CFD验证KVLCC 2船型在长短波下的附加阻力

内容需要下载文档才能查看 内容需要下载文档才能查看

OceanEngineering59(2013)240–273

ContentslistsavailableatSciVerseScienceDirect

OceanEngineering

journalhomepage:http://wendang.chazidian.com/locate/oceaneng

CFDveri?cationandvalidationofaddedresistanceandmotionsofKVLCC2with?xedandfreesurgeinshortandlongheadwaves

HamidSadat-Hosseinia,Ping-ChenWub,PabloM.Carricaa,HoKimb,YasuyukiTodab,FrederickSterna,n

ab

IIHR—Hydroscience&Engineering,TheUniversityofIowa,IowaCity,IA52242,USADepartmentofNavalArchitecture&OceanEngineering,OsakaUniversity,Osaka,Japan

articleinfo

Articlehistory:

Received19March2012Accepted2December2012

Availableonline15January2013Keywords:CFDEFD

Potential?owAddedresistanceHeadwaves

abstract

ThemotionsandaddedresistanceofKVLCC2atFr¼0.142and0.25withfreeand?xedsurgeinshortandlongheadwavesarepredictedusingURANSandvalidatedagainstEFDdataforFr¼0.142.Veri?cationstudiesshowtheresultsarefairlyinsensitivetothegridsizeandtimestep.CFDindicatesnosigni?cantdifferencebetweenfreeand?xedsurgewhileEFDpitchmotionandaddedresistanceareaffected.Addedresistancewaslargestwhenthebowrelativemotionhaslargestamplitudeandisabout1801outofphasewiththewaves.Thedecompositionofforcesandmomentsworkswellfor?rstharmonicsbutnotforhigherharmonicsoriginatedfromradiationinlongwavesanddiffractioninveryshortwaves.Maximumresponsesoccurneartheresonanceconditionandnearthemaximumwaveexcitationforcewhichisatl/L¼1.33andlongwavesforsurge/pitchandheave,respectively.Potential?owpredictionsformotionsandtheaddedresistancearefurtherfromthedatathanCFD.Local?owanalysesshowthataddedresistanceismainlyinducedbyhighpressureontheupperbowwhichiscorrelatedwithbowrelativemotion.Theunsteadywavepatternisanalyzedandthewake?owiscomparedwithPIVmeasurements.

&2012ElsevierLtd.Allrightsreserved.

1.Introduction

Thepredictionofaddedresistanceofashipinwavesisessentialtoevaluatetheshipperformanceinseaway.Also,itisofinterestasthemainfactorforthefuelconsumptioninseawaytomeettherequirementforminimumenergyef?ciencylevelmeasuredbyanEnergyEf?ciencyDesignIndex(EEDI)andEnergyEf?ciencyOperationalIndicator(EEOI)whichareregulatedbytheInternationalMaritimeOrganization’s(IMO)MarineEnvironmentProtectionCommittee(MEPC).Manyresearchershavedevelopedapproachestopredicttheaddedresistanceusingexperimental?uiddynamics(EFD)andpotential?ow(PF)andlatelycomputa-tional?uiddynamics(CFD).Sincetheaddedresistancehasasecondordernatureanditisbasedonmeanvalueofwaveforce,itsvalueisrelativelysmallcomparedtotheamplitudeoftheexcitationforce(Faltinsen,1990).Thusahighdegreeofaccuracyisneededbothintheexperimentsandcalculations.

EFDstudiesofaddedresistancehaveshowndependencyonshipmotions,speed,wavelength,waveheight,waveheading,hullform,bowshapeandbowrelativemotion.Theaddedresistancereachesthepeakwhentheheaveandpitchresponsesarerelativelylarge(FangandChen,2006;Dallingaetal.,2008)i.e.

n

Correspondingauthor.Tel.:þ13193355215;fax:þ13193355238.E-mailaddress:frederick-stern@uiowa.edu(F.Stern).

thecorrespondingwavefrequencyisneartheresonancecondi-tionandthewaveexcitationforces/momentsarelargewhichhappensatl¼1.33Lforpitchandverylongwavesforheaveandisfairlyindependentofshipspeedandshipgeometry(seeFig.1).Surgemotionhassmallin?uenceontheaddedresistancecoef?-cient;however,itspredictionisstillofimportanceasthemotionsmightbeslightlyin?uencedbysurge,heaveandpitchcoupling(Joncquezetal.,2008;Kashiwagi,1995,2009).Themagnitudeoftheaddedresistanceincreasesastheshipspeedincreasesandthenreducesforthehighspeeds(ZeraatgarandAbed,2006;GhaniandJulait,2008).Theaddedresistancehasquadraticdependenceonwaveamplitude(Journee,1976)andisoftenlargerinheadwavesthanthatinbeamwaves(ZeraatgarandAbed,2006;Dallingaetal.,2008).Thehullformwithlargerblockcoef?cienthaslargeraddedresistancewithapeakshiftedtolongerwavesaslargeblockcoef?cientreducesheaveandpitchnaturalfrequencies(seeTable1).Thebluntbowshapegenerallyprovideslargeraddedresistance(Blok,1983;Naito,2001;OriharaandMiyata,2003).Thebowrelativemotionhascorrelationwiththeaddedresistancesuchthatthepeakoftheaddedresistanceisnearthemaximumbowrelativemotion(Blok,1983;Grigoropoulosetal.,2000;Kashiwagietal.,2004).Kashiwagietal.(2004)alsoshowedthatthepeakvalueofnon-dimensionalbowrelativemotionreducesbyincreasingwaveheight.

Manyeffortshavefocusedonusingbothlinearandnonlinearpotential?ow(PF)foraddedresistance.Forlinearpotential?ow,

0029-8018/$-seefrontmatter&2012ElsevierLtd.Allrightsreserved.http://wendang.chazidian.com/10.1016/j.oceaneng.2012.12.016

H.Sadat-Hosseinietal./OceanEngineering59(2013)240–273241

Fig.1.Reviewofaddedresistanceandsurge,heaveandpitchwaveexcitation/totalforcesandmomentsfor?xed/freeshipinheadwavesfordifferentgeometries(Guietal.,2002;Journee,1992).

theaddedresistanceforceisestimatedfromvelocitypotentialand?uidpressuresolutioncorrectedwiththeperturbationmethodtoincludethehigherordertermsusingpressureintegra-tionmethod(Boese,1970;Salvesen,1978;Faltinsenetal.,1980),momentumandenergymethod(Maruo,1957,1963),andradiatedenergymethod(GerritsmaandBeukelman,1972).Thepressureintegrationmethodisanear-?eldmethodwhichcom-putestheaddedresistancefromhydrodynamicpressureintegra-tiononthebodysurfaceusingBernoulli’sequation,andaTaylorexpansionofthepressureaboutthemeanpositionoftheship.Theothertwoapproachesarefar-?eldmethodcomputingtheaddedresistancefromthegeneratedwavesenergyandmomen-tum?uxatin?nity.Thedetailsofthesemethodsandtheirimplementationsinthelinearpotential?owsolversarediscussedinFaltinsen(1990).

Thelinearpotential?owseakeepingsolversbasedonstriptheory,sourcedistributionmethod,panelmethodandmorerecentlyenhanceduni?edtheory(EUT)wereemployedwiththeabove-mentionedmethods.Journee(2001)appliedlinearstriptheorybasedsolvercalledSEAWAYwithGerritsmaandBeukel-man’sradiatedenergymethodandshowedtheaddedresistanceisunderpredictedforshortwavesasthe3Dbowwavediffractionisdominantandnottakenintoaccountforthestriptheory.FangandChen(2006)employed3-DsourcedistributionmethodwithSalvesen’smethod(Salvesen,1978)andLiuetal.(2010)usedfrequencydomain3DpanelmethodwithMaruo’smethod(Maruo,1957,1963)andbothshowedthat3Dmethodsprovidebetterpredictionofaddedresistancethan2D’s,especiallyinshortwaves.Bunniketal.(2010)reviewedtheaddedresistancepredictionfordifferentlinearsolversbasedonstriptheory,sourcedistributionmethod,panelmethodandconcludedthattheaddedresistanceisunderpredictedbymostofthesolversandheavemotionismoredif?culttopredictaccuratelythanthepitchmotions.Kashiwagi(1995,2009,2011)computedtheaddedresistancebyamodi?edversionofMaruo’sformulausingEUTtoconsidertheeffectof3Ddiffractedwaveatthebow.TheresultsshowedgoodagreementwithEFDdatabutrelativelylargeerrorsfortheshipwithlargeforwardspeedasthenonlinearitiesarenottakenaccountinEUTandMaruo’sformula.

Basedonpanelmethod,moreadvancedPFpredictionofaddedresistancehasbeendeveloped.The?rstorderandhigherordersolutionsarenotseparatedandtheyaresolvedthroughasinglesolutionprocess.Joncquezetal.(2008)usedAEGIRandcomparedtheaddedresistancefrompressureintegrationandmomentumconservationmethods.Itwasfoundthatbothmethodspredicttheaddedresistanceformostofthegeometriesexceptmomen-tumconservationmethodunderestimatestheaddedresistanceforthebulkcarrier.Zhangetal.(2009)http://wendang.chazidian.comMP-1isalinearpotentialcode,LAMP-2isapproximatenonlinearcodeforwhichthehydrostaticandFroude–Krylovforcesareappliedontheinstanta-neouswettedsurfaceofthebody,andLAMP-4includesthehighestnonlinearitylevelwiththeboundaryconditionsappliedontheinstantaneouspositionoffreesurfaceandbody.BothLAMP-1andLAMP-2showedverysimilarresultsforverticalmotionsandaddedresistanceandthepeakoftheaddedresistanceunderpredictedforS60andoverpredictedforWigleyhullforbothsolvers.Kimetal.(2010)appliedWISHandAEGIRtocomputetheaddedresistanceandshowedthatamongthecomponentsofaddedresistance,radiationismuchlargerthandiffractionaroundl/L$1butdiffractionwasslightlylargerinshorterandlonger

waves.

242H.Sadat-Hosseinietal./OceanEngineering59(2013)240–273

Table1

Addedresistanceandshipmotionsfordifferentgeometries.

CB

Fr

RAOmaxRAO

SR221C(Kashiwagi,2009)

0.803

0.15

z/Ay/AKCawz/A

Atl/L1.12831.06926.92442.16671.31349.61362.64441.70152.73331.60071.10931.0528.71721.4695–

8.08891.5488–

4.901910.32301.31751.206811.136712.27068.28035.73477.598910.0698–

15.253214.85238.53218.95069.39308.17049.105410.5477.90688.009215.647

fe(Hz)1.31.61.21.2011.3971.201a1.61.7951.8062.0121.331.51.151.33–1.331.5–1.51.11.24711.76461.21.21.1511.11.25–111.18*1.21.241.241.31.24*1.311.371.25*

fn_heave0.53860.47360.56631.33671.19291.33671.24841.14291.28781.18270.810.750.90.89–0.890.89–1.081.90321.80961.42871.85841.92371.08140.13000.12230.1129–

0.16200.16200.16420.17610.16250.16250.15800.15880.15610.15170.1580

fn(Hz)(ave.fe)fn_pitch0.5386

0.4736

DelftCatamaran(Castiglioneetal.,2009)0.20580.451.29101.1728

y/AK

0.60.75

KCS(Simonsenetal.,2008)

0.651

0.26

Cawz/Ay/AKz/Ay/AKz/Ay/AKCawz/Ay/AKCawz/Ay/AKCawCawz/Ay/AKCawCawCawCawCawCawCawCawCawCawCawCawCawCawCawCawCawCaw

0.86330.75

0.33

0.4

‘SR108(OriharaandMiyata,2003)0.5721

0.250.275

1.80961.4287

0.3

Bluntmodi?edWigley(Kashiwagi,2011)S175(Kimetal.,2010)

0.63440.561

0.20.150.20.25

0.15(towed)0.3(towed)0.3(propelled)0.250.2660.2830.2070.2220.250.1470.1650.25

––––––––––––––––

––––––––––––––––

FFG(McTaggart,1997)0.45

S60(McTaggart,1997)(ZakariaandBaree,2008)*0.6

0.7

0.8

a

OnlyCFDdataofFr¼0.45available;othersarebasedonEFDdata.

CFDhastheadvantageofpredictingaddedresistanceandshipmotionsandperformingnonlinearcomputationwithoutusingananalyticalformulaforaddedresistanceorempiricalvaluesforviscouseffect,butitisconsiderablymoreexpensivethanPFapproaches.OriharaandMiyata(2003)validatedtheaddedresis-tanceandmotionsforSR108containerandshowedalongandprotrudingbowreducestheaddedresistance.Simonsenetal.(2008)conductedsimulationsforKCScontainershowinggoodpredictionforthemotionsbutmeanand1stharmonicamplitudeofresistancewaspredictedby24%Dand80.6%D,respectively.Castiglioneetal.(2009)studiedtheaddedresistanceandresponseofhighspeedDelftcatamaraninheadwavesforseveralshipspeedsandshowedthereisapeakformotionsattheresonantfrequencyforallspeeds.Thepeakincreaseswithspeed,reachingtheirmaximumatthehighestspeed.TheaddedresistanceofKVLCC2inheadwavesisstudiedasoneoftheGothenburg2010workshoptestcasesbyDengetal.(2010),Moctaretal.(2010)andSadat-Hosseinietal.(2010a).Dengetal.(2010)predictedtheaddedresistanceandmotionsforl/L¼1.1and1.6byaverageofE¼17.7%DusingISISCFDRANSsolver.Theveri?cationstudyandvalidationformorewavelengthconditionswerelaterreportedinGuoetal.(2012)whichshowedgoodpredictionofthemotionsandtheaddedresistanceforallofthewavelengthconditionsstudiedthere.Moctaretal.(2010)validatedonlythemotionsforl/L¼0.6,1.1and1.6predictedbyOpenFOAMandCometRANScodes.ThesimulationswithfreesurgewerealsoconductedusingComet.Sadat-Hosseinietal.(2010a)predictedtheaddedresistanceandmotionsforKVLCC2atFr¼0.142usingCFDShip-IowaV4.5,anUnsteadyReynoldsAverageNavier–Stokes(URANS)code,forlongwaveregionforboth?xandfreesurgeandthepreliminaryresultswerecomparedagainstavailableEFDdata.

TheobjectiveofthispaperistovalidatethemotionsandaddedresistanceofKVLCC2tankeradvancingatFr¼0.142with?xedandfreesurgeinheadwavesforawiderangeofwavelengthconditionincludingveryshortwaves.Theveri?cation,naturalheaveandpitchfrequencies,effectsofhighershipspeedandtheconditionsformaximumaddedresistanceandshipmotionsareinvestigated.TheCFDresultsarecomparedwiththepotential?owpredictionsandthedecompositionoftotalforcesintoFroude–Krylov,diffractionandradiationisevaluated.

H.Sadat-Hosseinietal./OceanEngineering59(2013)240–273243

Lastly,thelocal?owisanalyzedandthewake?eldatpropellerplaneisvalidatedagainstPIVmeasurements.

3.Experimentalmethods

TheEFDdataareprovidedfromOUandINSEANforlongwavesandNTNUforshortwaves.TheEFDdataforlongwavesarewithfreeand?xedsurgeconditionsandusedinGothenburg2010workshop(Larssonetal.,2010)asbenchmarktovalidatethesimulations.TheEFDdataforshortwavesisforthemodelwith?xedsurge.3.1.Freesurge

ThefreesurgetestsareconductedinOsakaUniversitytowingtank.Thetankis100mlong,7.8mwideand4.35mdeep.Itisequippedwithdrivecarriage(7.4minlength,7.8minwidth,and6.4mheight)runningfrom0.01to3.5m/s.Itisalsoequippedwithplunger-typewavemakergeneratingregularandirregularwavesupto500mmheightandwavelengthof0.5–15m.Thewaveabsorberisasmall?xedgridironbeachatthebasin’send,withmovablebeachesalongitssides.Themodelistowedwithlightweightcarriageconnectedtomaincarriagebymeanofaspringtoallowthemodeltobefreeinsurgemotionwhileitisfreetoheaveandpitch,asshowninFig.3.Notethatthetowingmethodofthemodelmightbeimportanttostudynotonlytheaddedresistancebutalsothespeedloss(MinsaasandSverre,

2.KVLCC2geometry

TheexperimentsandsimulationsareconductedforbarehullKVLCC2appendedwithpropellershaft.ThemainparticularsoftheEFDandCFDmodelsareshowninTable2.TheexperimentsareperformedbyOU(OsakaUniversity),INSEAN(ItalianShipResearchInstitute)andNTNU(NorwegianUniversityofScienceandTechnology).OUandINSEANuseamodelmanufacturedofwoodwithscaleratioof1/100.TheexperimentconductedbyNTNUuseslargermodelwith1/58scaleratio.ForCFDsimula-tions,themodelwithscaleratioof1/320isemployed.TheCFDmodelincludesthepropellershaftsimilartoEFDmodels.ThebodyplanofthemodelisshowninFig.2.Thecoordinatesystemislocatedatcenterofgravity,withxpointingtowardtheaft(positivedownstream),ytostarboardandzupward.

Table2

Shipmodelproperties.

Realship

EFDmodelOU

ScaleLpp(m)Lwl(m)Bwl(m)Depth(m)Draft(m)Displ.(m3)KG(m)CBCMCW

LCB(%),fwdþKyy/LppKxx/B

1

320.0325.558.030.020.8312,62218.6

1003.23.2550.580.300.2080.3126210.186

INSEAN1003.23.2550.580.300.2080.3126210.1860.80980.99800.90003.480.250.4

NTNU58

5.51725.61211.00.51720.35861.60230.3207

3201

1.01720.18130.09380.0650.009540.0581CFDmodel

Fig.3.Testsetupforfreesurgecondition.

Fig.2.KVLCC2bodyplanandhullfrom.

244H.Sadat-Hosseinietal./OceanEngineering59(2013)240