Dynamic Loadings Due to Waves and Ship Motions

1 ,,*3** ,HonoraryMember,WebbInstituteofNavalArch itecture,GlenCove, ,Member,SUNYM aritimeCollege,FortSchuyler,NYCCoDy,iKht 1975byThesocietyofNavalA,ch tectsandMa,, ,En@eecsABSTRACTA fterconsiderationofthewaysthataship shullcanfail,areviewisfirstgivenofthemaj orlongitudinalbendingloads--withparticul aratten-tiontothepresentstatusofknowledg eofwavebendingmomentsCorrelationsaregive nbetweenpredictedtrends,us-ingboththeory andmodeltests,andfull-scaleobservationsV ibratoryresponses,bothtransient(slamming andwhipping)andcyclic(springing),aredisc ussed,aswellasloadingsresultingfromunevs nthermalgradientsFinally, ibratorywaveloads,tech-niquesforthetheor eticalcalculationofwane-inducedshearsand bendingmo-ments,collectionoffull-scalesh ipstressdata,andthecollectionandana-lysi sofoceanwaverecordsThesestudieshavebeenm adenecessarybythedrasticchangesinmerchan tshipcharac-teristics,particularlythelar gersizeofbulkcarriersa dthehigherspeedsofgeneralcargovessels, ailabletodayasitcanbeaD-plied-tonewshipd es~gnsthatareno;orsoonwillbeonthedra sdesignproblems.

2 ,:portedin1954(1)AmodelofaT-2tanker,(tth eDavidsonLaboratoryandr<jointedamidships ,wass ~tmeasured.(Firstmode-vibrationofthejoin tedhullwasalsoidentifiedandrecorded)Sinc ethisexperimentalworkprecededanyknownana lyticaltreatmentofthesubject,itwaswithso mesurprisechattheexperi-mentersnotedared uctioninbendingmo-mentfromthevaluescalcu latedbycon-ventionalquasi-staticmethods, asshowninFigure1.(Thisreductionwaslaterf oundtohavebeenexaggeratedatcertainspeeds becauseofdynamiceffectsinthemomentmeasur ements(2))Theanalyticaltreatmentofshipmo tionsandwaveloadsbyKorvin-Kroukovsky(3) (D Alembert)forcesovertheshiplengthasillust ratedbyFigure2,re-producedfromreference( 4)Theworkexplainedthereductionindynamicw avebendingmomentsonthebasisoftwofac-tors :thewell-known Smitheffect ,whichaccountsforthepressurereductionina wavecrestandincreaseinatroughresultingfr omtheorbitalmotionofwaveparticles.)

3 Andasecondeffectofcompar-ablemagn~tudere sultingfromship-waveinteraction(dampinga ndaddedmass)Furtherresearchhasestablishe dthatpitchingmotion,perse, ,asshownforexampleinthephotographpublish edasFigure25ofref-erence(1)Herethemodeli sshowninsaggingconditionwiththeloadwater linecompletelyoutofthewaterovertheen-tir elengthoftheship.(Staticbuoyancyf---M-1 305~;25 ~ +sforABSrz20(LongTank) @Qr*Tam x)15%\ *-Deriv9dfrom1954:+ +si ~ -Modd14+4-1955 Rqul(arWava +S~ ,!Spa-d, , { .:::# :12L !,, ,.,.,-------,, ,,.,,,,,,. -- , , ., .,,,,, ,,.,,,6,,,,li., , ,,, 0 ,.,. ,crestatbowisclearlylessthanhalfthenorma ldisplacement) ,experimentsbyAlDalzell(5) ,thegeo-metricalrelationshipbetweenwavea ndship--or ship/wavematching --isofprimesignificance~L.}

4 ,[-*- d!--- ,~ !,,.. !s,.,,7,.. ,..>7>~[1;..!.. Figure3, (derivedfromtestsinirrexularheadseas)Oth erresearch,boththeoreticalandexperimenta l,basbeenextendedCoincludelateralbending andtorsioninobliqueseas(6)(7)Excellentag ree-menthasbeenobtainedgenerallybetweent heoryandmodeltests, ,andfurtherdevelopmentsareneeded(8) ,asdefinedbytheirdirectionalspectra,This procedureyieldsshort-termstatisticswhich canbeintegratedoverseaconditiontoobtainl ong-termpredictions(9)Fall-scalestatisti caldataonwavebendingmomentshavebeencolle ctedoverperiodsof2to3yearsintheformofstr essesorstrains(10)(11)For-tunately,ithas beenfound thataship shull,eventhoughabuilt-upboxgirderrather thanahomogeneousbeam,followsthesimplebea mtheoryquitewell, ,measuredstressescanbeinterpretedasexter nalbendingmoments,withthehelpofsimpledoc kside calibrations , ,thishasprovedtobetheonlywaytomakecorrel ationsbe-tweenfull-scaleandmodel(ortheor et-ical).]]

5 Gi!.s.,,.), ,aswellasprac-tical, , , ,therearevibratoryeffectsresultingfromim pactandhigh-frequencywaveexcitationthatm ustbetakenintoco~. , (12)considersultimatefailureasthecomplet ecollapsebybucklingofthecompressionflang eandsimultaneoustensilefailureofthetensi onflange.!0y--y ~Y +,,!O,,..!..1>,,,1,2,.,,.0.,,..,,,,,,, ,.,,-,,,m a,,,w,.,,Im., ,0., , !--- However,itisclearthataconsiderablylessse veredamagewouldbeaseriousmatter:asindica tedbysuchfactorsasnecessityformajorrepai rs,interferencewithnormalshipoperationan dnon-water-tightnessHence,forourpurposew emayde-finedamageasastrutturaloccurrence thatinterfereswiththeoperationoftheshipt otheextentthatwithdrawalfromserviceforre pairisrequired,suchas.

6 -Excessivelocalhulldeflec- , ~ (orfailure)inmoredetail,thefirst,excessi vehulldeflectionisarareoccurrence,except locally, ,theymaybeexcessive,Loadsthatcancombinet othreatenhullfailurearestill aterbendingmoments,wave-inducedbendingmo ments(quasi-static),vibratory(highfreque ncy)loadsandthermaleffects(13)Secondisth epossibilityoffa-tiguecracking,whichseld omconstitutesfailurebutisimportantfortwo reasons:Fatiguecracks,whicharefairlyfre- quent,cangrowtothepointthattheymustberep aired, (14), Itisafavorablecircumstancethatfatiguecra ckspropagateveryslowlyinship sstructures .Cyclicloadstobeconsideredincludethesame loadsasmentionedabove, ,whichwasaser-iousproblemwithearlywelded shipsdur-ingWorldWar2,waslongagobroughtu ndercontrolbyinsuringsatisfactory notch-toughness ofshipbuildingsteel,aswellasbyeliminatin gseveredesignstressconcentrationsandbyim provingweldingtechniques,inspection, ,andthereforethephilosophyhasbeenoneof fail-safe , ,thatsincefatiguecrackingdoesnotthreaten thelifeoftheshipa dbrittlefracturecanbecontrolled.

7 Theprimarycriterionofrationalshipstructu raldesignshouldbeoneofultimatestrencth-- avoidinzexcessivedeflectionthr~ughbuckli ng-orplasticflow(15)Accordingly,ultimate strengthandthecorrespondingbendinzloadsw illbegivenparticularattent~ ~tionedabovetheseconsistof:Stillwaterben dingmomentsWave-inducedbendingmoments(au asi-static)Vib;atorybendingmomentsTherma leffectsSTILLWATERLOADSA lthoughthispaperisconcernedprimarilywith waveloads, ,whileevenlargervariationscanbefoundfrom onevoyagetothenextreturnvoyage( ,loadedandballasted) ,therehavebeentwoapproaches:(1)Makecalcu lationsforallex-tremeconditionsofloading poss-ibleintheship slifetimeanddesignforthelargesthoggingan dsaggingconditionsexpected.

8 (2)Setupreasonable,attainableconditionso floadingandestab-lishmaximumallowablehos +zin~ :-pr;-videguidanceinformation,oracompute rprogramtocalculatebendingmoments; ,ifnotexplicitlyadoptedandvigorouslyfoll owed, ,moderncontainershipssuchastheSL-7(16), ,wheretypicalcon-ditionssuchasfullload,b allast,lightload, erifiedbyservicerecords--ofboth theaveragevalueofbendingmomentandthestan darddeviationforeachbasiccondition,Thisa pproachisdesirableM-4becauseshipsinaccor danceareneverloadedexactlywiththedesigne r ,anditisconsistentwithan overallprobabilisticapproachtode-sign(13 )Aneffectpeculiartohigh-speedshipsistheb endingmomentcreatedbytheship , , ~g~ ,however, ~ scrud,ewayofad-justingforincreasinglengt hwastoas-sumeanincreaseinallowablestress Althoughsuchaprocedureisnotlogicalnorcon sistentwithCivilEngineeringpractice(whic h ouldconsidera aria-tioninloadwithconstantallowablestre ss)theresultingstrengthstandardswerereas onablesolongasshinsdidnotincreasedrastic allyi;size,~ ,wavebend-ingmoment ,&ccL1/3,itiseasytoshowthatthisisequival enttoass~minga a eheightccL2/3(andcon-stantallowablestres s)Asamatteroffactadeii~;m;a.

9 ~e~~~tt~LYn~t~ti~ , , :aspreviowlynoted,newresearchtechniquesh adbecomeavail-able,includingmodeltesttec hniquestomeasurewavebendingmoments,theo- reticalmethodsofcalculatingmotionsandben dingmomentsinbothregularandirregularwave s,newdataonwavepatternsinspectralform,an dfull-scaleshipstresscollectionprogramsT hese ewaePrOache~havebeenshown(13)tofittogeth erInacons~stentprobabilisticpictureofwav eloadingsonships, ,however, .stheABSswpectedthattheextrapolationofcu rrentdesignstandardstotankersintherangeo f800-900feetwasleadingtoexcessivelyse-ve rerequirements, , (17),onthebasisofasinglese-verestormseas pectrum,thatonacom-parativebasis,designw aveheightforfulltankersincreasedslo lyabo ~ ,SubsequentworkatWebbundertheguidanceofa specialPanelonLargerVesselsoftheABSN avalArchitectureConnnitteemadeuseofwaves pectraofdifferentlevelsofseverity,consid er-ingthefrequencyofoccurrenceofeachintb eNorthAtlantic,Althoughthisworkwasbasedo nprobabilisticpredic-tionsfordifferentsi zesofgeometricallysimilarships, !

10 Sincetheprobabilitylevelfordesignwavehei ghttrendswasselectedtocorrespondtovalues ofwaveheightknownfromexperiencetobesat-i sfactoryat600-footlength--includ-ingstil l ~ordropabov~ ,effectivewaveheight,he, (orsinusoidal)wavewhoselengthisequaltoth atoftheship, hichbyconventionalstaticM-5 Figure5,,.LqJf ,chasaconvenientphysicalinter-pretationi ntermsofconventionalwavebendingmomentcal culationsmadebynavalarchitects,Lislength ,Bisbreadth,CWiswaterDlanecoefficient0is massdensityand gistheaccelerationofgravity,Substituting theaboveexpressionforstaticwavebendingmo ment,Kcancelsout,andhe= Sincetheirregularwavebendingmomentabovei scentinual1yvaryingfromoneseaconditionto another, ,forexample,thevalueexpectedtobeexceeded oncei