^-..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 WAVE EXCITING FORCES ON FLOATING STRUCTURES

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

Dit proefschnft is goedgekeurd door de promotoren Prof. Dr. Ir. A. J. Hermans Prof. Ir. J. Gerritsma

STELLINGEN I Bij het bepalen van de laag frekwente bewegingen van een in golven afgemeerde konstruktie dient rekening gehouden te worden met viskeuze effekten in de hydrodynamische reaktiekrachten. II Het relatieve bewegingsprincipe geeft inzicht in de krachten uitgeoefend door variabele dwarsstroom op een varend schip en is een goed uitgangspunt voor het bepalen van de grootte van deze krachten. Ill De afmetingen van een semi-submersible kunnen zodanig geoptimaliseerd worden dat de bewegingen met golf frekwenties minimaal zijn. Het verdient overweging eenzelfde optimalisatie toe te passen met betrekking tot de tweede orde golfdriftkrachten. IV Spelende kinderen die een bal in het water gooien maken dikwijls gebruik van 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. V Door sommige onderzoekers is beweerd dat golfdriftkrachten op een afgemeerde konstruktie in stroom en golven afhankelijk zijn van de stroomsnelheid. Onderzocht moet worden of het juist niet de stroomkrachten zijn die veranderen door de aanwezigheid van golven. VI 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

aanbeveling bij het ontwerpen van dergelijke systemen ook hier aandacht aan te besteden. VIII De laag frekwente golfdriftkracht op een afgemeerde konstruktie in onregelmatige golven kan, mits aan bepaalde voorwaarden wordt voldaan, voorspeld worden aan de hand van gegevens over de gemiddeide driftkracht in regelmatige golven. IX Een schip varend in recht voor- of recht achterinkomende golven kan grote slingerbewegingen vertonen bij golf frekwenties die ver buiten de eigen slinger frekwentie van het schip liggen. X Op grote schepen dienen gegevens aan boord aanwezig te zijn met betrekking tot de gemiddeide golf-, wind- en stroomkrachten. Deze gegevens zijn van belang voor het geval dat een schip ten gevolge van averij geassisteerd moet worden door sleepboten. XI Meer jarenplannen voor onderzoek kunnen belemmerend werken op de kreativiteit van de onderzoeker. XII Het is aan te bevelen de voor de Nederlandstalige leek nogal lachwekkende benaming 'driftkrachten' te vervangen door een uitdrukking die minder associaties met menselijk gedrag oproept. Dissertatie J. A. Pinkster Delft, 8 oktober 1980

Aan Martha en de kinderen

CONTENTS I. II. INTRODUCTION PAST DEVELOPMENTS CONCERNING T H E COMPUTATION OF MEAN AND LOW FREQUENCY WAVE FORCES II. 1. Introduction 11.2, Historical review II. 3. Conclusions HYDRODYNAMIC THEORY 111.1. Introduction 111.2. Co-ordinate systems . 111.3. Motion and velocity o f a point on the hull of the body 111.4. Fluid motions and boundary conditions III.4.1. Boundary conditions within the fluid, at the free surface and on the sea floor III. 4.2. Boundary conditions on the body III.4.3. Boundary conditions at infinity 111.5. Pressure in a point within the fluid III. 6. Second order wave force and moment 111.6.1. Second order wave force 111.6.2. Second order wave moment III. 7 . Conclusions 1 8 8 8 15 19 19 22 23 24 24 26 30 30 32 32 35 37

III.

IV.

EVALUATION OF T H E SECOND ORDER WAVE EXCITING FORCES ... 38 IV. 1. Introduction 38 IV.2. The quadratic transfer function 38 IV.2.1. General 38 IV.2.2. Evaluation of the components dependent on first order quantities 41 IV.2.3. Contribution of the second order potential .... 42 IV.2.3.1. General 42 IV.2.3.2. Approximation for the contribution of the second order potential 45 IV.2.3.3. Comparison between exact results and the approximation 50 IV.2.4. Symmetry of the quadratic transfer functions .. 53 IV.3. Frequency domain representation of the mean and low frequency forces in irregular waves 57

IV.4. Time domain representation of the mean and low frequency second order forces IV. 5 . Conclusions V. COMPARISON BETWEEN RESULTS OF COMPUTATIONS AND ANALYTI CAL RESULTS ON THE MEAN WAVE DRIFT FORCE IN REGULAR WAVES V. 1. Introduction V.2. Computations V.2.1. General V.2.2. Motions and mean horizontal drift force V.2.3. Components of the mean horizontal drift force .. V.2.4. Mean vertical drift force V. 3. Conclusions COMPARISON BETWEEN COMPUTATIONS AND MEASUREMENTS OF THE MEAN SECOND ORDER FORCE IN REGULAR WAVES VI. 1. Introduction VI.2. Model tests VI.2.1. General VI.2.2. Model test conditions VI. 3. Computations VI. 4. Results of computations and measurements VI. 5 . Conclusions DETERMINATION OF THE QUADRATIC TRANSFER FUNCTION OF THE LOW FREQUENCY SECOND ORDER FORCES VII.1. Introduction VII.2. Model test set-up VII.2.1. General VII.2.2. Realizations of two systems of restraint VII.3. Model tests VII.3.1. Generation of waves VII.3.2. Test procedure and duration of measurements .. VII.4. Analysis of results of measurements of the low frequency longitudinal force in head waves VII.4.1. Regular wave groups VII.4.2. Irregular waves VII.5. Computations VII.6. Comparison between computations and experi ments

59 60

61 61 61 61 67 69 71 72 73 73 76 76 79 82 88 102 104 104 105 105 Ill 116 116 120 121 121 122 122 125

VI.

VII.

VII.7. Approximation for the low frequency force in irregular waves VII.8. Conclusions VIII. APPLICATION OF THEORY TO DYNAMIC POSITIONING OF A VESSEL IN IRREGULAR WAVES VIII.1. Introduction VIII.2. Theoretical prediction of the effect of wavefeed-forward VIII.3. Generation of the wave-feed-forward control signal VIII.4. Positioning system VIII.5. Model tests VIII.5.1. General VIII.5.2. Results of tests in irregular waves VIII.5.3. Results of tests in irregular waves and current VIII.6. Conclusions IX. CONCLUSIONS 160 162 163 145 151 155 155 155 139 137 137 132 135

APPENDIX A - COMPUTATION OF THE FIRST ORDER SOLUTION FOR THE VELOCITY POTENTIAL AND BODY MOTIONS APPENDIX B - CROSS-BI-SPECTRAL ANALYSIS REFERENCES NOMENCLATURE SUMMARY SAMENVATTING ACKNOWLEDGEMENT 166 173 182 196 200 2 02 204

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 concerning 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 Blenkarn [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 TIME in sec.

200

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.

E
c
LJ

O tr ID

I/) Li.

o
UJ Q _ _ >

a.
<

A =PERIOD
Fig. 1-2

NATURAL PERIOD OF SURGE MOTION OF WAVE GROUPS. REF [l-3]

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.

CURRENT! CONTROL SIGNAL

WAVES MEASUREMENTS

THRUSTER ALLOCATION

VESSEL

<r
WIND MODEL WIND FORCE

WIND

PREDICTED MEASUREMENTS

\
VESSEL MODEL

-=6

r \
7' FORCE < DEMAND VESSEL STATE ~ ~

STATE CORRECTION

CURRENT FORCE

CURRENT MODEL

WAVE MODEL

SETPOINTS

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 oncoming waves. Such effects are of importance in judging the minimum static stability requirements for such platforms. From observations in reality and from the results of model tests it has been found that large, deep floating storage vessels can carry out low frequency heave motions in irregular waves which are of the same magnitude as the heave motions with wave frequen cies.

50 60
*c

70 80 90

Q. UJ O _J UJ UJ

X .-

_.

100 r 110 0 40 80 120 TIME i n sec. 160 200

Fig. 1-4

Depth record showing effect of suction force on sub marine under waves. Quartering sea, wave height = 18 ft., vessel speed = 0 knot. Ref. [1-4]-

From the foregoing it can be seen that, depending on the kind of structure or vessel considered, one or more of the six components of the mean and low frequency second order wave forces in irregular waves can be of importance. In order to be able to evaluate the influence of such forces on the performance or be haviour of a structure the most reliable method available, which can take into account in a relatively straightforward way those factors which are deemed of importance for the behaviour of a

system, is by means of model tests. In many practical cases suf ficient insight in the complex behaviour of, for instance, a large tanker moored to a single point mooring system is still lacking for reliable prediction of the motion behaviour and forces in the mooring system to be made by means other than physical model testing. 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 the effectiveness of control systems for dynamically positioned vessels, time domain simulations, which take into account the equations of motion of the vessel and the behaviour of external loads such as the mean and low frequency wave drift forces, are carried out. In such cases, due to the complexity of the control system and the objectives of the study, it is more practical to make use of simplified equations describing the environmental forces and the reaction of the structure or vessel to external forces than to simulate the characteristics of the control sys tems during a model test. See for instance Sjouke and Lagers [1-6], Sugiura et al [1-7] and Tamehiro et al [1-8]. For such simulation studies accurate numerical data on the behaviour of the mean and low frequency wave forces are desirable, so that meaningful results can be given regarding the systems under inves 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 study such a theory is developed based on potential theory. The final expressions are valid for all six degrees of freedom and are obtained through direct integration of the fluid pressures acting on the instantaneous wetted surface of the body. The final expressions are evaluated using an existing computer program based on three-dimensional linear potential theory. Numerical results are compared with analytical results obtained for a simple shaped body using a different theory. Experimental results for different, more practical shapes of vessels and structures are compared with results of computations. It is shown that the expressions obtained for the mean and low frequency second order wave forces can be used to gain more insight in the mechanism by which waves and

structure interact to produce the forces. It is also shown that the insight gained using the method of direct integration can be used to enhance the positioning accuracy of dynamically positioned vessels in irregular waves. This is effected through the use of a wave-feed-forward control signal based on the instantaneous re lative wave height measured around the vessel.

II. PAST DEVELOPMENTS CONCERNING THE COMPUTATION OF MEAN AND LOW FREQUENCY WAVE FORCES iil:_._i{_.t_r2*lu_^22 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 FroudeKryloff 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 evalui 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 L j z S k e n 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 hydrodynamic 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. Hi 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 twodimensional 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. 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

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.

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.$ with: * = *(X,t) (III-l) (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, - the potential: * = e $ ( 1 ) + e 2 * ( 2 ) + 0(e 3 ) - the wave elevation: el (III-3) for instance:

. = .(0) + _. U)

_V2) + OU 3 )

(III-4)

- the motion of the object: X = x(0)

+

_x ( 1 )

+ E

x(2)

0(.3)

(III-5)

where the affix denotes the static value, indicates first (2) order variations and 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 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 (III-6)

series with respect to a small parameter E only be carried out up to and including the second order. In order to develop a consis tent theory it is assumed that the low frequency motions induced by the low frequency second order forces are of second order and small in relation to the first order motions. From measurements of low frequency motions of moored vessels it is known that in the range of practical wave heights this assumption is in some cases strictly speaking incorrect. This is due to the large dynamic mag nification of the motions resulting from small damping for low frequency motions. The theory, however, is based on the assumption of infinitesimal wave height (e--0) in which case, in spite of low system damping, the low frequency motions induced by the low fre quency second order wave drift forces are always small in relation to the first order motions. In this chapter the hydrodynamic boundary problem for the potential $ will be formulated to first and second order. If the potential $ is known the pressure in a point in the fluid may be determined using Bernoulli's equation: P = P 0 " Pg* 3 - P*t ~ %p | V* | 2 + C(t) where: p X^ t p g = atmospheric pressure = vertical distance of the point below the mean water surface = time = mass density of the fluid = 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 S and for the moments: M = -// p. (X x N) .dS S (III-8) (III-7)

C(t) = a function independent of the co-ordinates

(III-9)

_>.

in which: p S N X = fluid pressure = total wetted surface of the body = outward pointing normal vector of dS = co-ordinates of dS The numerical method used to finally evaluate the fluid forces and moments will be discussed in chapter IV. Ill 1 . 2 i_Co-ordinate_sys terns Use is made of three systems of co-ordinate axes {see Fig 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. axis vertically upwards in the mean position of the oscillating vessel. The surface of the hull is uniquely defined relative to this system of axes. A point on the surface has as position the vector x. The orientation of a surface element in this system of axes is defined by the outward pointing normal vector n.

dS = a surface element

Fig. III-l

Systems of co-ordinates.

The second system of co-ordinate axes is a fixed 0-X.-X--X, system with axes parallel to the G-x.-x--x, system of axes with the body in the mean position and origin 0 in the mean free sur face. The third system of co-ordinate axes is a G-Xj-Xl-Xi system of axes with origin in the centre of gravity G of the body and axes which are at all times parallel to the axes of the fixed 0-X1-X 2-X_ J 3 system. The angular motions of the body about the body axes are de noted by the Eulerian angles x., x 5 and x^. m^.3^_Motion_and_velocity_gf _aj2oint_gn_the_hu^_of_the_body If the body is carrying out small amplitude motions in six degrees of freedom under the influence of oscillatory first order and low frequency second order wave forces the position vector of a point on the hull of the body relative to the fixed system of 0-X -X2-X-. axes is: x = x ( 0 ) + _x ( 1 ) where X
+

_2X(2)

(111-10)

denotes the mean position vector with: X<> = X ( 0 ) + x (III-ll) g denotes the first order oscillatory motion with:

and X

(D

(D g

+ H

(D

(III

.12)

where a is the oscillatory first order angular motion vector with components xi , xi and xi respectively and X^ is the oscillatory first order motion vector of the centre of gravity of the body. Similarly the second order low frequency motion is: X<2> = X<2>
+

a(2) x x

(III-13)

where a (2. is the low frequency second order angular motion vector

(2) with components xl

(2) , x , -

(2) and x,

(2) respectively and X is the

low frequency second order motion vector of the centre of gravity. The velocity V is: .l . 2 f 0 . V = X = EX^ ' + X 1 ' where: x"(1) = V ( 1 > = * ( 1 ) g
+

(111-14) (111-15)

^{1) x x

and

, 2)

= V ( 2 ) = *(2> + ^ ( 2 ) * x

(IH-16)
and

in which the components of the angular velocity vectors a a are x^ , x^ , x^ and x. , xi and x^

respectively.

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 pointing normal vector n. Relative to the fixed system of co-ordi nate axes 0-X.-X--X, and the G-X'-X'-X' axes the normal vector of a surface element becomes: N = N(0) +
EN (1)

+ _2N(2)

(111-17)

where it is found that: N(0) = n N(1) = aCl) x n -(2) -(2) N l z ' = aK s x n (111-18) (111-19) (111-20)

miii_myi^_iI.2ti2l__^S^_b2yS^^Y_20^iti22^
III.4.1. Boundary conditions within the fluid, at the free surface and on the sea floor The fluid domain is bounded by the free surface, the sur face of the body and the sea floor. Assuming that the fluid is inviscid, irrotational, homogeneous and incompressible the fluid motion may be described by means of the velocity potential *:

= e4(1) + E 2 $ ( 2 )

(111-21)

The potentials are defined relative to the fixed system of 0-X1~ X-X axes with: $ = *(X,t) (111-22)

where t denotes time and X the position vector of the point under consideration. The potential $ must comply with the following boundary conditions: - Everywhere within the fluid domain the equation of continuity must be satisfied or: V2$ = 0 (111-23)

In order to satisfy this requirement to first and second order it follows that: V2$(1) = 0 V2$(2) = 0 (111-24) (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: g^i1* + *iiJ = 0
Att

on X-. = 0
3

(111-26)

The boundary condition is satisfied to second order if:

<+ *__)-2-<l,--_l> + < l, <*x;x 3 + i ^

on X 3 = 0 (111-27) The boundary condition at the sea floor, which states that to first and second order no fluid particles shall pass through this boundary or: V$(1).nb = 0 V$ ( 2 ) .n, = 0
b

(111-28) (111-29)

where n, is the normal vector of a point on the surface of the sea floor. III.4.2. Boundary conditions on the body In general the boundary condition on the body states that 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 passes through the hull. This boundary condition has to be satis fied at the instantaneous position of the body surface and is as follows: V*.N = V.N (111-30)

Taking into account equations (III-3) and (111-14) through (III29) 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: V* ( 1 ) .n = V ( 1 ) . n (111-31)

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: 7_ ( 2 ) .n= (V (1) - V* ( 1 ) ).N ( 1 ) + V ( 2 ) . n (111-32)

Equations (111-31) and (111-32) have to be satisfied at the instantaneous position of the surface of the body. Assuming that the motions are small and applying a Taylor expansion similar con ditions may be posed on the potentials at the mean position of the surface. The first order boundary condition becomes: V* (1) .n = V U ) . n The second order boundary condition becomes: (111-33)

V*(2).n = -(XC1).V).V*(1).H+ (V(1> - v V ^ . N ^

+ V(2}.n

(111-34)

where the additional term in equation (111-34) arises from the second order correction to equation (111-31) when applying the Taylor expansion to the velocity V$ In equation (111-33) and (111-34) the potentials and their derivatives have to be taken at the mean position of the body. We may decompose * in the following way: . (1 > = _(1> w
+

. j 1 ' + .<1J d b

(111-35)

in which $ is the first order potential associated with the un disturbed incoming waves. Substitution of equation (111-35) in boundary condition (111-3 3) gives the following: (V$ (1) + V$< 1} + V**1*)-n = V ( 1 ) .K w d b Since the expression is linear * into two parts: V*i1J.n = - 7 * ( 1 ) .K d w (111-36)

we may decompose this equation

(111-37)

and:

V** 1) .n = V ( 1 ) . n b

(11138)

Equation (11137) defines the diffraction potential < J x which compensates the normal velocity components due to the undis turbed incoming waves at the motionless body surface. Equation (11138) defines the body motion potential *. which must be in troduced in order to satisfy the first order boundary condition on the body oscillating in still water. From the solution of the dif combined with the potential of the undis (1) turbed incoming waves $ the socalled first order wave exciting W (1) forces are found. The body motion potential $^ is used to deter mine the hydrodynamic reaction forces known as added mass and damping for unit amplitude acceleration and velocity of the body motions. From the first order wave exciting forces the added mass and damping coefficients and the equations of motion of the body, the unknown first order motions and hence the total first order potential $ can be determined; see for instance Van Oortmerssen of equa [1112 3. Substitution of the first order potential $ fraction potential $!:

tion (11135) in the nonhomogeneous second order free surface boundary condition of equation (11127) shows that the second or der potential has, in general, the following components:

* (2) = *<2)
+

^V

*iV + *T
bb ,(2) bd
+

* ( ^ + ih ) +
wb db (HI39)

ww dd X2) + ,(2> dw bw

wd (2) 0

where the first nine components on the righthand side are poten tials which are particular solutions to the following type of boundary condition at the mean free surface, e.g.:

g*(2> ww x3

*<2> 2V* U) .V* {1) w w w tt t

*(1)(*(1) w t wx3x3

i*Cl) ) g w ttx3
(11140)

(2 ) The last potential *' is a potential which satisfies the homoge neous boundary condition: 9*n 2) x 3
+ $

n2)
fct

(11141)

(2 . $x is therefore an "ordinary" potential which satisfies the lin earized free surface condition. We will simplify equation (111-39) by putting: .(2) = .(2) + *<2)
W U

(111-42)

in which $ (2) represents the sum of the first nine components on W (2) the right-hand side of equation (111-39). ** ing wave potential $ . We will decompose 3x may be regarded as as follows: the second order equivalent of the first order undisturbed incom

*n2' = *i 2 )
0 a

*H2>
b

f" 1 - 4 3 '

Both these potentials satisfy the linearized free surface condi tion of equation (111-41). Substitution in equation (111-42) gives .(2 = .(2> w
+

. < 2 ) + *' 2 ) d b

(111-44)

Substitution of equation (111-44) in the second order boundary condition (111-34) gives:
(2) 2} 2) C1} C1) (V$ + V$i + V** w a b )-n = -(X .V).V$ .n +

+ (V C 1 ) - V ( 1 ) ) . N ( 1 )

V(2).K (111-45)

which may be decomposed in:

v . ' 2 ) X = x v . ( 2 ) - (x ( Xv>.v* ( 1 >}.n

d w + (V C 1 ) - V $ C 1 ) ) . N { 1 ) V $ ^ 2 ) .n = V C 2 ) .n

(111-46) (111-47)

Equation (111-46) defines the second order diffraction potential (2) Sx which firstly compensates the second order velocity components of $ (2) and the second order correction to the first order velocity V* , which results from the first order motion X ' in

a direction along the normal n to the surface and secondly the second order velocity component of the difference between the first order velocity V fluid velocity V* N of the body surface and the first order in a direction along the first order normal

. From the solution of the second order diffraction potential (2) (2) $ combined with the undisturbed second order potential $ the low frequency second order wave exciting forces are found. Equa (2) tion (11147) defines the second order body motion potential $/ tion on the body carrying out low frequency second order motions in still water. This potential satisfies the same boundary condi

which must be introduced in order to satisfy the boundary condi

tion as the first order body motion potential *.;

. The only dif

ference is that the motions are low frequency and of the second order in magnitude. The same techniques may therefore be employed in solving $i as used in solving $i . This means that $J; may be expressed in terms of hydrodynamic reaction forces for unit amplitude of motion velocity and acceleration of the body, known as added mass and damping. better

III.4.3. Boundary conditions at infinity For the potentials *^ 1 ) , *^ l j and *(![2)# *^ 2 ) a radiation condition, which states that at a great distance from the body the waves associated with these potentials move outwards, must be sat isfied. This restriction imposes a uniqueness which would not oth (2) erwise be present. Since the components of $ are particular solutions to the free surface boundary condition (11127), which is defined over the complete free surface, a radiation condition need not be imposed. IIIi5i_Pressure_in_a_gint_within_the_fluid If the velocity potential * is known the fluid pressure in a point is determined by Bernoulli's equation: P = P 0 " pgX 3 p* t *fp|V*|2 + C(t) where: (11148)

_n

p X3 $ C(t) t p

= = = = = =

atmospheric pressure vertical distance below the mean free surface velocity potential a function independent of the co-ordinates time mass density of the fluid.

In Bernoulli's equation p Q and C(t) may be taken equal to zero without loss of generality, see ref. [III-l]. Assuming that the point is carrying out first order wave frequency motions X and low frequency second order motions X about a mean position X and applying a Taylor expansion to the pressure in the mean position the following expression is found: .p- p(0) where: - hydrostatic pressure: P ( 0 ) = -pgx<> - first order pressure: p ( 1 ) = -pgx'1' - p.*1' - second order pressure: (111-51) (111-50)
+

_p(1) +

(2)

(IH-49)

P ( 2 ) = X P | V * ( 1 > | 2 - p . , ! 2 ' - p t x ' X v . ' 1 ' ) - pgX<2>

(111-52)

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 31

the potentials are taken at the mean position of the hull which is alternately within and outside the actual fluid domain. This ap pears to be permissible if the potential functions are sufficient ly "smooth" at the boundaries, see ref. [III3]. This is assumed to be satisfied in this case. IIIi^i_Second_order_waye_force_and_mgment III.6.1. Second order wave force In determining the second order wave force consideration must first be given to the choice of the system of axes to which will be referred. Since in general we are concerned with the slow wave drifting force induced motions of bodies in the horizontal plane we have chosen to determine the wave drifting force along the axes of the GX'X'X' system of coordinates. See Figure IIIl. The fluid force exerted on the body relative to the GX' 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 centre of gravity G of the body, follows from: F = // p.N.dS S (11153)

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, 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:

_ _

P = - J . (P <0) + BP (1 ' + E 2 p ( 2 ) )(n + E N (1 > + E 2 N< 2 ')dS S 0 -/J (p (0) + E p U ) + E 2 p ( 2 ) ) ( K + EN(1> + E 2 N ( 2 ) )dS
S

(111-54) = F(0)
+

EF(1) +

E2F(2) +

0(e3)

(111-55)

The hydrostatic force F follows from integration of the hydro static pressure p over the mean wetted surface S : ?(0) tt vx; (0) UJ F^' = pg jj .n.dS = (0,0,pgV) J ; S, o The total first order oscillatory fluid force F (111-56) follows from:

F ( 1 ) =-// ( p U , X + p ( 0 ) .N ( 1 ) )dS (111-57) S 0 = -JJ p ( 1 ) .n.dS + a ( 1 ) x (0,0,pgV) (111-58) S 0 The first part of this expression is the total first order fluid force relative to the body axes G-x -x2~x . The second order force is found by integration of all products of pressure p and normal vector N which give second order force contributions over the con stant part S n of the wetted surface and by integration of first order pressures over the oscillating surface s: F ( 2 ) =-/{ ( p U ) . N ( 1 ) + p (2) .n S 0 -// p (1) .n.dS s Taking into account that: N(1) = a(1) x n (111-60)
+

p (0) .N (2) )dS

. (111-59)

Since angular displacements are the same for all surface elements dS, the first part of the first integral becomes: -// p C l ) . N U ) . d S = a U ) x -// p(D .H.dS S S 0 0 (111-61)

The integral in this expression corresponds with the first term in equation (11158) which is the total first order fluid force rela tive to the body axes Gx.x 2 x 3 . Equation (11161) indicates that a second order force contribution relative to the GX'XIX' sys 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 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: a(1J x (0,0,mg) = a ( 1 ) x (0,0,pgV) (11162)

Adding this component to equation (11161) gives: a ( 1 ) * {"// p U ) . n . d S + a S

U )

x (0,0, pgV) } = a C l ) x F ( 1 }

(in-63:
where F is the total first order fluid force including the hy

drostatic restoring force, the wave exciting force and the hydro dynamic reaction force. See equation (I1158). Consequently, ac cording to Newton's law, we may put: FCl) = M.X(1) from which it follows that: a' 1 ' x F ( 1 ) = a ( 1 ) x (.f,
(IJI_6S)

(11164)

The second part of the first integral in equation (11159) involves straightforward integration of the pressure p (2) as given in equation (11152). The third part of the first integral is a second order hydrostatic component: // p ( 0 ) . N < 2 ) . d S = a<2> x /J p <0>.n. dS so = a
(2) S

0 (11166)

x (0,0,pgV)

The second integral in equation (11159) over the oscilla ting surface is solved by substituting p
_ A

from equation (11151)

and writing the surface element dS as: dS = dX3.dil Also taking into account that at the waterline: P*t1J
=

(11167)

P9CU)

(11168)

this integral becomes:

c(1)
/ J (1) X3WL (pgX. + pgc(1))n.dx .dl (11169)

which r e s u l t s i n :
- / fcpgC*11 WL .n.d* (111-70)

in which c

is the relative wave elevation defined by:

41' = ^(1) -*__.

2 F ( 2 ) J fcpgt^.n.di + a U ) x (M.*(1>)+ g WL

(III 71)

The final expression for the total second order force thus becomes:

// { ^ P | V * ( 1 ) | 2 P*t2) " P(x(l).v*1!l))}K.ds + S 0 // pgX*2).n.dS + a ( 2 ) x (o,0,pgV) S 0 (11172) III.6.2. Second order wave moment The moment about the axes of the GXjX'X' system of co ordinates follows from: M = "J/ P (X1 * N) .dS S (11173)

35

The derivation is analogous to that followed for the force. The final expression for the second order wave moment is: M ( 2 ) = - / *_pg<;(1) . (x x n).dl r WL
2

+ a ( 1 ) x (l.a (1) ) +

-jl
.

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

(x x n)-dS - // -pgX* 2) .(x x n).dS
S

0 (111-74)

Equations (111-72) and (111-74) give the total second order forces acting on a vessel, thus including the wave exciting force and the hydrodynamic and hydrostatic reaction forces. In most cases prime interest is focussed on the second order wave exciting forces and moments. It will be clear from the aforegoing that the second or der hydrodynamic reaction forces are contained in the contribu(2) tions due to the total second order potential * . The hydrostat ic reaction forces are contained in the last parts of equations (111-72) and (111-74). Taking into account equation (111-44) the second order wave exciting force and moment become:
(1 F (2) = - / W 1 , r2 . n . d * + a U ) * <M.S >> + g

WL

-// { - % P | ? ( 1 ) | 2 -p(*< 2 > SQ t - p(X ( 1 ) .V*^ 1 } )}n.dS

f >) d t

(111-75) 2 M ( 2 ) = - / J _ p g C U ) . (x x n).dl r WL SQ t + a ( 1 ) x (i.a (1) ) +

2) 2) -// {-%p|V (1) | 2 " P(*J, + *i ) + W d

- p ( X ( 1 ) ,V#^ Z ) )}. (x x n).dS (111-76) The hydrodynamic reaction forces due to motions induced by second order forces may be expressed in terms of added mass and damping forces as has been shown in the aforegoing.

w
H P O P H H 0 p, CL fD H

ro
P 0 Hi tn
ti

rt 3" P tn

rt P 0 3

rr ^.
P
TJ M rt 3* fD Pi p 3 d 3" P O 3 H 3 p rt D. 3 ^Q

ro X
c n

c n r+ P

Pi fD 3 rt

Hi O

ro H
c n
p 5 P <3 fD

rt 3 fD 0 Hi

& ^ < 3
g
P
H

c n c n ro 0
0 3 Pi

c 3 ^ H
H 0

fD O 0 3 Pi

tr
0 Hi

a P
c n
'

0 H

ro
ti

ft fD 3 rt P P P1 H H H 1 to ^]

3
c n
P pj 0 ft 3 rt 3* P 0 rt P 0 3 0 Hi 0 H Pi

3
c n
H 3

fD H 0 3 tn rt C tJ 0
el t3 3 fD

o_
rt 3" fD 0 H Pi

c n 3

3 C 3

ro >< o 0
o in n ro c
P
0 Hi 3 Pi

a p
o c n rt c n ro a

ro o ff pj ro 3 " 3
rt 0

o pi
P 3 rt 3J

ro 3
rt 3* H 0 3

Hi P H

TJ H Pi 0 rt p 0 tu pi 0 P C rt p 0 3 p 3 P rt p 0 3

TJ 0 ft fD 3 ft p P M

IH IH IH 1 lJ 1' 1

rt 3 P rt fD CL

tr
Hi H 0 Hi rt tP id H Hi tu rt 0 P 0 fD 3 3 0 p 3 Pi rt 3" 0 H Pi C D H

3 C l_. H Pi ip (U H 3

fD tQ C P rt p 0 3

tn P ri p

ro H
0 H Pi

3" 3 P lQ C fD tn

ro tn
rt 3J rt 3* rt tr Hi 0 p* M 0

o fD
Pi

o rt c
g

TJ 0 rt fD 3 ft P P H TJ H 0 P 3 C D

= < fD
c n c n P

c fD < e n ro Hi H ^ ~ K: H ^ tr 0 H
C T ) fD 0 0 3 p_

ro ro
T3 0 rt fD 3 rt P Pi1 I

*. ro 3 c Cu
3 rt P rt p Hi 0 H O

ro H
c n c n H, T) ro fD 0 0 H s p M O ii o ro 3 0 0 c n c n rt C D 3 3 Ip c n rt P p c n & 0 H 0

ro ro ro
*z
tr
3 Pi ,

fD X f. p ft p 3 ^P p 3 pi

ro H

ro ro X
O

s' ro

n ro c n o c c n tr P ro C D M c n o fD 0 ro rpt 0

10 13 IO IH IC icn IP 10 13 icn

O 0 3 rt H p

H rt ro r3t" 1 3 c n to 0 c ro ro a. E U H ft 3" tr <! ' P fD Q 0 Di 0 g c 3 H 3 0

o ro
c n
<e- tr er M ro ro i 0 c n

13 H P 0 rt P 0 (D
OJ

tr c rt
DJ H tu tn p

. H <
fD X P O 0 3

TJ 0 rr C D 3 rt H P H

<
c n c n
3 0

tr
3" P <P fD 3

a p
*-. H
H M 1 J Ul

tr o ft c p rt
rt 0 ft 3* p 0 3 tn 0 3

n pi ro H

p 0 3 tn

tu
0

H H 1 4*. cn Pi Pi H

ro g 0 ro c tn PJ
c n fD
P H 3

c 3

^ 0 tr 3 K p tr ro p 3 0

Pi

tr ro c n & p ro o Hi p tr rt Hi rt o 3
Pi 0 P. Pi

3
H 3

3 Cu ro ro
3
PJ

TJ I1 C D

& H
c n

Hi H

Hi Hi H tu 0 rt p O 3

rt p 3 ifl

< n 0 ro
c n
3

tu 3 rt

0 3*

3 3 P P rt c n 0 < 3 P ft . ro H i P P i c n tr H t, & ro ro ro P p. Cb H P Hi tn tn c n ro 0 r t T . pi Z tn H I P 3 0 3 0 < rP 0 P ro ro t H Pi c n c n P 3 c n ro ro P 3 H X rt ft P 3 o Hi 3 p 3 Pi 0 P*

K Hi P H tn rt 0 K
LQ

n p)
tr tr
Hi H 0 tn 3* 0 3 M

tr fl)
0 3 rt 3 H ft rt 3^ Pi Hi fu 0 (D rt P P H

rt 3 fD TJ 0 rt

rt P 0 3

rt 3" P rt P 3

H P tn p 3 ip p i1 P1

ro pi
0 3 I

P 3 P.

Pi

rt

0 H

P rt

H P O ft P 0 3 TJ O ti

C tn <D 0.

P TJ TJ H O X p

C _ 0 3 rt Eu P 3 tn

c n c n
P

c H n C
0

ro 3 ro ro 3 < tu
o

P rt P 0 3

pi p Hi Hi P

ro

. . M

Pi P

3 P
t

th O H H 0

*d p 3 Pi P 3 iQ

ro *< M ro 3 = c 3

H H
-*1 CTs

3 TJ |
p 3 iQ

ro rt ro H
g p

rt P 3 iQ hh 0 H

0
T3 P H rt

c _

ro cn

S
3 TJ
0 3 rt P P I* Pi P P Hi rt fD H P ft 3 P rt p 0 3

* p rt 3 p 3

rt 3 fD

tr
e*

n a c n ro tn 3 o ft 3 H
3 P" P

TJ 0 ti p ft 3d P rt

rt P 0 3 P 3 ^Q C 3 tn 3 0 Hi 0 3

ro P J ro H 3 fD
tu g

c 3
3
rt 0 P 3 rt 3d rt 3* Pi P H 0 3

< a t o ro

Tl H H 3 O P TJ P_

ro P rt c c n ro 0 ro P.

c n

ro 3
rt 3" ft

*" 3 ro *<
rt 3 P

ro rt
3 0 pi

ro 3 c
P

3 P

rt C H

ro
c n p
3 rt tn rt H 0 3 rt 0 Pi P 3

P tn

0 Hi

ro o rt ro rt
rt
DJ

< fD
ua g tr

rt 3

ro p
s

tr

Hi P

tn

3 0 3 ro tn tr P ro "< *
1
i .

ft P P H

h tn
rt P i i 0

ro c n ro
0 0 tn H 3 TJ 0 1 H rt

p 0 3 tn

c n c n ro c n fD
3

ro ro 3
P ro 3 ro Tl ro 3 ro ro ^ < 3 X
rt 3J

rt 31 C D

rt 3

tr ro
3 I

tr c n C D
3 O O 0 3 Pi P 1

Di C D ft

ro ro H
< ro
H 1 P rt 0 I
rt 0

ft 3

D.
H 1

ro iQ ro c ro pi rt o o H rt 0 3 0 3

*< a K

ro

rt P H

ro 0 n *<

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 transfer functions by means of which it is possible to express the second order wave exciting forces in the frequency domain in terms of force spectra or in the time domain as time histories of second order forces. The components of the transfer functions for the second order forces which depend on first order quantities can be evalu ated using an existing method of computation based on three-dimen sional linear potential theory of which a brief account will be given. The contributions due to second order potential effects will be determined by an approximation using results of computa tions based on the same method. Comparisons of this approximation with some exact results will be given. Only the low frequency sec ond order forces are treated here. The same procedure applies to the low frequency second order moments. IYi2i_The_guadratic_transfer_function IV.2.1. General In this study the total quadratic transfer function is split up in contributions arising from the following components of equation (111-75): I : First order relative wave elevation -^pg M.2 / C* .n.d WL (IV-1)

II

Pressure drop due to first order velocity -j!

S

-^p|V* ( 1 ) | 2 .K.dS 0

(IV-2)

III :

Pressure due to product of gradient of first order pressure and first order motion -// -p(X ( 1 ) . W
S 1 }

) .n.dS

(IV-3)

IV

Contribution due to products of first order angular motions and inertia forces a ( 1 ) x (M.X ( 1 ) ) g (IV-4)

Contribution due to second order potentials -// -P ( *w 2 ) w SQ t

+

*i 2) )-n.dS d t

(IV-5)

The procedure to obtain the quadratic transfer functions of the forces dependent on first order quantities (I, II, III and IV) will be illustrated by taking the low frequency part of the longi tudinal component of the force contribution due to the relative wave elevation: 2

F[2)

F<2)(t) = - J hpgK^
WL

(t,l)

.n

(D.&l
(IV-6)

in which: C (t,) = time dependent relative wave elevation in a point I along the waterline n ( S l ) = direction cosine of a length element dJL in longitudi nal direction. In irregular long-crested waves the elevation, to first order, of the incoming undisturbed waves - referred to the mean position of the centre of gravity of the floating body - may be written as: (1) C Mt) =
N

Z
i=l

(1) c \1} .cos(u.t + E. )

x

ii

(IV-7)

39

The first order relative wave elevation at a point on the waterline of the body may be written as follows:
1] tlr

(t,) =

Z ^ll) .r,ll) 1 = 1 ^ 1

' U ) .cos{w t + E. + E ^ ()} x 1 (IV-8)

in which: c,. e _ . O J. - amplitude of i-th regular wave component = random phase uniformly distributed over 0 - 2TT = frequency of i-th component relative wave elevation at point in the waterline e (} = phase angle of the relative wave elevation at point related to the undisturbed wave crest passing the cen tre of gravity. Substitution of (IV-8) in equation (IV-6) leads to: F.l/;(t) =
1

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

Z Z Ci=l j=l
N

-C- i ; .P._.cos{U. -to.)t + (E, - EJ. ) } + x J

J J

N (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

+ high frequency terms (IV-9) where P. . and Q. . are the in-phase and out-of-phase components of the time independent transfer function with: P.. = P(to,,co.) =
1=1 J

/ ^pgc; U).tl ( ) - C O S { E () + r WL i j i - E r _ ( } } n i U ) -d
3

(IV-10)

Q*+ = Q(w. ,to.) = - / ^pgCi U)-Cl U J . s i n U ^ () + iJ r ^ J WL i j i - E r> (A)}n 1 ().d

1 1

(IV-11)

/in

Taking the low frequency part of the square of the wave elevation given by equation (IV-7) results in: C
U )

(t)0=
1

N N Z Z ^C n U ) .c5 1 ) .cos{(D. -u.)t + (E. - E . ) } x ] J J i=lj=l IV-12)

Comparison with equation (IV-9) shows that P,. and Q.. are trans fer functions which give that part of the wave drifting force which is in-phase and out-of-phase respectively with the low fre quency part of the square of the incident wave elevation. It will be clear that similar developments can be made for other contributions to the wave drifting forces which depend only on first order quantities. The total in-phase and out-of-phase transfer functions are found by simple summation of the contribu tions from the five components. The wave drifting forces may thus be presented as transfer functions which, as can be seen from the aforegoing, are a function of two frequencies. In general the qua dratic transfer functions will also be functions of the direction of the waves. IV.2.2. Evaluation of the components dependent on first order quantities Evaluation of these components of the quadratic transfer functions of the low frequency wave drifting forces requires de tailed knowledge of the first order vessel motions and fluid mo tions. As can be seen from equations (IV-10) and (IV-ll) knowledge of the first order amplitude and phase transfer functions as a function of the wave frequency are sufficient to evaluate these components of the quadratic transfer functions. A numerical method by means of which such detailed infor mation may be obtained (using a distribution of sources over the mean wetted surface of the body) has been developed by Boreel [IV-1] and Van Oortmerssen [IV-2], A brief description of this method is given in Appendix A. The computer program DIFFRAC based on the theory given in this appendix has been used to evaluate 41

these components of the quadratic transfer functions. IV.2.3. Contribution of the second order potential IY^.2^3il^_General When a body is floating in a regular wave group consisting of two regular waves with frequencies O J . and u. part of the second order wave exciting forces are due to second order velocity poten2 )as was already indicated in chapter III. The tials $ { 2 ) and $ ( ' W d (2) second order potential < i > has to be determined taking into ac count the following boundary conditions at the mean free surface:
(2 g*w > + (1) 1) * (2) _ Wt wtt = - 2 _ V + l ) * lt - ( 4 lX l 3X3 + ^g * t tttX x ) 3

' 3 (IV-13) in which $ represents the total first order potential contain

ing the contributions from the undisturbed incoming waves, dif fracted waves and waves due to first order body motions. The cor responding boundary condition for the second order diffraction (2) is as follows: potential * m

g$i 2 ) d x3

$i 2 > = 0 d tt

(IV-14)

(2) also has to satisfy The second order diffraction potential $, the following boundary condition at the surface of the body in the mean position: V . ' X n = -V. ( 2 ) .n - { ( X ( X v ) . V _ ( 1 , } n d w
+

<V(1> - ( 1 ) ) N ( 1 > (IV-I5)

Besides these conditions other boundary conditions discussed in chapter III have to be satisfied. However, for the discussion in this section only the above conditions are relevant.

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 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 (2) chapte chapter III, the potential * can be split into the following parts:

*:2) * * *__> + <v * <v <v

in which: (2) * ww

= second order potential associated with the undis

c

turbed incoming first order wave potential

43

(2) x(2) , .' ' $ (2) ' wd

= second order potentials associated with the outgoing

diffraction waves and waves due to body motions (2 ) $, , = second second order order potentials potentials associated associated with with in interactions 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 of sources is necessary. The general procedure using this numeri 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-homoge44

neous free surface condition equation

(IV-13) leads to a set of

simultaneous equations which are linear in the unknown source strength of the free surface sources which is solved by standard (2 ) techniques. This solves the unknown potential $ . Having solved W (2) (2) * J > the right-hand side of the boundary condition for < _ x at the mean position of the hull of the vessel equation (IV-15) can be {2) solved by the method described in Appendix A. evaluated and Sx (2) (2) Having solved both potentials $ and $-. the contribution of c J w a these potentials to the second order forces can be evaluated by means of equation (IV-5). The procedure given here indicates that for the second or der potentials results can be obtained using numerical approxima tion techniques. It will be appreciated, however, that the compu tational effort to obtain results for this contribution will be considerable since the above procedure must be repeated for all relevant combinations of frequencies ux and o s _ . Advantage can, how(2) ever, be gained of the fact that the basic solution of < J x need only be obtained once for every series of combinations of ax and ax which yield the same difference frequency m - ux. For the pre sent work a different, more simplified approach has been followed in order to approximate the contribution of the second order poten tial to the second order forces. IYi2I_3_L2_L_Aproximatign_f or_the_contribution_of _the_second_order potential The approximation is based on the assumption that the major part of the low frequency second order force due to the second or der potential is the wave exciting force component due to the con(2) tribution $ of the undisturbed incoming waves to the second orww der potential, thus assuming that the first order diffraction and body motion potentials 0, and 0, are small relative to the un b c i (1) disturbed wave potential $ . This means that in the right-hand side of the free surface boundary condition of equation (IV-13) ' of only terms involving the first order velocity potential *
t

the undisturbed incoming waves remain. The second order potential which satisfies this boundary condition and the boundary condition at the sea floor of equation (111-29) as well as the equation of 45

continuity (11125) has been given by Bowers [IV3]. We now con sider a regular wave group travelling in the positive X direction consisting of two regular waves with frequencies _j. and to, with ox > a). The first order velocity potential associated with these waves is: <J> w
1 J

2 c ( 1 ) g cosh k (X th) = Z 1 ,\ ._ t o . cosh k.h

sin(k.X_ cu. t + E . ) ' i l l 1' (IV17)

The low frequency component of the second order velocity potential associated with these waves is as follows: , 2) 9( ww _ ~ - .I h -r < u > AA .. . n rC. .L . *i 3 13
2 c o s h { ( k k i - ;n )(x3+ rn cosh(k. k . ) h h ) }

sin{ (k i k . )X X ((ix a).) t + (._ e )} IV18 in which A . . is a coefficient depending on to. , t o . and the w a t e r ij i 3 depth h: A. . = ^ g 2 (OJ. t o . ) 2 i J B. . + C. . 2J 3J (k k . J g tanh(k. k )h i J i J (IV19) in which: i 2 k. ^ = oi. cosh B. . i , 2 k. tu. cosh a

k.h

k .h (IV2 0)

ij

2k.k. (to. cu.) (1 ttanh k.h tanh k.h) 3 i J _ 3 ^uij (IV21)

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.

DUE TO $

(2)

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: c o = kg tanh kh (IV-2 3)

The incoming waves due to the low frequency second order potential have a wave number equal to k. - k. and wave frequency equal to 47

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 c o . - c o - of the second order waves by selecting a different value for the constant of gravity: (toi -co. ) ( k i - k . ) tanh(k -k.)h

ij

(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 frequency t u . - c o by simply applying the factor:

ij

<IV"26>

48

to the initial force. This does not complete the transformation, however, since, besides satisfying the requirement that wave num ber and wave frequency be equal, the amplitudes of the potentials must be equal. After the alteration of the constant of gravity the transformed potential of the first order regular wave is: C C l ) g.. coshUk. k.)(X +h)} * = T^ ^r T"n^re sin{(k. k .) X, + (oi. -m.) cosh(k. k.)h i 3 1 (u^ OJ. )t + {e e_.) } (IV2 7) The amplitude of the second order potential is given in equation (IV18). Equality of the amplitudes means that:
r(D0
a 9

= cCl)CU)A

(IV28)

This means that the first order wave amplitude must be selected so that:
C(D
a

,C(D (D*n("i-V
J .

(IV.29)

1 J

^13 is determined for a value of unity for

The first order force F r,

. Since forces are proportional to the wave amplitude equation in order to give the required second order force

(IV29) gives a second correction factor which has to be applied

D the t to force F (2) F

(2)
Fr:!

^ MA ^
= n . 2 2J ij

C^ux)
L F^ i; (IV30)

which taking into account equation (IV22) gives:

PJ 2 ) = f j .F ( 1 )
where: fj = ^ ~
L

(IV31)

(IV32)

49

IYi.2i3^3^_Comp_arison_between_the_exact mation It can be shown that this method of approximation gives exact results in two simple cases and gives a reasonable approxi mation for a third, more practical, case. The first case concerns the second order pressure due to the second order potential in undisturbed irregular waves in a point X_ = -a below the still water l^vel. The second order pres sure is: pC2) = - P ^ 2 ) (IV-33)

For the low frequency component given in equation (IV-18) the am plitude of the pressure is: (2) (I) (1) P i 3 = PC^C^A^K-^)
C O s h { (k

i " V ("a + h ) } coa(k3 )h (IV-34)

For the first order potential the pressure follows from: P U ) = -P*^ 11 (IV-35)

The amplitude of the pressure using a first order potential compo nent of the type given in equation (IV-17), unit wave amplitude z, and wave number k. - k. is: a 1 3 , . cosh{ (k. -k . ) ( a +h) } K p ' = pg 7-7^ J: rr (IV-36) M *a ^ coshfk. - k . ) h Using the coefficient f. . given in equation (IV-32) gives the fol lowing approximation for the second order pressure amplitude: , ,-,, c o s h U k . - k . ) (-a + h) } 1, 1, Pi.' - P c l ^ A l j ( 1 - j ) coshdc^-k^h
f9 i M

(IV-37)

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 motions) are in this case zero. 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 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 frequency component of the total second order potential consists of a contribution associated with the undisturbed incoming waves and a contribution due to the outgoing diffraction waves. Since the approximation gives the exact value for the second order poten tial associated with the incoming waves it also gives the exact value for the second order potential associated with the outgoing waves and hence the approximation is also the exact value. The third example concerns the two-dimensional case of a 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 der potential to the low frequency second order sway force in reg ular wave groups in deep water. The method of approximation pre 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. The coefficient f.. of equation (IV-32) becomes for deep water:
^(UJ.

(those due to diffraction and body

f. . = - - ij

-to.) 3g

, . , 1J c- ' i 3

(IV-38)

The results are presented in the form of the amplitude of the low frequency second order forces due to the second order potential for a range of combinations of c o . and LO . which are the frequencies of two waves making up a regular wave group. Faltinsen's results are compared with the approximation in Table IV-1.

51

WAVES

FALTINSEN
0.59 0.59 0.72 0.84 0.95
1 . 12
LO./d/g

0.72 0.02
0

0.84 0.01 0.03

0

0.95 0.02 0.01 0.03

0

1 . 12

co. /d/g

0.13 0.10 0.04 0.06

0

(2)

pgLc^j 1 '

APPROXIMATION
0.59 0.59 0.72 0.84 0.95 1.12
t o ./d/g 0

0.72 0.03
0

0.84
0 . 11

0.95 0.20 0.11 0.04

0

1.12 0.33 0.26 0.18 0.10

0

co. /d/g

0.03
0

.2)

pgi^cj^

Table IV-1

Low frequency drifting forces on a cylinder in beam seas due to the second order potential.

S?

Comparison of the results shows that near the line O J . = cu. the approximation is good but for larger differences between co. and O J . the approximation is considerably higher than Faltinsen' s values. Further study will be required to determine the reason for the large differences which occur at higher frequencies. At the present, however, it may be tentatively concluded that the method of approximation gives the right order of magnitude to the low fre quency forces due to the second order potential for difference fre quencies which are not too large. For the cylinder in beam seas large differences between Faltinsen's results and the approximation occurred for values of the non-dimensional difference frequency greater than about 0.1. 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 semisubmersibles 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 O J . is as fol lows : C ( 1 ) (t) = 2 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:

( 2 )

(t)

2 Z i=l 2 i=l

2 Z j =l 2 I j=l P

( 1 )

c!
J

-1

cos{(u). -

(E, - E . ) }
J

c{1)c!1)Qii.sin{(toi -to,)t J J J + C r
( 1 ) 2

(Ei-e.)} J

= c r

{ 1 ) 2

11

P + '*22

+ C

l1)c_|a)
1 1

CP

12

+ P

21

c o s { (to

i "a32)t

{E

1 ~

) J

+ t^1*^

CQ

12 ~ Q 2 l ' '

s i n {

Cw1 - c o 2 ) t

(j

-E2)} (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 c o 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 floating objects in regular or irregular waves this is often ex 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 c o ,- O J 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 P2 and the difference of the out-of-phase functions Q _ and Q 2 l * (IV-40) it is seen that the transfer functions do From equation

not appear in isolation but rather in pairs. In general, the inphase and out-of-phase components of the quadratic transfer func tions as determined from equations binations of UJ (IV-1) through (IV-4) for com and cu2 will be so that, for instance: P(to2,to1) (IV-42) (IV-40) depends on transfer

P( ( _ ) . _ , ( i ) ) f 2

However, since the force as given in equation

the sum or difference of the components of the quadratic relations are valid: P(to ,to2) = P(u)2,to1) Q(to1,u)2) = - Q ( w 2 , u 1 )

functions these may b e so reformulated that the following symmetry

(IV-43) (IV-44)

The approximation for the force due to the second order potential (2) $ as given in equation (IV-31) is only defined for co. > to.. In order to conform with the definition given in equations (IV-43) and (IV-44) for the quadratic transfer function the in-phase and (2) out-of-phase parts of the force component due to $ become: P.. = ^P. ( 2 ) 13 13 P. . = P. . 31 13
Q

co. > to. 1 3

(IV-45) (IV-46)

ij

^Qij)

>

(IV-47) (IV-48)

Q.. = -Q. . ji *ij In these equations P.. of equation and Q..

represent the in-phase and out-of-

phase components of the second order force as determined by means (IV-31). This transformation was applied to the exam ple concerning the free floating cylinder in beam waves. The inphase component PtuiwUx) of the quadratic transfer function of the total second order force takes the form of a matrix which is sym-

metrical about the diagonal for which co, is equal to LO? while the out-of-phase component Q(to ,LO?) is asymmetrical about the diagonal. The second order force given in equation (IV-40) may also be writ ten as follows: F ( 2 ) (t) = r { 1 ) 2 P + r(1)2 P +

+ 2c| 1 } C _ j 1 5 .T 12 .cos{ (co1 -co2)t + ( E X - E 2 ) + l2 > (IV-49) in which T

T

is the amplitude of the quadratic transfer function:

=

12

T((J 1 ,LO 2 )

= yP2(co1,oo2) + Q 2 (LO 1 ,UJ 2 )

. . . .

(IV-50)

It also follows that: T(to1,U)2) = T(to2,to ) , is a phase angle defined by: tan E I 2 = tan E ( V I O 2 >
e

(IV-51)

Qfui-^tO = - p^J^y

(IV-52)

12 9 i v e s the

phase angle of the low frequency part of the second

order force relative to the low frequency part of the square of the wave elevation of equation (IV-39), which is as follows: C
U )

(t) =

2 2 Z Z ^.ci1,C-1).cos{(co. -io.)t + (E, - E . ) } 3 x J x J i=l j=l

2 + hK{2l) 2 + c{ 1 ) C 2 ( 1 ) .cos{(u) 1 -io2)t +

= HK[1)

+ (e-^ ~ 2 ) }

(IV-53)

__v__3__ FrgyeD__l25}*i-_:_-_:2ES2^ forces_in_irregular_waves We consider the case of a body floating in irregular waves which are characterized by a spectral density or energy density S ( c o ) where: S (oj).doj = ^ (I) 2 (IV-54)

The wave elevation in a point is a normally distributed process with zero mean and a probability density given by:

P ( -<1>,

= _ _ _ . e /2"rrm0

2m

(IV-55)

where:
oo

mn =
U

S (co).dto = area of the spectrum

C

. . . .

(IV-56)

The mean second order force is found by putting OJ. = co. in equation (IV-9): (2) mean
N

(l) 2

(l) 2 (IV-57)

1=1

i=l

Passing from a discrete to a continuous formulation taking into account equation (IV-54) gives: (2) 2) [
L

(t) = 2

S (to) .P(to,Lo) .da.

(IV-58)

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
*
0

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

t.

(IV-59)

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

= frequency of the low frequency second order force P(to+y,co), Q(to+y,to) = in-phase and out-of-phase components of the quadratic transfer function T(to+u,uj) = amplitude of the quadratic transfer function' S(y) ~ spectral density of the force.
J.

Besides knowledge of the mean and spectral density of the second order forces, knowledge of the distribution function of the force is of interest. In general, however, the distribution func tion of a second order force in irregular waves of the type as given by equation (IV-9) cannot be given. An indication can, how ever, be given of the type of probability distribution function involved by inspection of a quantity which is closely related to the second order force. This is the low frequency part of the square of the wave elevation as given in equation (IV-12). It can be shown (see ref. [IV-6]) that the low frequency part of the square of a normally distributed signal is of the exponential type with a probability density function: now m 0 e

P(=low ' = -

. 1.

(IV 61

" >

In this equation nt- is the area of the wave spectrum given by equa tion (IV-56). The distribution function as given by equation (IV61) is shown in Figure IV-2.

= .ft

m0

,r< 1 > 2 .

low

Fig. IV-2

Probability density of low frequency part of the square of the wave elevation.

IVi4^_Time_domain_representatign_of_the_mean_and_low_frequency second_order_forces According to Dalzell [IV-7] the low frequency second order forces can be computed given the quadratic transfer function and the time record of the wave elevation using the following relation ship:

F ( 2 ) (t> = r
oo

r
oo

g'^it^i.^'it-tji.

. ( 1 ) (t -t 2 ) ,dt1.dt2 (IV-62) where: (2) (t ,t) = quadratic impuls response function gv

.
t l
( 1 )

t )

= time dependent wave elevation = time shifts.

,t2

The quadratic impuls response function g (2) (t.,t-) is derived from the following expression:

(2)(

i
2TT

;~r
- -

(iuiti-iu2t2)

(2 )
. G (LO1,OJ2) .dto. dux

(IV63) in which: G (2) ( c o ,OJ ) = complex quadratic transfer function = P(io ,(_)2) + iQ(u)1,uJ2) and: P(co , U J 2 ) , 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 elevation records. The applicability of this technique has been demonstrated extensively and convincingly by Dalzell [lV7] using quadratic transfer functions for the second order forces obtained from tests in irregular waves using crossbispectral analysis techniques. IV.5. Conclusions In this chapter it is shown that the mean and low frequency second order forces and moments may be expressed in the frequency domain in terms of quadratic transfer functions which are depen dent on two frequencies. Physically, the quadratic transfer func tions contain information on the mean and amplitude of the low fre quency second order forces on the body floating in regular wave groups consisting of two regular waves. It has been indicated how these quadratic transfer functions are evaluated and an approxima tion for the contribution due to second order potential effects has been discussed. Dalzell [lV7] has demonstrated that on the basis of the quadratic transfer functions time records of the sec ond order forces can be generated. In this study therefore atten tion will be focussed on the frequency domain results only. If such results can be accurately predicted by computations then time do main results, which are often required for simulations, can also be generated with good accuracy. (IV64)

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 sional case of a free floating hemisphere in infinitely deep water, In order to make a complete comparison results of first order wave loads, added mass and damping and motions are also compared. Y._.2-_.-Commutations V.2.1. General For the computations the mean wetted surface of the body is approximated by 206 facets as shown in Figure V-l.

TOTAL NUMBER OF WATERLINE LINE ELEMENTS: 36

TOTAL NUMBER OF FACETS ON H U L L : 2 0 6

./ _/

Fig. V-l

Distribution of facets on the sphere and the distribu tion of line elements on the waterline. 61

to

p 3

rt 3

3 O
(_o a\

TS P H

0 0

ip p-

3
prt

c n
P P P
tn 3 P. rt 3 C D

P
0 3 H rt 3 rt 3 H
ti ti

a ro c n
3 3 H pi H P is H P3 TJ P 3 p-

ro
3 P C D 3

cn C H rt cn 0 rt p3 tp H H C D rt 0

3 o 3
P H P X P-

rt 3 H 0

ft P H

TJ C rt

H P Pi P-

3 P

ro

Ip 3

c
< ro
3 P3 3 C D P U3 pri

c C g ft ro ro H
P

P H
H

a
ro
p3 ip

< ro
c n < C D C D c n o
0 3 Pi

ro ua pro ro n > c ro
0 Hi H P 3 C D

TJ 3 C D H

P3 O 0

3 ro

iq p-

o
TS H O 3 H C D

o 3
rt 3 C D Pi P X

c p P
P

S g
P
tn tn ft H

c n 3 ro

< ro < ro
ro ro
P 3 Di Pi P

ro 3

^ 0

Hi Ptp

pi

H tn

3 ro c n
0 r+ H P 3

c c n

TS 0

P P

Pi H PHi ft

p3

c n
P

ro 3

0 H pi p 3 tQ

P P 3 UJ '

< . . ro 3 H c H ro < 1 ro tr p < 1

O H Pi 3 P. ^P

H C D Hi

*. Pc n Hi
H C D

g g ro
ft 0 ft 0 rt 3 C D

< ro
H 3

c n ro o o 3
p 3 M C D

n ro
pi C D TJ ft 3

S P ti

4 C D cn rt H P O ft P 0 3

c ro
rt 0 0 0

C O

>U

< C D
c n pP

c g
0 3 P TJ ti

a ro
g
TJ p 3 ip 3 C D P 3 Di p 3 P O

ro

ro

0 Hi H H TJ TJ P O 0

ro ro 3 ro c n
P
3 p ti P
hrj PCP

ro
rt PHi

< C D c *-i
H Hi H P H 0 3 Di p. prt PZ 0

Hi 0 H 0 C D TJ 0

-P C H C D

< ro
ro ro c n
H H H 0 H Pi rf P0 3

TJ H P ft C Pi C D tn

ft 3

P H C D

iQ P-

H 3

rt 0

P-

c n 3
rt 3 0 Hi H 3

a a ro < C D o rt
ro o ro
i

3 C
Di p H, Hi Di C D C D TJ

3 ro
TJ rt TJ C rt

p 3

n 0 C H P g c n ro TJ c n C ro ci
H 3

ro
C D 3

ip P-

c n
K < C D 3

< c I 3
ro 3 c n P c n tn ro
TJ p H C D Pi

ro ro 0 p H 0 ro ro P 3 r t P TJ g n CD p C 3 3 < rt ti ro P ft H c n P S P
P
0 3

< ro
, * H

H ft PO P H

^ <
ci ro ro
0 Hi Hi 0 H P 3 C H p H P H 3 0 H H H
PJ

rt 3 C D

3 0 H PN 0 3 ft P H

<p p-

a.

c n c n ro tn c n ro o ro c n ro
3 rt P rt C D H TJ rt 3

< C D
H

a H
rt 3 H 3

c n < 3 TJ ro 3 3 tr ro f t <

Ptp C H C D

P 3 Pi

0 ft P0 3 to

< i *.
ro ro
p C D H 0 0 0 H Di P 3 Pi P 3 D-

ft 3 C D

< P
c ro ro
tr ro Hi
Hi P H tn rt

rt i{ P 0 ft P 0 3

rt P rt P 0 3 cn

ro ro
Hi P H
r-> C D

< ro
H

ro
H

0 Hi

ro
P
H3 O H

a P
3

pHi ft

P3

< l c pi
p3 tp 4 P

* c Di <
TS 0

Di H PHi ft

< 1 < i c n to c H
P H p P 3 Pi 1 Ul 0 H C D H

O 0 3 rt H P-

-j ft 3 C D 3 0 3 C D

3 ro ro ro c n 3 < TJ P ro
t,

*j P tQ C H C D tn 3 rt p P H tn

o
0 P 3 3 0 ft

ft

g ro ro c n ro \-> <
3 rt

rt 3 C D ft p 0 H Di rt p Ul

< I ro tn i ' ft n 3 3
H 0 C ip 3 H C D <P C H P H

er

P to

Hi 0

> o

tr

p 3 P H K rt P0 P H

H3 3 C D C D P O 3

ro
3 3 cn rt 3

0 3

0 Hi P ft 3

n o ro tn c n

rt 3 C D

rt p0 3 tn

Hi 0 H 0 C D

*j Pip C H C D

s
tc
p

ft 3 H-

P H

3 pH C D

3 13 P3 ip

i C P 3 rt Prt

ft 3 Hi p0 Hi ft 3 rt

c n

P H C D

ip p-

ro ro Hi
P

< 1
ro
p. H P Hi ft H

c ro

C rt C D Pi

< to

O 0 3 rt H P

< ro

rt tn

rt 3 P 3

0 K Pi P 3 C.

< o ro 3
3 C D ro tr ro H ro S g o rpt O rt

ro o
0

tr ro o

O 0

3 rt 3

CO

ft 3

prt p0 3

ro o pC D 3 rt cn

w P

rt 0 ft P H Hi PH C O rt

< C D
<
P
0 Hi 0 Hi b

ip p-

c n
p 3

P O P H

p 0 3 tn

Hi P H tn rt

c p
ro
0 Hi O 0 3

<P P-

> . p S H 3 c P < p p H ro < tn 3 P ro < ip ro ro 3 Hi 3 pi

c n
0 M H H

H 3

< C D
c n pTJ C ft C D Pi O t. P< C D H p3

P 3 Pi p rt p 3 ip

P3

ro ro 3 0 < ro ro 0 3 P ro 3 3

s H . p < 1 ro P tr < < < 3 UJ N 1 ro < ro c r t . cn 0 Pi P ro ro C D H p H C D rt t. H ^* H 0 0 0 ro ro c X 3 tn 3 3 tQ ro a ro < ^ < ro

ro o o 3
ft 3 C D Hi 0 H C _

rt 3 s 3 p Pi ro < 0 ro H s P i ro p ti ro !_, H ro ro H < H Hi P

ro ro 3

0 ft

3 ft Prt

TJ H 0 Di C 0 ft ft 3 C D

. a C D
rt 3 C D Hi 0 H 0

ft 3

ro

P3

0 0

><
c n
cn C 0 Hi rr 3 C 3 Di PTJ 3 P Ip

53 0

ro

^ P-

c n C D
0 C D 3 rt H C D H p ft p0 3 H3 3 C D H C D tp C H P H

< ro CD c n &
c P
P H rt H P

tp rt 3

P ft C H C D

c *
ro tr
0 ft 3 C D Hi 0 C H J3 C C D 3 P

ro g C D
C D ip H P tp P 3 H

p 3

0 H Di C D h

rt P O 3

p 3

0 H

ro
P
3 rt p0 3

H) 0

rt 3 0 Pi

o 3
TJ 3 P 01 C D C H ip

<
H 3 C D H 3

< C D
Uj 0 H ip P

P 0.

0 Hi

H 3 C D

0 0

t.

J2 C P rt p0 3

lp C H

p_

pH P 3 H C D

rt H P 3 tn Hi C D H

ro c n
P P
3 Di 3 C D 3 Pi P 3 <P H

<
ro ro tr ro 3
pi rt C H

H C D

P 3 P H

c n c n

ro

TJ 0 3 C D 3 ft

H 0 3 <P 1 0 H C D

o
0

ro n ro p tn o p P 3 TJ g
TJ 0 3

c n

*< ft

0 Hi

0 H Pi C D H

n ro

ro H i c n c

Hi P H

PJ

ro Hi
H Hi

< C D c n ti 3
P H H ft 3 P cn P 3 Di

H 0 P D.

p 3 Hi P 3 P rt

ro c n

TS C ti

TJ H 0 X P

3 ft

ro

*
ro ro c n o ft

ro 3
pi

c H
ro c n
3

ft tn

TJ O 3 C D 3 rt tn

< i ,^ UD H < 1
ro
P
3 C D P

P0 P H

c H
c n
w

n
rt

O P

ro
H

P rt

c 5 f^ ro

p3 P H C D

H ro

0 Hi

i-3 3 C D

ft 3 C D

c H c n 5 P .->. rpt 0 0 0 ro p 0 c n < r t rt C D H 3 H 0 3 d p 3 TJ p TJ H ro c n < 1 3 3 p p 0 c n 0 c n 0 i-n 3 3 ti 0 H P <P C C D H r t 3 i H c n c n co ro P i p 3 C D 0 D i ro ro ro ro Hi ft H Pi

ro ro
C rt 0 M OJ

c n
P
P

0 Hi ft

c n C D c n ro c n ro
3
0 3 CD

*<
Pi C D rt 3 C D TJ

>
ro
ro
1 H CD

P Di Pi TJ TJ C D 3 1

p 3

ro
Di

Pi C D TJ

ro
pi

ro
rt 3 3 3 Pi

*"
CD

ft 3

O 0

ro

3 0 H PN 0

H C D ip C H P H

3 1

< ro <

ro

>< ro 3
rt

25
COMPUTED

500*

2.0

250

.(1) 1.5 '1a (1>

pvKt; a
1 . 0

0.5

0.5

1 . 0 ka

1 . 5

2.0

2.5

Fig. V-2

First order wave exciting force in surge.

2.5

500

2.0

250

.(1)
3a

1 5 U--* -f

pgAc'J'
1 . 0

0.5

0.5

1 . 0 ka

1 . 5

2.0

2.5

Fig. V-3

First order wave exciting force in heave.

63

1.00
ANALYTICAL COMPUTED

0.75

.
0.50

n/pv

ii/pvu>

0.25

0 0 Fig. V-4

0.5

1 . 0 ka

1 . 5

2.0

2.5

Added mass and damping coefficient for surge.

1.00

0.75
a

33/pV

0.50

0.25
b

33/p7W

0.5

1 . 0 ka

1 . 5

2.0

2.5

Fig. V-5
f,A

Added mass and damping coefficient for heave.

2.5

500

2.0

COMPUTED
250

<-<
(1) 1a

1.5

(1) 1.0

0.5

0.5

1.0 ka

1.5

2.0

2.5

F i g . V-6

First

order surge motion t r a n s f e r

function

2.5

500

2.0

" * ^ ^ _

250"

1.5*3a 0)
1.0

0.5

0.5

1.0
ka

15

2.0

2.5

Fig. V-7

First order heave motion transfer function. 65

oI~~

^S-

1 COMPUTED

-0.25 mean

\
1

Pga

(1)'

-0.50

11

-0.75

-1.00 0

0.5

1 . 0

ka

1 . 5

2.0

2.5

Fig. V-8

Mean second order horizontal drift force.

1 . 0
&r

n
- ! -^*J.J-S-

0 mean

nr m
TOTAL

Pga

(1)2. -1.0

\ \ \ .

nl

-2.0

-3.0

0.5

1 . 0 ka

1 . 5

2.0

2.5

Fig. V-9

Components of the computed mean second order horizontal drift force.

3.0

AN

2.0
y y

//

\
\
\ \

1 . 0 'mean

,. r

Pga C a

/ > '." r > * k

<

\ V v
*
N

i^

m
'i,m

**
1.0

\ ^ \ \ \
0.5

> y

~-A

TOTAL

\
1 . 0

J /

2.0

1 . 5 ka

2.0

2.5

Fig. V10

Components of the computed mean second order vertical force.

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 seen in Figure V7. The sphere is following the wave motion com pletely without creating noticeable disturbance. As the wave fre quency increases the heave motion increases and at the same time the mean horizontal drift force increases. The maximum heave mo tion occurs at a slightly lower wave frequency than the maximum horizontal drift force. In this range of frequencies the effects due to diffraction and body motions on the wave drift forces are increasing. At higher frequencies the body motions decrease con tinuously to become zero at frequencies tending to infinity. In the range of these frequencies the effects of body motions on the drift force decrease rapidly and in the limit only effects due to diffraction remain. As can be seen from Figure V8 the mean drift 67

force at high frequencies is of the same order of magnitude as the peak value and at increasing frequency tends to some limit. The value of the mean horizontal drift force for wave fre quencies tending to infinity is easily found by taking into con sideration that for very high wave frequencies the wave action, due to the very small wave length, is confined to a thin layer near the waterline of the stationary object. In this case the form of the hull may be replaced by a vertical wall which totally re flects the short waves. The circumference of the waterline may be considered as short sections of straight vertical walls by which the incoming wave is reflected as is shown in Figure V-ll.

REFLECTED WAVE

WAVE CREST OF INCOMING WAVE

Fig. V-ll

The mean horizontal drift force in short waves.

The contribution of such a section of the circumference of the waterline to the mean horizontal drift force is due to components I and II of equations (IV-1) and (IV-2) and is found to be as follows: dF. mean The total mean force is found to be: = -i.pgC(1) . (sin2 6)-dX9 (V-l)

'i
mean

-^P9 a

(1)2

+a

. (sin

2 B),dx

(V-2)

-a

This equation has been given previously by a number of authors. See for instance Maruo [V-2]. It is applicable to floating objects of arbitrary shape for determining the high frequency limit of the horizontal wave drifting forces. Taking into account that at high frequencies there are no waves behind the body the following re sults are found for the sphere: 2

mean

2 (li iJ = -f pga^

(V-3)

which means that the high frequency limit of the non-dimensional 2 force given in Figure V-8 equals - ^ . V.2.3. Components of the mean horizontal drift force The computed components I through IV of the mean horizontal drift force are given in Figure V-9. These components cannot be compared with analytical results since the analytical results given by Kudou were determined using a different theory and are given only as a total force. Inspection of Figure V-9 reveals that the total force consists of components which differ considerably both in sign and magnitude. Contribution I, due to the relative 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 ity, is different in sign and of about the same magnitude as the total. Contribution III is smaller and different in sign to the total force and also different in character since it tends to zero at high frequencies. Contribution IV, due to products of angular motions and accelerations, remains zero for all frequencies. At high wave frequencies only contribution I and II remain. This is in agreement with the previous discussion on the high frequency limit of the total force. It is of interest to discuss some as pects of the various components. __:2l_liihy:ti2_}__[ The general equation for this contribution is as follows:
fi.2

*I = - / i.pgCr mean WL

- n - dJl mean

(V-4)

f,Q

The integrand in this case always represents a pressure increase acting inwardly at the waterline. The sign of this contribution in Figure V-ll is in the direction of propagation of the waves in dicating that, as may be expected, the relative wave elevation on the incoming wave side is larger than on the shadow side of the sphere. In general, therefore, if it is seen that an object float ing in waves exhibits a large difference in relative wave height on both sides of the object, it may be expected that this contri bution will be large. _r22thution_II The general equation for this contribution is as follows:
F

II mean

~M -*P|V* (1) |2.n.dS S_ mean

(V-5)

The integrand in this case always represents a pressure decrease acting outwardly on the mean wetted surface of the hull. In gen eral, the fluid velocity V* tends to be largest on the incoming wave side. This results in a mean force component which contrary to intuition, is directed into the waves. 2Gi_ikution_iii The general equation in this case is: F = -// -p(X U) .V*^ 1) ).n.dS mean SQ mean (V-6)

Since this is a mixed product of first order motion and pressure gradient it is not possible, in general, to predict the sign of this quantity. The sign depends on the phase angles of both quan 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 dependent on the motions of the body and on pressure gradients. At very high wave frequencies the motions and hence this component tend to zero, while at very low frequencies it is the pressure gra dients which tend to zero and consequently the force also.

0 0
Pi Hi

H 3 H 3
Pi Pi

P
rf p rt
C C D W 3 0

Hi

Pi

P in

ro
P 3 <

ft

0 rt

TJ H C D

< TJ C D ro H P
O P cn

rt 3 C D

0 H

ro
ft

0 0 3

P 3

P P

>

3 TJ
P3
0 C D

c n
H
cn
rt O
3 C D 3
Pi

rt

\K

o ^ < (1
K ^ 3
rt 0
P 0 rt C N C D H 0

3 C D 3 rt

o 3 ^ <
3 C D H

*< > Ui
c n n 3 ro H P
C D H P
ft

TJ 0 3 C D 3 rt

C D H P H

P 3

n < ro ro H p
c n
ft Di

C D 3

Hi

ro c ro 3
P rt

ft

C D H H P 3 C D

c n p Q 3 c o H p P ro H

(n
C 3

c n
3 C D

Hi

rt 3 rt 0

ro
o P
H
Hi p

H tn
O P

P3 3 C D 3

ft

< C D
c n
C D
ti

ro
rt 3

UJ Ptp C H tn p 3 O C D

< to * ts ro P
< C D pi a ro TJ

ro
c n
C D p 3 II PI
H

ro
P 0 3

u d

P tn

3 O

tp H P

rt P 0 P H

< ro H
3
rt 3
0
ft ft Hi p ft ft

>3 3 C D

N C D H 0

H P 3 C rt P 0 3

<

0 0 3

0 H 0

ua

ro < C D
c n
H 0 3 P 3 P P

ft

0 rt p 0 3

rt 0

S ro
3
C
ti

ro ro n 0 3 < i 6
< C D
c n
H 3
ft ft

a c n
P rt rt 3 C D

3 P ip 3

rt 3

ro 3

i ro H i P P ro H P ro ro tr 3 3 0 3 rt r t t p Di ro ro t(A ro H *< p i 3 a Pi

P
P
PJ

rt
ft Hi

rt O
O P 3

ti

P
3 C D

0 rt p 0 3

P O P H H

P 3 P.

rt P 0 P H H

TJ
Pi

c n
ft kP
PJ

^ <
rt 0
N

H 0

3 C D

3 C D

rt 3 P

o c n c n
p tp 3
p 0 3

a
c n ro c n
0 0 X H C D O

t, ro P 3

H O P H

< 3
0 H O C D

3 in 3 10 ro I13 tt Ip IH IPro 3 13 IC ro H Ift P1 IPP 10 13 ro 1 U3 I H C K P rt p0 3

C 3
Di

ro 3 ro
P p
Hi

ro Di
c n
H

H P 3 C D

Di

rt O
P rt

3 ^ P

P 3

*S
*:
P
C D
Hi

TJ H C D tn

*< & Q c n rt
P

TJ 0 3 C D 3 rt
H C D H P p H 0 3 tp

ro & < a H
rt
3
p

TJ H 0
0 H.
ft

S
C D H 0

ro

" ^,

Hi

tn P
0

Di

rt

s: p
r+ <! 3 C D d c n
Hi ft

3
C D
0
Hi Hi

rt 0

Pi

c n C
K

c C D ro n rt
P
0
Hi

TJ 3 P tn

O H

c ro 3
>
3
p

c 3
P. tn
P 3 tp
ft

P 3 rt 0

N C D H 0

0 rt P 0 3 tn

P. 0

C O

ro H P
pti

s
Di

ft

p rt 3 prt 3
0

X|:

ro ii
t,
3
p

Pi

.
rt 3

Hi

tp

< C D
P
3

H P

*j

t 3 t P
3 ip
M

rt H K
ft

0 H

h
tt
0
ft

C D

H rt

0 H

s: P < C D
rt 3

n 0.
C D

Q 0 3 c 3 P TJ c n 0 ro 5 3 0 ro 3 3 C D n p ro o 3 f t H i 3 ro tn rt
3

tp

h1

3 Ptn Q 0

ft

&
H C D

Pi p

^ ro H > rt ro
tP

e n c n 3
Hi ft

H P H

ro
3
0

ro
c n *
0

h
0

< C D
H

ro
K
H 0 0 0 0 0

rt 3 P tn

tn

ro ro TJ ro 3
0 rt P 0 3 tn

^* 3 ro P
3

3 ^ 0
3

P ri 3

Hi

* P ro 3 < C D rt ro P
uP

0 P H H

*>
p p

0 H

Hi

c p
H C D 3 0

o ft
rt 3
3

c n ro ro h ro H H 0 3 C D P ro 0 p 3 < 3 3 ro rPt TJ 0

c p
3
ft

^
0 3
ii-

ft

3 H
ft p Hi p

3 P ip 3

rt 3 C D

3 0 0
Hi

ro
3
P 3
Pi Pi ft

Hi

3
K
Pi p

S
3 0

C D P tn
N 0 3

<
n
P H
rt
0 H
PJ
ft

S w, ro p c n
H

P O rt

rt 3 C D

< C D ro
P rt P
ri

3
TJ 0 3

ro 3 c
Hi ft pUl

ro P >P c n n 0 < c C D C D
c n c n
P H H 3 H
P Ht ft Hi

0 3 tn

ro 3

TI 0 3

o rf
P H

Hi

ro ro 3 3
ft p

ro P
3 0

ti

<! ro

P
Hi

3
Di

ro H
<
P N

Hi

ro
H 3 C D

e n
3 0 rt

P H

< ro H *< tn
c n
3 P0 H O C D 3 C D tn

c n o

H C D ^P C C D 3 0 P

ro
iQ

ft

a ro H

H tO

c n
3

c n
p

> rt
tp C D 3 C D P 3
p

P 0 P H

n n C D c 3
0 H
ft

ro H ro
p tn 6 3
p-

0 P

o
3
ft

c n

Hi

ro ro c ro 3 0 p P 3 ro O tn ro r t ro H
P 3 C D

ro
rt 3

O H O

^ P
c n
C H H

C D H

ro H ro H
*<
3
C D
Pi

0 M O

P
rt
p ft P

3 0 0

rt in
p-

< ro ii ro
0 0 P H 1

C P 3 rt P rt

ro
0 3 tn

ro
TJ 0 1

P rt
p

< 1 ro o tn 3
tp p
~j

rt 3 C D

< ro

3 TS 6 Hi H O t, 0 f t rf C D H p. P 3 3 3 *Q D H ro C ro ro P C vP c n Pi 3 Hi rt tn r t ro c 3 3 P ro 3 H, i < o ro >S P C D p n H to C ft H p 3 C D 0 ro D 0 c n ro C ft tn 0 o p r t rt tt H 0 rt 3 3 Pi P 0 P C D 3 r t r t ro P 3 ro C D 0 n c n 3 < TJ H ro C D ro ro < D P P P 3 TJ C ft 3 3 C D H H ** 3 0 0 Di p 1 O ro 3 1

S c n 3 < I i

P P rt

0 3 tn

ro

N c n C D TJ Pi fi 3 tn 0 C D H 0 P C D 3 p rt C D 3 P P P 0 3 tn H tp C ro H P 3 H Hi P P tp 3 H C H Pi C D H KQ P 3 P c If 0 Ul rt ro 3 3 P p 0 0 0 p rt 3 p H U) ro C D tn 0 3 P c n d c n 3 H H Pi ft 3 0 0 3 ft tp 0 0 3 C Pi ro p 3 H *< c n 1

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 based on Maruo's theory demonstrates the equivalence of these for mulations with respect to the total force. Inspection of the various contributions to the mean second order force reveals that the total force consists of contributions which differ both in sign and magnitude. From the results found for the mean horizontal force on the sphere it is seen that the general characteristics of the total force are also present in contribution I due to the relative wave elevation (see Figure V-9) although contribution I is greater than the total by about a fac tor 2. In the next section, in which experimental data on the mean horizontal drift forces on stationary vessels are compared with computed results, attention will also be paid to the relative im portance of the various contributions.

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; - a rectangular barge; - a submerged horizontal cylinder. Main particulars and body plans of these vessels are given in Table Vl-1 and in Figure VI-1. The hull forms of the first three vessels encompass the majority of vessels in use by the offshore industry either as per manent storage vessels, drill ships, drilling platforms, semi-sub mersible crane vessels, derrick barges and lay barges. For these vessels the first order motions and mean horizontal drift forces and yawing moment were determined for a range of wave frequencies and three wave directions, i.e. head waves (180 ) , bow quartering waves (135 ) and beam waves from the starboard side (90). The submerged horizontal cylinder is representative of a submersible hovering just below the surface or of horizontal sub merged elements of a semi-submersible. For this vessel the first order vertical motions and the mean second order vertical forces were determined for a range of wave frequencies and the same wave directions as mentioned above.

T*

Tanker Designation Symbol Unit

Semlsubmersible

Rectangular barge

Submerged horizontal cylinder

Calcu Calcu Calcu Tested Calcu lated Tested lated Tested lated Tested lated Scale ratio Length (between perpendiculars) Breadth Draft Displacement volume Centre of gravity above base Metacentric height Transverse gyradius in air Transverse gyradius in water Longitudinal gyradius in air Vertical gyradius in air Natural period of heave Natural period of roll Natural period of pitch Water depth

m m m m m m m m m m sec. secsec. m
3

1:82.5 310.00 47. 17 18.90 234,826 13.32 5.78

1:40 100.00 76.00 20.00 36,925 8.64 16.76 7.92 17.48 30.55 46.60 30.89 41.74 21.8 22.4 19.8 40.0

1:50 150.00 50.00 10.00 73,750 10.00 16.23

1:30 75.60 8.40 12.60* 4,034 4.20 0.00 2.94 2.94

PP

B T V KG GM
k

*K

17.02 77.57

14.77 17.02 77.50 79.30 11.7 14.2 10.6

43.47 30.89

24.16 39.00

20.00 24.20 39.00 39.00 10.4 12. 1 9.4 50.0

xx

16.80 16.80

16.80 16.80

V
k

zz Tz
T
T

11.8 14.2 10.8

21.3 21.3 19.5

12.0 9.3

75.0

0 Wd

82.5

* Distance between base and mean still water surface.

Table VI-1

Main particulars and stability data-

TANKER

SEMISUBMERSIBLE

e
310.0

>
20 tc r>.

sl o
o
20 Ol to

o
O
100.0

O
C

"

Si i
10-6

BARGE CYLINDER

f ^
75.6 150.0

-*-a.

_ _ f

F i g . VI-1

The v e s s e l s .
75

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 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 position. The mean mooring force is equated to the mean second or der drift force. In chapter II and chapter III it was shown that the second order forces are dependent on first order quantities. First order quantities are, for instance, the body motions with frequency equal to the wave frequency. 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 applies to the vessel is a constant force equal and opposite to the mean second order wave drift force. The mooring system shown in Figure VI-2(a) makes use of a falling weight to counteract the mean wave drift forces.

WEIGHT
^~X

a)

SPRING
b )

Fig. VI-2(a) and 2(b)

Systems of restraint

< ro tn
tn rt 3 p Hi H 0 O H cn P H cn rt p 3 Pi C

3 ro
3 H rt 3 Qj cn H 0 H Pi

cn

rt 3 C D

c
P

0 H

ft 3 C D

tp

ro c ro
3 H O 0 3 P TS ti Hi Hi P 0 p

H. Hi

tn

C cn rt

ro
P
p

rt p 0 3 tn

H C D cn

a ro o ro
Hi 0 H O ft 3 C D

c H
ro Z 3

Hi P H cn rt

a ro o ro ro a
C TJ cn C D rt

ft 3 P cn P H C C D

< ro
rt K TJ H cn cn

ro < C D *< cn
rt TJ P cn rt

_ _ < 3 ro CD
3 ro
P
3 H rt P P 3

rt 3 C D tp p P3 3 P cn

H H) 0 H H 3

rt H P 3 tn Pi

tn C D p 0 3 Pi

c 0
ro 3 3 n <
ft 3 C D Hi H 3
1

3 C D

ro ro . 3
pi tn Hi Hi O H O C D H 3 cn 0 Hi rt H 3 P cn 3 P

o
P1 0 rt 3 P 3 Pi

P 3 Pi D. H P Hi ft

0 Hi

P H

*< c n ti ro 3 o ro 3 TJ 3 TJ < ro p
tp 3 rt Pi P cn P Pi

tn TJ H P 3 tp TJ 0 cn p rt p 0 3

Hi p H H p 3 "P

P 3 Pi

H,

3 C D n ro s: C D 3
<
3 P cn rt 3 p cn cn P tp 3 rt

3 O

rt 3 P rt r+ C H P H 0 H Pi C D H cn TJ H P 3 tp

iji

ro ^ H ro
> _ 3
P cn H p Hi rt

P CO p 3 o

Ui H 3 P cn

Hi H p H P

0 H Pi C D H

vQ C C D tn rt p 0 3

ro ro o
H ft 3 P rt 3 P ft rt 0 0 H P 3 tQ

3 P ft C H P H

> 3 ro 3
H P rt C Pi C D

ro a < ^ cn
rt

a
H C D

>. H ro ro ro 3 ro cn
H Hi 3 ft 3 tn H p
LQ

Hi 0 H O TS H C D cn

P cn C H C D Di rt 3 C D P H P

rt 3

rt 0

ro tn 3
0 0 P 3 P P cn 0 Pi P 3 P.

p.

P H P tn C D cn

ro 3

3 C D 3 0 C D

rt 3

rt P

ft 3 C D

ro 3 ro ro cn

< H ro ^< tn Hi

P 3 O 0 H TJ 0 H P rt P 3 tp 0 0 H P 3 tp H C D tQ C P H

Hi Hi

ft 3

< ro ro ro cn a ro

ro
0 Hi 0 Hi rt 3

H 3

P tn

C D ft 3 0

ro ro 3 o
Z
ft 0 P TJ TJ P 3 D

p* 0 rt 0 0

c
ro c n ro 3 3
tn 3 0 cn K cn rt 0 rt H 0 3

0 ft

&
Di C C D

3 ro Q) ro 3 P 3

s; ro

Hi H C D *P C C D 3 O P C D 3 0 P 0 p, C D H ft 0 0 rt p O 3 cn

^< tn .

TJ M p 0 rt P 0 C D P 3 ft P tp P 3

ro tn

TJ H P 0 rt p 0

o 0 3 ft P P 3 cn

TJ 0 cn p rt p 0 3

ro 3 3 P ro c n 3 ft o
3 P H P O ti

ro ro tn 3 a ^ p ro
cn 3 0 3 rt

TJ 0 tn Ul p 3 H C D

3
Hi H rt 3

* 2 3
rt 3 ft P 3 cn P cn rt 3 H 3 rt 3 rt 3

0 H Di C D H

rt 0

O Hi

0 Hi

O C D 3 C D H P H H K

0 cn 0 p H H P rt P 3 ip

tn

.. ro
H 3 P cn

cn rt p Hi Hi

cn TJ H P 3 tp cn

P 3 3 O rt O 0 3 rt 3 p H tn

0 0 H P 3

ro ro 3
P p 3 H C D IP

rt 0

z 3
3 < ro cn ro
0 3 tn H H cn rt cn

C D

ro n ro ro ro cn . H tn
C D Hi Hi 3 C D rt 0

ft 3

rt 3 P 3

. ^ tn
rt

TJ P H rt

ro

ro ro H
3 P 0 3 H H

ro

H H

z p

0 0 H P 3 tp

ro p
3 ip O Hi

ro
3 P cn 3

ua ro
0 ft P 0 3 cn 0 Hi 1 M rf tr rt C H 3 Pi p Hi P H TJ P H ti

^ p
C tn tp C 4 C D

tp C D 3 C D H P H tn TJ H P 3 <P cn TJ H p 3 ip o 0 3 cn rt P 3 rt

C D K C D

0 3 Pi

rt 3 C D

p 3 Pi C P C D Pi

P 3

* ^ 3 o

tn TS H p 3 tp

0 3 rt P P 3 cn TS H p 3 <p cn

ft 3 C D

H C D tn rt H P P 3 rt

Hi H

TJ 0 p 3 rt

ft 3

ro

ft 3 H C D C D

0 H Pi C D H

0 0 H P 3 XI

tp

ro

<

3 ro 3 ro 3 ro n
K* p cn p

ro Z 3 ro CD rt Z H < H ro ro CD H ^< p o ti tp p 3 ro H 3 rt
p. 3

rt 3 C D

Hi 0 H

c ro

rt 3 C D

0 cn o p H P1 P r+ p 3 ^P

c H
ro
0 H

ro
rt 0

rt P

H cn

cn O

rt 3 C D P
ti

ro ro
3

H p cn rt p o cn

* c cn
C D Pi Pi

Hi H C D >D

z 3
< H
X ti p o 3

H 3 C D O P

3 0 o
*<

Hi 0 H 0 C D

ft 3

Cfl K2 tn rt

3 0 p p 3

3 ro
H
\~J

o 0 3 cn rt P 3 rt tn

c ro 3 ro H
P 0 rt 0 ^Q C P H

0 3 P H P O ti T) P H ft

ro a ro

ro a p n rt Ho

3 P o 3

ro c 3
rt 0 H C D tn cn

0 Hi

3 C D . a
rt 3 C D ft 3 H p Hi rt P

3 0 rt

rt 3 P (1 rt 3

ro cn
rt p 3

O 3

^
cn rt ^P C p rt C D

P 3

Hi 0 H 0 C D cn

ro ro 3 cn ro tn
0 Hi cn 3 rt 3 cn
K:

ft 3 H

H p 3 ip

ro ro
<
pi P H C D ft 3 C D C D cn cn

TJ 0 tn p rt p 0 3 cn rt 3 C D ft 3 C D P Hi Hi C D

0 0 H H C D O ft

ro 3 3 < 3 ro
cn rt rf Ul p H

0 3

0 H!

0 0 H P 3 tP

P H C D

0 0 3 cn rt P 3 rt cn

ro s: P < 3 ro o cn o ro
r. C D cn C rt 3 p tn

0 0 3 cn rt P 3 rt Hi P O C M rt H K

3 Hi 0

rt 3 C D

n rt

ro < C D

H P 3 0 0 H P 3 tp

<

0 Di C D H

Z p H H

rt 3

tn

ro . 3 tn 3 ^^ p < ro
H_ 3 P cn

ro
o Hi

0 Hi

P H

3 ro H 3 C D
P 3

ft 3 C D

c o ro n p
Z
3 p o 3 cn 0 C D 0 0 3 rt 3 ft 3 cn K cn ti

3 P tn

3 C D

ro
P

rt 3 C D

< tn

Hi p H Ui ft

P cn ft p o cn 5 C D 3 cn C H p ti 3 cn cn

s P
p tn rt 3 Hi P 3 Pi

ro ro ^P c p < ro
H P 3 H P

ro 3
O 0 H H 0 P. C D P*

ro
3 cn p 0 3

H P H

s:

ro H
H 0 ft

ti

5 P

<
p rt 3

ro 3 ro Z ro 3

0 3 0 tn C D 3 O 3 O cn C D 3

ro ro < cn
LP

H C D <P C H P H r. rt 3 p rt C D C O 0 3 cn rt P 3 rt Hi H O

P 3

S ro o Hi <. 3 ro 3 Z
tn 0 Hi ft rt 3

Hi P X 0 Hi rt 3 P H C D

0 Hi

3 tp

Hi P H tn ft

C D cn

tn C

TJ 0 tn p rt p 0 3 cn

ro ro 3
P H rt 3 P rt

pi

Hi 0 H

rt 3

tn K

ro ro 3 ro 3 ro

3 C D tP H P

ua
0 rt P O 3 cn 0 0 H P 3 tp

o H

ro n 3

P 3 rl C D H C D tn ft

ro p ro
P cn p 3 tp 3 ft p cn M P 3 C D cn P

ti

3 ro 3
cn p 0 3 0 0 H H C D 0 rt rt 3

c g
0 Hi

rt

< C D .
0 H

ro 3 P P < p ro 3 ro
P
H 1
3 3 cn rt 3 p 3

*< TJ
C D cn C D

P (1 3

p 3

ro o H i 3 P ro CD H _ 3 C O
cn C D O 0 3 Di P TJ TS H P C D tn

n
3 Hi 0 H 0

ft 3

ro ro ro cn ro
rt 3 P cn

P H

rf 0

ro ro ro a

p 3 H

Di C D H

H 3 P tn

rt 3

rt 3 P rt

cn rt P tp P rt 0 H cn

P
3 0

^J ~~J

ft 3 C D

n ro c n n ti 3 ro 3 ro 0 ro 3 cn P ro 3

ro 3

C S

0 1

3 1

ro

ro

p 3

< ro

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

3 a)

b)

Fig. VI-3(a) and 3(b)

Drift force measuring set-up.

This situation applies to a vessel for which it is required to determine the mean horizontal drift forces. The mooring line runs horizontally from the vessel to the mooring point with the vessel in the rest position. The three force transducers shown in the figure are: - a force transducer measuring the longitudinal force component of the mooring line force relative to the body axes (transducer 1); - a force transducer measuring the tension in the mooring line (transducer 2); - a force transducer measuring the horizontal component of the mooring line force relative to an earth-bound system of axes (transducer 3 ) . Since it is required to determine the horizontal drift force the true force will be measured by transducer 3. In order to determine the validity of measuring the force by transducer 1 or transducer 2 we assume that the vessel is car rying out oscillatory motions which induce the angles a and $ be tween the mooring line and the horizontal plane and the longitudi78

nal axis of the vessel (see Figure VI-3(b)). If the force in the mooring line is measured by transducer 2 is F^, then the relation ship between this force and the force F. is: F 3 = F 2 cos a or: (VI-1)

F0 = F . , I
2 3 cos a Assuming that the angle o t is small we may write: cos a = 1 - ^sa2 from which it follows that: F 2 = F3(l + ha2)

(VI-2)

(VI-3)

(VI-4)

From this it is seen that the error in F is in the order of F^.^ot . Normally the angle a will be of the same order as the an gular motions of the vessel (pitch, roll and yaw) which, except for the case of resonant roll motions, will normally have ampli tudes less than about 0.1 rad- Assuming this value of a the error in F will, at most, be of the order of F^ x 0.005 or about one half percent of F_. In the same way it can be shown that the error in F is of about the same magnitude from which it can be con cluded that the forces may be measured in any of the three men tioned positions provided the angular motions of the vessel are not largeVI. 2.2. Model test conditions Model tests were carried out in regular waves for three directions, i.e. head waves, bow quartering waves and beam waves for a range of regular wave frequencies. The test conditions with respect to the wave direction/ wave frequencies and wave ampli tudes are given in Table VI-2 through Table VI-5.

7q

Wave frequency t o in rad./sec. 0. 178 0.267 0-267 0.267 0.357 0.443 0.443 0.443 0.532 0.623 0.713 0.713 0.713 0.804 0.887 Table VI-2

Wave amp lit ude c,

in metres

.\/"f'
0.446 0.669 0.669 0.669 0.895 1. 110 1.110 1. 110 1.333 1.562 1.787 1.787 1.787 2.016 2.223

Wave direction Wave direction Wave direction 90 degrees 135 degrees 180 degrees 2. 14 1.21 2.11 2.86 1.86 1.23 1.98 2.02 2. 19 1.20 2.06 2.11 1.76 2.14 1.21 2. 11 2.86 1.86 1.23 1.98 2.83 2.02 2.19 1.20 2.06 2.76 2. 11 1.76 2.14 1.21 2.11 2.86 1.86 1.23 1.98 2.83 2.02 2. 19 1.20 2.06 2.76 2. 11 1.76

Test conditions for model tests with a tanker. Wave amplitude in metres a Wave direction Wave direction Wave direction 9 0 degrees 135 degrees 180 degrees 0.97 2.15 3.18 0.91 1.01 1.99 2.83 1.10 1.18 1.14 1.04 0.94 0.98 0.80 0.83 0.97 2.15 3.18 0.91 1.01 1.99 2.83 1.10 1.18 1.14 1.04 0.94 0.98 0.80 0.83 2.61 4.49 5.09 1.02 1.96 2.79 0.98 1.02 1.97 1.03 1.04 1.76 0.98 1.61 0.78 0.81 0.64 0.43

Wave frequency
O J in rad./sec.

0.300 0.300 0.300 0.455 0.455 0.455 0.525 0.6 30 0.630 0.630 0.72 7 0.785 0.805 0.910 0.920 0.920 1.047 1.047 1.060 1.168 1.200 1.200 1.278 1.339 1.350 1.398 1.570 Table VI-3

-\/Z'
0.550 0.550 0.550 0.834 0.834 0.834 0.963 1.155 1. 155 1.155 1.333 1.440 1.476 1.668 1.687 1.687 1.920 1.920 1.944 2.142 2.200 2.200 2.344 2.455 2.476 2.564 2.879

Test conditions for model tests with a semi-submersible

Wave frequency a . in rad./sec. 0.300 0.400 0.500 0.600 0.700 0.800 0.900 1.000 1. 100

.\K' '
0.620 0.827 1.033 1.241 1.447 1.654 1.861 2.067 2.274

Wave amplitude ix in metres a Wave direction Wave direction Wave direction 90 degrees 135 degrees 180 degrees 0.81 0.94 0.96 0.98 1.09 1. 16 1.08 0.94 0.79 0.81 1.07 0.96 0.98 1.09 1.16 1.08 0.94 0.79 0.81 1.07 0.96 0.98 1.09 1.16 1.08 0.94 0.79

Table VI-4

Test conditions for model tests with a rectangular barge.

Wave frequency a ) in rad./sec.* 0.760 0.881 1.081 1.260 1.530 1.880 2.167

-\/'f
0.968 1.122 1.377 1.605 1.949 2.395 2.760

Wave amplitude r in metres a Wave direction Wave direction Wave direction 90 degrees 180 degrees 135 degrees 0.55 0.73 1.06 1.08 1.20 0.75 0.50 0.55 0.73 1.06 1.08 1.20 0.75 0.50 0.55 0.94 1.06 1.08 1.20 0.75 0.50

Table VI-5

Test conditions for model tests with a submerged horizontal cylinder.

ft 1

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 Figures VI-4 through VI-8.

FACET SCHEMATISATION

TANKER

TOTAL 3 0 2 FACETS

WATER LINE SCHEMATISATION TOTAL 74 ELEMENTS

*I 1PH 1*

1 1 1

1 1

1 1

1 1

1 . -

Fig. VI-4
BT

Facet and waterline element distribution of a tanker,

FACET SCHEMATISATION SEMI-SUBMERSIBLE TOTAL 216 FACETS

WATER LINE SCHEMATISATION TOTAL 72 ELEMENTS

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 .

FACET SCHEMATISATION

BARGE

TOTAL 138 E L E M E N T S

-e

WATER LINE

SCHEMATISATION

TOTAL 4 8 ELEMENTS
H 1 H

-.Ih

\-

Fig. VI-6

Facet and waterline element distribution of the rectan' gular barge

FACET SCHEMATISATION CYLINDER TOTAL 286 FACETS

Fig, VI-7

Facet distribution of the submerged horizontal cylinder. 85

FACET SCHEMATISATION SEMI - SUBMERSIBLE TOTAL 360 FACETS

WATER LINE SCHEMATISATION TOTAL 144 ELEMENTS

Fig. VI-8

Facet and waterline element distribution of the semisubmersible: 360 facets.

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 Vessel o i in rad./sec.

Wave frequency
/ * * _

.vz
i
i

Vessel t o in rad./sec. 0.300 0.450 0.480 0.500 0.516 0.539 0.560 0.600 0.650 0.700 0.780 0.800 0.900 1.000 1.100

-\/'r
/"

Tanker

0.079 0. 112 0.189 0.266 0.354 0.444 0.523 0.600 0.713 0.758 0.887

0. 198 0.281 0.474 0.667 0.887 1.113 1.312 1.505 1.788 1.901 2.224

Rectangular barge

0.620 0.930 0.990 1.035 1.070 1. 116 1.160 1.240 1.340 1.450 1.610 1.660 1.860 2.070 2.270

Semi submersible

0.098 0. 146 0.200 0.300 0.400 0.500 0.600 0.702 0.800 0.900 1.000 1.050 1.100 1.150 1.200 1.250 1.300 1.400 1.500

0.180 0.268 0.36 7 0.550 0.733 0.916 1.100 1.290 1.466 1.650 1.830 1.925 2.020 2.108 2.200 2.292 2.380 2.570 2.750

Submerged horizontal cylinder

0.421 0.698 0.836 1.080 1.325 1.528 1.872 2.160

0.5 36 0.889 1.064 1.376 1.687 1.946 2.385 , 2.752

Table VI6

Frequencies used for computations.

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.
YJiii__,5Hi_2f_22^Eyt5ti2_}_22__r___^2^2{___

The results of computations and measurements with respect to the first order oscillatory motions of and about the centre of gravity are given in Figures VI-9 through VI-13. The results are presented in the form of non-dimensional frequency response func tions of the amplitudes of the motions to a base of the non-dimen sional wave frequency. The phase angles of the motions are given in degrees, also to a base of the non-dimensional wave frequency. A positive phase angle indicates that a motion reaches its posi tive maximum value before the crest of the undisturbed incoming regular wave passes the centre of gravity of the vessel. The posi tive direction of motions and forces are in accordance with the rectangular system of axes G-X'-X'-X" shown in Figure III-l. Comparison of computed and experimental data on the first order motions shows that the motion amplitudes of all four vessels are generally well predicted by the computations. Significant dif ferences occur in roll and sway motion amplitudes near the natural frequency of the roll motions of the barge and the tanker. These are mainly due to the fact that the computations predict a larger roll motion due to the omission of viscous effects in roll damping in the computations. Due to the sway-roll coupling computed sway motions also differ somewhat from the results of measurements. 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 measurements at these frequencies are, however, exaggerated due to the way angular motions are made non-dimensional.

O._.D 500 SURGE

TANKER COMPUTED MEASURED (ASC ENDING WAVE AMPLITUDE o .500*

500

"

vy ^
;

tl)

2 SEMISUBMERSIBLE

500 SURGE HEAVE

500 PITCH

500

7.

10

// \

*.

zz
Vot

1.0 (it *5a

rf
*

' >v

d11

*3a

0.5

0.5

*v
V^YW,

I*
<S>
4 0

w^ 2

X,
500

BARGE 500 SURGE

e

500* HEAVE

/.
1.0
o
(1)

*
ID

/.
0

_UlL

3a
t f

1 Wo
Fig. VI9

T,

2
7 V3

First order motions in head waves.

SQ

TANKER C OMPUTED o.a . MEASURED (ASC ENDING WAVE AMPLITUDE o _ . ) 500 HEAVE

500 SURGE

500

i l

"1

V
lo

4 11! 3a

. A^
x

tf'

\ &

>n_n . V

2 SEMISUBMERSIBLE

..
500

500 SURGE
1

500

SWAY

fi " ^ 7

VV

NvTL^n

1.0

*2a

TLo n f

2 BARGE 500

500
SWAY

500

SURGE

kw

0 (D

.
^2a
,tii

r P

\
0 1.0

CI

\
0

KJ

-^r.

\ ^

Fig. VI10

First order linear motions in bow quartering waves.

TANKEhf COMPUTED o.a.o MEASURED (ASC ENDING WAVE AMPLITUDE <=.* 1 500

500

500

"O

2 SEMISUBMERSIBLE

500 i n ROLL

1 PITCH

i5O0* YAW

500

J
O 1.0 5a kff

/ .
1.0

\ /

rw
f "* n o

1.0

if1'
4a

05

0.5

.
Mo, 4 O

BARGE 500 500

500

ROLL

^1 V

?'

1.0

4a

__'

y
\ ~
0 Nlo.o.

Fig. VI11

First order angular motions in bow quartering waves 91

TANKER COMPUTED s.4.o MEASURED (ASC ENDING WAVE AMPLITUDE o . . o ) 500* i SWAY 1 1 1500

500*

1 j

2a

I
V
o

500
SWAY

500 HEAVE ROLL

500

"L
;
____.

1.0

_t

ff
0.5

__ *E

\ \

IX:
SWAY

IkfflP.
BARGE

500*
o HEAVE

500* ROLL

i500

O Oj

*.

K<T

10

10
,1)

r >
*3a
x x 4a

X 11 '

*2a

< \
0.5

IfV
! \

s.
4 0
4 %

\/I

^7T 9

Fig. VI12
Q->

First order motions in beam waves

CYLINDER ( (i=180*> COMPUTED MEASURED 500* HEAVE

500

\
10 0 1.0

" F
0.5

*';*

z __v.

\ \ \

- 9 -

500
PITCH

500
HEAVE

1.0
(1)

\ \

!_3_L

0.5

Fig. VI-13

First order vertical motions of the cylinder in head waves, bow quartering waves and beam waves. 93

In terms of absolute differences between the measured roll angle and the roll angle computed for the wave amplitude concerned the differences are comparable to those found at higher frequencies. The phase angles of the first order motions are generally reasonably well predicted. When motion amplitudes are small dif ferences in the phase angles are somewhat increased. In such cases the harmonic analysis technique, by means of which the amplitude 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. The results of computations and experiments on the mean second order forces are given in Figures VI14 through VI17. 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 TANKER MEASURED ( A S C E N D I N G WAVE A M P L I T U D E o. f l ._ SEMISUBMERSIBLE )

BARGE

1 *

n^

f^
t>

fi S I

1 f. 1
COMP UTED ^ 3 6 0 FW E T S

pa

_ fl

V^

Fig. VI14

Mean longitudinal drift forces in head waves

94

o _.o TANKER

COMPUTED MEASURED ( ASCENDING WAVE AMPLITUDE o 4.0 ) SEMI-SUBMERSIBLE

-2

cm
._

>

-01

01 ct

T COMPUTED

- L -

0 . 0 0-^

P
i
I 4

'0

J\

, / y f 3eoFACETS

/TANKER
0

'

SEMI-SUBMERSIBLE

BARGE

J
t

2
M

_
Aj

Ea. " 1 A /COMPUTED L - | 360 FACETS 0 ,

0

c c

- ^

j
/

jo

c
0

>
TANKER

-^

of

a J

SEMI-SUBMERSIBLE

BARGE

2.5

25

IS

1.. Jv

V
-2.5

An s "*ij V

COMPUTED . 360FACETS

>
0 a

A I ^
1

0 0

_.___._

-2.5

-25

Fig. VI-15

Mean longitudinal and transverse drift forces and yawing moment in bow quartering waves. 95

o.a.t. TANKER

COMPUTED MEASURED (ASCENDING WAVE AMPLITUDE 0.4.0 ) SEMI-SUBMERSIBLE BARGE

0 9

5.0

A- f
0

5.0

t
25
/ COMPUTED X - 360 FACETS

2.5

2.5

_
0
j "

>

__A^/

0 t>

Fig. VI-16

Mean transverse drift forces in beam waves.

CYLINDER COMPUTED MEASURED 11=180

li>135

U = 90

r> 2

J
IS > 0

M
\
4 (D
1

A-

V 0

^ 0

;
H=135*

'

< )

"

,>
li=90"

<

\i
V Fig. VI-17

rV

11 = 1 6 0 '

\ /

Mean vertical drift force and trim moment on the cyl inder in head waves, bow quartering waves and beam waves.

96

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 in Figure VI-16 shows a large peak in the computed results which is not present in the experimental results. The mean yaw moment given in Figure VI-15 shows that at the natural roll frequency computed and experimental results are opposite in sign. The com puted results show a marked negative peak which again is not pre sent in the experimental results. For the barge similar abrupt changes in the computed drift forces and yawing moment are found near the natural roll frequency. It is felt that, as is the case with the overprediction by the computations of the first order roll motions, the large discrepancies in the mean drift forces may be also due to the fact that viscous effects in the roll damping are not included. In order to check, qualitatively, the influence of the roll damping on the results of the mean drift forces, computations were repeated for different values of the linear roll damping at the natural roll frequency of the tanker. Computations were carried out for roll damping values of 1, 3 and 5 times the potential roll damping. The results of computations are shown in Table VI-7 in which the mean values of the measured data, given at this frequen cy in the figures, are also given. From the results given in this table it may be concluded that increasing the roll damping tends to effect both first and second order quantities in such a way that better agreement with experimental data is obtained. The re 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 are less linear than has been assumed here.

97

FREQUENCY:

w = 0.444 rad./sec. ,1/3 = 1.113

Calculations Wave direction in deg. Roll Unit dampin g m u l t i p l i c a t i o n factor Measurements

Description

V--

uC

V--

u.

V--

ur.

MOTIONS Sway Heave Roll Surge Sway (leave Roll Pitch Yaw 90 90 90 135 135 135 135 135 135

*._xu

^zx *axz *!ZX *.xx 1 1

-.i ". ' -UXZ *._XZ

0.35 1.38 62.50 0.25 0 . 19 0.56 7 . 12 0.48 0.25

140 35 3 141 91 290 9 85 253 348

0.63 1.38 1 5 . 19 0.25 0.25 0.56 1.73 0.48 0.26

255 353 172 91 268 9 117 253 356

0.70 1 . 38 8.46 0.25 0.26 0.56 0.97 0.48 0.26

263 353 176 91 268 9 121 253 357

1 . 12 1.35 9 . 36 0.28 0.41 0.68 5.18 0.71 0.26

267 2 248 112 273 33 253 271 8

*axz
F2AP9V
1 / 3

MEAN DRIFT FOBCES AND MOMENT Transverse force Longitudinal force Transverse force Yaw m o m e n t 90 135 135 135 ci
1 ) 2

33.628 -0.274 -0.152 -0.783

9.650 -0.203 0.233 0.405

5.431 - 0 . 193 0.221 0.505

1.210 -0.203 0.443 1. 143

v^'V 2
F
2

A^

1 / 3

1 ) 2

v^ 2 7 3 ^ 2

Table VI-7 Influence of linear roll damping on the motions and mean drift forces and moment on a tanker at the natural roll frequency. 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 i n 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 semisubmersible contribution IV is practically zero except for the very low frequencies near heave and pitch resonance. For this vessel
I nr.

TANKER

BARGE

-2.5

25

\f
0

\
N.

25

Fig- VI-18

Components of the computed mean longitudinal drift forces in head waves.

CYLINDER

n m
li = 90*

TOTAL H=180

25

A
i

\
0

/'^>
0

\ s

y ^ '

25

-2.5

Fig. VI-19

Components of the computed mean vertical drift forces on the cylinder in head waves and beam waves.
lni

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 contribution arises from relative wave elevations around a waterline . 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 waves. The results shown in Figure VI-19 show that in this case the relative importance of the contributions to the total force can vary quite significantly. In head waves contribution II is dominant and contributions III and IV have only minor effect on 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 the mean horizontal force due to contribution II was directed op posite to the total force. VIi5i_Conclusions From the results presented in this section it can be con cluded that the mean wave forces on bodies of arbitrary shape can be computed with reasonable accuracy based on the method presented in this study. It was generally recognized that the mean forces acting on full forms such as barges and ships could be computed with the aid of potential theory and the supposition that the forces are basically a second order phenomenon. The correlation shown here between the results of computations and measurements of the mean wave forces acting on a semi-submersible and a submerged cylinder indicates that the same theory can also be used to pre dict these forces on more slender forms. 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 drift force in head waves indicate that, contrary to expectation, for equal displaced volumes the mean force acting on the semi-

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 semisubmersible 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___lt2<__-_-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 VII.2.1. General In this section the specific requirements, which must be met in order to be able to measure the low frequency second order wave forces on a vessel, will be discussed. From the expressions derived for the second order forces in chapter III it can be concluded that, in order to arrive at the correct value of the forces, the model restraint must be so that 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

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

' "~v

100%

l \\
\

II
\ \ \

UJ C O

z O
Q.

\ \ \

LU Q _

(/)

r. o/

\I

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.

WAVES

SERVO SYSTEM

VESSEL

POSITION

CONTROLLER

Fig. VII-2

Block diagram of dynamic system of restraint.

This diagram represents a possible control system for a vessel moored in waves. The waves exert forces on the vessel which in turn cause motions which contain wave frequencies and low frequen cies due to the drift forces. The motions are measured and fed to the control unit which filters out the high frequency part of the motion signal and gives only low frequency force commands to the servo system which exerts the required force on the vessel. In some 107

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
H> WAVE FORCE SPRING VESSEL Fm(t) MEASURED FORCE

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 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-A )
where: X
U

2 2

vV

= U)/(D

t u = frequency of force excitation = natural frequency of the horizontal motion of the moored vessel = /c/ni = stiffness of the mooring system = virtual mass of the vessel = non-dimensional damping factor = b//cm = damping coefficient.

c m v b

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

X=

L O

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 proportionaldifferential 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
1 1?

by the application of more sophisticated control systems than used for this investigation. From the point of view of practicality of results obtainable by the present system it can be stated that, since the natural frequencies of the horizontal motions of a moored tanker are generally lower than 0.05 rad./sec., it is possible to measure accurately those low frequency horizontal wave forces which are important for the low frequency horizontal behaviour of such vessels. In the aforegoing the feed-back control characteristics of 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. As stated previously the model tests with the semi-submers 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 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 and aft mooring line, each incorporating a force transducer and a linear spring. The restoring force of the mooring system in the longitudinal or surge direction amounted to 513 tf per metre dis placement for full scale. From a surge motion decay test in still water the natural frequency of the surge motion as induced by the system of restraint (mooring system) amounted to 0.4 rad./sec. full scale.

i i i

O
l~^m^=r-

O O

O O

WAVES 180 SPRING

c^pJWW-! FORCE TRANSDUCER

O
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 SURGE HEAVE PITCH

..ti)

la

I
\

1.0

I
1 (ii in r a d sec-

10

kt 0.5

0.5

Viz\
Fig. VII-8

Influence of mooring system on the amplitudes of the motions of the semi-submersible in regular head waves

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.

N STIFF MOOR ING SYSTEM

-20

I
\

fi

( N J

cg
w c q
i^_fi

-10

A,
Fig. VII-9 waves.

(A) i n rad.sec.-1 longitudinal

Influence of mooring system on the mean

drift force on the semi-submersible in regular head

115

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 VII.3.1. Generation of waves 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 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.
3 3 0 0 m (FULL SCALE).

MOTOR
CD-,

Fig. VII-10 116

Wave and Current Basin

_*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 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 ure VII10. By driving section A regular waves of frequency 1 0 and B in Fig

at a constant frequency to produce regular

and section B to produce

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

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.

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 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.

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

-:1 O ) ^ 0 . 5 6 0 r a d . s e e-1 r' U)2=0.535rad.sec

_1 CU = 0 . 3 5 7 rad.sec -1 00.pO.382 rad.sec.-' 2

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

-1 0^ = 0 . 6 7 4 ^ . s e e -1 r ' CU2= 0.624 rad.sec.

_1 OJ^ = 0 . 4 9 6 r a d . s e c-1 r 1 0J2= 0 . 4 4 6 r a d . s e c :

50

100 sec R e g u l a r wave groups

Fig. VII-11

Cu in rad.sec."

Fig. VII-12

Spectra of irregular waves for tests with the tanker.

* _ ^ w l / 3 = 3.31m

.
u
0) lA

=8.1 sec.

cy.

y
Fig. VII-13 submersible.'

0.5

U) i n rad.sec.-1

1 . 0

Spectrum of irregular waves for test with the semi-

.
<P
rt

J _l 0J P1 P

rt

&
Hi H O P

0 Hi

OJ

rt tr OJ ri

rt H 0> H

n n ro
H

Hi O H i H H Hi 0 H

fl)

rt tr

0>

3 ft

fl) 3 p

Hi O H

n ro
P
PJ

z 0>
ti

rt tr p

O H O W cn i

rt

Q 0 3 tn p

ro cn ^<
C O U)

3 Oi P> tn fD

a ro
tn OJ H rt P 1 1

3
0 p TJ rt P Hi rt P 3

H s er DJ ro fD <

n o
ro
0 H KJ 0i 3

ro tn
ft tr
H ro

<
3

<: C D
Pi rt 3" 3 P P cn rt O Hi Hi 0 ii

P t. P P

Hi ii C D

rt fD Ul rf

< tf
H rt 0> 3 0J P1 rt 0

cr p
H

H fD Xi

c p

3 P 3 p g

cr ro ro er ro ro
U3 rt rt tn

ro c
^d 0 H

Hi 0 H O fD cn ft 3*

3
fl>
Z 0J <

rt 0

a ro a ^. c
0J ft tp p H H Hi H fD

n
fi Pi fD

ro tn a er ro

o
tn ft p H H rt

tp H O C 13 cn

C O

0J p. H P fD Pi O P ft

Hi H C D

OJ 3 OJ H K tn

rt fD cn rt

0 Hi

O 0> 3 3 O ti

fl> tn
0 Hi OJ 3 OJ H M rt K rt H
h>

p.

r o
tr ro

3 p

ro 3 < N ro
ti

fl) ro Z H n p fD 3 ro p H O ro a 3 ti cn H ro B P ro

tf c ro 3 n <; n
cn rt fD 0J O

p tn

p, P ii

ft rr ft)

P C D 3 O P (D

er fD 3 p
3 P rt

rt

5 o* ro cn ro
H 0 3 ip to

p i cn rt tr

o > *< E M ro n
pi

c n 3 c 3 s: cr fD ro & ft
P 0 3 13 O 3 Pi Z p rt OJ H H P 3 ip 0 rt p O 3

i tn TJ fD O rt H 0> P p* 0 3 ip

0 H H fD w 13 0 3

fl)

cn rt

P H

3
H

Cb C H OJ rt P 0 3 M

0) rt

OJ

rt P 0 3

rt fD 3

3 O rf

O Hi

& Oi 3 3 ro \K n ro o c ro tn
tn rt tn

P O 3 rt 01 p 3

a p er ro 3 ro 3
p i Pi P H O 0 H H

tf ti Z P p ro ro rt w 3 er rf o er ro ro
rt

H 0> rt P O 3 ft H 0) 3 tn P fD 3 rt

ro cn

ro
P P1

ro 3 ro ro tp
ii OJ H ii

er ro cn

ro

0 i O O

o >
Z 0)

ro
PJ

ii C D tn C

*<
3 3 0> 3 Pi

a p

3 fl>

H rt

Ui p

z p
rt 3
CTi

3 fD H fii ft O *i tn

cr ro

cr ro
Q* P O Hi

O 0 ii H,

0 3

rt cn

tf tf ro
3 0 p 0 ft

tr g ro
H i rt Oi p 3 Hi P H to rt

P 3 rt fD H fD cn rf

r o cn
Hi H 0 3 H QJ 3 ip

TJ H fD tn fD 3 ft

Ti

p to 3 fD PJ 3 tn

rt O

rt tr 0 pi M

PJ

cr *<
tr
P H
cn p 0J ft

rt i OJ M

ro

Hi H fD

Pi C H P ft P O 3

13 H O O

p H

P. P H OJ rt P O 3

ro ro
Pi Hi 0J O Hi H Z P ti
rt

ro cr ro
tr 0J

fD tn

<

ro
PJ

ro
p ii OJ 3 P.

er fD
O Hi N 0 Hi 3 c

13 0 3

O 3

cr tr vy cr 1
P fD Pi P tn rt TJ O 3 tn rt cn

PI

tn rt 0i rt p o

rt

cn p ip 3 P M cn

PJ

OJ 0

cn P Hi Hi P P P

N fD ii O

fl> 3
0

P 3 ip

cn rt p p* H

o c ro p,
P
rt tr
H

rt 3" fD O H Pi

O O 3

tr fD n ro er tn ro
P
3 rK H

0 Ui O P H P* 0> rt p O 3 cn 0i 3 OJ H K tn p tn

c n n r t ro er 0 er a r t & H ro rt K 3 er rt O ro ro ro cn H ro rt cn a ro
z p
rt 3 OJ

< p
0 p H tr

tn C p rt P

< cr ro
PJ

Z
OJ H P

OJ tn 0J tn

ro
PJ

n ro tn
rt tn P 3 P H

p 3 0 m

p P1 M

c ro
0 Hi

rt

H 03 0

rt O Hi 0 Hi co o N 3 cn

rt 0

*< ro
M 0 M O p 3

0 O 3 0 fD Pi 3 fD P

C p M H

P* O

H P

p cn

& fD ro er
ti

fD

cr
P
P3 tr fD

n 0 3
0 H O tn cn l

tr 0 c

& ro ro B ii H ro
Hi d H

fl) ro
p 3 00 H

H p 3 IP

Pi C H OJ ft P 0 3

OJ rt O

i3 tr fD

tr tr p
z
cn 13 P H ii P rt
PJ

0 Hi o

!*
to p 0J P"

tr fD
P
'p rt O

Pi

K fD

TJ

P
PJ

0 3

rt

Hi H fD

tf p ro ro
OJ

o ro ^ ro
<P

Pi

(t

0 p rt i cn "P Hi H P ^P P

P1 OJ 0 ?r p 3 <P g 0 rt p O 3 cn

. ro o rt ro
O li H OJ H fD H ip p pi P H 0J

tf

tf

ft tr OJ rt

p 3 Hi p 3 P rt fD

P P H P

H fD P H Ul fD 0 0 3

er ro . t_n
p p Oi H rt

er ro ro ^ c
P
H

Hi O P1 P" O

er ro

O Hi

z tn
..
OJ H Hi O H O

rt

rt fD cn rt ti 0) Pi o 3 O

3 O Z H fO

ro cn
rt tn

3 ro
fl) cn
OJ tn P Hi

p" N) O O O d H P M

H O tfl P M rt tn

IP fD

&

P fD 3 O

*<
Z P < ro
Hi H 0) rt
ti

fD cn

0J ft

p OJ rt

ro o 0J ^< ro a er 0 c
H to 3 ?r fD H p 3

D 0) H N fD pJ P 0 (D rt 3" O

O H.

0 Hi

C O Ui rt P O 3 0J

tn (D

e 3
tn

H rt M

W P 3 O (D

P 3

0 cn Q p H M 0J rt p 0 3 cn

Z z

3 fD
OJ Hi

p 3 PJ 3
PJ

ro 3 fD
H

ii

P P H. OJ rt p 0 3

P< Ci o o

<
H 0J Pi

ft P" fD

cr o P O ro 3 < ro o
0 P H fD rt 3"

ft

N P H 0

0J C O

cr rt ro
Hi H Hi C H P1 .p P fD 3 O P
ti

tr p

\cn o tn ro 3 ro er O a cn
Hi 0 H O

O cn 0 p p> pj p rt p O 3 cn

tn

Hi H

tn P p 3^

ro

D < ro tn < ro cn ro M
H 3 p cn 3 H 0J cn ti

0 tn p p H H P rt p 0 3 tn pi p*

< ro c
OJ rt rt

3 rt tn 0* tn P H fD

3 3 ro 3
rt cn

ro ip c H
0J

H H 1 1

P H, O tn M i

fl)

Hi 0 H

cr

p i tn p P
M

Hi 0> M

O H Pi C D H

0) p 3 fD

h p

rt 3" fD H

ip

c n P ro
C P1 M

13

c n

0) Ui

tf

ip H O P 13 tn

0 rt P O 3 Ul

s:
OJ

P P fD 3 P

OJ cn fD in

ro
PJ

0 Ul 0 H M P1 OJ rt p 0 3 Ul
J 3 \cn P cn C O Pi P ro Hi O n Hi >3 H . cr & P Hi ro ro n p

tf ro fD O c n i rt

P 3 Hi H P (D 3 O C D OJ H fD 3 rt

cr *<

C P H fD 1

^ >

>

rt 3" O

Hi O >i O fD tn

p Hi

er ro ro ro in tf cn c p fD ro
3 1 ip l

ro
0 O H 1

Ui 0 0J H fD

P H H fD iQ P M OJ H

er ro
P

rt

ro c n i

0J

<: ro

cr ro

* ro
H

Z 01 tn

< ro ro

cn

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 -100tf

^/WVVvw^ x v A\./\r

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

TANKER FORCE

WAVE

^w1/

=10.3m

f = 13.3sec.

100sec.

-t

SEMI-SUBMERSIBLE FORCE

WAVE

w 1

^ = 3.3m

T =8.1 sec.

100sec.

-t

Fig. VII-15

Longitudinal drift forces in irregular head waves. 123

The transfer functions are given in matrix form of which the two axes represent the two frequency components of a regular wave group, 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. 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 - c o - 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 spectively. In order to be able to compare results of computations 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./ sec. no problem exists since at this frequency computed data are also available. These are the computed data on the diagonal of the matrix of the quadratic transfer functions.

"2

^ \

0.354 2.0

0.444 8.7 7.0

0.523 10.4 20.8 12.4

0.600 24.5 19.4 16.4 14.0

0.713 10.0 25.7 8.3 9.5 8.6

0.803 38.4 12.1 14.3 18.0 4.3 9.2

0.887 37.5 35.2 14.2 14.9 6.7 4.7 8.6

0.354 0.444 0.523 0.600 0.713 0.80 3 0.887

2 T in tf/m Frequencies in rad./ sec.

Table VII-1

Quadratic transfer fiinction of longitudinal force on the tanker in head waves.

"2

^ \ 0.5 0.6 0.7 0.8

0.5 0

0.6 8 1

0.7 11 6 0

0.8 7 7 7 11

0.9 8 5 15 23 25

1.0 20 14 3 20 21 20

1.1 15 19 18 13 9 21 24

0.9 1.0 1.1

T 1 2 in tf/m 2 Frequencies in rad /sec.

Table VII-2

Quadratic transfer function of longitudinal force on the semi-submersible in head waves.

from cies cies onal

For the semi-submersible the experimental data obtained cross-bi-spectral analysis are valid for difference frequen of 0 rad./sec. and 0.1 rad./sec. The data for these frequen are found on the diagonal and the first row next to the diag of the computed data given in Table VII-2.

YII__i_Q^E__Ei2_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. The experimental data for the tanker obtained by cross-bispectral analysis of data from tests in irregular waves show a rather irregular character. This is ascribed to the parameter set tings used during the cross-bi-spectral analysis.

COMPUTED
A

REGULAR WAVES IRREGULAR WAVES: w 1 / 3 = 2 . 7 m ; T= 8.2sec. = 4.8m; f = 10.1 sec. = 7.3m ; T= 12.0sec. = 10.3m ; f = 13.3sec.

30.0

. " J

15.0

U)1 0>2

in rad.sec1

Fig. VII16(a)

Quadratic transfer function of the low frequency longitudinal drift force on the tanker.

126

COMPUTED REGULAR WAVE

GROUPS = 2.7m = 4.8 m = 7.3m --10.3m t= 8.2 sec. f = 10.1 sec. f =12.0sec f =13.3 sec.

IRREGULAR WAVES: w 1 /

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

OJ

c
C\J

15.0

f}^\

0.5 L _ _ in rad.sec.-1

1 . 0

Fig. VII-16(b)

Quadratic transfer function of the low frequency longitudinal drift force on the tanker. 127

COMPUTED REGULAR WAVE GROUPS IRREGULAR WAVES: twy3 = 2 . 7 m ; f = 8 . 2 sec

= 4.8 m ; f =10.1 sec. = 7.3 m ; T =12.0 sec. = 10.3 m ; f =13.3 sec. 30.0

Oh 0J2 ~ 2 Fig. VII16(c)

i n rad.sec.'

Quadratic transfer function of the low frequency longitudinal drift force on the tanker.

128

-40

(a)

0.5 (A) in rad.sec.

-1

1.0

-40

-30

U)-, - U ) 2 = 0.1 rad.secr1

COMPUTED (TOTAL) FROM C.B.S. (TOTAL) (2) COMPUTED WITHOUT $

OJ

E ^ c
oi

-20

(b)

-10

0.5
U)-i + tx)p

1.0 *L_ m rad.secr1

i

Fig, VII-17(a) and 17(b)

2 Quadratic transfer function of the low frequency longitudinal drift force on the semi-submersible. 129

LO = .!

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 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 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 irregular sea state is about ten times larger than in the lowest sea state. Bearing this in mind it can be concluded that the re sults obtained from the various tests in irregular waves on the quadratic transfer function agree reasonably well. The experimental data obtained for the semisubmersible are smoother than those obtained for the tanker. This is in part due to the increased test duration and in part due to the different parameters used during crossbispectral analysis of these results (see Appendix B). In general the experimentally obtained data from tests 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 due to the method used to generate the regular wave groups in the basin. The wave groups were only realized in a relatively narrow field as indicated in Figure VII10 while the computations assume that these will be longcrested. As a result of this it is possible that the corresponding second order forces were not fully developed during the experiments.

_/!

The computed data for the tanker for difference frequencies of 0.025 rad./sec. and 0.05 rad./sec. given in Figures VII-16(b) and 16(c) show a sharp increase for lower values of the mean wave frequency. No experimental data are available at these frequencies to confirm this trend. Examination of the contributions due to the 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 by the non-linear second order potential. Examination of the computed data for the semi-submersible reveals that also here some influence of the non-linear second or der potential is found. In this case for a difference frequency of 0.1 rad./sec. the experimental data appear to confirm the existence of the contribution due to the non-linear second order potential. In Table VII-3 the computed quadratic transfer function for the semi-submersible is given without the influence of component V. The contribution due to component V is given in Table VII-4.

^ \ 0.5 0.6 0.7 0.8 0.9 1.0 1.1

0.5 0

0.6 1 1

0.7 1 1 0

0.8 5 3 3 11

0.9 6 6 10 21 25

1,0 3 5 3 18 21 20

1. 1 7 12 8 8 11 20 24

T 1 2 in tf/m 2

Frequencies in rad /sec.

Table VII-3

Quadratic transfer function of longitudinal force on the semi-submersible in head waves without contribu tion due to second order potential.

1 \ ^

*2 0.5 0.6 0.7 0.8 0.9 1.0

0.5 0

0.6 7 0

0.7 12 6 0

0.8 11 10 6 0

0.9 3 6 9 6 0

1.0 19 12 0 8 7 0

1. 1 12 15 15 6 6 7 0

2 in tf/m

1. 1

Frequencies in rad /sec.

Table VII-4

Contribution of second order potential to the qua dratic transfer function for the longitudinal force on the semi-submersible in head waves.

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 quency force at difference frequencies greater than zero are domi nated by component V. At higher wave frequencies components I through IV tend to dominate the results. These components are due to products of first order quantities which become large when ship motions increase and/or first order diffraction effects increase. When components I through IV dominate the results the first of these components generally is the largest of these, as was already indicated in chapter VI. VII^7^_AEr^ximation_for_the_low_freg In this chapter some results have been given on the low fre quency second order longitudinal force in head waves. The results of computations will be used in a discussion concerning a method for approximating the low frequency components of the second order 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 :
T

12

= T

^ 1 ,to 2 ) a T(

ux + ux 1 ,2 ,

to. + ux *) 2

(VII-2)

We may check the assumption for the case of the tanker and the semi-submersible by inspection of Figures VII-16(a), (b) and (c) and Figures VII-17(a) and (b). Equation (VII-2) implies that in Figures VII-16(a), (b) and (c) the computed curves of T.- given for the difference frequencies of 0.025 rad./sec. and 0.05 rad./sec. should be the same as the computed curve given for zero difference frequency. For mean fre quencies below 0.4 rad./sec. this is clearly not the case. At higher mean frequencies the discrepancies are less but nevertheless increase for the larger value of the difference frequency. The dif ferences below a mean frequency of about 0.4 rad./sec. are mainly caused by the force contribution V due to the second order non linear potential. This contribution is zero in regular waves. In Figures VII-17(a) and (b) in which results are given for the semi-submersible it is seen that for mean frequencies below about 0.8 rad./sec. the correct value of T
2

for a difference fre

quency of 0.1 rad./sec. would be underestimated if it were replaced by the value found for zero difference frequency. At mean frequen cies higher than 0.8 rad./sec. the differences are small. In Fig ure VII-17(b) computed results are also shown for the case that the contribution V is neglected. Comparison with the total force including this contribution shows that for mean frequencies higher than about 0.8 rad./sec. the influence of this contribution is small. At these mean frequencies the low frequency force is domi-

rt

3 0 ii p Pi 3

H*

pj H

tn P p 3 O J ft
Hi

ro & ro
rt
pi Hi

p 3 3 O H H ri p 3
PJ

P 3

tn
ti ti

P >

3 pi P K

Cr 0 < t ro 31 r H N g o ro ro 0*1
3 O Z 0) H
PJ

ro tf
P

H fD

c n cr ro ro o rt
tn C D P 0 p rt Hi 0 H O C D M

tr ro & er pc o
H C D P

O 3

rt

rt p O 3

er ro
P

c n pC D 3

tf o
ro Hi

tf
fl)
3 O Phj H Q rt

p
>3

p 3 P P P P<

n p cr ro tf n rt er
PJ
rt 0 3 P1 3

p. O 3 C O TJ 0 Hi
pj

tf ro er pi
Hi O Hi
ti

P Ptn

a ro
3 tn

h P

0) 3

*t 0

Hi p 3 P rt p 0 3

.
H
to H
ti

< ro ro rt 3 er ro
O rt rt rt O H rt

Hi 0 H P

Pi

0 M

ro c n fl) n 3
rt 3J

tr ro <

i . >3 ro tn p " er
P
TJ P> H rt 3 0 ft O P C D

p 3

0i H

Q ro tf ro

Pi P 3 Pi 3 rt

a ro tn c n cr ^< . ro o NJ ro tf 0 ro p ( J l rt ro Pti 3
P
rt 13 p-

rt O J rt p0 3

< Q ro 3 p-

3 O rt PO 3

tf

P< P H ro

er ro

rt
3 Pi ti

NJ

p 3

ro p) er 3 ro ro
Z DJ tn O 0 3 P.

Hi 0 H. P P ft PO 3 tn

h>

PJ ro ro Z cr Ii

*4 O H

ro

rt tn tr cr O ro 3 ro p ro H C D ii tr 3 H ro ro H 0J O pi ro ro Ul 3 tf P tr tn C D r t p H P 1 ro c n P- P J OJ Qi ro 0 0 3 rt H Pp3 to

fD P

ro ro
tn
PJ

*i H 0 3

P H rt tn

P_ 0J

p 3 rt 3 H Hi O H H ft O i TJ pj 3

rt

er p .
H

. ^
> < * ro

Hi 0 H

cr J ro <i ro
3 O O H

o Z < ro o > ro cr
T3 H C D Pi P O rt

ft 3* P Vi

Pi tn

Qj

3 n
in ti Pi

ro

Hi H C D

rt 0

Hi

ro tf
0 O 3

z pi
aro
tn P O

er 0 n

tf

ti

er ro fD < n 1 er ro
P PC D tn H p pi

er ro ro cr ro ro tn n
0
Hj

ro 3 n p-

ro *< c n rt 0 3 er *-. er a ro r t ro 0 *< er

tr

Ul rt P Pi

pi

O J

3 13 H C D

n K a
(t OJ tn Pi p0 3 3 pj K O H Pi

tn

^ r t ro i i *1 c n cr tr ro ro ro ro < tf H P tf H P ro < H i H ro tn er o
tn C Hi O H P fD P3 H C D

<

O J H

0 H

1 Ul

cr p ro
H Hi H

rt O

< ro c n .
tr ro
Hi O H P

p N 0 3 rt 0) H pi tn in p 3 TJ rt p 0 3

ro o

rt PO 3 tn 0 H rt fD H H rf p3

CC

Hi 0 H

0 3 Pi

p) H

P H rt

n < ro c n 0
P PJ O 0 3 tn

tn

* h Hi a
Hi H C D

Oi

p tt

n ro ro tn o p 3 Q : c H
tn P< Hi PJ O fD H P Pi C D Pi

ro

ro

tf P

C D

H O

Hi O H O

<! p

0 H Pi

n ro c n

pi ptn p P tn

s O J
c n ptf
0 3 P M fu Prt PH H C D

tr < ro c n
Hi

tn

< ro tn
H O

Hi 0 H O 0 PJ 3 3 O rt

ro to

ro ro

O J ft

ro

H O p

n & 3 tf
P C D 3 P

ro

rt 3" C D

ro ro Ti ro <! 0 rt ro tn fD ro P1 f t c n er
P rt
Hi

P p< rt to

. G c n p ro er "j
3^ 0 Z C D

Hi 0 p 3 Pi

0 3

ii

tr

cr ro ro x

ft

tf er
H

rt tn p Cb fD

i 0

p 3

3 0 rt

cr ro
*<:
tn

er

P >

p tp 3"

Pi P 0J

OJ rt

M H 1 to P1

ft

tf
< C D

s: tf ro r t

c n &

i>

3 ro

3 O 3 1 H P 3

tf c1
P 01 H

pH tn rt ptn 0 O H

<:

z tr
o
tn
PJ

er H i ro o H. c n ro p"
C O
x

pj Hi O H 0 ii tt

c n c n rt ro tr p. ro H er ro ro H i H O ro H .. ro

ro er H
pPi

pO

Ii pj

s: P < ro s: cn 0) <: C D 13

c n
pHi rt

o Q 3 13 c rt
rt 3" P> rt P 3 p3 ti TJ H PJ rt p-

PJ

z
c n ft ro
Hi O ii 3 Oi 3 3 0) Pi

a ro ro ro ro H i pi 3 fD < H H H p H H tf ro < ro 13 H i < pi 0 ro tf 0 rt H P< P cr cr ro CD ro

tn p3 ppP H p 0 3 TJ O 3 C D 3 rt

3 p P1 p pi 0 p> 3

01 3

ro
P>

Hi

H
fll

*< TJ cr ro er a
P
H 13 pi

ro H 0 n tr cr tf ro p 0 3 P ro in 0J H tr 3 ^ 1 P PJ PJ ro o ro pi p p H pi rt P 0 p o > ro tn Cb tf 3 cr H p H i 0 Dt ro ro ro p P< 3 PJ < <! H 0) tf fD ro pi rt 3 ti P PJ Ul 0 ot er tf -< tf

O O 3 P M rt PJ X p3

13 H

0) rt pi 3 P (D

PJ

0) H C D

< *
P

o tf ^
rt
PJ

3 n o
01
pi

OJ rt p O 3

<! ro

PJ

3 P P C D

Di

Hi H

c n

3 rt P pi P>

fi O <J pPi C D rt

p-

rt p

cr ro P m p3 Pi rt P-

a Qi
H P>

tt tf pj ft

pi

pi H 0> rt p.

H O P 13 C O

Hi 0 H H P H

ro a p o
P1 P1

tr er ro

t3

3 ro ro r ro tf t X c rf tr fD

0 H P N 0 3 rt Pi H

P C O

c n ro O tf H.

\ . ti c n er H i ro n H i tf P fl) ro P- c n O Hi H H H n ro M H ro ro c n Z ft pi M rt 0 tf P er ^< ti i~3 P er < f t z 0 ro H ro ro cr ro ro 3 M n ro H i H CD C er O P H Pi c n ro rt ft P13 pi ro ro ro 33 0 ro tf H cr CD M 3 tn tn P> O PJ c ro r t .3 pp i ro p* rt 3 P> 3 P 3 n Hi H H P 0 P PJ p LJ. & 3 H i Qi i i _ _ ro H C D Pi O ft O ro 3 tn h O P P 0) Hi 3 Hi cr r t TJ o ft TJ H P O O o r t l i H i HP H O J o r ro rt t O H * cr f t tf ft fc P3 ro c 3 O o H O ro p3 H i t i O 0 . ro pi Hi Hi PJ H ro n pi H P* a ft H H ft C D Q J H er ro ro ro tn 0) O \ ro r tr tf H tn t PJ rt c n ro P tf Ui H ro ro o C O Hi pj \tn o z c n PJ < er fl) H r t O 0 ro ro i > rt 0 0 Z o ^ tn P er p i 3 0 H O J . ro rt ro
P3 rt O 0 3 rt Pi

rt 3" p

5 C D

er ro
P rt H p 3 O rt

0) M P1

c n

pi 0

C D 3

0 pi

o ro tr
0 Hi O H

pj

3 O K

<

^<

3 p

cr ro
rt to OJ pi

O H
PJ

tn

er er ro 0 p.
P

0) rt

TJ H

ti

P ft PO 3

^< 3
rt

ro er pj ro p-

ro Hi
Hi

P fi

< o p ti rt rt ro z H er ro p rt < in PJ tn 3 tr ro H ro o > tn

pi

ro

PJ

Hi

ro ro c n

ft

p P1 P1

z 0J

P" P 3 Pi

OJ

tn

3 OJ i

ro

rt

z
CO

ro

H fD tn P1

rt 0

tr fD

ro

er ro

3 Pi 3 rt

P 0 3 ft H P

0 H PN 0 3 rt 0) M

o CO < ro o C O n ro n n c n ro
tn

3 fD

3 O rt

c i
ro

tr
p

c n
0
pi

ro
pP" rt H P
Pi

3
3 pl

fD H

C O pi 3

er ro
to C D X O C D 3 X M

0 ti

c n rt ro ro t er ro > < : cr r 0 fl) i i O ro ppp

P
1

pN 0 3 rt P> P1 Pi pi

ro
Pi

H 0

z
Q :
i

ro

Hi H

ro ro o c n rt c M c n r t
3 ro
i

p,

O J

P rt tn

P 3

ro

O Hi

frequency components of the force in irregular waves.

2 2 4 6 8

2k h

0.00

0.08 0.00

0.09 0.06 0.00

0.08 0.07 0.05 0.00

12 TfpgL

2k

2h Amplitude of low frequency second order transverse force in regular wave groups on a submerged cylinder in beam waves.

Table VII-5

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

a method for approximating the low frequency forces in irregular 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. In the case of a submerged horizontal cylinder it was found that in regular beam waves the mean horizontal force is equal to zero. The low frequency force in regular wave groups was, however, found to be non-zero, thus indicating that in some cases the mean force in regular waves cannot be used to predict the low frequency force in irregular waves.

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 nique which employs ship mounted propulsion units to counteract environmental forces due to wind, waves and current acting on the vessel, thereby maintaining as closely as possible some desired position in the horizontal plane. The last decade has seen a steady increase in the number of vessels which are stationed at sea by means of dynamic positioning systems. Up to now most dynamic positioning systems were used for positioning drilling ships in deep water where conventional anchor ing systems were considered to be too cumbersome. Nowadays dynamic positioning is also being used for diving support vessels and maintainance and survey vessels. This increasing interest in dynamic positioning systems stems from the need for a means of maintaining the vessel's positioning which is quick, accurate and versatile and does not interfere with systems, such as pipelines, lying on the sea floor. The propulsion units used for dynamic positioning can be either fixed, tunnel mounted controllable pitch propellers, azi muthing right-angle drive units with propellers externally fixed to the vessel, or vertical axis propellers. The magnitude and di rection of the thrust produced by such units are governed by a control system which has as input the position error of the vessel relative to the required position and heading in the horizontal plane (feed-back control). Also, in most cases, the instantaneous wind speed and direction measured from the vessel are used to
1 37

OJ OJ

13
H O
Hi

cn

o
C D 0) ft p

PJ

ro 3
Z
3
rt rt pp-

Z
0
Hi Hi Hi
Hi

rt Tt 0 tn

rt ro
Hi

ti

Z
pp-

3
0J p-

tf er ro
Pi p rt P

P H H

OJ

3 3
pi P1 pDJ
PJ

cr ro ro tf
O 3 3

pi

H ro o
H H

pi

fli

T5 P1 C D

ft P-

a o

ro n o
H
rt

0 H 0 3
Hi

Ul OJ rt

fli P1

< fl) ro tf
rt

tn
1 H O

H H

*4 O H
H P TJ

ft P H 3

pj

z
> Ul
H
ti OJ 0J rt <! H rK

ro
p 0 3

ft

Hi ii

P" P> H

C O rt QJ

i3

<
fD ii H

Pi

er ro < ro
3 rt H O

0 3 t+ P

0 3

er
pUl

tr ro o
rt 13
fli

ro
tn r+

O 3 tn

ro ro tf pi i
P
rt
3 3

tn tn

ft 0J P3

P Hi ft

"<
tr
PJ

C D
PJ

<
H 1
i
flj JJ

ro
P tf tn
pHi

P 3

PJ

Hi

P rt

ro ro pj o
3
QJ P< rt i Hi

tn

0 3 C D

tf
P* 0
p-

tn
3

er ro to ro
cn cn
0 0 3 O
rt Hi t+
Pi
-1

M C D 0J Pi

<
13
3
C O pi Qi rt

Z
Z 0 0 3 X
ti

3 ro *<
3 O
01 0

ro tr
po cr *:

ro o 3 rt c n ro QJ a

3
rt ro
ti

0
P tt P p

Hi Hi

H 0 3

c n o Ij
tn
X
M M

Tl

ro ro
tn
i

3
O

er P ro ro 3 o O 0 *< rt 3 er r t o ro H p (t
Z
H C D QJ

13 0
3
H
ti

3 3 tf
13 C D H
Pi Hi pti ti

0 0 0 3 cn 5 0) H

< ro
tn cn tn
PJ P1 ft O

T3 P
ft

tf
3
Hi

tn
O

0J H POJ rt P-

er ro
TJ O tn
prt p-

tn 0

c
tn tn
p

3 H
Hi
fli

cr ro ro o > ro
tn
O

*<
tn

tf
X

H fD

cn O 3 cn

ro c n
fl) Hi fl)
0 3 3 rt tn 0 0 3
ft

O 3
pj

rt < ro ro 3
H

er er ro ro
p-

13 O 3
H 0

3 tn
pi

rt 0 er p 3 er ro ro r t H i f t 10 Oi tf cr 0> P 3 ro < 3 0 3 a H ro ro ro ro tr QJ QJ 3 P 0 c n 0 3 3" 0 p TJ ro p r t cn cn 0 << OJ TJ 3 ro tr 3 a OJ 3 H 0 Pi 3 p 3 ro P

< p er p
0
Hi
pj

ft

H 0 H

Pi

pi

p N

P*

0 p 3

rt P

in
3
OJ I_J.

C O p ft p

0 3

*
P1 p PJ P P

ft

0 3

< c n 3
tn
Pi
IJ.

QJ

Hi

ro
Pi 3 P C D

Ul

fl) o

o
3

tf tf
tn
O O 3
rt pj
1

ro n er
tn
pti

3
tn

3
C D
QJ ft

er t i ro er n pro
< C D
H

P C D 3

ro
p3

TJ H 0 13 P

rt-

PJ

ro 3 er ro
O 3 H O

er er ro er ro
Hi PH

cn cn

ro
0>
H-

tn

3
p-

0J

tt

p Hi P P ti

0 ft

z 3
cr ><
rt
ft P Qi rt P1

Pi

ft PJ P*

H 3

m
rt
H

* P 3
H H O H 3 H P tn

PJ

0 3

OJ

3 tn

ft

TJ 0 C O p-

c fl)
0 H

.
Z rt ro cr

tn

tt

P3

tf
H

P 0

n ro ro o 3 c n ptf
3
OJ P1 C O ft p-

er ro ro o
0 3
Ul

Hi PH ft

rt O

*< er p-

H H 0 H

O 3

tn

C 3

Ptt

P 3 C D O P tn

3
rt
0

ft p

ro tf c ro
P 0 3 tn P

H 3 C D

Z o

OJ pi

TJ H 0

tn

<
fli

h-1

13 H 0 13 C1

Pi

pi

3 tn

~ . P

pi

o Hi
3 O
ii
PJ

Z p-

fl)
3 Pi

n <
3
Qi p

in
P H

0 3 in 3 P.

Hi

er p 3 s c n 0 H i 3 tn ro H p 3 C D tf f t tf er & ro ro ro o er & O Hi rt ro f t o > er H ro

O O
Hi

ft pj

p ti

0 3

0
fli ro 3 tr M n ro <

QJ P1 P1 rt Pi t. C D

Hi

0 H P

o rt
tn
O O
C D ii
V

P C D 3

er

tr H ro ro tf P ro

rt

O 1
Hi

pj

ro P i tf fl)

Hi
ti

rt

rt

P 3 0
Hi
ti
i

KJ tn

rt p

r _

S
ro <
cn rt C D
O
Hi ti
PJ

0 0 3 rt H O

o po tf c n 3 3 rt fo PJ o M O 3 tn H ro * a 3 ro f+ C H tn er p* Z o 3* pP

ti

ro c n
_ Pi tf ro
H tn

p. ft

O Qi

C D 0 C

tf
tn TJ rt
p-

C O

3
13 H O 13 P1
P

fl) er H i tf ro ro P n ro t 3 n r pP n tn < pC D ro ro H c n ti ^< pi er ro cr H ^< CD Hi H tn tf fl) tpr- ro 3 tf P 13 ro C D H 3 3 QJ pi n Q p i p3 ro ^ c n rt f t ro pC D P

Hi

Tt H O TJ C1
P pi

pi

0 3

OJ

er ro cn ro n n flj o TJ 3 o > Pi tr p tr er CD ro o p 3
P
PJ 3 H H 1 NJ

er
3

tn cn

<
pp-

O
M

TJ
ii

O 0

< rt ro cr
P1 pi

Hi Hi

ro er ro ro ai
Hi

z pj
O 3 O 3
PO 0J
pi

o H 3 < C D
PJ

2
H
M

z p0 H P O TJ P p 3 p p

Pi

er ro 3 o tf pi C O Z ro tr ro Z 3 < H ro
3 3

li 0 C D

< ro

ft

fl)
H
PJ

Hi

rt 3* A > 1 C D Tl
h*
i

rt
Hi PJ O rt P P

Hi

tr r t ro er

C D fD

M
ti

^<
P

3 tf cr tf
P 0)
H

tr p-

P(t

er ro t i tn er tr < P Z ro tn ro tr p tn p c n H n tn ro ro cr ro H P< pi c Z p OJ o p H P1 c n M 3 3 QJ tf er
fli

Pi <
fi

QJ

< 0 P1 ro H p c n Pi rt D ^< c n C ro n rt M tr 0
3

Pi p.

3 Prt

tn

O p

z
pi
PJ

to

er ro
P-

0 H O 0 3

tf ro ro tn P- Z
3

tr

p0

tf

Hi ii C D

p-

O H P

ro ro
3 13 P

z pM T.
p 3

rt 0

ti

0 H

PJ

3 0>

0 rt

P1 0 P

0
rt
PJ

er ro
P tn

ft

0 0 3

p N

ro ti ro
P
P
ti

r cr tf
K

< ro c n c n C D ro 3 er
0

TJ H

PJ

ft

Q > H

tf ^< er 3 ro tf

cr

tr

ro < $

3 P

-s er
ppC O

H C D

P-

3 Pi tf
3

er
p-

li

3 p

3
pft pft

Pi

tr ro
3

pi

0 3 rt

pj

H 0 P

ro ii 3

tn

ft P

ro a <

n er ro

rt

*<

0)

c n 3

O 3 fD 3

p
ft

tn

tf er P Pro *< ro
cn
H 3 O 3
rt p-

rt tn
p3

PJ

ro
tn
pi

O
tt

s 3
1

z p

0 0

rt H

PJ p

0 0 3

PJ

rt

pi

ro

tf

er ro r t Q tn ro H tr Z er p * H i H TJ C D ro ro H3 O H tf 3 PM p- 0 C D C P3 0 tf PUl Prt 0 P H 3 c fl) ro O TJ H 3 in < O Hi H 0 er 1 QJ i P ro < H Pcn 3 < < 0 PP> O ro ro ro P 1 ro a 0 r t 3 fl) z o p - 13 H > 0 H i 0 3 0 Ptf Hi 3 3 H i H 13 fl) tf pPC D tf 3 1 tt a OJ P f t O Hi tn P cn rt tn o ro ro QJ QJ ti ro P 1 ipicn H tr er p3 C D ro OJ P1 ti tf tf pP p er n- ro p - t i CO 0 3 D cn *< er C tf ro P i rptrt P rt QJ O tr TJ H i PH ro 0 O O p. ^< a
Pl
Pi

C P PJ n H tt ro ro fl) < fl) f t 3

PJ

er ro *<
3 0 O
I p-

O H P

Hi

a ro o
0 0 p 3 rf

ro

13 H O

er

<

Q Z

H i 1

tf
rt
O
tt Hi

Pi

Pi

Pi

ti
H 1 0 Z P

Oi

<
P >
3

rt
tn O J tn rt O J

er rt &
tn P
ti

rt ro pi
tn p

rt p

ti p

Q
PJ
r

cr
13
X TJ C D

fi

rt
|NJ

er

ro rt ro c rt Hi

rt P P>

< fl)
P

3
i3

P 3 O J M tn p O J rt P 0 3 tn

O J rt p 1' 1 |t3

Hi ft

P*

Hi

z ro OJ tf < pi ro fi i a Hi
O O 3 rt

fli

P 3 C D

n er rt ro O1 <! ro
3 i
Hi

K IH IH IH 1

rt z O J rt

rt p 0 3

C D tn TJ

H 3

< ro
tn p
0J

Hi 0 H

< ro H i z rt QJ ro < tf

p 3

C D C D

Pi

er ro rt ro rt

C D

Hi

0 ii 0

tf tf
ti

Pi

p 0

ro
QJ
Hi Pi

13

O J rt

z
ier

tn rt D fl) er C P tf ro tf 3 O J * QJ ii tf M P i t i ro cn 3 er p ro OJ 3 H r t tf

P M1 P 3 PJ H

&
rt
iro

Hi

ro rt P ro cn er ro in
i 0
iro 10

cr ro ro Hi ro ro a
Hi Hi

0 3

Z p 3 Q< 1

C D
fli

pi H M tn tn rf

cn n p
3 C D 3
ll

ti

rt ro tf cn er
M P

p
ip

p)
3
1

Pi Pi

er < ro H
p
10) IP i

Z a ro t i H i fli C D ro 3 p* o < er o ro 3 ro *< ro rt p < z ro 3 & s: n fll H C rt QJ 0 rt ro cn p < 3 P i 0 < ro TJ Qi tr ro

in
3 C D ti

ro
rt

ro n rt in ro
Hi

C D C D

rt p 3

ft

ft

0 3

3 O P

<
irt iro
QJ
ipj

rt z
p

Pi

1
Hi

ro <
tr
fli

flj

o >
Pi rt
3 3
Pi
fli

H H 1 3 C D

fli

P<

tn 0 3 C D C D 3

& rt
fl)
IA Ui
ft Hi

P PJ 3
OJ

P.

rt ro
oi tn
P1

ft

ro cn n
DJ H.
Pi Hi

Pi

H P

3 X C D

P 0 3 C O rt
O 3 0

Hi ft

ro rt er ro n ro
in
Hi

M P

Q
tn P1 0

er ro rt
< ro c n
Hi

!"* H

ti

OJ

Hl

tf c
IP IO

tr ro
0

ro r t cn cr ro
p 0 3

c n ro rt fli ro 3 rt

tr ro 3 3 a ro fll Pi rt 0 cr tn ro H p r t < tr cr < 3 C D rt p* ro er ro to f t ro CO er P ro ro rt o t t H ro 0 er tn TJ rt 3 ro rt P* rt ro p- rt to 3 < 3 p ro P p O J ro cr n 0 3 P er P rt 0 cn rt a . o cr pj

TJ
0 s

rt Z
P

Hi

PI

rt

p.

H 3

3 TJ C

er fl)

pi

ft

0 H 0

3 C D rt rt

p 3 ti

ti

rt ro

0 rt p 0 3

ro
0)
QJ

3* Oi ft

z
0
Hi

P1

tn

3 C D ft rt

ii 0) ft p 0 rt p 3 C D

0
Hi
PJ

Q
i
Qj

QJ

ro h ro tr
O J
H
Hi pj

0 1 ii C D tn TJ O 3

*
Pi 3 tn
O 0

c n ro rt rt H ro ro t i pi z ro p r t n H er pi fi 3 tr rt cn ro p 3 c n ro < M rt tn _ 3 e z ro ti Hi ft H p Ul ro o _ _ p i Ui P3 O J X er 5 tf er H 0 13 p i ro f t p- ro C D 0 H ro ro tr tn 3 PJ tn 3 rt H i ro J 13 P P z ro 3 tO o pt p PJ flj P rt 3 rt cr tf QJ p Oi f t ro tn H n p) 3 QJ ti P 3 3 H 13 o ft PJ r t P rt P TS 13 p 0) pj er cr rt ro c n 1 P 3 O J TJ 0 ro ro PI H rt PJ rt H 3 n r t *< cn P i a p < r t er Q J ro 3 cr p er ro 0 n r t Z rt 3 < 0 ro H i O p rt P Pi 0 a fl) H 3 tn 0 < p i ro O J rt* M n tn r t tn H i r t H i H r t p r t TJ 0 P tr tn C D PJ 3 D P1 M rt tf o C

z
tn

tn TJ p tn p 3
Hi

ti

z < ro
tn P" tn 0 p
P O

rt ro rt
tn p rt

p tn

er
P

Hi

Hi

ft
QJ

to rt P O J rt

to C D P 3

IP 10 13 1 10 IHi 1 lit

rt 3
p.

rt rt ro ro

rt P0 3
ft P-

rt

0 H 0

rt >i Oi 3

OJ

Hi er H. ro C D er ro ro o cn tf

Z C D
i3

Pi
ti

tf
i P
ti IHi IHi

ro ro a
ier
rt Oi tn

rt er p ro Ul

3 fl) cr p ro 3 rt tf tr
X C D

< fl)

ro tf rt H H i n ro rt
3 C D tn tn 0
Hi

a
0 3 Z P tn

Hi

P
Hi

tr ro p i er p) ro a rt ro Hi Ul tn 0 p ro QJ ii p 3 c n rt c o 0 *-* Hi H er
H P tn

o ro ro
3
ti

00

oi H

Z er
rt ro
H 3
OJ
Hi

C D
0

rt
ej pj h a
0 3

fl)

3 O

3 O J

rt

tf cr ro
iro i iro

O J 3
Pi

C D C D 3

^<
ft

er ro rt rt ro ro tf . P ro PJ 3 n rt
OJ

Pi

tf n

ro

p p

0 3 rt H P

Hi

P OJ

0 3 rt H P

tr

cr

*<

<

er
rt

pi rt

rt Z

<_l.

er ro ro ro TJ p . per ro ro 3 Hi 0 tf o 3 Hi
C D 0

fl) rt
0
Hi

3
QJ P1 H

tn

P rt p 0 3 O J H
0

Hi

O J

z
tr
ft

pi pHi Hi

pi

cr
O H 0
iro
I_J.

p tn
N

P rt O. er

<
C D
P

J9!
Pi

0 ii 0

ro
Ifli

rt
Hi

P rt P 0 3

C 3 0 rt p 0 3

fi

p rt p 0 3 Ul

er ro n 0J ro < cn ro
i

Ch P

cn
\<
11
IHi iro iro ipi

er ro Z rt z
O J t.
Pi

cn ro H i ro tf rt Pi P ro ro tf 0 3 P 3 0 ro <: 3 Z O 0) z K < tu H i ro < i ro 0 Hi rt p . O ro ro t i ro a H . Hi i 0) Hi ft n

iro in Irt l 10 IHi 1 IS

< rt ro
tn
Pi

< ro ro rt
C D 3
Pi

TJ P
H,

i K 3

tn

ft

er c tr
P >
Qj

tr
fl) ro
pi
fli

* tn . tr H 0 fl) H s rt tr 3 P ro Hi n H 0 ft ro H ,
rt p 3

rt p
Hi

tn

Qi

H O J 3 to

c c o
pi

ti

P 0 3
rt

6 c
11
3
Hi

ro tf
0 H P C D tn

cr ro

p> P tn P 0 3

p tn ft tn

Hi

tf ro ro rt 0 tr 3 ro PJ ro Hi
C D C D P. 1

ro tn cn c n
0 Po

p H p rt K

Hi

M rt tn

er ro 3 tf ro a
IHi 10 IH
M

rt pi p

rt 0

rt

ro
O

ro rt
p

ti

er

fl) tf c p rt ro p.

n
C D
Pi

P-

Hi

0 0 H,

p3 P" K O J H1 P
Z 0)

0 0 3 rt

C D

Pi

p> 0 3 p 3 P tn

< Hi fl) ro c n o

rt

rt

Z 0)

rt

er ro er ro tr ^< < ro tn
i

tr

3 tf rt p < X C D < ro ro QJ ro c n rt rt 3 tn ro C D QJ p i P P 1 cn 3 P C a *fl rt tn o P ro ro 3 tf a TJ P tf P p p r t rt C D < 3 P ft ro < 3 13 P H rt 0 1 p ro 3 P 3 < tn

P3

t.

tn C D P*

ro tf

0 3
Pi

0 3

Hi

pi

LJ KO

p
p ro

rt ro tf
5 0J H P<

Hi

< er ro H ro ro 1

rt 0 rt P)

c c n

rt

\z Ifll
Irt
IQi

m O J tr
H

p 0 p tn

< ro
i
ti

^<
Hi

P.

rt z QJ rt

tf ro
C D pft

3 PJ

rt p cr

er Z >? er D 0 ro Pi C < rt z rt ro 3 ro ti QJ p< a P rt er ft p tr PI p H i OJ rt < n rt P ro H i tf ro 3 0 z rt OJ f t pi < n 3 er ro ro P i ro tn Oi ro n

H

3 PJ K

Pi

ro rt cn Ul < ro ro ro
Pi

0 0
Hi

Pi

tr <

P 0 3 rt >i 01 P

P rf P 0 3

P rt p

3 p i

tf p p cr

P 1

P rt P 1

01 i

tn

er ro

P 0 3 TJ P I

CD
PJ

TJ 0 tn pl

tr
1

3 P1

0 3

3 p 1

tr ro rt ro rt

Pi

rt Hi ft Hi rt O 3 P 3 Pi Hi Hi 0 Hi 0

ft

_ >
ft Z p P *J pi 0 3 Hi

rt ro C O
ft P H

H 3

p li

rt

rt

rt p 0 3

Hi C D O ft

rt C D 3

p rt

z
o
TJ H p rt P ft O rt P P1 P 3 f TJ H O Z tn fD

rt

er p
0 3 cn p Pi C D

TJ P

tr P ro rt rt 0 3

ft p.

cr er ro
H 0 3 P_ P
P
1

tn

rt ti <: ro
0 3
ti

c n er ro Z < ro i rt p
P
PJ

0 H 3

tr rt rt 0 3
rt rt 3" C D

0 Hi

3 tn

ro

Z p rt

p p* p

rt

0 Qi C D M

er ro ro cr ro tn <
1

PJ

Z P tn

0 Hi

rt p 0 3

o rt rt ro o cn ro tn Z ro rt rt 0 ro 3 ro tn 0
0 Hi 1 TJ tr P tn

cr ro a i < C D Hi
in tn P Hi 0

p p 3 C D Pi

er p rf ro tn tf rt fi

tf p < C D

rt p 3 C D

^ < 3

ft

O Hi

Hi 0 pi Pi Q Z H 3

p Qi P 3 P H

rt

0 Hi

a * p er ro rt ro tn
P
M

3 rt ro P ro

tf
0 H P

P 3

QJ

rt

pi

tn

ft

er ro rt ro
P
H H H 1 ro LO LH 0

8
rt tn p C O

C D

er 0 rt a
rt p 0 3

Pi p tn P P M 10

ro er 0 ro rt

c n S C D rt fl)

ro n
p

P 0 3 tn

H Q 2

p rt p 0 3 P H

P Qi

z ro z ro
rt ro 3 P

n P

rt

<

rt tf P ro M P P rt rt* rt p ro < H p ro 3 tf z p < z ro p < C D ro cn P" ro < M

1

TJ 0

t_i.

rt ro ro cn
tn tr

rt p 0 3

ti

er ro

c n

< cr ro ro p z H P p < Hl 3 ro P tn i Hi rt ro P pi ro 3 ro rt Pi cr i ^ < P 3 Hi o 0 ro 0

1

P H

p tn

Hi

3 0 rt

0 Hi

p rt p 0 3

2 p

rt &

rt 3" C D

0 0 3 fi

rt fl) tf rt P 0 ro Hi
Hi rt p Qi

c n rt fl) Qi

ti

TJ TJ P1 p.

ro Qi

tf

tf p < ro 3

ro X ro P o H
M K H ft P TJ H P

0 3

< P

er ro n

Hi

TI

C D 3 P

cn P

ft p 3

0 P rf

Z p P> pi

P M P

3 0 Pi

ft

p*

rt p 0 3

* ro rt ro < ro tr ^ <
P

ro *: H tf p 3 tr P 3 0 M fl) rt 3 p
P C D tn

tr ro H rt ro p rf

er Hi ro rt < rti ro 0 c n rt c n 0 ro M ro tn P rt ^ ro 0 rt tn P H tr C O P rt rPt tf 3 ro f t rt . H P < H

0 Hi

rt P p

z p
tf p
p 3 Hi p

P3

C D tn

p cn p 3

N 0 3 ft P P1

c n P 3

tn C D ft 1 p 13

DJ 3 Pi
ti

rt p 0 3 in

3 0

n
ft O

rt r. r t ro H cr 3 rt rt 0 ti P rt P er tf p 3 er tn cn < < C D H o H P rt tn H tn C D 1 ro ^ rf Hi cr 0 0 ro O rt 0 rt n ro rt 0 ro 3 ro rt g 13 0 rt a 3 p J cr < P P P c rf 3 z tf P 0 3

QJ

ft

rt ro ro tn rt tf ro C D p 3 p 0 ro rt rt 3 QJ < P ro ro H ^< r t 3 P K S 0 rt P ro in < o cr rt C D 0 ro M 3 H i TJ o P C D 0 0 3 3 rt 3 0 tn 0 H H P ro rt tn r t Q c rt fl> J ro M ro pi tf ro Hi p X

3 tn 0 3

0 Hi rt Hi
ti

P> P

TJ C D

Hi

rt
o

QJ

ro 3
p tt

rt rt p
P
C D rt rt

ro er P

3 z ro p < 3

0 rt P

rt ro Hi 0 a 0 Hi p rt t i o ro rt er a ro ro pi rt P a ro p M o p Hi P 3 Hi rt C D tf rt rt < C D ti ro er 3 ro O r0t fl) 3

rti 0

P rt P

. ro *.

P rt 0 Hl 0

er ro f t 3 cr 0 ro QJ ro < w ro in rt er ro ro tf r t P 0 ro I 3
1

< fl) Ul

tn

ro M

Z er P ro rt tf tn < LJ. C D P PJ tf 1 P ti c p Hi C D H i pi er D tf rt 3 ro ro C C D ro rt ro H i pi P J 0 M a i 1 P rt ro Hi 0 c n H i tr 0 P rt tr r0t P p 0 0 P tn rt 3 2 rt ** Ul p Ul P tn ti 0 rt er . c 0 p i 0 3 ro i3 Ul H i ft ro P i rt tr n tc rt 0 0 0 ro ro pi PJ & 3 z P P 13 ro ro Hi n rt Pi < Kcn 1 p i ro tn 3 rt l tf

rt P H M

fl) P" *.

TJ TJ P* p 1

*<

tf P r t 3 rt P er ro p a ro 3 tf P P H rti 0 ro C D 3 P tf 3 P P rt o tf Hi P < rt p P t rt rP o 3 ct C ro 0 0 z rt tn Hi PJ tf P 3 p P p pi 3 Z ro n rt 3 C D pi P 0 fl rt 0 M rt 0 rt tr ro rt H i ro ro p o 0 tn 3 P er rt ti UJ ^> tr o o tn 0 pi er p ro Hi p ro ro . tn 3 ft 0 rt* C D p rt P tn 3 3 p rt 3 0 t i tf fD ro er P tn P p ti P ro 3 rt p 3 er H P P *< i ft rt H f i P H P er ro 0 z p 0 ro p n < er 3 f0t rt PJ ro ro tf p rt 3 3 rt P P rt o p tf ro o rt 3 0 3 er . 3 ro t t CD UJ P cr H i rti 3 P_ ft Hi ro rt 0 O ro ro rt i3 rt rt ro tf o rf er ro ro er cn C D ro 3 c ro er 3 ro rt 3 P QJ P rt rt C D er *< tr rt 0 P rt N ft rt rt <: P C D P ro 3 P* Hi 3 rt ro 0 0 0 tf 3 < ro z rt ro p i ro 3 rt rt rtl H p p ro Hi p 3 0 rt ro Hi 0 ro z 0 C D rti tf ro 3 0 Pi C rt rt fl) p ro ro P i O rt Hi 3 3 1 0 o ro C D < rt C D rt tf P rt r t cr cr 0 < C D ro z 3 ro p ro rt O 3 ti tn er tr H i ro H i ro QJ P P ro P rt tf 3 tn C D Hi P ro Z tr r t P 0 O P ro 0 rt ro tf it ft P < rt ro rt Hi C D p p P P P ro o 3 3 0 p 0 0 n rt* rt rt rt 3 Pi 0 3 3 Qi er p p TJ er rt ^ < ro P 3 Hi 0 0 ro < *. 3 H i tf H P 0 P P ro f t 0 3 3 ft ro < rt TJ rt 1 3 1 3 ro H i z LTI t i p 0 h rt rt er tf P ro f i O ro < H1i P er rpt tn rt rt ro 13 P ' ro P Hi P O ro . tr tn 0 PJ P QJ ro H i ro p P C D P P 3 rt . < 3 tn 3 O P rt ro 6 fD P P H 0 0 PI tf < ro H P P H i er 3 rP C D t 0 Z ro ro rPt f t p ti Hi P 0 tf ro P i tn P 0 0 3 tf er 3 T J er C D C D P P P ro rt (t 0 0 P o rt P i C D pi p tr o 3 ro P p rt 0 C D 0 rt 0 0 ro 3 3 3 3 0 p P rt p rt 3 cn p rt r t rt z tf p 3 ,^ ro rt ro tn P" ^ rpt p H cn C D o ^ < PJ O 13 < tr 0 3 C 1 Hi ti 0 P tr tf ro 0 rt rt Ul P Hi 3 rt P i P f t rt P O P P < rt 3 rt 3 P p C D 0 0 tr 3 3 p P tn H 3 3 3 ro o Hi P tf ro Hi Hi p ti ro O tr r t p i 0 0 er H O > < rt P P ro H i cr rt rt 3 P 3 *< E 0 tr 0 p TJ C c n ro 3 Hi 3 Hi p 0 3 H t 3 P n ro c rt rt H 13 V ro o rt cn rt rt p p 3 rt er 0 p 3 rt Hi 0 P e cr O P tn 3 tn 13 NJ rt fl) P ro H i 0 0 3 * p i CD O ^ < (t p p rt Hi H P 0 rti 3 P 3 p rt tr cr ft o Pi 0 tf ro tf ro rt tr 3 0 0 er ro * tn t i P 3 rt P c ro er ro rt rt rt P PJ P ro r t 3 1 c C D _ _ ro 1 l
X

TOTAL
U

\ ^

0.354 2

0.444 9 7

0.523 10 21 12

0.600 25 19 16 14

0.713 10 26 8 10 9

0,803 38 12 14 18 4 9

0.887 37 35 14 15 7 5 9

0.354 0.444 0.523 0.600 0.713 0.803 0.887 T

in tf/m 2

Frequencies in rad./sec.

CONTRIBUTION I
\
W

"

2 ^ \ 0.354 0.444 0.523 0.600 0.713 0.803 0.887

0.354 10

0.444 19 25

0.523 27 32 32

0.600 23 31 34 31

0.713 21 26 27 29 23

0.80 3 20 28 30 29 27 25

0.887 2i 28 27 27 25 28 25

2 T. in tf/m Frequencies in rad./ sec.

TOTAL - 0.4 * CONTRIBUTION I

* 2 ^ \ 0.354 0.444 0.523 0.600 0.713 0.80 3 0.887 T._ in tf/m Frequencies in rad./sec. 2 0.354 2 0.444 9 3 0.523 15 10 1 0.600 20 10 13 2 0.713 15 19 5 6 1 0.803 31 5 3 11 7 1 0.887 37 24 15 10 6 7 1

Table VIII-1

Longitudinal force in head waves. 141

TOTAL
\ . "2
w

\ ^

0.354 1

0.444 18 9

0 .523 25 19 16

0.600 32 38 38 17

0.713 14 14 36 11 13

0.80 3 39 16 31 10 8 12

0.887 29 5 27 15 3 4 10

0.354 0.444 0.523 0.600 0.713 0.803 0.887 T

in tf/m 2

Frequencies in rad./sec.

CONTRIBUTION I

\ ^

0.354 5

0.444 10 25

0.523 21 33 54

0.600 22 24 50 40

0.713 18 17 47 42 22

0.80 3 14 13 36 33 34 23

0.887 17 8 43 35 26 34 22

0.354 0.444 0.523 0.600 0.713 0.80 3 0.887

T. 2 in tf/m Frequencies in rad./sec.

TOTAL - 0.4 * CONTRIBUTION I

\
W

" 2

l ^ \

0.354 1

0.444 21 1

0.523 17 6 6

0.600 32 30 19 1

0.713 16 10 19 18 5

0.803 39 12 18 13 13 2

0.887 30 5 10 6 8 9 2

0.354 0.444 0.523 0.600 0.713 0.80 3 0.887 T

in tf/m 2

Frequencies in rad./sec.

Table VIII-2 142

Longitudinal force in bow quartering waves

TOTAL \
w

"

2 ^ \ 0.354 0.444 0.523 0.600 0.713 0.803 0.887

0.354 1

0.444 35 5

0.523 9 22 33

0.600 13 50 63 79

0.713 34 42 3 57 91

0.803 19 69 24 15 58 90

0.887 24 116 21 8 6 60 95

in

tf/m

Frequencies in rad./ sec.

CONTRIBUTION I 0.354 * 2 ^ \ 0.354 0.444 0.523 0.600 0.713 0.80 3 0.887 T _ in tf/m Frequencies in rad./ sec. 2 5 43 61 35 159 102 49 44 165 259 11 134 73 192 298 17 41 26 50 221 297 15 79 40 41 69 230 309 0.444 0.523 0.600 0.713 0.803 0.887

TOTAL - 0 . 3 * CONTRIBUTION I \ .
U w

\ .

0.354 0

0.444 46 13

0.523 6 43 2

0.600 17 47 15 1

0.713 31 29 20 13 2

0.803 21 65 19 21 17 1

0.887 21 103 17 5 23 16 2

0.354 0.444 0.523 0.600 0.713 0.80 3 0.887 T

2 in tf/m Frequencies in rad./sec.

Table VIII-3

Transverse force in bow quartering waves. 143

TOTAL
\
W

0.354 364

0.444 7085 1527

0.523 9284 16655 2554

0.600

0.713

0,803

0.887

2 ^ \ 8809 7551 7577 782 2 T

?

0.354 0.444 0.523 0.600 0.713 0.803 0.887

14 33 3052 5870 5124 1193

4448 5067 2937 2363 4576 200

2544 12269 2994 2445 3574 4998 562

in

tfm/m ir rad. / sec.

Frequencies

CONTRIBUTION I
o _
2

^ \

0.354 1881

0.444 1993 11633

0.523 3637 16368 9066

0.600 1611 26015 5377 1909

0.713 3742 3456 11610 14844 4366

0.803 1669 9644 6747 12883 13830 678

0.887 1872 7040 2808 2498 14818 13751 1003

0.354 0.444 0.523 0.600 0.713 0. 803 0.887

2 T,2 in tfm/m in r a d . / sec. Frequencies

TOTAL - 0 . 3 * CONTRIBUTION I 0.354 2 ^ \ 0 . 354 0.444 0.523 0.600 0.713 0.803 0.887 T.-j i n tfm/m in r a d . / s e c . 261 Frequencies 2 200 7589 1963 8223 15856 166 8634 8282 6081 209 2515 2434 2626 677 117 3989 3250 3330 2854 1402 404 2685 11398 2292 3146 2 861 1196 0.444 0.523 0.600 0.713 0.803 0.887

Table VIII-4 144

Yaw moment in bow quartering waves

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: F, (t) -C 1 l
1

/ a.pgc'15 WL

-n .d

(VIII-1)

- transverse force:
F

2(t)

"C2

/ ^P^l^ WL

- n 2 .d

(VIII-2)

- yaw moment: M,(t) w -C_ J J in which C , C

X
_ _ 1

/ ^pgc* 1} WL
* J

. (x.n, -x 9 n.) -dA i / z i

. . . .

(VIII-3)

and C

are gain factors which express the ratio

between the total wave drift forces and the contribution due to the relative wave elevation. The above equations contain line integrals around the waterline which involve the geometry of the hull form at the waterline and the instantaneous values of the first order relative wave ele vation. It must be remembered that C is the wave elevation as measured relative to the vessel at the waterline. Its value at any 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.

MEAN WATER LINE ON VESSEL

Fig. VIII-1

Wave probe measuring relative wave elevation t, . 145

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. The measured relative wave elevation at a point along the waterline contains first and higher order components: C r = C^ 1 } + e 2 ^ 2 ) + The square of < ; gives: .r2 = e2C<1)2 + e3.'11.'2' + 0 ( _ 4 ,
+

(VIII-4)

(VIII-5)

From this it follows that the lowest order in the square of r (1) involves only t as is required for the elevation of equations r 3 (VIII-1) through (VIII-3). Components of order e and higher are generally small and of high frequency and will be neglected here after. In irregular waves the first order relative wave elevation in a point along the waterline may be written as: C* a ) (t) = Z ci.1' -cos(o) t +e.) i=l i (VIII-6)

Squaring this expression gives: ll)2 ZI ' (t) = r (1) (1) 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 (1) (1) Z Z %c; ' C J . X ) . c o s { ( w . - u . ) t i - l 3=1 i j i J
N N N N N N

+ (e. - e . ) } + i - 3

(1) (}1 ) } + 2 Z ^C; K1 . c o s { ( u , + u . ) t + (EX + e . ) } r r i=l j =l i j ! 3 -i -3 (VIII-7) 146

This shows that the square of the measured wave elevation will in general contain mean and low frequency components corresponding to the difference frequencies and high frequency components correspond ing to the sum frequency of the frequency components in equation (VIII-6). In Figure VIII-2 time records of ci M t ) and the square of C (1. (t) are shown schematically. In this figure the dotted line indicates the low frequency part of the square of the wave eleva tion.

<t)

Fig. VIII-2

Relative wave elevation and square of relative wave elevation.

For the model tests equations (VIII-1) through (VIII-3) were evaluated by replacing the integrals by simple summations of the following type: F.(t) - C . K _ .pg Crn Xt).n. n .A. n n=l (VIII-8)

Al

F2(t) - C 2

K I n=l

kpgZrn

(t)-n

.A n

(VIII-9)

K M-, (t)ft*-C. Z

n=l

n~ ^pgc rn (t) . (x, In 2n

- 2n X. In n, ) .n A J t (VIII-10)

in which K index n

= number of wave probes = denotes wave probe under consideration = measured relative wave elevation of n wave probe = co-ordinates of n line form wave probe and the centre of a

ln'

2n

straight line element approximating the local water= length of n waterline element = direction cosines of the nth waterline element

M
n,

, Pr,

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

In'

2n

SERVO UNIT

WAVE AND CURRENT DIRECTION Fig. VIII-3 Position of wave probes measuring relative wave ele vation. 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, does not cause appreciable phase lag in the low frequency compo nents. With this type of signal this does not form a problem be 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 trol system based on the position error signal. In such cases the high frequencies coincide with the wave frequencies. 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. From this figure it is seen that the phase lag remains less than 45 degrees for frequencies up to 0.21 rad./sec. full scale. The 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 ward signals range from about 0.6 rad./sec. upwards, which is twice the lowest frequencies present in the irregular waves with the 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 1.0 rad./sec. At this frequency the filter amplitude is 40%.

tf) -180 CJl a

TJ

(/) - 9 0
(/) <
I
0.
LU

c
4.

i_ IOO

<
Z> Z LU P P

50

<

10-3

10 - 2

10 -1 FREQUENCY in rad.secr
1

10"

10'

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-feedforward.

1 0 N H U 0

0) 0)

TJ

n tf

41
> i

a c tf tf
0) T3

rfl

fl)
r.

fl) rt
O J
0 41 P Cn

O J

rrt

> u u u
r.

r-\

O J tn
PH

Pi. Qi. re

tf tf

rd C Cn

1 0

0 N

T3 O

tf

4J O

SH

rt r. fl) tf

SH

tf tf
in 4
J 11 m
CM
(V

fl) c
V. 0)
H

P
r t

C A 41 H M rH fd p H TJ 3 4> ri tr P

0)

ri

tn

tf

_ _ >
c p P 0 0 pi M < D 3

rH 0 in

O fO a)

0)

tf

tn P ri

tf tf c 3 0 tf
fl) rt
tn
it

>1

o
0

tf
C A
rrt 0) C A

>
tf) tf) rt 0)
H O J tn to O) H

iH 0) tn

in rH td P tn

tf c

r t

r t

TJ

tf tf ri 3

4H 0

cn td

p 0) 3 D1 P

tf o

0)

c c

rd TJ H fl) Ch P Ui rt M ffl TJ tf

fl) tf o

tf)

rfl

e 0)

fl) fl) tf) > ffl tf

QJ

H rd

P O 0

rt

TJ

VO 1 H H H

SH

tn

ai
i

O
HP
O)
*"

0)

o
y y / /

E
CO

oi
m
%i~n
i_ji

//
7f I /

Pi 3 1

>
3 P. 0 H Qi H

tf
til

EH

H 0 iH

rt rd s tf . ro c u > >
(fl tf c

TJ 0)

4H

{Ij 0

tf

H 0

u tf o

tf

tf

>

td 0)

P O N H

tf

rj ri

TJ ai

0) TJ P

0)

rt

SH

tf
0

4J 0)

tf]

tf)
tf}

tf c
tP P
r t

H 0)

ei
S

_
y

A
y
^y

0)
rfl 3

T3
OJ

<~~

rt
ffl ii

*.\

3 D* C D U

u rt
4 H 0

El 0)1
PI

rt tf
LOI >4 ml l Pi Ci HI PI Ol

41 0 _ _ MH tf P 0 O fd 0) H > i Cn <t PH Pi

tf tf

N rt rt tf rt EH

P U 0) P

3 a C A

i TJ 0) 0)

ro M 0) C 0) tn

rt tf

rrt

C 0 O

0)

tw O

tf

tf tf

< L >

tf 3 tf)

> i

3 tf . c tf) 0

i 01

tn td

0)

tf

tf rt o

tf tf

TJ 0 ri

u tf t n a) rt rt tf H h H rt P c tf ra ri

< L )

0)

e
0) 0

)H

tf o
rH

> rd 3

rrt

TJ 0)

0 H 0)

T3 P

rH rfl

rt tf Q J EH

fl) tf

tf
0

C D cn TJ 0

0)

0)

tf) o

c c
ftJ

tf tf a
C A tn P rd 4J 0 C D
W
PH

fd C tn r-i tn

> i r-\

tf

SH

OH

m
to Q J

0
r.

tf cn c rt
u
rt tf)
3

TJ
rrt

w * tf m* tf

a) tl) iH 0) TJ ttJ 0)
3 0 U

TJ C rd

H 0 H

H id

Cn

> tf
7.3S gUJ Ul ((T])
^S

ri

tf
0

tn p rfl TJ

0) 3 1

rt >.

r t

H rd C CP

tf

C 0

TJ C 0) Of 0) TJ C

tf c a)

0 fd

TJ

> tf c

U)

0)

rt tf tf

P O N

(A Q J

ri

r t

iH 0

u c

0) Oi

0 rd H

E fl>

0 EH

tf

O J

Q J 3 O**

tn Q J ri O P Q J 3 P1 0)

3 0 tf

in

V 4 rd

ro

cn iH 0)

>1

rt

tf

U)

rH fd

tn ri TJ

ri

. tf

tf tf

fl)

SH

rH

tf

iH

tf

tf

0)

.H tf

P 3

tf

0 H

>

rd

td

rrt

tf
rti 411 rti
LD 1 H H Pi
CAI

> .

Ol P.I i i *rl CP
r t

TJ ft] 0) P IA C H

f[J

rt tf tf tf
r t

TJ 0) Cn

fl)

rt

r *

cn

tf tf c
0 3
rrt
-\

C A cn 0) td to
H

H O

rt tf rt
Ptl Hi HI >l
EM

tf
r t

Ul 0
QJ

p 3 0 H rfl

0 N ri

tf c

01

tf
H

rd iH 0 04 irt 0

>

(1)

rt tf

tf tf
u 0 tf

P 0

1 Ti 0) 0) 4H

u rd tf

H c

tf

rt

TI

tf

p 0)

td

>

U oi in

H rd C tP

tf)

0) tn Ql

Tl P ffl

0)

tf tf

tf tf

tf

SH

0)

>

r4

Q J

ri

m tf

tf c

0)

tf)

ri

u c >
r.

P 0 m i H H H 0

.*

0 rd

tf

0) 0) )H

TJ 1 H rd

ai g Q) U rd

Q J

rt

C A

tn Q J

tu H

rrt

Ul 0) tn

ri

tf tf

s > tf

C A 0) P )H 0 M H

tf

1 TJ 0) 0) M H

tf tf

0)

c 0 rt tf

Pi tn ri TJ

tf tf

V. 0

rH td

QJ.

tf

ri tf tf 4J >1 O rt 4J C tf) O 0) rt tf) tf O 3 O 0 cr P i < u rH 0) > fl) u tf) ri rd > tf tf rd tf 0) tf <-i c > u fl) fl) f f l p

C A O

rt

tf

Ul

r t

rt < U tf

ffl

U rfl

s e

tf

l_ O) IL) T3

4) 4.

-180

I Ui LJ to - 9 0

<
X

a.

<D O

to 100
L

O 50
UJ

P P

<

10 -3

10 - 2

10 -1

10v
1

10'

FREQUENCY i n rad.sec.-

Fig. VIII-6

Low-pass frequency characteristic of the damping term,

The part of the control signal which was proportional to the dis placement remained unfiltered. The contribution of this part to the total feed-back control signal was weak and mainly served to limit the mean displacement of the vessel. 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 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 out in irregular waves the feed-back system parameters remained 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.

SURGE

OL

50

100

TIME i n sec. Fig. VIII-7 Surge motion decay after an initial displacement in still water.

For each sea condition a model test was carried out twice in the same wave train, once with and once without the wave-feed forward control signals. A block diagram of the control system is shown in Figure VIII-8.

WAVES WAVE FEED FORWARD

-0

SERVO SYSTEMS

VESSEL

POSITION

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 the test set-up it also was not possible to determine afterwards which value had actually been used.

VIII.5. Model tests VIII.5.1. General The model tests were carried out in irregular waves for the 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. The spectra of the irregular waves are shown in Figure VIII-5. The wave and current directions are defined in Figure VIII-3. During a model test, which was carried out for a time dura 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 VIII-13).

155

WITHOUT WAVE-FEED-FORWARD WITH WAVE-FEED-FORWARD 5 rnr WAVE

-2ml-

Fig. VIII-9

Surge motions in irregular head waves. Significant height 4.9 m.

"5 rnr

WAVE

TIME i n sec. Fig. VIII-10 Sway and yaw motions in irregular bow quartering waves. Significant height 4.9 m.

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 not necessarily result in a corresponding increase in the thrust to be applied to the vessel. The mean wave drifting forces are not affected by the con trol system used as is demonstrated from the results given in Fig ures VIII-11 through VIII-13. In Figure VIII-12 it is seen that 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 seen that the spectral density of the yaw moment is increased. In terms of lateral forces applied at the end of the vessel the abso lute value of the moment is small.

WITHOUT WAVE-FEED-FORWARD WITH WAVE-FEED-FORWARD

SURGE ^LONGITUDINAL FORCE

C\J

5F 2 8 t f

- 1 x104

S ^A 1mearr-

3 x
t/)

Li en

''

~ ^

0.25 to in rad.sec - 1

^ -

0.25

Fig. VIII-11

Spectra of low frequency surge motion and force in irregular head waves. Significant height 4.9 m. 157

WITHOUT WAVE-FEED-FORWARD WITH WAVE-FEED-FORWARD SURGE

1) tn CM

LONGITUDINAL FORCE
o _ _ _ 4) 4
CM

-|1x10

imean=-10tf -12 t t 0

0 SWAY TRANSVERSE FORCE

2 mean

1x10^

C\J

OJ _1_

YAW

YAW MOMENT

r\

u
tn

^110 tf.m
\

1x10

\ 3mean =\200 t f . m \ \
t/)

0.25

0 C O in rad.sec. -1

0.25

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-

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. 158

WITHOUT WITH SURGE

u 1 0
04

WAVE-FEED-FORWARD WAVE -FEED-FORWARD LONGITUDINAL FORCE

CM

10

1x10'

E
li-

x uo

imean-41 tf

3 8 t t

SWAY

TRANSVERSE FORCE

CM

u tu tn

10

1 x 1 0 5 _

if)

OO X

10 O YAW 0 u
4J

if"

YAW MOMENT M3 m e a n s 2 3 6 0 t f ' m

QJ (0 CM O) OJ TD

2x10

tn o CM b
CM

(0 X

y-

I/)

2450tf.m\ 0 0 0.25 O C O in rad.sec. - 1 0.25 0

ro

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

irreg4.9m. 159

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

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 feed-back system alone could reduce low frequency surge motions to almost minimal values. Adding the wave-feed-forward control signal brought about only minimal changes. The corresponding surge motion 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 and Figure VIII-13. For the test in irregular head waves, shown in Figure VIII-11, the surge motion damping was reduced (see surge motion decay test in Figure VIII-7: 180 ) , showing more clearly the effect of wave-feed-forward on low frequency surge motions. VIII.5.3. .Results of tests in irregular waves and current Tests were carried out with and without wave-feed-forward in bow quartering irregular waves (135 ) with a significant wave height of 4.9 m and a mean period of 10.2 sec. and a stern quarter ing current (45 ) of about 1 knot. The spectra of the low frequency 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 irregular waves are in both cases according to the wave spectrum given in Figure VIII-5. The results given in Figure VIII-14 show that current does not affect the control signal. This is to be expected since the wave-feed-forward signal is determined from wave elevation signals which do not change appreciably for the normal values of the cur rent speeds encountered. From the results it is also seen that the low frequency forces and motions are not appreciably different from the results given in Figure VIII-13, which indicates that the influence of current on the low frequency wave drift forces is, in this case, not great. From the results of without current it appears frequency part of the sway cy yaw and surge motion by forward. 160 the tests in irregular waves with and that it is possible to reduce the low motion by about 70% and the low frequen about 50% through the use of wave-feed

WITHOUT WAVE-FEED-FORWARD WITH WAVE-FEED-FORWARD SURGE LONGITUDINAL FORCE

u
C M

10-

to

1mean="24tf 0

Li. 10

SWAY

TRANSVERSE FORCE

u m
CM

10 ,

>

2mean

= 179tt

-1x10'

C M

CM X

0
u m

X
0

\ \ 177 tt \ \ \ \
1

\ N

to

LL

CM

s
^-

YAW

YAW MOMENT

<M O) 0) T3

5-

- 2x10

C M C M

(0 X

o\

0.25

zxx

3mean=-840ttm / -600tt.m

to

CO

0.25

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 i r t . 161

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 in mind, however, that for the experimental determination of the wave-feed-forward signals only eight wave probes were used- It may be expected that increasing this number will lead to more accurate evaluation of the low frequency component of the contribution due to the relative wave elevation, thereby increasing the accuracy of the control signals. Y_______.^__:2G_=iyi22 In this chapter it has been shown that, as a result of the theory developed in this study, whereby the mean and low frequency wave drift forces on floating objects are determined through di rect integration of all pressure contributions to the second order forces over the wetted part of the hull, expressions are derived which, after numerical evaluation, lead to conclusions regarding the applicability of hitherto unknown methods to improve the accu racy of dynamic positioning of vessels in waves. It is possible to predict on the basis of the results of computations the effectiveness of a wave-feed-forward control sig nal with respect to the degree in which such a signal can compen sate the instantaneous mean and low frequency wave drift forces acting on a vessel. Theoretical computations can be used to deter mine the values of the gain factors which are inherent to the wavefeed-forward method. The results of model tests indicate that, even without prior knowledge of such gain factors, wave-feed-for ward can function effectively.

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_lH2i22 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 t o t and cos tut. Descrigtion_of_the_theory First order wave loads and motions The ship is considered as a rigid body, oscillating sinusoidally about a state of rest, in response to excitation by a longcrested 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 by: ik(x C = CQ e cos a + x sin a) - ioit (A-2)

where: Cn = amplitude of the wave k = wave number = 2TT/A, where X is the wave length a. ~ angle of incidence. The flow field can be characterized by a first order veloc ity potential: *(x1,x2,x3,t) = 4i(x1,x2,x3)e
1U)

(A-3)

The potential function $ can be separated into contributions from all modes of motion and from the incident and diffracted wave fields:
< > j = ito C0(<t>0 + * 7 > " iu Z *j
C

(A

~4)

The incident wave potential is given by: cosh k(x. +c) ik(x. cos a + x 9 sin a) e *n = ~ vT-^3 (A-5) T 0 v cosh kd in which: 2/g . v = c o c = the distance from the origin to the sea bed d = water depth a = angle of incidence of the waves. The cases j = 1,....,6 correspond to the potentials due to the mo 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 the Laplace equation which satisfy the linearized free surface condition and the boundary conditions on the sea floor, on the body's surface and at infinity. The potential function 4 > can be represented by a continuous distribution of single sources on the boundary surface S:
i cn

> _ 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

for j = 1, 2 , - . . . , 7 (A-6) where: Y-(x,x,x-.,a,a,a 3 ) = the Green's function of a source, singular in a 2 , a 2 , a 3 a., a-,, a-.
a

= the vector describing S = the complex source strength.

j(ai'a2'a3]

For the Green's function a function is chosen which satis fies the Laplace equation and the boundary conditions on the sea bottom, in the free surface and at infinity. This function is giv en by (see Wehausen and Laitone [A-2]): v 1 = - + r r + PV .

2 ( + v)e -?d .cosh (a3 + c) .cosh (x3 + c)

/

sinh d - v cosh d 2TT (k 2

J 0 (R)d +

+ i

2 - v ) .cosh k (a, + c) .cosh k (x + c) = ~ J n (kR) u k^d - v d + v (A-7)

in which: = \/ ( x l- a l 1 R
=

) 2 + (x2 - a 2 ) 2 + (x3 - a 3 ) 2
2

\/ ( x l- a l )

+ (x2 ~ a 2 ) 2 + (x3 +2c + a 3 ) 2 + U2-a2)

2

. . . (A-8)

^(x.-a^

John [A-33 has derived the following series for y, the analogue of (A-7): - k2 Y = 2 / T 9 v = cosh k(a k d - v d + v
2

which is

+ c) .cosh k(x-, + c) .
J

i aa

{YQ(kR)

-i

<kR)>

i = l dy.

4(y 2 l

+V2)
2

+ dv

- v

cos y . ( x 3 + c ) . c o s

y^ (a., + c) .K (y . R) (A-9)

w h e r e y.

are the p o s i t i v e + v = 0

solutions

of: (A-10)

Vi. t a n ( y . d )

Although these two representations are equivalent, one of the two may have preference for numerical computations depending on the values of the variables. In general, equation (A-9) is the most convenient representation for calculations. When R = 0 the value of K becomes infinite; therefore equation (A-7) must be used when R is small or zero. The unknown source strength function o must be determined such that the boundary condition on the body's surface S is ful filled. Due to the linearization this boundary condition is ap plied to the surface in its equilibrium position S 0
n

-^ j (x 1 ,x 2 ,x 3 ) + 47 // aj(a1,a2,a3) . S 0 " 3n Y j' x l' x 2' x 3' a l' a 2' a 3^ dS

3 for

'

'6

' IF

*0

^r j = 7
(A-ll)

n through nfi are the generalized direction cosines on S , defined by:

169

n1 = cos (n,x,) n 2 = cos (n,x2) n~ = cos (n,x-J n. = x , n > - x,n4 2 3 3 2 n5 = x 3 n i - X l n 3 n 6 = xxn2 - x2nx To solve equation (A-6) numerically the surface S is sub divided into a number of finite, plane elements on which the source strength is constant. The boundary condition is applied in one con trol point on each element being the centre of the element. The integral equation (A-6) then reduces to a set of algebraic equa tions in the unknown source strengths. In general, the Green's function y may be computed with sufficient accuracy as if the source strength is concentrated in the centre (control point) of each element. When, however, the influence of an element on its own control point is evaluated y has a singularity of the type 1/r, which can be removed by spreading the source uniformly over the panel. When the influence of a panel on a control point, which is at a close distance of this panel and not lying in the same plane, is considered the source is spread uniformly and integrated 3 numerically to obtain its contribution to <\> or 3n* After solving the equations for the source strengths the first,'order potential function is known. The pressure on the sur face S can then be found from Bernoulli's theorem. The linearized hydrodynamic pressure is given by: p(xn ,x~ ,x v t) = -p -^r = 'l'~2'~3 3t
= {pio 2 2 C 0 (<J>0 + tj>?) + po) E ax
J J

(A-12)

C.)e

-iut (A-13)

j=l

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

X k pa)2 C 0 e " l u t // ( 0 + * 7 ) n k dS S 0 (A14) The oscillating hydrodynamic forces (k = 1, 2, 3) and moments (k 4, 5, 6) in the k F = pto2 direction are: Z c e'i(iit i =i i J : // *.1 n s0 s> dS (A15) According to common practice the hydrodynamic forces are repre sented by means of added mass and damping coefficients: a k . = p Re {jj ^
S

n k dS}

A ( 16)

0 A ( 17)

b k j = poo Im {// $ . . n k dS}

S

where: a, . = the added mass coefficient in the kmode due to motion in the jmode b, . = the damping coefficient in the kmode due to motion in the dam jmode. Finally, the motion response to first order excitation is computed by means of the well known equations of motion in the frequency domain: 2 { t o (M, t o t + e .) + K . + a, . ) .sin(tot + e .) + K b, . . to.cos ( j! 3 *J J 3 3
6

+ 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

M . is an inertia matrix. Since the origin of the system of axes coincides with the centre of gravity of the ship in its rest posi tion it is found that:
m 0 0 0 m 0 0 0 0 0 0 m 0 0 0 0 0 0 0
C

0 0 0

kj

0 0 0

-J46
0 I,. :A-I9)

-I 64

whe re: m I, th = moment of inertia in the k " mode = mass of the ship

I, . = product of inertia.

172

APPENDIX B - CROSS-BI-SPECTRAL ANALYSIS Introduction The computer program which was used to analyze the inputoutput relationship between the waves (input) and the low frequency mooring or restraining forces (output) is based on the cross-bispectral 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 intention of this appendix is to give a "feel" for the processes involved, rather than to give a condensed version of the specialist's point of view as given by Dalzell. We assume that the input (wave elevation) can be written as follows: (1) Z in which: t o . e. Zt N = frequency in rad./sec. = random phase, uniformly distributed from 0 - 2TT = amplitude of component with frequency t o . = time = a large number.
} N

(t)

Z
i=l

(1) z; } .sin(w.t + e. )
i
J -

(B-l)

The foregoing expression represents a zero-mean, normally distri buted, stationary random signal. The square of the wave elevation is: Z
K

(ll2
}

(t) = { Z i=l
N

... 7 Ci .cos(w.t + E . ) T
x x

N (1) (1) Z Z cz\ .costto.t +e. ) .cos(io.t +e.) x J i=l j=l i i 3 3

= J l jl ^ ^ C ^ c o s U t o . t o . J t + ( S i j ) ) +
N N (1) (1) + Z Z J f c J t\ .cos{(to + to.)t + (e. + E . ) } J x J i=l j=l J

(B2) The low frequency part of the square of the wave elevation is:
C

2 low

(t) =

%ci1,C_{1,.cos{(ioi <o.)t + J J i=l j=l

(E.E,)}
J

(B3) It is assumed that the output (wave drift force or moment) contains only low frequencies and is closely related to the low frequency part of the square of the wave elevation: F (2) (t) = N Z N Z C.(1)c(1)P.,.cos{(to. to,)t + (E. E , ) } +
i j ij J J J

i=l

j=i

+ Z

i=l j = i

Z\

(l) } (1) 1J
J

Z\

Q.,.sin{(to. to.)t + (e. .)}

LJ

J j

J J

(B4) in which P.. and Q. are inphase and outof phase quadratic trans fer functions dependent on the frequencies t o . and to.. The problem is to determine P.. and Q.. for arbitrary val
J-J /jv ij

ues of to. (t) and the output x and to. given that the input ? ^ known as time records. The ouput is a signal with F (2) (t) are only low frequency oscillatory components. From equation (B4) it can be shown that the time record of any component of F (2) (t) with chosen frequency Aw is the sum of contributions from components of which the difference frequencies are equal to the chosen frequency Aio:

(2) *' to

(t) =

._.

N Z

( l' ) CC.-(i 'D P - . - ; - c o s { A O t + (E. - - ) } +

i J

J-3

J - J

N + Z

c | 1 } C-1)Qi--sin{Atot +

(^-Ej)) (B-5)

i n which i and j a r e chosen such

to.

that: (B-6)

to. = Ato

Equation (B-5) becomes: r N F< 2} ( t) = [ Z A t o cf^C^^Pij.COStE.-Ej) +

+ Qx< .sintE^ -Ej) } cos A t o t+ f Z L i=l C^d^Q^.cos.E.-E.) J J J

- P^.sin.E.-E.)} sin Atot (B-7) 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 Q.. values. It is not possible to determine the value of individ ual P..*s or Q..'s from the signal. This is a result of the fact that F v ' (t) is a double summation. The foregoing indicates that in order to determine the quadratic transfer functions for required combinations of t o . and t o . it is necessary to find a way to extract from the time record of the output information which is essentially in the form of a single summation of oscillatory components with amplitudes which are in themselves not a summation of components. For instance, it is possible to generate a time signal U(t) of the following type:

175

Utt) =

N I U ..cos{ {ta -to. )t + (^-e. )} i=l 1 3 J J

(B-8)

where: ta. + t o-= t o , = some chosen fixed frequency EX , ex = random phases of i and j frequency components of wave elevations of equation (B-l). If the signal is cross-correlated with the output F tion (B-4) it follows that: (2) (t) of equa

i +T/2 e>\ Rpn(T) = Iim i / U(t) .F u '(t +x)dt (B-9) tU l T-* -T/2 (2) By this operation only those components of F (t) will be identi fied Which correspond with the components of equation (B-8). Con tributions from all other components will disappear since they are not correlated to the components of U(t). The actual computation of the transfer functions starts with a transformation of the input signal C(t) according to:
+ Tm

/ cos tokx.C(1) (t - T ) - C ( 1 ) (t +t)dT (B-10) T ~ m This represents the Fourier transformation of the product:
.
( 1 )

(_-.)..

( 1 ,

(t

.,

in T cok T

which: = time shift = some chosen fixed frequency = maximum time shift (maximum number of lags multiplied by sampling interval).

It will be clear that the output is a function of the chosen fre quency tok and time t. Substitution of the expression (B-l) for the wave elevation gives the following result for the inner product:

176

C(l)(tT).CU)(t+t) = (1) i; Z Z . G (1) .COS{li). (t T ) + . }.COS{(0. (t + T ) + E . } 3 J 1 J i=l i=i

N N
N N

hzl(D.d) t\ .cos{ (to +to.)x + {ta to. )t +

* J J

i=l j = l

+ ( _ E . E X ) } + high frequency components (Bll)

M u l t i p l i c a t i o n by cos O XT g i v e s :
COS ( O
k

T.CU) (t -T) . C

U )

Ct + T )

N Z
. i=l

N Z
. j=l

hZ(1)
i

c i 1 } .cos[{io,
J
J

(to.

+CO .)}T

+
J

(to. - t o . ) t
J - J
J

"

J-J

(. E X ) ] + high frequency components (B12)

Disregarding the high frequency components expression (B10) be comes : (1) (1) Z Z hz CU, .cos{((o. u>.)t + x D J i=l j=l m
+ ...)} 1 J / COS{(0 ((0. + tO. ) }T.dT + K 1 J
N N

" m
N N (1) (1) Z Z ^C C. sin{ (to. to.)t + J J x i=l j=l

+ (ex ) }
J

__

m / sin {to, ( t o . + to.) } T .dT K m

J J

(B13)

Inspection of this expression shows that the contributions which arise for the cases that t o . = to. + to. will dominate, so that the K 1 J outcome is of the following type:

2x

(l).(D J x ^i C^i f k ) -co8{( U i - U j ( i # k ) )t + U - j ( i f l c ))}

(B-14)

The outcome is a signal which contains only those difference

fre

quency components of the wave elevation which have as sum frequency: tok = to. to. (B-15)

which is what was needed in order to be able to identify correspond(2) ing components in the output F v ' (t). The above expression appears to increase as x increases. It must be remembered, however, that m the processes involved are stochastic. The output of the above ex pression is finite. 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 pression (B-14) is given for three values of the sum frequency to,. 10

w1/ 3 "5.5m

T
u
CM.
ffl

=12 sec.

( / )

~
0.5

1 . 0

U) in rad.sec-*1 Fig. B-l Wave spectrum.

178

+1

c o s cok x ( t - x ) . (t*x) dx

-x m
0)u = 0.8 rad.secr 1

-15 m sec. L

-1 t u k = 1.2 rad.sec;

15 m 2 sec. r-

-15 m 2 sec. WAVE (t)

-2 m L

TIME in sec.

100

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

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 t o , val ues result in signals with much less low frequency components. This is explained by examination of the wave spectrum given in Figure 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 rence of wave groups, has little energy for the highest and lowest 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 t o , . This operation results in a signal which for arbitrary sum frequency t o , supplies the time record containing the low frequency part of the square of the wave elevation (only for those difference frequencies which have t o , 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 t o . and t o . which have c o . 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 bispectral terminology, the crossbispectrum of input, input and output. The crossbispectrum is valid for the chosen sum frequen cy ta, and contains information for the range of difference frequen cies from zero upwards. Any chosen difference frequencies contain information on the transfer function P.. and Q.. for unique values of t o . and to.. The values of the transfer functions are finally de termined from the following type of expression:
GU

i' u j s ^ v i . . )
h
1

(B 16)

where: C F (toi,io.) = crossbispectrum of inputinputoutput S ((D.) G(toi,to.) 180 = wave spectrum (scalar spectrum) = quadratic transfer function of output.

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 crosscorrelation 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. 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. 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. Bhattacharyya, Rameswhar: "Dynamics of marine vehicles", Wiley Series on Ocean Engineering, 1978. 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. Sjouke, J. and Lagers, G.: "Development of dynamic posi tioning", Paper No. 1498, O.T.C, Houston, 1971. Sugiura, M., Nekado, Y. and Matsui, M.: "An experiment of the dynamic positioning system of a semi-submersible plat form", Interocean, 1973. 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. Van Oortmerssen, G.: "The motions of a moored ship in waves", N.S.M.B. Publication No. 510, 1976.

[1-2]

[1-3]

[1-4]

[1-5]

[1-6]

[1-7]

[1-8]

[1-9]

1 flO

[1-10] Aral, S., Nakado, Y. and Takagi, M.: "Study on the motion of a moored vessel among the irregular waves", J.S.N.A. Japan, Vol. 140, 1976.

ft-.

REFERENCES (CHAPTER II) [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

[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. Goodman, T.R.: "Forces on a hovering slender body of revolution submerged under a free surface", Developments in Mechanics, Pergamon Press, New York, 1965. 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. 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.

[11-22]

[11-23]

[11-24]

[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", Schiffstechnik, 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 the first and second order wave forces on bodies oscil lating in regular waves", Second International Conference on Numerical Ship Hydrodynamics, University of California, Berkeley, 1977.

[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.

[11-38] Wahab, R.: "Wave induced motions and drift forces on a floating structure". Report No. 1865, Netherlands Ship Research Centre, TNO, Delft, 1974. [11-39] Huse, E.: "Wave induced mean force of platform in direc tion opposite to wave propagation", Interocean, 1976.

REFERENCES (CHAPTER III) [III-l] Stoker, J.J.: "Water waves", Interscience Publishers Inc., New York, 1957. Van Oortmerssen, G.: "The motions of a moored ship in waves", N.S.M.B. Publication No. 510, 1976. 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.

[III-2]

[III-3]

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. Pinkster, J.A.: "Low frequency phenomena associated with vessels moored at sea". Paper SPE 4837, European Spring Meeting of SPE-AIME, Amsterdam, 1974.

[VII-2]

REFERENCES (CHAPTER VIII) [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. Sjouke, J. and Lagers, G.: "Development of dynamic posi tioning". Paper No. 1498, O.T.C, Houston, 19 71.

[VIII-2]

REFERENCES (APPENDIX A) [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) [B-l] Dalzell, J.F.: "Application of cross-bi-spectral analysis to ship resistance in waves", Report SIT-DL-72-1606, David son Laboratory, Stevens Institute of Technology, 1972. [B-2] Dalzell, J.F.: "Some additional studies of the application of cross-bi-spectral analysis to ship resistance in waves", Report SIT-DL-72-16 41, Davidson Laboratory, Stevens Insti tute of Technology, 19 72. [B-3] Dalzell, J.F.: "The applicability of the functional poly nomial input-output model to ship resistance in waves". Report SIT-DL-75-1794, Davidson Laboratory, Stevens Insti tute of Technology, 1975.

NOMENCLATURE L f f i x ( 0 J ' ( 1 ) ' ( 2 ) ' ( 3 ) affix denotes whether a quantity is of first, second, third order, etc. A. . , B. . , C . 13' 13' 13 C.t) coefficients dependent on wave frequencies to. and t o. pressure contribution independent of the co ordinates of the point under consideration force vector relative to the GX'X'Xl sys tem of axes l ' F2' F 3 = (2) =(2) =(2 F l ' F2 ' F3
F

components of the force vector mean values of the second order force compo nents centre of gravity of a body

(2) Gxrx2x3

complex quadratic transfer function body axes with origin in centre of gravity G, x.axis positive in forward direction, xaxis 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

GX 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 I 46 0 I, " A 6 and Gx 3 axes respectively

4'

V h

mass moments of inertia of the body about the Gx , Gx

46

product of inertia length of a vessel or cylinder

mass matrix m 0 0 0 m 0 0 0 m

M M., M-, M 3 N

moment vector relative to the G-X'-X'-X' sys tem of axes components of the moment vector outward pointing normal vector of a point on the surface of a body relative to the 0-X X

2~X3

or

G_X

l~X2~X3

svstems

of

axes

*__ T. . 13

Qij

components of the quadratic transfer function dependent on t o . and to. amplitude of the quadratic transfer function dependent on t o . and to. wave spectrum

S
s

total wetted surface of the hull mean wetted surface of the hull surface element of S axes static or mean waterline on the hull of a body position vector relative to the fixed system of axes components of X position of the centre of gravity of the body relative to the fixed system of axes vector operator or S

dS V WL X

velocity vector relative to fixed system of

l'

2'

V, V

volume of the mean submerged part of a body Laplace operator Principle Value area of waterline of a sphere position vector relative to the G-X!-X'-X' system of axes

v2
PV A X*

a d e g (2) g h

radius of sphere radius of cylinder 2.718 (constant of natural logarithm) constant of gravity second order impuls response function water depth

i
k d m rrx n n., n ? , n..
n

/=T
wave number = 2TT/A line element of the waterline mass of body in vacuum area of wave spectrum outward pointing normal vector of a point on the body relative to the body axes G-x 1 ~x - x 3 components of n coefficients depending on to. and t o pressure time amplitude of an oscillatory quantity position vector of a point on the hull of the body relative to the body axes components of x chapter VI: angle of mooring line appendix A: angle of incidence of waves; 0 represents stern waves, 90 represents head waves represents waves from starboard beam and 180

ii ' f ii

p t ua x x., x ? , x-. a

a 6 F E_.

angular motion vector mooring line angle a small quantity << 1 random phase angle of i frequency component

i a ft

phase angle between wave and some oscillatory quantity u

Z C a Z C. Z , /j T A A y

wave elevation wave amplitude of a regular wave relative wave elevation amplitude of i frequency component

significant wave height mean wave period wave length nondimensional frequency frequency of low frequency part of the second order forces

p x $ <J> to to. a y a.., a k*il' ^33 x x. , x J , x, , xJ: , x

mass density of water time lag velocity potential dependent on coordinates and time t part of velocity potential independent of time wave frequency
th r 1 frequency component

source strength Green's function added mass fpr surge and heave motions damping for surge and heave motions first order surge, sway and heave motions first order roll, pitch and yaw motions

SUMMARY In this thesis the mean and low frequency second order wave drift forces on bodies moored or stationed in waves are analyzed. Expressions are derived for the second order forces based on direct integration of pressure acting on the wetted part of the body. It is shown that the second order forces in irregular waves may be determined from knowledge of the mean forces in regular waves and the low frequency forces in regular wave groups. In order to calculate the mean and low frequency forces on bodies of arbitrary shape use is made of a three-dimensional linear potential theory computer program. The form of the body is approx imated by a distribution of plane facet elements representing a source distribution. For a hemisphere the results of computations of the mean second order horizontal forces in regular waves are compared with analytical results. This comparison demonstrates the accuracy of the computations and the equivalence of the expressions for the second order forces developed in this thesis with respect to an already existing expression based on momentum and energy con siderations. In order to demonstrate the validity of the present theory with respect to realistic hull forms, results of computations of the mean second order forces in regular waves are compared with results of experiments on a tanker, a semi-submersible, a rectan gular barge arid a submerged horizontal cylinder. For the first three hull forms the mean horizontal forces are compared. For the submerged cylinder the mean vertical forces are compared. The correlation found between results of computations and experiments confirms the general applicability of the theory for predicting the second order forces on a wide range of hull 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 crossbi-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 wavefeed-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 lichaam afgemeerd of gepositioneerd in golven. Uitgaande van integra tie van drukken over het natte oppervlak van een lichaam worden uitdrukkingen 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 potentiaal 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 beschouwingen. Met het doel de geldigheid aan te tonen van de in dit proef schrift gegeven theorie met betrekking tot meer realistische rompvormen worden resultaten van berekeningen van de gemiddeide tweede orde golfkrachten in regelmatige golven vergeleken met experimenteel 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. Een nadere analyse van de komponenten van de gemiddeide tweede orde krachten toont aan dat voor drijvende konstrukties de horizontale krachten gedomineerd worden door een bijdrage die afhankelijk is van de relatieve golfhoogte ter plaatse van de water lijn van de konstruktie. Voor de tanker en de semi-submersible worden de resultaten van berekeningen van de laag frekwente driftkracht in regelmatige golfgroepen vergeleken met resultaten verkregen uit modelproeven in regelmatige golfgroepen en onregelmatige golven. De meetresul taten verkregen uit proeven in onregelmatige golven zijn geanaliseerd door middel van kruis-bi-spektrale methoden. Uit de vergelijking 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

ACKNOWLEDGEMENT I am grateful to the Board of Directors of the Netherlands Ship Model Basin for permission to publish this research in the form of a thesis. I am indebted to the participants in the industry supported Wave Drift Research Program for their kind permission to make use of part of the model test data obtained from Phase I of this pro gram. My sincere thanks to the members of the Ocean Engineering Department for their patience and co-operation which contributed much to the realization of this work.