THE DRIFT FORCES AND. .NT ON SHIPS IN WAVES
by J.. N. Newman
David Taylor Model Basin Washington,. 'D, C.
June .965
ARCHIEF
y.
SchpLbouwbnde
Technische Hogeschool
ABSTRACT
The second-order steady: horizontal forces and vertical moment are
derived for a freely floating ship in regularwaves. The fluid is assumed to be ideal and the motion is linearized Momentum relations are used to
derive general results for an arbitrary ship or other body., in terms of the Kochin function, or far-field velocity potential of the body. Explicit
results are derived for slender ships.,. based .upon the ássumptions of.. slender body theory. Cbmputations. are made for a series 60 hull and are comparedwith experiments.. The analysis of the vertical moment permits
the prediction of stable heading angles in oblique waves, .and it .is shown
that unless the waves, are very short., the ship will be stable only in beam waves.
F' X F g H(e) n
r
t Cartesian coordinatesHeading angle of ship to waves (head sea =180°) Free surface elevation
Incident wave amplitude
:Shjp beam
Longitudinal force
Lateral force
Gravitational accéleration
Kochin function, defined by equation (20)
Waterline moments (n 0, 1, 2,)
Angular momentum vector
Wave number, K. = w2 1g
Ship, length
Linear .- momentum vector.
Vertical (yaw) moment
Unit normal out of. fluid
Flui4 , pressure
Polar radiús, :
'Position. vector'
Time ' '
Norma1. velocity, of' control surface "Fluid velOcity vector
Heave amplitude Pitch amplitude
e Polar angle, .e =.tan (y/x)
X Wavelength
p Fluid. density
+ Velocity potential
INTRODUCTION
When a ship is floating in the presence of an incident wave system,
pressure forces and thoments will be exerted on the ship and will influence its behavior. These forces and moments will include not only the conven-tional unsteady exciting forces, which give rise to the oscillations of the ship in waves, but also higher-order steady forces due to various non-linear effects., These latter forces are generally.too small to influence
the oscillatory thotions of the ship but they can nevertheless be important
in certain circumstances, particularly in considering those degrees of freedom of the ship which do not have a static restoring force (surge,
sway, and yaw). If the ship has a substantial forward speed then
effective restoring forces exist in these modes, through the actions of the propeller, leeway angle, and rudder, respectively, but at zero or very low
speeds large excursions will result from even a small wave-induced force
if it is steady over a long interval of time. We shall refer to this situation loosely as Hdriftingtl, and to the drifting forces and moment as
those second-order steady components of the wave-induced force in the horizontal plane and of the corresponding vertical (yaw) moment.
Clearly the drift velocity of the ship will be governed by the mag-nitude and direction of the drift force, while the drift moment will
govern the heading angle of the ship with respect to the waves. The prediction
of the transverse drift force and the drift moment are of importance, for example, in the design of bow-thrusters [1], if it is desired to maintain the mean position and heading of an unrestrained ship at zero speed in
predicting the stableheadtug angle of.the ship.
A theoretical analysis of drift forces has been .provided by Maruo [2], but the corresponding moment was not. included
a4
no.coputations were ma4e On the other hand, Suyehiro [31 analysed the yaw moment.in order. to predict t!e stable heading angle of a drifting ship, .but in that analysishydrÓ-dynamic effects were. neglected entirely in favor of gyrostatic moments
From the experimental viewpOint an .thvestiga.tion. was undertaken by; Spe,ns and Lalangas [4],inwhich meaSurements.were made of.the transverse drift .force and the drift momént on aSeries 60 model atone heading angle
In,the present...paper.we shall.derive general analy4cal equa.tions for. the drift forces and moment on an atbitrary body, and specific formulae for a slender ship. Calcúlations will be presented to illustrate.the
application of these results to a Series60 hull form, and comparison will be made with.the experiments of Spens and lalangas. This cOmparison of
théGry -ad-perimeñt :is only fair, and additional .experimental . conf irma
tionis needed.
Our analysis makes use Of the usual techniques of. linearized water
wave, problems, assuming the f luid to beincoesstble aúd inviscid and the wave amplitudes to be smalicompared to the wavelength and
character-istic ship dimensions. MOmentum relations are utilizéd to express the drift forces and moment in terms of. the farfield disturbance of the ship, thus avotding the necessity to determine the secondorder.disturbance in
We write
wh e r e
THE DRIFT FORCES AND MOMENT FOR' ARBITRARY. BODIES
vZ
I)
JReL
' V(xJy 2)is the (linearized) potential of. the incident.wave.system'and
+B is the disturbance due to the presence of. the body. Here g is the acceleration of gravity, K =
uP/g
is the wave number, A is the wave amplitude, andthe.angle ofwave incidence relative to the x-axis.
The potential +B is determined as the solution of.the boundary
condition
(1)
(2)
(3)
We consider, a floating body with no forward velocity, on the surface of an ideal incompressible fluid andin the, presence Of an incoming, incident
plane prorèssive.wave system. Let (x,.y, z)be Cartesian coordinates with
z = O the
plane.
of.the undisturbed free surface and the z-axis positiveupwards. The fluid motion is assumed to be harmonic in time with the
frequency of. encounter w/2rr -J
The fluid vefocity canbe represented by. the gradient of a velocity
--dt(7»,
-s+sps.t.s,oQ
7Ccc(Ili) \)
P( co:(ny,,/
(6)
±I+u
(4)
on.the ships surface,.where is the .normal.velòcity of.thatsurface, .and of the boundary condition
--o
o,O
(5)
Equation (5). is the. linearized free surface condition (cf.Wehausen and Laitone [5]).
In addition. +B.st.5a.sfy
equation throughoutthe fluid domain,.a radiation condition of outgoing waves at infinity, and
yield vanishing velocity, as z -. -.
general the solution of this boundary value.problèrn is unknown and
we must restrict ourselves to. simplified bodiessuch as-thin or slender
ships in order to find
+6 explicitly. - Before doing so, however, we can obtain certain general results for arbitrary bodies, and óur.task is
simplified considerably by employing the. far-fiéld asymptotic form of. the
potential .+.valid.at.large distances from the body.
In order to..relate the mean force, and moment acting on the body to
the far-field potential w consider the rate of change of. linear and angular. momentum within the fluid domain bounded by the shipVs wetted surface, free.surface, SFS , and a control-surface at.inf.inity,
As deriyed in Appendix 1.the rate of change of horizontal.linear momentum is given by .
and the rate of change of the vertical component of angular momentum is
dt 2
Here p is the total. flUid pressure, p is the density, eos(n, x) and cosn,, y) are the direction cosines with the direction of the unit.. vectoi'
taken as positive out .of the closed surface
5B +
+ S , isthe normal velocity of the surface., Ir is the position vector, and the
subscripts x, y, z denote componènts of vectors, 0m the body surface
SB and' the free surface SF thé normal velocities of the body and fluid
(p(rxnp(irxv)(vuj]ìs
' (7)are equal','or
V - U
=.0 We.shall take the control surface S to ben
.n'
' ' 'fixed, so on tl± surface U= O The contribution fromthe pressure t.er ofl 'S is just-the desired force or moment:
cg
ot)
¿Ç
c.oi(y)
C/S
,i.=ffpo1rxc1
Finally, on 'th'e free surface 'the pressure v4nishës. Thus we obtain the föllowing results for the above forces:
-(8)
(9)
Next we také time averages of the above equations, to ôbtain the mean
forces and moment. There is no contribution from the. last terms since the -motion is periodic arid there can be no. net increase of momentu.in the
control. volume from one cycle to another Thus
(Ir)(n,)
p(iV)
Vj0Lc
-
IS
[f
-\JV,
j 0Ls-
f1x)
VV]
Nèwman [6]).-- We now take the control surface. S
-. - .
-R?
-
ff
(i, j*f
V 7
Equations (l+. -.1.5) can also e derived from conservation, of energy. (cf.
to consist of a circular cylinder (.12)
(-13)
of large tad-ius about.the z-axi and extending from the free surface. down
to z.=-
(The.contribution from.a horizontal..closure Of canbe neglected sinëe the fluid veloàityis required to vanih .there)
(14)
('15)
We take (R, 8,.z) as polar coordinates, with x=R cos e and
y=R,sin 9
. After.making the appropriate substitutions..in (14.-16) wethen obtain, the results
SiMs
\,v
e1Gh
The above -equations are exact since no. linearization has. been employed,
.but.to.proceed further.we must assume that the wave height is small., and
we shall only retain contributions to the forces and moment which are of
se.cond. order in the incident wave amplitude.
The pressüre is giien .by Bernoulli's equation
8 -V9
ç,
e)]RcIe
(17)(19)
and the velocity components are
t,
(.21)
V
1
(22)+
The far-field potential. 'is of the form
where
expfr[
I,? c.s,'(2-eJ]
and where
*
and. thus, it follows that
/7'('ir#9)
¿'p (c *('kfri7r/4) .*
(23)
(25)
Here H(e) is the Kochin function and the integral is over the submerged
surface of the ship. Equation (25).is derived iñ Appendix IL
Since in the far-field both and +B' have exponential..dependance
on thecoordinate. z ,it.is convenient first.td càrry out the..z-integration
d2
11Loe1 j-o()
-
Tvi
J2o
*
The infinite lower limit does not contribute to the hydrostatic pressure
-.afterintegrating with respect to e
H(e=
-
d
(26)in.equat.ions (17-19).. For-this purpose we 'note that the upper limit of
-the z-íritegration is the free surface
-'
-
Z:
Thus,.rieglecting.terms. of order
?
8KJ0
LLRae
r !#
-j
Ld,e
e r2rr
±_ ì±±*.
¿ g1-J0L L,
-k ++ Re
cJ L ¿,e
RI&
RK°
R6
where we have used the fact that
+
-s
an4 asterisks denote.the complex conjugate.
Finally it is necessary to substitute the far-field potential (23 25) in (27 -29). Since(27.- 29) are qudratic in the. potential
(.29)
and its derivatives, thé restílt. will ïnvole terms which te quadtatic in .the potentials -and ..sep4rately plus cross-terms. involving products
of and . The contribution from alone must vanish since there
can be no forces or. moment associated with. the.undisturbed .iñcident wave
and
performing
the necessary operatiOns. Thus, we need only consider the cross terms and the terms. which àre quadratic inintegrals are
G) j co.c
9 c/
f
IC9ir
Lf)A
R
(cosfl#c.os9)
H(n9)
exp
fíkR 1'-
C-ii&- e)j-tí7r/j 01e
Q (R
'/aJ
(30)J/Hfrt9))2
,9 19
f4I4
(/<)¼.
RJ«(g
) H(-e)
exp'fi<RtI-o(fl-o)]*L1rÌdO
O(R)
The moment integral must be.treated more carefully. 'First let us 'consider
the cóntribution from 4uadratic terms in
JO
-
Rej
i--'2:o
j
rQir
í:_
.ij
ye) d
9ir
owhere a prime denotes the derivative with respect to e Next we consider
the cross terms involving
4B and RfL
aì
-df Rer2' #:
jo L ç2ir-
(ì.,+ coiee) H'(r*9)
xp{/RL;-
cofl9)]+7ri/í S9
Combining these results it follows that
,y
#r 8-
]1ede
;;
/1 (9)fr/ 19) S
-
2Tr 1A)
.rL (/-)1
H'(ir*)
exp{keL/-ô',,4-eJJ
1r/
d9
o(A7-'2)
The integrals in (3Q - 32) involving R can be evaluated by the method of stationary phase (cf. Stoker, [7]). Thus
e(e) ex? f-d*-e cs(/' 9))
(2ff
\YZ=
)
[e
* ff/4
+ e
--
o((KRy)
n)J
(32)Applying this asymptotic approximation to (30 32) we.obtain the forma.la.e
=
/H9)j c9d&
-
A c&
Ir»
H(rifl.)
(33)
'
f2ir
.2./H(6))
cì9
pA
S/
Ih H (7r*f2)
I
F(9)W'(6) c1
where Hv(rt + ) is to be interpreted as -iLi.«Je
= 1-r + ,These are the desired results for the drift.forces and moment on an
arbitrary body.
The force eqt.lations (33 - .34) can be put.in a
tore
compact.form,ollowingMaruo [2], if we assume that there is no net work being done on
the bödy by external forces. Then from conservation of energy, (cf..
Stoker [7]) .
(34)
Re /7"(7rfl)
(35)
and direct reduction of this integral, by the same methods employed above, leads to the relation
Im
i/i-,'('e))
1&
4,rA)
'4
Substituting (36) in (33 - 34) it follows that
pk2
ir=
;; [
/H) (c
'9
/
7tr
JW(&J)
Let uß now âpply the results derived, in the preceeding sect.on. to a
slendér.' ship. The Kòchin fûnction is
and the potential satisfies the boundary condition
-
ç,
'(40)where 'u is thé. normal velocity, on the body. For asleñder.'body y. n
z .are o(e) and moreover
«
(cf. Néwman and Tuck [8]). Thus inthe limit -the Kochin function is
!
)s
= LIQ()e1
(41)
where Q =
- dl
is the flux through the. transverse section Cf the six amplitudes of the ship's motion are C.e iwt (j = 1, .6
for surge,. sway, heave, roll, .pitch,.and yaw,.respectively) it.follows
[8] that
THE FORCESAND'MOMENT''ON4.SLENDER.BQDY
g
xp
fr(
tiyrt')]
d
=
and thus that
H)
-1
(42)'Here B(x) is the beam of the ships waterline. Since , B(x) =,O(e)
H(9) = O() and the error in (43) is O(2 log. s) From (43)
I HfrJ
- -
Re fß(r) (-x)
/Çwce.fl
ax
j
f5(x) (-x-) &&
,]
z'(r-)&
+o(
(44)but. fróm (36) 'the left-hand-side of (44) must be Q(s) , so that the'
integralj. in
(44)
must vanish, to order e , unless work is done on the shipby external forces. This conclusion can be checked directly, for
uñrestrained motions in waves, where the first order equations of motion are [Ö]
f 2x (xr)dx
(46)The drift forces
T
and can be found conveniently, from ('37) ànd(38') if, we write
.Sincethis-nan even functionof e it follows that
3r
(C*s
f
g()[ik
rJ
añd, similarlyk
3fe()
[
-,r+7
(47)4
k'f fß()[I
, ,fe
*
/
J
e'
CO9
k3
fe
f (J [e
(i-j
(48) lií3esei
oC
tL 4-
k;4,
We note the similarity of these integrals with the damping coefficients for
a slender ship. Jndeed if = 900 the integrals in (48) are identtcal to. those occurring.in the damping coefficients for pitch and heave in
ca1mwater. Thus it follows that f or.beam wàves
f&'t [A -tC4S(2
'''j d'
i.
g3
-
2/(i
from the fact that in beam waves a .slènder shi,' follows the orbital motion of the wave.
We turn finally to the drift.momentacting on a slender ship. For (35) we. need,, from (43)
-
weotfl]alx
(/oJ.
1)k) OE35.
Refx&x(-)e'
J
(49)where = heave dmping coefficient
B35 = pitch-heave cross-coupling damping coefficient
= pitch damping coefficient
However for free motion n beam waves, = iA and
=0 from the
solution of (45 - 46), and thus (49) vanishes, Physically this results
(50)
In general. this intregral is non-zero, unless the. ship is restrained. Thus
there is a çontribution. to the d.tift moment for a slender ship..of the same
order as the beam, or
0()
::
[ß
.r, - 1;
) e' '
where the constants I are the waterline moments
n
f()
Substituting (52 .53). in (51), we obtain, the expressiön
r -zJ
[kr
-
f) co
(i)rO (c
a)
In the simplest case of a ship with.rectangular wateplane,.so that B(x) = B.= constant ,.the constants I are
BL
Il O
= L'/12
(53)
'(54) For free motions in waves we can substitute the first order
displacements and in (51) . From (45 46) it follows that
where L is the 1ength Thus, in this special case,
Here
/
fl /
SL
cflf
L
3
a
-
k2g
,(/-spherical Bessel function.
The.essential features of (55) are.readily determined by. considering the product
j1(x)
j2(x) For small x-g
(x8't')
X3
Fòr large X
ii
)á2()()
j(z) =
(z) is theand j1 j2 = O at each of .the zeros of j1 and j2 The f irst few. such
zéros are, approximately,
x=4.49,5.76, 7.72,9.09, 10.90,12.32,
Thus we can draw the following conclusions regarding the dtift moment
0) =0
for 0<<
For short-er :wave lengths will change sigti.in,quadrant 0-< < 900 A typical graph of the drift moment as a .fuction of
of the form shown in Figure. 1.
Thé requirement that -the ship. be stable is that
and
O <. < 4,49
orIL>
0.70is therefore
M = O
and. the firstz
non-vanishing dèrivative is positive. Thiscondition is met by = 90
at allwavelengths, só that the ship is ahays stable in.bearnseas. For
X/L> 0.70 this is the only stable condition, but for 0.55.< .< 0.70 it.is alsostable in head or followingseas. For shorter wavelengths points of. stability will occur at intermediate angles. This solution may
be.compared with the studyof Suyehiro [3},who.considered the non-linear
gyrostatic coupling between roll. and pitch but ignored all hydrodynamic phenomena. Oi this basis Suyehiro concluded that ships:would be stable only in beam seas provided the frequency of èncounterdid not. satisfy the
inequality
w
while it would be stable only in head waves if
w
<w<w
roll pitch
Here w and w denote the natural frequencies in the two modes.
roll pitch
If the frequecy of encounter is equal to one of these Suyehiro predicts that the ship will be neutrally stable at all headings Experimental
evidence is given by Suyehiro with a small model but the data is dimensional
and the model length is not given. Thus it is conceivable that Suyehiro9s
experiments are in agreement with the present analysis, as illustrated in
Figure 1, and the agreement ;with his own theory is fortuitous This seems
likely to be the case for practical ships which are relatively slender, since the inertial forces considered by Suyehiro will then be of much higher order (in terms of the slenderness.parameter e) than the moment
which we have analysed This suggestion is confirmed by the experiments of Spens and Lalangas [4], who computed the gyroscopic yawing moment considered by Suyehiro, and found it to be very small compared with their experimental observations.
Finally we note the following symmetry properties for an arbitrary slender ship, with respect to heading angle, which are readily derived
from (47 - 54): MZ( =
=
-=(-)
-= y(iT =(rr -) (56) (57) (58)CALCULATIONS FOR THE SERIES 60 MODEL
In order to determine the drift forces and moment for a realistic ship hull, a computer program was written for the I.B.M. 7090 digital computer, based upon equations (47), (48), and (54). Values of the heave and pitch amplitudes were obtained from the first order equations of motion (52, 53). Calculations were then made for the series 60, .60 block coefficient hull which was tested by Spens and Lalangas [4]. The results for the two forces
and the moment are shown in Figures 2 - 4, plotted against the wavelength
ratio X/L , and for various heading angles. Headings are only given in one quadrant, ranging from head to beam waves, since the results for other quadrants can be inferred from the symmetry equations (56 58).
The two forces have been non-dimensionalized by dividing bythe force
pgB2L(KA)2 , while the moment has been non-dimensionalized with respect to pgBLA9 . Effectively, the forces are non-dimensionalized with respect to
the wave slope, while the moment is with respect to the wave height. The reason for this inconsistency lies in the fact that for very short waves
(i.e. large K) the lateral drift force tends to infihity, for all heading angles, and the longitudinal drift force is unbounded for head and
following seas.
For head seas the longitudinal drift force is always negative, as is
evident from equation (37), and it attains a maximum absolute value for wavelengths of one-half to three-quarters of the ship length. For oblique
the shorter wavelengths No experimental data. are available
for
coparison. with this component of the drift force, although in fact this corresponds to the mean increased, wave resistance dùe to the waes, at zero forward speèdThe lateral drift force vanishes for. head and beamwaves. . (Thé latter. is a conequence of the sleñder bódy approximation, together with the fact that foi- beamwaves the ship will follow the orbital motion of.the wave
it-self, and will, therefore not influence the inc.dent.waves.to.the.. same extent
. for pther ..eadingsj The lateral orce is only signflant fòr.very short wavelengths., dropping off. shatply, as the wavelength approaches t:he ship
length. It can b.e shown that the limiting value of.thi.s force coefficient,
as the wavelength tends to zero, is
Experimental values of the lateral drift force are available, from Reference
4,or the heäding angle f l2O°ànd for X/L equal to one-half and one.
Thse are shown by the circles in i.gure 3. Two differént::experimental procedures were èmployed
and
for.the shorter wavelength these ledto
resultswhich differ by.a factorof two, but the higherpair of t.hese experimental points is in fairly good agreement with the theory. At.the..l.onger wavelength the theoretical force has fallen off much more rapidly than the corresponding
experimental points. .
Figure 4 shows the non''dimensionalized drift moment for two heading angles. Other heading angles áre easi.iyinfèrredfroth the fact.that,,f'ro
function only of K'cos Thus, .exceptfor.the
smB
factor, the drift.momentdepends on the heading. angleonly insofar as it affects the apparent wavelength in the direction parallel to the shipUs axis This moment vanishes for -head or beam waves, as in the case of the lateral. force The drift
moment oscillatesabout zero for short wavelengths, before risingto a
maxii and then téndirig toero. Figure 4 also showsthe corresponding experimentalPoints froth ef-érence 4 At the shorterwavelength, .,X/L= the agreement is excellent,but forthe longer.wavelength, X/L= l.,.the theoretical prediction is again substantially lower than the experiment
On the whole the experimeital confirmation in Figures 3 and 4 is nOt very satisfactory Possible reasons for this may rest with either the theory
ok/the experiments. Regarding the heory, it may be noted that the drift forces and moment aré only significant for very. short wavelengths, and it -is
known, [8], that the slender body theory does not give satisfactory predic-tions of the pitch and heave mopredic-tions under these circumstances On the other
hand,.the agreement is best at the shorter wavelength, where the drift force
and moment are much larger, and it seems plausible that at the longer
wave-length,,where the force and moment are much smaller, the experimental data maybe less reliable. It.is hoped that furthér experiments can be carried
out in order- to clarify this situation.
The results of Figure 4 can be used to predict the stable heading angle of:the Series 60 block 60 hull in waves, when it is ùnrestrained,.as
discussed in the preceeding section for the rectangular ship Itwill be
wavelength rat-ioexceeded OJO. From the value .of. - (X/L).sec. at the last
zeroin Figure 4.,.it.is apparent that the corresponding value of X/L for the Series 60 hull is 0.53. Thus for practical purposes in the ocean it would seem that the only stable position of significance is that of barn
seas!
ACKNOWLEDGEMENT
The author is indebted to Miss Evelyn Woolley. for the preparation of
the computér.program and the calculation of the results shown inFigures 2.-4. These computations. were performed on the iB.M. 7090 computer of the
REFERENCES
L. "Some Aspects of Bow-Thruster Design," by George R. Stuntz,Jr., and
Robert. J. Taylor, Trans. SNANE.,-Vol. 72,1964.
2. "The Drift of a Body Floating on Waves,".by Hajime Maruo, Journal of Ship
Research, Volume 4, No. 3,Dec. 1960, pp. 1.- 10,
"Yawing of Ships Caused by Oscillàtion Arnongs. Waves;" by K. .Suyehiro
Trans. Institution of Naval Architects, Volume 62,1920, pp. 93-101.
4. "Measurements of the Mean Lateral Force an4 Yawing Môthent on a Series
60-Model in Oblique Regular Waves," by- Paul G. Spens and Petros A. Lalangas, Davidson Lab. Report 880, June 1962..
.5 "Suface Waves," by J. V. Wehausen and E.V. Laitone, Handbuch der Physik, Volume. 9,. 1960, Springer-Verlag, Germany.
6. "The Damping.and Wave Resistance of a Pitching and Heaving Ship;" Journal
-of Ship Résearch, Volume 3, No. 1, June 1959, pp. 1 - 19.
-.7. "Water Waves" by J. J. Stoker, Interscience, ew York, 1.957. "Current Progress in the: Slender:Body Theory for Ship Motions," by
J. N. Newman and E. 0,;Tück, Fifth Symposium on Naval Hydrodynamics,
Bergen Norway, Sept. 1964.
9. "Theory of Bessel Functions," by G. N. Watson, Cambridge. U. Press, Second
and the continuity equation
(A2)
Here 'p is the fluid.pressure, ,p the density, .g is the gravitatJonl
acceleration and z is the. vertical coor4inae, dire&ted ..Siteards.
The rate of change of. linear.momentum in the volume V is
fT v
APPENDIX I
The Rate f: Change of Linear and Angular Momentu
Here we wish to derive equations (6) and (7).. We consider the fluid
in a volume V , bounded by. a closed, surface S , with unit outward normal
We allow this volume to vary with.time, and denoté the normal
velocity, of the surface S by U
The flu-id is assumed to be inviscid and incompressible. Thus the
velocity vector
V(x, y,
z, t), satisfies Eu1ers equations of motionVd
pß'VS
j+(vv)
y]
(Al)
However, from (A2),
(v/.v)v
= 7.V)VV(wv)= v)V
(A4)where the operatôt V operates on all factors to its right. Thus the volume integral, can be transformed into a .sutface-.integai, using GusS
theorem:
.Th [?'f' +(V/'in)v3
vus
where V
=Vi
. Wenote from (A3) that the contribution from the gravitational potential gz is a vector in. the vertical direction, and 'thus the horizontal components of the rate 'of changé of lineàr'momentumneed notinclude this term.
Next ve perform a similar analysis of the tate of change of.angular
momentum. Thus
dt
L pfffúi
Io
1ff
íx
JoV/) U d S
4f
»[V(r/ri) $
y)
*
flffrV
4V
(-uJ
:(A5)VLc
(A6)Since vxir
= O . Similarly,x(VV\Y (V)(xV
+Vx
.\V(irx\y)
Since %/
x (\Y'V
)r= YxV
= O and thela.stequality follows as in(Ä4). Thus the volume integral can againbe reduced from GaussV theorem:
»JJ
((rx;n(#/p?)
(\Yin(ir\v)7a/S
pffrx \V)
Uds
Howeyer.=
(vv)(irv)
Vx(V.)
-
=(7.\v/)
(wxV)
(47)-
ffr
+ (V
-U
(ir\V)JdS
(A9)The similarity, of (A9) with (A5) i.s readily. apparent. We note that the term involving the gravitattonal ,otentiál may be deleted if we consider the vertical component of (A9), since \Vz is vertical and therefore the term
G
where
APPENDIX II
The Far-Fie1d Asymptotic Expansion of the Velocity Potential
FrOmGreens theorem, the .potential.at any;.point..in-the f:luid.due tö..
the presence of the ship. is
(AlO) s3
where the integration is over, the wetted surface of the.ship, .i is the .unit.normal into the shp,..and denotes the GreenVs function or. source
potential satisf.ying the free surface. condition. From Wehausen and Laitone
[5],.equation (l3..17),.this can be written in.the form
4t-4
(.t.I<O k (1) JA
+
1TLI,./(1).
L
r
(y-)+ (.!.Ç]1"
)J
I/R
L(x2
is the Bessel function of the third kind (Hankl function), and K
We desire a far-field approximation to thevelocity potential, which
will..be obtained by substituting (All) in (AlO) and assuming.that the
parameter KR is large. First.let us consider the asymptotic approximation
of the Green's function itself. Since K(kR) is exponentially, small. for
, the factor (I + K2Y1 can be expanded in.powers of E , and.
since
k,(..kJ) dA
s:
ô(
'
it.. follows that
R
) 1(JR)V
.The.integral over k has, the value (cf. Watson' [.9], 13.21(10))
,(1R)k
and since
r
9- O(r-)
=it. follows from (Al2) that
KeI<
H' (Ks
-
K RY
Finally substituting the well-known asymptotic expansion for the Hankel function (ibid f7.2 (1))
=
2(py'a
e
K() ,.
i(k/?,.r,4)
-(A14)
We are interested in the situation when the field point (x,
y,
z) is in the far-field and the source point(,
, ) is on the ships surface.Thus we set
x= R
cos.;eY = R:
si4e
and note that
[(RG
e
¿(R
)J
.
o(R'
Thus, after substituting (A14) in (AlO) it follows that
(Als)
where
is the Kochin function.
Equation Al5) is the well known far-field asymptotic approximation for
the velocity potential, in terms of. the Kochin fuñction. It is important to
note thàt..the error associated with thisapproximation is samli of.order
R
3/2
.0
0.6
0.4.
0.2
-0.4
-0.6
0O
LEGEND
X
---= 0.63
i'
/
I,
\
%.i
I
g I I I F i. I I I . I I Ii
30°
60°
900
1.200
150°
Figure i - Drift Moment of Rectangular Watorplane as a Function of Heading Angle
(Note /3 = O corresponds to. following waves)
8O°
'('J
4
0.3
0.2
0.I
j
N
N
a'
-0.3
-0.5
o
LEGEND
120°
1500
1650
180°
-0.4
s s s s 'ri
0.5
LO
1.5
X
I
-.
.'
6"y
,
/55517
0035
0.30
0.25
0.20
>C%J
LL.0-0.
Oslo
0.05
Figure 3 - Lateral Drift Force Coefficient for Series 60 Block 0.60 Hull
(Circles denote experimental points from Reference 4, for /3 = 120 degrees)
LEGEND
¡05°
1200
135°
1500
¡65°
0.5
I.0
'.5
L
N
0.35
Ó.30
0.25
cO.2O
-J
0.I5
OJO
0.05
0.5
1.0
1.52.0
L
Figure 4 - Drift Moment Coefficient for Series 60 Block 0.60 Hull