SSC-205

STRUCTURAL DESIGN REVIEW OF LONG,CYLINDRICAL, LIQUID-FILLED INDEPENDENT

CARGO TANK BARGES

This document has been approvedfor public releaseand sale;its

distributionis unlimited.

SHIPSTRUCTURECOMMITTEE

I970

MEMBER AGENcIES,

UNITED 5TAT. S COAST GUARD

NAVAL 5$+1, SYSTEMS COMMAND

MILITARY ,..4 TRANSPORTATION SERVICEM& R,,, ME ADM, t4,ST RAT10N

AMERICAN BUREAU 0, SHIPPING

ADDRESS CORRESPONDENCE To:

S. CR ETA-,

5.4P 37 RUCTUF?E COMMITTEE

“.s. COAST G“Ae D HEADQUARTERS

WAS. H4NGTON, D.C. 2059t

1970

Dear Sir:

The possibility of transporting liquid chemical cargoes inlarge tank barges on the open sea has necessitated an assessment ofthe state of the art in barge-tank design, to determine what furthertheoretical and experimental development is required.

Herewith is a final project report containing the reviewand recommendations of the study.

Sincerely,

m~~udActing Chairman, Ship StructureCommittee

SSC-205

FinalReport

on

ProjectSR-184,“ChemicalTank-BargeDesign”

tothe

ShipStructureCommittee

STRUCTURALDESIGNREVIEWOFLONG,CYLINDRICAL,LIQUID-FILLEDINDEPENDENTCARGOTANK-BARGES

by

C.W.BascomGeneralDynamicsGroton,Connecticut

under

DepartmentoftheNavyNAVSECContractNOO024-68-C-5419

This documenthasbeenapprovedfor public release andsale; its distribution is unlimited.

U.S.CoastGuardHeadquartersWashington,D.C.

1970

“.., —

ABSTRACT

Thisreportdescribesa programofanalyticalresearchtodeterminetheavailabilityofreliablemethodsfor thedesignoflong,largediameter,cylindricaltanksandtheirsupportsfortransportationofliquidsandlow-pressureliquifiedgasesinbargesforserviceonriversoratsea. Loadingconditions,ex-istingdesign/analysismethods,materialconsiderations,andacomputermethodforpredictingstressesarepresented.

Themajorconclusionoftheworkperformedisthatdesiqnproceduresfor riverbargetanksupto20feetindiameterarewellestablishedandthatno failuresdueto inadequatedesignpracticehavebeenreportedsincerefrigeratedtankswentintoserviceabouttenyearsago. Thepresentmethodfordesigningriverbargetanksisa logicalstartingpointfordeterminingthestructuralconfigurationoflargetanksforoceanicservice,butmoredetailedanalysisofloadsandresultingstressesshouldbeperformedforthisapplication.

Severalareasinwhichtheoreticalorexperimentaleffortisneededareidentified:(1)investigationoftank-saddle-bainteraction,(2)investigationoffatiguecriteriafor cyclicloading,(3)investigationof bucklingcriteria,(4)analyticaandexperimentalinvestigationofslamming,and(5)experimentaverificationofstressesinafull-scaletank.

ii

CONTENTS

INTRODUCTION.. . . . . . .. . . . . . . . . . . . . .. . . + .

APPROACH. . .

CONCLUSIONSAND

TANKBARGELOAD:

. . . . . . . . . . . . . . . . . . . . . . . . .

RECOMMENDATIONS.. . . . . . .. . . . . . . .. .

NGS.. . . . . . .. . . . . . . . . . . . . .. .

STRUCTURALDESIGN/ANALYSISOFTANKBARGES.. . . . . . . . . . . .

EVALUATIONOFSTRESSESINEXISTINGANDPROJECTEDDESIGNS. .. . .

DISCUSSIONOFMATERIALSANDCONSTRUCTIONOFPRESSUREVESSELSFORBULKTRANSPORTOFLIQUIDCARGOESONBARGES.. . . . . . .. .

ACKNOWLEDGEMENTS.. . . .. . . . . . . . . .. . . . .. . .. .

REFERENCES. . . . .. . . . . . .. . .. . . . . . . ...”” ””

APPENDIXA - INVESTIGATIONOFTHEENVIRONMENTOFLARGEOCEAN-GOINGBARGES.. .. . . . . . . . . . . . . . . . . . . . .

APPENDIXB -ANALYSISOFRINGSTIFFENERS. . . . . . . . . .. . .

APPENDIXC -OUTLINEFORSTRAINGAGEINSTRUMENTATIONOFATANKBARGE.. . . . . . .. . . .. . . . . . . . . . . . ● ● ● “

APPENDIXD - DISCUSSIONOFAPPROACHFORTANK/BARGESLAMMINGTESTS.. . . .. . . . . . . . . . .. . . .. . . . . . . . . .

iii

SHIPSTRUCTURECOMMITTEETheSHIPSTRUCTURECOMMITTEEisconstitutedtoprosecu

gramto improvethehullstructuresofshipsbyanextensionofkningtodesign,materialsandmethodsoffabrication.

CaptainJamesB,McCarty,Jr.,USCG- ActingChairChief,OfficeofMerchantMarineSafety

U.S.CoastGuardHeadquarters

CaptainW,R.Riblett,USN Mr.E.S.DillonHead,ShipEngineeringDivision Chief,DivisionofSNavalShipEngineeringCenter OfficeofShipCons

MaritimeAdministCaptainT.J.Banvard,USN Mr,C.J.L.SchoeMaintenanceandRepairOfficer VicePresidentMilitarySeaTransportationService AmericanBureauofS

SHIPSTRUCTURESUBCOMMITTEETheSHIPSTRUCTURESUBCOMMITTEEactsfortheShipStruc

technicalmattersby providingtechnicalcoordinationforthe degoalsandobjectivesoftheprogram,andby evaluatingandintesultsintenmsofshipstructuraldesign,constructionandoperat

NAVALSHIPENGINEERINGCENTERMr.J.B.O’Brien- ActingChairmanMr.J.B.O’Brien- ContractAdministratorMr.G.Sorkin- MemberMr.H.S.Sayre- AlternateMr.1,Fioriti-AlternateMARITIMEADMINISTRATIONMr.F.Dashnaw- MemberMr.A.Maillar- MemberMr.R,Falls- AlternateMr.W.G.Frederick- Alternate

AMERICANBUREAUOFSHIPPING

Mr.S. G,Mr.F.J.

OFFICEOF

Mr.J.M.Dr.W.G.

Stiansen- MemberCrum- Member

NAVALRESEARCH

Crowley- MemberRauch- Alternate

MILITARYSEATRANSPORTATION

Mr.R.R.Askren- MemberLt.J.G.T.E.Koster,USN-

SERVICE

Member

U.S.COASTGUARDLCdr.C.S.LoosmoCdr.C.R.ThompsoCdr.L.C.MelbergUCdr.L.A.Colucci

NAVALSHIPRESEARCMr.A.B.Stavovy-

NATIONALACADEMYOF

&A

S

Mr.A.R.Lytle,LiaMr.R.W.Rumk8,LiaMr.M.L.SellersL

AMERICANIRONANDST

Mr.J.R.LeCron,LiBRITISHNAVYSTAFFMr.H.E.Hogben,LiCdr.D.Faulkner,RC

WELDINGRESEARCHCOU

Mr.K.H.KoopmanLMr.C.Larson,Liai

LISTOFILLUSTRATIONS

Qk3!z94-14-2

4-34-+,

5-15-25-35-45-5

5-65-75-85-95-1o5-11

5-12

5-13

6-16-26-36-46-56-66-76-86-9

A-1A-2A-3A-4A-5A-6A-7A-8A-9

TrendinFundamentalFrequenciesvs. BargelengthBargeVelocityandDistancevs. TimeforConsbntDecelerationof1.5g

AcousticPressurevs. SpeedforVariousBargesAverage1/1OHighestSlammingPeakPressureBendingStressvs. ThicknessforVarious200-FootTanksBendingStressvs. ThicknessforVarious300-FootTanksBendingStressvs. ThicknessforVarious400-FootTanksThreeRepresentativeTankDesignsNondimemionalBucklingCurveforCircularTubesinCompression

Configurationofa TypicalRiverBargeConfigurationofaTypicalOcean-GoingTankBargeTypicalRiverBarge~ctionalViewTypicalOcean-GoingTankBargeSectionalViewTypicalTankCharacteristicsandSaddleSupportsTypicalloading,Shear,andMomentDiagramsfora

GroundedRiverBargeTypicalImading,Shear,andMomentDiagramsforanOcean-GoingTankBarge(SaggingCondition)

TypicalLoading,Shear,andMomentDiagramsforanOcean-GoingTankBarge(HoggingCondition)

TypicalStructuralModelforComputerAnalysisTypicalTankReinforcementAdapWbletoFabrication

E%E

1417191926272829

323434353536

37

39404344

200-FootTank– NormalOperation—ImgitudinalStress(o= O)47200-FootTank- NormalOperation—HoopStress(d=O) 47200-FootTank—Grounding- LongitudinalStress(@=O) 48200-FootTank- Grounding- HoopStress(6=O) 48400-FootTank– NormalOperation—longitudinalStress(0=O)52400-FootTank– NormalOpe=tion- HoopStress(o=o) 52InvestigationofCycliclaadingforFatigueAnalysisof400-FootTank 56

Pitchvs. Speed(NeumannSea) 77BOW Accelerationvs. Speed(NeumannSea) 77RollinRandomBeamSeas(MostProbableRollAngle) 77RollinRandomBeamSeas(Average1/1OHighestAngle) 78WaveLengthsAlongU.S.Coasts 82WaveHeightsAlongU.S.Coasts 82WavePeriodsAlongU.S.Coasts 83FrequencyofEncounterat10Knots 83ComparisonofNaturalPeriods 85

v

LJ-’LISTOFILLUSTRATIONS,~:o:lt“(i)

NaturalPeriodsofVibrationofFluidsContainedinaCircularCylinderofRadius20Feet,Length400Feet

SimplifiedSloshingMathematicalModelLoadingonaRingStiffenerTypicalLocationofExtensiveInstrumentationTypicalInstallationofPermanentPhotoelasticMaterialon

InsideofHullPressureDistributionOvertheWedge

LISTOFTABLES

MomentsandShearsforVariousTankConfigurationsValuesofAnforCalculatingFrequencyAverage1/1OHighestRelativeVelocityatSlammingStation

8687

9198

98

102

SummaryofTankConfigurationsSelectedforAnalysisof GroundingandSagging/Hogging

SupportIaadSummaryandMomentDistributionSupportImadandMomentSummary- SaggingConditionSupportbad andMomentSummary- HoggingConditionAppliedIaadsforNormalOperatingConditionsMaximumStressin200-FootTankBasedonComputerAnalysisIacalLongitudinalBendingStressatShell-StiffenerIntersection,400-FootTank(Normalhads)

MaximumStressesin400-FootTankBasedonComputerAnalysis

ComparisonofMaximumMid-SurfaceStressesMaximumCircumferentialStressinStiffeningRingImcalPeakStressCycleintheShell-RingIntersectionMaterialPropertiesPhysicalPropertiesofGasesBargeCharacteristicsWaveDataDeterminingFrequencyof EncounterNaturalPeriodsofMotionBargesand1/2FullTanks

101520

303841424649

50

51535457616976818486

vi

Section1 ~b

INTRODUCTION

Thetrendinthedistributionoflargevolumesofindustrialgaseshasbeentowardtherefrigeratedmodeoftransportandstorage.Independenttankbarges‘haveproventobebothpracticalandeconomicalandthismodeoftransportisbeingconsideredforcoastalandoceanicservice.

Dimensionsofriverbargesarelimitedtoapproximately10feetindraft,53feetinwidth,and300feetinlength;thedraftdimensionis controlledbyriverdepth,andlengthandbeambyriverlocksize. Thesedimensionslimitthemaximumcapacityofthebargetoapproximately3,000tons. Thisinturnlimitsthediameterofthecargotankstoapproximately20feetandthelengthtoabout250feet. Twotanksareusuallymountedsidebysideonthebarge,andsupportedonfrom7to13saddles.Stiffenersareinstalledatthesaddlestoaccommodatethehighreactionloadsatthesepoints.Typicaltanksarefabricatedof1/2-inchcarbonmanganesesteel. IXMignpressuresareaslowas4to10psianddesigntemperaturesareapproximately-30°F.Thetanksarecoveredwithapproximatelythreeinchesofinsulation.Redundantrefrigerationplantsandsafetyvalvesareprovidedtoensureagainstoverpressurizingthetanksduetovaporizationofthefluid.

Fromthestructuralpointofview,theselargeriverbargetankshaverelativelysmallthickness-to~iameterratios,i.e., theyarequitethinwalled.Theyoperateessentiallyatatmosphericpressures,andreactionforcesratherthanpressurestressesgovernthedesign.Theempiricalprocedurefordesigningthesetanksforreactionforcesisbasedonexperimentalworkwithstationarytankshavingjusttwosupportsandrelativelyheavierwallsthantoday’slargeriverbargetanks.

Independenttankbargesareenvisionedinthenearfutureforcoastalandtransoceanicservice.Verylargebargesinthe20,000-tonrangeareeconomicallyattractive.Sinceoceanorcoastalbargeswillnotbesubjectedtothedimensionallimitationsofriverbarges,tanksaslargeas40feetwideand400feetlongareenvisioned.

Twomajorquestionsariseconcerningthedesignoflargeoceanservicetankbarges.1. Whatloadingconditionsareapplicabletothedesignoflargecylindrical

tanksforoceanservice?2. Istheempiricalproceduredevelopedforsmaller,heavierwalled

stationarytanksapplicabletothelargersizes? Ifnot,whatis themostreliableprocedure,orwhatfurtherworkisneededtoderiveanadequateprocedure?

Interpretiveanswerstothesequestionshavebeentheprimeobjectiveofthisthree-monthstudy.Thisobjectiveis statedmorefullyinthescheduleofthecontractasfollows:“Analyticalresearchshallbeundertakentodetermineavailabilityofreliablemethods

-

-2-

forthedesignoflong,cylindricaltanks,andtheirsupports,forthetransportationofliquidsandlow-pressureliquifiedgases.inbargesonriversoratsea. Theworkshallinvolve:

1.

2.

3.

4.

Descriptionoftheloadsandloadingconditionswhich-mustbeconsideredintankdesign,forsizesupto40feetindiameterand400feetinlength.Determinationoftheanalyticmethodspresentlyavailableforuseinthedesignoftanksandtheirsupports,wheninstalledinbarges.IJ%erminationofthemostreliablemethodorcombinationofmethodspresentlyavailabletoextendsuchdesign,fromthestandpointofsafety,economy,andefficientdesign,tothelargertanks.Determinationofthoseareasinwhichtheoreticalorexperimentalworkisneeded.‘‘

-Li-

Section2

APPROACH

SixbasictaskswereperformedinordertoaccomplishthefourobjectivesstatedintheIntroduction.

TASKA-TASKB-TASKC-TASKD-TASKE-

TASKF-

InvestigationofChemicalTank/BargeOperatingConditions.InvestigationofTank/BargeLoadings.InvestigationofTank/BargeDesign”Characteristics.EvaluationofStressesinExistingandProjectedDesigns.EngineeringInvestigationofMaterialsProblemsAssociatedwithIargeTanks.PreparationofanInterpretiveReportIncludingRecommendationsforResearchinMajorProblemAreas.

Backgrounddatafortheabovetaskswasobtainedbyreviewingtheliteratureandbycontactingpersonnelinthebargeindustry.Theliteraturereviewisreflectedinthelistofreferences.Muchhelpfulbackgroundmaterialwasobtainedbycontactingregulatingbodies,surveyors,designers,buildersandoperatorsoftankbarges.(SomeofthemanyhelpfulcontactsmadeinthecourseofthestudyarelistedintheAcknowledgements.)

InordertoperformTasksCandD, itwasexpedienttoworkwithspecifictarddbargeconfigurations.Sinceexistingdesignsareofaproprietarynature,twohypotheticaldesigns- onerivertypeandoneoffshoretype- wereselectedforexamination.Thefollowingprocedurewasusedtodeterminetankand’bargecharacteristics.

Configurations,i.e., wallthicknessesandnumberandspacingofstiffeners,weredeterminedforvarioustanklengthsanddiameters.Thesewereaccomplishedbydeterminingreactionsduetotankdeadweightfromelementarystructuraltheory.Dynamicforcesduetopitch,roll,andheavewereaccountedforbyapplyingadynamicloadfactortothestaticforces. Densityofthefluidinthetankswasas-sumedtobe42poundspercubicfoot,whichisrepresentativeofseveralliquifiedgasesnowbeingtransported.Tankshavingfrom2to11saddleswereconsidered.Threefamiliesoftankswereinvestigated:20-footdiametertanks,200feetlong;3O-footdiametertanks,300feetlong;and40-footdiametertanks,400feetlong.Thewallthicknessforeachdiameter,lengthandsupport(numbe~ofsaddles)con-figurationwasdeterminedbyassumingthatthelgoverningcriterionwasthebucklingofshortcylindricalcolumnsasdefinedbyZick. Forselectedlengthsanddiameters,curvesofcriticalstressvs. thicknesswereplottedforconfigurationshavingfrom2to11’supports.Fromthecurves,representativetankwallthicknesseswereselectedforariverbargeandanoffshorebarge,basedonlowerlimitsoftankwallthicknessconsideredpracticalforfabrication.

Rough,structuraldesignswerealsopreparedofariverbargeandanoffshorebargewhichwouldaccommodatethepreviouslyselectedtanks.Thepurposeofthiseffort

-4-wastoobtaintheweight,stiffnessandbuoyancycharacteristicsofthebargeforuseinevaluatingbarge/tankinteractionduetowaveactioninthecaseofanocean-goingtankbargeandgroundinginthecaseofarivertankbarge.

Thetwotank/bargeconfigurationswerethenanalyzedfortheloadingconditionsestablishedinTaskB. Oftheseloadingconditions,themostsevereisgroundingforariverbargeandsagging/hoggingforanoffshorebarge.Reactionforcesweredeterminedfo thesesevereconditionsusingtheiterativeprocedureoutlinedbytheCoastGuard.1.Thmmethodisbasedontheassumptionsthatreactionforcesareprimarilydependentonbendingstiffnessandtheeffectsofshearstiffnessarenegligible.

Stressesintheareaofthesaddleswerethenevaluatedbytwomethod%themethodofZicklwhichisnowcommondesignpracticeandthemethodofKalnins, amoresophisticatedcomputerapproach.Thecalculatedstressesineachcasewerecomparedwithallowablestress. Thisanalysisdemonstratedproceduresofthetwomethodsandcomparedresults,ratherthanevaluatedthehypotheticaldesigns.

-5-

Section3

CONCLUSIONSANDRECOMMENDATIONS

Thissectionsummarizesthemostsignificantconclusionsoftheinvestigationandgivesrecommendationsforfurthertheoreticalandexperimentalwork.Conclusionsre-gardingspecificloadsanddesign/analysisproceduresarecontainedinsections4through7.

Design/analysisproceduresforlow-pressure,refrigeratedtanksforserviceonriversarewellestablished.Asurveyofdesigners,regulatorybodies,builders,surveyors,andoperators“indicatesthatnomajorfailuresduetodesigninadequacyhaveeverbeenreportedsincethistypeofbargecameintoserviceabout10yearsago. Inviewoftheexcellentoperatinghistoryandrecordofrivertankbarges,thedesignproceduresforriverbargetanksofupto20feetindiameterforriverbargeapplicationareconsideredadequate.Inmanycasesoperatorsspecifystructuralstrengthinexcessofregulatorybodyrequirements,

Designproceduresforriverbargesaregenerallyapplicablefordeterminingthebasicconfiguration%oflargertankscontemplatedforoceanservice.Thisconclusionisbasedonthegoodagreementbetweenmidsurfacestressescalculatedbytheestablishedempiricalprocedureandamoresophisticatedcomputeranalysis.How-ever,theempiricalproceduredoesnotgivestressesatenoughpointstofullysatisfyinputsforanalysisofcyclicloadsontanksforoceanserviceandamoredetailedstressanalysiswillberequiredforthiscase. Furthermore,theoreticalpredictionsofstressinlarge,thin-walledmultisupportedtanksshouldbeverifiedexperimentally.Theoreticalpredictionshavebeenverifiedonlyonsmaller,heavy-walledtaokssup-portedonjusttwosaddles.

Iangtanksforoceanservicewillbesubjectedtocyclicloadsandrelativelylargedeflectionsasthebargesagsandhogsduetooceanwaveforces. Cyclicloadsarenotassignificantinriverbargesandtherefore,criteriaforevaluationoftheseloadshavenotbeenestablished.Criteriaforcyclic-loadingandfatigueevaluationincludingfactorsforeffectsofsurfaceimperfectionsshouldbeestablishedforoceantankbarges.Thebargetank/deflectionswillcauseinteractionbetweenthebargestruc-ture,thetankstructure,thesaddlestructure,andthesaddleinsulation.Thespringconstantofthesaddleinsulationmaterialisnonlinearwhichgreatlycomplicatesexactpredictionoftheinteraction.

Duringtheinitialthreemonthswork,severalspecificproblemareasneedingfurtheranalyticalandexperimentalresearchwereidentified.Theseproblemsaregenerallynotapplicabletorivertankbargesnowinservice,butapplytothelargeroceanbargesenvisionedforthefuture.

3.1 EXPERIMENTALANDANALYTICALANALYSESOFANAS-BUILTTANKTheforemostproblemconfrontingthedesigneristhequestionofadequacyofdesign/analysistechniquesavailabletohim. Ourinvestigationtodatehasshowngoodagreementbetweenthesimplifiedapproachnowusedforriverbargesanda moresophisticatednumericalanalysisprocedureofpointsonthetankwherethesimpli-

— — -.

-6-

fiedmethodapplies.Neithermethod,however,hasexperimentaldatatoverifyresultsinthelargersizesenvisionedforoceanbarges.

Atankbargeinthebuildingstageshouldbeinstrumentedwithstraingagesforthepurposeofcheckinganalyticalresults.Thegagescouldremainonthetankafteritisputintoserviceforaspecifiedtimeandrecordingsmadeofstressesundervariousloadingconditions.Furtherdiscussionofanexperimentalprogramiscon-tainedinAppendixC.

3.2 FATIGUEANALYSISInthecourseofthestudy,itbecameevidentthatspecificexperienceindesigninglargethin-walledtankssubjecttocyclicloadsencounteredinoceanserviceisverylimited.Itwasalsoevidentthatthesimplifiedstressanalysisproceduresapproxi-matemembraneormid-fibrestressatselectedpointsonly.Amorecornprehensiveexaminationofstresses,bothinsideandoutsidethetankwall,isnecessaryforafatigue.analysis.Also,allowablestresslimitsforfatigueanalysisoftankmaterialshavenotbeendetermined.Datamayexist,and,if so, itmustbecollectedandrelatedtothetank/bargeapplication.Ifdatadoesnotexist,thenexperimentalworkwillbenecessary.

3.3 BUCKLINGANALYSISThereiswidedivergenceinthecriticalcompressivebucklingstressesdeterminedfrommethodscontainedintheliterature.Forexample,thecriticalbucklingstressasdeterminedbythemethodofTimoshinkoisgreaterbyafactorof2thanthevaluedeterminedbythemethodofZick. Thisareacertainlyneedsfurtherinvestigation.Amoreextensivereviewoftheliteratureandaninvestigationofbucklingcriteriadevelopedforotherapplicationsareproposed.Amodeltestprogrammaybenecessaryifnoapplicabledataisavailable.

3.4 SLAMMINGINVESTIGATIONSlammingisamajorareaofconcerninthedesignofallhullsforoceanservice.Oceantankbargesarenoexceptionwhereslammingloadsappeartoaffectthetankaswellasthebargehullitself. IntheareaoftheforwardrakebuIkhead,slammingmaycauselargedeformationofthehullwhichistransmittedupintotheforwardtanksaddle.Thetanksaddleisseparatedfromthehullbyalayerof insulationwhichmaycushionslammingloads,buttowhatextentthisoccurshasnotbeendetermined.Experimentalworkwithspecificmodelbargehullsshouldbeunder-takentodeterminepressuredistributions.

Thenextstepwouldbetoapplythesepressurestothehulltankstructurewithproperboundaryconditionstodeterminethehull/lankinteraction.Theproblemappearstobequitecomplexbutnotimpossibletosolveutilizingtoday’scomputertechnology.AfurtherdiscussionofamodeltestprogramforinvestigatingtankbargeslammingiscontainedinAppendixD.

3.5 TANK/BARGEANDSADDLEINTEWCTION

.

Theeffectofsaddleflexibilityonsagging/hogginganddynamicloadsshouldbedetermined.Thisproblemcouldbeapproachedbyutilizingamatrixstructuralanalysisprocedure.Thetankandbargewouldeachberepresentedby(n+1)

— -. — —

-7-

stiffnessmatrices,wherenisthenumberofsaddles.Eachsaddlewouldconsistoftwostiffnessmatrices:onefortheinsulationmaterialandoneforthesaddlestruc-ture.Stiffnessoftheinsulationmaterialwouldbedeterminedfromthemanufacturers’dataorfromtesting.Severalanalyseswouldbeperformedtoevaluatetheeffectofhardandsoftsaddles.Uniformload~duetoweightandvariablebuoyancyloadswouldberepresentedbyatleastthreeconcentratedloadsbetweeneachsupport.

.

-8-

Section4

TANKBARGELOADINGS

Refrigeratedcargobargesareoperatedatatmosphericpressure,withthedominantloadbeingcausedbyreactionsatthesupportsduutocargoweightratherthanbypressure.Furthermore,theverylargetanksenvisionedforocean-goingbargeswillbesubjectedtodynamicforces,inadditiontotheweightofthecargo,asthebargepitches,rolls,andheavesinaseaway,andtosaggingandhoggingforcesasthebargehulldeflectswhilewavespassunderit. Saggingandhoggingarecyclicloadsthatcausethefati~uestrengthofthetanktobeanimportantconsideration.Vibrationofthetankcausedbywavemotionmayoccurif naturalfrequenciesofthetankareclosetothewaveencounterfrequenciesofocean-goingbarges.

Riverbargesarenotsubjectedtothelarge,dynamicreactionloadsoftheirocean-goingcounterpart.Themostsevereloadonariverbargeiscausedbysupportre-actionsinthegrouncledcondition.Reference4discussessomedesigntechniquesandregulationsforriverbargestransportinghazardouscargoes.

Itis commonpracticeinthedesignoftankstoassumethatthesaddlereactionforcesandlongitudinalbendingmomentsmaybeobtainedbyaniterativeprocessutilizinga modelofanelasticbeam(thetank)mountedonanotherelasticbeam(thebarge).Inherentinthisprocedurearetheassumptionsthatthesaddleanditsfoundationareinfinitelyrigidandthatthemomentinthetankisalwaysacertainpercentageoftheoverallbarge/tankbendingmoment.Actually,thetanksaremountedonthermalinsulation(20#urethanefoam)whichalsocushionsthetankandhelpstodistributepeaksaddleloadstoadjacentsupports.Thus,theassumptionofrigidsupportsisconsideredtobeconservative.Theassumptionthatthetankcarriesacertainper-centageoftheoverallbendingmomentis consideredreasonableif thetankishelddownonthesaddles,if thetankstiffnessisnot,lessthanaboutonethirdofbargestiffness,andif theneutral=es ofthebargeandtankareseparatedbylessthanaboutonehalfofthetankradius.Theseconditionsaresatisfiedintypicalindependenttankbargedesigns.

Theprobabilityofaseveregroundingonapinnacleattheforwardrakebulkheadisextremelysmall,andthisfactisacknowledgedintheCodeofFederalRegulationbytheallowanceofastressequaltotwothirdsoftheultimatetensilestress.Inviewoftheseverityofthespecifiedgroundingloadandthesmalllikelihoodthatitwilloccur,amoresophisticatedapproachfordeterminingtank-barge-saddleinteractioninthedesignofriverbargesdoesnotappeartobenecessary.

If, ontheotherhand,tanksaredesignedasstructuralmembersofoceanbargessubjectedtomillionsofcycIesofsaggingandhogging,theneffectssuchassaddleflexibilitymayhavemoresignificance.Analysisof thetankmountedonthebarge,includingtheflexibilityofthesupports,ispossibleutilizingastiffnessmatrixapproach.Hgwever,aproblemariseswhendeterminingtheflexibilityofthefoaminsulationmaterialwhichseparatesthetankfromthesaddle.Datawhichadequatelydescribestheelasticand/orplasticcharacteristicsoftheinsulationmaterialapparentlydoesnotexistintheliterature.

-9-

Duetotheuncertaintyoftheelastic/plasticpropertiesofthesaddles,andthetimeandexpenseinvolvedinformulatingacomputermodel,theeffectsofsaddleinter-actionwereidentifiedasaproblemareaforfurtherinvestigationratherthanpursuedfurtherinthisstudy.Thetankloadsusedforevaluationofstressanalysisproceduresweredeterminedinthisinvestigationbytheiterativeprocessdescribedinsection4.3.

4.1 CARGODEADWEIGHTREACTIONLOADS(STILLWATER)Thisloadis commontobothriverandoffshorebargesandisquiteeasytoobtain.Reactionloadsmaybedeterminedusingelementarystructuraltheory,or theymaybeapproximatedinsymmetricaldesignsbydividingthetotalweightofcargoandtankbythenumberofsupports.Table4-Igivesreactio\loadsforvarioussizetankconfigurationswithhemisphericalheadand42lb/ft fluid,usingamomentdistributionmethod.

TheAmericanBureauofShipping6usesaconvenientmeansofapproximatingweightperfootofcargotank:

W.

where:w=t=R=

Sp.Gr.=

R2(256~ +196Sp.Gr.)

weightperunitlength(lb/ft)tankthickness(in.)tankradius(ft)specificgravityoffluid(dimensionless)

Formultiple-supportedtanksonevenlyspacedsaddles,andwhenthelengthofover-hangsapproachesonehalfthelengthofeachspan,thesaddlereactionloadinpoundsisqualtotheproductoftheweightperfootandthesaddlespacing.

Withthesaddlereactionandweightperfootknown, shearandmomentdiagramsmaybeplottedforuseincalculatingtankstresses.

4.2 DYNAMICLOA~Themostsignificantdynamicloadiscausedbyaccelerationofthemassofthecargotankanditscontents.Thisloadismaximumif thetanksareassumedtobefull. ~,thetanks~e assumedtO b onlypartiallyfull,sloshing10adswillbepresent.Dynamicloadsundereachconditionarediscussedinthefollowingpara-graphs.

4.2.1 FULLYLOADEDCONDITION—TheCodeofFederalRegulations,Title46Chapter1, subparagraph38.05-2,specifiesthefollowing:

“Cargotanksinvesselsinocean,GreatLakes,lakes,bays,andsounds,orincoastwiseserviceshallbedesignedtowithstandthefollowingdynamicloadings:1. Rolling30°eachside(120°)in10seconds.2. Pitching60halfamplitude(240)in7seconds.3. HeavingL/8O halfamplitudein8seconds.‘‘

—

Table4-1, MomentsandShearsforVariousTankConfigurations(Weight:42lb/ft3)DIAMETER/ NUMBER OVERHANGSUPPORTSPACING MOMENT(lb-R) SHLENGTH(ft) SUPPORTS (ft) (ft) (x10-q (x

20/200 23456‘789

1011

30/300

40/400

23456789

101123456789

1011

43.7028.2020.8116.4913.6511.6510.169.018.097.34

65.442.231.224.620.517.415.213.411.010.187.356.341.532.927.323.320.318.016.214.7

107.0769,0950.9840.4033,4428,5425.9922.0719.8217.98

160103.576.360.550.043.037.333.129.927.1

213.8137.9101.880.666.857.049.744*I39.635,9

14.15.883,202.011.381.01,762.601.484.399

71.7329.3316.3110,147,045.093.873.012.031.71

230.595.852.132.722.516.412.59,807.946,53

2111

6432211111

..

-11-

Theseconditions\vPreinvestigatedinTaskAandfoundtobereasonable.AppendixAgivestheresultsofthisim’estimation.

Usingthe.abovethreeconditionsforpitch,roll,andheave,togetherwiththechar-acteristicsofthebarge,adynamicverticalloadfactormaybeapproximatedfromelementaryequationsofharmonicmotion,resultinginthefollowingmaximumvalues:

Gv=p +r’hg (4-1)

and

where: Gv=p=r=h=g.

P =

r =

h=

with: Q=

d=

L=

verticaldynamicloadfactor(dimensionlesss)pitchacceleration(ft/sec2)rollacceleration(ft/sec2)heaveacceleration(ft/sec2)gravity= 32.2ft/sec2

();T 27 1Sin6“

()~r 2

dsin30°10

()2LT ~8 80

variabledistancefromlongitudinalcenterofgravityofbargetothesaddleinquestion(ft)distancefromtheverticalcenterofgravityofthebargetocenterofgravityofthetank.lengthofbarge(ft)

Thedynamicloadfactor,G,maybeapplieddirectlytoeachofthestillwatersaddlereactionstoapproximatethedesignload,Thedynamicloadfactorshouldalsobeappliedtothehydrostaticpressureinthetank.

4.2.2 PARTIALLYLOADEDTANKS- Sloshingloadswillbeprevalentinpartiallyfilled,unbaffledtanks.Whentheyarefilledtoornearcapacity,fluidwillactalmostasasolidmass,anddynamicloadsduetopitch,rollandheavewillbetransmittedtothesupports,asdescribedpreviously.Whentanksarealmostempty,theforceonthesupportswillbegreatlyreducedduetothenegligibleamountofmassofthefluid.Thepredictionof loadsduetosloshingisdifficultintherangeoffluidca-pacityfrom90to10percent.Sloshinginliquidfue tanksofmissileshasbeentreatedquiteextensivelybytheaerospaceindustry.$ However,themethodsde-velopedformissilesdonotappearapplicabletotankbargesforseveralreasons:(1)themotionsofthetankbargevarymorethanthoseofthemissile;(2)thefluidmassisvariableinthemissiletank,wherwsmassisconstantinchemicaltankbarges;and(3)theorientationofthetanksisverticalwithmissilesbuthorizontalinthecaseofchemicaltankbargas.

—

-12-

ApreliminaryinvestigationoffluidsloshingwasperformedinAppendixA (sectionA-3)withtheconchsionthatanadequatemethodofpredictingsloshingloadsinun-baffledtanksdoesnotappeartoexist. Thissituationcanbeovercomethroughapro-gramofexperimentalandanalyticalresearch;however,justificationforsuchapro-gramisquestionablewhenthepracticalaspectsareconsidered.Mostoceanicbargeoperationswillconsistofone-waytripswithtanksfull,andreturntripswithtanksempty,Ifpartialloadsarebeingconsidered,thensloshingmaybegreatlyreducedbytheinstallingofbafflingofthetanktrucktypeinthetanks.

4.3 GROUNDINGMIADSTheconditionforgroundingisspecifiedinTitle46,Chapter1, Paragraph98.03-25oftheCodeofFederaIRegulations.Groundingloadsonthetankwilldependontherelativestiffnessbetweenthetankandbargeandwhetherornotthetankishelddownonthesaddles.Thetanksupportsmaybedesignedsoastocontributetothestrengthofthebarge.Ifthisisthecase,thenthesupportloadsmaybedeterminedbyconsideringthetankasabeamonanelasticfoundation- thebarge.TheCoastGuard*hasformulatedthisanalysiswhichisessentiallyasfollows:Thebarge.isassumedtobegroundedattheforwardrakebulkhead.Aloadingcurve,shearcurve,andmomentcurve,suchasthose~howninsection6, maybeobtainedusingthefollowingprocedurewhichisquotedfromreference8.

“Startingfromtheforward(grounded)end,thetotalbargemomentiscomputedateachsaddle.Thisisdonebysummingthemomentsduetobargehullweight,tankloading,groundingforceandbuoyancy.

APCSitivemomentisonethatplacesthedeckofthebargeincompression,whileforcesarepositivedownward.Thebuoyancycurveisassumedtovarylinearlyfromzeroatthegroundingpointtoa maximumattheafterraketangencypoint.Themomentinach tankabreastis thencomputedonthebasisoftheproductoftheratio

ItankI +1barge tankx numberoftanksabreast

andthetotalbargemomentateachsaddle,exceptthatthemomentattheendsaddleiscomputedasthoughtheoverhangingsectionwereacantilever.Thetankweightis thendividedbythetanklength,andtheresultingweightperfootisassumedtobeevenlydistributed.Sincethemomentisknownateachsaddle,alongwiththedistributedloadbetweensaddles,thesheartotheleftandrightofeachsaddleiscomputedandcombinedtogivethereactionateachsaddle.Thesereactionsaremultipliedbythenumberoftanksabreasttogetthetotaltankreactionateachsaddlelocation.Acheckwillshowthatthesumof thereactionsequalstheweightofthetanks.Thecycleis thenrepeateduntilthesolutionconverges,theonlyvariationbeingthatthesaddlereactionsareusedtocomputethetotalbargemomentinlieuoftheuniformlydistributedtankloadingusedinthefirstcycle.”

4.4 SAGGINGANDHOGGINGLOADSSaggingandhoggingloadsaredeterminedinamannersimilartogrounding.Thebuoyantforce,howeverisobtainedbybalancingthebargeonatrochoidalwavewith

-13-

awavelengthequaltothebargelength.Thepeaksofthewave.areplacedontheendsofthebargetocreatesaggingandthemidpointofthebargetocreatehogging.(Thisprocedureisaxplainedinreference9.) had, shearandmomentcurvesfortypicalsaggingandhoggingsituationsareshowninsection6.

Saggingandhoggingloadsintanksmaybepracticallyeliminatedbyusingtwoormoretanksendtoendratherthanonelongcontinuoustank.Thisconfigurationmaybeadoptedif largenegativeforcesarepredictedinthesagging/hogginganalysis.Ifthisapproachisused,thebargestructuremustbedesignedtocarrytheentirebendingmomentingrounding,sagging,andhogging.

4.5 FATIGUEContinuoustanksrepresentasignificantportion(1/3to1/2)oftheoverallstructureofanindependenttankbarge.Furthermore,thetankstructureis locatedinagoodpositiontocontributetothebendingstrengthofthebarge.Thus,itiseconomicallyattractivetodesignthetankstocarryaportionoftheoverallbendingmoment.

Toaccomplishthis,thesaddlesmustbedesignedtotransmitloadsbetweenthetankandbargeinsuchamannerthatthetwostructuresactasanintegratedstructure.Whenthestructureis integrated,kmththetankandthebargemustresistthecyclicloadsofsaggingandhogging.Reference10presentsanengineeringapproachtolow-cyclefatigueofshipstructures.Theconclusionofthisreportis thatmostofthebendingcyclesexperiencedbyashipstructureinducelownominalstresses.There-forefatigueofthemainstructuralgirders,perse, isnotofprimeconcern.How-ever,lowstressintensitiesaremagnifiedbyunavoidablediscontinuitiesinlocalareaswheretheyieldstrengthmaybereachedorexceeded.Thus,low-cyclefatigueisarealproblemincertainlocalizedareasoftheshipstructure,

Fromapreliminaryexaminationoftypicaltankstruttures,itappearsthattheareaofthestructureintievicinityofthesaddlesmaybea troublespot,particularlyifcorrosivefluidsarecarriedinthetank.Inthisarea,heavystiffenersarejoinedh therelativelythintankwall,creatinggeometricaldiscontinuities.Furthermore,residualstressesanddiscontinuitieswilloccuraroundtheweldsrequiredtojointhestiffenertothetank.

Itappearsthattherearenoguidelinesavailabletothetankdesignerwhichwillassisthiminaccountingforgeometricaldiscontinuities,weldtreatment,andstresscorrosion,andthusdesigninga tankwhichwillresistcyclicloading.Solutionstotheseproblemsarenecessarybeforethetankcanbeutilizedasastructuralmemberofanocean-goingbarge.4.6 FORCEDVfBmTIONLOA.IETheforcedvibrationloadsonatankmaybesignificantifthenaturalfrequencyofthetsmk/bargeisclosetothefrequencyofwaveencounters.Thefrequencyofwaveencountersmay,ofcourse,bechamgedoperationallybyreducingspeedorbychangingthecourseofthebarge.Thisloadingconditionshouldbecheckedespeciallywhenverylongtanksandbargesarebeingconsideredforhigh-speedoperation.Figure4-1showsthetrendinbargefrequencyversuswaveencounterfrequenty.Asbargesapproach600to700feetinlength,natualfrequenciesapproachthewaveencounterfrequency.Reference11presentsasimplifiedmethodofcalculatingthefirstfivefrequenciesofatank/bargeonanelasticfoundation,asfollows:

-14-

(Ml,Iassumedconstant)

WaveEncounterFrequencyat10Knots

1 1 1400 500 600 700

BargeLength(Feet)

Figure4-1. TrendinFundamentalFrequenciesvs. BargeLength

-15-

where: fn=

An=

E=1=

‘1 =

frequency(cyclespersecond)

coefficientformodeoffrequency(table4-II)

modulusofelasticity@sf)momentofinertiaofbargeandtank(ft4)

()lb/see%massperunitlength—ft2

tank/bargelength(ft)

(4-2)

Table4-II. ValuesofAnforCalculatingFrequency

MODE SHAPE VALUE

1 22.0/ \

2

3

4

5

4.7 PRESSURELOAm

/

Pressureloadsonrefrigeratedcargotsnksgenerallyintamkdesign.However,pressurewillcausestressaccountedfor intheoverallevaluation.

61.7

121.0

200.0

298.2

willnotbeofmajorsignificancecomponentswhichmustbe

TheCodeofFederalRegulations,Title46,Chapter1, Support38.05-3(g),statesthat“Cargotanksinwhichthetemperatureismaintainedbelowthenormalat-mospherictemperaturebyrefrigerationorotheracceptablemeans,shallbede-signedforapressureofnotlessthan110percentofthevaporpressureoftheliquidatwhichthesystemismaintained.” Thisisthegeneralruleformostfluids;how-ever,inthecaseofammoni~whichisessentiallyatatmosphericpressureduringrefrigeratedtransport,theCodeinSubpart98.2510(d)statesthat25psigmustbeaddedtothetransportpressure.

Inadditiontothevaporpressure,thehydrostaticpressureshouldalsobeconsideredinthe.designoflargetanks.Ina4O-foottankcarrying42lb/ft3fluid,forexample,hydrostaticpressureonthebottomofthetankisabout11.6psig.

— — .

’16-

4.8 TEMPERATUREANDTHERMALLOADSTemperatureandthermalloadsareimportantconsiderationsinthedesignofre-frigeratedtnks. Thermalloadswillbeofmajorsignificanceif thetankscarryverylowtemperature(lessthan-150°F)gasessuchasliquifiednaturalgases,oxygen,andnitrogen.Highlyspecializeddesignsarerequiredforverylowtempera-tureapplications,andextensiveheattransferanalysesarerequiredtopredictthermalloadings.Thermalloadsonlowtemperature(above-15O°F)appli-cationsarenbtgenerillyaproblemifthetanksareproperlyinsulatedandiftheyaregraduallycooledduringloadingoperations.

Thedesigntemperatureforlowtemperatureapplicationsis moreimportantasabasisformaterialselectionthanforpredictionofthermalstresses.Thedesignapproachtothermalstressesshouldbetominimizethemthroughproperinsulationandbyinstallationofspraynozzlesorotherdevicestocoolthetankgradually.“Lowtemperature”steelsaresuitedtotarkswithambienttemperaturesdownto-150°F. Liquif’iedgaseshavingtemperaturesbelow-150”willrequirespecialdesignsforinsulationanduseofcryogenicsteelsoraluminum.Moreinformationonmaterialselectionforlowtemperaturesisgiveninsection7.2.

ThedesignorservicetemperatureusedintheselectionoftankmaterialmaybedeterminedbythemethodspecifiedintheCodeofFederalRegulations,Title46,Chapter1, Subpart38.5-2(b):

“(b) Theservicetemperatureistheminimumtemperatureatwhichthecargois loadedand/ortransportedinthecargotank.However,theservicetemperatureshallinnocasebetakenhigherthangivenbythefollowingformula:t5=t -0.25(tw- t~)wwhere: ts = servicetemperature

tw~ boilingtemperatureofgasatnormalworkingpressureoftankbutnothigherthan+320F

tB= boilingtemperatureofagasatatmosphericpressure.‘‘

“(d) Heattransmissionstudies,whererequired,shallassumetheminimumambienttemperaturesof 00Fstillairand320Fstillwater,andmaximumambienttemperaturesof1150Fstillairand900Fstillwater.”

4.9 COLLISIONLOADSTheCodeofFederalRegulations’requirementforcollisionshockloadsof1.5gappearstobereasonable.Figure4-2showsthestoppingdistance,bargevelocity,andstoppingtimesfor1.5g,assumingconstantdeceleration.Theserelationshipswereobtainedfromtheelementarytheoryofdynamics.Thestoppingdistancesandtimesappeartobeconservati~’einlightofthelargemomentumofloadedbargesandtheamountofdeformationcommonlyexperiencedinbargecollisionsorgrounding.

-17-

18.0

16.0

14.0

4.0

2.0

0 0.1 0.2 0.3 0.4 0.5 ().6TimtoStop(seconds)

Figure4-2. BargeVelocityandDistancevs. TimeforConstantDecelerationof1.5g

4.10 LOADSONTANKS/BARGESDUETOSLAMMINGShip6lamming,ingeneral,isaproblemareawheremuchinvestigationisbeingperformed.Thepurposeof includingthisbriefsummaryoftheproblemisonlytoindicatethenatureoftheproblemtoreadersunfamiliarwiththeproblem.Slammingintheareaofthebowmaycauselargeareasofdeformation.Ifthedeformationisneara saddle,damagetothetankmayoccur. Evidenceisavailablewhichindicatesthats~ammingofbargescancausesignificantstructuraldamage.However,theproblemofsurfaceshipslamminghasyettobecompletelysolved.CertiinresultsJus~ul‘n‘hecaseofflatbottombarges,areavailable.

Thefollowingparagraphspretierdseveralofthelatesttheoreticaltreatmentsoftheslammingproblem.Fromtheresultspresentedinfigures4-3and4-4, itisclearthattheorydoesnotpredictveryexactvalues.Figure4-4shouldbeindicativeofthemagnitudesofslammingloadsthatwillactuallybeencountered.

Inarecentpaper12Verhagenpresentedthefollowingexpressionform~imumimpactpressure:

P =Cpvcmax a

where: p = densityofwater(’bzc’ )V= relativevelocityofcraftwithrespecttowater(ft/see)

-18-Ca= speedofsoundinair(ft/see)

C= anundefinedconstant

TakingthevalueofCas1, theresultingpressures,basedonrelativevelocitiesfromtheshipmotionsprogram(table4-III),arepresentedinfigure4-3.

Assumingapocketofairbetweentheboatandwater,Verhagen’sexactexpressionformaximumpressureis:

-’m&.—8 pa

where: -Y’P=oCa =B=M=P =

()g[:(;)2MP+M]“C:V12gasconstant Pa =atmosphericpressure ‘1 ‘speedofsoundinair PI =beamofboatmassofboat hl =

densityofwater t =o

densityofairwatervelocityattopressureatto

heightofpocketatto

timewhenairpocketis sealed

Thisyieldsvaluessimilartothoseoffigure4-4.

-19-

1(WindSpeed– 30 Kn[}ts

100

t

Length(feet)

240440 (1 Tank)

340450

2 4 6 8 10 12BargeSpeed(knots)

500

400

.= 300‘a

200Figure4-4 Average1/10Highest

SlammingPeakPressure

100

0

Figure4-3. AcousticPressurevs. SpeedforVariousBarges

/t

/44oFeetI J 45oFeet/’

1 /’/’0“/,//////“— A,9 — 0° DeadRise

/“- ---- 3°DeadRise

(WindSpeed– 30 Knots)

1 12 4 6 8 10 12

BsrgeSpeed(knots)

-20-Table4-llL Average1/1OHighestRelativeVelocityatSlammingStation

(3O-KnotWind)

BARGELENGTH(FEET)SPEED 440 450 340 240(knots) (ft/see) (ft/see) (ft/see) (ft/see)

o 12.416 12.307 11.970 9.8813 14.942 15*002 15.293 13*7996 19.176 17.563 18.572 18.6529 23.954 19.962 21.486 23.742,

12 28.128 22.082 23.945 *28.301

ACOUSTICPRESSURE(Pmm= pwaterx Cairx vfps)

BARGELENGTH(FEET)SPEED 440 450 340 240(kaOts) (Psi) (psi) (psi) (psi)

o 192.27 190.59 185.37 153.023 231.39 232.32 236.83 213:696 296.96 271.98 287.61 288.859 370.95 309.13 332.73 367.67

12 435.59 341.96 370.81 438.27

Chuang13presents,fortheimpactpressureof“a 20-inchx 26.5-inchrigidflatbottombody:‘‘

‘1.4t/T ~t

P(t)= 0.72V2 e sin—T

where: 7 is impactdurationtime

SinceChuangwasunabletoscalethistolargerbodies,andsinceVerhagenfeelsadependenceonV2isnotgoodforallweights,thisexpressionisnotconsideredfurther.

—

-21-

Chuangalsocalculatedtheimpactdurationtime.

B— = O.045secforabarge(beam= 100feet)2Ca

Inreference13Chuangpresentsaseriesofequationsforthemaximumpressureduetoslamming.

P = 4.5 V for0°deadrisehullmaxP = 4.11 V1”6for3”deadrisehullmsx

Theseareplottedinfigure4-6.

NotethatiftheconstantCinVerhagen’sacousticpressurewerechosenas0.29ratherthan1, Verhagen’sexpres~ionwouldbeidenticaltotheaboveexpressionfor0°deadrise.

Slammingstudiesdoneonadestroyerindicatethattheactualslammingpressuresatthebowofabarge(wherethereisprobablysomedeadriseangle)probablyfallinthersngebetweenthesolidanddottedcurvesoffigure4-6.

Sincetestdatainthisareaisquitesparse,furtherworkonbargeslammingshouldbeundertaken.Anoutlineforanexperimentalprogramforthiseffortis inAppendixD.

-22-Section5

STRUCTURALDESIGN/ANALYSISOFTANKBARGES

Thestructuraldesignprocessforlargetanksmaybebrokendownintofourstraight-forwardstepsoncethecapacityandtankloadingshavebeenestablished.Thisprocessis essentiallythatwhichispresentlyusedforsmallerriverbargetanksanditmaybesummarizedasfollows:

1.

2*

3.

4.

Frompreviouslydeterminedtankdiameter,tanklength,pressure,weightanddynamicloadings,determinethebasictankconfiguration,i.e. , aneconomicalcombinationoftankwallthickness,numberofsupports,sndsupportspacing.Determineareaandmomentofinertiarequirementsforstiffenersandthick-nessofwearplatesduetoweightanddynamicreactions.Determinesupprtreactionforthegroundedconditioninthecaseofriverbargesand/orsaggingandhogginginthecaseofoceanbarges.Checklocalstressesintheareaofsaddlesduetogroundingand/orsagging/hoggingreactions.

If, atanypointinthedesign,thestressesexceedallowablelimits,scantlingsmaybeincreasedandthedesigncontinuedfromthatpoint.

Inthecaseofoceanbarges,thedesignershouldalsodetermineifthealternatingstressintensityisbelowtheendurancelimitforthepredictednumberofcyclesofhoggingsndsagging.

Indeterminingthebasicsizeofoceantank/barges,thefundamentalbendingfrequencyofthetsnk/bargestructureshouldbecalculatedaccordingtotheproceduregiveninsection4.6. Thefundamentalfrequencyofthecombinedbarge/tankstructureshouldbegreaterthantheforcingfrequencyofthewaves.ForcingfrequencymaybeestimatedusingthemethodgiveninAppendixA.

5.1 EXISTINGDESIGN/ANALYSISPROCEDURESTheU.S.CoastGuard2sndtheAmericanBureauofShipping6bothfurnishguidanceonthedesign/analysisoftankbarges.T~seguidelinesuse,asabasisfordeterminingstressesatsaddles,themethodof Zick. BothsourcesalsorefertothemethodofBrownellandYoung(reference14)whichis essentiallythesameasZick’smethod.Tankbargedesigners,almostwithoutexception,usetheCoastGuardproceduretoanalyzestressesintanks.Theauthor,inreviewingtheliterature,didnotuncoveranyotherdirectlyappl~5ablesimplemethodofanalyzingtankstressesintheareaofsaddlesupports.Rffren discussestheproblemofringstiffenersinmoredetailthanZickorBrownellandYoung,andthediscrepancieswhichoccurappeartobeminor.

TheCoastGuardandseveraldesignagentsutilizeacomputerprogramstocalculatesaddlereactionsofindependenttqnkbargesinthegroundedcondition.Usinganitera-tiveprocess,theprogramdeterminesthetotal(tankplusbarge)momentandthetankmomentateachsaddle,togetherwiththecorrespondingverticalreactions.(Theiterativeprocessmayalsobeperformedwiththeaidofadeskcalculator.) With

-23-

thetankmomentsandverticalreactionforcesknown,thestressesduetolongitudinalbending,circumferentialbending,directcompression,andtangentialshearmaybedeterminedusingthemethodofZick.ThestressesmaythenbecomparedtodesignallowablestressesasrecommendedbyZickorspecifiedintheCodeofFederalRemulations.

Anumericalmethodfordeterminingstressesinshellsofrevolutionsubjectedtoaxi-symmatricornon-symmetricreasure,bandloads,ringforces,andringmomentshasbeenpublishedbyKalnins.% ThismethodhasbeenprogramedforcomputationontheUNIVAC1107computer(reference16). Duringthestudythisprogramwasutilizedtoobtainacomprehendivestressdistributionintheareaofthesaddle.Stressesweracomputedontheinside,outside,andmid-surfaceofthetankwall.AdiscussionoftheanalysisandacomparisonwithresultsduetotheZickmethodaregiveninsection6.

Design/analysisofocean-goingbargessubjectedtosaggingandhoggingmaybeap-proachedinamannersimilartotheprocedureforgroundingcalculations.Specificcriteriaforcomparingalternatingsaggingsndhoggingstressestodesignallowable,basedonfatiguetheory,donotappeartoexist,nordoesaprecedentexistforincludingdynamicloadsinafatigueevaluation.Anapproachtothisproblemis discussedinsections4.5and6.4.

Experienceindesign/analysisofoceanbargesappearstobequitelimited.Inarecentsurveyoftheindustry(reference17),onlyonedesignforindependent,cylindricalocean-goingtank/bargeswasidentifieiLThisdesignwasforabargeofapproximately20,000tons,withtwotanks,eachabout30feetindiameterand300feetlong.Thetwotankshadoriginallybeendesignedtobecontinuouswithmultiplesupports.How-ever,thisdesignresultedinnegative(lift-off)forcesinthesaggingandhogginganalysisandthedesignwasmodifiedbyincreasingthenumberoftankstofour,eachsupportedononlytwosupports.

Asecondocean-goingtank/bargeconsistingofthreeintersectingcylinderswasre-portedtobeintheconstructionstagebutdesigndetailswerenotavailable.

&2 RATIONALEFORDETERM~NGT~ WALLT~CKNESSANDNUMBERANDSPACINGOFSUPPORTS

InEection4thetankandsaddleloadsduetotankandcargoweightaredeterminedfortanks200,300and400feetlongandwithconfigurationscontainingfrom2to11saddlesupports.Uniformcircularcross-section,uniformlyspacedsupports,hemis-phericalendenclosures,andanoveralllength/diameterratioof10wereassumedinthecalculationoftheseloads.Amethodwhichisanalogoustathatof Zickwasemployedtodeterminetherequiredtankthickness.Largetanksshouldbedesignedsothatthetankisreinforcedbycircularstiffeningringsplacedeitherdirectlyoveroradjacenttothesupports.5.2.1 CALCULATIONOFSTRESSES– Becauseofthelowinternalpressureassociatedwithrefrigeratedcargoes,thecircumferentialtensilestressinthetankisnotnecessarilythebasisfordeterminingther~uiredtankthickness.ThisisadeparturefromtheproblemwhichZickinvestigated.Themaximumcircumferentialtensilestressoccursatthebottomofthetankandis causedbyuniformintermilpressureplushydrostaticpressureduetotheweightofliquidenclosed.From

ClassicalMembraneisgivenby

(uh)~=

-24-ShellTheory,themaximumcircumferential(hoop)tensilestress

.mt (5-1)

whereP*is themaximuminternalpressure.Itisalsonecessarytoexamineotherprimarytankstressessothatthecriticalstressconditioncanbedetermined.BasedonClassicalBeamTheory,thelongitudinalbendingstressdistributioninthetank(at6’=O,180°asdefinedinfigure6-1)isgivenby -

(5-2)

Substitutionof themaximumbendingmoment,M*,whichoccursatthesupports,intoeq. 5-2yields

Thetransverseshearing

us(x)= ~7rrt

(5-3)m~t

stress distributioninthetank(atO=90°) isgivenby

(5-4)

Substitutionofthemaximumshearforce,V*,whichagainoccursatthesupports,intoeq. 5-4yields

(5-5)

5.2.2 ALLOWABLESTRESSLIMITS—Zickplacesthefollowinglimitsonprimarytankstresses:

a. oncircumferentialstress,theallowableworkingstressforthematerial.b. Onlongitudinaltensilestress,theallowableworkingstress.c. Onlongitudinalcompressivestress,thesmallerofone-halfyieldstressor

thevaluegivenby

(5-6)

whichaccordingtoZickis “basedupontheacceptedformulaforbucklingofshortsteelcylindricalcolumns.7‘

d. Onshear stress, 80percentoftheallowableworkingstress.

5.2.3 DETERMINATIONOFCONTROLLINGSTRESSMAGNITUDE- Basedonthelimitsonprimarytankstresses(section5.2.2), itwasfoundthatthelongitudinalcompressivestressis criticalwhendeterminingtankthickness.Figures5-1,5-2,and5-3showthevariationofmaximumlongitudinalbindingstresswiththicknessforeachofthetendifferentsupportconditionsfora 200-,300-,and400-foottankre-spectively.Superimposedoneachgraphis thevariationofallowablelongitudinalcompressivestresswiththickness(accordingb eq.5-6).

Withthehelpoftheseparametriccurves,typicaldesignsfora200-foot,300-foot,and400-foottankmaybedetermined.

-25-

&md&&::::; “C (InThouumhofHI)

12

11

10

9

8

7

6

5

4

3

1

1

0

—

—

—

—

L

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9Thickneu,t (lrmf@

Figure5-1, EendinEStreamVH.Thickn8ssforVarious200-FootTankConfigurations

CompressiveStress,aBendingStress,U. c (inThousandsofPSI)

12

11

10

9

8

7

6

5

4

3

2

1

0

Mm.AllowableCompressiveStress

\ 2SurmOrtu

I I I I I I 1 1

0.1 0.3 0.5 0.7 0.9 1.1 1,3 1.5 1.7 1.sThickness,t (Inches)

Figure5-2. BendingStressvs. ThicknessforVarious300-FootTankConfigurations

-27-

&#&s&~r~’ ‘c (inThousandsofPSI)‘0

12

11

10

9

8

7

6

5

4

3

1

1I

10U.i WA u.> U.* u. o w.0 ,.” 9

Thickness,T(Inches)

Figure6-S, BandingStrefmVU,Thiokne~sforVarious400-FootTankConfigurations

— —

-28-

5.2.4 DYNAMICU3ADS– Dynamicloadfactorsduetoshipmotions(heaving,pitching,etc.)aregiveninsection4. Thesefactorsreflectthemostseverecombina-tionofshipmotions.Toaccountforthesedynamicloads,thestressesinthetankduetocargoandtankweightmustbemultipliedbyoneplusthedynamicloadfactor.

Theuniforminternalpressureforwhichthetanksmustbedesignedisdiscussedinsection4.7.Thispressurecontributestauniformlongitudinaltensilestressinthetanks.Thevalueofthistensilestressisafunctionofthetankradius.Tocalculatethemaximumlongitudinalcompressivestresslevelitisnecessarytosubtractthisuniformtensilestressfromthebendingstress.

5.2.5 CHOICEOFSUPPORTCONFIGURATIONS—Afive-supportconfigurationwasarbitrarilyselectedasrepresentativeofthin-walledtankswithminimumsup-ports. ThisselectionconformswiththeAmericanBureauofShippingregulationswhichstatethatdistancesbetweensupportsshouldnotexceedtwicethetankdiameter.

Therequiredtankthicknessfora200-foot,300-foot,and400-foottankwasthendeterminedforafive-supportconfigurationbyadjustingthebendingstresscurvesandlocatingtheirintersectionwiththeallowablelongitudinalcompressivestresscurve;theintersectionspecifiestheminimumthicknessrequiredtomeetthestresslimit.

5.2.6 REQUIREDTHICKNESS—Thefollowingtatithicknessesweredeterminedforafive-supportconfiguration:

Configuration Thicknessa. 2oo-foottank O.2inchesb. 3Oo-foottank 0.4inchesc. 4oo-foottank O.65inches

Standardfabricationpracticefor20-footdiametertankscallsforawallthicknessof5/16inchorgreater.Itwasthereforearbitrarilydecidedthatforthe200-foottanktheminimumwallthicknesswouldbeincreasedto5/16inch,althoughtheoreticallythethicknesscouldhavebeenO.2inches.Utilizingthisdesign,foursupportswouldprovidesufficientstrength.

5.2.7 DETERMINATIONOFSTIFFENERSIZE– Designoftheringstiffenerswasbasedontheanalysisof180°arbitrarysaddlesupports(AppendixB). Thecir-cumferentialstressesinthestiffeneraregivenby

(’Q -0.48~ 0.043~ r.— -A I/c (5-7)

compress.+0.158Q +0.043~r

@o)tensile= A I/c

Zicksetsthefollowinglimitsonthecircumferentialstiffenerstress:a. orIb. on

.—

compressivestress,one-halfyieldstress;tensilestress,allowableworkingstress.

and

-29-

Atthispoint,itwasnecessaryh designateatypicalmaterialinordertadeterminetherequiredstiffenerstrength(basedonthepreviousstresslimits).Forthispurpose,carbonmanganesesiliconsteel,A516Gr65,waschosen.Itsmaterialstrengthpropertiesare65,000psiultimatestress,35,000psiyieldstress,and16,250psiallowableworkingstress.Knowingthelimitsoncircumferentialstressinthestiffener,therequiredcross-sectionalareaandsectionmodulusaredeterminedfromeq. 5-7.

Asummaryofthetankgeometryanddesignstressversusallowablestressesispresentedintable5-Iforeachofthethreetanksizesstudied.Figure5-4showsthethreerepresentativetankdesignsdrawntothesamescale.Thetankconfigurationsarerepresentativeofminimumrequirementsforthicknessandnumberofsupports.Flexibletanksareconsideredtobedesirabletoprevent“liftoff”ingroundingandsagging/hoggingconditions.

~ 200’u u u u 120”

K- 612”+(a) ‘

180”

(b)

~ 400’r

t-- “’”~ “’”--l 240”

Figure5-4.

(c)

ThreeRepresentativeTankDesigns

-30-

Table5-L Summaryof TankCmrfigurationsSelectedforAnalysisof GroundingandSagging/Hogging

OVERALLTANKLENGTH200FT 300FT 400FT

TankRadiusNo.ofSupportsSupportSpacing~lidthofSupportsTankThicknessStiffenerSect.Mod.StiffenerCross-Seet. Area

Circum.TankStressDesignAllowable

Long. Tens.TankStressDesignAllowable

Long. !Xnnp.TankStressDesignAllowable

120in.4

612in.12in.

5/16in.390in.3

170in.2

6,900psi16,250

5,60016,250

1,7604,730

TransverseShearTankStressDesign 4,350Allowable 13,000

Circ.Tens.Stiff.StressDesign 13,100Allowable 16,250

Circ.Comp.Stiff.StressDesign 16,900Allowable 17,500

180in,.5

725in.18in..4 in.

1875in.3

450in.2

10,600psi16,250

6,87016,250

2,3704,130

6,89013,000

13,30016,250

17,30017,500

240in.5

968in.25in.

.65in.6670in.3

1000in.2

11,100psi16,250

7,60016,250

3,9104,920

8,55013,000

13,30016,250

18,500*17,500

*Exceedsallowableinactualdesignstiffenersizewouldbeincreased.

-31-

5.3 DESI~,NF(3RBUCKLINGInbetweenstiffeners,thecylindricalshellis subjectedtolargecompressivebendingstressduetothenatureoftheloading.Considerationmustbegiven,therefore,tothepossibilityoftheshellbuckling.Thecriticalcompressivestress,u , hasbeencommonlyacceptedasbeing1.3timesthecompressivebucklingstr& (duetouniform

18 Fl~gge’sresultwasaxialcompression).SuchavaluewasobtainedbyFlfigge.foraparticularshellandbucklegeometryandisnotgenerallytrue,asis showninreference19. Theresultsofthisstudyshowedthatthecriticiiaxialcompressivestressduetobendingisnotmorethan10percentgreaterthanthecriticalstressforalongshellunderuniformaxialcompression,unlesstheshellisextremelyshort(lL/r< 0.15).Forrelativelylargelength-to-radiusratios(L/r)andradii-to-thicknessratios(r/t), reference19showsthat:

“Cr‘e) ; (5-8)

Theseresultsshowthatlinearbucklingofacircularcylindricalshellduetoasymmetric(non-uniform)axialcompressivestressdistributionwillalwaysoccurataloadlevelwherethemaximumlocalaxialcompressivestressequalstheuniformaxialcompressivestressforbuckling.

Sincethepresentstudyisdirectedtoverythincylindricalshells,initialdeviationfromtheidealcylindricalsurfaceshouldbeconsidered.(Thesemaycausebucklinatastresslevellowerthanthetheoreticalelasticbucklingstress.) Timoshenko2%presentsanempiricalformulaforcalculatingtheultimatestrengthofcylindricalshellsunderaxialcompressionwhichconsiderstheeffectofinitialimperfections.Thisformulaisgivenas:

‘Ult [1t-7 r0.6 ~-10 ~E

1+0.004~uYP

(5-9)

whereu istheyieldstrengthofthematerial.YP

Considera400-foottankwitha 20-footradius,m 80-footsp~ betweensupports,andashellthicknessofO.65inch.Applyingeq. 5-8andeq.5-9yieldsthefollowing

Assume E = 30x 106psi, v= O.3,andtr =35,000psiYPTheoreticalelasticbucklingis:

30x 106 0.65

“r= [~(’:@d“2= “4“200psi

—

-32-

FromTimoshenko(reference20):

= 30X106‘Ult

“Ult= lo,700psi

-7 ,240—-lo ()G5

1+.004()

30x 10635X103

Othertest results, givenin reference21,haveyieldedresultssimilartothoseob-tainedbyTimoshenko.Figure5-5,obtainedfromreference21,showsanondimen-sionalplotofthetheoreticalelasticbucklingcurveandanempiricalcurvebasedontestdata.AscanbBseeninfigure5-5,cylinderswithaslendernessparameter,

hr ofO.064orlesscanbestressedtotheiryieldstresswithoutbucklingE ~’

whereascylinderswithlargerslendernessparameterswillbuckleatlowerstresses.

1.0

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

;y,::: y+

~- O.16+0.20\ YPt

‘YP

\(Experimental)r

\

\

\

I I 1 I I I I I 10.1 0.2 0.3 0.4 0.5 0.6 0.7 0.S 0.9 1.0

aNondimensionalSlendernessParameter~ ~Et

Figure5-5. NondimensionalBucklingCurveforCircularTubesinCompression

—. -. .— — —

-33-

Applyingthiscurvetoourdesigncase,weobtain:

h L . 357000Psi ~.~oin.Et 6 0.65in. = 0.43

30x 10 psi

Andsince

Fcr E t—=.16—– r‘YP ‘YP

then

u = 13,000psiCr

Timoshenko’sresultsappearh givealowervalueofcriticalbucklingstressforthiscaseand,therefore,hisistherecommendedproceduretofollow.Ontheotherhand,applyingZick’scriticalbucklingstressformulatoourdesigncaseyieldsacriticalcompressivestressof4,920psi. Thisnumberappearstobehalfof thevaluegivenbyTirnoshenko’sempiricalequation.

5.4 REACTIONLOADSDUETOGROUNDINGSAGGINGANDHOGGINGInsection5.2therationaleforselectingwallthicknessandthenumberandspacingofsupportsforhypotheticaltanks200,300,and400feetlongwaspresented.Inthissection,thehypotheticaldesign/analysisprocedurewillbecontinued.

Itwasconcludedinsection4.3thatgroundingisthemostsevereconditionforriverbargesandthatsaggingandhoggingloadsaremostcriticalinoceanbarges(section4.4). However,inordertoanalyzetheseconditions,theW bargestructuremustfirstbeanalyzedasawhole.Itwasnecessarythereforetopreparepreliminarydesignsof bargestothepoint,whereweight,buoyancyandoverallbendingstrengthmaybedetermined.Forexpediency,the200-foottankandacorrespondingbargewereselectedforthegroundingcalculations,andthe400-foottankandcorrespondingbargewereselectedforthesagging/hogginganalysis.BothconditionswereanalyzedusingtheCoastGuardprocedureaidedbyadeskcalculator.

5.4.1 CONFIG~ATIONANDCHARACT’EHSTICSOFBARGES– Thegeneralcon-figurationandcharacteristicsofriverandoceanbargesareshowninfigures5-6and5-7.Typicalsectionalviewsforthepurposeofdeterminingbendingstrengthareshowninfigures5-8and5-9.Thetankdimensionsareshowninfigure5-10.

5.4.2 RESULTSOFGROUNDINGCALCULATION—The200-foottankconfiguration,loadingdiagram,shearcurve,andmomentcurvesareshowninfigure5-11.Asum-maryofforcesandmomentsactingonthesaddlesisgivenintable5-II.

5.4.3 RESULTSOFgiveloading,shear,

-34-

SAGGING/HOGGINGCALCULATION– Figures5-12and5-13andmomentcurvesforsaggingandhoggingofthe400-foottank.

Tables5-IIIand5-IVsummarizetheforcesandmomentsac~ingatthesaddles.

‘[ I 1I I )1I\% I i

I .==~===.–.–_=* ======~====== ==<==<’III

II

I ‘]II /1

I “k.-.— — ———— _ _____ __ ———— Y i–-’ ILOA– 270’

MlLoadDisplacement–3,000LTons

//”––––––––– –––-–-–-–– ––––---.\

———— ——

Figure5-6. ConfigurationofaTypicalRiverBarge

T (jT’,/ ‘..\1’ 1

1’96’ ‘\, ,/’-.-L.—-—L.—----- ----- ---- .—-———---- ------ -- .-—-—————---...—.., ...-—..——.————— .—,----- .---- ----- ------ —-—-————-——------ —---- ----- --- —-. “%

/ \\1’

L 1

iI!, ,;‘.FullLoadDisplacement–21,050L.Tons

\—--———IL / ;---- \\38’

+ ,~-’~--

\.18’ .~’-90’4 LOA–459’ ——..

LWL-450

Figure5-7. Configurationofa TypicalOcean-GoingTankBarge

-35-

1

t’

~,\\

I.—+—.

Ek 1

[Eu

u LLl IA w u LLlul IA Lu0000 00000

BargeStructure:Momentof Inertia: 39,850in.?x ft2SectionModulus: 7,520in.z ft

Figure5-b. ‘lypjc~]RiverBargeSectionalViewI

+.— ——

I

BargeStruct~e:MomentofInertia: 732,813in.2xft2SectionModulw: 30,933in.2xft

Figure5-9. T~icalOcean-Going”BargeSectionalView

—

’36-

OCEANTANK/BARGEEquivalentLength

3884”- ——————+

.

Saddle L- ’06, ---L-- ,O.,,–-l- ,..6,+ ,0.,,-’lSpucing .—.—— .— LOA400’ /

MomentofInertiaforOneTank: I=196,036in.2xft2WeightTWOTankx 1260L.Tons

RIVERTANK/BARGE

MomentofInertiaforOneTank I=11,808in.2xft2WeightTwoTank% 260L.Tons

#

—

Figure5-10. TypicalTankCharacteristicsandSaddleSupports

-37-

WeightCuwe16128404812162024

400300200100

0100200300400500600‘looSoo900

40,000

30,000

20,000

10,000

Stern 270’

I 34.510

Figure5-11.

i 2 j 4NumberofSaddles

TypicalLoading,Shear, andMomentaGroundedRiverBarge

IGroundingForce

.r50’—

Stem

— 5Q’—

Diagramsfor

-38-

Table5-II. SupportLoadSummaryandMomentDistribution

BargeGroundedatFwdRake CALCULATIONSFORS.WATERBhd.TwoTanks.S. Water35Ft3/Ton. TANKS&13W LOADCONDITIONSSaddleSupportsI thru4 AS.UNIFORM

13JSTRIBUT.1st 2nd 3rd 4thAllTons@2240lbs(L. Tons.) LOAD Distrib.ofTanks&DWasSaddleLoads

ForcesAtSaddleSupportNo.(L. Tons) —

ThesearethereactionAFT1forcesatthesaddlesforfullloadintwo20ft. dia.tanks;nomi- 2nallength200feet.

3

FWD4

MomentsAtSaddleSupport~(mx Tons)

Momentsshownare37%ofthetotalbendingmoment.(ftx tons) AFT1

2Momentsatsaddles1&4 3arethesameandcon- FWD4stantduetoconstant

450267m381506m142298m

298141G5072563392y&96773

2981215527251m397600G48

298G

298112G536251G397609

lG39

298G

3,500 3,500 3,500 3,5009,900 12,83013,900 14,300

12,800 16,300 17,600 18,0503,500 3,500 3,500 3,500

cantileverdesignoftankendg.

-39-

i80 DisplacementCUWG60 WeightCurve

40---

\T20 4691

‘~ * &3.24---

b DifferenceCuWe $ 6 ; ‘%.~~j~”;- ! Id1% I ~ +-4’‘arge

1s rm-! 1●“,-0

0A.P

5

10

15

20--

/(,ShemCurve

5 6 ‘1 8 9PP.

1’285,592

99,957

h

1 1 I 1

[ #1+ 80.6’+80.6’+ : 80!’6’

8 9Saddle #2

80.6 :.;.#3 #4 *<

Binge --l

Figure5-12. TypicalLoading,Shear,andMomentDiagramsforanOcean-~i~TankBarge@aggingCondition)

Tanks

-. —

-40-

100L DisplacementCurve80-60-

z& ~o 1-~ 20-:

0

-‘%A.P.41 2 3

L -Saddle #l” #2.-——..——.- ..-—-—~EquivalentLength388.4Feet

18t 1600.1

200,00‘T~8,447

~10F.P.

\

E LL-‘,65,741 65,956

:mG.- 56,>u 34,803t? 19.161 , ~I~–-~

,,:--- ...+...8 ..74‘0“p”#3IiquivalenlLength386.4Feet

Figure5-13. TypicalLoatig,Shear,andMomentDiagramsforanOcean-GoingTankBarge (HoggingCondition)

——-.

-41-

Table5-III.SupportLoadandMomentSummary– SaggingCondition

TwoTanksS.Water35Ft3/Ton.

CALCULATIONSFORS.WATERLOADCONDITIONS

SaddleSupports:1thru5. TANKS&DW 1st 2nd 3rd 4thAllTons@2240Lbs AS.UNIFORM(L. Tons) LOAD Distrib.ofTanks&DWasSaddleLoads

ForcesAtSaddleSupportNo.(L. Tons)

16553198

ThesearethereactionAFT1forcesatthesaddlesforfullloadintwo40-ftDtanks,nominallength400ft. 2

3

4

FW135

MomentsAtSaddleSupportNo.(FtXTonS) —

Momentsshownare35%ofthetotalbendingmo-ments(ftx tons) AFT1

2Momentsforsupports 31&5areconstantduetocantileverdesignat 4endsoftanks. AFT5

4853844

238532291657165733142385844

3229319816554853

/

165528874542115522643419177817783556226411553419288716554542

1655300446591038230833461734173434682308103833463004

23,501 23,501 23,50171,400 46,329 55,697

100,000 65,934 78,86171,400 46,329 55,69723,501 23,501 23,501

-42-

Table5-IV.SupportLoadandMomentSummary—HoggingCondition

TwoTanksS. Water35Ft3/Ton. LOADCONDITIONSSaddleSupports:1thru5. TANKS&DW 1st 2ndAllTons@2240Lbs AS.UNIFORM 3rd 4th(L. Tons) LOAD Distrib.ofTanks&DWasSaddleForces

ForcesAtSaddleSupportNo.(L. Tons)

16551816

ThesearethereactionAFT1 347’1forcesatthesaddlesforfullloadintwo40-ft.D. 2226tanks,nominallength 1699400ft. 2 3925

23432243468616992226392518161655

FWD5 3471

16552109G193318093742223322334466180919333742210916553764

MomentsAtSaddleSupport~/

//

1655188635412156175239082290~458017522156390818861655E

/

Momentsshownare35% /ofthetotalbendingmo-ments. AFT1 23,501

2 40,000Momentsforsupprts 3 65,9561&5areconstantduetocantileverdesignat 4 40,000endsof tanks. FWD5 23,501

23,501 23,50126,405 34,38243,456 56,10226,405 34,38223,501 23,501

-43-

Section6

EVALUATIONOFSTRESSESINEXISTINGANDPROJECTEDDESIGNS

Insection5therationalefordesignoftypicalindependenttanksforriverandoceanbargeapplicationswasdiscussed.TWOhypotheticaldesigns,basedonasimplifieddesign/analysisapproach,wereselectedformoredetailedinvestigationofstressesinthetanks.

Theobjectivesofworkdescribedinthissectionaretoevaluatethesimplifieddesign/analysistechniqueswithrespecttotheaccuracyofpredictingstressesduetocriticalloadsandtoexaminethevalidityofanindividualstreBscriteriaapproach.Toac-complishthesegoals,therepresentativedesignsweresubjectedtocomputeranalysisusinglinear,thinshelltheoryapplicabletonon-symmetricallyloadedshellsofrevolution.Acomputerprogramwasusedforthisanalysis.Atypicalstructuremodelusedforthecomputeranalysisisshowninfigure6-1. Themodelrepresentsa theoreticalconfigurationandisnotintendedasapracticalconfigurationforpurposes

t----x UniformAx3alForceduetoUniform

ment

SaddleReactionLoadonRing

Figure6-1. TypicalStructuralModelforComputerAnalysis

—

-44-

ofconstruction.Themodeldoesrepresenttheproperareaandmomentof inertiaforpurposesofanalysis.Atypicalconfigurationisshowninfigure6-2whichismoreadaptabletofabricationbutwhichhasthesamegeneralpropertiesasthetheoreticalmodel.

T—36’”

SectionMoclulusE:6,940in.3

-., - SaddIeIn8ulatingMaterial

Figure6-2. TypicalTankReinforcementAdaptabletoFabrication

6.1 STRESSANALYSISOF200-,300-,and400-FOOTTANKCONFIGURATIONS6.1.1 ANALYSISOF200-FOOTTANKCONFIGURATION– A200-foottankisrepresentativeofthesizepresentlybeingusedoninlandwaterways.Themostcriticalloadingconditionencounteredisgrounding.However,groundingoccursinfrequentlycomparedtothelengthoftirneinnormaloperation.Thereforethemaximumstressesencounteredinthegroundedconditionneednotbelimitedtotheallowableworkingstress(whichisapproximatelyonefourthofthedtimatestressfortherangeofmaterialspresentlyusedintaukconstruction),

Thus,the200-foottankwasanalyzedfortheloadconditionsofno~maloperation(includingadynamicloadfactor~forwhichitwasdesignedandfortheloadconditionsofgrounding.Thissecondanalysiswasperformedtoindicatewhetherornottheuseofconservativestressallowableinthedesignprocedurecancompensatefortheincreasedstresslevelswhichtypicallyoccurinthegroundedcondition.

Thecomputermodelconsistsofasectionofthetankequaltohalfthelengthbetweensup~rtsoneithersideofastiffeningring.The180°saddlesupportis replacedbytheassumedreactionloaddistribution(giveninAppendixA)actingonthestiffeningring.

Theshellis loadedbyuniforminternalpressureandhydrostaticpressureduetotheweightofcontainedliquid,Theboundaries,whichareatmidspan,arenotaffected

—

-45-

bythelocalbendingstressatthestiffeningrings(supports).Therefore,amembranestateofstresswasassumedinthetankattheselocations.Ithasbeenshown,18however,thatthismembranestateofstresscanbeconservativelydeterminedbyabeamanalogy,i.e. , analysisofthecylindricaltankandliquidasabeamsupportingadistributedload.Thelongitudinalnormalstressdistributionvariesaboutthecir-cumferenceasthecos 0.Thenetresultofthisstressis thelongitudinalbendingmoment.Thein-planeshearstressdistributionvariesaboutthecircumferenceasthesin0.Thenetresultofthisstressis thetransverseshearforce.Throughthebeamanalogy,theappropriatestressboundaryconditionsatmidspanforthecomp-uter modelareobtained.

Themostcriticalstressregion(andalsotheregionmostcrudelyanalyzedinpre-viouswork)isatthestiffeningringsoverthesupports.

Highlocalbendingstressesshouldbeexpectedinthisregionbecauseofthehighcon-centrationofloadappliedtothetankbythesupportsandthelargechangeinstiffnessfromthetanktothestiffeningring.Toreducethislocalstress,thetankwallwasreinforcedonbothsidesofthestiffener;atthestiffener,thethicknessofthetankwastripled.~hethicknesswasthentaperedoverasix-inchlengthtothenormaltankthickness.Thesizeandshapeofthereinforcementarebasedonasmallparametricstudyperformedonthe400-foottankmodelwhichispresentedundertheanalysisofthe400-foottank.

Thetwoloadconditionsanalyzedareshownintable6-I. Forthegroundingcondition,theloaddistributionisnotsymmetric,therefore,onlythemostseverelyloadedsup-portwasanalyzed.Fornormaloperation,eachsupportis loadedapproximatelythesameamountandtheassumptionofloadsymmetryabouteachsupportwasmadeinthedeterminationofloads.Therefore,theanalysisappliestoallof thesupportsfortheconditionofnormaloperation.

Thelongitudinalandcircumferentialstressdistributionsontheinnerandoutershellsurfacesat@=O,intheregionofthestiffeningring(support),arepresentedinfigures6-3through6-6forthenormalandgroundedconditions.Highlocalbendingstressesattheshell-stiffenerintersectionarepresentinboththelongitudinalandcircumferentialdirections.Table6-11liststhevariationoflongitudinalnormal,circumferentialnormal,andin-planeshearstressatselectedpointsaroundthecircumferenceintheshellandinthestiffeningringforbothloadconditions.Thecomparisonofresultsofthesimplifieddesign/analysistechniqueandthemoresophisticatedcomputerapproachisgiveninsection6-2.

6.1.2 ANALYSISOF300-FOOTTANKCONFIGURATION—The300-foottankalsowasanalyzedfortheconditionofnormaloperation(plusdynamicloadfactor).Theanalyticprocedureis thesameasthatemployedforthe200-foottank.Thestressdistributionandlocationsofcriticalareasaresimilartothoseobtainedforthe200-foottank.Becausenoadditionalconclusionscanbedrawnfromtheanalysis,thedetailsareomitted.

-46-

LoadCondition

Table6-L AppliedLoadsForNormalOperatingCondition(PlusDywmicLoadFactor)

200I?TTANK

PressureDist. InTanka. Uniformb. Hydrostatic

Normal(PlusI)yn.LoadFactor)

10psi3.96 (1+cos 0)psi

SaddleReactionLoada. -90°~ e ~90° -427.6COS 8 I)d

b. 900<0 <270° 0

B.C.s. AtMidSpan

Left Right Symmetryof Loading

AxialForceResultantdueto uniform 600lb/in.pressureactingonendclosures 749cos @lb/in.

AxialForceResultantduetoBendingMomentacrosssection o

In-PlaneShearForceResulkmtduetoLiquidWeight

PressureDistributioninTanka. Uniformb. Hydrostatic

SaddleReactionLoada. -90°~ e ~ 900b. 90°<0 <270°

400I?TTANK

10pai

10.03(1+cos 0)psi

-966cos19psio

Grounded

10psi3.23(1+00SO)psi

456.8cosOpsio

BoundaryConditionsatMid-SpanLeftEnd Ri~htEnd SymmetryofLoadingAial ForceResukntduetouniform 1200lb/inpressureactingonendclosures

AxialForceResultantduetoBending 2840cos.9lb/in.Momentacrosssection

NoSymmetryofLoading600lb/in. 600lb/in.

-1131cos8lb/in. -1598cos9lb/in.

zOOsinOlb/iu. -794sinolb/in.

In-PlaneShearForceResultantdue otoLiquidWeight

-47-

rl

! 1

10

5

0

-5

-10

-15

CItii

IIHI !III I‘xi“Xo

-20

Figure6-3. 200-FootTank—NormalOPeration—Longitudinal“Stress(o=0).

-3Crxlo 44” ,,h 12” 44” t— 6J

15r10

5

0

-5

-10I

-15

-20t

iI

%o

II %0IIII

Flgur@8-4,200-FootTank- NormalOperation- HoopStress(O=0)

— —-

-48-

1 I““0-3-+”+-1 I10

5

0

-5

-10

-15

-20

Figure6-5. 200-FootTank—Grounding– LongitudinalStress(6=O)

44” 44”ux1o-3- A 12” ~1 1—

15I

10

5

0

-5

-lo

-15

p, u. IUoi I I

‘(90

II IIIII‘% 1d

-20

Figure6-6.dOO-FootTank—Groundiug- HoopStress{O=0)

—— — — --J

-49-

Table6-II. MaximumStressesin ZOO-FOOLTmkBasedonComputerAnalysisNORMALOPERAIION GROUNDED

* IJ=fi e=90° 0=180°(9=27008:-0 e=90° 0 180°0.:270°

Max.Long. Stressesin iShell(atintersectionwith mstiffeningring) o

Mu.CircumferentialStressesinShell(away mfromshell-ringinter-sectionnegligiblebending)

Max.ShearStressinShell(atshell-ring mintersectionnegligibletwisting)

N1.H.Circumferential iStressinStiffeningRing m(otherstressesnotcritical)o

6,700 -M,520-1,720 1,925-8,050-3,180

7,220 5,550

0 4,400

11,870-16,630-2,500-1,300

-16,95014,050

6,190 4,520 5,500 5,183 14,120 5,1835,550 1,925-lo,500 1,!)2014,400 1,9204,910 -3,180-18,300-3,100+15,600-3,100

3,840 5,550 6,750 5,270 3,780 5,270

0 4,400 0 4,750 0 4,750

15,270 -16,63012,800-17,60916,300-17,6001,150 -1,300-2,900-1,400 1,350-1,400

-13,000 14,050-18,30014,900-13,70014,900

i- insiclesurfacem middlesurfaceu outsidesurface

6.1.3 ANALYSISOF400-FOOTTANKCONFIGURATION– Chemicaltanksaalargeas400feetinlengthand40feetindiameterhavebeenenvisionedasefficientcarriersforthelong-haultransportationofchemicals.Becauseoftheincreaseinsizeandtheeffectoftheopen-seaenvironment,thetechniquespresentlyemployedindesignofsuchtanksaresuhjecttomuchscrutiny.

The400-foottankwassnalyzedfortheconditionofnormaloperation(plusdynamicloadfactor)followingtheapproachusedforthe200-foottank.Thefirstcomputermodelofthe400-foottankhadnoreinforcementattheshell-ringintersection.Locallongitudinalbendingstresswasnearlyequaltatheyieldstress;successivereinforce-mentgeometriesloweredthisstressto14,000psi.Thelocallongitudinalbendingstressforvariousreinforcementgeometriesispresentedintable6-III.Theappliedloadsonthe400-foottankundernormaloperatingconditions,includingthedynamicloadfactor,arepresentedintable6-I.Thelongitudinalsndcircumferentialstressdistributionsonthei~er andoutershellsurfacesat0=Ointheshell-ringintersectionregionarepresentedinfigures6-7and6-8.Again,highlocalbendingstressesinthelorfgitudinalandcircumferentialdirectionsarepresentattheshell-ring,intersection.Thevariationsinlongitudinal,circumferential,andin-planeshearstressesatselectedpointsonthecircumferenceintheshellandinthestiffeningringarepresentedintable6-Iv.

Theeffectofopm-seaenvironmentis averyimportantfactorindeterminingsafestresslevelsinthetankbecausethetank-bargesystemis subjectedtocyclicloadingduetothenatureofwavemotion.Thealternatingconditionsofsaggingandhoggingresultinwidefluctuationsindeformationateachmintinthetank.Becauseofthis

-50”Table6-III.LocalLongitudinalBendingStressatShell-StiffenerIntersection,

400-FootTank(NormalLoads)PEAK PEAKINSIDE OUTSIDE

FtEINFORCE~NT STRESS STRESS

None

Constant(1O“Lon~1.3“Thick)

Taper (10”Long;2. O“to .65” Thick)

Taper(12.5“Lon&1.95”to .65”Thick)

+27,000

+19,000

+11,000

+10,000

-34*000

-22,500

-18,500

-13,500

cyclicloading,itisnecessarytoconsiderthefatiguepropertiesoftankmaterialsand’todete~mlne,byanalysis,thestressfluctuationsduringoneloadcycleinordertodesignforthedesiredtanklife.Afatigueanalysisofthe400-foottank,basedonstressfluctuationsduetosaggingandhoggingcycles,is illustratedinsection6.4.

6.2 COMPARISONOFDESIGNTECHNIQUESTOCOMPUTERANALYSISINTHEPREDICTIONOFSTRESSLEVELSINTHETANK

6.2.1 200-FOOTTANKCONFIGURATION– The200-foottankstresslevelspre-dictedby”thesimplifiedapproach,accordingtothesimplifieddesignanalysispro-

&ce ureof Zick,aregivenintable5-I.Stresslevelswerealsoobtainedby”computeranalysisforthe200-foottank.Resultsofthecomputeranalysisarepresentedintable6-II. Thesimplifieddesign/analysistechniqueusedtodeterminestresslevelsin thetankis basedonClassicalBeamTheoryandClassicalMembraneShellTheory.Thestress variation,throughthe taukthickness,is notaccountedfor inthis approach.Onlythestress levelat themid-surfaceofthe shell is predicted.Stressdistributionthroughthetankthicknessis assumedconstant.Thecomputeranalysisis basedonClassicalThinShellBendingTheorywhichpermitsstress variationthroughtheshell

-51-

Table6-N. MaximumStressesin40&FootTankBasedonComputerAnalysis

NORMALOPERATINGCONDITION

* 9=0 0==90° e=180° 0=270°

Max..LongitudinalStressinShell-(atshell-ringintersection)

Max.CircumferentialStressinShell(awayfromshell-ringintersection;negligiblebending)

Max.ShearStressinShell(atshell-ringintersection;negligible.twist)

Max.CircumferentialStressinStiffeningRing(mostcriticalstress)

i 9,900m -3,9500 -13,800

m 11,700

m o

i 10,700m -3,7500 -18,400

6,300 6,9501,850 7,550

-4,600 8,200

7,700 3,700

8,800 0

-17,300 15,500-1,550 1,55013,500-12,500

6,3001,850

-4,600

7,700

8,800

-17,300-1,10013,500

*i =insidesurfacem=middlesurfaceo =outsidesurface

— —

-52-

15

10

5

0

-5

-10

-15

-20

‘-’e-

Figure6-7.

15

10

5

0

-5

-lo

-15

-20

Ufi

oXo

II II II I

Ofi I I

aXo Uxo

400-FootTank—NormalOperation—LongitudinalStress(0=0)

--=+=+2’’-b-=-% o II %0

iIIIII

iIIII’00 ,

400-FootTank- NormalOperation– HoopStress(O=0)

— — —

“53-thickness.Thus,localshellbendingeffects(asopposedtooverallbeamtypebending)intheshell-ringintersectionregioncanbedetermined,Examinationoftable6-11illustratesthisfact.

Table6-Vshowsthat thereis verygoodagreementfor themid-surfacestresses.Therefore,it is concludedthat thesimplifieddesigntechniquesare adequateto de-terminethemid-surfacestress levelsin thetank.Sincetheeffectoflocalshellbendingis presentonlyin theshell-ringintersection(support)region,thesimplifieddesigntechniquescanaccuratelypredictthestress levelsthroughoutthetankexceptin thesupportregions. Fromfigures6-3and6-4, wesee that theappreciableeffectsoflocalshellbendingextendonlyabout30inches oneither sideofthe supportregion(fromthemiddleof thestiffeningring). Therefore,approximatelyonetenthofthe612-inchspanfrommid-supportto mid-supportexperienceslocalshellbending.

In thecomputermodel,the tankwallwasreinforcedat theintersectionwiththestiffeningring. Thismodificationwasmadebecauseoftheextremelyhighlocalstresses ontheinsideandoutsidesurfacesofthetankwall. Nowherein thesimplifieddesign/analysisprocedurewasthis stress effectaccountedfor. Locallongitudinalbendingoftheshellis dueto themismatchin stiffnessbetweenthestiffeningringandshellandalsoduetothehighconcentrationofloadactingonthestiffeningringfromthesaddlesupport.

Table6-V. ComparisonofMaximumMid-SurfaceStresses

TYPEOFSTRESS DESIGNPREDICTIONCOMPUTERANALYSIS

Circum.Stress 6,900psi 7,220psiLong.TensileStress 5,600psi 5,550psiLong.Compress.Stress -1,750psi -1,720psiShearStress

Theanalysispresented

4,350psi 4,400pSi

inAppendixBfordeterminingthesizeofthestiffeningringrequiredh carry thesupportloadis basedonClassicalThinShellBendingTheory,simplifiedtathecaseofaring.Thisanalysisneglectsanystiffnesscontributionduetotheattachedshellandshouldthereforebeconservative.It doesaccountforstressvariationthroughtheringthickness.Zick’sequationforthestressinringstiffenersisexactinform,butdiffersinthenumerical~oefficients;UckgivesCo-efficientsforthecasesof120°aud150°saddles.h thecomputerrndel, in orderhminimizetheeffectofcircumferentialbendingofthestiffeningringontheshell, thestiffeningringwassymmetricallyplacedabouttheshellmid-surface,However?indoingthis,thestiffnesscontributionoftheshellinresistingthesupportloadisalsominimized.Therefore,theringalonemustresistthesupportload”;thisisalsotheassumptionmadeinAppendixB.Thedesignvaluesofcircumferentialstressinthestiffeningringwerecomparedwiththoseobtainedbycomputeranalysisandtheresultsconfirmedthiseffect(table6-VI).

— ———

-54-

Table6-VI. MaximumCircumferentialStressinStiffeningRing

DESIGNVALUECOMPUTERVALUE

CompressiveStress -16,90013Si -16,950PsiTensileStress 15,200psi 15,270psi

Again,forcircumferentialstressinthering,exaellentagreementlashuwnbetweenthesimplifieddesign/analysispracedureandthemoresophistioatqdaomputeranalysis.

6.2.2 300-FOOTAND400-FOOTTANKCONFIGURATIONS- Acomparisonofdesignstressvalueswiththoseobtainedfromcomputeranalysisforthe300-footand400-foottankssubstantiatestheconclusionsreachedincomparisonsforthe200-foottank.Forthe400-footconfiguration,acomparisonoftable5-Iwithtable6-IVshowsgodagreementinstressvaluesatpointswherestresscanbeobtainedbythesimplifiedapproach.

Todemonstratetheadvantageofpositioningthestiffeningringssymmetricallywithrespecttotheshellmid-surface,acomputeranalysisofthe400-foottankconfigura-tionwithinternalstiffeningringswasperformed.Thisimbalsnaeofstiffnessabouttheshellmid-surfaceinductihighlocallongitudinalandcircwnfarantialbendingstressesintheshell.Themsximumabsolutestressincreasedfrom14,000psito23,000psi.Inaddition,tbeexpecteddecreaseinstressinthestiffeningringsduetotheassistanceoftheshellinrea@tingcircumferentialbqndingwasminor- approxi-mately5percent.

6.3 ALLOWABLESTRESSCRITERIAForthepurposeofdesigningtypicalchemicaltanks,theindividualstresslimitsgivenbyZickwereemployed.Thecqmputeranalysiswhichwasperformedpresentsamorecomprehensivelookatthevariousstresslevelsatallpointsinthetankmodel.

Basedontheresults ofthis analysis, it appearsunnecessarytotransformthestressstate at eachpointin theshell toprincipalstresses alongprincipaldirections.Exam-inationof tables6-11and6-IVindicatesthatat0= O,180°,thelongitudinalstressis maximumwhiletheshearstress is zero. Themaximumcircumferentialstressoccursat 0= O;as just indicated,theshear stress thereis zero. Ontheotherhand,shear stress is maximumat 13=90°, 270°.Attheselocationsthelongitudinalstressis minimum(longitudinalstressduetooverallbeambendingiszerointhe90°-270°plane). Itis concludedthatthemaximumnormalstressesalongthecoordinatedirec-tionsandthecorrespondingmaximumshearstresscanbecomparedtoindividualstresslimitswithoutthelikelihoodofoverstressalongaprincipaldirection.

AccordingtotheCodeofFederalRegulations,theallowablestresslimitsforatankinthegroundedconditionaretwothirdsoftheultimatestress.Fortherepresentativematerialusedinthisanalysis,theallowablestresswouldbeontheorderof40,000psi.Asshownintable6-II, themaximumstresslevelsinthe200-foottankinthegroundedconditionarewellbelowthisallowablestress.Itshouldbenotedthatthemostsevereriseinstressinthegroundedconditionoccurredinthelongitudinalstress.

.— —

-55-

Thisis theresultofanincreasedoverallbendingmomentactingonthemostseverelyloadedsupport.

Inthegroundingcondition,themaximumlongitudinalcompressivestressis increasedfrom-1,720psito-10,500psi.Zick’slongitudinalcompressivestresscriterionisbasedonbuckling,andvaluedat-4,700psi.Thephenomenonofbucklingprecludesthatoffailurebyyielding,andthereforeitwouldseemthatthetankinthegroundedconditionwouldbeindangerofbuckling.However,asdescribedearlier,theultimatestresswhichwillproducebucklingofsuchacylindricalshellhasbeendeterminedtobe-10,700psi,basedonananalysisbyTimoshenkowhichaccountsforinitialshellimperfections.Thisdiscrepancycertainlydeservesadditionalinvestigation.

Noattempthasbeenmadetojudgethemagnitudesofthestresslimitswhicharebasedontheallowableworkingstressofthetankmaterial.

6.4 DISCUSSIONOFFLUCTUATINGSTRESSESDUEToSAGGING/HOGGINGINLARGETANKS

Asdiscussedpreviously,theuseof400-foottanksforthetransportationofchemicalsincoastalwatersoropenseaintroducestheadditionaleffectofcyclicloadingduetowavemotion.Amaximumofabout7.2millioncyclesofhoggingandsaggingmaybeexpectedduringaten-yearlifeperiod.Ofthe7.2millionwaveencountersonlyaverysmallnumberofwaveswhichinducemaximumbendingwilloccur.Duringcyclesofhoggingandsagging,thelongitudinalstress dueto bendingfluctuatesaboutthestressconditionofnormaloperation.

Toillustratethemagnitudeofstressfluctuationsundersevereconditionsina400-foottank,theloadconditionsforthecasesofhoggingandsagging(determinedinsection5)wereutilized.Notingthatthemaximumvariationinbendingmomentsoccursatthemiddleofthetank,acomputeranalysiswasperformedforbothhoggingandsaggingforthesectionoftankhalfwaybetweensupportsoneithersideofthecentersupport.Theanalysisisanalogoustothatperformedforthenormaloperatingconditionofload.Havingdeterminedthestressdistributionforbothextremecases,thestressintensity(range)andmsximumstressatcertaincriticallocationswerecalculatedandarepresentedinfigure6-9andtable6-vII.Thestresslevelsshownrepresentextremeconditionsandwilloccurorbeexceededonlyasmallpercentageofthetime.

-,

-56-

TopofTank ~ +11,350psi

BottomofTank -7,650psi

TopofTank

Condition

BottomofTank ~-3,850psi

c)Sagging

‘~”BottomofTank ~+15,850psi

LongitudinalMembraneStressCycleintheShellattheIntersectionwiththeWiffeningRing (dynamicloadfactorusedinallcases)

Figure6-9. Inv@at@itIonofCyclicLoadingforFatigueAnalysisof400-FootTank

-57-

Table6-VII. LocalPeakStressCycleintheShell-RingIntersection

INSIDESURFACE OUTSIDESURFACE0=0 e=90° 8=18000=27000=0 0=90° @=180°0=270°

E%i2!uiLongitudinal 3,081Stress

NormalLongitudinal 6,090Stress

&@!&Longitudinal 21,800Stress”

6,190 10,131

6,090 6,930

6,036 -9,100

6,190 -18,342-2,493 +12,508-2,493

6,090 -13,820-2,400 +&180-2,400

6,036 9,540-2,340 -14,860-2,340

— —

’58-Section7

DISCUSSIONOFMATERIALSANDCONSTRUCTIONOFPRESSUREVESSELSFORBULKTRANSPORTOFLIQUIDCARGOESONBARGES

Thetransportofliquidcargoesintankbargesis regulatedbyfederallawforinter-stateandoverseasshipment.Thedesign,constructionandinspectionoftanksisgovernedbyTitle46oftheCodeofFederalRegulations.

CertaincargoespossessingdangerousorlethalpropertiesarelimitedbytheCodeastothemaximumvolumeorweightwhichcanbecarriedinasingletank.Thisisbecauseatankofsuchavolumeorweightis themsximumthatcouldreasonablyberecoveredwithoutexcessivedangertopersonsandpropertyshouldthetankbargesuffera casualty.Suchcargoes,andtheirtanks,areofconcerninthisstudyonlytotheextentthatsafehandlingorequipmentforrecoveryfollowingabargecasualtymightbeimprovedinthefuture.

Thisdiscussionis, therefore,directedprimarilytowardsrequirementsfor tankswhichcontainotherthanlethalliquidsandwherenoregulationsrelativeto thefluidpropertiesgovernthesize ofthetank. It is, furthermore,primarilyconcernedwithliquidsat subatmospherictemperaturesandatmosphericpressurewherepressurevesseldesignmustbeemployedandtemperatureeffectsonmaterialsmustbeconsidered.

Thissectionisnotanabstractofr~uirementspertainingtopressurevesselsperse, butanappraisalofthoserequirementswhichcouldaffecttheconstructionofverylargecargotanksofpressurevesseldesign.

7.1 DESIGNIMATERIALCONSIDERATIONSLargecylindricalpressurevesselsforcargoesbeingtransportedatsomewhataboveatmosphericpressure,theoretically,canhaveverythinwalls.However,atzeropositivepressure,thestaticpressureheadofthecargoonthelowerportionofahorizontalunstitfenedtankcanbesufficienttoeffectconsiderabledeformation.Forinstance,thepressureexertedbytheweightofpropanegasonthebottomofahorizontaltank40feetindiameterwillbeontheorderof145lbspersquarefoot.Thecylinderalsocandeformfromitsownweightifunsupported.Therefore,todesignatankwhichwillremainessentiallycylindricalandbeofminimumweight,abalancemustbeestablishedbetweenthenumberandsizeofstiffe~ers(internalorexternalframes), shellthickness,andmaterial.Thisis, inessence,thepurposeofthisstudy.Forlargetanks,thesectionmodulus,includingdepthandsizeofstiffenersandshellthickness,determinestherigidityofthestructureandthemaximumstressesatthesupportingsaddles.Designseekstokeepstresseslow,hence,materialstrengthpropertiesaresecondarytosectionsize.Forthisreason,high-strengthquenchedandtemperedsteelsmaynotbeeconomicallyjustifiable.However,sincetheyhaveexcellentnotchtoughnessWdretaintheirpropertiesofductilityandtou@-nessacrossweldedjointswithoutstressrelief,thehighermaterialcostmaybeoff-setbyfabdicationeconomies.

Minimumshellthicknessof 5/16inchis specifiedintheCodeofFederalRegulationsforcertainhazardousanddangerouscargoes(inparagraphs38,39and40ofSub-

.59-

chapterD).Forverylargetanks,designindicatesagreaterthicknessisnecessaryforreasonablestresslevels,thusthislimitationis ofnoconsequence.Sincethehigheststresses onthetankare foundat thesaddlesupportsbywhichthetankisattachedto the barge, localthickeningof theshellin theseareas willreducereactionstresses withminimumoverallincreasein tankweight.

Exceptwherecorrosion-resistantmaterialsare usedwhichare notaffectedbythecontainedfluids,or thecargois noncorrosive,mosttanksrequireacorrosionallowanceaddedb thedesignshell thickness.Paragraph52.05-12ofSubchapterFspecifiestheadditionof1/6 thedesignthickness,or 1/16inchwhicheveris less.Thisadditionis generallyofsmallconsequencewithrespecth theoargucapacityof the tank.

7.2 MATERIALSFORLOW-TEMPERATUREAPPLICATIONSTheCodeofFederalRegulations,Title46,Chapter1, Su~hapterF, “MarineEngineering,” 1968edition,designatesallowableferriticmaterialsfor low-tempera-ture servicein Table51.24.1. ThetablereferstoASTMspecificationsA300-5E,A333-63T?A334-63T,A350-61TandA352-60T.Whilethese s~cificationsare.all currentat this date (1968),eachhasbeenupdated,and incaseofA300,revisedto a considerableextent.Therefore,whendesigninga pressuretank, it mustbedeterminedwhetherthe CFRis to befollowedto the letter or thelatest revisionoftheapplicablespecificationis to beused.

TheU.S. CoastGuardNavigationandVesselInspectionCircular,No.7-67,dated9November1967,isacomplete(current)guidefor theuseofall steels (ferritic andaustenitic) in all forms– plate, shapes,castings,fastenings,andm forth– forlow-temperatureservicefromambientto below-320°F.Itappearsespeciallyvaluablein that, toa great extent,it doesnottie materialsto specifications,butnoteschem-istries andheat treatmentsrequiredtoprovidethestrengthandtoughnessfor theservicetemperature.Thus, thedesigneris free h choosethematerialbestsuitedto his applicationwithonlyprudentandreasonablerestrictionsinvoked.Nonferrousmaterialsare notincludedi4 circular 7-67, butmaybeusedat anylowtemperatureuponapprovsloftheapplicationbytheCoastGuardCommandant.Thisis specificallynotedinCircular7-67andinParagraph51.01-85,‘~Alternative~abrials” ofSub-chapterF, Title46-Chapter1 oftheCodeofFederalRegulations.

Withrespecttotoughnesspropertiesofferriticsteelsforlow-temperatureapplica-tions,ther~uirementthatfinegrainmeltingpracticebeemployedinmakingthesteelisuniversallyprescribedinspecifications.Also,Circular7-67emphaticallystatesthatwhereCharpyimpacttestingisusedtoevaluatenotchtoughness,theV-notchspecimenonlyisacceptable.Itis statedthereinthatcorrelationhasbeenes-tablishedbetweenthenil-ductilitydrop-weighttestandtheCharpyV-notchtestandthateitherofthesetwomethodsmaybeused.

CertainoftheASTMspecifications, includingA300-63T,specifytheCharpykeyholeimpacttest.WhereanASTMspecificationisusedtadesignateferriticsteelsfo,rbargeor shiptanks,andCharpyVorkeyholeimpactspecimensarecalledforbythespecification,thedesignermustindicatethattherequirementsofCircular7-67forCharpyV-notchteststakeprecedence.

Thosematerialswhichdonotundergoa ductile-to-brittletransitionwithdecreasingtemperature,suchas aluminumandausteniticstainlesssteel, are, in general,exemptfromimpacttesting. The.curveofimpactenergyversustemperatureisanearlyhor-izontallineto-3200F, thereforeimpacttests providenousefuldata.

Circular7-67, Paragraph4C, notesthatevaluationofmetaltoughnessis a fieldundergoingcontinueddevelopment.Thisis a referenceto thepresentemphasisonfracturemechanicswhichseeksto establishquantitativemeasurementsofmetalre-sistanceto brittlefailureandmathematicalanalysesofthisphenomena.Newtestiugtechniquesare beingevolvedwhichpromiseto bemoresignificantthanthepresentimpacttests, andthesewillberecognizedbytheCoastGuardas wellas otherCodebodiesastheyarerefinedandstandardized.

Thechemicalcharacteristicsofcertaincargoestransportablein steel tanksprohibittheuse ofsomeof the low-alloy,hightensilesteels. Ethleneoxideandpropyleneoxide,for instance,canbesafelycarriedin carbonsteel or austeniticstainlesssteeltanks, butcannotbe carriedin tanksofcopper-bearinglowalloysteel, suchasLukens’LT-75or U. S. Steel’sT-1, bothexcellent,low-temperaturesteels, becauseofreactivitywiththecopper.Thisparticularprohibitionis notedin Section40ofSubchapterD, Title46oftheCodeof FederalRegulations,andillustratesthat thenatureofthecargomustbecarefullyconsideredwhenselectingtankmaterial, inadditionto requirementsfor low-temperaturemechanicalproperties.

Table7-Iisacompilationofferriticsteels,stainlesssteelsandaluminumalloyplatespecificationsfromtheAmericanSocietyforTestingandMaterials,togetherwithpertinentmechanical/physicalproperties,whichcanbeusedforsub-ambienttemperaturecargotanks.Thesematerialscoverthetemperaturerangefromambienth below-400°F.

7.3 FORMINGREQUIREMENTSThemethodofformingparts for pressurevessels, in general,is notrestrictedbyCoastGuardor AM regulations.However,in order to takeadvantageofallowedmechanicalstress reliefprof;duresfor completedtanks,parts canbecoldformedtnonly4 percentplasticstrain . If individualparts, suchasheads,arestrainedmorethan4percentthroughcoldworkwhen.beingfabricated,theymustbethermallystressrelievedbeforeassemblyintothetankinorderthatthetankmaybemechanicallystressrelieved.Itisnotclearif thislimitonplasticstrainappliestocoldformingofnonferrousmetals;butwherea metalis knownto strainharden,it maybeassumedthat thermalstress reliefofseverelycoldworkedparts is required.

Tolerancesare appliedto theformedsectionsas directedbyParagraphs56,01-50and56.01-75ofSubchapterF ofChapter1, Title46CFR. Outofroundnessislimitedto 1percentof themeandiameter.Ina 40-footdiametertank, thisamountsto~4.8 inchesfromthetruediameter,Generallyspeaking,this is a generoustoleranceat this diameter.Submarinehullsof this orderofmagnitudeindiameterare heldtoout-of-roundnesstolerancesofless than1 inch.

Mismatchofabuttingedgesshallnotexceed1/4oftheplatethickness,or1/8inchforlongitudinaljointsand1/4inchf~rcircumferentialjoints,whicheveris less.Thistolerancerequirescareinthefabricationofthecylinders,since,intherollingofplates,itispossibletaformcylinderswhichvarybymorethan1/2inchindi-

Table7-1 MaterialProperties

MODU-WORK- LUSING OF COEFF. THERMAL SPEC.

HEAT STRESSELAS- OF CONDUC- HEAT@TREAT-20TO TICITY THERMALTI TY% TENSILE YIELD 70*F DENSITY

DESIG- CONDI- 650 x 10° CHARPY~PANSI~N(FT~HR) STRENGTHSTRENGTHBTU L3SNATION ALLOY TION (Pm) (PM) v-NoTcH(IN./IN./ F) (FT’F) (PSI) (PSI) (LB/°F) (CuIN.)

03B 21/2-31/2%Normal-17,500A203B NiSteel izedA203E 21/2-3l/2~ Normal-17,500

NickelSteel

A353 9%NickelSteeI

A410 Chrome-Copper-Nickel-AluminumAlloy

A516Gr55Carbon-Manga-neseSiIiconSteel

A516Gr60Carbon-Manga-neseSiliemSteel

A51~Gr65Carbon-Manga-neseSilicwlSteel

ized

Double 23,750Normal-ized&temperedNormal-15,000ized

Normal-13,750ized

Normal-15,000ized

Normal-16,230ized

29-30

29-30

29-30

29-30

30

30

:0

15ftlba@-75°F15fflbs@-150°F

25ftIbs@-32WF

15ft lbs@-l!30°F

15ft lbs@-5WF

151tlbs@-50°F

15ftlbs([h-50°F

.0000064 24.2

.0000064 24.2

.00000527 15.2[-58to32-F)

. oooooG’1

.0000064

,0000064

.ouooo(i-I

70,ooo-85,00070,ooo-85,000

100,ooo-120,000

60,000

55,ooo-G5,000

60,000-72,000

(i5,000-77.000

40,000

40,000

75,000

30,000

30,000

‘J.J,Oofj

35,[)00

0.11-0.12,283

0.11-0.12.283

0.11 .283

u,11 ,283

0.11 .283

0.11 .283

(),11 .283

AI

Table7-1(Cont.)MODU-

WORK- LUSING OF COEFF THERMAL SPEC.

HEAT STRESSELAS- OF CONDUC- HEAT[!TREAT-20TO TICITY THERMALTIVITY TENSILE YIELD 70°F DENSI

DESIG- CONDI- 650 x 100 CHA1{PYEXPANSION(FT2JHR) STRENGTHSTRENGTHBTU LBSNATION ALLOY TION (PSI) (PSI) V-NOTCH(IN,:IN./°F) (FT/°F) (Ps[) (PSI) (LBJ°F) (CVIN.)

A516Gr70Carbon-Manga-neseSiliconSteel

A517,all Chrome,grades Nickel

andmolyb-denumlowalloys

A537GrA Carbon-Manga-nese-SiliconSteel

A537GrB Carbon-Manga-neae-SiliconSteel

A538GrA 18NickelMarag-ing

SteelA538GrB 18

NickelMarag-ing

Steel

Normal-17,500 30izecl

Q&T 28,750 30(150°F&lower)

Normal-1’7,500ized

Q&T 20,000

Precip- 52,500itationhard-ened

Precip- 60,000itationhard-ened

30

30

26.5-27.5

26.5-27.5

15ft lbs .0000064@-50°F

30

15ft lbs ,00W.)064 30@-50°F

15it lbs ,0000064@-75°F

Byagree- ,00000597mentbe-tweenmillandpur-chaserByagree- ,00000597mentbe-tweenmillandpur-chaser

30

30

11.3

11.3

70,000 38,000 O.lL ,283

115,ooo- 100,000 0.11 0.283135,000 ~,289

70,ooo-90,000

80,000-100,000

210,000

240,000

50,000 0.11 .284

60,000 0.11 .284

200,000- — 0.29235,000

230,~00- — O.29260,000

-63-

mmml

-Fwml

mmaa“

I1

Zz-Om- m“r-oNM

.

m,+ .-.

4.

,-! .

xo00m“m

o0a0“a

Am0003o000’mN

c-E00000

.

z0is,4,4

ww“ 1.5

N

l-lI

c-

. .

o 3 0Lcl 0 mt- Ln t-m“ .m“c-l m :

0000-r-

0mC-J

Table7’-1(Cont.)

MODU-WORK- LIEING OF COEFF. THERMAL SPEC.

HEAT STRESSELAS- OF CONDUC- HEAT@TREAT-20TO TICITY THERMALTIVTY1 TENSILEYIELD 7(.)°F DENSITY

DESIG CONDI- 650 x 10° CHARPYEXPANSION(FT/HR) STRENGTHSTRENGTHBTU LBSNATION ALLOY TION (PSI) (PtH) V-NOTCH(IN./lN./°F) (FT/°F) (PSI) (PSI) (LB/gF) (CUIN.)

A240 18-8Type302 Cr-Ni

304 Stain-316 lese321 Steel347

A240 18-8Type304LLOW

316LCarbonStain-lessSteel

A240 ChromeType410 Stain-

430 leseB2095052Alumi-

numAlloy

B2095083Alumi-numAlloy

B2095086Alumi-numAlloy

B2095456Alumi-numAlloy

B2096061T6Temper,Welded

An-nsaled

An-nwhd

An-nealed

o-Temper

o-Temper

o-Temper

o-Temper

o-Temper

18,750

17,500

16,250

6,250

10,000

8,700

10,500

6,000

28-30

28-30

29

10.1

10.3

10.3

10;3

10.0

Ductileto .0000092cryogenictempe.

Ductileto .0000092cryogenictemps.

Ductileto .0000055cryogenictemps.Ductileto ,000012cryogenictempe.Ductileto .000012cryogenictempe.Ductileto .000012cryogenictempe.Ductileto .000012cryogenictemps.IWctileto .000012cryogenictemps.

9.5

9.5

14

80

68

73

68

99

75,000

70,000

65,000

25,000

40,000

35,000

42,000

24,000(Welded)

30,000

25,000

30,000

9,500

18,000

14,000

19,000

0.12

0,12

0.11

—

—

—

—

, 287-,292

,287-,292

.28

.097

.096

.096

.096

.098

-65-ameter.Forinstance,a40-footdiametercylindermayactuallybemorethan1/2inchtoosmallor toolarge.Carefulmeasurementofeach.y,olledplateandweldjointsetup,withallowancesfororrestraintofcontraction,willcircumventproblemsinthisrespect.

7.4 WELDINGCONSIDERATIONSWeldingproceduresforallmaterialsusedintankswhichwilloperateatlowtem-peraturesmustincludeconsiderationofnotchtoughness.Inadditionb thetensileandbendtestspecimensusedtoqualifyaweldingprocedure,specimensarerequiredfortoughnesstestingbyeithertheCharpyV-notchimpacttestorthenil-ductilitydropweighttest.Similarly,productionweldtestingperformedinaccordancewithSection56ofTitle46,Chapter1, CodeofFederalRegulations,mustincludeoneofthesetoughnesstests.

Weldingfillermetalis restrictedtothosecompositionscapableofpassingimpactordrop-weighttests.ASTMspecificaticmA233(MildSteelCoveredArcWeldingElec-trodes) showsthattheExx12,13,14,20and24classificationelectrodescannotbeusedsincenotoughnessrequirementis imposedonmetaldepositedbytheseelec-trodes.Similarly,forlow-alloysteel,flux-covered,arcweldingel-trodes,onlytheExx15,16and18classifications(lowhydrogenandlowhydrogenironpowdercoatings),exceptallE70= classes,arerequiredtopassanim@cttest.Thus,onlythesewouldbeacceptable.

Barewireforsubmergedarcweldingandinertgasshieldedwelding(ASTMA558andASTM599,respectively)isalsoclassifiedtoincludeorexcludenotchtoughnesstests.Onlythosegradessubjecttoimpactrequirementscompatiblewiththeplatesteelfor&edesignservicetemperaturecanbeusedfOrbV-tempeHttUr13bh.

Forradiographicqualitywelding,as is required for Class I andClass 11tanks,a certain amountofweldingelectrode control is required, especiallywithregardto moisture contentin the coveringof flux-coatedelectrodes and surface cleanlinessof spooledautomaticweldingwire. This is especially true for the higher tensilestrength electrodes, whereabsorptionof atmospherichumiditycan lead to entrap-mentof hydrogenin theweldswithsubsequentunderbeadcrackingandhydrogenflakes, or “fish eyes. “ Preventionofhydrogenentrapmentrequires bakingof theelectrodes andholdingthemin a heatedovenuntil theyare to be used. In some cases,especially in shipyards, electrodesmayremain out’of the ovenin the welder’sposses-sion for a limited time ordy– four to six hours —before beingrebakedto drive outabsorbedmoisture.

Foraluminumtanks, it mustberememberedthatthestrengthoftheweldgovernsthestrengthof thestructure. Themagnesium-alloyed,corrosionresistant, 5000series aluminumshavethebestweldstrength-to-platestrengthratio. Thealloy6061,andotherheat-treatablealuminums,whilecapableofhighstrength,are in-effectivesincetheweldsoflargetankscannotbesuitablyheattreated.Intheas-

ld d diti ld i 6061 ll l ti l kTh Ch V t ht t

-66-

7.5 STRESS-RELIEVINGCONSIDERATIONSParagraph56.01-70in Subchapte~F onMarineEngineeringin theCodeofFederalRegulationsstates thatall ClassI pressurevessels (unlessspecificallyexemptedbyothersectionsofthesubchapter)shallbestress relieved.Thefollowingparagraphsindicatethatonlythermalstress relievingis allowed.Thiswouldappeartoeffectivelylimit thesize oftanksfor ClassI servicesincethermalstress relievingofverylargetankscanbeextremelyexpensive.It’a furnacewhichwouldaccommodatethediameterof thetankdoesnotexist, onemustbebuiltandits costaddedto thetankcost.Furthermore,sincelargetanksare assembledonthebargehullbecausetheyaretoounwieldytohandlewhenassembled,circumferentialweldsjoiningstress-relievedcylindricalsectionsmustbelocallystress relieved.Thiscanpresentformidableproblemsofuniformheatapplicationandcontainmentto attainstress-relievingtemperatures.

ClassIIvessels,whichare theprimaryconcernofthis study,mustalsobestressrelievedin manyinstances.Exceptionsto thestress-relievingrequirementsexisthowever,whichmakeconstructionofjumbotankspractical.Stressrelievingis re-quiredofmildsteel(carbon-manganese-silicontype)onlyif theshellthicknessex-ceeds1.25inches(assuminglargetanksover20feetindiameter).AHOYsteelunderO.58inchinthicknessis exemptfromstressrelief.Mechanicalstressrelief,ef-fectedbyhydrostaticpressurization,ispermittedbyMerchantMarineTechnicalNote7-64of3December1964forClassIiandClassIIIvessels.Sincemechanicalstressreliefis theonlypracticalmethodtoemployonverylargetanks,thelimitsprescribedforitsusebyTechnicalNote7-64shouldbeconsidered.

Thefirstlimitationnotedis thattheyieldstrengthofthematerialmustbelessthan80percentofthetensilestrength.ThislimitationmightapplytomanyquenchedandtemperedsteelssuchasASTMA517,A542andA543.

Aspointedoutearlier underDesignConsiderations,theuseofhigh-strengthsteelmaynotbeattractivewheresectionmodulus,tominimizedeflectionandstress, isrequiredrather thantensilestrengthtopermithighstress. If, ontheotherhand,unwantedweightcanbeeliminatedbyhigherallowablestress, theuseofquenchedandtemperedsteels mightbeverydesirable.Insucha situation,andin thecaseofa verylarge tank, theprohibitionagainstmechanicalstres6 reliefof thesematerialsandthepresentrequirementto thermallystress ralievealloysteeloverO.58inchmightbecircumventedonthebasisofexperienceandtestingwhichindicatesthesesteelswillperformsatisfactorilywithoutanystress relief.

In1966,MPRAssociatespublishedareportentitled“TechnicalJustificationforUseofNi-Cr-MoQuenchedandTemperedSteelinClassBNuclearVessels.” Thisre-portwasaimedat theuseofQandT steel for nuclearreactorcontainmentvesselswhich,becauseoftheir size, shouldbemadeas lightas possible.Thefunctionofthecontainmentvesselis toprovidemaximumsafetyin theeventofmalfunctionofa nuclearreactorbycontaininga suddenrise inpressure. Thesevesselscannotbe

-67-

benefitedbythermalstressrelief.Toughnessis loweredandatendeneytowardheat-affectedzonecrackingdevelops.Thefactthatmanylargestructuressuchassub-marines,bridges,penstocksandstoragetankshavebeenconstructedis evidencethatthestressreliefrequirementcanbewaived.

Stressreliefis, inlargemeasure,desirabletoenhancefatigueresistancewhichmaybeanimportantfactorinatankonanocean-goingbargesubjecttowaveaction.Inthisrespect,itmaybenotedthatinthepresenceofflaws,high-strengthsteelsof80,000to100,000psiyieldstrengthhaveanendurancelimitofabout25,000psi,aboutthesameascarbonsteels.However,acarbonsteelstructuredesignedtoaworkingstressof15,000psiwillbeoverdesignedwithrespecttoitsendurancelimit.Aquenchedandtemperedsteeldesignedto.25,000psiworkingstresswillstillhaveaninfinitefatiguelife,assumingadequateweldingcontrolandnondestructivetestingisperformedtoassurefreedomfromgrossflaws.Quenched‘andtemperedsteelsof100,000tO 125,000tensile strengtharenotpronetostresscorrosionproblemsortobrittlefracture.Hence,ifdesigned,constructedandtestedtomini-mizefatigueresistance,theycouldbeexpectedtoperformsatisfactorilywithoutstressrelief.Itmaybenotedthatthenormalized9percentnickelASTMA353steelhasbeenapprovedunderASMECodeCase1308forfabricationtol-1/4-inchthicknesswithoutstressrelief.

Thesecondlimitationofmechanicalstressreliefis thatthedesigntemperatureshallnotexceed115°F. Thisstudyisconcernedwithlow-temperaturecargoes;therefore,theimplicationsofthisrestrictionarenotgermainetothepresentcase.Similarly,thisstudyis concernedwith“cargoestransportedatpressureslessthan100psi,andmechanicalstressreliefis generallynotrequired.

Afurtherrestrictiononmechanicalstressrelievingisthatthecargoescarriedshallhaveaspecificgravityof1.05orless.VeryfewcargoestransportableinClassIItankswouldfallinthiscategory,thustheeffectofthisrestrictionisnotconsideredsignificant.

Certaindetailsofconstruction,especiallyinreinforcedopenings,cannotbesatis-factorilystressrelievedbymechanicalmeansbecauseofinherentnotches(partialpenetrationnozzleweldsandsinglebevelweldswithnon-removedbackingstrips).DesignmusttakenotetoavoidsuchdetailswhicharespecificallynotedinMemorandum7-64.

Vesselsmustbedesignedtoeliminatestressconcentrationswhichmightleadtoex-cessiveplasticdeformationor, possibly,tofailureunderstress-relievinghydraulicpressure.

Lastly,unlessanextensivestressdeterminationisperformedwithstrainindicatorsduringthestress-relievingoperation,operatingpressureis limitedto40percentofthemaximumdesignpressure.Generallyspeaking,sincetheoperatingpressurei l t th t h i f th i t d d ft k id di thi

-68-thermalstressreliefdoes. Therefore,theuseofmaterialssusceptibletostresscorrosioncrackinginthepresenceofcertainfluidsmustbecarefullyanalyzedbeforewaivingthermalstressreliefinfavorofmechanicalstressrelief.

7.6 NONDESTRUCTIVETESTINGCONSIDERATIONSConstructioncostscanbeaffectedbytheamountofnondestructivet~stingspecifiedfora ClassIItank.Creditisgiveninweldefficiencyforradiographicinspectionupto100percentforflushground,radiographed,andthermallystress-relievedweldsinClassIvessels.Spotradiographyofoneareaper50feetofweld,usedinClassIIvessels,reducesdesignjointefficiencyto90percentwhereweldreinforce-mentis removed,andlowerstheallowableworkingstrengthoftheplate,requiringadditionalplatethickness.Itthereforebecomesnecessarytocomparethecostofadditionalplateweight(includingeffectofsuchadditionalweightoncargocapacity)andadditionalweldingwiththecostofcompleteradiography.Vesselweightisusuallyaverysmallpercentageofcargoweight,andradiographyis relatively=pensive.

Anadditionalefficiencyofonly5percentcanberealizedforfullradiographiccoverageforClassHvessels,and,therefore,wherespotradiographyisacceptable,itisgenerallypreferredto100percentcoverage.Inordertotakeadvantageofprovisionsformechanicalstressrelievingoftanksoperatingat-20°Forlower,spotradiographymustbe-tendedtoincludejunctionsbetweenlongitudinalandcircumferentialweldsandfor20timestheplatethicknessineachdirectionofweldfromthejunction.Itshouldalsobenotedthatcreditinweldefficiencyisnotgivenformechanicalstressrelievingasitis forthermalstressrelieving.

7.7 CARGOCHARACTERISTICS

Toassess thedesignproblemsassociatedwithlargetankbarges,itisnecessarytoknowthephysicalcharacteristicsofthosematerialswhicharecurrentlybeingcarriedintanksormightbeusedinsufficientquantitiest~warrantsuchtransporta-tion.

Table7-II,entitled“PhysicalPropertiesofGasses,‘‘ isacompilationofsomeofthemoresignificantphysicalcharacteristicsofthosegasseswhicharenowtransportedintheliquidstateorwhichmightreasonablybesocarriedinthefuture.TheCoastGuardtoxicratingandclassificationarealsoincludedinthetableforthosegasseswhichhavebeenso ratedandclassified.ThisdatawascompiledfrombothCoastGuardandcommercialpublications(references4, 7, and2.2through27).

TheImilingpointofthecargomustbeconsideredintankdesignsinceitwillinfluencethechoiceoftankmaterial,degreeofstressrelief,andweldingpractice.Thespecificgravityofthecargointheliquidstatewill,inpart,determinethestructuralloadingonthe barge. Theflammabilityofthecargois a measureofthehazarddueto leakagefromthetank.

TheCoastGuardtoxic isanindexofthehazardta fromcontact

Table7-IL PhysicalPropertiesofGases

Et-m3

-ma

.4L

-CDa I I I I I i I I I I 1 Im]C,* Mcnoli*CbmkalItitq -s12,# 0,810 12,5 74 B@nim

. . P019mO,mas .lll.:/3m pm

DGzmT8=Mw c.0. Chdnuiia

AUI1

Table7-II.(Cent’d)

I

vAmn

VAmm ET* CEt’1’lCALTEMP.fmr A m

TAMXMATUIAL m. %&’uRr (P,L%%% 1.talj PR-C

41]Ammbh~ IClOmleil*.~.i 170 ‘“ 10.743 no H.o I s

.,

-71-Thecriticaltemperatureandpressureisausefulindexofstability.Ifthecargoachievesthisstateandthetemperatureisexceeded,therewillbeachangeofstatefromliquidtogasregardlessofpressureincreases.Thisdata,combinedwithliquidandvaporphasedensities,definesthesafetyblow-offrequirementswhichwouldbeneededtoprecludetankrupture,shouldcriticalconditionsbeexceeded.

Table7-IIshowsthatthereareavaryingnumberofphysicalcharacteristicsassociatedwithmaterialswhichareorcouldreasonablybecarriedinrefrigeratedtanks.Ac-cordingly,a basicdecisionmustbemadeinitiallytodesignallsuchrefrigeratedtanksforthemostsevereservice,designtanksforoneor twosimilarcargoes,ordefinerangesofcharacteristicsanddesignforthemostsevereservicewithinthatrange.

-72-

ACKNOWLEDGEMENTS

Theauthorwishestothankthefollowingfortheircontributionstotheperformanceofthisstudyandtothepreparationofthispaper.

Mr. A.T.

Mr. R.C.

Mr. G.F.

Mr. R.J.

Mr. R.S.

Mr. R.A.

Fahlman,NavalArchitect,ElectricBoatGwin,Hydrodynamicist,ElectricBoatLeon,StructuralfiesearchEngineer,ElectricBoatMorante,StructuralResearchEngineer,ElectricBoatSnow,MaterialsEngineer,ElectricBoatToher,EngineeringSupervisor,ElectricBoat

Mr.DelBreit,BreitEngineeringCo.CDR.R.L. Brown,U.S.CoastGuardMr.

Mr.

Mr.

GeorgeDrake,NavalArchitectJohnEstes,Bethlehem/BeamountShipbuildingDiv.JohnFoley,AmericanBureauofShipping

LCDR.DeanFrankenhauser,U.S.CoastGuardCDR.R.C. Hill,U.S.CoastGuardMr.WilliamE. Hill,PortHoustonShipyardMr.SbnhopeHopkins,CanalBargeLinesCDR.J.L. Howard,U.S.CoastGuardMr.R. Johnson,GulfportShipyardMr.Jamesl?.Kanapaux,U.S.SalvageCorp.CDR.W.D.Markle,U.S.CoastGuardCDR.C.E.Mathieu,U.S.CoastGuardMr.WalterMichele,Friedy&GoldmanMr.R.W.Rumke,NationalAcademyofMr.ArthurStout,Jr., ToddShipyards

Sciences

CDR.C.R. Thompso~,U.S.CoastGuardMr.HarryTownsend,U.S.SalvageCorp.

-73”REFERENCES

1.

2.

3.

4.

5.

6.

7.

8.

9.10.

11.

12.

13.

14.

15.

16.

17.

18.19.

“StressesinLargeHorizontalCylindricalPressureVesseIsonTwoSaddleSupports,” byL.P. Zick,WeldingResearchSupplmmmtiSeptember1951.GuidefortheStructuralAnalysisofIndependentTankBarges,U.S.CoastGuard,January1966.‘‘AnalysisofSheikofRevolutionSubjectedtoSymmetricalandNon-sysmm6tri-calLoads,‘‘ N.Kalnins,JournalofAppliedMechanics,September1964.“DesignConsideationsforBargesTransportingHazardousCommodities,‘‘G.C. SteinmanandT.E. Carman,MarineTechnolon,Vol.3,No.3, July1966.ElementaryStructuralAnalysis.J. B, Wilbur,andC. H.Norris,McGrawHillBookCo,, NewYork,1958.TentativeGuidoforDeterminingScantlinmofIndependentCargoTanlw,AmericanBureauofShippingTechnicalStaff,February1963.“DesignandDevelopmentoftheSaturnMissileSystern,‘‘ MarineTechnolo~,Vol.3, No.3, July1966.ComputerSolutionofSaddleReactionforIndependentTankBargesGroundedona Pinnacle,U.S.CoastGuard,undated.PrinciplesofNavalArchitecture,J.P. Comstock,editor,SNAME,1967.AnEngineeringApproachtoLow-CycleFatigueofShipStructures,J. VastaandP.M.Palermo,TheSocietyofNavalArchitectsandMarineEngineers,74TrinityP1.Place,N.Y., N.Y.MechanicalVibrations,J, P. DenHartog,4thEditionMcGrawHillBook,Co.,NewYork,1956.“TheImpactofaFlatPlateonaWaterSurface,‘‘ J.H.G.Verhagan,JournalofShipResearch,December1967.ExperimentalInvestigationofRigidFlat-BottomBodySlammin~S.L. Chuang,DavidTaylorModelBasinReport2041,September1965.ProcessEquipmentDesire,L.E. BrownellandE.H.Young,JohnWiley&Sons,Inc.1959.“Self-EquilibratedRingLoadingofa StiffenedCircularCylinder,‘‘ E.M.Q.@ren, EuropeanShipbuildin~No.6, 1965.ShellsofRevolutionStatickadin~GeneralDynamicsElectricBoatReportNo.U411-64-046,December1964.

TripReport-GeneralDynamicsElectricBoatdivision,ChemicalTankBargeStudyContract,byC.W.Bascom,Autuat6, 1968.StressesinShells,W,Fliigge,Springer-Verlog,NewYork,Inc., 1966.“Onthe BucklingofCircularCylindricalShellsUnderPureBending,‘‘ P. SeideandV.I. Weingarten,JournalofAppliedMechanics,ASME,March1961,

-74-

REFERENCES(Cent’d)

20. TheoryofElasticStability,TimoshenkoandGere,SecondEdition,McGraw-HillBookCo.NewYork,1961.

21. “SteelDesignManual,‘‘ R.L. BrockenbroughandB.G.Johnston,UnitedStatesSteelCorp.ADVSS27-3400.01,July1968.

22. “MechanicalStressRelievingProcedures,‘‘ MerchantMarineTechnicalNoteNo.7-64,U.S.CoastGuard,3December1964.

23. itBulkLi~~idCarwes,c~~sificationandHazardRatingLis~!‘‘U.S.coastGuardCircular10-64,December1964,

24. “PropertiesofNickelSteelPlatesatkw Temperature@,‘‘ U.S.SteelCorp.,1960.

25. “ UwTemperature,LiquifiedGasTransportation,‘‘ MCietyofNavalArchitectsandMarineEngineers,1961.

26. “HandbookofChemistryandPhysics,‘‘ TheChemicalRubberCo.45thEdition,1964-1965.

27. TheAnalyticalChemistryofIndustrialPoisons,Hazards,andsolvents,M.B.Jacobs,SecondEdition,1949.

-75-AppendixA

INVESTIGATIONOFTHEENVIRONMENTOFLARGEOCEAN-GOINGBARGEH

A-1. INVESTIGATIONOFDYNAMICLOADFACTORSFC)ROCEANBARGESTheCodeofFederalRegulations,Title46,Subpart38.05-2,A-1statesthatoceanbargesshallbebuilttowithstandthefollowing“DynamicLoading:‘‘

1. Rolling30°eachside-120°in10seconds2. Pitching60halfamplitude-24” in7seconds3. HeavingL/80halfamplitude- L/2Oin8seconds(L= lengthofbarge)

TOinvestigateetheserequirements,thefollowingexpressions,whichdefinethemaximumvaluesofroll,pitch,andheavefromthelimitsgivenabove(assumingsinusoidalmotion),maybewrittenas:

Roll ~(t)=.: sin(: t i- g) Radians

Pitch @(t)= ; sin(~ t + O.) Radians

Heave z(t)+ sin(:t+Zo) Feet

Tocomputeaccelerationatapoint,thedistancefromthecenterofgravityismultipliedbythesecondderivativeofrolland/orpitch,thenheaveaccelerationsareadded.Forexample,combinedpitchandheave,inphase,yieldbowaccelera-tionsof ,.. ”

“$“,: ;\,.,.,.

‘ccBow=[$)2k] $’(~fbACCBOW= O.04952L (inft/sec2)

Byemployingstandardshipmotionsequations,itispossibletopredictthemotionsofvariousbargedesignsinrealisticseaways.ThesemotionsareusedtoevaluatetheCoastGuardStructuralRequirements.

Themethodsfordeterminingpitchheave,androllandassociatedresultsforfourbargedesigns(tableA-I),aredescribedinthefollowingparagraphs.FromfiguresA-1,A-2,A-3,andA-4,itis evidentthattheCoastGuardstructuralrequirementsareadequateforallnormalweatherconditions.Bargesdesignedforconstantheavyweatheroperation(windspeedgreaterthan3(1knots)mayrequiremorerigorousspecificationshowever.

Inthe case of the 24O-footbarge,itis evidentthat,thelowdeadweight-to-displace-mentratioresultsinexcessivemotions.FiguresA-1andA-2alsoindicatethatinheavyweatherasmallerbargewillIMingreatertrouble.Thisshouldb expected.

-76-NoteintableA-I thatthe440-footbargecarriesorIlyonetank.

TableA-I. BargeCharacteristicsNUMBEROFTANKS

1 2 2 2Diameter(ft)(tank) 40 40 30 20Length(ft)(tank) 400 400 300 200Barges

LOA(ft) 442 454 344 244LWL(ft) 440 450 340 240Beam(ft) 50 96 76 53Height(ft) 35 40 30 25Draft(ft) 26 23.6 17.3 12.2

iii (ft) 0.8 24.3 17.4 11,2Displacement 11,900 23,940 10,990 3,660(tons)

Inordertopredictthepitchingandheavingmotionsofthebarge,usewasmadeofthedivision’sSurfaceShipMotionsCornputEWProgram.AcompletedescriptionofthisprogrammaybefoundinreferenceA-2. Thefollowingisageneraloutlineofthewayinwhichtheprogramwasemployed.

Thebargewasdividedintoanumberofdi6creteseotions.Throughtheuseofstandardshipmotionequwlions,theforces@ restdtingmotionsontheentireshipwerecalculatedbysummingtheeffectsofthesecticms,h ordertobetterdefinethemotionofeachbargeinanoceanenvironment,arandomsea(Neumann)wasmathematicallyformulat~.Thisform~ationallowsastudyoftheresponseofeachcrafttovariousinputwaves.Throughtheuseoftheestablishedtheoryoflinearsuperposition,there~ponsesineachcomponentwavecanbesummedtogiveare-sponsespectrumfortheentirerangeofwaveswhichmightbeencounteredinwocean.

Sincetheaccelerationduetocoupledpitchandheavemotionis oftengreatestatthebow,theaveragevalueofthe1/1Ohighestbowaccelerationwasdetermined.+USOcalculatedwastheaveragevalueofthe1/1Ohighestpitchamplitudes.ThesehavebeenpresentedinfiguresA-1ad A-2.

-77-

1(WindSpeed–“3oKnots)

— Length(feet)

240

C.G.Spec.(All),-340440450

I2 4 6 8 10 12

CraftSpeed(knots)

FigureA-1. Pitchvs. Speed(NeumannSea)

I (WiidSpeed– 30Knots)rC.G. (4so)

.—— ——— ——— —— ——___ _

C.G.(440)

Length(feet)240

440

340

450

2 4 6 8 10 12CraftSpeed(koots)

FigureA-2. BowAccelerationvs. Speed(NeumannSea)

00

##

# 76’BeamBarge#

4968BeamBarge

10 20 30(WindSpeed(knots)

FigureA-3. RollinRandomBeamWas(MostProbableRollAngle)

-78-

56t — Neumann%*

52

48

u40E

./’--.-

/Pkm*OnSe* /1

10 20 3WindSpeed(knots)

$Barge53’Samn

76’Beam

96”Esam

)

FigureA-4. RollinRandomBeamSeas(Average1/10HighestAngle)

Rollmotionsforeachbargewerecalculatedinthefollowingmanner.Thebasicrollequation(referenceA-3)maybewrittenas

c+ C2 & C3@=C4Cos w

wheretheconstantsC~, C2,C3,audC4 maybefunctionsoffrequencyofencounterandvariouscraftparameters.Expressionsforthesemaybefoundinreference2.

Definethefollowing

KC3UndampedNaturalFrequency:w~ ~ ~IC

DimensionlessDampingCoefficient:w=K%

TuningFactor: ~~timn

whereu = frequencyofwaveencounter

[-1/2

MagnificationFaotor:p= (1- A2)2+ v2A21TherollequationcanWsolvedtoyield(referenceA-4)

~=--pe “where:

+/A = theratioofrollamplitudetowaveamplitudedistanceofcenterofbuoyancybelowwaterline

-79-UsingtheexpressionsforCl, C2,C3(referenceA-3),themagnificationfactorcanbewrittenasfollows:

where:

H=T=

G—M=(J.

g=

c#=A@=

barge~eambargeheightbargedraftbargemetacentricheightfrequencyofencountergravitationalconstsntconstant= 1.9forabargefunctionofBandw suchthat:

2 2A@=.7x (; :)2

WB~ ~) : 0.4when( -

2 2A@=0.25x(: ~ ) when( ; :) > 0.4

Inordertousetheprincipleofsuperpositionandachieveaseaspectrumresponse,thefollowingproductis formed.

~2R’[f]’’ [%]

where(_&) istheamplitudeseaspectrum.

IntegratingRwithrespecttofrequencyyieldsthetotalresponse,TR. Therearemanystatisticalvaluesthatmaybecalculatedfromthevalueoftotalresponse.Amongtheseare:

MostProbableValue=Average1/1OHighest=

Bothwerecomputedforeachbargeoceans.

0.707x @1.8 X6

inthefollowingtwomathematicallysimulated

NeumannSea 725.—H2 VK2U2

(*, . +eLIJ

PiersonSea9.7X104

I? “T+’(~)=+

MVK= windspeedin lmots

A-2. INVESTIGATIONOFWAVECONDITIONSENCOUNTEREIIALONGTYPICALBARGEROUTESANDEXPECTEDFREQUENC~SOFENCOUNTERFORUSEINHOGGINGANDSAGGINGCALCULATIONS

GiventheseadatapresentedintableA-II(referenceA-5),theexp~ctedvaluesoffrequencyofencounterinheadseascm bedetermined.Sincetheseadataareintermsofwaveperiod,somemanipulationsmustbeperformedtoyielddataintermsofwavefrequency.Then,by.includingtheeffectofboatspeed,avalueoffrequencyofencounteriscalculated.ThisispresentedinfigureA-5.

FiguresA-6,A-7,andA-8presentwavedatafortheb~geroutesconsideredthusfar.

Inordertobeabletocalculatehoggingandsaggingofthebarges,werequirefrequencyofencounterforwavelength= boatlength;thefollowingrelationsuppliesthis:

(useLWLforboatlength)

f incycles/seeis presentedin tableA-IIIforvariouslengthsands~eds, Torelatethis to sea data, findthepercenttimethatwavelengthequalsboatlength(figureA-6).Thisgivespercentofthetimethat the abovefrequencieswilloccur.

TableA-IL WaveData

PACIFIC120A2~ GULFCOAST AWAVE PEIU2ENT WAVE PERCEWI WAVE PERCENT WAVE PERCENT WAVE PERCEHEIGHT occuRENm PERIOD OCCURENCEHEIGHT OCCURENCE PERIOD 0CCUIU3NCEHEIGHT OCCUREm) (SOc) {ft) (see) m)

1.0 4.79 CALM+ 3.46 1.0 15.71 CALM+ 6.39 1.0 7.131.5 6.01 <5 2.9.52 1.5 20.86 <.5 59.19 1.5 13.73.0 21.M 6-7 29.19 3.0 31,19 6-7 22.32 3.0 26.35.0 21.90 8-9 20.27 5.0 18.98 8-9 7.70 5.0 21.96.5 16.71 10-11 10,ti 6.5 7.11 10-11 2.09 6.5 11,88,0 10.19 12-13 3.97 8.0 3.89 12-13 0.71 f?.o 7.549.5 7.28 14-15 2.09 9.5 1.23 14-15 0.11 9.5 4.2811.0 4.79 16-17 0.917 11.0 0.71 16-17 0 11.0 2,5213,0 3.62 18-19 0.306 13.0 0.149 18-19 0,15 13.0 1,4614.0 2.09 20-21 0 14.0 0.037 20-21 0.34 14.0 1,6916.0 0.25 >21 0.356 16.0 &027 >21 0.97 16.0 0,201?.5 0.31 17.5 0.037 17.5 0.1819.0 0.61 19.0 0 19.0 0.3621.0 0.25 21.0 0 21.0 0,3122.5 O.io 22.5 0 22.5 0.0624.0 0 24.0 0.074 24.0 0.1725.5 0.05 25.5 0 25.5 0.10

N~: (FromoceanWaveStatieticHbyfbglx?nad Iunb). 1967Edition.

+ - Inclde,n~lurllmuwn.“

IMaamforavwrage“YearROurdCcditions”

-82-

60

r

-y55 I — EasternPmiflc50 i —--GulfCoast

45 ----- WesternAtlantic---

40 1

u 35 i?? n

5

0

K1..-.‘i-.IIi------

I

100 300 500 700 900 10001300WWOLength(feet)

FigureA-5. WaveLengthsAlongU.S. Coasts

36

32

2a

40

12345678 9 10 11 12 13 14 IS 16WnveHeight(feet)

,

FigureA-6.WaveHeightsAlongU.S. Coasts

60

55

50

1s

10

50

-83”

60 ~._._

Ss L1 ! — EsctemPcdfic50

I‘-- GulfCoast-‘-- WesternAthntic

45 .......i

10

5

0

■I.ib.-L.-

i-qi--_L.+D-..

— &S-6 8 10 12 14 1616WwePeriod(Secon&)

4128 18432851373910001300WsveLength(Feet)

FigureA-7.WavePeriodsAlongU.S.Coasts

r r7.....—I — Eti.rnPmAfic

— - — Gulf~~ U------I

t----- W-ternAtl,ntic

I I1 *--OresterThm—.—. —— .2sCpc

.---,LJrd,l,,l,,lO .02,04.06

FigureA-8.

.08 .10.12 .14 .16 .1S .20 .22 .24 .26 .28 .30FrequencyofEncountm(CPS)

Frequency of Encounter at10Knots

—

-$4-Assumingthata44O-footbargeisoperated300daysperyearat8knots,thefollowingestimateis madeofsaggingandhogg@cyclesin1()years,usingdatafromfigureA-6andtableA-III.

Cycles= 0.2x [ 10yrsx 300days/yearx 86400see/dayx 0.13861=7.2x106

TableA-III. IMtermintngFrequencyofEncciunter

LENGTH SPEED@O@)(ft) 2 4 6 8 10 12440 0.1156 0.1233 0.1309 0.1386 0.1463 0.1540450 0.1142 0.1217 0.1292 0.1367 0,1442 0.1518340 0.1327 0,1426 0.1526 0.1625 0,1725 0.1824240 0.1602 0.1743 0.1883 0,2024 0.2165 0.2306

A-3. INVESTIGATIONOFEFFECTSOFFLUIDSLOSHINGTheterm“freesurfaceeffect”isgiventotheproblemofaship’sreducedstabilitycausedbythesloshingofaliquidcargo.StandardtextssuchasreferenceA-6treatthissubject.Thesimplestsolutiontothisproblemistoreducetheamountoffreesurface.Inthecaseofchemicalbarges,thisis achievedbyfillingthetankstocapacity.ItshouldbenotedthatCoastGuardregulationsdonotrequireaminimumtankfullness.Itmaybeadvisabletorequirethattanksbefilledtocapacityoremptied~tolessthmaboutone-fourthcapacity.Operatingathalfcapacityresuitsinthegreatestfreesurfaceeffect.

Theeffectoffluidmotiononbargemotioncanbedividedintotwoparts:a. actualforces impartedto thebargebythefluid,andb. changeincenterofgravityandmomentofinertiaoftheloaded

bargeduetomotionoftheload.

Anorderofmagnitudecalculationindicatedthat, for the expectedboatmotions,thecenterofgravitymovesonlyoneor twopercentofthebargelength. Fromthis itwasconcludedthat changesin thelocationofthe centerOfgravityandmomentofin-ertia maybeneglectedin calculatingbargemotions.

Theforcesimpartedtothebargebythefluidaredifficulttocalculatesincetheexactfluidmotionisnotl-mown.Itispossible,however,tomakesomegeneralremarksconcerningtheeffectoffluidmotionsonbargemotions.Clearly,if thefluidhadneithermassnormotion,thebargemotionwouldbeunaffected.Also,ifwaveandfluidhatllral Peritis were farbelowthtiofthebarge,therewouldbelittleeffect.Asystemforcedatafrequencyfaraboveitsownnaturalfrequencywillnotrespond.Maximumeffectsduetofluidmotionwilloccurwhenthenaturalperiodsoffluid,barge,andwavearethesame.Largeeffectswillbeobservedwhenbargeandfluidhavethesamenaturalperid ofmotion.IntableA-IVandfigureA-9thenaturalperiodsofbargeandfluidmotionarecomparedwiththeaveragenaturalpericdofthewavesfoundalongthebargeroutes.

—

’85-

Thecaseofhalf-filledbargesatzerospeedhasbeenpresented.InfigureA-1Othenaturalperiodofthefluidispresentedforvariousfillingdepths.‘-7 Althoughthenaturalperiodofbargerollfallsgenerallyabovethemostcommonaveragewaveperiod,thenaturalperiodofthefluidtransversemotionfallsamongthemostcommonwaves.

Itcanbeconcludedthatfluidmotionsmaynothavealargeeffectonbargerollmotions.Theeffectonbargesurgingandpitchingisnotyetclear. Thefluidshouldhaveverylittleeffectonbargeheaving.

Another,lessobvious,effectoffluidmotionsisthepossibilityoffatigueloadingfromthecyclicsloshingmotions.Althoughthesloshingloadsmaybesmall,thepossibilityoflargefatigueloadingontankwallsandsupportsshouldbeinvestigated.

I I60

55hAverAwofAUWWH

50

45

40

15

1050

I I a 1012 141618 202224 26WwaPeriod(seconds).

FigureA-9.ComparisonofNaturalPeriods

Agreatdealofworkhasbeendoneinvestigatingsloshinginverticaltankssuchasrockets,Judgingfromthelackofavailablereferencematerial,verylittlehasbeendoneinthefieldofhorizontaltanks.Thetwoproblemsaresomewhatsimilarbutdetailsofmathematicalmodelingareverydifferent.

Themathematicalmodel(figureA-II), basedonreferencesA-8andA-9,representsasimplifiedapproach.Evenso, ityieldsnonumericalresultsduetothelackofex-perimentalparametersandactualtestinformation.Alsoneededis amethodforapplyingexperimentalresultstofull-scalebarges.Themathematicalmodelrepre-sentsafirstapproachandispresentedasanindicationofwhatmustbedonetosolvethesloshingproblemanalytically.

-86-TableA-IV. NaturalPeriodsofMotionBarges&1/2FullTsmks

BARGE-LENGTH~

440

450

340

240

ROLLPERIOD

25.4

9.47

9.0

7.9

10010987

6

s

4

3

2

1!Jd7

6

5

4

3

2

1

FLUIDPERIOD ~EC~Nl13)LONGI

stMODE36

36

21*9

17.9

CJDINALMOTION

~12.8 8.4

12.8 8.4

8.25 5.6

6.74 ‘4.6

TRA1stMODE

4.3

4.3

3.73

3.06

VERSEM(ZndMODE

2.3

2.3

1.99

1.62

—

* -~**.

LongitudinalMotion

T~ansverseMotions2ndMode

I I0.2 0.4 0.6 0.8 1.

FluidDepth/TankDiameterFigureA-10.NaturalPeriodsofVibrationofFluidsContained

CircularCylinderofRadius20ft, Length400ftina

rION3rdMODE

1.75

1.75

1.52 ,

1.24

-87-

m .—------K ——— ——

FigureA-n-.SimplifiedSloshingMathematicalModel

ConsiderthefluidinatankpitchingatO(t)= a sinuttobeapointmassactedaspringanddamper.Theequationofmotionis (x= O is atcenteroftank):

[ti+r~cos O+tic0s6+ mgsinO=O1Forsmall13sinO~ O cos @~ 1 andwehaveasimpleequation:

ti+rii-kx+mgasinut =0kx+~ 2+—X=n-l m -g a sinwt

Letx=Asinut+Bcosut~=uAcos~t-m Bshwt~ =ti2Asin@ - U2BCosLot

ru-ti2Asinmt -L02BcosLOt+— Acosut-~ Bsinwt+~ Asintitm

+k~ Bcoswt= -gasinot

1

2-wA-~13+ ~A=-ga ‘

!

A~~-u2)=-ga+~B

I—

fl

—

k-U2B-1-~A+ ~B=O A% = \(W2-~)B

onby

[-

(:2:-U)(—-LO2)~ti1

(ga~)-— B = -gr; B.ru n-l’ f _2km2+ti4+&—m

m’ m 2m

Thus,ourexpressionforfluidmotionhas

mA J

become:

x(t) =Astiut+Bcosut

~2~2A= -km)g~

B= rurnga(k2 - 2kmti2+ ti4m2+ r2u2) (k2- 21Cm~2+ u4m2+ r2w2)

ForcesontheendofthetankwillbeFT= fi

2 2‘T =-mu Asinut - mmB COStit

or FT =

=> ‘T =

Ifwehadappropriatevaluesforr, m, sndk,wewouldnowhaveasolution,sincetheCoastGuardspecificationsgivevaluesforaandu. Variousmethodsforevalu-atingmhavebeendevelopedinthecaseofverticalcylindersandmaypossiblybeapplicabletothisproblem.

-J

-89-

REFERENCES

A-1. Codeof FederalRegulations,Revisionof1January1968,Title46,Parts1-146.

A-2. ADescriptionoftheTheoryandOpera-tionoftheSurfaceShipMotionsComputerProgramNumber0344,R.C. Uhlin,ElectricBoatdivisionReportU411-68-020,April1968.

A-3. FundamentalsoftheBehaviorofShipsinWaves,IR.G.Vossers,PublicationNo.151aoftheN.S.M.B.

A-4. “TheRollingandPitchingofaShipatSea.ADirectComparisonBetweenCalculatedandRecordedMotionsofa ShipinSeaWaves,‘‘ D.E. CartwrightandL.J. Rydill,InstituteofNavalArchitects,Vol.99No.1, January1957.

A-5. OceanWaveStatistics,HogbenandLumb,1967Edition,A-6. PrinciplesofNavalArchitecture,J.P. Comstock,Editor,SNAME,1967.A-7. InvestigationoftheNaturalFrequenciesof FluidsinSphericalandCylindrical

Tanks~J. L. McCartyandD.G. Stephens,NASAT.N. D-252,May1960.

A-8* A“MonographonTestingforBoosterPropellantSloshingParameters,D.M.Eggleston,ConvairDivisionReportGDC-BTD67-089,June1967.

A-9. MeasurementsoftheUnsteadyForcesActingonaCircularCylindricalTankContainingLiquidDuringHarmonicMotion,P.R. Guyett,R.A.E. TR67098,April1967.

L.

-90-

AppendixB

ANALYSLSOFRINGSTIFI?ENERS

Analysisofcircularringssubjecttoappliedsupportloadsandappliedtankandliquidweightloadswasperformed;1800saddlesupportsofwidthhwereassumed.Theloaddistributionontheringduetothesupportwasassumedtobepr = pl cos Ofor. .lfll~ 90°and Ofor90°40<270”. TheloaddistributionduetothetankandliquidweightwasassumedtohePO. p2sinO(seefigureB-l). Thetotalsystemofappliedloads must be self-equilibrating. Assumingthat the ring widthquals the saddlewidth, h,

_2L‘1 = hir2r

and

where~ isthetotalloadsupportedbythesaddle.

Itisnecessarytoexpandthesaddleload,p , ina Fouriercosineseriestohavetheproperformforanalysis.Performingthe~ecessaryintegrationsyields:

I2Q ~+irPr=-— ~cos e+; ~cos4Qhr2r cos20-15

+~35COS6@-~cos 80+ :Coslo e-... I(B-1)

Theweightloading,PO,alreadyhastheformofthen = 1termofseries:

GP~”-— sinorrh82

FromFlugge‘-1 (p219,Eq.13a-c),settingPx= 0,~ S —aX2

governingdifferentialequationsoflinearthinshellbendingtheoryintermsofvandw,thecircumferentialandradialdisplacements

2C& dw ‘r3r .0dB2+iii+ D

aFouriersine

(B-2)

~ Oyieldsthe

forcircular-ringsrespectively

(B-3)

.>-

,_

-91-

(a)

F@ureB-1, Load@gonRingStiffener

.

where ==-(1-V2)

andk.&_

12r2

-92-

12d4w _+2d2w ●w ‘rr.—= od04 d62 D (B-4)

(B-5)

(B-6)

@nAppendixB, t refers tostiffenerthickness,i.e., outerradiusminusinnerrzdius.)

Differentiationofeq. B-2withrespectto13andsubtractionofeq. B-1fromthere-sultyields:

C&v+d05

Substitutionofeqs.

+*=& [& (Pr)+ PO1 (B-7)‘o ‘D

B-1andB-2forprandPOgives:

$sinz9+fi sin40

Let w be aFouriercosineseries:m

W=z WnCosI1on=o

(B-8)

(B-9)

Differentiationofeq. B-9, substitutionintoeq. B-8, andequatingcoefficientsofthesametrigonometricfunctionsyieldsthefollowingsolutionforw. (NotethatW.isdeterminedfromeq. B-4.)

~ (~)+: (~) [--$ cos2ew = - D;l+k)

+~2Cos40 2-—cos69 + — COS 80

(15)3 (35)3 (63)3

2_— Cosloo...(99)3 1

Similarly,assumev isa Fouriersineseries:

(B-1O)

—. —

.93.

Differentiationofeqs.B-1OandB-11,substitutionoftheseresultsandeq.B-2intoeq.B-3,andequatingcoefficientsofthesametrigonometricfumctionsyieldsthefollowingsolutionforw

() ( )[2 g sfi6+&2Q,~=. c 1D hrr kD — sinZe

h?r2r (3)31 Cos48 + 1 1-— sin60 - Cosse

(2)(15)3 (3)(35; (4)(63)3

1+— Cosloe-.. .1 (B-12)

(5)(99)3

FromFltigge(p214,eq9a-h),thecircumferentialforceandmomentresultantsintermsofv andwdisplacementsare:

.O=:(*+.) +; [’+$1

Me . ‘—[1

w+d~r2 d02

whereEt3K= ~

(B-13)

(B-14)

(B-15)12(1-V4)

Substitutionofeqs. B-10andB-12forv andwintoeqs. B-13andB-14yield:

+~2 2

COS 60 - — COS 80 + — Cosloe-... 1(E-16)(35)2 (63)2 (99)2

M@. - ‘2(?!%),p [$ .0s2, - ~ .0s4012r2+t2h~

+~2 2

COS 60 - — Cosse -1-— Cosloe-.. .1(B-17)(35)2 (63)2 (99)2

Thecircumferentialstressintheringattheinnerandoutersurfacesisgivenapproxi-matelyby

(B-18)

—

-94-

Substitutionofeqs.B-16andB-17intoeq.B-18andassociating(M)withthecross-sectionalareaand(ht2/6)withthesectionmodulusofthestiffeningringyields:

at,0 = (),

_.~ + ‘r‘e = .043$ (B-19)

at0 =180:

,158zj● 043 $ (13-20)

Themaximumcompressivestressoccursat6 = O;themaximumtensilestressoccursatO= 180°.

REFERENCES

B-1.StressesinShells,W.Fliigge,Springer-Verlag,NewYork,Inc., 1966:

-95-

AppendixC

OUTLINEFORSTFWNGAGEINSTRUMENTATIONOFATANKBARGE

C.1. INTRODUCTION

Generally,strainrecordinginstrumentsareusedtodescribestrainmagnitude,direction,anddistributioninareasofcomplexstructureswhichmaynotbereducibletomathematicaldescription,or toverifythepresenceandmagnitudeofcertainstrainswhichhadbeenpredictedbytheoreticalanalyses.

Themainobjectivehereis tadescribe,ingeneralterms,presentlyavailablein-strumentationproceduresapplicabletoanexperimentalstressanalysisoflargetankssupportedandtransportedbybarge.

C.2. DISCUSSIONAsaprerequisitetotheinstallationofanyinstrumentation,thefollowinginformationisnecessary:

1.2.

3.4.

5.

6.

Typeoffluidwhichwillbecarriedduringthetest.Geographiclocationoftheworksiteandthetimeofyearproposedforin-stallation.Typeandavailabilityofelectricalpowerforsiteworkandtesting.Numberandsizeoftankpenetrationsavailabletopermitinternalinstrumen-tationinstallationandtheexitingofsignalloads.Totalnumberofinformationchannelstoberecordedandspecificlocationsatwhichstraingagesandothertransducersarerquired.Durationofthetestprogram.

Thefollowingtestconditionsarerepresentativeoftheexperimentalprogram1. Strainda~willberecordedduringfillingandemptyingoftheti or tanks.2. Straindatawillberecordedduringactualoperationof‘thebargeinsea

statesif thebargeselectedisforoceanservice.3. Bow-slamming,heaving,twisting,andpitchwillberecordedasafunction

ofstrain,pressuredistribution,accelerationsandtime,relativetoseastateandfomvardvelocityifthebargeisforoceanservice.

4. Straingageandtransducerdatawillberecordedusingdynamicsinstrumen-tation.

C.3 GENER4LAREASOFINTERESTGeneralareasof interesthavebeenidentifiedandshownonfigureC-1. Straingageswillbeapplied,bothinternallyandexternally,intheareaofthesupportsaddle.Measurementsofstraindecaybetweensaddlesupportswillbeobtainedbyanarrayofstraingagesontheinnerandouterskinofthetank.

-96-

FigureC-2indicatesthegeneralareasofinterestonthebargehull. Pressuretrans-ducers,velocityandaccelerationinstruments,andstraingageswillbeusedtoobtainanunderstandingoftheforcesactingonthebarge/tankduringactualoperation.

C.4 PROPOSEDINSTRUMENTATIONTosimultaneouslyrecorddynamicdatafroma largenumberof signalsources,multichannellightbeamoscillcigraphsor taperecordersmaybeused.Thesedevicesrecordaprocessedorconditionedsignalwhichisdevelopedbyperipheralequipment.Ingeneral,thefollowingequipmentisneededtoobtainanintelligiblesignalfromastraingageorothertypeoftransducer:

1. Powersupply2. Amplifierandsignalconditioner3* Recorder4. Interconnectingcable.

C.5 TESTFACILITIESANDCONSIDERATIONSsincea testofthistyperequiresconsiderabletimeandexpen8e,adequatefacilitiesarerequiredtoprotectexpensiveequipment.Asuitablestructure,centrallylocatedwithrespecttoheavilyinstrumentedareas,canreducetheamountofcableusedandresultinasignificantreductionininstallationcosts. Thepossibilityofmorethanonerecordingstationshouldbeconsideredifheavilyinstrumentedareasareseparatedbyahundredfeetormore. Secondary,unmannedrecordingstationscanbesyn-chronizedtothemainrecordingstationoroperatedindependently.

Animportantconsiderationinthecostevaluationofmultichanneldynamicrecordingsystemsisthefrequencylevelwhichthesystemmustrespondtoandrecord.ThefrequencylevelwillbewithinOto50Hz. Theoutputsignalfromastraingagemaybedirectlyrecordedbycertainlightbeamoscillographswithoutamplification,thuseliminatingoneofthecostlycomponentsoftherec~rdingsystem.However,ade-cisiontodowithoutamplificationmustbejustifiedbytheoreticalanalysis.

Finally,itis importanttohaveaclearunderstandingofthetypeofdataobtainablefromstraingages.Ifasingle-elementstraingageisusedatapointofinterest,thestraindatadoesnotpermitacalculationofthemaximumprincipalstressunlessthestrainfieldisamaximumanduniformlyuniaxialinthedirectionofthegage,suchaswouldoccurinacontrolledtensilespecimentest.

Atwo-elementorbiaxialstraingagemaybeusedtocalculatemaximumandminimumprincipalstressesatthepointofinterestwhenthebiaxialgageorientationis identicaltathebiaxialstrainfield.

Athree-elementstraingagewillallowmaximumandminimumprincipalstressesanddirectionstobecalculatedfromrawdataforthepointofinterestwithouttheneedforspecificorientationwithrespecttathetestspecimenptrainfield.

Hence,thethre-elementstraingage(triaxialstrainrosette)is mostsuitabletodetermineprincipalstressesanddirectionsinstructuressubjecttocombinedloading,suchastwistingmdbending.

— -1

-97-

a

1l—

—

I

FigureC-1. TypicalLocationofExtensiveInstrumentation

FullScale II

1

I

I

FigureC-2. TypicalInstallationofPermanentMaterialonInsideofHull

— —

-.

Photoelastic

-98-AppendixD

DISCUSSIONOFAPPROACHFORTANK/%AItGESIAMMINGMODELTESTS

D.1 PROPOSEDBOWPRESSUREMEASUREMENTSINWAVESThisdiscussionoutlinesaprogramofbargemodelfabricationandtowingtestsinvariousregular,sinusoidalwaveconditions.Thepurposesofthetestsaretameasurethebowimpactpressuresandnormalvelocitiesandtodefinethepitchingandheavingamplitudesofthebarge.Duringtheprogram,thetheoreticalpeakpressures,basedonthemeasurednormalvelocities,willbecalculatedandcomparedwiththee~eri-mentallydeterminedvalues.Completionofthisproposedprogramwillrequirethefollowingspecificitemsofwork

D.1.1 WOKSTATEMENT1. Fabricateonewoodenmiddlebody-afterbodybargemodeltaapproximately

1/25scale.Themodelwillprovidefortheattachmentofvaryingbowcon-figurationsandwillbefabricatedinaccordancewithlinessuppliedbythecustomer.

2. Fabricatethreeseparatewoodenbowconfigurationstothesamescaleasthemiddlebody-afterbodymodelofItem1inaccordancewithlinessuppliedbythecustomer.

3. Installtowingstaffandpitch,heave,andverticalaccelerationinstrumentationinthemiddlebody-afterbodysection.Installninecrystal-typepressuretrans-ducers,andnormalvelocityinstrumenbtioninthebowsection.Normalvelocitieswillbeobtainedbyelectricallyintegratingverticalaccelerationmeasurements.

4. Ballasteachmodelh thepredicteddisplacements(lightandheavy),centerofgravitypositions(longitudinalandvertical),andlongitudinalradiiof~ration.Calibratealltestinstrumentation.

5. Conducttowingtestswitheachmodelattwodifferentdisplacementsandatonevelocityintowavesoffourdifferentscaleheights.Testswillbeconductedinsixdifferentwavelengthsforeachwaveheight.Approximately144testswillberequired.Timehistoriesofthewaterimpactpressuresatninestations,normalvelocitiesatthreelongitudinalstations,verticalaccelerationsatbowandcenterofgravity,pitchangles,heavedisplace-ments,towingvelocity,andwavecontmmswillberecordedsimultaneouslybyaminimumoftwooscillographicrecorders.Thetimereferencesofthetworecorderswillbepreciselycorrelated.

6. Reducealltestdatatoengineeringtits andexpandtofull-scaleproportions.

7. Usingthenormalvelocitiesoccurringateachpressuretransducerlocationduringtheinstantofpeakpressure,calculatethetheoreticalmaximumimpactpressureusingmethodsdescribedinreferencesD-1andD-2.

8. Preparegraphicalpresentationsofthepeakimpactpressures(measuredandcalculated),pitchandheaveamplitudes,normalvelocities,andvertical

-99-accelerations.

9* Prepareandsubmitamodeltestreportpresentingthetestandtheoreticalresults,representativetestphotographs,anddiscussionsofthetestingtechniques,instrumentation,andresults.

D.2 SLAMMINGThefollowingcommentsaremade concerningthediscussionofslamming.

1. Itwillbeuseful,whilemakingslammingtests,toinvestigatethefollowinga. Effmtofon-boardfluitiontaintngtanks.b. Effectofsurgeduetotowingbyasurgingtug.

2. ThemethodofstaticstrengthdesignisconcernedmainlywithseaplanedeEignpracticewhichmayproveadequate.

D.3 DEC~SIONTheDTMBReportNo.1994(referenceD-3)hasbeenreviewed,aswerereferencesD-1,-2, -4 and-5. WebelievethattherisetimeanomaliesfoundinTable3ofreferenceD-3areamanifestationofthetrappedairproblemdiscussedinreferencesD-1mdD-4. Also,theirinstrumentationsystemfrequencyrespmwe(1200Hz)wasrelativelylowcomparedtothesystemresponse(200KHz)describedinreferenceD-1.

Systemresponsesof200KHzrequiretheuseofcathoderayoscilloscopesandstreakcamerasfordatarecording.Theuseofthisequipmentonourtowingtankcarriageisnotfeasible,particularlyinviewofthefactthatwewouldbeinterestedinnineseparatepressuremeasurements.However, wecanachieve-5000HZsystemresponseutilizingoscillographrecordingtechniques.Itisproposed,therefore,thatthebowpressuresatninespecificlocationsbemeasuredbycrystal-typepressuretransducerswith200KHzfrequencyresponseandrecordedbyoscillographsincorporatinggalvan-ometershavingfrequencyresponsesof 5000”HZ. Thus, an overall system responseof 5000HZwillresult.Crystal-typetransducersaredesirable~ecauseoftheirsmall(O.20Sinch)diameterpressurefaceandlowsensitivitytoaccelerationforces.Itisanticipatedthatmaximumfull-scalebargenormalvelocitiesof10to20fpsmaybeincurredinwaves.Sinceabargeis relativelylightlyloaded,thebowwillnotcutthroughthewaterandmaybeexpectedtodeceleratesignificantlyunderthegrowingpressurearea.Inanyevent,Wagner*suggeststhattheoutboardedgeofthewettedwidthsweepsoutward4ataspeed,

VNc=—

p(c)

wherethefunctionp(c)isgivenbyapowerseriesfornon-wedgeshapes,andby~tanp forwedges.

*Germanmathematician

-1oo“Thus,foranarbitraryhulldeadriseof20andafull-scalenormalvelocity,V of15fps,theresultingspeedis N’

.c=;x15x- 0.0309 x 12=9,150in./seefullscale

Thepointvalueofthepeakpressurewould,therefore,remainonaO.208inchdiameterpressuretransducerfor23microseconds~orthefull-scalecase,and115microsecondsfora1/25thscalemodelwithFroudescaling.A5000Hzsignalhasaperiodof200microseconds,whichisequivalenttoaquartersinewaverisetimeof 50microseconds.Thus,a zerowidthpeakpressurecorrespondingtatheaboveexamplecouldconceivableybemeasuredatmodelscale,butnotfullscale.Whatthepressurepickup“measures‘‘ isdependentuponthewidthofthepressurepeakaswellasuponthespeedwithwhichittraversesthepickupdiameter.Anexampleofthevariationofthiswidthwithwedgedeadriseis showninfigureD-1,whichisareproductionofafigurefromreferenceD-2.

Lnviewoftheabove,it is furtherproposedthat thepeakpressuresbecomputedbythe methodsofreferencesD-1, D-2, andD-5for comparativepurposes.Tofacilitatethesecomputations,thenormalvelocitiesincurredduringimpactwillbemeasuredatthethreelongitudinalbargestations,correspondingtothethreetransverselinesofpressuretransducers,byelectricallyintegratingtheverticalac-celerations.Thepressuremeasurementswillberecordedsirnultaneouslywiththenormalvelocitymeasurementsonthesameoscillographtapeandwill,therefore,becorrelatedbytheprecisiontimereference.Thenormal velocityattheinstantofpeakpressureateachpressurestationcantherebybeobtained.

Duringeachmodeltest,measurementsofthebargepitchingandheavingamplitudesandtheverticalaccelerationsatthebowandcenterofgravitywillbeobtained.Also,recordingsofthetowingvelocityandthewavecontours(measuredoutboardofonespecificbargestation)willbeobtainedduringeachtest.

Itis believedthatthis combinedmodeltest datawillfullydescribetheeffectsofwaveimpactsonthevariousbowconfigurations.

Oncethis data hasbeenobtained,theproblembecomesoneofestablishingtherelationbetweenstaticstrengthdesignandthehighlytransientloadingsimposedbylowdeadriseimpacts.Ourpastexperiencewithfull-scaleseaplanesoperatinginwaveshasbeenthatthemaximummeasuredpressuresarefarinexcessofthosewhichthebottomplatingcouldsustainstaticallyoveranysignificantarea. ItmaybeofinterestthatoneoftheauthorsofreferenceD-5laidthefoundationin1947fortheapproachofreferenceD-6. Thisisa methodwhichstillfindswideusetodayonhydrofoilsaswellasseaplanes.Itisbelievedthatasimilarcorre~tionofstructural‘~successes~~and“failures”withtheorymayberequiredforbarges.

—

18

16

14

12

10

6

4

2

-1o1-——

IIIIIIii

II/rI— FromDetailedAualysis I

/- —— FromExpanding-PlateAnalogy

//(withspray-rootareaestimated) /

40° \.-- —- ----- ---—- >

~ LL--1 I0 I0 0.2 0.4 0.6 0.8 1.0 1.2

X/G

FigureD-1. PresswreDistributionOvertheWedge

.—

-1o2-REFERENCES

D-1. JournalofShipResearch,Vol.11, No.3, September1367.

D-2. “ThePenetrationofa FluidSurfacebyaWedge”,JohnD. Pierson,StevensInstituteofTechnologyReportNo.381,July1950.

D-3. “Two-DimensionalExperimentsontheEffectofHullFormonHydrodynamicImpact”,MargaretD. OchiandFrankM.Schwartz,DavidTaylorModelBasinReport1994,May1966.

D-4. Journal of ShipResearch. Vol. 11, No. 4, December1967.

D-5. “ATwo-DimensionalStudyof the Impactof Wedgeson a WaterSurfaceforthe BureauofAeronautics”, R. L. Bisplinghoffand C.S. Doherty,MassachusettsInstitute of Technology,Departmentof AeronauticalEn-gineering, ContractNo. NOa(s)-9921,March20, 1950.

D-6. MilitarySpecification,AirplaneStrengthand RigidityWaterand Handlingbads for Seaplanes, MIL-A-8864(ASG),18May1960.

UNCLASSIFIEDSecurityClassification

DOCUMENT CONTROL DATA-R&D(SewrityClase.,ficationof titln, body of sbstract and indexing annotationmust bo ontereclwhen the overall report IS classified)

1. ORIGINATINGG AcTIUl?Y (Co.-por#te aufhor)GeneralDynamicsCorporation

28. REPORTSECURITYCLASSIFICATIONUnclassified

ElectricBoatDivision zb GROUP

1. REPORTTITLE

STRUCTURALDESIGNREVIEWOFLONG,CYLINDRICAL,LIQUID-FILLEDINDEPENDENTCARGOTANK-BARGES

!DESCRIPTIVENOTES(Typedmpd .sndinchsiwdates)

FinalReport5.AUTHOR(S)(Lasfname,firsr name, initial)

Bascom,C.W.

;.REPORTDATE 7a. TOTALNO.OF PAGES 7b, NO, OF REF5

May1970 102 27ja. CONTRACTOR GRANT NO. 9=. ORIGINATORS REPORT NUMBER(S)

NOO024-68-C-5419 U413-68-125b. PROJECT NO,

SF013-03-04;Task2022,SR-184c. 9b. OTHERREPORTNO(S)(Anyoth.rnumbersthetmaybeaaaigned

thisreport}

d. SSC-205I10,AvAILABILITY/LIMITATIoN NOTICES

Distributionofthisdocumentisunlimited.

Il.SUPPLEMENTARYNOTES 12.SPONSORINGMILITARYACTIVITYNavalShipSystemsCommand

II13,ABSTRACT

Thisreportdescribesa programofanalyticalresearchtodeterminetheavailabilityofreliablemethodsforthedesignoflong,largediameter,cylindrlca”tanksandtheirsupportsfortransportationofliquidsandlow-pressureliquifiedgasesin barges forservice on rivers or at sea. Loading conditions, existingdesign/arialysiSmethods, material considerations,and a computer method forpredictingstressesarepresented.

Themajorconclusionoftheworkperformedisthatdesignproceduresforriverbargetanksupto20feetindiameterarewellestablishedandthatnofailuresduetoinadequatedesignpracticehavebeenreportedsincerefrigeratedtankswentintoserviceabouttenyearsago.Thepresentmethodfordesigningriverbargetanksisa logicalstartingpointfordeterminingthestructuralconfigurationoflargetanksforoceanicservice,butmoredetailedanalysisofloadsandresultingstressesshouldbeperformedforthisapplication.

Severalareasinwhichtheoreticalorexperimentaleffortisneededareidentified:(1) investigationoftank-saddle-bargeinteraction,(2) investigationoffatiguecriteriaforcyclicloading,(3) investigationofbucklingcriteria,(4) analyticalandexperimentalinvestigationofslamming,and(5)experimentalverificationofstressesina full-scaletank.

&

DD,W%1473 UNCLASSIFIEDSecurityClassification

— — —.

SecurityClassification4. KEYWORDS

INSTRUCTTONSL ORIGINATINGACTIVITY.Enterthenameandaddressof thecontractor,subcontractor,~rantpe,DepartmentOfDE+fenseactivityorotherorganization(corporateauthor)issuingthereport.2a. REPORTSECUHTYCLASSIFICATION.Entertheover-all securityclassificationof thereport.Indicatewhether“RestrictedDate” is included Markingis to be in accordantewithappropriatesecurityregulations.2b. GROW Automaticdowngradingis specifiedin DoDDi-rective5200.10andArmedForcesIndustrialManual.Enterthegroupnumber.Also, whenapplicable,showthatoptionalmarkingshavebeenusedforGrouP3 andGroup4 as author-ized.3. REPORTTITLB. Enterthecompletereporttitlein allcapitalletters. Titles in all cases shouldbe ~classified.If a meaningfultitlecamotbe selectedwithoutclassifica-tion,showtitleclassificationin all capitalsinparenthesisimmediatey foilowingthetitle.4. DESCRIPTIVENOTES If appropriate,enterthetypeofreport,e.g., interim,prowess, s~mmw, ~n~ufJror final.Givetheinclusivedateswhena specificrePortingperiodiscovered.5. AUTHOR(SIEnterthename(s)of authotis)as shownonor inthereport.Entm1astname,ffist Dame,mid~einitial.If military,showrankandbranchof service. ThenameOftbeprincipalauthoris anabsoluteminimumreq~rem@n~6. REPORTDATE Enterthedateof thereportas day,month,year;ormonth,year. If mOrethanonedateaPPearsonthereport,usedateof publication.7s. TOTALNUMBEROFPAGES Thetotalpagecountshouldfollownormalpaginationproced~es,i.e., enterthenumberof pagescontatiinginformation7b. NUMBEROF REFERENCESEnterthetotalnumberofreferencescitedinthervort.8a. CONTRACTORGRANTNUMBERIfappropriate,entertheapplicablenumberof thecontractor wantunderwhichthe report was written.8b,SC,&Ed.PROJECTNUM9ER.Entertheappropriatemilitarydepartmentidentification,=.uchaswOjeCtnumber,subprojectnumber,systemnumber%tasknumber)etc.9a. ORIGINATOR’SREPORTNUMBER(S):Entertheoffi-cial reportrmmberby whichthedocumentwillbe identifiedandcontrolledby the originating activity. This numberQstbe uti queto thisreport.9b. OTHERREPORTNUMBER(S):If thereporthasbeenassignedanyotherreportnumbers(e~fherby fheoriginatororbythesponsor),alsoenterthisnumber(s)-1o.AVAILA131LITY/LIMITATIONNOTICESEnteranylim-itationsonfurtherdisseminationof thereport,otherthanthos~

LllROLE

A‘#T

LINK❑ROLE WT

LllROLE

cWT

imoosedby securityclassification,usingstandardstatementssu~has:

(1)

(2)

(3)

(4)

(5)

“QualifiedrequestersmayobtaincopiesOfthisreportfromDfJG”,#ForelW announcementanddisseminationOft~sreportby DDCis notauthorized.”,,u, s. Governmentagencies mayobtaincopies‘fthisreportdirectlyfromDDC. OtherqualifiedDDCusersshallrequestthrough

.1,,,U. s, ~litary agencies mayobtaincopiesof thisreportdirectlyfromDDC Otherqualifisdusersshallrequestthrough

,,,

“AU distributionof thisreportis controlled.@al-ifiedDDCusersshallrequestthrough

, ,,

If thereporthasbeenfurnishedto theOfficeof TechnicalServices,Departmentof Commerce,for saleto,thepublic,indi-catethisfactandentertheprice, if knowm11. SUPPLEMENTARYNOTES Useforadditionalexplana-torynotes.12. sPONSORINGMILITARYACTIVITY Enterthenameofthedepartmentalprojectofficeor laboratorysponsoring(Paring for) theresearchanddevelopment.Includeaddress.13. ABSTRACT:Enteranabstractgivinga briefandfactualsummaryof thedocumentindicativeof thereport,eventhoughit mayalso 3ppearelsewhereiIIthebOdyOfthetechnicalr@-port. If additionalspaceis required,a continuationsheetshalbe attached.

Itis highlydesirablethattheabstractof classifiedrePOfibe unclassified.Eachparagraphof theabstractshallendw“ithanindicationof themilitarysecurityclassificationOfthein-formationin theparagraph,representedas (TS),(s), (c). orW.

Thereis no limitationon thelengthof thenbstract.How-ever,thesuggestedlengthis from150t~ 225words.

14, KEYwORDS:Keywordsaretechnicallymeaningfulte~$or shortphrasesthatcharacterizea reportandmaybe usedaSindexentriesforcatalogingthereport.Keywordsmustbeselectedso thatno securityclassificationis required.~denti-fiers,suchas equipmentmodeldesignation,tradename,mi~itaprojectcodename,geographiclocation.maybe usedaSkeywordsbutwillbe followedby anindicationof technicalcon-text. 77he assignmentof links,ralcs,andwei~htsis optional.

UNCLASSIFIEDSecurityClassification

— — .-—,,

SHIP RESEARCH COMMITTEEMaritime Transportation Research Board

Division of EngineeringNational Academy of Sciences-National Research Council

N%i.s prog’ec?thas been conducted under the guidance of Adui.SOYZ$Group 1.[,.sh.i.ph’t?search Connni.tte?e.The (!omrnittee has cognizance >f ShipStrueturw Conrnititceprojectsin materials, design and fabrication as reLating 10 ir,~provedship ztwc turea. This

responsibility entaik r~ecormwzdir,g research ob,je~-tives,tuses,

prepax,i~gproject prOspec-evaluating proposals, providing liaison atidtechwica1 gutdance, revieui.w?

project reports, and stimulatiry productive avenues of research.

Mil,M. L. SELLERS, ChairmanNaval ArchitectNewport News Shipbuilding

and Dry Dock Company

DR. H. N. ABRAMSON (1,11)Director, Dept. of Mechanical SciencesSouthwest Research Institute

MR. W. H. BUCKLEY (1,11)Chief, Structural Criteria and LoadsBel1 Aerosystems Co.

DR. D. P. CLAUSING (III)Senior Scientist, Edgar C. Bain

Laboratory for Fundamental ResearchU.S. Steel Corporation

MR. D. P. COURTSAL (11,111)Project ManagerDravo Corporation

MR. A. E. COX (1,11)Senior Program ManagerNewport News Shipbuilding

and Dry Dock Company

MR. J. F. DALZELL (I), CoordinatorSenior Research ScientistHydronautics, Incorporated

DR. w. D. DOTY (III)Senior Research ConsultantU.S. Steel CorporationApplied Research Laboratory

MR. F. D. DUFFEY (III)Welding EngineerIngalls Shipbuilding Corporation

PROF. J. E. GOLDBERG (1,11)School of Civil EngineeringPurdue University

MR. J. E. HERZ (1,11)Chief Structural Design EngineerSun Shipbuilding and Dry Dock Company

MR. G. E. KAMPSCHAEFER, JR. (III)Manager, Application EngineeringARMCO Steel Corporation

PROF. B. R. NOTON (II)Department of Aeronautics

and AstronauticsStanford University

MR. W. W. OFFNER (111)Consulting Engineer

PROF. S. T. ROLFE (III), coordinatorCivil Engineering DepartmentUniversity of Kansas

PROF. J. WEERTMAN”(II,III)Walter P. Murphy Professor

of Materials ScienceNorthwestern University

CDR R. M. WHITE, USCG (1,11)Chief, Applied Engineering SectionU.S. Coast Guard Academy

PROF. R. A. YAGLE (11), CoordinatorDepartment of Naval Architectureand Marine Engineering

University of Michigan

coR D. FAULKNER, RCNC (I,II)Staff Constructor OfficerBritish Navy Staff

(I) = Advisory(11) - AduLsvq

(1~1) = Advisory

MR. R. W. RUMKE, Execwtive SecretaryShip Research Committee

GPoup I, Ship .StrainMeasurement & .Analysistir[q 11, ship stm.lwtwl1 Uewiqn(;roupIII, Metallurgical Studies

SSC-191,

SSC-192,

SSC-193,

5~c_194,

SSC-155,

SSC-196,

SSC-197,

SSC-198,

ss~-lgg,

SSC-200,

SSC-201,

SSC-202,

SSC-203 ,

SSC-204.

and P. O.

!

,