Ocean Engineering 192 (2019) 106239

Contents lists available at ScienceDirect

Ocean Engineering
journal homepage: www.elsevier.com/locate/oceaneng

Theoretical background and application of MANSIM for ship

maneuvering simulations
Omer Faruk Sukas a, *, Omer Kemal Kinaci a, b, Sakir Bal a
a
Faculty of Naval Architecture and Ocean Engineering, Istanbul Technical University, Turkey
b
Department of Naval Architecture and Marine Engineering, Istanbul Technical University - Turkish Republic of Northern Cyprus Education Research Campuses (ITU-
TRNC), Turkey

A R T I C L E I N F O A B S T R A C T

Keywords: In this study, a new developed code, MANSIM (MANeuvering SIMulation) for ship maneuvering simulations and
Hydrodynamic derivatives its theoretical background are introduced. To investigate the maneuverability of any low-speed ship with single-
DTMB5415 rudder/single-propeller (SPSR) or twin-rudder/twin-propeller (TPTR) configurations, a 3-DOF modular mathe­
KVLCC2
matical model or empirical approaches can be utilized in MANSIM. In addition to certain maneuvers of ships
MMG
Sensitivity analysis
such as turning or zigzag, free maneuver can also be simulated. Input parameters required to solve the equations
Hull-rudder interaction of motion can be estimated practically by several empirical formulas embedded in the software. Graphical user
interface of the code has simple design to enable users to perform maneuvering calculations easily. In addition to
results such as advance, transfer, tactical diameters etc. on the user interface, simulation results can also be
analyzed graphically; thus it is possible to examine the variation of kinematic parameters during simulation.
Using the code, maneuverabilities of a tanker ship (KVLCC2) and a surface combatant (DTMB5415) is investi­
gated and computed results are compared with free running data for validation purpose. MANSIM can be ad­
vantageous for parametric studies and it is a valuable tool especially for sensitivity analysis on ship maneuvering.
The software is available online at www.mansim.org. The effects of variation of hydrodynamic derivatives and
rudder parameters on general maneuvering performance of ships are investigated by performing sensitivity
analyses. It is found that linear moment derivatives and rudder parameters are highly effective in maneuvering
motion. Another interesting outcome of this study is identification of the significance of third order coupled
derivatives for DTMB5415 hull. Effects of linear derivatives on maneuvering indices are also investigated by
MANSIM. Results show that there is not a linear relationship between hydrodynamic derivatives and maneu­
vering indices.

method is more practical than the former. However, accuracy of these

calculations directly depends on selected mathematical model and hy­
1. Introduction drodynamic derivatives involved (Guo and Zou, 2017; Toxopeus et al.,
2018; Sukas et al., 2019). In the recent literature, current trend to ex­
Predicting maneuverability of ships is considered as one of the most press the hydrodynamic forces and moments is adopting either Abko­
challenging topics in ship hydrodynamics. Maneuvering simulations are witz model (Abkowitz, 1964) or MMG model (Ogawa and Kasai, 1978;
generally conducted by using CFD-based or system-based (SB) methods. Yoshimura, 2005; Yasukawa and Yoshimura, 2015). In Abkowitz model;
CFD-based method can be defined as a direct simulation of the actual hull, rudder and propeller are considered as one rigid body, and equa­
maneuvering motion, including the steering rudder and rotating pro­ tions of motion are defined by third-order Taylor series function. On
peller (Bhushan et al., 2009; Carrica et al., 2013; Broglia et al., 2015; contrary to Abkowitz model, MMG model is a simplified mathematical
Ohashi et al., 2018; Duman and Bal, 2019). From practical perspective, model that decomposes total hydrodynamic force and moment acting on
this approach is not feasible as it requires enourmous computational the ship into hull, rudder and propeller components. One of the biggest
power to perform full time-domain simulations. On the other hand, SB advantages of MMG model is that it allows to take the
methods include the solution of equations of motion for every time step hull-rudder-propeller interactions into account. There are various
using the previously calculated hydrodynamic derivatives. The latter

* Corresponding author.
E-mail address: [email protected] (O.F. Sukas).

https://doi.org/10.1016/j.oceaneng.2019.106239
Received 25 April 2019; Received in revised form 22 June 2019; Accepted 29 July 2019
Available online 19 September 2019
0029-8018/© 2019 Elsevier Ltd. All rights reserved.
O.F. Sukas et al. Ocean Engineering 192 (2019) 106239

Abbreviations NR Yaw moment due to rudder around z axis ðN mÞ

o0 x0 y0 z0 Earth-fixed coordinate system
Ad Advance o xyz Ship-fixed coordinate system
CFDB CFD-Based r Yaw rate around z axis at midship ðrad=sÞ
EMP Empirical S Wetted surface area of ship ðm2 Þ
FR Free Running tP Propeller thrust deduction factor in maneuvering motions
GUI Graphical User Interface ð Þ
LCG Longitudinal Center of Gravity tR Steering resistance deduction factor ð Þ
MMG Maneuvering Modelling Group TP Thrust of propeller ðNÞ
NR Number of Rudders u; v Velocities in x and y axis at midship ðm=sÞ
OA Overshoot Angle uP Longitudinal inflow velocity to propeller ðm=sÞ
SB System-Based uR ; vR Longitudinal and lateral inflow velocity components to
SDA Steady Drift Angle rudder, respectively ðm=sÞ
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
SPSR Single Propeller-Single Rudder U Ship velocity at midship, U ¼ u2 þ v2 ðm=sÞ
STD Steady Turning Diameter VA Initial speed of ship (knots)
STS Steady Turning Speed VT Steady turning speed (knots)
SYR Steady Yaw Rate wP0 Effective wake fraction at propeller position in straight
TD Tactical Diameter motion ð Þ
TPTR Twin Propeller-Twin Rudder wP Effective wake fraction at propeller position in
Tr Transfer maneuvering motion ð Þ
x’H Non-dimensional longitudinal position of acting point of
Symbols
additional lateral force ð Þ
AB Submerged bow profile area ðm2 Þ
aH Rudder lateral force increase factor ð Þ x’P ; y’P Non-dimensional longitudinal and lateral position of
propeller from midship ð Þ
AR Profile area of movable part of mariner rudder ðm2 Þ
B Ship breadth ðmÞ x’R y’R Non-dimensional longitudinal and lateral coordinate of
CB Block coefficient ð Þ rudder position, respectively ð Þ
c Rudder chord length ðmÞ XH Surge force due to hull in x axis ðNÞ
d Ship mean draught ðmÞ XR Surge force due to rudder in x axis ðNÞ
DP Propeller diameter ðmÞ XP Surge force due to propeller in x axis ðNÞ
fα Rudder lift gradient coefficient ð Þ YH Sway force due to hull in y axis ðNÞ
FN Rudder normal force ðNÞ YR Sway force due to rudder in y axis ðNÞ
Fr Froude number ð Þ αR Effective inflow angle to rudder ðradÞ
H Rudder span ðmÞ βP Geometrical inflow angle to propeller in maneuvering ð Þ
βR Effective drift angle at rudder position ðradÞ
g Gravity, g ¼ 9:81 ðm=s2 Þ
β Ship drift angle ðradÞ
Iz Yaw moment of inertia around z axis (kg m2 )
γR Flow-straightening coefficient of sway velocity for rudder
Jz Added yaw moment of inertia around z axis (kg m2 )
ð Þ
JP Propeller advance ratio ð Þ
δ Rudder angle ðradÞ
k0 ; k1 ; k2 Propeller open water characteristics for expressing KT ð Þ
ε Ratio of effective wake fraction in way of propeller and
KT Thrust coefficient ð Þ
rudder ( )
l’R Flow-straightening coefficient of yaw rate for rudder, l’R ¼ η Ratio of propeller diameter to rudder span ð Þ
lR =L ð Þ κ An experimental constant for expressing uR ð Þ
L Overall length of ship ðmÞ Λ Rudder aspect ratio ð Þ
m Ship mass ðkg; tÞ ρ Water density ðkg=m3 Þ
mx , my Added mass due to ship motion in x and y directions, τ Static Trim ðmÞ
respectively ðkgÞ ψ Ship heading angle ðradÞ
nP Propeller revolution ð1=sÞ
NH Yaw moment due to hull around z axis ðN mÞ

maneuvering simulation results using MMG model with different this issue can be handled if PMM motion parameters are selected
modified versions (Fang et al., 2005; Kang et al., 2008; He et al., 2016; properly according to ITTC Recommendations (ITTC 7.5-02-06-02,
Yasukawa et al., 2019). 2014). Recent study has shown that changing the advancing speed of
There are several prediction methods proposed in literature to ship in PMM tests has a significant effect on the hydrodynamic de­
determine the hydrodynamic derivatives in MMG models (Sukas et al., rivatives in MMG model (Zhang et al., 2019). In addition to these
2017a, 2017b). For example, Yasukawa and Yoshimura (2015) carried methods, Liu et al. (2017) presented an integrated empirical maneu­
out circular motion tests (CMT) to obtain the hydrodynamic derivatives vering model for inland vessels. Furthermore, all hydrodynamic de­
and it was noted that CMT is a suitable method since it has zero fre­ rivatives and propeller/rudder parameters in MMG model were
quency of motion which reduces the uncertainties for hydrodynamic estimated by various regression formulas selected from literature. On
forces and moment. PMM (Planar Motion Mechanism) tests are also the other hand, system identification techniques have also been used for
widely used to determine the hydrodynamic derivatives (Cura-­ estimation of the hydrodynamic derivatives (Zhang and Zou, 2013;
Hochbaum, 2011; Obreja et al. 2010; Sakamoto et al., 2012; Yoon et al., Sutulo and Soares, 2014; Yoon et al., 2015; Xu et al., 2019). In order to
2015; Duman and Bal, 2017). Accuracy of results based on the selection predict maneuverability of any ship using system-based approach, these
of PMM motion frequency and amplitude may even change. However, methods can be utilized to obtain the hydrodynamic derivatives,

2
O.F. Sukas et al. Ocean Engineering 192 (2019) 106239

Table 1
STD CB B 194 Hc Hc
Range of ship parameters used by Lyster and Knights (1979). ¼ 4:19 203 13:0 þ 35:8 ðST 1Þ þ 3:82 ðST 2Þ
L δ L δ Ld Ld
Parameters Single Propeller Ships Twin Propeller Ships AB
þ 7:79
Min: Max: Min: Max: Ld
(1)
LPP ½m� 54.86 329.18 76.20 225.55
CB 0.56 0.87 0.42 0.62 TD STD VA
¼ 0:910 þ 0:424 pffiffiffi þ 0:675 (2)
δ 10.00 45.00 10.00 35.00 L L L
B= L 0.11 0.18 0.06 0.20
τ= L 0.00 0.05 0.01 0.03 Ad TD
¼ 0:519 þ 1:33 (3)
Hc= Ld 0.01 0.04 0.01 0.02 L L
pffiffi
VA = L 0.20 1.00 0.25 2.20
Tr TD
¼ 0:497 0:065 (4)
L L
propeller/rudder parameters. VT TD
In this study, a new user-friendly ship maneuvering code called ¼ 0:074 þ 0:149 (5)
VA L
MANSIM (MANeuvering SIMulation) based on standard MMG mathe­
matical model (Yasukawa and Yoshimura, 2015) is introduced. The code where,
includes several empirical relations suggested by various researchers ST ¼ 1 if dc � Hc, ST ¼ 2 if. dc > Hc
working on the topic and is for those who would like to have a funda­ For TPTR ships, empirical formulas for turning maneuver indices are
mental background of the maneuvering abilities a ship during early given as follows:
design stages. The primary aim here by developing such a code is to
make the maneuvering predictions of ships easier by using a simple user STD CB B 188 Hc VA
¼ 0:727 197 þ 4:65 þ 218 ðNR 1Þ þ 1:767 pffiffiffi
interface. The software allows to simulate the turning and zigzag ma­ L δ L δ Ld L
neuvers of ships. In addition, a free maneuver option is also available þ 25:56
AB
(6)
with unlimited number of rudder angles which means one can simulate Ld
free maneuver of any ship in MANSIM by importing a.txt file that in­
TD STD
cludes rudder angles with respect to times. MANSIM displays the ¼ þ 0:14 (7)
L L
maneuvering results such as advance distance, transfer distance, tactical
diameter etc. Input parameters of the mathematical model, such as hy­ Ad TD
drodynamic derivatives and coefficients related to the propeller and ¼ 0:514 þ 1:1 (8)
L L
rudder, can also be estimated by several empirical formulas embedded
in the software. The empirical approach provided by Lyster and Knights Tr
¼ 0:531
TD
0:357 (9)
(1979) may also be preferred as a second option to have a basic un­ L L
derstanding of maneuvering abilities of a ship.
VT TD
Following this section, section 2 presents the theoretical background ¼ 0:028 þ 0:543 (10)
VA L
of MANSIM including the empirical approach suggested by Lyster and
Knights (1979) and MMG models for SPSR and TPTR ships. Section 3 Here, Ad, Tr, TD, STD and VT represent the turning maneuver indices
gives the empirical formulas embedded in software to estimate input and are advance, transfer, turning diameter, steady turning diameter
parameters of MMG model. GUI of code was briefly introduced and and steady turning speed, respectively. L, B, d and CB are the main
sample screenshots were given in Section 4. In section 5, maneuver­ particulars of ship. AB is the area of submerged bow profile, τ is the static
ability of two benchmark ships (KVLCC2 and DTMB5415) was examined trim and VA is the initial approach velocity of the ship. Parameters
for validation purposes and the results predicted were compared with related to rudder are rudder angle (δ), span length (H), chord length (c)
the free running test data. In addition, the effect of variation of hydro­ and number of rudders (NR).
dynamic derivatives and rudder parameters on general maneuvering
performance was investigated by performing a parametrical sensitivity 2.2. Modular mathematical model
analysis. Finally, section 6 includes brief summary and possible future
studies about MANSIM. In MMG model, hydrodynamic forces and moment acting on the ship
are divided into different parts (contributors) such as hull, rudder(s) and
2. Theoretical background propeller(s). Major advantage of this method compared to traditional
approach (Abkowitz-type) is including interaction effects of hull-rudder
In MANSIM, maneuvering performance of ships can be predicted by (s) and hull-propeller(s). MANSIM contains certain assumptions due to
using either the empirical model provided by Lyster and Knights (1979) the implementation of MMG model (Yasukawa and Yoshimura, 2015).
or the mathematical models presented by Khanfir et al. (2011) and These assumptions are given below.
Yasukawa and Yoshimura (2015). Further details are explained in the
following sub-sections. � Hydrodynamic forces and moment acting on the ship treat quasi-
steadily.
2.1. Empirical approach � Cruise speed of the ship is sufficiently low so that wave-making
resistance is ignored.
Range (maximum and minimum values) of parameters for ships used � Metacentric height (GM) is relatively large, thus effects of roll-
by Lyster and Knights (1979) are given in Table 1. Empirical formula­ coupling are negligible.
tions have been derived based on the model experiments.
Turning maneuver indices of SPSR ships can be calculated by the In the present version of MANSIM, mathematical models for SPSR
following semi-empirical expressions: (Yasukawa and Yoshimura, 2015) and TPTR (Khanfir et al., 2011) ships
are available to predict the maneuvering performance in calm water
condition. Following sections present the coordinate system,

3
O.F. Sukas et al. Ocean Engineering 192 (2019) 106239

positive while rotating to starboard side. u and v denote velocity com­

ponents in x and y directions, respectively. r is the yaw rate that can also
be defined as r ¼ ∂t .
∂ψ
The speed of ship is indicated as U ð ¼
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u2 þ ð vÞ2 Þ.
In MANSIM, hydrodynamic forces and moment, mass, added mass,
moment of inertia, added moment of inertia and other kinematical pa­
rameters are non-dimensionalized as given in Table 2.

2.2.2. Mathematical model

Maneuverability of only low-speed ships in the horizontal plane with
sufficiently large GM is considered in the mathematical model of
MANSIM. Ship motions withn six degree-of-freedom (6-DOF) were
reduced to 3-DOF under the following assumptions:

� Vertical motions (heave, pitch and roll) are ignored. Then w ¼ p ¼

q ¼ w_ ¼ p_ ¼ q_ ¼ 0.
� Ship was assumed to have symmetry over xz plane, then yG ¼ 0.

In this case, 3-DOF motion equations become:

� �
m u_ rv xG r2 ¼ Fx
Fig. 1. Coordinate system used in calm water.
m½v_ þ ur þ xG r�
_ ¼ Fy
IzG r_ þ mxG ðv_ þ urÞ ¼ Nz (11)
Table 2
Forces and moments on the right-hand sides of these equations can
Non-dimensionalization of ship and kinematical parameters.
be written separately:
Parameters Non-dimensionalized by
Fx ¼ mx u_ þ my vr þ X
X; Y 1 2 ρU2
=
Ld
Fy ¼ my v_ mx ur þ Y
N 1 2 ρU2 L2 d
=

u; v U
Nz ¼ Jz r_ þ N x G Fy (12)
r U=L 3-DOF motion equations then become;
m, mx , my 1 2 ρL2 d
=
�
Iz , Jz 1 2 ρL4 d
= ðm þ mx Þu_ m þ my vr xG mr2 ¼ X
�
m þ my v_ þ ðm þ mx Þur þ xG mr_ ¼ Y
�
IzG þ xG 2 m þ Jz r_ þ xG mðv_ þ urÞ ¼ N (13)
Table 3
X, Y and N can be divided into sub-components due to the modular
Hydrodynamic force due to propeller for single and twin propeller ships.
structure of MMG model:
Definition Single-Propeller Twin-Propeller Ships
Ships X ¼ XH þ X R þ XP
Surge force due to propeller XP ¼ ð1 tP ÞTp XP;S ¼ ð1 tP;S P;S Y ¼ YH þ YR
P P ÞTp
(s) N ¼ NH þ NR (14)
Thrust of propeller(s) Tp ¼ ρn2P D4P KT TP;S
p ¼ ρn2P D4P KP;
T
S

Thrust coefficent KT ¼ k0 þ k1 JP þ KP; S

¼ k0 þ k1 JP;S P;S 2 where the subscripts H, R and P refer to hull, rudder and propeller,
T P þ k2 ðJP Þ
k2 J2P respectively. 3-DOF motion equations can be written in matrix form as
Propeller(s) advance ratio JP ¼
uP uP;S follows,
nP DP JP;S
P ¼ P
nP DP;S
Inflow velocity to propeller uP ¼ uð1 wP Þ uP;S
P
P
¼ ð1 wP;S P;S
P Þðu þ yP rÞ
Acc ¼ M 1 ⋅ðP C ⋅ VelÞ (15)
(s)
Wake fraction of propeller 2
wP ¼ wP0 expð wP;S P;S P;S
P ¼ wP0 expð 4ðβP Þ Þ where Acc matrix includes acceleration terms (Acc ¼ ½ u_ v_ r_ �T ), P
(s) 4β2P Þ
forces and moments (P ¼ ½ XH þ XR þ XP YH þ YR NH þ NR �T ) and
Geometrical inflow angle βP ¼ β x’P r’ βP;S
P;S ’
r T
P ¼β ðx’P Þ
Vel the velocity terms (Vel ¼ ½ u v r � ). M 1 is the inverse of mass
matrix and C is the Coriolis matrix. Mass matrix M and Coriolis matrix C
are given as follows:
non-dimensionalization and the mathematical models for SPSR and 2 3
TPTR ships, respectively. m þ mx 0 0
6 7
M¼6 4 0 m þ my xG m 7
5 (16a)
2.2.1. Coordinate system and non-dimensionalization 0 xG m 2
IzG þ xG m þ Jz
The basic dynamic of motion is described by using Newton’s second
law of motion. Thus, two different coordinate systems can be defined for 2 � 3
0 m þ my r xG mr
a maneuvering ship: earth-fixed coordinate system (O X0 Y0 Z0 ) and C ¼ 4 ðm þ mx Þr 0 0 5 (16b)
ship-fixed (o xyz) coordinate system as shown in Fig. 1. xG mr 0 0
Here heading angle ψ refers to angle between x and x0 axis. The
difference between ship’s heading and actual course direction (velocity Using all these matrices, Eqn. (15) can be written explicitly as;
vector at COG) is drift angle, β ¼ tan 1 ð v =uÞ. The rudder angle, δ is

4
O.F. Sukas et al. Ocean Engineering 192 (2019) 106239

Table 4
Rudder parameters required to calculate rudder normal force(s).
Definition Single-Rudder Ships Twin-Rudder Ships

Rudder normal force FN ¼ 0:5ρAR U2R fα FP;S P;S P;S P;S P;S 2
N ¼ 0:5ρAR ðUR Þ f α sin αR
sinαR
Rudder lift gradient coefficient 6:13Λ 6:13ΛP;S
fα ¼ f P;S
α ¼
Λqþffiffi2:25
ffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P;S
Λq ffiffiffiffiffi2:25
ffiffiffiþ ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Inflow velocity to rudders UR ¼ u2R þ v2R 2 2
UP;S
R ¼ ðuP;S R Þ þ ðvR Þ
P;S

� � !!
Effective inflow angle to rudder(s) v yP;S
αR ¼ δ tan 1 R αP;S P;S
γP;S P;S 1 R
uR R ¼ δ R βR tan
xP;S
P
Effective inflow angle to rudder(s) in maneuvering βR ¼ β l’R r’ βP;S ðl’R Þ r’
P;S
R ¼β
vffiffiffi(
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffis
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi!)
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi vffiffiffi8
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi0
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi1
ffiffiffiffi9
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Longitudinal inflow velocity to rudder(s) u 2 u 2
u 8KT u >
u >
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi >
>
uR ¼ εuP tη 1 þ κ 1þ 2 1 þ ð1 ηÞ u < B
u
u 8K P; S C =
πJP uP;S
R ¼ εP;S uP;S
P t
u η 1 þ κ P;S Bt1 þ
@
T
2
1 C
A> þ ð1 ηÞ
>
>
: πðJP;S p Þ >
;
!!
Lateral inflow velocity to rudder(s) vR ¼ γR βR y P;S
vP;S
R ¼ uP;S
R tan γR βR
P;S P;S
tan 1 RP;S
xP

Table 5
Empirical relations to estimate mx , mY and JZ .
YH ¼ Yv v þ Yr r þ Yvvv v3 þ Yvvr v2 r þ Yvrr vr2 þ Yrrr r3
Reference Empirical Formula
NH ¼ Nv v þ Nr r þ Nvvv v3 þ Nvvr v2 r þ Nvrr vr2 þ Nrrr r3 (18)
Clarke et al. mx ¼ m*0:05
(1983)
Zhou et al.
� �
d
�
L where all coefficients here (Xvv , Yvvv , Yvrr , Nrrr , etc.) are known as hy­
mY ¼ m 0:882 0:54CB 1 1:6 0:156ð1 0:673CB Þ þ drodynamic derivatives or maneuvering coefficients.
(1983) B B
� � � ��
d L d d L d
0:826 1 0; 678 0:638CB 1 0:669
B� B � B BB B ��2 2.2.2.2. Propeller forces and moment. Propeller surge force (XP ) is
Zhou et al. 1 L
(1983) JZ ¼ m
100
33 76:85CB ð1 0:784CB Þ þ 3:43 ð1 0:63CB Þ
B calculated for SPSR (Yasukawa and Yoshimura, 2015) and TPTR
(Khanfir et al., 2011) ships with the parameters given in Table 3. Note
that side force (YP ) and yaw moment (NP ) due to propeller are neglected
as they have smaller magnitudes compared to those of hull and rudder
Table 6
components. The superscripts ‘P’ and ‘S’ in Table 3 represent port and
Empirical formulas for hydrodynamic derivatives for surge force (X).
starboard propellers, respectively.
Reference Empirical Formula
Here, tP is propeller thrust deduction factor in maneuvering motion,
Lee et al. (1998)
XVV ¼ 0:0014 0:1975d
ð1 CB Þ L nP is propeller rotation rate and DP is propeller diameter. k0 , k1 and k2
B d represent open water characteristics of propeller for expressing KT . wP0
Yoshimura and Masumoto (2012) CB
XVV ¼ 1:15 0:18 denotes the effective wake fraction at propeller position in straight
L=B
Yoshimura and Masumoto (2012) CB
XVVVV ¼ 6:68 þ 1:1 motion and x’P is non-dimensional longitudinal position of propeller
L=B
from midship.
� �
Lee et al. (1998) d L
Xrr ¼ 0:0027 þ 0:0076CB
B d
Yoshimura and Masumoto (2012) CB
Xrr ¼ 0:085 þ 0:008 xG my 2.2.2.3. Rudder forces and moment. Forces and moment related to
L=B
Lee et al. (1998) L rudder (XR ; YR ; NR ) for single-rudder ships are calculated based on
Xvr ¼ ½m þ 0:1176my ð0:5 þ CB Þ�
d
Yoshimura and Masumoto (2012) CB rudder normal force (FN ), rudder angle (δ) and hull-rudder interaction
Xvr ¼ my 1:91 þ 0:08
L=B coefficients (tR ; aH ; xH ) by the following equations (Yasukawa and
Yoshimura, 2015):

2 3 2 32 � 3
u_ 1=M11 0 0 XH þ XR þ XP þ v m þ my þ rxG mr
4 v_ 5 ¼ 4 0 M33 =detM M32 =detM 5⋅4 YH þ YR ðm þ mx Þur 5 (17)
r_ 0 M23 =detM M22 =detM NH þ NR xG mur

The first equation is solved separately, while the others need to have
a coupled solution. Moment of inertia around z axis in the mass matrix M XR ¼ ð1 tR ÞFN sin δ
is not given by the user, but assumed as approximately
Iz ffi mð0:25Lpp Þ2 . External forces and moment given in the right-hand YR ¼ ð1 þ aH ÞFN cos δ
side (RHS) of Eqn. (13) are explained in the following sub-sections. �
NR ¼ xR þ aH x’H FN cos δ (19)
2.2.2.1. Hull forces and moment. Hydrodynamic forces and moment
acting on both SPSR and TPTR hulls are expressed as follows (Yasukawa where tR is steering resistance deduction factor. aH and x’H are rudder
and Yoshimura, 2015): lateral force increase factor and non-dimensional longitudinal position
of acting point of aH from midship, respectively. For twin rudder ships,
XH ¼ X0 ðuÞ þ Xvv v2 þ Xvr vr þ Xrr r2 þ Xvvvv v4 hydrodynamic forces and moment due to rudders are calculated by

5
O.F. Sukas et al. Ocean Engineering 192 (2019) 106239

Table 7 Table 8
Empirical formulas for hydrodynamic derivatives for sway force (Y). Empirical formulas for hydrodynamic derivatives for yaw moment (N).
Reference Empirical Formula Reference Empirical Formula
� �
Kijima et al. (1990) 2d B Kijima et al. (1990) d
Yv ¼ 0:5π þ 1:4CB Nv ¼ 2
� L L � � L �
Lee et al. (1998) d B L Lee et al. (1998) d L
Yv ¼ 0:4545 þ 0:065CB Nv ¼ 0; 23 þ 0; 0059
� L �L d L d
Yoshimura and 2d CB Yoshimura and d
Yv ¼ 0:5π þ 1:4 Nv ¼ 2
Masumoto (2012) L L=B Masumoto (2012) L
Clarke et al. (1983)
" � �2 � �# Clarke et al. (1983) � �2 � �!
d B L d d L
Yv ¼ π 1 þ 0:4CB Nv ¼ π 0:5 þ 2:4
L d d L ! L d
� �2 Smitt (1971)
Smitt (1971) d L
� �2
d L
Yv ¼ π 1:59 Nv ¼ π 0:62
" L d L d
Norrbin (1970) � �2 � �# �!
d CB B L Norrbin (1970) � �2 �
d CB B L
Yv ¼ π 1:69 þ 0:08 Nv ¼ 0:64 0:04
" L π d d π
L π d d
Inoue et al. (1981) � �2 � �# !
d 1:4 B L Inoue et al. (1981) � �2
d 2 L
Yv ¼ π 1þ CB Nv ¼
" L π d d π
L π d
Khattab (1984) � �2 � � # "
d 2:3 1:466 B 0:00102 L 2 L Khattab (1984) � �2
d 1:758 0:00768 CB L2
Yv ¼ π þ CB Nv ¼
L π π d π d d π
L π π Bd
" � �
Ankudinov (1987) d 2
� � �
B 2 B � � !#
Yv ¼ π Ky 0:25 CB 1:5CB þ 0:0008 L 2 L
L d d
!# π " d d
Ankudinov (1987) � �2 � �#
L B d d CB B L
3:45 if CB > 5 Ky ¼ 5 ; else Ky ¼ 1 Nv ¼ π 0:75 0:04
�d d CB B�
� L π d� d
Lee et al. (1998) d L Lee et al. (1998) d L
Yvvv ¼ 0:6469ð1 CB Þ þ 0:0027 Nvvv ¼ 0:0348 0:5283ð1 CB Þ
B d B d
Yoshimura and L Yoshimura and Nvvv ¼ ½ 0:69CB þ 0:66�
Yvvv ¼ 0:185 þ 0:48
Masumoto (2012) B Masumoto (2012)
Kijima et al. (1990) B Kijima et al. (1990) 2d
� �
2d 2
Yr ¼ ðm þ mx Þ 1:5CB Nr ¼ 0:54
� L � L
þ
L
Lee et al. (1998) B L Lee et al. (1998)
" � �2 #
Yr ¼ 0:115CB þ 0:0024 d d L
L d Nr ¼ 0:003724 þ 0:10446 1:393
Yoshimura and B L L d
Yr ¼ mx þ 0:5CB Yoshimura and
� �2
Masumoto (2012) L 2d 2d
" �# Masumoto (2012) Nr ¼ 0:54 þ
Clarke et al. (1983) � �2 �
d B B L " L L
Yr ¼ 0:5 þ 2:2 0:080 Clarke et al. (1983) � �2 � �#
π
L L d d d B B L
Nr ¼ π 0:25 þ 0:039 0:56
Smitt (1971)
" � �2 # L d L d
d L " � �2 #
Yr ¼ π ð 0:32Þ Smitt (1971) d L
L d Nr ¼ 0:21π
Norrbin (1970)
" � �2 � �# L d
d CB B L " �#
Yr ¼ π 0:645 þ 0:38 Norrbin (1970) � �2 �
d CB B L
L π d d Nr ¼ π 0:47 0:18
Inoue et al. (1981)
" � �2 !# L π d d
d L " #
Yr ¼ ð 0:5Þ π Inoue et al. (1981) � �2 �
d 1:04 4 d
�
L
L d Nr ¼ π
Khattab (1984)
" � �2 � � # L π π L d
d 1:0328 0:11 B 0:00004 L 2 L "
Yr ¼ π CB Khattab (1984) � �2
d 1:3192 CB
L π π d π d d Nr ¼ π 0:68228
Ankudinov (1987)
" � �2 � � � �2 !!# L π π
d B d L
Yr ¼ π 0:3 CB Yv π � � !#
L L� L d 0:00019 L 2 L
�
Lee et al. (1998) d L π d d
Yrrr ¼ 0:0233CB þ 0:0063 " � �
B d Ankudinov (1987) d 2
� � �
B 2 B
Yoshimura and Yrrr ¼ 0:051 Nr ¼ π Ky 0:03 CB 0:15CB þ
L d d
Masumoto (2012)
� � !#
Kijima et al. (1990) ð1 CB Þ L B d
Yvrr ¼ 5:95d 0:5 if CB > 5 Ky ¼ 5 ; else Ky ¼ 1
B
Lee et al. (1998)
� �
d L �d d � CB B
Yvrr ¼ 0:4346ð1 CB Þ Lee et al. (1998) d L
B d � Nrrr ¼ 0:0572 þ 0:03CB
� L d
Yoshimura and L Yoshimura and
�
0:25CB
�
Yvrr ¼ 0:26ð1 CB Þ þ 0:11 Nrrr ¼ 0:056
Masumoto (2012) B Masumoto (2012) L=B
Kijima et al. (1990) CB Kijima et al. (1990)
� �
Yvvr ¼ 1:5d 0:65 CB
B Nvrr ¼ 0:5d 0:05
Lee et al. (1998)
�
d
�
L � B �
Yvvr ¼ 0:1234CB 0:001452 Lee et al. (1998) d L
B d Nvrr ¼ 0:0005 þ 0:00594CB
Yoshimura and Yvvr ¼ 0:75 B d
Yoshimura and L
Masumoto (2012) Nvrr ¼ 0:075ð1 CB Þ 0:098
Masumoto (2012) B
" #
Kijima et al. (1990) �
B
�2 �
B
�
Nvvr ¼ 57:5 CB 18:4 CB þ 1:6
L L
(Khanfir et al., 2011): Lee et al. (1998)
" � � � � #
B B 2 L
� Nvvr ¼ 1:722 þ 22:997 CB 77:268 CB
XR ¼ ð1 tR Þ F PN sinδP þ FSN sinδS � � L L d
Yoshimura and 1:55CB
Nvvr ¼ 0:76
� Masumoto (2012) L=B
YR ¼ ð1 þ aH Þ FPN cosδP þ F SN cosδS
� �
NR ¼ xR þ aH x’H F PN cosδP þ F SN cosδS area of movable part of rudder, Λ is aspect ratio of rudder and η is ratio of
� propeller diameter to rudder span. Note that equation provided for
þ ð1 tR Þ yPR FPN sinδP þ ySR F SN sinδS (20)
rudder lift gradient coefficient (fα ) is suggested by Fujii and Tuda
Here yPR
and ySR
are the offsets of rudders from the ship’s centerline. (1961).
Parameters required for prediction of rudder normal force(s) (FN ) during
maneuver are given in Table 4. 3. Empirical equations
Here, ε; κ, lR and γ R are unknown rudder parameters that need to be
estimated empirically, numerically or experimentally. AR is the profile All input parameters should be inserted to obtain full maneuvering

6
O.F. Sukas et al. Ocean Engineering 192 (2019) 106239

Table 9 Table 14
Empirical formulas for wake fraction coefficient in straight motion, wP0 . Empirical formulations embedded in MANSIM for estimating ε.
Reference Empirical Formula Reference Empirical Formula
� � � �
Kijima et al. wP0 ¼ 0:5CB 0:05 Kijima et al. (1990) B 2 B
(1990) ε ¼ 156:2 CB þ 41:6 CB 1:76
L L
Harvald (1983) wP0 ¼ w1 þ w2 þ w3 Lee and Shin (1998) 2d
ε ¼ 2:3281 þ 8:697CB 3:78 þ 1:19CB 2 þ
b L
w1 ¼ a þ � �2
cð0:98 CB Þ3 þ 1 2d 2d
0:05 292 82:51CB
w2 ¼ L L
100ðCB 0:7Þ2 þ 1 Yoshimura and Masumoto ε ¼ 2:26*1:82ð1 wP0 Þ
0:00756 (2012)
w3 ¼ 0:18 þ
DP
þ 0:002
L
B B B
a ¼ 0:1 þ 0:149; b ¼ 0:05 þ 0:449; c ¼ 585 5027 þ
L L L
� �2 Table 15
B
11700 Empirical formulations embedded in MANSIM for estimating κ.
L
Reference Empirical Formula

Lee and Shin (1998) κ ¼ 0:6=ð 2:3281 þ 8:697CB 3:782d=L þ 1:19CB 2 þ

Table 10 292ð2d=LÞ2 82:51CB 2d=LÞ
Empirical formulas for thrust deduction factor, tP . Yoshimura and κ ¼ 0:55=ð2:26 *1:82ð1 wP0 ÞÞ
Reference Empirical Formula Masumoto (2012)
Yoshimura and Ma κ ¼ 0:55 0:8CB B=L
Kulczyk (1995) tP ¼ 0:27 (2003)
Harvald (1983) tP ¼ t1 þ t2 þ t3
e1
t1 ¼ d1 þ
f1 ð0:98 CB Þ 3 þ 1
t2 ¼ 0:02
� � Table 16
DP
t3 ¼ 2 0:04 Empirical formulations embedded in MANSIM for estimating l’R .
L
B Reference Empirical Formula
d1 ¼ 0:625 þ 0:08
L � �
B Kijima et al. (1990) l’R ¼ 2xR
e1 ¼ 0:165 0:25
L
B
� �2
B Lee et al. (1998) l’R ¼ 2xR
f1 ¼ 525 8060 þ 20300
L L Yoshimura and Masumoto (2012) l’R ¼ 0:9
Yoshimura and Ma (2003) B
l’R ¼ 1:7CB 1:2
L

Table 11
Empirical formulations embedded in MANSIM for estimating aH . performance of a ship in MANSIM. In case no input data is available for
Reference Empirical Formula hydrodynamic parameters of hull, rudder and propeller; empirical re­
Aoki et al. (2006)
lations embedded in MANSIM can be applied. These empirical equations
aH ¼ 3:349CB 2 3:293CB þ 1:059
are taken from various studies in related literature. From practical point
Lee et al. (1998) aH ¼ 2:78CB 1:922
of view, empirical formulas may be useful to assess the order of mag­
Yoshimura and B
Masumoto (2012)
aH ¼ 3:6CB
L nitudes of parameters in the preliminary design stage. The empirical
Quadvlieg (2013) aH ¼ 0:627CB 0:153 formulas embedded in MANSIM are given in the following sub-sections.
Lee and Shin (1998) 2d
aH ¼ 11:4036 þ 40:94CB 81:11 31:69CB 2 þ
L 3.1. Added mass and added moment of inertia
� �2
2d 2d
90:76 þ 79:47CB
L L Empirical formulas embedded in software for the estimation of
added masses mx , mY and added moment of inertia JZ are given in
Table 5. All formulas provided are based on the main particulars of ship
Table 12 such as; m; L; B; d and CB . It should be noted that mx is advised to be
Empirical formulations embedded in MANSIM for estimating x’H . taken approximately as 3 6% of ship mass (m) in Clarke et al. (1983),
where it is taken as 5% of ship mass in MANSIM.
Reference Empirical Formula

Aoki et al. (2006) x’H ¼ 0:45 3.2. Hydrodynamic derivatives

Lee et al. (1998) x’H ¼ 1:68CB 1:968 þ 0:5
Yoshimura and Masumoto (2012) x’H ¼ 0:4 In this section, the empirical formulas existed in the software to es­
Lee and Shin (1998) B
� �2
B timate the hydrodynamic derivatives are presented. Table 6, Table 7 and
x’H ¼ 6:054 þ 58:18 148:44
L L Table 8 show the empirical formulas provided for derivatives related to
surge force (X), sway force (Y) and yaw moment (N), respectively. Total
resistance coefficient (X’0 ) is calculated by Holtrop method (Holtrop,
Table 13 1978). Note that the accuracy of empirical relations may change with
Empirical formulations embedded in MANSIM for estimating tR . range of ship parameters and mathematical model used in the corre­
Reference Empirical Formula sponding study. Note that all empirical equations given in Tables 6–8
have been rearranged according to non-dimensionalization procedure
Aoki et al. (2006) tR ¼ 0:629CB 2 þ 0:605CB þ 0:129
used in MANSIM.
Kijima et al. (1990); Lee et al. (1998) tR ¼ 0:45 0:28CB
Yoshimura and Masumoto (2012) tR ¼ 0:39

7
O.F. Sukas et al. Ocean Engineering 192 (2019) 106239

Table 17 steering and x’H represents the application point of this lateral force
Empirical formulations embedded in MANSIM for estimating γR . component in longitudinal direction during steering. Empirical formulas
Reference Empirical Formula for rudder force parameters are based on the main particulars of ship
Kijima et al. (1990)
� � � � such as L;B;d;CB . Alongside of hull-rudder interaction coefficients, there
B 2 B
γR ¼ 22:2 CB þ 0:02 CB þ 0:68 are some necessary coefficients (ε;κ;l’R ;γ) to be known according to MMG
L L
Lee et al. (1998) B model to predict rudder normal force (FN ). Empirical relations for these
γR ¼ 2:7236CB þ 0:021
L
Yoshimura and B coefficients given in related literature are presented in Tables 14–17.
γR ¼ 2:06CB þ 0:14
Masumoto (2012) L
� �
Lee and Shin (1998) 2d 4. Graphical user interface
R ¼ 23:708
γþ 83:84CB þ 173:72 þ 71:64C2B þ
L
� �2 � �
2d 2d
157 261:11CB Graphical user interface (GUI) of MANSIM provides an easy utiliza­
L L � �
2d tion to use mathematical models and empirical approaches, and has
γR ¼ 6:8736 16:77CB þ 3:5687 þ 4:68C2B
L been designed to be a practical tool for ship maneuvering simulations.
Mathematical models embedded in MANSIM are available for SPSR
� �2 � �
2d 2d
253:14 þ 74:83CB
L L (Yasukawa and Yoshimura, 2015) and TPTR (Khanfir et al., 2011) ships
to simulate turning, zigzag and free maneuvers in calm water. Mathe­
3.3. Self-propulsion parameters matical models for SPSR and TPTR ships were explained in detail in
Section 2.2. In addition to mathematical models, turning maneuver of
Empirical relations used to estimate wake fraction coefficient in single-propeller and twin-propeller ships can be predicted based on the
straight motion (wP0 ) and thrust deduction factor (tP ) are given in empirical relations provided by Lyster and Knights (1979). Details of
Table 9 and Table 10, respectively. Empirical equations provided for this approach are given in Section 2.1. A flow diagramme of GUI of
self-propulsion parameters are based on the main particulars of ship and MANSIM is shown in Fig. 2.
propeller such as L; B; CB and DP . Apart from these parameters, open Main solver code and functions of MANSIM were developed in
water characteristics of the propeller (k0 ; k1 ; k2 ) and the propeller rev­ MATLAB environment. GUI was created with MATLAB Guide Layout
olution (nP ) must be known. Self-propulsion parameters given in MMG editor which allows user to design different kinds of user interfaces with
model can also be obtained by traditional engineering approach as some basic tools (menus, toolbars, buttons, sliders, etc.). In the input
explained in Kinaci et al. (2018). section of MANSIM, parameters required for mathematical model can be
imported from a premade “.txt” file using the corresponding icon in
3.4. Rudder parameters toolbar instead of filling the text boxes one by one. Alternatively, if user
has no input parameters except the main dimensions of ship, all inputs
Empirical relations embedded in MANSIM for estimation of hull- related to hull, propeller(s) and rudder(s) can be calculated automati­
cally by MANSIM using available empirical formulas embedded in
rudder interaction coefficients (tR ; aH ; x’H ) are given in Tables 11–13.
software. Obtained outputs can be examined on user interface or can be
Here, tR is the deduction factor of rudder resistance due to existence of
exported as “.dat” file. It is also possible to visualize the trajectory of
ship hull, aH denotes the factor of lateral force acting on the hull during

Fig. 2. Workflow scheme of MANSIM.

8
O.F. Sukas et al. Ocean Engineering 192 (2019) 106239

Fig. 3. Sample screenshot of input section of 3DOF-MMG approach in MANSIM.

ships during turning/zigzag/free maneuvers as 2D animation. Sample 5.1. Simulation of turning and zigzag maneuvers of KVLCC2
screenshots of input and output sections of MANSIM are shown in
Figs. 3–4. Pop-ups near some parameters in the input screen are used for Turning and zigzag maneuvers of full-scale KVLCC2 tanker have
selecting empirical formulas embedded in the code. Note that the pa­ been simulated by MANSIM by using hydrodynamic derivatives, rudder
rameters of propeller and rudder in TPTR option have double values and propeller parameters given in Yasukawa and Yoshimura (2015) who
different from SPSR configuration, since TPTR ships may have different have conducted circular motion and rudder force tests for the 1/110
values for each parameter of rudder and propeller. scaled model of KVLCC2. Free running results are also available for this
ship and these tests have been carried out by MARIN for the 1/45.7
5. Application of MANSIM to benchmark ships scaled model (ftp://ftp.forcetechnology.com). Maneuvering results of
35� and 35� turning circle, and 10=10, 10= 10, 20=20, 20= 20
In this section, turning and zigzag maneuvers computed by MANSIM zigzag maneuvers predicted by MANSIM were compared with results of
were validated for two benchmark ships, namely, KVLCC2 and free running tests for full-scale KVLCC2. Input parameters of MANSIM
DTMB5415 hulls. KVLCC2 tanker is a SPSR ship, while DTMB5415 has a for the prediction of maneuvering performance are given in Table 18.
TPTR configuration. Mathematical models for these type of ships are Comparison of results for port and starboard turnings are shown in
available in MANSIM and these models are described in section 2.2. Figs. 5 and 6. Note that both results are in good agreement.
Available experimental and computational results for hydrodynamic Turning maneuver indices such as advance (Ad), transfer (Tr),
derivatives, rudder force and self-propulsion parameters were used to tactical diameter (TD), steady turning diameter (STD), steady yaw rate
compare the turning and zigzag maneuvers for both ships. Numerically (SYR), steady turning speed (STS) and steady drift angle (SDA) were
obtained hydrodynamic derivatives have free surface effects taken into obtained for the full-scale KVLCC2 tanker and shown in Table 19. It can
account since both ships have relatively high Froude numbers (Kinaci be said that the results calculated by MANSIM agree well with the free
et al., 2016). Simulation results were compared with free running data running data. History of kinematical parameters was underpredicted
available in literature. Furthermore, influence of variation of hydrody­ slightly except steady yaw rate that has a perfect match. Largest dif­
namic derivatives and rudder parameters on maneuvering indices such ference is around 20% in speed reduction. However, this discrepancy
as advance, tactical diameter, overshoot angles, etc. were investigated can also be attributed to the differences in propeller rotation rate. Cal­
systematically by performing a comprehensive sensitivity analysis. culations are based on the self-propulsion point of the full-scale ship

9
O.F. Sukas et al. Ocean Engineering 192 (2019) 106239

Fig. 4. Sample screenshot of output section of turning/zigzag maneuver in MANSIM.

Table 18
Inputs of MANSIM for prediction of maneuvering abilities of full-scale KVLCC2.
Main Particulars

Lpp ðmÞ 320 BðmÞ 58 dðmÞ 20.8 CB 0.81

GMðmÞ 5.71 xG ðmÞ 11.1 rðm3 Þ 312622 Fr 0.142
Hydrodynamic Derivatives
X’0 0.022 m’x 0.022 Y’vrr 0.391 N’r 0.049
X’vv 0.040 Y’v 0.315 Y’vvr 0.379 N’rrr 0.013
X’vvvv 0.771 Y’vvv 1.607 m’y 0.223 N’vrr 0.055
X’rr 0.011 Y’r 0.083 N’v 0.137 N’vvr 0.294
X’vr 0.002 Y’rrr 0.008 N’vvv 0.030 J’z 0.011
Components of Propeller Force
DP ðmÞ 9.86 tP 0.22 k1 0.275 k0 0.293
nP ðrpsÞ 1.53 wP0 0.35 k2 0.139 xP 0.48
Components of Rudder Forces and Moment
HR ðmÞ 15.8 λ 1.827 x’H 0.464 l’R 0.71
x’R 0.50 aH 0.312 ε 1.09 γR (βR < 0) 0.395
AR ðm2 Þ 112.5 tR 0.387 κ 0.50 γR (βR > 0) 0.64

whereas the free running tests are conducted by self-propulsion point of Maneuvering indices of zigzag motion are considered as 1st and 2nd
model ship. This is most likely the primary reason of differences in the OAs for 10= 10, 10=10, 20= 20 and 20=20 maneuvers, and
turning circle trajectories. It can also be stated that the results of comparison of results are given in Table 20. Scale effect could be a
empirical approach seem to be in accordance with experiments. reason for these discrepancies since the hydrodynamic derivatives and
Comparison of various type of predicted zigzag maneuvers with free other parameters have been obtained at model scale of ship. Further­
running data for full-scale KVLCC2 is shown in Figs. 7–10. Agreement in more, a more precise prediction method may be required to improve the
the first overshoot angles (OA) are better than the second overshoot accuracy of results instead of using empirical approaches for the terms of
angles for both 10= 10 and 10=10 zigzag maneuvers. OAs in 20= added mass and added moment of inertia. It should also be noted that
20 and 20=20 zigzag maneuvers are shown in Figs. 9 and 10 and general the accuracy of prediction was found to be strongly related with the
trend of zigzag motion agree well with the free running data. It can also initial conditions (approach speed, rudder angle, propeller rate, etc.) in
be deduced from these figures that the phase shifts in trajectories are zigzag maneuvers.
generally caused by the mismatch in rudder execution times.

10
O.F. Sukas et al. Ocean Engineering 192 (2019) 106239

Fig. 5. Comparison of 35� turning maneuver results of KVLCC2 by MANSIM with free running (FR) data (Fr ¼ 0:142).

5.2. Simulation of turning and zigzag maneuvers of DTMB5415 rudder and propeller parameters in CFD analyses. Because all necessary
rudder and propeller parameters of DTMB5415 hull were estimated by
After performing a maneuvering simulation for KVLCC2, a system- CFD analyses [3] for port turning. Additionally, it was assumed that
based simulation was also carried out for full-scale DTMB5415 surface these parameters are identical/symmetrical with the starboard turning.
combatant by using MANSIM to predict its turning and zigzag maneu­ Another possible reason for the discrepancies in Figs. 12–13 may be
verabilities. Hydrodynamic derivatives and other parameters related to caused by neglecting the roll-coupled effects for DTMB5415 hull. Yaw
the propeller and rudder were computed by CFD for the 1/46.588 scaled rate of ship was slightly overestimated by MANSIM which leads to a
model of DTMB5415 hull (Sukas et al., 2019). Since DTMB5415 has a smaller turning trajectory prediction. Speed reduction percentages
TPTR configuration, propeller and rudder parameters may show dif­ during maneuver were estimated higher than the free running results for
ference due to asymmetric flow around the control surfaces during both side turnings. Turning maneuver indices of full-scale DTMB5415
maneuvering motion. For validation, system based (SB) simulation re­ are given in Table 22. The results obtained by empirical approach also
sults of 35 and -35 turning maneuvers, and 20/20 zigzag manuever seem to be underpredicted as compared to the experiments.
were compared with those of free running tests carried out by MARIN Predicted heading/rudder angles and trajectory for 20= 20
(ftp://ftp.forcetechnology.com). Input parameters of MANSIM for zigzag maneuver were also compared with free running data and shown
full-scale DTMB5415 hull are given in Table 21. Parameters with su­ in Fig. 13. Despite a good agreement of the predicted overshoot angles
perscripts “S” and “P” indicate the values for starboard and port sides, with the experiments, there is a discrepancy in trajectories due to the
respectively. Predicted results of trajectory, yaw rate and speed loss early execution time of second and third deflections of rudder in the
were compared with experimental data for 35 and 35 turning ma­ system-based simulation. Maneuvering indices of zigzag motion were
neuvers and shown in Figs. 11 and 12. compared in terms of 1st and 2nd OAs and given in Table 23. As
According to the results shown in Figs. 12 and 13, free running data mentioned previously, accuracy of results can be influenced by slight
seems to have larger difference in port and starboard turnings than those differences with the experimental procedure such as initial conditions of
of MANSIM. This discrepancy in experiments can be due to asymmetry the model. It can also be noted that the results by MANSIM are based on
between the values of rudder parameters for port and starboard turn­ self-propulsion point of the full-scale ship whereas free running tests are
ings. On the other hand, there is a large relative difference between conducted by self-propulsion point of the model ship. This can be
MANSIM and free running data in starboard turning for DTMB5415 hull. another reason for the differences in the turning and zigzag trajectories
Such difference is most likely due to symmetry assumption for the of DTMB5415 hull.

11
O.F. Sukas et al. Ocean Engineering 192 (2019) 106239

Fig. 6. Comparison of 35� turning maneuver results of KVLCC2 by MANSIM with free running (FR) data (Fr ¼ 0:142).

Table 19
Comparison of turning maneuver indices of full-scale KVLCC2.
Maneuvering Indices δ ¼ 35� δ ¼ 35�

MARIN MANSIM Yasukawa SB (2015) Lyster EMP (1979) MARIN MANSIM Yasukawa SB (2015) Lyster EMP (1979)
FR SB FR SB

Ad ( ) 3.11 3.10 3.56 2.76 3.25 3.10 3.67 2.76

Tr ( ) 1.22 1.23 1.51 1.30 1.36 1.35 1.58 1.30
TD ( ) 3.08 2.90 3.59 2.75 3.34 3.16 3.71 2.75
STD ( ) 2.48 2.05 – 2.06 2.54 2.31 – 2.06
SYR ( ) 0.30 0.30 – – 0.29 0.28 – –
STS ( ) 0.36 0.29 – 0.35 0.38 0.32 – 0.35
SDA (deg) 19.83 21.51 – – 18.59 20.24 – –

5.3. Sensitivity analysis of hydrodynamic derivatives and rudder and zigzag maneuverabilities. 35� turning and 20� = 20� zigzag
parameters maneuvers were selected as sample cases to be examined. Influence of
variation of hydrodynamic derivatives/rudder parameters on maneu­
Standard MMG model used in MANSIM consists total of 17 hydro­ vering indices such as advance (Ad), transfer (Tr), tactical diameter
dynamic derivatives in the maneuvering equations of motion. In addi­ (TD), steady turning diameter (STD) and overshoot angles (OA) were
tion, propeller and rudder parameters are included into these equations investigated. Original value of parameters was increased separately by
to simulate free running tests. Utilizing the user interface of MANSIM, 25% and these values were used in the mathematical model. Sensitivity
parametrical studies such as sensitivity analysis can be performed analysis was carried out by using the index proposed by Sen (2000). This
readily. Effect of any parameter on general maneuvering performance of analysis was adopted to examine the variation of maneuvering indices
ships can be investigated in detail. Here, a sensitivity analysis was per­ caused by changes in each hydrodynamic derivative and rudder
formed for KVLCC2 and DTMB5415 ships to investigate the effect of parameter. Sensitivity index S is represented as given in Eqn. (21):
each hydrodynamic derivative and rudder parameters on the turning

12
O.F. Sukas et al. Ocean Engineering 192 (2019) 106239

Fig. 7. Comparison of predicted heading angle, rudder angle and trajectory with free running (FR) data in 10= 10 zigzag maneuver (Fr ¼ 0:142).

Fig. 8. Comparison of predicted heading angle, rudder angle and trajectory with free running (FR) data in 10=10 zigzag maneuver (Fr ¼ 0:142).

Fig. 9. Comparison of predicted heading angle, rudder angle and trajectory with free running (FR) data in 20= 20 zigzag maneuver (Fr ¼ 0:142).

13
O.F. Sukas et al. Ocean Engineering 192 (2019) 106239

Fig. 10. Comparison of predicted heading angle, rudder angle and trajectory with free running (FR) data in 20=20 zigzag maneuver (Fr ¼ 0:142).

by the total value of this indice. For example, percentage of the effect of
Table 20
X0 in advance is calculated by,
0

Comparison of predicted maneuvering indices of full-scale KVLCC2 with free

running data in zigzag motions. SX0
%SX0 ;Ad ¼ 100⋅ 0 (22)
Maneuver OAs (deg) MARIN-FR MANSIM-SB Yasukawa-SB (2015) 0 SAd
10= 1st 9:5 7:5 8.8
Here; %SX0 ;Ad denotes the effect of X0 in percentages, SX0 is the
0

10 2nd 15 9:4 12.6 0 0

sensitivity index of X0 and SAd is the total sensitivity index value for
0

10= 10 1st 8:2 5:3 5.8

2nd 21:9 14:1 20.5 advance index. Sensitivity of each parameter as a percentage for each
20= 1st 15:1 14:2 16.1
maneuvering index for KVLCC2 and DTMB5415 are given in Tables 24
20 13.3 11:8 14.6
and 25. In these tables, parameters which have greater value than or at
2nd
least equal to 10% were considered to be highly effective as shown in
20= 20 1st 13:7 11:1 11.8
bold form. Parameters, which are between 3 and 10%, were considered
2nd 14.9 15.5 19.7
to have mediocre at best as shown in underlined form.
According to Table 24, hydrodynamic derivatives related to surge
force X have lower effect on the obtained results than that of sway force
ðR R* Þ=R*
S¼ (21) and yaw moment derivatives. Only resistance coefficient X’0 has a
ðH H * Þ=H *
moderate effect on transfer, tactical diameter and second overshoot
where, R* and H* denote the original values of maneuvering index and angles. First order derivatives seem to have a strong influence on ma­
corresponding hydrodynamic derivative/rudder parameter, respec­ neuvers especially on zigzag indices, where N’v has the highest impact
tively. R and H represent increased values by 25%. After calculating S, among all. High order and coupled derivatives have lower effect on
all these values of parameters are summed up and a total value for each maneuvering indices except N’vvr . Terms of added mass and added
indice is obtained. Then, sensitivity index for each parameter is divided moment of inertia have a moderate effect on first overshoot angle, while

Table 21
Inputs of MANSIM for prediction of maneuvering abilities of full-scale DTMB5415.
Main Particulars

Lpp ðmÞ 142 BðmÞ 19.06 dðmÞ 6.15 CB 0.507

GMðmÞ 1.95 xG ðmÞ 0.652 rðm3 Þ 8424.4 Fr 0.248
Hydrodynamic Derivatives
X’0 0.016 Y’v 0.294 Y’vvr 1.506 N’rrr 0.048
X’vv 0.182 Y’vvv 1.174 m’y 0.108 N’vrr 0.218
X’rr 0.028 Y’r 0.047 N’v 0.162 N’vvr 0.800
X’vr 0.093 Y’rrr 0.052 N’vvv 0.225 J’z 0.008
m’x 0.007 Y’vrr 0.784 N’r 0.045
Components of Propeller Force and Moment
DP;S
P ðmÞ
6.15 tP;S
P
0.210 kP;S
0
0.398 kP;S
1
0.299
� ’ P;S �
nP;S
P ðrpsÞ
1.65 wP;S 0.073 kP;S 0.141 P;S
x’P , �y � 0.462, 0.244
P0 2 P
Components of Rudders Force and Moment
� P;S �
HP;S
R ðmÞ
4.38 aP;S
H
0.086 εP;S 0.93; 1.00 x’R
P;S
, �y’R � 0.472, 0.267

λP;S 1.26 tP;S

R
0.440 κ P;S 0.62; 0.70 γP;S
R (βR < 0) 0.53; 0.37

AP;S 2
R ðm Þ
15.4 x’H
P;S 0.437 l’R
P;S 0.944 γP;S
R (βR > 0)
0.37; 0.53

14
O.F. Sukas et al. Ocean Engineering 192 (2019) 106239

Fig. 11. Comparison of 35� turning maneuver results of DTMB5415 by MANSIM with free running (FR) data (Fr ¼ 0:25).

only m’y has a moderate effect on steady turning radius. Rudder pa­ variation of parameters in MMG model. Terms of added mass and added
rameters have also major influence on maneuvering indices. Particu­ moment of inertia have relatively low effect than the other parameters
larly, variation of ε and κ greatly affects the turning and zigzag especially in turning motion. Thus, they are generally estimated based
maneuvers of ship. It can briefly be stated for sensitivity analysis of on empirical formulas or charts. Results of sensitivity analysis for both
KVLCC2 tanker that the overshoot angles are more sensitive to the ships are summarized in Table 26.
variations of hydrodynamic derivatives and rudder parameters than
those of turning maneuver. In addition, advance distance and steady 5.4. Effect of linear derivatives on maneuvering indices
turning diameter seem to be the least and most influenced indices by the
variation of parameters in turning maneuver, respectively. In the previous section, sensitivity analyses showed the significance
Table 25 shows the sensitivity indices of turning and zigzag ma­ of each parameter on maneuvering indices. However, how these pa­
neuver of DTMB5415 hull. Note that the parameters of port and star­ rameters change maneuvering indices are still unknown. This section is
board rudders of DTMB5415 are increased together. Similar to KVLCC2 devoted to the effect of linear derivatives on maneuvering indices of
hull, N’v was found to be the most dominant linear derivative in both KVLCC2 to reveal how they change turning and zigzag maneuvers. To do
turning and zigzag maneuvers of DTMB5415 hull. Nonlinear derivatives this, linear derivatives (Yv , Yr , Nv and Nr ) were changed by a consider­
able amount and results were plotted. Only one derivative was changed
have small effects except N’rrr which affects the manuevers moderately.
at a time to particularly show corresponding effect. Some interesting
Coupled derivatives related to the yaw moment (N’vvr ; N’vrr ) have rela­
observations were noted:
tively major impact as compared to those of surge and sway forces. X’vv ,
X’rr , X’vr , Y’rrr and m’x have almost no influence on the indices of turning � There is an optimum value for Yv to minimize advance and steady
maneuver. Rudder parameters have a significant effect on the maneu­ turning diameter.
vers rather than the hydrodynamic derivatives. Additionally, similar to � Yr also has an optimum value for minimizing steady turning
KVLCC2 case, advance distance was found to be the least influenced diameter.
index from the variation of hydrodynamic derivatives and rudder pa­ � Decrease in the magnitude of Nv results in increase in some
rameters. On the whole, it can be noted for both ships that the de­ maneuvering indices such as advance, transfer, steady turning
rivatives based on yaw moment are highly influential on turning and diameter and tactical diameter of turning circle test. On the other
zigzag maneuvers. Indices of zigzag motion are oversensitive to the

15
O.F. Sukas et al. Ocean Engineering 192 (2019) 106239

Fig. 12. Comparison of 35� turning maneuver results of DTMB5415 by MANSIM with free running (FR) data (Fr ¼ 0:25).

Fig. 13. Comparison of predicted heading angle, rudder angle and trajectory of full-scale DTMB5415 with free running (FR) data in 20= 20 zigzag maneuver
(Fr ¼ 0:25).

16
O.F. Sukas et al. Ocean Engineering 192 (2019) 106239

Table 22
Comparison of turning maneuver indices of full-scale DTMB5415.
Maneuvering Indices δ ¼ 35� δ ¼ 35�

MARIN FR MANSIM SB Carrica CFDB (2013) Lyster EMP (1979) MARIN FR MANSIM SB Carrica CFDB (2013) Lyster EMP (1979)

Ad ( ) 2.71 2.59 2.90 2.55 3.19 2.40 2.90 2.55

Tr ( ) 1.46 1.35 1.58 1.15 1.33 1.17 1.58 1.15
TD ( ) 3.65 3.48 3.87 2.83 3.60 3.14 3.87 2.83
STD ( ) 3.66 3.53 – 2.69 3.75 3.19 – 2.69
SYR ( ) 0.38 0.41 – – 0.39 0.43 – –
STS ( ) 0.75 0.72 – 0.62 0.74 0.68 – 0.62

Table 23 Table 25
Comparison of predicted maneuvering indices of full-scale DTMB5415 with free Sensitivity analysis of hydrodynamic derivatives and rudder parameters of
running data in 20= 20 zigzag maneuver. DTMB5415 in turning and zigzag maneuvers.
Maneuver OAs (deg) MARIN-FR MANSIM-SB Carrica-CFDB Parameters Turning Maneuver Indices Zigzag Maneuver Indices

20= 1 st 4:70 5:12 7.30 Ad Tr TD STD 1st OA 2nd OA

20 2nd 4.80 6:34 7.20
X’0 0.82 0.00 0.00 0.00 2.15 2.85
X’vv 0.00 0.00 0.00 0.00 0.66 0.67
X’rr 0.00 0.00 0.00 0.00 0.83 0.67
Table 24 X’vr 0.00 0.00 0.00 0.00 0.83 1.34
Sensitivity analysis of hydrodynamic derivatives and rudder parameters of Y’v 2.45 4.16 0.00 0.71 2.48 1.68
KVLCC2 in turning and zigzag maneuvers.
Y’vvv 0.82 0.83 0.00 0.36 0.33 0.84
Parameters Turning Maneuver Indices Zigzag Maneuver Indices 0.82 1.66 0.35 0.36 1.49 0.50
Y’r
Ad Tr TD STD 1st OA 2nd OA Y’rrr 0.00 0.00 0.00 0.00 0.33 0.50
Y’vrr 2.45 2.49 0.35 0.71 0.66 0.50
X’0 2.74 2.94 3.17 1.68 0.88 3.99
Y’vvr 3.27 1.66 1.75 1.79 1.49 0.67
X’vv 0.46 0.00 0.32 0.42 0.35 0.20
N’v 13.88 10.80 14.39 14.29 17.36 15.10
X’vvvv 0.46 0.00 0.32 0.84 0.35 0.40
N’vvv 1.22 0.83 1.05 1.07 0.83 0.17
X’rr 0.46 0.00 0.32 0.84 0.35 0.40
N’r 11.84 9.14 10.88 11.07 8.60 7.55
X’vr 0.46 0.00 0.00 0.00 0.35 0.20
N’rrr 3.27 3.32 3.51 3.57 0.50 2.01
Y’v 0.00 5.88 2.54 0.84 8.13 5.59
N’vrr 6.94 5.82 7.72 7.14 2.98 3.52
Y’vvv 0.46 0.88 0.00 0.42 0.18 1.00
N’vvr 9.39 9.97 12.63 12.14 3.31 4.70
Y’r 0.46 3.82 2.22 0.42 5.30 2.59
m’x 0.00 0.00 0.00 0.00 0.83 0.67
Y’rrr 0.46 0.00 0.00 0.00 0.35 0.20
m’y 0.00 0.83 0.35 0.00 1.49 0.67
Y’vrr 0.46 2.06 0.32 0.42 0.53 0.80
J’z 1.22 0.00 0.00 0.00 4.13 3.02
Y’vvr 0.46 0.88 0.00 0.00 0.18 0.40
aH 0.82 0.00 0.35 0.00 0.00 1.17
N’v 14.16 13.53 13.02 6.95 34.45 29.14
x’H 0.00 0.83 0.35 0.36 0.33 0.50
N’vvv 0.46 0.00 0.32 0.00 0.53 0.40
tR 3.27 2.49 3.16 3.21 0.17 0.50
N’r 15.53 12.35 12.38 8.21 15.02 13.97
ε 10.20 0.00 2.46 2.50 16.03 18.29
N’rrr 0.46 0.88 1.27 2.11 0.00 0.40
κ 0.82 0.83 0.00 0.36 3.14 6.21
N’vrr 1.37 2.06 2.54 2.95 0.18 1.00
γR 15.92 26.32 24.21 23.93 17.02 15.27
N’vvr 2.74 4.71 6.98 6.53 1.41 3.59 10.61 18.01 16.49 16.43 12.07 10.40
l’R
m’x 0.46 0.88 0.00 0.42 2.30 1.20
m’y 0.46 0.00 2.22 4.84 1.94 0.20
J’z 1.83 0.00 0.32 0.00 3.36 4.39 above have primary importance for turning and zigzag tests. Graphs for
aH 4.11 2.94 2.54 2.53 1.41 2.79 Yv are given in Fig. 14, Yr in Fig. 15, Nv in Fig. 16 and Nr in Fig. 17.
x’H 3.65 2.06 2.22 2.53 1.77 2.40
tR 0.00 0.88 0.95 2.11 0.53 0.80 6. Conclusions
ε 31.96 26.76 27.30 28.00 6.01 10.18
κ 11.42 9.71 11.75 19.79 3.71 6.59 In this study, a user-friendly ship maneuvering code based on MMG
γR 2.74 3.82 4.44 4.21 6.71 0.80 mathematical model has been introduced. Graphical user interface of
l’R 1.83 2.94 2.54 2.95 3.71 6.39 code allows to make an easy and simple changes to hydrodynamic de­
rivatives or propeller/rudder parameters. This software provides a basis
for researchers to experiment with all coefficients/parameters and to
hand, it enhances zigzag test by decreasing 1st and 2nd overshoot have a better understanding of ship maneuvering phenomena which
angles. involves a dynamic and complex background. It also contains many
� Decrease in the magnitude of Nr results in decrease in some empirical relations suggested by many researchers in the field of ship
maneuvering indices such as advance, transfer, steady turning maneuvering. MANSIM is considered to be helpful especially for sensi­
diameter and tactical diameter of turning circle test. Nevertheless, tivity analysis on maneuvering. A SPSR and a TPTR ship have been
overshoot angles in zigzag test considerably increase. investigated and parameters that have significant effect on each
maneuvering indice have been found. Findings can be listed as follows:
Observations from these graphs may be elaborated but points listed

17
O.F. Sukas et al. Ocean Engineering 192 (2019) 106239

Table 26
Parameters that have high or mediocre impact for turning circle and zigzag maneuvers.
Impact Level Turning Maneuver Zigzag Maneuver

KVLCC2 DTMB5415 KVLCC2 DTMB5415

High Mediocre High Mediocre High Mediocre High Mediocre

Hydrodynamic derivatives N’v , N’r N’vvr N’v , N’r , N’vvr Y’v ; Y’vvr , N’rrr , N’vrr N’v , N’r Y’v , Y’r , J’z N’v N’r , N’vvr ,
N’vrr , J’z
Rudder parameters ε, κ γR ε, γR , l’R tR ε κ, γR , l’R ε, γR , l’R κ

Fig. 14. Effect of Yv on the behavior of turning circle test indices (a) through (c). On zigzag test (d).

Fig. 15. Effect of Yr on the behavior of turning circle test indices (a) through (c). On zigzag test (d).

Fig. 16. Effect of Nv on the behavior of turning circle test indices (a) through (c). On zigzag test (d).

� Nv and Nr are highly effective on maneuvering indices. Although the last two statements are only applicable for DTMB5415
� ε is a highly effective rudder parameter in ship maneuvering. ship, it is considered that they are valid for TPTR ships in general.
� Coupled terms have a non-negligible importance for DTMB5415 However, more research is needed to support these statements. An
ship. additional subsection was devoted to the effects of linear derivatives on
� Rudder parameters have higher importance in zigzag motion of maneuvering indices. It was found that there was no linear relationship
DTMB5415 ship. between the linear derivatives and maneuvering indices.

18
O.F. Sukas et al. Ocean Engineering 192 (2019) 106239

Fig. 17. Effect of Nr on the behavior of turning circle test indices (a) through (c). On zigzag test (d).

Currently, MANSIM does not consider external disturbances. Effects Khattab, O., 1984. Multiple Regression Analyses of the Hydrodynamic Derivatives for
Maneuvering Equations. BSRA Report, W1090, December.
of wind, wave and current are also to be included into the software.
Kijima, K., Katsuno, T., Nakiri, Y., Furukawa, Y., 1990. On the maneuvering performance
Future studies can also add quadratic model to the code, which is only of a ship with the parameter of loading condition. J. Soc. Nav. Archit. Jpn. 168,
working with cubic model in the current form. Furthermore, 3-DOF 141–148.
MMG model will be expanded to a 4-DOF model that includes roll- Kinaci, O.K., Gokce, M.K., Alkan, A.D., Kukner, A., 2018. On self-propulsion assessment
of marine vehicles. Brodogradnja 69 (4), 29–51.
coupled effects. Kinaci, O.K., Sukas, O.F., Bal, S., 2016. Prediction of wave resistance by Reynolds-
Averaged Navier-Stokes based computational fluid dynamics approach. J. Eng.
Acknowledgements Maritime Environ. 230 (3), 531–548.
Kulczyk, J., 1995. Propeller-hull interaction in inland navigation vessel. Trans. Built
Environ. 11, 73–89.
This study is a part of PhD thesis of the first author, under progress at Lee, H.Y., Shin, S.S., Yum, D.J., 1998. Improvement of prediction technique of the ship’s
Istanbul Technical University. First author also would like to thank Mrs. maneuverability at initial design stage. J Soc Naval Archit Korea 35, 46–53 (in
Japanese).
Hulya Sukas for helping with design and general layout of MANSIM. Lee, H.Y., Shin, S.S., 1998. The prediction of ship’s manoeuvring performance in initial
design stage. Practical Design Ships Mob. Units 633–639.
References Liu, J., Hekkenberg, R., Quadvlieg, F., Hopman, H., Zhao, B., 2017. An integrated
empirical manoeuvring model for inland vessels. Ocean. Eng. 137, 287–308.
Lyster, C.A., Knights, H.L., 1979. Prediction equations for ships’ turning circles. Trans.
Abkowitz, M.A., 1964. Lectures on Ship Hydrodynamics Steering and Maneuverability.
NECIES 95, 217–232.
Technical Report, Technical Report Hy-5. Hydro and Aerodynamic Laboratory,
Norrbin, N.H., 1970. Theory and observation on the use of a mathematical mode for ship
Lyngby, Denmark.
maneuvering in deep and conned waters. In: Proceedings of the 8th Symposium on
Ankudinov, V., 1987. Controllability Assessment Using Prediction Techniques. Design
Naval Hydrodynamics. Pasadena, CA, 807–904.
Workbook on Ship Maneuverability, SNAME, New Jersey.
Obreja, D., Nabergoj, R., Crudu, L., Pacuraru-Popoiu, S., 2010. Identification of
Aoki, I., Kijima, K., Furukawa, Y., Nakiri, Y., 2006. On the prediction method for
hydrodynamic coefficients for maneuvering simulation model of a fishing vessel.
maneuverability of a full scale ship. J. Jpn Soc. Nav. Archic. Ocean Eng. 3, 157–165.
Ocean. Eng. 37, 678–687.
Bhushan, S., Xing, T., Carrica, P., Stern, F., 2009. Model and full-Scale URANS
Ogawa, A., Kasai, H., 1978. On the mathematical model of manoeuvring motion of ships.
simulations of Athena resistance, powering, seakeeping, and 5415 maneuvering.
Int. Shipbuild. Prog. 25 (292), 306–319. December.
J. Ship Res. 53 (4), 179–198.
Ohashi, K., Kobayashi, H., Hino, T., 2018. Numerical simulation of the free-running of a
Broglia, R., Dubbioso, G., Durante, D., Di Mascio, A., 2015. Turning ability analysis of a
ship using the propeller model and dynamic overset grid method. Ship Technol. Res.
fully appended twin screw vessel by CFD. Part 1: single rudder configuration. Ocean.
65 (3), 153–162.
Eng. 105, 275–286.
Quadvlieg, F., 2013. Theoretische Berekening Van Simulatie Modellen Voor
Carrica, P.M., Ismail, F., Hyman, M., Bhushan, S., Stern, F., 2013. Turn and zigzag
Binnenvaartschepen Ten Behoeve Van Maatgevende Manoeuvres (In Dutch). Tech.
maneuvers of a surface combatant using a URANS approach with dynamic overset
Rep. Maritime Research Institute Netherlands (MARIN), Wageningen, The
grids. J. Mar. Sci. Technol. 18, 166–181.
Netherlands.
Clarke, D., Gedling, P., Hine, G., 1983. Application of manoeuvring criteria in hull design
Sakamoto, N., Carrica, P.M., Stern, F., 2012. URANS simulations of static and dynamic
using linear theory. Trans. R. Inst. Nav. Archit. 125, 45–68.
maneuvering for surface combatant: Part 1. Verification and validation for forces,
Cura-Hochbaum, A., 2011. On the numerical prediction of ship’s manoeuvring
moment, and hydrodynamic derivatives. J. Mar. Sci. Technol. 17 (4), 422–445.
behaviour. Ship Sci. Technol. 5, 27–39.
Sen, D., 2000. A study on sensitivity of maneuverability performance on the
Duman, S., Bal, S., 2017. Prediction of the turning and zig-zag maneuvering performance
hydrodynamic coefficients for submerged bodies. J. Ship Res. 44, 186–196.
of a surface combatant with URANS. Ocean Sys. Eng.-An Int. J. 7 (4), 435–460.
Smitt, W.L., 1971. Steering and manoeuvring full scale and model tests (Parts-2).
Duman, S., Bal, S., 2019. A quick-responding technique for parameters of turning
European Shipbuilding 20, 1.
maneuver. Ocean. Eng. 179, 189–201.
Sukas, O.F., Kinaci, O.K., Bal, S., 2019. System-based prediction of maneuvering
Fang, M.C., Luo, J.H., Lee, M.L., 2005. A nonlinear mathematical model for ship turning
performance of twin-propeller and twin-rudder ship using a modular mathematical
circle simulation in waves. J. Ship Res. 49 (2), 69–79.
model. Appl. Ocean Res. 84, 145–162.
Fujii, H., Tuda, T., 1961. Experimental research on rudder performance (2). J. Soc. Nav.
Sukas, O.F., Kinaci, O.K., Bal, S., 2017. A review on prediction of ship maneuvering
Archit. Jpn. 110, 31–42 (in Japanese).
performance, part 1. GMO J. Ship Marine Technol. 210, 37–75 (In Turkish).
Guo, H.P., Zou, Z.J., 2017. System-based investigation on 4-DOF ship maneuvering with
Sukas, O.F., Kinaci, O.K., Bal, S., 2017. A review on prediction of ship maneuvering
hydrodynamic derivatives determined by RANS simulation of captive model tests.
performance, part 2. GMO J. Ship Marine Technol. 210, 76–105 (In Turkish).
Appl. Ocean Res. 68, 11–25.
Sutulo, S., Soares, C.G., 2014. An algorithm for offline identification of ship maneuvering
Harvald, S.A., 1983. Resistance and Propulsion of Ships. John Wiley & Sons.
mathematical models from free-running tests. Ocean. Eng. 79, 10–25.
He, S., Kellett, P., Yuan, Z., Incecik, A., Turan, O., Boulougouris, E., 2016. Maneuvering
Toxopeus, S., Sadat-Hosseini, H., Visonneau, M., Guilmineau, E., Yen, T.G., Lin, W.M.,
prediction based of CFD generated derivatives. J. Hydrodyn. 28 (2), 284–292.
Grigoropoulos, G., Stern, F., 2018. CFD, potential flow and system-based simulations
Holtrop, J., 1978. Statistical data for the extrapolation of model performance tests. Int.
of fully appended free running 5415M in calm water and waves. Int. Shipbuild. Prog.
Shipbuild. Progress. 25 (588), 122–126.
65, 227–256.
Inoue, S., Hirano, M., ve Kijima, K., 1981. Hydrodynamic derivatives on ship
Xu, H., Vassani, H., Soares, C.G., 2019. Uncertainty analysisof the hydrodynamic
manoeuvring. Int. Shipbuild. Prog. 28 (321) (May).
coefficients estimation of a nonlinear manoeuvring model based on planar motion
ITTC, 2014. Recommended Procedures and Guidelines, 7.5-02-06-02 rev. 04, Captive
mechanism tests. Ocean. Eng. 173, 450–459.
Model Test Procedure.
Yasukawa, H., Sakuno, R., Yoshimura, Y., 2019. Practical maneuvering simulation
Kang, D., Nagarajan, V., Hasegawa, K., Sano, M., 2008. Mathematical model of single-
method of ships considering the roll-coupled effect. J. Mar. Sci. Technol. 1–17. htt
propeller twin-rudder ship. J. Mar. Sci. Technol. 13, 207–222.
ps://doi.org/10.1007/s00773-019-00625-4.
Khanfir, S., Hasegawa, K., Nagarajan, V., Shouji, K., Lee, S.K., 2011. Manoeuvring
Yasukawa, H., Yoshimura, Y., 2015. Introduction of MMG standard method for ship
characteristics of twin-rudder systems: rudder-hull interaction effect on the
maneuvering predictions. J. Mar. Sci. Technol. 20, 37–52.
manoeuvrability of twin-rudder ships. J. Mar. Sci. Technol. 16 (4), 472–490.

19
O.F. Sukas et al. Ocean Engineering 192 (2019) 106239

Yoon, H., Simonsen, C.D., Benedetti, L., Longo, J., Toda, Y., Stern, F., 2015. Benchmark Yoshimura, Y., Ma, N., 2003. Manoeuvring prediction of fishing vessels. In: International
CFD validation data for surface combatant 5415 in PMM maneuvers-Part 1: force/ Conference on Marine Simulation and Ship Maneuverability (MARSIM’03).
moment/motion measurements. Ocean. Eng. 109, 705–734. Kanazawa, Japan. Aug.
Yoshimura, Y., Masumoto, Y., 2012. Hydrodynamic database and maneuvering Zhang, C., Liu, X., Wan, D., Wang, J., 2019. Experimental and numerical investigations of
prediction method with medium high-speed merchant ships and fishing vessels. In: advancing speed effects on hydrodynamic derivatives in MMG model, part 1: Xvv, Yv,
International Conference on Marine Simulation and Ship Maneuverability Nv. Ocean. Eng. 179, 67–75.
(MARSIM’12). Singapore. Apr. Zhang, X.G., Zou, Z.J., 2013. Estimation of the hydrodynamic coefficients from captive
Yoshimura, Y., 2005. Mathematical model for manoeuvring ship motion(MMG Model). model test results by using support vector machines. Ocean. Eng. 73, 25–31.
In: Workshop on Mathematical Models for Operations Involving Ship-Ship Zhou, Z., Yan, S., Feng, W., 1983. Manoeuvring prediction of multiple-purpose cargo
Interaction, Tokyo, Japan, pp. 1–6 (August). ships (in Chinese). Ship Eng 6, 21–36.

20