Documente Academic
Documente Profesional
Documente Cultură
Hiroshi Isshiki
Institute of Mathematical Analysis,
Osaka University,
Yamada-oka 2-1, Suita,
Osaka, 565-0871, Japan
Introduction
From ancient times, human beings have used ships with propulsion devices. Some devices use hydrodynamic drag and some use
hydrodynamic lift. As well, some creatures moving in air and
water also use drag and/or lift for their propulsion. Birds use wind
energy by using their own wings to fly up and down rapidly.
Fishes also take advantage of fluid flow energy by using their own
fins to move around in the water. The behaviors and movements
of their wings and fins are useful references for naval architects to
develop new devices to decrease ship motions in waves and to
utilize wave energy for an auxiliary propulsion device. As one
example, to stabilize a roll motion in waves, the anti-rolling fin
system has been utilized for a long time, which is equipped at the
midship. To make more advanced ocean-going ships, it is useful
for naval architects to investigate further the mechanics of flying
and swimming of creatures.
General Overview
The state of the art in bow wing research, which has been
confirmed by experiments and/or alternative verifications, is summarized in 22 topics, which follow. The theory behind much of
this research is laid out in Sec. 2.
1.1 Ships in Sea. Generally speaking, a ships resistance increases in waves, as compared with calm water. This phenomenon
is well known to naval architects as the resistance increase or
added resistance of a ship in waves. Waves also produce ship
motion; namely, a ships body receives increased kinetic energy.
In winter, a season punctuated by increased large waves, this kinetic energy greatly increases. Utilization of this kinetic energy
for ship propulsion, i.e., converting this kinetic energy into ship
propulsion, would serve as a great energy-saving measure. In the
North Pacific and North Atlantic Ocean, there exists enormous
natural wave energy, especially during the winter season. This
wave energy causes various kinds of ship motion which sometimes lead to serious damage of a ship, with sinking being the
worst case.
1.2 Examples of Wave Devouring Propulsion. In 1895, Linden 1 filed a British patent on a ship which has a flexible wing at
the bow and moves by utilizing wave energy British Patent No.
Transmitted by Associate Editor W. Shyy.
Fig. 1 Picture of Autonaut designed by Linden from Pearsons Magazine, Dec. 1898
theory the free-surface effect is neglected. Grue et al. 21 extended Wus theory and included the free-surface effect. Their
theories were comparable with the experiments of Isshiki et al.
22.
Their experimental results and theoretical considerations are
given for the scenario of a wing placed alone in waves, as shown
in Fig. 3. In this case, wave energy is absorbed and converted into
propulsion by the wing itself. The relationship between wave
length and free-running speed is measured, as shown in Fig. 4.
The results seem to agree.
When the wavelength is 3 m and height is 12 cm, i.e., the wave
slope is 0.04, the maximum speed, 0.65 m / s, is measured. The
theory provides a fair explanation of the experimental results.
1.6 Effective Wing Form. Naito et al. 23 researched the
optimum form of the wing with the simple surface panel method
based on SQCM Kataoka et al. 24 in still water. Calculations
were performed for a wing with an aspect ratio 4 NACA.0012.
Because two wings are set up symmetrically at both sides of the
bow, the performance calculation for wings was determined by
using the mirror-image method applied to one side of the model.
The number of panels on the wing surface is 780. The number of
panels in the span direction is 13. The number of panels of chord
circumference is 60.
1.6.1 Tapered Wing. From a structural-strength viewpoint, the
tapered wing, in which the chord length of each section becomes
shorter from the point of the wings fit with the ships body toward
the wing tip, is advantageous. They compared this to the symmetrical wing rectangular wing, in which the center chord length
is the same as the length at the point of wings fit with the ship
body. As the taper ratio is reduced, the lift coefficient increases
1.6.2 Swept-Back Wing. The swept-back wing, which is defined as having its wing tip pointing downstream along the direction of the chord, was examined and the desired result achieved.
The taper ratio was about 0.4 and the swept-back ratio was about
0.1. Even if the optimum tapered and swept-back wings are applied, the increase of the lift coefficient is only about 2%. Therefore, at this initial stage of experimental and theoretical investigation, a designer may use the rectangular wing in assessing the
effect of the wing on the propulsive performance of the ship with
wings in waves. The optimum wing form is shown in Fig. 5. This
is a tapered and swept-back wing, which is recommended from a
theoretical point of view. A lift coefficient of this wing is slightly
larger than that of a rectangular wing. This wing is used in the
experiments mentioned below. The model is equipped with a wing
on both sides, where the wing is at a forward position of 2.072 m
of the center of gravity and at a depth of 0.16 m.
That is, even with the fixed bow wing, the wing effect appears to
have two effects, one on ship motions and the other on ship resistance. Later, the investigated results on the controlled bow wings
are described. The resistance increase of the ship with the wings is
smaller than the one without the wings. At the mean wave period
of 9 s actual ship scale, where ship motions are large, the performance of the ship with the bow wings is superior.
The effectiveness of bow wings is illustrated in Fig. 11, most
notably in the frequency region where the relative motion between
the ship and the wave is large. Three different wings shown in
Fig. 12 are used in these experiments.
1.10 Effective Position of the Wings. In Fig. 13, relations
between the longitudinal position of a wing and the thrust generated by a wing are shown for various wavelengths. The vertical
axis shows the thrust, and the horizontal axis shows the wing
position. Figure 13 is noteworthy, as it demonstrates that the best
position to obtain thrust is, generally, in front of the bow, and the
worst position is at midship, where thrust can hardly be obtained.
Furthermore, the results indicate that the optimum wing position
moves forward as the wavelength increases.
Naito et al. 28,29, additionally indicated that positioning the
wings in front of the bow also decreases the resistance increase in
waves. This is presented in Fig. 14.
1.11 Stall. When a ship speed is low, the vertical component
of inflow velocity to the bow wings becomes large in comparison
with the horizontal component. As a result, the fluctuation of the
attack angle easily exceeds the stall angle of 15 deg. The attack
angle of wings in wave can be estimated by measuring the wave
and calculating the ship motions. When the attack angle regularly
exceeds 15 deg, the wing is thought to be in the stall. In Fig. 15,
the attack angle of the bow wings is shown as reaching the 15 deg
maximum in the range of 0.9 / Lpp 1.6. To avoid stall, intermittent wing control is recommended so that when the attack
angle does not exceed the stall angle the wings are not be con-
trolled, and when the attack angle exceeds the stall angle the
wings are controlled in order to reduce the attack angle. To calculate the attack angle, the bow bottom pressure method is used,
as introduced below.
1.12 Passive Control of Wings. The stall effect has a significant influence on the thrust; to avoid stall, methods for controlling
the bow wing are needed. Isshiki proposed a spring system that
added a restoring moment to the wing, as shown in Fig. 16.
From the results in Fig. 16, we know that an optimum constant
of moment spring exists. This fact indicates the necessity of considering stall effect when correctly estimating thrust generated by
a wing, and when studying active wing control. If the optimum
coefficient of the moment spring can be determined, then passive
control is recommended, because an active control system is typically expensive. The results show that, while wing thrust is affected by moment spring stiffness, resistance increase is not. It is
therefore sufficient to consider only the effect of moment spring
stiffness on thrust when determining the moment spring coefficient. Figure 16 indicates the existence of an optimum moment
spring stiffness for thrust generated by a wing. The horizontal axis
is a nondimensional coefficient of moment spring. From this figure, the value of Kp can be decided as about 0.5, which depends
on wave frequency. However, it must be noted that, because an
irregular wave includes many wave frequency components, it is
difficult to determine the stiffness value uniquely. This problem is
discussed below.
Fig. 17 Total resistance ratio of theoretical to experimental results for various wings. Fn = 0.15 Rn = 3.30 106: Initial attack
angle= 0 deg. Case A = without wing, Case B = NACA0012,
chord= 0.1 m, span= 0.2 m, Case C = NACA0012, chord
= 0.082 m, span= 0.246 m, Case D = NACA0024, chord= 0.1 m,
span= 0.2 m.
Fig. 16 Effect of moment spring stiffness on thrust generated
by the wing and resistance increase of ships in waves Fn
= 0.2, / L = 1.2. The wing is set in front of the bow, Kp is the
coefficient of spring stiffness.
angle, and the wing installation position can be examined. Therefore, an optimum condition of the wing can be predicted because
there is little wing resistance in still water.
Miyata et al. 30 and Ogiwara et al. 31 have succeeded in
reducing the resistance of a ship in calm water by attaching a wing
to the hull, though in their case, the wing was attached to the hull
just above still water level near the bow, which is fairly different
from that of the bow wings discussed in the WDP case.
Currently, the optimum position of bow wings that do not increase total resistance has not been found. In order to find the
wing position, where the drag of the wings is small and the efficiency of WDP is high, further research must be done.
1.14 Measurement of Pressure Working on the Wing. Nine
pressure gauges were arranged on the wing of NACA.0024, to
measure the vertical force acting on the wing. Figure 19 shows the
time history, at position A, of the differential pressure between the
upper face of the wing and the lower face / Lpp = 1.0. By using
the attack angle obtained through previous experiments, the pressure is calculated by considering the effect of the unsteady influence. This calculated result is represented by the solid line in the
figure. The dotted line indicates the pressure value corresponding
to the attack angle 15 deg. The value measured on the boundary
of this dotted line falls into disorder, and shifts more than the
calculation values. That is, the stall occurs in the range of the
attack angle exceeding 15 deg.
A0 = exp
Fig. 19 Comparison of the time history of the measured pressure acting on the wing between calculation and experiments
Fn = 0.2, / Lpp = 1.0
1 C2d Cv2
+
2 2d v2
1.1
M wK mV t
ct
1.2
Here
Km =
1
,
1 + M w/M
2
c
8
Mw =
1
Vt
= 0.144 103
t
c
1.3
0.673
1.4
M,M w,c: mass, added mass, and chord length of flat plate
Vt: entering velocity into water
1.5
cb2
2
1.6
M t = Pmax
Here, b is the span length and the bow wing slamming condition is
defined as when the bending moment at the fitted part of the wing
becomes equal to the yield stress of mild steel. The critical value
of relative speed Vt max that generates this bending moment can be
determined by the next formula in conjunction with Eq. 1.2.
Vt1.673
max =
c0.673
Mt
M wK m
1.7
dead space of the bow is not very wide for stowing wings. Kawatani 33 has proposed one system for stowing wings in the narrow bow portion, which is an improved model of the anti-rolling
fin system. In this system, the starboard wing is piled up against
the portside wing, as shown in Fig. 22.
1.19 Evaluation of Wings Effect in Actual Sea. In Isshikis
study 25, the JapanNorth Pacific route in winter is supposed
as the sea condition, and the experimental results shown in Fig. 6
are used.
A wave of period 8 s is chosen as a design wave, since the
power of this wave is the highest at 28.0 ps/ m or 20.6 kW/ m.
The wavelength and height of the wave are 99.8 and 3.13 m,
respectively, and the frequency of occurrence is 26.8%. According
to the model test results, wave energy is utilized efficiently when
the ship length is smaller than the wavelength.
The effective horse power EHP, in calm water, of the ship
with bow wings ship length is 80 m is compared with the EHP
of the ship without wings in the design wave. For example, if the
output of the main engine and the propulsive efficiency are assumed to be 700 ps and 0.7, respectively, the speed of the ship
with the wing in the design wave becomes 12.1 kts. On the other
hand, the speed of a ship without the wing is 11.4 kts in calm
water and 7.1 kts in the design wave. This suggests that the effect
of the wing is significant.
Tests were conducted at sea, using a small fishing boat 15.7 m
in length. The principal particulars of the wing were 1.05 m in
chord, 3.8 m in span equal to the ship breadth, and 1.65 m in
submergence, as described in NACA 0015. The pivot point of the
wing is placed at 1 / 4 of the chord length, namely at the lift center
in the steady state. Although the theory overestimates the effects
of the wing, it explains the experimental results qualitatively.
A large wing at the bow induced no serious abnormal phenomena, and no increase of turning radius was observed. The wing
support system was composed of a hydraulic lifting device and a
flexible wing support by a hydraulic spring. The hydraulic cylinder and accumulator altered the spring constant. In spite of unfavorable weather conditions small waves, a free-running speed of
2.5 kts was obtained. If this velocity is scaled up to a ship of 80 m
in length by Froudes law, this corresponds to 5.6 kts. In freerunning tests, the ship automatically adjusted its direction to
match the wave direction. Furthermore, speed increased about
0.75 kt in following seas. Effects on the reduction of pitch were
also verified.
Transactions of the ASME
1.20 Ship Motions Caused by the Forced Bow Wings Motion in Still Water. Experiments of the forced wing motion were
carried out in still water 34 to confirm the relationship between
wing motion and ship motion. Namely, Rz, R, and
R shown in Sec. 2, Eq. 2.14 must be confirmed. In these
experiments, the wing amplitude 0 was 10, 15, and 20 deg,
and the phase difference between both side wings was 0deg and
180deg; i.e., in the former case heave and pitch motions occurred, and in the latter case roll motion occurred.
The frequency response functions of heave and pitch are shown
in Figs. 23 and 24. The ship motions are almost a linear function
of the wing motions in the range where the wing amplitude is
relatively small. In Fig. 25, the results of the rolling motion are
shown with nondimensional vertical and horizontal axes. There is
sured resistance increase, and the attack angle are shown in Fig.
27. It has been established that there are strong relationships
among wing motion, wave motion, and ship motion, that reduce
the resistance increase of the ship. That is, the phase is important
so that if one misses phase control, contrary to natural expectation, ship resistance increases even more.
The attack angle shown in Fig. 27 is obtained indirectly by
t =
vWt zt lW t
t + t
U
U
1.8
Fig. 27 Relationship among controlled bow wings, added resistance, and attack angle in head waves. Measured added
resistance added resistance without the wings in waves
thrust generated with the wings and other effects.
Fig. 29 Relationship between the bottom pressure and the attack angle Fn = 0.284, head wave
Fig. 30 Comparison of attack angle of bow wings between calculated and quasi-experimental values
1
vWt zt lW t
LSt cbU2CL
t + t
2
U
U
2.1
Here, the dot above a variable indicates time differentiation. Naturally, when t = 0, the lift of the fixed bow wings is obtained.
Moreover, CL is the lift inclination coefficient by which the induced drag is considered. It is defined as
Fig. 32
C L =
2
.
1 + c/b
2.2
In the case of two-dimensional wings i.e., where b tends to infinity, CL = 2. The first term of the right-hand side of Eq. 2.1
shows the lift obtained when the wing advances in waves at a
constant inclination. The second and third terms show the lift
obtained when the wing advances in still water while heaving and
pitching. Then, the varying lift Lunt is given by Eq. 2.3,
which is based on Eq. 2.1 by multiplying each lift with its corresponding oscillation influence coefficient COW, COH, and COP,
respectively.
zt lW t
1
vWt
COH
Lunt = cbU2CL COW
2
U
U
COPt + t
2.3
Fig. 33 Coordinates of the bow wing
k
COW = J0 + J2 i + J0 iJ1Ck
2
k
COH = i + Ck
2
COP =
k =
c
,
2U
2.5
k
k
i+
i + 1 Ck
2
2
2c
2g
2.4
+ 2Lunt + FSFt
2.6
: angular frequency
2.9
dFPt = zt lWt
dFPt = zt lW t + Ut
dFPt = zt lW t + 2U t
2.10
where M W, mW, and CF denote the mass, the added mass, and the
damping coefficient of one wing, respectively. Therefore, FSF is
obtained using Eqs. 2.3, 2.9, and 2.10.
Tt = Ltsint + t + t
Dtcost + t + t
1
Dt
= CLDt cU2 sin t + t + t tan1
2
Lt
2.7
where
CLDt =
Lt2 + Dt2
cU2/2
2.8
For a small attack angle t one can use Eq. 2.3 as the lift
Lt in Eq. 2.7 and neglect the drag because the lift is much
larger than the drag Lt Dt. When t is small, one
can use Ck, but when t becomes larger, the oscillating wing
stalls a phenomenon called dynamic stall. When dynamic stall
occurs, the wing performance changes; see Naito et al. 26 for
details of wing performance after dynamic stall. If one can control
the attack angle so that a stall does not occur, one can produce a
larger thrust.
In this theory, the distinction between the wing motion t and
the attack angle t must be noted.
2.2 Motions of the Ship. The force caused by controlled bow
wings influences ship motions. Here, FSF is the force of the controlled bow wings acting on the ship. The equation for the bow
wings force is
264 / Vol. 58, JULY 2005
+ cbUCLCOWvW + cbU2CLCOP
2.11
If the motions of the two bow wings are controlled with a phase
difference, an unsymmetrical motion rolling and yawing occurs.
This fact indicates that by using the bow wings one also can
control the rolling motion. In this research, yawing motion is neglected. The coupled motions of heave and pitch, and the single
motion of roll are considered.
Here, by setting up the equation of ship motions caused by the
wings, the equations of ship motions with controlled bow wings
are shown.
Heave and Pitch
A11z + A12z + A13z + A14 + A15 + A16 = Fz FSF 2.12
A24 + A25 + A26 + A21z + A22z + A23z = M + lWFSF
2.13
Roll see the coordinate system in Fig. 33
a37 + a38 + awd + a39 = m + awf t A f U2CLay Gt
2.14
where
awf t = UCLy Gaekd f eikL/2 cos
awd = UCL y G
cy cy c + B/2dy c
and where the coefficients A11 A26, a37 a39 and the external
force and moment due to waves FZ, M , and M can be obtained
by the strip method. One can calculate ship motions with the
controlled bow wings by adding Eq. 2.11 to Eqs. 2.12 and
2.13.
, and to be
Here, one can define z, , ,
Fourier transforms of zt, t, t, t, and t. In the frequency domain, the motion of a ship with controlled bow wings is
R z R z
= R R
R R
2.15
the controlled bow wings is presented. The figure shows the desired value of the attack angle. If wings could be controlled to
match expectation, a larger thrust could be obtained as shown in
Fig. 35. This is a significant difference, and indicates the effectiveness of the controlled wings.
Based on these calculations, a simple concept of bow wing
control is proposed, as shown in the block diagram of Fig. 36. The
upper line of the diagram shows that the input is the bottom pressure pt and the output is the attack angle t of the fixed wings
estimated from the previous research results, which are shown in
Fig. 29. The lower line shows that the input is the same and output
is the desired attack angle t. The control signal, which makes
the difference of two output signals zero, is put into the controller
G. The frequency response functions are labeled as Hp
and Hp, respectively.
The wings are controlled such that the actual wing motion must
always coincide with the desired wing motion. The joint frequency response function Hp of this whole system is written
as
Hp = GHp Hp
3.1
The impulse response function hpt of the system i.e., the input
of the bottom pressure pt and the output of the desired attack
angle outt, is obtained by Fourier transformation as
hpt = J1Hp
3.2
outt =
hptpt d
3.3
Fig. 39 Short-term prediction of the mutual relationship between nominal speed loss and significant wave height. The
ship does not have a wing stowing system.
Simulated results are shown in Fig. 38. Pitch motion is nearly the
same between the fixed wing and the controlled wing, but the
attack angle became larger. Therefore, the controlled bow wings
generated larger thrust than the fixed wings shown in the lowest
figure in Fig. 38.
Our expectation for the controlled bow wings is great based on
these findings.
Until now the bow wing system has had two major disadvantages:
1. Wave impact load negatively impacting the wings and the
resistance increase of the wing;
2. Economical advantage of bow wings has not been
demonstrated.
For the first point, the wing stowing system is proposed. Conceptualization of this system is not difficult, and would depend on
the ship owners decision.
For the second point, we offer the following example: When
ships navigate in the North Pacific Route from Japan to the West
Coast of the U.S.A., by using information of sea conditions on the
route and knowledge of the bow wing system, we can predict a
significant economical advantage of the fixed bow wing system in
the area of nominal speed loss, specifically involuntary speed loss,
as shown in Fig. 39. In lower significant wave height, the speed
loss of a ship with wings is larger than a ship without wings as a
result of wing resistance. On the contrary, in increased or higher
significant wave heights, the ship with wings shows advanced
performance.
Since this result can easily be transformed into fuel oil consumption, the financial merit may be evaluated.
Conclusions
1. The bow wing system in waves, either the fixed or controlled system, produces the effect of an auxiliary propulsion
and serves as a reduction damper of ship motion.
2. In the case of a fixed system, a designer must consider the
most effective sea condition.
3. For practical application, the stored system of bow wings is
recommended, in order to avoid the negative effects of the
wing system.
4. The theories introduced here qualitatively explain the effects
of wavelength, wing form, wing size, wing position, spring
Transactions of the ASME
stiffness, control on thrust, resistance increase, and ship motion. Future work should include quantitative estimation of
the effects of the bow wing system.
5. The effects of controlled bow wings on ship motion and
propulsion are discussed. More detailed investigations are
desired.
6. The wave energy conversion system where wave energy is
absorbed from ship motion as kinetic energy, and the kinetic
energy is then converted into propulsion through the thrust
generated by a wing, seems to be ineffective when the
wavelength is shorter than the ships length. In order to apply this kind of wave energy into propulsion of a large ship
in waves, an introduction of new technology, such as an
active control system for a wing, may be necessary.
According to the reviewed results, the possibility of wave energy utilization in ship propulsion seems promising.
Currently many ships have been fitted with anti-rolling device
systems. If this system could be remodeled into a bow wings
system, the effect would not only decrease roll, but also decrease
heave and pitch and reduce speed loss in the ship.
Nomenclature
fluid density
U ship speed
L pp ship length
lW distance between the C.G. of a ship and the
rotation center of a wing
dW depth of wing from the still water line
c wing chord length
b wing span length
k wave number
k reduced frequency
0 wave frequency
encounter wave frequency
t 0eit+kx: incident waves
Wt i0ekz0eit+kx: vertical orbital velocity
zt z0eit+s: heave motion
t 0eit+s: pitch motion
t 0eit+: roll motion
t 0eit+: wing motion
t 0eit+: attack angle
dFPt vertical motion at the bow
References
1 The Naval Architect, 1979, Wave Energy for Propelling CraftNothing
New, pp. 239.
2 Jakobsen, E., 1981, The Wing Propeller, Wave Power for Propulsion, 2nd
International Symp. On Wave & Tidal Energy, BHRA Fluid Engineering, pp.
363368.
3 Terao, Y., 1982, A Floating Structure Which Moves toward the Waves Possibility of Waves Devouring Propulsion, Journal of KSNAJ, 184, pp. 5154.
4 Maruo, H., 1963, Resistance in Waves, 60th Anniversary Series, SNAJ, 85,
pp. 67102.
5 Ulysses, A. P., 1958, A Study of the Sea Behavior of A Mariner-Class Ship
Equipped With Antipitching Bow Fins, DTMB Report 1084.
6 Abkowitz, M. A., 1959, The Effect of Anti Pitching Fins on Ship Motion,
Soc. Nav. Archit. Mar. Eng., Trans., 62, pp. 210252.
7 Stefun, G. P., 1959, Model Experiments With Fixed Bow Anti-Pitching Fins,
J. Ship Res., xx, pp. 1423.
8 Becker, L. A., and Duffy, D. J., 1959, Strength of Anti-Pitching Fins and Ship
Motions Measured on USS COMPASS ISLAND EAG153, DTMB Report
1282.
Shigeru Naito graduated and gained his Doctorate of Engineering from Osaka University in 1975. He has
worked since 1975 as a Lecturer and Associate Professor, and for the last 10 years as a Professor of Marine
Systems Engineering, Department of Naval Architecture and Ocean Engineering, Graduate School of Engineering, Osaka University. Naitos principal current research topics are seakeeping performance of ships,
added resistance of ships in waves, utilization of wave energy, and wave generation and absorption. He has
been a committee member at several international conferences such as the International Towing Tank
Conference (ITTC), International Society of Offshore and Polar Engineering (ISOPE), etc. And, he organized several international conferences in his research field. He became President of the Japan Society of
Naval Architects and Ocean Engineers this year.
Hiroshi Isshiki graduated from the Department of Naval Architecture, University of Tokyo in 1965 and
obtained his doctoral degree from the same university, carrying out research on Variational Principles in
Applied Mechanics in 1972. After working as a visiting scholar at College of Engineering, Seoul National
University, Korea, he joined the Technical Research Institute, Hitachi Zosen Corp. in 1973. He established
IMA. (Institute of Mathematical Analysis) in 2001. He is currently interested in the theory of high accuracy
long baseline kinematic GPS positioning, algorithms for measuring sea bottom crust movement, and the
theory of hovering flight of a dragonfly.