Sie sind auf Seite 1von 16

SHIP BOW WAVES

Delhommeau Gerard, LMF-UMR 6598 CNRS, Ecole Centrale de Nantes, France


Noblesse Francis, David Taylor Model Basin, NSWC-CD, West Bethesda, USA
Guilbaud Michel, LEA-UMR 6609 CNRS, Université de Poitiers, France

SUMMARY
An analytical and experimental study of the bow wave generated by a ship in steady motion is reported. Simple analytical
expressions are given for the height of water at the stem, the height of the bow wave, the distance between the crest of the
bow wave and the ship stem, and the shape of the bow wave. Furthermore, a simple criterion that defines if a ship in steady
motion generates an unsteady bow wave is given. Finally, a simple, fully-analytical, nonlinear theory that explicitly defines
an overturning ship bow wave in terms of main design parameters is given. In spite of their remarkable simplicity, these
analytical results are in reasonable agreement with experimental observations and measurements for a number of non bulbous
hull forms with wedge-shaped bows, and for a rectangular flat plate towed at a yaw angle and a heel angle.

1. INTRODUCTION A last question-of considerable importance for practical


applications to ship design is: what is the relationship
Free-surface flow due to a ship that advances at constant between the main characteristics of a ship bow wave (wave
speed along a straight course in calm water, i.e. a ship in height and location, steadiness, geometry of overturning bow
steady motion, is considered in this study. More wave) and the main design parameters (ship speed, draft,
precisely, the study considers ship bow waves, arguably waterline entrance angle, and flare angle) that define a ship?
the most conspicuous, complex, and important feature of In other words, how do things work?
free-surface flows about a ship. Elementary questions The object of this study is to provide simple fully-analytical
related to a ship bow wave are: what is the height of the answers to the basic questions listed above.
bow wave? What is the distance between a ship stem The analytical results given in the study are compared with
and the crest of the bow wave? What is the height of the experimental observations and measurements for the wave
water at ship stem? What is the shape of the bow wave? generated by an inclined flat plate. A relatively large set of
experimental measurements was available for the analysis
A ship in steady motion is usually assumed-notably for reported in this study as a result of cheap experiments
numerical-calculation and analytical purposes-to generate performed by the authors with a rectangular flat plate that
a steady bow wave, but this is not always true. Indeed, was towed at various immersion depths D, speeds U, yaw
common observations show that a ship in steady motion angles α and heel angles γ . The usefulness of this substitute
can generate an unsteady bow wave. More generally, for a systematic series of ship models is validated by the
steady motion of a body through a fluid at rest does not results. This cheap substitute to a ship bow form made it
necessarily result in a steady flow; a classical example of possible to perform measurements for a broad range of the
unsteady flow generated by steady motion of a body is critical parameters U, D, α and γ ; and to derive simple
the von Karman vortex sheet that can be observed (under “cause-and-effect” relationships between basic design
some conditions) behind a bluff body. Thus, parameters (U, D, α, γ) and flow features (bow-wave height
basicquestion is: when does a ship in steady motion Zb and water height ZS at ship stem) of importance for ship
generate an unsteady bow wave? design. In spite of their remarkable simplicity, the analytical
Common observation of the bow wave generated by a results are in reasonable agreement with experimental
ship in steady motion, in the “steady bow-wave regime,” observations.
shows that an overturning thin sheet of water is typically
generated by a ship bow. Accurate prediction of this 2. EXPERIMENTAL STUDY
highly nonlinear, turbulent, two-phase, wave-breaking
flow is problematic. Furthermore, CFD methods suited to A series of experimental observations and measurements
compute such complicated free-surface flows (e.g. VOF were performed in the towing tank of Ecole Centrale de
method) may be overly complicated and not efficient Nantes with a rectangular flat plate 0.782 m long and 0.5 m
enough for many practical applications, notably for high immersed at a draft D=0.2 m or 0.3 m. The flat plate
hydrodynamic design at the preliminary and early stages. was towed at speeds U=1.5, 1.75, 2., 2.25, 2.5 m/s with the
Thus, the question here is: can the main features of the draft Froude number varies from 0.6 to 1.8. The values of the
overturning thin sheet of water that is typically generated incidence angle are αE =10-15-20-25-30-45-60-75 and 90°
at the bow of a fast ship be predicted without resorting to and the values of the flare angle are γ=0-10-15-20, 30 and
overly complex numerical methods? 40°. For each run, numerical pictures have been taken. In
order to obtain a good accuracy for the shape of the contact

Keynote 1 Delhommeau
line, circular points of 5 mm diameter with horizontal
and vertical spacings of 2 cm are located on the plate. All
values of incidence angles and velocities are summarized
on Fig. 1. Each point on this figure corresponds to 6 flare
angles. The color pictures (Fig 2a) are first transformed
into a grey scale, then corrected for geometric distorsion
and projection by a model of second order camera. The
result is a 2D picture (Fig 2b) which can be directly
digitalized. Accuracy and repetitivity of measurements
are obtained by 2 series of runs for α=γ=20° and U= 2
m/s where several pictures are taken during each run. The
first series is done when water is at rest for 3 hours and
the second one 20 mn later. The shape of the bow wave
is given in Fig. 3 and numerical results in table 1. It can Fig. 2b: 2D picture ready for digitalization
be seen that accuracy of measurements is better on wave
height than on location of the wave crest.
140
130
2
120
x x x 110

x x x UNSTEADY 100
1.5 90
x x x x x
80
z (mm)

x x x x x x
70
x x x x x x x
FD

1 60
x x x x x x x x
50
40
4.4 tan(αE)/cos(αE)-1
30
0.5 Exp Nov 2005 α=γ=20°; U=2 m/s
x Exp Jun 2006 20 traits ou traits plus symboles creux : eau calme (série 1)
Exp Dec 2006 symboles eau non calme (série 2)
10
0
0 0 50 100 150 200 250 300 350 400 450 500 550 600 650 700
0 10 20 30 40 50 60 70 80 90 x (mm)
αE (deg)
Fig. 3 : Repetitivity of measurements
Fig. 1: Summary of experiments on a flat plate.

mm Series1 Series2 All


zmax 129,40 129,54 129,46
Min 128,20 127,82 127,82
Max 130,33 130,52 130,52
σ 0,67 0,84 0,77
X(zmax) 151,20 138,36 145,21
Min 145,89 130,84 130,84
Max 156,96 150,67 156,96
σ 4,61 6,01 8,21

Table 1 : Accuracy of measurements


Fig. 2a: Original color picture
3. BOW-WAVE HEIGHT

A simple analytical expression for the height (above the


mean free-surface plane) Zb of the bow wave generated by a
ship that advances at constant speed U in calm water is given

Keynote 2 Delhommeau
in Noblesse et al. (1) . This expression directly defines zb
= Zb g/U2, where g is the acceleration of gravity, in terms
2.4
of the ship speed U, draft D and waterline entrance angle
2αE. Specifically, elementary fundamental theoretical
2.2 / (1+FD)
αe = 10 °
considerations (dimensional analysis, and asymptotic 2 αe = 15 °
αe = 20 °
behaviors in limits αE → 0 , D → 0 and D → ∞) yield

(Zb g / U 2) cos αE / tan αE


Z g C Z tan α E 1.6
zb = b = (1a)
U 2 1 + FD cos α E
1.2
with FD = U / gD (1b)
This expression is validated by comparison with 0.8
measurements, which determine the constant CZ as
C Z ≈ 2.2 (1c) 0.4
The simple analytical expression (1) is in excellent
agreement with experimental measurements for 0
0 0.2 0.4 0.6 0.8 1
wedge-shaped ship bows. Expression (1) is also in good FD / (1+FD)
agreement with measurements for the Wigley hull and
the Series 60 model, and similar ship-bow forms, like Fig. 5: Height (Zb g/U2) cosαE / tanαE of bow waves for a flat
Larrarte struts (5), especially if the simple procedure plate in heel and drift. The solid line is the approximation
given in Noblesse et al. (1) is used to define an effective 2.2/(1 + FD)
draft D and waterline entrance angle 2αE. This agreement
between experimental measurements and theoretical 4. UNSTEADY BOW-WAVE CRITERION
predictions can be verified in Fig. 4 , where the
normalized bow-wave height (Zb g/U2 ) cosαE / tanαE is For a steady free-surface flow observed from a Galilean
depicted as a function of the draft-based Froude number system of coordinates (X Y, Z) attached to a ship that
FD given by (1b) . Experimental measurements for eleven advances along a straight path with constant speed U in calm
ship hulls are shown in Fig. 4, where the solid line is the water, the velocity of the total flow (uniform stream opposing
approximation (1). These results are also in good the forward speed of the ship + flow due to the ship) is (Vx -
agreements for a flat plate towed in heel and drift as U, Vy , Vz ) . Here, (Vx, Vy , Vz) is the flow due to the ship.
shown on Fig. 5. Furthermore, the X axis lies along the ship path and points
toward the bow, and the Z axis is vertical and points upward
2.4 with the mean free surface taken as the plane Z=0 . The
x Standing 5° Bernoulli relation
Ogilvie 7.5°
2
Standing 10°
P − Patm U2 (Vx − U 2 ) + Vy2 + Vz2
Ogilvie 15°
+ gZ + (2) =
Karion 20°
ρ 2 2
(Zb g / U 2) cosαE / tanαE

Waniewski 26°
1.6 x Karion 10° applied at the free surface, where the pressure P is equal to
+ Wigley hull
x Series 60 the atmospheric pressure Patm, shows that an upper bound for
x x Larrarte 10°
1.2 x
x Larrarte 20°
the free-surface elevation Z = E is :
x +
x+++
x+++ Eg / U 2 ≤ 1 / 2 (3)
x + ++
0.8 This Bernoulli constraint is satisfied by expression (1) for the
bow-wave height Zb if FDB (α E ) ≤ FD where the function
0.4
FDB (α E ) is defined as
0
0 0.2 0.4 0.6 0.8 1
FD /(1+FD) FDB (α E ) = 0 if α E ≤ α EB (4a )
Fig. 4: Height (Zb g/U ) cosαE / tanαE of bow waves for
2 tan α E
FDB (α E ) = 2C Z - 1 if α E > α EB ( 4b)
eleven ship hulls. The solid line is the approximation cos α E
2.2/(1 + FD)
with α EB = sin −1 ( (C Z ) 2 + 1 − C Z ) ( 4c )
i.e. α EB ≈ 12°30 for C Z ≈ 2.2 ( 4d )

Keynote 3 Delhommeau
2

4.4 tan(αE)/cos(αE)-1
Experiments

1.5
FD

1 3 2

0.5
1

UNSTEADY
Point 1, FD=0.58, α= 20 °,γ=0°
0
0 10 20 30
αE (deg)

Fig. 6: Unsteady bow-wave region defined by the


Bernoulli constraint (3) and the related boundary (4).

Thus, the Bernoulli constraint (3) is satisfied for every


value of FD if α E ≤ α EB , but is only satisfied for a
sufficiently high value of the Froude number FD if
α E > α EB . For FD < FDB (α E ) , the Bernoulli constraint
does not permit a steady-flow solution, except at rest and
unsteady flow must be expected. Expressions (4) show
that a ship with waterline entrance angle 2αE smaller than Point 1, , FD=0.58, α= 20, γ=20 °
approximately 25°, i.e. with a sufficiently fine waterline,
may be expected to generate a steady bow wave at any
speed. However, a ship having a fuller waterline can only
generate a steady flow if the ship speed is higher than the
critical speed (4b). The unsteady flow region defined by
the Bernoulli boundary (4) is shown in Fig. 6.
Fig. 6 also shows three points, identified as points 1, 2
and 3, where experimental observations of the bow
waves due to a rectangular flat plate were made.
Specifically, a rectangular flat plate, immersed at a draft
D = .3 m, was towed at an incidence angle αE and a flare
angle γ with speed U. The incidence angle αE is equal to Point 2, , FD=0.87, α= 20 °,γ=0°
20° for points 1 and 2, and to 15° for point 3 . The speed
U is equal to 1 m/s (FD =0.58) for point 1, and to 1.5 m/s
(FD =0.87) for points 2 and 3. Two flare angles γ were
considered: γ =0° (for which the rectangular flat plate is
vertical) and γ = 20°. The bow waves at point 1, which is
located inside the unsteady-flow region defined by (4),
were unsteady for both γ = 0° and γ = 20°. At point 3,
located well outside the unsteady-flow region, the flat
plate generated steady overturning thin sheets of water
for both γ = 0° and γ = 20°. At point 2, located slightly
outside the unsteady-flow region, the flat plate generated
an unsteady bow wave for γ = 0° and a steady
Point 2, FD=0.87, α= 20 °,γ=20°
overturning bow wave for γ = 20°. Thus, these
experimental observations of bow waves generated by an
inclined flat plate agree with the theoretical predictions
given by the Bernoulli boundary (4).

Keynote 4 Delhommeau
energy qb2 / 2 defined by (7) and (1) are linear functions of
FD
0≤δ = ≤1.
1 + FD
Expression (7) yields
2C Z tan α E
FD = -1
1 − qb2 cos α E
(8)
−1 (C Z ) 2 CZ
with sin ( +1 − ) ≤ αE
(1 − qb2 ) 2 1 − qb2
Point 3, FD=0.87, α= 15 °,γ=0°

0.9
0.8
0.7
1.5
0.6
0.4

qb=0
FD

0.5
Point 3, FD=0.87, α= 15 °,γ=20° UNSTEADY

Fig. 7: Pictures of bow wave for different cases


0
0 10 20 30 40 50 60
2 αE (deg)
Nondimensional coordinates (x, y, z) and flow velocities
(vx , vy , vz ) are defined in terms of the ship speed U and Fig. 8: Curves FD defined by (7) for given values of the
the acceleration of gravity g as kinetic energy qb at the crest of a ship bow wave
( x, y, z ) = ( X , Y , Z ) g / U 2 (5a) This expression defines curves FD (α E ) that correspond to
(v x , v y , v z ) = (V x , V y , V z ) g / U 2
(5b) specified values of the kinetic energy qb at a bow wave crest.
In the special case qb =0, (7) is identical to the unsteady-flow
At a point of the free surface z = ζ (x,y) , the Bernoulli
relation (2) with P = Patm and expressions (5) yield boundary (3) shown in Fig. 6. The curves FD (α E ) defined
by (8) are depicted in Fig. 8 for several values of qb in the
q = (v x − 1) 2 + v 2y + v z2 = 1 − 2ζ (6)
range 0 ≤ qb < 1. These curves are roughly parallel to the
where q obviously stands for the magnitude of the total curve qb =0 that borders the unsteady bow wave region. In
flow velocity (flow velocity due to the ship plus uniform the limit qb → 1, (8) yields αE =0 and 0 ≤ FD. Curves
stream opposing the ship speed) . At a wave crest, (6) associated with increasing values of the kinetic energy qb at a
then yields : bow wave crest may be presumed to correspond to steady
2C Z tan α E bow waves that are increasingly more stable.
qb2 = 1 − 2 z b ≈ 1 − (7)
1 + FD cos α E
On Fig. 9, we have reported the wave height for different
This expression yields zb → 0 and qb → 1 in the high- incidence angles up to 90° and heel angles γ up to 40°
speed limit FD → ∞. Thus, the potential energy zb is null together with the Bernoulli bounds for two speeds. The real
and the kinetic energy qb2 / 2 is equal to 1/2 in this limit. draft is: T = D − (1 − cos γ ) H , where H=0.615m is the height
Expression (7) also yields zb → 1/2 and qb → 0 in the of the rotation axis in heel and D=0.2m. So the Froude
low-speed limit FD = FDB (α E ) , where FDB (α E ) is given number FD = U / gT is a function of the heel angle γ for a
by (4). Thus, the kinetic energy qb2 / 2 is null along the given speed. A remarkable conclusion is that these curves
unsteady-flow boundary defined by the Bernoulli show that the simple formula (1) is available up to the
constraint (3). The potential energy zb and the kinetic Bernoulli bound. So, this formula can be used up to the
maximum height of the wave.

Keynote 5 Delhommeau
of the flow velocity at the bow along the horizontal t axis and
the vertical z axis are then equal to q stem cos β ≈ cos β and
1.6
U=1.25 m/s
1.5 q stem sin β ≈ sin β , respectively, here, (9) was used.
1.4
A simple approximation to a ship bow-wave profile may be
1.3
1.2
obtained by assuming that a water particle that passes
1.1 through the bow (t = 0, z = 0) roughly follows a path that is
determined by Newton’s equations
2

1
(1+FD)gZM / U

0.9
d 2 t / dθ 2 = 0 and d 2ζ / dθ 2 = −1
0.8
0.7 γ = 0° This elementary Lagrangian analysis, which obviously
0.6 Theory+Bound ignores interactions among water particles, shows that the
γ =10°
0.5 γ =15° path of a water particle is defined by
0.4 γ =20°
0.3 γ =30° t = θ cos β ζ = θ sin β − θ 2 / 2 (10)
Theory+Bound
0.2 γ =40° Here, a water particle is assumed to be located at the ship
0.1 Theory+Bound
bow (t = 0 , z = 0) at time θ =0 .
0
0 10 20 30 40 50 60 70 80 90 The parametric equations (10) yield
α°
ζ = t (sin β cos β − t / 2) / cos 2 β (11)
Thus, one has ζ = 0 for t = 0 and t = 2 sin ß cos ß. Equation
(11) shows that the highest value of ζ is reached for t = tb =
sin ß cos ß and is given by 2zb = sin2ß. It follows that :
1.8
U=1.50 m/s
1.6 t b = t b ( z b ) = 2 z b (1 − 2 z b ) (12)
1.4
These relations and the change of variable t = t0 + tb , which
places the origin t0 =0 at the bow-wave crest and orients the t0
1.2 axis toward the ship bow (instead of the ship stern, as for the
2
(1+FD)gZM / U

1
t axis) , show that (11) can be expressed as
ζ = z b (1 − t 2 / t b2 ) with - t b ≤ t 0 ≤ t b (13)
0.8 γ = 0° 0
Theory+Bound Expression (13) defines a family of parabolic ship bow waves
0.6 γ =10°
γ =15° that is entirely defined in terms of the height zb of the bow
γ =20°
0.4
γ =30°
wave. This simple one-parameter analytical family of bow
Theory+Bound waves is depicted in Fig. 10 for zb = 0.05 , 0.1 , 0.15 , 0.2 ,
0.2 γ =40°
Theory+Bound 0.25 (top) and for zb = 0.25 , 0.3 , 0.35 , 0.4 , 0.45 , 0.49
0
0 10 20 30 40 50 60 70 80 90
(bottom) . In the limit zb = 0.5, (12) predicts that the width 2tb
α° of the bow wave vanishes, and the wave (13) becomes a
vertical wall of water.
Fig. 9: Comparison of theoretical and experimental bow-
wave height for high incidences.
0.5
5. SIMPLE ANALYTICAL BOW WAVES
zb = 0.05
zb = 0.10
The Bernoulli relation (6) shows that the magnitude of 0.4 zb = 0.15
zb = 0.20
the flow velocity at a ship stem, where the free-surface zb = 0.25

elevation ζ is small, is given by 0.3


q stem ≈ 1 (9)
ζ

Define a horizontal axis that is tangent to the mean ship 0.2


waterline at a ship bow, and points toward the ship stern.
Let t = T g/U2 stand for the distance along the t axis,
0.1
measured from the bow (intersection of the ship stem
with the mean free-surface plane z =0). Thus, the ship
bow is located at t = 0 and z = 0. Furthermore, define the 0
-0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6
nondimensional time θ =Θ g/U, and let ß stand for the t0

angle between the horizontal mean free-surface plane and


the (total) flow velocity at the ship bow. The components

Keynote 6 Delhommeau
0.5
zb = 0.25
zb = 0.30
zb = 0.35
0.4 zb = 0.40
zb = 0.45
zb = 0.49

0.3
ζ

0.2

0.1

0
-0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6
t0

Fig. 10: Family of parabolic bow waves defined by (12)


and (13) for zb =0.05 , 0.1 , 0.15 , 0.2 , 0.25 (top) and
zb =0.25 , 0.3 , 0.35 , 0.4 , 0.45 , 0.49 (bottom) . Fig. 12: Distance tb = Tb g/U2 between a ship stem and bow-
wave crest for five ship hulls.
Thus, the bow wave defined by (12) and (13) becomes a
vertical wall, clearly unstable, as the wave height zb
approaches the upper bound zb = 0.5 allowed by the The expression for the distance between a ship stem and
Bernoulli constraint (3) for steady flows. This property bow-wave crest given in Noblesse et al. (1) shows that an
provides further insight into the unsteady-flow boundary alternative expression for tb in (13) is
(4).
The bow-wave half-width 0 ≤ tb ≤ 1/2 defined by (12) is CX
depicted in Fig. 10 for 0 ≤ zb ≤ 1/2 . Fig. 12 also shows t b = t b ( FD ) = with C X ≈ 1.1 (14)
1 + FD
the experimental measurements for five ship hulls
considered in Noblesse et al. (1). These measurements
are limited to the range zb < 0.3. Furthermore, the Expressions (13) and (14) are based on both elementary
experimental measurements for the Series 60 model are fundamental theoretical considerations (dimensional analysis,
shown in Noblesse et al. (1) to be somewhat marred by limits αΕ → 0, D → 0 and D → ∞) and experimental
considerable scatter (much larger than for the Wigley measurements. Specifically, (13) and (14) follow from
hull). theoretical considerations, but involve constants CZ and CX
Nevertheless, the theoretical predictions and that are determined from experimental measurements.
experimental measurements shown in Fig. 11 are in Expression (14) for the distance tb between a ship stem and
reasonable agreement. bow-wave crest is depicted in Fig. 12 with the experimental
measurements for five ship hulls considered in Noblesse et al.
(1) . Fig. 13 compares the alternative parabolic bow waves
1.2
tb (zb) defined by (14), with tb taken as the functions tb (zb) or tb (FD)
Standing 5°
Ogilvie 7.5° given by (13) or (14), to experimental measurements of bow
1
+
Standing 10°
Wigley
waves due to a rectangular flat plate immersed at a draft D =
Series 60
Larrarte 10°
0.3 m and towed at yaw angles αE = 10° or αE = 20°, heel
0.8
Larrarte 20° angles γ = 0° , 10° , 15° , 20°, and speeds U = 1.5 m/s or 2
2
Tb g / U

0.6
m/s (FD =0.87 or 1.17). Theoretical predictions of the bow
+
wave height and of the location of the bow wave crest given
0.4
++ +
++ +
by (2) and (13) or (14) agree reasonably well with
+ + + experimental observations.
0.2

0
0 0.1 0.2 0.3 0.4 0.5 0.6
2
Zb g / U

Fig. 11: Distance tb = Tb g/U2 between a ship stem and


bow-wave crest for five ship hulls. The solid circle is the
approximation (12)

Keynote 7 Delhommeau
U = 1.5 m/s , αE = 10 ° U = 2.0 m/s , αE = 20 °
0.05 0.18
tb(FD) tb(FD)
tb(zb) tb(zb)
γ = 0° 0.16 γ = 0°
γ = 10° γ = 10°
0.04 γ = 15° γ = 15°
0.14
γ = 20° γ = 20°
0.12
0.03
0.1
Z (m)

Z (m)
0.08
0.02
0.06

0.04
0.01

0.02

0 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0 0.1 0.2 0.3 0.4 0.5 0.6
X (m) X (m)

Fig. 13: Comparison of the simple analytical bow-wave


approximation given by (12) and (14) with experimental
measurements
U = 1.5 m/s , αE = 20 °
0.12
tb(FD)
6. COMPOSITE BOW WAVE
tb(zb)
γ = 0°
0.1 γ = 10° However, the simple analytical bow waves given by (13)
γ = 15°
γ = 20° with (12) and (14) do not agree with experimental
0.08 measurements beyond the bow wave crest. These
observations suggest that interactions among fluid particles,
Z (m)

0.06
ignored in the elementary Lagrangian analysis, are more
important in the “recovery zone” past a wave crest than in the
0.04
“build-up zone” between a ship stem and bow wave crest.
Thus, the parabolic bow wave defined by (14) with (13) may
0.02
be used for −tb ≤ t0 ≤ 0 but not for t0> 0. The ship bow wave
is then considered here aft the wave crest, i.e. for t0> 0. An
0
0 0.1 0.2 0.3 0.4 0.5 0.6 obvious analytical approximation for a bow wave aft the crest
X (m) 2
is an elementary wave with wavelength 2πU /g , i.e.
U = 2.0 m/s , αE = 10 ° ζ = z b cos t 0 with t 0 ≥ 0. Here, t0 = 0 at the crest of the bow
0.08 2
tb(FD)
tb(zb) wave and t0 = T g/U . The change of variable
γ = 0°
γ = 10°
γ = 15°
t b = τ b 1 − σ S − t in the foregoing complementary
0.06 γ = 20°
parabolic and sinusoidal approximations yields the composite
bow wave :
Z (m)

0.04

ζ = z b (σ S + 2 1 − σ S t / t b − t 2 / t 2 for 0 ≤ t ≤ t C
 b
 (15)
0.02 ζ = z b cos(t − t b 1 − σ S ) for t ≥ t C
with
0
0 0.1 0.2 0.3 0.4 0.5 0.6 tC = tb 1 − σ S
Here, σS=zS/ zb defines the ratio of the elevation zS of the free
X (m)

surface at the ship stem over the height zb of the wave crest.
Expression (15) yield ζ= zS at a ship stem t=0 and ζ = zb at a
bow-wave crest t C = t b 1 − σ S .

Keynote 8 Delhommeau
considerations) show that one has
7. WAVE HEIGHT AT SHIP STEM ZS g /U 2
= Ψ0α /(1 + F N
)
D

The height of water at the stem of a ship hull—with a Thus, the function
nonbulbous wedge-shaped bow—that advances at C N (α ) = (1 + F N ) Z S g / U 2 (16)
D
constant speed in calm water is considered using two
distinct methods: (i) a theoretical-experimental approach can be expected to be independent of δN , i.e. of FD. The
in which elementary fundamental theoretical exponent N in (16), which specifies the variation of the water
considerations (dimensional analysis and rudimentary height ZS at a ship stem with respect to the ship speed U and
asymptotic considerations in thin-ship, shallow-draft and draft D, is now determined— using experimental
deep-draft limits) are used in conjunction with measurements—from the condition that the function CN
experimental measurements for simple hull forms and a defined by (16) is nearly independent of FD .
rectangular flat plate towed at several yaw and heel The variations of C1, C2, C3, C4 with respect to FD are
angles; and (ii) thin-ship theory, i.e. a fully-analytical considered for seven cases. Specifically, we compare
approach. Both of these two methods yield simple experimental measurements, obtained by Larrarte (5) at Ecole
expressions that define the rise of water at a ship stem Centrale de Nantes, for two strut-like models that have
explicitly—ab initio and without calculations—in terms rectangular framelines and sharp-ended parabolic waterlines
of the ship speed, draft, and waterline entrance angle. with entrance angle 2 α = 20° or 40°, results of
The theoretical-experimental expression and the thin-ship measurements, obtained by the authors in the towing tank of
expression are in good agreement except at low Froude Ecole Centrale de Nantes, for a rectangular flat plate
numbers, and are also in reasonable agreement with immersed at a draft D = 0.3 m or 0.2 m and towed at several
experimental measurements. speeds U in the range 1m/s ≤ U ≤ 2.5 m/s , yaw angles α =
10° , 15° , 20° , 25°, and heel angles γ = 0° , 10° , 15° , 20°
7.1 THEORETICAL-EXPERIMENTAL APPROACH and measurements for the Wigley hull, obtained at the Univ.
of Tokyo and the Ship Research Institute (Japan) and
The height ZS of the free surface at the stem of a ship, reported in Kajitani et al. (6)
with a nonbulbous wedge-shaped bow, that advances at Hull Wigley Strut 10° Strut 20° BN
constant speed U in calm water mostly depends on U, the
gravitational acceleration g, the hull draft D, and the B1 -0.185 -0.203 -0.539 -0.309
entrance angle 2α at the waterline. Viscosity and surface
B2 -0.103 -0.124 -0.346 -0.191
tension, variables related to the overall ship geometry
(e.g. beam/length and draft/length ratios), and the flare B3 0.028 -0.069 -0.214 -0.085
angle γ are expected to have a secondary influence.
Dimensional considerations then show that ZS g/U2 is a B4 0235 -0.028 -0.117 0.030
function of α and the draft-based Froude number FD.
Thin-ship theory indicates that ZS is approximately
proportional to α in the limit α → 0. Thus, Ψ=(Zsg/U2)/α Plate 10° 15° 20° 25° BN
is expected to be a function of the draft-based Froude
B1 -0.111 -0.114 -0.229 -0.378 -0.208
FDN
number FD, or the related variable δ N = which
1 + FDN B2 -0.048 -0.036 -0.104 -0.168 -0.089
varies in the range 0 ≤ δ N ≤ 1 as 0 ≤ FD ≤ ∞ . The
exponent N is to be determined in the manner explained B3 0.058 0.097 0.107 0.113 0.113
further on.
ZS vanishes in the shallow-draft limit D → 0, i.e. in the B4 0.241 0.328 0.472 0.815 0.464
high-Froude-number limits FD → ∞, and is finite in the Table 2: Slopes B1 , B2 , B3 , B4 for the Wigley hull, the
deep-draft limit D → ∞, i.e. in the low-Froude-number Larrarte 10° and 20° struts, and the flat plate.
limits FD → 0. The simplest function that satisfies these N 2 2.5 3
two boundary conditions is the linear function CSN 0.89 0.97 1.06
Ψ = Ψ0 (1 − δ N ) = Ψ0 /(1 + F N )
D
These simple basic theoretical considerations Table 3: Constants in (18) determined from experimental
measurements for the Wigley hull, the Larrarte 10° and 20°
(dimensional analysis and rudimentary asymptotic
struts, and the flat plate .

Keynote 9 Delhommeau
Least-square linear fits are given under the form center and bottom rows of Fig. 15 correspond to N =2 , 2.5 or
CN = AN + BN FD (17) 3 . The experimental data are those previously reported.
A null value of the slope BN in (17) indicates that CN is These functions are depicted as functions of 0≤FD≤2 .
independent of FD. The values of the slopes B1 , B2 , B3 ,
B4 are listed in Table 2 .
N = 2 CS=0.89
1.2 Plate α = 10°
Plate α = 15°
The average slopes BN for the Wigley hull and the two Plate α = 20°
1 Plate α = 25°
Larrarte struts, and for the flat plate at four yaw angles, Strut α = 10°
Strut α = 20°
are also indicated in Table 2 . The slopes BN are shown in

ZS g/U 2 cosα / tanα


0.8 Wigley
N=2
Fig. 14 as functions of 1 ≤ N ≤ 4 for the seven cases
considered in Table 2 . Fig. 14 and Table 2 show that the 0.6

slopes B1 and B2 are negative for the seven cases 0.4


considered here. The slopes B2 and B4 are positive for
five of the seven cases. The experimental measurements 0.2

reported in Table 2 and Fig. 14 indicate that the least


0
variation with respect to speed in (16) is obtained for the 0 0.2 0.4 0.6 0.8 1
FD
1.2 1.4 1.6 1.8 2

exponent N in the range 2 ≤ N ≤ 3 .

N = 2.5 CS=0.97
1.2 Plate α = 10°
Plate α = 15°
Plate α = 20°
1 Plate α = 25°
1 Strut α = 10°
Plate α = 10° Strut α = 20°
Plate α = 15°
ZS g/U 2 cosα / tanα

0.8 Wigley
N=2.5
Plate α = 20°
Plate α = 25°
0.6
Strut α = 10°
0.5 Strut α = 20°
Wigley 0.4
Fit
0.2
BN

0
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
FD

-0.5 1.2 S Plate α = 10°


N = 3 C =1.06 Plate α = 15°
Plate α = 20°
1 Plate α = 25°
Strut α = 10°
Strut α = 20°
ZS g/U 2 cosα / tanα

Wigley
-1 0.8
0 1 2 3 4 5 N=3
N
0.6

Fig.14: Slopes B1 , B2 , B3 , B4 in the linear fit to the


0.4
normalized water rise CN for the Wigley hull, the Larrarte
10° and 20° struts, and the flat plate. 0.2

The three possibilities N = 2 , 5/2 , 3 are then considered. 0


0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
FD

The foregoing analysis, based on elementary Fig.15 Normalized water height (Zs g/U2) cos α/ tan α and
fundamental theoretical considerations and experimental
functions C S /(1 + F ) with N=2, 2.5, 3 for the Wigley
N N
measurements, yields the simple analytical expression D
hull, the Larrarte 10° and 20° struts, and the flat plate.
C SN tan α
Z S g /U 2 = with 2 ≤ N ≤ 3 (18)
(1 + F ) cos α
N 7.2 FULLY-ANALYTICAL APPROACH
D
The function tanα / cosα provides a better fit at high A complementary analysis of the thin-ship limit α → 0 is
incidence angles that the function α and has the same now considered. Specifically, we consider steady free-surface
behaviour in the limiting case α → 0. flow about an infinitely-long thin wedge defined by
The values of CS for N =2 , 2.5 or 3 are listed in Table 3 . Y = ±X tanα
C SN with -∞ <X ≤ 0 , -D ≤Z ≤ 0 . (19)
The functions are compared to the normalized Within the context of thin-ship theory, which is appropriate
(1 + F N )
D in the limit α → 0 now considered, the flow about the wedge
water height (ZS g/U2) cosα / tanα in Fig. 15. The top, (19) can be represented in terms of a distribution of sources,

Keynote 10 Delhommeau
with (constant) strength 2 tan α , over the semi-infinite (20) for the function E(FD) in (20a) are depicted in Fig. 16
strip (19) in the X, Z plane, and is defined in terms of a together with experimental measurements for the Wigley
Green function G(ξ- x, η- y, ζ, z). Here, (ξ, η,ζ) and (x, y, hull, two struts that have sharp-ended waterlines with
z) stand for coordinates of a flow-field point and a source entrance angle 2α = 20° or 40°, and a flat plate towed at a
point. The source point (x, y, z) is located in the plane y yaw angle α = 10° , 15° , 20° , 25°.
=0 with -∞ <X ≤ 0 and -d ≤ z ≤0 . The z-axis is vertical The theoretical-experimental expression (20a) and the thin-
and points upward, and the mean free surface is taken as ship expressions (20b)–(20d) are nearly identical for FD > 0.8
the plane z = 0. Furthermore, the coordinates (ξ,, η, ζ) but are significantly different for FD < 0.6 . The theoretical-
and (x, y, z) and the draft d of the wedge are experimental expression (20a) and the thin-ship expressions
nondimensional in terms of the gravitational acceleration (20b)–(20d) are in reasonable agreement with the
g and the speed U of the wedge. Thus, one has experimental measurements shown in Fig. 16. In fact, the
d = D g/U2 = 1/FD 2 . experimental measurements are fairly evenly distributed
Within the thin-ship approximation, the nondimensional around the theoretical-experimental approximation (20a) and
free-surface elevation Z g/U2 is given by the thin-ship expressions (20b)–(20d).
Zg ∂φ (ξ , η , ζ = 0) 0 0 ∂G
= = 2 tan α ∫ − d dz ∫ −∞ dx
U 2 ∂ξ ∂ξ 1.8 Plate α = 10°
Plate α = 15°
The Green function can be expressed in the form given in 1.6 Plate α = 20°
Appendix 3 of Noblesse (7). Plate α = 25°
Strut α = 10°
Thus one has 1.4 Strut α = 20°
Z S g tan α U Wigley
ZS g/U 2 cosα / tanα

= E ( FD ) with FD = (20a) 1.2 Theory-Exp.

U 2 cos α gD Integral
Series
1
Series+Correction
where E(FD) stands for the function Fontaine
2 2 0.8
1 − e −(sin t ) / FD
High-Froude Approx.
2 π /2
E = ∫ 0 dt (20b) 0.6
π sin t
A simple approximation of integral (20b) is given by the 0.4

(enhanced) high-Froude-number asymptotic 0.2


approximation
πE 1 2/3 19 / 45 0
≈ + + + 0 0.2 0.4 0.6 0.8 1
2 1 + FD2 (1 + FD2 ) 2 (1 + FD2 ) 3 2 2
FD / (1+FD)
(20c) Fig.16: Comparison of the six alternative expressions (20) for
26 / 105 601 / 4725 1502 / 31185
+ + + ... the function E(FD) in (20) with experimental measurements.
(1 + FD2 ) 4 (1 + FD2 ) 5 (1 + FD2 ) 6
A low-Froude-number correction can be added to the The two approaches that have been used in this study— thin-
high-Froude-number approximation (20c) to obtain a ship theory, and a theoretical-experimental method in which
practical approximation to the integral (20b). E.g., the elementary fundamental theoretical considerations are used in
expression conjunction with experimental measurements—may seem
πE 1 2/3 19 / 45 26 / 105 overly simplified and even exotic to the CFD generation.
≈ + + + + Thus, it may be useful to note here that simple analytical
2 1 + FD (1 + FD )
2 2 2
(1 + FD )
2 3
(1 + FD2 ) 4
methods can yield useful results. In particular, analytical
601 / 4725 1502 / 31185
+ + 4.16e −13FD −0.26 (20d) methods can provide simple “cause-and-effect”
(1 + FD )
2 5
(1 + FD )
2 6 relationships—often of critical importance for practical
applications, notably at preliminary and early design stages—
is essentially identical (20b), except for very low FD and that could only be derived from numerical simulations if a
huge number of systematic parametric calculations can be
may be used in practice instead of (20b).
performed.
The one-term high-Froude-number approximation Finally, an amusing result of the high-Froude-number
πE ≈ 2 / FD2 (20e) slender-body studies of Sclavounos (10) and Fontaine et al.
(8) (9) and the thin-ship analysis considered here is that these
given by the thin-ship approximation (20c) and the
three studies yield the high-Froude-number approximations:
high-Froude-number approximation
1 1 2
πE ≈ 1 / FD2 (20f) Z S ≈ CDα with C = , , .
2π π π
given by Fontaine et al. (8) (9) are in agreement, except
for the factor 2 in (20e). The six alternative expressions

Keynote 11 Delhommeau
17. The vector m is colinear with the projection onto the
8. FLOW AT SHIP-HULL AND FREE-SURFACE mean free-surface plane z =0 of the unit vector n =(nx, ny, nz)
CONTACT-CURVE normal to the ship hull; bottom of Fig. 17 . The vectors n and
m point outside the ship. The vector t is tangent to the ship
hull surface and, on the positive side y>0 of the ship hull
considered here, points toward the ship bow; top of Fig. 18 .
t m One has
Y α>0 t = (cosα, -sinα, 0) m = (sinα, cosα, 0)
n = (sinα cosγ, cosα cosγ, -sinγ) (22a)
X with −π / 2 ≤ α ≤ π / 2 and − π / 2 ≤ γ ≤ π / 2 .
2αE
In the bow region, the angle α between the unit vector t and
the x axis is positive. The flare angle γ between the normal
vector n to the ship hull and the mean free-surface plane z =0
is positive for a typical hull form, as in bottom of Fig. 17, and
negative for a tumble hull.
The unit vector
Hull Free surface s = t × n = (sinα sinγ, cosα sinγ, cosγ )
= m sinγ + k cosγ (22b)
k µ>0 is tangent to the ship hull and points upward, as in bottom of
Z Fig. 17. Here, k = (0, 0, 1) is the unit vector along the vertical
γ z axis; see bottom of Fig. 17.
Y s z
The velocity component v along the unit vector n normal to
γ m
the ship hull is null. Thus, the flow velocity is given by vtotal
= u t + w s = u t + w t × n , where the velocity components u
n and w along the unit vectors t and s tangent to the hull are
determined by the free-surface boundary conditions (21b)
and (21c) .

Fig.17: Definition sketch The total-flow velocity at the contact curve can be shown to
be given by
Steady free-surface flow about a ship in the “steady
bow-wave regime” is now considered. A simple vtotal = u t + w n × t
analytical theory of overturning ship bow-waves is 1 − 2ζ
developed. A main element of this theory is a fully- vtotal = cos(γ + µ ) t + (23)
cos (γ + µ ) + ζ t2 cos 2 µ
2
nonlinear analysis of the steady inviscid flow along the
contact curve be-tween the ship hull and the free surface. 1
Thus, surface-tension and viscosity effects are ignored ζ t cos µ n × t
cos (γ + µ ) + ζ t2 cos 2 µ
2
here. However, no other approximation is made, and the
analysis is exact for steady inviscid flows.
The ship-hull boundary condition and the kinematic and
dynamic boundary conditions Here, −π / 2 ≤ µ ≤ π / 2 stands for the unknown angle
n (v x − 1) + n v y + n v z = 0
x y z
(21a) between the free surface and the mean free-surface plane z=0,
as shown in bottom of Fig. 17. Expression (23) follows from
v z = (v x − 1)ζ x + v y ζ y (21b) exact boundary conditions (for steady inviscid flows), at the
actual locations of the ship hull and the free surface, and thus
(v x − 1) + v + v = 1 − 2ζ
2 2
y
2
z (21c) is exact.

at the free surface z = ζ(x,y) are presumed to hold along If tanγ tanµ << 1 , one obtains the approximation
the contact curve between the ship hull and the free vtotal = u t + w n × t (24)
surface. The three boundary conditions (21) provide three
1 − 2ζ 1
algebraic equations (two linear equations and a quadratic vtotal = cos γ t + ζ t n × t
equation) that can be used to determine the three velocity cos γ
2
+ ζ t2 cos γ + ζ t2 2
components vx , vy and vz in terms of ζ, ζx and ζy . Two This approximation which is independent
orthogonal unit vectors t =(tx, ty, 0) and m =(-ty, tx , 0) of the unknown angle µ may be expected to hold except near
that lie in a horizontal plane are defined; see top of Fig.

Keynote 12 Delhommeau
a ship stem or stern where γ and/or µ can be large. The hull along the ship-hull/free-surface contact curve. This step
approximation (24) defines the nondimensional is an elementary Lagrangian analysis of the motions
total-flow velocity at the free-surface and ship-hull of fluid particles that leave the ship hull at the
contact curve in terms of the ship speed U and flare angle flow-detachment curve (ship-hull/free-surface contact curve.
γ , and the elevation ζ of the contact curve. A particle of water leaving the hull at contact point (t,
m,ζ) follows a trajectory given by
u , µ* = 0 ° t '−t
1 u , µ* = 10 °
u , µ* = 20 °
m' = m + w sin γ
0.8 u , µ* = 30 °
u
(25)
w , µ* = 0 °
t '−t t '−t
0.6 w , µ* = 10 °
w , µ* = 20 °
z' = ζ + ( w cos γ − )
0.4 w , µ* = 30 ° u 2u
0.2
Initial coordinates (t, m,ζ) are not independant. In particular,
u,w

0
along a flat plate with a flare angle γ, we have : m = ζ tan γ .
-0.2
γ = 15°, z b = 0.2 Expressions (25) define the detached sheet of water that
-0.4
leaves the ship hull along the flow-detachment curve in terms
-0.6 of the location of the flow-detachment curve and the related
-0.8 velocity components u and w. These velocity components are
-1 defined by (24) in terms of the ship speed, the hull geometry,
-1 -0.5 0 0.5 1
t0 / tb and the location of the flow-detachment curve. Thus, the
detached sheet of water generated at a ship bow is explicitly
u , µ* = 0 °
determined in terms of the ship speed, the hull geometry, and
1 u , µ* = 10 °
u , µ* = 20 °
the location of the flow-detachment curve.
0.8 u , µ* = 30 ° Thus, the projections of the paths of water particles on the
w , µ* = 0 °
0.6 w , µ* = 10 ° horizontal plane (m , t) and the vertical planes (k , t) and (k ,
w , µ* = 20 °
0.4 w , µ* = 30 ° m) are a straight line and parabolas, respectively, as
0.2
expected.
The water trajectory defined by (25) intersects the mean free-
u,w

0
surface plane z =0 for
-0.2
t z = 0 − t m z =0 − m
γ = 20°, z b = 0.1 = =
w sin γ
-0.4
u (26)
-0.6

-0.8 w cos γ + w 2 cos 2 γ + 2ζ


-1
-1 -0.5 0 0.5 1
Thus, the variables t‘ and m’ in (25) vary within the ranges
t0 / tb t z =0 ≤ t' ≤ t
Fig. 18 : Effect of angle µ between the free surface and m ≤ m' ≤ m z = 0 if w > 0 (27)
the mean free-surface plane z = 0 on the velocity m z =0 ≤ m' ≤ m if w < 0
components u and w defined by (23), (13) and (12) for
If w>0 , the water trajectory reaches a top height for
(top) zb=0.2 and γ = 15° and (bottom) zb=0.2 and γ = 20°.
t top − t mtop − m
= w cos γ = (28a)
The effect of the angle µ is illustrated in Fig. 18, where u w sin γ
the angle µ in (23) is taken as and the top height is given by
µ = µ * t 0 / t b with - 1 ≤ t 0 / t b ≤ 1 z top − ζ = ( w cos γ ) 2 / 2 ≈ (uζ t ) 2 / 2 (28b)
and µ* =0° , 10° , 20° , 30°. Fig. 11 shows that the
velocity components u and w are not drastically affected Thus, the maximum height z top reached by water particles
by the angle µ . Thus, the approximation µ =0 and the that leave the ship-hull/free-surface contact curve at a height
related expression (24) may be used in practice. z =ζ is significantly larger that ζ only if | ζt | is large, e.g. near
a ship stem where the contact curve is tangent to the ship
9. LAGRANGIAN ANALYSIS OF DETACHED stem; see Noblesse et al. (1991) . Expression (28b) yields z top
BOW SHEET
= ζ if w =0 , e.g. at a crest of the flow-detachment curve.
Fig. 19 shows a comparison of the analytical and
The next main step in the theory of overturning ship
experimental bow waves for four cases that illustrate the
bow-waves developed here is the determination of the
effect of the speed U, the incidence angle αE , and the
shape of the detached sheet of water that leaves the ship

Keynote 13 Delhommeau
analytical calculations appear to be in reasonable agreement
with the experimental observations, and to predict the effect
U = 1.5 m/s , α = 20 ° , γ = 20 °
of the speed U, the incidence angle αE and the flare angle γ
upon the main geometrical characteristics of the overturning
0.2
bow wave approximately correctly.

Z
0.1
0

0.1
0 U (m/s) αΕ (°) γ (°) Zbexp(m) Zb th(m)
0.2 0.2 0.1 0
0.3
0.7 0.6 0.5 0.4

X
1.5 20 20 0.089 0.104
1.5 10 20 0.049 0.048
2.0 20 20 0.145 0.160
2.0 10 10 0.147 0.160

U (m/s) αΕ (°) γ (°) Tbexp(m) Tb th(m)


1.5 20 20 0.061 0.066
1.5 10 20 0.093 0.113
2.0 20 20 0.147 0.167
2.0 10 10 0.153 0.167
Table 4 : Bow-wave height and location

A comparison of experimental and theoretical values of the


bow-wave height Zb and distance Tb from the leading edge of
the flat plate is shown below in table 4 .
Experimental measurements and theoretical predictions are in
reasonable agreement.

10. COMPARISON WITH MEASUREMENTS


U = 2.0 m/s , α = 20 ° , γ = 20 °

The first case of comparison is a rectangular flat plate


0.2 immersed at a draft D = .3 m. and towed at speeds U =1.5
m/s and 2 m/s (draft-based Froude numbers FD =0.87 and
Z

0.1

1.17) , with incidence angles αE = 10° and 20°, and flare


0

0.1

angles γ = 0°, 10°, 15°, 20°. The shape of the composite


0
0.2 0.2 0.1 0
0.5 0.4 0.3
0.7 0.6
X
wave is compared with measurements on Fig. 20. This wave
is composed of two parts. The first part is parabolic and takes
into account the ship stem height. The second part is
sinusoidal, with a wave-length based on Froude number.

U = 1.5 m/s , αE = 10 °
0.06 Sinusoidal
γ = 0°
γ = 10°
γ = 15°
0.05 γ = 20°

0.04
Z (m)

0.03

0.02

Fig. 19: Experimental observations and analytical


predictions of overturning bow waves due to a flat plate 0.01

at speed U, incidence αE and flare angle γ.


0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.6

flare angle γ . Specifically, analytical predictions and X (m)

experimental observations are shown for αE = 20° , γ =


20° and U =1.5 m/s (top) or U =2 m/s (bottom) . The

Keynote 14 Delhommeau
U = 1.5 m/s , αE = 20 °
0.12 Sinusoidal
0.03
γ = 0° Maq10 Fn= 0.203
0.11 γ = 10° Fn=0.250
γ = 15° Fn=0.297
0.1 γ = 20° Fn=0.344
0.09 Fn=0.391
Fn=0.417
0.08 0.02 Exp Fn=0.203
Exp Fn=0.250
0.07 Exp Fn=0.297
Z (m)

Exp Fn=0.344

Z/L
0.06 Exp Fn=0.391
Exp Fn=0.417
0.05

0.04 0.01

0.03

0.02

0.01

0 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0 0.1 0.2 0.3 0.4 0.5
X/L
X (m)

U = 2.0 m/s , αE = 10 °
0.08 Sinusoidal
γ = 0° 0.08
γ = 10°
0.07 γ = 15° Maq20 Fn= 0.203
γ = 20° Fn=0.250
0.07 Fn=0.297
0.06 Fn=0.344
0.06 Fn=0.391
Fn=0.417
0.05
Exp Fn=0.203
0.05 Exp Fn=0.250
Z (m)

0.04 Exp Fn=0.297


Exp Fn=0.344
Z/L

0.04 Exp Fn=0.391


0.03 Exp Fn=0.417
0.03
0.02
0.02
0.01
0.01
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
0
X (m) 0 0.1 0.2 0.3 0.4 0.5
X/L

U = 2.0 m/s , αE = 20 °
0.2 Sinusoidal
γ = 0° 0.03 Wigley Hull
0.18 γ = 10°
γ = 15°
γ = 20°
0.16
0.025 Fn= 0.250
Fn=0.267
0.14
Fn=0.289
0.02 Fn=0.316
0.12 Fn=0.354
Z (m)

Fn=0.408
0.1 Exp Fn=0.250
Z/L

0.015 Exp Fn=0.267


0.08 Exp Fn=0.289
Exp Fn=0.316
0.06 Exp Fn=0.354
0.01
Exp Fn=0.408
0.04

0.02 0.005

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
0
0 0.1 0.2 0.3 0.4 0.5
X (m) X/L

Fig. 21 : Comparison of theoretical and experimental bow-


Fig. 20 : Comparison of theoretical and experimental
wave height for ship hulls.
bow-wave height for a flat plate.
The half bow-angles are 10 and 20°. Corrections of effective
The second comparison is made on strut-like models
draft and bow angle for practical ship hulls are given in
simulating real hulls. These models have been used by F.
Noblesse and al (1). The Froude number indicated in Fig. 21
Larrarte (5) for her PhD Thesis. The 2 models have
identical extremities (amphidromic profile), a length is relative to the ship length ( Fn = U / gL ).
L=1.5 m, a draft D=0.20 m with rectangular sections.

Keynote 15 Delhommeau
The shape of models Μaqα is given by: Three main limitations of the results given here need to be
3 2  L noted. (i) The theory of overturning ship bow waves does not
y=− x tan α + x tan α x ∈ 0,  provide a complete description of the flow. In particular,
2L  3 although the theory provides reasonable predictions of the
B(α ) L L L main characteristics (horizontal extent and height) of the thin
y= = tan α x∈ , 
2 6 3 2 sheet of water that is generated by a ship bow, it provides no
The last hull is Wigley hull. Measurements come from information about the thickness of the sheet. (ii) The
cooperatives experiments of Washington Workshop in expressions for the bow-wave height and shape, and the
1983 (6). related unsteady-bow-wave criterion and overturning-bow-
Results for practical hulls show a reasonable agreement wave theory assume a sharp wedge-like ship bow. Thus,
for the shape of the wave and a good prediction of the these results cannot be applied to ships with rounded or
wavelength up to a Froude number of 0.3. Above this bulbous bows. (iii) The simple theory of overturning bow
number, the estimated wavelengths given in (15) have to waves given here does not predict the occurrence of
be corrected for high-speed effects. overturning if the flare angleγ is null, i.e. for wall-sided ship
hulls. Indeed, within the theory developed here, the flow-
CONCLUSION separation curve along which a detached thin sheet of water
is generated becomes a streamline in the special case γ =0 .
Several basic questions pertaining to the bow wave
generated by a ship in steady motion have been REFERENCES
considered and simple analytical results given. These
results are now summarized. 1. NOBLESSE F., HENDRIX D., FAUL L., SLUTSKY J.
The height of a ship bow wave is explicitly defined in ‘Simple analytical expressions for the height, location, and
terms of the ship speed U, draft D and waterline entrance steepness of a ship bow wave’, J. Ship Research, 50, 360-
angle αE by expression (1). This simple analytical 370, 2006.
expression is in good agreement with experimental 2. NOBLESSE F., HENDRIX D., KARAFIATH G, ‘When is
measurements. Expression (1) for a ship bow-wave the bow wave of a ship in steady motion unsteady?’, 21st Il
height and the Bernoulli constraint for steady flows were Workshop on Water Waves and Floating Bodies,
used to obtain a criterion (4), which predicts when a ship Loughborough, UK, pp. 133–135, 2006.
in steady motion generates an unsteady bow wave. Two 3. DELHOMMEAU G., GUILBAUD M., NOBLESSE F.
expressions defining a family of parabolic ship bow ‘Flow at a ship-hull and free-surface contact curve at a ship
waves are given and compared to experimental bow and overturning bow wave’, 8th Numerical Towing
measurements for a flat plate. A composite bow-wave Tank Symp., Varna, Bulgaria, pp. 6.1–6.6, 2005.
(15) is deduced. Two expressions of the wave height at 4. DELHOMMEAU G., GUILBAUD M., NOBLESSE F. ,
stem are given (18) (20). ‘A simple theory of overturning ship bow waves’, 21st Il
The flow velocity at the contact (flow-separation) curve Workshop on Water Waves and Floating Bodies,
between a ship hull and the free surface is given by the Loughborough, UK, pp. 33–36, 2006.
simple analytical expression (23). The approximation 5. LARRARTE F., ‘Etude experimentale et theorique des
(24), which does not involve µ may be used in practice profils de vagues le long d’une carene’, These de Doctorat,
instead of (23) . The detached sheet of water (overturning Univ. de Nantes et Ecole Centrale de Nantes, 1994.
bow wave) that is commonly generated by a ship bow in 6. KAJITANI H., MIYATA H., IKEHATA M., TANAKA
the “steady-bow-wave regime” is explicitly determined H., ADACHI H., NAMIMATSU M., OGIWARA S. , 2nd
in terms of the ship speed, the hull geometry, and the DTNSRDC Workshop on Ship Wave-Resistance
location of the flow-detachment curve. Fig. 20, 21 and 22 Computations, David Taylor Naval Ship Research and
show that predictions given by this simple theory are in Development Center, MD USA, 1983.
reasonable agreement with experimental measurements 7. NOBLESSE F., ‘The near-field disturbance in the
for a flat plate and practical ship hulls. centerplane Havelock source potential’, 1st Il Conf.
The simple analytical expressions obtained in this study, Numerical Ship Hydro., Washington DC, 481-501, 1975.
using elementary fundamental considerations and 8. FONTAINE E., FALTINSEN O.M. (1997) ‘Steady flow
analysis, provide explicit relationships between main near a edge shaped bow’, 12th Il Workshop on Water Waves
characteristics of a ship bow wave (wave height and and Floating Bodies, Carry-le-Rouet, France, 75-78.
location, steadiness, geometry of overturning bow wave) 9. FONTAINE E., FALTINSEN O.M., COINTE R. , ‘New
and main design parameters (ship speed, draft, waterline insight into the generation of ship bow waves’, J. Fluid
entrance angle, and flare angle) that define a ship. These Mechanics, 321, 15-38, 2000.
“cause-and-effect” relations may be useful for practical 10. SCLAVOUNOS P. , ‘On the intersection near a fine ship
applications to ship design, and should illustrate the bow’, 29th Symp. Naval Hydro., Natl Academy Press, 934-
value of analytical methods. 945, 1995.

Keynote 16 Delhommeau

Das könnte Ihnen auch gefallen