• Nie Znaleziono Wyników

An axis-symmetrical dock in waves

N/A
N/A
Protected

Academic year: 2021

Share "An axis-symmetrical dock in waves"

Copied!
41
0
0

Pełen tekst

(1)

ARCHIEF

SEOUL NATIONAL UN W. RSITY

COLLEGE OF ENGINEERING

DEPARTMENT OF NAVAL ARCHIÌECTURE

AN 'AXIS-SYMMETRICAL DOCK

IN WAVE

by

H. Isshiki and J.H. Hwang

SHIP HYDRODYNAMICS LABORATORY

Report No. 73-1

DeIfL

(2)

by

H.

Isehiki and J.

H. Ifwang

Report No. 73-1

SHIP HYDRODYNAMICS LABORATORY

Department of Naval Architecture

College of Engineering

Seoul National University

Seoul, Korea

(3)

Abstract

Introduction ....

Linearized theory of motion of an axis-symmetrical

dock 3

Variational formulation of the boundary value

problem

lo

k. Rayleigh-Ritz procedure for the variational

formulation 1k

5. Numerical calculations and discussions 22

RefQrences 23

Appendix A 25

Appendix B 27

Tables 29

(4)

1. Introduction

Motions of an axis-symmetrical dock freely floating on a free surface of water are important for designs of

artificial islands or ocean platforms. The hydrodynamic

properties of a floating semi-sphere were discussed by

Havelock (1955) and Barakat (1962), and those of'

semi-spheroias by Kim (1965) and Sao, lVlaeda & Hwang (1971). By H. ISSHIKI AND J. H. HWANC

Department of Naval Architecture, College of Engineering,

Seoul National University, Seoul, Korea

Linearized motions of an axis-symmetrical dock freely

floating in a regular plane wave are discussed. An

extension of the Bessho variational principle (Bessho (1968)) is derived to obtain a numerical procedure for a solution of the boundary value problem associated with

the fluid motion.

The added mass and the damping coefficients for a circular dock in vertical (heave) and horizontal (surge) oscillations are evaluated numerically, and the results

(5)

Miles & Gilbert (1968), Garrett (1971), Miles

(1971)

and

Black, Nei & Bray (1971) discussed wave forces on a circular

dock. In the present work, a general variational formulation

for radiation and scattering of water waves by an

axis-symmetrical dock is obtainedas an extension of the Bessho.

variational principle (Bessho

(1968),

Isshiki

(1970, 1972),

Mizuno

(1970),

Sao, Maeda & Hwang

(1971)).

Sao et al. also discussed heave oscillations of a

circular dock on the basis of the Bessho variational principle

and slender body assumptions. But their method seems to be

out of their assumptions for this problem.' In this paper,

the same problem is dealt with from a little different point

of view. An extension of the Besaho variational principle

is obtained, and the Rayleigh-Ritz procedure is derived based

on the variational formulation and an admissible potential

similar to that by Miles & Gilbert (1968).

The added mass and the damping coefficients are.

calculated for a circular dock in heave and surge oscillations. The numerical results seem to be reasonable, and are

compatible with the results for a sphere and epheroids.

Near the end of the preparation of this paper, the

authors received an.interesting papBr from Dr. K. J. Bai

(Bai (1972)J which also discusses linear water wave problems from the standpoint' of variational calculus. He regards

the Sommerfeld radiation condition as a boundary condition' on a circular cylindrical surface which is assumed to be

located far from 'a body, and approximates the

boundary valu

problem in an infinite region by a problem in a finite

fluid region within the cylindrical surface.

(6)

Let an axis-symmetrical dock be freely floating on a free surface of water as shown in figure 2.1. Q(x,y,z)

Fig.21

is a right-handed Cartesian coordinate system fixed in the

space. x- and y-axes are taken on the calm water surface, and z-axis is directed vertically upwards. (r.,,.z) refers to the space-fixed cylindrical öoordinate system Buch that

= r cos Q

(

y=r sin Q.

Let G be the center of mss of the dock, and be on the axis of the dock. At an initial instant of time, the dock

is assumed to be in its hydrostatic equilibrium with z-axis coincident with the axis of the dock. Furthermore, the equilibrium is assumed to be stable.

Let the free surface elevation %(x,y)e_it

of a

regular plane wave, which travela.in water of the depth d

along x-axis..from left to right, be approximated by

= 0 e

where

Jm(1C coB mQ,

i = imaginary unit = circular frequency

ampiitude of the free surface elevation

for m=O

for m = 1,2,3,

= Bessel function of the fist kind of order

2.1):

(7)

YO X,y, z

. coshk(z+d) jm (kr) cos 1n, 'O d cosh kd . m m

whére g is the gravitational acceleration.

Since is on the axis of the axissymmetrical dock, themotiön of the dock ,de to the regular plane 'wave (2.2). is assumèd 'to be confined to xz plane. If is

sufficiéntiy small compared with a' principal dimension'

of the doc:k, the. motion of the dock may be assumed tobe small from the assumption of the stable hydrostatic

equilibrium of the dock Then, the motion of the fluid

also bécornes small. The linearized theory Is assumed in

the followings. Let

ÍGe_t

, ZGe_'* and

be the stationary harmonic parts of .the horizontal, the vertical and the angular displacements of G respectively (figure 2.2), and 9(x,y,z) be the velocitr potential .

Fig.2.

of the fluid motion. Then, theunknowns X ,

and Y are the, solutionsof the following linearized

boundary value problem (Wehausen & Laitone (1960)):

(.+

2+)0

òx ày òz

nx(idXG)

+

nz(_iZG)

-

_'

cosh. k(z+d) eC O 'cósh lcd in

The velocity potentia.10(x,y,z)

e_t

of the regular incident wave (2.;2) may be approximated by

(8)

2

( )5=O

on (2.5è)

y=o

at z = -d (2.5d) 4J ( ik )(',) «-'

o()

as

r +,

(2.5e) and 2N

XG = °?

Ss dS (2.6a) + S% ) ZG =

1P Ss

n2 dS (2.6b)

= iclp J5{xn2. - (z_zG)nXJ.ffas, (2.6e) where

p = density of the fli4d

D = mean region occupied by.the fluid S = man wetted surface of the dock

F5 = mean free surface (i.e., calm water surface) V = mean displaced volume of the dock

S = mean water plane area of the dock

=

(n)

= unit normal of S into the fluid = height of G above calm water surface

M = mass of the dock

'G = moment of inertia o the dock corresponding to the angular displacement

I

= J (Z_2G) dV

=

Js

2

dsw.

(2.5e) is the Sommerfeld radiation condition.. Since (2.5b) can be written as

(9)

= - zn =

Substituting (2.8) into (2.6), the following equations are obtained: 2M X =

iøpçoFox

-

idpO1DI

d2pxaFxx -

d 1OX+ zGFxx)

(-d2M + pgS)Z.

= 4poFoz -

PODZ

-(_2

+ pg I + pg

)O

=

-ip0(F

+ ZGFOX) - + ZGPJJX) - d2pXG(PxO + ZGFXX) - + ZGP + ZGFOX +

¼2

(2.8)

(2.9)

Qn S (2.10) (2.11)

y

=

- Z

+ nZ(icZG) + (xn

- zn)(_i

,5 may be decomposed as = o

o +

+ (_iÖx

- ZGIÓG)X

+ (_iccZG)$Z + where - j cosh k(,z+d eu1 cosh lcd

00

-

ii cosh k(z+d) '-" cos mO, cosh kd m=O m and

D and are harmonics in

D which

satisfy the free surface condition, the bottom condition, the Sommerfeld radiation condition and

ô A

ônF'D

A

(10)

F1j=j1fjdS=jiidS

(I = o,D,X,Z,e; J =

It must benoticed that and F1 (I,J = D,X,Z,)

are determined by th geometrical characteristics of 3

Because of Green's integral theorem, the hydrodynamic

force (I,J = D,X,Z,e) is equal to . i

-=

Js[$I

_.J

-] dS = O

(i,J =D,X,Z,).

Let X , Z and be the displacements of the

point fixed to the dock which has the mean coordinates (x=y=z=0)...±t' then follows that

XG=X_zG

ZG=Z

From (2.10), .-iX, $ and -i.

are the.

velocity, potentials due to the simple harmonic oscillations of the dock in still water corresonding to the

displacements X , Z and respectively. From the

pressure equation and (2.12),

-2p{x,z,OJF1

(I,J = X,

z,e) is the J-th component o± the hydrodynamic reaction due to:the I-th motion (I = X,Z,). Let

be the component f the hyd.rodynamlc reaction with repeet

tothe center of mass G coxresponding to the dispiaòement (XG,ZG,©G). Then, the relations between F1 and

can easily be obtained as . , . .

(2.12)

(2.13)

(11)

GZZ = ZZ

9 = PlrI + ZF)( + z&F + ZG' xx

= F

+ ZGXX

GX = + ZGFXX,

where other components are equal to zero from the geometrical and the kinematical syminetricities.

Re[-pF13]

and Im{-pF1:l

1) (I,J = x,Z,) are

called the added mass and the damping matrices corresponding to the motion (X,Z,), since

'j

= {-d2X,-d2Z,-d} Re [-pP1 I

+ {icX,id'Z,ic$O} Im[-dpF1]. (2.16)

Let F (I,J = 'X,Z,) be the complex conjugate of P1

It then follows that

2k im[F1j] = PIJ

-*

*ò$I

= (O (21L * = 11m dz t rdQ [

-rJ-d

JO

i âr J'

since, from (2.10),

$1/n

(I = x,z,) is a real

valued function on S . Hence,

Im[F1]

(I,J = X,Z,e)

can be determined by the assymptotic behaviour of the corresponding potentials. Let the assymptotic form of

i) Re[ J and Im[ J mean to take the real and the

imaginary parts of the quantity in the square bracket.

(2.17) (2.15)

(12)

(I

= x,z,ÇJ)

, vhen r tends to infinity, be given as

A1()

cash k(z+d) eiOE

r

d cash k

substituting (2.18) into (2.17), 2i Im[P1j] can be

derived as

2ig2 d cosh kd sinh kd 2i Im[P1j3

2(cosh k5' + 2k (27

j

4(Q)

(I =

x,z,)

represents the wave

excitation force to the dock fixed in the space,

and

_iccp0P01 is called the FroudeKrylOV force. Prom Greèns

integral theorem,

OIDI

can be written as

o$ DI =

505D

1

dS

J3[$o

- i (0 (27t sI =

hm

dz \ rdQ[ 5

-r+DoJd

JO

Substitution of (2.9) and (2.18) into (2.20) leads to the following expression of the wave excitation force:

2j/3E. g e2 ilt/4

d cash kd sinh kd

'0IDI

2(cosh kd)2

(

2k

A(1L).

This relation is called the HaskindNewman relation

(Newman (1962)).

(2.18)

(2.19)

(2.20)

(13)

3. Variational formulation of the boundary value problem

Prom discussions in 2, it can be deduced that. a

boundary value problem of the following type must be solved to determine the wave forces and the hydrodynamic reactions:

2 2 2

mD

(3.la) ay öz on S (3.,lb)

(Ô2)

on

at z=-d

(3.ld) jf ( - ik ) $ -..-0() as

r -oo

(3.,le)

Bessho (1968) and leshiki (1970, 1972) obtained an variationál expression for the hydrodynamic force associated with the

boundaryvalue. problem (3.1). In the followings, an extension of their theory will be discussed.

Devide the region D into two subregions D and DU

as shown in figure 3.1, that is

I

U

D=D +D.

I

I

The intersection of S and D is denoted by S , and

U U

I

U

that of S and D by S . Let $ and $ refer to

I

U

the velocity potential $ in D

and D

respectively. Then, the boundary value problem (3.1) may be formulated

F1S3.1

(3.2)

I

in terms of $ and a2 U $ + o2 as follows: r in in on on

I

D U D

I

S (3.3a) (3. 3b) Ox2 + dy2 Oz2

(14)

The boundary value problem (3.3) can be transformed into a

I

variational problem. Let L{ ,$ ;f] be a functional of

z ir

the argument functions $ and $ such that

L [$'

's;

±' = (

4+

$ dD +

7=I,TIJS

-f (

'_ dF8 F8 Oz g ,. =1 ! ôz dxdy

+_1z

:ir( JD,-D Or d(DtDR)

It then follows that

'L=2

A=1,xJD

ôx2 Oy + 2

t: I

(

'-

f)

dS On - 2

1

(

'-

$'

dF Oz g dD a 2 = I

= o

C ik )

3' o()

on at as

I

D on z = -d r -,co, and

IiI

D,-D D : (3 3c) (3.3d) (3. 3e) (3. 3f) (3.:g) A,Í

and the conditions

a

.;p

-on the intersecti-on of =

a

1

=p

(15)

:. z (O (27t.

- uni

I dz I rd@ (

Á1I)

p

rooJ-d

Jo

+ 2

JD',.D

Á

à(D'i-J?). (3.5)

It is easily seen that (3.3a-d) and (3.3g) are the stationary conditions of the functional L under the subsidiary

conditions (3.3e) and (3.3f). Hence, the following variational problem is equivalent to the boundary value

problem (3.3): L[,3',?;f] = stationary, (3.6a) under .ff ( -

ik )

t'-, O() as r

-tao

(36b)

I

I

Z = on D,-D . (3.6e) X

It must be noticed that L[

,$

;f] is a variational

expression for a quantity -Çf dB , that is

iS

= -

dS, (3.7)

,=I ,fl

JS

when $ (¼=x,1!) is the solution of the variational problem (3.6).

If the natural conditions (3.3a), (3.3c) and (3.3d) of the variational problem (3.6) are also considered as the subsidiary ones, the following variational problem is

then obtained:

+ 2

L1

dxdy

(16)

undé r

M

-.1

.

t

E ,

;f] =.L[

,

;f]

2f) t

dS

J'cD

Ç?-stationary, a2

ay

az

2 =

I

a 1

l$1r

= O

.,,ç

-. ¡k )

'....'o()

.1

E

mD

E in. D on

as

r

-too,

I

If on D,-..D

The natural cönditions of this variational

problem are

as

follows:

,Ön

(3.8g)

= on

D',-D.

(3.8h)

The

variational problem (3.8). is an extension of the

Bessho

variational formulation (Bessho 1968).

-..

I

U'

If the continuity of the normal velocity. on

D,-D

is assumed, a complementary variational fprmulation of.

(3.6); oai be ôbtained as follows:

E. ,on S (3..8a) (3.8b) (.3.8e) (3.8d) (3.8e) (3,.8f)

(17)

-I

:ii

-

'-i

I

-

r $fl) d(D',-.D)

Jii ,-j

= stationary, (3.9)

under (3.3e) and (3.3g). (3.3a-d) and (3.3f) then become the natural conditions of this variational problem.

Corresponding to (3.8), it then follows the following complementary formulation: * i t M 1 ,$ ;f] = M[ , ;f] - 2

Ç

(

''_

$) d(D'r?)

JD,.D

= stationary,

under (3.3a), (3.3e-e) and (3.3g). Then, (3.3b) and (3.3f) are given as the natural conditions of (3.10).

4. Rayleigh-Ritz procedure for the variational formulation

An admissible function for the variational problem (3.8) will be obtained in the followings, and a Rayleigh-Ritz approximation based on the admissible function will

be discussed. Let $

00

= Em jm cos inQ (3.10)

(18)

I(cXr)F J(ftr) H(1) (pr)

3t'(r,z)

= E

eos mQ

t

in D K in D 1

[e' + em+/2)71]

1/27cjr 1

[ei{_12m)/41

+

eiß«2m+1)/4}

s/2JLr

i i{r-7i(2m+]j/4}

Vmre

(4.4) (4.5) 4.1)

be a solution of the Laplace equation:

r2

(4,2)

It then follows that

j; j; + ) 3b1

= o.

The solution of this equation by the separation of the variables are as follows:

Km()1

(cos cXz Im(atr)) sin

dz,

where

= real parameters

'm' Km = modified Bessel functions of order m

H1)

= Hankel function of the first kind of order

m

I(cfr)

,

Jm(r)

are regalar at r = O , and Km(cfr)

unbounded at r = O . When r tends to

infinity, these functions have the following assymptotic

expansione:

e0

J(r)

1r

f

e

H(r)J

I

e,

(19)

PrOm the propertis at r 0, cO.

and

the condition at

d ,it may be assumed as follows:

)bX(r,z)

-)(r,z)

K(cr) cos c«z+d

I(cI'r) cos '(z+d)

The free surfacé cond tio. (3.3e) then yields the following eigen-value equations for and.

(4.7a) has an infinite number of roots

and the solutiön of (4.7b) is equal to the wave number of shallow water k (see (2.3)). Hence, 3b1(r,z) may be

assumed as

X'

(r,z)

= A

K(r)

cos 1(z+d)

B' H(kr) eoshk(z+d)

m m

coshkd

where': = linear combination of o(

tan. cd

= -d2/g

ß tanhpd=

where A1

(i=l,2,

. )

and

Bm

are integral constants,

and r0 is the radius of the

cylinder

DK.

(figure

.1).

From (3.8f),

3C(r,z) must be,a function such that

. . )b(r0,z).= 3b'(r Let '(r,z) be decomposed as z, Ku

(-r,z)

= 'th (r,z) + 'm (r,z), = r) cosh ß(z+d)

and-i

(46)

cosh 1ß'(z+d).

i:

ce.>Q

ß>O.

.1 K

on D,-D

I

On

DpD

4,8)

4.9)

(.4 .lOa) LiJ ,z) = C)

(20)

Then, it may be assumed that rio

I(c1r)

'"m (r,z) = A I

r)

i=1 where B mi 00 cosh k1(z+d) "m (r,z) =

B, J(k1r)

cosh k d 1=1 mi

are integral constants,

and

=

m,i1O

Hence, 1//(r,z) may be assumed as

00

I(.r)

= Ami

i (OEr)

cos c1(z+d)

+ B' H1)(kr0)

J(kr)

cosh k(zd) m J (kr0) cosh kd in ori cosh k (z+d)

+ T

B. Jrn(kmjr)

coshk.d

In the followings, an approximation based

on the

variational problem (3.8) and the admissible functions

(4.8) and (4.14) will be considered.

For the sake of simplicity, a numerical procedure will be discussed for a circular dock of the radius a

(figure 4.1). And r0 is assumed to be equal to a (i = cosh k(z+d) + B J (lcr) in in

coshkd

where

r

A . = A . (i = 1,2, ) mi ini (i B J (kr0) = B H (kr0). (4.:llc) in in m m

Let j . be the ith zero point of the Bessel function

:

in, i

= o

0<i1<i

< . (4.12)

m m,2

Then, (r,z) may be written as

1,2, )

cos

c(z+d)

(4.14).

(21)

Substituting (4.1) and (4.15) into the functional

I

M[5 ,,S ;f]

M

-

2f'(z,) ) Jr=a

adQdz

N[$',?;f]

1-(d-h)JO

ta(27t + J

[(

-

-

2f (r,e) ) ]z=-(d-h rdrd9

joJo

1-(d-h)

12n[M'

1

J-d

j

r är ir=a ad9dz, (4.16)

the following expression for M is obtained:

z 00 z u z ut.,

M[5 ,5

;f] = Em(_i) Mm[3'.t'm;fm,fm.J, where

Mm[XE!

Y'm,Y'm;fm,fm]E

Io

[

-

a ( - - 2f(z) ) b1 ]r=a dz - J-(d-h) u 2f(r) ) rn]z=-(d-h) rdr (-(d-h) Y' + a I

[]r=a dz

J-d

1_(d_h)

J-d

In r=a dz.

-a

Substitution of (4.8) and (4.14) into (4.17b) leads to the following expression of Mm : (4.17a) (4.l7b) Let f f =J be expanded on S

I

(z,Q)

as

I

(fm(z) cos m on on z S S (4.15) = E jm m

1f (r,)

(22)

-U

i

n

-

IA .,B ,Bmi;fm,fm]z z n

r

it

in'- mi in

00 00 z r

r

00

zr

=)-

[A,A]

A

i1 j=1

ni ni ij miAmj +

>-

i=1

[A,B]1 AB

r1

I

12

00

I

It

I

U + 'B ,B

i

(B '- ni

m'

ru +

y-

[B,BJ. B

B

i

m ini i=1 00 00 r ir

I

it 00 00 +

E:

AmiBmj +

E

i=1 j=1 1=1 1=1

[BB] BBJ

00 X

I

I

I

U Ir

-

2

y-

[AJ± Ami - 2 [Bm] B

-

2 >

[B1i

, (4.18) i=1 1=1

where [AA]

, etc. are given in appendix A.

Because of Green's integral theorem, the following relations

hold:

[AA]

=

=

[BB]1

(4.19)

From (4.17a)

and (4.18),

the stationary condition of

the functional M is given as

Mm[Ai,3,Bi;fX,fU

]=

stationary

(4.20)

Hence, the following linear

X

I

algebraic equation for the

unknowns

Ami B

and

Bmj is obtained:

r

r-

x

r

r

r

rit

E x

'T

2[Am,A ijAmj + [A ,B ] B +

y-

[A,B:li

.B

= 2[Am]i in

mirn

[t

AB

I-

Z

I

X

I

I

E it

I

. +

2[a,

]B +

[B ,B

J B

= 2

mjm

mm

=i ru in

j

mj [B'] (4.21) for in = 0,1,2, 00

YT [A»].iAT.

j

=1 +

[I

U

I

It Z ir E

BmBmiiBm

+ = 2[Bm]i

I,

I

U where i = 1,2, . IA I. ,

[b]

and IB

i.

are

rni

(23)

'-determined according to the conditions on S . These conditions are given iii table 4.1, where "heave", "surge" and "pitch" refer to the vertical (Z) motion, the

horizontal (X) notion and the angular () thotion about y-axisrespec±ively ((2.10)). By using the formulae in appendices A-B and table 4.1, ,

[B]

and

for these motions can be written exlicitely as shown in

table 4.2. In numerical calculations, the infinite series

in (4.21) are truncaed byfinte tèms according to an

approximation:

I

=.O for

i'P

E (4.22)

= O for

where P and Q are appropriate integers determined by the

numerical convergence.

The added mass and damping can be approximated as

follows. From (2.12), (2.16) and (3.8), these quantities

p

e-= j.[ M[ ,$ ;f] ],.

pIm)

-

I

E

where (.$ ,$ ) is the exact solution of the variational

problem (3.8). i'rom (4.17a), it then follows that added mass damping. where

m=1

P

Re}[

],

apIm

for. heavé

for surge or pitch.

4.23a)

(4. 2 3b)

can be written as

added mass

-pRe

(.

j=.

f. dS E $

dB]

(24)

I.

Let Ami (i. = 1,2,...P) B and Bmj (ï 1,2,....Q)

be ail approximate solutionof (4.21) based on the assumption

I.

n

(4.22), and ,)" ) the corresponding approximation of'

(i,,y) . Then, may be approxImated by

Mm[31 ,3

;f,f]

, that is

ME,;f] M[',4

,3bm

From. (4.18) and (4.20), the following expression for

Mm[)I ,Y'

,f,f]

is obtained.

i z ir z n

z x n z

n

mrn

m'

Ç]

=

M[A

Bm ,Bmj;fm,fm]

___ T

T

Î

= T

[A1]i Ami -

[Bj

The óoupling force such as the surge (pitch) component in

itch (surge), is also easily obtained (,Issh±ki (1970)).

If the numerical convergence is satisfactory, the radiation

V V.

z.

wave amplitude may be calculated from Bm. by using the

assymptotic expansion of H(kr)

((4.5)) and the linearized dynamical condition bn the free surface:

_jV

. -idt V

=0

at z=0,

where is the free surface elevatiom. From the

Haskind-Newman. relation (.2.21), the waver excitation force

will bé easily' caicu.lated.l). V

i) From .a standpoint.6f thé variational calculus,, the

radiation wave amplitude br the wave excitation force should be. approximated by using the radiation and the

catteringpötentja1s (Isshjki. (1970))..

(4.24)

[B] B:

. (4.25)

(25)

5. Numerical calculations and discussions

The added mass and damping of a circular dock (figure 41)

were calculated on the basis of the numerical procedure

developed in

54.

The numerical calculations were carried out by the computer IBM 1130 at Seoul National University,

and the single precision (six significant figures) were adopted.

The added mass and the damping coefficients are defined

as follows i

added mass coefft. = added, mass/Ep2(d-h)]

2 (5.1)

damping coefft. = damping/['p7a (d-h)].

In tables

5.1-2,

the numerical tendancies of the added mass

T.Lb.

5.1

and the damping coefficients with increasing P and Q are

Tab. 5 . 2 shown, and seém to be satisfactory. From tables

5.3-4,

Tab. 5. 3

the convergence of Bm seems also to be reasonable.

Tab .5. 4

In figures 5.1-2, the frequency dependencies of the

Fig. 5.1 added mass and the damping coefficients are shown. The

Fig. .5. 2 free surface effects seem to be compatible with the results

of a semi-sphere by Havelock

(1955),

Barakat

(1962)

and spheroids by Kim

(1965),

Sao, Maeda & Hwang

(1971).

The

horizontal force on the dock ( d/a =

0.75,

h/a =

0.25, 0.5

) due to wave ( k =

1.32

) was compared with the correporiding results by Garrett

(1971),

and the agreement was satisfactory.

The authors are deeply indebted to Mr. 1. H. Kim

and Mr. K. P. Rhee for their unlimited cooperations in the

(26)

R}FERENCES

BAI, K. J.

1972

A variational method in potential flows

with a free surface. Dissertation, Univ. of Calif.,

Berkeley.

BARAKAT, R.

1962

Vertical motion of a floating sphere

in a sine-wave sea. J. Fluid Mech.

13, 540-56.

BESSHO, N.

1968

Boundary value problems of an oscillating body floating on water. Memoirs Defense Acad. Japan

8, 183-200.

BLACK, J. L., NEI, C. C. & BRAY, N. C. G.

1971

Radiation and scattering of water waves by rigid bodies. J. Fluid

Mech.

46, 151-64.

GARRETT, C. J. R.

1971

Wave forces on a circular dock.

J. Fluid Mech.

46, 129-39.

HAVELOCK, T. H.

1955

Waves due to a floating sphere making periodic heaving oscillations. Proc. Roy. Soc. Lonion, Ser. A

251, 1-7.

ISSHIKI, H.

1970

Variational principles associated with

surface ship motions. Korea-Japan Seminor on Ship

Hydrodynamics, Soc. Naval Arch. Korea.

ISSHIKI, H.

1972

(under publication) Variational principles

associated with ship motions in waves.

KIN, W. D.

1965

On the harmonic oscillations of a rigid

body on a free surface. J. Fluid Mech.

21, 427-51.

MILES, J. & GILBBRT, F.

1968

Scattering of gravity waves

(27)

MIZÌJNO, T. 1970 On sway and roll motion of some surface-piercing bodies. J. Soc. Naval Arch. Japan 127,

55-70.

NEWMAN, J. N. 1962 The exciting forces on fixed bodies in

waves. J. Ship Res. 6(3),

10-7.

SAO, K., MAEDA, H. & HWANG, J. H.

1971

On the heaving oscillation for a circular dock. J. Soc. Naval Arch. Japan 130, 121-30.

WFHAUSEN, J. V. & LAIT0NF, E. V.

1960

Surface waves.

Encyclopedia of Physics vol.

9, 446-778,

Berlin: Springer-Verlag.

(28)

Appendix A.

K(.5)

O cos c(1(z+d) cos c(z+d) dz [AmAm]ij =

ca

Km(ctia)

d

+ Im(cia)Im(ci.ja)

r Im(ctjr)

i(cr) dr sin

cth

cos d.h

I(ca) ((dh)

cos d(z+d) cos

oz+d) dz

- dia

Im(ja)

Jd

K(ca)

i

=

{ka H'(ka)

+

dja H1(ka)

K(da)

}coèh kd

(O .

cosh k(z+d) cos c(z+d) dz

Jd

(

k81h

{_

I(1a)

Ç

r I(cX1r) dr cosh kd COS cXh

cosh kh - °i cosh kd

i

((dh)

cosh k(z+d) cos ci(z+d) dz}

H1(ka)

J(ka)

cosh kd

(O i (cosh k(zd))2 dz

[B,B1

= ka

H1)'(ka) H1)(ka)

(cosh kd)2 )d

ra

r

(J

(kr))2dr

sinh kh cash kh

)o

cosh kd cash kd.

i

(cosh k(z+d))

ka J(ka)

(cosh kd)2

Jd

2dzJ.

H1»(ka)

2 )

sin cX1h )

( ka

J(ka)

+

o(ja J(1c5

(29)

a cosh kh

sinh k.h

((dh)

H1)(ka)

.1 cash k(z+d) cash k(z+d) dz}

J(ka)

)d

=

-

Im(ja)

: r

mmj

Im(çr) dr

sinh kmih cosh k h

'(k

mj cosh kmjd cas

ch -

csin

ch

cosh krnd

I'(ca)

- (

;5

J(kmja)

+ cX.1a )'

r(dh)

i I cosh kmj(Z+d) cas ç(zd) dz cosh kmjd

)d

r a sinh k .h cosh k r J (k .r) J (k .r) - Ifli m

LBmBmJij =

_kmi m mi m m ar cash kmjd cosh kmjd

- kmia

J(k1a) JmCmj

cosh kmid

Cosh

kd

((dh)

cosh k (z+d) cosh k .(-z+d) dz

d

m

(o

[A].

= a I f' (z) cas c(z+d) dz

)(dh)

m +

r f(r) I(t1r) dr cas

cth

[Bt,BU]. sinh r

mmi

kh

cash kh

)

dr ( kmi

cosh kd cosh kd

- (

ka

J(kmja)

i

+ k

cash kd cosh kmjd

(30)

(O

I

[B ] = a

HW(ka)

i )_(d_h cosh k(z+d) dz m m cosh kd

cosh ich

H»(ka)

(a. + )

r f(

dr cosh kd r (a cosh k .hmi

LB ]

= r f (r) Jm(kmjr) dr còsh k .d

mi

m mi

Formulae for the integrals fcos 1(z+d) cos

(zd) dz

,

r Im(c:ir) Im(cjr) dr , etc. are given in appendix B.

Appendix B.

cos cos cXz dz

i

( sin cXz cos cz

- Cos cXZ Sin Z )

{

-z i

+

sin 2cz

for c=.c(.

Jcosh kz cos cxz dz

( k sinh kz cos jZ + cXcosh kz sin

1z )

(31)

Çcash kz cash

kz

dz i

( k .cosh k .z sinh k .z - k .cosh

k .z sinh k . )

2 2 mi mi mj mi

kmi_kmj

for k1 kmj

z i

+ 2k cosh kmjZ sinh kmjZ

for km= k.

I

(cx.r)

Im(jr) dr

j

mi

{_cXjlm(cxjr) Im_i(jr) jI_1(cxjr) Im(cxjr)}

for

r Jm(kr) I(cx.r) dr

mit1)

- I_1(1r) 'm+l

r

- k2+

{jJm

Imi(jr) - k Jm_i(1C1

Im(jr)}

r Jm(kmjr) Jm(1Cmj1 dr

rk1_k

kinjJm(1cmi

mlmj

- kmiJmi(kmir) Jm(kmjr)

2

for kmi4: kmj

t{ mmiY2

- m-1mi

m+1miJ

for kmj=

m+1

rm+i

I(c.r)

dr r

I1(c.r)

rn+i

(rm+1

j (kr) dr

-)

(32)

TABLE 4.1. Kinematical condition on

S

ft(z,)

?(r,Q)

m

f(z)

f(r)

heave(Z)

0

-1

0 0

-1

surge(X)

cos

O

i

-1/2

0

(33)

TAL$L

4.2.

I

[Bu] and [Bm]1 heaveZ); m = O surge(X); m = i jAm] i a 11(d.1a) -( s in

a -

sin ci1h )

-

(Ç a) co s [B - cosh kh

H,1)(ka)

la

H(ka)

i sinh kdcosh kd sinh Ich J1(ka) cosh kd J0(ka) cosh kd r LB

J.

a

coshk0h

o J1(k01a)

- ç

cosh

k0d

pitch(s);

rn = i z [Am]i cos cL d cos cL h 1a ci. + cj21 - sin

cth

)

I

a

12(cL1a)

+

Ç

I1(ci.a) 005 IBm]

i

H(ka)

d sinh kd i h sinh kh i cosh Ich

cosh kh

H(ka)

coshkd;2-1COShkd;2COhkd)

a + osh kd J1(ka)

J2(k1a)

cosh k11h 2k1 .cosh k1d

(34)

TABLE 5.1. ConvergenCe of theadded mass and the damping coefficients (heave);

(2a/g

= 1, d/a =6, h/a = 4).

TABLE 5.2. Convergence of the added mass and thè damping cbefficints (surge.);

= 1, d/a 6, h/a. 4. ). 5.. added mass .coefft. 0.2462 0.2609 damping coefft. 0.00422

0.00393

6 added mass coefît.

0.2519

. 0.2665 - -a-... damping coefft. 0.00412 0.90383 i 5 -adde.d mass coefft. .

0.5411

0.5415

-damping coefft.

0.-5346

.

0.5351

6 added mass coefft.

0.549

0.5433

damping òoefft.

0.5354

.

.

.0.5359.

(35)

TABLE 5.3. convergence of B

/

cosh kd (heave);

(2a/g

= 1, dIa = 6, h/a = 4).

5 6 i P (O.2821O.4959i)

.io

(O.2787+0.4900i) .10 2 (0.2720+0.47861) (O.2686+O.4727i) .iO

TABLE 5.4. Convergence of B

/

cosh kd (surge);

(2a/g

= 1,

d/a = 6, h/a = 4).

i.

2

5 _(o.4273+0.1541i).102 _(o.4275+0.1543i).102

(36)

01

7

d

(37)
(38)

I

region D

I I

I.

t regionu

I I I t

'reg jDi

I

J

FIGURE 3.1. Division

of D into

subregions

D1

and

DU.

region

region D

(39)

z

4 I

region D

region

D11.

h

y

rjIs,'

region

/

FIGURE 4.1. Circular dock,

(40)

1.0

08

.0.6.

0.4

.02

o

i

2

3

4.

o'4a/g.

FIGURE 5.1.

Added mass and damping

coeff. for a circular dock in heaveZ).

(41)

0.5

0.

02

o'2cig

FIGURE. 5.2

Addéd mass and damping

.cöeff for a circular dock in

surge (X).

Cytaty

Powiązane dokumenty

ке спрямування даної освітньої галузі, що, через свою специфіку, вимагає особистого контакту творчого наставника з

82 W oryginale podana fraza napisana większymi literami oraz podkreślona podwójną kreską. 83 W oryginale podana fraza napisana większymi literami oraz podkreślona

Jest rzeczą oczywistą, że badając dzieje Światpolu, autor sięgnął przede wszystkim do materiałów źródłowych Związku — archiwaliów (niekompletnych niestety), źródeł

We consider time-delay linear fractional dynamical systems with multiple, constant delays in the state described by a fractional differential equation with a retarded argument of

An infinite family of T -factorizations of complete graphs K 2n , where 2n = 56k and k is a positive integer, in which the set of vertices of T can be split into two subsets of the

Then, as a counter-example to uniqueness in the range 0.5 &lt; m &lt; 1.0, Cosserat [12] showed for the circular domain, that a non-unique solution for the displacement boundary

[r]

In the case of the Laplace equation we can transform the basic problem into an integral equation on the boundary of the domain by using the single or double