.

THE SOCIETY OF NAVAL ARCHITECTS AND MARINE ENGINEERSWI Pavonla Avenue, Suite ~, Jersey Ctty,New Jersey 07306 USA

Pap pressnmdat tlm hkrine SWCUA lns~tim, Maintenan=, and MmhwirtgSpnpc6iumShetatw”Naiiwal Hotel,Mington, Wginia, krch 1S-19, lW1

StructuralOverview,

Maintenance for New and Existing Shim:Fatigue Cracking and Repairs

R.G.Bea,R. Pollard. R. Schulte+trathaus and R, Baker..Universi~ of California, Berkeley, California

ABSTRACT

This paper summarizes the objectives, approach,and organization of a joint industry - government spon-sored cooperative research project focused on developmentof engineering technology that can lead to improvements inswuctural maintenance for new and existing tankers, Theproject is being conducted by the Department of Naval %-chitecture and Offshore Engineering in behalf of nineteenparticipating organizations representing governmentregulatory bmlies, classification societies, new-build andrepair yards, and ship owners and operators. Initial resultsfrom several of the stu~les that comprise this project areSun-lmarized.

INTRODUCTION

A two-yea international JointIndustryProjecton“StructuralMaintenanceforNew and ExistingShips”(SMP) was initiated by the Department of Naval Architec-ture & Offshore Engine@g at the University of Californiaat Berkeley (UCB ) in June 1990, The project has twotechnical gcxk

To develop practical tools and proceduresfor analysis of proposed ship structuralrepairs in order to minimize time andmaterials within the constraints ofregulatory and class requirements andprudent engineering practices, and

To prepare guidelines for the cost-effective design and construction of lower-maintenance ship structures which alsofacilitate future inspections and repairs.

The Joint Indusn-y Project was formed in responseto two recent trends in the shipping industry. The firsttrend is the well-documented ageing of the existing “fleetofmerchant ships, particularly of tankers. Such ageing isleading to significantly increased scopes of sn-uctural re-pairs and thtir associated costs and days out of service,

The second trend is the recent boom in ship con-struction. Many new ships are bging designed <byship-building yards and are being reviewed by owners, classifi-cation sccieties and government agencies. The heightened

“ interest in double bottom and double hull configurationsfor ships has generated new concerns related to theirsmtcturdreliability ~d futui-emaintenance costs.

In order to better meet the challenges posed bythese recent rrends, the research in this project has been fo-cused on two primary aspects of s~cmral maintenance:

Fatirme effects on the performance ofcriti;al internal structural components ofexisting and new ship hulls (includinghigh strength steel, reduced scantlingdesigns), and

Corrosion effects on the critical internalstructures of existing and new ship hulls.

The goal of the SMP is todevelop engineeringprocedures and PC. based computer programs that can as-sist ship owners, operators, classification agencies, andgovernment agencies in accomplishing effective and eft3-cient sn-uctural maintenance and life extension for ageingships. -Each procedure and program will be verified byapplying these to real-world problems.

This project is directed toward improving engineer-ing technology to make realistic fitness for purpse evalua-tions of ship hull smuctures , and to help, answer the twokey questions:

1) How should I fix critical internalstructural components ?

2) How can I design better critical internalstructural components ?

In addition to its technical objectives, this projecthas important organization objectives. The project is in-tended to provide a common, neutral ground for the con-structive interaction between ship owners and operators,ship classification societies, governmental agencies andship building and repair yards. The development of in-formed consensus approaches to the problems associatedwith structural maintenance of existing ships and designof new ship hull structures is expected to provide signifi-cant benefits to the ship industry.

This project is one of four inter-related cooperativeresearch projects being conducted by the Department ofNaval Architecture & Offshore Engineering. The titles,sponsors, and objectives of the other three projects me asfollows:

Frror in ODeWi.ws of Marine Svstm -Sponsoroi by the Sea Grant Program and sevenindusrnal - government agency sponsors, the ob-jective of this project is to develop and verify engi-neering analysis produres to assess alternativesto reduce the effects of human and organization er-rors in operations of tankers.

II-A-1

Ine Str~ritv Pro~ra~fkl.!Wl - Sponsored by the Ship StructuresCommittee, the objective of this project is to de-velop a procedure for definition of advanced ma-rine structural integrity programs based on currentdevelopments in Airframe Structural IntegrityPro-grams (ASIP).

. .and llltv of T- StrucW

- Sponsored by the Maritime Administration andconducted under the auspices of the UCB NationalMaritime Enhancement Institute, the objective ofthis project is to assess the impact of inspection andrepti programs on the reliability of tanker struc-tures.

The objectives of these four projects are fwused ondevelopment of a comprehensive engineering technologyfor the improved maintenance, design, and operation ofshipping fleets, with a specific focus on very and ultralarge crude carriers (VLCCS and ULCCS), including bothstructural and non-suuctural (e.g. human and organizationfactors) aspects.

PROJECT “ORGANIZATION

There are four major organizational components inthe SMP. The first component is the project sponsors andparticipants. At the present time, there are nineteen na-tional and international organizations including ship own-ers and operators, ship construction and repair yards,classification societies and government agencies that com-prise the first component. Table 1 lists the names of theparticipating organizations.

The second organizatiomcomponent is the Project -Technical Committee (ITC). Each of the project sponsorsand participants are represented on the PTC. The PTC ischaired by the American Bureau of Shipping (ABS). Thepurpose of the PTC is to provide the project investigatorswith directions on technical goals and “objectives, with in-formation and data to assist the project, and to monitor theproject budget and schedule.

The third organization component is the Office ofResearch Services and Sponsored Projects Office at theUniversity of California at Berkeley, This component isresponsible for the contracting, invoicing, and project ac-counting.

The fourth organization component is the projectinvestigators. Table 2 summarizes the names and respon-sibilities of the project investigators.

BACKGROUND

Since the turn of the centuty, ship hull design hasseen significant progress. Examples of this progress in-clude materials technology, seaway loading technology,and computerized loadings and souctural analysis. Theeconomic benefits of this progress to ship owners and op-erators have been far-reaching.

The advancement of shipbuilding technology alsobeen significant. Automated plate cutting and welding,CAIXAM, zone outfitting, and mmlular construction aresome of the tools avaiIable to shipbuilders that were notavailable in the past, and that have resulted in the ship-building revolution.

In stark conuast to the two success stories ofde-sign and construction is the field of ship repair and mainte-nance. Today’s vessels sre repaired pretty much the sameway as their predecessors were at the turn of the century.Steel weights and coating areas are mostly calculated byhand. Most ship maintenrutce records are kept on paper.Much inspection, maintenance, and repair is still done by“rule of thumb. ”

From this history, one might conclude that shipmaintenance and repair is a relatively less expensive. and’less important undertaking than ship design or construc-tion. Anyone in the marine business knows that just theopposite is tie. A ship is built and constructed in a veryshofi period of time, but it still needs to be maintained for20 years or more. As the ship gets older, the maintenancerequirements and costs increase significantly. In the lateryears of a ship’s life, many difficult questions are raisedwith regard to the desirability and feasibility of extendingits life.

Given the present state of ship design and comstruction technology and the present state of the world’sageing fleet of ships, it is time to further develop and up-grade the technology of structural maintenance of new andexisting ships. The following factors influence these de-velopments.

A Iarge proportion of the worlds tanker fleet is ap-proaching the age of 15 years. Steel renewal re-quirements are increasing with a consequent in-crease in time out of service. The process. of in-specting a vessel, writing the repair specification,making repair drawings, calculating the steelweights and coating areas, etc., is v“erylabor inten-sive. The workload is increasing with each pass-ing year as the fleet grows older.

Life extension of existing ships beyond 20 years isbecoming more atwacrive as the cost of new build-ing skyrockets. Yet, it is difficult to conduct aproper economic analysis of the two alternativesbecause estimating the future repair requirementsand costs for existing ships is a time-consumingtask.

There is increased public scrutiny of ship opera-tions, particularly tanker operations. There is nosuch thing as an insignificant oil leak in today’sworld.

Tanker structure information data bases are difficultto assemble, analyze, and rerneve to detect andmonitor dangerous trends in pitting, cracking, cor-rosion, and coating failures.

Given a decision to buiId new ships, there me sig-nificant concerns that the standards and promduxesused to design and constructthese ships will resultin a vessel that can be efficiently and effectivelymaintained. Even in some recently constructedULCCS, there are unsettling indications that in thequest for lower structural weights and initial coststhat reasonable levels of durability have been sacri-ficed.

The mare general use of higher tensile steels, useof lighter scantlings of higher tensile steel, and therequirements for hull structures that will havegreater degrees of safety against cargo losses givengrounding and collisions indicate a greater need topay more attention to detail design, design to facili-

II-A-2

tate inspections and maintenance, providing suffi-ciem stnictuml reserve, and providing robust hullstructures that can tolerate defects and damagewithout significant 10SM+in capaciry and safety.

The SMP seeks to bring ship suuctuml repair tech-nology to a level commensurate with today’s engineeringneeds and capabilities. It is intended to help quip organi-zations with powerful,’, yet practical, analytical tools forship repair and maintenance. Based on the experience withthe past generation of ships, the project also is intended tohelp develop guidelines for new builds that c~ result inhull structural systems that will have higher degrees of in-speaability, maintainability, and.dttmbiliry.

Most importantly, the prpject is intended to providea common ground for ,tie interaction between ship owners,ship classification swieties, governmental agencies, andship buildhtg and repair yards (ship industry). The experi-ence of these groups will provide the basic starting pointsfor this project. The development of informed consensusapproaches to the problems asswiated with repairs toexisting ships and design of rtcw ship hull suuctures tofacilitate inspections and ,maintenance is a key objective ofthis project.

A principal objective of this project is to betterequip the ship industry to extend the useful lives of exist-ing ships and to define the characteristics for design ofnew builds that will profit from the lessons of the past. Torealize this objective, the industry needs to pull togethertoward a comrtion set of goals. Ship owners and operatorsmust take the initiative to manage and not be managed.Public initiated “legislated naval. architecture” must beavoided. Development of guidelines for improved designand repair are,key aspects of this management, and will bea key aspect of this project.

Navat architects and ship maintenance-repair engi-neers need to have better guidelines and twls to accom-plish their work. Development of improved guidelinesfor both existing ships and new builds-to help better min-imize corrosion and fatigue cracking, problems, and devel-opment of computer programs to assist these engineers area key aspect of this project,

Ship builders and repair yards have responsibilityfor quality construction and repair. They must have thetechnical tools and other resources required to deliver thenecessaq quality. This project is intended to providesome of the technical tools that can assist in improvementsin repairs. and design of critical smuctural components innew builds.

Ship surveyors, classification agencies, and gov-ernmental agencies have ,responsibility for quatity inspec-tion, verification, ind encouragement of the ship industryto do what is right for the industry and the societies that itserves. This means helping mamtain the economic viabil-ity of a critical industry, and defining those guidelines andrequirements tba~will result in acceptable perfmrnance bythe indum-y. This project is intended to contribute to thedevelopment of such guidelines and requirements.

Experience with the maintenance and life extensionof existing ships has indicated that there are two key prob-lems that must be addressed if maintenance costs are to bemanaged within acceptable limits, and if the structural reli-ability of the hulls are to be maintained.

1) Corrosion of critical internalcomponents, and

structural

2) Fatigue cracking of critical internalstructural components.

In many cases, fatigue and corrosion have beeninter-related. In some cases (in particular in many hightensile steel, lighter scantling, HTS/LS , ships), designand construction methods have exacerbated fatigue andcorrosion problems.

Fatigue cracking and corrosion are a major concernbecause of their potential effects on hull and tank leak in-tegrity, and their pxential effects on the structural capacityof the hull. Repairs are costly , and sometimes, are inef-fective. Evaluating how to best repair cracks, and howand when to repair .cormsion in the most cost effectivemanner is not an easy task. More definitive guidelines andanalytical took that can help the surveyor, inspector, andrepair engineer make such decisions are badly needed.Development of such tools is a primary objective of thisproject.

While fatigue artd corrosion maintenance havesome very important ramifications with regard to life ex-tension of an ageing fleet, they also have some potentiallycritical implications with regard to consuuction of newships. If properly and well understood, experience witholder ships can provide some important insights into im-proved engineering and construction methods for new-builds. Development of guidelines for the improved de-sign of critical so-ucturat details and components of hulls isa prim~ objective of this project.

A substantial base of technology pertaining to theobjectives of this project has been developed by organiza-tions such as the International Ship Suuctures Congress[1,2], the Tanker Structure Co-operative Forum [3], theShip Structures Committee [4-10], the American Bureau ofShipping [1 1-15], and others [16-39].

Inaddition, many ship owners and operators havedeveloped advanced methods for maintaining their fleets[24,25, 39]. A starting point for each of the efforts in thisproject is this base of technology; fully utilizing availableengineering and operating experience. For example, theAmerican Bureau of Shipping and other organizations havedeveloped some very sophisticated analytical tmls to per-form fatigue and smength analyses. Several ship ownersand operators have developed and implemented advancedinspection and maintenance data archiving software.

A major problem in ageing vessel maintenance islmating structural failures and severe corrosion. Timingof inspections, access to critical areas, and coverage ofcritical areas with rust and cargo residues provide otherobstacles to disclosing cracks and corrosion. This projectis intended to address, but not necessarily solve such“realities.” One of the benefits of the development ofcondition survey data bases on ship hulls that have com-mon characteristics is to improve insights on when andwhere such problems might exist, and how they might bemost effectively found. Analysis of the hull sn-ucture canindicate which suuctural components need KObe watchedmost closely.

Experience with life extension of mtuine structuresindicates that the most severe problem is usually the lack ofdefinitive information on the current condition of thestructure. This is a problem of how, when, and where toinspect. This is a problem of starting and maintaining acomplete and accessible data bank on the structure, Thisexperience also indicates that if the su-ucture cannot be ef-

11-A-3

fecrively inspected, then it is highly likely that it cannot beeffectively maintained. It is not possible to realisticallyevaluate the needs for repairs and evaluate the safety andintegrity of the hull structural system without definitive in-formation on the condition of the critical components thatcomprise the system. Tlwgreater the knowledge about thecritical components, then the more realistic is the evahsa-tion, and the more effective the repairs and maintenance.

TECHNICAL APPROACH

Six inter-related studies define this research project(Table 3). The fatigue and corrosion damage evaluationsconstitute the basic studies in the project. These evalua-tions, however, cannot be completed without defining theboundary loading and fixity conditions of the leal detailswhere damage has occurred. Such boundary loads andconditions will be developed in Study 3.

Based on results from Studies 1-3, repair strate-gies and guidelines will be developed in Studies 4 and 5.Finally, software packages for personal computers withdmmmentation will be developed in Study 6. The follow-ing sections will describe in more detail the content of eachof these studies

1- F~e Ev~.

v

The objective of this study is to develop and verifyengineering approaches to assess fatigue effects on the per-.formance characteristics of critical structural details intanker hulls, including the effects of inspection, mainte-nance and repair. This study is addressing both mild steeland HTS/LS steel hull’ structural elements and systems.This study is organized into six tasks (,Table 3).

Stutlv 2- Corresion Damage Ev~

The objective of this study is to develop and verifyengineering approaches to evaluate internal corrosion ef-fects (general and pitting) on the structud srrength andleak integriy characteristics of critical (to strength and leakintegrity) components comprising existing ship hulls andnew builds.

Greatly accelerated corrosion rates have been ob-served in Imalized areas of low structuml rigidity in shiphulls. This appears to be due to the corrosion products(rust scale) being flaked off by the flexing of the compo-nent. This effmt is believed to be particularly important inHTS/J..S ships. This study is investigating the relationshipbetween lwal flexue of hull structures and corrosion rateswith the goal of recommending limits to local flexibility forboth mild and HTSI’LS. The study is organized’into ninetasks (Table 3).

This study plays a key role in that it provides inputand support to-the fatigue and corrosion damage effectselements of the project .“ The over all objective is to de-velop a reliable but simplified and practical andyticrd toolthat will enable the engineer to make the necessmy struc-tural system .performance evaluations rapidly and ‘with ac-curacies sufficient to make gcNMengineering decisions onrepairs and maintenance strategies.

The analysis of the interaction between critical in-ternal suuctttrai details,- e.g., brackets, and adjacent struc-tural components, e.g., webs and stiffened plate panels,must provide (a) an accurate and efficient model of the

load-displacement behavior of the detail in conjunctionwith the adjacent structural components, and (b) the smessdistributions at the element level for the fatigue, corrosionand repair evaluations. The study is organized into twoprincipal tasks (Table 3).

The successful completion of Task 1 and Task 2will provide the foundation for the the development ofi (a)a Iibrruy of iypical generic structural detail modules con:sisting of the detail and the adjacent shucture of sufficientextent to model the detail’s boundary conditions, (b) a cor-responding library of module loadings, and (c) the PCsoftware necessq to implement the analysis. These stepswill k carried out during following stages of the project.

The objective of this work is the development ofsimple and reliable procedures. To this end, much effort isbeing devoted to proving and calibrating the simplifiedmodels.

Such’ 4 Fatigue and Corrosion Reuair Ass- .rwlls -

The objective of this study is to develop and verifywith ship service data engineering guidelines for the eval- .uation of fatigue and corrosion repairs to critical swucturalcomponent: of existing ships, and to develop generalguidelines for new builds to help maximize inspectabilityand minimize repairs,

The fatigue and corrosion studies will provide ana-lytical abilities to project furure crack propagation and cor- ~rosion rates ‘and effects. The corrosion study also willprovide background on limits that should be placed on theflexibility of components to reduce corrosion rates.

The work of the Tanker Structural Co-operativeForum (TSCF) provides a vah.tabIestartingpointfor thiseffort [3]. As well, the special reports developed byCommittee V.3 (Service Experience - Ships), of the Inter-national Ship & Offshore Suuctgres Structures Congress[1, 2, 36] provide important direction for this effort. Inparticular, the TSCF has documented frequently occurringfatigue damage, and s~ategies to repair that damage. Anobjective of this study will be to continue and extend theTSCF developments. The study of fatigue and corrosionrepair assessments for existing ships is organized into fivetask (Table 3):

GuWmes for New Sh~. . .-

V

The Ship Structures Committee has initiated are-setwch project being conducted at UCB on the topic of de-velopment of Marine Structural Integrity Programs(MSIP) for ships. The project is addressing new buildship life-cycle phases, structural and non-structural(operational) aspects, inspections and quality conmol, andinter-relationships of design of new VLCCS and ULCCSand M.SIP.

In addition to a practicil approach that can be usedto develop life-cycle MSIP for new builds, the project isintended to define a gentil purpose computer based in-formation and evaluation system to assist in the life-cyclemanagement of the structural integrity of ships.

As a basis for the development”of MSIP, the studyis reviewing the U.S. Air Force’s Airtlame Structural In-tegrity Rogrsrn and the comparable program of the FederalAviation Administration. Results from the Ship StructuresCommittee sponsored resemch project will be incorporated

... ...

II-A-4

intothis project as appropriate to the objectives of thisproject.

This study is organized into four tasks that are fo-cused on improving the durability characteristics of criticalstructural elements (Table 3).

v 6 - Develop~nt of Srdlm.re and ArInlL

‘f?sisstudy, unlike the other technical studies, is fo-cused at providing the background, standards and supportso that the computer cmles developed by many program-mers will be of uniformly high quality, will permit easymmlification and be user friendly. As such, this study willprovide a uniform foundation and standard interfaceswhich will serve as a reference for all of the studies.

There are importaht reasons for developing thesestandards early and adhering to them strictly. In the past15 years we have experienced at least 5 different genera-tions of computer technology and there appears to be noend in sight of this rapid development. This means it iscertain that the,cding developed on the current generationof machines will be used .on future generations of fasterand more powerful machines. It is imperative, thereforethat the code have”portability and not be dependent onquirks or specialized features of any particular hardwareconfiguration.

Further, it “is anticipated that the computer pro-grams produced as part of this study will be “living” doc-uments. “That is the f~st version delivered will be a base-line to modified, expanded and improved with time. Thismeans that in all likelihood several different programmerswill contribute to the cede. Setting of standmds prevents acode from “becoming programmer dependent and thereforelost when he leaves (or graduates). It should also be rec-ognized that programming to a standad requires consider-ably more effort than programming in one’s own style andthat the initial programming costs are t@refore higher.

As a result, several global aspects of these stan-dards have been defined. The programs will be written inthe FORTRAN language, using the conm-ucts embodiwl inthe 1977 revisions (also known as FORTWN/77). Noneof the many machine dependent additions to this languagewill be included. The use of this “plain vanilla” FOR-TRAN is essential to solving the poncability issue and, forinstance, will allow usage of the codes on both currentIBM PC and Apple Macintosh equipment . This study isorganized into two tasks (Table 3).

TECHNICAL RESULTS TO DATE

The following sections of this paper summarizesome of the impormnt results developed during the first sixmonths of the project.

As an initial step in all of the studies that comprisethis project, a significant number of field trips have beenmade to participate iri unscheduled surveys and repair op-erations, scheduled surveys, and scheduled drydock repairoperations involving VLCCS and ULCCS. Understandingthe realities of what goes on inside the ballast and cargotanks of a VLCC or ULCC has been a sobering experiencefor the project research investigators and research assis-tants alike.

Examples of imporiant observations include:

“ Corrosion can interact with welded stiffeners,coatings,and tank wall flexibility to cause fatiguecracking.

● Drainage holes intended to prevent ballast wateraccumulations in tanks can become clogged withcorrosion products, debris, and sediment.Concentrated areas of corrosion develop in theseareas. Anodes in the bottom of ballast tanksfrequently are covered with sediment, reducingtheir effectiveness. This sediment severelyhampers inspections of the critical sh-ucturalelements in the bottoms of ballast tanks.

● Inspections of cargo and ballast tanks to deter-mine the Imations and causes of fractures and cor-rosion are exmemely challenging, pardculwly whenrafting the cmgo tanks wearing breathing apparatusmust be used. The quality of inspections is verydependent on the experience and diligence levels ofthe surveyor and on the time and facilities devotedto the survey.

●It is impossible to perform inspections that willreliably disclose the presence of all signifscan tcracks. Survey reports are frequently lacking in-formation on cracks in highly corroded areas,which are replaced in a major overhaul. The prob-ability of detection (POD) ctuwes used in presentfatigue analyses [e.g. 13, 34] are not realistic forpresent inspections of critical stmctural details inVLCC and ULCC tarlkS.

● Many severe corrosion and fatigue crackingproblems can be dwectly traced to bad initial de-sign, construction, and maintenance practices. Oursurveys indicate that problems with high strengthsteel elements seem not to be a problem with thematerial, but rather with its proper use (design),construction (welding, fitting, cutting), and main-tenance (corrosion prevention).

● Repairs to critical internal sbwctural details is adifficult and demanding task for ship owners, op-erators, repair yards, surveyors, and inspectorsaJike. There is no reasonable consensus on what,how, and when to repair. The general lack ofreadily retrievable and analyzable information onrepairs and maintenance frustrates repair andmaintenance tracking. Many fracture repairs ap-pear to be ineffectual. Veeing and welding cracksthat have occurred early in the life of the shipseems to be ineffective; they quickly develop again.Attempts to make temporary repairs (e.g. coldpatching) sewe too long can result in costly downtime due to unexpected cargo losses.

Database development has been a key aspect of theinitial phase of three of the studies (Fatigue, Corrosion,Repairs). In this development, a general purpose com-puter program, FoxPro, has been used. The databases de-veloped with this software have been designed to be fullycompatible with the more comprehensive tabulsw andgraphics capabilities of the CATSIR 3.0 (Computer AidedTanker Structure Inspection and Repair) system being de-veloped by several of the participants in this project [40,41].

Studv 1 - Fatirwe Dama~e Evaluation$

~. Thefields included in the fatigue database are summarized in

II-A-5

-$,.

Table 4. “TIWdevelopment of this database format identi-fied sewed important problems and constraints:

● At the present time it is not feasible to use a gemeral ccnmiinates yslem for all the different classesof tankers involved. Since such a coordinate sys-tem would resolve most of the encountered prob-lems especially in combination with the use ofCADICAM systems, this topic will be addressed infuture research.

●. Within the scope of this database, the reoccur-rence of a crack cannot be determined. Ineffectiverepairs cannot be documented. This is a majordrawback for the rep~ database.

● The type of crack tid the location within a detailhave to be described by a set of key words. Thesekey words were developed with the help of A13Sand are also USMIin the conrosion database.

The location, type, and chwacteristics of a crackwe determined as follows:

● Lorigitudinal position: Frame number.

● Vertical position: Definition of three zones(lower third, middle third, upper third) of Sideshell, Longitudinal Bulkhead, and TransverseFraming.

. .

● Horizontal position: Port or” Starboard andzones.

GDetailed position: Key wordsfordescriptionofcracks.

~. Ten ships with a total number of3,629 cracks have been included into the database. Asummary of these ships and the number of cracks found isgiven in Table 5.

For 6 ships, survey reports submitted by the pardc-ipants have been analyzed in order to achieve the necessaryinformation for the database input. These survey reportsvaried greatly in the amount and detail of information in-cluded For some of the database fields the survey reportsdid not contain sufficient information. A database submit-ted by one participant included 4 ships of the same classwith a total number of 1989 cracks.

Due to the relatively small amount of data included in thepresent database, only basic statistical analyses were per-formed to show analysis approaches and data trends. Thisanalysis has been performed first for all ships included inthe database and then for the 4 ships of the same class,which are mentioned above.

ki!!!k. One important result of the databasedevelopment has been to give operators as well as surveycompanies a better understanding as to what should be in-cluded in future sumey reports. In order to find trends andthe most probable lwations of cracks the percentage of oc-currence of cracks in different types of suuctural compo-nents has been evaluated:

● Cracks in the side shell Lon~itud inal~accounted for about 4270 of all cracks.About 10 % of all cracks were found ineach the Bottom

. .onmtudlnal s and in the

Figure 1 shows the number of cracks as a functionof the time until detection (date of swey - date of deliveryof vessel) for all ships. This graph shows that a relativelyhigh number of cracks occurs early in the life of the ves-sels. These cracks can berelated to bad initial design. Laterin the lifetime of the vessels the effects of fatigue begin toshow.

The number of cracks per tank is presented in Fig-ure 2 for the 4 ships of the same class to show the longi-tudinal disrnbution of the cracks. Most cracks occur aftmidships in tank no. 4.

.The distribution of cracks over the height of theships can be seen in Figure 3 for all ships and in Figure 4for the 4 ships of the same” class. Here the number ofcracks is shown for the different zones, which are estab-lished in the database. For side shell cracks most crackswe found in the middle third of the height.

In general the chosen format of the database hasproven to be sufficient to enable detailed analysis of the in-put data.

A review ofthe existing approaches to fatigue and fracture mechanicshas been started. In this review, a primary emphasis hasbeen given. to analyses associated with defective or dam-aged welded details. A summary ofboth the conventionalstress range - numbers of cycles to failure (S-N) curveapproach and the fracture mechanics (F-M) approach hasbeen prepared.

The work has addressed development of a hybridS-N / F-M analysis that would permit practical analyses ofdefective or damaged welded de[ails. For the calculationof the residual life of cracked details a fracture mechanicsapproach will be used to establish a set of S-N curves fordifferent crack lengths. This set of curves will be compat-ible whh the design S-N curves for untracked details. Themain effon of the next months will be to determine the de-tails of this approach. Professor Stig Berge, a visitingprofessor working on this project (Table 2), has providedsignificant guidance IOdevelopment of this approach.

As a part of the fatigue study, the use of predictedfatigue crack growth behavior in the updating of fatiguedesign life has been inve:cigated [42]. Based on experi-ence and experimental faugue crack growth tests, the rela-tionships between developed crack size and remtining fa-tigue life has been characterized. These analyses haveestablished a definitive link between a conventicmal S-Nfatigue analysis model and a fracture mechanics analysism~el. ~ls has pticuIwlyimportantramificationsinde-velopment of acceptability criteria for cracked internalsuucmral details, :avoiding the zero tolerance syndrome.

The analyses have demonstrated the critical impor-tance of defining realistic probability of detection (POD)curves based on practical ship inspection methmk. Thework has been extended to include a cost - benefit model toevaluate alternative strategies for inspections, maintenance,and repair (IMR) [431.

&udv 2 . Cor rosion Damaze F.valuat ions

The data for this study was provided by the projectparticipants in the form of the gauging report po~ion ofsurveys conducted on the vessels during their servicelives. Surveys me typically conducted every three to fiveyears, as dictated by classification smieties, or the opera-torsown internal maintenance philosophy (which ever is

II-A-6

/.

scwner). The reports can range in detail from simple beltgirth gaging, to full surveys of major details in all tanks.The number of gagings might range “froma few hundred toseveral thousand. These are then compiled irt binders,typically ordered by tank or detail type.

The corrosion rate is de.termined by the environment that the element is exposedto. What is important is more than just the relative amountof salt present in the water. The composition of the corro-sive, while certainly important, is not necewrily the mostimportant factor in determining the corrosion’ rate. Forbaltast tanks one might say that over a huge sample of ves-sels in the same trade the composition of the ballast is thesame. Yet we can expect to see vastly different corrosionrates in ships which have heated cargo and those without.There are in fact innumemble differences in the conditionsin which corrosion takes place, some crucial, some lessso.

When confronted with this “problem, the only re-course was to turn to the literature for an outline of whatswas considered important in the determination of corrosionrate. From the literature, and consultation with industry, alist of the inprtaot factors was compiled.

The next step was to consult with industry repre-sentatives to detm-mine, of the factors considered impor-,tant, those cute might-expect to find reasonable amounts ofdata. This was accomplished by means of a questionnairewhich was sent to each of the participants. This question-naire asked only what types of data the participants hadand could release to the project. For example, it is of inter-est to know the humidity of the ballast tanks when thetanks are not in ballast. An estimate can be made, but accu-rate, high quality data is generilly not available.

Table 6 summarizes the list of the important factorsfor which reasonable amounts. of quality data exists.These factors became the basis for development of the cor-rosion data “base.

We next developd a second questionnaire. Weasked the participants to provide the investigators with ex=actly the information they indicated they could provide, foreach of their ships they wished to see in the database.From this, a list of the ships-for which sufficient amountsof data were available was compiled, and the effort madeto obtain the gauging reports. The resulting database in-cludes those ships for which we were able to get these re-ports.

De elor)w.v The amount of cor-rosion data on even a single ship makes the developmentof a data base a large bookkeeping problem the sort ofproblem that is best suited to a ‘database management sys-tem. If the data is organized in a rational fashion, analysiscan be performed by simple search and average routines.Once the relevant data is input then, work can begin on ananalysis. This is where the difficulty in this sort of worklies. It is vital in the beginning the database is constructedin such a way that all the important data is in fact included,and included in such a manner that it lends itself to anaJy-sis.

The corrosion related factors were separated intothree main types: Ship specific data, Tank specific data,and Incident specflc data.

Ship specific data - data which are assumed toapply to all gagings in atl tanks for all surveys of asingle ship. They include” ship size, date of build,

cargo type (crude or prcduct), double side, doublebottom, class wiety, trade route (it is uue that this”may change over the life of the ship), and the unitsthe surveys are taken in.

Tank specific data - including tank type, time inballast (for ballast tanks), time. in cargo (for cargotanks), corrosion protection system, fresh or saltwater ballast, clean or dirty ballast, sulphur, water,and wax content of cargo, presence of beatedcargo, IGS gas quality (Yosulphur), and method oftank washing.

Incident data - an incident of corrosion is de-fined as a location where a gauging was taken.Thus every gauging represents a corrosion inci-dent, and every gauging from the smwey is in-cluded in the”data base. The incident data includes:ship age at sutvey, the type of corrosion, the typeof detail the corrosion is gauged at, and some rela-tive location in the tank of the gauging.

Table 7 lists the types of detailswhich were consider~ in this study. Depending upon thelevel of complexity one wished to consider, the list mightbe ei~er longer or shorter. This pardcular list was decidedupon as it closely matched one which was used by theTSCF [3]. In developing this list, one of the considera-tions was that the list must lx compatible with that used inthe fatigue portion of this project. The list of details thatstudy used was-much more exhaustive than the one usedhere. For example, brackets of any type are not included,nor are some details such as centerline girders. It was feltthat the large increase in the degrees freedom implied bythe larger list of dettils would mean a diminished samplesize for each of the analysis, and so diminished confidencein the results. It is important in this type of study, becauseof the variable nature of corrosion, to obtain the largestpossible sample size, so that any statistics developed accu-rately reflect reality. The TSCF list of basic details waschosen as a basis for one which would satisfy both the re-quirements of brief generality, as well as compatibilitywith the fatigue study.

The location of the gaugingis given simply in the longitudinal, and verdcal frames aseither forward-middle-aft or upper-middle-lower respec-tively , a format which is used in the fatigue study, andwas chosen here for that reason. Any more detailed, orrather specific, manner of identifying l~ation would havemeant the same increase in degrees of freedom as dis-cussed above in reference to the detail types, and so waslimited for this reason. Even still, the list of 22 standarddetails, along with this cmle of 9 locations, implies 198pssible detailhation pairs

There are a large number ofpossible combinations of”all the factors included in thedatabase. The example that will be developed here will bethe combinations of what were indicated by the participantsto be the most important factors: tank type, detail type, andlmation.

These factors were separated into two groups of‘keys’.The ilrst, KEYI was a combination of the tank typeand detail type. The second, KEYII was a combination ofthe tank type and location. KEYI has 88 possible combi-nations (4 tank types x 22 detail types), while KEYII has36 possible combinations (4 tank types x 9 location pairs)KEYI allows one to examine the relative expected corro-sion rates over a set of details throughout the tank, whereKEYII allows one to examine how one might expect those

II-A-7

.,.

corrosion rates to change as lmation in the tank changes.For example the corrosion rate of Longitudinal BulkheadStiffenerWebs (L13LW) in ballast only tanks might be aconcern.

After determining the expected corrosion rate overzLILBLW% in the tank (Figure 5), one would examine theKEYII curves to determine how corrosion changes as onemoves either forward, tit, down, or up in the tank (Figure6).

. The data for the corrosion portionof the study came,. for the most part, from the gaugingportion: of survey reports. These reports are intended toreflect the cu~nt condition of the suucture in the tank.The repmts gre often not intended to allow one to under-stand how the condition of the structure is changing withtime. The owner/operator may not be interested in under-standing how corrosion rate is changing, having more thanenough to worry about in simply maintaining the vessel.Because of this, no consistent, coherent effort has beenmade to insure that the data, the gaging portions of thesumey reports; are collected to further this effort. Oftengagings’ are not taken at the mine location in each survey,giving no rime continuity to the data making it difficult tounderstand then how the corrosion will vary through time.

As well, data for localized corrosion is somewhatmuddled. Different firms, depending upon their mainte-nance philosophies regmding localized corrosion, collectdata on the various forms of corrosion (pitting, grooving)in different manners, whether it is simply counting thenumber of pits in a tank; or identifying one gauging astaken in a pit. No indusuy standiud methml of evaluatingthe corrosion damage by localized is used. This has madetheeffom to analysis Imalized corrosion in the same man-ner as general corrosion difficult, if not impossible. An al-ternate method to deal with localized corrosion mus[ be de-veloped.

Sudv 4Dl.HllS -

Fatigug and Corrosion R~

The general smategy used in repairing a vessel isbased on the following considerations.

The design life of the vessel. Typically fortankers this is approximately 20 years. As the ves-sel ‘approaches the end of “economic life, the opera-tor general “will spend less money for repairs andmaintenance. The emphasis will be on makingminimal repairs. needed to keep the vessel in class.

Second hand values as determined by thesupply and demand for tonnage for a ves-sel .of a particular size. The current and antic-ipated demand for tonnage is dictated by the do-mestic arid-international oil markets. Another ma-jor factor is the cost for new builds which has hadan economic substitutional effect on second handvalues which has recently received a lot of atten-tion. The rise in second hand values encouragesship, owners to invest more in maintaining theircu~ent ships and taking a longer term approachtow@ repairs. The object of this effort is to delaythe purchase of expensive new builds.

Future plans of the company for retentionof the ship. Marketing and refining logisticschange with time, Maintenance expenditures forsteel and coatirtg repairs are reduced when theoperator decides that the vessel may no longer fit

in their Iogistics.strategy. Oil companies with U.S.flag tanker operations are faced with the projecteddecline of the Alaska North Slope crude oil tradedue to decreasing production in that field.Independent tanker operators of U.S. flag vesselsalso face this issue.

Availability of funds for maintaining andrepairing vessels. During the t%st half of the1980’s the tanker owners and operators faced eco-nomic crisis. Huge financial. losses by both oilcompany and independent operators alike reducedthe availability of cash for repairs snd maintainingtheir vessels. Owners were forced to make mini-mum investments for repairs and maintenance.

Environmental issues. Increased internationalconcern over environmental issues particularlytanker oil spills have prompted ship owners to in-crease their efforts in maintaining hull shwcturrd in-tegrity. ...-

k@.& The initial effo!m in this study were focused ondevelopment of a tabulru and graphic computer data basesystem for recording and amalyzing steel and coating re-pairs. Two general observations have developed duringthe efforts to compile repairs databases on six tankers:

1) Data on repairs is extremely difficult toretrieve and evaluate, and

2) There is a wide diversity of practices inaccomplishing repairs.

Development and analysis of the data base canmake problem areas readily apparent by reducing largeamounts of repair data into types of structural members re-paired and by location in the vessel. Figure 7 shows acomparison of two double hull tankers of the same class.The data base indicates the l~ation and types of structuralmember repaired. This gives a graphic summmy to theships operator of the areas that need the most attention infuture inspections.

At the initiation of this study three operators haddeveloped crack / steel repair data bases for tracking frac-tures. The sophistication of these data bases varied from arelatively simple paper file system to a computer data basewith graphical displays containing ships drawings. Onlyone operator was tracking coating repairs data in a database.

Tanker opetators in general are not mting full useof computers as tools in tracking their repair expendituresand maintenance documentation. Genemlly, there is thelack of organization in engineering files for rerneving in-formation quickly on steel and coating repairs. Much in-formation including visual and ultrasonic surveys reportsis missing or extremely difficult to retieve due to poorrecord archiving.

Many ship owners and operators have very infor-mal systems for tracking the details of maintenance of agiven ship. Documentation ranges from a coherent historyof reasonably detailed shipyard repair reports on crack re-pairs, st=l renewals, and coatings and anodes maintenanceto scattered shipyard invoices that define gross tonnagesand areas., The dccumentanon varies widely as a functionof the diligence of the owner and opemtor, and as a func-tion of the ship’s life. Maintenance documentation devel-

..

.......

.-

11-A-8

,..

oped during the first five years of a ship’s operationfrequently cannot lx retrieved by the ftiteenth year.

Dcwumentation of crack repairs frequently cannotbe tracked from one repair to mother repai cycle. Thus, itbecomes impossible to e@.luate the effectiveness of giventypes of repairs. The problem’ of documentatiott of crackrepairs is further complicated by corrosion. In manycases, if corrosion is extensive, cracking will not be notedit will only be noted that the detail or section needs to bereplaced. In several cases, we have found that cracks thatwere to be repaired in a certain manner were not repaired atall or were repaired in a manner different from that speci-fied in the repair repott

Similar problems exist with figard to maintenrutceof coatings and anod~s. Details of locations and the coat-ing break downs and the procedures used to repair thebreak downs are frequently not documented. Coating re-pairs wiIl be noted in terms of total area, the coating usedin”the repair, and the cost per unit area. This does notmake it possible to rrack the effectiveness of coating re-pairs nti the basic durability characteristics of the originalcoatings. Similar statements apply to anodes.

Some operators have begun the development andimplementation of computer based maintenance trackingsystems. This study has made use of components of onesuch system identified as CAIRS (Computer Aided In-spection and Repair System) version 3.0. Given acquisi-tion of the required data by ship owners and operators, andits input to this system, tils system promises to revolution-ize maintenance macking. Perhaps, most importantly, thesystem will permit timely analyses of the effectiveness ofrepairs. Improvements in maintenmtce documentation is.badly needed if ship maintenance is to be improved.

Proced~ The inspection processprior to the vessel entering th~ shipyard varies dependingupon the owner. Several months before the vessel isscheduled for the repair yard, an iititial visurd survey isconducted by the ships staff, the shoreside technical staffand an independent surveyor. A gaging survey may ilsobe conducted to quantify the degree” and extent of steelwastage,

Based on the results of the survey, a repair plan iswritten up and an estimate is made of the cost. The repairplan is then submitted to shipyards for bids. The contractis then awarded to the shipyard which makes the best of-fer. Once the ship enters the shipyard, visual and gagingsurveys are again conducted. These ‘secondary surveysusually reveal additional repair items since all the tanks wetree of cargo and ballast. Repairs are then made on itemslisted in the repair contract as well as any additional itemsdiscovered during the repair operations.

During the repair phase shipymd time and budget-ing have a tremendous influence on the type of repairsmade. ‘If the work falls behind schedule or it budgetedfunds m+ redirected for more critical needs, changes in therepairs approach will be”made from the original repairspecifications drawn at the ot%ce. For example, to re-welda fracture and omit the installation of fabricated reinforce-ment brackets. After repairs are completed finalization ofaccounts usually occurs long after the ship has depruted.

See 1 and Coa tin~ Re r)air (’)bservatio~.This study has provided the opportunity to survey a largenumber of ships that have been repaired or were being re-paired. Repati”o@ervanons are being documented in the

TSCF [3] format. The following summarizes some of ourinitial obsmations concerning steel and coating repairs.

: Not all repairs are sound from a naval architecturalstandpoint. Many operators make repairs using experi-enced based rules of thumb approaches. In many cases,cracks begin to reappear during thenext inspxtion.

Often there are differences in the repairs proposedby the office technicat department and what is actuallydone at the shipyard. This is due to either differences inopinion or budget and time constraints at the shipyard.Many of the repairs resulted in m-cracking.

Not all cracks are or can be repaired when they arefound. Given present day inspection procedures andmethods, it is highly unlikely that all signKlcant cracks canbe discovered. However, significant attention is given tothe side shell and tank top structural elements. Cracks inthe side shell and in the major structural members are re-paired using tempor~ (e.g. end drilling cracks) or per-manent methtis (13gures 8- 10). In many cases, it hasbeen observed that cracking is initiated by corrosion (e.g.grmving corrosion in tank stiffener welds) or exacerbatedby corrosion.

A common cracking problem in tankers is at theintersection of the side shell longitudinal at the webframes. In one class of ships, three ship operators triedthree different approaches in bracket and detail design tosolve such problems. One set of details were repairedthree different times. Cracking started during the first fewyears of operations of these ships. Causes can be traceddirectly to improper design, ignored or unknown loadingsand loading effects, and poor construction.

Corrosion protection philosophies vary greatlybetween tanker operators with regttrd to the use of tankcoatings and anodes. Each operator has different hktoriesof uial and error approaches that has evolved into their cor-rosion protection philosophies. Surface preparation of thecoating areas during the initial coating of the newly builtvessel seems to be the key ingredient in getnng the maxi-mum life for tank coatings. Coverage of anodes in ballasttar@ with sediments accumulate in the tanks seems to bea key problem decreasing the effectiveness of antics.

Repairs of cracks and coatings varies widely. Re-pairs of cracks can range from temporary cold patches tocomplete re-design of the detail and replacement of steel inthe vicinity of the detail. Welding cracks is a popular re-pair that the data developed in this project indicates fre-quently must be repeated within a short perimi of time.

Drilling the ends of the cracks is a frequently usedtem@rary repair measure that is used until the ship cart betaken into the drydwk. Repairs of these cracks can rangefrom simple welding to addition of reinforcing. elements.Again, the data developed during this project indicates thatmany of these repairs must be repeated in subsequent dryddings. In one case, a series of side-shell longitudinalcracks has been repaired four times, and each time a differ-ent repair procedure has ken tried.

Many of tbe repairs identifiedby the TSCF [3] arenot followed. We have seen repairs identified by theTSCF as being unsuccessful being used in current repairs.There is a wide variety of opinions on how repairs shouldbe made, ranging from very high quality to very low qual-ity. Our data indicates that high cost repairs do not neces-sarily tm@ate to high durability repairs.

II-A-9

Also we have found that repatis accepted by oneClassification Swiety surveyor or Coast Guard inspectorfor a given ship at given time and location will not be ac-cepted by another for the same ship at a different time andlccation. We have dso found that repairs specified by theowner: operator maintenance personnel frequently will bemodified in the shipyard due to budget and time limita-tions. In many cases, very little engineering or structuralanalysis goes into the specification of repairs, even in thecase of critical shuctural elements.

tv Gu@lmgs for New -. . .

Two general observations concerning durabilityguidelines for new ships have developed during this pro-j=

1) The primary problems with current shipstructures does not seem to be focused intheir capacity characteristics; rather it ,seems to” be focused in their durabilitycharacteristics. Due to the large degrees.of redundancy, ductility, and capacity, thestructural system generally is very robust,i.e., it is very tolerant of localized damageor defects,

2) The majority of the durability problemsseem to be focused in the need for im-provements in design of critical structuraldetails and in improvements in corrosionprotection and maintenance for these de-tails.

Experience developd during this study indicatesempirically, based, hand-book design, and in some casesanalysis based design of critical structural details in mildand high tensile steel construction is not developing sufi%ciently durable structural systems. Conventional swessrange - numbers of cycles to failure (SN) structural analy-sis procedures have been highly developed in the marineindustry and these should be employed in design of criticalstructural details. Design’of many of these details dms notrecognize the specific, construction procedures that will beused to build the ship, and the problems of inspections andrepairs (maintenance).

Similarly, experience is indicating that well de-signed, applied, and maintained corrosion systems canprovide the protection necessary for critical structural de-tails. Improvements are needed in coatings and cathodicprotection systems, and design of compatible structural -coating systems. The major problems are showing up inimproperly designed, applied, and maintained corrosionsystems, and incompatibilities between suuctural and cor-rosion protection systems (e.g. sediment covered anticsin ballast tanks, flexible bulkheads coated with stiff coat-ings, corrosion cells set up between the parent material andthe weld heat-affected zone).

Experience developed during this project has sug-gested three key teclmicat strategies in design for durabil-ity

1) Damage Tolerant Design . design of aship structure that is forgiving in its abil-ity to be tolerant of defects, flaws, and”damage and is able to maintain the criticalaspects of capacity and redundancy.

2) High Quality Production - design andmanufacturing processes and procedures,

and inspection methods that will assure ahigh quality ship’ structure.

3) High Quality Maintenance - painstakingattention to inspection, maintenance, andrepair/replacement. of critical structuredetails throughout life to maintain the im-portant aspects of capacity and redun-dancy.

Developments in de-sign for durability include explicit requirements snd .prwe-dums for design of critical sh-uctural details and systemsfor: a) Repeated loadings, b) Constructability c)Inspectability, d) Repairability, and e) Corrosionprotection (coatings, cathodic, maintenance),

The primary objective of design for durability is tocreate an efficient ship sh-ucture devoid of unanticipatedcosily maintenance and out of service requi.remenrs Theextent of design for durability represents a trade-off be-tween initial costs and long-term operating costs. me ob-jecuve is to make a sufficient initial investment in durabilityquality to forestall escalation in future maintenance andout-o f+semice costs.

Infotrnation developed in Study 2 indicates that fa-ti~e problems develop most frequently because of ignoredor inaccurately characterized loadings, poorly designedconnections (e.g. inappropriate or no analyses, high stressconcentrations, bad--load Vansfer mechanisms), poorlyconstructed systems, and poorly maintained systems (e.g.corrosion allowed to initiate or exacerbate fatigue).

Connections with low stress concentration factors,accurate determination of sustained and cyclic stratrunghistories, use of ductile and fatigue resistant materials(including weldments), robust (damage tolerant) systemdesigns, construction and maintenance quality assuranceand control, and perceptive design methmis are the key de-fenses against fatigue darnage or low durability structuresystems. Fatigue cracking data developed during thisproject indicates that much more care has to lx taken in thedesign and construction of structural details to minimizeshess concentrations when using high tensile steels.

Design for durability includes not only assessmentof the effects of repeated loadings, but as well the.associ-ated aspects of design for constructability, corrosion pro-tection, inspectability, and repairability. Design for con-structability is intended to help assure hat the ship struc-ture system that is designed can be effectively (high likelj~hmd of teaching quality objectives) and efficiently (lowest, -reasonable cost) consh-ucted. This requires that the designand constnsction procedures and plans be thoroughly andproperly integratd.

Design for inspectabiliry is intended to help assurethat the ship structure system can be adequately inspectedand sumeyed, during the construction phase and duringthe.operations - maintenance phase. The reliability of in-spectability is directly connected with the design for re-peated loa&ngs. Given that the degree of inspectability ofthe structural system is low, either during conirruction oroperations - maintenance, “thenthe requirements for defecttolerance (robustness) in the system are increased

It is here that important questions ,should be raisedconcerning how ship smsctures are presently designed.Designs are focused on creation of minimum weight sys-tems. These emphasize the use of thin plates (to containcargo and ballast, and exclude sea water) reinforced by a

..........

II-A-1O

multitude of frames and stiffeners (to provide stiffness andstrength). Consideration of design for highly automatedfabrication provides important additional consmints on thestructural configurations and assemblages.

Primary attentionneedstobe directed to rczogni-tionof the very limited degrees ‘of inspectability of thestructural system, rather than assuming that inspectionscan or will be done with a high degree of detection and ac-curacy. This would tend to constrain the design of thesystem to ‘use of thicker plates and fewer frames and stiff-eners. Design for inspectability should also address prc+visions to facilitate human access and inspections. Adop-tion of seater spacings for members to facilitate access,avoiding blind spots in the structural arrangements, andproviding access facilities (openings, ladders, walkways,removable staging systems) for emering important parts ofthe structure. Cleaning, degassing, and lighting systemsalso need to be provided. In addition, design for in-spe-ctability should address development of and provisionsfor remotely operat~ inspection systems and insh-umenta-tion systems.

Design for repairability should include explicitconsideration of hbw the system can be repaired whenthere is damage or defects or when the system must bemaintained. Too often, in the relative coinfort of the de-sign office, it is assumed that the critical structural detailscan be easily accessed, damaged or defective elements re-moved, and repairs made. Planning must be done at thedesign stage on how repairs and-maintenance will be done.Again, this requires proper and thorough integration of therepair yard and maintenance objectives and capabilitieswith the other design objectives.

The information developml in Srudy 1 indicates thata key element in design for durability is corrosion protec-tion, particularly for the critical internal soucrural elementsassociated with cargo and ballast tanks of VLCCS andULCCS. Experience indicates that the most severe corro-sion rates can be expected in billast tanks. The corrosioneffects may be the worst when the ballast tanks are emptyor partially full. In this phase, cathodic protection can notprotect the metal not covered by water. Cathodic protec-tion efficiency can ‘be reduced by sediment cover in thebottoms of the tanks. Corrosion can be exacerbated byadjacent heated cargo tanks.

Corrosion is also a problem in the cargo tanks.Generally, these tanks experience more of the pitting ~peof corrosion rather than general wastage. Tank washingand the area under loading line outlets can act to removecoatings and the protection provided by waxy crude car-gos. Breakdown of coatings in the under-deck area ofcargo tanks can be very severe; ‘Coating breakdowns andpartially coated arkai can act to accelerate local corrosion:

Coatings and cathodic protection are practical pro-tective measures. Design that eliminates or minimizestraps for water and sediment, and provides tiour or ero-sion protection must be encouraged. Coatings must beproperly designed to match the projected expected setviceand maintenance, and flexibility of the components to beprotected. They must be properly appIied, cured, andmaintained. Similar statements regard the design, installa-tion, and maintenance of cad-mdicprotection systems.

ACKNOWLEDGEMENTS

The successful initiation of this project and its ini-tial progress hasbeenduetotheeffortsofmany people

andorganizations. In particular, the assistance and leader-ship provided by Bob Ternus, John Balczewski, MarkBuetzow, Rong Wang, and Kirsi Tikka, Chevron Ship-ping; John Cordon Y.K. Chen, American Bureau ofShipping; Anil Thyamballi (formerly ABS, nowIMODCO); John Ferguson, Lloyds Registery of Ship-ping; Dick Whiteside and Dave Wltmer, B.F’.Oil Com-pany; Al Delli Paoli and Stuart Lawrie, Exxon Co. Inter-national; Tom Hagner and Dick Bell, Amoco, Paul Cojeenand Mike Parmelee, U.S. Coast Guard are gratefully ac-knowledged.

REFERENCES

[1] International Ship & Offshore Structures Congress,“Proceedings of the Ninth International Ship SticturesCongress,” Geneva, Italy, Sept. 1985.

[2] International Ship & Offshore Sh-uctures Congress,“FY6ceedings of the Tenth International Ship & OffshoreStructures Congress,” Vol. 1-3, Lyngby, Denmark, Aug1988.

[3] Tanker Structure Co-operative Forum, “GuidanceManual for the Inspection and Condition Assessment ofTanker Structures,” Issued by International Chamber ofShipping Oil Companies International Marine Forum,Witherby & Co. Ltd., London, 1986.

[4] Jordan, C.R., and Knight, L.T., “Further Survey ofIn-Service Performance of Structural Details,” ShipSmtctures Committee, SSC-294, 1980.

[5] Herring, L.C. Jr,, Titcomb, A. N., “Investigation ofInternal Corrosion and Corrosion Conrrol Alternatives inCommercial Tankships,” Ship Structures Committee,SSC-312, 1981.

[6] Munse, W.H., Wilbur, T.W., Tellalian, M.L, Nicoll,K., and Wilson, K., “Fatigue Characterization of Fabri.catedShip Details for Design,” Ship Structures Commit-tee, SSC-318, 1983.

[7] Buckley, W.H., “A Study of Extreme Waves andTheir Effects on Ship Structures,” Ship Structures Com-mittee, SSC-320, 1983.

[8] Kaplan, P., Benatar, M., Bentson, J., and Achtarides,T.A.;’ “Analysis and Assessment of Major UncertaintiesAssociated with Ship Hull Ultimate Failure,” Ship SUUc-tures Committee, SSC-322, 1984,

[9] Bumside, O.H., Hudak, S.J., Oelkers, E., Chan, K.,Dexter, R. J., “Long-Term Corrosion Fatigue of WeldedMarine Steels,” Ship Structures Committee, SSC-326,1984.

[10] Mansour, A.E., “Tutorial Summary on Sh-uctural Re-liability Theory Directed at the Marine Industry,” ShipStructures Committee Contract on Reliability of MarineSrrucmres, Contract No. DTCG23-86-20054, 1989.

[11] Stiansen, S.G., and Chen, I-WI., “Application ofProbabilistic Design Methuls to Wave Load Prediction forShip S~ctures Analysis,” SNAME T&R Bulletin 2-27,Jan. 1982.

[12] Thayamballi, A.K., Chen, Y-K, and Chen, H-H,“Deterministic and Reliability Retrospective StrengthAssessments of Oceangoing Vessels,” SNAME AnnualMeeting, Nov. 1987.

II-A-11

m

[13]Paliou,C.,and Shinozuka,M., “Reliabilityand [29] Mansour, A.E., Yang, J. M., and Ren, D. H.,Durability of Marine Structures,:” Final Technical+Report, “Experimental Investigation of Shi~ Hull Ultimate-ABS Conrract No. CUO0396501, Columbia Umversity;1986.

[14] Chen, “Y.N., and Mavrakis, S.A., “Closed-FormSpecrral Fatigue Analysis for Compliant Offshore Struc-tures,” Journal of Ship Research, Vol. 32, No. 4, Dec.,1988.

[15] Thayamballi, A., Chen Y-K., and Liu, D., ““FractureMechanics Based Assessment of Fatigue Reliability inShip Structures,” SNAME, Ship Structure Symposium’84, Oct. 1984.

[16] Faulkner, D., and Shinozuka, M., (Editors),“Design, Inspection, and Redundancy in Marine Struc-tures,” National Academy of Science Press, 1984.

[17] Sores, Guedes, “Probabilistic Mcdels for Load Ef-fects in Ship Snuctures,” Report UR-84-38, Trondheim,June 1984.

[18] Smith, C.S., “Influence of “Local Compressive Fail-ure on Ultimate Longitudinal Strength af a Ship’s Hull,”Proc. Conf. on Practical Design in Shipbuilding(PRADS), Tokyo, 1977.

[19] Kohler, P.E., and Valsgard, S., “Advanced ShipStructural Design and Maintenance,” Paper Series No.88PO04, Det Norske”Veritas, June 1988.

[20] Skam, K.T., Valsgard, S., Kohler, P.E., and Murer,C.,“How Low Can Steel Weight Go with Safety andEconomy ?,” Paper Series No. 87, Det Norske Veritas,June 1987.

[21] Bea, R.G., and Smith, C. E., “AIM (Assessment,Inspection, Maintenance) and Reliability of Offshore Plat-forms,” Proceedings, Marine Structural Reliability Sym-posium, SNAME, SY-23, Oct. 1987.

[22] White, G.J., and Ayyub, B.M., “Reliability Methodsfor Ship Structures,” Naval Engineers Journal, May 1985.

[23] Moan, T., “Suuctural Risk Assessment for Designand Operation,” Proceedings, Offshore Safety Conference,London, Oct. 1987.

[24] Weber, P.F., “Structural Surveys of Oil Tankers,”Exxon Paper Presented at Joint Meeting Institute of MarineEngineers and Royal Institution of Naval Architects,Transactions,. Institute of Marine Engineers, Vol. 96,1984.

[25] Exxon Corporation Position Paper, “Large Oil TarikerSh-uCturalSurvey Experience,” June 1982.

[26] MOW, T., “Safety of Offshore Suuctures,” Proceed-ings, ,Fourth Int. Conf. on Applications of Statistics andProbability in Soil and, Suuctural Engineering, 1983-

[27] Bee, C, Jorgensen, I., Mathiesen, T. Chr., andRuiren; E.M.Q.,’’Reliability Design Criteria for ShippingIndustry,” Det Norske Veritas, Pub. No. 93, 1976.

[28] Kristoffersen, K., and Brakas, M.J., “Methodologyand Computer Analysis for Optimization of Maintenance,”Paper Presented at The Asian Maintenance ManagementConference, Singapore, April 1986.

Stre;gth Using Large S~ale Maiels,” Report to the Msr-itime Administration, Dept. of Naval Architecture & Off-shore Engineering, Univ. of Ca., Berkeley, Feb. 1987. .-. ...

[30] Mansour, A.E., “Ultimate Strength of a Ship’s HullGirder in Plastic and Buckling Modes,” Ship StructuresCommittee, SSC-299, 1980.,

[31] Akit% Y.,”’’Reliability of Ships in Collapse, Fatigue,and Corrosive Damages,” Proceedings, First Jnt. Sympo-sium, Ship’s Reliability ’85, 1985.

[32] Smith, C.S., and Dow, R.S., “Residual Strength ofDamag~ Steel Ships and Offshore Structures,” Journal ofConstructional Steel Research, Vol. 1, No. 4, Imndon,Sept. 1981.

[33] Adamchak, J-C., “~STk A Program forl%timatingthe Collapse Moment of a Ship’s Hull Under LmtgitudinalBending,” David Taylor Naval Ship Res. and Devel.. Ctr..Rpt. 82/076, Oct. 1982.

[34] Madsen, H. O., Tallin, A. G., and Kirkemo, F.,“Probabilistic Fatigue Crack Growth Analysis of OffshoreStructures, with Reliability Updating Through Inspection,”proceedings, Marine Structural Reliability Symposium,SNAME SY-23, Oct. 19g7.

[35] Wirsching, P.H., and Chen, Y.N., “Considerationsof Probability-Based Fatigue Design for Marine Shmc:tures,” Ibid.

[36] International Ship Smuctures Congress, Report ofCommittee 111.3,“Fabrication ~d Service Factors,” Pro-ceedings, Ninth Int. Ship, Struct. Cong., Geneva, Italy,1985.

[37] Akita, Y., “Lessons Learned from Failure and Dam-age of Ships,” Joint Session 1, Eighth Int., Ship Struct.Cong., Gdansk, 1982.

[38]” Purtell, T.W., Mielke, T. C., and Bn.rsseau, J.P.,“Marine Structural Casualty Study,” Report to the MruineInspection Program Casualty Review Council, U.S. CoastGuard, April 27, 1988.

[39] Allan, R.C.,.Bird, J., and Clarke, J.D., “Use of Ad-hesives in Repair of Cracks $? Ship Structures,” Materials.and Science Technology, Insutute of Metals, Institution ofMechanical Engineers, Vol 4, London, Oct. 1988.

[40] Donnelly, J. W., “CA~SIR Computer Aided TankerSmuctures Inspection andRepair Planning,” Proceedings,Marine Structural Inspection, Maintenance, and Monitor-ing Symposium, March 1991.

[41] Balczewski, J. T.; and Temus, R. A., “ComputerAided Tariker Structure Inspection and Repair (CATSIR),”Proceedings, Mmine Structural Inspection, Maintenance,and Monitoring Symposium, March 1991.

[42] Cramer, ~. H., and Bea, R. G., “Fatigue ReliabilityModel for Inspection, Updating, and Repair of WeldedGeomehies,” Proce.tiings, Marine Suyctmal Inspection,Maintenance, and Monitoring Symposium, March 1991.

[43] Cramer, E. H., Hauge, L. H., “A Maximum UtilityMtiel for Structures Subject to Fatigue Crack Growth,”Proceedings, Marine Structural Inspection, M@tenance,and Monitoring Symposium, March 1991.

II-A-12

Sporisors and Pa;i!!~a~ts in tJCB SMF’

Sector

Government

Classification

Shipyard

Organization

U.S. Coast GuardMilitary S&lift CommandMaritime Administmtion.Naval S~a Systems Command(ship structures commit-tee)

:mlmi;mlw$#u of shipping

Lloyds Registery of ShippingGermanischer Lloyd

LisnaveJurong Shipyard Ltd.Ishikawajima-Harima Heavy Industries Co. Ltd.Mitsubishi Heavy Industries Ltd.Daewoo Shipbuilding& Heavy Machinery Ltd

Ammo Transport Co.Arco Maine Inc.B,P. Oil CompanyExxon Company InternationalChevron Shipping Co.Mobil Shipping and Transpon Co,

Owner/Operator

Table 2 ‘Project Organization

Project Responsibility Name, OrganizationI

~onsulting to All Prof. Alaa Mansour, UCBitudies Y. K. Chen, ABS

Study 1 - Fatigue Prof. Robert Bea, UCBProf. Stig Berge, U. of Trondheim, NorwayRolf Schulte-Strathaus, Research AssistantEspen Cramer, Research Assistant

:tudy 2 - Corrosion Prof.Robert Bea, UCBRob Pollard, Research Assistant

$tudy 3 - Interaction of Prof. Randolph Paulling, UCBDetails with Adjacent Jim $kear, Research AssistantStructure

Prof. Robert Bea, UCBStudy “4 - Repairs Rokt Baker, .Reieirch Assistant

Martin Cepauskas, Research Assistant

study 5 - New Build . Prof. Robert Bea, UCBGuidelines Research Assistant to be Appoin ted

Study 6 . “Software Prof. William Webster, UCBDevelopment Scott Morris, ~O@TI rningAssistant

II-A-13

+-

Table 3Summary of Tasks Comprising SMP Studies

Study Tasks

laskl - Gather,mchive (computer data bases), and assess (trends,statistics) data from pmticipants and from the literature on fatigue andcorrosion damage to ship hull structure elements, with pardcular emphasisgiven to darnage develo~ in primary sh-ucturrdcomponents of largetankers subjected to severe s@ce conditions. Evaluate this data to deter-

1 mine important characteristics of fatigue crack initiationmid propagation.~- Review, critique, and document existing approaches to evalua-

Fatigue tions of fatigue effects on the srength chamctmistics of welded details andhull compcments, with pardculw attention to defdcrive / damagedcomponents and higher stiengih steel~ this will include evaluations of thefatigue damage causes and characteristics .summarizd in the data basestudy;~-Develop a general approachtoevaluatetheeffectsoffatigueonpropagatingbown defects(existingcmcki)incriticalinternalhulldetails,utilizingprobabilisticstress-numberofcycles(S-N) and fracturemechanics approache~_ - Based on a review of ship maintenance records, informationwhich has btin supplied by pardcipants, and the technical literature onhull connection details, establish a fatigue classification system for thecritical internal suuctural details, and establish fatigue strength limitsconsidering both the structural analysis procedures and S-Nclassifications.Task ~ - Characterize the interactions between capacity of single comp~nents (similar to standard S-N curves) and hull service strength to definetheir criticality;H - Develop a PC-based program that will permit the practicalevaluations of fatigue crack propagation, and.effects on componentstrength and leak integrity.

~ - Gather, archive ~cornputer data bases), and assess (trends,statistics) data from parhclpants and from the literature on corfosiondarnage to ship hull structure elements, with psrticuku emphasis given todamage developed in critical internal snuctural components of largetankers subjected to normal service condirnons. Evaluate this set of data todetermine important chmacteristics of corrosion (local and genet@ internal

2 corrosion sates as function of time, tank contents, and ship routes),corrosion conmol systems and their performance, and effects of corrosion

Corrosion on component capacity. Evaluate this data to determine imptantcorrelations of corrosion and fatigue damage, and corrosion and hull-component flexibility. This task will be performed by the same personnelthat perfo~ $e assmiatcd task in the fatigue study.~- Crttlctily examine and report on allowable corrosion limits usedby Classification Societies to assess the rteed for steeI renewals.m - Evaluate the influence of flexure of a component on itscorrosion rate based on analysis of data from in-service vessels;Task 4- Study the feasibility of an experimental program whichexamines the effect of flexure on corrosion rates, including the design ofan experimen~ necess~ equipment, tid budge~~ - Develop guidelines and .mcommended limits on local flexibilityof comlxments for new consmscbonx~ - Develop guidelines for existing vessels to mitigate flexure-related corrosion of an area by increasing its lccal stiffness;Task 7- Develop theoretically and empirically based guidelines for theevaluation of the effects of corrosion on the strength and leak integrity ofcritical internal structural details mid define appropriate wastage allowancehl-nit&= - Develop PC-baswl software to analyze specfic corrosionproblems based on the guidelines developed in Tasks 5-7, showingcorrosionratesversustime,memberflexibilityandIcxationand~ ofprotection.Thesoftwarewl-givetheevaluatorinsights into where,when, and with what rates corrosion darnage can be expcted in the criticalinternal components comprising tanker hulls.lask-!l - Verify the analytical methods and software with sevemlcorrosion/hull flexure case studies.

r

‘>_.

II-A-14

,- -.--

Table 3Summary of Tasks Comprising SMP Studies

,,,.

Studv I Tasks

3

Interactionof Details

withAdjacentStructure

IaSILl-Determine the extentofstructure(rnddesize)necessarytoaccuratelymodelthelcmlelement-globalstructureinteraction.Classificationsocietypracticesusuallyrequirethatamwhde consistingofthreeorfourtankspacesononeside of the center line be modded for theproper three dimensional analysis of transverse suen~h of tankerstructures. Such a lsrge and detailed mcdel is probably unnecessary forthe lmal response analysis of concern in this project-

A major initird part of the work will be concerned with defining themodule of adjacent suucture of miniium extent necessary to accuratelydepict the bmiat-y conditions and loading of the detail underconsideration. While this mdde is ex~cted to be much smaller than thefour tank lengths mentioned above, in the early part of the project, it willbe necessary to carry out one or mot-elarge scale computations embracingthe entire ship, followed by several smaller scale computations focussedon progressive] y smalIer mcdules in order to identify the minimummtiule extent snd to verify and calibrate the conclusions.

The structural components and their response will, in general, be mmlelldby standard elastic finite element methods. The mdule structural mdelwill be designed sround a system of substructures representing the hullcomponents adjacent to the detail. These components will in most cases,consist of panels of plating and associated stiffeners. The sh-uctural detaile.g., a bracket, will then be modelled by a fine finite element mesh inorder to fccus on the lcwal m-uctural response. The load input to theanalysis will be the aforementioned subsh-ucture bdy and boundaryloads,

w -In parallel with the determination of appropriate mcdule size,which affects the bound~ conditions of the suucturai derail underconsideration, it will be necessary to develop’s procedure for generatingappropriate boundary loads on the detail. Such loads will accuratelyreflect the local internal reactions due to the ship’s weight and buoyancydkrnbution and the seaway effects. The module sizing study of Task 1will lx conducted in such a manner that it will lx possible to track thebehavior of boundary loads on the succession of module sizes as we foauin on the local detail.

The loading system will reflect the weight, load and ballast distributions,the disrnbution of buoyancy and seaway effects drawn from the rule loadsspecified by classirlcauon sccieties as well as current ship motions andloads computational pt-ccedures~

/-

II-A-15

Table 3Summary of Tasks Comprising SMP Studies

Study Tasks

w - Gather,archive(~ andmain-framecomputerdatabases),sndassess(trends,statistics)datafromparticipantsandfromtheliteratureoncorrosionandfatiguedamagerepairsto ship hull structure elements;

4 ~ - Review, evaluate, organize, summarize, and document existingapproaches to repair of fatigue damages to critical internal hull structural

Repairs details (mild and HTS steel); and develop critical detail fatigue designguidelines to help rninirnk maintenance cmts associatwl with new builds;~ - Review, evaluate, organize, summarize and document existingapproaches to repair of corrosion damages to critical internal hull structuraldetail% and develop corrosion design and protection guidelines to helpminimize maintenance costs associakid with new build~= - Bas~ on given unit costs (repair, out-of-se~ice), repair timeand tonnage esnmates, and the Kkelihmis of repair effectiveness, developeconomics based prcoxlures for evaluation of alternative repak programs(Figures 2, 3);Wl@ - Develop a pC-basti program that will assist analysesof theeffectiveness of pro~sed fatigue and corrosion damage repairs (theseanatyses will address the local effectiveness of proposed repairs contrastedwith the global effectiveness addressed by the globtd stmctural analyses).

@l@ - Define information that should be provided for durability design5“ mcludmg the hfe-cycle structural integnty plan, design criteria damage

tolerance plan, durability. development approach, materials selection andfabrication, and operations - maintenance plans.

~ew-BuildGuidelines ~- Define improvements in structural analyses me[hods that can .

lead to acceptable durability characteristics of the ship’s critical elementsincluding analyses explicitly addressing darnage toletance and durability.

W - Define requirements for testing of critical sn-uctural componentsto demonstrate adequate capacity, dumbility, and damage tolerance, andin-service monitoring of critical elements in VL.CCSand ULCCS aimed atimproving understandings of loadings and oWrations procedures effectson loadings.

w - Define the characteristics of an industry-wide computer database system for archiving design and construction information, operationssuwctural tracking and maintenance tracking of the paformancecharacteristics of the critical internal structud elements. .

~ - development of standards for the coding of the programs, the6 writing of documentation, and the selection of appropriate database

softwareSoftware

M - development of an appropriate user interfaces for the cmlesdevelopwl m the various studies.

‘.,..._ .’

.-......-

.

II-A-16

----,/

Table 4Descriptive Fields for Fatigue Data Base

Field Name Input

Vessel lD NumberlTank ID Number and Lmabon (S,CY)Memkr cracked Char-+”- ~-’~ ISwev m ngtP

=avE=

m.

+-%%

1 aL LG1 J AGUI

. ..J— , _..: of surveyving no. I NUIkr:s %“,an Height

,U lank

I

IFrame no. NumLlDistance from Frame “DecimalNumberCrack Tym. Set of Keywords I

I. .

I (Ot3tionto-editormodifynew I] ke~words when needed-

Cwk 1.enrrth [ I e.nnn units I

Crack Llas mbeDate of Su

—

lrv

.. ... 1 -. . . . . . .

Ss I Kur :r, Yearnwey I Month, Year

‘.ionth, YearWracter Fieldlet of Keywords

, :haracter FieldICausefor Damam 1:

Table 5Summary of Tankers Included in the Data Base

Hull Tvne I.-

1DWT I Year Built Number of

Cracks,

DoubleHull 39,000 1977 168Double HuH 39,000 -1975 24Double Bottom 188,500 1979 327Double Bottom 188,500 1980 177Single Hull 70,200 1972 639Single Hull 35,700 1973 321Single Hull 153,200 1977 651Single Huh “153,200 ~ 457Single Hull 153,200 1977 413Sirwle Hull 153.200 1976 467

II-A-17

.,!FF

Table 6Corosion Factors Included in Corrosion Data Base

Ship SizeDelivery DateCargo T~Double J30ttom(Y~)Double Side (Y/N)Class SocietyTrade RoutelTank Location

Tsnk TypeTne in Cargo “”Time in Ballast ~~Corrosion ProtectionSystemBallast Type

Tank Tcmperature2”Tank Humidity2

Cargo Sulphur (%)Ctigo Water (%)Wsx in cargo (Y/N)Heated Cargos (%)Tank WashingIGS (Y/N)Corrosion T~Cornxled DetailDetal kation

1-Seldom well defined2- For Ballast tanks, often an estimate

Table 7List of Critical” Internal Structural Details Included in Corrosion Data Base

I DETAIL I CODE

DeckPlatingDeck Long.Web :[WDeck Long. Flange DLFSide PlatingSide lan~. Web SLWSideLon~. Flange SLFBottom Plating BPBottom l-m-m.Web BLWBottom Lon~. Flange BLFLong.BHD Plating LBPLong. BHD Long. WEBLBLWLong. BHD Long. Flange LBLFT-BHD Plating T.BPT-BHD Stiffe~er Web TBSWT-BHD Stiffener Flange TBSF.Horizontal Girder WebHorizontal Girder Flange

I Vettical Girder Web I VGWVerncal Girder Flange VGFT-Web PlatingT-Web Flange TWFSwash BHD Plating SBP

,------

-...,.

II-A-18

boor-””’”””’ J

600

500

400

300

200

100

01 2 345 6789101112131415

Ship Age Fears]

Figure 1 Number of Cracks vs. Time until Detection

700F

,-.,

,,

Figure 2 No of Cracks per Tank (4 Ships same Class)

II-A-19

---

----1

Side Shell L-BHD Transverse

Zone8

Figure 3 No. of Cracks per Zone (All Ships)

---

I5oot

1

Side Shell L.BHD Transverse

Zones

Figure4 No. of Cracks per Zone

(4 Ships of the same Class)

II-A-29

1

O.B

0.6

Relative Fre~ency

0.4

0.2

0

0.08

0.06

Average WastageRate (m/w)

0,04

0.02

0

0

o -g -e-h-a..II A

i

F==l– Approx. Dist.

o 0.04 0.08 0.12 0,16 0.2Wastage Rate (snn. yr)

Fig. 5 “PDF for KEYI: Ballagt only . LBLW

Aft Iiiddle ForwardLongitudinalLocation

Fig. 6 Wastage Rate for KEYII: Ballast Only . Location Paira

II-A-21

ITPE 1

TYPE 1

1I I

w

IL!3!L II

-H%I!-.S-J

‘w

,. . .....

Figure 7 Comparison of Crack Locationsof two Class Vessels

‘,

II-A-22

.OCATION: Tratwersebulkheadvertical sttlfenerlntetilan at tank rop or mu~e

Mall

SXAMPLENo. 1: Cracks at Vefica[ stiffenerweld and tank top @ute

TYPICAL DAMAGE

+

,4TlERTICAL STIFFENER

I \ TANK TOP

IFRACTURE

v

8

A

h 3

I \TRANSVERSE

RAT HOLE1,

BULK{EAD

REPAIR

T/,.,,,~

//?,”,,

l\PLATE’ INSERT

FACTORS CONTRIBUTING TO DAMAGE

1.

2.

3.

4.

Poor detail design due to-lack of tripping brackets.

Weld undercuts and excessive root openings.

Rat hole under tank top is too large creating stress area.

Mis-alignment of veflical bulkhead stiffeners’ and Iongitudinals underthe tank top,

Figure 8 Repair Case Study 3

..

EXAMPLE No. 3: Cracks and wcstoge

TYPICAL DAMAGE

LONGITUDINALBULKHEAD KNUCKLE LINE

CRACKS

\

). 1

1

v/

LONGITUDINAL

It longitudinal stflener

REPAIR

p~TE INSERT

\

/

I

FACTORS CONTRIBUTING TO DAMAGE

1. Grooving corrosion wastage and fatigue,

2. Dynamic seaway loads i ship motion of forward end of ship.

3. High stress area at intersection of knuckle line caused accelerated coatingbreakdown and corromon along with fatigue.

Figure 9 Repair Case Study 4

-.\

)CATION: ~ /&w#-d ofIm@udlmli bwM30d SepuofhgCorgom

(AMPLE No. 4: CrocksInlongitudinalbulkhead akmg tomda Of~itwlnals

TYPICAL DAMAGE

WEB FRAME

/ LONGITLIDINALS

3AL~sT TANK LOkGITUDINALBULKHEAD

REPAIR

WEB FRAME

(“PIATEINSERT

--&

—i[~l—— —__ ___ --——

. —

LONGITUDINAL

- -&q --

1

I

CARGO TANK

‘ACTORS CONTRIBUTING TO DAMAGE

Grooving corrosionand fatigue.

!. Deflection of longitudinalbulkhead underload accelerating coating breakdown and fatigue.

Figure 10 Repair Case Study 6

II-A-25

DISCUSSION

Grea White

First, I’d like to say that I was impressed with thethoroughness with which you’ve attacked this immenseproblem. It is going to b extremely helpful in the longterm. I’m interested in your discussion on corrosion. -Approximately ten years ago, when I was working forExxon, we were doing a similar internalsurvey of ships.What we found was thatthe surveyors would go in thereand because the structureis so large, we were gettingcorrosion measummentaout which were almost useless.There wasnoreal way of knowing, with thefew corrosionmeasurementsthey had takenon thishuge am, what thereal values were. What I want to ask is how are youaddressingthatproblem? Are you preplanningan areatocover and a number of sites, or Iaving it up to theinqxztors to look for the worst corrosion spots? Is themsome means to take the information you’re getting andcame up with some soti of contldence intend on thewastage experienced and expected?

Robert Bea

First, on the &ta for these 1970 to early 1980s ships wewere, of course, having to work with the &ta the opem-tors gave us so there’s not much of a chanm to, I’ll call i~

rwtructurehistory. One of theprincipal thingstheprojecthad to go through initially was looking at repetitive sur-veys within the ship thatwere of high enough quality to ‘-enter into the database, We encountered the very sameproblem you addmsed, much of the dataquality was notsufficient to resolve the questions thatwe were trying toanswer. So thatdatawas called, “ti you, but no thankyou.” The datathatwas selecti hasbeen repted, it ishigh quality, it’s been done in a number of i&nticallocathns acres a number of surveys. So we have someinfonnadon to help us with data con6&nce. I think thatwe have noted in that&velopment thatarea bit, I’ll callit disarming,is thatwhenever an areais heavily ccmoddiwhetherurnot it’s fracturingvery seldom shows up in theSurveyorrqmts. That steel is identifmdfor renewal. Thecracking information is very generrdlydefwient bewseof that.

On the other point thatyou addressed,as we go forwardin theinsqxztionprmes improving when, where andhowwe inspecg I thinkit is importantto &fine how we archivethatmassive data. Rob Pollard hastpnt many Saturdaysand Sun&ys inputting almost a million data points intothe database. There’s got to be an easier way. I thinkit’stime for computers,perhapsdigitalvoice systems,they’rebeing usedinmany otherindustrk We’relooking atthatwithin a component of thisprojec~

, .-

-’..

II-A-26