Nomenclature
i . Introduction
When we design high speed displacement type ships, we think of making the resistance small at first. The resistance consists of frictional resis-tance and residual resisresis-tance. When a ship cruises
Read at the Spring Meeting of Kansai Society of Naval Architects, Japan, May. 27, 1994, Received June 10. 1994
\
UniversityTRDI, Japan Defense Agency
at high speed, the residual resistance takes most part of her resistance. Moreover wave making resis-tance is the largest in the residual resisresis-tance. lt is known from recent studies that diverging wave com-ponent is large in the wave making resistance at high speed. So it is important to reduce diverging wave component.
On the other hand, it is also important for high speed ship to reduce spray resistance. There is a the-
-ory that it
is necessary to make the pressure zeroITEM DIMENSIONLESS ITEM DIMENSIONLESS
wetted length:Lw
derisity:p
coordinates
wave number
swell-up potential
wave making resistance
residual resistance another expression of each resistances 2 i (z', s", z') breadth:B veloc it y: V displacement
water plane area
pressure
spray resistance
water head resistance trim rise A B/LWL i Ç" V (z,y,z) = g(LwL/2) K (LLW/2)2 AÇq = = V2 A' P' A = LWL/2 D p pV3 D5 = = pV2(LWL/2)3 pV(Lw/2)2 n' DH "H DT pV2(LwL/2)2 D' pV2(LWL/2)2
r=tanr'
h'h=
r = pgV'/L.L222-
Z69
J. Kansai Soc. N. A.. Japan, No.222. September 1994A Theoretical Study on High Speed Ship
Hull Forms*
(Application for Practical Hull Forms)
By Masatoshi BESSHO** (Member) and Shun SAKUMA*** (Member)
When we design high speed displacement type ships, it is the most important to reduceher re-sistance. The faster a ship cruises. the larger wave making resistance becomes. This is the rea-son why designers want to search hull forms which give small wave making resistance.
More-over spray resistance is also large at high speed. Andit is known from recent studies there is water head resistance in a ship which has a vertical stem and a transom stern.
Hull forms which give minimum resistance components are studied, but obtained forms are strange and unrealistic. This report represents improvement of such strange forms and
applica-tion for practical hull forms. The obtained hull form has small spray at the bow. And this
form is enough to be realized.
at the edge of her water plane area in order to de-sign spray free hull forms. But there are few stud-ies what these hull forms become and how we make these forms.
A new resistan component is studied recently. It is water head resistance4. This component ex-ists in a ship which has a vertical stem and a tran-som stern. This component is studied very well on two-dmensional problem3'5. But we cannot ob-tain good result on three-dimensinal prob1em8.
We think that residual resistance consists of three components described before. But how do we design a hull form which makes these components
minimum? The authors try to obtain such
opti-mum hull forms on the condition which her water plane area is rectangle. The obtained hull form has very large difference between cruising draft and
still water draft. As the result of it, she cannot
run in free condition. This is caused by using pres-sure distribution which gives too small wave mak-ing resistance. But this hull form shows nice
perfor-mance at fixed condition. That is, there is no
steep wave which causes bow spray'°.
In this report, we obtain an optimum hull form in the practical region. We use optimum pressure distribution obtained from statical pressure
distri-bution. And the obtained pressure distribution
does not give so small wave making resistance. As the result of it, present hull form has very small
differen between cruising draft and still water
draft. Moreover curvature around bow is similar to previous optimum hull forms. So the bow spray is also diminished.
i Formulation of shallow draft ships The coordinate system is shown as figure 1. In this figure, F shows water plane area. CF shows forward periphery of F, and CA shows aft
periph-Fig. i Coordinate system
58-ery of F. We define p as pressure distribution, A as strength of swell - up potential4. In the linear theory, surface elevation is shown as follows
c
=
-
,y - ThO)dd17+ J A(i')S(z - F, Y - 77p, O)diip (1)
Here T and S are kernel functions. In order to ob-tain numerical solution, it
is necessary to
inte-grate these functions in a divided
panel. This method is shown in referernces79. 1f we give hull offset to , equation (1) is regarded as integralequation. So we can obtain p and A for given hull forms by solving
it.
Then Kochin function isshown as follows H(k,9) = JJ p(z,y)e °'
'dzdy,
(2) F H(k, 9) = H(k,8)-i cos 9J A(yp)e" cøtø+yp ImD)dyF C,
So we can obtain wave making resistance.
Dw = - f
H(Ksec2 99)2 sec5 8d9 (4)ir io
And we define as strength of singularity of pres-sure at bow
= ):!,
[(z/i
()2]
So spray resistance is
= o2(y)dy. (6)
Then water head resistance is
=
i I
A2(y)dy. (7)2 Jc,
-Thus we can know various things due to solving in-tegral equation (1)8). But if we give p and A to
right hand of equation (1), we can obtain corre-
-sponded surface elevation. So if p and A is opti-mum, obtained surface elevation is regarded as
hull offset of an optimum hull form.
Now let us discribe conditions of an optimum hull form. At first, we can easily find from equa-tion (7) that water head resistance becomes O by setting A0. Secondly, let us think of spray resis-tance. we can also easily find from equation (5)
Ctuinzed
pressure
di stri buti ori
Usirig infli.ence function
calcu)us of vari atioris
If water piare area is
1.rectarigie 2. ellipse
3.rectangle+eilipse.
If water piare area is arbitrary.
Fig. 2 Methods for pressure optimization
Now let us consider detail of influence function. \Ve can integrate equation (4) with respect to & on the condition of equation (9). So we can obtain Wave making resistance as follows
D =
.JJJJP(z,y)P(e,'T7)P_5(KT, Ki
,, 0)deddd1, (10)here P_ is very complex function defined by
Bessho2 shown as bellow.
59--P_5(x,y,0) = fcos(zsece)cos(ysec9sin8)
xsec59d8 (11)
Now let us consider the variation of D, that is,
¿1Dm, when
varies p+p.
Dw±Dw =
P_s(Kz -,
0)ddi7dzdy 2K2+ ¡f
p(z,y)dzd,x ir JIF JJ p(,i)P_5F (12)Here the second term of right hand is zlDw. We can define influence function as r.
2JC,
r(z, y) = ir Jjp1 p(, i7)P_5(K ¿, KjT, 0)
xddi7 (13)
So the variation of wave making resistan is repre-sented as follows
=
il
y)r(x, y)dzdy (14) This equation is very intresting. If we can findpres-sure distribution which gives r = constant, r
goes to the front of integral signs of equation
(14). In general, we optimize pressure distribution on the condition which the displacement is con-stant. So right hand of equation (14) becomes O.
Namely, it shows the pressure is optimum and
gives minimum wave making resistance. This is the reason why we use influen function for pres-sure optimization. Wave making resistance is
Dw
.ffp(z,y)r(,)dzd
(15)It is convinient for numerical computation to inte-grate P_5 in a devided panel.
K2
___
Jj(z,y; 77) =
P 5(K,KjT,0)
ir
Jill
xdedi7dzdy (16)
-The integration method of this equation is
repre-
-sented in references79. The integration of influ-ence function in a panel is represented as follows us-ing Ji
Hi 6 that spray resistance becomes O by using
rsure distribution which gives p=O at the edge tvater plane area. This condition is very
effec-. We verify that this condition realizes spray
t hull forms in our previous work.Thus we can optimum conditions for water head resistance
spray resistance shown as follows
p(z,y)=O 0nCA+CF (8)
A(y)=0 OflCF (9)
or a next step, we must optimize pressure dis-:ributiofl which gives small wave making
resis-hice.
How to obtain optimized pressure distribution There are two methods for pressure optimization shown in figure 2. Firstly, it is calculus of varia-tions. It is very effective when water plane area is one of three shapes shown in figure. We use this method on rectangle water plane area in our previ-ous work. Secondly it is the method using influ--'nce function. It is effective when water plane area is arbitrary. We use statical pressure distribution of series 64 model no.4793 as initial value. So we
optimize pressure distribution using influence func-tion.
Now let us vary pressure distribution using this function. In this work, pressure is varied in propor-tion to this funcpropor-tion.
p(z,y) =
cjr(z,y) + C2
(18)The negative sign of C1 shows characteristic of pre-sent method. If P is negative in a panel, pres-sure is increased. If positive, prespres-sure is decreased.
We assume that displacement is constant. So we can find easily,
ILp(z,)dzdy
=
_c1JJr(z,y)dxdc3JJdzd=o (19)
Substituting equation (18) for equation (14),=
C1 J' r(, )r(, y)dzdj+
c2JJr(z,)dzdy
(20)Each integration about r and
r* is defined as
followsA1 =
JJ r(,)dzdy
=(21)
Here s,,,,,
is area of a panel.
A3 =JJ r(z,,)r(,y)dzdy
=r.2(,
) (22)A3
=JJ r(z,y)dzdv= (23)
So the reduction of wave making resistance is
repre-sented as follows
¿D cA2_AlA3,1 C1
Dv,
-
aHere a is arbitrary positive number and A is wa-ter plane area. Following condition is necessary in order to reduce wave making resistance.
A3
A3> -
iwSo we can easily obtain coefficients C1 , C2.
If we use these coefficient,
ip can be obtained.
Thus pressure is optimized by iterative computa-tion from equacomputa-tion (17) to equacomputa-tion (26).
3. The computed results
The optimization is carried out at K=2.5. Fig-ure 3 shows behavior of wave making resistance and optimization parameter as iterative number
in-creases. Though we designate a =0.4, ¿1D/D
does not become 0.4. This is because C1, C2 is con-stant over the water plane area.
Fig. 3 Behavior of parameters for iterative computation
Figure 4 shows influence function and surface ele-
-vation which are computed with given pressure dis-tributions. The given pressure distributions are statical pressure distribution which is obtained with hull ofset and the computed pressure distribu-tion which is modified from statical pressure distri-bution using the iterative computation method de-scribed before. Firstly let us consider variation of pressure distribution. There are not so obvious dif-ference as iterative number increases. Observing longitudinal distribution, we find pressure shifts
gradually to fore and aft end. This trend agrees
with optimum C
curve theory'. But it is
suspi-cious that wave making resistance is reduced so much, because center of lift shifts to aft end
gradu-ally.
Secondly let us observe influence function distri-bution. We can find easily that the curves become
= JJ
r(,)dd
= 2 (17) C1 = aß, C3 =
ZETR
-L
PRESSURE DISTRIBUTION (BODY PLRN)
(c-2.500 LRIIDDR. .525 PuRe- .2894
990
PR55URE DISTRIBUTION (BOOT PLRNI -2.5DO LAF1BDA- .128 rIlRx- .1471
I :x0.990 2: 0.952 3: 0.875 4: 0.769
5: 0.631 6: 0.469 7: 0.289 8: 0.098 9: -0.098 10:-0.285 11:-0.469 12:-0.531 13: -0.759 14:-0.578 5:-0.052 16:-0.990
BURFRCE ELEYRTION IZETRI BOOT P1..RN
ZI1RX. .07664
PRESSURE DI8TRIBUTION (PROFILEJ
(c-2.500 LRrISCR. .128 PflRX- .1471 ZETA O ZETA O -L
PRESSURE OIBTRIBUTION PROFILE(
((-2.500 LRflBDR- .125 PIRX- .0394 a:y0. 017 b:0.049 c:0.075
d: 0.101 e:0. 118 f:0. 125
2 -2 0
SURFACE ELEVATION IZETRI PROFILE
¡flAc. .07594
INFLUENCE FUNCTION (GRrlrlRc PROFILE
0LiRe- .00L3L P14-5.55
L-L
statical pressure
distribution
INFLUENCE FUNCTION IQArrnR PROFILE
0LIRe- .001009 RM-2.37
6 times iterative computation
Fig. 4 Optimization of pressure distribution and obtained surface elevation
oath and thier absolute value becomes small as rative number increases. But distribution shape not so different. This is because the water plane
a is fixed.
astiv we compute surface elevation. It is notice-e that aft and fornotice-e notice-end bnotice-ecomnotice-e dnotice-enotice-epnotice-er as
itera-
-61--tive number increases. In present work, we draw body plan using 6 times iterative surface elevation.
The obtained body plan is shown in figure 5. There are unevenness in the computation result. So we carry out fairing boldly. The curvature of fore
part is similar to previous forms. \Ve think the
3 14 IS IS
i&
I 2 3 4LL
5618 1211 log e ¡ d/1
/11r:
fi' V. 13141516 ¿pcá.4
1 2 34 l2f
\
IO 6 7 8 -L 5 0.531 9: 0 098 3: -0. 159 2: 0.952 6: 0.469 10:-0.289 14:-0,818 3: 0.878 7: 0.289 11:-0.459 15:-0.952 4: 0.769 9: 0.098 12:-0.531 16:-0.990 L a:y0. 017 d: 0.101 b:0.049 c:0.078 e:0. 118 f:0. 126 -L O O 2 0 2SURFRCE ELEYÑTION LZETR( BOOT PLRN SURFACE ELEVATION (ZETA! PROF ILE
L = 1500mm B = 192mm V = 0.01538 FP 2' f,
/ /
s'Fig. 5 Body plan of the obtained hull form
spray
free condition gives such curvature. We
draw the bow shape such as a destroyer shown in figure.
There is small spray strip in fore part. This is be-cause previous optimum hull forms have much lager spray strip. It is caused by negative pressure for transverse direction. There is no negative part in present pressure distribution. So we have found spray strip is not necessary in present case.
Finally we compute displacement of this hull
form. The cruising draft is almost equal to still
water draft.
Fig. 6 Observation of bow spray (present hull form, Fn0.447)
62-0.04 0.02 -0.04 20 lo -o 4 . ExperimentThe experiment is carried out at circulating wa-ter channel in West Japan Engineering Laboratory Co., ltd. The items are measurement of resistance and dipping and spray observation.
Now let us cosider bow spray. The result is
shown in figure 6. We can find easily there is no steep wave which causes bow spray. This trend is the same as previous optimum hull forms. Figure 7 shows bow spray observation of series 64 model no.4793. Though she cruises at lower speed, there is high steep wave. Obseving these two photoes. we can find it is very effective to use spray free con-dition which is eq. (8) in spite of her ship type wa-ter plane area. There is a little wave such as mus-tache from the tip of present hull form. This is be-cause the entrance angle is smaller than Kelvin
an-computed trim
£ computed rise
measured trim trim of series 64 no.4793
measured rise
L
rise of series 64 no.4793
rrieasured residual resistance measured residual resistance
of series 54 no.4793
computed residual resistance
Fig. 7 Observation of bow spray
(series 64 no.4793, Fn 0.447) Fig. 9 Residual resistance
trim
5 x rise o
.0.02
0.2 0.3 0.4 0.5
froude number
Fig. 8 Trim and rise
-0.5 04
0.2 0.3
r
O-1
p
O
PRESSuRE DISTRIBUTION (BOOT PLRN(
N2.500 LRIISUR- -128 PIIRI- .1980
-1
PRESSURE DISTRIBUTION (BOUT PLRNI
N-2.500 LRflSQR- .128 PTIRX. .1415
Fig. 11 C and C curves of present hull form and
series 64 model no.4793
gle. It is necessary to round the tip of the water plane area in order to vanish this wave.
For a next step, let us consider trim and rise. It
is shown in figure 8. Trim is larger than that of
series 64 model no.4793. This is because cetner of floatation goes to aft end. Rise is also large. This is because diminished spray causes low bottom pres-sure around the bow. We can compute trim and rise at only design speed. We cannot succeed in
solv-ing integral equation (1) perfectly.
Lastly let us consider residual resistance.
lt is
shown in figure 9. Though we attempt to reduce wave making resistance using influence function,
ex-perimental data is very negative. Why is not wave making resistance reduced ? Firstly we verify that
pressure can be optimized truely with influence
func-zion. Figure 10 shows this result. We compute opti-mum pressure with both calculus variations and in-fluence function on ellipse water plane area which ad the same displacement as present hull form.
\VO different methods shows almost same result.
-1 0
X
PRESSURE DIETRIBUTION (PROFILE!
K-2.500 LR(1BOR- .128 PuRI- .1335
PRESSURE DISTRIBUTION (PROFILE)
c-2.500 LRI1SDR- .128 P(1RX- .1415
Fig. 10 Pressure optimizations using calculus variations and influence function on same ellipse water plane area
33-Using influence function Iterative number is 9.
Using calculus of variations Trans. series number is 1. Longi. series number is 4.
So we think present method is effective. But we can not obtain the complex pressure distribution which is obtained easily with calculus variations,
though we carry out the iterative computration
many times. This fact shows that it is effective to use influence function when we improve pressure
dis-tribution slightly.
Secondly we check C,, and C, curves. It is shown in figure 11. C. curve of present hull form is the same as that of series 64 model no.4793. But C,, curves are different. C,, curve of present hull form is affected significantly by C curve. We think wa-ter plane area is bad. This fact shows it is impor tant to optimize water plane areá before we try to optimize pressure distribution. We think these are the reasons why wave making resistance is not
re-duced.
Finally we computes residual resistance. As de-scribed before, we cannot solve integral equation (1) perfectly. Though the computation is not per-fect, there will be some swell-up potential.
Be-cause if we add water head resistance to wave mak-ing resistance vhich is obtained from statical
pres-sure distribution, the
results shows about the
same order as measured residual resistance. But the detail is not yet known.
Conclusions
The present work shows some results shown as bellow.
If we design optimum hull forms using pressure
making resistance, she cruise at a draft which is
almost equal to still water draft.
The spray free hull forms are obtained on the con-dition that pressure is zero at the edge of water plane area even if her water plane area becomes slender at fore part. And such water plane area is very effective for designing bow shape
properly
-It is convinient to use influence function in order to improve pressure distribution. And it is neces-sary to use calculus variations in order to obtain
optimum pressure distribution.
It
is very important to optimize water plane
area inorder to search more optimum hull
forms. But we don't know yet how the water
plane area is optimized.
Acknowledgement
The authors express their
gratitude to Mr.
Kiyoshi Morohoshi and Mr. Shigeo Watanabe, Sum-itomo Heavy Industries. They gave us the chance of this experiment and very effective discussions.
Bessho, M., An essay to the theory of ship
forms in view of the wave-making resistance
theory, Memoirs of the Defense Academy,
Japan, vol.111, No.1, 1963, 19-38
Bessho, M., On the fundamental function in the theory of the wave-making resistance of ships,
Memoirs of the
Defense Academy, Japan, vol.111, No.1, 1964, 19-38Bessho, M., On the two-dimensional theory of wave resistan of a shallow draft ship,
Sci--64
entific and engineering reports of National De-fense Academy of Japan, 29, 1991, 87-93 Bessho, M., On the theory of wave resistance of a shallow draft ship, Scientific and engineer ing reports of National Defense Academy of Japan. 29, 1991, 95-101
Bessho, M., On a consistent linearized theory of the wave-making resistance of ships, 16th Georg Weinbium Memorial Lecture, 1993
Bessho, M., Sakuma, S., Trim and sinkage
of two-dimensional shallow draft ships, Jour-
-nal of Kansai Society of Naval Archi-tects,218, 1992, 69-78
Bessho, M., Sakuma, 5., A numerical solu-tion of three dimensional gliding plates,
Jour-nal of Kansai Society of Naval Architects,
218, 1992, 79-91Bessho, M., Sakuma, S., On the theoretical
study of the residual resistance of high speed shallow draft ships, Journal of Kansai Society
of Naval Architects, 219, 1993, 43-47
Bessho, M., Sakuma, S., A numerical solu-tion of three dimensional gliding plates, Pro-ceeding of FAST'93, 1993, 963-974
10' Bessho, M., Sakuma, S., A theoretical study on high-speed ship hull forms (optimization problem), Journal of Kansai Society of Naval Architects, 221, 1994
11) Hugh. Y. H. Yeh, Series 64 resistance experi-ments on high - speed displacement forms, Ma-rine technology, vol.2, No.3, 1965, page 248