Vous êtes sur la page 1sur 257

Group Design Project

Concept Design of a
Fast Sail Assisted
Feeder Container Ship
Aaron Burden,
Thomas Lloyd,
Simon Mockler,
Lorenzo Mortola,
Ie Bum Shin,
Ben Smith
Supervisor: Grant E Hearn
Academic year 2009/2010

Main Report

Acknowledgements
Thegroupmemberswouldliketothankanumberofpeoplewhohavehelpedinthecompletionoftheproject.
Firstlytooursupervisor,ProfessorGrantHearnwhohasguidedtheworkfromdayoneandpushedusinall
aspects of the project. Also to Ian Campbell, the 2nd supervisor whose advice regarding and supervision of
towingtanktestingprovedinvaluable.
Other members of the WUMTIA who have provided support are Dickon Buckland, Martyn Prince and Chris
Harmer.ThanksalsotoDr.DominicTauntonforhisadviceontanktestingandresultsprocessing.Manythanks
to the University of Southampton and SouthamptonSolent Universityfor allowing us to use theirrespective
testingfacilities,andtoMikeTudorPoleforhishelpwiththeuseofthewindtunnel.Supervisioninthetowing
tankwaskindlyprovidedbyDr.MingyiTanandMs.SallyDenchfield,andisgreatlyappreciated.Manyhoursof
manufacturingtimehavecontributedtothesuccessfullydeliveryofthetestrigandshipmodelsthankstothe
membersoftheEDMCnamelyMikeSellwood,TomRoberts,GordonMills,PhilHerring,DaveWilliams,Richard
DoolerandMikeStreet.
The kind support of Lloyds Register is gratefully acknowledged, both for contribution to the budget which
allowedtheprojecttoachievealargeamountofphysicaltesting,andtheadviceprovidedbymanyemployees:
KimTanneberger,DavidTozer,EdFort,RobDickie,RhodaWilsonandespeciallyDr.FaiChengforrecognising
thepotentialoftheprojectearlyon.WinddatawaskindlysuppliedbyDr.ElizabethKentoftheNOCS.ABB
arethankedforprovidingazipodproductdata.Finally,thanksgotoChristianMashofBorchardLinesforhis
insightintotheoperationoffeedercontainershipsandtheimpactthishadonthedesign.

Concept Design of a Fast Sail Assisted Feeder Containership

Report Contributions

Section:
1.
2.
3.
3.1
3.2
3.2.1
3.2.3
3.2.2
3.3
4.
4.1
4.1.1
4.1.2
4.1.3
4.1.4
4.1.5
4.2
4.2.1
4.2.2
4.2.3
4.3
4.4
4.4.1.
4.4.2
4.4.3
4.4.4
5.
5.1
5.1.1.
5.1.2

5.1.3
5.2
5.2.1
5.2.2
5.2.3
5.2.4
5.2.5
5.2.6

Sectiontitle:
Introduction
DesignSpecification
InitialDevelopmentofFastFeederContainerShip
Basisships
Massandpoweringestimate
Massestimate
Powering
Stability
Mainpropulsionandmachineryoptions
HydrodynamicDesignDevelopment
Hullformdesign
Designforhydrodynamicperformance
HullAbasishullform
HullBbasishullform
Bulbousbowoptimisation
Results
Towingtankmodeldesignandmanufacture
Introduction
Modeldesign
Modelmanufacture
Testingmatrixandprocedure
Analysisofresults
Uprightcondition
Sailingcondition
Addedresistanceinwaves
Uncertaintyanalysis
SailSystemDesign
Initialdesign
Conceptreview
Design
Winddata
Theoreticalperformance
Windtunneltesting
Introduction
Modeldesign
Modelmanufacture
Experimentalsetup
Windtunnelpostprocessing
Analysisoftheresults
ComputationalfluiddynamicsstudyofMultiwing
system
PerformancePredictions
Designselection
Sailingperformance

5.3
6.
6.1
6.1.1

ii

Workundertakenby:
TL
BS,SM&LM

LM&IBS
IBS&LM
LM&IBS
LM&IBS
LM
LM&IBS

Writtenby:
TL
BS

AB

BS
BS
BS
TL

AB
AB
SM
AB
AB

TL
TL
TL
TL
AB
AB
AB
AB
BS

IBS
LM&IBS
BS
LM

LM&IBS
IBS
LM
LM
LM
LM&IBS

IBS

IBS

LM

LM

AB&TL
BS&SM
BS&SM
BS&SM
TL

AB
AB
SM
AB
AB
TL
TL,SM & LM
TL
AB&TL
AB
AB
AB
AB
BS

IBS
LM&IBS
BS
LM

Report Contributions

6.1.2
6.1.3
6.2
6.2.1
6.2.2
7.
7.1
7.1.1
7.1.2
7.1.3
7.1.4
7.1.5
7.1.6
7.1.7
7.2
7.2.1
7.2.2
7.2.3
7.2.4
7.3
7.3.1
7.3.2
7.3.3
7.4
8.
9.

Propulsiveefficiency
Justificationofhullformchoice
Designfeasibility
Roundtripevaluation
EconomicfeasibilityofMultiwingsystem
DesignDevelopment
Seakeeping
Modelling
Results
Theoreticaladdedresistance
Absolutemotions
Cargosecuring
Motions
Rolldampingduetosails
StructuralDesign
Midshipscantlings
Globalstrength
Finiteelementmidshipsectionmodel
Finiteelementsailrigmodel
Stability
Intactstability
Damagestability
Freeboardandtonnage
Layoutandarrangement
Conclusions
RecommendationsforFutureWork

TL
TL

TL
TL

TL
LM

AB
AB
AB
BS
BS
BS
LM

SM
SM
BS
IBS

IBS
SM
AB
BS
TL
TL

TL
LM

AB
AB
AB
BS
BS
BS
LM
SM
SM&BS
BS
IBS
IBS
SM
AB
BS
ALL
ALL

iii

Concept Design of a Fast Sail Assisted Feeder Containership

Table of Contents
ListofFigures...............................................................................................................................................vi
ListofTables..................................................................................................................................................x
Nomenclature............................................................................................................................................xiv
1.

Introduction..........................................................................................................................................1
1.1 Background.............................................................................................................................................1
1.2 Aimsandobjectives................................................................................................................................2
1.3 Outcomeandmainachievements..........................................................................................................2

2.

DesignSpecification...............................................................................................................................3
2.1 Operatingprinciples................................................................................................................................3
2.2 Regionanalysis........................................................................................................................................4
2.3 Portanalysis............................................................................................................................................4
2.3.1 Throughputpredictions...................................................................................................................4
2.3.2 Portrestrictions...............................................................................................................................5
2.4 Routeanalysis.........................................................................................................................................6
2.5 Operationalanalysis...............................................................................................................................8
2.6 Economicandenvironmentalanalysis...................................................................................................9
2.7 Summary...............................................................................................................................................12

3.

InitialDevelopmentofFastFeederContainerShip...............................................................................13
3.1 Basisships.............................................................................................................................................13
3.2 Initialmass,poweringandstabilityestimates......................................................................................15
3.2.1 Massestimate................................................................................................................................15
3.2.2 Powering........................................................................................................................................17
3.2.3 Stability..........................................................................................................................................18
3.3 Mainpropulsionandmachineryoptions..............................................................................................19
3.3.1 Propulsor........................................................................................................................................19
3.3.2 Plantandfuels...............................................................................................................................21

4.

HydrodynamicDesignDevelopment....................................................................................................25
4.1 Hullformdesign....................................................................................................................................25
4.1.1 Designforhydrodynamicperformance.........................................................................................25
4.1.2 HullAbasishullform.....................................................................................................................26
4.1.3 HullBbasishullform.....................................................................................................................27
4.1.4 Bulbousbowoptimisation.............................................................................................................29
4.1.5 Results............................................................................................................................................31
4.2 Towingtankmodeldesignandmanufacture.......................................................................................35
4.2.1 Introduction...................................................................................................................................35
4.2.2 Modeldesign.................................................................................................................................35
4.2.3 Modelmanufacture.......................................................................................................................38
4.3 Preparationsfortesting........................................................................................................................39
4.4 Analysisofresults.................................................................................................................................42
4.4.1 Uprightcondition...........................................................................................................................43
4.4.2 Assessmentofhydrodynamicforcesresultingfromuseofsails...................................................49
4.4.3 Addedresistanceinwaves.............................................................................................................53
4.4.4 Uncertaintyanalysisoftestingdata..............................................................................................55

5.

SailSystemDesign...............................................................................................................................58
5.1 Initialdesign..........................................................................................................................................58
5.1.1 Conceptsreview.............................................................................................................................58
5.1.2 Design............................................................................................................................................60
5.1.3 Theoreticalperformance...............................................................................................................66
5.2 Windtunneltesting..............................................................................................................................67

iv

Contents
5.2.1 Introduction...................................................................................................................................67
5.2.2 Modeldesign.................................................................................................................................67
5.2.3 Modelmanufacture.......................................................................................................................69
5.2.4 Experimentalsetup........................................................................................................................71
5.2.5 Windtunnelpostprocessing.........................................................................................................72
5.2.6 AnalysisoftheResults...................................................................................................................78
5.3 ComputationalfluiddynamicsstudyofMultiwingsystem.................................................................86
5.3.1 Meshgenerationandphysicalmodel............................................................................................86
5.3.2 Simulations....................................................................................................................................86
6.

PerformancePredictions.....................................................................................................................89
6.1 Designselection....................................................................................................................................89
6.1.1 Sailingperformance.......................................................................................................................89
6.1.2 Propulsiveefficiency......................................................................................................................97
6.1.3 Justificationofhullformchoice...................................................................................................101
6.2 Designfeasibility.................................................................................................................................101
6.2.1 Roundtripevaluation..................................................................................................................101
6.2.2 EconomicfeasibilityofMultiwingsystem..................................................................................106

7.

DesignDevelopment.........................................................................................................................109
7.1 Seakeepinganalysis............................................................................................................................109
7.1.1 Modelling.....................................................................................................................................109
7.1.2 Results.........................................................................................................................................109
7.1.3 Theoreticaladdedresistance.......................................................................................................111
7.1.4 AbsoluteMotions........................................................................................................................113
7.1.5 Cargosecuring.............................................................................................................................113
7.1.6 Motions.......................................................................................................................................117
7.1.7 Rolldampingduetosails.............................................................................................................121
7.2 Structuraldesign.................................................................................................................................125
7.2.1 Midshipscantlings.......................................................................................................................125
7.2.2 Globalstrength............................................................................................................................128
7.2.3 Finiteelementmidshipsectionmodel........................................................................................130
7.2.4 Finiteelementsailrigmodel.......................................................................................................144
7.3 Stability...............................................................................................................................................148
7.3.1 Intactstability..............................................................................................................................148
7.3.2 Damagestability..........................................................................................................................150
7.3.3 Freeboardandtonnage...............................................................................................................152
7.4 Layoutandarrangement....................................................................................................................152

8.

Conclusions.......................................................................................................................................155

9.

FutureWorkandRecommendations.................................................................................................158

References...............................................................................................................................................159
AppendixABasisShips...........................................................................................................................167
AppendixBHullFormDevelopment.......................................................................................................168
AppendixCModelDesign,ManufactureandTestingPreparation...........................................................175
AppendixDTowingTankResultsProcessing...........................................................................................179
AppendixESailDesign............................................................................................................................186
AppendixFWindTunnel.........................................................................................................................191
AppendixGPerformancePredictions......................................................................................................210
AppendixHDesignFeasibility.................................................................................................................213
AppendixISeakeeping...........................................................................................................................216

Concept Design of a Fast Sail Assisted Feeder Containership


AppendixJPredictedRollDamping.........................................................................................................219
AppendixKStructuralDesign..................................................................................................................222
AppendixLStability................................................................................................................................232
AppendixMGeneralArrangement..........................................................................................................238
AppendixNBudgetSummary..................................................................................................................239

List of Figures
Figure2.1Primaryworldtraderoutes[NOAASatelliteandInformationService(2009)]....................................3
Figure2.2Shippingnetworkwithout(left)andwith(right)transhipment[PSASingaporeTerminals(2009)]...3
Figure2.3Predictedgrowthinregionalcontainershipmarketsby2020............................................................5
Figure2.4Flowchartusedtodeterminetherequiredshipcapacityandspeedforaparticularroute.............10
Figure3.1PBasafunctionofthecubeofshipspeed(Vs3).................................................................................14
Figure3.2Typicalpoddeddrive(ABBVOseries)andassociatedmachinery[ABB(2009)]................................20
Figure3.3ContrarotatingpropellerpodlayoutonfastRoPaxship[Levander(2009)]....................................21
Figure3.4ExampleLNGelectricContrarotatingpropellerpodpropulsionsystemarrangement:plantare
showninbluewithgeneratorsandmotorsinred[Levander(2002)]................................................24
Figure4.1VariationofoptimumLCBwithCB,Schneekluth&Bertram(1987)..................................................27
Figure4.2HullAbasissectionalareacurve........................................................................................................28
Figure4.3BodyplansfortheaftsectionsofHullB,MichletbasisandGodzillaoptimised(left);and
variationinwaveresistancecoefficientCWwithshipspeed(right)..................................................29
Figure4.4Schematicshowingbulbousbowdimensiondefinitions...................................................................31
Figure4.5VariationofCTat25knotsforallbulbvariants.................................................................................32
Figure4.6BreakdownofresistancecomponentsasafunctionofFnforbothHullAandHullB......................33
Figure4.7EffectivepoweragainstshipspeedforHullAandHullB..................................................................34
Figure4.8(a)Turbulencestimulationstuddimensions;and(b)comparisonoftheITTCandWUMTIA
recommendedpositions.....................................................................................................................38
Figure4.9Schematicshowingtotaltrimmomentcorrectionlevers and ....................................................41
Figure4.10Illustrationofmisalignment(left);andyawmomentcorrection(right)..........................................44
Figure4.11CorrectedmodelresistanceforHullAandHullB............................................................................44
Figure4.12VariationofCTSwithsurfaceroughness...........................................................................................46
Figure4.13Effectivepowercomparison............................................................................................................47
Figure4.14Breakdownofresistancecomponentsfor(a)HullA;and(b)HullB................................................48
Figure4.15ComparisontoHoltropeffectivepowerforHullAandHullB.........................................................48
Figure4.16ComparisonofindividualresistancecomponentstothoseobtainedusingtheHoltrop
regression.(a)CTSCAAS;(b)CWS;(c)CFSand(d)CVS...........................................................................49
Figure4.17Illustrationofresistancecomponentsduetoheelandleewayasafunctionofsideforce.............51
Figure4.18Evaluationofsideforceandresistance,(a)HullA15.9knots;(b)HullB15.9knots;(c)Hull
A25.5knots;(d)HullB25.5knots................................................................................................52
Figure4.19NondimensionaladdedresistanceprofilesforHullAandHullBatbothtestedspeeds...............54
Figure4.20Schematicoftestsystemgroupedintoareasofuncertainty[ITTC(2008c)]...................................55
Figure4.21Comparisonofgeneratedwavefrequencytorequestedwavefrequency.....................................57
Figure5.1Polarplotfordifferentsailrigdesigns...............................................................................................59
Figure5.2Multiwingsystemandgeneralarrangementofcargoship[Walker(1985)]....................................60
Figure5.3Aluminiumaircraftwingloadingagainstwingweightperunitarea.................................................61
Figure5.4VariationofInducedDragCoefficientwithaspectratio....................................................................63
Figure5.5Multiwingsystem.............................................................................................................................63
Figure5.6Comparisonofdragcoefficientagainstliftcoefficientsquaredfordifferentflapchordpositions
(left)and;NACA0015polarplotforflapangles22.5and45degrees(right)....................................64
Figure5.7Seasurfacewindfor2002:magnitudeasscalarmean[NOAASatelliteandInformationService
(2009)]................................................................................................................................................65
Figure5.8WindbystrengthanddirectionfortheCaribbean(annual)..............................................................65
Figure5.9Windtunnelmodelassembledview..................................................................................................69

vi

Contents
Figure5.10Modelassembledinthewindtunnel..............................................................................................70
Figure5.11Experimentalsetupinwindtunnelcontrolroom............................................................................71
Figure5.12Sixcomponentdynamometerinthelowspeedwindtunnel[Campbell(2009)]............................71
Figure5.13Containerslayoutanddimensionsforsailhullinteractiontests....................................................73
Figure5.14Dynamometermisalignmentfromportandstarboardtacks(left);andvariationofbasedrag
withgritsize(right)[Braslow&Harris(1966)]...................................................................................73
Figure5.15Windagedata:wind(left)andship(right)axescoefficientsofthesupportingstructure.
Boundarycorrectionsareapplied......................................................................................................75
Figure5.16Lift(left)anddrag(right)curveswithvaryingReynoldsnumber;wingspacing100%ofthe
chordzerostagger,zeroflapdeflection.............................................................................................76
Figure5.17(a)LiftcurveextrapolationtofullscaleRe;and(b)DragcurveextrapolationtofullscaleRe.......77
Figure5.18Multiwingliftcoefficientexperimentaldatawithchangeinspacing(left),andBiplanetheory
data(right)..........................................................................................................................................78
Figure5.19Lift(left)anddrag(right)coefficientagainstAoAatfourspacings.................................................79
Figure5.20Dragbreakdowntrendat100%and120%spacing.........................................................................79
Figure5.21VariationinbiplanestaggerpositionliftconstantGwithgap/chordratio.....................................80
Figure5.22Aerodynamiccoefficientsanddragbreakdownforthreestaggerconfigurations..........................81
Figure5.23Flatsailconfiguration:dragcoefficientagainstliftcoefficientsquared(left)and;liftanddrag
coefficientagainstAoA(right)............................................................................................................81
Figure5.24Shipaxescoefficientsformaximumwingspacingspacing.Theplotshighlightthebenefices
derivingfromthepresenceofthecontainersunderthebaseofthewing.......................................82
Figure5.25Shipaxescoefficients[lift(left);drag(right)]oftheselectedconfigurations;themarker
indicatesthechangefromthesailingpointfromwindwardtobeamreaching................................83
Figure5.26ComparisonofMultiwingrigwithotherexistingrigs.Thegraphshowsonlythesailingpointat
maximumliftcoefficient....................................................................................................................84
Figure5.27SingleandMultiwing(WingA)comparison(left)and;Singlefoilcomparison,liftcurve
(ExperimentalandXfoildataisforRe=235,000whilsttheNACAdatapointsareforReef=
2.76x106(right)...................................................................................................................................85
Figure5.28Meshgenerationofwholedomain(left)andsinglewing(right)....................................................87
Figure5.29Aerodynamiccharacteristicsof100%chordspacingwith22.5degreeflapmodel........................87
Figure5.30StreamlinesatanAoAof:eightdegrees(topleft),tendegrees(bottomleft);andpressure
coefficientateightdegrees(right).....................................................................................................87
Figure5.31StreamlineforforwardandaftMultiwingsystematapparentwindangleof30degrees.............88
Figure6.1Flowdiagramforperformancepredictionprogram.Thesourcecodeisnotincludedinthe
appendicesforpracticalreasons.ItwillbeincludedintheCDromattachedtothemainreport....91
Figure6.2HeelangleHullA,VS=25knots,VT=30knots.Thestepinthesailcurveisthechangeinsail
coefficientsCLandCD.Ahigherwindspeedischoseninordertoappreciatethehullwindage
component.........................................................................................................................................93
Figure6.3Variationinrolldampingwithtruewindangle(Singaporemeanwindspeedof5.9ms1)..............93
Figure6.4Variationinleewayanglewithtruewindangle(Singaporemeanwindspeedof5.9ms1)..............93
Figure6.5Variationineffectivedraughtwithleewayangle,atVS=15knots...................................................94
Figure6.6Variationineffectivedraughtwithleewayangle,atVS=25knots...................................................95
Figure6.7Poddimensionsusedindragestimate[ABB(2009)].........................................................................98
Figure7.1Mappingof11LewissectionstoHullA(Shipsectionswhite;Lewisformsgreen)...................110
Figure7.2ComparisonofnumericalandexperimentaladdedresistanceHullA..........................................111
Figure7.3ComparisonofnumericalandexperimentaladdedresistanceHullB..........................................112
Figure7.4Modellingassumptionsrequiredfortheanalysisofthecargosecuringarrangements[Lloyd's
Register(2009),Part3Chapter14]..................................................................................................114
Figure7.5Containerstacksconsideredincontainersecuringanalysis............................................................115
Figure7.6Diagrammaticrepresentationofthecargosecuringarrangements[Lloyd'sRegister(2009)]........116
Figure7.7PolarplotofSubjectiveMotionvariationwithwaveheadingheightforthebridgedeckand
officersloungeat25knots...............................................................................................................119
Figure7.8Probabilityofslammingasafunctionofshipspeedandwaveheight............................................119
Figure7.9Probabilityofslammingasafunctionofwaveperiodandshipspeedfora7.5msignificantwave
height.Solidlinesindicatewavesthatexistintheinvestigatedseaareasanddashedlines
indicatewavesthatdonotexistintheinvestigatedseastates.......................................................120
Figure7.10MotionInducedIndex(MII)forvariouswaveheights,waveheadingandspeeds.......................121

vii

Concept Design of a Fast Sail Assisted Feeder Containership


Figure7.11Rollreductionfor15and25knotsshipspeedatdifferentapparentwindangles........................123
Figure7.12ReductionofRollInducedYawComponent.Shipspeed25knots................................................124
Figure7.13Structuralarrangementandprincipalpanelswithinthemidshipsection.....................................126
Figure7.14Coffindiagramforstructuremassnormalisedbyamidships(left)andresultinglightship
distributionof(right).....................................................................................................................130
Figure7.15DistributionofstillwatershearforceandSWBMintheloadedarrivalcondition(left)and
ballastarrival(right).........................................................................................................................130
Figure7.16Theextentsofthefiniteelementmodel(left);andthemodellingofstiffenedplatesusinga
combinationofbeamandplateelements(right).............................................................................131
Figure7.17Thefiniteelementmodelshowing:thefinemeshregionattheintersectionofthemastand
crossdeck(left);andtheapplicationofloadstothetopofthemast(right)...................................133
Figure7.18Locationsofgroundspringsusedtorestrainthemodelatitsends..............................................133
Figure7.19Applicationofendmomentstotheindependentpointsattheendofthemodel(left);andthe
meshattheintersectionofthetransversebulkheadandsidetanks(right)...................................134
Figure7.20Applicationofhydrostaticandcargopressures(left);andcorrectionloads(redarrows)
requiredtoensurenonetverticalforceonthemodel(right).........................................................135
Figure7.21Locationsofareasofinnerbottom(17,18)andbilgeboxtop(15)platingthatdonotmeetthe
LRSDAacceptancecriteriadefinedinTable7.18(redareas)fortheinitialanalysisinbeamseas
condition(left);andthestressconcentrationcreatedattheintersectionofthecrossdeckand
innerhull(14)fortheheadseascondition(right)[Areasinblueindicatestructurethatmeets
theLRSDAacceptancecriteria]........................................................................................................139
Figure7.22Structuralimprovementsmadetoshipstructureasaresultoftheinitialanalysis......................139
Figure7.23Locationsofareasofkeelplating(1)thatdonotmeettheLRSDAacceptancecriteria(red
areas)fortheinitialstructureanalysis(left);andthemodifiedstructureanalysis(right)inbeam
seascondition[AreasinblueindicatestructurethatmeetstheLRSDAacceptancecriteria].........140
Figure7.24Locationsofareasofthesideshell(4,5)thatdonotmeettheacceptancecriteriaofTable7.19
(redareas)(redareas)forthebeamseacondition[Areasinblueindicatestructurethatmeets
theLRSDAacceptancecriteria]........................................................................................................140
Figure7.25Structuralimprovementsmadetomaststructureasaresultoftheinitialresults,withthe
additionofbracketsatthemastandcrossdeckintersection(left);andinternalringstiffening
(right)................................................................................................................................................141
Figure7.26Stressinthefinemeshzoneatthebaseofthemastwhichdoesnotmeettheacceptance
criteriaofTable7.19(redareas)fortheinitialanalysis(left);andthemodifiedanalysis(right)in
theheadseascondition[AreasinblueindicatestructurethatmeetstheLRSDAacceptance
criteria].............................................................................................................................................141
Figure7.27Highstressconcentrationregionsaroundthemastandcrossdeckintersectionfortheinitial
analysis(left);andthemodifiedanalysis(right)fortheheadseascondition..................................142
Figure7.28Upwindmode(left)anddownwind/storagemodel(right)inANSYS............................................145
Figure7.29Upwindmodelsolutioninoperationalcondition(left)andworstcondition(right).......................145
Figure7.30Downwind/storagemodelsolutioninoperationalcondition(left)andworstcondition(right)...146
Figure7.31Thestressdistributionbefore(left);andafter(right)modification..............................................147
Figure7.32Bucklingmodeofdownwind/storagemodel.................................................................................147
Figure7.33DerivedwindheelingleverwithtypicalGZcurve..........................................................................150
Figure7.34Locationofdamagezonesalongtheship,alignedwithwatertightsubdivision...........................151

FigureA.1LOAasafunctionofcargocapacity(left);andBasafunctionofcargocapacity(right).................167
FigureA.2Draughtasafunctionofcargocapacity(left);andLOA/Basafunctionofshipspeed(VS)(right)..167
FigureA.3Deadweightasafunctionofcargocapacity...................................................................................167
FigureB.1ComparisonofresultsusingHoltropregressionandThinShipTheory..........................................171
FigureC.1ViewsofHullA:bulbousbowshapeandinternaldetail(left);andsternshape(right).................175
FigureC.2ViewsofHullB:bulbousbowshape(left);andsternshape(right)................................................175
FigureC.3(a)HullAsternsectionduringmilling;(b)HullAatprefinishingstage;(c)Bulkheadsandjoin
offorwardandaftmodelsectionsduringgluing;(d)ComparisonofsternformsofHullsAandB;
(e)Turbulencestimulatorlocationsandfairingissuesatintersectionofbulbandhull;(f)Heel
fittingandtrimmingmomentrailattachedtoHullA.......................................................................177
FigureC.4Totaltrimmomentcorrectionappliedforrangeofmodelspeedstested.....................................178
FigureD.1Variationinempiricaldragcorrectionwithtripstudlocation.......................................................179

viii

Contents
FigureD.2ResistanceestimatesofHullAandHullBwitherrorbars.............................................................183
FigureD.3YawmomentvariationwithyawangleofHullAandHullBwitherrorbars.................................183
FigureD.4Errorbarsonwaveprofilemeasurements.....................................................................................185
FigureD.5FrequencyspectrumsforHullAaddedresistancetests................................................................185
FigureD.6FrequencyspectrumsforHullBaddedresistancetests................................................................185
FigureE.1SkySailsSystem...............................................................................................................................186
FigureE.2NYKSuperEcoShip2030(left)andE/SOrcelle(right)concepts....................................................186
FigureE.3ExtrapolationofdataforWingsail:effectivepower(left);andaveragepropulsiveforce(right)
atanaveragewindspeedof15knots..............................................................................................187
FigureE.4Forcesactingonsailassistedship[Satchwell(1989)]....................................................................188
FigureE.5Windbystrengthanddirectionforthewholeworld(annual).......................................................190
FigureE.6WindbystrengthanddirectionforSouthEastAsia(annual)..........................................................190
FigureE.7WindbystrengthanddirectionfortheNorthAtlantic(annual).....................................................190
FigureF.1CNCmillingmachine(left);andhotwirecutter(right)..................................................................192
FigureF.2Modelmanufacture:wingwithstock(left);andbase(right).........................................................192
FigureF.3Sailcontainerinteractionatapparentwindangleof60degrees(WingA,maximumchord
spacing)Theboxeswhichcoverthebasehadtobetrimmedtoavoidanycontactwiththe
dynamometerlink............................................................................................................................198
FigureF.4Sailcontainerinteractionatapparentwindangleof109degrees(WingB,60degreesof
stagger).............................................................................................................................................198
FigureF.5Flatsailconfiguration(WingC)fordownwindperformance..........................................................199
FigureF.6Windagedataofrigsupportingstructure,nointeractions............................................................199
FigureF.7Variationinsolidblockagecorrectionparameters 1and 1.......................................................200
FigureF.8LowReynoldsnumberdragcurveforCLARKYtypeairfoil[Marchmann&Werme(1994)]In
particularthecurveontherightforRe200,000showssimilartrendtotheMultiwinginthe
speedcalibrationruns......................................................................................................................202
FigureF.9LowReynoldnumbertesting.Va=6ms1,AoA=14degrees.Thefirstcolumnoftufts(20%
chord)isattachedtothewing,whilstseparationstartat30%ofthechord.Notehowthe
alignmentofthesecondandthirdcolumnoftuftschangefromFigure125,unfortunatelythe
pictureisnotclearenough,howeverobservingthevideoitispossibletoseethetuftsvibrate....203
FigureF.10LowReynoldsnumbertesting.Va=8ms1,AoA=14degrees.Herethetuftsarealignedwith
theflowandareattachedtothesurface.Observingthebottomrowispossibletonotetheend
vortexcomingoffthewing;thefirsttwotuftsofthisrowareseparatingfromthesurfacedueto
theinterferenceofthebottombar.................................................................................................203
FigureF.11ContainerSailinteractionstudy.thegapbetweentheboxesandthebaseofthewing
replicatesthefullscalegapof2metres.Notetheredwirewhichwasusedinsteadofthesmoke
tovisualizetheflownearthegap;unfortunatelythisintrusivemethodhadlimited
effectiveness....................................................................................................................................204
FigureF.12NACA0015maximumliftcoefficient(left);basedragcoefficient(right)atdifferenteffective
ReynoldsNumbers[Jacobs&Sherman(1937)]...............................................................................204
FigureF.13StandarddeviationofwingperformancecoefficientsforWingA(left);andWingB,60degrees
ofstagger(right)...............................................................................................................................208
FigureF.14StandarddeviationofwingperformancecoefficientsforWingA,flapat45degrees(left);and
WingB,30degreesofstagger(right)...............................................................................................208
FigureF.15StandarddeviationofwingperformancecoefficientsforWingC................................................209
FigureF.16Boundarylayer(left)andfirstwall(right)thicknesses.................................................................209
FigureG.1ManufacturersVOseriespodeddrivespecifications[ABB(2009)]..............................................211
FigureG.2L50DFenginedrawing[WrtsilShipPowerTechnology(2009a)]...............................................212
FigureH.1Speedprofileforfastfeederandcomparisonshipsoverfortnightlyoperatingperiod................213
FigureH.2Powerprofileforfastfeederandcomparisonshipsoverfortnightlyoperatingperiod................214
FigureH.3Freightratebasis[Hansa(2009)]...................................................................................................215
FigureI.1HeaveRAOat25knots(left);and15knots(right)..........................................................................216
FigureI.2RollRAOat25knots(left);and15knots(right)..............................................................................216
FigureI.3PitchRAOat25knots(left);and15knots(right)............................................................................216
FigureI.4PolarplotofSMvariationwithwaveheadingheightforthebridgedeckandofficersloungeat
20knotsforHullA............................................................................................................................217

ix

Concept Design of a Fast Sail Assisted Feeder Containership


FigureI.5PolarplotofSMvariationwithwaveheadingheightforthebridgedeckandofficersloungeat
15knotsforHullA............................................................................................................................217
FigureI.6PolarplotofSMvariationwithwaveheadingheightforthebridgedeckandofficersloungeat
10knotsforHullA............................................................................................................................218
FigureJ.1Rolldecaycurve...............................................................................................................................219
FigureJ.2Frameofreferenceforrolldampingcalculations...........................................................................219
FigureJ.3Variationinaerodynamicrolldampingincidencechangescoefficientcomponentwith
apparentwindangle.........................................................................................................................220
FigureJ.4Variationinaerodynamicrolldampingairspeedchangescomponentwithapparentwind
angle.................................................................................................................................................220
FigureK.1HighstressregioninthebottomshellthatdonotmeettheacceptancecriteriaofTable7.18in
theinitialanalysis(left);andthemodifiedanalysis(right)forthestillwatercondition..................230
FigureK.2HighstressregioninthebottomshellthatdonotmeettheacceptancecriteriaofTable7.18in
theinitialanalysis(left);andthemodifiedanalysis(right)fortheheadseascondition..................230
FigureK.3HighstressregioninthebottomshellthatdonotmeettheacceptancecriteriaofTable7.18in
theinitialanalysis(left);andthemodifiedanalysis(right)fortheobliqueseascondition.............230
FigureK.4Stressinthefinemeshzoneatthebaseofthemastwhichdoesnotmeettheacceptance
criteriaofTable7.19fortheinitialanalysis(left);andthemodifiedanalysisinthebeamseas
condition...........................................................................................................................................231
FigureK.5Stressinthefinemeshzoneatthebaseofthemastwhichdoesnotmeettheacceptance
criteriaofTable7.19fortheinitialanalysis(left);andthemodifiedanalysisintheobliqueseas
condition...........................................................................................................................................231
FigureL.1Stabilitymodelusedintheanalysis(refertoAppendixMforarrangementandfunctionof
spaces),showing:(a)distributionofinternalsubdivision;(b)dispositionoftankspaces;and(c)
allocationofinternalcompartments................................................................................................232
FigureL.1Fullloaddeparture(left)andarrival(right)GZcurves....................................................................234
FigureL.2Ballastdeparture(left)andarrival(right)GZcurves.......................................................................234

List of Tables
Table2.1Identificationofregionalhubportsbyminimumcombineddistancetoallsatelliteports..................4
Table2.2Predictedcontainershipmarketgrowthby2020.................................................................................5
Table2.3Summaryofportrestrictions................................................................................................................6
Table2.4Summaryofrequiredshiprangeforcitedhubports............................................................................6
Table2.5Distancetravelledonagivenheadingfortheregionsbeingconsidered.............................................7
Table2.6Breakdownofbasisshipsbyyearbuilt.................................................................................................7
Table2.7Summaryofrequiredspeed................................................................................................................10
Table2.8Summaryofrequiredshipcapacity....................................................................................................11
Table2.9PredictedCO2emissionsin2020.........................................................................................................11
Table2.10Numberofshipsrequiredin2009and2020byregion....................................................................12
Table2.11Averageshipsizebyregion...............................................................................................................12
Table3.1Summaryofregressionanalysisresults..............................................................................................13
Table3.2Summaryofmassestimatesbasedonscalingbasisships..................................................................17
Table3.3Summaryofmassestimates...............................................................................................................17
Table3.4Assumedoperationalprofileforfuelandtankmassestimates.........................................................18
Table3.5Capacityandemissionpertrip............................................................................................................18
Table3.6Summaryofinitialstabilitycheck.......................................................................................................18
Table3.7Summaryofprincipleparticulars........................................................................................................19
Table3.8EstimatedemissionsreductionusingLNGcomparedtoMDO[Levander(2008a)]............................23
Table3.9Comparisonofmarinefuelsbydensity,lowerheatingvalueandcost[MDOandMGOprices
obtainedonline,correcton23rdMarch2010,Singapore;LNGpriceestimatedbyLevander
(2008b)]..............................................................................................................................................23
Table4.1BasisestimatesandGodzillaoptimisationconstraintsonprincipalparticulars.................................29
Table4.2Topologyofbulbvariants....................................................................................................................30
Table4.3Summaryofprincipalparticularsandhydrostatics.............................................................................35

Contents
Table4.4SummaryofSouthamptonSolentUniversityTowingTankfacility.....................................................35
Table4.5Summaryofmodelscalehullsatdesignspeed..................................................................................37
Table4.6Componentmassesusedinestimationofmodellongitudinalcentreofgravityandpitchgyradius.37
Table4.7Timeallowedperruninminutesforbothconservativeandoptimisticschedules........................40
Table4.8Wavefrequenciesinputintowavemaker...........................................................................................42
Table4.9ModelmisalignmentcorrectiontoyawmomentHullA..................................................................43
Table4.10Sensitivityofformfactorduetovaryingsurfaceroughness............................................................46
Table4.11NondimensionaladdedresistanceforHullAandHullBatbothtestedspeeds.............................54
Table4.12Thereductionofuncertaintyinexperimentalresultsduetorepeats..............................................56
Table4.13Errorinwaveheightbetweenrequestedandmeasuredwaves......................................................57
Table5.1Typicalaerodynamicparametersofsomebasicrigtypes.[Schenzle(1985)].....................................58
Table5.2Sailrigaerodynamicperformance,2Dsectiondata[Bergeson&Greenwald(1985)].......................59
Table5.3Totalsailareaestimation....................................................................................................................60
Table5.4Mastandmainstockbendingmoment,deflectionandmaximumstressusingaliftcoefficientof
2.2.......................................................................................................................................................61
Table5.5Multiwingsystemdimensions...........................................................................................................63
Table5.6Meanwindspeedbyregion................................................................................................................65
Table5.7Theoreticalthrustbenefitprediction(VS=15knots,VT=16.5knots)...................................................66
Table5.8Initialthrustreductionprediction.......................................................................................................67
Table5.9Windtunnelmodelweightestimation...............................................................................................68
Table5.10Windtunneltestmatrix.Rigtestedatarangeofanglesofattackuptostall..................................72
Table5.11Windtunnelwakeblockagecorrection............................................................................................75
Table5.12Separateddrag,CDs...........................................................................................................................76
Table5.13Threestaggerconfigurationinpercentagechord............................................................................80
Table5.1Designchoice.WingAcorrespondtomaximumchordspacing,zerostaggerwhilstWingB
correspondstoastaggerangleof60degrees...................................................................................83
Table5.15Entryflowcharacteristicsofforwardandaftrigwithtotalliftanddragcoefficient.......................88
Table6.1AveragedaddedresistanceforHullAsailingat25knotsinSingapore(valuesinN).........................90
Table6.2AveragedaddedresistanceforHullBsailingat25knotsinSingapore(valuesinN).........................90
Table6.3CalculationoftheprobabilityofheadwavesforSingaporeregion....................................................90
Table6.4VoyagedetailsforSingaporearea......................................................................................................92
Table6.5PerformancesummaryforbothhullformsinSingaporeregion........................................................92
Table6.6PerformancesummaryforbothhullformsinCaribbeanregion........................................................92
Table6.7Roundtripspeedweightings..............................................................................................................95
Table6.8Percentagethrustreductionweightedforroundtrip........................................................................95
Table6.9AnnualNorthAtlanticvoyagesimulation...........................................................................................96
Table6.10Referencesusedforestimatingpropulsivecoefficients...................................................................97
Table6.11Electricalefficiencycomponentsofthetransmissionefficiency......................................................97
Table6.12Summaryofdimensionsusedinestimationofpoddeddrivedrag(HullAusesVO1800andHull
BusesVO2100)[ABB(2009)].............................................................................................................98
Table6.13Summaryofvaluesusedinestimationofinstalledpropulsivepowerforbothhulls.......................99
Table6.14SavingsinpowerrequirementforHullsAandBbasedonannualandseasonalwindconditions,
Singaporeregion(powermarginnotincluded)...............................................................................100
Table6.15Hydrodynamicandaerodynamiccomponentsofswayforceandyawmoment...........................102
Table6.16Summaryofdatausedforcargohandlingsimulation[Kalmar(2009)]..........................................103
Table6.17Summaryofsimulatedvoyagetomeetfortnightlycontainerdemand.........................................104
Table6.18Summaryofpowerrequirementsforfastfeederandcomparisonshipsoverfortnightlyperiod.104
Table6.19Summaryoffuelconsumptionandcost,andemissionsforfastfeederandcomparisonships
overfortnightlyperiod.....................................................................................................................105
Table6.20Summaryoffastfeederperformanceagainstcomparisonshipsusingdesignindices..................105
Table6.21Summaryofdailycostestimatescomparingfastfeederconcepttotwotypicalexistingships....106
Table6.22BuildingandrunningcostfortwoMultiwingrigs.........................................................................106
Table6.23AnnualbenefitsfromtheuseoftheMultiwing............................................................................107
Table6.24NPVanalysisforHullA.operatinginSingaporeregion..................................................................107
Table6.25NPVanalysisforHullB.Singapore..................................................................................................107
Table6.26NPVsensitivityanalysisforHullB,seasonalschedule....................................................................108

xi

Concept Design of a Fast Sail Assisted Feeder Containership


Table6.27NPVanalysisincludingthrustreductionsduetomotiondamping.Theassumedrateofreturnis
5.31%................................................................................................................................................108
Table7.1Summaryofabsolutemaximummotionsat25knotsforHullA......................................................113
Table7.2Summaryofabsolutemaximummotionsat15knotsforHullA......................................................113
Table7.3Acceptancecriteriaforcontainersecuring.......................................................................................115
Table7.4Maximumcompressiveforceintieroneoftheconsideredcontainerstacksandthemotion
conditiontocausefailure.................................................................................................................116
Table7.5Averagecontainermassthatcausesfailureofacontainerstack.....................................................116
Table7.6Subjectivemagnitudescale...............................................................................................................118
Table7.7Summaryofmaximumshipspeedincertainseastatestopreventexcessivemotions...................119
Table7.8Longitudinallyeffectivepanels..........................................................................................................126
Table7.9Designloadingsusedininitialscantlingderivation,externalandinternallocaldesignloadsfor
individualpanelsandmaximumglobalbendingmoments..............................................................126
Table7.10Platethicknessbydirectcalculationandminimumrulerequirements..........................................127
Table7.11Minimumrulerequirementsforlocalsectionmodulusofstiffenersandattachedplating...........128
Table7.12Requiredsectionmoduliandselectedwebprofilesfortransversestiffening................................128
Table7.13Globalpropertiesofthemidshipsection........................................................................................128
Table7.14RuleGlobalstrengthcriteria...........................................................................................................129
.Table7.15Rulederivedantisymmetricgloballoads......................................................................................129
Table7.16Combinedstresscalculations..........................................................................................................129
Table7.17Summaryofoptimummastdimensions.........................................................................................136
Table7.18Acceptancecriteriaforprimarystructure[Lloyd'sRegister(2006b)].............................................137
Table7.19Maximumpermissiblestressesforfinemeshanalysis[InternationalAssociationof
ClassificationSocieties(2008)].........................................................................................................137
Table7.20Finiteelementmodelvalidationbystructuremass........................................................................137
Table7.21Themastdeflectioninthedynamicloadcasesfortheinitialandmodifiedanalysis.....................142
Table7.22Errorinpeakstressduetothechoiceofmodellingschemeforthethreemotionloadcases........143
Table7.23TheoreticalandANSYSsolutioninoperationalandworstweathercondition...............................146
Table7.24Sumaryofmaximumdeflectionandstresscomparingoriginalandmodifiedsailrigdesigns.......148
Table7.25Effectofchangingmeshsizeonmaximumdeflectionandstress...................................................148
Table7.26Hydrostaticsoffourloadconditionsinequilibriumcondition.......................................................149
Table7.27valuesusedincalculationofwindheelinglever.............................................................................149
Table7.28Summaryofstabilitycharateristicsforanalyisdshipconditions....................................................149
Table7.29LY2Monohullsailingshipcriteria...................................................................................................150
Table7.30Damagecasesconsideredinthestabilityanalysisandtheirconstituentlongitudinaldamage
zones.................................................................................................................................................151
Table7.31Calculatedsubdivisionindexes.......................................................................................................152
Table7.32Summaryoffreeboardrequirements.............................................................................................152
Table7.33Tonnagemeasurements.................................................................................................................152

TableB.1HullA;Totalresistancecoefficientforallbulbvariants...................................................................168
TableB.2HullB;Totalresistancecoefficientforallbulbvariants...................................................................169
TableB.3Breakdownofcoefficientsofresistanceandeffectivepowerforoptimisedhulls..........................170
TableC.1Conservativeestimateofpossibletestsduringthreedays(totalof64runs).................................178
TableC.2Estimateofmaximumrequiredtestsduringthreetestingdays(totalof80runs)..........................178
TableD.1Exampledragcorrectioncalculationsforbothhulls.......................................................................179
TableD.2HullAeffectivepowerandresistancecomponentbreakdown.......................................................180
TableD.3HullBeffectivepowerandresistancecomponentbreakdown.......................................................180
TableD.4Resistancebreakdowninsailingcondition......................................................................................181
TableD.5Summaryofuncertaintyintotalresistanceandsideforcemeasurementsfortestsinsailing
condition...........................................................................................................................................184
TableE.1Sailmastdimensions........................................................................................................................187
TableE.2Mastbendingmoment,deflectionandmaximumstressusingdrivingforceestimation................187
TableF.1Estimationofmaximumforceexperiencedbyindividualwindtunneldynamometer....................191
TableF.2Stockbendingmoment,deflectionandmaximumstressforwindtunnelmodel...........................191
TableF.3Maststructurecalculationforwindtunnelmodel...........................................................................191
TableF.4Summaryofwindtunnelcalibrationmeasurements.......................................................................197

xii

Contents
TableF.5Bodyshapefactorsusedforwakeblockagecorrection[ESDU(1980)]...........................................201
TableF.6ModelfrontalareaatzeroAoA........................................................................................................201
TableF.7Calculationofwakeblockagecorrectionfactorwb.........................................................................201
TableF.8Measuredliftanddragcoefficientwithwingspacingof100%chord(noflapangle;nocontainer
interactions).....................................................................................................................................205
TableF.9Measuredliftanddragcoefficient,andlifttodragratio,forwingspacingsof100%and75%
chord(22.5degreeflapangle;nocontainerinteractions)...............................................................205
TableF.10Measuredliftanddragcoefficient,andlifttodragratio,forwingspacingsof50%and120%
chord(22.5degreeflapangle;nocontainerinteractions)...............................................................206
TableF.11Singlewingwith22.5degreeflapanglewithoutcontainers.........................................................206
TableF.12Measuredliftanddragcoefficient,andlifttodragratioforwingspacingof120%ofchordand
flapofangleof22.5degrees............................................................................................................206
TableF.13Measuredliftanddragcoefficient,andlifttodragratioforwingspacingof120%ofchordand
flapangleof45degreeswithcontainersat90to119degreesapparentwindangle.....................207
TableF.1430degreestaggerconfigurationwithwingspacingof120%ofchordwith22.5degreesflap
anglewithcontainersat90to119degreesapparentwindangle...................................................207
TableF.1560degreestaggerconfigurationwithwingspacingof120%ofchordwith22.5degreesflap
anglewithcontainers90to119degreesapparentwindangle.......................................................207
TableF.16Extremestaggerconfigurationwithwingspacingof120%ofchordwithcontainersat90to119
degreesapparentwindangle...........................................................................................................207
TableF.17Aerodynamiccharacteristicsofwingspacingof100%ofchordwith22.5degreeflapangle
model...............................................................................................................................................209
TableG.1ProbabilityofsignificantwavelengthtowaterlinelengthforSingaporearea...............................210
TableG.2ProbabilityofsignificantwaveheightandwavelengthtowaterlinelengthfortheCaribbean
area..................................................................................................................................................210
TableG.3AveragedaddedresistanceforHullAsailingat15knotsinSingaporevaluesinNewtons..........210
TableG.4AverageaddedresistanceforHullBsailingat15knotsinSingaporevaluesinNewtons............210
TableG.5AverageaddedresistanceforHullAsailingat15knotsintheCaribbeanareavaluesin
Newtons...........................................................................................................................................211
TableG.6AveragedaddedresistanceforHullBsailingat15knotsintheCaribbeanareavaluesin
Newtons...........................................................................................................................................211
TableG.7L50DFengineseriesdimensionsinmillimetres,andweights[WrtsilShipPowerTechnology
(2009a)]............................................................................................................................................212
TableH.1Summaryofplantspecificationsusedinvoyagesimulation[dataobtainedfromWrtsilShip
PowerTechnology(2009a)andMANDiesel(2010)].......................................................................213
TableH.2Emissionsestimatesusedinvoyagesimulation[dataobtainedfromWrtsilShipPower
Technology(2009a)andMANDiesel(2010)]...................................................................................213
TableJ.1Rolldampingtimeseries...................................................................................................................219
TableK.1PanelparticularsderivedfromtheNG254basisship......................................................................222
TableK.2Panelplatingdesignstressesforthemidshipsection.....................................................................223
TableK.3Rulebucklingchecksforpanelplateandlongitudinalstiffeners.....................................................224
TableK.4Assumedextentofmajorstructuralelementsalongthelengthoftheshipandstructural
distributionfactor............................................................................................................................227
TableK.5Lightshipmassdistribution(includingcrewandstores)..................................................................228
TableK.6Tankmassdistributionforeachloadingcondition(fortanknamesandpositionsseethegeneral
arrangement,AppendixM)..............................................................................................................229
TableK.7Fullloadcargomassdistribution(zerocargomassassumedinballastcondition).........................229
TableL.1Summaryofvaluesusedinintactstabilityanalysis.........................................................................233
TableL.2Intactstabilityresultssummary,applyingIMOA.749(18)Ch3and4.9code..................................235
TableL.3Longitudinalextentsofdamagezonesandlocationoflongitudinalbulkheadconsideredfor
damagelimitation,measuredfromtheaftextentandthesideshellrespectively.........................236
TableL.4Probabilitiespiforalldamagecases,withintermediateresultsforsinglezonedamages..............237
TableL.5Probabilitiessiforalldamagecasesandthethreeloadingconditions...........................................237
TableN.1Budgetsummary.............................................................................................................................239

xiii

Concept Design of a Fast Sail Assisted Feeder Containership

Nomenclature
Symbols:

Sailarea
Windtunnelcrosssectionalarea
contraction
Attainedsubdivisionindex
Windagelateralarea
Mouldedbreadthorbeam
Distancefromcentreofbuoyancy
totransversemetacentre
yawdampingcoefficientinduced
byroll
Shipmodelcorrelationallowance
Blockcoefficientatdesigndraught
Blockcoefficientat0.8
Dragcoefficient
Dragcoefficientwindagecorrection
Crossflowdragcoefficient
Hull/superstructuredrag
coefficientforwindage
Dragcoefficientduetoliftor
effectiveinduceddragcoefficient
Minimumorparasiticdrag
coefficient
Totaldragcoefficientatmaximum
lift
Frictionalresistancecoefficient
Frictionalresistancecoefficientat
15oC
Frictionalresistancecoefficientat
measuredwatertemperature
Liftcoefficient
Liftcoefficientwindagecorrection
Liftcoefficientatzeroangleof
attack
Prismaticcoefficient
Inducedresistancecoefficient,
motiondamping
Totalresistancecoefficient,roll
damping
Totalresistancecoefficient
Totalresistancecoefficientat15oC

Totalresistancecoefficientat
measuredwatertemperature
Viscousresistancecoefficient
Waveresistancecoefficient
Waterplaneareacoefficient
Sailthrustcoefficient
Sailheelingcoefficient
Mouldeddepth
Driveforce
Measuredmodeldrag
Propellerdiameter
Saildrag

xiv

,
,

Waterdepth
Rackingforcesattheithtierofa
containerstack
Residualforcesattheithtierofa
containerstack
Saildriveforce
Aerodynamicsideforce
Verticalcomponentofaerodynamic
sideforce
Froudenumber
Effectiveaspectratio
Biplaneliftconstantduetochange
inangleofattack
Biplaneliftconstantduetosection
curvature
Metacentricheight
Accelerationduetogravity
Windlever
Effectiverigheight
Heelingforce
Slidingforceattopofmotionsin
transversedirection
Slidingforceatbottomofmotions
intransversedirection
Verticaldistancefromcentreof
windagelateralareatohalfmean
draught
Highspeed(whensubscript)or
heightofgritstrip
Significantwaveheight
Shipsectionsecondmomentof
areaaboutthehorizontalneutral
axis
Structuralelementsecondmoment
ofareacontributionaboutthe
transverseneutralaxis
Shipsectionsecondmomentof
areaabouttheverticalneutralaxis
Sectionalsectorialmomentof
inertia
Slidingforceattopofmotionsin
longitudinaldirection
Slidingforceatbottomofmotions
inlongitudinaldirection
Coveragefactorforstandard
deviation
Verticalcentreofbuoyancyabove
keel
Verticalcentreofgravityabove
keel
Wavenumber
Equivalentspringstiffnessofa
containerstackattheithtier
Lashingstiffness

Nomenclature

Reynoldsnumber
Shipscale(whensubscript)or
wettedsurfacearea
Transverseshearforceattheith
tierinacontainerstack
Longitudinalshearforceattheith
tierinacontainerstack
Stiffenerspacing
Probabilityofsurvivalfollowing
damagecase
Draught
Designscantlingdraught
Effectivedraught
Heaveperiod
Tensioninalashingattheithtierin
acontainerstack
Pitchperiod
Rollperiod
Thrustdeductionfactor
Platingthickness
WindSpeed
Apparentwindspeed
Advancespeed
Modelspeed
Shipspeed
Truewindspeed
Windagedriveforce
WindageHeelingforce
Wakefraction
Structuralelementdistancefrom
verticalneutralaxis
Stiffenersectionmodulus
Structuralelementdistancefrom
transverseneutralaxis
Heightoftransverseneutralaxis
frombase

Pitchgyradius
Constantforgritstripcalculation
Lengthbetweenperpendiculars
Loadlinelength
Lengthoverall
Rulelength
Saillift
IMOsubdivisionlength

Lengthatwaterline
Modelscale(whensubscript)
Windheelingmoment
Numberoftestruns
Stillwaterbendingmoment
Staticcargotorque
Horizontalwavebendingmoment
Wavehydrodynamictorque
Waveverticalbendingmoment
Resistancemass
Trimmingmoment
Yawmoment
Meansquareoffrequency
spectrum
Cargocapacity(numberofTEU)
Numberofpropellerblades
Totalaerodynamicrolldamping
Aerodynamicrolldamping,
incidencechangescomponent
Aerodynamicrolldamping,air
speedchangescomponent
Propellerrps
Numberofholds
Numberofstacks
Numberoftiers
Brake(installed)power
Deliveredpower
Effectivepower
Verticalforceatcornerpostsof
containersinastackattheithtier
Probabilityofdamagecase
occuring
Windforceinonetransverseendof
acontainerinastackattheithtier
Dynamicpressure
Airdrag
Addedresistance
Averageaddedresistance
Averageaddedresistanceincluding
weightforshipheading
EffectiveReynoldsnumber
Frictionalresistance
Resistanceduetoheel
Inducedresistance
Requiredsubdivisionindex
Totalresistance
Uprightresistance
Viscousresistance

xv

Angleofattack
Apparentwindangle
Truewindangle
Massdisplacement
Displacementofacontainerstack
attheithtier
Positionoftorsionneutralaxis
belowsectionbaseline
Measuredwaveamplitude
Significantwaveamplitude
Hullefficiency
Propelleropenwaterefficiency
Relativerotativeefficiency
Trimangle
Modelscalefactor
Wavelength
Structuralmassvariationalongthe
shiplength

Concept Design of a Fast Sail Assisted Feeder Containership

Coefficientoftippingorslidingfor
calculationofMII
Kinematicviscosity
Nondimensionalhydrodynamic
rolldamping
Nondimensionalaerodynamic
rolldamping
Massdensityofwater
Massdensityofair
Summation

FEM
FEU

Addedresistancecoefficient

IRR
ITTC

FP
FPP
GL
HFO
IMO

Bucklingstresscorrectedfor
plasticityeffects

Midshipscantlingdesignstress
Permissibledesignstress
Stressduetolocallyappliedloads
Nominalstress

Longitudinaldirectstress

Transversedirectstress

Materialyieldstress

Acronyms:
AoA
AE
AP
AR

CO
CNC
CPP
CRP
CSR
DF
DoF
DWT
EEDI
FC
FE

LCB
LCF
LCG
LHV
LNG
LR
LR SDA
LY2
MBtu
MCA
MCR
ME
MDO
MGO
MII
MSI
NACA

Shearstress
Heelangle
Wavespectrumasafunctionof
waveencounterfrequency
Rollangle
Shipheading
Leewayangle
Yawangle
Wavefrequency
Waveencounterfrequency
Normalisedsectorialcoordinate

Volumetricdisplacement

Angleofattack
Auxiliaryengine
Aftperpendicular
Wingaspectratio
Aftpropeller(whensubscripted)
Carbondioxide
Computernumericallycontrolled
Controllablepitchpropeller
Contrarotatingpropeller
CommonStructuralRules
Dualfuel
Degreeoffreedom
Deadweightmass
EnergyEfficiencyDesignIndex
Fuelconsumption
Finiteelement

NPV
NO
PM
PPP
RAO
SECA
SF
SFC
SM
SO
SWBM
TEU
TR
TF
VCB

xvi

Finiteelementmethod
Fortyfootequivalentunit
(container)
Forwardpropeller(when
subscripted)
Forwardperpendicular
Fixedpitchpropeller
GermanischerLloyd
Heavyfueloil
InternationalMaritime
Organisation
Internalreturnrate
InternationalTowingTank
Conference
Longitudinalcentreofbuoyancy
Longitudinalcentreofflotation
Longitudinalcentreofgravity
Lowerheatingvalue
Liquefiednaturalgas
LloydsRegister
LloydsRegisterShipRightPrimary
StructureforContainerShips
designprocedure
TheLargeCommercialYachtCode
MetricBritish unit
MaritimeCoastguardAgency
Maximumcontinuousrating
Mainengine
Marinedieseloil
Marinegasoil
Motioninducedinterruption
Motionsicknessindex
NationalAdvisoryCommitteefor
Aeronautics
Netpresentvalue
Nitrogenoxides
Particulatematter
Performancepredictionprogram
Responseamplitudeoperator
RequiredminimumvaluefromLRs
Rules(whensubscript)
SOx EmissionControlArea
Sideforce
Specificfuelconsumption
Subjectivemagnitude
Sulphuroxides
Stillwaterbendingmoment
Twentyfootequivalentunit
(container)
Thrustreduction
Turbulencefactor
Verticalcentreofbuoyancy(from
keel)

Introduction

1. Introduction
1.1

Background

Shipping is currently estimated to contribute 3.3% towards total global emissions [Buhaug et al. (2009)]. Of
primary concern is carbon dioxide (CO2), a greenhouse gas (GHG) linked to global warming. An additional
considerationisthepredictedgrowthofshipping.Increasesinemissionsofbetween150%and250%by2050
areestimatedifmeasuresarenottakentoimproveenvironmentalsustainability[Buhaugetal.(2009)].
Whilstlegislationtoguaranteethelongtermenvironmentalsustainabilityofshippingisnotcurrentlyinplace,
theindustryiscertainlyaddressingthisissuefromadesignperspective,seeWrtsilShipPowerR&D(2009).
The adoption of the IMO Energy Efficiency Design Index [IMO (2005a)] as a means of assessing the
environmental performance of existing or future designs is a step towards environmental sustainability
becoming a key consideration in ship design and operation. However, economic viability remains the main
driverindesign.Reluctanceofshipownersandoperatorstoadoptnewtechnologiesorchangeoperational
procedure[Mash(2009)]couldslowprogressinthedevelopmentoflowcarbonshipping.
Regulation of other ship exhaust emissions is also tightening, characterised by the recent MARPOL Annex VI
amendments[IMO(2005b)].Theseglobalregulationslimitthelevelsofcertainemittedsubstances,including
sulphuroxides(SOx),whichcauseacidrain;andnitrogenoxides(NOx)andparticulatematter(PM),whichcan
worsenorcauseconditionssuchasemphysemaandcancer.
Thus there is a strong focus on the maritime industry to address the issue of its emissions, despite shipping
remaining the most carbonefficient form of transport [Buhaug et al. (2009)]. The challenge of the naval
architectisdiversifying,withpressurefrombothinsideandoutsideoftheindustrytoproduceenvironmentally
consciousships,whilemaintainingtheprofitabilityandfamiliarityexpectedbyowners.
Containershipsareoneofthelargestcontributorstoglobalshippingemissions[Buhaugetal.(2009)],largely
duetotheirhighspeed.Thiscanbeupto26knotsforthelargestships[WrtsilShipPowerR&D(2009)].The
efficiencyofscaleprinciplealongsideincreasingdemandofcontainerisedtransportationhasseenanincrease
intheaveragesizeofcontainerships.Thelargestarecapableofcarrying14,000twentyfootequivalentunits
(TEU), representing a high transport efficiency. Wrtsil Ship Power R&D (2009) (p.6) state that a 10%
increaseinshipsizewillincreasetransportefficiencybya45%.
Whilepostpanamaxshipsrepresenthightransportefficiency,smallerfeedervesselspresentanopportunity
forsignificantimprovementinbothefficiencyandsustainability.Theseshipsmakeupalargeproportionofthe
globalcontainershipfleet,anddistributecargofrommainlinehubstoregionalports,notaccessiblebylarger
vessels.Measuressuchasincreasingshipsize,improvingcargohandlingandemployingjustintimearrival
techniquesarejustsomeoftheefficiencyimprovementssuggestedforthisshiptype[WrtsilShipPowerR&D
(2009)].
Furtherimprovementsinefficiencycanbeachievedthroughtheuseofauxiliarypropulsionsystems.Theseare
particularly applicable to smaller ships, where the benefits can be greater. They include: sail systems, kites;
flettner rotors; and air bubble systems. While the latter is a newer technology, windassisted propulsion of
merchant vessels has been a popular topic of research and development in the past [Smulders (1985);
Bergeson&Greenwald(1985)]mainlyduetohighfuelprices.Whileinterestinwindpropulsionhasrecently
beenseentoreflectbunkerprices,itshouldberememberedthatinthenottodistantpast,windpowerwas
the primary means of merchant vessel propulsion. The use of windassistance devices is becoming more
commonasameansofreducingemissionsaswellassavingfuelcosts.Designsexistbothconceptually[NYK
(2010)]andoperationally[SkySails(2010)].Thustheapplicationoftechnologiessuchasthesehasthepotential
toimproveboththeefficiencyandenvironmentalsustainabilityoffeedercontainerships.

Concept Design of a Fast Sail Assisted Feeder Container Ship


Thisreportdetailsthedesignofsuchavessel,takingintoaccountbotheconomicandenvironmentalfeasibility.
A general approach in reducing emissions and fuel costs can be achieved by lowering operating speed, this
project adopts a rather different philosophy. That is that an increase in speed, rather than decrease, is the
answer to improving the efficiency of feeder services. By optimising ship design and operation, a viable
conceptissoughtthatcanmeetbothfuturetradepredictionsandenvironmentaltargets.

1.2

Aims and objectives

The aim of this project is to design a fast, sailassisted feeder container ship that can be considered
economicallyandenvironmentallysustainableinthefuturecontainertrademarket.Onlyoneiterationofthe
designspiralistobeconsideredduetothelimitedtimeavailable.
Thiswillbeachievedbyaddressingthefollowingmainobjectives,to:

Makeanassessmentoffuturecontainertradeandderivesuitableshipparticulars;

Investigate technologies and methods for improving both transport efficiency and environmental
sustainability;

Developtheconceptthroughtheuseofnumericaltechniquesandphysicalmodeltesting,focussing
onminimisingshipresistanceandthuspowerrequirement;

Complete a feasibility study to assess the proposed concept against existing ships and the
requirementsderivedintheinitialstagesoftheproject;

Conductafullconceptdesignstudytoassesstheimpactofthenoveldesignfeaturesutilisedonvessel
performance and prove the viability of the concept in all design areas. This will require considering
seakeeping,structuraldesignandstability.

The chapters that follow deal with each of these objectives in turn, which correspond to the natural
progressionofthedesignprocess.

1.3

Outcome and main achievements

Theresultoftheprojectisafastfeedercontainershipthathasbeendesignedtoenteroperationin2020.The
design performs favourably in terms of both transport efficiency and emissions levels compared to existing
vessels operating on the routes considered. Emissions reduction targets have been exceeded, implying
relevantlegislation(MARPOLAnnexVI)isalsocompliedwith.Thisisduelargelytotheuseofliquefiednatural
gas(LNG)asfuel,whichcontainslesscarbonthantypicalfueloils,andhaszerosulphurcontent.Thethrust
benefitprovidedbytheinstalledMultiwingsailsystemislowattheshipsservicespeedof25knots.However,
thefeasibilityofthedesignreliesonalowpowerrequirementat15knotsspeed,whichisinfactlowerthan
typical existing feeder ships,despite an increase in cargo capacity ofapproximately40%. The ability to self
berth and load/unload independently of shoreside cranes also plays a major factor in the viability of the
design. The unconventional design has also been rigorously assessed in terms of structural design, motion
analysesandstability,thusimprovingtherobustnessoftheproposedfastfeederconcept.

Design Sp
pecification
n

2. Design
D
Sp
pecificatio
on
2.1

Operating principlees

Theeefficienciesoffscaleprincip
plehasseentthesizeofco
ontainerships increaseoveerthepastdecade,atrend
d
that is set to continue. Thiss poses prob
blems to wo
orld shipping as the num
mber of portss capable off
mmodatingthesemegaco
ontainershipssislimiteddu
uetodraught,,lengthandccargohandlinggrestrictions..
accom
The solution to this problem liees with the operation
o
of mega
m
contain
ner ships on m
mainline EasttWest routess
1
ment of contaainers from the regional hub ports to
o
betweeen regional hub ports (Fiigure 2.1). The transhipm
smalleersatellitepo
ortsiscarriedoutusingsmaallerfeederco
ontainershipss,aconceptthatiscommo
onlyknownass
ahub
bandspoken
network[KoiYYuNg&Kee(2008)].

Figure2.1Primaryworldtraderroutes[NOAA
ASatelliteandInformationSService(2009)]
Ahub
bandspoke networkhasaanumberofaadvantagesovvertraditionaalnetworks.FFirstlyitactsaasameansto
o
transp
port cargo fro
om satellite ports
p
to a hub
b port for traanshipment to
o a mainline shipping routte. Secondly,,
becau
use it is quickker for a conttainer to be transhipped
t
through a hub
b port than to
o wait for a direct
d
ship, itt
provid
des quicker interregional transport off cargo between satellite ports. Geneerally, it meaans far fewerr
journeeys are requ
uired, which makes hub and spoke networks eco
onomically and environmentally moree
sustaiinable.Foreexample,conssideraregionwithsixteenregionalportts.Withoutttranshipmentt,ifeachportt
conneectedtoeveryyotherporttthenumberofjourneysisrrepresentedb
bytheleftof Figure 2.2.H
However,with
h
transh
hipmentthessamenumberofportscouldbeconnecte
edmoreefficiientlyasshow
wnintherighttofFigure2.2
2
[PSASSingaporeTerrminals(2009))].

Figu
ure2.2Shipp
pingnetworkwithout(left))andwith(rigght)transhipm
ment[PSASinggaporeTermin
nals(2009)]
Theadvantagesof transhipmentthroughhub
bportsmeanttthetranship
pmentmarketttripledover theperiodoff
nsultantsLtd((2006)]atrendthatissetttocontinuem
makingitanin
ntegralpartoff
19952005[Ocean ShippingCon
thefu
uturecontaineershippingmaarket.

1
Tran
nshipmentisthemovementofcontainerrstoaninterm
mediatelocationbeforebeiingmovedtotheir
requirreddestinatio
on.

Concept Design of a Fast Sail Assisted Feeder Container Ship

2.2

Region analysis

An initial assessment of the current container ship market was made, using various research papers,
professionalsocietypublications(suchasTheNavalArchitect)andcontainershipoperatorwebsites.Seven
regions(theCaribbean,Mediterranean,MiddleEast,FarEastaswellasSouthEastAsiaandtheEastandWest
coastsoftheUSA)wereidentifiedaspotentialareaswhereahubandspokenetworkcouldbeapplied.The
EastandWestcoastsoftheUSAwerelaterrejectedduetothestrongcompetitionfromlowcostrailtransport,
making the use of feeder container ships less economically and environmentally viable. Within the five
remainingregions,arepresentativesampleofthe60largestportswereselectedforfurtheranalysis.
Inparticular,potentialhubportsforthe2020containershipmarketwereidentifiedbasedontheminimum
combineddistancetoeveryotherportintheregion.Thisminimisesthetotaldistancethatcargoneedstobe
transported,improvingthetransportchainwhenconsideredasawholeratherthanonaroutebyroutebasis.
For a more detailed analysis, consideration needs to be paid to the location of the hub port in relation to
mainline services in order to optimise the entire transport chain. For simplicity only the location of the port
withintheregionhasbeenconsideredinthisanalysis.AsummaryoftheregionalhubportsisgiveninTable
2.1.
Table2.1Identificationofregionalhubportsbyminimumcombineddistancetoallsatelliteports
Region
SouthEastAsia
Caribbean
FarEast
Mediterranean/MiddleEast

Hubport
Singapore2
Kingston3
Busan
GioiaTauro4

Country
Singapore
Jamaica
SouthKorea
Italy

Totaldistance/n.miles
11836
7870
5033
13014

Inselectinghubportsconsiderationwaspaidtoselectingportsthatalreadyhadsufficientcapacitytohandle
largevolumesofcargo.Thisassumptionisimportantasitisdeemedimpracticaltobuildapurposebuiltport
facilityfromscratchwithintheconsideredtimeframe.Itwouldbeexpectedinthelongertermthathubport
facilitieswouldbecustombuilt(eitheronlandorfloatingoffshore)inoptimumlocationstooptimizetheentire
transportchain.

2.3

Port analysis

2.3.1

Throughputpredictions

In order to dimension a suitable ship for the 2020 market it is necessary to gain an understanding of the
containershipmarketatthattime.Inordertoassessthis,historicportthroughputdata(intermsofTEU)was
obtainedfromDegerlund(2004;2006;2008)aswellasportwebsitesforaperiodbetween1995and2008for
the ports identified in Section 2.2. A linear regression analysis was carried out and used to predict port
throughputin2020foreachofthe60portsbeinginvestigated.Thepercentageincreaseon2009levelshas

Ocean Shipping Consultants Ltd (2006) found that Singapore has a 73.3% share in the South East Asia
transhipmentmarket.ContainerthroughputatSingaporeroseby151%over19952005witha113%increase
intranshipmentoverthesameperiod.Thisprovidesadditionaljustificationforitsselectionasahubport.
2

KingstonisidentifiedasahubportbyFrankel(2002)whoinvestigateshowtheuseofmegacontainerships
onmainlineserviceswillaffecttransshipmentintheCaribbean.
4

GioiaTauroisidentifiedasahubportbyFrancesetti&Foschi(2002)whodiscussestheuseofhubandspoke
networksintheMediterranean.

Design Specification
beenaveragedonaregionalbasisandispresentedinTable 2.2.Theauthorhasusedhisdiscretiontoremove
itemsofdatathatappearedtobeanomalouswhenmakinglinearregressionestimates.
Table2.2Predictedcontainershipmarketgrowthby2020
Region
Caribbean
Mediterranean
FarEast
SouthEastAsia
Average

Predictedpercentageincrease
83.12%
100.37%
159.33%
67.39%
102.55%

Thehistoricalthroughputofcontainersforthecitedregionstogetherwithpredictionsto2020areprovidedin
Figure2.3.
TEUThroughput/MillionTEU

200
180
FarEast

160
140
120
SouthEsAsia

100
80

MiddleEast

60
40

Mediterranean

20

Caribbean

0
1995

2000

2005

2010

2015

2020

2025

Year
FarEast(Historical)
MiddleEast(Historical)
Caribbean(Historical)

FarEast(Prediction)
MiddleEast(Prediction)
Caribbean(Prediction)

SouthEastAsia(Historical)
Mediterranean(Historical)

SouthEastAsia(Prediction)
Medierranean(Prediction)

Figure2.3Predictedgrowthinregionalcontainershipmarketsby2020
TheseestimatesofcontainershipmarketgrowthareconsistentwithresearchofOceanShippingConsultants
Ltd(2006)whichpredicteda102126%increaseinworldcontainerthroughputover20042015andafurther
1824%increaseover20152020.
TheestimatesofTable 2.2containerthroughputby2020arederivedusingalinearregressionanalysis.Thisis
deemed conservative compared to estimates made by other some authors [Cheng (2009)], who base their
estimatesuponanexponentialgrowthrate.Thepredictionisalsorathersimplisticinthattheestimateisnot
linked to the growth oftheworld economy (or the recent financialcrises), nor does it accountfor scenarios
thatchangetradepatterns(suchastariffs)orsuddenshockstothetradepatterns.

2.3.2

Portrestrictions

One of the key parameters in the design and operation of feeder container ships is the ability to operate
betweensmallsatelliteportsthatmayhavelengthanddraughtrestrictions.Inordertodeterminetheport
restrictiondimensionalrestraints,maximumberthlengthandwaterdepthdatawascollectedfromappropriate
portwebsitesandDegerlund(2004;2006;2008).AsummaryofthisresearchispresentedinTable2.3.

Concept Design of a Fast Sail Assisted Feeder Container Ship


Table2.3Summaryofportrestrictions
Draught/m
5

10

12

13

14

15

6
7
8
9
10
12
13
14
15

No.ofports
0
0
1
4
5
8
5
13
4
9

%ofports
0.00
0.00
2.04
8.16
10.20
16.33
10.20
26.53
8.16
18.37

Berthlength/m no.ofports %ofports


160 170
2
4.08
170 180
1
2.04
190 200
0
0.00
210 220
0
0.00
220

46
93.88

FromdiscussionswithChristianMashofBorchardLinesLtd,afeedercontainershipoperator,itwasidentified
that shorter beamy ships have advantageous operational benefits over larger ships. In ports with length
restrictions,twolongshipscannotberthatthesametime.Consequently,smallershipsareallowedtojump
thequeuemakingthemmoreprofitable.Thisimprovesefficiencybyreducingportwaitingtimes.
A third port restriction that can restrict a ship size is air draught. Due to the relatively small size of the
proposed ship this constraint is viewed as less onerous, although insufficient data is available to warrant a
completeinvestigation.

2.4

Route analysis

TheDatloyvoyagemanagementsystemusedbymajorshippingcompaniesworldwidepermittedvoyageroutes
between all of the satellite ports and theircorresponding regional hubport tobe investigated. The Dataloy
voyage management system holds data on more than 7,200 ports and maritime locations, and over 69,000
waypoints to plan ship voyages accurately [Dataloy (2009)]. Hence it was possible to assess the distance
between ports and the distance travelled at different headings on a particular route by post processing the
longitudeandlatitudepositionofthewaypointsalongtheroutesconsidered.Thisdatapermitsidentification
ofasuitableshiprangeandallowssailingperformanceestimatesfortheshipfromtheanticipatedapparent
winddirection.
AsummaryoftherequiredshiprangesfromacitedhubporttothesatelliteportsisprovidedinTable2.4.
Table2.4Summaryofrequiredshiprangeforcitedhubports
Range/nm
No.ofroutes
0
16
1000
1000 1500
12
1500 2000
11
2000 2500
8
2500 3000
3
3000 3500
2
3500 10000
0

%ofroutes
30.77
23.08
21.15
15.38
5.77
3.85
0.00

Cum.sum
30.77
53.85
75.00
90.38
96.15
100.00
100.00

ThedistancetravelledataparticularheadingfortheCaribbean,SouthEastAsiaandNorthAtlanticroutesis
summarised in Table 2.5. The Caribbean and South East Asia are presented as the results of the initial
economic analysis (discussed in Section 2.6) these regions are the most viable for the exploitation of
transhippment.

Design Specification
In order to also address the feasibility of sail assisted ship propulsion it is pertinent to investigate the ship
performance in areas of the world with more favourable wind conditions. For this reason, twelve north
transatlanticroutesfromvariousportsontheEastcoastofCanada,USAandtheCaribbeantoRotterdamand
Gibraltar have been considered. Within these routes, the route that provides the most favourable wind
strengthandapparentwindangleshallbeselectedtoassesstheshipperformanceinidealconditions.For
thisreasonTable2.5alsocontainssummaryroutedataforNorthAtlanticroutes.
Table2.5Distancetravelledonagivenheadingfortheregionsbeingconsidered
Heading
/degrees
029
3059
6089
90119
120149
150179
180209
210239
240269
270299
300329
330359

Distancetravelled/nm
SouthEastAsia Caribbean
827
1120
1330
2307
1281
1630
1163
1058
2424
366
3377
781
827
1134
1330
2307
1281
1630
1163
1058
2424
366
3377
781

Combinedtotal
/nm

%oftotal

1947
3637
2911
2221
2791
4158
1961
3637
2911
2221
2791
4158

5.51
10.29
8.24
6.28
7.90
11.76
5.55
10.29
8.24
6.28
7.90
11.76

NorthAtlantic
distancetravelled
/nm
1332
1265
11463
23043
2257
4301
1334
1265
11463
23043
2257
4301

%of
total
1.53
1.45
13.13
26.39
2.58
4.93
1.53
1.45
13.13
26.39
2.58
4.93

The cascade principle in the operation of container ships says that a new container ship will operate on
mainline services until being replaced with a new ship. Once replaced, the old ship will be used on feeder
routes until it becomes uneconomical and is scrapped. This progression of container ships from mainline to
feeder to scrapping makes feeder ships older and less efficient on average. Hence it is an ideal area of the
markettomakeefficiencysavings[Tozer(2009)].
Fromananalysisof174servicesacrossthefiveregionsbeingconsideredstatisticsregardingtypicalshipsize,
voyagetime,speedandageservingeachportwereinvestigated[Degerlund(2004;2006;2008)].Asummary
oftypicalageofshipsoperatingontheroutesbeingconsideredisshowninTable2.6.Otherstatisticsgathered
inthisexercisearediscussedinSection2.6.
Table2.6Breakdownofbasisshipsbyyearbuilt
No.of ships
Yearbuilt
1960 1970
1
1971 1975
4
1976 1980
16
1981 1985
20
1986 1990
9
1991 1995
20
1996 2000
49
2001 2005
29

%of ships
0.68
2.72
10.88
13.61
6.12
13.61
33.33
19.73

Ifcontainershipservicelifeis30yearsthenby2020shipsconstructedbetween1990and2000,thatis46.94%
of the feeder fleet, will need replacing. Assuming these ships carry an equivalent share of the total TEU
throughput,theamountofcargotheproposedfeedershipwillhavetocarryinatypicalyearcanbeestimated
by assuming it will replace a certain percentage of these scrapped ships. Initially a 20% target share of the

Concept Design of a Fast Sail Assisted Feeder Container Ship


2020 new build market has been set, which gives a target market share in the 2020 feeder container ship
marketintheorderof10%.Thisassumptiontakesnoaccountofthevariationsinscrappingandbuildrates
overtime,whichisheavilyinfluencedbythegrowthordeclineoftheglobaleconomy.

2.5

Operational analysis

Further discussions with Mash (2009) of Borchard lines Ltd provided insight regarding the feeder container
shipsrequirementsfromanoperatorsperspective.
Fuelefficiencyiscrucialnotjustbecauseofrisingfuelcosts.Theamountoftonnageavailabletocharterallows
operatorstoselecttheshipsizethatmostsuitstheirmarketsandtheirservicephilosophy.Generally,more
frequentservicessatisfyshippersdemandsforashorttransittime.Bigshipsareonlyeconomicalifyoucanfill
them.Thus,shipsize,speedandfuelefficiencyarecrucialandwillvaryheavilyfromroutetoroute.
Congestion inbusyports can be up to fivedays and is amajor problem for operators. Typically on a 28 day
voyage20daysisspentatsea,eightdaysinportandusuallytwotothreedaysislefttoaccountforcongestion.
Evenwiththisallowance,shipsstillneedtosailfasterthantheirdesignspeedtocatchup,whichincreasesfuel
costs.Badweatherhassignificantinfluencesontheabilitytomaintainschedulesandmanyshipsareworse
thanothersatkeepingspeedinroughweather.Ideallyshipsneedtobeabletooperateatconstantspeedup
to force five. Currently some ships reduce speed to less than five knots in bad weather when other similar
sized/poweredshipscanstillproceedatupwardsof12knots.Shipsspend75%oftheirtimeataservicespeed
of90%MCRandsometimesoperateatreducedspeede.g.incaseofgoodweather,fastturnaroundinportetc.
andthenMCRisatleast70%becauseanylowerthanthisenginesincurproblems.Inwinter,shipswillspenda
greaterproportionoftheirtimeatfullspeed.Performanceofashipovertimeismonitoredandifashipdoes
notperform,thenitisreplacedbymoreefficientdesigns.Someshipyardsinvestlittlemoneyintoachievingan
efficienthullandpropellerdesign.
InordertocomplywithSOxEmissionControlAreas(SECAs)shipsneedtobeabletoburnbothhighandlow
sulphurfuels.Currentlyownersareunwillingtoinvestinfittingscrubberstoreduceemissionsgiventhecost
versusreturn,becausetheydonotpayforthefuelandtimechartersarenotlongenoughtoachieveasaving.
Regular bottom cleaning and low friction coefficient paints are regularly used to improve efficiency and can
improve performance by two to three nautical miles per ton. Sails and other structures that protrude on a
containershiptendtogethitanddamagedduringcargooperationsandanythinghightechtendstogowrong
andcostsmoney.Ifthecrewcaneffectrepairswithouttheneedforshorebasedtechnicianstojointheship
thenthissavesoncostsanddelays.
ShipsthataremanoeuvrablewithbowandsternthrusterssaveontugusagewhichcanbethousandsofEuros
a time. Ships come under different rules depending on the gross tonnage. For instance, ships with a gross
tonnagegreaterthanfifteenthousandtonnesrequirepilotstopasstheMessinastraitsaddingcostsanddelay.
Shipsneedtobeabletotakehighcubecontainersinallpositionswithoutlossofslots,withanabilitytoload
cargoformorethanonedestinationineachbaybutabletodischargeinanyorder.Openhatchshipsareideal
for this. Otherwise split hatch covers are used. Container stacks need to be capable of supporting high
containerweightswhicharenotreducedbecauseofrackingforces.Thisisadrawbackofopenhatchdesigns.
Forexampleifyouloadsixhigh30metrictonne20ftcontainersthenthebottomcontainercancollapse.

Design Specification

2.6

Economic and environmental analysis

Akeycriterionincontainershipoperationistodeploytherightsizeshiptocaterforthedesiredmarketwhile
minimisingexcesscapacity,sincetherightsizetonnagehassignificantimplicationsonfreightrate.Thereisno
one ship to suit all markets. Thus, in order to determine a suitable capacity and speed for the fast feeder
conceptananalysiswasconductedencompassingalltheresearchdiscussedinSections2.2to2.5.
Somekeyassumptionshadtobemadeinitiallytocontrolshipoperationssoastooptimisethetransportchain
efficiency as a whole, and not just on a ship by ship basis. The assumptions made can be summarised as
follows,that:

In2020allcontainersinaregionaretranshippedthroughahubportwithdirectreturnfeederservices
fromhubportstosinglesatelliteports,i.e.therearenodirectservicesbetweensatelliteports,circular
routesorbacktrackingroutes.Frankel(2002)foundhubandspokenetworkstobethequickestand
mosteconomicallyefficientmethodoftransportingcargo;

Thenumberofsailingsperweekistoremainconstantat2009levels,tomaintainthesamelevelof
consumer service. The number of sailings is calculated by assuming that operators switch with
immediate effect to a hub and spoke network with no services between satellite ports (i.e. all
containersaretranshippedthroughahubport).Thisassumptionwillessentiallyincreasethesizeof
feedercontainershipsinlinewiththeincreasedgrowthofthecontainershipmarketspredictedbythe
portanalysis;

The number of ships required on a regional basis is to be halved in order to reducecongestion and
timespentwaitingtogetintoport.Thisistobeachievedbyincreasingshipspeedandporthandling
capabilities;

The proposed ship will not experience congestion and will control the loading and unloading of all
containerstominimisedelays;

Theinitialtargetmarketsharewillbetocarryapproximately10%ofthetotalcontainerthroughputon
allroutes.

Inordertodeterminesuitabledimensionsfortheship,the2009TEUthroughputwasdividedbytheaverage
ship size on a given route to set the number of sailings, assuming a 90% ship utilization [Noteboom &
Vernimmen(2009)].ThemarketgrowthandtargetmarketsharewasusedtocalculaterequiredTEUcapacity
perweekperroute.Ontheassumptionthenumberofsailingsaweekonarouteistoremainconstantanew
shipsizecouldbedetermined.Ontheassumptionthat:

theshipwillhavetomakeatleasttwovoyagesinthetimeittakesacurrentshiptomakeone;

theroutelengthisknown;

anassumedunloadingtimecanbemadebasedonshipsize;

theshipspeedcanbedetermined.Sincethecongestionandunloadingtimesvariesfromporttoporttheexact
number of basis ships the ship will replace is adjusted using the authors discretion to yield realistic and
achievablevalues(i.e.replacing3basisshipsfor1conceptship,or3basisshipsfor2basisshipsinsteadof2for
1 etc. on some routes). The process followed by the author to determine the ship capacity and speed is
presentedgraphicallyinFigure 2.4.Theboxesingreyindicateinputstotheprocesswhereasthehighlighted
boxesindicateprocessoutputs.

Concept Design of a Fast


F
Sail Asssisted Feedeer Container Ship

Figure2.4Flowcha
2
artusedtodeeterminetherrequiredship capacityandspeedforapaarticularroute
e
Theprocess showninFigu
ure 2.4wasappliedtoeachroutetode
etermineanoptimumship sizeandspee
edthat
ptions.Statisticswerethengeneratedttoindicatetheeproportiono
ofshipsofaccertain
satisfiedthe listedassump
marisedinTab
ble2.7andTable2.8.
sizeandspeeed.Theseressultsaresumm
Tab
ble2.7Summaryofrequiredspeed
Speed/kn
nots No.of ships
s
0 16
39
9
16 17
5
17 18
12
2
18 19
4
19 20
16
6
20 21
8
21 22
1
22 23
3
23 24
0
24 25
1
25 26
0
26 27
8
27 28
0
28 29
2
29 30
0
35 36
1

%of ships
s
38.8
89
5.0
05
12.1
12
4.0
04
15.6
66
8.0
08
1.0
01
3.0
03
0.0
00
1.0
01
0.0
00
8.0
08
0.0
00
2.0
02
0.0
00
1.0
01

Cum. sum
s
38.8
89
43.9
94
56.0
06
60.1
10
75.7
76
83.8
84
84.8
85
87.8
88
87.8
88
88.8
89
88.8
89
96.9
97
96.9
97
98.9
99
98.9
99
100.0
00

Arathersimplisticestimatteofshipspeedassumesthattheshipaadvancesatitsoperationalspeedfromleaving
Clearlyapropo
ortionofthe timeisspentatslowspeeeduponleavin
ngand
porttoarrivingattheirdestination.C
henmanoeuvvring.Thus,ttheservicesp
peedoftheproposedshipswillbe marrginally
enteringporrtandalsowh
larger than the
t values qu
uoted in Table 2.7. Usingg all the inforrmation discussed and thee ship require
ements
analysis for specific routees it was conccluded that a
a service spee
ed of 25 knotts would be aappropriate. This is
nroutesofintterest.
approximateelytenknotsfasterthanexiistingshipson
Theshipcap
pacityofTablee 2.8suggestssthat1200to
o1300TEUw
wouldbesuitaableforalarggeproportion ofthe
2020contain
nershipmarkeet.Thusthisiisthetargetccapacityforth
heproposedn
newdesign.

10

Design Specification
Table2.8Summaryofrequiredshipcapacity
Capacity/TEU
No.of ships
0
0
100
100 200
0
200 300
2
300 400
0
400 500
1
500 600
1
600 700
2
700 800
2
800 900
6
900 1000
2
1000 1100
10
1100 1200
9
1200 1300
11
1300 1400
6
1400 1500
4
1500 1600
2
1600 1700
2
1700 1800
2
1800 1900
0
1900 6000
40

%of ships
0.00
0.00
1.98
0.00
0.99
0.99
1.98
1.98
5.94
1.98
9.90
8.91
10.40
5.94
3.96
1.98
1.98
1.98
0.00
39.11

Cum. sum
0.00
0.00
1.98
1.98
2.97
3.96
5.94
7.92
13.86
15.84
25.74
34.65
45.05
50.99
54.95
56.93
58.91
60.89
60.89
100.00

Detailed analysis of the South East Asia and the Caribbean routes revealed that the optimum ship size was
similar.TheoptimumshipsizesfortheMediterranean/MiddleEastandFarEastroutesarecomparable,buta
lotlargerthantheshipfortheCaribbeanandSouthEastAsiamarkets.Asaconsequenceoftheseverydistinct
differencesitwasdecidedtofocusondevelopingaconceptualdesignfortheCaribbeanandSouthEastAsia
markets.
ThebasisshipforaparticularoperationalareawillbedesignatedCaseA.Theoptimaldesignwillbereferred
toasCaseB.Ineachcase,andforeachregionofinterest,itisnecessarytoassesstheaverageweeklyCO2
emissionsexpectedfor2020,whentransporting10%ofthetotalregionalcontainerthroughput.Acomparison
oftheemissionsfromtheproposedoptimumandbasisshipisgiveninTable2.9.
Table2.9PredictedCO2emissionsin2020
Region
SouthEastAsia
Mediterranean
Caribbean

MilliontonneCO2 perweek
CaseA CaseB %increase
976
1262
129.32
609
982
161.40
243.6
294
120.71

Table 2.9indicatesthatbyreducingtheshipnumbersgivesariseinCO2ofbetween20%and60%duetothe
increase in speed. However, these increases are expected. The aim of this investigation is to reduce
congestion and improve transport efficiency, a strategy that does not reflect well on a tonnes of CO2 per
nauticalmilemeasurementscheme(suchasthatusedfortheEEDITheriseinemissionsfromincreasingspeed
willberecoveredusingforexample,auxiliarypropulsionintheformofrigidwindsails;lowfrictionantifouling
coatingsandalternativefuels.
The predictions of Table 2.9 are rough estimates of average ship speed over the route distances, but are
adequateforaninitialcomparisonofthetwooperationalstrategies.Thereisinsufficientbasisshipdataon
routes in the two regions not presented (Far East and Middle East) to make a suitable analysis but it is

11

Concept Design of a Fast Sail Assisted Feeder Container Ship


suspectedduetothenatureofthecontainerlinebusinessthatthetrendwillbesimilartotheMediterranean.
ThegreaterincreaseinemissionsassociatedwithMediterraneanroutesislinkedwiththelargeroptimumship
size.
AcomparisonofthenumberofshipsrequiredbythetwoassumptionsisgiveninTable2.10.
Table2.10Numberofshipsrequiredin2009and2020byregion
Region
SouthEastAsia
Mediterranean
Caribbean

CaseA
120
80
28

No. ofships
CaseB %difference
44
36.67
45
56.25
12
42.86

ItisevidentfromTable 2.10thatasignificantreductiononshipnumberscanbeachievedwiththeproposed
operationalassumptions.Furthermore,Table 2.10indicatesthattheshipreductioninSouthEastAsiaandthe
Caribbeanismuchgreaterthanthe50%settarget.Thelikelycauseforthisistheaverageshipsizeissmaller
andolderthantheotherregionsbeingconsidered.Thiswasdiscoveredbyanalysisoftheroutespecificbasis
shipdata.Duetolackofportregulationetc.containershipoperatorsusetheirolderandmoreinefficientships
onSouthEastAsiaroutes.Thismeansthatgreatersavingscanbemadeintermsofoperationalefficiencyby
replacingtheseships.IntheMediterranean,thereverseistrue.Thetighterregulationsgiverisetoshipsthat
are larger, newer and more efficient. This makes it much harder to producethe same efficiencysavings. A
summaryoftheaverageshipsizesoperatingintheregionsbeingconsideredcurrentlyandpredictedin2020
aregiveninTable2.11.
Table2.11Averageshipsizebyregion
Region
Averageshipsize/TEU (2009)
SouthEastAsia
890
Mediterranean
1547
Caribbean
955

Averageshipsize/TEU(2020) %increase
1303
146
2161
139
1088
113

2.7

Summary

Theproposedfastfeederdesignwillhaveaservicespeedof25knotsandacargocapacityof1300TEU.Thisis
deemedtobesuitableforalargeproportionoftheroutesconsideredinthisinvestigation.Itisevidentthat
theconceptismostsuitedtotheSouthEastAsiaandCaribbeanregions,wheregreaterefficiencysavingscan
bemadeduetothelengthoftheroutes(upto1500miles)andtheinadequacyoftheshipscurrentlyserving
theseroutes.Theseregionswillbethefocusoftheremainderofthisinvestigation.Furthermoretheuseof
sailassistedshipsontransatlanticpassageswillalsobeconsidered.

12

Initial Development of Fast Feeder Container Ship

3. Initial Development of Fast Feeder Container Ship


3.1

Basis ships

In any design study it is important to assess existing ships undertaking similar operations. In Chapter 2
economicandoperationalassessmentsofexistingshippinglinesfacilitatedtheidentityofafeasibleconcept.
Thereafter, basis ships are a good source for the preliminary estimation of a number of basic design
parameters. A database of around 170 ships has been created from data primarily sourced online [van
Duivendijk (2009) and Svendsen & Tiedemann (2009)] and secondarily from Degerlund (2004; 2006; 2008).
Whilstonlinesourcesprovideaplentifulsupplyofdata,theyareoftenincompleteandpotentiallyinaccurate.
However,asthebasisshipdataisonlyusedtoprovideinitialestimates,theexistenceofinaccuraciesissimply
noted.Discretionhasbeenusedtoilluminateorcorrectdatathatisobviouslyincorrect.Somemorereliable
sourcesofdatahavebeenfoundinpublishedtechnicalpapers[Kimetal.(2003),Sipila&Brown(1997)and
BMTNigelGee(1998)].Theseprovidedetailedtechnicalinformation,includingindicationsoflayout,however
thesesourcesarescarceandthusdonotconstitutealargeproportionofthedatabase.

Informationextractedfromsourcesincludes,butisnotlimitedto:

Cargocapacity(

Maximumcapacityofrefrigeratedunits(Reefers);

Lengthoverall(

Beam( );

Draught( );

Maximumspeed;

Operatingspeed;

Yearbuilt;

);
);

Installedpower(

Machinerytypes;

Deadweight(

);
).

Dimensions and capacities given are generally maximum values allowing for an estimation of overall
dimensions.Itisunfortunatethatoperatingspeedisonlygivenforanumberofships,asthisisofparticular
interest.Primarilythebasisshipdatahasbeenusedtoestimatetherequiredprincipaldimensionstofulfilthe
capacity requirement identified in the initial economic evaluation. The basic dimensions will primarily be a
, and against ,(FigureA.1
functionofthecargocarryingcapacityofanycontainership.Byplotting
andFigureA.2),itisclearthetrendislinear.

/ and /
Itisnotunusualatthisstagetousedimensionlesshullformparameters.Howeverifoneplots
against capacity there is little variance and hence the correlation of the linear regression is weak. Thus this
approachisavoided.TheresultsoftheseregressionsaresummarisedinTable3.1.
Table3.1Summaryofregressionanalysisresults
/
/
/
/
/
/

0.04391
0.005608
0.002236
0.093637
14.547
1.671

13

105.5358
17.898958
6.161147
4.738531
529.94
707.4

162.619
25.189
9.070
7.079
18381.160
25401.980

Concept Design of a Fast Sail Assisted Feeder Container Ship


Asexpectedthetrendsofthedatasetssuggestthatasthecargocapacityofacontainershipincreasesthenso
mustallofitsprincipaldimensions.Whenincreasingthedimensionsoneshouldtakenoteofanyoperational
restrictionsplacedontheshipsuchasdepthinports,lengthofberthsandminimumairdraught.Theinfluence
ofchangingtheparametersuponshipperformanceneedtobeappreciated,andhencetheeffectonefficiency
andcosts(includingthoseassociatedwithconstruction).

Mostcontainershipshavesimilarhullformtopology.Withintheregressionanalysesthisfactisreflectedby
thenumberofdatapointsnearthemean.However,shipswithexcessivebreadthordraught(Figure A.1and
Figure A.2),comparedtotheircargocapacityexist.Thisisinterpretedasinefficientdesignortheimpactof
restrictionsontheotherprincipaldimensionsleadingtoagrowthinunrestrictedparameters.

/ .This
Inadditiontocapacity,thehullformshapeisafunctionoftheshipoperatingspeed,particularly
isduetominimisingresistanceofahullform,seeSection 4.1.Generallyfasterships,forwhichwavepattern
/ to minimise this
resistance constitutes a larger proportion of the total resistance, will have larger
componentofresistance.ThisisevidentinFigure A.2withthelowerspeedsethavingthisgeneraltrendand
/ . Ideally the slenderness ratio
/
the higher speed ships showing a more significant increase in
wouldbeusedbutalmostallofthebasisshipdatalacksinformationregardingthedisplacementandwaterline
length.RegressionofthedatainFigureA.2leadstotheresultspresentedinTable 3.1.Onecancomparethe

and giving
6.456.
resultfromthisregressiontotheresultobtainedfromtheregressionsfor
Thereisan8.8%decreasefromtheoriginalresult,suchadifferenceisexpectedwhendealingwithregression
analysisofdata.

Estimatingthemassofashipisimportant.Fromthebasisshipdatabaseinformationregardingdeadweightis
abundant.Asdeadweightisalmostpurelyafunctionofcargocapacityitcanbeestimatedusingtheregression
formula of Table 3.1. Similarly the installed power can be estimated by deriving a regression equation
dependentuponcubeoftherequiredmaximumspeed.ThisgivesthelineartrendofFigure3.1.Itisimportant
tonotethesignificanceofthebuildyearfieldwhenfilteringthisdata.Shipsinthedatabasebuiltinthe1970s
and1980s,whichmustbenearingorhavecometotheendoftheiroperationallife,tendtohavelargeinstalled
powersforthespeedachievedandthecapacityofcargocarriedwhencomparedtomoremodernships.This
indicates inefficient ship design. These data points have been eliminated from this analysis as they do not
providedataofrelevancetofuturedesigns.
25000
y=1.671x 707.4
20000

PB /kW

15000

10000

5000

0
0

2000

4000

6000

8000

10000

VS 3 /knots3
Figure3.1PBasafunctionofthecubeofshipspeed(Vs3)

14

12000

Initial Development of Fast Feeder Container Ship


Thelowerspeedshipsof15knotscannotbeexcluded,sincedatarelevanttothe25knotdesigntargetarevery
limited,see[Kimetal.(2003),Sipila&Brown(1997)andBMTNigelGee(1998)].

Further analysis could be carried out using the basis ship data. However it is important to remember the
approximatenatureofthisanalysisandhenceitsapplicability.Thederiveddataaretobeusedtoaidfurther
developmentofthedesignbutwillnotbeusedasabsolutevaluesfordesignparameters.Absolutevalueswill
be determined using technical and logical design methods. In short, extensive basis ship analysis is not an
efficient use of time. The basic design parameters to be used are those already suggested by the initial
economicanalysisnamely,shipspeedandcargocapacity.

3.2

Initial mass, powering and stability estimates

Theprocessesinwhichpreliminarydimensionsaredeterminedforacontainershipareverydifferenttothatof
aconventionalship.Sincecontainershipscarryasignificantproportionoftheircargoondeck,cargovolumeis
indeterminate,whichmeansitisnotpossibletobasethedesignontherequiredcargovolume.Thedesignis
principally driven by stability considerations, which controls the limiting height to which containers can be
stacked. To help stability and maximise capacity the top container slots are reserved for lightly loaded
containers.Heavycontainersarelocatedlowerdownandtypicallyballastwateriscarriedpermanentlytohelp
improve stability. For a given stacking height there will be a minimum ship breadth in order to guarantee
stability.Withinthis,toensurelongitudinalandtorsionalstructuralstrength,aproportionofthebreadthhas
to be devoted to structural decks. Thus the number of container tiers determines the number of container
rowsinthebreadth.Lengthisdeterminedbythetechnicallydesirablelengthtobreadthratioandalsolimiting
port constraints. These dimensions are heavily influenced by ship speed, because of its effect on CB and
installedpower[Watson&Gilfillan(1977)].
Withthisinmindthedeterminationoftheshipdimensionsisprincipallyfoundtoreflectmultiplesofcontainer
height,widthandlengthwithanallowanceforsailstowage,machinery(determinedfrombasisships),foreand
aft peak volumes, bilge volume and fuel volume. The requirement for hydrostatic equilibrium is heavily
dependentonshipdimensionsandcarryingcapacity.
Usingtheinitialprincipalparticularsasastartingpoint,aniterativeapproachisadoptedtodeterminesuitable
dimensions.Theapproachcansummarisedasfollows:
1.

2.
3.
4.
5.
6.

Definetheshipbeamanddepthasafunctionofcontainerdimensionsandthenselectshiplengthin
termsofthenumberofcargoholdsrequiredtoachievetherequiredcapacitywithconsiderationto
relevantportrestrictions;
Estimateshippowerrequiredtoobtaindesignspeedusingbasisshipdata;
EstimateshipmassusingtheWatsonandSchneekluthempiricalmethodsaswellasbasisshipscaling
methods;
Demonstrateabalanceofweightandbuoyancyatadraughtwithintheportrestrictions.
Ensuretheshiphasadequatestability;
Alterdimensionsandrepeatuntilasatisfactorydimensions,balanceofforcesandstabilityisachieved.

Whatfollowsisabriefdescriptionofsomeofthemethodsemployedinthecitedsteps.

3.2.1

Massestimate

Aninitialestimateofshipmasswasmadeduringeachiterationusingthreedifferentmethods,namely:

15

Concept Design of a Fast Sail Assisted Feeder Container Ship

A general empirical estimate using the method described by Watson & Gilfillan (1977) based on
Lloydsequipmentnumeralwithconstantsprescribedforcontainerships;

A more detailed empirical estimate described by Schneekluth & Bertram (1987) to estimate the
lightshipmassofcontainerships;

By scaling of basis ships with corrections allowing for a change in dimensions and a change in
scantlings.

Thetwoempiricalmethodsuseregressionequationsbasedonstatisticaldatafrombasisshipstoprovidean
estimateoflightshipmassforvariousweightgroups,whichincludethe;

Weightofhullsteel;

Weightofsteelsuperstructureanddeckhouses;

Weightofequipmentandoutfit;

Weightofengine(propulsionplant);

Weightofcranes;

Weightmargintoaccountformaterialtolerance,designdetailsandtolerancesindesign;

Total deadweight including payload, ballast water, provisions, fuel, lubricants, water, persons and
personaleffects.

The method given by Schneekluth & Bertram (1987) is more detailed and more specifically tailored to the
designofcontainershipsandincludesestimatesfor:

Lashings,cellguidesandrefrigerationequipment;

The breakdown of engine mass into separate masses to account for engine, gearbox, generators,
propeller,shaftingandmiscellaneousitems;

Themassoftanks;

Thedependenceoftheshipmassonsizeandtypeofcontainercarried(TEU,FEU,reeferetc.).

When employing empirical methods in ship design it is important that the ship particulars are within the
prescribedlimitsoftheoriginaldatacollected.Generallytheformoftheproposedshiplieswithintheselimits,
exceptforthehighFroudenumberrequired.However,intheabsenceofmoreaccurateempiricalmethodsthe
proceduresfollowedhereweredeemedadequateasaninitialestimate.Fortheauthors,thegreatestconcern
ofuncertaintywastheestimateofmasses.Theempiricalequationswerebasedoninstalledpower,andthe
relativelyhighspeedoftheconceptdesignismuchlargerthanbasisships.
A key criterion in the determination of machinery mass is the choice of propulsion system. At the point of
determininginitialdimensionsanumberofdifferentpropulsionsystemswereunderconsideration,aswellas
the use of alternative fuels such as LNG and hydrogen. Each will require different machinery installations
requiring different space and mass allowances. Initially an LNG plant with a contrarotating pod has been
assumed. This initial mass estimate was an attempt to second guess the results of more detailed propulsion
researchtofollow.
Theempiricalestimatesofshipmasswerebackedupwithlightshipestimatesprovidedbythescalingofbasis
ships following the methodof Molland (2008). The method includes corrections to the lightship mass of the
basis ships due to changes in dimension through the determination of mass per metre of length, depth and
beam,: this is then multiplied by correction factors to obtain a dimensional changes in lightship mass. A
correctionisthenappliedinasimilarmannertoaccountforthechangesinscantlings.Thelimitationofthis
methodwasbeingabletofindbasisshipswithsimilardimensionsandknownlightshipmass.Themethodof

16

Initial Development of Fast Feeder Container Ship


basis ship scaling was applied to four ships of the International Shipping Corporation (2010). The results of
whicharesummarisedinTable3.2.
Table3.2Summaryofmassestimatesbasedonscalingbasisships
Ship

Lightship
/tonnes

LOA
/m

B
/m

D
/m

IntraBhum
MaerskAlabama
JavaSea

9825
6120
1702
3283

182
155
91
116

30.55
25
14.7
18

16.2
9.52
4.97
7

Dimensional
correction
/tonnes
2601
+586
+2411
+2409

Scantling
correction
/tonnes
1021
+166
+812
+782

Total
correction
/tonnes
3623
+752
+3224
+3191

Lightship
estimate
/tonnes
6202
6872
4926
6474

AsummaryofthemassestimatesproducedbythethreedifferentmethodsisgiveninTable3.3.Themaximum
variationinmassestimatebetweenthevariousmethodsemployedis14%.
Table3.3Summaryofmassestimates
Method

Steelmass
/tonnes

WatsonandGilfillan(1977)
SchneeluthandBertram(1985)
Scalingbasisships

4670
4326

Outfit
Machinery
mass
mass
/tonnes /tonnes
1430
1121
2489
1144

Lightship
Deadweight Total
mass
/tonnes
/tonnes
(tonnes)
7366
12842
20208
8962
12842
21629
6119
12842
18961

As well as providing empirical estimates of mass, the method of Schneekluth & Bertram (1987) provides
empiricalestimatesforverticalcentreofgravityofthevariousweightgroups(i.e.shipsteelmass,outfitand
propulsion). When combined with known estimates of the vertical centre of gravity of the deadweight
componentsaninitialestimateofshipcentreofgravitycanbemade,thusallowinganassessmentoftheship
stability.

3.2.2

Powering

Withintheempiricalmethodsemployed,thereareequationstoprovideestimatesofmachinerymass.These
areprincipallyafunctionofinstalledpowerandthusrequireaninitialestimateofpoweringtobemade.In
ordertogainaroughapproximationofinstalledpowerthebasisshipdatawasusedtoestimateshippoweras
relationship(Figure 3.1).Theestimateisrathercrudeasitdoesnot
afunctionofshipspeedusinga
accountforhullformchanges,howevertherewasagoodfittothebasisshipdata;theshipinstalledpower
wasfoundas25MW.ThisisveryclosetotheBMTfastfeederNorasia[BMTNigelGee(1998)].
Aspreviouslymentioned,itwasanticipatedthatinordertomeetemissionreductiontargetsanalternativefuel
suchasLNGorhydrogenwouldneedtobeemployed.InitiallyitisassumedthatLNGwouldbethepropulsion
fuelalthoughallowancesweremadetoselectanotherfuelsourcelaterinthedesignprocess.Anestimateof
fuelandtankmass,andtankvolume,weremadeassumingtheuseofLNGandavoyageprofile.Arangeof
3000 nautical miles (nm) was assumed, with 2% of voyage time spent manoeuvring, 23% of voyage time in
port, 46% of voyage time at lower operational speed and 29% of time at service speed [Mash (2009)]. An
auxiliarypowerallowanceof1.5MWwasassumedtoberequiredthroughouttheentirevoyage.Asummaryof
the assumed voyage profile is given in Table 3.4 A summary of the fuel and tank mass estimate with an
estimateofemissionsandacomparisontostandardMDOisgiveninTable3.5.

17

Concept Design of a Fast Sail Assisted Feeder Container Ship


Table3.4Assumedoperationalprofileforfuelandtankmassestimates
Service
speed
25.00
1537.00
61.48
29.00
25.03
1538.60

Speed/knots
Dist.travelled/n.miles
Timespent/hours
%timespent
Powerreq./MW
Energyreq./MWh

Slow
speed
15.00
1462.89
97.52
46.00
4.60
448.60

Cruise
auxiliary

163.25

1.20
195.90

Manoeuvring

4.24
2.00
5.00
21.20

Port
Total
auxiliary

3000.00
48.76 212.00
23.00

0
26.23
0 2204.45

Table3.5Capacityandemissionpertrip
Mass/tonnes
Tankmass/tonnes
CO2/tonnes
NOx/tonnes

LNG
343
302
1090
4

MDO
479

%diff.
28.47

1597
29

31.75
87.83

Table3.5indicatesthattheuseLNGforpropulsionleadstosubstantialemissionsandfuelmasssavings.

3.2.3

Stability

Aninitialstabilitycheckwasconductedusinganapproximatestabilityformula[Molland(2008)].Theminimum
limitingGMTwasestimatedat0.5metresandthemaximumGMTtobelimitedbyaminimumrollperiodoften
seconds(whichisequivalenttoamaximumGMTof1.2681metres).
The water plane area coefficient (CWP) was estimated using the Baker, MunroSmith and BSRA methods
[Molland (2008)]. The sensitivity of the solution to these parameters investigated. The Baker method was
usedinthefinalestimatesofstability,asthiswasdeemedthemorerepresentativeoftheproposedhullform5.
TheKBoftheshipwasestimatedusingMorrishsformula.Asummaryofthefinaliterationstabilitycheckfor
theproposedconceptisgiveninTable3.6
Table3.6Summaryofinitialstabilitycheck
Parameter Methodemployed
CWP
Baker
CWP
MunroSmith
CWP
BSRA
KB/m
Morrish'sFormula
BMT/m
BML/m
KG/m
GMT/m
Tr/s

Value
0.7988
0.6700
0.7020
5.5500
7.2910
8.010
12.956
0.601
14.520

AsummaryoftheinitialprincipalparticularsoftheshipisgiveninTable3.7.

5
ThiswasvalidatedlaterwithwaterplaneareacalculatedfromtheMaxsurfmodel,whichwasveryclosetothe
Bakerestimate.

18

Initial Development of Fast Feeder Container Ship


Table3.7Summaryofprincipleparticulars
Particular
LOA /m
LWL /m
B /m
D /m
T /m
CB
CB0.8D
/tonnes

Dimension
170.70
155.40
26.19
18.97
9.00
0.57
0.64
21402.00

3.3

Main propulsion and machinery options

The two main elements of the propulsion system are the propulsor and the plant. When choosing an
appropriatepropulsionsolution,itisnecessarytotakeintoaccount:therangeofoperatingspeeds;theoverall
efficiency of the system; initial cost and bunker prices; complexity and ease of maintenance; manoeuvring
requirements;andemissions.Potentialoptionsforboththepropulsorandplantareconsiderednext.

3.3.1

Propulsor
Fixedpitchpropeller

This is the most common propulsion device amongst merchant ships, especially for low speed operation. It
providestheoptimumefficiencywhenashipoperatespredominantlyatonespeedonly.Inaddition,fixedpitch
propellers(FPPs)featuresimpledesignandthusreduceconstructionandmaintenancecosts.
Controllablepitchpropeller
Thecontrollablepitchpropeller(CPP)providesanumberofadvantagesovertheFPP:firstly,onecanachieve
greater control over the thrust generated, aiding in manoeuvring situations; this also allows (where
appropriate)thepropulsionmachinerytostayrunningataconstantspeed,tomaintainefficiency,evenwhen
variablethrustisrequiredduringavoyage;lastly,thebladescanbefeatheredtoreducehydrodynamicdrag,
shouldthepropellernotberotating.SincethebossdiameterincreasesslightlycomparedtotheFPP,a23%
reduction in efficiency can be expected [Molland (2009)]. An additional consideration is the complex
mechanismusedtocontrolthebladespitch,whichmeansaCPPrequiresmoremaintenancethananFPP.
Waterjets
Awaterjetconsistsofapumpwhichdrawswaterthroughaductingsystembeforeexpellingitathighvelocity
nearthewaterline,thusgeneratingthrust.Waterjetsarebestemployedonshipsoperatingatspeedsbetween
30and45knots[Molland(2009)],andsufferfrompoorefficiencyatlowspeeds.Highmanoeuvrability,aswell
asstoppingability,canbeachievedusingdeflectorunits.Itshouldbenotedthattheweightofacomplete
waterjet unit is generally more than that of a propeller arrangement, although no rudder is required, thus
savingweightandappendagedrag.Waterjetsalsorequiresignificantmodificationofthesternlayoutofthe
ship,whichmayimpactotherdesignconsiderations.
Azimuthingdrives
Thispropulsionsolutioniscommonlyseenintwomainforms,namelytheZdriveandthepoddeddrive,each
ofwhichisdiscussedseparatelynext.

19

Concept Design of a Fast Sail Assisted Feeder Container Ship


(i)

Zdrive

Allowing high levels of manouevrability, the Zdrive is socalled due to the shape of the mechanical linkage
betweenplantandpropulsor.Bevelgearstransferpowerthroughashaftmountedwithinthebearingofthe
azimuthing unit to the propeller. This device is a popular installation on tugs, where ducted units are often
usedduetohighthrustloadings.Theyhavetheadvantageofrelativelylowdragsincetheunitissmallerthan
anequivalentpoddeddrive.However,thecomplexgearingcanleadtoanincreaseinmaintenancecosts.They
arenotbestsuitedtouseincombinationwithelectricpropulsion,duetotheshaftingarrangementandmain
plantlocation.Thussomeoftheefficiencygainsandlayoutflexibilityofelectricpropulsionarelost.
(ii)

Poddeddrive

ThispropulsionarrangementisadevelopmentoftheZdriveconcept,mountinganelectricmotorinsidethe
podhousingtopowerthepropeller,insteadofusingamechanicaldrive.Thisimprovestheflexibilityofpower
distributionandmachinerylayout.Alargerangeofsuchdevices,upto23Megawatts(MW)installedpower
per pod is available [Carlton (2007)]. They are suitable even for large ships where manoeuvrability is
important,suchascruiseships.Generallythepropellerismountedonthefrontofthepod,offeringacleaner
inflowtothepropellerduetoanabsenceofshaftingandbracketing,resultinginhigherefficiency.Reductions
inappendagedrag(byremovingshaftingbracketsandrudders)areoffsetbythedragofthepodunititself.A
further consideration is the large mass supported by the stern framing, which includes a large amount of
additionalmachineryasseeninFigure 3.2.Anadditionalconsiderationisthecomplexbearingunitsrequired
torotatethepods,whicharerequiredtotransmitlargethrustloadingstothehullandasaresultcansuffer
fromlowservicelives.Wheremultipleunitsareused,theirinteractioncancausesignificanthullvibrationas
wellasdamagetothepodhousingsthemselves,andthuscarefuldesignisessential.

Figure3.2Typicalpoddeddrive(ABBVOseries)andassociatedmachinery[ABB(2009)]

20

Initial Development of Fast Feeder Container Ship


Contrarotatingpropellers
Acontrarotatingpropeller(CRP)consistsoftwopropellersmountedalongthesameshaftaxisbutrotatingin
oppositedirections.Oneshaftturnsinsidetheother.Thissystemhastheadvantageofrecoveringsomeofthe
rotational energy lost in the propeller slipstream. The aft propeller is generally smaller to accommodate
slipstreamcontraction,andthetwopropellersmayhaveadifferentnumberofbladessoastoavoidinducing
large hull vibrations. Improvement in efficiency is of the order of 57% [Molland (2009)]. Their mechanical
complexitymakesthemunpopularforuseonmerchantships.
AdevelopmentoftheCRP,popularisedbyWrtsil[Levander(2002)]istheCRPpodconcept,whichmounts
the aft propeller on a podded drive, as presented in Figure 3.3. This removes the complex shafting
arrangementandallowstheadvantagesofusingapoddeddrivetobecombinedwiththeefficiencygainsofa
CRP. For a ship with two or more distinct operating conditions, such as is being considered here, this
arrangementhastheadditionaladvantageofallowingtheforwardpropellertobefeatheredwhenoperatingat
lower speeds without any loss of manoeuvrability. This propulsion layout is popular in fast RoPax designs.
RoPaxshipshaveasimilarfineformto,andoperateatspeedsclosetothefastfeederconcept.Fuelsavingsof
16% have been quoted by operators using this system, having switched from a twinscrew layout [Levander
(2002)].Adisadvantageofthisarrangementistheimpingementoftheforwardpropellerslipstreamontothe
podhousingduringcoursekeeping.Thiscouldinducesignificanthullvibrationorsurfacedamage.

Figure3.3ContrarotatingpropellerpodlayoutonfastRoPaxship[Levander(2009)]

3.3.2

Plantandfuels
Dieselengines

Dieselenginesarethemostcommonlyseenplantinmerchantships.Theyaregenerallyclassedaslowspeed
and medium speed. Low speed diesels offer the best efficiency in terms of specific fuel consumption (SFC),
with values as low as154grams per kilowatthour [Carlton(2007)], although165 g.(kWh)1 isa more typical
figure[MANDiesel(2010)].Poweroutputsofalmost100MWareseenfromthelargestunits.Theseengines
arebestsuitedtodirectdrivearrangementswithconstantoperatingspeedformuchofthevoyagetomaintain
optimalefficiency.Consequentlytheyarenotsuitableforusewithpoddeddrives.
Other disadvantages of lowspeeddiesel enginesconcern the size and weight of the plant. The dimensions,
particularly height, of a typical unit place significant restrictions on layout. In addition, low speeds typically

21

Concept Design of a Fast Sail Assisted Feeder Container Ship


weigh more than other plant of equivalent power output, and have the further limitation that this weight
cannotberedistributedeasilyforthepurposesofmassdistributioncalculations.Asanexample,atypicallow
speed engine of suitable size for application in this design, the MAN MEC8, rated at 25 MW [MAN Diesel
(2010)]hasanoverallmassof820tonnesandaheightof13metres.Bywayofcomparison,amediumspeed
dieselpropulsionsolution,giving26.6MW[WrtsilShipPowerTechnology(2009a)],weightsapproximately
450tonnesandhasamaximumheightoffivemetres.Sinceitconsistsoffourseparateunits,thissolutionalso
providesflexibilityintermsofengineroomlayoutandmassdistribution.
Aswellasthecitedadvantages,mediumspeedengineswouldprovideconsiderableredundancyandflexibility
inpoweroutputtothefastfeederconcept.Forashipoperatingatmultiplespeeds,demandingvastlyvarying
power requirements, the medium speed engines readily facilitate demands and maintain the efficiency of
runningathighmaximumcontinuousrating(MCR)byturningunwantedplantoffandrunningwithfewerunits
atnormaloperatingconditions.Thiswouldnotbepossiblewithalowspeedenginewheretypicallyonlyone
unit is installed. Medium speed also allows auxiliary load to be handled more easily, without the need for
separategenerators,andthuslendsitselftouseinconjunctionwithelectricpropulsionpowerdistribution,also
allowingamoreflexiblelayoutofmachineryspaces.Themaindisadvantageofmediumspeeddieselenginesis
theirSFC,whichistypically180g.(kWh)1,or9%higherthanthecomparablelowspeedengine.
SteamTurbine
The steam turbine features low weight and space requirements, as well as low maintenance costs. The
freedom to burn a variety of fuels in order to generate steam can be attractive in some cases, such as LNG
carriersusingboiloffgasfromthecargotankstofueltheboiler.However,steamturbineshavepoorSFCwhen
comparedtodieselengines,andassuchcannotbeconsideredthebestoptionfromanemissionsstandpoint.
GasTurbine
Gas turbines are popular in naval ships, where sprint ability is important, and cruise ships, which take
advantageofthevastspacesavingspossibletoincorporateadditionalcabins.Gasturbinesofferhighpower
toweightandpowertovolumeratios,typicallyfourtimeshigherthanmediumspeeddiesels[Carlton(2007)].
Other advantages are low maintenance, low manning requirements, and lower nitrogen oxide (NOx) and
sulphuroxide(SOx)emissions,burningmarinegasoil(MGO)fuel.Ifusedinconjunctionwithdieselengines,in
a socalled Combined Diesel and Gas (CODAG) arrangement, gas turbines can provide rapid increases in
deliveredpowerwhenrequiredforhighspeedtransit.Sincegasturbineefficienciesarelowwhennotrunat
thedesignMCRandtheirpoweroutputsareextremelylarge,theyareswitchedoffforlowspeedcruisingand
dieselsaloneareused.Onedisadvantageofthissystemhoweveristheneedtocarrytwofueltypes.
Itisalsonotedthatoperationinregionsofhighambienttemperaturereducestheefficiencyofgasturbines,
the reverse of which is true for diesel plant. Since the fast feeder concept is to be designed primarily for
tropicaloperation,thisfactorpresentsalargedisadvantage.
Hydrogen
Theideaofusinghydrogenfuelinhighspeedmarineapplicationsisacommonone.Veldhuis(2007)proposes
ahighspeedhydrofoilassistedfeedercontainershippoweredusingmodifiedgasturbinesburninghydrogen
fuel.Akeypointisthereductioninmassofhydrogenfuelrequired,estimatedas2.8timeslesscomparedto
normal gas turbine fuel. This allows a significant improvement in transport efficiency for highspeed ships.
However, this solution is recommended for power requirements greater than 36 MW, and as such is not
consideredsuitableforthisconcept,wheretheshipisclassedasfastratherthanhighspeed.
Fuelcells
Fuelcellsuseachemicalreactionbetweenhydrogenandoxygentoproduceenergy,withthemainemission
beingwater.Inthecaseofthesolidoxidefuelcell,thefuelusedismethaneandwater.Fuelcellshavehigh

22

Initial Development of Fast Feeder Container Ship


efficiencies compared to engines, with greater than 50% electrical efficiency [Wrtsil Corporation (2010)].
Despite their potential to drastically reduce emissions, in terms of merchant shipping, fuel cells are still
essentiallyataprototypestage,withsmall20kWunitsinstalledonsomeships.AlthoughfiveMWunitsare
planned [Wrtsil Corporation (2010)] these are far from being operational and thus current commercially
availablecellswouldnotevenprovideadequateauxiliarypowerforatypicalmerchantship.Whetherornot
fuelcelltechnologywillhavematuredsufficientlyby2020toprovidesignificantpowerishardtopredict,yetit
is almost certain that using fuel cells as the sole means of power would entail large volume and weight
penaltiesonshipswithlargepowerrequirements.
Naturalgas
Liquefied natural gas (LNG), the main constituent of which is methane, offers large savings in emissions,
particularlysulphuroxides(SOx)andparticulatematter(PM)[Levander(2008a)].Table 3.8summarisesthese
benefits compared to Marine Diesel Oil (MDO). The use of a dualfuel engine, such as the Wrtsil 50DF
[WrtsilShipPowerTechnology(2009b)]allowsLNGfueltobeburnt,andoffersmuchhigherefficiency(SFC=
135 g.kWh1) compared to diesel operation, meaning a lower fuel mass is required. Wrtsil Ship Power
Technology(2009b)alsoclaima50%increaseintheservicelifeofenginecomponentswhenusingLNGinstead
ofMDO.Thisisasignificantadvantageintermsofreducingmaintenancecostsandincreasingtheservicelifeof
theship.
Themaindisadvantageofthisfuelisitslowdensity,eveninliquidform,whichdoublesthestoragevolumefor
an equivalent operational range. Whilst Levander (2008b) advocates the use of such engines on container
shipsinordertoreducecoastalandportemissions,thereisnoreasontouseLNGastheprimaryfuel,aslong
astherequiredstoragevolumeisfeasible.
Table3.8EstimatedemissionsreductionusingLNGcomparedtoMDO[Levander(2008a)]
Emission
CO2
NOx
SOx
PM

Percentagereduction
2530
85
100
99

Fuel price makes up a large part of the operational costs of a ship, and with fast ships being particularly
sensitive to fluctuations in bunker price, this must be considered at an early stage in the design. Table 3.9
comparesthethreemainfuelsunderconsiderationhere,includingcurrent(2010)bunkerprices.
Table3.9Comparisonofmarinefuelsbydensity,lowerheatingvalueandcost[MDOandMGOprices
obtainedonline6,correcton23rdMarch2010,Singapore;LNGpriceestimatedbyLevander(2008b)]
Fuel
type

Density
/kgm3
900
875
410

LHV
/MJ.kg1
42.7
42.7
28.0

LHV
/kWh.kg1
11.9
11.9
7.8

Cost
/US.ton1
643
656
465

Cost
/USD.MBtu1
15.9
16.1
10.0

ThecostperMetricBritishunit(MBtu)givesarelative costperenergycontent,andshowsLNGtobevastly
superiortothemorecommonfuelsinthisrespect.Levander(2008b)alsonotesadownturninthepriceofLNG
circa January 2008, yet it is nave to assume that the price of LNG will continue to decrease into the 2020
market.Therearemanyfactorsthatmakeithardtopredictfuturefuelprices,suchaspoliticalrelationswith
oilproducingnations,yetasthedrivetowardscleanershippingcontinuesLNGmaybecomemoredesiredand

6
http://www.bunkerworld.com/prices/

23

Concept Design of a Fast Sail Assisted Feeder Container Ship


as such, more expensive, before production and bunkering infrastructure can develop sufficiently to
accommodatedemand.Whilstitisvastlybeyondthescopeofthisreporttoinvestigatefuturetrendsinfuel
prices, it is assumed that LNGfuel will be commonplace in the marine industry in 2020, and thus a viable
option.Inaddition,anyeconomiccalculationsinvolvingfuelpriceswillusecurrentfiguresforsimplicity.
Electricpropulsion
Electricpropulsionusesmainenginestodrivegeneratorsthatprovideelectricalpowerwhererequiredonthe
ship through a distribution system. This is particularly useful on ships with large hotel loads, such as cruise
ships,andthosepoweredbypoddeddrives,wherepoweristransmittedtoamotorinthepodhousing.Such
propulsion systems are often referred to as dieselelectric, or DFelectric, where dualfuel main engines are
used. Since there is no mechanical connection between main engine(s) and propulsor, the designer has
flexibility in terms of the location and layout of the propulsion machinery, which can improve the cargo
capacityoftheship.Itshouldbenotedthatefficiencygainsfromrunningthemainenginesatconstantrating
are offset by losses in the electrical distribution system. Wrtsil Corporation (2009a) claim potential
reductionsinfuelconsumptionofbetween58%inadditiontoareductionininstalledpowerof10%owingto
moreefficientloaddistributionthroughoutoperation.
Anexampleofacompletepropulsionsystemutilisinganumberofthetechnologiesdiscussedinthissectionis
illustratedinFigure3.4.

Figure3.4ExampleLNGelectricContrarotatingpropellerpodpropulsionsystemarrangement:plantare
showninbluewithgeneratorsandmotorsinred[Levander(2002)]
Summary
Considering all the propulsor and plant options discussed here, the main driver in selecting an appropriate
solutionistheneedtomaximiseefficiencyatbothserviceandsailingspeedsandreduceemissionslevels.The
mostefficientpropulsionsystemforthisshipisthusapoddeddriveorCRPpodarrangement,whichhasthe
addedadvantageofprovidinghighlevelsofmanoeuvrabilityandlayoutflexibilitythroughtheuseofanelectric
powerdistributionsystem.IfthissystemwerecombinedwithplantburningLNGfuel,itisanticipatedthatthe
emissionsdeficitpresentedinTable2.9canbeeasilyrecovered,basedonthevaluesgiveninTable3.8.Thisis
beforeanyotherformofemissionsreductionorimprovementinefficiencyhasbeenmade.

24

Hydrodynamic Design Development

4. Hydrodynamic Design Development


4.1

Hull form design

The hull form development must fulfil the operational requirements of the shipping routes selected whilst
being economically and environmentally efficient. Fortunately, on a fundamental level, these requirements
complement one another; if a ship is efficient in terms of reducing operational costs by reducing fuel
consumptionthenitmustbesaidthatthiswillalsoreducetheoperationalenvironmentalimpact.Ultimately
theoverallefficiencyisconsideredtobeacombinationoftheefficienciesofthehullformhydrodynamicsand
thepropulsionsystem,whichareofcourseinterdependent.
Toincreasethecomprehensionofthisstudytwohullformsaretobedevelopedtohousedifferentpropulsion
system designs. The first (Hull A) is designed for a contra rotating CPP azimuthing pod combination. The
second design (Hull B) is developed to accommodate twin azimuthing podded propulsors. These propulsion
optionshavebeenselectedasaresultofaninitialevaluationofpropulsionandmachineryoptions(seeSection
3.3).Itisexpectedthatthesetwoconceptswillperformbestintermsofefficiencyandoperationalflexibility.
Onewouldexpectthehydrodynamicperformanceofthesetwohullstobesomewhatdifferent,asaresultof
thesignificantlydifferentstandardsternformsneededtohousetherelevantpropulsors.Thishighlightsthe
needtoassesstheoverallefficiencyofboththehullandpropulsor.

4.1.1

Designforhydrodynamicperformance

This section outlines the design of Hull A and Hull B to optimise hydrodynamic efficiency and to provide a
performancepredictionforeachform.Thehullformsareoptimisedtoreducethehydrodynamicdragincalm
water,aimingtominimisetheenergydemandtoattaintheoperationalspeedof25knots.Theproportionof
time a ship spends operating in calm water is of course minimal. Thus, the performance in waves is an
importantconsideration.Optimisingashipforaddedresistanceinwavesismuchmorecomplexandnumerical
methods are only useful when supported by model tests. Both of the forms will be analysed to assess
performanceinwavesinSections4.4.3and7.1.3.
It is important to carry out hull form design early in the project as most subsequent design stages are
dependent on the finalisation of the ship lines. One must however be confident in the design in terms of
creating something that is optimal as it is difficult to make subsequent changes to the lines. In terms of
economics,designdecisionsmadeinthefirstfewweeks oftheprojectaccountfor70%to80%ofthecosts
associatedwiththedesign[Harries(2006)].Itisthereforeimportanttohaveconfidenceintheproceduresand
methodsappliedatthisstageoftheproject.Whilstthissectionfocusesonhydrodynamicperformanceitis
important to consider other implications of the optimisation. The preferred design solution to improve
hydrodynamicsistoobtainanoptimumformthatneitherrequiresadditionalproductioncostsnorentailsany
operationalcomplications[Harriesetal.(2006)].
Hydrodynamicoptimisationisnottrivialbyanymeans.Harriesetal.(2006)statestorealisea1%increasein
speedoneneedstolowerthetotalresistancebyaround4%to5%,providedpropulsiveefficiencyisconstant.
Thislevelofresistancereductionissignificant,especiallyattheearlystageofaconceptdesignwhenextensive
timecannotbesparedfordetailedhullhydrodynamicanalysis.Notethatwithregardstothisconcepttheaim
istoreducethepowerrequirementnotincreasethespeed,thebenefitsareequivalent.
The primary source for the estimation of the calm water resistance is the Holtrop regression [Holtrop &
Mennen(1982)andHoltrop(1984)].Thisisaregressionmethodbasedon334shipmodelexperiments,carried
outatMARINintheNetherlands,andresultsfromseatrials.Thehullformstestedintheseinstancesarenot
vastlydissimilartothefastfeedershipforms.Hencethisregressionprovidesrealisticestimatesofthenaked

25

Concept Design of a Fast Sail Assisted Feeder Container Ship


hull resistance for each design. This is contrary to others expectations who comment that such regression
analysestheyareoutdatedandthusfoundtounderestimatetheresistanceofmodernships[Bertram(2000)].
The Holtrop regression is however the best method available; allowing comparison of hull form designs.
UltimatelythefinalvaluesofresistancewillbedeterminedfromthemodeltestspresentedinSection4.4.

4.1.2

HullAbasishullform

Although the propulsion system utilised with this design is somewhat unique there are not that many
differenceswithotherfastcontainershiphullforms.Thereforealibraryhullformofalargercontainership
hasbeenusedasastartingpoint.Thishullformhasbeenscaledtomatchtherequireddimensionsspecifiedin
Section 3.2.Thefullnessofthehullformisadjustedandthedraughtcalculated,withrestrictionsinmind,to
givetherequireddisplacementandblockcoefficient .TheHognersternformhasbeenmodifiedtoensure
thereisenoughspaceavailableforthepropulsionsystemandtheassociatedcontrols.Thebowform,including
bulbous bow, have at this stage been left to resemble the library hull. Initially the mid body has been
designedtomaximisetheamountofparallelmidbody,toattaintherequireddisplacement,andsotoincrease
the cargo capacity. The cargo delimitation on a container ship is discrete and thus a small change in hull
dimensionscanresultinlargechangestothecargocapacity.
Theresistancewasevaluatedandfoundtobemuchlargerthanexpectedfrominitialestimates,almosttwice
theexpectedvalue.Thisdrasticincreaseinresistancewouldmeanthattheshipwouldnotbeeconomically
viabletooperate.Itwasclearthattheshipcouldnotoperateefficientlywiththislevelofparallelmidbodyat
25 knots. In particular, the prismatic coefficient was observed as being very large, this is defined as
/

. Increasing

increases

increases the resistance of a hull form and it is expected that as Froude number

shoulddecrease[Schneekluth&Bertram(1987)].Asthemaximumsectionalarea

isfixedby

the preliminary dimensions and the waterline length


cannot change significantly it must be the
displacement thatischanged.Thedisplacementmusthencebereducedslightlybyreducingthelengthof
the parallel mid body. Hence the entrance and run increase. These modifications reduce from 0.765 to
0.620.ThisisconsistentinlinewiththerecommendationsofComstock(1967)(Figure56p345)atthisFroude
number.Theresultofthisisareductioninresistancetoavaluecomparabletobasisships(seeSection 3.1).
Thisinevitablyleadstoalowercargocarryingcapacity.Thehullformchangeshavebeenaccommodatedsoas
nottoloseawholerowofcontainers.Thelossindisplacementis992tonnes,whichequatesto100TEU.This
cargocapacityreductionstillleavesthedesignwithinthelimitsforeconomicoperation.
Itispossibletoequatethischangeinhullformshapetochangesinblockcoefficient ratherthan .Thisis
arbitraryhoweverandtheauthorfeelsthepointisbetterillustratedintermsof .Adecreaseinvolumehas
thesameeffectonthefrictionalresistancebyreducingthewettedsurfacearea.Asidefromthistheshipis
becomingmoreslenderintheentranceandtherunandhencethedisturbanceofthefreesurfaceisreduced.
Physically, due to a reduction in the flow pressure, the wave pattern resistance is reduced. Prior to the
reductiontheparallelmidbodywaslargeandhencetherewasalargegradientbetweenthestem/sternand
themidbodyresultinginhighpressureintheseareas.Thisconstitutesalargewavepatternresistance.When
themidbodylengthwasdecreasedthisgradientreducedsignificantly,henceloweringthewavepatternand
hasbeenincreasedbyaroundtwometres.Thiswillof
overallresistance.Inadditionthewaterlinelength
course increase the friction resistance, but this is far outweighed by the reduction in the wave pattern
has been achieved by
resistance. This can also be attributed to a lower value of . The increase in
closerto
.
adjustingthestemandsternprofiles,thusbringing
Whenoptimisingthehullformatconstantdisplacementoneconsiderstheeffectofmovingthelongitudinal
centreofbuoyancy(LCB),henceadjustingthedistributionofthedisplacementvolume.Schneekluth&Bertram
(1987) state that the optimal position of the LCB is a function of and , and provides an estimation

26

Hydrodynamic Design Development


procedure.Howeverforreducedresistancethelongitudinalcentreofgravity(LCG)mustcoincidewiththeLCB
to facilitate level keel operation. The position of the LCB often needs to be a compromise between the
hydrodynamicoptimumpositionandcoincidewiththemassdistributiondependantLCG.Figure4.1illustrates
howtheoptimumLCBvarieswith .

Figure4.1VariationofoptimumLCBwithCB,Schneekluth&Bertram(1987)
2.88metres
The block coefficient for the basis hull is CB 0.547; this equates approximately to LCB
from amidships. On observation of the LCB value in Table 4.3 it can be seen that this value is closer to
amidships. The LCB has been moved forward from the optimal location to accommodate the ships
unconventional layout. Ultimately, having superstructure and large generators positioned forward, the LCG
will be located further forward than in a conventional container ship. With the position of LCB nearer to
amidshipsitshouldbepossibletoadjustthemassdistributionsotheLCGcoincideswiththispointandthusthe
ship can operate on an even keel. The final position of LCB presented in Table 4.3 is not random but a
consequenceoffinetuningthedisplacementdistribution.Itiscommonpracticetodesignthesectionalarea
curvetobesmoothexceptforoneinflectionbetweentheforwardperpendicularandthestem[Harriesetal.
(2006)],togenerateahullthatexhibitsgoodresistancecharacteristics.Finaladjustmentstothehullformhave
been made to achieve this and Figure 4.2 exhibits this with inflections at the forward and aft end. This
illustratesarapidchangeinsectionalareaintheseregions.
Designconsiderationsotherthanhydrodynamicoptimisationhavebeenmade.Theaccommodationblockisto
belocatedforwardtoimproveaerodynamics,cargostowageefficiencyandlineofsight.Thebowhassufficient
flaretoproduceadeckareaconsistentwithhousingtherequiredsuperstructure.Whilstincreasedflarecould
resultinheavyslammingtheadditionalbuoyancywillacttoreducetheamplitudeofthepitchmotions,which
isimportantgivenmostoftheworkingspacesarelocatedattheforwardend.Aseakeepinganalysisofthe
designisprovidedinSection7.1.

4.1.3

HullBbasishullform

TheformforHullBwasdevelopedfromthe optimisedversionofHullA,tomaintainabasisofcomparison.
Theonlynotabledesigndifferenceistheselectedpropulsionmethod,sotheaimwastooptimisetheformfor
usewithtwinazipods.Inpractice,thismeantthattheoptimisationeffortwasfocusedontheaftform.

27

Concept Design of a Fast Sail Assisted Feeder Container Ship

250

Sectionalarea/m2

200

150

100

50

0
90

60

30

30

Longitudinalpositionfromamidships/m

60

90

Figure4.2HullAbasissectionalareacurve
Given the draught restriction (Section 3.2) the pods were to be mounted relatively high above the baseline.
This,coupledwithadesiretomaximisepodefficiency,(Section3.3.1)byallowingthemtooperateincleanflow
outside of the turbulent wake required a gradual rise of stern aft of amidships. The computational time
availablefortheanalysiswaslimitedbymodelproductionconsiderations,sothedesignprocesswasdrivenby
asimpleminimumresistanceoptimisation.ThiswouldbecarriedoutthroughuseoftheAustralianfreeware
packagesMichletandGodzilla.
Software
Michlet [Cyberiad (2009)] provides the user with predictions of the hull form wave resistance by applying
Michells thin ship theory based wave resistance integral from a user defined hull underwater form. The
method cited was implemented by Tuck et al. (1999). Tuck & Lazaukas (2008) have demonstrated that it
remainsagoodgeneralindicatorofwaveresistancetrends,correlatingwithexperimentalandCFDresults,soit
wasfeltthatitwouldbeareasonableindicatorofhullperformance.Estimatesoftotalresistanceareobtained
byuseofITTC57correlationlineandauserinputviscousformfactor.
Thefreewarepackageissuppliedwithanintegralgeneticalgorithmaddon,Godzilla.Anexplanationofthe
searchmethodologyemployedisprovidedbyTuck&Lazaukas(1996).
Methodology
Thepracticalapplicationofthesoftwarerequiresdefinitionoftheunderwaterformofthebasishull.Sincere
entrant sections at bow and stern could not be modelled, the match was based on maximum waterline
dimensions,formcoefficientsandtheunderwatermidshipsection.
Once satisfied with this initial form, the optimisation procedure was constrained with regards to overall
dimensions and form parameters (see Table 4.1) and the parametric definition of the forward form was
effectively fixed. While some particulars such as were very tightly constrained to maintain compatibility
withtheforwardform,otherssuchasLCBand wererelativelylooselyfixed,soastogivethebestpossible
freedom to the optimisation. The principal dimensions were allowed to vary slightly to permit a broad
populationrange.Theoptimisationalgorithmwasthenrunwiththeobjectiveofminimisingtotalresistanceat
speedsof25and15knots.

28

Hydrodynamic Design Development


Table4.1BasisestimatesandGodzillaoptimisationconstraintsonprincipalparticulars
155
25.6
8.75
0.54
0.58
4

4000

Particularsandconstraints
/m

/m

/m

% from

midships
2
/m

Seedvalue
(157.0)
(25.7)
(8.94)
(0.554)
(0.620)

160
25.8
9.0
0.57
0.67
0

(2.0)

5000

(4679)

Results
The aftbody plan of the optimised form from Godzilla as compared to the basis approximation and the
improvementinwaveresistancedemonstratedareshowninFigure 4.3.Itmustbestressedatthispointthat
theresistancecomparisonisforcomparativeandqualitativepurposesonlyandisintendedtobeindicativeof
theoptimisingtrend,notoftheperformanceofthefinalHullBorofanyquantitativedecrease,especiallyin
thespeedregionabove20knots.
Wave (pattern) resistance
coefficient, CW / kN

0.025
Optimised

0.020

Basis
0.015
0.010
0.005
0.000
0

a)

b)

10
15
20
Ship Speed, VS / knots

25

Figure4.3BodyplansfortheaftsectionsofHullB,MichletbasisandGodzillaoptimised(left);andvariationin
waveresistancecoefficientCWwithshipspeed(right)

4.1.4

Bulbousbowoptimisation

Now that basis hulls have been developed to exhibit good resistance characteristics one can focus on the
optimisationofthesedesigns.Comstock(1967)(Table18,p.346)statesthatforshipsoperatingattheselected
Froude number a bulbous bow is beneficial. Bulb design is ship dependant and is not trivial, as a badly
designedbulbwillincreaseresistance.Unfortunatelyitisimpossibletoassociatestandarddesignguidelines
forabulbousbow.ToselectbulbsforthebasisformsdevelopedinSections 4.1.2and 4.1.3anoptimisation
processwasimplementedtogeneratebulbsthatleadtoincreasedhydrodynamicefficiency.
InaccordancewithHarriesetal.(2006),aparentbulbwasdeveloped.However,thebulbdesignhasnotyet
beenconsideredandsoinsteadofcomparingtoonebaselinehullthetwohullformswithanumberofbulb
optionsaretobecomparedwithoneanother.Itisimportanttonotethatitisonlythebulbousbowthatis
beingoptimised.Themidbodyhasalreadybeendesignedtoreduceresistanceasfaraspossible,subjectto
required cargo capacity, and the stern forms are optimised to exhibit good flow characteristic into the
propulsors.Thustheoptimisationwaslimitedtothebulbassignificantsavingscanbemadewithlittleeffect
onthehydrostatics.Kracht(1978)suggeststhattherearenosignificantadverseeffectsofabulbousbowin

29

Concept Design of a Fast Sail Assisted Feeder Container Ship


termsofmotionsandaddedresistanceinwavesbelowBeauforteightwind,sothebulbwasoptimisedforcalm
waterresistance.
A bulbous bow reduces the wave pattern and wave breaking resistance. For fast slender ships the main
reduction is in the wave pattern resistance [Hoyle et al. (1986)] due to interference effects, as the wave
resistance constitutes a large proportion of the total resistance. The bulb must create a cancellation effect
between the wave systems created by the bulb andhull. A ship moving through a free surface creates two
wavesystemsattributedtotwohighpressurelocationsintheflowaroundthehull,atthebowandstern.The
purposeofthebulbistomodifythelocationandmagnitudeofthehighpressureatthebow.Thelongitudinal
positionofthebulbisameasureofthephasingbetweentotwowavesystems.Thesizeofthebulbisrelated
to the amplitude of the waves generated at the bow [Kracht (1978)]. Unfortunately it is only practically
possible to optimise a bulb for one speed. Complete cancellation of wave peaks and troughs from the two
systems can be achieved with the correct phasing of the bow and stern wave systems. In reality complete
cancellationisunlikely,andinattemptingthisonemayhavetocreatealargebowwavesystemthatwillcreate
significantdragbeforecancellationeffectstakehold.Henceforcancellationthewavesystemgeneratedatthe
sternmustalreadybereasonablysmall.
Wavebreakingresistanceismuchmoredifficulttoquantify.However,itsoccurrenceisapparent.Itisbroadly
associatedwithwavescreatedbytheshipbreakingandsprayfromthesewavesresultinginanetenergyloss
downstreamthatmustbeattributedtoaresistancecomponent.Theonlywaytoreducethiscomponent,in
terms of bulb design, is to ensure that the bow waves are not steep enough to break [Kracht (1978)]. The
introductionofabulbonashipincreasesthefrictionalresistanceduetotheincreasedwettedsurfacearea.
However,itisexpectedthatthiswillbesmallcomparedtothereductioninwaveresistance,especiallyfora
fastship.Abulbdoesnot,ingeneral,affectcoursestabilityormanoeuvrabilityandistheidealplacetolocate
bowthrusters[Kyriazis(1996)].
Fifteen bulb combinations of varying dimensions, as defined in Table 4.2, which have been analysed. The
dimensions have been chosen to cover the range that is practically achievable in terms of construction and
operation and based on proportions that have been observed on basis ships. Figure 4.4 describes how the
dimensionsaredefined.
Table4.2Topologyofbulbvariants
/ %

Bulbvariant
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

4.5
4.5
4.5
3.9
3.9
3.9
2.9
2.9
2.9
2.0
2.0
2.0
1.0
1.0
1.0

/
7.168
7.168
7.168
6.253
6.253
6.253
4.689
4.689
4.689
3.126
3.126
3.126
1.563
1.563
1.563

30

/ %
25.6
21.4
18.3
25.6
21.4
18.3
25.6
21.4
18.3
25.6
21.4
18.3
25.6
21.4
18.3

/
6.712
5.600
4.800
6.712
5.600
4.800
6.712
5.600
4.800
6.712
5.600
4.800
6.712
5.600
4.800

Hydrodynamic Design Development


Bystudyingdifferentbulblengthsandbreadthsitwillbepossibletoassesswhentheinterferencerelationship
betweenthewavesystemsisdestructive.Thisallowsforthemostsuitablebulbforeachofthehullformsto
beselected.Sincethebulbdesignisspecifictoaparticularshipdesigntheoptimalbulbwillnotbethesame
forbothHullAandHullB.ThebulbdimensionshavebeendefinedbasedonamethodpresentedbyKracht
andthebulbbreadth
(1978)wherethelengthofthebulb isnondimensionalisedwithrespectto
withrespectto (Table4.2).Thebulbdepthisconstantandalwaysbelowthewaterline.

Figure4.4Schematicshowingbulbousbowdimensiondefinitions
ThebulbcrosssectionsfitintoacategorytermednablatypebyKracht(1978).Thistypeofbulbhasitscentre
ofvolumetowardthetopofthebulbgivingfavourableseakeepingcharacteristics,itisforthisreasonthatthis
is the most common type of bulb and why it has been chosen for development in this concept. In previous
studies,Percivaletal.(2001),Kyriazis(1996),Hoyleetal.(1986)andKracht(1978)havealldrawnsignificant
benefitsfromusingnablatypebulbsandsoitseemsreasonabletodevelopnablabulbsforthetwobasishull
forms.

4.1.5

Results

TheHoltropregressionseriesisusedtoanalysethecalmwaterresistanceofthetwohullformsforeachofthe
15bulbcombinations.ThisregressionanalysisisimplementedthroughthesoftwarepackageHullspeed,part
oftheFormationSystemsdesignsuite.Hullspeedtakesthe3DNURBSsurfacefromMaxsurfandcalculatesthe
particulars of the ship required to carry out the regression. In terms of the bulb definition the Holtrop
regressionseriesusesthetransversebulbareaandthebulbdepthfromthekeel.Implementingtheregression
seriesrequirestheapplicationofmanyequationstodeterminethecoefficientsofresistance.Onimplementing
this for all 30 hull form variants the results may be represented in terms of the total resistance coefficient.
Figure4.5presents foreachhullforanoperatingspeedof25knots.

31

Concept Design of a Fast Sail Assisted Feeder Container Ship


3.80
HULLA

CTS x1000

3.75

HULLB

3.70

3.65

3.60

3.55
1

7
8
9
Bulbvariant

10

11

12

13

14

15

Figure4.5VariationofCTat25knotsforallbulbvariants
Caution is required when comparing dimensionless resistance. In particular we note that resistance is non
dimensionalisedwithrespecttowettedsurfacearea.Thevariationinbulbdimensionsmeansthewettedarea
isnotconstantwhichwillaffectthefrictionalresistance.Fortunatelytheareafluctuationinallofthesecasesis
small and presentation of the total resistance, instead of the total resistance coefficient, indicates the same
trendasobservedinFigure 4.5.Fulldetailsofthevariationin overarangeofspeedsforeachvariantare
essentially represents the variation of as the
given in Table B.1 and Table B.2. The variation of
frictional (andviscous )isalmostconstant.ThisisduetotheimplementationoftheITTC57correlation
. Hence one can deduce that variation of the resistance at the design speed is due to
lines to estimate
changesinwavepatternresistance.
Figure4.5showsthattheoptimalbulbvariantforHullAisnumber9andforHullBisnumber6.Figure4.5also
4.80 , have a low resistance compared to other bulb
indicates that bulbs 3, 6, 9 and 12 for which
variants of the same length. This suggests that these bulbs are leading to destructive interference and
reductionsinthewavepattern.Fromthesebulbstheoptimumforeachhullisdecidedbythephasingthey
4.689 )than
introducebetweenthebowandsternwavesystems.HullAisoptimalforashorterbulb(
6.253 ).Thisisafunctionofthedistancesbetweenthesourcesandthewavelengthgenerated
HullB(
bythosesources.

32

Hydrodynamic Design Development


5.0

CT,CW,CF,CV x1000

4.0

HULLA Ct

HULLA Cw

HULLA Cf

HULLA Cv

HULLB Ct

HULLB Cw

HULLB Cf

HULLB Cv

3.0

2.0

1.0

0.0
0

Fn

0.1

0.2

0.3

0.4

Figure4.6BreakdownofresistancecomponentsasafunctionofFnforbothHullAandHullB
Figure 4.6presentsthebreakdownoftheresistanceintoitscomponents.TheHoltropregressionessentially
determinesthetotalresistanceusing
1

(4.1)

ThewherethefrictionalcoefficientisdeterminedusingtheITTC57correlationline
0.075

(4.2)

Theformfactors 1
,determinedintheregressionanalysis,equal1.1878forHullAand1.1819forHullB.
The coefficient of wave making resistance and the correlation allowance are also determined in the
regressionanalysis.Thecorrelationallowancerepresentsthedifferencesinfullscaleresultsfromtowingtank
0.416694andforHullB
0.414195.
measurementsandseatrials.ForHullA
From Holtrop & Mennen (1982) there are two additional resistance terms, accounting for the additional
pressureresultingfromthebulbousbowandanimmersedtransom.Inthiscasetheyarebothzero.
AsalreadyobservedfromthebulboptimisationHullAhasalowertotalresistancecoefficient.Havingstudied
thelinesplans,thisissurprising.AsHullBwouldbeexpectedtohavealowernakedhullresistance.Itisonly
whenthedimensionalformoftheresistanceisplotted,Figure 4.7,thatthelowernakedhullresistanceofHull
B is observable above 20 knots of advance. Reverting toTable 4.3, one notes that Hull A does have a larger
meanthateachdesignwillnothaveidenticalshipspeedsat
wettedsurfaceandtheslightdifferencesin
identicalFroudenumber.HencethiswillcausesomeshiftingofthecoefficientofresistancecurvesinFigure
4.6.FulldetailsofthecoefficientbreakdownandeffectivepowerofbothhullsisgiveninTableB.3.

33

Concept Design of a Fast Sail Assisted Feeder Container Ship


20000
HULLA

16000

HULLB

PE /kW

12000

8000

4000

0
0

10 V /knots
S

15

20

25

30

Figure4.7EffectivepoweragainstshipspeedforHullAandHullB
InadditiontousingtheHoltropregression,thewavemakingresistancewasdeterminedusingThinShiptheory
[Michell (1898)] via Hullspeed. Thin Ship theory vastly overestimated the total resistance, although the
slendernessand arewithintheclaimedrangesofapplicability[FormationDesignSystems(2005a)].
Figure B.1 shows a comparison of the results from the regression and Thin Ship theory. There is good
correlationatlowspeed.Thisislikelyduetothefactthatthewavepatternresistanceisnegligible.Thereis
increasing disagreement with higher , where the Thin Ship Theory results should become more realistic.
Duetopoorrepresentationattheoperating theresultsfromThinShipTheoryhavebeenneglected.
ThelinesplansfortheoptimisedHullAandHullBarepresentedinAppendices B.3and B.4respectively.The
principalparticularsandhydrostaticsarepresentedinTable 4.3.Thepowerpredictionallowsotheraspectsof
thedesigntoprogressincludingthedesignofmodelsfortowingtanktesting(Section 4.2),whichareusedto
verifythecalculatedresistanceresults.

34

Hydrodynamic Design Development


Table4.3Summaryofprincipalparticularsandhydrostatics
/
/
/
/
/
/
/
/
S/
/
/

/
/
/
/
/

HullA
170.7
160.09
26.2
18.97
8.94
20466
19966.88
158.47
4847.188
203.315
3224.507
0.62
0.547
0.604
5.833
5.213
7.236
242.742
33.051
316.98

/
/

/
.

HullB
170.7
160.09
26.2
18.97
8.72
20344
19847.58
160.301
4633.287
203.335
3113.071
0.609
0.55
0.893
4.214
5.082
6.533
233.462
31.909
303.126

4.2

Towing tank model design and manufacture

4.2.1

Introduction

Modeltestswerecarriedoutinordertosupporttheexistingdesignwork,andvalidatenumericalpredictions
ofhullformresistanceandaddedresistanceinwaves.Inaddition,testswereundertakentoassesssailhull
interaction and allow comparison of the sailing performance of each hull form, which cannot be so readily
achievedusingthenumericalresistancepredictionmethods.Inordertoguidethedesign,manufactureand
testing of the models, ITTC recommended procedures were followed as closely as possible, to improve
confidenceinthetestresults.
ThefacilityavailableforconductingthetestswastheSouthamptonSolentUniversity(SSU)TowingTank,the
principaldimensionsandcapabilitiesaresummarisedinTable4.4.
Table4.4SummaryofSouthamptonSolentUniversityTowingTankfacility
Dimension
Length/m
Breadth/m
Depth/m
Max.carriagespeed/ms1
Frequencyrangeofwavemaker/Hz

Value
60.0
3.7
1.85
4.5
0.25 2.00

4.2.2

Modeldesign
Choiceofscale

There are a number of considerations to be made in selecting an appropriate scale for a ship model.
Maximising model size is important in order to reduce the percentage experimental error and give more
accurateresultswhenusingresistancescalingtechniques.Howeverpracticalrestrictionsareimposedby:

35

Concept Design of a Fast Sail Assisted Feeder Container Ship

The size of the tank: breadth is most important in this case to avoid tank wall interference when
testing in waves. Depth is less important since the Froude number ( ) is likely to be low. This is
)onlybecomescriticalforamodellengthof
confirmedbythefactthatthedepthFroudenumber(
17.38metres,ascalculatedinAppendixC;

Themaximumcarriagespeed:duetoFroudescaling,therequiredmodelscalespeedincreaseswith
modelsize,andthismuststaybelowthesafeoperatinglimitofthecarriage;

Budget constraints: the choice, quantity and price of material all have to be considered within any
financiallimitations;

Manufacturingcapability:Thetoolsandexpertiseavailabletothedesignermustbetakeninaccount.

ITTC(2005b)provideafunctiontoensurethat,whentestinginwaves,tankwallinterferencedoesnotoccur.
Thisrequiresconsiderationofthelikelywavelengthstobetested,whicharenormallytakenasapercentageof
theshiplength,andthereforeunknown.Theprocessadoptedwasto:

Assumeascalefactorandmodellengthfromthefullscaleparticulars;

Calculatetheupperlimitofwavefrequencythatwillcausetankwallinterference;

Simultaneouslycalculatetherequiredwavemakerwavefrequenciesandcheckthattheylieabovethe
establishedlimit;andwithinthecapabilitiesofthewavemakeritself;

CalculatetherequiredcarriagespeedforthehighestFroudenumbertobetestedandcheckagainst
themaximumcarriagespeed.

Following this approach, assuming the wavelengths of 50, 100, 150 and 200 % of the model length, the
maximummodelsizethatcanbetestedhasawaterlinelengthofsixmetres,correspondingtoacarriagespeed
of approximately 2.5 metres per second. Whilst this gives an agreeable scale factor of 26.4, the model
displacementwouldbeoftheorderof1.1tonnes,whichissimplynotpractical,andwouldmakehandlingand
transportationdifficult.
Toreducethesizeofthemodels,attentionwasshiftedtowardsmaterialselectionandmanufacturingprocess.
TheUniversitysEngineeringDesignandManufacturingCentre(EDMC)wasapproachedtocarryoutthework,
sincelabourcostwasnotchargeddirectlytotheproject,andthelocationmadecommunicationeasier.Based
onpreviousexperience,itwasrecommendedthatthehullsbemanufacturedusinghighdensityfoam(HDF)cut
on a ComputerNumerically Controlled (CNC) milling machine. This was considered cheaper and less labour
intensive than the proposed alternative, using fibrereinforced plastic (FRP), and could offer a higher quality
surfacefinish.ThesetwomanufacturingmethodsarerecommendedbyITTC(2002),whoalsosuggestwaxand
woodaspotentialconstructionmaterials.Thesehadhoweverbeendiscountedsincetheformerwouldbetoo
heavy,andthelatertoohardtoconstructinareasofhighhullcurvaturesuchasatthebulbousbow.
HavingchosenHDFastheconstructionmaterial,themainlimitationwasthesizeofthematerialblockthatthe
CNCmillingmachinecouldaccommodate.Thiswaslimitedto1.25metresinlength,andthusitwasdecided
thattheoveralllengthofbothmodelsmustfitwithintwofoamblocksof1.25metrelength,soastoreducethe
total cost and machining time. Allowing 50 millimetres of waste material at either end of the block for the
)ofthemodelswasrestrictedto2.3metres.Usingthis
machiningheadtomanoeuvre,theoveralllength(
limit, a scale factor of 1:75 between ship and model scale was derived. Details of the two models are
summarisedinTable4.5.

36

Hydrodynamic Design Development


Table4.5Summaryofmodelscalehullsatdesignspeed
/m
/m
/m
/
@25knots
/ms1

HullA
2.113
0.344
0.119
48.5
0.326
1.484

HullB
2.127
0.344
0.116
48.2
0.324
1.483

Usingtherecommendedsupplier,Homeblown,thecostoffoamrequiredwasestimatedalongwithadditional
materials.ThisisincludedaspartofthebudgetsummaryinAppendixNThetotalestimatedcostofmaterials
forthetwotowingtankmodelswas1045.45.Notethatinordertosavecosts,120kg/m3densityfoamwas
specifiedfortheaftsections,while200kg/m3densityfoamwasusedforthebowsectionswherethedetailof
thehullformswasmoreintricate,becauseofthebulbousbow.
Massdistributionfortestsinwaves
Inordertoobtainaccurateresultsofaddedresistancewhentestinginwaves,thepitchgyradius( )hadto
was
be modelled to represent that expected on the fullscale ship. As recommended by ITTC (2005b),
andthenecessarymassdistributioncalculatedusing:
estimatedas0.25

(4.3)

where is the component mass, and and are the longitudinal and vertical centroids of . This
requirementalsohadtobebalancedagainsttheneedtoachievethecorrecttrim.Theprocesswascarriedout
forHullAonly,assumingHullBtobesimilar,usingaspreadsheet.Thefoamcrosssectionwasassumedsolid
atbowandsterninordertoretainstrengthinthesefinerregionsofthehullform,whilstthemidbodywas
designed to be hollowed out to save weight and allow more flexibility in the positioning of the ballasting
weights(Figure C.1andFigure C.2).Bulkheadswerelocatedeithersideofthetowpostplateinordertogive
the midbody torsional strength. The mass and centroid of the hollowed sections was estimated using the
areainspectiontoolinAutoCAD.Theballastingweightmassesusedwerebasedontheactualweightsthat
wereavailableattheSSUTowingTank.AbreakdownofthemasscomponentsisgiveninTable4.6.
Table4.6Componentmassesusedinestimationofmodellongitudinalcentreofgravityandpitchgyradius
Component
Foam
Plywood
Heelfitting
Towpost
Resinandpaint
Ballastweightforward
Ballastweightaft
Ballastweightaft
Ballastweightaft
Totalmass

Mass/kg
10.9
0.225
1.5
1
2
2x9.07
9.07
5
0.665
48.5

are0.0042metresaftand
Basedonthis,thepredictedvaluesoflongitudinalcentreofgravity(LCG)and
).Thesecomparefavourablywiththerequiredvaluesof0.0040metresaftand0.534
0.577metres(or0.27
metresrespectively.Whilstitisrecognisedthatthisapproachearlyoninthedesignisveryapproximate,since
thetruemassdistributionofthemanufacturedhullisnotknownaccurately,itprovidessomeconfidencethat

37

Concept Design of a Fast Sail Assisted Feeder Container Ship


the model may be ballasted in the towing tank to the correct draught and trim, and still achieve a pitch
gyradiusclosetothevalueassumed,thusimprovingconfidenceintheresultsofanyseakeepingtests.
Turbulencestimulationstuds
Turbulencestimulationstudswerefittedtothemodeltoapproximatetheeffectofafullscaleboundarylayer.
Initiallyitwasplannedtofitthestudsatoneinchintervalsaroundthegirthinaccordancewiththemethoddue
toHughes&Allan(1951)asrecommendedbyITTC(2008b).However,onconsultationwithCampbell(2010b)
aft
itwassuggestedthatthesestudpositionsweretoofarforwardandthatauniformarrangementat10%
oftheforwardperpendicular(FP)wouldbemoreappropriate.Dimensionsofthestudsusedandacomparison
ofthealternativepositionsareshowninFigure4.8.

Figure4.8(a)Turbulencestimulationstuddimensions;and(b)comparisonoftheITTCandWUMTIA
recommendedpositions
Modelplans
Thetwomodelsweredesignedusingthe3DNURBSsoftwarepackageRhino,usingthehullformdefinitions
exportedfromMaxsurf.InRhino,theinternalshapeofthemodelswasdefinedsoastoallowthehollowing
out codes for the CNC milling machine to be created. In addition, 2D plans were also drawn up using
AutoCAD,describinginternalfeatureswhichwerenotclearfromthe3Dfiles.2Ddrawingsand3Dmodelsof
bothhullsareincludedinAppendixC.

4.2.3

Modelmanufacture
Stagesofmanufacture

Themanufacturingprocesscanbesummarisedinthefollowingstages:
1.

DrillholesinfoamblockstomountonCNCmillingmachine.

2.

Sizefoamblockstocorrectbreadthanddepth,andhollowoutallinternalpockets.

3.

Cutsternhullformsofbothmodelsfromabove.Cutjigsawlockingholesinkeel.

4.

Cutbowhullformsofbothmodelsfromportandstarboardsides.Thistakeslongersincethefoamblock
must be reoriented halfway through the cutting process due to the threeaxis limitation of the CNC
millingmachine.Cutjigsawlockingholesinkeel.

38

Hydrodynamic Design Development


5.

Removeallexcessmaterialmanually.Cutjigsawlockingpinstojointhetwohalvesofthehulls.

6.

Gluethetwohalvesofthehull,clampingthemovernight.

7.

Fillsurfacedeviationswithfillerandsandhullsmoothwith400gritwetanddrypaper[ITTC(2002)].

8.

CoathullsinathinlayerofAmpreg22resintosealandaddstrength.Leavetodryovernight.

9.

Furtherlightsandingfollowedbytwocoatsofgreycarbodyprimerspraypaint.

10. Mark design waterline using a height gauge on a flat surface. Install turbulence stimulators in desired
location.
11. Useepoxyresintoglueinplywoodplateformountingheelfitting.
HighmanufacturingtolerancesarespecifiedbyITTC(2002)soastoachievethecorrectdisplacementduring
testing. Due to the high accuracy of the CNC milling machine, both models were produced to within the
required onemillimetreinlength,beamanddepth.Theresinandpaintfinishdidnotaddsignificantlytothe
overall dimensions. The draught marks were measured using a height gauge, with an accuracy of one
hundredth of a millimetre, although they are marked by hand using a permanent marker pen which has a
thicknessof0.5millimetres,thuslosingsomeofthisaccuracy.Photographsofthemodelsatvaryingstagesof
completionareincludedinFigureC.3.
Modelquality
Whilst ITTC procedures were taken into account, the models produced did not adhere perfectly to the
recommendations. The main concern was the qualityof the surface finish exhibited, especially onthe stern
sectionsofbothmodels.Numerousporesinthefoamwerenoticed,duetothelowqualityoffoamspecifiedin
as a costsaving measure, which would contribute to increased drag during testing and thus inaccuracies in
results.Despitetheapplicationoffillerandcoatsofresinandpaint,theseholescouldnotbeentirelyremoved
andthusthemodelsurfacefinishissubstandardintermsofITTCrequirements.Whilstanallowanceismade
forsurfaceroughnessinthepostprocessingoftestresults,quantifyingtheeffectoftheseporesonthedatais
notwelldocumentedandassuchisnotaddressedinthisproject.
Anadditionalconcernregardedthehullfairness,especiallyatthejoinofthebulbousbowwiththehull.Hereit
wasnotedthatthedesiredcontinuityofhulllineswasnotexhibited,andthatflowseparationwaspossible,
whichwouldagainleadtoinaccuraciesinthemeasureddata.However,thisissuecannotbeattributedtothe
manufacturing process, due to the high accuracy of the CNC milling machine, and is instead traced to the
original design of hull lines in Maxsurf, where perfect fairing was not achieved. The effect of this on the
measuredresistanceisagainhardtoquantify,orevenqualify,withouttheuseofflowvisualisationtechniques
andmoredetailedanalysisthanispossiblewithinthescopeofthiswork.

4.3

Preparations for testing


Testmatrix

The main considerations in planning the required tests were the total time available, which was limited to
threedays,andthetimerequiredforeachrun.Bothconservativeandoptimistictestschedulesweredrawn
up,withthetimeallowedforeachrunmodifiedbetweenthetwo,asshowninTable4.7.Themaindifference
betweenthetwoschedulesistheabilitytocompletethesailingconditionrunswithbothportandstarboard
leewayangles.AfullbreakdownoftheconservativeandoptimisticschedulesisgiveninTableC.1andTable
C.2.Notethatbothschedulesallowextratimesothatthenumberofrunswouldnotbecompromisedifdelays
wereexperienced.Duringactualtestingthissparetimewasusedtotestthemodelsinthesailingconditionat
anadditionalspeed.

39

Concept Design of a Fast Sail Assisted Feeder Container Ship


Table4.7Timeallowedperruninminutesforbothconservativeandoptimisticschedules
RunType
Calmwater
Sailingcondition
Inwaves

Conservative
10
15
20

Optimistic
7
10
15

The models were prepared for testing following the recommendations of the ITTC (2002,(2008b), requiring
that:

Theybeloadedtowithin0.2%ofthedesigndisplacement.Thiswascarriedoutbyfirstweighingthe
modelwithoutballastandthenapplyingtheappropriateamountofballastingweight;

Themeanofthedraughtsforwardandaft,portandstarboardbewithintwomillimetresofthedesign
draught;

Thetowforcebeappliedatthe

andinlinewiththepropellershaft.

Itisrecognisedthattheaccuracywithwhichtheseguidelineswerefollowedwasnotentirelysatisfactoryfor
commercial testing purposes. Acknowledging the recommendations when preparing the models for testing
wasconsideredimportantinordertogivethebestpossibleconfidenceinthequalityofresultsobtained.
Momentcorrections
Variouscorrectionswereappliedtothemodelsinordertosimulatethefullscaleshipmoreclosely,including
compliancewiththeITTCrequirementsalreadydescribed.
(i)

Skinfrictioncorrection

Theskinfrictioncoefficientatmodelscaleisalwayslargerthanforthefullscaleship,astheReynoldsnumber
isloweratmodelscale.Ifthetowingpointisabovethecentreofwettedarea,thiswillresultinabowdown
,iscalculated
trimmomentwhichmustbecorrectedbyapplyingballastaft.Theskinfrictioncorrection,
asshowninEquation(4.4).

(ii)

(4.4)

Thrustlinecorrection

In addition, if the model is towed from a point which above the thrust line, a second correction aft of the
centreofgravityhastobeappliedtocorrecttheresultingbowdowntrimmoment.Ifthetotaldragofthe
modelisknownbeforehand,thetotalcorrectioncanbecalculatedas(refertoFigure4.9):

,
(4.5)
where and are the lever arms between the towing point, and thecentre of wetted area and thrust line
respectively.
The total trim moment correction was calculated for theappropriate range of test speeds, as seenin Figure
C.4.Thecorrectionswereappliedusingaslidingweightmountedonaguiderailscrewedtothemodelsdeck
(seeFigureC.3(f)).

40

Hydrodynamic Design Development

Figure4.9Schematicshowingtotaltrimmomentcorrectionlevers and
(iii)

Saildriveforcecorrection

Thedriveforcegeneratedbythesailswillalsocauseabowdownmoment;torepresentthischangeintrim,
ballastisaddedforwardofthecentreofgravityofthemodel.Thedriveforcegeneratedbythesailsatthe
samescaleofthemodelcanbecalculatedusingaveragevaluesforapparentwindangleandspeed;thedrive
force is then multiplied by the levers from the centre of effort to the centre of gravity to find the required
correctionof15.5Nm.
(iv)

Sailheelforcecorrection

When operating at an angle of heel the aerodynamic sideforce will have a vertical component acting
sin .Thisforceissimulatedbyaddingaweightontopofthetowingpost,
downwardsandequalto
howeverduetothesmallmagnitudeofthepredictedangleofheel,thecorrectionamountstoonly3.2N.
Measurementsystem
The experimental setup at the SSU towing tank consisted of a WUMTIA single tow post dynamometer
constrainingthemodelinsurge,swayandyaw,andaLargeYachtheelfittingwhichisaheelandtrimable
towpostfitting.Whentestingthemodelsintheuprightcondition,resistanceandsideforcemeasurements
wereobtainedfromthedynamometer,withyawmomentmeasurementstakenfromthetowpostandheave
measured from the position of the model relative to the dynamometer. For tests in the heeled and yawed
conditions, the tow post assembly was rotated within the dynamometer to the desired leeway angle and
ballast shifted in the model to achieve suitable heel before locking the heel fitting in that position, thus
minimising the roll moment and removing any influence of the flexibility of the system on the angle of heel
[Campbell&Claughton(1987)].
Whentestinginwavesaswordtypewaveprobewasusedtomonitortheformofthegeneratedwaves.The
probewasmountedasfarfromthemodelaspossibletowardsthefrontrightcornerofthecarriagesoasto
avoidmeasuringawavesignalfromthemodelsownwavepattern,andtoavoidthewavepatternoftheprobe
impingingonthemodel.
All data acquisition was throughtheWUMTIA software LASSO, via an analogue to digital converter. For the
waveprobetraceaChurchillsignalprocessingunitwasalsoemployed.Calibrationofthesystemcomponents
wasconductedonthecarriage,duringwhichtimeitbecameapparentthattherollmomentoutputfromthe
heelfittingwasunsteady.Asitwasjudgednoncriticaltothetesting,rollmomentmeasurementswerenot
recorded.
Wavemaker
Thewavefrequenciestobeinputintothewavemakercomputerwerecalculatedusingsimplewavetheoryand
thevaluesofmodellengthgiveninTable4.5.Fromthedispersionrelation,assumingdeepwater7:
2

,Bertram(2000)

41

(4.6)

Concept Design of a Fast Sail Assisted Feeder Container Ship


andthewavefrequency( )inHertzisderivedtobe:
1
2

(4.7)

The resulting frequencies input are given in Table 4.8. The wave amplitude was constant for all runs, and
chosenasanapproximatelyonequarteroftheshipdraught,equaltothreecentimetres.Onlyonevaluewas
testedsincetheresultscanbescaledtoanywaveamplitudeduringpostprocessing.
Table4.8Wavefrequenciesinputintowavemaker
Wavelength

/%
50
100
150
200

Frequency/Hz
HullA HullB
1.209 1.215
0.855 0.859
0.698 0.702
0.604 0.608

4.4 Analysis of results


Throughoutthecourseoftestingatotalof112measuredrunswerecompleted.Ofthese83aredeemedtobe
usefulforfurtheranalysis.Theusefuldataisalmostequallydividedbetweeneachhull,withHullAandHullB
having43and40runsrespectively.Increasingthesizeofthedatasetandrepeatingrunsincreasesconfidence
intheresultsbyreducingrandomandsystematicerror.
Themajorityoftherunsthatdonotyieldmeaningfulresultsrepresentrunscorrespondingtounsatisfactory
calibrationofthedynamometer.Someproblemswereencounteredwhentryingtoachieveanalmostconstant
zero value between runs. This led to recalibration until the problem had been rectified, wasting valuable
testing time. In some cases corrections have been applied to results. When an unexpectedly large drag
measurementisobservedduetoashiftingzerovalue,anacquisitioncanbemadewiththemodelinthesame
conditionbutwithzerocarriagespeed.Thiseffectivelydefinestheactualzeroandhencethedragisobtained
bydifferencingthemeasureddragandthemeasuredzero.Althoughthiscorrectionhasbeenappliedinafew
instancesthisisnotanefficientuseoftime.Thustimewasspentrecalibratingtoavoidthis.Theparameters
measuredbythevariouschannelsofthedynamometerandoutputbyLASSOareasfollows:

Modelspeed,

Dragforce,

Sideforce,

Trimangle, ;

Heave;

Rollmoment;

Yawmoment,

Waveamplitude, ;

Dragzero.

;
;

Not all of these data channel outputs are used in the analysis. This is either because these values are not
neededortheresultsarenotreliable.Thisisprobablyduetotheinstrumentationonthedynamometerand
the heel fitting not working correctly. In some cases connection problems to the signal box occurred.

42

Hydrodynamic Design Development


Fortunately,allthevariablesthatareneededfortheanalysisproducereliableresultsi.e.
providesufficientinformationrequiredforthepostprocessingoftheresultdata.

and

4.4.1

Uprightcondition
Processingofmodelscaledata

Testing procedure and data analysis have been carried out for the upright condition in line with ITTC
recommended procedures [ITTC (2008b)]. The upright condition results do indicate a misalignment of the
model from the centreline of the tank. All the tests in this condition are required to be carried out at zero
leeway angle, however, the results show nonzero yaw moment and hence indicate model misalignment.
Alternatively this could suggest asymmetry in the model about the centre plane. The alignment was simply
conductedbysightingthemodelfromtheendofthetankandfromthecarriage.Thisisnotthemostrigorous
approach,butwasthebestmethodavailableduringtesting.Acorrectionneedstobeappliedtotheresistance
resultstoaccountforthismisalignment.Fortunatelythiscanbeachievedwithrelativeeaseasthemodelhas
been tested at a number of leewayanglesto assess theeffects of thesail system(Section 4.4.2). From this
datathemisalignmentcanbedetermined.ConsiderHullAasanexample.Table4.9showsthemeasuredyaw
momentatarangeofleewayanglesforthehighspeedsailingcondition(approximately25knotsshipspeed).
Table4.9ModelmisalignmentcorrectiontoyawmomentHullA
/

5.0
2.5
0.0
2.5
5.0

11.07
5.97
0.32
6.04
11.32

10.16
5.07
0.58
6.94
12.22

/
11.36
5.68
0.00
5.68
11.36

8.159E03
7.266E03
6.878E03
7.749E03
8.820E03

8.340E03
7.284E03
6.900E03
7.702E03
8.956E03

Oneexpectsalinearrelationshipbetweenyawmomentandleewayanglepassingthroughtheorigin.Figure
4.10 (left) shows the actual relationship. One can appreciate the linearity, however the trend does not pass
through the origin (further illustrated by the equation of the best fit line) indicating the magnitude of the
misalignment.ThecorrectedyawmomentinTable4.9isobtainedbyaddingafactortotheyawmomentdata
so that the best fit line [Figure 4.10 (left)] intersects the origin. To account for the fact that this data only
representsonespeedtheresistanceisnondimensionalisedusing

.
2

(4.8)

Thisallowsforthecorrectiontobeappliedtoanyspeed.Theyawmomentisthenadjustedtocoincidewith
the new best fit line, giving symmetry port and starboard. Now this corrected yaw from the best fit line is
andafourthorderpolynomialisfittedthroughthedata[seeFigure 4.10(right)].
plottedagainst
,
This allows for the
values on the best fit line to be obtained. These values of
are subsequently
averagedtoobtainsymmetryforpositiveandnegativeyawmoment.Theseaveragevaluesareplottedanda
quadraticpolynomialisfittedthroughthedata.Itisthisequationthatprovidesthecorrectiontotheupright
,resultingfromleeway,isgivenintheform
cases.

(4.9)

where , and areconstantsdeterminedduringthecurvefittingoperation.Thusthecorrectedcoefficientof


totalresistanceis

43

Concept Design of a Fast Sail Assisted Feeder Container Ship

(4.10)

as isthecorrectedcoefficientoftotalresistancewhentheyawmomentiszero.Thenclearlythecorrected
modeldragisgivenby
1

(4.11)

9.0

15.0
y=2.271x 0.904
R=0.998

10.0
5.0
0.0
5

5.0

CTM measuredx103

Measuredyawmoment/Nm

y=5E05x4 +0.000x3 +0.018x2 0.043x+6.890


R=1

8.5

8.0

7.5

7.0

10.0

Leewayangle/degrees

15.0

Correctedyawmoment/Nm

6.5
12.0

8.0

4.0

0.0

4.0

8.0

12.0

Figure4.10Illustrationofmisalignment(left);andyawmomentcorrection(right)
Figure 4.11 presents the corrected model resistance over the range of tested speeds and allows an initial
comparisonofHullAandHullBbeforeanyscalingprocedureshavebeenimplemented.
7.0
6.0
HULLA

5.0
RTM /N

HULLB
4.0
3.0
2.0
1.0
0.0
0.2

0.4

0.6

0.8

1.0

1.2

VM /ms1

1.4

1.6

Figure4.11CorrectedmodelresistanceforHullAandHullB

ImplementationofITTC1978PerformancePredictionMethod
TheITTC78PerformancePredictionMethod[ITTC(2008d)]iscurrentlytherecognisedprocedureforscaling
modeltestresultsforcommercialshipstopredictfullscaleperformance.Intermsofresistancethismethod
dividesthecomponentsofresistanceasfollows
1

44

(4.12)

Hydrodynamic Design Development


where
1

0.044

10
5.68

0.000125;

0.6
1
2

and

(4.13)
(4.14)

10 ;

(4.15)

(4.16)

represent,incoefficientform,thewavemakingresistance;resistanceallowanceforhullroughness;correlation
allowancebetweenmodelandshipandthestillairdragresistancerespectively. isdeterminedusingthe
ITTC57correlationlinedefinedinEquation(4.2).TheITTC78methodhasbeenassessedandisreasonableto
applyinthisinstance,exceptforthefollowing,that:

Anallowanceformodelroughnesswillneedtobeintroduced;

Areasonableformulationforthestillairresistancemustbederived.
Allowanceformodelroughness

Duetobudgetconstraintsandlimitationsinmanufacturingtechniquesthemodelsproducedwerenotglass
smooth as is assumed in the ITTC 78 method. Consequently if one commences the analysis without
accounting for model roughness the power requirement is overestimated. This problem is apparent when
observingtheformfactordeterminedfromtheProhaskaplot[ITTC(2008d)].
affect the form factor and ultimately the ship resistance. To account for the model
Variations in
must be introduced. In this case it does not
roughness, enabling a realistic power estimate, a factor
make sense to use Equation (4.14), the formula proposed by Townsin (2003), as this is applied at full scale
is
alongside the correlation allowance. If one applies it without the correlation factor at model scale
predominatelynegativewhichmakesnophysicalsense.Applyingitwiththecorrelationfactoratmodelscale
isnotreasonable;itwouldmeanusingthecorrelationallowancetwiceandapplyingittoadjustmodelscale
values before scaling. For these reasons a formula has been used from a previous revision of the ITTC 78
method,knownastheBowdenDavisonequation:

105

0.64

10 .

(4.17)

Thefinalcomplexityisindeterminingorestimatingthemodelsurfaceroughness.Measuringtheroughnessis
notimpossible,butrequiresspecialistmethodsthatarenotavailabletotheauthors.Havingaroughnessvalue
fortheHDFwouldnothelpmattersasitiscoatedinresinandpainted,asexplainedinSection4.2.3,andthusa
standardvalueisofnouse. Thisprocessisincrediblyeasyfortheshiphullroughnessasavaluequotedby
80.5 10 metres).
InternationalPaints[Willsher&Solomon(2010)]isused(
The model roughness then must be estimated. It is expected that it must lie in the range 0.1 50 10
metreswhichrepresentstheroughnessboundarybetweenplateglassandbaresteelplate[Molland(2009)].
Intheabsenceofanydefiniteinformationitseemsreasonablethatastheformfactorisameasureofhulldrag
changessowilltheformfactor(determinedfromtheProhaskaplot),that
canbevaried
andthatas
in the range defined to achieve a reasonable form factor. As it is believed that the hull forms are well
representedbytheHoltropregressionitisexpectedthattheformfactorsdeterminedinSection 4.1.5should
be fairly close to the actual values. These values are used as an approximate target to determine the

45

Concept Design of a Fast Sail Assisted Feeder Container Ship


roughness of the model. Table 4.10 illustrates the sensitivity of the form factor when varying the surface
roughness.
Table4.10Sensitivityofformfactorduetovaryingsurfaceroughness

10
Holtrop
1.188
1.182

1
HullA
HullB
HullA
HullB

0.1

1.349
1.351
4.048
10 @25knots
3.297

1.229
1.230
3.777
3.030

1.178
1.179
3.662
2.918

1.145
1.145
3.587
2.844

6
1
Prohaska
1.120 1.083
1.120 1.082
3.530 3.445
2.788 2.705

10

30

50

1.055
1.054
3.382
2.643

1.032
1.031
3.330
2.592

0.911
0.909
3.057
2.325

0.851
0.849
2.920
2.192

Althoughtheregressionanalysisvalueisthoughttobereasonableitwouldbequestionabletoblindlyselect
theroughnessvaluegivingtheclosestformfactorresultoftwomicrons.Itishoweverexpectedthatmodel
surfaceroughnessistowardsthelowerendofthesuggestedrange.Twomicronsseemsaverylowroughness
forthesurface,andhencetobemoreconservativearoughnessofthreemicronswasselectedforthemodel
surface.
isnotstraightforward,as
Theultimatedependenceoftheshipresistanceonthechangesin 1
and
whenthelatterisincreasingtheformerisdecreasing.Inthedeterminationof [Equation(4.13)]theterm
mustbereplacedwith

.Itmustofcoursedependonthemagnitudesof 1
and
as
towhetherthereisanetincreaseordecreasein .Thecomplexityisfurtherincreasedbythedependenceof
on 1
and .Thesimplestwaytoevaluatethisthenistostudythenumericalresultsatthedesign
speed. This data is presented in Table 4.10 alongside the form factors. This variance is better illustrated
graphically(Figure4.12).ThenondimensionalformprovidesnorealcomparisonbetweenHullAandHullBas
the wetted surface areas are somewhat different. The resistance of each hull is however decreasing as the
model roughness is increased. This is a consequence of the scaling procedure and illustrates that the ship
resistance significantly varies with the model roughness, highlighting the importance of formulating a
reasonableestimate.

CTS x103 @25knots

4.5
HULLA

4.0

HULLB

3.5
3.0
2.5
2.0
0

10

20
30
Modelroughness,kSM x1x106

40

50

Figure4.12VariationofCTSwithsurfaceroughness
Afinalcorrectionthathasbeenappliedtothemodelresistanceaccountsfortheadditionaldragresultingfrom
the trip studs. This has been done in accordance with a procedure outlined by Molland et al. (1994) and is
(seeAppendixD.1).
representedbyafurtherincreaseof

46

Hydrodynamic Design Development


Correctionofstillairresistanceexpression
TheITTC78methodprovidesaformulaforthedeterminationofthestillairresistanceoftheship[Equation
(4.16]. If one considers the structure of this equation and assesses the dimensions it is fundamentally
when this coefficient should be dimensionless. In
incorrect. It has dimensions of
/
particular it should represent an air drag nondimensionalised by the standard hydrodynamic parameters
1
asitissimplysummedwiththeothercoefficients[Equation(4.16)].Ifonefollowsthesestepsthe
2
airresistancecoefficientisrepresentedas

(4.18)

For all the analysis in this section, Equation (4.18) replaces Equation (4.16). Attempting to utilise Equation
(4.16) further confirms its incorrectness as the results are unreasonably large. is chosen in line with
Molland(2009)andtakenas0.5,whichisalowvalueassociatedwiththeaerodynamiccharacteristicsofhaving
isdeterminedfromthelayoutas
aslopedforwardsuperstructure.Thetransverseareaabovethewaterline
and
616.23 ; the air and sea water density are calculated for a temperature of 25 as 1.1855
respectfully and the wetted surface areas are given in Table 4.3. This results in
1023.38
7.364 10 forHullAand
7.7304 10 forHullB.
Uprightconditionresults
Onecannowbegintoevaluatetheresistancecharacteristicsofthehulldesigns.Theeffectivepowerisgiven
overtherangeoftestedspeedsinFigure4.13,fulldetailsareprovidedinTableD.2andTableD.3.
25000
HULLA

PE/kW

20000

HULLB

15000
10000
5000
0
5

11

14
17
VS/knots

20

23

26

Figure4.13Effectivepowercomparison
ThefirstconclusiontodrawfromFigure 4.13isthattheeffectivepowerofHullBislessthanHullAacrossthe
entirespeedrange.ThisisatrendthathasalsobeenobservedinSection 4.1.5andisduetothedifferencein
(naked) wetted surface area between the two hulls. In terms of dimensionless resistance components, by
comparing results in Figure 4.14, Hull Bs total resistance coefficient is the lesser over the Froude number
range;theconverseofwhatwasobservedusingtheHoltropregression.

47

Concept Design of a Fast Sail Assisted Feeder Container Ship

4.5
4.0

HULLA Cts

HULLA Cws

HULLA Cfs

HULLA Cvs

3.5
HULLB Cts
HULLB Cfs

3.0

HULLB Cws
HULLB Cvs

2.5

3.0

CT,CV,CF,CW x103

CT,CV,CF,CW x103

3.5

2.5
2.0
1.5

2.0
1.5
1.0

1.0
0.5

0.5

0.0

0.0
0.05

0.15

0.25

Fn

0.05

0.35

0.15

0.25

0.35

Fn

Figure4.14Breakdownofresistancecomponentsfor(a)HullA;and(b)HullB
.The
FromFigure 4.14onecanappreciatetheeffectoftheformfactor,thedifferencebetween and
reasonHullAhasalarger isduetothe beingaroundtwicethemagnitudeofthe forHullB. and
aresimilarforeachhullduetothefactthat
issimilar,theonlyvariableaffecting ,andalsotheform
factorsareverysimilar.ThevariationofthesecomponentswithFroudenumberisnotsubstantial.Thewave
resistance never constitutes more that 50% of the total resistance. It is somewhat expected that for the
conceptoperatingatthedesignFroudenumberthatviscousandwavepatternresistancearecomparablein
magnitude,asobserved.OneexpectsHullBtohavelowerwavemakingresistanceasitisdesignedtoreduce
the wave pattern generated by the stern, see Section 4.1.3. It may be noticed that the viscous and wave
patterncomponentsdonotseemtosumtogivethetotal.OneshouldconsultEquations(4.15)and(4.16)and
and arenotpresentedinFigure4.14howevertheystillcontributeto .
noticethat

25000
HULLA Holtrop
HULLB Holtrop
HULLA
HULLB

20000

PE /kW

15000
10000
5000
0
0

10

15

VS /knots

20

25

30

Figure4.15ComparisontoHoltropeffectivepowerforHullAandHullB
Figure 4.15 and Figure 4.16 present comparisons between resistance measurements obtained from models
isnotincludedasnoallowance
testsandthoseestimatedinSection 4.1.5.Notethatintheseinstances
for air resistance is accounted for in the Holtrop regression. From Figure 4.15 it can be seen that the
predictionsandthemeasurementsarefairlysimilarespeciallyatlowspeed.ThepredictionforHullAisbetter
thanitisforHullB.ItislikelythatHullBisnotwellrepresentedbytheHoltropregressionasitssternformis
notstandardformerchantships.Thiswouldaffectthevalidityofthecoefficientofwaveresistance,asseenin
Figure 4.16(b),andsuggestswhythepowerestimateforHullAisclosertothemeasuredvaluethanforHullB

48

Hyydrodynamic Design Deevelopmentt


(Figurre 4.15).Italsoseemsreasonablethat theHoltroprregressionoveerpredictstheresistanceffurtherasthee
speed
d increases. At
A higher speeed the propo
ortion of resisstance attribu
uted to the w
wave pattern increases and
d
henceeasitisthisp
partofthetottalresistancethatisbeingo
overestimated
d,thedifferen
ncebecomesgreater.Thiss
canbeeobservedinFigure4.16(aa)and(b).
HULLLA

HULLB

HULLLA Holtrop

HULLB Holtrop
H

HULLA
H
H
HULLA
Holtrop
2
2.0

3.6

1
1.6

CW x1000

CTSCAAS x1000

3.2

1
1.2

2.8

0
0.8

2.4

0
0.4

2.0

0
0.0
0.05

0.15

Fn

0.25

0.35

0.05

0.15

Fn
n

0.25

0.35

2.00

1.80

1.70

1.90
1.85

CVS x1000

1.65
1.60
1.55
1.50

1.80
1.75
1.70

1.45

1.65

1.40

1.60
0.05

0.15

0.25

Fn

HULLA
HULLB
HULLA Holtrop
HULLB Holtrop

1.95

HULLA
HULLB
HULLA Holtrop
p
HULLB Holtrop
p

1.75

CFS x1000

HULLB
B
HULLB
B Holtrop

0.35

0.05

0.15

F
Fn

0.25

0.35

Figurre4.16Comparisonofind
dividualresistancecompon
nentstothoseobtainedusin
ngtheHoltropregression.
CAAS;(b)CWS;((c)CFSand(d)CVS
(a)CTSC
Theree is a small difference
d
in the
values obtained from the preedictions and the model tests.
t
This iss
unexp
pectedasinb
bothinstancesITTC57fricctionlinehassbeenused. Itisthought thatHullspeeedmusthavee
used theincorrect
,ormoreelikely,adiffferentvalueo
ofthekinemaaticviscosity, asnoaccoun
ntistakenforr
urface tempeerature, when
n determining the ship Reyynolds numbeer. This of co
ourse affects the values off
sea su
,w
whicharefurttheraugmentedfromoneaanotherbyth
hefactthattheformfactorrsusedineach
hmethodaree
different. One caan also see how
h
the Holttrop form factors are diffferent for eaach hull. Other than thee
nitudesthetrrendsofthereesultsareveryysimilar.
differencesinmagn
henselectingtthemostsuitaablehullform
m(seeSection
n6.1).Itisimportanttobee
Theseeresultsareconsideredwh
clearinthemindofthereaderthatthetestin
nginthissectiononlyconsiderstheperfo
ormanceofth
henakedhull..
efits currentlyy
Although there aree appreciablee differences in the performance of thee hulls there may be bene
hatcanoutweeighthis.
unacccountedforth

4.4.2

Assesssmentofhyd
drodynamicfforcesresulttingfromuseeofsails

Anam
mountofheelandleeway willbeexperriencedinallw
windconditio
ons,exceptfo
orrunningdeaadbeforethee
wind,byanyshipthatusessailsforpropulsio
onwhenunderway.Heelan
ndleewaypro
oduceforcesaandmomentss

49

Concept Design of a Fast Sail Assisted Feeder Container Ship


that balance the aerodynamic sideforce generated by the sails [Kirkman & Pedrick (1974)]. The underwater
bodyoperateslikealiftingsurfacewhichgeneratesanamountofsideforceSFwhichbalancestheforceofthe
sails.Associatedwiththisliftforcethereisaninduceddragcomponent.Itisimportantthatthehullformis
able to generate enough sideforce at small leeway angles in order to minimise this component and hence
increase the thrust reduction from the sails. These forces must be quantified by running the models at a
numberofheelandleewayangles.Thisofcourseisnotastandardpracticeinashipconceptdesignandhence
knowledgemustbetakenfromthedesignofsailingyachtswithregardstotestingproceduresandevaluation
ofperformance.Theforcesresultingfromthechangesinorientationarenamelyaresistancecomponentanda
sideforcecomponent,actinginorthogonalplanes.Theresistancemeasuredofcourseisthetotaldragforce
howeveraportionofthisisattributedtotheuseofsails.Themagnitudeofthesideforcewilldeterminehow
muchtheorientationoftheshipchangesduetoaforceonthesailsandhencethemagnitudeoftheinduced
resistancethatisexperienced.Beingabletoestimatethisisveryimportantasitwillallowforanassessmentas
towhetherthebenefitofthesailsoutweighstheinducedresistanceduetotheleewayandacomponentdue
toheel(Section6.1.1).
CorrectionshavebeenmadetothemeasurementsofresistanceandsideforcesimilartothosemadeinSection
4.4.1.Inthisinstancetheresistancehasbeencorrectedsothatthevalueatzeroheelandzeroleewayisthe
same as the upright resistance measured (after being corrected for misalignment) in Section 4.4.1. The
sideforcehasbeencorrectedsothatitiszerointheuprightcondition.Inanyinstancetheseerrorsaresmall
withthemeasuredsideforceatzeroheelandzeroleewayat15.9knotsbeing0.02N.Thesecorrectionshave
beenappliedbyconsideringalinearfitofdatathroughagraphofresistanceversussideforcesquaredforeach
heelangle.Therearemanyareasofuncertaintyandsourcesoferrorthatcouldhavecausedtheseproblems.
Uncertainty isquantified inSection 4.4.4. Sources of error identifiedby the authors are most likely resulting
fromthedynamometersetup,seeSection4.4.1.Alignmentisanunavoidableproblem.Asasinglepostsystem
is used, which is fairly flexible, the leeway angle set when the model is static is not necessarily the angle
undertakenwhentheforcesaremeasured.Thehydrodynamicforcescanbesufficienttodisplacethemodel
fromthesetleewayangle.Thisproblemisamoresignificantforlargermodels[Claughtonetal.(1998)].There
canalsobeproblemsifthereisinterferencebetweensensorsdesignedtomeasureaparticularforcebyother
forces.Itisthoughtthatthiscouldcertainlybeaproblemwiththeyachtheelfittingused,astheresultsfrom
some of the channels are meaningless. Although these are not the channels needed for the sideforce and
resistance,thisinterferencecouldhaveaffectedthemeasurements.
TheresistanceisbrokenupintocomponentsassuggestedbyCampbell&Claughton(1987),
,

where istheuprightresistance; istheinducedresistanceduetoleewayand


heel.ThisbreakdownisperhapsbetterunderstoodwhenconsideringFigure4.17.

50

(4.19)
istheresistancedueto

Hydrodynamic Design Development

Figure4.17Illustrationofresistancecomponentsduetoheelandleewayasafunctionofsideforce
,byobtainingtheintersectionofthebestfitlinewiththe
axis
Forthecaseofzeroheel
andbysubtractingtheuprightresistance theerrorcanbedetermined.Thisiscorrectedbyadjustingthe
datapointsuntilthebestfitlineinthiscasegoesthrough .Thisisappliedtotheothertwoheelcasesatthis
speedandanothercorrectionisdeterminedfortheotherspeedtests.Notethatalthough iszeroatzero
heeldoesnotmeanthatitincreasesasashipisheeled.Atsmallanglesofheelanegative canbeobserved.
Campbell&Claughton(1987)reportdifferencesinmeasureddataduetothismethodofbreakingdownthe
components and states It should not be assumed that the measured data will conform neatly to this
descriptionofresistance.
Scalingproceduretofullscale
A scaling procedure is adopted using Froude scaling, as suggested by Campbell & Claughton (1987). This is
appliedto and asthefullscalevalueof isalreadyknownfromSection 4.4.1.TheFroudescalingis
implementedbyexpressingtheresistancesinnondimensionalcoefficientform.Itisnotreallyimportanthow
thesearenondimensionalisedasthereisnointerestinthecoefficientvalue.Itissimplyusedtoobtainaship
scaleforce,say

,
2

(4.20)

where canbetheprojectedarearelativetotheincidentfloworthewettedarea.Inanycase,whenone
appliesFroudescalingtheresultis

and

whereisthemodelscalefactor.Similarly

(4.21)

aregivenby

and

respectively.

51

(4.22)
(4.23)

Concept Design of a Fast


F
Sail Asssisted Feedeer Container Ship
Reesults
All results presented
p
aree those at full scale. Tab
ble D.4, App
pendix D, gives the valuess of the resistance
componentsandsideforceeforeachheelandleewayycombination
ntestedatbo
oththehigh(2
25.5knots)an
ndlow
presentedasproposedinFFigure4.18.
(15.9knots)speeds.Thesseresultsarep

6.4E+05

RTOT /N

5.9E+05

0 HU
ULLA
2.5 HULLA
H
5 HU
ULLA

5.4E+05

4.9E+05
0.0
0E+00

2.0E+11

4.0E+11

SF2 /N2 6.0E+11

8.0E+11

1.0
0E+12

5.2E+05

RTOT /N

4.7E+05

4.2E+05

0 HULLLB
2.5 HU
ULLB
5 HULLLB

3.7E+05
0.0
0E+00

1.0E+11

2.0E+11

SF2/N2 3.0E+11

4.0E+11

5.0
0E+11

RTOT /N

2.5E+06

2.0E+06

0HULLA
2.5 HULLA
H
5 HU
ULLA

1.5E+06
0

2E+12

4E+1
12

SF2/N2 6E+12

8E+12

1E+13

4E+1
12

SF2/N2 6E+12

8E+12

1E+13

2.10E+06

RTOT /N

0 HULLB
B
2.5 HULLLB
5 HULLB
B

1.60E+06

1.10E+06
0

2E+12

ofsideforceaandresistancee,(a)HullA15.9knots;(b
b)HullB15.9
9knots;(c)Hu
ullA
Figure4.18Evaluationo
25.5knots;(d
d)HullB25.5
5knots
FromFigure4.18itcanbeeseenthatthetrendsareaapproximatelyylinearasiseexpected.How
wever,HullBseems
orHullAthe resistancevalueatzerosideforcetend
dstofallbelow
wthelinearffit,this
tofitthistreendbetter.Fo

52

Hydrodynamic Design Development


hasbeenobservedbyCampbell&Claughton(1987).Itisdifficulttodrawanyconclusionaboutthisfromthe
resultsasthereareonlythreedatapointsforeachdatasetandhenceitcannotbeclearwhichofthepointsis
infactstrayingfromthelineartrend.Itshouldbenosurprisethatthesideforceandresistancegetlargerwith
speedandleewayangle.Itisalsoseen,andsomewhatexpected,thatthesideforceandresistancearelower
forHullBthantheyareforHullAatthesamespeed,thisisduetotheHognersternofHullA.Alargeside
forceisdesirableasitwillpreventsignificantleewayanglesexistingwhentheshipisusingthesails.However,
this accompanies a large resistance which is not desirable. Hence evaluating hull form performance is not
trivialandisdependantonthewindconditionsandthesaildesign.ThisisconsideredindetailinSection6.1.1
Thatsaidonecanquicklycomparetherelativemagnitudeoftheresultsforeachhull.IfonecomparesFigure
4.18 (a) and (b) for the low speedcase,bearing in mind that sideforce is squared, the difference in the side
force between the hulls is around twice the difference in the resistance. This suggests then that Hull A will
performwell,astheincreaseinsideforce,whichisdesirable,isapproximatelytwicetheincreaseinresistance,
which is not desirable, and hence the benefits may outweigh the drawbacks. For the high speed case,
comparingFigure4.18(c)and(d),thereisnotaverysignificantdifferenceinthesideforcebetweenthetwo
hulls, however, Hull A still has a significantly larger resistance than Hull B. This suggests that HullA will not
performsowellathighspeed.
In most instances the resistance is greater, at high leeway, for the low heel angles. This is thought to be
attributabletothewettedsurfaceareachangesastheshipsheel.Thishasnotbeenexamined.However,it
wouldseem(fromthehighleewayresults)thatitisreducingastheshipheelincreases.
Asidefromuncertaintiesanderrorsourcesintheexperimentalprocedure(discussedinSection4.4.4)thereare
some reasons to believe that the analysis method could be affecting the results. An allowance for model
roughnesshasonlybeenmadeinthedeterminationof howeverroughnesswillcertainlyaffecttheother
components of resistance and the sideforce. Hence the resistance and the sideforce will be overestimated.
Thiscanonlyberectifiedbyaddingsomeallowance,howevernostandardallowanceisknowntotheauthors.
Thisispossiblyaconsequenceoftheincreasedcomplexityofthenatureoftheviscousboundarylayerinthese
conditions.TheFroudescalingmethodusedissomewhatstandard,howeveritisnotedthatitisdifficultto
develop a scaling procedure to take full account of viscous flow differences between model and full scale
[Campbell&Claughton(1987)].

4.4.3

Addedresistanceinwaves

Theresistanceofthemodeladvancinginwavesofprescribedlengthhasbeenmeasured.Thisallowspossible
performanceinwavesandhencetheoperationalcapabilityincertainweatherconditionstobeassessed.The
addedresistanceatmodelscalehasbeendeterminedusingEquation(4.24)forvaluesatthesamespeed.

(4.24)

The added resistance is presented in a nondimensional form that has been proven by StromTejsen et al.
(1973) to give consistent results when implemented on model test results of varying scale. This gives
confidencethatthiscoefficientformcanbeusedtoscalevaluesofaddedresistancetoshipscaleandforthe
necessarywaveamplitude.Theaddedresistancecoefficientisthusdefiedas

(4.25)

Addedresistancehasbeentestedattwospeeds;alowspeed(HullA15.9knots;HullB16.0knots)anda
highspeed(HullA25.5knots;HullB25.6knots).Thesespeedsaresimilarandaretobeassumedthesame
forpurposesofhullcomparison.Thedifferencessimplyarisefrominaccuracyinthecarriagecontrolbetween
setandachievedspeed.Table4.11presentstheresultsforthetestedsituations.

53

Concept Design of a Fast Sail Assisted Feeder Container Ship


Table4.11NondimensionaladdedresistanceforHullAandHullBatbothtestedspeeds
HullA

/%
52.05
104.37
155.60
220.72
50.31
101.13
151.42
196.91

Speed
Low

High

0.527
3.923
1.606
0.691
1.107
3.415
3.522
0.656

HullB

/%
48.21
103.25
153.92
181.25
49.74
100.09
149.86
194.71

1.086
3.732
1.853
0.394
1.164
2.963
4.627
1.433

Figure 4.19 represents Table 4.11 graphically. The data points have been joined with straight lines as
insufficientwavelengthswereexaminedtorealisticallyrepresenttheaddedresistanceprofile.Attemptshave
beenmadetotestatwavelengthswherethepeakintheaddedresistanceexists,howeverthishasnotquite
been achieved and due to restrictions in the testing schedule there has been no time to rerun other
conditions.Itseemslikelythatthepeaksineachofthecasesexistsomewherebetweenthesecondandthird
datapoint.Asanapproximationoneexpectsthepeakstoexistforthelowspeedcaseatawavelength120%
of ship length and for the high speed case 140% of the ship length for both hulls. This trend of decreasing
wavelength peak with decreasing speed is also noticed by Salvesen (1978). A numerical assessment of the
addedresistancewillbecarriedoutinSection 7.1.Thisshouldprovidesomevalidationofthelocationsofthe
peaks to help the assessment of the efficiency of the fast feeders operations (Chapter 6). In any case one
expectsthemagnitudeoftheaddedresistancepeakstobelargerathighershipspeed.
Themagnitudeoftheaddedresistanceisessentiallyameasureoftheseakeepingcharacteristics(heaveand
pitch)ofeachofthehulls[Wilson(1985)]whenoperatinginaspecifiedseaway.Inaqualitativesense,these
resultscanbeusedtocomparetheseakeepingqualitiesofeachhull.Oneshouldbearinmindthedifferencein
theshipspeedbetweenthetwohulls,howevertheeffectofthisisexpectedtobesmall.
At low speed Hull A seems to performs better at low wavelengths (below 80%) but not as well at larger
wavelengths(above155%).Athighspeedforwavelengthsbelow100%theperformanceofthehullsissimilar
with Hull B performing marginally better. For wave lengths greater than 160% Hull B performs better. A
comparisoncannotbemadeinthewavelengthrange100160%astheprofileisnotsufficientlydefined.
5
HULLA low
speed
HULLA high
speed

aw

3
2
1
0
50

80

110

140
/LWL /%

170

200

230

Figure4.19NondimensionaladdedresistanceprofilesforHullAandHullBatbothtestedspeeds

54

Hyydrodynamic Design Deevelopmentt


4.4.4

Uncerrtaintyanalyssisoftestinggdata
Uncertaiintyanalysis((resistanceesttimates)

ncertaintyanaalysiswasconductedontheecalmwaterresistance,sid
deforceandyyawmomentexperimentall
Anun
resulttsfollowingth
heITTCProced
dureforUnceertaintyAnalysisofResistan
nceTests[ITTTC(2008c)].Theproceduree
makesanestimateefortheeffecctofvariousu
uncertainties intheexperim
mentalproced
dureontestrresultsdueto
o
merror),seeFiigure4.20.
bias(ssystematicerror)andpreciision(random
Theuncertaintyinaresistancetestcanbebro
oadlygrouped
dintofivecom
mponents,nam
mely:

w
includees errors in manufacturing
m
g, deformation
n during testiing (includingg
the modeel geometry, which
theadditionofballast),theeffecto
ofresidualwavesfromprevviousrunsan
ndrunningatttitude.Thesee
dsurfaceareaafromthenominalwetted
dsurfaceareaa
uncertainttiesleadtovaariationoftherealwetted
whichisoneofthemosstimportantp
parameterinrresistanceesttimates;

themodelinstallation, includingaliggnment,weigght,draughtaandtrimverification.Misaalignmenthass
uencesonressistance,trimandsinkage measurementswhileweigghtandbuoyaancyhavethee
directinflu
direct influences on th
he wetted surrface area and displacemeent. Variation
n of weight, displacement,
d
,
differencesin expectedwaaterdensityandtemperatu
ureaswellass
draughtandtrimcanaariseduetod
ballastingerrors;

n, which occu
urs because most
m
instrum
mentation is liinear. The linear regression equationss
Calibration
used during calibration
n leads to an
n element off uncertainty.. There is also a level of uncertaintyy
dwiththemassesuseddurringcalibration[ITTC(2008a)];
associated

Direct meeasurement, which


w
arises due to the uncertainty in the time hisstory of the sampled
s
dataa
whichisaaffectedbylen
ngthoftimessignal,samplingrateandttowspeedmeeasurement. Althoughthee
resistanceetestissteadyy,themeasurredresistance
ewillvarydueetoturbulenttboundarylaayerflow,hulll
wake, testt rig vibration
ns, drift of the measureme
ent system, fluctuating po
ower supply and electronicc
noiseetc.Theaverageresistanceisobtainedbyaaveragingthetimetraceofresistanceovverthecoursee
ofthemeaasuredrun.TThelevelofthisuncertaintyycanbereduccedconsideraablybyrepeattingruns;

Datareduction,arisingfromblockageandtankwaalleffects.

Figure4.20Schem
maticoftestsyystemgroupedintoareaso
ofuncertainty[ITTC(2008c))]

55

Concept Design of a Fast Sail Assisted Feeder Container Ship


TosimplifytheuncertaintyanalysisithasbeenassumedthatsincethemodelwasmanufacturedtowithinITTC
procedures [ITTC (2002)] noaccountof errors from machining tolerances need tobetaken into account. In
addition to this, no allowance has been made for uncertainties arising from extrapolation to full scale,
turbulencesimulation,blockageandwalleffectsorscalingofformfactor.
The process of determining the uncertainty in total resistance coefficient8 involves calculating the maximum
amountofuncertaintyarisingfromsystemerror(biaslimit)andalsomeasurementuncertainty(precisionlimit)
assumingthereisnoerrorcorrelationsbetweenvariables.Asummaryoftheprocessusedtodeterminethe
uncertaintyoftheresistancetestestimatesisgiveninAppendixD.
Plotsofcalmwaterresistanceagainstspeed,sideforceagainstyawangleandyawmoment9againstyawangle
withassociatederrorbarsaregiveninAppendixD.Thecoveragefactor, ,hasbeenchosenas1.645toyielda
90%confidencelevelintheresults.
It was found that there is a high level of uncertainty surrounding the experimental results being typically
between600%atslowspeedsand60%athighspeeds,withtheaveragebeing200%forHullAand250%for
hull B. The uncertainty is dominatedby the random errors associated with theprocessing of the time trace
into values of resistance, with the system bias typically accounting for only a few percent of the total error.
This uncertainty could be reduced significantly by repeating runs more than once. For example, as a
verificationexercisehullAwasrunat19knotsonthreeoccasionsasacalibrationexercise.Table 4.12shows
howtheuncertaintyofeachtestrunonitsownisbetween86and132%,however,whenconsideredtogether
theuncertaintyisonly6.7%.
Table4.12Thereductionofuncertaintyinexperimentalresultsduetorepeats
CT (15degC)
UCT15degC (ResistanceCoefficientCT ,%ofCT15degC)
Run1
0.00708
116.933
Run2
0.00617
85.666
Run3
0.00680
131.873
Combined
0.00668
6.686
Uncertaintyanalysis(waveprobes)
In a similar manner to the calm water resistance estimates bias and precision limits were calculated for the
experimentalmeasurementsofwaveprofileinordertoaccesstolevelofcertaintywiththeaddedresistance
runs. A summary of the process to determine the uncertainty of the wave probe measurements is given in
AppendixD.
Thesignificantwaveheightwascalculatedfromthemeansquareofthewaveelevationandcomparedtothe
valuerequestedfromthewavemaker.TheerrorvaluesaregiveninTable4.13.
Itisevidentthatthereisadeviationoftherequestedwaveheightfromtherequestedvalueatthelowerwave
frequencies.ThisisconsistentwithobservationsofotherresearchersusingtheSSUtowingtankwavemaker
whoputtheerrorsdowntoadifferencebetweentheinputtransferfunctionandthetransferfunctionusedby
thewavemaker.Thedeviationinwaveheighthasbeenincludedasacorrectionintheanalysisoftheadded
resistanceofthetwohulls.

8
Acorrectionisrequiredtocorrect toatemperatureof15Cusing

wherewaterpropertiesatdifferenttemperaturesarecalculatedwith[ITTC(2006)].
9

Yawmomenthasbeennondimensionalisedwithrespectto0.5

56

Hydrodynamic Design Development


Table4.13Errorinwaveheightbetweenrequestedandmeasuredwaves
Wavelength/
%LWL

VS/
knots
15
25
15
25
15
25
15
25

50
100
150
200

HullA
Measuredwaveheight/
cm
2.893
2.870
3.335
3.369
4.013
4.061
4.848
4.773

Error/
cm
1.041
1.078
1.085
1.018
1.200
1.083
1.133
1.159

HullB
Measuredwaveheight/
cm
2.888
3.047
3.567
3.634
4.213
4.294
4.823
4.918

Error/
cm
0.813
0.927
0.977
0.980
0.999
0.999
1.030
1.060

AspectralanalysiswasconductedonthewaveprobetimetracesusingafastFouriertransformtoyieldthe
frequencyspectrum.ThefrequencyspectrumforalloftherunsforthetwohullscanbefoundinFigure D.4,
withthepeakfrequenciessummarisedinFigure4.21.

Wavefrequency/radsec1

Shaft(Input)

Shaft(Measured)

Pod(Input)

Pod(Measured)

8.0
7.0
6.0
5.0
4.0
3.0
50

100
150
Wavelength/%LWL

200

Figure4.21Comparisonofgeneratedwavefrequencytorequestedwavefrequency
The measured wave frequencies are very close to those requested with the maximum deviation from those
requestedbeing4.4%althoughmostaretypicallywithin2%oftherequestedvalue.

57

Concept Design of a Fast Sail Assisted Feeder Container Ship

5. Sail System Design


5.1

Initial design

5.1.1

Conceptsreview

Thebasicconceptforthesailsystemisto designasailplanthatcanprovideasubstantialamountofthrust
reduction at the design speeds of 15 and 25 knots. To design such a system, the requirements need to be
outlinedfirst.Acasedesignstudyonwindassistedcargoships[Smulders(1985)]istakenasaguidetoidentify
thekeyfeaturesthatanefficientwindpropulsionsystemshouldinclude:

a10%thrustreductionat15knotsand3%at25knots;

guaranteedstability;maximumheelingoffivedegrees;

nointerferencewithloadingandunloading;

strengthstandardsmaintained;

adequateprofitability;

highreliabilityandlowmaintenance;

norestrictionsondesignrequirementsoftheshipinotherrespects.
Typicalsailsystems

Based on the design requirements an investigation of existing typical and innovative sail systems was
undertaken.TypicalsailrigsareillustratedalongwiththeiraerodynamicparametersinTable5.1.
Table5.1Typicalaerodynamicparametersofsomebasicrigtypes.[Schenzle(1985)]

TS

Rig Traditional
type
Square
Rig

0.13

0.65

0.9

RS

TG(MS)

BM

SW(MR)

WS(WF)

FR

Modern
Square
(DYNA)

Rigid
Square
Rig

Traditional
Gaff
(MOD.G.)

Bermuda
Mainsl.

Sail Wing
(Mastrol)

WingSail
(TE.FLAP)

Flettner
Rotor
(Surface)

0.1
(0.07)
0.46
(0.43)

0.58
1.5

2.0

0.10
(0.07)
0.35
(0.370.27)
1.1
(1.21.5)
1.6
(23)

0.10

1.5

0.5

MS(DS)

1.5

58

0.05
(0.06)

0.08

0.02

0.280.23

0.360.26

1.01.3

1.7
(1.6)

23

23

0.270.17
(0.650.45)
1.11.4
(1.82.0)
23

0.6
(0.2)
4.05.6
(1.33.2)
7.010.0
(2.23.2)
69
(23)

Sail System Design


Among the rig systems considered, Sail Wing and Wing Sail with trailing edge flap show the highest
maximumliftbuttheformerhaslowerdragatmaximumliftresultingingreaterliftdragratio.Howeverthe
twodimensionalsectiondata,showninTable 5.2togetherwithliftdragratiotrendssummarisethedifferent
characteristics.TheWingSailwithplainflapshowsamaximumliftcoefficientof2.0anddragcoefficientat
maximumliftof0.051whichissignificantlysmallerthanthosevaluesseeninTable5.1.

Table5.2Sailrigaerodynamicperformance,2Dsection
data[Bergeson&Greenwald(1985)]
CD1
1.5
1.5
1.5
1.6
1.0
2.0
10.0

Rigtype
StayedForeandAft
UnstayedCat
RigidSquare
SailWing
WingSail(noflap)
(plainflap)
FlettnerRotor

CCM
0.091
0.084
0.122
0.035
0.037
0.051

CDo
0.092
0.063
0.107
0.075
0.022
0.011
0.113

Figure5.1Polarplotfordifferentsail
rigdesigns

Throughinvestigationtwodesigncandidates,SailWingandWingSailarechosenbasedontheirsuperiordrag
lift ratio performance. The Sail Wing can be improved with the aid of a slat forward of the leading edge
[Fujiwaraetal.(2003)].Sincethesesystemshavemainlybeenusedforgeneralcargoshipswithspeedupto15
knotsitisnecessarytostudymoreinnovativeandefficientsystemsforreasonablereductionathighspeed.
Innovativesailsystems
Inrecentyears,therehasbeenanincreasinginterestinkiteassistedpropulsionsuchastheSkySailssystem.
Thebasicideaofthissystem,asdepictedinFigure E.1,istousealargekitetogeneratethrustwithminimum
heelingforceusingmorefavourablehighaltitudewindconditions.Thecompanyclaimsthatsavingsof10to
35% in annual fuel consumption are possible, and that upwind sailing is possible from 50 degrees true wind
angle.Othermajoradvantagesofthissystemclaimedbythecompanyaretheflexibilityinoperationdueto
thelaunchandrecoverysystemandasimpleinstallationprocess.However,limits ontheoperationofkites
near the shore or when approaching the port of call may arise in the near future, as demonstrated on 15th
September2009,whenanearmissincidentoccurredwithahelicopterinthesouthernNorthSea.Duetothe
highservicespeedofthefeederandresultinglowaverageapparentwindangles(lessthan50degrees),itwas
decidednottoconsiderkitesfurther.
TheNYKSuperEcoShip2030andE/SOrcelleconceptswerealsoexamined.Bothdesignsfeaturetechnologies
intendedtoimprovethepropulsionandcargohandlingefficiency.IntermsofwindpropulsiontheNYKSuper
EcoShipconceptisdesignedfordownwindsailing,asdeclaredbythecompany,andusesarigidwingsystem
withafoldingmechanismandsolarpanels.ItisinterestingthatboththeEcoShipandSkysailsconceptshavea
foldablesailmechanismwhichmaybeapplicabletothefastfeederconcepthoweverthereislackofthesail
systemdescription.TheseconceptdesignsareillustratedinFigureE.2.
Finally,theMultiwingsystemdesignedbyWalkerWingSailSystemLtd.(seeFigure5.2)wasinvestigated.The
designer of this system claims a maximum thrust coefficient above 3.0 whereas that of soft sail is only 2.0
[Walker(1985)];NauticalInnovationServicesundertookaneconomicevaluationofWingSailandMultiwing
SystemoncargoshipsandtheanalysisconcludedthataMultiwingsystemproducestwiceasmuchpowerper
unitsailareathanawingsailatanaveragewindspeedof15knots[Smulders(1985)].

59

Concept Design of a Fast Sail Assisted Feeder Container Ship

Figure5.2Multiwingsystemandgeneralarrangementofcargoship[Walker(1985)]

SincethesaildesignistobeoptimisedforupwindsailingduetothehighoperationalspeedafoldableMulti
wingsystemissuitableforthewindpropulsionsystemasitallowsretractionduringstormyweatherandcargo
handling. This type of system generally has smaller overall dimensions compared to single wing system
thereforepreventingexcessiveheightoftheinstallation.InthisprojectaconventionalNACAsectionforthe
wingswasusedwhenassessingperformanceofdifferentwingconfigurationcases,alsoavoidingpatentright
issueofWalkerWingSail.Thefollowingsectionswilldiscussthesizingofthesailplanandthechoiceofthe
wingsection.

5.1.2

Design
Targetsailarea

Theinitialdesignbeganwithanestimationofthesailarearequiredtogivethedesiredthrustreduction.Atthe
preliminarydesignstageanaveragepropulsiveforcepersailareaisusedbasedondataprovidedbySchenzle
(1985).Sincethedatarangeis10to15knots,linearextrapolationisappliedtoobtaindataupto30knots.
Based on the extrapolated data (Figure E.3), the sail area required for the thrust reduction requirement is
calculated at different speeds using the hull resistance calculation from Section 4.1.5. The results are
summarisedinTable5.3,givingatotalrequiredsailareaof945squaremetresat25knotsshipspeed.
Table5.3Totalsailareaestimation
Required

/kW
/knots
5
189.00
10
1030.00
15
4160.00
20
11300.00
25
23200.00
30
40089.60

Required
/kW
157.44
857.99
3465.28
9412.90
19325.60
33394.64

Hull / Hull /
Thrust
fromwind/ Totalsail
reduction/% kW
area/m2
knots
MW
100
157.44
1181.07
5.25
0.189
27
231.66
914.56
10
1.03
10
346.53
928.28
15
4.16
5
470.65
954.07
20.25
11.3
3
579.77
945.33
24.75
23.2
2
667.89
910.80

60

Sail System Design


Structuralconsiderations
Themaximumbendingmomentisestimatedfromtheassumedmaximumliftcoefficientforatwodimensional
wingsectionwithflap,obtainedfromTable5.1.Thisisthenusedinasimplebeamanalysisofamaststructure
attachedtothedecktodeterminethedimensionsofthemast,deflectionandmaximumstresses,inorderto
identify any relevant constraints. Prior to this analysis the extrapolated driving force was used for the
calculation(AppendixE.2).
Table5.4Mastandmainstockbendingmoment,deflectionandmaximumstressusingaliftcoefficientof2.2
Max.stress/
Max.deflection
%Max / Yield

Windspeed Max.bendingmoment /
/knots
MNm
MNm2
/mm
stress
Mast
30
1.62
8.13
0.073
5.8
70
8.82
44.27
0.395
31.6
30
0.51
30.00
0.090
21.4
Main
stock
70
0.84
163.34
0.500
116.7

Table 5.4 shows the maximum deflection of the mast to be 0.395 metres and maximum stress of 31.6% of
loweryieldstresslimitof5083Aluminiumalloyat70knotwindspeed.Asimilaranalysisiscarriedoutforthe
main wing stock assuming all wind loading is transferred to the stock. Using pinnedpinned beam with
uniformlydistributedload,theestimateddeflectionandmaximumstressatawindspeedof70knotsare2.5
metresand116.7%ofloweryieldstresslimitrespectively.Thisshowsthatthemastprovidesgoodstructural
rigidityandstrengthbutthewingstructurerequiresstockswithhigherstrengthandadditionalstiffeners.The
dimensionscanbeseeninTableE.1.
80

Wingweightvsloading
Wingspan(2030m)

Wingweight/kgm2

70
60
50
40

y = 0.010x + 0.733

30
20
10
0
0

1000

2000

3000

4000

5000

6000

Wingloading/Nm2

7000

Figure5.3Aluminiumaircraftwingloadingagainstwingweightperunitarea
To estimate the weight of the wing, a database of 61 different types of aircraft wing is taken as reference
[Svoboda (1999)]; the wing weight of seven regional aircraft with 20 to 30 metre wing span is used; the
averageofthese,26.44kgm2,resultsinasinglewingweightof4.1tonnes.Theaveragedwingloadingofthese
structuresisequivalenttothewingloadingat90knotswindspeedfortheMultiwingsystem.Howeverasthe
sailwinghasamuchsimplerstructureandadditionalstructureonthetop,thiswingweightcanbereduced.
Forfurtheranalysistotalwingweightof10tonneswasassumed.
Dimensionalconstraintsandsailstowage
Toavoidissueswithcargohandlingandtoimprovethesafetyandreliabilityofthesailsystemconcept,therig
hadtobedesignedtobestoredeitherbelowthedeck,orfoldedondecksothatthemaximumheightisbelow
thelevelofthetopofthecontainerstackswhennotinoperation.Alternativemeansofachievingthiswere
consideredasbelow.

61

Concept Design of a Fast Sail Assisted Feeder Container Ship

Afixedrigidmastwith:
o

the mast fixed at double bottom level and the sails able to retract fully by being lowered
downaroundthemastintothehold;

the mast fixed at deck level and the sails telescoped / folded onto deck when not in
operation;

thesailsmadeofasoftmaterialthatcanbeeasilyloweredandstoredondeck.

Arigidmastwhichisjackablewhere:
o

the jacking system is located at deck level and the mast is lowered into a cavity below the
jackingsystemandthesailsaretelescoped/foldedontothedeckwhennotinoperation;

thejackingsystemislocatedatmiddepthintheholdandthesailandmastareloweredand
storedinthehold.

Themastandsailsarebothcapableoffolding/telescoping.

Jackingsystemsarecommonplaceintheoffshoreindustryandfallintothreecategories,summarisedbelow.

Rackandpiniontheseareusedondrillingrigsandcanhaveaveryhighcapacity(say10,000tonnes
each),arecapableofcontinuousrunning,arerelativelyfast,butareveryexpensive.

Linearjacktheseuseholesinthelegsandlinearjacksthatmoveapininthelegs,andthenthelegis
locked, the pin is removed, the jack retracted and then the process starts again. This is a slow,
cheapersystemusedon'lowergrade'drillingjackupsandmaintenance/production/accommodation
jackupswhichcannotjustifythehighcostofarackandpinionunit.

Wires these use deck mounted winches and wires and sheaves to raise the deck against the legs.
Thisisaquickandcheapsystem,usedoncoastaljackupswithlowloads.

Thesailsystemunderconsiderationisnotasubstantialstructurecomparedtooffshorerigsandthusthemost
realistic system for the hoisting of the sails will be a wire system. A benefit of storing the sails between
container stacks is that during cargo operations, the container cell guides can be employed as a cover to
protectthesailsfromdamageduringcargooperations.Moreconsiderationofthefoldingofthesailsshallbe
given, along with consideration of the necessary ship structure required to support the sail system and the
vessellayout(Section7.2.3andAppendixMrespectively).
Inadditiontothehoistingsystemtherewillhavetobesomeformofhydraulicsinordertooperatethesails
andorientatethemwiththeprevailingwinddirection.Spacewithinthevesselclosetothesaillocationwill
needtobeallocatedforthesemachineryrequirements.
The design of the connection between the sail and the ship should be such to avoid excessive loads on the
deck;anotherissuethatoftenarisesduringthedevelopmentofwindassistedships[Tanneberger(2009)].This
shallbeconsideredlaterinthereportwithafiniteelementanalysisinvestigation(Section0).
Assumingthesailswillbestowedwithinthehold,dimensionalconstraintsforthespanandchordofthewing
sail can be defined. From the preliminary dimensions and deck arrangement, the clearance between the
doublebottomandthetopofthecontainersisestimatedas30metres.Thesumofthechordlengthofthe
wingisthenlimitedtothemouldedbeamof26.7metres.

62

Sail Systtem Design


n
Geometrryofwingsaill
ngsailasitafffectstheloaddistribution
n
Thetaaperratioisaanimportant factortoconsiderwhendesigningawin
asweellastheinduceddrag.In thedevelopm
mentofthisrigg,onlytherectangularplanformwasco
onsideredduee
totheemethodofm
manufactureo
ofthewings(SSection5.2)andtheaddedcostthatpro
oducingtapere
edwingswith
h
flapsw
wouldhavein
ncurred.The aspectratio (AR)isanothe
ergeometricaalproportion whichmustb
beconsidered
d
whendesigningassailplan;thisratioaffectsth
heefficiencyo
ofanywingorrsoftsailandcontrolsthemagnitudeoff
uced drag duee to the lift. A high aspectratio increases the total aerodynamicc
the lifft coefficient and the indu
force coefficient when
w
sailing close
c
to the wind [Marchaj (2003)] wh
hilst it reducees the perforrmance when
n
ngdownwindorwhenthefoilsoperateinthestalledcondition.Th
heaspectratio
oisdefinedas
runnin
AR

span length
.
d length
chord

(5.1)

1.40

LiftCoefficient,CL

1.20

AR1

1.00

AR2

0.80

AR3
AR4

0.60

AR5
AR6

0.40
0.20
0.00
0
0.00

0.05

0.10

0
0.15

Ind
ducedDragCoefficient,CDi
D

0.20

4Variationo
ofInducedDraagCoefficienttwithaspectrratio
Figure5.4

o rigid wing sails


s
had a relatively low aspect
a
ratio of
o between tw
wo and three [Bergeson &
&
Early prototypes of
nwald(1985)],intheattem
mpttomaintaingooddown
nwindcharactteristics.The fastfeederco
oncepthasto
o
Green
mainttain a service speed of 25 knots; thereffore the downwind perforrmance is lesss relevant as the apparentt
windspeedwillbelower.Theadvantagesinselectingahighaspectratiioareclear;however,anin
nitialestimatee
d dimensions of the supporting structu
ure had to be addressed to identify any
a structurall
of thee weight and
constraints.Theraatiowassetto
ofour,enough
htoensurelowdragcharaccteristicsontherelevantco
ourses.
d
constraints, co
ombined with
h the choice of
o aspect ratiio and target sail area deffine the wingg
The dimensional
dimen
nsionsasshow
wnbelowinTTable5.5.
Table5.5Multiwingssystemdimen
nsions
2
No.ofsyystem
No.ofwingspersysteem
3
Height/m
26.5
Wingspaan/ m
25
Chord/m
m
6.25
AspectRatio
4
TaperRaatio
1
Totalsaillarea/m2
937.5
Width/m
m
13.86
Extremewidth/m
21
Wingweeight/ tonnes
10
Linkagesstructureweigght/tonnes 2

Figu
ure5.5Multtiwingsystem
m

63

Concept Design of a Fast Sail Assisted Feeder Container Ship

WingSectionDesign
The commercial code XFoil is used to create a database of NACA four digits section at various Reynolds
numberswhichcovertherangeoffullscaleapparentwindspeeds.Initiallynoflapsareincludedandthepolar
plotsareexaminedbylookingatthe:

maximumliftcoefficientforbeamreachperformance;

bestlifttodragratioat(

liftcoefficientafterstallaccountingforincidencechangesalongthespanduetorollingoftheship;

thicknesstochordratioforstructuralintegrityandmastallowance.

)forwindwardsailing;

NACA0014,0015and0016sectionsareselectedascandidates.Fortheadditionofthetrailingedgeflap,hinge
points of 70%, 75% and 80% of the chord length were compared on the same basis as above. The plots
showedthatthemaximum / ratiowouldbeachievedbythe0015sectionwiththeflapdeflectedat22.5
degrees,whilstthemaximumliftforbeamwindsisachievedwithadeflectionof45degrees.
x/c0.7flap22.5
x/c0.8flap22.5
x/c0.75flap45

x/c0.75flap22.5
x/c0.7flap45
x/c0.8flap45

x/c0.8flap22.5

x/c0.8flap45

2.20

0.20
0.18

2.00

0.16
0.14

1.80

0.10

CL

CD

0.12
1.60

0.08
1.40

0.06
0.04

1.20

0.02
0.00

1.00
1

CL

0.1

0.2

0.3

0.4

CD

Figure5.6Comparisonofdragcoefficientagainstliftcoefficientsquaredfordifferentflapchordpositions
(left)and;NACA0015polarplotforflapangles22.5and45degrees(right)
WindData
StatisticaldataforwindstrengthwasobtainedfromNOAASatelliteandInformationService(2009).Thesea
winddataisprocessedforanelevationoftenmetresabovetheseasurfacebyblendingdatafromuptosix
satellites to reduce the subsampling aliases and random errors. The data is complete from 1987 to the
presentdaywithatimeresolutionvaryingfromsixhourly,dailytomonthly,withdataona0.25degreegrid
whichgives1,036,800dataitemspertimesnapshot.

64

Sail Systtem Design


n

Figgure5.7Seasurfacewindfor2002:magnitudeasscaalarmean[NO
OAASatelliteandInformatiionService
(2009))]
Monthly mean values of easteerly and northerly compon
nents of speeed for the yeear 2008 were taken and
d
ddirection.Scatterdiagram
msofaverageespeedanddirectionweree
proceessedtoobtainabsolutewiindspeedand
then produced forr worldwide and
a region sp
pecific wind statistics
s
on a month by month basis and a yearlyy
bbean showin
ng a clear pre
evailing wind
d
cumulative total. A summary of yearly wind strength for the Carib
plotsshowinggthevariation
nofwindstrengthwithdire
ectionforthee
directtionisshown inFigure 5.8.Equivalentp
wholeeworld,South
hEastAsiaandtheNorthA
AtlanticaregivveninAppend
dixE.

8000
Numberof 6000
Observatio
ons 4000
2000
01

60008000
4

40006000

20004000

10

02000

13
WindSpeeed/ms1

16
19

270 315
22

135 180
25

0 45

WindHeaading
/degreees

Figure5.8Windbystreengthanddire
ectionfortheCaribbean(an
nnual)
Theyyearlymeanw
windspeedan
ndstandardd
deviationforttheregionsbeeingconsideredandthew
wholeworldiss
giveninTable5.6.
Table5.6
6Meanwind
dspeedbyreggion
Reg
gion
Caribbeaan
SouthEaastAsia
NorthAttlantic10
WholeW
World

Meeanwindspeeed/m/s
5.962
5.802
8.827
6.108

Sta
andarddeviatiion
4.652
4.647
6.572
4.8525


10
Bassedonseaareeas23,24,25and16only[NOAASatellitteandInformaationService(2009)].

65

Concept Design of a Fast Sail Assisted Feeder Container Ship


5.1.3

Theoreticalperformance

Itisnecessarytomakeanevaluationofthesailingperformanceatanearlystageoftheprojectasitenables
thedesignertoquantifythedecreaseinbrakepowerthatcanbeachievedtomaintainthefeederataconstant
speed.AnotherimportantaspectthatneedstobeaddressedistheeffectthattheMultiwingsystemhason
thedynamicsofthevesselwhilstatsea.APerformancePredictionProgram(PPP)wascreatedinFORTRAN90;
theunderlyingprincipleatthebaseofthecalculationsisthatundersteadysailingconditions,thehorizontal
andverticalforcesandmomentsactingonhullandsailsarebalanced.Thecodeusestheaerodynamicdata
obtainedfromXFoiltocalculatetheaerodynamicforcesgeneratedbythesails.Theangleofheelcanbealso
estimated using the initial hydrostatic data, whilst heave and pitch are neglected in this analysis as less
significant. The thrust reduction is computed combining the hydrodynamic drag data from Section 4.1. In
Section6.1.1,theprogramwillbemodifiedtoassessthevesselperformanceusingtheexperimentalresults.
AerodynamicIndex
The code consist of three main loops for ship speed , true wind speed and true wind direction ; the
apparent wind speed and the apparent wind angle are calculated as shown in Appendix E. The lift
2.1 and the base drag coefficient is taken from the XFoil database as
coefficient is taken as,
toobtain
0.019.Theaerodynamiccoefficientsarethentransformedintotheshipaxiscoefficients
theforwardthrustandtheheelingforce.Atthisstageoftheprojectonlyempiricalrelationshipsareavailable
toestimatetheleewayangle;thesearebasedonhydrodynamicderivativesofoldhullformswithfixedrudders
andhencecannotbeappliedtothefastfeeder.Theformulausedtocalculatetheleewayangleistakenfrom
Schenzle(1985)andisshowninAppendixE3.InSection6.1.1,towingtankresultswillbeusedtocomputethe
leewayangletoahigherdegreeofaccuracy.
ThrustReduction
The thrust generated by the sails in the direction of motion can be regarded as an effective reduction in
resistance.Sincethespeedoftheshipandthewindspeedareknownateverytime,itispossibletocalculate
the reduction in brake power requirement, given that the total hydrodynamic resistance of the hull is also
known.Table 5.7presentsthetypicalthrustreductions,heelandleewayforarangeoftruewindangles.The
thrustreductioniscalculatedas,
.

Table5.7Theoreticalthrustbenefitprediction(VS=15knots,VT=16.5knots)
/deg
20
40
60
80
100
120
140
160
180

/knots
17
16.2
14.9
13.2
11.2
8.7
6.1
3.5
1.8

/deg
11
22
33
45
57
70
86
110
180

/deg
2.50
2.26
1.85
1.34
0.83
0.41
0.14
0.01
0
average:

66

/deg
0.008
0.007
0.006
0.004
0.003
0.001
0
0
0

Thrustreduction /%
HullA
HullB
0.0
0.0
7.5
7.3
16.4
16.0
20.3
19.8
19.2
18.7
14.3
13.9
7.8
7.6
2.4
2.3
0.1
0.1
9.8
9.5

(5.2)

Sail System Design


Theestimateofthethrustreductionisdoneassuminganaveragewindspeedof16.5knots,asthewinddata
presentedinSection5.1.2wasnotyetavailable.Thereductionsarethenaveragedovertherangeoftruewind
angles asshowninTable 5.8.Onelimitationtothisassessmentisthelackofanydragdataforthemastand
thesupportingstructure;alsotheshipresistanceusedinthecalculationdoesnotincludeadditionalresistance
duetoair,waves,sideforceandheel.Ontheotherhandthesailingpointiskeptfixedforallheadings,whilsta
gaininthrustwouldbeobtainedbyoptimisingthesailincidenceforeachcourse(seeSection6.1.1).
Table5.8Initialthrustreductionprediction
Shipspeed/knots
10
15
20
25

Thrustreduction/%
27
10
5
3

5.2

Wind tunnel testing

5.2.1

Introduction

The preliminary estimate of the performance of this Multiwing system relies on the accuracy of the
aerodynamic coefficients. Testing a model of the rig in the wind tunnel enables the verification of the
predictionsandtoassessthepotentialofthistypeofrigbyoptimisingthespacingbetweenthewingsandthe
staggerangle.Thedesignersbelievethatanincreaseinliftcoefficientcanbeachievedthankstothesloteffect
whichshouldincreasetheefficiencyoftheupwindwings[Chret(2000)].Testingthemodelalsoallowsthe
assessmentofthemagnitudeofthewindageforthisparticularstructurehenceincreasingthefidelityofthe
finalprediction.Arecentstudyonthefeasibilityofasailassistedbulkcarrier[Fujiwaraetal.(2005)],suggested
thatsailsailandsailhullinteractionsneedtobeinvestigatedwhendesigningwindassistedships.Theaimof
thisistoassessthechangeinperformanceduetothesurroundingenvironment,thusimprovingthevalidityof
theresults.Duetothedimensionalconstraints,outlinedinSection5.2.2,therighasbeentestedinisolation
and successively in the presence of containers; the aim of this is to investigate the potential change in
performancecausedbytheadditionofabluffbodyupstream,andareflectionplaneclosetothebaseofthe
sails.Acomputationalfluiddynamics(CFD)studywillprovidetherequiredsailsailinteractions,assessingthe
sheddingeffectbetweenthetworigswhensailingtowindwardandoptimisingtheangleofattack(AoA)ofthe
individualwings.(seeSection5.3)

5.2.2

Modeldesign

A scaled model of the Multiwing system is to be designed and tested in the wind tunnel. A number of
considerations need to be made. Firstly, the model must be geometrically identical. Since the geometry is
simpleitisnotanissue.Secondlytodeterminethescaletheconstraintsderivedfromthewindtunnelsizeand
thedynamometercapacityneedtobeinvestigated.Theinitialideawastodesigntworigsandinvestigatethe
interactionbetweenthem,butduetoinsufficientbudget,complexityoftestingandtimeconstraintsasingle
rig of larger scale has been designed. The model scale is determined based on the following constraints
[Claughton&Campbell(1994)]:

thewindtunnelsizelowspeedsection:4.6metreswideby3.7metreshighby3.7metreslong;

amaximummodelheightoftwometres;

amaximumtestwindspeedofeightmetrespersecond;

individualdynamometerlimitof1000N;

67

Concept Design of a Fast Sail Assisted Feeder Container Ship

theneedtoadjusttheflapsmanuallymaximumheightof2.3metres;

modelstrength;

modelweight.

The rationale behind the choice of the model scale is to maximise the dimensions of the model under the
constraintsoutlinedabove,sothatthetestsareperformedatahigherReynoldsnumber,thusincreasingthe
qualityoftheresults.Thescalewasfinalsetto1:15.Tominimisetheweightofthewingsectionsandflaps,low
densityfoamisusedforthemanufacture,whilstthereststructureismadefromaluminium.Whendesigning
themast,theheightwasmaximisedinordertohaveenoughclearancetofitthecontainersbelowthesails.To
avoidexcessiveheelingforceduetothehighercentreofeffort,themastwaslimitedto0.4metres.Theweight
of the system is calculated from the volume and density of the components. This weight produces
approximatelyequallydistributedverticalforceof76.62Noneachdynamometerblock.
Table5.9Windtunnelmodelweightestimation
Item
Weight/ kg Number Totalweight/ kg
Wing
0.766
3
2.30
StockA(main)
1.283
3
3.85
StockB(flap)
0.499
3
1.50
Wingplate
0.245
6
1.47
Topbar
1.285
1
1.28
Bottombar
2.677
1
2.68
Mast
3.193
1
3.19
Base
3.513
1
3.51
Mastfittingring
0.373
1
0.37
Transversefittingrod
0.936
1
0.94
Longitudinalfittingrod
1.338
1
1.34
Mastkey
0.500
1
0.50
Bolts,screws
0.500
1
0.50
Overalltotalweight
23.43

Usingthemaximumliftcoefficientof2.10obtainedfromXFoil,predictionsoftheresultantmaximumvertical
forceonthesidedynamometersduetomomentproducedbyliftiscalculated.ThisisshowninAppendix F.1,
wherethemaximumresultantforceonanindividualdynamometerremainsunderthemaximumallowancefor
windspeedsuptoninemetrespersecond.
Finally,thestrengthofthemodelisevaluated.Sincethegeometryofthecomponentsissimple,Eulerbeam
theorycanbeapplied(see AppendixF.1).ForStockAandB(Appendix F.2)themaximumdeflectionsare4.2
mm and 40.0 mm respectively, while the mast deflection is negligible. It is noticed that Stock B has large
maximumdeflection.Howeveronceassembled,tensionwillbeactingonthestocksotheadditionalsupport
structure at midspan is considered sufficient to restrain this deflection. The maximum stresses are
considerablylowerthantheyieldstressandsoplasticdeformationisnotexpected.Theresultsaresummarised
in Appendix F.1. For the model manufacturing drawings BSI (2007) are used (F.2). The overall view of the
assembledmodelisillustratedinFigure5.9.

68

Sail System Design

Figure5.9Windtunnelmodelassembledview

5.2.3

Modelmanufacture
Wings

Thefirstcomponentstobemanufacturedwerethewings.ItwasdecidedtouseStyrofoam,lowdensityfoam
usedforinsulationandreadilyavailablefromlocalsuppliers.Thischoicewasdictatedbythelowcostofthe
rawmaterial,thesimplicityofthemanufactureusingtheEDMChotwirecutter(FigureF.1)andtherestrained
weight. DXF drawings of the main section and the flaps were imported to the numerical cutter; due to the
limited length of the wire, the sections had to be cut in three parts and then joined together with epoxy
bonding.Asamplesectionwascreatedtoverifythecutterstoleranceforcuttingtheinternalfeaturesofthe
section.Toachieveagoodsurfacefinish,anessentialfeaturetoensurelowskinfrictiondragcharacteristics,a
highbuildprimerandnonvinilicpaintwereappliedonthetrialwing,butbothhadtheeffectofcorrodingthe
surface.Thealternativesolutionofcoveringtheentiresurfacewithathinplasticcoverwasdiscardeddueto
thepresenceofthetrailingedgeflapandthecorrespondinggapwhichwouldexistalsoatfullscale.Wetand
drypaperwaseventuallyusedtosmooththesurface.Tocutthecircularslotforthestocksthehotwirecutter
cut through the leading edge along the chord until the pivot point at 20% of the chord. Due to the high
sensitivityofthismaterial,inertbodyfillerwasusedtofillthegap,thenwetanddrypaperwasusedtoremove
theexcessfiller.
Structure
ThesupportingstructureofthewingswasmanufacturedintheEDMCusingCNCmillingmachines(Figure F.1)
whichoperatefromafileinputconvertedfromCADsoftware.Thealuminiumbasestructure,supportingthe
mastandusedtoconnectthedynamometersbar,wasweldedbytheEDMCusingtheTungstenInertGas(TIG)
technique (Figure F.2). The mast and transverse bars were assembled together using standard fasteners. A
viewofthefinalassembledmodelisshowninFigure5.10.

69

Concept Design of a Fast Sail Assisted Feeder Container Ship

Figure5.10Modelassembledinthewindtunnel

FlowStimulation
Themaximumspeedachievableinthelowspeedsectionofthewindisthiseightmetrespersecond,implyinga
test Reynolds number of 230,000 which is too low to stimulate the transition of the flow from laminar to
turbulent at the same location that would occur operating at the full scale Reynolds number. The scope of
inducingflowtransitionaffectstheaccuracyofthedata,inparticularthemagnitudeofliftanddrag.Thereare
differentmethodstotriptheflowthatareusedincommonpractice,theonechosenhereforitssimplicityand
repeatability,isthecarbondurumgritstrip.XFoilisusedtoidentifythenaturaltransitionpointoftheNACA
0015atthefullscaleReynoldsnumber.AtanAoAoffivedegreesthetransitionpointisfoundtobeat11%of
thechorddecreasingto2%at15degrees.Thelocationofthegritstripisthereforefixedat10%ofthechord
fromtheleadingedgeandthewidthatfivemillimetres,assuggestedinBarlowetal.(1999).Themethodused
heretocalculateofthesizeofthegritisgivenbyBraslow&Harris(1966):
12

(5.3)

istheReynoldsnumberperfootbasedonthefreestreamvelocity; isaconstantbasedonthegrit

where

roughness.ThisisalsoafunctionofthelaminarReynoldsnumberbasedonthefreestreamvelocityandthe
distancefromthetripstriptotheleadingedge.Theheight isderivedas0.0075incheswhichconvertstogrit
number80.
FlowVisualisation
Toverifytheefficiencyofthegritanattemptwasmadetouseparaffin;intheturbulentflow,theoilapplied
afterthestripwouldhaveevaporatedquickerthanbeforethestrip.Howeverduetotheporosityofthefoam
thismethodcouldnotbeappliedastheoilsoakedintoodeeplytoevaporateinareasonabletime.Tovisualize
the flow wool tufts were used instead; a smoke machine was prepared to visualize the flow characteristic
beforeandaftertheadditionofthecontainers,butonthedayofthetesting,itwasfoundunusableduetoa
problemwiththeoilpump.

70

Sail System Design


5.2.4

Experimentalsetup
Measurementsystem

The measurement system consists of the control room equipment (Figure 5.11) and the dynamometer
mountedinthewindtunnel(Figure5.12).

Amplifier(1);
Filter(2);
Analoguetodigitalcard(3);
Acquisitionsoftware(LASSO)(4);
Roofcameraandremotecontrol
system(5);
Laptopfortemperaturepressure
acquisition;
Sidevideocamera.

Figure5.11Experimentalsetupinwindtunnelcontrolroom

SixComponentsDynamometer:
Channel1:Driveforcestarboard;
Channel2:Heelingforceport;
Channel3:Heelingforceforward;
Channel4:Verticalforceforward;
Channel5:Verticalforcestarboard;
Channel6:Verticalforceport.

Figure5.12Sixcomponentdynamometerinthelowspeedwindtunnel[Campbell(2009)]
Calibration
TheacquisitionsoftwareusedforthisexperimentistheLASSOsuiteprovidedbyWUMTIA.Thisallowsuser
definedchannelstobecreatedinwhichthesignalsareprocessedtogivethedesiredoutput;thisalsoallows
corrections to be made for the interactions between forces and moments in the dynamometer. Extensive
calibrationwasnecessarybeforetheexperimentbegan;thiswasdonetoavoidthenecessityofrepeatingthe
calibrationprocedureinthesuccessivedaysoftesting,hencesavingtime.Theprocedurecanbesummarized
asfollows:
1.

Mountcalibrationrigandpositionpulleyatthedriveforcelink;

2.

Takezeroreading;

3.

Appendonekilogramandacquiredata;

4.

Controlmeasuredvalue;

5.

Repeatsteps3and4uptoatotaloffivekilos;

6.

Remove1kilo,acquireagainandcomparewithpreviousmeasurement;

7.

Repeatstep6fortheremainingweights.

71

Concept Design of a Fast Sail Assisted Feeder Container Ship


The rig is then positioned at the heeling force starboard link and the procedure is repeated. The first
calibrationshowedanerrorof1.9%forthedriveforceand0.5%fortheheelingforce.Theexpressionsforthe
userchannelwerehencemodifiedtocompensatefortheerror.Intheseconditerationtheerrorwasreduced
to0.5%and0.1%respectively.
TestMatrixandProcedure
Theoriginalscheduleforthetestingwasmodifiedduetoadelayinthemanufacturingoftherig;thefirstday
oftestingwaspostponedtothe17thofFebruaryandtheexperimentcouldresumeonlyonthe23rdofthesame
month.Thisgaveplentyoftimetoimprovethetestmatrixwhichwasupdatedattheendofthefirstday.
Table 5.10showsthefinaltestmatrix.Excludingthecalibration,thetotalnumberofusefulrunsis116;the
remaining59runsarerepetitionsforzeroingerrorsandproblemswiththesignaloftwochannels.

Table5.10Windtunneltestmatrix.Rigtestedatarangeofanglesofattackuptostall
Day
1
1
2
2
2
2
2
3
3
3
3
3
3
3
3

Rundescription
calibration
speedcalibration
windage
Wingspacing100%chordlength
Wingspacing70%chordlength
Wingspacing50%chordlength
Wingspacing110%chordlength
Wingspacing110%chordlength
Wingspacing110%chordlength
Wingspacing110%chordlength
110%30degreesstagger
110%60degreesstagger
flatsailconfiguration
windage
singlefoil
Totalruns

No.ofruns
66
21
5
21
15
19
15
15
13
10
11
10
11
6
3
241

Speed/ms1

4,6,8
8
8
8
8
8
8
8
8
8
8
8
8
8

Interaction

no
no
no
no
no
no
yes
yes
yes
yes
yes
yes
yes
no

Windtunnellayoutforsailhullinteractions
Thedimensionalconstraintsderivedbythelargescalefactorofthemodel,allowedcoveringonlyapartofthe
testing that would be required to achieve a complete sailhull interaction experiment. It was decided to
replicate the stack of containers using large cardboard boxes. Due to the mast height having to be limited
becauseoftheexcessiveheelingmomentthatwouldhavegenerated,onlythelastthreerowsofcontainers
couldbesimulated.Thegapbetweenthetopoftheboxesandthebaseofthesailsreplicatesthefullscale
clearanceof2metres;thisvalueisbasedontheinitialestimatesofthefullscalerigdeflection.Theheadings
thatweresimulatedinthewindtunnelarebasedontheaverageapparentwindanglesseenbythefastfeeder
sailsatanadvancespeedof15and25knots,coveringmainlyreachingandbroadreachingcourses.FigureF.3
andFigureF.4showthelayoutwhensimulatingcontainers.

5.2.5

Windtunnelpostprocessing

Throughout the duration of the experiment it was possible to cover all the relevant configurations and
additionalrunsthatwerescheduled,despitethetimelostincalibratingtheequipmentandrepeatingsomeof
theruns.ThecorrectionoftherawdatatoinvestigatethefreeairperformanceofthisMultiwingsystemis

72

Sail System Design


now addressed. The drive and heeling force of the supporting structure are subtracted from the measured
dataleavingtheforcesduetothewingsonly.

Figure5.13Containerslayoutanddimensionsforsailhullinteractiontests
Thisenablesthedatatobenondimensionalisedandtoobtaintheliftanddragcoefficientsofthewingswhich
arethenusedforcomparingthedifferentconfigurations.Thewingdatamustthenbecorrectedtoaccountfor
theclosedenvironmentofthewindtunnelandscaleeffectsasbothaffecttheforcesmeasured.
Dynamometermisalignment
Figure 5.14(left)showsanaverageerrorofbetween3.5%and18%betweenportandstarboardtacksforthe
liftanddragrespectively.ThedifferenceismarginalatmoderateAoAsandincreasestowardsthestallregion.
Thiskindofuncertaintycanhavemultiplecauses;suchasasmallmisalignmentofthedynamometerorsurface
defectsduetothepresenceofthewooltuftsonlybeingononesideofthewings.
Cdstbd
Clstbd
Clcorrected

Cdport
Clport
cdcorrected

0.021

1.20

0.020
CDO min

CL;CD

1.00
0.80
0.60

0.019
0.018
0.017

0.40

0.016

0.20

0.015

0.00
5

7.5
10
AoA/degrees

12.5

15

0.0000

0.0025

0.0050
0.0075
h/inches

0.0100

Figure5.14Dynamometermisalignmentfromportandstarboardtacks(left);andvariationofbasedragwith
gritsize(right)[Braslow&Harris(1966)]
Gritdrag
The addition of the grit strip will increase the base drag of the wing sails; at zero AoA, the drag of lifting
surfacesisdominatedbytheskinfriction,hencetheincreaseindragwillbeproportionaltotheroughnessor
thegritsize.Thisincreasecanbecorrectedusingpublisheddataonthevariationofbasedragcoefficientwith
variablegritsize[Braslow&Harris(1966)].ThecurveinFigure 5.14(right)showsannetincreaseindragata
gritsizeof0.003inches;thisdemonstratesthatnoturbulentboundarylayerhasdevelopedupstreamofthe

73

Concept Design of a Fast Sail Assisted Feeder Container Ship


grit.Afterthispointtheriseindragislowerandconstant.Thegritsizeusedinthisexperimentis0.007inches
andthedifferenceindragisobtainedextrapolatingthecurvebacktozerogritsize,asshownbythedashed
line.Thecorrectionissimplythedifferencefromthedragat0.0075inchesandtheextrapolatedzerogritsize
0.01998 0.0194 0.0006.Thepresenceofthewooltuftswouldalsocontributetoincreasing
dragCD
theroughnessoftheliftingsurface,andthereforeacorrectionshouldbeapplied;howevernoprocedurewas
foundintheliteraturetoaccountforthisaddedroughness.The1.5millimetrethicktuftswereusedinasparse
mannerandonlyonthewingvisiblefromthecontrolroomofthewindtunnelandfromthecameras;therefore
theeffectonthewingsailperformanceshouldbelimited.
Windage
To calculate the aerodynamic coefficients of the wing in isolation, the drive force and heeling force of the
supportingstructurealoneismeasuredat0,5,10and15degreestothewinddirection.Thesevaluesarethen
subtractedfromthetotalforcesmeasuredwiththewings;withintherangecovered.Thedatashowsalinear
trend,thereforelinearinterpolationisusedatintermediateangles.

sin

(5.4)

and

cos

(5.5)
5
Here A is the full scale sail area and is the dynamic head in millimeters of water. The pitt tube which
measuresthedynamicheadislocatedinthehighspeedsectionofthewindtunnel,wherethecontractionin
as the head is proportional to . The
cross sectional area is five, therefore must be corrected by
windagedatacanalsobetransformedintothewindaxessystemandnondimensionalisedwithrespecttothe
sail area. Thus the windage can be added to the forces due to the sails when assessing the different
configurationswiththePerformancePredictionProgram(PPP)(seeSection6.1.1).

Onlythewindagedatafromtheinteractionslayoutispresentedinthissection.Figure F.6showsthewindage
datawithoutthecontainers.Asonlyafewanglesweretested,thewindagecoefficientshadtobeextrapolated
asshownbythedottedlineinFigure 5.15(left).Apeakat90degreesapparentwindanglecanbeobserved;
this is readily explained if we observe Figure F.4. As the wind moves aft and forward of the beam the
containersgraduallyshedthecontributionofthebaseandmastuntilonlythetransversebarsandthetopend
ofthemastcontributetotheforcesmeasured.Whilstalotofworkhasbeencarriedouttoassessthescale
effects of lifting surfaces, and standard procedures have been outlined, very little is known on scaling
proceduresofwindagedata.Ifmoretimehadbeenavailable,themastshouldhavebeenremovedtomeasure
theforcesduetothebasealone;thiswouldhaveallowedusingthecorrectCLandCDforthemastinthePPP.It
isimportanttonoticethatthemodelscalestructurewillbedifferentfromthefullscale,asthemastwouldbe
larger,thereforetheadditionofthebasemaycompensateforthisdifference.
Downwashcorrection
Thecomponentofinducedflowintheliftdirectionatthemodelischangedduetothefinitedistancefromthe
wallsofthewindtunnel[Barlowetal(1999)].Closedsectiontunnels,likethe15x12lowspeedsectionat
theUniversityofSouthamptontendtoincreasetheincidenceofthewingsandthemeasureddrag.Appendix
F.5showstheequationsusedtocalculatethiscorrection.
Solidblockagecorrection
Solidbodyblockageisafunctionofthevolumeofthebodyandthecrosssectionalarea.Continuityofmass
flowthroughthechannelrequiresincreasesintheflowvelocitiesnearthebodycomparedwiththevelocities

74

Sail System Design


inunconfinedflow;thesurfacestressesarehencehigherandthedataneedstobecorrectedtoaccountfor
this.
Cx

Cdwindage

Cy

0.0450

0.045
0.040

0.0350

0.035
Cx ,Cy

CL ,CD

0.030
0.025
0.020

0.0250

0.0150

0.015
0.010

0.0050

0.005
0.0050 60

0.000
60

80
/deg

100

120

80

100

120

/deg

Figure5.15Windagedata:wind(left)andship(right)axescoefficientsofthesupportingstructure.Boundary
correctionsareapplied
The Herriot formula is used to estimate the magnitude of the correction [Barlow et al. (1999)]. Table 5.11
summarizestheresults;notethatwindtunnelcrosssectionalarea ,istakenasthegeometricareaminusthe
areaoftheboundarylayeratameanthicknessof0.1metres.
Table5.11Windtunnelwakeblockagecorrection
Item
Wings
Mast
Boxes
Blockageratio%
2.2
0.5
4.0
Wingsvolume/m3
0.0839
0.003
0.87
Windtunnelarea
16.1

Boundarylayerthickness/m
0.1
Boundarylayerthicknessarea/m2
1.48
Jetarea/m2
14.62
K1
1.034
0.962
2.42
t1
0.794
0.791
0.831
Solidblockagecorrections
0.00123 0.00004 0.0313

AppendixF.5showsthecalculationof

1 and 1 whicharecorrectionfactorswhichrelatethemodelgeometry

tothewindtunnelarea(Figure F.7); thecorrectionfortherunswherenocontainersareusedisverysmall,


howeverthevolumeoftheboxeshasalargereffectdespitethedecreaseinfrontalareaduetotheboundary
layer thickness; the volume of the base also does not contribute to the blockage correction as it remains
embeddedintheboundarylayer.Theblockagecorrectionforthewindagedataiscalculatedusingasimplified
formulaforunusualshapesaccordingtoBarlowetal.(1999),(seeAppendixF.5).
Wakeblockagecorrection
Therationalebehindthiscorrectionissimilartothesolidblockage,thewallsofthesectionconstrainthewake
behind the wings thus causing an increase in drag; this correction is based on the drag measured after
and increases proportionally to the wake size [Campbell (2009)].
is the function of the
separation,
model body shape used to calculate this correction, Appendix F.5 shows its derivation which follows the
procedureoutlinedin[ESDU(1980)].

75

Concept Design of a Fast Sail Assisted Feeder Container Ship


Speedcalibrationruns
Thechoiceofthescalehasplayedafundamentalroleinobtainingsensibleresults.ThetestReynoldsnumber
alwaysdictatestheaccuracyoftheliftanddragdatawhentestingliftingsurfaces;toinvestigatetheeffectsof
varyingReynoldsnumber(Figure 5.16)andtoverifythatthetestReynoldsnumberishighenoughtoobtain
gooddata,itwasdecidedtotesttherigatfour,sixandeightmetrespersecond.

0.80

0.40

Re
235,00
0

0.70
0.60

0.25

CD

0.40

Re
117,00
0

0.30
0.20

Re
176,000

0.20
0.15
0.10

0.10

0.05

0.00

0.00

Re
235,000

0.30

Re
176,00
0

0.50
CL

0.35

5
10
AoA/degrees

Re
117,000

15

10

AoA/degrees

15

20

Figure5.16Lift(left)anddrag(right)curveswithvaryingReynoldsnumber;wingspacing100%ofthechord
zerostagger,zeroflapdeflection
117,000thedragattendegreesislowerthanthebasedragatzerodegrees.At
176,000the
At
same behavior is observed at an angle of attack of ten degrees. Barlow et al. (1999) suggests that laminar
separationwithperiodicalvortexsheddingisthereasonofthisanomaly,whichhasbeenencounteredalsoin
thetestingofotherairfoilsasreportedinMarchmann&Werme(1994)(seeAppendix F.6).Flowvisualization
overthedownwindsurfaceofthewingshasenabledtheauthortoobservethebehaviourofthetuftsatthe
differentspeeds.Figure F.9andFigure F.10showsthatatthelowerspeedsamildseparationoccursat30%
of the chord, whilst at 8 ms1 the flow separates at approximately 80% of the chord. Table 5.12 shows the
magnitude of the separated drag used in the wake correction. At
235,000 this component is only
presentat15degreesAoA,wherethefoilstalls.Atthelowerspeeds appearsfrommoderateAoAs.The
liftcoefficientcurveincreasesproportionallywithReynoldsnumber;asthewindspeedincreasesthegradient
oftheliftcoefficientcurverises,butthemaximumliftcoefficientisstillsmallerascomparedtothepublished
data on the 0015 section [Jacobs & Sherman (1937)] and to XFoil. This leads to the conclusion that scale
effectspartiallylimitedtheperformanceoftheWingsail.Havingoutlinedtheproceduretoevaluatethefree
airperformanceofthesail,anattemptwillbemadetoextrapolatethedatatohigherReynoldsnumberswhich
correspondstoaveragewindspeedsfoundinservice.
Table5.12Separateddrag,CDs

/degrees
0
5
10
13
14
15

Reynoldsnumber
117,000 176,000 235,000
0.0
0.0
0.0
0.043
0.087
0.0
0.041
0.102
0.0
0.043
0.087
0.0
0.009
0.092
0.0
0.082
0.081
0.048

76

Sail System Design


ExtrapolationtofullscaleReynoldsnumber
The method commonly used is to extrapolate wind tunnel results to full scale is based on the effective
rather than the test Reynolds number; this accounts for the dynamic pressure
Reynolds number
fluctuationsacrossthetestsection.

iscalculatedbymultiplyingthetest

bytheturbulencefactor

As suggested by Campbell (2010a) the addition of a grid increases the turbulence in the tunnel; Sinmons &
Salter(1934)measuredtheturbulencefactordownstreamofagridusingahotwireanemometer.Themean
38
from the grid) was found to be 5.92. It follows that
value in the free stream region (
1.39 10 using

235 10 .Researchconductedsofarsuggeststhattherigwillperformbetterat
higherReynoldsnumbers.Barlowetal.(1999)givesanoverviewofthemethodsusedbyaerodynamiciststo
obtainfullscalecharacteristicsofwings.Thesecanbeappliedonlywhenextensivepublisheddataisavailable
for the required section. A large number of tests were conducted at NACA for common airfoil sections and
hence an attempt can be made to correlate lift and drag data to higher Reynolds numbers. The average
apparent wind speeds for the selected routes are calculated using the PPP (see Section 6.1.1) the reference
Reynolds number is 6.1 10 . A turbulence factor of 1.7 is assumed [Choi (2004)] relating to an effective
10.3 10 . The plots shown in Appendix F.8 are taken from Jacobs & Sherman
Reynolds number
8.3

(1937) and summarise the NACA 0015 performance at Reynolds numbers up to

10 , close

enough to the target value. At full scale, the maximum lift coefficient is increased by 0.4 whilst the base
drag decreasesby0.0015.Figure5.17(a)showsthepolarplotfor
1.39 10 ;ahorizontallinecan
bedrawnthroughtheestimatedfullscalemaximumliftcoefficient
,thelinearpartofthetestliftcurveis
.The
extendedwiththesameslopeandthecurvedpartofthepolaristhenraiseduptocoincidewith
dragcurveismodifiedbyplotting versus [Figure 5.17(b)]theinduceddragisthensubtractedtoleave
; this is then increased to
in the same way as the lift curve. The base drag is then
the base drag
.
decreasedbyanamountestimatedinAppendixF.8andtheinduceddragisaddedbackupto
2.00

2.00

CLmax

1.80

1.80

extr.

1.60

1.20

1.40

1.00

1.20

CL

CL,CD

CD0 extr

FinalCD

1.60

exp.

1.40

CLmax

0.80

CD tunnel

CD0 tunnel

1.00

0.60
0.80

0.40

0.60

0.20
0.00

0.40
0

10
15
AoA/degrees

20

25

0.0000

0.2000

CD

0.4000

0.6000

Figure5.17(a)LiftcurveextrapolationtofullscaleRe;and(b)DragcurveextrapolationtofullscaleRe
Thelimittotheresultobtainedisthattheangleofmaximumliftincreasesexcessively,andthestallwouldalso
bemoreabrupt[Jacobs&Sherman(1937)];moreoverthedatausedfortheextrapolationregardstheairfoil
sectionperformanceandnotthewing.Forthisreason,theresultspresentedinthenextsectiondonotinclude
the extrapolation process. However, the designers believe that this result can give an indication of the
potential increase in lift for the full scale rig. For the application of the proposed wing sail system, the
improvementwouldregardonlybeamreachingandreachingcoursesasthebestaerodynamicdragatlower
AoAremainsunchanged.

77

Concept Design of a Fast Sail Assisted Feeder Container Ship


5.2.6

AnalysisoftheResults
Wingspacingeffect

Inthissetofexperimentstheliftvariationofthemodelwithdecreasingwingspacingisinvestigated.General
BiplaneTheory[Munk(1923)]providesliftvariationwithspacingduetoAoAandcurvatureeffects,basedon
Prandtls vortex theory using twodimensional inviscid flow. A similar trend was anticipated, however direct
comparisonisnotpossiblesincethepressuredistributiononthewingswasnotmeasuredinthetests,anddue
tothepresenceofathirdwing.TheliftvariationduetochangesinspacingisplottedinFigure5.18(right)from
Biplane Theory. The constants and are related to lift due to AoA and curvature respectively and are
defined by the relationships; lift produced by AoA,L 2AqG sin and lift produced by curvature L
CL /2.Assumingthereisnocurvaturetogeneratelift,
1.0indicatesthelift
2SqG sin ,where
atinfinitespacingand
0.5atzerospacing.Thisisbecausezerospacingrepresentsasinglewingwiththe
referenceareaoftwowings.
TheexperimentaldatainFigure 5.18(left)showstheliftcoefficientvariationwithspacingatthreeanglesof
attack.Allthreetrendlinesdepictalossinliftcoefficientasthegapdecreases.AtanAoAoftendegreesthe
trendisparabolicasinGeneralBiplaneTheory[Munk(1923)];showing22%,23%and24.6%decreaseinliftas
thespacingdecreasesfrom100%to50%chordlengthatanAoAof15,10,5degreesrespectively.Thisverifies
thatthedecreaseinspacingresultsinareductioninliftandthatthistrendisproportionaltothedecreasein
AoA.

1.3

AoA 15
1.2

B0

0.9
1.1

AoA 10
0.8

G,G0

CL

1.0
0.9

0.7

0.8

AoA 5

0.7

0.6

0.6
0.5

0.5
0.7 Gap/chord 1

0.4

1.3

Gap/chord

Figure5.18Multiwingliftcoefficientexperimentaldatawithchangeinspacing(left),andBiplanetheorydata
(right)
In Figure 5.18 (left) the relationship between lift coefficient and AoA is illustrated at four different spacings.
TheresultsshowthatapartfromzeroAoAtheliftincreaseswithincreasedspacing.Thestallingangleranges
from16to18degreesandbeforestalltheliftincreaseslinearly.Thedragcoefficientvariationdoesnotshow
clear trend with changing spacing, however there is no significant change in overall trend. At 50%and 75%
spacingthedragcoefficientvariationissimilarbutbetween100%andthemaximumspacing(120%)thereis
substantial difference in drag. At a wing spacing of 100% of the chord length, the drag coefficient is lower
throughouttherangeofAoA,showinganaverage13%reduction.Theaverageliftisonly0.8%greaterat120%
wingspacingandthemaximumliftisat16degrees.100%spacingshows0.2%greaterliftcoefficientresulting

78

Sail System Design


in a higher / . Biplane Theory [Munk (1923)] indicates that a wing spacing close to 100% is the most
efficient. In the drag breakdown plot, Figure 5.20, it can be seen that the 100% spacing shows better
performance in terms of liftdrag ratio. The aerodynamic characteristics of the proposed sails system are
illustratedinFigure5.19.
Cl50%
Cl100%

Cd50%
Cd100%

Cl75%
Cl120%

1.25

Cd75%
Cd120%

0.35

1.15

0.30

1.05
0.25

0.85

CD

CL

0.95

0.20

0.75
0.15

0.65
0.55

0.10

0.45
0.05

0.35
0

10

15

20

AoA/degrees

10

15

20

AoA/degrees

Figure5.19Lift(left)anddrag(right)coefficientagainstAoAatfourspacings
0.35

100%

120%

0.30

CD

0.25
0.20
0.15
0.10
0.05
0.15

0.35

0.55

0.75

0.95

1.15

1.35

CL2

1.55

Figure5.20Dragbreakdowntrendat100%and120%spacing
Staggereffect
Attheendofthefirstsetofexperiments,modelcontainerswereaddedtoinvestigatetheinteractionwiththe
sails. As no time was available to repeat all the initial configurations,only the maximum chord spacing was
testedagaininordertoprovidedataforacomparisonwiththeresultsoftheisolatedwing.Themaximumwing
spacingconfigurationtestedwiththecontainerswillbereferredtoasWingA
GeneralBiplaneTheory[Munk(1923)]alsoinvestigatedtheeffectofstaggerposition.Figure 5.21showsthe
variation of for nonstagger biplane and small, medium and large stagger positions at three different

79

Concept Design of a Fast Sail Assisted Feeder Container Ship


gap/chordregions.Theplotshowsaslightincreasein henceincreasinginliftbutthetrendisnotconsistent,
showing a linear increase in lift at gap/chord ratio of around 1.1 and 1.4 and as the stagger to chord ratio
increases, shows a nonlinear trend at a gap/chord ratio about 0.7. It can be said that the effect due to
spacingchangeisdominantwherethereisaslightincreaseinliftforstaggerpositionassummarisedinGeneral
BiplaneTheory[Munk(1923)].
Duringtestingthestaggerpositionandspacingarelinkedsincethedistancebetweenthewingstocksarefixed
atmaximumspacing.Thetwostaggerconfigurationsindicatedwithrelativeangletothetopandbottombar
canberedefinedrelativetopercentagechordlengthasshowninTable5.13.
Table5.13Threestaggerconfigurationinpercentagechord
Stagger
0
30
60

Gap / chord
1.2
1.039
0.6

Stagger / chord
0
0.6
1.039

TheexperimentalresultssummarisedinFigure 5.22Aerodynamiccoefficientsanddragbreakdownforthree
staggerconfigurationsshowthat,Stagger60producesthehighestliftanddragcoefficient;Fromnowonthis
configurationwillbereferredtoasWingB.Thestaggeredcasesalsohavehigherliftcurveslopewithinthe
linearregion.IntermsofliftdragratiothedragbreakdownplotinFigure 5.21illustratesthatbelowCL2of1.8
Stagger30givesbestperformanceandbeyondthisWingBshowsbetterperformance.Aninterestingresultis
WingBshowshigherliftcoefficientthanStagger30althoughtheamountofdecreaseinspacingandincrease
instaggerpositionrelativetothechordlengthwasthesame.Thereforewiththeexistenceofthethirdwing
the stagger effect is greater than the spacing effect in lift coefficient variation. However the overall
performanceofthesestaggeredconfigurationsvarieswiththerangeofliftcoefficientproduced.
1

0.68

0.9

0.78 0.53

0.83 0.6
0.49

0.25

0.20

Staggertochordratio

0.8

0.32

0.7

Non staggeredtrend

0.6

0.5
0

0.4

0.8

1.2

1.6

Gap/chord

Figure5.21VariationinbiplanestaggerpositionliftconstantGwithgap/chordratio

80

Sail System Design


Stagger30Cl
Stagger60Cl
stagger0Cl

Stagger30Cd
Stagger60Cd
Stagger0Cd

Stagger30

Stagger60

Stagger0
0.60

1.80
1.60

0.50

1.40
0.40

1.00

CD

CL,CD

1.20

0.30

0.80
0.60

0.20

0.40
0.10

0.20
0.00
0

10

AoA(degrees)

15

0.00

20

CL2

Figure5.22Aerodynamiccoefficientsanddragbreakdownforthreestaggerconfigurations
Flatsailconfiguration
Thisconfigurationwasoriginallysettorepresentcontinuoussinglewingshape,aligningthethreewingsinline.
However,duetoscrewsandnutsfortheflapsthecompletealignmentcouldnotachieved.Theaerodynamic
coefficients and drag breakdown results illustrated in Figure 5.23 show lower lift coefficient and higher drag
coefficientthantheotherconfigurationsandtheliftdragratioisalsopoor.HoweverWingCstalledatvery
high AoA behaving as a conventional sail. This configuration may be used for downwind sailing at low
operationalspeed(seeAppendixF.3).
0.95

1.70
1.50

0.75

1.30

0.55

CL,CD

CD

1.10
0.90
0.70
0.35

0.50
0.30

0.15

0.10

15
25
35
AoA/degrees
Figure5.23Flatsailconfiguration:dragcoefficientagainstliftcoefficientsquared(left)and;liftanddragcoefficientagainst
AoA(right)
CL2

81

Concept Design of a Fast Sail Assisted Feeder Container Ship


ContainersSailInteractions
Theexperimentsontheeffectofwingspacingwereperformedwiththeriginisolation,whilsttheinvestigation
onthestaggereffectwascarriedoutsimulatingthepresenceofthecontainerstack.Theinteractionbetween
thesailsandtheflatsurfaceincreasedtheliftbecausethetopofthecontainersworkedasareflectionplane,
decreasingtheendvortexesonthewingsandhenceincreasingtheoverallefficiencyoftherig.

1.50

1.50

1.00

1.00

CY

Cx

0.50

0.50

0.00

0.50

nointeraction

nointeraction

1.00

interaction

interactions
0.00

1.50
0

50

100

150

200

50

100

150

TrueWindAngle /deg

TrueWindAngle /deg

Figure5.24Shipaxescoefficientsformaximumwingspacingspacing.Theplotshighlightthebenefices
derivingfromthepresenceofthecontainersunderthebaseofthewing
Whenthewingwastestedinisolation,theendvortexesgenerateacrossflowatthelowerandupperpartof
thewing,thuscontributingtotheinduceddragandreducingtheefficiencyofthewing.Oncethereflection
planeisadded,theflowbecomesuniformalsoattheextremitiesthusincreasingtheefficiency,duetoviscous
effectsthisgaininperformanceextendstowardsthetopofthewing[Chret(2000)].Theeffectiverigheight,
whichthemagnitudeoftheinduceddragdependson,isfoundtoincreaseby8.6%withtheadditionofthe
containers; for the maximum wing spacing case, increased from 2.21 metres to 2.4, 38% more than the
geometric centre at midspan. Appendix F.7 shows the attemptto visualize the flow in the wind tunnel. The
resultofthisinvestigationprovedtobebeneficialtothescopeofthefeederconceptasthemaximumthrust
coefficient , was found to increase by 15% (Figure 5.24), thus improving the efficiency of the auxiliary
propulsion.
DesignSelection
While examining the polar plots for the different configurations, the best performing arrangements can be
readilyidentified.Astheefficiencyandhenceviabilityofthefeederhasbeenassessedsimulatingavoyageon
designated routes (Section 6.1.1), it is necessary to follow the same approach for the selection of the best
configuration;thiswillenablethechoiceofthemostsuitablesailingpointandwingconfiguration.Figure 6.1
showsaflowdiagramrepresentingtheprogramusedtopredictthefeedervoyageperformance.Anadditional
loopiscreatedtorunthecodeforeverysailingpointandconfigurationtestedinthewindtunnel.Thechoiceis
mainlybasedontheaveragethrustgeneratedfortheroutes.Asexpected,thebestresultsaregivenbyWing
A,andWingB.

82

200

Sail System Design


Table5.14Designchoice.WingAcorrespondtomaximumchordspacing,zerostaggerwhilstWingB
correspondstoastaggerangleof60degrees
Parameter
CLmaxat18degreesofattack
CDatCLmax
CD0
(CL/CD)maxatfive degreesAoA
Minimumaerodynamicdragangle/deg.
Centreofeffort/%span
Effectivecentreofeffort/%span

1.80

WingA
1.423
0.426
0.062
7.7
7
35
111

WingB
1.608
0.553
0.079
5.6
10
33
106

1.8

WingB

1.60

1.6

1.40

1.4

WingA

1.20

WingB

1.2

1.00

WingA

CY

CX

1.0

0.80

0.8

0.60

0.6

0.40

0.4

0.20

0.2

0.00

0.0
0

60

120

180

Apparentwindangle /deg

60

Apparentwindangle /deg

120

Figure5.25Shipaxescoefficients[lift(left);drag(right)]oftheselectedconfigurations;themarkerindicates
thechangefromthesailingpointfromwindwardtobeamreaching
at
The best windward sailing point for both the configurations is at 14 degrees, despite having
is5.2and4.3forWingAandWingBrespectively,but
fivedegreesangleofattack.Atthisangle
thehigherliftcoefficientovercomesthedisadvantageofhavingasmalleraerodynamicdragangle.Figure5.25
shows only the positive values since along the abscissa the curve is almost symmetrical. A negative heel
coefficientat
170 isnotrealistic.Whenrunningdownwindtheflatsailconfiguration(WingC)would
beusedinthestalledcondition,unfortunatelythisconfigurationwastestedonlyupto30degreesbecausethe
heelingforcewouldhaveincreasedtonearhalfofthedynamometerlimitandhencenoriskwastaken.At30
1.63.
degreesthisconfigurationhas
Whenselectingthefinaldesign,performanceisthemaindriver;however,thedimensionalenvelopeofthetwo
configurationsisalsoconsidered(SeeAppendixF.4).InthiscaseboththecriteriafavourthechoiceofWingB.
Figure 5.26comparestheshipaxesliftcoefficientfordifferentsailplansatthemaximumliftcoefficientsailing
point.TheMultiwingsailandtherigidsquareriggeneratesimilarthrustandbotharemoreefficientthanthe
Dynarig.However,itisnecessarytorememberthatthedatatakenfromTable 5.1reflectsthecharacteristics
oftherigsatthetimeofpublicationandthereforedoesnotincludethedevelopmentsofthepast20years.
Thehybridrectangularsails[Fujiwaraetal.(2005)],outclasstheformerrigtypesintermsofmaximumthrust
generated.InthegraphthethrustcoefficientofWingBisalsoplottedusingthesailcoefficientsafterthefull
scaleextrapolation.

83

Concept Design of a Fast Sail Assisted Feeder Container Ship

DynaRig

RigidSquare

multi WingBextr.

HybridRectangularSail

MultiWingB

2.50

2.00

Cx

1.50

1.00

0.50

0.00
15

35

55

75

95
115
TrueWindAngle /deg

135

155

175

Figure5.26ComparisonofMultiwingrigwithotherexistingrigs.Thegraphshowsonlythesailingpointat
maximumliftcoefficient
Singlefoil
Following the last day of testing, the containers and the side wings were removed in order to measure the
forcesofasinglefoilinisolation.Figure5.27comparesthesailcoefficientsbetweenthesingleandMultiwing
arrangements.TheslopeoftheliftcurveislowerfortheMultiwingresultinginahigherstallangle.Thebase
drag is lower for the single wing but for angles of attack greater than seven degrees, the Multiwing shows
lowerdrag.ThemaximumliftoftheMultiwingisalsoslightlylowerthanthesinglewing.NotethatinFigure
5.27,WingAdoesnotincludetheinteractioneffects.
ThissectionalsoaddressesthecomparisonofthewindtunneldatawiththatgeneratedusingXFoilandfrom
publisheddataforthesamesection.However,duringthetestingthewingwasleftwiththeflapdeflectedat
25 degrees. The consequence is that less experimental data is available for the validation. Sears & Liddel
2.76 10 ,buttheflapis30%
(1942)containsdatafortherequiredflapdeflectionof25degreesata
ofthechordlengthandnot20%asforthesailtested.InFigure5.27theXFoilpolarseemstooverestimatethe
liftsincetheflaplengthis20%ofthechordlength,howevertheXFoilanalysisisbasedonanidealgeometry
wherethegapattheflaphingepointissealed;thisincreasestheperformanceandbalancesoutthedifference
inflaplength.OntheotherhandagapwasintentionallyleftatthehingepointoftheMultiwingmodel,to
replicatethefullscalerig.Thereforethelowerdragcurveofthesinglefoilcouldbeexplainedbythedifferent
flaplengths.
TheliftcurveoftheexperimentalsinglewinghassimilargradientinthelinearregiontotheXFoilandNACA
polars. However, the angle of stall for the NACA polar is considerably higher than for XFoil and the tested
wing,thiscanbepartlyexplainedbythehigherReynoldsnumberoftheformerpolar.Alargerdiscrepancyis
found in the drag curves; the base drag of the experimental foil is quite low and the curve gradient is also
higher.PartofthiserrorcanbeexplainedbythefactthatbothXFoilandNACAdatareflecttheperformance
ofsectionswithidealsurfacefinish;thesurfacefinishofthemodeltestedinthewindtunnelwascertainlynot
idealduetothematerialusedandthepresenceofthetufts.Theleadingedgegeometrymayhavecontributed
tothisdecreaseinefficiencyduetothesandingprocessrequiredtoremovetheexcessfiller.

84

Sail System Design

Multi Wing

Cdexp.
ClXFoil

SingleWing

Clexp.
CdXFoil

2.00

1.40

Single

1.80

1.20

1.60
1.40

1.00

CL,CD

CL , CD

Multi

0.80
0.60

1.20
1.00
0.80
0.60

0.40

Single

0.40

Multi

0.20

0.20

0.00

0.00
0

10

15

20

10

15

AoA/degrees

AoA/degrees

Figure5.27SingleandMultiwing(WingA)comparison(left)and;Singlefoilcomparison,liftcurve
(ExperimentalandXfoildataisforRe=235,000whilsttheNACAdatapointsareforReef=2.76x106(right)
UncertaintyAnalysis
ThesamerationaleasthetowingtankuncertaintyanalysisaddressedinSection 4.4.4canbeappliedtoassess
theaccuracyofthewindtunnelmeasurements.Hereonlytheprecisioncomponentofthetotaluncertaintyis
evaluatedasthenumberofrepeatedrunswastoosmalltoprovideasignificantbiasanalysis.Appendix F.10
showsthepolarplotaftertheerrorbarsareapplied.
Sourcesofuncertaintyincludemodelgeometry,whichincludesdimensionaltoleranceofthehotwirecutter,
the addition of grit and polyfiller to seal the leading edge gap, and the presence of the tufts. The result is
variation of the surface roughness which in turn increases the drag, hence altering the performance. A
correctionforthegritstripshasbeenapplied,whilstthetuftswereremovedaftertheirinitialtests.Further
uncertaintyarisesfromthetimetraceofthesampleddata,duetoanomalieswithtwoofthechannelsusedto
derivetheoutput;thisuncertaintywasreducedbyrepeatingruns.

Discussion

TheinitialconceptofMultiwingsystemwastargetedforupwindsailingperformance.Theproposedsystem
hasbeendevelopedbyconsideringwingspacing;magnitudeofstagerandflaplengthtooptimisetheWalker
Multiwingsail[Walker(1985)].Toensurethestructuralintegritysimplebeamtheoryhasbeenusedatboth
shipandmodelscale.Thesailplangeometryhasbeenoptimisedwithindimensionalconstraints.ThroughX
Foil aerodynamic analysis and structural considerationNACA0015 with 20%chord length trailing edge flap is
selectedforthewingsection.Usingtheobtainedsectionaerodynamiccharacteristicsandaveragewinddata
fortheoperatingregionsanoverallperformancewasanalysedanda3%thrustreductionat25knotsand10%
at15knotsofshipspeedispredictedasdefinedintheinitialrequirements.Howevertheperformanceestimate
was based on two dimensional section coefficients of a single NACA 0015; any further improvement in the
qualityandfidelityoftheprojectionsnecessitatedmodelscalewindtunneltesting.
Throughoutthedurationoftheexperiment,ithasbeenpossibletocompletethetestmatrixoriginallyplanned.
The results of the performance when varying wing spacing suggested that the optimal gap corresponds to
100%ofthechordlength.TheresultswerefoundtoagreewithGeneralBiplaneTheory,despitetheadditional
wing.

85

Concept Design of a Fast Sail Assisted Feeder Container Ship


Thesecondsetofexperimentsprovedthattheadditionofanangleofstaggerincreasestheefficiencyoftherig
thankstothesloteffectwhichfavourstheperformanceoftheupwindwings.Alongwiththissecondseriesof
tests, boxes were positioned under the wings to replicate container stacks on the fast feeder. From this
experiment the efficiency of the sail plan was found to increase due to a reduction of the induced drag
component.Theresultspresentedinthischaptercontributetotheassessmentoftheoveralldesignfeasibility
presentedinSection 6.2.Themethodsandresultspresentedarenotcommonintheliteratureandtherefore
actasausefulresourceaidingfutureresearchanddesign.

5.3

- Computational fluid dynamics study of Multi-wing system

Computational fluid dynamics (CFD) is potentially an alternative method to model experiments for
performanceprediction,intermsoftimeefficiency.Duetodifficultiesinsimulatingrealisticflowconditions
experimentallysomedegreesofuncertaintyexistintheresults.UsingCFDonecanvisualisetheflowaround
objects.InthisprojecttwodimensionalCFDisusedtoundertakefurtherinvestigationoffullscaleMultiwing,
particularlyfromanoptimisationperspective,whichwasnotavailableduringthetesting.Sincetheanalysisis
in two dimensions conducted at fullscale and has a different environmental conditions to the wind tunnel,
hencetheresultsmaynotmatch.Tocarryoutthisanalysis,thecommercialsoftwareStarCCM+isused.

5.3.1

Meshgenerationandphysicalmodel

In mesh generation fifteen prism layers are used for the boundary layer and three sizes of unstructured
tetrahedralmeshformaindomainasshowninFigure5.28.Usingaflatplateempiricalformulaetheboundary
layer thickness of 0.12 metres and first layer thickness of 0.95 millimetres were obtained assuming wall
function,Y+of50andinletvelocityof15.432ms1equivalentto30knots[Tuetal.(2008)].Finermeshingis
placedaroundthewingsectionsandinthedownwindregiontocomputetheturbulentflowinthewake.The
samemethodhasbeenappliedtoalloftheCFDmeshesused.
DuetohighReynoldsnumber(6.2 10 )oftheflowturbulentmodellingisrequired.Intheoptimisationof
the Maltese Falcon sail rig [Doyle et al. (2002)] realisable kepsilon and SpalartAllmaras turbulence models
showed little difference in results so SpalartAllmaras model was used to increase the efficiency of the
iteration.InthisMultiwinganalysisfulloptimisationwasnotconsideredsoforthe physicsmodelkepsilon
turbulentwithSegregatedflowisusedduetothegeometricalchangearoundtheflapswhichmaycausemore
turbulentflowinthewakeregion.

5.3.2

Simulations
Wingspacingof100%chordandflapangleof22.5degrees

The above defined model showed best liftdrag ratio in the wind tunnel testing in the nonstaggered
configurationsmoredetailedcharacteristicsofthisconfigurationareinvestigatedusingCFDmodelling.
Theoverallliftanddragcoefficienttrendsaresimilartothewindtunnelresultsbuttheindividualwingsshow
different trends around stalling angle. The drag values show insignificant change at AoA of less than eight
degrees and at higher AoA they start to increase at different rates, as seen in Figure 5.29. This implies the
stallingofonewingchangestheflowregimearoundtheotherwings.Intermsofliftdragratiowithreasonable
liftgeneratedeightdegreesAoAshowsthebestperformancebutthemaximumliftisproducedattendegrees
AoA.
InFigure 5.30increaseinseparationonthewingsattendegreesAoAcanbeseen.Sincetheliftanddragvary
at different rates Multiwing can be optimised by adjusting the AoA of each wing for best liftdrag ratio or

86

Sail System Design


maximumlift.Furthermorethestaggeredconfigurationsandtheflapangleofthewingscanbeoptimisedfor
differentsailingconditions.

Figure5.28Meshgenerationofwholedomain(left)andsinglewing(right)
1.5

CL,CD

1.0
TopwingCL
TopwingCD
MiddlewingCL

0.5

0.0
0.0

2.0

4.0

6.0
8.0
AoA/Degree

10.0

12.0

Figure5.29Aerodynamiccharacteristicsof100%chordspacingwith22.5degreeflapmodel

Figure5.30StreamlinesatanAoAof:eightdegrees(topleft),tendegrees(bottomleft);andpressurecoefficientateight
degrees(right)
Forwardandaftsailsysteminteraction
In the upwind sailing condition the air flow behind the forward sail system may affect the aft system
performance. Based on the performance prediction the average apparent wind angle of 30 degrees was
selectedfortheanalysis.Atthisangletheflowinteractionbetweentherigsmaybemoresignificantthanin
othersailingconditions.Twodimensionalrigsaremodelledtohaverelativeapparentwindangleof30degrees
andAoAoftendegrees.Thedomainsizewasincreasedby20%overallfromthesinglerigmodelandshiftedto
accommodatetworigsensuringsufficientwakeregion.Tomeasuretheinteractiontheentryflowangleand
magnitude, total lift and drag coefficient and streamline of forward and aft system are obtained in post

87

Concept Design of a Fast Sail Assisted Feeder Container Ship


processing.Averticallineprobewasderivedatfivemetres(80%chordlength)beforeeachrigwith19metre
length(300%chordlength).
Fromthehorizontalandverticalflowvelocitycomponentsthemagnitudeandangleofflowalongtheprobes
werecalculated.InTable5.15themagnitudeandangleoftheentryflowaresummarisedandtheresultsshow
1.53% flow velocity reduction and 6.04% reduction in AoA. In aerodynamic characteristics the forward rig
producedlessliftandmoredragthantheaftrigduetostalling.InthestreamlineinFigure 5.31separationis
smallerintheaftrigduetoreductioninflowvelocityandAoA.ThisimpliesthatinupwindoperationtheAoA
oftheaftrigcanbeincreasedtoproducemoreliftwhentheforwardrigisinoptimumsailingcondition.
Table5.15Entryflowcharacteristicsofforwardandaftrigwithtotalliftanddragcoefficient
Forwardrig
Aftrig
Reduction/%

Flowvelocity/ms1
15.66
15.42
1.53

Flowangle/degree
17.89
16.81
6.04

CL
0.96
1.03
6.80

CD
0.143
0.127
12.60

Directionoftravel
30

Apparentwind

Figure5.31StreamlineforforwardandaftMultiwingsystematapparentwindangleof30degrees

88

Performance Predictions

6. Performance Predictions
This Chapter deals with the prediction of the aerohydrodynamic performance of the fast feeder. The
experimental results from both Chapters 4 and 5 are combined in conjunction with estimates of propulsive
efficiencyallowingfinalhullformselectionandcomparisonofperformanceagainstexistingvessels.

6.1

Design selection

6.1.1

Sailingperformance

The program described in Section 5.1 can now be expanded to encompass the aerohydrodynamic data
obtainedintheexperiments.Initially,thealgorithmperformstheaerodynamiccalculations,wheretherigand
hullwindagearecombinedwiththeforcesgeneratedbythesailstocomputetheresultingangleofheeland
leeway.Thecodethencalculatesthetotalresistanceoftheshipwhichisdefinedas:

(6.1)

where
istheaddedresistanceinwavesaveragedovertheselectedroutes,andR AAS istheairdragofthe
abovewaterstructure.ThecodeisthenusedtocomparethedifferentconfigurationsoftheWingsailtested
inthewindtunnel;thisenablestheassessmentoftheefficiencyoftherigonanoperationalbasisratherthan
justfromthepointofviewofpureaerodynamicperformance.

RigWindage

TheshipaxiscoefficientsoftherigwindagepresentedinSection5.2.5aredimensionalisedwithrespectthesail
areaandtheresultingforcesareaddedtotheforcesresultingfromthesails.
Windlever
ThewindheelingmomentisestimatedusingEquation(6.2)[IMO(2008)].
1
2

(6.2)

Thecodeinterpolatestheprojectedareaforthecorrespondingapparentwindangle;agustinessfactorof1.23
is included to represent fluctuating wind speed. The angle of heel induced by the wind load is once again
calculatedseparately,andthenaddedtothesailcomponent.
Addedresistanceinwaves
The annual wave data for Singapore and the Caribbean are initially converted into probability of occurrence
can be transformed
regarding significant wave height and zero crossing period. Using Equation (4.6),

forbothHullAandHullB(TableG.1andTableG.2).Onthisbasis,thenondimensionaladded
into
,canbematchedtotheseastateanddimensionalised
,whichwastestedoverarangeof
resistance
forthespectrumofwaveamplitudeusingEquation(4.25).Theaimofthisistocalculateaninstantaneous
addedresistanceforSingaporeandtheCaribbean,whichcanbeaddedtothecalmwaterresistanceinthePPP.
Table 6.1 and Table 6.2 shows
for both hulls operating in the Singapore region at 25 knots speed,
AppendixGshowsthesummaryfortheCaribbeanandthe15knotspeedcasefortheSingaporeregion.

89

Concept Design of a Fast Sail Assisted Feeder Container Ship


Table6.1AveragedaddedresistanceforHullAsailingat25knotsinSingapore(valuesinN)
Wavelength/wettedlength
12
20
30
42
0.0
0.0
686.8
3149.2
0.0
0.0
584.8
1838.8
0.0
0.0
41.4
71.4
0.0
0.0 1339.1 5271.3

HullA,25knots
/ /m
2.5
1.5
0.5
Total

55
3689.2
1531.6
35.7
5658.8

71
89
109 130 Total
2317.8 1009.7 302.0 76.3 11231.1
719.9 227.2 54.4 0.0 4956.6
9.1
2.5 0.0 0.0 160.1
3046.8 1239.4 896.9 76.3 16347.9

Table6.2AveragedaddedresistanceforHullBsailingat25knotsinSingapore(valuesinN)
Wavelength/wettedlength
12
20
30
0.0 137.9
1751.5
0.0 176.1
1491.4
0.0
25.7
105.5
0.0 339.8
3348.4

HullB,25knots
/ /m
2.5
1.5
0.5
Total

42
3831.4
2237.1
86.9
6155.3

55
71
89
3703.0 2150.7 896.3
1537.4 668.0 201.7
35.8 8.4
2.2
5276.2 2827.1 1100.2

109
283.9
51.1
0.0
335.0

130
Total
86.8 12841.6
0.0 6362.7
0.0 264.6
86.8 19468.8

Onlysignificantwaveheightsupto2.5metresareconsideredasthisrepresents89.3%ofobservations.Some
of the wave length ratios from the sea charts fall outside of the range of waves covered in the towing tank
are therefore extrapolated for /LWL < 50. Linear interpolation is used to find
testing; the values of
withintherangecoveredbytheexperiments.Noaddedresistancecanbeappreciatedat/LWL<30and
/LWL<20forHullAandHullBrespectively.At25knotsshipspeed,theaveragedragamountsto1.1%and
1.7%oftheuprightresistanceforHullAandHullBrespectively.Alargediscrepancyintheperformanceofthe
twohullsisobservedatashipspeedof15knots(see AppendixG)wherethepercentageofaddedresistance
amounts to 1.6% and 4.9% of the upright resistance for Hull A and Hull B respectively. The average added
resistance is then weighted for the probability of head seas, assuming that the waves will follow the same
direction as the wind. To calculate the weighting factor the probability of head winds relative to the ship
heading ismultipliedbyafactorwhichaccountsforthenumberofmilesthefeedersailsateachheading.
ThecalculatedweightingcorrespondingtotherouteintheSingaporeregionisshowninTable6.3.

(6.3)

Table6.3CalculationoftheprobabilityofheadwavesforSingaporeregion
Course( )
045
4590
90135
135180
180225
225270
270315
315360
SUM

Distance/miles
1716
1721
1454
5511
1716
1721
1454
5511
20803

distance/total distance
0.082
0.083
0.070
0.265
0.082
0.083
0.070
0.265
1.00

P(bow waves) Weight

0.1326
0.2258
0.1659
0.0546
0.0915
0.1217
0.1101
0.0979

0.0109
0.0187
0.0116
0.0145
0.0075
0.0101
0.0077
0.0259

1.00

Weight=0.107

Inthealgorithm,theinstantaneousresistanceinwavesisthereforeestimatedas

90

0.107.

Performance Predictions
Inducedandheeledresistance
Theresultsoftherunswherethemodelsweretestedatanglesofheelandleewayareinputintothecodeand
using a double interpolation technique it is possible to match the hydrodynamic and aerodynamic sideforce
andknowingtheextentofvesselheeldeterminingthemagnitudeoftheleeway.Thisprocessisillustratedin
Figure6.1.
PPPflowdiagram

VS,VT,,CL,CD

HullWindage

RigWindage

SailcoefficientCX,CY

ThrustFxSideforceFYHeel

Induceddrag

Addedresistance

/heeleddrag

Leeway

Changesail
configuration

Calmwaterdrag/airdrag

ThrustReduction

RollDamping

Output

Figure6.1Flowdiagramforperformancepredictionprogram.Thesourcecodeisnotincludedinthe
appendicesforpracticalreasons.ItwillbeincludedintheCDromattachedtothemainreport.
Voyagesimulation
The performance of the feeder is to be assessed over the selected routes in the Singapore and Caribbean
regions. The output file of the PPP describes the performance of the proposed hull forms in arbitrary wind
conditions.Thisisthenfedintoaspreadsheetwhichaveragesthisdataoverthe24selectedroutescombining
theprobabilityofwindspeedanddirectionrelativetotheshipheading.Table6.4showsoneoftheroutesfor
Singapore. The average thrust reductions, angle of heel, leeway and motion damping can be calculated for
individualroutesorfortheentirearea.Thefollowingcalculationsareundertakenusingthemostefficientsail
configuration,WingB(seeSection5.2.5).

91

Concept Design of a Fast Sail Assisted Feeder Container Ship


Table6.4VoyagedetailsforSingaporearea
Waypoint

Speed/
miles Course/deg Reversecourse/deg
knots

SailUse

Position

Singapore

1.2 100.5 1.09 100.7

Singapore(Port)

EofSakijang(LH)

SofRaffles(LH)

Malacca(Strait)

Malacca(Strait)

25

25

1.6

90.04

270.04

25

1.9

155.04

335.04

1.1 103.4 1.05 103.3

25

9.6

266.15

86.15

1.1 103.3 1.07 103.2

25

17.5

283.01

103.01

2.2 101.5 2.10 101.3

25

119.7

298.76

118.76

OneFathomBank

2.5 100.6 2.30 100.7

25

59.2

289.55

109.55

N/A

3.6 99.3 3.75 99.2

25

105.3

312.94

132.94

Belawan(Port)

3.5 98.5 3.71 98.3

25

40.7

267.80

87.80

Belawan

3.5 98.4 3.28 98.3

25

8.8

185.57

5.57

1.2 103.5 1.09 103.7


1.1 103.5 1.08 103.7

(6.4)

The annual average includes all routes and the probability matrix reflects annual wind data. The seasonal
averagereflectssummerandwinterwinddataandisbasedononeroutewhichisselectedforitsfavourable
winddirectionrelativetotheshipheading.Thespreadsheetcanbeusedtoaverageothervoyagedatawith
resultsgiveninTable6.5andTable6.6.
Table6.5PerformancesummaryforbothhullformsinSingaporeregion
Hull A

HullB

Annual

Speed

<4

15

Seasonal
25

<4

15

Annual
25

<4

15

Seasonal
25

<4

15

25

Thrustbenefit/%

80.0 2.5 0.9 80.0 4.1 1.5 81.0 3.5 1.1 81.0 5.9 1.9

Angleofheel/deg.

0.1 0.2 0.6 0.1 0.3 0.7 0.1 0.2 0.6 0.1 0.3 0.7

Leewayangle/deg.

0.0 0.3 0.9 0.0 0.4 1.0 0.0 0.4 1.3 0.0 0.6 1.4

Inducedheeleddrag/%RT 0. 0.7 0.3 0.0 0.7 0.3 0.0 1.6 1.3


RollDamping/%

1.7 1.5

9.4 16.3 30.1 10.0 20.9 31.1 9.4 16.3 30.1 9.4 20.9 31.1

Table6.6PerformancesummaryforbothhullformsinCaribbeanregion

Speed/knots

HullA
HullB
Annual
Seasonal
Annual
Seasonal
<4.0 15.0 25.0 <4.0 15.0 25.0 <4.0 15.0 25.0 <4 15.0 25.0

Thrustbenefit/%
Angleofheel/deg.
Leewayangle/deg.
Inducedheeleddrag/%RT
RollDamping/%

70.0
0.1
0.0
0.0
7.5

2.0
0.2
0.3
1.4
16.0

0.9
0.5
0.8
1.1
32.1

70.0
0.1
0.0
0.0
8.5

92

3.4
0.3
0.4
1.6
16.4

1.1
0.7
1.0
1.3
30.6

72.0
0.1
0.0
0.0
7.5

2.8
0.2
0.4
1.5
16.0

1.0
0.5
1.3
1.4
32.1

72.0
0.1
0.0
0.0
8.5

4.9
0.3
0.5
1.7
16.4

1.4
0.7
1.4
1.5
30.6

Performance Predictions

Angleofheel /degrees

2.5

2.0

Sails

1.5

1.0

0.5

Nosails
0.0

0
10

30

50

70

90

110

130

150

170

190

TrueWindSpeed/m/s

Figure6.2HeelangleHullA,VS=25knots,VT=30knots.Thestepinthesailcurveisthechangeinsail
coefficientsCLandCD.Ahigherwindspeedischoseninordertoappreciatethehullwindagecomponent
40
35

RollReduction/%

25knots
30
25

15knots
20
15
10
5
10

30

50

70

90

110

130

150

170

190

TrueWindAngle /degrees

Figure6.3Variationinrolldampingwithtruewindangle(Singaporemeanwindspeedof5.9ms1)
2.5

HullB
25knots
Leeway /degrees

2.0

HullA
25knots

1.5

HullB
15knots

1.0

0.5

HullA
15knots

0.0
10

30

50

70

90

110

130

150

170

190

TrueWindAngle /degrees

Figure6.4Variationinleewayanglewithtruewindangle(Singaporemeanwindspeedof5.9ms )

93

Concept Design of a Fast Sail Assisted Feeder Container Ship


Figure 6.2showstheplotoftheheelangleatawindspeedof30knots,thelimitforthesafeoperationofthe
sailsatashipspeedof25knotsarisingfromthestructurallimitations(seeSection 7.2.3).Notehowtheangle
ofheelremainsbelowthelimitdefinedinSection 5.1.thesmallamountofheelisexplainedbythesmallsail
area, meaning that additional sideforce could be taken by the hull without compromising stability and
generatingexcessivecontainerlashingloads.Theaverageangleofheelisquitesmallduetotheaveragelow

windspeedsof12knots,theheelangleofHullBisthesameasforHullAsincethemetacentricheights
areassumedtobeequal.
The method used to calculate the roll damping is outlined in Section 7.1.7. Figure 6.3 proves that positive
dampingismaintainedthroughouttherangeoftruewinddirections.Furtheranalysisofthemotiondamping
actionofthesailsisprovidedinSection7.1.7.
ObservingFigure6.5itispossibletonotethattheaverageleewayofHullBishigherthanforHullA,asthishull
form generates less hydrodynamic sideforce. This is due to the Hogner stern which acts a lifting surface,
whilsttheflatsternofHullBdoesnotcontributesignificantlytotheproductionhydrodynamiclift.Thestern
arrangementofHullAalsogeneratesahigherinducedresistance.Duetoflowseparationattheafterendthe
differenceincombinedheeledandinducedresistancebetweenthehullswhengeneratingthesamesideforce
is16%,meaningthatHullBgeneratessideforcemoreefficiently.Figure 6.5andFigure 6.6confirmthisview
showingtheplotoftheeffectivedraughtforthetwohulls;thisisadirectmeasureoftheefficiencyofthehull
formandiscalculatedasshowninEquation(6.5).At15knotsthetwohullformshavesimilarbehaviour,whilst
at 25 knots, Hull A as a much higher induced drag component, as noted in Section 4.4.2. It is necessary to
remember that the models were tested without appendages, therefore it is not possible to estimate the
increaseininducedorheeleddragresultingfromtheadditionofthepods,thusnodefinitiveconclusioncanbe
drawnonwhichhullperformsbetterwhensailingtowindward.
(6.5)

HullA,heelangle=0
HullB,heelangle=0

HullA,heelangle=2.5
HullB,heelangle=2.5

HullA,heelangle=5
HullB,heelangle=5

EffectiveDraught/metres

6
5
4
3
2
1
0

0
0

0.5

1.5

2.5

3.5

Leeway /degrees
Figure6.5Variationineffectivedraughtwithleewayangle,atVS=15knots.

94

4.5

Performance Predictions
HullA,heelangle=0
HullB,heelangle=0

HullA,heelangle=2.5
HullB,heelangle=2.5

HullA,heelangle=5
HullB,heelangle=5

10

EffectiveDraught/metres

9
8
7
6
5
4
3
2
1
0

0
0

0.5

1.5

2.5

3.5

Leeway /degrees

4.5

Figure6.6Variationineffectivedraughtwithleewayangle,atVS=25knots
Inordertoobtainauniquereductionfigurefortheentirevoyage,itispossibletoaveragethethrustreduction
for different speeds using the round trip breakdown shown below in Table 6.7. Originally two different
schedulesweresimulatedtoinvestigatetheeffectsofdelaysontheaveragereductions.Insuchacase,the
feederwouldslowdownto15knotstoincreasethepropulsivebenefitsfromthesails;howeverthedifference
in thrust reduction between schedules was found to be only 0.2% and therefore only the schedule which
reflectsarealisticloadingandunloadingtimeisincludedinthereport.
Table6.7Roundtripspeedweightings

Speed/knots Hours Weight


Approachingport&manoeuv.
<4
2.24 0.027
Slowspeed
15
45.92 0.55
Highspeed
25
35.84 0.43
Port
0
28
0
Total
112

Thechoiceofincludingathirdlowspeedisbasedfromthepossibilityofusingthevectoredthrustofthesails
formanoeuvring,thusreducingtheuseofthebowthruster.ThisisthecasefortheMVAshingtonwhichused
a Walker Wingsail system fitted on top of the superstructure removing the need to use of tugs [Satchwell
(1986)].Thenecessityofhavingbowthrustershoweverremainsasthesailsystemwouldnotbeabletosway
thefeederstraightintothewind.Atlowerspeedsthethrustreductionisunrealisticallyhigh;thisisexplained
bythefactthatnoexperimentaldragdataisavailableforthatspeed,thereforetheHoltropestimateusedto
computethereductions,maynotbeveryaccurateatsuchlowspeeds.Forthesereasons,thereductionsatthe
manoeuvringspeedsarenotincludedinthemaineconomicanalysis.
Table6.8Percentagethrustreductionweightedforroundtrip
Singapore
Caribbean
Annual Seasonal Annual Seasonal
HullAthrustreduction/% 3.9
5
3.3
4.2
HullBthrustreduction/%
4.6
6.4
3.9
5.4

Table 6.8 summarises the thrust reduction for the selected areas of operation, the reductions in thrust are
considerablysmallrespecttotheinitialestimateofSection 5.1.3.At15knotstheestimated10%reduction,

95

Concept Design of a Fast Sail Assisted Feeder Container Ship


hasdecreasedto26%whilstat25knotsthe3%estimatehasdecreasedto12%.Thefactorsthatdetermined
thesechangesaresummarisedbelow:

experimentalCDwasfoundtobelowerthantheoreticalestimatethusincreasingthethrustreduction;

experimental calm water resistance at 15 knots is higher than the original Holtrop estimate thus
decreasingthethrustreduction;

additionofwindagedatadecreasestherigpropulsiveefficiency;

additionofinducedandheeldragdecreasesthesailpropulsivebenefits;

loweraveragewindspeedfortheselectedareaslimitsthethrustreductions.
NorthAtlanticroutes

Oneofthefactorswhichlimitedthethrustreductionsistheaveragewindspeed.Theaimofthissection,isto
undertake an environmental sensitivity analysis and show how the performance of the feeder (mainly fuel
consumptions), would improve if it was to operate in adifferent area. Using data from the initialeconomic
analysisithasbeenpossibletoidentify12routesintheNorthAtlanticregionwhichwouldbesuitableforthe
fastfeeder.Table 6.9presentsthevoyagedetailsfortheselectedroutesincludingtheimprovementinthrust
reductionrespecttotheannualestimateforSingapore(Table 6.1).Theincreaseisduetothehigheraverage
truewindspeed(+34%),whichinturnincreasestheapparentwindangleby16%andmovestheaveragesailing
pointtowardsbeamreach,wherethemaximumthrustisgenerated.

Table6.9AnnualNorthAtlanticvoyagesimulation
Speed/knots
Thrustreduction/%

HullA
HullB
15
25
15
25
4.73 1.66 6.4 2.15

Improvement
Angleofheel/deg.
Leewayangle/deg.

+2.2 +0.7 +2.9 +1.0


0.32 0.69 0.32 0.69
0.44 1.04 0.58 1.22

Inducedheeleddrag/%RT 1.68 1.33 1.84 1.45


RollDamping/%

17.95 30.5 17.95 30.5

A significant improvement of the thrust reduction is observed at the lower speed, whilst at 25 knots the
reductions remain quite low. This result enables the designers to draw some conclusions about the
effectivenessoftheproposedrigandthelimitstoitsapplicationforthefastfeederconcept.Thereductions
achievedatthelowerspeedcanpartiallyjustifytheuseofsails,howeverhighersavingswouldberequiredto
maketheuseofsailsprofitable(seeSection 6.2.2forfurtherdetails).Ifaseconditerationofthedesignspiral
was undertaken the sail area would have to be increased and the possibility of adding a third sail may be
considered. Both solutions would imply radical changes to the fast feeder concept in terms of safety and
capacity. Increasing the span and the chord of the wings would exclude the possibility of stowing the rigs
below the deck; hence the structural analysis would change to assess the loads on the ship structure. The
additionofathirdfoldablerigwoulddecreasethecargocapacityandthewholeeconomiccasewouldhaveto
berevised.Decreasingtheservicespeedwouldcertainlyimprovetheviabilityoftheexistingsailsystem,but
theinitialspecificationpreventsthis.Section 6.1.2addressesthecomparisonofthetwohullformdesignsin
termsofpropulsiveefficiency.

96

Performance Predictions
6.1.2

Propulsiveefficiency

The efficiency of the propulsion system must be estimated in order to predict the total installed power
requirement of a ship, in this case providing a major measure by which to compare the two designs. This
requires estimates of the performance of the propeller and transmission system, as well as the flow in the
wakeandthedragofthepoddeddrives.Inthiscase,propellerperformanceiscalculatedusingthecommercial
software WINPROP, transmission efficiencies taken from manufacturers guidelines and all other calculations
madeusingempiricalmethods.
Generaltheoreticalbasis
Therelationshipbetweeneffectivepoweranddeliveredpower(atthepropeller),isgivenby:
.
Thequasipropulsivecoefficient(

(6.6)

)isdefinedas:
1
1

is the efficiency of the propeller without the presence of the hull, whilst
interaction between hull and propeller. Values of ,

and

(6.7)
and

account of the

for each hull are estimated using empirical

formulae.ThereferencescontainingthesearesummarisedinTable6.1.
Table6.10Referencesusedforestimatingpropulsivecoefficients
Coefficient

HullA
Holtrop&Mennen(1978)
Henschke(1965)
Holtrop(1984)

HullB
Holtrop&Mennen(1978)
Schneekluth(1988)
Holtrop&Mennen(1978)

The coefficient

is calculated using WINPROP which uses standard series propeller charts to derive

performance data for a given set of input parameters. The Wageningen B4.70 series data is used for four
bladedpropellers,whileB5.75isusedforthefivebladedpoddedpropeller.Theseserieswerechosenasthey
provide a relatively high BAR and therefore are less likely to encounter problems involving cavitation.
However,detaileddesignofthepropellersisnotcarriedoutinthiscase.Havingobtainedallthepropulsive
coefficients,thebrake(orinstalled)powerisestimatedby:

(6.8)

The transmission efficiency ( ) in the case of electric propulsion is made up of a number of component
efficiencies,withthevaluesgiveninTable 6.11takenfromthosequotedbyABB(2009).Note issimplythe
productofthefirstfivevalues.
Table6.11Electricalefficiencycomponentsofthetransmissionefficiency
Efficiencycomponent
Frequencyconverter
Generator
Motor
Switchinggear
Transducer
Overalltransmission

97

Value
0.990
0.970
0.970
0.999
0.995
0.926

Concept Design of a Fast Sail Assisted Feeder Container Ship


Since both the propulsion systems proposed utilise two motors powering two propellers, the transmission
efficiencyisassumedthesameforbothdesigns.
Accountingforpoddeddrives
Thesimpledesignprocedureoutlinedintheprevioussectionmustbemodifiedtoaccountforthedragofthe
podded drives for both hulls. The procedure followed is that outlined by ITTC (2008e), which estimates the
dragofthepod(s)andaccountsforthisbyreducingthevaluesof

and

attributedtothepropeller.Note

thisisaslightdeparturefromstandardperformanceestimates,whichnormallyincludeappendagedraginthe
totalresistanceoftheship.However,thereisaconsiderabledifferencebetweenapropelleroperatingbehind
a standard merchant hull form and one mounted on a pod, which this method must account for. The pod
dimensionsareestimatedusingABB(2009)andareassumedthesameforbothHullAandHullBsinceaneven
powerdistributionwasinitiallyassumedforHullA.Whererequiredvalueswerenotavailable,dimensioning
ratiossuppliedbyMollandetal.(unpublished)wereemployed.Figure G.1andFigure G.2providethevalues
usedinallcalculations.Notethat,havingcheckedforadequateclearancesonthegeneralarrangement,the
largestpossiblepropellerdiameterwasselectedinbothcasessoastomaximiseefficiency.
Table6.12Summaryofdimensionsusedin
estimationofpoddeddrivedrag(HullAuses
VO1800andHullBusesVO2100)[ABB(2009)]
Symbol

Figure6.7Poddimensionsusedindragestimate
[ABB(2009)]

Dimension

VO1800

VO2100

Propellerdiameter 3.85.9 4.46.4


Poddiameter
2.33
2.68
Podlength
10.10
11.60
Strutchord
5.76
6.60
Strutlength
2.59
2.97
Strutthickness
1.73
1.99

Using these values and the ITTC 57 skin friction line, the drag of the pod can be calculated based on three
components: pod body drag; pod strut drag; and interference drag. Any component due to lift is neglected
sincethepropellerisassumedtobelightlyloadedatthedesignspeedof25knots.Duringthisprocess,the
valueofinterferencedragobtainedprovedunreliable,andfartoolarge.Therefore,basedonvaluespresented
by ITTC (2008e) the interference drag was assumed to be equal to the pod body drag as a conservative
estimate.Notethatthismethodtakesintoaccountthedifferenceinthevelocityimpingingonthepodinside
andoutsideofthepropellerdiameter.
Accountingforcontrarotatingpropellers
Themainconsiderationwhenpredictingcontrarotatingpropellerdesignistheincreaseininflowvelocityinto
theaftpropeller.Inthiscase,thiswascalculatedinthesamewayasthevelocityimpingingonthepod,based
onMollandetal.(unpublished):

1 ,

(6.9)

;
0.5
0.5/ 1
0.15/
; is the propeller
where is the advance speed, equal to 1
;and istheadvancecoefficient,equalto /
.Notethisservesto
thrustcoefficient,equalto /
increasethetotalpoddrag.vanGunsteren(1971)describesanalternativeapproach,usingabladeelement

98

Performance Predictions
momentum(BEM)modeltoevaluatecontrarotatingpropellerperformance,andoptimisepropellerdiameter.
Whilst this would have been the preferred method, it was considered too complex and could not be
successfullyimplementedinthetimeavailable,andassuchwasdisregarded.
Theseconddesignissuetoselectingthedistributionofpowerbetweentheforwardandaftpropellers.Praefke
et al. (2001) recommend that the designer keep the forward and aft blade passing frequencies the same to
avoid broad band excitation and optimise efficiency by eliminating swirl from the slipstream using a torque
identity.Thisimplies:
,
,

(6.10)

Sinceitisdesirabletominimisethesize,andthereforedrag,oftheinstalledpod,anduponfindingmostCPPs
/
= 0.8 was chosen. Taking into account the propeller speed and
to be fourbladed, a ratio of
shaftpowerratingsofthepodsin(seeFigure G.1),thisassumptionallowedthesmallerVO1800modeltobe
selectedforHullA.Finally,theforwardpropellerwassizedtohavealargerdiameterthantheaftone,taking
accountofslipstreamcontraction.
HavingderivedthepowerdistributionbetweenforwardandaftpropellersfortheHullA,asuitableCPPcould
be selected. Based on delivered power requirement and suitability for fast ships, the Wrtsil LIPS EHUB
4E1300waschosen,providingupto16MWdeliveredpower,andafavourablebossdiameter/overalldiameter
ratioof0.2[WrtsilCorporation(2008)].
Estimatinginstalledpowerrequirement
Applying the theoretical approach described, the required installed power of each ship for calculated.
ImportantvaluesaresummarisedinTable6.13.
Table6.13Summaryofvaluesusedinestimationofinstalledpropulsivepowerforbothhulls
Shaft

,
,

90%
/

fwd
aft
18
18
0.07965 0.07965
0.2883 0.2883
0.9975 1.1658
6.3
5.9
4
5
0.700
0.657
0.847
0.990
11.371
9.093
0.926
0.926
12.280 9.821
25.42

Pod
25
0.09822
0.1263
0.9600
6.4
4
0.706
0.664
11.676
0.926
11.792
27.12

Anoperationalpowermarginof15%isassumedbasedontherecommendationITTC(2005a).Auxiliaryload
wasassumedfromBuhaugetal.(2009)whoestimatetheauxiliarypowerrequirementof1000to2000TEU
container ships as one generator of985kW runningat60%utilisation equivalent to approximately 0.6MW.
However,WrtsilShipPowerR&D(2009)claimoverallsavingsofatleast10%ininstalledpowerwhenusing
electricpropulsionwhichislargerthanthetotalestimatedauxiliarypowerrequirement.Thereforethisfigure
isnotincludedinthetotalinstalledpowerestimate,althoughitcanbeseenthattheplantselectedwouldbe
abletoprovidethispowerifnecessary.

99

Concept Design of a Fast Sail Assisted Feeder Container Ship


Based on the final values of installed power in Table 6.13, plant were chosen for each concept. As the
calculated installed power is the assumed maximum power requirement, it is desirable not to oversize the
installed plant so as to reduce efficiency, yet there are obvious limitations in the number of units available
withinacertainpowerrange.ConsideringtheWrtsildualfuel50DFrange[WrtsilShipPowerTechnology
(2009a)],thefollowinginstallationsareproposed:

HullA:2x6L50DF@5.70MWand2x8L50DF@7.60MWtotalling26.6MW.

HullB:2x6L50DF@5.70MWand2x9L50DF@8.55MWtotalling28.5MW

FullspecificationsoftheseplantareincludedinFigure G.2andTable G.7.Itcanbeseenthattotalinstalled


powerforHullAisclosertothetargetvaluethanforHullB,aswellasbeinglower.Inaddition,a40tonne
savingintotalweightforthefourunitsisseenwiththeformerconfiguration.Aswellasimprovingefficiency,
theuseoffourplantinsteadoftwolargerunitsincreasesoverallsystemredundancy,althoughapenaltyispaid
intermsoftotalweightandoverallfootprintarea.
Offdesigncases
Basedontheestimatesofthrustreductionprovidedbythesailsystem(seeTable6.5),theanalysisofthemain
propulsionsystemwasrepeatedforspeedsofboth15knotsand25knots.Theseareconsideredoffdesign
casessincethepropulsionsystemhasbeenoptimisedtoprovide100%ofthepropulsivepowerrequirement
attheservicespeedof25knots.Asummaryofthepowersavingsintermsofbothpowerrequirementand
percentagereductionsisgiveninTable6.14.
Table6.14SavingsinpowerrequirementforHullsAandBbasedonannualandseasonalwindconditions,
Singaporeregion(powermarginnotincluded)
Windcondition

Nowind

Speed/knots

15

HullA
HullB

Annual

25

15

25

Seasonal

%reduction
15

25

15

25

%reduction
15

25

3.94 22.10 3.88 22.02 1.52 0.36 3.82 21.97 2.28 0.59
4.66 23.58 4.38 23.45 6.01 0.55 4.32 23.37 7.30 0.89

The percentage reduction in power requirement is larger for Hull B than Hull A. However, the power
requirementforHullBisstillhigherinallcases.ThevaluespresentedinTable6.14areusedinSection6.2.1as
partofthevoyagesimulationtoassessthetotalemissionsoftheshipduringatypicalroundtrip.
Discussion
Itmustberememberedthatthemethodusedtoestimatethepropulsiveefficiencyofbothhullformsisbased
onempiricalformulaeandstandardseriespropellercharts.Whilsttheaccuracyofthemethodemployedisnot
fullyquantifiable,itisrecognisedthattheestimatespresentedaresensitivetoboththeassumptionsthathave
beenmade,andthechoiceofformulaeandinputvaluesused.
This is particularly true for Hull B where the novel hull form means the empirical estimates of propulsive
coefficientsfortwinscrewshipsmaynotbeappropriate.However,withalackofanyotherdata,thesewere
consideredthebestestimateavailabletothedesigner.Amoredetailedinsightintothewakevelocityprofile
wouldallowimprovementsinthedesignandoptimisationofthepropulsors,yetthiswouldrequiretheuseof
specialisedequipmentsuchaswaketraverseprobes,whichwerenotavailabletotheproject.

100

Performance Predictions
6.1.3

Justificationofhullformchoice

Since the concept is required to improve efficiency and reduce environmental impact compared to existing
ships,themaincriteriaforselectingahullformistheinstalledpowerrequirement,aspresentedinTable6.13.
ThisleadstoHullAbeingchosenforfurtherdevelopment,sinceitsinstalledpowerrequirementisestimated
tobeapproximately6%lowerthanHullB.ThisisduetothetwinpodsofHullBprovidingsignificantlymore
dragthantheCRPpodpropulsionarrangementofHullA.HullAhasalsobeenshowntoperformbetterwhen
operatinginwaves,againreducingoverallfuelconsumption.
Although it has not been modelled here, the CRPpod propulsion arrangement of Hull A has an additional
advantage;whenoperatingatlowspeed,theCPPcanbedisengagedandfeatheredsoastoreducedrag,with
all the propulsive power delivered through the pod. Due to the twin pod layout of Hull B this cannot be
achieved and thus both pods must be run at low power and thus efficiency than the single pod of Hull A. It
shouldbeemphasisedthatbothconceptshavenotbeenoptimisedforlowspeed(15knots)operationsinceit
isassumedthattheinstallationofCPPsofpoddeddrivesisnotpossibleduetothehighcomplexityofsucha
system.Ifthiswerepossible,thenbothofthepropulsionarrangementsconsideredcouldbeoptimisedfortwo
speedoperation.
Whilstthischoicehasbeenbasedonthosefactorsalreadymentioned,itshouldbenotedthatHullBperforms
significantly better under sail, as discussed in Section 6.1.1. However, this thrust reduction is not large
enoughtocompensateforthedifferenceinpowerrequirementalreadydescribed.Iftheshipweredesigned
tooperateatalowerservicespeed,ormorefavourablewindconditionsweremodelled,thenthechoiceofhull
formmaybedifferent.

6.2

Design feasibility

6.2.1

Roundtripevaluation
Manoeuvrabilityinport

Animportantfeatureofthedesignisitsabilitytomanoeuvreintoaberthwithouttheuseoftugboats.This
aimstosaveonberthingcostsaswellasreducewaitingtimeswhenenteringport(seeSection2.5).Therefore
therequirementsetwasthattheshipbesuitablyequippedtobeabletomoveinpureswayandpureyawat
slow speed (three knots), and appropriate machinery be installed to achieve this. While the azimuthing
poddeddrivecanbeusedforthispurpose,abowthrustermustbespecifiedtoaugmentit.
The approach recommended by Palmer (2010) was to equate the total force and moment in sway and yaw
respectivelytotheequivalentforceandmomentprovidedbythemanoeuvringdevices.Thisrequiresassessing
boththehydrodynamicandaerodynamiccrossflowdragoftheship.Thefollowingassumptionsweremadein
thisanalysis,that:

Theshipismovinginpureswayoryawatavelocity,orangularvelocityequivalentto,threeknotswith
nocurrenteffects;

Thewindspeedistakenastheprevailingconditionfortheregionsunderinvestigationof5.3metres
persecondattenmetresabovethewatersurface.Thewindheadingisassumedtobebeamonwith
theshipmovingintothewindi.e.theworstcasescenario;

Shallowwatereffectsareneglectedandcalmwaterisassumed;

Thesailsystemisfoldedawayandthusprovideszerodrag.

101

Concept Design of a Fast Sail Assisted Feeder Container Ship


The hydrodynamic sway force and yaw moment are calculated by adapting the method of Faltinsen (1998)
whichisdesignedforassessingcurrentloads,usingastriptheoryapproach,thus:

1
2

(6.11)

and,

1
2

The change in crossflow drag coefficient (

(6.12)

) along the length of the hull was estimated from Faltinsen

(1998) and adapted based on values recommended by Hoerner (1965) to correct for the effect of increased
bilgeradius.Thehullwasdividedinto20sectionsof7.924metreslength,withthedraught
measuredat
eachfromthegeneralarrangementdrawing.
Theaerodynamicdragwascalculatedassumingtheabovewatersideonformtobeasimplerectangleof165.8
metresby19.74metres,withadragcoefficienta2.025takenbyinterpolatingvaluesfromengel&Cimbala
(2006)forrectanglesofvaryingaspectratioandwindgradientprofileaccordingtoMolland(2009).Theresults
aresummarisedinTable6.6.
Table6.15Hydrodynamicandaerodynamiccomponentsofswayforceandyawmoment
Hydrodynamic Aerodynamic
component
component
Swayforce/kN
823.7
169.3
Yawmoment/kNm
2292.4
160.9

Total

993.0
2453.2

ToequaltheforcesandmomentsestimatedinTable6.15,thefollowingprocedurewasused:
1.

Dividetherequiredthrustbetweenthebowthrusterandthepod;

2.

Calculate the actual thrust achieved and power required by the pod propeller using the method
describedinSection6.1.2.Sincethepodisoperatingat90degreestothecentreline, , and are
assumedtobezero;

3.

Select a bow thruster diameter based on manufacturers specifications [Rolls Royce (2008)] which
produces close to the required thrust, calculated using actuator disc theory [Molland (2009)].
Comparerequiredthrusttoachievedthrust.Ifdiscrepancy,modifythrustbalancebetweenpodand
thrusterandrepeat.Ifsatisfactory,compareyawmoment;

4.

Modify yaw moment by moving bow thruster position forward or aft. Check bow thruster location
againstlayoutrestrictions;

5.

Estimatetotalpowerrequirement.Forthepod,assume
0.8[Bertram(2000)].

,andbowthrusterefficiencytobe

Having achieved an agreement between required and device thrust of less than 1%, the following power
requirementswerederived:2949.0kWforthepod;and2642.6kWforthebowthruster.Thisresultsinatotal
propulsivepowerrequirementformanoeuvringof5591.7kW.Notethatthispowerrequirementcanbemet
usingoneofthe6L50DFplant,ratedat5.7MW,runningat100%MCR.
ThebowthrusterinstalledistheRollsRoyceTTCPof2.8metrediametergivingapowerof2650kW.Thissize
ofbowthrusterismorethantwicethepowerofatypicalunitinstalledonastandardfeedership,asmaybe

102

Performance Predictions
expected for such a manoeuvrable ship, yet its diameter meets the requirement to be located at least one
diameterbelowthestillwatersurfaceinordertoavoidventilation[Palmer(2010)].
Cargohandlingtime
Reducingcargohandlingtimeisalsokeytoimprovingtheefficiencyofthedesign.Therequirementistobe
abletoselfunload,andfasterthanexistingships,whichcanspendupto48hoursbeingunloadedbyquayside
cranes.Timeisalsosavedsincetheshipwillnothavetowaitforacraneequippedberthtobecomeavailable.
Itisassumedthattheshipisequippedwithtwogantrytypecranes.Informationontheoperatingspeedsof
similarcontainerterminalgantrycraneswasusedinestimatingthetotaltimeatberth(seeTable6.16).
Table6.16Summaryofdatausedforcargohandlingsimulation[Kalmar(2009)]

Trolley Gantry Hoist


1

Speed/m.min
Ave.dist.permove/m
No.ofmovespertwoTEU

90.0
28.0

270.0
26.4
2.0

90.0
20.6
4.0

Inaddition,thefollowingassumptionsweremaderegardingtheloading/unloadingpattern,that:

TheaverageloadperliftisequivalenttotwoTEU;

Eachcraneunloadsanentirebaybeforemovingtothenext,whichislocatedonaverageonequarter
ofthelengthofthecargoareaaway;

TheaveragedistancemovedintheverticaldirectionperliftisequivalenttoeightTEU;

Theaveragedistancemovedinthelateraldirectionperliftisequivalenttohalftheshipbeamplus13
metresforclearanceontothequayside;

A10%marginisassumedforconnectingthecontainer(s)tothegantry.

Basedonthevaluesandassumptionspresented,thetotalcargohandlingtimeforthefullyloadedshipof1270
TEUis25.9hoursbasedontwoloadingandunloadingsequencesandanaveragehandlingtimeof0.28minutes
perTEU.Thetotaltimereducesto23.3hoursiftheshipisonlyutilisedto90%.Inamorerealisticscenario
wheretheshipisrequiredtousethehubportsownquaysidecranes,thetotalcargohandlingtimeincreases
to28.2hoursbasedonatimeof0.36minutesperTEUatSingapore[MaritimeandPortAuthorityofSingapore
(2009)].
Voyagesimulation
Inordertoprovideanassessmentofthesuccessoftheconceptintermsofbotheconomicsandemissions,a
comparison was made with what were assumed typical existing ships operating in both the Singapore and
Caribbeanregions.BasedontheanalysispresentedinSection2.6andassumingthateachshipoftheproposed
designwouldreplacetwoexistingships,itwaspredictedthattheconceptcouldmeetpredicted2020container
tradedemandbysailingthreeroundtripsforeverytwosailedbyanexistingship.Table 6.17illustratesthis
overa14dayperiod,assumingautilisationfactorof90%forallships.
The demand in Singapore is met easily, and the utilisation factor can reduce to 84% before a shortfall in
delivered TEU is seen. This is a negligible shortfall in the Caribbean region which is assumed allowable,
althoughanincreaseinutilisationfactorof0.5%wouldseethisnegated.

103

Concept Design of a Fast Sail Assisted Feeder Container Ship


Table6.17Summaryofsimulatedvoyagetomeetfortnightlycontainerdemand
Singapore
comparison
801
2
2
168
3204
15.2

AverageTEUcapacityat90%utilisation
No.ofshipsoperatingonroute(minimumratio)
No.ofroundtripspershipperfortnight
Averagetimeperroundtrip/hours
TotalTEUcarriedto/fromspokeportperfortnight
Speed@90%MCR/knots

Caribbean
comparison
860
2
2
168
3440
17

Fast
feeder
1143
1
3
112
3429
25

Abasicvoyagesimulationwasmadeassumingtheoperationalprofileofthetypicalshiptobesimilartothat
statedbyMash (2009). Cargo handling time was estimated assuming 0.36 minutesper TEUat the hub port
[MaritimeandPortAuthorityofSingapore(2009)]and1.1minutesperTEUatthespokeport[ChittagongPort
Authority(2010);VietnamSeaportsAssociation(2010)].Thisresultsintotalcargohandlingtimeperroundtrip
of 40 hours and 45.9 hours for the Singapore and Caribbean ships respectively. Power requirement and
installedplantwereselectedusingtheequationspresentedinTable3.1andexaminationofbasisshipdata.
Theoperationalprofileofthefastfeederwascalculatedtoincludetheestimatesofcargohandlingtimeaswell
aspowersavingsduetotheuseofthesailsystematspeedsofboth15and25knots,calculatedfromSection
6.1.1 assuming typical annual wind conditions. Note that added resistance in waves of the fast feeder was
modelled directly from testing results, while no such effect was included for the comparison ships. It was
assumedthatanyslackinthescheduleoftheseshipswastakenupwaitinginportusingonlyauxiliarypower,
whilethefastfeederexperiencesnodelaysduetoweatherorwaitingtime.Notethatsincethedifferencesin
thrust reduction due to sails and distance of route between the two regions was negligible, the fast feeder
voyage was modelled identically in both regions. The speed and power profiles modelled are included as
Figure H.1andFigure H.2.Table 6.18summarisestheoverallpowerrequirementsforthefortnightlyperiod
simulated.
Table6.18Summaryofpowerrequirementsforfastfeederandcomparisonshipsoverfortnightlyperiod

Singaporecomparison
Time
ME
AE
Operating
/
power
power
Mode
hours
/kW
/kW
Cargohandling
78.0
0
591
Manoeuv./waiting 44.2
0
591
Lowspeed
53.5
5656
591
Highspeed
160.3
7272
591

Caribbeancomparison
Time
ME
AE
/
power
power
hours
/kW
/kW
91.7
0
591
51.2
0
591
48.3
6567
591
144.8
8443
591

Fastfeeder
Time
ME
AE
/
power
power
hours
/kW
/kW
77.3
0

6.7
5600

154.6
4880

97.4
22100

Notethatthecomparisonshipsareassumedtohaveapowerrequirementformanoeuvringequivalenttotheir
normalauxiliaryload.Althoughthismayseemlow,thisaccountsfortheuseoftugsinberthingoperations.
Notealso,thefastfeederhaszeropowerrequirementduringcargohandlingsincecoldironingisassumed,
wherebyshoresidepowerisprovidedtotheshipwhenberthed.
Based on the profile assumed in Table 6.18 and the values of SFC and emissions specified in Table H.1 and
Table H.2respectively,anestimateofthefuelconsumptionandemissionsofeachshipcouldbemade.Fuel
costswereestimatedusingvaluesfromTable 3.9Notethatthecomparisonshipsareassumedtobeburning
MDOfueltocomplywithMARPOLAnnexVIregulations[IMO(2005b)],whilstthefastfeederconceptusesLNG
fuel exclusively. The percentagedifference valuespresented are the savings predicted when operating the
fastfeederasopposedtothecomparisonshipineachregion.

104

Performance Predictions
Table6.19Summaryoffuelconsumptionandcost,andemissionsforfastfeederandcomparisonshipsover
fortnightlyperiod
CO2
NOx
NOx
CO2
Cost
Cost
/tonnes /%diff. /tonnes /%diff. /kUSD /%diff.
39.7
38.3
89.2
381.5
51.5
2100.3
42.2
40.0
89.7
397.7
53.4
2190.5

4.1

185.0

1266.5

FC
FC
/tonnes /%diff.
SingaporeComp.
593.3
32.9
CaribbeanComp. 618.5
35.6
FastFeeder
398.2

Table 6.19 shows that the overall GHG (CO2) emissions reduction target of 30% has been exceeded in both
regions,meaningthefastfeederisdeemedenvironmentallysustainablewithinthescopeofthisanalysis.Note
thatthisestimatedoesnotaccountforanyemissionsincurredduringbuildorscrappageandintheproduction
andtransportofbunkerfueltoitspointofdeliverytotheship.Themajorityofthissavinginemissionsisdue
totheuseofLNGfuelandreducingthenumberoffortnightlysailingsfromfourtothree.Itcanalsobeseen
thatthereductioninNOxemissionsestimatedisinlinewiththevaluespresentedinTable 3.8.Thismeansthe
ship complies with MARPOL Annex VI Regulation 13. In addition, Regulation 14 on the Annex, limiting SOx
emissionsto4.5%,iseasilymetsinceLNGfuelhaszerosulphurcontent.Buhaugetal.(2009)notethatmost
shipscurrentlyalreadycomplywithRegulation14,whichcomesintoforcein2020.However,thereiscurrently
littleincentiveforshipownerstoreducesulphuroxideemissionsfurtherusingdieselfuels,evenafterthisdate.
Thus the fast feeder concept is considered highly environmentally sustainable in this respect, removing the
shipscontributiontoacidraincompletely.
Theperformanceofthefastfeederwasalsojudgedagainstthecomparisonshipsusingperformanceindices:
thetransportefficiencyindex(TEI)andtheIMOEnergyEfficiencyDesignIndex(EEDI).Modificationshavebeen
madetotheformulationsofthesemeasurestoaccountforthevoyageprofilesgiveninFigure H.1andFigure
H.2.
Table6.20Summaryoffastfeederperformanceagainstcomparisonshipsusingdesignindices.

TEI

modifiedTEI

TEUknots/kW

Singaporecomparison 6.88
Caribbeancomparison 7.12
Fastfeeder
4.48

3.44
3.56
5.61

percentage
percentage
EEDI modifiedEEDI
improvement
improvement
gCO2/tonnenm
(modifiedEEDI)
(modifiedTEI)
63
24.15
56
27.05
58
22.63
62
31.41

14.82

11.84

Table6.20showsthatunlessthedecreaseinthenumberofshipsisaccountedfortheTEIforthefastfeederis
notfavourable.However,sincethefastfeederhasbeendesignedtoreplaceafeederservice,ratherthana
specificship,itisreasonabletoincludethismodificationtotheTEI.Theenvironmentalperformanceofthefast
feederintermsoftheEEDIisbetterthanthatofthecomparisonshipsevenwithoutthemodificationsmadeto
the index. Note that the percentage improvement in EEDI is significantly larger in each case than the
percentagereductionsinCO2presentedinTable 6.19.Thisisbecausetheincreaseinspeedofthefastfeeder
servestoimprovetheindexvalue,sincethecalculationismassofCO2perunitspeed,asopposedtopurely
massofCO2emitted.
UsingcostdatasuppliedbyOceanShippingConsultantsLtd(2010)anestimateofthetotaldailycostsofeach
of the ships could be made in order to assess the economic viability of the proposed concept. The values
presentedarelinearlyinterpolatedtotheTEUcapacityofeachshipusingthedataprovided,andarebasedon
ownershipcostsratherthancharter.ThedailycostspresentedinTable 6.21representtwocomparisonships
againstonefastfeeder.Dailycapitalchargeisbasedonnewbuildpricesof21.85and23.45millionUSdollars
respectively for the Singapore and Caribbean comparison ships. The newbuild price of the fast feeder was

105

Concept Design of a Fast Sail Assisted Feeder Container Ship


interpolatedfromthedataprovided,butafactorwasincludedtoaccountfortheincreaseincapitalcostdueto
thenovelsystemsinstalledontheship,includingthepoddeddrive,LNGfueltankandretractablesailsystem.
Thusatotalbuildcostof45millionUSdollarswasused,sincemoreaccurateestimatesoftheadditionalcosts
werenotavailable.
Table6.21Summaryofdailycostestimatescomparingfastfeederconcepttotwotypicalexistingships
SingaporeComparison CaribbeanComparison
(twoships)
(twoships)
dailycapitalcharge
18016
19332
manning
2226
2396
repairandmaintenance
976
1048
insurance
582
626
admin/other
890
956
fuel
27250
28407
totalcost
49940
52765
Cost/USDperday

FastFeeder
18623
1588
696
416
635
13214
35172

The daily cost saving based on a typical fortnightly period is 29.6% and 33.3% in case of the Singapore and
Caribbeanregionsrespectively.Thisrepresentsasignificantcostsavingtotheowner,especiallyconsidering
thisestimateincludesanestimateoftheadditionalconstructioncostsassociatedwiththefastfeederdesign.
Based on this analysis, the fast feeder will only become uneconomical if the price of LNG fuel more than
doubles,toUSD958andUSD1000pertonnerespectivelyinSingaporeandCaribbeanregions.

6.2.2

EconomicfeasibilityofMultiwingsystem

ThissectionprovidesanestimateofthecapitalandoperatingcostsfortheMultiwingsailsystem;asimpleNet
PresentValue(NPV)analysisisalsocarriedouttoassesstheprofitabilityofusingthesailsovertheassumed
timespanafterwhichthefeederissold.Bergeson&Greenwald(1985)summarisesbuildingandrunningcosts
of different rig configuration including a rigid wing sail of surface area 280m2, the cost of which is scaled
accordingtothesailarea.Toobtainthecostsatthepresentday,anannualnationalinflationrateistakenas
2.91%[FederalReserve(2010)].Thefiguresforbuildingcostsareavailableforasingleprototyperig,arigout
ofaproductionrunoftenandoneoutof100,producedinU.S.Inthefollowinganalysisthecostofonemodel
outof10isusedasareference.
Table6.22BuildingandrunningcostfortwoMultiwingrigs
Capitalinvestment. Annualmaintenance
Costin1985/kUSD

881

10.6

Costin2010/kUSD

1806

21.6

Cost(including50%marginforhoistingsystem)/kUSD

2709

32.3

Asthedatareflectsthecostofawingsailfittedondeckandtherequiredmachinerytooperateit,afactorof
1.5isincludedtoallowfortheextracostofthehoistingsystem,consistingmainlyofthegeneratorsrequiredto
powerthelift.
The annualbenefits due to the use of thesails are calculated from the associated power reductions derived
using Winpow. As the reductions for Singapore and the Caribbean are very similar, only Singapore is
considered in this economic analysis. Table 6.23 shows the calculated benefits; the breakdown between
annualandseasonalreflectsdifferentroutesandwindconditionsasexplainedinSection 6.1.1.Thefollowing
assumptionshavebeenmade:

106

Performance Predictions

71roundtripsperyear;

SFCof430g.(kWh)1;

CostofMDOtobe575USD.ton1[Hinrichs(2009)];

Fullloadconditiononly.

Table6.23AnnualbenefitsfromtheuseoftheMultiwing
HullA
Annual
Seasonal
Reduction/kW
60
80
90
130
Hoursperroundtrip
46
36
46
36
Fuelsaved/t
1.2
1.2
1.8
2.0
Savingsperroundtrip/USD 681 709 1022 1152
Annualbenefit/kUSD
48.36 50.33 72.55 81.79
TotalBenefit/kUSD
99
154

HullB
Annual
280
130
46
36
5.5
2.0
3179 1152
225.71 81.79
308

Seasonal
340
210
46
36
6.7
3.2
3860 1861
274.07 132.12
406

NetPresentValue(NPV)analysis
The Net Present Value appraisal method is used to verify whether the operation of the Multiwing would
generateaprofitattheendofanassumedprojecttimespanof15years.Thecalculationwillbeundertaken
consideringtheMultiwingsystemasaseparateentity,asifitwastoberetrofittedonanexistingvessel.NPV
isequaltothedifferencebetweenthenetglobalbenefitandtheinvestmentcostdiscountedonthesamebase
[Mange (2006)]. In this analysis, the rate of return is the interest that could be earned on an alternative
investmentwithsimilarrisk.Itisessentialtoselectasensiblevalueasthiswilldeterminehowthediscounted
cashflowiscalculated.Ahypotheticalsituationcanbeassumedwherealargeshippingcompanycommission
the building of ten fast feeders, the interest rate the company would receive is the corporate bond rate of
5.31%[FederalReserve(2010)].TheInternalRateofReturn(IRR)isthediscountratewhichmakestheNPV
equal to zero (the breakeven point), and reflects the risk involved in the project [Mange (2006)]. The
calculationisbasedonthecostsandannualbenefitsestimatedintheprevioussectionanditreliesonthesame
assumptions.
Table6.24NPVanalysisforHullA.operatinginSingaporeregion
Annual Seasonal
NPV
2034
1469
Paybackperiod/years 27.44
17.5

Table 6.24showsanegativeNPVforHullA,withapaybackperiodwellabovetheassumedprojecttimespan.
A sensitivity analysis on the discount rate used and fuel prices confirms this view. Table 6.25 shows the
projectionsforthesecondhullform.Inthiscasetheannualsimulationresultsjustabovebreakeven,whilst
theseasonalschedulewillgeneratesomeprofit.HowevertheIRR,foundtobe1%,reflectsthelowprofitability
ofthisinvestment.
Table6.25NPVanalysisforHullB.Singapore
Annual Seasonal
NPV
88
1092
IRR 0.057
0.095

AsensitivityanalysisbasedonHullBfortheseasonalschedule,isalsoincludedinTable 6.26toimprovethe
economicmodel.Notethatthefuelcostisafundamentalfactorasitdeterminesthemagnitudeoftheannual
benefitsfromthesails.Whilefuelpricesareexpectedtogrowby2020,thusincreasingtheprofitability,there
arenomeansofestimatingthepricefluctuationsoversuchatimespan.

107

Concept Design of a Fast Sail Assisted Feeder Container Ship


Table6.26NPVsensitivityanalysisforHullB,seasonalschedule

375
475
Fuelcosts/kUSD 575
675
775

3.31%
6
831
1656
2481
3305

Rateofreturn
4.31% 5.31% 6.31% 7.31%
178 345 495 631
591
374
177
0
1359 1092 850
631
2128 1810 1522 1262
2897 2528 2195 1893

Theeconomicassessmentsconductedsofarhaveonlyincludedthebenefitsderivedfromthethrustreduction
generated by the sails. The reduction achieved at a ship speed of 25 knots has proven to be too small to
generate enough return on the investment. Section 7.1.7 outlines a procedure from which the thrust
reductions derived from the motion damping due to the sails is estimated. This reduction has not been
includedinthemaineconomicassessmentasthemethodusedassumesaseriesofparameterswhicharenot
directlyrelatedtothefastfeederhullform.Nonetheless,thedesignersareoftheopinionthatifthispotential
dragreductionwasfurtherinvestigatedandproven,theuseofsailsatthisspeedwouldbereadilyjustifiedand
considerablesavingswouldbegenerated.Table6.27showstheNPVanalysisbasedonannualbenefitsderived
fromwindpropulsionandmotiondamping.Theannualbenefitsfrommotiondampingarefoundcalculating
the fuel saved when sailing in quartering seas. The thrust reductions are weighted with the probability of
quarteringwindsgivingatotalreductionof2.14%and1.93%at25and15knotsrespectively.

Table6.27NPVanalysisincludingthrustreductionsduetomotiondamping.Theassumedrateofreturnis
5.31%

NPV

Seasonal
Annual
HullA HullB HullA HullB

1736 2302 3973 4976

TheapplicationoftheWingsailisnowaprofitableinvestment.Sincethisanalysisaimtoassesstheabilityof
sailstogenerateprofit,excludingthebenefitsderivedbythereductionsinemissions,thepenaltyarisingfrom
thedecreaseincontainercapacityduetothefoldablesystemmustalsobeconsidered.88TEUswhicharelost
toaccommodatethesails;usinganestimatedfreightrate(AppendixH.2)of4316USDitispossibletocalculate
the annual loss assuming a 60% utilisation; the annual figure amounts to 13.48 million USD, well above any
projectedprofitfromthesails.

108

Design Development

7. Design Development
7.1

Seakeeping analysis

Aseakeepinganalysishasbeencompletedtoestimatetheextentofthemotionsexperiencedbytheship.This
willaffecttheoperabilityoftheshipintermsofsafetyandcomfortofcrewandregardingstructuralintegrity,
duetopossiblehighslammingloads.Itistheformerreasonthatisofprimaryconcern.Thesuperstructureof
theshipissituatedattheforwardend,whereitislikelythatlargeaccelerationswillbeexperienced.Thisis
wherethecrewspendthemajorityoftheirtime.Alongsidethisaccelerationsareevaluatedatthesailmast
locations to aid structural analysis (Section 7.2.3) and where the most vulnerable containers are located to
allowanassessmentofcargolashingloads(Section7.1.5).

7.1.1

Modelling

TheanalysishasbeencarriedoutusingStripTheoryimplementedusingthesoftwareSeakeeper.StripTheory
is commonly used in Naval Architecture to assess a ships motions. It essentially simplifies the ship by
representing it as a number of two dimensional transverse sections (21 have been used in this case). The
hydrodynamiccoefficientsaredeterminedatthesesectionsusingconformalmappingandintegrationiscarried
outalongtheshipslengthtogivetheglobalhydrodynamiccoefficientsformingtheequationsofmotion.A
detailedexplanationofStripTheoryisgivenbyLewis(1989).Asthismethodmakesasignificantsimplification
oftheshipshullformthereareclearlylimitationstoitsapplicability.BetterresultsareobtainedatlowFroude
numbershowevergoodcorrelationisclaimedupto
0.53[FormationDesignSystems(2005b)],butthe
qualityofresultsisalsodependantonhullformtype.
The hull is defined by a number of geometric sections determined from the surface definition in Maxsurf.
Lewis sections are mapped to these for use in the analysis. Two other main inputs are details of the mass
distributionandthedefinitionofseaspectra.Themassdistributionoftheshipisdefinedbyaverticalcentreof
gravityandapitchandrollradiusofgyration.TheVCGistakenfromthemassestimateincludingafactorfor
freesurfaceeffects,10.377metresabovethekeel.Thepitchradiusofgyrationistakenas25%oftheoverall
lengthandtherollas35%ofthebeam.Thesearesimplyestimatesbasedoncommonpracticeandexpected
values [Loukakis & Chryssostomidis (1975)]. The seaway is represented by the ITTC or Two Parameter
Bretschneider spectrum [Formation Design Systems (2005b)]. This requires the user to input characteristic
waveheightandzerocrossingperiod.Thecharacteristicwaveheightissettoonemetre,asaccelerationscan
be scaled linearly for the desired wave height, and the zero crossing periods have been defined from 3.5 to
11.5 seconds in intervals of one second. This range has been chosen in consideration of the annual wave
statisticsintheproposedoperatingareas[Hogben(1986)].

7.1.2

Results

Accelerationsaretobepresentedasandwhennecessarytocarryoutanyfurtheranalysis.Inthissectionthe
responses in heave, roll and pitch will be discussed. These are best represented byRAOs which are plotted
againstwavefrequencyinFigureI.1toFigureI.3forboth15and25knotsshipspeed.
StudyingFigureI.2onecanseeverylargerollresponsemotions.At25knotsthemaximumrolloccursinbeam
seas,seeminglyreasonable.At15knotslargerollmotionsarepredictedinsternquarteringseas.Lewis(1989)
statesthatrollmotionsareworstinlargeshipsinquarteringseaswhenunderway,makingthisareasonable
prediction.Themagnitudesoftherollmotionhoweverseemexcessive.ThemaximumRAOat25knotsis4.25
and for 15 knots is 7.0. These represent large magnitude motions that, especially at 15 knots, would make

109

Concept Design of a Fast Sail Assisted Feeder Container Ship


operation impossible if not capsize the ship, depending on the wave environment. At 15 knots in stern
quarteringseasmultiplepeaksareobservedsuggestingresonantfrequenciesthatareveryclosetogether.This
does not seem likely. Less is known about transverse motions and roll is the most difficult to predict being
strongly coupled with sway and yaw [Lewis (1989)]. Furthermore Strip Theory is not able to predict roll
motionsofthisamplitudeastheywillcertainlybeoutsidetheassumptionsofthelinearanalysis.Evaluating
such motions would require time domain simulations. It is possible that Seakeeper is over predicting the
magnitudeoftherollmotions.Peaksintherollresponseareobservedathighfrequencyinsternquartering
seas. Lewis (1989) (p79) states that severe roll can occur in lighter seas with shorter period if the ship is
underwayandtheseasandcomingfromthequarterwhichcoincideswithwhatisseenfromtheresults.In
addition, for a ship with roll damping less than 5%, as in this case (see Section 7.1.7 for naked hull roll
damping),itisnotunusualtoobservepeakrollRAOsgreaterthanten[Lewis(1989)].Onthisvein,itispossible
thattherollpredictionisnotinaccuratebuttheperformanceofthehullispoor.Thisseakeepinganalysisdoes
notaccountforanyadditionaldampingfromthesails.Section7.1.7estimatesthisincreaseindamping,which
significantly improves the roll motions potentially eliminating the need for any additional stabilising
mechanisms.
Pitchmotionsseemreasonablebothintermsofmagnitudeandtheeffectofwaveheading.Atbothspeedsin
headandfollowingseastheresponseissimilarinmagnitude(peaksatapproximately1.2)asitisinsternand
bow quartering seas (peaks at approximately 1.0) and the magnitude of the response in beam seas is
significantlylower(peakatapproximately0.6).Thisbehaviourisexpected.
At 25 knots there are large heave motion peaks in head (1.4), bow quartering (1.5) and beam seas (1.25).
These results are larger than expected, especially in bow quartering and beam seas. It is possible that this
couldbeduetopushingthelimitsofStripTheorysapplicabilityathighFroudenumberasthesameproblems
inheadandbowquarteringseasarenotobservedat15knotsalthoughthelargeamplitudemotionsinbeam
seasexistalmostidentically.TheseproblemscouldberesultingfromthepoorfittingoftheLewissectionsin
theforwardandaftregionsoftheship.Lewissectionshavenotbeenabletofitthehighcurvatureshapeat
thebowandstern,hencethesectionsarenotrepresentingtheformaccurately.ThiscanbeseeninFigure7.1
andisoccurringduetoLewisformsbeingrestrictedtobehorizontalatthekeelandverticalatthewaterline.
This has resulted in changes in the volume distribution that affects the magnitude of the motions at all
headings.Atbothspeedstheresponseforfollowingandsternquarteringseasdecaysquicklyfromavalueof
onebutthenincreasesandpeaksagainathigherfrequency.

Figure7.1Mappingof11LewissectionstoHullA(Shipsectionswhite;Lewisformsgreen)

110

Design Development
7.1.3

Theoreticaladdedresistance

Atheoreticalpredictionoftheaddedresistancehasbeencomputedinheadseastoprovideavalidationand
comparisontotheresultspresentedinSection 4.4.3.ForthisreasontheanalysisiscarriedoutforbothHullA
andHullB,seakeepingonlyconsideredHullA.Seakeeperprovidesthreeanalysismethodsfordeterminingthe
added resistance in waves; two methods by Gerritsma & Beukelman (1972) and one by Salvesen (1978).
Salvesen(1978)providesacomparisonoftheproposedmethodagainstthemethodofGerritsma&Beukelman
0.49 formoperatingat
0.35.Salvesen
(1972)andresultsfromexperimentsforadestroyerhull
(1978)(Figure 7.2andFigure 7.3)showsthattheproposedmethodgivesresultsthatarealotclosertothose
from the experiments than the results obtained by the method of Gerritsma & Beukelman (1972). This
destroyer form is not significantly different to the proposed hull forms and hence this method is employed.
This method is defined by Wilson (1985) as a hull pressure method and uses a distribution of sources and
doublets along the centreline of the ship to represent the flow around the hull. The velocity potential has
threecomponentsduetothe

incomingwavefield;

thesingularitiesincalmwater;

andthewavesystemofthesingularities.

Theaddedresistanceishencegivenbythesumofsixcomponentsduetothe

pureheavingoftheship;

thepurepitchingoftheship;

thewavediffractionoffthehull;

thecouplingbetweenheaveandpitch;

thecouplingbetweenheaveandthereflectedwave;

andthecouplingbetweenpitchandthereflectedwave.

Exp 15.912kts

Exp 25.497kts
Num 15.912kts

Num 25.497kts

AW

4
3
2
1
0
0

50

100

/LWL /%150

200

250

Figure7.2ComparisonofnumericalandexperimentaladdedresistanceHullA

111

Concept Design of a Fast Sail Assisted Feeder Container Ship


7
Exp 16.005kts
6

Exp 25.649kts
Num 16.005kts

Num 25.649kts

AW

4
3
2
1
0
0

50

100

/LWL /% 150

200

250

Figure7.3ComparisonofnumericalandexperimentaladdedresistanceHullB

Figure 7.2 and Figure 7.3 illustrate a comparison of the numerical and experimental solutions for Hull A and
HullBrespectively.Theinputspeedsforthenumericalanalysismatchthoseofthetestedruns.Theresultsare
relativelysimilar.However,generallytheresultsfromthenumericalmethodarelarger.TheresultsforHullB
matchbetter.AsnotedinSection 4.4.3thepeaksoftheaddedresistanceintheexperimentaldatahavenot
beencaptured.Onecannotthenbesurehowthemagnitudesofthepeaksfromthetestswouldcompareto
thenumericaldata.Itisprobablethatthepeakswillcoincideintermsofwhatwavelengththeyoccurat.The
trendinthepeaklocationasspeedincreasesasmentionedinSection 4.4.3canbeseenclearlyinbothsetsof
numericaldata.Thematchbetweenthecorrelationoftheexperimentalandnumericaldataatwavelengths
lessthanthepeakaregoodhoweveratwavelengthsgreaterthanthepeakthecorrelationisnotsogood.The
numericaldatadoesnotshowasharpdecreaseinaddedresistanceathigherwavelengthsthanthepeakadded
resistance. The experimental data shows this trend which is somewhat expected, from studying examples
presentedbySalvesen(1978).
The differences discussed arise from a combination of errors and limitations in both methods. The
determination of the numerical added resistance is strongly dependant on the heave and pitch motions
[Wilson(1985)]andthusthewaytheyaredeterminedcanhaveagreateffectontheresults.Althoughmost
computationsdetermineshipmotionsusingStripTheorytheaccuracyofthesolutionwillhavesevereeffecton
theaccuracyoftheaddedresistance.StromTejsenetal.(1973)hasexploredthissubjectandfoundthat,in
somecases,vastdifferencesintheaddedresistanceoccur;especiallynearthepeakvalue.Thisislikelytobea
particularprobleminthenumericalanalysiscarriedoutinthisthesisasunexpectedresultshavebeenobserved
fortheheaveRAOsinSection7.1.2.Alltheforcescomprisingthetotaladdedresistancearedependentonthe
waveamplitudesquaredandhence,unlikethemotions,thesolutionisnonlinear.Thismeansthatanyerrors
incurred in the numerical or experimental analysis are also nonlinear and hence their sources are more
difficult to identifyand their magnitudes larger. There is also the question ofthe comparabilityofthepitch
gyradiususedineachcase.InthenumericalanalysisthepitchgyradiushasbeenusedasdefinedinSection
7.1.1.Intheexperimentsthemodelwasballastedinanattempttomatchthishoweverduetorestrictionsin
timethemodelwasnotswungwhichistheonlyreliablemethodindeterminingthemodellightship(model
withoutballastweightsandfittings)gyradius.Anysmalldifferencesinthegyradiiwillaffectthemotionsand
duetotheaddedresistancesstrongdependenceonthisandthenonlinearnatureoftheproblemcouldpose
largediscrepanciesintheresults.

112

Design Development
7.1.4

Absolutemotions

The Hull A roll, pitch and heave RAOs where used with a Bretschnider wave spectrum as part of a spectral
analysistocalculatethemaximumroll,pitchandheavemotionsfora7.5msignificantwaveheightat15and25
knotsforallwaveheadings.Therollandpitchangleareimportantinthedeterminationofforcesduetoship
motion for both the container securing analysis (section 7.1.5) and also the finite element analysis (Section
7.2.3).Asummaryoftheheave,pitchandrollresponseinvariousseastatesisgiveninTable7.1andTable7.2.
The highlighted values correspond to the maximum significant motion and seem reasonable and within the
limitsofempiricalvaluesgivenwithLloydsrulesformaximumpitchangle(notgreaterthan8o)androllangles
(between22oand30o)[Lloyd'sRegister(2009)].
Table7.1Summaryofabsolutemaximummotionsat25knotsforHullA

m0/m2
Heave
sig.motion/m
m0/2
Roll
sig.motion/
m0/2
Pitch
sig.motion/

Waveheading/deg
135
90
45
1.796 1.531 1.057
5.361 4.950 4.113
0.133 7.676 4.859
1.460 11.082 8.817
0.876 0.407 1.971
3.744 2.552 5.615

180
1.097
4.190
0.000
0.000
0.802
3.581

0
0.779
3.530
0.000
0.000
1.835
5.419

Table7.2Summaryofabsolutemaximummotionsat15knotsforHullA

m0/m2
Heave
sig.motion/m
m0/2
Roll
sig.motion/
m0/2
Pitch
sig.motion/

180
0.671
3.277
0.000
0.000
0.810
3.601

Waveheading/deg
135
90
45
1.189 1.947 0.414
4.362 5.581 2.574
0.408 7.676 52.852
2.553 11.082 29.080
0.844 0.329 0.669
3.675 2.296 3.273

0
0.187
1.732
0.000
0.000
0.717
3.387

7.1.5

Cargosecuring

Itisestimatedthateveryyear10,000containersarelostfromcontainerships[Podsada(2001);Frankel(2002)]
whichraisesenvironmentalandsafetyconcerns.Inaddition,thereisaneconomiccosttotheoperatorarising
fromthelossofcargo.
Throughoutthedesignprocessthemainfocushasbeenontheoptimisationofhydrodynamicresistancewith
little attention focussed on the ships seakeeping characteristics. Since the proposed layout is somewhat
differenttotraditionalcontainershipsananalysiswasconductedtodeterminetheabilityofthecargotoresist
forces due to static gravity forces; inertial forces generated by accelerations due to roll, pitch and heave
motionsoftheship;windforces(takenatawindspeedof40metrespersecondacrossthebeam)andforces
imposedbythecargosecuringarrangements
The basis of the method is prescribed by Lloyd's Register (2009) Part 3 Chapter 14; Regulations for Cargo
Securing Arrangements which the author has modified to account for different stacking heights and lashing
arrangements.
Theanalysisisconductedbasedonthemostseverecombinationoftheforces,wheretheresultantforceacting
on a container at tier ( ) is the vectorial summation of the individual directional components of all forces
actingatagiveninstant.

113

Concept Design of a Fast


F
Sail Asssisted Feedeer Container Ship
A numerical finite elemeent model waas created using a spreadssheet which replaced a sttack of up to
o eight
w up to 16
6 elements (d
depending on
n the stack size being considered) whe
ere the
containers and lashings with
7
stiffnessofttheelements wasrelatedttotheequivalentstiffness ofthecontainerandlashiings(Figure 7.4).A
global stiffneess matrix forr the containeer stack and an equilibrium
m condition for
f the system
m was then derived
d
(Equation(7.1)).UsingfforcescalculattedusingLloyyd'sRegister (2009)ateacchtierinthe containerstaackthe
ner stack at each
e
tier; the tension in the lashings ( ); residual forces transmitted
deflection off the contain
laterally thro
ough the staack (( ,
cos
), where is the lashing angle); and the add
ditional
verticalforceesduethepreesenceofthelashingscould
dbedetermin
ned.

Figure7.4
4Modellinggassumptionsrequiredfortheanalysiso
ofthecargoseecuringarranggements[Lloyyd's
R
Register(2009
9),Part3Chap
pter14]

0
0

0
0
0

0
0
0

0
0

(7.1)

betweentierssofcontainerrstoresistne
egative
Themodelaassumesthat thesuitablellockingdeviceesarefittedb
(separation)forcesandstacksofcontainerswithinthecargohold
dareconnecteedtogetherw
withdoublesttacking
orceshaveno
otbeentransfe
erredfromneeighbouringsttacksthrough
hthese
cones[MacGregor(2010)],however,fo
connections..
This last assumption lead
ds to an undeerestimate off the racking force in the lower tiers b
but will still provide
p
enassumedth
hatforceson thecontainerstack
correctrepreesentationof thecompresssiveforces.Itthasalsobee
willbegreatterinthetran
nsversedirecttionthanlonggitudinallyand
dthusfailureinthisplaneeisthefocus ofthis
investigation
n.
Sixcontainerrstackswereconsideredasspartoftheaanalysis,whossepositionsontheshipareeindicatedinFigure
7.5.Thestacckswereselectedtoberepresentative ofthestacks whichareexp
pectedtoexp
periencetheh
highest
loadingduettotheirdistan
ncefromtheccentreofflotaation,centrelineandrollceentreandalso
ostackingheigght.

114

Design Deevelopmentt

Figure7.5Containerstacksconsiderredincontaineersecuringan
nalysis
For each of the sixx stacks the transverse
t
an
nd longitudinaal sliding forcces and verticcal forces imp
posed on thee
uetothemottionofthesh
hip11wherecaalculatedfor sixseparatem
motioncondittions,namelyy
containerstacksdu
withttheshipexperriencinga:
1
1.

2
2.

3
3.

Rollingmo
otionwith
o maximumroll
m
motion(descending)andm
maximumheaavemotion(deescending);
o maximumroll
m
motion(ascending)andmaaximumheavemotion(asccending);
Pitchingco
onditionwith
o maximumpitc
m
dmaximumheeavemotion((descending);
hmotion(desscending)and
o maximumpitc
m
hmotion(asccending)andm
maximumheaavemotion(ascending);
Combined
dconditionwith
o 0.71x[maximu
0
umrollmotio
on(descendingg)andmaximumpitchmottion(descendiing)];
o 0.71x[maximu
0
umrollmotio
on(ascending))andmaximumpitchmotio
on(ascendingg)].

Them
maximumtenssioninthelasshings,rackinggforce,comp
pressiveforceinthecornerpostsandsh
hearforcesin
n
the sttack for each of the six motion
m
conditiions at each tier
t were com
mpared to acceptance critteria given byy
AmericanBureauo
ofShipping(19
988)andLloyd
d'sRegister(2
2009)andwhiichissummarrisedinTable7.3.
Tab
ble7.3Acceeptancecriteriiaforcontaineersecuring

MaximumLoad
M
d/tonnes
TEU
FEU

oninlashings
Maxximumtensio
Racckingforceoncontainerend
Verrticalforcesatteachbottom
mcorner,tensiion
Verrticalforcesatteachcornerpost,compression

12
15
25
86.4

12
15
25
86.4

The container
c
secu
uring arrangements for thee stacks considered were chosen to bee representative of currentt
containersecuring arrangementtsusedonco
ontainerships[ClassNK(20
008)]andissh
howndiagram
mmaticallyforr
oninFigure7..6.
thesixxstacksunderconsideratio


11
Bassedonthemaaximumrollan
ngleandpitch
hanglescalcullatedfromtheeseakeepinganalysis(seeSSection7.1).

115

Concept Design of a Fast


F
Sail Asssisted Feedeer Container Ship

Figure7.6Diagramm
7
maticrepresentationoftheecargosecuringarrangemeents[Lloyd'sR
Register(2009
9)]
Lashingtensions,rackingfforcesandmaaximumcomp
pressiveforcesforthesixcontainerstacksbeingconsidered
v
given in Table 7.3 assuming
a
the containers all weigh the design
were checkeed against thee allowable values
nominalweigghtof9.3ton
nnesperTEU.Itwasfound
dthatallcrite
eriaweremettexceptforth
hreestacks,n
namely
thetwostaccksinholdsixxandstacktw
woinholdsevventhatfailed
dincompresssionatthelowesttierinaatleast
onemotionccondition,asssummarisedinTable7.4.
Table7.4Maximum
mcompressiveeforceintieroneoftheconsideredconttainerstacksaandthemotio
on
ntocausefailu
ure
condition

Maxcom
mpressiveforcce/tonnes
86.31
Stackk1
Hold8
H
Stackk2
85.90
81.17
Stackk1
H
Hold7
Stackk2
88.37
86.78
Stackk1
H
Hold6
Stackk2
93.48/87.96
6

Failluremotion

Combinedtop
C
p
Combinedtop
C
pofmotions
Combinedtop
C
p/rolltopofm
motions

um mean con
ntainer load to
t cause failure of the con
ntainer
A more usefful analysis is to determinee the maximu
stackinanyo
ofthefailurecriteria.Theresultsfromtthisinvestigationaresumm
marisedinTable7.5.
7
containermasssthatcausessfailureofaccontainerstack
Table7.5Averagec

1
Stack1
Stack2
2
1
Stack1
Stack2
2
1
Stack1
Stack2
2

Hold8
Hold7
Hold6

Maximum
mContainer
Mass/ttonnesTEU1
1
10.8
1
11.0
9.8
9.0
9.2
8.5

Failure
Criteria
Compressio
on
Compressio
on
Compressio
on
Compressio
on
Compressio
on
Compressio
on

116

FaiilureMotion
Combineedtopofmotions
Combineedtopofmotions
Combineedtopofmotions
Combineedtopofmotions
Combineedtopofmotions
Combineedtopofmotions

Failuree
Tier
1
1
1
1
1
1

Design Development
Table 7.4andTable 7.5showsthattherearestructuralconcernswiththecontainerstacksprincipallyinlower
tiersofthehighercontainerstackswhicharesusceptibletocompressivefailure(buckling).Thereareanumber
ofsolutionstothisproblemincludingtheuseofhatchcovers;thecarefulloadingofcontainerstoensureonly
lightcontainersareusedinsusceptiblestacksandplacingheavycontainersatthebottomofastack.Careful
cargoloadingalsohasbenefitsinreducingrackingandshearforceswithincontainerstackswhichmaybecome
importantifanonhomogeneouscontainermasswithinastackisconsidered.
If hatchcovers were used for the second stack in the hold six it would be possible to increase themaximum
nominalcontainermassto15.4tonnesperTEUwithinthecargoholdand14.7tonnesperTEUondeck,which
isasignificantimprovementonthe8.5tonnesperTEUwithouthatchcovers.
As mentionedSection 07.1.2, thequality of the roll RAOgeneratedbySeakeeper is somewhat questionable.
Sinceforcesoncontainerstacksaredependentonrollangleitispossiblethatsomeofthehighforcescanbe
attributedtotheinaccuraciesintheseekeepinganalysis.Inaddition,anypossiblebeneficialinfluenceofthe
sailsonrollmotionhasnotbeenconsideredwhendeterminingtheforcesonthecontainerstacks.

7.1.6

Motions

OneofthedesignparameterslaidoutinSection2.5isthatafeedercontainershipmustbeabletokeepspeed
inseastatesuptoforcefive.Itisthusnecessarytodetermineatwhatpointtheshipwillhavetoreducespeed
inordertoensurethecrewssafetyandabilitytofunction.
Ashipunderwayinaseawaywilltypically,duetothephaselagbetweenheaveandpitchmotions,showthe
lowestmotionsatalocation30%ofthewaterlinefromthestern,where,themotionwillbe3050%lessthan
other parts of the ship. This is one of the deciding factors for conventional container ships having an
accommodation block located near to this point [Keuning et al. (2001)]. A key concern with a forward
accommodation block position is the effect of the ship motions on the comfort and ability of the crew to
perform their duties as the motions will be greater than the motions within a conventionally located
accommodation block. There are several measures that can be used to assess the influence of motions on
crewcomfortandabilitytofunction,namelythe:

MotionSicknessIndex(MSI)whichpredictsthepercentageofapeoplewhowillfeelseasickinatwo
hourperiodofsustainedverticalmotionsofgivenmagnitudeandfrequency;

Subjective Magnitude (SM) quantifies how severe a motion feels relative to a reference motion of
0.6gat1.0Hz;

Motion Induced Interruption (MII) indicated the number of times a crew member will have to stop
worktoholdontoasuitableanchoragetopreventlossofbalanceduetoslidingortipping;

Probabilityofslamming;

Probabilityofdeckwetness;

Probabilityofexceedingacertainacceleration(e.g.0.6gatthebow).

To assess the safety and ability of the crew to function the absolute and relative values of acceleration at
various positions around the accommodation block were investigated and used to estimate the limiting sea
stateandspeedcombinationsusingthecrewperformancemeasureslistedabove.Thepositionsinvestigated
were;
1.
2.

thebridgedeckatcentreline
thebridgedeckonwing

117

Concept Design of a Fast Sail Assisted Feeder Container Ship


3.
4.

theofficerslounge
themooringdeck

TheaccelerationswereevaluatedforHullAatspeedsof10,15,20and25knotsforallwaveheadingsfrom
headseastofollowingseasinincrementsof45degrees.Theanalysiswasconductedforallcombinationsof
waveheightandperiodthatwereexpectedtooccurintheoperationalseaareas[Hogben(1986)],assumingno
timeisspentinshelteredwaterclosetoislandsorbanks.
TheMSIisapoorindicatorofperformancefortheshipbecauseseasicknessisdependentonmorethanjust
verticalmotionsastherearesecondaryfactorssuchassmellsthatcontributetothefeelingofseasickness.In
addition,theMSIisderivedfromasmallsampleofmalestudenttestsubjectsandthusisnotrepresentativeof
crewmemberswhoareaccustomedtoworkingatseaandwhowillnotsufferundulyfromseasickness[Lloyd
(1989)].Forthesereasonstheresultsofthisanalysishavenotbeenpresentedinthisreport.
AmuchmorerelevantmeasureofmotionsisSM,whichquantifiesamotionintofivecategories(Table7.6).
Table7.6Subjectivemagnitudescale
SMValue

Description

Under5

Moderate

5to10

Serious

10to15

Severe

15to20

Hazardous

Above20

Intolerable

AmotionequivalenttotenontheSMscaleisgenerallydeemedunacceptableasthisrequirescrewtosupport
themselves by holding on to a suitable anchorage [Lloyd (1989)]. The SM values at combinations of speed,
waveheadingandwaveheightwereinvestigatedandplottedonpolarplotsfortheofficersloungeandbridge
deck12.Thesetwopositionswereinvestigatedbecauseasalthoughthemotionsonthemooringdeckwillbe
moresevere;itisnotexpectedthataccesstotheseareaswillberequiredinthemostextremeseastates.A
typicalpolarplotshowingtheSMvalueforvariousheadingsandwaveheightsat25knotsisshowninFigure
7.7.PolarplotsofSMforspeedsof10,15and20knotsaregiveninAppendixI.
ItisevidentfromthepolarplotsinFigure 7.7and AppendixIthatthemostseveremotionisexperiencedin
head and bow quartering seas and at wave heights above 6.5 metres where the SM in the officers lounge
reachesunacceptablelevels.AtnocombinationofspeedorwaveconditiondoestheSMmagnitudeonthe
bridgedeckexceedtheimposedlimitoften.Fromananalysisofthepolarplotsitspossibletodefinelimiting
speedsassociatedwithwaveheightanddirectioncombinationstopreventtheSMexceedingten.Theresults
fromthisanalysisareshowninTable7.7.
Slammingisanimportantphysicaleffectbecauseitcausesdeclarations,localstructuraldamageandtransient
vibratorystresses(whipping)[Lewis(1989)].Theprobabilityofslammingwasinvestigatedforallcombinations
ofwaveheight,waveheadingandspeedsbeingconsidered(Figure7.8)wherethecriticalvelocityforslamming
wascalculatedbasedonthetheoryofOchiforamerchantshipas3.76meterspersecond[Lloyd(1989)].

12
ThepresentedSMisbasedonthewaveperiodthatcreatestheworstmotionforagivenwaveheight.This
waschosenbecausewaveheightisamucheasierwaytodescriberealseaconditionsthatwaveperiod.

118

Design Deevelopmentt

Waveheading
W
/ degrees
135

Officers Lounge

SMat25Knotts
180
15

Wavehead
ding
/degrrees
1
135

10

Brid
dgeDeck

90

90

45

7.5m
m

h1/3

6.5m
m

h1/3

5.5m
m

h1/3

4.5m
m

h1/3

3.5m
m

h1/3

2.5m
m

h1/3

1.5m
m

h1/3

4
45
dline bridge decck
Solid
Brokkenline officerslounge
0

Figuree7.7Polarp
plotofSubjectiveMotionvvariationwithwaveheadinggheightforth
hebridgedeckkandofficers
5knots
loungeat25
Table7.7Summaryofmaximumshiipspeedince
ertainseastatestopreventexcessivemo
otions
Limitingspeed

h1/3

Hea
adseas Bow
wquarteringsseas

Beam,sternquarteriingandfollow
wingSeas

5
5.5metresand
dbelow

25
5knots

25knots

25knots

6
6.5metres

20
0knots

20knots

25knots

7
7.5metres

15
5knots

10knots

25knots

10%
Probabilityofexceedance

9%
8%
7%
6%
5%
4%
3%
2%
1%
0%
10

15

20
0

25

ShipSp
peed/knots
7.5
5m h1/3

6.5m h1/3

5.5m

h1/3

4.5m
m

h1/3

Figure7.8Prob
babilityofslam
mmingasafu
unctionofship
pspeedandw
waveheight
g
as expected
e
exceept for the drop in slamm
ming probability at 25 knotts. On closerr
Thesee results are generally
inspecctionofthep
probabilityofsslammingasaafunctionofw
waveperioditwasobserveedthatat25knotsthereiss
aninccreasedprobaabilityofslam
mmingoverth
heprobability at20knotsifallwaveperriodsareconssidered(from
m
3.5to
o11.5second
ds)asshown inFigure 7.9. However,att25knotsthewaveperiodsthatcause
ethegreatestt

119

Concept Design of a Fast Sail Assisted Feeder Container Ship


probabilityofslammingdonotoccuratthelargesignificantwaveheightsintheconsideredseaareasandthus
theprobabilityofslammingisreducedat25knots.ThisresultissimilartotheresultsofBonafouxetal.(2001)
whonoticedthesamepatternwheninvestigatingtheMSIoffastferries.
Asuitableacceptableprobabilityofslammingisintheorderof34%[Harriesetal.(2003)],thusoperationof
theshipispossibleinallseasupto6.5msignificantwaveheightwithanacceptablelevelofslamming.

Probabilityofslamming

0.14

25Knots
20Knots
15Knots
10Knots

0.12
0.1
0.08
0.06
0.04
0.02
0
6

9
10
Waveperiod/s

11

12

Figure7.9Probabilityofslammingasafunctionofwaveperiodandshipspeedfora7.5msignificantwave
height.Solidlinesindicatewavesthatexistintheinvestigatedseaareasanddashedlinesindicatewavesthat
donotexistintheinvestigatedseastates
TheMIIgivesanindicationofthenumberoftimesacrewmemberwillhavetostopworkduetotippingor
slippingandthustheirabilitytoworkeffectively.Theanalysishasbeenconductedbasedontransverseand
verticalmotion(althoughlongitudinalmotioncouldalsohavebeenconsidered)oftheshipforallwaveheight,
waveperiodandwavedirectioncombinations,.TheMSIcoefficient, ,fortippingisaratioofstancewidthto
apersonsCOG(definedas0.25inthisinvestigation)[Lloyd(1989)]andforsliding,isthecoefficientoffriction
betweenacrewmembersshoesandthedeckwhichforadryweatherdeckis0.7[Grahametal.(1992)].The
MIIforthemooringdeckwasjudgedasthemostadequatelocationtoapplythiscriterion,theresultsofwhich
areshowninFigure7.10.
Theprobabilityofdeckwetnesswasinvestigatedandfoundtobe1.56%fora7.5msignificantwaveheightand
0.42%for6.5msignificantwaveheightwhichisconsideredtobesmallandnotaconstraintonoperationwhen
comparedtocriteriasuchasslammingandSMandthushasnotbeenpresentedinthisreport.
Theprobabilityofbowaccelerationwascalculatedbutbecauseitdidntyieldanymoreinformationthanhas
alreadybeenobtainedfromothermeasureshasnotbeenpresentedinthisreport.
From the results presented, it is evident that the ship should be able to operate safely in seas up to 5.5m
significant wave height, equivalent to force 67 on the Beaufort scale (providing there is sufficient installed
powertoovercometheaddedresistanceandthecrewcantoleratereasonableshipmotion)whichmorethan
satisfies the design requirement set in Section 2.5. The problem with the measures discussed in this
investigationtomeasurecrewcomfortisthatthemeasuresaresubjectiveandthelimitsofwhatsacceptable
willvarywidelyfrompersontoperson.

120

Design Development
9
8
7

BowQuatering
Seas

MII/min1

6
5

HeadSeas

4
3
2

BeamSeas

1
0
0

Significantwaveheight/m
25Knots(HeadSeas)
20Knots(Headseas)
15Knots(HeadSeas)

25Knots(BowQuateringSeas)
20Knots(BowQuateringSeas)
15Knots(BowQuateringSeas)

25Knots(BeamSeas)
20Knots(BeamSeas)
15Knots(BeamSeas)

Figure7.10MotionInducedIndex(MII)forvariouswaveheights,waveheadingandspeeds

7.1.7

Rolldampingduetosails
Natureofrolldamping

SofartheWingSailhasbeenconsideredexclusivelyaspropulsionauxiliary.However,thissailsystemwillalso
act as a motion dampingdevice whichdecreases the roll andyaw motions of the feeder in resonant waves.
TheJapaneseexperienceinthe80swithsailassistedcargoshipssuggeststhatontopofthefuelsavingsdue
tothethrustgeneratedbythesails,thereisanotherfivetotenpercentdecreaseinresistancewhensailingto
windwardassociatedwithareductionofcoupledroll/yawmotions.Theotherwiseunexplainabledecreasein
drag is a direct effect of the reduced yaw motion in quartering seas and consequent decrease of induced
resistance. Research conducted at the University of Southampton in the 80s by Satchwell (1986) provides
guidancefortheestimateoftheaerodynamicloadsandassociatedrolldamping,howeververylittleisknown
aboutthecorrelationofthisreductioninmotiontothedecreaseindragandhencefuelsavings.Inthissection
thereductionincoupledroll/yawforthefeederisestimatedusingtheliftingsurfacetheoryandpublisheddata
basedonliftinglinetheory[Glauert,H.(1930)].Theresultsarethenusedtoestimatethepotentialreduction
inresistancewhichwouldoccurwhensailingtowindward.
Liftingsurfacemethodandliftinglineanalysis
The first estimate of the aerodynamic roll damping was carried out applying the lifting surface method,
howevertheresultswerefoundtooverestimatethedamping.AssuggestedbySatchwelltheaccuracyofthis
methodmaybelimitedbytheuncertaintyrelatedtothehullsuperstructure,vortexsheddingandturbulence
atsealevelshowstheformulaeusedandplotsofthedampingcomponents.
Theliftinglinemethodproposedhereneglectsstartingvortexeffectsandanydistortiontotheplanartrailing
vortex wake. This quasistatic analysis accounts for unsteady flow effects by a sequence of steady flow

121

Concept Design of a Fast Sail Assisted Feeder Container Ship


and

calculations [Satchwell C.J.(1986)].

are the dampingcomponents dueto incidencechanges

andairspeedchangesrespectively,AppendixJshowshowthesequantitiesarederived.
Controlsystem
Figure 7.11 shows that in strong following, winds the damping becomes negative. This is a result of the
airspeedchangescomponentwhichbecomesnegativeasthewindmovesaftofthebeam.Howeverdueto
150. Figure J.3
the low operational apparent wind angle the negative damping is present only when
showsthatincidencedampingisalwayspositiveifstallisavoided.Asimplecontrolsystemcouldbeusedto
avoidstallifthechangeinangleofattackduetotherollingmotionisknown.Againliftinglinetheorycanbe
used to estimate the changes in angle of attack along the span of the wing and the result could be used to
calculatetherequiredmargintoavoidstall[Equation(J.7)].
Rollreduction
Anumberofcomponentscontributetothetotalhydrodynamicrolldamping.Theseincludetheskinfrictionof
thehullincontactwithwater,momentsarisingfromseparationoftheflow,hydrodynamicsideforce,andthe
moment due to the creation of waves [Lewis (1989)] . The resulting hydrodynamic damping of the feeder,
which has been estimated in the previous section, can be used to describe the rolling of a ship as follows
suggestedbySatchwell(1986).

(7.2)

isalinearhydrodynamicdampingandtheterm

isintroducedtoincludeanynonlinear

effects,appropriatetoashiprollingthroughafullcycle.ThesolutionofEquation(7.2)forthemaximumangle

isproportionaltothewaveslopeand1 .Theaerodynamicdampingequivalent

ofheel

byidentifyingthedampingterminEquation(7.1)andequatingthisto
2

isfound

(seeAppendixJ).
(7.3)
(7.4)

From Appendix I.1, roll resonant frequencies for beam and quartering waves are found at
280
30,for15and25knotsrespectively.ObservingtheseastatedatainAppendixG[Hogben(1986)]we
notethattherearenoobservationsfortheresonantwavelengthsat25knots,howeverat15knotsthepeaks
30 accounts for 31 and
in quartering beam coincide with waves, but at 15 knots resonant roll at
19.5% of the sea state for Singapore and Caribbean respectively, suggesting that for the selected routes the
dampingactionofthewingsailwillbeparticularlybeneficialatthelowspeed.

122

Design Development

30

60

90

120

150

180

60

15Knots
50

RollReduction%

40
30
20
10
0

11

16

21

26

10
20

TrueWindSpeed/ms1

60

25knots
50

RollReduction%

40
30
20
10
0

11

16

21

26

10
20

TrueWindSpeed/ms1

Figure7.11Rollreductionfor15and25knotsshipspeedatdifferentapparentwindangles

YawComponent
Clayton&Sinclair(1989)giveguidanceonhowtoestimatetheyawaerodynamicderivativeofawingsailat
different headings. These are used to calculate the corresponding yaw damping coefficient component
.Asnolateralmotionswereincludedintheseakeepinganalysis,ayawdamping
inducedbyrollmotion
coefficient is assumed as 0.15 from Das et al. (2006), comparing this typical value with the aerodynamic
component,itisreadilyseenthatasignificantdampingcanbeachieved.

123

Concept Design of a Fast Sail Assisted Feeder Container Ship

30

60

90

120

150

180

0.10
0.09

YawReduction%

0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0.00
5

10

15

20

25

30

ApparentWindSpeed /ms1

Figure7.12ReductionofRollInducedYawComponent.Shipspeed25knots

Resistancereductionduetomotiondamping
InthesimulationofthevoyagesperformedinSection6.1.1theaveragereductioninrollmotionforresonant
wavesisfoundtobe16%and30%at15and25knotsboatspeedrespectively.Inadditiontothis,theaverage
reductionoftheyawcomponentinducedbyresonanthasalsobeencalculated.Withthisinformationisnow
possibletopredictadecreaseininducedresistanceinresonantwaves.Theexperimentalinduceddragdata
canbeusedtosimulatetheresistanceassociatedwithasinusoidal2.5degreesyawmotion;FromSatchwell
(1986)atimeaveragedinducedresistancecoefficient isdefinedasfollows:
1

where

250

(7.5)

istheperiodofcoupledroll/yawmotion,assumedtobe9seconds[Das&Das(2007)].Thenumerical

0.25. Using the same expression in Equation (7.5) the total resistance of the vessel
integration yields
1.20,thustheyawingmotioncouldcontributeupto
sailingat25knotscorrespondstoacoefficientof
around 20% of the total resistance. Similarly, the reduction at a ship speed of 15 knots is found to be 18%.
Removing this yaw motion when sailing to windward would eliminate this source of drag. The expectance
probabilityofsailinginquarteringandbeamwavescalculatedinSection6.1.1canbeusedtoapplyaweightto
thisreductiontoaveragethevaluefortheselectedroutesandusetheresultingvalueforaneconomiccasein
Section 6.2.2. For Singapore area the averaged benefit is found as2.14% and 1.93% at 25 and 15 knots
respectively.Thisreductionisnotincludedinthemaineconomicanalysisasitreliesonassumedhydrodynamic
yaw damping coefficients and yaw displacement periods. Future work could involve a detailed seakeeping
analysisofthefeederandoptimisetheaerodynamicmotiondampingandtheassociatedresistancereduction
for the resonant peaks, and aerodynamic pitch damping could be also considered for a decrease if added
resistanceinwaves.

124

Design Development

7.2

Structural design

7.2.1

Midshipscantlings

It was recognised that only a preliminary design was practical or desirable within the scope of the concept
design.Thusthedesignwaslimitedtoprimarymembersandassociatedstiffeningatthemidshipsection,(to
facilitate the finite element modelling of a hull module centred amidships). The approach consisted of a
combinationofdirectcalculationbasedondesignloads;rulebaseddesignfromLloydsRegistersRulesforthe
Classification of Ships [Lloyd's Register (2009)] (herein the Rules) and design from basis. Aside from initial
arrangementandplatethicknesscalculations,thestructuralworkwascarriedoutwiththeassistanceoftheLR
softwarepackageRulescalc.TheuseofRulescalcfacilitatedanincreaseinthespeedandaccuracyofstructural
property calculations, and improved the fidelity of Rules checks with a more accurate representation of the
shipsection.
Arrangement
The cue for the design was taken as the Nigel Gee and Associates Ltd. design, NG254, for which a midship
sectionandrelatedstructuralinformationwereavailable[Gee(2007)].Basedonthebroadsimilaritiesofthe
designsoperationalenvelopesanumberofkeyassumptionsfortheconceptweremadeatthisstagerelating
tothestructuralarrangement.
(v)

A duct keel in the double bottom would facilitate maintenance access and provide space for
powertransmissiontothemotorsatthestern.

(vi)

Sidegirdersinthedoublebottomwouldbeneglectedinfavourofthebilgeboxandductkeel.

(vii)

Conventional side structure with stringer decks including an underdeck walkway and high
scantling torsion box to resist stresses at deck level would be retained this latter being a
requirementparticularlystressedbytheRules.

(viii)

355 meganewtons per square metre) was to be used


LR H36 grade high tensile steel (
throughoutthemidshipsection.TheuseofH36throughoutthesectionratherthanonlyatdeck
andbottomallowedageneralreductioninstructuralmass,inadditiontonegatingconstruction
complexitiesarisingfromthejoiningofhightensileandmildsteelstrakes.

(ix)

Thescantlingdraughtusedinpressureloadingandrulescantlingcalculationswastakentobethe
freeboardsummerloadwaterlinecalculatedattheseconddeckfrommaximumdepth,leadingto
9.52metres.Theuseof inthedesignstageeffectivelyservesasabuiltinsafetyfactorfor
theshipinitsoperationalcondition.

(x)

Commercial bulb flats [Corus (2002)] were adopted as the secondary longitudinal stiffening
members throughout the section, as while manufactured Tsections can incur significant cost
savings [Blomqvist & Forrest (2000)], it was felt that standard bulb sections would improve
shipyardviability.

DrawingontheseassumptionsarisingfromthebasisdesignandthefurtherrequirementsoftheRules(Part4,
Chapter8),theprinciplestructuralmemberswerefinalisedasshowninFigure 7.13anddividedintoprimary
panels as in Table 7.8. The stiffener distribution between and along primary longitudinal members was
assumedtobethesameasthatfortheNG254,withoneadditiononpanels7and13toaccountforareduction
intheoveralldepthandarelativeincreaseintheheightofthewalkway.

125

Concept Design of a Fast Sail Assisted Feeder Container Ship


Table7.8Longitudinallyeffectivepanels

No.
1
2
3
4
5
6
7
8
9
10
11
12

Function
Keel
BottomShell
Bilge
SideShell1
SideShell1
SideShell1
SideShell1
Sheerstrake
MainDeck
InnerHull1
InnerHull1
InnerHull1

Note1

Figure7.13Structuralarrangementandprincipalpanels
withinthemidshipsection

Note2

No.
13
14
15
16
17
18
19
20
21
22
23
24

Function
InnerHull1
InnerHull1
BilgeBoxTop
BilgeBoxSide
InnerBottom2
InnerBottom2
DuctKeelSide
Girder
Stringer
Stringer
Stringer
Stringer

Thesideshellandinnerhullbelowthe
torsionbox,whilebehavingasasingle
panelinglobalterms,weredividedinto
individualstrakestofacilitatevarying
thickness.
Similarly,theInnerBottomabovetheDuct
Keelwastreatedseparatelytotherestof
thecargoregion.

Loads
Theinitialdesignpointwastakenfromverticalbendinggloballoadsforthefullyloadedconditioncombined
withlocalhydrostaticandcargoloadings.Thelocalhydrostaticloadingswerecalculatedfromfirstprinciplesfor
thescantlingdraught.Hydrodynamicloadsduetomotionsandwaveswereestimatedfromempiricalformulae
from the rules, while cargo loads were based on assuming a maximum averaged container loading of 14
tonnes and uniformly distributed throughout the hold. A working value of was obtained from the initial
massestimateadjustedfortheshipinballastcondition(maximumhoggingbendingmoment),whilethedesign
wasobtainedfromtheRules.ThedesignloadsaresummarisedinTable7.9.
valuefor
Table7.9Designloadingsusedininitialscantlingderivation,externalandinternallocaldesignloadsfor
individualpanelsandmaximumglobalbendingmoments

Panel
1
2

Hydrodynamic
32.5
32.5

Hydrostatic
95.7
95.7

Panel
10
11

Ballasthead
78.2
48.1

/kNm2
Cargoloads

32.5

88.6

12

22.0

4
5
6
7
8

32.5
32.3
32.2
30.6
27.3

37.9
11.8

15
16
17
18
19
20
21

27.0

26.1
46.0
78.2

41.6

46.2
46.2

Externalloads,

/kNm2

Scantlings

Internalloads,

Globalloads
/MNm

814.6
814.1
649.0

InitiallymodellingthemidshipsectionwithscantlingsderivedfromtheNG254,thelongitudinalstressesacting
was obtained for a uniformly
on plating elements were calculated by means of Equation (7.6), where

126

Design Development
loadedplateasinEquation(7.7),definingelementsasapanelofwidth andwithalengthofoneframebay.
Theplateshapefactor istakenas0.5forarectangularplatewithalledgesfixed[Young&Budynas(2002)].
(7.6)

(7.7)

Comparing this design stress to the permissible longitudinal stress from the Rules,
157 , panel
thicknesseswerethenreduced,balancingtheeffectonlocalstressvalueswiththatonthesectionmodulusat
deckandkeel.Oncesatisfiedwiththesevalues,thestructuralarrangementwasinputintoRulescalcformore
detailedanalysis.
Verification
TheinitialscantlingsweresubjectedtoanumberofcheckstoensurecompliancewiththeRules.
(i)

MinimumplatingthicknessrequirementswereobtainedfromPart4,Chapter8,Section5ofthe
Rulesandcomparedtoderivedscantlings(seeTable7.10).

(ii)

StiffenerandattachedplatingsectionmoduliwereobtainedfromPart4,Chapter8,Section6of
theRulesandusedtoselectappropriatestiffenerprofiles(seeTable7.11).

(iii)

Critical buckling stresses were calculated for plates and stiffeners in accordance with the Rules
andcomplianceverified(AppendixK.1).
Table7.10Platethicknessbydirectcalculationandminimumrulerequirements
Panel
No.
1
2
3
4
5
6
7
8
9
10
11
12

/mm /mm
18.0 19.0
16.0 12.0
14.0 12.0
11.0 10.0
11.0 10.0
11.0 10.0
11.0 10.0
17.5 11.0
17.5 13.5
11.0
7.5
7.5
7.5
7.5
7.5

FAIL
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS

Panel
No.
13
14
15
16
17
18
19
20
21
22
23
24

/mm /mm
9.0
7.5
17.5
13.5
10.0
9.0
12.0
10.0
10.0
8.5
8.5
8.5
10.5
10.5
12.0
10.0
12.0
9.0
9.0
9.0
9.0
9.0
17.5
9.0

PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS

TransverseStructure
The transverse structure was obtained similarly to the longitudinal stiffening by design from rule. Required
section moduli were calculated for the transverse framing on the shell plating. Suitable Teesection builtup
profileswerethenfound,assumingaconsistentflangeplateof200x10mm.Thesescantlingswereassumedto
berepeatedontheinboardplatingofthedoublehullstructure.Nearthemaindeck,scantlingswereincreased
aboverulerequirementsinaccordancewiththeincreasedtransversestiffeningobservedontheNG254,itwas
expectedthatthiswouldaidinresistingwarpingdistortionsatdecklevel.TheresultsareshowninTable7.12.

127

Concept Design of a Fast Sail Assisted Feeder Container Ship


Table7.11Minimumrulerequirementsforlocalsectionmodulusofstiffenersandattachedplating

Panel
/cm3
No.
1
2
3
4
5
6
7
8
9
10
11
12

270.1
268.9
288.9
200.8
196.3
149.7
117.8
103.8
311.8
200.3
189.5

Stiffener
profile
/mmxmm
650x12
220x10
220x10
240x10
200x8.5
200x8.5
180x9
200x10
200x10
240x10
200x8.5
200x8.5

Panel
/cm3
No.
13
14
15
16
17
18
19
20
21
22
23
24

5.0
13.5
147.5
150.9
412.5
491

220.4

19.8

Stiffener
profile
/mmxmm
180x8
200x10
180x8
180x8
280x10.5
260x10
200x10
200x10
none
180x8
180x8
200x10

Table7.12Requiredsectionmoduliandselectedwebprofilesfortransversestiffening
Panel
No.
1
2
3
4
5
6
7
8

/cm3
1401.5
1857.8
1797.6
610.4
524.3
407.0
325.5
325.5

WebProfile

/mmxmm
/cm3
730x12
1411.0
850x12
1863.0
840x12
1823.0
500x10
610.4
510x8
526.2
450x8
419.3
450x10
504.4
450x12
586.2

WithallrulestrengthchecksdeemedsatisfactorythescantlingswerefinalisedasshownintheMidshipSection
drawinginAppendixK.2andthedesignprogressedtomoredetailedstructuralanalysis.

7.2.2

Globalstrength

Once satisfactory scantlings had been established throughout the midship section a further check on the
structural design was carried out in the form of a global strength analysis. This consisted in the first part of
checkingtheglobalpropertiesagainstLRlongitudinalstrengthrequirementsdetailedinPart3Chapter4and
further,ContainerShipspecificrequirementsinPart4Chapter8.Onceglobalpropertieswereconfirmedwithin
permissiblelimits,operationalbendingmomentenvelopesweregeneratedandthegloballoadsforaselection
ofoperationalconditionsassessed.TheglobalpropertiesofthemidshipsectionareshowninTable7.13.
Table7.13Globalpropertiesofthemidshipsection
Global
Properties

/m4
96.89

/kgm
18,143

/m4
234.70

/m
7.611

/m3
8.53

/m3
12.731

/m
2.539

/m6
20,124

Rulecriteria
Globalstrengthpropertiesofthemidshipsectionwereassessedagainsttwosetsofrulecriteriaintwodistinct
conditions,aheadseaswaveconditionasthatusedinthescantlingderivationconsistingofmaximum and

128

Design Development
waveinduced
;andanobliqueseaconditionincorporatingtheeffectsofwaveinducedhorizontalbending
momentandtorque.Theheadseaswaveconditionisappliedasacheckontheabilityofthestructuretoresist
pureverticalbendingwithoutexceedingaspecifiedstresslevel.Thecriteriaforthecheckaresummarisedin
Table7.14.
Table7.14RuleGlobalstrengthcriteria
Criteria

Rule
/m3

Minimumhullsectionmodulus,

5.114
4

Minimumhullmomentofinertia,
Permissiblestress, /Nmm2
PermissibleSWBM, /MNm

/m

PASS

57.95
218.1
1365.4

PASS
PASS

While vertical bending checks were passed, the open section design of container ships means that they are
considerably more susceptible to antisymmetric distortions and stresses and this is accounted for in Lloyds
Rules.Usingruleprescribedbendingmomentsandtorques(summarisedinTable 7.15),thecombinedstress
valueatapositionwithinthesectioniscalculatedbymeansofEquation (7.8),inwhichthewarpingstresses
havebeenestimatedbywith[Magnuckietal.(2004)]and

and

wereacquiredfromRulescalc.Results

areshowninTable7.16.
Table7.15Rulederivedantisymmetricgloballoads.
/MNm
284.3

GlobalLoads

/MNm
69.9

/MNm
37.0

0.6

(7.8)

Table7.16Combinedstresscalculations

/m
/m
/m2

DeckEdge
Inboard
11.39
11.58
206.72
186.4
PASS

Bilge
Extreme
11.40
4.72
62.84
85.0
PASS

All global strength criteriabased only on the midship sectionhad been successfullypassed, so the next step
wastoconfirmthatthelikelyloadingconditionsfortheconceptasawholewerewithinpermissiblelimits.To
do this, the mass estimate derived in Section 3.2 was revised to include a more accurate description of the
structuralmasstotalanddistribution.
Hullsteelmass
Thehullsteelmasswasestimatedbymeansofacoffindiagram,thederivationofwhichtracedthecontinuity
and extent of major structural features. Specifically, the elements considered were the shell girth, strength
deckbreadth,doublebottombreadthandinnerhulldepth.Thecontributionsfromtheseelementswerethen
summated along the length of the ship and normalised to unity. After being corrected for scantling taper
allowancesattheforeandaftendsoftheshipfromtheRules,thisdistributionwasmultipliedthroughbythe

129

Concept Design of a Fast Sail Assisted Feeder Container Ship

1.2

350

1.0

300

/ tonnes.m-1

Normalised Structural
Distribution Factor,

valueofsteelmasspermetre( )amidshipsandintegratedtogivethelongitudinalstructuremass.Transverse
structuralmasswasestimatedfromthemassofaframeamidshipsanddistributedbythefulldepthsectional
areaofthehullform.Thenormalisedcoffindiagramandtheresultinglightshipmassdistributionareshownin
Figure7.14.

0.8
0.6

Longitudinal
Tapered
Transverse

0.4

250
200
150
100

0.2

50

0.0

0
0

50
100
Distance from AP /m

150

50
100
Distance from AP / m

150

Figure7.14Coffindiagramforstructuremassnormalisedbyamidships(left)andresultinglightship
distributionof(right)
Operationalconditions

SWBM

MS / MNm

400
200
0

200

Shear
Force

400
600
800
0

50
100
Distance from AP / m

MS / MNm

600

1000
800

Shear Force / MN

20
15
10
5
0
5
10
15
20

800

20
15
10
5
0
5
10
15
20

SWBM

600
400
200
0
200

Shear
Force

400
600

800
1000

150

50
100
Distance from AP / m

Shear Force / MN

From the revised hull mass estimate and using the load conditions specified in Appendix K.3, SWBM values
werecalculatedfortheshipinoperationtoconfirmthembeneaththemaximumpermissibleSWBM,seeFigure
7.15.

150

Figure7.15DistributionofstillwatershearforceandSWBMintheloadedarrivalcondition(left)andballast
arrival(right)
NeithertheloadednortheballastconditionproducedSWBMmaximaapproachingthepermissiblerulevalue
fromTable 7.14soassuchtheglobalstrengthcheckwasconsideredpassedonallcountsandthestructural
arrangementwasnowreadyformoredetailedanalysis.

7.2.3

Finiteelementmidshipsectionmodel

Theuseoffiniteelementmethods(FEM)inthedesignofshipshasbecomecommonplace,givingthedesigner
morepowertocontroltheoverallanddetaileddesignofashipstructure.FEMareparticularlyusefulinthe
structuralassessmentofnovelshipstructures(suchassailassistedships)whichhavegeometryorloadingsthat
aredifferenttothosenormallyexperiencedbyashipandisamandatoryrequirementfortheirclassification
[Lloyd'sRegister(2006a)].

130

Design Development
In order to assess the structural integrity of the proposed sail system and its effect on the ship structure an
FiniteElement(FE)modelofashortlengthoftheproposedconceptincludingonetransversebulkheadandsail
system mast has been modelled in ANSYS. The FE model has been used to perform a structural strength
assessmentoftheproposedmidshipstructureinstillwater;thestresseswithinthehullduetoextremeloading
onthesailsystem;thestressesduetodynamicwavebendingandtorqueloads.
Modelling
The modelling of the sail system and ship section was guided by International Association of Classification
Societies (2008) and Lloyd's Register (2006b) which provides guidance on the modelling of structural
components;restraintrequirements;loadingandacceptancecriteriarequiredinordertoconductastrength
assessmentusingFEM.
A28metresectionoftheshipwasmodelledwhichincludedthefulllengthoftheshipsparallelmidbodyplusa
small allowance at either end. The procedure involved modelling the entire primary and secondary support
structure,withplatesmodelledusingshellelementsandstiffenersandridersmodelledwithbeamelements.
Thebeamandplateelementswhereconnectedtogetherbetweentheirnodeswithrigidlinkswhicharebeam
elements with very high rigidity and very small mass. This effectively welds the beam and plate elements
together.Brackets,smalllighteningholesandsmallradiuseshavenotbeenmodelledasrecommendedbythe
CSR.

Figure7.16Theextentsofthefiniteelementmodel(left);andthemodellingofstiffenedplatesusinga
combinationofbeamandplateelements(right)
Themasthasbeenassumedtobefixedrigidlytothecrossdeckwiththesailsbeinglowereddownthemast
andfoldedtobestoredondeck(AppendixM).Fromthelistofpossiblesailstowagearrangementsgivenin
Section 5.1.2thisisdeemedthemostpracticalmethodoffixingthesailtotheshipbecauseitminimisesthe
totalbendingmomentwhichinturnreducesmastscantlings;minimisesmasttipdeflectionandhastheleast
interferenceontheshipstructure.Thesamehightensilesteelasthehullhasbeenusedforthemast,thisisa
changefromtheinitialmaterialselectionofAluminium5083becauseofthedifficultyinweldingaluminiumto
steel.
Thesailsupportingbulkheadandcrossdeckwasinitiallyassumedtobeofequivalentdimensionstothatofa
standard watertight bulkhead with the addition of additional stiffening on the reverse side to create a cross
deckthreemetresinwidthratherthan1.25metresforastandardbulkhead.Effectivelythiscanbeseenas
twostandardwatertightbulkheadsbacktoback,astructuraldrawingofthemodelledbulkheadcanbeseenin
AppendixK.2.
FEMdiscretisesastructureintoelementswithanexactrepresentationoftheshipstructureonlyachievedwith
an infinite number of elements with the accuracy of the solution tending towards reality as the number of

131

Concept Design of a Fast Sail Assisted Feeder Container Ship


elementsisincreased;howeverthiscomesattheexpenseofcomputationtime.InordertomakeFEMafast
and reliable method of strength assessment, acceptancecriteria is given based on an assumed mesh size by
providing a suitable safety margin to take account of the coarseness of the mesh. The mesh for the hull
structure was developed based on the guidelines for a structural strength assessment given in the CSR and
LRSDAsotheacceptancecriteriagivenintheLRSDAprocedurescouldbeapplied.Themeshingprocesscan
besummarisedasfollows:

Ameshisgeneratedthatfollowsthestiffeningsystemasfaraspracticable,hencerepresentingthe
actualplatepanelsbetweenstiffeners;

Oneelementisusedbetweeneverylongitudinalstiffener.Longitudinally,theelementlengthisnot
greaterthantwolongitudinalspaces;

Oneelementisusedbetweeneveryverticalstiffenerontransversebulkheads;

One element is used between every web stiffener on transverse and vertical web frames, cross ties
andstringers;

Atleastthreeelementsareusedoverthedepthofdoublebottomgirdersandfloors,transverseweb
frames,verticalwebframesandhorizontalstringersontransversebulkheads;

Theaspectratioofplatesiskeptasclosetooneasispracticallypossible;

Theplateelementsarealldefinedinthesameclockwisedirectiontoensurethenormalsareinthe
samedirection;

The rotation of the local element coordinate system for beam elements is chosen to orientate the
elementcorrectlyintheglobalcoordinatesystem.

Inadditiontothemeshforthestrengthassessmentalocalfinemeshwascreatedtoassessthestressesarising
around the intersection of the mast to the cross deck from which suitable structure could be designed to
preventfailureofthemastunderextremeloadingevents.Themeshforthisareawasgeneratedwithguidance
from the CSR and involved creating a mesh with a resolution of 50 millimetres square extending for ten
elementsinalldirectionsfrominvestigatedarea(Figure 7.17).Theotherhighstressconcentrationareasina
containershipsuchastheintersectionofthedeckandhatchcoamingswiththetransversebulkheadswhich
are normally investigated using a fine mesh when analysing a container ship structure (LRSDA procedures,
(2006))havenotbeenconsideredinthisinvestigation.
Thesailmastwasmodelledfromtheconnectionwiththecrossdecktothebottomofthesailsusingthesame
finemeshsizeusedfortheintersectionofthemasttothecrossdeck.Thenodesatthetopofthemastwere
connectedtoanindependentpointatthecentreofthemastsection.Thispointwasthenusedasalocationto
applyforcesandmomentstoaccountfortheloadingonthesails(Figure7.17).
ThemodelrestraintwasselectedinthemostminimalwaypossiblefollowingtheguidelinesoftheLRSDAand
CSR to prevent rigid body motion while minimising the amount of additional forces applied to the model.
Groundspringelements(anelementwithoneendconstrainedinallsixdegreesoffreedom(DoF),andstiffness
inglobal orglobal degreeoffreedomdependingonthestructuralmembertowhichitisattached)were
usedtoconstraintheintersectionofthedeckandinnerandoutersideshellinglobal directionandthedeck
andtransversefloorsintheglobal direction(Figure7.18).Thestiffnessofthegroundspringswascalculated
usingtheLRSDAprocedurestoprovideanequivalentshearstiffnessoftheshipatthemodelends.

132

Design Development

Figure7.17Thefiniteelementmodelshowing:thefinemeshregionattheintersectionofthemastandcross
deck(left);andtheapplicationofloadstothetopofthemast(right)

Figure7.18Locationsofgroundspringsusedtorestrainthemodelatitsends
Thenodesateitherendofthemodelwereconnectedusingrigidlinkelementstoanindependentpointlocated
atthemodelsneutralaxis.Oneendthemodelwasconstrainedatitsindependentpointin translation.Ina
similar manner to the independent point at the top of the mast, moments to replicate the global bending
moment of the ship where applied to the independent points at the model ends. The application of end
momentsinthismannercreateslargestressconcentrationsatthemodelendsneartheconnectionoftherigid
linksandtheshipstructure.Inordertominimisetheeffectofthesestressconcentrationsonthestressresults
themodelledsectionwasextendedbeyondtheparallelmidbody.
The modelling of the ship section was performed with the aid of Microsoft Excel in order to generate the
requiredANSYSinputfile.Thisgavetheauthortheabilitytomakedimensionalchangestotheshipstructure,
changeplatethicknesses,sectiontypesandnumberofstiffenersquicklyinordertoaccesstheaffectofthese
parametersonthestrengthoftheshipsection.ItalsoallowedtheauthortomodeltheshipsectioninANSYS
withoutknowingthescantlingsgiveninSection7.2.1whichcouldbeinputafterwards.

133

Concept Design of a Fast Sail Assisted Feeder Container Ship

Figure7.19Applicationofendmomentstotheindependentpointsattheendofthemodel(left);andthe
meshattheintersectionofthetransversebulkheadandsidetanks(right)
Loading
Sincethereisinsufficienttimetoconductafullanalysisofthestructureinalltheloadcasesrequiredforafull
strengthassessment,theLRSDAprocedures,theCSRandLloyd'sRegister(2008)havebeenusedforguidance
to derive load cases that the author feels are representative of the most extreme loading the ship will
experience. The loading conditions used to investigate the structural strength of the ship and mast were
developed to investigate the ship in the full load departure loading condition13 in four motion conditions14,
namely:
1.
2.
3.
4.

Staticallyinstillwater;
Dynamicallyinheadseas;
Dynamicallyinbeamseas;
Dynamicallyinobliqueseas.

Theloadingappliedtothemodelforthefourmotionconditionscanbebroadlybedividedintothefollowing:

Hydrostaticpressureincludingtheeffectoftheshipheelandtrim;

A uniform pressure applied to the bottom of the cargo holds representative of the weight of the
containerstacks.Inthedynamicloadcasesthecargopressureisfactoredbytheverticalacceleration
at the container stacks centre of mass (as calculated by the seakeeping analysis). The affect of
longitudinalandtransverseaccelerationsonthecargohavenotbeenconsideredforsimplicity;

Moments applied to the ends of the model so that the maximum still water bending moment is
achievedwithinthelengthofthemodelledsection;

Dynamicverticalwavebendingmomentsandhorizontalwavebendingmoments;

Hydrodynamicandcargotorqueappliedatthemodelends;

Correction loads distributed across the intersection of the transverse web frames and deck plate to
ensurenonetverticalforceonthemodel;

Windloadingpressureonsailmast;

13
Thefullloaddepartureconditionhasbeenselectedforanalysisasthisloadingconditionproducesthe
higheststillwaterbendingmoment.
14
Theheadseas,beamseasandobliqueseasintheLRSDAandCSRforassessingtheglobalstrengthofthehull
havebeenassumedtobeequivalenttothemaximumpitching,rollingandcombinedmotionsloadcases
prescribedbyLloyd'sRegister(2008)CodeforLiftingAppliancesintheMarineEnvironmentwhichhasbeen
usedtoderivetheextremeloadingonthesailsystem.

134

Design Development

Inertialloadingonsailmastduetoshipmotions;

Forcesandmomentsappliedtotopofmasttoaccountforwindloadingandinertialloadingonsails
duetoshipmotions.

Figure7.20Applicationofhydrostaticandcargopressures(left);andcorrectionloads(redarrows)required
toensurenonetverticalforceonthemodel(right)
Forcestoaccountforwindloading,inertialloadingduetotheshipmotionsandoperationalloadingonthesail
and mast were calculated using a procedure to access the strength of cargo handling equipment [Lloyd's
Register(2008)]forthefourloadingconditionsbeingconsideredusingthemaximumrollandpitchanglesfrom
Section7.1.4.Loadingonthesailswasdeterminedandappliedasanequivalentsetofforcesandmomentsto
theindependentpointatthetopofthemast.
ThemainchangefromtheprocedureofLloyd'sRegister(2008)rulesforcargosecuringarrangementswasin
the selection of design wind speed. Lloyd's Register (2008) defines a design wind speed of 63 metres per
second,howeveritwasdecidedthatthisisexcessivesincetherulesareprescribedforanoffshoreinstallation
whichwillseefarharsherconditionsthantheship.Insteada50yearreturnwindspeedwasinvestigatedas
defined by International Electrotechnical Committee (2005) for a ten minute gust. International
ElectrotechnicalCommittee(2005)definesasimplerelationshipfora50yearreturnwindspeedassimplyfive
timestheyearlymean.Thisisequivalenttoa23.2metrepersecondwindspeedintheCaribbeanandSouth
EastAsiaanda24.2metrespersecondwindspeedworldwide.Awindspeedof32.6metrespersecondhas
beenchosenandusedforthecalculationofwindloadingswhichisequivalenttowithstandingaforceeleven
storm and is more onerous than the International Electrotechnical Committee (2005) 50 year return wind
speed.Thisisdeemedamorethanadequatewindspeedinwhichtoaccesstheshipwithitssailsinthefully
extendedposition.
Themaximumverticalwavebendingmoment,horizontalwavebendingmomentandcargoandhydrodynamic
torque are applied to the motion load cases using the design load combination factors given in the LR SDA
procedures.Thesearefactorsappliedtothemaximumwavebendingmomentinheadseastoestimatethe
wavebendingmomentsinthebeamandobliqueseaconditions.
The additional loading on the ship structure due to the passing of a wave is usually accounted for with an
additional positive (wave crest) or negative pressure (wave trough) superimposed onto the hydrostatic
component.Theadditionalloadingimposedbygreenseas15(waterondeck)istakenaccountofwithrelevant

FromSection7.1.4itwasfoundthattheprobabilityofdeckwetnesswasverysmallandthussupportsthe
decisiontoneglectthisloadingconditionfromtheanalysis.
15

135

Concept Design of a Fast Sail Assisted Feeder Container Ship


pressure loading. To simplify the analysis and due to time constraints the effect of these dynamic wave
pressureshavenotbeenconsidered.
Using a similar argument to neglecting dynamic wave components, the assessment of the structure in
combinationsofdifferentcargoloadingconditionsanddamagedconditionshasnotbeenconsidered,forafull
analysistheseloadcaseswouldneedtobeundertaken.
Thedimensionsofthemodelledmastsectionwereassumedtoconsistofacircularcylinderofgiventhickness
with vertical stiffening running its entire vertical height. The principle mast dimensions; plate thickness;
diameterandstiffenersectionwerechosenbasedonthemaximumdesignloadsusingabeamtheoryapproach
andalimitingstressequaltoyieldatthefootofthemast.Sincethemastandsailloadingisdependentonsail
andmastweightaniterativeoptimisationprocedurewasadoptedinwhichthemastdiameter,platethickness
andstiffenersection(chosenfromalistof450standardandcustomsections[Corus(2002;2007)])waschosen
in order to minimise the mast weight. A summary of the optimum mast dimensions from a beam theory
approachisgiveninTable7.17.
Table7.17Summaryofoptimummastdimensions
Parameter
Mastdiameter/m

Dimension
2.00

Mastplatethickness/mm

10.00

Stiffenerdepth /mm

100.00

Stiffenerwidth/mm

150.00

Stiffenerwebthickness/mm

25.00

Stiffenerflangethickness/mm

25.00

Numberofstiffeners

18.00

Mastmass/tonnes

14.14

Thebucklingcapacityofthemastwascalculatedusingbeamtheoryandfoundtobe413MNwhichismore
thanadequatetoresistthecalculatedmaximumverticalloadonthemastof0.596MNintheheadseasmotion
condition.
Acceptancecriteria
Theacceptancecriterionwhichdetailstheallowablestresseswithindifferentpartsoftheshipstructurewas
taken from the LR SDA procedures and is summarised in Table 7.18. Results have been considered for a 20
metresectionofthemodelwithnoresultsconsideredforthefourmetresateitherendforreasonsdiscussed
previously.
) have been considered, although within LR SDA
For this investigation only the von Mises stresses (
procedures acceptance criteria for longitudinal ( ), transverse ( ) shear ( ) and buckling ( ) stresses are
given. There are also prescribed stresses for beam elements which have not been evaluated in this
investigation for simplicity. It is anticipated that general conclusions regarding the suitability of the ship
structurecanbeobtainedfromtheplateelementvonMisesStressresults,thusjustifyingthissimplification.
Themaximumvaluesofallowablestresswithinthefinemeshzone,aroundtheintersectionofthemastand
crossdeckaredifferentfromtheacceptancecriteriaforprimarystructureandweretakenfromtheCSRand
aresummarisedinTable7.19.

136

Design Development
Table7.18Acceptancecriteriaforprimarystructure[Lloyd'sRegister(2006b)]
Structuralitem

AllowablevonMisesstress(

Bottomshellplating
Doublebottomgirders
DeckPlating
Doublebottomfloors
Sideshelllongbulkhead
Sidestringers
Sidetransverses
Transversebulkheadplating
Crossdecktransverses

0.63
0.63
0.75
0.75
0.75
0.75
0.75
0.75
0.75

16

(
(
(
(
(
(
(

)/Nmm2
223.65)
223.65)
266.25)
266.25)
266.25)
266.25)
266.25)
266.25)
266.25)

Table7.19Maximumpermissiblestressesforfinemeshanalysis[InternationalAssociationofClassification
Societies(2008)]
Loadingcondition AllowablevonMisesstress(

Elementstress

1.75

621.25)

Stillwater

1.36

318.24)

Dynamicloadcase

1.5

532.5)

Stillwater

1.2

426)

Elementnotadjacenttoweld Dynamicloadcase
Elementadjacenttoweld

)/Nmm2

Modelvalidation
The finite element model was verified by comparing the modelled mass to the expected model mass. A
summaryofthetotalmassofthefiniteelementmodelisgiveninTable7.20.
Table7.20Finiteelementmodelvalidationbystructuremass
No. ofelements

Mass/kg

Plates(shell93)

90387

561437

Beams(beam4)

35770

137252

20

1.19

Total

698690.19

Structuralcomponent(elementtype)

Restraintelements(Link8)

The expected model mass excluding the additional structure added to the reverse side of the transverse
bulkhead is 658.490 tonnes. This seems reasonable since there will be mass associated with the increased
scantlingsofthetransversebulkhead,ifthiswasconsidereditissuspectedthatthemodelmassandshipmass
wouldbeverysimilar.
The mast deflection was verified by fully restraining the at mast deck level in all directions so it could be
considered a fixedfree beam. The tip deflection under a known load was checkedto see if was reasonable
whencomparedtoabeamtheorycalculation.Thetheoreticalpredictionofdeflectionfrombeamtheorywas
found to be 3.965 centimetres. The tip deflection predicted by ANSYS was found to be 6.4535 centimetres
whichisofcomparativemagnitudetothebeamtheoryestimate.

16
istheyieldstressofsteel,whichforH36hightensilesteelis355MNm2.

137

Concept Design of a Fast Sail Assisted Feeder Container Ship


Thefinalcheckofthemodelgeometryforerrorscamefromcarefulanalysisofstressresults.Itwasfoundin
theinitialstagesoftheanalysisthatmanyareasofthestructurewereshowingunexpectedresults.Onfurther
investigationoftheseareasitwaspossibletotracetheseerrorstoincorrectlydefinedgeometryandmodelling
errorswhichweresubsequentlyfixed.
Hullstructureresults
From an initial analysis conducted using the ship scantlings defined in Section 7.2.1 it was found that there
wereanumberofoverstressedregionsinthehullstructure,namely:

Thekeelplating(1)forallloadcases(Figure7.23andAppendixK.4)

Theinnerbottom(17,18)andbilgeboxtop(15)platingaroundtheintersectionswiththeductkeel
side(19)andtransversewebframesfortherollloadcase(Figure7.21);

Theintersectionofthecrossdeckwiththeinnerhull(14)intheheadseascondition(Figure7.21);

Thesideshell(4,5)attheintersectionwiththestringers(22)andtransverseframesinthebeamsea
loadcase(Figure7.24).

The author made changes to the hull structure upon analysis of the initial results which principally involved
strengthening the pipe trunk. It was observed in Gee (2007) that in the development of the Norasia fast
feederthattherewassomeadditionalverticalbulkheadstiffenersbetweenframes.Theauthoraddedthese
additional structural items to the FE model to relive some of the high stresses in the keel plating (1). The
additional structure consisted of plating extending between the centre girder and duct keel side (19) and a
vertical stiffener on the duct keel side (19) spaced three between every transverse frame (Figure 7.22 and
AppendixK.2).
Closer inspection of the failure region of the inner bottom (17,18) in the beam seas load case revealed that
these areas were only marginally overloaded. It is suspected that a factor in the presence of these stress
concentrationsisduetothewaythestructurehasbeenmodelledandcargoloadsapplied.Inreality,therewill
be additional reinforcement around the location of container corner posts to support the container stacks.
Thismeansthecargoloadscouldbetreatedasaseriesofpointmassesatthecontainercornerpostsandnot
asauniformpressureloadashasbeenusedinthisanalysis.Forthesereasonstheseoverstressedareasshall
not be focussed on in this investigation as the overloaded areas are likely to be due to the limitations of
neglectingthefoundationsunderthecontainerstacksandthewayinwhichthecargoloadshavebeenapplied
andnotduetoadeficiencyinthehullstructure.ItwasalsocommentedoninSection 7.1.2thattheshiproll
motionispoorlypredictednumerically.SincetheloadingontheFEmodelisderivedfromthesemotions,this
maybehavinganadverseeffectonsomeoftheloadingconditionswhichisturnshowingthatsomeareasof
thestructurefailtomeettheacceptancecriteriainthebeamsealoadcasewheretherollmotionisdominant.
The intersection of the cross deck and inner hull (14) is a common location to get high stresses within a
container ship [Lloyd's Register (2006b)]. As a result this is usually the focus of detailed FE analysis in the
design of a conventional container ship. These stress concentrations are usually prevented with detailed
designoftheintersectionandbecausethefailureregionissmallandonlyoccursintheheadseasloadcaseis
notconsideredanissueforthisinvestigation.Theprincipalconcernwiththisinvestigationismorewithglobal
strengthanddetaileddesignofthemastcrossdeckintersectionandnotthedetaileddesignofhatchcorners.

138

Design Development

Figure7.21Locationsofareasofinnerbottom(17,18)andbilgeboxtop(15)platingthatdonotmeettheLR
SDAacceptancecriteriadefinedinTable7.18(redareas)fortheinitialanalysisinbeamseascondition(left);
andthestressconcentrationcreatedattheintersectionofthecrossdeckandinnerhull(14)fortheheadseas
condition(right)[AreasinblueindicatestructurethatmeetstheLRSDAacceptancecriteria]

Figure7.22Structuralimprovementsmadetoshipstructureasaresultoftheinitialanalysis.
Figure 7.23 shows the typical failure region in the keel plating (1) for the initial analysis and the affect the
structuralimprovementshadinreducingthestressestowithinacceptablelevels.Theresultspresentedarefor
thestillwaterloadcase,resultsforthedynamicloadcasescanbefoundinAppendixK.4.
Inadditiontothekeelplate(1)andtheinnerbottom(17,18),thesideshell(4,5))onthesubmergedsideofthe
shipinthebeamseaconditionfailedtomeetthecriteriaoftheLRSDAprocedures(Figure7.24).Thefailurein
theseareasisverysimilarinformtothefailureregionswiththeinnerbottom(17,18),withthefailureregion
occurring at the intersection of the side shell plating (4,5) with the transverse frames and stringers (22) and
wereonlymarginallyoverloaded.Itisanticipatedthatthesefailureregionareconnectedtotheinaccuraciesin
therollresponsepredictedbytheseakeepinganalysisproducingamaximumrollanglewhichislargerthan
reality and thus imposing higher pressures on the side shell than will be experienced in reality, and for this
reasonshallnotbeconsideredinthisinvestigation.

139

Concept Design of a Fast Sail Assisted Feeder Container Ship

Figure7.23Locationsofareasofkeelplating(1)thatdonotmeettheLRSDAacceptancecriteria(redareas)
fortheinitialstructureanalysis(left);andthemodifiedstructureanalysis(right)inbeamseascondition[Areas
inblueindicatestructurethatmeetstheLRSDAacceptancecriteria]

Figure7.24Locationsofareasofthesideshell(4,5)thatdonotmeettheacceptancecriteriaofTable7.19
(redareas)(redareas)forthebeamseacondition[AreasinblueindicatestructurethatmeetstheLRSDA
acceptancecriteria]
Maststructureresults
Fromananalysisofthefinemeshregionaroundthemastintersectionwiththecrossdeckandmastforthe
initialmodel,itisevident;andnotunexpectedthat;thereisasignificantstressconcentrationformedatthe
intersectionofthemast,crossdeckandunderlyingverticalsupportstructureandinthelowerpartsofthemast
plating(typicalresultsfortheheadseasconditionareshowninFigure7.26resultsfortheotherloadcasesare
presentedinAppendixK.4).ThestresseswithintheremainderofthemastwerewithinthelimitsoftheLRSDA
proceduresandtheseconclusionswereconsistentacrossallthedynamicmotionconditions.
As a result of the initial analysis of the mast structure the author set about making a series of structural
improvementstothemaststructure.Fouriterationsweremadearoundthedesignlooptoarriveatamast
structure that met the requirements of the LR SDA procedures acceptance criteria given in Table 7.19. A
summaryofthestructuralimprovementsmadetothemaststructureafterthelastiterationofthedesigncycle
aresummarisedasfollows:

Increased mast plating thickness around bottom 1 metre section from ten millimetres to twenty
millimetres(althoughforpracticalpurposesthewholemastwasmodelledastwentymillimetreplate)
torelievehighstressareasinthemastplatingatthebottomofthemast;

140

Design Development

Additionofhorizontalringsofplating250millimetresdeepand17.5millimetresthickspacedat1000
millimetreintervalsverticallywithinmast(Figure7.25);

Additionof30millimetrethicksupportbracketswith120millimetrewide,30millimetrethickridersat
thefourquadrantsofthemastintersectionwithcrossdeck(Figure 7.25)toreducethemagnitudeof
the stress concentrations, reduce the stresses in the lower mast plating and reduce the mast tip
deflection.

StructuraldrawingsofthefinalmastscantlingsincludingthesechangescanbefoundinAppendixK.4.

Figure7.25Structuralimprovementsmadetomaststructureasaresultoftheinitialresults,withtheaddition
ofbracketsatthemastandcrossdeckintersection(left);andinternalringstiffening(right)
Thehighstressregioncreatedattheintersectionofthemastandcrossdeckfortheheadseasconditioninitial
analysiscanbeseeninFigure 7.26.Inaddition,figuresshowingtheeffectofthemodifiedmaststructureon
reducingthemagnitudeofthestressinthemastandcrossdeckplatinghavealsobeenpresented.Withallof
thepresentedresultstheregionsofhighstressarereasonablysymmetricalinsizeandshapeoneithersideof
the mast with one side showing a higher peak stress depending on the direction of loading. Results for the
beamandobliqueseasloadcasescanbefoundinAppendixK.4.

Figure7.26Stressinthefinemeshzoneatthebaseofthemastwhichdoesnotmeettheacceptancecriteria
ofTable7.19(redareas)fortheinitialanalysis(left);andthemodifiedanalysis(right)intheheadseas
condition[AreasinblueindicatestructurethatmeetstheLRSDAacceptancecriteria]
ItcanbeseenfromFigure 7.26andAppendix K.4thattheadditionalstructurehashadasignificanteffecton
reducingthemagnitudeofthestressaroundtheintersectionofthemastandcrossdeck.Therearestillsmall
areas where the requirements of the LRSDAprocedures acceptance criteria have not been met. The head
seasloadcomparativelycaseshowsamuchlargerfailureregion(Figure 7.26)thanthebeamandobliquesea

141

Concept Design of a Fast Sail Assisted Feeder Container Ship


loadcasespresentedinAppendixK.4.Itistheopinionoftheauthorthattheseareascouldbeeliminatedwith
increaseddimensionsofthebracketsatthefourquadrantsofthemastintersectionwiththecrossdeckwhich
due to time constraints have not been made within this project. It has however been demonstrated that
significant improvements can be made with the addition of very small amounts of additional supporting
structure.
Themasttipdeflectionwasinvestigatedfortheconsideredloadcasesandcomparedtoanallowablelimitof
50 centimetres at the mast tip to prevent the mast striking the container stacks. The results of this
investigationareshowninTable7.21.
Table7.21Themastdeflectioninthedynamicloadcasesfortheinitialandmodifiedanalysis
Loadcase

Initialdeflection/m

Modifieddeflection/m

%reduction

Headseas

0.406

0.116

71.8

Beamseas

0.267

0.0726

73.1

Obliqueseas

0.335

0.124

63.0

ItcanbeseenfromTable 7.21thatatalltimesthedeflectionofthemastiswithinthelimitstopreventthe
maststrikingthecontainerstacks.Itcanalsobeseenthatthemasttipdeflectionisreducedbybetween63%
and 73% with the addition of the additional structure at the base of the mast which is a significant
improvement.
Discussiononstressconcentrations
Fromtheresultspresentedforthemastintersectionwiththecrossdeckitisevidentthattherearesomevery
extremestressconcentrations(Figure 7.27).Itiswellknownthatstressconcentrationsarenotwellmodelled
by FE programs which, is in part due to neglecting brackets and radiuss when modelling. It can be seen in
stresscontourspresentedinFigure7.27showhowthestressconcentrationincreasesveryrapidlyasthestress
concentrationisapproached(leftofFigure7.27).

Figure7.27Highstressconcentrationregionsaroundthemastandcrossdeckintersectionfortheinitial
analysis(left);andthemodifiedanalysis(right)fortheheadseascondition
Ainfluentialfactorinthemodellingofthesestressconcentrationsisthechoiceofmodellingscheme,beit,a
solid,shellofshellbeamhybridmodel,thelatterofwhichhasbeenusedtomodeltheshipstructurewhilea
shell modelling scheme has been used for the mast. Of these methods the solid model is the most
representative of the ship geometry, but it is also the most computationally expensive, where as the shell
beam model is the most efficient but least representative of reality. A subinvestigation conducted by the
authortoinvestigatetheeffectofmodellingschemeonthestressconcentrationfactorcreatedbetweenthe

142

Design Development
mastandthecrossdeckplaterevealedthatthepeakstressofthesolidmodelwas4752%ofthatoftheshell
model.ThesubinvestigationwasconductedusingFEmodelswhichincludedthefullextentofthemast,an
area of cross deck plating around the mast foundation and the top supporting plating underneath using the
sameloadcasesasusedforthemainFEinvestigation,seeTable7.22.
Table7.22Errorinpeakstressduetothechoiceofmodellingschemeforthethreemotionloadcases
Peakstressforaconsidereddynamicloadcase
Roll
Pitch
Comb
18.85
19.02
22.65
9.23
9.02
11.84
24.91
25.90
39.70

Modelling
Scheme
Shell
Solid
Shellbeam

TheseresultsindicateastohowmuchANSYSisoverestimatingtheconcentrationcreatedattheintersectionof
themastandthecrossdeckduetothechoiceofmodellingscheme.Forfurtheranalysisofthemaststructure
carefulconsiderationtothelocationofthestressconcentrationsandthereasonsfortheiroccurrencewould
needtobemadetoensurethestressesarenotaresultofthechosenmodellingscheme.Itissuspectedbythe
author that a significant proportion of the peak stresss observed in the presented fine mesh results can be
attributedtothiseffect.
Discussion
FEMhasbeenusedinthisinvestigationtoinvestigatetheglobalstrengthoftheproposedhullform.Anumber
of deficiencies were found and either solved with additional structure or attributed to modelling errors that
cannotberesolvedwithinthetimeframeofthisproject.Theintersectionofthemastwiththecrossdeckhas
beeninvestigatedandsuitableimprovementssuggestedtopreventfailureofthemastunderextremeloading.
Itappearsfromananalysisofalltheresultsthattherearenosignificantinfluencesofloadingonthesailson
theultimatestrengthofthehullandthatwiththerightstructuraldesigntheproposedconceptisviable.There
are however a number of local areas of structure which still fail to meet the requirements of the LR SDA
procedures.Itishopedthatwithfurtherworkthesesmallareascouldbeeliminatedwithincreasedscantlings
andcarefuldetaileddesignofkeyareas.
Theanalysisconductedisonlyasmallportionofwhatisrequiredforafullanalysisandverificationofthehull
structure.Inafullanalysisofthehullstructuremanymoreloadcasesanddetailedinvestigationsintocritical
areasofthedesignwouldneedtobeundertaken.
Therearemoreconsiderationstothestructuraldesignoftheproposedsailsystemthanjustultimatestrength.
Windiscyclicinnaturewhichcreateslargefluctuatingstressesinthemastandhullstructureandcombined
withsignificantstressconcentrationsaroundthemastintersectionwiththecrossdeckmeantherearelikelyto
be problems from fatigue. It is likely that before any ship will use sails for propulsion an in depth fatigue
analysis will have to be undertaken using FEM to redesign the mast cross deck intersection to improve its
fatigueperformance.
IthasbeendemonstratedinthisinvestigationhowFEMisausefultoolthatcanbeusedtoprovethereliability
ofanovelfeatureofashipstructure,aprocedurewhichisamandatoryrequirementfortheclassificationofa
shipwithunusualornovelstructure.Therearehowever,intheopinionoftheauthor,anumberoflimitations
withtheuseofFEMforshipdesign,thatinclude:

Theresultsareverysensitivetomannerinwhichthemodelisrestrained;meshed;loadsareapplied
andresultsobtained.Throughoutthisinvestigationeveryefforthasbeenmadetoeliminateerrors
asmuchaspossible,however,anumberofsimplificationsandassumptionshavehadtobemadeto
completetheprojectwithinthetimeframewhichwillallhaveaninfluenceonthefinalresults;

143

Concept Design of a Fast Sail Assisted Feeder Container Ship

There is a large potential for errors in modelling, principally due to user error due to the large
numberofconsiderationsthatneedtobemadethroughouttheprocess;

Theprocessisslowandlabourintensivecomparedtoconventionalrulebaseddesign.Thishasbeen
demonstrated by the two design methods running side by side in this thesis. There are specialist
classificationsocietiescomputerprogramssuchasGLShipModel[GermanischerLloyd(2010)]that
canbeusedtospeeduptheprocesswhencomparedtotraditionalFEprogramssuchasANSYS.

TherearehoweveranumberofbenefitstotheuseofFEMforshipdesign,namely:

Areductioninscantlingstoachieveamoreefficientdesign(bothsteelweightandproductioncosts)
[GermanischerLloyd(2010)].

The ability to investigate fatigue life, buckling capacity and vibration characteristics along with
ultimatestrength.

7.2.4

Finiteelementsailrigmodel

InthepreliminarydesignstageEulerbeamtheorywasappliedtovalidatethestrengthofmajorcomponentsof
Multiwingstructure.Howevertheassembledstructurehassomedegreeofcomplexityandthisnecessitates
use of alternative methods. Therefore to evaluate the structural rigidity of the rig, finite element methods
(FEM)usedinthemidshipsectionanalysiswillbereintroduced.
The rig model of Multiwing system will be simplified to increase the efficiency maintaining the general
structural features. The rig is to be modelled for two conditions which represent upwind and downwind or
storageconfiguration.Fortheloading20knots(15knotsfordownwind)and70knotsofwindspeedareapplied
for the average operational condition and for the worst weather condition respectively. The results will
illustrate the maximum deflection, stress and stress concentration for each case and buckling behaviour in
storageconfiguration.
Modelling
TheANSYSmodelofthefullscalesystemconsistsofshell,beamandpipeelements.Sincethemastandstocks
areofconventionalpipeshape,thePIPE16elementisusedandforthetopandbottombarBEAM4isapplied.
To simplify the structure thin plates are created to represent the wing surfaces using SHELL63 and the
connectivityoftheseelementsisverified.ThemodelgenerationisillustratedinFigure7.28.
Thentherealconstantsarecalculatedandappliedtothemodelwithmaterialpropertiesof5083aluminium
alloyand0.4%carbonsteelconsidered[Calvert&Farrar(1999)].Intheinitialmodelling5083aluminiumalloy
isusedforallthecomponentstoensurethelightweightofthestructure.
Loading
TheloadingappliedtothestructureiscalculatedusingEquation(F.1)withappropriateliftcoefficient.Forthe
wingsectiontheliftcoefficientof2.20inupwindconditionfromXFoilanalysisandliftcoefficientofflapplate,
1.28 is used in downwind condition. In this analysis it is assumed that the wind loading on the mast and
connecting structure is negligible therefore only the pressure on the wing plate is to be applied. From the
equation the pressures applied are 138 Pa, 45 Pa for the upwind and downwind operational condition and
1689Paand982Paforupwindanddownwindworstweathercondition.Thepressureisappliednormaltothe
platesurfaceandevenlydistributed.FortheboundaryconditionsofthemastendsallDoFareconstrainedin
bothcasesandthetopofthestructureisfreeend.

144

Design Development

Figure7.28Upwindmode(left)anddownwind/storagemodel(right)inANSYS

Results
Having applied nodal solution the results in four cases are shown in Figure 7.29 and Figure 7.30. Since the
displacementandstressdistributiontrendareidenticalinbothwindconditions,foroperationalwindspeed,
stress distribution is displayed, and for worst weather, the displacement is displayed. The results are
summarised in Table 7.23 together with theoretical prediction of a single main wing stock using Euler beam
equationsforfixedfreeendwithevenlydistributedloadwhichismorereasonablethanthepreliminarydesign
analysissincethetopbarisnotconstrained.

Figure7.29Upwindmodelsolutioninoperationalcondition(left)andworstcondition(right)

145

Concept Design of a Fast Sail Assisted Feeder Container Ship

Figure7.30Downwind/storagemodelsolutioninoperationalcondition(left)andworstcondition(right)

Table7.23TheoreticalandANSYSsolutioninoperationalandworstweathercondition
Model
Operational
Worst
weather

Upwind
Downwind
Upwind
Downwind

Theoreticalestimation
Max.deflection/
Max.stress /
m
MPa
0.68
68.73
0.22
22.54
8.26
845.00
4.80
492.00

ANSYSsolution
Max.deflection/
Max.stress/
m
MPa
1.641
146.0
0.576
51.2
20.100
1790.0
12.574
1120.0

Theworstweatherconditionresultsinexcessivedeflectionandstresssincethetargetmaximumdeflectionand
maximumstresswereonemetreandhalftheyieldstressrespectively.TheANSYSsolutionproducedresultsof
morethantwicethemaximumallowabledeflectionandstress.Thisrequiressomemodificationofthemodels.
Modification
To maximise the efficiency the simplest and straightforward modifications will be undertaken before
considering design modifications involving additional structure or remodelling. Two modifications, changing
materialwithhigherYoungsmodulusandincreasingdimensionsareundertaken.Twocomponents,thestocks
andbottombararemodified.
A direct way to reduce the stress concentration on the stock is to increase the diameter from 0.45 to 0.7
metresandthethicknessofthewallfrom0.03to0.05metreswithinthedesignallowanceforwingfitting.This
resultsin6.2timeshighersecondmomentofareahencesmallermaximumdeflectionandstressareexpected.
By observation of the upwind model deflection it is noticed that small inclination of the bottom bar may
produce large deflection on the top bar. One degree inclination of the bottom bar results in 0.43 metres
deflection on the top bar. Therefore the bottom bar second moment of area needs to be increased. By
changingthematerialfromaluminiumto0.4%Csteelandincreasingthewidthfrom1.5to2metresandthe
thickness from 0.09 to 0.15 metres the second moment of area is improved in all directions. Figure 7.31
illustratesthechangeinstressdistribution.
It can be seen that after modification the stress is distributed to mast and other components. The results
summarised in Table 7.24 shows significant improvement in both maximum stress and deflection. Although
themodificationdoesnotsatisfythetargetvaluesitwasclearlyobservedwhichstructuralelementsarecritical
soitcancontributetheefficiencyoffurthermodification.Themodificationinvolvessignificantweightincrease
duetochangeindimensionandmaterialbutsincesecondmomentofareacanbeincreasedusingstiffeners

146

Design Development
andgeometricmodificationwithoutsignificantincreaseinweighttheoriginalweightestimationusingexisting
aluminiumwingdataisusedinotherstructuralanalysisofthisproject.

Figure7.31Thestressdistributionbefore(left);andafter(right)modification
The last case to investigate is buckling in modified downwind/storage model. When the stored rig elevates
collision may happen due to obstacles on the way. Therefore this buckling analysis can determine the
maximumspeedorlimitofliftingforcetoavoidbuckling.Tosimulatebucklingunitforce,1Nisappliedonthe
top of each stock. Using the buckling factor the buckling load of 1480 kN is calculated which is bigger than
theoreticalvalueof821kNforasinglewingplate.Figure7.32illustratesthebucklingphenomenon.

Figure7.32Bucklingmodeofdownwind/storagemodel
Thefigureshowsthebucklingoccursontheinnerplatesofthesidewingplates.Thismaybecausedbythe
deflection of the top and bottom bar since there is no gap between the plate and the bars in the model.
Thereforewhendesigningtherealrigstructureareasonableamountofclearancebetweenthesetwoelements
needstobetakenintoaccounttoavoidwingbuckling.Thebucklingloadcalculatedmaynotbevalidinreal
wingsectionbuttheproceduremayberequiredtodeterminetheliftingspeedandforcelimitoftherig.

147

Concept Design of a Fast Sail Assisted Feeder Container Ship


Table7.24Sumaryofmaximumdeflectionandstresscomparingoriginalandmodifiedsailrigdesigns
Originaldesign
Max.
Max.stress
deflection/m
/MPa
Upwind
1.641
146.0
Downwind
0.576
51.2
Upwind
20.100
1790.0
Downwind
12.574
1120.0

Model
Operational
Worst
weather

Max.
deflection/m
0.245
0.079
3.002
1.716

Modifieddesign
%of
Max.stress
%of
original
/MPa
original
14.9
31.60
21.6
13.7
9.76
19.1
14.9
387.00
21.6
13.6
213.00
19.0

Modelvalidation
To validate the model generated the results are reproduced in different mesh sizes. Four different element
numbers are used as shown in Table 7.25. The smallest element number 251 produced larger maximum
deflectionandstressbutfromelementnumber339thechangeinvalueisinsignificant.As451elementsare
usedinbothmodelstheerrorduetomeshsizewillbeinsignificanttotheoverallanalysis.
Table7.25Effectofchangingmeshsizeonmaximumdeflectionandstress
No.ofElement Max.deflection /m Max.stress /MPa
251
0.283270
36.7
339
0.245389
31.6
451
0.245273
31.6
592
0.245278
31.6

7.3

Stability

7.3.1

Intactstability

For the intact stability of the fast feeder concept, IMO (2008b) regulations were applied and additional
criterions from LY2 [MCA (2007)] were used to look at the differences due to the sail system. However it
shouldbenotedthatthefastfeederconceptisnotrequiredtomeettheLY2criterionsinceitisindifferent
category. The analysis was undertaken using a commercial software, Hydromax using the built in criteria
analysistool.
Tankandcompartmentdefinition
Priortotheanalysis,thetanks,compartmentsandmainitemsweredefinedtodeterminethelongitudinaland
verticalcentreofgravitybasedontheweightestimation.Thismodel,definedusingHydromax,isillustratedin
FigureL.1.Thenfourloadcaseswerecreatedensuringthatthedraftrangelieswithinthemaximumdraftlimit
of 9.5 metres for operating in some ports and the minimum draft limit for complete emergence of the
propellers which requires 6.4 metres includingclearance. The four load cases are summarised in Table 7.26
withhydrostaticsinequilibriumconditionandthefullitemlistsareinAppendixK.3.
IMOintactstabilitycriteria
The IMO codes used for the intact stability comprise Annex International Code on Intact Stability, 2008
Chapter 3.2 for all ships and Chapter 4.9 for container ships greater than 100 metres. With coefficients
calculated in Appendix L.2, the required criterions were obtained and updated on the IMO criteria database
ready for the large angle of heel analysis. To obtain the wind heeling lever, lw1 or severe wind and rolling
criteria,thewindageareaandheelingleverarerequired.Thereforetheworstcaseisassumedwhereallsix
wingsareexposedtothewinddirectionhavingthelargestarea.

148

Design Development
Table7.26Hydrostaticsoffourloadconditionsinequilibriumcondition
Fullload
departure
8.942
20483
8.871
9.013
8.947
0.143
158.719
4821.082
3236.991
5.219
9.874
7.251
0.0512

DraftAmidships/m
Displacement/t
DraftatFP/m
DraftatAP/m
DraftatLCF/m
Trim(+vebystern)/m
WLLength/m
WettedArea/m2
WaterplaneArea/m2
KB/m
KGfluid/m
BMt/m
Trimangle(+vebystern)/degree

Fullloadarrival
8.935
20416
9.04
8.829
8.927
0.211
158.097
4804.151
3217.485
5.207
9.805
7.218
0.0757

Ballast
departure
7.004
14506
6.483
7.525
7.014
1.042
158.239
3954.269
2831.904
4.078
8.434
8.099
0.3735

Ballastarrival
7.011
14522
6.618
7.405
7.018
0.787
157.779
3949.572
2819.949
4.079
8.55
8.059
0.282

Table7.27valuesusedincalculationofwindheelinglever
Area/m2
1664.75
246.67
118.35
21.6
151.44
908.64
3035.73

Hull
Superstructure
Container Bay
Mast
Sail (singlewing)
Totalsailarea
Worst condition

Z/m
14.00
23.82
23.82
24.37
42.39
42.39
18.44

Fourloadcaseswereanalysedandsatisfiedallcriteriainthecodes.TheGZcurvesforallconditionsaregiven
inAppendixL.2.TheresultsaresummarisedinTable7.28andfurtherdetails,includingpercentagemargin,are
giveninTableL.2.
Table7.28Summaryofstabilitycharateristicsforanalyisdshipconditions
Condition

InitialGM@

0 /m

Max. GZ/m

(Max.GZ)/deg

Fullloaddeparture

2.596

2.520

59.1

Fullloadarrival

2.620

2.576

59.1

Ballastdeparture

3.743

4.180

68.2

Ballastarrival

3.588

4.061

68.2

LY2intactstabilitycriteria
Sincethereisnostabilitycriteriadevelopedforsailassistedcontainerships,LY2formonohullsailingvessels
[MCA(2007)]wasusedtoinvestigatethedifferencebetweenIMOandLY2criteria.Therearethreecriterions
from LY2, which involve the range of positive stability, angle of equilibrium and angle of downflooding as
showninTable7.29.

149

Concept Design of a Fast Sail Assisted Feeder Container Ship


Table7.29LY2Monohullsailingshipcriteria
Code
Criteria
/degree
11.2.2.1.2 Rangeofpositivestability
Shallbegreaterthan
90
11.2.2.1.3 Angleofequilibriumderivedwindheelingarm Shallbegreaterthan
15
11.2.2.1.3b Angleofdownflooding
Shallnotbelessthan
40

Therangeofpositivestabilityissatisfiedinallloadconditionsbuttheothertwocriterionarenotsatisfied.This
is due to nonwatertight walkway compartments along the side of the ship, which cause low downflooding
angle.Therelativelyhighdownfloodinganglerequiredmaybeduetothenatureofsailingyachtswhere20to
25 degree heel is normal upwind sailing condition which is not acceptable in most commercial vessels. The
main difference between the severe wind in IMO and rolling criteria and LY2 is the former uses the actual
windage area, wind heeling lever and dynamic pressure whereas the latter derives the heeling lever from
downfloodingangleor60degreeswhicheverisleastasshowninFigure7.33andthefollowingequations.

Figure7.33DerivedwindheelingleverwithtypicalGZcurve
Thederivedwindheelinglever(dwhl)atangleisequalto0.5 WLO cos . whereWLO GZ cos . .
Noting that: WLO is the magnitude of the actual wind heeling lever at zero degrees which would cause the
vesseltoheeltothedownfloodingangle or60degreeswhicheverisleast; istheangleatwhichthe
derivedwindheelingcurveintersectstheGZcurve.

7.3.2

Damagestability
Compartmentdefinition

Bulkheads were defined in accordance with the requirement from Part 3,Chapter 3 of Lloyds Rules [Lloyd's
Register (2009)] that the ship have at least eight watertight bulkheads over its length. For consistency of
loadingitwasdecidedthatallcargoholdbulkheadswouldbeassumedfullywatertight,givingatotalofnine
fullbulkheads.AwatertightbulkheadwasalsolocatedundertheengineroomfloorforwardsoftheLNGplant
room.
The damage stability of the concept was assessed against IMO probabilistic damage stability criteria from
SOLAS Chapter II1, Part B1 [IMO (2009)]. This relies at its core on the ship being divided into a series of
longitudinally defined damage zones which in this case were chosen to align with the major transverse
bulkheads.AsaresultoftheminorsubdivisionforwardoftheLNGspaces,theengineroomcompartmentwas
includedintwodamagezones.ThedefinitionoflongitudinaldamagezoneswithisshowninFigure7.34.

150

Design Development

Figure7.34Locationofdamagezonesalongtheship,alignedwithwatertightsubdivision
Probabilisticcalculations
SOLASprobabilisticstabilityconsidersavesseltobesuitablysubdividedifitsattainedsubdivisionindex,isnot
lessthanarequiredvalue[seeEquation(7.9)]whichisdependentonthesubdivisionlength .Theattained
indexiscalculatedbyassessingtheprobabilityofoccurrenceofaparticulardamagecase( )andsubsequent
survival( )atthreeoperationaldraughts,thedeepestsubdivisiondraught ,apartialsubdivisiondraught
and a light service draught
[Equation(L.3)].

, and summating weighted contributions of the partial subdivision indices

128
152

1
0.4
where

0.60335

0.4

0.2

(7.9)

(7.10)

, ,

Calculationsforadamagecasemaybebasedonasingledamagezone,oracombination.Whilethedamage
zones account only for transverse subdivision along the ship, the factor incorporates the influence of
longitudinal, based on an IMO statistical analysis. The probability of survival, , is then calculated from the
positive righting lever range in the damaged condition, with a reduction factor accounting for horizontal
subdivision.
Theattainedindexwilltendtoincreaseasmoredamagecasesareinvestigated,thoughthecontributionsfrom
larger damage lengths will clearly reduce as reduces. Anticipating that the length of the engine room
compartment might lead to too low a survival probability to attain the required index, a number of
combinationswereassessedasshowninTable7.30.
Table7.30Damagecasesconsideredinthestabilityanalysisandtheirconstituentlongitudinaldamagezones
DamageCase,
Zone(s)

1
1

2
2

3
3

4
4

5
5

6
6

7
7

8
8

9
9

10
10

11
11

12
8,9

13
9,10

14
10,11

Theprobabilities and werecalculatedinaccordancewiththeprocedurefromSOLASChapterII1,PartB1


[IMO(2009)]giveninAppendix L.3,withtheresultingindexvaluessummarisedinTable 7.31.Itisclearthat
with the initial compartment definition, the required subdivision index is not only attained but substantially
exceeded, suggesting that a lesser extent of subdivision might be possible without jeopardising the damage
stabilityoftheship.

151

Concept Design of a Fast Sail Assisted Feeder Container Ship


Table7.31Calculatedsubdivisionindexes
0.603
0.784
0.737
0.804
0.838

7.3.3

Freeboardandtonnage
Freeboard

ThefreeboardfortheconceptdesignhasbeendeterminedinconjunctionwiththeLoadLineConvention[IMO
(1966/1988)]. This convention essentially prescribes the safe loading limit of a ship and markings that
representwherethemaximumdraughtsarelocatedhavetobedisplayedontheshipshull,foreachoperating
water type, as can be seen in the General Arrangement (Appendix M). This has been done to conform to
internationalregulations,howevertheresultsisnotcrucialtothedevelopmentoftheconcept.Thisisdueto
themaximumallowabledraughtbeingmuchlargerthantheshipisdesignedtooperate.Thefreeboardsare
summarisedinTable7.32.
Table7.32Summaryoffreeboardrequirements
Summerfreeboard/mm

4488.5

Summerdraught/mm

14500.0

Tropicalfreeboard/mm

4186.4

Winterfreeboard/mm

4790.6

Freshwaterfreeboard/mm

4160.9

Tonnage
Tonnage is of more interest as it is these values that port dues and other costs are often based in the
operational maritime environment. It is therefore often desirable to minimise the tonnage. However, the
determination of this is prescribed by the International Tonnage Convention [IMO (1969)] which aims to
preventconflictingmeasurementsoftonnageforaparticularship.Itshouldbenotedthatdifferenttonnage
certificatesexistfortransitthoughttheSuezCanal,forexample.Thisisessentiallyforthegainofthatstate
andneednotbeconsideredforthisconcept.UsingtheInternationalConventiontherearetwomeasuresof
tonnage, gross and net. Gross is essentially a measure of the entire enclosed volume of the ship and net a
measure of the cargo volume, although for a container ship this does not include containers stowed above
deck.Bothofthesetonnages,althoughbeingmeasuresofvolume,aredimensionlessandaresummarisedin
Table7.33.
Table7.33Tonnagemeasurements
Grosstonnage

18463.7

Nettonnage

8284.6

7.4

Layout and arrangement

Whatfollowsisabriefexplanationofareasofthevesselarrangementnotdiscussedinthepreviouschapters.
The design methodology behind the layout was to locate the accommodation and engine room forward to
maximise the container stowage and minimise aerodynamic drag through the use of an aerodynamic

152

Design Development
accommodationblockshape.Allthemainexternalfeaturesoftheaccommodationblock,suchaslifeboatsand
external stairs are enclosed within the outer skin of the accommodation block to reduce aerodynamic
resistance.Crewingandspacerequirementswithintheaccommodationblockweresizedusingbasisvessels,
Watson & Gilfillan (1977), International Labour Organisation (1970) and a Nigel Gee concept design vessel
[CentrefortheCommercialDeploymentofTransportationTechnologies(2006)].
Duringthearrangementofthevesselitwasfoundthattheoveralllengthofthevesselhadtobeincreased(by
two metres) to fully accommodate the azipod and allow for a bulwark at the bow. This arose due to
insufficient consideration to the dimensions of the azipod and final layout at the initial design stages. This
dimensional change will only have small influences on the rest of the concept as the underwater form
remainedunaffectedwiththemostnotablechangebeingasmallincreaseinstructuremass.
Thecargostowagearrangementswerearrangedwithineighthatchcoverlesscargoholdswithtwoadditional
rowsofreefercontainersondeckjustaftoftheaccommodationblock.Theuseofhatchcoverlessholdswere
usedbecauseitplacesnorestrictionsonthestowageofhighcubecontainers17[Mash(2009)]forwhichthereis
increasingdemandduetotheirgreatertransportefficiency.Currently29%of20ftcontainersand55%of40ft
containersarehighcubes.By2020itispredictedthat80%ofcontainerscouldbehighcube[Lloyd'sRegister
(2006a)].
ThestorageofLNGwaslocatedwithinatankunderneaththeenginespaceallowingfora5.4metrescofferdam
to the side shell (Section 7.3.1) and assumed to be unrestricted by shape requirements. The use of
containerisedLNGwasalsoconsideredbutneglectedduethelargervolumeandresultinglossincargospace.
Theunusualformoftheengineroomduetoitspositionwithintheshipandpropulsionmachineryinstallation
warranted special consideration to the engine room layout. Machinery was selected from manufacturers
websitestocoverthemaingroupsof:
o

machinery,including
o

engines;

powertakeoffs;

exhaustsystemandsilencer;

electricmotors;

electrical,including
o

switchboardandcycloconverters;

transformers;

frequencyconverter;

systems,including
o

coolingsystems;

freshwatersystem;

wastesystem;

lubricatingoilsystem;

refrigerationsystem;

17
Ahighcubecontaineris2896mminheightratherthan2591mmforastandardcontainer[HapagLloyd()].

153

Concept Design of a Fast Sail Assisted Feeder Container Ship


o

firefightingsystems;

ballastsystems;

LNGprocessingplant;

engineroomcrane;

Spaces were also allocated to a machinery control room, electrical and engineering workshops; spare gear
store;changingandlaundryareas.ThelayoutwasarrangedinamannertodeterminethefeasibilityofaLNG
diesel electric propulsion system located in this location and not intended as a detailed arrangement.
Sufficient space has been left around the various machinery items to account for the smaller items of
machinerythathavenotbeenconsidered.
Amachineryarrangementforthethreemachineryspacesisillustratedonthegeneralarrangementdrawingin
AppendixManddemonstratesthatadequatespacehasbeenachievedwithaforwardlocatedaccommodation
blocktoaccommodatealloftherequiredmachinery.
Thegeneralarrangementandtankplanalongwithadiagrammaticillustrationofthedecklayoutsisgivenin
AppendixM
To demonstrate the feasibility of the cargo handling equipment, diagrammatic representation of the gantry
crane operation around a folded sail system is shown in Appendix M. Little attention has been paid in this
thesistothedetailedengineeringofthesailsystemfoldingmechanism.Atthisstageaninitialindicationofa
likelyfoldingmechanismhasbeengivenwiththedetaileddesignafutureconsideration.

154

Conclusions

8. Conclusions
Thefastfeederconcepthasbeenshowntobebotheconomicallyandenvironmentallyfeasiblebasedonthe
analysesdetailedinthisreport.Thebestimprovementinperformanceovertypicalexistingshipsisseeninthe
Caribbean region. Here a 42% reduction in CO2 emissions has been predicted as well as a 33% reduction in
daily costs, whilst meeting the market share for the 2020 container market initially assumed. It should be
notedthattheassumptionofmaintainingaconstantnumberofsailingshasnotbeenadheredtoo.Inorderto
replacetwoexistingshipswithonefastfeederthenumberofsailingshasdecreasedby25%soasnottounder
utilisethevessel.TheperformanceoftheshipintheSingaporeregionisalsofavourableandonlyslightlylower
intermsofCO2andcostsavingscomparedtotheCaribbean.
It is noted that numerous environmentally conscious ship concepts are currently being proposed within the
maritime industry [Byklum (2010); NYK (2010); Wrtsil Corporation (2009b)], and as such the fast feeder
reflects changing attitudes towards ship design. Of particular interest is the novel container ship concept
developedbyDNV[Byklum(2010)]whichreflectsmanyoftheunconventionaldesignfeaturespresentinthis
report.Theseincludelowblockcoefficient,LNGfuelledengines,poddeddrivesandimprovedcargohandling
efficiency. This leads to the conclusion that the fast feeder concept has adopted design and operational
approachesthatwillbecommerciallyviableinthefuture,furtherjustifyingthisstudy.
The main conclusions regarding the achievements of the project are summarised as follows. Regarding
hydrodynamicmodeltesting:

Asidefromthequantifieduncertaintytheresultsfromthemodeltestsaregiveanaccuratenakedhull
formperformancepredictionfollowingappropriatepostprocessing.Noaccounthasbeentakenfor
appendages.Intermsofuprightresistancethisisaccountedforbyadjustingthepropulsiveefficiency.
The effect on side force and induced resistance when the ship is under sail is not considered. It is
expectedthattheeffectoftheappendagesissignificant,duetothelargesizeofthepoddeddrive(s);

ThetestingofthetwoshipmodelsisjustifiedbythenovelformofHullB.Validationofthenumerical
resistancepredictionswasimportantsinceHullBwaspoorlyrepresentedbytheHoltrop.regression;

Model testing of both hull forms in waves was important to evaluate voyage performance more
realistically.Thisalsoprovidescomparisontonumericalmethodswhosevalidityhasbeenfoundtobe
questionable;

Inhindsightthenumericalpredictionoftheaddedresistanceinwavesshouldhavebeencarriedout
prior to model testing. This may have allowed more reasonable estimation of the wave lengths to
test,possiblyenablingthecaptureofthepeakinaddedresistanceforeachhullatbothtestedspeeds;

Regardingaerodynamicmodeltesting:

Thechoiceofthemodelscaleplayedafundamentalroleinobtainingaccurateresults.Theanalysis
proves that the forces generated by the sails are highly dependent on the Reynolds number and a
furtherincreaseinthesailcoefficientwouldbeexpectedforthefullscalerig;

Theinvestigationofoptimumsailrigconfigurationhasbeenvaluableinincreasingtheliftgenerated.
The results ofthe spacing optimisation agree with General Biplane Theory [Munk (1923)] whilst the
investigationonthestaggereffectshowsthattheperformanceoftheMultiwingcanbeimprovedby
thesloteffectcreatedbetweentheupwindanddownwindwings;

Examining sailcontainer interaction has also led to a benefit in terms of performance, due to the
reduction oftheinduceddragcomponent.Thishasbeenpredictedasa15%increaseingenerated
thrust;

155

Concept Design of a Fast Sail Assisted Feeder Container Ship

The limitations of the wind tunnel testing meant detailed rigrig interaction could not be achieved.
Theseeffectsarenotfullyaccountedforintheoverallpredictionofperformance.ThustheuseofCFD
has been valuable in showing that the two rigs must be independently controlled to maintain
optimumoverallsailsystemperformance;

Regardingconceptfeasibility:

A detailed study into 2020 container market has been carried out. This analysis does not include
extrapolationpastthisdate.Assuch,thegrowthinthecontainermarketthroughouttheoperating
lifeofthefastfeederhasnotbeentakenintoaccountwhenassessingfeasibility.However,sincethe
comparison ships are operating in the same environment, this simplification is not considered too
onerous.

Thethrustprovidedbythesailsislow,asexpected,especiallyathighspeed.Anadditionallimitation
onthesailrigwastherequirementtofoldinsidethehullsoasnottohindercargohandling.Thisis
largelyduetotheshiptype,andnonretractablesailscouldhaveprovidedlargersailarea.However,
theinfluenceofthesailsonshipstabilityistheultimatelimitation;

Theattainedthrustreductionduetotheauxiliarysailpropulsionisnotaslargeasinitiallyestimated.
At25knotsa3.0%reductionwaspredictedwhereasonly12%reductionwasachieved;at15knotsa
10.0%reductionwaspredictedwhereasonly26%reductionwasachieved,dependingtotheroute.
Althoughthewindtunnelresultsshowimprovedrigefficiency,intermsofdrag,comparedtoinitial
estimates, other factors that were crudely accounted for or disregarded in the initial design stages
have resulted in a net reduction in the magnitude of thrust reduction. These factors include
estimationsofcalmwater,inducedandaddedresistance;

Thefinalconceptcanoperatesafelyinconditionsupto30knotstruewindspeed(forceseven)and
6.5 metres significant wave height, without affecting the scheduling of the service. This is
considerablybetterthanexistingfeederships(Section2.5);

Account has been taken of the typical operation of feeder services and attitudes of operators. It is
acknowledgedthatmanyofthedesignandoperationalconceptsintroducedinthisstudywillnotbe
feasible immediately. However, the conservative nature of the maritime industry has not been
allowedtohinderthedesigndevelopment,anditisassumedthatmanyofthetechnologiesproposed
will be commonplace in 2020. It is noted that marine LNG terminals already exist in both the
Singapore and Caribbean regions [Wrtsil Corporation (2009b)], so in this respect, the adoption of
LNGasamainpropulsionfuelisconsideredentirelyfeasible;

While the performance evaluation of the fast feeder concept is considered to be realistic, since the
probabilityofwindandwaveconditionsisincludedinthePPP,noaccountistakenfortheseeffectson
the performance of the comparison ships. This means that efficiency gains may have been under
predicted,sincethemarginincludedinthecomparisonshipsschedulehasbeenassumedtobedueto
waitingtimeinport,notweatherdelays;

Theeconomicfeasibilityanalysisisconsideredtobefairlysimplistic.Thisisduetoalackofcostdata
forthefastfeeder,especiallyrelatingtotheaddedconstructioncostattributedtothepoddeddrives
and LNGelectric propulsion system. Thus the confidence placed in the results of the economic
assessment is not as high as that for the environmental analysis. The operation of the Multiwing
provestobeprofitablewhenthebenefitsderivedfrommotiondampingareconsidered;

156

Conclusions
Thedesignimplicationsofthenovelconceptwerenotfullyrealisedattheoutsetandwerenotaccountedfor
earlyonintheproject.Theseinclude:

The unconventional mass distribution of the ship due to the sternmounted podded drives and
forwardengineroomleadstohogging.However,thestructuraldesignpassesallclassrulesregarding
globalstrength.Inaddition,thisleadstodifficultiesingettingthevesseltofloatonanevenkeelasis
desirableforefficientoperation.

TheoptimisationofthehullformforusewithsailsTheperformanceofthecandidatehullformsin
thisconditionwasnotunderstooduntilaftertheresultsofthemodeltestinghadbeenanalysed.Itis
anticipatedthatagreaterthrustbenefitcouldhavebeenachievedifthishadbeenconsideredduring
hullformdevelopment.Howevernumericaltechniquestoevaluatethiswerenotavailable.

An additional benefit of the sail system was also not initially foreseen. The ability of the sails to
providemotiondampinghastheadvantageofreducingrollandyawmotions,whichinturnreduces
inducedresistancewhensailingtowindward,aswellasstructuralandcargoloads.Sincetheanalysis
ofmotiondampingwasnotcarriedoutuntillateoninthedesigndevelopment,theseeffectscould
notbeincorporatedintotheresistanceandseakeepingpredictions,orFEAmidshipsectionmodel.

Computational techniques, namely CFD and FEA, have been used to great advantage in the
demonstration of the fast feeder concept viability. This is important, since in such unconventional
designs simple design methods may not be rigorous enough. Novel design aspects are not always
covered by rules and regulations that the vessel must adhere to and thus more intense numerical
methodsmustbeappliedtoprovetheconceptviability.

Onefullloopofthedesignspiralhasbeencompleted,asproposedattheoutsetoftheproject.However,this
meansthatthefinaldesigncouldbefurtherrefinedbyiteratingthroughthespiral.Thefocusinthisreporthas
beenonreducingresistancethroughhullformoptimisation,yetanadditionalpassthroughthedesignspiral
could focus on reducing structural mass. This could allow deadweight to be maximised or ship size to be
reduced,bothofwhichwouldimproveefficiency.
Fromareflectivepointofview,theorganisationof,andconstraintsontheproject,havedirectlyaffectedthe
outcome.Thishasmanifestedinbothpositiveandnegativeeffects.

Theefficientorganisationandearlyallocationoftaskswithinthegrouphasmeantabroadscopeof
work has been completed in the time available. The use of a quality checking system has also
improvedthequalityofthefinaldesign.

The submission of model plans by week 7 meant that the hull forms developed are not necessarily
considered the most appropriate in hindsight. The limited time available resulted in unsatisfactory
hull fairing. Ideally, more detailed design would have been carried out during the hull form
development. This could have included: design for sailing performance; detailed propulsion
calculationstoinformsterndesign(avoidingtheproblemsencountered(Section7.4);andtheuseofa
geneticalgorithmtorapidlyoptimisethehullformbasedonmultiplerequirements.

157

Concept Design of a Fast Sail Assisted Feeder Container Ship

9. Future Work and Recommendations


The recommendations listed can be divided under twomain headings: those that would have benefited this
project but were not carried out; and those that are outside the scope of this work but would enhance the
robustnessoftheproposedconcept.
Significantimprovementscouldhavebeenmadetotheinvestigationby:

Conducting hull form optimisation for performance in waves, as well as reduced vessel motions. In
addition, the PPP has not accounted for added resistance in stern quartering seas, and thus the total
magnitude of added resistance over a typical voyage may be underestimated. A reduction in added
resistance,aswellaspitchmotions,maybeevaluatedfromananalysisofthepitchdampingduetothe
sails. This has not been carried out here since the appropriate inputs to the theoretical model are not
available.

Incorporating roll damping predictions into cargo securing and midship FE analyses, providing more
realisticresults.

Incorporating the concept of weather routing into the PPP so as to optimise the voyage simulation in
termsofminimisingpowerrequirement.

Making improvements in manufacturing techniques. The main concern is in refining the surface finish
exhibitedbythemodelssoastoreducetherelianceonroughnesscorrectionsandincreaseconfidencein
thetestresults.

Researching and quantifying other methods for improving efficiency such as hull coatings [Willsher &
Solomon(2010)].

Investigatingoftheeffectofthegapsizebetweenthecontainersandrigonthrustincrease.Itisexpected
that further reducing the gap will increase the thrust, based on the containersail interaction results
alreadyobtained.

CompletingmorecasesinvestigatingtheinteractionbetweentheforwardandaftsailrigsusingCFD.This
willallowagreaterunderstandingofthelevelofinteractionforarangeofwindspeedsandangles.The
results of such an investigation could also be incorporated into the PPP to improve performance
prediction.

Improvementsinperformancethroughmoredetaileddesignthatwerenotpossibleinthisprojectinclude:

Reiterationofthedesignspiral.Thiswouldaimtooptimisethehullforminthesailingconditionaswellas
forcalmwaterresistance.

Detailed numerical simulation or physical testing of the propulsor hull interaction. This could include
model selfpropulsion tests; wake traverse measurements; tests with appendages (podded drives); and
CFDsimulationsoftheinflowintothepropellers.

MoredetailedinvestigationandunderstandingofthefeasibilityofusingLNGasamarinefuel.Inaddition,
theenvironmentalcostincurredinproducingandtransportingthefueltothebunkeringlocationshouldbe
accounted for. The viability and environmental impact of using biofuels could also be considered. It
follows that the life cycle emissions of the fast feeder, including construction and scrappage should be
estimated.Thishasnotbeenconsideredatallinthiswork,howeveratrulyenvironmentallysustainable
shipshouldtakeaccountofallemissionsincurredthroughoutitslife.

158

References

References
ABB(2009)ProjectGuideforAzipodVOandVISeries.
AmericanBureauofShipping(1988)GuideforCertificationofContainerSecuringSystems.
Andersen,P.andKristensen,H.O.(2009)DesignforEfficiencyofPerformance.Proceedingsofthe10th
InternationalMarineDesignConference,Trondheim.
Barlow,J.B.,Rae,W.H.andPope,A.(1999)LowSpeedWindTunnelTesting.Thirdedition.Wiley,NewYork.
Bergeson,L.andGreenwald,K.C.(1985)SailAssistsDevelopments19791985.JournalofWindEngineeringand
IndustrialAerodynamics,19(13),pp45114.
Bertram,V.(2000)PracticalShipHydrodynamics.ButterworthHeinemann,Oxford.
Blomqvist,P.A.andForrest,D.L.(2000)LaserCuttingandWeldingofDDG51StructuralShapes.TheJournalof
ShipProduction,16(4),pp18392.
BMTNigelGee(1998)NorasiaDataSheet.Availablefrom:
http://media.bmt.org/bmt_media/resources/29/NG254Norasia.pdf,accessed15/10/2009.
Bonafoux,J.,Dudson,E.andSherwood,D.(2001)AnEvaluationoftheEffectofHullFormChoiceonthe
OperabilityofFastFerries.Proceedingsofthe6thInternationalConferenceonFastSeaTransport
(FAST),Southampton.
Braslow,A.L.andHarris,R.V.(1966)UseofGritTypeBoundaryLayerTransitionTripsonwindTunnelModels,
NASA.
BSI(2007)EngineeringDrawingPractice:AGuideforFurtherandHigherEducationtoBS888:2006,Technical
ProductSpecification,
Buhaug,.,Corbett,J.J.,Endresen,.,Eyring,V.,Faber,J.,Hanayama,S.,Lee,D.S.,Lee,D.,Lindstad,H.,
Markowska,A.Z.,Mielde,A.,Nelissen,D.,Nilsen,J.,Palsson,C.,Winebrake,J.J.,Wu,W.andYoshida,
K.(2009)IMOSecondGHGStudy2009,InternationalMaritimeOrganisation.
Byklum,E.(2010)QuantumAContainerShipConceptfortheFuture,ContainerUpdate.
http://www.dnv.com/industry/maritime/publicationsanddownloads/publications/dnvcontainershipup
date/2010/12010/Quantumacontainershipconceptforthefuture.asp.Accessed:19/04/2010
Calvert,J.R.andFarrar,R.A.(1999)AnEngineeringDataBook.Palgrave,Basingstoke.
Campbell,I.(2009)YachtExperimentalTechniquesLectureNotes,UniversityofSouthampton.
Campbell,I.(2010a).Personalcommunication:DiscussioninUniversityofSouthamptonwindtunnel.
22/02/2010.
Campbell,I.(2010b).Personalcommunication:SSUtowingtank.24/02/2010.
Campbell,I.andClaughton,A.(1987)TheInterpretationofResultsfromTankTestson12mYachts.Proceedings
oftheEighthChesapeakSailingYachtSymposium,pp91107.
Carlton,J.S.(2007)MarinePropellersandPropulsion.ButterworthHeinemann,Oxford.
engel,Y.A.andCimbala,J.M.(2006)FluidMechanicsFundamentalsandApplications.McGrawHill,New
York.
CentrefortheCommercialDeploymentofTransportationTechnologies(2006)PentamaranConceptDesign
Lifton/LiftoffContainership,CaliforniaStateUniversity.Availablefrom:
http://www.ccdott.org/transfer/projresults/2004/...31/task%202.31_1.pdf,accessed20/04/2010.
Cheng,F.(2009).Personalcommunication:Sustainableshippingtalk,UniversityofSouthampton.15/10/2009.
Chret,B.(2000)LesVoiles:Comprendre,Rgler,Optimiser.Gallimard,Paris.
ChittagongPortAuthority(2010)TurnAroundTimesofVesselsDuringtheFiscalYear20072008.Available
from:http://www.cpa.gov.bd/new/index.php?option=com_content&view=article&id=95&Itemid=64,
accessed09/04/2010.
Choi,E.C.C.(2004)ReportontheStatusofCodificationonWindLoadinginSingapore.Proceedingsofthe
WorkshoponRegionalHarmonisationofWindLoadingandWindEnvironmentalSpecificationsinAsia
PacificEconomies.Availablefrom:http://www.wind.arch.t
kougei.ac.jp/APECWW/Report/2004/SIN.pdf,accessed13/04/2010.
ClassNK(2008)GuidelinesforContainerStowageandSecuringArrangements.
Claughton,A.andCampbell,I.(1994)WindTunnelTestingofSailingYachtRigs.Proceedingsofthe13th
InternationalHISWASymposiumonYachtDesignandYachtConstruction,Amsterdam.pp89106.
Claughton,A.,Wellicome,J.andShenoi,A.(1998)SailingYachtDesign:Theory.AddisonWesleyLongman
Limited,Harlow.
Clayton,B.R.andSinclair,F.M.(1989)MotionDampingofShipsFittedwithMarineAerofoils.Transactionsof
theRoyalInstitutionofNavalArchitects,131pp12133.

159

Concept Design of a Fast Sail Assisted Feeder Container Ship


Comstock,J.P.(1967)PrinciplesofNavalArchitecture.TheSocietyofNavalArchitectsandMarineEngineers,
NewYork.
Corus(2002)CorusSpecialProfilesBrochure:BulbFlatsforPlateStiffening.Availablefrom:
http://www.corusgroup.com/file_source/StaticFiles/Business%20Units/Special%20Strip/Bulbflat.pdf,
accessed10/02/2010.
Corus(2007)CorusAdvanceSectionsBrochure.Availablefrom:
http://www.corusconstruction.com/file_source/StaticFiles/Construction/Library/Sections/Advance_br
ochure.pdf,accessed10/02/2010.
Cyberiad(2009)Michlet.http://www.cyberiad.net/michlet.htm.Accessed:23/10/2009
Das,S.K.andDas,S.N.(2007)ModelingofCoupledRollandYawDampingofaFloatingBodyinWaves.
MathematicalProblemsinEngineering.
Das,S.K.,Das,S.N.andSahoo,P.K.(2006)InvestigationofSway,RollandYawMotionsofaShipwithForward
Speed.Proceedingsofthe5thInternationalConferenceonHighPerformanceMarineVehicles,
Tasmania.
Dataloy(2009)DataloyDistancetable.http://www.dataloy.com/.Accessed:31/11/09
Degerlund,J.(2004)ContainerisationYearbook2004.Informa,London.
Degerlund,J.(2006)ContainerisationYearbook2006.Informa,London.
Degerlund,J.(2008)ContainerisationYearbook2008.Informa,London.
Doyle,T.,Gerritsen,M.andIaccarino,G.(2002)OptimizationofYardSectionalShapeandConfigurationfora
ModernClipperShip.Proceedingsofthe17thInternationalHISWASymposiumonYachtDesignand
YachtConstruction,Amsterdam.pp7383.
EngineeringSciencesDataUnit(1980)ESDU80024BlockageCorrectionsforBluffBodiesinConfinedFlow.
EngineeringSciencesDataUnit(1995)ESDU76028ALiftinterferenceandBlockageCorrectionsforTwo
dimensionalSubsonicFlowinVentilatedandClosedWindtunnels.
Faltinsen,O.M.(1998)SeaLoadsonShipsandOffshoreStructures.CambridgeUniversityPress,Cambridge.
FederalReserve(2010)FederalReserveStatisticalRelease.
http://www.federalreserve.gov/releases/h15/data.htm.Accessed:16/04/2010
FormationDesignSystems(2005a)HullspeedUserManual.
FormationDesignSystems(2005b)SeakeeperUserManual.
Francesetti,D.G.andFoschi,A.D.(2002)TheDevelopmentofHubandSpokeSystemsintheMediterranean:a
ProblemofSustainableDevelopment,Availablefrom:http://ssrn.com/abstract=385183or
doi:10.2139/ssrn.385183,accessed14/03/2010.
Frankel,E.G.(2002)TheChallengeofContainerTranshipmentintheCaribbean.ProceedingsoftheIAME
Panama2002InternationalSteeringCommittee,Panama.Availablefrom:
http://www.eclac.cl/Transporte/perfil/iame_papers/papers.asp,accessed14/03/2010.
Fujiwara,T.,Hearn,G.E.,Kiramura,F.andUeno,M.(2005)SailsailandSailhullInteractionEffectsofHybrid
SailAssistedBulkCarrier.JournalofMarineScienceandTechnology,10(2),pp8295.
Fujiwara,T.,Hirata,K.,Ueno,M.andNimura,T.(2003)OnAerodynamicCharacteristicsofaHybridSailwith
SquareSoftSail.Proceedingsofthe13thInternationalOffshoreandPolarEngineeringConference
Hawaii.pp32633.
Gee,N.(2007)NewDevelopmentsinContainerVesselDesign,BMTNigelGee,Availablefrom:
http://www.ngal.co.uk/Documents%20&%20Resources/?/126/2865/2865,accessed03/04/2010.
GermanischerLloyd(2010)GLShipModel.Availablefrom:
http://www.germanlloyd.org/en/maritime_software/15450.php,accessed11/04/2010.
Gerritsma,J.andBeukelman,W.(1972)AnalysisoftheResistanceIncreaseinWavesofaFastCargoShip.
InternationalShipbuildingProgress,19pp28593.
Graham,R.,Baitis,A.E.andMeyers,W.G.(1992)OntheDevelopmentofSeakeepingCriteria,.Naval
EngineeringJournal,104(3),pp25975.
Harries,S.(2006)FundamentalsofAdvancedHydrodynamicDesign.TheNavalArchitect,April2006,pp4854.
Harries,S.,Heimann,C.A.J.andHochkirch,K.(2006)AdvancedHydrodynamicDesignofContainerCarriersfor
ImprovedTransportEfficiency.ProceedingsoftheDesignandOperationofContainerShips
Conference,London.pp12129.
Harries,S.,Heimann,J.andHinnenthal,J.(2003)ParentoOptimalRoutingofShips,TechnicalUniversityof
Berlin.Availablefrom:
http://www.futureship.net/downloads/2003_NAV_ParetoOptimalRoutingofShips.pdf,accessed
01/04/2010.
Henschke,W.(1965)SchiffbautechnischesHandbuch,VerlagTechnik.

160

References
Hinrichs,R.(2009)CostofMDOinSingapore.HANSAInternationalMaritimeJournal,146(6),p.15.
Hoerner,S.(1965)FluidDynamicDrag.publishedbytheauthor,BrickTown,NewJersey.
Hogben,N.(1986)GlobalWaveStatistics.UnwinBrothers,Feltham,Middlesex.
Holtrop,J.(1984)AStatisticalReanalysisofResistanceandPropulsionData.InternationalShipbuilding
Progress,31(363),pp27276.
Holtrop,J.andMennen,G.G.J.(1978)AStatisticalPowerPredictionMethod.InternationalShipbuilding
Progress,25(603),pp25356.
Holtrop,J.andMennen,G.G.J.(1982)AnApproximatePowerPredictionMethod.InternationalShipbuilding
Progress,29(335),pp16670.
Hoyle,J.W.,Cheng,B.H.,Hays,B.,Johnson,B.andNehrling,B.(1986)ABulbousBowDesignMethodologyfor
HighSpeedShips.TransactionsoftheSocietyofNavalArchitectsandMarineEngineers,94pp3156.
Hughes,G.andAllan,J.F.(1951)TurbulenceStimulationonShipModels.TransactionsoftheSocietyofNaval
ArchitectsandMarineEngineers,59pp281314.
InternationalMaritimeOrganisation(1966/1988)InternationalConventiononLoadLines,1966,asAmended
bytheProtocolof1988
InternationalMaritimeOrganisation(1969)InternationalConventiononTonnageMeasurementofShips.
InternationalMaritimeOrganisation(2005a)MEPC/Circular.471InterimGuidelinesforVoluntaryShipCO2
EmissionIndexingforUseinTrials(29July2005)AnnexInterimGuidelinesforVoluntaryShipCO2
EmissionIndexingforUseinTrialsAppendixCalculationofCO2indexbasedonoperationaldata.
InternationalMaritimeOrganisation(2005b)ResolutionMEPC.132(53)AmendmentstotheAnnexofthe
Protocolof1997toAmendtheInternationalConventionforthePreventionofPollutionfromShips,
1973,asModifiedbytheProtocolof1978RelatingThereto(Adoptedon22July2005).
InternationalMaritimeOrganisation(2008a)ExplanatoryNotestotheSOLASChapterII1Subdivisionand
DamageStabilityRegulations,ResolutionMSC.285(86).
InternationalMaritimeOrganisation(2008b)InternationalCodeonIntactStability,ResolutionMSC.267(85).
InternationalMaritimeOrganisation(2009)InternationalConventionfortheSafetyofLifeatSea.
InternationalAssociationofClassificationSocieties(2008)CommonStructuralRulesforDoubleHullOilTankers
(CSR).Availablefrom:
http://www.iacs.org.uk/document/public/Publications/Common_rules/PDF/CSR_Double_Hull_Oil_Ta
nker_pdf778.pdfaccessed30/03/2010.
InternationalElectrotechnicalCommittee(2005)IEC61400121:WindTurbinesPart211:Power
PerformanceMeasurementsofElectricityProducingTurbines.
InternationalLabourOrganisation(1970)ILO133InternationalLabourConference(ILO)ConventionNo.133
ConventionConcerningCrewAccommodationonBoardShip(SupplementaryProvisions),PartII.Crew
AccommodationRequirementsArticle5
InternationalShippingCorporation(2010)FleetSpecs.Availablefrom:http://www.intship.com/wp
content/uploads/2009/09/Container_Specs_ISH.pdf,accessed17/03/2010.
ITTC(2002)RecommendedProceduresandGuidelines:7.5010101ModelManufactureShipModels.
ITTC(2005a)RecommendedProceduresandGuidelines:7.5020301.5TestingandExtrapolation,Propulsion,
Performance,Prediction.PowerMargins.
(2005b)RecommendedProceduresandGuidelines:7.5020702.1TestingandExtrapolationMethods,Loads
andResponses,Seakeeping.
ITTC(2006)RecommendedProceduresandGuidelines:7.5020103TestingandExtrapolationMethods,
GeneralDensityandViscosityofWater.
ITTC(2008a)RecommendedProceduresandGuidelines:7.5010301UncertaintyAnalysisInstrument
Calibration.
ITTC(2008b)RecommendedProceduresandGuidelines:7.5020201TestingandDataAnalysisResistance
Test.
ITTC(2008c)RecommendedProceduresandGuidelines:7.5020202TestingandExtrapolationMethods,
GeneralGuidelinesforUncertaintyAnalysisinResistanceTowingTankTests.
ITTC(2008d)RecommendedProceduresandGuidelines:7.5020301.4Performance,Propulsion1978ITTC
PerformancePredictionMethod.
ITTC(2008e)TheSpecialistCommitteeonAzimuthingPoddedPropulsion.Proceedingsofthe25thITTC.
Availablefrom:
http://ittc.sname.org/proc25/assets/documents/VolumeI/Volume1_5PropulsionCommittee.pdf,
accessed23/03/2010.

161

Concept Design of a Fast Sail Assisted Feeder Container Ship


Jacobs,E.N.andSherman,A.(1937)AirfoilSectioncharacteristicsasAffectedbyVariationinReynoldsNumber,
NACAReport586.
Kalmar(2009)AutomaticStackingCraneSystems:MaximiseYourCapacity.Availablefrom:
http://www.kalmarind.com/source.php/1261239/ASCSystems_0909.pdf,accessed27/03/2010.
Keuning,A.,Toxopeus,S.andPinkster,J.(2001)TheEffectofBowShapeonaFastMonohull.Proceedingsof
the6thInternationalConferenceonFastSeaTransport(FAST),Southampton.p.197.
Kim,E.C.,Koh,C.D.,Lee,C.J.,Kim,S.Y.andChoi,Y.R.(2003)Developmentofa30knotClassHighSpeed
ContainerShipCHUNGHAE30.ProceedingsoftheDesignandOperationofContainerShips
Conference,London.pp10511.
Kirkman,K.L.andPedrick,D.R.(1974)ScaleEffectsinSailingYachtHydrodynamicTesting.Transactionsofthe
SocietyofNavalArchitectsandMarineEngineers,82pp77101.
KoiYuNg,A.andKee,J.K.Y.(2008)TheOptimalShipSizesofContainerLinerFeederServicesinSoutheastAsia:
aShipOperatorsPerspective.MaritimePolicyandManagement,35(4),pp35375.
Kracht,A.M.(1978)DesignofBulbousBows.TransactionsoftheSocietyofNavalArchitectsandMarine
Engineers,86pp197217.
Kyriazis,G.(1996)BulbousBowDesignOptimisationforFastShips,MassachusettsInstituteofTechnology.
Levander,O.(2002)LNGfuelledRoPaxVessels.TheShipPowerSupplier:WrtsilCorporation,pp1015.
Availablefrom:http://www.wartsila.com/Wartsila/global/docs/en/ship_power/
media_publications/marine_news/2002_2/levander.pdf,accessed04/03/2010.
Levander,O.(2008a)AlternativeFuelandMachineryTechnologiesforFerries.ProceedingsoftheRoRo2008,
Gothenburg.Availablefrom:http://www.wartsila.com/Wartsila/global/docs/en/ship_power/
media_publications/presentations/alternativefuelandmachineryconceptsforferries
presentation.pdf,accessed27/02/2010.
Levander,O.(2008b)AlternativeFuelsforShips.ProceedingsoftheHllbaratransporter2008,Stockholm.
Availablefrom:http://www.wartsila.com/Wartsila/global/docs/en/ship_power/
media_publications/presentations/alternativefuelsforshipspresentation.pdf,accessed28/02/2010.
Levander,O.(2009)NewConceptsinFerryPropulsion,WrtsilCorporation.Availablefrom:
http://www.wartsila.com/Wartsila/global/docs/en/ship_power/media_publications/brochures/produ
ct/general/2009/Ferrypropulsion2009.pdf,accessed23/03/2010.
Lewis,E.V.(1989)PrinciplesofNavalArchitecture.Volume3.Secondedition.TheSocietyofNavalArchitects
andMarineEngineers,JerseyCity.
Lloyd'sRegister(2006a)ContainersPushSizeandShapeBoundaries.
Lloyd'sRegister(2006b)NewShipRightProcedureforContainerShipsHelpstoAssessHullStresses(LRSDA).
Lloyd'sRegister(2008)CodeforLiftingAppliancesintheMarineEnvironment.
Lloyd'sRegister(2009)LloydsRegisterRulesandRegulationsRulesandRegulationsfortheClassificationof
Ships.
Lloyd,A.R.J.M.(1989)Seakeeping:ShipBehaviourinRoughWeather.EllisHorwoodSeriesinMarine
Technology,
Loukakis,T.A.andChryssostomidis,C.(1975)SeakeepingStandardSeriesforCrusierSternShips.Transactions
oftheSocietyofNavalArchitectsandMarineEngineers,83pp67127.
MacGregor(2010)LashingSystemsDataSheets.Availablefrom:http://www.macgregorgroup.com/?id=7176,
accessed21/03/2010.
Magnucki,K.,Szyc,W.andStasiewicz,P.(2004)StressStateandElasticBucklingofaThinwalledBeamwith
MonosymmetricalOpenCrosssection.ThinWalledStructures,42(1),pp2538.
MANDiesel(2010)MarineEngineIMOTierIIIProgramme.Availablefrom:
http://viewer.zmags.com/htmlCat/index.php?mid=rpwhph,accessed23/03/2010.
Mange,E.(2006)ShortSeaShippingCostBenefitAnalysis.ProceedingsoftheEuropeanTransportConference,
Strasbourg.
Marchaj,C.A.(2003)SailPerformance.AdlardColesNautical,London.
Marchmann,J.F.andWerme,T.D.(1994)ClarkYPerformanceatLowReynoldsNumbers.Proceedingsofthe
VA.AIAA22ndAerospaceScienceMeeting,Reno.
MaritimeandPortAuthorityofSingapore(2009)PortStatistics.Availablefrom:
http://www.mpa.gov.sg/sites/global_navigation/publications/port_statistics/port_statistics.page,
accessed09/04/2010.
Mash,C.(2009).Personalcommunicationviaemail,BorchardLinesLtd.13/11/2009.
MCA(2007)LY2TheLargeCommercialYachtCode,MaritimeandCoastguardAgency.Availablefrom:
http://www.ukshipregister.co.uk/msn_1792_edition_2.pdf,accessed14/04/2010.

162

References
Michell,J.H.(1898)TheWaveResistanceofaShip.PhilosophicalMagazine,45(5),pp10623.
Molland,A.F.(2008)ShipDesignandEconomicsLectureNotes,UniversityofSouthampton
Molland,A.F.(2009)MarinePropulsionEngineeringNotes,UniversityofSouthampton.
Molland,A.F.,Turnock,S.R.andHudson,D.A.(unpublished)ShipResistanceandPropulsion.Cambridge
UniversityPress,Cambridge.
Molland,A.F.,Wellicome,J.F.andCouser,P.R.(1994)ShipScienceReport71:ResistanceExperimentsona
SystematicSeriesofHighSpeedDisplacementCatamaranForms:variationofLengthDisplacement
RatioandBreadthDraughtRatio,UniversityofSouthampton.
Munk,M.M.(1923)GeneralBiplaneTheory,NACA.
U.S.DepartmentofCommerce(2009)GlobalOceanObservations.Availablefrom:
http://www.aoml.noaa.gov/phod/goos.php,accessed14/03/2010.
Noteboom,T.E.andVernimmen,B.(2009)TheEffectofHighFuelCostsonLinerServiceConfigurationsin
ContainerShipping.TheJournalofTransportGeography,17(5),pp32536.
NYK(2010)NYKSuperEcoShip2030.http://www.nyk.com/english/csr/envi/ecoship.htm.Accessed:
19/04/2010
OceanShippingConsultantsLtd(2006)IstheGrowthoftheEastAsianContainerPortMarketsSustainable?
OceanShippingConsultantsLtd(2010)DeepSeaContainershipTradingCosts,obtainedfromDavidTozerof
Lloyd'sRegisterwiththepermissionoftheauthors.
Palmer,A.(2010).Personalcommunication,viaemail.19/03/2010.
Percival,S.,Hendrix,D.andNoblesse,F.(2001)HydrodynamicOptimisationofShipHullForm.AppliedOcean
Research,23(6),pp33755.
Podsada,J.(2001)SeaCargo:BeachBountyorJunk?.
http://news.nationalgeographic.com/news/2001/06/0619_seacargo.html.Accessed:14/03/2010
Praefke,E.,Richards,J.andEngelskirchen,J.(2001)CounterRotatingPropellersWithoutComplexShaftingfor
aFastMonohullFerry.Proceedingsofthe6thInternationalConferenceonFastSeaTransport(FAST),
Southampton.pp110.
PSASingaporeTerminals(2009)Transhipment.http://www.singaporepsa.com/corporate/transhipment.html.
Accessed:
RollsRoyce(2008)KamewaPropulsionBrochure.Availablefrom:http://www.rolls
royce.com/marine/downloads/propulsion/prop_broc.pdf,accessed10/12/2008.
Salvesen,N.(1978)AddedResistanceofShipsinWaves.JournalofHydronautics,12(1),pp2434.
Satchwell,C.J.(1986)MarineAerofoilMotionDampingandRelatedPropulsiveBenefits,Universityof
Southampton.
Satchwell,C.J.(1989)WindshipTechnologyanditsApplicationtoMotorShips.TransactionsoftheRoyal
InstitutionofNavalArchitects,131pp12133.
Schenzle,P.(1985)EstimationofWindAssistancePotential.JournalofWindEngineeringandIndustrial
Aerodynamics,20(13),pp97109.
Schneekluth,H.(1988)HydromechanikzumSchiffsentwurf.Koehler,
Schneekluth,H.andBertram,V.(1987)ShipDesignforEfficiencyandEconomy.ButterworthHeinemann,
Oxford.
Sears,R.I.andLiddel,R.B.(1942)TunnelInvestigationofControlSurfaceCharacteristics.VIA30PercentChord
PlainFlapontheNACA0015Airfoil,NACAReport454.
Sinmons,L.F.G.andSalter,C.(1934)ExperimentalInvestigationandAnalysisoftheVelocityVariationsin
TurbulentFlow.ProceedingsoftheRoyalSocietyofLondon,145(854),pp21233.
Sipila,H.andBrown,A.(1997)ApplicationoftheSlenderMonohulltoHighSpeedContainerVessels.
Proceedingsofthe4thInternationalConferenceonFastSeaTransport(FAST),Sydney.pp24753.
SkySails(2010)SkySails:TurnWindintoProfit.http://www.skysails.info/.Accessed:19/04/2010
Smulders,F.(1985)ExpositionofCalculationMethodstoAnalyseWindPropulsiononCargoShips.Journalof
WindEngineeringandIndustrialAerodynamics,19(13),pp187203.
StromTejsen,J.,Yeh,H.Y.H.andMoran,D.D.(1973)AddedResistanceinWaves.TransactionsoftheSocietyof
NavalArchitectsandMarineEngineers,81pp10943.
Svendsen,J.andTiedemann,J.(2009)ContainershipInfo.http://www.containershipinfo.com/.Accessed:
11/11/2009
Svoboda,C.(1999)AluminiumStructuralMemberComponentWeightasaFunctionofWingLoading.Aircraft
Design2,2(4),pp23137.
Tanneberger,K.(2009).Personalcommunication:MeetingatLloyd'sRegister71FenchurchStreet,London.
16/12/2009.

163

Concept Design of a Fast Sail Assisted Feeder Container Ship


Townsin,R.L.(2003)TheShipHullFoulingPenalty,InternationalMarineCoatings.
Tozer,D.(2009).Personalcommunication:MeetingatLloyd'sRegister71FenchurchStreet,London.
16/12/2009.
Tu,J.,Yeoh,G.H.andLiu,C.(2008)ComputationalFluidDynamics:APracticalApproach.Firstedition.
ButterworthHeinemann,Oxford.
Tuck,E.andLazaukas,L.(2008)DragonaShipandMichell'sIntegral.ProceedingsoftheXXIIInternational
CongressonTheoreticalandAppliedMechanics,Adelaide.pp12.
Tuck,E.O.andLazaukas,L.(1996)UnconstrainedShipsofMinimumTotalDrag,UniversityofAdelaide.
Tuck,E.O.,Lazaukas,L.andScullen,D.C.(1999)SeaWavePatternEvaluation,Part1Report:PrimaryCodeand
TestResults(SurfaceVessels),UniversityofAdelaide.
vanDuivendijk,L.(2009)ContainerShips.http://www.duivendijk.net.Accessed:10/11/2009
vanGunsteren,L.A.(1971)ApplicationofMomentumTheoryinCounterrotatingPropellerDesign.International
ShipbuildingProgress,18(206),pp35972.
Veldhuis,I.(2007)ApplicationofHydrogenMarineSystemsinHighspeedSeaContainerTransport.Thesis,
UniversityofSouthampton.
VietnamSeaportsAssociation(2010)ShippingSchedules.Availablefrom:
http://www.vpa.org.vn/english/information/info_schedure.htm,accessed09/04/2010.
Walker,J.G.(1985)AHighPerformanceAutomaticWingsailAuxiliaryPropulsionSystemForCommercialShips.
JournalofWindEngineeringandIndustrialAerodynamics,20(13),pp8396.
WrtsilCorporation(2008)WrtsilControllablePitchPropellers.Availablefrom:
http://www.wartsila.com/Wartsila/global/docs/en/ship_power/media_publications/brochures/produ
ct/propulsors/controllablepitchpropellerswartsilabrochure.pdf,accessed24/03/2010.
WrtsilCorporation(2009a)FlexibilityinaSmallPackage.Availablefrom:http://www.wartsila.com/
Wartsila/global/docs/en/about_us/in_detail/1_2009/flexibilityinasmallpackagearticle.pdf,
accessed23/02/2010.
WrtsilCorporation(2009b)ShippingintheGasAge.Availablefrom:
http://www.wartsila.com/Wartsila/global/docs/en/ship_power/media_publications/brochures/shippi
nginthegasagebrochure.pdf,accessed19/04/2010.
WrtsilCorporation(2010)FuelCellsforMarineApplications.Availablefrom:
http://www.wartsila.com/Wartsila/global/docs/en/products_services/fuel_cells/wartsila_fuel_cells_f
or_marine_applications_leaflet.pdf,accessed23/03/2010.
WrtsilShipPowerR&D(2009)BoostingEnergyEfficiency.Availablefrom:http://wartsila.eu/Wartsila/
global/docs/en/ship_power/energyefficiency/boostingenergyefficiencypresentation.pdf,accessed
01/03/2010.
WrtsilShipPowerTechnology(2009a)Wrtsil50DFMainData.Availablefrom:http://www.wartsila.com/
Wartsila/global/docs/en/ship_power/products/2009/maindatawartsila50df.pdf,accessed
26/02/2010.
WrtsilShipPowerTechnology(2009b)Wrtsil50DFProductGuide.Availablefrom:
http://www.wartsila.com/wartsila50DF,accessed24/02/2010.
Watson,D.G.M.andGilfillan,A.W.(1977)SomeShipDesignMethods.TransactionsoftheRoyalInstitutionof
NavalArchitects,119p.45.
Willsher,J.D.andSolomon,T.(2010)OperationalandEnvironmentalBenefitsofFoulReleaseCoatings.
ProceedingsoftheShipDesignandOperationforEnvironmentalSustainability,London.pp5154.
Wilson,P.A.(1985)AReviewoftheMethodsofCalculationofAddedResistanceofShipsinaSeaway.Journal
ofWindEngineeringandIndustrialAerodynamics,20(13),pp18799.
Young,W.C.andBudynas,R.G.(2002)Roark'sFormulasforStressandStrain.Seventhedition.McGrawHill,
NewYork.

164

Appendices

Appendices
AppendixABasisShips...........................................................................................................................167
AppendixBHullFormDevelopment.......................................................................................................168
B.1Bulbousbowoptimisation.................................................................................................................168
B.2Numericalresistanceprediction........................................................................................................170
B.3HullAlinesplan..................................................................................................................................173
B.4HullBlinesplan..................................................................................................................................174
AppendixCModelDesign,ManufactureandTestingPreparation...........................................................175
C.1Design.................................................................................................................................................175
C.2Towingtankmodelplans...................................................................................................................176
C.3Manufacture......................................................................................................................................177
C.4Testingpreparation............................................................................................................................178
AppendixDTowingTankResultsProcessing...........................................................................................179
D.1Turbulencestimulationcorrection....................................................................................................179
D.2Uprightresistance..............................................................................................................................180
D.3Sailingcondition.................................................................................................................................181
D.4Uncertaintyanalysis...........................................................................................................................182
AppendixESailDesign............................................................................................................................186
E.1Conceptreview..................................................................................................................................186
E.2Design.................................................................................................................................................187
E.3Theoreticalperformance....................................................................................................................188
E.4Windstatistics....................................................................................................................................190
AppendixFWindTunnel.........................................................................................................................191
F.1Designandmanufacture....................................................................................................................191
F.2Windtunnelmodelplans...................................................................................................................193
F.3Windtunnelcalibrationresults..........................................................................................................197
F.4Sailcontainerinteractionfigures.......................................................................................................198
F.5Windtunnelcorrections.....................................................................................................................199
F.6LowReynoldsnumbertesting............................................................................................................202
F.7Flowvisualisation...............................................................................................................................203
F.8ExtrapolationtofullscaleReynoldsnumber.....................................................................................204
F.9Aerodynamiccharacteristicsfromwindtunnelresultsaftercorrection...........................................205
F.10Uncertaintyanalysisrandomerrorcomponent............................................................................208
F.11ComputationalfluiddynamicsstudyonMultiwingsailsystem......................................................209
AppendixGPerformancePredictions......................................................................................................210
G.1Sailingperformance...........................................................................................................................210
G.2Propulsion..........................................................................................................................................211
AppendixHDesignFeasibility.................................................................................................................213
H.1Performanceindices..........................................................................................................................214
H.2Freightrateestimate.........................................................................................................................215
AppendixISeakeeping...........................................................................................................................216
I.1PredictedRAOcurves..........................................................................................................................216
I.2Subjectivemotion(SM)polarplots.....................................................................................................217
AppendixJPredictedRollDamping.........................................................................................................219
J.1Nakedhullrolldamping......................................................................................................................219
J.2Liftingsurfacemethod........................................................................................................................220
J.3Liftinglinemethod..............................................................................................................................221

165

Concept Design of a Fast Sail Assisted Feeder Container Ship


J.4Controlsystem....................................................................................................................................221
AppendixKStructuralDesign..................................................................................................................222
K.1Midshipscantlingcalculations............................................................................................................222
K.2Typicalsectionstructuraldrawings....................................................................................................225
K.3Globalstrength...................................................................................................................................227
K.4FEmidshipsectionmodelresults.......................................................................................................230
AppendixLStability................................................................................................................................232
L.1Stabilitymodeldefinition....................................................................................................................232
L.2Intactstability.....................................................................................................................................232
L.3Damagestability.................................................................................................................................236
AppendixMGeneralArrangement..........................................................................................................238
AppendixNBudgetSummary..................................................................................................................239

166

Appendices

Appendix A Basis Ships


36

210
190

32

28
150

B/m

LOA /m

170

130
110

24
y=0.005608x+17.898958
R=0.438420
20

y = 0.043911x + 105.535807
R = 0.575463

90

16
0

500

1000

1500

2000

2500

3000

NC/TEU

500

1000

1500

2000

NC/TEU

2500

3000

FigureA.1LOAasafunctionofcargocapacity(left);andBasafunctionofcargocapacity(right)
12

10
y=0.093x+4.738
R=0.201

9
10

L/B

T/m

8
7
8
6

y=0.002236x+6.161147
R=0.470685

6
0

500

1000

1500

2000

NC/TEU

2500

3000

10

15

20

25

VS /knots

30

35

FigureA.2Draughtasafunctionofcargocapacity(left);andLOA/Basafunctionofshipspeed(VS)(right)
35000

Deadweight/tonnes

30000
25000
20000
15000
10000
y=14.547x 529.94
R2 =0.7871

5000
0
0

500

1000

1500

NC/TEU

2000

2500

FigureA.3Deadweightasafunctionofcargocapacity

167

12

13

14

15

168

0.1370 3.105 3.074 2.843 3.043 2.984 2.806 3.041 2.965 2.746 3.103 3.048 2.767 2.728 2.652 2.684

0.1566 3.055 3.024 2.805 2.996 2.939 2.770 2.994 2.921 2.713 3.050 2.998 2.732 2.696 2.624 2.654

0.1761 3.018 2.988 2.785 2.964 2.911 2.753 2.961 2.893 2.700 3.013 2.964 2.718 2.686 2.620 2.647

0.1957 3.006 2.979 2.793 2.957 2.909 2.765 2.954 2.891 2.715 3.003 2.959 2.736 2.710 2.651 2.675

0.2153 3.028 3.006 2.837 2.984 2.942 2.814 2.981 2.926 2.767 3.030 2.991 2.793 2.776 2.726 2.747

0.2348 3.082 3.066 2.916 3.044 3.009 2.898 3.040 2.994 2.855 3.090 3.059 2.888 2.882 2.842 2.860

0.2544 3.183 3.174 3.046 3.150 3.125 3.033 3.147 3.111 2.995 3.201 3.179 3.040 3.048 3.020 3.035

0.2740 3.358 3.358 3.253 3.331 3.317 3.247 3.327 3.304 3.211 3.388 3.377 3.273 3.301 3.287 3.298

0.2935 3.516 3.524 3.437 3.496 3.490 3.436 3.490 3.475 3.402 3.555 3.552 3.477 3.525 3.522 3.527

0.3131 3.588 3.601 3.529 3.572 3.572 3.532 3.565 3.558 3.501 3.633 3.636 3.584 3.643 3.648 3.650

0.3327 3.654 3.675 3.621 3.642 3.653 3.629 3.637 3.641 3.602 3.710 3.722 3.695 3.762 3.779 3.782

0.3523 3.821 3.855 3.826 3.815 3.840 3.841 3.813 3.832 3.819 3.898 3.922 3.933 4.014 4.049 4.052

0.3718 4.150 4.204 4.206 4.152 4.199 4.233 4.154 4.196 4.217 4.258 4.302 4.365 4.472 4.534 4.536

0.3914 4.661 4.741 4.784 4.674 4.749 4.825 4.679 4.749 4.816 4.811 4.881 5.011 5.159 5.255 5.254

10.5

12.0

13.5

15.0

16.5

18.0

19.5

21.0

22.5

24.0

25.5

27.0

28.5

30.0

0.1174 3.157 3.126 2.887 3.091 3.032 2.849 3.090 3.013 2.787 3.156 3.102 2.811 2.771 2.693 2.727

9.0

0.0978 3.195 3.167 2.927 3.128 3.071 2.888 3.127 3.052 2.827 3.198 3.145 2.853 2.814 2.737 2.771

11

7.5

10

0.0783 3.207 3.183 2.954 3.143 3.090 2.917 3.143 3.073 2.859 3.214 3.166 2.887 2.851 2.777 2.811

6.0

1000

0.0587 3.195 3.176 2.975 3.139 3.094 2.943 3.139 3.080 2.892 3.205 3.164 2.919 2.889 2.825 2.855

4.5

0.0391 3.160 3.147 2.995 3.117 3.084 2.970 3.118 3.074 2.932 3.170 3.140 2.954 2.933 2.884 2.907

3.0

0.0196 3.172 3.166 3.083 3.150 3.132 3.070 3.151 3.127 3.050 3.180 3.164 3.063 3.052 3.026 3.038

0.0049 3.560 3.560 3.538 3.556 3.551 3.535 3.557 3.551 3.530 3.566 3.561 3.535 3.533 3.526 3.529

1.5

TableB.1HullA;Totalresistancecoefficientforallbulbvariants

0.4

/knots

Bulbvariant

B.1 Bulbous bow optimisation

Appendix B Hull Form Development

Concept Design of a Fast Sail Assisted Feeder Container Ship

Bulbvariant

12

13

14

15

169

0.1557 3.150 3.232 2.944 3.257 3.164 2.934 3.286 3.187 2.906 3.355 3.277 2.924 2.854 2.747 2.787

0.1751 3.104 3.178 2.911 3.202 3.115 2.902 3.229 3.137 2.877 3.290 3.218 2.892 2.829 2.731 2.767

0.1946 3.082 3.149 2.906 3.171 3.091 2.897 3.198 3.113 2.877 3.253 3.186 2.892 2.838 2.749 2.782

0.2140 3.093 3.156 2.937 3.174 3.102 2.928 3.201 3.127 2.914 3.252 3.192 2.931 2.889 2.812 2.840

0.2335 3.137 3.197 3.003 3.210 3.146 2.993 3.239 3.175 2.987 3.286 3.234 3.008 2.980 2.915 2.939

0.2529 3.227 3.288 3.120 3.292 3.238 3.108 3.324 3.274 3.114 3.371 3.329 3.141 3.130 3.080 3.100

0.2724 3.386 3.449 3.309 3.443 3.400 3.294 3.481 3.446 3.314 3.528 3.498 3.350 3.364 3.330 3.345

0.2919 3.531 3.594 3.476 3.579 3.544 3.458 3.622 3.598 3.490 3.668 3.646 3.533 3.570 3.549 3.557

0.3113 3.596 3.659 3.561 3.638 3.611 3.541 3.683 3.669 3.581 3.728 3.713 3.628 3.677 3.667 3.672

0.3308 3.659 3.726 3.649 3.696 3.679 3.629 3.745 3.744 3.677 3.792 3.787 3.732 3.791 3.794 3.798

0.3502 3.820 3.896 3.846 3.853 3.849 3.825 3.911 3.927 3.890 3.963 3.971 3.957 4.033 4.057 4.060

0.3697 4.137 4.228 4.215 4.164 4.179 4.189 4.237 4.279 4.280 4.299 4.326 4.369 4.476 4.527 4.527

0.3891 4.629 4.740 4.771 4.649 4.686 4.740 4.742 4.818 4.866 4.819 4.869 4.985 5.137 5.225 5.221

12.0

13.5

15.0

16.5

18.0

19.5

21.0

22.5

24.0

25.5

27.0

28.5

30.0

0.1362 3.210 3.300 2.993 3.324 3.227 2.983 3.355 3.252 2.952 3.433 3.353 2.974 2.900 2.786 2.830

10.5

0.1167 3.268 3.367 3.046 3.388 3.290 3.036 3.422 3.316 3.004 3.509 3.427 3.031 2.955 2.836 2.883

9.0

0.0973 3.310 3.415 3.090 3.433 3.335 3.080 3.468 3.364 3.048 3.562 3.483 3.080 3.005 2.885 2.934

11

7.5

10

0.0778 3.320 3.426 3.113 3.439 3.348 3.103 3.475 3.378 3.072 3.572 3.499 3.109 3.039 2.923 2.972

6.0

1000

0.0584 3.286 3.386 3.107 3.393 3.315 3.098 3.427 3.343 3.072 3.519 3.457 3.109 3.049 2.946 2.991

4.5

0.0389 3.215 3.294 3.080 3.298 3.239 3.074 3.324 3.262 3.055 3.400 3.354 3.087 3.042 2.964 2.999

3.0

0.0195 3.181 3.225 3.107 3.226 3.195 3.104 3.242 3.208 3.095 3.286 3.262 3.114 3.092 3.049 3.068

0.0049 3.526 3.536 3.505 3.538 3.529 3.505 3.543 3.534 3.504 3.556 3.549 3.511 3.507 3.496 3.500

1.5

0.4

/knots

TableB.2HullB;Totalresistancecoefficientforallbulbvariants

Appendices

1.5
3.0
4.5
6.0
7.5
9.0
10.5
12.0
13.5
15.0
16.5
18.0
19.5
21.0
22.5
24.0
25.5
27.0
28.5
30.0

/knots

0.0196
0.0391
0.0587
0.0783
0.0978
0.1174
0.1370
0.1566
0.1761
0.1957
0.2153
0.2348
0.2544
0.2740
0.2935
0.3131
0.3327
0.3523
0.3718
0.3914

3.050
2.932
2.892
2.859
2.827
2.787
2.746
2.713
2.700
2.715
2.767
2.855
2.995
3.211
3.402
3.501
3.602
3.819
4.217
4.816

0.169
0.281
0.361
0.406
0.433
0.439
0.435
0.434
0.448
0.487
0.561
0.667
0.824
1.056
1.262
1.374
1.487
1.716
2.125
2.734

2.074
1.881
1.781
1.714
1.665
1.626
1.595
1.568
1.545
1.525
1.507
1.491
1.477
1.463
1.451
1.440
1.429
1.420
1.411
1.402

HullA
2.464
2.235
2.115
2.036
1.977
1.932
1.894
1.862
1.835
1.811
1.790
1.771
1.754
1.738
1.724
1.710
1.698
1.686
1.675
1.665

170

3.4
26.3
87.6
205.3
396.5
675.6
1057.1
1559.0
2208.6
3046.7
4133.6
5536.0
7382.9
9887.5
12884.9
16091.2
19858.5
24994.9
32460.8
43234.2

/kW
0.0195
0.0389
0.0584
0.0778
0.0973
0.1167
0.1362
0.1557
0.1751
0.1946
0.2140
0.2335
0.2529
0.2724
0.2919
0.3113
0.3308
0.3502
0.3697
0.3891

3.104
3.074
3.098
3.103
3.080
3.036
2.983
2.934
2.902
2.897
2.928
2.993
3.108
3.294
3.458
3.541
3.629
3.825
4.189
4.740

0.242
0.439
0.582
0.666
0.701
0.702
0.687
0.669
0.664
0.683
0.734
0.818
0.950
1.152
1.331
1.427
1.527
1.734
2.110
2.671

2.071
1.879
1.778
1.712
1.663
1.624
1.593
1.566
1.543
1.523
1.505
1.489
1.475
1.462
1.449
1.438
1.428
1.418
1.409
1.400

HullB

2.448
2.220
2.102
2.023
1.965
1.920
1.882
1.851
1.824
1.800
1.779
1.760
1.743
1.727
1.713
1.700
1.687
1.676
1.665
1.655

TableB.3Breakdownofcoefficientsofresistanceandeffectivepowerforoptimisedhulls

B.2 Numerical resistance prediction

3.1
23.0
76.4
176.1
372.9
763.8
1430.0
2651.5
3335.3
5371.4
8968.4
10344.1
14912.0
15168.2
19221.4
23627.2
27803.8
37000.8
54638.0
80225.6

/kW

Concept Design of a Fast Sail Assisted Feeder Container Ship

Appendices
30000

25000
HULLA Holtrop
HULLB Holtrop
20000

HULLA ThinShip

PE /kW

HULLB ThinShip
15000

10000

5000

0
0

10

VS/knots

15

20

25

FigureB.1ComparisonofresultsusingHoltropregressionandThinShipTheory

171

30

Concept Design of a Fast Sail Assisted Feeder Container Ship

172

Appendices

Appendix C Model Design, Manufacture and Testing Preparation


C.1 Design
DepthFroudenumbercalculation
DeptheffectswillbecomeincreasinglyimportantasthedepthFroudenumber,definedas
,
approaches and exceeds a value of one. Assuming the critical value of

(C.1)
, the required model speed is

,whichisequalto4.26ms .BasedonaFroudenumberscalingapproach,

andthecorrespondingscalefactoris

(C.2)

9.12.ThisequatestoamodelLWL of17.38metresforHullA.

3Dmodelviews

FigureC.1ViewsofHullA:bulbousbowshapeandinternaldetail(left);andsternshape(right)

FigureC.2ViewsofHullB:bulbousbowshape(left);andsternshape(right)

175

JIGSAW
CONNECTOR
4 X JIGSAW PIECES CNCMACHINED FROM 200 kg/m
FOAM TO FIT CUT-OUTS IN
MODEL PARTS.

30

A'

A'

142

A''

415

1138

200

1138

598

30

30

12

149

361

72

SEE DETAIL

60

174

174

30

HULL A MODEL

60

100

551

251

60

60

60

251

114

SCALE 1:5

SECTIONS ON
A-A' & A'-A''

48

72

100

DETAIL
SCALE 1:2
174

SEE DETAIL

175

60

297

149

B'

B'

1138

1138

30

30

551

FORE AND AFT SECTIONS OF HULL


MODELS TO BE CNC MACHINED WITH
'JIGSAW' DETAIL. SECTIONS TO BE
ALIGNMENED WITH JIGSAW PIECES AND
JOINED WITH SILICON ADHESIVE.

60

60

60

120

662

251

B''

415

142

251

30

30

HULL B MODEL

TOWING TANK MODEL PLANS


Designed by
SM

SECTIONS ON
B-B' & B'-B''

Materials
Units
mm

120 kg/m
FOAM

200 kg/m
FOAM

Checked by
TL

APPENDIX C.2

Date
19/04/10

Scale
1:10

Edition
A

Sheet

University of Southampton
University Road
Highfield
Southampton
SO17 1BJ

1 of 1

Appendices
C.3 Manufacture

a)

b)

c)

d)

e)

f)

FigureC.3(a)HullAsternsectionduringmilling;(b)HullAatprefinishingstage;(c)Bulkheadsandjoinof
forwardandaftmodelsectionsduringgluing;(d)ComparisonofsternformsofHullsAandB;(e)Turbulence
stimulatorlocationsandfairingissuesatintersectionofbulbandhull;(f)Heelfittingandtrimmingmomentrail
attachedtoHullA

177

Concept Design of a Fast Sail Assisted Feeder Container Ship

Totalmomenttoapply
aft./Nm

0.200
0.150
0.100
0.050
0.000
0.10

0.30

0.50

0.70

0.90

1.10

VM/ms1

1.30

1.50

FigureC.4Totaltrimmomentcorrectionappliedforrangeofmodelspeedstested

C.4 Testing preparation


TableC.1Conservativeestimateofpossibletestsduringthreedays(totalof64runs)
Day

Time

1
1
1
1
2
2
2
2
2
3
3
3
3

08:3012:00
12:0013:40
13:4015:40
15:4017:00
08:3009:00
09:0011:00
11:0013:00
13:0014:40
14:4016:40
08:3009:00
09:0013:00
13:0014:00
14:0017:00

Rundescription

No.ofruns No.ofspeeds Waves?

Calibrationandballasting
HullA:Calmwater
HullA:Sailingcondition,calmwater
HullA:Addedresistanceinwaves
Calibrationcheck
HullA:Addedresistanceinwaves
Changeover
HullB:Calmwater
HullB:Sailingcondition,calmwater
Calibrationcheck
HullB:Addedresistanceinwaves
HullsA/B:Lowspeed(Prohaska)tests
Sailingconditionatadditionalspeed?

10
7
4

10
7

12
6

10
1
2

6
1

2
3

no
no
yes

yes

no
no

yes
no
no

TableC.2Estimateofmaximumrequiredtestsduringthreetestingdays(totalof80runs)
Day

Time

08:3012:00 Calibrationandballasting

12:0013:10 HullA:Calmwater

10

10

no

13:1016:40 HullA:Sailingcondition,calmwater

15

no

16:4017:00 HullA:Lowspeed(Prohaska)tests

no

08:3009:00 Calibrationcheck

09:0012:00 HullA:Addedresistanceinwaves

12

yes

12:0014:00 Changeover

14:0015:10 HullB:Calmwater

10

10

no

15:1017:00 HullB:Sailingcondition,calmwater

11

no

08:3009:00 Calibrationcheck

09:0009:40 HullB:Sailingcondition,calmwater

no

09:4012:40 HullB:Addedresistanceinwaves

12

yes

12:4013:00 HullB:Lowspeed(Prohaska)tests

no

13:0017:00 Sailingconditionatadditionalspeed?

Rundescription

No.ofruns No.ofspeeds Waves?

178

Appendices

Appendix D Towing Tank Results Processing


D.1 Turbulence stimulation correction
Themeasureddragwerecorrectedforthedragduetoturbulencestudsbymeansofthemethodoutlinedby
Molland (1994), wherein a correction based on model boundary layer thickness in laminar and turbulent
regionsisappliedtothemeasureddragtoallowaccuratescalingwiththeITTCcorrelationline.Calculatingthe
and
respectively, the correction may be obtained as
drag contribution for both flow regions as
isthefrictionaldragactingonthestimulationstudsthemselvesand
indicatedinEquation(D.1),where
thefrictionaldragaccordingtotheITTCcorrelationline.

(D.1)

Although in the yawed sailing condition tests the boundary layer thickness at the studs will clearly vary
betweentheleewardandwindwardsides,itwasassumedforthepurposesofthecorrectionthatthiseffect
was to an extent selfcancelling and that the upright resistance correction could be applied with relative
confidence.ThecorrectionsobtainedfromthemethodareshowninTableD.1.
TableD.1Exampledragcorrectioncalculationsforbothhulls

/ms1
0.940
1.515
0.936
1.510

HullA
HullB

/N
2.414
6.758
2.254
5.981

/N
1.608
3.796
1.522
3.599

Correction
/N
/%
0.048
2.00
0.131
1.94
0.050
2.23
0.136
2.28

/N
0.016
0.057
0.015
0.055

/N
0.097
0.199
0.098
0.201

/N
1.437
3.386
1.347
3.180

Havingobtainedapositivecorrection,itwasassumedthatthetransitiontoturbulentflowwasinfactoccurring
forwardsofthepositionconsidered.Retrospectively,itisthuspossibletoexaminetheeffectofstudlocation
ondragcorrectiontobeapplied,(seeFigureD.1).

Drag correction / N

0.2

HullA

0.15

HullB

0.1
0.05
0
0.05
0.1
2

10

Studlocation/%LWL aftFP

FigureD.1Variationinempiricaldragcorrectionwithtripstudlocation
The slight difference between the two hulls is to be expected, due to the increased submerged length and
differing surface area of Hull B. From this result it is proposed that abetter representation of full scale flow
aroundthehullmighthavebeenobtainedfromplacingthetripstudsfurtherforwards,specificallywithinthe
lengthofthebulbousbow.

179

Concept Design of a Fast Sail Assisted Feeder Container Ship


D.2 Upright resistance
TableD.2HullAeffectivepowerandresistancecomponentbreakdown
/

5.07
0.066
6.22
0.081
7.15
0.093
10.14
0.132
11.32
0.148
13.34
0.174
13.45
0.175
13.45
0.175
13.90
0.181
15.82
0.206
15.91
0.208
17.83
0.233
17.91
0.234
19.39
0.253
21.22
0.277
23.23
0.303
25.50
0.333

1000
4.370
3.397
2.841
2.352
2.685
2.788
2.840
2.796
2.993
2.903
3.038
3.144
3.083
3.435
3.317
3.452
3.661

1000
1.934
1.027
0.516
0.141
0.512
0.668
0.723
0.679
0.887
0.840
0.978
1.121
1.062
1.440
1.352
1.518
1.758

1000
1.717
1.671
1.642
1.570
1.549
1.518
1.516
1.516
1.510
1.486
1.485
1.465
1.464
1.450
1.434
1.419
1.403

1000
/

1.966
192.73
1.914
277.05
1.880
350.49
1.798
828.13
1.774
1318.86
1.738
2235.42
1.736
2335.58
1.736
2299.61
1.729
2719.87
1.702
3886.61
1.701
4138.93
1.677
6024.36
1.676
5986.72
1.660
8463.05
1.642
10710.04
1.625
14625.44
1.607
20524.07

TableD.3HullBeffectivepowerandresistancecomponentbreakdown
/

6.26
7.26
7.26
10.20
11.44
13.53
15.85
16.01
17.95
19.50
21.37
23.33
25.57

1000
3.297
2.545
2.330
2.333
2.231
2.411
2.540
2.438
2.765
2.954
2.844
2.848
2.921

0.081
0.094
0.094
0.132
0.148
0.174
0.204
0.206
0.232
0.251
0.276
0.301
0.330

1000
0.929
0.225
0.010
0.125
0.060
0.296
0.478
0.379
0.744
0.961
0.881
0.915
1.019

180

1000
1.669
1.637
1.637
1.568
1.546
1.514
1.485
1.483
1.462
1.448
1.432
1.417
1.402

1000
1.911
1.875
1.875
1.796
1.770
1.734
1.700
1.698
1.675
1.658
1.640
1.623
1.605

256.76
309.34
283.56
786.89
1060.49
1895.32
3212.65
3174.35
5082.24
6957.38
8811.47
11481.95
15509.72

25.5

15.9

181

2.5

2.5

/
0
2.5
5
0
2.5
5
0
2.5
5
0
2.5
5
0
2.5
5
0
2.5
5

/
4.926E+05
0.000E+00 5.409E+05
6.245E+05
4.918E+05
8.368E+02 5.235E+05
6.146E+05
4.931E+05
4.535E+02 5.219E+05
5.993E+05
1.531E+06
0.000E+00 1.746E+06
2.160E+06
1.547E+06
1.657E+04 1.791E+06
2.325E+06
1.585E+06
5.412E+04 1.804E+06
2.419E+06

HullA
/

HullB
/

/
/
/
0.000E+00
0.000E+00
4.505E+05
2.988E+04 0.000E+00
9.009E+05
1.219E+05
7.618E+04
0.000E+00
3.707E+05 3.750E+05 1.437E+04 3.929E+03
9.627E+05
9.998E+04
9.012E+04
0.000E+00
3.504E+05
2.031E+04 3.120E+03
9.171E+05
9.384E+04
0.000E+00
0.000E+00
1.082E+06
1.312E+05 0.000E+00
2.163E+06
5.278E+05
1.483E+05
0.000E+00
1.149E+06 1.153E+06 1.208E+05 2.583E+03
2.731E+06
4.970E+05
2.461E+05
0.000E+00
1.333E+06
9.685E+04 2.183E+04
3.032E+06
4.341E+05

TableD.4Resistancebreakdowninsailingcondition

/
0.000E+00
4.822E+04
1.319E+05
0.000E+00
4.926E+05 3.169E+04
1.228E+05
0.000E+00
2.885E+04
1.062E+05
0.000E+00
2.157E+05
6.294E+05
0.000E+00
1.531E+06 2.435E+05
7.772E+05
0.000E+00
2.186E+05
8.341E+05

D.3 Sailing condition

/
3.750E+05
4.049E+05
4.969E+05
3.789E+05
3.933E+05
4.789E+05
3.781E+05
3.984E+05
4.719E+05
1.153E+06
1.284E+06
1.681E+06
1.150E+06
1.271E+06
1.647E+06
1.175E+06
1.272E+06
1.609E+06

/
0.000E+00
3.345E+05
6.691E+05
6.005E+04
3.160E+05
6.831E+05
5.686E+04
2.943E+05
6.530E+05
0.000E+00
1.333E+06
2.665E+06
1.620E+05
1.295E+06
2.888E+06
1.790E+05
1.407E+06
2.868E+06

Appendices

Concept Design of a Fast Sail Assisted Feeder Container Ship


D.4 Uncertainty analysis
Nomenclature
ITTCbasednomenclaturepertinenttoappendix,otherparametersasperprincipalnomenclature.

Resistancecoefficientbiaslimit
Resistancemassbias

kg

Resistancecalibrationbias

kg

kg
kg

Resistancecurvefitbias
Resistanceloadcell
misalignmentbias
Resistancetowingforce
inclinationbias
Wettedsurfaceareabias

Assumederrorinhullformbias
Errorindisplacementbias
Watertemperaturebias

m2
m2
o
C

Biasonwaterdensity

kgm3

cm

Biasonwaveprofile
measurement
Biasonscaleplacement
Biasonmarkerplacement
Biasonmarkerreapplication

Biasonwaveelevationreading

cm

kg
m2

cm
cm
cm

Biason position
Precisionlimitonresistanceco
efficient
Precisionlimitonwaveprofile
measurement
Watertemperature
Uncertaintyofresistanceco
efficient
Sensitivitycoefficientforresistance
mass
Sensitivitycoefficientforwetted
surfacearea
Sensitivitycoefficientforspeed
Sensitivitycoefficientfor position
Sensitivitycoefficientforwater
density
Sensitivitycoefficientforwater
temperature
Sensitivitycoefficientonwave
amplitude
Measuredwaveamplitude
Significant waveamplitude
Standarddeviationofresistanceco
efficientmeasurement
Standarddeviationofwaveprofile
measurement

cm

m/Ns2
m2
1/ms1
cm1
m3kg1
o 1

cm1
cm
cm

Resistancetestsuncertaintyestimation
Thetotaluncertaintyisacombinationoftheprecisionlimit(

)andthebiaslimit(

)i.e.

(D.2)

where,
and
Thebiaslimitisthesumoferrorsfromspeedvariation(
andwettedsurfaceareavariation(

),resistancemass(

)andwaterproperties(

Theresistancemassbiaslimitincludeseffectofcalibration(
)andtowforceinclination(
)i.e.
(

(D.4)

),curvefitbias(

),loadcellmisalignment

.
The s are sensitivity coefficients obtained from the differential of the total resistance coefficient,
respecttorequiredparameteri.e.

)whichincludeserrorinhullformanderrorindisplacementi.e.

(D.3)

182

(D.5)
, with

Appendices
2
0.5
1
0.5
and

0.5
|0.0638

15C

0.5

0.017

0.0001897

|.

(D.6)

Withintheprecisionlimitequation, isstandarddeviationon , iscoveragefactorwhichcontrolsthelevel


ofconfidenceintheresultand isnumberofruns.Inthisanalysisthecoveragefactorwaschosentoyielda
90%confidencelevelintheresults(
1.645).
Uncertaintyresultsresistancetests
Thefollowingfigurescontainexperimentalresultsforresistance,sideforceandyawmomentwithappropriate
errorbars.

12

TotalResistance/N

TotalResistance/N

10
8
6
4
2
0
2
4
0.0

0.5

1.0
Modelspeed/m/s

1.5

12
10
8
6
4
2
0
2
4
6
0.0

2.0

0.5

1.0
1.5
Modelspeed/knots

2.0

FigureD.2ResistanceestimatesofHullAandHullBwitherrorbars

12
10
8
6
4
2
0
2
4
6
8
10
12
14

12
10
8
6
4
2
0
2
4
6
8
10

M/N.m

M /N.m

YawAngle/degrees

2
0
2
YawAngle/degrees

FigureD.3YawmomentvariationwithyawangleofHullAandHullBwitherrorbars

183

Concept Design of a Fast Sail Assisted Feeder Container Ship


TableD.5Summaryofuncertaintyintotalresistanceandsideforcemeasurementsfortestsinsailing
condition
VS/ / /
knots deg deg
0
0
2.5
5
0
15.9 2.5 2.5
5
0
5
2.5
5
0
0
2.5
5
0
25.5 2.5 2.5
5
0
5
2.5
5

HullA
RT /N Error/N SF / N
2.488
4.355
0.020
2.628
4.726
0.750
2.827
3.406
2.155
2.485
3.453
0.090
2.585
4.143
0.788
2.801
3.476
2.191
2.493
2.805
0.129
2.586
3.454
0.746
2.770
3.852
2.089
8.054
3.418 4.700
7.172
3.778 1.553
6.790
4.006
0.486
7.625
3.756
2.551
8.704
3.609
6.069
6.913
4.434
0.790
7.686
3.840
3.161
8.951
3.774
6.912
7.082
3.418
0.600

Error/N
5.129
4.752
4.432
4.922
4.235
4.502
4.504
5.834
4.106
4.260
5.121
3.457
3.546
3.041
3.071
2.255
2.562
2.861

RT /N
2.157
2.227
2.445
2.188
2.208
2.411
2.184
2.232
2.407
5.725
5.953
6.805
5.716
6.037
6.928
5.855
6.067
6.867

HullB
Error/N SF/N Error/N
3.495
0.171 4.682
3.098
0.523
2.848
2.274
1.414
2.850
2.560
0.048 2.863
2.382
0.559
2.746
2.403
1.429
2.929
3.136
0.047 4.158
3.312
0.516
4.415
2.829
1.366
4.416
3.178
0.525 4.498
3.945
2.290
3.446
3.416
5.848
4.152
5.206
0.200 6.151
3.504
2.485
6.586
3.519
6.262
5.083
3.033
0.152 5.701
3.486
2.759
7.110
3.149
6.222
8.793

Waveprobeuncertaintyestimation
Theuncertaintyinwaveprofilemeasurementcanbeexpressedasthesumoftheprecisionandbiaslimits,i.e.

(D.7)

where
and
The bias limit on the wave profile,
(

),markreapplication(

, consists of bias limits from scale placement (

),waveelevationreading(

(D.8)
), marker placement

)andlongitudinalpositioninthetank( )bias(

i.e.
.

(D.9)

(D.10)

Thesensitivitycoefficientsonwaveprofileand positionare
and

184

Appendices
TypicalerrorbarsonawaveprofiletracearegiveninFigureD.4.

WaveAmplitude/cm

4
3
2
1
0
1
2
3
4
0.00

0.50

1.00
Time/s

1.50

2.00

FigureD.4Errorbarsonwaveprofilemeasurements
ThefrequencyspectrumsfortheaddedresistancetestrunsaregivenintheFigureD.5andFigureD.6
1200
1000

zz /cm2s

800
600
400
200
0
0.50

1.00

1.50
2.00
WaveFrequency/Hz

50%L15Knots
150%L15Knots

50%L25Knots
150%L25Knots

2.50

100%L15Knots
200%L15Knots

3.00
100%L25Knots
200%L25Knots

FigureD.5FrequencyspectrumsforHullAaddedresistancetests
1400
1200

zz/cm2/s

1000
800
600
400
200
0
0.50

1.00

1.50

2.00

2.50

3.00

WaveFrequency/Hz
50%L15Knots

50%L25Knots

100%L15Knots

100%L25Knots

FigureD.6FrequencyspectrumsforHullBaddedresistancetests

185

Concept Design of a Fast Sail Assisted Feeder Container Ship

Appendix E Sail Design


E.1 Concept review

18

FigureE.1SkySailsSystem

19

FigureE.2NYKSuperEcoShip2030(left)andE/SOrcelle(right) concepts

18
Pictureobtainedfrom:
http://www.skysails.info/fileadmin/user_upload/documents/Dokumente/SKS_Broschueren/EN/EN_Technolog
y_Information.pdf,accessed15/04/2010.

19
Picturesobtainedfrom:NYKSuperEcoShip2030
http://www.nyk.com/english/release/31/NE_090422.html,accessed15/04/2010;E/SOrcelle
http://www.ecofriend.org/entry/esorcellecargoshipgetsitsjuicefromsunwindandwater/,accessed
15/04/2010.

186

Appendices
E.2 Design
Sailareaestimation
60

0.8
0.7

Fx/A/Nm2

PE /A / kWm-2

55

y=0.024x+0.0133

0.6
0.5
0.4
0.3
0.2

50
45
y=0.2571x+52.714

40
35

0.1

30

0.0
10

15

20

25

10

30

Vs /knots

15

20
Vs/knots

25

30

FigureE.3ExtrapolationofdataforWingsail:effectivepower(left);andaveragepropulsiveforce(right)atan
averagewindspeedof15knots
StructuralDesign
TableE.1Sailmastdimensions
Dimensions (Decktobase)
Innerradius/m
0.5
Outerradius/m
0.7
2
Sectionarea/m
1.26
SecondmomentofArea /m4 0.14
Decktosystembase/m
11
Weight/kg
36907

Liftforceiscalculatedusingthefollowingequation:
1
AV CL
L
2
Themomentleveristhedistancefromdecktosailrigsystembase,tenmetres.

(E.1)

TableE.2Mastbendingmoment,deflectionandmaximumstressusingdrivingforceestimation
Vs/knots Max.bendingmoment/Nm Max.stress/Nm2 Max.deflection/m %Max/Yieldstress
5
334074.40
1676518.69
1.36E03
1.20
10
252222.49
1265753.10
1.03E03
0.90
15
249444.70
1251813.04
1.02E03
0.89
20
249629.80
1252741.98
1.02E03
0.89
25
240656.89
1207712.34
9.80E04
0.86
30
225429.05
1131292.97
9.18E04
0.81
*Maximumdeflection
WL /3EI

187

Concept Design of a Fast Sail Assisted Feeder Container Ship


E.3 Theoretical performance

FigureE.4Forcesactingonsailassistedship[Satchwell(1989)]

Nomenclature
Nomenclaturepertinenttoappendix,allotherparametersdefinedinprincipalnomenclature

Saildrag
Drivecoefficentshipaxis)
Heelingcoefficent(shipaxis)
Totalaerodynamicforce
Totalhydrodynamicforce

Sailthrust
Aerodynamicsideforce
Saillift
Truewindangle

tan

(E.2)

0.005

(E.3)
(E.4)

188

(E.5)

Appendices

(E.6)

FormulafromSchenzle(1985)areusedtoestimatesideforce,rearrangedforleeway.Thevalues 2and
0.25areconstantsselectedforhullswithfixedrudderandoperatingpropellertoaccountforanoffsetin

lift.

| |

(E.7)

Theangleofheelisestimatedbyrearrangingthefollowingequation:
sin

189

(E.8)

Concept Design of a Fast


F
Sail Asssisted Feedeer Container Ship
E Wind
E.4
d statistics
200000300000

30
00000
Numberof
20
00000
O
Observations
1
100000

100000200000
0100000

01
4
7
0
10
13
WindSpeed/ms1 16
19

270 315
5
2
22

135 180
25

WindHeading
/degrees

0 45
4

FigureE.5Windb
bystrengthanddirectionfo
orthewholew
world(annual))

Numberof
Observations
O

30004
4000

4000
3000
2000
1000
01

20003
3000
10002
2000
01000
0
5
9
13

WindSpeed/ms1

17
21

180 215
25

ndHeading
Win
/
/degrees

0 45
4

bystrengthan
nddirectionfo
orSouthEastA
Asia(annual)
FigureeE.6Windb
1500020000

20
0000
1
Numberof 15000
1
Observations 10000
O
5000
01

1000015000
50001
10000
05000
0

5
9
13
ms1 17
WindSpeed/m
21

180 225
25

0 45
5

WindHeading
W
/degrees

FigureE.7Windbyystrengthand
ddirectionforrtheNorthAtlantic(annuall)

190

Appendices

Appendix F Wind Tunnel


F.1 Design and manufacture
TableF.1Estimationofmaximumforceexperiencedbyindividualwindtunneldynamometer
Wind
speed/ms
1

5
6
7
8
9

Max.Side
force/N
68.45
98.57
134.16
175.23
221.78

Moment Verticalforcedueto Weight


/Nm
Moment/N
/kg
91.38
261.09
23.43
131.59
375.97
23.43
179.11
511.73
23.43
233.93
668.39
23.43
296.07
845.93
23.43

Verticalforcedue
toweight/N
76.62
76.62
76.62
76.62
76.62

Maximum
resultantforce/
N
337.70
452.58
588.35
745.00
922.54

Windtunnelmodelstructuralanalysis
WL /3EI(where is
Themastisassumedtobeafixedfreeendbeamwithapointload.Thus
maximumdeflection,Wispointload,LM ismastlengthandIisthesecondmomentofareaofthemast)and
WL(whereM isthemaximumbendingmoment).Thesailwingstocksaremodelledasfixed
M
wL /384EI (where w is a uniformly distributed
fixed end beam with evenly distributed load. Thus:
wL /12.
load)andM
TableF.2Stockbendingmoment,deflectionandmaximumstressforwindtunnelmodel
Component

Sailarea
/m2

Force/
Nm1

Max.bending
moment/Nm

Max.stress
/MNm2

StockA(Main)
StockB(Flap)

0.6972
0.1394

44.2671
8.8534

10.2880
2.0576

5.4570
97.0307

Max.
deflection
/m
0.00042
0.03970

Yield
Stress
/MNm2
285.0
285.0

%Max
/Yield
Stress
1.9148
34.0459

TableF.3Maststructurecalculationforwindtunnelmodel
Shear Bending
force moment
/N
/Nm
222

110

Weight
above
mast/
kg
14

Vertical
force/
N
138

Max.
bending
stress/
MNm2
3.2

Max.
deflection
/m
9.47E05

191

Max.
normal
stress/
MNm2
0.063

Max
resultant
stress/
MNm2
3.3

Yield
stress/
MNm2
285

%
Max.
yield
stress
1.15

Concept Design of a Fast Sail Assisted Feeder Container Ship


Manufacture

FigureF.1CNCmillingmachine(left);andhotwirecutter(right)

FigureF.2Modelmanufacture:wingwithstock(left);andbase(right)

192

Appendices

F.3 Wind tunnel calibration results


TableF.4Summaryofwindtunnelcalibrationmeasurements
Weight/ Heelingforce/ Error/ Driveforce/ Error/ Heelingforce/ Error/ Driveforce/ Error/
kg
N
%
N
%
N
%
N
%
1
9.93
0.7
10.32
3
9.95775
0.4
10.01546
0.2
2
19.92
0.4
20.60
3
20.00216
0.0
20.21857
1.1
3
29.65
1.2
30.91
3
29.98211
0.1
30.30577
1.0
4
39.65
0.9
41.37
3
39.82752
0.4
40.09646
0.2
5
49.51
1.0
51.59
3
49.78548
0.4
50.05789
0.1
4
39.75
0.6
40.79
2
39.98039
0.0
40.04002
0.1
3
29.86
0.5
30.40
1
30.21798
0.7
2
20.03
0.1
19.98
0
20.13961
0.7
1
10.05
0.5
9.78
2
10.06070
0.6
Average
0.5
1.9
0.1
0.5

197

Concept Design of a Fast Sail Assisted Feeder Container Ship


F.4 Sail-container interaction figures

FigureF.3Sailcontainerinteractionatapparentwindangleof60degrees(WingA,maximumchord
spacing)Theboxeswhichcoverthebasehadtobetrimmedtoavoidanycontactwiththedynamometerlink

FigureF.4Sailcontainerinteractionatapparentwindangleof109degrees(WingB,60degreesofstagger)

198

Appendices

FigureF.5Flatsailconfiguration(WingC)fordownwindperformance

F.5 Wind tunnel corrections


2.00
1.00
0.00
2

10

12

14

16

DF,HF/N

0
1.00
2.00
3.00
4.00
5.00

AoA/degrees

Df

Hf

FigureF.6Windagedataofrigsupportingstructure,nointeractions

199

Concept Design of a Fast Sail Assisted Feeder Container Ship


Nomenclature
Nomenclaturepertinenttoappendix,allotherparametersdefinedinprincipalnomenclature

m2
Windtunnelcrosssectionarea
Totalblockagecorrection
Wakeblockagecorrection
Solidblockagecorrection,windagedata
Dragcoefficient,downwashcorrection
Dragcoefficient,blockagecorrection

Solidblockagemodelgeometryfactor

Effectiverigheight
Apparentwindangle,downwashcorrection deg

Downwashcorrectionfactor

Solidblockagecorrection

Wakeblockagecorrectionfactor

Solidblockagetestsectionfactor

Separateddragcoefficient

Downwashcorrection[ESDU(1995)]

(F.1)

(F.2)

isacoefficientbasedontheratioofthespanofthewingstothewidthofthejetandthetestsectionshape.
0.113isthevalueforoctagonalsections[ESDU(1995)].
Solidblockagecorrection
FigureF.7showsthecalculationofthedimensionalratios and forthesolidblockagecorrection:
0.96

1.10
1.05

0.94

t1wing

0.92

t1mast

0.90
1.00

0.88
0.86

0.95

0.84

0.90

0.82

K1
k3

0.80
0.78

0.85
0.00

0.10

t/c,d/l

0.20

0.30

0.00

0.20

0.40

0.60

0.80

tunnelspan/tunnelbreadth

FigureF.7Variationinsolidblockagecorrectionparameters

1.00

and

(F.3)

Windagesolidblockage
1

4

200

(F.4)

Appendices
Wakeblockage
TableF.5Bodyshapefactorsusedforwakeblockagecorrection[ESDU(1980)]

TableF.6ModelfrontalareaatzeroAoA
Component Height/m Width/m Area/m2
Wingsx3
1.670
0.189
0.31563
Topbar
0.008
1.014
0.008112
Bottombar
0.010
1.014
0.01014
Mast
0.420
0.080
0.033600
Base
0.080
0.500
0.040000
Total
0.407482

TableF.7Calculationofwakeblockagecorrectionfactorwb
Item Wings

Topbar

0.319032

16.09500 16.09500

0.00811

/ 0.019821 0.000504

Bottombar Mast
0.01014

16.09500 16.09500 16.09500


0.00063

0.00209 0.00249

0.18900

1.01400

1.01400

0.08000

0.5000

1.68800

0.00800

0.01000

0.42000

0.0800

/ 0.11197 126.75000 101.40000 0.19048

6.2500

0.033600 0.04000

2.83216

Base

201

2.82667

2.4025

Concept Design of a Fast Sail Assisted Feeder Container Ship

(F.5)

(F.6)

,1

(F.7)

TotalBlockageCorrection

(F.8)
(F.9)
(F.10)

F.6 Low Reynolds number testing

FigureF.8LowReynoldsnumberdragcurveforCLARKYtypeairfoil[Marchmann&Werme(1994)]In
particularthecurveontherightforRe200,000showssimilartrendtotheMultiwinginthespeedcalibration
runs

202

Appendices
F.7 Flow visualisation

FigureF.9LowReynoldnumbertesting.Va=6ms ,AoA=14degrees.Thefirstcolumnoftufts(20%chord)is
attachedtothewing,whilstseparationstartat30%ofthechord.Notehowthealignmentofthesecondand
thirdcolumnoftuftschangefromFigure125,unfortunatelythepictureisnotclearenough,howeverobserving
thevideoitispossibletoseethetuftsvibrate.

FigureF.10LowReynoldsnumbertesting.Va=8ms1,AoA=14degrees.Herethetuftsarealignedwiththe
flowandareattachedtothesurface.Observingthebottomrowispossibletonotetheendvortexcomingoff
thewing;thefirsttwotuftsofthisrowareseparatingfromthesurfaceduetotheinterferenceofthebottom
bar

203

Concept Design of a Fast Sail Assisted Feeder Container Ship

FigureF.11ContainerSailinteractionstudy.thegapbetweentheboxesandthebaseofthewingreplicates
thefullscalegapof2metres.Notetheredwirewhichwasusedinsteadofthesmoketovisualizetheflownear
thegap;unfortunatelythisintrusivemethodhadlimitedeffectiveness.

F.8 Extrapolation to full-scale Reynolds number


1.70

0.0160

1.60

0.0140

1.40

0.0120

1.30

CDO

MaximumCL

1.50

1.20

0.0100

1.10
1.00

0.0080

0.90

0.0060

0.80
100000

1000000

10000000

Reef

1.00E+05

1.00E+06
Reef

1.00E+07

FigureF.12NACA0015maximumliftcoefficient(left);basedragcoefficient(right)atdifferenteffective
ReynoldsNumbers[Jacobs&Sherman(1937)]

204

Appendices
F.9 Aerodynamic characteristics from wind tunnel results after correction
WindSpeedCalibration
TableF.8Measuredliftanddragcoefficientwithwingspacingof100%chord(noflapangle;nocontainer
interactions)
U=4ms1
AoA/deg (Stbd)
0
5
10
13
14
15
U=6ms1
AoA/deg (Stbd)
0
5
10
13
14
15
U=8ms1
AoA/deg (Stbd)
0
5
10
13
14
15

CD
0.0102
0.0736
0.1312
0.1754
0.1752
0.2055

CL
0.0108
0.0625
0.1228
0.1508
0.1709
0.1454

CL2
0.0001
0.0039
0.0151
0.0227
0.0292
0.0211

CD
0.0249
0.1608
0.1526
0.3894
0.4130
0.4297

CL
0.0255
0.1556
0.1455
0.3603
0.3722
0.3893

CL2
0.0007
0.0242
0.0212
0.1298
0.1385
0.1515

CD
0.0221
0.0301
0.1167
0.2303
0.2682
0.3615

CL
0.0501
0.3022
0.5476
0.6786
0.7106
0.7069

CL2
0.0025
0.0913
0.2998
0.4605
0.5050
0.4997

WingA
TableF.9Measuredliftanddragcoefficient,andlifttodragratio,forwingspacingsof100%and75%chord
(22.5degreeflapangle;nocontainerinteractions)
chordspacing

AoA /deg
0
5
10
11
12
Starboard
13
14
15
16
17
0
5
Port
10
15
16

CL
0.4162
0.7000
0.9645

1.0634
1.1155

1.1987
1.2228
1.1966
0.4373
0.7077
0.9763
1.2018
1.2286

100%
CD
CL2
0.0695 0.1734
0.1089 0.4906
0.1654 0.9315

0.1956 1.1323
0.2100 1.2460

0.2469 1.4388
0.2630 1.4971
0.3265 1.4654
0.0856 0.1912
0.1257 0.5009
0.1887 0.9532
0.2863 1.4444
0.3269 1.5094

205

CL / CD
5.9869
6.4271
5.8328

5.4354
5.3125

4.8561
4.6496
3.6656
5.1091
5.6299
5.1744
4.1975
3.7578

CL
0.3602
0.6075

0.8971

0.9848
1.0262
1.0669
1.0981
1.1183
0.4118
0.6436
0.8847
1.0573

75%
CD
CL2
0.0700 0.1298
0.1156 0.3690

0.1711 0.8048

0.2012 0.9698
0.2116 1.0530
0.2226 1.1382
0.2347 1.2059
0.2736 1.2507
0.0983 0.1696
0.1145 0.4142
0.1809 0.7827
0.2618 1.1178

CL/CD
5.1448
5.2562

5.2441

4.8952
4.8487
4.7918
4.6797
4.0872
4.1880
5.6219
4.8893
4.0391

Concept Design of a Fast Sail Assisted Feeder Container Ship


TableF.10Measuredliftanddragcoefficient,andlifttodragratio,forwingspacingsof50%and120%chord
(22.5degreeflapangle;nocontainerinteractions)
chordspacing
AoA/deg
0
5
10
Starboard 12
15
16
17
0
5
10
12
Port
14
15
16
17
18

CL
0.3082
0.5272
0.7430
0.8210
0.9318
0.9786
0.8784
0.3875

0.7950
0.8677

0.9669
0.9926
1.0051
0.9973

50%
CD
CL2
0.0976 0.0951
0.0910 0.2782
0.1347 0.5527
0.1699 0.6749
0.1974 0.8693
0.2103 0.9590
0.3175 0.8123
0.0898 0.1503

0.1703 0.6328
0.1922 0.7547

0.2298 0.9425
0.2455 0.9970
0.2598 1.0267
0.2777 1.0191

CL/CD
3.1567
5.7936
5.5155
4.8323
4.7213
4.6546
2.7668
4.3143

4.6669
4.5146

4.2073
4.0439
3.8679
3.5916

CL
0.4019
0.6802
0.9417
1.0350
1.1714

0.4152
0.7111
0.9724
1.0952
1.1810

120%(maximum)
CD
CL2
0.0724 0.1617
0.1139 0.4632
0.1539 0.8879
0.1891 1.0726
0.2347 1.3738

0.0820 0.1726
0.1213 0.5063
0.1859 0.9468
0.2169 1.2009
0.2492 1.3966

CL/CD
5.5536
5.9712
6.1184
5.4746
4.9914

5.0650
5.8642
5.2306
5.0493
4.7400

TableF.11Singlewingwith22.5degreeflapanglewithoutcontainers
CL
CD
CL2
AoA/deg
0
0.505956 0.00665 0.2563191
Starboard 10
1.254967 0.197406 1.5769551
15
1.198690 0.310085 1.4921586

CL/CD
76.08051
6.357274
3.865683

TableF.12Measuredliftanddragcoefficient,andlifttodragratioforwingspacingof120%ofchordand
flapofangleof22.5degrees
AWAofcontainers/deg
AoA/deg(Stbd)
0
5
10
14
15
16
18
19

CL
0.402
0.736
1.091
1.333
1.385
1.423
1.421
1.366

90119
CD
CL2 CL / CD
0.062 0.16 6.519
0.095 0.54 7.738
0.179 1.19 6.081
0.253 1.78 5.271
0.271 1.92 5.112
0.308 2.02 4.621
0.375 2.02 3.792
0.426 1.86 3.203

206

CL
0.547
0.845
1.155
1.384

1.464

6979
CD
CL2
0.128 0.299
0.148 0.715
0.301 1.334
0.408 1.914

0.499 2.143

CL/CD
4.263
5.700
3.833
3.391

2.932

Appendices
TableF.13Measuredliftanddragcoefficient,andlifttodragratioforwingspacingof120%ofchordand
flapangleof45degreeswithcontainersat90to119degreesapparentwindangle
AoA/deg(Stbd)
0
5
10
14
16
18

CL
0.720
1.074
1.397
1.634
1.729
1.751

CD
0.175
0.266
0.337
0.417
0.486
0.572

CL2
0.519
1.153
1.953
2.671
2.989
3.066

CL/CD
4.124
4.034
4.144
3.916
3.558
3.063

WingB
TableF.1430degreestaggerconfigurationwithwingspacingof120%ofchordwith22.5degreesflapangle
withcontainersat90to119degreesapparentwindangle
AoA/deg(Stbd)
0
5
10
14
16
18

CL
0.381
0.747
1.095
1.368
1.406
1.442

CD
0.057
0.127
0.196
0.287
0.353
0.441

CL2
0.145
0.557
1.198
1.871
1.976
2.080

CL/CD
6.663
5.872
5.577
4.773
3.979
3.270

TableF.1560degreestaggerconfigurationwithwingspacingof120%ofchordwith22.5degreesflapangle
withcontainers90to119degreesapparentwindangle
AoA/deg(Stbd)
0
5
10
14
16
18

CL
0.472
0.855
1.243
1.517
1.573
1.608

CD
0.079
0.151
0.259
0.352
0.450
0.553

CL2
0.223
0.731
1.546
2.300
2.474
2.587

CL/CD
6.010
5.646
4.809
4.310
3.498
2.906

TableF.16Extremestaggerconfigurationwithwingspacingof120%ofchordwithcontainersat90to119
degreesapparentwindangle
AoA/degree(Stbd)
5
10
15
20
25
30

CL
0.534
0.716
1.003
1.296
1.528
1.631

207

CD
0.201
0.250
0.366
0.494
0.647
0.893

CL2
0.285
0.513
1.007
1.679
2.334
2.660

CL/CD
2.658
2.863
2.745
2.625
2.360
1.826

Concept Design of a Fast Sail Assisted Feeder Container Ship


F.10 Uncertainty analysis random error component
1.75

2.25

1.55
1.75

1.35
1.15

1.25

CL,CD

0.95

CL,CD

0.75

0.75

0.55
0.35

0.25

0.15
0.05

10

15

20
0.25

0.25

10

15

20

AoA/degrees

AoA/degrees

FigureF.13StandarddeviationofwingperformancecoefficientsforWingA(left);andWingB,60degreesof
stagger(right)
2.25
2.25
1.75
1.75

1.25

CL,CD

CL,CD

1.25

0.75

0.75

0.25

0.25

0.25

10

15

20

0.25

10

15

20

AoA/degrees

AoA/degrees

FigureF.14StandarddeviationofwingperformancecoefficientsforWingA,flapat45degrees(left);and
WingB,30degreesofstagger(right)

208

Appendices

2.00
1.80
1.60
1.40

CL,CD

1.20
1.00
0.80
0.60
0.40
0.20
0.00
5

15

25

35

AoA/degrees

FigureF.15StandarddeviationofwingperformancecoefficientsforWingC

0.12

1.5

0.10

1.4
Thickness/mm

Thickness/m

F.11 Computational fluid dynamics study on Multi-wing sail system

0.08
0.06
0.04

1.3
1.2
1.1
1.0

0.02

0.9

0.00

0.8
0

y1

Positionfromleadingedge/m

Positionfromleadingedge/m

FigureF.16Boundarylayer(left)andfirstwall(right)thicknesses

TableF.17Aerodynamiccharacteristicsofwingspacingof100%ofchordwith22.5degreeflapanglemodel
AoA/degree
0.0
4.0
8.0
10.0
12.0

CL
0.74
0.96
1.11
1.14
1.05

Total
CD
0.03
0.05
0.08
0.13
0.22

CL/CD
21.33
19.67
13.94
8.75
4.72

CL
0.75
1.01
1.17
1.25
0.91

Topwing
CD
CL/CD
0.03 23.12
0.05 21.79
0.08 15.00
0.14
9.22
0.21
4.46

209

Middlewing
CL
CD
CL/CD
0.78 0.03 22.15
0.96 0.05 18.90
1.13 0.08 14.03
1.21 0.11 11.46
1.16 0.21
5.39

Bottomwing
CL
CD
CL/CD
0.69 0.04 18.96
0.92 0.05 18.47
1.03 0.08 12.82
0.96 0.15 6.41
1.09 0.25 4.35

Concept Design of a Fast Sail Assisted Feeder Container Ship

Appendix G Performance Predictions


G.1 Sailing performance
TableG.1ProbabilityofsignificantwavelengthtowaterlinelengthforSingaporearea
Wavelength/wettedlength
/ /m
6.5
5.5
4.5
3.5
2.5
1.5
0.5
Total

12
0.0000
0.0000
0.0000
0.0003
0.0017
0.0100
0.0371
0.0491

20
0.0000
0.0003
0.0010
0.0050
0.0214
0.0758
0.0995
0.2031

30
0.0003
0.0010
0.0043
0.0170
0.0595
0.1406
0.0895
0.3123

42
0.0007
0.0017
0.0063
0.0237
0.0671
0.1089
0.0381
0.2465

55
0.0007
0.0020
0.0053
0.0177
0.0414
0.0478
0.0100
0.1249

71
0.0000
0.0010
0.0030
0.0087
0.0170
0.0147
0.0017
0.0461

89
0.0000
0.0003
0.0017
0.0033
0.0053
0.0033
0.0003
0.0144

109
0.0000
0.0000
0.0003
0.0007
0.0013
0.0007
0.0000
0.0030

130
0.0000
0.0000
0.0000
0.0003
0.0003
0.0000
0.0000
0.0007

Total
0.0016
0.0063
0.0220
0.0768
0.2150
0.4018
0.2762
1

TableG.2ProbabilityofsignificantwaveheightandwavelengthtowaterlinelengthfortheCaribbeanarea

Wavelength/wettedlength
/

/m

12

20

30

42

55

71

89

109

130

Total

6.5

0.000 0.000 0.000 0.000 0.001 0.000 0.000 0.000 0.001 0.002

5.5

0.000 0.000 0.000 0.001 0.002 0.001 0.000 0.000 0.002 0.007

4.5

0.000 0.000 0.002 0.005 0.005 0.004 0.002 0.000 0.007 0.026

3.5

0.000 0.002 0.011 0.020 0.018 0.010 0.004 0.001 0.025 0.090

2.5

0.001 0.012 0.044 0.059 0.040 0.017 0.005 0.001 0.077 0.256

1.5

0.004 0.044 0.098 0.083 0.037 0.011 0.002 0.000 0.147 0.427

0.5

0.013 0.041 0.040 0.017 0.004 0.001 0.000 0.000 0.076 0.191

Total

0.018 0.099 0.195 0.186 0.107 0.043 0.013 0.004 0.334 0.874

TableG.3AveragedaddedresistanceforHullAsailingat15knotsinSingaporevaluesinNewtons
HullA15knots
/

Wavelength/wettedlength

/m

12

20

30

42

55

71

89

109

130 Total

2.5

0.00 0.00 0.00 0.00 2047.1 1997.0 1034.2 326.7 62.4 5467.5

1.5

0.00 0.00 0.00 0.00 849.9 620.3 232.7 58.8 0.0 1761.7

0.5

0.00 0.00 0.00 0.00 19.8

Total

0.00 0.00 0.00 0.00 3140.1 2625.1 1269.5 970.4 62.4 7259.4

7.8

2.6

0.0

0.0 30.2

TableG.4AverageaddedresistanceforHullBsailingat15knotsinSingaporevaluesinNewtons
HullB15knots

Wavelength/wettedlength

sigwaveheight/m 12 20 30

42

55

71

89

109

130

Total

2.5

0 0 720.94 3282.92 3799.00 2397.35 1046.76 312.79 60.84 11620.60

1.5

0 0 613.86 1916.83 1577.20 744.59 235.52 56.30 0.00 5144.30

0.5

0 0 43.42

Total

0 0 1378.22 5274.23 5412.97 3151.33 1284.90 369.09 60.84 16931.5

74.48

36.76

210

9.40

2.62

0.00

0.00 166.68

Appendices
TableG.5AverageaddedresistanceforHullAsailingat15knotsintheCaribbeanareavaluesinNewtons
HullA,15knots Wavelength/wettedlength
/m

71

89

109

130

Total

12

20

30

42

55

2.5

0.0

0.0

0.0

585.7

1979.8 1995.7 968.9 326.5 14343.3 19028.5

1.5

0.0

0.0

0.0

294.9

659.3 450.8 162.8 29.4

9900.6 10907.9

0.5

0.0

0.0

0.0

6.6

7.9

566.2

Total

0.0

0.0

0.0

909.4

2870.0 2714.1 1568.3 1260.4 25653.3 30507.1

3.1

0.0

0.0

570.7

TableG.6AveragedaddedresistanceforHullBsailingat15knotsintheCaribbeanareavaluesinNewtons
HullB15knots Wavelength/wettedlength
/

/m

11.9 19.7 30.0

42.0

54.8

71.0

89.0

109.0 130.0

Total

2.5

0.0 0.0 611.2 3323.5 4202.9 2740.6 1121.9 357.6 15996.7 28354.3

1.5

0.0 0.0 488.4 1673.7 1399.6 619.1 188.5 32.2 11041.9 15443.3

0.5

0.0 0.0 22.2

Total

0.0 0.0 1121.8 5034.6 5619.2 3364.0 1310.3 389.8 27670.1 44509.8

37.3

16.8

4.3

0.0

0.0

631.5

G.2 Propulsion
Propulsorandplantselection

FigureG.1ManufacturersVOseriespodeddrivespecifications[ABB(2009)]

211

712.2

Concept Design of a Fast Sail Assisted Feeder Container Ship

FigureG.2L50DFenginedrawing[WrtsilShipPowerTechnology(2009a)]
TableG.7L50DFengineseriesdimensionsinmillimetres,andweights[WrtsilShipPowerTechnology
(2009a)]
Engine
6L50DF
8L50DF
9L50DF

A
8115
9950
10800

B
3580
3600
3600

C
2850
3100
3100

D
3820
3820
3820

212

F
1455
1455
1455

Weight/tonnes
96
128
148

Appendices

Appendix H Design Feasibility


TableH.1Summaryofplantspecificationsusedinvoyagesimulation[dataobtainedfromWrtsilShip
PowerTechnology(2009a)andMANDiesel(2010)]

Typical(Singapore)

Typical(Caribbean)

FastFeeder

ME
MANMCC86cylinder MANMCC87cylinder Wrtsil6L50DF/8L50DF
power/kW
8280
9660
2x5700;2x7600
SFC@MCR/g.kWh1
177
177
135.2
SFC@partload/g.kWh1
175
175

AE
MAN5L21/31
MAN5L21/31

power/kW
950
950

SFC@MCR/g.kWh1
190
190

1
SFC@85%MCR/g.kWh
188
188

TableH.2Emissionsestimatesusedinvoyagesimulation[dataobtainedfromWrtsilShipPower
Technology(2009a)andMANDiesel(2010)]

LNG(MCR) LNG(partload) MDO(MCR) MDO(partload)


1

CO2/g.kWh
NOx/g.kWh1

430
1.4

450
2

630
12

630
11.5

SingaporeComparison

CaribbeanComparison

FastFeeder

30

Speed/ms1

25

20

15

10

0
0

24

48

72

96

120

144

168

192

216

240

264

288

312

336

Voyagetime /hours

FigureH.1Speedprofileforfastfeederandcomparisonshipsoverfortnightlyoperatingperiod

213

Concept Design of a Fast Sail Assisted Feeder Container Ship

SingaporeComparison

CaribbeanComparison

FastFeeder

25000

Brakepower/kW

20000

15000

10000

5000

0
0

24

48

72

96

120

144

168

192

216

240

264

288

312

Voyagetime /hours

336

FigureH.2Powerprofileforfastfeederandcomparisonshipsoverfortnightlyoperatingperiod

H.1 Performance indices


Nomenclature
Nomenclaturepertinenttoappendix,previouslydefinedsymbols

Carbonfactor(gCO2/tonnefuel)

Numberifships

Energyefficiencydesignindex

Transportefficiencyindex

highspeed(subscript)

weightingforhighspeedoperation

lowspeed(subscript)

weightingforlowspeedoperation

Transportefficiencyindex

This simple measure of ship efficiency is often defined as the ratio between volume rate of cargo and the
powerrequirement,thus

(H.1)

No account is taken in Equation (H.1) for the number of ships carrying the cargo, or operation at different
speedsorpowerrequirements.Itisdeemedappropriatetomodify
,
,

214

(H.2)

Appendices

EnergyEfficiencyDesignIndex

Thismeasureofenvironmentalperformanceisdefinedas

(H.3)

wherethesummationisusedtotakeintoaccountalloftheinstalledengines.Thecalculationistobecarried
outforadesignloadconditionatashipspeedandpowerratingcorrespondingto75%MCR.Rulesaccounting
for auxiliary engines are also defined [Andersen & Kristensen (2009)]. Taking a weighted approach when
calculatingtheEEDIleadstothefollowingmodifiedformulation:
,

.
,

(H.4)

ThecarbonfactorvaluesusedforLNGandMDOare2931200and320600respectively[IMO(2005a)]Whilst
this version of the EEDI cannot be used to compare the fast feeder to values published for existing ships, it
providesamoreaccurateassessmentofperformanceagainstthecomparisonshipswithinthecontextofthis
report.

H.2 Freight rate estimate


9000
8000

Rate/$

7000
6000
5000
4000
3000
2000
0

1000

2000

3000

4000

5000

6000

TEU

FigureH.3Freightratebasis[Hansa(2009)]

215

Concept Design of a Fast Sail Assisted Feeder Container Ship

Appendix I Seakeeping
I.1 Predicted RAO curves
1.4

1.6
1.4

1.2

135deg

1.2

90deg

1.0

45deg
0deg

0.8

1.0

HeaveRAO

HeaveRAO

180deg
135deg
90deg
45deg
0deg

180deg

0.6

0.8
0.6
0.4

0.4
0.2

0.2

0.0

0.0
0.2

0.7

1.2

0.2

1.7

0.6

Wavefrequency/rads1

1.4

1.8

2.2

2.6

Wavefrequency/rads1

FigureI.1HeaveRAOat25knots(left);and15knots(right)
8

4.5

180deg
135deg
90deg
45deg
0deg

3.5

RollRAO

3.0
2.5

180deg
135deg
90deg
45deg
0deg

7
6

RollRAO

4.0

2.0

5
4
3

1.5
2

1.0

0.5

0.0
0.2

0.7

1.2

0.2

1.7

0.6

Wavefrequency/rads1

1.4

1.8

2.2

2.6

Wavefrequency/rads1

FigureI.2RollRAOat25knots(left);and15knots(right)
1.2

1.2

PitchRAO

0.8
0.6
0.4
0.2

180deg

1.0

135deg
90deg

0.8

PitchRAO

180deg
135deg
90deg
45deg
0deg

1.0

45deg
0.6

0deg

0.4
0.2

0.0

0.0
0.2

0.7

1.2

1.7

0.2

Wavefrequency/rads1

0.7

1.2

Wavefrequency/rads1

FigureI.3PitchRAOat25knots(left);and15knots(right)

216

1.7

App
pendices
I.2 Sub
bjective motion (SM) polar
p
plots

Waveheading
/degrrees

SM
Mat20Knots
0
15

Waveheading
/degrees

Officers Lounge

BridgeDeck
45

45
5
10
7.5m h1/3
1
6.5m h1/3
1

5.5m h1/3
1
4.5m h1/3
1
90
9

90

3.5m h1/3
1
2.5m h1/3
1
1.5m h1/3
1

135

13
35
Solidline BridgeDeck
unge
Brokenlline OfficersLou
180

FigureI.4Polarplo
I
otofSMvariattionwithwaveheadingheiightforthebrridgedeckand
dofficersloun
ngeat20
k
knotsforHull
A

Waveheeading
/degrees

SSMat15Knots
0
15

Waveheading
/degreees
Deck
BridgeD

Officers Loun
nge
45

45
10
7.5m h1/3
1
5

6.5m h1/3
1
5.5m h1/3
1

9
90

90

4.5m h1/3
1
3.5m h1/3
1
2.5m h1/3
1
1.5m h1/3
1

135

135
Solidline BridgeDeck
unge
Brokenlline OfficersLou
180

FigureI.5Polarplo
I
otofSMvariattionwithwaveheadingheiightforthebrridgedeckand
dofficersloun
ngeat15
k
knotsforHull
A

217

Conceptt Design of a Fast Sail Assisted


A
Feeeder Contaiiner Ship

Waveeheading
/ degrrees

SSMat10Knots
0
15

Waveheading
/ degrees
BridgeDeck

Officers Lou
unge
45

45
5
10
7.5m h1/3
1
5

6.5m h1/3
1
5.5m h1/3
1

90
9

90

4.5m h1/3
1
3.5m h1/3
1
2.5m h1/3
1
1.5m h1/3
1

135

13
35
eck
Soliidline BridgeDe
Bro
okenline OfficerrsLounge
180

FigureI.6Polarplo
I
otofSMvariattionwithwaveheadingheiightforthebrridgedeckand
dofficersloun
ngeat10
k
knotsforHull
A

218

Appendices

Appendix J Predicted Roll Damping


J.1 Naked hull roll damping
TableJ.1Rolldampingtimeseries
Time/s
12.13
26.61
41.04

Peakangle/deg
25.67
18.74
13.68

Period/s

14.48
0.043
14.43
0.043
14.53
0.043

FigureJ.1Rolldecaycurve

FigureJ.2Frameofreferenceforrolldampingcalculations

219

Concept Design of a Fast Sail Assisted Feeder Container Ship


J.2 Lifting surface method

1
2
1

2

(J.1)

(J.2)

(J.3)

5.0

4.0

NA

3.0

2.0

1.0

0.0
5

25

45

65

85

105

125

145

165

185

Apparentwindangle / degrees

FigureJ.3Variationinaerodynamicrolldampingincidencechangescoefficientcomponentwithapparent
windangle
0.06

0.04

NAV

0.02

0.00
5

25

45

65

85

105

125

145

165

185

0.02

0.04

0.06
Apparentwindangle / degrees

FigureJ.4Variationinaerodynamicrolldampingairspeedchangescomponentwithapparentwindangle

220

Appendices
J.3 Lifting line method
Thecomponentsofrolldampingwerecalculatedbasedonaregressionanalysisofa20metrerectangular
planformaerofoil.Thecomponentsaregivenby
1
2

1
2

and
where 1 and
foundas:

(J.4)
2 ,

(J.5)

2 are a function of the span/air gap ratio. The total aerodynamic damping coefficient is

(J.6)

J.4 Control system


Thecalculatedliftcoefficientissubtractedfromtheoperationalliftcoefficienttoobtaintherequiredmarginto
avoidstall:
|

221

|.

(J.7)

Concept Design of a Fast Sail Assisted Feeder Container Ship

Appendix K Structural Design


K.1 Midship scantling calculations
Nomenclature
Designbendingstress
Criticalbucklingstress
Elasticbucklingstress(from
Part3,Chapter3,Section7of
theRules)

Arrangement
TableK.1PanelparticularsderivedfromtheNG254basisship

Panel
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Note1

Panel Vertical
length, depth,

Function

Keel
BottomShell
Bilge
SideShell
SideShell
SideShell
SideShell
Sheerstrake
MainDeck
InnerHull
InnerHull
InnerHull
InnerHull
InnerHull
BilgeBoxTop
BilgeBoxSide
InnerBottom
InnerBottom
DuctKeelSide
Girder
Stringer
Stringer
Stringer
Stringer

Height
above
base,

/m

/m

/m

/mm

/mm

2.000
7.538
5.749
2.627
2.602
2.492
2.629
2.911
1.708
2.989
2.593
2.491
2.629
2.911
2.188
2.682
6.949
1.669
2.602
1.895
1.205
1.308
1.708
1.708

0.139
0.967
4.738
2.599
2.593
2.491
2.629
2.911

2.989
2.593
2.491
2.629
2.911

2.682

0.000
2.602
1.895
0.000
0.000
0.000
0.000

0.050
0.258
3.078
7.047
9.641
12.185
14.745
17.515
18.970
6.852
9.643
12.185
14.745
17.515
5.358
4.017
2.676
2.676
1.375
1.729
5.358
7.949
13.131
16.359

965
580
575
525
434
498
438
485
342
598
432
498
438
485
547
894
535
477
651
632
600
436
427
342

19.0
16.0
16.0
15.0
12.5
12.5
18.0
20.0
20.0
12.0
10.0
10.0
16.0
20.0
10.0
12.0
10.0
10.0
15.0
12.0
11.0
10.0
18.0
20.0

No.
Stiffener
Stiffeners profile
/mm
1
12
9
4
5
4
5
5
4
4
5
4
5
5
3
2
12
3
3
2

2
3
4

Keelstiffenerissingleflatbarpositionedoncentreline,allotherprofilesareCorusbulbflats.

222

650x12
240x10
240x10
220x10
180x8
180x8
180x10
200x11
200x11
220x10
180x8
180x8
180x10
200x11
200x10
220x10
220x10
200x11
220x10
220x10

200x11
180x10
200x11

Appendices
Directcalculations
TableK.2Panelplatingdesignstressesforthemidshipsection
Panel
No.
1
2
3
4
5
6
7
8
9
10
11
12

/Nmm2

/m4

/Nmm2

137.5
125.5
116.5
64.1
39.1
38.5
13.6
12.0
0.0
145.5
67.4
41.0

2.600
9.391
3.074
0.068
0.147
0.722
2.757
6.672
5.513
0.090
0.124
0.606

242.1
227.4
180.9
75.7
61.9
95.0
104.2
139.4
146.7
159.8
90.2
97.6

Panel
No.
13
14
15
16
17
18
19
20
21
22
23
24

/Nmm2

/m4

/Nmm2

93.3
112.3
99.0
78.8
36.9
95.6
174.4

2.509
6.672
0.195
0.603
2.875
0.650
2.076
1.105
0.087
0.000
1.016
3.214

90.6
127.3
127.4
164.2
168.7
148.5
123.9
177.9
208.5
0.3
69.1
112.0

Rulecalculations
Rule checks for buckling strength of plate strakes and stiffeners were conducted in accordance with Part 3,
Chapter4,Section7oftheRules.TheDesignbendingstress wascalculatedbymeansofEquation(K.1),and
byEquation(K.2).
criticalbucklingstresses
,forstructuralmembersabovetheneutralaxis,or
, forstructuralmembersbelowtheneutralaxis
1

223

(K.1)

(K.2)

Concept Design of a Fast Sail Assisted Feeder Container Ship


TableK.3Rulebucklingchecksforpanelplateandlongitudinalstiffeners
PlateBuckling
Panel
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

Design

/Nmm2
187.3
182.4
114.7
41.7
41.7
51.3
81.6
114.5
131.8
41.7
41.7
51.3
81.6
114.5
60.1
92.2
124.4
124.4
155.6
147.1
60.1
41.7
62.5
100.8

Elastic

/Nmm2
219.7
282.1
275.6
247.6
281.8
258.4
280.3
327.3
341.3
216.0
193.1
146.4
242.1
327.3
158.6
92.8
166.1
204.1
174.4
185.5
202.1
190.2
196.9
341.3

PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS

Elastic

/Nmm2
347.2
295.4
300.2
318.3
297.6
293.7
277.5
271.7
286.9
316.3
306.1
303.7
283.9
271.7
270.9
224.4
331.6
328.9
288.3
284.0

284.1
284.7
286.9

224

StiffenerBuckling
Torsion

/Nmm2
348.7
308.8
315.1
310.2
315.0
313.0
318.4
329.7
330.9
310.4
300.1
296.0
309.7
329.7
299.2
264.9
274.1
282.4
292.3
295.2

299.3
298.1
331.3

Web

/Nmm2
224.2
330.3
337.5
334.2
334.6
334.6
340.4
340.6
340.6
334.2
334.6
334.6
336.2
340.6
325.7
325.7
319.6
330.6
340.6
334.6

325.7
325.7
340.6

PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS
PASS

PASS
PASS
PASS

TYPICAL WATERTIGHT BULKHEAD


AMIDSHIPS
17.5 PL (EH) x 342
200 x 10 BF

DETAIL OF VERTICAL STIFFENERS


BETWEEN FRAMES POSITIONED ON A
MAX. SPACING OF 1000mm

1000 x 10
400x10
/100x10

8 PL

400x10
/100x10
.

17.5 PL (EH) x 435


200 x 10 BF

7.5 PL x 518
200 x 8.5 BF

12 PL
400x10
/100x10

10 PL

17.5 PL (EH) x 427


200 x 10 BF

11 PL x 518
200 x 8.5 BF

210 X 10 BF
1250 x 13 / 600 x 20

16359 AB

9 PL x 427
180 x 8 BF

7.5 PL x 518
200 x 8.5 BF

10 PL

11 PL x 538
180 x 9 BF

12 PL

9 PL x 538
180 x 8 BF

17.5 PL

9 PL x 460
180 x 8 BF

13.5 PL

10 PL x 600
180 x 8 BF

12 PL

PL
22 (DH
0x )x
10 57
BF 4

10 PL x 580
280 x 10.5 BF

10.5 PL x 651
200 x 10 BF

12 PL x 632
200 x 10 BF

2676 AB

14

10 PL
10 PL

7949 AB

5358 AB

12PL x 894
180 x 8 BF

8.5 PL x 473
260 x 10 BF

10540 AB

11 PL x
5
240 x 1 74
0 BF

11 PL x 518
240 x 10 BF

11387

10 PL

13131 AB

11 PL x 518
200 x 8.5 BF

200 x 10 BF

TYPICAL FRAME AMIDSHIPS

17.5 PL (EH) x 435


200 x 10 BF

18970 AB

12 PL

BASELINE

650 x 12 FB
1419
16 PL x 580
220 x 10 BF

19 PL (EH)

ALL UNMARKED TRANSVERSE STIFFENING 100 x 10 FB


FB - FLAT BAR
BF - CORUS SHIPBUILDING FLAT BAR
ALL STEEL LR GRADE H36, YIELD STRESS 355 N/mm

8949

TYPICAL FRAME & TRANSVERSE BULKHEAD


AMIDSHIPS
Designed by
SM

Checked by
BS

APPENDIX K.2

Date
19/04/10

Scale

Edition
A

Sheet

University of Southampton
University Road
Highfield
Southampton
SO17 1BJ

1:100
1 of 2

3000

MAST STIFFENER
(CUSTOM SECTION)

17.5 PL (EH)

30 PL (EH)

100

150

25 PL

2000

11000

17.5 PL

30 PL / 120 x 30 30 PL / 120 x 30
30 PL (EH)

20 PL
30 PL (EH)

SAIL SUPPORTING TRANSVERSE BULKHEAD


(ALL UNMARKED SCANTLINGS AS PER TYPICAL
TRANSVERSE BULKHEAD)

SAIL MAST AND SUPPORTING BULKHEAD

Designed by
BS

Checked by
SM

APPENDIX K.2

Date
19/04/10

Scale
1:200

Edition
A

Sheet

University of Southampton
University Road
Highfield
Southampton
SO17 1BJ

2 of 2

Appendices
K.3 Global strength
Nomenclature

Longitudinalstructuralareanormalisedtoamidships

Hullmassestimate

The structural area distribution factor used to calculate the hull coffin diagram is shown in Table K.4, along
with the assumed structural element extents from which it is derived. The tapered factor is obtained by
applyingthethicknesstaperratiosfromPart3,Chapter3,Section2oftheRules.
TableK.4Assumedextentofmajorstructuralelementsalongthelengthoftheshipandstructural
distributionfactor

/m
2.98
0.00
4.00
8.01
12.01
16.01
24.02
32.02
40.03
48.03
64.04
80.05
96.06
112.07
120.08
128.08
136.09
144.09
148.09
152.10
156.10
160.10
170.40

Shell
Girth
/m
38.71
39.30
40.41
42.40
53.42
53.63
52.92
51.90
52.39
54.07
56.43
56.89
55.04
50.19
48.45
47.55
46.91
46.71
46.63
46.23
45.48
45.39
0.00

Deck
Beam
/m
6.60
6.60
6.60
6.60
6.60
6.60
6.60
6.60
6.60
6.60
6.60
6.60
6.60
6.60
26.19
26.19
26.19
25.23
23.58
20.89
17.48
12.99
0.00

InnerHull
Depth
/m
7.21
9.70
10.14
10.49
11.08
16.06
16.06
21.25
27.23
27.23
27.23
27.23
27.23
27.23
24.09
24.09
24.09
11.08
10.26
9.24
8.09
6.66
0.00

227

Bottom
Breadth
/m

0.00
9.68
10.94
13.45
15.97
18.49
19.80
22.42
22.42
22.42
19.91
17.55
15.20
12.84
4.76
4.76
4.76
4.76
4.76
0.00

0.464
0.491
0.505
0.526
0.714
0.771
0.787
0.846
0.925
0.952
0.996
1.000
0.984
0.919
1.028
0.999
0.973
0.776
0.753
0.717
0.670
0.617
0.000

0.317
0.343
0.365
0.394
0.543
0.603
0.656
0.750
0.872
0.952
0.996
1.000
0.984
0.919
0.963
0.874
0.791
0.591
0.553
0.509
0.460
0.414
0.317

Concept Design of a Fast Sail Assisted Feeder Container Ship

Loadconditions

The load conditions used in the global strength analysis were based on those required by the IMO Intact
Stability Code IMO (2008b) as they most adequately represented the likely behaviour of the ship. These
consisted of fully loaded and ballast conditions, modelled at both departure (with full stores and fuel) and
arrival (with 10% stores and fuel remaining). Subject to propeller immersion and restrictions on maximum
draught,theloadconditionswereballastedtoattainsuitabletrim.
TableK.5Lightshipmassdistribution(includingcrewandstores)
Item

Quantity

FwdSail
AftSail
Fwdmast
Aftmast
Crew
Stores
Accom.Tier1
Accom.Tier2
AccomTier3.
AccomTier4
AccomTier5
Crane
Crane
Podpropulsionmodule
SteeringModule
SlipRingUnit
CoolingAirUnit
HydraulicPowerUnit
OilTreatmentUnit
GTU+AIU+LBU+ACU
Podpropeller
Shaftpropeller
Refrigerationsystem
Shaftmotor
Shaft/m
BowThruster
AftEngine
ForwardEngine
Misc.engineweight
Weightmargin
Hullmass
Anchors
Chain

1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
1
1
1
2
1

UnitMass
/tonnes
14.00
14.00
18.00
18.00
1.8.00
50.00
93.35
91.72
85.17
86.80
65.51
126.00
126.00
78.00
44.0
3.00
7.00
4.50
0.60
0.50
5.00
5.00
81.00
6.50
0.77
7.50
93.00
148.00
641.00
135.50
4577.02
4.32
36.93

228

TotalMass
/tonnes
14.00
14.00
18.00
18.00
1.80
50.00
93.35
91.72
85.17
86.80
65.51
126.00
126.00
78.00
44.00
3.00
7.00
4.50
0.60
0.50
5.00
5.00
81.00
6.50
0.77
7.50
186.00
296.00
641.00
135.50
4577.02
8.64
36.93

Long.Arm
/m
79.290
23.315
79.290
23.315
0.000
147.480
148.480
145.440
142.830
139.660
136.480
125.730
115.200
0.000
0.000
1.130
2.500
2.500
2.500
2.500
3.700
6.290
130.000
16.740
10.000
137.780
118.400
133.740
133.740
160.000
80.430
164.100
164.100

Vert.Arm
/m
43.470
43.470
24.470
24.470
0.000
20.650
20.650
23.241
25.832
28.423
31.014
0.000
0.000
4.500
11.700
11.700
11.700
11.700
11.700
11.700
4.000
4.000
12.900
4.000
4.000
2.767
12.930
12.940
12.90
15.290
11.510
0.000
0.000

Appendices
TableK.6Tankmassdistributionforeachloadingcondition(fortanknamesandpositionsseethegeneral
arrangement,AppendixM)

Fullloaddeparture
Unit
Mass/
Mass/ Quantity
tonnes
tonnes
329.112
98% 322.530
123.643
98% 121.170
273.418
98% 267.950
50.437
98% 49.428
407.295
98% 399.149
167.432
10% 16.744
355.310
0%
0.000
192.236
0%
0.000
625.370
0%
0.000
532.754
0%
0.000
281.832
0%
0.000
316.282
0%
0.000
494.182
0%
0.000
271.502
0%
0.000
297.024
0%
0.000
445.340
0%
0.000
192.336
0%
0.000
280.386
0%
0.000
303.812
0%
0.000
814.392
0%
0.000
234.908
0%
0.000
143.300
0%
0.000
1305.876
0%
0.000
187.594
0%
0.000
104.681
0%
0.000
187.026
0%
0.000
187.026
0%
0.000
374.052
0%
0.000
277.829
0%
0.000
9756.400
1177

Tank
FW
LNG1
LNG2
LOT
FPT
DBT1
DBT2
TS1
LWT1
DBT3
LWT2
TS2
DBT4
LWT3
TS3
DBT5
LWT4
TS4
DBT6
LWT5
TS5
DBT7
LWT6
TS6
DBT8
LWT7P
LWT7S
APT
Total

Fullloadarrival
Mass/
tonnes

Quantity
9%
9%
9%
9%
98%
98%
0%
0%
0%
98%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%

29.620
11.128
24.608
4.539
399.149
164.084
0.000
0.000
0.000
522.098
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
1155.2

Ballastdeparture
Quantity
98%
98%
98%
98%
0%
0%
98%
0%
98%
98%
98%
0%
98%
98%
0%
98%
98%
0%
98%
98%
98%
98%
98%
98%
98%
98%
98%
98%
98%

Mass/
tonnes
322.530
121.170
267.950
49.428
0.000
0.000
348.200
0.000
612.860
522.100
276.200
0.000
484.300
266.070
0.000
436.430
188.490
0.000
297.740
798.100
230.210
140.430
1279.800
183.840
102.588
183.285
183.285
366.570
272.272
7933.800

Ballastarrival
Quantity

9%
29.620
9%
11.128
9%
24.608
9%
4.539
98% 399.149
98% 164.080
98% 348.200
98% 188.390
98% 612.860
98% 522.100
98% 276.200
0%
0.000
98% 484.300
98% 266.070
0%
0.000
98% 436.430
98% 188.490
0%
0.000
98% 297.740
98% 798.100
98% 230.210
98% 140.430
98% 1279.800
98% 183.840
98% 102.588
98% 183.285
98% 183.285
98% 366.570
98% 272.272
7994.300

TableK.7Fullloadcargomassdistribution(zerocargomassassumedinballastcondition)
Item
Hold10
Hold9
Hold8
Hold7
Hold6
Hold5
Hold4
Hold3
Hold2
Hold1

Mass Long.Arm Ver.Arm


/tonnes
/m
/m
677
5.302
10.582
1054
18.742
9.468
1284
34.012
8.853
1484
47.542
8.237
1692
61.062
7.621
1734
74.582
7.621
1773
89.932
7.621
1442
103.532
8.853
809
117.142
10.470
418
127.672
20.212

229

Mass/
tonnes

Concept Design of a Fast Sail Assisted Feeder Container Ship


K.4 FE midship section model results
Hullstrengthresults

FigureK.1HighstressregioninthebottomshellthatdonotmeettheacceptancecriteriaofTable7.18inthe
initialanalysis(left);andthemodifiedanalysis(right)forthestillwatercondition

FigureK.2HighstressregioninthebottomshellthatdonotmeettheacceptancecriteriaofTable7.18inthe
initialanalysis(left);andthemodifiedanalysis(right)fortheheadseascondition

FigureK.3HighstressregioninthebottomshellthatdonotmeettheacceptancecriteriaofTable7.18inthe
initialanalysis(left);andthemodifiedanalysis(right)fortheobliqueseascondition

230

Appendices
MastStrengthResults

FigureK.4Stressinthefinemeshzoneatthebaseofthemastwhichdoesnotmeettheacceptancecriteria
ofTable7.19fortheinitialanalysis(left);andthemodifiedanalysisinthebeamseascondition

FigureK.5Stressinthefinemeshzoneatthebaseofthemastwhichdoesnotmeettheacceptancecriteria
ofTable7.19fortheinitialanalysis(left);andthemodifiedanalysisintheobliqueseascondition

231

Concept Design of a Fast


F
Sail Asssisted Feedeer Container Ship

Append
dix L Sta
ability
L Stability model definition
L.1
d

a)

b)

c)

delusedintheeanalysis(reffertoAppendixMforarran
ngementandffunctionofsp
paces),
FigureL.1Stabilitymod
showing:((a)distribution
nofinternalssubdivision;(b
b)dispositionoftankspacees;and(c)allo
ocationofinte
ernal
com
mpartments

L Intactt stability
L.2
Nom
menclature
Nomenclaturrepertinenttoappendix.O
Otherparamettersdefinedin
nprincipalno
omenclatureo
orreporttext.
Mouldedbreadth
Bread
dthofshiponthewaterlineeathalf
mean
ndraught.
Mean
nwidthofhattchcoamingsw
within /4
forwaardandaftofamidships.
Formfactorfor .

m
m

Mouldeddepthoftheship,corrrectedfor
olumeswithinthehatch
definedpartsofvo
mings.
coam

232

Mean draugh
ht.
Meanheightofhatchcoam
mingswithin
midships.
/4forwardandaftofam
mingwithin
Length ofeacchhatchcoam
/4forwardandaftofam
midships.
ntreofmassaabovebase,
Heightofcen
correctedforrfreesurfaceeffect,notto
betakenasleessthan .
Lengthofship.

m
m
m
m

Appendices
Calculationofcoefficients
TableL.1Summaryofvaluesusedinintactstabilityanalysis
/
/
/
/
/
/
/

9.482
18.970
0.000
0.000
26.190
0.000
159.890
9.760
0.551
0.806
25.847
24.420

/
/
Mouldeddepthiscalculatedas

18.97.

(L.1)

Theformfactoristhusderived:
1000

233

0.109916.

(L.2)

Concept Design of a Fast Sail Assisted Feeder Container Ship

Fullloadarrival

steadywind

Fullloaddeparture

steadywind

gustwind

InitialGMT

gustwind

InitialGMT

GZ / m

GZ / m

GZcurves

Severewindandrollingwindheeling

Severewindandrollingwindheeling
0

2
30

30

60

90

120

Angle of heel / degrees

150

180

30

30

60

90

120

Angle of heel / degrees

150

180

FigureL.2Fullloaddeparture(left)andarrival(right)GZcurves
Ballastdeparture

steadywind
InitialGMT

GZ / m

GZ / m

gustwind

Severewindandrollingwindheeling

2
30

30

60

90

120

150

180

Angle of heel / degrees

steadywind

gustwind

InitialGMT

Severewindandrollingwindheeling

Ballastdeparture

30

30

60

120

Angle of heel / degrees

FigureL.3Ballastdeparture(left)andarrival(right)GZcurves

90

234

150

180

Area1/Area2

235
0.382
15.114

>

>

4.9.2.1:Area0to40

4.9.2.2:Area30to40

4.9.2.3:MaximumGZat30orgreater

4.9.2.4:ValueofmaximumGZ
4.9.2.5:AreaunderGZcurveto
downflooding

0.3

3.1283

8.3423

4.6925

100

80

16

0.15

25

0.2

1.7189

4.9.2.1:Areato30

4.9Containershipslargerthan100m

Angleofsteadyheel
Angleofsteadyheel/Deckedge
immersionangle

3.1.2.4:InitialGMt

3.2.2:SeverewindandrollingCriteria

3.1.2.2:MaxGZat30orgreater

3.1.2.3:AngleofmaximumGZ

3.1.2.1:Area30to40

5.1566

m.deg

m.deg

m.deg

m.deg

deg

deg

m.deg

m.deg

172.032

2.5840

2.5840

17.8346

39.6537

21.8192

497.48

12.37

2.20

2.687

58.20

2.584

17.8346

39.6537

21.8192

3.1.2.1:Area0to40

m.deg

3.1513

3.1.2.1:Area0to30

1038.19

576.44

761.33

470.11

375.33

364.98

397.48

84.54

86.25

1691.33

132.80

1192.00

937.56

668.99

592.39

161.743

2.4840

2.4840

16.8892

37.3356

20.4465

492.17

13.06

2.40

2.514

59.10

2.484

16.8892

37.3356

20.4465

970.11

550.26

728.00

439.88

347.55

335.73

392.17

83.67

85.00

1576.00

136.40

1142.00

882.56

624.04

548.83

308.705

4.2130

4.2130

24.8619

55.2814

30.4195

486.43

8.62

2.30

3.773

67.300

4.213

24.8619

55.2814

30.4195

1942.43

1002.88

1304.33

694.74

562.66

548.26

386.43

89.22

85.63

2415.33

169.20

2006.50

1346.38

972.05

865.30

276.85

3.9160

3.9160

22.7360

50.4755

27.7395

472.63

9.55

2.60

3.431

67.300

3.916

22.736

50.4755

27.7395

1731.69

925.13

1205.33

626.78

505.05

491.15

372.63

88.06

83.75

2187.33

169.20

1858.00

1222.71

878.85

780.26

Margin/%

Ballastarrival

Margin/% Actual

Ballastdeparture

Margin/% Actual

Fullloadarrival

Margin/% Actual

Actual

Value

A.749(18)Ch3Designcriteriaapplicabletoallships

Unit

Fullloaddeparture

Code

TableL.2Intactstabilityresultssummary,applyingIMOA.749(18)Ch3and4.9code

Pass

Pass

Pass

Pass

Pass

Pass

Pass

Pass

Pass

Pass

Pass

Pass

Pass

Pass

Pass

Status

Concept Design of a Fast Sail Assisted Feeder Container Ship

Concept Design of a Fast Sail Assisted Feeder Container Ship

L.3 Damage stability


Nomenclature
Nomenclaturepertinenttoappendix.Allotherparametersdefinedinprincipalnomenclatureorreporttext.
Damagezonenumber

1, 2,

Aftextentofdamagezone
Forwardextendofdamage
2
zone
Probabilityofdamagebasedon
1, 2
transversesubdivision
Longitudinalbulkheadnumber

infromshell,
0atshell
Distancefromsideshellto

longitudinalbulkheadsinzone

1, 2,

Reductionfactorbasedonlongitudinal
subdivision
Equilibriumheelanglefordamagecase
Downfloodingheelanglefordamage
case
Maximumpositiverightingleverupto
angle
Reductionfactorbasedonlongitudinal

subdivision

Calculationof
Theprobability foraparticulardamagecaseiscalculatedbyeither
1, 2

1, 2,

1, 2,

(L.3)

whenthedamageinvolvesasinglezone,or
1, 2

1, 2

1, 2
1

1, 2

1, 2,

, 2

, 2

1, 2,
,

ifthedamageinvolvestwoadjacentdamagezones.Inallcases,
1 , 2 and

Valuesofand

1, 2,

,
(L.4)

, 2

1, 2,

0.

werecalculatedbythemethodinSOLASChapterII1,PartB1,

Regulation 71 for all damage zones. The longitudinal extents and longitudinal bulkhead position for each
damagezoneareshowninTableL.3withthelowestvaluefor beingusedforthecalculationofvaluesof
1, 2,
inaccordancewiththeprocedureoutlinedintheIMOsexplanatorynotestodamagestability
[IMO(2008a)].TableL.2showstheresultingprobabilities obtainedforalldamagecasesbyEquations(L.3)
and(L.4)asappropriate.
TableL.3Longitudinalextentsofdamagezonesandlocationoflongitudinalbulkheadconsideredfordamage
limitation,measuredfromtheaftextentandthesideshellrespectively
Zone
1
2
3
4
5
00.00
12.39
25.92
41.19
54.71
1
2 12.39 25.92 41.19 54.71 68.23
n/a 3.06 3.37 1.27 1.43

6
68.23
81.79
1.49

236

7
81.79
97.15
1.43

8
97.15
112.00
3.33

9
112.00
129.61
5.72

10
11
129.61 156.83
156.83 170.70
5.81 n/a

Appendices
TableL.4Probabilitiespiforalldamagecases,withintermediateresultsforsinglezonedamages
Damage
Zone(s)
Case
1
1
2
2
3
3
4
4
5
5
6
6
7
7
8
8
9
9
10
10
11
11
12
8,9
13
9,10
14
10,11

1, 2

1, 2,

0.0508
0.0345
0.0440
0.0345
0.0345
0.0347
0.0445
0.0416
0.0586
0.1395
0.0588
0.1905
0.2871
0.2578

1.0000
1.9936
1.7327
1.0364
1.1475
1.1818
0.9333
1.8000
1.7262
1.0742
1.0000
1.0000
1.0000
1.0000

0.0508
0.0688
0.0763
0.0358
0.0396
0.0410
0.0415
0.0749
0.1011
0.1499
0.0588
0.0144
0.0361
0.0492

Calculationof
Forcargoships, iscalculatedfromthefinalequilibriumstageoffloodingonly,ignoringintermediatestages
andneglectingtheinfluenceofheelingmoments.Probabilityofsurvivalatthefinalstageisdefinedasshownin
Equation(L.5).Theheelangleandrightinglevervaluesobtainedforalldamagecasesanddraughtsareshown
inTableL.5alongwiththeresultantprobabilities.

0.05

16

1,

(L.5)

30 .

25

, 25

where

30

0,

TableL.5Probabilitiessiforalldamagecasesandthethreeloadingconditions
Deepestsubdivisiondraught, Partialsubdivisiondraught,
Case
/
/
/
/

1
0.0 17.4 1.176 1.0000 0.0 20.4 2.175 1.0000
2
0.4 13.4 1.185 0.9494 0.0 19.4 2.325 1.0000
3
3.2 13.8 1.143 0.9022 1.9 19.1 2.254 1.0000
4
3.0 15.7 1.204 0.9439 0.4 19.3 2.337 1.0000
5
2.7 14.7 1.134 0.9306 0.5 19.0 2.224 1.0000
6
3.0 13.9 1.134 0.9085 0.6 17.7 2.182 1.0000
7
3.2 12.6 1.061 0.8755 1.5 16.4 2.018 0.9824
8
0.8 14.3 1.153 0.9584 0.6 18.0 2.223 1.0000
9
4.1 12.1 0.725 0.8409 2.1 14.8 1.539 0.9439
10
4.7 10.6 0.656 0.7793 2.4 13.4 1.436 0.9106
11
0.0 17.7 1.396 1.0000 0.0 19.2 2.269 1.0000
12
0.7 12.7 1.161 0.9306 0.5 16.0 2.217 0.9921
13
5.8 8.6
0.582 0.6468 3.5 10.8 1.299 0.8219
14
4.6 10.7 0.672 0.7858 3.2 11.4 1.274 0.8461

237

Lightservicedraught,
/
/

0.0 27.2 2.269 1.0000


1.9 25.8 2.653 1.0000
3.0 26.0 2.693 1.0000
1.4 26.1 2.668 1.0000
0.3 25.4 2.439 1.0000
0.4 24.6 2.390 1.0000
0.5 24.1 2.332 1.0000
1.2 25.5 2.520 1.0000
0.0 24.2 1.996 1.0000
0.0 21.9 1.873 1.0000
0.0 26.4 2.475 1.0000
0.9 24.1 2.550 1.0000
1.2 19.2 1.796 1.0000
0.7 19.8 1.749 1.0000

Appendices

Appendix N Budget Summary


TableN.1Budgetsummary
Item

Towingtankmodels

Windtunnelrig

Hireofwindtunneland
towingtank
Transport
Photocopying,printing&
binding

Deductions

Homeblown HighDensityFoam(2x
120kg.m3+2x200kg.m3)
SP'Ampreg22'Resin
Araldite
HighBuildprimerspraypaint
Carwingmirror
Grippaperandtape
Bluefoamforwings
Bolts
Boxesforcontainers
Aluminiumrodforflapstock
Aluminiumplate
Aluminiumtube
Aluminiumflatbar
Aluminiumrod
Aluminiumrod
Polyfiller
Windtunnelhire(academicrate)
SSUtowingtankhire
TraintickettoLloydsRegisterinDecember
TraintickettoLloydsRegisterinJune
TransporttoSSUtowingtank
Printingreport
PrintingGAdrawings
Initialbudget
Elevatorpitch
Lloydssponsorship
Windtunnelhire
Towingtankhire
Individualcontributionfortrainticketsto
Lloyds
Individualcontributiontomodelparts
IndividualcontributionfortransporttoSSU
towingtank
Individualcontributionforprinting

Qty

Priceper
item/

Subtotal /

1432.00

1432.00

1
3
3
1
1
2
1
1
1
1
1
1
1
1
1
3
3
6
6
12
3
3
1
1
1
3
3

31.98
4.10
4.32
3.99
16.10
23.00
26.98
59.80
11.50
58.80
46.00
10.99
32.82
11.75
6.56
1716.00
300.00
27.80
27.80
2.00
16.00
12.00
780.00
160.00
523.52
1716.00
300.00

31.98
12.30
12.96
3.99
16.10
46.00
26.98
59.80
11.50
58.80
46.00
10.99
32.82
11.75
6.56
5148.00
900.00
166.80
166.80
24.00
48.00
36.00
780.00
160.00
523.52
5148.00
900.00

12

27.80

333.60

32.61

195.64

4.00

24.00

6
14.00
84.00
Totalspent
8310.14
Deductions
8148.76
Projectdeficit
161.38

Thedeficitof161.38indicatedinTableN.1isattributedtoanunderestimateforthepriceoftheHomeblown
High Density Foam which was initially estimated to cost 1045.45 but cost 1432 when ordered, adding
386.55 to the project costs. A total of 18 days of manufacturing time were required by the EDMC for the
manufactureofthetowingtankmodelsandwindtunnelrig.Anadditional15daysofmanufacturingtimewas
requiredbyteammembersonmodelmanufacture.TherequiredmanufacturingtimeintheEDMCwasgreater
thantheinitialprojectallocationoffivedays,however,itwasagreedinadvancethattheEDMCwouldbeable
toundertaketheworkonourbehalf.

239