^-..W.

cxv

LOW FREQUENCY SECOND ORDER

WAVE EXCITING FORCES ON
FLOATING STRUCTURES

J.A.PINKSTER
LOW FREQUENCY SECOND ORDER WAVE EXCITING FORCES ON
FLOATING STRUCTURES
 LOW FREQUENCY SECOND ORDER Dit proefschnft is goedgekeurd
door de promotoren
WAVE EXCITING FORCES ON
FLOATING STRUCTURES Prof. Dr. Ir. A. J. Hermans
Prof. Ir. J. Gerritsma

PROEFSCHRIFT
TER V E R K R U G I N G VAN DE G R A A D VAN D O C T O R IN DE
T E C H N I S C H E W E T E N S C H A P P E N AAN DE T E C H N I S C H E
H O G E S C H O O L TE D E L F T , OP G E Z A G VAN DE R E C T O R
MAGNIFICUS,
VOOR EEN C O M M I S S I E A A N G E W E Z E N D O O R HET C O L L E G E VAN
D E K A N E N TE V E R D E D I G E N OP W O E N S D A G 8 O K T O B E R 1980
TE 14.00 U U R

DOOR

J O H A N N E S ALBERT PINKSTER
SCHEEPSBOUWKUNDIG INGENIEUR
G E B O R E N TE E M S W O R T H

H. V E E N M A N E N Z O N E N B . V . - W A G E N I N G E N
 STELLINGEN aanbeveling bij het ontwerpen van dergelijke systemen ook hier aandacht aan te
besteden.
I
VIII
Bij het bepalen van de laag frekwente bewegingen van een in golven afgemeerde
konstruktie dient rekening gehouden te worden met viskeuze effekten in de De laag frekwente golfdriftkracht op een afgemeerde konstruktie in onregel-
hydrodynamische reaktiekrachten. matige golven kan, mits aan bepaalde voorwaarden wordt voldaan, voor-
speld worden aan de hand van gegevens over de gemiddeide driftkracht in
II regelmatige golven.
Het relatieve bewegingsprincipe geeft inzicht in de krachten uitgeoefend door IX
variabele dwarsstroom op een varend schip en is een goed uitgangspunt voor het
bepalen van de grootte van deze krachten. Een schip varend in recht voor- of recht achterinkomende golven kan grote
slingerbewegingen vertonen bij golf frekwenties die ver buiten de eigen slinger
Ill frekwentie van het schip liggen.
De afmetingen van een semi-submersible kunnen zodanig geoptimaliseerd X
worden dat de bewegingen met golf frekwenties minimaal zijn. Het verdient
overweging eenzelfde optimalisatie toe te passen met betrekking tot de tweede Op grote schepen dienen gegevens aan boord aanwezig te zijn met betrekking tot
orde golfdriftkrachten. de gemiddeide golf-, wind- en stroomkrachten. Deze gegevens zijn van belang
voor het geval dat een schip ten gevolge van averij geassisteerd moet worden
IV door sleepboten.

Spelende kinderen die een bal in het water gooien maken dikwijls gebruik van XI
tweede orde golfdriftkrachten om deze weer binnen bereik te drijven. Dit is, tot
nu toe, het enige nuttige gebruik van dit verschijnsel wat bekend is. Meer jarenplannen voor onderzoek kunnen belemmerend werken op de
kreativiteit van de onderzoeker.
V
XII
Door sommige onderzoekers is beweerd dat golfdriftkrachten op een
afgemeerde konstruktie in stroom en golven afhankelijk zijn van de Het is aan te bevelen de voor de Nederlandstalige leek nogal lachwekkende
stroomsnelheid. Onderzocht moet worden of het juist niet de stroomkrachten benaming 'driftkrachten' te vervangen door een uitdrukking die minder
zijn die veranderen door de aanwezigheid van golven. associaties met menselijk gedrag oproept.

VI Dissertatie J. A. Pinkster
Delft, 8 oktober 1980
Doordat energie in golven nauw verwant is aan het verschijnsel van de laag
frekwente golfdriftkrachten kan dit gebruikt worden voor het verminderen van
laag frekwente horizontale bewegingen van konstrukties afgemeerd in golven.

VII

Systemen voor het opwekken van energie uit golven zijn veelal gekoncipieerd op
basis van het gedrag van het golfoppervlak. Daar de energie van golven
gedragen wordt door de waterdeeltjes onder dit oppervlak verdient het
Aan Martha
en de kinderen
 CONTENTS IV.4. Time domain representation of the mean and low
frequency second order forces 59
I. INTRODUCTION 1 IV. 5 . Conclusions 60

II. PAST DEVELOPMENTS CONCERNING T H E COMPUTATION OF MEAN V. COMPARISON BETWEEN RESULTS OF COMPUTATIONS AND ANALYTI­
AND LOW FREQUENCY WAVE FORCES 8 CAL RESULTS ON THE MEAN WAVE DRIFT FORCE IN REGULAR
II. 1. Introduction 8 WAVES 61
11.2, Historical review 8 V. 1. Introduction 61
II. 3. Conclusions 15 V.2. Computations 61
V.2.1. General 61
III. HYDRODYNAMIC THEORY 19
V.2.2. Motions and mean horizontal drift force 67
111.1. Introduction 19
V.2.3. Components of the mean horizontal drift force .. 69
111.2. Co-ordinate systems . 22
V.2.4. Mean vertical drift force 71
111.3. Motion and velocity o f a point on the hull of
V. 3. Conclusions 72
the body 23
111.4. Fluid motions and boundary conditions 24 VI. COMPARISON BETWEEN COMPUTATIONS AND MEASUREMENTS OF THE
III.4.1. Boundary conditions within the fluid, at the MEAN SECOND ORDER FORCE IN REGULAR WAVES 73
free surface and on the sea floor 24 VI. 1. Introduction 73
III. 4.2. Boundary conditions on the body 26 VI.2. Model tests 76
III.4.3. Boundary conditions at infinity 30 VI.2.1. General 76
111.5. Pressure in a point within the fluid 30 VI.2.2. Model test conditions 79
III. 6. Second order wave force and moment 32 VI. 3. Computations 82
111.6.1. Second order wave force 32 VI. 4. Results of computations and measurements 88
111.6.2. Second order wave moment 35 VI. 5 . Conclusions 102
III. 7 . Conclusions 37 VII. DETERMINATION OF THE QUADRATIC TRANSFER FUNCTION OF THE
IV. EVALUATION OF T H E SECOND ORDER WAVE EXCITING FORCES ... 38 LOW FREQUENCY SECOND ORDER FORCES 104
IV. 1. Introduction 38 VII.1. Introduction 104
IV.2. The quadratic transfer function 38 VII.2. Model test set-up 105
IV.2.1. General 38 VII.2.1. General 105
IV.2.2. Evaluation of the components dependent on first VII.2.2. Realizations of two systems of restraint Ill
order quantities 41 VII.3. Model tests 116
IV.2.3. Contribution of the second order potential .... 42 VII.3.1. Generation of waves 116
IV.2.3.1. General 42 VII.3.2. Test procedure and duration of measurements .. 120
IV.2.3.2. Approximation for the contribution of the VII.4. Analysis of results of measurements of the low
second order potential 45 frequency longitudinal force in head waves 121
IV.2.3.3. Comparison between exact results and the VII.4.1. Regular wave groups 121
approximation 50 VII.4.2. Irregular waves 122
IV.2.4. Symmetry of the quadratic transfer functions .. 53 VII.5. Computations 122
IV.3. Frequency domain representation of the mean and VII.6. Comparison between computations and experi­
low frequency forces in irregular waves 57 ments 125
 VII.7. Approximation for the low frequency force in
irregular waves 132
VII.8. Conclusions 135

VIII. APPLICATION OF THEORY TO DYNAMIC POSITIONING OF A

VESSEL IN IRREGULAR WAVES 137
VIII.1. Introduction 137
VIII.2. Theoretical prediction of the effect of wave-
feed-forward 139
VIII.3. Generation of the wave-feed-forward control
signal 145
VIII.4. Positioning system 151
VIII.5. Model tests 155
VIII.5.1. General 155
VIII.5.2. Results of tests in irregular waves 155
VIII.5.3. Results of tests in irregular waves and
current 160
VIII.6. Conclusions 162

IX. CONCLUSIONS 163

APPENDIX A - COMPUTATION OF THE FIRST ORDER SOLUTION FOR THE

VELOCITY POTENTIAL AND BODY MOTIONS 166

APPENDIX B - CROSS-BI-SPECTRAL ANALYSIS 173

REFERENCES 182

NOMENCLATURE 196

SUMMARY 200

SAMENVATTING 2 02

ACKNOWLEDGEMENT 204
I. INTRODUCTION
Stationary vessels floating or submerged in irregular
waves are subjected to large, so-called first order, wave forces
and moments which are linearly proportional to the wave height
and contain the same frequencies as the waves. They are also sub­
jected to small, so-called second order, mean and low frequency
wave forces and moments which are proportional to the square of
the wave height. The frequencies of the second order low frequen­
cy components are associated with the frequencies of wave groups
occurring in irregular waves.
The first order wave forces and moments are the cause of
the well known first order motions with wave frequencies. Due to
the importance of the first order wave forces and motions they
have been subject to investigation for several decades. As a re­
sult of these investigations, methods have evolved by means of
which these may be predicted with a reasonable degree of accuracy
for many different vessel shapes.
This study deals with the mean and low frequency second
order wave forces acting on stationary vessels in regular and
irregular waves in general and, in particular, with a method to
predict these forces on basis of computations. Knowledge concern-
ing the nature and magnitude of these forces is of importance
due to the effect they have been shown to have on the general
behaviour of stationary structures in irregular waves.
The components of mean and low frequency second order wave
forces can affect different structures in different ways and al­
though of the same origin have even been called by different
names. The horizontal components of the mean and low frequency
second order wave forces are also known as wave drift forces
since, under the influence of these forces, a floating vessel
will carry out a steady slow drift motion in the general direc­
tion of wave propagation if it is not restrained.
The importance of the mean and low frequency wave drift
forces from the point of view of motion behaviour and mooring
loads on vessels moored at sea has been recognized only within
the last few years. Verhagen and Van Sluijs [I-l], Hsu and Blen-
karn [1-2] and Remery and Hermans [1-3] showed that the low fre­
quency components of the wave drift forces in irregular waves
could, even though relatively small in magnitude, excite large
amplitude low frequency horizontal motions in moored vessels.
It was shown that in irregular waves the drift forces contain
components with frequencies coinciding with the natural frequen­
cies of the horizontal motions of moored vessels. Combined with
the fact that the damping of low frequency horizontal motions of
moored structures is generally very low, this leads to large
amplitude resonant behaviour of the motions. See Figure I-l.
RECORD OF AN IRREGULAR SEA
RECORD OF SURGE MOTION
20 m r
-20 m
100 200
TIME in sec.
Fig. I-l Low frequency surge motions of a moored LNG carrier in
Irregular head seas.
Remery and Hermans [1-3] established that the low frequency com­
ponents in the drift forces are associated with the frequencies
of groups of waves present in an irregular wave train. See Fig­
ure 1-2.
 CURRENT! WAVES
MEASUREMENTS
CONTROL
THRUSTER SIGNAL
VESSEL
ALLOCATION

E
c WIND PREDICTED
<r MEASUREMENTS
LJ
O \
WIND
tr FORCE
ID
I/)
Li.
WIND
MODEL N
VESSEL
MODEL -=6 STATE
CORRECTION
o
UJ
r\
Q
__> 7—' CURRENT
FORCE < FORCE
a. DEMAND V \
VESSEL
< STATE ~~ CURRENT WAVE
MODEL MODEL

NATURAL PERIOD OF SURGE MOTION O

A =PERIOD OF WAVE GROUPS. REF [l-3] SETPOINTS

Fig. 1-2 Surge motions of a moored barge in regular head wave
groups. Ref. [1-3].
Dynamically positioned vessels such as drill ships which
remain in a prescribed position in the horizontal plane through
the controlled use of thrust generated by propulsion units are
also influenced by mean and low frequency wave drift forces. The
power to be installed in these vessels is dependent on the magni­
tude of these forces. The frequency response characteristics of
the control systems must be chosen so that little or no power is
expended to compensate the large oscillatory motions with wave
frequencies, while the mean and low frequency horizontal motions
caused by the mean and low frequency drift forces should be re­
duced to values commensurate with the task of the vessel. This
has led to the development of sophisticated control systems.
See Figure 1-3.
Fig. 1-3 Block diagram for typical dynamic positioning system.
The vertical components of the second order forces are
sometimes known as suction forces. This term is generally applied
in connection with the mean wave induced vertical force and
pitching moment acting on submarine vehicles when hovering or
travelling near the free surface. It is shown by Bhattacharyya
tl-4] that in extreme cases the upward acting suction force due
to waves can cause a submarine vehicle to rise and broach the
surface, thus posing a problem concerning the control of the ve­
hicle in the vertical plane. See Figure 1-4.
The vertical components of the second order wave forces
have also been connected with the phenomena of the steady tilt of
semi-submersibles with low initial static stability as indicated
by Kuo et al [1-5]. Depending on the frequency of the waves it
has been found that the difference in the suction forces on the
floaters of a semi-submersible can result in a tilting moment,
which can cause the platform to tilt towards or away from the
system, is by means of model tests. In many practical cases suf­
oncoming waves. Such effects are of importance in judging the
ficient insight in the complex behaviour of, for instance, a
minimum static stability requirements for such platforms. From large tanker moored to a single point mooring system is still
observations in reality and from the results of model tests it lacking for reliable prediction of the motion behaviour and forces
has been found that large, deep floating storage vessels can in the mooring system to be made by means other than physical
carry out low frequency heave motions in irregular waves which model testing.
are of the same magnitude as the heave motions with wave frequen­
cies. Simulation techniques based on numerical computations are
becoming of increasing importance in the design phase of many
floating structures however. For instance, in order to evaluate
50 the effectiveness of control systems for dynamically positioned
vessels, time domain simulations, which take into account the
60 equations of motion of the vessel and the behaviour of external
•*- loads such as the mean and low frequency wave drift forces, are
c 70 carried out. In such cases, due to the complexity of the control
X system and the objectives of the study, it is more practical to
.-
Q. 80 make use of simplified equations describing the environmental
UJ
O forces and the reaction of the structure or vessel to external
_J
UJ 90 forces than to simulate the characteristics of the control sys­
UJ
_«.
tems during a model test. See for instance Sjouke and Lagers
100 r [1-6], Sugiura et al [1-7] and Tamehiro et al [1-8]. For such
simulation studies accurate numerical data on the behaviour of
110 the mean and low frequency wave forces are desirable, so that
0 40 80 120 160 200
meaningful results can be given regarding the systems under inves­
TIME in sec.
tigation. See for instance Van Oortmerssen [1-9] and Arai et al
[1-10]. In order to produce numerical results, however, a theory
must be available on which calculations can be based. In this
Fig. 1-4 Depth record showing effect of suction force on sub­ study such a theory is developed based on potential theory. The
marine under waves. Quartering sea, wave height = final expressions are valid for all six degrees of freedom and
18 ft., vessel speed = 0 knot. Ref. [1-4]- are obtained through direct integration of the fluid pressures
acting on the instantaneous wetted surface of the body. The final
From the foregoing it can be seen that, depending on the expressions are evaluated using an existing computer program based
kind of structure or vessel considered, one or more of the six on three-dimensional linear potential theory. Numerical results
components of the mean and low frequency second order wave forces are compared with analytical results obtained for a simple shaped
in irregular waves can be of importance. In order to be able to body using a different theory. Experimental results for different,
evaluate the influence of such forces on the performance or be­ more practical shapes of vessels and structures are compared with
haviour of a structure the most reliable method available, which results of computations. It is shown that the expressions obtained
can take into account in a relatively straightforward way those for the mean and low frequency second order wave forces can be
factors which are deemed of importance for the behaviour of a used to gain more insight in the mechanism by which waves and
structure interact to produce the forces. It is also shown that II. PAST DEVELOPMENTS CONCERNING THE COMPUTATION OF MEAN AND LOW
the insight gained using the method of direct integration can be FREQUENCY WAVE FORCES
used to enhance the positioning accuracy of dynamically positioned
vessels in irregular waves. This is effected through the use of
i£il:_._i{_.t_r2*lu£_^22
a wave-feed-forward control signal based on the instantaneous re­
lative wave height measured around the vessel. In this section, in which a review is given of developments
in the past concerning theories which may be used to predict the
second order wave forces, theories concerning the prediction of
the added resistance of ships travelling in waves will also be
taken into account, since the physical aspects are the same in
both cases. In fact the added resistance is simply the longitudi­
nal component of the mean second order wave forces for the case
of non-zero forward speed. Indeed, initially emphasis was placed
on obtaining good estimates of the added resistance in waves of
vessels with forward speed. Only in recent years, due to the enor­
mous increase in the number of vessels being moored at sea, have
theories been developed which did not have to take into account
the effect of forward speed which is of great importance for the
added resistance. Most of the work carried out in the past has
been concerned solely with the mean second order wave forces on a
vessel or structure travelling or stationary in regular waves.

Maruo [Il-l] and Gerritsma [II-2] show that on basis of

this information the mean component of the second order wave force
can be determined in irregular waves. As shown by Dalzell [II-3]
the low frequency component of the second order wave forces on
bodies in irregular waves can, strictly, only be determined from
knowledge of the low frequency excitation in regular wave groups
consisting of combinations of two regular waves with different
frequencies. The low frequency wave force will then have the fre­
quency corresponding to the difference frequency of the component
regular waves. As will be seen in this section only in recent
times have attempts been made to determine these components of
the second order forces.

Ii_i.:___Historical_reyiew

The existence of non-zero mean components in the total wave

force acting on a floating vessel was first noted by Suyehiro

8
[II-4] who, from experiments, found that a vessel rolling in reg­
ular beam waves was subjected to a mean sway force. Suyehiro
contributed this force to the capability of the vessel to reflect
part of the incoming wave.
Watanabe [II-5] gave an expression for the mean sway force
in regular waves based on the product of the first order roll mo­
tion and the Froude-Kryloff component of the roll moment, which
indicated that the phenomenon involved was of second order. Re­
sults of Watanabe's calculations accounted for about half of the
mean forces measured by Suyehiro.
Havelock [II-6] gave a similar second order expression for
the mean longitudinal component of the second order wave force or
added resistance on vessels in head seas involving the Froude-
Kryloff parts of the heave force and pitching moment and the heave
and pitch motions. This expression was used to estimate the in­
crease in resistance experienced by a vessel travelling into head
waves. The results obtained using Havelock's expression generally
overestimate the added resistance at pitch resonance and under­
estimate the added resistance in the range of short wave lengths,
where diffraction effects become more important. Watanabe's and
Havelock's expressions for the mean second order wave forces in
regular waves neglected diffraction effects.
Maruo tlI-7] presented expressions for the longitudinal
and transverse components of the mean horizontal second order wave
force on stationary vessels in regular waves. The theory is valid
for two and three-dimensions and is exact to second order within
potential theory. It is based on the application of the laws of
conservation of momentum and energy to the body of fluid surround­
ing the vessel. The final expressions derived are evaluated based
on knowledge of the behaviour of the potential describing the
fluid motions at great distance from the body. Numerical results
given by Maruo are, however, limited and do not give satisfactory
verification of the applicability of the theory since no correla­
tion is given with experimental results.
Kudou [II-8] has given analytical results on the mean hor­
izontal wave force on a floating sphere in regular waves using
Maruo's [II-7] theory and shows reasonable correlation between
computed and measured data.
Newman [II-9] rederived Maruo's three-dimensional expres­
sions for the horizontal force components and extended the theory
by including an expression for the mean yaw moment. The expres­
sions were evaluated using slender body assumptions and results
of computations compared with experimental results given by Spens
and Lalangas [U-lOl. Through lack of sufficient experimental data
no final conclusions could be drawn regarding the validity of the
theory.
Faltinsen and Michelsen [11-11] modified Newman's expres­
sion and evaluated their result by using a computer program based
on three-dimensional potential theory using a distribution of
singularities over the surface of the body. Results of computa­
tions compared with experimental results of the mean horizontal
force on a box shaped barge in regular waves showed good agree­
ment .
Recently Molin [11-12] modified Maruo's expression for the
horizontal force and evaluated it using a numerical fluid finite
elements method of computing the potential describing the fluid
motion. The modification to the original formulation lies in the
change of the surface of integration. Molin used the mean surface
of the vessel while Maruo applied asymptotic expansions valid at
great distance from the vessel. Molin's results compare well with
experimental results on the mean longitudinal and transverse force
and yawing moment on a stationary tanker in head, beam and bow
quartering regular waves.
Kim and Chou [11-13] have made use of Maruo's [II-7] ex­
pression for the two-dimensional case of a vessel in beam seas to
derive the mean sway force on stationary vessels in oblique waves.
Comparisons made by Faltinsen and L^ken [11-14] with results ob­
tained by other methods and from experimental results with the
method of Kim and Chou indicate that the method can show large
10
deviations.
Joosen [11-15] has determined, by application of slender
body theory, the added resistance of ships using Maruo's [II-7]
expression. The final result is similar to that found by Havelock
[II-6]. In Joosen's case the added resistance is independent of
speed.
Lee and Newman [11-16] have given expressions to determine
the mean vertical force and pitching moment acting on deeply sub­
merged slender cylinders. The method is based on momentum consid­
erations. No computed results are given.
Karppinen [11-17] has developed a method to determine the
mean second order wave force and moment on semi-submersible struc­
tures based on three-dimensional potential theory. Karppinen as­
sumes that the structure may be subdivided into slender elements
which do not interact. The total mean forces and moments are found
by summation of the contributions of the elements. The mean force
on each element is determined from momentum considerations in a
manner similar to that given by Lee and Newman [11-16]. Karppinen
gives computed results for a semi-submersible. No comparisons are
made with experimental results. Mean forces on simple elements
are compared with results obtained by others.
Lin and Reed [11-18] have presented a method, based on mo­
mentum consideration and through the use of an asymptotic form of
the Green's function valid at a large distance, for the mean hor­
izontal second order force and yaw moment on ships travelling at
a constant speed in oblique regular waves. No results of computa­
tions are given.
An approximative theory for the added resistance in regular
waves is given by Gerritsma and Beukelman [11-19]. In this method
the mean force is derived by equating the energy radiated by the
oscillating vessel to work done by the incoming waves. The expres­
sion obtained has been applied to the case of ships travelling in
head seas and the correlation between the computed and measured
added resistance is good. Strip theory methods are used to evalu-
i i
ate the final expression. No experimental data are given for the
case of a stationary vessel.
Kaplan and Sargent [11-20] have proposed to use the ap­
proach of Gerritsma and Beukelman to the case of oblique seas. No
comparisons with experimental data are given by these authors.
Ogilvie [II-21] developed expressions based on two-dimen­
sional potential theory for the mean second order vertical and
horizontal wave forces on submerged circular cylinders fixed, free
floating or with forced motions in regular beam waves. The problem
is solved analytically and the results are exact within potential
theory. No assumptions are made regarding the slenderness of the
cylinders. No comparisons are given with experimental results.
Goodman [11-22] has determined, by direct integration of
pressure acting on the hull, the mean vertical force acting on a
submerged cylinder in regular beam and head waves for wave lengths
in the order of the diameter of the cylinder. No comparisons are
given between computed and experimental results.
Salvesen [11-23] has derived expressions for the total mean
and low frequency second order wave force and moment on floating
structures which is three-dimensional and exact to second order
within potential theory. The expressions were derived through in­
tegration of pressure over the hull surface. The final results,
however, make use of the asymptotic behaviour of the velocity po­
tentials at great distance from the body. The theory was applied
to the case of stationary vessels and to vessels with forward
speed in regular waves. In order to finally evaluate the expres­
sions slender body assumptions were applied. Comparisons made by
Faltinsen and Lf6ken [11-14] with other theories show that the
slender body assumptions can scarcely be applied in many practical
cases.
Dalzell and Kim [11-24] have computed the mean and low
frequency components of the second order forces on a vessel using
Salvesen's [11-23] equation for the mean force in regular waves.
Comparisons are given between computed and measured data which
show reasonable qualitative agreement.
Ankudinov [11-25], [11-26] gives expressions for the mean
second order force and moment on stationary ships and the added
resistance of ships travelling in regular waves in deep or shallow
water. The theory is exact and based on integration of pressure
on the body's wetted surface. The expressions of Havelock, Maruo
and Newman are derived as particular cases of Ankudinov's final
expressions. No numerical results are given on the mean force on
stationary vessels. The added resistance of ships travelling in
waves is computed using strip theory methods and compared with
experimental data for deep and shallow water. The results compare
reasonably well.
Based on direct integration of pressure, Boese [11-27] ap­
proximated the added resistance of ships in regular waves from the
relative wave height and the product of heave force and pitch mo­
tion. The final expressions were evaluated by strip theory meth­
ods. Results of computations agree reasonably with experimental
data.
Pinkster [11-28] gave an expression based on direct inte­
gration of pressure for the mean and low frequency second order
horizontal wave force on a vessel in irregular waves. This expres­
sion included the components used by Boese [11-27]. Using strip
theory methods, only the same components could be evaluated. Re­
sults of computations of the mean and low frequency surge motions
of a vessel moored in irregular head seas were compared with ex­
perimental results and showed reasonable agreement.
Pinkster and Van Oortmerssen [11-29] presented results of
computations of the mean longitudinal and transverse force and yaw
moment on a stationary free floating rectangular barge in regular
waves based on the method of direct integration of pressures.
Evaluation of the complete expressions given, which are exact
within potential theory, requires accurate and detailed knowledge
of the flow around the hull. This was determined using a numerical
three-dimensional sink and source technique utilizing Green's
functions. See Boreel [11-30]. The results of computations were
13
compared with experimental results and good correlation was found.
Computed results of the mean vertical force are also shown. No
comparisons are given with experimental data for this component.
Faltinsen and L0ken [11-31] presented a two-dimensional
method based on potential theory to compute the mean and low fre­
quency components of the second order transverse force on cylin­
ders floating in beam seas. The method takes into account the
force contribution arising from the second order non-linear veloc­
ity potential as well as the usual components arising from pro­
ducts of the first order quantities. The expressions obtained are
exact within potential theory and results of computation of the
mean and low frequency transverse force on a number of cylinders
with different forms and breadth to draft ratios are presented by
Faltinsen and LjzSken in [11-32]. No comparisons were given with
experimental results.
Pinkster and Hooft [11-33] and Pinkster [11-34], [11-35]
extended the method of direct integration to include the low fre­
quency components of the second order wave forces on stationary
free floating bodies in regular wave groups. The contribution
arising from the second order, non-linear, potential is included
using an approximation based on the transformation of a first
order wave exciting force. The approximation for this component
is compared with two-dimensional exact results given by Faltinsen
and LjzSken for the case of a floating cylinder in beam waves.
Pinkster [11-34] compared results of computations of the mean
longitudinal wave force in regular head waves on a semi-submers­
ible with experimental results. The comparison indicates that po­
tential effects rather than viscous effects dominate in the second
order force on semi-submersibles.
Pinkster [11-35] computed by the method of direct integra­
tion the low frequency component of the second order longitudinal
force on a semi-submersible in head waves and compared the results
with experimental results obtained from tests in irregular head
waves using cross-bi-spectral analysis techniques as developed by
Dalzell [II-3]. The agreement was reasonable.
14
 Bourianoff and Penumalli [11-36] determine the total hydro-
dynamic force including the first order force and the second order
mean and low frequency forces by means of time domain solution of
the Euler hydrodynamic equation coupled with the rigid body equa­
tion of motion for the ship. The method allows non-linear treat­
ment of ship-wave interaction and arbitrary two-dimensional geom­
etry. Furthermore ship motions are calculated in regular or ir­
regular waves and the effect of arbitrary mooring forces can be
included. Results of computations are compared with experimental
results regarding the low frequency motions of a vessel in irreg­
ular beam waves. The correlation is reasonable but computation
time exceeds real time by a factor of about four.
Pijfers and Brink [11-37] developed expressions by means
of which the mean horizontal wave force on semi-submersible struc­
tures consisting of slender elements could be determined. The
method is based on the use of Morison's equation and the relative
motion concept to determine the wave loads on the structural ele­
ments. Results of computations indicate that the viscous drag
plays an important role in the mean force. In regular waves the
mean force as determined by Pijfers and Brink is not a quadratic
function of the wave height. No comparisons with results of ex­
periments are given. Previously Wahab [11-38] presented a similar
method to that of Pijfers and Brink. The results of computations
were compared with limited data from experiments. However, no
general conclusions could be drawn.
Huse [II-39] has given an expression for the mean horizon­
tal force on semi-submersibles from which a qualitative indication
is drawn regarding the influence of viscous effects. Comparisons
are made with experimental results for two semi-submersibles.
For the computation of the mean force the restriction of long
waves relative to the platform dimensions is imposed.
H i 3^. .Conclusions
Resuming the foregoing it can be seen that the theories,
developed in the past, may be grouped in four main categories:
15
Potential theories which deduce the mean second order forces
based on momentum and energy considerations applied to the body
of fluid surrounding the vessel. The change in momentum (or
moment of momentum) of the fluid is equated to the mean force
(or moment) acting on the vessel. These theories generally
make use of knowledge of the far-field behaviour of the poten­
tials describing the fluid motions. Theories in this category
are due to:
- Maruo [II-7]
- Newman [II-9]
- Faltinsen and Michelsen [11-11]
- Molin [11-12]
- Kim and Chou [11-13]
- Joosen [11-15]
- Lee and Newman [11-16]
- Karppinen [11-17]
- Lin and Reed [11-18]
The theory of Maruo, Newman, Faltinsen and Michelsen and Molin
are three-dimensional and exact to second order within poten­
tial theory. Their basic expressions do not impose restrictions
on the hull form. Other methods in this category make use of
slender body assumption. The theory of Lin and Reed includes
the effect of forward speed.
Potential theories which deduce the mean and in some cases also
the low frequency second order forces and moments through di­
rect integration of the fluid pressure acting on the wetted
part of the hull. In a number of these cases the final expres­
sions are, by application of Gauss's theorem, transformed to
equivalent expressions which have to be evaluated on a ficti­
tious boundary at great distance from the vessel, thus making
use of the asymptotic or far-field behaviour of the potential
describing the flow. Theories in this category are due to:
- Watanabe [II-5]
- Havelock [II-6]
- Ogilvie [11-21]
- Goodman [11-22]
- Salvesen [11-23]
16
- Dalzell and Kim [11-24]
- Ankudinov [11-25], [11-26]
- Boese [11-27]
- Pinkster [11-28], [11-34], [11-35]
- Faltinsen and Ljziken [II-31]
- Bourianoff and Penumalli [11-36]
Of the theories for the mean second order forces, those due to
Ogilvie and Faltinsen and Lj6ken are two-dimensional and exact
to second order. The theories of Salvesen, Ankudinov and Pink­
ster are three-dimensional and exact to second order. The the­
ories which have been used to determine the low frequency part
of the second order forces are those due to Dalzell and Kim,
Faltinsen and Ljiken, Pinkster and Bourianoff and Penumalli. The
theory of Dalzell and Kim makes use of slender body assumption
and is approximative. The theory of Faltinsen and Lj6ken is two-
dimensional and exact to second order in basic formulation and
in the- results obtained. The theory of Pinkster is three-dimen­
sional and exact to second order in basic formulation and, for
the greater part, in the results obtained. The theory due to
Bourianoff appears to be fully non-linear in basic formulation
and in the results obtained. This theory is, of all theories
discussed here, the only one solved in time domain.
In the following chapter the hydrodynamic theory for the
general three-dimensional case of a body floating in arbitrary
wave conditions will be treated. Expressions will be derived for
the mean and low frequency second order wave forces for six de­
grees of freedom based on the method of direct integration of
pressure over the wetted hull.
From the review of work already published in this field it
would appear that similar derivations may have been given by other
authors. This is, however, not the case. With respect to the meth­
od of direct integration of pressure, partial results of the same
nature as given in chapter III have been given by Ogilvie [11-21]
and Boese [11-27]. In neither case has the general hydrodynamic
theory been discussed or have the complete and general expressions
for the mean and low frequency second order forces been derived.

Potential theories which deduce the mean second order forces

by equating the damping energy radiated by the oscillating ves­
sel to work done by the incoming waves. These theories are ap­
proximative and in all cases make use of slender body assump­
tion. Theories in this category are due to:

- Gerritsma and Beukelman [11-19]

- Kaplan and Sargent [11-20]

Approximative theories which make use of Morison's equation and

the relative motion concept. These methods apply typically to
semi-submersible structures which are assumed to consist of
slender elements. These theories are due to:

- Wahab [11-38]
- Pijfers and Brink [11-37]
- Huse [11-39]

17 18
III. HYDRODYNAMIC THEORY
III.1. Introduction
In this section the hydrodynamic theory which forms the
basis for computations of the mean and low frequency second order
wave drift forces on floating or submerged objects will be treated.
The theory is developed based on the assumption that the fluid
surrounding the body is inviscid, irrotational, homogeneous and
incompressible. The fluid motions may then be described by a ve­
locity potential $ from which the velocity field can be derived
by taking the gradient:
V = V.$ (III-l)
with: * = *(X,t) (III-2)
in which X, t are respectively the position vector relative to a
fixed system of rectangular co-ordinate axes and time.
For an arbitrary case the motions of the body and the po­
tential $ are unknown quantities which have to be determined tak­
ing into account certain boundary conditions applicable to the
flow and the equations of motion of the body. In accordance with
classical hydrodynamic theory - see for instance Stoker [III-l] -
it will be assumed that the velocity potential $ of the flow and
all quantities derivable from the flow, such as the fluid veloc­
ity, wave height, pressure, hydrodynamic forces and the motions
of the object, may be expanded in a convergent power series with
respect to a small parameter z, for instance:
- the potential:
* = e $ ( 1 ) + e 2 * ( 2 ) + 0(e 3 ) e«l (III-3)
- the wave elevation:
. = .(0) + _. U) + _V2) + OU 3 ) (III-4)
- the motion of the object:
X = x(0) + _x ( 1 ) + E
2
x(2) + 0(.3) (III-5)
where the affix denotes the static value, indicates first
order variations and (2) the second order variations, etc.
In waves the first order quantities are oscillatory quan­
tities with wave frequencies. In the most general case second
order quantities, besides containing low frequency components,
also contain high frequency components with a frequency in the
order of twice the wave frequencies. For some problems, for in­
stance hull vibrations, the high frequency components of the sec­
ond order wave forces may be of interest. In that case the excit­
ing forces can be obtained by taking the high frequency components
of the second order forces. Force and motion components of this
type are, however, of no consequence for the problem at hand and
will therefore be left out of consideration in this study.
It will be understood hereafter that first order quantities
are oscillatory with wave frequencies, while second order quanti­
ties are restricted to low frequencies with frequencies lower than
the wave frequencies.
In the following quantities are of second order if preceded
2 2
by £ . If, as in many cases, the e or E are discarded this will
be due to the fact that the expression involved will contain only
first or second order quantities. In such instances first order
quantities will be recognizable by the affix and second order
quantities by the affix (2) or by the fact that a component is the
product of first order quantities with affix . For instance,
the pressure component:
H.p|V*(1)|2 (III-6)
is recognized as a second order quantity.
For the derivation of the second order wave forces on an
object in waves it is sufficient that the expansions in a power
series with respect to a small parameter E only be carried out up
in which:
to and including the second order. In order to develop a consis­
p = fluid pressure
tent theory it is assumed that the low frequency motions induced
S = total wetted surface of the body
by the low frequency second order forces are of second order and
dS = a surface element
small in relation to the first order motions. From measurements
N = outward pointing normal vector of dS
of low frequency motions of moored vessels it is known that in the
X = co-ordinates of dS
range of practical wave heights this assumption is in some cases
strictly speaking incorrect. This is due to the large dynamic mag­
The numerical method used to finally evaluate the fluid
nification of the motions resulting from small damping for low
forces and moments will be discussed in chapter IV.
frequency motions. The theory, however, is based on the assumption
of infinitesimal wave height (e-»-0) in which case, in spite of low
Ill 1.2 i_Co-ordinate_sys terns
system damping, the low frequency motions induced by the low fre­
quency second order wave drift forces are always small in relation
Use is made of three systems of co-ordinate axes {see Fig­
to the first order motions.
ure III-l). The first is a right-handed system of G-x -x -x., body
axes with as origin the centre of gravity G and with positive G-x.
In this chapter the hydrodynamic boundary problem for the
axis vertically upwards in the mean position of the oscillating
potential $ will be formulated to first and second order. If the
vessel. The surface of the hull is uniquely defined relative to
potential $ is known the pressure in a point in the fluid may be
this system of axes. A point on the surface has as position the
determined using Bernoulli's equation:
vector x. The orientation of a surface element in this system of
axes is defined by the outward pointing normal vector n.
P = P 0 " Pg* 3 - P*t ~ %p | V* | 2 + C(t) (III-7)

where:
p» = atmospheric pressure
X^ = vertical distance of the point below the mean water surface
C(t) = a function independent of the co-ordinates
t = time
p = mass density of the fluid
g = gravity constant.

The fluid forces acting on the body are determined by the

method of direct integration using the following basic equation
for the forces:

F = -// p.N.dS (III-8)

S
and for the moments:

M = -// p. (X x N) .dS (III-9)

S Fig. III-l Systems of co-ordinates.

_>.
 The second system of co-ordinate axes is a fixed 0-X.-X--X, (2) (2) (2) —(2)
system with axes parallel to the G-x.-x--x, system of axes with with components xl , x,- and x, respectively and X is the
the body in the mean position and origin 0 in the mean free sur­
low frequency second order motion vector of the centre of gravity.
face.
The velocity V is:
— — —.l . 2— f 0 .
The third system of co-ordinate axes is a G-Xj-Xl-Xi system V = X = EX^ ' + £ X 1 ' (111-14)
of axes with origin in the centre of gravity G of the body and
where: x"(1) = V ( 1 > = * ( 1 ) + ^{1) x x (111-15)
axes which are at all times parallel to the axes of the fixed g
0-X1-X 2-X_3 system.
J
and , ¥2) = V ( 2 ) = *(2> + ^ ( 2 ) * x (IH-16)
The angular motions of the body about the body axes are de­
in which the components of the angular velocity vectors a and
noted by the Eulerian angles x., x 5 and x^.
a are x^ , x^ , x^ and x. , xi and x^ respectively.

m^.3^_Motion_and_velocity_gf _aj2oint_gn_the_hu^_of_the_body
The orientation of surface elements of the hull of the body
relative to the body axes G-x.-x^-x-, are denoted by the outward
If the body is carrying out small amplitude motions in six
pointing normal vector n. Relative to the fixed system of co-ordi­
degrees of freedom under the influence of oscillatory first order
nate axes 0-X.-X--X, and the G-X'-X'-X' axes the normal vector of
and low frequency second order wave forces the position vector of
a surface element becomes:
a point on the hull of the body relative to the fixed system of
0-X -X2-X-. axes is:
N = N(0) + EN
(1)
+ _2N(2) (111-17)
(0) (1) 2 (2)
x = x + _x + _ X (111-10)
where it is found that:

where X denotes the mean position vector with:

N(0) = n (111-18)
(0)
X<°> = X + x (III-ll)
N(1) = aCl) x n (111-19)
g
and X denotes the first order oscillatory motion with: -(2) -(2) —
N l z ' = aK s x n (111-20)
S (D . 5 (D + H (D „ ; (III .12)
g miii_myi^_iI.2ti2l_§_^S^_b2yS^^£Y_£20^iti22^
where a is the oscillatory first order angular motion vector
III.4.1. Boundary conditions within the fluid, at the free surface
with components xi , xi and xi respectively and X^ is the
and on the sea floor
oscillatory first order motion vector of the centre of gravity of
the body. Similarly the second order low frequency motion is:
The fluid domain is bounded by the free surface, the sur­
face of the body and the sea floor. Assuming that the fluid is
X<2> = X<2> + a(2) x x (III-13) inviscid, irrotational, homogeneous and incompressible the fluid
motion may be described by means of the velocity potential *:
where —
a (2. is the low frequency second order angular motion vector
 = e4(1) + E 2 $ ( 2 ) (111-21)
<+ *__)-2-<l,--_l> + < l, <*x;x 3 + i ^
The potentials are defined relative to the fixed system of 0-X1~
on X 3 = 0 (111-27)
X„-X axes with:
The boundary condition at the sea floor, which states that to
$ = *(X,t) (111-22) first and second order no fluid particles shall pass through
this boundary or:
where t denotes time and X the position vector of the point under
consideration. V$(1).nb = 0 (111-28)

The potential $ must comply with the following boundary V$ ( 2 ) .n, = 0 (111-29)
conditions: b

where n, is the normal vector of a point on the surface of the

- Everywhere within the fluid domain the equation of continuity sea floor.
must be satisfied or:
III.4.2. Boundary conditions on the body
V2$ = 0 (111-23)
In general the boundary condition on the body states that
In order to satisfy this requirement to first and second order the relative velocity between the fluid and the body in the direc­
tion of the normal to the body be zero. This means that no fluid
it follows that:
passes through the hull. This boundary condition has to be satis­
fied at the instantaneous position of the body surface and is as
V2$(1) = 0 (111-24)
follows:

2 (2)
V $ =0 (111-25)
The boundary conditions at the free surface. The (unknown) free
surface is a surface of constant pressure and the velocity com­
ponent of the fluid normal to the free surface is equal to the
velocity of the surface in the same direction. The latter state­
ment implies that no fluid particles pass through the free sur­
face. The boundary conditions on the moving free surface may be
expressed as boundary conditions, which must be satisfied on the
mean, fixed free surface. According to Stoker [III-l] the bound­
ary condition is satisfied to first order if:
V*.N = V.N (111-30)
Taking into account equations (III-3) and (111-14) through (III-
29) and grouping powers of e results in the first and second order
body boundary conditions.
The boundary condition for the first order potential *
on the body, which states that, to first order, there is no rela­
tive motion between the fluid and the body surface in the direc­
tion of the outward pointing normal vector N, is as follows:

g^i1* + *iiJ = 0 on X-. = 0 (111-26) V* ( 1 ) .n = V ( 1 ) . n (111-31)

A- tt -3

The boundary condition is satisfied to second order if: The boundary condition for the second order potential $ (2)
states that, to second order, the relative velocity in the direc-
tion of the outward pointing normal N be zero or:
and: V** 1) .n = V ( 1 ) . n (111­38)
(2) (1) (1) (1) (2)
b
7_ .n= (V - V* ).N +V .n (111-32)
Equation (111­37) defines the diffraction potential <Jx
which compensates the normal velocity components due to the undis­
Equations (111-31) and (111-32) have to be satisfied at the
turbed incoming waves at the motionless body surface. Equation
instantaneous position of the surface of the body. Assuming that
(111­38) defines the body motion potential *. which must be in­
the motions are small and applying a Taylor expansion similar con­
troduced in order to satisfy the first order boundary condition on
ditions may be posed on the potentials at the mean position of the
the body oscillating in still water. From the solution of the dif­
surface. The first order boundary condition becomes:
fraction potential $!: combined with the potential of the undis­
turbed incoming waves $ (1) the so­called first order wave exciting
V* (1) .n = V U ) . n (111-33) W
(1)
forces are found. The body motion potential $^ is used to deter­
The second order boundary condition becomes: mine the hydrodynamic reaction forces known as added mass and
damping for unit amplitude acceleration and velocity of the body
V*(2).n = -(XC1).V).V*(1).H+ (V(1> - v V ^ . N ^ + motions. From the first order wave exciting forces the added mass
and damping coefficients and the equations of motion of the body,
+ V(2}.n (111-34) the unknown first order motions and hence the total first order
potential $ can be determined; see for instance Van Oortmerssen
where the additional term in equation (111-34) arises from the [111­2 3. Substitution of the first order potential $ of equa­
second order correction to equation (111-31) when applying the tion (111­35) in the non­homogeneous second order free surface
Taylor expansion to the velocity V$ boundary condition of equation (111­27) shows that the second or­
der potential has, in general, the following components:
In equation (111-33) and (111-34) the potentials and their
derivatives have to be taken at the mean position of the body. * (2) = *<2) +
^V +
*iV + *T +
* ( ^ + ♦ih ) +
ww dd bb wd wb db
We may decompose * in the following way: X2) + ,(2> ,(2) (2) (HI­39)
+ + +
dw bw bd 0
. (1 > = _(1> + . j 1 ' + .<1J (111-35) where the first nine components on the right­hand side are poten­
w d b tials which are particular solutions to the following type of
in which $ is the first order potential associated with the un­ boundary condition at the mean free surface, e.g.:
disturbed incoming waves. Substitution of equation (111-35) in
boundary condition (111-3 3) gives the following: g*(2> + *<2> ­ ­2V* U) .V* {1) + *(1)(*(1) + i*Cl) )
ww w w w w
x3 tt t t wx3x3 g w
ttx3
(V$ (1) + V$< 1} + V**1*)-n = V ( 1 ) .K (111-36) (111­40)
w d b
Since the expression is linear * we may decompose this equation (2 )
The last potential *' is a potential which satisfies the homoge­
into two parts: neous boundary condition:

V*i1J.n = - 7 * ( 1 ) .K (111-37) 9*n 2) + $

n2) =
° (111­41)
d w x
3 fct
 (2 .
a direction along the normal n to the surface and secondly the
$x is therefore an "ordinary" potential which satisfies the lin­
earized free surface condition. We will simplify equation (111-39) second order velocity component of the difference between the
by putting: first order velocity V of the body surface and the first order
fluid velocity V* in a direction along the first order normal
.(2) = .(2) + *<2) (111-42) N . From the solution of the second order diffraction potential
W U (2) (2)
in which $ (2) represents the sum of the first nine components on
$£ combined with the undisturbed second order potential $ the
W low frequency second order wave exciting forces are found. Equa­
(2) (2)
tion (111­47) defines the second order body motion potential $/
the right-hand side of equation (111-39). ** may be regarded as
which must be introduced in order to satisfy the boundary condi­
the second order equivalent of the first order undisturbed incom­
tion on the body carrying out low frequency second order motions
ing wave potential $ . We will decompose 3x as follows:
in still water. This potential satisfies the same boundary condi­
*n2' = *i 2 ) +
*H2> f" 1 - 4 3 ' tion as the first order body motion potential *.; . The only dif­
0 a b ference is that the motions are low frequency and of the second
Both these potentials satisfy the linearized free surface condi­ order in magnitude. The same techniques may therefore be employed
tion of equation (111-41). Substitution in equation (111-42) gives in solving $i as used in solving $i . This means that $J; may
be expressed in terms of hydrodynamic reaction forces for unit
.(2» = .(2> + . < 2 ) + *' 2 ) (111-44) amplitude of motion velocity and acceleration of the body, better
w d b known as added mass and damping.
Substitution of equation (111-44) in the second order boundary
condition (111-34) gives: III.4.3. Boundary conditions at infinity

(V$w( 2 ) + V$i
a
2}
+ V** 2) C1} C1)
b )-n = -(X .V).V$ .n +
For the potentials *^ 1 ) , *^ l j and *(![2)# *^ 2 ) a radiation
condition, which states that at a great distance from the body the
+ (V C 1 ) - V » ( 1 ) ) . N ( 1 ) + V(2).K waves associated with these potentials move outwards, must be sat­
isfied. This restriction imposes a uniqueness which would not oth­
(2)
(111-45) erwise be present. Since the components of $ are particular
solutions to the free surface boundary condition (111­27), which
which may be decomposed in: is defined over the complete free surface, a radiation condition
need not be imposed.
v . ' 2 ) X = x v . ( 2 ) - (x ( Xv>.v* ( 1 >}.n +
d w IIIi5i_Pressure_in_a_£gint_within_the_fluid

+ (V C 1 ) - V $ C 1 ) ) . N { 1 ) (111-46) If the velocity potential * is known the fluid pressure

in a point is determined by Bernoulli's equation:
V $ ^ 2 ) .n = V C 2 ) .n (111-47)
P = P 0 " pgX 3 ­ p* t ­ *fp|V*|2 + C(t) (111­48)
Equation (111-46) defines the second order diffraction potential
(2)
Sx which firstly compensates the second order velocity compo- where:
nents of $ (2) and the second order correction to the first order
velocity V* , which results from the first order motion X ' in

■_n
p = atmospheric pressure the potentials are taken at the mean position of the hull which is
X3 = vertical distance below the mean free surface alternately within and outside the actual fluid domain. This ap­
$ = velocity potential pears to be permissible if the potential functions are sufficient­
C(t) = a function independent of the co-ordinates ly "smooth" at the boundaries, see ref. [III­3]. This is assumed
t = time to be satisfied in this case.
p = mass density of the fluid.
IIIi^i_Second_order_waye_force_and_mgment
In Bernoulli's equation p Q and C(t) may be taken equal to zero
without loss of generality, see ref. [III-l]. III.6.1. Second order wave force

Assuming that the point is carrying out first order wave In determining the second order wave force consideration
frequency motions X and low frequency second order motions X must first be given to the choice of the system of axes to which
about a mean position X and applying a Taylor expansion to the will be referred. Since in general we are concerned with the slow
pressure in the mean position the following expression is found: wave drifting force induced motions of bodies in the horizontal
plane we have chosen to determine the wave drifting force along
.p- p(0) + _p(1) + E
2
P
(2)
(IH-49) the axes of the G­X'­X'­X' system of co­ordinates. See Figure
III­l.
where:
The fluid force exerted on the body relative to the G­X'­
- hydrostatic pressure: X'­XI system of axes, which is the system with axes parallel to
the axes of the fixed system 0­X ­X2~X but with origin in the
P ( 0 ) = -pgx<°> (111-50) centre of gravity G of the body, follows from:

- first order pressure: F = ­// p.N.dS (111­53)

p ( 1 ) = -pgx'1' - p.*1'
- second order pressure:
P ( 2 ) = X P | V * ( 1 > | 2 - p . , ! 2 ' - p t x ' X v . ' 1 ' ) - pgX<2>
In the above the derivatives of the potentials have to be taken in
the mean position of the point.
We have assumed that the point is moving within the fluid
domain. The same expression will be used to determine the pressure
on a point on the hull of the body. This means that derivatives of
S
(111-51) where S is the instantaneous wetted surface and N is the instanta­
neous normal vector to the surface element dS relative to the
G­X'­X'­X' system of axes. N is given by equation (111­17) and p
by equation (111­49).
The instantaneous wetted surface S is split into two parts,
(111-52) viz.: a constant part S up to the static waterline on the hull
and an oscillating part s between the static waterline on the hull
and the wave profile along the body. See Figure III­l.
Substitution of the pressure p as given by equation (III­
49) and the normal vector N as given by equation (111­17) gives:

31 ■ _ ­ _
 P = - J . (P <0) + BP (1 ' + E 2 p ( 2 ) )(n + E N (1 > + E 2 N< 2 ')dS +
The integral in this expression corresponds with the first term in
S
0 equation (111­58) which is the total first order fluid force rela­
-/J (p (0) + E p U ) + E 2 p ( 2 ) ) ( K + EN(1> + E 2 N ( 2 ) )dS tive to the body axes G­x.­x 2 ­x 3 . Equation (111­61) indicates that
S a second order force contribution relative to the G­X'­XI­X' sys­
(111-54) tem of axes arises from rotation of the first order fluid force
relative to the body axes. In the same way the gravity force act­
= F(0) + EF(1) + E2F(2) + 0(e3) (111-55) ing on the body relative to the body axes must be accounted for in
the second order force. This force relative to the body axes is:
The hydrostatic force F follows from integration of the hydro­
static pressure p over the mean wetted surface S : ­a(1J x (0,0,­mg) = a ( 1 ) x (0,0,pgV) (111­62)

F^' = pg jjtt vx;

?(0) (0)UJ -
.n.dS = (0,0,pgV) (111-56) Adding this component to equation (111­61) gives:
; J
S,o
The total first order oscillatory fluid force F follows from: a ( 1 ) * {"// p U ) . n . d S + a U )
x (0,0, pgV) } = a C l ) x F ( 1 }
S„

F ( 1 ) =-// ( p U , X + p ( 0 ) .N ( 1 ) )dS (111-57) (in-63:

S
0
= -JJ p ( 1 ) .n.dS + a ( 1 ) x (0,0,pgV) (111-58) where F is the total first order fluid force including the hy­
S drostatic restoring force, the wave exciting force and the hydro­
0
The first part of this expression is the total first order fluid dynamic reaction force. See equation (I11­58). Consequently, ac­
force relative to the body axes G-x -x2~x . The second order force cording to Newton's law, we may put:
is found by integration of all products of pressure p and normal
vector N which give second order force contributions over the con­ FCl) = M.X(1) (111­64)
stant part S n of the wetted surface and by integration of first
order pressures over the oscillating surface s: from which it follows that:

a' 1 ' x F ( 1 ) = a ( 1 ) x („.f«», (IJI_6S)

F ( 2 ) =-/{ ( p U ) . N ( 1 ) + p (2) .n + p (0) .N (2) )dS +
S
0
-// p (1) .n.dS . (111-59) The second part of the first integral in equation (111­59)
s involves straightforward integration of the pressure p (2) as given
Taking into account that: in equation (111­52). The third part of the first integral is a
second order hydrostatic component:
N(1) = a(1) x n (111-60)
­// p ( 0 ) . N < 2 ) . d S = a<2> x ­/J p <0>.n. dS
Since angular displacements are the same for all surface elements ­so S
0
(2)
dS, the first part of the first integral becomes: = a x (0,0,pgV) (111­66)

-// p C l ) . N U ) . d S = a U ) x -// p(D .H.dS (111-61) The second integral in equation (111­59) over the oscilla­
S S ting surface is solved by substituting p from equation (111­51)
0 0

■_ A
and writing the surface element dS as: The derivation is analogous to that followed for the force. The
final expression for the second order wave moment is:
dS = dX3.dil (111­67)
2
M ( 2 ) = - / *_pg<;(1) . (x x n).dl + a ( 1 ) x (l.a (1) ) +
r
Also taking into account that at the waterline: WL

-jl {-.P|v.(1)|2 - P*< 2) - Pii 1 1 1 ^' 1 1 )] .

1J = U)
­P*t P9C (111­68)
. (x x n)-dS - // -pgX* 2) .(x x n).dS
this integral becomes: S
0
(1) (111-74)
c
­ / J (­pgX. + pgc(1))n.dx .dl (111­69)
(1) Equations (111-72) and (111-74) give the total second order forces
X
«■ 3WL acting on a vessel, thus including the wave exciting force and the
which r e s u l t s i n : hydrodynamic and hydrostatic reaction forces. In most cases prime
interest is focussed on the second order wave exciting forces and
- / fcpgC*11 .n.d* (111-70) moments. It will be clear from the aforegoing that the second or­
WL der hydrodynamic reaction forces are contained in the contribu-
in which c is the relative wave elevation defined by: (2)
tions due to the total second order potential * . The hydrostat­
ic reaction forces are contained in the last parts of equations
41' = ^(1) -*_£_. (III 71)
- (111-72) and (111-74). Taking into account equation (111-44) the
second order wave exciting force and moment become:
The final expression for the total second order force thus
becomes: F (2) = - / W 1 , r2 . n . d * + a U ) * <M.Sg(1>> +
WL
2
F ( 2 ) ­ ­ J fcpgt^.n.di + a U ) x (M.*(1>)+ -// { - % P | ? » ( 1 ) | 2 -p(*< 2 > + f «d > ) +
g
WL SQ t t
­// { ­ ^ P | V * ( 1 ) | 2 ­ P*t2) " P(x(l).v*1!l))}K.ds + - p(X ( 1 ) .V*^ 1 } )}n.dS
S
0 (111-75)
­// ­pgX*2).n.dS + a ( 2 ) x (o,0,pgV)
S 2
0 M ( 2 ) = - / J_pgCU) . (x x n).dl + a ( 1 ) x (i.a (1) ) +
r
(111­72) WL

-// {-%p|V« (1) | 2 " P(*J,

W
2)
+ *i
d
2)
) +
III.6.2. Second order wave moment SQ t t

- p ( X ( 1 ) ,V#^ Z ) )}. (x x n).dS

The moment about the axes of the G­Xj­X'­X' system of co­ (111-76)
ordinates follows from:
The hydrodynamic reaction forces due to motions induced by second
M = "J/ P­ (X1 * N) .dS (111­73) order forces may be expressed in terms of added mass and damping
S
forces as has been shown in the aforegoing.

35
 IV. EVALUATION OF THE SECOND ORDER WAVE EXCITING FORCES

IV^l ^Introduction

In the previous section general expressions for the second

order wave exciting forces and moments have been obtained based on
the method of direct integration of pressure acting on the wetted
surface of a body. The expressions obtained are, however, not in
a form which is easily used for practical applications. In this
section it will be shown that the second order forces may be ex­
pressed more conveniently in terms of time independent quadratic
w £ 0 H» rt ft rt rr ^­. TJ rt p­ Pi TJ Hi TJ Pi rt & 0 g cn P IH
fD
O
P­
H
p,
CL
P
O
3"
P­
fD
3
P­
0
P
3
H
H
M
ro
3*
fD
cn
r+
fD
3
H
Pi
O
c
0
ft ro
H
3
fD
^3< Hi p­
3
ro
0
cn IH
IH transfer functions by means of which it is possible to express the
tn rt 3 P­ rt fD H­
0 H fD ro P­
d H X
p­
0
rt
3 P 3 P 0 1­
second order wave exciting forces in the frequency domain in terms
3 H
3"
1
to rt
cn 3
p­
D. 3
ft
cn •5 3
p­
0
rt
rt
p­
3 cn l­J
Pi tr 0 a P 0 ro ^Q O • 0 P Pi 1'
ro C
3 Hi P
tn
P1 Hi P
cn
^] >< o ti­
3 0
tu H
p­ P» <3
fD
o P­ 0 P
pj H
1 of force spectra or in the time domain as time histories of second
^ in
■ — '
0 0 P C 0 3 0
cn 3 ti­ fD Hi pi rt H 3
H ro o rt • 0 3 ro 3 M H H 3 cn 10
order forces.
o_ 3 a
p­
ff » pj tQ Hi Pi o P­ p­ fD 0 ro cn 0 Pi 0 13
fD 0 ro 3 " 3 C 0 t­3 3" H o rt P 0 X 3 P Hi
ro rt IO
H £ tn P 3­ rt 0 3 cn pi 3J 3 3 f. o p­ H ft 3* IH
rt rt tn P­ rt cn ro a p­ rt rt P­
3 3 0 Pi
p­
fD 3" H
lQ
» ft H­
3 el­ 3 IC
icn
TJ C C c ip P fD Pi ro 3* pi s' ro cn
TJ
0
rt
rt tJ l_. fD (U ri­
0 CT)
ro C 0 tn Hi rt
3J
rt p­
ro 0 c ro H IP­ The components of the transfer functions for the second
3 tr H < H­ p­
3 fD Hi H fD
tn
H 0 3* 3 3 o o cn
0 10

3
fD P
rt
fD
CL
ro 3 en
Hi
0 H Pi rt
tr
p* ro ro ^P ro fD cn tr P ro 13
icn order forces which depend on first order quantities can be evalu­
ft
0 H ^H­~ K: 0 ro T3 *. ro M X Hi Pi CD M cn o
= Hi ^ tr 3 ro 0 H ro 0cn cn H , T ) 0 0 fD
ro rpt­
P­ P­ 0 H p_ rt rt ­s fD 0 0 H H tr 0 g ated using an existing method of computation based on three-dimen­
P 3 £ ro r3t" H rt cn fD ro tP o p­ O M i­i ro O *z 3 €, O
H
13
CD
<
id
c
cn 1 3
ro 0H at o. ro
0
H
P
H
3
rt
3
Pi
c
Cu
0
3
3
Ip
0
3
c
rt
n
CD
cn
cn
ro
cn rt
3 P­
Pi
P rt
3
sional linear potential theory of which a brief account will be
rt
O H fD EU Pi Hi P­ 3
& P­ cn p ­ 3" •< 3 P­
0
0 P
TJ
ft
P­
tr <!
fD
•—' 3Q" CD tu Pi1
I­
rt rt
P­
H cn
fD
0 0 —. ro H ft given. The contributions due to second order potential effects
3 0 0 • H o 0 P­ 0 3 0 3 P cn tr HP­i
Pi
rt
H
rt
P­
0
rr
0
3
c
3
Di
H
g
0 a
ro
<e- tr c
rt
p­
tr o
ft
n
pi rt
p
3
cn &
Cb
ro ro
p.
ro H
t,
P will be determined by an approximation using results of computa­
p­ CD
0 OJ tr 3" <P p­ er « M­­i ro ro r t p­ ro 0 *-. ro 0H i tn tn cn
tr (D 3
rt
— . DJ ­< P­ fD
3
Hi 0 cn p ­ 0 H tr H pi Z I­
T . r t
1
tions based on the same method. Comparisons of this approximation
c H H cn Hi c ■ Pi ^ 0 3 ft ro H H P tn
rt
p­
tu H­
P
H
H
•< fD
X g
ro
0
H
tu
3
Pi tr 3
0
K
3
tn
cn &
p­
3*
ro o
M
1
3 0 < P­
P ro ro r t H
0 3
0
with some exact results will be given. Only the low frequency sec­
tn H 1 P­ p­ p­ Hi ­J
0
p­
O ro c
tn
0
rt
Pi ro tr
p­ Hi
tr
rt Ul
cn cn Pi £ ond order forces are treated here. The same procedure applies to
3
tn
• 4*.
cn
0 cn
rt
PJ
p­
H 0 3 tu rt cn
fD
cn ro P­ ro 3
3 3 3 ­< 0 tu 0 H H P «— P X 3 H
the low frequency second order moments.
■ • — •
H p­ Hi P­ H­ rt ft
TJ
I­1
&
H­
cn O g 3
rt
3*
tn
£
p­
P 0 P 3 o Hi
3
Cu CD
3
rt
3
rt
H 3 n ro p­ i­1
O o
3
3 P 3
P*
Pi p­ 0
rt
3

rt
C_
P­
ifl ro 0 cn • n P1
ft
P­
ro 3
rt
0 P
K 0
Hi
3" ro TJ 3 p)
3
3 Pi pi P. Pi P­ H rt
P 3 3 0 P 0 PJ pi *d ip 0 Pi ro 3 « c_
tr
fl)
TJ
TJ
fD rt
Eu
3 P­
H
rt tn
c
rt
ro
P­
rt
p­
Hi
p­
3 tr Hi
tr
ro
3 0
P.
0
3
.—
M
. P
3
rt
ro
iQ 0 ro
cn
*£
p­ IYi2i_The_guadratic_transfer_function
cn P­ 3 P­ Hi I­­ H rt
H t­h P­ 0
rt 3
H
Hi rt P­
Pi
P­
ro H0 tn
TJ Pi
H
TJ
p­
H hh T3
C
tn
O O 3
tn
3
fu
0 O
ti­
ro *< 3
g
p­
0 P
S 3
p­
X H P­ 3 n 3 ro 3 3* M | H H
<D
0. 3
p­ H
0 rt
rt
3
0
K
Tl
H
0
(D
P
H
cn C
H
iQ <
tu o
0
£
ro
3
=
P
Hi
-*1
CTs
iQ
•—
3
P
n
ro cn
art 3
IV.2.1. General
t H­ ft rt 3 rt rt rt tn
P
rt
3 ro PJ LQ M 0
P­
c •— • p­
­ 3 TJ
P­ fD ro 3 tr ■e* P­ 3^ c 3 3 fD o
ft
0
P­ 3 3 H O 0
a t ­o­ <ro Pi ro P» TJ ft
3d
P Pi H P­ 0 H­ 3 ti­
0
3
^Q C
3
fD
tu g
P­
TJ
c
3 ­ 3
c rt
ro 0cn ro
0
3 P
I­* P­
cn rt
ft 3 3 P
3 03
p­
ro
3 In this study the total quadratic transfer function is
P. rt ft ft
tn *" 3 ro P_ Pi ro 3­ rt 3" P­ 0 ro
0 ro 3 rt ro P 0 P­ rt P" 3 P C ro tn Hi tr P tn 1
P­ split up in contributions arising from the following components of
Hi o 0 £ 3 H 3 3 0 c P rt P­ < H ro "< *» P

rt
rt
p­
£
3
ro
rt
0
pi
rt
3
*< rt ro r3t
rt
H­
cn
p­
tn 3 fD tr
ro
Hi
P­
tn
ro
rt
31 P
i .

rt
H
equation (111-75):
d CD ft
rt 3
3 0 cn P­ O 3* iQ 0 Pi g tr p­
s
h ro tr Di 3
ro 3 cn ro • cn 0 ro c ro 3 ua ro 3P ro tn
rt
3 cn CD
ft ro 3
cn tn fD DJ 3 pi Tl ro tr CD ro
ro H­
0
0 tn
H
3
TJ
0
Pi
P­
o
rt
rt
H­
o
0
rt
0
ro
X
ro
3
^< 3 £
P 0
rt
3J
*<
a
3
O
ro
H
rt
0
0
n I : First order relative wave elevation
rt 3 c 1 1 3 0 3 i i < H P K ro i rt *<
D. 3 I 1 rt 0 P
H ro ro
1 I H M.2
-^pg / C* .n.d£ (IV-1)
WL
II : Pressure drop due to first order velocity
The first order relative wave elevation at a point £ on the water-
line of the body may be written as follows:
-j! -^p|V* ( 1 ) | 2 .K.dS (IV-2)
S
0
tlr1] (t,£) = Z ^ll) .r,ll) ' U ) .cos{w t + E. + E ^ (£)}
III : Pressure due to product of gradient of first order 1 = 1 ^ 1 x
1
pressure and first order motion
(IV-8)
in which:
-// -p(X ( 1 ) . W 1 }
) .n.dS (IV-3)
S
c,. - amplitude of i-th regular wave component
0
IV : Contribution due to products of first order angular e_. = random phase uniformly distributed over 0 - 2TT
motions and inertia forces OJ . = frequency of i-th component

C'(1) '(£) = transfer function of the amplitude of the first order

a ( 1 ) x (M.X ( 1 ) ) (IV-4)
g relative wave elevation at point £ in the waterline

V : Contribution due to second order potentials e (£} = phase angle of the relative wave elevation at point £
related to the undisturbed wave crest passing the cen­
( 2) + 2) tre of gravity.
-// -P *w *i )-n.dS (IV-5)
w d
SQ t t
The procedure to obtain the quadratic transfer functions of the Substitution of (IV-8) in equation (IV-6) leads to:
forces dependent on first order quantities (I, II, III and IV)
will be illustrated by taking the low frequency part of the longi­ F.l/;(t) = Z Z C- -C- i ; .P._.cos{U.
x
-to.)t
J
+ (E, - EJ. ) } +
1 J J
i=l j=l
tudinal component of the force contribution due to the relative
N N
wave elevation: (1) (1)
+ Z Z C, -Ci . Q ^ - s i n U w . -ui.)t + (e. -e.)} +
J J J x J
i-l j=l
2
F[2) « F<2)(t) = - J hpgK^ (t,l) .n (D.&l
WL + high frequency terms
(IV-9)
(IV-6)
in which:
where P. . and Q. . are the in-phase and out-of-phase components of
C (t,£) = time dependent relative wave elevation in a point I
the time independent transfer function with:
along the waterline
n (Sl) = direction cosine of a length element dJL in longitudi­
P.. = P(to,,co.) = / ^pgc; U).tl ( £ ) - C O S { E (£) +
1=1 J r
nal direction. WL i j i

- E r _ ( £ } } n i U ) -d£
In irregular long-crested waves the elevation, to first order, of 3
(IV-10)
the incoming undisturbed waves - referred to the mean position of
the centre of gravity of the floating body - may be written as:
Q*+
iJ
= Q(w. ,to.) = - / ^pgCi U)-Cl U J . s i n U ^ (£) +
^ J WL i r
j i
N
(1) (1)
C Mt) = Z c \1} .cos(u.t + E. ) (IV-7) - E r> (A)}n 1 (£).d£
x
i=l ii 11
(IV-11)

39
/in
Taking the low frequency part of the square of the wave elevation these components of the quadratic transfer functions.
given by equation (IV-7) results in:

2 N N IV.2.3. Contribution of the second order potential

U )
C (t)0= Z Z ^C n U ) .c5 1 ) .cos{(D. -u.)t + (E. - E . ) }
1 x ] J J
i=lj=l
IY^.2^3il^_General
IV-12)
When a body is floating in a regular wave group consisting
Comparison with equation (IV-9) shows that P,. and Q.. are trans­ of two regular waves with frequencies OJ . and u. part of the second
fer functions which give that part of the wave drifting force order wave exciting forces are due to second order velocity poten-
which is in-phase and out-of-phase respectively with the low fre­ tials $ { 2 ) and $ ( '2 )as was already indicated in chapter III. The
W d
quency part of the square of the incident wave elevation. (2)
second order potential <i> has to be determined taking into ac­
count the following boundary conditions at the mean free surface:
It will be clear that similar developments can be made for
other contributions to the wave drifting forces which depend only g*w(2 > + * (2)
wtt = - 2 _ V »
(1)
_ W t1 ) + * l tl ) - ( 4 l Xl 3 X 3 + ^ g* t tttX
x )3
on first order quantities. The total in-phase and out-of-phase '3
transfer functions are found by simple summation of the contribu­ (IV-13)
tions from the five components. The wave drifting forces may thus
be presented as transfer functions which, as can be seen from the in which $ represents the total first order potential contain­
aforegoing, are a function of two frequencies. In general the qua­ ing the contributions from the undisturbed incoming waves, dif­
dratic transfer functions will also be functions of the direction fracted waves and waves due to first order body motions. The cor­
of the waves. responding boundary condition for the second order diffraction
(2) is as follows:
potential * - m
IV.2.2. Evaluation of the components dependent on first order
quantities g$i 2 ) +
$i 2 > = 0 (IV-14)
d d
x3 tt
Evaluation of these components of the quadratic transfer
functions of the low frequency wave drifting forces requires de­ The second order diffraction potential $,(2) also has to satisfy
tailed knowledge of the first order vessel motions and fluid mo­ the following boundary condition at the surface of the body in the
tions. As can be seen from equations (IV-10) and (IV-ll) knowledge mean position:
of the first order amplitude and phase transfer functions as a
function of the wave frequency are sufficient to evaluate these V . ' X n = -V. ( 2 ) .n - { ( X ( X v ) . V _ ( 1 , } n + <V(1> - « ( 1 ) ) N ( 1 >
d w
components of the quadratic transfer functions.
(IV-I5)
A numerical method by means of which such detailed infor­
mation may be obtained (using a distribution of sources over the Besides these conditions other boundary conditions discussed in
mean wetted surface of the body) has been developed by Boreel chapter III have to be satisfied. However, for the discussion in
[IV-1] and Van Oortmerssen [IV-2], A brief description of this this section only the above conditions are relevant.
method is given in Appendix A. The computer program DIFFRAC based
on the theory given in this appendix has been used to evaluate

41 42
 As can be seen from equation (IV­14) the second order dif­
(2)
fraction potential <2>, satisfies the free surface boundary condi­
tion which is applicable to normal first order potentials. This
means that in principle the same numerical techniques as indicated
in A ppendix A can be used to determine this potential for arbi­
trary body shapes.
The boundary conditions at the mean position of the body
given in equation (IV­15) contains first order contributions which
are known on the basis of the first order solution and an unknown
second order contribution due to the "undisturbed" second order
(2 .
potential $ ' which has first to be determined taking into ac­
r
w of sources is necessary. The general procedure using this numeri­
count the non­homogeneous free surface boundary condition (IV­13).
(2)
This boundary condition prescribes the behaviour of * over the
complete mean free surface. No elementary solutions for this po­
tential are known for the general three­dimensional case. This is
due to the complexity of the right­hand side of the free surface
boundary condition equation (IV­13).
In the three­dimensional case of a vessel floating in a
wave field consisting of two regular waves with frequencies ui. and
ox approaching from the same direction the total first order po­
tential $ (1) will contain contributions of the two long­crested
incoming regular waves and a complex pattern of cylindrical out­
going waves due to diffraction and body motion effects. The right­
hand side of the non­homogeneous free surface condition of equa­
tion (IV­13) contains products of potentials associated with these
long­crested incoming and outgoing cylindrical waves. As shown in
chapte (2)
chapter III, the potential * can be split into the following
parts:
*:2) ■ *™ * *__> + <v * <v ♦ <v ♦
in which:
(2)
* = second order potential associated with the undis­
c
ww
turbed incoming first order wave potential
43
x(2) , .'
(2)' = second order potentials associated with the outgoing
diffraction waves and waves due to body motions
$ (2)' $,(2, ) = second
second order
order potentials
potentials associated
associated with
with in
interactions
wd bd ^
between incoming waves and outgoing waves.
(2)
Of these potentials an analytical solution is only known for 0 .
The other potentials may be written as complicated two-dimensional
Fourier integrals. Evaluation of such integrals, however, presents
a considerable computational problem. Therefore we prefer a de­
scription in terms of a source distribution. In order to solve the
(2 )
second order potential $ a considerable increase in the number
* w
cal technique will be described here briefly. Firstly, a source
distribution is defined over the mean wetted surface of the vessel
and over the mean free surface in the vicinity of the vessel. In
the numerical method described in Appendix A the Green's function
chosen for the elementary sources which are distributed over the
mean surface of the vessel satisfies the homogeneous free surface
condition equation (IV-14). The source distribution over the free
surface cannot make use of the same formulation for the elementary
source since in that case the non-homogeneous free surface condi­
tion equation (IV-13) cannot be satisfied. Instead we may choose
the elementary source function of the type, which corresponds
to the first two terms of equation (A-7) given in Appendix A.
This type of source satisfies only the equation of continuity
equation (II1-23) and the kinematic condition equation (I11-2 9)
at the sea floor. The extent of the distribution over the free
surface will be the result of a compromise between the magnitude
of the error in the results due to the truncation some distance
away from the vessel and the increase in computation times. Second­
ly, the first order solutions obtained for two regular waves with
frequency oi, and o)2 are used to compute the values of the second
order free surface conditions of equation (IV-13) at the centre
of the sources distributed over the free surface. Since the source
distribution over the hull of the vessel satisfies the homogeneous
free surface condition equation (IV-14) these do not contribute
to the source strengths in the free surface. The source strength
in the free surface may therefore be solved without consideration
of the source distribution over the hull surface. The non-homoge-
44
neous free surface condition equation (IV-13) leads to a set of
continuity (111­25) has been given by Bowers [IV­3]. We now con­
simultaneous equations which are linear in the unknown source
sider a regular wave group travelling in the positive X ­direction
strength of the free surface sources which is solved by standard
(2 ) consisting of two regular waves with frequencies _j. and to, with
techniques. This solves the unknown potential $ . Having solved
W ox > a)­. The first order velocity potential associated with these
(2) (2)
*J> the right-hand side of the boundary condition for <_x at the waves is:
mean position of the hull of the vessel equation (IV-15) can be
evaluated and Sx {2) solved by the method described in Appendix A. 2 c ( 1 ) g cosh k (X th)
(2) (2) <J> 1 J
= ­ Z — 1 ,\ sin(k.X_ ­cu. t + E . )
Having solved both potentials $ and $-. the contribution of w ._ to. cosh k.h ' i l l —1'
J c
w a
these potentials to the second order forces can be evaluated by (IV­17)
means of equation (IV-5).
The low frequency component of the second order velocity potential
The procedure given here indicates that for the second or­
associated with these waves is as follows:
der potentials results can be obtained using numerical approxima­
tion techniques. It will be appreciated, however, that the compu­ ,9( 2 ) _~ - Ih 2
-r <. n rC.
u > AA .. c o s h { ( k
i-k„)(x3+ h ) }

tational effort to obtain results for this contribution will be ww . .L . *i 3 13 r—n——;—n

cosh(k. ­k.)h *

considerable since the above procedure must be repeated for all

relevant combinations of frequencies ux and os_ . Advantage can, how- • sin{ (k i ­ k . )X X ­ ((ix ­ a).) t + (£._ ­ e ■ )}
(2)
ever, be gained of the fact that the basic solution of <Jx need
only be obtained once for every series of combinations of ax and IV­18
ax which yield the same difference frequency m - ux. For the pre­
in which A . . is a coefficient depending on to. , to . and the w a t e r
sent work a different, more simplified approach has been followed ij i 3
in order to approximate the contribution of the second order poten­ depth h:

tial to the second order forces. „ B. . + C. .

A. . = ^ g 2 2J 3J
2
(OJ. ­ t o . ) ­ (k ­ k . J g tanh(k. ­ k )h
IYi2I_3_L2_L_A£proximatign_f or_the_contribution_of _the_second_order i J i J i J
potential (IV­19)
t in which:
The approximation is based on the assumption that the major i 2 , 2
part of the low frequency second order force due to the second or­ k. k.
der potential is the wave exciting force component due to the con- B. ■.^ = oi. cosh
i k.h a
tu. cosh k .h
(2)
tribution $ of the undisturbed incoming waves to the second or-
ww
der potential, thus assuming that the first order diffraction and (IV­2 0)
body motion potentials 0, and 0,b are small relative to the un­
ci (1)
disturbed wave potential $ . This means that in the right-hand 2k.k. (to. ­cu.) (1 ­t­tanh k.h tanh k.h)
Q = 3 i J _ 3
side of the free surface boundary condition of equation (IV-13) ij ^uij
only terms involving the first order velocity potential * ' of
the undisturbed incoming waves remain. The second order potential (IV­21)
which satisfies this boundary condition and the boundary condition
at the sea floor of equation (111-29) as well as the equation of

45 46
 The low frequency component of this second order potential
represents a long wave which is induced by the presence of the reg­
ular wave group. The phase of this long wave relative to the regu­
lar wave group is such that it has a trough where the wave group
attains its maximum wave height. This is shown in Figure IV-1.
(2)
DUE TO $
ENVELOPE OF GROUP
Fig. IV-1 Wave due to second order potential of a wave group.
The potential associated with such a wave does not satisfy the
boundary condition on the body which for the simplified case is
assumed to be equivalent to the normal first order boundary condi­
tion. This means that in the right-hand side of equation (IV-15)
first order contributions are neglected.
As shown in chapter III the second order diffraction poten-
(2)
tial $ satisfies the equation of continuity, the boundary con­
dition at the sea floor, the radiation condition and the homoge­
neous free surface condition:
g$
x3 +
*tt =
° (IV-22)
This last condition gives rise to the well known dispersion rela­
tionship:
co = kg tanh kh
The incoming waves due to the low frequency second order potential
have a wave number equal to k. - k. and wave frequency equal to
UJ. -co.. These waves do not obey the dispersion equation (IV-23).
If the incoming waves have a frequency of to. -co. then the
diffracted waves have the same frequency but the wave number will
be according to the relationship:
(LO. - C O . ) 2 = kg tanh kh (IV-2 4)
In order to simplify the situation we allow the diffracted
waves to have the same wave number k. - k. as the incoming waves.
This means that differences will occur in the diffracted waves
further away from the vessel. Close to the body the situation will
be similar since the boundary condition at the body still has to
be satisfied. The reason for this alteration in wave number of the
diffracted waves will be apparent from the following.
We have reduced the problem to the situation where we have
to determine the wave exciting force on the body due to a wave
which has a velocity potential as given by equation (IV-18) while
we allow the diffracted waves to have the same wave number and
frequency as the incoming waves. This is solved by considering the
ordinary first order wave exciting force F on the body in a
regular wave with wave number k equal to k, - k. in an ordinary
gravity field with g as constant of gravity. For such a case the
associated wave frequency LO will be in accordance with the disper­
sion relationship of equation (IV-23). The frequency of this wave
can be made equal to the frequency co. - co - of the second order
waves by selecting a different value for the constant of gravity:
(toi -co. )
g =
ij ( k i - k . ) tanh(k ± -k.)h (IV-25)
Since the wave exciting force is proportional to the constant of
gravity the initial force F with wave with frequency co, which
follows from equation (IV-23), becomes a second order force with
(IV-2 3) frequency tu. - co • by simply applying the factor:
n =
ij ^ <IV"26>
47 48
to the initial force. This does not complete the transformation,
IYi.2i3^3^_Comp_arison_between_the_exact
however, since, besides satisfying the requirement that wave num­
mation
ber and wave frequency be equal, the amplitudes of the potentials
must be equal. After the alteration of the constant of gravity the
It can be shown that this method of approximation gives
transformed potential of the first order regular wave is:
exact results in two simple cases and gives a reasonable approxi­
C C l ) g.. coshUk. ­k.)(X +h)} mation for a third, more practical, case.
* = ­ T^ ^­r T"n—^—re sin{(k. ­ k .) X, +
(oi. -m.) cosh(k. ­k.)h i 3 1 The first case concerns the second order pressure due to
- (u^ ­OJ. )t + {e± ­ e_.) } the second order potential in undisturbed irregular waves in a
point X_ = -a below the still water l^vel. The second order pres­
(IV­2 7)
sure is:

The amplitude of the second order potential is given in equation

pC2) = - P ^ 2 ) (IV-33)
(IV­18). Equality of the amplitudes means that:

r(D0 For the low frequency component given in equation (IV-18) the am­
a 9 ±
\ = cCl)CU)A (IV­28) plitude of the pressure is:
C O s h { (k (
(2) (I) (1) i"V "a + h ) }
This means that the first order wave amplitude must be selected Pi3 = P C ^ C ^ A ^ K - ^ ) coa£(k3 )h

so that:
(IV-34)
C(D ,C(D (D*n("i-V (IV.29)
a J. 1 M • ■
J
^13 For the first order potential the pressure follows from:

The first order force F is determined for a value of unity for

P U ) = -P*^ 11 (IV-35)
r, . Since forces are proportional to the wave amplitude equation
(IV­29) gives a second correction factor which has to be applied
The amplitude of the pressure using a first order potential compo­
toD the
t force F in order to give the required second order force
(2) nent of the type given in equation (IV-17), unit wave amplitude
F
z, and wave number k. - k. is:
a 1 3
(2) ^ MA
^ C^­ux) , . cosh{ (k. -k . ) (-a +h) }
Fr:! = n .— 2 ­2J ­L­ F^ i; (IV­30)
ij
pK ' = pg 7-7^ J—: rr (IV-36)
M
*a ^ coshfk. - k . ) h
which taking into account equation (IV­22) gives:
Using the coefficient f. . given in equation (IV-32) gives the fol­
PJ 2 ) = f ± j .F ( 1 ) (IV­31) lowing approximation for the second order pressure amplitude:

f9 i M, ,-,, c o s h U k . - k . ) (-a + h) }
where: Pi.' - P c l 1 , ^ 1 , A l j ( » 1 - » j ) coshdc^-k^h

f±j = — ^ ~ L
(IV­32) (IV-37)

49 sn
which equals the exact value given in equation (IV-34). The reason
for this is that other contributions to the exact value which are
neglected in the approximation (those due to diffraction and body
motions) are in this case zero. F / \ WAVES

The second case concerns the horizontal low frequency wave

drifting force, due to the second order potential, acting on a
vertical wall in deep water. It can be shown that the approxima­
FALTINSEN
tion is also equal to the exact result in this case. The reason
for this is that the first order incoming waves and the first or­
der outgoing waves are identical (total reflection) and the low 0.59 0.72 0.84 0.95 1 . 12 co. /d/g

frequency component of the total second order potential consists

of a contribution associated with the undisturbed incoming waves 0.59 0 0.02 0.01 0.02 0.13
and a contribution due to the outgoing diffraction waves. Since 0.72 0 0.03 0.01 0.10
the approximation gives the exact value for the second order poten­ (2)
0.84 0 0.03 0.04
tial associated with the incoming waves it also gives the exact pgLc^j 1 '
value for the second order potential associated with the outgoing 0.95 0 0.06
waves and hence the approximation is also the exact value. 1 . 12 0

The third example concerns the two-dimensional case of a LO./d/g

free floating cylinder in beam waves as presented by Faltinsen and
L0ken [lV-4]. These authors solved the second order problem and
gave numerical results on the contribution of the total second or­
APPROXIMATION
der potential to the low frequency second order sway force in reg­
ular wave groups in deep water. The method of approximation pre­
0.59 0.72 0.84 0.95 1.12 co. /d/g
sented here was applied to the same case using results given by
Vugts [IV-5] on the first order sway force in regular beam waves.
0.59 0 0.03 0 . 11 0.20 0.33
The coefficient f.. of equation (IV-32) becomes for deep water:
0.72 0 0.03 0.11 0.26
^(UJ. -to.) , . , .2)
1J 0 0.04
f. . = - -± - 3- C c- ' (IV-38) 0.84 0.18
ij g i 3 pgi^cj^
0.95 0 0.10

The results are presented in the form of the amplitude of the low 1.12 0
frequency second order forces due to the second order potential
for a range of combinations of co. and LO . which are the frequencies to ./d/g
of two waves making up a regular wave group. Faltinsen's results
are compared with the approximation in Table IV-1.
Table IV-1 Low frequency drifting forces on a cylinder in beam
seas due to the second order potential.

51 S?
 Comparison of the results shows that near the line OJ . = cu. 2 2
( 2 ) ( 1 ) n
the approximation is good but for larger differences between co. F (t) = Z Z x c! P cos{(u). - u ) t + (E, - E . ) } +
J
i=l j =l -1 3 J

and OJ . the approximation is considerably higher than Faltinsen' s

values. Further study will be required to determine the reason for 2 2
+ £ I c{1)c!1)Qii.sin{(toi -to,)t + (Ei-e.)}
the large differences which occur at higher frequencies. At the i=l j=l J J J J
present, however, it may be tentatively concluded that the method
{ 1 ) 2 ( 1 ) 2
of approximation gives the right order of magnitude to the low fre­ = cr P + Cr P +
i 11 2 '*22
quency forces due to the second order potential for difference fre­
quencies which are not too large. For the cylinder in beam seas
+ C
large differences between Faltinsen's results and the approximation l1)c_|a) CP
12 + P
21 }
- c o s { (to
i "a32)t + {E
1 ~ E
2 ) J +

occurred for values of the non-dimensional difference frequency

greater than about 0.1. + t^1*^ 1 1 CQ
12 ~ Q 2 l ' ' s i n {
Cw1 - c o 2 ) t + (£j -E2)}

For floating bodies the approximation will give best results
when the contribution to the second order potential of the first
order diffraction waves and waves due to body motions are negligi­
ble. This requirement is satisfied more by vessels such as semi-
submersibles than by ordinary ship or barge shapes. It will be
found, however, that when first order diffraction and body motion
effects increase the influence of the first order contributions to
the total second order forces will become dominant, so that the
increase in the error of the component due to the second order po­
tentials still remains small relative to the total force.
IV.2.4. Symmetry of the quadratic transfer functions
The first order wave elevation in a regular wave group con­
sisting of two regular waves with frequency ux and OJ . is as fol­
lows :
2 floating objects in regular or irregular waves this is often ex­
C ( 1 ) (t) = Z c { 1 } .costoi.t + £ i )
i=l
= cj^-cosdu.t +E.) + C 2 1 J .cos(oj2t + E 2 )
(IV-39)
The second order force associated with such a wave train has the
following form:
(IV-40)
From equation (IV-40) it is seen that the second order force con­
tains two constant components. Each of these components represents
the constant force which would be found if the wave train consisted
of a single regular wave with frequency co or oj? respectively.
This shows that, although the force is a non-linear phenomenon,
the constant or mean second order force in a wave train consisting
of a superposition of regular waves is the sum of the mean forces
found for each of the component waves. The quadratic transfer
function:
P =
ll P(to1/^2) (IV-41)
gives the mean second order force in regular waves with frequency
to,. In literature dealing with the mean second order forces on
pressed as a function dependent on one frequency co. . The above
equations show that the transfer function for the mean or constant
part is, however, only a specific case of the general quadratic
transfer function P (LO. /to2) for the force in regular wave groups.
Besides the constant parts the second order force contains
low frequency parts with a frequency corresponding to the differ­
ence frequency co, - OJ of the component regular waves. It is seen
that the amplitudes of the in-phase and out-of-phase parts depend
on the sum of the in-phase quadratic transfer functions P - and metrical about the diagonal for which co, is equal to LO? while the
P2 and the difference of the out-of-phase functions Q _ and Q 2 l * out-of-phase component Q(to ,LO?) is asymmetrical about the diagonal.
From equation (IV-40) it is seen that the transfer functions do The second order force given in equation (IV-40) may also be writ­
not appear in isolation but rather in pairs. In general, the in- ten as follows:
phase and out-of-phase components of the quadratic transfer func­
tions as determined from equations (IV-1) through (IV-4) for com­ F ( 2 ) (t) = r { 1 ) 2 P + r(1)2 P +
binations of UJ and cu2 will be so that, for instance:

+ 2c| 1 } C _ j 1 5 .T 12 .cos{ (co1 -co2)t + ( E X - E 2 ) + £ l2 >

P ((_)._,(i)2) f P(to2,to1) (IV-42)

(IV-49)
However, since the force as given in equation (IV-40) depends on
the sum or difference of the components of the quadratic transfer in which T is the amplitude of the quadratic transfer function:
functions these may b e so reformulated that the following symmetry
relations are valid: T
12 =
T((J 1 ,LO 2 ) = yP2(co1,oo2) + Q 2 (LO 1 ,UJ 2 ) . . . . (IV-50)

P(to ,to2) = P(u)2,to1) (IV-43) It also follows that:

Q(to1,u)2) = - Q ( w 2 , u 1 ) (IV-44) T(to1,U)2) = T(to2,to ) (IV-51)

The approximation for the force due to the second order potential £,« is a phase angle defined by:
(2)
$ as given in equation (IV-31) is only defined for co. > to.. In
Qfui-^tO
order to conform with the definition given in equations (IV-43) tan E I 2 = tan E ( V I O 2 > = - p^J^y (IV-52)
and (IV-44) for the quadratic transfer function the in-phase and
(2)
out-of-phase parts of the force component due to $ become: e
12 9 i v e s the phase angle of the low frequency part of the second
(2)
order force relative to the low frequency part of the square of
P.. = ^P. co. > to. (IV-45)
13 13 1 3 the wave elevation of equation (IV-39), which is as follows:

P. . = P. . (IV-46) 2 2 2
U )
31 13 C (t) = Z Z ^.ci1,C-1).cos{(co. -io.)t + (E, - E . ) }
3 x J x J
i=l j=l
Q =
ij ^Qij) u
i > u
j (IV-47)
2 2
= HK[1) + hK{2l) + c{ 1 ) C 2 ( 1 ) .cos{(u) 1 -io2)t +
Q.. = -Q. . (IV-48)
ji *ij
+ (e-^ ~ £ 2 ) }
In these equations P.. and Q.. represent the in-phase and out-of-
phase components of the second order force as determined by means
(IV-53)
of equation (IV-31). This transformation was applied to the exam­
ple concerning the free floating cylinder in beam waves. The in-
phase component PtuiwUx) of the quadratic transfer function of the
total second order force takes the form of a matrix which is sym-
__v__3__ Fr§gyeD£¥__l25}§*i-_:_-_:2E£§§S2^
where:
forces_in_irregular_waves

T2(u)+y,u)) = P2(to+y,uj) + Q (u)+y,to) (IV-60)

We consider the case of a body floating in irregular waves
which are characterized by a spectral density or energy density
and:
S (co) where: y = frequency of the low frequency second order
2 force
(I)
S (oj).doj = ^ (IV-54) P(to+y,co), Q(to+y,to) = in-phase and out-of-phase components of the
quadratic transfer function
The wave elevation in a point is a normally distributed process T(to+u,uj) = amplitude of the quadratic transfer function'
with zero mean and a probability density given by: S„(y) ~ spectral density of the force.
J.

Besides knowledge of the mean and spectral density of the

P ( -<1>, = _ _ _ . e 2m
° (IV-55) second order forces, knowledge of the distribution function of the
/2"rrm0 force is of interest. In general, however, the distribution func­
tion of a second order force in irregular waves of the type as
where: given by equation (IV-9) cannot be given. An indication can, how­
oo ever, be given of the type of probability distribution function
mn = { S (co).dto = area of the spectrum . . . . (IV-56) involved by inspection of a quantity which is closely related to
U C
0
the second order force. This is the low frequency part of the
The mean second order force is found by putting OJ. = co. in equation square of the wave elevation as given in equation (IV-12). It can
(IV-9): be shown (see ref. [IV-6]) that the low frequency part of the
square of a normally distributed signal is of the exponential type
N
(2) (l) 2 N
(l) 2 with a probability density function:
mean 1=1 i=l

(IV-57) now
(
. 1.2 1 e
m
0 (IV 61
Passing from a discrete to a continuous formulation taking into
P =low ' = - • " >
account equation (IV-54) gives:
In this equation nt- is the area of the wave spectrum given by equa­
tion (IV-56). The distribution function as given by equation (IV-
[(2)
2)
(t) = 2 / S (to) .P(to,Lo) .da. (IV-58)
L 61) is shown in Figure IV-2.
mean

The spectral density of the low frequency components of the force

in equation (IV-9) is found to be as follows:

S (y) = 8 / S (to+y).S (co).T (to+u,u.) . du.

* 0 t. Q

(IV-59)

= .ft
 (2)(¥ _ i ;~r (iuiti-iu2t2)
2TT -■» - «
(2 )
. G (LO1,OJ2) .dto. ­dux

m0 (IV­63)
in which:
G (2) (co ,OJ ) = complex quadratic transfer function
,r< 1 > 2 . = P(io ,(_)2) + iQ(u)1,uJ2) (IV­64)
low and:
P(co ,UJ2), Q(co ,0).) = in­phase and out­of­phase components of the
quadratic response function.

From equation (IV­62) it is seen that if the quadratic impuls re­

sponse function g (2) (t ,t ) is known the time record of the low
frequency second order forces can be computed for arbitrary wave
Fig. IV-2 Probability density of low frequency part of the square elevation records. The applicability of this technique has been
of the wave elevation. demonstrated extensively and convincingly by Dalzell [lV­7] using
quadratic transfer functions for the second order forces obtained
IVi4^_Time_domain_representatign_of_the_mean_and_low_frequency from tests in irregular waves using cross­bi­spectral analysis
second_order_forces techniques.

According to Dalzell [IV-7] the low frequency second order IV.5. Conclusions
forces can be computed given the quadratic transfer function and
the time record of the wave elevation using the following relation­ In this chapter it is shown that the mean and low frequency
ship: second order forces and moments may be expressed in the frequency
domain in terms of quadratic transfer functions which are depen­
F ( 2 ) (t> = r r g'^it^i.^'it-tji. dent on two frequencies. Physically, the quadratic transfer func­
— oo —oo
tions contain information on the mean and amplitude of the low fre­
. £ ( 1 ) (t -t 2 ) ,dt1.dt2 quency second order forces on the body floating in regular wave
groups consisting of two regular waves. It has been indicated how
(IV-62) these quadratic transfer functions are evaluated and an approxima­
where: tion for the contribution due to second order potential effects
gv(2) (t ,t„) = quadratic impuls response function has been discussed. Dalzell [lV­7] has demonstrated that on the
basis of the quadratic transfer functions time records of the sec­
( 1 )
. ( t ) = time dependent wave elevation
ond order forces can be generated. In this study therefore atten­
t l ,t2 = time shifts. tion will be focussed on the frequency domain results only. If such
results can be accurately predicted by computations then time do­
The quadratic impuls response function g (2) (t.,t-) is derived from main results, which are often required for simulations, can also
the following expression: be generated with good accuracy.
V. COMPARISON BETWEEN RESULTS OF COMPUTATIONS AND ANALYTICAL
RESULTS ON THE MEAN WAVE DRIFT FORCE IN REGULAR WAVES

Vili_Introduction

In chapter III and chapter IV the theory and method of com­

putation of the mean and low frequency second order wave drift
forces on floating objects through direct integration of pressure
has been treated. In this section results of computation of the
mean wave drift forces in regular waves will be compared with an­
alytical results obtained by Kudou [V-l] who made use of an ex­
pression given by Maruo [V-2], This expression is based on energy
and momentum considerations. The results concern the three-dimen­ 3 cn P- p- P P 3 iq o o £
to
p rt cn 0 rt 3 TS ft ip P
3 rt 3
P rt tn rt TS 4 rt c Pi
sional case of a free floating hemisphere in infinitely deep water, 3 3 C 0 3 c Cf t P p- 0 H TJ H 3 3 P 0ro p - 0 H 0 CD 3 p P P­
a ro H 3 H g H
< rt ro 3 O cn tHp CD P. CD H P < H O 3 cn CD ti­ ti­ X
cn rt TJ 0 ro ro ro p- CD 0
< cn 3 ro pp ­i 3 TJ rt ro ro
In order to make a complete comparison results of first order wave H ro cn C
c H
a Pi 3 o ua H 3 3 ro P 3 P CD CD 3 H P H S H H
>
P rt Ip P H 3 p- ro p- P H ro 3 cn o cn P - 3 CD n P­ P
loads, added mass and damping and motions are also compared. Pi Hi 3 ro 3 H 3 P- p-
< • 3 X p- cn P TJ CD 0 ro tQ 3 p O ti­ pi pi
P- P- P pi tn ro Hi 3 r+ ro ip P- cn 0
< H 3 3 o 3 ft 3 ft ro CD
c H
c tp
< .—. P ft H 3 H g ro P­ Pi o rt M P­ H TJ ro 3
cn c ro 3 H H 3
*. P 3 £ TS U3 ft r—i 3 ro 0 3 CD 0 ft ro
H ro <1 CD Hi P- 3 P- CD P g 0 p- P C
<1 O P. ^P CD 3
a 3 rt

Y._.2-_.-Commutations
P ro tr p
CO ro 3 >U
Hi
ro
3
0
0
H
-P
C
cn
Hi
3
P
<
CD TJ
cn
p- <
ro
H
CD
Pi
CD UJ
H
Pi 0
c
P
0
3 3
ft
0 p­
ro
TJ C
0n
0
rt 3
hrj
a tn ■—' rf rt
* — ■
0 ro H H 0 H ro 0 ro H P ro 3 P H g T
a TJ cn P-
• P-
Hi P- CD P- ■
P 3
< • CD H O H Pi TJ P Di 3 cn ro TJ
H CD rt N cn CP ro P p-
rt <D 0 0 ro 0 ti­ 3 CD i cn C
ft CD iQ H H 0 0 P- Hi
c p. cn C Z 3 Hi H TJ H 3 ro CD p C Di rt rt
V.2.1. General 3 P- ft 3 3 cn <I c *-i cn P- 0 TJ 3
o cn H cn TJ ti­ P p­ i­{ P
ro ip
< P- rt ft 3 ro 3 ro < K p rt rt Hi ro ro cn H, P­ rt
P- ro O
^ 3 P <p a. 0 cn P tn CD 3 p H 3 *j ro o ,— * H o € H Hi 0 P­
cn < 3 P P- CD H p- ft cn
tn
CD 3 CD ro P­ 3 H H
< 0 P H ro ft 0

For the computations the mean wetted surface of the body TJ ro

3 3 tr
H tp
C O
<
CD
P
3
P-
<i ft
3 <
P
3 P
H
Pi
Hi
tQ rt
p­
P
ro
<I 3 rt
CD
ci
ro
p
H
H
ro
P­ 3
cn
ro f t •< H a 3
0
3 P CD H rt K • P­
C
H tn —
i ■ H
Di 0
3
Di 0 H Pi
*. P H TJ ■P 3
is approximated by 206 facets as shown in Figure V-l. p- 3
c 3 rt
CD to ■ rt rt
H H 3 H H CD H
—' 3 0 ro 0
ro 0
* P- rt Hi P-
<l P Pi 3 ro ro P- 3 tn tn tn H H 3 H 0 P
• c Hi
<1 H ft 3 H- 3 ro 0 ro rt
n ft H PJ ro Hi Hi 3
P Di ft P- -j 3 D-
a P tr i £ P
<i cn p­ 3 3 CD ro cn ft rt 0 3
0 >—■ tr Hi *j *■ O P 3 ro
Hi
C H 0 0 CD H <P H
c 3 3 H P 0
er Hi o 0 P- H 3 P 0 H to H CD ft
c c < 0 C P

TOTAL NUMBER OF WATERLINE TOTAL NUMBER OF

P
to 0
0
n rt
rt
p-
H
0
ip
C
£
3 c
pi
1
Ul
13
P-
TS
0
0
H
3
rt <
P
3
TJ
Di
ro P
P
H
O
0
C
ip
H
P s
p
rt
tn
3 3 CD
P
3
C tr
LINE ELEMENTS: 36 FACETS ON H U L L : 2 0 6 ro 3 o 3 CD p- p- ■ 3 cn CD P- H 3 t, 3 3 ■ tr 3 ro
0 H
p-
C 4 H
< 0 O
ro CD 3 CD H 3 ip rt
c rt Pi P rt ro ro 3 3 CD
0
Hi
ft
3
tn
rt
tn P
H
CD tp ►3
3 O
rt
p-
rt
3 ^cn
ro CD S —H . •s
p •<
to
H
P­
tc
p
o >­. o
CD
S P­ p
H- 0 ro
<1 ft ft ro 0 0 ro P
P
<1
Pi
tr <1 <
p. 0
3 c P O H
ft cn
P
ft
ip
p- CO 3 3 ro 3 3 €
<
ro UJ N ro
H
P­ <
p­ ro p
H
CD Hi Hi 0 Pi ro
< rct •£. cn Hi 3 tn < P H
X
3
ro w
P
H <
CD
ip
p-• ro ro P- Hi 0 H CD H
P
H CD rt
ro
t. p­ < ro H P­
ro ro
\ 3 3 0 H p- Hi ro H ro c 3 ^­* ft ip
3 Hi 3
TJ c
p
<P
P- P <
ro
H
3 ro 0 CO o tn
X
ro a ro
0 0
3 t Q
0
3 Hi cn CD H H P
pi

3 3 P- 3 ro 3P 3 rt p- ft rt O P cn p­ p­ P­ H ro ro ft
H
ft <
CD CD 3 0 < rt rt 3 0
0 0
M
ip
H 3
0
0
P- 3
Pi 3
P- 0
TJ
O 3 ro 0 3 P 0 p­ H Hi <
ro
^< H
C_ P­ tp
J3
C
C
H c
ft
Di
rt ft ro 3 cn
C
ft t. rt Hi cn p- Hi 3 P ft <
ro CD rt CD CD
C
. 0
3 CD P< cn 0 p- H 3
H
CD ro o cn &
tp
P­
0
>< 3 J2 CD
CD ft
ip 3 3 3
• H)
ft
• 53 ro C
^ TJ 0 Pi CD
a
CD
P rt tn tp ►b o CD 3
* P 0

0 H
0
0
P
rt
P-
lp
0
3
o
3 TJ
H p-
3
3
rt
3
CD H
3
H
3
CD
Hi
0
Hi
CD
C
H c
P
3
rt Hi
P
H
ft
3
H
ro H »■< Uj
0
n
i
Hi 3 P 0 p- C CD p_ 3 cn H p CD H H P rt H 0 ro CD 3 H p­
cn CD ft ■ p-
/ CD 3 3 0 H 3 P C
H C
C
3 p-
TJ 0
Hi
H
0
P­ C
Hi
P
3
CD H
ft
3
Hi
/ t. P TJ 3 ro ft 0 01 3 ro tr H 0 3

./
P H
CD
H O
3
cn cn H CD ip
ro
n
Ip
0
Hi
Di
P-
0
3
H3
3
P
cn
«
P
cn c c
rt 0
P
H
3 P­
H
PJ
ro
Hi
3
P­
P­
3
_/ < *< <i Pi
5 P
ro cn ft CD
,^ UD
0 CD P
P
ro rr
cn
rt
H
CD ro
rt <D .->.
C
p­
3H
p­
0 0
3
0 cn
rt
H
ro ro
Hi H
H
cn P­
rt
c P- 3 H Hi H 3
3 P 3 C ro TJ p p­
0
cn <1 3 TJ H
d
cn o O ro
3
c
H
ft
0 rt
tn <1 • ft 3
<P
Pi 3 ro H p 3 3 cn 03 cn ft n H
5 tn
P
H
H
ro Di tr 0
• i-n 0
3
p
ti­
0
H c
H P
P
cn T
H i-3 3
CD
0
cn 3 3 ro 3
P <P C
cn H H cn co — r t 3CD ro PCDi rt 0 cn CD *< T
f^ 0
-—• 3 H
ro
Fig. V-l Distribution of facets on the sphere and the distribu­
ro
H
p-
3
H
ro
Hi CD
O
P-
N P P
3
CD
CD
3
a ro ro ro i p H i ro
cn w ro ro P i
0
ft
Di
H
cn Hi ro
cn rt
Pi
CD
i ft 0
3
0 H
<
ro
P p- 0 C rt OJ P P ft 3 ro
tion of line elements on the waterline. *" 3
ro 1
3
1
CD
< i M ro 3
ro
0
1
H
CD
3
CD
CD TJ
CD

61
 25 500*
1.00
• COMPUTED ANALYTICAL
2.0 250 • COMPUTED

0.75

.(1) 1.5 a
'1a . n/pv
„(1>
pvKt; a 0.50
1.0

b
ii/pvu>
0.5 0.25

O 0.5 1.0 1.5 2.0 2.5 0

ka 0 0.5 1.0 1.5 2.0 2.5
ka
Fig. V-2 First order wave exciting force in surge.
Fig. V-4 Added mass and damping coefficient for surge.

2.5 500
1.00

2.0 250

0.75
.(1) 1 5 U-»-*— -f
3a a
pgAc'J' 33/pV
0.50 /
1.0

0.5 0.25

b
33/p7W
0.5 1.0 1.5 2.0 2.5
ka 0 0.5 1.0 1.5 2.0 2.5
ka

Fig. V-3 First order wave exciting force in heave.

Fig. V-5 Added mass and damping coefficient for heave.

63 f,A
 2.5 500

• COMPUTED oI—•~~• °—^S- 1

2.0 250
• COMPUTED

—•—<•—-< ► _ *
-0.25 \

1.5 mean
1
(1) (1)'
«1a Pga£
(1) -0.50
11
1.0

\ (•
-0.75
0.5

-1.00
0 0.5 1.0 1.5 2.0 2.5
0 0.5 1.0 1.5 2.0 2.5
ka ka

F i g . V-6 First order surge motion t r a n s f e r function Fig. V-8 Mean second order horizontal drift force.

2.5 500
1.0
n
" * ^ ^ _ , _ - &r
2.0 250"
0 -• • !► »-^*J.J-S-
nr
m
mean
(1)2. \ TOTAL
1.5- Pga£ \
-1.0 \
*3a
„0) . nl
1.0

-2.0 X Z
A •

0.5

-3.0
0 0.5 1.0 1.5 2.0 2.5
ka
O 0.5 1.0 15 2.0 2.5
ka
Fig. V-9 Components of the computed mean second order horizontal
Fig. V-7 First order heave motion transfer function. drift force.

65
 3.0
force at high frequencies is of the same order of magnitude as the
peak value and at increasing frequency tends to some limit.
AN
2.0 \
i
i The value of the mean horizontal drift force for wave fre­
// \
y
y
y
quencies tending to infinity is easily found by taking into con­
1 \
1.0 ,. r \ sideration that for very high wave frequencies the wave action,
due to the very small wave length, is confined to a thin layer
V
\­v­—
'mean
/ >
Pga C a i^ m near the waterline of the stationary object. In this case the form
r '.­"
■ ■ » •—<
>* •• k •
*•
'i,m of the hull may be replaced by a vertical wall which totally re­
N i — — —
~-A TOTAL
** ­ I
flects the short waves. The circumference of the waterline may be
y
> n
­1.0
\ y
considered as short sections of straight vertical walls by which
^
\
\ /
/ the incoming wave is reflected as is shown in Figure V-ll.
\ J
\ /
­2.0
0.5 1.0 1.5 2.0 2.5
ka

Fig. V­10 Components of the computed mean second order vertical

force. REFLECTED
WAVE
WAVE CREST OF
INCOMING WAVE
V.2.2. Motions and mean horizontal drift force

From these results it is seen that the computed results of

both first and second order quantities agree well with analytical
results. When viewing the general characteristics of the mean hor­
izontal drift force given in Figure V­8 it is seen that for low
wave frequencies the force is zero. In this condition the sphere
has a first order heave amplitude equal to the wave amplitude as Fig. V-ll The mean horizontal drift force in short waves.
seen in Figure V­7. The sphere is following the wave motion com­
pletely without creating noticeable disturbance. As the wave fre­ The contribution of such a section of the circumference of the
quency increases the heave motion increases and at the same time waterline to the mean horizontal drift force is due to components
the mean horizontal drift force increases. The maximum heave mo­ I and II of equations (IV-1) and (IV-2) and is found to be as
tion occurs at a slightly lower wave frequency than the maximum follows:
horizontal drift force. In this range of frequencies the effects
due to diffraction and body motions on the wave drift forces are dF. = -i.pgC(1) . (sin2 6)-dX9 (V-l)
increasing. At higher frequencies the body motions decrease con­ mean
tinuously to become zero at frequencies tending to infinity. In
The total mean force is found to be:
the range of these frequencies the effects of body motions on the
drift force decrease rapidly and in the limit only effects due to = (1)2 +a
(sin. 2B),dx (V-2)
'i -^P9£ a • -f :
diffraction remain. As can be seen from Figure V­8 the mean drift
mean -a

67
This equation has been given previously by a number of authors.
The integrand in this case always represents a pressure increase
See for instance Maruo [V-2]. It is applicable to floating objects
acting inwardly at the waterline. The sign of this contribution
of arbitrary shape for determining the high frequency limit of the
in Figure V-ll is in the direction of propagation of the waves in­
horizontal wave drifting forces. Taking into account that at high
dicating that, as may be expected, the relative wave elevation on
frequencies there are no waves behind the body the following re­
the incoming wave side is larger than on the shadow side of the
sults are found for the sphere:
sphere. In general, therefore, if it is seen that an object float­
2
2 ing in waves exhibits a large difference in relative wave height
P = -f pga^(li
iJ
(V-3) on both sides of the object, it may be expected that this contri­
mean
bution will be large.
which means that the high frequency limit of the non-dimensional
2
force given in Figure V-8 equals - ^ . _r22t£±hution_II
The general equation for this contribution is as follows:
V.2.3. Components of the mean horizontal drift force
F =
II ~M -*P|V* (1) |2.n.dS (V-5)
The computed components I through IV of the mean horizontal mean S_ mean
drift force are given in Figure V-9. These components cannot be
compared with analytical results since the analytical results The integrand in this case always represents a pressure decrease
given by Kudou were determined using a different theory and are acting outwardly on the mean wetted surface of the hull. In gen­
given only as a total force. Inspection of Figure V-9 reveals that eral, the fluid velocity V* tends to be largest on the incoming
the total force consists of components which differ considerably wave side. This results in a mean force component which contrary
both in sign and magnitude. Contribution I, due to the relative to intuition, is directed into the waves.
wave elevation, is about twice as large and of the same sign as
the total force. Contribution II, due to the square of the veloc­ £2Gi_£ikution_iii
ity, is different in sign and of about the same magnitude as the The general equation in this case is:
total. Contribution III is smaller and different in sign to the
total force and also different in character since it tends to zero F = -// -p(X U) .V*^ 1) ).n.dS (V-6)
at high frequencies. Contribution IV, due to products of angular mean SQ mean
motions and accelerations, remains zero for all frequencies. At
high wave frequencies only contribution I and II remain. This is Since this is a mixed product of first order motion and pressure
in agreement with the previous discussion on the high frequency gradient it is not possible, in general, to predict the sign of
limit of the total force. It is of interest to discuss some as­ this quantity. The sign depends on the phase angles of both quan­
pects of the various components. tities. In the case of the sphere the mean force due to this com­
ponent was also directed into the waves. This force component is
__:2l_li£ihy:ti2_}__[ dependent on the motions of the body and on pressure gradients.
The general equation for this contribution is as follows: At very high wave frequencies the motions and hence this component
tend to zero, while at very low frequencies it is the pressure gra­
fi.2 -
*I = - / i.pgCr - n - dJl (V-4)
dients which tend to zero and consequently the force also.
mean WL mean

f,Q
 Y.i_______:2._2i]_.-_.i22s-

From the agreement obtained between numerical and analyti­

cal results it may be concluded that the potential accuracy of
numerical methods is sufficient for the prediction of both first
order oscillatory quantities and the mean second order drift
forces.

The agreement between results obtained using the method of

direct integration of pressure and the results given by Kudou
0 £ P TJ < TJ rt Hi Pi p­ P­ 3 H Pi TJ ro < P­ O W Hi rf 0 CD P P > ►3 i n based on Maruo's theory demonstrates the equivalence of these for­
0 P 0 H ro CD 3 0 C rt in CD *< C 0 3 •t o 3 P 3 H ro 0 H 3 P cn 3 10
3 ft
CD
rt CD H
rt
P
\K
CD H o
(1­
^< N
3
rt
• CD 3
CD P-
< cn 0 ro 3 3 P
H
Pi n ro I13t t mulations with respect to the total force.
TJ p­
cn n •£ * 0 C D K ^ Pi ft
ro 3
o H Q cn p­ < ro P CD > rt 3 3 H cn 3 c H ■»
3 H P Ip IH
3
CD
H
P­
3 c
H
o
p
P­
3
ro
H p­
0
Hi
0
rt
H
0
Pi
C
Ui 0 rt
UJ
• <
CD pi tn
ro
3
rt
O
P­
3 P
CD
3
CD
H
^< ro
3
IP-
13
3
ft
3
CD
P­ ro H
(n rt
rt
P­
cn
3
cn « ro H­
tn
ft
3
Di P-
tp
P3
3
ts a
■
ro
TJ
p­
3
O
P­ N
C
rt
rt o P 3
CD H
ud ro IC
H Ift Inspection of the various contributions to the mean second
C ro ro ft

P­ 0
tp
H
Hi 3 0 <
CD
0
rt
< >­3
3
rt CD ro C
H
ro 3
P CD
3
O
CD
ro CD
H
P­ 3
P­ rt
P­
0
< P1 IP-
P 10 order force reveals that the total force consists of contributions
ft 0 ro P ro 0 P cn 0 cn 3
tn 3 0 P H H H p­ H CD H Hi rt ro n a 0 3 ip 3 3 CD CD 13
3
ro H i PP i­ 0 < ft 0 rt rt cn P 0 0 < cn P­ rt 3 p­
« ro p ro 1 which differ both in sign and magnitude. From the results found
ua ro CD 3 P­ 3 P­ O 3 0 P <i ■3■6 CD rt 3 P ti­
cn • 3 3 U3 I H
C K
ro
3
tr 3P­ 3ro rt­ rHt
0
0
rt
O
P P
0
P
S ro ft ft
3 H
H
ft
0
3 P­
CD 3 ro Pt,
3 H rt II P for the mean horizontal force on the sphere it is seen that the
3 ft C a Hi
ro D i ro (A 3ro p ­
tt p­ H 3 H TJ TJ 3 rt CD o ti­ H­ cn cn Pi <3 0 3 3 rt
H *< p i
P
a
P
0
3
H P. H 0
3
H CD 3
P H p
ro O
P
ft
3 P
p­
tp
p cn ro H
O ft
p­ PI p-
0
general characteristics of the total force are also present in
rt­ H rt
Pi
ti­ H
^< TJ
*< CD
0
O ft CD H
& H CD 3
0
3 rt CD 3
cn
H 3
w cn PJ Hi
contribution I due to the relative wave elevation (see Figure V-9)
ro P­ O ro C H & 3 c P H 0 < TJ 0 H ro 0 "
Di 3 3 rt 3 P *S CD Q rt n cn P P 3 P a TJ H 0 S CD O 0 Hi
• CD P Di O 0 3 s: tn € rt ro rt ft tp H H 3 0 H. ^P N kP
Q 0 3 X O
H
although contribution I is greater than the total by about a fac­
tn rt Di p cn 3 PJ P­ c p­ P Di
c CD c 3 P TJ
rt rt N P­ H C *: P 0 ■£ r+ <! K CD Hi tn d ft
ro H ro »5 cn 0 ^, tor 2.
H 3 0 CD 0 P. H P CD p­ 3 CD p- rt ro n 3 3 0 3 0 ro 3CD ft

c P­ H 3 0
Pi
CO ro H
Hi
rt CD rt 0 rt p­ rt n 3 s 3
3 3 N 0 tn t P. rt P 3 s: 3 Hi Hi P cn H p­ ro o 3 X|: P-
ro . 3 Hi tp tn 0 3 « P 0 3 0 K > ro 3 Hi ft tp ^1 tn
H i­i 0 P p- rt tn rt
<
CD P 0
Pi
tP H
h tp ft
< tt
n ip Hi Hi ft h­
Q In the next section, in which experimental data on the mean
P *j P t, rt P 3 ti­ CD 3 0.
M
0 rt ­ ft Di
­—
H ^ • CD 0 H Pi 3 3 H ro ro cn ro CD ro ft
h 3 < p­ 3 ro ^* 0
rt ro H & * p­ p­
ro P h ro cn *£ 0 P­ CD 3 tn ro TJ 3 3 horizontal drift forces on stationary vessels are compared with
p­
0
H >
rt
ft
0 H
en P ro H H 3 H H 0
CD P­ ro 0
3 0 ro tn H 0 ro ro
P
^0
< 3 CD cn ro 0 K rt N cn 3 computed results, attention will also be paid to the relative im­
P P Q Hi CD r t tP 3 p 3 < 3 P Hi uP 0 0 P­ CD TJ Pi 3 3
H ri­ 3 rt *> 0 rt c o 3 ro rPt TJ 3 c 0 0 H 0 fi 3 tn ro
H P­
ip
3
CD
H Hi
H
P
O
3
CD
p
H
ft
3 rt
0
3
ft
3
p
3
3
TI
3
TJ
0 3
tn
0 CD
H 0
3
ft
portance of the various contributions.
^ ft
£ p­ p­
< c
3
3 3 P S ro
P >P
rt CD 0
3
0 3 0
3
ro
3
CD
H
ro ft
P­
0
3
0
3 p
P
rt
CD 3
p-
0
ro Hi
P
<
CD
cn n
0 0 tn
w, ro
p­ ii- rt
Hi
H CD 3 P Ul
S < c ft Hi
ro ro
€ P o CD CD c 3 Hi N 3 P­ cn p­ p­ P­ 3 3 CD P P 0 3 •■

3 P <! rf
H
Hi 3
O
TS
6 P 0 0 tn • n
P
H ro ri­ ft ^P tn H tp
ro ti­
ro Hi
t, 0 f t 3 3 Pi cn cn C ro H P C
P ro rf CD H p. P 3 3 Pi ft K P H H rt • p­ P CD 3 H
3 H Hi 3 *Q ro ro H ro CD P H 3 en H 3 P Hi tp P
H H P C vP cn P i 3 3 H Pi £ PJ 0 H cn 0 3 H C H
< p­
ro rt ro c tn rt Hi 0 p­ P ft H H 3 3 o P­ Pi CD H
ro 3 £ 3 ro 3 » P­ Di Hi 3 *< 3 P­
a CD 0 • ro KQ P 3
H ro c < o 3 ro >S PCi H, ■< 0 CD tn P- Ht
ro rt cn P c If 0
ft
• ro CD p­ n H to H tn cn ft H cn > Ul
ro rt
P­ 3 H ro p­ 0 3 CD ft P O N H­ tp rt P 3 3 P­
0 « 0 0 cn ro £ CD f t n CD h ro 0 Hi iQ tO CD o p 0 0 0
P p­ P­ o tn 0 n •• c H P O C 3 p­ 0 p­ rt 3
3
H 3 ro p­ r t rt tt 0 H CD 3 0 cn H P CD 3 ro H ro p­ U)
O tn rt 3
P­ CD 3
P 3 0 Pi H ft
• ro O 3 P­ ^ ft H CD tn 0
Hi
ro r t CD
ro rPt r3t ro ro 3
CD
< ►6 rt
ro rt
P­
3 P
H
CD
H
ro
H
cn • 3 P
3
0
M
0 n H ro 3 3 P
d cn
P 3 cn 3 < TJ H ro P i­i p­ rt cn H 3 H H Pi
—
O rt ro ro
P
CD ro
P 3 TJ CD
< ft 0 rt p-
in
ro tn p­ C *< CD rt ft 3 0
0 3
<1
ro rt
P
3 3
P­ 0 p­
0
ro o Pi p­
S ft
** CD H H 0 3 0 tp tn 3 P­ 0 ft tp 0 0 ~j

s S
■ — ■
3 p 3 0 0 3 TJ P p- p­ P 3 3 C
Di ro Pi
CD < ro 3p­ 1 O tn 0 H cn 3 <i rt­ tn p­ 3 H *<
ro 1 1 1 I ro cn 1
VI. COMPARISON BETWEEN COMPUTATIONS AND MEASUREMENTS OF THE MEAN
SECOND ORDER FORCE IN REGULAR WAVES

¥!._.:_. .i_l2l_roduction

In this chapter experimental results on the first order

oscillatory motions and mean second order forces and moments for
different vessel forms floating in regular waves will be compared
with results of computations. The purpose of this comparison is
to demonstrate the validity of the theory with respect to physical
reality. For this purpose the following vessels were selected:

- a tanker;
- a semi-submersible;
Seml- Rectangular Submerged
- a rectangular barge; Tanker submersible barge horizontal
Symbol Unit cylinder
Designation
- a submerged horizontal cylinder.
Tested Calcu­ Calcu­ Calcu­ Calcu­
lated Tested lated Tested lated Tested lated
Main particulars and body plans of these vessels are given in
Scale ratio - - 1:82.5 1:40 1:50 1:30
Table Vl-1 and in Figure VI-1.
Length (between m 310.00 100.00 150.00 75.60
L
perpendiculars) PP
The hull forms of the first three vessels encompass the Breadth B m 47. 17 76.00 50.00 8.40
majority of vessels in use by the offshore industry either as per­ Draft T m 18.90 20.00 10.00 12.60*

manent storage vessels, drill ships, drilling platforms, semi-sub­ Displacement volume V m3 234,826 36,925 73,750 4,034

mersible crane vessels, derrick barges and lay barges. For these Centre of gravity KG m 13.32 8.64 7.92 10.00 4.20
above base
vessels the first order motions and mean horizontal drift forces Metacentric height GM m 5.78 16.76 17.48 16.23 0.00
and yawing moment were determined for a range of wave frequencies Transverse gyradius m - 14.77 - 30.55 - 20.00 2.94 2.94
k
in air *K
and three wave directions, i.e. head waves (180 ) , bow quartering
Transverse gyradius m 17.02 17.02 43.47 46.60 24.16 24.20 - -
k
waves (135 ) and beam waves from the starboard side (90°). in water xx
Longitudinal gyradius
in air V m 77.57 77.50 30.89 30.89 39.00 39.00 16.80 16.80

The submerged horizontal cylinder is representative of a Vertical gyradius in air k

zz
m - 79.30 - 41.74 - 39.00 16.80 16.80

submersible hovering just below the surface or of horizontal sub­ Natural period of heave Tz sec. 11.8 11.7 21.3 21.8 - 10.4 - -
merged elements of a semi-submersible. For this vessel the first Natural period of roll T
* sec- 14.2 14.2 21.3 22.4 12.0 12. 1 - -
order vertical motions and the mean second order vertical forces Natural period of pitch T
0 sec. 10.8 10.6 19.5 19.8 9.3 9.4 - -
Water depth Wd m 82.5 40.0 50.0 75.0
were determined for a range of wave frequencies and the same wave
directions as mentioned above.
* Distance between base and mean still water surface.

Table VI-1 Main particulars and stability data-

T*
 YI-iZ^Model^tests

VI.2.1. General

In order to measure the mean second order wave drift forces

on the model of a vessel in regular waves, attention has to be paid
TANKER SEMI­SUBMERSIBLE
to the measuring system employed. This generally takes the form of
a mooring system in which force transducers are incorporated which
measure the mooring force necessary to keep the vessel in a mean
e >
sl o o O position. The mean mooring force is equated to the mean second or­
20 o
der drift force. In chapter II and chapter III it was shown that
tc
r>. the second order forces are dependent on first order quantities.
First order quantities are, for instance, the body motions with
310.0
20
Ol
to
o O C frequency equal to the wave frequency.
100.0

If it is required to measure the mean second order forces

on a free floating vessel the mooring system must be such that the
first order motion, etc. are not influenced by the system. In the­
■ "

ory this requires that the only force which the mooring system
Si i i applies to the vessel is a constant force equal and opposite to
10-6
the mean second order wave drift force.

BARGE
The mooring system shown in Figure VI-2(a) makes use of a
CYLINDER falling weight to counteract the mean wave drift forces.
-*--
f
^
75.6 a. WEIGHT
«
150.0
^~X y a)

__f
SPRING

b )

F i g . VI-1 The v e s s e l s . Fig. VI-2(a) and 2(b) Systems of restraint

75
 shown in Figure VI-3(a). In this figure only one mooring line is
shown.

2 3

a)

< rt tn 3 rt rt cn 0 ft H tn p­ rt 0 H. cn Hi O p P­ ft P tn < rt rt TJ H­ « cn Hi 0 rt tp P3
ro H CD C 3 p­ c H 3 ro c 3 3 H Hi *< P­ H 3 3 H TJ P ro K 0 cn P p ro 3 p­ 3
tn P p cn CD 0 H a CD tp Hi Pi ro Pi cn H a Hi O P­ cn H H H TJ cn < 3__ H H CD < P­
tn
ro
3
tn
0
3
rt
vQ
3
tn
CD
cn
ro
H 3 c
H
Hi
P­
C
o 3
ro
H
ft
3
ti­
ro
cn
rt
ro o
H H
0
H
cn o P­
3
C
CD
ro ro
a
p­
rt
C
TJ
ro CD H
p­
rt
P TJ
CD
3
cn
b)
H Pi Pi 3 C P P 0 ro P CD 3 O TJ 0 ro tp ■
rt p­ cn Qj 3 3 P­ P cn
CD CD rt p­ r+ Hi ro 3 3 H) 3 CD 3 ro 3 cn H *<
P c
0 0 tn o
P
3 3
0
Hi
ft
C
H
ro
p­
C 0 cn
0
H
D.
H
0
H P TJ 3 > 0 H ro 3
0
rt TJ P
"P
rt CD cn
3 ro H 3 rt 3 Pi P H £ 3 3 H H TJ H Pi P­ Z P TS P1 0 H 3 ■
H 3 < Pi ■
tQ rt
Pi H, Pi CD p­ O rt iji P P n­ •< P O H 3 CD Hi 3 rt ti­ 0 0 3 a P P­ ro P­ C ro
Ui CD P 0 ro 3 H < H ft H ro P­ P­ H ft P­ P­ ro ^ H 0 rt TJ H CO 3 rt TJ p­ cn >­_ P­3
Hi H cn 3 CD rt rt ro ^< 3 . 3 cn n 3 H ro P­ 0 3 M p­ p­ Pi C 0 tp P 3 H
H­ P C rt 3 3 P­ Hi tn Hi tp 3 Hi 3 tp 1 pi 3 H ro p Hi 3 Pi cn 3 Pi P­ ro 3
>. H Hi H P 3 0 ro 3 H • Hi CD H H 0 0 O tp P­ 0 rt o rt CD p­ rt < cn 3 P
ro ro 0 CD H ro CD ro CD p ro 3 P­ cn rt H o cn H Hi 3 s: rt ro 3 rt P ro cn
H P­ ro cn TS cn p­ 3 o tn CD P­ 3 CD < p­ 3 cn 3
a Di
tn 3 ft Hi
*P O
CD
H
H
•P
TJ
0
p­
O
CD
0
3
0
Hi ^<
tQ
P­ rt
3
ft ^cn ft o Fig. VI-3(a) and 3(b) Drift force measuring set-up.
rt
TJ
0
O
ro ■ p. CD CD 3 Hi p*
C
CD 3 H CD c
CD H rt H
O
3 ft rt ft
H
tn cn tp
0
CD
0
rt 3 3
P
H 3
0 P
cn P 3
0 tn tn 3 cn ft CD ro 0 3 CD CD cn 3 P 0 H cn P P 3 rt rt K 3 H­ p­ 0 P­ . tp rt­ P P
Ul • 3 0 s; O H 3 ro 0 0 0 P 3 P­ cn P­ ro
3 cn rt P­ 0 cn p, ro ft ro ro 3 H
ft p­ 0 P 0 3 ft ro P­ P 0 3 P­ rt rt P D­ 3 TJ cn ro 3 rt­ ft 3 CD cn H 0 tn 3 a P
3 3 H H tn & ro
Q)
h ro
tn
H < 0 ro p­ cn P­ 3 cn H ro Di 3 H
tn
Hi ^ p­
O
ti­
This situation applies to a vessel for which it is required to
ro H 3 Di
rt 3
0
Hi
H
K
ro ro cn 0 rt
p­
3 P. P P­
3
o 3 tn
3
3 C
CD
P £
P­
0
ft
3
3 rt
3 *
CD
tp
CD
H 0
O
Hi
rt
3 p­ ..
a
0
ro
0 Hi
ip
rt
3 P
tp
P
cn
0
Pi 0 Z
TJ
TJ H
0
H zp ­ H
p­ 3
ro 3
0 ro
ro
H determine the mean horizontal drift forces. The mooring line runs
0 CD tn ft P 3 Hi Hi Hi rt Hi TJ 3 0 o cn ro ro £ 3 ft ro H P­ H LQ
ro cn Z p­ 3 p­
0 P­ 3 rt
z3 H ro 3 3 Pi tn H
rt
3 . P
H
ro cn 0 . rt
H 3 P­ 0 p 3 H 3
rt
ro CD 3 P
cn
cn horizontally from the vessel to the mooring point with the vessel
H • CD 0 CD n ro C TS 2 rt ro o 3 0 H 3 tp Z H < ip rt
P­
3
CD
»
H
P 0
CD
K
tn
rt
0
3 3
rt
3
P
CD
H
p­
3 3
ro ro
3
tn
H
3
rt cn
TJ
p­
H
cn
rt
H
3
3
ro ro
ft P­
3
3 rt
3
tn O
rt cn
O
0
H ro ro CD
^< p ­ o H O 3
p­
o in the rest position. The three force transducers shown in the
ua •• CD CD rt P­ 3 P1 cn P­ rt Hi cn
H. ^tn • ro
H 3 H Pi 0 Pi 3 TJ H P tp ti­

P­
* rt
ft
P
P­
0 tp 3 cn
TS
3 H
P­
cn P p­
3
P 3
rt
H
CD
< rt
cn
^
p­
cn tn 0 3
rt­
3 3
p­ 3
rt
ro H rt
ro figure are:
TJ
ft
P
P­
0 H
P­
Hi 3 ro 0 ^ r+ ro
cn
ro ro p. 3 *£
0 3 3 3 H H •< o H ft 3 p­ ip p­ IP tp C CD 3 ro 3 3 3 3
p­
3 ft
3
H
3 CD rt Pi
CD
3
■XI
ro
tp rt
0
3
Hi
H
p­
3
P­
0
<P 3 3
o 0
3
^P
TJ
P c
H
tn
ro
0
H»
C
4
tn
ro
Hi
Hi
CD 0
3
3 p
ro
rt
P­
H­
cn
cn
O P
CD c
cn
P­
o
- a force transducer measuring the longitudinal force component of
rt 3 CD rt Hi Hi H c 3 cn CD <p 3 0 rt Pi 3 H P H CD a ro a ro p­ 3 0 CD 3
• ro CD 3 0 0 Cfl ro CD rt >D cn cn 0 C p­ P T) ti­ H
z n p­ n cn rt CD 3 rt ti­ 0 Pi the mooring line force relative to the body axes (transducer 1);
CD H H Hi K2 3 P c 3 H Hi H P H­ 3 < ro rt Ho K* 0 «. 0 3 ro 3 O
H
3
3
0
TJ
0 0
0
CD ft
0
H
tn
rt­
0
p­
3
0
3
rt
ro
3
p­
3
0
Hi
cn
rt
3 *< P
O
H
ft
ro s: cn
^Q P
p­
o
H
1
X
ti­
• Hi
P­
ro
Pi
0
Hi a ft
rt P
(1­
cn
rt
cn
rt
3
^
o
3
- a force transducer measuring the tension in the mooring line
CD o tn 0 3 0 ro ro 0 tn O P P rf ti­ C < 3 3 M ro O H 3 ^P P Hi

ft
H p­
rt
H P­ ro CD
cn
3 cn H P­ rt ft 3
rt
Hi tr ro P P ro
cn
o ­3—. 3tn C c rt p­ ro C rt p­ 3
rt
cn 0 (transducer 2);
p­ H 3 P­ P ro 3 3 Hi ro H 0 H o r. M 3 3 Hi p­ 3 3 ro
3
H
3
ip
p­
0
CD
O rt
3
0 Z
0
3
0
H!
3
tP
H
CD
tn CD CD cn CD
n Hi
P­
cn 5 rt rt
H
ro
CD
cn
•^^ p ­
• <
rt H
CD
CD rt ro
^P
rt
CD
ro
H rt
<
CD
g
- a force transducer measuring the horizontal component of the
ro 3 ft 3 Di p­ 0 3 < P rt p­ ft CD 0 3 pi C ro rt tn O P c 3 CD ro H s:
ro H
P­
cn CD CD H
H
rt
3
rt
3
tn 0 Hi 0 CD
cn
H
CD
cn H
Ui
p­
o
3
cn ft
p­
tn 3
\~J
rf
H_ H 0 cn 0
3 •s H
p­ O 0 <P 3
cn
P
H
ro
H
mooring line force relative to an earth-bound system of axes
< H •< 3 0 3 K P 0 P. C
TJ
0
3
ro
0
Hi
P
H
3
0 3 3
ro CD tn
ti­
0
tn
P­
3
H
P­
cn
ro O
rt
3
ft cn C
H
3
CD cn
­< Ul P­
cn p­
Hi
P­
rt
3
cn
rt
•< P­
3
H
H
CD
P*
H
P
p­
0 P­
H
(transducer 3 ) .
tn ■ »
c o P CD Hi 5­ ro CD rl­ 3 H 3 ro o o p K:
3 p tn 3 ro P Hi H ro H 3 3 S
p­ ft ro n tn P­ P 3 3 CD tp •« O H Hi ti­ 3 cn ro H cn Pi 3 H P­ 0 r. < P­
rt Hi 3 p­ 3 H < H cn 3 PCDi ro CD rt­ P ro •< rt C rt O c ft ro £ rt CD (1­ Since it is required to determine the horizontal drift force the
p­ P­ CD 0 3 3 CD tn CD P tn CD cn Z CD ro ft P ro 3 cn 3 rt ro 3 g ­. 3 P *< cn 3
0 X Hi tp ro tP ft cn H C tn
ft
K
cn
3
p­
3
_
H 3 Z
p­
3 3 cn LP
p­
rt
CD
ro 3 p­
tp
ti­
rt
P
p­
<
CD
TJ
CD
rt
P­ ft
true force will be measured by transducer 3.
3 ro Hi
tn
ro H
P­
• ro n ti­ cn
ro rt cn tn
ro ro
3
0
Hi
0
3 cn tp
cn pi 0 rt 3 o
H
3 o
CO
3 P o
Hi
<. 3 3
ft cn
3 H
cn
­ 3
H 3 K ua H p­
rt
0 ro 3 0 0
rt
3 3 CD TJ ro 0
Hi
0
p­
ro Hi 3
P
rt
ro
P rf n ro cn n p­ Di 3 3 Hi 3 CD 0 O TS 3 Z 0 rt n­
H 0 ro ti­ 3 3 CD P­
tn rt
3
rt
P rt 0 P­ 0 rt 0
3
H
P­
M
P­ p­
P
cn
ft H p­
cn
0 3 0
H
3 p­
3
3 P
cn
0 P In order to determine the validity of measuring the force
ro 3 ro 0 H H P H 3 0 O H 3 H 3 ro ro ro H
cn
3
ft ro 3 cn ro rt P­ ro CD 3 3 CD P 0
^J 3 P ro • 3 CS
3
ro
3
ro
P
H
CD P
rt
3
a
3
cn 3
Di
tn CD 0 P 3 ro rt
0
< by transducer 1 or transducer 2 we assume that the vessel is car­
~~J CD 3 3 0 3 ro tp cn rt H cn 3 p­ ro
1 1 1 ro 3 rying out oscillatory motions which induce the angles a and $ be­
tween the mooring line and the horizontal plane and the longitudi-

78
nal axis of the vessel (see Figure VI-3(b)). If the force in the Wave frequency Wave amp lit ude c, in metres
mooring line is measured by transducer 2 is F^, then the relation­
ship between this force and the force F. is:
to in
rad./sec. .\/"f' Wave direction Wave direction Wave direction
90 degrees 135 degrees 180 degrees

0. 178 0.446 2. 14 2.14 2.14

F 3 = F 2 cos a (VI-1) 0.267 0.669 1.21 1.21 1.21
0-267 0.669 2.11 2. 11 2.11
0.267 0.669 2.86 2.86 2.86
or: 0.357 0.895 1.86 1.86 1.86
0.443 1. 110 1.23 1.23 1.23
0.443 1.110 1.98 1.98 1.98
F0 = F., — I — (VI-2) 0.443 1. 110 - 2.83 2.83
0.532 1.333 2.02 2.02 2.02
2 3 cos a 0.623 1.562 2. 19 2.19 2. 19
Assuming that the angle ot is small we may write: 0.713 1.787 1.20 1.20 1.20
0.713 1.787 2.06 2.06 2.06
0.713 1.787 - 2.76 2.76
cos a = 1 - ^sa2 (VI-3) 0.804 2.016 2.11 2. 11 2. 11
0.887 2.223 1.76 1.76 1.76

from which it follows that: Table VI-2 Test conditions for model tests with a tanker.

F 2 = F3(l + ha2) (VI-4) Wave frequency Wave amplitude £ in metres

a

-\/Z'
OJ in Wave direction Wave direction Wave direction
From this it is seen that the error in F„ is in the order of rad./sec. 9 0 degrees 135 degrees 180 degrees
F^.^ot . Normally the angle a will be of the same order as the an­
0.300 0.550 - - 2.61
gular motions of the vessel (pitch, roll and yaw) which, except 0.300 0.550 - - 4.49
for the case of resonant roll motions, will normally have ampli­ 0.300 0.550 - - 5.09
0.455 0.834 0.97 0.97 1.02
tudes less than about 0.1 rad- Assuming this value of a the error 0.455 0.834 2.15 2.15 1.96
in F will, at most, be of the order of F^ x 0.005 or about one 0.455 0.834 3.18 3.18 2.79
0.525 0.963 0.91 0.91 0.98
half percent of F_. In the same way it can be shown that the error 0.6 30 1.155 1.01 1.01 1.02
in F is of about the same magnitude from which it can be con­ 0.630 1. 155 1.99 1.99 1.97
0.630 1.155 2.83 2.83 -
cluded that the forces may be measured in any of the three men­ 0.72 7 1.333 1.10 1.10 -
tioned positions provided the angular motions of the vessel are 0.785 1.440 - - 1.03
0.805 1.476 1.18 1.18 -
not large- 0.910 1.668 1.14 1.14 -
0.920 1.687 - - 1.04
0.920 1.687 - - 1.76
VI. 2.2. Model test conditions 1.047 1.920 - - 0.98
1.047 1.920 - - 1.61
1.060 1.944 1.04 1.04 -
Model tests were carried out in regular waves for three 1.168 2.142 0.94 0.94 -
directions, i.e. head waves, bow quartering waves and beam waves 1.200 2.200 - - 0.78
1.200 2.200 - - 0.81
for a range of regular wave frequencies. The test conditions with 1.278 2.344 0.98 0.98 -
respect to the wave direction/ wave frequencies and wave ampli­ 1.339 2.455 0.80 0.80 -
1.350 2.476 - - 0.64
tudes are given in Table VI-2 through Table VI-5. 1.398 2.564 0.83 0.83 -
1.570 2.879 — — 0.43

Table VI-3 Test conditions for model tests with a semi-submersible

7q
 Yli^i^Computations

The distribution of facet elements on the mean wetted part

of the hulls of the vessels and, for the surface vessels, the dis­
tribution of the line elements around the waterline are given in
Wave frequency Wave amplitude ix in metres Figures VI-4 through VI-8.
3 a
a. in
rad./sec. .\K' ' Wave direction Wave direction Wave direction
90 degrees 135 degrees 180 degrees

0.300 0.620 0.81 0.81 0.81

0.400 0.827 0.94 1.07 1.07 FACET SCHEMATISATION TANKER
0.500 1.033 0.96 0.96 0.96
0.600 1.241 0.98 0.98 0.98 TOTAL 3 0 2 FACETS
0.700 1.447 1.09 1.09 1.09
0.800 1.654 1. 16 1.16 1.16
0.900 1.861 1.08 1.08 1.08
1.000 2.067 0.94 0.94 0.94
1. 100 2.274 0.79 0.79 0.79

Table VI-4 Test conditions for model tests with a rectangular

barge.

Wave frequency Wave amplitude r in metres

a
a) in
rad./sec.*
-\/'f Wave direction Wave direction Wave direction
90 degrees 135 degrees 180 degrees

0.760 0.968 0.55 0.55 0.55

0.881 1.122 0.73 0.73 0.94
1.081 1.377 1.06 1.06 1.06
1.260 1.605 1.08 1.08 1.08
1.530 1.949 1.20 1.20 1.20 WATER LINE SCHEMATISATION
1.880 2.395 0.75 0.75 0.75 TOTAL 74 ELEMENTS
2.167 2.760 0.50 0.50 0.50

*—I 1—PH 1—* 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 .-

Table VI-5 Test conditions for model tests with a submerged

horizontal cylinder.

Fig. VI-4 Facet and waterline element distribution of a tanker,

ft 1 BT
 FACET SCHEMATISATION SEMI-SUBMERSIBLE
TOTAL 216 FACETS

FACET SCHEMATISATION BARGE

TOTAL 138 E L E M E N T S

-e ?

WATER LINE SCHEMATISATION

TOTAL 72 ELEMENTS

WATER LINE SCHEMATISATION

TOTAL 4 8 ELEMENTS

H 1 H -.—I—h

■H 1 r H 1 1 1 1 \-

F i g . VI-5 F a c e t and w a t e r l i n e element d i s t r i b u t i o n of t h e s e m i -

s u b m e r s i b l e : 216 f a c e t s . Fig. VI-6 Facet and waterline element distribution of the rectan'
gular barge
 FACET SCHEMATISATION SEMI - SUBMERSIBLE
TOTAL 360 FACETS

FACET SCHEMATISATION CYLINDER

TOTAL 286 FACETS

WATER LINE SCHEMATISATION

TOTAL 144 ELEMENTS

Fig. VI-8 Facet and waterline element distribution of the semi-

Fig, VI-7 Facet distribution of the submerged horizontal cylinder. submersible: 360 facets.

85 86
 The computations were carried out for the same wave direc­
tions as the model tests and covered the range of wave frequencies
tested. The wave frequencies for which computations were carried
out for the four vessels are given in Table VI­6.
Wave fre quency Wave frequency
Vessel
Additional input data for the computations with respect to the
vessels are given in Table VI-1. In the case of the semi-submers­
ible some differences occur between the model data and input data
for computations with respect to the position of the centre of
gravity and the transverse radius of gyration in air. The influ­
ence of these differences on the results of computations was small,
however.

.vz Vessel

-\/'r
/" — ■ ­i
i ­i / **_

oi in to in
rad./sec. rad./sec. YJiii__,5®§Hi£§_2f_22^Eyt5ti2_}§_22__r_™__§§^2^2{___§

0.300 0.620
0.450 0.930 The results of computations and measurements with respect
0.079 0. 198 0.480 0.990 to the first order oscillatory motions of and about the centre of
0. 112 0.281 0.500 1.035
0.189 0.474 0.516 1.070 gravity are given in Figures VI-9 through VI-13. The results are
0.266 0.667 0.539 1. 116 presented in the form of non-dimensional frequency response func­
0.354 0.887 0.560 1.160
Tanker 0.444 1.113 Rectangular 0.600 1.240 tions of the amplitudes of the motions to a base of the non-dimen­
0.523 1.312 barge 0.650 1.340 sional wave frequency. The phase angles of the motions are given
0.600 1.505 0.700 1.450
0.713 1.788 0.780 1.610 in degrees, also to a base of the non-dimensional wave frequency.
0.758 1.901 0.800 1.660 A positive phase angle indicates that a motion reaches its posi­
0.887 2.224 0.900 1.860
1.000 2.070 tive maximum value before the crest of the undisturbed incoming
1.100 2.270 regular wave passes the centre of gravity of the vessel. The posi­
tive direction of motions and forces are in accordance with the
0.098 0.180 rectangular system of axes G-X'-X'-X" shown in Figure III-l.
0. 146 0.268
0.200 0.36 7
0.300 0.550 Comparison of computed and experimental data on the first
0.400 0.733
0.500 0.916 order motions shows that the motion amplitudes of all four vessels
0.600 1.100 0.421 0.5 36 are generally well predicted by the computations. Significant dif­
0.702 1.290 0.698 0.889
Semi­ 0.800 1.466 Submerged 0.836 1.064 ferences occur in roll and sway motion amplitudes near the natural
submersible 0.900 1.650 horizontal 1.080 1.376 frequency of the roll motions of the barge and the tanker. These
1.000 1.830 cylinder 1.325 1.687
1.050 1.925 1.528 1.946 are mainly due to the fact that the computations predict a larger
1.100 2.020 1.872 2.385 roll motion due to the omission of viscous effects in roll damping
1.150 2.108 2.160 , 2.752
1.200 2.200 in the computations. Due to the sway-roll coupling computed sway
1.250 2.292 motions also differ somewhat from the results of measurements.
1.300 2.380
1.400 2.570
1.500 2.750 At very low wave frequencies non-dimensional amplitudes of
the measured roll motions of the tanker become increasingly larger
than the computed values. The differences between computations and
Table VI­6 Frequencies used for computations.
measurements at these frequencies are, however, exaggerated due to
the way angular motions are made non-dimensional.
 TANKER
COMPUTED TANKER
O._.D MEASURED (ASC ENDING WAVE AMPLITUDE o„
500 — C OMPUTED
500 .500*
SURGE o.a . MEASURED (ASC ENDING WAVE AMPLITUDE o _ . )
500° 500 500
SURGE HEAVE

■
z
2 i l
■ " • ^
•
vy
; ^ • "1
/ 0 4 . A^
tl)

V x
11!
3a

tf'
^ 0

\ &

lo

4 0 2

SEMI­SUBMERSIBLE
>n_n « . V i
0 2 °.. 1
4

500° 500° 500° SEMI­SUBMERSIBLE

SURGE HEAVE PITCH
500 500 500
•

zz
SURGE SWAY

\ J
7. 0 1­0 // \
■
*­.
0 1.0 rf ' ­>v
• •
•1

VV
• (it "^ 7­fi 0 4
•
0 1.0

*3a *5a
11
d • *2a

0.5 0.5

t
*v I*
V^YW, <S> „
4 0
Vot
2
w­^ 4 0
*
X, TLo n f NvTL^—n
BARGE 2

500 500* 500° BARGE

SURGE HEAVE
500 500 500

e
1
/. SURGE SWAY

•­*­ /.
^ / / •

ID ••
0 1.0
o
• ­ * ■

kw 0 2 r \ 0 1.0

P
(1)
_UlL
x
3a
r .­■­
(D
■ t f
^2a
\ \ ,tii

1W o
CI

L
\ \
©T,
0
2
7 V3
4 °0

0 i
KJ -^r.
4 0

• \ ^

Fig. VI­9 First order motions in head waves. Fig. VI­10 First order linear motions in bow quartering waves.

SQ
 TANKEhf TANKER
■COMPUTED ■COMPUTED
o.a.o MEASURED (ASC ENDING WAVE AMPLITUDE <=.*­° 1 «s.4.o MEASURED (ASC ENDING WAVE AMPLITUDE o . ­ . o )
500 500° 500 500* i 1 1 1500° 500*
SWAY

• 1 j

I
2a

V o

4 "O 2

SEMI­SUBMERSIBLE

500° i 1 1 i5O0* 500 500 500° 500

n ROLL PITCH YAW SWAY HEAVE ROLL

J
rw
•

1.0 \
/ .
O 1.0 0 1.0
"L ;
0 1.0 _ 0 2
J
if1' / 5a ____§.
« _t

4a __£
kff ff *E

05 0.5 0.5

k
\
. \
4 O
f "* n o
2
M°o, IX: A 0
—L£ IkfflP­.
4 °0

BARGE BARGE

500° 500° 500° 500* 500* i500°

ROLL SWAY HEAVE ROLL
o

O Oj
• •
^1 V ?' 0 1.0 0 1­0 0 10
►*. K<T ■ ■ r • • >•
•­A
V X 11 '
*2a *3a xx , 1 )
4a
4a
«__' <
\ 0.5

y
IfV
\ ~ ! \ •/
s.
0
—Nl­o.o. ~ 4
4 °0 4 %

«\/I ^7T
9

Fig. VI­11 First order angular motions in bow quartering waves Fig. VI­12 First order motions in beam waves

91 Q->
 CYLINDER ( (i=180*> In terms of absolute differences between the measured roll angle
COMPUTED
MEASURED
and the roll angle computed for the wave amplitude concerned the
500* 500
HEAVE differences are comparable to those found at higher frequencies.

\• The phase angles of the first order motions are generally

10 0 1.0
reasonably well predicted. When motion amplitudes are small dif­
*';*
ferences in the phase angles are somewhat increased. In such cases
"F \
the harmonic analysis technique, by means of which the amplitude

z __v.
0.5 \ and phase angle of the motions are determined from measured data,
\
tends to give less correct results since at low motion amplitudes
the influence of errors in the measurements become more important.
0 - 9 — -
. 4
The results of computations and experiments on the mean
500 second order forces are given in Figures VI­14 through VI­17.
PITCH
These results are given in non­dimensional form making use of the
displaced volume of the vessels. For equal displacement volumes
the results for the various vessels are directly comparable. The
forces and moments are given to a base of the non­dimensional wave
frequency.

■COMPUTED
O.H.B MEASURED ( A S C E N D I N G WAVE A M P L I T U D E o. f l ._ )

TANKER SEMI­SUBMERSIBLE BARGE

500
HEAVE
­2 ­2 ­2
■

1.0
•

1 * 1f. pa

1
(1) ­1 ­1 ­1
!_3_L \
COMP UTED
\ ^ 3 6 0 FW E T S

0.5

n^
f^ ­t> fi S I
_ fl

«V^

Fig. VI-13 First order vertical motions of the cylinder in head

Fig. ­VI­14 Mean longitudinal drift forces in head waves
waves, bow quartering waves and beam waves.

93 94
 COMPUTED
o _.o MEASURED ( ASCENDING WAVE AMPLITUDE o 4.0 )

TANKER SEMI-SUBMERSIBLE COMPUTED

o.a.t. MEASURED (ASCENDING WAVE AMPLITUDE 0.4.0 )

TANKER SEMI-SUBMERSIBLE BARGE

-2 0
M

t
cm 9
3
.£_ 5.0 5.0

A- f
> 0
01
ct
-01 0 / T
0
/
- L - COMPUTED
, / y f ° 3eoFACETS 2.5 1 2.5 25

P '0
_ c
•J\ _
/ COMPUTED
X - 360 FACETS

__A^/
0 . — 0 0-^
0 t>
0
j •"
> 4 4 0

' SEMI-SUBMERSIBLE BARGE

/TANKER
I
i
4 J
0
Fig. VI-16 Mean transverse drift forces in beam waves.
2
_
M t °

Aj
E- CYLINDER
a.
A /COMPUTED
COMPUTED
L - | 360 FACETS MEASURED

j
" 1
0 jo 11=180° li>135° U = 90°
0 ,

c
/
0

/
a J
of
A-
M
0 cc - ^
4
-^ r> 2
0 2
3
•> 0

2.5
TANKER

25
SEMI-SUBMERSIBLE BARGE
J \
4 °(D

1
V 0

; ' °< ) "

c

,>
^ 0

•<

11 = 1 6 0 ' H=135* li=90"

IS 0
1..
Jv
a

V s "*ijV A n
COMPUTED
. 360FACETS

0
>
0 a

IA^ 0 0 IS > 0 rV 0
\ °/
_.___._
\i
V
-2.5 -2.5 -25 0 s *
1
• ^

Fig. VI-17 Mean vertical drift force and trim moment on the cyl­
Fig. VI-15 Mean longitudinal and transverse drift forces and inder in head waves, bow quartering waves and beam
yawing moment in bow quartering waves. waves.

96
95
The results given in Figure VI-14 through Figure VI-17 indicate
that in general the mean second order forces and moments are rea­
sonably well predicted with some notable exceptions. For the tan­
ker the mean transverse force in beam waves given in Figure VI-16
and the mean yaw moment in bow quartering waves in Figure VI-15
show a large difference between computations and measurements at
the non-dimensional wave frequency corresponding to the natural
roll frequency. At this frequency the mean transverse force given
FREQUENCY: w = 0.444 rad./sec.
in Figure VI-16 shows a large peak in the computed results which
,1/3
is not present in the experimental results. The mean yaw moment = 1.113
given in Figure VI-15 shows that at the natural roll frequency
computed and experimental results are opposite in sign. The com­ Calculations

puted results show a marked negative peak which again is not pre­ Wave
Roll dampin g m u l t i p l i c a t i o n factor
Measure-
ments
Unit
sent in the experimental results. For the barge similar abrupt Description direction
in deg.
changes in the computed drift forces and yawing moment are found 1
V-- C
uC V-- E
u. V-- E
ur.
near the natural roll frequency. It is felt that, as is the case
MOTIONS
with the overprediction by the computations of the first order
roll motions, the large discrepancies in the mean drift forces may *._xu
^zx
Sway 90 0.35 140 0.63 255 0.70 263 1 . 12 267

be also due to the fact that viscous effects in the roll damping Heave 2

*axz
90 1.38 35 3 1.38 353 1 . 38 353 1.35

are not included. Roll 90 62.50 141 1 5 . 19 172 8.46 176 9 . 36 248

*!ZX
*.xx
Surge 135 0.25 91 0.25 91 0.25 91 0.28 112

Sway 135 0 . 19 290 0.25 268 0.26 268 0.41 273

In order to check, qualitatively, the influence of the roll 1 1
damping on the results of the mean drift forces, computations were
(leave 135
-.i ". ' 0.56 9 0.56 9 0.56 9 0.68 33

repeated for different values of the linear roll damping at the

Roll 135
-UXZ 7 . 12 85 1.73 117 0.97 121 5.18 253

*._XZ
*axz
Pitch 135 0.48 253 0.48 253 0.48 253 0.71 271
natural roll frequency of the tanker. Computations were carried Yaw 135 0.25 348 0.26 356 0.26 357 0.26 8

out for roll damping values of 1, 3 and 5 times the potential roll
MEAN DRIFT FOBCES AND MOMENT
damping. The results of computations are shown in Table VI-7 in
which the mean values of the measured data, given at this frequen­ Transverse 90 F2AP9V 1 / 3
ci 1 ) 2
33.628 9.650 5.431 1.210
force
2
cy in the figures, are also given. From the results given in this
table it may be concluded that increasing the roll damping tends
Longitudinal
force
135
v^'V -0.274 -0.203 - 0 . 193 -0.203

1 / 3 1 ) 2
to effect both first and second order quantities in such a way Transverse 135 F 2 A^ ^ -0.152 0.233 0.221 0.443
force
273 2
that better agreement with experimental data is obtained. The re­ Yaw m o m e n t 135 v^ ^ -0.783 0.405 0.505 1. 143

sults also show, however, that if the roll damping is increased

to such an amount that the computed roll motions agree with the
measured roll motions this does not necessarily result in agree­
ment in the computed and measured drift forces. This discrepancy
may be a result of the phenomenon that viscous damping effects Table VI-7 Influence of linear roll damping on the motions and
are less linear than has been assumed here. mean drift forces and moment on a tanker at the natural
roll frequency.

97 98
 With respect to the results obtained for the semi-submers­
ible it is noted that the mean drift force in head waves given in
Figure VI-14 shows rapid fluctuations with the wave frequency,
which are not present in the corresponding results for the tanker
and the barge. Fluctuations are also present in the results for
the tanker and barge at other headings. In that case they are main­
ly due to the additional peak at the natural roll frequency of the
tanker and the barge. For the semi-submersible these fluctuations
are not due to such effects since the frequencies at which reso­
nance in roll, pitch or heave occurs are restricted to quite low
frequencies as can be seen from the results on the first order mo­
tions given in Figures VI-9 through VI-12.
For the semi-submersible the fluctuations in the mean drift
forces appear to be related to interaction effects between the col­
umns. In head waves the results given in Figure VI-14 show a marked
reduction in the mean drift force at a non-dimensional wave fre­
quency of 2.2. In beam waves the results given in Figure VI-16 show
a similar reduction at a non-dimensional wave frequency of about
1.8. The wave lengths corresponding to these frequencies for the
head waves and the beam wave case amount to 43 m and 62 m respec­
tively for the vessel size as given in Figure VI-l. These values
are quite close to the distance between the columns as measured in
the direction of the wave propagation which amount to 38 in and
60 m respectively. In such cases standing wave effects may occur
between the columns.
In order to check the quadratic relationship between the
mean second order forces and the wave amplitude, experiments with
the tanker and the semi-submersible were carried out in waves with
different amplitudes. The influence of the wave amplitude on the
various quantities at a number of wave frequencies can be seen in
the figures. The wave amplitudes are given in Table VI-2 and Table
VI-3. In general the quadratic relationship between the mean forces
and the wave amplitude is conformed with to a reasonable degree.
For the tanker there is a trend, however, which indicates that the
non-dimensional forces and motions reduce with increasing wave am­
plitudes. For the semi-submersible the influence of the wave ampli­
tude is less consistent.
For the computations the surface of the hull of the vessel
is approximated by a distribution of plane elements or facets.
These facets represent a distribution of source singularities
which each contribute to the potential of the flow about the ves­
sel. The choice of the number of facets used for the computations
is a compromise between the quality of the results obtained and
the costs of computations. Increasing the number of facets is ex­
pected to increase the quality of the results but the costs also
increase. The question as to whether a given number of facets will
yield satisfactory results can only be checked by repeating compu­
tations using more elements and comparing the results. In order to
show the influence of the number of facets on the results addition­
al computations were carried out for the semi-submersible at three
wave frequencies. For these computations 360 facets were used in­
stead of 216 facets. The distributions are shown in Figure VI-5
and Figure VI-8- From these figures it can be seen that for the
case of 360 facets the additional number of facets arises from the
finer distribution on the columns. The results of the additional
computations are indicated in the figures. In general the influence
of the number of facets is small. Some influence is found at the
highest wave frequency. The differences in the results using more
or less elements are, however, less than the differences found be­
tween measurements and computations, so that for this case 216
facets were sufficient to give satisfactory results.
In Figure VI-18 the computed components of the mean second
order longitudinal drift forces on the tanker, semi-submersible
and barge are given. The numerals indicate the components given
by equations (IV-1) through (IV-4). From the results given in this
figure it is seen that, as was the case with the hemisphere treated
in chapter V, contribution I due to the relative wave elevation is
dominant in all cases. Contribution II, due to the pressure drop
as a consequence of the fluid velocity, is in a direction opposite
to contribution I. Contributions III and IV are generally less im­
portant. The sign of these contributions is different for the three
vessels. For the tanker they are predominantly of the same sign as
the total force. For the barge the opposite is true. For the semi-
submersible contribution IV is practically zero except for the very
low frequencies near heave and pitch resonance. For this vessel
I nr.
 contribution III oscillates in sign with the wave frequency.

For the submerged cylinder the contributions to the total

mean vertical drift force are given in Figure VI-19 for the head
wave and beam wave case. Since the vessel is fully submerged no
TANKER BARGE contribution arises from relative wave elevations around a water-
-2.5 25
line . Hence contribution I is zero. Since the cylinder is circular
with the centre of gravity in the centre of the cylinder no roll
motions occur. Due to this effect contribution IV is zero in beam
\f waves. The results shown in Figure VI-19 show that in this case
0
~ the relative importance of the contributions to the total force
\
can vary quite significantly. In head waves contribution II is
N.

dominant and contributions III and IV have only minor effect on

25
the total. In beam waves, however, contributions II and III are of
the same order but opposite in sign. In both cases the total mean
force is directed upwards. The sign of contribution II is also up­
wards in both cases. It will be recalled that for surface vessels
Fig- VI-18 Components of the computed mean longitudinal drift the mean horizontal force due to contribution II was directed op­
forces in head waves. posite to the total force.

VIi5i_Conclusions
CYLINDER
I
n From the results presented in this section it can be con­
m
TOTAL
cluded that the mean wave forces on bodies of arbitrary shape can
H=180° li = 90*
be computed with reasonable accuracy based on the method presented
in this study. It was generally recognized that the mean forces

A
25

acting on full forms such as barges and ships could be computed

i
/'^> with the aid of potential theory and the supposition that the
\ forces are basically a second order phenomenon. The correlation
0 0
\ s y ^ '
shown here between the results of computations and measurements of
the mean wave forces acting on a semi-submersible and a submerged
-2.5
cylinder indicates that the same theory can also be used to pre­
25
dict these forces on more slender forms.
4 0
From the results some interesting observations with respect
to the mean horizontal forces acting on a semi-submersible can be
made. The results shown in Figure VI-15 on the mean horizontal
Fig. VI-19 Components of the computed mean vertical drift forces drift force in head waves indicate that, contrary to expectation,
on the cylinder in head waves and beam waves. for equal displaced volumes the mean force acting on the semi-

lni
submersible can be as large as the force acting on a rectangular
barge and considerably larger than the force acting on a tanker.
Furthermore, in beam waves the reverse is true. From Figure VI-16
it is seen that the mean forces on the semi-submersible are less
than on the barge or on the tanker. The mean force on the semi-
submersible in beam waves is of the same magnitude as in head
waves; however, the frequencies at which the mean forces are large
or small are different in head waves and beam waves. This means
that in irregular waves, if it is required to minimize the mean
forces acting on the vessel, it may be possible to let the peak of
the wave spectrum coincide with a minimum in the response function
of the mean force by altering the heading of the semi-submersible.
For the case of this semi-submersible, for instance, if the irreg­
ular wave spectrum is such that the non-dimensional frequency of
the peak of the spectrum is about equal to 1.8, the mean force on
the vessel will be smallest with the vessel turned beam-on to the
waves. If the non-dimensional frequency of the peak of the spec­
trum is appreciably lower, say about equal to 1.2, then the mean
force is smallest with the vessel head-on to the waves.
From the aforegoing discussion it can also be concluded
that, if a semi-submersible is to operate in a specific location
and under specific design conditions with respect to the irregular
waves, it is in principle meaningful and possible to investigate
the influence of the dimensions and layout of the vessel on the
mean second order drift forces with the aim of optimizing a design
from this point of view.
in.
VII. DETERMINATION OF THE QUADRATIC TRANSFER FUNCTION OF THE LOW
FREQUENCY SECOND ORDER FORCES
YII__I___l£t£2<__-_-2l_i2-}
Experimental determination of the low frequency components
of the second order wave forces acting on vessels in waves places
unusual demands on the system of restraint and the measuring sys­
tem employed. In this chapter the ideal characteristics of the
system of restraint are discussed and two possible realizations of
such systems are introduced. The ideal characteristics of the sys­
tem of restraint are only partly obtained by these two systems so
that it must be borne in mind that further development in this
field is necessary.
In order to demonstrate the type of results which may be
obtained experimentally and to verify the results of computations
some experimental results on the quadratic transfer functions for
the amplitude of the low frequency second order longitudinal force
on two vessels in head waves will be compared with results of com­
putations.
As indicated in chapter IV the quadratic transfer function
for the low frequency force corresponds with the low frequency
component of the second order force when a vessel is floating in a
regular wave group consisting of two regular waves with frequen­
cies u_ and Gux. There are two methods by means of which these re­
sults may be obtained from experiments:
1. From model tests in regular wave groups; the results are direct­
ly comparable with computed results.
2. From model tests in irregular waves; the time records of the
second order forces are analyzed by means of cross-bi-spectral
analysis techniques. The results of this analysis are directly
comparable with results of computations.
In this chapter some results obtained by both methods will be giv­
en. The cross-bi-spectral analysis technique employed here was
based on the method developed by Dalzell [VII-1]. The computations
of the quadratic transfer function are in accordance with the
method discussed in chapter IV.
Experiments and computations were carried out for the fol­
lowing vessels, which were also treated in chapter VI:
- a tanker;
- a semi-submersible.
The main particulars of these vessels are given in Table VI-1. For
the tanker model tests were carried out in regular wave groups and
in irregular waves. For the semi-submersible model tests were car­
ried out in irregular waves only. In all cases results are given
for head waves. For many practical cases, for instance in very
high seas, head waves represent the most important wave direction
for moored vessels.
YI_t__2i_Model_test_setzup
of restraint must allow the model to move freely at wave frequen­
cies while at the same time low frequency motions corresponding to
the low frequency wave drift forces must be fully suppressed. In
that case the forces in the mooring system will possess only mean
and low frequency components which will correspond to the required
forces.
The response of the mooring system is shown schematically
in Figure VII-1.
WAVE SPECTRUM
SPECTRUM OF LOW FREQUENCY
SECOND ORDER FORCE
IDEAL RESPONSE OF SYSTEM OF RESTRAINT

VII.2.1. General
' "~v

100% N

In this section the specific requirements, which must be \ l

met in order to be able to measure the low frequency second order \
\\ II
wave forces on a vessel, will be discussed. \
\
UJ \
CO \
From the expressions derived for the second order forces in zO \
chapter III it can be concluded that, in order to arrive at the Q.
(/) \
LU \
correct value of the forces, the model restraint must be so that \

\I
Q_

first order motions are not affected by the method of restraint.

In case the mean second order forces are to be determined no other
requirements have to be met by the mooring system. When it is re­
quired to measure the low frequency second order force an addition­
al requirement must be fulfilled, namely that the model does not
carry out motions with frequencies which coincide with the frequen­
cy of the second order forces. This requirement is analogous to the
case where it is required to measure first order wave loads. In
that case the captive model must be rigidly held so that it does
not carry out motions at wave frequencies. Failure to comply with
this requirement results in incorrect values of the forces due to
dynamic magnification effects following from the elasticity of the
system of restraint. For the case under consideration the system
r. o/„ s l ^--»_
FREQUENCY
Fig. VII-1 Ideal characteristic of the system of restraint.
In this figure the wave spectrum and the spectrum of the low fre­
quency second order forces in irregular waves are given schemat­
ically to a base of frequency. Also shown in this figure is a line
which indicates the idealized characteristics of the system of
restraint of the vessel. In the range of frequencies of the second
order forces the system is at 100%, indicating that it must sup­
press fully motion components of these frequencies. At wave fre­
quencies the system is at 0%, indicating that for these frequencies
the system should, ideally, not exert any forces on the vessel.
The system of restraint, in effect, must possess low-pass charac­
teristics. The high frequencies to be filtered out in this case
being the frequencies of the waves. Such characteristics may be
approximated by the application of a dynamic system of restraint
incorporating control and servo systems which react in the required
manner to the motions of the vessel in waves. In Figure VII-2 a
block diagram of such a system is given.
cases, instead of using an active system of restraint as described
here, a simple passive system consisting of linear springs may be
employed. Consider the case of a vessel restrained in the longitu­
dinal direction by means of linear springs. For simplicity we as­
sume the virtual mass and damping to be constant. The system of
restraint is shown schematically in Figure VII-3.
HYDRODYNAMIC
DAMPING
Fm(t) MEASURED FORCE
H>

WAVE FORCE
WAVES
SPRING

VESSEL

SERVO POSITION
SYSTEM VESSEL
Fig. VII-3 Schematic r e p r e s e n t a t i o n of a p a s s i v e mooring s y s t e m .

The waves e x e r t f i r s t and second o r d e r f o r c e s on t h e v e s s e l which

c a u s e i t t o e x e c u t e m o t i o n s . The f o r c e i s d e n o t e d by F ( t ) i n F i g -
ure V I I - 3 . In t h e mooring system a f o r c e t r a n s d u c e r i s mounted
which measures t h e mooring f o r c e F m ( t ) . I f t h e s h i p - m o o r i n g - s y s t e m
CONTROLLER i s assumed t o be l i n e a r t h e a m p l i t u d e r e s p o n s e f u n c t i o n of t h e
measured f o r c e i n r a t i o t o t h e wave f o r c e i s :

ma (VII-1)
V<l-A2)2 + vV
Fig. VII-2 Block diagram of dynamic system of restraint. where:
X = U)/(D
This diagram represents a possible control system for a vessel
tu = frequency of force excitation
moored in waves. The waves exert forces on the vessel which in
U = natural frequency of the horizontal motion of the moored
turn cause motions which contain wave frequencies and low frequen­
vessel = /c/ni
cies due to the drift forces. The motions are measured and fed to
c = stiffness of the mooring system
the control unit which filters out the high frequency part of the
m = virtual mass of the vessel
motion signal and gives only low frequency force commands to the
v = non-dimensional damping factor = b//cm
servo system which exerts the required force on the vessel. In some
b = damping coefficient.

107
The amplitude response function is shown in Figure VII-4.
V= 0.063
Fig. VII-4 Amplitude response of measured force.
From the figure we can see that at low frequencies of excitation
the measured force equals the wave force while for frequencies
above the natural frequency the measured force becomes progres­
sively smaller in ratio to the wave forces. In between these
regions considerable dynamic magnification is evident. We now
consider the case that the vessel is moored in irregular waves.
The wave spectrum and the spectrum of the low frequency second
order forces are shown schematically in Figure VII-5. Superim­
posed on these spectra is the amplitude response function of
the measured force (mooring force) in ratio to the wave forces.
As can be seen from this figure, if the stiffness of the mooring
system is sufficiently large so that the peak of the response func­
tion is between the frequencies of the second order forces and the
wave frequencies, the dynamic magnification of the measured force
is small in the range of frequencies of the low frequency second
order forces. At the same time the ratio of measured or mooring
force to wave force in the range of the first order motions and
i nq
first order wave frequencies can be small. This means that, al­
though a stiff system is employed, the influence of the mooring
system on the first order motions, although present, need not af­
fect appreciably the first order motions.
WAVE SPECTRUM
SPECTRUM OF LOW FREQUENCY
SECOND ORDER FORCE
A M P L I T U D E RESPONSE OF MEASURED FORCE
LO
X=
u>e
Fig. VII-5 Schematic representation of wave spectrum, drift force
spectrum and amplitude response of the measuring sys­
tem.
It will be evident that the applicability of such a system of re­
straint becomes greater in irregular waves with a narrow spectrum
and with decreasing mean periods. In such cases the separation
between the high frequency wave spectrum and low frequency force
spectrum becomes larger so that the influence of the peak of the
response function of the measured force is reduced* In that case
both the dynamic magnification of the measured low frequency force
and the effect of the mooring system on the first order motions
are reduced. A drawback in this system is that, although the first
I in
order motions at wave frequencies may be relatively unaffected,
the measured force F (t) will nevertheless contain force components
with wave frequencies which are commensurate with the first order
motions and the mooring stiffness. These wave frequency force com­
ponents can be large relative to the second order forces. Before
the results of such measurements are analyzed further the high fre­
quency force components are filtered out of the force signals.
VII. 2.2. Realizations of two systems of restraint
In the aforegoing the basic principles of two systems of
restraint of a vessel have been discussed. The experimental data
on the low frequency second order horizontal forces acting on the
tanker were obtained using the dynamic system of restraint. The
vessel was positioned in the horizontal plane by means of three
servo units arranged as shown in Figure VII-6.
WAVES 180
FORCE
TRANSDUCER
SERVO SYSTEM
Fig. VII-6 Dynamic system of restraint.
One servo unit was used to control the surge motion while two
additional units were used to control the sway and yaw motion.
Each servo unit could exert forces on the model through the light­
weight metal rod connecting the model to the servo units. Force
transducers, from which the required forces were obtained, were
incorporated in the connecting rods.
Basically, each servo unit worked independently as a posi­
tion control system with characteristics based on feed-back of the
motions of the vessel in the direction of the axis of the connect­
ing rods. The motion and motion velocity were measured within each
of the units and fed to analog control units with proportional-
differential type characteristics. The output of these units gov­
erned the forces applied by the servo units on the vessel. The
proportional-differential characteristics were so adjusted that
the control systems applied small restoring forces and heavy damp­
ing to the vessel motions for the low motion frequencies. Through
the use of a low-pass filter on the motion velocities the damping
characteristics were filtered out for higher (wave) frequencies.
The proportional part of the control was independent of the wave
frequency.
The final settings chosen for the control systems allowed
determination of low frequency wave forces on the vessel with fre­
quencies up to approximately 0.05 rad./sec. full scale. Above this
frequency dynamic magnification effects became evident. This was
due to phase lag introduced by analog filtering of the velocity
feed-back which introduced a resonance peak in the response of the
system at a frequency of about 0.11 rad./sec. At higher frequen­
cies than 0.11 rad./sec. the forces due to the system of restraint
rapidly reduced until only the relatively weak proportional re­
storing forces remained in the range of normal wave frequencies.
At these frequencies the effect of the system of restraint on the
first order motions was negligible.
The system of restraint used for the tanker did not fully
conform with the requirements discussed in section 2 of this chap­
ter. The discrepancies are evident from the relatively narrow band
(0 - 0.05 rad./sec.) of the low frequencies for which the low fre­
quency second order wave forces may be measured without apprecia­
ble dynamic magnifications. This problem will probably be reduced
11?
by the application of more sophisticated control systems than used
for this investigation. From the point of view of practicality of

O O O
results obtainable by the present system it can be stated that,
since the natural frequencies of the horizontal motions of a moored WAVES 180
tanker are generally lower than 0.05 rad./sec., it is possible to
SPRING
measure accurately those low frequency horizontal wave forces which
l~^m^=r- —c^pJWW-!
are important for the low frequency horizontal behaviour of such
vessels. FORCE TRANSDUCER

In the aforegoing the feed-back control characteristics of O O O

the dynamic system of restraint used for the model tests with the
tanker are discussed. In order to enhance the position keeping
characteri'stics of the system in the range of the low frequency
horizontal motions an additional control signal, based on the in­
stantaneous relative wave elevation measured at a number of points
around the tanker model, was generated and used as an additional
command signal for the servo units. This control signal consti­
tuted an approximation for the instantaneous value of the low fre­
quency second order horizontal wave forces on the vessel as derived
from instantaneous evaluation of equation (IV-1). The effect of
including this additional control signal will be discussed in chap­
ter VIII in connection with dynamic positioning of a vessel at sea
and will not be treated further here.
Fig. VII-7 Passive system of restraint.
In order to show the influence of the stiffness of the moor­
ing system on the wave frequency motions of the semi-submersible
the amplitude response functions of the heave, surge and pitch mo­
tions in head waves were computed for the stiff mooring system and
compared with the results computed for the free floating case, i.e
for the case with zero mooring stiffness. The motion response func­
tions are compared in Figure VII-8.
FREE-FLOATING
IN STIFF MOORING SYSTEM

As stated previously the model tests with the semi-submers­ SURGE HEAVE PITCH
ible were carried out using a passive system of restraint based on
stiff linear spring characteristics. The model tests were carried
out in head waves only, so that only in the longitudinal direction I
I
1.0 10
a stiff spring system of restraint was used. The set-up is shown
in Figure VII-7. The system of restraint consisted of a forward la
..ti) kt
and aft mooring line, each incorporating a force transducer and a
0.5 0.5
linear spring. The restoring force of the mooring system in the
longitudinal or surge direction amounted to 513 tf per metre dis­ \

Viz\
placement for full scale. From a surge motion decay test in still
water the natural frequency of the surge motion as induced by the 0
1
system of restraint (mooring system) amounted to 0.4 rad./sec. (ii in r a d sec-
full scale.
Fig. VII-8 Influence of mooring system on the amplitudes of the
motions of the semi-submersible in regular head waves

i i i 11 A
From this figure it is seen that the heave and pitch motions are
practically unaffected by the stiffness of the mooring system. At
lower wave frequencies the surge response is increased due to the
stiffness of the mooring system. This indicates that from this
point of view the stiffness of the mooring was somewhat too great.
It will be seen, however, that the second order wave exciting
forces are practically unaffected. Computations of the mean second
order forces in regular head waves on the free floating semi-sub­
mersible were presented in chapter VI. For some wave frequencies
the computations were repeated taking into account the stiffness
of the mooring system in the longitudinal direction.
In Figure VII-9 the mean longitudinal force in regular waves
is compared for the stiff mooring system and the free floating ves­
sel.
The results show that the influence of the stiff mooring system on
the mean second order force is small. This result seems to be in
contradiction with the statement made previously that if the first
order motions are affected the second order forces will be affected
as well. In the case of the semi-submersible, however, the force
level at the lower wave frequencies is already small indicating
that the vessel is only causing slight diffraction of the incoming
waves due to the small size of the members, such as the columns
relative to the wave length at these frequencies. In such cases the
disturbance created by the vessel motions also contributes little
to the forces in which cases, even though the motions increase, the
second order forces still remain small. On the basis of this result
it is concluded that also the low frequency forces will be only
slightly affected by the stiff mooring system.
YI_=i3i_Model_tests

o N STIFF MOOR ING SYSTEM VII.3.1. Generation of waves

I
fi
For both vessels model tests were carried out in the Wave
and Current Laboratory of the Netherlands Ship Model Basin. This
basin measures 60 m by 40 m with a variable water depth from 0 to

I
-20 \
1.10 m. Wave generators of the fixed stroke, variable frequency
type are disposed on two sides of the basin; see Figure VII-10.
For the tests only the wave generators on the short side of the
basin were used.
(NJ cg
w cq
i^_fi 3 3 0 0 m (FULL SCALE).

-10

A, J (A) in rad.sec.-1
MOTOR
CD-,

Fig. VII-9 Influence of mooring system on the mean longitudinal

drift force on the semi-submersible in regular head
Fig. VII-10 Wave and Current Basin
waves.

115 116
­_*§9!_ i __:­_■ .wave _p roups
Since the wave generators are of the fixed stroke, variable fre­
quency type it is not possible to generate regular wave groups in
a straightforward way. This would require wave generators of the
type which are variable both in frequency and stroke. In order to 0) 1 - 0 ) 2 = 0 . 0 2 5 r a d . s e e r-11 O ) ^ 0 . 5 6 0 r a d . s e e-1
r ' U ) 2 = 0 . 5 3 5 r a d . s e c-:1
generate regular wave groups consisting of two regular waves with
small difference frequency the bank of wave generators was split
into two independently driven sections denoted by A and B in Fig­
ure VII­10. By driving section A at a constant frequency to produce
regular waves of frequency 10 and section B to produce regular
waves of frequency (_.„ regular wave groups are created at the common
edge between these two wave fields. This is shown in Figure VII­10.
The frequency of the wave groups is equal to the difference fre­ _1 CU = 0 . 3 5 7 rad.sec -1
00.pO.382 rad.sec.-' 2
quency to. ­ w . The width of the overlap region between the two
fields of regular waves increases with the distance from the wave
generators. For the tests in regular wave groups the tanker model
was situated as indicated in Figure VII­10. Examples of the wave
elevation in wave groups measured at the location of the tanker
model are given in Figure V I I ­ 1 1 . Due to the method of generating
wave groups these were not long­crested.

CO- - U)2= 0 . 0 5 0 r a d . s e c r-1

1
r ' CU2= 0.624 rad.sec.-1
0^ = 0 . 6 7 4 ^ . s e e -1
Irregular_wayes
Irregular waves are generated by varying in a random manner the
frequency of the wave generators at a fixed stroke of the wave pad­
dles. The stroke of the wave paddles and the variations of the fre­
quency are chosen so that irregular waves are generated which con­
form with a given spectral density distribution or wave spectrum.
The irregular waves were in all cases long­crested. For the tests r 1 0J2= 0 . 4 4 6 r a d . s e c_1
OJ^ = 0 . 4 9 6 r a d . s e c-1 :
with the tanker four w a v e spectra were used. These are shown in
Figure V I I ­ 1 2 . For the tests with the semi­submersible only one
wave spectrum was used. This spectrum is shown in Figure VII­13.

I L
0 50 100 sec

Fig. VII-11 R e g u l a r wave groups

 ­. & rt rt 0 OJ Hi O 3 rt 0> Hi « Hi Hi Hi Q O 3 a tn rt H s <P rt­ 1 i
J rt tr H Hi n O H fl) tr O n H 0 0 0 P­ Oi ro OJ P­ er DJ n ro
_l OJ 0> n H O 3 fl) 3 H 0J ro H H 3 ro P> H H 3 ro fD
< 0 tn o P ft H
H rt W ft p­ P1 »P O tn cn tn ^< fD rt 0J Hi tr ro
z0> ri­
tr
P
t. Hi cn H 3 rt ro P fD ft p­ rt ti­ CO
p
TJ p. rt ro o
<: rt OJ p P ii i O fD P­ cr ro cn 3* PJ O 3 0 a n ro Pi H fl> 0
CD fD 3 P CD cr 0> Xi 3 ro Hi 3 fl) ro Hi P P 0 tn rt Hi c P­ H Z f
Ul OJ rt •< tf p­ 3 c p 0 n Z H cn 0> ro H H rt 3" H H ^»d fD 0J P
Hi rf H fD P i 3 p­ g CO ii p­ fD 3 ro p ­ rt 3 er H tn ro fD 0> 0 Pi ft < fD
H K cn 0 CD tn O H ro ro H O a 3t i ­ 0J rt ro fD tf rt H H ro a
CD Cb tn rt Hi 3 TJ ti­ ro 3 3 rt cn ro H P P1 0 w Z ti­ P P­ O 0) ^
tf C fl> O fD p. p p­ tr B P ro O ­< p* 13 p­ ro ro O rt P 3 U3
c H tn p, ft P­ O er 5 i ro 0 OJ OJ M 3 N rt 0 0 rt w 3 3 er ft tn ro
ro •rr (D rt fD rt o* cn Hi 3 ro rt ro tr 3 er rf o ro • P­
Cu in rad.sec." OJ P 3 H 3 3
3 rt p­ ii ft) cn H fl) P rt n» p­ cn 01 ro ip a K O fD fD
n P­ tn o> 0> 3 cn H & Oi P­ h­> cn 3 rt p­ ti­ p­ rt Pi 0 rt 3 H
•<; 0 rt ii rt P­ O rt 3 3 3 *< »« P­ • 3 er H Pi 3 er P H ro rt fii
3 0) P­ CD fD rf O ro \K rt EM fD 3 ro 0 C ip ro H o H tn ft
3 M tn 3 3 Hi n ro fD ro n s: cn c & c ro rt 3
Fig. VII-12 Spectra of irregular waves for tests with the tanker. OJ
rt 0
3 a fl> H­
P.
Ui H H pi rt 3 cr
0J
ro
H
P to ft H tn tn 0
O

*< O TJ
C
P 1
p­ rt rt
P
H 0 p­ fD •i fD O Hi ft zp ­ ro P­ P­ 13 rt
*i
tn
0 H 0 rt 3 tr OJ Ui tr 0> cn cr OJ 3 3 ro cn H P­ rt 3 0 3 O Z p­
cr ii fD 3 cn tf 0 0 rt O p­ 3 rf *< M g »i p­ fD O 3­ p­ 3 ip 3 p­ O OJ
ro H, tn pi PJ P­ P­ ro Pi § Hi rt ip tf 3 i Pi rt 3 H
rcno fD Ti­ p­ rt M 3 O H • H 3 0i Q* P­ Oi 3 P CTi cn n r t ro er ro
cn 3 er to O 3 P* Hi QJ fD 3 ro H p­ P ro O 0 er a r t cr
*_• P 13 ft fD O cr 0> PI H 3 PJ OJ H to 3 M 3 Hi tr & H ro ro tn
Hi 0 0 N 3 p tr rt rt 0 ip 3 H rt cn 0 0 3 K rt­ r t er tr C
^ w l / 3 = 3.31m Hi 3 cn OJ O fD fD p­ 3 ro tn K O 3 p­ ft P ro ro O ro 0J p­
P­ PJ rt 0 3 ii rt O n O Z Pi tn Hi O c tr ro er H H cn ro < rt
. f =8.1 sec. P P­ p­ o 0 O 3" O p 3 p­ N 0 p­ H § vy tn ro cn cn TJ rt
tn 0J p­ P
P­ 3 p* c fD fD 3 3 cn P1 ro Hi Hi tn 0J Pi & p­ O rt ft P 0 cr
fl> ip H H Pi C n 0 M p, n • tr ro ro B O rt­ Hi cr1 3 cn 3 p ro
u 3 03 3 p­ P* H m p­ 0 0 0 0 ii H ro Hi ro d P a H rK H PJ
0) rt rt 1
fD M cn P tn H fD
lA P 0 O P cr 3 H P3
tr
c ro Z fl) ro tr
cy. • 0 OJ *< P­ H z & p­ fD P O rt g O 3 rt rt H Pi P­ P H 0
0 • fD 3 K TJ N tn fD 0 Hi cn tr tn P­ ti­ tn P 3
i­3 OJ ?r 0 tr Hi Pi Hi fD o ro cn co rt ro to 3 zp ­ er rt p­ PJ •»•
tr tr p­ Hi !* fD H p­ tf H rt M l M o p­ 0 0 p­ p 3
fD 0 3 3 fD ft 3 P fD O 0 tr O O cn Hi Ul 3 0J P­ rt ti­ 00 Pi
p <P H O tf tf tr P­ P­ P p­ z O 3 Q fD P" H 3­ er H ro
rt rt . O Z C P OJ rt H H o P i cn cn p­ cn 0 P­ ro ^ ro 0 ro o
fD tfl H O fD rt fD P •p 'p cn Hi 0 H 13 0 H • ro OJ tn ^ 0J
cn p" D P fO Ui 3 • a n­ "P H p­ P M ro 3 ii <P .t_n rt p c ^<
rt N) 0) M & rt O n p­ rt ro P p> rt 0J o PJ ro p P p­ H ro
O H rt IP P­ *< fD W 0J OJ ti O O ^P pj rt rt ip p­ 3 H OJ a
P­ O N tn fD O cn P­ ft rt 0) d­ P p Z p­ H O p Oi er ?r H H
P fD 3 tn e 3 Pi o H ro rt P 0 OJ li pi H 0 fD P 3
H. t pJ O 0 0J (D H O N P­ • • P 3 p­ < 3 H OJ P c H rt Z fD
OJ P­ H. Hi cr o rt (D P 3 \cn ot n M O O ro cn fD H « H p­ OJ 0*

y
rt P< P­ O M H ­< 3 3 H D» to p­ 0 < tn
p­ Ci i—i P ft ro 3 ft 0 tn ro 3 cn Hi 0) ro z < 3 3 ro P
0 o < H, P" P­ O tr P 0 H (D tn H rt ti­ Hi 0J ro Hi tn H
3 o H O fD 0J cr p­ Hi p ■ 0J rt ro P ro er 0 < tn c P­ ip fD
H tn CO O rt tn H 3^ Pi 3" o H .p rt O H ro pi H OJ H 3
0.5 1.0 Hi 0 1 M P­ H 0) fl) Hi O 0 fD P 3" a O cn H p* H rt O ro
U) in rad.sec.-1 0 Ul 1—
i i
i 3 Hi Pi p­ h »P n C \cn PcnJ 3 fD ro cn ro 0J
cn Ui
fD P 3
rt
H 0 cr Hi 0> CD 3 p P OJ H Pi CO 3 M p­ iQ rt 13
H­ p­ H M H fD tf fD cn P1 ro • P O Hi ti­ • cn ti­ 0 P er tn cn
rt M cn i P a C 3 fD n O Hi P­ Her ro 0J M ro
3" P1 P cn (D 0) Hi P­ P in tn . >­3 H Hi ro ro ro H 3 PJ H OJ rt *£
Fig. VII-13 Spectrum of irregular waves for test with the semi- fD OJ ip 13 3 Ui O cr H ^ >» p cr & P­ in tf cn 3 ro fD H € cr ro
rt tf ro O >i *< fD > P Hi ro ro n c 0 P • 0J ro H
p­ fD O CD rt O 1 p­ M C p­
H OJ fD p­ O ro <: ro
submersible.' 0 cn i
rt
3" fD
tn
Hi ro P1 fD
3
H 3
1
ip
l
H
1
cin ro
3
Ul cn
O M
rt
ro
tests in irregular waves with the tanker corresponds to 950 to
1500 oscillations at wave frequency depending on the mean period
of the irregular waves. For the test with the semi-submersible the
test duration corresponded with about 2700 oscillations at wave
frequency.
YIII_4i_Analysis_of_results_of measurement s_of_the_low__freguency
VII.4.1. Regular wave groups
In Figure VII-14 some typical results of the low frequency
longitudinal force on the tanker in regular wave groups are shown.
In the same figure the corresponding wave elevation is also shown.
100 tf r
FORCE ^/WVVvw^ x v A\./\r
-100tf
100sec.
Fig. VII-14 Longitudinal drift force on the tanker in regular
wave group.
The time record of the longitudinal force contains a constant part
corresponding to the sum of the mean second order force due to
121
each of the regular wave components and a low frequency oscillatory
part arising from the combined action of the regular wave compo­
nents; see equation (IV-49). From the results of measurements the
quadratic transfer fiinction of the amplitude of the low frequency
force T,- is found by simply dividing the amplitude of the measured
a r e t h e am
low frequency force by 2C, C 2 » where £. and £* "
plitudes of the regular wave components.
VII.4.2. Irregular waves
In Figure VII-15 typical results on the low frequency com­
ponents of the longitudinal force in irregular waves on the tanker
and the semi-submersible are given with the corresponding wave
elevations. The time records of the longitudinal forces contain a
constant part corresponding to equation (IV-58) and low frequency
components.
In order to obtain results on the quadratic transfer func­
tion for the low frequency longitudinal forces cross-bi-spectral
analysis was applied based on methods developed by Dalzell [VII-1].
Due to the specialized nature a full discussion on cross-bi-spec­
tral methods is outside the scope of this work. For this we refer
to the above mentioned author. In Appendix B a brief discussion on
the method and some details on the analyses are given. The model
test results obtained for the tanker were analyzed in full accor­
dance with Dalzell's method. The results of the model tests with
the semi-submersible were analyzed using a slightly modified ver­
sion of Dalzell's method. The modification is discussed in Appen­
dix B.
YIl_L{ii_Com]2Utatigns
Computations of the quadratic transfer function of the am­
plitude of the longitudinal force in head waves were carried out
in accordance with the theory set forth in chapter II through
chapter IV.
For the tanker and the semi-submersible the quadratic trans­
fer functions are given in Table VII-1 and Table VII-2 respectively.
122
 The transfer functions are given in matrix form of which the two
TANKER axes represent the two frequency components of a regular wave group,
FORCE The data given in these tables represent the amplitude of the low
frequency forces as computed based on equation (IV-50). The data
on the diagonal {ta - &__) represent the amplitudes of the forces
for zero difference frequency which, except for the sign, corre­
spond to the mean drift force in regular waves. Since the transfer
functions are symmetrical about the diagonal, values are only given
for IM > to-. The results give the quadratic transfer function for
the force in tf/m to a base of wave frequency for the full scale.
WAVE ^w1/ =10.3m f = 13.3sec.
As can be seen from these tables the quadratic transfer
function for the tanker has been computed for frequency combina­
tions of which the smallest difference frequency w - co- is greater
than 0.05 rad./sec. The results obtained from model tests apply to
frequencies of 0 rad./sec., 0.025 rad./sec. and 0.05 rad./sec. re­
100sec. spectively. In order to be able to compare results of computations
-t
with experimental results the computed data were cross-faired and
interpolated at the difference frequencies of 0.025 rad./sec. and
0.05 rad./sec. respectively. For a difference frequency of 0 rad./
SEMI-SUBMERSIBLE
sec. no problem exists since at this frequency computed data are
FORCE also available. These are the computed data on the diagonal of the
matrix of the quadratic transfer functions.

0.354 0.444 0.523 0.600 0.713 0.803 0.887

"2 ^ \

0.354 2.0 8.7 10.4 24.5 10.0 38.4 37.5

WAVE . w 1 ^ = 3.3m T =8.1 sec. 0.444 7.0 20.8 19.4 25.7 12.1 35.2

0.523 12.4 16.4 8.3 14.3 14.2

0.600 14.0 9.5 18.0 14.9

0.713 2 8.6 4.3 6.7

0.80 3 T in tf/m 9.2 4.7

Frequencies in rad./ sec.
100sec. 0.887 8.6
-t

Table VII-1 Quadratic transfer fiinction of longitudinal force on

Fig. VII-15 Longitudinal drift forces in irregular head waves. the tanker in head waves.

123
 0.5 0.6 0.7 0.8 0.9 1.0 1.1
"2 ^ \
COMPUTED
0.5 0 8 11 7 8 20 15 A • ■ REGULAR WAVES
0.6 1 6 7 5 14 19 IRREGULAR WAVES: £ w 1 / 3 = 2 . 7 m ; T= 8.2sec.
0.7 0 7 15 3 18
= 4.8m; f = 10.1 sec.
= 7.3m ; T= 12.0sec.
0.8 11 23 20 13
= 10.3m ; f = 13.3sec.
0.9 25 21 9
2
T 1 2 in tf/m 30.0
1.0 20 21
Frequencies in rad /sec.
1.1 24

Table VII-2 Quadratic transfer function of longitudinal force on

the semi-submersible in head waves.

For the semi-submersible the experimental data obtained

from cross-bi-spectral analysis are valid for difference frequen­
cies of 0 rad./sec. and 0.1 rad./sec. The data for these frequen­ ."•J
cies are found on the diagonal and the first row next to the diag­
onal of the computed data given in Table VII-2. 15.0
c
YII__£i_Q°^E__Ei§2£_hetween_computati

The quadratic transfer functions for the low frequency lon­

gitudinal force in head waves on the tanker as obtained from compu­
tations and experiments in regular wave groups and irregular wave
groups are compared in Figures VII-16(a), (b) and (c). In Figures
VII-17(a) and (b) the computed data and experimental data for the
semi-submersible in irregular head waves are compared. For each
value of the difference frequency the data are given in one figure
In each figure the data are given to a base of the mean frequency
of the regular wave components.
U)1 ♦ 0>2 1
in rad.sec­1
The experimental data for the tanker obtained by cross-bi-
spectral analysis of data from tests in irregular waves show a
rather irregular character. This is ascribed to the parameter set­ Fig. VII­16(a) Quadratic transfer function of the low frequency
tings used during the cross-bi-spectral analysis. longitudinal drift force on the tanker.

126
 COMPUTED COMPUTED

REGULAR WAVE GROUPS REGULAR WAVE GROUPS

IRREGULAR WAVES: £ w 1 / = 2.7m t= 8.2 sec. IRREGULAR WAVES: twy3 = 2 . 7 m ; f = 8 . 2 sec

= 4.8 m f = 10.1 sec. = 4.8 m ; f =10.1 sec.

= 7.3m f =12.0sec = 7.3 m ; T =12.0 sec.
f =13.3 sec. = 10.3 m ; f =13.3 sec.
--10.3m
30.0 30.0

U),, - 0J2 =0.025 rad.sec- 1

OJ

15.0
c

i
C\J
/

f}^\
•
•

o 0.5 1.0

— L _ — £ _ in rad.sec.-1 Oh ♦ 0J2 ­
—~——— in rad.sec.­'
2

Fig. VII-16(b) Quadratic transfer function of the low frequency Fig. VII­16(c) Quadratic transfer function of the low frequency
longitudinal drift force on the tanker. longitudinal drift force on the tanker.

127 128
 -40
By, for instance, changing the filter settings for these computa­
tions a smoother set of data could be obtained. Essentially, the
results would, however, be the same. The tanker data also show that
the scattering of the experimental data from tests in irregular
waves is less for difference frequencies of 0.025 rad./sec. and
0.05 rad./sec. than for 0 rad./sec. This is probably related to the
fact that the number of oscillations in the low frequency force
(a) components with frequencies tending to zero also become zero. This
will tend to decrease the reliability of the results of the cross­
bi­spectral analysis.

Generally, the experimental data for the tanker from the

different tests in irregular waves show reasonable correlation con­
sidering the complexity in both the test set­up and the method of
analysis. It should also be remembered that the level of the actual
0.5 1.0
(A) in rad.sec. - 1 force varies considerably between the tests in the lowest and the
highest irregular waves. Since the low frequency forces vary qua­
dratically with the wave height the force level in the highest
-40 irregular sea state is about ten times larger than in the lowest
COMPUTED (TOTAL)
FROM C.B.S. (TOTAL) sea state. Bearing this in mind it can be concluded that the re­
(2)
COMPUTED WITHOUT $ sults obtained from the various tests in irregular waves on the
U)-, - U ) 2 = 0.1 rad.secr1 quadratic transfer function agree reasonably well. The experimental
-30
data obtained for the semi­submersible are smoother than those
obtained for the tanker. This is in part due to the increased test
OJ
E duration and in part due to the different parameters used during
^ -20 (b) cross­bi­spectral analysis of these results (see Appendix B).
c
oi In general the experimentally obtained data from tests
-10 in irregular waves compare reasonably well with the computed data
for both the tanker and the semi­submersible. The data from tests
with the tanker in regular wave groups are somewhat lower than
computed data and data from tests in irregular waves. This may be
0 due to the method used to generate the regular wave groups in the
0.5 1.0
U)-i + tx)p i basin. The wave groups were only realized in a relatively narrow
LO = .—! *L_ m rad.secr1 field as indicated in Figure VII­10 while the computations assume
2 that these will be long­crested. As a result of this it is possible
Fig, VII-17(a) and 17(b) Quadratic transfer function of the low that the corresponding second order forces were not fully developed
frequency longitudinal drift force on during the experiments.
the semi-submersible.

129 1 ■_/­!
 The computed data for the tanker for difference frequencies
M
of 0.025 rad./sec. and 0.05 rad./sec. given in Figures VII-16(b) \ 1 0.5 0.6 0.7 0.8 0.9 1.0 1. 1
and 16(c) show a sharp increase for lower values of the mean wave *2 \ ^
frequency. No experimental data are available at these frequencies 0.5 0 7 12 11 3 19 12
to confirm this trend. Examination of the contributions due to the
0.6 0 6 10 6 12 15
five components to the force given in equations (IV-1) through
(IV-5) shows that this effect is due to component V which is caused 0.7 0 6 9 0 15
by the non-linear second order potential. 0.8 0 6 8 6

0.9 2 0 7 6
Examination of the computed data for the semi-submersible T in tf/m
reveals that also here some influence of the non-linear second or­ 1.0 0 7
Frequencies in rad /sec.
der potential is found. In this case for a difference frequency of 1. 1 0
0.1 rad./sec. the experimental data appear to confirm the existence
of the contribution due to the non-linear second order potential.
Table VII-4 Contribution of second order potential to the qua­
In Table VII-3 the computed quadratic transfer function for dratic transfer function for the longitudinal force
the semi-submersible is given without the influence of component V. on the semi-submersible in head waves.
The contribution due to component V is given in Table VII-4.
Comparison of the results given in these tables and the
total given in Table VII-2 shows that for lower values of the fre­
quency of the regular wave components W and u>? the total low fre­
0.5 0.6 0.7 0.8 0.9 1,0 1. 1 quency force at difference frequencies greater than zero are domi­
U
2 ^ \
nated by component V. At higher wave frequencies components I
0.5 0 1 1 5 6 3 7 through IV tend to dominate the results. These components are due
0.6 1 1 3 6 5 12 to products of first order quantities which become large when ship
motions increase and/or first order diffraction effects increase.
0.7 0 3 10 3 8
When components I through IV dominate the results the first of
0.8 11 21 18 8 these components generally is the largest of these, as was already
0.9 25 21 11 indicated in chapter VI.
T 1 2 in tf/m 2
1.0 20 20
Frequencies in rad /sec. VII^7^_A£Er^ximation_for_the_low_freg
1.1 24
In this chapter some results have been given on the low fre­
quency second order longitudinal force in head waves. The results
Table VII-3 Quadratic transfer function of longitudinal force on
of computations will be used in a discussion concerning a method
the semi-submersible in head waves without contribu­
for approximating the low frequency components of the second order
tion due to second order potential.
forces in irregular waves.
 It is generally believed that the low frequency components
of the second order forces in irregular waves can be predicted
using the mean forces in regular waves only; see ref. [1-2], [1-3],
[1-10], [1-14] and [VII-2]. Within the framework of this study this
assumption implies that in a regular wave group with frequency com­
ponents u>, and a) the correct amplitude T - of the low frequency
component of the second order forces can be replaced by the value
for the amplitude found on the diagonal of the matrix of the qua­
dratic transfer function at the mean value of the two frequencies.
Thus :

ux + ux to. + ux
T = T 1
12 ^ 1 ,to 2 ) a T( ,2 , 2 *) (VII-2)
rt 3 H* Hi pj p­ tn 3 p­ cn cr P­ rt 0 p­ p rt tn tr Hi tn tn tr ti­ rt OJ ti­ 3 Hi P tf
Cr 03 •< p H 3 P ro 3 ro ro 3 p­ ii 3 Pi cr O ro 0 CD ro & er OJ ft er O H cn P
We may check the assumption for the case of the tanker and the ro 1 rH­t 3 ro P n P> o rt O 3 ro 3 H P & 0 p- rt ro rt CD p- CD
P P­ p­ 3 H O rt 3 3 p­ ro HCD O • 3 P p co p- Z P- tf 3 3
g N o rt tf P cr fD 3 h tr O ii tr 3 CD ro rt rt 0 0) pi O P tf o
semi-submersible by inspection of Figures VII-16(a), (b) and (c) ro ro 0*1 p­ p­ P< ro Hi tf a P ro < a H ro ro H 0J M H H 3 3 < Q 3 fl) K
0) *t I­ 0 < ro p. 0 P ro ro ro pi ro O ro ri­ Hi O ro ro 3p - cn 3 0
Ul tr CO p­ . P- 3 cn
and Figures VII-17(a) and (b). tn H O
■
3 0 3 ro H Pi 0 H Pi >­3 tf Ptn 3 O PJ
. •i 3 rt p­ P 0i M Q P er p rt CD 3 TJ er ro o 3 tn Hi 3 O P- P1fl)
Hi ro . >­3 3­ 3 ro H P J *4 Z ro 0 ro P 1 ro cn H PJ ro p i tf NJ tn H Pi Hi ro *< n
0 *i tn p " p­ ro ro O Hi OJ Qi ro P- • O 0 ro p (Jl pj ro 3 cn r t 03
H. H P NJ 3 ft Z cr I­i H Hi rt to H O P rt rt ro rt P tf rt ti­
3* ft H er P- ti­ 3 H er CD P er *-. er a
Equation (VII-2) implies that in Figures VII-16(a), (b) and P
ro
0
3
H
rt Hi P_
h­>
o P­
Pi
tn
fD
Qj
er ro
H
CD
0 ti­p)
er 3 Z
rt
O
0
3 P- rt ro fD
< er P p P» ro TJ ro r t ro
tn 0 0J Z Vi 3 0 X tf ti­ ro ro DJ P. O n ro P- pi rt ft 3 P> *< 0
(c) the computed curves of T.- given for the difference frequencies p­ rt . H crJ ro n P er < T3 3 n
1
P in CD O n H
rt
3 13 3 ti
3 er P Hi Ul tf zpi p­ 3 ro ro tn pi ro H O tn pj rt ><* ro tn \ . ti­ p- O p- ro Pi
of 0.025 rad./sec. and 0.05 rad./sec. should be the same as the p­^ r t ro ii rt P N 13 3 ro tn tn CD ►i 3 ^ • cn er HP-i ro O n PJ ro
*1 cn cr tr ro P OJ < 0 H n tn 0 in . Pi a- tn TJ H i tf fl) ro cn H i
O H P 3 H
ro ro ro < tf Pi pi
ro 3 CD K ro O p P­ ro P cr ro H H H n H ro M
computed curve given for zero difference frequency. For mean fre­ tf H H P ■< cn rt ­ o 3 3 O H O ro tn ft ro ro • H cn
r t
Z
f t
pi M Hi Hi
P ro < Hi H ro rt . 0) Hi 0 a TJ rt er tf P 0 er ^< O O
quencies below 0.4 rad./sec. this is clearly not the case. At H
p)
tn
P
OJ
H
0
H n
1
Ul
3 cr
p­
O
CC
H O
tfP­
0
H
3
Pi O
rt
p­ (t
Hi
O OJ
n
0
0
Hj
3
pj
P
ro
ti­
ro cr
i~3
z 0
ro er < f t
H ro ro
H
P
H
n
H H P n • •< tn N ro Hi o H 0 tr H tn 3 K 3 n ro H i H CD CO M er fD ro
higher mean frequencies the discrepancies are less but nevertheless rt ro ro ro O M CD 0 Pi 3 ro P o tn P
rt
H Pi
f t
cn ro cn
* tn o Hi H H Hi H <! H ro ro Pi
p-
H C cr P- pi ro ro 13 ro 3 P-
increase for the larger value of the difference frequency. The dif­ Oi p­ p­ 3 0 ro O 0 O p­ OJ Pi H 0 H P h ro ro 3 - 03 tf H cr CD M 3 p-
3
< tt 0 tn H to . p ro ro ft ro PJ Q 0 cn Pi Hi tn P> c ro r tO tn PJ
ro P rt O P 3 H Hi 3 «: 3 0 CD PJ ro .3 p- ro pp-i *— r t H
ferences below a mean frequency of about 0.4 rad./sec. are mainly tn n rt N 3" ro p< G Pi 0 i—i
tr a 3
tn
P Pi O x P> 3 3 3 n P CD P-
& tn ro CD rt c n 3 M p­ 3 H O Hi cn r t ro Ti­ H P 0 Hi P PJ H tf H
caused by the force contribution V due to the second order non­ p 3 p­ »i p­ to p­ ft H tp O O rt H c cn ro ro & 3 p- H i LJ. __ i i Qi P H
Cb 0 Pi 3 3 er P> 1 3" 3 ro ­ CD H ro tr <! 0 O ft ro H O
3 CD Pi M CD
_ cr fD . P­
s: tf OJ to 1 • tf Hi p. ro ro rt P P tn 0) h O fu tf
linear potential. This contribution is zero in regular waves. 3 ro H ro
0J rt cn P1 3 H er P PJ H tn fD cr r t TJ o
ft
Hi 3 Hi n c1
0 ft Hro Htf rt i—>
ro P­ "£j 0 CD O er ro tn P< o P O TJ O H P
rt n tr pcr­ tf ro0 er Hi & 3 3^ Z 3 fD ro H i ro P o rHt H i H- OJ r t l i ­ ro s: 01
0 ro in P 3 ro o 3 a § ro ro CD P H O P1 f t rt *£fc O H
ft
cr f t tf P H
tr 3 H ^0J H. ro ro H i pi 3 fD
< *<: < ro H cn er P- 3 ro c <
In Figures VII-17(a) and (b) in which results are given for ro o PJ PJ P1
ro o H H H ­ CD .. ro P
rt
o H OH i t3i ­ p-
O ro
3
ro
cn
s:
pi p­ p H pi Q pj tf p ­ H H p tn
Hi rt
tn
p-
.p i ro 0 O 0)
p P o> 0 rt ro 3 I­i ro < ro 13 H i <■ 0 cn PJ ro Hi Hi H n <:
the semi-submersible it is seen that for mean frequencies below tn Cb tf 3 cr « Hp ­ 13 pj pi ro tf 0 0 3 ro O 3" 3 Hi
a f t H HCDi f t H P* p- 13 CD
ro ro 0 Hi ro Dt c P< P rt H cr TJ p" H P> p- 0 H • er ro ro QJ ro fi tn
CO rt p- \ ro rtnt O tr tf H tn
about 0.8 rad./sec. the correct value of T 2 for a difference fre­ PJ P< 3
0)
p­
H
<
tf fD ro
<! rt
OJ z
PJ 0
cr
01
ro CD
3
ro
3 P­
O
3 3 x P
ii 0)
PJ cn rt ro P tf
O
<J 0

\tn o z
rt ti­ pi 3 P PJ rt <! H 3 P rt 3 3 CD p- Oi H tt • ro Ui H ro o p- p>
quency of 0.1 rad./sec. would be underestimated if it were replaced 0 er Ul ot tf p­ ro P­ PJ P­ P­ cn 0) 3 3 3 rt cr pj CO • cn PJ • Hi Pi 3
0) 0) pi H PJ O N H CD pi ft Pi rt -< PJ O ro • < er fl) H CD
by the value found for zero difference frequency. At mean frequen­ 13 rt H o H O 3 3 Hi 0 K cn P> pi ro cn t i X O ro ro O »i 0 >t r t a cr
H CD tf 0> P Di H 3 rt 3 ro TJ p- 3 er o ^ tn Pp i 0 0 Z r er rt
ro
ro Hi <* p­ rt 13 3 ro rt p­ rt P­ P Hi O tf H 3 P 0 . 0 H 3 OJ ro
cies higher than 0.8 rad./sec. the differences are small. In Fig­ a 0 pi p. CO ro rro tf Pi o 3" CO (D O H. P PJ tf M H
P-
H 13 pi Pi
5 rt er TJ
p­ H tr < o • X t c H 0) p­ ^ ii H n P P ro rt CD er er 0) H
rt P rt P- m N 3 rt P 3 cr O tn ro 0p. rt ro
ure VII-17(b) computed results are also shown for the case that o
rt ti­
0 p­
ro r t
rf
CD
tr fD
3
M
P1
cn pi
er H p- 3 p- 0 Pi P- rt O ro H pj
er zro ­ H
t­3
o­ 3
0
pi O
O
•< ^< Z
0
3
P
0
PJ
P1 ro o rt 0 3 ro H rt PJ 3 CO p- rt p-
the contribution V is neglected. Comparison with the total force rt p­ < in PJ ro pj K P" p­ p­ 3 P1
z0J 01 O 3 rt o to
p rt fD P pi N er 0
tr tn ro 3 ro PJ ro P­ ft P1 3 ft Hi pi
cn 0) •CO 3 CD P H cn 3 0 ro ti­
ro o> ro cn er P1 tr < OJ M ro X 3 rt ro 3 ro
including this contribution shows that for mean frequencies higher p
H
3
tn
Hi
H
fD
pi
i
3
Pi ro
Pi
3
H
P­ 0
O
H ro o n « 3 to O X ro rt M Pi
M OJ OJ ro tn rt rt ro 3 rt tr a n CO pi n p- H p- er CD ro ><: cr rt P> Q
than about 0.8 rad./sec. the influence of this contribution is 0 tn i H P1 0 tr H 3 O
ci p­ ro cn
ro
0 pi P l 0
p
fl) ii O
p-
ro p- P1 «:
z ro P fD fD rt ro P" Pi Pi 3
rt tn P1 i
small. At these mean frequencies the low frequency force is domi- tn
P
3
CO
rt pi ro
 a method for approximating the low frequency forces in irregular
frequency components of the force in irregular waves.
waves based on knowledge of the mean forces in regular waves was
discussed. The results of this discussion indicate that for surface
vessels the low frequency horizontal forces which are of importance
from the point of view of the low frequency motions (i.e.: force
components with frequencies near the natural frequency of the hor­
izontal motions of a moored vessel) can be predicted with reason­
able accuracy based on knowledge of the mean forces in regular
waves. Certain conditions with respect to the natural frequency of
the horizontal motions and the dominant period of the irregular
waves must be satisfied however.
2 4 6 8 2k h
In the case of a submerged horizontal cylinder it was found
2 0.00 0.08 0.09 0.08 that in regular beam waves the mean horizontal force is equal to
zero. The low frequency force in regular wave groups was, however,
4 0.00 0.06 0.07
12 found to be non-zero, thus indicating that in some cases the mean
6 0.00 0.05 TfpgL
force in regular waves cannot be used to predict the low frequency
8 0.00 force in irregular waves.

2k
2h

Table VII-5 Amplitude of low frequency second order transverse

force in regular wave groups on a submerged cylinder
in beam waves.

VI 1^8 ^Conclusions

In this chapter aspects were discussed of the system of

restraint necessary to determine experimentally the low frequency
second order wave forces on vessels in waves. Two possible reali­
zations of such a system were used for model tests. For the fre­
quency ranges in which the system behaved as required, reasonable
correlation between results of computations and measurements was
found. It should be stated, however, that further development of
such systems of restraint are necessary in order to increase the
frequency range of application.

On the basis of results of computations on the mean and low

frequency second order wave exciting forces, the applicability of
VIII. APPLICATION OF THEORY TO DYNAMIC POSITIONING OF A VESSEL IN
IRREGULAR WAVES

VIIIil^_Introductiori

In the previous chapters it has been shown that the mean

and low frequency second order horizontal forces acting on floating
vessels are dominated by the contribution from the relative wave
elevation at the waterline of the vessel. In this section it will
be shown that this knowledge may be put to practical use to improve
the positioning accuracy of dynamically positioned vessels at sea.

Dynamic positioning or station keeping of vessels is a tech­ ­1 OJ 13 cn P­ Z rt rt H ti­ T5 Z Tt OJ H O 3 0 Hi ft a ro n *4 ft H p- tr rt Hi Hi

o 3 ro p­ H OJ cr H ro er P1 CD 0 ro o 0J Hi H P- o H 0 O P P Hi
ro p- H CD
nique which employs ship mounted propulsion units to counteract 0 PJ 3
Ul
tf
3
H < ro
fl) ro p ­
0
pi
fli
pi
ro P
tn
0)
ft
tn
p­
Z
p­
Hi 3
rt
p-
3 rt
ro
tf
O
3
3
rt
H
o
3
t+
H H
3
TJ
p- rt o
O
3
ro ro
tf pi i
P» OJ fli tf •< < p­ er rt 3 H ft er P H H P rt er O tn P
environmental forces due to wind, waves and current acting on the 0 rt P1 P pj 3 0 ro 0 ro P­ Pi — O 0J ro ro fD O p- er M
PJ
p- 3 ro tr
ro H CO 3 3 tn 3 H 0 1 P- 3 ii P1 tn 3 ro CD < Ul 13 Qi 3 01
vessel, thereby maintaining as closely as possible some desired z ft
P"
P> P­ rt i­3 Pi 0 tn 3 Hi >
<!
H
pi
3 o O H • p- to 0J 0J ro rt O 0
Hi rt 0 H Hi QJ er "< < ro P p ro Ul H tn 0 *< 0 tf P* ro Pi H Z 3 *< X
ii 3 p­ CD 3 0 rK H P­ rt 3 ti o 3 0 cn P- OJ CO Z
position in the horizontal plane. er
ro ro r t £ H i f t 10
ft
PJ
p»
P­ 3 ro
ro
pj tr 1 tf er r t o 3 QJ Z H OJ cn pi 0J 0 0
tf cr 0> Oi Hi rt PJ ro i PJ P 3 ro H p( t tn CD P< CD rt rt rt­ < 0 0
P 3 ro <3 0 3 a 0 H cn P­ Tl Hi tn i i flj O p- tn Hi QJ O P- 3 ro ro 3 3
ro ro ro H ro tr Hi 0 tt p­ o J­J
< M T3 tf O H cn O O pj 13 rt
3 QJ cn P 0 QJ Hi 3 0 P­ 3 Ij cr tn ro P 3 Hi tn fD O ro 3 rt Hi O H
The last decade has seen a steady increase in the number of 0 3 3" 0 ro p TJ 0 0 tf 5 ro tn tn ft PJ 0 tf 3 X cn fl) H 3 0
<< p­ 0 cn cn r t 13 3 c 3 0) ro X cn *< P1 ft P cn O t+ Hifl) fl) Pi

vessels which are stationed at sea by means of dynamic positioning OJ TJ

tr 3
3
a OJ 3 H
ro 0
CO
0
<
3
ft
cn
•
13
CD
H
Pi
3 o> ro
M H
tn
ti­
O
Hi
tn er t i ­
ro er
CD
3 ti­
ro p - er H i tf 3
cn 3 ro ro P r t
-1

0 Pi 3 p ­ 3 ro P ­ p­ p­ ro H tn 3 - ro p- QJ n er p -
cn n ro tn 0
systems. Up to now most dynamic positioning systems were used for H 0 rt Pi pi ft 3 n H Hi tn er fli tn 3 ti­ tn 3 ft p- CD ti­ n rpt- 3 . 0
p­ P* p P­ 3 in p­ tf PJ er 0 p­ 0 3 er OJ ro < er P n 3
positioning drilling ships in deep water where conventional anchor­ N
0
Q
*
3
ft
0
3
•<
QJ fl) o
o
3 p­
0
ft
p­
cn 3
H tf
3 z3 Hi
TJ
0
Hi
H
ro pj
c
I_J.
0
O
3
tn
.
O
O
CD ro tn < p - H
ro CD ro er
ft
0>
3 ro Ul *P­ 3 0 3 Pi PJ P­ CO O Hi fl) H P1 3 Z rt H cn ro H-
ing systems were considered to be too cumbersome. Nowadays dynamic ft Hi Pi • PJ 3 OJ tn ft p­ 3 P­ cr P­ p- 3 P- *< H rt ro cr t i ­ ^< 3
PJ H 3 n tn tt er tn P >< ti­ rt H rt TJ er H pj H er pi Hi
positioning is also being used for diving support vessels and main- P* ro
tf
13
H
P3
er
tf P­
0
ro
H Z
ro ro
OJ Pi
CD P­
3 m
ro
H
p-
o
ft ft
er ro r t
O H
Q
0
P1 p-
tn
O
P1
H P
tn
ro cr CD P-
^<
H
H
o
o
3 ro H tr Z rt cn f3
3 c 0 p­ 3 3 o pi p­ P I­J. Qi Hi
tainance and survey vessels. This increasing interest in dynamic 0 ro 13 cn 0 Hsi CD p 3 o 0 rt 0 rt • • er p * H i H c n
p-
3 H tn tf rt o
ft 3 C1 3 Pi pi
o tn 0 3 er & ro TJ ro ro CD & O fl) tpr- ro o
positioning systems stems from the need for a means of maintaining p­
0
n
•<
P
in
tn
p­
ro H
3 CD Z
pi Hi ft
pj
3
Ul
tn
P­
ro
ro
o
er
H
3
O
P-
H3
p- 0
3
M
H tf
CD C
tf
3 rt­
H i tf
P 13
3
ro H
O 3
ro
3 p­ tf f t tf p­ tr H 3 ro P i 0 3 3 P- 0 tf P- OJ ro f t 3CD H a 3
the vessel's positioning which is quick, accurate and versatile and in Hi 0 3 P ti­ ro ro rt tf fl) 0 Qi p­ O rt Ul 0 P- P H P1
CO
o> er 3 QJ ro f+
0 3 QJ P CD er pj P H 3 3 tf ii c 3 fl) ro H ro n Q pp ii H tn
does not interfere with systems, such as pipelines, lying on the 0 H P1 o 3 rt r_ S ro fli rt P. P PJ O TJ H < 3 in p-
Hi P P rt O rt er r t CD 3 tr H ro CD Hi er H 0 1 O H i ft O Tt
Hi
ro 3 ^ * Z
ro 3 O P­ KJ er ro p ­ 0 n M O _­• tn ii QJ ro P •< P- O H cn rt­ f t P p-
sea floor. ti­ cn p. Qi 3 ro 3 C ro •< P1 Pi
rt
< cn 3 < 0
P-
O
p-
ro < rt
er 3ft tf cn o tf CO V
tf Pi ro O ro P> 13 3
CD
TJ pi O ro cr
ro t i ­ tn ro n pi 0 tn ro t. O ro P 1 ro a 0 r t H C1 3 0
er
< P Z ro
tr 0
3
n flj
o TJ
CO Z
ro tr
Hi
0
Hi ro »<
tn cn
H
tn
CD
O z o> 0 p - 133 fl) H
O
13
P P
er tn
< TJ P 1 H i
O ii 0 zpj
The propulsion units used for dynamic positioning can be ro tn tr
tn p­ p OJ
3
Pi
o>
tr Z 3
ro li
0
ti­
er CD
TJ rt
CD ■—i
O
Hi
tr
rt
0
Hi
tf H i 3 0 P-
3 3
P 1 pi
P 3 O
p- M
<
O
TJ
pi
H
H
P
<
ro
cn H n tn p­ p­ CD ro o 3 < ti­ PJ H 13 Hi fl) tf cn P- 3 CD P tf ro
either fixed, tunnel mounted controllable pitch propellers, azi­ ro ro cr ro < 0 P1 rt rt p­ H ro p- p- CD tf P-
1
3 p- O PJ M ro tn Hi
H P< pi ro Hp­ 3* tr r t Hi A> P H 3 3 tt a OJ P f t O O 0J p to ii
muthing right-angle drive units with propellers externally fixed c Z cn P i r t ro er PJ P1 PJ H tn P cn rt­ tn Hi 3 pi 3
P- z P- P- Z CD
o p ­1 OJ cn CD ^< Hi P O P­ 3 1 CD tr o ro ro QJ (t
O 0 3 tr
p-
tf
to the vessel, or vertical axis propellers. The magnitude and di­ H p­ P cn ro n CD < M rt ^< NJ 3 p- QJ ro P 1 ipi- ti­ p p p 3 P
3 3 M M rt fD ro ro i—■ Tl tf tr cn H er 3 p tn pi
3 0 ro
rection of the thrust produced by such units are governed by a
tf QJ tr 0 Pi cn TJ ti­ P­ 3 h­* cr p- OJ 3 CD ro P-
M PJ p Pi er 3
Pi
3
QJ
0 0
p­
3
0
H
r
cr
cn
CD
H
ro
er 3 er p.
3
0
►< tf
P1
p-
ti­ tf tf
er p - t i ­ P •p
rt
cn
T.
O -s
er H
3
P-
tf
3 p-
p
p-

control system which has as input the position error of the vessel 0>
ti­
er rt 0 p­
PJ
ro PJ M <
fli
fi­ tf g P ro P n- ro CO 0 3 p- CD rt p-
ft
3 ro
P1 ro 3 Q> N ti­ 0 $ P 0 P­ a 0) *< 3 er CD cn 3 fD p tf tn PJ tn
0 ft ft H 0 ro *• tr 3 n­ tn ro P <« H tf ro P i rpt- 0) 3 er P • PP J C
relative to the required position and heading in the horizontal tf 3 0 H cr 3 ii pi ft er rt PJ p- rt P rt *< ft P- n pi
^< 0 p­ rt 3 0 tn P­ ro K ti­ P­ CO rt O QJ tr TJ H i P- tn p- H tt ro ro 3
plane (feed-back control). Also, in most cases, the instantaneous s3
pj 0 P­ rt PJ p­ 0 3 er 3 ro *< H ro 0 O O O . 3 fl) < PJ

zp ­ 0 H p­ 3 3 PJ tf 13 ro tf ro p-
rt
cn
P-
H 3 3 fl) f t 3 p-
3 pi ro [ M H p- Pi 3 0 O
wind speed and direction measured from the vessel are used to rt 1 H 1 O ^< a l *< I Q
er •< 1
z Z

1 37
 tf Pi rt Pi rt Pi ti z 0 Oi O tn rt
tn rt K < O ro rt tt Hi rt rt rt er P> 3 tn tn OJ rt
P­ P­ ro ti p­
Q ro OJ cr fi­ O p­ er CD fl) IH H er X ro OJ O p­ CD P OJ rt & cP tn er
< 3 pi p­ < Z tf < rt 3 tf ro P­ tf IH 1 PJ TJ tn rt rt 0 tn P> cn 0 tr rt
ti­ ro
fl) OJ OJ Hi
ro H i pi ro fli n­ rt 3 * OJ IH r— 13 CD p P­ z ro 3 TJ ro 3 H i -
3 M rt ft fi i P­ er rt QJ tf ii 1­ rt­ rt H ro ro •
0 OJ ti­ 3 a ro s fll rt­ H Z r t rt­H i n o o H i tf rt tf
p­ z rt a Hi 3 ro O1 M ro t i ­ P i |NJ z ro p­ r t 3
pi n Hrt er P i rt u 0 3
P p ­ ro
rt ft p rt
z rt TJ
P
Hi ft Hi Hi P rt O 0 TJ ft P ft
P* Hi
< H i QJ ro p­ CD CD P cn 3 er p­ 1' tr rt 3 cn pi tn fi­ ro cr tn ro H ro CO <!0 rt O
CD
tr P li­ CD p­
rt er
p­
P 0 rt­ P 0 ro 0 rt P 3 er ro P 3 Hi H O 3 o
3 0 H ro 0 < tf 3 CD Pi p­ ■
ro O J 3 i­3 1 p­
cn ro • < 3 r t •< p­ _> rt r t ro3 3 f t
ro rt­ 3 rt rt
p­
rt ro H rt rt 3 ro p ­ a ro M P 1 p­ rt Pi
ii 3 H CD P tf Pi 0 tn H r t tf cr |t­3 M _» 3 e z Ulro r t t i ­ tn tr cr < 3 i rt­ ft
p. cr tn ti­ o cn 0 ro cn
tfP 3 P H P rt ft TJ
tr
rt P
P1
ti­
Hi 0 Z a ro t i ­ 1 OJ z p­ ■£ 0J ro ier ro ft H p­ o _ CD r t p* ro cn er H i Z 3 0 Z 0 ro Z ro 3 p­ rt* rt­ er P tn er
5_
Hi
tn
P ro 1­3 fli CD ro 3 Hi rt tf ro rt­ P­ ti­ iro P3 OJ X p i Ui
er er ro to ft ro ro0 p­H0 0 3
tn
ro p­
rt
P Z 0 3 Z p tn 0 Piro Prti 0 p­
ro ro P QJ P ro
M1 < p* o er o Hi 3 ro cn er ro 10 tf er H 0 13 p i ro f t p- ro « ro « CO erp ­ Z tr
H i ror t
3 i Qi
CD
P
tn
Hi P <: CD CD3
tf 3 < Hp ­ H
tn
ro Z tr
P 0 & ro 3 ro *< ro rt pi rt PJ QJ Pi ro Hi
in ro ro 0 CD H tr tn P ro rop * a p Q<r t
rt er PJ O *J 0 < ro 03 rt o P P tf ro 3 H ro P rt rt
M p­ ft Hi Hi P •< rt­ p ­ 3
3 M p­ < H ro H tn rt iro 3 3 rt H i ro tn rt­ o t t H
0
i p­ 1 tfp ­ ^3< rt 3 ro P­ 0 0 < rt ro
cn CD ro 3 & s: n z
fll M tf cn tn rf tr 3
Hi
ro in­
PJ

ro 3 tOtJ 13 P­ P­ z o p- ro 0 er tn TJ
p­ pi
Pi
Hi
3
H ier
p­ rf
3
CD
P
3
< cr 0
rt
i
Hi
pi rt rP­t o 3 ctC zp tf 3 o Hi CD P p­
n ti­ H C rt­ QJ 0 rt p p er ro CD n ip­ rt flj P 3 rt p­ PJ cr tf rt 3 ro ­ rt 0 CD CD <
CD
p ro
H
ro pz 3 3"
rt t i ­ ro 0 0z rt tf P_ 0 n p
p) er < ro cn p­ < 3 Pi rt ro fli 3 M ti­ rt­ in QJ p­ Oi f t ro tn H n p) 3 P* rt rocr Q
rt P i CD« ro
tn 3
a 3 P­ p­ rt CD
ro pH­i er tfp 3 p tn HP i PP­J < z rt­ P 3 rt Pi rt
p­
rt
p­
< ro rt­ < ro r t ro ro p cr er rt
Z 0 fi­ TJ
Pi
ro H < 0 ro TJ p­ 3 3 CD P 3 ll­ p­ 10) P 1
3 QJ 3 ti­ H o 13 p- rt 3 to H­ z Pi
ro H Qi tn 3 C 0 3 O Pi cn ro < IP­ PJ r t ft P rt­ P 3 p­ ro P p­P i 3 p P­ 1Q i
p rt Hi
*p ­ ro Hl
P­
3
tn
ro
i
er a ro • pp ii ZCD n rt
3 < P P» 0 ro HP­i 3 <
ft H ft 0 & rt­ 3 ro P­ rt rt­ fli 3 CD i pj TS 13 p­ 0) er cr rt ro cn ro cr OJ n 0 p ­ rt3 H iQ J cn er rt
P
pi rt Hi
ro < Hi i rti <
CD
3 P ro CD
cn
PJ
P­
ro
tn
3 tf ro
< o> 1 P­ 3 fli rt ro cn ro r t tr ro 3 CD 3 ro 3 OJ TJ 0 P1 ro ro PI H P 3 er P tn tf a 0 p­ ro 8 ro in
ro 0 •fl rt 0 M P"
3
ro 3 ft
rt
flj 3 P< 0 CD 3 P­ P. ft ■
cn cr Pi rt fli fl) irt PJ rt H 3 n rt 0 rt cnP tn sf t rt r t S
CD
rt CD
er H pi P
3
ro rt tn
0
H
0 cn rt­ 0rt tr ro rt H i ro 3 ro z
H X 3 Pi Hi
tr ro rt PJ 3 IA iro r t *< cn P i a p­ ►< r t er H i p­ QJ oi . Zp ­ ro 0 •< ro ro rt cn ro ro p o 0 < ­M— P < rt­ rt­ er p
Hi
P CD 0 CO !"*» ft
tf ro H O Hi 3 Ui ipj cr p­ QJ ro 3 er ro f t 0 3 tn o M
P ro er
0 ro
P0 rt
fl)
n
P
P
rt TJ cr r t P i
^< P3 Hi i
tn
ro
P­
£
P
0
ro
ro tn
3
P er ■
rt
P
rt LO
H H
H
ro f i ­ ro <
O
rt tn ro
ro rt Q • H ti­ c 3 Hi n rt O DJ OJ IP 0 n r t Z rt­ 3 < 0 ro
rt
3 Pi cr10 rt rt3 zro rt 0
P
M ■
rt tn
M

^­> ti­
tr o UJ o p­ LH Hi H P Hi O ro
M 3 OJ er Hl PI 3 0 er H. 3 0 IO a piP O p­ Pi 0 fl) H i er fli pj
P­ p .P p­
a • P
TJ ro rt cn o 0ro 0 er p0 tn H i ro ro ro
H P ti­ rt pi 0 0 0 1 tn 0
P­ TJ 0 fl) 0 pi CD ft p­ 1 ro n P1 Pi Hi in­ •< H 3 tn ro 0 ro P P
1
p­ tn ro
Q H P 3 p­
■ — 1

rt 0 tn rt­n ro er rt ro ro ­. rt rt . H i P
0 Qi 0 p
H­
3
3
rt­
C rt
p­ 3
H
0
ii
0)
rt
rt
3*
Oi
3
ti­
ii
CD rt
ro
tn
0
P 1
z
IP­
10 TJ
OJ rt* M
H tr
n tn r t tn H i
0 rt p­ r t Hi
rt
P1 < p­ TJ ro•
3 •2 ti­
er0
t_i. 3 ro rt H p 3 p­
^ ro ft tn
rt* 3 0 • r. r t 3 O rt ro
t r ro 0 CD ro ft ft ro tn p­ z 0 0) 13 M tn CD PJ 3 rt tf o» CD P ro tn z er rt
tn
H i p­0
P 3
P
p­
■p2­
tn P tf CO 0
tn
p CD
tn
rt P zro ro tf fD PI 0 <
ti­ ro p i 3 ft tn p­ H h TJ 3 tr Hi * < 1 Pi P" Hi QJ P rt O cn p­ H i ro tn rt tn3cn 0 0
Hi
0 rt p ­ rt 3
t i ­
H
­» cr er 3 rPt
er p) rt ■«
0 Q O ro O CD OJ H ro 10 3 0 tn tn p­ rt­ rt­ P- ft Hi tn 3 rt rt rt rt­ rt TJ 3 H & < p­ P 3 rt 3 0
p ­ tf rt 3 rt CD Z ro t i ­ p ­
ro a rt 0 Z QJ to 3 cn ro H i i IHi tn Hi Qj 0 rt ro ro 0 P- O ro tf ro ro p­
3"
CD
fl)
P fl)
fi­
ti­
TJ
P1
ro 3 0 ro ro
cn
H tr
CO
fD ro
tn P­
er P ro rt 0 tfP Ptn
ro Hi Hi ft rt to rt PJ P­ ro tf rt Hi 1 3 3 p. pj 3 < rt H H i tf ­. rt tf rt Q i
P a0
P
p.
Qi tf z Z
p­
3
tn
P P ti­ ro 3p ­ ti­
rt P tfP er 30
Hi Ul tn 0 er >i Hi CD P­ QJ Pi P ro ro lit OJ rt fl) cr p­ fl) n • P CD P iH
< er tr
rt rPt rt pH 3 er 3 er P tf P
(t
ro
X 3 ro 0 P ro n ft 0 Hi P> tf 3 P P *< P p­ 3 er 0 P
0 ro p rt H ro H. Oi OJ 3 tf ro tf ro ier rt er ro 3 tr ro rt ro PJ Z X 3 P­ 3 P­ P p P rt pi
ro ro ro f t ft rt­ H f i ­ P HP i H tn cn P­
tr pi p o
ii QJ p­ er 0 CD 3 Z 3 tf 0 3 P a iro p­
ro Ul rt­ tf Oi i­3 CD CD CD 0 3 P TI P r tcn 3 Hi ft
3 cn r t ro ro o cn tf CD Pi cr 3 0 ro i i P rt er ro tn tn OJ CD 3 n r t tn p­P­
M 3 rt
p­
rf ro rt rt er ro 0 z < H< 0 rt rt 0 0
co 0 Hi P ti­ ro •<: 3 Hi iro tf rt t i ­ ro ro TJ
3 tn 3 3 ^< OJ ro
H
*: 0 p P 0 *­ tfp tnCD
ro Hi
H. ro p ­ n­ p ­ 0 P­
o CD 3 p P rt 3
0 3 CD rt rt Z O 0 p­ < CD er p .
X p- fl) 3 tn tf tr Pi 0 ro 3 ro < HP < er 3 f t TJ P rt HH p­ rt
rt r t
*-* o Z fl) 3 er tr 3 3
IHi Pi ft p­ ft p­ rt
p* ro 3 rt 0 ro rt o P PJ ro ro tf 0 H tn tn
,­^ 3 p­
ro rt
Hi H oi 0 er rt rt 3 OJ er H Z 0) z K IHi < ro ro QJ ro ro 3 0 rt er Z >? er 3 0 OJP rt ro H QJ 3 < r t rt H M p­ rt 3 3 rt P­ ro CD 1 H CD o cn p ­
00 H 3 p­ H O •< pi P ro < tu H i •ej iro ro cn rt rt­ 0 Hi tf Hi pi QJ ro Pi CD 0 fl) rt HM p­
• rf 0 p ro P ro p­ P K P P ro ^ <1 P J t i ­ 130 tr
tf o ro rt p P­ Hi
*< rt tn ro ro < pj in 3 tn ro o 3 Hi p- P1 < rt z 3 p­ cn
P Hi H ^< r t 3P < CD
tf0 3p r t ro o o Qi rf Hi P
P 0 H tr cr C rt­ 3 i ro 0 h Irt CD QJ p i 0 CD 0
zJ Hi H rt ro » 3 N 0
<_l.
rt
OJ —* tn P­ P 3 Hi P rt ro Hi rt a l P­ P 1 cn 3 rt 0 Hi O Hi »< ro 0 J9!p z H P3 CD p­
ro tn rt er r t 3 . cr 0 0 Ul Hi rt 3 r t
pi . tr cr p­ rt 0 0 ■ O. er H ti­ ro p . O 10 3 n tr fP­t < ro
■ .
p< a QJ ti­
P
3 TJ
er Ztn 3
rt­ PS 0 K Hi
ro 3 ro t t 3CD UJ tr ro Ort 0 O
rt P P i 0P­
H 0 p tn P­ rt ii tf ro er ro t i ro 3 IHi P C a p> CD Pi ro rt P rt er
p­
ft P cn P ro rt­ P
tn
tf ro
in
0 a P cr ^< rt ­— 03 3tn P­
■

fl) H •s rt 0 p­ 0 ro Hi ro a HH i. OJ 1 *fl rt tn o P < i tn ft ft p­ tr PI tn P» trL J .

3 < < o cr 0 Hi p P_ ro H i rti0 f t HO i n ro rt Hi
3 3
tfP t i ­ C1D rt CD P
■ — ­

rt tr 3 p­ rt 3 0 ro fi O i 0) N IS P­ ro ro 3 tn rt CD P i H, er p­ H i OJ rt CD P 1 p­
PJ P
M
0 ro rt t i ­ o rt • 0 ro tf
P­ ro 0 er 3 rt­ fl) H Z Hi ft n Ifli tf a TJ P­ ro 3 K fl) 3 p­ < n­ rt ft
c
Hp i p­ Hi
3
H i TJ
ro rt
ro ro rt i­3 rt 3
g 13ro ti­ ro O tr
n Hi 3 ro Ch cn 0 tr tf n 0J 0 rt cr \< P tf P 0 tn ro 3 ro fli tn ro H i tf P­ 1 p­
CD er H i CD P­
o rt er a ro tf o rf
tf rt er H ><
H 0 Ul P 3 ro c ro < rt Hi p­ ro iro rt p­ p­ r t 3 Pci tn Qj pi p- ro 3 3 tf
0 r tro
rt 3 ro ro 0 C D 0 ro ro er ro c ro
cn
er p
p­
rt
p­ a 03 ro H i O P
ft ro H , rt ro 0 p­ cn ro z 0 3 11 CD < 3 P cn 3P- i 3 0 0
p
K rt ro CD 3 rt 3tn 30 pi rt ro 3 ro CD er 3
H tn Qi 6 Hi PJ ro rt i OJ H tf I_J. IHi ro ft tr f t Hi P" 0 z
pi rt OJ t f
13 rt
p­
pi 0 Hi M a
rt P0 ro Hi i
PJ
1 H­
0
ro P H
P
o
a
p­
ro
M 3 rt
3
P P QJ
ro 3
cr ■< PP­J
c P 3
p­
3
C TJ
3 0 Hi
OJ c ti­ P rt rt Hi
ro p­ Hi t. P c iro < 3 13 P­ P> tn 0 n 0 K 3 < n 3 er DJ 0 3 Hi c n tr rt tn r t QcJ p­ Hi P CD er *< rt
tr rt Hi rf
z tf t n 3 P
3 M er ro pi 0 rt < Hi p. co CD Pi CD 0 pi iro H rt 0 cn Po OJ rt rt ro ro P i ro 3 PJ P P P­ rt
to rt ro p­ er fl) ro CD tn 3 tf ipi 1 p­ ro 3 CD 0 H1 zQJ P- rt tn
3
in K 0
rt tr r0t Pp ­ 0 rt • fl> 3 Hi rt­ rt 0 rt p 13 ro o
rt
Pi P 0 ro ro tf CD rt­ ft rt rt CND <: P n 0 3 er p­ 3
Hi
ro
tn
p>
rt­
0 Z
ti­
er
CD cn o
tn rt
rt
tr
3
ro
P.
1
P
0 3 ­rt
ro
a
11
IHi
P
■
3 < tn P i H, Z P
p­ 0) ^< rt
3 p­
cr
ro
H Pi
Oi n
P ft 0 tr
3 tn
rt rUlt 2 ** tfp XM ro pi rt C<D P ro 3 ft 3
tn
0 QJ P
tn
0
3
rt
tn
0

H Pi rt CD t. 3 p­ 10 0 0 P P. 3
p­ Ul P Hi ti­ ro P* Hi 3 rt ro O 3 P rt
0 0) ro tf Hi er m 0» < P ro rt cn Ul er H i ro
ti­ tn
rt 0 P­> TJ Hi ti­ er 3O r t 0 0 tf 0 3 < rt ro rt
ro
rt 3 rt < 0 P* O c M O rt 3 ro tr IH OJ p 3 3 0 ro TJ PJ rf < ro ro ro ro f t rt­er . c 0 pi 0 P CD
ro er ro z rt ro p i ro ro QJ 3 P Hi 0 0
ro tf er ro H rt cn p­ >i P ^< tr tn P i 3 i 0 z tf P­ P 3 o tt Hi
pi
Hi
p P­ ro ro1 P) tr
Hi
rt tf 5 01 tn < \z
Ifll H ti­ TJ Hi tn p- 3 0 rt Pi p­ 0 3 cr ro
ro i­3
Ul
ro
Hi 3
ft
rt rt
p­
0
rt
P
rt 3
fl) 0 rtl H p p­
3
rt
rt
ro
p
.£*. ro P
P
3
0
rt
rti

LJ 3 rt P Pi ­< p­ 0J P ro Irt ro er P CD p- ft P- 3 p­ 0 i Hi
ro rtrt tr P i 0n tc rt P­ CD Hi ti­
rt 0 30 ­ HH ii 0p ­ ro rt ft o
KO P P 01 p­ cr H tn ro I CD l tr 1 0 3
0
ro ro
z ro < 3
P rt
rt er ro z CD rt
Hl
CD
Pi
rt tr
•£ *
1 1 i tn p­
ro
P< i
IQi
PJ
1 3 p­
1
QJ
ro < PPJ 6 & 3 z
0 0
pi
p
< 3 fl) TJ rt ro ro tf ro 3 rti rti rt
0
0
0 er ro
w in
ro P i ro P 13 ro fl) rt P­" TJ tf r t C Pi rt rt Pfl)i 0 0 tf c ro
ni Hi rt i <
P cn Ul P *. P* P ro pH­i ro rt rt rt
cr P­ rt
1
tn 1 p­ ro K
tn
H p­ ro 0
I
3 3 O cr 0 <
ro 3 rt M 1 3
1
o 1 0
CD
ro ro z 3CD __
1
M tf l *< ■<
3
 TOTAL TOTAL
w
\ . l 0.354 0.444 0 .523 0.600 0.713 0.80 3 0.887
0.354 0.444 0.523 0.600 0.713 0,803 0.887
U
2 \ ^ "2 \ ^

0.354 2 9 10 25 10 38 37 0.354 1 18 25 32 14 39 29

0.444 7 21 19 26 12 35 0.444 9 19 38 14 16 5

0.523 12 16 8 14 14 0.523 16 38 36 31 27

0.600 14 10 18 15 0.600 17 11 10 15

0.713 9 4 7 0.713 13 8 3
T 2 in tf/m 2 T 2 in tf/m 2
0.803 9 5 0.803 12 4
Frequencies in rad./sec. Frequencies in rad./sec.
0.887 9 0.887 10

CONTRIBUTION I CONTRIBUTION I

\ " l 0.523 0.354 0.444 0.523 0.600 0.713 0.80 3 0.887

0.354 0.444 0.600 0.713 0.80 3 0.887 »2 \ ^
W
2 ^ \

0.354 10 19 27 23 21 20 2i 0.354 5 10 21 22 18 14 17

0.444 25 32 31 26 28 28 0.444 25 33 24 17 13 8

0.523 32 34 27 30 27 0.523 54 50 47 36 43

0.600 31 29 29 27 0.600 40 42 33 35

0.713 23 27 25 0.713 22 34 26
2 T. 2 in tf/m
T. in tf/m 0.80 3 23 34
0.803 25 28
Frequencies in rad./ sec. Frequencies in rad./sec.
0.887 25 0.887 22

TOTAL - 0.4 * CONTRIBUTION I TOTAL - 0.4 * CONTRIBUTION I

\ " l
0.354 0.444 0.523 0.600 0.713 0.803 0.887 0.354 0.444 0.523 0.600 0.713 0.803 0.887
W
* 2 ^ \ 2 ^ \

0.354 2 9 15 20 15 31 37 0.354 1 21 17 32 16 39 30

0.444 3 10 10 19 5 24 0.444 1 6 30 10 12 5

0.523 1 13 5 3 15 0.523 6 19 19 18 10

0.600 2 6 11 10 0.600 1 18 13 6

0.713 1 7 6 0.713 5 13 8
2 T in tf/m 2
2
T._ in tf/m 0.80 3 2 9
0.80 3 1 7
Frequencies in rad./sec. Frequencies in rad./sec.
0.887 1 0.887 2

Table VIII-1 Longitudinal force in head waves. Table VIII-2 Longitudinal force in bow quartering waves

141 142
 TOTAL
TOTAL
\ " l
0.354 0.444 0.523 0.600 0.713 0.803 0.887 \ * 1
w 0.354 0.444 0.523 0.600 0.713 0,803 0.887
2 ^ \ W
2 ^ \
0.354 1 35 9 13 34 19 24
0.354 364 7085 9284 8809 14 33 4448 2544
0.444 5 22 50 42 69 116
0.444 1527 16655 7551 3052 5067 12269
0.523 33 63 3 24 21
0.523 2554 7577 5870 2937 2994
0.600 79 57 15 8
0.600 782 5124 2363 2445
0.713 2 91 58 6
in 0.713 2 1193 4576 3574
T tf/m
0.803 90 60 T ? in tfm/m
Frequencies in rad./ sec. 0.803 200 4998
0.887 95 Frequencies ir rad. / sec.
0.887 562

CONTRIBUTION I
CONTRIBUTION I
0.354 0.444 0.523 0.600 0.713 0.803 0.887
0.354 0.444 0.523 0.600 0.713 0.803 0.887
* 2 ^ \ o _ 2 ^ \

0.354 5 43 35 49 11 17 15
0.354 1881 1993 3637 1611 3742 1669 1872
0.444 61 159 44 134 41 79
0.444 11633 16368 26015 3456 9644 7040
0.523 102 165 73 26 40
0.523 9066 5377 11610 6747 2808
0.600 259 192 50 41
0.600 1909 14844 12883 2498
0.713 298 221 69
2 0.713 4366 13830 14818
2
0.80 3 T _ in tf/m 297 230 T,2 in tfm/m
Frequencies in rad./ sec. 0. 803 678 13751
0.887 309 Frequencies in r a d . / sec.
0.887 1003

TOTAL - 0 . 3 * CONTRIBUTION I
TOTAL - 0 . 3 * CONTRIBUTION I
w
\ . l
0.354 0.444 0.523 0.600 0.713 0.803 0.887
U
2 \ . 0.354 0.444 0.523 0.600 0.713 0.803 0.887
W
2 ^ \
0.354 0 46 6 17 31 21 21
0 . 354 200 7589 8223 8634 2515 3989 2685
0.444 13 43 47 29 65 103
0.444 1963 15856 8282 2434 3250 11398
0.523 2 15 20 19 17
0.523 166 6081 2626 3330 2292
0.600 1 13 21 5
0.600 209 677 2854 3146
0.713 2 2 17 23
0.713 2 117 1402 2 861
T in tf/m
0.80 3 1 16 T.-j i n tfm/m
Frequencies in rad./sec. 0.803 404 1196
0.887 2 Frequencies in r a d . / s e c .
0.887 261

Table VIII-3 Transverse force in bow quartering waves. Table VIII-4 Yaw moment in bow quartering waves

143 144
YIIIi3i_Generation_of_the_waye-feed-forward_c
The wave-feed-forward control signal is derived from the
continuous evaluation of the following equations for the longitu­
dinal force, the transverse force and the yaw moment:
- longitudinal force:
Another method to determine the relative wave elevation is
to install pressure pick-ups in the side of the vessel just below
the waterline. The pressure variations may be easily converted into
relative wave elevations making use of the fact that near the wave
surface pressure variations are due to hydrostatic effects. For the
experiments use was made of wave probes fixed to the side of the
vessel.

F, (t) « -C 1l / a.pgc'15 -n .d£ (VIII-1) The measured relative wave elevation at a point along the
1
WL
waterline contains first and higher order components:
- transverse force:
C r = £C^ 1 } + e 2 ^ 2 ) + (VIII-4)
(VIII-2)
F
2 (t) w
" 2 C
/ ^P^l^ - 2 .d£
n
WL
The square of <; gives:
- yaw moment:
.r2 = e2C<1)2 + e3.'11.'2' + 0 ( _ 4 , + (VIII-5)
1}
M,(t)
J
w -C_
J
/ ^pgc* . (x.n, -x 9 n.) -dA . . . . (VIII-3)
WL i / z i
From this it follows that the lowest order in the square of £
r
in which C , C and C are gain factors which express the ratio (1)
X __1 *J involves only t as is required for the elevation of equations
r
between the total wave drift forces and the contribution due to 3
the relative wave elevation. (VIII-1) through (VIII-3). Components of order e and higher are
generally small and of high frequency and will be neglected here­
The above equations contain line integrals around the water- after.
line which involve the geometry of the hull form at the waterline In irregular waves the first order relative wave elevation
and the instantaneous values of the first order relative wave ele­ in a point along the waterline may be written as:
vation. It must be remembered that C is the wave elevation as
measured relative to the vessel at the waterline. Its value at any C* a ) (t) = Z ci.1' -cos(o) t +e.) (VIII-6)
i=l i
point can therefore be measured directly by means of a wave probe
fixed to the side of the vessel. This is shown in Figure VIII-1. Squaring this expression gives:

ll)2 N N
(1) (1)
ZIr ' ( t ) = X Z ZI ZI . c o s ( u > . t + e . ) . c o s (to. t + e . )
i=l j=l i j i i 1 3

N N
MEAN WATER LINE ON VESSEL (1) (1)
= Z Z %c; ' C J . X ) . c o s { ( w . - u . ) t + (e. - e . ) } +
i - l 3=1 i j i J i - 3

N N
(1)}
(1)
+ 2 Z ^C;
r K1r} . c o s { ( u , + u . ) t + (EX + e . ) }
i=l j =l i j ! 3 -i -3

Fig. VIII-1 Wave probe measuring relative wave elevation t, . (VIII-7)

145 146
This shows that the square of the measured wave elevation will in
K
general contain mean and low frequency components corresponding to F2(t) « - C 2 I kpgZrn (t)-n .A£ n (VIII-9)
the difference frequencies and high frequency components correspond­ n=l
ing to the sum frequency of the frequency components in equation
K
(VIII-6). In Figure VIII-2 time records of ci M t ) and the square M-, (t)ft*-C. Z ^pgc rn (t) . (x, n~ - 2n
X. In
n, ) .nAJt
In 2n
of C (1. (t) are shown schematically. In this figure the dotted line n=l
indicates the low frequency part of the square of the wave eleva­ (VIII-10)
tion. in which
K = number of wave probes
index n = denotes wave probe under consideration
= measured relative wave elevation of n wave probe
X X = co-ordinates of n wave probe and the centre of a
ln' 2n
straight line element approximating the local water-
£. <t) line form
M = length of n waterline element
n
n, , Pr, = direction cosines of the nth waterline element
In' 2n
C , C , C^ = gain factors.

For the model tests eight wave probes were used. The position of
the wave probes are shown in Figure VIII-3. Equation (VIII-8)
through equation (VIII-10) were evaluated continually by means of
an analog computer.

WAVE PROBE
SERVO UNIT

Fig. VIII-2 Relative wave elevation and square of relative wave

elevation.

For the model tests equations (VIII-1) through (VIII-3) were

WAVE AND
evaluated by replacing the integrals by simple summations of the CURRENT
following type: DIRECTION

K
F.(t) « - C . _ .pg Crn Xt).n. n .A. n (VIII-8) Fig. VIII-3 Position of wave probes measuring relative wave ele­
n=l vation.

1 Al 148
 From equation (VIII-7) it will be clear that the output of
the control signals as generated by equations (VIII-8) through
(VIII-10) will contain mean and low frequency components correspond­
ing to the difference frequencies of the irregular waves and high
frequencies corresponding to the sum frequencies of the irregular
waves. As indicated in the introduction of this chapter thruster
control signals may not contain high frequencies from the point of
view of wear and tear of mechanical components. The wave-feed-for­
ward thrust control signals must therefore be filtered to eliminate
the high frequencies which in this case are sum frequencies. As was
already indicated in the introduction care must be taken to select
an analog filter which, while removing the sum frequency component, tf)
does not cause appreciable phase lag in the low frequency compo­ £-180
CJl
nents. With this type of signal this does not form a problem be­ a
TJ

cause the high frequencies are in the order of twice the wave fre­
quencies. This means that the demands placed on the filter are more
easily met in this case than in the case of a normal feed-back con­ (/) - 9 0
LU
trol system based on the position error signal. In such cases the (/)
<
I
high frequencies coincide with the wave frequencies. 0.

The amplitude and phase characteristics of the analog filter

through which the wave-feed-forward signals generated by equations
(VIII-8) through (VIII-10) were passed are given in Figure VIII-4.
c
From this figure it is seen that the phase lag remains less than 4.
i_ IOO
45 degrees for frequencies up to 0.21 rad./sec. full scale. The a
spectra of the irregular waves in which model tests were carried
out are given in Figure VIII-5. From this figure it can be deduced
that the sum frequency components in the unfiltered wave-feed-for­ < 50
Z>
ward signals range from about 0.6 rad./sec. upwards, which is twice Z
z
the lowest frequencies present in the irregular waves with the LU
P
P
longest mean period. The amplitude of the filter has at this fre­ <
quency reduced to 50%. For the wave spectrum with the lowest mean
period the sum frequency components have frequencies higher than 10-3 10 - 2 10 -1 10" 10'
1.0 rad./sec. At this frequency the filter amplitude is 40%. FREQUENCY in rad.secr 1

Fig. VIII-4 Low-pass frequency c h a r a c t e r i s t i c of t h e wave-feed-

forward.
 0) TJ 1 «
1 0) 4)
fl) • c 4. tf
c n ­a ­rt OJ rrt 0 T3 l_ 4J
0 tf c fl) 0 r-\
tf rd m 1 N O iO) O tf)
N rfl 0 0 IL) rfl
4­1 r.
> Pi. tf C •rt SH
T3 tf
fl)
•H P u Qi. Cn SH r.
o -180 fl)
U > i OJ Cn OJ re . ■ri tf a 0 tn rd TJ fl) tf)
0 tf u ­r. tn 0) tn O tf P H fl) > ffl •
tf u PH CA fO 0) •ri Ch P O tf VO
T3 0 fl) 4­1 •H rH a) tf tf Ui ­rt 1
0) fl) 4­1 P tf •H 0 tf
>1
o o •M ffle QJ 0) H
tf c ­M in d p 0) I
TJ TJ H
Ui tf rt
h
__
■ r t

tf V. rH 3 in tf 4H 0) c P H
§
■ r t

0) fd c 0 rH c tf 0 3 c HLJ O O >
in > 3 tf 0 td 0 tf D1 0 rd 0
p
tfto - 9 0 tf
4­» 0 ­H 0 > P o cn P u < tf TJ 0) 0)
■ri
■H pi TJ tf rrt iH tn 3 td SH P TJ rt
■J X
11 c
p
3
4­>
CA 0)
CA
0)
tn
■ r t

tn
TJ
TJ
3 •4H tf a.
O
N
rj
•ri
ai P 3
m fl) rt o u tf t n
CM (V
M •ri tf) tf) rd 0) H 3 td •H tf a) •rt •rt
ai O P <D tr •rt 0) 0) H s tf 0 0 0) SH 3 tf H h
H 0 3 P > tf 0 {Ij ro rt rrt O H
i « ­
HP •rt o Pi EH iH 0 M tf tw tf ­ ­rt P ■ric
tn H H 3 til tf m 0) C 0) O tf) tf ra
E £ i­t OJ 1 tf • c i C 0 tf <L> TJ
CO O) //
y /
3 tn H tf 4J tf] 0 TJ 0) O tf > i tf 0 <L) 0) C
oi ■*"
y / P. to (fl 0) H o 0) tn tf tf ri rt tf 3
. 0 O) tf tf) tP 0) 0) 0) c •rt QJ EH 0
7—f— H
m ei I / ro > c o 0P m tn tf 0 o T3 rH » tf
cu Qi 0 tf} 3 3
■ r t i td B 0 P rfl • in
* S
•A
> 0) N •rt ■P ■a tf 01 3 TJ H fl) tf » tn
%i~n i_ji
_— rfl T3 tf •rt tf U CA > • 0) 0) c P (A QJ ro
3 OJ tf rt EH 0) 0 rd rrt > i tf > U)
<D O QJ •ri V4
y
y 4­1 0 P § tf 3 fd r-\ 0 ­rtO N ■ri O rd
^y
rt MH tf . c )H C tf fd tf tfto 100u P
__
■ r t

<~~ ffl P 0 tf) 0 tf CD 0) tn c c tf£L iH c QJ cn

i­i O fd 0) o cn tf •r-i a) TJ 0) 0 QJ 3 1 iH
3 0) 0 0) c tf tn TJ E 0 tf 3 P 0)
D* e H rH
tf) o 0 C g fl> EH O** 0) tf
*.\ CD Cn <­t
> i
0 TJ TJ H 0) 0 OJ SH rH
■ V PH SH OH fl)
U ­rt c Pi m 0 0 a) C 0 Of 0) rd O. tf iH tf .H
u El tf ftJ a r.
tf tl) rd H 0) Oi H tf tf tf tf
­rt 0)1 CA to QJ iH tf TJ 0)
­PI cn w
* tf H H C C
>1
tf a
tn
■ri
P tf 5 > 0)
«4H LOI tn TJ id rd 0 •ri 3 0
0
c m* tf
0) Cn
c ­rt
rH TJ 5 0H0 rd tn
Ql
>4 P ­rt rrt
•rt ■ r t
C u 3
ml ­rt TJ 3 > ■ri
>. CP U) fd O
UJ tf) tf
rd l tf) ttJ 0 tn 0) td rrt
­rt r * tf ■rt TI P
> H Tl 0) tf
u Pi 3 0) U p 3 fl) cn u ■H tf c P
U rd P tf
4J Ci tf rfl 1 TJ CA rd c p td <
oi C ffl tf tf
0 ­HI tf tf rt tf 0) cn H tf 0) in tP O
CD PI 0 TJ 0 tf tf tf 0) O 1 3 SH tf ­rt tn tu
PH Ol c ■ r t Cn rd > rt Ti 0 0) c QJ CA QJ H CA
W ­rti TJ f[J tf rrt ■•-\
iH tf 0) > m ai tf ■ri
tf O
4­11 ft] td H 0 (1) P 0) g 4J O •rt
7.3S gUJ Ul ((T]) ­^S 3 r4
tf >1 ■rt
­rti 0) c tf 04 tf 0 4H QJ ■ri Q) 4J C tf) tf
LD CAI ­P 0 to c 01 irt tf o tf) TJ U O ­rt 0) -3 tf)
10 Ul 10 - 2 10 -1 10v 10'
1 Ol IA ­rt tf 0 tf 0 0) 1 rd tf O 3 O ■ r t

H
H
P.I
i H
C tf
­rt
■ r t

p
N
•ri
H u
c
P
0
.*
0
tf
H
0)
0)
H
rd
rrt

Pi <u
0
rH
cr P i
0)
rt
<U FREQUENCY in rad.sec.-1
Pi •i Ul 3 u ■ ■ri rd )H c tn > fl) u tf) tf
> *rl 0 0 m CA tf tf 0 •ri ■ri
> tf rd U
•I QJ 0 tf i Ul 0) 1 tf ­rt TJ tf ffl

. Ptl > H tf P TJ tf rd tf 0) tf U
CP Hi H r. 0 H ■ri )H 0) 0) V. rH <-i c > u rfl
HI rfl 0) H c 0 0) tf 0 td fl) fl) f f l p tf
■ r t

>l tf tn s >
tf MH MH tf tf M g O
EM
3 QJ.
s e

Fig. VIII-6 Low-pass frequency characteristic of the damping term,

The part of the control signal which was proportional to the dis­ For each sea condition a model test was carried out twice
placement remained unfiltered. The contribution of this part to the in the same wave train, once with and once without the wave-feed­
total feed-back control signal was weak and mainly served to limit forward control signals. A block diagram of the control system is
the mean displacement of the vessel. shown in Figure VIII-8.

The overall characteristics of the feed-back control system

were such that high damping was achieved in the low frequency re­
gion only. The horizontal motions of the vessel in still water
WAVES
after an initial displacement out of the equilibrium position are
given in Figure VIII-7 and demonstrate the high damping introduced
by the feed-back system. During the tests which were all carried WAVE
FEED -0
out in irregular waves the feed-back system parameters remained
FORWARD
unchanged, except for the test in head seas (180 ) where the damp­
ing for surge motion was reduced in order to show more clearly the
effect of wave-feed-forward.
SERVO POSITION
SYSTEMS VESSEL

SURGE

FEED-BACK
CONTROLLER

Fig. VIII-8 Block diagram of dynamic positioning system.

The values of the gain factors C , C and C-. of the wave-feed-for­

ward signals were adjusted on a trial and error basis as at the
time of execution of model tests it was not possible to predict
these values on the basis of computations. Due to limitation in
OL the test set-up it also was not possible to determine afterwards
which value had actually been used.
50 100
TIME in sec.

Fig. VIII-7 Surge motion decay after an initial displacement in

still water.
VIII.5. Model tests

VIII.5.1. General WITHOUT WAVE-FEED-FORWARD

WITH WAVE-FEED-FORWARD
The model tests were carried out in irregular waves for the
5 rnr WAVE
following sea conditions:
- In head waves (180°) with a significant wave height of 4.9 m and
a mean period of 10.2 sec.
- In bow quartering waves (135 ) with a significant wave height of
4.9 m and a mean period of 10.2 sec. with and without a current
of 1 knot from 45 degrees.

- In bow quartering waves (135 ) with a significant wave height of

2.6 m and a mean period of 8.2 sec. -2ml-

The spectra of the irregular waves are shown in Figure VIII-5. The Fig. VIII-9 Surge motions in irregular head waves. Significant
wave and current directions are defined in Figure VIII-3. height 4.9 m.

During a model test, which was carried out for a time dura­
"5 rnr WAVE
tion corresponding to 35 minutes full scale, the surge, sway and
yaw motions and the total longitudinal and transverse forces and
yaw moment exerted on the vessel by the servo units were measured
and recorded on F.M.-tape. The results of measurements were ana­
lyzed to determine the spectra of the low frequency components of
the motions and the forces as well as the mean values of the forces,
The mean values of the motions are not of importance in this case,
since this can be easily rectified in reality by the inclusion of
an additional control signal based on a time integral of the dis­
placements.

VIII.5.2. Results of tests in irregular waves

The results of the tests are presented in the form of exam­

ples of time traces of the horizontal motions (Figures VIII-9 and
VIII-10) and in the form of spectra of the low frequency components
of forces, moment and horizontal motions (Figures VIII-11 through TIME in sec.
VIII-13).

Fig. VIII-10 Sway and yaw motions in irregular bow quartering

waves. Significant height 4.9 m.

155 156
 From the results it is seen that, except for the surge mo­
tions in bow quartering waves (135°), the low frequency parts of
the horizontal motions are significantly reduced when applying
wave-feed-forward. It appears that a reduction in the motions need WITHOUT WAVE-FEED-FORWARD
not necessarily result in a corresponding increase in the thrust WITH WAVE-FEED-FORWARD
to be applied to the vessel.
SURGE LONGITUDINAL FORCE
The mean wave drifting forces are not affected by the con­
v
trol system used as is demonstrated from the results given in Fig­ 1)
o
___ 4)
ures VIII-11 through VIII-13. In Figure VIII-12 it is seen that
tn
CM
-|1x104 «
CM

the low frequency component of the sway force F„ does not change
significantly even though the sway motion itself is considerably
smaller when using wave-feed-forward. In the same figures it is X
F
seen that the spectral density of the yaw moment is increased. In imean=-10tf
-12 t t
terms of lateral forces applied at the end of the vessel the abso­
0 0
lute value of the moment is small. SWAY TRANSVERSE FORCE

u
1x10^ «
2 mean C\J

WITHOUT WAVE-FEED-FORWARD
WITH WAVE-FEED-FORWARD
OJ
_1_

SURGE ^LONGITUDINAL FORCE

YAW YAW MOMENT

r\
C\J 5- - 1 x104 ™
S u
tn
3 F 2 8 t f j ^110 tf.m
^A 1mearr- \ 1x107 E
Li­
x en \
t/) \
'' x
3mean =\200 t f . m
0 ~ ^ \ l ^ - — 1
\
0.25 0.25 \
t/)
to in rad.sec - 1

0.25 0 0.25
CO in rad.sec. -1

Fig. VIII-11 Spectra of low frequency surge motion and force in Fig. VIII-12 S p e c t r a of low f r e q u e n c y f o r c e s and m o t i o n s i n irreg-
irregular head waves. Significant height 4.9 m. u l a r bow q u a r t e r i n g waves. S i g n i f i c a n t h e i g h t 2.6 m.

157 158
 In Figures VIII-12 and VIII-13 it is seen that the surge
motion is hardly affected by wave-feed-forward. It appeared that
the damping of the low frequency surge motion was so large that the
WITHOUT WAVE-FEED-FORWARD feed-back system alone could reduce low frequency surge motions to
WITH WAVE -FEED-FORWARD almost minimal values. Adding the wave-feed-forward control signal
brought about only minimal changes. The corresponding surge motion
SURGE LONGITUDINAL FORCE decay test in still water is shown in Figure VIII-7, indicated by
135 , being the wave direction in the tests shown in Figure VIII-12
u CM and Figure VIII-13. For the test in irregular head waves, shown in
10 10 1x10'
04 Figure VIII-11, the surge motion damping was reduced (see surge
E motion decay test in Figure VIII-7: 180 ) , showing more clearly the
effect of wave-feed-forward on low frequency surge motions.
li-
x F
imean-38tt
uo -41 tf VIII.5.3. .Results of tests in irregular waves and current
0
SWAY TRANSVERSE FORCE Tests were carried out with and without wave-feed-forward
in bow quartering irregular waves (135 ) with a significant wave
u height of 4.9 m and a mean period of 10.2 sec. and a stern quarter­
tu
tn 10 1 x 1 0 5 ™_ ing current (45 ) of about 1 knot. The spectra of the low frequency
CM
parts of the motions and forces are shown in Figure VIII-14. The
results are similar to the results shown in Figure VIII-13. The
OO if" irregular waves are in both cases according to the wave spectrum
X 10
if) given in Figure VIII-5.

O 0 The results given in Figure VIII-14 show that current does

YAW YAW MOMENT
u not affect the control signal. This is to be expected since the
4J
M3 m e a n s 2 3 6 0 t f ' m tn wave-feed-forward signal is determined from wave elevation signals
u

A
o CM
QJ
(0 2x10 b
E which do not change appreciably for the normal values of the cur­
CM
O)
CM
rent speeds encountered. From the results it is also seen that the
OJ
TD low frequency forces and motions are not appreciably different
from the results given in Figure VIII-13, which indicates that the
y- \
(0 ro influence of current on the low frequency wave drift forces is,
X 2450tf.m\
I/) in this case, not great.
0 0
0 0.25 O 0.25
From the results of the tests in irregular waves with and
CO in rad.sec. - 1 without current it appears that it is possible to reduce the low
frequency part of the sway motion by about 70% and the low frequen­
Fig. VIII-13 S p e c t r a of low f r e q u e n c y f o r c e s and m o t i o n s i n irreg- cy yaw and surge motion by about 50% through the use of wave-feed­
u l a r bow q u a r t e r i n g waves. S i g n i f i c a n t h e i g h t 4.9m. forward.

159 160
 This is less than could be expected on the basis of results of com­
putations given in section 2 of this chapter. It should be borne
WITHOUT WAVE-FEED-FORWARD in mind, however, that for the experimental determination of the
WITH WAVE-FEED-FORWARD wave-feed-forward signals only eight wave probes were used- It may
be expected that increasing this number will lead to more accurate
SURGE LONGITUDINAL FORCE evaluation of the low frequency component of the contribution due
to the relative wave elevation, thereby increasing the accuracy
u
of the control signals.
CM
10-

Y_______.§^__:2G_=iy§i22§

In this chapter it has been shown that, as a result of the

X F Li.
to 1mean="24tf 10 theory developed in this study, whereby the mean and low frequency
i
wave drift forces on floating objects are determined through di­
0 0
SWAY TRANSVERSE FORCE rect integration of all pressure contributions to the second order
forces over the wetted part of the hull, expressions are derived
u which, after numerical evaluation, lead to conclusions regarding
m 10 F = 179tt -1x10' the applicability of hitherto unknown methods to improve the accu­
CM > 2mean CM
•, racy of dynamic positioning of vessels in waves.
E
\ \ 177 tt

X
\ \ CM
It is possible to predict on the basis of the results of
CM
X \ \ to
LL
computations the effectiveness of a wave-feed-forward control sig­
\ N
\ s nal with respect to the degree in which such a signal can compen­
1 0
0 ^-— i sate the instantaneous mean and low frequency wave drift forces
YAW YAW MOMENT
acting on a vessel. Theoretical computations can be used to deter­
u mine the values of the gain factors which are inherent to the wave-
m
<M 8 CM feed-forward method. The results of model tests indicate that,
O) 5- - 2x10
0) CM
even without prior knowledge of such gain factors, wave-feed-for­
T3
ward can function effectively.
M
(0 3mean=-840ttm

°zxx
X / -600tt.m CO

to
5
o\ 0.25 0 0.25
0
0) in rad.sec.- 1

Fig. VIII-14 Spectra of low frequency forces and motions in irreg­

ular bow quartering waves and 1 knot stern quartering
current. Significant height 4.9 irt.

161 162
IX. CONCLUSIONS
As a result of the investigations presented in this study
the following conclusions can be drawn:
1. The total low frequency hydrodynamic forces acting on a vessel
in waves may be considered as the sum of two parts:
- the mean and low frequency second order wave exciting forces;
- the hydrodynamic reaction forces resulting from the low fre­
quency motions induced by the low frequency wave forces.
Furthermore the mean and low frequency second order wave exci­
ting forces are independent of the low frequency motions. The
hydrodynamic reaction forces may be expressed in terms of added
mass and damping coefficients which are determined by means of
existing linear potential theory methods (chapter III).
2. Based on the method of direct integration of fluid pressure
acting on the instantaneous wetted part of the hull of a body
it is shown that the total second order wave exciting forces
contain five components. Four of these components may be eval­
uated using existing computation methods based on linear po­
tential theory. The fifth contribution depends on the solution
of the second order non-linear velocity potential. This con­
tribution may be approximated using results on the first order
wave exciting forces (chapters III and IV).
3. The second order wave exciting forces acting on a body in ir­
regular waves are the sum of second order force components due
to regular wave groups present in irregular waves. The second
order wave exciting forces may be expressed in the form of
quadratic transfer functions which may be used to compute the
second order forces in irregular waves in the frequency domain
or the time domain (chapter IV).
4. The comparison between results of computations using the meth­
od of direct integration of pressure and analytical results
obtained using an existing method based on energy and momentum
considerations demonstrates the equivalence of both methods
with respect to the total force (chapter V ) .
5. Except at wave frequencies and wave directions which result in
large amplitude resonant roll motions, the mean second order
horizontal wave forces and yaw moment on a tanker and a rec­
tangular barge can be predicted with good accuracy by means of
computations based on potential theory (chapter VI).
6. The good comparison obtained between experimental results and
computed results on the mean second order horizontal forces
and yaw moment acting on a semi-submersible and the mean sec­
ond order vertical force and pitch moment on a submerged cyl­
inder indicates that viscous effects are small even for bodies
consisting of slender elements (chapter VI).
7. The mean second order horizontal forces acting on surface ves­
sels such as a hemisphere, a tanker, a barge and a semi-sub­
mersible are dominated by the contribution due to the relative
wave elevation around the waterline. This contribution deter­
mines the sign of the total force. The second most important
contribution is due to the non-linear pressure contribution
in the Bernoulli equation. This contribution is generally of
the same order as the total force, but of opposite sign. The
contribution of the second order non-linear velocity potential
to the mean forces is at all times equal to zero. The remaining
two contributions, due to products of local pressure gradients
and motions and due to products of angular body motions and
inertia forces, vary in sign and are generally smaller in mag­
nitude (chapter VI).
8. The mean second order vertical forces on a submerged horizontal
cylinder are dominated by the non-linear pressure contribution
in the Bernoulli equation. The contribution due to products of
local pressure gradients and motions are also of importance.
Other contributions are zero or small compared to the first
two contributions (chapter VI).
9. The low frequency second order longitudinal force in irregular
waves on a tanker and a semi-submersible contains contributions
 arising from products of first order quantities and a contri­
bution due to the second order velocity potential. The relative
importance of these contributions depends on the wave frequen­
cies and the low frequencies of interest of the second order
force. At high wave frequencies, where first order diffraction
effects are large, the low frequency forces are dominated by
contributions arising from products of first order quantities.
At low wave frequencies, where first order diffraction effects
are small, the contribution due to the second order velocity
potential is relatively of greater importance. The importance
of this contribution becomes greater as the frequency of the
second order forces increases (chapter VII).
10. Existing methods for computing the low frequency second order
forces in irregular waves on floating structures, which rely
solely on the mean forces in regular waves, are applicable for
surface vessels provided that:
- the wave frequencies are sufficiently high to ensure that the
second order exciting forces are dominated by products of
first order quantities;
- the frequencies of interest of the second order forces are
low (chapter VII).
11. In the case of a long submerged horizontal cylinder in beam
waves computations of the low frequency horizontal force in
irregular waves cannot be based on the mean force in regular
waves (chapter VII).
12. The accuracy of station keeping of a dynamically positioned
vessel can be improved through the application of a wave-feed­
forward control signal, which is based on the relative wave
elevation measured around the vessel (chapter VIII).
APPENDIX A - COMPUTATION OF THE FIRST ORDER SOLUTION FOR THE
VELOCITY POTENTIAL AND BODY MOTIONS
I_}tE25_lH2£i22
This appendix gives a short account of the underlying theory
and method of computation of the first order velocity potential and
first order body motions for an arbitrarily shaped body floating
in regular, long-crested waves as given by Van Oortmerssen [A-l].
A brief review of the method is given here for the sake of com­
pleteness and due to its importance with respect to the computa­
tion of the mean and low frequency second order forces. Since this
appendix deals only with first order quantities the affix ,
which is used in the main body of this work to distinguish between
first and second order quantities, is deleted. Furthermore, in
keeping with ref. [A-l] use is made of the complex notation e 1 W
to denote oscillatory quantities instead of sin tot and cos tut.
Descrigtion_of_the_theory
First order wave loads and motions
The ship is considered as a rigid body, oscillating sinusoi-
dally about a state of rest, in response to excitation by a long-
crested regular wave. The amplitudes of the motions of the ship as
well as of the wave are supposed to be small while the fluid is
assumed to be ideal and irrotational. A right-handed, fixed system
of co-ordinates O-X.-X -X_ is defined with the origin in the mean
position of the centre of gravity of the body and the 0-X axis
4- Vi
vertically upwards. The oscillating motion of the ship in the j
mode is given by:
Xj = Cj e" l w t j = 1 ,6 (A-l)
in which £. is the amplitude of the motion in the j mode and ta
the circular frequency. The motion variables x , x ? and x^ stand
for the translations surge, sway and heave, while x., xc and x,
denote rotations around the 0-X., O-X and 0-X., axes respectively.
 The free surface at great distance from the ship is defined >_j (x l f x 2 ,x 3 ) = — // Pj (a a ,a 2 ,a 3 ) . Y j (x 1 ,x 2 ,x 3 ,a 1 ,a 2 ,a 3 )dS
by:
ik(x cos a + x„ sin a) - ioit for j = 1, 2 , - . . . , 7
C = CQ e (A-2) (A-6)
where:
where:
Y-(x,x„,x-.,a,a„,a 3 ) = the Green's function of a source, singular
Cn = amplitude of the wave
in a 2 , a 2 , a 3
k = wave number = 2TT/A, where X is the wave length
a. ~ angle of incidence. a., a-,, a-. = the vector describing S
a
j (a a
i' 2' 3 a ] = the complex source strength.
The flow field can be characterized by a first order veloc­
ity potential: For the Green's function a function is chosen which satis­
fies the Laplace equation and the boundary conditions on the sea
1U)
*(x1,x2,x3,t) = 4i(x1,x2,x3)e (A-3) bottom, in the free surface and at infinity. This function is giv­
en by (see Wehausen and Laitone [A-2]):
The potential function $ can be separated into contributions from
all modes of motion and from the incident and diffracted wave v1 = - + —
r r
+ PV .
fields:
°° 2 (£ + v)e -?d .cosh £ (a3 + c) .cosh £ (x3 + c)
<>
j = ito C0(<t>0 + * 7 > " iu Z *j C
j (A
~4) / J 0 (£R)d£ +
£ sinh £d - v cosh £d

The incident wave potential is given by: 2 2

2TT (k - v ) .cosh k (a, + c) .cosh k (x + c)
+ i = ~ J n (kR)
cosh k(x. +c) ik(x. cos a + x 9 sin a) k^d - v d + v u
e
*n = ~ vT-^3 • • • (A-5)
T (A-7)
0 v cosh kd
in which: in which:
v = co2/g
.
c = the distance from the origin to the sea bed ) 2 + (x2 - a 2 ) 2 + (x3 - a 3 ) 2
= \/ ( x l- a l
d = water depth
2
a = angle of incidence of the waves. 1 =
\/ ( x l- a l ) + (x2 ~ a 2 ) 2 + (x3 +2c + a 3 ) 2 . . . (A-8)

2 2
The cases j = 1,....,6 correspond to the potentials due to the mo­ R ^(x.-a^ + U2-a2)
tion of the ship in the j mode, while 4>7 is the potential of the
diffracted waves. The individual potentials are all solutions of John [A-33 has derived the following series for y, which is
the Laplace equation which satisfy the linearized free surface the analogue of (A-7):
condition and the boundary conditions on the sea floor, on the
2
body's surface and at infinity. The potential function 4> can be - k2
v
Y = 2-/T 9 = cosh k(a + c) .cosh k(x-, + c) .
represented by a continuous distribution of single sources on the k d - v d + v J J

boundary surface S:

i cn i aa
 2
4(y +V2)
{YQ(kR) -i J <kR)> + Z 2 l 2 n1 = cos (n,x,)
i = l dy. + dv - v
n 2 = cos (n,x2)
n~ = cos (n,x-J
cos y . ( x 3 + c ) . c o s y^ (a., + c) .K (y . R) (A-12)
n. = x-,n-> - x,n-
(A-9) 4 2 3 3 2
n5 = x 3 n i - X l n 3
w h e r e y. are the p o s i t i v e solutions of: n 6 = xxn2 - x2nx

Vi. t a n ( y . d ) + v = 0 (A-10) To solve equation (A-6) numerically the surface S is sub­

divided into a number of finite, plane elements on which the source
Although these two representations are equivalent, one of strength is constant. The boundary condition is applied in one con­
the two may have preference for numerical computations depending trol point on each element being the centre of the element. The
on the values of the variables. In general, equation (A-9) is the integral equation (A-6) then reduces to a set of algebraic equa­
most convenient representation for calculations. When R = 0 the tions in the unknown source strengths. In general, the Green's
value of K„ becomes infinite; therefore equation (A-7) must be function y may be computed with sufficient accuracy as if the
used when R is small or zero. source strength is concentrated in the centre (control point) of
each element. When, however, the influence of an element on its
The unknown source strength function o must be determined own control point is evaluated y has a singularity of the type
such that the boundary condition on the body's surface S is ful­ 1/r, which can be removed by spreading the source uniformly over
filled. Due to the linearization this boundary condition is ap­ the panel. When the influence of a panel on a control point, which
plied to the surface in its equilibrium position S is at a close distance of this panel and not lying in the same
0 plane, is considered the source is spread uniformly and integrated
n = 3£
numerically to obtain its contribution to <\> or
j -^ j (x 1 ,x 2 ,x 3 ) + 47 // aj(a1,a2,a3) . 3n*
S
0
After solving the equations for the source strengths the
Y x x x a a a dS for = x 6 first,'order potential function is known. The pressure on the sur­
" 3n j' l' 2' 3' l' 2' 3^ J ' '
face S can then be found from Bernoulli's theorem. The linearized
hydrodynamic pressure is given by:
3
n = *0
j ' IF ^r j = 7
p(xn ,x~ ,x v t) = -p -^r =
(A-ll) 'l'~2'~3 3t
2 2 -iut
n through nfi are the generalized direction cosines on S„ , defined = {pio C 0 (<J>0 + tj>?) + po) E ax C.)e
J J
by: j=l
(A-13)

Subsequently, the first order wave exciting forces and moments can
be found from:

169
 X k ­ ­pa)2 C 0 e " l u t // ( ♦ 0 + * 7 ) n k dS M . is an inertia matrix. Since the origin of the system of axes
S
0 coincides with the centre of gravity of the ship in its rest posi­
(A­14) tion it is found that:

The oscillating hydrodynamic forces (k = 1, 2, 3) and moments (k m 0 0 0 0

4, 5, 6) in the k direction are:
0 m 0 0 0

F = ­pto 2
Z c e' i(iit
// *.1 n dS 0 0 m 0 0
­
:i =ii J s0 ­ M
s> kj 0 0 0 0
-J46
(A­15)
0 0 0 C 0
5

According to common practice the hydrodynamic forces are repre­ 0 0 0 -I 64 0 I,.

sented by means of added mass and damping coefficients:

:A-I9)

a k . = ­p Re {jj ^ n k dS} A( ­16)

S
0 whe re:
m = mass of the ship
b k j = ­poo Im {// $.. n k dS} A( ­17) th
I, = moment of inertia in the k " mode
S
0 I, . = product of inertia.
where:
a, . = the added mass coefficient in the k­mode due to motion in
the j­mode
b, . = the damping
dam coefficient in the k­mode due to motion in the
j­mode.

Finally, the motion response to first order excitation is

computed by means of the well known equations of motion in the
frequency domain:

6
2
K . + a, . ) .sin(tot + e .) +K b, . . to.cos (tot + e .) +
£ {­to (M,
j­! 3 *J J 3 3

+ C j . . s i n t w t + e . ) H - = X k « s i n (tot + 6fe)

f o r k = 1 , . . . . ,6
(A-18)
i n which:
Xk = wave excited force in the k mode
£_■_ 6 k = phase angles.

1 *T 1 172
APPENDIX B - CROSS-BI-SPECTRAL ANALYSIS N N
= J l j­l ^ ^ C ^ ­ c o s U t o . ­ t o . J t + ( S i ­ j ) ) +
Introduction
N N
The computer program which was used to analyze the input- (1) (1)
+ Z Z JfcJ t\ .cos{(to + to.)t + (e. + E . ) }
J J x
output relationship between the waves (input) and the low frequency i=l j=l J
mooring or restraining forces (output) is based on the cross-bi- (B­2)
spectral method as given by Dalzell [B-l], [B-2] and [B-3]. For a
complete description of the method we refer to his works. The low frequency part of the square of the wave elevation is:
2 N N
The intention of this appendix is to give a "feel" for the C
low (t) = Z S
%ci1,C_{1,.cos{(ioi ­<o.)t + ( E . ­ E , ) }
J J J
processes involved, rather than to give a condensed version of the i=l j=l
specialist's point of view as given by Dalzell. (B­3)

We assume that the input (wave elevation) can be written as It is assumed that the output (wave drift force or moment) contains
follows: only low frequencies and is closely related to the low frequency
N
part of the square of the wave elevation:
(1) (1)
}
Z (t) = Z z; } .sin(w.t + e. ) (B-l) N N
i=l i J- F (2) (t) = Z Z C.(1)c(1)P.,.cos{(to. ­to,)t + (E. ­ E , ) } +
i=l j=i i j ij J ­ J J
in which:
N N
to. = frequency in rad./sec. + Z I
(l)} (1)
Z\ 1J
Z\ Q.,.sin{(to. ­to.)t + (e. ­£.)}
i=l j = i J­ j ­LJ J ­ j J ­ J
e. = random phase, uniformly distributed from 0 - 2TT
(B­4)
Z- = amplitude of component with frequency to.
t = time in which P.. and Q.■ are in­phase and out­of phase quadratic trans­
N = a large number. fer functions dependent on the frequencies to. and to..

The foregoing expression represents a zero-mean, normally distri­ The problem is to determine P.. and Q.. for arbitrary val­
buted, stationary random signal. The square of the wave elevation J-J /jv ij
ues of to.
x and to. given that the input ? (t) and the output
is:
F (2) (t) are only
^ known as time records. The ouput is a signal with
low frequency oscillatory components.
(ll2 N
... 7
ZK } (t) = { Z Ci .cos(w.t + E . ) T
i=l x x
From equation (B­4) it can be shown that the time record of
N N
any component of F (2) (t) with chosen frequency Aw is the sum of
(1) (1)
= Z Z c- z\J .costto.t +e. ) .cos(io.t +e.) contributions from components of which the difference frequencies
i=l j=l x i i 3 3
are equal to the chosen frequency Aio:
 N
N
(2)
*' (t) = Z C -( ±l ' )C.-(i 'D
P - . - ; - c o s { A £ O t + (E. - £ - ) } + Utt) = I U ..cos{ {ta± -to. )t + (^-e. )} (B-8)
to ._. i J J-3 J - J
i=l 1 3 J J
where:
N ta. + to - = to, = some chosen fixed frequency
+ Z c | 1 } C-1)Qi--sin{Atot + (^-Ej))
EX , ex = random phases of i and j frequency components of
wave elevations of equation (B-l).
(B-5)
(2)
i n which i and j a r e chosen such that: If the signal is cross-correlated with the output F (t) of equa­
tion (B-4) it follows that:

to. - to. = Ato (B-6) i +T/2 e>\

Rpn(T) = Iim i / U(t) .F u '(t +x)dt (B-9)
tU
T-* l -T/2
Equation (B-5) becomes: (2)
By this operation only those components of F (t) will be identi­
r N
F< 2} ( t) = [ Z cf^C^^Pij.COStE.-Ej) + fied Which correspond with the components of equation (B-8). Con­
Ato tributions from all other components will disappear since they are
not correlated to the components of U(t).
+ Qx< .sintE^ -Ej) }cos Atot +
The actual computation of the transfer functions starts with
f Z C^d^Q^.cos.E.-E.)
a transformation of the input signal C(t) according to:
L
i=l J J J + Tm
/ cos tokx.C(1) (t - T ) - C ( 1 ) (t +t)dT (B-10)
- P^.sin.E.-E.)} sin Atot
~Tm
This represents the Fourier transformation of the product:
(B-7)
( 1 ) ( 1 ,
. (_-.).. (t + .,
From equation (B-7) it can be seen that the amplitude of a frequen-
(2 .
cy component of F v '(t) contains information on a range of P.. and in which:
Q.. values. It is not possible to determine the value of individ­ T = time shift
ual P..*s or Q..'s from the signal. This is a result of the fact cok = some chosen fixed frequency
that F v ' (t) is a double summation. The foregoing indicates that T = maximum time shift (maximum number of lags multiplied by
in order to determine the quadratic transfer functions for required sampling interval).
combinations of to. and to. it is necessary to find a way to extract
from the time record of the output information which is essentially It will be clear that the output is a function of the chosen fre­
in the form of a single summation of oscillatory components with quency tok and time t. Substitution of the expression (B-l) for the
amplitudes which are in themselves not a summation of components. wave elevation gives the following result for the inner product:
For instance, it is possible to generate a time signal U(t) of the
following type:

176
175
 C(l)(t­T).CU)(t+t) = Inspection of this expression shows that the contributions which
arise for the cases that to. = to. + to. will dominate, so that the
N N K 1 J
Z Z £ .(1)
i;
G (1)
• .COS{li). (t ­ T ) + £. }.COS{(0. (t + T ) + E . } outcome is of the following type:
1 3 J J
i=l i=i
N
2x J x ^ i(l).(D
C^i f k ) -co8{( U i - U j ( i # k ) )t + U ± - £ j ( i f l c ))}
N N m
I £ hzl(D.d)
t\ .cos{ (to± +to.)x + {ta± ­to. )t +
i=l j = l ­* J J (B-14)

+ (_E. ­ E X ) } + high frequency components The outcome is a signal which contains only those difference fre­
quency components of the wave elevation which have as sum frequency:
(B­ll)
tok = to. + to. (B-15)
M u l t i p l i c a t i o n by cos O XT g i v e s :
which is what was needed in order to be able to identify correspond-
(2)
COS ( O k T . C
U )
(t -T) . C U )
Ct + T ) = ing components in the output F v ' (t). The above expression appears
to increase as x increases. It must be remembered, however, that
m
N N the processes involved are stochastic. The output of the above ex­
(1) 1 }
= Z Z hZ ci .cos[{io, - (to. +CO .)}T + (to. - t o . ) t +
. • . i J •"• J-J J - J pression is finite.
J
i=l j=l J J

+
(£. ­ E X ) ] + high frequency components Some examples are given of the output of the above expres­
sion in Figure B - 2 . The input is the wave elevation of which the
ordinary wave spectrum is given in Figure B-l. The output of ex­
(B­12)
pression (B-14) is given for three values of the sum frequency to,.

Disregarding the high frequency components expression (B­10) be­ 10

comes : £w1/ 3 "5.5m
N N
(1) (1) T =12 sec.
Z Z hz CU, .cos{((o. ­u>.)t + u
x D J ffl
i=l j=l CM.

m
E
+ .£.­£.)} / COS{(0 ­ ((0. + tO. ) }T.dT +
1 J K 1 J
"Tm
(
/)
N N
(1) (1)
­ Z Z ^C C. ­sin{ (to. ­to.)t +
x J J

z
i=l j=l
~ ^
m
+ (ex ­£■) } / sin {to,
K
­ (to. + to.) } T .dT 0.5 1.0
J __ J ­ J
m 1
U) in rad.sec-*
(B­13) Fig. B-l Wave spectrum.

178
 From the plots given in Figure B­2 it is seen that the mid­
dle value of the sum frequency results in a signal which contains
large amplitude low frequency components and that the other to, val­
ues result in signals with much less low frequency components. This
+1
is explained by examination of the wave spectrum given in Figure
m
/ c o s cok x £ ( t - x ) . £ (t*x) dx B­l. From this figure it can be deduced that the low frequency part
of the square of the wave elevation, which is related to the occur­
-x m
rence of wave groups, has little energy for the highest and lowest
0)u = 0.8 rad.secr 1 sum frequencies since in those cases the spectral density of the
waves and consequently the amplitude of wave groups is smaller than
for the middle value of to, .

-15 m sec. L This operation results in a signal which for arbitrary sum
frequency to, supplies the time record containing the low frequency
t u k = 1.2 rad.sec;-1
part of the square of the wave elevation (only for those difference
frequencies which have to, as sum frequency) . This solves the prob­
lem of determining the transfer functions P. . and CL­:­ By perform­
ing cross­correlation between this signal and the low frequency
output only those frequency components of the output corresponding
to the input will be identified. This means that the cross­corre­
lation function contains only information for those combinations
of to. and to. which have co. as sum frequency. The final result after
1 3 ­K
Fourier transformation of the cross­correlation function is the
cross­spectrum of the transformed input and output or, in cross­
15 m 2 sec. r-
bi­spectral terminology, the cross­bi­spectrum of input, input and
output. The cross­bi­spectrum is valid for the chosen sum frequen­
-15 m 2 sec. - cy ta, and contains information for the range of difference frequen­
cies from zero upwards. Any chosen difference frequencies contain
WAVE £(t) information on the transfer function P.. and Q.. for unique values
of to. and to.. The values of the transfer functions are finally de­
termined from the following type of expression:

-2 m L GU
i' u j» ■ s ^ v i . . ) (B 16)
-
h 1 Z J

where:
O 100
TIME in sec. C F (toi,io.) = cross­bi­spectrum of input­input­output
S ((D.) = wave spectrum (scalar spectrum)
G(toi,to.) = quadratic transfer function of output.
F i g . B-2 Time r e c o r d s of wave and t r a n s f o r m e d wave

179 180
 The afore given explanation on the cross-bi-spectral analy­
sis method can be used to indicate how the computational method
can be altered in order to save computation time and increase the
accuracy for identifying the quadratic transfer fiinction for the
low frequency wave drift forces.
Since we are restricting ourselves to low frequencies in the
output the computations which take place after the initial trans­
formation of the input (see equation (B-10)) can be performed using
a considerably increased sampling interval. This can be done after
the output of the transformation examples, which are given in Fig­
ure B-2, are low-pass filtered. Subsequent computation of cross-
correlation functions between the transformed input and the output
can be carried out using a greater value of the maximum time shift
between the signals thus allowing for potentially more accurate
determination of the quadratic transfer functions for very low fre­
quencies.
Analysis_of_model_tests_with_the_tanker_and^
The duration of tests in irregular waves with the model of
the tanker corresponded to 210 minutes full scale. The input (wave)
and output (forces) records were sampled at 0.8 sec. intervals.
The bi-spectral analysis of the digitized data were carried out
using 75 lags for the cross-correlation functions. In the analysis
the above described process by which the sampling interval is in­
creased was not applied.
The duration of the tests with the semi-submersible corre­
sponded with 360 minutes full scale. The time records of input and
output were digitized using a sampling interval of 1.5 sec. full
scale. The number of lags used for the cross-correlations amounted
to 30. After the initial transformation of the input (see equation
(B-10)), the transformed input and the output were low-pass fil­
tered and the remaining computations were carried out using 30 lags
and a sampling interval corresponding to 7.5 sec. full scale.
181
REFERENCES (CHAPTER I)
[I-l] Verhagen, J.H.G. and Van Sluijs, M.F.: "The low frequency
drifting force on floating bodies in waves", International
Shipbuilding Progress, Vol. 17, No. 188, April 1970.
[1-2] Hsu, F.H. and Blenkarn, K.A.: "Analysis of peak mooring
forces caused by slow vessel drift oscillations in random
seas". Paper No. 1159, O.T.C., Houston, 1970.
[1-3] Remery, G.F.M. and Hermans, A.J.: "The slow drift oscilla­
tions of a moored object in random seas". Paper No. 1500,
O.T.C, Houston, 1971.
[1-4] Bhattacharyya, Rameswhar: "Dynamics of marine vehicles",
Wiley Series on Ocean Engineering, 1978.
[1-5] Kuo, C., Lee, A., Welya, Y. and Martin, J.: "Semi-submers­
ible intact stability-static and dynamic assessment and
steady tilt in waves". Paper No. 2976, O.T.C, Houston,
1977.
[1-6] Sjouke, J. and Lagers, G.: "Development of dynamic posi­
tioning", Paper No. 1498, O.T.C, Houston, 1971.
[1-7] Sugiura, M., Nekado, Y. and Matsui, M.: "An experiment of
the dynamic positioning system of a semi-submersible plat­
form", Interocean, 1973.
[1-8] Tamehiro, M., Akasaka, N., Kasai, H. and Miwa, E.: "On
dynamic positioning system design in particular reference
to the positional signal filtering technique", J.S.N.A.
Japan, Vol. 142, 1977.
[1-9] Van Oortmerssen, G.: "The motions of a moored ship in
waves", N.S.M.B. Publication No. 510, 1976.
1 flO
[1-10] Aral, S., Nakado, Y. and Takagi, M.: "Study on the motion REFERENCES (CHAPTER II)
of a moored vessel among the irregular waves", J.S.N.A.
Japan, Vol. 140, 1976. [Il-l] Maruo, H.: "The excess resistance of a ship in rough
seas", International Shipbuilding Progress, Vol. 4,
No. 35. 1957.

[II-2] Gerritsma, J., Van der Bosch, J. and Beukelman, W.:

"Propulsion in regular and irregular waves", International
Shipbuilding Progress, Vol. 8, No. 82, 1961.

[II-3] Dalzell, J.F.: "Application of the fundamental polynomial

model to the ship added resistance problem". Eleventh
Symposium on Naval Hydrodynamics, University College,
London, 1976.

[II-4] Suyehiro, K.: "The drift of ships caused by rolling among

waves". Trans. INA, Vol. 66 (pp 60-76), 1924.

[II-5] Watanabe, Y.: "Some contribution to the theory of rolling",

Trans. INA, Vol. 80 (pp 408-432), 1938.

[II-6] Havelock, T.H.: "The drifting force on a ship among waves".

Philosophical Magazine, Series 7, Vol. 33 (pp 467-475),
1942.

[II-7] Maruo, H.: "The drift of a body floating in waves". Jour­

nal of Ship Research, Vol. 4, No. 3, December i960.

[II-8] Kudou, K.: "The drifting force acting on a three-dimen­

sional body in waves", J.S.N.A. Japan, Vol. 141, 1977.

[II-9] Newman, J.N.: "The drift force and moment on ships in

waves", Journal of Ship Research, Vol. 11, No. 1, March
1967.

[11-10] Spens, P.G. and Lalangas, P.A.: "Measurements of the mean

lateral force and yawing moment on a series 60 model in
oblique regular waves", Davidson Laboratory, Report 880,
June 1962.

I ft-. i ft/i
[11-11] Faltinsen, O.M. and Michelsen, F.: "The motions of large
structures in waves at zero Froude number". Symposium on
the Dynamics of Marine Vehicles and Structures in Waves,
London, 19 74.
[11-12] Molin, B.: "Computations of drift forces". Paper No. 3627,
O.T.C, Houston, 1979.
[11-13] Kim, C.H. and Chou, F.: "Prediction of drifting force and
moment on an ocean platform floating in oblique waves",
International Shipbuilding Progress, Vol. 20, No. 230,
1973.
[11-14] Faltinsen, O.M. and L^ken, A.E.: "Drift forces and slowly
varying forces on ships and offshore structures in waves",
Norwegian Maritime Research, No. 1, 19 78.
[11-15] Joosen, W.P.A.: "Added resistance of ships in waves".
Proceedings of the Sixth Symposium on Naval Hydrodynamics,
1966.
[11-16] Lee, C M . and Newman, J.N. : "The vertical force and moment
of submerged bodies under waves". Journal of Ship Research,
Vol. 15, No. 3, 1971.
[11-17] Karppinen, T.: "An approach to computing the second order
steady forces on semi-submerged structures", Report No. 16,
Ship Hydrodynamics Laboratory, Helsinki University of
Technology, 19 79.
[11-18] Lin, W.C and Reed, A.: "The second order steady force and
moment on a ship moving in an oblique seaway". Eleventh
Symposium on Naval Hydrodynamics, University College,
London, 19 76.
[11-19] Gerritsma, J. and Beukelman, W.: "Analysis of the resis­
tance increase in waves of a fast cargo ship", Report No.
334, Laboratorium voor Scheepsbouwkunde, Technical Univer­
sity of Delft, 1971.
[11-20] Kaplan, P. and Sargent, T.P.: "Motions of offshore struc­
tures as influenced by mooring and positioning systems",
BOSS 1976, Trondheim, 1976.
[11-21] Ogilvie, T.F.: "First and second order forces on a cyl­
inder submerged under a free surface". Journal of Fluid
Mechanics, Vol. 16, Part 3 (pp 451-472), 1963.
[11-22] Goodman, T.R.: "Forces on a hovering slender body of
revolution submerged under a free surface", Developments
in Mechanics, Pergamon Press, New York, 1965.
[11-23] Salvesen, N.: "Second order steady state forces and mo­
ments on surface ships in oblique regular waves", Sym­
posium on the Dynamics of Marine Vehicles and Structures
in Waves, London, 1974.
[11-24] Dalzell, J.F. and Kim, C.H.: "Analytical investigation of
the quadratic frequency response for added resistance",
Report SIT-DL-76-1878, Davidson Laboratory, Stevens Insti­
tute of Technology, 1976.
[11-25] Ankudinov, V.K.: "Non-periodical forces and moments on a
ship in waves". International Shipbuilding Progress, Vol.
16, No. 179, 1969.
[11-26] Ankudinov, V.K.: "The added resistance of a moving ship
in waves", International Shipbuilding Progress, Vol. 19,
No. 220, 1972.
[11-27] Boese, P.: "Eine einfache Methode zur berechnung der
Widerstandserhohung eines Schiffes im Seegang", Schiffs-
technik, Ed. 17, 1970.
[11-28] Pinkster, J.A.: "Low frequency second order wave forces
on vessels moored at sea". Eleventh Symposium on Naval
Hydrodynamics, University College, London, 19 76.
[11-29] Pinkster, J.A. and Van Oortmerssen, G. : "Computation of [11-38] Wahab, R.: "Wave induced motions and drift forces on a
the first and second order wave forces on bodies oscil­ floating structure". Report No. 1865, Netherlands Ship
lating in regular waves", Second International Conference Research Centre, TNO, Delft, 1974.
on Numerical Ship Hydrodynamics, University of California,
Berkeley, 1977. [11-39] Huse, E.: "Wave induced mean force of platform in direc­
tion opposite to wave propagation", Interocean, 1976.
[11-30] Boreel, L.J.: "Wave action on large offshore structures",
Conference on Offshore Structures, Institute of Civil
Engineers, London, 19 74.

[11-31] Faltinsen, O.M. and L(_>ken, A.E.: "Drift forces and slowly
varying horizontal forces on ships in waves", Timman Sym­
posium on Applied Mathematics, Delft, 1978.

[11-32] Faltinsen, O.M. and LjzSken, A.E.: "Slow drift oscillations

on a ship in irregular waves", Journal of Applied Research,
No. 1, 1979.

[11-33] Pinkster, J.A. and Hooft, J.P.: "Low frequency drifting

forces on moored structures in waves". Fifth International
Ocean Development Conference, Tokyo, 19 78.

[11-34] Pinkster, J.A.: "Wave drifting forces", Proceedings of

WEGEMT, Aachen, 19 79.

[11-35] Pinkster, J.A.: "Mean and low frequency wave drifting

forces on floating structures", Ocean Engineering, October
1979.

[11-36] Bourianoff, G.L. and Penumalli, B.R.: "Numerical simula­

tion of ship motion by Eulerian Hydrodynamic Techniques",
Second International Conference on Numerical Ship Hydro­
dynamics, University of California, Berkeley, 1977.

[11-37] Pijfers, J.G.L. and Brink, A.W.: "Calculated drift forces

of two semi-submersible platform types in regular and ir­
regular waves", Paper No. 2977, O . T . C , Houston, 1977.
REFERENCES (CHAPTER III)
[III-l] Stoker, J.J.: "Water waves", Interscience Publishers Inc.,
New York, 1957.
[III-2] Van Oortmerssen, G.: "The motions of a moored ship in
waves", N.S.M.B. Publication No. 510, 1976.
[III-3] Joseph, D.D.: "Domain perturbations: the higher order
theory of infinitesimal water waves", Archive for Rational
Mechanics and Analysis, Vol. 51 (pp 295-303), 1973.
iftq
REFERENCES (CHAPTER IV)
[IV-1] Boreel, L.J.: "Wave action on large offshore structures",
Conference on Offshore Structures, Institute of Civil
Engineers, London, 19 74.
[IV-2] Van Oortmerssen, G.: "The motions of a moored ship in
waves", N.S.M.B. Publication No. 510, 1976.
[IV-3] Bowers, E.C.: "Long period oscillation of moored ships
subject to short wave seas", Paper presented to R.I.N.A.,
August 19 75.
[IV-4] Faltinsen, O.M. and Ljbken, A.E.: "Slow drift oscillations
on a ship in irregular waves". Journal of Applied Research,
No. 1, 1979.
[IV-5] Vugts, J.H.: "The hydrodynamic coefficients for swaying,
heaving and rolling cylinders in a free surface", Report
No. 194, Laboratorium voor Scheepsbouwkunde, Delft, 1968.
[IV-6] Davenport, Jr., W.B. and Root, W.L.: "An introduction to
the theory of random signals and noise", McGraw-Hill,
New York, 1958.
[IV-7] Dalzell, J.F.: "Application of the fundamental polynomial
model to the ship added resistance problem". Eleventh
Symposium on Naval Hydrodynamics, University College,
London, 19 76.
190
REFERENCES (CHAPTER V)
[V-l] Kudou, K.: "The drifting force acting on a three-dimen­
sional body in waves", J.S.N.A. Japan, Vol. 141, 1977.
[V-2] Maruo, H.: "The drift of a body floating in waves", Jour­
nal of Ship Research, Vol. 4, No. 3, December 1960.
[V-3] Salvesen, N.: "Second order steady state forces and moments
on surface ships in oblique regular waves". Symposium on
the Dynamics of Marine Vehicles and Structures in Waves,
London, 1974.
REFERENCES (CHAPTER VII)
[VII-1] Dalzell, J.F.: "Application of the fundamental polynomial
model to the ship added resistance problem", Eleventh
Symposium on Naval Hydrodynamics, University College,
London, 19 76.
[VII-2] Pinkster, J.A.: "Low frequency phenomena associated with
vessels moored at sea". Paper SPE 4837, European Spring
Meeting of SPE-AIME, Amsterdam, 1974.
REFERENCES (CHAPTER VIII) REFERENCES (APPENDIX A)

[VIII-1] Balchen, J . C , Jenssen, N.A. and Sealid, S. : "Dynamic
positioning using Kalman filtering and optimal control
theory", Automation in Offshore Oil Field Operation,
North Holland Publishing Co., 1976.
[VIII-2] Sjouke, J. and Lagers, G.: "Development of dynamic posi­
tioning". Paper No. 1498, O.T.C, Houston, 19 71.
[A-l] Van Oortmerssen, G.z " The motions of a moored ship in
waves", N.S.M.B. Publication No. 510, 1976.
[A-2] Wehausen, J.V. and Laitone, E.V.: "Handbuch der Physik",
Vol. 9, Springer Verlag, Berlin, 1960.
[A-3] John, F.: "On the motions of floating bodies", Comm. on
Pure and Applied Mathematics, Part I: 2, 1949 and Part
II: 3, 1950.
REFERENCES (APPENDIX B) NOMENCLATURE

[B-l] Dalzell, J.F.: "Application of cross-bi-spectral analysis L f f i x ( 0 J ' ( 1 ) ' ( 2 ) ' ( 3 ) affix denotes whether a quantity is of first,
to ship resistance in waves", Report SIT-DL-72-1606, David­ second, third order, etc.
son Laboratory, Stevens Institute of Technology, 1972.
A. . , B. . , C . coefficients dependent on wave frequencies
13' 13' 13
to. and to .
[B-2] Dalzell, J.F.: "Some additional studies of the application
of cross-bi-spectral analysis to ship resistance in waves", C.t) pressure contribution independent of the co­
Report SIT-DL-72-16 41, Davidson Laboratory, Stevens Insti­ ordinates of the point under consideration
tute of Technology, 19 72. force vector relative to the G­X'­X'­Xl sys­
tem of axes
[B-3] Dalzell, J.F.: "The applicability of the functional poly­
components of the force vector
nomial input-output model to ship resistance in waves".
F
l ' F2' F 3
= (2) =(2) =(2 mean values of the second order force compo­
Report SIT-DL-75-1794, Davidson Laboratory, Stevens Insti­ F
l ' F2 ' F3
tute of Technology, 1975. nents

centre of gravity of a body

(2) complex quadratic transfer function

G­xrx2­x3 body axes with origin in centre of gravity G,

x.­axis positive in forward direction, x„­axis
positive to port side and x.­axis positive up­
wards

o-x r x 2 -x 3 fixed system of axes with origin in the mean

free surface, X­axis and X­axis in the hor­
izontal plane and X­axis positive vertically
upwards

G—X — X_ — X­j system of axes with origin in centre of grav­

ity of the body and axes parallel to O­X.­X­­
X­ system of axes

mass moment of inertia matrix

­I 46
0 0
­I A I,
46 " 6
mass moments of inertia of the body about the
4' V h G­x , G­x and G­x 3 axes respectively
product of inertia
46 ■

length of a vessel or cylinder

M mass matrix a radius of sphere
m 0 0 d radius of cylinder
0 m 0
e 2.718 (constant of natural logarithm)
0 0 m
g constant of gravity
M moment vector relative to the G-X'-X'-X' sys­
(2)
tem of axes
g second order impuls response function
M., M-, M 3 components of the moment vector h water depth

N outward pointing normal vector of a point on i /=T

the surface of a body relative to the 0-X - k wave number = 2TT/A
X
2~X3 or G_X
l~X2~X3 svstems of
axes
d£ line element of the waterline
components of the quadratic transfer function
*__• Qij m mass of body in vacuum
dependent on to. and to.
rrx area of wave spectrum
T. . amplitude of the quadratic transfer function
13
dependent on to. and to. n outward pointing normal vector of a point on
the body relative to the body axes G-x 1 ~x - x 3
wave spectrum
n., n ? , n.. components of n
S total wetted surface of the hull
n f
ii ' ii coefficients depending on to. and to-
s
o mean wetted surface of the hull
p pressure
dS surface element of S or S
t time
V velocity vector relative to fixed system of
axes ua amplitude of an oscillatory quantity

WL static or mean waterline on the hull of a body x position vector of a point on the hull of the
body relative to the body axes
X position vector relative to the fixed system
of axes x., x ? , x-. components of x

X X X components of X
l' 2' 3
a chapter VI: angle of mooring line
X position of the centre of gravity of the body appendix A: angle of incidence of waves;
relative to the fixed system of axes 0 represents stern waves, 90
vector operator represents waves from starboard
beam and 180 represents head waves
V, V volume of the mean submerged part of a body
a angular motion vector
v2 Laplace operator
6 mooring line angle
PV Principle Value
F a small quantity << 1
A area of waterline of a sphere
X* E_. random phase angle of i frequency component
position vector relative to the G-X!-X'-X'
system of axes
i a ft
£ ­ phase angle between wave and some oscillatory SUMMARY
quantity u
In this thesis the mean and low frequency second order wave
Z wave elevation
drift forces on bodies moored or stationed in waves are analyzed.
C„ wave amplitude of a regular wave Expressions are derived for the second order forces based on direct
a integration of pressure acting on the wetted part of the body. It
Z relative wave elevation is shown that the second order forces in irregular waves may be
C. amplitude of i frequency component
determined from knowledge of the mean forces in regular waves and
the low frequency forces in regular wave groups.
Z , /­j significant wave height

T mean wave period In order to calculate the mean and low frequency forces on
A wave length bodies of arbitrary shape use is made of a three-dimensional linear
potential theory computer program. The form of the body is approx­
A non­dimensional frequency
imated by a distribution of plane facet elements representing a
y frequency of low frequency part of the second source distribution. For a hemisphere the results of computations
order forces of the mean second order horizontal forces in regular waves are
compared with analytical results. This comparison demonstrates the
p mass density of water accuracy of the computations and the equivalence of the expressions
x time lag for the second order forces developed in this thesis with respect
to an already existing expression based on momentum and energy con­
$ velocity potential dependent on co­ordinates
and time t siderations.
<J> part of velocity potential independent of
time
In order to demonstrate the validity of the present theory
to wave frequency with respect to realistic hull forms, results of computations of
r
to. 1■ t h frequency component the mean second order forces in regular waves are compared with
results of experiments on a tanker, a semi-submersible, a rectan­
a source strength
gular barge arid a submerged horizontal cylinder.
y Green's function

For the first three hull forms the mean horizontal forces
a.., a added mass fpr surge and heave motions are compared. For the submerged cylinder the mean vertical forces
k*il' ^33 damping for surge and heave motions are compared. The correlation found between results of computations
and experiments confirms the general applicability of the theory
x , x­J , xJ: first order surge, sway and heave motions
for predicting the second order forces on a wide range of hull
x. , x, , x first order roll, pitch and yaw motions forms.

A detailed analysis of components of the mean second order

forces shows that for floating vessels the horizontal forces are
dominated by a contribution dependent on the relative wave eleva­
tion around the waterline of these vessels.
 For the tanker and the semi-submersible results of computa­
tions of the low frequency horizontal force in regular wave groups
are compared with experimental results obtained from model tests
in regular wave groups and irregular waves. The experimental re­
sults from tests in irregular waves are analyzed by means of cross-
bi-spectral methods. Results of this comparison indicate that, pro­
vided certain conditions are fulfilled, the mean second order force
in regular waves may be used to approximate the low frequency force
in irregular waves.
Finally, for a dynamically positioned vessel the results of
computations are used to demonstrate the effectiveness of a wave-
feed-forward control signal based on relative wave elevation mea­
surements for reducing low frequency horizontal motions induced by
drift forces in irregular waves. The results show that model tests
confirm the theoretical predictions.
201
SAMENVATTING
In dit proefschrift wordt een analyse gegeven van de gemid­
deide en laag frekwente tweede orde golfdriftkrachten op een li-
chaam afgemeerd of gepositioneerd in golven. Uitgaande van integra­
tie van drukken over het natte oppervlak van een lichaam worden uit-
drukkingen voor de tweede orde krachten gegeven. Aangetoond wordt
dat de tweede orde krachten in onregelmatige golven bepaaid kunnen
worden uitgaande van kennis van gemiddeide krachten in regelmatige
golven en de laag frekwente krachten in regelmatige golfgroepen.
Voor het berekenen van de gemiddeide en laag frekwente
krachten op willekeurig gevormde lichamen wordt gebruik gemaakt van
een rekenprogramma gebaseerd op drie-dimensionale lineaire poten-
tiaal theorie. De vorm van het lichaam wordt benaderd door middel
van een aantal elementen die een verdeling van bronnen voorstelt.
Voor een halve bol worden de resultaten van berekeningen
van de gemiddeide horizontale golfdriftkracht in regelmatige golven
vergeleken met reeds bekende, langs analytische weg verkregen resul­
taten. De nauwkeurigheid van de berekeningsmethode wordt hiermee
aangetoond, alsmede dat de in dit proefschrift gegeven uitdrukking
voor de driftkrachten qua resultaat equivalent is aan een reeds
bekende uitdrukking die gebaseerd is op impuls en energie beschouw-
ingen.
Met het doel de geldigheid aan te tonen van de in dit proef­
schrift gegeven theorie met betrekking tot meer realistische romp-
vormen worden resultaten van berekeningen van de gemiddeide tweede
orde golfkrachten in regelmatige golven vergeleken met experimen-
teel bepaalde resultaten voor een tanker, een semi-submersible,
een rechthoekig ponton en een ondergedompelde horizontale cilinder.
Voor de drie eerstgenoemde rompvormen worden de resultaten
voor de horizontale krachten vergeleken. Voor de ondergedompelde
cilinder worden de resultaten voor de gemiddeide vertikale kracht
vergeleken. De overeenkomst tussen de resultaten van berekeningen
en metingen bevestigt de algemene toepasbaarheid van de theorie
voor het voorspellen van de tweede orde krachten op een grote ver-
scheidenheid van rompvormen.
ACKNOWLEDGEMENT

Een nadere analyse van de komponenten van de gemiddeide

I am grateful to the Board of Directors of the Netherlands
tweede orde krachten toont aan dat voor drijvende konstrukties
Ship Model Basin for permission to publish this research in the
de horizontale krachten gedomineerd worden door een bijdrage die
form of a thesis.
afhankelijk is van de relatieve golfhoogte ter plaatse van de water­
lijn van de konstruktie. I am indebted to the participants in the industry supported
Wave Drift Research Program for their kind permission to make use
Voor de tanker en de semi-submersible worden de resultaten of part of the model test data obtained from Phase I of this pro­
van berekeningen van de laag frekwente driftkracht in regelmatige gram.
golfgroepen vergeleken met resultaten verkregen uit modelproeven
in regelmatige golfgroepen en onregelmatige golven. De meetresul­
My sincere thanks to the members of the Ocean Engineering
taten verkregen uit proeven in onregelmatige golven zijn geanali-
Department for their patience and co-operation which contributed
seerd door middel van kruis-bi-spektrale methoden. Uit de verge-
much to the realization of this work.
lijking blijkt dat, mits aan bepaalde voorwaarden wordt voldaan,
de gemiddeide driftkrachten in regelmatige golven gebruikt kunnen
worden om de laag frekwente krachten in onregelmatige golven te
benaderen.

Tenslotte worden voor een dynamisch gepositioneerd schip

de resultaten van berekeningen gebruikt om een voorspelling te
geven van de effektiviteit van een wave-feed-forward regelsignaal
die gebaseerd is op metingen van de relatieve golfhoogte voor het
verminderen van de laag frekwente horizontale bewegingen, die op­
gewekt worden door de driftkrachten in onregelmatige golven.
Resultaten laten zien dat de modelproeven in overeenstemming zijn
met de theoretische voorspellingen.

?03