A study on the development of energy saving device for the twin screw car ferry

(1)

A study on the development of energy saving device for the twin screw car ferry

2축 프로펠러를 가진 57m급

차도선의 연료절감 부가물 개발에 관한 연구

조선대학교 대학원

선 박 해 양 공 학 과

임 연 지

2013년 8월 석사학위 논문

년 월

석 박 사 학 위 논 문

성 명

[UCI]I804:24011-200000263889

(2)

A study on the development of energy saving device for the twin screw car ferry

차도선의 연료절감 부가물 개발에 관한 연구

2013년 8월 23일

조선대학교 대학원

임 연 지

(3)

A study on the development of energy saving device for the twin screw car ferry

차도선의 연료절감 부가물 개발에 관한 연구

지도교수 이

귀 주

이 논문을 석사학위신청 논문으로 제출함

2013년 4월

조선대학교 대학원

임 연 지

(4)

임연지의 석사학위논문을 인준함

위원장 Chalmers 대학교 교 수 Carl-Erik Janson(인) 위 원 조선대학교 교 수 이 귀 주 (인) 위 원 FlowTech 연구소 소 장 Leif Broberg (인)

2013년 5 월

조선대학교 대학원

(5)

Cont ent s Li stofTabl es

Li stofFi gur es ABSTRACT

I .I nt r oduct i on

Ⅱ. Desi gn ofHul lf or m andAppendages

Ⅲ.Devel opmentofEner gy Savi ng Devi ce

Ⅳ.Theor et i calCal cul at i on

A.FLOW SOLVERS B. Pot ent i al f l owsol ver 1. Boundar yl ayer met hod 2. RANSsol ver

C. MODELLI NGOFTHEPROPELLER D. val i dat i on

1.Resi st anceCal cul at i on Resul t 2.Sel fPr opul si on Cal cul at i on Resul t

Ⅴ.ModelTest

A.Testf aci l i t y B.Hul lmodel C.Pr opel l ermodel

D. Ext r apol at i on by I TTC78 and modi f i ed I TTC78 met hod

E.Resul tofmodelt est

Ⅵ.Check t heseakeepi ng per f or mance

Ref er ences

(6)

Cont ent s

Li stofTabl es ‥‥‥‥‥‥‥‥‥‥‥‥‥‥Ⅲ Li stofFi gur es ‥‥· ‥‥‥‥‥‥‥‥‥‥‥Ⅳ ABSTRACT ‥‥‥‥‥‥‥‥‥‥‥‥‥‥‥ 1

I .I nt r oduct i on· · ‥‥‥‥‥‥‥‥‥‥‥‥‥‥2

Ⅱ. Desi gn ofHul lf or m andAppendages· ‥3

Ⅲ.Devel opmentofEner gy Savi ng Devi ce‥7

Ⅳ.Theor et i calCal cul at i on‥‥‥‥‥‥‥‥· 11

A.FLOW SOLVERS ‥‥‥‥‥‥‥‥‥‥‥‥12

B.Pot ent i al f l owsol ver ‥‥‥‥‥‥‥‥‥‥‥12

1. Boundar yl ayer met hod · ‥‥‥‥‥‥‥‥‥‥13

2. RANSsol ver ‥‥‥‥‥‥‥‥‥‥‥‥‥‥13

C. MODELLI NGOFTHEPROPELLER

D. val i dat i on ‥‥‥‥‥‥‥‥‥‥‥‥‥‥‥· 13

1.Resi st anceCal cul at i on Resul t ‥‥‥‥‥· · ‥14

2.Sel fPr opul si on Cal cul at i on Resul t ‥‥‥‥‥15

(7)

Ⅴ.ModelTest‥‥‥‥‥‥‥‥‥‥‥‥‥‥41

A.Testf aci l i t y ‥‥‥‥‥‥‥‥‥‥‥‥‥‥41

B.Hul lmodel · ‥‥‥‥‥‥‥‥‥‥‥‥‥‥42

C.Pr opel l ermodel · ‥‥‥‥‥ ‥ ‥ ‥ ‥‥‥‥42

D.Ext r apol at i on by I TTC78

andmodi f i edI TTC78met hod ‥‥‥‥‥· · · · 44

E.Resul tofmodelt est ‥‥‥‥‥‥‥‥· ‥‥45

Ⅵ.Check t heseakeepi ng per f or mance‥· ‥51

Ⅶ.Concl usi ons‥‥‥‥‥‥‥‥‥‥‥· ‥‥72

Ref er ences‥‥‥‥‥‥‥‥‥‥‥‥‥‥· ‥‥· 73

(8)

Li stofTabl es

Table2-1Maindimensionofship···3

Table3-1Theconfigofenergysavingdevice···9

Table4-1Comparisonofresistancecomponents···14

Table4-2Comparisonofpropulsioncoefficient···15 Table5-1Maindimensionofshipandmodel···42

Table5-2Maincharacteristicsmodelpropeller···43

Table5-3Comparisonoftheresultsofself-propulsiontests···46

Table6-1Dimensionofship···53

Table6-2InputofRAO-response···55

(9)

Li stofFi gur es

Fig.2-1 Linesofcarferry···4

Fig.2-2modelingof Hull···4

Fig.2-33D modelofBracketFin1(BF1)···5

Fig.2-43D modelofBracket Fin2(BF2)···5

Fig.2-53D modelofBracketFin1+Longstator(BF1+LS)···6

Fig.2-63D modelofLongstator(LS)···6

Fig.3-1Pre-swirlstator···7

Fig.3-2StrutFin···8

Fig.4-1FourDifferentConfigurationofEnergySavingDevice···12

Fig.4-2ResistancecomponentsvsFn···14

Fig.4-3Fnvspropulsioncoefficients···15

Fig.4-4(a)Comparisonofpotentialflow streamlines (BareHull,sidebow view)···16 Fig.4-4(b)Comparisonofpotentialflow streamlines (BF1,sidebow view)···16

Fig.4-4(c)Comparisonofpotentialflow streamlines (BF2,sidebow view)···16

Fig.4-4(d)Comparisonofpotentialflow streamlines (BF1+LS,sidebow view)···17 Fig.4-4(e)Comparisonofpotentialflow streamlines(LS,sidebow view···17

Fig.4-5(a)ComparisonofStream line(BareHull,bottom sternview)···17

Fig.4-5(b)ComparisonofStream line(BF1,bottom sternview)···18

Fig.4-5(c)ComparisonofStream line(BF2,bottom sternview)···18

Fig.4-5(d)ComparisonofStream line(BF1+LS,bottom sternview)···18

Fig.4-5(e)ComparisonofStream line(LS,bottom sternview)···19

Fig.4-6(a)ComparisonofStream line(BareHull,persp.sternview)···19

Fig.4-6(b)ComparisonofStream line(BF1,persp.sternview)···19 Fig.4-6(c)ComparisonofStream line(BF2,persp.sternview)···20

(10)

Fig.4-6(d)ComparisonofStream line(BF1+LS,persp.sternview)···20

Fig.4-6(e)ComparisonofStream line(LS,persp.sternview)···20

Fig.4-7(a)Sternflow direction(BareHull,sternview)···21

Fig.4-7(b)Sternflow direction(BF1,sternview)···21

Fig.4-7(c)Sternflow direction(BF2,sternview)···21 Fig.4-7(d)Sternflow direction(BF1+LS,sternview)···22

Fig.4-7(e)Sternflow direction(LS,sternview)···22

Fig.4-8(a)Comparisonofthe_ /_ (BareHull)···22

Fig.4-8(b)Comparisonofthe_ /_ (BF1)···23

Fig.4-8(c)Comparisonofthe / (BF2)···23

Fig.4-8(d)Comparisonofthe / (BF1+LS)···23 Fig.4-8(e)Comparisonofthe_ /_ (LS)···24 Fig.4-9(a)Comparisonofthe__ (BareHull)···24

Fig.4-9(b)Comparisonofthe_ (BF1)···24

Fig.4-9(c)Comparisonofthe (BF2)···25

Fig.4-9(d)Comparisonofthe__ (BF1+LS)···25

Fig.4-9(e)Comparisonofthe__ (LS)···25

Fig.4-10(a)Comparisonofthe (BareHull)···26

Fig.4-10(b)Comparisonofthe__ (BF1)···26

Fig.4-10(c)Comparisonofthe__ (BF2)···26 Fig.4-10(d)Comparisonofthe__ (BF1+LS)···27

Fig.4-10(e)Comparisonofthe (LS)···27

Fig.4-11(a)Comparisonofthewakefraction(BareHull)···27

Fig.4-11(b)Comparisonofthewakefraction(BF1)···28

Fig.4-11(c)Comparisonofthewakefraction(BF2)···28 Fig.4-11(d)Comparisonofthewakefraction(BF1+LS)···28

Fig.4-11(e)Comparisonofthewakefraction(LS)···29

Fig.4-12(a)Comparisonofthevtr(BareHull)···29

(11)

Fig.4-12(b)Comparisonofthevtr(BF1)···29

Fig.4-12(c)Comparisonofthevtr(BF2)···30

Fig.4-12(d)Comparisonofthevtr(BF1+LS)···30

Fig.4-12(e)Comparisonofthevtr(LS)···30

Fig.4-13(a)Comparisonofthewake(BareHull)···31

Fig.4-13(b)Comparisonofthewake(BF1)···31

Fig.4-13(c)Comparisonofthewake(BF2)···31

Fig.4-13(d)Comparisonofthewake(BF1+LS)···32

Fig.4-13(e)Comparisonofthewake(LS)···32

Fig.4-14(a)ComparisonofWakein_{  } (BareHull)···32

Fig.4-14(b)ComparisonofWakein_{  } (BF1)···33

Fig.4-14(c)ComparisonofWakein_{  } (BF2)···33

Fig.4-14(d)ComparisonofWakein_{  } (BF1+LS)···33

Fig.4-14(e)ComparisonofWakein_{  } (LS)···34

Fig.4-15(a)ComparisonofWakein_{  } (BareHull)···34

Fig.4-15(b)ComparisonofWakein_{  } (BF1)···34

Fig.4-15(c)ComparisonofWakein_{  } (BF2)···35

Fig.4-15(d)ComparisonofWakein_{  } (BF1+LS)···35

Fig.4-15(e)ComparisonofWakein_{  } (LS)···35

Fig.4-16(a)ComparisonofWakein_{  } (BareHull)···36

Fig.4-16(b)ComparisonofWakein   (BF1)···36

Fig.4-16(c)ComparisonofWakein_{  } (BF2)···36

Fig.4-16(d)ComparisonofWakein_{  } (BF1+LS)···37

Fig.4-16(e)ComparisonofWakein_{  } (LS)···37

Fig.5-1PhotoofTowingTankof HiroshimaUniv.···38

Fig.5-2Hullmodel···39

Fig.5-3Propellermodel···40

Fig.5-4Photographsofmdeltest···43

(12)

Fig.5-5Calibrationofrpsvs_ ···43

Fig.5-6Modelwake(_)vs forwithandwithoutBF1···44

Fig.5-7_ vs_ forWithoutBF1andWithBF1···44 Fig.5-8_ vs_ forWithoutBF1andWithBF1condition···44

Fig.5-91-tvs_ forWithoutBF1andWithBF1condition···45 Fig.5-101-w vs_ forWithoutBF1andWithBF1condition···45

Fig.5-11_ vs_ forWithoutBF1andWithBF1condition···45

Fig.5-12_ vs_ forWithoutBF1andWithBF1condition···46

Fig.5-13_ vs forWithoutBF1andWithBF1condition···46

Fig.5-14_ vs_ forWithoutBF2andWithBF2condition···46

Fig.5-151-tvs_ forWithoutBF2andWithBF2condition···47

Fig.5-161-w vs_ forWithoutBF2andWithBF2condition···47

Fig.6-1CalculationresultofLoadingCondition···50

Fig.6-2Matrixesofship···51

Fig.6-3CalculatedRAO···53 Fig.6-4Seacondition···59 Fig.6-5Motionofship···63

(13)

ABSTRACT

A s t udyont hede ve l opme ntofe ne r gys a vi ngde vi c e f ort het wi ns c r e w c arf e r r y

Lim,YeonJi

Advisor:Prof.LeeKwi-joo,Ph.D.

DepartmentofNanalArchitecture

& OceanEngineering, GraduateSchoolofChosunUniversity 본 논문에서는 스트럿으로 지지된 2축 프로펠러를 가진 57m급 차도선의 연료절감 부가물개발에 관한 결과를 다루었다.본 차도선은 실제 운항 중인 선박이며 이의 선 미부 스트럿에 새로운 Pre-swirl타입의 새로운 연료절감 부가물인 BF1, BF2, BF1+LS,LS의 형상을 좌현과 우현에 대칭으로 부착하여 프로펠러 회전방향으로 인 해 비효율적으로 흐르는 유동을 통제하는 역할을 하게 하였다. 이는 기존의 Pre-swirlStator와 비교했을 때 프로펠러 반경의 반절길이의 수평 Fin과 부착하는 Fin의 수를 줄여 선미 부의 하중을 감소 시킬 것이며 제작비용이 절감되고 유지보수 가 용이할 것으로 예상된다.서로 다른 형태의 네 가지 다른 디자인은 쉽플로우를 이 용한 CFD이론계산에 의해 질적 분석이 수행되었고 그 중 두 가지 경우에 대하여 예 인수조에서의 모형시험이 수행되었다.선미부에 부착 된 Fin은 프로펠러에 흐름을 유 도하고 wake peak를 감소하여 프로펠러면에서의 wake를 균일하게 유도하였다.한 방향으로만 회전하는 프로펠러로 인해 좌현과 우현의 수행이 다르게 되고 비대칭 압 력분포는 선체에 악영향을 미치게 되나 후류 패턴을 바꾸어 프로펠러와 선미사이의 손실 에너지를 감소시키는 역할을 하게 하였다.기존의 stator와 비교하여 짧은 길이 의 Fin으로 기존의 것과 같은 효과를 나타낼 수 있으며 예인수조시험 결과는 계획속 도의 만재상태에서 2%의 동력을 얻을 수 있다는 결론을 얻었다.

(14)

Ⅰ.I nt r oduct i on

Recently the whole world is working forenvironmentalprotection and energy saving. The world has been implemented the Kyoto Protocol 2008 to 2012.

Although Korea excluded from the Kyoto Protocolsince 1990 according to the Economicgrowth increasein greenhousegasemissionsweremorethan twice.As a result,Korea was implemented until 2020 Green Growth law and Korea's greenhousegasreductiontargetis4% ascontrastin2005.

InternationalMaritime Organization were discussed ship fuelregulations on 22 March 2010.According to the Fueleconomy regulations (EEDI,Energy Design Index)a new technology ofvessels willlead to a new eco-friendly technology.

Shipoperationcausesmanykindsofenvironmentalpollution.InternationalMaritime Organization (IMO) has strengthened the regulation according to environmental pollution from vessels. Eco-friendly technology, ship design, various new technologiesforthereductionoffuelshouldbeinvestigatedaggressively.

The more energy itconsumesand energy saving technologies areessentialfor developments ofgreen ships.Solution to this problem is,in the early stage,to design ahullform having notonly theminimum resistancebutalsotheoptimum propulsive efficiency.However,the hullforms ofgeneralships bear long term experienceofoperationswhichisnotexpectedtobeimprovedradically.Thus,itis consideredmoreappropriatetoattachenergysavingdevicesratherthantoimprove thehullform directly.In thisthesis.appendagesattachedashiphasbeenstudied.

Through modeltests in the towing tank,the conditions are soughtin which the appendagescanreducetheshipresistance.Theeffectsoftheappendagesuponthe wake distributions at a propeller plane are also investigated to clarify the effectiveness ofthem on the equalizations ofthe distributions,which are closely related to the ship propulsive efficiencies.It is found that appendage attached vertically totheshipstern can accelerateflow in theregion,resulting equalization of the wake distribution.The results of CFD have been compared with the

(15)

experimentalresults,to validatetheefficiency ofappendage.To apply theresults developed to ship design,effects to the resistance characteristics and propulsive performances of the ship should be carefully investigated. Furthermore, the developed tools from each specific projects should be closely linked,verified and enhanced,iftobeapplicablepractically.

Ⅱ.Desi gn ofHul lf or m andAppendages

ThemainParticulars ofshipwhichhasbeenusedinthisthesisisshowninTable 2-1

Table2-1Maindimensionofship Ship Lengthbetweenperpendiculars 57m Lengthatloadwaterline 60.23m

Breadth 13m

Depth 2.9m

DesignDraft 1.8m

(16)

Fig.2-1 Linesofcarferry

Fig.2-2modelingof Hull

(17)

BF1hasbeenattachedontherightsideoftheY-shapedstrutwitha650mm lengthoftheplate,450*200mm size,witha100°anglefrom thecentralaxisof theshaftofthestatorisattached.3D designareconfirmedinFig.2-3～2-6

Fig.2-33D modelofBracketFin1(BF1)

Fig.2-43D modelofBracket Fin2(BF2)

(18)

Fig.2-53D modelofBracketFin1+Longstator(BF1+LS)

Fig.2-63D modelofLongstator(LS)

(19)

Ⅲ.Devel opme ntofEner gySavi ngDe vi ce

Despite ofcontinuous efforts ofscientistsand engineers in finding the betterhullform,the performanceimprovementbyhullform improvementitselfhasitslimitationsinmanycases.

So,theextensiveactivitiesonthedevelopmentofenergysavingdevicesarestillontheirprogressand utilizedintheactualships(Leeetal.,1992).

Oneofthemostapplicabletoenergysavinginshipswillbepre-swirlstatorforwhichvariationsare showninfollowingFig.3-1

Fig.3-1Pre-swirlstator

Conventionalpre-swirlstatorconsistsofseveralfinswhichhavealmostthesamespanlengthasthe propellerradiusandarefixedradiallyonthesternframeinfrontofthepropeller.

Thesettinganglesoftheconventionalhorizontalfinswerefixedtocontrolmainlytheflow near thefintip.Inthiscase, thefindidnotcontrolproperlytheflow fieldnearthefinrootbythe presenceoflongitudinalvortexduetotheshiphull,ratherdeterioratedthepropulsiveefficiency.By adopting thehorizontalfinsofalmosthalf-length ofpropellerradius,itmay contributetocan controlthe flow field only near the stern frame and then getalmostequaleffectto the conventionalfin.Furthermore,thebentplateontheedgeofstrutgeneratethecounterflow tothe propellerturningdirection. Thatis,asshowninFig.3-2,thisnew pre-swirltypeenergy-saving deviceconsistsoftwofins. Thisdeviceisnamed“StrutFin”,whichalsocanenhancethesafety foradriftobstaclesuchaswoodortimber,andreducemanufacturingcost.

(20)

Fig.3-2StrutFin

Againstthesternflow mechanism,thedesignofthesternfingenerallyplaysarole asflow guidingdevicemakingawakefieldneartheupperpartofthepropellerplane uniform by both deflecting flow toward propellerand reducing the wake peak at propellertopposition.

When thepropellerrotatesin onespecificdirection,itsperformancesonportand starboard side are differentand the pressure centershifts to the side where the propellerbladesmovedownwards.

Thisinturnleadstoaasymmetricpressuredistributiononthehullinfrontofthe propeller,whichinfluencestheboundarylayersonportandstarboardsidedifferently.

The idea in the design ofStrutFin is to reflectthe stern vortices which move upwardsintheregionthepropellerrotatesdownwards.Thisreflecting may reduce thewakenearpropellertipandcontributetotheequalizationofwakefield,thusmay leadtoapositivegaininpropulsionefficiency. Incaseofaclockwiseturningmore separation occurson portsideabovethepropellershaftwherethepropellerblades move upwards.The flow becomes turbulentand is even partly deflected in the opposite direction.From this pointofview developed appendage mightcontribute towards the reflection ofvortex flow which moves upwards in the region where propellerblades move downwards,by which the increase ofpropulsive efficiency mightbeattained.

ThedimensionsoffullscaleenergysavingdevicearepresentedinTable.3-1

(21)

BF1

Length : 0.65m Breadth : 0.2m

The Z-axis of rotation : 42˚

Origin : (1.65, 1.68, 1.5)

<Fin>

Length : 0.45m Breadth : 0.3m

The Z-axis of rotation : 100˚

Origin : (2, 2.25, 0.545)

BF2

Length:0.65m Breadth:0.2m

TheZ-axisofrotation :42˚

Origin:(1.65,1.68,1.5)

TheZ-axisofrotation :-42˚

TheX-axisofrotation :80˚

Origin:(1.68,2.7,1.5)

<Fin>

TheZ-axisofrotation :100˚

Origin:(2,2.25,0.545) Table3-1Theconfigofenergysavingdevice

(22)

BF1+LS

TheZ-axisofrotation:42˚ Origin:(1.65,1.68,1.5)

<Fin>

TheZ-axisofrotation:100˚ Origin:(2,2.25,0.545)

<LS>

(23)

LS

<LS>

Ⅳ.Theor et i calCal cul at i on

ComputationalFluidDynamics(CFD)iswidelyusedintheshipdesignprocess.In particularduring the initialdesign stage CFD has become an importanttool.It enablesthedesignertoevaluatealargernumberofhullalternativesandtherebya betteroptimizedandreliabledesignbeforethefinalvalidation.Itistruethatnotonly fornew buildingsbutalsoforexistingshipsandretrofittingofshipenergysaving devices.Thetoughcompetitionontheshipbuildingmarketcreateshighdemandson shortlead times and competitive designs.This mustbe metby developments of effectiveCFD toolsandintegrationwithCAD(MEPD,2009).

SHIPFLOW has developed steadily since the first version was released by FLOWTECH in 1992..Thisprogram isoptimized forship hydrodynamicsdesign.

GridsfortheRANS solversaswellasmeshesforXPAN arecreatedautomatically from thehullshape.Varioustypesofhullshapescanbehandled,suchasmonohulls, twinskeghulls,multihulls,sailingyachtsetc.InadditiontothisSHIPFLOW usesan efficient overlapping grid technique for use with appendages. Resistance and

(24)

propulsiondataarepresentedinthenavalarchitectswayandthesolversareadapted forhullgeometries.

CFD codemakescostdown fortheevaluation and prediction ofperformanceof ship.Comparative theoretical calculation has been performed by CFD code of Shipflow forfourdifferentdesignsofStrutFinasshowninFig.4-1

DesignBF1 DesignBF2 DesignBF1+LS DesignLS Fig.4-1FourDifferentConfigurationofEnergySavingDevice A.FLOW SOLVERS

Computations are performed with SHIPFLOW developed at FLOWTECH InternationalAB.Therearethreekindsofflow solversin SHIPFLOW.XCHAP is aRANS solverforsteady incompressibleflow,XPAN apotentialflow solverand XBOUND isan integralmethodforthinboundary layers.Thesolverscanbeused separatelyorincombinationdependingontheneeds.

The zonalapproach is an efficienttechnique formany applications.The methods arein thiscaseappliedin asequence.Thefreesurfaceand thedynamictrim are first computed by XPAN,thereafter the boundary layer on the fore body by XBOUND and finally the flow around the stern and in the wake by XCHAP.

Alternatively,the moregeneralglobalapproach can beused wherethecomplete flow domainiscomputedbyXCHAP.

B.Potentialflowsolver

Thepotentialflow methodXPAN isanon-linearRankinesourcepanelmethod[4]. It uses higher order panels and singularity distributions and a non-linear free surfaceboundaryconditions.

The method can handle liftand induced drag by adding dipoles to the lifting surfacesandtrailingwake.A Kuttaconditionisappliedtofindthestrengthofthe

(25)

boundcirculation.

Dynamic sinkage and trim are computed during the iterative procedure for the non-linear free surface boundary condition. During each iteration the ship is repositionedandthepanellizationofthehullandfreesurfaceisregenerated.

1.Boundarylayermethod

XBOUND isafirstorderintegralmethod[5].Itcomputestheboundarylayeralong potentialflow streamlines.The flow can be laminar or turbulent.The method includesamodelforthetransitionfrom laminartoturbulentflow.

2.RANS solver

XCHAP solves the steady incompressible Reynolds Average Navier-Stokes equations.There are two available turbulence models the k-w SST [6]and the ExplicitAlgebraic Stress ModelEASM [7,8].The EASM takes the non-isotropy into account using algebraic expressions for the Reynolds stresses containing non-linear terms in the mean strain and rotation rates.The modelis a good compromise between performance and the ability to predictthe importantvortex flow in the stern wake and is therefore the standard modelin the program.No wallfunctionsareusedandtheequationsareintegrateddowntothewall.

C.MODELLINGOFTHEPROPELLER

Tosimulatetheeffectofthepropeller,body forcesareintroduced in acylindrical component grid in the overlapping grid. When the flow passes through the propelleritslinearandangularmomentum increasesasifithadpassedapropeller ofinfinitenumberofblades.Theforcesvary inspacebutareindependentoftime andthereforeapproximatingapropellerinducedsteadyflow.

D.validation

A validationofthemethodispresentedwithcomparisonsofresistance,openwater testandselfpropulsionsimulationsatmodelscale.

Theresistancewascomputedusing thecoarseapproachinSHIPFLOW.Thewave resistancewascomputedbythepotentialflow moduleXPAN.

The open water characteristics are required in order to make the full scale extrapolation using the modified-ITTC78 procedure.SHIPFLOW can automatically compute an open waterThe speed of the propeller was automatically adjusted during the selfpropulsion simulation such thatthe propellerthrustbalanced the

(26)

resistance of the hull corrected for the towing force.The towing force was computed according to themodified-ITTC78procedureand includethemodel-ship correlationandroughnessallowance.simulationforasequenceofadvanceratios. 1.ResistanceCalculationResult

kandctcomparedtobarehullwerealllargevalueexceptBF1+LS.

ResistancecalculationresultsarecomparedinTable.2andinFig.4-1 Table4-1Comparisonofresistancecomponents

Barehull BF1 BF2 BF1+LS LS k 0.201 0.202 0.222 0.200 0.204 Cw 1.046*10^-3 1.046*10^-3 1.046*10^-3 1.046*10^-3 1.046*10^-3

Ct 5.266*10^-3 5.270*10^-3 5.341*10^-3 5.263*10^-3 5.279*10^-3

Fig.4-2ResistancecomponentsvsFn

(27)

2.SelfPropulsionCalculationResult

1-w compared to barehullwere alllarge value.In case of 1-t,compared to barehullwere allsmallvalue exceptBF2.BF1 BF1 + LS,and LS compared to barehull_ washighly calculated.BF1caseof_ and_,theresultswereboth highcomparedtobarehull._ hadthegreatestresultsinthevalueoftheBF2and

_ hadthegreatestresultsinthevalueoftheBF1.

SelfpropulsioncalculationresultsaresummarizedinTable.4-2andinFig.4-3 Table4-2Comparisonofpropulsioncoefficient

1-w 1-t ηH ηO ηR ηD

Barehull 0.819 0.894 1.093 0.721 1.020 0.804 BF1 0.814 0.905 1.111 0.721 1.026 0.821 BF2 0.808 0.877 1.085 0.721 1.040 0.811 BF1+LS 0.817 0.904 1.107 0.721 1.026 0.818 LS 0.818 0.905 1.106 0.721 1.026 0.818

Fig.4-3Fnvspropulsioncoefficients

(28)

Fig.4-4(a)Comparisonofpotentialflow streamlines(BareHull,sidebow view)

Fig.4-4(b)Comparisonofpotentialflow streamlines(BF1,sidebow view)

Fig.4-4(c)Comparisonofpotentialflow streamlines(BF2,sidebow view)

(29)

Fig.4-4(d)Comparisonofpotentialflow streamlines(BF1+LS,sidebow view)

Fig.4-4(e)Comparisonofpotentialflow streamlines(LS,sidebow view)

Fig.4-5(a)ComparisonofStream line(BareHull,bottom sternview)

(30)

Fig.4-5(b)ComparisonofStream line(BF1,bottom sternview)

Fig.4-5(c)ComparisonofStream line(BF2,bottom sternview)

Fig.4-5(d)ComparisonofStream line(BF1+LS,bottom sternview)

(31)

Fig.4-5(e)ComparisonofStream line(LS,bottom sternview)

Fig.4-6(a)ComparisonofStream line(BareHull,persp.sternview)

Fig.4-6(b)ComparisonofStream line(BF1,persp.sternview)

(32)

Fig.4-6(c)ComparisonofStream line(BF2,persp.sternview)

Fig.4-6(d)ComparisonofStream line(BF1+LS,persp.sternview)

Fig.4-6(e)ComparisonofStream line(LS,persp.sternview)

(33)

Fig.4-7(a)Sternflow direction(BareHull,sternview)

Fig.4-7(b)Sternflow direction(BF1,sternview)

Fig.4-7(c)Sternflow direction(BF2,sternview)

(34)

Fig.4-7(d)Sternflow direction(BF1+LS,sternview)

Fig.4-7(e)Sternflow direction(LS,sternview)

Fig.4-8(a)Comparisonofthe / (BareHull)

(35)

Fig.4-8(b)Comparisonofthe_ /_ (BF1)

Fig.4-8(c)Comparisonofthe / (BF2)

Fig.4-8(d)Comparisonofthe_ /_ (BF1+LS)

(36)

Fig.4-8(e)Comparisonofthe_ /_ (LS)

Fig.4-9(a)Comparisonofthe (BareHull)

Fig.4-9(b)Comparisonofthe__ (BF1)

(37)

Fig.4-9(c)Comparisonofthe__ (BF2)

Fig.4-9(d)Comparisonofthe_ (BF1+LS)

Fig.4-9(e)Comparisonofthe__ (LS)

(38)

Fig.4-10(a)Comparisonofthe__ (BareHull)

Fig.4-10(b)Comparisonofthe (BF1)

Fig.4-10(c)Comparisonofthe__ (BF2)

(39)

Fig.4-10(d)Comparisonofthe__ (BF1+LS)

Fig.4-10(e)Comparisonofthe (LS)

Fig.4-11(a)Comparisonofthewakefraction(BareHull)

(40)

Fig.4-11(b)Comparisonofthewakefraction(BF1)

Fig.4-11(c)Comparisonofthewakefraction(BF2)

Fig.4-11(d)Comparisonofthewakefraction(BF1+LS)

(41)

Fig.4-11(e)Comparisonofthewakefraction(LS)

Fig.4-12(a)Comparisonofthevtr(BareHull)

Fig.4-12(b)Comparisonofthevtr(BF1)

(42)

Fig.4-12(c)Comparisonofthevtr(BF2)

Fig.4-12(d)Comparisonofthevtr(BF1+LS)

Fig.4-12(e)Comparisonofthevtr(LS)

(43)

Fig.4-13(a)Comparisonofthewake(BareHull)

Fig.4-13(b)Comparisonofthewake(BF1)

Fig.4-13(c)Comparisonofthewake(BF2)

(44)

Fig.4-13(d)Comparisonofthewake(BF1+LS)

Fig.4-13(e)Comparisonofthewake(LS)

Fig.4-14(a)ComparisonofWakein   (BareHull)

(45)

Fig.4-14(b)ComparisonofWakein_{  } (BF1)

Fig.4-14(c)ComparisonofWakein_{  } (BF2)

Fig.4-14(d)ComparisonofWakein   (BF1+LS)

(46)

Fig.4-14(e)ComparisonofWakein_{  } (LS)

Fig.4-15(a)ComparisonofWakein_{  } (BareHull)

Fig.4-15(b)ComparisonofWakein_{  } (BF1)

(47)

Fig.4-15(c)ComparisonofWakein_{  } (BF2)

Fig.4-15(d)ComparisonofWakein_{  } (BF1+LS)

Fig.4-15(e)ComparisonofWakein_{  } (LS)

(48)

Fig.4-16(a)ComparisonofWakein_{  } (BareHull)

Fig.4-16(b)ComparisonofWakein_{  } (BF1)

Fig.4-16(c)ComparisonofWakein_{  } (BF2)

(49)

Fig.4-16(d)ComparisonofWakein_{  } (BF1+LS)

Fig.4-16(e)ComparisonofWakein_{  } (LS)

(50)

Ⅴ.ModelTest

A.Testfacility

-ModeltestsarecarriedoutinthetowingtankofHiroshimaUniv.

Thetowingtankhasthefollowingmainparticulars:

Basin LⅩBⅩD 100×8(partly10)×3.5m Carriage Speed 3m/s

Fig.5-1PhotoofTowingTankof HiroshimaUniv.

(51)

B.Hullmodel

ThedimensionsoffullscaleshipandmodelarepresentedinTable.5-1.

Table5-1Maindimensionofshipandmodel scale 17

ship model Loa 69 4.059 Lpp 57 3.353 Bmax 15 0.882 Bmould 13 0.765 Depth 2.9 0.171 draft 1.8 0.106

Fig.5-2Hullmodel

C.Propellermodel

-Manufacturedpropellermodelsareoutwardturningtwinscrew,fourbladed propellerwithcharacteristicsasshowninTable.5-2.

(52)

Table5-2Maincharacteristicsmodelpropeller Characteristic Value Diametermodelscale 0.1m

Diameterfullscale 1.7m PitchratioP/D atr/R =0.75 0.7

BladearearatioAD/A0 0.7

Fig.5-3Propellermodel

(53)

D.Extrapolation by ITTC78andmodifiedITTC78method

In case ofwithoutthe complex appendage mounted the predictions were made accordingtothe1978ITTC extrapolationmethod.

With the complex appendage mounted a somewhatmodified wake scaling was used.Themethodhasbeendiscussedatthemeeting in1999ITTC andtentatively acceptedforthecaseofevaluation ofpre-swirlstatorconcepts.Thewakescaling presumes thattests with the same propeller butwithoutthe stator have been performed as well.The difference between modeleffective wake with statorand the modelwake withoutstatoris considered as a potentialwake created by the stator.

Thehullpotentialwakeand thefrictionalwakearescaled asforthemodelwith -outstatoraccording totheITTC - 78method,towhichthestatorpotentialpart isadded.Theamount0.04representsthepotentialwakecreated by therudderat thelocationofthepropeller.

ThusisinthemodifiedITTC 1978extrapolationthefullscalewake.

_{ } _   _{ } _ 

  _{ }∆_{ }  _{ }  _{ }_{ } where,

"W"standsfor"withcomplexappendage"

"wo"standsfor"withoutcomplexappendage"

"m"standsfor"model"

"s"standsfor"shipscale"

"T"standsfor"thrustidentity"

Theform factorisbasedonthecasewithoutcomplexappendage.

(54)

E.Resultofmodeltest

Modeltesthasbeenperformedinthetowingtankwitha3.35m lengthmodelwhich scaleis1/17.

The fullscale wake is calculated from the modelwake_ ,and the thrust deductiont:

_    _    × _



__

(1)

Intheformula,thefactor0.04isusedtotakeaccountforruddereffect.Iffullscale wake

_ isgreaterthanmodelwake_ ,followingformulaisused.

_ =_ (2)

Stator angle of attack and position had been decided based upon the CFD calculation results inflow angle and direction of flow at stator position.Self propulsion testsresultsareanalysisby modified ITTC 78prediction method.The results ofthe propulsion testare summarized in Table.5-3 forthe speed of9.5 Knots.AndthephotographsofmodeltestarepresentedinFig.5-4.

(55)

< 9Knots>

< 10Knots>

Fig.5-4Photographsofmdeltest

Table5-3 Comparisonoftheresultsofself-propulsiontests Barehull BF1 BF2 Trustdeductionfraction(t) 0.141 0.141 0.179

Shipwakefraction(w) 0.176 0.185 0.228 Hullefficiency(ηH) 1.024 1.036 1.064

Relativerotative

efficiency(ηR) 0.983 0.983 0.963

_ 0.550 0.560 0.513

Fig.5-5Calibrationofrpsvs_

Fn

(56)

- 44 -

Fig.5-6Modelwake(_)vs forwithandwithoutBF1

Fig.5-7_ vs_ forWithoutBF1andWithBF1condition

Fig.5-8_ vs_ forWithoutBF1andWithBF1condition

Fn Fn

(57)

Fig.5-91-tvs_ forWithoutBF1andWithBF1condition

Fig.5-101-w vs_ forWithoutBF1andWithBF1condition

Fig.5-11_ vs_ forWithoutBF1andWithBF1condition

Fn

Fn Fn

(58)

- 46 -

Fig.5-12_ vs_ forWithoutBF1andWithBF1condition

Fig.5-13_ vs_ forWithoutBF1andWithBF1condition

Fig.5-14_ vs forWithoutBF2andWithBF2condition

Fn Fn

(59)

Fig.5-151-tvs_ forWithoutBF2andWithBF2condition

Fig.5-161-w vs_ forWithoutBF2andWithBF2condition

The results show the reduction of trust deduction coefficient(t) and wake fraction(w) for the case of appended hullby which hullefficiency(_)increased slightly.Therelativerotativeefficiency(_)decreased.Itseemsthatthislow _ is theeffectofrelativelyworseirregularwakefieldbytherotationofpropeller.

Fn

(60)

Ⅵ.Check t heseakeepi ng per f or mance

Purpose ofthis study is predicting the reponses ofa ship in a particularsea state.'Responses'being themotionsoftheship:roll,pitch,yaw,surge,sway and heave;as wellas the consequences ofthese motions such as bow slamming or Green water on deck,propeller emergence,crew and passengers suffering from motion sickness, loss of cargo, etc. The prediction of these responses is indispensible to determine how good a ship is with respectto sea-keeping.Poor performance in sea-keeping means differentthings fordifferentvessels.Reduced fuelconsumption andcargocapacity mean nothing oftheshipcannotperform the taskitwasintendedtointheoceansitwasbuiltfor.

Objectives ofthis study are calculate the response amplitude operators ofa ship for different loading conditions, configure the derived responses of a ship (slamming,propelleremergence,etc.),obtain awavescatterdiagram representative forthearea/voyagesharea ship operatesand predicttheresponsesofa ship in severalseastatesandunderstandwhattheymeantortheship'soperability.

The analysis sequence of used software,Octopus is shown in following Figure.The obtained design values may serve as the criteria which should not be exceeded during the transport or operation.The calculated models and design values can beused in OctopusOnboard giveonboard operational support using the same methods and results as used in design value calculation procedure.

(61)

This (statistical)data is specifically aimed atdescribing thewave environment. Eitherby means ofa voyage,which can be matched in space and time with a wave climate database,orby a scatterdiagram,orsimply a setofdesign sea states.

A wave scatter diagram by calculation shows the probability of a wave combinationofHsandTz.

Input of Loading Condition is as shown in Table and calculation result of LoadingConditionisasshowninfollowingFig.6-1.

(62)

Table6-1Dimensionofship

Name Value

Mass[T] 1142.28

Draft Aft[m] 1.80

Draft Fwd[m] 1.80

LCG[m] 29.96

ZCG[m] 0.87

[m] 4.55

[m] 15.06

[m] 15.40

GM[m] 9.78

GG’[m] 0.0

Fig.6-1CalculationresultofLoadingCondition

(63)

External condition can be defined in matrixes(6x6) for mass, added mass, damping and restoring coeffients.Added mass,damping and restoring coeffients matrixescan bedefinedforeachcombination ofspeedandencounterfrequency.If anexternalconditionisnotdefinedrelativetothecoordinatedsystem with0point at(APP,CL,BL)thanisnotnessecarytorecalculateallmatrixes.Thetranslationto (APP,CL,BL)canbegivenatthedefinitionofthecoordinatesystem.

Fig.6-2Matrixesofship

A hydrodynamicanalysisstartswith thecalculation ofahydrodynamicdatabase (HDB).Thehydrodynamicdatabasedoesnotdependonparameterslikeshipmass, viscous damping or spring restoring parameters.These become importantwhen RAO'sarecalculated.

RAO isan abreviation of'ResponseAmplitudeOperators'.TheRAO deliversthe

(64)

Name Value S e c t i o n a l

Loads

Plane XZ-Plane

2.337 0

Point response ( x , y , z )

coordinates[m] 50 0 25

C o m b i n e d response

Unit n/hr

Response Factor Derive

1 None

responsemotionperunitwaveheight.Calculating aRAO requiresahydrodynamic database. In the Hydrodynamic database draft, speed, heading, and frequency dependentparameters were stored.The RAO requires thatrestoring parameters, viscous damping,and a load distribution are known.Basic responses are always required.The basic responses are motions ofthe motion reference point(surge, sway,heave,roll,pitch,yaw)

InputofRAO-responseisasshowninfollowingTable6-2.

Table6-2InputofRAO-response

Shortterm statisticsiscalculatedforthereferenceperiodgiveearlierwhensetting up the configurations and reponses.The values displayed in the plots willbe differentifanotherreference timeis selected.MSIrefers to theMotion Sickness Studiesdoneby O'Hanlan andMcCauly taken intoISO2631.Slamming criteriaare defined according to the theory by Ochi.Note that when using the slamming operators, Slamming should also be inserted as a response in the response configurationsincluding therelativemotion at0.90L.Levelisused asamaximum value (criterion)forthe diagram.By defaultlevelhas the value ofcriterion as configuredinStatisticsConfiguration.

ThecalculatedRAO ispresentedinagraphicalwayasshowninfollowngfigures.

(65)

(66)

(67)

(68)

(69)

(70)

Fig.6-3CalculatedRAO

(71)

SEAWAY is a strip theory based program to calculate the hydromechanical parameters, wave forces, Respone Amplitude Operators, Statistics, and added resistanceduetowaves.Thesehadtobeinvestigatedearlierbydoingtanktests. Statisticsenableaseastatetobeproperlymodelledin asummation ofaselection ofrandom waves(frequency,height,etc).Via known RAO onecan calculatethe statistics of the reponse of the vessel, motion point, etc... Before statistical calculationscan beperformed,seastateshould bemodelled underCommon in the projecttree.Seastatescan bemodelledin aspectrum,scatterdiagram,orcan be importedform a3rdparty.

Theresultsarevisiblein anumericaland graphicalway,seetabs:General,Sea Condition,ShortTerm andLongTerm asshowninFig.6-4.

(72)

(73)

(74)

Fig.6-4Seacondition

(75)

(76)

(77)

(78)

(79)

(80)

Fig.6-5Motionofship

(81)

Ⅶ.Concl usi ons

Thefunction ofstern fin appended atthestern ofship isgenerally considered to equalizewakefield atthepropellerplaneby both directing flow to propellerand reducing the wake peak at top position. When the propeller rotates in one direction,itsperformancesofportand starboard sidesaredifferentand thecenter ofpressureshiftstothesidewherethepropellerbladesdownwards.Thisin turn leads to a non-symmetric pressure distribution on the hullin frontofpropeller, whichinfluencestheboundarylayersofportandstarboarddifferently.

The strutfin,which is designed to guide the flow forthe equalization ofwake field,showed1.7% ofDHP reductionatdesignspeed.

(82)

Ref er ences

1.L Broberg and M.Orych,'An EfficientNumericalTechnique to Simulate The PropellerHullInteraction',FLOWTECH InternationalAB,Sweden,2012

2.MEPC,'Interim GuidelinesonthemethodofcalculationoftheEEDIfornew ship', MEPC.1/Circ.681,2009.

3.Kwi-Joo Lee,Jung-Sun An and Sun-Hee Yang,‘A study on the Developmentof Energy-Saving Device“CrownDuct"‘,Dept.ofNavalArchitecture& OceanEngineering, ChosunUniversity,Gwangju,KoreaandBasicPlanningDepartment,SPPShipbuildingCo.,Ltd, Busan,Korea,2012

4.JANSON,C.-E.,'PotentialFlow PanelMethodfortheCalculationofFree-Surface FlowswithLift',Ph.D.Thesis,ChalmersUniversityofTechnology,1997.

5.BROBERG.L.,et.al.,'SHIPFLOW Users Manual',FLOWTECH International AB,Sweden,2011

6.MENTER,F.R.,'ZonalTwoEquation k-w TurbulenceModelsforAerodynamic Flows',24thFluidDynamicsConference,Orlando,AIAA paper-93-2906,1993

7. GATSKI, T., SPEZIALE, C.G., 'On Explicit Algebraic Stress Models for ComplexTurbulentFlows',JournalofFluidMechanics,254,59-78,1993

8.DENG,G.B.,Queutey,P.,Visonneau,M.,'ThreeDimensionalFlow Computation with Reynolds Stress and Algebraic Stress Models', Engineering Turbulence Modelling and Experiments 6,Rodi,W.,Mulas,M.,Editors,ELSEVIER,389-398, 2005