Ship Manoeuvring and Seakeeping PDF

LectureNotes
SHIP MANOEUVRING AND

SEAKEEPING
ByProf.Dr.Ing.ZouZaojian
AprilJune 2006
SchoolofNavalArchitecture,OceanandCivilEngineering
SHANGHAIJIAOTONGUNIVERSITY
UndergraduateCourse
SHIP MANOEUVRING AND
SEAKEEPING
(26lecturehours,1.5credits)
Instructor ProfessorZouZaojian
Contact Room113, TowingTank
XuHuiCampus
Telephone:62933645
Email:zjzou@sjtu.edu.cn
CourseWebSite http://naoce.sjtu.edu.cn/pna
SchoolofNavalArchitecture,OceanandCivilEngineering
SHANGHAIJIAOTONGUNIVERSITY
I
TableofContents
CHAPTER1:INTRODUCTION1
1.1ContentsoftheCourseandItsSignificance1
1.2RelationshipsofShipManoeuvringMotionandMotioninWaves1
CHAPTER2:BRIEFINTRODUCTIONTOSHIPMANOEUVRABILITY 1
2.1Definitionof ShipManoeuvrabilityandItsContents1
2.2Significanceof ShipManoeuvrability 2
CHAPTER3:EVALUATIONOFSHIP MANOEUVRABILITY 4
3.1BriefIntroduction4
3.2StandardManoeuvresandtheParametersEvaluatingShipManoeuvrability4
3.3IMOStandardsforShipManoeuvrability 7
3.4FullscaleTrials8
CHAPTER4:PREDICTIONOFSHIPMANOEUVRABILITY 8
4.2FreerunningModel Tests8
4.3Equationsof ShipManoeuvringMotion(MathematicalModels)9
4.4MethodsforDeterminingtheHydrodynamicForcesActingonaManoeuvring
Ship13
4.5PredictionBasedonMathematicalModelsandComputerSimulation16
CHAPTER5:LINEAREQUATIONSOFSHIPMANOEUVRINGMOTION16
5.2LinearHydrodynamicDerivatives17
5.3AnalysisofStability (inherentdynamicstability)20
5.4NomotoModelandK,TIndexes23
5.5AnalysisofTurningAbilityBasedontheLinearEquations29
CHAPTER6:CONTROLDEVICES31
6.1KindsofControlDevices31
6.2RudderandItsDesign33
CHAPTER7:MEASURESTOIMPROVESHIP MANOEUVRABILITY 36
CHAPTER8:BRIEFINTRODUCTIONTOSEAKEEPING 37
8.1Definitionof SeakeepingandItsContents37
8.2ShipMotionsinSixDegreesofFreedom 38
CHAPTER9:OCEANWAVESANDSTATISTICANALYSIS38
II
9.2RegularWaves38
9.3BasicTheoryofIrregularWaves40
9.4TheoreticalBasicofSpectrumAnalysis 40
9.5ResponseRelationshipsofLinearSystem42
CHAPTER10:ROLLMOTION43
10.1LinearTheoryof RollMotion44
10.2Roll Test50
10.3DeterminingtheHydrodynamicCoefficientsof RollMotion50
10.4Estimationof Roll Amplitude51
10.5RollControl Devices52
CHAPTER11:PITCHANDHEAVEMOTIONS53
11.1TheoreticalComputationforSeakeepingProblems53
11.2PitchandHeaveinIrregularWaves56
11.3DerivedResponsesduetoShipMotion56
11.4ShipMotioninObliqueWaves57
CHAPTER 12: DESIGN OF SEAKEEPING PERFORMANCES AND FULL
SCALETRIALS58
12.1SeakeepingPerformanceIndex58
12.2EffectsofShipPrincipal DimensionsandFormonSeakeepingPerformances 59
12.3FullscaleTrialsofSeakeeping59
REFERENCES 60
SHIPMANOEUVRINGANDSEAKEEPING
1
CHAPTER1:INTRODUCTION0.5hour
Demand:Knowthebasicconceptsofkinematicperformancesofshipsincalmwaterandinwavesknow
thesimilaritiesanddifferencesofshipmanoeuvring
1
motionincalmwaterandshiposcillatingmotion
inwaves,aswellasthe immanentrelationshipsbetweenthetwokindsof shipmotion.
1.1ContentsoftheCourseandItsSignificance
This course is composed of two parts: ship manoeuvring and seakeeping performances. Manoeuvring
and seakeeping performances are important hydrodynamic performances of ships. The course introduces
thebasicconceptsandbasictheoriesofship manoeuvrabilityandseakeeping performance,aswellasthe
correspondingcomputation methodsandthe methodsof fullscaletrialsand modelteststoevaluatethem.
Themaintaskofthecourseisbymeansofalltheteachingphasestoenablethestudentstomasterthebasic
conceptsandbasictheories,and thecorrespondingcomputationalandexperimentalmethods.Theemphasis
islaidontrainingtheability,especiallytheabilityofinnovationandpracticetoapplytherelevanttheories
andmethodstoanalyzeandsolvetheshipmanoeuvringandseakeepingperformancesrelatedproblemsin
shipdesign,toensurethattheshipisdesignedwithpropermanoeuvringandseakeepingperformances.
1.2RelationshipsofShipManoeuvringMotionandMotioninWaves
Thiscourse dealswithshipmanoeuvrabilityandseakeeping performance,the unsteadykinematicand
dynamicperformancesofship.Bothperformancesarerelatedtoshipunsteadymotions.Traditionally,ship
manoeuvrability deals with the ship manoeuvring motion in the horizontal plane in calm water, whereas
seakeepingdealswiththeshiposcillatingmotionsinsixdegreesoffreedominoceanwaves.Actually,there
is an immanent relationship between the two kinds of motions. Especially for a seagoing ship, it will
usuallytravelinwavesandneedtokeepitscourseorchangeitscourseinwaves.Ontheotherhand,aship
undergoing manoeuvring motion in waves will undergo oscillating motions in six degrees of freedom
simultaneously.Therefore,itisimportanttoinvestigatetheshipmanoeuvring andseakeepingperformances
conjunctly.
CHAPTER2:BRIEFINTRODUCTION TOSHIPMANOEUVRABILITY0.5hour
Demand: Know the basic concepts of ship manoeuvrability know the relation between manoeuvrability
andnavigationsafetyandeconomy.
2.1Definitionof ShipManoeuvrabilityandItsContents
Ship manoeuvrability is the ability of a ship to keep or change its state of motion under the control
actions,i.e.,tokeepthe straightaheadcoursewithconstantspeed,orto change thespeed,thecourse and/or
thepositionoftheship,accordingtothe intentionofthehelmsman.
Shipmanoeuvrabilityincludesthefollowingcontents:
(1) Inherent dynamic stability, also called straight line stability: A ship is dynamically stable on a
straightcourseifitcan,afterasmalldisturbance,soonsettleonanewstraightcoursewithoutanycontrol
actions, see Figure 2.1. The resultant deviation from the original course will depend on the degree of
inherentstabilityoftheshipandonthemagnitudeanddurationofthedisturbance.AsshowninFigure2.1,
foradynamicallyunstableship,itwillultimately enterinto an arbitrary unsteadyturningmotion.
1
ItismanoeuvringinBritishEnglishandmaneuveringinAmericanEnglish.
2
finalcourse
(2)Coursekeepingability,also calleddirectionalstability: Thecoursekeepingabilityistheabilityof
thesteeredshiptomaintainitsoriginalcoursedirection,seeFigure2.2.Ashipwhichhasinherentdynamic
stability can only maintain its original course direction under the control action. Also a ship which is
dynamically unstable on straight course can maintain its original course directionby frequently applying
thecontrolaction.
(3)Initialturning/coursechangingability:Theinitialturningabilityandcoursechangingabilityare the
abilityofshiptochangeitsheadingasresponsetoacontrolaction.A shipwithgoodinitialturningability
and/or coursechanging ability will quicklyget intoturning or change its original course after the control
action.
(4)Yaw checking ability: The yawcheckingability is the ability of the steered ship torespondto the
counterrudderactionappliedinacertainstateofturning.
(5) Turningability:Theturningabilityistheabilityof shiptoturnunderthehardoverrudderaction.
(6)Stoppingability:Thestoppingabilityistheabilityofshiptostopwithenginestopped(inertiastop)
orenginefullastern (crashstop)afterasteady approach atfullspeed.
2.2Significanceof ShipManoeuvrability
Ship manoeuvrability is directly related to navigation safety and economy. For a ship manoeuvring
undersevereenvironmentalconditionsand/orinrestrictedwater,marinedisastersmayoccuriftheshiphas
not adequate manoeuvrability. For example, for a ship with poor initial turning ability or poor turning
ability, collision with obstacles in the waterway or with the bank of the narrow waterway may not be
y
x
smalldisturbance
Figure2.2 Coursekeepingability
originalcourse
unstable ship
stable ship
y
finalcourse
x
smalldisturbance
Figure2.1 Inherentdynamicstability
originalcourse
3
avoided,asshowninFigure 2.3.
On the other hand, since a dynamically unstable ship or a ship with poor dynamic stability can only
keep its course by frequently repeated use of the control device, for such a ship not only the voyage is
longer than the planned one, as shown in Figure 2.4, but also more energy is needed to consume by the
controldevicemachine.
Actually,ship manoeuvrabilityitselfisnotthe only factorwhichaffectsasafenavigation.Othertwo
factors,theenvironmentalconditionssuchaswind, wavesandcurrent,andthe humanfactor,contributea
lottonavigationsafetybyinfluencingtheshipmanoeuvringindifferentmanner.Thoughthehumanfactor
isthemostimportantfactorwhichaffectsnavigationsafety,andtheeffectofenvironmentalconditionson
ship manoeuvrability is an important issue in study of ship manoeuvrability, only the inherent ship
manoeuvrabilitywillbedealtwithinthiscourse.
portB
x
portA
y
Figure2.4 Relationbetweenshipmanoeuvrabilityandnavigationeconomy
U
betterdynamicstability worsedynamicstability
external disturbances
originalcourse
(a) Relationbetweeninitialturningabilityandnavigationsafety
U
poorturningability
x
obstacle
goodturningability
poorinitialturningability
goodinitialturningability
y
x
y
Figure2.3 Relationbetweenshipmanoeuvrabilityandnavigationsafety
U
(b)Relationbetweenturningabilityandnavigationsafety
bank
bank
4
CHAPTER3:EVALUATIONOFSHIP MANOEUVRABILITY1.5hours
Demand: Know the various kinds of standard manoeuvres and the parameters used to evaluate ship
manoeuvrability know the definition of the parameters, the methods recommended to obtain the
parameters and the quantitative demands of the parameters in the IMO (International Maritime
Organization)StandardsforShipManoeuvrability.
3.1Brief Introduction
Ship manoeuvrability is an important hydrodynamic performance relating to navigationsafety. A ship
should have adequate manoeuvrability to ensure a safe navigation. To design a ship with good
manoeuvrability,itisimportantforustojudgeifashiphasgoodorpoormanoeuvrability.Forthispurpose
some standard manoeuvres are proposed. From these manoeuvres the significant qualities for evaluating
shipmanoeuvrabilitycanbederived.
3.2StandardManoeuvresandtheParametersEvaluatingShipManoeuvrability
3.2.1Turningtest
The turning test is performed to evaluate the turning ability. A turning circle manoeuvre is to be
performed to both starboard and port with 35 rudder angle or the maximum design rudder angle
permissibleatthetestspeed.Therudderangleisexecutedfollowingasteadyapproachwithzeroyawrate.
Thekinematicparametersandthepathofthemidshippointarerecordedduringthetest,seeFigure 3.1.
Theparametersobtainedfromthismanoeuvreforevaluatingtheturningabilityarethetacticaldiameter,
Tactical diameter
T
D
Driftangle
V
0
X
0
Y
180
changeof heading
Approachcourse
Advance
Rudderexecute
V
V
Transfer
90
changeofheading
Figure 3.1 Path of the midship point during turning
TurningradiusR
Pathof midshippoint
5
theadvanceandthetransfer.
3.2.2 Zigzagtest
Thezigzagtestis performedtoevaluatetheinitialturning,theyawcheckingand the coursekeeping
abilities. A zigzag test is to be initiated to both starboard and port and begins by applying a specified
amountofrudderangle.Therudderangleisthenalternatelyshiftedtoeithersideafteraspecifieddeviation
fromtheoriginalheadingoftheshipisreached
2
.
Usually two kinds of zigzag tests, the 10/10 and 20/20zigzagtests are performed. The 10/10
zigzag test uses 10 rudder angles to either side following a heading deviation of 10 from the original
course.The20/20 zigzagtestuses20 rudderanglesand20headingdeviationfromtheoriginalcourse.
The parameters obtained from this manoeuvre for evaluating the manoeuvring characteristics are the
overshoot angles
ov
y , the initial turning time to second execute t
i
and the time to check yaw, see Figure
3.2.
3.2.3Stoppingtest
Thestoppingtestisperformedtoevaluatethestoppingability.Afullasternstoppingtestisconducted
todeterminethetrackreachofshipfromthetime when anasternorderis givenuntiltheshipisstopped
deadinthewater.Thetrackreachisthelengthmeasuredalongthepathofmidshippoint,see Figure3.3.
2
ThezigzagmanoeuvrewasfirstproposedbytheGermanscientistGntherKempfin1932.
t
i
Port
Starboard
Time(s)
Headingangley
2
nd
overshootangle
2 ov
y
1
st
overshootangle
1 ov
y
0
Figure3.2 Timehistoriesofthe rudderangleandheadingangleduringzigzagtest
originalheading
(rudderangle)
, y (Deg)
Rudderangle
y (headingangle)
6
3.2.4Spiraltest
There are two kinds of spiral test, the direct spiral test, also called Dieudonns spiral test
3
, and the
reversespiraltest,alsocalledBechsreversespiraltest
4
,whichareperformedtoevaluatetheshipdynamic
stability and coursekeepingability.
Thedirectspiraltestisanorderlyseriesofturningcircletests.Therudderanglechangesinsequence
of+15+10+50 -5 -10 -15 -10 -50+5+10+15.Thetestresult
is expressed as steady turning rate versus rudder angle curve, as shown in Figure 3.4. This test is time
consumingandsensitiveto weatherconditions.Moreover,foradynamicallyunstableship, withinasmall
rangeofrudderangletheshipmaystillturnsteadilyintheoriginaldirectionalthoughtherudderisalready
deflected to the other side. In this case the curve of steady turning rate versus rudder angle will change
abruptly,asshowninFigure3.4(b).
Inthereversespiraltest,theshipissteeredtoobtainaconstantturningrate andthemeanrudderangle
requiredtomaintainthisrateismeasured.Thereversespiraltestprovidesamorerapidproceduretoobtain
the steady turning rate versus rudder angle curve than the direct spiral test, and can obtain the complete
curveforadynamicallyunstableship,asshowninFigure3.4(b).TheloopheightandloopwidthinFigure
3.4(b) reflect how unstable the ship is, therefore Figure 3.4(b) obtained from spiral tests can be used to
evaluatethe instabilityofship.
3
Thedirectspiraltestwasfirstproposed bytheFrenchscientistJeanDieudonn in19491950andpublishedin1953.
4
ThereversespiraltestwasfirstproposedbyMogensBechin1966andpublishedin1968.
Propellerrotationreversed Asternorder
Lateraldeviation
0
Y
0
X
Trackreach
Approachcourse
Headreach
Figure3.3 Pathofthemidshippointduringstoppingtest
Shipdeadinwater
7
3.2.5Pullouttest
Pulloutmanoeuvrewasproposedforevaluatingthedynamicstabilityofaship
5
.Itisconductedafter
completion of a turning circle manoeuvreto starboard or port with rudder angle of approximately 20
by
returningtheruddertothemidshippositionandkeepingitthereuntilasteadyturningrateisreached.Fora
stableship,thesteadyturningratewilldecaytozero,whereasforanunstableshiptheratewillbeacertain
value, the socalled residual turning rate, which is a measurement of the instability. Usually, the residual
turning ratestostarboardandportwillbedifferent,asshowninFigure3.5(b).
3.3IMOStandardsforShipManoeuvrability
Ships with very poor manoeuvrability mayresult in marine disasters and environmental pollution. To
avoidthatshipsarebuiltwith very poor manoeuvrability,InternationalMaritimeOrganization(IMO),the
international authoritative organization responsible for maritime safety and environment protection, has
madelongtermeffortsto developcriteriaand evaluation methods forship manoeuvrabilitysincelate60s
last century. In December 2002, IMO adopted the resolution MSC.137 (76), Standards for Ship
Manoeuvrability. The standards are to be used to evaluate the ship manoeuvrability and to assist those
responsible for ship design, construction, repair and operation to ensure that ships comply with the
standards.
5
Thepullouttestwasfirstproposed byRoyBurcherin1969.
Rudder returned tomidship
Rudder returned tomidship
r starboard r starboard
port port
t t
(a) Stableship (b)Unstableship
Figure3.5 Timehistoriesoftheturningrateduringpulloutmanoeuvre
20 10
(b)r~ curveforadynamicallyunstableship
Obtainedbyreverse
spiralmanoeuvre
20 10
r(Starboard)
Loopheight
(a)r~ curveforadynamicallystableship
r(Starboard)
Starboard
20 10
Port
Figure3.4 r~curveobtainedfromspiraltests
Loopwidth
Starboard
20 10
Port
8
According to the Standards for Ship Manoeuvrability, a ship is considered satisfactory if the
followingcriteriaarecompliedwith:
1. Turningability
Theadvanceshouldnotexceed4.5shiplengthsandthetacticaldiametershouldnotexceed5ship
lengthsintheturningcirclemanoeuvre.
2. Initialturningability
Withtheapplicationof10rudderangletostarboard/port,theshipshouldnothavetraveledmore
than2.5shiplengthsbythetimewhentheheadinghaschangedby10fromtheoriginalheading.
(It is equivalent to require that the initial turning time to second execute in 10/10 zigzag test
satisfiest
i
<2.5L/V,whereLisshiplengthand Visshipspeed).
3. Yawcheckingandcoursekeepingabilities
3.1 The first overshoot angle in 10/10zigzagtest should not exceed 10if L/V is less than 10 s
20 ifL/Vis30sormoreand5+1/2(L/V)degreesifL/Vis10sormore,butlessthan30s.
3.2 Thesecondovershootanglein10/10zigzagtestshouldnotexceed25ifL/Vislessthan10s
40 ifL/Vis30sormoreand17.5+0.75(L/V)degreesifL/Vis10sormore,butlessthan30s.
3.3 Thefirstovershootanglein20/20 zigzag testshouldnotexceed25.
4. Stopping ability
The track reach in the full astern stopping test should not exceed 15 ship lengths. However, this
valuemaybemodifiedbytheadministrationwhereshipsoflargedisplacementmakethiscriterion
impracticable,butshouldinnocaseexceed20shiplengths.
3.4Fullscale Trials
TodemonstratethecomplianceofanewbuiltshipwiththeIMOstandards,fullscaletrialsofstandard
manoeuvresaretobeconductedbeforedelivery.Thecharacteristicsaremeasuredfromthetestrecordsand
aretobe comparedwiththeIMOcriteria.
CHAPTER4:PREDICTIONOFSHIPMANOEUVRABILITY3.5 hours
Demand: Know the various kinds of freerunning model tests and master the methods to obtain the
parameters evaluating manoeuvrability from the tests know how to derive the equations of ship
manoeuvring motion know the various kinds of methods for determining the hydrodynamic forces
acting on a manoeuvring ship master the method of manoeuvrability prediction by means of the
mathematicalmodelandcomputersimulationtechnique.
4.1Brief Introduction
According to the IMO Standards for Ship Manoeuvrability, manoeuvring performance of a ship
shouldbedesignedtocomplywiththestandardsduringthedesignstage.Therefore,predictionoftheship
manoeuvringperformanceshouldbeconductedattheearlydesignstage.Atpresent,therearetwomethods
forthispurpose.Thefirstoneisthemethodoffreerunningmodeltests,andthesecondoneisthemethod
ofcomputersimulationusingthemathematicalmodels.
4.2FreerunningModel Tests
Freerunning model tests are conducted with a scale model. The standard manoeuvres are performed
andtheparametersevaluatingmanoeuvringcharacteristicsaremeasureddirectlyfromthetestrecords.This
kind ofmethodisthe most direct methodfor predicting the manoeuvring performanceandisregardedas
9
themostreliablepredictionmethod.However,themethodistimeconsumingandcostsalotofmoney,and
isinconvenientforapplyingatthedesignstage,sinceseriesof modeltestsareusuallyneededtoconduct.
Moreover,itshouldbenotedthatduetothesocalledscaleeffects,thereexistsusuallyadistinctdifference
betweenthepredictedperformanceandtheactualone.
4.3Equationsof ShipManoeuvringMotion(Mathematical Models)
With the rapid development of the computer technology and its successful application in ship
engineering, the method of computer simulation using the mathematical models becomes more and more
popular. It provides a convenient tool for predictingship manoeuvrability at the design stage. One of the
preconditions for applying this kind of method is modeling of the equations describing the manoeuvring
motion.
4.3.1Coordinatesystems
To investigate ship manoeuvrability by means of mathematical tool, two righthanded coordinate
systemsareadopted:the earthfixed coordinatesystem
0 0 0 0
o x y z andthebodyfixedcoordinatesystem
o xyz which movestogetherwiththeship,see Figure4.1. The
0 0 0
o x y planeandthe o xy planelie
ontheundisturbedfreesurface,withthe
0
x axispointingtothedirectionoftheoriginalcourseoftheship,
whereasthe
0
z axisandthe z axispointdownwardsvertically.Theanglebetweenthedirectionsof
0
x
axis and x axis is defined as the heading angle, y . At the moment as manoeuvring motion is getting
start,thetwocoordinatesystemscoincidewitheachother.Atanylatermoment,thepositionoftheshipis
determined by the coordinates
0G
x and
0G
y of the ship center of gravity in the earthfixed coordinate
system,and theorientationoftheshipisdeterminedby theheading angle y .
The ship manoeuvring motion in the horizontal plane is described by the speed V
v
of translational
motionandtheyawrate r y = & ofrotationalmotionaboutthe z axis.Thecomponentsofthespeed V
v
inthedirectionsof x axisand y axisare u and v ,respectively.Theanglebetweenthedirectionsof
speed V
v
and x axisisdefinedasthedriftangle, b.Obviouslywehave
with cos , sin . u V v V V V b b = = - =
v
4.3.2Mathematicalmodels
(driftangle) y
oG
x
oG
y
r
(yawrate)
Shipcenterofgravity
V
v
(shipspeed)
(rudderangle)
x
y (headingangle)
0
y
0
x
Original course
Figure4.1 Coordinatesystems
o
G
0
O
10
Here mathematicalmodelsmeantheequationsofshipmanoeuvringmotion.Theequationsofmotion
canbederived accordingtotheNewtonsecondlaw,the Newtonianlawofmotion.
The earthfixedcoordinatesystemisan inertialframe inwhichtheNewtonianlawofmotionisvalid.In
thiscoordinatesystem,wehavethe equationsofmotionasfollows.
0 0
0 0
0
,
,
.
G
G
zG
X mx
Y m y
N I y
=
=
=
&&
&&
&&
(4.1)
where
0
X and
0
Y arethecomponents ofexternalforceacting ontheshipinthedirections of
0
x axis
and
0
y axis,respectively
0
N isthe componentof externalmomentaboutthe z axis m isthemass
oftheship
zG
I isthe moment ofinertiaoftheshipabout z axis
0G
x && and
0G
y && arethe components
of acceleration inthedirectionsof
0
x axis and
0
y axis,respectively y && istheyaw acceleration.
Forpurposeofshipmanoeuvrabilityprediction,itismoreconvenienttousetheequationsofmotionin
the bodyfixed coordinate system. To establish these equations, the relations between the kinematic
parameters defined in the earthfixed and the bodyfixed coordinate systems are to be utilized. With the
originalofthebodyfixedcoordinatesystemlyingonthecenterofgravity,wehave
0 0
0 0
0
cos sin ,
sin cos ,
.
G
G
x x x y
y y x y
z z
y y
y y
= + -
= + +
=
orinreverse,
0 0 0 0
0 0 0 0
0
( ) cos ( ) sin ,
( ) sin ( ) cos ,
.
G G
G G
x x x y y
y x x y y
z z
y y
y y
= - + -
= - - + -
=
Denoting
0 0 0G
x x x = - ,
0 0 0G
y y y = - and
0 0
z z = ,wehave
0
0
0
cos sin ,
sin cos ,
.
x x y
y x y
z z
y y
y y
= -
= +
=
and inreverse,
0 0
0 0
0
cos sin ,
sin cos ,
.
x x y
y x y
z z
y y
y y
= +
= - +
=
Denotingthe components of external force in the directions of x axis and y axis by X and Y,
andthecomponentsofshipspeedinthedirectionsof x axisand y axisby
G
u and
G
v ,respectively,
wehave
0 0
0 0
cos sin ,
sin cos .
X X Y
Y X Y
y y
y y
= +
= - +
(4.2)
0
0
cos sin ,
sin cos .
G G G
G G G
x u v
y u v
y y
y y
= -
= +
&
&
(4.3)
11
Differentiating Equation(4.3)withrespecttotimeyields
0
0
cos sin sin cos ,
sin cos cos sin .
G G G G G
G G G G G
x u u v v
y u u v v
y y y y y y
y y y y y y
= - - -
= + + -
& & && & &
& & & & &
(4.4)
Substituting Equations (4.1) and (4.4) into Equation (4.2), we obtain the equations of motion in the
bodyfixedcoordinatesystem,withtheoriginalofthesystemlyingontheshipcenterofgravity,as
( ),
( ).
G G
G G
X m u v
Y m v u
y
y
= -
= +
& &
& &
(4.5)
Ontheotherhand,sincetheexternalmomentaboutthe z axisthroughthe centerofgravity issame in
the bodyfixed coordinate system as in the earthfixed coordinate system, we have the equation of yaw
motioninthebodyfixedcoordinatesystemasEquation(4.1).
Inpracticeitismoreconvenientwhentheoriginalofthebodyfixedcoordinatesystemdoesnotlieon
thecenterofgravity,butonthemidshippoint.Assumingthattheshipissymmetricalaboutitslongitudinal
centerplane,thecenterofgravityhasthecoordinates( , 0,
G G
x z )inthebodyfixedcoordinatesystemwith
the original lying on the midship point. In such a coordinatesystem the components of ship speed at the
centerofgravity,
G
u and
G
v ,and attheoriginal, u and v ,havetherelationsasfollows
, .
G G G
u u v v x y = = + &
Therefore, we obtain the equations of motion in the bodyfixed coordinate system with the original
lyingonthemidshippointintheform
2
( ),
( ).
G
G
X m u v x
Y m v u x
y y
y y
= - -
= + +
& & &
& && &
(4.6)
Ontheotherhand,since
2
0
,
G z zG G
N N Y x I I mx = - = + ,
where N istheexternalmomentand
z
I isthemomentofinertiaaboutthe z axis,respectively.From
the third equation of Equation (4.1) we have the equation of yaw motion in the bodyfixed coordinate
system withtheoriginallyingonthemidshippointasfollows
( ).
z G
N I mx v u y y = + + && & &
Asaresult,we havetheequationsof motioninthebodyfixed coordinatesystem,withthe originalof
thecoordinatesystemlyingonthemidshippoint,intheform
2
( ),
( ),
( ).
G
G
z G
X m u vr x r
Y m v ur x r
N I r mx v ur
= - -
= + +
= + +
&
& &
& &
(4.7)
where r y = & istheyawrateaboutthe z axis.
4.3.3Expressionsofthehydrodynamicforces
X , Y and N inEquation(4.7)arethecomponentsofexternalforceandmomentactingontheship.
These force and moment include the hydrodynamic force and moment in calm water as well as various
environmentalexcitingforceandmomentduetowind,wavesandcurrent,etc.Inthiscoursewewillonly
dealwith thehydrodynamicforceandmomentincalmwater.
12
There are two kinds of manner for expression of the hydrodynamic force and moment, one is that
introduced first by Prof. Martin A. Abkowitz from MIT in USA and the other one is due to the
MathematicalModelingGroup(MMG)fromJapan.
In1964,Prof.Abkowitzproposedamethodforexpressionofthehydrodynamicforceandmomentby
using the Taylor expansion. The hydrodynamic force and moment are expressed as functions of the
kinematicalparametersandtherudderangleintheform
( , , , , , , ),
( , , , , , , ),
( , , , , , , ),
X X u v r u v r
Y Y u v r u v r
N N u v r u v r
d
d
d
=
=
=
& & &
& & &
& & &
and then expanded in Taylor series about the initial steady state of forward motion with constant speed,
i.e.,
0 0 0 0 0 0 0
, 0, 0, 0, 0, 0, 0. u U v r u v r d = = = = = = = & & & Itresultsin
0
2
( )
1
( )
2!
1
( ) ,
!
n
X X X X X X X
X X u U v r u v r
u v r u v r
u U v r u v r X
u v r u v r
u U v r u v r X
n u v r u v r
d
d
d
d
d
d

= + - + + + + + +

+ - + + + + + +

+ + - + + + + + + +

& & &
& & &
& & &
& & &
& & & L L
& & &
(4.8a)
0
2
( )
1
( )
2!
1
( ) ,
!
n
Y Y Y Y Y Y Y
Y Y u U v r u v r
u v r u v r
u U v r u v r Y
u v r u v r
u U v r u v r Y
n u v r u v r
d
d
d
d
d
d

= + - + + + + + +

+ - + + + + + +

+ + - + + + + + + +

& & &
& & &
& & &
& & &
& & & L L
& & &
(4.8b)
0
2
( )
1
( )
2!
1
( ) ,
!
n
N N N N N N N
N N u U v r u v r
u v r u v r
u U v r u v r N
u v r u v r
u U v r u v r N
n u v r u v r
d
d
d
d
d
d

= + - + + + + + +

+ - + + + + + +

+ + - + + + + + + +

& & &
& & &
& & &
& & &
& & & L L
& & &
(4.8c)
where
0
X ,
0
Y and
0
N aswellasallthederivativestakethevalueattheinitialsteadystateofforward
motionwith
0 0 0 0 0 0 0
, 0, 0, 0, 0, 0, 0. u U v r u v r d = = = = = = = & & &
ThederivativesinEquation(4.8)arecalledhydrodynamicderivatives.Forsimplicitythesederivatives
areusuallyexpressedas
, , , , , , .
u v r u v r
X X X X X X X
X X X X X X X
u v r u v r
d
d

= = = = = = =

& & &
L
& & &
SubstitutingEquation(4.8)intoEquation(4.7),theequationsofshipmanoeuvringmotionarederived.
ThiskindofequationsiscalledAbkowitzmodel,orwholeshipmodel.
Another methodforexpressionofthehydrodynamicforceand momentwasproposedbytheJapanese
MathematicalModelingGroup,JMMG,inthelate1970slastcentury.Inthisexpressionthehydrodynamic
13
forceandmomentaredecomposedintothreeparts,i.e.,thepartsactingontheshiphull,thepropellerand
therudder,respectively,intheform
,
,
,
H P R
H P R
H P R
X X X X
Y Y Y Y
N N N N
= + +
= + +
= + +
(4.9)
wherethe subscriptsH, P and Rdenotethehull,thepropellerandtherudder,respectively.
ReferringtotheAbkowitzmodel,expressionsforthecomponentsofhydrodynamicforceandmoment
in Equation (4.9) in form of hydrodynamic derivatives as well as the coefficients accounting for the
interaction effects of the hull, propeller and rudder were proposed. Substituting Equation (4.9) into
Equation (4.7), the equations of ship manoeuvring motion are derived. This kind of equations is called
MMGmodel,ormodularmodel.
4.4MethodsforDeterminingtheHydrodynamicForces ActingonaManoeuvringShip
Obviously, as a preconditionto use the equations of motion to simulate the manoeuvring motion, the
hydrodynamic derivatives in the equations should be determined. There are various methods for this
purpose. Inthefollowing,thebasic methods whichcanbeusedto determinethe hydrodynamic forceand
momentactingonashipinmanoeuvringmotionwillbe briefly introduced.
4.4.1Captivemodeltests
Captivemodeltestsaretestsconductedwithascalemodelinshipmodelbasins,wherethemodelsare
forcedtomoveinaprescribedmanner.Thesetestsincludeobliquetowingtestinaconventionallongand
narrowtowingtank,asshowninFigure4.2,rotatingarmtestinarotatingarmfacility,asshowninFigure
4.3,andtheplanarmotiontestusingPlanarMotionMechanism(PMM)inalongandnarrowtowingtank
asshowninFigure4.4(a),(b)and(c),aswellasCircularMotionTest(CMT)inabigtowingtankorina
seakeepingandmanoeuvringbasin,seeFigure4.5.Byanalyzingtheforcesandmomentsmeasuredonthe
model,thehydrodynamicderivativescanbedetermined.
Itshouldbepointedoutthatcaptive modeltestssufferfromscaleeffectsasinthefreerunning model
tests,thereforeextrapolationcorrectionsareusuallyneeded.
TowingTank
Carriage
Model
V(speedofthecarriage)
Figure4.2 Obliquetowingtestintowingtank
V(speedofthecarriage)
14
(a) PMMinatowingtank
TowingTank
Xcarriagemovingalongthetowingtank
Ycarriage movinginthetransversaldirection
0
x
0
y
(b) Puresway test
Pathofthemidshippoint
w=r
Circularpathofmodel
Rotatingarm
Circularbasin
Figure4.3 Rotatingarmtestinrotatingarmfacility
V
v
15
4.4.2Systemidentification technique
The hydrodynamic derivatives in the equations of motion can also be determined from the fullscale
trialsorfreerunningmodeltestsbysystemidentificationmethods.Thedataofcontrolparameterssuchas
rudder angle are measured and used as input, whereas the data of kinematic parameters such as speed,
acceleration are measured and used as output, then the system parameters, i.e., the hydrodynamic
derivatives in the dynamic model, the equations of motion, are determined by means of the parameter
identificationmethods.
4.4.3Semitheoreticalandsemiempiricalmethods
Thehydrodynamicderivativescanalsobedeterminedbysemitheoreticalandsemiempiricalmethods.
Thesemethodsincludethemethodsofdatabaseandtheestimationmethodsusingsemiempiricalformulae.
Systematicalcaptive modeltestsneedtobeconducted,andthedataof measuredhydrodynamicforceand
moment are to be gathered, from which a data base or semiempirical formulaecan be derived. Then the
data base and the semiempirical formulae can be used to estimate the hydrodynamic derivatives in the
equationsofmotion conveniently.
4.4.4Numericalmethods
Xcarriagemoving longitudinally
Ycarriagemoving transversely
Figure 4.5 CircularMotion Test
0
x
0
y
(c) Pure yawtest
Figure4.4 PMMtestsinatowingtank
16
The hydrodynamic force and moment acting on a ship in manoeuvring motion can be calculated by
theoreticaland/ornumerical methods. Traditionally,simple methodsbased on potentialtheorysuchasthe
lowaspectratiowingtheoryandtheslenderbodytheoryareusedtocalculatethehydrodynamicforceand
moment.Withtherapiddevelopment of computertechnologyand numerical methodsinthefield offluid
dynamics,ComputationalFluidDynamics(CFD)techniquehasbecomemoreandmorepopularinthefield
of ship hydrodynamics. Completethreedimensional numerical methods are now availablefor calculating
the hydrodynamic force and moment acting on a ship in manoeuvring motion. These 3D numerical
methodscanbeclassifiedintotwo categories,i.e.,themethodsbasedonpotentialtheory,suchasBoundary
Element Method (BEM), also called panel method, and the viscous flow methods, such as the method
based on solution of ReynoldsAveraged NavierStokes (RANS) Equations by means of, for example,
FiniteVolumeMethod(FVM).Greatprogresshasbeenachievedinthisaspectduring the lastdecade.
4.5Prediction Basedon Mathematical ModelsandComputerSimulation
Compared with the method of freerunning model tests, the method of numerical simulation using
mathematical models isausefulandconvenient method for predictingship manoeuvrabilityatthe design
stage. Once the hydrodynamic derivatives are determined, the equations of ship manoeuvring motion,
which are a set of differential equations, can be solved by numerical methods such as the RungeKutta
methodstodeterminethecomponentsofmanoeuvringmotion ( ), u t ( ) v t and ( ) r t atanytime t .Then
the position and the orientation of the ship can be determined by the coordinates of the ship center of
gravityin theearthfixed coordinatesystemandthe heading angle of the ship, which can be obtained by
numericalintegrationsasfollows
0 0 0 0
0 0 0
( ) ( ) , ( ) ( ) ( ) ( )
t t t
G G G G
x t x t dt y t y t dt t t dt y y = = =

& & & (4.10)
where
0
0
( ) ( ) cos ( ) [ ( ) ( )]sin ( ),
( ) ( ) sin ( ) [ ( ) ( )]cos ( ),
( ) ( ).
G G
G G
x t u t t v t x r t t
y t u t t v t x r t t
t r t
y y
y y
y
= - +
= + +
=
&
&
&
(4.11)
In this way, any standard manoeuvres can be simulated by a computer and the manoeuvring
characteristicsoftheshipcanbepredicted.
CHAPTER5:LINEAREQUATIONSOF SHIPMANOEUVRINGMOTION4 hours
Demand: Know the physical meanings and characteristics of the linear hydrodynamic derivatives in the
equationsofship manoeuvring motion masterthe methodofanalyzingthecoursekeepingabilityand
turning ability based on the linear equations of ship manoeuvring motion know how to derive the
Nomoto model and its application in course control problems understand and master the relationship
betweentheK,TindexesintheNomotomodelandshipmanoeuvrability.
5.1BriefIntroduction
As stated in the last chapter, the equations of motion, i.e., the mathematical models, can be used to
simulateanystandardmanoeuvres,and hence topredict ship manoeuvrability. Inthis chapter we will use
thesimplifiedequations,thelinearequationsofshipmanoeuvringmotionto analyze shipmanoeuvrability.
As shown in Equation (4.8), the hydrodynamic force and moment acting on a ship in manoeuvring
17
motion are expressed in series, with the hydrodynamic derivatives as coefficients. Under the assumption
that the changes in velocities and accelerations , , , , , u u U v v r r u u v v r r D = - D = D = D = D = D = & & & & & &
duetothemanoeuvringmotionandtherudderangle d aresmall,thehighordertermsintheexpressions
of hydrodynamic force and moment can be neglected. This results in the linear equations of ship
manoeuvringmotionasfollows
2
0
0
0
( ) ( ) ,
( ) ( ) ,
( ) ( ) .
G u v r u v r
G u v r u v r
z G u v r u v r
m u vr x r X X u U X v X r X u X v X r X
m v ur x r Y Y u U Y v Y r Y u Y v Y r Y
I r mx v ur N N u U N v N r N u N v N r N
d
d
d
d
d
d
- - = + - + + + + + +
+ + = + - + + + + + +
+ + = + - + + + + + +
& & &
& & &
& & &
& & & &
& & & & &
& & & & &
(5.1)
For consistency, the terms
2
,
G
vr x r in Equation (5.1) should be neglected and for the samereason,
( ) ( ) ur u U U r u U r Ur Ur = - + = - + . On the other hand, the hydrodynamic force and moment at
theinitialsteadystateofforwardmotion
0 0 0
, , X Y N shouldvanish.ItresultsfromEquation(5.1)
( ) ,
( ) ( ) ,
( ) ( ) .
u v r u v r
G u v r u v r
z G u v r u v r
mu X u U X v X r X u X v X r X
m v Ur x r Y u U Y v Y r Y u Y v Y r Y
I r mx v Ur N u U N v N r N u N v N r N
d
d
d
d
d
d
= - + + + + + +
+ + = - + + + + + +
+ + = - + + + + + +
& & &
& & &
& & &
& & & &
& & & & &
& & & & &
(5.2)
5.2LinearHydrodynamicDerivatives
Thelinearhydrodynamicderivativeshaveexplicitphysicalmeanings.Theymeantheforceormoment
induced by a unit component of velocity or acceleration in manoeuvring motion, while the other
components keepvanished.Mathematicallytheyarethe gradients ofthecurveofforceor moment versus
thecomponentofvelocityoraccelerationattheorigin.Obviously,forshipswithportstarboardsymmetry,
the linear hydrodynamic derivatives , ,
v r
X X , ,
v r
X X
& &
X
d
and ,
u u
Y Y
&
, ,
u u
N N
&
should vanish due to
the symmetry of the flow about the ship centerplane, as shown in Figure 5.1 for
v
X as an example.
Therefore,weobtainthelinearequationsof shipmanoeuvringmotion asfollows
( ) ( ) 0,
( ) ( ) ( ) ,
( ) ( ) ( ) .
u u
v v G r r
G v v z r G r
m X u X u U
m Y v Y v mx Y r mU Y r Y
mx N v N v I N r mx U N r N
d
d
d
d
- - - =
- - + - + - =
- - + - + - =
&
& &
& &
&
& &
& &
(5.3)
Sometimesitismoreconvenienttousethenondimensionalformoftheequationsofmotion.Dividing
the first two equations of Equation (5.3) by
2 2
1
2
L U r and dividing the last equation by
3 2
1
2
LU r
respectively,where r isthemassdensityofwater, L istheshiplengthand U istheshipspeedatthe
0 =

=
v
X
X
v
v
X
Figure5.1 Typical X versus v curve
18
initialsteadystate of forward motion,we obtainthe nondimensionalform ofthelinearequations ofship
manoeuvringmotion
( ) ( ) 0
( ) ( ) ( )
( ) ( ) ( )
u u
v v G r r
G v v z r G r
m X u X u U
m Y v Y v m x Y r m U Y r Y
m x N v N v I N r m x U N r N
d
d
d
d
- - - =
- - + - + - =
- - + - + - =
&
& &
& &
&
& &
& &
(5.4)
where thenondimensionalparametersandcoefficientsaredefinedas
2
3 3 2
2
3 2
2
2
4 3 2 2
4 3
, , , , , 1
1 1 1
2 2 2
, , , ,
1 1
2 2
, , , , ,
1 1 1
2 2 2
, ,
1 1
2 2
u u
u u
v v
v v
G r r
G r r
v v
v v z
X X m uL u
m X u X u U
U U
L L L U
Y Y vL v
Y v Y v
U U
L L U
x Y Y Y rL rL
x Y r Y r Y
L U U
L LU L U
N N I
N N I
L LU
d
d
r r r
r r
r r r
r r
= = = = = =
= = = =
= = = = = =
= = =
&
&
&
&
&
&
&
&
&
&
&
&
&
&
5 5 4 3 2
, , , .
1 1 1 1
2 2 2 2
z r r
r r
N N N
N N N
L L L U LU
d
d
r r r r
= = =
&
&
InEquation(5.3),thehydrodynamicforcesproportionaltotheaccelerationsareinertiaforces,whereas
those proportional to the velocities are damping forces. Correspondingly, the coefficients , ,
u v
X Y - -
& &
r
Y -
&
and
v
N -
&
arecalledaddedmassesandcanbeexpressedas
11 22
, m m ,
26
m and
62
m and
r
N -
&
is called added moment of inertia and can be expressed as
66
m whereas the coefficients
, , ,
u v r v
X Y Y N - - - - and
r
N - arecalleddampingcoefficients.Ontheotherhand,thecoefficients Y
d
and N
d
are calledcontrolderivatives.
Accordingtotheanalysis of the flowabouttheship, wecan determinethemagnitudeandthesignof
thelinearhydrodynamicderivativesasfollows:
u
X
&
isthelongitudinalforceinducedbyaunitlongitudinalacceleration.Thisforcetendstoarrestthe
acceleration.Thatmeans,forapositiveaccelerationtheforcewillbenegative,andviceversa,asshownin
Figure 5.2. As a result, the gradient of theX ~u & curve at the origin is negative, which means
u
X
&
is
negative.
v
Y
&
is the lateral force induced by a unit transverse acceleration. Similar as
u
X
&
, we can find
v
Y
&
is
negative,seeFigure5.3.Ontheotherhand,sincetheforcesactingontheforehalfpartandontheafthalf
0 <
u
X
&
u &
X
Figure5.2 Typical X versus u & curve
19
partoftheshipinducedbythetransverseaccelerationareinthesamedirection,asshowninFigure5.4,the
absolutevalueof totalforcewillbelarge,therefore
v
Y
&
haslargemagnitude.
From intuition we can image that
v
Y
&
is larger than
u
X
&
. Actually, for a conventional ship the
magnitudesofthese twohydrodynamicderivativeshaveapproximatelythemagnitude
(0.05 ~ 0.15)
u
X m -
&
(0.9 ~1.2) .
v
Y m -
&
u
X and
v
Y arethelongitudinalforceandlateralforceinducedbya unitlongitudinalvelocityanda
unit lateral velocity, respectively. Similar analysis can be made for
u
X and
v
Y as for
u
X
&
and
v
Y
&
to
findthat
u
X and
v
Y arenegative,and
v
Y haslargemagnitude,whereas
u
X hasmoderatemagnitude.
r
N
&
and
r
N are the yaw moments induced by a unit yaw acceleration and by a unit yaw rate,
respectively.Sinceapositiveyawaccelerationorapositive yawratewillinduceanegativeyaw moment,
wecanfind
r
N
&
and
r
N arenegative.Moreover,sincethe momentsacting onthe fore half partand on
theafthalfpartoftheshipinducedbytheyaw accelerationorbytheyawrateareinthesamedirection,the
absolutevalueoftotalmomentwillbelarge,therefore
r
N
&
and
r
N havelargemagnitude.Figure5.5and
Figure5.6showthecaseof
r
N
&
asanexample.
r
Y
&
and
r
Y are the lateral forces induced by a unit yaw acceleration and by a unit yaw rate,
respectively.Sincetheforcesactingontheforehalfpartandontheafthalfpartoftheshipinducedbythe
yawaccelerationorbytheyawrateareintheoppositedirection(seeFigure5.6asexample),thetotalforce
willbeverysmall(Forashipwithforeandaftsymmetry,itshouldvanish),andthesignofthetotalforce
will be the same as the larger one of the fore and aft forces, therefore
r
Y
&
and
r
Y have very small
magnitudeand uncertainsign. Asimilaranalysis canbe madefor
v
N
&
and
v
N asfor
r
Y
&
and
r
Y with
conclusion that
v
N
&
and
v
N have very small magnitude and uncertain sign (see Figure 5.7 as example
for
v
N
&
).
0 <
r
N
&
N
r &
Figure5.5 Typical N versus r & curve
0 <
a
N
0 <
f
Y
0 <
f
N
y
(Localacceleration r x v & & = )
r &
x
o
0 >
a
Y
N
f
Y
a
Y
x
v &
Y
y
o
Figure5.4 Forceonashipwithacceleration v &
0 <
v
Y
&
Y
v &
Figure5.3 Typical Y versus v & curve
Figure5.6 Momentonashipwithyawacceleration r &
20
Theresultsoftheanalysisabove are summarized inTable5.1.
Table5.1Magnitude and sign ofthelinearhydrodynamicderivatives
item
u
X
& u
X
v
Y
& v
Y
r
N
& r
N
r
Y
& r
Y
v
N
& v
N
magnitude moderate moderate verylarge verylarge verylarge verylarge verysmallverysmallverysmallverysmall
sign negative negative negative negative negative negative uncertain uncertain uncertain uncertain
Ontheotherhand,controlderivatives Y
d
and N
d
denotethecontrolforceandmomentinducedbya
unit rudder angle. With rudder angle to starboard defined as positive, a negative force and a positive
moment will be induced by a positive rudder angle. Hence Y
d
is negative, whereas N
d
is positive.
Obviously, rudder is to be designed to give as large control force and moment as possible, therefore Y
d
and N
d
should beaslargeaspossible.
5.3 AnalysisofStability(inherentdynamicstability)
It can be seen that in Equation (5.3) the first equation is uncoupled from the last two equations.
Therefore,itcanbesolvedseparately.Ontheotherhand,thelasttwoequationsarecoupledeachotherand
hencehavetobesolvedtogether.
ThefirstequationofEquation(5.3)canberewrittenas
( )
( ) 0.
u
u
X d u U
u U
dt m X
-
- - =
-
&
(5.5)
Equation (5.5) is an ordinary differential equation of first order with constant coefficients, whose
solutionisgiven by
3
3
,
t
u U c e
s
- = (5.6)
where
3
/( )
u u
X m X s = -
&
,and
3
c isanintegralconstanttobedeterminedbytheinitialcondition.
Since 0
u
X < and 0,
u
m X - >
&
we have
3
0. s < Therefore, as , t the term on the right
handside ofEquation(5.6)willvanish,resultingin . u U Thatmeans,thelongitudinal motionofthe
shipwillfinallytendtotheoriginalsteadystatewith aconstantforwardspeed.
Toanalyzetheinherentdynamicstabilityoftheship(i.e.thedynamicstabilitywithoutcontrolactions),
weset 0 d = andassumethatthelateralmotionandyawmotionareinducedbyasmalldisturbance.We
obtainfromthelasttwoequationsofEquation(5.3)
Y
a
Y
f
Y
x
v &
y
o
Figure5.7 Momentonashipwithacceleration v &
f
N
a
N
21
( ) ( ) ( ) 0,
( ) ( ) ( ) 0,
v v G r r
G v v z r G r
m Y v Y v mx Y r mU Y r
mx N v N v I N r mx U N r
- - + - + - =
- - + - + - =
& &
& &
& &
& &
(5.7)
whichcanberewrittenas
( ) ( ) ( ) ,
( ) ( ) ( ) .
v v G r r
G v v z r G r
m Y v Y v mx Y r mU Y r
mx N v N v I N r mx U N r
- - = - - - -
- - = - - - -
& &
& &
& &
& &
(5.8)
SolvingEquation(5.8)for v & and v yields
( )[( ) ( ) ] ( )[( ) ( ) ]
,
( ) ( )
[( ) ( ) ] [( ) ( ) ]
.
( ) ( )
v G r z r G v r G r
v G v v v
v G r z r v r G r
v G v v v
m Y mx U N r I N r mx N mU Y r mx Y r
v
Y mx N N m Y
Y mx U N r I N r N mU Y r mx Y r
v
Y mx N N m Y
- - - + - + - - + -
=
- - -
- - + - + - + -
=
- - -
& & & &
& &
& &
& &
& &
& &
&
(5.9)
Differentiatingthefirstequationof Equation (5.9)withrespecttotimeandlettingtheresultequaltothe
secondequation,weeliminate , v v & fromEquation(5.9)andobtain
0, Ar Br Cr + + = && & or 0,
B C
r r r
A A
+ + = && & (5.10)
where
( )( ) ( )( ),
( ) ( ) ( )( ) ( )( ),
( ) ( ).
v z r G v G r
v z r v G r v G r G v r
v G r v r
A m Y I N mx N mx Y
B Y I N N mx Y m Y mx U N mx N mU Y
C Y mx U N N mU Y
= - - - - -
= - - + - + - - - - -
= - - + -
& & & &
& & & &
Equation(5.10) is an ordinary differential equation of second order with constant coefficients, whose
solutionisgiven by
1 2
1 2
,
t t
r r e r e
s s
= + (5.11)
where
1
r and
2
r areintegralconstantstobedeterminedbytheinitialcondition,
1
s and
2
s aregiven
by
2
1
2
1
4 .
2
B B C
A A A
s
s

= - -

(5.12)
Similarly,ifwe eliminate , r r & fromEquation(5.7),weobtain
0, Av Bv Cv + + = && & or 0,
B C
v v v
A A
+ + = && & (5.13)
withthesolution
1 2
1 2
,
t t
v v e v e
s s
= + (5.14)
where
1
v and
2
v areintegralconstantstobedeterminedbytheinitialcondition,
1
s and
2
s aregiven
by Equation(5.12).
As we can see, if both
1
s and
2
s are negative real numbers or imaginary numbers with negative
real parts, we will have
1
0
t
e
s
and
2
0
t
e
s
as . t That means, after a small disturbance,
0 r and 0 v as , t theshipwillfinallytendtoanewstraightcoursewithconstantforward
speed.Atthemeantime,theshipwillhaveaheadingdeviationfromtheoriginalheadingwhichisgivenby
22
0
0
( ) ,
t
r t dt y =
(5.15)
where
0
t isthemoment when r vanishes.Obviously,ashipwithbetterinherent dynamicstability will
tendtoanewstraightcoursefasterandwillhavesmallerheadingdeviation.
From the above analysis, it can be seen that a ship affected by a small disturbance can not keep its
originalcoursewithoutcontrolaction.Itcanonlyturnintoanewstraightcoursewithaheadingdeviation
fromtheoriginalcourseifithasinherentdynamicstability.Contrarily,iftheshipisunstable,theyawrate
givenbyEquation(5.11)willnotvanishas , t thereforetheshipwillturnintoanunsteadyturning
motion.SeeFigure2.1forreference.
Both
1
s and
2
s are negative real numbers or imaginary numbers with negative real parts is the
criterionforinherentdynamicstability.Furtheranalysisofthestabilityismadeinthefollowing.
Forarbitrary
B
A
,if 0,
C
A
< then
2
1
1
4 0
2
B B C
A A A
s

= - + - >

and
2
2
1
4 0
2
B B C
A A A
s

= - - - <

.
Therefore,toensureboth
1
s and
2
s arenegativerealnumbersorimaginarynumberswithnegative
realparts,wemusthave 0.
C
A
>
Then,assuming 0,
C
A
>
(1) if
2
4 0,
B C
A A

- >

1
s and
2
s are two different real numbers. Since
2
4 ,
B C B
A A A

- <

both
1
s and
2
s are positive real numbers if 0
B
A
< both
1
s and
2
s are negative real numbers if
0.
B
A
> Inthiscase,toensurethatboth
1
s and
2
s arenegativerealnumberswemusthave 0.
B
A
>
(2)if
2
4 0,
B C
A A

- =

1
s and
2
s arerealnumbersandequal:
1 1
2
B
A
s s = = - .Inthiscase,to
ensurethatboth
1
s and
2
s arenegativerealnumberswemusthave 0.
B
A
>
(3)if
2
4 0,
B C
A A

- <

1
s and
2
s aretwodifferentimaginarynumberswiththerealpart
2
B
A
- .
Inthiscase,toensurethatboth
1
s and
2
s areimaginarynumberswithnegativerealpartwemusthave
0.
B
A
>
In conclusion, to ensure both
1
s and
2
s are negative real numbers or imaginary numbers with
23
negativerealparts,wemusthave 0
B
A
> and 0.
C
A
>
Actually, by requiring
1 2
0 s s > and
1 2
0 s s + < (which means that both
1
s and
2
s are
negativerealnumbersorimaginarynumberswithnegativerealparts)itismore straightforwardtoseethe
criterionforinherentdynamicstability:
2 2
1 2
1 1
4 4 0
2 2
B B C B B C C
A A A A A A A
s s

= - + - - - - = >

.
2 2
1 2
1 1
4 4 0 0
2 2
B B C B B C B B
A A A A A A A A
s s

+ = - + - + - - - = - < >

.
Since
v
m Y -
&
and
z r
I N -
&
are positive and very large, whereas
G v
mx N -
&
and
G r
mx Y -
&
are
smallwithuncertainsign,itcanbeconcludedthat
( )( ) ( )( ) 0.
v z r G v G r
A m Y I N mx N mx Y = - - - - - >
& & & &
Moreover, since
v
Y - ,
z r
I N -
&
,
v
m Y -
&
and
r
N - are positive and very large,
v
N ,
G r
mx Y -
&
,
G
mx U,
G v
mx N -
&
and
r
Y aresmallwithuncertainsign,itcanbeconcludedthat
( ) ( ) ( )( ) ( )( ) 0.
B Y I N N mx Y m Y mx U N mx N mU Y = - - + - + - - - - - >
& & & &
Therefore,thecriterionforinherentdynamicstabilityisreducedto
( ) ( ) 0.
v G r v r
C Y mx U N N mU Y = - - + - > (5.16)
Thecorrespondingnondimensionalinequality is
( ) ( ) 0.
v G r v r
C Y m x U N N m U Y = - - + - > (5.17)
Inequality (5.16) shows that the criterion for inherent dynamic stability includes only four linear
hydrodynamicderivatives.Thesederivatives are also called the stability derivatives. A ship withinherent
dynamicstabilityhasapositive C.Thelargerthemagnitudeofthepositive C,thebetteristhedynamic
stability.
Since 0
v
Y - > and 0
r
mU Y - > ,dividing Inequality (5.16)by ( )
v r
Y mU Y - - yields
0,
r G v
r v
r v
N mx U N
l l
Y mU Y
-
- = - >
-
(5.18)
where
,
r G v
r v
r v
N mx U N
l l
Y mU Y
-
= =
-
are the moment arms due to the forces induced by the yaw motion and the sway motion, respectively.
Inequality(5.18)showsthatthe force dueto the yaw motion mustact ontheshipbeforethe forcedue to
theswaymotion,inordertomake theshipdynamicallystable.
5.4NomotoModelandK,TIndexes
Themathematicalmodelsdiscussedin4.3arethesocalledhydrodynamicforcemodels.Nowwewill
deriveanotherkindofmodels,theresponsemodels,fromthehydrodynamicforcemodels.Suchmodelsare
24
mainlyusedtostudythecoursecontrolproblemsortodesignautomaticcontroldevicessuchasautopilots.
We will also see that although this kind of models has simpler form, the coefficients of the models have
explicitrelationwiththemanoeuvringcharacteristicsandcanbeusedtoevaluate shipmanoeuvrability.
ThelasttwoequationsofEquation(5.3)arerewrittenintheform
( ) ( ) ( ) ,
( ) ( ) ( ) .
v v G r r
G v v z r G r
m Y v Y v Y mx Y r mU Y r
mx N v N v N I N r mx U N r
d
d
d
d
- - = - - - -
- - = - - - -
& &
& &
& &
& &
(5.19)
SolvingEquation(5.19)for v & and v yields
( )[ ( ) ( ) ] ( )[ ( ) ( ) ]
,
( ) ( )
[ ( ) ( ) ] [ ( ) ( ) ]
.
( ) ( )
v G r z r G v r G r
v G v v v
v G r z r v r G r
v G v v v
m Y N mx U N r I N r mx N Y mU Y r mx Y r
v
Y mx N N m Y
Y N mx U N r I N r N Y mU Y r mx Y r
v
Y mx N N m Y
d d
d d
d d
d d
- - - - - - - - - - -
=
- - -
- - - - - - - - -
=
- - -
& & & &
& &
& &
& &
& &
& &
&
(5.20)
Differentiatingthefirstequation ofEquation(5.20)with respecttotimeand lettingtheresultequalto
the second equation,weobtain
1 2 1 2 3
( ) , TT r T T r r K KT d d + + + = +
&
&& & (5.21)
where
1 2
1 2
3
( )( ) ( )( )
,
( ) ( )
( ) ( ) ( )( ) ( )( )
,
( ) ( )
,
( ) ( )
( )
v z r G v G r
v G r v r
v G r v r
v v
v G r v r
G v
m Y I N mx N mx Y
TT
Y mx U N N mU Y
Y I N N mx Y m Y mx U N mx N mU Y
T T
Y mx U N N mU Y
Y N N Y
K
Y mx U N N mU Y
Y mx N
KT
d d
d
- - - - -
=
- - + -
- - + - + - - - - -
+ =
- - + -
- +
=
- - + -
- - +
=
& & & &
& & & &
&
3
3
( )
,
( ) ( )
( ) ( )
.
v
v G r v r
G v v
v v
N m Y
Y mx U N N mU Y
KT Y mx N N m Y
T
K Y N N Y
d
d d
d d
-
- - + -
- - + -
= =
- +
&
& &
Comparingtheexpressionsofthecoefficients , A , B C inEquation(5.10)andthecoefficients
1 2
, TT
1 2
T T + in Equation (5.21), noting that
1
s and
2
s are given by Equation (5.12), we can find the
relationshipsbetween
1
s ,
2
s and
1
, T
2
T asfollows
1 2
1 2 1 2
1 1 1 C
A TT T T
s s

= = = - -

and
1 2
1 2
1 2 1 2
1 1 T T B
A TT T T
s s
+
- = + = - = - + -

.
Equation(5.21)istheresponsemodelofsecondorder.Itexpressestherelationbetweentheyawmotion
andtherudderangleandcanbeusedtodescribetheyawmotionresponsetothecontrolaction.
Assuming that the ship is approximately foreandaft symmetrical, Equation (5.21) can be further
reducedtothe responsemodeloffirstorder.
Forashipwhichisapproximately foreandaftsymmetrical,wehave
0,
G
x 0,
r
Y 0,
r
Y
&
0,
v
N 0
v
N
&
.
Itfollowsthat,
25
,
r
N
K
N
d
-
1 2
,
v z r
r v
m Y I N
T T
N Y
- -
+ - -
& &
3
,
v
v
m Y
T
Y
-
-
&
1 2 3
.
z r
r
I N
T T T
N
-
+ - -
&
Ontheotherhand,fromthethirdequationofEquation(5.3)wehaveapproximately
( )
z r r
I N r N r N
d
d - - =
&
& ,
orrewritten in theform
z r
r r
N I N
r r
N N
d
d
-
- + = -
&
& .
Denoting
1 2 3
T T T T = + - ,wehave
Tr r Kd + = & . (5.22)
Equation (5.22) is the response model of first order. It was firstly derived by Japanese professor K.
Nomoto in 1957 by using the method of Laplace Transformation. Therefore, the response model of first
orderisalsocalledNomoto model. , K T inthis modelarecalledmanoeuvrabilityindexes,since,aswe
canseeinthefollowing,theyhave explicitrelationswithmanoeuvringcharacteristics.
Theresponse modelsofsecond orderandfirst order canbe usedto predictthe yaw motion under the
controlaction.Inthefollowingwewillusetheresponsemodeloffirstordertoinvestigatetheyawmotion
responseoftheshiptotherudderaction.
It is assumed that the ship is initially in its straight course witha steady forwardspeed the rudder is
orderedtoturntoacertainangle
0
d duringthetime
0
t withauniformrudderturningrateandthenheld
fixed,seeFigure 5.8.Thechangeofrudderanglewithtimeisgivenby
0
0
0
0 0
, 0
,
t t t
t
t t
d
d
d d

=

= >
(5.23)
(1) Atthefirstphase,
0
0 t t ,itresultsfromEquations(5.22) and(5.23)
0
0
1 K
r r t
T Tt
d
+ = & (5.24)
Equation(5.24)isan ordinarydifferentialequationoffirstorder,whosesolutionisgivenby
(Starboard)
t
0
Timet(s)
0
Figure5.8 Timehistoryofrudderangle
26
0
0
( ) ,
t t
T T
K
r e e t T c
t
d
-
= - +

(5.25)
where c isanintegralconstant.Accordingtotheinitialcondition 0 r = at 0 t = , c isfoundas
0
0
.
K
c T
t
d
= (5.26)
SubstitutingEquation(5.26)inEquation(5.25)resultsin
0
0
( ).
t
T
K
r t T Te
t
d
-
= - + (5.27)
At
0
t t = ,itresultsfromEquation(5.27)
0
0
0
0
( ).
t
T
K
r t T Te
t
d
-
= - + (5.28)
IntegratingEquation(5.27)withrespecttotime t resultsin
2
2 0
0
(1 ) .
2
t
T
K t
Tt T e
t
d
y
-
= - + -

(5.29)
DifferentiatingEquation(5.27)withrespecttotime t resultsin
0
0
(1 ).
t
T
K
r e
t
d
-
= - & (5.30)
(2) Atthesecondphase,
0
t t > ,itresultsfromEquations(5.22)and(5.23)
0
1
.
K
r r
T T
d
+ = & (5.31)
Equation(5.31)isadifferentialequationoffirstorder,whosesolutionisgivenby
0
,
t t
T T
r e K e c d
-
= +

(5.32)
where c isanintegralconstant.Accordingtotheinitialcondition,Equation(5.28), c isfoundas
0
0
0
1 .
t
T
K
c T e
t
d

= -

(5.33)
SubstitutingEquation(5.33)inEquation(5.32)resultsin
0
0
0
0
(1 ) .
t t
T T
K
r t T e e
t
d
-
= + -

(5.34)
IntegratingEquation(5.34)withrespecttotime t (from
0
t to t )resultsin
0
0
2
2 0 0
0 0
0
( ) ( ) ( 1) .
2
t t t
T T
t
K t
t rdt t t T T e e
t
d
y y
-
= + = - - + -

(5.35)
InderivingEquation(5.35)wehaveutilized
0
( ) t y obtainedfromEquation(5.29).
Ontheotherhand,differentiatingEquation(5.34)withrespecttotime t resultsin
27
0
0
0
( 1) .
t t
T T
K
r e e
t
d
-
= - & (5.36)
Since 0
z r
I N - >
&
and 0
r
N < , we have 0 T > . Therefore 0
t
T
e
-
as t . It results form
Equation(5.35),(5.34)and(5.36)thatas t
0
0
( )
2
t
K t T y d

- -

,
0
r Kd and 0. r &
This means, as t the ship will trend into steady turn with a steady yaw rate
0 0
r Kd = and the
headinganglechangingwiththetime t linearlyas
0 0 0
0 0
( ) ( ) .
2 2
t r t
K t T K t T y d d

= - + = - -

Denoting theship speed in steady turningphaseas
0
V and the turning radius as R, thediameter of
theturningcircle,
0
2 D R = ,canbedeterminedby
0 0
0
0 0
2 2 .
V V
D
r Kd
= = (5.37)
From Equation(5.27)and(5.34)wecanseethattheyawrateisproportionalto K.Atacertainrudder
angle,larger K willgivelargeryawrate.Moreover,inthesteadyturningphasetheyawrate
0 0
r Kd = ,
which means larger K will give larger steady yaw rate
0
r and hence smaller diameter of the turning
circle
0
D . Therefore, the parameter K can be used to evaluate the turning ability. Larger K implies
betterturningability.
Ontheotherhand,asdiscussedabove,foranormalship, 0 T > .Atacertainmoment t ,smaller T
willresultinsmaller
t
T
e
-
(
t t
T T
t
T e e
T
-
- - )asindicatedbyEquation(5.30),thismeansa
largeryawacceleration r & attheinitialturningphase.Therefore,smaller T impliesbetterinitialturning
ability or coursechanging ability. Moreover, since 0
t
T
e
-
as t for a smaller T,
t
T
e
-
will
vanish faster. It means, as indicated by Equation (5.34), the ship will tend to steady turning faster. Once
Pathof midshippoint
0
r
0
V
Figure 5.9 Turningradiusinsteadyturningphase
TurningradiusR
0
2 D R =
28
again,thisindicatesthatsmaller T willgivebetterinitialturningability.
Smaller T, actually, will give also better dynamic stability on a straight course. To show this, we
assume that a ship is originally traveling in a straight course and then affected by a small disturbance.
Without control action, the yaw motion of the ship is described by the following equation according to
Equation(5.22)
0. Tr r + = & (5.38)
ThesolutionofEquation(5.38)isgivenby
,
t
T
r ce
-
= (5.39)
where c isanintegralconstantdeterminedbytheinitialcondition
0
r r = at 0 t = ,resultingin
0
c r = .
Equation (5.39) describes the yaw motion of the ship after a small disturbance. It can be seen that as
t theyawratewillvanishafterthedisturbancesmaller T willmaketheyawratevanishingfaster
andhencegiveabetterdynamicstabilityonastraightcourse.
Itshouldbearinmindthattheanalysisaboveusingtheresponsemodeloffirstorder,Equation(5.22),
isbasedontheassumptionof 0,
G
x 0,
r
Y 0,
r
Y
&
0,
v
N 0
v
N
&
.
As discussed above, the parameter K can be used to evaluate the turning ability, whereas the
parameter T canbeusedtoevaluatethe dynamicstabilityonastraightcourse,theinitialturningability
and the coursechanging ability. For good ship manoeuvrability we wish the ship having larger K and
smaller T. However, as we will see in the following, a large K usually accompanies a large T or
contrarily,asmall K accompaniesasmall T.
Werecallthe definitionofthemanoeuvrabilityindexes
,
r
N
K
N
d
-
z r
r
I N
T
N
-
-
&
where N
d
is the control moment derivative due to the rudder angle,
r
N is the damping moment
derivativeduetotheyawmotion,
z
I isthemomentofinertiaoftheshipand
r
N -
&
istheaddedmoment
ofinertia.
To obtain a large K for good turning ability, we wish a small
r
N . This will, however, result in a
large T andhenceapoordynamicstabilityonstraightcourse andapoorinitialturningability atthe same
time. Therefore, we can not improve the ship manoeuvrability only by changing the value of damping
moment for yaw motion. On the other hand, obtaining large control moment derivative by designing the
rudder may result in good turning ability without loss of other manoeuvring performances. Therefore,
designingtheruddertoobtainlargecontrolmomentistheeffectivewaytoimproveshipmanoeuvrability.
To obtain a small T for better dynamic stability on straight course and initial turning ability, one
mighttrytodesigntheshipwithsmallmomentofinertiaandsmalladdedmomentofinertia.However,the
momentof inertiaandaddedmomentof inertiaaremainlydeterminedbytheprincipaldimensionsandship
form coefficients whicharechosenaccordingto othership performances.Hencethere isnobigspacefor
choosingthe principal dimensions and ship form coefficients to obtain small T. For large ship with full
form such as VLCC, T is usually very large due to the large inertia. Therefore, such ships are usually
dynamicallyunstableandhavepoorinitialturningabilityandcoursechangingability.
As stated above, the manoeuvrability indexes K and T can be conveniently used to evaluate the
manoeuvringperformancesofships.Forcomparisonpurpose,itismoreconvenienttousenondimensional
manoeuvrability indexes. Since K has the dimension of 1/ s and T has the dimension of s, the
nondimensionalformsoftheseindexesaregiveby
29
,
L
K K
U

=

.
U
T T
L

=

5.5 AnalysisofTurningAbilityBasedontheLinearEquations
In last subsection, the response model of first order is used to analyze ship yaw motion and the
relationshipbetweenthemanoeuvrabilityindexes K, T andshipmanoeuvrability.Inthissubsectionwe
willusethelinearhydrodynamicforcemodelto analyze theturningmotionofships.
It is assumed that a ship originally travels with a steady forward speed in a straight course, and the
rudderisorderedtoacertainrudderangletostarboardorportwithaconstantrudderturningrateandthen
heldfixed,seeFigure5.8.Ultimately,the shiptendsto asteady turningmotion.
Thewholeprocesscanbedividedintothreephases:Therudderturningphase,thetransitionphaseand
thesteadyturningphase.Inthefollowingwewillanalyzethemotioninthethreephasesindetailbyusing
thelasttwoequationsofEquation(5.3):
( ) ( ) ( ) ,
( ) ( ) ( ) .
v v G r r
G v v z r G r
d
d
d
d
- - + - + - =
- - + - + - =
& &
& &
& &
& &
(5.40)
(1) Rudderturningphase
At this phase,
0
0 , t t , the rudder angle d is given by
0 0
/ , t t d d = where
0
d is the rudder
angle applied and
0
t is the time of rudderturning. The control force and moment of the rudder will
produce a sway acceleration and a yaw acceleration. But the time of rudderturning is so small and the
inertiatheshipissolarge thatatthisphasenonotable swayspeedandyawrate areinduced.
Therefore,wehaveattherudderturningphase
0
0
0, 0 0, 0. t v r v r
t
d
d = & &
ItresultsfromEquation(5.40)
( ) ( ) .
( ) ( ) .
v G r
G v z r
m Y v mx Y r Y
mx N v I N r N
d
d
d
d
- + - =
- + - =
& &
& &
& &
& &
(5.41)
FromEquation(5.41)weobtain v & and r & asfollows
( ) ( )
,
( )( ) ( )( )
( ) ( )
.
( )( ) ( )( )
z r G r
v z r G v G r
v G v
v z r G v G r
I N Y mx Y N
v
m Y I N mx N mx Y
m Y N mx N Y
r
m Y I N mx N mx Y
d d
d d
d
d
- - -
=
- - - - -
- - -
=
- - - - -
& &
& & & &
& &
& & & &
&
&
(5.42)
Equation (5.42) shows that the sway acceleration v & and the yaw acceleration r & change with the
rudderangleproportionallyattherudderturning phase. NotingthatthedenominatorinEquation(5.42)is
thecoefficient A in Equation(5.10),whichisusuallypositive,since
( ) ( ) ( ) 0,
( ) ( ) ( ) 0,
z r G r z r
v G v v
I N Y mx Y N I N Y
m Y N mx N Y m Y N
d d d
d d d
- - - - <
- - - - >
& & &
& & &
forapositiverudderangle,theswayacceleration v & willbenegativeandtheyawacceleration r & willbe
positive. This means, at the rudderturning phase, a positive rudder angle will create a negative sway
30
accelerationandapositiveyawacceleration.
Assumingthat
G
x ,
r
Y
&
and
v
N
&
are very smallandtherelevanttermscanbeneglected,wehave
.
v z r
Y N K
v r
m Y I N T
d d
d d d =
- -
& &
& & (5.43)
Equation(5.43)showsthatlarger K andsmaller T will givealarger , r & henceabetterinitialturning
ability.
(2) Transitionphase
Atthisphase,
0
d d = 0, 0 v r & & 0, 0 v r .ItresultsfromEquation(5.40)
0
0
( ) ( ) ( ) ,
( ) ( ) ( ) .
v v G r r
G v v z r G r
d
d
d
d
- - + - + - =
- - + - + - =
& &
& &
& &
& &
(5.44)
At the transition phase, the control force and moment remain unchanging, whereas the kinematical
quantitiesoftheshipchangewithtime.WecansolveEquation(5.44)bynumericalintegrationmethod,for
example,bytheRungeKuttamethod,toobtainthekinematicalquantities ( ), ( ), v t r t & & ( ) v t and ( ). r t
Attheinitialtransitionstage,theshipmoveslaterallytotheoutside(othersideastherudderdeflection)
underrudderforceactionandturnstoinside(thesamesideastherudderdeflection)underruddermoment
action.Withthedevelopment of ( ), ( ), v t r t & & ( ) v t and ( ), r t thehydrodynamicforceand momentacting
ontheshiphullincreasegradually.Ultimately,thehydrodynamicforceandmomentactingonthehulland
the control force and moment due to the rudder angle reach equilibrium, the ship then enters to the next
phase,i.e.,thesteadyturningphase.
(3) Steadyturningphase
At this phase,
0
d d = 0, 0 v r = = & &
0 0
, , v v r r = = where
0
v and
0
r are the steady sway
velocity andthe steadyyawrate.ItresultsfromEquation(5.40)
0 0 0
0 0 0
( ) ,
( ) .
v r
v G r
Y v mU Y r Y
N v mx U N r N
d
d
d
d
- + - =
- + - =
(5.45)
SolvingEquation(5.45)for
0
v and
0
r ,weobtain
0 0
0 0
( ) ( )
.
( ) ( )
.
( ) ( )
r G r
v G r v r
v v
v G r v r
mU Y N mx U N Y
v
Y mx U N N mU Y
Y N N Y
r
Y mx U N N mU Y
d d
d d
d
d
- - + -
=
- - + -
- +
=
- - + -
(5.46)
NotingthatthedenominatorinEquation(5.46)isthecoefficient C discussedinsubsection5.3,we
obtain thesteadyturningdiameterasfollows
0 0
0
0 0
2 2 2 ,
v v
V V C
D R
r Y N N Y
d d
d

= = =

- +

(5.47)
where
0
0
V
R
r
= isthesteadyturningradius,
2 2
0 0
V U v = + isthesteady ship speed,seeFigure5.9.
Thenondimensionalsteadyturningdiameterisgivenby
31
0 0
0
0 0
2 2
2 ,
D V R
D
L L r L r

= = = =
(5.48)
where,with
2 2
0 0
, V U v U = +
0
0 0
0
.
v v
r L Y N N Y L
r
V R C
d d
d
- +
= = =
(5.49)
SincefromEquation (5.46)
0 0
,
v v
Y N N Y
r
C
d d
d
- +
= (5.50)
we can see that at a certain rudder angle, a positive C with larger absolute value, which means a better
dynamicalstability,willresultinasmallersteadyyawrateandhencealargersteadyturningdiameter.This
indicatesthatabetterdynamicalstabilitywillusually accompany aworseturningability,andviceversa.
Moreover, since
v
Y is negative with large absolute value and 0, N
d
> we have 0
v
Y N
d
- > and
since 0 Y
d
< and
v
N is small with uncertain sign, it follows that
v
N Y
d
is small with uncertain sign.
Hence,
v v
Y N N Y
d d
- + hasthesamesignas
v
Y N
d
- has. Thatmeans, 0
v v
Y N N Y
d d
- + > .
Therefore,if 0 C > (shipisdynamicallystable),forapositiverudderangle(ruddertostarboard),we
have
0
0 r > (turning to starboard) whereas for a negative rudder angle (rudder to port), we have
0
0 r < (turningtoport).Thissituationiswhatweexpect.
On the other hand, if 0 C < (ship is dynamically unstable), for a positive rudder angle (rudder to
starboard), we have
0
0 r < (turning to port) whereas for a negative rudder angle (rudder to port), we
have
0
0 r > (turningtostarboard).That means,forashipwhichis dynamically unstable,withrudderto
starboard, the ship will turn to port whereas with rudder to port, the ship will turn to starboard. This
impliesthattheshipisuncontrollable.
CHAPTER6:CONTROLDEVICES1.5hours
Demand: Know the various kinds of control devices know the geometrical and hydrodynamic
characteristicsofrudderandmasterthebasicmethodofrudderdesign.
6.1KindsofControlDevices
Shipmanoeuvrability istheabilityof ashiptomaintainasteadycourseortochangethestateofmotion
includingthedirectionandthespeedofmotionundercontrolactions,whichisexertedbyacontroldevice.
Hence controldevice isavitalelementtoshipmanoeuvrability.
In general, control devices can be classified into two categories: active control devices and passive
controldevices.Activecontroldevicesconvertenergyintocontrolforcesactivelyinamoredirectmanner,
whereas passive control devices produce control forces passively by absorbing energy from the ambient
water.Passivecontrol devicesareusuallyusedforcommon manoeuvringpurpose,whereasactivecontrol
devicesareinstalledonshipsoperatedatlowerspeedsorwithspecialmanoeuvrabilityrequirements.
Themostpopularactivecontroldeviceislateralthruster,whichconsistsofatransversetunnelwithan
impellerinthemiddleandproducesalateralforcebyjetreaction,seeFigure6.1.Itmaybeinstalledatthe
boworatthestern.Alateralthrusteratthebow,thebowthruster,isgenerallyeffectiveatlowershipspeeds
atwhichtherudderisleasteffective.Athighershipspeeds,theflowaroundthehulldistortstheoutflowof
bowthrusterhencethethrusterbecomespracticallyineffective.Ontheotherhand,alateralthrusteratthe
stern,thesternthruster,isinboundarylayerflowandiseffectiveathighershipspeeds.
32
Other active control devices include azimuthing rudder propeller, also called Zdriver, which is a
azimuthingductedpropellerhingedinaverticalpositionandisusedasacombinedpropulsionandsteering
system, see Figure 6.2, and steering nozzle, which consists of a main propeller and a steering nozzle
combinedwithafixedorflappedfin andcreatessteeringforce bymeansofadeflectedjet,see Figure 6.3.
Thesimplestand most commonly used passivecontrol deviceisrudder,whichcanbefurther divided
intotwogroups,conventionalrudderandunconventionalrudder.Conventionalrudderhassimpleformand
structureandisusedforcommonmanoeuvringpurpose,seeFigure6.4.Unconventionalrudderusuallyhas
higher hydrodynamic performance and produces high lift in some active manner. It is derived from
BECKERKortnozzle
Figure 6.3 Steeringnozzleasanactivecontroldevice
(a)withfixedfin (b)withflapped fin
rudderpropellerusedonatug
Figure 6.2 SCHOTTEL rudderpropeller
sternthruster
Figure 6.1 Lateralthrusterasanactivecontroldevice
bowthruster
33
conventionalrudderandisusedformeetingspecialmanoeuvrabilityrequirements.
Unconventionalruddersinclude rotatingcylinderrudder,flaprudder,etc.
Arotatingcylinderrudderincorporatesaverticalrotatablecylinderonthe forwardedgeoftherudder,
seeFigure6.5.Whenthecylinderisrotating,circulationisgeneratedandaliftisproducedwhichisnormal
tothestreamandthecylinder.
Aflaprudderisakindofallmoveablerudderwithatailflap,seeFigure6.6.Theflapcanturnrelative
tothemainpartoftherudderandsomakethecamberoftheruddervariable.Inthiswayhigherliftmaybe
produced.
6.2RudderandItsDesign
Figure 6.6 Flaprudderasaunconventionalrudder
BECKERflaprudder
w
Figure 6.5 Rotatingcylinderrudderasaunconventionalrudder
BECKERrotating cylinder rudder
Figure 6.4 Conventionalrudderasapassivecontroldevice
34
In this chapter we will only deal with rudders and will concentrate our attention on the conventional
rudderanditsdesign.
6.2.1 Rudder types
Themaintypesof conventionalrudderinclude:
Simple type, rudder post simple type, fully balanced balanced, with fixed structure semibalanced
underhungrudderspaderudderseeFigure6.7.
6.2.2 Ruddergeometry
A typical allmovable rudder is shown in Figure 6.8. The parameters describing rudder geometry
include rudder height (span) b, chord length c , maximal thickness of section t , rudder stock position
d,sectionshape,andrudderarea
R
A .
Since most rudders are neither rectangular in shape nor of uniform thickness, the mean values of the
geometricalparametersaretobeused:
stockaxis
Figure 6.8 Atypicalallmovableconventionalrudder
f
b rudder height atthe leading edge
t
b rudder height atthe trailingedge
t
c chordlength atthetopside
b
c chordlength atthebottomside
t
t thickness ofsectionatthetopside
b
t thickness ofsectionatthebottomside
d mean distance between the leading edge
andthestockaxis
t
c
b
c
t
b
f
b
b
t
t
t
d
Figure 6.7 Ruddertypes(fromPrinciplesofNavalArchitecture,VolumeIII,SNAME,1989)
35
Meanrudderheight
2
t f
b b
b
+
= Meanchordlength
2
t b
c c
c
+
= Meanthickness .
2
t b
t t
t
+
=
Theratio / b c isthegeometricaspectratiotheratio / t c isthethicknesschordratio.Therudder
areacanbetakenas
R
A b c = .
6.2.3 Rudderforceinfreestream
A rudder with rudder angle (angle of attack) a in free stream of velocity U is considered. A
resultant hydrodynamicforce F
v
acting on therudder will be induced. This force acts on a single point
calledthecenterof pressure,CP.Theresultantforce canberesolvedintoaliftcomponent, L ,normalto
the direction of free stream and a drag component, D, parallel to the direction of the free stream. The
resultant force can also be divided into a component normal to the centerplane of the rudder and a
componentin the centerplane,seeFigure6.9.
Denotingthecomponentnormaltothecenterplaneas
N
F ,wehave
cos sin .
N
F L D a a = + (6.1)
Theproductofthisnormalforcetimesthedistanceofthecenterofpressurefromtheaxisoftherudder
stockgivesthehydrodynamictorqueabouttherudderstock,whichisofimportancetorudderdesign.
For comparison of rudder forces and moments acting on rudders of different sizes and operating at
differentspeeds,itisconvenienttoexpresstherudderforcesandmomentsinnondimensionalform:
2 2
2
,
1 1
2 2
cos sin .
1
2
L D
R R
N
N L D
R
L D
C C
U A U A
F
C C C
U A
r r
a a
r
= =
= = +
(6.2)
Obviously, rudders are to be designed to give as large lift coefficient
L
C as possible, and as small
dragcoefficient
D
C aspossible,sothattheruddersareofhighefficiency inhydrodynamicperformance.
6.2.4 Rudderdesign
Thebasicconsiderationsinrudderdesignincludethefollowing aspects:
(1) Ruddertype.Forthespecialapplicationpurpose,decisionistobemadeifactiveorpassivecontrol
device, conventional or unconventional rudder should be chosen. Also among different types of
Figure 6.9 Rudderforceinfreestream
a
U
F
v
CP
L
D
N
F
d
36
conventionalrudderonepropertypeisto be chosen.
(2) Thelocation ofrudder.Therudderis usually placedabaftthepropelleratthestern. The propeller
stream is to be utilized to improve the rudder effectiveness and hence the controllability at slow
speedofship.
(3) Number of rudder. Especially for inland waterway ships, due to the higher requirement of
manoeuvrability on one hand and thelimitation of water depth on the other hand, usually two or
threeruddersare neededtogive enoughrudderarea.
(4) Rudderarea(ruddersize,height).Inprinciple,rudderareashouldbeaslargeaspossible.However,
rudder size is usually limited by the space at stern. Moreover, larger rudder area requires more
powertobeconsumedbyturningtherudder.
(5) Ruddersectionshape.Differentruddersectionshapeshave differenthydrodynamiccharacteristics
andhencegivedifferentruddereffectiveness.Someruddersectionshapeswithhighhydrodynamic
performancewereproposedbasedon systematicalmodeltests.
CHAPTER7:MEASURES TOIMPROVESHIP MANOEUVRABILITY0.5hour
Demand:Knowtheeffectsofshipformandappendagesetc.onshipmanoeuvrabilityknowthemeasures
tobetakentoimprovemanoeuvrabilitybeabletoapplytheknowledge inpracticalshipdesign.
Traditionally, manoeuvrability has received little attention during the ship design stage. Analysis of
manoeuvrabilityisusuallyperformedlateinthedesign,whentheprincipaldimensionsandshipformhave
beendeterminedaccordingtorequirementsofotherhydrodynamicperformances.Withtheadoptionofthe
IMOsStandardsforShipManoeuvrability,manoeuvrabilityhasgraduallybecomeanimportantiteminthe
ship designspiral.AccordingtoIMOsstandards,shiphastobedesigned withadequate manoeuvrability.
Therefore, it is important to choose the principal dimensions and ship form by making a compromise
amongshipmanoeuvrabilityandotherhydrodynamicperformancesatthedesignstage.
Tosimplifytheanalysis oftheeffects of principaldimensionsandshipformonship manoeuvrability,
we recallthedefinitionsofmanoeuvrabilityindexes
,
r
N
K
N
d
-
z r
r
I N
T
N
-
-
&
thestabilitycriterion
( ) ( ) 0
v G r v r
C Y mx U N N mU Y = - - + - >
andthe steadyyawrate andsteadyturningdiametergivenby
0
0 0 0
0
, 2 .
v v
Y N N Y V
r D
C r
d d
d
- +
= =
As discussed in Chapter 5, larger K gives a better turningability smaller T gives better dynamic
stability on a straight course larger K and smaller T give a better initial turning ability and
coursechanging ability. Apositive C with larger magnitude gives better dynamic stability on a straight
course. Based on this knowledge we can analyze the effects of the hydrodynamic coefficients on
manoeuvrabilityasfollows.
Since K is proportional to N
d
, a larger N
d
will give a larger K and hence a better turning
ability. This means, the rudder is to be designed to give as larger yaw moment as possible. This can be
achievedby designingrudder with high efficiency andlarger rudder area. Actually, such a rudder at zero
37
rudderanglewillplayarollofstabilizingfin,andhenceisalso favorablefordynamic stabilityonastraight
course.
Since K and T are inversely proportional to
r
N , a larger
r
N will give a smaller K and a
smaller T,andhenceaworse turning abilityandabetterdynamicstabilityonastraightcourse.Therefore,
toobtainabetterdynamicstabilityonastraightcourse,thelateralareaofthecenterplaneoftheshipshould
be as full as possible at both ends of the ship. On the other hand, to obtain a better turning ability, the
centerplaneoftheshipshouldbefullinthemidshipandbethinatbothends.
Since T is proportional to
z r
I N -
&
, a large
z r
I N -
&
will give a large T and hence a worse
dynamicstability onastraight course. Therefore,the massdistributionoftheshipshouldbeconcentrated
nearthemidshipsforabetterdynamicstabilityonastraightcourse.
Since
G
x is small, and
v
Y ,
r
N are negative with large magnitudes,
v r
Y N is positive with large
magnitude and ( )
v G r
Y mx U N - - is also positive. Moreover, since
v
N ,
r
Y are small with uncertain
sign,
r
mU Y >> ,
r
mU Y - is positive, the sign of ( )
v r
N mU Y - is the same as
v
N . If
v
N is
positive, ( )
v r
N mU Y - willbepositive,and C willbe positiveandlarge. Therefore,to obtainalarge
and positive C,
v
Y and
r
N should be negative with as large magnitudes as possible, and
v
N should
bepositivewithaslargemagnitudeaspossible. Toensurethat
v
N ispositive,apositivelateralvelocity
v shouldinduceapositiveyawmoment N,andanegativelateralvelocity v shouldinduceanegative
yaw moment N. This means, thelateral area of the centerplane of the ship should be fuller at the stern
thanatthebow.Forthisreason,atrimatthesternwillresultin abetterdynamicstability.
Since Y
d
is negative and N
d
is positive with rudder deflectionto starboard defined as positive, we
canseefromEquation(5.50)thatapositive
v
N willresultinasmallersteadyyawrateandhencealarger
steadyturningdiameter,whichimpliesaworseturningstability.Inotherwords,inordertoobtainabetter
turning ability,
v
N shouldbe negative. Here we see once again that the course stabilityand the turning
abilityare inconflictwitheachother.
CHAPTER8:BRIEFINTRODUCTION TOSEAKEEPING0.5hour
Demand:Knowthebasicconceptsofseakeepingandtherelationbetweenseakeepingperformancesanda
safeand comfortablenavigationknowtheshipmotionsinwaveswithsixdegreesoffreedom.
Thesecondpartofthiscourse,seakeeping,dealswithship motionsin oceanwaves, whichiscontrary
toshipmanoeuvringmotionsincalmwaterasdiscussedpreviously.
8.1Definitionof SeakeepingandItsContents
Seakeeping performance concerns the ability of seagoing ships to carry out a particular mission in a
given sea condition, speed and heading. A ship traveling in rough seas will undergo irregular oscillating
motions. These motions will not only affect the effectiveness of the ship in attaining its mission, but also
mayresultindangerofstructuraldamageorcapsizing.Therefore,goodseakeepingperformanceisvitally
importantforasafeandeffectivenavigation.
Seakeepingperformanceincludestheshipperformanceofmotionsatseas,aswellasthatofthederived
responses,suchas:
Slamming:asuddenrelativemotionbetweentheshipandtheambientwater,especiallyatthebow,see
Figure8.1.
Shippingwaterondeck:seeFigure8.2.
Propellerracing:propelleroutofwaterandrunrapidly,seeFigure8.2.
Speedloss:including(passive)speedlossduetoaddedresistanceinwaves,increasedpropellerloading
and reduced propulsive efficiency due to propeller racing, and the socalled voluntary speed loss, i.e., to
38
lowthespeedactivelyinordertoreducethemotionsinducedbywaves.
8.2ShipMotionsinSixDegreesofFreedom
A ship floating on the free surface andtravelingin rough seas will undergo oscillating motions in six
degrees of freedom. These are three oscillating translations and three oscillating rotations, as shown in
Figure8.3.Itisworthwhiletocomparethemanoeuvringmotionsdiscussedpreviously,whicharemotions
insingledirection.
CHAPTER9:OCEANWAVESANDSTATISTICANALYSIS3.5hours
Demand: Know the causes, classifications, statistic characteristics and expressions of ocean waves
understand and master the theories of regular and irregular waves master the basic concepts and
methods of spectrum analysis master the response relation of linear systems be able to apply the
methods of spectrum analysis and frequency response to prediction of the statistic characteristics of
shipmotioninoceanwaves.
9.1BriefIntroduction
Ocean waves constitute the environment of seagoing ships and arealso the cause of ships oscillating
motionatseas.
Ocean waves are characterized by their irregularity, both in time and in space. The ship motions
induced by ocean waves are also irregular. However, according to the principle of superposition, the
irregularmotionsofashipinoceanwavescanbedescribedasthelinearsuperpositionoftheresponsesof
the ship to all the wave components, which are regular and have different lengths, amplitudes and
propagating directions, wheretheamplitudesareassumedassmall.Forthisreason, we willstartwiththe
introductionoftheregulargravitywavesofsimpleform inthischapter.
9.2RegularWaves
One of the regular gravity waves is the cosine wave, whose wave profile (free surface elevation) is
yaw
pitch
roll
surge
heave
sway
Figure 8.3 Shipmotioninsixdegreesoffreedom
x
y
z
o
slamming
greenwaterondeck propellerracing
Figure 8.1 Slamming(impactatbow) Figure 8.2 Waterondeckandpropellerracing
39
givenby
( , ) cos( ),
a
x t kx t z z w = - (9.1)
where
a
z isthewave amplitude, k isthewavenumber,and w isthecircularfrequencyofthewave.
This cosine wave is a harmonic wave propagating in the positive x direction. Figure 9.1 shows the
wavecontouratacertainmoment
i
t .Ifweobservethefreesurfaceelevationatacertainposition
j
x ,we
candrawthetimehistory ofthefreesurfaceelevationasshowninFigure 9.2.
Therelationsbetweenthewaveparametersareasfollows
2
k
p
l
= ,where l isthewavelength
2
T
p
w = ,where T isthewaveperiod.
Itfollowsthat
c
k T
w l
= = ,
where c isthewavepropagationspeedorthephasevelocity(Itshouldbepointedoutthatonlythewave
formmoveswiththisphasevelocity,notthe waterparticles).
ThewavecontourdescribedbyEquation(9.1)isacontourofequalpressure(equaltotheatmospheric
pressure). It can be shown that the contour of equal pressure at any depth
0
h is also cosine curves with
waveamplitude
0
0
2
h
kh
a a
e e
p
l
z z
-
-
= ,see Figure 9.3.
Figure 9.3 Wavecontourofequalpressure
x
z
l
a
z
o
c
0
kh
a
e z
-
0
h -
Figure 9.1 Wavecontourattime
i
t Figure 9.2 Freesurfaceelevationatposition
j
x
x
z
l
a
z
o
c
j
x
z
t a
z
T
o
40
Thecontourofequalpressure atdepth
0
h isdescribedby
0
2
( , ) cos( ).
h
a
x t e kx t
p
l
z z w
-
= - (9.2)
Thatmeans,thewaveamplitudeofthecontourofequalpressureat
0
z h = - decaysproportionallyto
0
kh
e
-
.Thepressureat
0
z h = - isgivenby
0
2
0
( ) cos( ).
h
a
p g z gh g e kx t
p
l
r z r r z w
-
= - - = + - (9.3)
ThelasttermoftherighthandsideofEquation(9.3)denotesthedifferenceinpressureduetothewave
elevation z at
0
z h = - which decays proportionally to
0
kh
e
-
. This difference is due to the socalled
Smitheffect(SmithW.E.,1883).
9.3BasicTheoryofIrregularWaves
Oceanwavesarecharacterizedbytheirirregularity,bothintimeandinspace.Mathematically,however,
they canbe describedasarandomorstochasticprocess withstatisticalsteady, orstationarycharacteristic
appearanceunder shortterm statistically stationary conditions. Therefore, probability and statistics theory
canbeusedto analyze thewavecharacteristics.
Accordingtotheprincipleofsuperposition,irregularwavescanbedescribedasalinearsuperposition
of infinite number of simple, regular harmonic wave components having various amplitudes, lengths,
periodsorfrequenciesand directions of propagation.Hence,anirregularwavepropagatinginthepositive
x direction canbewrittenas
1 1
( , ) cos( ),
i ai i i i
i i
x t k x t z z z w e

= =
= = - +

(9.4)
where
ai
z ,
i
k,
i
w,
i
e are the waveamplitude, wave number, wave frequencyand stochastic phase of
the i th wave component. Since
i
e is stochastic, the wave contour
i
z of the i
th
wave component is
stochastic.
If all the regular wave components propagate in the same direction, longcrested waves which are
twodimensional irregular waves, as described by Equation (9.4), are obtained. On the other hand, if the
wave components propagate in different directions, shortcrested waves which are threedimensional
irregularwavesareobtained.Thelateristhemoregeneralcase.
9.4TheoreticalBasicofSpectrumAnalysis
Theenergyinatrainofregularwavesconsists ofkineticenergyassociatedwith the orbital motion of
water particles and potential energy resulted from the change of water level in wave hollows and crests.
Under one wave length l, the kinetic energy and the potential energy per unit breadth of a wave as
described byEquation(9.1)aregivenby
2
1
,
4
k a
E g r z l =
2
1
4
p a
E g r z l = .
Thetotalaverageenergyperunitareaof free surfaceis
2
1
2
k p
a
E E
E g r z
l
+
= = .
Since ocean waves can be regarded as the linear superposition of simple, regular harmonic wave
components, the energy of ocean waves can be obtained from the summation of energy of the wave
41
components. The total average energy per unit area of free surface for wave components of frequencies
( ,
i i
w w w + D )isgivenby
2 2
1 1
( )
2 2
i i
i i
ai ai
E g g
w w w w
w w
r z r z
+D +D
= =

.
Denotingthisenergyby ( )
i
gS
z
r w w D ,where S
z
isthesocalledwavespectrum,wehave
2
1
2
( )
i
i
ai
i
S
w w
w
z
z
w
w
+D
=
D
.
Here ( )
i
S
z
w denotes the wave energy density at
i
w. S
z
shows the distribution of energy of the
irregularwavesamongthedifferentregularcomponentshavingdifferentfrequencies.
As 0 w D , ( ,
i i
w w w + D )
i
w Therefore,as 0 w D ,
2
1
( )
2
i ai
S d
z
w w z = .
The wave spectrum shows clearly which wave components composing the irregular ocean waves are
important,seeFigure9.4.
Itcanbeshownthatthewavespectrumandthevarianceofthewavecontourhave asimplerelationship.
The wave contour of irregular waves is given by Equation (9.4). At the origin 0 x = , we have the free
surfaceelevationchangingwithtime t asfollows
1 1
(0, ) cos( )
i ai i i
i i
t t z z z w e

= =
= = -

(9.5)
where
i
e is stochastic phase and can take any value among (0, 2 ) p hence, the probability density
functionfor
i
e is
1
( )
2
i
f e
p
= .
Theaveragevalue(expectation)ofthefreesurfaceelevation is
( )
2 2
1 1 1
0 0
1
( ) cos( ) ( ) cos( ) 0.
2
i ai i i i i ai i i i
i i i
M M t f d t d
p p
z z z w e e e z w e e
p

= = =
= = - = - =

Figure 9.4 S
z
versus w curve
S
z
w
o
i
w
( )
i
S
z
w
42
Thevariance,
2
z
s ,ofthe freesurfaceelevationis
( ) [ ]
2
2
2
1 1
0
2 2
2 2
0
1 1 1
0 0
( ) cos( ) ( ) ( )
1
cos ( ) ( ) ( ) .
2 2
i ai i i i i i
i i
ai
ai i i i i
i i i
D D t M f d
t d S S d
p
z
p
z w z
z s z z w e z e e
z
z w e e w w w w
p

= =

D
= = =
= = = - -
= - = = D =

That means, the area under the curve ( ) ~ S
z
w w is equal to
2
z
s , the variance of the free surface
elevation.
Ingeneral,wecandefinethen
th
ordermomentoftheareaunderthe curve ( ) ~ S
z
w w withrespectto
theverticalaxisat 0 w = asfollows
0
( ) .
n
n
m S d
z z
w w w

=
(9.6)
For 0 n = ,wehave
0 2
0
0
( ) m S d
z z z
w w w s

= =
(varianceof z )
For 2 n = ,wehave
2 2
2
0
( ) m S d
z z z
w w w s

= =

&
(varianceof z
&
)
For 4 n = ,wehave
4 2
4
0
( ) m S d
z z z
w w w s

= =

&&
(varianceof z
&&
).
Thevariance
2
z
s can beusedtoestimatethestatisticalcharacteristicsofthefreesurfaceelevation,for
example,theaveragewaveamplitude 1.25
a z
z s = thesignificantwaveamplitude
1
3
2.00
a
z
z s = .
9.5ResponseRelationshipsofLinearSystem
In this subsection we will deal with the relationship between the wave spectrum and the response
spectrum of ship motions in irregular waves. This relationship can be used to predict the statistical
characteristicsofshipmotionsinoceanwaves.Undertheassumptionthattheshipintheoceanwavesisa
linear system which converts the energy from ocean waves into the energy of ship motions, the response
spectrumofshipmotionscanbederivedfromthewavespectrum.
Ship motions in ocean waves are induced by the ocean waves. Regarding the ship as a linear system,
theinputtothesystemistheoceanwavesandtheoutputofthesystemistheshipmotions,seeFigure9.5.
Figure 9.5 Shipinwavesasaenergytransferring system
Output Input
Ship
( ) t z ( )
j
t x
Oceanwaves Shipmotions
43
Assuming that the wave amplitudes and the amplitudes of ship motions are small, the principle of
superposition holds and the induced irregular ship motions can be regarded as a linear superposition of
infinite number of simple, regular motions, each of which are induced by a simple, regular wave
component.
Input: cos( ),
i ai i i i
k x t z z w e = - +
Output: [ ] cos( ),
i i ai i i i i
L k x t x z x w e e = = - + + D
where L isthelineartransferoperator,
ai
x istheamplitudeofshipmotion(itcanbesurge,sway,heave,
roll,pitchoryawmotion),
i
e D isthephasedifference.
Denotingtheratioofamplitudesas
( ) .
ai
i
ai
Y
xz
x
w
z
= (9.7)
Thisratioiscalledthetransferfunction,orthe ResponseAmplitudeOperator(RAO).
Theshipmotion isobtainedbylinearsuperposition
1 1 1
( ) cos( ) ( ) cos( ).
i ai i i i i i ai i i i i
i i i
t k x t Y k x t
xz
x x x w e e w z w e e

= = =
= = - + + D = - + + D

(9.8)
Similartothedefinitionofwavespectrum,wecandefinethe response spectrumofshipmotionsas
2
2 2 2 2
1 1 1
( ) ( ) ( ) ( ) ( ) .
2 2 2
i ai i ai i ai i i
S d Y Y Y S d
xz xz xz xz z
w w x w z w z w w w = = = =

Itfollowsthat
2
( ) ( ) ( ). S Y S
xz xz z
w w w = (9.9)
Equation(9.9)expressestherelationshipbetweenthewavespectrumandtheresponsespectrum.Itcan
beusedtosolvethefollowingthreekindsofproblems.
(1)Thewavespectrumandtheresponseamplitudeoperatorareknown,theresponsespectrumistobe
determinedinordertopredictthestatisticcharacteristicsofshipmotions.Mostproblemsareofthis
kind.
(2)Thewavespectrumandtheresponsespectrumareknown,theresponseamplitudeoperatoristobe
determined.Modeltestsareusuallyconducted inmodelbasinforthispurpose.
(3)Theresponsespectrumandtheresponseamplitudeoperatorareknownforashiptravelingatseas,
thewavespectrumoftheseasistobedetermined.
CHAPTER10:ROLLMOTION4hours
Demand:Masterthebasictheoryandmethodforlinearrollmotionmasterthemethodsofrolltestswith
modelinregularwaves,tobeableto determinetheresponseamplitudeoperator forroll motion from
the tests know and master the methods of model tests and empirical formulae for determining the
hydrodynamic coefficients of roll motion know the estimation methods for roll amplitude know the
variouskindsofrollcontroldevicesandunderstandtheprinciplesofrollcontrolwiththesedevices.
Ship motions in six degrees of freedom can be classified into two categories, i.e., the longitudinal
motions in the vertical plane and the transverse motions in the horizontal plane. Surge, heave and pitch
belongtothefirstcategory,whereassway,rollandyawbelongtothesecondcategory.Amongthesemodes
44
of motion, roll, heave and pitch are most likely to be of large amplitudes due to the existence of the
restoring force and moment, and hence are most important for seakeeping performance. In the present
chapterwewilldealwiththerollmotion,whereasinthenextchapterwewilldealwiththeheaveandpitch
motions.
10.1Linear Theoryof Roll Motion
We start the analysis from the roll motion in calm water. A ship floating freely on the free surface is
considered.Theshipisinitiallyforcedtoheelwithanangle
0
q ,andthenlettorollfreely.
10.1.1Rollmotionincalmwaterwithoutdamping
First,weconsiderthesimplestcaseofrollmotionincalmwaterwithoutdamping.Themomentsacting
ontheshipincludetherestoringmoment
( ) M Dh q q = -
andtheinertialmoment
6
1
( ) ( )
x x x
M I I I q q q = - + D = -
&& && &&
where D is the displacement, h is the transverse metacentric height ( GM), q is the rolling angle
1 x x x
I I I = + D ,
x
I is the mass moment of inertia about x axis,
x
I D is the added mass moment of
inertiaseeFigure10.1.
Theequationofrollmotion is
1
0,
x
I Dh q q - - =
&&
or
1
0.
x
Dh
I
q q + =
&&
Since
1
0
x
Dh
I
> ,wedenoteitas
2
1 x
Dh
n
I
q
= andobtain
2
0. n
q
q q + =
&&
Thisisanordinarydifferentialequationofsecondorder,whosesolutionisgivenby
6
According toNewtonianlawofmotion, .
x x
I I Dh q q q = -D -
&& &&
Herewedenotetheterm ( )
x x
I I q - + D
&&
asthe
inertialmoment.
Figure 10.1 Shipinrollmotion
G
q
M Dhq = -
B
h
45
1 2
( ) cos sin t c n t c n t
q q
q = + ,
where
1
c and
2
c aretheintegralconstantstobedeterminedbythe initialcondition.
Hereweusetheinitialcondition
0
q q = and 0 q =
&
at 0 t = ,andobtain
1 0
c q = ,
2
0 c = .Hence
0
( ) cos . t n t
q
q q = (10.1)
Equation (10.1) shows that the roll motion in calm water without roll damping is harmonic with a
constantamplitude equalto
0
q and anaturalfrequency n
q
oranaturalperiod T
q
:
1
x
Dh
n
I
q
=
1
2
2 .
x
I
T
n Dh
q
q
p
p = = (10.2)
10.1.2 Rollmotionincalmwaterwith damping
Actually, there exist damping forces against the roll motion. Here we consider only the linear
component of roll damping moment under the assumption of small roll motion. The linear roll damping
momentisexpressedby
( ) 2 , M N
q
q q = -
& &
where 0 N
q
> isthedampingcoefficient.Inthiscasetheequationofrollmotionbecomes
1
2 0,
x
I N Dh
q
q q q - - - =
&& &
or
1 1
2
0
x x
N Dh
I I
q
q q q + + =
&& &
.
Denoting
1 x
N
I
q
q
n = and
2
1 x
Dh
n
I
q
= ,weobtain theequationofrollmotion
2
2 0 n
q q
q n q q + + =
&& &
. (10.3)
The solutionofEquation(10.3)isgivenby
1 1 2 1
( ) ( cos sin )
t
t e c n t c n t
q
n
q q
q
-
= + , (10.4)
where
1 x
N
I
q
q
n = ,
2 2
1
n n
q q q
n = - , (10.5)
1
c and
2
c aretheintegralconstantstobe determinedbytheinitialcondition.Fromtheinitialcondition
0
q q = and 0 q =
&
at 0 t = ,wehave
1 0
c q = ,
2 0
1
. c
n
q
q
n
q =
Therefore,from Equation(10.4)weobtain
0 1 0 1
1
( ) ( cos sin )
t
t e n t n t
n
q
n q
q q
q
n
q q q
-
= + . (10.6)
46
Denoting
0
cos
m q
q q e = ,
0
1
sin
m
n
q
q
q
n
q q e = ,wecanrewriteEquation(10.6)as
1
( ) cos( )
t
m
t e n t
q
n
q q
q q e
-
= - , (10.7)
where
2
0 0
1 1
1
m
n
n n
q q
q q
n
q q q

= + =

and
1
arctan
n
q
q
q
n
e

=

.
From Equation (10.7) it can be seen that theamplitude of roll motion is
t
m
e
q
n
q
-
which decays with
.
t
e
q
n -
The frequency of roll motion is
2 2
1
n n n
q q q q
n = - < accordingly, the period of roll motion
1
T T
q q
> . Thephasedifferenceofrollmotionis
q
e .
As , t 0,
t
e
q
n -
resultingin 0. q Forabetter motion performancethe rollangleshould
decaytozeroasfastaspossible.Therefore,
q
n shouldbeaslargeaspossible.Since
1
/
x
N I
q q
n = ,large
N
q
isfavorable forgoodseakeepingperformance.
10.1.3Rollmotioninregularwaves
Now we proceed to analysis of linear roll motion in regular waves. We consider only the case of
quarteringseas,sincerollmotionsareworstformostlargeshipswhentheyaretravelinginquarteringseas.
Thekeypointintheanalysisistheappropriateexpressionofthedisturbancemomentduetothewaves.
Weassumethattherollangleissmall,sothatthelineartheoryofrollmotionholds.
AsshowninFigure 10.3,thewavecontourisdescribedby
( , ) cos( )
a
y t ky t z z w = - .
z
o
m
a
z
y
a
( )
m m
M Dh a a =
Figure 10.2 Rollmotionincalmwaterwithdamping
t
m
e
q
n
q
- q
t
o
0
q
0
q -
Figure 10.3 Shipinregularwaves
47
Theslopeofwavecontourattheorigin 0 y = isgivenby
0
tan sin( ) sin
a a
y
d
k t k t
dy
z
a z w z w
=
= = - - = .
Hence,assumingthefreesurfaceelevationissmall,wehave
0
tan sin sin
a
k t t a a z w a w = = ,
where
0 a
k a z = isthe amplitudeofthewaveslope.
Similar to the moment due to the restoring force, we assume that the disturbance moment due to the
wavescanbeexpressedinthe form
( )
m m
M Dh a a = ,
where
m
a istheeffectivewaveslopewhichisdefinedas
0
sin
m m
t a a w = , with
0 0 m
K
q
a a =
where K
q
isacoefficientdependingontheshipform.
Then weobtain theequationofrollmotioninregularwavesintheform
1
2 0,
x m
I N Dh Dh
q
q q q a - - - + =
&& &
(10.8)
wherethesign ofthewave disturbancemomentis positive, becauseitis the momentthat makestheship
rolling.
Equation (10.8)canberewrittenas
1 0
2 sin
x m
I N Dh Dh t
q
q q q a w + + =
&& &
.
Usingthedefinitionsof
q
n and n
q
,wehave
2 2
0
2 sin
m
n n t
q q q
q n q q a w + + =
&& &
. (10.9)
Equation(10.9)isanordinarydifferentialequationofsecondorderwithnonzerorighthandside.This
equationhasthesolutionintheform
*
( )
c
t q q q = + ,
where
c
q is the solution of Equation (10.3) given by Equation (10.7) and
*
q is a special solution of
Equation(10.9)whichhasthefollowingform
*
( ) sin( )
a
t t
qa
q q w e = + , (10.10)
where
a
q istheamplitudeofroll motioninducedbywaves, w isthefrequencyofinducedroll motion
(equaltothe wavefrequency),
qa
e isthe phase differenceoftheroll motionrelativetothatofthewave
slope.
As , t wehave 0
c
q and
*
q q .Therefore,as , t weobtain
*
( ) ( ) sin( )
a
t t t
qa
q q q w e = + . (10.11)
FromEquation(10.10)wehave
*
( ) cos( )
a
t t
qa
q wq w e = +
&
and
* 2
( ) sin( ).
a
t t
qa
q w q w e = - +
&&
(10.12)
Since Equation (10.10) is a special solution of Equation (10.9), substituting Equations (10.10) and
48
(10.12)intoEquation(10.9),andcomparingthecoefficientsof sin t w and cos t w onbothsidesofthe
equation,we have
2 2 2
0
2 2
[ 2 sin ( ) cos ] ,
2 cos ( )sin 0.
a m
n n
n
q qa q qa q
q qa q qa
q n w e w e a
n w e w e
- + - =
+ - =
(10.13)
FromthesecondequationofEquation(10.13)wehave
2 2
2
tan
n
q
qa
q
n w
e
w
-
=
-
, or
2 2
2
arctan
n
q
qa
q
n w
e
w
-
=

-

(10.14)
2 2 2 2
2
sin
( 2 ) ( ) n
q
qa
q q
n w
e
n w w
-
=
- + -
2 2
2 2 2 2
cos .
( 2 ) ( )
n
n
q
qa
q q
w
e
n w w
-
=
- + -
(10.15)
FromthefirstequationofEquation(10.13)wehave
2 2
0 0
2 2
2 2 2 2
.
( ) cos 2 sin
( ) ( 2 )
m m
a
n n
n
n
q q
q qa q qa
q q
a a
q
w e n w e
w n w
= =
- -
- + -
(10.16)
Denoting
n
q
q
w
L = and ,
n
q
q
q
n
m = where
q
L isknownastuningfactor,wehave
0
2 2 2 2
(1 ) 4
m
a
q q q
a
q
m
=
- L + L
2
2
arctan
1
q q
qa
q
m
e
- L
=

- L

. (10.17)
Wedenotetheratio
a
q to
0 m
a as K
qa
,thatis
2 2 2 2
0
1
(1 ) 4
a
m
K
qa
q q q
q
a
m
= =
- L + L
, (10.18)
K
qa
iscalledmagnificationfactor.
InthefollowingwewilldiscusssomespecialcasesaccordingtoEquations(10.17)and (10.18).
(1) As 0
T
n T
q
q
q
w
L = = (i.e., n
q
w << or T T
q
>> ),wehave 1 K
qa
and 0
qa
e .
Inthis case,
0 a m
q a .Theshiphasarollamplitudeequaltotheamplitude ofeffective waveslope,
whereas the phase difference is zero. This casecorresponds to a small ship traveling on large waves, see
Figure 10.4.
(2) As
T
n T
q
q
q
w
L = = (i.e., n
q
w >> or T T
q
<< ),wehave 0 K
qa
, 180
qa
e -
o
.
Inthiscase, 0
a
q .Theshiphasarollangleapproximatelyequaltozero.Thiscasecorrespondstoa
bigshiptravelingonsmallwaves.Theship respondshardlytowaves,seeFigure10.5.
(3) As 1
T
n T
q
q
q
w
L = = (i.e., n
q
w or T T
q
),wehave
1
2
K
qa
q
m
, 90
qa
e -
o
.
Inthiscase,
0
2
m
a
q
a
q
m
.Since
q
m isusuallyverysmall,
a
q willbeverylarge.Thiscorrespondsto
49
the case of resonance. Therefore, as 1
q
L ( n
q
w or T T
q
), the socalled resonance will occur.
Thissituationshouldbeavoided.
Actually,therollamplitudeisverylargenotonlywhen 1
q
L intherangeof 0.7 1.3
q
< L < ,i.e.,
in the socalled resonance zone, the roll amplitude is also large. Hence it should be avoided that a ship
travelsintherangeof 0.7 1.3
q
< L < .
FromEquation(10.18)itcanbeseenthatinordertoreducetherollresponsetowavedisturbance,
q
m
should be aslargeaspossible.Since
1
1 1 1
x
x x x
N N I N
n n I I Dh DhI
q q q q
q
q q
n
m = = = = ,
it is obvious that large N
q
, small h and small
1 x
I will result in large
q
m , and hence small roll
amplitudeinwaves.
Small h will not only result in large
q
m , but also large natural period T
q
. Therefore, the
favorablenessofsmall h forseakeepingperformanceisdoublefold.Thatmeans,fromthepointofview
of seakeeping, h should be as small as possible. However, from the point of view of ship hydrostatic
performance,transversestability maybedamnified, if h istoosmall. Therefore, h shouldbeproperly
chosenby makinga compromise betweenseakeeping performanceandhydrostatictransverse stability. In
principle, h shouldbeassmallaspossibleunderthepreconditionof an enoughtransversestability.
10.1.4 Rollmotioninirregularwaves
Further we cananalyzethe statistical characteristics of linear roll motion of ships in irregular waves.
Thetransferfunction,ortheresponse amplitudeoperator(RAO),forrollmotionis
( )
a
a
Y
qz
q
w
z
= .
Since
0 a
k a z = ,
0 0 m
K
q
a a = and
0 a m
K
qa
q a = ,wehave
0
0
( )
a
m
a
k
Y K k K K
qz qa qa q
q
w a
z a
= = = . (10.19)
Fordeepwater
2
k
g
w
= ,wehave
2
( ) Y K K
g
qz qa q
w
w = . (10.20)
Figure 10.4 Smallshiponlargewaves Figure 10.5 Largeshipon smallwaves
or n
q
w <<
T T
q
>>
or n
q
w >>
T T
q
<<
50
Equations (10.19) and (10.20) show clearly the relationship between the response amplitude operator
(RAO)andthemagnificationfactor.
If we know the wave spectrum ( ) S
z
w , once we have determined the transfer function ( ) Y
qz
w , we
canobtaintheresponsespectrumofrollmotionaccordingtoEquation(9.9)andthenpredictthestatistical
characteristicsofshiprollmotioninocean waves.
10.2Roll Test
Model testisanimportant methodforinvestigatingseakeeping performance.Sincethehydrodynamic
phenomena in roll motion are too complicated, especially when nonlinear effects are dominating, pure
theoreticalornumericalmethodsusuallycannotgiveresultsaccurateenough,andmodeltestisregardedas
themostreliablemethod topredictrollmotion.
Rolltestswithscalemodelsareusuallycarriedoutin towingtankorseakeepingmodelbasin.Thereare
mainlythreekindsofrolltests
(1) Roll test in calm water.The purpose is to determine the natural period, as well as hydrodynamic
coefficients,especiallythedampingcoefficients.
(2) Rolltestinregularwaves.Thepurposeistodeterminethetransferfunctionofrollmotion.
(3) Roll test in irregular waves. The purpose is to determine the transfer function of roll motion or
wavespectrum.
10.3DeterminingtheHydrodynamicCoefficientsof Roll Motion
Hydrodynamiccoefficientsofrollmotioncanbedeterminedbyempiricalformulaeorbymodeltestsin
model basin. In the following the method of model test for determining the parameters and damping
coefficientsofrollmotionwillbebrieflyintroduced.
Rolltestincalmwatertodeterminethenaturalperiodandthedampingcoefficients:
Rolltestwithscale modelintowingtank or modelbasinisconductedincalmwater.Therollangleis
recorded,asshowninFigure10.6.
Fromtherecordtheamplitude
i
q andtherollperiod
1
T
q
aremeasured.
1
T
q
isdefined as
1
2 2 2 2
1
2 2 2
1 1
T
T
n
n n
q
q
q
q q q q q
p p p
n m m
= = = =
- - -
.
Since
q
m is very small,
1
T
q
is approximately equal to T
q
. The natural period T
q
is an important
parameterevaluatingtheseakeepingperformance.
Denoting
1 i i i
q q q
+
D = - ,
1
( ) / 2
mi i i
q q q
+
= + ,wecandrawthe q D versus
m
q curveasshownin
Figure 10.6 Timehistoryof the rollangle inrolltest
i
q
q
o
t
0
q
1 i
q
+
1
T
q
51
Figure 10.7. Weexpressthisrelationshipas
2
m m
a b q q q D = + , (10.21)
wherethecoefficients a and b canbedeterminedbycurvefittingtechnique.
CorrespondingtotheexpressionofEquation(10.21),thedampingmomentcanbeexpressedas
( ) 2 , M N W
q
q q q q = - -
& & & &
(10.22)
where N
q
and W are the coefficients of linear and nonlinear components of damping moment,
respectively.
Accordingto the relationbetweentheworkdonebythedampingmomentandthechangeof rollmotion
energy, we can prove that there exist relations between the coefficients a , b and the coefficients N
q
,
W asfollows
2
2
Dh
N a
n
q
q
p
= and
2
3
4
Dh
W b
n
q
=
orinreverse,
1
2
x
n T
a N N
Dh I
q q
q q
p
= = and
2
1
4 4
.
3 3
x
n
b W W
Dh I
q
= =
Fromthese relationswecandetermine N
q
and W from a and b.
10.4Estimationof Roll Amplitude
Asrollamplitudeincreases,thenonlinearcomponentofdampingmomentbecomesdominating.Inthis
casethedampingmomentcanbeexpressedas
( ) M W q q q = -
& & &
.
Correspondingly,the q D versus
m
q curvecanbeexpressed intheform
2
m
B q q D = .
Letting the work done by the nonlinear damping moment equal to that done by an equivalent linear
dampingmoment,wecanderivethe equivalentlineardamping coefficientas
1
8 2
2
3
m m x
N W I B
q
q w q w
p p
= = .
Figure 10.7 q D versus
m
q curve
q D
m
q
o
52
Accordingly,
1
1 1
2 2
2 2 2 .
m x
m
x x
N I
B B
n n I n I n
q q
q
q q q q
n q w w
m q
p p
= = = =
When resonanceoccurs,wehave 1
n
q
w
= .Thenthemaximalrollamplitude canbeestimatedas
0 0 0
2 2
2
m m m
a
m a
B B
q
a a a
q
m
q q
p p
= = , or
0
2
m
a
B
pa
q = .
10.5Roll Control Devices
For motions witheffects ofrestoringforces or moments, suchasheave,rolland pitch,resonancemay
occur when the natural frequency equal to the wave frequency. At resonant frequency the magnification
factor for roll motion can be very large due to very small roll damping. This implies severe roll motion
whichhas great negative effects onseakeepingperformances. For example, large roll motions maycause
thegoodsshifting,resultinginlossoftransversestabilityoritmaydeteriorate theworkingconditionofthe
crew.Therefore,itisnecessaryforustotake measurestoreduce ortocontroltheroll motionresponseto
wavedisturbance.Forthispurposeantirollingdevices,orthesocalledrollstabilizers,canbeused.Inthe
followingtheantirollingdevicesincommonusageareintroduced.
Antirolling devices can be classified into three groups, the passive, the controlledpassive and the
activeones.
The most commonlyused passive antirolling device is bilge keel, which is mounted on the ship hull
near the bilge and is used to increase the roll damping, see Figure 10.8. Bilge keel is the simplest
antirollingdevice andmostofseagoingshipsare equipped withit.
The passive antirolling tank is another one of passive antirolling devices. This includes free surface
tankandUtubetank,seeFigure10.9.Withcarefuldesignthefluidmotioninthesetankscausesamoment
which counteractspartofthedisturbancemomentbywaves.
(a) Free surfacetank (b) Utube tank
Figure 10.9 Passiveantirollingtanks
Figure 10.8 Bilgekeelaspassiveantirollingdevice
bilge keel
53
The controlledpassive roll tank is a controlledpassive antirolling devices where the fluid motion in
thetankiscontrolledinsomemanner.
The most popular active antirolling deviceis thestabilizingfins which are mounted on bothsides of
ship hull,seeFigure10.10.This kind ofantirolling deviceis generallyusedon warshipsorluxuryliners
whichhavespecialrequirementsofrollperformance.Whenthefinsaresetatoppositeanglesofattack,the
liftscausedbytheflowsoverthemformamomentaboutthelongitudinalaxis.Theanglesofattackcanbe
controlledtocreatea momentagainsttheroll motion.When notbeingusedtoreduceroll motionthe fins
areusuallyretractedto avoid thedragactingonthefins.
Theactivetankstabilizerisanotherkindofactiveantirollingdevice.Thisdeviceusesaflowpumpto
force the water in the tank to flow from one side of the ship to another sideto create actively a moment
againstrollmotion.
CHAPTER11:PITCHANDHEAVEMOTIONS4hours
Demand:Knowandmasterthetheoreticalandnumericalmethodsforanalysisofshipmotionsknowthe
method of spectrum analysis and the methods of model test for determining the transfer function for
pitch and heave motions know the various kinds of derived responses due to ship motions know the
basicconceptsofshipmotioninobliquewaves.
Pitch and heave motions are longitudinal motions in vertical plane. These motions are affected by
restoringforceandmoment,andmayrespondwithlargeamplitudeswhenresonanceoccurs.Althoughthe
responsesofpitchandheavemotionsarenotsoseriousasthatofrollmotion,thederivedresponsesdueto
thesemotionshaveremarkablenegativeeffectsontheseakeepingperformance.
11.1 TheoreticalComputationforSeakeepingProblems
11.1.1Briefintroduction
Forthelongitudinalmotionstheviscouseffectsareusuallysmallandcanbeneglected.Forthisreason,
hydrodynamic characteristics of pitch and heave motions are usually investigated by theoretical and
numericalmethodsbasedonpotentialtheory.
Traditionaltheoreticaland numerical methodsincludethestriptheoryandslenderbodytheory.Based
ontheassumptionthattheshipisslender,i.e.,thelongitudinal dimension oftheship is much largerthan
the transverse and vertical dimensions, the threedimensional flow around the ship is transformed into
twodimensionalflowaroundthesectionsoftheship,whichsimplifiestheproblemgreatly.
More advanced theoretical and numerical method is the threedimensional method such as Boundary
ElementMethod(BEM),orthesocalledpanelmethod,whichisfirstlyappliedforlinearmotionsofsmall
amplitude with linearized boundary conditions on the free surface and on the mean wetted body surface,
then for nonlinear motion of large amplitude with nonlinear boundary conditions on the free surface and
body surface. The unsteady problem of ship motion is traditionally solved by using threedimensional
Figure 10.10 Stabilizingfinsasactiveantirollingdevice
stabilizingfin
54
method in frequency domain more actually it is solved in time domain, allowing to predict the
instantaneousshipmotionwithlarge amplitudes.
11.1.2Coordinatesystems
Ashiptravelingin forward motionwithaconstantspeed U atseasisconsidered. At thesame time
the ship undergoes an oscillating motion of small amplitude in six degrees of freedom about its mean
position.
The jth mode of motion is denoted by
j
x , where 1, 2, 3 j = correspond to translating motions
surge, sway and heave, respectively and 4, 5, 6 j = correspond to rotating motions roll, pitch and yaw,
respectively.
Forseakeepingproblem,threerighthandedcoordinatesystemsareadopted,seeFigure 11.1:
(1) Earthfixedcoordinatesystem
0 0 0 0
o x y z - .Aframeof referencefixedontheearth,with
0
x axis
pointingtothedirectionofshipforwardmotionand
0
z axispointingupwards.
(2) Movingcoordinate system o x y z - . Aframe of reference moving with thespeed U . At the
initialmomentthemovingcoordinate systemcoincideswiththeearthfixedcoordinatesystem.
(3) Bodyfixed coordinate system o xyz - . A frame of reference fixed on the ship and moving
together with the ship. This coordinatesystem has oscillating motions in six degrees of freedom
relativetothemovingcoordinatesystem.
11.1.3Forcesactingontheship
Here we will only consider the cases in linear theory, where the amplitudes of waves and the ship
motionsareassumedtobesmall.
Theforcesactingonaship traveling atseasinclude:
(1)Hydrostaticrestoringforces
The i th componentofhydrostaticrestoringforces(nethydrostatic forces)isgivenby
6
1
Ci ij j
j
F C x
=
= -
, ( 1, 2, , 6 i = L ) (11.1)
where
j
x is the jth mode of motion
ij
C is the hydrostatic restoring force coefficient which
corresponds to the i th component of hydrostatic forces induced by the unit jth mode of motion the
minussigninEquation(11.1)meanstheforceisagainstthemotion.
yaw
pitch
roll
surge
heave
sway
x
y
z
o
Figure 11.1 Coordinatesystemsusedinshipmotionproblem
x
0
z
0
x
0
o
z
o
Ut
U
55
Assumingthattheshipissymmetricalaboutitslongitudinalcenterplane,thehydrostaticrestoringforce
coefficient
ij
C isgivenby
( )
( )
22
11
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0
0 0 0 0 0
0 0 0 0
0 0 0 0 0 0
w w f
ij
B G
w f B G
gA gA x
C
g S z mgz
gA x g S z mgz
r r
r
r r

-
=

+ -

- + -

,
where r isthemassdensityoffluid g isthegravitationalacceleration
w
A istheareaofwaterplane
f
x isthe longitudinalcoordinateofthecentreofflotation isthedisplacement m isthemassofthe
ship
B
z and
G
z are the vertical coordinates of the centre of buoyancy and the centre of gravity,
respectively.
11
S and
22
S are givenby
2 2
11
2
w
A L
S x dxdy yx dx = =

2 3
22
2
3
w
A L
S y dxdy y dx = =

,
where ( ) 0 y x > isthesemibreadthofwaterplane.
It should be noted that the matrix of
ij
C is symmetrical, which means
ij ji
C C = .
ij
C has to be
positive toensuretheshiphashydrostatic stability.
(2)Hydrodynamicforces
Hydrodynamicforcesarecomposedoftwoparts:theradiationforcesduetotheoscillatingmotionsof
theshipandthewaveexcitingforcesduetothediffractedwavesandtheincidentwaves:
The i th componentofradiationforcesisgivenby
6
1
( )
Ri ij j ij j
j
F A B x x
=
= - +

&& &
, ( 1, 2, , 6 i = L ) (11.2)
where
ij
A istheaddedmasscoefficient,
ij
B isthedampingcoefficient.
The i th componentofwaveexcitingforcesis
D I
Ei i i
F F F = + , ( 1, 2, , 6 i = L ) (11.3)
where
D
i
F is the diffraction exciting forces due to the diffracted waves and
I
i
F is the socalled
FroudeKrylovexcitingforcesinducedby theincidentwaves
7
.
11.1.4Linearizedequationsof shipmotion
FromtheNewtonianlawofmotionweobtainthelinearizedequationsof shipmotionasfollows
7
Theconceptof approximate excitingforcesduetoincidentwaveswasfirstintroducedintheclassicalworkonshiproll
motiondoneby Froudein1861andgeneralizedtoshipmotionsinsixdegreesoffreedombyKrylovin1898.
56
6
1
D I
ij j Ci Ri i i
j
M F F F F x
=
= + + +

&&
, ( 1, 2, , 6 i = L ) (11.4)
where
ij
M isthemassmatrix(inertiamatrix)givenby
0 0 0 0
0 0 0
0 0 0 0
0 0 0
0 0 0
0 0 0
G
G G
G
ij
G x xz
G G y
G xz z
m mz
m mz mx
m mx
M
mz I I
mz mx I
mx I I

-

-
=

- -

-

-

.
Itcanbeseenthat
ij
M isasymmetricmatrix,i.e.,
ij ji
M M = .
Substitutingtheexpressionsofforces,Equations(11.1)~(11.3),intotheequationsofmotion,Equation
(11.4),weobtain
6
1
[( ) ]
D I
ij ij j ij j ij j i i
j
M A B C F F x x x
=
+ + + = +

&& &
, ( 1, 2, , 6 i = L ) (11.5)
ij
M ,
ij
C and
I
i
F can be calculated directly, whereas the added mass coefficient
ij
A and the
damping coefficient
ij
B are to be determined by solving the radiation problem the diffraction exciting
forces
D
i
F aretobedeterminedbysolvingthediffraction problem.
11.2PitchandHeaveinIrregularWaves
Thestatisticalcharacteristics of pitchandheave motionresponsestoirregularwavescanbeestimated
by spectrum analysis. The key point is to determine the response spectrums or the response amplitude
operators(RAO).
Forpitchmotion:
( )
a
a
Y
fz
f
w
z
= ,
2
( ) ( ) ( ) S Y S
fz fz z
w w w =
where
a
f isthepitchamplitude.
Forheavemotion:
( )
a
Z
a
Z
Y
z
w
z
= ,
2
( ) ( ) ( )
Z Z
S Y S
z z z
w w w =
where
a
Z isthepitchamplitude.
( ) Y
fz
w and ( )
Z
Y
z
w can be obtained from model tests or theoretical and numerical methods. After
that,the responsespectrumsofpitchandheavemotionscanbedeterminedandthestatisticalcharacteristics
ofpitchandheavemotionscanbe estimated byspectrumanalysis.
11.3DerivedResponsesduetoShipMotion
57
Derived responses from motions of notable amplitude in six degrees of freedom have important and
negativeeffectsonseakeepingperformance.Theseresponsesaremostlydueto thepitchandheavemotions
and include:
(1) Localmotions,i.e.,verticalorhorizontalmotions,velocitiesand accelerationsatspecialpoints
(2) Relativemotionsbetweenalocationintheshipandtheencounteredwaves
(3) Shippingwaterondeck(greenwaterondeck)
(4) Slamming
(5) Yawingandbroachingto
(6) Addedresistanceandpoweringinwaves
(7) Wavebendingmomentsandloadsonship.
Alltheseresponsesareusuallyrelatedtononlineareffects.
11.4ShipMotioninObliqueWaves
Considering that a ship travels witha constant forwardspeed U at an angle m to regular waves, as
shown in Figure 11.2. 0 m =
o
corresponds to following sea, 90 m =
o
to beam sea and 180 m =
o
to
headsea.
Thewavevelocityrelativetotheshipis
cos
e
C C U m = - .
Thewaveperiodrelativetotheshipis
cos
1 cos 1 cos
e
e
T
C
T
U U
C C U
C C
l
l l
m
m m
= = = =
-
- -
.
e
T iscalledtheperiodofencounter. Thecorrespondingfrequencyofencounterisdefined as
2 2 2 2 2
( cos ) cos cos cos .
e
e
C U C U kU kU
T T
p p p p p
w m m m w m
l l l
= = - = - = - = -
Fordeepwater
2
/ k g w = ,weobtain
2
2
cos
e
e
U
T g
p w
w w m = = - . (11.6)
30
o
quartering
30
o
bow
60
o
head
60
o
following
Figure 11.2 Shiptravelingatanangletowaves
U
o
y
x
C
m
60
o
beam
58
Therefore, if a ship travels with a forward speed U at an angle m to regular waves, the effective
wave frequency is the frequency of encounter. In all formulas containing the wave frequency w, w
should be replaced by the frequency of encounter
e
w . Correspondingly, T should be replaced by
e
T .
Forexample,when rollresonanceoccurs,wehave
e
T T
q
= or
e
n
q
w = . Therefore,when
cos
e
T T
C U
q
l
m
= =
-
or
2
cos
e
U n
g
q
w
w w m = - = , (11.7)
rollresonancewilloccur.
Notingthattherelationsamongthewaveparametersare
C
T
l
= ,
2
T
p
w
= and
2
k
p
l
= C
k
w
=
Indeepwater
2
/ k g w = ,resultingin
g
C
k
w
w
= =
2
2
gT gT
C T l
w p
= = =
2
T
g
p
l = and
2
2
C
g
p
l = .
If a ship travels in deep water on regular waves of wave length l, wave velocity C and wave
frequency w, the ship velocity U and the angle m can be chosen according to Equation (11.7) to
avoidrollresonance.
CHAPTER 12: DESIGN OF SEAKEEPING PERFORMANCES AND FULLSCALE
TRIALS2hours
Demand: Know the various kinds of seakeeping performance indexes to be considered in ship design
knowtheeffectsofshipprincipaldimensionsandshipformonseakeepingperformancestobeableto
applytherelatedknowledgeinshipdesigntoimprovetheseakeepingperformancesknowthemethods
offullscaletrailsofseakeepingandknowhowtoanalyzetheresults.
12.1SeakeepingPerformanceIndex
Seakeepingperformance ofashipdepends notonly onitsformandthe principaldimensions,butalso
on the environment in which it is operated. In general, the methodology for assessing seakeeping
performancedependsonfourfactors:
(1)Mission.Missionsareassignedtotheship,whichmaybesubdividedintothreecategories:
(a)Porttoporttransportationofgoodsorpeople
(b)Militarymissionscarriedoutentirelyatsea
(c)Commercialmissionsatsea,suchasfishing,oildrilling,etc.
(2) Environment. Waves, wind speed, which are defined as a function of both location and time, and
quantifiedbyseastatenumbers.Significantwaveheight,H
1/3
,andmodalwaveperiod,T
m
,arethetwo
parameterscurrentlyusedtocharacterizetheseastate.
(3) Ship responses. Ship motion responses are described as a function of sea state numbers ship
velocityand ship heading angle relative to waves will affect ship motion performance and determine
themaximumattainablespeedinwaves.
(4)Seakeepingperformancecriteria.This isakeyelementin developinga methodology forassessing
ship operational performance at seas and these criteria determine whether or not a mission can be
carriedout.Examplesofseakeepingperformancecriteriaareasfollows:
(a) Absolute motion amplitudes: Roll angle, pitch angle, vertical displacement of points on flight
59
deck,etc
(b)Absolutevelocitiesandaccelerations: Verticalacceleration,lateralacceleration,etc
(c) Motions relative to sea: Frequency of slamming, frequency of emergence of a sonar dome,
frequencyofdeckwetnessprobabilityofpropelleremergence.
Prescribedlimitingvaluesaregivenforthesecriteriatomakecertainiftheseakeepingperformance
isacceptableorunacceptable.
12.2EffectsofShipPrincipal DimensionsandFormonSeakeepingPerformances
Theory shows that the effects of longitudinal motions, i.e., surge, heave and pitch, and transverse
motions,i.e.,sway,rollandyaw,canbeconsideredseparately.Forthisreason,theeffectsofshipprincipal
dimensions and form on the seakeeping performance will be discussed separately for the two kinds of
motion,
12.2.1 Factorsaffectingpitchandheavemotions
Ingeneral,forconventionalmonohullshipsinheadseas,thelongertheship,thelesstheaveragewave
excitationthatmeans,longershiplengthisfavorable.
Asthedraftincreases,slammingisreduced.
With increasing breadthtolength ratio, there is a distinct increase in damping. Therefore, from the
pointofviewofrollmotionresponse,largerbreadthtolengthratioisfavorable.
Withincreasinglengthtodraftratio,therelativebowmotionisreduced.Hencelargerlengthtodraft
ratioisfavorableforgreenwaterondeck.
12.2.2 Factorsaffectingrollmotion
The first consideration in design to reduce roll motion is to introduce artificial damping by using
antirolldevicelike bilgekeelswhicharethemostsimplestandyeteffectiverollcontroldevice.
Sinceresonanceeffectsare veryimportantinrolling,itisdesirableto designfora naturalperiodthat
avoidsresonancezone.
As discussed in Chapter 10, smaller transverse metacentric height is favorable for roll motion
performance.
Besidestheantiroll devicesintroducedinChapter6,rudderscanalsobe usedasrollcontrol devices,
although they are mainly used for changing the course and controlling the heading of the ship. This
applicationhastheadvantageofmakinguseofanexistingsystem,thusreducingthecost.
12.3FullscaleTrialsofSeakeeping
Tobe completed.
60
REFERENCES
1. M.A.Abkowitz,LecturesonShipHydrodynamicsSteeringandManeuverability,ReportNo.Hy5,
HydrodynamicsDepartment,HydroogAerodynamiskLaboratorium,Lyngby,Denmark,May1964
2. M.A.Abkowitz,StabilityandMotionControlofOceanVehicles,Cambridge,Mass.,TheMITPress,
1969
3. J.N.Newman,MarineHydrodynamics,Cambridge,Mass.,TheMITPress,1977
4. Edward V. Lewis (Editor), Principles of Naval Architecture (Second Revision), Vol.III, Motions in
WavesandControllability,PublishedbyTheSNAME,USA,1989
5. International Maritime Organization (IMO), Standards for Ship Manoeuvrability, Resolution
MSC.137(76),December2002.
6. International Maritime Organization (IMO), Explanatory Notes to the Standards for Ship
Manoeuvrability, MSC/Circ.1053,December2002.
7. Sheng Zhengbang, Liu Yingzhong (Editors in chief)Principles of Naval Architecture (Vol.II) (in
Chinese)TheShanghaiJiaoTongUniversityPress,2004.
8. Feng Tiecheng, Zhu Wenwei, Gu Shuhua (Editors), Ship Manoeuvring and Oscillating Motions (in
Chinese), TheNationalDefenseIndustryPress,1989.

Ship Manoeuvring and Seakeeping PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ship Manoeuvring and Seakeeping PDF

Uploaded by

Copyright:

Available Formats

LectureNotes

SHIP MANOEUVRING AND

You might also like