Ship Dynamics

Noverdo Saputra, S.T., M.Eng

[email protected]

2017
Contents

1 Introduction 1

1.1 What is Ship Dynamics? . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.2 Why is the Ship Dynamics of a Vessel Important? . . . . . . . . . . . . 1

1.3 Aim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

1.4 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

1.5 Recommended Reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2 Ship Motions in Six Degrees of Freedom 4

2.1 Axis System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

2.2 Coupling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2.3 Motion Resonance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

3 Basic Statistics 7

4 Description of Regular and Irregular Waves 9

4.1 Regular Waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

4.2 Dispersion Relationship . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

4.3 Sinusoidal Wave Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

4.4 Wave Encounter Frequency . . . . . . . . . . . . . . . . . . . . . . . . . . 12

4.5 Characteristics of an Irregular Sea . . . . . . . . . . . . . . . . . . . . . . 16

4.6 Statistical Analysis of an Irregular Sea . . . . . . . . . . . . . . . . . . . . 18

4.6.1 Probability Density Function . . . . . . . . . . . . . . . . . . . . . 18

4.6.2 Probability Distribution Function . . . . . . . . . . . . . . . . . . . 19

i
CONTENTS

4.6.3 Significant Wave Height and Wave Amplitude . . . . . . . . . . . . 20

5 Wave Spectra 22

5.1 Spectral Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

5.2 Standard Wave Spectra . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

5.2.1 ITTC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

5.2.2 JONSWAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

5.2.3 Pierson-Moskowitz . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

5.3 Wave Encounter Spectrum . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

6 Ship Response 30

6.1 Ship Response as a Linear System . . . . . . . . . . . . . . . . . . . . . . 30

6.2 Equations of Motion for a Ship in a Seaway . . . . . . . . . . . . . . . . . 30

6.3 Response Amplitude Operators . . . . . . . . . . . . . . . . . . . . . . . . 31

6.4 Response Spectra and RMS . . . . . . . . . . . . . . . . . . . . . . . . . . 34

7 Summary 37

ii
Chapter 1

Introduction

1.1 What is Ship Dynamics?

So far you will only have considered ships in a static case. The study of the bahaviour
of ships in a seaway is split, at present, into two distinct areas. The first is called
seakeeping, where the ship motion is excited by some known wave force. The second
is manoeuvring, which is carried out assuming that the sea is calm. Generally these
two areas are treated separately. The key to the study of both of these topics is how the
environment the ship is in (i.e. the sea) is modelled.

Seakeeping can be defined as (thefreedictionary.com):

1. The ability of a vessel to navigate safely at sea for prolonged periods during stormy
weather
2. The study of the response of a ship or floating object to waves

Both of these aspects are important, but the former cannot be assessed without being
able to do the latter and it is the latter which these notes address.

The behaviour of a ship in a seaway is generally modelled via frequency analysis. The
behaviour of a ship is initially assessed in a set of regular waves. An irregular ocean
seaway can then be built up by the summation or superposition of a set of regular waves;
similarly the response of the ship in each regular wave can be added together to give
the total response in the irregular sea. The superposition of regular waves to give a new
wave form is called Fourier Analysis.

1.2 Why is the Seakeeping Ability of a Vessel Important?

The seakeeping ability of a vessel needs to be assessed, not only to demonstrate how
it will behave under the normal operating conditions for which it is designed, but also
to model the effect of phenomena such as breaking waves, water on deck, slamming,
1
Seakeeping Lecture Notes 2017 - Chapter 1: Introduction

extreme accelerations and abnormally high wave loads (bending moments and shear
force).

Figure 1.1: Why is the Ship Dynamics of a vessel important?

1.3 Aim

To provide a fundamental understanding of the methods associated with modelling the

ocean environment and some basic ideas associated with the dynamics of a ship moving
in a seaway.

1.4 Objectives
1. To understand the different motions a ship can experience in a seaway
2. To understand the basis of a random seaway (a regular wave) and how it is defined
with respect to a moving vessel
3. To understand how a seaway is modelled using statistical methods (i.e. wave
spectra)
4. To understand how to model the dynamic behaviour of a rigid vessel in waves

2
Seakeeping Lecture Notes 2017 - Chapter 1: Introduction

1.5 Recommended Reading

A. R. J. M. Lloyd (1998) Seakeeping: Ship Behaviour in Rough Weather

E. V. Lewis (Ed.) (1989) Principles of naval architecture vol. 3: Motions in waves

and controllability, SNAME

3
Chapter 2

Ship Motions in Six Degrees of

Freedom

2.1 Axis System

The axis system of a ship is defined about its centre of gravity. Figure 2.1 demonstrates
this. The x axis is the longitudinal axis of symmetry with positive towards the bow.
The y axis acts in a positive direction to port. The z axis acts in a positive direction
vertically.

Figure 2.1: Ship body axis

A ship is a complex system which is subject to six modes of motion (degrees of freedom)
about the centre of gravity (Figure 2.2). These consider that a ship is a rigid, rather
than flexible, body and are often referred to as rigid body modes. The six degrees of
freedom are (always numbered in the same order):
4
Seakeeping Lecture Notes 2017 - Chapter 2: Ship Motions in Six Degrees of Freedom

1. Surge

2. Sway

3. Heave

4. Roll

5. Pitch

6. Yaw

Motions 1 - 3 are considered linear (translational) motions whilst motions 4 - 6 are

considered rotational motions.

Figure 2.2: Six degree of freedom motions [REF: IACS Common Structural Rules for
Bulk Carriers and Oil Tankers 2014]

2.2 Coupling

Generally a ship is vertically symmetrical. In this case the 6 modes can be de-coupled
into two sets, vertical and lateral. Vertical modes are Heave, Pitch and Surge whilst
lateral models are Roll, Sway and Yaw. How the modes couple is dependent on the
second moments of inertia of the vessel: Ixx , Iyy and Izz which are all large magnitude
and positive, as well as the non-zero Ixy and Iyx . The remaining second moments of
inertia (Ixz , Izx , Iyz and Izy ) can be approximated to zero because of port-starboard
symmetry and are therefore ignored.

Heave-Pitch Coupling Heave and pitch couple because the ship is not fore-aft sym-
metrical. Therefore as the ship heaves, the force at the aft end of the vessel is different
to the force at the forward end of the vessel. This results in a net rotational moment
about the y axis, otherwise known as pitch. Surge can be decoupled from heave and
pitch because it is a weak motion and for the purposes of the linear system being
considered we are dealing with small amplitude waves only, making the amplitude of
surge motion neglible.
5
Seakeeping Lecture Notes 2017 - Chapter 2: Ship Motions in Six Degrees of Freedom

Roll-Sway Coupling If there is a net sway force on the vessel such that the sway force
to starboard is not equal to the sway force to port then this will result in a net rotational
moment to port or starboard about the x axis, otherwise known as roll.

Sway-Yaw Coupling If the sway force on the forward end of the vessel is not equal
to the sway force on the aft end of the vessel then this will result in a net rotational
moment of the vessel about the z axis, or yaw.

It should be noted that if a ship normally operates at an angle of heel then all six degrees
of freedom are coupled.

2.3 Motion Resonance

Each of these modes of motion occurs due to an external disturbance force acting on the
vessel.

A vessel subjected to surge, sway or yaw will not return to its original position follow-
ing the application of an external force or moment but instead will be subjected to a
permanent alteration of course or attitude. This means that there is no resonant fre-
quency associated with these three modes of motion. The modes of roll, heave and pitch,
however, all contain a hydrostatic component which returns the ship to its equilibrium
position following a disturbance. Therefore these three modes have a resonant frequency,
and a degree of damping, associated with them.

Vertical Motions: Heave and pitch have a resonance associated with them, are mod-
erately damped (see Figure 2.3) and are strongly coupled. Surge has no resonance and
is weakly coupled.

Lateral Motions: Roll has a resonance associated with it and is lightly damped (see
Figure 2.3). Sway and yaw have no resonance associated with them.

Figure 2.3: Different levels of motion damping

6
Chapter 3

Basic Statistics

This chapter deals with statistics related to random variables. It is a summary of the
aspects of the topic required to obtain an understanding of seakeeping.

Consider a sample space containing n observations of a physical event. Each observation

has a numerical value X(n) such that the number of possible outcomes of the event is
< X(n) < . The Probability Distribution Function (or Cumulative Distribution
Function) F (x) is the probability of any X(n) being less than or equal to any real number
x

P [X(n) x] = F (x). (3.1)

Note that F () = 0 and F () = 1. Therefore we can also say that

P [X(n) > x] = 1 F (x) (3.2)

and

P [x < X(n) x + 4x] = F (x + 4x) F (x). (3.3)

The Probability Density Function (PDF) is defined as

dF (x)
f (x) = (3.4)
dx

with the area under the curve being equal to

Z
f (x)dx = F () F () = 1.0. (3.5)

7
Seakeeping Lecture Notes 2017 - Chapter 3: Basic Statistics

The probability P [x < X(n) x + 4x] is found from the PDF as

Z x+4x
P [x < X(n) x + 4x] = F (x + 4x) F (x) = f (x)dx. (3.6)
x

The mean value of a random variable X is

Z
E[X] = xf (x)dx = (3.7)

and the mean square is

Z
2
E[X ] = x2 f (x)dx = m0 . (3.8)

The variance of the random variable with respect to its mean value is

2 = E[(X )2 ] = E[X 2 ] 2 . (3.9)

If the random variable has a zero mean (as a regular wave has) then 2 = m0 . The
positive root of the variance, is the standard deviation or the root mean square value.

Two typical distributions in common use for Seakeeping - the Gaussian and Rayleigh
distributions, are summarised in Table 3.1.

Table 3.1: Typical Probability Distribution Functions

Distribution Gaussian (Normal) Rayleigh
 2
  
x2
Probability Density Function f (x) = 1
2
exp (x)
2 2
f (x) = x
2
exp 2 2

Wide band data

Uses Narrow banded data Peak values
White noise data

Graphical form

It should be noted that other distributions exist (e.g. the Weibull distribution) which
can be used for extreme rather than linear sea states, but these are not covered here.

8
Chapter 4

Description of Regular & Irregular Waves

The key to solving the problem of seakeeping is to determine how to model the environ-
ment (i.e. the sea). The chosen technique is to use a Fourier series and the method of
superposition. This is a frequency domain analysis and allows the summation of a series
of regular waves of many different wave frequencies and amplitudes in order to simulate
an irregular (or random) wave system (Figure 4.1). Conversely, if a record of an irreg-
ular sea at a particular point is available then it is possible to, using Fourier analysis,
decompose the irregular wave into the individual regular wave components that produce
that particular sea state.

Figure 4.1: Addition of regular waves of various frequencies to form irregular seas

4.1 Regular Waves

A typical regular wave (Figure 4.2) moving left to right along the x-axis with the z-axis
being the vertical axis (positive upwards) has the following properties:

Phase : This determines the regularity of the wave. A sinusoidal (regular) wave has
a repetition or phase of 2 radians.
9
Seakeeping Lecture Notes 2017 - Chapter 4: Description of an Irregular Sea

Figure 4.2: A typical regular (sinusoidal) wave profile

Wavelength : This is the horizontal distance between consecutive points in phase on

the wave (e.g. consecutive crests, troughs, zero up-crossings or zero down-crossings). It
is typically quoted in metres.

Wave period T : This is the time elapsed between equivalent successive locations on
the wave passing a fixed point. For example the time between two crests passing a fixed
point. Measured in seconds.

Wave phase speed, c: The velocity of an individual wave crest moving in the x
direction, passing a fixed point in space. This is measured in metres/second and can
also be referred to as wave speed, wave velocity or wave celerity. It is calculated as

c= (4.1)
T

Wave frequency, f : A measure of how often a wave crest (or trough) passes a fixed
point. Measured in 1/seconds.

1
f= (4.2)
T

Wave frequency can also be quoted as a wave radian frequency, measured in rad/s.

2
= 2f = (4.3)
T

Wave phase speed can therefore also be quoted as:

c= (4.4)
2
10
Seakeeping Lecture Notes 2017 - Chapter 4: Description of an Irregular Sea

Wave number: The wave number is defined as

2
k= (4.5)

which allows the wave phase speed to be expressed as

c= . (4.6)
k

Wave steepness: The wave steepness relates the wave height to the wave length,

H 2a
= = (4.7)

This expression is sometimes expressed in terms of radians so

2a
= = ka (4.8)

4.2 Dispersion Relationship

Waves travel on water at different speeds. The speed at which they travel depends
on their length and the water depth in which they are travelling. This is referred to
as dispersion. The dispersion relationship relates wave phase speed, wave number and
water depth d as

g g
c2 = tanh (kd) = tanh (kd) (4.9)
2 k

where g is the gravitational constant (= 9.81ms2 ). In terms of this is

2 = gk tanh (kd). (4.10)

A wave is considered to be in deep water when d > /2 such that d is sufficiently large
that d and tanh kd 1.0 and c2 = g/k = 2 /k 2 . In shallow water tanh kd kd
and c2 = g/k kd = gk. This means that all waves propagate at the same wave phase
speed in shallow water (independant of their wavelength) which is proportional to the
square root of the water depth.

A useful table relating all the quantities describing a regular wave in deep water in terms
of their formulae can be found on page 27 of A. R. M. Lloyd (1998) Seakeeping: Ship
behaviour in rough weather.
11
Seakeeping Lecture Notes 2017 - Chapter 4: Description of an Irregular Sea

4.3 Sinusoidal Wave Theory

The wave elevation of a regular, sinusoidal wave is defined as

= A cos (kx t). (4.11)

At a fixed point in time, e.g. t = 0 the wave elevation becomes

= A cos (kx) (4.12)

whilst for a fixed point in space, e.g. x = 0 it becomes

= A cos (t). (4.13)

Waves which are propagating in a single direction are termed uni-directional. If a ship
is travelling at an angle from the predominant wave system direction (measured from
the direction of propagation) then the wave elevation relative to a datum point becomes

= A cos(kx cos + ky sin t). (4.14)

The pressure in a wave can be estimated from

P = g(d + ) + Patmostpheric (4.15)

and the energy under a regular wave is

gA2
E= . (4.16)
2

4.4 Wave Encounter Frequency

So far we have dealt with the properties that describe the wave. However what is also
important is the direction at which the moving vessel encounters the waves. Figure 4.3
shows a ship travelling in waves which are at an oblique angle to the path of the ship.
We assume that the ship is maintaining a steady speed, U and a straight line course.

We can define the apparent velocity of the ship relative to the wave direction of propa-
gation as

c U cos . (4.17)

Since we know that T = /c (Equation 4.1) we can define the apparent period of
encounter of the waves by the ship as
12
Seakeeping Lecture Notes 2017 - Chapter 4: Description of an Irregular Sea

Figure 4.3: Waves at an oblique angle to a travelling ship

2
Te = = (4.18)
c U cos e

Therefore
c U cos
e = 2 (4.19)

which can be expanded into

2c 2U cos
e = (4.20)

Since we know that = 2c/ (Equation 4.4) and k = 2/ (Equation 2.5) then we
can rewrite this expression as

e = kU cos . (4.21)

This is known as the encounter frequency of the wave for the ship and it is this frequency
that causes the ship to oscillate and experience motions.

If we look at the expression for encounter frequency in deep water then we know that
c2 = g/k = 2 /k 2 (Equation 4.6) and and therefore 2 /g = k. Hence the expression for
encounter frequency can be rewritten for deep water to be
13
Seakeeping Lecture Notes 2017 - Chapter 4: Description of an Irregular Sea

2
e = U cos . (4.22)
g

Whether the encounter frequency is less than, greater than or equal to the wave frequency
is dependant on the direction at which the waves are being encountered. In head seas
= 180 degrees (the ship travels directly into the waves) and

2U
e = + . (4.23)
g

Therefore in head seas each wave frequency will correspond to one wave encounter fre-
quency.

In following seas the heading angle is = 0 degrees (the ship travels with the waves)
and
2U
e = (4.24)
g

In following seas the same wave encounter frequency may occur for multiple (up to three)
different wave frequencies depending on whether the ship is overtaking the waves or vice
versa (Figure 4.4). Positive encounter frequencies correspond to the ship overtaking
the waves, whilst negative encounter frequencies correspond to the waves overtaking the
ship.

Figure 4.4: Example e calculation allowing for wave direction, U = 19.07kts

Figure 4.5 is a diagram detailing wave directions and the associated relationship between
wave frequency and encounter frequency.
14
Seakeeping Lecture Notes 2017 - Chapter 4: Description of an Irregular Sea

Figure 4.5: Influence of wave direction on e

15
Seakeeping Lecture Notes 2017 - Chapter 4: Description of an Irregular Sea

4.5 Characteristics of an Irregular Sea

We assume that an irregular seaway is represented by the summation of a series of

regular (sinusoidal) waves all of which have a random amplitude and phase. If a regular
wave is represented by the equation

(x, t) = A cos (kx t + ) (4.25)

where k is the wave number, the wave frequency and the wave phase then an irregular
sea is represented by the Fourier series

(x, t) = A1 cos (k1 x 1 t + 1 )+A2 cos (k2 x 2 t + 2 )+...+AN cos (kN x N t + N )

(4.26)

or

(x, t) = N
n=1 An cos (kn x n t + n ) (4.27)

where N is the number of regular waves making up the seaway.

An irregular sea can be classed as narrow banded or wide banded (Figure 4.6).

Figure 4.6: Wide and narrow banded sea states

A narrow banded sea state has a peak wave period Tp (the time between successive
crests on the wave) approximately equal to the zero crossing period Tz (the time between
successive upward or successive downward zero-crossings) and includes only a small range
16
Seakeeping Lecture Notes 2017 - Chapter 4: Description of an Irregular Sea

of wave frequencies (Note that for a regular wave Tp = Tz ). A wide banded sea state
has Tp << Tz and includes a large range of wave frequencies.

Figure 4.7 illustates the properties that make up an irregular sea:

Peak wave period Tp : The time between successive peaks (or crests) on the wave
record. This quantity varies with time.

Zero-crossing period Tz : The time between successive upward or successive downward

zero-crossings of the wave record. This quantity varies with time.

Wave elevation : The elevation of the wave from the zero level at any time.

Apparent wave amplitude a : The distance from the zero level to any wave peak.

Apparent wave height Ha : The distance between any successive peak and trough.
Can be measured trough-to-peak or peak-to-trough.

Figure 4.7: Characteristics of an irregular seaway

The more regular wave components used to generate an irregular seaway, the more truely
random the seaway becomes. This is illustrated in Figure 4.8.

Figure 4.8: Effect of increasing the number of regular wave components making up an
irregular sea

17
Seakeeping Lecture Notes 2017 - Chapter 4: Description of an Irregular Sea

4.6 Statistical Analysis of an Irregular Sea

When carrying out a statistical analysis of an irregular sea a suitably long wave record
has to be used in order for the analysis to give an accurate picture of the statistical
properties of the sea state. Typically at least 100 crests and troughs are required,
requiring a wave record of at least 30 minutes. The longer the wave record, the more
accurate the statistical analysis will be. The wave elevation of a suitably long wave
record will follow the Gaussian distribution with a zero mean value.

4.6.1 Probability Density Function

Data sampled from a wave record can be plotted as a Probability Density Function
(PDF). This can be plotted for any sampled data - wave period, amplitude, height or
length, obtained from the irregular sea. For example, for a set of wave amplitude data,
assuming that the data has been sampled and grouped into ranges such that N samples
are available and the number of samples within each range of data (range width ) is
ni then the probability of the wave amplitude occurring within a particular range is

ni
(4.28)
N

and the probability density function is

ni 1
f (i ) = . (4.29)
N

It should be the case that

ni
= 1.0 (4.30)
N

and therefore the probability density function can be found as a percentage by multipling
f (i ) by 100. Example PDF data for sampled wave amplitude of N = 1868 and = 0.2
and the associated PDF plot is given in Figure 4.9. The area under the PDF curve will
add up to 1.0.

The data given in the table in Figure 4.9 can also be used to calculate three other
quantities - the mean , variance m0 and standard deviation 0 (also called the root
mean square or RMS) of the data:

ni
= N
i=1 i (4.31)
N

ni
m0 = N 2
i=1 i (4.32)
N
18
Seakeeping Lecture Notes 2017 - Chapter 4: Description of an Irregular Sea

Figure 4.9: Calculation and plot of PDF for wave amplitude

0 = m0 (4.33)

4.6.2 Probability Distribution Function

An alternative method of presenting the data is as a Probability Distribution Function

or Cumulative Distribution Function. This is the probability of any sample being less
than or equal to a particular value X, and is calculated as the summation of all prob-
abilities up to and including P (X). So for the data shown previously in Figure 4.9,
the corresponding Probability Distribution Function is shown in Figure 4.10 with the
Probability Distribution Function calculated as

ni
F (i ) = i1 . (4.34)
N

Figure 4.10: Calculation and plot of Probability Distribution Function for wave ampli-
tude
19
Seakeeping Lecture Notes 2017 - Chapter 4: Description of an Irregular Sea

The discrete data for the wave record can be compared to the standard (continuous)
distributions covered in Chapter 3. Therefore the Rayleigh probability density function
for peak wave amplitudes will be

2
 

f () = exp . (4.35)
m0 2m0

where is the observed positive peak wave amplitudes and m0 the mean square of the
wave elevation. The probability distribution function will be

2
 
F () = 1 exp (4.36)
2m0

These relationships are compared to the sampled data on Figures 4.9 and 4.10 (repre-
sented by the red line). The smaller the range width for which the data is sampled, the
closer it will appear to the continuous Rayleigh distribution.

The mean of the peak wave amplitudes, and probability of the peak wave amplitude
exceeding a particular value can therefore be found from the Rayleigh distribution using
the relationships in Chapter 3.

4.6.3 Significant Wave Height and Wave Amplitude

We can obtain, for a particular wave record, all the values of Ha and a that occur in that
record, and calculate a mean value (note that mean values can also be obtained for wave
periods in the same manner). However there are two further quantities can be calculated
which are commonly used to describe an irregular sea. These are the significant wave
height and significant wave amplitude. Significant wave amplitude 1/3 is the mean
value of the highest one-third of the sampled measurements of a whilst significant wave
height H1/3 is the mean value of the highest one-third of the sampled measurements
of Ha . Essentially this means that for N measurements of either wave height or wave
amplitude ranked from largest to smallest, then the significant value in each case is the
mean of the first N/3 values in the list. The two quantities are related as

H1/3 = 21/3 . (4.37)

It should be noted that sometimes additional values of H1/10 and 1/10 are calculated to
give a more in-depth analysis of the statistics of a wave record. These are calculated in
the same way, but using the highest one-tenth of the sampled values.

The Rayleigh probability density function can be used to find an alternative expression
for significant wave amplitude and wave height. Defining this parameter as the average
value of the biggest one third of the highest observed peaks, we can calculate
20
Seakeeping Lecture Notes 2017 - Chapter 4: Description of an Irregular Sea

2
Z  
1
= exp d (4.38)
3 a1/3 m0 2m0

where a1/3 is the value of wave amplitude where the top one-third of the observations
start. Therefore


2
 
1
= exp (4.39)
3 2m0 a1/3

and
a21/3
!
1
= exp (4.40)
3 2m0

Hence
a21/3 = 2m0 ln 3 (4.41)

and

a1/3 = 1.482 m0 . (4.42)

Thus the significant wave amplitude is the average of the wave amplitude values between
a1/3 . Using the equation for the average value of a probability distribution,
the average of the highest one-third values is

2
Z  

1/3 = exp d. (4.43)
a1/3 m0 2m0

If this integration is performed, then we find that

1/3 = 2 m0 (4.44)

hence

H1/3 = 4 m0 . (4.45)

We can also define wave parameters using the n1 th significant values, which are found in
the same manner. The general equation is

p
2
1/n = 2m0 ln n 1/n = 2m0 ln n. (4.46)

For example, another common value to calculate is 1/10 = 2.55 2m0 .

21
Chapter 5

Wave Spectra

5.1 Spectral Analysis

If we take a simple sine wave = A sin (t) then we can find the mean square value for
that wave as

T T
A2
Z Z
1 2 2
m0 = A sin (t)dt = (1 cos (2t))dt (5.1)
T 0 2T 0

using trigonometric identities, where T = 2/ is the period of one wave. Therefore

2/
A2

1
m0 = t sin (2t) (5.2)
2T 2 0

which simplifies to give the mean square of a regular wave as

1
m0 = A2 (5.3)
2

and the root mean square as

r
A2
= m0 = = 0.707A. (5.4)
2

If we carry out the same analysis for a set of frequencies, all of which are multiples of
such that the seaway is constructed from a Fourier Series as

= N
m=1 Am cos (mt) (5.5)
22
Seakeeping Lecture Notes 2017 - Chapter 5: Wave Spectra

then

2
2 = N
m=1 Am cos (mt) = N N
m=1 Am cos (mt) n=1 An cos (nt) (5.6)

which can be written as

2 = N N
m=1 n=1 Am An cos (mt) cos (nt). (5.7)

Therefore the mean square is

Z T
1
m0 = N N Am An cos (mt) cos (nt)dt (5.8)
T m=1 n=1 0

which, using trigonometric identities and substituting in T = 2/ becomes

2/    
nm
Z
N 1 n+m
m0 = N Am An cos t + cos t dt (5.9)
2 m=1 n=1 2 0 2 2

Hence


0 n 6= m
m0 = 1 N 2 (5.10)
2 m=1 Am n=m

showing that the mean square of the free surface is the summation of the individual mean
square values for each individual regular wave making up the sea surface. It should be
noted that phase has no direct relevance for the mean square value as it does not change
the frequency content, it just changes the shape of the free surface. Therefore for this
relationship it can be ignored.

In reality a sea state is not composed of a set of evenly spaced frequencys , 2, 3...N
but is made up of an unlimited set of frequencies r where r is not an integer. Therefore
the free surface can be represented, in its simplest form, as

Z
(t) = Ar cos (r t)dt (5.11)
0

with the mean square value still being the summation of all the individual mean square
values:

1
m0 = 2 = A2 (5.12)
2 r=1 r
23
Seakeeping Lecture Notes 2017 - Chapter 5: Wave Spectra

We can plot the distribution of the mean square value as a function of frequency. Since
we know that the energy in a regular wave is E = gA2 /2 then we can see that the
mean square value for a regular wave, A2 /2 is proportional to the energy present in that
regular wave. Hence the plot of distribution of the mean square with wave frequency is
referred to as the Wave Energy Spectrum or Wave Spectrum, given the symbol S(). A
typical plot is in Figure 5.1.

Figure 5.1: A typical wave energy spectrum

A wave spectrum can be used to find the wave amplitude associated with a particular
wave frequency by evaluating the energy present at that wave frequency. Looking at
Figure 5.1, if we take p = r 4/2 and q = r 4/2 then we can say that the
energy contained in the sea state between p and q is g multiplied by the area under
the graph. Therefore

E(r ) = gS(r ) 4 . (5.13)

We can equate this to the energy in a regular wave as

g 2
= gS(r ) 4 (5.14)
2

Hence
24
Seakeeping Lecture Notes 2017 - Chapter 5: Wave Spectra

2 = 2S(r ) 4 (5.15)

and therefore the amplitude of the wave which represents the energy content of the
component wave at frequency r (associated with frequency band 4) is
p
= 2S(r ) 4 . (5.16)

The total energy in the sea state (equal to the mean square value m0 ) is therefore

Z
1 1
m 0 = 2
r=1 r = r=1 2S(r ) 4 = S()d (5.17)
2 2 0

as 4 0. m0 is referred to as the zeroth moment of the wave energy spectrum.

We can also find the nth moment of the wave energy spectrum as

Z
mn = n S()d. (5.18)
0

This can be used to find the key wave periods, hence frequencies, of a particular sea
state. The zero crossing wave period and frequency are defined as

r
m0 2
T0 = 2 = (5.19)
m2 0

whilst the average wave period and wave frequency are

m0 2
T1 = 2 = (5.20)
m1 1

We can relate the peak, zero crossing and average wave periods as

TP = 1.2957T1 = 1.40769T0 (5.21)

which means therefore that T1 = 1.08643T0 . Note that the peak wave period Tp can
also be found from dS()
d = 0. For the ITTC two-parameter spectrum, for example (see
4

4
Section 5.2.1), we find that p = 4B/5 and therefore Tp = 2/ 0.8B. The value of the
wave spectrum that corresponds to p is found by substituting p into the formula for
S().
25
Seakeeping Lecture Notes 2017 - Chapter 5: Wave Spectra

5.2 Standard Wave Spectra

Standard wave spectra exist which are used for open ocean and coastal estimates of wave
climates. Some of the more common ones are reproduced here.

5.2.1 ITTC

The International Towing Tank Committee (ITTC) two parameter spectrum or Bretschnei-
der spectrum (recommended for use in open ocean) is given by:

A B
S() = 5
exp 4 (5.22)

where A is calculated as
H1/3
A = 172.75 (5.23)
T14

and B as

691
B= (5.24)
T14

If only significant wave height, H1/3 , is known then A and B approximate to 8.11
103 g 2 and 3.11/H1/3
2 respectively. This wave spectrum is in most common use and is
recommended by the International Towing Tank Committee. Figure 5.2 demonstrates
how the spectrum changes appearance with a change in (a) significant wave height and
(b) peak wave period.

5.2.2 JONSWAP

The Joint North Sea Wave Project (JONSWAP) spectrum is recommended for fetch
limited (coastal) waters and is given by:

g 2 p4 a
S() = exp (5.25)
5 4

where p is the peak radian wave frequency of the spectrum, and

0.0081 0.01 (5.26)

( p )2
a = exp (5.27)
2p2 2
26
Seakeeping Lecture Notes 2017 - Chapter 5: Wave Spectra

Figure 5.2: Variation in the ITTC wave spectrum with changes in (a) significant wave
height and (b) peak wave period


0.07 p
= (5.28)
0.09 > p

5
= (5.29)
4

is the peak enhancement factor, generally taken as 3.3 for typical conditions. However
it can range between 1.0 and 7.0. Figure 5.3 demonstrates how the spectrum changes
appearance with a change in the peak enhancement factor. You should note that a
JONSWAP wave spectrum with = 1.0 reduces to the ITTC two-parameter spectrum
for the same significant wave height and peak wave period.

5.2.3 Pierson-Moskowitz

The Pierson-Moskowitz wave spectrum accounts for the fact that wave formation is
heavily influenced by the wind strength, therefore including the wind speed U in the
calculation:

8.1 103 g 2 g 4
S() = exp 0.74 (5.30)
5 U

where g is the acceleration due to gravity.

27
Seakeeping Lecture Notes 2017 - Chapter 5: Wave Spectra

Figure 5.3: JONSWAP wave spectra for a range of values of peak enhancement factor

5.3 Wave Encounter Spectrum

The wave enery spectra discussed so far have been for waves of frequency . However, if
a ship is travelling in the waves, with a forward speed of U then the relevant frequency
that the ship sees becomes the wave encounter frequency,

2U
e = cos . (5.31)
g

The total energy present in the wave spectrum remains the same regardless of whether
the ship is stationary or moving. Therefore we can say that

Z Z
m0 = S()d = S(e )de (5.32)
0 0

which means that

S()d = S(e )de (5.33)

and therefore the wave encounter spectrum can be calculated from the wave spectrum
as

S()
S(e ) =  de  . (5.34)
 
d

28
Seakeeping Lecture Notes 2017 - Chapter 5: Wave Spectra

From the expression for wave encounter frequency we can obtain

de 2U
=1 cos (5.35)
d g

and therefore

S()
S(e ) =  . (5.36)
2U
1 cos 

g

Figure 5.4 demonstrates how a typical wave encounter spectrum will change with in-
creased ship speed.

Figure 5.4: Demonstration of change in wave encounter spectrum with ship speed

29
Chapter 6

Ship Response

6.1 Ship Response as a Linear System

We can consider a ship as a linear system. The wave is the input to the system and the
motion response of the ship is the output from the system. Therefore we can assume
that (for small amplitude motions) the total motion of the ship can be calculated as
the summation (linear superposition) of the motion responses to each individual regular
wave making up the seaway. This linear system is illustrated in Figure 6.1. For a
simple regular wave input of = 0 cos (e t) the motion response of the ship will be
= i0 cos (e t + i ) where i = 1, 2, 3 are the linear (translational) motions, in metres,
and i = 4, 5, 6 are the rotational motions, in radians. i gives the phase between the
input wave and each output motion.

Figure 6.1: A ship as a linear system

6.2 Equations of Motion for a Ship in a Seaway

The three resonant responses (heave z, pitch and roll ) can be represented as a simple
2nd order equation of motion
30
Seakeeping Lecture Notes 2017 - Chapter 6: Ship Response

m + c + k = F (, ) = F0 exp (it) (6.1)

where c is the damping force, k is the restoring force and m is the displacement or
mass moment of inertia. is the motion in question, either heave, pitch or roll.

The solution to this equation is

F0 cos (t + )
=p (6.2)
(k m 2 )2 + c2 2

with the phase, , being  

1 c
= tan (6.3)
k m 2

We normally express the motion in terms of the natural frequency 0 and non-dimensional
damping associated with it:

r
k
0 = (6.4)
m

c
= (6.5)
2m0

If we define a non-dimensional frequency as /0 then we can re-write the solution to

the equation as

F0
0 = p (6.6)
(1 )2 +
2 (2)2

where 0 is the amplitude of the relevant motion.

6.3 Response Amplitude Operators

From the motions of a ship in regular waves we can obtain the transfer function, or
Response Amplitude Operator (RAO) that relates the ship motion at each wave fre-
quency to a wave of unit amplitude at that frequency. The RAO, H is therefore the
non-dimensionalised motion, non-dimensionalised with respect to wave amplitude for
linear motions and wave slope for rotational motions:

(
i
0 i = 1, 2, 3
Hi = i (6.7)
k0 i = 4, 5, 6

31
Seakeeping Lecture Notes 2017- Chapter 6: Ship Response

Essentially this represents the proportion of the wave amplitude or the wave slope am-
plitude which is converted into a motion by the ship. This data can be plotted for
each motion, wave direction and ship speed. Assuming a linear system (hence small
amplitude waves) the magnitude of the RAO is independant of wave amplitude. In a
nonlinear system (i.e. large amplitude waves) then the magnitude of the RAO becomes
dependant on wave amplitude, particularly around resonance.

Figure 6.2 shows typical heave and pitch RAOs in head seas, including the influence of
ship speed.

Figure 6.2: Influence of speed on heave and pitch RAOs in head seas

32
Seakeeping Lecture Notes 2017 - Chapter 6: Ship Response

A typical presentation of RAO data is reproduced in Figure 6.3 for all six degrees of
freedom and five different heading angles.

Figure 6.3: Typical presentation of RAO data for all six degrees of freedom
33
Seakeeping Lecture Notes 2017 - Chapter 6: Ship Response

6.4 Response Spectra and RMS

The RAOs derived in the previous section can be used to obtain the Motion (Response)
Spectra for the ship in irregular waves. If we define a particular response spectrum Ei ()
then we can write (in the same way as for a wave spectrum)

p
i () = 2Ei () 4 (6.8)

Since we know that

()
H() = (6.9)
0 ()

then we can say that

p p
2Ei () 4 = 2S() 4 Hi () (6.10)

and therefore for motion i

Ei () = Hi2 ()S() (6.11)

where E() is the motion response spectrum in an irregular sea, H() is the response
of the ship to regular waves - the RAO or transfer function, and S() is the wave
spectrum for the irregular sea. This relationship can also be written in terms of the
wave encounter frequency, e which is a more realistic representation of the response
spectrum and recommended for use where possible:

Ei (e ) = Hi2 (e )S(e ). (6.12)

Figure 6.4 shows the relationship between wave spectrum, RAO and response spectrum
for a frigate at zero speed, for heave on the left hand side and pitch on the right hand
side.

As with the wave spectrum, we can also obtain the mean square values for the response
spectrum, for each motion and its associated velocity and acceleration. For example the
mean square values of heave motion, heave velocity and heave acceleration are evaluated
as:

Z  Z 
m0z = Ez (e )de = Hz2 (e )S(e )de (6.13)
0 0

Z
m0z = e2 Ez (e )de (6.14)
0

34
Seakeeping Lecture Notes 2017 - Chapter 6: Ship Response

Figure 6.4: Wave spectrum, RAO and response spectrum for heave (left) and pitch
(right)

Z
m0z = e4 Ez (e )de (6.15)
0

The root mean square values and significant (one-third) values of the motion can then
be evaluated as before, as can the probabilities of the motion exceeding a particular
value. For example, for heave:

H1/3z = 4 m0z (6.16)
35
Seakeeping Lecture Notes 2017 - Chapter 6: Ship Response

z2
 
P (Z > z) = exp (6.17)
2m0z

36
Chapter 7

Summary

Seakeeping involves the study of the response of a ship subject to a known wave
force. It is important for the assessment of how a ship will behave under normal
operating conditions, and when subject to phenomena such as breaking waves,
slamming and water on deck.
A ship is subject to six coupled degree of freedom motions (both linear and rota-
tional): (1) Surge, (2) Sway, (3) Heave, (4) Roll, (5) Pitch and (6) Yaw. These are
known as rigid body modes and can be de-coupled into two sets - vertical (Heave,
Pitch and Surge) and lateral (Roll, Sway and Yaw) assuming that the vessel is
vertically symmetrical.
An irregular sea can be modelled by the linear superposition of regular wave com-
ponents of varying frequency, amplitude and phase. The more components used
to create an irregular sea, the more truely random that sea becomes.
The component waves making up an irregular sea can be represented as Wave
Spectra. Standard Wave Spectra exist which can be used to generate an irregular
sea with a defined peak wave period and significant wave height.
The response of a ship in an irregular sea can be modelled by the linear super-
position of the responses to the individual component waves which when added
together create that irregular sea.
It is important to consider the encountered wave frequency, and the encountered
wave spectrum, when modelling the response of a vessel. These depend on the
ship speed and the angle at which the ship encounters the waves.
A Response Amplitude Operator (RAO) can be calculated for a ship for each of the
six motions at any heading angle and ship speed and a range of wave frequencies
(or encountered wave frequencies) by non-dimensionalising linear motions by wave
amplitude, and rotational motions by wave slope.
A Response Spectrum for a ship can be calculated from the Wave Spectrum (or
Encountered Wave Spectrum) and the RAO (or Transfer Function) for a particular
motion.

37