CRITICAL EVALUATION of Low Energy Ship Collision Damage- Theory & Design Methodologies

7/30/2019 CRITICAL EVALUATION of Low Energy Ship Collision Damage- Theory & Design Methodologies

1/45

SSC- 285CRITICAL EVALUATIONOF LOW-ENERGY SHIPCOLLISION ~DAMAGETHEORIES AND DESIGNMETHODOLOGIESVOLUME 11: LITERATURESEARCH AND REVIEW IS

Th is document has been app rovedfo r pub l ic release and sale ; i t s

d ist r ibut ion is unl imi ted.

SHIP STRUCTURE COMMllTEE1979
http://toc-cd-2.pdf/


2/45

MemberAgencks:lhitdSMESCcmtGumdNavalS-soystemsommand~ti W commandMmtimeAhinistmtion

UnitdstobGdWml survey/&qwmBUIWUof%@in9 *

Adti Gmeqwnden= to:%Xetury,hip shlcture ComnlitkUS.Ccmt,%OIU Hmdqudem,(G-f@4W&@n, D.C.20590

StiuctureCommitteeAn InteragencyAdvisoryCommittee

DedicatedtoIrnprovhgtheStictureofShipsSR-1237

APRIL 1979

Interest in structural protection from collision, grounding orstranding ranges from nuclear-powered vessel design to minordamage resulting in pollution. The interest involves economics,safety of life and property, and conservation of the environment.In view of the existence of a body of prior research, the ShipStructure committee has conducted a project to critically eval-uate this prior effort and determine whether at least one of the ,available design methods or a combination of methods can be usedwith confidence for minimization of collision damage and protec-tion of the vessel.,This is Volume 11 of the final report of the project and containsthe annotated bibliography. Volume I contains the analysis, results,and recommendations. J@@k@

Rear Admiral, U.S. Coast GuardChairman, Ship Structure Committee


3/45

SSC-285FINAL REPORT

onProj ect SR-1237

Col l i si on Damage and Dtrandi ng

CRI TICAL EVALUATION OF LOWENERGYSHI P COLLI SI ON-DAMAGE THEORI ES

AND DESIGN METHODOLOGI ES

VOLUME I I : LI TERATURE SEARCH AND REVI EWby

P. R. Van Mat er, J r .J . G. Gi annott i

GI ANNOTTI & BUCK ASSOCI ATES, I NC.Wth contri buti ons byN. J ones andP. Genal i s

underDepart ment of Transport ati onUni ted States Coast GuardContract No. DOT-CG-63738-A

Tkis &cument has been app~ovad fop p-&Zie ~elease andsaLej- its distribution is unlimited.

U. S. Coast Guard HeadquartersWashi ngton, D. CJ1979


4/45

TochniFolopwthcumontgtionago1.R.mrt He. 2. Gov-rnmmnt Accos*ion No. 3. I?ocipimnts Catalog No..SSC-285 ; . ,>

.4. Ti*l*~ritica?d~%!!~ationf Low-Energy CollisionDamage S.Ropwt Dmtmheoriesand Design Methodologies. Volume 11: Litera- July 19784.Pe;formiow Orgonizet, eri Cod.ure Search and Review7. Adig,l, )

a. Pwfwtning Organization Repor*Ne.

.R. Van Mater, Jr., J. Giannotti,N. Jones, P. Genalis9. P.rlermJn~ Orgrnizotiem Nmo and Addrog! 10. Work Unit No. (TRAIS)~iannotti& Buck Associates, Inc. ,,~711Sarvis Avenue, Suite 402 11.COWact ~, G,otNe.:iverdale,Maryland 20840 DOT-CG-63738-A

13. Typo of Roportond P-ried Cow-rod12, spensmrin~A qncYNomo end Addr*l~:ommandanttG-FCP-2/71) TechnicalReport[.S. Coast Guard 28/2/77 - 28/7/7800 Seventh Street, S. W.Washington,D. C. 11S. Su pp lo mo tne ry No !o s;ontractmonitored by: Ship Research CommitteeNational Academy of Sciences

Washington,D. C.16 . AbstractThis report is Volume II of a two-volumereport prepared under Ship StructureCommitteeProject SR-237, CriticalEvaluationof Low Energy CollisionDamageTheories and Design Methodologies.The material containedherein is the result of one of the tasks of the project whichcalled for conductinga literaturesearch and review of documents relevant to low-energy colllsiondamage. The various data resources used were identified;the stateof the art in ship collisionresearch was summarized;an annotatedbibliographyofthe key documentswas preparedand a list of referenceswhich are considered to berelevant to the problem was developed.Volume I contains the actual assessmenttheories and design methodologiesalongresearch.

of the various low-energycollision damagewith recommendationsfor their use and future

17. Iioywwh - lE. Difitrikuti.n Ststomomt;hipCollisions This documenthas been approved for;tructuralAnalysis public release and sale; its distribution;tructuralDamage is unlimited.;hipStructures;tructuralFailures19. S* wr i t ~Cl rOsi f . (* f Miar* ~t ) 20. t iw ri t y Cl m. si l. (of thi spo~) 21.o. of P* , * * 22. PricmUnclassified Unclassified 38

hIh00TF17~.7 (B-72) ii


5/45

The SHIPprogram to improveby an qxtension ofconstruction.~M H. H. Bell

SHIP STRUCTURECOM241~STRUCTUREC~I~is coast~tutedto-prosecutea rssearcbthehull structuresof shipsqnd othermarinestructuresknowledgepertaining to design,materialsand methodsof

(Chaimun) Hr. M. Pi tk inChi ef, O ff i ce of MerchantMarineSafetyU. S. Coast GuardHeadquartersMr. P. M. Pale-oAssistantfor StructuresNaval ShipEngineeringCenterNavalSea SystemsCormandMr. W. N. HannanVice Presidentberican Bureauof Shipping

AssistantAdministrator forCmmncrcialDevelopmentUaritime AdministrationMr. R. B. KrahlChief,Branchof MarineOil qndGas OperationsU. S. GeologicalSutieyHr. C. J. WhitestoncChiefEngineerMilitary Sealift Comand

Cotmnittee

LCDRT. H. Robinson,U. S, coast Guard(Secretary)SHIP STRUCTURXSUBCOMMITTEE

The SHIP STRUCTURESUBCOMMITTEEacts for the Ship Stkuctureon technicalmattersby providingtechnicalcoordinationfor thedeterminationof goalsand objectivesof the program,and by qvaluatingandinterpretingthe resultsin termsof structuraldesign,construction andoperation.

U. S. CWST GUARDCdr.J. C. CardLcdr S. H. DavisCapt C. B. GlassDr. W. C. DietztivAL SW SYSTEMSMr. R. ChiuMr. R. JohnsonMr. G. SorkinMr. J. B. OBrien

COMMAhT

(ContractsAdmin.)MARITW ADMINISTRATIONHr. F. J. DashmawPk. N. O. HamerMr. F. SeiboldMr. x4.ToumaNATIOK4LAMDEMYSHIPRESEARCHMr. O. H. Oakley

OF SCIENCESCONMITTEE- LiaisonMr. R. W. Rumke - Liaison

SOCIE~ OF lthVALARCHITECTS&MARINEENGINEEEc3Mr. A..B. Stavovy.LiaisonWELDINGREStiCH COUNCILMr. K. H. Koopman - Liaison

MILITARYSEALI~ COwNDMr. T. W. ChapmanMr. A. B. Stavovykir, D. SteinMr. J. TorresenAMSRICANBUREAUOFDr. H. Y. JanMr. D. LiuMr. 1. L. StemMr. S. G. Stiansen

SHIPPING

(Chaiman)U. S. GEOLOGI~ SURVEYHr. R. GiangerelliMr. J. GregoryINTZRNLTIOHALSHIP STRUCWS CONGRESSProf. J. H. Evans - LiaisonAMERICANIRON& STEEL INSTITUTEHr. R. H. Stc?me - LLaisonSIXTSUNIV.OF NEW YORKFLARITIYEOLLEGEDr. W. R. Porter- LiaisonU. S. COASTCU~ AL4DEMYCaptW. C, Nolan- LiaisonU. S. l&iVALCADEMYDr. R. Battacharyya- Liaison

U. S. MERCHAhT MARIh%ACADEKYDr. Chin-Bea Kim - Liaison

iii


6/45

METRIC CONVERSION

Convmims 10 Metric Measms Ap pr oxi mat e Co nve rsi ons f ro m M et Ii c M~asu/es

-.. .-. ..

When Vau Know Muti[plv b! To Fb nd10 f ind SvmbolLENGTH

LENGTH mill mewrs 0.04centimeters 0.4.lmlers 3.3nW*ls 1.1kilometers 0.6

. x !!~. .-- z.- 5 . ... . . . . . .. . :

i nc h . s .2.s Cm,lmwwsCeul,nzu,,meter hklwnem,

.,cmmh,,,

m?n?n?h.,>ha

9k!,t

n I.,1mlIII;3l,?

c

deM 30ads 0.9,n,ltis 1.6

AREAAIIEA -. n- s rj .a rc cm lb nmcm 0,16square nwl. rs 1.?squ rm k il aa h tw s 0.4hectwes il0,00U n? I 2,5

, ,1 !, ., ,, . U ,L IW >.,,4,.71,. ,,,,1.s ,, !, ., , . ,,!,4,.. ,lcrrs

$Ilmvu mchils 6.5Se,,,.mti tact 0.02W ar. v a,d s 0.0square nu~es 2.6, It ,l . s 0.4

5UA,0 C,ml,mm,,, Ip l. ,,u m!x lx swL~. .ClcrsSi+,lwl! k(lmw!tl!,sI,,! .1,,,,+

.-.- s-P< MASS ~weight}MASS {weight)

ohm.. 2ap ln, ,t i s 0.45Awl la,, 0.9

[ 2W0 m l) VOLUME

-.-. 5. .-.. . z.Clrml. 0.035knlwlrams 2.2t omhc , I 1OOOkg ) t.1

Cwr,c, .k,hm,,,,ms1!,, ! ,, . . . ..- rn VOLUME

1*P I#!,l.,M,,m& 5Ibsp %.,tll. spmws 15!1 Ot 11,4 I Amms 30c cups 0.24P pmls 0.47Ill qua ,1s 0.95Qat gallms 3.s,,3 C ll bi c ( em 0.03yd3 c!,b!c ydrds 0.76

TEMPERATI JRE [ ex sc t)

mi 11,I,lws UWIlle,s 2.1Iiwfs 1.06J,Ie , s 0.2Gcub , c nw te rs 35cu b ic nm le rs 1.3

..

TEMPERATURE (e ~a cl )Celsius 9, 5 [ Ih e=- Iempemiure add 32)f FahrenlwI 5/9 IJIIU

tcnbpwatur* sublmcli~32)

CalsusWw+mra lore

01

F 32 986-40 0.40 00 l,, ,,, ,,; J, II,120

160 200t; 1 1 I 1 t 1

-g: -20 0 20 40 60 00 h37 D


7/45

TABLE Ol?CONTENTSVOLUME 11

PAGE1.O INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . ..12.0 LITERATLJRE SEARCH AND REVIEW. . . . . . . . . . . . . . . . . 1

2.1 SUMMARYOF PAST AND CURRENT WORK IN SHIP . . . . . . . . 2COLLISION RESEARCH

2.2 U. S. AND FOREIGN DATA SOURCES . . . . . . . . . . . . . 72.2.1 United States. . . . . . . . . . . . . . . . . . 72.2.2 Germany. . . . . . . . . . . . . . . . . . . . . 82.2.3 Japan. . . . . . . . . . . . . . . . . . . . . . 92.2.4 Italy. . . . . . . . . . . . . . . . . . . . . . 9

2.3 ANNOTATED BIBLIOGRAPHY . . . . . . . . . . . . . . . . . 93.0 BIBLIOGRAPHY OF DOCUMENTS RELATED TO SHIP COLLISION DAMAGE. . 29

v


8/45


9/45

INTRODUCTIONThis report is Volume 11 of a two-volume report prepared under ShipStructure Committee Project SR-237, Critical Evaluation of Low-Energy Collision-Damage Theories and Design Methodologies.The material contained herein is the result of one of the tasks ofthe project which called for conducting a literature search and reviewof documents relevant to low+nergy collision damage. The variousdata resources used were identified; the state-of-the-art in shipcollision research was summarized; an annotated bibliography of thekey documents was prepared and a list of references which are con-sidered to be relevant to the problem was developed.Volume I contains the actual assessment of the various low-energycollision-damage theories and design methodologies along with recommendations for their use and future research.

O LITERATURE SEARCH AND REVIEWA search of the pertinent U. S. and foreign literature resulted inthe development of a bibliography of 428 citations. Copies of 192documents were obtained for review and possible use as data sourcesfor the study in question. The methods used in the search haveincluded:

a Reference citations in technical papers and reportsq Manual search of indexes including

Government R&D Report Index- MRIS Index- Engineering Index Oceanic Index

a Computerized search using NTIS, Lockheed DIALOG and ASDMS(Advanced Ship Data Management) Systems

her half the citations represent various papers on structural analy-sis, plastic analysis, limit analysis, etc. which have been referredto in the various literature on ship structural problems. It becamequite evident from the collected references that there is an abundanceof literature on analytical methods for treating structural componentsin the plastic range, however, the literature on the synthesis of thesemethods to analyze the overall ship structure is much more limited.Only a few documents deal with the problem of low-energy collisions.TO suggest the availability of information, a breakdohn on the documentscollected is given below:

1


10/45

1, NUMBER OF DOCUMENTSDOCUMENT TQEIC. ::. CITATIONS HELD

Structural analysis; theory and experiment 220Collisions, research and design 73Collisions, lowenergy analysis 6Collisions, statistics 26Collisions, model tests 18Collisions, vehicles 10Ship Structures, general 43Ship Structures, ice strengthening, ice breakers 13Ship Structures, slamming loads 15Miscellaneous 4

Total ZZ

933162350138112

1922.1 SUMMARY Ol?PAST AND CURRENT WORK IN SHIP COLLISION RESEARCH

Recently Professor Jones (34) wrote a survey paper on the collision protection of ships. Excerptsof this paper are included aspart of this summary followed by an update of such a survey basedon recent and ongoing work being conducted by the Maritime Administration in cooperation with the German Government.

Many articles have been published over the past decade on variousaspects of automobile, train and bus collisions, and some of these(e.g. (55)) have demonstrated clearly the important contributionswhich structural plasticity can make to developments in this area.However, the complexity of an automobile or train collision, whichinvolves many nonlinear effects (e.g. large strains, strainratesensitive material behavior, etc.), is a serious obstacle to thefurther development of both theoretical and numerical procedures(58). Moreover, the response of a structure which is subjected todynamic loads can be quite different to the behavior when theloads are applied statically, as observed in experimental investigationswhich have been reported in References (33) and (42) onthe longitudinal impact of motor coaches. Thus , the full-scaletesting of automobiles and even aircraft (e.g. (60)) has played animportant role in,the prediction and improvement of crashworthinesscharacteristics [47, etc). This is sometimes a very expensive andtimeconsuming procedure, a situation that has resulted in recentstudies on the feasibility of scalemodel experiments (31, 32) whichare attractive because a systematic variation of the relevant param-eters can be undertaken. However, caution mustbe exercised whenscaling the experimental results up to full size, although Holmesand Sliter (32) did obtain encouraging agreement between the experimental behavior of scaled models and fullsized vehicles. Duffey(18) has shown that the influence of material strain-rate sensi-tivity cannot be properly scaled when a model and a prototype areconstructed of identical materials. Further theoretical objectionsmay well be encountered when attempting to correlate the responseof models and prototypes which involve fracture.


11/45

Lee and Wierzbicki (41) are currently utilizing Martins theoremsin dynamic plasticity in order to obtain bounds on the responseof certain components in automobiles. These simple methods appearpromising for the preliminary design of bumpers, doors and internalenergy absorbing devices, etc.

The foregoing comments are intended to provide a very brief over-view of the many experimental and theoretical investigations whichhave been conducted into the crashworthiness of vartous land-basedvehicles.The ship collision problem is more complicated than the automobilecollision problem for a number of reasons, not least of which isthe enormous amount of kinetic energy possessed by some ships onthe high seas. Thus , the majority of investigations which have beenpublished in this area were conducted on experimental models. Thelimitations on scaling referred to in a previous section acutelyapply when scaling up the experimental results on models in orderto predict the behavior of fullsized ships due in part to hydrodynamic interactions which strongly affect the end result.Akita and Kitamura (3) observed that the bow structure of a strik-ing ship plays a very important role during a collision between twoshtps. The included stem angle, rake and framing of a bow clearlyare important, but the ratio between the strength of the bow of astriking ship and ~he strength of the side of a struck ship has amajor influence on the partition of energy exchange between the twoships. Generally speaking, a stiff bow (e.g. icebreaker) wouldabsorb very little energy so that most of the kinetic energy lostduring impact must be absorbed by the side of the struck ship. Onthe other hand, a weak bow may absorb most of the kinetic energylost during a collision leaving the side of the struck ship essen-tially undamaged. Incidentally, Cheung has suggested a design fora soft bow in Reference (14).~Despite the great deal of care and attention that authors havedevoted to the experimental work performed in Italy (10,63), WesternGermany (67 to 85 ), United Kingdom (96), and Japan (1-3,5-7 ), itwas nevertheless necessary to make compromises. Some experimentshave been conducted statically, or with rigid bows, while othershave utilized a primitive structure for the side of a struck ship.As far as we know, all experimental investigations have examinedthe symmetrical case in which the striking ship impacts at rightangles in the central region between two adjacent transverse bulkheads in the midship section. Incidentally, Akita, et al ( 8) usetheoretical arguments to show that this case is more severe thaneither eccentric or oblique collisions. However, this is not alwaystrue because McDermott, et al (46) have shown that less energy isabsorbed before hull rupture when a vessel strikes near the transverse bulkhead of an oil tanker. Another aspect which has notreceived too much attention is the influence of added mass. Minorsky(48) assumed that the virtual increase in mass of the struck ship

3


12/45

due to entrained water was 0.4 times the mass of the struck ship,since.previous studies on the transverse vibrations of hulls indeep water indicated that the liquid added mass was approximatelythis value. In fact, the theoretical results which are presentedin Figure 1 of Reference (48) show that the kinetic energy lostduring a collision is relatively insensitive to the actual valueof the virtual mass. The largest discrepancy according to Minorskystheory accurs when the mass of the striking ship is larger thanthe mass of the struck ship. For example, when the mass of thestriking ship is double the mass of the struck ship, then the kin-etic energy lost during a collision when the added mass is neglectedis two-thirds of the value when the added mass equals the mass ofthe struck ship. The theoretical results for the amount of addedmass recommended by Minorsky lie about mid-way between these twocalculations. More recently, Akita, et al ( 2 ) conducted someexperimental tests on a ship model and obtained good agreementwith a simple theoretical approach which predicts the added massduring a right-angled collision. It turns out that the added massof the model is about 40 percent of its mass when the duration ofimpact is short. However, the actual value of added mass dependson the duration of impact and on the functional form of the externalforce. In order to provide some guidelines on how short the duration of a typical collision must be in order that an added mass of40 percent of the mass of a struck ship is appropriate, it wouldbe worthwhile to conduct some additional tests and to properly scalethem up to typical full-sized ships.

lT.heprotective structural arrangements which have been examined inall the studies on nuclear powered ships, oil tankers and the singlestudy on a L.N.G. (liquified natural gas) carrier (12) are similarand utilize either the normal structural designs for these vesselsor a slight,modification which includes additional decks specifically designed to absorb the kinetic energy lost during a collision.However, it is clear that tha design requirements for these variousships are different. Clearly, the bow of a striking ship must notbe allowed to penetrate the containment vessel of a nuclear-poweredship. Presumably a similar design requirement would be used for aLNG carrier, except that a number of cargo tanks would requireprotection. The entire length of an oil tanker requires protectionso that it is only feasible in this case to provide protection forminor collisions.The simple semianalytical formula of Minorsky (48) was developedby neglecting the influence of those members which have little depthin the direction of penetration. For example, Minorsky retained theinfluence of decks and transverse bulkheads in the struck vessel, anddecks, longitudinal bulkheads and a portion of the shell plating inthe striking vessel. However, the actual energy absorbed by the struckand the striking ships was not calculated, presumably because of thedifficulty in estimating the failure loads of the various structuralmembers involved in a collision. In order to circumvent this diffi-culty, Minorsky introduced a resistance factor which is related to


13/45

the volume of material located in the damaged portions of thestriking and struck ships. Minorsky plotted this resistance factoragainst the kinetic energy lostduring a collision and observed thatthe data from a number of actual ship collisions essentially col-lapsed onto a straight line. The design formula developed byMinorsky (48) is in fact the equation of this straight line.The simple theoretical method of the Naval Construction ResearchEstablishment (N.C.R.E.) (94) was developed for a rigid bow andonly considered the influence of the deck plates and bottom of astruck ship, which were assumed to have crippling stresses whichwere 90 percent of the corresponding 0.3 percent proof stress.Nevertheless, Belli (10) has recently summarized the experimentalwork which has been conducted in Naples since 1961 and found thatthe N.C.R.E. method gave good predictions provided appropriateallowance was made for the rigid bow approximation.

The design procedures due ~o Minorsky (48) and N.C.R.E. (96)neglect the influence of the shell plating in the struck shipand are therefore expected to be more appropriate for major collisions. In this connection it should be remarked that essentiallyonly bending energy would be absorbed by a flat plate when it isperforated by a rigid wedge which has the same assumptions andthe simple theoretical procedure given in Reference (104 for athin plate perforated by a circular drift. However, it is quiteclear that the behavior of the shell plating of oil tankers assumesvital importance during collisions if the cargo is to be contained(i.e., perforation prevented). McDermott, et al (46) have developeda structural analysis for minor tanker collisions which focuses onthe behavior of the shell plating in the struck ship. It turns outthat typically between 2/3 and 9/10 of the total energy absorbedduring a minor collision is absorbed as membrane tension in thestiffened hull plating. However, it is remarked in Reference (105)that the strength of beams and rectangular plates are very sensi-tive to the magnitude of the inplane displacements at the supportsand some specific expressions are derived in Reference (35) whichcould be developed further to assess the importance of inplanedisplacement in tanker collisions.

Akita, et al ( 2) observed that there were- two major types offailure in transversely framed side structures which were penetratedstatically with rigid bows. A deformation type of failure occurredwhen the strain directly below the bow was less than about 0.3,while crack-type failures were associated with larger strains. Itappears from some dynamic tests on similar structural arrangements,which were ~eported by Akita, et al ( 2), that the energy absorbing mechanisms and f?acture types were similar to those observedin the corresponding static tests. However, the energy absorbedin a dynamic test was larger than that which was absorbed in thecorresponding static tests, a circumstance which was attributedto the influencebe remarked thatin a ship duringphenomenon which

of material strain-rate sensitivity. It shouldthis increase in capacity might not be realizeda collision because this is a highly nonlinearis very sensitive to size. Moreover, the

5


14/45

influence of material strainrate sensitivity cannot be properlyscaledup from a model to a full-sized structure which is madefrom the same material (18). Furthermore, it appears that noinvestigations, except Reference (35), have been undertaken toexamine whether the structural response of ships may be consideredto be static, or whether it is necessary to retain the influenceof inertia forces. It was suggested in Reference (106)that thestructural response of a panel in amarine vehicle during a severeslam could be accurately predicted with a static analysis, provialed the duration of the pressure pulse is longer than the funda-mental period of elastic vibration. Indeed, encouraging agreementwas obtained between the theoretical predictions according to astatic analysis and some experimental results which were recordedon a onequarter scale model of a section of the bottom of a U. S.Coast Guard cutter. However, the inertia terms must be retainedwhen the duration of a pressure pulse is short. It was shown inReference (35) that the structural response of the shell platingof the particular tanker design considered in Reference (46) couldbe predicted with sufficient accuracy using a static analysis.It would therefore appear worthwhile to develop further thesesimple ideas in order to provide guidelines which indicate whenstatic analysis could be used with no sacrifice in accuracy,although it is likely that the retention of inertia terms wouldbe unavoidable when analyzing even minor collisions of high-speedmarine vehicles.Akita, et al (2 ) and Arita (7 ) have developed approximate theoretical procedures which consider the energy absorbed in the shellplating, as well as various other members for both the deformationand crack failure modes which occur during ship collisions. Thesetheoretical predictions agree reasonably well with some experi-mental results recorded on idealized models which are reported inReference ( 2) and are further discussed in References (12) and(95). It appears that these theoretical analyses should bridgethe gap between the analysis of McDermott, et al (46) for minorcollisions in which the membrane energy of the shell plating isdominant and the analyses of Minorsky (48) and N.C.R.E. (96) formajcm collisions in which the membrane tension in the shell plat-ing of the struck ship is neglected. However, there are a largenumber of different assumptions in these various analyses so thatthe theoretical methods in References (2 ) and (7 ) do not agreewith each other and appear to neither reduce to Reference (46)for minor collisions nor to References (48) or (96) for majorcollisions.

It should be remarked at this juncture that relatively little isknown about the fracture of structural members which are subjectedto large dynamic loads. Apparently, the only investigation whichis relevant to ship collisions, is the experimental study onbeams by Menkes and Opat (107) and the subsequent theoreticalanalysis of the same problem which appears in Reference (108).It is clear that much further work is required on the fracture

6


15/45

of rectangular plates and grillages as well as in the effects ofsize before the theoretical procedures of References (2 ) and( 7) and others can be used with confidence for parameters whichdo not lie within those of the experimental tests. Indeed, it isnot anticipated that a theoretical method will be developed inthe near future which can predict accurately the structuralresponse during a collision between two ships. The chief virtueof the various available theoretical methods is that they allowa comparison of various designs and suggest the most favorablecollision protection arrangements.Current work in structural response of colliding ships is beingconducted by Genalis, Minorsky, and through a contract, by Hydronautics senior analysts. Of primary concern is the estimation ofloads which occur at impact, their duration, their magnitude,distribution and area of application. Previous work is beingevaluated and new techniques are being considered. For example,Faulkners work (22) is being applied. These loads will then beutilized in the numerical analysis of several structural configurations using sophisticated finite-element computer programsPartially due to high cost of such analyses, a comprehensivestudy of available numerical analyses techniques is being carriedout to establtsh the most suirable one.Simultaneously, a longer range plan is pursued where a collisionsynthesis model will be produced based on the individual behaviorof structural components and the statistics of overall structurebehavior.

2.2 U. S. AND FOREIGN DATA SOURCES2.2.1 United States

A. M. Rosenblatt & Son/U. S. Steel (1971-1975)A series of studies were conductedt sponsored by theU. S. Coast Guard intended to develop a methodologyfor the analysis of minor collisions. In addition anumber of collision inspection reports are available.See References 46, 102, 103.

B. MARAD (Current)Under MARAD sponsorship George G. Sharp, Inc. is usingmethods developed by Professor Reckling (University ofBerlin) to predict forces observed in GKSS collisionexperiments. In a parallel MARAD funded effort, Hydronautics, Inc. has developed a finite-element model ofthe GKSS energy resisting barrier to predict the elastoplastic response to known input dynamic loads. SeeReferences 49 thru 53.

7


16/45

c. Gibbs & Cox, Inc. Design ManualAt the time of the design of the N. S. SAVANNAH inthe late 1950s by George G. Sharp, Inc. an independent study at Gibbs & Cox, Inc. was funded by MARAD.The product of this study was a design criteria manualfot nuclear-powered ships. See Reference 109.

D. U. S. Coast Guard Casualty ReportsThese are reports of all collisions which eitherinvolve vessels of U. S. registry or occur in U. S.water. The reports are maintained by the U. S. CoastGuard QffTce of Merchant Marine Safety.

E. U. S. Naval Safety Center (Norfolk, Virginia)This office maintains records of all collisionsinvolving U. S. Navy vessels.

1?. Massachusetts Institute of Technology , Department ofOcean EngineeringAt the present time the Department of Ocean Engineeringof M.I.T. has a contract from the Ship Structure Committee to gather and monitor R&D work in the area ofship collision damage both in the U. S. and abroad.The SSC project title is Surveillance of Ship Collision/Stranding Research Studies (SR1246). The PrincipalInvestigator is Professor Norman Jones.

G. University of Rhode Island, Department of OceanEngineeringN.M.R.C. sponsored a graduate research project at theDepartment of Ocean Engineering of the University ofRhode Island. Model tests simulating ship collisionswere conducted in the elastic range. Accelerationsand velocities were measured at two points. Impactoccurred on transverse steel bulkhead. An added masscoefficient of 0.39 was inferred. See Reference 11O.

2.2.2 GermanyUnder the supervision of the Geselischaft fur KernenergigieverWertung in Schiffbau und Schiffahrt (GKSS), Geesthacht-Tesperhude, a series of dynamic collision tests were con-ducted from 1967 to 1975. Three of the tests were conductedwith absorption barriers of the OTTO HAHN type> nine testswith resistant barriers with various bow configurations.The latter fall within the low-energy collision definition.,

8


17/45

2.2.3

2.2.4

(no shell rupture). Investigators in these efforts areMr. G. Woisin and Dr. Letnin of GKSS and ProfessorReckling of the University of Berlin. See References67 through 86.JaDanDr. Y. Akita of the Japanese Classification Society hasreported on collision research in Japan. Interest in theproblem, as in the case of GKSS, stemmed from interest innuclear powered ships as far back as 1958. Experimentswere conducted as early as 1963 with the greatest activityoccurring during the 19661969 time frame. See References1-3, 5-7, 54, 59, 66.ItalvDuring the mid 1960s the Italians, principally under thedirection of Professor F. Spinelli of the University ofNaples conducted a total of 24 tests on collisions ofvarious configurations. Test results have been reportedtogether with an analytical treatment. See References10, 61-65.

2.3 ANNOTATED BIBLIOGRAPHYThis section gives brief summaries of those documents which wereconsidered to be key sources of information for the study on low-energy collisions. A number of the summaries presented are basedon those given in the following two reports which describe the state-of.the-art Up to the time they were written. These are:(a) Tanker Structural Evaluation, M. Rosenblatt & Son, Inc.,

Report No. 2087 prepared for the Department of Transportation,U. S. Coast Guard under Contract DOT-CG-1O,6O5A, April 1972.

(b) Report on Ship Collision Study, Present Situation Survey,George G. Sharp, Inc., Report 5516 prepared for Babcock &Wilcox Company and the U. S. Maritime Administration, November1975.

Additional summaries are presented to cover key publications writtensince the time the above two documents were published and to includeother older documents which are relevant.1. Preliminary Analysis of Tanker Collisions and Grounding

U.S.C.G. Office of Research & Development - Project 713112,by David M. Bovet, January 1973.This report presents the results (of 51 collisions and 13grounding) of a preliminary analysis of tanker collision andgrounding data. Statistics are presented for the geographicallocation of collision. Collisions are analyzed in terms of

9


18/45

vessel size, vessel speed at time of occurrence, angle ofcollision, depth of penetration and geographical location.Correlations of penetration depth with striking ship speed,momentum, and energy are attempted. A brief analysis oftanker grounding is presented. Diagrams are given, and acomputer analysis program is exhibited.The sample used is small and the size of struck vessels isalso limited. A large percentage were struck beam on and ahigh percentage of collisions were in harbors or approaches.The method needs extension with more data and a broaderrange of vessels.

2. Tanker Structural Analysis for Minor Collision, by J.McDermott, R. Kline, E. L. Jones, N. Maniar and W. P. Chiang,SNAME, 1974.Mathematical models and experiments were studied for bendingand buckling followed by membrane stretching with and withoutweb frame failure up to hull rupture. Both single and doubleskinned ships are studied.It is shown that for minor collisions membrane tension providesa large part of the energy absorption. Comparative values aregiven in tabular form for impacts at different points of thespan between web frames.Even though the paper attempts to simplify the problem by hedging it in by numerous assumptions, one gets the definiteimpression that it is difficult to contain the problem withinbounds: the degree of restraint exerted by web frames on thesupported panel is not obvious, the support provided by theinner hull of a double hull ship would seem to depend on webspacing, depth of cofferdam and waterplane angle of strikingbow, which is not brought out. The allowable penetration andeffective resistance to rupture is a function of strike location in the plan view (with respect to webs) and in elevation(with respect to decks); this is not very clearly brought out.The authors conclude that the method followed does not lenditself to becoming a design tool. The paper treats only thefirst instant of a major collision.

3. Ship Collisions at Varying Angles of Incidence, Report No.N.C.R.C./N. 163, by F. H. Haywood, Naval Construction ResearchEstablishment, St. Leonards Hill, Dunfermline, Fife, February1964.The paper presents the results of a mathematical analysis ofthe energy absorbed in an inelastic collision for given initialvelocities and masses at various angles of collision. Theformulae used are developed and discussed. The results arepresented Tn graphical form and indicate that:

10


19/45

A. For collisions amidship, and if the displacements of thetwo ships are nearly equal , maximum energy absorptionoccurs at a collision angle of:(i) 90 when the struck ship has zero initial velocity.

(ii] 160 to 180 when the struck ship has an initialvelocity in the range of one half to double the velo-city of the striking ship.

B. For collisions at bow or stern of struck ship, if the displacements of the two ships are nearly equal, the maximumenergy absorption occurs at a collision angle of approximately 180 regardless of the relative masses of the twoships.

c. For collisions at bow or stern of the struck ship, themaximum values of rotational energy absorption for allcombinations of ship speeds occur at collision anglesranging between 70 and 90 and is at the maximum whenthe struck ship has zero initial speed.

D. For amidship collisions, the energy absorbed is maximumat collision angles ranging between 160 and 180 andincreases with velocity of the struck ship.

This paper presents a convenient tool for evaluating the relative energy in ship collisions for various combinations of shipweight, collision course and point of impact on the struck ship.Assuming the mathematical me~hod of analysis presented can beverified by spot check experimental results, this paper offersa simple inexpensive method of establishing an optimum modeltest program.After completion of the model tests the formulae presentedin this paper could be calibrated to provide a convenientmethod of comparing and evaluating intermediate combinationsof ship weights, collision courses and relative velocitiesnot covered by the model tests.

4. A Theoretical Note on Ships Collisions, by J. H. Haywood,Report No. R.445, Naval Research Establishment, St. LeonardsHill, Dunfermline, Fife, February 1961.A ship collision is analyzed theoretically in the case of aship striking a stationary ship amidship and at right angles.Calculations are carried out assuming the.collision force iseither constant or varies linearly with depth of penetration.The total work done, collision force, penetration, duration ofcollision, and energy partition are examined including theenergy absorbed by transverse vibration of the ship as a beam.

11


20/45

This.,paper may be the only one trying to solve the equationsof motions, which are greatly simplified by limiting the solu-tion to a particular case. The solution and the conclusionsfrom the investigation is applicable only to the cases ofdirect, central, and symmetrical impact. Since the reactorcompartment is usually near the stern, some of the conclusionsdo not apply. Also treating the ships hull as a uniform beamis not realistic. Even though this paper is still too simpleto be practical, it is much better than the approach byCastagneto and the others.Haywoods conclusion about zero collision force and zero workdone is not practical. This conclusion is obtained onlybecause he has used the constant force assumption. Accordingto all experiments the force at small duration is mostlylinear and then becomes more or less constant. Practically,ships involved in collisions never behave like perfectlyelastic rigid bodies.

5. An Analysis of Ship Collisions with Reference to Protectionof Nuclear Power Plants, by V. U. Minorsky, Journal of ShipResearch, October 1959.Ship collisions are assumed to be almost wholly inelastic. Arelationship for kinetic energy lost in the collision isdeveloped based on ship displacements, speed V of strikingship and added mass of water assumed to be .?0 that ofstruck ship.An empirical linear relationship was found to exist for high-energy collisions between lost kinetic energy and a resist-ance factor RT which includes the volume of structural memhers which are edgeon to direction of collision, such asdecks, flats, etc. in both ships, longitudinal bulkheads ofstriking vessel, and a component of striking vessel shelltaken at .70 of sheli area 5 ft. deep in way of deck.An allowance of one-third increase in RT is made for theforward speed of struck vessel.The method allows an approximate calculation of depth ofpenetration into ships of conventional design, or of thecritical speed for which a known bow will reach a certainpenetration into a given ship side.It is implicit that the collision occurs approximately atmidships where the energy spent in the collision and hencethe penetration are at a maximum value.This easily applied approximate method is useful in the caseof collisions involving conventional ships where collisionprotection is of the absorbent type.

12


21/45

6. Collision ?roblems for Nuclear Ships, by G. Woisin, Hansa1964, No. 10. ,,

This paper is a review of the state of knowledge of the sub-ject (1964) covering the dynamics of collision includingresearch on added mass, a discussion of the SAVANNAH collision analysis and the energy spent in elastic deformation.A reviewis made of all of the experimental work and of thevarious collision protections schemes. Much of this materialis covered in other papers by the same author.There is little that is new in this paper but it has value asan overview of the subject at the time it was written.

7. Collision Considerations in the Design and Construction ofthe SAVANNAH, by J. A. Dodd & S. MacDonald, Motor Ship,November 1960.The paper reviews the SAVANNAHs characteristics, containmentvessel support and scantlings, shielding and collision pro-tection including the method of calculating critical speed.The paper is a convenient summary but does not add anything new.

8. Estimating the Decelerations Sustained in Ship Collisions, byG. Woisin, Schiff und Hafen 13 (1961) November.The author discusses the importance of establishing the deceleration(also acceleration) of ships in a collision and indicates it can becalculated simply if the total penetration is known from the speeds andmasses using the relationship established in the SAVANNAH study, withand without the resistance of the water. He also examines the extremecase of a ship ramming a rigid quay wall. Finally, he calculates theduration of the deceleration using a graphical method.The author states that the calculated decelerations are toohigh, approximate and uncertain. It would appear that relationships between penetration and time for given vessels wouldbe better obtained from tests.

9. Research on the Collision Resisting Construction of the Sidesof a Nuclear-Powered Ship, (Report No. 3) by H. 01, T. Harima,H. Iizuka and G. Kataoka, Mitsubishi Nippon Heavy IndustriesTech. Review .4 (1963).Loads were applled by a swinging weight, also statically by ascrew jack to ship side models built to a 1/15 scale. Modelswere of two types, S1 with 3 decks and transverse framing, andS2 with 2 decks and additional frames and longitudinal stringers.Several types of bows were used including B1 simulating a rigidbow. In some experiments (Group 1) only the side model wasdeformed; in others (Group 2) only the bow, and in Group 3

13


22/45

both bow and side models were deformed. Relationships weredeveloped between energy absorption and structural deforma-tion. The following were measured or investigated:1.2.3.4.5.

The1.

2.

3.4.

The

Velocity of bow model and deformation of side model.LoadLocal strains (by means of strain gage)High-speed camera pictures of destructive process.Final configurations of models after experiments werecompleted. (Extent of damage).results were:If both bow and side are destructible, the load - deforma-tion relation for each is almost the same as for collisionof bow with a rigid wall or of the side with a rigid bow.The maximum load on a bow colliding with a rigid wall isthe maximum load on the bow shell panel forward enclosedby decks and frames.There is an optimum thickness for shell plating whichdepends on several variables.When bow is stronger than side shell, the energy is absorbedalmost entirely by the side structure.derivation of equations is not obvious and the theory

does not seem easy to apply, but useful work is reported.10. Research on the Collision Resisting Construction of the Sides

of a Nuclear Powered Ship, Report No. 2 by T. Harima, S.Yamada and Y. Tokuda, Mitsubishi Nippon Heavy Industries Technical Review 2, 1961.The paper gives test results for the impact resistance of beams,flat plates and stiffened plates. The effects of strain hard-ening, strain speed and variation in yield stress are studied.Some of the results are:1.2.

3.

4.

5.

Absorbed energy = Plastic moment x Bend angleLoad-deflection relationship for simple plate fixed atboth ends can be calculated considering tensile stressonly in the axial direction; with stiffened plate thecalculation must include the bending stress too.The load deflection relationship can be calculated takingthe simple square plate as a circular membrane and thestiffened square plate as a grid.The above calculations must take into consideration, strainspeed and strain hardening as variable coefficients.Absorbed energy per unit deflection favors HT overMs , but energy to rupture favors MS over HT.

14


23/45

11. Equivalent Added Mass of Ships in Collision, by S. Mptorla,,M. Fujino, M. Sugiura & M. Sugita, JSNA, December 1969.The authors have calculated and verified by experiments thatthe added mass of the ship which is .40 at first is not constant but varies during the collision, increasing with durationof impact. They calculate the acceleration at the end of thecollision dividing the external force by an equivalent addedmass.The highest values of added mass are for the case of a softstructured ship struck at low speeds with a considerable amountof penetration. In the case of a ship with strong collisionprotection struck at high speed with relatively low penetration,the added mass is very close to the .40 assumed for the N-S.SAVANNAH.

12. Model Testing with Collision Protection Structures in ReactorW* by G Woisin? Schiff Und Hafe~* July> 1972The author states that he is concerned only with dynamictests, with only slightly simplified models. He is not concerned here with the influence of the water surrounding twoships in collision. The test stand is described, followed bya discussion of scale effects.It seems questionable to state that successively greaterimpacts on the same model can be produced without incurringa strain hardening effect different from that which would beproduced for a maximum impact. There is no clue to thebrittleness scale effect.

13. A Study on Collision of an Elastic Stem with the Side Structureof a Nuclear Ship, by Yoshio Akita & Katsuhide Kitamura,B.S.R.A. No. 35300, Sot. of Nav. Arch. in Japan, 1972.Collisions test results of 1/10 scale models are compared withcalculated values, using the Minorsky method. Tests of onemodel of hull side structure of the struck vessel were made inconjunction with six (6) bow models of the striking ship featuring variations in framing and scantlings. Two of the bowmodels were transversely framed and four (4) were longitudi-nally framed. The six (6) bows were graded from soft tohard based on their ability to resist deformation in a collision. The transversely framed models were on the soft endof the scale, and the longitudinally framed with heavy scantlings was.at the hard end.The portion of the collision energy absorbed by the sideand the bow modelsvalues calculated by

is measured and compared with theoreticalthe Minorsky method.

15


24/45

This paper presents information on model test results andconcepts having direct application and/or provides guidanceon the design of a suitable protective structure in way ofthe reactor.

14. A Study of Similarity Laws of Impact Damage, in particularof Ship Collisions and Collision Model Tests, by G. Woisin,Schiff und Hafen 20, 1968.Laws governing similitude are listed and it is brought outthat it is desirable to satisfy Kicks law concerning stressesand strains as well to have equal strain velocities; at thesame time, if the impact is sudden, there are inertia forcesto consider which introduce a Newtonian dynamic similarity;however, for the forces exerted on water, the liquid beingthe same for ship and model, ~is-cosityforces are the same,and so are gravity forces. All this leads to the conclusionthat not all conditions can be met by one set of similitudelaws. The author goes on to explain that some of the rela-tionships can be neglected as being of little significance.Some of the considerations that cannot be neglected are:1. Temperature (brittle fracture)2. Strain hardening3. Molecular structureThese will introduce scale effects.

Empirical formulae are given that reconcile full-scale resultswith model tests to various scales and a corrected similaritylaw for tests in the dry is developed.

The author proposes that similarity laws be verified comparingships, statistically similar cases, and models, so as torefine the relationship taking into account various scaleeffects.It is pointed out that at small scales more energy must beapplied--as much as 110% more at 1/15 scale than for fullscale--because of strain hardening.

Experimental verification of many of the relationships pre-sented is missing. It is suggested that wellknown casehistories and statistical data be applied to:1. Actual ships2. Partly ships, and partly models3. Model tests

16


25/45

The difference due to strain hardening alone at scales of1/7.5 and 1/15 instead of (1/2)3 = 1/8 = 0.125 is about 38%higher or 0.174. The differe~ce between full scale and 1/15scale is about 110%.

15. Analysis of World Merchant Ship Losses, by W. J. Beer,RINA,March 28, 1968A general statistical summary of various types of ship lossesof vessels insured by Lloyds over the period 1949 thru 1966,grouped by ship size for both tankers and general cargo ships.The losses are categorized as follows:1. Wrecked2. Foundered3. Collision4. BurntThis paper provides a broad general picture of the causes ofship losses over the period of 1949 - 1966 but contains littletechnical data applicable to collision analysis.

16. The Distribution of Collisions in Japan and Methods of Estimating Collision Damage, by Yahei Fujii and Hiroyuki Yamanouchi,B.S.R.A. No. 35,299.Presents data on all recorded collisions along the coastlineof Japan during the period 1966 1968. The data is categorized by ship size and by location of collision; i.e., inharbors and outside of harbors. The latter category is furthersubdivided into several major coastal areas.The statistics presented in this paper indicate that most ofthe collisions in Japanese waters involve small vessels ofunder 500 tons. No attempt is made to correlate the colli-sions with their probable cause other than to grade them byship size and geographic location, as noted above. However,the paper presents a rather sophisticated mathematical procedure by which existing collision statistics could be adjustedfor variations in ship size, traffic density, etc.

17. Research on the Collision Resisting Construction of the Sidesof a Nuclear Powered Ship, by Kagami, et a~, MitsubishiNippon Heavy Industries Technical Review, Fol. 2, No. 1, 1961.Models to 1/20 scale of a 45,000 DWT tanker with nine (9)different side structures were hit by a pendulum simulatingthe bow of a 45,000 DWT tanker at 5 knots. Results werecompared qualitatively for damage depth, stresses and impactforces developed.

17


26/45

The experimenters conclude that the side shel-1should be mademubhstronger than the striking bow, that the reactor wall-andthe side shell should not be structurally connected, that a ,grid is the best stiffening for the side plating, and that itis effective to increase the side plating thickness.The tests do not include any variation in kinetic energy; thekinetic energy is quite low. Some of the configurations areof no interest. The experimenters in their conclusions overlooked the significance of the very high impact forces,oftest T1 and the corresponding acceleration from the standpointof the reactor control rod design.

18. Destructive Energy in Ship Collisions, by E. Castagneto,Technics Italiana 27, No. 10, December 1962.Formulae are developed to obtain the energy developed atimpact, both direct and off center, showing the advantage ofplacing the reactor aft. Also the proportion of energy spentin elastic deformation of the hull is studied as well as therelative strength of anticollision structures. The addedmass of water is discussed and experimental results are pre-sented.These results were:1. Experiments are advisable to determine added mass of water

in the case of ships other than tankers.2. The energy lost in elastic deformation is small (less

than 5% of total).3. The strength of collision protection should not be so great

that the collision will cause excessive stresses in the hull.4. Off-center impacts produce less energy loss than impacts

at the e.g. of struck ship and it is preferable to placethe reactor at the stern.

19. Studies on Collision Protective Structures of Nuclear PoweredW* @ ~. Akita and 3d Nuclear Ship Research Committee.Shipbuilding Research Association of Japan Report No. 71.This paper is a summary of all Japanese research work oncollisions between 1966 and 1970, representing work by 13panels combining private industry and government organizations.A large part of the work consists of attempting to derive equa-tions, based on theory and experiment, with which to analyzecollisions. Other subjects investigated deal with the addedmass of water, distribution of absorbed energy in a collisionas a function of relative stemside shell strength, stem angleand collision angle variation and the relative strength ofvarious side structures.

18


27/45

This-paper represents the results of a good deal of useful.experimental work. The so


28/45

21. Strength of Huge Tankers in Collision, by Toshiro Suhara,et al, Journal of Society of Naval Architects of Japan,Vol. 128, 1970.This paper reports the results of static penetrating testsusing l/15scale tanker models. The model of the struck shipwas a portion of the wing tank of a 400,000ton tanker. Themodel of the striking ship was the bow of a 100,000ton tanker.Two solid bow models were used, one was a normal bulbous bowstem, the other was a vertical stem with wedgeshaped crosssection. Damage patterns were observed during the staticpenetration of the bow into the wing tank, and load penetrationcharts were obtained.The experiments showed that a deck strut and its adjoiningstructural members such as shell plate, transverse wing, andhorizontal girder which withstand the collapse of the strutprovide the largest portion of the total resistance forceagainst penetration. By assuming that a strut and its adjoining members maintain constant resistive force after buckling,an approximate method of calculating the load versus penetration was obtained. This approximate method of calculatedenergy absorbed was about 10% less than that measured duringthe experiment.The total energy available in the testing machine was lessthan that necessary for collapsing the model when the verticalstem wedge was used. A gas cut about 5 feet long was made inthe shell to initiate failure. The profile of the bulbousbow was such that the stem at the main deck punched throughthe shell of the struck ship first and the bulb provided aconcentrated load which initiated local failure. In neithertest was the total potential membrane tension plastic energyin the stiffened side obtained - in the vertical stem case asa result of the gas-cut, and in the bulbous bow case as aresult of the punching action.

22. The Probability of Vessel Collisions, by T. MacDuff, OceanIndustry, September 1974.Based on Channel width and stopping distances in the case forgrounding in the Straits of Dover, and on ship speed, meanspacing, length and angle to stream in the case of collisions,the author develops the mathematical probability of randomgrounding and collisions, i.e. without the benefit of anynavigational aids. He then compares these with the actualfrequency of groundings/collisions, and ascribes the differ-ence to causation probability. He goes on to apply thisconcept to the possible collision of a ship with a North Seaplatform.

20


29/45

23.

24.

The Safety of the Nuclear Reactor on Merchant Ships, by FrancoSpinelli, Tecnica Italiana, pp. 797-812, 1963, Technics Navale.Relative to a scale-model study of ship crash phenomena, atable of scaling laws was prepared, based on the velocitiesand pressures being the same in the model and the prototype.The ratio of the prototype energy to the model energy was considered proportional to the cube of the dimension ratio; thedimension ratio was assumed to be the same in any direction.Criteria for Guidance in the Design of Nuclear Powered MerchantUir@ w Gibbs L Cox? ~ncYprepared for the Office of Researchand Development, Maritime Administration.This 3volume paper treats virtually all aspects of ship designwhich may be considered unique to a nuclear-powered ship. Sec-tion 4 on Collision Barrier is relevant to the subject ofthis project. It covers the following:i. Probability of ship penetrating a collision barrierii. Design of absorbent, resistant and combination collision

barrieriii. Mechanics of collisioniv. Model test of simple wood absorbent collision barrierv. Energy absorbing characteristics of conventional ships

structurevi. Calculation of the force required to crush the bow of a

Mariner class shipUnder (i) the paper provides large quantities of statisticaldata on collisions. Based on certain selected data it suggestsa method for determining the number of collisions per year inwhich a container would be penetrated and the criteria fordesigning an absorbent collision barrier.Under (ii) the paper outlines the approach to the design of anabsorbent barrier with particular emphasis on wood, laminatedsteel and wood, and steel barriers resembling conventionalship structure. It also discusses design of resistanttypebarriers, however, it recognizes the lack of much essentialdata which could result in developing an overly conservativedesign.Under (iii)the paper develops an equation for the total energytransfer in collision. It reasons that it is possible to analyzethe collision on the basis of an inelastic collision in a fric-tionless medium. In a numerical sxample the mass of the struck

21


30/45

.

vessel is doubled to account for entrained water while noincrease is made to the mass of the striking vessel for theentrained water.With respect to steel structures, Gibbs & COX state that Theuse of steel structures to absorb energy by collapsing andrupturing is a possibility. However, it has not been possibleto find a reliable rational approach to calculating the energyabsorbing characteristics of even relatively simple steelstructures. (This appendix) reports an analysis of collisionsbetween conventional ships and an attempt to obtain a correla-tion between volume of steel structure demolished and energyabsorbed. Various other correlations with energy absorptionwere attempted but none appeared superior to metal volume.The force ~equired to crush the bow of a Mariner is calculatedassuming (1], the bow acts as a column in collapsing and (2),the bow plating and stiffener collapse. The force valuesreported are 25,000 tons and 19JO00 tons.It should be pointed otitthat the statistical data on the worldfleet of ships and collisions data used by Gibbs & Cox is outdated.Gibbs & Cox data is mostly pre-1958 when the ships were relativelysmaller and slower.

25. Collision Protection of Nuclear Ships - A Survey of the Stateof the Art, by Odo Krappinger, University of Michigan, College?f Engineering, May 1966.The paper surveys significant literature published since designof the N. S. SAVANNAH and concludes that only modest progresshas been made in the field of collision protection of nuclearpower plants. The paper attempts to organize the problem ofcollision protection, however, its treatment is uneven, andsome major aspects are given only cursory coverage. Whilethere is no new information presented herein, this paper is ahelpful summary for those looking for an introduction to theproblem.

26. A Scale-Model Study of Crash Energy Dissipating Vehicle StrucLures, by G. C. Kao and G. C. Chan, Wyle Laboratories ResearchStaff, Huntsville, Alabama, March 1968.Relative to a scale-model study of vehicle crash phenomena, atable of gravity scaling laws was prepared, based on acceleration, modulus of elasticity, stress, and strain being thesame in the model and the prototype. The ratio of the prototype energy to the model energy was considered proportional tothe square of the dimension ratio; the dimension ratio wasassumed to be the same in any directton. This assumption forenergy sc?ling is compatible with membrane-force energy, ratherthan bending energy.

22


31/45

27. Tanker Structural Evaluation, by M. Rosenblatt & Son, Inc.Contracti No. DOT-CG1O,6O5A, April 1972. ,.,.The purpose of this study was to examine existing tanker structural arrangements, determine those design features whichhave a significant influence on cargo protection, and providethe Coast Guard with a means of evaluating the relative effectiveness of various systems in preventing leakage after collision.A review of pertinent literature and collision histories wasconducted; boundary considerations were established; and analyticprocedures were developed which provide for an assessment ofthe plastic as well as elastic energy absorbed in a minor collision with an unyielding ships bow.Seven collision cases were studied. The striking ship was aT-2 type with an unyielding bow with either 15 stem rake ora vertical stem. The struck ship was a version of a 120,000-DWT tanker varied to include changes in shell material, changesin scantlings, variations in hit location, and single versusdouble hulls.

The analytical procedures developed in this study are forestimating the plastic and elastic energy absorbed by thestructure of a conventional longitudinally framed tankerstruck at its center of gravity in a right angle encounterwith an unyielding bow having a verttcal stem or a stem rakeof 15. The following conclusions are drawn based on theapplication of this procedure to minor collision cases (i.e.,where the collision will cause oil leakage) in which a 120,000-DWT tanker and its variations are struck by a T2 type tanker.It should be noted that the conclusions drawn relative todouble-hull ships are based on two simple hulls without theemployment of any special energy absorbing material locatedbetween them as the use of special interconnecting systemssuch as honeycombed structure.1. The procedure is effective in ranking tanker structure

from the viewpoint of cargo containment protection affordedin the event of minor collisions.

2. Collision energy absorbed in elastic deformations ofoverall ship structure will be negligible for practicalcollision situations. For elastic energy absorption tobecome significant, the struck ship must have exception-ally strong side structure, so that high collision forcesare generated and the striking ship is brought to rest ina period of time substantially shorter than the fundamental period of transverse vibration of the ship.

3. Single-hull ships are more efficient absorbers of collisionenergy than double-hull ships. The double hull is superiorto the single hull in the case of a punching or tearingaction where little energy is absorbed and the inner hullmay remain intact and prevent leakage.

23


32/45

4.

5.

6.

7.

8.

-Themost efficient way to increase the ability to absorbcollision energy is to increase the thickness of shellplating. If a double hull is desired, the most efficientplacement of material is to make the outer hull as heavyas possible and make the inner hull as light as possibleconsistent with hydrostatic and other design requirements.The spacing required to insure interaction of the two shellplating systems in the double-hull case is so small as tobe impractical from a construction standpoint. This doesnot take into account the possible use of such unorthodoxsystems as honeycombed structure.The shape of the bow of the striking ship has a significantinfluence on the energy absorbed. The greater the verticalextent of side shell which can be engaged, the greater theenergy required for failure.The effect of ambient temperature on rupture is significantsince no plastic energy can be assigned where the temperature is below the transition temperature.Although there may be a large relative increase in energyabsorption possible through increases in shell thicknessand strength, the collision energy absorbed before ruptureby conventional tanker structure is quite small. Withinthe collision cases examined, a T2 at 20,000 tons displacement and a speed of 3 knots would rupture the structuralconfiguration absorbing the largest amount of energy.Therefore, tankers of the same size are no~ likely to varyin their cargo containment capability after collision unlessradical increases in hull weight are accepted or unlessinnovative non-structural containment systems are used.

28. On the Collision Protection of Ships, by N. Jones, InternationalSeminar on: Extreme Load Conditions and Limit Analysis Procedures for Structural Reactor Safeguards and ContainmentStructures, Berlin, 1975.This article provides a brief survey on the literature avail-able on the collision protection of ships. It starts off witha brief discussion of the current state of knowledge on thecollision protection of landbased vehicles. It then reviewsthe experimental and theoretical investigations dealing withthe collision protection of various types of ships. Variousenergy absorbing methods are then discussed with the emphasisplaced on their suitability for the protection of ships involvedin collisions. The behavior of the honeycomb (hexagonal cell)structures is then investigated in some detail. Finally, alternate structural arrangements in ships which utilize hazardousenergy-absorbing systems are suggested.From this report, it appears that honeycomb structures providea feasible alternative to deck structures which are presentlyused to achieve protection of ships in collisions. The featureof the honeycomb panels are explored in various ways. A design

24


33/45

which utilizes both sides of the hull is proposed for a nuclear-powered ship involved in a collision. In some circumstances,a nest of tubes might be advantageous over the honeycomb panels,therefore this idea is described and its energy-absorption proper-ties examined. It is suggested that the honeycomb panels orthe nest of tubes could be used in conjunction with currentdesigns to provide additional protection.It is also recommended that supporting experimental evidenceon structural characteristics be obtained before using thedesigns in a ship or marine vehicle.

29. Ship Casualty Analysis, by V. U. Minorsky, George G. Sharp, Inc.November 1975.127 monthly casualty return sheets 1964 1974 from the Liverpool Underwriters Association were analyzed for ships over2,000 gross tons. Casualties were studied world wide, withspecial interest in casualties deriving from collisions andmore specifically for those along proposed nuclear ship routes.It was found that collisions and grounding remained constantfor this period while world fleet increased. In the periodthere were 831 ships in collision, 850 grounding and 977 fires.

30. Development of a Collision Protection Structure for NuclearPowered Ship, by G. Woisin, Institut fur Anlagentechnik.Recently, nuclear containership studies.were carried out withcontainerships having a small breadth. Because of this, thecollision barrier for the reactor compartment changed from theenergy-absorbing type to a resisting type. The former typeconsisted mainly of decks extending at least 1/5 the shipsbreadth into the ships sides. The resisting type consistedlargely of grillages fitted to the inner frame of the outershell of the ship extending to only 1/12 of the ships breadth.The new design was tested with 8 model exponents with the samemodel testing facility of GKSS.The tests showed that adequate protection against collisioncould be demonstrated by the resistance type barrier, Type 2(18 mm thick web plate). Because of the very minor penetra-tion depth protection can be achieved by carrying the grillagestructure from the outer shell to a longitudii~al wing bulk-head located approximately .07 B inboard rather than the .2 Bminimum safety distance required for energy absorption typebarriers.

31. The Collision Tests of GKSS. bv G. Woisin. GeesthachtThis paper describes the GKSS collision tests inbow models were impacted against ship side shell

which shipmodels. Three

25


34/45

tests were made using side-shell models with barriers ofthe energy-absorbing type. Nine tests were made with side-. ..shell models of the energyresisting type. Various bow configurations were used. The side-shell models were fixed toan encastered beam whose elastic properties simulated thoseof a ship. The report describes the applicable scaling laws,the test technique, the mechanics of the collision processand the results of the tests.The tests showed that it is possible to design an energy-resistant type barrier which can successfully resist virtu-ally any conceivable collision by constructing an egg-cartongrillage structure between the shell and the protected com-partment, in this case a reactor compartment. The testsdemonstrated that when a resistant-type barrier may be reducedfrom .2B, required for energyabsorbing barriers, to .08B.Thus the width of the compartment protected, whether it bededicated to cargo or a reactor is increased by 40% from.60B to .84B. Furthermore the barrier compartments could beused for carrying liquids.The report also asserts that a further advantage of this typeof construction is the watertight integrity of the ship which isalmost entirely preserved after a collision fn the reactor zone.

32. Report on Ship Collision Study Present Situation Survey, byV. Minorsky, George G. Sharp, Inc., November 21, 1975.Collision research since the SAVANNAH is briefly reviewed. Abibliography of 74 papers is given in four categories: (a)Research (41 papers); (b) Statistics and Probabilities (13papers); (c) Avoidance (13 papers); (d) Miscellaneous(7 papers). Of these, 14 were translated and 34 abstracted,the abstracts being included in the report.

33. On the Development of Design Criteria for Collision Resistance,by Richard J. Burke, S.U.N.Y. Maritime College, May 11, 1978..

This paper deals with collisions involving a ship carrying ahighly poisonous, flammable or explosive cargo which couldresult in widespread property damage and personal injury ina radius extending far beyond the colliding vessels. Adetailed account of the collision between the PACIFIC ARESand the YuYO MARU is given. A summary of research on themechanics of collision is then presented. This includedMinorskys high-energy collision approach and the rotationaleffects of an eccentric collision where the amount of energyabsorbed by the structure is reduced because some of it ispassed to rigid-body rotation. Also discussed Minorskysadded mass value as compared to the actual value which dependson the velocity of the striking ship and depth of penetration.Model testing was next discussed and design variations such assoft bow versus rigid bow, etc. The fact that a ships side

26


35/45

34.

35.

structure can absorb energy in more than one mode is discussed.Akitas work is briefly examined and the importance at low-energy collision research is expressed. The author identi-fies a number of factors that are significant with regard tothe amount of plastic energy absorbed. These are:1.2.3.4.5.6.

Relative strength of side shell plating and weakness ofweb frames.Position of impact point relative to strong transversestructural members.Shape of the bow of the striking vessel and verticalextent of encounter.Ductility and transition temperature of side shell plating,and the ambient temperature.Strength and type of transverse connections between outerhull and inner hull, if any.Angle of incidence at the point of impact.

The probabilisitc views of collisions are then dealt with andthe development of design criteria for collision resistance isexamined. The criteria developed would have to evaluate risk.Design criteria which are based on a probabilistic theory musthave statistical data. The value of data collected thus farfor detailed analysis is limited. On a national level, thesecriteria would be established by a governmental agency, butsuch an effort must involve all segments of industry.While not a highly technical paper, as such, it does give a goodpicture of the methodology presently being incorporated in shipcollision analysis. The difficulty in establishing design criteriais that presently no easy definition of acceptable risk is availa-ble. In addition, the criteria should be established on aninternational scale. The nature of such criteria depends in alarge measure on the broadly defined economics of the situationin question. However, the benefits should not be ignored.The Collision.Protection of Nuclear-Powered Merchant Ships,Lloyds Register of Shipping, May 10, 1967.This paper is a review of the work done on the philosophy of colli-sion protection with a view to assessing the data available todesign suitable collision protection for a large nuclear-poweredPolar Icebreaker.Tanker Structural Analysis for Minor Collisions, Final Report,M. Rosenblatt & Son, Inc., Report No. CG- D- 7Z-76, Dec=b= 1975-This report describes the work accomplished during the course ofthe project on the Evaluation of Tanker Structure in Collision.The intent of the report is to present the investigations performedin evaluating the phenomena that contribute to the ability of alongitudinally framed ship, particularly a tanker, to withstand aminor collision. A minor collision is defined as one in which the


36/45

The final output of the study is an analytical procedure andits numerical application for estimating the plastic energyabsorbed by longitudinally framed ships when involved in eithright angle or oblique collisions. A static analysis is emplThe plastic energy analysis has indicated that the most signicant energy-absorbing phenomena are membrane tension in the sshell (the most important), membrane tension in the deck, sheof web frames, and plastic bending of the sideshell. Componestructures tests and investigations of actual collisions wereperformed. Parametric analyses are also presented which consof the numerical application of the plastic energy analysis pcedure to six collision incidents in which a 120,000 DWT (andvariants) is struck by a 20,000 ton displacement ship. Anothobjective of the project was to perform an investigation of nrigid bows to propose methods of evaluating their significancLimitations of the procedure are recognized as: (1) the pro-cedure employs static analysis; (2) the striking bows are assinfinitely rigid; (3) damage to the structure does not extentthe bilge area; and (4) the possibility of the striking bow iiately cutting or punch-shearing the shell of the struck shipnot considered.

36. Ship Collision Dynamics and the Prediction of the Shock Envirment for Colliding Ships, by Michael Peter Pakstys, Universitof Rhode Island, 1977.The objective of this paper is to develop improved methods topredict the shock environment of two surface ships involvedin a collision accident at sea. The collision consideredinvolves the bow of the striking ship colliding at a rightangle with the midship bulkhead of a stationary vessel. Thecollision velocity is assumed to be low enough so that theship structure deformations are essentially elastic. Thetechnical approach involves integrated analytical and experi-mental developments. A comprehensive computer program wasdeveloped to simulate the ship collision process. Ship modelcollision tests were performed in a large laboratory water taShock acceleration values for several values of collisionvelocity were obtained for right-angle ship collision for twoinstrumented ship models about the same weight. On the basisof those limited collision tests with floating models, thefollowing conclusions can be drawn on the basis of the obser-vations and the measurements.


37/45

2.

3.

4.

5.

6.

7.

The maximum shock response on the struck ship need notoccur at the impact region. In the test case, the peakacceleration at the end bulkheads was 70% higher than that the center bulkhead where the collision force was acting during the contact with the striking ship.The collision load duration is determined by the decel-eration pulse of the striking ship and was found to be,independent of collision velocity.The peak acceleration on the struck and striking shipmodels and the peak collision force were found to bevarying essentially linearly with the collision velocityThis indicates that the ship models and the water mediumexhibited linear behavior, for the range of collisionvelocities used.The added mass coefficient for the struck ship vibrationcorresponded to the theoretical value for high frequencyvibration of 0.39 for the specific rectangular section.In the initial collision phase (when maximum shock accelerations occur) the water supporting the floating shipsacts essentially as a frictionless medium. Thereforeany in-air collision tests where the struck ship is con-strained would not give valid results for shock responseThe water medium is also slow to dissipate the vibrationof the struck ship since it takes over 100 cycles toreduce the vibration amplitudes by 50%.The tests have provided valid laboratory test data toestablish some experimental verification of the analyti-cal approach and computer program results developed bysenior authors for prediction of the shock environmentship collisions.

BIBLIOGRAPHY OF DOCUMENTS RELATED TO SHIP COLLISION DAMAGEThe documents contained in this bibliography have been selected fromtotal of 192 which were reviewed in order to determine their relevancto the problem of ship collision damage prediction. The key documenthave been briefly summarized in the Annotated Bibliography of SectionQ.L.3.


38/45

BIBLIOGRAPHY

5.

/.,iet_rich,R., Untersuchung Zur Beurteilung eines Flugzeugabsturzes aufalit:Seitenwand eines Nuklearen Containerschiffes, GKSS 75/E/13,Cesel]schaft fur Kermenergieverwertung in Schiffbau und Schiffahrt mbH.(;e,esthacht,1975.i~odd, J. A. and MacDonald, S., Collision Considerations in the Designand Construction of the SAVANNAH, The Motor Ship, 333-335, 1960.Dul:fey3 T. A. , Scaling Laws for Fuel Capsules Subjected to Blast,lm~):lctand Thermal Loading, Proc. Intersociety Energy ConversionConr. , 1971, Paper S.A.E. 719107.Eugelhrecht, T. W., Design Data for the Preliminary Selection ofHoneycomb Energy Absorption Systems, Hexcel Tech. Service Bull. 122.Euratom, B. V. und G. L., Technische Beurtoilung der Sicherheit derN. S. SAVANNAH, Bericht und 2 Erganzungsberichte, Brussel 1961 und 1962Ezra, A. A. and Fay, R. J., An Assessment of Energy Absorbing Devicesfor Prospective [Jse ~n Aircraft Impact Situations, Dynamic Response of


39/45

Faulkner, D., A Review of Effective Platings for Use in the Analysisof Stiffened Plating in Bending and Compression, Journal of ShipResearch, March 1975, Volume 19, No. 1, pp 117.Fujii, Y. and Yamanouchi, H. , The Distribution of Collisions in Japanand Methods of Estimating Collision Damage, Journal of Navigation,London, England, Vol. 26, No. 1, January 1973, pp 108-113.Green, R. J., Methods for Determining the Collision Performance ofSome Energy Absorbers in Automobile Bumper Applications, Denver Res.DRI 2581, 1973.Grimes, C., A Survey of Marine Accidents with Particular Reference toTankers, The Journal of Navigation, Vol. 25, No. 4, pp 496510, 1972.Guida, Aurelio, Analysis of the Similarity of Comparability of Resultsfrom Collision Experiments, Tecnica Italiana (Italian Technology),Trieste, 1964.Harima, T., Yamada, S., und Tokuda, Y., Research on the CollisionResisting Construction of the Sides of a NuclearPowered Ship,(Report 2), Mitsubishi Nippon Heavy Ind. Tech. Rev. 2 (1962), p. 147(Nr. 2) (BSRA - Translation No. 1373).Harine, T., Yamada, S., and Tokuda, y., Research on the Collision-Resisting Construction of the Sides of a NuclearPowered Ship (ReportNo. 2), Mitsubishi Nippon Heavy Industries Technical Review, Vol. 2,1963, p 147 (No. 2).Hawkins, S., Levine, G. H., Taggart, Robert, A Limited survey of ShipStructural Damage,r Ship Structure Committee, 1971. AD733085, SSC22f~.Haywood, J. H., A Theoretical Note on Ship Collisions, NCRE Report#R445, February 1961.Holmes, B. S. and Colton, J. D., Scale Model Experiments for SafetyCar Development, Int. Auto. Eng. Congress, Detroit, January 1973,Paper S.A.E. 730073.Holmes, B. S. and Sliter, G. E., Methods, Application and Cost Effec-tiveness of Scale Model Studies of Automobile Impacts,r Stanford Res.Inst., Project PYu2766, Final Tech. Rep., 1974.Johnson, W., Impact Strength of Materials, Arnold (London) and CraneRussak (U. S.), 1972.Jones, N., On the Collision Protection of Ships, paper presented a~the International Seminar on Extreme Load Conditions and Limit AnalysisProcedures for Structural Reactor Safeguards and Containment Structures,Berlin 1975.Jones, N., Written discussion of Reference (46) Trans. SNAllE,Nov. 1974.Jones, Norman, Damage Estimates for Plating of Ships and Marine Vehicl,,-,M.I.T. Report No. 7713, May 1977.Kagami, K. u.a., Research on the Collision-Resisting Construction ofShips Sides, Nuclear Ship Propulsion International Atomic Energy Agencn,,Wien 1961, S. 485/502 (deutsch: Hansa 1962 S. 2174/2180).Kline, R. G., and Clough, R. W., The Dynamic Response of Shipsr Hullas Influenced by Proportions, Arrangement, Loading and StructuralStiffness, Presented at the Spring Meeting, SNAME, July 1967.Krappinger, Ode, Collision Protection of Nuclear Ships - A Survey ofthe State of the Art, University of Michigan, College of Engineering,May 1966.


40/45

40.41.

42.

43.44.45.

46.

47.

48.

49.

50.51.52.

53.

54.

55.56.

57.58.

Kurtzemann, B., Collision Similitude Tests on Ship Models, NouTechniques Maritimes 1973, Paris.Lee, E. H. and Wiersbicki, T., Bounds on Deflections of Stiff Ifaces in Vehicles During Collisions, Stanford University Reportpreparation.Lowe, W. T., A1-Hassani, STS, Johnson, W., Impact Behavior of SScale Model Motor Coaches, Proc. I. Mech, E., Auto. Div. Vol. 1409-419, 1972.Macduff, T., The Probability of Vessel Collisions, Ocean IndustrVol. 9, No. 9 pp. 144-149, September 1974.McDermott, J. F., Chlorine Barge Collision. Structural AnalysiM. Rosenblatt and Son, Inc., New York, Project No. CG-7251.O.6,McDermott, J. F., Tests to Verify the Plastic Behavior of StiffHulls of Struck Tank Ships, Report 46.019-200(1), U. S. Steel AResearch Laborator, June 13, 1973.McDermott, J. F., Kline, R. G., Jones, E. L., Maniar, N. M., Chiw. P., Tanker Structural Analysis for Minor Collisions, Trans.Paper 10, November 1974.Miller, P, M., Hardware Development and Production Feasibilityrelated to Vehicle Structures and Exteriors Research, Symp. onVehicles Safety Research Integration, U. S. Department of Transption, Washington, D. C., 19-35, 1973.Minorsky, V. U., An analysis of Ship Collisions with ReferenceProtection of Nuclear Power Plants,r Journal of Ship Research, 11959.Minorsky, V., Study of Collision Effects, George G. Sharp, IncJanuary 19, 1977, MA-RD-920-77058.Minorsky, V., Probability of Collisions on Selected Routes, GeSharp, Inc., September 30, 1976. MA-RD-920-77021.Minorsky, V. U., Executive Summary Ship Accident Studies, GeorSharp, Inc., New York, New York, March 4, 1977.Minorsky, V. U., Ship Casualty Analysis, George G. Sharp, Inc.Report No. 5516-5 for Babcock and Wilcox Co. and U. S. Maritimeistration, November 20, 1975.Minorsky, V. U., Ship Collision Study Present Situation Survey,G. Sharp, Inc. Final Report 5516 for Babcock and Wilcox Co. andMaritime Administration, November 21, 1975.Motora, S., Fugino, M., Sugiura, M., and Sugita, M., EquivalentMass of Ships in Collisions, Selected Papers from the Journal oSociety of Naval Architects of Japan 7, 1971, 138-148, from J.S.Japan 126, December 1969.Perrone, N., Crashworthiness and Biomechanics of Vehicle ImpactDynamic Response of Biomechanical Systems, A.S.M.E. , 1-22, 1970.Reckling, K. A., The Contribution of Elasto- and PlastomechanicResearch Into Ship Collisions, translation: Beitrag der ElastoPlastomechanik zur Untersuchung von Schiffskollisionen.Robbins, R. A., Casual Factors of Collisions at Sea, AD-AO09 9Naval Postgraduate School, Monterey, California, March, 1975.Saczalski, K. J., Structural Problems Associated with the Predi


41/45

59.

60.

61.

62.63.

64.

65.66.

67.

68.69.

70.

71.

72.73.74.75.

76.77.

Sakai, T. and Ushioda, S., Investigation into the Strength of ShipHull in Collision, Trans. of West Japan Naval Architects Association,No. 155, November 1964, Vortrag am 13.5, 1964 (U).Singley, G. T., A Survey of RotaryWing Aircraft Crashworthiness,Dynamic Responses of Structures, Ed. G. Herrmann and N. Perrone,Pergamon Press, 179-223, 1972.Spinelli, F., Isti

Documents

CRITICAL EVALUATION of Low Energy Ship Collision Damage- Theory & Design Methodologies