See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/315935520

Simulation-Based Analysis of Ship Motions in Short-Crested Irregular Seas

Article  in  Journal of ETA Maritime Science · January 2017

DOI: 10.5505/jems.2017.83803

CITATIONS READS

3 1,211

3 authors, including:

Kıvanç Ali Anil Devrim Bülent Danışman

Istanbul Technical University Istanbul Technical University
5 PUBLICATIONS   11 CITATIONS    36 PUBLICATIONS   144 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

waterline parabolization View project

ESTHETICS (Energy Saving Techniques for Energy Efficient Vessels and Emission Reduction towards Green Shipping) View project

All content following this page was uploaded by Devrim Bülent Danışman on 29 May 2017.

The user has requested enhancement of the downloaded file.

 Original Research (AR) Anıl et al./ JEMS, 2017; 5(1): 19-38

Received: 16 October 2016 Accepted: 30 December 2016 DOI ID: 10.5505/jems.2017.83803

Journal of ETA Maritime Science

J EMS
OURNAL

Corresponding Author: Kıvanç Ali ANIL

Simulation-Based Analysis of Ship Motions in Short-Crested Irregular

Seas
Kıvanç Ali ANIL1, Devrim Bülent DANIŞMAN1, Kadir SARIÖZ1
1
İstanbul Teknik Üniversitesi, Gemi İnşaatı ve Deniz Bilimleri Fakültesi, anil@itu.edu.tr; bulent.
danisman@itu.edu.tr; sarioz@itu.edu.tr

Abstract
Demonstration of the seakeeping calculation results other than polar diagrams and Cartesian plots
is important during the initial and detail design stages of naval platforms due to the necessity of
numerical simulations (time series data) for the design and validation of the systems on board. These
time series simulations are called as “real time computer experiments”. Similar simulation algorithms
for ship motions and wave elevation are also used by ship-handling simulators for realistic visualization.
The goal of this paper is to create a basis for the simulation-based analysis of ship motions and wave
elevation for future design and validation studies for both the naval platform itself and the systems on
board. The focus of this paper is the clarification of the theoretical background of this process, i.e. all
formulations required to create and validate a ship motion and wave surface simulation are given in
detail. The results of this study may also be used in ship-handling simulators or helicopter landing on
ship simulations.

Keywords: Seakeeping Simulation, Ship Motions, Short-Crested Irregular Seas.

Kısa Dalga Tepeli Karışık Denizlerdeki Gemi Hareketlerinin Simülasyona Dayalı

Analizi

Öz
Denizcilik hesap sonuçlarının polar diyagramlar ve Kartezyen grafikler dışında gösterimi, deniz
platformlarının başlangıç ve detay tasarım aşamalarında gemideki sistemlerin tasarımı ve doğrulaması
için sayısal simülasyonların (zaman serisi veri) gerekliliği nedeniyle önem arz etmektedir. Bahse konu
zaman serisi simülasyonları “gerçek zamanlı bilgisayar deneyleri” olarak adlandırılmaktadır. Gemi
hareketleri ve dalga yüksekliği için benzer simülasyon algoritmaları, gemi kullanma simülatörleri
tarafından gerçekçi görselleştirme için de kullanılmaktadır. Bu makalenin amacı, deniz platformunun
kendisi ve üzerindeki sistemler için gelecekteki tasarım ve doğrulama çalışmalarında kullanılmak üzere,
gemi hareketlerinin ve dalga yüksekliğinin simülasyona dayalı analizi için bir temel oluşturmaktır.
Makalenin odak noktası bu sürecin teorik altyapısının açıklığa kavuşturulması olup, bir gemi hareketini
ve dalga yüzey simülasyonunu oluşturmak ve doğrulamak için gerekli tüm formülasyonlar ayrıntılı
olarak verilmektedir. Çalışmanın sonuçları gemi kullanma simülatörlerinde veya gemiye konuşlu
helikopter simülasyonlarında da kullanılabilir.

Anahtar Kelimeler: Denizcilik Simülasyonu, Gemi Hareketleri, Kısa Dalga Tepeli Karışık Deniz.

19
 © UCTEA The Chamber of Marine Engineers Journal of ETA Maritime Science

1. Introduction Language (VRML) by Adanır et al. [8].

Demonstration of the seakeeping The current simulation works real
calculation results is important during the time and has no time limitation (infinite
initial and detail design stages of naval simulation duration). Time step of the
platforms. In order to confirm the fulfillment “real time simulation” is the time passed
of the mission criteria sets for the platform, since the last frame rendered, i.e. depends
polar diagrams and Cartesian plots of root on the frame rate. The simulation can be
mean square (rms) or significant values accelerated or decelerated using predefined
of responses are commonly used, but large or small fixed time steps for analysis
numerical simulations (time series data, purpose. The simulation has also real
time history) are also needed for the design time controllable parameters for the ship
and validation of the systems on board. and the environment. The controllable
These systems mainly include active roll ship parameters are the speed and rudder
stabilizer fins, active motion interceptors, angle. The controllable environment
rudder roll stabilization, helicopter parameters are the sea area, sea-state
securing & traversing, helicopter visual and wave direction. Sea areas are defined
landing aid, and all sensors & weapons. by appropriate sea-spectra. Sea-states
The simulated time series of ship motions are defined by the World Meteorological
and wave elevation may also be used for Organization (WMO) standard sea-state
the naval platform itself to derive the shear code (Douglas Scale).
force and bending moment distribution of Ship responses in short-crested irregular
ship sections using ship hydroelasticity seas are calculated for the selected sea-state
theory, as investigated by Belik [1]. These and sea-spectrum (sea area). There is also a
time series simulations are called as long-crested irregular seas option for the
“real time computer experiments” [2, 3]. analysis purpose. Generation of the time
Similar simulation algorithms for ship series data for the seakeeping simulation
motions and wave elevation are also used (ship responses) is performed by the similar
by ship-handling simulators for realistic superposition algorithm indicated by Perez
visualization, as presented by Yeo et al. [4]. [9], using pre-calculated frequency domain
The results of this study may also be used Response Amplitude Operators (RAOs) for
in mooring simulations [5] or helicopter the selected sea-state and sea-spectrum.
landing on ship simulations [6]. The algorithm is based on the St. Denis and
Pierson’s [10] application of the principle of
2. Methodology superposition to the ship motion problem.
An in-house ship motion and wave The three Cartesian coordinate systems
surface simulation (in short-crested defined by Beck et al. [11] are used
irregular seas) code is developed in a to determine the ship motions for the
C++ programming language which can seakeeping simulation. These are the earth-
be controlled by the user in real time fixed ( x0 , y0 , z 0 ), moving(inertial) ( x , y , z
on screen. The visualization is achieved ), and the body-fixed ( x , y , z ) system. The
using Object-Oriented Graphics Rendering x0 − y0 plane of the earth-fixed system and
Engine (OGRE) Software Development Kit the x − y plane of inertial system coincide
[7], as shown in Figure 1. The simulation with the calm water level. For the body-
uses the basic methodology of the real time fixed coordinate system which moves with
uncontrollable ship motion simulation all the motions of the ship, the x − y plane
(in long-crested irregular seas) which coincides with the ship’s calm water-plane,
was written in Virtual Reality Modelling with z axis normal to it upwards, i.e. the

20
 Anıl et al./ JEMS, 2017; 5(1): 19-38

Figure 1. Ship Motion and Wave Surface Simulation

origin is in the calm water-plane at any purpose. The wave surface is defined in the
convenient point between fore and aft in earth-fixed coordinate system. The wave
the center plane, such as amidships. In this visualization is performed using the add-
study the z axis is through the center of on library of OGRE Software Development
gravity of the ship, as Salvesen, Tuck and Kit named “Hydrax Version 0.5.1” [13]. In
Faltinsen [12] specified. The orientation order to create an infinite wave surface,
of the body-fixed system relative to the “Projected Grid” option of Hydrax based on
inertial system gives the translational and Johanson [14] thesis is used. New Hydrax
rotational motions of the ship. The inertial subroutines are written for the desired sea-
to earth-fixed system transformation is spectra.
given by: As a beginning of a future maneuvering
x0= x + VS t
simulation study, smooth turning of the ship
is achieved by using the interpolated values
y0 = y
(1) of tactical diameter with respect to the
z0 = z rudder angle and ship speed. The heading
where VS , and represents the ship angle and the orientation of the earth-
speed and time respectively. There are fixed and inertial coordinate system are
several cameras (views) in the simulation updated when the rudder angle is changed.
located on all coordinate systems, which Coordinate system transformation for the
makes analysis of ship motions possible maneuvering simulation is not in the scope
from different angles. The bridge view of of this study.
the body-fixed system can be used for a The current simulation does not fit
ship-handling simulator. simulation interoperability standards like
The time series data for the short-crested High Level Architecture (HLA).
irregular sea surface is generated by the
well-known superposition algorithm that 3. Theoretical Background
follows St. Denis and Pierson [10], for the The focus of this paper is the explaining
selected sea-state and sea-spectrum. There of the theoretical background of a ship
is also a long-crested option for analysis motion and wave surface simulation.

21
 © UCTEA The Chamber of Marine Engineers Journal of ETA Maritime Science

Although the readers are aware of most Table 1. Sea-States, after [15] & [27]
the formulations given in this paper, all Significant Mean Values of
formulations including Sea-State, Sea- Sea- Wave Height Significant Wave
Spectra, Directional Spectrum, Derived State Range Height Range
Responses which are required to create such (in meters) (in meters)
a simulation are given in the following sub- 0-1 0,00 – 0,10 0,05
sections. 2 0,10 – 0,50 0,30
3 0,50 – 1,25 0,88
3.1. Sea-State
4 1,25 – 2,50 1,88
Sea-states are defined by the WMO
standard sea-state code (Douglas Scale) i.e. 5 2,50 – 4,00 3,25
in the WMO Code Table 3700 [15]. This code 6 4,00 – 6,00 5,00
table is used for recording the sea-state by an 7 6,00 – 9,00 7,50
observer. The wave height values recorded 8 9,00 – 14,00 11,50
by the observer refer to the well-formed
9 ≥ 14,00 ≥ 14,00
wind waves of the open sea [15]. On the
other hand, the definition of the “significant
wave height” according to STANAG 4154 [16]
where ω is the circular wave frequency in
is as follows: “If all the wave heights (peak to
radians per second, A and B are constants.
trough) of a wave record are measured, the
If no information except the significant
significant wave height is the mean value of
wave height H 1 / 3 is available about the sea
the highest one-third of all the wave heights.
area then,
It is approximately equal to the wave height
estimated by an observer”. The statistical (3)
term “significant waves" is first introduced
by Sverdrup and Munk [17] as “the mean Here, g is the acceleration of gravity. If
height of the highest one-third of all the average period T1 is also known then,
waves” (cited by Longuet-Higgins [18]).
H1/2 3 691
Sverdrup and Munk [17] implied= that “a A 173 = 4
, B (4)
careful observer who attempts to establish T1 T14
the character of the higher waves will tend Equation (2) along with Equation (3) is a
to record the significant waves”. Therefore form of the “Pierson-Moskowitz Spectrum”
the wave height range of this sea-state code which was first introduced by Pierson
(WMO Code Table 3700) can be regarded as and Moskowitz [20] in 1964 and called as
significant wave height range for the naval “ITTC One Parameter Spectrum”. Equation
engineering purpose, as shown in Table 1. (2) along with Equation (4) is a form of
the “Bretschneider Spectrum” which was
3.2. Sea-Spectra first introduced by Bretschneider [21] in
The random nature of an irregular sea 1959 and called as “ITTC Two-Parameter
can be described mathematically by the sea- Spectrum”. This spectrum is also called
spectrum formulas. 12th International Towing as “The Modified Pierson-Moskowitz
Tank Conference (ITTC) [19] recommended Spectrum”, but the reason is not clear since
the following spectral formulation for open the Bretschneider Spectrum is devised
ocean wave conditions; before the Pierson-Moskowitz Spectrum
[22].
A  B  (2) 15th ITTC [23] made an amendment
=S (ω ) exp  − 4 
ω5  ω  to the above spectral formulation and

22
 Anıl et al./ JEMS, 2017; 5(1): 19-38

recommended the use of only Equation The JONSWAP Spectrum in terms of

(2) along with Equation (4) as a standard modal wave period is as follows [26];
spectral formulation for open ocean wave 
 2 
conditions. 15th ITTC recommended the exp − 1  ωTm −1
 


following spectral formulation which is a  2σ 2  2π






(9)
 
S J (ω ) = 0.658 S (ω ) 3.3  
form of the “Joint North Sea Wave Project
(JONSWAP) Spectrum [24]” for fetch-
where,
limited situations;
2
 0.206ωT
1 −1 

exp − (5)
2σ 
S J (ω ) = 0.658 S (ω ) 3.3

  (10)

where,
The mean values of the significant
wave height range can be used to define
(6) the sea-state as shown in Table 1, but the
corresponding modal wave periods for
each sea-state should be known to define
the sea area. 18th ITTC [27] gives the range
Here S ( ω ) is the “ITTC Two-Parameter and most probable modal wave periods of
Spectrum” which was defined by Equation North Atlantic, North Pacific, and Northern
(2) along with Equation (4). 17th ITTC [25] Hemisphere for each sea-state.
made a replacement for the above fetch- ITTC One-, ITTC Two-Parameter and
limited spectral formulation as a draft JONSWAP sea-spectra at “sea-state 6” for
recommendation, but it was not adopted the Northern Hemisphere are shown in
as a final recommendation by the 17th ITTC Figure 2. The dimension of the sea-spectra
seakeeping committee. is “squared length multiplied by time”, since
The significant wave height H 1 / 3 and the spectral ordinates are the measure
average period T1 are the main input of the of the average squared value of the wave
above recommended spectra. On the other amplitude of the corresponding frequency
hand, STANAG 4154 [16] recommended abscissa [10]. The area under the spectrum
to use the significant wave height and the curve is equal to the “variance” of the large
modal wave period Tm to characterize irregular wave time history which has a
the nature of the seaway. The modal wave zero mean [26];
period corresponds to the maximum wave
energy, i.e. the peak frequency (modal wave ∞

frequency) of the spectrum curve [16, 26]. m0 = ∫ S (ω ) dω (11)

The relation between the average and the
0
modal wave period is as follows [26]; This is why the sea-spectrum is also
Tm = 1.296 T1 (7) named as “variance spectrum” by Beck
et al. [11]. Since the irregular wave time
The constants (4) become as follows, history has been agreed to hold a zero
using Equation (7); mean, variance is equal to the “mean square
H1/2 3 1949 value” [28].
A 488
= , B (8) The “standard deviation” or “rms value”
Tm4 Tm4
is equal to the square root of variance [28].
The significant wave amplitude ζ 1/ 3 is

23
 © UCTEA The Chamber of Marine Engineers Journal of ETA Maritime Science

equal to the double rms amplitude. Hence defined as follows [9];

the following equation is valid for the
S (ω=
,ν ) M (ν − µ ) S (ω ) (14)
significant wave height [26];
where M (ν − µ ) is the spreading
2ζ 1/ 3 =
H1/ 3 = 2 × 2 m0 =
4 m0 (12) function and S ( ω ) is the unidirectional

spectrum. Here µ is the primary (dominant)
The peak frequency of the spectrum and ν is the secondary wave direction
curve equal to the modal wave frequency relative to the earth-fixed coordinate
ωm , as mentioned before and therefore the system. The range of the secondary wave
modal wave period is as follows; directions is defined as follows [26];
2π
Tm = (13) −ν max ≤ ν − µ ≤ ν max
ωm (15)

It is obvious that one can derive where ν max is the wave spreading angle.
the same values of the significant wave The spreading function distributes the total
height and modal wave period from a sea- energy of the unidirectional spectrum over
spectrum curve, which has been calculated the range of secondary wave directions
using above spectral formulas in terms of from −ν max to ν max [26]. Therefore, the
the significant wave height and modal wave variance of the irregular wave time history
period. (11) becomes as follows [10] (cited by
Perez [9]);
3.3. Directional Spectrum ∞ ∞ ν max + µ
The sea-spectra described above = m0 ∫= S (ω ) dω ∫ ∫ S (ω ,ν ) dν dω (16)
is unidirectional, i.e. can be used for a 0 0 −ν max + µ

description of a long-crested irregular This equation means that volume

sea. In order to describe a short-crested under the directional spectrum surface is
irregular sea it is required to have a multi- equal to the variance of the irregular wave
directional spectrum. Multi-directional time history with has a zero mean. The
or shortly “directional” spectrum can be volume under the directional spectrum
surface is also equal to the area under the

Figure 2. Sea-Spectra at “Sea-State 6” for the Northern Hemisphere

24
 Anıl et al./ JEMS, 2017; 5(1): 19-38

corresponding spectrum curve. Following In this study, spreading index m is 2, and

equation can be written from Equation the spreading angle ν max is π 2 , which is
(16); the most common practice to represent
ν max + µ ν max + µ
S (ω ) = (17) short-crested seas for ship design purposes.
(ω ,ν ) dν
∫ +µ S= ∫ M (ν − µ ) S (ω ) dν
−νmax −ν max + µ Therefore, Equation (14) becomes as
The above equation can be rewritten as; follows [26];
2
ν max
=S (ω ,ν ) cos 2 (ν − µ ) S (ω ) (25)
S (ω ) = ∫ M (ν − µ ) S (ω ) d (ν − µ ) (18) π
−ν max
The directional ITTC Two-Parameter sea-
The spreading function is defined as follows spectrum derived from Equation (25) at
[26]; “sea-state 6” for the Northern Hemisphere
 π  is shown in Figure 3.
M (ν − µ ) D cos m 
= (ν − µ )  (19)
 2ν max 
where D is a constant and m is the
spreading index (a positive integer). If we
set;
π π
ν=′ (ν − µ ) ⇒ dν=′ d (ν − µ ) (20)
2ν max 2ν max

Equation (18) becomes [26];

π
ν max
2ν max
2ν max
S (ω ) = ∫ D cos m (ν ′ ) S (ω ) dν ′ (21)
π π
2ν max
( −ν max )

The constant D can be written from the Figure 3. Directional ITTC Two-Parameter
above equation as follows [26]; Spectrum at “sea-state 6” for the Northern
π 1 π 1 Hemisphere
D =
2ν max π 2 4ν max π 2
(22) 3.4. Generation of the Time Series
∫
m
cos (ν ′ ) dν ′ ∫0 cos (ν ′)
m
dν ′
−π 2 So far it has been shown that the sea-
Integration of the above equation for spectrum formulas can be used to define
different values of m yields; the random characteristics of an irregular
 1 sea. A similar spectral formulation can
 for m = 0
2ν max be used to investigate ship responses to

 π irregular waves since both are random
 for m = 1 processes. Under the assumption of
 4ν max
D= (23) linearity for the ship motions, St. Denis and
 1 ⋅ 3 ⋅ 5 m π
for m odd > 1 Pierson [10] used the linear superposition
 2 ⋅ 4 ⋅ 6 ( m − 1) 4ν max
 theory and the spectral analysis techniques
 2 ⋅ 4 ⋅ 6 m 1
for m even > 0 developed in other disciplines and
1 ⋅ 3 ⋅ 5 ( m − 1) 2ν max
introduced an innovative method which
Following equation is also valid for even associates the spectral density of ship
values of m other than zero, where responses to the input sea-spectrum
n = m 2 [9]; to determine the statistics of the ship
( 2n−1) n !( n − 1)! motions in irregular seas. In this method,
2 π π
=M (ν − µ )
π ( 2n − 1)!
cos 2n (ν − µ ) for −
2
≤ν − µ ≤
2
(24) the principle of superposition allows the

25
 © UCTEA The Chamber of Marine Engineers Journal of ETA Maritime Science

ship responses in irregular waves to be the approaches zero [11];

sum of the responses to regular waves of 1 2
all frequencies of the input sea-spectrum. ζ n = S (ωn ) δω (27)
Therefore, responses can be represented by 2
the spectral density functions (motion or Therefore, each component amplitude can
response spectra) similar to the sea-spectra be found as follows if δω is small enough
and the area under these motion spectra [11];
gives the variance of the corresponding ζ n = 2 S (ωn ) δω (28)
response amplitude [11, 12, 29]. Price
The time series data for a short-crested
and Bishop [28] described further irregular wave surface can be generated
developments and details about St. Denis with the following algorithm as indicated
and Pierson’s [10] approach [11, 30]. by Perez [9];
Following these studies, generation of N M
ζ ( x0 , y0 , t )
= ∑∑ ζ cos (ωnt + ε n ,m − kn ( x0 cosν m − y0 sinν m ) ) (29)
the time series of numerical simulations can n 1=
= m 1
n ,m

be performed. Based on this method, from Here ν m is the each secondary wave
any irregular wave time history, resultant direction relative to the earth-fixed
sea-spectrum can be derived. The inverse coordinate system;
of this procedure is also available, i.e.
µ −ν max ≤ ν m ≤ µ +ν max (30)
superposition of a large number of sine or
cosine wave components assumed to have where, if ν m < 0 , then ν=
m
ν m + 2π .
a definite frequency and a random phase- For a directional spectrum (14), component
shift angle produces an irregular wave time amplitude can be found as follows [11];
history [26, 11]. The same back and forth
process is also valid in the ship responses. ζ n ,m = 2 S (ωn ,ν m ) δωδν (31)
The time series data for a long-crested
irregular wave surface defined in the earth- Long- and short-crested irregular wave
fixed coordinate system can be generated surfaces derived from above formulations
with the following algorithm as indicated at a time are shown in Figure 4.
by Perez [9]; The time series data for the ship
N
responses is also calculated based on
ζ= ∑ ζ cos (ω t + ε − k ( x cos µ − y
( x0 , y0 , t ) n n n n 0 0 sin µ ) ) (26) the St. Denis and Pierson [10]. The input
n =1
wave spectrum and the RAOs for the ship
Here ζ n , ε n , k n are the wave amplitude, are required for this purpose [29]. St.
random phase-shift angle, and the wave Denis and Pierson [10] defined the RAO
number of each component respectively, as “the amplitude of the response to the
and µ is the wave direction relative to the wave amplitude of unity”. In other words,
earth-fixed coordinate system. Following RAO is the non-dimensional motion
equation is valid for the amplitudes of amplitude. Linear motion (surge, sway,
the wave components in the limit as δω heave) amplitudes are divided by the wave

Figure 4. Long-Crested (Left), Short-Crested (Right) Irregular Wave Surface

26
 Anıl et al./ JEMS, 2017; 5(1): 19-38

2
amplitude ζ , angular motion (roll, pitch, Sηi (ωe ) = H i (ωe ) S (ωe ) for i = 1, 2, , 6 (35)
yaw) amplitudes are divided by the wave
slope amplitude kζ . The RAO is also named Here i is the motion index for surge, sway,
as “transfer function” in other disciplines. It heave, roll, pitch yaw respectively. Similar
is the ratio of wave amplitude or wave slope to the sea-spectra, the area under these
amplitude “transferred” to the response motion spectra gives the corresponding
amplitude by the ship which is supposed to motion amplitude variance [26];
be a “system” [26]. Essentially, the modulus ∞

(absolute value) of the complex transfer mi 0 = ∫ Sηi (ωe ) dωe (36)

function (non-dimensional complex motion 0
amplitude) is defined as the RAO [11]. Alternative formulation for the motion
Consequently, the argument of the complex spectra in wave frequency domain is as
transfer function is the phase-shift angle follows [26];
of the response. However, sometimes the 2
Sηi (ω ) = H i (ω ,VS , χ ) S (ω ) for i = 1, 2, , 6 (37)
term RAO is used for the complex transfer
function itself. Note that, the RAO in Equation (37) is equal
On a moving ship, the responses are to the RAO in Equation (35), which is just
experienced at the encounter frequency, for the demonstration of the domain;
not the wave frequency. In deep water H i (ω ,VS , χ ) = H i (ωe ) (38)
following equation is valid for the encounter
frequency [26]; The ordinates of motion spectrum in the
ω 2VS wave frequency domain (37) have no
ωe= ω − cos χ (32)
g physical meaning, but the areas under both

Here, g is the acceleration of gravity, VS motion spectra, i.e. the motion amplitude
is the ship speed and χ is the “heading variances are equal [26];
angle” or “encounter angle” which is the ∞ ∞

ship course relative to the primary wave mi 0 = ∫ Sηi (ω ) dω = ∫ Sηi (ωe ) dωe (39)
0 0
direction;
Similar to (27), the following equation can
=χ course − µ (33) be written for the motion amplitude in the
where, if χ < 0 , then χ= χ + 2π . limit as δω approaches zero;
Since 9th ITTC [31] it has been agreed that 1
ηi ,n = Sηi (ωn ) δω
2
(40)
heading angle of 0° represents following 2
seas and 180° represents the head seas. The Therefore, motion amplitudes for each
sea-spectrum should be also transformed component can be found as follows if δω
into the “encounter spectrum”, which has is small enough;
the same area under the spectrum curve,
to cover the range of encounter frequencies ηi ,n = 2 Sηi (ωn ) δω (41)

on board. The encounter spectrum can be
calculated as follows [26]; In short-crested seas, component amplitude
can be found as follows, similar to Equation
g
S (ωe ) = S (ω ) (34) (31);
g − 2ω VS cos χ
The product of the encounter spectrum ηi ,n,m = 2 Sηi (ωn ,ν m ) δωδν (42)

with the square of the RAO of the desired where;
2
response gives the corresponding motion Sηi (ωn ,ν m ) = H i (ωn ,VS , χ m ) S (ωn ,ν m ) (43)
(response) spectrum [29];

27
 © UCTEA The Chamber of Marine Engineers Journal of ETA Maritime Science

ship can be generated as follows;

Here χ m is the heading angle, i.e. the ship
course relative to the each secondary wave η AV ( t ) = η3 ( t ) − x sin (η5 ( t ) ) (48)

direction ν m ; where x is the distance from the center
χ m course −ν m
= (44) of the body-fixed coordinate system. Note
that sine of a small angle is equal to the
where, if χ m < 0 , then χ=
m
χ m + 2π . angle itself in radians. Therefore, the above
From (31) and (43), Equation (42) can be equation becomes as follows;
rewritten as follows;
η AV ( t ) = η3 ( t ) − xη5 ( t ) (49)
ηi ,n,m = ζ n,m H i (ωn ,VS , χ m ) (45)
Likewise, the vertical component of the
The time series data for the seakeeping angular pitch motion RAO at any point of
simulation (ship responses) in short- the ship becomes as follows;
crested seas can be generated with the
H 5V (ωe ) = x k H 5 (ωe ) (50)
similar algorithm as indicated by Perez
[9], using pre-calculated frequency domain where k is the wave number.
RAOs for the selected sea-state and sea- Superposition of the amplitudes of two
spectrum; simple harmonic motions, acting in the
N M “same” direction is given by the following
=ηi ( t ) ∑∑ ηi,n,m cos (ωe n,m t + σ i ,n,m + ε n,m ) for i =
1, 2,3
n 1=
= m 1
(46)
equation [33];
N M
ηi ,n,m kn cos (ωe n,m t + σ i ,n,m + ε n,m )
=ηi ( t ) ∑∑ for i =
4,5, 6
A= A12 + A22 + 2 A1 A2 cos (θ1 − θ 2 ) (51)
n 1=
= m 1

Here σ i , n , m is the component phase-shift
where A1 , θ1 and A2 , θ 2 are the amplitudes
angle of the response. Encounter frequency
and phases respectively. The resultant
for each heading angle (44) is as follows
phase is given by the following equation
ωn 2VS (47) [33];
ωen,=
m ωn − cos χ m
g  A sin θ1 + A2 sin θ 2 
The time series data for the seakeeping θ = arctan  1  (52)
simulation (ship responses) in long-crested  A1 cos θ1 + A2 cos θ 2 
seas can be also generated by the above Since RAO is the non-dimensional motion
formula with M = 1 and dν = 1 [9]. amplitude, the superposition of heave and
vertical component of pitch RAOs acting
3.5. Derived Responses in the “opposite” direction becomes as
The time series of the derived follows;
responses like absolute or relative motions, H AV (ωe ) = H 3 + H 5V − 2 H 3 H 5V cos (σ 3 − σ 5 ) (53)
2 2

accelerations; vertical distortion, shear

The above equation is for the RAO of the
force and bending moment distribution of
absolute vertical motion at any point of the
ship sections can be obtained by the “real
ship in long-crested irregular head seas.
time computer experiments”. The excessive
The resultant phase becomes;
motions, deck wetness, slamming or
emergence at any location can be observed  H 3 sin σ 3 − H 5V sin σ 5 
σ AV = arctan   (54)
and analyzed from the relative motion time  H 3 cos σ 3 − H 5V cos σ 5 
 
history.
The absolute longitudinal, lateral
As an example, the absolute vertical
and vertical motion formulations for six
motion involves heave and pitch motions
degrees of freedom are given by Lloyd [26].
only for the ships in long-crested irregular
For example, the RAO and the phase of the
head seas. The time series data for the
absolute vertical motion which involves
absolute vertical motion at any point of the

28
 Anıl et al./ JEMS, 2017; 5(1): 19-38

heave, pitch and roll motions are given by the absolute and relative vertical motion
following equations; time series data (49)&(62) once and twice
respectively.
H AV (ω=
e) P112 + P122 (55)
The expressions are slightly different
 P12  for the hydroelasticity theory [32], as
σ AV = arctan   (56) investigated by Belik [1]. For the symmetric
 P11  response of ships in long-crested irregular
where; head seas, the RAO for the vertical
P11 =H 3 cos σ 3 + H 4V cos σ 4 − H 5V cos σ 5 distortions of a ship section is expressed as
(57)
P12 =H 3 sin σ 3 + H 4V sin σ 4 − H 5V sin σ 5 follows [1, 3];
If there is no roll motion, the above six n

degrees of freedom formulation yields

w( x) = ∑ p w ( x)
r r
(63)
to the Equation (53) and (54) using
r =0

where x is the distance from the aft
appropriate trigonometric identities. The perpendicular of the ship, pr represents
corresponding absolute motion spectrum the complex quantities called “principle
can be found similar to Equation (35); coordinates” and wr ( x ) represents the
2
S AV (ωe ) = H AV (ωe ) S (ωe ) (58) “principle mode shapes” of the symmetric
vibration. The above superposition
The absolute vertical velocity and algorithm needs care similar to Equation
acceleration RAOs can be written as follows; (51) and (53), since the polar form of
complex quantities has amplitudes and
H ′AV (ωe ) = ωe H AV (ωe ) phases.
(59)
H ′′AV (ωe ) = ωe2 H AV (ωe ) The principle coordinates consist of
“rigid body” modes ( r = 0,1 ) and the
The relative motion RAO and phase distortion modes ( r = 2, 3,  , n ) [32]. The
formulation is also given by Lloyd [26]. For bodily principal coordinates for the rigid
instance, the RAO of the relative vertical modes are the heave and pitch motions of
motion is as follows; the ship respectively;
H RV (ωe ) =
H AV − 2 H AV ζ cos (σ AV + Q ) + ζ 2
2
(60) p0 = H 3 (ωe )
(64)
LBP
where Q = k x cos µ . For the ships in head
seas Q = − k x . The phase of the relative
p1 =
2
(
sin k H 5 (ωe ) )
vertical motion is given by the following where LBP is the ship’s between
equation. perpendiculars and k is the wave number.
The principle mode shapes for the rigid
 H AV sin σ AV + ζ sin Q 
σ RV = arctan   (61) modes ( r = 0,1 ) are shown in References
 H AV cos σ AV + ζ cos Q  [1, 2, 3, 32]. Since mode shapes are a
 
The time series data for the relative function of x LBP , for the rigid body modes
vertical motion at any point of the ship can only, Equation (63) becomes the absolute
be generated as follows; vertical motion similar to Equation (53).
Similar to Equation (46), the time series
η RV ( t ) = η3 ( t ) − xη5 ( t ) − ζ ( x , t ) (62)
data for the principle coordinates can be
The time series data for the absolute generated by the following algorithm [2];
and relative vertical motion velocity and m n

acceleration at any point (bow, midships (

w ( x, t ) = ∑ ∑ ζ j pr , j wr ( x ) cos ωe j t + arg ( pr , j ) + ε j ) (65)
j =1 r = 0
or stern) can be derived by differentiating

29
 © UCTEA The Chamber of Marine Engineers Journal of ETA Maritime Science

4. Validation The corresponding circular frequency in

th
Validation of the time series of wave radians per second for the k coefficient is
elevation and all responses, including the as follows [1, 34];
derived ones can be performed using the 2π k
spectral analysis technique which includes ωk = (72)
TL
discrete Fourier transform (DFT), and
smoothing algorithms, as explained by The DFT is valid for frequency up to the
Belik [1] and Newland [34]. Nyquist frequency. From Equation (66) and
For the spectral analysis, the time step (69) following relation can be written to
of the simulation should be selected as find the maximum value of the k [34];
follows [1, 34]; 2π k π
= ω* = max
(73)
∆t =
π (66) ( N + L ) ∆t ∆t
ω * As shown in the above equation, the
Here ω * is the “Nyquist frequency” in maximum value of the k for the validation
radians per second, which is the highest should be as follows;
wave or encounter frequency used for the N+L
kmax = (74)
generation of the time series. The time 2
length of the simulation with N discrete Therefore, the DFT algorithm for the
data points becomes as follows [1]; validation is defined as follows [1, 34];
TN= N ∆t (67) 1 N + L −1  2π nk  N+L
(75)
N+L ∑
= Xk xn exp  −i  for k = 0,1, , 2

The number of discrete data points n  N +L

should be a power of two as a condition of th

The k real Fourier coefficient is given
the fast Fourier transform (FFT) algorithm. as follows [1, 34];
Therefore the time series data must be * 2
extended by adding L additional zeros as = Sk X= k Xk Xk (76)
amplitude [1, 34]; *
where X k is the complex conjugate of
N+L= 2p the X k . The ordinates of the continuous
(68)
spectral estimate are given as follows [1,
where p is
a positive integer. The extended
34];
time length of the discrete data points with
the additional zeros becomes as follows; N + L TL
S ′ ( ωk ) = S (77)
N 2π k
TL = ( N + L ) ∆t (69) Here ( N + L ) N is the correction factor

The DFT algorithm for the N + L data for the additional zeros. The smoothing
points is defined as follows [1, 34]; of the spectral estimate ordinates “to
increase their statistical reliability” can be
1 N + L −1  2π nk 
N+L ∑
=Xk xn exp  −i
 N +L
 for
= k 0,1, , N + L − 1 (70) performed with the following algorithm [1,
n

Here X k is the
th
complex Fourier 34];
coefficient, xn is the n discrete data point
th
1 s

of the time series, including the additional

S (ωk ) =
2s + 1 q∑
=− s
(
S ′ ωk + q ) for k = 0,1, ,
N+L
2
(78)
zeros, and i is the imaginary unit of a where 2 s + 1 is the span of the smoothing
complex number. The above algorithm (70) process;
can also be performed using the built in
MATLAB commands “nextpow2” and “fft”; N+L
2s + 1
= = T β TL β e (79)
N N e
X k = (1 2 ^ nextpow2 ( N ) ) *fft ( xn ) (71)

30
 Anıl et al./ JEMS, 2017; 5(1): 19-38

Here β e is the “effective bandwidth”, i.e. the accuracy of the derived spectral density
frequency interval for the smoothing function as Newland [34] stated. For the
2s + 1 further validation studies the length of each
βe = (80) time series is 1800 seconds and α = 0.45.
TL
For the short-crested wave data, the spectral
The dimension of the effective bandwidth
validation can be performed separately for
is Hertz (Hz), since the frequency step for
each secondary wave direction component.
the spectral estimates in Equation (78)
For example, the component series for
is 2π TL in radians per second, as shown
0 and ν − µ = 75 with δν = 15 and
 
ν −µ =

by Equation (72). Belik [1] expressed the
the corresponding spectral validations are
effective bandwidth as the following;
shown in Figure 7 and Figure 8 respectively.
2π (81)
βe = 2 ( 0 < α < 1) Heave and pitch motion time history (in
α TN
long-crested seas) of a DTMB Model 5415
Here as α tends to zero the precision representing the preliminary design of the
of the spectral estimate in the effective DDG-51 hull form [35] (in forward motion
bandwidth increases. The above algorithm with the Froude number of 0.41), derived
(78) can also be performed using the built from Equation (46), shown in Figure 9
in the MATLAB command “smooth with and Figure 10 was validated like the wave
default “moving” method”; time history. The maximum encounter
S (ω ) smooth ( S ′ (ω ) , 2 s + 1, 'moving') (82) frequency, i.e. the Nyquist frequency is
In conclusion, according to Price 2.50 radians per second. The motion time
and Bishop [28] (cited by Belik [1]) the series in short-crested seas can also be
“physically realizable one sided spectral validated separately for each secondary
density function” is as follows; wave direction component.
The absolute vertical motion time
φ (ω ) = 2S (ω ) (83)
histories at bow and stern of a frigate
For the validation, the spectral density generated by Equation (49) and the
functions derived from time series data spectral validations are shown in Figure 11
with above process were compared to the and Figure 12.
original pre-calculated spectra which were
used for the generation of the time series at 5. Summary and Conclusions
350 frequency abscissae. The time series In this study, the simulation-based
were generated using ITTC Two-Parameter analysis method (a.k.a. the real time
Spectrum at “sea-state 6” for the Northern computer experiment method) of ship
Hemisphere. motions and wave elevation was discussed
The long-crested irregular wave with the procedure of spectral validation.
elevation time histories generated by For this purpose, visually simulated time
Equation (26) at a fixed point and the series of ship motions and wave elevation
corresponding spectral validations are were generated for the demonstration and
shown in Figure 5 and Figure 6. The analysis of the pre-calculated seakeeping
maximum wave frequency, i.e. the Nyquist results. Validation of the time series of
frequency is 1.60 radians per second. wave elevation and all responses, including
The length of the first time series is 5400 the derived ones were performed using
seconds and α = 0.26. The length of the the spectral analysis technique which
second time series is 1800 seconds and includes discrete Fourier transform (DFT),
α = 0.45. The length of the time series and smoothing algorithms. It is found
and the effective bandwidth changes the that this analysis method can be adapted

31
 © UCTEA The Chamber of Marine Engineers Journal of ETA Maritime Science

Figure 5. Long-Crested Irregular Wave Time History (5400 seconds) and Spectral Validation (RMS
Amplitude = 1.25 m, Significant Amplitude = 2.5 m)

Figure 6. Long-Crested Irregular Wave Time History (1800 seconds) and Spectral Validation (RMS
Amplitude = 1.25 m, Significant Amplitude = 2.5 m)

32
 Anıl et al./ JEMS, 2017; 5(1): 19-38

Figure 7. Short Crested Irregular Wave Time History Component (for ν − µ =

0 and δν = 15 ) and


Spectral Validation

Figure 8. Short Crested Irregular Wave Time History Component (for ν − µ =

75 and δν = 15 ) and
 

Spectral Validation

33
 © UCTEA The Chamber of Marine Engineers Journal of ETA Maritime Science

Figure 9. The Heave Motion Time History (In Long-Crested Seas) and Spectral Validation (RMS
Amplitude = 1.12 m, Significant Amplitude = 2.24 m)

Figure 10. The Pitch Motion Time History (In Long-Crested Seas) and Spectral Validation (RMS
Amplitude = 0.0266 radians, Significant Amplitude = 0.0532 radians)

34
 Anıl et al./ JEMS, 2017; 5(1): 19-38

Figure 11. The Absolute Vertical Motion Time History at Bow (In Long-Crested Seas) and Spectral
Validation (RMS Amplitude = 2.63 m, Significant Amplitude = 5.26 m)

Figure 12. The Absolute Vertical Motion Time History at Stern (In Long-Crested Seas) and Spectral
Validation (RMS Amplitude = 1.50 m, Significant Amplitude = 3,00 m)

35
 © UCTEA The Chamber of Marine Engineers Journal of ETA Maritime Science

to design and validation studies for both [3] Başaran, İ. (2005). Responses of
the naval platform itself and the systems Ships in Extreme Sea Conditions,
on board effectively. The analysis of the Thesis (M.Sc.), İstanbul Technical
simulation results provides sufficient University, Institute of Science and
information about the extreme motions Technology, İstanbul. Retrieved: 12
and accelerations. Observation of the March 2016, from: http://hdl.handle.
deck wetness, slamming or emergence net/11527/11939.
at any location is also feasible from the [4] Yeo, D.J., Cha, M. and Mun, D.
relative motion time history. Utilizing (2012). Simulating ship and buoy
the ship hydroelasticity theory [32], the motions arising from ocean waves
vertical distortion, shear force and bending in a ship handling simulator.
moment distribution of ship sections can be Simulation: Transactions of the
acquired by these computer experiments. Society for Modeling and Simulation
Furthermore, the results of this study International, 2012:88(12):1407-
may also be suitable for the ship-handling 1419. Retrieved: 07 June 2013,
simulators or helicopter landing on ship from: http://sim.sagepub.com/
simulations for more realistic visualization. content/88/12/1407.full.pdf+html.
As indicated by Belik and Gökmen [2] [5] Menteş, A., Yetkin, M., Türkoğlu N.,
the importance of the simulation-based Yalçın, İ. , Akyıldız, H. and Helvacıoğlu,
analysis is its practical and innovative İ.H. (2013). Açık Deniz Petrol
nature. The desired length of time series Platformu Modellemesinde Yapay
data for any ship response, including the Sinir Ağları Yaklaşımı. Journal of ETA
derived ones can be generated without Maritime Science, 2013:1(2):1-8.
any cost and restrictions except the [6] He, B., Wang, W., Zhang, Y. and Liu,
computer performance. Comparison of the L. (2016) A Simulation Model of
performance of different hull forms in the a Helicopter Landing on a Ship.
same irregular sea condition is possible International Journal of Simulation
with this simulation technique. If supported Systems, Science & Technology,
by experimental results, a database of the 2016:17(25):5.1-5.7.
responses from the different type of hull [7] OGRE3D (2013). Open Source 3D
forms can also be generated. Moreover, the Graphics Engine. Retrieved: 04
real time simulation-based analysis method February 2013, from: http://www.
may prevent the over-design issues and ogre3d.org/.
allow the refinement of the design criteria [8] Adanır, İ.R., Schoenwaelder, B. and
for both the naval platform itself and the Schumann, L. (2002). Ship Motion
systems on board. Simulation., Project. The University of
Michigan Virtual Reality Laboratory
References (VRL) at the College of Engineering,
[1] Belik, M.Ö. (1982). Symmetric Ann Arbor. Retrieved: 05 May 2012,
Response of Ships in Regular and from: http://www-vrl.umich.edu/
Irregular Waves, Doctoral thesis, sel_prj/wave_ship/.
University College London. [9] Perez, T. (2005). Ship Motion Control,
[2] Belik, M.Ö. and Gökmen, M. (1995). Course Keeping and Roll Stabilisation
Gemilerin Aşırı Deniz Şartlarında Using Rudder and Fins. London:
Davranışının Bilgisayarla Simülasyonu. Springer-Verlag.
Deniz Kuvvetleri Komutanlığı, 2inci [10] St. Denis, M. and Pierson, W.J. (1953).
Teknik Sempozyumu, İstanbul, 1995. On the Motions of Ships in Confused

36
 Anıl et al./ JEMS, 2017; 5(1): 19-38

Seas. Transactions, Society of Naval [18] Longuet-Higgins, M.S. (1952). On the

Architects and Marine Engineers, Statistical Distribution of the Heights
1953:61:280-357. of Sea Waves. Journal of Marine
[11] Beck, R.F., Cummins, W.E., Dalzell, J.F., Research, 1952:11(3):245-266.
Mandel, P. and Webster, W.C. (1989). [19] 12th International Towing Tank
Motion in Waves. In E.V. Lewis, (Ed.), Conference (1969). Technical
Principles of Naval Architecture Decisions and Recommendations.
(2nd revision, Vol. 3) (pp. 1-190). Rome, Italy.
Jersey City, NJ: The Society of Naval [20] Pierson, W.J. and Moskowitz, L.
Architects and Marine Engineers. (1964). A Proposed Spectral Form
[12] Salvesen, N., Tuck, E.O. and Faltinsen, for Fully Developed Wind Seas Based
O. (1970). Ship Motions and Sea on the Similarity Theory of S. A.
Loads. Transactions, Society of Naval Kitaigorodski. Journal of Geophysical
Architects and Marine Engineers, Research, 1964:69(24):5181-5190.
1970:78:250-279. [21] Bretschneider, C.L. (Ed.). (1959). Wave
[13] González, X.V. (2008). Hydrax. Variability and Wave Spectra for Wind
Version 0.5.1. Ogre3D add-on library. Generated Gravity Waves (Technical
Retrieved: 04 February 2013, from: Memorandum No.118). Beach Erosion
http://www.ogre3d.org/tikiwiki/ Board, US Army Corps of Engineers.
Hydrax. [22] Michel, W.H. (1999). Sea Spectra
[14] Johanson, C. (2004). Real-time Revisited. Marine Technology,
water rendering, Introducing the 1999:36(4):211-227.
projected grid concept, Thesis (M.Sc.), [23] 15th International Towing Tank
Department of Computer Science, Conference (1978). Recommendations
Lund University, Lund Sweden. of the Executive Committee and the
Retrieved: 29 May 2013, from: http:// Technical Committee and Panels.
fileadmin.cs.lth.se/graphics/theses/ Hague, Netherlands.
projects/projgrid/ projgrid-lq.pdf. [24] Hasselmann, K., Barnett, T.P., Bouws,
[15] WMO Manual on Codes (Publication E., Carlson, H., Cartwright, D.E.,
No.306) (2015). International Enke, K., Ewing, J.A., Gienapp, H.,
Codes, Volume I.1, Annex II to the Hasselmann, D.E., Kruseman, P.,
WMO Technical Regulations, Part Meerburg, A., Müller, P., Olbers, D.J.,
A - Alphanumeric Codes, 2011 ed., Richter, K., Sell, W., Walden, H. (1973).
Updated in 2015. Retrieved: 02 June Measurements of Wind-Wave Growth
2016, from: http://www.wmo.int/ and Swell Decay During the Joint
p a g e s / p ro g / w w w / W M O C o d e s / North Sea Wave Project (JONSWAP).
WMO306_vI1/VolumeI.1.html. Ergänzungsheft zur Deutschen
[16] STANAG 4154 (2000). Common Hydrographischen Zeitschrift, Reihe
Procedures for Seakeeping in the A(80), Nr. 12. Retrieved: 05 June
Ship Design Process (3rd Ed.), NATO 2016, from: TU Delft Institutional
Military Agency for Standardisation. Repository, http://repository.tudelft.
[17] Sverdrup, H. U. and Munk, W. H. nl.
(1947). Wind, Sea and Swell: Theory [25] 17th International Towing Tank
of Relations for Forecasting (H.O.Pub. Conference (1984). Report
No. 601). U.S.Navy Department of Seakeeping Committee,
Hydrographic Office. Retrieved: 13 Recommendations Concerning the
June 2016, from: http://archive.org. Work of the Information Group and

37
 © UCTEA The Chamber of Marine Engineers Journal of ETA Maritime Science

the Technical Committees Adopted by

the 17th ITTC. Göteborg, Sweden.
[26] Lloyd, A.R.J.M. (1989). Seakeeping:
Ship Behaviour In Rough Weather.
Chichester, West Sussex: Ellis
Horwood Limited.
[27] 18th International Towing Tank
Conference (1987). Report of
Seakeeping Committee. Kobe, Japan.
[28] Price, W.G. and Bishop, R.E.D. (1974).
Probabilistic Theory of Ship Dynamics.
London: Chapman and Hall.
[29] Beck, R.F. and Reed, A.M. (2001).
Modern Computational Methods for
Ships in a Seaway. Updated version of
a paper prepared as an invited lecture
for the Twenty-third Symposium on
Naval Hydrodynamic.
[30] Newman, J.N. (1978). Theory of
Ship Motions. Advances in Applied
Mechanics, 1978:18:221-283.
[31] 9th International Towing Tank
Conference (1960). Seagoing Qualities
of Ships. Paris, France.
[32] Bishop, R.E.D. and Price, W.G. (1979).
Hydroelasticity of Ships. Cambridge:
Cambridge University Press.
[33] Flexible Learning Approach to
Physics (1994) Physics Module P5.1,
Simple Harmonic Motion, The Open
University. Retrieved: 22 May 2016,
from: http://www.physics.ox.ac.uk/
teach/FLAP/FLAP/P5.1.pdf.
[34] Newland, D.E. (1993). An Introduction
to Random Vibrations, Spectral and
Wavelet Analysis (3rd ed.). New York:
John Wiley & Sons.
[35] Ratcliffe, T.J., Mutnick, I. and Rice,
J. (2001). Stern Wave Topography
and Longitudinal Wave Cuts
Obtained on Model 5415, With and
Without Propulsion (NSWCCD-
50-TR-2000/028). West Bethesda,
Maryland: David Taylor Model
Basin Naval Surface Warfare Center,
Carderock Division.

38

View publication stats