safety of hsmv

7/29/2019 Safety of HSMV

1/41

1

The Specialist Committee on

Safety of High Speed Marine Vehicles

Final Report and

Recommendations to the 22nd ITTC

1 GENERAL

1.1Membership and MeetingsThe 22nd ITTC Specialist Committee to

study the Safety of High Speed Marine Vehi-cles initially comprised eight members andhad four recommendations from the 21st ITTCon which to base its work.

Three of the members (Dand, Doctors andKeuning) had served on the High SpeedMarine Vehicles Committee of the 21st ITTC,the remainder being newcomers. Dr. Ian Dandwas appointed Chairman by the 21st ITTC andthe Committee elected Dr. Lex Keuning to bethe Secretary. Unfortunately, due to ill health,Dr. Keuning had to withdraw from the Com-mittee towards the end of its working life, andDr. Dand assumed the dual role of Chairmanand Secretary for the time remaining.

The membership was therefore:

Dr. I. W. Dand, UK (Chairman)Dr. J. A. Keuning, Netherlands (Secretary)Prof. H. H. Chun, KoreaProf. L. Doctors, AustraliaProf. G. Grigoropoulos, GreeceMr. P. Grzybowski, PolandProf. M. Takaki, JapanDr. S. Vogt Andersen, Denmark.

The Committee met four times in the three

year period since the last conference as shownin Table 1.

Other communications between members

were accomplished by means of e-mail, faxand telephone.

TABLE 1

Date PlaceNo. At-tending

April 1997 London 5

July 1997 Sydney 4

December 1997 Lyngby, Denmark 8

November 1998 Athens, Greece 6

1.2Recommendations from the 21st ITTCThe following recommendations for the

work of the Committee were made by the 21st

ITTC:

Study the dynamic instabilities of highspeed craft and develop procedures to sol-ve problems relating to high speed roll,

pitch and directional stability anomalies. Develop by means of test procedures and

computer codes, information on dynamicinstability which can be used to improvecoverage of this topic in the IMO HighSpeed Craft Code.

Catalogue incidents and accidents to highspeed passenger-carrying vessels to identi-fy trends and areas of hydrodynamic in-adequacy


2/41

2

Develop full-scale test procedures to de-fine and determine high speed craft safety.

2 THE SAFETY OF HIGH SPEEDCRAFT AND THE ITTC

In the short period of time since the lastITTC the fast ferry has established itself as afixture in the maritime scene. While there arestill technical challenges to be solved in itsdesign and operation, the concept is now suf-ficiently mature to warrant a comprehensiveoverview of its impact on the maritime popu-lation. It is the impact on safety which hasreceived the most attention, reflecting a world-

wide concern about safe practices, safe opera-tions, indeed the growth of the safety culture.

National administrations have carried outFormal Safety Assessments of fast ferry op-eration and the International Maritime Organi-sation, IMO, is already in the process of re-viewing and revising its Code of Safety forHigh Speed Craft. In view of the fact that thisCode was published comparatively recently,the fact of its review is a measure of both thespeed at which this sector of the marine com-

munity is developing and also concerns for theconsequence of any accident involving such avessel.

Two disciplines are involved in any safetyassessment, one being technical, the other op-erational. The link between them is some-times hard to see, but it is clear that more andmore hydrodynamic test establishments are

becoming directly involved with the relation-ship between safe operation and design. TheCommittee is therefore of the opinion that,whereas some of the disciplines used in thestudy of safety may be foreign to ITTC mem-

bers, they nevertheless have a place in its de-liberations. The Committee therefore saw itsrole as:

indicating some broad areas of study inwhich operational safety and vessel dy-namics combine,

helping to forge the links between thevarious disciplines involved. This may bedone by providing information for thoseinvolved in model experiments to deter-

mine whether a given craft design will beoperationally safe or not.

In our deliberations, aimed at satisfyingthese goals, it became apparent that the re-commendations made by the last Committeehad, in some degree, been overtaken by the

pace of development. Some were no longerappropriate, while others needed more work,due to inadequacies in present databases.(Relevant accident statistics are the prime ex-ample.) In what follows, therefore, attemptswill be made to adhere as closely as possibleto the guidelines provided by the Recommen-dations, but some deviation will be inevitable.

We have assumed safety to embrace thecomplete avoidance, (or minimising the con-sequences) of an accident, where an accidentis defined (Werenskiold 1998) as

A sudden unintended event that could re-sult in serious injury, or death, to passengers,crew or third parties.

It is well known that the main cause ofmost accidents in transport systems is humanerror and it has been shown that accidents inthe operation of fast ferries are no exception.Therefore, in order to minimise accidents, theoperation, as well as the design of the vesselmust receive detailed consideration.

As an example of this, it has been found(IMO 1997) that of the accident types towhich a fast craft is likely to be exposed, thatof collision is by far the most serious. Toavoid a collision some fundamental featuresof the vessel are needed:

it must be sufficiently controllable to bepredictable in its response and behaviourso that not only the crew, but also ap-

proaching vessels must be easily able todeduce its behaviour in the short term,

as a corollary of this, the vessel must besufficiently dynamically stable,

it must have sufficient power to be able tomake an appropriate speed in the prevail-ing conditions,

it must be sufficiently manoeuvrable.In addition it must have a sufficient re-

serves of stability to survive should a collisionoccur.

Virtually all of these points come within

the ITTC area of expertise. Therefore theymust be of interest to the members. But thedemands of safe operation go a little further


3/41

3

than mere study of the technology behind col-lision avoidance. In the above test, judgmen-tal words such as sufficient and adequateoccur without any definition. It is easy to

conduct a tank test to demonstrate manoeu-vrability, but it is more difficult to demonstratewhether the degree of manoeuvrability is sat-isfactory, sufficient or adequate.

It was on this crucial point that the Com-mittee agreed that a worthwhile part of itsoutput would be a list of criteria which helpedto define the words sufficient and ade-quate. This would allow the results of ex-

periments (the techniques for which are beingstudied by another committee) to be assessed

from a safety perspective.

A final general point to be made relates toan awareness of the safety of the environmentwhich has intensified since the recommenda-tions for this committee were formulated. Forhigh speed vessels this impacts two main ar-eas:

the power required to attain the servicespeed and hence the type of fuel used andthe exhaust emitted,

the wash generated.Both of these come directly or indirectly

within the remit of this Committee, and hencethe ITTC. Exhaust emissions can be harmfulin the long term and wash from fast vesselshas become a significant nuisance at best, anddanger at worst, to those on or near the shorewhen a fast craft passes. Hydrodynamic es-tablishments have been in the forefront ofstudying wash, both to understand and mini-mise it. By the use of model tests, an efficientvessel design can be obtained which can sig-

nificantly affect the power used and, therefore,the exhaust emitted.

With these considerations in mind, themain body of this report was produced. Anattempt has been made to adhere to the origi-nal recommendations, but some additionalmatter has been added.

Accordingly, the main body of our reportcomprises the following sections:

an assessment of the types of accidents towhich fast vessels are prone, to discernany obvious trends,

an overview of the important aspect ofhuman factors and their relation to safety.This involves internal and external vesselcontrol, communications and navigation

equipment, theories related to dynamic stability. The

hope is that, with such theories, craft maybe designed that behave in a safe and pre-dictable manner in order to minimise un-wanted problems for the human operator,

the behaviour of vessels in extreme situa-tions. This is the ultimate test of safe ves-sel design and links directly to its dynamicstability,

full scale tests, in recognition of the re-quirement of the IMO High Speed Craft

Code that compliance with the code mustbe demonstrated at full scale, wash, including the studies that have been

made both to understand and minimise itby design and operation,

criteria to link model and full scale tests toacceptable levels of safety.

In compiling this, the Committee has takennote of the following Conferences:

FAST 97, Sydney, July 1997, SURV IV High Speed Marine Craft, Safe Design and

Operation, NSF, Bergen, 1996, HSMV 99, Capri, March 1999

3 RESEARCH INTO THE SAFETY OFHIGH SPEED CRAFT

3.1Casualty AnalysisIntroduction

Any discussion of safety-related mattersmust be based on of the magnitude of the

problem, and indeed, if there is presently aproblem at all. The way in which this is doneis frequently by the assiduous collection ofcasualty data followed by its detailed analysis.This provides information on casualty rates,numbers of fatalities, type of vessel involvedetc. It can also, in some cases, give indica-tions of what was the main technical cause ofan accident.


4/41

4

The fact that fast passenger-carrying ves-sels are comparatively new on the maritimescene means that little historical casualty datais available. This problem is compounded bythe fact that there does not appear to be manyorganisations collecting, collating and pub-lishing such data on a regular basis.

However it is also becoming clear thathigh speed passenger-carrying vessels have todate maintained a generally good safety record,their casualty rates in port and harbour areasoften being an order of magnitude less thanthose of conventional ocean-going vessels.

It is against this background that this sec-tion of the report is presented. It had been ho-

ped that an analysis of casualties to high speedcraft would reveal particular problems of dy-namics and hydrodynamics which would lendthemselves to research and solution by ITTCmembers. The lack of much relevant datamade this an impossible goal in the life of thiscommittee. However, some relevant datawere found, and these are now discussed.

Categorisation of Accidents

It is clearly important to catalogue acci-dents to high speed passenger-carrying vesselsso that trends and areas of hydrodynamic in-adequacy may be identified.

In Figure 1 the worst passenger ferry dis-asters are categorised by cause (Lloyds Mari-time Information Service 1997). It is apparentthat wrecks and strandings dominate. Figure 2shows cruise ship total losses for the same pe-riod. The greatest number of accidents aredue to fire, while those due to collision andcapsizing are reduced. Accidents to conven-tional ships such as tankers and cargo vesselsexhibit almost the same trends.

Figure 1. Major Passenger Ferry Accidents,1963-1996.

Therefore, it appears that accidents to highspeed passenger ferries are related to their dy-namics and operation. Unfortunately thiscasualty data did not reveal the mechanism

behind the accidents and/or the relationshipbetween them and the hydrodynamic forcesacting on the vessels.

Figure 2. Cruise Ship Total Losses, 1963-

1996.

Fast Ferry Casualties

Some fast ferry casualties have occurredsince the last ITTC; a sample is given in Table1.


5/41

5

TABLE 2

Vessel(s) Date Type Craft Type

Hai Yang and

Man Boon

April

1997

Collision Catamaran/

conventional

ferry

Superferry 2 October

1997

Multiple

collision

Monohull

Flores May

1998

Mechanical

failure

Jetfoil

Laura June

1998

Grounding Hydrofoil

Sunnhord-

land/Kingtor

June

1998

Collision Catamarans

It is of interest to note that several of theseaccidents were collisions, emphasising theconclusion drawn by IMO (1997) that the riskof collision is the most significant hazardfaced by high speed ferries. The chance ofsevere injury is emphasised by the accident tothe Jetfoil which stopped abruptly from speedcausing injuries to 122 passengers.

3.2Human Factors and Navigation Equip-mentHuman Factors

Introduction. Due to its very nature, highspeed navigation sets demanding requirementsfor the crew. Fast ferries often operate in ar-eas of high traffic density and in restrictedwaters, while patrol boats and SAR vessels

usually navigate close to the shore, factorswhich make the stress on navigators high.The person in command of a HSMV has less(often much less) time to take an appropriateaction than that of a conventional displace-ment vessel.

Dynamically supported craft (planing, hy-drofoils, air-cushion vehicles), multiple hulls,and atypical propulsion and steering devices(e.g. waterjets) aggravate the problem further.

Navigating an HSMV is not like navigating an

ordinary vessel it is a different skill. Thus,in the interests of safety, special attention must

be paid to suitable training for HSMV crews.

Simulators in Training. Ship handlingsimulators are extremely useful in training

navigators of fast craft (Hagman and Ahlman,1996, Kaplan, Rmeling and Tveit, 1995). Anexample of such a simulator was presented byKaplan, Rmeling and Tveit (1995). Thefunctioning of a simulator can be based oneither model test data or a mathematicalmodel describing the hydrodynamic forcesacting on a vessel. The latter solution requiresgood qualitative and quantitative understand-ing of the hydrodynamic phenomena associat-ed with HSMV operation, but is not limited tothe particular vessel for which model tests

were performed.

Apart from its flexibility, cost saving andease of use, simulator training makes it pos-sible to create dangerous emergency situationsor to simulate unusual events likely to occurduring normal ship operation. Trainees can begiven the opportunity to cope with them by

proper instruction.

Simulators in Behavioural Studies. Ship

handling simulators may also be used to ac-quire knowledge of shiphandling and the deci-sion-making processes of fast craft navigation(Hara, 1991, Nagasawa, Hara, Nakamura andOnda, 1993, Hammer and Hara, 1990, Imazu,1995).

In the study presented by Hara (1991) theimportance of various factors affecting thesafety of navigation was determined by ap-

plying the Analytic Hierarchy Process to theresults of interrogations of a number ofHSMV navigators. The subjective level ofdifficulty of different encounter situations wasassessed and showed large individual differ-ences in navigators reactions. During simu-lation they paid particular attention to otherhigh speed vessels.

Hammer and Hara (1990) indicate thatknowledge of the intention of another vessel isthe most important factor in assessing the riskof collision. Thus communication directly

between vessels, or via a VTS centre, must beconsidered as very important. Transponders

may be used to interchange such informationautomatically.


6/41

6

An interesting approach is presented in apaper by Nagasawa, Hara, Nakamura andOnda (1993) where navigation is treated as atrade-off between risk of collision and physi-cal loss and the navigators reaction is a resultof a trade-off between his mental and physicalloads. Suitable measures are proposed to pre-sent these notions quantitatively. The termBlocking Coefficient is introduced, which pre-sents the overall risk of collision taking intoaccount all possible manoeuvres in a givensituation. Physical losses are assessed by thenavigators subjectively by the application oftheir preferred order of manoeuvres.

Results of the aforementioned studies canbe used to elaborate suitable training proce-

dures for both navigators and VTS operatorsor even influence the present regulations.

HSMV Safety Reporting System. Hu-man errors are the cause of many accidents;information on these is vital and must be col-lected and analysed. This can be done by thecollection and analysis of incident data.

Navigators should be encouraged voluntarilyto report all safety incidents and hazardoussituations, not necessarily attributed to human

errors. In return, confidentiality should beguaranteed. Such a system to record aviationsafety incidents has existed for many yearsand proved to be useful (Rosenthal, 1997).

Team Work and Man-Machine Inter-face. Since the workload on a HSMV navi-gator is very large, one person cannot dealwith all aspects of ship handling. Hagman andAhlman (1996) state that the bridge should bemanned with two experienced officers, onewho pilots the ship while the other assists andclosely follows progress on the route. The

piloting officer should always see that the co-pilot is well informed continuously aboutplanned actions.

The efficiency of crews can be increasedduring Bridge Resource Management courses(Wahren, 1996, Huth and Firnhaber, 1997)concentrating on communication, team build-ing, workload management and bridge re-sources. Their main target is to use all avail-able technical resources and to deploy humanresources in an optimal manner. A similarkind of training is commonplace in commer-cial aviation.

Navigation Equipment

Integrated Bridge Systems (IBS). Inte-grated Bridge Systems were created as a resultof the application of Risk Analysis and Man-Machine Interface principles together withmethods of ergonomics. Systems of this kindare described by Kristiansen and Tomter(1992), Lazarevic, Kuzmanic and Lakos,(1997), Pedersen, Ljungberg, Franck andBrathen (1992) or Kaplan, Rmeling andTveit (1995). The basic idea of their design isthat the navigator has all the controls and dis-

plays related to all aspects of ship operationeasily accessible from his seat. Informationfrom many sources is processed and filteredadequately for the current mode of operation(open water navigation, docking, evacuationetc.) and only the relevant portion is presentedto the navigator in an easy to perceive form.Time consuming activities, such as determi-nation of position and course, route planning,

position prediction, plotting of other vesselspositions and courses, are made automaticallyor are computer-supported.

Controllers and Predictors. Some highspeed passenger-carrying vessels now carrysophisticated controllers, some of which havea predictive capability. Kallstrom (1996,1997) describes the system developed for theStena HSS vessel which provides a high de-gree of automatic control almost up to the

berth. Even swinging manoeuvres are con-trolled automatically. The controller also hasa predictive capability which shows, on a

screen, the probable path the vessel will takeif the present control state remains unchanged(Figure 3). With the ability to look aheadfrom 30 to 120 seconds, such a system makesa significant contribution to safety by givingan early indication of whether the vessel isstanding into danger.


7/41

7

Figure 3. HSS Track Showing Predicted Po-sitions.

The speed at which a navigation situationcan change in a fast vessel and the need to beappraised of potential hazards in its path haveled to research into intelligent navigation aidswhich combine prediction with an expert sys-tem and chart information. Of particular valuein shallow and congested waters, such a sys-

tem gives the operator of a fast craft warningsof not only potential collisions, but also haz-ards such as shoal waters, piled beacons andother fixed structures. Use of electronic chartsand a mathematical model of the vessels be-haviour combine to produce a safety tool whi-ch owes much to tank test results. Such a toolformed one output from the Brite EuramSPAN (Safe Passage and Navigation) projectdescribed by Doyle (1999).

Electronic Chart Systems (ECS). Alt-

hough ECS are standard components of IBS,Hughes (1997) points out that according to the

present regulations they cannot replace paper

charts and therefore should not be used asprime aids to navigation. He also draws at-tention to deficiencies of existing ECS andlists requirements for the reliable operation of

such systems.

Observation and Detection. Apart fromX and S band radars, Hellstrm, Blount, Ot-tosson and Codega (1991) include image in-tensifiers and an infrared gyro-stabilised cam-era in the necessary observation equipment.

Transponders. Shipborn transponders(Automatic Identification Systems) may be

used to provide automatic data transfer be-tween ships and between ship and shore (e.g.VTS centre). It is especially important withfast craft that information relevant to naviga-tion is quickly and reliably transferred to allinterested parties without putting an additionalworkload on navigators.

Some benefits of using transponders are asfollows:

Ships can broadcast their intentions,thereby eliminating the danger of misin-terpretation (Crichton and Redfern, 1996).(See also Section 3.2.1.3).

In the areas where VTS radar coverage ishindered by the geography of the areaships may continually broadcast abouttheir position (Foxwell, 1995).

All traffic participants may share the sameradar picture, transferred to them from theVTS centre (Foxwell, 1995).

GPS corrections may be transferred toships (Foxwell, 1995).

When a ship is lost in an accident the re-cording of transmitted data may be treatedas a remote black box providing infor-mation for SAR operations and enablinglater analysis of events (Heikkil, 1996).

So far two systems of marine transpondershave been tested (Heikkil, 1996):

Digital Selective Calling (VHF DSC). Ship-Shore Ship-Ship transponder (4S).

Virtual Prototyping (VP). Virtual pro-totyping can model and visualise the processof building, testing and operating a prototype


8/41

8

system. Jons and Schaffer (1995) point outthat this technology might be applied duringthe development of fast ferries (and possiblyany HSMV).

For instrumentation VP might be used totest different bridge configurations in order toselect the best solution.

VP would be also an ideal tool during ini-tial crew training by providing the means totest bridge procedures before the actual ship isconstructed.

Vessel Traffic Services (VTS)

The IMO guidelines for VTS define it asfollows:

Any service implemented by a competentauthority, designed to improve safety and effi-ciency of traffic and the protection of the envi-ronment. It may range from the provision of

simple information messages to extensivemanagement within a port or waterway.

VTS cover areas of high marine traffic

density, quite often with additional naviga-tional difficulties (e.g. due to geographicalfeatures).

No extensive research on VTS withHSMV particularly in mind has been done;nevertheless VTS applicability in the areaswhere such vessels operate seems obvious.

According to the findings of the COST301 project carried out by the EEC countries,VTS should be designed to collect, process,

present to operators, disseminate to users,store and print out data of all types relating tothe marine traffic situation which it monitors(Degr, 1995). A broad description of manyaspects of VTS operation is provided by Bell(1990).

Moore (1993) draws attention to someproblems associated with VTS operation,namely communication with the ships crewand operation procedures.

Training of VTS Operators.

Many aspects of training are dealt with byBarber (1990). Simulators of VTS centresmay be used during training. These simula-tors might be linked to ship handling simula-tors (Distributive Interactive Simulation might

be used for this purpose - Jons and Schaffer,1995) in order to provide both VTS operatorsand navigators with suitable feed-back duringtraining and to smooth their co-operation.The need for a proper training may be illus-trated by the fact that, during one series ofsimulator tests, all situations leading eventu-ally to grounding were assessed too optimisti-cally by the trainees (Heikkil, 1996). Thiswould have produced disastrous consequences

in real life. The same source points out thatship performance may deteriorate with the

presence of VTS advice.

Mariners Reluctance

Many mariners see VTS as detracting fromtheir authority (Boisson, 1994, Zade, 1994).For many, VTS conflicts with the powerfultraditions of the freedom of usage of the seas(Degr, 1995).

Shore-Based Pilotage

Since it would be highly impractical for aHSMV to take a pilot on board in situationswhen his advice is desirable, the concept ofShore-Based Pilotage may be applied (Huthand Firnhaber, 1997). According to the defi-nition adopted by the International MaritimePilots Association Shore-Based Pilotage is anact of pilotage carried out in a designated area

by a pilot licensed for that area from a posi-tion other than on board the vessel concernedto conduct the safe navigation of that vessel.

Marine Traffic Simulation

Marine traffic simulation is a valuable toolused to investigate the marine risk, and hencesafety, levels in congested waters. Traffic israndomly generated at one or more gates andcaused to move along prescribed routes within

prescribed speed bands. Each vessel is as-sumed to be within a domain which moves

with it. If another vessel enters this domainan encounter is deemed to have taken placeand is logged. Encounter statistics may be


9/41

9

used to indicate levels of marine risk and theeffects of changes in operation on safety maythen be judged by carrying out what if sce-narios on the simulator.

Various operating rules are built into suchsimulators and many of these relate to thesafety of low speed manoeuvres in wind,waves and current. The link to ship hydrody-namics occurs at this point and increasing useis now made of manoeuvring simulation withtraffic simulation in marine safety studies.

Conclusions

The following conclusions are drawn withregard to human factors and HSMV safety:

There is a need to interchange informationbetween ships about their intentions.

Training of navigators and VTS operatorswith the use of simulators will becomemore common. Thanks to VirtualPrototyping, the training of navigatorsmay start before the actual ship is con-structed.

Connecting bridge simulators with VTScentre simulators would provide bothnavigators and VTS operators with realis-tic feedback and refine their co-operation.

An HSMV bridge should be manned withno less than two officers.

The training of HSMV navigators shouldnot be limited to navigation skills only butshould encompass team work, communi-cation skills, coping with difficult andstressful situations, optimum use of avail-able technical means etc. (these being theaim of Bridge Resource Management

training). The same applies to VTS op-erators.

Time-consuming activities of HSMVnavigators should be automated (or becomputer supported) in order to allow thenavigators more time for assessment of thesituation and decision-taking.

Electronic Chart Systems should be madesufficiently reliable to be approved as the

prime aids to navigation. Vessel Traffic Services will develop and

become more common for HSMV.

As a result of these conclusions, the fol-lowing is recommended:

A Navigation Safety Reporting Systemshould be set up in order to improve op-erational procedures, and identify areas

needing investigation, by studying inci-dents and accidents in which fast craftwere involved.

3.3Dynamic Stability

Introduction

In the report of the HSMV Committee tothe 21st ITTC, a substantial section was devot-ed to the dynamic stability of high speed craft.Some types of instability were identified andmodel test techniques to study them were out-lined.

The connection between dynamic stabilityand safety is so self-evident that it was be-lieved by this Committee that an analysis ofhigh speed marine accidents would soon re-veal dynamic instability as a major cause.

With the limited information that the Com-mittee has been able to obtain, it has becomeclear that this is not the case. It has further

become clear that only limited research hasbeen carried out on the topic worldwide sincethe last ITTC.

This suggests that either:

dynamic stability problems are not a majorcause of accidents, or

they are captured at the design and modeltest stage, or

they are eliminated by the provisions ofthe HSC Code (IMO 1996).

It is suspected that the last two of these as-sumptions are correct in that dynamic insta-

bilities can (and should) be identified duringthe design and testing process and the opera-tional and other limits imposed by IMOshould prevent the vessels operating in extre-me conditions.

The 21st ITTC High Speed Marine Vehicle

Committee identified some dynamic instabil-ities and listed them in its report (ITTC, 1996).


10/41

10

In addition, Dand (1995, 1998) listed the fol-lowing:

Loss of GM due to Wave System Course Keeping Bow Diving and Plough-In Porpoising Chine Tripping Take Off Spray Rail Effect of Critical Speed.

Dand (1998) discussed the physicalmeanings of these and they are listed here inaccordance with his definitions.

Loss of GM due to Wave System

If a displacement vessel moves fast enough,its wave system can be characterized by crestat bow and stern combined with a troughamidship. If the hull form is fine enough, theloss of buoyancy and waterplane area caused

by the trough can cause an apparent loss oftransverse GM. The vessel may then loll overto one side or other. In some cases this heelmay couple into yaw and the vessel begin toturn. The turn may induce further heel, whichinduces further yaw and so on. Taken to ex-

tremes, the vessel may suffer a catastrophicheel and turn, leading to capsize.

Many researchers have been studying thisinstability for both HSMV and conventionalships, Washio and Doi (1991) studied the dy-namical stability characteristics of HSMVrunning in calm water and in waves. Further-more, they proposed a hull form with im-

proved transverse instability developed fromthe experimental results.

Lewandowski (1998) showed that coupled

roll-yaw-sway dynamic stability of hard chinehulls in the planing regime developed atFnv>2 (Fnv: Volume Froude Number). Hefound that the traditional transverse stabilityinvolving the metacentric height GM or therighting arm GZ is meaningless, because thecenters of dynamic and hydrostatic force aredistinct.

Afremov and Smolina (1995) performedsystematic model tests and analytical studieson relevant hydrodynamic characteristics of a

ship series. They also provided a mathemati-cal model of a ships lateral motion in a sea-way. They provided a reliable prediction of

high-speed ship performance and found waysof ensuring stability and eliminating the pos-sible development of dangerous roll angles.Course Keeping

Poor dynamic stability about the verticalaxis gives rise at best to poor course-keepingand at worst to loss of control. Although

problematic in themselves, these tendenciesbecome more serious from a safety perspec-tive as speed increases. In extreme cases acalm water broach can occur if the yaw in-stability couples into heel.

Rutgersson (1998) proposed test proce-dures originally used to characterize the calmwater maneuvering performance of ships.

This study of stability problems in followingwaves for high speed monohulls will be car-ried out as a joint Finish and Swedish research

program. The test results will be used forvalidation of mathematical models.

Bow Diving and Plough In

Bow diving and plough-in at wave speedoccur when a vessel, moving into or with awave system, comes off one wave and

ploughs into the next. For a high speed pas-

senger ferry carrying vehicles, the need foradequate buoyancy forward (possibly a prob-lem for catamarans with fine fore-bodies) to-gether with adequate bow door arrangementsis essential if the vessel is not to be engulfed.

Porpoising

Porpoising is a well-known pitch instabil-ity which affects high speed planing craft. Itis now amenable to elimination by design, butwas the cause of catastrophic accidents tosome early high speed vessels.

Chine Tripping

Chine Tripping is experienced on planinghard chine monohulls when turning. The chi-ne may dig in and cause a powerful and sud-den heeling moment. In extreme cases thevessel could roll over at high speed. No refer-ences to work on this topic have been found.

Take Off

High speed catamarans may experiencelarge aerodynamic lift forces and pitch mo-ments at speed or in waves. This is caused by


11/41

11

air flow over and under the bridge deck which,in extreme cases, could cause the vessel to liftfrom the water and rotate in pitch. At presentsuch behavior is confined to high speed, light

weight, vessels. No references to work on thistopic have been found.

Spray Rail Engulfing

A high speed vessel receives a not incon-siderable amount of lift forward from thespray rails, which, on some designs, maydouble as fenders. Tank tests have shown that,once a certain speed is exceeded, these maycease to deflect the bow wave or spray and

become engulfed. When this happens the bow

may drop, to the accompaniment of largesheets of green water thrown into the air in theregion of the forebody. Speed may reduce atthe same time. In extreme cases bow divingmay occur. No references to work on thistopic have been found.

Effect of Critical Speed

Recent tank tests and full scale trials on ahigh speed passenger catamaran in shallowwater have identified some loss of directionalstability when moving at or near the criticalspeed. This speed is defined as

where h is the water depth. It is aboutthe speed at which solitary waves or solitonsare shed and hydraulic jumps created. Wheth-er or not solitons are shed, it is common forthe vessels own wave system to be charac-terized at such speeds by a high, and often

breaking, following wave, similar in form to ahydraulic jump. This is situated just astern of

the vessel and its upstream influence gives atendency to broach. This is perhaps analogousto the loss of directional stability experienced

by aircraft flying through the transonic speedrange, and has been felt not only by the pre-sent day vessels in shallow water, but also byhigh speed torpedo boats passing through anarrow, shallow, harbor entrance at speed inWorld War 2.

Dynamic Stability of Multihulls

Renilson and Anderson (1997) developeda mathematical model to predict the behaviourof a high speed catamaran in following seas,with particular reference to bow diving. The

surge force from the wave, and the vesselsheave and trim in the wave, were calculatedusing a quasi-steady assumption. These cal-culations, together with the validity of thequasi-steady assumption for this situation,were checked using a partly captive model. Inaddition, model tests were conducted to de-termine how the resistance, heave and trimvaried with speed in calm water. From theresults of the mathematical model, situationswhere bow diving would occur were identi-fied. The effect of the additional drag of the

submerged cross structure was included andthe resulting motions studied.

Dynamic Stability of Monohulls

Renilson and Tuite (1996) proposed afour-degree-of-freedom model, accounting forsurge sway, roll and yaw, which incorporatedthe effect of heel. They calculated the coeffi-cients required to simulate the behaviour of a

high-speed round-bilge planing hull in fol-lowing seas, including roll/yaw coupling.They concluded that the region of broachingof a fast planing vessel decreases as the ves-sel's metacentric height is increased.

Spyrou (1996a, b) analysed the dynamicstability of displacement ships in quarteringregular waves focusing on the specific condi-tions leading to broaching. He carried outsteady-state and transient analyses in the sys-tem's multidimensional state-space in order toidentify all existing limit sets and locate at-

tracting domains. Then he focused on the de-velopment of roll and the eventuality of cap-sizing and set out a multi-degree method ofglobal analysis based on transient maps. Hisstudy established a connection between speed,heading, automatic control parameters andcapsizing. In a more recent paper, Spyrou(1997) devised an elaborate nonlinear modelfor manoeuvring in waves to identify non-linear phenomena that govern large-amplitudehorizontal ship motions. His analysis un-veiled a sequence of phenomena leading to

cumulative broaching, which involves achange in the stability of the ordinary periodicmotion on the horizontal plane, a transition


12/41

12

towards sub-harmonic response and, ulti-mately, a sudden jump to resonance.

Lewandowski (1998) developed a methodto predict the dynamic roll stability of hardchine planing craft. Starting with the equa-tions of motion, an equation governing smallroll perturbations is developed. The roll re-storing moment acting on the hull is evaluated

by considering static and dynamic contribu-tions. The contribution of rudders and skegs,which is significant for this type of craft, isalso determined. Lewandowski(1997) alsodeveloped a method to evaluate the coupledroll-yaw-sway dynamic stability of hard chinehull in the planing regime. Expressions forthe linear stability derivatives are presented as

functions of geometry and loading, speed, trimangle, and wetted keel and chine lengths. Astability criterion is derived, and the effects oflength/beam ratio, loading, LCG position,deadrise, and appendages on stability are ex-amined. A simple method to check the trans-verse dynamic stability of a proposed design is

presented.

Katayama and Ikeda (1995, 1996c) inves-tigated experimentally transverse stability lossand the dynamic instability of pitch-excited

rolling in planing craft advancing in calm wa-ter. Using a database of measured three-component hydrodynamic forces they con-cluded that although the period of rolling in-duced by pitching is almost constant, roll be-comes unstable when the pitch is a multiple ofhalf the roll period.

Celano (1998) recently investigated theporpoising stability of planing craft experi-mentally. His tests included hulls with higherdeadrise angles, more typical of craft nowemployed for high-speed military purposes.

Two models of actual full-scale craft, com-plete with performance enhancing featuressuch as lifting strakes, trim tabs and variabledrive angle were tested. These additions werefound to have a profound effect upon condi-tions for the inception of porpoising. Estab-lished planing hull analysis methods wereaugmented with techniques developed duringthe course of the study to provide a basis fromwhich to design and outfit high-speed, heavilyladen planing hulls with respect to porpoisingstability. The longitudinal dynamic instability

in calm water, (porpoising), of a personal wa-ter craft was investigated experimentally at upto Fn=6.0 by Katayama and Ikeda (1996a,

1996b, 1997). The criteria of occurrence ofporpoising are predicted using a linear stabil-ity theory. Porpoising motion is also estimatedusing a non-linear time domain simulationmethod. Measured restoring, added mass anddamping coefficients are used in thesemethods and the predicted criteria and simu-lated motion are in good agreement with themeasurement.

To reveal the cause of porpoising instabil-ity, the measured forces are analysed. Theresults show that coupled heave and pitch re-storing coefficients have a different sign andare of the same order as the other coefficientsin the high speed range. This means that por-

poising is caused by a self-excited oscillation

due to the energy exchange between heaveand pitch motions.

Dynamic Stability of Wing-in-Ground Ef-fect Craft

With recent political changes in theEastern Bloc countries, WIGs, or ekranoplans,were revealed to the western world. Sincethen, the recasting of a rather old WIG con-

cept received much interest worldwide in thelast few years, this being reflected in a seriesof four international conferences, see Pro-ceedings (1995, 1996, 1997, 1998). Recently,the commercialisation of WIGs has been ex-

plored in a number of countries such as Russia,Australia, Taiwan, Germany, USA and Korea.Bogdanov (1995a) noted that legislation forekranoplans could fall under the jurisdictionof either or both of the International MaritimeOrganization (IMO) and the InternationalCivil Aviation Organization (ICAO), depend-ing on the mode. There seems to be littleopen literature available on the subject of thesafety of wing-in-ground-effect (WIG) craft,or ekranoplans. The vast majority of the pub-lished research is centred on the aerodynamicsof these craft, concentrating on the question ofthe increased aerodynamic efficiency whenoperating close to the "ground" or the surfaceof the sea. However, there are a number of

papers on the stability of WIGs, see Kumar(1972), Staufenbiel and Schlichting (1988),Gera (1995), Fuwa et al (1995), Riley (1995),Hall (1994), Rozhdestvensky (1996), Del-

haye(1997), Chang, Paik and Chun (1998).Chang et al (1998) derived the static and dy-namic longitudinal stability equations and the


13/41

13

stability of a 20 passenger WIG design, basedon a number of wind tunnel experiment data.A general conclusion from these references onthe stability is that due to the very proximity

to the free surface, a WIG can be inherentlyunstable if the aerodynamic centers are mis-matched. For stability, the vertical aerody-namic centre of a WIG should be ahead of itsaerodynamic centre in pitch.

Some publications address the question ofthe optimal shape of a WIG.Rozhdestvenskyand Savinov (1998) investigated the optimaldesign of 2-D wing sections in extremeground effect. Suzuki et al produced 2-D op-timal wing shapes with a high lift performance

and satisfactory longitudinal static stabilityusing a potential panel and SQP method. Kimand Chun (1998) developed a method to pro-duce optimal 2-D wing sections in a tail wingcombination with the design constraints whichsatisfy static stability and showed that an op-timal wing section could have a larger lift for-ce than that of the original wing.

3.4Seakeeping and Extreme Motions

Introduction

The dynamic behaviour of high speed craftin a seaway is directly related to its safety.Accordingly a separate section has been pre-

pared in which recent advances, both theoreti-cal and experimental, in the behaviour of fastcraft in a seaway are presented. This supple-ments the discussion of dynamic stabilitygiven in section 3.3 above.

Analytical Modelling of HSMV SeakeepingResponses

Although, during the last decade, much ef-fort has been directed toward developinganalytical tools for the reliable estimation ofHSMV motions, there is still a long way to goto devising an adequate standard. There aretwo major problems in this respect:

the appropriate modelling of all the factorsthat affect the dynamic behaviour of mod-

ern types of HSMV at the higher speedrange

their non-linear behaviour.Furthermore, lateral motion responses are

highly influenced by the strongly non-linearbehaviour of roll, while viscous effects shouldbe taken into account for its estimation.

The current trend in the analytical model-ling of HSMV seakeeping responses is thedevelopment of time-domain computer codes,which take into account some of these factorsanalytically, empirically or semi-empirically.Disregarding some of these factors should be

carried out with great care, however, but ad-vances in computer speed and power providea powerful tool for this approach.

Monohulls. De Kat and Paulling (1989)developed a time-domain numerical model

based on the impulse response function pro-cedure proposed by Cummins (1962) to de-termine the large amplitude motions of steeredvessels subjected to severe sea conditions. DeKat, et al (1994) used this model to investigateintact ship survivability of frigates in extremewaves and concluded various criteria shouldsupplement those currently in use; these arelisted in Section 3.7 below.

Payne (1990) developed a three-dimensional time-domain computer programfor planing hulls in random, head or followingseas. His hydrodynamic coefficients are

based upon a two-dimensional strip theory forstandard displacement vessels.

Lai (1994) developed a linearized vortex-

lattice method with special epsilon-modeltreatment for jet spray and presented an exten-sive study of three-dimensional planing hy-drodynamics.

Wang (1995) used the mixed Eulerian-Langragian scheme developed by Longuet-Higgins and Cokelet (1976) to simulate nu-merically fully non-linear, free-surface flowsin the time-domain. He used an EulerianBoundary Element Method, with a desingu-larized source distribution to handle the

boundary-value problem. Both this methodand the source-doublet panel method devel-oped by Markew (1991) have been proven to


14/41

14

be stable, efficient and robust time-steppingschemes for fully non-linear free surface

problems. In the sequel, a Lagrangian FiniteDifference Scheme was used to satisfy the dy-namic and kinematic free-surface conditions.The author compared the proposed numericalsimulation with experimental results for thecase of flare slamming and deck wetness. Fi-nally, he used the source-doublet method tostudy the problems of the free drop of a flared

body in calm water and of three-dimensionalplaning. Zhao and Faltinsen (1993) proposeda similar method to study water entry of two-dimensional bodies.

Ulstein and Faltinsen (1996) generalisedSedov's (1940) two-dimensional theory for the

unsteady problem of a flat plate that enterswater with constant fall velocity to cope alsowith a heaving planing plate. They proposed atime-domain theory under the assumptionsthat gravity can be neglected in the near fieldand the immersion of the flat plate is small.

Wu and Moan (1996) presented a linearand non-linear hydro-elastic analysis of theseakeeping responses of semi-displacementvessels using appropriate body boundary con-ditions of flexible modes. The total response

was decomposed into linear and non-linearparts. The linear part was evaluated by anextension of the high-speed strip theory fornon-planing ship hulls presented by Faltinsenand Zhao (1991). The non-linear part camefrom the convolution of the impulse responsefunctions of the linear ship-fluid system andthe non-linear hydrodynamic forces. It wasconcluded that the hydro-elastic effect inlinear extreme responses is insignificant andthat the non-linear influences are more signifi-cant the larger the Froude Number.

Bertram and Yasukawa (1996) overviewedthe currently available Rankine Source Meth-ods (RSM) and compared them with GreensFunction Methods (GFM). They concludedthat hybrid methods matching a RSM near-field solution to GFM far-field solutions areattractive for overcoming problems in the fre-quency domain for encounter frequencieslower than 0.25g /U (where g =9.81 m/sec2

and U =ship speed).

Jiang et al (1996), applied and extended

recent developments in dynamic system ana-lysis to the study of highly non-linear shiprolling motion and capsizing in random beam

seas. They also defined safe and unsafe areasin the phase plane of the unperturbed systemmodel to distinguish the qualitatively differentship motions of capsize and non-capsize. Thecorrelation between phase space flux and cap-size was investigated through extensivesimulation

Savander (1997) expanded the two-dimensional impact model of Vorus (1996) toa three-dimensional steady planing formula-tion. In both models, gravity was excludedfrom the free-surface boundary conditions. Inaddition, this model could be interfaced to aseakeeping simulation model.

Multihulls. Chan (1994) presented twomodels for the calculation of ship motions andwave loads acting on high-speed catamarans.He used both a three-dimensional oscillatingand a three-dimensional oscillating-pulsatingsource distribution method. Although bothmethods provide similar results, only the latter

predicts some experimentally-observed waveload phenomena.

Fang and Her (1994) described a time-domain method to model non-linear SWATH

ship motions in large longitudinal waves.They took into account both large non-linearmotions of the ship and non-linear viscousforce. The time simulation technique washandled by a fourth-order Runge Kuttamethod and the cubic spline method was usedto interpolate the corresponding hydrody-namic coefficients at each time step. Theanalytical results were experimentally verified.

Kvaeldvold (1994) investigated, analyti-cally and numerically, slamming against thewetdeck of a multihull vessel in head waves.He used a two-dimensional, asymptoticmethod valid for small local angles betweenthe undisturbed water surface and the wetdeckin the impact region. The disturbance of thewater surface and the local hydro-elastic ef-fects in the slamming area were accounted for.Interaction effects between local loading andglobal rigid ship motions were partially in-vestigated. He concluded that loads andstresses on the wetdeck are strongly influ-enced by the elasticity of its structure.

Van't Veer (1997) devised a three-dimensional Rankine panel method to calcu-late the steady and unsteady velocity potential


15/41

15

around a twin-hull vessel. He compared hismethod with experimental and numerical re-sults, using a two-dimensional strip theory, upto a Froude Number of 0.60. He concluded

that trim and sinkage should be included in themotion calculations if they are significant.

Kring et al (1997) used the non-linear,time-domain, three-dimensional RankinePanel method to simulate the seakeeping be-haviour of semi-displacement monohulls andcatamarans. They took into account the tran-som flow of semi-displacement hulls and theflow interaction effects between catamarandemihulls, as well as the non-linear effects offlare and the submergence of overhanging

structures, which can also be significant.Furthermore, their code can predict extremestatistics, and model passive and active ridecontrol systems.

SES. Ulstein (1995) studied numericallyand theoretically in the time domain the verti-cal plane motions of a SES in low sea states,focusing on non-linear air leakage underneaththe seals and the coupling between the flexiblestern bag and the air cushion pressure. He as-sumed a large forward speed compared to therelative vertical velocity between the bagstructure and the water surface. He then cal-culated in a simplified way the hydrodynamiccoefficients due to the side-hulls and, since theflexible bag behaves hydrodynamically as anunsteady 2-D planing surface, he solved the

problem by combining the solution for thatsurface with an integral equation for the wet-ted length of the bag. The latter is a generali-sation of what Wagner (1932) did for slam-ming. He concluded that the increase in theheight to length ratio of the flexible stern seal

bag reduces the vertical accelerations.

Sebastiani and Valderazzi (1997) formulat-ed a computational procedure for the time-domain simulation of a SES moving in waves.Among the main innovations of theirseakeeping procedure is a detailed treatmentof the non-linear dynamics of the air cushion,accounting for its interaction with the flexibleseals, developed on the basis of experiments.

Hydrofoils. Van Walree et al (1991) de-vised a design tool for hydrofoils encompass-ing a time-domain simulation of their power-

ing and seakeeping characteristics. The sys-tem was linked to a simulation model of theride control system. The authors announcedalso the initiation of an extensive experimen-

tal research program to determine the hydro-dynamic characteristics of the craft.

Ohtsubo and Kubota (1991) presented anew method for calculating vertical motionsand wave loads of large high-speed ships withhydrofoils. Ship motion was predicted by astrip method taking into account the effects ofthe non-linear hydrodynamic forces and dy-namic lift of the hydrofoils. They stressed theimportance of accurately predicting the dy-namic lift of submerged hydrofoils travelling

close to the free surface under waves. Ex-perimental verification of the method in thecase of a hydrofoil catamaran was provided.

Experimental Investigations

Monohulls. Suhrbier (1978) investigatedexperimentally, by means of captive and ra-dio-controlled tests, the roll stability of a

semi-displacement craft at high speeds. A re-duction of roll stiffness at speeds correspond-ing to Froude numbers above 0.6 to 0.8 wasreported. Spray rails can be used to overcome

problems of heeling at these speeds if themetacentric height cannot be increased. Testscarried out with two geosims did not produceany indication of scale effect problems on thestability characteristics. In a more recent pa-

per, Suhrbier (1995) investigated the effect ofpropeller cavitation on propeller/hull interac-tion and the dynamic stability of planing craft.He found out that loss of running trim due totrim flap ventilation and collapse of suctionforces can lead to dynamic instability and pos-sibly broaching. Thus trim flaps should besealed at their connection to the transom.

Fuwa et al (1982) used a radio-controlledmodel to reproduce broaching on a smallhigh-speed boat in a model basin. They thencarried out a large series of tests with captivemodels to determine the coefficients of theequations describing manoeuvring motion inwaves. A method of system identification wasused for this. The optimal form of the ma-neuvering equation was derived and the dy-namic balance of each component of the


16/41

16

equation was examined. To carry out theformer task they analyzed measured data byminimizing Akaike's Information Criteria (Sa-gara, 1981) to control the accuracy of the largenumber of parameters to be identified usinglimited data. Broaching always occurs when aship is under surf-riding conditions on thefront slope of a wave and the velocity U be-comes almost equal to that of the wave propa-gation VW. The authors concluded also thatthe most likely conditions are /L=2.0, head-ing of encounter=20 to 30 and U cosVW. Experimental results have been verified

by full-scale tests with satisfactory agreement.

Kan et al (1990) carried out extensive testsof a containership in quartering waves for a

set of encounter angles at speeds correspond-ing to Froude numbers up to 0.37. Most cap-sizing occurred at encounter angles between20o and 40o, to the lee side and at high speeds.The latter two conclusions were attributed tosurf-riding or asymmetric non-linear surgingmotion. They also examined numerically non-linear phenomena such as the fractal capsizing

boundaries in the initial value plane as well asthe control parameter plane.

Lundgren (1993) studied experimentally,

(by carrying out captive model tests), and nu-merically (by using a time-domain computercode), the operational limits for two high-speed monohulls operating in extreme seaconditions. He concluded that modelling ofgreen water on deck is crucial for the accuratecalculation of large rolling angles and thatcapsizing seemed to occur at wave heightscausing relative motion to exceed the free-

board midships, even if the metacentric heightis sufficient.

Francescutto et al (1994) provided experi-

mental evidence of the strong non-linear ef-fects that are present in the rolling motion of adestroyer in beam seas. A jump from low os-cillation to a higher resonant state was ob-tained by means of a shock while wave exci-tation was kept constant. The analysis of the

phase lag between excitation and rolling con-firmed that the jump was due to a bifurcation.These phenomena can be modelled by meansof a bifurcated solution of the non-linear roll-ing equation.

MacFarlane and Renilson (1995) carriedout tests on a model of a planing hull form inregular and irregular waves, to investigate the

limits of applicability of linear theory for pre-dicting vessel motions in irregular head seas.The experimental results were compared withthose calculated using a standard linear striptheory combined with linear superpositiontheory and with those obtained from linearsuperposition theory combined with the ex-

perimental results in regular waves. Theyprovided plots of the applicability areas forboth of the above cases. Grigoropoulos andLoukakis (1995) carried out similar tests forthree planing hull forms, concluding that stan-dard strip theories are inadequate to predictthe seakeeping behaviour of planing craft.

Keuning and Pinkster (1997) investigatedexperimentally and analytically variants of a

semi-planing fast patrol boat within their"Enlarged Ship Concept". The idea behindthis concept is to create variants of a basicdesign with increased length and with equal

payload, in order to improve calm and roughwater performance. The results of their con-cept were confirmed. They concluded that

peak values of the vertical accelerationsshould be used in the limiting criteria ratherthan significant values.

Grigoropoulos et al (1997) investigated

experimentally the seakeeping characteristicsof a double-chine parent hull form, (with widetransom, bottom warp and L/B ratio of 5.5), ofthe NTUA series of planing hulls, extrapolatedto a 108-metre car/passenger ferry and to a65-metre passenger ship. Experimental re-sults for the vertical seakeeping responses inhead sea states at Froude Numbers of 0.34 and0.68 were compared with analytical results

based on Salvesen, Tuck and Faltinsens(1970) strip theory for conventional ships andthe time-domain method for planing hullforms proposed by Payne (1990). While strip

theory results in an excessive overestimationof the vertical ship responses, Payne's method

provides reasonable results. In a more recentpaper, Grigoropoulos and Loukakis (1998)presented experimental results for three mem-bers of the NTUA series with L/B ratio rang-ing between 4 and 7 at the same speed inseven sea states. They concluded that verticalship responses are relatively small overall anddo not differ greatly when grouped with re-spect to Froude Number, despite the large dif-ferences in the parameters involved. In addi-

tion, they derived two rules of thumb applica-ble at pre-planing speeds and at the tested seastates: Pitch (RMS) [deg] 0.22 HS [m] and


17/41

17

C.G. acceleration (RMS) [g] 0.03 HS [m].Despite the approximate nature of these crite-ria, they can be useful in determining the op-erability of a particular ship in a given sea en-

vironment. Finally, since mean added resis-tance in waves seems to be relatively smalland relatively constant, a modest increase inresistance at the design speed of the order of15% is sufficient to account for the added re-sistance.

Summary

It is clear that the topic of fast craft be-

haviour in a seaway has received attention inrecent years. Its direct relation to safety hasnot been addressed in all cases, but some in-formation leading to safe design and operationhas been revealed.

There is more to be done, however, and itis apparent that more tests, at model and fullscale, should be carried out to investigate theeffect of GM and CG position on course sta-

bility and capsize. This, of course, impingeson the domain of the ITTC Stability Commit-

tee and indicates an area of overlap within theITTC organisation.

3.5Full Scale Tests

Types of Full Scale Tests

With the emphasis of the IMO Code ofSafety for High Speed Craft (the HSC Code(1996)) on the demonstration of a satisfactorylevel of operational safety on the craft itself,testing at full scale is required in a number ofkey areas.

The following is a selection of the maintypes of testing required by IMO:

manoeuvring and controllability, with par-ticular emphasis on the effects of systemfailures (Chapter 17 and Annex 8 of theCode)

conformity with acceptable accelerationlimits (Annex 3)

evacuation trial (paragraph 4.8) structural loading trials (paragraph 3.6) inclining experiments (paragraph 2.7) seat tests (Annex 9)

An overview of the approach to safetyadopted in the HSC Code has been given byWerenskiold (1997). In this paper he dis-cussed the parts played in high speed craftsafety by the International Safety Manage-ment (ISM) Code (which became mandatoryfor most types of ships in July 1998) andFormal Safety Assessment (FSA). The formeris a formal framework for the safe manage-ment and operation of ships by setting rulesfor the organisation of company managementin relation to safety and pollution prevention,together with the implementation of a com-

pany safety system. The FSA is another for-mal framework which identifies the hazards towhich a vessel is exposed, the risks which thisimplies and the means to control them.

Werenskiold shows how these conceptsand others may be combined with the HSCCode to provide a means for total safety as-sessment. Central to this is the need to de-monstrate that certain safety criteria have beenmet; the way this is achieved is, in many cases,

by means of the full scale tests mentioned

above. Those of most interest to members ofITTC are now considered.


18/41

18

Evacuation Trials

Marine evacuation systems (MES) areused on passenger-carrying high speed craft.

In many cases they consist of an inflatableslide with some form of floater at its foot tocollect the evacuated passengers; a number ofthese are discussed in Safety at Sea Interna-tional (January 1997). Regardless of their in-novative qualities, they must be able to de-monstrate an ability to deploy safely in the

prevailing environmental conditions. Tanktesting could supplement full scale trials forthis, and indeed has been used for this purposewith some offshore escape systems. Whiletank tests can be used to demonstrate the ef-fects of bad weather on the MES, the Code

only requires the full scale demonstration tobe carried out in calm conditions within a har-bour.

Werenskiold (1996) discusses the applica-tion of performance requirements to assessand quantify the safety and reliability of thewhole evacuation process of high speed craft.In his opinion, the safety margins proposed,and demonstrated in recent HSMV accidents,are far too small.

Sea Trials

Of direct interest to ITTC members are thesea trials required by the HSC Code to dem-onstrate, among other things, safe manoeu-vring and control together with compliancewith safe acceleration limits.

Dogliani, Capizzi and Lauro (1997) dis-cussed this whole area and, by an analysis of

the HSC Code, designed new types of test tosupplement the more usual manoeuvring trialsas described, for example, in the ITTC Codesof Practice. These arose from the followingrequirements:

measurement of acceleration in the hori-zontal plane to verify compliance withcriteria in Annex 3 of the code duringstopping manoeuvres, slam starts and highspeed turns

measurement of cruise performance duringnormal operation (low sea states) and wor-st intended condition (high sea states).The operating levels have to be established

by full scale tests in different sea condi-tions with various headings

demonstration of the effects of failures ormalfunctions according to a Failure Modeand Effect Analysis (FMEA) as specifiedin Annex 4 of the Code.

These highlighted the need to upgrade andupdate standard manoeuvring trials codes in-dicating that additional numbers and types oftrial were necessary. The authors split the re-quired full scale trials into four groups:

a) Speed/Power TrialsConventional speed/power trials are unaf-

fected by the new requirements, but the de-termination of a safe maximum speed with

one or more propulsion units disabled is nownecessary.

b) Manoeuvrability TrialsAn increased interest in emergency stops

and emergency manoeuvres, with their highattendant acceleration levels, has led to theneed to ensure that acceleration levels remainwithin the limits set in Annex 3 of the Code.

The remaining turning, zig-zag and stop-ping trials required by the IMO Manoeu-

vrability Standards (1993) are unaffected, alt-hough good instrumentation and control isrequired if meaningful zig-zag tests are to becarried out at high speed. An alternative

practical means to indicate directional stabilityis the pull-out test. Although more compre-hensive information on directional stability isobtained from spiral tests, it may be remarkedin passing that the usual drawbacks of spiraltests long timescales and large amounts ofsea-room would be less likely to apply tohigh speed vessels.

c) Seakeeping TrialsFull scale trials specifically to evaluate

seakeeping have been a requirement for navalvessels in the past, but have not usually beencommon for merchant vessels. This situationis now changed, with commercial high speedvessels being required to satisfy (and demon-strate compliance with) the acceleration limitsof Annex 3 of the Code. It is necessary todetermine:

speed/power relations in various seastates,and at various headings

speed loss at a constant engine setting


19/41

19

whether motions and accelerations inrough seas satisfy Annex 3.

Dogliani et al propose a test in a given

seastate whereby a pentagon-shaped course isfollowed, a wave-measuring buoy being

placed at the centre of the pentagon (Figure 4).This provides responses from seas on the head,

beam and quarters with additional measure-ments being taken in beam seas to assess theeffectiveness of any ride control systems. It isrecommended by the authors that the vesselshould stay on each side of the pentagon for20 minutes (with at least 10 minutes of dataacquisition) before changing heading. Wavedata for the complete trial are collected from

the central wave buoy.

Figure 4. Pentagonal Track for seakeepingTrial.

Similar tests are proposed in the NATO

Seakeeping Trials Procedure (1994) whileMarintek (Holden, 1991) gives an alternativeTrials Procedure for High Speed Craft.

In the former, the pentagonal course ofDogliani et al is extended to an 8-sided coursewhich resembles a straight-sided crescent. Inthe latter a comprehensive set of trials proce-dures is given, covering calm water trials ma-noeuvring trials, rough water trials and analy-sis/reporting procedures.

d) FMEA TestThe purpose of these tests is to determine

the safe limits of operation taking account ofthe possible effects of equipment failure.

They can be carried out during the manoeu-vring trials by, for example, simulating thesudden failure of the propulsion line, a mal-function in appendage retraction or loss of lift

for an ACV. The resultant change in behav-iour is measured.

Directional Stability Tests

A simple test, which can be carried atmodel and full scale, was developed by Wer-enskiold (1993, 1995). It is discussed at somelength in the report of the HSMV Committeeof the 21st ITTC and consists of running amodel or vessel, at an initial angle of heel, in a

straight line. If the heel increased beyond agiven limit at speed, then the vessel is likelyto be directionally unstable. The detailed cri-teria are given in section 3.7 below.

Wash Measurements

In Section 3.6 of this report, the problemsposed by the wash of HSMV are mentioned.Full scale measurements of wash, while not at

present compulsory, have been carried out by

a number of investigators as already men-tioned.

Tethered or floating wave height measur-ing devices have not been generally successfulfor this purpose, and, if possible, wave probesshould be attached to a suitable fixed mount-ing. Care should be taken to measure thewave probe position in relation to the track ofthe passing vessel; this is usually done usingDGPS for both vessel track and probe position.It is important to indicate the time at whichsome part of the vessel passes the wave probe

and its passing speed. Ideally such testsshould be carried out at times of slack water,

but if this cannot be arranged, the mean waterlevel should be recorded against time. Finally,some idea of the bathymetry local to the testarea should be obtained.

Summary

Safety requirements for high speed craft

mean that more comprehensive full scale testsand measurements are now required. This


20/41

20

means that existing Codes of Practice for trialsat sea must be revisited and revised.

3.6Wash

The Problem

Any vessel moving at high speed on thewater surface must inevitably cause a distur-

bance. This will hold true whether the vesselis of the displacement, semi-displacement or

planing type and whether it be hull-borne,foil-borne or supported on a cushion of air.Some of these will cause less disturbance thanothers, but all will transfer some energy to thewater in which they move and various formsof disturbance will result.

By far the most obvious is the wave washcreated by all vessels moving on the watersurface. The advent of fast passenger-andfreight-carrying vessels has brought with it anincrease in wash nuisance around the world; inan increasing number of areas this has had animpact on marine safety. This arises when fastvessels pass near to shoaling waters or coastalareas where their wave system comes ashore.

There has been an increase in incidents inwhich long-period waves of large amplitudehave come ashore after the passage of a fastvessel on an otherwise calm day. This hascaused people on beaches to be knocked down,

people in small boats to be inconvenienced, or,in extreme cases, capsized, and, in areaswhere sea walls abound, has given rise to sig-nificant wave reflections and interference.Ships loading or off-loading in harbours have

been caused to move excessively (either bythe direct effect of the wave wash or by the

seiche which they trigger) and large vessels inapproach channels can be induced into a no-ticeable roll motion by their passage (see, forexample, Seaways 1998)

All of these manifestations have implica-tions for safety and attempts are being made

both to understand the causes of the phenome-non and to minimise its effect by design andoperation.

The Cause

The normal free-wave system of low speedsurface vessel is well-known; it comprises di-verging and transverse wave systems. Inshallow water, the following changes occur inthis familiar wave system as speed increases:

the diverging waves may leave the vesselat ever-increasing angles as the criticalspeed, at a Froude Depth Number of unity,is reached (Havelock 1908)

the waves increase in height and steepen solitary waves (solitons) may be formed

near the critical speed at super-critical speeds, the transverse

wave system disappears and the divergingsystem remains

near the critical speed a large followingwave may arise, similar in nature to a hy-draulic jump. In some circumstances thiscan affect handling due to a reduction indirectional stability.

A further effect can occur if a vessel isturning, when its wash on the inside of theturn can become steeper due to a wave fo-cussing effect.

Wash approaching a beach will alsosteepen, and possibly break, causing, at best,

inconvenience to those on the beach, and atworst, danger to the same people by washingthem off their feet.

Although the largest waves are likely tooccur when vessels operate in shallow waterconditions, it should be noted that waves fromhigh speed vessels in deep water can radiateinto shallower water near the coast and stillcause wash nuisance. Perhaps the most in-sidious form of this occurs when large long-

period waves ride up a beach unexpectedly onan otherwise calm day. These may have been

caused out at sea by a fast vessel which willhave long-since disappeared by the time itswash reaches the shore. If the waves reachland where there is no beach, but a rocky sho-re or seawall, they may reflect, adding to thewave activity just offshore.

Studies of this phenomenon have beenconfined in the main to vessels whose abso-lute speed is high and whose Froude Depth

Number is around the critical value. However,some investigators have considered lower

speed vessels in inland waters whose wash notonly causes a nuisance and possible danger,but also erodes the banks of the waterways.


21/41

21

A comprehensive study by the DanishMaritime Authority (DMA) (1998) deals withthe former while that of Gadd (1994) deals

with both the former and latter with regard topleasure craft. The DMA report notes thatwaves from large, fast, ferries are character-ised by long periods of around 8 to 10 secondswhich may be contrasted to the 4 to 5 second

periods of conventional ferries. They alsoprovide a plot (Figure 5) which indicates azone of speeds near the critical value withinwhich vessels should not operate if they are toavoid excessive wash. It is of interest to notethat 40 knot vessels operating in the shallowopen waters of some seas (the southern North

Sea for example) are operating within thiszone.

Figure 5. Critical Speed Zone for Exces-sive Wash.

Gadd (1994) demonstrated a compara-tively simple method to compute the wavesystem of a fast planing vessel. He represent-ed the planing hull by a surface pressure dis-tribution whose length and breadth are such

that their product equals the waterplane areaof the boat and whose depression, created atrest, equalled the immersed volume of the

boat. The resulting wave system (Figure 6)gave remarkably good predictions of meas-urements made for a variety of pleasure craft(Figure 7).

Figure 6. . Predicted High Speed Craft WaveSystem.

Figure 7. Wash Prediction: Comparison of

Full Scale and Computer Results.

Other measurements of wave wash havebeen made by Renilson and Lenz (1989) whocarried out model and full scale experimentsto investigate the effect of hull shape on washheight. This was followed (Renilson et al,1991a, 1991b) by the development of a lowwash ferry for service on an Australian River.This demonstrated the value of wave resis-tance calculations in estimating the impor-tance of the wash generated.

In Europe the SPAN research project hasinvestigated the behaviour of fast catamarans


22/41

22

in very shallow water. Wave wash measure-ments were made with towed models in a tankand were observed with a free-running modelin open shallow water. Solitons were ob-served in the towing tank around the criticalspeed (Figure 8); they were created at themodel and ran ahead of it down the tank. In-terestingly, they were also observed in theopen water tests when they were shed aheadof the model, their wave fronts forming a cir-cular wave of appreciable energy. Onceformed, these moved within the basin inde-

pendently of the models wave system andreflected off the walls. Also observed in thesetests (and repeated at full scale) was the largefollowing wave at speeds near critical whichhad similarities to a hydraulic jump.

Possible Solutions

Solutions to the wash problem of highspeed craft have been partially successful. Ithas been found impossible at present to pre-dict accurately the far-field wash climate fromnear-field measurements taking account of

water depth, beach and coastline and the routeof the craft. This appears to be a non-stationary Boussinesq (soliton) wave problemwhich has so far resisted solution.

There is therefore a need, in the first in-stance, to determine a reliable decay law forthe far-field wash of high speed vessels.

Solutions to the wash problem are notcomprehensive, but some attempts have beenmade. These fall into the following broadcategories:

speed and route restrictions hull design remedial measures on shore, including

warnings.

These are now considered by resortingwhere possible to published information, alt-hough at present this is understandably sparse.

However, before going so, it is importantto consider by what criteria the adequacy ofany remedial measures may be judged.

Wash Criteria

Various authorities are contemplating washcriteria. Usually they are of two main types:

wash height restrictions comparison with already-accepted wash.

Wash height restrictions have been pro-posed by the Danish Maritime Authority(1998) who set the acceptable height of a long

Figure 8. Measured Wash of High speed Catamaran Showing Solitons


23/41

23

period wash wave at 0.35 metres in 3 metreswater depth. This has been suggested as alimit for the time being so that future highspeed ferry routes can be assessed for both

safety of navigation, leisure activities on thecoast and protection of coastal structures. Asimilar criterion was suggested for a highspeed catamaran operation on the RiverThames.

If an absolute criterion is not set, the washgenerated by high speed vessels is often com-

pared with that generated by conventional fer-ries in the same waters. This is done on theassumption that conventional ferry wash is at

best accepted, and at worst tolerated, so that

its use as a criterion does not increase marinerisk.

Speed and Route Restrictions

Speed and route restrictions are made inthe belief that, by changing speed (often byreducing it) and changing the route (often bymoving it further offshore), wash problemswill be ameliorated.

Unfortunately this is often complicated bythe fact that reducing speed may move thevessel from one wavemaking regime to onethat is worse, while the persistence of some ofthe low frequency shallow water waves (ordivergent wave groups) makes route selectiondifficult.

Speed restrictions to minimise wash havebeen used for fast ferry operation on the RiverThames, in Hong Kong Harbour, and in theapproaches to Harwich in the North Sea. For

low speed leisure craft speed restrictions arealso imposed on many recreational waterwaysfor both safety reasons and environmental

protection (Motor Boat and Yachting, 1997).

In many areas the problems associatedwith speeds near the critical Froude Depth orLength Numbers has been recognised andspeeds are regulated accordingly. As alreadymentioned, Figure 5 shows the restrictedspeed/depth relationship suggested by theDanish Maritime Authority (1998) so that ves-

sels avoid the critical speed range. In manycases high speed vessels are able to pass rap-idly through this range because their opera-

tional speeds are past the critical speed or themain Froude Number hump. Service speedswhich lie within the range are to be avoided.

Speed restrictions are also imposed onhigh speed vessels using fairways occupied bylarge displacement vessels constrained bytheir draught. This is to prevent unnecessaryloss of underkeel clearance from roll andheave motion induced by the wash of the highspeed vessel.

Although speed restrictions can in somecases be beneficial, there are understandablystrong commercial pressures against theirwidespread adoption. Clearly their existence

works directly against the whole purpose ofhigh speed marine transport; too many or toosevere speed restrictions tend to eliminate anyadvantages gained.

Hull Design

As it is the hull of the vessel which pro-duces the waves, it is natural to assume that,

by changing its design in some way, wave

wash can be reduced at source.

For low speed pleasure boats, and stimu-lated by full scale measurements (May andWaters 1986), attempts were made to usevarious well-known devices to eliminate orreduce hull-generated waves. Firth (1991)and Ship and Boat (1997) described develop-ments carried out at BMT using bulbous bowsand wave shelves to cancel waves. This ap-

proach met with some success and was laterextended by the work of Gadd (1997) inwhich the bulbous hull form was combined

with a tunnel stern incorporating a large,slow-speed high-efficiency propeller. Thisenvironmentally-friendly design exhibited re-duced wash characteristics by comparisonwith conventional cabin cruisers of similardisplacement.

The benefits of cylindrical bows on canalor river narrow boats were demonstrated byMcGregor and Ferguson (1993) when signifi-cant reductions in wash were produced com-

pared to boats of conventional form.

For faster vessels Doctors (1997) andDoctors et al (1991) have considered ways of


24/41

24

reducing the wave resistance and wash of rivercatamarans by hull shape changes. Althoughsome of this work is purely theoretical, a lowwash design was produced for use in Australia(Janes High Speed Marine Craft 1996). Thisexploited, among other things, the significanteffect of displacement-length ratio on washand resistance; the lower is this ratio, the low-er the resistance and the lower the wash height.

This finding was confirmed in the SPANstudy for high speed in shallow water. Gener-al design changes were made to the demi-hullof a high speed catamaran, but displacement-length ratio remained constant. The changeshad small effects on wash and resistance atsupercritical speeds; reducing the displace-

ment-length ratio, on the other hand, had anoticeable effect.

Remedial Measures on Shore

While attempts are made to reduce ship-generated waves at source, other remedialmeasures to reduce their impact ashore havealso been developed. An indication of highspeed vessel wash on the safety of those on

shore is given by the warning notices placedon vulnerable beaches. These warn people tobeware of the wash of fast ferries either pass-ing or slowing down nearby.

Protection of river banks vulnerable towash erosion has been carried out by variousmeans of which the reed bed has proved to bea useful wave damping device. Shoaling

beaches usually cause the waves to break, butif solitons are created they may not be elimi-nated by this means and pass up, or evenacross, the beach. In some cases more perma-nent beach protection is being developed.

Summary

There is little doubt that the wash generat-ed by high speed vessels causes a possiblesafety problem world-wide. This is now gen-erally recognised and remedial measures aresought.

The role of ITTC members in providingunderstanding of the phenomenon and devel-oping the means to minimise its effect by

design and operation will be significant in thefuture, thereby ensuring that marine safetylevels are not compromised.

3.7Safety CriteriaIn this section a number of Rules of

Thumb are gathered together which shouldallow ITTC members to assess whether a par-ticular design will be dynamically stable andsafe. They have been collected from the openliterature and are grouped into two main areas: design guidelines model test criteria.

It is hoped that they will help to answersome of the questions relating to the appropri-ate magnitude of a given dynamic quantityrather than just its presence or absence. Inother words, it is hoped that these will indicatehow much dynamic stability (say) is needed,rather than simply stating that the craftshould be dynamically stable.

Design Guidelines

Monohulls: The region of broaching of afast planing vessel decreases as the vesselsmetacentric height is increased (Renilson andTuite, 1996).

For a fast planing craft, although the peri-od of rolling induced by pitch is almost con-stant, roll becomes unstable when the pitch

period is a multiple of half the roll period(Katayama and Ideka, 1995, 1996c).

To avoid dynamic roll instability in plan-ing hulls (Petersen and Werenskiold, 1997):

Increase L/B ratio Avoid trim at rest and keep running trim

above 3o at Froude number Fn > 1.5 Keep warp abaft amidships below 15o Use entrance angles less than 30o and Raise the vertical position of the rudders.

For frigate-like vessels operating in severesea conditions, the following design guideli-nes are suggested (De Kat et al, 1994):


25/41

25

The righting lever in calm water shouldremain positive up to a heeling angle of atleast 90o

The minimum area under the rightinglever curve for unlimited operation in thedesign condition must be between 1.00 m-rad (for a vertical prismatic coefficient CVP= 0.55) and 0.67 m-rad (for CVP = 0.70).Linear interpolation should be used forintermediate CVP values

The minimum dynamic stability between30o and 40o for unlimited operation in thedesign condition must be between 0.13 m-rad (for CVP = 0.55) and 0.10 m-rad (forCVP = 0.70). Linear interpolation should

be used for intermediate CVP values.

Trim flaps should be sealed where theyjoin the transom to prevent directional insta-bility and broaching (Sububier, 1995).

Yaw/heel coupling can be reduced bymeans of spray rails, if metacentric heightcannot be increased, to reduce directional in-stability in planing or semi-displacement craft(Sububier, 1978).

Suitable positioning of the lcg and centreof lateral resistance can provide good handlingqualities (Dand and Cripps, 1995). For a highspeed rescue boat a range of from +3% to -5%was recommended where a positive quantityrefers to the point of application of the linearmotion sway force being ahead of the lcg.

The most likely conditions for broachingin a following sea are (Sagara, 1981):

/L=2.0 encounter heading =20 to 30 VcosVw where V is craft speed and Vw

wave speed.

With fast monohulls in high seas, capsiz-ing occurs at wave heights causing relativemotion to exceed the freeboard amidships,even if the metacentric height is sufficient(Lungren, 1993).

For operability estimates of high speedmonohulls at pre-planing speeds in waves, thefollowing approximations may be used (Grig-oropoulos and Loukakis, 1998):

RMS pitch ()0.22 Hs (m) RMS acceleration at cg(g)0.03 Hs (m)

added resistance 15%.SES: An increase in the height-to-length

ratio of the flexible stern sealing bag/skirt re-

duces vertical accelerations in low sea states(Ulstein, 1995).

WIG Vehicles: Due to the proximity ofthe ground plane, a WIG can be inherentlyunstable if the aerodynamic centres are mis-matched (see Section 3.3 for references)

Multihulls: In calm shallow water at highspeeds, the directional stability of a modelcatamaran was found to be improved at super-critical speeds and degraded in the trans-

critical region (SPAN study 1999).

Model Test Criteria

Calm Water Directional Stability (see21st ITTC HSMV committee report).

Tow

safety of hsmv

Documents