"beat" frequency which will be thesum and difference of the twoOscillator/Demodulator RF

frequencies.

System Cable Length and Junction

Boxes Eddy Probe Transducer Systems area "tuned" length, and several systemlengths are available. Length ismeasured from the Eddy Probe tip tothe Oscillator/Demodulator, and ismeasured electrically which canslightly vary the physical length. For

example, SKF and Bently Nevada probe systems are available in 5, and10 meter system lengths. Care must be taken to insure that the proper system length is ordered to reach therequired Junction Box.

Grounding and Noise Electrical noise is a very seriousconsideration when installing anyvibration transducer, and special care

needs to be taken to preventunnecessary amounts of noise. Asmost plant electrical noise is 60 HZ,and many machines running speed isalso 60 HZ, it is difficult to separatenoise from actual vibration signal.Therefore, noise must be kept to anabsolute minimum.

Instrument Wire A 3-wire twisted shielded instrument

wire (ie; Belden #8770) is used toconnect each Oscillator/Demodulator to the Signal Conditioner in theMonitor. Where possible, a singlerun of wire from theOscillator/Demodulator (JunctionBox) to the Monitor location should be used. Splices should be avoided.



STI also recommends that the EddyCurrent Transducer(s) observe anintegral part of the shaft, as it possible for a nonintegral part toloosen creating a false reading. As an

example, sweated on Thrust Collarshave been known to come loose,causing normal thrust readings whilethe machine is experiencing acatastrophic failure.

External Mounting External mounting is the preferablemethod of mounting, and can becompleted when the end of the shaft

is accessible

through acover plateor end plate.Care must betaken toinsure thatthe thrust bearing is on

the same end of the machine so thatthe measurement will not be affected by thermal growth of the shaft. The

STI Model CMCP912 Dual AxialProbe Mounting Adapter with NPTis available for this application.

Advantages of External Mounting:

Eddy Current Probereplacement while machine isrunning.

Usually good viewing surface Gap may be changed while

machine is running.

Disadvantages of ExternalMounting:

May not be close to thrust bearing.

The gauge of the selected wiredepends on the length of theinstrument wire run, and should beas follows to prevent loss of highfrequency signal:





Red -24 VDC

Black Common

White Signal

Common Point Grounding To prevent Ground Loops fromcreating system noise, systemcommon, ground and instrumentwire shield must be connected toground at one location only. In mostcases, the recommendation is toconnect commons, grounds andshields at the Monitor location. Thismeans that all commons, groundsand shields must be floated or not

connected at the machine.

Occasionally due to installationmethods instrument wire shields areconnected to ground at the machinecase and not at the monitor. In thiscase, all of the instrument wireshields must be floated or notconnected at the monitor.

Conduit

Dedicated conduit should be provided in all installations for bothmechanical and noise protection.Flexible metal conduit should beused from the Eddy Probe to theOscillator /Demodulator junction box, and rigid bonded metal conduitfrom the junction box to the monitor.



Internal Mounting Internal mounting is accomplished by installing the Eddy CurrentTransducer either directly throughthe thrust bearing or with a custom

designed bracketviewing thethrust collar directly or anearbyshoulder onthe shaft.Care must be

taken in locating and tieing down thetransducer cable(s) to prevent

damage. If an existing exit hole fromthe case does not exist, one will needto be drilled and tapped above the oilline.

Advantages of Internal Mounting:

Usually good viewing surface Close to thrust bearing.

Disadvantages of Internal Mounting:

No access to transducer whilemachine is running.

Cables must be tied downdue to "windage".

Transducer cable exits must be provided.

Care must be taken to avoidoil leakage.

Measurement Conventions

There are several importantmeasurement conventions that must be decided upon prior to installationand calibration of the system. It issuggested that the conventions used be common throughout a plant toavoid confusion.

Calibration All Eddy Current Systems (Probe,Cable and Oscillator Demodulator)should be calibrated prior to beinginstalled. This can be done by using

a STI CMCP601 Static Calibrator, -24 VDC Power Supply and a DigitalVolt Meter. The Eddy Probe isinstalled in the tester with the targetset against the Probe tip. Themicrometer with target attached isthen rotated away from the probe in0.005" or 5 mil increments. Thevoltage reading is recorded andgraphed at each increment. The SKFor Bently Nevada probe systems will

produce a voltage change of 1.0VDC +-.05 VDC for each 5 mils of gap change while the target is withinthe Probe's linear range.

Gap Extreme care must be taken to insurethat the thrust position Eddy Probesare gapped properly. Failure to gapthe transducer properly will result inthe allowed mechanical thrust range

being outside the transducer's linear range.

Prior to gapping the Eddy Probe, theallowable wear, float zone andassociated alarm levels should beover-laid onto the Probes calibrationwork-sheet. This will insure that allalarms are within the linear range of the Eddy transducer and provide theoptimum installed gap.

It will be necessary to mechanically bar the shaft to a known position,usually against the active thrust shoe(see Measurement Conventions).The Eddy Probe can then be gappedand the DC voltage documented.



Thrust Position Signal ConditioningCards and Display Scales areavailable in many different ranges tosatisfy most applications, specialsare available on request. The most

common thrust position full scalerange is +40-0-(-)40 mils which issmaller than the maximum range of 90 mils for SKF or Bently Nevadaeddy current probe systems..

STI CMCP540 Thrust PositionTransmitter/Monitor modules have a jumper so that upscale and down-scale may be changed to reflectactual movement of the shaft. Care

needs to be taken to document active(normal) thrust direction of the shaft,and the actual measurement directionof the Eddy Probe Installation.

Zero (0) Active Shoe When using the "Zero Active Shoe"method of gapping the Eddy CurrentTransducer, the shaft is mechanically

barred in theactive or

normaldirectionthrough thefloat zoneuntil it isagainst theactive thrust

shoe. The Eddy Probe andMonitoring System are thencalibrated to Zero (0). This methodof calibration provides more system

range in the active direction, andwhen the machine is operatingnormally with no wear of the thrust bearing the monitoring system willread "0".

Installation Checklist

1. Mounting Type, InternalExternal

2. Number of Transducers,

Single or Dual3. Linear Range4. Close to Thrust Bearing

(Thermal growth)5. Target Integral to Shaft6. Target Material, 4140 Other 7. Smooth Target Area8. Size of Target Area9. Active Thrust Direction10. Eddy Probe Installation

Documented

11. Junction Box Location(s)12. Metal Conduit (Junction Box


Box to Eddy Probe)14. Correct Instrument Wire15. Shielding Convention,

Monitor or Machine16. Calibration17. Gap Set and Documented

Field Wiring



There are a number of items to consider wheninstalling a permanently mounted vibrationmonitoring system. One of the most criticalconsiderations is the selection and installation

of the wire connecting the vibrationtransducers mounted on a machine to theassociated Monitoring Instrument. Along withthis wiring, there are a number of ExternalMonitor Connections that requireconsideration. All of this Monitor Systeminterconnection is often referred to as the"Field Wiring".

If the Field Wiring is not designed andinstalled in an appropriate manner, Noise or

Line Interference can be induced into theVibration Monitoring System. Since theinduced noise is normally an alternatingwaveform, the Monitoring Instrument willinterpret this signal as false vibration. TheVibration Monitoring System will then notfunction as a reliable and credible protectionor information system.

Low Level Signals

Vibration monitoring systems utilize the lowlevel (voltage) output signals available fromvibration transducers. These signals representthe actual vibration or motion of themachine's shaft or bearing housing.

The following table provides a comparison of vibration transducer signal levels expectedfrom a machine running at 3600 RPM.

Transducer

Type

Expected

Vibration

Transducer

Scale Factor

Output

Level Non-Contacting

1.0 mil 200 mV/mil 200 mV

Velocity 0.2 in/sec 1080 mV/in/sec

216 mV

Accelerometer 0.1 g 100 mV/g 10mV

10 mV

Red - Power

Black - Common

Clear - Signal

Recorder Outputs

The cable from the VibrationMonitor's Recorder Outputs to anyRecording Device should be atwisted pair cable. These cablesshould be stranded, individuallyinsulated, shielded, and overall jacket. Before installing thesecables, the manuals for both theVibration Monitor and theRecording Device should bereviewed for proper connections of Signal, Common, and Shield.

Junction Boxes

An important piece of hardware for installing Vibration Transducers areJunction Boxes. These should belocated at the machine for mountingthe Non-Contacting Pickup's SignalSensors and interfacing terminal

strips.

A Junction Box is also the transition point of Flexible Conduit to theVibration Transducer and RigidConduit to the Monitor. TheJunction Boxes should be installedclose to the monitored point andwithin the length of the VibrationTransducer's extension cable. TheJunction Boxes should be mounted

in a convenient location for serviceability. They should not bemounted under machine skirts or other inaccessible locations whenthe machine is running.

Following is a list of available SKF-



As can be seen from this table, the signallevel available from these transducers is quitelow. Base Line Noise especially in Power Plants, can be as high as 200 mV if the

Vibration Monitoring System is not properlyinstalled. Since a running speed of 3600 RPMis the same as 60 Hz, any noise induced onthe Vibration Monitoring System by a power source will be interpreted as 1 times runningspeed vibration. These transducer signallevels are also frequency dependent. Whenthe machine's designed running speed isincreased, in general, the expectedDisplacement level will decrease, theexpected Acceleration level will increase, and

the expected Velocity level will remainconstant.

Noise Sources

Noise or Line Interference can be induced in aVibration Monitoring System in a number of ways. However, there must first exist a sourcefor the induced noise. There are numerousnoise sources available in an industrial or power generation plant:

AC Power Transients Ground Differentials Switching Circuits High Voltage Circuits Improper Load Balance

Noise can be induced in a VibrationMonitoring System through Electrostatic(Capacitive), Electromagnetic (Inductive) or Conductive Coupling (Direct Connection).

All noise will be induced in the monitoringsystem through one or more of its externalconnections or Field

CM NEMA 4X Junction Boxes.

Eddy Probes

Two Channel (2Drivers)

P/N CMCP-150-02

Four Channel (4Drivers)

P/N CMCP-150-04

Six Channel (6Drivers)

P/N CMCP-150-06

Accelerometer

6"H x 6"W x 4"D P/N CMCP260-01

8"H x 6"W x 4"D P/N CMCP-260-02

10"H x 8"W x 6"D P/N CMCP-260-03

Following is a list of availableExtension Cable Lengths for location distance from theassociated transducer.

5mm Eddy Probe (meters) 5 or 10

8mm Eddy Probe (meters) 5, 10, or 15

LCV-100 Low Cost Velocity(feet)

10, 30 or 50

LCV-150 Velocity Pickup(feet)

Up to 100

793V Velocity Pickup (feet) 16, 32, or

64

Accelerometer 16, 32, or 64

Conduit

For a quality Vibration MonitoringSystem installation, it is critical thatconduit be utilized on the VibrationTransducer and its associatedInstrument Wiring. The use of

conduit greatly reduces the possibility of induced noise or lineinterference on the signal path. Theconduit system should be dedicatedsolely to the Vibration MonitoringSystem, and no other wiring of anyclassification should be in the same



Wiring.

AC Power

The AC Power source for the VibrationMonitoring System or any other ElectronicInstrument needs to be a "CLEAN" source.This implies that the source must be free of power surges and transient voltages. A power conditioning device, such as a Sola® Power Line Conditioner, can be installed to alleviatethese power source problems. For installations with a switchable power source,an Uninterruptable Power Supply (UPS) will be required.

A Straight Blade 3-Wire Grounding Power Receptacle (Nema 5-15R) is required for the power connection. This AC Power Sourceshould have a manually operated switchdevice, or circuit breaker, in line. A separatelysourced Service Power Receptacle should belocated near the Vibration Monitoring Systemfor test equipment power. The Power Sourceshould be checked for the following voltagetolerances at 50/60

Hz.

110 VAC Instrument Power

Neutral to Ground 0 VAC

Line to Neutral 105-126 VAC

Line to Ground 105-126 VAC

220 VAC Instrument Power

conduit.

Cable Trays, Wire Ways, or Instrument Trays are anunacceptable alternative to

dedicated conduit. This conduitmust be routed as far as possibleaway from any power cables. Thisis also the case when the VibrationMonitor is installed in a cabinet.The Instrument Wire for thetransducers must be separated asmuch as possible from both Power and Relay Contact cables. Allconduit must be installed andgrounded in compliance with the

appropriate

Articles of the National ElectricalCode, in effect at time of installation.

Instrument Wire Conduit and Power Cable Conduit parallel runs should be avoided when possible. When parallel runs cannot be avoided, thefollowing spacing should be used.

Length Of Run

120/240VCircuits

480-6900VCircuits

0-100' 2' 4'

100-250' 4' 8'

250-400' 6' 12'

400-550' 8' 16'

This recommended spacing is basedon a 500 Ampere Circuit and can beadjusted proportionally for other

loads. However, a minimumspacing of (1) foot should bemaintained. At conduit cross over locations, a minimum spacing of one (1) foot should be maintained.

Rigid Metal ConduitRigid Metal Conduit (IMC)



Line-1 to Line-2 207-242 VAC

Line-1 to Ground 104-121 VAC

Line-2 to Ground 104-121 VAC

Relay Connections

When a Monitor's Relay Contact Connectionsare used for annunciation or shutdown, thesedriven circuits must be free of noise or voltage transients. If these circuits present problems, Slave Relays should be utilized toisolate the Vibration Monitoring System fromthis known noise or transient sources.

Startup Connections When Startup Contact Connections are usedto initiate a Monitor's Startup feature, the

drive circuit (dry contact) must be free of noise or voltage transients. Again, if thesecircuits present problems, Slave Relaysshould be utilized to isolate the VibrationMonitoring System from this known noise or transient sources.

Transducer Instrument Wiring The Instrument Wire from the VibrationTransducer to its Monitor should be either atwisted pair or triad cables depending on the

Transducer's requirement. These cablesshould be stranded, individually insulated,shielded, and overall jacket. The shields or drain wires must be insulated or isolated fromeach other and the conduit. The use of multi-conductor cable with a single shield isstrongly discouraged due to its susceptibilityto induced noise and line interference.

The gauge or thickness of the InstrumentWire is determined by the distance between

the Vibration Transducer and Monitor. Longlengths of Instrument Wire acts as a low passfilter, and will attenuate high frequencysignals. This situation can be a problem whenmonitoring gear mesh frequencies, blade passage, or roller element bearings with ahigh frequency accelerometer.

continuously bonded made of ferrous (magnetic) material must beused between the VibrationMonitoring system and the JunctionBoxes located at the machine.

Flexible Metal ConduitFlexible Coated Metal Conduit(Sealtite® or Liquatite®)continuously bonded may be usedfrom the Junction Box to theVibration Transducer or machineentry point provided it is made of ferrous (magnetic) material.

When installing Instrument Wire

Conduit, the Conduit must not beoverfilled with Instrument Wires.As a rule, only 40% to 50% of theConduit's cross sectional areashould be filled with InstrumentWires. This fill ratio allows easier installation of the Instrument Wireswith some future expansioncapability. Following is a crossreference table of recommendedcables, conduit size, and number of

cables installed in the conduit.

NUMBER OF CABLES PER CONDUIT

CONDUIT: 1/2" 3/4" 1" 1-1/2"

2" 2-1/2"

CABLE

8760 3 6 11 26 43 61

8762 4 8 13 31 51 73

8761 6 11 17 43 69 99

8770 3 5 9 21 35 50

8772 4 7 11 27 44 64 8771 5 8 14 34 56 80

Grounding/Shielding

A "Single Point Grounding" schemeshould be utilized when installing a



The following table offers a guideline to helpselect the proper Instrument Wire gauge(AWG).

Transducer Type Length of Cable Runs

<200' <1000' >1000' Non-Contacting 22 AWG 20 AWG 18 AWG

Velocity 22 AWG 20 AWG 18 AWG

Accelerometer 20 AWG 18 AWG --

The following table is a partial list of Belden® Cables that should be used for theInstrument Wire. These part numbers can becross referenced to equivalent cables fromother manufactures. These cables are polyethylene insulated, twisted, with beldfoilshield and drain wire, and PVC jacket.

Belden Part Numbers

Pair Nom. O.D. Triad Nom. O.D.

18 AWG 8760 0.22" 8770 0.25"

20 AWG 8762 0.20" 8772 0.22"

22 AWG 8761 0.17" 8771 0.19"

A color code convention should be used wheninstalling the Instrument Wire as outlined in

the following list.

Vibration Monitoring System. Thisscheme of grounding means that allgrounds are connected or tied downat one location. It is highlyrecommended that for a Vibration

Monitoring System installation theSingle Point Ground should be atthe Monitor not at the Machine. Ona large machine or where multiplemachines are being monitored,substantial ground differentials(potentials) can be found betweentransducer locations.

All Instrument Wire shields must begrounded at one end of the cable,

and the other end left floating or notconnected. The Instrument Wireshould be grounded at the VibrationMonitoring System. If the shield isnot grounded, the shield will become an antenna increasinginduced noise on the signal path. If the shield is grounded at both ends,it will allow ground differential(potential) current (ground loop) toflow through the shield seriously

increasing induced noise andvoltage transients.

Checklist

AC Instrument Power o Proper Voltageo Power Groundo Common to Ground

<5 Voltso System Neutral to

Ground = 0 Volts Junction Boxes Used and

Accessible Solid Ferrous Metal Conduit

(Monitor to J-Box) Flexible Conduit (J-Box to

Transducer) Instrument Wire



o Individual WiresUsed

o Insulated, Shielded,Twisted

o Proper Gauge

Common Point GroundScheme

Shield Grounded One EndOnly

Typical Layout Drawing

Accelerometer Transducers and Installation

Accelerometers have been a popular choicefor rotating machinery vibration monitoring.They are a rugged, compact, light weight

transducer with a wide frequency responserange. Accelerometers have been usedextensively in many machinery monitoring

applications. This transducer is typicallyattached to the outer surface of machinery.Generally this machinery will have parts that

generate high frequency signals, such as,

rolling element bearings or gear sets.

As can be seen in the figure above, themounting method also has an effect on theoperating

frequency range of an accelerometer.By design,

accelerometershave a naturalresonance which is

3 to 5 times higher

than the advertised high end frequency



The application and installation of an

accelerometer must be carefully consideredfor an accurate and reliable measurement.

Accelerometers were designed to bemounted on machine cases. This will provide

continuous or periodic sensing of absolutecase motion (vibration relative to free space)in terms of acceleration.

Theory of Operation

Accelerometers areinertial measurementdevices that convert

mechanical motion to an

electrical signal. This

signal is proportional tothe vibration's

acceleration using the piezoelectric principle.Inertial measurement devices measuremotion relative to a mass. This follows

Newton's Third Law of Motion: A body actingon another will result in an equal action onthe first.

Accelerometers consist of a piezoelectriccrystal and mass normally enclosed in a

protective metal case. As the mass applies

force to the crystal, the crystal creates acharge proportional to acceleration. The

charge output is measured in pico Coulombs

per g (pC/g) terms where g is the force of gravity. Some sensors have an internalcharge amplifier, while others have an

external charge amplifier. The chargeamplifier converts the charge output of thecrystal to a proportional voltage output inmV/g terms.

Current or Voltage Mode

This type of accelerometer has an internal,

low-output impedance amplifier and requiresan external power source. The external

power source can be either a constant

current source or a regulated voltage source.This accelerometer is normally a two wire

transducer with one wire for power and

signal, and the second wire for common. Thistype of Accelerometers have a lower

response. The frequency response range islimited so that a flat response is provided

over the operating range. The advertisedrange is achievable only by bolt mounting.Any other mounting method will adversely

affect the natural resonance, and in turn the

usable frequency response range.

Sensitivity

Accelerometers utilized for vibrationmonitoring are usually designed with a

sensitivity of 100 mv/g. Accelerometers can

be supplied with a wide range of sensitivitiesfor special applications such as structural

analysis, geophysical measurement, or veryhigh frequency analysis.

Frequency Range

Accelerometers are designed to measurevibration over a given frequency range. Oncethe particular frequencies of interest for a

machine are known, an accelerometer maybe selected. Typically, an accelerometer formeasuring machine vibration will have a

frequency range from 1 or 2 hertz to 8 or10k hertz.

An accelerometer is used on machines whenhigh frequency measurements are desired. Interms of energy sensed by the transducer,

acceleration will have larger amplitudes asthe frequency increases. At low frequencies,the acceleration amplitudes may be quite

small giving a false impression of anacceptably operating machine.

Calibration Piezoelectric accelerometers can not berecalibrated or adjusted. Unlike a velocity

pickup, this transducer has no moving parts

subject to normal wear. Therefore, theoutput sensitivity does not require periodicadjustments to correct for wear.

An accelerometers has internal components

which can be damaged from shock or

overheating. When an accelerometer issuspect, a simple test of the transducer's

bias voltage will help determine whether it

should be removed from service. Anaccelerometer's bias voltage is the DC



temperature rating due to the internalamplifier circuitry. Signal cable lengths up to

500 feet have negligible effect on the outputsignal quality. Longer cable lengths willreduce the effective frequency responserange.

Charge Mode

Charge mode accelerometers differ slightlyfrom current or voltage mode types. Thissensor has no internal amplifier and

therefore have a higher temperature rating.

An external charge amplifier is supplied witha special adapter cable which is matched to

the accelerometer. Field wiring is terminated

to the external charge amplifier. As withcurrent or voltage mode accelerometers,signal cable lengths up to 500 feet have

negligible effect on the output signal quality.Longer cable lengths will reduce the effectivefrequency response range.


Mounting There are three mounting methods typicallyused for monitoring applications: boltmounting, glue, and magnets.

The bolt mounting method is the best

method available for permanent mountingapplications. this method is accomplished viaa stud or a machined block. This method

permits the transducer to measure vibration

according to the manufacturer'sspecifications. The mounting location for the

accelerometer should be clean and paint

free. The mounting surface should be spot-faced to a surface smoothness of 32 micro-inches. The spot-faced diameter should be

10% larger than the accelerometer diameter.

Any irregularities in the mounting surfacepreparation will translate into improper

measurements or damage to theaccelerometer.

The adhesive or glue mounting method

provides a secure attachment withoutextensive machining. However, this

mounting method will typically reduce the

operational frequency response range. This

component of the transducer's output signal.The bias voltage is measured with a DC volt

meter across the transducer's signal outputand common leads with power applied. Atthe same time, the power supply voltage

should also be checked to eliminate the

possibility of improper power voltageaffecting the bias voltage level.

Instrument Wire The following table is a partial list of Belden®

Cables that should be used for the

instrument field wiring. These part numbersmay be cross referenced to equivalent cables

from other manufacturers. The listed cables

are polyethylene insulated, twisted, withBeldfoil shield, drain wire, and PVC jacket.

Belden® Part Numbers P/N Nom. O.D.

18 AWG 8760 0.22"

20 AWG 8762 0.20"

22 AWG 8761 0.18"

Common Point Grounding

To prevent Ground Loops from creating

system noise, system common, ground andinstrument wire shield must be connected to

ground at one location only. In most cases,

the recommendation is to connect commons,grounds and shields at the Monitor location.This means that all commons, grounds, and

shields must be floated or not connected atthe machine.

Occasionally, due to installation methods,instrument wire shields are connected toground at the machine case and not at the

monitor. In this case, all of the instrument

wire shields must be floated or not connectedat the monitor.

Conduit Dedicated rigid conduit should be provided in

all installations for mechanical and noise

protection. The conduit should be metal, andin direct contact with each segment. All

metal junction boxes and fittings should be in

direct contact with the conduit. With thistype of installation, a single ground point can



reduction is due to the damping qualities of the adhesive. Also, replacement or removal

of the accelerometer is more difficult thanany other attachment method. Surfacecleanliness is of prime importance for properadhesive bonding.

The magnetic mounting method is typically

used for temporary measurements with aportable data collector or analyzer. Thismethod is not recommended for permanent

monitoring. The transducer may be

inadvertently moved and the multiplesurfaces and materials of the magnet may

interfere with or increase high frequency

signals.

be established.

To facilitate removal of the accelerometer, a junction box with a barrier terminal strip

should be located close to the transducer.

The rigid conduit should be attached to the junction box, and the final run to thetransducer can be metal flexible conduit.

Accelerometer Checklist

1. Accelerometer Type, Current or

Voltage Charge

2. Sensitivity

3. Frequency Range

4. Calibration


6. Grounding7. Rigid Conduit

8. Location(s) Documented

Velocity Transducer Installation



The velocity pickup is a very popular transducer or sensor for

monitoringthevibrationof rotating

machinery. This type of vibrationtransducer installs easily onmachines, and generally costs lessthan other sensors. For these tworeasons, this type of transducer isideal for general purpose machineapplications. Velocity pickups have

been used as vibration transducerson rotating machines for a very longtime, and they are still utilized for avariety of applications today.Velocity pickups are available inmany different physicalconfigurations and outputsensitivities.

Theory of Operation When a coil of wire is moved

through a magnetic field, a voltageis induced across the end wires of the coil. The induced voltage iscaused by the transferring of energyfrom the flux field of the magnet tothe wire coil. As the coil is forcedthrough the magnetic field byvibratory motion, a voltage signalrepresenting the vibration is produced.

Signal Conventions A velocity signal produced byvibratory motion is normallysinusoidal in nature. In other words,in one cycle of vibration, the signalreaches a maximum value twice inone cycle. The second maximumvalue is equal in magnitude to the

Sensitivity Some velocity pickups have thehighest output sensitivities of anyvibration pickup for rotating machineapplications. The sensitivity will vary

from manufacturer to manufacturer.The higher output sensitivity isuseful in situations where inducedelectrical noise is a problem. Thelarger signal for a given vibrationlevel will be less influenced by thenoise level. Some velocity pickupswith their sensitivities are listed below:

Sensitivity

STI LCV100

500

mv/in/sec

Frequency Response Velocity pickups willhave differing frequencyresponses depending onthe manufacturer.However, most pickupshave a frequencyresponse range in theorder of 10 to 1000 hz.

This is an importantconsideration whenselecting a velocity pickup for a rotatingmachine application. The pickup's frequencyresponse must be withinthe expected vibrationfrequencies of themachine. Due to thesupport spring for the bobbin., a naturalmechanical resonanceoccurs at the low end of the frequency responsecurve. This resonance iseither damped by the oilcontained within thesensor, or with a shunt



first maximum value, but opposite indirection. By definition velocity can be measured in only one direction.Therefore, velocity measurementsare typically expressed in zero to

peak, RMS units. RMS units may bespecified on permanent monitor installations to allow correlationwith information gathered from portable data collectors.

Another convention to consider isthat motion towards the bottom of avelocity transducer will generate a positive going output signal. In other words, if the transducer is held in its

sensitive axis and the base is tapped,the output signal will go positivewhen it is initially tapped.

Construction The velocity pickup is a self-generating sensor requiring noexternaldevicesto produce

avibrationsignal.Thistype of sensor is made up of threecomponents: a permanent magnet, acoil of wire, and spring supports for the coil of wire. The pickup is filledwith an oil to dampen the springaction.

Due to gravity forces, velocitytransducers are manufactureddifferently for horizontal or verticalaxis mounting. With this in mind,the velocity sensor will have asensitive axis that must beconsidered when applying these

resistor across the coil'sleads.

Calibration No calibration of the

velocity pickup isnecessary, however, onan annual basis thesensor should beremoved from service for a calibration verification.Verification is required because velocity pickupsare the only industrialvibration sensor whichhas internal moving parts

that are subject to fatiguefailure.

This verification shouldinclude a sensitivityresponse versusfrequency test. This testwill determine if theinternal springs anddamping system havedegraded due to heat and

vibration. This testshould be conductedwith a shake tablecapable of variableamplitude and frequencytesting.

Instrument Wire The following table is a partial list of Belden®Cables that should be

used for the instrumentfield wiring. These partnumbers may be crossreferenced to equivalentcables from other manufacturers. The listedcables are polyethylene



connected at the monitor.

Conduit Dedicated rigid conduitshould be provided in all

installations for mechanical and noise protection. The conduitshould be metal, and indirect contact with eachsegment. All metal junction boxes andfittings should be indirect contact with theconduit. With this typeof installation, a single

ground point can beestablished.

To facilitate removal of the velocity pickup, a junction box with a barrier terminal stripshould be located closeto the transducer. Therigid conduit should beattached to the junction

box and the final run tothe pickup can be metalflexible conduit.

Velocity PickupChecklist

1. Velocity Sensor Type

2. Number of Sensors

3. Sensitivity4. Frequency Range5. Correct

Instrument Wire6. Rigid Conduit7. Grounding8. Location(s)



Documented9. Calibration

Check



Reciprocating Compressors

Reciprocating Compressors are utilized in all manufacturing industries. Because these machines are

capable of providing high pressure along with variable loading, they are favored for many gas processapplications. The total quantity of positive displacement reciprocating engines, pumps and compressors far

exceed the number of centrifugal units.

Past studies within the Hydrocarbon Processing Industry (HPI) indicatethat the maintenance costs for reciprocating equipment areapproximately 3.5 times that of centrifugal equipment. Substantialsavings in maintenance costs and an increase in run time may beachieved through basic monitoring of some if not all of the followingReciprocating Machine parameters.

1. Frame Vibration

2. Rod Drop

3. Rod Run-out

4. Crosshead Vibration

5. Main Bearing Vibration

6. Valve Temperature

Frame Vibration

The most important vibration parameter of a successful monitoring program is Frame Vibration. Whenproperly applied, monitoring Frame Vibration will help prevent catastrophic failures. In the event of a failure,the damage to a reciprocating machine can be reduced.

For the greatest benefit, a Frame Vibration Monitoring system should be wired to an automatic machine trip.To decrease the possibility of a false trip, two (2) case mounted accelerometers should be mounted on theframe relatively close to each other. The outputs from the two transducers are signal conditioned, and their trip circuits are "AND" voted. In other words, before a trip is initiated, both of the transducers with their monitors must sense a high vibration level.

The 786A Industrial Accelerometer is highly recommendedfor this application. It’s low frequency response of .8 Hz, a

standard scale factor output of 100 mV/g, and relative lowcost are desired features for this application. Since thistransducer is Piezoelectric (Solid State Design) based, it isless susceptible to cross axis signal distortion and wear commonly a problem with other Velocity Transducers.

The two transducers should be perpendicular to the shaft, and oriented in the direction of the Piston travel.The transducers must be mounted on a surface that allows direct transmission of the Frame's Vibration. Thetransducers must be mounted on a surface that allows direct transmission of the Frames Vibration. Avoidmounting the transducers on a cover or door.



The 786A Industrial Accelerometer should be interfaced to a CMCP530A Velocity Monitor. This combinationoffers numerous advantages over an OEM supplied “EARTHQUAKE” Switch.

1. Low Frequency Response

2. Solid State Design (Long Life)

3. Indication of Vibration Levels

4. Warning and Shutdown Alarms

5. Shutdown "AND" Voting

6. Fault Detection

7. Vibration Diagnostic Capability

Monitoring the Frame Vibration of a Reciprocating Machine offers the following major benefits:

1. Prevent Catastrophic Machine Failures

2. Reduction of Machine Damage

Rod Drop

The vast majority of Reciprocating Compressors are designed with horizontal Cylinders and Pistons. This isprimarily due to foundation requirements and the popularity of opposed-balanced machine designs.

The force of gravity causes the Piston to "RIDE" more in the bottom of the Cylinder than in the top. In turn,this causes the Piston to wear more in the "DOWN" direction. Machine manufactures provide wear or rider

rings to provide a replaceable wearing surface. For lubricated Cylinders, glass embedded Teflon may beused. For non-lubricated Cylinders, Teflon may be used.

The wear or rider rings are allowed to wear sacrificially. They are rotated or replaced before damage to theCylinder lining occurs. There are several methods used to determine when to replace or rotate the rings.One method is to operate a new machine for a given number of hours or days. Then a valve is removed,and the wear is measured by using a feeler gauge. A calculation is then performed with this information. Theresults determine the length of time the machine can be safely operated with periodic inspections of therings. Obviously, this is a very frustrating method of performing preventative maintenance.

Currently, one popular safety device for detecting Rod Drop is a unit mounted under the rod at a gapdetermined by the allowable wear of the wear ring. When the rod contacts the safety unit white metal isworn through allowing instrument air to escape. This in turn causes a pneumatic flag on the control panel to

change status.

There are several disadvantages to the above-mentioned methods of Rod Drop detection:

1. A real trend of ring wear cannot be established with a short amount of operating time.

2. Since the machine must be shut down, halting production, periodic inspections for ring wear areexpensive.



3. A change in processed gas, load changes, and foreign matter can cause an extreme change in ringwear rate.

For several years, Eddy Probe systems have been utilized to measure Rod Drop. This method of Rod Dropmeasurement has been gaining positive recognition with Reciprocating Machine users. This is especially

true on larger machines, or when the customer has become frustrated

with the previously mentioned methods.

To measure Rod Drop with an Eddy Probe system, the probe isinstalled in the vertical direction viewing the rod. The preferredinstallation would have a probe bracket adapted to the packing glandplate, mounted internal to the distance piece. As an alternate solution,some users have used the CMCP801 Eddy Probe Housing, providingan external adjustment (through the distance piece) of the probe gap. As the Eddy Current field emitted from the probe tip will penetrate therod surface 15 mils, it is important that the observed rod behomogenous in nature and free of any surface irregularities. The EddyProbe system is interfaced to a CMCP545 Position Transmitter tomeasure the probes DC output (Probe Gap). . The CMCP545 will

provide a 4-20 mA output that is proportional to the DC Gap Voltage. If a CMCP545A Monitor is used, twolevels of alarms with corresponding Alert and Danger relays are provided. By trending the DC Gap voltagesfrom the eddy probe, it is possible to measure the average horizontal running position of the piston rod. Thismethod of Rod Drop measurement offers advantages over the previously described methods:

1. An immediate trend of ring wear can be established.

2. The periodic inspections that require a machine shutdown and disassembly are eliminated.

3. Wear rate changes can be observed.

4. Both Warning and Shutdown alarms can be provided.

Monitoring the Rod Drop of a Reciprocating Machine using an Eddy Probe offers the following benefits:

1. Prevents Cylinder and Piston damage caused by the Piston contacting the liner.

2. Stops unnecessary periodic inspections that require a machine shutdown with the associated lostprocess time.

3. Scheduling down time to replace or rotate wear rings within the limitations of a plant's schedule.

Rod Run Out

Whereas Rod Drop is a measurement of rod position, Rod Run Out is a measurement of the rod's actualdynamic motion as it travels back and forth on its stroke. Another term for this measurement is RodDeflection.

One method to make this measurement is to mount a dial indicator in the distance piece riding on the pistonrod. The machine is then barred through a complete cycle. Indicator readings are taken in both the verticaland horizontal directions during the machine's cycle.

The amount of Rod Run Out is highly dependent on the cylinder alignment with the Crosshead. Due toinherent looseness in the Crosshead and thermal growth of the machine, higher readings of Rod Run Out



4. To reduce effects on plant production.

Crosshead Vibration

The Crosshead of a Reciprocating Machine is made up of several major components: Crosshead Bed,Crosshead, Slippers and Crosshead Pin. The purpose of the Crosshead is to transform the circular motion

of the crankshaft into linear motion for the rod and piston.

The Crosshead slides on a lubricated babbitted surface much like a standard journal bearing. However, theCrosshead slides back and forth instead of in a circular motion like a shaft. Clearance between theCrosshead and the babbitt surface may be in the range of 10 to 25 mils. As crankshaft rotates, theCrosshead is driven to slide on either the upper or lower babbitt surface. As the clearance between theCrosshead and babbitt surface increases, the Crosshead vibration increases.

In a compromise to measuring both vertical and in-line with the cylinder, experience has shown that a 786A

Industrial Accelerometer, mounted on a 45 angle block, located in-line with the piston travel, will adequatelymeasure Cross Head vibration. The 786A Accelerometer would be connected to a CMCP530(A) VelocityTransmitter (Monitor).

By processing the vibration signal in Peak Acceleration instead of RMS detection, we can measure highamplitude short duration “peals” or “events” that appear periodically. The Cross Head mounting location in -line with Piston travel will see the mechanical transfer of energy caused by impacts resulting frommechanical looseness on the compressor cylinders, such as loose rod nuts and loose bolts. Liquid in theprocess can also be detected. Competitive offerings may refer to this measurement as Rod Impactmonitoring.

Crosshead Pin lubrication can also be diagnosed with this measurement. The Crosshead Pin connects theconnecting rod from the crankshaft to the crosshead. This pin is force lubricated through "reversal" whichallows oil between the pin and its bushing. With each stroke, the oil is forced out. If reversal does not occur,the pin and its bushing will fail rapidly.

Monitoring the Crosshead Vibration with an Accelerometer offers the following benefits:

1. Assures that the Crosshead to babbitt surface clearance is within acceptable limits.

2. By analyzing the vibration waveform, Crosshead Pin reversal can be confirmed.

3. Machine shutdowns for repair can be scheduled.

Main Bearing Vibration

Main Bearing Vibration has not been proven to be a popular approach for monitoring ReciprocatingCompressors. Several end users have had problems with broken crankshafts, which they thought werecaused by unusual bending of the crankshaft. In one documented case, machinists had over tightened thedrive belts powering the cooling fan on a reciprocating engine. This caused unnecessary bending of thecrankshaft.

Currently, several end user’s have installed X and Y Eddy Probe systems on the main bearings of large12,000 HP Reciprocating Compressors. This installation very nearly resembles that of a standard centrifugalcompressor. However, both probes view the crankshaft from the bottom bearing cap-mounted 90

oapart. As

the lubricating oil cools the main bearing, no unusual measures needed to be taken on this installation.



The Eddy Probes are connected to a CMCP540 Vibration (Displacement) Monitor to measure radialvibration. The full-scale range of the monitor is based on the bearing clearance. As in all Reciprocating

Compressor applications, the entire bearing clearances are used.

A cost effective compromise in lieu of installing X, Y Eddy Probes is tomount a 786A Industrial Accelerometer in-line with the crankshaft

centerline and interface it with a CMCP530(A) Velocity Transmitter (Monitor).

Valve Temperature

According to industry studies, valve failures account for 41% of theproblems associated with reciprocating machinery.

In a Reciprocating Compressor, the valves are a pressure actuated"Poppet" variety. Every machine manufacturer has favorite types of valvesfor different applications. These valves operate utilizing a delta or

differential pressure technique. The opening and closing of a valve occurs when the delta pressure is lessthan the force of their return springs.

When a valve begins to fail, it usually begins to leak the process gas. This causes the process gas to be re-compressed, further heating the gas. This higher temperature process gas can be detected using atemperature transducer. This transducer can be mounted through the valve cover plate measuring the gastemperature near the valve. In some installations, the transducer is simply imbedded in the valve cover plate, or within a valve cover plate bolt. This mounting method is preferred when the process gas isexplosive, which is usually the case.

The measured temperature of the process gas is thencompared to the measured temperature of the processgas at the same type valve, suction or discharge, and thesame stage of compression. Measured temperaturedifferences of 4 to 20

oF can indicate a problem with a

valve.

Monitoring the temperature of the process gas at eachvalve offers the following benefits to the customer:

1. Provides an early indication of a problem with avalve reducing the possibility of machinedamage.

2. If a problem with a valve occurs, only the bad valve needs replacement increasing machine runtime. This approach does not limit the life of all the valves to the shortest life of a valve.

3. Trending the valves' conditions can allow for scheduled machine shutdowns.

2. Help prevent damage to the cylinder lining that occurs when valve parts are ingested by themachine.

Each RTD is connected to a CMCP560(A) RTD Temperature Transmitter (Monitor).

Checklist



1. Target Material (Rod Drop & Run Out)

2. Rod Drop Full Scale Range

3. Machine Speed/Low Frequency Transducers

4. Transducers Mechanically Protected

5. Flexible Conduit (J-Box to Transducer)

6. IMC Conduit (J-Box to Monitor)


8. Instrument Wire Shielding Conventions

9. Calibration

10. Eddy Probe Gap Set

11. CMCP500 Series Monitors

About the Authors

Frank Howard, Jr. Mr. Howard founded Sales Technology, Inc. (STI), a manufacturer's representative of industrial electronic products, in March 1989. He is currently serving as CEO and President of STI. From1976 through 1989, Mr. Howard worked for Bently Nevada Corporation, an international supplier of high-speed electronic monitoring devices, with his last position being that of acting Vice President of Sales.

David Gallagher is Vice President of Sales Technology, Inc. (STI), and CEO and President of Reliability

Direct, Inc. He has spent the past 23 years in the field of Rotating Equipment, and has authored manypapers and conducted numerous seminars on predictive technologies and machinery reliability. He is amember of the Vibration Institute and ISA, and holds a BS degree from Flagler College.

Cooling Tower Fans

Cooling Tower Fans have applications in all industries. They remove heat from other materials, usually water. A cooling tower may consist of one cell or many individual cells ina single structure. A s ingle plant may have multiple structures. Cooling Tower Fans comein two basic types. The first type has the motor mounted to the side of the cell and uses a

jackshaft to drive the gearbox in the center. The second type has both the motor and

gearbox centrally mounted in the cell. The mechanical components of a Cooling Tower Fanare made up of, Motor, Jackshaft (Optional), Gearbox, and Fan Blades.

The motor speed for a Cooling Tower Fan is usually 1800 RPM. Fan speeds are much slower and determined by the diameter of the bladesto keep the blade tips subsonic. Average cells are 14 feet to 28 feet in diameter. Fan speeds of 90 to 230 RPM are normal for cooling tower fans. In cooler climates, the fans can be reversing to prevent freeze during the winter.

The most common Cooling Tower Fan failure involves the gearbox or fan blades and are catastrophic in nature. In many cases, this type of failure leaves the gearbox and fan blades lying in the cooling water pond at the bottom of the tower.



Many Cooling Tower Fans were equipped by the OEM (Original Equipment Manufacture) with a earthquake detection type devices mountedclose to the cell. Although questionable to start with for this application, over time most of these devices have quit working due to corrosion or neglect.

Permanent monitoring is the recommended solution for preventing cooling tower cell failures. Periodic measurement programs rarelyproduce the desired result. In most cases, the location and environment discourages any attempts to take data or visually inspect the fan.

Permanent monitoring of a cooling tower fan cell requires direct monitoring of the most important component by mounting a transducer directly on the gearbox. Additional system protection can be added by mounting transducer(s) to the motor. Additionally, knowledge of theexpected vibration frequencies is important.

1. Fan Speed 90-300 RPM2. Blade Pass (Fan Speed X # of Blades)3. Gear Mesh Frequencies4. Motor Speed5. Bearing Frequencies

Gear Vibration

The central component of a cooling tower is the gearbox. This is often a high maintenance source due to aerodynamic loading from the fan,excessive loading on the gear teeth, and improper alignment of the gear to the motor.

The gearbox is always located within the cooling water stream requiring special consideration for mounting a vibration sensor. Theenvironment is usually caustic due to chemicals added to control the pH level of the cooling water.

A vibration sensor must be able to measure the expected gear mesh frequency, blade passage, and bearing defect frequencies. Other frequencies of interest include fan balance and motor alignment.

Gear and bearing defect frequencies tend to be the highest frequencies, while the fan balance frequencies are the lowest to be monitored.The other frequencies are found scattered between these extremes.

Monitoring the gearbox can be accomplished by installing a single low frequency accelerometer (SKF CMSS793V) on the gearbox in ahorizontal orientation perpendicular to the jackshaft. The output signal should be routed to CMCP500 Series Monitors. One monitor shouldbe specified to measure velocity for higher frequencies like gear mesh. The second monitor should be specified to integrate the signal todisplacement terms for lower frequencies such as blade speed and blade pass.

(1) CMSS793V Accelerometer 100mV/In/Sec Output.(1) CMCP692-32-01-02-01 Armored Extension Cable.

(1) CMCP200-01 Mounting Pad 1/4"(1) CMCP230 Mounting Stud(1) CMCP220 Acrylic Adhesive Bypac(1) CMCP530-100V-02P-00-00 Velocity Monitor (1) CMCP535-100V-30-00-00 Displacement Monitor

Motor Vibration

Motor vibration frequencies of interest include motor unbalance, rotor bar defects, output shaft alignment, and bearing defect frequencies. Acomplete continuous monitoring approach should include one accelerometer per bearing location. Mounting orientation for theaccelerometers should be horizontal at the bearing. As the motor speed is usually 1800 RPM economical standard accelerometers such asthe SKF=s CMSS786 can be used. A more economical approach would be to mount a single accelerometer at the motor output shaft bearinglocation only.

(2) CMSS786 Industrial Accelerometer

(2) CMCP692-16-01-02-01 Armored Extension Cable(2) CMCP200-01 Mounting Pad 1/4"(2) CMCP230 Mounting Stud(2) CMCP220 Acrylic Adhesive Bypacs(2) CMCP530-100A-02P-00-00 Velocity Monitor

Start Up Considerations

All CMCP500 Series Monitors are provided with relays, reset terminals and a trip multiply function. Cooling Tower Fans experience largeamounts of vibration during start-up. Either timer logic or a contact to energize trip multiply needs to be provided if the monitors are wired toshutdown.



DCS/PLC Systems

The CMCP500 series monitors can be equipped with Modbus or TCP/IP communications. I/O can be specified to provide operators withvibration levels, alarm set points, discrete alarms and the ability to energize the trip multiply and reset functions. A complete system usingModbus or TCP/IP can be specified providing operators with screens and software.

Predictive Maintenance Program

A QualityRotatingMachineryPredictiveMaintenance(PM) Programfor your facility will

return your investment many timesover. This type of PM Program is based on periodic vibrationmeasurements of your RotatingMachinery. In many cases, a ReturnOn Investment (ROI) of less thanone year (six months typical) is quitecommon. A Contract PM Programeliminates the initial capitalinvestment required to conduct your own PM Program.

There are three classifications of machinery maintenance methods:Breakdown, Preventative, andPredictive Maintenance. Eachmethod has its own associated costsand benefits.

Breakdown Maintenance, by its ownnature, is the most expensive methodof plant maintenance. This method

has no scheduled maintenance until amachine destroys itself, and it must be replaced at great cost. Themachine breakdown often brings the production process to an immediatehalt. Breakdown Maintenance hashigh costs in manpower, replacement parts, and lost production.

1. Review machinery andlayout.

2. List all applicable machinery.3. Review current and past

problems.4. Determine machinery

classifications.5. Determine the number of

data points.6. Determine time needed to

collect data.7. Determine data collection

schedule.8. Start collection of the

information required tocomplete a ProgramJustification.

After all the information has been

acquired to complete a "PM ProgramJustification Form", a complete proposal will be sent to you for approval and comment.

PM Program Implementation A PM Program can be implementedin several ways. One of which will be suitable to your facility, currentknowledge, and budget. You maychose to make a capital investment

for the hardware, software, andtraining, or a complete MachineryPM Program may be contracted fromSTI. A contract PM Program reducescapital investment to zero, and permits the PM Program to start in atimely fashion. When you are readyto acquire the appropriate hardware,



Preventative Maintenance, the nextlogical method, relies on a periodicinspection the machines. During theinspection, machine damage is foundand corrected. This method requires

a large inventory of replacement parts prior to the machine'sinspection. PreventativeMaintenance has a lower associatedcost because manpower can be planned in advance.

Predictive Maintenance involvesmonitoring the machine's vibrationcharacteristics or symptoms todiagnose its condition. This method

relies on the machine's condition toaccurately schedule the repair interval. The machine's conditionalso determines the requiredreplacement parts. PredictiveMaintenance has the lowest cost of the three methods with the highest possible savings.

A Machinery PM Program is beneficial to all industries that have

rotating machinery on their site suchas:

Petroleum Chemical

Power Generation Co-Generation

Pulp and Paper Water Treatment

Building Services HVAC

Mining Food Processing

Typical Rotating Machinery

commonly included in a MachineryPM Program include but are notlimited to the following:

ElectricMotors

Pumps

Fans Gear Boxes

software, and training to proceed onyour own, the complete PM Program package can be turned over to you.

Once you have determined that a

contract PM Program works best for you, an STI vibration engineer or technician will return to your facility.

They will do a complete facilitysurvey to identify machinery, locatedata collection points, possiblydetermine bearings, and formulatethe data collection route. After thisinformation is acquired, they will

complete the following off site usinga state of the art PM ProgramSoftware Package:

1. Program facility andmachinery information intoPM Software.

2. Program appropriate bearingand alarm information intosoftware.

3. Establish trending and

exception list reporting.4. Establish Stage 1 to Stage 4

bearing failure criteria for allmachinery.

Once the survey information has been properly programmed, the STIvibration engineer or technician willschedule a return trip to your facilityto do the following:

1. Collect initial baselinevibration data.

2. Review overall vibration datafor immediate action.

3. Provide a written report.

Scheduled PM Program Once the initial survey and initial



Turbines Compressors

Paper Machines

ReciprocatingMachines

Blowers Machine Tools

Chillers Conveyors

PM Program A RotatingMachinery PMPrograminvolves thescheduledcollection of

vibration data from the variousmachines. All rotating machinery

vibrates, and this vibration willincrease as the equipment wears. The periodically collected vibration datacan be analyzed by a trainedengineer or technician to determinethe cause and the required machineryrepair. An initial machine inspectionis usually done to determine themachinery parts that are subject towear.

PM Program Benefits Quality PM Programs from manyindustries have shown that RotatingMachinery PM Programs will:

Reduce Capital Investment Reduce Machinery

Depreciation Reduce Machinery

Breakdowns Increase Machinery Life Increase Maintenance Staff Productivity

Reduce Dissatisfied/LostCustomers

Reduce Penalties Reduce Unnecessary

Machinery Repairs Reduce Rework

baseline information is programmed,the STI vibration engineer or technician will return on a scheduled basis. For a reliable PM Program,the normal schedule is either once or

twice a month. During these visits,vibration data will be collected alongwith a visual inspection. After thevisit, you will receive the following:

Before leaving the facility:

1. Exception report showingitems requiring immediateattention.

By delivered report:

1. Exception report of itemsrequiring future attention.

2. Report of all machinesshowing categories Stage 1to Stage 4 bearing failure.

Advantages to Contract PM

Programs There are several important

advantages in contracting your initialMachinery PM Program with STI:

1. Immediate Results By usinga highly trained andexperienced STI vibrationengineer or technician,immediate results are usuallyapparent. After several visits,the credibility and reliabilityof the program will be

proven.2. No Capital Investment A

quality PM program requiresstate of the art DataCollectors, Software,Computer Hardware, andTraining. No capitalinvestment is required for a



Reduce Scrap Reduce Warranty Claims Increase Credibility and

Reliability Reduce Overtime

Increase Safety and ReducePenalties

Reduce Injuries Reduce Power Consumption Reduce Spare Parts Inventory Reduce Defects on New

Machinery Reduce "Wrong" Repairs Reduce Insurance Costs

PM Program Survey

To see if a PM Program is beneficialto you, STI will survey your facility.STI will then help you complete a"PM Program Justification Form".This no-charge survey will include:

contract PM Program.3. No Training Required

All of STI's engineers andtechnicians have many years of

experience in the field of machineryvibration. Therefore, your maintenance personnel do notrequire training to initiate a PMProgram. If you desire, your maintenance personnel canaccompany STI personnel duringtheir surveys and learn how a PMProgram works.

PM Program Justification

A. Direct Machinery Costs 1. Labor

a. Regular Labor for Unscheduled Repairs _______ Hr. x ______ $/Hr. = $ _________

b. Overtime Labor for Unscheduled Repairs _______ Hr. x ______ $/Hr. = $ _________

c. Regular Labor for Avoidable Repairs _______ Hr. x ______ $/Hr. = $ _________

d. Overtime Labor for Avoidable Repairs _______ Hr. x ______ $/Hr. = $ _________

Total Labor Costs (1) $ _________

2. Parts and Materials c. Good Parts Replaced $ _________ d. Premium Cost Parts $ _________ e. Replacement Machinery $ _________



Total Parts and Materials (2) $ _________ Total Direct Costs Labor/Parts and Materials (1 + 2) $ _________

B. Indirect Costs 0. Lost Production

_______ Hr. x ______ $/Hr. = $ _________ 1. Outside Repair Services $ _________ 2. Excessive Parts Inventory $ _________ 3. Cost to Maintain Standby Equipment $ _________ 4. Excessive Insurance Costs $ _________ 5. Out of Specification/Scrap Material $ _________

Total Indirect Costs $ _________ Total Potential Cost Reduction Direct + Indirect Costs$ _________

C. STI PM Program Costs

0. PM Program Survey $ _________ 1. Initial Setup and Baseline $ _________ 2. Scheduled Data Collection

____Visits/Yr. x ______ $/Visit = $ _________

Total PM Program Costs for One Year $ _________

D. Summary 0. Total Direct and Indirect Costs (From Above) $ _________ 1. Machinery Maintenance @ 35% of Total Potential Reduction (Line 1 x 0.35) $

_________

2. Savings generated by 50% Reduction of Machinery Maintenance (Line 2 x 0.50)$ _________ 3. STI PM Program Costs (From Above) $ _________ 4. Annual Savings (Line 3 - Line 4) $ _________

PAYBACK 5. (Line 4 / Line 3) Years _________

PM Programs Part-1

This application note will provide a basis for selection of the machinerylist of a Predictive Maintenance(PM) data collector program.

Creating a PM machine list is alogical process that will provide asmooth flow for data collection.

After the final list has been generatedthe next step is to decide upon themeasurement points for eachmachine.

MEASUREMENT POINTS

Measurement points are the actual



Selection of the eligible machineswill determine the size of thedatabase. The measurement pointsfor each machine will define thelocations at which data will be

collected. Measure ment parameterswill characterize how the data will be collected and trended.

CLASSIFICATION CRITERIA

Generating, the list of machines toinclude in the PM program shouldget considerable attention. Theimmediate decision might be toinclude everything in the plant.

Indiscriminately including allmachines in the plant could result inunnecessary data collection. Certainmachines may not require datacollection if they are only operatedinfrequently such as stepper motorsor standby equipment. Thesemachines should be monitored usingspecially developed testing programs.

Any machine which is operatingcontinuously should be a candidatefor the machine list. The machinelist, which will most likely begin on paper, should include the machinename, horsepower, bearing type(s), plant location, and operating speed.Optionally, other items may beadded to this list, such as bearingoperating temperatures, fluid flowrate, pressures, etc. If available, the

maintenance and economic historyof each machine should be gathered.

After this information has beenassembled the machine list can besub-divided into three categories.These categories are critical,

locations at which data is to becollected. Each machine should haveall of its bearings included in the listof points. Generally, each bearingshould have vibration data collected

in three directions: vertical,horizontal, and axial. Due tomachine construction and other considerations, the axial vibrationdata may not need to be collected ateach bearing.

An often overlooked supplementaryinput which can be incorporated intomany databases is inspection codes.The human senses of the data

collector operator should not beoverlooked, since the operator isworking at the machine. An operator will observe certain situations whichwill not be identified by a vibrationmeasurement. They are usuallyentered manually during the datacollection process. Inspection codescan reveal important informationabout the condition of the machinewhich the vibration data could

overlook.

Finally, now that the variousmachines and the measurement points are defined the only remaining bit of information needed is howeach machine will be analyzed. Thisinvolves defining the measurement parameters for each measurement point.

MEASUREMENTPARAMETERS

At each measurement point certaintypes of data must be collected.Vibration data may be collected indisplacement, velocity, acceleration,or spike energy units. Machine



essential and general purpose.

Critical should be candidates for permanent instrumentation due totheir production impact should they

fail. Economic factors, such asreplacement cost or maintenanceexpenses, should be considered to justify permanent instrumentation.These machine should always bemonitored.

Essential machines usually providemajor support for the production process and may be partially spared.These machines should be

monitored, but due to their supportfunction may be candidates for permanent instrumentation. If theyare not permanently monitored theyare candidates for the machine list.

General purpose machinery willinclude all other machinery in the plant. This category will most likely be a much larger list than the other categories and quite possibly

consume the bulk of the maintenance budget. These machines aredefinitely candidates for the machinelist.

Each category of the machine listcan be further sub-divided intomachine which operate nominallyand those which are troublesome.The troublesome machines should beimmediate candidates for the final

machine list. They are goodcandidates to "cut your teeth" on,since the probability of improvements in operation are quitehigh once corrective action is takenthat is based upon collected data.These machines should occupy themajority of the newly started PM

construction and operatingconditions will dictate which type toinclude.

Most data collector based PM

programs are implemented onmachinery with rolling element bearings. These bearings willgenerate specific frequencies relatedto the bearing condition. Many datacollectors offer a measurement parameter called spike energy whichcan provide early detection of bearing defects. Thus, machines withrolling element bearings should havespike energy data collected along

with at least one of the other vibration data types.

Certain machines have couplingswhich transmit axial vibration fromone case to another and may notneed to have axial vibration datacollected at each bearing. Processvariables, such as temperature, pressure, or flow rate, may be addedto the measurement parameters list

where provisions have been madethat allow the data to be inputautomatically by the data collector.

In addition to bearing condition, proper measurement parameter selection will provide informationabout the balance state, alignmentcondition, and the general conditionof the machinery internals.

Criteria Checklist

1. Machine Classification2. Measurement Points3. Measurement Parameters



program with weekly or bi-weeklymonitoring cycles.

General purpose machines which arenot troublesome and essential

machines, along with theunmonitored critical machines, can be listed for routine monitoring on amonthly cycle. When thetroublesome machines are operatingnominally they can be monitored ona monthly cycle.

PM Program Part-2 Routes

This application note will provide a basis for creation of a PredictiveMaintenance (PM) data collector route.

Establishing an efficient PM datacollector route is a logical processthat will provide a smooth flow for data collection. A haphazardlycreated route can result in improper

data collection, missedmeasurements, duplicatedmeasurements, and excessive datastorage requirements. Detailedinvestigation of the plant layout andmachine construction will berequired to create the machine route.

This application note assumes that amachine list has been developedwhich defines each machine, the

measurement points for eachmachine, and the measurement parameters for each point. See theapplication note "PM Program - Part1" for information regardingmachine list development.

ROUTES

It will, also, define the time requiredfor completion of the data collectionfor each route. These criteria maydictate creation of a new route basedupon the machinery access or timerequired for data collection. Now isthe time to massage the route(s) before implementation of the datacollection process.

EXAMPLE

The diagram shows a route whichstarts in the pump room, progressesto the compressor room, and finishesin the fan room. Route constructionoften involves clustering themachines into groups based upontheir locations relative to each other.Other items such as enclosures,access panels, or piping may

influence the route layout. After theroute has been created a practicewalk through may identifyadjustments that may be required.

Data collection in the pump room begins with Pump #3. Due to the proximity of Pump #2 and Pump #3



Once the list of machines, themeasurement points, andmeasurement parameters have beentabulated a data collection route can

be created. A plant layout diagramwill be necessary to determine the progression from machine tomachine and point to point.

A plant with a large machine listmay be split into several routes.Each route will be a logical progression from measurement pointto measurement point. The progression may not necessarily flow

from one machine to the next in aconsistent way. Certain routes mayhave the data collected based uponthe proximity of the measurement points, while other routes may becreated which encompass machineson a single floor or part of the plant.

When a route has been developed itshould be walked through. Thismeans following the route

instructions exactly, progressingfrom each point to the next. Thiswalk through will determine whether the route is efficient for datacollection or requires slightmodifications. It will identify other requirements such as required accessclearances or ladders needed to getto each measurement point.

the data collection progress from one pump to the other with Pump #1 being the last machine to monitoredin the pump room. Notice that for Pump #3 and #2 the data is collected

from the pump towards the motor while the data is collected in thereverse order for Pump #1.

Also, notice that the motor inboardmeasurement of the compressor machine train was taken on the southside instead of the north side. If nointerfering structures were presentthis reading could have beenobtained on either side, whichever

side has the easiest access.

Data collection in the fan roomfollow the same sequence for eachfan with this route ending at thesouth exit door of the fan room.

Route Checklist

1. Machine List2. Route(s) Developed

3. Practice Walk Through



Reverse Dial Alignment

Mis-alignment can be the most usualcause for unacceptable operation andhigh vibration levels. New facilities or new equipment installations are often plagued by improper alignmentconditions.

Many methods are used to alignmachinery. The simplest method isusing a straight edge to bring themachines into rough alignment. Arough alignment is necessary due tothe range limitations of the dialindicators. A popular method used for years and is still in use today is therim and face method. This method produces acceptable results, but is lessaccurate than the reverse dial indicator

method covered in this applicationnote. Additionally, reverse dialalignment does not require removal of the coupling to collect data.

WORKING WITH DIAL

INDICATORS

Choosea graphscalelargeenoughto provide

sufficientaccuracy

when all calculations are completed.Lay out the physical dimensions of the machine train and show thecollected data on the graph for completeness.

On thegraph

show thelocationsof themachinefeet, thedialindicator sweep plane, and the



Dial indicators are available in many physical sizesand ranges. For most alignment

applicationsthe smaller sized indicatorsshould be usedto reduce indicator bar sag. Dialindicators should be chosen that havea range of 0.100 inch and accurate to0.001 inch.

Indicator readings,

and manyother typesof readings,are

expressed in several units. A readingof 1/1000" is equivalent to 0.001 inchand is commonly expressed as 1 mil.

A common convention used whenreading dial indicators is that whenthe indicator plunger is moved toward

the indicator face the display shows a positive (+) movement of the dialneedle by sweeping the needleclockwise. As the plunger is strokedaway from the face a negative (-)reading is displayed by sweeping theneedle counterclockwise. Negativemovements of the dial needle may beconfusing if the indicator is notobserved carefully throughout therotation cycle of the machine shafts.

Another convention to employ is thatwhen all readings are recorded, theyshould be interpreted recorded byviewing from the stationary machineto the moveable machine. Thisconvention is necessary to distinguishright from left readings during data

location of the power transmission points. Note that for this example, the power transmission points do notcoincide with the sweep path of thedial indicators.

Whenactually plottingtheverticaloffsets,start withthestationar

y set first. Remember that the data

collected represents twice the actualdifferences between the shaftcenterlines. A positive result is plottedabove the line representing thestationary machine centerline. When plotting the moveable machinereadings a rule of thumb is thatopposite signs are plotted on the sameside of the stationary machinecenterline on the graph.

For simplicity's sake, the horizontalreadings are plotted separately in thisexample. A better approach is tocombine the vertical and horizontal plots ontoa single plot.

Just as was performedfor the

vertical plots, thehorizontal plot has a scale reference, the fieldcollected data, and other importantdimensions shown for completeness.Also, note that the stationary machinecenterline is labeled right and left.



collection process and will be appliedwhen the calculations and graphs aremade to decide upon the actual movesrequired at the moveable machine.

PRE-ALIGNMENT CONDITIONS

Prior to any alignment activity anextensive list of items must bechecked to ensure acceptable results.The obvious item involves ensuringthat the machine shaft axes areroughly aligned within 50 mils. Other inspections should include:

1. The foundation base plate is

adequate and that grout has been installed properly.

2. All machine feet are in fullcontact with the foundation baseplate or supports.

3. Piping is not inducing strainonto the machine cases. Pipingshould be aligned to within "of their flanges.

4. The machine feet bolt holeshave enough clearance to

conduct alignment activity.5. The coupling faces are axially

aligned; the axial spacing between the coupling faces ascorrect.

6. Coupling radial runout is lessthan 2 mils.

7. Coupling face runout is lessthan ½ mil.

8. The moveable machine has asinitial shim pack installed; the

shim pack should becomprised of one thick spacer and one or two smaller shims.Too many shims will act like aspring causing additional problems.

9. The required amount of coldoffset compensation is

Thesame processof plotting

the dataisfollowed, but

note that the field collected data has been adjusted so that the right sidereadings have a magnitude of zero tomake plotting easier. The oppositesign rule of thumb also applies for horizontal plotting.

After all the data has been plotted, therequired corrections of the moveablemachine can be obtained directly fromthe graph. Obtaining this informationin this manner eliminatesmathematical errors possible usingother methods. Graphical presentations allow experimentationand study of many possibilities for

correctingalignment.

For thisexample,the motor outboardfeet needto belowered 17mils and

the inboard feet must be lowered 12.5mils for a perfect alignment with the

pump.

This alignment condition assumes thatthe pump does not thermally grow asit operates. If some thermal growth isanticipated, then this information can be plotted on the graph as an offset of the stationary machine's centerline



available; this may beavailable from the machinemanufacturer.

10. The alignment bracketry isappropriate to the activity

required.

MOUNTING DIAL INDICATORS

Many commercially available reversedial indicator alignment kits havemodular bracketry which canencompass the majority of applications. However, someapplications will require a custom bracket. Regardless of the type of

bracket used, the amount of bracket or bar sag should be documented so thatthis information is available to beincluded in the calculations.

Bar sag can be easily documented byinstalling the bracket and dialindicator, intheidentica

larrangement to be used on the machine, ontoa pipe. Zero the indicator while it ison top of the pipe. Now rotate the pipe180 so that the indicator is at the bottom position. The indicator willnow display twice the amount of bar sag.

Mounting the indicator onto the

bracketry should be performedcarefully so that the indicator plunger axis is perpendicular to the machineshaft axis to ensure accurate readings.An error of only 10 will produce a16% error in the indicator reading.

TAKING READINGS

and appropriate moves of themoveable machine can be obtained by projecting the stationary machine'scenterline over to the moveablemachine's location.

The horizontal movements for thisexample are 35.5 mils to the left at themotor outboard and 16.5 mils to theleft at the inboard motor feet.

Graphically plotting the results makesthe movement computations easier, but prior to any moves the followingtopic "ACCEPTANCETOLERANCE" should be considered

because the existing alignment may beacceptable. Re-alignment of amachine with acceptable alignmentconditions is a waste of time and isonly a practice exercise at best thatmay produce a worse alignmentcondition.

ACCEPTANCE TOLERANCE

Determining whether the existing

alignment condition is acceptable or the actual machine moves resulted inan acceptable

alignment condition can be quantified by referencing the chart at the end of this application note. This chart may be applied to all machine andcoupling types. The chart takes intoaccount the coupling span and themachine operating speed.



After the bracketry is firmly attachedand the dial indicators are installed,four reading locations are required.These

locationsare alongthecircumference of theshaft or coupling inthe path of theindicator plunger. They are top, bottom, right, and left. These location

are to be separated by 90 of shaftrotation. Marking these locations withan indelible marking pen is adequate.Another approach is to use a commontwo axis trailer level attached to acoupling face or other surface todetermine when the shaft has beenrotated 90 . Placing four pieces of tapeequally spaced around thecircumference of the shaft will work,as long as the tape is not in the path of

the indicator plunger.

Before any readings can be taken thedial indicators must be set. A simpletest of rotating the machine shaftthrough an entire 360 sweep willverify that the indicator plunger tip isin complete contact with the shaft.When the indicator is at the toplocation the indicator should be resetto display zero. This is accomplished

by rotating the outer bezel of theindicator until the dial face, which isattached to the bezel, shows "0" under the needle.

Collecting the data is simply a matter of rotating the machine shaft in 90increments and noting the dial

The key to applying the chart is todetermine the locations at which the power is transmitted. For gear typecouplings the power transmission points are the gear teeth on each

coupling half. For diaphragm typecouplings the power transmission points are the coupling faces.

The locations of the power transmission points should be notedon the graphical plot. Depending uponthe data collection method and thecoupling type,. the power transmission points may not coincidewith the coupling faces or the dial

indicator sweep path. Following arethe calculations necessary todetermine the alignment accuracy:

Alignment Accuracy = Maximum (X,Y)/DwhereX = ( XV2 + XH2 )½Y = (YV2 + YH2 )½

XV and XH = amount of offset,

vertically and horizontally, at the power transmission point on thestationary machine.

YV and YH = amount of offset,vertically and horizontally, at the power transmission point on themoveable machine.

Maximum (X, Y) = larger of X or Y,calculated above.

Plotting the resultant alignmentaccuracy on the chart will determinewhether the existing alignmentcondition is acceptable or whether the proposed correction moves will produce acceptable results.



indicator readings with their signs (+or -).

If only one dial indicator setup isavailable, the bracketry must be

relocated to the other coupling or shaft and the sweep should berepeated. Remember, that all readingsshould be collected while observingfrom the stationary machine to themoveable machine to maintain rightand left consistency.

ACCURACY VERIFICATION

Collecting the necessary data is

simple enough, but will be entirelyuseless without some form of accuracy verification. Each time thedial indicator is rotated to the toplocation it should display a reading of zero. If it does not then something hasmoved during the rotation: indicator, bracket, clamping mechanism,machine. Correct the problem and

startover.

Another test,whichcan be performed as thedata iscollected,

is to verify that the sum of the top andthe bottom readings should equal the

sum of the left and right readings.

CALCULATIONS

As the dial indicator is swept aroundthe circumference of a coupling or shaft it displays twice the difference between the projected centerline of

MOVING THE MACHINE

Moving a machine is, in many cases,difficult due to their size and weight.Extremely heavy machines, such as

power plant generators will requirehydraulic jacks. Most other machinescan be moved using jacking screws,which are rigidly attached to thefoundation base plate, and pry bars tolift the machine.

Prior to any horizontal move a dialindicator should be installed tomonitor each foot along one side of the machine for horizontal

movements. Vertical movements willrequire an indicator on each foot on both sides of the machine. Verticallyoriented indicators should be observedas the machine foot bolts are re-torqued. The displayed indicationshould not change by more than 1-2mils, indicating that all feet aresupporting the machine equally.Finally, after the bolts are re-torquedthe jacking bolts should be backed out

so that they do not influence thenatural thermal growth as the machineheats. Other machines, such as gear boxes, turbines, and compressorsshould have dowel pins installed atstrategic locations to control thethermal growth direction.

The best choice for shim material isstainless steel. This material is verystable and is easy to maintain. Carbon

steels should be avoided because itwill rust and eventually compromisethe machinery alignment. Synthetic or plastic shim material should beavoided for industrial applications because it is easily damaged andunder heavy load will deform whichcompromises the alignment condition.



the indicator's attachment point andthe measured shaft centerline. Thisargument applies for both the verticaland horizontal readings.

Thus, the sum of the vertical andhorizontal readings must be divided by two to represent the actualdifferences in the two shaftcenterlines. Remember to observe thesigns of the indicator readings closelyto prevent errors in these calculations.

Two vertical offset numbers and twohorizontal offset numbers will beobtained; one set representing the

readings while the bracketry isinstalled on the original shaft andanother set representing the readingswhile the bracketry is installed on thesecond shaft.

Horizontal calculations sometimes present some confusion because oneside does not start at zero. Adding or subtracting the magnitude of the rightside reading to both sides will force

the right side to zero.

GRAPHING THE RESULTS

Presenting the calculated results in agraphical format will assist invisualizing the required machinemoves. Although any size graph scaleis adequate for this process,

The shims used for industrialapplications should be large enough toadequately support each foot.Commercial shims are available invarious dimensions. These shims are

precut and dimensioned to standardthicknesses which are labeled on asmall tab. These shims are easy toinstall and are difficult to mix up. If shims are manufactured in the fieldthey should be large enough tosupport the machine foot and all edgesshould be smoothed to eliminate burrs. Kinked or otherwise damagedshims should be discarded and newones obtained. The shims, the base

plate surface, and bottoms of themachine feet should be clean and freeof defects prior to installing anyshims.

WHICH MACHINE MOVES?

Generally, the stationary machine hascertain constraints which make inimpractical to move it. Pumps haverigid piping attached, generators have

complex cooling systems, and gear boxes are relatively sensitive to anyorientation other that flat and level.When these machine types are movedthe attached systems must berelocated to eliminate sources of strain.

Multiple case machine trains, such asdual compressors driven by oneturbine, pose another problem. All

three machine shafts must operate co-linearly to function efficiently. Bystudying the graphical plot of thecurrent alignment and the desiredalignment it may prove most effectiveto move the center machine case,instead of moving two or threemachines.



expanding the scale as large as possible will improve the accuracy of the move calculations because themeasured differences of the shaft or coupling centerlines are projected out

to the locations of the moveablemachine's feet.

As the figure shows, two sets of dialindicator readings are collected. Thereadings taken on the stationarycoupling are located above thestationary machine (pump) and thereadings collected on the moveablecoupling are located above themoveable machine (motor). When

plotting these readings start with thestationary readings and then proceedto plot the moveable readings.

Directly plotting the measuredreadings will display a representationof the existing mis-alignment. Thedesired alignment condition can bedrawn onto the graph. The desiredcondition should include any offsetcompensation so that when the

machine train is operating under normal conditions the alignment iswithin acceptable tolerances.

Alignment Checklist

1. Pre-alignment Conditions2. Shim Materials3. Indicator Bracket Sag

4. Graph Materials

Field Application Note

Documents

Transcript of Field Application Note