0,U.1mc, THE SOCIETYOF NAVALARCHITECTSANO MARINEENGINEERS. ‘* 74 TrinityPlace,New York,N.Y.,10006

.+’6

010

~ P,wrt.bemsmt,dattheShipStructureSmPO$iuInW,,Mn@n,D.C.,O.t.b.,6.8,1975

~+% x;S**L#

Ships–Yesterday’s Technology–Today’sSome Tanker ExperienceD. J. Szostak, Member, Marine Transport Lines, Inc.,New York, N.Y.

0 COPYMt 1975bYTheSOCIetYof NavalArch,tect,and Ma,,., E.~,nee,,

ABSTRACT

This paper describes,ina practical

manner, some operatingexperiencewithtankersbuiltinthe mid 19601swhen finiteelement analysistechniqueswere notbroadlyapplied, A comparative review of the trans-

verse strength of certain vessels is present-ed from the view of rules in effectand design

approaches availableatthetime ofdesignversus rulesand analysistechniquessub-sequentlydevelopeda“d applied, As a resultofthe experiencedescribeditis recomrne”d-ed thatproper attentionbe paidt“ allcriticalareas of ship!sstructnre. A plea is madeforthe use ofrealisticdesign loads reason-

ably representative of practical conditions tobe expected at sea. A few rather obvioas butoften ignored recommendations regarding

detailsare presented. The paper furtherdescribes some experience with engine roomdonble bottotn deflectionsand bow they were

quantified.

INTRODUCTION

The drawing boards normally ccmta.inmore interestingproducts than the assembly

lines, especially in periods of rapidly

expanding technology. For obvious reasonsthe latestconceptual design is not immediately-

available We bmy yesterdays product in

today!s marketplace.This is particularlytrue of ships. A

contract signed in tmid 1974 promises deliv-ery in mid 1977 of a ship builtin accordance

with rules in effectat the time of contract

signing. Detail design a“d plan development,material procurement and vessel assembly:

not only is each of these phases of ship con.strmctiootime co”surning, but one follows

another. Meanwhile. researchers and

designersfindmore rigorousapplicationofdesigntheoryand technology, Sometimes

prmdence requiresimmediate adoptionof suchdevelopments when thenew technologydiscoversand overcomes some undesirablefeaturepreviously,thomghunwittingly,incor-

poratedina constructioncontract. zx,,ptwhere safety01’designdeficiencyconsider-ationsarise,itisimpracticalto continuallyupdatethe designinac.orda.”ce withtheselatesttechnologicaldevelopments sincecostoverrmns and latedeliverieswould beguaranteed.

Generally speakingshipsare builttomeet requirements ofa classificationsociety.Classificationsocietyrulesare usuallybasedcm pasthistorya“d thereisan understandablelagin updatingthese rulesto TefLectthelatesttechnology. Ship desigmrs as wellas thetechnicalstaffsofthe Classificationsocietiesare usuallya stepahead ofthe ruleschanges.The existingrules,therefore,shwld repTe-se”tthe minimum designcriteria.

As requirements expand and applicationsdevelopwhich precede tbe experiencefactorsofthe rules,basictheorybecomes moreimpoz’tautin solvingdesignproblems. Thesedesignsbecome thebasis ofthe experienceneeded to npdatethe regulationsa“d the sta”-dard designtechniques.

But what oftheinnovativedesign? Whatis tobe done toassure thatthetheories

appliedare ind..d practical? what factorofsafetyshouldbe used? Mo?. importantly.toashipowner, what shouldbe done with existingdesignsthatexhihitproblem. ? One possible

approach isto instrument shipsto evaluateexistingdesignsand to obtainor refinedata

fordesignwe. Researchers and desigwersoftenmeet with oppositionf.om ship.mvnersand operatorscm thisapproach. Like theauthori“ Reference 1 theybemoan the scarcity

G-1!----

CASE I . Plan of Cargo Area

PrincipalPa rticnl.ars Metric English

Length B.P. 221.0 725,-0,8

Breadth MLD 33.2 109,-0,$

Depth MLD 16.6 54t-6~~

DraftMLD 12.2 40,-0,,

Deadweight 60,000 T

Figure 1 - Plan and Particulars of 60,000 DWT Tanker

of volunteers,butprobably fordifferentreasons.

Commercial shipsare builtwith a verydefinitepnrpose in mind and most owners ofsuch ship.are notinterestedintransformingthese intodesignlaboratoriesdue to fear of

eOssibleinterferencewith OeeratiOnsandconsequentcosts. Those me.mbers oftheR & D community interestedin fallscaledatagatheringmust convincethe shipovmers ofthebenefitsthe owners willderive$1’ornvolunteeringtheuse oftheirshipsforbasicresearch, designdevelopment, or whateverotherlegitimatereasons might exist. Welllaidoutexperimentationwhere the research-ers recognizethevagaries ofreallifeat seaand are willingto properlyplanand coordi-natethe executionof datagatheringwould beacceptedhy most shipowners.

For the most part ifproblems developin existingdesignsanalyticalapproaches areneeded to understandand overcome them.This paper dealswith certainstructural

problems which occurred in specifictankersdesignedand constructedinthe mid 1960-s.Itisthe intentofthispaper to describe,instrictlya practicalmanner, the natureoftheproblems and how theywere handled. Thetheoryand the detailsofitsapplicationarenotpresentedbut are referencedwhereappropriate. A comparative review ofthe

prOblem structuresispresentedfrom thepointofview ofrulesin effectand designapproaches availableat thetime ofdesignversus rulesand analysistechniquesavail-

ableat tbetime ofthevessel repair.Two emmples referredto as Case I and

Case IIdeal withtransversering strength.Case IIIdescribesa problem ofdoublebottomflexibilityin largetankers.

CASE I

Rapid development in tanker sizefromabout 50,000 DWT tons to in excess of300,000 DWT tons dnringthe 19601swasfosteredby a combinationofvarious develop-ingtechnologiesincludingshipyardproductionmethods, welding engineering,materialsapplication,compnter sciencesand structuraldesigntechniqtte.s. Intbe designa“d constric-tionofmany tankersbuiltdnringthistimevarious stagesofthesetechnologieswere

appliedsometimes withless than desirableresults,

The lag inproperlyapplyingavailabletechnologyoftenmanifestsitselfin the formof operatingproblems. Numerous tankerswere reportedtohave developedcracks

particularlyinthe transversering stt.acture.These cracks had been attributedto variomscauses includingimproper stiffeningagainstbuckling,inadequateshear area, insufficientweld area and poor detailarrangement, oft.”compounded by poor workmanship.

Vesml

The firstexatnple to be sitedinvolves a

crude oiltanker delivered from the builderin

1966. The designand constructionwere inaccordance with the 1964 rulesofa majorclassificationsociety.

The vesselwas bmiltas a 60,000 DWTton ship,accornodationsaft. The dimensionsand generaltankplan are indicatedinFigure 1.

Typicaltankers ofthatera had fewcargo tanks, each quitelarge. This vesselhad twelvecargo tanksand two midship

permanent water ballasttanks. permanentwater ballasttankswere speciallycOa@d aSwere thebottom fearfeetof.11cargo tanksintendedto carry sea water ballast.

Incident

In mid 1968 the vessel,enroatefromthePersian Gulfto Europe with cargo tanksfulland midship ballasttanks empty,encounteredextreme sea conditionsand windsofhurricaneforce. The storm was sointensein thearea thatat leastone vesselwas lostand numerous otherswere reportedin trouble. Wave heightsof20-30 ft.abovethedeck were reported. Considerablenuisancedamage occurred on deck and above:derrickbooms were bent, accornodationladdersand davitswere damaged, pipesupportswere distorteda“d items on top ofthemidship deck house were .everelYdamaged.

Figure 2 - Face Plateon Deck Transverse

Figure 3 - Strut

Fignre 4 - Strut-Swa sh Bmlkhead

.,G3

—

‘U$m_m—m Figure 5

Typical Damage to Transverse Ri w

Fignre 6 - Face Plate on Bottom Transverse Figure 7 - SideshellSat-In

G4

.--——

Because oftheheightto which thedamage was in evidenceitis concludedthatsignificantgreen water washed at least4meters m’ more above themain deck. Theheightofthewaves relativetothe deck wasprobablyinfluencedhy heavy rollingofthevesselduringthe storm.

W

The seriousdamage which instigateddetailedanalysisofthe structureofthe shipoccarred intheport midship permanentwater ballasttank. Alltransversewehframes inthistankwere damaged. The webplate8 and the faceplatesatthe gunwalecorner, the strut,and the bottom transversene.ar thebilgecorner were deformed andcracked. The extentoftbe damage to thesetransversemembers can be seen inthesketchesand photosof Figures 2 tbrn6.

Figm-e 7 shows the externalevidenceofthe damage, the side shellhavingbeen setinshout one meter.

Design and Construction- 1964

This vesselwas designedaccordingtomethods and rulesin existencein 1964. Tbescantlingsofthetransverseringaccordingto 1964 rulesand as builtas are indicatedinFigure 9. At tbe time ofplan developmentthe scantlingswere checked by beam theorymethods and fonndtohe withinacceptablelimits.

The transverse strengthofthehullwasbased, generally,on theassumptions thatthe vesselwould be inthe followingcondi-tions:

1. The “esselafloatwith one centertank in structuralt.stcondition.

2. The vesselafloatwithwing tanks,port k starboard,in structuraltestcondition.

3. The VeSSd would be operatedin

accordance with tbe conditionssetoutin thevessel,s loadingmanual,encounteringwaves presumed bytbe rulesofthe classificationsociety.

The heightofthewanes supposed by theclassificationsocietyrulesatthetime ofdesignwere based on the following:

Hw= 1,026 xLw04

Where, Ffw= Wave HeightLw = Length ofShip

For thisvesselwhose Lw is 221 meters,

Hw = 8.9 meters

Therefore the draftused forthe stndyoftransverse strengthwaD tbemean draftineach loadi~gconditionplus or minus 4.45meters. With a fullload draftof 12.2 meters,the effectivedraftbecomes 16.65m or wavecrestat deck at side.

The loadingdistributionsused foranalysisofthetwo testconditionsand at seaconditionsnotedaho.reare illustratedinFigure 8.

StructuralT.stCondition4.15 MeterEffectiveHead

-8%23=%E1216.6 Meter EffectiveHead (atDeck at Side)

Figure 8 - Loading DistributionsAasnmed at Time ofDesign

Analysisand Repair - 1968

Are the failnresbrieflydescribedaboveattributableto faultydesigncriteria? Werethe assumptions regardinghead and vesselloadingproper? Was thebasic designandanalysismethod adequate? In short,was thetechnologyofthetime up to thetask ofpro-ducinga vessel suitableforitstime ?

As a resultofthe aforementioned dam-

age a detailedrea~alysisOfthetransversestructu?ewas accomplished. Du~ing the fotmbriefyears from the vesselgsdesigntmtilthedamage, considerableprogress had been madeinthe applicationofmatrix methods totheanalysisofcomplex structures. The re-analysiswas accomplished by two-dimensionalframe analysisutilizing“Stiff”essmethod”of matri~ analysis. Cmnp@ ationswere per.formed using‘rIcesStrudl- 111and “Stress”computer programs. The background oftheanalyticaltechniquecan be foundin References2 th~ough 11. Three dimensionalanalysiswasalso ccmdnctedtakingintoaccountthe relative

G5 f---

—-

Figure 9 - Case I - Midship SectionAs Built

displacement of the longitudinalstrengthrnernbers in determining the stresses in each

structural member ofthetransversering.This was done accordingto Reference 12.

Loading conditionsused inanalysisincludedalsotheactualsea conditionsre-

pOrtedby the vesselduringtbe stomn thatcaused the damage in order to obtaina rea-sonablefailureanalysisas well as trans-verse strengthreanalysis.

Ithas alreadybeen notedthatduringtheextxeme sea ccmd~tionsdamage was causedto thevesselwhich indicatedgreen vm.terinexcess of4 meters above the deck. TheMaster,s reporthad indicatedthatseas ashighas almost 9 meters bad occu=red. Anaverage head of4.88 meters (16feet)ofwater, actingon the frillbeam ofthevesselwas used inthe reanalysis. The rnaximmm

Pressure head inthiscOnditiOnis 21.48meters, The wave heightinthiscase can beexpressed as:

Hw . (Depth+ 4.88 - Draft)x 2=(16.6 +4.88 -12.2 )x2= 18.56m

When compared withthepreviously

mentioned wave heightbased on class rulesof 8.9 mete?., thisis indeedan extremelyabnormal sea condition.

Not yet consideredare tbe dynamiceffectswhich undoubtedlyexist. Generallyspeaking,whenever a vesselencountersabnormally extreme weather conditionssnchas a hurricanenecessary measmes wonld betakenwithoutdelaytoproperly orientthevesselto avoidwashing ofwaves over thebroadside ofthevessel. However, itisconceivablethata vessel,when strnckby alarge wave from one side,may heel greatlyand be struckagainby another largewavebeforeproper correctiveactioncan be taken.In a case where thisvesselencountersaclassificationsocietyassumed wave heightof8.9 meters when itisheeling20° to o“e sidein fullloadcondition,itsdraftis:

Draft= 12.2+ B tan 20° + 4.45 . 22.7m

-Z

When tbe vesselis struckby a wave 6.4meters inheightwhen itisheeling20° theeffectivehead is 21.48 meters, the same a$thatused inthe reanalysis.

+G6

The above briefexample was presentedto show the reasonablenessofusinganeffectivehead ofabout 5 meters over thedeck in structuralanalysesofthistype.

Tbe tnostsignificantloadingconditionsnsed in tbe reanalysiswere those of Figure 8and those of Figure 10.

The transverse strengthoftbe ship’shullintbeheeled statewas examined by theuse ofa sophisticatedform ofthreedimens-ional analysisofReference 13.

Draft Rough Sea Condition,FullLoad,

Upright~ HeelingCOnditiOns

Figure 10 - Loading ConditionsAssumedAt Time ofReanalysis

Tbe followingresultsofthe reanalysisindicatedtbe following:

1. Tbe asswrmd classificationwaveheightwas probablyunrealisticconsideringthe extreme seas towhich vesselsare subjected.

2. The computed stresseswbicbresultedfrom the assumed extremeloadingconditionindicatedthattbelack of continuityof structureinthe faceplatesofthebottom anddeck transfereemembers and i“the strutw.s themost importantcause of failurewhen thevesselwas exposed to repeatedsevereirnp.actloads caused by tbe forceoflarge seas breakingagainsttheship.

3. The tmbalanceddesignofthe strut,theupper faceplatewidthbeingthree times thewidth oftbelowerfaceplate,was presmned to haveadded to thefixedend bendingmoments which are, theoretically,consideredtobe large,than foranniform cross section. This isdiscmsed in Reference 10.

At tbetime of reanalysistbe 1968classificationsocietyrulesand recentnoticeswere appliedto thisdesignto compare theexistingvesseltotbe then c=rrentdesigncriteria.This is shown in Figttre11. Whencompared with the original scantlings of

Figure 9 itmay be seen thattbeprincipalchanges occur in the sizeofweb platethick-ness, relocationoftrippingbracketsat tbestrut,and a balanceddesignofthe strut. Tbelack ofcontinuityinthe faceplatearrange-rnentremains, however.

Itis understoodthatthe classificationsocietieshad consideredtbeproblem of

PraCtiCal design head criteriaand concludedthata designcondition“singa draftequaltotbe depthat sidewould be adequate. As arestdtexternalpressures greaterthantbedepthat sidewere notused when meetingclasscriteria.

Tbe reanalysisoftbetransverse struc-turewas completed, ofcourse, withtbeideain mind of repairingand preventingareoccnrre”ce ofthedamage. Utilizingasnmch oftbe existingdesignas possible,reinforcementofthetransverseringwasaccomplished accordingto Figure 12.

The principalpointstonoteinthereinforcementare tbe removal ofthe faceplatethicknessdiscontinuities, thebalancingofthe strutdesign,theuse offlatbar panelbreakers to rednce bucklingpossibilities,andthe relocationofthetrippingbracketsattbestrntends.

The vesselhas been operatingfreeofstructuralproblems sincethe reinforcementswere completed.

CASE 11 - TRANSVERSE STRENGTH

This second emmple referstotbe same

generalsubjectOftransversestrengthOfatankerbniltintbe mid 1960!s. Unliketheexample in Case I, thisvesseldidnot fail

priOrtO structuralreanalysisand reinfOrce-rnent. As a resultofthe Case I failureandnumerous reportsof structtmalfailuresintankers ofthisvintageitwas deemed prudentto conductan investigationintothe vessellsstructnraladequacy usingthe latestavailabletechniquesand applyingextreme, but realisticassumptions regardingdesignhead.

Vessel - 1965

This is an example ofa classof 95,000DWT Tankers constrictedin Japan in 1966accordingto classificationsocietyrulesof1965,

Tbe plan view and principaldimensionsare givenin Fignre 13.

The designreflectstbetypicaltrendsatthe time offew largecargo tanks,midship

permanent ballasttanks,and reduced scant-lingnotationin the record. The designdifferedsomewhat from tbemmmal tankerarrangement in thatthe centerand wing tankswere allofthe same width.

Tbs scantlingsoftbetransversering,asbuilt,are shown in Figure 15.

G7k

Figure 11- Case I - Midship SectionUtilizing1968 Rules and Notices

Figure 12 - Case I - Midship SectionAs Repaired

G8

tie,4 C,T, 1?W.B.T NO’? C.T. No. 1 CT

I: do,+ C.T No 3 C.-r Not C.T NO ! C.T ‘+,%E0.

PJO.&cT PVI.67 140.2Cr floICT

Case 11Plan of Cargo Area

Principal Particulars Metric English

Length (B.P. ) 264.6 8688-01,LW L 269.2 8831-0,!Beam (MLD) 38.95 127,-9,,Depth (MLD) 18.90 621-o,,Draft (extreme) 13.3 431-7 112~,Deadweight 95,000

Figure 13 - Plan and Particularsofa 95,000 DWT Tanker

The crite~iafor checkingthetram -verse strengthofthisdesignat thetime itwas developedwas essentiallythe same asthatdescribedin Case I. The structuraltestconditionsas wellaB the most criticalcases ofthe vessel,s expectedloadingpatternwere analyzedin accordance withclassificationeocietyrecommendations.

Tbe classwave heightformulaHw=I.026 x Lw 0.4 resuitedina wave heightof9.3 meters. The effectivehead daringtheworst assmned operatingconditionfullyLadenapplyingthe classwave is:

Head . Draft+ Hw-z-

= 13.3+ 9.3 = 18 metersT

This assumed worst effectivehead isalmost one meter below the deck at side.

StructuralAnalysis - 1968

The detailedframe analyseshy the

Xcugh Sea Condition- 4.88 Meter HeadOver Main Deck

m=.gure 14 - Loading ConditionAssumed At

The Time of Reanalysis

methods pre.rionslyomtlinedin Case I wereperformed. These includedthetwo dirnen.sional,’displacementmethod,,and thethreedimensionalanalysisconsideringthe relativedeflectionofthewing tanks. The analysisofthe V.sselintheheeled conditionw.s notdone.

The variousloadingconditionsstudiedwere identicalto Case 1 assuming water to be4.88rn or 16 feetover the main deck, centertanks frilland wing tanks empty, Figure 14shows the worst conditionconsidered.

The resultsofthe reanalysesindicatedthathigh combined stres~esexistedinthedeck transversesand high shear stressesexistedinthebottom weh near thebilge. Thestrntindicateda propensityto buckle.

Correctiveminforcenmnt was under-takenas indicatedin Figure 16.

The pointsto noteare the correctionofthe discontinuousfaceplateofthe deck trans-

verse, panel stiffenersi“the strutand theinstallationofmore shear area in the form ofa diagonalbilgebracket,

CASE III- DOUBLE BOTTOM FLEXIBILITY

This thirdexample ofthelaginthe

applicationoflateSttechnologicaldevelop-ments to thepracticalrealm dealswith the

problem of entineroom doublebottom flexi-bilityin largetankers. A* tankers rapidlyincreasedin deadweighttormage duringthe1960’s,shiptslength,beam, depth,draftandeven block coefficientwere increased. Tbeincreasein shipsbeam and draft,slightincreaseinblock coefficientand the refine-ment ofweldingtechnologyresultingin moreefficientand lighterstructures,alIcontri-butedtowsrd greaterflexibilityin engineroom

?---

—.,.

G9

.—

1 1 I

298 X11 Fe,

kid w

Figure 15 - Case 11 - Midship SectionAs Built

I

Figttre16 - Case II - Midship SectionReinforcements

G 10

.

steelwork.Contrary to the trend of greater flexi-

bilityof ship,s structure was the trend in

greater stiffnessof the main p.oPmlsionrnachinet’yassociated with higher powers

applied to single shafts and lower mainengine RPM. The relativeinccvmpatibility

became more and more significantas the sizeand power of vessel, increaeed, Serious

physical problems were the resmlt: impropershaftingalignment, damaged main redaction

gears and main engine crankshaft andhearing damage. These problems are caref-ully analyzedand reportedupon inReferences 14 and 15.

This example refersto the classof

95,000 DWT tankers describedin Case IIabove, The planview and dimensions are

givenin Figure 13.Shortlyafterdeliverysignsofpitting

were noted on the forward helixesofthe mainhullgear and low speed pinions. The wearw.s attributableto faultyshaftalignment.Only aftercom+idera,bleinvestigationcouldit

be confirmed thatthe doublebottom d,fo,~-ationwas markedly greaterthanhad been

pre,umed duringthe d..ignand cOnstru.tiOnstagesofthese vessel,.

Design k Ccmstrmction. 1965

During the design stagesofthisclassofvesselstheanticipatedhulldeflectionwascalculatedusingthe conventionallonghandlongitudinalstrengthcomputationmethods fordifferentconditionsofloading. The resnltsare illustratedin Fi.gnre17 and show a fairly

typicaltanker deflectioncurve. ltcan beseen thatthe characteristichogging conditionsexistin lightshipa“d ballastconditions,ThefuR loadconditionshows almost no deflectionaft,but some saggingforward. The effectofempty midship ballasttanksis evidentinthisfigure.

An enlargement ofthe engineroomportionOfthehullgirderdeflectioncurve ispresentedin Figure 18. From thefigureitcan be seen thatthe engineroom deformsessentiallylinearly. There is indicatednpto Zmm upward deviationfrom a linear

projectionfrOm the sternframe tOthe engineroom forward bulkheadinthe fullyloadedand ballastco”diticms,However, the relativedeflectionofthe engineroom doublebottomfrom thelinearbetween thosetwo operatingconditiom doesn-tchange. For example,forward of Frame 29, thelocationofthe mainreductiongears, the deviationfrom thelinearis 2mm inbothloadedand ballastconditions.The conclusionwas to neglectthe change indeformationofthe hullas a girderforprac-ticaloperatingconditions.Proper shaftalignmentforthe ballastconditionwasconsideredtobe validforallthe operating

conditionsbetween ballastand fullload.Local deformationsdue to variationsin

draftwere neglected. Ithas properlybeen

pointedoutin References 14 and 15 thatasship sizesincreasedfrom 50,000 DWT to100,000 DWT and upwards littleconsiderationwas givento the draft-engineroom beamrelationship.The leugthofthe doublebottomfloorsisa directfunctionofthebeam oftheship. When treatedas a beam withan evenlydistributedload,hydrostaticpreesnre d~e todraft,the deflectionisproportionalto thefourthpower ofthe floorlength. All cdherthimg.sbeing constant,doublingt~ beam ofa vesselresnltsin deflectionsbeing increased16times.

With the cl.ssificationsocietyrd.srecptiringdoublebottom depthto increaseas afunctim ofvesselbeam and draft,actualdeflectionsprobablyvary somewhat 1.ssthanthe cube ofthebeam, With the donblingofvesseltsbreadththisstillrepresentsan,,~nhealthy,’increasein deflectionby a factor

ofalmost 8.

Analysis- 1970

Investigationintothe gear problems andrationalizationof shaftingalignmentwas

directedtoward determinationof steelworkdeformation, This was carriedoaton a two

p~Onged frOn~ 1)analyticaland 2) fullscaleinstrumentation.

The importance of determiningthelocaldeflectionofthe doublebottom undev thebearing foundationswas recognized, A finiteelement analysistechniquewas appliedutilizinga newly developedcomputer pro-grammed Matrix Method ofStructuralAnalysisof Framed Structures. The resultofthisanalysisofthe steelwork as a complexbox stru.tm-eindicatedan expecteddeflectionpatternatthe shiptscenterlineto be asindicatedin Figure 19 fora nniform load

equivalenttO 6.1 meters (2ofeet)draftvariation.The calculateddeflectionincreas-ed from O attheafterpeakbulkheadto O.47mmattheafterend ofthe gear case, O,85mm atthebullgear forward bearingto a maximumof 3,88mm at Frame 43, wellforward ofthe

gearingand lineshafting.Eqre~sed interms ofdeflectionat tbebullgear forwardbearingper footofvesseldraftafttheexpecteddeformationamonnted to O.42mm/ 10feet.

In order to corroborateresultsofthefiniteelement analysisso thatadequatecompensation would be made forlocaldeform-ationin way ofthelineshaftsnpportpoints,actualmeasurement ofthe deflectionofthestructm’ewas undertaken. This was acccun-plishedby two independentmethods:

1) A pianowire was stmmg from theforward bulkhead ofthe engineroom tothe

G 11 f---

AP F54nlm

—. _ LIGHT SI+\P

&oSEE FIG,19 BALLAST \ \

5b

fULL

&-. _-— - /

Fignre 17 - DeflectionofHullGirder

2+4.LIGH7SHIP+ BI%LI-AST

s FLILl-(j LINEAR VALUE II

50

I33

gl)

(:) -1

Figure 18 - Deflectionof HullGirder inWay ofEngine Rocnm

afterpeak bulkheadat theafterend oftheengine room alongthe vesselcenterline.This set-upis illustratedin Figure 20.Micrometers and clock gages were positionedto measure movement in way ofthe gearcasingas the vesseldraftvaried duringloadingand discharging.

2) A laseTinstrumentwas installedonthe vesselcenterlineat the engineroom afterbulkheadand directedtoward the engineroomforward bulkhead. At intermediatepointsalongthe vesse1°scenterlinethe deflectionofthe doublebottom in relationto thelaserbeam was measured. The laserdevice

prOved to b. highlysensitiveto vesselvibra.tionsand the resultswere somewhat less

Figure 19 - Deflectionat ~ DLIeto

6.1Meter DraftChange

L FULL-EMPTY—.—$ WIRE -=-+ ,=ULL-bhLUST _+—

FULL-HALF LOAD‘*-Z3

Fls F40

Figure 20 - Measured ~ Deflection

repeatablethan forthepianowire. Not with-standingthis,thelaserproved usefulincorroboratinghulldeflectiontrends.

The resuitsofthe shipboardmeasure -ments are givenin Fignre 20. Not onlywerethe deflectionsconsiderablygreaterthan

AFT DRAFT VARIATION- FEE1

Figure 21 - Deflectionat BullGearForward Bearing

calculated,butthe pointofmaximum deflec-tionwas much furtheraftthan anticipated.Inthiscase themaximum deflectionoccurredatthe main gear casing.

The deformationatthe bullgear forwardbearingas a functionoftheafterdraftvari-ationisplottedin Figm’e 21. The slopeof1.65mm per 10 feetof draftis almost fourtime, the calculateddeformationofO.42mml10 feetpreviouslymentioned. The canses forthepoor analyticalr.suitsare notknown butare as.mmed tobe primarilyda. to faultybonndary conditionassnmptions and disregardofthe effectsofinterconnectingstructuresand pillars.

The lineshaftingofthisclassofvesselswas realignedtakingintoaccountreactionsdue to verticaldeformationofthe localstructureas wellas thewhole shipas a

girder.

Todayns Tankers

In view ofthese variousparametersworking to createcertainincompatibilitiessbetween thehullon one hand and tbemain

propulsiveequipment on the other,One Oftwo choices seem, available:

a) Reduce the stiffnessofthe shaftingand therebyadjustthe equipment totheflexibilityofthe strncture,or

h) increasethe stiffnessofthe fonnda-tionsand doublebottom structureand therebyadaptthe structureto the reduced flexibilityofthe machinery.

Tbe lattermethod is recommended.Serious considerationshouldbe giventoincreasingthe depth ofthe engineroomdoublebottom floorsand to increasetbeeffectiverigidityoftbe enginefoundations.Mounting thepropulsionplantas faraftas

practicablewilllocatethe equipment wherethe floorbreadthisless and thus reduce theeffectof doublebottom deflection.Reference14 and 16 have pertinentdiscussionson thispoint.

Sunmnary and Reconmnendatiorw

Continuedcooperationand conmnunica-tionamong builders,operators,designersand researchers are needed particularlyinareas offullscaleresearch to continuepracticalapplicationofadvanced technology.

Three specificcases givingtwo basic

problems encounteredwithtankersdesignedand builtinthe mid. 1960,s have been present-ed. With referenceto transverse strengthone case of severe damage was described,evaluationofwhich ledto reinforcementoftbestrncture. Utilizingadvanced structuralamlysis techniquesreanalysisofotherexistingvesselsledto theirreinforcement.As a resnltofthe experiencesdescribed,thefollowingare indicated:

~) Itis recommended dnringvesseldesignto utilizethe latestfiniteelementanalysistechniquesto determine stresslevelsin allmajor areas ofvessel structnre.

2) Use realisticloadingassumptions,being cognizantofthe extreme conditionstowhich the ocean environment snbjectsvessels.Pres sure head assurnptio”sapproximately5 meters above themain deck are consideredappropriate.

3) Major problems oftenoriginatedue to lack ofproper attentionto ratherbasicand apparentlyminor details.

a) Maintaincontinuityof struc-ture and avoid abruptchanges in section.

b) Strutdesigns shouldfavorbalanced or uniform cross sections.

c) Adequate panel stiffness

againstbucklingsbO~d be prO~ded.

The relativeflexibilityoflarge shipengineroom doublebottoms conpledwithhigh

p.wer, low RPM main propulsionequipmentresmlt8 ina basicincompatibilitybetweenshipsmachinery and itssupportingstructure.Itis recommended thatthe doublebottoms oflargetankersbe designedto rninirnizehulldeflectionsby deepeningdoublebottoms andincreasingthe effectiverigidityoffoundations.

G 13 f---

REFERENCES

1. Robe,t A. Baker, Editor, “A Stre8s

Analysisofa StraplessEveningGown and Other Essays foraScientificA g.”, PrenticeHall,In. , 1963.

2. .7.H. Argyris, ,,Recent Advances inMatrix Methods ofStructuralAnalysis,’, The McMillan Co. ,1964.

3. ~. ~, Adam, , ,fimtes on Stressesin

Tanker Meznber e<<, InstitutionofNaval Architectsin London,March 1950.

4. E. Abraham sen, ,,Recent Develop-

ments inthe PracticalPhilosophyof ShipStructuralDesigncf,SNAME SpringMeeting,Paper No. 5, 1967.

5. A.S, Hall,,’Frame Analysis,!,

John Wiley & Son, Inc. 1961

6. I,Ice. Strud. 1, The StructuralDesign

Language EngineeringUsersMannal,,,Department ofCivilEngineering,Cambridge, Mass.

7. StructuralEngineeringSolver(Stress)Program ofthe IBM 1130,Model 2B (1130-EL-03X).

8. F. Bleich,,,BucklingStrengthofMetal Strictures<<,McGraw-HillBook Co. , Inc., 1952.

9. B.G. Johnston,“Guide to DesignCriteriaforMetal CornpressionMember,c , John Wiley & Son,hlC. 1966.

10. T. Wah, Editor,,tGuidefortheAnalyeisofShipStructures”,U.S. Dept. of Commerce, Officeof TechnicalServices,Washington.25,.D. C.

11. K, Bre.ler, ,,Strength of Plate

Girder in Shea,”, TransactionsASCE, volume 128, Part II,1963.

12. M, Mori & mbers, llOnthe Trans-

verse Strengthof OilTankers”,JournalofS.N.A.J. , Vol. 121,1967.

13. M, Mori & Oth.rs, “On the Trans-

verse Strength of Oil TankersUnder the Heeled Ccmdition”,

Journal of S.N.A. J., Vol. 122,

1967.

14. G. VOlcy, “Damages to Main BearingsDue to ShaftingA1igmnent”,

I.M.A. S., London, 1969.

15. G. VOlcy and Others, “Deformabilityofthe Engine Room of LargeT?.nkersr’, InternationalShip-buildingProgress, Vohtrne21,August 1974, No. 240.

16. Bureau Veritas,“Recommendations

Designed to Limit the Effectsof VibrationsOnboard Ships’,,May 1971.

G 14

D1SCUSS1ON

Huynh d.. B.., Visitm

This paper is, in several aspects, verylaudable. The e.thor sbo.ld be particularlycommended for tbe very practical manner in whichthe damages have been analyzed and the correc-tive measures explained. Indeed, most unfor-tunately, today’s ship designers having to dealwith .stremendous amount of computer’s printouts, seldom can indulge into lengthy meditationover ...s.1 relationships between loads andstructural behavior.

In regard to different questions raised bythe author a few answers can be provided. Firstof .11 the shipowners sbo.ld not be over-anxiousof b uying ~fyesterdayqsproduct in today,smarket place.“ Indeed, Classification Societiesalways place the highest priority into structuralintegrity regardless of Rules changes. For obvi-ous practical reasons Classification Rules .a.-not be updated on a daily basis to reflect tbelatest in house or otherwise acquired techno–logical improvements. However, when reviewingthe r easonableness of a given design, BureauVeritas members always apply tbe most up to datetechnique known to them. h. a matter of fact,the “ Shadow Rules” reflecting tbe latest changes(to b e published) are simultaneously used withthe existing R .1.s for comparison purposes.This practice most certainly is also adopted byothex Societies.

16ithreferences to case 1 transversestrength damages this discusser would like to:

(a) fully agree with the author’s judge-ment regarding t he unbalanced design of thestrut.

(b) seek furtherclassificationco.c,rn.ing conclusion(2)of page G7. Did the 1968reanalysis, include investigation of the trans-verse ring’s behavior under dynamic loads(shocks)? What was the e..tbor’sjudgement ofthe quality of workmanship particularly in areasof disccmtinuities in deck and bottom transversesface plates?

(c) Call the author’s attention to theusually lack of significanceof tbe mentionedtest condition (center tank full, wing tankempty) regardless of the vessel,s draft overthe behavior (stress) of transverse membersThi. loading condition is primarily aimed atchecking the scantlings of the strut(s) ltthus vo”ld be interesting to know whether a re-run of t he transverse analysis with reinforce-ments made only to the $tr.t,s scantlimgs wouldmodify the stress distribution.

(d) In connection with the sbo”e, deEer–mine whether che damages presented the same

degree of severity in way of transverse rings

where the struts may possibly be reinforced by

transverse end stringers of the wing tank. trans–“,,s, bulkheads.

(e) disagree, however respectfully, withthe author regarding the absolute necessity ofavoiding discontinuity in tbe face plates. in-deed, several designs of this type, have been

proven s.cc..sfu1 in servic.. The deck, =dbottom transverses are stressed differently thanthe side,, or bulkhead,s transverses. In obviousarea., more section is needed for shear (eitherby effect of external pressure or by forces d“,to the relative deflection between the side shelland tbe longitudinal bulkhead) Thus weight andcost constraints compel one to optimize thescantlings of the face plates. The importantaspect is to carefully provide for a smoothstress flow by proper tapering as well .s .!de-q.ate tripping brackets. The discontinuity issomewhat dramatized in this ,,ro.ndedface plate,’design as opposed to the Europe.. straight de-sign (with transition brackets) Each of bothdesigns b.. its own merit. The obvious i“con-venie”ce in tbe design of this vessel resides ina difficult stabilization of the strut i. way ofthe Confection with tbe side!. and bulkhead,stra”s”erses

(f) ask tbe a.tbor to provide for more in-formation regarding the percentage of stress i“.crease owing to higher pressure bead, particularlysbeari”g stress imposed by bigber relative de-flection shelllb.lkhead in bottom, strut a“ddeck tra”sverse$, Indications relative to theheeled condition should be very much instructive.Similar calculations performed by this discussero“ OBO of 250,000 d.dtr.+”gehave failed to i“di-cate needed rei”forceme”ts, taking into accountthe scantlings as required by other loading con-ditions (full or light ballast)

Finally the author,s recomme”datio” (3.)

page G13 is wholeheartedly agreed to. It issomewhat strange to notice in several designsthat a.pproacbto panel stiffening .agai”stb“ck-li”g bas bee” co”d.cted in .11 but the mostnatural way consisting (as w.. adopted in tbeauthor’s repairs) in fitting stiffeners parallel,close to the face plate (1/3 of web!s height forinstance) where axial stress is expected to behigh a“d vertical flat bars in areas where .sbearstress is at its peak;eg,between the above “panelbreaker” and the primary 1oIi8it”dinal stiffeners.

.

AUTHORS‘ CLOSURE

Mr. H.ynh d.. B..’. comments are .sppreci-ated

l.lithreference to Item (b) of his discus-sion, no shock analysis was done during there-.analysisof either of the subject cankersStatic equivalents were used reflecting theestimated head experienced by vessel A in heavyseas. Much work has been done and a lot hasbe.,,written about assumed practical heads forprimary strength determination. ue.allymeasured stresses are compared against stressescalculated from theoretical waves. The authorbelieves more work can and should be done inthe area of assumed prsctical head for localstrength wherein actual stresses are similarlycompared t. those theoretically derived.

The structural test conditions mentionedi. paragraph (c) is considered a useful and

G-16

inexpensive way to check the structural design.Obviously an 8 foot head in tbe center tank withwing tanks empty is not quite the same .s sucha head loading the vessel externally as is ac-complished by wave action.

TO answer paragraph (d) all webs werefractured. The influence of the transversebulkheads could be noted only inasmuch as thefm’wardtuxt and aftermost webs showed less dis–plarame”t than the center web. The authoragrees with the discusser’s comments regardingtbe importance of proper tapering. It is, inthe author’s opinion, a method recommended foravoiding discontin.ities.

Flat bar stiffeners in the web were deter–mined necessary as . result of the heeled candi-tion by a factor of 1.2 to 2.0 than in similarloading conditions with the vessel unbeeled.