Vous êtes sur la page 1sur 182

_______________________________________________________________________________________________________________

FAKULTET FOR INGENIRVITENSKAP OG TEKNOLOGI NTNU TRONDHEIM


NORGES TEKNISK-NATURVITENSKAPELIGE UNIVERSITET
FACULTY OF ENGINEERING SCIENCE AND TECHNOLOGY NTNU TRONDHEIM NORWEGIAN
UNIVERSITY OF SCIENCE AND TECHNOLOGY

_______________________________________________________________________________________________________________

TMR7

Experimental Methods in
Marine Hydrodynamics

Sverre Steen

Revised August 2014

MARINTEKNISK SENTER

INSTITUTT FOR MARIN TEKNIKK

MARINE TECHNOLOGY CENTRE


TRONDHEIM, NORWAY

DEPARTMENT OF MARINE TECHNOLOGY

Page1

CONTENTS
1

INTRODUCTION..................................................................................................3
1.1
1.2

GENERALMODELLINGLAWS..............................................................................6
2.1
2.2
2.3
2.4
2.5
2.6

General...................................................................................................................................55
RigidModels...........................................................................................................................55
ElasticModels.........................................................................................................................57

CONVENTIONALSHIPTESTING.........................................................................61
6.1
6.2
6.3
6.4

Generaldescriptionofequipment.........................................................................................28
Strainanddisplacementmeasurements................................................................................28
Positionmeasurements..........................................................................................................32
Accelerations..........................................................................................................................34
PressureTransducers..............................................................................................................35
Velocities................................................................................................................................38
ForcemeasurementsDynamometers..................................................................................41
WaveMeasurements..............................................................................................................42
DataAcquisition......................................................................................................................44
SamplingFrequency...............................................................................................................48
LengthofRecords...................................................................................................................50
Calibration..............................................................................................................................53
Zeroing....................................................................................................................................54

PHYSICALMODELLING......................................................................................55
5.1
5.2
5.3

Introduction............................................................................................................................14
TowingTanks..........................................................................................................................14
CavitationTunnel....................................................................................................................18
OceanLaboratories.................................................................................................................20
Generationofenvironment....................................................................................................21

INSTRUMENTATION.........................................................................................28
4.1
4.2
4.3
4.4
4.5
4.6
4.7
4.8
4.9
4.10
4.11
4.12
4.13

Geometricalsimilarity..............................................................................................................6
Kinematicsimilarity..................................................................................................................6
Dynamicsimilarity....................................................................................................................7
ScalingRatios..........................................................................................................................10
Hydroelasticity........................................................................................................................11
Cavitation................................................................................................................................13

EXPERIMENTALFACILITIES...............................................................................14
3.1
3.2
3.3
3.4
3.5

Background...............................................................................................................................3
Whymodeltests.......................................................................................................................4

General...................................................................................................................................61
Towingandpropulsiontestsintowingtank..........................................................................61
Cavitationtunneltests............................................................................................................65
Maneuveringtests..................................................................................................................67

SEAKEEPINGTESTING.......................................................................................71
7.1

General...................................................................................................................................71

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page2
7.2
7.3
7.4
7.5

OFFSHORESTRUCTURETESTING......................................................................77
8.1
8.2
8.3
8.4

Deliverytrials........................................................................................................................103
Shipmonitoringsystems......................................................................................................110

ERRORANALYSIS............................................................................................113
12.1
12.2
12.3

13

General...................................................................................................................................85
Statictests..............................................................................................................................85
Decaytest...............................................................................................................................85
RegularWaveTest..................................................................................................................88
IrregularWaveTest................................................................................................................91

FULLSCALEMEASUREMENTS.........................................................................103
11.1
11.2

12

Testingoffloatingoffshoreshipsandplatformswithmooringandflexiblerisersystems...83
Testingofhydrofoilships........................................................................................................83
Testingoffloatingoffshorewindturbines.............................................................................83
Challengesinhybridmodeltesting........................................................................................83

ANALYSISOFMEASUREDDATA........................................................................85
10.1
10.2
10.3
10.4
10.5

11

General...................................................................................................................................77
TestRequirements..................................................................................................................77
Deepwaterstructuresrequirements.....................................................................................78
TestProcedure........................................................................................................................80

REALTIMEHYBRIDMODELTESTING................................................................82
9.1
9.2
9.3
9.4

10

TestRequirement...................................................................................................................71
Testsetup..............................................................................................................................72
TestProcedure........................................................................................................................73
Tankwalleffects.....................................................................................................................75

Introduction..........................................................................................................................113
Uncertaintyanalysis.............................................................................................................113
DiscussionofErrorSources..................................................................................................119

MODELTESTSVSNUMERICALCALCULATIONS..............................................127
13.1
13.2

General.................................................................................................................................127
ModeltestsforValidationofNumericalCalculations..........................................................128

14 REFERENCES...................................................................................................130
15 INDEX.............................................................................................................132
ANNEXA ExampleofReportingfromModelTest
ANNEXB ExampleofModelTestSpecification
ANNEXC ViscousSurgeDampingofFloatingProductionVesselMooredatSea
ANNEXD ErrorAnalysisofExperiments.LecturenotebyS.Ersdal
ANNEXE ITTCstandardforpoweringperformanceprediction

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page3

INTRODUCTION
ThiscompendiumhasbeenpreparedforthecourseExperimentalMethodsinMarine
Hydrodynamics.Partsofthenotesarebasedonearlierlecturenoteswithinthisfield;seeHuse
(1999)andWalderhaug(1983).ExtensiverevisionsofthecompendiumwrittenbyAarsnesin2001
weremadebySteenin2004,2005,and2006,followedbysmallerrevisionsin2010and2012.
AlthoughthenameofthiscourseisExperimentalMethodsinMarineHydrodynamics,wewill
mainlybetalkingaboutmodeltesting,sincemostexperimentsinmarinehydrodynamicsaremade
inmodelscale.Also,modeltestinginvolvesmanyinterestingissues,likescalingandmodelling.Full
scaletestingishandledasaspecialcase,seechapter11.
Throughoutthetext,manyreferencesaregiventosupplementaryliterature,anditis
recommendedtoconsultthoseforamoreindepthtreatmentofspecialtopics.Agoodtextbook
thatcoversmostofthetopicsintheselecturenotesatanintroductorybutstillmorethoroughlevel
isthebookbyDunn(2005).

Figure 1.1 Model tests in Peerlesspool in London in 1761

1.1

Background
Experimentalfacilitiesformodeltestingofshipshavealongtradition.Improvedresistance
performanceoftheshipswasearlythemaindrivingforcebehindthedevelopmentofshipmodel

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page4
testing.ItisknownthatLeonardodaVinci(aboutyear1500)carriedouttestswith3models
ofships,allwithequallength,butwithdifferentforeandaftshape.Basedonhisexperimentshe
wasabletogiverecommendationaboutwhichshapegivesthehighestspeed.LaterSamuelFortey
(16221651)alsodidtestswithshipmodelsandin1721EmanuelSwedenborggaveadetailed
proposalforshipmodeltestingintroducingtheprinciplewithfallingweightfortowingofthe
models.Inthiswayhewasabletoachieveaknownandconstanttowingforce.In1761this
principlewasusedinPeerlesspoolinLondonasshownonthepicturegiveninFigure1.1.Atthat
timenoscalinglawswereavailabletopredictfullscalebehaviourandonehadtoassumethatthe
winnerwasthebestalsoinfullscale.
WilliamFroude(18101879)isoftengiventhehonourforthemethodofreallyusingmodeltesting
forshipdesignbythedevelopmentofamethodforscalingfrommodelresistancetotheactualship
resistance.Thismayberight,butseveralotherworksfromthesametimealsocontribute
significantlytothisdevelopment.Theestablishingofthescalingmethodsshouldthereforemorebe
regardedasaresultoftheincreasinginterestandactivitieswithinthisfield.
FroudestowingtankwasbuiltinSouthEnglandinca1870andisregardedasthebeginningof
modernmodeltesting.ThemaindimensionofthetankwasLxBxd=85mx11mx3m.Itwas
equippedwitharailintheroof,whichcarriedthedynamometers.Maximumspeedwas5m/s.
Shortlyafterthistankwasestablished,severalothertankswerebuiltinEngland,Germanyand
elsewhere.ThetowingtankinTrondheimwascompletedin1939withdimensionLxBxd=170mx
10.5mx5m,whichwasanormaltanksizeatthattime.
Later,thedevelopmentwithinshiptechnologyhasinitiateddevelopmentandbuildingof
specialisedfacilitiesascavitationtunnels,manoeuvringandseakeepingbasins.Duringthelast20
25yearstheneedsfromtheoffshoreindustryhavepushedthisdevelopmentevenfurther,and
complexlaboratorieswiththepossibilityoftestingstructuresinrealisticconditionsincludingwind,
currentaswellasmultidirectionalwaves,havebeenbuilt.Anexampleofthistypeoflaboratoryis
theOceanBasinatMARINTEK.
Differenttypesoffacilitiesaredescribedinmoredetailsinchapter3.
Afurtherdescriptionandreviewofthehistoryanddevelopmentofshipmodeltestingcanbefound
inSNAME(1967)andinStoot(1959)

1.2

Whymodeltests
Hydrodynamicmodeltestingwillbasicallyhavethreedifferentaims:
1. Toachieverelevantdesigndatatoverifyperformanceofactualconceptsforshipsand
othermarinestructures
2. Verificationandcalibrationoftheoreticalmethodsandnumericalcodes
3. Toobtainabetterunderstandingofphysicalproblems.
Alltheaimscanbeassociatedtotheoftenverycomplicatednatureofproblemsconnectedtothe
interactionbetweenfixedandfloatingstructuresandthemarineenvironment.
Aim1isspeciallythecaseiftheanalysisisverycomplicatedforwhichverifiednumericaltoolsare
notavailable.Modeltestcanbeusedtoinvestigateeffectsofsimplificationsusedasbasisfor

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page5
analyticalornumericalmodels.Inthiswaymodeltestresultscanbeusedtoassist
developmentofmorereliablenumericaltools.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page6

GENERALMODELLINGLAWS
Physicalmodelsareintendedtorepresentthefullscalesystemascloseaspossibleata(much)
smallerscale.Tobeabletodeterminetheproperpropertiesofthemodelweneedmodellingor
scalinglawsthatensureasimilarbehaviourinmodelandfullscale.
Dimensionalanalysiscanbeusedtoderiveagroupofmeaningfuldimensionlessquantitiesfor
applicablevariables.Thisisparticularlyusefuliftheproblemiscomplex.Typicallyallthevarious
quantitiesassumedtobeofimportanceforacertainphenomenaislisted.Afunctionalrelationship
betweenthedifferentparametergroupsisthenestablishedforallflowgoverningquantities.The
scalinglawsareobtainedbytakingtheratioofthedifferentforces.Adetaileddescriptionof
dimensionalanalysiscanbefoundinTaylor(1974).Aderivationofthemostcommon
dimensionlessvariablesusedinfluiddynamics,usingBuckinghamsPitheoremisfoundinchapter5
ofWhite(2005).
Toachievesimilarityinforcesbetweenthemodelscaleandfullscalesituationthefollowing
conditionsmustbefulfilled:

Geometricalsimilarity

Kinematicsimilarity

Dynamicsimilarity

Inthefollowingtheserequirementswillbediscussed.Amorecomprehensivediscussionabout
modellawsisgivenbyChakrabarti(1998)

2.1

Geometricalsimilarity
Geometricalsimilarstructuresinmodelandfullscalehavethesameshape.Thismeansthata
constantlengthscalebetweenthemexist:

LF / LM

whereLMandLFareanydimensionsofthemodel/fullscalestructure.Therequirementtoequal
lengthratioforalldimensionsdoesnotapplyonlytothestructures,butalsotothesurrounding
environment.Atthefirstviewthisseemstobeaneasyrequirementtosatisfyforpracticaltesting.
Howeverthisneednotbetheactualsituation.Forexampletheactualsurfaceroughnessofaship
cannotbeaccuratelymodelled.Anotherexampleisthealmostunrestrictedextentofthe
surroundingwaterforasailingship(exceptforwaterdepthinsomecases).Thissituationisnot
possibletoreproduceinmodelscale,whichimpliesthatphysicalboundariesalwayspresentin
modeltestingcaninfluencethetestresults.

2.2

Kinematicsimilarity
Theratiosbetweenvelocitiesinmodelscalehavetobeequaltothecorrespondingratiosinfull
scale.Thisimpliesthatflowwillundergothegeometricalsimilarmotionsinbothcases.Asan
exampletheratiobetweentheforwardspeedofashipandtherotationalspeedofthepropeller
hastobethesame:

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page7

VF
VM

nF (2 RF ) nM (2 RM )

VF
V
M JF JM
nF DF nM DM

or

whereVistheshipspeed,nistherateofrevolutionofthepropeller,Risthepropellerradius,Dis
thepropellerdiameterandJistheadvancecoefficient.

2.3

Dynamicsimilarity

2.3.1 Forces
Dynamicsimilarityisachievedifwehavethesameratioatmodelscaleandfullscaleforthe
differentforcecontributionspresentintheproblem.Inprinciplethefollowingforcecontributions
willbeofimportance:
1. InertiaForces,Fi
2. Viscousforces,Fv
3. Gravitationalforces,Fg
4. Pressureforces,Fp
5. Elasticforcesinthefluid,Fe.
6. Surfaceforces,Fs.
Inaddition,forelasticmodelstheelasticrelativedeformationsmustbeidenticalinmodelandfull
scale.
Wewillusethefollowingdifferentphysicalquantitiestocharacterisethedifferentforce
contributions;physicallength;L,velocity;U,fluiddensity;,gravitationalacceleration;g,andthe
fluidviscositycoefficient;.ThefollowingdependenceofthephysicalparametersL,U,,gand
willexistforthedifferentforcecontributions:
InertiaForces:

Fi

dU 3
dU dx 3
L
L U 2 L2
dt
dx dt

ViscousForces:

Fv

dU 2
L UL
dx

GravitationalForces:

Fg gL3

PressureForces:

Fp pL2

ElasticfluidForces:

Fe v E v L2

SurfaceForces:

Fs L

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page8
wherevistherelativeelongation(compression),Evisthevolumeelasticityandisthe
surfacetension.

2.3.2 FroudeNumber
Thedynamicsimilarityrequirementappliedontheratiobetweeninertiaandgravityforcesgives
thefollowingrelation:

Fi
U 2 L2 U 2

Fg
gL
gL3
Appliedonmodelandfullscalethisrequirementgives:

UM 2 UF2

gLM gLF
UM
UF

FN
gLM
gLF

whereFNistheFroudenumber.Geometricalandkinematicsimilarity,andequalityinFroude
numberinmodelandfullscalewillthereforeensuresimilaritybetweeninertiaandgravityforces.
Sincesurfacewavesaregravitywaves,thisimpliesthatequalityinFroudenumbershouldgive
equalityinwaveresistancecoefficient.

2.3.3 ReynoldsNumber
Equalratiobetweeninertiaandviscousforceswillgive:

Fi U 2 L2 UL UL

Re
Fv
UL

whereReistheReynoldsnumberand=/isthekinematicviscosity.EqualityinReynolds
numberbetweenfullscaleandmodelscalewillthereforeensurethattheviscousforcesare
correctlyscaled.

2.3.4 Machsnumber
Theelasticityofwaterwillinfluencethepressuretransmissioninwaterandwillthereforebe
importantforsometypeofmodeltesting.Equalratiobetweeninertiaandelasticforcesgives:

Fi
U 2 L2

Fe v E v L2
Usingthegeometricalsimilarityrequirementthatvareequalinmodelandfullscalethis
requirementgives:

U 2 L2
U 2 L2

2
2
v Ev L M v Ev L F

UM
UF

Mn
Ev , M
Ev , F

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page9
whereMnistheMachnumberand Ev isthespeedofsoundinwater.

2.3.5 Webersnumber
Theratiobetweeninertiaandsurfacetensionforcesisgivenfrom:

Fi
U 2 L2 U 2 L

Fs
L

Similarityrequirementforthisforceratioinmodelandfullscalewillnowgivethefollowing
requirement:

U 2 L
U 2 L

M F

whichgives:

UM

( L) M

UF

Wn

( L) F

whereWnistheWebersnumber

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

2.4

Page10

ScalingRatios
Thefollowingdimensionlessquantitiesarecommonlyusedfortestingofshipandoffshore
structures:
Symbol

DimensionlessNumber

ForceRatio

Definition
UL

U
gL

Re

ReynoldsNumber

Inertia/Viscous

FN

FroudeNumber

Inertia/Gravity

Mn

MachsNumber

Inertia/Elasticity

Wn

WebersNumber

Inertia/Surfacetension

St

Strouhallnumber

fv D

KC

KeuleganCarpenterNumber

Drag/Inertia

U AT

EV
U

TheStrouhalNumberisnotderivedfromaforceratio.fvisthevortexsheddingfrequencyandStis
thenondimensionalvortexsheddingfrequency,whichagaindeterminetheoscillationfrequency
ofthetransverseliftforcesactingonacylinderwithcrossdimensionD.
TheKeuleganCarpenterNumberisdeterminedfromforceratiobetweendragandinertiaforces
forthecasewithoscillatingflowpastacylinder.TistheperiodofoscillationandUAisthevelocity
amplitude.EqualKCinmodelandfullscaleisforexampleachievedifthesameratiobetweenwave
heightandcylinderdiameterisused.
Inpracticaltestingitwillnotbepossibletosatisfysimultaneouslythedifferentscalinglaws.For
exampleshipsandoffshorestructuresareformostpracticalsituationinfluencedbysurfacewave
effects,eitherfromincomingwavesorwavegeneratedbytheforwardspeedormotionsofthe
structure.Gravitationalforceswillgovernthesurfacewaveformation.Thisimpliesthatforthese
conditionsequalityinFroudenumberinmodelandfullscalemustbeachieved.Ifviscousforcesare
importantfortheactualsituation,therequirementofequalityinReynoldsnumbershouldin
principlealsobesatisfied.Thisisnotpossibletoachieve.Theviscousforceswillnotbecorrectly
scaledandinthescalingprocessfrommodeltofullscalethiseffecthastobeevaluated.
OtherpracticallimitationsforachievingequalityinRearemodelsizeandnecessarymodelspeed.
TherequirementofconstantUL(assumingconstant)willformostcasesbeimpossibletoachieve.
Inconventionalmodeltestingofshipsandoffshorestructures,physicalscalingandtestexecution
aremostcommonlycarriedoutbasedonFroudeScaling.TheeffectofdifferentReynoldsnumberis
accountedforbydifferentscalingprocedures.Atypicalexampleisshipresistancetests,where
scalingmethodsforcorrectingforeffectofdifferentReynoldsnumberiswellestablished.Forother
applicationsnoestablishedmethodexistsforaccountingfortheeffectofReynoldsnumber.This
willbediscussedinmoredetailsinchapter12.2.
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page11
AssumingFroudescalingisappliedandgeometricalsimilaritywithscaleratio LF / LM ,
fromtheequalityinFroudenumberwehave:

UM
UF

gLM
gLF

UF UM

LF
UM
LM

Theotherphysicalparameterscannowbederivedfromthedimensionalanalysissfollows:

F 3
MM
M

Structuralmass:

MF

Force:

FF

Moment:

MF

Acceleration:

aF aM

Time:

tF tM

Pressure:

pF

F 3
FM
M

F 4
MM
M

F
pM
M

Theratio F M isincludedtoaccountforpossibledifferenceinfluiddensitybetweenfullscale
andmodelscale(usuallyseawaterinfullscalerelativetofreshwaterinthetesttank).

2.5

Hydroelasticity
Inhydroelasticproblemsthehydrodynamicforcesareinfluencedbytheelasticdeformationofthe
structure.Thisdeformationisgovernedbytheinertiaforcesandelasticforcesinthestructure.The
modellingoftheelasticpropertiesofstructureswillthereforegiveseveraladditionalproblems
comparedtothemodellingofwaveinduceddynamicresponseofrigidstructures.Exampleswhere
correctlyscaledelasticbehaviourofthemodelwillbeimportantisspringingandwhippingofships,
anddynamicbehaviourofmarinerisersandmooringlines.
Additionalrequirementstotheelasticmodelcanbesummarisedasfollows:

Correctlyscaledglobalstructuralstiffness

Structuraldampingmustbesimilartofullscalevalues

Themassdistributionmustbesimilar.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page12
Geometricalsimilaritybetweenmodelandfullscaleforanelasticstructurewillrequirethat
theelasticdeformationsaresimilar.Toillustratethiswewillconsiderthedeflectionofacantilever
beamasanexample.Thedeflection,,isgivenfrom:

FL3

EI

whereEIistheflexuralrigidityandFisthehydrodynamicforcewhichcanbeexpressedas:

F CU 2 L2
whereCisaforcecoefficientdependentonFN,Reetc.Therequirementofsimilarityindeformation
inmodelandfullscalegives:

F
LF

M
LM

F M

Usingtheaboveequationsthisrequirementissatisfiediftheratio:

C U 2 L4

EI

isequalinmodelandfullscale.Assumingequalforcecoefficientanddensityweobtainthe
followingrequirementtothestructuralrigidity:

U 2 L4
U 2 L4
5

EI F EI M
EI
EI

F
M

Ifalldimensionsofthecrosssectionalshapeofthebeamarescaledgeometricalsimilar,the
momentofinertia,I,willsatisfytherelation:

IF IM 4

WearethanleftthefollowingrequirementtotheYoungsmodulus,E:

EF EM

ThisimpliesthattheYoungsmodulusforthemodelmustbe1/timesthevalueofthefullscale
structure.
Itshouldbenotedthatthetwolastequationsisnottoberegardedasrequirementstothemodel.
ThebendingstiffnessrequirementisgivenforEI.Inpracticalmodeltestingtherequirementgiven
toscalingofEIisoftensatisfiedbymanipulatingthedifferentparametersbyapplyingother
materials,otherwallthickness,orbymodifyingthestructuralbuildupofthebeam.Theouter
geometry,whichisexposedtothehydrodynamicforces,hastobemodeledgeometricallycorrect.
Alsotherequirementtocorrectmodelingofmassdistributionandstructuraldampinghastobe
satisfied.Thiswillbefurtherdiscussedaspartofthephysicalmodeling,seechapter5.3.
Similarresultswillbefoundfortheaxialandtorsionstiffnesses.Therequirementfortheaxial
stiffnesscaseis:

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page13

( EA) F ( EA) M 3
whereEAistheaxialstiffness.Thisrelationgivesequalstraininmodelandfullscale.Thecross
sectionalarea,A,willsatisfytherelation AF AM 2 ,whichgivesthesamerequirementtothe
Youngsmodulusasshownabove.

2.6

Cavitation
Ifcavitationoccurs,dynamicsimilarityalsorequiresthatthelawofequalcavitationnumberis
accountedforintheexperiments.Therequirementisthatthecavitationnumbergivenas:

( gh p0 ) pv

1/ 2 U 2

hastobethesameforthemodelasinfullscale.p0istheatmosphericpressure,ghisthe
hydrostaticpressureandpvisthevapourpressure.Tosatisfythisrequirementacavitationtunnel,
withpossibilitytolowertheatmosphericpressurehastobeapplied.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page14

EXPERIMENTALFACILITIES

3.1

Introduction
Thedifferenttypeofexperimentalfacilitiesusedforshipsandoffshorestructurescanbe
categorizedasfollows:

Towingtanks,conventionalandfacilitiestailormadeforspecificpurposes.

Cavitationtunnels

Oceanbasins

Usuallywewillfindtwoormoredifferenttypeoftestsfacilitiesateachresearchandtesting
institution.ForexampleatMARINTEKthereisthreetowingtanks,acavitationtunnelandanocean
basin.InFigure3.1anoverviewofthetestfacilitiesatMARINTEKispresented.
Theexperimentalfacilitiesfortestingofshipandoffshorestructuresarenotonlythephysical
tank/basinwherethetestsareexecuted.Thetestingfacilitieshavealsotocoverdifferent
additionalfunctionsasworkshopsforconstructionandbuildingofmodels,instrumentation,
simulationofenvironmentandsoftwareandtoolstorecordandanalyzethemeasureddata.A
typicallayoutofatowingtank,includingutilityfunctions,isshowninFigure3.2.Fortestingof
realisticbehaviorofstructuresinaseaway,equipmentforgenerationofwind,wavesandcurrent
andefficientwaveabsorptionareofvitalimportance.

3.2

TowingTanks
Thefirsttowingtankswerebuiltforperformingtowingandpropulsiontests.Thelengthofthe
towingtankhastobelongenoughtogetasufficientlongtimewithsteadyflowconditionsfor
measurementsoftowingandpropulsionforces.Therequiredsizewillthereforebedependenton
typeofshipstobetested,scaleratioandforwardspeed.Todayalargenumberoftowingtanks
exist,morethan200areinregularuse.Thelengthofthetowingtanksisfrom20mtomorethan
1000m.
Thesmalltanksaretypicallyconnectedtoteachingandresearchinstitutions.Theverylongtanks
aremainlyconnectedtonavalactivities.AnexampleisthehighspeedtankatDavidTaylorNaval
ShipResearchandDevelopmentCentrewithdimensions900mx6.4mx3mwithamaximum
towingcarriagespeedofupto50m/s.Thistankwasbuiltespeciallyfortestingofhighspeedships.
Theconstructionofthistankwasadirectresultoftherequirements:Navyin50kn,definedin
about1960asatargetfortheUSNavy.AsimilarfacilityalsoexistsinSt.Petersburg.
AtypicalsizeforcommercialworkingtowingtanksisLxBxd=250mx10mx5m.Typicalshipmodel
lengthis58m.Thissizeoffacilitiesseemstorepresentareasonablecompromisebetweencostfor
tankconstruction,costformodelmanufactureandoperationalcosts(whichtogetherdetermine
thecostofmodeltesting)andtherequiredscaleratioandcorrespondingaccuracythatcanbe
achieved.ThesizeofthelargehydrodynamiclaboratoriesattheMarineTechnologyCentreis
showninFigure3.1,withmoredetailsaboutthetowingtanksgiveninFigure3.3.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page15
Almostalltowingtanksuseatowingcarriagetomovethemodeltroughthewater.A
typicaltowingcarriagedesignisshowninFigure3.3.Thetypicalmaxcarriagespeedis10m/s.
Duringcalmwatertowingandpropulsionthemodeliskeptfixedinsurgeswayandyaw,butfreeto
heaveandpitch.

Figure 3.1 Overview of test facilities at MARINTEK

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page16

Figure 3.2

A typical towing carriage design.

Tobeabletoperformseakeepingtestsorothertypeoftestinginsurfacewaves,manytowing
tanksareequippedwithawavemakeratoneendofthetank.Generationofwavesandtypeof
wavegeneratorsarediscussedinmoredetailsinchapter3.5.1.Inordertopreventreflectionsof
wavesfromtheoppositeside,awavebeach,whichisabsorbingthewaveenergy,hastobe
installedatthissideofthetowingtank.
Ithasalsobeenconstructedtowingtanksthatarehighlyspecialisedforagivenpurpose.An
exampleistheDutchandChinesevacuumtanks,wheretheentirespaceabovethetankis
evacuatedandairpressuredownto0.04barcanbeachieved.Thepurposeofthistypeoffacilityis
todopropulsiontestswithsurfaceeffectsandcompleteshipmodelpresent,atthelowpressure
requiredforequalityincavitationnumber.
AnotherexampleofaspecialisedtankistheicetanksinHamburgandHelsinki.Iceismodelledby
freezing,usinghighsalinitywaterandchemicalstocontrolthemechanicalprioritiesoftheice.
Thesetanksareusedfortestingoficebreakersandoffshorestructuresexposedtotheactionof
driftingice.
ThetowingtankatMARINTEKwascompletedin1939withdimensions170x10.5x5m.Later,in
1978,extendedto260mwherethedepthoftheextensionis10.0m.Thelayoutofthetowingtank
isshowninFigure3.4.Thetowingcarriageisofconventionaltype.Fortestingofhighspeed
vesselsthecarriageisequippedwithaFreetoSurgerig.The8mlongrigismountedinfrontof
thetowingcarriageasshowninFigure3.4.Usingthisrigthewinddisturbanceatthepositionofthe
modeliseliminatedandthemodelisallowedtofreelysurge,heaveandpitchduringwavetesting.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page17

6
7

28

10
4

5
3

13
2

9
10
11
12
13

Ship model manufacturing


shop
Trimming tank
NC milling machine for
model production
Instrumentation workshop
10.5

5.6
10.0

11

Model store
Drawing office
Reception
Tank II

13.5

1
2
3
4
5

TANK I

14
15

Carpenter workshop
Propeller model
manufacturing shop
Cavitation laboratory
Dock gate
Wave absorber, Tank I
and Tank I+III
Wavemaker, Tank III
and Tank I+III
Wave absorber, Tank III

12

15

TANK III

14

1
39

85
260

Towingtankdata
TankI

TankII

TankIII

TankI+III*

Length:
175m
25m
85m
260m
Width:
10.5m
2.8m
10.5m
10.5m
Depth:
5.6m
1.0m
10m
5.6/10.0m
Tot.weightcarriage:
20tons
0.2ton
15tons
20/15tons
Wheelbase:
11.04m
3m
11.04m
11.04m
Speedrange:
0.0210m/s
0.051.75m/s
00.9m/s
0.0210m/s
2
2
2
Max.acceleration:
1m/s
1m/s
1m/s
1m/s2
Modelsizerange:
8m
1m

8m
Doubleflap
Doubleflap
Singleflap

Wavemaker:
Regularand
Regularand
Regularand

irregularwaves
irregularwaves
irregularwaves

0.9m
0.3m

Max.waveheight:
0.9m
0.85sec.
0.253sec.

Waveperiodrange:
0.85sec.
1:10
1:8

Max.wavesteepness:
1:10
Wavespectra:
Computergenerated

* Tank I and III can be used separately and also as one long tank (Tank I + III) by removing the gate (12)
and wave absorber (15). In Tank I + III either of the two carriages can be used.
Figure 3.3

Towing Tanks at MARINTEK.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page18

Figure 3.4

3.3

Free to surge rig in front of towing carriage.

CavitationTunnel
Cavitationtunnelsaredesignedtobeabletotestpropellersandotherliftingsurfacesata
sufficientlylowpressuretoachievecorrectcavitationnumber.Mostcommercialmodeltest
institutionshaveoneormorecavitationtunnels,andabout100tunnelsistodayinregularuse.
Awiderangeofdifferentsizecavitationtunnelsexist,fromsmallsizetunnelsforresearchand
education,withtestsectionareaoftypically0.25x0.25m,toverybigcirculatingwatertanks,with
testsectiondimensionupto3x6mandlengthof11m(theBerlintunnel).Atypicalsize
conventionaltunneliswithcirculartestsectionwithdiameterofabout1mMaximumflowspeed
atmeasuringsectionisusually1020m/s.Largetunnelsoftenhavetestsectionsallowingfor
mountingofacompleteshiphullmodel.
Fortunnelsthatcannotallowtestingofentireshiphullmodels,anafterbodymodeloftheshipis
oftenappliedtoproducecorrectinflowtothepropeller,andmeshscreensareusedtoproducethe
specifiedwakedistribution.Thebenefitofusingafterbodymodelsandmeshscreens,insteadofa
completemodel,isthepossibilityofhavingthemeshscreensimulatefullscalewake,notonly
modelscalewake.Whentestingtheentireshipmodel,onlymodelscalewakemightbetested.
Somecavitationtunnelsareofthefreesurfacetype.Suchtunnelscanbeusedfortestingofhigh
speedpropellersoperatinginfullorsubmergedcondition.Thistypeoftunnelsisespeciallywell
suitedforstudyingventilationproblemsforpropeller,waterjetsandfoilsections.Largetunnels
withfreesurfaceenabletestwithnormalshipmodels.
ThecavitationtunnelatMARINTEKisshowninFigure3.6.Thediameteroftheworkingsectionis
1.2mandthelengthofworkingsectionis2.08m.Maximumwatervelocityis18m/s.The
minimumworkingpressureis0.1atm.Afterbodymodelsandmeshscreensareapplied.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page19

Figure 3.5

Cavitation Tunnel at MARINTEK.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

3.4

Page20

OceanLaboratories
Theoceanlaboratoriesareingeneralconstructedfortestingofoffshorestructuresandfor
seakeepingandmanoeuvringtestingofships.
Incontrasttothetraditionaltowingtankstheoceanbasinsandseakeepinglaboratoriesmakeit
possibletocarryouttestswithanywaveheading(obliquewaves)forshipswithforwardspeed.For
manoeuvringtestsitisrequiredwithatowingcarriagewithcontrolledmotionsinbothlongitudinal
andtransversedirection.Thisisachievedbyasubcarriage,whichisconnectedunderneaththe
mainlongitudinalmotioncarriage(seeFigure3.7,takenfromtheseakeepinglaboratoryatSSPA,
Sweden).Combinedwiththelargewidthofthesefacilities(typically3050m)thearbitrary
horizontalmotionrequirementtendstomakethecarriagesystemcomplexandheavy.
Basinspurposebuiltforoffshoretestingisformostcasesbuiltafter1980.Foroffshoretestinga
largecarriagesystemisnotrequired.Oceanlaboratoriesareusuallyequippedwithadvanced
systemsforgenerationofwaves,oftencapableofgenerationofbothlongcrestedand
multidirectional(orshortcrested)wavesaswellaswindandcurrent.Inthiswayitispossibleto
givearealisticrepresentationofthemarineenvironmentalconditions.

Figure 3.6

Carriage system in Ocean Basin (from SSPA, Sweden).

Examplesofcommercialoceanlaboratoriesfortestingofcoastalandoffshorestructuresare:
MARINTEK,Trondheim;

LxB=80mx50m,d=010m

MARIN,Netherlands;

LxB=45mx36m,d=010.5m,pitincentrewithd=30m

HydralicLab,Ottawa,Canada:

LxB=50mx30m,d=3m

OTRC,TexasA&M:

LxB=45.7mx30.5m,d=5.8m,pitwithd=16.8m

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page21
Somebasinsareequippedwithafalsebottomthatcanbesettodifferentdepths.Inthis
waytheactualwaterdepthcanbecorrectlymodeledinthetestsetup.Thispropertyalsoenable
mountingofmodels,mooringsystemandothersubseaequipmentonadrybottom,whichafter
installationofthesystemtobetested,islowereddowntothewantedwaterdepth.Thislargely
simplifiesthepreparationworkforthetest.
TheoceanlaboratoryatMARINTEKisfittedwithtwosetsofwavemakers.Alongthe50msideof
thebasinthereisadoubleflapwavemakercapableofgeneratinglongcrestedwaves.Alongthe80
msidethereisamultiflapwavemakerconsistingof144individuallycontrolledflapsforgeneration
ofshortcrestedandlongcrestedwaves.Waveabsorptionbeachesareinstalledonthetwo
oppositesidestoreducetheproblemswithwavereflections.Currentcanbemodeledindirection
alongthebasin(inwavedirectionofthedoubleflapwavemaker).Thewaterdepthisadjustable
from0m(surfaceposition)to10mbymovingthefalsebottom.Amoredetaileddescriptionofthe
OceanLaboratoryatMARINTEKisgivenbyHuseandTrum(1981)andNaeser.(1981).

3.5

Generationofenvironment
Reliablemodeltestingrequirescontrolledgenerationofwind,wavesandcurrentinbothtimeand
spacetoachievearealisticandwelldefinedenvironment.Commonlyusedequipmentfor
environmentgenerationisdescribedinthefollowing.

3.5.1 Wavegenerationandabsorption
Therearetwomainclassesofwavegenerators,thehorizontaldrivenflaptypewavemakerandthe
verticaldrivenwedge(plunger)typewavemaker.Inmoderntestfacilitiesalmostonlytheflaptype
isused.TwoexamplesofflaptypewavemakersareshowninFig3.8.Thefirstoneisadoubleflap
wavemakerasinstalledinthetowingtankandintheoceanbasinatMARINTEK.Hydraulic
actuatorsareused.TheotheristhesingleflapwavemakerasinstalledinMarineCybernetic
Laboratory(MCLab)atMARINTEK.Thiswavemakeriselectricallydriven.Therearsideoftheflap
maybeeitherdryorwet.Thedoubleflaptypeisusuallyusedfordeeperwater.Bythepossibility
ofusingtheupper,thelowerorbothflapsincombinationsforthedoubleflaptype,itispossibleto
generatewaveswithaminimumofdistortionforlargerwavelengthrangethanwhatispossiblefor
asingleflapsolution.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page22

Figure 3.7

Examples of flap type wave maker; single flap and double flap.

Othertypesofwavemakersarepistontype(asmallsketchisshowninFigure3.9),andpneumatic
wavemakers.Pneumaticwavemakersusevariableairpressureinachamberabovethewaterat
theedgeofthebasintocreatewaves.DavidTaylorModelBasininWashingtonDChasapneumatic
wavemakerintheiroceanbasin,exceptforthattheprincipleislittleused,andisconsidered
inferiorrelativetoflaptypewavemakers.

Figure 3.8

Wave-maker theory, Wave height to stroke ratio as function of relative depth.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page23
Thegenerationofwavesarecontrolledbythefrequencyandamplitudeoftheflap.In
Figure3.9therelationbetweenflapstroke,S,andwaveheightH,isshownforflaptypewave
maker.Theresultingwaveheightisshownasafunctionoftheparameterkhwhere k 2 is
thewavenumber,isthewavelengthandhisthewaterdepth.Theresultsarebasedonwave
makertheoryseee.g.DeanandDalrymple(1984).Thisratiobetweenthemechanicaldisplacement
oftheflaptothewaveamplitudeisthetransferfunctionofthewavemaker.
Aregularwaveelevationcanbegeneratedusingthetransferfunctiontodeterminetherequired
controlsignaltothewavemaker.InFigure3.10themaximumwaveheightforregularwavesas
functionofwaveperiodisshownforthedoubleflapwavemakersinthetowingtankatMARINTEK.
Itisobservedthatincreasingwaveperiod(andhencewavelength),givesdecreasingmaximum
waveheight.

Figure 3.9

Maximum wave height as function of wave period, Double Flap wave maker at
MARINTEK

Thegenerationofirregularwavesiscontrolledbyaninputsignalbasedontheselectedwave
spectrumcombinedwiththetransferfunctionofthewavemaker.Thecommonlyusedassumption
thattheseasurfaceelevationisastationaryGaussianprocesswithzeromeanisapplied.The
surfaceelevationasfunctionoftime,(t),canthanberepresentedbyafinitenumberofFourier
components:
N

(t ) an cos( nt n )
n 1

wherenisthephaseangleofcomponentncreatedfromarandomphasegenerator.Random
phaseisnecessarytoeliminateanycoherentfeaturesdevelopinginthewavesignal.anisthe
FourieramplitudeofcomponentndeterminedfromtheinputwavespectrumdensityS()as:

an 2 S ( n )

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page24

isthefrequencyintervalforeachcomponent.Itisimportantwithasufficientnumberof
componentsinthewavegenerationtoavoidrepetitionofthewavesignalandaliasing,see
Newland(1975)forfurtherdetails.Typically,2000componentsareused.
FormostcasesthewavespectraaregeneratedaccordingtotheJONSWAPformulation:

S ( ) g 2 5 exp 1.25( 0 ) 4 exp ( 0 )

( 2 0 2

where:

0.07 for 0
0.09 for 0
0 isthespectralpeakfrequency
isthepeakednessparameter.
InFigure3.11atypicalexampleoftheoreticalJONSWAPspectrumandmeasuredwavespectrumin
thewavetankareshown.Theagreementinenergydistributionisseentobeverygood.

Figure 3.10

Typical example of theoretical JONSWAP spectrum and measured wave spectrum in


the wave tank

Multidirectional(orshortcrested)wavescanbegeneratedusinganarrayofflapsalongonesideof
thebasin.Therearetypicallyaboutahundredindividualflaps.Themultiflapwavegeneratorscan
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page25
alsobeusedtogeneratelongcrestedwaveswithanarbitrarywaveheading.Togenerate
shortcrestedwavethedirectionalspreadingfunctionmustbespecifiedinadditiontotheenergy
densityspectrum.Thisgivesthefollowinggeneralisationforthesurfaceelevationasfunctionof
spaceandtime,(x,y,t):
N

( x, y, t ) a n cosk n ( x cos n y sin n ) n t n

n 1

whereannowistheFourieramplitudeofcomponentnincludingthespreadingfunction:

a n 2 S ( n ) D( n , n )
Toavoidreflectionandwavebuildingupinthebasin,anefficientwaveabsorptionsystemisalso
essential.Themostusedsystemforwaveabsorptioniswavebeaches.Anexampleisshownin
Figure3.12.Theshapeisparabolicwhichhasbeenfoundtobemoreefficientforalargewave
periodrangecomparedtoastraightbeach.Reflectedwaveheightoflessthan5%oftheincoming
waveheightwilltypicallybeachievedwiththisbeachdesign.Inoceanbasinswaveabsorbersare
usuallymountedonthesideswithoutwavemakers,typicallywithtwosideswithwavemakersand
twosideswithwaveabsorbers(asinMARINTEKoceanbasin).
Intowingtanksthemainproblemwillbereflectionoftheshipgeneratedwavesystemfromthe
tankwalls.Itisnotpracticaltomountabeachalongsidethetankwallandtransversewaveswillbe
generated.Fortestswithforwardspeedthisisusuallynotaproblem,sincethereflectedwaveswill
hitthetestareaafterthemodelhasleft.Fortestswithzeroorverylowforwardspeed,the
problemofwavereflectionsmustbetakenveryseriously.Forlongtestruns,likeistypicalforatest
inirregularwaves,somekindofwaveabsorptionalongthetankwallisrequired.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page26

Figure 3.11

Upper; Wave absorber, beach type. Lower; measured wave reflection from beach, in
% of incoming wave amplitude.

3.5.2 Windgeneration
Windgenerationismostconnectedtostationarymodeltestsinoceanbasins.Forfreelymoving
modelswindisnoteasilyappliedinpracticaltesting.Thewindinthebasinisusuallygeneratedby
meansofabatteryofportableelectricalfans.Thefansareplacedsomedistancefromthetesting
areatoachieveahomogeneouswindspeeddistributionatthepositionofthetestmodel.Thewind
directioncanbechangedbymovingthepositionofthefans.
Twodifferentmethodsforcalibratingthewindspeedarecommonlyused:
1. Froudescalingofwindspeed,i.e: U M ,Wind U F ,Wind

2. Usingprecalculatedwindforceactingonthemodelandtuningthemodelwindspeedto
thisforceisachieved.
Forthefirstcasethewindspeediscalibratedatthepositionofthemodel,butwithoutthemodel
present.Usingthisprocedurerequireaveryaccuratemodellingofthemodelsuperstructureto
obtainreliablewindforces.Thelastprocedurerequiresthatreliablewindforceestimatesare
availableonbeforehand.Ifthisisthecasethescaleeffectsonwindforcescanbeavoided.Usually
thisprocedurewillgiveabout20%higherwindspeedthanthespeedestablishedfromFroude
scalinglaw.
Theeffectofwindwillbeimportantforalmostalltypesofmooredstructures.Thewindspeedisin
generalnonsteadyandthedynamiceffectsofthewindscanbeanimportantexactionsourcefor
resonancemotionsofmooredstructureandinspecialcasesforrollmotionsofships.Thedynamic
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page27
effectofwindcanbesimulatedinthetestbycontrollingthepowertothefans.A
frequentlyusedwindenergyspectrumforoffshoreapplicationsistheNPDspectrum.
Thefrequencyrangeofawindspectrumisquitebroadbanded,oftencoveringarangefrom0.005
to1Hz.

3.5.3 Current
Tosimulatecurrenttwodifferentapproachesarecommonlyused:
1. Directgenerationofcurrentinthebasinusingpumpingofwater
2. Towingofthemodelsetupwithspeedequaltothecurrentspeed.
Thefirstapproachrequireslargepumpswithrecirculationducts.Forbasinswithafalsebottom(as
atMARINTEK)thepumpscanrecirculatethewaterunderthefalsebottom.Alsoexternalpiping
(outsidethebasin,asusedinMCLabatMARINTEKandatMarinintheNetherlands)canbeusedfor
recirculationofwater.
Localcurrentcanbegeneratedbyplacingportablecurrentgeneratorsinfrontofthemodel(in
principleasforthewindgeneration).Howeverforthismethoditisdifficulttoachieveareasonable
stationarycurrentfieldatthepositionofthemodelduetolargeeddiesofbackflowingwater.The
effectofcurrentonwaveforceswillnotberealisticallyaccountedforbythisproceduredueto
largespacevariationincurrentfieldbetweenthewavemakerandmodel.
Theeffectofcurrentisespeciallyimportantformooredstructures,bothduetothedirectforces
duetothecurrentandduetotheinteractionbetweencurrentandwaves.Theinteractionbetween
currentandwaveseffectcanlargelyinfluencethewavedriftforcesandhenceinfluencemean
offsetandforcesaswellastheslowdriftmotions.Forthecasewithmooredstructuresthe
simulationofthecurrenteffectbytowingthemodelwithaspeedequaltothecurrentspeedisnot
apracticalsolution.Forthiscaseabasinwithdirectcurrentgenerationwillberequired.
ThegenerationofcurrentspeedisbasedontheFroudescalinglaw.Thisisnecessarytoproperly
representthewavecurrentinteractioneffects,butitmayintroducesomescaleeffectsforthe
currentforcesduetodifferenceinRenumber,resultinginpossiblydifferentflowregimesinmodel
andfullscale.
Currentcalibrationofspeedandprofileshouldbeperformedwithoutthemodelinthebasin.
Velocityfluctuationswillalwaysbepresentinbasingeneratedcurrent(inrealfullscalecurrent,
fluctuationswillalsobeobserved).Largefluctuationsincurrentmayrepresentanexcitationsource
forslowlyvaryingresonanceoscillationsandthemagnitudeshouldthereforebeaslowaspossible.
Astandarddeviationforthecurrentfluctuationofabout5%ofmeancurrentistypicallyachievedin
basinswithclosedrecirculationsolutions.Correctmodellingofthecurrentfluctuationsmightbe
importantforthedynamicsofdeepwatersystems,butwestillnotknowenoughaboutthis,andno
modeltestingfacilitiescurrentlyhavepossibilitiestocreatecontrolledcurrentfluctuations.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page28

INSTRUMENTATION

4.1

Generaldescriptionofequipment
Alargerangeofdifferenttypeofmeasuringequipmentisusedfortestingofshipsandoffshore
structures.Usually,instrumentsaredesignedtogenerateananalogvoltageorcurrentsignalwhich
islinearlyproportionaltothemeasuredparameter.Nonlinearcharacteristicsoccurinrarecases.
Instrumentswithdigitaloutputareincreasinglyused,butanalogoutputisstillpreferred,inorder
toavoidthecomplexitiesofdealingwithdifferentdigitalsignalprotocols.
Thesystemrequiredforperformingmeasurementsincludesthefollowingcomponents:

Thetransducers

Amplifiers

Filters(analogand/ordigital)

ADconverter

Datastorageunit

Cablingbetweenthedifferentcomponents

TRANSDUCER

Figure 4.1

AMPLIFIER
& SIGNAL
CONDITIONER

A/D
CONVERTER

COMPUTER DATA BUS

RESPONSE

AtypicalsetupisshowninFigure4.1.Itiscommonpracticetousetwoormoreindependent
computersystemsforoperatingthetankfacility.Onemachineisusedforrealtimegenerationof
controlsignalforthewavemakerandanadditionalmachineisusedforthedataacquisitionand
analysis.Additionalmachinesmightbeusedforcontrolofruddersorothercontroldevices,orfor
controlofthecarriage.

Schematic of typical set op of a data acquisition system for model testing

Afurtherdescriptionofinstrumentationandtransducersrelevantformodeltestingcanbefoundin
Olsen(1992).Adetaileddescriptionofmeasurementtechniquesforfluidmeasurementsisgivenby
Goldstein(1983)

4.2

Strainanddisplacementmeasurements
Themostusedmethodsforstrainanddisplacementmeasurementsinmodeltestingarebasedon
thefollowingprinciples:

Resistivetransducers,basedonchangeofresistanceduetostrain;straingauges

Inductivetransducers

Capacitancetransducers.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page29
Inadditiontoadirectmeasureofstrainanddisplacements,thesetypeoftransducerare
alsocommonlyusedasthebasisforpressurecells,forcetransducers,velocitymeasurementsand
accelerometers.Theseapplicationswillbediscussedseparately.

4.2.1 Straingauges
Thestraingaugemeasuredtheelongationinthematerialtowhichitisglued.Straingaugesare
commonlyusedinanumberofdifferenttypesoftransducers.Examplesofstraingauge
constructionsareshowninFigure4.2.ThethreadsareusuallyCUNialloys.
Theuseofstraingaugesisbasedonthattheelongationofthestraingaugewillchangethe
resistance.Thegaugefactork,isdefinedas:

R
L

R
L

whereRistheresistanceandListhelengthandrepresentthechangeoflengthorresistance.The
factorkistypicallyabout2formetallicmaterials.Increasingthefactorkwillincreasethe
sensitivityofthestraingauge.kvaluesuptoabout100200canbeachievedusingpiezoresistive
materials.
Theelongationofthestraingaugesususuallymeasuredinmicrostrain,S=106S,where

S L L .

Figure 4.2

Examples of strain gauges designs

Tomeasurethechangeofresistanceoverastraingauge,a
Wheatstonebridgecircuitisused.AbasicWheatstone
bridgecircuitcontainsfourresistances,aconstantvoltage
input,andavoltagegage.ForagivenvoltageinputVin,the
currentsflowingthroughABCandADCdependonthe
resistances,i.e.,

Vin V ABC V ADC

I ABC R1 R2 I ADC R3 R4

Figure 4.3 Wheatstone bridge circuit

ThevoltagedropsfromAtoBandfromAtoDaregivenby:

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page30

V AB I ABC R1
V AD

Vin
R1
R1 R2

Vin
R4
I ADC R4
R 4 R3

ThevoltagegagereadingVgcanthenbeobtainedfrom:

V g V AB V AD

Vin
Vin
R1
R4
R1 R2
R 4 R3

R1 R3 R2 R4
V

R1 R2 R4 R3 in

Nowsupposethatallresistancescanchangeduringthemeasurement.Thecorrespondingchange
involtagereadingwillbe:

V g V g

R1 R1 R3 R3 R2 R2 R4 R4
V
R1 R1 R2 R2 R4 R4 R3 R3 in

Ifthebridgeisinitiallybalanced,theinitialvoltagereadingVgshouldbezero.Thisyieldsthe
followingrelationshipbetweenthefourresistances:

R1 R3 R2 R4
V 0
R1 R2 R4 R3 in

Vg

R1 R3 R2 R4

or

R1 R4 1

R2 R3 r

Wecanusethisresulttosimplifythepreviousequationthatincludesthechangesinthe
resistances.DoingsoresultsinthesolutionforthechangeinVg:

V g

1 r

R1 R2 R3 R4

1 Vin

R
R
R
R
1
2
3
4

whereisdefinedby:

1 r
1
R
R3
R1 R4

r 2
R1
R4
R3
R2

Moreover,whentheresistancechangesaresmall(<5%),thesecondorderterm is approximately
zero and can be ignored. We then have:

V g

1 r

R1 R2 R3 R4

Vin

R2
R3
R4
R1

whichisthebasicequationgoverningtheWheatstonebridgevoltageinstrainmeasurement.The
coefficient

1 r 2

iscalledcircuitefficiency.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page31
Inpractice,oneoftenusesthesameresistancevalueforallfourresistors,R1=R2=R3=R4=R.
Notingthatr=1inthiscase,thechangeinvoltagecanbefurthersimplifiedto

V g

R1 R2 R3 R4
Vin
4R

DifferenttypesofWheatstonebridgecircuitsareusedtomeasurethechangeofresistanceovera
straingauge.Averysimpleexampleformeasurementofforceisshowninfigure4.4.Onestrain
gaugeismountedtoeachsideofabeam.Twodummyresistances(usuallyintegratedinthe
amplifier)areusedforbalancingthebridge.Thissetupiscalledahalfbridge.Aconstantvoltageis
usedasexcitation,Vin.Theforcegivesrisetoanelongationofstraingauge1andcompressionof
straingauge2.Thisintroducesanunbalanceinthebridgeandavoltagecanbemeasuredatthe
exitatVg.Inafullbridgecircuit,allfourbranchesofthebridgearestraingauges.Mountingtwo
straingaugesoneachsideofthebeaminFigure4.4inafullbridgearrangementwouldgivetwice
thesensitivityofthehalfbridgearrangement.
B
Force K

R
-

R
+
R

Strain
1
gauges

G Vg

Side view

Figure 4.4
gauges.

Front view

Vin

Examples of half-bridge circuit for measurements of change of resistance of strain

4.2.2 Inductivetransducers
Theinductivetransducersarebasedonthevoltageinducedbyamovablecore.Anexampleofan
applicationisshowninFigure4.5.TheshownsystemiscalledLVDT(linearvariabledifferential
transformer).OnesetofthecoilisexcitedbyACvoltageandtheinducedvoltageismeasuredin
thesecondset.
Thistypeoftransducerisavailableinawiderangeofsizes,frequencyrangesandaccuracys.Itis
usedfordirectpositionmeasurements,butalsoasbasisforpressurecellsforcetransducers,
velocitymeasurementsandaccelerometers.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page32

Figure 4.5

LVDT transducer for displacement measurements.

4.2.3 Capacitancetransducers
Thecapacitancetransducersconsistoftwocloselyspacedplatesoraplatesuspendedbetweena
pairofouterplatesasshowninFigure4.6.Theplatesareconductiveandrelativemovement
betweentheplatesintroduceavariationofthecapacitance

F
F

Figure 4.6

Capacitance transducers for displacement measurements.

Thistypeoftransducersrequireamuchsmallerdrivingforcecomparedtoinductivetransducers,
buthaveahighernoiselevelandarethereforelessfrequentlyusedinpracticalmodeltesting.

4.3

Positionmeasurements
Typicalpositionmeasurementsofinterestforfloatingstructureswillbethe6degreesoffreedoms
rigidbodymotionsofship/platformsandmotionsofmooringlinesandrisers.Otherexamplesare
measurementsofdeflectionsofelasticmodelsasforspringingandwhippingresponseofships.

4.3.1 OpticalandVideosystems
Forfreerunningmodelsandmooredstructurestheglobalmotionsaremeasuredbyopticalor
videobasedsystems.Foropticalsystemminimum3lightemittingdiodesarelocatedonthemodel.
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page33
Forvideosystemsballshapedreflectorsmountedonthemodelisused.Onshorecameras,
minimum2,areusedforreadingthepositionofeachdiode.Basedontheinstantaneousposition
(x,yandz)ofeachofthe3diodes,themotionsin6DoFaredetermined.
Theaccuracyofthemeasuredmotionswillforbothopticalandvideosystemsbeoftheorderof
+/1mmforposition(inmodelscale)and+/0.05degreesforroll,pitchandyaw.Formostcases
thisisanacceptableaccuracyfortherigidbodyvesselmotions.

4.3.2 Gyros
Therollandpitchmotionscanalsobemeasuredusinggyros.Theprinciplebehindthegyrois
showninFigure4.7.Arotatingmasskeepstheinnerpartofthegyroinacontinuoushorizontal
position.Thenextlinkcanbetiltedaboutoneaxisandtheangleismeasuredusinga
potentiometer.Theouterlinkcanbetiltedaboutanaxisperpendiculartothefirstaxisandthe
anglecanbemeasuredinthesameway.Fromthemeasuredanglesandtheknownsequenceof
theanglestherollandpitchmotionsareuniquelydetermined.
Thegyroisarobusttoolandiscommonlyusedbothinmodeltestingandininstrumentsappliedin
fullscale.However,sinceitinvolvescomplexmechanicalcomponents,itisfairlylargeandfairly
expensive.Thesizelimitstheuseinmodels,andthecostlimitstheusebothinmodelandinfull
scaleexperiments.

Figure 4.7

Principle of a gyro for measurement of roll and pitch.

4.3.3 Potentiometer
Lowfrictionpotentiometerscanbeusedformeasurementofmotioninonedirection.ANylonline
isconnectedtothemodel,thenpassedaroundthepulleyonthepotentiometerspindleand
tensionedbysprings.Themotionofthemodelwillthereforebedirectlytransducedintoavoltage
signalbythepotentiometer.
Acommonlyusedsetupformeasurementsofheaveandpitch(trim)fortowingtestsisshownin
Figure4.8.Thepotentiometersareusedformeasurementsofthemotionsbetweentheshipmodel
andthetowingcarriage.Thesetupisusedbothformeasurementsofrunningheaveandtrimin
calmwatertestingandformeasurementsofwaveinducedmotionsinheadseawaves.
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page34

Figure 4.8
towing tests.

Example of using potentiometers for measurements of heave and pitch motions in

4.3.4 Positionbasedonmeasuredforceandacceleration.
Measurementofpositioncaninprinciplebeobtainedfrommeasuredacceleration(seebelow)bya
doubleintegrationofthemeasuredaccelerationsignal:

x(t ) a (t ) dtdt At B

Ascanbeseen,positionbasedonintegratedaccelerationcannotgiveinformationaboutmean
levelorpossibleconstantdriftoftheposition.Increasingperiodofoscillationwillgivereduced
accuracyofthederivedposition.Forpracticalapplicationsthismethodarethereforeusuallylimited
tomeasurementsofthewavefrequencypartofthemotions.
Anotherindirectwaytoestablishthepositionisusingmeasuredforceincombinationwithalinear
spring:

x (t )

F (t )

Themethodrequiresaspringconnectionbetweenthemodelandafixedpoint(e.gtowingcarriage,
seabedetc.).Itisnecessarythatthespringstiffnessissufficientlowtoavoidanyinfluenceonthe
dynamicbehaviorofthemodel.

4.4

Accelerations
Measurementsmadebyaccelerometersarebasedontheratiobetweenforce,massand
acceleration:

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page35

a (t )

F (t )

Amasscanbeconnectedtoabeam.Whenexposedtoaccelerationthebeamwillbedeflectedby
theinertiaforces.Thedeflectionofthebeamisproportionaltotheacceleration.Straingaugescan
beusedformeasuringthedeflectionofthebeamandhencetheaccelerationisobtained.
Anothertypeofaccelerometersisbasedonthepiezoelectricaleffect.Apiezoelectricalmaterialis
amaterialwhichwhendeformedproducesanelectricalfield.Thevoltagegeneratedisproportional
tothesurfacepressureapplied.Combinedwithamassthisgivesasignalproportionaltothe
acceleration.TheprincipleisshowninFigure4.9.Thechargeistransferredtovoltageinacharge
amplifier,butsomeofthechargeleaksout.Thisgivestheaccelerometeralowerlimitforwhich
frequenciesthatcanbecovered.Thistypeofaccelerometerscanthereforeonlybeusedfor
dynamicmeasurements.

Figure 4.9

Piezo electric material exposed to surface pressure.

Theresonanceofthemassspringsystemmayinfluencethemeasurementofaccelerations.For
frequencieswellbelowtheresonancefrequencythemasswillfollowthemotionsofthehousing
andwehavealinerrelationbetweentheaccelerationandthesignalout.Forfrequenciesinthe
resonanceregionthemasswillbeexitedandthesignaloutwillbefrequencydependent.The
dynamicamplificationswillbedependentofthedampingofthesystem,butingeneral
accelerometersshouldonlybeusedformeasurementsofresponseswithoscillationfrequencies
wellbelowthenaturalfrequencyoftheaccelerometer.AccelerometersbasedonPiezoelectricity
canbemadeverystiffwithresonancefrequencyhigherthan500KHz.Thismakesthemusefulfor
applicationsofmeasurementsofresponseduetoimpactloads.Toincreasethesensitivityofthe
accelerometerthemassmustbeincreasedorthestiffnessreduced.Accelerometerswithlow
naturalfrequencywillthereforebemoresensitivethantheaccelerometerswithhighresonance
frequency.
Theaccelerometersaddweighttothestructureanditisthereforeimportanttoensurethatthat
weightissufficientlylowtoavoidanyinfluenceonthedynamicbehaviour.Theweightofthe
accelerometercanbemadeverysmall,typicallydowntoafewgrams.

4.5

PressureTransducers
Pressuremeasurementsaremostlyperformedusingpressurecells.Threetypesofpressurecells
arecommonlyused:

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page36
1. basedonpiezoelectricity
2. basedoninductivetransducers
3. basedonstraingauges.
Pressurecellsarebasicallyforcemeasurementsoverasmallarea.Typicaldimensionsofpresser
cellsusedformodeltestingisD=210mm.Thedifferenttypesofpressuremeasuringdevicesare
illustratedinFigure4.10.

Figure4.10

Schematicrepresentationofthemostcommontypeofpressuretransducers.

(a):Capacitancetransducer,(b):Piezoelectricand(c):Straingauge
Pressurecellsbehaveinmanywayssimilarasanaccelerometerandtheresonanceofthemass
springsystemmayinfluencethemeasurements.Thedynamicamplificationswillbedependentof
thedampingofthesystem,butingeneralpressurecellsshouldonlybeusedformeasurementsof
responseswithfrequencieswellbelowthenaturalfrequencyofthecell.Straingaugetypecells
respondstodisplacementsfromdcto5kHz.Itisthereforewellsuitedformostpracticalmodel
testing.Pressurecellsbasedonpiezoelectricitycanbemadeverystiffwithresonancefrequency
uptomorethan500kHz.AnexampleofthistypeofpressurecellisshowninFigure4.11.This
transduceristhereforewellsuitedtomeasurementsofpressurebehaviorwithverylowrisetimeas
willbethecaseforslammingpressuremeasurements.Anexampleofmeasuredslammingpressure
fortheimpactofaflat,elasticplatetowardsawavecrestisshowninFigure4.12.Itisobserved
thatclosetothecenteroftheplatebottomwherethewavecresthit,therisetimeislessthan
0.0001sandthedurationofthepeakextremelyshort.Consequentlyapressurecellwithveryhigh
resonanceperiodwillberequiredtoaccuratelyreproducethispressurebehavior.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page37

Figure 4.10

Piezoelectric pressure transducer for pressure measurements.

Figure 4.11
Example of measured slamming pressure. Impact of a horizontal circular cylinder
towards calm water surface.

4.5.1 Measurementofpressuredistribution
Inthelateryears,pressuresensingfilmhasbeendevelopedbyseveraldifferentcompanies.The
filmisbasicallyamatrixofsmallpressurecellsintegratedintoaflexibleplasticfilmthatcanbe
appliedtocurvedsurfaces,andwilleffectivelyreturnthepressuredistributionoverthesurface.
Thepressurecellsaremadeofalayerofsemiconductingmaterialwherethedegreeof
conductivitydependsonthepressureappliedtothematerial.Twocompaniesthatdevelop
pressuresensingfilmareTekscanhttp://www.tekscan.com/andPressureProfileSystems
http://www.pressureprofile.com/.Thistechnologyhasmainlybeendevelopedfordry
applications,liketestinganddevelopmentofcarseats,sportsequipment,andsimilar.Thus,itisnot
straightforwardtoapplyittomarinehydrodynamicsproblems,butthepossibilityofeasily
measuringthepressuredistribution,notonlypointpressures,meansthatthistechnologyis
probablygoingtobeappliedalsotomarinehydrodynamicsinthefuture.
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page38
Aslightlysimilartechnologyispressuresensitivepaint(PSP).Thecolorofthepaintis
changingwiththepressure,whenaspeciallightsourceisused.Thisisalsoarecentmeasurement
technique,developedintheformerSovietUnionandknownintheWestsinceanadvertisementin
AviationWeeklyin1990.Itisprimarilyusedinwindtunneltestingofaircraft.Thepressure
sensitivepaintissensitivenotonlytopressure,buttotemperatureandoxygencontentintheflow,
somethingthatcomplicatestheapplication.

4.6

Velocities
Thevelocityinapointcanbeobtainedbyastraightforwardintegrationofmeasuredacceleration
orbyaderivationofmeasuredposition.Bothmethodsarecommonlyusedforvelocity
measurementsofstructuralcomponents.
Formeasurementoffluidvelocitydifferentprinciplesarepossible:

Basedonmeasurementsofpressure,e.g.pitottubes

LDV

Ultrasonictransducers

Bymeasuringrateofrevolutionofasmallimpeller.

Thetwofirstmethodsarediscussedinthefollowing.

4.6.1 Pitottubes
Thepitottubesensoriscommonlyusedformeasurementofthewakesurveys,flowthroughwater
jetsetc.ThePrandtlpitottubeisshowninFigure4.13.Thepressuredifferencebetweenthetotal
pressureheadatthefront(inpos.Ainthefigure)andthestaticpressureattheside(atpositionB)
ismeasuredbyadifferentialpressurecell.Basicallythepitottubeisapressuredifferencemeasure,
butthevelocityisobtainedfromthewellknownrelation:

p 1 2 U 2

Toimproveaccuracythistheoreticalrelationisnotused,insteadthecalibratedrelationbetween
pressureandvelocitywillbeapplied.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page39

Figure 4.12

Prandtl pitot tube.

Tocovermorepositionsinthesamerun,severalpitottubescanbemountedtogetherasshownin
Figure4.14.
Usingpitottubeswithfiveholeslocatedindifferentangularpositiononasphericalhead,the
velocityinallthreedirectionscanbemeasured,incaseofwakesurveys;axial,tangentialandradial
velocity.

Figure 4.13

Pitot tube arrangement for measurement of velocity in several positions.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page40

4.6.2 LaserDopplerVelocimetryLDV
LaserDopplerVelocimetry(LDV)hasbeenusedformorethan20yearsformeasurementsofflow
aroundshipsandpropellers.Theusehasbeenmainlyforvalidationofpredictiontoolsforvelocity
distributionaroundliftingsurfaces,withinboundarylayersandforwakeflow.GenerallyLDV
measurementsaretimeconsumingandexpensivetoperformandtheuseforcommercialmodel
testinghavebeenverylimited.SeeITTC,(1996)forafurtherdiscussionforapplicationofLDVin
modeltesting.
LDVusestheDopplershiftinthereflectedlightfrequency(color)todeterminethevelocityand
directionoftheflow.Byusingtwobeamsfromdifferentdirection,the3Dvelocityvectorina
singlepointcanbemeasured.Tomeasurevelocityindifferentpositions,thetransmittingand
receivingopticsmustbemoved(seeFigure4.14).Thisiscommonlydoneusinganautomatic
traversingsystem.
OneofthemainbenefitsofLDVisthatitisanonintrusivemeasurementtechnique,whichmeans
thatonedoesnothavetoputanysensorsintotheareawhereonewantstomeasure,thusoneis
notinterferingwiththeflowfieldofinterest.LDVdoalsogiveveryquickresponse,whichmeans
thatitissuitableformeasurementofturbulenceandsimilarlyrapidlychangingflowphenomena.
Formeasurementofaveragevelocityinonedirection,itisgenerallyrecommendedtousepitot
tubesorsimilartechniques.
ForLDVtoworkthereneedstobelightreflectingparticlesdissolvedinthewaterinorderto
providelightscattering.Ifthereisnoparticlesinthewaterthelightfromthelaserwillnotbe
reflected.Infact,oneisnotmeasuringthewatervelocity,butrathertheparticlevelocity.There
mightbesufficientdirtinthewaterfromthestart,butitiscommontohavetoapplyparticlesfor
thepurposeofLDVmeasurement.Thisprocessiscalledseeding.Properseedingisoneofthekeys
tosuccessfulapplicationofLDV.Whentestinginlargefacilities,likealargetowingtank,seeding
mightbeoneofthemainchallenges.Seedingtheentiretankisdifficultandexpensive.Seeding
locallyintheareaofmeasurementmightdisturbtheflow,anditmightbedifficulttoobtaina
reasonablyhomogeneousdistributionofparticles.

Flow
Laser
HeNe
Ar-Ion
Nd:Yag
Diode

Beamsplitter
(Freq. Shift)
Achrom. Lens

PC

Receiving optics
with detector

Transmitting
optics
Gas
Liquid
Particle

Signal
processing
Spectrum analyser
Correlator
Counter, Tracker

Figure 4.14

Achrom. Lens
Spatial Filter
Photomultiplier
Photodiode

Signal
conditioner
Amplifier
Filter

Principles of a LDV measurement system

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page41

4.6.3 ParticleImageVelocimetryPIV
ParticleImageVelocimetry,orPIV,isanonintrusive,opticaltechniqueformeasurementofvelocity
vectors.AsforLDVlaserlightisused,andsimilartoLDVthewaterneedsparticlesseedingto
providelightscattering.WhileLDVmeasuresvelocityinasinglepointatatime,PIVmeasuresthe
velocityinanarea,soitisconsideredafieldmeasurementtechnique.PIVusesalaserlightsheet,
createdbyputtingaspreadinglensinfrontofthelaserbeam.Bytakingtwostereophotographs(or
onedoubleexposurestereophotograph)withveryshorttimeintervalthevelocityoftheparticles
canbedeterminedfromhowfartheyhavemoved.Byusingstereophotographs,themovementin
space,notonlyinasingleplane,canbedetermined.PIVhasbeeninuseforalongtime,butitis
theadvanceindigitalhighspeedvideoandimageanalysisthatquiterecentlyhasmadePIV
interestingtoapplyinregularhydrodynamicsresearch.Inthetimeoneusedfilmbasedpictures
andmoreorlessmanualanalysis,PIVwasextremelytediousandtimeconsuming.Now,semi
automatinganalysisofthedigitalvideoimagesmeansthatlargeamountsofmeasurementdatacan
begeneratedfairlyquickly.Also,theaccuracyhasbeengreatlyincreased.PIVisincreasinglyused
insteadofLDV,sinceitismuchfastertomapavelocitydistributionactuallythedistributioninan
areaiscapturedatoneinstant,andnotoversometime,aswithLDV.

Figure 4.15

Principles of a PIV measurement system

4.7

ForcemeasurementsDynamometers
Theloadcellsusedinmodeltestingareoftendesignedtofitaspecificpurposeandtocoverthe
expectedrangeofloadsduringthetests.Forexamplefortowingtests,veryhighaccuracyin
measuredforcewillberequiredandthetransducerwillbetailormadetomeasureforceinone
directionwithahighaspossibleresolution.Thisiscalledaresistancedynamometerandisstandard
equipmentinatowingtank.InFigure4.15adynamometerformeasurementsofpropellerthrust
andtorquedirectlyonthepropellerhubisshown.Thethrustismeasuredbyaninductiveposition
transducerandthetorquebystraingaugeswhichmeasuresheardeformationonahollowpartof
thepropellershaft.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page42

Figure 4.16

Dynamometer for measurements of propeller thrust and torque.

Transducersformeasurementsofforcesinoneormoredegreesoffreedomareoftenproducedas
apurposemachinedpiecewithstraingaugesgluedtothebasematerial.Atypicaldesignfor
measurementofforcesin3d.o.fs(axialforceandshearforces)isshowninFigure4.17.

Figure 4.17

Example of force transducer based on strain gauges. Axial forces and shear forces
are measured.

Themachiningofthematerialgivesareaswithhighshearstresseswherethestraingaugesare
mounted.Bycarefuldesignofthetransducerthecrosstalkcanbekeptataminimum.Cross
talkmeanscouplingeffectsbetweenthedifferentdegreesoffreedom.Forexampleforthe
transducerinFigure4.17,ifapuretensioninxdirectionisapplied,themeasuredresponsefrom
thetransducerinyandzdirectionisaresultofcrosstalk.Fortransducerdesignitisimportant
withminimumcrosstalk.
Using3transducerofthistypemountedbetweentwoplatesa6d.o.ftransducerisobtained.

4.8

WaveMeasurements
Waveelevationismostcommonlymeasuredbymeansofwaveprobesoftheconductive(or
resistance)type.Avoltageisappliedontoparallelrods.Theresistanceisdeterminedbythelength

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page43
oftherodswhichiswetted(orsubmerged).Bymeasuringthecurrentduetotheapplied
voltageacrosstherodsthewettedpartandhencesurfaceelevation,isdirectlyachieved.Arraysof
threeormoreofthistypeofsensorscanbeusedtomeasurethedirectionaldistributioninshort
crestedwaves.
Waveprobesarealsousedformeasurementofrelativemotionsbetweenthestructureandthe
watersurface.Thewavegaugemaythenbemountedonthestructureandtherelativemotionin
thepointisdirectlyobtained.ExamplesoftransducersolutionsforthispurposeareshowninFigure
4.18.Bothrodsandconductivetapegluedtothemodelareusedforthispurpose.Rodsare
preferableatzerospeedduetoeasiermountingandcalibration,whileconductivetapeisoften
requiredformodelswithforwardspeed,toavoidresistanceandwatersprayfromtherods.

Figure 4.18

Example of wave transducers for measurements of relative motions.

Formeasurementofwavesathighforwardspeeds,thewirebasedprobesdontworkwell,dueto
thesprayandwavemakingofthesurfacepiercingwires.Runupinfrontofthewiresand
ventilationbehindthewiresleadstolargeerrorsinthemeasurements.Forforwardspeedsofmore
than2m/s,thewirebasedprobesshouldnotbeused.Alternativesaremainly:

Ultrasoundwaveprobes

Servoneedlewaveprobes

Ultrasoundwaveprobesworksbysendingoutahighfrequencysoundpulse,andmeasuringthe
timeittakesbeforethereflectedsoundwavereachestheprobe.Thetechniquehasbeenusedfor
alongtimeforlevelmeasurementintanksanddams,butitisnotuntilfairlyrecentlythat
instrumentswithsufficientaccuracyfortowingtankwavemeasurementhasbeenavailable.The
UltraLabsystemsfromGeneralAcousticsisinpracticaluseinthehydrodynamiclaboratoriesatthe
MarineTechnologyCentre.Someofthesesystemshavelimitationswithrespecttoforwardspeed
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page44
andwavesteepness,whilethemostadvancedsystemscoverallconditionsofpractical
interestfortowingtankwork.
Servoneedlewaveprobesconsistofasensingneedlemountedonaservomechanism.Thesensing
needlemeasuresthedegreeofcontactwiththewater,andtheservomechanismmakessurethe
needlehasafairlyconstantsubmergence.Then,theactualwaveheightisfoundbymeasuringthe
positionoftheneedle.Thisisafairlycomplexandfragileinstrument,andtheonlyreasontouse
suchaninstrumentisthecapabilityofwavemeasurementathighforwardspeed.

4.9

DataAcquisition
Thedifferentinstrumentsappliedinthetestsetupareconnectedbycablestoadataacquisition
systemtorecordthemeasureddata.AschematicofadataacquisitionsystemisshowninFig4.1.

4.9.1 Amplifiers
Atypicalanalogtransducersignaliswithoutputinmicrovolts.Itisthereforeamplifiedbyan
amplifier,usuallyto+/10Voltrange.
Dependingontypeoftransducerdifferenttypeofamplifiersisused.Amplifiersforanalogsignals
fromstraingaugeandsimilartypeoftransducerscanhavebridgeexcitationandbridgebalancing
builtin.Thismeansthattheamplifierwillprovidethecurrentandvoltagerequiredforthe
measurementbridge(seesection4.2.1Straingauges),inadditiontoamplificationoftheoutput
signal.Piezoelectricaltransducersrequireachargeamplifier.Inductiontypetransducersrequire
ACexcitation,whichmeansthatthedrivingcurrentofthetransducerisACratherthanDC.Strain
gaugetransducerscanuseeitherACorDCexcitation.

4.9.2 Filters
Filterscanbeeitheranalogordigitalfilters.Analogfiltersareappliedbeforethesignalsare
convertedtodigitalunitsbytheADconverterandmaybebuiltinasanintegratedpartofthe
amplifier.ThedigitalfiltersareappliedaftertheADconversionandcanbeimplementedeitherasa
digitalcircuitorassoftwareinthecomputerusedforanalysis.
Thefiltersremovesignalsatcertainfrequencybands.Dependingonwhichfrequencybandis
removedtheycanbesplitintothreeclasses:

Lowpass,i.e.forremovingofhighfrequencycomponents

Highpass,i.e.forremovingoflowfrequencycomponents

Bandpass

thedifferenttypesareillustratedinFigure4.19.
Analoglowpassfilteriscommonlyusedforremovingofnoiseasthenoiseisusuallyappearingata
significantlyhigherfrequencythanthephysicalmeasuringsignal.ToavoidNyquistphenomena(se
discussionbelow)thecutofffrequencyoftheanaloglowpassfilteredshouldbesettobelower
thanhalfthesamplingfrequencyfS..Thegeneralrecommendationistosetthecutofffrequency
muchlowerthanhalfthesamplingfrequency1/10ofthesamplingfrequencyisthepreferred
value,butitmightmeanthathighsamplingfrequenciesarerequired,andthereforecompromises
areoftenmade.
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page45

Amplitude
Ideal characteristic
Real characteristic

Low pass filter

High pass filter

Band pass filter


Frequency
Figure 4.19

Illustration of different type of filters.

Lowpassfiltersappliedinrealtimecausesatimedelay.Thismightbeillustratedwithanexample
ofaverysimplelowpassfilter,whichiscreatedbycreatingthelowpassfilteredsignalfroman
averageoftheunfilteredsignaloveracertainperiodoftime.Intheexample,showninFigure4.20,
asinusoidalsignalwithperiodof30secondshashadanothersinusoidalsignalwithperiod1.2
secondssuperimposed.Thefilterisimplementedasarunningaverageover2.4seconds,seenin
Figure4.20asthetimeittakesbeforethefilteredsignalappears.Itisclearlyseenfromthefigure
thatthefilteredsignallagsbehind,withhalftheaveragingtime.Thelongertheaverageperiod,the
lowerthefilterfrequency,andthelongerthetimedelay.Iftheaveragingwindowcouldhavebeen
symmetricallyplacedrelativetothetimeofthefilteredvaluethedelaywouldbeavoided.Thiscan
beachievedwhenfilteringanexistingdataset,butnotwhenfilteringinrealtime,sinceitwould
requiretheabilitytolookintothefuture.
Itshouldbenotedthatfilteringisinpracticenotperformedbysimpleaveraging.Thereisalarge
varietyofmethods,whichisnotcoveredhere.SeeforinstanceDunn(2005)foradiscussionof
filteralgorithms.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page46

Figure 4.20 Illustration of time-delay due to filtering

4.9.3 AnalogtoDigital(AD)Conversion
Mosttransducertypes,includingthelargegroupofstraingaugebasedtransducers,areanalog
transducers,meaningthattheyproduceananalogoutputsignal.Torepresentthissignalina
computerithastobedigitized.TheprocessofdigitalizationisperformedintheADconverter.
TheADconverterisusuallyaphysicallyseparateunit;aboxoranextracardforthecomputer.The
lineinpartofacomputersoundcardisanADconvertercustommadeforsound,butreallyquite
similartotheADconvertersusedforothermeasurements.TheADconverterunitwilloftenalso
haveapossibilityfordigitaltoanalogconversionDAsimilartothelineoutofacomputersound
card.TheDAconvertermightforinstancebeusedtoproduceasignalforacontrolsystem.TheAD
convertermightbeintegratedwiththemeasurementamplifier.ThisisthecasefortheHottinger
MGC+digitalmeasurementamplifiersinuseinthehydrodynamiclaboratoriesattheMarine
TechnologyCentre.
TheADconverterunitwill,dependingontheactualmodel,haveacertainnumberofchannels,
whichmeanshowmanysignalscanbeconvertedsimultaneously.LimitationsintheADconverter
willdeterminehowfastthedatacanbesampledthesamplingfrequency.Itmeansthatfor
experimentsrequiringveryfastsampling,specialADconverterequipmentmighthavetobe
acquired.
OneshouldalsobeawarethatsomeADconverterssampleallchannelsatexactlythesameinstant,
whileotherswillsamplethechannelssequentiallyduringthesamplinginterval.Ifforinstance10
channelsaresampledat10Hz,thesequentialADconverterwillsampleasinglechannelevery
1/100second,sothateachchannelissampledevery1/10second,butnotatthesametime.Thisis
usuallynotaproblem,sincethesamplingfrequencyshouldbemuchhigherthanthefrequencyof
thephenomenatobestudied.
TherangeoftheADconvertermeanswhatinputvoltagevaluesareaccepted.MostADconverters
acceptvaluesintherange10V.Iftheinputsignalexceedstherange,anoverflowoccurs,resulting
inaninvalidvalue,andpossiblyinanerrorsituationthatdisturbstheotherchannelsand/or
disruptsthemeasurement.
AnotherimportantfeatureoftheADconverteristheresolutionhowmanybitsareusedto
representonesample.An8bitADconverterwillrepresenttheanalogvaluebychoosingthe
nearestofonly28=256differentvalues.12bitADconvertersarecommonforthelessexpensive
models,producingarangeof4096differentvalues.Expensivemodels,liketheMGC+,use20bit,
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page47
giving1048576differentvales.EspeciallyforlowresolutionADconvertersitisvery
importanttouseasmuchaspossibleoftherangeoftheADconverter.Ifyouhavea8bitAD
converteranduseonly10%oftherangeyougetaneffectiveresolutionofonly25differentvalues,
andthatwillseriouslyimpactyourmeasurementaccuracyyougetasmuchas100/25=4%error
onlyontheADconversion!Touseasmuchoftherangeaspossible,youshouldselectatransducer
withacapacitythatisclosetowhatyouwillmeasure,butwhereyouareprettysurethatthe
capacityisnotgoingtobeexceeded.Thismeansyouneedtoknowapproximatelythemagnitude
ofyourmeasuredvaluespriortotheexperiment.Inadditiontoselectingthecorrecttransducer,
youcanalsoadjusttheamplificationoftheamplifier.Thereasonwhyyoushouldnotjustusea
largecapacitytransducerandturnuptheamplificationisthatthiswillalsoamplifythenoiseinthe
measurementchain,resultinginapoorsignaltonoiseratiointheresultingmeasurement.

4.9.4 Wirelessdatatransmission
Examplesofcaseswherewirelessdatatransmissionisrequiredinatestsetuparemeasurements
onarotatingshaft,forinstanceapropellershaft,andmeasurementsonboardasmallshipmodel.
Therearevariousalternativesandcurrentlyarapiddevelopmentoftechnologyandavailable
products.Afundamentalquestionisatwhatstageshallthedatatransmissiontakeplacebefore
oraftertheA/Dconversion?Forstraingaugeandsimilarsensors,thetransmissionshouldtake
placeaftertheamplification.Thismeansthattheamplifiershouldbeplacedonboardthemodelor
therotatingshaft.Thisagainmeansthatforthesmallestmodels,andindeedfortherotatingshaft,
theamplifiermustbeofaspecial,verysmallkind.Previously,itwasmostcommontotransmitthe
amplifiedanalogsignal,forinstanceusingananalogradiolink.Nowitismorecommontoinclude
theA/Dconverterintheinstrumentationpackageonboardthemodelandtransmitthedigital
signal.Wewillprovidetwoexamplesofthis.
Forafreerunningshipmodeltobeusedinseakeepingtests,MARINTEKisusingaMGC+digital
measurementamplifieronboardthemodel.Theamplifierisrunningon24VDCpower,suppliedby
twomotorcyclebatteriesmountedinseries.TheMGC+iscommunicatingwiththecomputerbyan
ordinaryIPconnection.Usuallythisconnectionisbycable,butincaseofthefreerunningmodel
thecableisreplacedbyawirelesslinkofstandard802.11b/gtype.Theonlydrawbackofthisset
upisthesizeandweightoftheMGC+,whichmeansitcannotbeusedforreallysmallmodels.Also,
theamplifierisbynomeanswaterproof,whichmeansyoucaneasilydestroythisverycostlypiece
ofequipmentifthemodelgetsflooded.
Formeasurementofforcesonapropelleronathrusterorpod,NTNU,MARINTEKandthe
electronicscompanyNorbithavedevelopedameasurementsystemusingonwirelesstransmission.
Inthissystem,asmallamplifierandA/Dunitismountedtotherearendofthepropellershaft.In
thecentreofthisunitisaninfrared(IR)sender/receiverunitofthesametypeusedinmobile
phonesforthephonetocommunicatewithaPCoranotherphone.Directlyinfrontoftherotating
amplifierandA/DunitisasimilarelectronicsunitthatcontainsthesametypeIRsender/receiver.
Powerissuppliedtotheunitontherotatingshaftbymeansofacoilwithonerotatingandone
stationarypart,asseeninFigure4.20.Figure4.21showsthephysicaldimensionsoftheunit.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page48

Figure 4.21

Principle of wireless data transmission from rotating shaft.

Figure 4.22
Sketch of electronics units for wireless data transmission from rotating shaft. The
end of the rotating shaft is seen to the right. To the left is drawn the electronics unit seen from the
rear

4.10

SamplingFrequency
Therecordingofthedataduringthetestisdonebysamplingofthesignal,i.e.thesignalisrecorded
andstoreddigitallyinthecomputerwithacertaintimeinterval,h,betweeneachsampling.1/his
calledthesamplingfrequencyfS.Therequireddatastoragecapacityforeachmeasuringchannelis

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page49
determinedbytheproductTfSnbit.,whereTisthetotalmeasuringtimeandnbitisthe
numberofbitsrequiredtostoreeachnumber.Thesamplingfrequencywillthereforedetermine
thedatastoragecapacityrequired,andthesizeoftheresultfilesisamajorreasonnottoselecta
higherthannecessarysamplingfrequency.
Thesamplingfrequencyhastoreflectthetimedependencyoftheprocesstoberecorded.Ithasto
bequickenoughtocoverallvariations.Ontheotherhandusinghighsamplingfrequencyincreases
theamountofdatacollectedthroughthetestsanddataprocessingandstoragewillbedemanding.
InFigure4.22isshownthatsamplingcanleadtogenerationoferroneouscomponentsinthe
record.Arelativelyhighfrequentsinusoidalissampledatasamplingfrequency1/h.Theapparent
signalstoredinthecomputerisamuchmorelowfrequentsinewave,whichdidnotexistinthe
originalsignal.Theinformationintheoriginalsignalislostandcannotberestoredinthe
subsequentanalysis.

Figure 4.23

Illustration of consequence of too low sampling frequency

Inordertodefinetheamplitudeandphaseofapuresinewavewetheoreticallyneedaminimumof
2samplesprcycle.Inotherwordsthehighestfrequencythatcanbedeterminedfromthesignalis:

fC

2h

ThisfrequencyiscalledtheNyquistfrequencyofthesamplingprocess.
IftheoriginalsignalcontainscomponentsabovefCtheywillbefoldedbackintothelow
frequencypartofthespectrumandappearthereaserroneous,nonphysicalcomponents.Thisis
showninFigure4.23.Twomethodscanbeusedtoavoidthiserror:
1. ChooseasufficientlyhighfSthatallpossiblefrequenciesinthesignalarecorrectly
recordedinthesampling
2. Applyanaloglowpassfilteringofthesignal,removingallsignalcomponentsat
frequencyabovefCbeforethesignalissampled.
Thelattermethodismostcommonlyappliedinpractice.Forpracticalmeasurements,itis
recommendedtosetthecutofffrequencyofthelowpassfiltersignificantlylowerthantheNyquist
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page50
frequency.Therecommendationistosetthelowpassfiltercutofffrequencyto1/10of
thesamplingfrequency.Sincethelowpasscutofffrequencymustbesufficientlyhightoretainall
frequenciesofinterest,followingtherecommendationleadstohighsamplingfrequencies.Thus,
compromisesmightbemade,butitisstronglyrecommendedtokeepthefiltercutofffrequency
wellbelowtheNyquistfrequency.

fc
f

Figure 4.24

Effect of Folding

Wavefrequencymotionswillfollowthewavefrequency,andamoderatesamplingfrequencyis
required.Typically,alowpassfilterfrequencyequaltominimum10timesthewavefrequency(or
peakfrequencyincaseofirregularwaves)shouldbeappliedforthispurpose.
Impactloadsisanexamplewhereveryhighsamplingfrequencyisrequired.Thisisespeciallythe
caseifthelocalriseangleatthepositionofimpactbetweenmodelandwavesurfaceiscloseto
zero.Anexampleofmeasuredpressureonahorizontalplatedroppedverticallytowardsawave
crestisshowninFigure4.11.Thetestswerecarriedouttosimulateimpactloadsagainsta
catamaranwetdeck.Therisetimeisseentobeabout0.0001sandasamplingfrequencyof100
kHzwasrequiredtocoverthepressurepeak.Thetimedurationoftheimpactisveryshortanditis
associatedwithveryhighlocalpressurepeaks.Thiswillalsorequirethattheamplifiershavea
sufficientlylargefrequencyrangewithlinearrelationbetweenmeasuredandamplifiedresults.For
thetransducersverylowrisetimeandhighresonancefrequencyarerequired.
Thisexampleillustratethatthepropertiesofequipmenttobeused,includingsensors,amplifiers,
filtersanddataacquisitionunithastobecarefullyselectedbasedontheactualbehaviourateach
application.Agoodunderstandingofthephysicsofthetestssituationisthereforeofvital
importancetodecidetypeofequipmentthatshallbeusedandhencetoachievereliableandhigh
qualitytestresults.

4.11

LengthofRecords
Requireddurationofthetestswilldependontypeoftesting.Forregularwavesashortdurationof
testisusuallyused.Typicallyonly10cyclesisusedfortheanalysis,butmakesurethattheactual
testissufficientlylongtoestablishasteadystateresponse.Fordecayteststherequirementsto
durationissimilar.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page51
Forirregularwavesthestatisticalaccuracyofmeasureddataisimprovedbyincreasingthe
lengthoftherecord.Therequireddurationdependsonthefollowingparameters:

Theperiodofthemostlowfrequentphenomenawhichoccurinthetests

Thesystemdamping

Therequiredstandarddeviationofthequantitiesdeterminedbythestatisticalanalysis

Incommercialtestingthereisaverysimpleruleofthumbthatthetestdurationshouldpreferably
beatleast100timestheperiodofthemostlowfrequentphenomenatobeinvestigated.For
testingwhereonlythewavefrequencyphenomenaareofinterestthisgivesarequiredtimeof
typically1000sfullscaletime.Forcaseswhereslowdriftmotionsareimportant,naturalperiodsin
theorderof15mincanbeexpected.Thiswillrequiretestdurationupto10hours.Inpractice,
often3hourdurationoftestisusedforthiscase.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page52
Referringtothespectralanalysisrequirementshowsthatthereexistcertainrelationships
betweenthefollowingquantities(seeBendatandPiersol(1966)forfurtherdetails):

Bh

T Nh

where:
B;

thedesiredfrequencyresolutionintheresultingspectrum(inHz)

h;

samplinginterval(ins)

m;

numberoftimeshiftsinthecalculationoftheautocorrelationfunction(fromwhich
thespectrumiscalculated,seechap9.5.2)

N;

numberofsamples

T;

lengthofrecording(ins)

normalisedstandarddeviationofthespectrumcalculation

Letusconsidermodeltestswithmeasurementsoftheheavemotionsofashipinirregularwavesas
anexample.Assumingthefollowingdata:

SpectralpeakperiodTP=10s.

ResonanceheaveperiodT3=8s.

Scaleratio=40.

Forthiscasewecanassumethattherewillbealmostzeroenergyoftheresponsespectrumfor
frequenciesabove0.25Hz.TheNyquistfrequencycanthenbesettofC=0.5Hz.Wantedfrequency
resolutionintheresultingspectrumissettoB=0.02Hz.Thisgives:

Samplinginterval

1
1 .0 s
2 fc

Numberoftimeshifts

1
50
Bh

Numberofsamples

Lengthofrecord

T Nh 2222s 22 min

2222

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

4.12

Page53

Calibration
ThedataacquisitionprocessdescribedsofargivesasignalinVolts(ormillivolts).However,whatis
neededisasignalintherelevantphysicalunit.Toobtainalinkbetweenthemeasuredvoltage
signalandthecorrectphysicalvalue,weusuallyapplycalibration.Theprocessofcalibrationisto
applyaknownloadtothetransducerandmeasuretheoutputvoltage.Thisisdoneforanumberof
differentloads,coveringtheexpectedrangeofmeasuredvalues.Then,acalibrationcurvecanbe
drawn.Foreaseofapplication,mosttransducersaremadesuchthattheywillgivealinearrelation
betweenloadandoutputsignal.Thecalibrationfactorwillthenbetheslopeofthestraightlinefit
tothemeasuredcalibrationdata(seeFigure4.24).
Tobeabletoaapplyaknownloadinaneasyandwellcontrolledmanner;thecalibrationofeach
transducerisusuallydonebeforethetestissetupbeforethetransducerismountedintheset
up.Whendoingthecalibrationthisway,onemustmakesurethatthemountingofthetransducer
inthetestsetupdoesntalterthecalibrationfactor.Also,becauseoneismeasuringtheamplified
signalinVolts,theamplifiersettingswillinfluencethecalibration.Thus,oneshouldmakesureto
usethesameamplifiersettings,andpreferablyalsothesameamplifierasintherealtestsetup.
Itisfurthermorerecommendedtoperformacheckofthecalibrationswhenthetestsetupis
completed.Thisensuresthatthecalibrationfactorshaventbeensignificantlychangedbythe
mounting,andalsothatallconnectionsarecorrect.Anexampleofsuchacheckiswhentheship
modelisconnectedtothecarriageandtheresistancedynamometerinpreparationforatowing
test,aknownforceisappliedtothemodelinthedirectionoftowingandtheresultingoutputof
theresistancedynamometeriscomparedwiththeknownforce.

Z Force

25

load [N]

20
15
y = 30.829x + 0.1564
10

R = 0.9997

5
0
0.0

0.2

0.4

0.6

0.8

1.0

output [V]

Figure 4.25 Example calibration curve for a force transducer. The calibration factor obtained from
this curve is 30.829 N/V.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

4.13

Page54

Zeroing
Thecalibrationdescribedabovewillensurethattheproportionalitybetweentheoutputsignaland
thecorrectphysicalpropertyiscorrect(rememberthatthecalibrationfactoristheslopeofthe
curverelatingoutputinVoltstothephysicalunit).However,wealsoneedtomakesurethatthe
absolutevalueisknown.Thisisdonebydefiningareferencelevelandreferringallmeasurements
tothislevel.Typically,thereferenceleveliszero,andinthefollowing,weassumethatthisisthe
case.
Beforethetest,wemeasurethezerolevelofallrelevanttransducers.Forinstanceforaresistance
test,wemeasurethezerolevelforresistance,carriagespeed,andsinkageforeandaftwiththe
modelatrest.Theoutputmeasuredinthisconditionisstored,andlatersubtractedfromthereal
measurementsatthewantedforwardspeeds.ThedataacquisitionprogramsinuseattheMarine
TechnologyCentrehavezeromeasurementfunctionalitybuiltin.

4.13.1 Bridgebalancing
Thezeromeasurementoutlinedaboveislikeanordinarymeasurement,involvinganaverageover
atimeselectedbytheexperimenter.Inadditiontothiskindofzeroing,themeasurementamplifier
usuallyhasadditionalmeansofzeroingtheoutput.Themostbasiconeisbridgebalancing,whichis
toadjustthebuiltinresistancesinordertomakesurethattheWheatstonemeasurementbridge
forthetransducergiveszerooutputintheselectedcondition.Itisimportanttobalancethebridge
beforestartingthemeasurements,asthismakessubsequentmeasurementsmoreaccurate.
Inaddition,theamplifiermighthaveatareorzeroingfunction,whichinprincipleissimilartothe
zeromeasurementdescribedabove,exceptthatitbasesthezeroreadingonasingle,oravery
limitednumberofsamples.Thus,thiskindofzeroislessaccurate,duetotheriskofenvironmental
disturbances.

4.13.2 Temperaturedrift
Zeroreadingsaredonemanytimesduringameasurementcampaignbecausemanytransducers
haveatendencytodrift.Driftinatransducermeansthatforaconstantload,theoutputchanges
slowlywithtime.Zeroingmustbeperformedsufficientlyoftentoreducetheerrorcausedby
transducerdrifttoanacceptablelevel.Ifthedriftisquickcomparedtothelengthofa
measurement,itmeansthetransducerisnotsuitableforthetest.Transducerdriftforstraingauge
basedtransducersisusuallyrelatedtotemperaturechangesinthetransducer.Whenthe
temperatureincreases,thematerialtowhichthestraingaugeisattachedwillbeelongated,sothe
straingaugewillgetareadingthatisfalse.Astraingaugeemitssignificantamountsofheat
comparedtoitssmallarea,soshortlyafterthetransducerisconnectedtotheamplifier(orthe
amplifierturnedon),thetransducerwilltypicallydriftquitesignificantlyduetoheatingupofthe
transducermaterial.Thedriftduetoheatfromthestraingaugewillreachastableequilibrium
withinafewminutes(mightbeuptoabout30minutes),butwhentheenvironmentaroundthe
transducerchanges,forinstanceduetoachangeoftheairvelocityaroundthetransducer,the
equilibriumischangedanddriftwillagainoccur.Especiallyfortransducersimmersedinwater,the
changeofcoolingofthetransducerduetochangesinwatervelocitycancauseseriousproblemsfor
theaccuracy.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page55

PHYSICALMODELLING

5.1

General
Designandconstructionofeffectivetestmodelsisanimportantpartofthemodeltestingprocess.
Typicalmaterialsusedforconstructionofshiphullmodelsarewax,wood,aluminum,glass
reinforcedplastic(GRP)andfoammaterial.UseofGRPwillrequirethataplugismadewiththe
exactfinishedshapeofthehull.Thetypeofconstructionmaterialtobeusedwilldependonshape,
modelsize,weightandstrengthrequirement,andcost.Forshiplikemodelscomputercontrolled
millingmachinesarecommonlyusedformodelproduction.
Foralmostalltypeofmodeltesting,veryaccuratemodelingofthepartofthemodelgeometrythat
isexposedtowaterwillberequired.Fortestingwhereviscousforcesareimportantalsothe
roughnessofthesurfacehastobecarefullycontrolled.
Theselectionofscalewillbelimitedbyseveralfactors:

Theexperimentalfacilityavailable

Requirementsfrominstrumentationandphysicalmodelingoftheactualstructure

Requiredscaletoavoidseverescalingeffects,veryoftenconnectedtoviscousforces
(Reynoldsnumber).

Costofmodelsandtestexecution

Ingeneralatoosmallscalewillgiveproblemswithscaleeffectsandmeasuringaccuracy,while
largemodelswillbecostlyandcanbedifficulttohandle.Ifthemodelsgetlargecomparedtothe
sizeofthetesttank,blockageeffectsarelikelytooccur,whichmeansthatthetankwallsinfluence
theresults.Requirementstocorrectlyscaledmassdistributionand(whenrelevant)structural
elasticitywillbemoredifficulttosatisfyforsmallmodels.

5.2

RigidModels
Fordynamictestingwherethemodelisfreelymovingandtheinertiaforceswillbeofimportance,
themassdistributionneedstobecorrect.Inpracticaltermsthefollowingrequirementsshouldbe
satisfied:

Totalmass

Momentofinertiaexpressedthroughthegyrationradius,rxxryyandrzz

Longitudinalandverticalpositionofcentreofgravity

Ifinternalloadsshallbemeasured,themassdistributionmustalsobecorrectly
modelled

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page56
Themassisdeterminedsimplybyweighingofthemodel,orbyensuringthatthemodelis
floatingatthecorrectwaterline.
ApracticaltoolfordeterminingthepositionofCoGandmomentofinertiaofamodelistousethe
pendulumtest.Themodelisarrangedinapendulum,freetorotateaboutapointA,asshownin
thesketchinFigure5.1.Theperiodofaphysicalpendulumisgivenfrom:

T0 2

Mgh

whereMisthemassofthemodel,IisthemomentofinertiareferringtothepointA,Gisthe
positionofcenterofgravity,histhedistancefromAtoGandgisgravity.Theaboveequationcan
berewrittenas:

4 2 I MgT0 h 0
2

Figure 5.1
Sketch of a Pendulum for determination of centre of gravity and moment of
inertia

InthependulumtestonefirstmeasuretheperiodT0.Thentwoadditionalmasses,m,arefittedat
eachendofthependulumatadistanceafromtherotationcentreA.ThenthependulumperiodT1
ismeasuredwiththeadditionalweightsfitted.Wenowhave:

4 2 ( I I ) ( M 2m) gT1 ( h h) 0
2

where I 2ma 2 and h

2mh
.Insertingintotheaboveequationgives:
( M 2m)

4 2 ( I 2ma 2 ) MmgT1 h 8 2 ma 2
2

Wenowhavetwoequationswithtwounknowns,Iandh.Thesolutionis:

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page57

8 2 ma 2

2
2
Mg (T1 T0 )
2

2ma 2To

2
2
T1 T0

Toobtainthemomentofinertiaofthemodelreferringtothecentreofgravitywehave:

I m I Mh 2

Theradiusofgyrationcanbeadjustedtotherequiredvaluesbymovingtheinternalballast
weights.Itisthereforeimportanttohavesufficientamountofballastweightstobeabletoachieve
thecorrectweightdistribution.
Whenthepositionofcentreofgravityisavailable,themetacentricheightisalsoeasilyobtained.It
iscommonpracticetomeasurethemetacentricheightalsobyastaticinclinationtestafterthe
modelisinthewater.Ifanydeviationfromtheresultsfromthependulumtestisobserved,the
inclinationtestisusuallyconsideredtobemostaccurate.
Whenthemodelisdividedinseveralsectionsformeasurementofgloballoads(shearforcesand
bendingmoments)ingiventransversecuts,itisnotsufficientthatthetotalmodelhasthecorrect
massandmomentofinertia.Forthiscaseeachsectionoftheshipmusthavecorrectmassand
momentofinertiatogetcorrectforces.Thependulumtestshavethentobecarriedoutforeach
section.

5.3

ElasticModels

5.3.1 General
Fortestconditionswheretheelasticdeformationofthemodelisimportanttheelasticproperties
ofthemodelneedtobecorrectlyscaled.Exampleswherethiswillbethecaseare:

Marinerisers(bendingstiffness)andloadinghoses(bendingstiffnessandaxial
elasticity)

TethersforTensionLegPlatforms(bothaxialandbendingstiffness)

Mooringlines(axialstiffness)

Springingandwhippingofships,bothmonohullsandcatamarans(bendingstiffnessand
torsionalstiffness)

Floatingbridges

Fishfarmingplants

Seismiccables

Thisimpliesthatforoffshoreandcoastalstructuretestingoneormoreelementswillusually
requiremodellingofelasticproperties.Foratestcasewheretheelasticityisnotimportant,the
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page58
modelismadesufficientstifforasstiffaspossibletoavoidanyartificiallyhydroelastic
effectsinthemodel.
Thegeneralscalinglawstobeappliedforelasticmodellingisdiscussedinchap.2.5.Asoutlined,a
directgeometricalscalingusingmaterialwiththesamemodulusofelasticity,givesamodelthatis
timestoostiff.Thisimpliesthatinpracticalmodellingtheproblemistogetthemodelsufficiently
elastic.
Forthedynamicresponseofanelasticstructurethemassdistributionandstructuraldamping
shouldalsobecorrectlymodelled.Itisthereforearequirementthatthemodellingdonotintroduce
additionaldamping.Materialsthatshowhystereticbehaviourshouldthereforenotbeusedfor
modelling.Othersourceforunwanteddampingcanbefrictionalforcesbetweendifferent
componentsinthebuildupofthemodel.

5.3.2 SlenderStructureModelling
Afrequentlyusedmethodformodellingoflongandslenderstructures(eg.riserortethers)isto
applyasteeloraluminiumcorewithdimensionsdeterminedtogivecorrectbendingstiffness.
Aroundthecoreisfittedbuoyancymaterialtoobtainthecorrectoutergeometryandweight.Using
thismethoditisimportanttoavoidmotionsandresultingfrictionbetweenthecoreandthe
buoyancymaterialtoavoidintroducingartificialstructuraldamping.InFigure5.2anexampleof
modellingofalongandslendertubebridgeusingthismethod,isshown.Frictionforcesbetween
coreandouterbuoyancyareinthiscaseavoidedbyconnectingthebuoyancyfoamtoanouter
pipe,whichisweldedtothecore.Thegapbetweenthebuoyancyelementsensuresthatthe
bendingstiffnessisnotinfluencedbythebuoyancyelements.

Figure 5.2

Example of elastic modelling of marine riser.

5.3.3 MooringLinesmodelling
Theaxialelasticityofmooringlinescanbemodelledbyintroducingaxialspringsinoneormore
positionsalongthelines.Thiswillbeimportantespeciallyformooringsystemincludingfibreropes,
butalsoforconventionalmooringsystemdesignsconsistingofchainand/orwiresegments,the
axialelasticitycanbeimportantforthetotalrestoringstiffnessandshouldthereforebeaccurately
modelled.

5.3.4 Shipmodelling
Forelasticmodellingofships,threedifferentsolutionshavesofarbeenused.(SeeMaeda(1991)):
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page59

Backbonemodel

Fullyelasticmodel

Hingedmodel

InFigure5.3thetwofirstalternativesareshownschematicallyforamonohull.Forthebackbone
modelling,theelasticityofthemodelisrepresentedbyanelasticbeamtowhichrigidsegmentsare
connected(similartothesolutiondiscussedabove).Usingthismodellingtechniqueitisrelatively
easytomodelthestiffnessandmaterialsassteeloraluminiumwithstableandwelldocumented
propertiescanbeused.Furtheritiseasytomodifythestructuralproperties,andthestructural
dampingislow.Oneproblemwiththismodellingisthegapbetweenthedifferentsections.They
maybeclosedusinganelasticmembrane,butitisdifficulttoavoidtransferoftensiontroughthe
membrane.Ifthegapsareopen,eachsectionhastobebuiltwatertight.Furtherthedynamic
pressureinthegapsmaytosomeextentinfluencetheresults.Forshipmodelswithforwardspeed
thegapswillgiveadditionalresistanceduetotheinfluenceontheflowfieldaroundtheship.

Figure 5.3

Methods for modelling of an elastic monohull

Thefullyelasticmodelisbuiltupusingcrosssectionswithoneortwolayerswithdifferentelasticity
asshowninFigure5.3.Glassfibreresinincombinationwithafoammaterialcanbeused.The
thicknessoftheinnerresinlayercanbevariedtoachievethecorrectelasticity.Thismethodof
modellingavoidsthegapproblems,butitisdifficulttoachievethecorrectbendingstiffness
distribution.Further,theuseoffoammaterialintroducessomehystereticeffectsandthestructural
dampingmaybetoohighforthemodel.Fortestingofspringingresponsethisiscriticalforthe
results.Forwhippingresponsethestructuraldampinglevelislesscritical.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page60
Whilethebackboneandfullyelasticmodelisbestsuitedformonohullsthehingedmodel
solutionhasalsobeenusedforcatamarans,seeHermunstadet.al.(1995).Anexampleofelastic
modellingofacatamaranusingthehingedmodelisshowninFigure5.4.Eachhullisdividedinto
threerigidsegmentsthatareconnectedbysprings.Thespringsareslendersteelbeamswith
dimensionsdeterminedtogivecorrectbendingandtorsionalstiffness.Thetwohullsareconnected
troughthreetransversespringsasshowninthefigure.ThehullsegmentsaremadeofFoam/GRP
similartowhatisusedforstandardrigidmodelproduction.Tomakethesegmentsstiff,an
aluminiumframeismountedwithineachsection.Rubbermembranesareattachedbetweeneach
segmenttomakethemodelwatertight.
Ingeneral,arequirementforcorrectlyscaledelasticityofashipmodelwillsignificantlyincreasethe
complexity(andhencethecost)formodelproductionandalsofortestexecution.Inaddition
detailedinformationabouttheelasticpropertiesofthefullscaleshipwillberequired.

Figure 5.4

Hinged model of an elastic catamaran

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page61

CONVENTIONALSHIPTESTING

6.1

General
Conventionalshiptestingincludesthefollowingtypeofmodeltesting:

Towingtestforresistancemeasurements

PropellerOpenWaterTesttodeterminethepropellercharacteristicsinopenwater
(withoutashipmodelpresent)

Propulsiontestintowingtank,eithertodeterminethepropulsionpowerdirectly,orto
findthepropulsionfactors(thrustdeduction,effectivewake,andrelativerotative
efficiency)whendoneincombinationwithresistancetestandpropelleropenwater
test

Wakesurveytowingtestwherethevelocitydistributioninthepositionofthe
propellerismeasured.Thewakedistributionresultingfromthistestisimportantinput
toacavitationtestsandforthepropellerdesign.

Streamlinepainttestusuallyperformedwithworkingpropeller(s)todeterminethe
flowpatternaroundthehull,inordertodetectflowseparationandoptimiseinclination
ofpropellershaftbrackets,positionofbilgekeels,anddirectionoftunnelthruster
openinggrids.

Cavitationtestsforidentificationofpropellercavitationproblems.

Manoeuvringtestsforcheckofmanoeuvrability

6.2

Towingandpropulsiontestsintowingtank
Thetraditionalpurposeofshipmodeltests,andstillthemainactivityofcommercialtowingtank
testing,istopredictthespeedpowerrelationforashipbeforeitisbuilt.
Allshipbuildingcontractscontainstrictrequirementstothespeedthenewshipshalldoforagiven
enginepower.Failuretomeetthecontractualspeedrequirementleadstoheavyfinesfortheyard,
andultimatelythattheshipownermayrefusetotakeownershipofthevessel.Thus,itisimportant
fortheshipyardtodeterminebeforethebuildinghasstartedhowfasttheshipwillgoonthetrials.
Evenwhenmaindimensionsarefixed,thespeedperformanceofthedesigncanbeimproved
significantly(upto20%onpower)bycarefuldesignofbow(includingbulb!),sternandpropulsion
arrangement.Thus,itiscommontomakemodificationstothedetailsoftheshipdesignbasedon
theobservationsduringthemodeltests.
Conventionaltowingtests,propulsiontestingandscalingofresultstofullscalearediscussedin
detailsinthebasiccourseinHydrodynamics.ReferenceismadetoSteen(2011).Ashortsummary
willbegivenhere.

6.2.1 Towingtestsforresistancemeasurements
Inatowingtest,themodelisconnectedtothetowingcarriagethroughaforcetransducer,
commonlycalledresistancedynamometer,sothatallforcesinthedirectionoftowingistakenby
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page62
thetransducer.Inaddition,themodeliskeptonastraightcoursebysocalledtrimposts,
whicharedevicesthatonlyprovideslateralforces,butleavesthemodelfreetoheave,trim,and
surge.Usually,thetrimpostshaveanarrangementformeasurementofthesinkageforeandaft.
Thespeedisusuallymeasuredasthespeedofthecarriage.Asketchofthesetupfortowingtests
isshowninFigure6.1

Towing Carriage
Measurement of:
Model resistance RTm
Model speed
Sinkage fore and aft

Resistance
Dynamometer

Ship model
Flexible connection
Figure 6.1

Test set-up for towing tests

Theprocedureforthetestistoacceleratethemodeltowantedspeed,thenmeasurethe
resistanceatconstantspeedforatleasttenseconds,andthengraduallystopthecarriage.For
heavymodels,itiscommonpracticetohavealockingmechanism(clamp)tounloadthe
resistancedynamometerduringaccelerationanddeceleration.
Intheresistancetest,themodelisusuallyequippedwithrudderandpropellershaft(s),butwiththe
propeller(s)replacedwithdummybossesmadeoflead,sothattheweightofthedummyboss
equalstheweightofthepropellermodel.Furthermore,themodelisequippedwithaturbulence
stimulator,tomakesurethattheflowoverthehullisturbulentalsoatlowspeeds.Theturbulence
stimulatorisusuallyplaced5%ofLppaftofFP.Theturbulencestimulatoriscommonlya1mm
diametertreadgluedtothehullnormaltotheflowdirection.Forhighspeedmodels,studsare
commonlyusedforturbulencestimulation.Studsare23.5mmdiametercylindersmounted
normaltothehullsurface,protruding22.5mmfromthesurface,withabout2025mmdistance
betweenthestuds.Sandstripsarealsoused.Thatisa510mmwidestripofsharpedgedsandwith
grainsizeofabout0.5mmwhichisgluedtothehull.Turbulencestimulationmightinadditionbe
requiredonappendicesandseparatepartsofthehullprotrudingoutoftheboundarylayerofthe
hull,likeasonardome,orthekeelofasailingyacht.Onappendices,sandstripsarethemost
commonformofturbulencestimulation.

6.2.2 Propulsiontests
Inthepropulsiontest,themodelisthesameasfortheresistancetest,exceptthatthedummy
bossesarereplacedwithgeometricallyscaledmodelpropellers.Thepropellersaredrivenby
electricmotors.Betweentheelectricmotorsandthepropellersaremountedpropeller
dynamometers,whichmeasurepropellerthrustandtorque.Inaddition,thepropellerrevsare
measured.SeeFigure6.2forasketchofthesetup.
Forconventionallyshaftedpropellers,thepropellerdynamometersarelocatedinsidethehull.
Thus,thefrictioninthepropellershaftsealsandbearingswillinfluencethemeasurementoftorque
(sinceonemeasurethetorqueinsidethehull,butwanttoknowthetorquedeliveredtothe
propeller).Thus,itiscommontodomeasurementsofthesocalledidletorque,whichisthetorque

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page63
measuredwithonlythepropellerbosspresent.Thisisdoneforseveralpropellerspeeds,so
thatcorrectionsofthemeasuredpropellertorquecanlaterbedonebyinterpolation.
Sincethefrictionalresistancecoefficientislargerinmodelthaninfullscale,itisrequiredtopartly
towthemodel,inordertoobtainthecorrectloadingofthemodelpropeller.Thiscanbedonein
twodifferentways:
1. ApplicationofaconstantforceFD,byusingaweightandpulleysystemasindicatedin
Figure6.2.Thepropellerrevshastobeadjustedsothatthemodelreachesthespecified
speed.Thisiscalledcontinentalmethodofpropulsiontest.
2. Themodelisconnectedtotheresistancedynamometerinthesamewayasinthe
resistancetest.Thespeedisthendecidedbythecarriagespeed.Thepropellerrevsare
variedtoobtainatowingforceequaltoFD.Itisusuallyrequiredtodoaboutthreerunsin
thetankateachspeedtofindthepropellerrevsthatgiveatowingforceequaltoFD.Thisis
calledtheBritishmethodofpropulsiontest.Sinceitrequiresmoreruns,itislessused.
However,thefactthatitgivesresultsfordifferentvaluesofFDmeansthattheresultscan
easilyberescaledtodifferentscaleratiosanddifferentpoweringpredictionmethods.
ThetowingforceFDisfoundfromthedifferenceinresistancecoefficientsinmodelandfull
scale.Thisdifferenceisprimarilyduetothedifferenceinfrictionalresistancecoefficients,but
differencesinairresistance,transomsternresistance,andappendageresistancemightalsobe
included.

Towing Carriage Measurement of:


Torque Q
Thrust T
RPM
Model speed
Sinkage fore and aft
Dynamometer

Figure 6.2

Tow rope FD

el. motor

Test set-up for propulsion tests. Continental method

6.2.3 Propelleropenwatertest
Thepurposeofthepropelleropenwatertestistomeasuretheperformanceofthepropelleralone,
withoutthehullpresent.Whencombinedwiththeresultsoftheresistanceandpropulsiontests
onecanidentifytheinteractioneffectsbetweenthepropellerandhull,likeeffectivewake,relative
rotativeefficiencyandthrustdeduction.
Inthepropelleropenwatertest,thepropellerismountedonapropelleropenwater
dynamometer,whichislikeanextremelyslenderthrusterwithpullingpropeller(seesketchin
Figure6.3).Thepropellerisequippedwithadummypropellercap.Theresistance(thrust)and
torqueofthedummypropellercapandpropellerhubismeasuredinseparaterunsandsubtracted
fromtheresults,sothatoneeffectivelygetstheperformancecharacteristicsofthepropeller
blades.Onemeasuresthepropellertorque,thrust,andrevs,aswellasthespeed.Testsareusually
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page64
doneatconstantpropellerrevs,varyingthespeedfromzerotoaspeedthatgiveszero
propellerthrust.

Measurement of:
Torque Q
Thrust T
Rate of revolutions n
Speed V

Water speed V

Figure 6.3

Test set-up for propeller open water tests.

0.7

T
n2 D4
Q
KQ
n2 D5
K J
T
2 KQ
KT

KT
10* KQ

0.6

etta0

KT, 10*KQ, 0

0.5

0.4

0.3

0.2

0.1

0
0

0.2

0.4

0.6

Advance number J

Figure 6.4

V
nD

0.8

Propeller open water diagram, as resulting from a propeller open water test.

Theresultsoftheopenwatertestarecompiledintoanopenwaterdiagram,asshowninFigure
6.4.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page65
Forshipswithductedpropellers,theductsarenotpresentintheresistancetest,but
includedinthepropelleropenwatertests.Inthepropelleropenwatertest,theductthrustis
measuredseparately.
Forshipswiththrusterpropulsion,theentirethrusterunitisusedforthepropelleropenwatertest.
Thus,thethrusterunititselfisconsideredasthepropulsor,andtheresistanceofthethruster
bodyshouldnotbeincludedintheresistancemeasuredintheresistancetest.

6.2.4 Poweringperformanceprediction
Thepurposeoftheresistance,propulsionandpropelleropenwatertestsismainlytodetermine
therequiredshaftpowerofthefullscaleshipatdifferentspeeds.TheITTCstandardforpowering
predictionisgiveninAnnexE.Adetaileddescriptionofpoweringperformancepredictionisalso
foundinSteen(2011).Thedetailsofthepoweringpredictionmethoddiffersbetweentowing
tanks,sothatforinstancethemethodinuseatMARINTEKisnotexactlythesameasthemethod
describedinAnnexE.Abriefoutlineoftheperformancepredictionmethodisgivenhere.
Withtheresultsfromthementionedtestsavailable,thepoweringperformancepredictionisdone
asfollows:
1. Themodelscalevaluesofthepropulsivecoefficientsthrustdeductiont,relativerotative
efficiencyR,andeffectivewakewarefoundfromanalysisofthepropulsiontestresult,
usingalsotheresistanceandopenwatertestresults.
2. Fullscalepropulsivecoefficientsaredeterminedbyassumingthatthrustdeductionand
relativerotativeefficiencyarefreeofscaleeffects,whilethewakeisscaled(atleastfor
singlescrewvessels).
3. Theresistanceisscaledtofullscale,byseparatingviscousandresidual(orwave)
resistance,andscalingthetwopartsdifferently.
4. Thepropelleropenwaterdiagramisscaledtofullscale.(Thisisnotdonebyallmodel
basinsMARINTEKisusingthemodelscaleopenwaterdiagramalsoinfullscale)
5. Thefullscalepropulsionpointisdeterminedbyinterpolationinthefullscalepropeller
openwaterdiagram.ThepropulsionpointmeansthevalueofJ,whichbyuseofthefull
scaleopenwaterdiagramgivesthevalueofKQ.
6. Thefullscalepropellerrevsandtorquearecalculatedfromthefullscalepropulsionpoint
valuesofJandKQ.Thenpoweriscalculated.

6.3

Cavitationtunneltests

6.3.1 General
Testingofcavitationproblemswillrequireafacilitythatgivesthepossibilitytotestwithcorrect
cavitationnumber.Cavitationtestingfacilitiesisdescribedinchapter3.3.Thepurposeofcavitation
testingcanbeoneormoreofthefollowingreasons:

Cavitationinducederosionofpropellerblades.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page66

Effectofcavitationonpropulsionefficiency

Vibrationsandnoise.

Thepressurefluctuationscausedbythecavitationwillintroducepressurevariationsonthehull,
whichagaincanleadtohullvibrationsandgenerationofnoise.Anotherproblemisthatthe
cavitationinducedpressurepulsescanintroducenoise,whichisdisturbingtopassengers,or
affectingacousticpositioningsystemorothertypeofinstrumentation.Thistypeofnoisecanalso
beaproblemforfishingvesselsbythedirecteffectonthefish.

6.3.2 Modellingrequirements
Thecavitationtesthastobecarriedoutinsuchawaythatthemainforcesastrustandtorqueare
similarinmodelandfullscale.Inprincipalthiswillrequiregeometricalsimilarityofbothpropeller
shapeandshiphull.Duetothelimiteddimensionsofmostcavitationtunnels,therewillbea
problemwiththetunnelwalleffectsandwiththefreesurface.
Kinematicsimilarityrequiresthattheinflowvelocitydistributiontothepropelleriscorrectandthat
thetestsarerunatsameadvanceratioasinfullscale:

JM

VM
V
JF F
nM DM
nF DF

whereVisthespeedofadvanceofthepropeller,nistherateofrevolutionandDisthepropeller
diameter.
Forthedynamicsimilarityrequirement,itisoftenassumedthatthepropellerissufficientlywell
submergedtoavoidsurfacewaves.TherequirementtoequalFroudenumbercanthanbe
abandoned.TheReynoldsnumberhastobehighenoughtoensureturbulentboundarylayeratthe
propellerblades,astheactualflowregimewillinfluencethecavitation.Asapracticallowerlimit,

Re

UL

5 105 isoftenused.Lisheretakenasthechordlengthofthepropellerbladeat0.7

radius.
Toensurethesameriskofcavitationinmodelandfullscale,equalityincavitationnumberis
requiredasdiscussedinchapter2.6.
TheWebernumber(seechap2.3.5)mayalsohaveaneffect,especiallyforthestartupof
cavitation.Closelyrelatedistherequirementofacertainamountofcavitationnucleiinthewater,
seeHuse(1999).Thephysicsofincipientcavitationissuchthatifthewateriscompletelypure,the
cavity(cavitationbubble)mustnecessarilystartasaninfinitelysmallbubble,withinfinitelysmall
radiusofcurvatureandthusinfinitelylargeinsidepressureduetothesurfacetension.Butwithan
insidepressurelargerthanthesaturatedvapourpressurethecavitywillcollapse.Thus,infinitely
smallcavitiescannotexist,andsincelargercavitieshavetostartasinfinitelysmallones,they
shouldalsobephysicallyimpossible.Innaturalseawatertherewillalwaysbeacertainamountof
cavitationnucleisandthecavitationprocesscandevelop.Itisthereforeveryimportantthatthe
gascontentinthewaterinthecavitationtunnelissufficientlyhightoprovideanadequatelevelof
cavitationnuclei.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page67

6.3.3 TestProcedure
Thestandardprocedurefortestingincavitationtunnelscanbesummarisedasfollows,seealso
Huse(1999):
1. Chooseflowvelocityintestsectionbasedonactualadvanceratio,J.
2. Installaftbodymodelandadjustwakefieldbymeshscreens
3. Installpropellermodel
4. Withatmosphericpressureinthetunnel,adjustpropellerrateofrevolution(and/or
flowvelocity)untilthepropellerthrustiscorrectaccordingtoopenwaterand
propulsiontestsinthetowingtank.Thisiscalledthethrustidentityprinciple.
5. Keepingflowvelocityandrateofrevolutionfrombullet4constant,reducethetunnel
pressureuntilthespecifiedcavitationnumberisachieved.
6. Donecessarycavitationobservationandmeasurements.
Duringthetestthefollowingparametersaremeasured:

Propellerrateofrevolution

Thrust

Torque

Statictunnelpressure

Waterspeedinthetestsection

Inadditiontotheabovemeasurements,visualobservationofthepropellercavitationisan
importantpartofthetests.Thevisualobservationsaredonebystroboscopicilluminationat
differentangularpositionsoftheblades.Photographyandhighspeedfilmorvideorecordingis
oftenusedfordocumentation.Basedonexperienceonecanalreadyfromthevisualobservations
drawconclusionswhethertheactualcavitationisacceptableornot,and,ifrequired,which
modificationsofthepropellergeometryshouldberecommendedtoimprovetheperformancewith
respecttocavitation.Forinstance,experienceisthaterosioniseasilyproducedbybubble
cavitation,particularlyifthebubblescollapseonthebladesurface,whilelimitedamountsofsheet
andvortexcavitationareconsideredharmless.

6.4

Maneuveringtests

6.4.1 General
Thepurposeofmanoeuvringmodeltestsistwosided:
1. Directtestofthemanoeuvringbehaviouroftheshipwithactiveuseofrudderduringthe
test
2. Establishthehydrodynamicquantitiesrequiredintheinthemanoeuvringequationsfrom
thetests
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page68
Foritem1,alargespacewillberequiredandthesetypesoftestsarecommonlycarriedout
inseakeepingbasins.Afreerunning,selfpropelledmodelwithsteeringgearandpossibilityof
automaticruddercontrolisused.Maneuveringtestsareusuallycarriedoutaccordingtospecified
standardmaneuvers.Duringthetestsaprescribeduseofrudderisappliedasinput.Thehorizontal
motionsofthemodel(i.e.surge,swayandyaw)andtheactualrudderanglearemeasuredduring
thetests.Thetestsresultsareusedfordeterminationofthemaneuveringperformanceofthe
model.Themostusedstandardmaneuversarethezigzagtest,thespiraltestsandturningcircle
test.
TherequirementtoscalingformaneuveringtestsistouseFroudescalinginadditiontothe
geometricalsimilarityrequirements.Fordynamictestsalsotheinertiadistributionmustbe
correctlyscaled,seechap5.2.Possiblescaleeffectswillbeintroducedfromviscousforces.
Themostcommonandreliablemethodtodeterminethecoefficientsinthemanoeuvring
equations,seeBerg(2000),isstillmodeltestresults.Thecoefficientsfrommodeltestcanbe
obtainedeitherfromanalysisoftestsresultswithfreerunningmodelsorfromtestwithfixed
modelwheretheforcesandmomentsactingonthemodelismeasuredduringthetests.Fortesting
withfixedmodelthreedifferentapproachesisused:

PlanarmotionMechanism(PMM)tests

Towingtanktests

Rotatingarmtests

Thesemethodsarediscussedinthefollowing.

6.4.2 TowingTanktests
Inatowingtankthemodelcanbetowedatconstantspeedandyawangle.Themodelmustbefree
topitch,rollandheave,butrestrainedinswayandyaw.Theswayforcesandyawmomentare
measuredasafunctionoftowingspeedandyawangle.Boththelinearandquadratictermsare
obtainediftestsarecarriedoutforseveraltowingspeedsandyawangles.Theyawanglesare
usuallylimitedtobewithin10degreesasthisisconsideredasarealisticmaximumdriftangle
duringmanoeuvringoperations.
Similartestcanalsobecarriedoutfordifferentrudderangles.Therudderforcesaremeasured.
Fromthesetestsalsotheinfluenceoftherudderonyawmomentandswayforcescanbe
established.

6.4.3 PlanarMotionMechanism(PMM)
Thisisthemostusedmethodfordeterminationofthehydrodynamiccoefficientsinthe
maneuveringequations.PMMismountedtothetowingcarriageinatowingtankoraseakeeping
basin.AtypicalexampleofPMMsetupisshowninFigure6.5.Thetankhastobesufficientlywide
toavoidinterferencebetweenthemodelandthetankwall.Forshallowwaterteststhewater
depthshouldbecorrectlyscaledasthebottomeffectmayinfluencethemeasuredforces.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page69

Figure 6.5

Planar Motion Mechanism

ThemodelisusuallyconnectedtothePMMsuchthatitisfreetoheaveandpitchandfixedinroll.
Greatcareisrequiredwhenaligningthemodelandthisshouldbecheckedbeforeandafterthe
test.
Thefollowingsequenceoftestsshouldbecarriedout:
1. Obliquetowingtests,
2. Pureswaytests
3. Pureyawtests
4. Yawingwithdrift.
Duringtheteststhehullforcesinsurgeandswaydirectionandtheyawmomentarerecorded.If
rudderisappliedtherudderliftforcesshouldalsobemeasured.Theoscillationfrequencyshould
beselectedsufficientlowtobefreefromfrequencydependence.Furtherinterferencewithtank
wallmustbeavoided.Thedriftangle,driftanglespeedandswayspeedshouldbevariedtocover
therealisticrangeduringmanoeuvringconditions.
AfurtherdescriptionofrecommendedtestsprocedureforPMMisgivenbyITTC(1996),
ManoeuvrabilityCommittee.
PMMtestsareregardedasthemostreliableandcontrolledmethodfordeterminationofthehull
forcesasinputtothemanoeuvringequations.ThemaindrawbackwithPMMtestingisthe
complexityandcostconnectedtorunningthesetests.Toestablishacompletesetofhullforces
dataalargenumberoftestcasesarerequired.

6.4.4 Rotatingarm
Inrotatingarmteststhemodelismovedinacircularpathbyamechanicaldrivenrotatingarm.The
principleisshowninFigure6.6.Themodelisrestrainedinroll.Itispossibletovarytheradiusof
rotationandalsothedriftanglecanbevaried.Withthissetupthemodelwillhaveacorrect
combinationofspeedandradiusforaturningcircletest.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page70
Duringtheteststheswayforceandyawmomentaremeasured.Withthistestmethodthe
swayandyawcoefficientstogetherwiththedifferentnonlinearcouplingtermscanbeestablished.
Inadditionsurgeandheelmomentduringturningcanbemeasured.

Figure 6.6

Rotating arm Planar Motion Mechanism

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page71

SEAKEEPINGTESTING

7.1

General
Conventionalshiptestingasdescribedintheprevioussectioncoverthecalmwatertestingfor
ships.Thisiswhatweoftencalltheclassicalmodeltesting.Inthischaptertestingofshipsinwaves
willbediscussed.
Forseakeeping,atermthatcanbeinterpretedasshipbehaviourinroughweather,thefollowing
typeoftestresultswillbeimportant:
1. Shipmotions(heave,rollandpitch)andaccelerations
2. Slammingloads(bottomslamming)
3. Greenwaterondeck
4. Addedresistanceandspeedreductioninwaves
5. Globalloadsinhullbeam
Thefirstitemisimportantforcomfortevaluation(crewandpassengers)andevaluationofcriteria
forvoluntaryspeedreduction.Slammingloadsareimportantforstructuralstrengthbothlocally
andglobal(whipping)andwillinfluencevoluntaryspeedreduction.Greenwaterisimportantfor
safetyandpossibleloadingondeckhousesandequipment.Addedresistancedeterminedthe
involuntaryspeedreductionandisanimportantparameterforfuelconsumption.Optimisationof
resistanceperformancebasedononlycalmwaterresistancecanendupwithadesignsolution
whichisnottheoptimum,astheshipwillseewaveconditionswhereaddedresistancegivesa
significantcontributiontothetotalresistance,inalargepartofthetotaloperationtime.Fora
generaldiscussionofseakeepingperformance,seeLloyds(1989).
Seakeepingtestscanbecarriedoutintowingtanks(equippedwithwavemakerorinseakeeping
basins/oceanlaboratories.Towingtanktestingislimitedtoheadandfollowingsea,whileinlarger
basinsarbitrarywaveheadingcanbecovered.

7.2

TestRequirement
Fortestinginwaves,gravityforceswillbethegoverningforcecontributionandtheFroudescaling
hastobeapplied.Forscalingofwavesthisimpliesthatthewaveheightfollowsthegeometrical
scaleratioandthewaveperiodfollowsthesquarerootofthescaleratio,seealsochapter2.4.
Forseakeepingtestingthedynamicmotionsofthemodelarethekeyresultsandhenceinertia
forceswillbeimportant.Thisimpliesthatthemassdistributionofthemodelneedstobecorrectly
scaled,asdiscussedinchapter5.2.
Forseakeepingtestingthemaintopicofinterestisusuallythewavefrequencyresponse.If
slammingforcesisapartofthepurposeoftestingonehastoensureadequateequipmentand

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page72
sufficientsamplingfrequencytocovertheveryfastresponseofthistypeofloading,see
alsothediscussioninchapter4.10.
Themodelsusedforseakeepingtestingareusuallyconsideredtobeinfinitelystiff.Ifwhipping
and/orspringingresponsesareapartofthetestpurpose,alsotheglobalelasticpropertiesofthe
hullmodelneedstobecorrectlyscaled.Thisisdiscussedindetailinchapter5.3.Forthese
resonancefrequencyphenomenatheresponsecanbeconsideredtobehighfrequency(HF)and
wellabovetherangeofthewavefrequency.

7.3

Testsetup
Possibleprinciplearrangementsforconnectionsbetweenthemodelandcarriageareillustratedin
Figure7.1.Connectiontypea)isonlyusedformeasurementofwaveexcitationforces.For
alternativeb),c)andd)themodelisfreetoheaveandpitch,butwithdifferentlevelofconstraining
thesurgemotions.Intowingtanktesting(onlyheadandfollowingseas)alternativec)andd)are
mostcommonlyused.Alternatived)willrequireaselfpropelledmodel.Forthiscasethesurge
connectionhastobewithverylowfriction.Thiscanbeachievedbyusinghingedmechanism,
restrainedsideways,butfreetomoveverticallyandlongitudinally,connectedatbowandsternof
themodel.
ForSeakeepingbasintestingalternativee)isthemostcommonlyused.Inthissetuponlythe
poweringandinstrumentationcablesareconnectedbetweenthemodelandcarriage.The
connectionhastobesufficientlysofttoavoidinfluenceontheshipmotions.Thisalternativewill
requireaselfpropelledmodelwithheadingcontrol.Acarriagesystemthatisabletofollowthe
modeltrackwillalsoberequired.Fortestinginarbitrarywaveheadingacarriagewithasub
carriagefortransversemotionswillberequired.
Alternativef)isacompletelyfreerunningmodel.Allinstrumentationandpoweringsupplyhaveto
becarriedbythemodel.Themeasurementsmustberecordedonboardortransmittedby
telemetry.Thissetupdoesnotrequireanycarriagesystem.Howevertheequipmentonboardis
heavy,especiallybatteriesforpropulsionpower,andwillthereforerequireaquitelargemodel.
Alternativef)ismainlyusedfortestingmodelsoutsidelaboratories,onlakesandinthesea.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page73

Figure 7.1

Possible arrangement for mounting the model under the carriage


(From Lloyd (1989))

7.4

TestProcedure
Forseakeepingtesting,bothregularandirregularwavesarecommonlyused.Regularwavetesting
isnormallyusedforobtainingthemotionRAOsandaddedresistanceatthewavefrequencyandis
oftenusedforverificationofnumericalcalculations.Irregularwavesareusedtostudythe
performanceandestablishdesignvaluesatactualconditionsincludingstatisticaldistributionand
extremevalues.
Thewavecalibrationshouldbecarriedoutpriortothetestingandwithouttheshipmodelinthe
tanktoavoidwavereflectionfromthemodelonthewavemeasurements.Inalongandnarrow
towingtankthewavecharacteristicscanvaryalongthetanklength.Thisisinparticularaproblem
forshortandsteepwaves.Thisvariationisconnectedtothestabilityofthewavesasthey
propagatedownstreamthetankandalsototheeffectofthetankwalls.Duringthewave
calibrationthewavemeasurementsshouldthereforebecarriedoutindifferentpositionsalongthe
tanktodeterminepossiblepositiondependencyforthewaveheight(forregularwaves)and
spectralshape(forirregularwaves).

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page74
Ifthephaseinformationbetweenthewaveelevationandmotionresponseofthemodelis
required,itisnecessarytomeasuretheincidentwavebyawavegaugemountedatthecarriagein
aknownpositionrelativetothepositionofthemodel.Thedifferenceinpositionbetweenthewave
gaugeandmodel(measuredinthewavedirection)hastobeaccountedforwhencalculatingthe
relativephase.Itisalsoimportanttolocatethewavegaugeinapositionwherethedistortionof
thewavecausedbythemodelisavoided.
Bothselfpropelledandtowedmodelareusedinseakeepingtests.Intowingtestsaprescribed
towingspeediskeptconstantduringtherun.Forselfpropelledmodelstwodifferentprocedures
canbeapplied:

Constantpropulsionpowerappliedduringtherun.Thecarriagespeedadjustedto
followthemodel

Constantcarriagespeed.Propulsionpoweradjustedduringtheruntofollowthe
carriage.

Themeasurementsduringthetestwilldependontheactualpurposeofthetests.Atypicalsetof
measurementsduringaseakeepingtestwillbe:

Towingresistance(whentowingthemodel)

Thrust,torqueandrateofrevolution(forselfpropelledmodels)

Speed

Shipmotions;alwaysheaveandpitch,butalsosurge,sway,rollandyawmaybe
measured,dependingonsetup.

Verticalacceleration;typicallyatFP

Relativemotions;typicallyinbowandsternareas

Greenwaterondeck;measuredbywavegaugesondeckorbyforcetransducersto
measureamountofgreenwater

Impactloads;bottomslamming,bowflareorotherexposedlocations.Measuredby
pressurecellsorforcepanels.

Videorecording

Foramodelrunningatawaveheading,,( 180 isheadsea)theencounterfrequency, e ,is


givenfrom:

2U
g

cos

where isthewavefrequencyandUtheshipspeed.Whentestinginfollowingwavesthe
encounterfrequencymaybeverylow.Thiscanintroduceapracticalproblemasthenumberof
wavesencounteredduringonerunalongthetankwillbelow,andanumberarepetitionrunswith
differentrealizationsoftheincomingwaveswillberequiredtogetaproperconfidenceofthe
statisticaldistributionoftheresponses.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

7.5

Page75

Tankwalleffects
Theshipmodelwillgenerateitsownwavesystem,bothduetodiffractedwavesandradiated
wavesduetotheshipmotions.Thesewaveswillreachthetankwallandbereflectedbacktothe
modelasshowninFigure7.2.Ifthemodelspeedislowthereflectedwaveswillhitthemodelas
shownontheupperdrawinginFigure7.2,andtheresponseofthemodelwillbeinfluenced.

Figure 7.2

Illustration of tank wall effects. (From Lloyd (1989))

Thephasevelocityoftheradiatedwavesis:

ce

2e

Thetravellingdistanceisfromtankcentre(assumingthemodelinthisposition)tothetankwall
andbacktothemodel,i.e.equalthebreadthofthetowingtank,B.Thetravellingtimeisthen

tw B

ce

.Interferencewilloccurifthemodelmovesadistancelessthanitsownlength,LM

duringthetimetw.i.e.if:

U Ucrit

LM

tw

Combinedwiththeequationforencounterfrequency, e ,thefollowingrelationisobtainedforthe
criticalmodelspeedinheadseawaves:

U crit

2 LM
g
1
1
2
B

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page76
Here,isthewavecircularfrequency.Itfollowsthatthecriticalspeedincreaseswiththe
wavelength.TheresultingcriticalFroudenumberisshowninFigure7.3forheadseaandfollowing
waves.Forfollowingwavestankwallinterferencewillalwaysoccurformodellengthsgreaterthan
B/4,whichisusuallythecase.
Thetankwalleffectasanerrorsourcewillbefurtherdiscussedinchapter10.

Figure 7.3

Critical Froude number for tank wall effects. (From Lloyds (1989))

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page77

OFFSHORESTRUCTURETESTING

8.1

General
Inthischaptertestingofoffshorestructureswillbediscussed.Typicalshiptestingasdescribedin
thetwoprevioussectionscovershipsdesignedandusedforconventionaltransportationpurposes.
Thetermoffshorestructuresisinthispresentationinterpretedasallotherapplications.
Exampleswillbemoored,floatingplatformsandshipsappliedforproductionand/orstorageof
oil/gas,fixedstructures,risers,mooringsystems,floatingandsubmergedbridges,fishfarming
structures,etc.
Testingofoffshorestructurescoverawiderangeofmodeltestingfromsimpledecaytestincalm
waterintendedtodeterminehydrodynamiccoefficientstosetupswherecomplexplatform
structuresincludingmooringandrisersystemexposedtowaves,windandcurrentaremodelled.
Typicalobjectivesfortestingofoffshorestructuresincludes:
1. Determinehydrodynamicdamping,addedmass,motionRAOsandother
hydrodynamicquantitiesusedasinputtonumericalsimulations.
2. Verifyconceptstoensurethatallimportantphysicalphenomenaareproperly
understoodandaccountedforduringthedesignphase.
3. Establishoperationallimits.
4. Generatedataforcalibrationandverificationofnumericaltoolsforactualtypeof
structureandloadingconditions.
Fornewtypeofconceptsitem2willoftenbethemostimportant.Manyexamplesexistwhere
modeltestinghasrevealedunexpectedproblemswithnewconcepts.Insomecasestheproblems
areofsuchnaturethattheymaketheconceptunfeasible.Inothercasestheproblemsmaybeof
suchanaturethattheycaneasilybesolved,onceyouareawareofthem.
InAppendixBanexampleofaspecificationfortestingofanoffshorestructureispresented.The
actualcaseisaturretmooredfloatingproductionandstoragetanker(FPSO).

8.2

TestRequirements
Testingofoffshorestructuresisformostcasescarriedoutforfixedormooredstructures,which
meanszerospeed.Problemswithwallreflectioneffects(seechap.7.5)willthereforemaketheuse
ofrelativelynarrowtowingtanksveryquestionableforthistypeoftesting.Alsotherequirements
towindandcurrentgenerationareingeneralnotavailableintowingtanks.Oceanlaboratorieswill
thereforeberequiredinmostcasesforoffshoretesting.
Therequirementstoscalingofmodelsandwavesaretoalargeextentsimilartoforseakeeping
testing(seechapter7.2).Requirementstomodellingofmooringsystemsaregiveninsection5.3.3.
Themotionresponsesthatcanbeobservedforoffshorestructuresaresplitintodifferent
frequencyregimesdependentonexcitationsourceanddynamicpropertiesofthestructures:

Wavefrequency(WF)motions,i.e.motionstakingplaceinthefrequencyrangeofthe
usedwavespectrum.Theexcitationismainlylinearwaveforces.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page78

Lowfrequency(LF)motions,i.e.motiontakingplaceatresonantfrequencywell
belowthewavefrequencyrange.Theexcitationismainlynonlinearwaveforces(wave
driftforces),butalsothedynamicwindandforsomespecialcasescurrentinduced
oscillationcanbeimportantexcitationsources.Thislowfrequencyresponseisalso
oftencalledslowdriftmotion.

Highfrequency(HF)motions,i.e.motionstakingplaceatresonancefrequencieswell
abovetherangeofthewavefrequency.Theexcitationismainlynonlinearwaveforces
includingimpactloads.

Thelowfrequencymotionswillbeimportantforalltypeofmooredstructures.Dependingonthe
waterdepth,typicalresonanceperiodscanbeintherange40500s.Thelowfrequencymotions
areoftenfoundtobethemainsourcebothforplatformmotionsandmooringforces.
ThehighfrequencymotionshavebeenaproblemforTensionLegPlatforms(TLP).Theaxial
stiffnessoftheverticaltethers(usuallysteeltubes)connectingtheTLPtotheseabedintroducesa
resonancefrequencyinheave,rollandpitch.Typicalresonanceperiodswillbe2.54s,whichwill
bewellbelowthewaveperiodrangeforwaveexcitation.AnotherexampleofHFmotionsis
whippingandspringingofships.

8.3

Deepwaterstructuresrequirements
Thediscussionofscalinghassofarbeenlimitedtocaseswhereitispracticallypossibletoachieve
geometricalsimilarity.Anexamplewhereitisdifficultorimpossibletocorrectlymodeltheactual
geometryofthesystemisoffshorestructuresindeepwater.Deepwaterisusuallydefinedaswater
depthsgreaterthan1000m.
Traditionally,offshorestructuretestinghasbeencarriedoutatascaleofabout1:50.Ithasbeen
shownthatwhenusingstandardequipment(wavemakers,instrumentation,etc.)thescaleratio
canbeextendedtoabout1:100withoutseverereductionofqualityofmeasurement.The
maximumwaterdepthinexistingtestfacilitiesis10m,seechapter3.4.Thislimitsthepractical
waterdepthto5001000mforstandardtestinginexistingtestbasins.Basinswithmuchlarger
depthsarenotconsideredfeasibleofeconomicalreasons.Itshouldbenotedthatalsothe
length/breadthhavetobeincreasedaswelltobeabletomodelthehorizontalextentofthe
mooringsystem,seeFigure8.1
Differentsolutionsoftheproblemhavebeensuggested;seeStansberget.al.(2000)andBouchner
et.al.(1999):
1. Useofultrasmallscalemodeltesting(=1:>>100).
2. Passiveequivalentmooringsystemortruncatedhybridsystem.
3. Solutionswithactivecontrolsystems,basedonrealtimesimulationofthemooring
systemresponse
4. Outdoortesting
Forcase1thecompletesystemismodelledandnonumericalcalculationswillberequired.
Howeverincreasingscaleratiowillincreasethescaleeffects,especiallyforcaseswhereviscous
forcesareimportant.Physicalmodellingwillbedifficult.Notethatmassandloadscaleswiththird

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page79
powerofscalefactor( 1 3 ).Asenexamplescale1:200givesthat1ginmodelscale
correspondsto8tonnesinfullscale.Windandwavecurrentgenerationwillbecomeincreasingly
problematicandwaveswillbesignificantlyinfluencedbysurfacetension.

Exististing
basins

Mooring lines

Deep water

Figure 8.1

Required basin size for testing in deep water

Forcase2anequivalentandmoreshallowwatermooringsystem,withpropertiesassimilaras
possibleasthedeepwatersystem,isestablishednumerically.Theequivalentortruncatedsystem
isusedinthetest.AnexampleofmooringsystemtruncationisshowninFigure8.2.Themooring
stiffnesscanusuallybecorrectlymodelledbythismethod,butmooringlineandriserdynamicswill
notbecorrect.Thismethodwillthereforerequireextensivenumericalsimulationsafterthetests,
wherethedynamiccouplingoftheplatformandmooring/risersystemistreated.Thisprocessis
illustratedinFigure8.3.Usingthismethodconventionalmodelscaleratioscanbeusedtocover
applicationsinverydeepwater.Thedisadvantageisthatonehastorelyonnumericalcalculations
(calibratedtowardstheshallowwatertests)toestablishdesignvaluesfortheactualdeepwater
system.

Figure 8.2

Illustration of truncation of mooring system. Left; Full system, Right; truncated


system. ( from Stansberg et.al. (2000))

Case3isinprinciplesimilartocase2usingatruncatedsystem,butnowwithactivenumerical
controlatthemooringlineintersectionswiththebasinbottom.Theintentionwiththecontrolisto
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page80
simulateasimilarbehaviourasthepartofsystemnotmodelledwouldhavegenerated.
Complexonlinesimulationsystemswillberequiredforthisapproachandtheperformanceofthe
systemwillbedependentoncorrectlysimulatedcontrolactions.Alsothissolutionwilltherefore
havetorelayontheaccuracyofnumericalcalculations.
Case4withoutdoortestingmaybeafeasiblesolutionforsomecasesandhasbeenused,e.g.for
testingoflongrisersinacurrent.Howeverthefundamentalproblemwiththisapproachiscontrol
ofenvironmentalconditions.

Figure 8.3

Illustration of hybrid verification procedure (from Stansberg et.al. (2000))

8.4

TestProcedure
Foroffshorestructuretesting,irregularwavesaremostfrequentlyused.Inatypicaltestsetup
bothtestsinshortcrestedandlongcrestedwavesareoftenincluded.Forgenerationofdatafor
verificationandcalibrationofnumericalmethodslongcrestedwavesareusuallyused.
Asmentioned,testingofoffshorestructurescancoverawiderangeofcomplexityintestsetups,
fromsimpledecaytestsincalmwatertomultibodyarrangementsincludingmooringandriser
systemexposedtowaves,windandcurrent.Thetestprocedurewillvaryaccordingly.Fortestingof
mooredstructuresthemainstepsinthetestprocedurecanbesummarisedasfollow:

Environmentcalibrationtests,includingwaves,windandcurrent.Carriedoutwithout
themodelinstalled.Thecalibrationhastobecarriedoutatthecorrectwaterdepth
andfortheactualcombinationofwaveandcurrentconditions.Duringthecalibration
themeasurementsarecarriedatthemeanpositionofthestructure.

Installationofmooringandrisersystemandmodelinthebasin.Inbasinswithmovable
bottom,mostofthisworkwillbedonewithbottominsurfaceposition.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page81

Staticcalibrationtestsforverificationandcalibrationofsystemproperties,
includingheeltestofplatformmodelformetacentricheightverification,pullouttests
forverificationofmooringandrisertensionandgeometry,etc.

Decayteststoverifynaturalperiods,dampingandgeneraldynamicperformance,
typicallyincludingplatformheave,rollandpitchmotionsaswellassurgeandsway
motionsformooringsystem.

Testsinthedefinedwave,windandcurrentconditions.

Testsareoftencarriedoutfordifferentloadingconditionsoftheship/platformtocovertheactual
workingrange,typicallyfullyloadedandballastconditionwillbecovered.Asresponseofmoored
structurewillbesensitivetotheheadingoftheenvironment,severalcombinationsofwindwaves
andcurrentheadingsareusuallycoveredbythetestprogramme.
Thepurposeoftestsisoftentoverifythesystemperformanceintypicaloperatingconditionsand
toestablishthedesignvaluesoftheresponsesinextremeenvironmentalconditions.In
combinationwithdifferentenvironmentaldirectionsandloadingconditionsacomprehensivetest
programwithasignificantnumberoftestswillberequired.
Themeasurementsduringthetestwilldependontheactualpurposeofthetests.Atypicalsetof
measurementsfortestingofamooredplatform/shipwillbe:

Mooringlinetensioninallmooringlines

Risertensionandshearforces/bendingmomentsatplatformconnection.

6degreeoffreedomplatformmotions.

Accelerationsatspecifiedpoints

Relativemotionsandwaverunupatcolumns/bowarea

Forshipstructures;Greenwaterondeck,Impactloads;bottomslamming,bowflareor
otherexposedlocations.

ForPlatforms;Impactloadstowardsdeckstructuresandotherexposedareas.

Videorecording

Typicaldurationoftestsinirregularwaveswillbe36hours(fullscaletime).
Analysisofthemeasurementsisdiscussedinchapter10.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page82

REALTIMEHYBRIDMODELTESTING
Realtimehybridmodeltestingmeansmodeltestinginrealtimecombinationwithnumerical
simulation.Theideaisnotreallynew,butthepossibilitytorealizetheideaisjustrecentlyoccurred,
duetothedevelopmentofsimulationsandactuatortechnology.
Theconceptofrealtimehybridmodeltestingisbestexplainedbyanexample;Passiveantiroll
tanksareextensivelyusedonmanytypesofoffshoreships,sincetheyprovideeffectiveroll
dampingregardlessoftheforwardspeedoftheship.Thesloshingflowinsidethetankishardto
calculatebynumericalmethods,andtheviscouseffectsrequireamodelofthetankwhichisfairly
large(morethanameterinlength).Thereisofcourseaclosetwowaycouplingbetweenthe
sloshingwatermotionsintheantirolltankandtherollingmotionsoftheship.Therollmotionsof
theshipcanbesimulatedfairlyeasilyandaccuratelybyindustrystandardprogramslikeShipX
Vesim,Shipmo,orWases.Thus,onehavesetupahybridmodeltestingschemewheretheship
motionsaresimulatedusingtheshipsimulatorVesim(whichisusingmotioncalculationsfrom
ShipXVeres),themotionscalculatedbyVesimarefedtoarigholdingthemodeloftheantiroll
tank,whichthenshakesthetankaccordingtothesimulatedshipmotions,andthentheresulting
forcesfromthetankontherigaremeasuredandfedbackasinputtothesimulatorinthenext
timestep.ThisisillustratedinFigure9.1below.

Figure 9.1Principle sketch of hybrid testing of passive anti-roll tank

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page83
Therearemanytypesoftestswhichmaybenefithugelyfromuseofrealtimehybridmodel
testingtechniques.Herearesomeexamples:

9.1

Testingoffloatingoffshoreshipsandplatformswithmooringandflexiblerisersystems.
Currently,thewaterdepthneedstobemodelledinthesamescaleastheplatformfortheriser
andmooringsystemdynamicstobemodelledcorrectly.Thisplacesseverelimitsonpossiblemodel
scalesforverificationofdeepwaterinstallationsorrequirestheconstructionofextremelycostly
deepwatermodelbasins.Thecurrentsolutionistotesttruncatedsystems,butthisisnotreallya
satisfactorysolution.Inahybridmodeltestingschemeonewouldforinstancemodeltheplatform,
simulatethemooringandrisersystems,applythecalculatedforcesontheplatforminrealtime
andfeedtheplatformmotionsbacktothesimulations.

9.2

Testingofhydrofoilships.
ThehydrofoilsrequirehighReynoldsnumberandcorrectlymodelledcavitationforcorrect
performance,whiletheshiphullrequiresFroudescalingandafreewatersurface.Together,this
requiresalargetowingtankwithcontrolledairpressure(vacuumtank)withaveryhighspeed
carriage,somethingthatdoesntexist.Thehybridmodeltestingsolutionwouldbetotesttheship
hullinatowingtankwhilesimulatingtheforcesfromthefoilsystem,and/ortotestthefoilsystem
inahighspeedcavitationtunnelandsimulatetheshipperformance.

9.3

Testingoffloatingoffshorewindturbines.
Duetowindgustsandthemotionsofthefloatingturbinetower,theforcesfromtheturbineonthe
towerarehighlydynamic.Inaddition,theelectricgeneratorandthepitchcontrolsystemofthe
turbineareimportantfortheresponse.Allthismightbemodelledinalargeoceanbasinwitha
powerfulwindgeneratingsystem,butitrequiresalargemodelandacorrespondinglyhugebasin.
Thehybridmodeltestingsolutionisforinstancetotestthetowerinanormallysizedwavebasin,
simulatetherotorandgeneratordynamicsandapplythesimulatedforcesfromtheturbineonthe
turbinetowermodelinrealtime.

9.4

Challengesinhybridmodeltesting
Hybridmodeltestingasoutlinedabovemightseemlikeaverysimpleandattractivetechnique,but
therearegoodreasonswhythistechniqueisstillnotinroutineuseinmostmodelbasins.The
initialexamplewiththeantirolltankissimplerthantheotherthreeexamplesbecauseitismotions
andnotforcesthatareappliedonthephysicalmodel.Applyingspecifiedforcesaremore
challenging.Theelectrical(orhydraulical)actuatorsthataretypicallyusedwillprovideaspecified
position.Itisthenpossibletoplaceforcetransducersbetweenthemodelandtheactuator,and
thenhavealocalcontrollooptocontroltheactuatortogivethewantedforce,butitadds
complexitytothesystem,andtheinevitablesmalltimedelaysinsuchacontrolloopmight
introduceunwanteddynamicsandeveninstabilitytothesystem.Figure9.2showsanothersetup
forahybridtestofafloatingoffshorestructure,wheretheactuatorsareapplyingcalculated
displacementstotheendsoftruncatedmooringlines.Inthismanner,oneavoidstheproblemof
applyingknownforces,butontheotherhand,onegetstheadditionalproblemofsimulatingthe
dynamicresponseofthetruncatedmooringlines.
Intherealtimehybridtestingcontrolloopthestateofthemodelisfirstsampled,thentherequired
actuationiscomputedbasedonsimulationofthepartofthesystemnotincludedinthemodeltest.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page84
ThesamplingrequiresuseofalowpassfiltertoavoidNyquistdownfolding,andadditional
filteringmightberequiredtoremovenoisefromthesignal.Asdiscussed

Mooring lines

Basin

Mooring line servo units

Full water depth

Floater motions

Output motions
of mooring line end

A/D
Mooring line tension

Mooring line
servo unit
D/A

Control system

Control
signals

Power

T
h
e

Figure 9.2 Possible solution for hybrid model test of floating offshore structure.

Timedelayisalsoaproblemforthesimulationpart.Timescalesas 1/ ,sotimegoesquickerin
modelscale.Thus,thesimulationsmustbefairlysimpletobeabletoruninrealtimeinmodel
scale.Thismeansthatonemighthavetomakespecial,simplifiedsimulationmodelsforthe
purposeofhybridmodeltesting.Thisisforinstancethecaseforthefloatingmooredoffshore
structure,whereafullsimulationofthedynamicresponseofthemooringandrisersystems
certainlywontruninrealtimemodelscale.Thisproblemisbecomingsmallerascomputer
calculationspeedisincreasing,butitwillcontinuetobealimitationfortheforeseeablefuture.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page85

10

ANALYSISOFMEASUREDDATA

10.1

General
Themeasurementsfrommodeltestswillberecordingsoftimehistoriesfromallapplied
transducersinthetestsetup.Theanalysisofmodeltestdatawillthereforebeconcentratedon
timeseriesanalysis.Thetypeofanalysistobeperformedwilldependonwhichtypeofteststhat
havebeencarriedout.Typicaltypeoftestswillbe:
1. Statictests
2. Decaytests
3. Regularwavetests
4. Irregularwavetests
Analysisofeachofthedifferenttypeoftestsisdescribedinthefollowing.

10.2

Statictests
Withstatictestswemeanteststhatareexpectedtogiveaconstantmeasuredvalue.Typical
examplesofsuchtestswillbe:

Towingtestsandpropulsiontestsincalmwater

Pullouttests

Forsuchteststhemeanvalueistheonlytestresultofinterest.Howeversomecareisrequiredto
avoidincludingthetransienteffectsduringstartup.Itwillthereforeberequiredtoremovethis
partbeforethemeanvalueiscalculated.Plotofthemeasuredtimehistorywillgiveagoodcontrol
thatsteadystateconditionhavebeenreached.

10.3

Decaytest
Decaytestswillgiveimportantinformationaboutnaturalfrequencies,addedmassanddampingof
adynamicsystem.AnexampleofmeasuredresponsefromadecaytestisshowninFigure10.1.
Letusconsideraonedegreeoffreedomsystemwithnonlineardamping.Thedifferentialequation
describingthemotionsis:

Mx B1 x B2 x x Cx 0

whereMisthemass(includingaddedmass),Cistherestoringstiffness,B1isthelineardamping
andB2isthequadraticdampingterm.Thenaturalfrequencyoftheundampedsystemisgivenas:
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page86

Fromameasurednaturalperiodofthesystemandaknownrestoringstiffnessthetotalmasscan
becalculatedfromtheaboveequation.Theaddedmassisnowdeterminedbysubtractingthe
knownstructuralmassofthesystem.

Figure 10.1

Example of measured response from a decay test

Inthefollowingthemethodfordeterminationoflinearandnonlineardampingisoutlined.The
equationofmotionscanbedividedbyMandwegetthestandardform:
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page87

x p1 x p2 x x p3 x 0

Theanalysisofthisequationisbasedonwellknownsolutionofalinearoscillatingsystemin
combinationwiththetechniqueofequivalentlinearization.Equivalentlinearizationimpliesthatthe
nonlineardampingtermisreplacedwithalinearterm.Theequivalentlineartermisdetermined
fromtherequirementofequaldampingenergypercycle.Thisrequirementissatisfiedusing:

pEQ p1

8
x0 p2
3

wherex0isthemotionamplitudeoftherelevantcycleandistheoscillationfrequency.The
linearizedequationofmotioncannowbewrittenas:

x pEQ x p3 x 0

Assumingxi1andxiaretwosucceedingamplitudes,thelogarithmicdecrementisdefinedas:

ln(

xi
)
xi 1

Theamplitudesxiandxi+1shallbespacedonedampednaturalfrequency,Td,apart.Thedamping
ratio isdefinedastheratiobetweenactualandcriticaldamping:

p
p

pcr 2 M 0

Thegeneralrelationbetweenthelogarithmicdecrementandthedampingratioisgivenas:

0Td 2

1 2

Forlowdampingratios,typically < 0.2,thisrelationmightbeapproximatedby:

Theequivalentdampingcoefficient,pEQ,isgivenfrom:

pEQ 2 M 0

2C

pEQand cannowbeobtainedforeachcyclefromthemeasuredlogarithmicdecrement.The
resultsfor foreachperiodareplottedversusthemeanamplitude(meanoftwosuccessive
8 xi
.Anexampleisshownin
amplitudes)oralternativelyagainstthetheequivalentvelocity
3
Figure10.2.p1isfoundfromthefigurefromtheintersectionwiththeabscissaandp2isfoundfrom
theslopeofthecurve.Oneshouldavoidusingthefirstoscillation,duetotransienteffects,andthe
smallestamplitudesatthetailofthedecay,duetoinaccuracy.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page88

Figure 10.2

Example of measured damping from a decay test

Thenaturalfrequencyofthedamped,freelyoscillatingsystem,d,isgivenfrom:

d 0 1 2

where0isthenaturalfrequencyoftheundampedsystemdefinedaboveandistherelative
dampingofthesystem.

10.4

RegularWaveTest
RegularwavetestsarenormallyusedforobtainingtheRAOsofmotionsandloads.Theanalysisof
regularwavesrequiresthattheinputwaveissufficientlyclosetosinusoidalandthatthewaveis
stableandstationarylongenoughtoobtainsteadystatevesselmotions.Steepregularwavesmay
tendtobeunstableastheypropagatefromthewavemakeranddownthetank.Wavebreakingor
transferofenergytootherwavefrequenciesmayoccurandtheinputwavereachingthevesselis
nolongersinusoidal.Tankwallreflectionwillbeanotherproblemfortestinginrelativelynarrow

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page89
tanksandintroduceanonstationaryeffect.Wavereflectioneffectsasanerrorsourcewill
befurtherdiscussedinchapter12.
Itisimportanttoensurethatasteadystateresponsehasbeenachievedforthetimewindowused
fortheanalysis.Forwaveperiodsclosetoanyresonanceperiodofalightlydampedsystemseveral
oscillationswillberequiredtoachieveasteadystateresponse.Carefulinspectionofthemeasured
responsetimehistorywillthereforeberequiredbeforeperformingtheanalysis.
Assumingaharmonicinputwavesignal:

x(t ) X A sin(1t )

where X A isthewaveamplitudeand 1 isthewavefrequency.Foralinearsystemthemeasured


responsesignal,y(t),willalsobeaharmonicsignal.Theamplitude YA ,canbeeasilyobtainedfrom
themeasuredmaximaandminimaofthe y (t ) signal.Thephaseangle ,isobtainedfromthe
phaselagbetweentheinputwavesignalandthemeasuredresponse.TheRAO(Response
AmplitudeOperator)ortransferfunctionisdefinedastheratiobetweentheresponseamplitude
andinputwaveamplitude:

RAO

YA

XA

Asdiscussedabovetheinputwavemaynotbecompletelyharmonic.Also,thetestedsystemmay
introducenonlinearityseitherduetononlinearwaveloadingeffectsormechanicalnonlinearities
(e.g.mooringforcesorstructuralnonlinearitysinthemodel).InpracticalmodeltestingFourier
analysisofthemeasuredresultsisalmostalwaysused.FromtheFourieranalysistheamplitudeand
phaseofthedifferentharmoniccomponents, Yi and i ,areobtained.Thefundamental
componentisatthefrequencyofwaveexcitation 1 andthehigherordercomponentsareat
frequencies n n1 , wheren=2,3.TheRAOsarenowdefinedastheratiobetweenthe
fundamentalcomponentofthemeasuredresponseandthewave:

RAO

Y1

X1

InFigure10.3atypicalexampleofplotsofmeasuredresponsesinregularwavesareshown.The
caseconsideredisverticalwaveforcesactingonafixedhorizontalcircularcylinderwithdiameter
D=0.5m.Thefirstcaseshowstheresponseforasmallwaveheight,H=0.04m.Thesystemisclose
tolinear.Thesecondcaseisforthesamewaveperiod,butwithwaveheightincreasedtoH=0.4m.
Strongnonlinearitiesareobserved.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page90

Figure 10.3

Examples of measured responses in regular waves. Vertical wave forces on


horizontal circular cylinder. a: Input wave for H=0.04 m, b: Vertical forces for
H=0.04 m, c: Vertical force for H=0.4 m

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page91
TheresultsfromtheFourieranalysisarealsoincludedintheFigure10.3.Forthefirstcase
almostallenergyisconcentratedonthefundamentalfrequency(e.gtheinputwavefrequency).
Forthelastcasesignificantcontributionsatthehigherharmoniccomponentsareobserved.The
inputwavesignalisseentobealmostcompletelysinusoidal.
Testinginregularwaverequireonetestforeachfrequency.Itwillthereforeberelativelytime
consumingtoobtainacompletetransferfunctioncoveringanactualfrequencyrange.Typically10
15testsatdifferentfrequencieswillberequired.

10.5

IrregularWaveTest
Irregularwavesareintendedtorepresentarealisticseastateandarethereforeusedforstudying
theactualresponsesincludingnonlinearphenomenaashighfrequency(HF)andlowfrequency
(LF)responses,impactloadsandsurvivabilityinextremeseastates.Toachieverealisticwave
frequencyresponsetheenergyspectrum(i.etheenergyfrequencydistribution)oftheinputwaves
mustbecorrect.Forthelowfrequencymotionsitisalsorequiredthatthewavegroupspectrumis
correctlymodelled.Thewavegroupspectrumisthespectrumoftheenvelopecurveofthewave
train,seeFigure10.4.

Figure 10.4

Envelope curve of the wave train.

10.5.1 Generalaboutstochasticprocesses
Themeasuredtimehistoryfromirregularwavetestswillbeastochasticprocessthatisa
continuousfunctionoftime.Itisthereforenecessarytodiscusssomecharacteristicsthatclassifya
randomprocess.
Stationarity:
Testinginirregularwavesisalmostexclusivelybasedonthattheinputwavesignalisastationary
process,whichmeansthatallthestatisticalproperties,includingprobabilitydistributionand
densityfunctionoftheprocessareunchangedbytime.Thismeansforexamplethatfortheprocess
x(t):

E x(t ) E x(t )

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page92
whereE[x]istheexpectedvalueofx.
Homogeneousprocess
Itiscommonlyassumedinanalysisofwavesthattheprocessisnotonlystationaryintime,butalso
homogeneousorstationaryinspace.Thisassumptionimplythatnochangewouldbemadeofthe
statisticalpropertiesifthepositionofthemeasurementswhereshiftedtoanewlocation.
Ergodic
Astationaryprocessiscalledergodicwhenitisallowedtoreplacetheaveragingoverspacebyan
averagingovertime.Thisimpliesforexamplethatforstationaryandergodicprocessx(t)wecan
write:

E x1 (t ) E x2 (t ) E x(t )

whereE[x1]istheexpectedvalueofxatposition1.Thisassumptionmakesitpossibleto
interchangetheexpectedvalueoffunctiong(x)withthetemporalaverageofg(x):

Lim 1 T / 2
E g ( x)
g ( x)dt
T T T / 2

Aconsequenceoftheergodicpropertyisthatthewaveregistrationatonesinglepointcanbeused
forcharacterisationoftheseastate.
Inthefollowinganalysiswewillassumethatthemeasuredresponseisastationaryandergodic
stochasticprocess.Applyingtheseassumptions,largelysimplifythestatisticaltreatmentofthe
measuredstochasticprocesses.Detaileddescriptioncanbefoundintextbookstreating
stochasticallyanalysis,e.g.PriceandBishop(1974),BendatandPiersol(1966).
Theautocorrelationandcrosscorrelationfunctions
Theautocorrelationfunction R xx ( ) correlatesthevalueofthestochasticprocessxattimettoits
valueatalaterstage(t+).Inthisway R xx ( ) givesameasureofthecorrelationofthesignals
dependentonthetimelag.Theautocorrelationfunctiongivesameasureoftheperiodicityofthe
timeseries.AtimeserieswithwhichisperiodicwithperiodTwillhaveapeakinthe
autocorrelationfunctionat Rxx ( T ) .Theautocorrelationfunctionisdefinedas:

Rxx ( )

Limit 1 T
x(t ) x(t )dt
T T 0

Appliedontwodifferentsignalsx(t)andy(t)wherex(t)isthereferencesignal(waveinput)and
y(t)isthemeasuredresponse,thecrosscorrelationfunctionisgivenfrom:

Rxy ( )

Limit 1 T
x(t ) y (t )dt
T T 0

Thisisausefulexpressionforachievingarelationbetweenthewaveinputandmeasuredresponse
inamodeltestsetup.
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page93

10.5.2 SpectralAnalysis
Inanalysingmodeltestresultsitwilloftenbeveryconvenienttotransformthemeasuredtime
historysignalx(t),tothefrequencydomain.Thiscanbeobtainedbyaspectralanalysisinwhich
thespectraldensityofthesignalisdeveloped.ThespectralanalysisisusuallydonebytheFast
FourierTransform(FFT)technique.
Assumingtheprocesstakesplaceoverthetimeinterval (0 t T , T ) .Thepowerspectrum
ofx(t), S xx ( ) isdefinedastheFouriertransformoftheautocorrelationfunction R xx ( ) :

S xx ( )

xx

( )e ( i ) d

Thecrossspectrumbetweentwosignalsx(t)andy(t)wherex(t)isthereferencesignal(wave
input)andy(t)isthemeasuredresponseisgivenfrom:

S xy ( )

xy

( )e( i ) d

Thelineartransferfunctionbetweenresponseandwavecanbeestablishedfromthecross
spectrumbetweentheresponseandtheinputwavesignalandthewavespectrumasfollows:

H ( )

S xy ( )
S xx ( )

H ( ) isthecomplextransferfunction.TheResponseAmplitudeOperator(RAO)canbedefinedas
themodulusof H ( ) andthephase,(),isthephaseangleof H ( ) .
AlternativelytheRAOcanbeobtainedfromtheinputwavespectrumandmeasuredresponse
spectrumdirectlyas:

H ( )
2

S yy ( )
S xx ( )

Usingthisformulationnophaseinformationisobtained.Theprinciplesofthisrelationare
illustratedinFigure10.5.
InFigure10.6anexampleofresultsfromaspectralanalysisofmeasuredpitchmotionsareshown.
Themeasuredwavespectrum,Pitchresponsespectrum,RAOandrelativephasebetweenthewave
elevationandPitchmotionsareshown.TheRAOandrelativephaseareobtainedfromthecomplex
transferfunction H ( ) .
Thereferencewaveusedinthecalculationofthetransferfunctionshouldbetakenasthewave
measuredatthereferencepositionwithoutthemodelpresent.Thisisinordertoavoidtheeffect
ofthewavesystemgeneratedbythemodel.Fortestwithastationarymodel(e.g.fixedstructures
ormooredvessels)thisisstraightforwardtoobtain,butfortowedorfreerunningmodelsthe
referencepositionisnotwelldefined.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page94

Figure 10.5
RAO.

Example of relation between input wave spectrum, measured response spectrum and

Thenthmomentsofthespectrumisdefinedas:

mn n S ( )d

Thedifferentmomentscanbeinterpretedasfollows:
n=0:

m0=variance(2)oftheprocessormeansquareresponseorareaofspectrum

n=2:

m2=secondmomentofthespectrumormeansquarevelocityofresponse

n=4:

m4=meansquareaccelerationofresponse

Thefollowingparameterscanbecalculatedfromthespectralmoments:

m0

Standarddeviationofresponse:
Significantvalueofresponse: x1

4 m0

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page95
Averageperiodofresponse:

T1

Averagezerocrossingperiod: T2

m0

m2

m0

m1

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page96

Figure10.6

Exampleofresultfromspectralanalysis(fromMARINTEKReport).

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page97
AtypicalexampleofresultsfromspectralanalysisforamodeltestsetupisshowninTable
10.1.Foreachofthemeasuringchannelsalltheabovediscussedparametersarelisted.

Table10.1

Exampleoftabulatedresultsfromspectralanalysis(fromMARINTEKReport).

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page98

10.5.3 Statisticalanalysis
Thestatisticalanalysisofthemeasuredsignalsconsistsincalculatingthestatisticalparametersand
thestatisticaldistributionoftheprocess.Theprobabilitydistributionfunctionorthecommutative
probability,P(x),istheprobabilitythatageneralvalueoftheprocessx(t)islessthanorequalthe
valueofxbeingconsidered.Thisiswrittenas:

P ( x ) P ( x (t ) x)

Theprobabilitydensityfunctionisdefinedfromtherelation:

p ( x)

dP ( x )

dx

orinotherwords:Theprobabilitythata<x(t)<bisgivenbytheprobabilitydensityfunctionsuch
that:
b

P (a x (t ) b) p ( x )dx

Twodifferentprobabilitydistributionsareofinterestinthestudyofrandomwavegenerated
responses:

Thedistributionoftheprocessitself,e.g.thedistributionofthewaveelevationx(t)and
themeasuredresponsey(t).

Thedistributionofamplitudes;e.g.distributionofthewaveamplitudesxAand
measuredresponseamplitudes,yAinthetests.

ThewaveelevationforrandomwavesiscommonlyadaptedasaGaussian(ornormaldistributed)
process.ThewaveamplitudescanthenbeassumedtofollowtheRayleighdistribution.The
amplitudedistributionisthemostinterestingquantitiesinthemeasurementofwaveinduced
motions.Thisdistributionwillthereforebediscussedinsomemoredetails.
ThecumulativedistributionfortheRayleighdistributionisgivenas:

1 x 2
X
P ( x) 1 exp
for x 0
2 X

Here X isthemeanorexpectedvalueofx(t),definedas:

X E x xp( x)dx

X2 isthevarianceofx(t),definedas:

2
X2 E x X E x 2 X 2

ForameasuredtimeserieswithNsamplesthemeanvalueandthevariancearecalculatedas
follows:

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page99

X2

1
N

x
i 1

1 N
( xi X )
N 1 i 1

CurvesshowingthecumulativedistributionandtheprobabilitydensitydistributionfortheRayleigh
distributionareshowninFigure10.7.

Figure 10.7

The Rayleigh distribution (a) cumulative distribution, (b) probability density.

AsmentionedabovethewaveamplitudescanbeassumedtofollowtheRayleighdistribution.For
waveinducedlinearresponsesthesameassumptioncanbemade.Howeverforabroadrangeof
practicalimportantproblemscoveredbymodeltestingofshipsandoffshorestructuresthe
responseisnotlinear.Nonlinearitiescanbeintroducedbeseveralreasons.Examplesare:

Surfacezonenonlinearitycausedbynonverticalstructuresidesorstructureinandout
ofwater

Excitationfromhigherorderwaveforces(driftforces)

Nonlineardragforces

Nonlineardamping(importantifresonanceoscillationoccur)

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page100

ImpactLoads

Nonlinearrestoringstiffness(importantformooredstructure)

Forallthesesituationsthedistributionoftheresponseamplitudeswill/maynotfollowtheRayleigh
distribution.ThemoregeneralWeibulldistributionisthereforecommonlyusedforfittingthe
measureddatatoaknowndistribution.Toachievethisfitthestatisticaldistributionderivedfrom
themeasurements,P(xA),isplottedinWeibullscale.ThisisachievedbyusinglogarithmicaxisforxA

andtheP(xA)axisplottedas ln ln(1 P( xA )) .UsingtheWeibullscaleplottingwillalso

emphasisethetailofthedistribution,whichwillgovernthepredictionoftheextremeamplitude
values.FromplottingthecumulativedistributionsinWeibullscaleitispossibletoanalysewhether
largeextremevaluesmeasuredinthetestsaresimplyresultsofstatisticaluncertainties,orresults
fromamoresystematictrend.
TheWeibulldistributionisgivenas:

1 x A X G
P ( x A ) 1 exp


G

whereasbefore X isthemeanvalueoftheprocessx(t),isthestandarddeviationofthe
processandGistheshapeparameterdescribingtheslopeoftheWeibullcurve.ForG=2givesthe
RayleighdistributionandG=1givestheExponentialdistribution.
Anexampleofplotsofmeasuredcumulativeprobabilitydistributionofresponseamplitudesis
showninFigure10.8.ThestraightlineshowninthediagramsrepresentG=2,i.etheRayleigh
distribution.ItisseenthatforthefirstcasethedistributionfollowscloselytotheExponential
distribution,whileforthesecondcasethedistributionisclosetoRayleighdistributed.

Figure10.8

ExamplesofcumulativeprobabilitydistributionsplottedinWeibullscale.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page101
Inadditiontothecumulativeprobabilitydistributiondiscussedaboveanumberof
statisticalparametersareobtainedfromthestatisticalanalysis.Theyaresummarisedinthe
following.
SignificantValues:

Significantmaxima; xA1/ 3 :

themeanofthehighestonethirdofthecrestto
zerovaluesofxA,

Significantminima; xA1/ 3 :

themeanofthehighestonethirdofthetrough
tozerovaluesofxA,

Significantdoubleamplitude;2 x A1/ 3 :

meanofthehighestonethirdofthemaximum
tominimumvaluesofxA,

Maximum/MinimumValues:

MaximumValue; x A, MAX :

Measuredmaximumvalueintherecord.

MinimumValue; x A, MIN :

Measuredminimumvalueintherecord.

Largestdoubleamplitude; 2 xMAX : Largestmeasuredcresttotroughvalueinthe

record.
SkewnessandtheKurtosis.
TheSkewnessisdefinedasthethirdcentralmomentoftheprocess:

1 E ( x X )3

m3

X3

Theskewnessprovideanindicationofthedegreeofasymmetryintheprobabilitydistribution
aboutthemeanvalue.ForaGaussiandistributiontheskewnesswillthereforebezero,butnonzero
foraRayleighdistribution.
TheKurtosisisdefinedas:

2 E ( x X )4 3

m4

X4

TheKurtosisiszeroforaGaussianprocess.
AtypicalexampleofresultsfromstatisticalanalysisforamodeltestsetupisshowninTable10.2.
Foreachofthemeasuringchannelsalltheabovediscussedparametersarelisted.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page102

Table10.2

Exampleofresultsfromstatisticalanalysis(fromMARINTEKReport).

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

11

Page103

FULLSCALEMEASUREMENTS
Fullscalemeasurementsareintheselecturenoteshandledastheexception,ratherthantherule.
Thismightatfirstseemabitawkward,sincefullscaleisclosertothetruesituation.However,
testinginfullscaleisusuallytimeconsuming,verycostly,andeveniftheuncertaintyofscalingis
avoidedtheuncertaintyintroducedbyanuncontrollableenvironmentmightmaketheresultsless
reliable.
Fullscaletestsareusuallycarriedoutinthreedifferenttypesofcircumstances:

Deliveryofnewbuildings(Seatrials)

Ifaspecialproblemhasarisen,likepropellernoiseorexcessivefuelconsumption

ForResearchpurposes

Wewillnowdiscusseachofthesethreecases,beforesomegeneralconsiderationsaboutthe
specialchallengesoffullscaletrialsaremade.
Inadditiontothiscomesshipmonitoringsystems.Bythis,wemeansystemsthataremonitoring
aspectsoftheperformanceoftheshipduringnormaloperation.Shipmonitoringiscoveredina
separatesectionattheendofthischapter.

11.1

Deliverytrials
Inalmostallshipbuildingcontractstherearespecificrequirementsforthespeedtheshipshall
obtainatcertainenginepowerlevels.Therearealsoveryspecificandgraveconsequencesfor
deviationsfromthecontractedspeed.Iftheshipspeedissmallerthanthecontractspeedby,
typically,0.2knots,theshipyardwillhavetopayafeetotheshipowner.Ifthedeviationislarge,
forinstance1knot,orevenassmallas0.5knots,theshipownerhastherighttorefusetotake
ownershipofthevessel.Ontheotherhand,iftheshipisfasterthancontracted,theyardusually
getsabonus.Onemightthenthinkthatthistemptstheyardtoquotealowspeed,orspecifya
largeenginetoreachthespecifiedspeed,butdoingthiswillmostlikelyresultintheyardlosingin
thecompetitionforgettingthecontractinthefirstplace.Withthisbackgrounditiseasyto
understandwhydeliverytrialsareimportant.Itisalsoeasytounderstandwhycarefulmodel
testingtodeterminethetrialspeedbeforetheshipisbuiltissocommon.Ifthemodeltestshows
thattheshipisnotgoingtomeetthespeedrequirementitisstilltimetodofinetuningoflinesor
propellerarrangementtoimproveperformance.Aftertheshipisbuiltitistoolate.
Deliverytrialswillalwaysincludespeedtrials.Usuallydoubleruns(runningthesamestraighttrack
inbothdirections)atnotlessthanthreedifferentenginepowerlevels.Othertypesoftests
commonlyperformedaremaneuveringtrials,toensurethattheshipfulfillstheIMOrequirements
formaneuverability.Forhighspeedcrafts,thenewIMOHighspeedcraftcodealsoincludes
requirementsforseakeepingtests(seeIMO:2000HSCCode(IMO185E),Annex9).Inaddition,
therewillusuallybealotoftestsrelatedtothecorrectperformanceofonboardsystems,butthese
testsareoflittleinteresttohydrodynamicists,andwillnotbediscussedhere.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page104

11.1.1 Organizationofdeliverytrials
Whenparticipatinginaseatrialitisimportanttobeawareoftheorganizationandwhois
responsible,sowewillgiveabriefoverview:TheShipbuilderisresponsiblefororganizingand
executionoftheseatrials.TherewillnormallybeappointedaTrialLeader,whoisemployedor
hiredbytheShipbuilder.TheTrialLeaderisresponsibleforallphasesofthetrial,butlikeallgood
executiveshemightdelegateexecutionofdifferenttrialstodifferentpersons.Whenyou
participateintheseatrialsitisveryusefultolocateandcontacttheTrialLeader.
Thenthereareusuallyseveralshipmasterspresent.Theshipyardwillhaveatrialmastertobe
responsibleforthehandlingoftheship.Theshipownerwillusuallyhaveoneormoreshipmasters
present,usuallythemastersthataregoingtosailontheshipafterdelivery.
Measurementsareperformedbypersonnelengagedbytheshipbuilder,orbythirdparties(like
MARINTEK).

11.1.2 Speedtrials
Thepurposeofthespeedtrialsistodeterminethespeedpowercharacteristicsoftheshipatthe
contractualconditions,whichisusuallydeep,calmwater,nowind,seatemperatureof15degC,at
aspecifieddraught,usuallydesigndraught.Sincethetestscanrarelybecarriedoutunderthese
conditions,resultswillusuallyhavetobecorrectedforthedifferencebetweenspecifiedandactual
conditions.
ProceduresforhowtodoseatrialsaregivenintheinternationalstandardISO19019andinthe
ITTCstandardprocedureProcedureforthePreparationandConductofSpeed/PowerTrials.A
shortsummaryisgivenhere:
Measurements:

Thespeedismeasuredby(D)GPSorbyclockingthetimeusedtotravelameasuredmile.
TheshaftpowershouldbefoundfromRPMandshafttorquemeasuredbystraingauge
measurementonthepropellershaft(s),orbyspecialtorquemeters.Inpractice,theextra
costofshaftmeasurementsisoftenavoidedandpowerisfoundfromfuelconsumption
readings,butthismethodreliesonenginemanufacturerdataonspecificfuelconsumption,
whichisoftennotveryreliable.

Watertemperatureshouldbemeasured.Waterdensityshallpreferablybemeasured,but
isoftenfoundfromthetemperatureandstandardtables,basedonassumedsalinity.

Windspeedanddirectionshouldbemeasured.Onboardwindindicatorsmightbeutilized,
butitispreferabletobringaportablewindmeterforanextracheckastheship
instrumentsmightnotbeproperlycalibrated.

Waterdepthshallbenotedandincludedinthereport.Foundfromechosounderorcharts.

Waveconditionsshallbedocumented.Wavemeasurementsbywavebuoyarepreferable,
butseldomdone.Instead,visualobservationandestimationofwaveheightanddirectionis
made.Thebestwaytodothisistohaveseveralexperiencedpersonsdotheirown
estimates,andthentaketheaverage.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page105

> 5 min and 1 mile


Steady Approach

Steady Approach
> 5 min and 1 mile

Figure11.1

Recommendedtrackforconductingspeedtrials

Figure11.2

Measurementofshafttorque

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page106

Figure11.3

Measuredmile

Performance:
Thespeedtrialsshouldalwaysbedonerunningthesametrackinbothdirections.Aftercorrecting
theresultofeachrunforwindandwaves,theresultsareaveragedinordertocorrectfortheeffect
ofcurrent.
Itisimportantthatthecourseandspeedhavereachedsteadyconditionsbeforetherunstarts.This
isachievedbyasuitableapproachrun,asindicatedinfigures11.1and11.3.
Useofrudderduringtherunshouldbeminimized.Iftheshipisreasonablydirectionallystable,itis
besttoturnofftheautopilotandrunwiththerudderfixed,orcontrolledbyanexperienced
helmsman.Ifuseofrudderisrequireditshouldbenotedinthereport.
Environmentalconditions:
Typicalcontractualconditions:

Seastate:
o

Nowaves

Inpractice:Beufort1(Waveheight0.1m)

Wind
o

Nowind

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page107
o

Inpractice:Beufort2(Windspeed6knots)

Waterdepthh
o

Deep,

Inpractice:

6.0

and

(AMismidshipsectionarea)

Current
o

Nocurrent

Nopracticallimitforwhencorrectionsaremade.Useofdoublerunsmeansthat
correctionsarealwaysincluded

Maxacceptabletrialconditions:

Seastate
o

Preferablyseastate3

Ultimatelyseastate5(oruptoseastate6forshipswithL>100m)

Wind
o

Beufort6(20knots)(forshipswithL>100m)

Beufort5(forshipswithL100m)

Waterdepthh
o
o

6.0

and

2 (AMismidshipsectionarea)
Smallerdepthsrequirecorrectionsforshallowwater

Current
o

Currentofmorethanafewknotsisunacceptable

Whenenvironmentalconditionsareabovethecontractualconditions,correctionswillhavetobe
done.ProceduresfordoingsuchcorrectionsaregiveninInternationalStandardISO15016,andin
ITTCstandardprocedureProcedurefortheAnalysisofSpeed/PowerTrialData.Correctionsfor
smalldifferencesindraught,differenceinwatertemperatureanddensity,wind,andshallowwater
arestraightforwardtoapplyandfairlyaccurate,whilecorrectionsforwavesanduseofrudderis
difficultandnotveryreliable.Fordetails,youarereferredtotheabovementioneddocuments.The
ITTCstandardisfairlyclosetotheISOstandard,butusingbothmightbehelpful.

11.1.3 Bollardpulltests
Testingofbollardpulliscommonaspartofdeliverytrialsfortugsandanchorhandlers.Avery
ruggedpullmeterisapplied.Theimportantpartoftheprocedureistominimiseeffectsof
boundaries(quay,bottom)(seeFigure11.4),andlookforeffectsofwaterrecirculation(seeFigure
11.5).

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page108

Goodlocation

Poorlocation

Figure11.4

Goodandbadlocationsforbollardpulltest

Figure11.5

Selectingtheproperpartofthetimeseriesfromabollardpulltest.

11.1.4 Manoeuvringtrials
Trialtypesandexecutionarethesameasinmodelscale.Standardtestsarezigzag,turningcircle
andstoppingtest.
Measurements:

(D)GPSpositionmeasurement

Gyrocompasscourse

Rudderangle

Propellerrevs

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page109

11.1.5 Highspeedcraft
Highspeedcraftwilltypicallydospeedtrialsthesamewayasconventionalships.ThenewIMO
Highspeedcraftcodeincludesrequirementsforseakeepingandmaneuveringtests(seeIMO:2000
HSCCode(IMO185E),Annex9).Thesetestscanbesummarizedasfollows:

Stopping
o

Normalstopfrommaxspeedtozero

Emergencystop

Crashstop

Cruiseperformanceintwoseastates
o

Normalconditions

Worstintendedconditions

Measurementsofaccelerations,speed,waveheading

Failuretests
o

Checkthattheship,crewandpassengersarenotatriskifforinstancethesteering
fails

11.1.6 Wavemeasurements
Oneofthemaindrawbacksoffullscaletestingistheproblemofcontrolandmonitoringofthe
environment,andespeciallythewavecondition.Controlofthewavesisofcourseoutofthe
question.Wavemonitoringisalsosurprisinglydifficult.Thereareseveralalternatives:

Visualobservationcheapandeasy,butnotreliable.Especiallywavelengthismissing.

WavebuoyThebestalternative,butwhenthereisnostationarybuoyinthearea
deployingandespeciallyrecoveringthebuoywilloftenrepresentaseriouspractical
challenge.

WaveradarThereareseveraldifferentconceptsonthemarket.Dedicatedmicrowave
radarsarebigandusuallymuchtoocostlyforordinaryseatrials.SystemsusingtheXband
radaralreadyontheshipareusedonshipstosomeextent,buttheyhavehadproblems
measuringthecorrectwaveheight.Distributionofdirectionandwaveperiodisfairlygood
onthesesystems.OneexampleofsuchaproductisWavex,manufacturedbyMirosA/Sin
Asker,Norway.Thecomplexityofinstallationofsuchsystemsmeansthattemporary
installationforthepurposeofseatrialsisnotanattractiveoption.

Bowmountedaltimeter.Measurestherelativedistancebetweenthebowandthewaves.
Requiresthenaccuratemeasurementoftheverticalmotionsatthebow.Effectonthe
incomingwavesfromthepresenceoftheshipmightbeaproblem.Fairlyreliableheight
andperiod,butwavedirectionmustbevisuallyestimated.

Forvisualobservations,theBeufortwindandwavescaleisuseful.However,itshouldbenotedthat
therelationbetweenwindandwaveheightisonlyvalidforopensea.Withlimitedfetch,thewaves
willbelower,shorterandsteeper.Waveheightandperiodcanalsobedirectlyestimatedfrom
windspeedandfetchlength.
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page110
Sea Description term
Beufort state Wind
0
0 Calm
1
0 Light air
2
1 Light breeze
3
2 Gentle breeze
4
3 Moderate breeze
5
4 Fresh breeze
6
5 Strong breeze
7
6 Near gale
8
7 Gale
9
8 Strong gale
10
9 Storm
11
9 Violent storm
12
9 Hurricane
13
9 Hurricane
14
9 Hurricane
15
9 Hurricane

11.2

Wave
Calm
Ripples
Small wavelets
Large wavelets
Small waves
Moderate waves
Large waves
Large waves
Moderately high waves
High waves
Very high waves
Exceptionally high waves
Exceptionally high waves
Exceptionally high waves
Exceptionally high waves
Exceptionally high waves

Wind sp. [knots] Wave height [m]


min
max
Probable Max

0
1
3
6
10
16
21
27
33
40
47
55
63
71
80
89

1
0
0
3
0.1
0.1
6
0.2
0.3
10
0.6
1
16
1
1.5
21
2
2.5
27
3
4
33
4
5.5
40
6
7.5
47
7
10
55
9 12.5
63
11.5
16
71
14
16
80 >14
>16
89 >14
>16
99 >14
>16

Table11.1
TheBeufortwindandwavescale.Seealso
http://ioc.unesco.org/oceanteacher/resourcekit/M3/Formats/Codes/SeaState.htm

Shipmonitoringsystems
Shipmonitoringsystemsaresystemsthatmonitorcertainaspectsoftheperformanceoftheship
duringnormaloperation.Thus,thisisgenerallysystemsthatarepermanentlyinstalledonboard.
Thisisafieldinrapiddevelopment,drivenbytechnologicalprogressandtheincreasingneedfor
shipownerstogetdetailedandreliableinformationabouttheperformanceoftheirfleet.Examples
ofsuchmonitoringsystemsarediscussedbelow.

Figure11.6

Hullmonitoringsystem

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page111

11.2.1 Hullmonitoring
Thisismonitoringofstrainatstrategiclocationsinthehullgirder.Theprimarypurposeisto
monitorthedevelopmentoffatiguedamage.Itismandatorytohavesuchasystemonbulkcarriers
above20000tonnes.Thesystemwillmonitorthelocalstrainandcomputetheexpectedremaining
fatiguelife.Itisalsousedtoavoidoverloadingwhenloadingandoffloading.Therearemany
suppliersofsuchsystemsandthetechnologyiswelldeveloped.Straingaugesarebyfarthemost
commonsensortechnologyforthispurpose,butfiberopticshasalsobeenused.Figure11.6shows
anexampleofhullmonitoringsystem.

11.2.2 Performancemonitoring
Monitoringofspeedandpowerperformanceisperformedinordertooptimizetheoperationand
maintenanceofmerchantships.Theshipownerwantstomonitorhowthepowerrequirement
increaseswithtime,sothatoptimumintervalsforcleaningofhullandpropellercanbeachieved.
Thepowersuppliedtothepropellerismeasured,preferablybyatorsion(orpower)meterinstalled
onthepropellershaft.SpeedismeasuredbyspeedlogandGPS.Itisthespeedthroughwater,
whichiswhatthespeedlogissupposedtomeasure,thatisneededfortheperformance
evaluation.However,thespeedlogmeasurementsarenotoriouslyunreliable,sincetheyare
locatedinthevicinityofthehull,sothattheyareinfluencedbytheboundarylayerandlocal
pressurefield.Sincethelocalpressurefieldmatters,wavesandshipmotionswillcausenoiseinthe
measurement.Marinegrowthonthesensorsisalsoacauseofmeasurementerror.
Amajorproblemwithperformancemonitoringistocorrectthemeasurementresultsforeffectsof
wind,waves,currentandloadingcondition.Tohelpinthecorrection,extrameasurements,like
shipmotions,incomingwaves,windetc.mightbeaddedtothesystem.
Traditionally,themonitoringofspeedandpowerperformancewasmadeintermsofnoonreports
madebythecrewandsenttotheheadofficebyfaxortelex.Increasingly,systemsthat
automaticallylogdataatmuchhighersamplingintervalsthanonceperdayareinstalled.The
systemsmightcollectthedataonboardorsendthemtoshorebysatellitelink.Thereisarapid
technologicaldevelopmentinthisfield.

11.2.3 HEMOSHealthMonitoringSystem
Anovelkindofmonitoringsystemismonitoringtheperformanceofonboardsystems,withtheaim
ofdetectingpossiblefaultsandtoseehowthesystemsareoperated.Aprimaryexampleofsucha
systemistheHEMOS,developedbyRollsRoyceMarine,inclosecooperationwiththeshipowner
Farstad.AsisillustratedonFigure11.7,thesystemiscollectingdatafrompropulsionunits,engines,
tunnelthrusters,aswellasoperationaldatalikespeed,headingandpositionandsendingthese
datainrealtimetoRollsRoyceofficesonshore.Thus,thisisprimarilyasystemthatallowsthe
equipmentsupplier,inthiscaseRollsRoyce,tomonitortheperformanceoftheirequipment.By
doingthis,theequipmentsuppliercanlearnmoreabouthowtheirequipmentisactuallyused,and
theymightpotentiallydetectfailingequipmentbeforeitcompletelystopsworking.Theideaforthe
systemistakenfromtheairenginebusinessofRollsRoyce,whereasimilarsystemhasbeenin
operationformanyyears,onmostoftheenginessoldbyRollsRoyce.CompetitorsofRollsRoyce
areworkingonsimilarsystems.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page112

Figure11.7

HEMOSHealthMonitoringSystembyRollsRoyceMarine

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page113

12

ERRORANALYSIS

12.1

Introduction
Itisimportanttokeepinmindwhenperformingallkindsofphysicalexperimentsthatthereis
inherentuncertaintyinallmeasureddata.Atestresultisreallyjustoneexampleoftherangeof
possibleoutcomesoftheexperiment.Iftheexperimentisrepeateditisveryunlikelythatexactly
thesamemeasurementresultisobtained.Itisofcourseveryimportanttoknowhowfarthe
measuredvaluecouldbefromtheunknown,truevalue.Toknowthis,onehastoperformsocalled
uncertaintyanalysis.Systematicuseofuncertaintyanalysisinshipandoffshoremodeltestingisa
relativelynewphenomenon,whichhasarisenduringthelast10years.Agooduncertaintyanalysis
isespeciallyimportantwhenusingexperimentstoverifycomputercodesoratheoreticalmethod.
VerificationofCFDcodesbymodelexperimentshasbeenadrivingforceinincreasingthe
awarenessofuncertaintyinconventionalhydrodynamictesting.
Whendesigninganexperimenttoansweracertainquestion,forinstanceiftheproposedbulb
designchangeofashipwillactuallydecreasethefuelconsumption,oneshouldalwayskeepin
mindwhatistheuncertaintyoftheexperimentcomparedtotherequiredaccuracyoftheanswer
tothequestion.Forthementionedexample,ifthemeasuredresistancereductionis2%andthe
uncertaintyofmeasurementis3%,wehaveinfactnotdocumentedaresistancereduction.To
documentaresistancereductionof2%,wewillneedtoperformanexperimentwithanuncertainty
thatissignificantlylessthan2%.
Inthischapterwewillstartwithashortintroductiontouncertaintyanalysisofexperimentsin
general.Thissection(11.2)isreliesheavilyonthelecturenotebySveinErsdal,whichisincludedin
AnnexD.ForanindepthtreatmentofthissubjectwerecommendthebookbyColemanandSteele
(1999).DetailsofuncertaintyanalysisoftowingtanktestsarefoundinLongoandStern(2005)and
intheITTCRecommendedProcedures,whichcanbefoundattheITTCPermanentwebsite:
http://ittc.sname.org/documents.htm
Afterthesectionongeneraluncertaintyanalysis,wegoontodiscussspecificerrorsourcesof
specialimportanceinexperimentalhydrodynamics.

12.2

Uncertaintyanalysis

12.2.1 BasicConcepts
Theaimofanuncertaintyanalysisistogiveaquantitativemeasureofhowreliableameasuredor
calculatedvalueis.Theworderrorisusedforthedifferencebetweenameasurementresultand
thetruevalue,whileuncertaintyisthestatisticalrepresentationoferror.
Theuncertaintyisusuallyquantifiedintermsoftheconfidenceinterval.A95%confidenceinterval
offorinstance2Nmeansthat95%ofallreadingsofaparticularmeasurementwillbewithin2N
fromthetruevalue.Inotherwordsitmeansthattheprobabilitythatthetruevaluewillbewithin
theconfidenceintervalis0.95.
Twotypesoferrorsareconsideredintheanalysis:biasandprecisionerrors.Biaserrorsare
systematicerrors,errorsthatarenotrevealedbyrepetitionoftheexperimentwhileprecision
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page114
errorsarescatterintheresults,foundbycomparingtheresultsofrepeated
measurement.Thereisreallynocleardistinctionbetweenthetwo,sincetheamountofbiaserror
dependsonhowlargepartoftheexperimentisrepeated.Forinstance,incaseofatowingtestofa
shipmodel,thebiasandprecisionerrorswillbeverydifferentiftherepetitiononlyincludes
runningthemodelseveraltimesinthetank,orifitinvolvesbuildinganew,butsupposedlyequal
modelandtestingitinanotherfacility.Itisstillimportanttodistinguishbetweenthetwotypesof
uncertainty,andwhenperformingtheuncertaintyanalysis,wemustchoosethereplicationlevel,
thatishowlargepartoftheexperimentalsetupthatisremadeaspartoftherepetition.Including
morefactorsintherepetitionreducesthebiaserror,andthatisagoodthing,sincethebiaserror
cannotbemeasured,butmustbeestimatedbysomekindofqualifiedguess.Calibrationisthe
keytoreductionandquantificationofbiaserror.Bymeasuringaknownquantitywiththesame
testsetupasusedontherealexperiment,onecanatleastintheory,eliminatethebiaserror.

12.2.2 Calculationofprecisionerror
Theprecisionerrorcanbecalculatedfromrepeatedmeasurements,andwewillshowthe
procedurehere.Forastart,itiscommontoassumethatifameasurementisrepeatedinfinitely
manytimesthemeasuredvalueswillfollowaGaussiandistributionaroundamean.TheGaussian
distributioniscalledtheparentdistribution.TheGaussiandistributionisgivenas:

1
f X
e
2

2 2

Whereisthemeanandisthestandarddeviation.ForNsamples,themean X isgivenas:

1 N
Xj
N
j 1

AndthestandarddeviationSxisgivenas:

1 N
XjX
N 1
j 1

Sx

Itshouldbenotedthatthemeanisitselfnormallydistributedwithmean.Thestandarddeviation
ofthemeansdependsonthenumberofsamplesaccordingtothefollowingrelation:

SX

SX

Fortheparentdistributiontheconfidenceintervalofasampleisgivenby:

Prob X j t X j t

Whereistheconfidenceinterval,typically=0.95.t1.96foranormaldistributionwhen=0.95
.Forafinitenumberofsamplesthestandarddeviationoftheparentdistribution,,isunknown.
Alsotisthenunknown.Thisequationcanberewrittenas:

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page115

Xj

Sx

Prob t

Wherethevariable

Xj
SX

israndomandfollowsaStudentstdistributionwithN1degreesof

freedom.Thevalueoftmaythenbefoundfrom:

t F 1 12 (1 )

WhereF1()istheinverseofthecumulativedensityfunctionforthetdistribution.Notethat
F1isafunctionofdegreesoffreedomN1.Nisstillnumberofsamples.Therelationshipbetweent
andnumberofsamplesisgiveninFigure12.1.Itisseenthatwhenthenumberofsamplesincrease
thevalueoftgoestowardsthevalueoftforanormaldistribution.tcanbecomputedinExcelby
thefunctionTINV(10.95;N1)(inthiscasefor95%confidence).
Theprecisionlimitforasampleisnoweasilyfoundfrom:

Px t S x

Notethatthisistheprecisionlimitofonesampleofanexperiment,butthattocalculatethis
precisionlimityouneedtohavemorethanonesample.Wewilldiscussthismorethoroughlylater.
TheprecisionlimitofthemeanofNrepetitionsisgivenby:

PX t

SX

8
95% confidence

99% confidence
6

Weight t

5
4
3
2
1
0
0

10

15

20

25

30

Degrees of freedom (N-1)

Figure12.1

TheweighttforestimatingconfidenceintervalsusingStudentstdistribution

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page116
ItisseenfromFigure12.1thatthevalueoftdecreasesrapidlywithincreasingnumberof
samplesforlessthanabout10samples.Fromthepreviousequationitisseenthatthiswillresultin
rapiddecreaseintheprecisionlimitofthemeanofthesamples.Thus,repeatingexperimentsupto
about10timesisanefficientwayofdecreasingtheprecisionlimit.Theprecisionlimitofone
sampleshallideallynotchangewithchangingthenumberofsampleswhenNisincreasingtwill
decrease,butSXwillincreaseduetomoresamples.ThatisthepointoftheStudentstdistribution,
tocompensateforthefactthatSXwillbemuchsmallerthanwhenthenumberofsamplesislow.
However,theuncertaintyofSXwilldecreasewithincreasingnumberofsamples.Thus,tocalculate
asensiblevalueforSXyouneedseveralsamples.
Soinaworldwithoutsmallbudgetsandnotenoughtimeallexperimentswouldberepeatedat
least10times,sowecouldcalculatetheprecisionerrorandalsoreducetheuncertainty.Inpractice
thisisseldompossible.Itisalsousuallynotrequiredtorepeattheexperimentinordertoreduce
theuncertainty.Inordertocalculatetheprecisionerroritisstillrequiredtohaverepeatedtests.
Thisisoftensolvedbyrepeatingonlyoneofmanyconditions,andassumethattheprecisionlimit
foundforthisconditionisrepresentativealsofortheotherconditions.Forinstanceforatowing
testwithashipmodelyouwouldonlyrepeatoneofthespeeds,notall,andonlyatoneloading
condition.Itcouldalsobearguedthatyoucouldusetheprecisionlimitfoundforonelargetanker
modelforallsimilarlysizedtankermodels,andsoon.

Example:Calculationofprecisionlimitforatowingtest
Thetowingtestwithashipmodelisrepeated15timesforonespeed.
Thestandarddeviationofthemeasuredresistanceforthe15testsisSX=0.185N.
Thisgivesaprecisionlimit PX S X t15 0.185 2.145 0.396 N
Theaverageresistanceofthe15testsis41.65N.Theuncertaintyoftheresistancemeasurement
(ofasingletest)isthen:

0.396
0.0095 0.95%
41.65

Thestandarddeviationofthemeanofthe15testsis

0.185
0.0478 .
15

Thisgivesaprecisionlimitforthemean PX S X t15 0.0478 2.145 0.102 ,whichgivesan


uncertaintyof

0.102
0.002 0.2% .
41.65

Thus,itisseenthatanumberofrepeatedtestsisrequiredtocalculatetheprecisionerrorofa
singletest,andthattheuncertaintyofthemeanofseveraltestsissignificantlysmallerthanthe
uncertaintyofasingletest.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page117

12.2.3 Chauvenetscriterionforrejectingoutliers
Datasamplesthatdeviatesignificantlyfromthemajorityofthedatasamplesarecalledwildpoints,
oroutliers.Ifnophysicalexplanationisfound,thedatapointmightstillbedisregarded,evenifno
particularerrorsourceisidentified.Anacceptedcriterionforrejectionofasample,whenno
particularreasonfordeviationisfound,isChauvenetscriterion.Itstatesthatsampleswithina
bandaroundthemeanwithprobabilityofexceedencelessthan 1 21N shouldberetained.This
limitcanbeexpressedasaweightonthestandarddeviation:

tchauvenet F 1 12 (1 p)
Where p 1 21N isthelimitprobabilitybasedonNsamples.NotethatinthiscaseF()isthe
cumulativedensityfunctionofthenormaldistribution,alsoforsmallvaluesofN.Valuesoftchauvenet
isplottedinFigure12.2.
Sampleswithhigherdeviationfromthemeanthan:
X j X tchauvenet S x

maythenbedisregarded.

Figure 12.2

The weight tchauvenet in Chauvenets criterion for rejection of outliers

12.2.4 Estimatingbiaserrors
Incontrasttoprecisionerrors,thereisingeneralnostraightforwardwayofquantifyingbiaserrors.
ITTChasprovidedanexamplecalculationofuncertaintyofatowingtest,whichcanbeusedasan
exampleofhowtoestimatebiaserrors.Agoodexamplethereistheestimationofthebiaserror
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page118
relatedtotheaccuracyofthegeometricalshapeofthemodel.Itisassumedthatthe
modelingaccuracyis1mm,andthatthemodelisballastedtocorrectweight,notdraught.The
maximumerrorindisplacementcanbecalculatedfromthemodelingaccuracy,andtheeffecton
resistanceofincreasedordecreaseddraughtduetoerrorinvolumeiscalculated.Thisthen
representthebiaserrorofmodelingaccuracyonmodelresistance.
Animportantsourceofbiaserroriscalibration.Ifthecalibrationofasensorisincorrect,thiswill
turnoutasabiaserror.Uncertaintyofcalibrationfactorscaninturnsbedividedinprecisionand
bias,wheretheprecisionerrorcanbefoundfromrepeatedcalibrations.Biaserrorsofthe
calibrationcanbehardtodetermine,anditcouldbetemptingtoignoreit.Goodcalibration
procedureswillensurethatcalibrationbiaserrorsaresmall.
Anothersourceofbiaserrorisblockageeffect,andtheprocedureusedtocorrectforblockage
effect,ifthatisapplied.Inthiscaseitseemsreasonabletosearchtheliteratureforalternativeways
ofcorrectingforblockage,comparethem,andmakeaqualifiedguessoftheuncertaintybasedon
thecomparison.

12.2.5 ErrorPropagation
Whencalculatingtheuncertainty,oneshouldconsiderthefinalendresult,andbyfinalendresult
wemeantheanswertothequestioninitiallybeingasked.Ifthequestioniswhatisthetotal
resistanceofaparticularshipmodel?,askedforinstanceaspartofvalidationofaCFDcode,one
neednotincludethescalingtofullscaleintheuncertaintyanalysis.Ifthequestioniswhatisthe
shipresistance?,alsotheconversiontofullscalemustbeincluded.Theequationsrelatingthe
measuredquantities(intheaboveexamplethatmightbethenumberofpulsesfromthecarriage
speedpulsecounterandthevoltagereadfromthedynamometerstraingauges)withthefinalend
result,arecalleddatareductionequations.Thedatareductionequationsaffecthowthe
uncertaintyofindividualfactorsinfluencesthetotaluncertainty.Thedatareductionequations
mightalsocontributetothebiaserrors,incasetheyarenotanentirelytruerepresentationofthe
realworld.Anexampleofthisisthescalingofshipresistancetofullscale;somethingthatwe
knowisonlyanimperfectassumptionofarealitytoocomplextomodelentirelycorrect.
Wewillnowbrieflyshowhowtheinfluenceonthefinalendresultofthedifferenterrorsourcesis
calculated.Thereductionequationcaningeneralbewrittenas:

X f r Y1 , Y2 ,..., YN ,
WhereXistheresultforwhichtheuncertaintyissought,fristhefunctionalrelationandYiisthe
parametersonwhichXdepends.AssumingthatasmallchangeinaparameterYiresultsinasmall
changeinX,Taylorexpansiongives:

X
1 2
X
Yi
Yi 2 O(Yi 3 )
2

2
Y

Y
i 1
i
i
N

X
X

X
Yi

X
i 1 Yi
i

Fromthistheinfluencecoefficientisdefinedas

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page119

Yi

andtheelementalerroristhen:

ei X Yi i Yi
Yi
Theelementalerrorsarecalculatedindependentlyforbiasandprecisionerrors.Forprecision
errors Yi canbefoundfromrepeatedmeasurements,whileforbiaserrors Yi mustbe
estimated,asdiscussedpreviously.Therecommendedpracticeistocalculate Yi to95%
confidencelimit.Calculationoftheinfluencecoefficientsrequiresknowledgeofthemathematical
relationbetweentheparametersandtheresult.Ifelementalerrorsourcesareindependentthe
combinedeffectoftheerrorsarefoundbysummationas:

ei 2

i 1

Tofindthetotalerror,biasandprecisionerrorsmustbecombined.Approximately95%coverageof
thetruevalueisreachedbyusing:

e eS2 eB2
WhereeSisprecisionerrorandeBisbiaserror.SimplesummationofeSandeBgivesabout99%
coverage.
If
X isthemeasurementandeisthetotalerrorfoundusingtheaboveprocedurewitha
confidenceintervalof95%,thenweknowthatthetruevalueXisfoundintheinterval

X
X e
withaprobabilityof95%.Thus,theerroreisgiveninthesamephysicalunitasthemeasurementX.
Itiscommontoquotearelativeerror,definedas:

e
er
X

12.3

DiscussionofErrorSources
Performinguncertaintyanalysisasbrieflyoutlinedabove,andasdescribedbyforinstanceColeman
andSteele(1999),mightbeverytimeconsuming,easilymoretimeconsumingthantheexperiment
itself.Inroutinetesting,oneusuallyreliesonpastexperienceofagreementbetweenmodeltest
resultsandfullscaletrials,whichgivesagoodindicationoftheinherentuncertaintyoftheend
result.Duetothedifficultiesinestimatingthebiaserrors,thisgutfeelingfortheuncertainty
mightbejustasreliableastheresultsofaformaluncertaintyanalysis.However,theformal
uncertaintyanalysiswillprovidedocumentation,anditwillgiveinformationaboutwhatarethe
mostimportantcontributionsforthetotaluncertainty,thusgivinginformationaboutwhatpartsof

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page120
theexperimentalsetupthatshouldbeimproved.Fornoveltesttypes,wherethereisno
largedatabaseofexperiencetoprovideproofthatthetestgivesvalidresults,itisindeed
recommendedtoperformanuncertaintyanalysisasanintegralpartoftheplanningandsettingup
oftheexperiment.
InadditiontoColemanandSteele,theproceedingsandRecommendedProceduresofITTCprovides
specificguidanceonhowtoperformuncertaintyanalysisofstandardshipmodeltests.Asastarting
pointwewillinthefollowingdiscusssomeofthemostimportanterrorsourcesofmodeltestingof
shipandoffshorestructures:
1. Scaleeffects
2. Incorrectmodellingofstructures(geometry,weightdistributionetc)
3. Incorrectmodellingofenvironment
4. Instrumentationandmeasurementserror
5. Errorinanalysisandinterpretationoftestsresults.
Thelasterrorsourcearenotlimitedtomodeltests,butareequallyimportantforevaluationof
numericallyobtainedresultsandforresultsfromfullscalemeasurements.
Itshouldbenotedthatexceptitem4Instrumentationandmeasurementserrorinthelistabove,
thisismostlysourcesofbiaserror.Theprecisionerrorsareeasiertoidentify,butharderto
generalise,sincetheyusuallydependontypesoftransducersanddetailsoftheexperimental
facilitiesandtestsetup.

12.3.1 Reynoldsnumberscaleeffects
Asdiscussedinchapter2thedifferentrequirementstoachievedynamicsimilarityinmodelandfull
scalecaningeneralnotbesatisfiedsimultaneously.Testingofshipsandoffshorestructuresarefor
mostpracticalsituationinfluencedbysurfacewaveeffects,eitherfromincomingwavesorwave
generatedbythemotionsofthestructure.Gravitationalforceswillgovernsurfacewaveformation.
ThisimpliesthatfortheseconditionsequalityinFroudenumberinmodelandfullscalemustbe
achieved.Froudescalingisthereforeassumedasbasisforthediscussionofscaleeffects.
ApplyingFroudescalingthedifferenceinfullscaleandmodelscaleReynoldsnumberwillbe
(assumingthesamefluidviscosity):

3
ReF
LU
F F 2
ReM LM U M

Usingascaleratioof1:50givesafactorforReynoldsnumberequaltoabout350.Thedifferencein
Reynoldsnumberisthemainsourceforscaleeffectsinmodeltestingofshipandoffshore
structures.Theimportanceofthescaleeffectsandhowtoaccountforthemwilldependonthe
actualtypeoftests.Scalingoftowingresistanceofshipsistheclassicalscalingproblemanditis
probablytheonlycaseforwhichareasonablyrationalandprovenscalingprocedureexists.The
scalingproceduresusedtodayforshipresistanceisaresultoflargeeffortsoveraverylongperiod
oftime.ThistopicisdiscussedindetailinthebasiccoursesinHydrodynamicsandwillnotbe
treatedhere.ReferenceismadetoSteen(2011).
Inthefollowingscaleeffectswillillustratedbyconsideringtwodifferentcases,thewake
distributionatpropellerpositionofashipandthedragforcesoncylindricalelements.
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page121

12.3.2 ScalingofWakedistribution
DuetothedifferencesinReynoldsnumberinmodelandfullscaletheremaybescaleeffectsonthe
wakedistribution.Differenceinwakefractionwillinfluencepropellerthrustandruddernormal
forces.Itisthereforeimportanttoaccountforthescaleeffects.Apracticalmethodtoscalethe
wakedistributionhasbeensuggestedbySasajimaandTanaka(1966).Thebasicideasofthe
methodare:

Theviscouswakefieldisobtainedbysubtractingthepotentialwake(obtainedfrom
calculationsorexperiments)fromthemeasuredtotalwake

Theviscousmodelwakeshouldbecorrectedinsuchawaythatwhengoingfrommodel
tofullscalethemomentuminthewakeshouldbereducedbythesamefractionasthe
viscousresistance.Thisisobtainedbymovingallpointsonthewakecontoursinthe
maptowardsthecentreplanebytheratiooffullscaletomodelscaleviscous
resistance

Addthefullscaleviscouswakefieldtothepotentialwake(notinfluencedbyscale
effects)toobtainthecompletefullscalewakedistribution.

AnexampleofmeasuredandscaledwakedistributionisshowninFigure12.3.(fromHuse,1974).

Figure 12.3

Original model wake distribution (left) and scaled wake distribution (right).

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page122

12.3.3 Scalingofdragforces
InFigure12.4andFigure12.5thedragcoefficientforasquaredshapedcylinderwithsharpcorners
andacircularcylinderareshownasfunctionofReynoldsnumber.ThedragcoefficientsCDisgiven
fromthedragforceaccordingto:

dFD 1 2 C D DU 2

Theresultsareforsteadyflow.Forthecylinderwithsharpcorners,Figure12.4,itisseenthatthe
dragforceisalmostindependentofRe.Thisimpliesthatforstructureswithsharpcornersthescale
effectswillbenegligibleandFroudescalingcanbedirectlyappliedwithoutcorrections.
ForthecircularcylindercaseapronouncedeffectofReisobserved.Fourdifferentflowregimesare
defined.ThedefinitionoftheseregimesisincludedinFigure12.5.(fromSarpkayaandIsaacson
(1981)).ThescaleeffectscaninprinciplebeevaluatedusingtheresultsgiveninFigure12.5by
comparingtheCDvalueformodelscaleandfullscaleRevalues.Forfullscaleconditionswewillfor
mostcasesbewithinthePostSupercriticalflowregime(Re>3*106),whichgives C D 0.7 ,see
Figure12.5.InmodelscalewewillusuallybewithintheSubcriticalflowregime,whichgives

C D 1.2 .Thisimpliesthatforthiscasetheviscousdragforcesaresignificantlyoverpredictedin
themodeltests.Thescalingerrorisobtainedastheratiobetweenthemodelscaleandfullscale

C D values.
ThedragforcewillinadditiontotheReynoldsnumber,dependonanumberofparameters,
includingsurfaceroughness,KCnumber,shapeofthestructure,orientationofstructurerelativeto
flowdirection,interactioneffectbetweendifferentmembersetc.TheRedependencywillin
generalbedifferentdependingontheactualcondition.Thisimpliesthatthesimplescalingmethod
appliedaboveforsteadyflowpastacircularcylindercanonlygivearoughestimateoftheactual
scaleeffects.Insummarynorationalprocedureexistsforaccuratescalingofdragforcesfora
generaltestcondition.

Figure 12.4

Drag coefficient for squared cylinder with sharp corners in steady flow.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page123

Figure 12.5

Drag coefficient for circular cylinder in steady incident flow.

12.3.4 Scaleeffectsonresponses
Ingeneralscaleeffectsmaybeimportantforallcaseswhereviscousforcesgivesasignificantforce
contribution,eitheraspressuredragorasskinfrictionforces.Tomakeanevaluationofpossible
errorsourcesfromscaleeffectsitisthereforenecessarytohaveagoodunderstandingofthe
physicsoftheactualhydrodynamicproblem.
Forwavefrequencymotionsonfloatingstructurestheexcitationforcesarenormallydominatedby
massforcesandnotbydragforces.Ifthefrequencyrangeisfarfromresonancefrequencies,the
motionresponsewillnotbeinfluencedbydampingforcesandthescaleeffectswillbenegligible.
Closetoresonancefrequency,dampingisimportantandthescaleeffectsforthewavefrequency
motionscanbeimportant.
ForHighfrequencyandLowfrequencymotionsofstructureswearedealingwithalowdamped
systematresonance.Forsuchsystemsthemotionresponseswilldependonthedampingandscale
effectswilloccurifviscousforcesrepresentanimportantdampingcontribution.Thisisoftenthe
case.InANNEXCapracticalproceduretoquantifythescaleeffectsonLowFrequencymotionsofa
mooredvesselisgiven(HuseandMatsumoto,(1989)).
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page124

12.3.5 Errorsduetostructuralmodelling
Itisofpracticalreasonsnotpossibletoincludealldetailsandmechanicalpropertiesofafullscale
structureintoamodel.Simplificationshavetobemade.Errorsduetostructuralmodellingcanbe
spitintotwodifferentcategories:
1. Modellinginaccuraciesthatincludeerrorsinshape,weightdistributionetc.
2. Modellingsimplifications.
Modellinginaccuraciescanbecontrolledbycarefulcheckingofthemodelpriortotesting.Itis
commonpraxistocheckallmaindimensionsofthemodelaswellasweigh,draft,trimand
metacentricheight(byinclinationtests)beforethetestprogramisstartedandanydeviationcanbe
identified.
Itismoredifficulttoassesstheeffectofmodelsimplifications.Asanexample,themooringsystem
offloatingstructureswasintheearlystageoftestingofoffshorestructure,oftenmodelledwith
horizontal,abovewatersprings.Theideawastogetthecorrectrestoringeffect.Lateronitwas
foundthatthedynamicsofthemooringlinescouldgivealargecontributiontothelinetensionand
alsogiveanimportantcontributiontothelowfrequencydamping.Theusedsimplificationin
mooringsystemmodellingcouldthereforeintroducealargeunderestimationthemaximum
mooringlineforcesandalsooverpredicttheslowdriftmotions.
Ingeneraloneshouldalwaystrytomodelthepropertiesofthefullscalestructureascloseas
possible.Ifsimplificationsareintroduced,possibleconsequencesforthetestresultshavetobe
carefullyevaluated.

12.3.6 ErrorsduetoEnvironmentalmodelling
Modellingoftheenvironmentinmodeltestingcanrepresentanerrorsourceduetothefollowing
factors:
1. Waves(waveheightandperiod),windandcurrentmodelling
2. Physicallimitationsofthetestfacility.
Forwaves,windandcurrentthevariationinspaceandtimecanrepresentanimportanterror
source.Themostimportantphysicallimitationsofthetestfacilityarethetankwallsandthelimited
waterdepth.

12.3.6.1

Waveparametersandspectralshape

Whenpropagatingfromthewavemakerthepropertiesofthewavesmaychangedownstream.This
isinparticularaproblemforshortandsteepwavesandmostpronouncedinalongandnarrow
towingtank.Duringthewavecalibrationthewavemeasurementsshouldthereforebecarriedout
indifferentpositionsalongthetanktoverifythewavedataalongtheentiretesttrack.
Fortestinginirregularwavestheshapeofthewavespectrumisanimportanterrorsource,see
GuedesSoares(1990).Characterisationoftheusedspectrumonlybythestandardparameters,Hs,
Tpandisthereforenotsufficientanditisrequiredtoensurethatthemeasuredspectrumshapeis
inagreementwiththetheoreticalshape.Forcomparisonwithnumericalcalculationsthemeasured
spectralshapeshouldbeusedinthecalculations.
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page125

12.3.6.2

TankWallEffectsWavereflection

Themodelwillgenerateitsownwavesystem.Fortestingofstationarystructuresdiffractedwaves
andradiatedwavesduetotheship/platformmotionswillbegenerated.Thesewaveswillreachthe
tankwallandthewavegeneratorandwillbereflectedbacktothemodel.Inthiswayatransverse
wavesystemwillgraduallybedeveloped.Forshipswithforwardspeed,thestationarywavesystem
duetotheforwardspeedwillalsobereflectedfromthetankwalls.
Thetankwallinterferenceeffectscanhaveaveryimportanteffectontheexperimentalresults.In
Figure12.6resultsfrommeasurementsofverticalwaveforcesactingonafixedhemispherein
regularwavesareshown,seeZhaoet.al(1988)forfurtherdetails.Thetestwascarriedoutina
conventionaltowingtank.Theintentionwiththetestwasvalidationofanumericalcode.
Calculationswerecarriedoutbothwithandwithouttankwalleffects.Itisseenfromthefigurethat
thetankwalleffectsareveryimportant.Atsomewavefrequencies,correspondingtothe
resonancefrequenciesofthetransversewavesysteminthetank,theverticalforcedropstoavalue
thatisanorderofmagnitudelowerthanthepredictedvaluewithoutwallreflections.Includingthe
tankwalleffectinthenumericalmodelisseentogiveaverygoodagreementbetweentestsresults
andcalculations.Thisexampleclearlyillustratestheimportanceofusingequivalentconditions
whencomparingmodeltestresultsandnumericalcalculations.

Figure 12.6

Comparison between numerical and experimental results for first order vertical
wave forces on a hemisphere Effect of tank wall interference

TheresultsinFigure12.6clearlyshowthelimitationofusingconventionaltowingtanksfortesting
ofstationaryorlowspeedmodels.Ifusedforthistypeoftestinggreatcareisrequiredinthe
interpretationofthetestresults.Usingalargebasinwithwavebeachesattwosideswilllargely
reducethisproblem,butthereflectedwavesfromthewavemakerandthesidewithoutabeach
maystilltosomeextentinfluencetheresults.

12.3.7 Instrumentationandmeasuringerrors
Theaccuracyoftransducerswillbeinfluencedbytheactualmeasuringrange,linearity,possible
hysteresis,tendencyofdrift,calibrationaccuracyetc.Itisthereforeveryimportanttoensurethat
theusedinstrumentationintestsetuphavetherequiredproperties.Choiceofinstrumentationhas
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page126
tobemadebasedonagoodunderstandingofthegoverningphysicalandhydrodynamic
phenomenatobetested.
Transducersareofteninfluencedbythetemperature.Forslowvariationsintemperaturethiscan
introducedriftinthemeasuredsignal.Thiseffectcanbecontrolledbyregularlyupdating
measurementstowardsaknownvalue(oftenbasedonzeroreadings).Ifthetemperaturechanges
abruptlyitwillbemuchmoredifficulttoidentifythetemperatureeffectonthemeasurement.An
exampleismeasureofimpactpressure.Whenthepressurecellhitthewaterthetemperature
changeabruptlyduetodifferenceinairandwatertemperature.Thetemperaturechangeoccursat
thesametimeastheimpactpressureanditisverydifficulttosplitbetweenthetwocontributions.
Directuseofthemeasuredsignalcanthereforegivetotallymisleadingresultsforthiscase.
Manysensorsbasedonstraingaugesexperienceanotabledriftinthesignalshortlyafterthe
amplifierhasbeenturnedon.Thisisduetotheheatproducedbythecurrentgoingthroughthe
straingauge,whichiscausingsmallthermaldeformationsofthemetaltowhichthestraingauges
areglued.Tominimisesucherrors,azeroreadingshouldbemadesometimeafterturningonthe
amplifiers.
Carefulcalibrationproceduresareimportanttominimisemeasurementerrors.Atcommercial
modeltestinstitutionsstandardproceduresforinstrumentcalibrationisincludedasapartofthe
QualityAssurance(QA)system.
Asaroughestimatethefollowingmeasuringaccuracycanbeapplied(basedonmodelscale):

Waveprobes:

1mm

Accelerometers:

0.5%ofmeasuredvalue

Forcetransducers:

15%ofmeasuredvalue

Opticalpositionmeasurements

0.5mm

Opticalanglemeasurements

0.1deg.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page127

13

MODELTESTSVSNUMERICALCALCULATIONS

13.1

General
InTable12.1anoverviewofthemostimportantqualitiesofphysicalmodels(whichmeansmodel
testing)andnumericalmodelsareshown.ThetableistakenfromAage(1992).
Qualities

PhysicalModels

NumericalModels

Representation

VeryGood

Limitedbyavailabletheoriesand
computerpower

Accuracy

Good

Goodwithinvaliditylimits

ScaleEffects

Yes

No

Reliability

VeryGood

Riskofhumanerrors

Credibility

VeryGood

Primafacienotgood

Flexibility

NotGood

Good

Execution

Long

Lowwithstandardprograms

Cost

High

Developmentcosthigh

Table12.1

PhysicalversusNumericalModels(fromAage(1992))

Thecommentslistedinthetablerepresentgeneralevaluationofthecapabilitiesofthetwo
differenttoolstoobtainreliableresults.Therankingwillofcoursedependontheactualcase,the
complexityoftheproblemandhowappropriatetheavailabletestfacilityandthenumericalcode
arefortheactualcase.
Themainadvantageofmodeltestsisthepossibilitytomodelverycomplicatedsituations.
Assumingarealisticmodelling,onecanbequitesurethatallimportantphysicalphenomenaare
properlycovered.Thisisinparticularlyimportantfornewconceptsorfornewapplicationsof
establisheddesignsolutions.Themainproblemswithmodeltestsarethelackofflexibilityin
changingdesigncondition,scaleeffects,andcosts.
Fromthetableitisconcludedthatmodeltestinghasanimportantroletoplayintheanalysisof
complexdesignsappliedingeneralenvironmentalconditions.
Asnumericalcalculationsevolveandbecomemoreaccurateandeasiertouse,theimportanceof
modeltestingforroutineverificationoftheperformanceofshipsandoffshorestructurebecomes
smaller.Calculationsareslowlytakingovertheroutinework,buttheneedforverificationof
numericalmethodsandthedevelopmentofnewandtechnologicallydemandingstructuresmeans
thatexperimentsandmodelbasinsarealmostasimportanttodayastheywere30yearsback.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

13.2

Page128

ModeltestsforValidationofNumericalCalculations
Validationofanumericalcodemeanstocheckthatthecomputerprogramisconsistentwiththe
physicalreality.Forthispurposemodeltestsplaysanimportantrole.
Tocarryoutacomputercodevalidationusingmodeltestresultsweneedtoidentifyerrorsources
bothforthemodeltestsandforthenumericalmodel.Inprincipleanuncertaintyanalysisshouldbe
appliedtobothnumericalanalysisandexperimentalresults.Aprocedureforuncertaintyanalysis
forseakeepingcalculationsisgivenbyFaltinsenandSvendsen(1990).Theerrorsourcesinthe
numericalcalculationsareclassifiedasnumericalerrors(meaningerrorsrelativetothetheoretical
basisoftheprogram),physicalerrors(whichiserrorscausedbythetheoreticalassumptionsand
simplificationsusedasbasisforthetheory)andhumanerrors.Theeffectofeacherrorsourcemust
besystematicallyinvestigatedbynumericalcalculations.
Itisnotpossibletodefineadetailedgeneralvalidationprocedure.Thevalidationprocedurewillbe
dependentontheactualproblemtobeconsidered,thenumericalcodetobevalidatedandwhat
typeofexperimentaldatathatisavailable.Apossibleprocedureforvalidationofacomputer
programforseakeepinganalysisisoutlinedinthefollowing.
Stepsinavalidationprocedureforwaveinducedshipmotions:
1. Equalmodelloadingcondition;Ensurethatthefollowingparametersisthesameinthe
modeltestsandnumericalcalculations:

Geometry

Vesseldraftandtrim

Metacentricheight(GMTandGML)

Radiusofgyrationrxxryyandrzz

2. Equalenvironmentaldata;Examinetheenvironmentaldatausedinmodeltests.

WaveheightHandperiodTatpositionoftestingforregularwaves.For
irregularwavesHs,TPandspectralshape.

Possibledeviationinwavedataalongtesttrack.

Effectofwaterdepth?Ifyes,shallowwatertobeincludedincalculations

Testresultsinfluencedbydiffractedandreflectedwaves,includingtankwall
effects?

3. Equaltestcondition;Ensurethatthemodeltestsandcalculationsarecarriedoutfor
thesametestcasesincludingparametersas:

Forwardspeed

Waveheading

Transienteffectsinmodeltests(andalsoincalculationsiftimesimulations)

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page129
4. Naturalperiods;Comparecalculatedandmeasuredresonanceperiodsin
heave,rollandpitch.Causestopossibledeviationstobeexplained(masstermsor
restoringterms,ifdecay;dampingcontributions).
5. Influenceofdifferenceinconditionsbetweenmodeltestsandprototype;Establishby
numericalcalculationstheeffectofpossibledifferencesbetweenmodeltestconditions
andcalculationconditioncaseontheresults.
6. Influenceoferrorrangesonresults;Establishbynumericalcalculationstheeffectof
errorrangesforthedifferentparametersonthefinalresults.
7. Comparisonofwaveresults;
RegularWaves:

RAO

Phase

Addedresistance/speedreduction

Irregularwaves:

Standarddeviation

Statisticaldistribution

Extremevalues

RAOs

Addedresistance/speedreduction

(Greenwater,ImpactLoadsetc.,dependingonactualcase)

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

14

Page130

REFERENCES
Aage,C.(1992),RelevanceofModelTestinginIrregularSeasandCurrent,Proc.ofOMAE1992.
Bendat,J.S.andPiersol,A.G.(1966).MeasurementsandAnalysisofRandomData,JohnWiley&
Sons,Inc.
Berg,T.E.(2000),ShipManoeuvring,DepartmentofMarineHydrodynamic,NTNU.
Buchner,B,Wichers,J.E.W.anddeWilde,J.J(1999),FeaturesoftheStateoftheArtDeepwater
OffshoreBasin,OffshoreTechnologyConference,paperOTC10814,Houston,Texas
Chakrabarti,S.K(1999).ModelingLawsinOceanEngineering,inDevelopmentsinOffshore
Engineering,WavePhenomenaandOffshoreTopics,EditedbyJ.B.Herbich,GulfPublishing
Company,ISBN0884153800
Coleman,H.andSteele,W.G.(1999).ExperimentationandUncertaintyAnalysisforEngineers,
SecondEdition,WileyInterscience,ISBN0471121460
Dean,R.G.andDalrymple,R.A.(1991).WaterWaveMechanicsforEngineersandScientists,in
AdvancedSeriesonOceanEngineeringVolume2.WorldScientific,USA
Dunn,P.F.(2005),MeasurementandDataAnalysisforEngineeringandScience,McGrawHill.
ISBN0072825383
Goldstein,R.J(1983).(ed.)FluidMechanicsMeasurements,HemispherePublishingCompany,
Washington,USA
GuedesSoares,C.(1990),EffectofSpectralShapeUncertaintyintheShortTermWaveInduced
ShipResponses,AppliedOceanResearch,Vol12,pp.150159.
Hermundstad,O.A,Aarsnes,J.V.andMoan,T.(1995),HydroelasticAnalysisofaFlexible
CatamaranandComparisonwithExperiments,FAST95,Lybeck,Germany,VolI,pp487500
Huse,E.(1999).ExperimentalMethodsinMarineHydrodynamics,DepartmentofMarine
Hydrodynamic,NTNU.
Huse,E.(1974).Effectsonafterbodyformsandafterbodyfinsonthewakedistributionofsingle
screwships.,NSFIreportR31.74..
Huse,E.andMatsumoto,K.(1989).ViscousSurgeDampingofFloatingProductionVesselsMoored
atSea,8thOMAEConference,TheHague.
Huse,E.andTrum,A.(1981).NHLOceanLaboratoryDesignPhilosophy,Int.Symp.on.
HydrodynamicinOceanEngineering,Trondheim,Norway
ITTC(1996).ReportfromPropulsorCommittee,21theint.towingtankconference,Trondheim,
Norway
ITTC(1996).ReportfromManoeuvrabilityCommittee,21theint.towingtankconference,
Trondheim,Norway

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page131
ITTCSymbols&Terminology,RecommendedProceduresandQualitySystemsmanual:
http://ittc.sname.org/documents.htm
Lloyds,A.R.J.M.(1989),Seakeeping,shipbehaviourinroughweather,EllisHoreoodLimited.West
Sussex,England
Longo,J,Stern,F.(2005),UncertaintyAssessmentforTowingTankTestsWithExampleforSurface
CombatantDTMBModel5415,JournalofShipResearch,Vol.49,No.1,March2005,pp.5568
Maeda,H.(1991).Modellingtechniquesfordynamicofships,Phil.TransR.Soc.London,Vol334,
pp339355.
Naeser,H.(1981).NHLOceanBasinCapabilitiesandLimitations,Int.Symp.on.Hydrodynamic
inOceanEngineering,Trondheim,Norway
Newland,D.E.(1975).RandomVibrationsandSpectralAnalysis,Longman,London
OlsenO.A.(1992).Instrumenteringsteknikk,TapirTrondheim.
Sasajima,H.andTanaka,I.(1966),OntheEstimationofWakeofShips11thInternationalTowing
TankConferenceProceedings,Tokyo.
SNAME(1967).PrinciplesifNavalArchitecture,ThesocietyofNavalArchitectsandMarine
Engineers,NewYork.
Sarpkaya,TandIsaacson,M.(1981),MechanicsofWaveForcesonOffshoreStructures,Van
NostrandReinholdCompany.
Stansberg,C.T.,ritsland,O.andKleivan,G.,(2000),VERIDEEPReliableMethodsforLaboratory
VerificationofmooringandStationkeepinginDeepWater,OffshoreTechnologyConference,
paperOTC12087,Houston,Texas
Steen,S.,(2011),MotstandogPropulsjon.Propellogfoilteori,KompendiumUK201199i
TMR4247MarinTeknikk3.InstituttforMarinTeknikk.
Stoot,W.F.(1959)SomeAspectsofNavalArchitectureintheEighteenthCentury,Transactionof
InstitutionofNavalarchitects,101,31
Taylor,E.S.(1974).Dimensionalanalysisforengineers,ClarendonPress,Oxford
Walderhaug,H.(1983),EksperimentellMarinHydrodynamikk,DepartmentofMarine
Hydrodynamic,NTNU.
White,F.M.(2005)FluidMechanics,fifthedition,McGrawHill,ISBN0071243437
Zhao,R.Faltinsen,O.M.,Krokstad,J.andAanesland,V.(1988),WaveCurrentInteractionEffects
onLargeVolumeStructures,BOSS88,Trondheim,Norway.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

15

Page132

INDEX

6DoF,35
accelerometer,36
ADconverter,30,46
ADconverter,48
range,48
resolution,48
addedmass,79
Addedresistance,73
afterbodymodel,20
amplifier,30,46
frequencyrange,52
risetime,52
analog,30
atmosphericpressure,15
autocorrelationfunction,94
BendatandPiersol,54,94
Berg,70
biaserror,115,119
blockageeffect,57
bollardpull,109
Bouchner,80
boundarylayer
turbulent,68
bridgebalancing,46,56
bridgeexcitation,46
Britishmethod,65
BuckinghamsPitheorem,8
bulkcarrier,113
calibration,29,55
calibrationcurve,55
calibrationfactor,55
capacitancetransducer,34
carriage,74
catamaran,59
cavitation,15,68
pressurepulses,68
cavitationnumber,15,18,20
cavitationtests,63
cavitationtunnel,15,16,20
centreofgravity,57
CFD,115,120
Chakrabarti,8
chargeamplifier,46
Chauvenetscriterion,119

ColemanandSteele,115
commutativeprobability,100
confidenceinterval,115
continentalmethod,65
Crashstop,111
crossspectrum,95
crosstalk,44
crosscorrelationfunction,94
current,29
currentgenerator,29
DAconverter,48
dataacquisition,30,46
decaytest,87
digital,30
dimensionlessquantities,12
Dopplershift,42
Dunn,5
dynamicsimilarity,8,9,68
dynamometer,43
elasticfluidforces,9
elasticmodel,13,34,61
Backbonemodel,61
Hingedmodel,61
elasticmodels,9
environmentcalibration,82
equivalentdampingcoefficient,89
ergodicprocess,94
error,115
Excel,117
Farstad,113
FastFourierTransform,95
fatiguelife,113
FFT,95
fieldmeasurementtechnique,43
filter,30
filters,46
bandpass,46
highpass,46
lowpass,46
flapwavemaker,23
Floatingbridge,59
floatingplatforms,79
folding,51
Froudenumber,10,68,122

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014


Froudescaling,12,13,28,70,122
Froude,William,6
fullbridge,33
Gaussiandistribution,116
Gaussianprocess,103
GeneralAcoustics,45
geometricalsimilarity,8,14
geometry,60
globalloads,59
Globalloads,73
Goldstein,30
GPS,110
gravitationalforces,9
Greenwater,73
gyrationradius,57
gyro,35
Gyrocompass,110
halfbridge,33
HEMOS,113
Hermunstad,62
Highspeedcraft,111
homogeneousprocess,94
Hottinger,48
hullmonitoringsystem,113
hullvibrations,68
Huse,5
hydrodynamicdamping,79
hydroelastic,60
hydroelasticproblems,13
hydroelasticity,13
icetank,18
IMO,105
HighSpeedCraftcode,105
impactloads,39,52,80
inductivetransducer,33
inertiaforces,9,57
ISO,109
ITTC,42,71,109,115,119,122
JONSWAP,26
KeuleganCarpenternumber,12
kinematicsimilarity,8
kinematicviscosity,10
Kurtosis,103
LaserDopplerVelocimetry,42
LDV,40,42
seeding,42

Page133
Lengthofrecord,54
LeonardodaVinci,6
logarithmicdecrement,89
lowfrequencymotions,80
Machsnumber,10
Maeda,60
manoeuvring,69
manoeuvringtests,63
Marin,29
MarineCyberneticLaboratory,23
mass,57
distribution,57
MCLab,29
measurementbridge,32
metacentricheight,59
momentofinertia,57
monohull,59
mooringline,34
mooringlines,13
Mooringlines,59
mooringsystem,60
multidirectionalwaves,26
Newland,26
noise,68
normaldistribution,116
NPDspectrum,29
Numberofsamples,54
Nyquistfrequency,51
oceanbasin,16
oceanlaboratory,22
offshorestructures,79
Olsen,30
operationallimits,79
opticalsystem,34
ParticleImageVelocimetry,43
Peerlesspool,5
pendulumtest,58
phaseangle,91
piezoelectric,37
pitottube,40
PIV,43
seeding,43
PlanarmotionMechanism,70
PlanarMotionMechanism,70
PMM,70
positionmeasurement,34

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014


potentiometer,35
powerspectrum,95
Prandtlpitottube,40
precisionerror,115
precisionlimit,117
Pressurecell,38
pressureforces,9
Pressuremeasurements,37
PressureProfileSystems,39
pressuresensingfilm,39
pressuresensitivepaint,40
pressuretransducers,37
PriceandBishop,94
probabilitydistribution,100
propeller
diameter,68
speedofadvance,68
thrust,43
torque,43
propellerdynamometer,64
propulsiontest,63
PSP,40
RAO,79,90,95
Rayleighdistribution,100
regularwave,90
replicationlevel,116
resistancedynamometer,63
resistancemeasurement,63
ResponseAmplitudeOperator,95
Reynoldsnumber,10,68,122
riser,34
risers,13,59
RollsRoyce,113
Rotatingarmtests,70
rudder,69
samplingfrequency,46,48,50
SamuelFortey,6
sandstrip,64
SarpkayaandIsaacson,124
Sasajima,123
scaleeffects,122,125
scalingratios,12
seatrials,106
Seatrials,105
seakeeping,73
seakeepingtest,18

Page134
seismiccables,59
servoneedlewaveprobe,46
Shipmonitoringsystem,112
Shipmotions,73
signaltonoiseratio,49
Significantvalue,96
similarity
dynamic,8,9
geometrical,8
kinematic,8
Sixdegreesoffreedom,34
Skewness,103
slamming,39,73
Slowdriftmotion,80
slowdriftmotions,29,53
spectralanalysis,95
speedlog,113
speedtrials,106
springing,13,59
standarddeviation,116
Standarddeviation,96
Stansberg,80
stationaryprocess,93
Steen,S.,63,67,122
stochasticprocess,93
stoppingtest,110
straingauge,31,33
Strouhalnumber,12
Studentstdistribution,117
studs,64
surfaceforces,9
Swedenborg,Emanuel,6
Tanaka,123
tare,56
Tekscan,39
temperaturedrift,56
TensionLegPlatforms,59,80
testduration,53
towingcarriage,18
towingtank,16
towingtest,63
transducer,30,31
resonancefrequency,52
transducerdrift,56
transferfunction,25
TrialLeader,106

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014


trimpost,64
turbulencestimulator,64
turningcircle,110
UltraLab,45
ultrasoundwaveprobe,45
uncertainty,115
uncertaintyanalysis,115
vacuumtank,18
validation,130
vapourpressure,15
variance,96
velocitymeasurement,40
nonintrusive,43
ultrasonic,40
videobasedsystems,34
viscousforces,9
voluntaryspeedreduction,73
wakefraction,123
Walderhaug,5
wave
amplitude,91
frequency,91
waveabsorption,27

Page135
wavebeach,18,27
wavedriftforces,29
waveelevation,44
Wavefrequency,79
wavefrequencyphenomena,53
wavemaker,18,23
waveprobe,44
wavereflection,27
wavespectrum,25
Webernumber,68
Webersnumber,11
Weibulldistribution,102
Wheatstonebridge,31,33
whipping,13,59,73
White,8
Windgeneration,28
Youngsmodulus,14
zerocrossingperiod,97
zerolevel,56
Zeroing,56
Zhao,127
zigzag,110

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page136

ANNEXA
EXAMPLEOFREPORTINGFROMMODELTEST
Thecontentofamodeltestreportwilldependonthetypeoftesting,theextentofthetestsand
requirementstoreportingformtheclient.Atypicallayoutofthemodeltestreportmaybeas
follows:

Summary

Summaryofthetestresultsandthemainconclusionfromthemodeltest.

Introduction

Generalinfoabouttheproject

Objectivesofthetest

DescriptionofTestSetUp

Modelscale

TestFacilities

Testarrangement

Modeldescription;maindimensions,loadingconditionsetc.

Modelcalibrationandverification

MeasurementsandDataAcquisition

Descriptionofusedinstrumentation(typeoftransducers).Positionoftransducersonmodelor
inbasinetc.

Notionandsignconventions

Coordinatesystemappliedforthemeasurements

Dataacquisition,filtering

Measuringaccuracy

Controlandcheckroutines

EnvironmentalConditions(onlyrelevantfortestinginwaves)

Wave,windandcurrentdatausedinthetests(whenrelevant)

Calibrationofenvironmentalconditions

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page137

TestProgramme

Detaileddescriptionofthetestprogramforeachtypeoftest(statictests,decaytests,regular
wavetestsorirregularwavetests)withnumberingoftests.Allrelevantinputparameterstobe
specified(i.ewave,windandcurrentcondition,speed,loadingconditionetc.)

DataAnalysis

Scalingofresultsfrommodeltofullscale

Filteringappliedduringthemeasurementsandanalysis

Analysiscarriedoutandmethoddescriptionfordifferenttypeoftests(statictest,decaytests,
wavetestsetc.)

PresentationanddiscussionofResults

Presentationofmainresults(DetailedresultsusuallyinAppendixes)

Interpretationonevaluationofresultsformainparameters.

Evaluationofscaleeffects

APPENDIXES
Willtypicallyincludemoredetailedinformationfromtheteststhanwhatisgiveninthemain
report.Exampleswillbe:

Detaileddescriptionofcalibrationproceduresandresults

Wavecalibrationtests

Descriptionofanalysismethods

MeasuredTimeseries

Detailedresultsfromanalysis(filtering,spectralandstatisticalanalysis)

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page138

ANNEXB
EXAMPLEOFMODELTESTSPECIFICATION

Introduction
Modeltestswithaturretmooredtankershallbecarriedout.ThetankermodelisastandardNorth
Seashuttletanker.
Thetestsshallcoverenvironmentalconditionscorrespondingtotypicaloperationalconditionsas
wellasdesignconditionbasedonthe100yearsreturnperiodstorm.
ThemodeltestwillbeperformedatMarintek,Trondheim.

Objectives
Themainpurposeofthemodeltestswillbeverificationofthepresentdesign.Themotionsand
seakeepingperformanceofthetankeraswellasthemooringloadsaredeterminedforboth
operatingandextremeenvironmentalconditions.Theresultswillalsoserveasinputtocalibration
andverificationofthetheoreticalandnumericaldesignanalyses.

The objectives of the model tests with the moored tanker can be summarized as follows:

Toverifynumericalpredictionsofsystemglobalresponses,maximumoffsetsandmooring
linetensions;

Toverifynumericalpredictionsofthevertical,horizontalandmomentloadstransferred
fromtheTurretsystemtothetanker.

Tomeasureandestimatethegreenwatereffectsandpossiblebottomslammingloads

Toestablishthemaximumrollandpitchmotions

Toquantifythedampingofthelowfrequencymotion;

Forcesinthemooringsystem

Toidentifythemostcriticalcombinationofmetoceanloadsandvalidatethedesignofthe
mooringsystem;

Toinvestigateanypotentialinterferenceproblemsbetweentankermooringsystemand
flexibleriser;

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page139

Modeldescription
Themodelsshallbeconstructedtoalinearscaleof1:50.
Thewaterdepthisxxxm.
Thetotaltestsetupincludethefollowingcomponents:

Tanker

Turret

MooringSystem

RiserSystem

TankerDescription
Maindataforthetanker:
Parameter
LengthbetweenPerpendiculars,LPP
Lengthoverall,LOA
DraftatF,dFP
DraftatAP,dAP
BreadthMoulded,B
TransverseWindArea
LongitudinalWindArea
Freeboard,maindeckfMD
FreeboardatBow,fBOW
FreeboardatTrunkdeck,fTD
Displacement
RollRadiusofGyrationrxx
PitchRadiusofgyrationryy
YawRadiusofgyrationrzz
Transversemetacentricheight,GMt
Table1
MainParticularsofTanker.

Loaded

Ballast

Boththeloadedandballastconditionshallbeused(seealsoTable1):
The tanker model shall be constructed in accordance with the Yard drawings. The main dimensions,
mass and inertia properties of the tanker shall be accurately modeled. The tanker model shall match
the prototypes characteristics including the superstructure.
Themodelwillbefittedwithequipmentformeasuringbendingmomentsandshearforces,attwo
stationsofthetanker,atmidshippositionandaftoftheTurret.Bendingmomentsshallbe
measuredrelativetothepitch,rollandyawaxesofeachstationandshearforcemeasurementsshall
includebothhorizontalandverticalcomponents.Theinstrumentationshallbedesignedsoasnotto
distortthemodelsglobalresponses.
LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page140
TurretModel
The turret model will be constructed to scale in accordance with the turret design and the line
attachment positions will be according to design drawings. The position of the turret will be at
position xxx* Lpp from FP.
The loads shall be measured in 5 degrees of freedom; forces in x-, y- and z-direction and moment
about x- and y-axis. Hence the turret will be free to rotate about the vertical z-axis.
MooringSystem
Themooringsystemconsistof8evenlydistributedmooringlines(i.e.45betweeneachline).
Mooringline1shallbedirectedtowardslargewavemaker.
Eachoftheeightmooringlineshasthefollowingcomposition:
Segmenttype

Length

(fromanchor)

Chain
WireRope
Chain
WireRope

Nominal
diameter
Mm

Axialstiffness
EA
kN

Weightinair
kN/m

Submerged
weight
kN/m

Linepretension

xxxkN(fairleadatdepth15m)

Horisontaldistancebetweenanchorandfairlead

xxxxm

Waterdepth

xxxm

All mooring lines shall be physically modeled in terms of mass, submerged weight and total leg
elasticity.
RiserSystem
TherisersystemconsistsofoneoffxxflexibleriserinSteepWaveconfiguration
Thefollowingparametersdescribetheconfiguration:

Horizontaldistancebetweentopandbottomendconnection :

xxm

Lengthofbarerisersectionbelowbuoyancysection

xxm

Lengthofbuoyancysection

Lengthofbarerisersectionabovebuoyancysection

xxm

Netsubmergedweightofbaresection

xxxkN/m

Netbuoyancyofbuoyancysection

xxxkN/m

Outerdiameterofbareriser

xxxm

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

xxm

Page141
The riser configuration will be modeled geometrical correct from the turret table to the sea
bottom with the above weight distribution.

TestProgram
StillWaterTests
Thefollowingstillwatertestsshallbeperformed:
(a) Staticoffsetteststoverifystiffnesscharacteristicsofthemooringsystem.
(b) Extinctionteststomeasurethelowfrequencysurgedampingandnaturalperiod.
(c) Pitch,heaveandrollnaturalperiodsforallloadingconditions.
(d) Currentloadtesttoverifyloadsappliedbyspecifiedcurrentvelocity.
(e) Windloadteststesttoverifyloadsappliedbyspecifiedwindvelocity.
IrregularWaveTest
In Table 2 tentative test program for the irregular wave are specified. The following notations are
used:
HS :

Significant wave height

TP :

Spectral Peak period

Wave Spectrum Peaked-ness parameter

WAVE : Wave heading relative current


UW :

Wind speed, 1 hour mean, at 10 m above sea level

Wind : Wind heading relative current


UC :

Current speed

Thewindandwaveheadinggiveninthetablerepresenttheheadingrelativetocurrentdirection.
(fixedinthelaboratory).ThewindspectrumistobetakenastheNPDspectrum.
Intact mooring system is assumed except for one test condition where the most loaded line shall be
broken.
For offloading conditions, sufficient combinations of directions of wind, current and wave direction
shall be investigated to ensure the critical cases are found.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page142
Test
No

Waves
HS

Table2

TP

Wind

WAVE

WAVE

UW

Testdefinition,irregularwavetests.

Current
UC

Loading
Condition

Instrumentationanddataacquisition
Instrumentation
Theinstrumentationandmeasurementsduringthetestsshallasaminimuminclude:

Waveelevationat.

Currentvelocityprofile

Referencewindvelocity

Relativewaveelevationat...

Deckwetness(greenwater)atthe..

Sixdegreeoffreedommotionsoftanker.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page143

Verticalaccelerationat.

Globalbendingmomentsandshearforcesintwotransversecuts

Tensioninmooringlinesandriser.

Dataacquisition
DatashallbesampledatminimummodelscalerateofxxxHzfortheirregularwavetests.Forthe
measurementofpressureandimpactforcesasamplingrateofxxxHzshallbeappliedfortests
usingthedesignseastates.
Thedatarecordingdurationinirregularwavesshallbelongenoughtoachievesufficientlysmall
bandwidthinthespectralanalysis.Thetestdurationshallnotbelessthan3hoursprototypetime.
Asufficientlapseoftimeshallbeallowedbetweenteststoavoidanydistortionofthenew
generatedenvironmentbytheprevioustest.

Analysis
Allcalculationsofspectralanalysisandmooringlinetensiontransferfunctionsshallbedocumented
withbandwidth,recordlength,samplingestimates.Alldataanalysisshallbeprovidedinboth
tabularandplotformat.Statisticalandspectralanalysisshallbecarriedoutonallchannels.
Statisticalanalysisshallincludemean,standarddeviation,zerocrossingperiods,andminimumand
maximumvalues.Allresultsshallbepresentedinprototypescale.

Requiredtestaccuracy
1) Mooringsystem
Extinctionteststocalibratestiffnessandnaturalperiodsofthemooredsystem.Themeasured
stiffnessandnaturalperiodsshallbewithinx%.
2) CurrentandWindGeneration
The current velocity profile over the top x % of the water depth shall be modeled in still water.
Wind simulation will be performed by fans. The speed of fans shall be adjusted so that the mean
value of analytical or measured data obtained from wind tunnel test is matched for the various
tested wind headings.
3) WaveGeneration
Acceptable tolerances on the values of significant wave height (Hs), and spectral period (Tp),
shall be x %. Wave envelope spectral shall be provided to verify the absence of unrealistic
wave reflection in the basin.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page144

TestDocumentation
Thefinalreportshallcontain,asaminimum:

Detaileddescriptionoftestsetupsandprogram;

Descriptionofmodels;

Descriptionofwind,waves,andcurrentmodeling;

Resultsofstaticdeflection,extinctionandcurrenttests;

Zerocrossingstatisticsofallchannels;

Responseamplitudeoperatorplotsofallrelevantchannels(includingphaseangles);

Weibullextremevalueplotsofallchannels;

Mean,standarddeviation,minimumandmaximumvaluesofallchannels;

Selectedtimehistoriesofrepresentativeportionsofeachtest;

Firstandsecondorderwavesspectracomparingmeasuredandtheoretical;

Highandlowfrequencynumericalfilteringofmainloadsandmotions.

StillPhotographyandVideo
Stillphotographsshallbeprovidedtankermodel,themooringlegsandforcetransducers,thetank
setupsincludingwaveprobes,andthewindandcurrentmeasurementpositions.
Videorecordingofallwavetestsbythefollowingcameras:

Abovewatercamera(roofandside)

Twounderwatercameratoobservethemooringlinesandriserresponse

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page145

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Page146

ANNEXC
VISCOUSSURGEDAMPINGOFFLOATINGPRODUCTION
VESSELMOOREDATSEA

Huse,E.andMatsumoto,K.8thOMAEConference,1989,TheHague.

LecturenotesinExperimentalMethodsinMarineHydrodynamics,issuedAugust2014

Annex C

Annex D


      
 !"# %$ '&%()) "*+)

,.-0/!132547698
:<;>=?; 2A@CB"DFE0G9D?DIH

J KML!NFOAPNRQ>LSQ>OANRQ7OTPUN
XV WZYS[]\_^`bac[]deY5fgfh`bf.[]di[bjlknmg\_m7\omqpg`srt\lutYv[xwzyi[bdtph\{p|[}ph\lutY^~Y5[bmgyZfhY`]aRWi`}fhYAj_\o[]Zj_Yv[x^~Y3[bmgyZfgY3`bf.5[]joTyZjo[}phY5u}[bjlyZY
\omAxv`bfh^[]j_jlkphWZ\_mS\omSZfgY3mY5dzpgY5C[bmS[<A`bdn9nYAdiAY\_dzpgY5fgu}[]jYb r<ztsVXWZ\omS^~Y5[]dImphWi[}ppgWiYifg`ti[]Z\_j_\{pkpgWI[}pSpgWZY
phfgyZYu}[]j_yZY\omSU\{phWZ\_dpgWiYrt\lutYAd\_dzpgYAfhu}[]jF\_mi znSdpgWZYa`tjlj_`}U\ldirpgWZY`bf|YAfhfg`tfvU\_jlj>9YyImY3pg`nY5mhTfh\_IYspgWZY
n\omph[]dITYx9YTpYAY5d[^~Y5[tmyZfhY5`tf5[]joTyZjo[}phY5u}[]j_yZY[bdipgWiYxpgfhyZYt9iynpyZdizdi`}UdRIu}[bjlyZYtvVXWZY.`tfhyZdiAYAfgph[b\ldzpk
^[ke 9Y!yimgY5a`bfXpgWiYxmp|[}pg\ompg\oA[]jcY5mpg\_^~[]pgY`]aFpgWiYxmg[b^YSphWZ\ldiri
j_j7[]di[bjlknmg\_mWZYAfhYxfhYTaYAf|mUph`<mg`A[bjlj_Y5ph\l^~YU\_mgYsYA"9YAfh\_^Y5dzphm5I\ YtUpgWiYs^~Y5[tmyZfhY5wzyi[bdzpg\lpk<\omv^~Y3[bmgyZfgY3[}pS[
mg\ldirbj_Y0jl`n5[}pg\_`bd![]pn\l9Y5fgY5dtpcpg\_^Y3mA>VXWZYq[bj{phYAfhdi[}ph\lutYbmg[b^~ZjlYAph`]mg[b^~ZjlY>YA"9YAfh\_^Y5dzphm53\_d!UWZ\o|WpgWZYwzyi[]dzpg\lpkSu[bfg\_Y5m
\_dCmgi[bAYsZyZpdi`]p\ldpg\_^Y~^[k\ldC^[]fh\ldZY~[]iZjl\oA[]pg\_`bdim9YsphWZY~\ldzpgY5fgY3mph\ldirea`bf![bZZj_\_5[}pg\_`bdIm[tmphWZYAfh^~`"ZkzdI[]^~\_
Zfh`b9YAfgpg\_Y5mS`ba.X[}pgY5f![]diC^[}pgY5fg\o[]j7ifg`tIY5fph\lY3m`ba.^[}phYAfh\_[bj~Z`tfSpgWZY~ph`}U\ldZr<ph[bdZYA"9YAfh\_^Y5dzphm!T`tdimg\_nY5fgY3WZY5fgY
mgyi|Wu[bfg\o[}ph\l`tdimX[]fhYdZ`]pvA`bdimg\_ZYAfhY5
7g )>9SvZ"q7vn7n7RZZ0
VX`xpk"9Y5mX`]aYAfhfg`tfhm.[]fhYSA`bdimg\_ZYAfhY5\_dphWZY[]di[bjlknmg\_m5XAx[bdi~?g|To}YAfhfg`tfhm5VXWZY!n\l?YAfhYAdiAY\om\lj_jlyImphfh[]pgY5\ld
irtyZfhYp|[]tYAdafg`t^.`tjlY5^~[bd[]dizphYAYAj_Y5bttT0VXWZY!irtyZfgY![bj_mg`\lj_j_yimphfh[]pgYphWZY]]>YAfhfg`tfX[tmXpgWiYxmyZ^`]a>Z\o[bm
[bdieZfhY5A\_mg\_`bdeY5fgfh`bf|mA ^~[b\ld[bmhmyi^Zpg\_`bd\_dephWZ\omU[]di[bjlknmg\_m.\_m.phWi[}pU\la[]d<\ldnIdZ\{phYdzyi^s9YAfX`]ac^~Y3[bmgyZfgY5^~YAdzphmYAfhY
p|[]bY5dphWZYxZfhY5A\_mg\_`bdYAfhfg`tf`byij_<a`bj_j_`}
[e[]yImgmg\_[bdZ\_mpgfh\liynpg\_`bdR.VXWZYxZ\o[bmXY5fgfh`bfX`byZjo<pgWZY5dIYphWZYsn\l9Y5fgY5diTY
9YTpYAY5d~pgWZY^~Y3[]d~`]a?pgWi\_m.n\_mpgfh\_Zynpg\_`bd[bdipgWZYpgfhyZYu}[]j_yZYb>VXWZYUphfg`tyZZj_YUU\{phW~pgWZY3mYnYTidi\{ph\l`tdim0\om7`]aRA`byZf|mYphWi[}p
phWZYpgfhyZYu}[]j_yZY!\_mXyidZ"dZ`}UdRimg`\_dZf|[bph\_AY![[bj{phYAfhdi[}ph\lutY!nYTidi\{ph\l`td\omXdZY5Y5nY3
V`n`~pgWi\_mpgWZYsA`bdiAYAnp`]a>fgY5Zjl\oA[]pg\_`bdj_YAubY5jc\omU\_dtphfg`nnyITY5R.WZY5dfgY5IY3[}ph\ldZre[bdYTn9YAfh\l^~YAdzpUphWZYxfhYAZj_\_5[}ph\l`td
j_YAutYAj0nY3mgAfg\_IY3mUWi[}p!YAn[tphjlk\omfhYA9Y5[}phY5cWZYAfhY~\ldpgfhYAY~j_YAutYAjomA p!pgWiY5YAfh`]phWC`tfhnY5ffgY5Zjl\oA[]pg\_`bdCpgWZY^~Y3[bmgyZfgY3
wzyi[bdtph\{pk\_m.[bmhmyi^Y3[]img`bj_ynphYAj_kT`tdimph[]dzp0\_d~pg\_^~Ybnm`phWi[}p.`bdZj_ku}[bfg\o[}ph\l`tdim>\ldiWZYAfhYAdzpq\_d~pgWZY^~Y5[bmgyZfhYA^~YAdzpqmgknmphYA^
\lp>mgYAj_ubY3mA`bdzpgfh\_ZynpgY3mRph`vphWZYYAfhfg`tf5ci`bfFYTZ[]^~Zj_Yb]5[]j_\lifh[]pg\_`bdx`baZjl`z[bsAYAj_j_m9YTa`bfhY^`tyZdzpg\_dZrv\lp7\ldspgWZYYTn9YAfh\l^~Y5dtp|[]j
mgYTphyZmgWZ`tyZj_rb\_ubYAYAfh`]phW`tfhnY5ffhYAijl\oA[]pg\_`bdRxdpgWZ\om!A[tmYpgWiYphfgyiYu}[]j_yZY\_m"dZ`}Udc?mg`Z\_[tm[]diZfgY3T\om\_`bdA[bd
9YY3mph\l^[]pgY5e\ldpgWZYdZ`bfh^[]j?mY5dimYt pif|mpg`tfhZYAf0fhYAZj_\oA[}ph\l`tdej_YAubY5jZphWZY\_dimpgfhyZ^~YAdzph[]pg\_`bd[bdiemgYTphyZ\om0i"Y3"Zynp
phWZY~pg\_^~Y\_mfhyZdZdZ\_dZrI~VXWi\_m!A`bfhfgY3m9`bdiimvph`fgY5IY3[}pg\_dZrpgWiYYTnIY5fg\_^~YAdzp!bY5YAZ\_dZr<pgWZY\_dimpgfhyZ^~YAdzph[]pg\_`bdT`bdImp|[]dzp5
UY5IY3[}phY5CfgyZdIm\_d)[eph`}ph[bdZ\om![rt`"`"YTZ[]^~Zj_Yb~x pgWj_YAubY5j7fgY5Zj_\_5[}pg\_`bdC\ldITj_yinY5m![bjlj>`bpgWZY5fS9`tmhmg\lZj_Y|Wi[]dirbY5m
UWZY5d~fgY5IY3[}pg\_dZrphWZYUYTn9YAfh\l^~YAdzp3]a`bfh^|WI[]dZrt\ldZr[ZfhY5mhmyifgYUTY5jlj]ph`!IY5fa`tfg^~\_dZr![!m\_^~\ljo[]f7YAnIY5fg\_^~YAdzp0\lde[bdZ`]phWZYAf
jo[]9`bf|[}ph`bfhkbzU\{phWZ\{?YAfhYAdzp9YAf|m`tdZdZYAj
"\_diTYSphWZYZ\_[tmYAfhfg`tf.\_dephWZYAYAfh`]phWj_YAutYAj?fgY5Zj_\_5[}pg\_`bd<yimgyi[]j_jlke5[]d<IY![b5T`tyZdzpgY5a`bfX\ld<pgWiY5[]j_\lifh[]pg\_`bde^~`nnYAj
[^~`bfhYXZf|[bph\_5[]jZn\ompg\_diph\l`tdIYAp.Y5YAd~ZfhY5T\omg\l`td[bdiZ\_[tm>YAfhfg`tf7\_mq[bmFa`bj_j_`}mAFVXWZYZfhY5A\_mg\l`tdY5fgfh`bf>\om7rb\_ubY5d"kspgWZY
u}[]fh\o[}pg\_`bd`]a0if|mpS`tfhnY5ffhYAijl\oA[]pg\_`bdUWZ\_j_YZ\o[bmSY5fgfh`bf\_mS\_diTj_yinY3C[]ps phWC`tfhnY5ffhYAijl\oA[]pg\_`bdR!VXWZ\omS\l^~^~Y5n\o[}phYAj_k
fhYAutY5[]jpgWiY^[b\ldZfh`bZj_YA^`baqYAu}[]j_yi[]pg\_dZrphWZY~Z\o[bmYAfhfh`bf3I\lpsA[bd\_dCZf|[bTpg\oTYdZ`]p9Y^~Y3[bmgyZfgY3.mph\l^[]pgY5m^syImp
9Yi[bmgY5C`bdYTn9YAfh\lY5diTYe[bdiC^`tfgY~`tf!jlY3mgmY3nyiA[]pgY3CrtyZY5mhmgY5m5dpgWiY`]pgWiYAf!Wi[bdiphWZYeZfhY5A\_mg\_`bdY5fgfh`bf!5[]dIY
Y3mph\l^[}phY5<pg`[fhY5[bmg`bdI[]Zj_YSZYArbfhYAYS"kfhYA9Y5[]pg\_dZrpgWZY!YA"9YAfh\_^Y5dzp5
7 X>S!0FZq7v?.USvcv?gqnRZF7
apgWZY^~Y5[tmyZfhYA^~Y5dtpX\omXfgY5IY3[}pgY3[bd<\_dnidZ\lpgYd"yZ^IY5fU`]acph\l^~Y3mA"phWZY^~[b\ld[bmhmgyZ^~npg\_`bd<WZYAfhY\_m.pgWI[}pXpgWZY!^~Y3[bmgyZfgY3
u}[]j_yZYxU\_j_j9YxdZ`tfg^[]ji`tf[]yImgmg\_[bdRIZ\_mpgfh\liynpgY3[]fh`byZdI[e^Y3[]dRVXWi\_mUZ]gTI}oAn}\_m|Wi[]f|[bTpgY5fg\_AY3"k
phWZY!^~Y5[]d[]dipgWZYxmph[bdiZ[bfhnYAu"\o[}ph\l`td>i[bdiYTnZfhY5mhmY3e"k
C0 c } Y}b T l
3
WZY5fgY Cn\ompgWiY!Zfg`ti[]i\lj_\{pk~phWi[}pv[~mg\ldirbj_YS^~Y3[bmgyZfgY5^~YAdzpU'U\_jljcjl\_YIYAp.Y5YAd[bdi<Sn>VXWZY!^~Y5[tmyZfhYT
^~YAdzpX\_m`baT`tyZfhmgYvdiYAubY5f.fhYA9Y5[]pgY3e[]d<\ldnIdZ\{phYdzyi^s9YAfX`]aph\l^~Y5m50VXWZYmp|[}ph\_mpg\ovifg`tIY5fph\lY3mq`bac\ldzphYAfhY5mp\om0phWZYAfhYTa`tfgY

F\_rbyZfhYtg}| Z~|ngTI!!
}}bAl  > blsg|b] .~qg|b} }. 0I I~|A
 g|b}  }  Rg! #"0}l })}9%
$I|Tl' &()+*,)
i[tmY3`bd<[A!v9lXZ|9nl]]pgWi[]pXA`bdimg\omp`]ac[!IdZ\{phYd"yZ^IY5f.`bamh[]^~Zj_Y5mqZfh[Udafh`b^phWZYI[]fhYAdzp.n\omphfg\_Zynph\l`tdR
i`bfXpgWiYxmg[b^ijlY9`bZyZjo[}ph\l`tdphWZY!^~Y5[bd\omXrb\_ubY5d<"k
- / . 0
b
02143
UWZY5fg5Y -\_mFd"yZ^s9YAf7`baImh[]^~Zj_Y5mU^~Y3[bmgyZfgY5^~YAdzphm|R 0 7VXWZYZfgY3T\om\_`bd\ldInYT]`tf>mph[bdiZ[bfhnYAu"\o[}pg\_`bdc}`]aiphWZYUmg[b^~ZjlY
9`bZyij_[]pg\_`bd\_mZYTidZY3zk
.
687 ; 9:: / 0 < >
?t
-=<
02143
`bfh"\ldZrU\lpgW[mh[]^~Zj_Yx9`bZyij_[]pg\_`bdRnphWZYs^~Y3[]d\ldY3wzyi[}ph\l`tdbXU\_j_jnY5IY5di`bdphWZYsmgYTpv`ba7mh[]^~Zj_Y5mUph[]tYAdafh`b^
phWZY~i[]fhYAdzp!I`tZyZjo[}ph\l`tdRVXW"yimSphWZY~^~Y5[]d\lpsmY5jlutY5mS\_mdi`bfh^~[bjlj_kn\omphfg\_ZynphY5U\{phW^Y3[]dC^Y3[]d`]a.pgWZY~i[bfgY5dtp
n\ompgfh\lZyZpg\_`bdIX[bdi<mph[bdiZ[bfhnYAu"\o[}ph\l`td @fgY5k"mgA\_r5btt
6 7 6A7
C Bz
Z`tfXpgWZY!I[]fhYAdzpn\_mpgfh\_Zynpg\_`bdphWZYxT`tdnInY5diTY\_dtphYAfhu}[]j`]a7[~mg[b^~ZjlY\omUrb\_ubY5d"k
D E!FG} 0 <IHK
JI J 0 L HR0N M
b
UWZY5fgOY HP%+ Q + Ra`bf[di`bfh^~[bjcn\omphfg\_Zynph\l`td[]diT`tdnInY5diTY!j_YAutYAAj MS Q znn\ YphWZYszt A`bdnIZYAdiAY!\ldzpgY5fgu}[bjqi`bf
phWZYmh[]^~Zj_Yv9`bZyij_[]pg\_`bdRzpgWiYmph[]dIZ[]f|nYAu"\o[}ph\l`tde\_m.yZdZ"dZ`}Ud~phW"yim.pgWZYu[bjlyiYv`]8a H7\om.dZ`]prb\_ubYAdc>.wzyi[}ph\l`tdbqA[bd

9Y!fgY5Ufg\lpphYAd

0 < J HNM
Rt
6 7
UWZY5fgYspgWZYu}[]fh\_[bZj_Y
  \_mfh[bdin`t^[bdia`tjlj_`}m[zpgyInYAdzp5 m Hvn\omphfg\_Zynph\l`tdRU\lpgWK- <nY5rbfhYAY5mv`]a0afhYAY3n`b^
VXWZY!u}[bjlyZY`ba H.^~[kIYa`tyZdi<afg`t^ C@fhYAknmgA\_r<3btt

H0 3 . MR 
]
UWZY5fgY 3  \_mFphWZYU\_dzutYAf|mY.`]a9pgWZYUAyZ^syZjo[}ph\lutYUnYAdimg\lpk!ayZdiTpg\_`bda`bf>pgWiY Hn\omphfg\_Zynph\l`tdRb[brt[]\_dsU\lpgW%- <UnY5rbfhYAY5m
`baFafhYAY3n`b^XF\lrtyZfgYsrt\lutY5mUu}[bjlyZY3m`ba Ha`bfStz []dItzMT`bdZInYAdITYx\_dzpgYAfhu}[]jomAXZ`tf - ?tpgWZYxu}[bjlyZYOHX
\_m
[tATY5nph[bZj_Y~dzpgYAfhdi[]pg\_`bdi[bj?V`}U\_dZrV[bdZ.`bdZaYAfhYAdiAY3btt>VXWZY!ZfhY5A\_mg\_`bdjl\_^~\{pUa`tf[mh[]^~Zj_Y\_mUpgWZY5d
D 7 H 6A7
t
[bdiea`tfXpgWZY!^~Y3[]d`]a -'fhYA9YTph\{ph\l`tdim
6A7
D7  H
t
D E!FGH J

3.5

12

/ .0 2.5

30
10!"$#%&15'(*)+%, 1 20. 3 25
F\_rbyZfhYn54 Z qA n8H4T]ST } |}.tT?hS IA }}  76 )98;: |} bA9|=<?>?>%)+)!:
CBbos|}hi}9}x')98DM: |}.tT?h T}se9]~]F}oTn} E (+ )GF
0

|} bA9|A@Xg] 9}

7 IH JLKv7UqcvM?inRiqON|0ZQPzc9ZgSRqUZRi
TS[}p|[<mh[]^~Zj_Y5mvpgWi[]p[bZ9Y5[]f|mSm\_rbdZ\lI5[]dzpgj_k`bynpS`]aqj_\ldiYU\lpgWCpgWZY^[VU`tfg\lpk`]aqmh[]^~Zj_Y5m[]fhYsA[bjlj_Y5 7 lI`t\ldzphm5?`bf
]nA|>aIphWZYfhY5[bmg`bda`bf.myi|Wn\omhTfhYAi[bdiTkx\om7`tzu"\_`byim7[bdiubY5fg\lI[]ijlYtbmyI|W9`b\_dzphm05[]d~9Yvn\omfhYArz[]f|nY5cdeA[bmgY5m
UWZY5fgYpgWiYfhY5[tm`td\_mWI[]f|nYAfpg`YTnZjo[]\_dRR[]d[b5TYAZph[]ijlY~Tfh\{phYAfh\l`tda`bffhY UY3pg\_`bdC`]a.[mh[]^~Zj_Y\ompgWZY.Wi[byZubY5dZYTp3 m
Afg\lpgY5fg3 \_`bdR>VXWZ\omUmp|[}phY5m.pgWi[]pmg[b^ijlY3mXU\{phWZ\ld[i[]di[]fh`byidiphWZY^Y3[]d<U\{phWifg`ti[]Z\_j_\{pk~`]aFYTZTY5Y5nY5diTYSj_Y5mhmqphWi[]d
< > . mWZ`tyZjo<9Y!fhYTph[b\ldiY57VXWZ\omUj_\l^~\lpvA[]d9Y!YTnZfhY5mhmY3<[tmU[YA\_rbWzpU`bd<pgWiYxmp|[]diZ[bfhnY5uz\o[}ph\l`tdR
HXWY%Z=[%\=]^Q]_0 3 .a`?
5t

UWZY5fgY `< < > 3. \om.pgWZYSjl\_^~\{pUZfh`bI[]Z\_jl\lpk~i[bmgY5e`bd -'mg[b^~ZjlY3mA7v`]pgYpgWI[}pX\ldpgWZ\omXA[bmgbY X  .\om.pgWZYTyZ^yZjo[}pg\_ubY


nY5dimg\{pkayidiph\l`tdC`baqpgWiYdi`bfh^~[bj7n\_mpgfh\_Zynpg\_`bdcc[]jom`a`bfxmg^[]j_jFu}[bjlyZY3m`]a -d c0[bjlyZY3mS`]a HXWYeZf[e\=]^9]_[bfgYZj_`]pphY5C\ld
irtyZfh Y ?i
n[b^ijlY3mUU\{phWWZ\lrtWZYAfZYAu"\_[]pg\_`bd<afg`t^ pgWiY!^Y3[]d<pgWi[bdR
gg 0 < gg  H WYeZf[e\=]^9]_ 6 7
b3
^[kpgWZY5d9YSZ\_mgfgY5rt[bfhnY3t[bdidZY5 ^~Y5[bd[]di<mp|[]dii[]f|nYAu"\o[}pg\_`bd<A[]dIYa`byZdIei[tmY3`bdphWZYSfgY5^~[b\ldi\ldZrZ[}p|[Z
VXWZY!ifg`nTY3nyZfhYS5[]d`bdZj_k9Y!9YAfga`bfh^Y3`tdiTYt
?

4.5
4




3.5



3 

12

/ .0 -

3
2.5
2
1.5
1
0.5
0
10

1
10

()e!"10)
I+.
"   

3
10

4
10

F\_rbyZfhYO?Z 4 Z q T  " H WYeZf[e\=]^Q]_ Z} }A9A A A}


T}sgA|A]s}nA|
7  )>RZZ0ctZFZq
VXWZYsZfgY3T\om\_`bdj_\l^~\lphm\ldY3wzyi[}ph\l`tdt[]diY3wtyI[}pg\_`bd)tU\om`tdZj_ku}[]j_\_a`tfphWZYsZfhY5A\_mg\_`bdi[]fgp`]a>phWZYsY5fgfh`bf3nUWZ\_jlY
phWZYsZ\o[bmSA[bdZdZ`bpvIYspgfhY5[]pgY5pgWZ\omvX[kbVXWZ\_mva`bj_j_`}mUafh`b^nYTidi\{ph\l`tdR9pgWZYZ\_[tmvY5fgfh`bfn`"Y5mdi`]pS|WI[]dZrtYxUWZYAdpgWZY
^~Y5[tmyifgY5^Y5dzp\om!fhYA9Y5[]pgY5R`bpgY~pgWI[}p!pgWiYeWZ\_rbWZY5fpgWZYefhYAijl\oA[]pg\_`bdjlY5ubYAjq`]aXfgY5IY3[}pgY3^~Y3[bmgyZfgY5^~YAdzphm5RphWZY~aYAYAf
yZdizdi`}UdsZ\o[bmYAfhfg`tfhmRpgWZY5fgYX[]fhYq\_dxphWZYXfgY3myij{p|mAFVXWZ\_mF^~Y5[bdimpgWi[]pFY5fgfh`bf|mphWi[}p7^yimp>9Y.Y3mph\l^[}phY5\{aipgWZYXYTn9YAfh\l^~Y5dtp
\omXIY5fa`tfg^~Y3`tdZjlk<`bdiAYbn^~\_rbWzpUIY!\_diTj_yinY3\ld[mph[}ph\_mpg\oA[bjR[]di[bjlknmg\_m\laphWZY!YTn9YAfh\l^~Y5dtpU\omUfhYA9Y5[}phY5
@dZ`}Ud[]di<.Y5jljRZYTidZY3Z\_[tmYAfhfg`tfhm5tj_\_bY!T`tdimp|[]dzppgYAfh^mX\ld[~A[bjl\_Zf|[}pg\_`bd^`nnY5jI[]fhYyimgyi[bjlj_keT`tfgfhY5TpgY5ea`bf
[bdidi`]p!\ldITj_yinY5\_dphWZYYAfhfh`bf[]di[bjlknmg\_m5VXWiYfhY5mp5Ra`bf!YTZ[b^ijlYpgWZYyidiTY5fp|[]\_dtpk`baqphWZYeT`tdimp|[]dzppgY5fg^\_dCpgWZY
5[]j_\lZf|[}ph\l`td^~`nnYAj^syimpIY~Y5mpg\_^[}phY5\_dCmg`b^~Y[kt!VXWZY[bZZfh`t[t|W\_mph`<Y3mp|[]Zj_\_mgWC[Z\o[bmSYAfhfg`tfvj_\l^~\lp MphWi[}p
U\lpgW[rt\lutYAdT`bdZInYAdITYfgY5ZfgY3mY5dzphmpgWZY~yZdizdi`}UdZ\o[bm5xd`bf|nYAfSpg`Y5mph[bZj_\_mgWphWZ\_mj_\_^\lpsmg\l^~Zj_\{IY5CY5mpg\_^~[]pgY3mA
YA"9YAfh\_YAdiAYS`tfU^~`bfhY`tfUjlY3mgmXY3nyiA[]pgY3<rtyZY5mhmgY5m^syimp9Yx[]ZZj_\_Y5_

!#"$&%7
L ')(*"+P &%FQ>LxO "Q ,.-/"L '0,<NFOAN
VXWZY!I[]f|[]^~YTphYAf[bTpgyi[bjlj_ke^~Y5[tmyifgY3\ommgYAjon`b^ pgWZYidI[]jRfhY5mgyZjlp`]aFpgWZYphY5mp5.i`bfUYAn[b^~ZjlYtZUWZY5d^~Y3[bmgyZfg\_dZra`bf|TY
U\lpgW[mpgf|[]\_dCrz[]yZrtY5mSphWZYei[bfh[b^YApgY5f[tphyi[]j_jlkC^~Y5[bmgyZfhY5C\_m[]dYAj_Y5Tpgfh\_5[]j7fhY5mg\_mph[bdiTYt2 1X[bmgY5C`bdpgWiY|Wi[]dirbY~\ld
fhY5mg\_mph[bdiTYt][a`tfhAY.\om7A[bj_AyZjo[}pgY3][]dIxphWZYAdyimgyi[]j_jlk!di`bd~n\l^~Y5dim\_`bdI[]j_\l5Y5FVXWZYidi[]jnfhY5mgyZjlpF\omFpgWiYAfhYTa`bfhYX[ayZdIpg\_`bd
`baimY5ubYAf|[]jti[]f|[]^~YApgYAf|m539`]phWs\_dZWZY5fg\lp\_dxphWZY.^~Y3[bmgyZfgY5^~YAdzp^~YTphWZ`ns[bdix\_d!pgWiY.ZfhYTaY5fgfhY5x[kpg`SZfhY5mgYAdzpcphWZY.fhY5mgyZjlp5
Xg 3 cv9Zq5 4XvZq
VXWZYayidiph\l`tdi[]jfhYAjo[}ph\l`tdimWi\l\_mU5[]j_jlY3<pgWZYg|]iT}C7 6Ai}}[]di<\_mYTnZfhY5mhmY3
9 8  : 39; : > ; Q Q Q ; : .
3b
UWZY5fgYX\_mFphWZYU^~Y5[tmyZfhYA^~Y5dtpmg[b^ijlY[bdipgWZYXfhY5nyIpg\_`bdY5wzyi[]pg\_`bd 9 8 \_mq[vayidiph\l`td`ba -i[bfh[b^YApgY5fhm : 3 QQ Q<: .
0n[b^~ZjlY3m?a`bfcpgWZY.YTn9YAfh\l^~YAdzp|mR\_d!j_[]pgYAfF|Wi[]ZpgYAfc`]aimgyi|W!i[]f|[]^~YApgYAf|m[]fhYq5[]j_\lZf|[}ph\l`tdSa[bph`bf|mA5pg`})mgIY5Y5![bdi^~`"ZYAj
n\_^~YAdimg\_`bdimZynp[bj_mg`j_Y5mhm`b"u"\l`tyimqa[bph`bf|mjl\_bYpgY5^9YAf|[}phyZfhY[bdiphWZYAfh^~`nnk"di[]^~\ovZfh`b9YAfgpg\_Y5m`baX[}pgY5f5>VXWZYSif|mp
mpgY5<`bacpgWiYSY5fgfh`bfX[]dI[]j_k"mg\om.\ompg`~Y3mp|[]Zj_\_mgW<[j_\_mpU`]aFi[]f|[]^~YTphYAf|m.pgWi[]pX^yimpX9Y!T`tdim\onY5fgY3>dphWZYa`bj_jl`}U\_dZrpgWZY3mY
i[bfh[b ^~YTpgY5fhmX[bfgYnYAdi`]pgY3)Tl TI]Th}A}ng|T
mhmgyZ^~\ldZrphWi[}p[m^[]j_j|Wi[]dirbYS\ld[I[]f|[]^~YTphYA=f :?>FfhY5mgyZj{p|m\ld[mg^[]j_jR|Wi[]dirbYv\_d<pgWZY^~Y5[tmyZfhY5u}[]j_yZYbnVF[kzj_`bf
YA"I[]dimg\l`tdrt\lutY5m5
?t
@ BA : @ >9C : > A >: >D@> C : > > FEHG C :>KI J
A
A
P @ A @ :L>
Bz
:
A 9> C

ifg`t^ pgWZ\om5npgWZY iT?h|5 AAiX\_mvnYTidiY5[bm






[bdiephWZY!YAj_YA^~Y5dtp|[]jRY5fgfh`bf\omA

A : @ >

>


3b

Rt
A : @ >9C : > > C : >
d^~`tmpvA[tmY3mpgWZYx[bim`tjlyZpgY!u}[]j_yZY!`]a C : > \_mUdi`]A pv"dZ`}UdR0v[}pgWiYAf\lpv\omUrb\_ubY5d[tm[ZfgY3T\om\_`bdj_\l^~\lpi[bmgY5`td[
A`bdnIZYAdiAY\_dzpgYAfhu}[]j>[bmnY3mgAfg\_IY3\_dphWZY~ZfgY5u"\l`tyimmgY5Tpg\_`bdRxVXWiYfhY5A`b^~^~YAdinY3Zf|[bTpg\oTY\_mpgWiYztT`tdnInY5diTY
\_dzpgYAfhu}[]jnUWZYAfhYSpgWZY!ifg`ti[]Z\_j_\{pke`bacphWZYx[bTpgyi[bj C : > 9YA\_dZrj_Y5mhmX`bfXY5wzyi[bjpg`pgWiY!Y5mpg\_^~[]pgY3<u}[bjlyZY\omtz
: > N
D 0 D >
e ]
C
`tfXa`bfUZ\o[bmXY5fgfh`bf
: >  >  >
5t
C
WZY5fgYOD>0[]dI  >0[]fhYpgWZY!ZfhY5A\_mg\_`bd[]di<Z\o[bmXj_\l^~\lpv`]aFi[]f|[]^~YTphYA+f :L>7fgY3m9Y5Tpg\_ubY5jlkt
X ZZq vi0v? R0FZq
a[]j_jnYAj_YA^~YAdzp|[]jZY5fgfh`bf>m`tyZf|TY5m7[bfgY\_dinYA9YAdInYAdzp5tpgWiYT`b^Z\_dZY5YT?Y5Tpq`bdpgWZYU^~Y3[bmgyZfgY3su}[bjlyZY!YAfhfg`tfFifg`ti[]rz[}pg\_`bd9
\omA
; 9:: /> . > > >
5t
143
VXWZY!ZfhY5A\_mg\_`bd[]di<Z\o[bmXYAfhfh`bf|m[bfgYA[bj_AyZjo[}pgY3<mgYAI[]f|[}pgY5jlk?
]t
; 9:: /> . > D > >
>

 



143
.

; 9:: /> >  > >


n3
143
UWZY5fgY!mgyZimhTfh\lZp D\omUnYAdZ`bpgY3mXZfgY3T\om\_`bd[]dI  n Y5dZ`]phY5mUi\_[tmXYAfhfg`tf5
VXWZYph`]ph[bjYAfhfg`tfX^~[kIYa`tyZdi"k
>
>
bb
AT`tfhZ\ldZrph`C.`bj_YA^[]d[]di)zphYAY5jlY5bttvpgWi\_m!rt\lutY5m![bZZfh`"\_^[}phYAj_k[ttT`}ubY5fh[brbY`]apgWZYphfgyiYu[bjlyiYbVXWZY
[bj{phYAfhdi[}ph\lutY\omXpg`mgyZ^MphWZY!YAfhfh`bf|m

 ?t
VXWZ\omrb\_ubY3mbzT`}ubY5fh[brbYx\lapgWZYZ\o[bm[bdiZfgY3T\om\_`bdCYAfhfg`tfhmS[]fhY`]apgWZYmh[]^~Y~`bf|nYAf![]dICztA`}ubYAf|[]rtYx\la`tdZY~\_m
dZY5rbj_\lrt\lijlYA`b^~i[bfgY3pg`phWZY`]pgWiYAf3dzpgYAfhdi[]pg\_`bdi[bj.V`}U\ldZrCV[bdZ .`bdnaY5fgY5diTY5tbt!fgY3T`b^~^~YAdIZmx9`]phW)pg`CIY
fhYA9`bfgpgY3nZynp`tdZjlkepgWZYa`tfg^~YAfUU\_j_jRIYxZ\_mhTyimhmgY5<\ldjo[}pgY5f|Wi[]ZpgYAf|m5
a @ \_m.pgWiYY5fgfh`bdZY5`byimq^~Y5[bmgyZfhYA^~YAdzpX[]di \om.pgWZYpg`bph[bj9Y5fgfh`bfqafg`t^ Y5wzyi[}ph\l`tdtbTtphWZYSpgfhyZYvu}[bjlyZYS\om.a`byZdi
\_dpgWZY!\_dzpgYAfhu}[]j
@
! Bz
U\lpgW[~ifg`ti[]Z\_j_\{pk`]a0tzXVXWi\_mv`]a0T`tyZfhmgY!fhY5wzyZ\_fgY3mXpgWi[]pvphWZYstzMA`bdn9nYAdiAYx\_dtphYAfhu}[]jR\omvyImY3pgWZfh`byZrtWZ`byZpUpgWZY
[bdi[]j_knm\omAp\omq[bj_mg`s[tmgmgyZ^~Y5phWi[}p.pgWZYZ\_[tmqY5fgfh`bf7\ommk"^~^~YTpgfh\o]zUWZ\_|W\_mqdZ`]pX[]j_[knm>phWZY5[bmgYbFaRdZ`]p3"[xWZ\lrtW[]dI
j_`}j_\l^~\lp^syimp9Y!Y5mph[]ijl\omWiY5afg`t^MY5wzyi[]pg\_`bdbt
Y5fgYpgWiYY5fgfh`bfWi[bmvpgWiY~mh[]^~YyZdZ\lp![tmphWZY~^~Y5[bmgyZfhY5u}[bjlyZY\lpmgYAj_ubY3mA d[]jlpgY5fgdI[}pg\_ubY\_mph`nYAidZY~[gAl] }
Ag}"
8
bb

@
mg`pgWi[]p
@ 8
 Rt
VXWZY!fhYAjo[}ph\lutYY5fgfh`bfU5[]dIYx5[]joTyZjo[}phY5<zkyim\_dZr[gAl} }x .iA9|h3 ATIT
A @
8>
9]
@ A : >
\_dY5wzyi[}ph\l`tdimbt.vZ3



 

 

 







L!N$%
 #"Q %
%  %7N
VXWZYxa`bj_jl`}U\_dZrempgYAImv[]fhY!j_\_mpgY3\lddzpgY5fgdI[}pg\_`bdI[]jcV`}U\ldireV[bdZ.`tdnaYAfhYAdITYe3btt\_d`bf|nYAfUph`eTfhY5[]pgYs[bdyidiTY5f
p|[]\_dtpk\ldzpgY5fgu}[bj?a`bf[^~Y5[bmgyZfhYA^~YAdzp3
tX\_nY5dzpg\lak[bjljY5fgfh`bfXm`tyZf|TY5m
ZUnYTphYAfh^\_dZYphWZY!\ldIn\lu"\onyi[bjRZfhY5T\omg\l`td<[]dI<i\_[tmXYAfhfg`tfhmqa`bfUY3[b|WYAfhfg`tfXmg`byifhAY
?iUnYTphYAfh^\_dZYphWZYxmY5dimg\{ph\lu"\lpke`]aFpgWZY!Y5difgY3myij{pXpg`~Y5fgfh`bfUmg`byZf|TY3m
BIUTfhY5[}phYSpgWZYph`]ph[bjZfhY5T\omg\l`td<\_dzpgYAfhu}[]j
ZUTfhY5[}phYSpgWZYph`]ph[bjZ\o[bmUyZdITYAfgph[b\ldzpk
RiUT`b^Z\_dZYpg`bph[]ji\_[tmU[]di<ZfhY5T\omg\l`td<Y5\{phWZYAf"k[tZn\lpg\_`bd`bf!g3}. M 6Ai}gx^YApgWZ`nY3wzyi[}ph\l`tdbb
nUnY5Aj_[bfgYfhY5mgyZj{p|mafg`t^ Bz RzmY5i[]f|[}phYAj_kb
VXWZY!IfhmpUp.`mpgY5im[]fhYpgWiYpg`byirbW`bdZY3mA I`zmgmg\_ZjlYx[bZZfh`t[t|W<\omXpg`emph[bfpU\lpgWphWZYxfhY5nyiTpg\_`bdY5wzyi[]pg\_`bd[]dI
phWZYAdT`tdim\onY5f.\lphmXi[bfh[b^~YTpgY5fhm[]dIeUWi[]pX^~\_rbWzpX\l d
IyZYAdiAYSpgWZY5^0VXWiYAd<pgWZYfhYAZj_\_5[}ph\l`tdj_YAubY5j?^syimpUIY!T`tdimg\_nY5fgY3
\_d`tfhZYAfpg`)n\omph\ldZrtyZ\omWIYAp.Y5YAdZfgY3T\om\_`bd[]di Z\o[bmYAfhfg`tfhm5VXWZYwzyZY5mpg\_`bd\_mpk"Z\oA[bjlj_k)\lavpgWiYYAnIY5fg\_^~YAdzp\_m
fhYA9Y5[]pgY5RZ[]diUWZ\o|W<yidiTY5fp|[]\_dtph\lY3mU[]fhY!T`}ubY5fgY3"kfhYA9YTph\{ph\l`tdR
VXWZY\l d
iyZYAdITYT`"Y eA\lY5dtp|mv\oma`byZdI"ki[]fgpg\o[]j>nY5fg\_u}[}pg\_`bd`]a0phWZYfhY5Zyiph\l`tdY5wzyi[]pg\_`bdRY5\{phWZYAf[]di[bjlkzph\_5[]j_jlk\la
9`tmhm\_Zj_Y`bfUd"yZ^~YAfh\_5[]j_jlk"k|Wi[]dZrt\ldir\_dZZynpI[]f|[]^~YTphYAf|mqph`e[dzyi^Y5fg\oA[bjZfh`"AY5nyifgYt
VXWZYVUVvfhYA9`bfgp0yimY[smj_\lrtWzpgj_kn\l9Y5fgY5dzp0[k!ph`sA[bj_AyZjo[}pgYXphWZYUZfhY5A\_mg\l`td~YAfhfg`tf>pgWi[bdphWZY`tdZYU`bynphjl\_dZY3[]9`}ubYb
[]pgWZY5fXpgWi[bdY5mph[]ijl\omWi\ldZr[ZfhY5A\_mg\l`tdj_\l^~\lp Da`tfUY5[t|W<i[bfh[b^~YTpgY5+f :ei[pg`bph[bjZfgY3T\om\_`bdj_\l^~\lp\_m5[]joTyZjo[}phY5
.
6 8 ][ _ ; 9:: / >7> >
]t
> 143
VXWZY!ifgY3T\om\_`bd<YAfhfg`tf\ompgWiYAdrb\_ubY5d<"k
H 6 8 ] [ _
]t
UWZY5fgKY H  \omephWZYCYA\_rbWzpi[bmgY5`td[ zt T`tdnInY5diTYC\ldzphYAfhu[bjS[bdi  \ompgWZYd"yZ^s9YAf`]a!ph\l^~Y3mphWZYCpgY5mp\_m
fhYA9Y5[]pgY5RXV`idI H  afg`t^[empgyInYAdzp5 mUpSn\_mpgfh\_Zynpg\_`bdC[]d[ubY5fh[brbYnYArtfgY5Y!`]a>afhYAY5Z`b^\_mvfgY3wtyi\lfhY5i.Y5\lrtWtphY5"k
phWZY^[]rbdi\{phyinYe[]dICnY5rbfhYAY~`]aafhYAY5Z`b^m`]apgWZY\_din\_u"\_Zyi[]j0ifgY3T\om\_`bd\ldInYTnY5m 6 > eVXWZ\om!\omxn`tdZY~zkpgWZYYAjo|Wn
n[]pphYAfgpgW"[b\{phY![]ZZfh`n\l^[]pg\_`bdR
G 6 > > J >
8 ][ _
 ?bt

, 
VXWZ\omn`di`]pmgYAY5^Mpg`rt\lutYx[~m\_^~Zjl\lI5[}pg\_`bdA`b^~i[bfgY3eph`eA[]joTyij_[]pg\_dZre[~ZfhY5T\omg\l`td<j_\l^~\lpa`bfvY5[b|WYAj_YA^~Y5dtp|[]jYAfhfg`tf
mg`byZf|TYtnm`pgWi\_mU^~YApgWZ`n\_mdZ`]p[]iZjl\_Y5WZY5fgYt

   'AO ""! &%
%  %7N
VXWZY!`bf|<^`nnY5jR\omUyimgY5<\ld[]pXj_Y5[tmpUphWZfhYAYxn\l9Y5fgY5dzpvT`bdzphYT"phmX\_dpgWZ\omXpgWiY5mg\_m5qF\lf|mpUpgWZY5fgY!\ompgWZYsZW"k"mg\oA[]j^~`nnY5j_m
yimgY5\ldpgWZYYTn9YAfh\l^~YAdzp|mA\ YspgWZYWi[bfhnX[]fhYs[tpgyI[]j_jlkZynp\_dpgWiYX[}phYAfS\ldC`bf|nYAfvpg`^~Y3[bmgyZfgYsa`bf|TY`bfSfgY3m9`bdimgYb
vYT"ppgWZY5fgY\_mphWZY~^~[]pgWZY5^[}pg\oA[bj>^`nnY5j7yimgY5ph`nY5mhTfh\l9Y~m`t^YZW"knm\oAm5?I`zmgmg\_Zjlk`tdmgYAutYAf|[]jjlY5ubYAjom5sVXW"yimpgWZY
W"knnfh`"ZkzdI[]^~\_a`tfhAYU^~`"ZYAj9\_m.[^~[]pgWZY5^[}pg\oA[bji^~`nnYAjIyimgY5~\_dpgWZYm\_^syZjo[}ph\l`td`]a?pgWiY^~`]pg\_`bd`baR[xA[bZjlYtz[Wi\lrtWZYAf
j_YAutYAjR^[}phWZYA^[]pg\oA[]jc^`nnY5jXF\ldI[]j_jlkpgWZY5fgY!\omv[~d"yZ^~YAfh\oA[]j^~`nnY5jIWZYAfhY!yimgY5pg`eZY5mhTfh\l9YpgWZYs\l^~Zj_YA^~YAdzp|[}pg\_`bd`ba
[^[}phWZYA^[}ph\_5[]j>^~`nnYAjRphWzyIm\ldiAjlyIn\ldir<phWZYd"yZ^~YAfh\_5[]j>^~YTphWZ`nZmx[]dI`bpgWZY5f[bZZfh`n\l^[}ph\l`tdim^~[tnYpg`[bTpgyi[bjlj_k
mg`bj_ubY!phWZYsY3wzyi[}ph\l`tdimrt\lutYAd"kpgWZY^~[]pgWZY5^[}pg\oA[bjc^~`nnYAjom5 `tIYAayZj_jlk\lpU\_jljF9YAjlY3[]fUafh`b^pgWiYT`tdzpgYT"pSUWi[}pS\_m
[tpgyI[]j_jlke^~Y3[]dzpU\ldY5[t|W5[bmgYb
VXWZ#Y
I`}n\o[]rtfh[b^%\_dIrbyZfhY B!\_mq[bZ[bnpgY3safh`b^dzpgYAfhdi[]pg\_`bdi[bjiVc`}U\_dZr!VF[]dZ.`bdnaY5fgY5diTYx5bttF[]dIZY5mhTfh\l9Y5m
phWZY^~`"ZYAj_\ldZr[bdiu}[bjl\oZ[}ph\l`tdCifg`nTY3mgm5`bpgYpgWZYeZ\_mpg\_diph\l`tdIYAp.Y5YAdubYAfh\lIA[]pg\_`bdC[bdiCu[bjl\oZ[]pg\_`bdRcpgWZY~a`tfg^~YAf
9YA\_dZr~pgWiYxZfh`"AY5mhmX`]a7|WZY5|"\_dZrpgWi[]pUpgWZYs\l^~Zj_YA^~YAdzphY5T`nnYx[tpgyI[]j_jlk<rb\_ubY5mU[phfgyiY!fgY5ZfgY3mY5dzph[}ph\l`td`baphWZYx^[}phWZYT
^[}ph\_5[]j?^~`nnYAjm|nUWZ\_j_YvphWZYj_[]ppgY5fX\_m.pg`~utYAfh\{akpgWI[}pXpgWZYSd"yZ^~YAfh\_5[]j?^~`nnYAj?\omU[]d<[tnY5wzyi[]pgYfgY5ZfgY3mY5dzph[}ph\l`tde`]acpgWZY
ZW"knmg\_5[]j9fgY3[]j_\{pkt>`byZrtWZjlkemg9Y5[]"\_dZrxphWZYubY5fg\lIA[]pg\_`bdeU\_j_j\ldpgWZYSa`bj_j_`}U\ldZr9YSn`tdZY"keT`b^~i[bfg\omg`bd~pg`[bdi[]j_kzpg\oA[]j
fhY5mgyZ jlphma`bfmg9Y5T\o[]j?phY5mpvA[tmY3mAnUWZ\_j_YSu}[bjl\on\lpk`]aFpgWiY!dzyi^Y5fg\oA[bj?^`nnY5jR\omUi[bmgY5<`bdT`t^~i[]fh\_mg`bdpg`~YTn9YAfh\l^~Y5dtp|mA
T`tdnInY5diTY\_dzpgYAfhu}[]jX`]apgWZYd"yZ^~Y5fg\oA[bj^~`nnY5jU\_m~Y3mp|[]Zj_\_mgWZY3 \_d phWZYmg[b^~YX[k[bma`tf~pgWZYYTn9YAfh\l^~Y5dtp|[]j
fhY5mgyZjlphm5.T`tdimg\_nY5fg\_dZrpgWZYd"yZ^~YAfh\_5[]jX^~`nnYAjv[CfhY5nyiTpg\_`bdY5wzyi[]pg\_`bd[bdiY5mpg\_^~[]pg\_dZrjl\_^~\{p|m~a`bf~Y5[t|W i[bfh[b^~YTpgY5f5
qfgfh`bfXm`tyZf|TY5m.ph`IYxA`bdimg\_ZYAfhY5[bfgY\_din\oA[]pgY5\_dIrbyZfh Y Bi





Reality
Neglected Physical
Effects

Mathematical
Models

Representation
errors

Discretization
errors

Programming
errors

Verification

Numerical
model

Validation

Conceptual
Model

Computer
Code
Numerical
inaccuracies
Computation

Results

Comparison with
- analytical results
- Other numerical results
- experimental results
(benchmarks)

F\_rbyZfhY BI3bA ~}?O }}t}}s9g3||5g g||!  T?b}}iU IA9]]9] 4i7 a4i]'"0} TAgA9|
&(),)

g JC.>i RUc7v X!cic.


d A`b^~i[bfg\_dZr^~Y5[tmyifgY3[]di A[]joTyij_[]pgY5fgY3myZjlphmxpgWZYn\omph\ldIpg\_`bdIYAp.Y5YAd^~Y3[bmgyZfgY5^~YAdzpsYAfhfh`bf|ms[]dI)^~`"ZYAj_\ldZr
Y5fgfh`bf|m>[]fhYdi`]pq[]j_X[k"mF`tzu"\_`byim5>VXWZ\_m0\_m>phWZYA[tmYXa`tf7i[]f|[]^~YApgYAf|m>dZ`bp0YTnZj_\_A\{phjlkmp|[}pgY3\ld~phWZYUfgY3nyiTpg\_`bd~Y5wzyi[]pg\_`bdR
mg`~pgWi[]pv[~d"yZ^~YAfh\_5[]j^~`nnYAj\_mUdiY5TY3mgmh[]fhkph`Y5mpg\_^~[]pgYxmgYAdimg\lpg\_uz\lpkb.dpgWZY!a`tjlj_`}U\_dZr~i[]f|[]^~YTphYAf|mX`]aFpgWZ\omUpk"9Y!U\lj_j
9Y~pgfhY5[]pgY5[bm^~`nnY5jl\_dZrYAfhfg`tfhmUWZ\_jlY~phWZYpgY5fg^'^Y3[bmgyZfhYA^~YAdzpYAfhfg`tfhmSU\_jlj79YfgY3mY5fgutY5a`bf!yZdiAYAfgph[b\ldzpg\_Y5mS\_dpgWZY
[tpgyI[]j^~Y5[tmyZfhYA^~Y5dtp3

0j_`]pgpg\_dZrYAnIY5fg\_^~YAdzph[bjfhY5mgyZjlphmS[]diYAfhfg`tfhmU\_dphWZY~mg[b^YZj_`]phmS[bmd"yZ^~YAfh\_5[]jFfgY3myij{p|m[bdiYAfhfh`bf|mrb\_ubY3mv[rb`"`n
\_^~ZfgY3mgmg\_`bd`ba0pgWZY~`}utYAfhj_[b`]a0phWZYp.`I9iynp!^\_rbWzpfgY3wzyZ\lfhY[<j_`]p`]aqZj_`]p|mA dC[bj{phYAfhdi[}ph\lutYbwzyi[]dzph\{p|[}pg\_ubYX[k`ba
ZfhY5mgYAdzph\ldZrphWZYxT`t^I[]fh\_mg`bde9YTpYAY5d^`nnY5jc[]dIYA"9YAfh\_^Y5dzp\_mUmgyZrbrtY5mpgY3eph`9Yb
 C H 6 W| >. > 
]
 ?Z3
] 7
vYAfhY 6 W \om>pgWZYmp|[]diZ[bfh~nY5u"\_[]pg\_`bd`]a?pgWZYn\{?YAfhYAdITYU9YTpYAYAdphWZYd"yZ^~YAfh\_5[]ji[bdiYTn9YAfh\l^~YAdzp|[]ji[bdimYAf3! H>\_m7pgWZY
YA\_rbWzp[]dimgWZ`tyZj_A`bfhfgY3m9`bdiepg`~phWZYxYA\_rbWzpyimgY5\ldphWZYxYAfhfh`bfXY5mpg\_^[}pgY3mA ] 7 \_mXphWZYxY3mph\l^[}phY5YTn9YAfh\l^~Y5dtp|[]j
Y5fgfh`bfv[bdi !phWZYY3mph\l^[]pgY5^~`"ZYAj_\ldZr<YAfhfg`tf5VXWZYsa`tfg^yZjo[\_mSi[tmY3`bdC[T`t^~^Y5dzp"k.`bj_YA^[]d[bdizphYAY5jlY
5btt`td fhYArtfgY3mgmg\l`td)yZdiAYAfgph[]\_dzpkb)VXWZY\_ZY5[\ompgWi[]pphWZYn\_mhTfhYAI[]diAk9YTpYAY5d ^~`nnYAjU[bdiYA"9YAfh\_YA^~dzp^syImp
9YA`bdimg\onYAfhY5ph`brtWZYTphYAfU\{phWpgWiY<^~Y5[tmyifgY5^dzp[]di^`nnY5jl\_dZrYAfhfg`tfhm5apgWZYf|[}pg\_` ]~\omyidZ\{pk)`tfj_Y5mhm5>pgWZY
u}[]j_yZY~Y5mpg\_^[}phY5afg`t^pgWZY~^~`nnYAj7\omU\lpgWZ\_dpgWZYT`t^sZ\_dZY3yZdITYAfgph[b\ldzpk`baqpgWiY^~Y3[bmgyZfgY5^~YAdzp5~VXWZ\om\_mSphWZY~IY3mp
A`b^~i[]fh\om`tdpgWi[]pA[]d<IY`tnph[b\ldZY3eU\lpgW<pgWZYYA"9YAfh\_^Y5dzph[]jZ[]ph[~mgYTp5n[bdiephWZY^~`"ZYAjRmgWZ`byZjophWZYAd9Y[tAAYAnp|[]Zj_Yb
i`bf ]sjo[]fhrbY5fpgWi[bdyZdZ\lpkphWZYe^~`nnYAj0U\lpgWphWZYej_`}.Y3mpxf|[}ph\l`rt\lutY5mphWZYIY3mpfgY5ZfgY3mY5dzph[}ph\l`tdC`baXpgWZYe^~Y3[bmgyZfgY3
u}[]j_yZYtVXWZ\omfhZfhY5mgYAdzph[]pg\_`bd.`tfgnmX9Y5mp\la>pgWZYxY5fgfh`"Y\ldphWZYx^~Y5[tmyifgY3i[}ph[Z`zY3mUdZ`bpvu}[]fhkepg`^syi|Wafh`b^A[bmgYpg`
5[bmgYb

<L  '   "O !$
Q # D% +P "Q &! L =P , '5O "" &%
dphWZ\_mvYTZ[]^~Zj_Y!phWZYxj_\{a p39`tfUpgf|[]dImutYAf|mYSa`bf|TY3mAi`bd[~fh\lrt\_Akzj_\_dinYAf[}pvj_`}[bdZrbj_Yx`ba7[]pp|[b|nmU[]fhY!^~Y5[bmgyZfhY5UVXWZY
YA"9YAfh\_YA^~dzph[]j.mYApgyZ)\omxmgWZ`}Ud\ldirtyZfhYn<VXWiYAfhY\_m[a`bf|TYemgYAdimg`bf!\_d)Y5[t|WY5di`]aXpgWiYpgY3mpmgY5Tpg\_`bdRpgW"yim!pgWZY
ph`]ph[bj?a`bf|TYS\_mX[~mgyZ^M `]ap.`^~Y5[tmyZfhYA^~Y5dtp|mA7VXWZYa`bf|TY!mgYAdimg`bf|m[]fhYS[mp|[]dii[]f|mpgf|[]\_drt[byZrbY3mi[bmgY5<nY5mg\lrtd<yImY3
YAzphYAdimg\_ubYAj_k"(k ' UVU @~FUWZ\lj_Y[]j_j.utYAj_`"A\{ph\lY3m[bfgYYAdiA`"ZYAf`bynphZynphmafh`b^pgWiY5[]fhfg\o[]rtY5m5VXWZY<^~`nnY5jX[bm


VF[]Zj_Y~bU]g!TA|x!ZsT0A|T}
cYAdZrbpgW

Z ,?^
T\_[b^YApgY5f

i zn^
TfhkYA\_rbWzp 
b $ ] @r
[]pgYAfZYAdimg\{pk  btZ  @r ^ > I
Tkzdc7uz\omhT`tmg\lpk
b t
U3 ^ ]m

RY5dZr]phWph`en\o[]^~YTphYAfUf|[}pg\_`
?tZ
zpgfhyiphyZf|[]j^[bmhmpg` ZnY5^[bmhmXf|[}pg\_`   
i BR

U`tyZrbWidZY5mhm  ]S Q  5
|f []d9`]phWC\_dphWZY~j_[bfgrtYxph`}U\ldZrph[bdZ[]di\ld('C[]R!dI`bpgWCj_[bI`tfh[]pg`bfh\_Y5mv[ Y5\_ZYAWZY5\ld('xUYAfhYsyimgY5a`bf
[b^ijl\lIA[]pg\_`bdR3ijlpgY5fg\_dZrS[]dix[bdi[]j_`brpg`Z\lrt\{p|[]j"T`bd"utYAf|m\_`bd`]ai[bjljz^~Y5[tmyZfhYA^~Y5dtp|mA jlpgWZ`tyZrbWsmh[]^~Y.^~[bbYq[bdi!^~`nnYAj
p`n\l?YAfhYAdzpUyZdZ\lphmXYAfhYvyimgY5>d<pgWZYSph`}U\ldZrsph[bdZe[mh[]^~Zj_YafgY3wtyiYAdiAk`]aFbb v![]di<[xijlpgY5fUTynph`]afhY5wzyZYAdITk
`ba3b !YAfhYyimgY5RqZ`tfXpgWZYa`tjlj_`}UyZYTnIY5fg\_^~YAdzphmX\_dpgWZY ' pgWZYxmh[]^~Zj_\ldirafhY5wzyZY5diTk[tmXfgY3nyiTY3eph`3b
vbZIj{phYAfh\ldZr<[}p Bz b jljca`bf|TYxmg\lrtdi[]jomUYAfhY!fgY3T`bf|nY3\_ad c7`bjlphm5IUWi\lj_YsA[bjl\_Zf|[}ph\l`td[bdi[]j_jZ[]ph[[bdi[]j_knm\omUYAfhY
9YAfga`bfh^~Y5<`bd[
XyIm\_dZr ' V  1 


y
x
B

F\_rbyZfhYSZ54 Z 5tT8}nI|n9tTSZx|} t T>AhA} }hhTAZA] i}"  "A9s3bi}


]9~n!"$#b b
^~I`tfp|[]dzpi[bfh[b^YApgY5fhm[]fhYrb\_ubYAd\_d<p|[]Zj_Y~b
dpgWZYXph`}U\ldZrph[bdZ? Y3[b|WfgyZd.Y5fgY9YAfga`bfh^Y3s\_dI`bpgW~n\lfhY5Tpg\_`bdima`tf7Y5[b|WT`t^sZ\_di[}ph\l`tds`baIubY5jl`nA\{pkx[bdis[bdZrbj_Yb
"\lfhYA9YTpg\lpg\_`bdIm[bfgYepgWZYAa`bfhY[u}[]\_jo[]Zj_Ya`bfphWZY5mgYi[}ph[I`t\ldzp|mA)VXWZ\omU\lj_jXIYyimgY5ph`C\_j_jlyimpgf|[}phYphWZYn\{?YAfhYAdiY5TY
9YTpYAY5d[t mp[]di[en di<`bf|nYAfXfhYAZj_\_5[}ph\l`td[bdi[]j_knm\omA
%Xg 3 cv9Zq
4XFZq
VXWZYZfhY5mgYAdzphY5efhY5mgyZjlphm`]apgWZYYA"9YAfh\_^Y5dzphm\_mphWZYjl\la pT`"YT?Y5A\lY5dzpU[tmX[sayidiph\l`td<`baF[]dirbj_Yv`][]ph[t|' &)T`"Y eT\_YAdzp|mA
TYAdZ`bpg\_dZrSpgWZYU^~Y3[bmgyZfgY3xa`tfhAY5mFafh`b^phWZYXp`a`tfhAYUmY5dim`tfhmFU\lpgWmgyZIY5fhmhTfh\_nphm []di)
(!a`tf>a`bfh[bfh[]di[}a pqmgYAdimg`bf

fhY5mgIY3ph\lutYAj_kb"pgWiY!fgY3nyiph\l`td<Y3wzyi[}ph\l`td<\om

>

 &

Z 

3
>  >  

?tb
?+?t
[bdifhYAZfhY5mgYAdn

V`srtYAdZY5fh[bjl\_AYt]pgWiYfhY5nyIpg\_`bdY5wzyi[}ph\l`tdimqA`bdzph[b\ld^Y3[bmgyZfhY5a`bf|TY3mA"[sZkzdI[]^~\_vZfhY5mhmyZfhY! >3  >


p|[}pg\_ubY!n\_^~YAdimg\_`bdim50VXWZYYAfhfg`tfUm`tyZfhAY5mX[]fhYrbfh`byZ9Y5<[b5T`bf|n\_dZrpg`~pgWi\_m5
%X CZc"q 7

7g Z N|0 Nhqi?R7"icc.n


i`bfXpgWiYa`bf|TY^~Y5[bmgyZfhYA^~YAdzp|mAzpgWiYAfhYx[]fhYvphWZfgY5Y!fgY5Zj_\_5[}pg\_`bdIm.j_YAutYAjomU[u}[]\_j_[bZjlYt
VXWZY<mgYAdimg`bf|mxYAfhY5[]j_\lZf|[}phY5)[]a pgYAf[bmhmY5^sZj_kb 1.kfh`]p|[}pg\_dZrphWZY<mgYAdIm`tfs[]di)pgWZYmI[bTY5fx\lpX[bms^`tyZdzpgY3
`bdR9"dZ`}Ud.Y5\lrtWzphmU`baX3}]trtfh[b^mU.Y5fgYx[bZZj_\lY3\ldCn\lfhY5Tpg\_`bdimZ bil]Z_5tZ ! BzZt[bdi9]nYArtfgY5Y5mUU\lpgW
fgY3m9Y5Tp!pg`pgWiYeI`zm\lpg\_ub Y [}n\omA<VXWZ\_mxrt[ubY~^~`bfhYpgWi[bd 3]9`b\_dzphm!`tdUWi\_|Wpg`iyZ\ljo)[ TM5[]j_\lZf|[}ph\l`td
^`nnY5jR\_diT`tfg9`bf|[}ph\ldZr~Tfh`tmhmp|[]j_~[bdi[]j_\lrtdZ^~YAdzpUYAfhfg`tfhm5
^e
 3 3   >  >
 ?Bz
YAfhY ^ \_mspgWiYe"dZ`}Uda`bf|TYe\_dn\_fgY3ph\l`td 7 [bfgYepgWZYe^~Y3[bmgyZfgY3ub`tj{p|[]rbY3m\ldpgWZYp`n\_fhY5ph\l`tdimx`ba
pgWZYmY5dim`tf[]di  3 [bdi  > []fhYxT`"Y eT\_YAdzp|mUa`byidi 3 "> kepgWiYsj_Y5[bmpvmhwtyI[]fhYZpv^~YTphWZ`nUVXWZYxfhY5mgyZj{p\_m BemYAphm`ba
T`"Y eT\_YAdzp|mA0VXWZ\omX\omXpgWZY!5YAfh`]phW<fhYAZj_\oA[}ph\l`td<j_YAutYAj
VXWZYx^~Y5[tmyZfhYA^~Y5dtp|mU[]fhYSpg\_^YmY5fg\_Y5mXUWZY5fgYphWZYpg\_^~YsmgYAfh\lY3mU[]fhY[]pvj_Y5[bmp!3~mgY5A`bdiZmXj_`bdiriqVXWZYa`tfhAY5mUyImY3
\ldpgWiYx[]di[bjlknmg\_mX\ompgWZY!^~Y3[]d`]aFpgWZ\omXpg\_^~Y!^~Y5[bmgyZfhYA^~YAdzp|m[bdifgY5ZfhY5mgYAdzphmX[b mpUfhYAZj_\oA[}ph\l`td<j_YAutYAj
pj_Y5[tmpUm`t^~Yv`]aphWZYpg\_^Y!mgYAfh\lY3mq[bfgYfgY5IY3[}pgY3emg\{ph\l^~Y5m5"fgY5ZfhY5mgYAdzpg\_dZrs[n die`tfhnY5fqfhYAZj_\oA[}ph\l`tdej_YAubY5jIU\lpgW
?ZYArbfhYAY3mS`]aqafhYAY3n`b^ Y5fgY~u}[]fh\o[}pg\_`bdIm\_d[]j_\lrtdZ^~YAdzpx[]diCubY5jl`nT\lpk[]fhY~\ldITj_yinY5C\_dCphWZYZ[}p|[mYAp5mg`<pgWZY
ZfgY3T\om\_`bdj_YAubY5jRmWi`byZjoIY!WZ\_rbWiYAfa`bfXphWZ\_mvA[bmgYb

0fhY5A\_mg\_`bdj_\l^~\lphm[]fhYrb\_ubY5d\ldp|[]Zj_YSZFi`bfAY5fg`bpgW`bf|nY5f0fgY5Zj_\_5[}pg\_`bd~phWZYSu[bjlyiY\_mqpgWiYtzT`bdZInYAdITY\_dzpgYAfhu}[]j
`ba7pgWZYA[]j_\_Zfh[]pg\_`bd^~`nnYAjVXWiYb mpv`bf|nYAfjlY5ubY5jcyimgY5mvpgWZYsa`bfh[bfhfhyZdim`bdphWZYsphWZ\_fhi[kb9UWZ\lj_YxphWZYZ dI`bf|nYAf
\_diTj_yinY3m.[bjlj9^~Y5[tmyZfhY5Z[}p|[Z0"\li[}ph[sI`t\ldzp|mq\om.[tmgmgyZ^~Y5a`bfX[]j_j9[bdZrbj_Y5m.\ldepgWZ\omXA[bmgYb>Z`tfqphWZYSWZ\lrtWZYAf.`bf|nYAf.jlY5ubY5j_m
phWZY!ZfhY5T\omg\l`td\omUrb\_ubY5d<"k?
6 7
D7  H
 ?tb
dph[bZj_YspgWZY!^~Y5Z\_[bd\omUrt\lutYAdfh[]pgWZY5fXpgWi[bdpgWZYx^~Y3[]dRqVXWiY!^Y3n\o[]d\omXpgWZY!u}[bjlyZYa`tfUWZ\o|WpgWZYxifg`ti[]Z\_j_\{pk
`baFYAnAYAY3nYAdiAY\_mvbz7d`]phWZYAfU`bf|Zm5nWi[]jlaphWZYxA[bj_AyZjo[}pgY3j_\_^\lphm[]fhYIY5jl`}pgWZ\omUj_\l^~\lp5I[bdiWi[]jlaFpgWZY!j_\l^~\lphm[]fhY
[bI`}utYx\{p3s"\_diTYfgY5j_[]pg\_ubY5jlkaYA u[bjlyiY5m~^~`tmpgj_k &]zv[bfgYwtyi\{phYWZ\_rbWCT`b^~i[bfgY3pg`<pgWiYj_\l^~\lpa`bfS^`zmp!A[tmY3mA
phWZY^~Y5n\o[]d\_mq^`tfgYUfhYAifgY3mY5dtp|[}ph\lutY.phWi[]depgWZY^~Y3[]dR7Zfh`b^\_dim9Y5Tpg\_`bd`ba9phWZYvf|[ Z[}p|[\{p[tm>a`byZdI~pgWi[]p0a`bfb mp
`tfhnY5fXfgY5Zjl\oA[]pg\_`bd<pgWZY!u}[bjlyZY!n`"Y3mUdZ`]pUu}[]fhkeU\lpgW[]dirbj_YS`baF[]pph[t|?"pgW"yimXphWZY!u}[]fh\_[]pg\_`bd\omU^[]\_dZj_knyZYph`~ubYAj_`nT\lpkb
Z`tf & ] phWZYx^~`nnYAj[tmX`bimgYAfhubY3eph`e`tmhT\_j_j_[]pgY![]diphWZYsmph[bdiZ[]f|nYAu"\o[}pg\_`bd`baFphWZY!ph\l^~YsmgYAfh\lY3mX\ldITfhY5[bmgY5m
"k[a[tpg`tf7iutY`tf0^~`bfhYbFVXWZ\_m0\_m7fhY
IY5phY5\_d~pgWZYU^[}u}[]j_yZYU`]a9pgWZYvZfgY3T\om\_`bdj_YAubY5j>Z`tf0mY3T`tdi`bf|nYAf>fhYAZj_\_5[}ph\l`td
phWZYZ\lfhY5Tpg\_`bdfhY5mgyZjlphmU[bfgY^syi|WWZ\_rbWZY5f.phWi[]dYTn9Y5TpgY5Rni[]fgpg\oTyZjo[]fhj_kem\_diTYpgWiY5mgYSu}[bjlyZY!mgWZ`byij_<IYj_Y5mhmXmY5dim\lpg\_ubY
ph`mg^[]j_j>u[bfg\o[}ph\l`tdimS\ld[]dirbj_Y`ba[}pgph[t|pgWi[bdphWZYkzn\_fhY5ph\l`tdRVXWi\_mS^[k\_^ijlkyidimphY5[tnk
i`}%I[}pphYAfhda`tfSpgWZY
WZ\_rbW[]dirbj_Y5m5
VXWZYWZ\_rbWZY3mpXj_\l^~\lphmU[]fhYa`byZdiea`bfpgWiYSWi\lrtWZY5mp^~Y3[bmgyZfgY3eu}[]j_yZY3mAnmg`sphWZYfgY5j_[]pg\_ubYvY5fgfh`bf\om.^~`tfgYT`tdimp|[]dzpqphWi[]d
phWZYsZfhY5A\_mg\_`bdjlY5ubY5j_mv\ldiZ\_5[}pgY3mAV.k"Z\oA[]j_j_kbZphWZYsifgY3T\om\_`bdj_\l^~\lp\omv[]9`byZptM`]a>phWZYs^~Y5[tmyifgY3u[bjlyiY5m5I[}pjlY3[bmpv\ld
phWZY!kzn\_fhY5ph\l`tdR
VXWZYen di`bf|nYAfSjlY5ubY5jF\omWZY5fgY\ldITj_yinY5a`bfT`t^~i[]fh\_mg`bd`bdZj_kbn\ldiAYpgWZY5fgY\_m[mg^~[bjlj7n\l?YAfhYAdiAY\_dmg9YAY5\ld
a`tfgX[]f|[]di)I[b|X[]f|ZmsfgyZdImAFpgWiYfhY5mgyZj{ph\ldirZfhY5T\omg\l`td)jl\_^~\{p|ms\omxph`"`WZ\_rbWRWZYAdY3mph\l^[]pg\_dZrphWZYeidI[]jYAfhfg`tf
afh`b^fgY5IY3[}pgY3fgyZdImAipgWZYsA`"Y eT\_YAdzphmv[]fhY5[]joTyZjo[}phY5<a`bfS[]j_jcZ[]ph[mgYAfh\lY3mAnphWZYAdphWZYx^~Y5[bd[]diphWZY!ZfhY5T\omg\l`tdj_YAutYAj
\omXY5mph[bZjl\omgWZY5
nY5dtph\{IY5m`tyZf|TY5mX`baZ\o[bmXY5fgfh`bf|m\ld<pgWiYa`bf|TY^~Y5[bmgyZfhYA^~YAdzp|mU[]fhY
!#"%$&('<) 3+* "\_diAYpgWZYph`]ph[bja`bf|TY`tdAkzj_\_dinYAfU\omU^~Y5[tmyifgY3Z[]j_j\ldiYAfgpg\o[sa`tfhAY5mUnyiYpg`~u"\lZf|[}ph\l`td<\_dinyITY5
zkphWZYS5[]fhfg\o[]rtY[]die\{p|m.nfh\_ubYmgknmphYA^ []fhY\_diTj_yinY3e\_dpgWiY^~Y5[tmyifgY5^Y5dzphm5 1ynp.pgWZY3mY[]fhYv[bmhmgyZ^~Y5~ph`sWi[utY
AYAfh`e^~Y5[bdRiY5dipgWiYsYT?Y5Tp\om\_diTj_yinY3\_dpgWZYZfgY3T\om\_`bdY5fgfh`bf|mX\ldp|[]Zj_YZUVXWZYxYT?Y5Tp`ba>u"\_Zf|[}pg\_`bd`tdpgWZY
[ubYAf|[]rtY5W"knnfh`nnk"di[]^~\oZ[b^~Z\ldirsa`tfhAY5mXnYA9YAdIZmU`bdpgWZY!`bfhn`bdiYS"kphWZY!Tk"j_\ldiZYAfU`bd<pgWiY
IyZ\_RZUWZ\o|W
[]rt[b\ldnY5IY5diZmx`tdpgWZYe[tpgyI[]j7^~`]ph\l`td`ba.phWZYAkzj_\_dinYAf3ifg`t^ nY55[kpgY3mp`]apgWZY^~`nnYAjq^`tyZdzpgY3C`tdpgWZY

VF[]Zj_YZhhAo}]9)A<{  xh}zT}?I~|]}ZI& - } T]g| ~|}|nh TI  5 }Tl3A T]9


] bl! ?Tit|+  & ( ?}Xhb?|bAl 
UY5Zjl\oA[]pg\_`bd n[b^~ZjlY3m YA\_rbWzp
ktn\_fgY3ph\l`td
Tn\_fgY3ph\l`td
cYAubY5j

p ^\_d ^[} ^~Y5Z\_[bd


^~\_d ^[} ^~Y5Z\_[bd
Z pgW
]

Z Z
i z]
t mp
Bzbb
i zn i Ri R Z  Bt
i z] Z !Bt Z +R+R
n di
R
Z $  ?i Ri3 Z b+ R
i +B Z ?t Z ?t
1\_[tmYAfhfg`tfhm
cv\_Zf|[}pg\_`bd
)
Z Z_3 i ttt b
U3 Z bb Z bZ
.k"jl\_dinY5f0di  ) >3 i ttb BI B U5 
(
"pgfhynp[bbY  ) I
(
i tZ3
 ^ 
'Y3[bmgyZfgY3<i`bf|TY
?G" ?b$  Z RZ3 Bi b! U5   b bZ Z ?+Rb
A []fhfh\_[brbY!phWZY~jl`}Y5mpY5\lrtYAdnafhY5wzyZY5diTk\oma`byZdipg`9Y~ tsVXWZY~WZ\_rbWZY3mp!mgWZY5ZZ\ldZr<afhY5wzyZYAdITk\omva`tyZdia`bf
n e^ }mph`} m9YAY3C[bdiI?bZYArbfhYAY3mvkz[ []dirbj_YbsVXWZYY5mpg\_^[}phY5Cu[bjlyiY\om \ BSQ < RSQ ,ReUWZY5fgYpgWZY~j_`}
u[bjlyiYT`tfgfhY5mgI`tdiZmpg`CmyZ?Tfh\lpg\oA[]j
i`}zpgfh`byiWi[]j0d"yZ^IY5f 6 _ Q b[bdipgWiYeWZ\_rbWjl\_^~\{pT`bfhfhY5mgI`tdiZmSpg`
pgf|[]dimhTfh\{ph\_5[] j
I`}! 6 _ S Q bI[]jlpg\_dimY5d5tbzFn\ldiAYUpgWiYv\_dejl\_dZY`zmgA\lj_jo[}pg\_dZr!a`tfhAYUWi[bm0pU\_AYUpgWiYafhY5wzyZYAdITk
`]a9pgWZYUut`bfgpgYTsmgWZY3Zn\_dZri}phWZYAfhYX\_m7[S9`tmhm\_Z\_jl\lpkx`ba9^[}p||WZ\_dZrjl`z[bs[bdiYA\_rbY5dnafgY3wzyZYAdiAka`bf>phWZYUYT"pgfhYA^~YUA[tmY3mA
SmgA\lj_j_[]pg\_`bd`]aZphWZYX^`nnY5j"[tm`timgYAfhubY5R39`]phWsu"\omyi[bjlj_kx[]dIx\_dsphWZYqa`tfhAY.^~Y3[bmgyZfgY5^~YAdzphm5a`bfFubY5jl`nT\lpg\_Y5m>[bI`}utY
b !^ }m[]di[]dZrtjlY3m.[bI`}utY]xZYArbfhYAY3mA>Z`tfqphWZYSfgY5^~[b\ldi\ldZrxpgY5mphmdZ``tmhT\_jljo[}ph\l`td`]aphWZY^~`nnY5j9X[bmq`bimgYAfhubY3
[]di<pgWiY!^`bpg\_`bd`]aFpgWiYxTk"jl\_dinY5f\om[bmhmyi^Y3j_\_^\lpgY3pg`pgWiYsZ_^^Tj_Y5[bfh[bdiTY`]apgWZYsZ\ld`baphWZYxmY5dimg`bfXpg`
pgWZYZpgpg\_dZr\ldpgWZYpgY3mp!Tk"j_\ldInYAf3sZ`tfvphWZYA[tmY`]a c csi[bj{ph\ldImY5d5bttrb\_ubY[mg\l^~Zj_YfhYAjo[}ph\l`td9YTpYAY5d
nfh[brT`"Y eT\_YAdzp[]dI[b^~Zjl\lpgyInY!`]aF`tmhT\_jljo[}ph\l`tdR

%.    Z 7
 ?+Rt
)
 YAfhY  )!\omUpgWiYsnf|[]rT`"Y eA\lY5dtpU\{phWZ`bynpuz\_Zf|[}ph\l`tdR  \omXpgWiYs[]^~Zj_\{phyinYs[bdi e\omUphWZYsTk"j_\ldInYAfvZ\_[b^YApgY5f5
jlpgWZ`tyZrbWpgWiYeTyZfhfgY5dzp!5[bmgY\om!dZ`b5p cv csRpgWZ\om!fhYAZfhY5mgYAdzp![]dyZZ9YAf!j_\l^~\lp5VXWiYi\_[tm!nyZY~pg`u"\_Zfh[]pg\_`bdA[bd
pgWZY5dIY!YA"ifgY3mgmgY5
 ) 3   
 ?$]
 Z 7 ) Q b+ B9 )
WZYAfhAY )!\omXphWZY!^~Y5[bmgyZfhY5ea`bf|TYb

 $   $    +& '  ) > * V`~YAdimgyZfhYvphWi[}p[bjlj?a`tfhAY5mX[bTpgyi[bjlj_krb`"Y5mphfg`tyZrbWpgWZYSa`tfhAYSmgYAdimg`bfU[rt[b<9YTpYAY5d
pgWZY.pgY3mp7Tk"j_\ldiZYAf>[]dixpgWZYmyZiI`tfp>\_mdZY3TY5mhmh[]fhkbcVXWZY9`byZdii[]fhkjo[kbY5fRpgWi\_|"dZY3mgmF[}ppgWZY.a`bfh[bfhrz[]x\omca`byZdi
pg`  P, Q x^^n[]9`byZpqpgWiYmh[]^~Y[tm7pgWiYsv^~^ rz[]`b9YAdi\ldZrI>p.\om0phWZYAd[bmhmyZ^~Y3pgWi[]p.phWZYp|[]dZrtYAdzpg\o[]j
i`}
ubYAj_`nT\lpk[]p0pgWZYvrt[]~\omqmg^[]j_ji[]diphWi[}p.I`bpgYAdzph\_[bjZa`bf|TY3m>afh`b^mgyiZZYAde|Wi[bdZrbYU\_den\o[]^~YTphYAfq5[]d9YdZYArtjlY3phY5
VXWZYs[]dI[]j_k"mg\om.\omXphWZYAdj_\l^~\lpgY3pg`\_diTj_yinYafh\_Tpg\_`bdi[bja`bf|TY3mUnyZYph`jl`}m9YAY3
i`}
9YTpYAY5dpgWZY!`tynpgY5fZjo[}pgY
`]a>pgWZYmyZiI`tfpv[bdi<pgWZYxifg`bpgfhyim\_`bd[}pUphWZYsAkzj_\_dinYAfvYAdiXVXWZ\omU\omv[bmhmgyZ^~Y5pg`e9Ysmg\l^~\_j_[bfXpg`[eQ T%|Wi[]didZYAj

i`}!mgYAYfh\lrtWzpI[]fgpx`]airbyZfh%
Y RZVXWiYWZYA\_rbWzpx`]apgWiY|Wi[]didZYAj7\om  ^~^pgWZYZfhY5[]pgW G~ B^~^ pgWZY
jlY5dZr]phWe`]aphWZYifg`bpgfhyim\_`bd~`tdpgWZYSTk"j_\ldiZYAfqY5diiT"[bdi~phWZYvj_YAdZrbpgW c >3 > "\ YWi[]jlapgWiYT\_f|TyZ^aYAfhYAdITY`bapgWZY
Tk"jl\_dinY5f50ifg`t^ 1jlY5u"\ldim!5ttnnVF[]Zj_Y ?] B"7phWZYxmgWZY5[bf.a`tfhAY`bd<pgWZY!X[]j_j_mU[bfgYt

 ?bt
   ` 3 < ` >
YAfhY ` 3 < ` > \_m~phWZYZfhY5mhmyifgYnfh`bpgWZfh`byZrtW phWZY|Wi[bdZdZY5j VXWZYZfhY5mhmgyZfgYn\l?YAfhYAdiAY`tdpgWZYyZImphfgY3[]^
[]diCjlY5YAX[]f|3 mg\_ZY`baqpgWiY~Ak"jl\_di> nY5fxA[]dCIYY3mph\l^[]pgY5C"kpgWZYnk"di[b^~\_ZfhY5mhmyifgY`]a0phWZYeTfh`tmhm
i`} mg`phWi[}p
` < ` P  m\_dc ( ` ("g \{phWY3wtyI[]j
I`}`td9`]pgWCm\onYs`ba7pgWZYZfh`]pgfhyimg\l`tdRZphWZYxph`]p|[]ja`bf|TYx`tdpgWZY
Tk"3jl\_dinY5f> YAdI> \om

 ) >  +G G  Cmg\ld &> >


?bt

$  ! &! &#" "  &)VXWZYvyimY`]aR[xYAj_jIZ`"AyZ^~YAdzpgY3e[]dIpgY5mpgY3~pg`"`tjij_\ltY ' V  1   mWZ`tyZjofgY3nyiAYUpgWZYd"yZ^s9YAf


`]aZZyZrzmc\_d!phWZY.A`b^~ZynphYAf[bdi[]j_knm\omcmhTfh\_nphm53[bdiphWZYAfhYq\omRdZ`[kvpg`SwtyI[]dzpg\la kSpgWZY.I`zmgmg\li\lj_\{pkS`]aZiyZrtm5 dZ`]phWZYAf
wtyiY5mpg\_`bd~\_m0UWZ\_|W~i[bfp7`baIphWZY^~Y5[tmyZfhY5spg\_^~YmY5fg\_Y5mFph`!yimgYU\ld[]dI[]j_k"mg\omA j_jZfgY3T`tfhn\_dZrzmYAfhYUmp|[]fgpgY39YTa`tfgY
5


Force Sensor

A
1 mm Gap

2D channel flow

0.1 mm
Gap

A-A

F\_rbyZfhYORZ3bAFbA} I4 qT|A]A9
A

VF[]Zj_YO?ZhhAo})}?<A!  AT}'& e}CgCTAgA9| Al3T #+  8 }Tl3A T}?]blx ?T{ib|+


?}|b?|bAl 
UY5Zj_\_5[}pg\_`bd n[]^~Zj_Y5m CY5\lrtWtp
V`}
mg9YAY5 ^']m
cYAutYAj

p
b  t
n
n
t mp
Bzbt
Z ] U3 t U5  b b U5  b t U5 
1\o[bmXYAfhfh`bf|m
UY3m\onyi[bj
i`}  )
i t B
SYAdZY5fh[]pgY5<X[ubY5m  )  RZ U3  Z bt$  i ib
i z! B

& (

pgWZY[bphyi[]jfgyidR9mg`phWi[}p[]j_jRpgf|[]dIm\_YAdzpvYA9Y3phmS[]fhY!ZfhY5mgYAdzpv\_dpgWiYxpg\_^~YmgYAfh\lY3mAVXWiYxi[]fgpv`ba7pgWZYspg\_^YmY5fg\_Y5m
yimY3\ld)[ubY5fh[brb\_dZr<[bfgY~\onYAdzph\{iY3i[bmgY5`bdCpgWiYpg`}ubYAj_`nT\lpkbU\{phW[<ph\l^~YnY5j_[kafg`t^ T`tdimp|[]dzp!ubY5jl`nT\lpk
[tmv[b|Wi\lY5ubY5ph`<[tmgmgyZ^~Y5mpgY3[bnkmph[}phYxa`tfhAY5m`baq[}pSjlY3[bmpmgY5A`bdiZm5vifg`t^u"\omyI[]jc\_dimgIY3ph\l`td`]aqmgYAj_Y5TpgY3
a`bf|TYSpgf|[bAY5mXdZ`pgf|[]dIm\_YAdzphmXYAfhYS`bimgYAfhubY3e9YAkb`tdi<pgWZ\omXph\l^~YbiZynpdZ`ayZfgpgWZY5fv[]dI[]j_k"mg\om.YAfhY!n`bdiYb
 
"   &! !&'  ) I * dZjo[]didZ\ldirxpgWiYSYAnIY5fg\_^~YAdzphm.\{pXX[bm.Y5mpg\_^~[]pgY3pgWI[}ppgWZYSZfhY5mhmyZfhYvfhY5A`}ubY5fgk\ldpgWZY
[bbYX`]aR[!mg\ldZrtjlYmpgfhynpqmgWZ`byij_9YvwzyZ\_|YAdZ`tyZrbWmg`pgWi[]p0\{pn\_~di`]pqTfhY5[]pgY[Sa`tfhAY`tdpgWZYUphY5mp.mgY5ph\l`tdR>Zfh`b^
fgyZdImU\lpgWAY5fg`[bdZrbj_Y!\{p\_mmY5YAdpgWi[]ppgWiYAfhYx\_mv[em^[bjljRutYAfgpg\oA[bj?a`bf|TY!`bdphWZY!a`tfgX[]f|<mgYAdIm`tf5VXWZ\_m\_mv^`zmp
jl\_bY5jlknyZY~pg`phWZYeT`t^sZ\_dZY5YA9Y3px`]aX[]j_j7pgWifgY5Y~mpgfhynp|mA"\_diAY~pgWZYYA9Y3px`]aU[]dirbj_Y`baX[]pp|[b|\om!yZdiAYAfgph[]\_dR
\{p5[]dZdi`]pvmg\l^~Zj_k9YsmgyZnpgf|[bTpgY3afh`b^ pgWZYsfgY3myZjlphm5ZZyZpv\_m\ldiAjlyInY5[tmU[]dZ\o[bmUyZdITYAfgph[b\ldzpk<jl\_^~\{p|mA.pv`bdZj_k
\l d
IyZYAdiAY5mXa`tfhAY\ldpgWiY!Tn\lfhY5Tpg\_`bdR
Z`bfpgWiYkzZ\lfhY5Tpg\_`bd)\lp\omxa`tyZdi)phWi[}ppgWZY<ZfhY5mgYAdiAYe`]aUphWZY<mpgfhynphmWi[utY[^V[ U`tfs\_ d
iyZY5diTY<`bd)phWZY\_dZ\lpg\o[]j
ub`bfgpgYA<a`tfg^[]pg\_`bd`bdphWZYAk"jl\_dinY5fvdZ`zmYtVXWZ\_mvYT?Y5Tp\ommpgyiZ\lY3"kA`b^~i[]fh\_dZrfg`bph[]pg\_`bd\_dkt[pg`efhY5mgyZjlphm
afg`t^ fh`]ph[]pg\_dZr\ldZ\lph|W\_d<phWZYxn\omgAyimhm\_`bd`]apgWZY!fhY5mgyZjlphm5
VXWZYf|[]dirbY[]dI^~Y3n\_[bd`]a?pgWiYZ\o[bm7j_\l^~\lphm.[]fhYUrb\_ubYAd~\_d~ph[bZjlYSn>Z`bf0pgWZY^~`zmpqI[]fgp5]phWZYZ\o[bm0YAfhfg`tf7\_m0[bd`bf|nYAf
`ba^[]rtdZ\{phyinY!j_`}.Y5fpgWi[bd<phWZY!ZfhY5A\_mg\l`tdj_\l^~\lphm5
%XI H CZc"q 7

7g Z N|0 7 vZc""i


VXWZYxZkzdI[]^~\_SZfhY5mhmgyZfgYS\omUrb\_ubY5d"k
!   >
CBtt
va>pgWiY5mgYxi[]f|[]^~YApgYAf|mX`bdijlk<pgWZYm9YAY3\_mv^Y3[bmgyZfhY5i[bdi[]jomg`fgY5ZfgY3mY5dzpUpgWZYxjo[]fhrbY3mpyidiTY5fp|[]\_dtpk<U\{phWfgY3m9Y5Tppg`
Z\o[bm5Zmg`phWZYx[]dI[]j_k"mg\om.U\_j_jcT`tdiTY5dtphfh[]pgYS`bdpgWi\_m5
VXWZYspg`}m9YAY3\omv^~Y5[tmyifgY3<"k[eA`byZdzpgY5fv^~`byidtphY5`tdpgWZYA[bfgfh\_[brbYf|[]\_j_m5VXWZYsifgY3T\om\_`bdY5fgfh`bfU\omvrt\lutYAd"k
Y3wtyI[}pg\_`bd) ?tb[]dij_\_mpgY3a`bfxt mpv`tfhnY5fUfgY5Zj_\_5[}pg\_`bd\_dph[bZjlY ?ZUVXWZYsmgIY5Y5\omvmgjl\_rbWzphjlkn\{?YAfhYAdzpv`tda`bfh[bfh[]dI
i[t|"[bfhfgyZdZyZY!pg`~phWZYsmgIY5Y5A`bdzpgfh`bj`td<phWZYs5[]fhfg\o[]rtYbZmg`phWZYAfhYx\omXpg`eaYAZ[}p|[~I`t\ldzp|mXpg`e\_diAjlyiZYsn di`bf|nYAf
fhYAZj_\oA[}ph\l`td<WiYAfhYb
VXWZY!Z\o[bmXY5fgfh`bf|mX[]fhYb
 &+!   " '  )  * VXWZYeX[]\lpg\_dZrph\l^~Ye9YTpYAY5d)fgyZdImxX[bm!^~\_dZ\l^yZ^ B^~\ld"ynphY5m5>i[bmgY5`bd)YAnIY5fg\_YAdiAY[]dI
uz\omgyi[]jR`timgYAfhu[]pg\_`bd<`]aFpgWZY!X[ubY3m.\_dpgWZYp|[]dZ? dY3mph\l^[]pgY!`]aFpgWiY!jlY5ubYAj`baFfhY5mg\onyi[]j
I`}
A[bd<I[bmgY5`bd<a[]f
b

[bbY~pgWiYA`bfhkbVXWZY^Y3[]d)utYAj_`"A\{pknYT9T\lp[}pspgWZYAYAdzpgY5fs`baU[ij_[bdZY`tf}TX[]bY\omxrt\lutYAd"k 1jlY5uz\_dim
5bzn"R  RG]
  
3
C Bi3
 , Q

Z`bf[bd[}n\{mgkz^~^~YTphfg\oI`nnkpgWZY!fhYAjo[}ph\l`td<\om


3 S Q ?z
C Bzb

   
YAfhY \omvpgWZY^[}n\l^yZ^ubY5jl`nT\lpk\ldphWZY[bbY pgWZY~|Wi[]f|[bTpgY5fg\omph\_![bfgY3[~`bdUWi\_|W \omI[bmgY5?[]di 
pgWZYafhYAY!3 mpgfhY5[]^ ubYAj_`nT\lpkb7qwzyi[]pg\_`bdim BIS[bdi B"fg YAaYAf|mqph`[A`"`bf|n\ldI[}pgY!mgknmphYA^ ZnY3eph`phWZY!^~`}u"\ldZr9`"Zkb
"U\{p||WZ\_dZrdZ`}
pg`[~Zjo[]dZYsZnY5U\lpgWfhY5mgIY3pph`pgWiY
iyZ\o[bdidZ`bfh^[]jph`pgWiYn\_fgY3ph\l`td`]a7^~`]pg\_`bdc9[tmU\ld
T!vpUpgWZY5`bfhkb"pgWiYxn\{?yimg\l`td`]acphWZY!X[]bYS9Y5T`t^~Y5m[ayZdiTpg\_`bd`]aFpg\_^~Y!"kpgWZYxmgyZimpg\lpgyZpg\_`bd
_
 H  
C B,?t
3  , Q
 
  H

C B+Bz

%


S
Q
z
?

3
   
a`bf7Zjo[]dZYU[bdi~[}n\{mgkz^~^~YTphfg\o.[bbY5mfhY5mgIY3pg\_ubY5jlktdpgWi\_m>i"Y3sij_[bdZYpgWZYUutYAj_`"A\{pk 3 \omFpgWZY[tpgyI[]j"ubY5jl`nT\lpk
[}p7pgWiYTYAdzphYAf0j_\ldiYbbdi`]p7pgWiYUubYAj_`nT\lpksnYAIT\lp5zm\_diTYXphWZYU9`"ZkxZnY5~A`z`tfhZ\ldi[]pgYX^~`}ubY3m>U\{phW~[SutYAj_`"A\{pk  U\lpgW
fgY3m9Y5TpUpg`pgWZYi"Y3<mgknmphYA^0VF[]Zj_ Y Bj_\omp|mXpgWZY!fhY5mgyZjlpg\_dZrubYAj_`nT\lpka`bfXphWZYxTk"j_\ldInYAfx[}n\{mgkz^~^~YTphfg\o5[bdi<[
mphfgynpZjo[]dZYX[}p Hg BtmAVXWZYpg`}mgIY5Y5\_mvmYAppg`  Q ^ }m[bdi\{pv\om[tmgmgyZ^~Y5<pgWI[}pUpgWiYsn\l9yimg\_`bd\ld
pgWZYZnY3Zjo[]diY\omXdZ`]p[]9Y3pgY3"kphWZYSa[bTpphWZYSpgWZY^~`]ph\l`td<`bacphWZY!Tk"jl\_dinY5f[bTpgyi[bjlj_kWi[bmXmpg`tZIY3<[]pm`t^Y
pg\_^Y H ! Bti
VF[]Zj_OY Bi . ~]|h }i] }T_5A # 3  |T   InT

VcY3mpS.k"j_\ldiZYAf i i     Q ?t  > Z b3z
zpgfhynp il
Z ZS^
Z b+B


VXWZYfgY3myij{pa`bfphWZYempgfhynpx\_m!pgWZY5dyImY3\_dph[]ijl'Y ?i~VXWiYAfhY\omx[bj_mg`phWZY9`tmhm\_Z\_jl\lpk`ba.fhY5mg\_nyI[]j
i`}nyZY~pg`
fgY5^~[b\ldi\ldZr~X[ubY3mA"Zynpvmg\_diTYSpgWiY5mgYxT`byij_<dZ`]p9Y!`bImY5fgutY5phWZ\omU\_mXdiYArbj_Y5TpgY3
$      &'  )  * VXWZYnYAnphW)[}pxUWZ\o|WphWZYeTk"j_\ldInYAf!\omSpg`}Y5C\_m![<phfh[tnY~`]aXIYAp.Y5YAdfg\_rb\on\lpk`]a.pgWZY
^`nnY5j[]di~phWZYS\l d
IyZYAdiAY`bapgWZYafhYAYmgyZfga[bAYbFdepgWi\_mXA[tmYpgWZYTk"j_\ldiZYAf[tm7pg`}Y5[}p[nYAZpgW`baZ Bx^"`bf
i0VXWZY![b^ijl\lpgyiZY!`]a>ubYAj_`nT\lpk[}p[rb\_ubY5dZYAnphW\om
Y
C Bzb
    
UWZYAfhY 
Q B^\_mphWZYpg`}MnY5npgWR   \_mphWZYT\_fhAyZj_[bfsafhY5wzyZY5diTk)`bapgWiYX[ubYt0[]di   \ompgWZY
[]^~Zj_\{phyinYS`]apgWiYX[ubYUY5jlY5u[]pg\_`bd[]pqphWZYmgyZfga[bTYtFVXWiYvYTnZfhY5mhmg\l`tdyimY3m.[tmgmgyZ^~Y5mq\ldZidZ\lpgYS[]pgYAfnYAZpgWR0VXWZY
[utYvd"yZ^s9YAfX\omrb\_ubYAdzk   Z >     UWZY5fgYvpgWZY^[}eX[ubYvjlY5dZr]phW\om  ]  a`bfpgf|[]dimgubY5fhmgYX[ubY3mq[]dI
  a`tfSZ\lutYAfhrbYAdzpX[ubY3mxI[]jlpg\_dimY5dC]t B" \_rbW[utY!dzyi^s9YAf|m^~Y5[bdimmgj_`}.Y5fvnY55[k<U\lpgWCnYAnphWR
m`~phWZY!I pgf| [] dImutYAf|mYS[utY5mXmgWZ`byZjoIYTfh\{ph\_5[]jdpgWZY!`bpgWZY5fWi[]dInpgWiYx`bdZj_kX[ubY3m.pgWi[]pT`tyZj_9Y!`bimgYAfhubY3
nyZfh\ldZrphWZY!YTn9YAfh\l^~YAdzpYAfhYSZ\lutYAfhrbYAdzpUX[ubY5mqafh`b^ phWZYxmphfgyZp50VXWZYxifg`tyinYd"yZ^s9YAfXa`tfXpgWZYxmpgfhynphmX\om!t + B
a`bf + Q '^ ]m!UWZ\o|W\_ms\ld)[f|[]dZrtY~UWZYAfhY~pgWZY<n\lutYAfhrbY5dtp!X[ubY3m![]fhY~Z`b^~\ldI[]dzp<I[]jlpg\_dimgYAd ]t B"2 1k
`bimgYAfhu"\ldZrspgWZYSX[bmgW`bdepgWZYSph[bdZX[]j_j_m5t\lpU[tm.A`bdiAjlyInY5epgWi[]p.phWZY[utYv[]^~Zj_\{phyinYdZY5ubYAfYTZAYAY5ZY5<sT^
VXWZY!Z\o[bmXj_\l^~\lpv\_mphWZYAdY3mph\l^[}phY5"k?
  Q ts^
C B,Rt
?

  >
C BG]
 )
 ?    Y
C Btt
3


VXWZYxmgY5A`bdi<i[bfh[b^YApgY5f\_dpgWZYnk"di[]^~\oifgY3mgmgyZfhYS\ompgWiYx[]pgY5fXZYAdimg\{pktqVXWZYxnY5dimg\{pk[bdiu"\_mhT`zm\lpke`]apgWZYs[]pgYAf
\_dpgWZYsph[bdZyimgY5a`tf5[]joTyZjo[}ph\ldZreZkzdI[]^~\_xZfgY3mgmgyZfhYx[]diYAk"dZ`bjoZmUd"yZ^IY5f[bfgY!rt\lutYAd\ldp|[]Zj_Ye tXVXWiY5mgYxu}[]j_yZY5m
[bfgY~i[tmY3C`td)[<^~Y3[bmgyZfgY3[]pgYAfphYA^~9YAf|[}pgyifgY`baS3 Q   []diCph[bbY5dCafh`b^ph[]ijlY3m!yimgY5[}p ' UvVU @~VXWZY
u}[]fh\o[}pg\_`bd`]a?phWZYXpgY5^9YAf|[}phyZfhY\ldphWZYUph[bdZs\om0utYAfhkxj_`}!tmg`pgWZYmh[]^~YXpgY5^~IY5fh[]pgyZfhYU\_mqyimgY5a`bfq[bjljIZ[knmA>VXWZYUZ\o[bmgY5m
[bfgYt
TvY5dim\lpk
 ) Z_


Tvk"di[b^\o7 cv\omgA`tmg\{pk n U5  
VXWZYZkzdI[]^~\_u"\omgA`tmg\{pk)\om`tdZjlkyimgY5UWZYAd5[]joTyZjo[}phY5UYAk"dZ`tj_Zmedzyi^s9YAf3[bdi\_mdi`]pe\_diTj_yinY3\_dpgWZYYAfhfg`tf
Y3mph\l^[}phYbU
%X  CZc"q 7

7g Z N|0  R.vc.)>vgv"q


%$ $ VXWZYUkt[[]dZrtjlYXX[bm>mYAp0^[]d"yi[]j_j_k!yim\_dZr![SZnY5Zfh`]phfh[tpg`tf5VXWZY[]dZrtjlYX[bm7mgYTp7"k!pgWZYmh[]^~YX9YAf|m`td
[]j_jcZ[kbnY5dimgyZfg\_dZrpgWi[]pUpgWZYxmh[]^~YfhYTaYAfhYAdITY[tmXyimgY5<Y5[b|W<pg\_^~Yb0VXWZY!i\_[tmXjl\_^~\{pvX[bm.phWZYAdY5mpg\_^[}phY5<pg`
 ) S Q b 
C Btt

UyZdZdZ\_dZr9`]pgWn\_fgY3ph\l`tdimSU\{phWC5YAfh`<[]dirbj_Y`ba[}pgph[t|fhYAutY5[bjlY3dZ`n\l9Y5fgY5diTY\ldTn\_fgY3ph\l`tda`bf|TYpgW"yimdZ`
YAfhfg`tfF\_d~Z\lph|W[]j_\lrtdZ^~YAdzpqT`tyZjos9YU`bimgYAfhubY3Fn\ldiAYmyI|Wsa`tfhAY5m>^[kx9YUubYAfhksmg^[]j_jb[ifgY3T\om\_`bdYAfhfg`tfFY3wzyi[]j
pg`~pgWiY!kt[ ZfhY5A\_mg\_`bd[]fhYS[bZZj_\lY3[bd"kzX[kb
 "  $&+#"%$ & '<)
* jlj>n\_^Y5dimg\l`tdim[]fhY!bY5npA`bdimph[bdtpvZyZfg\_dZre[bdi[]j_knm\omAZpgW"yimUphWZYAfhYxmgWZ`byZjo9YxdZ`ZfhYT
T\om\_`bdYAfhfh`bf|mA>VXWZY!j_YAdZrbpgW[]din\o[]^~YApgYAfU`baphWZY!phY5mpvmY3pg\_`bdYAfhYS^~Y3[bmgyZfgY3eph`e[]d[b5TyZf|[bAke`]a S Q s^~^
VXWZYAd
 )
Q }^~^
]t
%X % z
Svcv? JCX  qZ
VXWZY!\_ d
iyiYAdiAYA`"Y eT\_YAdzp9Y5T`t^~Y5m
@
> A
n3
: >
UWZY5fgY : > \_mpgWZYi[]f|[]^~YTphYAf|mX\ldY5wzyi[}ph\l`tdC ?,?t+ 1YAj_`}!ZphWZA YsI[]f|[]^~YTphYAf|mU[]fhYx\ldInYT"kphWZY[bjlZWI[]9YTpvf|[}phWZYAfUphWi[]d
d"yZ^IY5fhm5"pg`mgYAi[bfh[]pgYSafg`t^MYAfhfh`bfUm`tyZf|TYS\ldiZYTnY5m50VXWZY!mY5dimg\{ph\lu"\lpkpg`u}[]fh\o[}pg\_`bd<\ld[~mg\ldZrtjlYxmgYAdIm`tf\om
 Z 



Z A
bb




3
>   >   

3
>   >  

[bdi<mg\_^\_jo[]fXa`bfS  Z 0i`bfUY5fgfh`bfX\_d<phWZY!ubY5jl`nA\{pk~phWZYxmY5dimg\{ph\lu"\lpk\om
Z
 Z
q A  3   >    <v 3       
?t
I
A
>
>

a`tfUI`bpgWdZ`bfh^[]j?a`tfhAY5mX[]di^~`b^~Y5dtp|mA7VXWZYxmgYAdIm\lpg\_u"\{pkph`u}[]fh\_[]pg\_`bd\_dnYAdIm\lpke\om
Z
 Z
Wq A  3   >     < 3   >  >    
!Bz
A
>
>

[bdieIdi[]j_jlka`bfUu}[]fh\o[}pg\_`bd<\ldn\o[]^~YApgYAf
Z
 Z 
3
 < 3   
bb
 A
A
> O>  
>  > > 

[bdi<mg\_^\_jo[]fXa`bfUu}[bfg\o[}ph\l`td<\_djlY5dZr]phWR
VXWZYs[bdZrbj_Y!`]a0[]pph[t| & \_mUdi`]pv\_diTj_yinY3\ldphWZYxfhY5nyIpg\_`bdY3wzyi[}ph\l`tdimAIm`~phWZYsmgYAdimg\{ph\lu"\lpk<5[]dZdZ`bpv9YxYTnZfhY5mhmY3
phWZ\_m[kt 1.yZpm\_diAYpgWZYx[bdZrbj_Y`]a7[}pp|[b|<\_mUu}[bfg\_Y5"phWZYxfhYAjo[}ph\l`tdimWi\lIYAp.Y5YAdpgWiYa`bf|TYx[]dIphWZYx[]dirbj_YS`ba0[}pp|[b|
\omq"dZ`}Udeafg`t^ pgWZYSYTn9YAfh\l^~YAdzp3>VXWi\_m.`]aA`byZf|mgY\_m.phWZYfhY5[tm`td~a`bfXn`b\_dZrspgWZYSYTn9YAfh\l^~YAdzp|mq\_dephWZYvIfhmpij_[tTYb7VXWZY
mgYAdimg\lpg\_uz\lpk ]SA[]d<pgWiYAfhYTa`bfhYIY!Y3mph\l^[]pgY5"k[d"yZ^~YAfh\_5[]jRnY5fg\_u}[}ph\l`td<`baphWZY!^~Y5[bmgyZfhY5<fgY3myij{p|mAnU\lpgWfhY5mgIY3pUpg`
phWZY!rb\_ubY5d[bdZrbj_Y`ba>[]pp|[b|?
?


X 7n7scZZ0ctZF"
V`[ut`b\oen\_u"\_mg\l`td"kAY5fg`UWZYAdA[bj_AyZjo[}pg\_dZr[sfhYAjo[}ph\lutYvYAfhfh`bf3t`tdZjlk'&
[bfgYS\_diTj_yinY3e\_dphWZYSa`bj_jl`}U\_dZrsfhY5mgyZjlphm
F\_rbyZfhYbSmWi`}mFpgWZYfhYAjo[}ph\lutYXYAfhfg`tfhma`bf R5[]joTyZjo[}phY5i[tmY3s`td~pgWZYXa`tfgX[]f|sfhyZdim7pgWZY ?i f|~Z[kb>VXWZ\om7fhYAZfhY5mgYAdzphm7[
t mp.`tfhnY5fqfhYAZj_\oA[}ph\l`tde[]Zifg`z[b|WRFVXWZYvi[bfhm.rb\_ubY3m7phWZYA`bdzpgfh\liynpg\_`bdimqafh`b^pgWZYm`tyZf|TYvnY3mgAfg\_9Y5e[bI`}utYb>VXWZYvi[bfp
`baqpgWiYi[]f|mSfgY5ZfgY3mY5dzpg\_dZra`tfhAY[bdink"di[]^~\oifgY3mgmgyZfhY #}  gT  \ldiAjlyInY5m9`]phWifgY3T\om\_`bd[]diCZ\o[bmYAfhfh`bf|m
a`tf!pgWZY3mYm`tyZf|TY5m5<VXWiY<mg`bj_\_j_\ldZYrb\_ubY5m!pgWZYpg`]p|[]j.YAfhfg`tfafg`t^ Y5wzyi[]pg\_`bd bbTcphWZY<wzyi[tnf|[}pg\omgyZ^'`baXpgWZY<YAfhfg`tf
A`bdzpgfh\liynpg\_`bdim5
%



1 3"# )  %) $

1 3"#   %) $

Q 
#C* 
 D&
 )  C
 () 

45

 



   
 !
"
 

45
40
35

40
35
30

30




25

25

20

20

15

15

10

10

0
1

8 10 12 16 20 25 30
  



1 > # )  %) $

8 10 12 16 20 25 30
 

1 > #   %) $

45

40

40

35

35

30

30
25




25

20

20
15

15
10

10

0
1

8 10 12 16 20 25
 

8 10 12 16 20 25
 

F\_rbyZfhY " vAl] }Ag]|x|}T|} (  X]gbA!g9|]])]9] #o  0" ]lTn_}}o T]


%&

^>

F\lrtyZfhYsrb\_ubY3mXpgWiYsfhYAjo[}pg\_ubY!Y5fgfh`bf|mafh`b^pgWZY~n di`bf|nY5ffgY5Zj_\_5[}pg\_`bdcdphWZ\om5[bmgYs[eT`"Y A\lY5dzpv\_mA[bj_AyZj_[]pgY3


a`tfvY5[t|WfgyidR[]diphWZYsifgY3T\om\_`bdY5fgfh`bf\omi[bmgY5`tdpgWZYu}[]fh\_[]pg\_`bd`]a7phWZY5mgYsu}[bjlyZY3mAS"\{fhYA9YTph\{ph\l`tdim[bfgY[bmhmyZ^~Y3
a`tfv[]j_j[bdZrbj_Y5m5 1kfgY5IY3[}ph\ldZr~phWZY!phY5mpmY5fg\_Y5mYAfhfg`tfhmUnyiY!pg`eu}[bfg\o[}ph\l`td`ba>pgWZYsmgYTpS[]dZrtjlYx[bdipgWZYsfgY3m\onyi[bjRubY5jl`nT\lpk
U\_jljF9Ys\_diTj_yinY3\ldpgWZYZfgY3T\om\_`bdi[bfp39mg2`  ) []di  )
s[]fhYxdZ`bp\_diTj_yinY3\_dpgWZYZ\o[bmY5fgfh`bfUY3mph\l^[]pgY5m5vVXWZ\om\_m
phWZY^[b\ldCfgY3[bmg`bda`bffhYA9Y5[]pg\_dZrephWZYYAnIY5fg\_^~YAdz phm5!`]phYZ\{?YAfhYAdiAY\_dj_YArtYAdiIYAp.Y5YAdpgWZYp.`<Zj_`]phm5!Z`tf!Z dI
`tfhnY5fvfhYAZj_\oA[}ph\l`tdpgWZY5fgY\omv`tdZjlk`tdZYsZfhY5A\_mg\_`bdYAfhfh`bfv[bdi\lp\omSrb\_ubYAdmgYAI[]f|[}pgY5jlktVXWZYfgY5^[]\_dZ\ldir<mg`byZf|TY3mv[]fhYs[bjlj
Z\o[bmXY5fgfh`bf|m5
.`b^~i[bfg\_dZrpgWZYp.`<^~YTpgWi`"imA?phWZY<t mp`bf|nYAffgY5Zjl\oA[]pg\_`bdrt\lutY5mWZ\_rbWY5fgfh`bf|mXa`tfSj_`}%[]dirbj_Y5mS[]di[<j_\{pgpgj_Yj_`}
Y5fgfh`bf|mSa`bf^~Y5n\_yZ^[]diWZ\lrtW[]dZrtjlY3msA`b^~i[]fhY5ph`Cn di)`tfhZYAfxfhYAZj_\oA[}ph\l`tdR
7[bfph\_AyZj_[bfgj_k[]pspgWiYj_`}.Y3mp[bdZrbj_Y5m
\lpmgYAY5^~mxpgWi[]pxpgWiYYAfhfh`bfxnyZYpg`u}[bfg\o[}ph\l`td\_d [bdZrbj_Y`]av[]pph[t|\omxj_`}.Y5f!pgWI[]dpgWiYY5mpg\_^[}phYe\ld  )
i`bfxWi\lrtWZYAf
[bdZrbj_Y5m0pgWZYST`b^~i[bfg\omg`bd~\_m.fhY5[bmg`bdI[]Zj_YbtUWZ\_|WWZ`b9YTayZj_j_k\_din\oA[]pgY3m0phWi[}pdZ`sjo[]fhrbYUY5fgfh`bfqmg`byifhAY5mq[bfgYv`b^~\{pgpgY3\ldepgWZY
[bdi[]j_knm\omA
B

1 3"# )  %) $

1 3"#   %) $

18

18

16

16

14

14

12


12


10

Q 
D& 
#C* 
 D&
 )  C

 
   



   

6
4
2

6
4
2

0
1

8 10 12 16 20 25 30
  



1 > # )  %) $

8 10 12 16 20 25 30
 

14

14

12

12

10

10

1 > #   %) $

16



10



6
6
4

2
0

0
1

8 10 12 16 20 25
 

8 10 12 16 20 25
 

\_rbyifgY!Z T_} Th}|s|A|}   ? <]gbA!g9|}}]9} #|o 0" }_Tnl]]Co T]




&

^>

P N

! &%
#" +

 
5ttbT .b?| qn) #}9!T}?b|5 } @fg\_YArtYAf
0yiZjl\omgWZ\ldire.`t^~i[]d"kb
  "! #"$%&! ('*)+-,/.0   5bttT BTZA TI]]]921?|T} I # ?} #o T]
] ?|T|3b`td

\lj_YAk[]di"`bdImA
45!6.0
7 ##89 :2F5tbz $?|<;c3b}x] $ " I]9 @>=Ui}g$IiAngTi[]^sifg\onrbY?di\lutYAf|m\lpk
0fhY5mhm5
45!6.0
7 ##89 :2F]tB" knnfh`nnk"di[]^~\oAm`]a>WZ\_rbWnmgIY5Y5<^~[bfg\_dZYubY5WZ\oTj_Y5m5 RY5TpgyZfhYdZ`]phY5m5
dzpgYAfhdi[]pg\_`bdi[bj"V`}U\_dZrSVF[]di.`bdnaY5fgY5diTY3btt i}7 X Z U}?T?}A @I}t}}' g3|h]ngT|bdtphYAfhdi[}ph\l`tdi[]j
Vc`}U\_dZrV[bdZ.`tdnaYAfhYAdiAYb3Z pgWR
BDC0 E GFH
JIKL<c5tbtT v ]9|| ] ?|T   } Z }< RY5 M 3t`bd\_jlY5k<[]dI"`tdim5


&

Annex E

ITTC Recommended
Procedures and Guidelines
Performance, Propulsion
1978 ITTC Performance Prediction
Method

7.5 02
03 01.4
Page 1 of 10
Effective Date
2008

Revision
02

Table of Contents
1.

PURPOSE OF PROCEDURE

2.

DESCRIPTION OF PROCEDURE

2.1 Modifications of the 1978 ITTC


Performance Prediction Method.........2
2.2 Definition of the Variables...................2
2.3 Analysis of the Model Test Results .....3
2.4 Full Scale Predictions ...........................4
2.4.1 Total Resistance of Ship ............... 4

Edited by 22nd ITTC QS Group 1999


15th ITTC 1978 pp388 402
17th ITTC 1984 pp326 - 333
18th ITTC 1987 pp266 - 273
Date

2.4.2 Scale Effect Corrections for


Propeller Characteristics. ........................ 7
2.4.3 Full Scale Wake and Operating
Condition of Propeller .............................. 7
2.4.4 Model-Ship Correlation Factor ...... 8

3.

VALIDATION

3.1 Uncertainty Analysis ............................9


3.2 Comparison with Full Scale Results ...9
4. REFERENCES

Approved
15th ITTC 1978, 17th ITTC 1984
th
and 18 ITTC 1987
Date

ITTC Recommended
Procedures and Guidelines
Performance, Propulsion
1978 ITTC Performance Prediction
Method

7.5 02
03 01.4
Page 2 of 10
Effective Date
2008

Revision
02

1978 ITTC Performance Prediction Method


1.

PURPOSE OF PROCEDURE

The procedure gives a general description


of an analytical method to predict delivered
power and rate of revolutions for single and
twin screw ships from model results.

CF
CFC
CNP
CP

2.

DESCRIPTION OF PROCEDURE
CN

2.1 Introduction
The method requires respective results of
a resistance test, a self propulsion test and the
characteristics of the model propeller used
during the self propulsion test,
The method generally is based on thrust
identity which is recommended to be used to
predict the performance of a ship. It is supposed that the thrust deduction factor and the
relative rotative efficiency calculated for the
model remain the same for the full scale ship
whereas on all other coefficients corrections
for scale effects are applied.
In some special cases torque identity
(power identity) may be used, see section
2.4.4.

CR
CT
D
FD
J
JT
JQ
KT
KQ
KQT
k
kP
NP
n
nT

2.2 Definition of the Variables


CA
CAA
CApp
CD

Correlation allowance
Air resistance coefficient
Appendage resistance coefficient
Drag coefficient

P
PD, PP
PDT

Frictional resistance coefficient


Frictional resistance coefficient
at the temperature of the self
propulsion test
Trial correction for propeller
rate of revolution at power
identity
Trial correction for delivered
power
Trial correction for propeller
rate of revolution at speed
identity
Residual resistance coefficient
Total resistance coefficient
Propeller diameter
Skin friction correction in self
propulsion test
Propeller advance coefficient
Propeller advance coefficient
achieved by thrust identity
Propeller advance coefficient
achieved by torque identity
Thrust coefficient
Torque coefficient
Torque coefficient achieved by
thrust identity
Form factor
Propeller blade roughness
Number of propellers
Propeller rate of revolution
Propeller rate of revolution,
corrected using correlation factor
Propeller pitch
Delivered Power, propeller
power
Delivered Power, corrected
using correlation factor

7.5 02
03 01.4
Page 3 of 10

ITTC Recommended
Procedures and Guidelines
Performance, Propulsion
1978 ITTC Performance Prediction
Method

PE, PR
Q
RC

Re
RT
S
SBK
T
t
V
VA
w
wQ
wR
wT
Z

CF
CFC
wC
D
H
0
R

Effective power, resistance


power
Torque
Resistance corrected for temperature differences between
resistance- and self propulsion
test
Reynolds number
Total resistance
Wetted surface
Wetted surface of bilge keels
Propeller thrust
Thrust deduction factor
Ship speed
Propeller advance speed
Taylor wake fraction in general
Taylor wake fraction, torque
identity
Effect of the rudder(s) on the
wake fraction
Taylor wake fraction, thrust
identity
Number of propeller blades
Appendage scale effect factor
roughness allowance
Individual correction term for
roughness allowance
Individual correction term for
wake
Propulsive efficiency or quasipropulsive coefficient
Hull efficiency
Propeller open water efficiency
Relative rotative efficiency
Water density in general

Subscript M signifies the model


Subscript S signifies the full scale ship
2.3 Analysis of the Model Test Results
The calculation of the residual resistance
coefficient CR from the model resistance test

Effective Date
2008

Revision
02

results is found in the procedure for resistance


test (7.5-02-02-01).
Thrust TM, and torque QM, measured in
the self-propulsion tests are expressed in the
non-dimensional forms as in the procedure for
propulsion test (7.5-02-03-01.1).

KTM =

TM
M DM4 nM2

and

KQM =

QM
M DM5 nM2

Using thrust identity with KTM as input


data, JTM and KQTM are read off from the model
propeller open water diagram, and the wake
fraction

wTM = 1

J TM D M
VM

and the relative rotative efficiency

R =

KQTM
KQM

are calculated. VM is model speed.


Using torque identity with KQM as input
data, JQM is read off from the model propeller
open water diagram, and the wake fraction

wQM = 1

J QM D M
VM

VM is model speed.
In case of using torque identity the relative
rotative efficiency

R = 1.0
The thrust deduction is obtained from

ITTC Recommended
Procedures and Guidelines
Performance, Propulsion
1978 ITTC Performance Prediction
Method

t=

TM + FD RC
TM

where FD is the towing force actually applied


in the propulsion test. RC is the resistance
corrected for differences in temperature between resistance and self-propulsion tests:
RC =

(1 + k ).CFMC + CR R
(1 + k ).CFM + CR TM

where CFMC is the frictional resistance coefficient at the temperature of the self-propulsion
test.
2.4 Full Scale Predictions

2.4.1 Total Resistance of Ship


The total resistance coefficient of a ship
without bilge keels is
CTS = (1 + k )CFS + CF + CA + CR + CAAS

is the form factor determined from the


resistance test, see ITTC standard procedure 7.5-02-02-01.

- CFS

is the frictional resistance coefficient


of the ship according to the ITTC1957 model-ship correlation line

- CR

is the residual resistance coefficient


calculated from the total and frictional
resistance coefficients of the model in
the resistance tests:
CR = CTM (1 + k )CFM

The form factor k and the total resistance coefficient for the model CTM

Effective Date
2008

Revision
02

are determined as described in the


ITTC standard procedure 7.5-02-02-01.
The correlation factor for the calculation of
the resistance has been separated from the
roughness allowance. The roughness allowance CF per definition describes the effect of
the roughness of the hull on the resistance.
The correlation factor CA is supposed to allow
for all effects not covered by the prediction
method, mainly uncertainties of the tests and
the prediction method itself and the assumptions made for the prediction method. The
separation of CF from CA was proposed by
the Performance Prediction Committee of the
19th ITTC. This is essential to allow for the
effects of newly developed hull coating systems.
The 19th ITTC also proposed a modified formula for CA that excludes roughness allowance, which is now given in this procedure.
- CF is the roughness allowance
k 3

1
C F = 0.044 S 10 Re 3 + 0.000125
LWL

where
-k

7.5 02
03 01.4
Page 4 of 10

where kS indicates the roughness of


hull surface. When there is no measured data, the standard value of
kS=15010-6 m can be used.
- CA

is the correlation allowance.


CA is determined from comparison of
model and full scale trial results.
When using the roughness allowance
as above, the 19th ITTC recommended
using

CA = (5.68 0.6 log Re) 103


to give values of CF+CA that approximates the values of CF of the
original 1978 ITTC method. It is recommended that each institution main-

ITTC Recommended
Procedures and Guidelines
Performance, Propulsion
1978 ITTC Performance Prediction
Method

tains their own model-full scale correlation. See section 2.4.4 for a further
discussion on correlation.

7.5 02
03 01.4
Page 5 of 10
Effective Date
2008

Revision
02

scale. The following alternative methods are


well established:
1) Scaling using a fixed fraction:

- CAAS is the air resistance coefficient in full


scale
CAAS =

1
A
AVS2CDA VS
2
SS

where, AVS is the projected area of the


ship above the water line to the transverse plane, SS is the wetted surface
area of the ship, A is the air density,
and CDA is the air drag coefficient of
the ship above the water line. CDA can
be determined by wind tunnel model
tests or calculations. Values of CDA are
typically in the range 0.5-1.0, where
0.8 can be used as a default value.
If the ship is fitted with bilge keels of
modest size, the total resistance is estimated
as follows:
CTS =

SS + S BK
[(1 + k )CFS + CF + CA ] + CR + CAAS
SS

where SBK is the wetted surface area of the


bilge keels.
When the model appendage resistance is
separated from the total model resistance, as
described as an option in the ITTC Standard
Procedure 7.5-02-02-01, the full scale appendage resistance needs to be added, and the
formula for total resistance (with bilge keels)
becomes:
CTS =

SS + S BK
[(1 + k )CFS + CF + CA ] + CR + CAAS
SS

+ CAppS

There is not only one recommended


method of scaling appendage resistance to full

CAppS = (1 ) CAppM
where (1-) is a constant in the range
0.6-1.0.
2) Calculating the drag of each appendage separately, using local Reynolds
number and form factor.
n

C AppS = (1 wi ) 2 (1 + ki ) C FSi
i =1

Si
SS

where index i refers to the number of


the individual appendices. wi is the
wake fraction at the position of appendage i. ki is the form factor of appendage i. CFSi is the frictional resistance coefficient of appendage i, and Si
is the wetted surface area of appendage
i. Note that the method is not scaling
the model appendage drag, but calculating the full scale appendage drag.
The model appendage drag, if known
from model tests, can be used for the
determination of e.g. the wake fractions wi. Values of the form factor ki
can be found from published data for
generic shapes, see for instance Hoerner (1965) or Kirkman and Kletsli
(1980).

ITTC Recommended
Procedures and Guidelines
Performance, Propulsion
1978 ITTC Performance Prediction
Method

7.5 02
03 01.4
Page 6 of 10
Effective Date
2008

Revision
02

ITTC Recommended
Procedures and Guidelines
Performance, Propulsion
1978 ITTC Performance Prediction
Method

2.4.2 Scale Effect Corrections for Propeller


Characteristics.
The characteristics of the full-scale propeller are calculated from the model characteristics as follows:

7.5 02
03 01.4
Page 7 of 10
Effective Date
2008

value of which is set kP=3010-6 m. Rec0 must


not be lower than 2105.
2.4.3 Full Scale Wake and Operating Condition of Propeller
The full-scale wake is calculated by the
following formula using the model wake fraction wTM, and the thrust deduction fraction t
obtained as the analysed results of selfpropulsion test:

KTS = KTM KT
KQS = KQM KQ
where

wTS = (t + wR ) + ( wTM t wR )

P cZ
KT = CD 0.3
D D
KQ = CD 0.25

cZ
D

CD = CDM CDS
where

5
t 0.044

C DM = 2 1 + 2
1
2
c (Rec 0 )6 (Rec 0 )3

and

t
c

= 2 1 + 2 1.89 + 1.62 log


c
kP

(1 + k )C FS + C F
(1 + k )C FM

where wR stands for the effect of rudder on


the wake fraction. If there is no estimate for
wR, the standard value of 0.04 can be used.

The difference in drag coefficient CD is

CDS

Revision
02

2.5

In the formulae listed above c is the chord


length, t is the maximum thickness, P/D is the
pitch ratio and Rec0 is the local Reynolds
number with Kempfs definition at the openwater test. They are defined for the representative blade section, such as at r/R=0.75. kP
denotes the blade roughness, the standard

If the estimated wTS is greater than wTM,


wTS should be set as wTM.
The wake scale effect of twin screw ships
with open sterns is usually small, and for such
ships it is common to assume wTS = wTM.
For twin skeg like stern shapes a wake correction is recommended. A correction like the
one used for single screw ships may be used.
The load of the full-scale propeller is obtained from
KT
S
CTS
= S2
2
J
2 DS (1 t ) (1 wTS ) 2 N P

where NP is the number of propellers.


With this K T / J 2 as input value the full
scale advance coefficient JTS and the torque
coefficient KQTS are read off from the full scale

7.5 02
03 01.4
Page 8 of 10

ITTC Recommended
Procedures and Guidelines
Performance, Propulsion
1978 ITTC Performance Prediction
Method

propeller characteristics and the following


quantities are calculated.
- the rate of revolutions:
(1 wTS ) VS
nS =
J TS DS

(r/s)

- the delivered power of each propeller:


K
(kW)
PDS = 2 S DS5nS3 QTS 10 3

- the thrust of each propeller:


K
TS = T2 J T2S S DS4nS2
J

Effective Date
2008

Revision
02

In the following, several different alternative concepts of correlation factors are presented as suggestions. It is left to each member organisations to derive their own values of
the correlation factor(s), taking into account
also the actual value used for CA.
(1) Prediction of full scale rates of revolutions
and delivered power by use of the CP - CN
correction factors
Using CP and CN the finally predicted trial
data will be calculated from

nT = C N nS

(N)

(r/s)

for the rates of revolutions and


- the torque of each propeller:
K QTS
QS =
S DS5 nS2

PDT = CP PDS
(Nm)

(kW)

for the delivered power.

the effective power:


1
PE = CTS SVS3SS 10 3 (kW)
2

(2) Prediction of full scale rates of revolutions


and delivered power by use of CFC - wC
corrections

the total efficiency:


N P
D = P DS
PE

In such a case the finally trial predicted


trial data are calculated as follows:

the hull efficiency:


1 t
H =
1 wTS

2.4.4 Model-Ship Correlation Factor


The model-ship correlation factor should
be based on systematic comparison between
full scale trial results and predictions from
model scale tests. Thus, it is a correction for
any systematic errors in model test and powering prediction procedures, including any
facility bias.

KT
S
CTS + CFC
= S2
2
J
2 DS (1 t ) (1 wTS + wC ) 2 N P

With this KT/J as input value, JTS and KQTS


are read off from the full scale propeller characteristics and the following is calculated:
(1 wTS + wC ) VS
(r/s)
J TS DS
K
= 2 S DS5nT3 QTS 10 3
R

nT =
PDT

(kW)

ITTC Recommended
Procedures and Guidelines
Performance, Propulsion
1978 ITTC Performance Prediction
Method

At the point of KT-(J)-Identity the condition is


reached where the ratio between the propeller
induced velocity and the entrance velocity is
the same for the model and the full scale ship.
Ignoring the small scale effect KT on the
thrust coefficient KT it follows that J-identity
correspond to KT- and CT-identity. As a consequence it follows that for this condition the
axial flow field in the vicinity of the propeller
is on average correctly simulated in the model
experiment. Also the axial flow of the propeller slip stream is on average correctly simulated. Due to the scale effects on the propeller
blade friction, which affect primarily the
torque, the point of KQ-identity (power identity) represents a slightly less heavily loaded
propeller than at J-, KT- and CT-identity. At
the power identity the average rotation in the
slipstream corresponds to that of the actual
ship and this condition is regarded as important if tests on stator fins and/or rudders are to
be done correctly.
In this case, the shaft rate of revolutions is
predicted on the basis of power identity as
follows:
KQ
1000 CP PDS
3 =
2 3
3
J T 2 S DSVS (1 wTS )
KQ 0
J3

nS =

K
= Q3 RM
J T

(1 wTS ) VS
J TS DS

Effective Date
2008

Revision
02

nT = CNP nS

(3) Prediction of full scale rates of revolutions


and delivered power by use of a CNP correction
For prediction with emphasis on stator fins
and rudder effects, it is sometimes recommended to use power identity for the prediction of full scale rates of revolution.

7.5 02
03 01.4
Page 9 of 10

3.

VALIDATION

3.1 Uncertainty Analysis

Not yet available


3.2 Comparison with Full Scale Results

The data that led to 1978 ITTC performance prediction method can be found in the
following ITTC proceedings:
(1) Proposed Performance Prediction Factors
for Single Screw Ocean Going Ships
(13th 1972 pp.155-180) Empirical Power
Prediction Factor ( 1+X )
(2) Propeller Dynamics Comparative Tests
(13th 1972 pp.445-446 )
(3) Comparative Calculations with the ITTC
Trial Prediction Test Programme
(14th 1975 Vol.3 pp.548-553)
(4) Factors Affecting Model Ship Correlation
(17th 1984 Vol.1 pp274-291)

4. REFERENCES

(1) Hoerner, S.F. (1965) Fluid-Dynamic


Drag. Published by the author.
(2) Kirkman, K.L., Kletsli, J.W. (1980)
Scaling Problems of model appendages, 19th American Towing Tank Conference, Ann Arbor, Michigan

ITTC Recommended
Procedures and Guidelines
Performance, Propulsion
1978 ITTC Performance Prediction
Method

7.5 02
03 01.4
Page 10 of 10
Effective Date
2008

Revision
02