:r*t'-ai,r./J .*t'f '#" ?- T, fi"'"

OCE,di!n,En&3t,waa6Fttntri

Ocean EngineeringCopyright O 2011 Published by Elsevier Ltd. All rights reserved.

Manoeuvring prediction of pusher barge in deep and shallow water

A. Maimun, A. Priyanto, A.H. Muhammad, C.C. Scully, Z.I. AwaIPages l29I-1299 Volume 38, Issues 17-12,Pages 7277-1356 (August2011)

An lnternational Jou rnal

Editor-in-ChiefProfessor Atilla lncecikDepartment of Naval Architecture and Marine EngineeringA joint Department of the Universities ;iGi;g.;

"rd strathcrydeHenry Dyer Building

100, Montrose StreetGlasgow, c4 \LZUnited Kingdcrne-m ail : atilla.incecik@ na-me.ac.uk

EditorsM. E. McCormick (Annapolis, MD, USA)

Emeriti

R. Bhattacharyya (Annapolis, MD, USA)I

I

J 3 P Barltropl&rroarsities of Glasgow & Strathclyde,Eh*er,cw, UK

I EcrnitsasOlmtersity of Michigan, Ann Albor, Ml, USA

l. G- L Borthwicktniversity of Oxford. Oxfor.d, UK

I-lY. Chenl:conal Cheng Kung UniversityfaT an City,Taiwan, ROC

I Chybafiwersity of Hawaiiat Manoa,b.rolulu, Hl, USA

i ClaussEcfrnische UniversitHt Berlin (TUB),lerlin, Germany

t- Collettelniversity of Michigan, Ann Arbor, Ml, USA

Y. Cuihina Ship Scientific Research Centeriangsu, China

. DemirbileklS Army Engineer B&D Centerict<sburg. MS, USA

, Eatock Taylorhiversity of Oxford. Oxford, UK

. L Edgeorur Carolina State University.Jeigh, NC, USA

. E de O. FalcaoiversidadeTdcnica de Lisboa,t6oa, Portugal

trrantde centrale de Nantes,;rtes, France

Associate EditorsO. Gorenlsta nbu I Tech nica I U niversity,lstanbul, Turkey

M.A. GrosenbaughWoods Hole Oceanographic lnstitution,Woods Hole. MA, USA

C. Guedes SoaresU,niversidade T6cnica de Lisboa.Lisboa, Portugal

P J, HudsonUnited States Naw.W. Bethesda, MD, iiSA

R. HuijsmansTechnische Universiteit Delft,Delft,The Netherlands

D.-S. JengUniversity of Dundee.Dundee, UK

J.W. KimTechnip USA, Houston.TX, USA

M. H. KimTexas A&M,College Station, TX, USA

U. A. KordeSouth Dakota School of Mines andTechnology, Rapid City, SD, USA

P LinSichuan Universiry Chengdu,Sichuan, China

A. NaessNorwegian University of Science andlechnology (NTNU), Trondheim, Norway

M. A. S. NevesUniversidade Federal do Hio de Janeiro(UFRJ). Rio de Janeiro, Brazit

J. K. PaikPusan National University, Busan,South Korea

A. PapanikolaouNational Technical U niversity ofAthens (NTUA), Athens, Greece

P T, PedersenDanmarks Tekniske Universitet (DTU),Lyngby, Denmark

J. M. RaceNewcastle uponTyne,Newcastle Univeriity, UK

H. R. BiggsUniversity o{ Hawaii at Manoa,Honolulu, Hl, USA

R. J. SeymourUniversity of California.San Diego (UCSD).La Jolla. CA, USA

V. Sundarlndian Institute of Technology at Madras,Chennai. lndia

H. SuzukiUniversity of Tokyo, Tokyo, Japan

A. W. TroeschU niversity of Michigan,Ann Arbor, Ml, USA

R. W. YeungUniversity of California at Berkeley,Berkeley. CA. USA

S. C.YimOregon State U niversitv.Corvallis, OR. USA

Etr"IiliJr&i;""i,ftif",:J#x]*ii#ffiJ:,'#ffiff1t:tf:1#:xTl;:i,rn":]i:;j,.,"",,"ffitr#i]ttilER:

send change of addru',ito a;eii irgireerng. Ersevier ir.io.", suruicJoepartment,32sl Biverport Lane, Maryrand Heights,EEIGHTAND M'AIL|NG in USA by Mercury lnternational Limited, 365, Blair Road, Avenet. NJ 07001.

Elsevier Ltd.The Boulevard, Langford Lane. Kidlington, Oxford OX5 1cB, U.K.

Elsevier lnc.655 Avenue of the Americas, Newyork, Ny 100i0, U.S.A.

@ 201j Elsevier Limited. All rights ,u""_"j

Ocean [-,ngineering 38 {201 I ) 1291 -1 299

Contents lists available at ScienceDirect

Ocean Engineering

lournal homepage: www.elsevier.com/locate/oceaneng

Manoeuvring prediction of gusher barge in deep and shallow water

A. Maimun*, A. Priyanto, A.H. frrfKmqr-a{, C.C. Scully,Z.l. AwalDepartmenr of Marine Technalow, r*r, i**rir^ rnr,rrrr^f. IJniversiti Tektlologi Malaysia, Skudai 81i10, Mataysia

ARTICLE INFO ABSTRACT

This paper presents an experimental investigation on the manoeuvring characteristics of a pusher-barge system for deep (H/d > 3) and shallow watet (H/d=7.3) condition. Since, the operation of pusher-barge mairrly concentrates on confined waters, there is a ueed to predict and analyze the manoeuvringcharacteristic of the system for a safe and acceptable performance. A time domain simulationprogramme was developed for this purpose. A series of model experiments were carried out todetermine the hydrodynamic coefficients using a planar motion mechanism (PMM). The time domainsimulation shows the manoeuvring characteristic in thc form of turning circle trajectories and zig-zagmanoeuvre based on the hydrodynamic coefficients, which were derived based on experimental results.The manoeuvring characteristics in shallow and deep water conditions were compared through thesimulation results. A comparison of simulation results based on experimental and empirical drivencoefficients for both conditions shows that the experimental coefficients gave better manoeuvringcharacteristics for both turning circle trajectories and zig-zag

B*1?;t;,., Ltd. All righrs reserved.

.4tticle history:Received I6 November 201)9

Received in revised fornt27 October 2010.{ccepted 15 May 20'l 1

Editor-in-Chief: A.l. Incecik.{\,ailable online 25 lune 2011

Leywords:Pusher bargeTurning circleZig-Zag manoeuvringSirallow waterDeep water

1. Introduction

Presently, the most economical means of carrying goods in:nland waterways for Indonesia and Malaysia is through barges.

:{owever, the use of barges in towing mode could affect its safety,:s it has to manoeuvre in rivers of confined waters. In order to:rhance safety there is growing use of pusher-barge systems for:arriage of cargoes, which is an altemative mode of transportation:e inland waterways and coastal regions that offered a minimum:perating cost and safety. This system must have good manoeuvr-rg capabilities to maintain its intended course in inland water-uays, coastal area and in ports. The pusher barge must also be able:: stop within a reasonable distance or turn within the reasonable::ning path in order to avoid some hazardous conditions, such as

::ilision, ramming and grounding.lr,lanoeuvring characteristic of a pusher barge is dependent on

re parameters of the waterways such as bank shape and water:{.th (Lataire et al., 2007). Vantorre and Eloot (1996) compared:.:l:rent formulations of lateral tbrce and yawing moment with:,-cel experiment results for shallow water manoeuvring for all::i angles and found that a tubular formulation of the lateral force

--..j the yawing moment was needed to cover the whole range of:-i angles. According to Beukelman andJournee (2001 ), deduction:i',.;ater depth causes an increase of moment and lateral force,m-.:r will reduce the manoeuvring capability of a vessel.

" - :.:?sponding author. Tel.: +60 75534744i fax: +60 75566159.!-:;ii address: adi@fkm.utm.my (A. Maimun).

:i29-801 8/$ - see front matter o 201 1 Els?vier Ltd. All iights reserved: :::1 0.1 0l 6/j.oceaneng.20l 1 .05.01 1

Since steering and manoeuvring describes the pusher bargemotions on a horizontal plane, atime domain coupled equationmay be developed to describe the motions. The coefficients of thevarious terms in the equation are referred to as the hydrodynamicderivatives. These derivatives are dependent upon the hydrody-namic flow around the ship hull, which in turn depends on thegeometry of the submerged body of the hull (Wang et al., 2000).This research focuses on a simulation programme, which wasdeveloped based on the hydrodynamic coefficients to predict themanoeuvrability of the p[sher barge.

2. Mathematical model

The mathematical model for manoeuvring motion can bedescribed by the following equation of motion, using the coordi-nate system in Fig. 1.

x : + pLz d(m' + m)n -lpL2 d(m' + my)ru

Y:+pL2d(m'+m)n +lpL2d(m'+m.)ru (1)

N:1pL4d(tL+lL)i

where m, mr, and m, are the mass of ship, and added mass inx- and y-directions, respectivelyi lzz and Jo moment of inertia andadd moment of inertia around z-axis, respectively; B is Drift angleat the centre of gravity C.G. [P: -sin-I(r/U)l; / is dimensionlessturning rate l/:t\Llu)l; r and u are the turning rate and sway

A. Maimun et al. I Ocean Engineering 38 (2011) 1291 1299

Fig. 1. Co-ordinate system'

velocity, respectively; and L, d and u are the ship length, ship

draught and ship speed, respectivelyThe superscript {'} in the equations refers to the non-dimen-

sional quantities defined bY:

, m,mx'mvm',m*m'r:ffi, LJ'-:As shown in Fig. 1, (U) is the actual ship velocity that can be

decomposed in an advance vetocity (u) and a transversal velocity(u). The Pusher-Barge has also a rotation velocity with respect to

the z-axis. This axis is normal to the XY plane and passes throughthe Pusher-Barge centre of gravity (C.G). (B) is the angle between

U and the x-axis and it is called drift angle. ( Y ) is the Pusher-Barge

heading angle and (6) is the rudder angle. X,Y and N represents the

hydrodynamic force and moment acting on the mid ship of hull.These forces can be described separating into the followingcomponent from the viewpoint of the physical meaning.

X:xx*Xn*Xp Q)

Y:Yn+Yn+Yp (3)

N = Na*IVn*Np (4)

where, the subscripts H, P and R refer to hull, propeller and rudder,

respectively, according to the concept of MMG expression.

2.1. Forces and moment acting on Hull

Xs, Ys and Ns are approximated by the following polynomials

of f and /. The coefficients of polynomials are called hydrody-namic coefficients.

xH : lp Ldu2 (x'g,r' sin 0 + x',, cos2 fi)

y H = rpLdu2 {Y'110 + Y',r' +Y'u, frlPl +Y;/ lt' | +Y'BpB 03

+Y*r'3 +(Y'p1t,P+Y'p,t')0r') (5)

N H : tpL2 du2 (N's fr + N',r' + N', p {t I A 1 + N;r lr' I

+ N'p p p 03 + N'*rts + (N'p p,B + N'pnr' ) Al )

The calculation method of forces and moment induced bypropeller can be relerred in Appendix A

2.2. Determination of hydrodynamic coefficient

Success of the manceuvring prediction depends heavily on

the knowledge of the hydrodynamic coefficients and the ways

to estimate them. This research involves model testing and

empirical methods to deterrnine the hydrodynanlics coefficients

for the simulation. Hydrodynamic coefficients based on model

testing method are obtained by the mcasurement of forces fromplanaimotion mechanism (PMM). The empirical rnethod is based

on an approximation formulas to determine the hydrodynamic

coefficients. tn this case, the Kijima's formula will be used as a

reference. This formula is obtained semi-empirically from the

results of numerical calculations based on lifting surface theory

and model tests in full load condition. The shallow water

coefficients are obtained by multiplying a correction factor withcoefficients in deep water condition (Maimun et al', 2005)'

Dshw:f (h)Dd"p

where D.6,, is the derivatives in shallow water including ballast

and half toad ccnditions, Da., is derivatives in deep "vater

includ-

ing ballast and half load conditions and flh) is correcting factor

I(jima's approximatiorr formula is furrher explained in Appendices

B and C.

3. Experiment

The experiment was conducted in the 120 m long towing tank

at Marine Technology Laboratory of Universiti Teknologi Malaysia

(UTM). Captive model tests are carried out using a planar motion

mechanism (PMM) to determine the hydrodynamic coefficients ofthe pusher barge.

The model was connected to the PMM by means of one

longitudinal and two transverse force transducers fitted in ball-jointed rods, allowing heave' pitch and rol[' The transducer in

longitudinal direction is more sensitive than the transducers as

appll.a in transverse direction in order to increase the accuracy ofthe measurements.

3.1. Dota collection

In the experiment, the hydrodynamic forces and moments are

measured in both deep water and shallow water. The deep water

condition has a water depth to vessel draught ratio H/d greater

than 3 and in shallow water H/d of 1.3. The principal particulars of

these ship models are listed in Table 1.

Oscillatory model tests results are used to develop the hydro-

dynamic coefficients as input in mathematical modelling' Measure-

ments of hydrodynamic forces and moments are shown with ship

model from Figs. 2-9 as an example. The model rvill experience

oscillatory motion while being towed in the tank at constant speed'

The motion generated by PMM coulci be drift tests, pure sway' pure

yaw, and yaw with drift test. The model is also free to heave and

pit.h, Uut restrained in surge, sway, aiid yaw motions. The input

parameters of these tests are presented in Table 2.

Hull force and moment coefficients: Xs, Y1r and NH are deter-

mined based on measurement by force transducers on the PMM'

Table 1

Principal dimensions of pusher and barge.

Pusher-t arge

Full scale Model Full scale Model Fult scale Model

23582.33420.380.0950.06401

o.723

- laii;rs, 5.

:-E5l::r. a::::f,{'E

rL TLD

lhe-:_---:;il-e ,.es-

::6,a:

Lal (m) 29.76 0.59s2 9s

Lpp (m) 25 0.s 94.s

B (m) 10.208 0.2042 19

d (m) 3.7 0.0740 4.75

volume (m3) 542..147 0.004446 75c2.5cB 0.5746 0.5746 0.8797

1-9 117 .921.89 1 16.71

0.38 190.095 4.750.061s 't805.74

0.8797 0.723

Ecientsmodel

= froms baseriynamicdasa[l thetheorywJter

r vfithI

b.Ilastiodud-.EBT?flIi6es

g tankI{eysiadionbreof

d onein ball-minEts A5

rxlrof

rts are

PEteflltdeflrsof

lt&o-hggre-lb.liPUinCptryeedrr, tr,re

-:Ylr iptrp&tr-r ?l&ll

uEt3ELWt3t16rrtlEB

I /"

*/t

A. Maimun et d.l. / Ocean Engineering 38 t01I) t )9l -1299 1293

-0.'1

_r0

-0.2

-0.2 a o.2

li (RAD)

Fig. 2. Sway force (pure sway-deep water),

;-0

-o.2 0 0.2

tt (p!/.D)

Fig.4. Sway force (pure sw.ay-shallow water).

o.2

-0.2-0.4 -0.2 0 0.2 0.4

lt (RAD)

Fig. 5. Yaw moment (pur€ sway- shallow water).

yaw axis of the mathematical model. In this case, Matlab Simulinkprograms were used to create numerical simulation of the turningand zig-zag manoeuvres (ltijima et al., 2000).

The equations of motion in this time domain simulation arethen solved by numerical integration Dormand-prince Method(Maimun et al., 2005). This method is included in one of theMathlab ODE suites. The Mathlab ODE suites is a collection of fiveuser-friendly finite-difference codes for solving initial valueproblems given by first-order systems of ordinary differentialequations and their numerical solutions. The three codes (ode23),(ode45), and (odell3) are designed to solve non-stiffproblems andthe two codes (ode23s) and (odelSs) are designed to solve bothstiff and non-stiff problems. (Ode45) was used in the simulationprogramme with variable time steps integration to avoid errors insome critical condition.

-0.4 -4.4

.Iz

I

-44-

t-

)--)D'/-

)/__{

/

/

/--

I

.Ie

-0.02

-0.04 t--0.4 -0.2 0 0.2 a.4

f (RAD)

Fig. 3. Yaw moment (pure sway-deep water).

Forces and moments are made non-dimensionalby pLdI_Pl2 andp*du'/2, respectively, and plotted against drift angie, as shown infts. 6-9. Coefficients based on experimental and empiricalmthods are compared in Table 3. Comparison of manoeuvrabilittrd the pusher barge will take place once these coefficients areinported as parameters in the simulation.

{ Time domain simulation

The components of forces in the equation of motion wereolculated corresponding to the prescribed manoeuvring motions.fbe vessel's swept path can based on the input of hydrodynamicGmcients in the simulation. The swept path is by doubleiregrafing the acceleration of the vessel in surge, sway and

A. Moirnun et ol. I Ocean Engineeritlg i8 (2011) 1291 -1299

!r0

-1 01r'

fig. 6. sway force (pure yaw-deeP water)'

0.06

0.04

0.02

0

-0.02

-0.04

-0.06-2'1012

r'

fig.7. Yaw moment (pure yaw-deeP water)'

4.1. Sinrulation equation

The equations in the system of a three degree of freedom

motions ii used in the simulation and is based on the concept of

the modular manoeuvring model. Whereby each element con-

sisting of hull, propellers' and rudders- is represented by a self-

contained and discrete module. The surge, sway, and yaw force

equations are presented as follows:

x : ! p L2 d{m' + m!-\n -+P Lz d(m' + rn-r)nt

Y :$pL2d(m'+d)n +lpL2d(m'+m:,)ru (6)

N:tpL+d(t-+lL)i

The equations can be rewritten in an acceleration form:

it : (x+rpL2d(m' +nly)tv)lCpL2d(m' +m'))

!*0

-1 0'1r'

Fig.8. Sway force (pure yaw-shallow water).

-2-1 0',1 2

r'

Fig. 9. Yaw moment (pure yaw-shallow water)'

Table 2lnput parameters of PMM test.

a

.Iz

o.2

-0.1

-0.3

Model test Drift angle(dee.)

oscillator Phase angl€ ofamp (m) oscillator

arm (deg)

Drift test 0",4',8', 12",16" 0 0

Pure sway 0" 0.1-0.4 0

Pure yaw o" 0.1-0.4 33-02'

Yaw wirh drifr 4.' 8", 12', 16' 0.1-0'4 33.02"

, : (Y-rpLzd(m' +m'^)ru) lGpt) d1m' +m'r))

t:NlGpL4d(r'-+r))

The above set of Eqs. (7) could be integrated once to obtain

velocities, while displacement of motion is obtained with a single

(7)

/F

/_)/

,/--5

,-',/

/

/

//

/

,/

I/

A. Maimun et ol. / Oceon Engineering jB (2011) t2g1_12gg 1295

Table 3Comparison of hydrodynamic coefficients

Table 4Propeller and rudder parameters.

Symbol Hydrodynamic coeffi cients value Deep water Shallow warer-fiff: 1.3

Deep water (I/d > 3) Shallow water (h/d : 1.3)

Empiricalltl pMM Empiricallr I

Ship speed, Y (lmots) 7 7Number of propelfers 2 2P/D oJ4 o.t4Number of blades (Z) 4 4Diameter, D (m) 2.67 2.67Blade area ratio, EAR 0.461 0.461Revolutions, RIs (n) 2.0636 2.7415Wake traction (wp") 0.299 0.4236Irust deduction (t) 0.2 0.254Advance ratio Ue) 0.458 0.275Thrusr coefficient (Kr) 0.173 0.23a7" 0.525 0.33Thrust, I (kN) 76.825 193.29Cn,C1,C2 0.313-o;-0.2736;-0.1048 0.3139;-0.2736:-0.1O48Wake fraction (tvRo) 0.31 89 0.4395Rudder high (m) 2.5 2.5Rudd€r area (m2) 10.00 (Nf:2) 10.00 (Np=2)Numberofruddes 2 2

x'1tr

Y',,

Y'r

Yhp

| ,,.

Nh

NlNhn

N;Nhn,

N',,,ff

Yhw

Nhn

N;

0.017 40.3124

0.s967

0.25010

0

-1.750s2.0189

0.0383

-0.01880

0

-o.3262-0.1 190

'1.4556

0.0387

- 0.0c49

0.0086

0_1s7s

1.9588

2.4193

- 1.04080

04.4692

2.6048

0.2081

-o.07370

0

- 1.5483

- 0.s450

6.2548

2.4s420.0398

0.0578

0.01620.0s09

0.2646

0.00700.5269

-0.0141-0.1239

0.4934

0.0788

-0.03660.0r40

-0.0245- 0.1 750

-o.0254

0.0400.0509

1.2396

-o.43511.838s

0.02321.0400

3.5500

o.4172

-0.11260.0331

-0.0647-0.2193-0.o312 Table 5

Turning circle parameters based on experimental data.

hlr Experimental lMOstandards(m)

Comespond

Deep water AdvanceTacticaldiameter

1.3 AdvanceTacticaldiameter

<53] mor4.5L<590mor5,0 1

<531 mor4.5L<590mor5.01

254265

360410

YesYes

YesYes

integration to the velocities (Maimun et al., 2002):

u: I ttat andx: I udt

v: .f

r,atanat: I vat

r: ltarandt: lrdtr!:f ardrlt:f (8)

Turning circle in deep water

:!c

Tran:fer irrr)

Fig. 10. Comparison of simulated turning trajectories in deep water conditionberween empirical and PMM test [H/d=26.5, where ship speed is 7 Knots].

5.1. Compaison of empiical and expeimental results

The series of Figs. 10-12 displaythe empirical and experimentalresults obtained from the simulation programme in deep watercondition. The advance and tactical diameter of experimentalresult in Fig. 10 are smaller as compared to the empirical result'sdiameter. Figs. 11 and "12 show a higher value of overshootingangle for empirica! as compared io experimental result for 10/10and 2Ol2O zig-zag manoeuvre.

4.2. Simulation data

The simulation input data, which is a combination of hydro-dynamic parameters and forces of huli, rudder and propeller,must firstly be determined before the pusher barge simulation ofmanoeuvring pusher barge can begin. The effect of hydrodynamicforce, which incorporates the rudde( propeller and hull isincluded in the simulation together with equation of motion(Maimun et al., 2005).

Hydrodynamic force composes oi three components namely thebare hull, propeller, and rudder force components, which aredetermined by their respective source (Yoon and Rhee, 2003).Tables 4 and 5 show the propeller diameter and rudder area usedin the simulation. Prediction of resistance component was carriedout in Universiti Teknologi Malaysia towing tank. The resistanceand propulsion components are one of the parameters in thesimulation programme. Rudders area is calculated based on Detnorske Veritas (DnV) for minimum rudder area and the calculationof rudder area coefficient is referred to Crane et al. (1989). Forvessel with increased manoeuwability, rudder area coefficient areestimated 2-4 percent of L x d (Maimun et al., 2005).

5. Results

The simulation results illustfate a comparison of experimentaland empirical results of ship motion in both deep and shallowwater conditions. The result of ship motion is based on manoeuvr-ing ability test such as turning circle test and. zig-zag manoeuvre.

t

3 2oN

'u.t

ilnl

5i-l1:.1Dt3[i i:'nC101

t

I

I

A. Maimun et ol- I ocean Engineering 38 (201 1) 1291 -1299

.-!-f

tl

a

fi

.5

i0

:a

'!i

Fig. 11. Comparison of time histories of :tr,llo zig-zag for deep water [speed:

7 Knotsl.

20r?0 deg..l,'lanoeuvfe in deeP watel

''1.{: 700 $3i r;nli*i ,,n1

6i0

" r*'^

i

., i-,-lr'l:j

oi'

,oi

"i-:il *

'J

4

---:::..*--*--.*|.,-.'--...,.,,..-:.^ . . ,,t i', I

| ' i ,{i "ili ii ",, ,li ',"li ll',i,,,1i',,',1i,-I ill! ili :,| ,r, ri, 1! ili j,t \i i, ll ,l'r ,i i

I \; "'

ll il't i II ,,, .,1,,

t,..i i

i.i

.^--:'..^*-*,--)-.---,'l'

-*L*-

* 1

^-- tE^*-- Zf, i'J: 4tti tlil at-

Fig. 13. Comparison of simulated turning trajectories in deep and shallow water

based on emPirical equatlons'

4u

3tt

a11u )ffl

d:

1tr

. " " * " - " - " - " -".::. r* *:11: - *:

i- *i;;r*;] i

|- ll':::'i rr ;!+ ir' 'ir+' {:+ar

Fig.l2.Comparisonoftimehistoriesof20'/2ozig-zag[sPeed:7Knotsl.

5.2. Comparison of results Jor deep and shallow water conditions

5.2."1. Tuming abilitY-.-it. comp-a.ison of results in both deep and shallow water

.onJi i* are illustrated in Figs. 13-16. Figs. 13 and 14 show a

,*.fi.. ,arrnce and tactical d'iameter for deep water condition'

;il;".; the empirical result shows a slower response of tuming

trajectory in deep water as compared to shallow water'

5.2.2. Zig-zag manoeuYre

te f-O/f O zig'zagmanoeuvre result in Fig' -15 demonstrates a

f"rg".1r"ttnooi anite in shallow water' However' the difference

of overshoot angle between deep and shallow water is less

obvious in Fig. 16 in the 20/20 zig'zag manoeuvre'

6. Discussion

The simulation results of the manoeuvring capability of th€

push barge are dependent on the input of hydrodynamic

Fig.14. Comparison ot simulated turning trajectories in deep and shallow water

based on experimenHl results'

coefficients as well as the hull and rudder parameters' Thus the

;;;;;;;;;;;r results in tne experimental and empirical methods

are mainly due to some "i'r'Auting the experiment or model

;;;;;;il". Errors such "s

noi'e during data collection from the

experiment are normally p*it" ttowiver' the re-sult generated

il; il. simulation on the turning circle and zig'zag manoeuvr-

i;;1;;;; on "*p"'i-"ntal method illustrate better man-

oeuvring Properties as compared to em,pirical approach'

The turning circle result! have satisfied the IMO standards by

not exceeding the IMO advance and tactical diameter' However'

;lh" t-;;;;;;id not be said to the zig-zag manoeuvre result as the

overshoot angle for snartow wateri exieeds the IMO standards'

oneofthemaincausesofforexceedingovershootangleisdesign

dae*,r

riEfEEGq!

I

Turning circle in deep and shallow water (Empiricail

:fi

Turning circle in deep and shallow water {ExPeriment)

-, -,. --'

rl

t

A. Maimun et al. f Ocean Engineering 38 (201 I ) 1291 -l2gg

lCl!C Crg. nrnordyro ir diap and shallff wrtil Gilriricallar1a rcE. Llzno{u}ra ix dN,p andlhdlo*watcr {Err0in.nt)

I

:

I

tl:l,lrl

. -, j

.tt -

.T -....---**- . - ^* ---- *:-*-.. --^L-*-,-,-.^!. ,.t ;,11 t.i' flJ :,1-

iii.t.,i

_-l:-;:t,r:; l.i.-,,..; ---^Lrit1.rti, ii,.,t ts:l: - _ ,r,iat,n ::i Jrfl I

- tiilrialSfiij:i )

Fig. 15. Comparison of time histories of tO"l1O. zig-zag for deep and shallow warer (H/d=.t.3) lspeed:7 Knots].

2lA ir! li!r"i..usr jrd.tl ardshalto* *atil (Erp€.imrnil

:tl2Cd.q. r.nolu{rcind{rpanrshail.rivrt.r{Empiri(nii

g4

I

x

:"rc

I 131 iln ,: n, illi.r.r i:l

-:J.n.! !ta{i:?ni.

- - .I:i:.',,..1,::.=:l''- - .rati,l]:lli.-lFl:

,1'c

:l .l-l1

,l,, lj

iuI

--7r:ri

1sater

rcow water

Thus theI methodsor modelrfrom thegeneratedrlanoeuvr-Ger man-Lndards bYHowever,

sult as thesfandards.e is design

I

i'lrit' 1l;i ',1rl

,t'ttt;t:l;ti!ili

tit:,J

ir.a iti

:

-;trr:.1

iE:!:L(1t "/.- jj

I

-44cf'x

t ,-4: ito{, *ttt.::_ - 'q,irtrr-!111:l _l

i * - .E?::,1: ,1,t 4-!' ::

";,

, **aatat at :1e ,iaar aa'eat I

:-rkinr,: ;iit i:.:!qr:i1l :

:-,-.*Jlr?rj:et{"i jj :

-. .. ,:ta:,i irlu i,ii -.

Fig. 16. Comparison of time histories of 20"/20. zig-zag lspeed=7 Knots]

of the pusher-barge system. Thus, further studies on the hulldesign for the pusher-barge system might improve the zig-zagmanoeuvre result.

The comparison of experimental and empirical results for:urning circle shows that the experimental result has a better:nanoeuvring characteristic based on IMO standards.The zig-zag:nanoeuvre result shows that the empirical result has a higherlvershoot angle as compared to the experimental result. A:omparison of the 10'/10' zig-zag manoeuvre in Fig. 15 illustratesin irregular overshoot angle of the empirical result whereby the:rfference of the first and third overshoot is around 5".

The comparison result of rurning circle for both deep and:-:rallow water conditions in Fig. 13 shows deep water conditionras a tendency to have a smaller tuming circle diameter as

--Dmpared to the shallow water condition. A smaller turning circle:rdicates higher turning angle rate for a deep water condition.Ine turning circle results sholv that the deep water has a higher:curse changing capability as compared to shallow water condi-:on. The relationship of the turning circle test in the context:i push barge operation in conhned and normalty shallow water-d,'ays such as in canals and river inlet raises complication as the

vessel needs a good course changing capability to avoid ground_ing eliects and collisions.

Two standard time history of zig-zag manoeuvre result namelythe 1 0'/1 0' and,20 l20 for both shallow and deep water conditioniprovide estimation on the vessel ability to enter into tumingmotion. Results for both rudder angles signiry a higher overshootof heading angle for shallow as compared to deep water condition.However an increase of rudder angle indicates a small difference ofovershoot between the two conditions. The increment of rudderangle produces larger moment acting on the vessel, which countersthe opposite yaw motion. However, vessel's response towards thechange of rudder angle and moment depends on the flow fieldaround the hull. Constraints of flow field around the hull in shallowwater will reduce the rudder efflcienry.

The combination of turning circle and zig-zag manoeuvre testsuggests shallow water conditions have a significant impact onthe manoeuvring characteristic of push barge. Operations of apush barge in shallow water are confined to several limitationssuch as difficulty in course changing capability as well as lowrudder response. These limitations require a great deal of naviga_tion experience as well as high dependency of navigation aid.

fhe result ofthe simulation can be further utilized in the futureto analyze the correlation of ship manoeuvring capabilify withfinite depth condition by adding more ratio of water depth to shipdraught (H/d) condition. Since tlre simulation incorporated therudder parameters, a recommended future research of analyzingthe correlation of increment of rudder area to the manoeuvrability

. of pusher barge could be done with experimental validation

7. Conclusion

This research can be concluded as follows:

(a) The simulation result of zig-zag manoeuvr and tuming circleshows that the experimental derived coefficient tend to givebetter manoeuvring characteristic as compared to the empiricallyderived coefficient based on agreement with IMO standards.

(b) The comparison oi manoeuvring properries of deep and shailowwater coitditions show that a pusher barge has difficulty inmanoeuvring in shallow water due to constrains such as com_plexity in course keeping and high overshoot heading angle.

(c) This research can be further developed by determining thecorrelation of finite water depth and rudder area withmanoeuvrabilify of the pusher-barge by varying the waterdepth to ship draught tatio (H/d) as well as increasing rudderarea of the pusher barge.

Acknowledgement

The authors would like to express their sincere gratitude toProf. Chengi Kuo from the University cf Strathclyde, Gtasgow forhis invaluable advice and comments given during the preparationoF this paper.

Appendix A. Force and moment induced by propeller andrudder

Xp, Yp, Np and Xr, Yn, Na are expressed as the tollowing formulas

Xp : Ctp{1 -tp)n2 Df,tG{J o)

Yp:oNp: o

Xp

' iquuzwhere

KrUil: C1 +C2Jp+CdiJp: U cos 00-wp)l(nDp)wp:wnexp(-4.O0i)P''p: P-/pt'xi = -0.5where f, is the thrust reduction coefficient in straight forwardmoving, Cr, the constant, n the propeller revolution, D, the propellerdiameter, w, the effective wake fraction coefficienl at propellerlocation, w. the effective wake fraction coefficient of propellerin straight running, Kr is the thrust coefficient of a propeller force,

./p the advance coefficient, and C1, C2, and C3 are the constants forpropeller open characteristics.

The terms and the non-dimensional of ruder forces describesas the following.

Xp : -(1 -fp)F1y sindYR: -(1 f og)lycosdN6: -(xp*ogxs)Flycosd

A. Maimun er al. f Ocean EngineeringjS (2011) l29l_12gg

where xR is the the distance between the centre of gravity of shiand centre of lateral force (xp : xht) and xR represents the locatioof rudder (: -Ll2), xH is the distance between the centre cgravity of ship and centre of tateral force (xs :xrL),6 is ruddeangle, and tp, tp, as, and xs are the interactive foice coeff,cientamong hull, propeller and rudcier. Fiu is rudder normal force anrdescribed as the following:

F'N -- (AR/LdlcNUfrsinap, ", Fru

' N: lpt-d

where Ap is the rudder area, C17 is the gradient of the lifcoefficient of ruder, and can be approximated as the function orudder aspect ratio Kp.

Crv :6.13Kn/(& +2.25)

Up and aR represent the rudder inflow velocity and anglerespectively; they can be described as the following.

Urt:6-w*1211+G(s))s(s) : tlK12-(2-rQslsl(1 _s)2

4 : Dp/hpK : 0.6(1 -wp) /(1-wn)s : 1 .0-(1 -wp )U cos B /npWp:wpsWpfWp6aa: 6-! 0'n

B'*= B-2x:*r'xh = -o'5where Ap is the rudder area, hs the rudder height, Kp the aspectratio of rudder, Up the effective rudder inflow speed, cta theeffective rudder inflow angle, C the coefficient for starboard andport rudder, wp the effective wake fraction coefficient at rudderlocation, wp6 the effective wa[<e fraction coefficient at rudderlocation in straight forw,ard motion, p the propeller pitch, andl, the flow straightening coefficient.

Appendix B. Approximate formulae of hydrodynamiccoefficients in deep water by l(jima

These formulae are function of ship length (t), breadth (B).draught (d), block coefficient (CB) and aspect ratio of ship huti [:(k:2dlL). However, the prediction formulas proposed

-by Mori

(1995) are for four parameters, which are eo, e,o, oo and K toexpress the characteristics of aft hull shape. The coefficients eoand e! express fullness of aft run and oo is aft sections fullnesimetric, while K is the form factor. The parameters are defined byprincipal particulars with water plane area coefficient C* aniprismatic coefficient Cpo of aft hull between A.p.:

Ieo : "U{1-Ceo)

l)qB

E-,U 4 ' @+d)2

e'a=ea

fo,:l$-

I _Lpo

* : G * ffi -o.zz) <os5oo + o.4o)

The engagement of these parameters together with a modetrdatabase of 15 kinds ofships and 4g loading conditions generaresa series of approximate of hydrodynamic coefficients as follovis:

Yb : o.snkk + t.szst ( $\ o;' \1./Y'B-1m' + m;1 : o.25nk + 0.052e,,-o.4s7Y'pp: -7-79gcBoo*7.o5

1299A. Maimun et al. f Ocean EngineeringjB (2011) 1291-1299

o 22s({:\e',-o 12\D /

,,r..Jdt1-cBt\"--"1 B j

trrj(l_CB)1 ,12 -^_.(d(l_CB)) .,to44ilta# ).,1 -s 3?4t r;::'le"t 1227

u f

, ro uuu {{

A;Q}.;r}'-zr a r, {4# }.;r< + r soz]

L .'

= -o 54k-k2 -0.o477e',oK+O.0368- 0.15K-0.068

- : -0 4O86Cts+0.27(dr1-CBrl

-- -0 826 t arB::i

le -o ozg

r;pendix C. Approximate formulae of hydrodynamic

:-:efficients in shallow water

ihe approximate formulae for estimating the hydrodynamic

. :e acting on a ship in shallow water was proposed by (Kijima

,, ri., rSSOI These foimulas are obtained semi-empirically from thej...rlts

of numerical calculations based on lifting surface theory' andj:del

tests in full toad condition' The formulae for full load

:rdition can be applied to the case of shallow water by correctingl:: hydrodynarnii toefficients and coefficients' which have been

, .ginally oltained for deep water (Maimun et al" 2005)'-grtirnation

of the hydrodynamic force acting on a ship in

:-.allow water can be done as follows:

_ _..,:f (h)xD6",

".:ere D,6* is the coefficients in shallow water including ballast

; ,d hali ioad conditions, D7", the coefficients in deep water'

-.t,rJlng batlast and half lcad conditions, andflh) the correcting

''.aor, h:dlH (d is draught, H is water depth)'

The correcting factor J(h) for the effect of water depth is

:.r.'ided into two different equations, which are applied based

, ,, a .ertain hydrodynamic coefficients' The correcting factors flh)::e as followed:

:) Correcting factor, J(h) for hydrodynamic coefficients Y'O' Y'rr'

Y'pn, N'p and N'.:

f(h): t1l(1-h)nl-hwhere, for instance, Y'n, n:O.4CB (Bld): Y'or, n: -O'26C8 (Bl

d) + 1.7 4: Y',oo, n = - z.izcsls + t.B (5.7); N',r, n : 0'425 x cB x Bl

d; Ni, n: -7.14k+1.5r) Correcting factorflh) for other hydrodynamic coefficients:

f (h):1 +arh+azh2 +ath3

where for instance,(i) For, Y;-(m'+ml):''

0 | : -'s.5(CB(B I d))2 +26C8(Bld) -31'sa2 : 37 (CB(B I dDz - 1 8s CB(B I d) +230ca : - 3 8(CB(B I d))2 + 7s7 CB( B I d) - 2so

(ii) For, Yi,:at:-b.ts " 1os(1-cB)s

a2:7.16 x 105(1-CB)5a3- '1.28 x 105( l -CB)s

(iii) For, Y,,,,.:' ar-2.'is v lo4((d(1-cB)lD2)

-0.48 x fi4do-cB)iB+220a2:-4.o8 x 1oa((d(7 -CB)|B)2)

-0.75 xrc4d(J.-cB)lB-274o3: -9.08 x 104((d(1 -cB)lil'z)

*2.55 x 104d(J -cB)lB-1400(iv) For, Ni,r:

o1: -O.24 x 10'(1 -CB)+57a2:1.77 x 103(1 -Cts)-413as-_ _ 1.98 x 10s(1 -cB)+467

(v) For, N'":ot :- o.t so x 1 04((d( 1 - cByB)1 + 44Bd(1 - CB)l B * 2s

a2 : 1,.222 x 1 O4(( d( 1 - cB) | B)2) - 27 20d(1' CB) I B + 446

,r: -t.ZtA x 104((d(1 -Cl)lilz)+26s0d(1-CB)!B-137

(vi) For, Nrr,:o,:0.51 x rc2cB@lB)- )sa2-- -5.75 x rczcB@lB)+744a3:5.08 x rc2cq@lB)-143

(vii) For, Ni,,:at:0.4 v 10JcB(B/d)-88a2: -2.95 x rc3cB?ld)+645a3:3.12 x 1 o3cB(B/d)- 678.

References

Beukelman, W., Journee, J.MJ.,2001. Hydrodynamic transverse loads on ships in---d"ep

rnd ,titto* *ri"., HADMAR'2001, in: Proceedings ofthe 22nd lnterna-

tional Conference on Hydrodynamics and Aerodynamics l' Marine Engineering'

Crrn", C.f"- eO", H., Lands-burg, A', tg8g' Controllability Chapter lX Principle of

Naval Architecture Second Revision Vol lll'xUiri., i., rl"f.i.i, Y., Furukawa, Y., 2000- on the Prediction Method for Ship

Manoeuvrability, irr: Proceedings of lnternational Workshop on Ship Man-

oeuvrability, Hamburg Ship Model Basin, Germany, Paper No' 7'

<Ui"ri, ri., lrir.'iii, Y., Tsuiui, f., M"tsunrga, M., 1990' Prediction method of ship

manoeuvrability in deep and shallow waters MARStM and ICSM' vol 90'

Tokyo, JaPan.f-rt.i,.le.,"Vjntotre, M., Laforce,8., Eloot' K', Delefortrie, G',2007' Navigation in

Confined Waters: lnfluence o[ Bank Characteristics on ShiP-Bank Interaction'

in: Proceedings of Second International Conference on Marine Research and

Transportation, June 2007' lschia, Itali.

Maimun,^A., Atel i., Muhammad, A.H.' 2005' Manoeuvring of Pusher-Barge in' Deep and Shatlow Water, in: Proceedings ofTenth JSPS Marine Transportation

Engineering Seminar, Hiroshima, Japan'tuaimu"n, R., t-o-h, S.p., t-.", R., 2002. Minoeuvring Assessment of Fishing Vessel' in:

Proceedings of 7th JSPS Marine Transportation Engineering Seminar' Hir-

oshima, JaPan.fra".i, ft,l., iS"95. n note on hu!l form design (Part 24)' Journal oi Ship Science 48'

40-49 (ln JaPan)'vanJrre,u., eiolt,r., lsgo. Hydrodynamic Phenomena Affecting Manoeuvres at'-'-L"*

ip."a in shallow Navigation Areas, in: proceedings of 't.lth lnternational

Harbour Congres, Anrwerpen, Belgium, pp 535-546'

w"rg,'oJ- g"k.intouzis, 1., katory, rt/I., zooo' Prediction of ship hydrodynamic

d'erivltives in shallow water and restricted waters' lnternational ShiPbuilding

Progress 47 (452), 37 9 -396.voon,iix"-ntt"", p.i., zoos. ldentification of hydrodynamic coefficients in ship- - -

rrno"uuring equations of motion by estimation-before-modeling technique'

Ocean Engineering.