Oceanic Engineering International, Vol. 6, No. 2, 2002, pp.

74-115

Study Of Michell’s Integral And Influence Of Viscosity And Ship Hull

Form On Wave Resistance

A. Sh. Gotman1

1
D. Tech. Sci., professor Gotman A.Sh., Novosibirsk State Academy of Water Transport, Shchetinkina st., 33,Novosibirsk,
630099, Russia

e-mail: [email protected]

ABSTRACT

A systematical study of Michell’s integral was carried out and an investigation into a discrepancy between the linear theory and
experiment was conducted. Special attention was given to the influence of viscosity on the interaction of the bow and stern
systems of ship waves.
An experiment with two struts provided information about the mechanism of influence of viscosity of a fluid on the ship wave
resistance. This experiment argued against a popular opinion that the humps and hollows are absent in an experimental curve
because of the action of the boundary layer, the wake, and the hull sheltering effect. Our calculations confirm the hypothesis that a
certain part of the bow wave’s energy is wasted on turbulence of a flow around the moving ship and does not participate in the
interaction with the stern system of waves.
The new form of Michell’s integral in which the monotone part is separated from the oscillatory part made it possible to
determine the previously unknown peculiarities of this integral. For the first time analytical and experimental methods suggested
the reason why the wave resistance at low and high speeds should be investigated separately.
The new comparative criterion of choosing a ship hull form with the least wave resistance was obtained from the modified form
of Michell’s integral and then verified for models described by different equations. This method provided optimum forms that
were well coordinated with the known forms.
The study of the influence of the ship hull surface curvature on wave resistance produced a graph combining all possible ship
hull forms. It was shown that the “simplified” hull shapes with cylindrical fore bodies have the largest possible value of wave
resistance and why the design of fore bulbs is an extremely tricky matter.

1. INTRODUCTION Pavlenko [1937], de Sendagorta & Grases [1988], Sharma

[1969], Shearer [1951], Sretenskii [1937; 1941; 1977], Tuck
[1976; 1989], Wehausen [1957; 1963; 1973], Weinblum [1930;
It has been more than one hundred years since Michell’s 1932; 1950; 1952], Wigley [1926; 1931; 1936; 1937 – 8; 1942;
famous article about the wave resistance of thin ships was first 1944; 1948; 1962; 1963; 1967], and others).
published in 1898. More than one hundred years have passed, but in spite of all
Michell’s theory dealt with the problem of ship wave the efforts that have been made by mathematicians and
resistance and had been worked out by linearizing the hydrodynamicists over the world, we still have nothing better
boundary conditions on both the ship’s hull and the free for practical implementation than the Michell theory. That is
surface of the fluid. Many works have been devoted to the why we are forced to repeatedly return to the Michell integral.
study of Michell’s integral. (For example Birkhoff et al. It would be convenient if the Michell calculated curve of ship
[1954], Birkhoff & Kotik [1954], Cuthberg [1964] Emerson wave resistance coincided with the experimental one.
[1954; 1971], Guilloton [1946; 1951; 1952; 1962; 1965], However, we still don’t know the cause of the humps and
Doctors [1998], Havelock [1909; 1926; 1932; 1934a; 1934b; hollows present in the Michell curve that are absent in the
1935; 1948; 1951], Hsiung & Wehausen [1969], Inui [1957; experimental one for small Froude numbers. Nevertheless, the
1962; 1981], Keller & Ahlawalia [1976], Lunde [1951], Michell integral can be used in the design of ship hulls. This
Michelsen [1966; 1967; 1972], Newman [1964; 1976], has become the topic of a large body of research.

74 Oceanic Engineering International

 Michell himself carried out the first test of the theory. As his A. Musker [1989] showed that the calculated points for the
brother wrote, Michell had assembled a set of data, containing wave resistance that had been obtained by different authors for
known experimental measurements of the total resistance of the same model of a Series 60 hull using linear theories
water to the movement of real ships. Unfortunately, he didn’t differed significantly. The calculated points form a cloud
give examples of these data in his paper. about an experimental curve as shown in figure 2.
Wigley had carried out fundamental arithmetical and Musker thought that the cause of this lay in the linearization
empirical verifications of Michell’s integral at the end of the of boundary conditions, but it will be shown later in this paper
1920s. For this purpose Wigley had developed a large series that there is another cause.
of different analytical models with constant draft, a
length/beam ratio L/B = 10.67 and beam/draught ratio
B/T = 1.5. The dependence upon x and z was represented as a 2. BRIEF SURVEY OF THE KNOWN AND OBTAINED
product of a function of x with a function of z thereby RESULTS CONNECTED WITH THE LINEAR THEORY
simplifying the calculation of Michell's integral. At the same AND THE MICHELL INTEGRAL
time Weinblum, having developed a series of analytical
models, carried out similar research. Having discovered a
good qualitative result of Michell’s integral the investigators The necessity of considering the discrepancy between the
turned to a systematical study of the linear theory of ship wave predictions and experimental data led us to the conclusion that
resistance and Michell’s integral in particular. there is a need to examine all the assumptions of the Michell
The ultimate goal of our investigation was to reveal the theory. With this aim, a careful analysis has been conducted.
cause of the discrepancy between Michell’s theory and the
experimental data at low Froude numbers. In addition, there (a) Effect of linearizing the boundary conditions on both the
was a need for a theoretical method to design the ship hull with ship hull and a free surface of a fluid
the least wave resistance.
It would appear reasonable to assume that presence of the
This paper contains six parts.
sharp humps and hollows in Michell’s curve is due to
1. The first part includes a brief survey of some well known
linearizing the boundary conditions. In this connection it is
and new results related to the linear theory and Michell
interesting to see the results of the calculations of the ship
integral.
wave resistance when both linear and nonlinear theories are
2. The second part describes a new form of Michell’s
used.
integral in which the main monotone part is separated
As is evident from the graph in figure 1, all the curves have
from the oscillatory one, which is directly connected with
the same humps and hollows as the Michell curve. Such
the interference of bow and stern wave systems of the
features of the wave resistance, derived by nonlinear theory,
ship.
were revealed when new sufficiently powerful computers were
3. The third part presents the results of the calculation of
developed.
Michell’s integral, assessment of these results, and the
immediate deductions.
0,45
4. In the fourth part a new hypothesis of the turbulent action
0,4
of ship waves is presented. The hypothesis did arise from
the fact that ship waves create the turbulence of flow 0,35
Cw x 1000

around the moving ship, a certain amount of the bow wave 0,3
energy being wasted for this. 0,25
5. In the fifth part a practical way to shape the ship hull 0,2
forms with the least wave resistance, using new criterion 0,15
obtained on the basis of accounts from part 2, is described. 0,1
It is illustrated by clear examples (Appendixes B and C). 0,05
6. The sixth part contains an investigation of the influence 0
the hull surface curvature exerts on the wave resistance. It 0,15 0,19 0,22 0,26 0,29 0,33 0,36
is based upon the use of a mathematical model of the hull
Fn
with a developable surface and on the analysis of the
integrand of Michell’s integral. Experiment
Figures 1A – 21A in Appendix A show the results of both Michell
the calculations and the experimental curves of residual Doubl. model
resistance for 21 models of Wigley and Weinblum. There is a Doubl. Mod. + Lin. Integral
measured curve, Michell’s curve, the main part of Michell’s Doubl. Mod. + Lin. Int. +
integral, and the curve of wave resistance taking into account Sources on the free surface
the viscous effect.
Finally, the explanation of the discrepancy between the Figure 1. Comparison calculations of wave resistance of the
prediction of wave resistance based on linear theory and parabolic Wigley model (by H. Maruo & K. Suzuki [1977],
experimental data has been obtained. figure 3).

Oceanic Engineering International 75

 The important problem of the wave-making resistance of a of the free surface conditions. However, the results for such
surface-piercing body seems to require a more realistic studies vary from one investigator to another.
representation of the free surface than is provided by the Peters and Stoker [1957] were the first to call attention to
classical linearized theory. There is an essentially analytic the existence of a line integral in obtaining the Green function.
approach to the nonlinear flow past a finite ship hull – the Next, Wehausen [1963] obtained a line integral in the formula
second order perturbation about a thin ship representation. for ship wave resistance. Then Maruo [1966], Eggers [1966],
Formulations of this type have been invented or developed by Kotik & Morgan [1969], Brard [1972], Kusaka. [1976], Bessho
Sizov [1961], Wehausen [1969], Brard [1972], Maruo [1966], [1976; 1994], Kutazawa & Takagi [1976] addressed the
Yim [1968], Eggers & Choi [1975], Guilloton & Noblesse contribution of the line integral to the wave resistance of a
[1975], Dagan [1975] and others. Comparison between wave ship.
resistances computed by these methods and experimental data Newman [1976] wrote in the discussion that the second-
published by Gadd [1973] and by Hong [1977]. Chapman order potential was generally expressed as a surface integral
[1977] provides a brief review of the papers devoted to the plus a line integral, and much debate has occurred regarding
analytical methods solving non-linear ship wave resistance the line integral.
problems. Bessho [1976] summed up the foregoing results on the line
integral of the velocity potential of a surface piercing body as
follows:
a) its source-term is canceled out by the one resulting from
the change of the wetted surface of the ship;
b) the sum of its doublet equals the virtual volume change of
the displacement;
c) it is not important at high speed, but at low speed the
contribution of the line integral is essential.
In addition, analyzing the work of Baba & Takekuma
[1975], Bessho came to the conclusion that the uniqueness of
representation of the ship hull by hydrodynamic singularities is
linked with the line integral.
There is an interesting physical interpretation in Maruo’s
[1966] work of the line-integral term as a necessity to fulfill
Figure 2. A collection of computer predictions of wave the continuity condition. He wrote that one can assume that
resistance for the Series 60 hull using linearized theory taken the line-integral term is a consequence of the linearization of
from the two workshops differ only in their method of solution the problem, and one can also note that this additional
(A.Musker [1989], figure 1). distribution of singularities does not cause any appreciable
change in the position of the humps and hollows in the wave
Non-dispersive waves exist around a ship’s bow. These resistance curve.
nonlinear, non-dispersive waves in the field near the ship bow Kotik & Morgan [1969] wrote that there is uniqueness for
are mostly attributed to the ship’s resistance and are named submerged bodies, but not for the case of surface-piercing
Free Shock waves. The papers of Baba & Takekuma [1969], bodies. They defined an exact singularity distribution potential
Miyata et al. [1980; 1981] deal with this problem. at zero Froude number for uniform flow incident on a double
In order to examine the role of linearization of the boundary body. The authors proved that for submerged bodies although
conditions on the free surface, the wave resistance with exact such singularity distributions on the body were highly non-
boundary conditions was calculated. The goal of this research unique, the associated wave resistance is unique. This is not so
was to examine the effect of neglecting squares of the for surface-piercing bodies, and an attempt was made to restore
perturbation velocities in the boundary conditions. The field of uniqueness by introducing integrals over the waterplane.
perturbation velocities obtained by the Hess-Smith method was Brard [1967] wrote that there are two problems requiring the
introduced into the calculation of ship wave resistance. These inclusion of the line integral. First, it is necessary when the
calculations showed that neglecting the squares of the question of the representation of the ship hull by hydrodynamic
perturbation velocities in the free-surface condition doesn’t singularities arises. Secondly, it is necessary to take it into
lead to any significant errors. account when the body pierces the free surface of a liquid.
Kusaka [1976] has found out in the process of numerical
(b) The contribution of the line integral to ship wave integration that for bodies piercing a free surface the total
resistance density of all hydrodynamic singularities representing the hull
is equal to zero only with only the line integral taken into
The following investigation is devoted to a study of the role account writing that
of the line integral taken along the main waterline in 1. the line integral gives a serious contribution to the density
computing the value of the wave-resistance of a ship. There of sources of the main hull just between the free surface;
was reason to hope that including the line integral would make 2. it brings better prediction of the wave-making resistance;
it possible for one to take partly into account the non-linearity 3. the line integral should be taken into account for

76 Oceanic Engineering International

 mathematical consistency of the boundary value problem. 2υ f x
Kutazawa & Takagi [1976] have found that the line integral σ ( x) = . (4)
seems to characterize a surface piercing vessel. This appears at
1 + f x2
the beginning of the formulations of the second-order potential,
corresponding to the change of the wetted surface or the Taking into account that
integration on the free surface, but the final formula includes
no line integral. This fact seems to imply that even if the fx (5)
nx = ,
wetted surface changes chiefly in the high-Froude-number 1 + f x2
range, the classical first-order potential can be a good
approximation for a surface piercing vessel. We do not need to
consider the effect of the change of the wetted surface in the the expressions for P(θ) и Q(θ) have the form
second-order potential.
The deduction given below is from the work of Wehausen & 1 f x3
Brard.
P(θ ) = ∫
ko L 1 + f x2
cos(kx)dx,
(6)
The velocity potential with a line integral has the form
1 f3
Q(θ ) = ∫ x 2 sin(kx)dx.
1 ∂φ ∂G ko L 1 + f x
φ ( x, y , z ) =
4π ∫∫ (G ∂n − φ ∂n )dS
S (1) Finally, the calculated formula for the ship wave resistance
1 ∂φ ∂G with line integral turns out as (1) where
4π ko ∫L
+ (G −φ )dy at z = 0,
∂x ∂x
I (θ ) = ∫∫ f x ( x, z ) ⋅ e− pz cos( kx) dxdz ,
D
where
S is the ship hull surface, J (θ ) = ∫∫ f x ( x, z ) ⋅ e− pz sin(kx)dxdz ,
L is the main waterline, D

g
ko = , and
υ2
υ is the ship velocity, and L
2 f x3 ( x, o)
G is the Green function. P (θ ) = ∫ ⋅ cos(kx)dx ,
From here wave resistance becomes −L
2
1 + f x2 (7)
L 3
2 f x ( x, o)
4ρ g
π
2 Q(θ ) = ∫ 2
⋅ sin(kx)dx .
−L 1 + f x
Rw = ∫ {I (θ ) + J 2 (θ ) +
2
2
πυ 2
0

υ 4
(2) The results of our calculation are shown in figure 3.
2
[ P 2 (θ ) + Q 2 (θ )]−
g
2υ 2 dθ
[ P (θ ) I (θ ) + Q(θ ) J (θ )]} 3 , Parabolic model
g cos θ 0,006

0,005
where
0,004
1
I (θ ) = ∫∫ σ ( x, y, z ) ⋅ e − pz
cos(kx)dS , 2
Cw

0,003 4
S 68
10
J (θ ) = ∫∫ σ ( x, y, z ) ⋅ e − pz
sin( kx) dS , 0,002

S (3) 0,001
P (θ ) = ∫ σ ( x, y, z ) ⋅ nx cos( kx) dy,
0
L
0,15
0,19
0,23
0,27
0,31
0,35
0,39
0,43
0,47

Q(θ ) = ∫ σ ( x, y , z ) ⋅ nx sin(kx) dy.

L Fn
Measur. Coef. Cw
Brard [1972] has obtained the densities of sources σ(x, y, z) Michell
from the integral equation, but in our set the first
Michell+lin. Int.
approximation of density σ(x,y,z) is taken in the form given by
Michell, i.e. σ(x,y,z) = 2 υ cos(n,x). The first approximation of
the source density in the line integral in using the linear theory Figure 3. Comparison of the wave resistance coefficients with
has the following form. and without the line integral (Wigley parabolic model).

Oceanic Engineering International 77

 Such calculations showed that the contribution of this 0,006
integral is negligibly small at low Froude numbers, but it is not
so at the high Froude numbers. This result is in conflict with 0,005

the Bessho [1976] conclusion. 0,004

Cw
0,003
(c) Experimental verification of Michell’s integral
0,002
The situation with experimental data is no better than with 0,001
calculations. In figure 4 one can see the wide scatter of the
experimental data for the same parabolic Wigley model from 0

0,21

0,27

0,33

0,39

0,45
different towing tanks. This scatter is so wide that almost the

0,15
0,18

0,24

0,30

0,36

0,42

0,48
whole Michell curve with its humps and hollows lies inside of Fn
the region of experimental points. These data were taken from Michell
a paper by Bai [1979].
Exper. Max
Chen & Noblesse [1983] made a comparison of the results
Exper. Min.
of nine different computations with the results of eleven
experiments carried out in various testing basins. Chen & Figure 4. A comparison between the coefficient of wave
Noblesse [1983] showed that the divergence of computations resistance curve and experimental data of the parabolic Wigley
from experimental data varied with different authors in the model (Bai [1979b] Fig.1).
interval 0.266 < Fn < 0.482 from 5% to 28% with the worst
divergence from 12% to 32%.
The discrepancy between the predicted and experimental 0,0016
value of the ship wave resistance may be due to the beam to 0,0014
length ratio that does not fulfill the Michell’s condition. 0,0012
0,001
Consequently, it is interesting to consider the results of the
Cw

0,0008
experiments with plates in which the beam to length ratio is 0,0006
very small. 0,0004
To determine the length a vessel should be to be considered 0,0002
a “thin ship” Weinblum et al. [1952] tested a model in the 0
David Taylor model basin, which they have called “a rough
0,12
0,20
0,28
0,36
0,44
0,52
0,60
0,68
0,76
0,84
plate”. The body had a length to beam ratio L/B = 37.67. Fn
In 1969 in the Hamburg tank S. Sharma carried out tests of a
parabolic model with length L = 2m, draught T = 0.3m, and Measur. Cw
breadth B = 0.1m.
Michell
In these two last cases the experiments were carried out for
Froude numbers greater than 0.20.
In those times, when the authors carried out these tests, the Figure 5. Comparison of a calculated Michell curve with
calculation of Michell's integral represented significant the experimental curve, obtained by Weinblum et al. [1952].
difficulties because of the limited capacity of computers.
Therefore we have prepared new exact calculations. The 0,004
results of comparisons with experimental curves are given in 0,0035
figures 5 and 6.
0,003
It is evident from figures 5 and 6 that for the thin ship model
under consideration there is fair agreement between the 0,0025
Cw

theoretical wave resistance calculated from Michell’s theory 0,002

and the empirical values if Fn > 0.23. Consequently, the 0,0015
Michell theory works well at high Froude numbers without 0,001
taking into account the viscosity of a fluid. 0,0005
However, neither of these experiments covers speeds, for
0
which the Michell curve has humps and hollows, that are
0,20

0,27

0,34

0,41

0,48

0,55

0,62

0,69

0,76

absent on experimental curves. Wigley and other researchers

explained this phenomenon by the fact that the theories are Fn
Measur. Cw
carried out without taking account of viscosity of a liquid. For
Michell
the last decades many papers have been written, in which the
wave resistance of a vessel is determined taking into account
viscosity (for example, Hino [1989], Shahshahan & Landweber Figure 6. Comparison of Michell’s calculated curve with
[1990] and others). experimental curve obtained by Sharma [1969].

78 Oceanic Engineering International

(d) Experimental investigation of the mechanism of effects of layer and wake was eliminated. It can be expected that the
viscosity on the ship waves experimental curve of these tandem struts will have the same
humps and hollows as the calculated one. The results of this
The next investigation was undertaken to find the effect of experiment are shown in figure 7.
the fluid viscosity on the ship wave resistance.
The humps and hollows in the wave resistance curve are a
consequence of the interaction of the bow and stern wave 0,20
systems. That is why this investigation is based upon a new 0,15
form of Michell’s integral in which a main monotone part of
wave resistance has been separated from an oscillatory

Rw
0,10
interaction part, as outlined in Section 3 of this paper. This
0,05
investigation incorporated the results of the experiment which
had shown that the boundary layer, the wake, and the hull 0,00
sheltering effect of a moving ship do not influence ship wave

0,14
0,16
0,18
0,20
0,22
0,24
0,26
0,28
0,30
0,32
0,34
interaction. In this work we also built on the results obtained
by Japanese researchers upon studying the waves around a Fn strut
Rw of one
moving vessel (Baba [1969; 1975; 1976], Inui [1981], Miyata
[1980; 1981]). Doubled Rw of one strut
It was known that Wigley [1938] and Havelock [1909; Rw of tandem struts
1948] and others believed that the primary source of the humps
and hollows in a calculated resistance curve is connected with Figure 7. Results of the experiment with two struts.
the assumption of a perfect fluid. They thought that the
boundary layer and the wake affected the interference of the But the result did not turn out as expected. The
bow and stern wave systems and smooth out the humps and experimental curve was monotonic, like the ship’s ordinary
hollows. For example, Havelock [1926] wrote wave resistance curve. There are three curves in figure 7. The
lower curve shows wave resistance of a single strut, the upper
“The direct effect of viscosity upon waves already one shows doubled wave resistance of the single strut, and the
formed may be assumed to be relatively small; the middle curve is the wave resistance of the two tandem struts.
important influence is one which makes the rear portion This experiment also showed that the interference of the
of the model less effective in generating waves than the bow and stern wave systems begins at about Fn = 0.23. Before
front portion. We may imagine this as due to the skin the velocity reaches this Froude number there is a region of
friction decreasing the general relative velocity of the calm water between the two wave systems. When the speed of
model and surrounding water as we pass from the fore the movement is increased the length of the cross waves also
end to the aft end; or we may picture the so-called increases and results in the stern system of waves entering
friction belt surrounding the model, and may consider sequentially from the fourth to the first wave of the bow
the general effect as equivalent to a smoothing out of system. When the aft strut enters the second wave of the bow
the curve of the rear portion of model”. system, the interference starts progressing rapidly. When the
second strut enters the first wave of the bow system, the waves
And in the paper [Havelock 1935] he wrote: have merged to become one dipole like wave system.
This experiment lets us conclude that the cause of the humps
“It seems fairly certain that one of the main causes of
the difference between theoretical and experimental and hollows isn’t related to the boundary layer and the wake of
results is the neglect of fluid friction in the calculation the ship.
It was quite natural to carry out the numerical check of the
of ship waves, and further that the influence of fluid
friction may be regarded chiefly as one which makes experimental results with struts. Would it be that in the
the rear portion of the ship less effective in generating calculated curve of the wave resistance of the struts there
waves than the front portion”. wouldn’t be any humps and hollows? For simplicity the
equation y = b f(x) is submitted for surface of struts, where b is
In order to verify this assumption we carried out a simple the half-beam of the model hull.
experiment with two struts with an aircraft profile (l = 39mm, The coordinates x, ξ are accepted as the first strut, the
b = 24 mm, [Gotman 1989]). These struts were situated in the coordinates x+L and ξ+L as the second strut, L is the distance
positions of the fore and aft perpendiculars (between them between the struts, and l is the length of the strut. The Michell
L = 0.915m). During this experiment the Froude number integral is written as
Fr = н gl of struts was very high and it is connection each π l l
4 ρ g 2b 2 2 2 T 2 T
strut produced the only Kelvin wave system. Hence there was R= Re ∫ sec3 θ dθ ∫ ∫ ∫ ∫ f x' ( x, z ) ⋅
the required simulation of two ship wave systems from tandem πυ 2
0 −l 0−l 0
2 2
(7)
struts. Between these wave systems there is obviously no ship fξ (ξ , ς ) ⋅ e
' − p (ς + z )
⋅e ik ( x −ξ )
dxdzdξ d ς ,
hull. Therefore, the influence of a ship hull with its boundary

Oceanic Engineering International 79

 where wake change a field of speeds and pressure at the stern wave-
making point, resulting in a reduction of amplitudes of the aft
g g (8) wave system.
p= ; k= .
υ 2 cos 2 θ υ 2 cosθ But how does all this influence the interaction of the bow
and aft ship wave systems? For the viscosity of a liquid to
As the struts are identical in length and form it is possible to influence the interaction of waves it is necessary for the wave
write amplitudes of the bow system to essentially decrease as they
pass along the ship hull. Our calculations have shown that the
π kinematics coefficient of viscosity is too small and is
4ρ g 2 2
Rw = b 2 Re ∫ sec3 θ dθ × practically unable to significantly affect the amplitudes of
π υ 2
0 moving waves.
l l In the experiment with struts the boundary layer, wake, and
2 T 2 T the sheltering effect of the ship’s hull are not present. The only
∫ ∫ ∫ ∫ f x' ( x, z ) fξ' (ξ , ς ) ⋅ e − p (ς + z ) × assumption is that the wave flow around the ship becomes very
−l 0−l 0
2 2 turbulent because of viscosity. The hypothesis was made that
⎡eikx + eik ( x + L ) ⎤ ⋅ ⎡ eikξ + eik (ξ + L ) ⎤ dxdzdξ d ς the ship waves create the turbulence of flow around the
⎣ ⎦ ⎣ ⎦ moving ship, and that a certain part of the bow wave energy is
(9) wasted.
π
The turbulent effect of waves is well known and was
4 ρg 2 2

= b 2 × Re ∫ sec 3 θdθ × described by Shuleikin [1958]. Dobroklonskii [1947] and

πυ 2
0 Shutilov [1941] have studied this phenomenon of ocean waves
1
T 1
T
and have estimated the turbulence coefficient. Apparently,
2 2

( x, z ) f ξ1 (ξ , ς )e − p (ς + z ) ×
similar experimental and theoretical work should be done for
∫∫∫∫f
1
x ship waves.
− 12 0 − 12 0 When it became apparent that the turbulence of a ship’s
[e ik ( x −ξ )
+ e ik ( x −ξ ) + e ik ( x −ξ + L ) + e ik ( x −ξ + L ) dxdzdξς ] waves is associated with the process of wave damping and
energy dissipation we discovered a great many existing
π
8 ρg 2 2
⎡ ⎤ T T theoretical and experimental results which support this
= Re ∫ sec 3 θdθ ×⎢ ∫ e pz dz ∫ e − pς dς ⎥ hypothesis. For verification one can explore numerous papers
πυ 2 0 ⎣0 0 ⎦ on this problem.
1
2
1
2
The complex structure of free surface shock waves (FSSW)

∫ ∫ f (x, z ) f ξ (ξς )e (cos kL + 1)dxdξ

1 1 ik ( x −ξ ) was studied by Miyata [1980], Mori [1979], Shin & Mori
x
[1989] and others. The following investigation was focused on
− 12 − 12
a study of turbulence near the free surface in different forms by
many authors including Madsen & Svendsen [1983], K.
It is easy to see from expression (9), that there is a factor Hanjalic & Launder [1972], Launder et al. [1975], Hung &
(cos kL+1). This factor should be reflected in the form of Buning [1985], Hunt & Graham [1978], Nakano [1988],
humps and hollows in the calculated curve, like the ordinary Schofield [1985], Tryggvason [1988], Visbal & Knight [1984],
calculated wave resistance curve. From this expression it Zhao & Zou [2001], and many others.
becomes evident that the smaller the Froude number the more The structure of the bow waves has received considerable
humps and hollows expected in the calculated curve. We have attention in a review of Miyata and Inui [1984]. They
to accept this proof because there is no way to calculate the observed that the near-surface flow structure and wave pattern
wave resistance of struts for the large Froude numbers depended strongly on the Froude number and the bow
Fr = н gl , where the Michell curve decreases. configuration. They also demonstrated that the near-surface
Our active search for the cause of humps and hollows in the flow had transformed into a turbulent flow downstream of the
calculated wave resistance curves within the framework of the wave crest. Their velocity measurement showed that this
theory of a perfect liquid has not resulted in success. Wehausen transition involved a significant loss of energy. They wrote:
[2001], when discussing the results of the experiment with
struts, persistently repeated that viscosity does influence the “The wave system is not like that of Kelvin’s waves in
wave resistance of struts. It has pushed us to look for a the near field of the ship model. Two remarkable waves
different approach to solving this problem, viscosity of a liquid are originated from the forepart of the model, and their
being taken into account. crest lines are nearly parabolic or straight. The
The viscosity of a liquid, undoubtedly, exerts its influence appearance of the water surface is quite different in
on the ship wave systems. It was confirmed with experiments front and behind the water crest line. Behind the wave
by Baba [1969] and Miyata et al. [1980; 1981] which have crest the water surface seems to be turbulent… The
shown that the first two waves of a bow system are of shock bow waves evidently show intense nonlinearity in a
character and this is proven by the loss of total head in contrast wide range of advance speeds. Their appearance is very
to Kelvin waves. It is also clear that the boundary layer and similar to a turbulent bore or a hydraulic jump”.

80 Oceanic Engineering International

The turbulent character of the flow near the bow of ship was dissipation. The turbulence on the free surface is
shown in figures 2 and 3 by Inui [1981]. concerned with this dissipative behavior. …A kind of
Our hypothesis about a turbulent character of the bow ship viscosity involving turbulence also diffuses the energy
waves is confirmed also by the experiments of Dong et al. deficit produced at the wave front in the rear region.”
[1997]. They wrote
There is no way to list all the necessary work but one can
“ …evident is the tendency of the flow downstream of find many more references in review articles by Peregrine &
the bow wave crest to turn towards the body and then Svendson [1978] and Banner & Peregrine [1993].
turn outwards again as it reaches the next wave crest.
The latter outward trend has also been demonstrated by (e) The accuracy of numerical integration of Michell’s integral
Miyata & Inui [1984], as part of their attempt to make
an analogy between a bow wave and an oblique shock. Yet another important calculation has been carried out. In
… in most cases the near-surface becomes turbulent order to check the calculations for accuracy, Michell’s integral
while crossing the wave, in agreements with Miyata & was calculated by two methods for an analytical model. In the
Inui [1984] for ship waves as well as Peregrine & first case the double integrals were integrated exactly
Svendsen [1978] for two-dimensional waves”. analytically in terms of known functions and in the second case
they were calculated numerically. The results emerging from
Turbulence of the ship’s bow waves has received this analysis showed that for the accuracy of the numerical
considerable attention. This problem became the purpose of integration to coincide with the exact value, it was necessary to
study by Allesandrini & Delhommeau [1994; 1995; 1996; divide the length between perpendiculars into 1000 parts and
1999], Brocchini & Peregrine [2001], Dana Dabiri [2000], the draft into 200 parts.
Dummermuth et al. [2000], and many others. There was a In figure 8 we can see what we would have if these numbers
Workshop devoted to this problem in March, 2000 in were 250 and 10 respectively. It should be pointed out that in
California. Some interesting studies have been conducted at our calculation the ordinates of a ship hull were calculated
Southern Queensland University by Mei & Roberts [1995; exactly and were not taken from a lines drawing. In the latter
1998] and others. case the errors are bound to be much larger. It is very difficult
Some investigations of shock and breaking waves refer to get the necessary accuracy because the Michell integral is a
directly to the problem of turbulence. Miyata and Inui [1984] Stieltjes integral and hence the integrand and the surface of
wrote: integration must be smooth without any broken lines. Any
jumps in the derivatives of the surface hull equation initiates a
“A free surface shock wave is supposed to have four fictional Kelvin wave system, which is absent in reality. As a
developmental stages, namely, consequence we obtain more wave resistance from more
(i) formation of very steep nonlinear waves, subdivisions. If the number of frames and waterlines is very
(ii) breaking or damping of the wave crest and great, the jumps in the derivatives smooth out, and the account
occurrence of energy deficit, becomes exact.
(iii) diffusion of energy deficit with turbulence and
sometimes air entrainment on the free surface,
and 1. 00
(iv) formation of a momentum-deficient wake far 0. 90
behind.” 0. 80
0. 70
Further 0. 60
Cw

0. 50
“A part of the wave energy of the nonlinear steep wave 0. 40
is dissipated at the wave front and transformed into 0. 30
momentum loss far behind the ship; on the other hand 0. 20
another part of the wave energy is likely to be supplied 0. 10
to the dispersive linear wave system that propagates to 0. 00
0. 16
0. 18
0. 19
0. 21
0. 22
0. 24
0. 26
0. 27
0. 29
0. 30
0. 32
0. 34
0. 35

the far field. The nonlinear steepness of the nonlinear

wave is partly compensated by dispersive spread and is
Fn
partly eased by dissipation. Therefore, the waves of
ships possess both dispersive and dissipative Mi chel l
properties. The dissipation becomes dominant when N=10, L=50 by Si mpson
the free surface is turbulent and breaking or damping N=10, L=100 by Si mpson
of the wave crest is remarkably strong.”
N=20, L=100 by Si mpson
And again:
Figure 8. Comparison of results of numerical integration with
“It is supposed that a kind of viscous effect plays a different numbers of frames and waterlines (N is the number of
certain role in the process of wave damping and energy frames, L is the number of waterlines).

Oceanic Engineering International 81

 Hence, the use of numerical integration in wave resistance expression (10) we replace cos[k(x - ξ)] by the real part of exp
problems requests one to be careful. If there is a corner in the ik(x - ξ):
hull surface, the resulting wave is not fictional, of course.
π L L
4ρ g 2 2 2 z ( x) 2 z( x)
Rw = Re ∫ ∫ ∫ ∫ ∫ f x ( x, z ) ×
π υ 2
0 −L 0 −L 0
2 2 (14)
3. A NEW EXPONENTIAL FORM OF MICHELL’S
− p ( z +ς ) ik ( x −ξ ) dθ
INTEGRAL fξ (ξ , ς ) ⋅ e ⋅e d ς dξ dzdx .
cos3 θ

The humps and hollows of the calculated wave-resistance It is convenient to write this integral in relative coordinates
curve hamper the use of the Michell theory for analysis and where we have replaced x by 2x/L, z by z/T, y = f(x,z) by
optimization of the displacement ship hull. Seemingly, y = b•f(x,z) where b = B/2. Then
Michell’s integral doesn’t give such interference of wave
systems as experiments, and that is why, in order to make this gz g⋅L Tz
pz = = 2 ⋅ =
problem clear, it appeared necessary to separate its oscillatory υ cos θ υ cos θ L ⋅ T
2 2 2

and nonoscillatory parts. The traditional form of Michell’s 1 T z z

⋅ ⋅ = po ⋅ ,
integral gives no way of determining such a separation, and for Fn 2 ⋅ cos 2 θ L T T (15)
this purpose a new form has been obtained. gx g⋅L x
Michell obtained the wave-resistance integral in the form kx = 2 = . =
υ cosθ 2 ⋅ υ cosθ 2 L
2
π L L 1 x x
4ρ g 2 2 2 z ( x) 2 z( x) ⋅ = ko ⋅ ,
Rw = ∫ ∫ ∫ ∫ ∫ f x ( x, z ) fξ (ξ , ς ) 2 ⋅ Fn 2 cosθ L L
π υ 2
0 −L 0 −L 0 2 2
2 2 (10)
− p ( z +ς ) dθ where po = 1 T 1 In the further
e cos[k ( x − ξ )]d ς dξ dzdx . ⋅ , ko = .
cos3 θ Fn 2 ⋅ cos2 θ L 2 ⋅ Fn 2 cosθ
index “o” in po and ko falls.
After expanding the cosine of the difference and doing the Then the integral (14) can be written more conveniently in
regrouping, Michell found the integral in the following form the form
2
π 4 pg 2 ⎛ B ⎞ π
dθ
4ρ g 2 dθ Rw = ⎟ ⋅ T × Re ∫0 J 1 (k , p ) ⋅ J 2 (k , p ) (16)
2 2 2

2 ⎜
Rw = ∫0 [ I + J ] cos3 θ , (11)
2 2

π υ 2 πυ ⎝ 2 ⎠ cos 3 θ

where
where
1 z( x)
L − pz +ikx
I (θ ) ⎫ 2 z( x)
cos(kx) ⎫
(12)
J 1 ( x, k , p ) = ∫ ∫ f x ( x, z ) ⋅ e ⋅ dz dx ,
⎬= ∫ ∫ f x ( x, z ) ⋅ e − pz ⎬ dzdx, −1 0 (17)
J (θ ) ⎭ − L 0
sin(kx) ⎭
2 1 z( x)
− pς −ikξ
J 2 ( x, k , p ) = ∫ ∫ fξ (ξ ,ς ) ⋅ e ⋅ dς dξ .
where −1 0
ρ is the mass density of the liquid,
g is the acceleration of the force of gravity, For computation and analysis these expression may be
υ is the velocity of the ship, conveniently written as
L is the hull length, 1
z(x) is the equation of the zero buttock, J 1 ( x, k , p ) = ∫ J 3 ( x, p ) ⋅ e ikx dx ,
y = f(x,z) is the equation of the hull surface, −1 (18)
B is the beam, 1
T is the draft, and J 2 ( x, k , p ) = ∫ J 3 ( x, p ) ⋅ e −ikx dx ,
−1
θ is the angle between the direction of the moving ship and that
of a propagating wave, and where
p and k equals
z( x)
− pz (19)
g g (13)
J 3 ( x, p ) = ∫ f x ( x, z ) ⋅ e ⋅ dz .
p= 2 2
, k= 2
. 0
υ cos θ υ cosθ
The separation of the oscillatory part from the main
The new form of the integral is obtained when in the (nonoscillatory) part in the wave-resistance integral (16) may

82 Oceanic Engineering International

be easily demonstrated for a symmetrical Wigley model with m −1 g ( n ) (1) ⋅ eik
constant draft along the whole length of the ship hull. The σ3 = ∑ (−1) n ⋅
n =0 (ik ) n+1
equation of the surface of the hull may then be written as a
product of functions of z and x m −1 g ( n ) (1) ⋅ eik
∑ (−1) n , (26)
n =0 (−ik ) n +1
y = b f2 (x) f1 (z) (20)
m −1 g ( n ) (−1) ⋅ e −ik
σ4 = ∑ (−1) n ⋅
In this case the integrals (19) depends only upon p and is n =0 (ik ) n +1
equal to
m −1 g ( n ) (−1) ⋅ e −ik
∑ (−1) n .
1 n=0 (−ik ) n+1
J 4 ( p ) = ∫ f1 ( z ) ⋅ e − pz ⋅ dz , (21)
0 By regrouping the products in (26) of the resulting
series according to the expressions [(exp 2ik + exp (-2ik)]/2
and the integrals (18) are simplified and may be written as and [(exp2ik - exp(-2ik)]/2 we then have
products:
e 2ik + e −2ik e 2ik − e −2ik
J1 ( k , p ) =
1
J 4 ( p) ∫ f 2' ( x)eikx dx = J 4 ( p ) J 5 (k ), cos 2k = , sin 2k =
−1 (22) 2 2i
1
J 2 (k , p ) = J 4 ( p ) ∫ f 2' ( x)e −ikx dx = J 4 ( p) J 6 (k ) . which leads to a representation of the wave resistance in the
−1
form
These integrals are then integrated by parts, which gives the 4ρ g 2 B 2 2 (27)
Rw = ( ) ⋅ T ⋅ ( I1 + I 2 ) ,
final expression as either a series or perhaps polynomials. To πυ2 2
simplify the notation we have introduced
where
g ( x ) = f 2' ( x ) (23) π
2 dθ
I1 = ∫ J 42 ( p) ⋅F1 ( k ) ,
0 cos3 θ
The series may be written as
(28)
π
1 2 dθ
J 5 (k ) = ∫ g ( x) ⋅ eikx dx = I2 = 2 ∫ J 42 ( p) ⋅ (F2 (k ) cos 2k + F3 (k ) sin 2k ) .
−1 0 cos3 θ
(n) ik (n) − ik
∞ g (1) ⋅ e − g (−1) ⋅ e Here, the indexes “bo” and “b1” relate to the bow and “so”
∑ (−1) n ,
n =0 (ik ) n +1 (24) and “s1” relate to the stern of the hull. Then
1
J 6 (k ) = ∫ g ( x) ⋅ e− ikx dx = F1 ( k ) = k −2 [Gbo
2
+ k −2 Gb21 + Gso
2
+ k −2 Gs21 ] ,
−1 (29)
F2 ( k ) = k − 2 [Gbo ⋅ G so + k − 2 Gb1 ⋅ Gs1 ] ,
∞ n g
(n)
(1) ⋅ e −ik − g ( n ) (−1) ⋅ eik
∑ ( −1) . F3 ( k ) = k − 3 [Gbo ⋅ Gs1 − Gb1 ⋅ Gso ] .
n =0 (−ik ) n +1
To obtain the expressions Gbo, Gso, Gb1, and Gs1 the
If f2(x) is polynomial of degree m, its derivative of order following transformations are made with the sums in (25),
more m equals zero and in (24) there remain only m terms in which can be written in the form
the summations. The product of the series (or polynomials)
(24) gives terms containing the products m−1 g (n) (1)
∑ (−1)n = − ig(1)k −1 + g' (1)k −2 −
n=0 (ik )n+1
m −1 (n) m −1 (n) (30)
g (1) g (1) ig '' (1)k −3 − g ''' (1)k −4 − ig IV (1)k −5 + gV (1)k −6 =
σ 1 = ∑ ( −1) n ⋅ ∑ ( −1) n ,
n =0 (ik ) n +1 n =0 ( −ik ) n +1
(25) k −2[ g ' (1) − g ''' (1)k −2 + gV (1)k −4 − ...] −
m −1 g ( n ) ( −1) m −1 g ( n ) ( −1) k −1[ g(1) − g'' (1)k −2 + g IV (1)k −4 ...] ,
σ 2 = ∑ ( −1) n ⋅ ∑ ( −1) n ,
n =0 (ik ) n +1 n =0 ( −ik ) n +1

in which the exponential functions do not occur, and which m−1 g(n) (−1)
∑ (−1)n =ig(1)k −1 + g' (1)k −2 −
gives the main (nonoscillatory) part of Michell’s integral. The n=0 (−ik )n+1
expressions containing the exponential functions and that ig'' (1)k −3 − g''' (1)k −4 + ig IV (1)k −5 + gV (1)k −6 =
(31)
ultimately yield the oscillatory part of Michell’s integral are
k −2[ g' (1) − g''' (1)k −2 + gV (1)k −4 − ...] −
ik −1[ g(1) − g'' (1)k −2 + g IV (1)k −4 ...] .

Oceanic Engineering International 83

 For the sums at the stern perpendicular J 5 (k ) ⋅ J 6 (k ) = k −2 [Gbo
2
+ k −2Gb21 + Gso
2
+ k −2Gs21 ] −
k −2 [Gbo ⋅ Gso + k −2Gb1 ⋅ Gs1 ] ⋅ 2cos 2k − (36)
m −1 g ( n ) ( −1) m −1 g ( n ) ( −1)
∑ ( −1) n
n +1
, ∑ ( −1) n
n +1
(32)
k −3 [Gbo ⋅ Gs1 − Gb1 ⋅ Gso ] ⋅ 2sin 2k =
n =0 (ik ) n=0 ( −ik )
F1 ( k ) − 2[ F2 ( k ) cos 2k + F3 ( k )sin 2k ].
similar expressions are obtained.
In order to compress the writing it is convenient to introduce This product gives the formulas (27), (28), and (29), which
the following notations are the new form of Michell’s integral.
In the obtained expression the oscillatory part is separated
from the nonoscillatory part, and furthermore, the bow
m −1 g ( n ) (1)
∑ ( −1) n = k − 2 ⋅ Gb1 − ik −1 ⋅ Gbo , contribution is separated from that of the stern. The interaction
n=0 (ik ) n +1 of the bow and stern systems is determined by the sum of the
m −1 g ( n ) (1) second and third members of formula (36).
∑ ( −1) n = k − 2 ⋅ Gb1 + ik −1 ⋅ Gbo , Such representation of the integral is convenient for the
n=0 ( −ik ) n +1 (33)
investigation of the linear theory of wave resistance and of
m −1 g ( n ) ( −1) interaction of the bow and stern wave systems.
∑ ( −1) n = k − 2 ⋅ G s1 − ik −1 ⋅ G so ,
n=0 (ik ) n +1 If the ship’s drought is a variable quantity, the formula of
wave resistance has the form
m −1 g ( n ) ( −1)
∑ ( −1) n = k − 2 ⋅ G s1 + ik −1 ⋅ G so ,
n =0 ( −ik ) n +1 4ρ g 2 B 2 2
Rw = ( ) ⋅ T ⋅ (Q1 + Q2 ) , (37)
πυ2 2
where
where the derivative can be written as
Gb1 = g ' (1) − g ''' (1)k −2 +
gV (1)k −4 − ...(−1)r +1 g (2r −1) (1)k −(2r −2) , g(x,z) = fx(x,z) (38)
−2
Gbo = g(1) − g (1)k + ''
and then the integral (19) has another form
⎧⎪(−1)r g (2r ) (1)k −2r , if m = 2r + 1,
g IV (1)k −4 ... ⎨ z( x)
r −1 (2r −2)
⎪⎩(−1) g (1)k −(2r −2) , if m = 2r, Φ ( x, p ) = ∫ g ( x, z ) ⋅ e − pz dz =
(34) 0
(39)
Gs1 = g ' (−1) − g ''' (−1)k −2 + ∞ g z( n ) ( x, z ( x))e − pz ( x ) − g z( n ) ( x, o)
n
∑ ( −1) ,
gV (−1)k −4 − ...(−1)r +1 g (2r −1) (−1)k −(2r −2) , n=0 (− p) n +1
Gso = g (−1) − g '' (−1)k −2 +
where z = z(x) is the equation of the zero buttock.
⎧⎪(−1)r g (2r ) (−1)k −2r , if m = 2r + 1,
−4 Formulas (18) and (19), then, have another more
g (−1)k ... ⎨
IV
r −1 (2r − 2)
⎪⎩(−1) g (−1)k −(2r −2) , if m = 2r. complicated form because the derivatives of g(x,z) are
depending on x through x and through z(x). Instead of equation
(22) now
The last member of the series Gbo and Gso depend on
whether the degree m of the polynomial f2(x) is even m = 2r or 1
odd m = 2r+1. J 1 ( k , p ) = ∫ Φ ( x, p ) ⋅ e ikx dx ,
Now the series can be easily presented in the form −1 (40)
1
m −1 g ( n ) (1) ik J 2 ( k , p ) = ∫ Φ ( x, p ) ⋅ e − ikx dx ,
∑ ( −1) n e =[ k −2 ⋅ Gb1 − ik −1 ⋅ Gbo ] eik , −1
n=0 (ik ) n +1
m −1 g ( n ) (1) The main part has the form
∑ ( −1) n e − ik =[ k −2 ⋅ Gb1 + ik −1 ⋅ Gbo ] e − ik ,
n=0 ( −ik ) n +1 (35)
π
m −1 g ( n ) ( −1) − ik 2 dθ
( −1) n e =[ k −2 ⋅ Gs1 − ik −1 ⋅ Gso ] e − ik , (41)
∑
n=0 (ik ) n +1
Q1 = ∫ F1 (k , p) cos 3 θ
0
m −1 g ( n ) ( −1) ik
∑ ( −1) n e =[ k −2 ⋅ Gs1 + ik −1 ⋅ Gso ] e ik .
n=0 ( −ik ) n +1 and the trigonometric part is
π
Then the product of the integrals in equation (24) is obtained 2 dθ (42)
Q2 = 2 ∫ (F2 (k , p ) cos 2k + F3 ( k , p ) sin 2k ) ,
in the form 0 cos3 θ

84 Oceanic Engineering International

where F1(k,p), F2(k,p), and F3(k,p) have the form of equation length L = 4.5 m, beam B = 0.30 m and draught T = 0.12 m.
(29), but Gbo, Gso, Gb1, and Gs1 have been written in the form As it easy to see from the expressions above in section 3, in
the calculation derivatives of the equation of the hull surface
Gb1 = Φ'x (1, p) − Φ'''x (1, p)k −2 + ΦVx (1, p)k −4 − taken at the bow and stern perpendiculars occur. These
...(−1)r +1 Φ(2 r −1)
(1, p)k −(2r −2) , derivatives reveal the fundamental geometric and numerical
x
characteristics of the hulls and through them we can determine
Gbo = Φ(1, p) − Φ''x (1, p)k −2 + Φ IV
x (1, p)k
−4
the dependence of the wave resistance, as obtained from
⎧⎪(−1)r Φ(2 r)
x (1, p)k
−2 r
, if m = 2r + 1, Michell’s formula, upon the ship’s hull.
... ⎨
r −1 (2 r − 2)
⎪⎩(−1) Φ x (1, p)k −(2r −2) , if m = 2r, (43) The values of the calculated derivatives for all these models
are given in tables 3 and 4.
Gs1 = Φ'x (−1, p) − Φ'''x (−1, p)k −2 + ΦVx (−1, p)k −4 − The analysis of the results of these calculations allows one
...(−1)r +1 Φ(2
x
r −1)
(−1, p)k −(2r −2) , to draw the following important conclusions:
Gso = Φ(−1, p) − Φ''x (−1, p)k −2 + Φ IV −4 1) An evident first result that is obtained from the new
x (−1, p)k ...
form of Michell’s integral is related to the peculiarities
⎪⎧(−1) Φ x (−1, p)k , if m = 2r + 1,
r (2 r ) −2 r

⎨ of wave formation around the moving ship. Michell’s

r −1 (2r − 2)
⎪⎩(−1) Φ x (−1, p)k −(2r −2) , if m = 2r. integral is adequate to represent the influence of a hull
form on the wave resistance of ships. It is well known
that two systems of ship waves are generated from two
wave-making points near bow and stern, as though from
a source and sink. Of course, the whole ship hull takes
4. ANALYSIS OF CALCULATION RESULTS part in the wave making, but its effect reveals itself in
these points. It is easy to see from equations (30), (31)
and (34) that the influence of the ship hull on its wave-
The form obtained for Michell’s integral is convenient for
making is centered on the fore and aft perpendiculars
carrying through an analysis of the dependence of wave
where the derivatives of the equation of the main
resistance upon the form of the hull according to its principal
waterline are determined. This important property of
non-oscillatory part, since in this case there is excluded a non-
Michell’s integral is impossible to extract from its
adequate account of the interference of the bow and stern
traditional form [Michell 1898].
systems of ship waves.
2) If the waterlines near the ends of the hull are convex,
For better visualisation of the manner in which the non-
Michell’s curve passes above the experimental one. If
oscillatory part of Michell’s integral correlates with the
they are concave, Michell’s curve is located under the
experimental data, calculations were carried out for 21 models
experimental curve. If waterlines are straight, only
investigated in towing tanks by Wigley [1926; 1927; 1930;
slightly convex or only slightly concave in these
1931; 1942; 1944] and Weinblum [1930; 1932; 1950; 1952].
regions, Michell’s curve is near the experimental ones
The results of these calculations are shown in Appendix A.
for Fn < 0.29. From this we conclude that for good
Comparative calculations of the wave resistance were
agreement with the experiment it is necessary to take
completed for analytically given hulls of Wigley models
into account second derivatives of the equation of a hull
[1942; 1944]. Their equations have the form
surface, which are connected directly to curvature of
η = b (1 − ξ 2 )(1 + a 2 ξ 2 + a 4 ξ 4 )(1 − ς 2 ), where b is the this surface.
half-beam of the model hull. 3) If the Froude numbers are above 0.29 - 0.33, the
A mean curve of the wave resistance of the parabolic model computed values of the main and oscillatory parts of
was obtained as a result of analyzing these experiments. Baar Michell’s integral begin to increase by several orders of
& Price [1988] reworked the results of measurements of wave magnitude, and the Michell resistance becomes a small
resistance by means of pressures and with the aid of wave difference of large values. This process begins when
analysis. They obtained a curve that we have used for the stern wave-generating point enters the first wave of
comparison with computations. Data for the Wigley models, the bow system. In this case the ship length is as long
for which calculations and analysis have been provided, are as, or is smaller than, the length of the ship wave and
listed in table 1. the ship generates waves like a dipole. Such a wave
Model N43 differs from the others by a greater fullness, pattern is observed experimentally as well.
which is evident in figure 8A (Appendix A). Except for model It is the principal difference between the flow
2038C, V-shaped sections and sufficiently great draught around the ship hull at the low and the high Froude
characterize the rest of the Wigley models. numbers. The Froude numbers 0.29 – 0.33 are the
Almost all Wigley models are geometrically similar. Hence, boundary of the separation for a study of wave
calculations were also carried out for Weinblum models resistance. The behaviour of Michell’s curve at the low
[1932], which embrace a greater diversity of hull forms. The and at the high Froude numbers differs because the
data for Weinblum models are given in table 2. interaction of the bow and stern systems. It is an
All the models listed in table 2 have length L = 4.5 m, beam explanation why the low speed theory is needed. All
B = 0.45 m and draught T = 0.18 m. Only model 1093 has this confirms that the Michell integral is adequate to

Oceanic Engineering International 85

 represent the singularities of the flow around the ship errors of numerical integration. It is very difficult to get
hull. The new form of Michell’s integral makes it the necessary accuracy because the integrals from
easier to become aware of these properties than the equation (17) are Stieltjes integrals and hence the
generally accepted form. integrand and the surface of integration must be smooth
4) From equation (27) and the expression (13), (29) one without broken lines.
can see that the lowest-order Froude number in 10) The Michell theory gave very poor results for only three
Michell’s wave resistance is six. However, an of the 21 models. This is easy to explain for the Wigley
approximation of the main part curve by the polynomial model 2038C that has a very small draught
allows determining its degree depending on the Froude T/L = 0.03125, for this contradicts the requirements of
number in every case. The dependence of the main part Michell theory. Weinblum’s model 1112 has a bulb in
on the Froude number varies through a large range from the bow region, making it difficult to apply the Michell
3 to 10 at low Froude numbers. It is the influence of theory. It is very difficult to explain the divergence of
the hull form on the behaviour of the calculated curve. computation and experiment for the Wigley model N43.
5) Comparison of the behaviour of the experimental For all it appears that a confirmation of the
curves and the main part of the calculated ones for experimental results is necessary.
different hulls ( as seen in figures 9 through 12) reveals 11) The main part of Michell’s integral practically
that the relationship between these curves has been coincides with the experimental curve at low Froude
preserved as the Froude number varies. This latter result numbers (0.15 – 0.3) for models 1110 (figure 16A) for
allows taking the main part of Michell’s integral as a Froude numbers up to 0.26 for models 1097, (figure
basic guideline in deciding upon the ship hull form if 13A), 1098 (figure 14A), 1113, (figure 19A), 1136
the Froude number is between 0.3 and 0.33. (figure 21A), 829 (figure 1A), 1846b (figure 2A),
6) The higher order derivatives of the equation of the hull 1970b (figure 6A), and a parabolic model (figure 10A)
have a substantial influence on the components of the for Froude numbers that are smaller than 0.3 and 0.33.
wave resistance. In addition, their influence increases 12) The results of our investigations showed that comparing
with an increase in Froude number; the higher the order theory with the experiment in the example of a single
of the derivative, the more this influence begins to show model is not sufficient for judging different theories.
itself with greater values of the Froude number. The analysis given of Michell’s integral was developed
7) For the Froude numbers between 0.36 and 0.40, firstly to determine whether linear theory gives the main part
Michell’s integral agrees well with experimental values of the wave resistance, and secondly to analyze the reason for
and does not require special treatment. the non-adequate calculation in Michell’s integral of the
8) If the smoothness of the surface of integration is interaction of the bow and stern systems of waves. The answer
disturbed in the process of integration over the surface to the first question is given by the behavior of the main non-
of the hull, then fictitious systems of waves form, oscillatory part of Michell’s integral. The tests of 21 models of
making a contribution to the computed value of the various shapes showed that Michell’s integral gives the main
wave resistance, so that any numerical method must be part of the solution of the problem of the wave resistance of a
tested on analytically exact models. ship, correctly reflecting the pattern of wave formation. The
9) The analytical models of Wigley and Weinblum were study of the distortion of the pattern of interaction of wave
chosen with the goal of excluding from analysis the systems is the purpose of the following research.

Model number Length L, m Beam B, m Draught T, m a2 a4 Volumetric coeff. Cv

Parabolic 4.8766 0.4572 0.3048 0 0 0.444

829 4.8766 0.4572 0.3048 -0.2000 0 0.427
1805a 4.8766 0.4572 0.3048 -0.6000 0 0.391
1805b 4.8766 0.4572 0.3048 -1.0000 0 0.355
1846a 4.8766 0.4572 0.3048 0.6000 0 0.498
1846b 4.8766 0.4572 0.3048 0.2000 0 0.462
1970b 4.8766 0.4572 0.3048 0.4375 -0.4375 0.467
1970c 4.8766 0.4572 0.3048 0.8125 -1.3125 0.467
2038c 4.8766 0.5334 0.1524 -0.5000 0.400
N43 4.8766 0.4572 0.3048 η = (1-ζ )(1-ξ )(1+0.2ξ ) + ζ (1-ζ ) (1-ξ2)4
2 2 2 2 8

2130a 4.8766 0.4572 0.3048 η = (1-ζ2)(1-ξ2)(1+0.4375ξ2± 0.5ξ3- 0.4375ξ4)

Table 1. Basic characteristics of the Wigley models.

86 Oceanic Engineering International

Model number Equation of hull surface Wetted area S

1093 η = (1-ξ4) (1- 0.4ξ2)⋅(1- 0.5ζ⋅ξ2)⋅ (1-ζ3) 1.585

1097 η = (1-ξ4) (1- 0.4ξ2)⋅(1- 0.5ζ⋅ξ2) 2.385
1098 η = [(1-ξ4) (1- 0.4ξ2)⋅-0.5(ξ3- ⋅ξ5)]⋅ (1- 0.5ζ⋅ξ2)⋅(1- ς3) 2.390
1100 η = (1-ξ4) (1- 0.4ξ2)⋅(1- 0.5ζ⋅ξ2)⋅ (1- 0.564ζ4-0.436ζ8) 2.50
1110 η = (1-ξ4) (1- 0.4ξ2)⋅(1- ζ⋅ξ2)⋅ (1- 0.564ζ4-0.436ζ8) 2.45
1111 η = (1- 2.46ξ4 +1.46ξ6) (1- ζ⋅ξ2)⋅ (1- 0.564ζ4-0.436ζ8) 2.425
1112 η = (1-ξ2) (1- ξ2ζ +3ξ6ζ) (1- 0.564ζ4-0.436ζ8) 2.421
1113 η = (1+0.1ξ2 – 1.995ξ4 – 0.895ξ6) (1- ζ⋅ξ2)⋅ (1- 0.564ζ4-0.436ζ8) 2.491
1114 η = (1- 0.8285(ξ2 + ξ4 - ξ6) –0.1715⋅ξ8)⋅ (1- 0.564ζ4-0.436ζ8) 2.444
1136 η = (1-ξ4) (1- 0.4ξ2)⋅(1- ζ12)⋅ [1- 0.5(ζ+ζ3) ⋅ζ2] 2.614

Table 2. Basic characteristics of the Weinblum models.

Model number f’(1) f’’(1) f’’’(1) fIV(1) fV(1) fVI(1) ao

Parabolic -.20 -.20 0 0 0 0 10o38’

829 -0.16 0.00 0.48 0.48 0 0 8o00’
1805a -0.08 0.40 14.40 1.44 0 0 4o17’
1805b 0.00 0.08 2.40 2.40 0 0 0o00’
1846a -0.32 - 0.08 -1.44 -1.44 0 0 16o41’
1846b -0.24 -0.40 -0.48 -0.48 0 0 12o41’
1970b -0.20 0.15 3.15 13.65 31.65 31.65 10o38’
1970c -0.10 1.35 10.65 42.13 94.50 94.50 5o20’
2038c -0.10 0.30 1.20 1.20 0 0 5o20’
N43 -0.24 -0.40 -0.48 -0.48 0 0 12o41’

Table 3. Derivatives of the equations of the Wigley models.

Model number f’(1) f’’(1) f’’’(1) fIV(1) fV(1) fVI(1) fVII(1) fVIII(1)

1093 -0.160 -0.107 1.600 8.000 19.20 19.20 0 0

1097 -0.240 -.080 2.400 12.000 28.80 28.60 0 0
1098 -0.14 0.62 5.100 18.000 34.80 28.80 0 0
1100 -0.240 -0.080 2.400 12.000 28.80 28.80 0 0
1110 -0.240 -0.080 2.400 12.000 28.80 28.80 0 0
1111 -0.108 1.428 11.616 46.640 105.12 105.12 0 0
1112 -0.200 -0.200 0 0 0 0 0 0
1113 -.0.241 0.311 5.952 27.440 64.44 64.44 0 0
1114 -0.137 0.365 2.191 -0.974 -55.60 -286.1 -691.5 -691.5
1136 -0.240 -0.080 2.400 12.000 28.80 28.80 0 0

Table 4. Derivatives of the equations of the Weinblum models.

Oceanic Engineering International 87

 0,006

WB1093
WB1097
0,005
WB1098
WB1100
0,004 WB1110
WB1111
WB1112
Cw

0,003 WB1113
WB1114
WB1136
0,002

0,001

0
0,1505

0,1618

0,1731

0,1844

0,1957

0,2070

0,2183

0,2296

0,2409

0,2521

0,2634

0,2747

0,2860

0,2973

0,3086

0,3199

0,3312
Fn

Figure 9. Main values of the wave-resistance coefficients of the Weinblum models.

0,0035
WB1093
WB1097
0,003
WB1098
WB1100
0,0025
WB1110
WB1111
0,002 WB1112
Cw

WB1113
WB1114
0,0015
WB1136

0,001

0,0005

0
0,1505

0,1618

0,1731

0,1844

0,1957

0,2070

0,2183

0,2296

0,2409

0,2521

0,2634

0,2747

0,2860

0,2973

0,3086

0,3199

0,3312

Fn

Figure 10. Experimental values of the wave-resistance coefficients of the Weinblum models.

88 Oceanic Engineering International

 1,4
829*
1805A
1,2
1805B
1846A
1
1846B
1970B
0,8 1970C
Cw

2038C
0,6 N43

0,4

0,2

0
0,16

0,17

0,18

0,19

0,20

0,21

0,22

0,23

0,24

0,25

0,26

0,27

0,28

0,29

0,30

0,31

0,32

0,33
Fn

Figure 11. Main values of the wave-resistance coefficients of the Wigley models.

1,2

829*
1 1805A
1805B
1846A
0,8
1846B
1970B
Cw

0,6 1970C
2038C
0,4 N43

0,2

0
0,16

0,17

0,18

0,19

0,2

0,21

0,22

0,23

0,24

0,25

0,26

0,27

0,28

0,29

0,3

0,31

0,32

0,33

Fn

Figure 12. Experimental values of the wave-resistance coefficients of the Wigley models.

Oceanic Engineering International 89

5. HYPOTHESIS OF TURBULENT VISCOSITY interested in research of viscosity influence on both
APPLICATION progressive and ship waves. It is natural to assume that this
influence is of a different character. The progressive waves
keep their form even moving large distances. The ship waves
It has already been mentioned that we assumed from the appear from the circular waves generated by the source
beginning that viscosity plays an important role in the movement. Their length and amplitude vary during removal
experiment with struts. It is presupposed that in this from the source.
experiment viscosity influences a wave field by smoothing Let us consider progressive waves [Milne-Thomson 1960]
humps and hollows in an experimental curve of wave whose height on the surface of a ideal liquid is small (waves of
resistance. infinitesimal amplitude) and has the form
Thus, we took into consideration that the bow waves are
losing part of their energy while moving along the ship hull to η = a sin(mx – nt) (45)
its stern by the turbulence. It makes us take the turbulence into
account in wave resistance. But, a very simple method is where m is the wave number, determined by the formula
required for this purpose because the exact definition of the m = g/υ2, and n = g/υ is the frequency of waves at a speed of
ship wave’s turbulence represents an unsolvable task. distribution υ .
A lot of semi-empirical methods have been offered to take The account of the tangential stress arising in a viscous fluid
into account the influence of viscosity on ship’s wave is expressed as an exponential factor. The height of a wave
resistance. Havelock [1935] and Wigley [1937-8] were the first enters as
who began to enter some corrections with this aim. Havelock
[1935] assumed a reduction factor to the slope of the hull η = ao exp(-2ν m2t) ( sin(mx – nt) (46)
surface ∠y/∠x in the form β(x), which is unity at the bow and
decreases towards the stern. Wigley [1937-8] has divided the where t is the time after the moment of waves arising and is the
wave resistance of a ship into components such as kinematical coefficient of viscosity.
Cw = Cwb + Cwi + Cws where Cwb is the wave resistance due If ν = 1.22.10-6, the height of the second wave practically
to the bow wave system, Cws is the wave resistance due to the has no change because the exponential factor is close to unity,
stern wave system and Cwi is the wave resistance due to the (i.e. e–0.0007 ≈ 1). Even the height of the fourth wave will
interaction of the bow and stern wave systems. He entered the change very insignificantly, some e–0.0021 times. Hence, the
correction factors in the form Cw = Cwb + αβCwi + βCws. viscous properties of a fluid do not noticeably influence
Both α and β are constants less than unity. Then Wigley progressive waves.
assumed α = β. The correction for viscosity was obtained as A large body of research has exposed that the distinction
∆Cw = -(1-β2)(Cw – Cwb). between movement in ideal and in a viscous liquid shows itself
Wigley gave an empirical formula for the reduction factor as as the factor depending on the kinematical coefficient of
follows viscosity.
Lamb [1947] gives a factor in which viscosity is taken into
β2 = exp(-0.001 Fn-5) (44) consideration in the calculation of progressive waves

where Fn is the Froude number ψ VgL . He determined this (

exp − 2vg 2 t υ 4 ) (47)
factor by means of comparison with experimental data for
models with simple mathematical lines. If time t = L/υ, i.e. the time, necessary for a wave to pass
Unlike Wigley’s assumption, Emerson [1954] considered the distance equal to length of a ship hull, then
factor α a general correction factor in the wave system, which
is independent of β. Furthermore he employed another type of g2 L 2ν
viscosity effect that is a virtual extension of the stern. Inui exp(−2ν ) = exp(− ). (48)
[1957] brought out a similar idea. He regarded factor α to υ4 υ Fn5 L3/ 2 g 1/ 2
have no relation to the viscosity effect but to represent some
non-linear effect that appeared from the finite beam instead of Cumberbatch [1965] investigated the effect of viscosity on
infinitesimal one in Michell’s theory. He named it a self- ship waves. He has shown that the exponential decay rate is
interference coefficient and gave a semi-empirical formula the main effect of viscosity and that the damping of the
α = exp[-0.40 (B/L)Fn-2]. transverse wave system varies little, whereas the diverging
However, the question of how to enter these correction system is more heavily damped. He wrote that the main effect
factors of viscosity to the calculation expression is much more of the viscosity dependent term is an argument that gives rise
interesting. From this point of view the works devoted to the to an exponential viscous decay factor exp(-rνg Bo cosθ/v 3),
calculation of waves and wave resistance in a viscous fluid are where the factor (Bo cosθ)-1 is found to vary only slightly from
pursued further. the value 0.25 over most of the transverse wave of Kelvin’s
Research on influence of viscosity on waves and wave system, ν is a kinematical coefficient of viscosity, r is a
resistance has been done in different directions. We are distance from a source of perturbation.

90 Oceanic Engineering International

 Nikitin & Podrezov [1964] have carried out some very character of the flow is taken into account and due to the
important research on surface waves of a viscous fluid of impossibility of its exact determination one simply replaces v
infinite depth. They have obtained the expression of the wave for the coefficient of turbulent viscosity turb. For example, with
height in the form such a replacement one can see the diffusion of vortices in the
problem [Kochin et al. 1963]. According to this the
Qσ
3
8σ 5ν r
η= exp(− )× exponential factor in our case has the form
ρg 2
2π gr g3
3
k turb = exp(−vturb / 2 / L 2 / g 2 / Fn 5
1
π σ 2r 1 (49) (55)
sin[ − + σ t ][1 + o( )+
4 g λr
where νturb is half of the turbulence coefficient of viscosity.
4r
o( ) + o( g2 )]. This factor enters into the oscillatory interaction part of
gt 2 σ r Michell’s integral I2 of equation (28) through F2(k) and F3(k)
in the form
where σ =
g
.
υ F2 (k ) = k −2 [Gbo ⋅ Gso + k −2Gb1 ⋅ Gs1 ] ⋅ kturb ,
Using the following symbols δ we can write (56)
F3 (k ) = k −3 [Gbo ⋅ Gs1 − Gb1 ⋅ Gso ] ⋅ kturb .
8σ 5ν r 8 g 2ν r
δ= = . (50) but does not enter in the main part because the bow waves
g3 υ5 begin lose energy after their formation thereby creating flow
turbulence.
Then the factor taking viscosity into account has the form The coefficient for turbulent viscosity in our calculations
was determined by a trial and error method. In the process of
exp(-δ) = exp(- 8g ν r ).
2
(51) our calculation we had chosen the coefficient of turbulent
υ5 viscosity to obtain a good agreement with experimental curve
of wave resistance. This coefficient has changed in the
To compare the expression of progressive waves we have sufficiently tight limits for the different models.
It is known that the turbulent viscosity is a hundred
2π 2π g 2 r 2 g 2ν r thousand or even a million times more than the molecular one.
δ = 2ν m 2t = 2v( ) 2 t = 2ν ( ) = . (52)
I.A. Kibel, in a study by Fedjaevskii et al., provided us with a
λ 2πυ 2 υ υ5
real life description of turbulent viscosity when he wrote “the
coefficient of turbulent viscosity of air equals the coefficient of
Then exp(-) = exp(- 2g ν r ).
2
It is easy to see that the the usual molecular viscosity of syrup, and the associated
υ5 kinematical coefficient of turbulent viscosity of water (is equal
argument turns out to be 4 times less. to) the kinematical coefficient of molecular viscosity of shoe-
The ship waves appear to be the result of movement of a polish”. That is why it was not a great surprise when this
pressure impulse or a source. Their form is determined by coefficient turned out to be about 0.08 – 0.16 rather than
transformation of the circle waves when moving the source 1.22k × 10-6.
(Newman [1977], Pavlenko [1956]). For this reason in order It is important that the parameter of the Froude number
to take viscosity into account we turn to the solution of the degree is equal to 5. This factor, from equation (55), is in
Cauchy-Poisson problem found by Sretenskii [1941] by using agreement with the Wigley factor of β = exp(-A/Fn5) obtained
the Navier-Stokes equations. We used his formula (48), where from testing many models [Wigley 1938]. It is even more
η is a wave height important that the turbulent factor kturb is entered only in the
5 oscillatory interaction part of Michell’s integral from equation
ν g 2t
1 g St − 2 x4 gt 2 π (28). For this purpose the new exponential form of the Michell
η= e cos( − ) (53)
integral is used.
2 π x 32 4x 4
Our calculations were performed for many Wigley and
Weinblum models and were in a good agreement with
If we presupposed that time t, in which the bow wave experimental data. In Appendix A the symbol NT in figures
reaches the aft wave-making point, and substitute it in the 1A – 21A is the coefficient of turbulent viscosity νturb. In order
Sretenskii formula (48), we obtain the factor in the following to have a trustworthy method of calculating the ship wave
form resistance it is necessary to obtain a coefficient of turbulent
viscosity for the real ship hulls.
(
exp − v / 2 / L 2 / g 2 / Fn 5
3 1
) (54) To summarize, the purpose of our last research was to
determine if it is possible to include the correction factor that
It is clear that the coefficient of molecular viscosity does not accounts for the viscosity of moving ship waves when
influence the dissipation of water waves. That is why we calculating the wave resistance of the ship. The second aim
replace v for turb. There are some problems where the turbulent was to define the coefficient of turbulent viscosity by trial and

Oceanic Engineering International 91

error and by doing so either validate or disprove the hypothesis 1. y = b{(1 + a 1 x 2 + a 2 x 4 + a 3 x 6) ×
of the turbulence effect of bow waves. The third aim of the
research was to determine the range in which the coefficient of (1 − 0,564 z S 1 − 0, 436 z S 2) −
turbulent viscosity varies. 0,5( x 2 + a 1 x 4 + a 2 x 6 + a 3 x 8) ×
Our calculations confirm the hypothesis that a certain part of
the bow ship wave’s energy is wasted on turbulence of a flow ( z S 3 − 0,564 z S 3+ S 1 − 0, 436 z S 3+ S 2)},
around the moving ship and does not participate in the
interaction with stern system of waves. The plausible 2. y = b{(1 + a 1 x 2 + a 2 x 4 + a 3 x 6) ×
explanation for the absence of humps and hollows in the
(1 − 0,564 z S 1 − 0, 436 z S 2) −
experimental curves of wave resistance of different models at
low Froude numbers has been obtained. ( x 2 + a 1 x 4 + a 2 x 6 + a 3 x 8) ×
The complex structure of the bow ship waves is the factor ( z S 3 − 0,564 z S 3+ S 1 − 0, 436 z S 3+ S 2)},
which complicates the development of an exact theory of wave
resistance for displacement ships.
The given research allows us to make the following 3. y = b{1 + a 1 x 2 + a 2 x 4 + a 3 x 6) ×
conclusion: the solution of a problem of wave resistance is (1 − z S 1)[1 − 0,5( z S 2 + z S 3) x 2 ]},
very complex if it is to be derived by the solution of a
boundary value problem with exact boundary conditions. 4. y = b[(1 − x 2) −
Apparently, it is necessary to carry out this problem in two ( x 2 + a1 x 4 + a 2 x 6 + a 3 x 8) z S 3)] ×
stages. In the first stage it is necessary to find wave resistance
(1 − 0,564 z S 1 − 0, 436 z S 2),
in an ideal liquid, and after this to take into account the fact
that the waves on a free surface become turbulent under the
influence of viscous properties of a liquid. A more or less 5. y = b(1 + a 1 x 2 + a 1 x 4 + a 2 x 6 + a 3 x 8) ×
exact definition of the coefficient of vortical or turbulent
(1 − 0,564 z S 1 − 0, 436 z S 2),
viscosity is the main difficulty and must be the purpose of new
research.
6. y = b[(1 − z S 1)(1 + a 1 x 2 + a 2 x 4 + a 3 x 6) +
4
( z S 2 − z S 3)(1− x 2) },
6. THE COMPARATIVE CRITERION IN DECIDING ON
THE SHIP HULL FORM WITH THE LEAST WAVE 7. y = b{(1 + a 1 x 2 + a 2 x 4 + a 3 x 6 −
RESISTANCE
0,5( x 3 − x 5)(1 − z S 1) −
0,5[( x 2 + a 1 x 4 + a 2 x 6 + a 3 x 8 −
As noted above in part 3, comparison of the main values of 0,5( x 5 − x 7)]( z S 2 − z S 3),
the coefficients of wave resistance with the experimental ones
for the Wigley and Weinblum models showed that the relation
between the main values and the experimental ones for small 8. y = b(1 + a 1 x 2 + a 2 x 4 + a 3 x 6)(1 − z S 1).
Froude numbers was basically preserved. Therefore, the main
part of the new form of Michell’s integral can be used as a To test whether the minimum of wave resistance does exist,
comparative criterion in deciding on ship hull form with the the equation of the wide Weinblum model 1102 was used
least wave resistance. To ground such method for optimisation (η = (1-ξ4) (1- 0.4ξ2)⋅(1- ζ12)⋅ [1- 0.5(ζ+ζ3) ⋅ζ2]) in the form of
of ship hull forms there was a need to ascertain that there is a equation (3). While we change the coefficient a2 and exponent
minimum of wave resistance for changes of the hull equation the S1, other coefficients a1, a3, and exponents S2 and S3 change
coefficients. The analysis has verified that such minimum of as well provided that the block coefficient δ, the draft T and the
wave resistance does exist. length L between load waterline perpendiculars are constant.
The comparative criterion was chosen as the ratio of the From here there was revealed that minimum wave resistance
main part of Michell’s integral Rg to displacement of ship D can be obtained provided the underwater volume of the ship
hull moves as close to the bottom or the middle section as
K g = R g / D. (57) possible, that entails a diminution of the waterline entrance
angle.
As noted in the first part of this paper, the application of Figure 13 shows dependence of the Kg = Rg/D on both the
numerical integration to the wave resistance problems coefficient a2 and the Froude numbers for the model 1102.
introduces large errors. Hence, the comparative criterion for The thick line connects points of the minimal values of wave
the real ship hulls cannot be practical for computer resistance on different Froude numbers.
calculations. Therefore, the primary source of this trouble is To find the hull shapes with the least wave resistance
the need to present a ship hull in an analytical form. For our programs were carried out in which the forms of frames or
purposes, eight equations were worked out in the form waterlines were changed while the block coefficient δ, the

92 Oceanic Engineering International

draft T, and the relative length L/B were constant. In our case At low speeds (in the first range of Froude numbers) the
all the models had length L = 4.5, B = 0.45, T = 0.18, and optimum waterlines have a concavity along the draught and the
block coefficient δ = 0.6. The choice of the hull form was frames turn out V-figurative. The variety of the hull form
carried out for three ranges of relative speeds: obtained by all these equations for Froude numbers 0.16 – 0.23
• 0,15 ≤ Fn ≤ 0,20 (low speeds); demonstrated that the hulls with the least wave resistance have
similar forms. Even the fourth equation has produced the hull
• 0,21 ≤ Fn ≤ 0,26 (middle speeds); without a bulb. It gives concave shapes on small Froude
• 0,27 ≤ Fn ≤ 0,31 (high speeds). numbers and bulb shapes on high ones. At middle speeds (in
The above equations allow us to change the form of the the second range of Froude numbers) waterlines are almost
waterlines from concave up to convex, and to change the form straight and the frames become U-figurative. At high speeds
of the frames from V- figurative up to U - figurative and even waterlines turn out convex and the frames are U-figurative.
up to a bow bulb or to a midship side bulb. However, not all For Froude numbers between 0.26 and 0.27, calculations
these equations are equivalent. For example, the fourth were made for all equations (1 through 8). It was found out
equation gives both the traditional hull and the bulb-bow form. that, at these Froude numbers, the waterlines are always
The sixth equation gives the hull with a midship side bulb and convex. The traditional forms have U-figurative frames, but
traditional hull too. The remaining six equations describe with some moving aside from the bottom to a main waterline.
traditional hull forms of displacement ships. The sixth equation gives the least criterion Kg among all of
For this research two means are chosen. The first way them. The best form has a midship side bulb. One can see an
involves a variation of the form of waterlines at preservation of example of the optimal forms obtained from equation 6 at
the form of frames, and the second way consists of varying the Fn = 0.27 in appendix figure 6B.
form of the frames at the chosen character of waterlines. Models with a bulb-shaped bow (fourth equation) have the
greatest value Kg at the given Froude number. Despite the
different forms, other equations give almost equal values of
wave resistance (the hulls with the least criterion Kg are shown
in Appendix B).
Besides, the choice of the hull form with the smallest Kg
value was made (with the help of the fourth equation) for
different Froude numbers. As it is easy to see from Appendix
C, the hulls vary from traditional forms on Fn = 0.15, 0.18,
0.23 up to the form with a bulbous bow on Fn = 0.27.

7. A THEORETICAL STUDY OF SHIP SHAPES

CURVATURE INFLUENCE ON WAVE RESISTANCE

The analysis of the waterline’s curvature influence on the

wave resistance was made by means of the Michell integral in
the Pavlenko [1937; 1956] representation and in doing this the
mathematical model of the ship hull with developable surface
was used.
The method of constructing a lines drawing and designing
the ship forms by using developable surfaces had been carried
out by Gotman [1979; 1985]. The comparative towing tank
tests showed the possibility of designing ship forms with DS
(by using developable surfaces) without any loss in
Figure 13. The graphical displays of Rg/D as a function of hydrodynamic quality. The analytical representation of ship
coefficient a2 for the third equation of the hull of Weinblum’s hull shapes with DS is a suitable mathematical model for
model 1102 with δ held constant. conducting theoretical research in ship hydrodynamics.

These calculations have allowed us to make the following (a) The mathematical model of the ship hull with developable
conclusions. surface and Michell’s integral
The minimum exists only for coefficients of waterlines. For
the majority of the equations, a minimum is not present at a The analytical expression of the ship’s hull with a
change of the form of the frames. The monotonous change of developable surface is the system of five equations [Gotman
the degree S1, S2, and S3 results in either a monotonous 1979]. Two equations of this system are given by the parallel
reduction or in a monotonous increase of wave resistance. basis sections of the hull surface. The third equation is the

Oceanic Engineering International 93

equality of the inclination angles of the lines of tangency to and aft are considered. In this case the integral G(θ) vanishes
these basis sections, passing through the points of the and the expression H(θ) has the form
intersection of basis sections with the generatrices. The
remaining two equations are equations of this straight ⎛ sin k 0 cos k 0 − 1 ⎞
sin k 0
generatrix itself. In general, this system may be solved by the H (θ ) = L1 + L2 ⎜ + 2 ⎟+
iterative method, but when the waterlines are chosen as basis k0 ⎝ k0 ko ⎠ (64)
sections the equation of a hull surface is given in the form ⎛ sin k 0 2(k 0 − sin k 0) ⎞
L3 ⎜ + 3 ⎟,
y = ( D1 z 3 + D 2 z 2 + D 3z + D 4xz + D 5 x 2z + ⎝ k0 ko ⎠
(58)
D 6x z 2 + D 7 x 2 + D 8x + D 9) /( D10 + D11z ) 2 , where

if the equations of the basis waterlines are presented by second L1 = a1F1 + a2F2 + a3F3,
order polynomials
L2 = a4F2 + a5F3 (65)
y = m0 + m1 x + m2 x 2 , on t = t1 ,
(59)
y = n0 + n1 x + n2 x 2 , on t = t2 L3 = a6F3

where t = z/T is relative draught. The coefficients in equation while the functions F1 , F2 and F3 are equal respectively
(58) depend on the parameters of equation (59) and the draught
of the basis waterlines.
F1 =
1 − e− p0 L − p 0 T
T
( L) e − p 0T
L

,
The Pavlenko representation uses the equation of ship hull
p (T L )
2
2
instead of its derivative as in Michell’s integral. 0

Since y = f(x,z) is taken in the form of equation (58), the (66)

inner integrals of Michell’s integral can be integrated with 1− e − p 0T
L
F2 = ,
respect to x and z, giving p T0 L
{E i[− ( a 7 + 1) p 0 T ] −
p T
F 3 = a7e
a
7 L

4ρ g 4 ⎛ L ⎞ ⎛ B ⎞ 2
2 2 0
L
R= T ×
π υ 6 ⎜⎝ 2 ⎟⎠ ⎜⎝ 2 ⎟⎠ (60) E i (a 7 p 0 T L)},
π
2 dθ
∫ ⎡⎣G (θ ) + H (θ ) ⎤⎦ The parameters a1 through a7 algebraically depend on the
2 2
,
cos θ
3
o coefficients of equation (59) and for t = 0 are as follows:

where n 0 − m 0 + ( m 1 − n 1) ,
2
a1 =
t2 4 t 2 (m 2 − n 2 )
G (θ ) = L1 S1 + L2 S 2 + L3 S 3 ,
m 2 (m 1 − n 1) ,
(61) 2
H (θ ) = L1Q1 + L2 Q 2 + L3 Q 3. a 2 = m0 −
4 (m 2 − n 2 )
2

m 2 (m 1 − n 1) ,
Here the functions S1, S2, S3, Q1, Q2, and Q3 are trigonometric 2

functions from ko, and the functions L1, L2, L3 depend on a3 =

4 (m 2 − n 2 )
2
relative draft, po and the form of the hull. Here (67)
n1 m 2 − m 1 n 2 ,
a4 =
1 1 m2 − n2
p0 = , k0 = . (62)
Fn cos θ
2 2
2 Fn 2 cosθ m 2 ( m 1 − n 1) ,
a5 =
m2 − n2
The study is carried out in the dimensionless coordinates a6 = m2 ,
n2 t2 .
2x 2y z a7 =
ξ= , η= , ς= . (63) m2 − n2
L B T
(b) Analysis of Michell’s integrand
It is supposed that a ship has no cylindrical middle body and
the forebody may be described by equation (58). In order to In order to determine the role of every component H(θ) in
simplify the analysis, the ship which has a symmetrical fore expression (64) it is necessary to evaluate the order of the

94 Oceanic Engineering International

trigonometric functions with L1, L2 and L3 and then the demonstration let us transform expressions (67) in order to see
absolute values and signs of the last. the role of a7 in each parameter (with t2 = 1) in the following
Let us consider the trigonometric expressions for ko→ 0, i.e. way:
when the velocity increases to infinity. It is easy to see that for
2
ko→ 0 the limits are equal. ( m 1 − n 1)
a1 = n 0 − m 0 + a7 ,
4 n2
sin k o
= 1, m 2 (m 1 − n 1 )
2
lim
k o →0 ko a2 = m0 − a7 ,
4 n2
⎡ sin k cos k o − 1⎤ 1 (68)
m 2 (m 1 − n 1 )
o 2
lim ⎢ + ⎥= ,
k o → 0 ⎣⎢ ko 2
k o ⎦⎥ 2 a3 = a7 ,
4 n 22
⎡ sin k o (70)
2( k o − sin k o) ⎤ 3 n1 m 2 − m 1 n 2 ,
lim ⎢ + 3 ⎥= . a4 = a7
k o → 0 ⎢⎣ ko ko ⎥⎦ 4 n2
m 2 ( m 1 − n1)
a5 = a7 ,
Hence these trigonometric expressions are changing from 0 n2
to 1, ½, and ¾ , respectively. a6 = m2 ,
If the velocity of steady motion is small, i.e. ko→ ∞, then
n2 .
these expressions are vanishing. a7 =
Expressions L1, L2 and L3 can have different signs which m2 − n2
depend on the behaviour of the functions F1, F2 and F3 and also
on the parameters from equation (58) which are connected with
the coefficients of the ship hull surface equation. The
functions F1, F2 and F3 depend on ship speed and a relative
draught. When the speed increases and p = T/(L×Fn2 cos2θ)
vanishes then

1 − e − p − pe − p 1
lim F 1 = lim 2
= ,
p →0 p →0 p 2 (69)
1 − e− p
lim F 2 = lim = 1.
p →0 p →0 p

The behavior of the function F3 is considerably complicated

because it depends not only on p but also on the parameter a7
which is connected with a surface curvature. The study of the
function F3 has been carried out in the work by Gotman [1979;
1985]. Here only the graph of the function F3 in respect to p
and a7 is represented in figure 14.
The effect of the functions F1, F2, F3, L1, L2 and L3 on the
value of the Michell’s integrand can be analysed only if these
functions are not oscillating when θ changes from 0 to π/2.
The analysis showed that the functions F1, F2 and F3 are
monotonically decreasing on this interval of θ variation. It is
easy to see by means of the signs of their derivatives. The
behaviour of the function F3 depending on θ was verified by Figure 14. The function F3 in respect to p and a7.
the calculation. The behaviour of the functions F1 and F2
turned out monotonical and decreasing. Let us consider the expressions L1, L2 and L3. From
The monotonical behaviour of the functions F1, F2 and F3 in equation (64) L1 = a1F1 + a2F2 + a3F3.
the process of integration with respect to θ provides the In this expression F1, F2 and F3 are positive and F1 is
monotonical behaviour of the expressions L1, L2 and L3. The approximately half as much as F2. From figure 14 it is seen
calculations of the different F1, F2 and F3 showed that this that the function F3 can be made as small as one likes. Thus,
functions decrease without intersecting. The expressions L1, L2 the order of the value L1 is defined more by the order of
and L3 behave in the analogous way. member a2 F2 than a1 F1 due to two causes. Firstly, a1 is
One can see from expressions (64), (65), and (66) that the defined by the differences mo - no and m1 – n1 which for
function F3 and the form parameter a7 play the leading part in conventional vessels are usually small and, secondly, F1 is half
the formation of the expressions L1, L2 and L3. For visual as much as F2. Parameter a2 depends on relative width mo,

Oceanic Engineering International 95

which can’t be diminished either by a7 or by any other where a7 is positive. The upper and lower waterlines have the
coefficients. Consequently, it is necessary for the diminution same curvature in this part of the graph and it means that all
of the order-of-magnitude L1 to select the curvature of upper waterlines curvature of the bow end are convex or concave and
and lower waterlines so that a2 should become as small as the upper waterlines curvature is greater than the curvature of
possible. Taking into account that the convex waterline has the lower ones. This occurs frequently in the case of
m2 < 0 one can get a diminution of a2 by the diminution or the conventional ship forms.
change of sign of m2 for positive signs that can be achieved at If a7 tends to infinity (when the waterline form is constantly
the expense of the upper waterline concavity. along the draft) only the load waterline form may decrease the
Figure 14 shows that when a7 decreases F3 simultaneously integrand. In this case we have the “simplified” hull shape
decreases. Hence, as can be easily seen from the expressions with a cylindrical fore body and the largest possible magnitude
(70) the order-of-magnitudes of the value L2 = a4 F2 + a5 F3 is H(θ) because if the displacement and the placement of the fore
diminishing directly at the expense of a7 and through F3 . perpendicular have been given the coefficients of the
The third expression L3 = a6 F3 depends on F3 and on the waterlines m0, m1, m2 are single valued and the integrand has
parameter a6 = m2, i.e. only on the hull surface curvature, the largest value. It is just the case when one considers the
namely on the load waterline curvature. wave resistance of the struts with the finite draft.
The analysis that was carried out led to the following The interval -1 < a7 < 0 is of the greatest interest. For the
conclusions sake of study this interval is divided into three parts (-1 -
1) The integrand H(θ) depends on the expressions L1, L2 0.7),(-0.7 - 0.5) (-0.5 - 0). When a7 belongs to the first part of
and L3, which are connected with hull form and the interval the integrand has very large magnitude increasing
exchange monotonically in the process of integration in to infinity.
respect to θ from 0 to π/2 so that one can decrease the The third interval, when a7 is equal to values from -0.5 to 0,
value of Michell’s integral by decreasing the order-of corresponds to the hull forms with the hull shape discontinuity
magnitude of these expressions. near the free surface. In this interval the function F3 is
2) The influence of the ship principal particulars on the negative. In this case the function F3, instead of decreasing the
magnitude of the Michell integral appears through the integrand, actually increases it.
relative draft T/L that plays the same role as the inverse The middle part of the interval of the value a7 from -0.7 to
value of the square of the relative speed. -0.5 gives the possibility to decrease F3 along with the
3) By varying the coefficients m2 and n2 that define basic integrand. The negative value a7 can be obtained only in the
waterline curvatures one can decrease the value of case when the curvature of the upper and lower waterlines has
Michell’s integral. The parameter of the hull form a7 different signs or if the curvature of the lower waterlines is
connected with the above coefficients determines the much larger than that of the upper ones. The bulb can achieve
order-of-magnitude of expressions L1, L2, and L3. this situation. This part of the graph in figure 14 indicates that
4) It is impossible to find the optimal relationship between the choice of the bulb shape is a very fickle problem.
the coefficients m0, m1, m2, n0, n1, and n2 which would The left part of the graph of the function F3, when a7 < -1,
be good for all ships. Having the principal parameters, belongs to case when the signs of the curvature of the upper
relative speed, and the main parameters of the hull form and lower waterlines are equal, but the curvature of the lower
presented as block coefficient, sectional area curve etc., waterlines is much greater than that of the upper ones.
one must choose the relationship between m0, m1, m2, Hsiung [1981] was looking for the optimal ship form with
n0, n1, and n2 so as to reduce a2, a6, a7, and F3. minimum wave resistance by means of the technique of
5) The fact that the coefficients m0 and m2 influence the mathematical programming and has obtained the best forms
order-of-magnitude of the integrand confirms once when a wing-like bulb is at z = 0.25T with a7 = -0.75. In this
more the importance of load waterline shape while case from figure 14 we have a sufficiently large value of the
designing the hull form. integrand and hence a large value of the wave resistance. The
experiment, which was conducted on such hull shapes by
(c) The influence of the coefficients of the ship fore body on Hsiung [1981], showed that in this case the resistance increases
wave resistance greatly and the hydraulic jump takes place as on the shallow
water.
As shown above, it is necessary to reduce a7 and F3 for the
reduction integrand of the Michell integral. For this purpose
there are different possibilities.
It is easy to see in figure 14 that when a7 = -1 the function 8. CONCLUDING REMARKS
F3 goes to infinity. Hence, it is undesirable to choose such
combinations of the coefficients that define the curvature of the The investigations of Michell’s integral showed the limits in
waterlines, which give a7 = -1. In real practice it may occur if which this integral can be used for practical purposes. The
the ship has a flat bottom and such basic waterlines shapes that very important experiment with two struts gave the answer to
the developable surface pulled on them is tangent to the base the question about the influence of viscosity of a fluid on the
plan. ship wave resistance. This experiment argued against the
Let us consider the right part of the graph F3 in figure 14 popular opinion that the humps and hollows are absent in an

96 Oceanic Engineering International

experimental curve because of action of the boundary layer, REFERENCES
wake, and the sheltering effect of the ship hull.
Our calculations confirm the hypothesis that a certain part of
the bow ship energy is wasted on turbulence of a flow around Aldogan, H.J. 1979 A nonlinear wave resistance theory and
the moving ship and does not participate in the interaction of its application. Schiffstechnik, 26, (79), 116.
the bow and stern wave systems. Hence revealing the reason Allen, R.F. 1968 The effects of interference and viscosity in
for the discrepancy between the Michell’s and experimental the Kelvin ship-wave problem. J. Fluid Mech., 34, 417-
curve. 421.
The new form of Michell’s integral made it possible to Allesandrini, B. & Delhommeau, G. 1994 Numerical
determine the previously unknown peculiarities of this integral. calculation of three-dimensional viscous free surface flow
If the Froude numbers are above 0.29 – 0.33 the computed around a series 60 CB=0.6 ship model. CFD Workshop,
values of the main and oscillatory parts of Michell’s integral Tokyo, 1-12.
begins to increase by several orders-of-magnitude and the Allesandrini, B. & Delhommeau, G. 1995 A multigrid
Michell resistance becomes a small difference of large values. velocity-pressure-free surface elevation fully coupled
This phenomenon reflects the radically different character of solver for calculation of turbulent incompressible flow
wave resistance and wave patterns at low and high speeds of around a hull. 9th Intern Conf. on Numerical Methods on
ships. Laminar and Turbulent Flow, Atlanta, 1173-1184.
Apparently, the turbulence of a flow plays a significant role Allesandrini, B. & Delhommeau, G. 1996 A multigrid
in wave resistance of models and ships. The turbulent velocity-pressure-free surface elevation fully coupled
condition of the water during tests in the basin causes a great solver for calculation of turbulent incompressible flow
deal of disorder of the experimental data received in different around a hull. 21st Symposium on Naval Hydrodynamics,
tanks. The wide scatter of experimental data for the same Trondheim, 328-345.
parabolic Wigley model obtained from different towing tanks Allesandrini, B. and Delhommeau, G. 1999 A fully coupled
(figure 4) is attributable to different intensities of fluid Navier-Stokes solver for calculation of turbulent
turbulence during the test process. incompressible free surface flow past a ship hull. Int.
The comparative criterion in choosing the ship hull form Journal for Numerical Methods in Fluids, 29, 125-142.
with the least wave resistance was obtained and verified for Baba, E. 1969 A new component of viscous resistance of
eight models outlined by different equations. With the help of ships. J. Soc. Nav. Arch., Japan, 125, 23-34.
these equations the optimum forms were obtained which are Baba, E. 1975 Analysis of surface flow near the bow of flat
well coordinated with the known forms carried out in tanks. It ships. Japan Shipbuilding and Marine Engineering, 9 (2),
was shown that the hull forms with a midship side bulb give 5-19.
the least wave resistance. Baba, E. 1976 Wave breaking resistance of ships. Intern.
The study of the influence of the ship hull surface curvature Seminar on Wave Resistance, Tokyo, 75-92.
on the wave resistance combined all possible forms of ship Baba, E. & Takekuma, K. 1975 A study on free-surface flow
hulls. It was shown that the “simplified” hull shapes with a of slowly moving full forms. J. Soc. Nav. Arch., Japan 137,
cylindrical fore body have the largest possible value of wave 1-10.
resistance. It was shown also that the designing of the fore Bai, K.J. 1979a Blockage correction with a free surface. J.
bulb is very capricious undertaking. Fluid Mech.- London, 94, 433-452.
As a result of this research it is possible to tell that, in Bai, K.J. 1979b Overview of results. Proc. Workshop Ship
connection with the difficulty of the solution of a boundary Wave-resistance Computations, I.
value problem with exact boundary conditions on a free Banner, M.L. & Peregrine, D.H. 1993 Wave breaking in deep
surface and on a surface of the hull, it is possible to try to solve water. Ann. Rev. Fluid Mech., 25, 373-397.
the problem about wave resistance in two stages. First we find Beck, R.F. 1971 The wave resistance of a thin ship with a
the wave resistance of a vessel in an ideal liquid, and then we rotational wake. J.Ship Research, 15, 196-214.
must take into account the influence of the turbulent viscosity Bessho, M. 1961 On the formula of wave-making force acting
on the interaction of wave systems. on a ship. Journal of the Society of Naval Architects of
Japan, 110, 65-73.
Bessho, M. 1976 Line integral, uniqueness and diffraction of
wave in the linearized theory. In Proc. Intern. Seminar on
Wave Resistance, Tokyo, 45-55.
ACKNOWLEDGMENTS Bessho, M. 1977 On the fundamental singularity in a theory
of ship motion in a seaway. Memoirs Def. Acad., Japan, 17,
95-105.
The author is sincerely grateful to Prof. John Wehausen for Bessho, M. 1994 On a consistent linearized theory of the
his active discussion, fruitful criticism and also for scientific wave-making resistance of ships. J.Ship Research., 38, 83-
literature, which he sent to Novosibirsk. Thank you also to 86.
Prof. Neil Bose for his help with the preparation of the Birkhoff, G., Korvin-Kroukovsky, B.V., & Kotik, J. 1954
manuscript. Theory of the wave resistance of ships. Trans. Soc. Nav.

Oceanic Engineering International 97

 Arch. and Mar. Engin., 62, 359-385, disc. 396. Eggers, K.W.H. 1976 Wave analysis state of the art 1975.
Birkhoff, G. and & Kotik, J. 1954 Some transformations of Intern. Seminar on Theor. Wave Resistance., Tokyo, 93-
Michell’s Integral. Publ. Nat. Tech. Univ. Athens, 10, (26). 103.
Brard R. 1967 Research in ship hydrodynamics. Eggers, K.W.H. & Choi, H.S. 1975 On the calculation of
Schiffstechnik, 14, 3-10. stationary ship flow components. First Int. Conf. on
Brard, R. 1970 Viscosity, wake and ship waves. J. Ship Numerical Ship Hydrodynamics.
Research, 14, 207-240. Emerson, A. 1954 The application of wave resistance
Brard, R. 1972 The representation of a given ship form by calculations to ship hull design. Quart. Trans. of Inst. Nav.
singularity distribution when the boundary condition on the Arch., 96, 268-283.
free surface is linearized. J. Ship Research, 16, 79-92. Emerson, A. 1967 The calculation of ship resistance. An
Brocchini, M. & Peregrine, D.H. 2001 The dynamics of application of Guillotons method. Trans. Roy. Inst. Nav.
strong turbulence at free surfaces, Part 1. Description. J. Arch., 109, 241-248.
Fluid Mech. 449, 225-254 Emerson, A. 1971 Hull form and ship resistance. Trans.
Brocchini, M. & Peregrine, D.H. 2001 The dynamics of North-East Coast Inst. Eng. Shipb., 87, 139-150, disc. D27-
strong turbulence at free surfaces, Part 2, Free-surface D30.
boundary conditions. J. Fluid Mech. 449, 255-290. Evans, D.V. 1975 The transmission of deep water waves
Calisal, S. 1972 Effect of wake on wave resistance. J. Ship across a vortex sheet. J. Fluid Mech., London, 68, 389-401.
Research, 16, 93- 112. Fontaine, E., Faltinsen, O.M., & Cointe, R. 2000 New insight
Calisal, S.M. & Chan, J.L.K. A numerical Modeling of Ship into the generation of ship bow waves. J Fluid Mech. 421,
Bow Waves. J. Ship Research.33, 21-28. 15-38.
Chapman, R.B. 1977 Survey of numerical solutions for ship Fontaine, E., & Faltinsen, O.M. 1997 Steady flow near a
free-surface problems. In Proc. Second Int. Conf. on wedge shaped bow. Abstract for the 12th Intern. Workshop
Numerical Ship Hydrodynamics, Berkeley, 5-16. on Water Waves and Floating Bodies, Marseille 75-79.
Chen, C.Y. & Noblesse, F. 1983 Preliminary numerical study Gadd, G.A. 1973 Wave Resistance Calculations by
of a new slender-ship theory of wave resistance. J. Ship Guilloton’s Method. Trans. Royal Institution of Naval
Research, 27, 172-186. Architects., London, 115, 377-392.
Cuthberg, J.W. 1964 Integral hull form parameters and a high Gotman, A. 1985 The design of hydroconic ship hull shapes
speed approximation to Michell’s integral. J. Ship with high hydrodynamics quality. In Proc. 14th Scientific
Research, 7, 12-15. and Methodological Seminar on Ship Hydrodynamics,
Cumberbatch, E. 1965 Effects of viscosity ship waves. J. Varna, 2 (53), 1-11.
Fluid Mech., London, 23, 471-479. Gotman, A. Sh. 1998 The comparative criterion in deciding
Dagan, G. 1975 A Method of Computing Nonlinear Wave on the ship hull form with least wave resistance.
Resistance of Thin Ships by Coordinate Straining. Journal Colloquium Euromech 374., Poitiers, France, 277-288.
of Ship Research, 19, 149-154. Grosenbaugh, M.A. & Yeung, R.W. 1989 Flow structure near
Dana Dabiri 2000 Free-surface roughness correlations with the bow of a two-dimensional body. J. Ship Research, 33,
the near-surface turbulence. ONR 2000 Free Surface 269-283.
Turbulence and Bubbly Flows, Workshop Agenda, Calif. Guilloton, R.A. 1946 Further notes on the theoretical
Inst of Technology, San-Diego, 1-3. calculation of wave profiles and of the resistance of hulls.
Doctors, L.J. 1998 Modifications to the Michell’s Integral. Trans. Inst. Nav. Arch., London, 88.
Improvement predictions of ship resistance. In Proc. of the Guilloton, R.A 1951 Potential theory of wave resistance of
Twenty Seventh Israel Conf. on Mech. Eng., Technion City, ships, with tables for its calculation. Trans. Soc Nav. Arch.
Israel, 502-506. And Mar. Eng., London, 59, 86-122, disc. 123-127.
Dommermuth, D.G., Sussman, M., & Novikov, E.A. 2000 Guilloton, R.A 1952 A note on the experimental
The numerical simulation of turbulent free-surface flows determination of wave resistance. Quart. Trans. Inst. Nav.
using Cartesian-grid methods. ONR 2000 Free Surface Arch., London, 94, 343-362.
Turbulence and Bubbly Flows, Workshop Agenda, Guilloton, R.A 1962 Examen critique des methods dėtude
California Inst. Technology, San-Diego, 35, 1-7. theoretique des carėns de surface. Schiffstechnik, 9 (45), 3-
Dong, R., Katz, J., & Huang, T. 1997 On the structure of bow 12.
waves on a ship model. Journal of Fluid Mech., 346, 77- Guilloton, R.A 1965 La pratique du calcul des une isobars sur
115. une carėne linėarisėe. Bull. Assoc. Tech. Maritime et
Eggers, K.W.H. 1962 Uber die Ermittlung des Aėron., 65, 379-400.
Wellenwiderstands eines Schiffsmodells durch analyse Hanjalic, K & Launder, B.E. 1972 A Reynolds stress model
seines Wellensyastems. Schiffstechnik, 9, 79-84 and 93- of turbulence and its application to thin shear flow. J.
106. Fluid. Mech. 52, 609-638.
Eggers, K.W.H. 1966 Second order contribution to ship wave Havelock, T.H. 1909 The wave-making resistance of ships: a
and wave resistance. Trans. 6-th Symp. Nav. Hydrod., theoretical and practical analysis. In Proc. Royal Soc. of
Washington, 649-672. London., Ser.A. 82 (A 544), 276-303.

98 Oceanic Engineering International

Havelock, T.H. 1926 Wave resistance: some cases of Schiffstechnik, 24 (118), 178-213.
unsymmetrical forms. In Proc. Royal Soc. of London., Inui, T. et al. 1979 Non-Linear Properties of Wave Making
Ser.A. 110 (A 544), 233-241. Resistance of Wide-Beam Ships. Journal of the Society of
Havelock, T. 1932 The theory of wave resistance. In Proc. Naval Architects of Japan, Tokyo, 146, 18-26.
Royal Soc. of London, Ser. A, 138, 339-348. Inui, T, 1981 From Bulbous Bow to Free-Surface Shock
Havelock, T. 1934a Wave patterns and wave resistance. Wave-Trends of 20 Years Research on Ship Waves at the
Trans. of Royal Inst. Nav. Arch., 76, 430-446. Tokyo University Tank. J.Ship Research, 25, 147-180.
Havelock, T. 1934b The calculation of wave resistance. In Kajitani, H. 1965 The second order treatment of ship surface
Proc. Roy. Soc. London., Ser. A. 144 (856), 514-521. condition in the theory of wavemaking resistance of ships.
Havelock, T. 1935 Ship waves. The relative efficiency of bow Journal of the Soc. of Naval Architects of Japan, Tokyo,
and stern. In Proc. Roy. Soc. London. Ser. A, 149, 417-426. 118, 84-107.
Havelock, T.H. 1948 Calculations illustrating the effect of Keller, J.B. & Ahluwalia, D.S. 1976 Wave resistance and
boundary layer on wave resistance. Trans. RINA, 90, 259- Wave Patterns of Thin Ships. J.Ship Res. 20, 1-6.
271. Kinoshita, M. 1976 Wave resistance in a viscous fluid derived
Havelock, T.H. 1951 Wave resistance theory and its from momentum analysis. Int. Seminar on Wave
application to the ship problem. Trans. Soc. Nav. Arch. Resistance, Tokyo, 295–298.
Mar Eng., 59, 13-24. Kitazawa, T. & Takagi, M. 1976 On the second order velocity
Hermans, A.J.& Brandsma, F.J. 1989 Nonlinear Ship Waves potential of the thin ship. Int. Seminar on Wave Resistance,
of Low Froude Number. J.Ship Research, 33, 176-193. Tokyo, 255–261.
Hess, J.L. 1977 Progress in the Calculation of Nonlinear Kotik, J. & Morgan, R. 1969 The uniqueness problem for
Free-Surface Problems by Surface-Singularity Techniques. wave resistance calculated from singularity distributions
In Proc. Second Int. Conf. on Numerical Ship which are exact at zero Froude number. J. Ship Research,
Hydrodynamics, 278-284. 13, 61-68.
Hinze, J.O. 1959 Turbulence in introduction to its mechanism Kusaka, V. 1976 On the contribution of line integral to the
and theory. wave resistance of surface ships. In Proc. Intern. Seminar
Hino, T. 1989 Computation of free surface flow around an on Wave Resistance, Tokyo 249-254.
advancing ship by Navier-Stokes equation. Fifth Numerical Lamb, H. 1932 Hydrodynamics. Cambridge.
Conference on Ship Hydrodynamics. Part 1. Hiroshima, Larsson, L. & Baba, E. 1996 Ship resistance and flow
69-82. computations. Advances in Fluid Mechanics, 5, 1-75.
Hong, Y.S. 1977 Numerical Calculation of Second-Order Launder, B.E., Reece, G.L. & Rodi, W. 1975 Progress in the
Wave Resistance. J. Ship Res. 21, 94-106. development of a Reynolds-stress turbulence closure. J.
Hsiung, C.-C. & Wehausen, J.V. 1969 Michell resistance of Fluid Mech., 68, 537-566.
Taylor’s standard series. 12th Int. Towing Tank Conf., Lonquet-Higgins, M.S. & Stewart, R.W. 1961 The changes in
Rome, 176-180. amplitude of short gravity waves on steady non-uniform
Hsiung, C.-C. 1981 Optimal ship forms for minimum wave currents. J. Fluid Mech., London, 10, 529-549.
resistance. J. Ship Research, 25 (2), 95-116. Lunde, J.K. 1951 On the linearized theory of wave resistance
Hsiung, C.-C. & Shenyyan, D. 1984 Optimal Ship Forms for for displacement ships in steady and acceleration motion.
Minimum Total resistance. J.Ship Research, 28, 163-172. Trans. of the Soc. Nav. Arch. and Mar. Eng., 59, 25-76,
Hughes, B.A. & & Stewart, R.W. 1961 Interaction between Disc. 77-86.
gravity waves and a shear flow. J. Fluid Mech., London, Lurye, J.R. 1968 Interaction of free-surface waves with
10, 385-400. viscous wake. The Physics of Fluids 10, 261-265.
Hung, C.-M. & Buning, P.G. 1985 Simulation of blunt-fin- Madsen, P.A. & Svendsen, I.A. 1983 Turbulent bores and
induced shock-wave and turbulent boundary-layer hydraulic jumps. J. Fluid Mech. 129, 1-25.
interaction. J. Fluid Mech., 154, 163-185. Maruo, H. 1966 A note on the higher order theory of thin
Hunt, J.C.R. & Graham, J.M.R. 1978 Free-stream turbulence ship. Bulletin of the Faculty of Engineering Yokohama Nat.
near plane boundaries. J. Fluid Mech., 84, 209-235. University., Yokohama, 15, 1-21.
Inui, T. 1952 A New Theory of Wave-Making Resistance, Maruo, H. 1973 Ship waves and wave resistance in a viscous
based on the Exact Condition of the Surface of Ships. J. of fluid. J. Soc. Nav. Arch. of Japan, 134, 15-29.
Zosen Kiokai., Japan, 85, 29-44. Maruo, H. 1976 Ship waves and wave resistance in a viscous
Inui, T. 1953 A New Theory of Wave-Making Resistance, fluid. Int. Seminar on Wave Resistance., Tokyo, 217-238.
based on the Exact Condition of the Surface of Ships. J. of Maruo, H. & Suzuki K. 1977 Wave resistance of a ship of
Zosen Kiokai., Japan, 93, 11-21. finite beam Predicted by the Low Speed Theory. J. Soc.
Inui, T. 1957 Study on wave-making resistance of ships. Soc. Nav. Arch. of Japan, 142, 1-7.
Nav. Arch. Japan, 60th Anniversary Series, 2, 173-355. Maruo, H. 1977 Wave Resistance of a Ship with Finite Beam
Inui, T. 1962 Wave making resistance of ships. Trans. Soc. at Low Froude Numbers. Bulletin of the Faculty of
Nav. Arch. Mar. Eng., 70, 283-352. Engineer Yokohama Nat.University, 26, 59-77.
Inui, T. & Kajitani, H. 1977 A Study on Local Non-Linear Mei, Z. & Roberts, A.J. 1995 Equations for turbulent Flood
Free Surface Effects in Ship Waves and Wave Resistance. Waves: in a book K.Kirchgössner Structure and dynamics

Oceanic Engineering International 99

 of nonlinear waves in fluids. World Sci., 342-352. Noblesse, F. 1978 The steady wave potential of a unit source
Mei, Z., Roberts, A.J. & Zhenguan, Li 2002 Modelling the dynamics at the centerplane. J. Ship Research, 22 (2), 80-88.
of turbulent floods. To appear in SIAM J.~Appl. Math. Ogilvie, T.F. 1967 Nonlinear high-Froude number free-
Michell, J.H. 1898 The wave resistance of a ship. surface problems. J. Engineering Mathematics, Groningen
Philosophical Magazine.-Ser. 5, 45, 106-123. 1, 215-235.
Michelsen, F.G. 1966 Asymptotic approximations of Peregrine, D.D.H. & Svendson, I.A. 1978 Spilling breakers,
Michell’s integral for high and low speeds. Schiffstechnik, bores, and hydraulic jumps. In Proc. 16th Coastal Engng.
13 (66), 33-38. Conf. ASCE, Hamburg, 1, 540-550.
Michelsen, F.G. & Kim H.C. 1967 On the wave resistance of Peregrine, D.H. 1987 Recent development in the modeling of
a thin ship. Schiffstechnik, 14, 46-49. unsteady and breaking water waves. In Nonlinear water
Michelsen, F.G. 1972 The application of Gegenbauer waves. Intern. Union of Theor. And Appl. Mech. IUTAM
Polinomials to the Michell Integral. J. Ship Research, 16, Symp. (ed. K. Horikawa, H. Maruo), Tokyo, 17-27.
60-65. Peters, A.S. 1949 A new treatment of the ship wave
Milne-Thomson, L.M. 1960 Theoretical Hydrodynamics. resistance. Comm. Pure and Appl. Math., 2, 123-148.
London: Fourth edition. Peters, A.S. & Stoker, J.J. 1957 The motion of a ship, as
Miyata, H., Inui T. & Kajitani H. 1980 Free surface shock floating rigid body in a seaway. Communic. Pure and
waves around ships and their effects on ship resistance. J. Applied Mathematics, 10, 339-490.
Soc. Nav. Arch. of Japan, 14 (7) 1-9. Savitsky, D. 1970 Interaction between gravity waves and
Miyata, H. 1980 Characteristics of nonlinear waves in the finite turbulent flow fields. Eighth Symposium Nav.
near fields of ships and their effects on resistance. In Proc. Hydrodyn, Pasadena, 389-446.
12th Symp. on Naval Hydrod.,Tokyo, 335-351. Shahshahan, A. & Landweber, L. 1990 Boundary-layer
Miyata, H., Suzuki, A. & Kajitani, H. 1981 Numerical effects on wave resistance of a ship model. J. Ship
explanation of nonlinear nondispersive waves around a Research, 34, 29-37.
bow. Third Intern. Conf. On Numeric. Ship Hydrod., Paris, de Sendagorta, M. & Grases, J. 1988 A Method for
37-52. Calculating the Michell and Havelock Integrals. J. Ship
Miyata, H. & Inui, T. 1984 Nonlinear ship waves. Advances Research, 32, 19-28.
in Applied Mechanics, 24, 215-288. Sharma, S.D. 1969 Some results concerning the wavemaking
Miyata, H. & Kanai, A. 1996 Numerical Analysis of the of a thin ship. J. Ship Research, 13, 72-81.
Structure of Free-Surface Shock Waves about a Wedge Shearer, R. 1951 A preliminary investigation of the
Model. J. Ship Research., 40, 278-287. discrepancies between the calculated and measured
Mori, K. 1979 Prediction of viscous effects on wave wavemaking of hull forms. Trans. East Coast Inst. Of Eng.
resistance of ship in framework of low speed wave And Shipbuilders in Newcastle-upon Tyne, 67, 43-68, D21-
resistance theory. Memoirs of the Faculty of Engineering D34.
Hiroshima Univ. Serial No 21, 7 (1), 9-19. Schofield, W.H. 1985 Turbulent-boundary-layer development
Mori, K. & Nishimoto, H. 1981 Prediction of flow fields in an adverse pressure gradient after an interaction with a
around ships by modified Rankine-source method. J. Soc. normal shock wave. J. Fluid Mechanics, London, 154, 43-
Nav. Arch. of Japan, 150, 9-18. 62.
Musker, A.J. 1989 Stability and accuracy of a non-linear Shin, M. & Mori, K. 1989 On turbulent characteristics and
model for the wave resistance problem. Fifth Numerical numerical simulation of 2-dimensional sub-breaking
Conference on Ship Hydrodynamics.- Part 2., Hiroshima, waves. J. Soc. Nav. Arch. of Japan, 165, 1-7.
437-450. Sretensky L.N. 1957 Sur la resistance due aux vagues d’un
Nakano, T. 1988 Direct interaction approximation of fluid visqueux. In Proc. Symp. on the behavior of ship in a
turbulence in the wave packet representation. J. Phys. seaway, Wageningen, 2, 729-733.
Fluids, 31 (6), 1420-1438. Suzuki, K. 1971 On the boundary conditions of wave making
Newman, J.N. 1964 The asymptotic approximation of resistance theory. Journal Soc. Nav. Arch. of Japan, 130,
Michell’s integral for high speed. J. Ship Res. 8, 10-14. 11-20.
Newman, J.N. 1976 Linearized wave resistance theory. Int. Tatinclaux, T.C. 1970 Effect of a rotational wake on the
Seminar on Wave Resist., Tokyo, 31-43. wavemaking resistance of an ogive. J. Ship Research, 14,
Newman, J.N. 1977 Marine hydrodynamics. England. 84-99.
Noblesse, F. 1975 A Perturbation Analysis of the Tryggvason, G. 1988 Deformation of a free surface as a result
Wavemaking of a Ship with an Interpretation of Guilloton of vortical flows. The Physics of Fluids, 31 (5), 955-957.
Method. J. Ship Research, 19, 140-148. Tuck, E.O. 1974 The effect of a surface layer of viscous fluid
Noblesse, F. & Dagan, G. 1976 Nonlinear ship-wave theories on the wave resistance of a thin ship. J.Ship Research, 18,
by continuous mapping. J. Fluid Mechanics., London, 75, 265-271.
347-371. Tuck, E.O. 1976 An approximation to Michell’s integral. Int.
Noblesse, F. 1976 A note on second-order, thin-ship theory Seminar on Theor. Wave Resistance, Tokyo, 239-244.
by centerplane source distributions. Intern Seminar on Tuck, E. O. 1989 The wave resistance formula of J.H.
Wave Resistance, Tokyo, 263-267. Michell (1898) and its significance to recent research in

100 Oceanic Engineering International

 ship hydrodynamics. Journ. of Australian Math. Soc. Ser. Wigley, W.C.S. 1944 Comparison of calculated and measured
B., Australia, 30, 365-277. wave resistance for a series of forms not symmetrical fore
Ursell, F. 1988 On the theory of the Kelvin ship-wave source and aft. Trans. of the Royal Inst. of Naval Arch., 86, 41-60.
asymptotic expansion of an integral. In Proc. Royal Society Wigley, W.C.S., 1948 Lunde, J.K. Comparison Calculated
of London, Ser. A., London, 418, 81-93. and observed wave resistance for a series of forms of fuller
Ursell, F. 1990 On the theory of the Kelvin ship-wave source: midsection. Trans Inst. of Naval Arch., 90, 92-110.
the near-field convergent expansion of an integral. In Proc. Wigley, W.C.S. 1962 The effect of viscosity on wave
Royal Society of London, Ser. A, London, 428, 15-26. resistance. Schiffstechnik, 9 (46), 69-72.
Visbal, M. & Knight, D. 1984 The Baldwin-Lomax model for Wigley, W.C.S. 1963 Effects of viscosity on wave resistance.
two-dimensional shock-wave. AIAA Journal, 22 (7), 921- Int. Seminar on Theor. Wave Resistance, Ann Arbor, III
928. 1295-1310.
Wehausen, J.V. 1957 Wave resistance of thin ship. Symp. Wigley, W.C.S. 1967 A note on wave resistance in a viscous
Nav. Hydrod., Washington, 109-137. fluid. Schiffstechnik, 14 (10).
Wehausen, J.V. 1963 An approach to thin ship theory. In Yeung, R.W., Xing Yu 2000 Three-dimensional free-surface
Proc. Int. Seminar on Theor. Wave Resistance, Ann-Arbor, flow with viscosity: a spectral solution. Hydrod. in Ship
2, 821-852. and Ocean Eng. RIAM, Kyushu Univ., 87-114.
Wehausen, J.V. 1969 Use of Lagrangian coordinates for ship Yim, B. 1968 Higher Order Wave Theory of Ships. Journal
wave resistance (first- and second-order thin ship theory). of Ship Research, 12, 237-245.
Journal of Ship Research, 13 (1), 12-22. Yim, B. 1974 A Simple Design Theory and Method for
Wehausen, J.V. 1973 The wave resistance of ships. Advances Bulbous Bows of Ships. Journal of Ship Research, 18 (3),
in Applied Mechanics, 13, 93-245. 141-152.
Weinblum, G.P. 1930 Schiffsform and Wellenwiderstanden. Zhao, C. & Zou, Z. 2001 Computation of waves generated by
In Proc. Third Int. Congr. Appl. Mech., Stockholm, 449- a ship using an NS solver. Proc. 16th Intern. Workshop on
458. water waves and floating bodies, Hiroshima, Japan, 181-
Weinblum, G.P. 1932 Schiffsform und Wellenwiderstanden. 184.
Jarbuch der Schiffbautech, Jarbuch, Gessellschaft, 33, 419-
451.
Weinblum, G.P. 1950 Analysis of wave resistance. David
Taylor Model Basin, Navy Department, Report 710, RUSSIAN LITERATURE
Washington.
Weinblum, G.P., Kendrick J.J. & Todd M.A. 1952 Антимонов К.И. (1935) Теория Мичелла и ее приложение
Investigation of wave effects produced by a thin body – к расчету волнового сопротивления судов//Труды
TMB Model 4125. Navy Department, the David W. Taylor ВНИТОСС.- M., 1 N 4, 19 – 59.
Model Basin, Washington 7.DC, Report 840. Биркгоф, Г. 1963 Гидродинамика Изд. Ин. Лит. М.
Wigley, W.C.S. 1926 Ship wave resistance. A comparison of (Birkhoff, G. 1963 Hydrodynamics)
mathematical theory with experimental results. Trans. INA, Готман, А.Ш. 1995 Определение волнового сопротивления
68, 124-137. и оптимизация обводов судов. - Новосибирск: НГАВТ,
Wigley, W.C.S. 1927 Ship wave resistance. A comparison of Части 1 и 2 320 c. ( Gotman, A.Sh. 1995 Determination of
mathematical theory with experimental results. Trans. INA, the Wave Resistance and Optimization of Ship Hulls.
69, 191 – 210. Novosibirsk, NSAWT Parts 1 and 2 320 p.).
Wigley, W.C.S. 1930 Ship wave resistance. A comparison of Готман, А.Ш. 1979 Проектирование обводов судов с
mathematical theory with experimental results. Trans. INA, развертывающейся обшивкой: Ленинград,
72, 216-228. Судостроение, 192 с. (Gotman, A.Sh. 1979 Design of the
Wigley, W.C.S. 1931 Ship wave resistance. An explanation ship hull shapes with developable skin: Leningrad,
comparison of the speeds of maximum and minimum Sudostroenie 192 p.)
resistance in practice and in theory. Trans. North East Дмитриев, А.А., Бончковская, Т.В. 1953 К вопросу о
Coast Inst. of Eng. and Shipbuilders, 47, Part 4, 153-180. турбулентности в волне// Докл. АН СССР, XCI No 1,
Wigley, W.C.S. 1936 The theory of the bulbous bow and its 31 – 33 (Dmitriev, QA.A., Bonchkovskaja, T.V. (1953) On
practical application. Trans. North East Coast Inst. of Eng. the turbulence in a wave// Docl. Acad. Sci. SSSR, XCI No
and Shipbuilders. 52, 65-88. 1, 31 – 33)
Wigley, W.C.S. 1937-8 Effects of viscosity on the wave- Доброклонский, С.В. 1947 Турбулентная вязкость в
making of ship. Trans. of the Inst. of Engin. Ship Builders поверхностном слое моря и волнение// // Докл. АН
in Scotland, 81, 187-212. СССР, 58.(геофизика) No 7, 1345 – 1348 (Dobroklonskii,
Wigley, W.C.S. 1942 Calculated and measured wave S.V. 1947Turbulent viscosity in the surface layer of sea//
resistance of series of forms defined algebraically, the Docl. Acad. Sci. SSSR, 58 No 7, 1345 – 1348)
prismatic coefficient and angle of entrance being varied Костюков, А.А. 1959 Теория корабельных волн и волнового
independently. Trans. of the Royal Inst. of Naval Arch., 84, сопротивления: Судпромгиз. –Ленинград, (Kostjukov,
52-74.

Oceanic Engineering International 101

 A.A. 1959 Theory of ship waves and wave resistance: вызванного ее начальным возмущением (Potetjunko,
Sudpromgiz, Leningrad) 310 p. E.N., Philimonova, L.D. On character of an attenuation of
Костюков, А.А. 1972 Взаимодействие тел, движущихся в the elevation of free surface produced by initial
жидкости Судостроение.- Ленинград. (Kostyukov, A.A. perturbation.) ПММ Applied mathematics and mechanics//,
Interaction of bodies moving in a fluid: Sudostroenie, 40 2, 362 – 366.
Leningrad, 312 p. Сизов, В.Г. 1961 К теории волнового сопротивления
Кочин, Н.Е. 1949 О волновом сопротивлении и подъемной судна на тихой воде// Изв. АН СССР, Механика и
силе, погруженных в жидкость тел. (Kochin, N.E. 1949 машиностроение, 1 75 – 85. (Sizov, V.G. 1961 On
On the wave resistance and lift of bodies submerged in a theory of the wave resistance of a ship on the calm water//
fluid: Collected works 2 105 – 143). Izv. Acad. Sci. Dep. Techn. Sci., Mechanics and
Кочин, Н.Е., Кибель, И.А., Розе, Н.В. 1963 Теоретическая Engineering, 1 75 – 85.).
гидромеханика, часть II, ГИФМЛ,. Сретенский, Л.Н. 1937 Теоретическое исследо- вание
Лаврентьев, В.М. 1951 Влияние пограничного слоя на волнового сопротивления. Глава !. К теории
волновое сопротивление корабля (Lavrentjev, V.M. 1951 Мичелла//.-(Sretenskii L.N. Theoretical research about
Effects of boundary layer on wave resistance of wave resistance. Part I. To Michell’s theory.// Trans.
ship)//Докл. АН ССС (Doclady AN USSR), 80 № 6, 857 – CAGI) Труды ЦАГИ.- M., 319 3 – 14.
860. Сретенский, Л.Н 1941 О волнах на поверхности вязкой
Ламб, Г. 1947 Гидродинамик,.М.-Л. 928 c. жидкости.( Sretenskii L.N. On waves on a surface of a
Милн-Томсон, Л.М. (1964) Теоретическая гидромеханика viscous fluid//Trans. CAGI) //Труды ЦАГИ, 541
(Milne-Thomson, L.M. Theoretical hydromechanics) Сретенский, Л.Н. 1977 Теория волновых движений жидкости.-
Мир, 656 с. (Sretenskii L.N. The theory of wave movements of a liquid:) Изд.
Никитин, А.К., Подрезов, С.А. 1964 К пространственной 2-ое, переработанное.- М., 816 с.
задаче о волнах на поверхности вязкой жидкости Федяевский, К.К., Войткунский, Я.И., Фаддеев, Ю.И. (1968)
Гидромеханика.- Л. Судостроение.
бесконечной глубины(Nikitin, A.K., Podrezov, C.A. 1964 Хинце, Д.О. 1963 Турбулентность. Ее механизм и теория.- М.:
To a spatial problem of waves on a surface of a viscous ГИФМЛ 680 с.
liquid of infinite depth// ПММ (Applied mathematics and Шлихтинг, Г. 1956 Теория пограничного слоя: (Hermann
mechanics), 28 3, 452 – 463. Shlichting Grenzschicht theorie: (1951)). Иностр. Лит., М.
Павленко, Г.Е. 1937 Судно наименьшего сопротивления// Шулейкин, В.В. 1958 Физика моря- Изд. АН СССР,-М.,
Труды ВНИТОСС 2 Вып. 3, 28 – 62. (Pavlenko, G.E. (Shuleikin, V.V. Physics of the see: M. (1956))..
(1937) Ship of the least resistance// Proc. VNITOSS, 2 Шутилов, К.В. 1941 Опыт определения коэффициента
Issue 3 28 – 62). турбулентной теплопроводности морской воды в
Павленко, Г.Е. 1956 Сопротивление воды движению судов, природных условиях. (Shutilov, K.V. Experience of
М. (Pavlenko, G.E. Resistance of moving ships) determination of the turbulent heat-conductance of sea
Потетюнко, Э.Н., Филимонова, Л.Д. 1976 О характере water in the nature conditions// Izv. Acad. Sci.) Изв. АН
затухания возвышения свободной поверхности, СССР 4 –5 447 – 451.

102 Oceanic Engineering International

APPENDIX A: WAVE RESISTANCE COEFFICIENTS OF WIGLEY AND WEINBLUM MODELS

Wigley 829
1,6

1,4
1 2
4 6 8
1,2
10

1
Cw

0,8
M easur. coef. Rw

0,6 M ichell

M ain part Cw
0,4
With viscous effect
0,2 NT=.08

0
0,16

0,19

0,22

0,25

0,28

0,31

0,34

0,37

0,40

0,43

0,46

0,49

0,52
Fn

Figure 1A. Wave resistance coefficients of Wigley model 829.

Wigley 1805a
1,6

1,4
1 2
4 6 8 10
1,2

1
Cw

0,8

M easur. coef. Rw
0,6
M ichell
0,4
M ain part

0,2 With viscous effect

NT=.08
0
0,12

0,15

0,18

0,21

0,24

0,27

0,30

0,33

0,36

0,39

0,42

0,45

0,48

0,51

0,54

Fn

Figure 2A. Wave resistance coefficients of Wigley model 1805a.

Oceanic Engineering International 103

 Wigley 1805b (Shearer model 3201)
1,6

1,4
1
2
4
6
1,2 8
10

1
Cw

0,8

M easur. coef. Rw
0,6
M ichell
0,4
M ain part

0,2 With viscous effect

NT=.08
0
0,20

0,22

0,24

0,26

0,28

0,30

0,32

0,34

0,36

0,38

0,40

0,42

0,44
Fn

Figure 3A. Wave resistance coefficients of Wigley model 1805b.

Wigley 1846a
1,4

1,2
1 2 4
68
1 10

0,8
Cw

0,6 M easur. Coef. Rw

M ichell
0,4
M ain part

0,2 With viscous effect

NT=.08

0
0,20

0,21

0,22

0,23

0,24

0,25

0,26

0,27

0,28

0,29

0,30

0,31

0,32

0,33

Fn

Figure 4A. Wave resistance coefficients of Wigley model 1846a.

104 Oceanic Engineering International

 Wigley 1846B
1,6

1,4
1
2
4 68
1,2
10

1
Cw

0,8

M easur. coef. Rw
0,6
M ichell
0,4 M ain part Cw

0,2 With viscous effect

NT=.08

0
0,16

0,19

0,22

0,25

0,28

0,31

0,34

0,37

0,40

0,43

0,46

0,49

0,52

0,55
Fn

Figure 5A. Wave resistance coefficients of Wigley model 1846b.

Wigley 1970b
1

0,9
1 2
4
6
0,8 8
10
0,7

0,6
Cw

0,5
M easur. coef. Rw
0,4
M ichell
0,3
M ain part Cw
0,2
With viscous effect
0,1 NT=.08
0
0,16
0,17

0,18
0,19
0,2

0,21
0,22

0,23
0,24
0,25

0,26
0,27
0,28

0,29
0,3

0,31
0,32

0,33

Fn

Figure 6A. Wave resistance coefficients of Wigley model 1970b.

Oceanic Engineering International 105

 Wigley model 1970c
1,2

1 2 4 68
1
10

0,8
Cw

0,6
M easur. Coef. Rw

0,4 M ichell

M ain part
0,2
With viscous effect
NT=.08

0
0,16

0,18

0,20

0,22

0,24

0,26

0,28

0,30

0,32

0,34

0,36

0,38

0,40
Fn

Figure 7A. Wave resistance coefficients of Wigley model 1970c.

Wigley N43
0,6

0,5 1 2
4
6
8
0,4
10
Cw

0,3
M easur. Cw

0,2 M ichell

M ain part
0,1 With viscous effect
NT=.08

0
0,12

0,14

0,16

0,18

0,2

0,22

0,24

0,26

0,28

0,3

0,32

0,34

0,36

Fn

Figure 8A. Wave resistance coefficients of Wigley model N43.

106 Oceanic Engineering International

 Wigley model 2038
0,3

0,25 1 2
4
6
8
10

0,2
Cw

0,15
Measur. coef. Rw

0,1 Michell

Main part Cw
0,05 With viscous effect
NT =.08
0
0,16

0,18

0,20

0,22

0,24

0,26

0,28

0,30

0,32
Fn

Figure 9A. Wave resistance coefficients of Wigley model 2038С.

Parabolic model

0,006
1
2
0,005 4
68
10

0,004
Cw

0,003

M easur. Coef. Cw
0,002
M ichell

M ain part
0,001
With viscous effect
NT=.08
0
0,15

0,18

0,21

0,24

0,27

0,30

0,33

0,36

0,39

0,42

0,45

0,48

Fn

Figure 10A. Wave resistance coefficients of the parabolic Wigley model.

Oceanic Engineering International 107

 Wigley model 2130a (unsymmetric)
1,2

1
1
2
10 12 14 16 18 4
6 8
0,8 10

1
Cw

0,6

0,4 Measur. coef. Rw

Michell
0,2 Main part Cw

With viscous effect

0 NT =.08
0,16

0,18

0,19

0,21

0,22

0,24

0,25

0,27

0,28

0,3

0,31

0,33
Fn

Figure 11A. Wave resistance coefficients of Wigley model 2130a (unsymmetrical).

Weinblum 1093
0,0012

0,001
1 2
4
6 8
10
0,0008
Cw

0,0006

M easur. coef. Rw
0,0004
M ichell

0,0002 M ain part

With viscous
0 effect NT=.08
0,1505

0,1618
0,1731

0,1844
0,1957

0,2070
0,2183
0,2296

0,2409
0,2521

0,2634
0,2747

0,2860
0,2973

0,3086
0,3199
0,3312

Fn

Figure 12A. Wave resistance coefficients of Weinblum model 1093.

108 Oceanic Engineering International

 Weinblum 1097
0,003

0,0025
1 2
4
6 8
10
0,002
Cw

0,0015
Measur. Coef. Rw

Michell
0,001
Main. Part

0,0005 With viscous effect

NT =.08

0
0,1806

0,1919

0,2032

0,2145

0,2258

0,2371

0,2484

0,2597

0,2710

0,2823

0,2935

0,3048

0,3161

0,3274
Fn

Figure. 13A. Wave resistance coefficients of Weinblum model 1097.

Weinblum 1098
0,0035

0,003
1 2 4 6 8
0,0025 10

0,002
Cw

0,0015 Measur. Coef. Rw

Michell
0,001
Main part

0,0005 With viscous effect

NT =.08

0
0,1505

0,1656

0,1806

0,1957

0,2107

0,2258

0,2409

0,2559

0,2710

0,2860

0,3011

0,3161

0,3312

Fn

Figure 14A. Wave resistance coefficients of Weinblum model 1098.

Oceanic Engineering International 109

 Weinblum 1100

0,0035

0,003
1
2
4
6
0,0025 8
10

0,002
Cw

0,0015 Measur. coef. Rw

Michell
0,001
Main part
0,0005
With viscous effect
NT =.08
0
0,1505
0,1618
0,1731
0,1844
0,1957
0,2070
0,2183
0,2296
0,2409
0,2521
0,2634
0,2747
0,2860
0,2973
0,3086
0,3199
0,3312
Fn

Figure 15A. Wave resistance coefficients of Weinblum model 1100.

Weinblum 1110

0,003

0,0025 1
2
4
6
8
10

0,002
Cw

0,0015
Measur. coef. Rw

0,001 Michell

Main part
0,0005
With viscous efffect
NT =.08
0
0,1505
0,1618
0,1731
0,1844
0,1957
0,2070
0,2183
0,2296
0,2409
0,2521
0,2634

0,2747
0,2860
0,2973
0,3086
0,3199
0,3312

Fn

Figure 16A. Wave resistance coefficients of Weinblum model 1110.

110 Oceanic Engineering International

 Weinblum 1111

0,004

0,0035 1
2
4
6
8
0,003 10

0,0025
Cw

0,002
Measur. coef. Rw
0,0015
Michell
0,001 Main part Cw

0,0005 With viscous effect

NT =.04
0
0,1505

0,1656

0,1806

0,1957

0,2107

0,2258

0,2409

0,2559

0,2710

0,2860

0,3011

0,3161

0,3312
Fn

Figure 17A. Wave resistance coefficients of Weinblum model 1111.

Weinblum model 1112

0,0025
1 2
4
6
8

0,002 10

0,0015
Cw

0,001

Measur. coef. Rw
0,0005
Michell
0 Main part
0,1505

0,1656

0,1806

0,1957

0,2107

0,2258

0,2409

0,2559

0,2710

0,2860

0,3011

0,3161

0,3312

Fn

Figure 18A. Wave resistance coefficients of Weinblum model 1112.

Oceanic Engineering International 111

 Weinblum 1113
0,006

0,005
1 2
4
6

0,004 8
10
Cw

0,003

Measur. coef. Rw
0,002
Michell

Main part
0,001
With viscous effect
NT =.08
0
0,1505
0,1618

0,1731
0,1844
0,1957

0,2070
0,2183
0,2296
0,2409
0,2521
0,2634
0,2747

0,2860
0,2973
0,3086

0,3199
0,3312
Fn

Figure 19A. Wave resistance coefficients of Weinblum model 1113.

Weinblum 1114
0,003

0,0025
1 2
4
6
8
10

0,002
Cw

0,0015
M easur. coef. Rw

0,001 M ichell

M ain part
0,0005
With viscous effect
NT=.08
0
0,1505
0,1618

0,1731
0,1844

0,1957
0,2070

0,2183
0,2296

0,2409
0,2521
0,2634

0,2747
0,2860

0,2973
0,3086

0,3199
0,3312

Fn

Figure 20A. Wave resistance coefficients of Weinblum model 1114.

112 Oceanic Engineering International

 Weinblum 1136
0,0035

0,003
1 2 4 6
8
0,0025
10

0,002
Cw

0,0015 M easur. coef. Rw

0,001 M ichell

M ain part
0,0005
With viscous effect
NT=.08
0
0,1505

0,1656

0,1806

0,1957

0,2107

0,2258

0,2409

0,2559

0,2710

0,2860

0,3011

0,3161

0,3312
Fn

Figure 21A. Wave resistance coefficients of Weinblum model 1136.

Explanation of figures 1A-21A

In these figures are shown

1) Measured coefficient of wave resistance ( Measur. Coef. Rw );
2) Michell’s curve (Michell);
3) Main part of the Michell integral (Main part);
4) Value of the Michell integral taking into account the influence of turbulent viscosity on the interaction of bow and stern wave
systems (with viscous effect. NT = .08 or some other value). Here NT is the half the coefficient of turbulent viscosity.

Oceanic Engineering International 113

APPENDIX B: THE SHIP HULL SHAPES WITH LEAST WAVE RESISTANCE ON Fn = 0.27.

1 1
2 2
4 4
6 6
8 8

10 10

Figure 1B. Optimum hull form by equation (1). Figure 2B. Optimum hull form by equation (2).
Rgd = 2.4366. Rgd = 2.5991.

1 1
2 2
4 4
6 8
6
8
10

10

Figure 3B. Optimum hull form by equation (3). Figure 4B. Hull form with bulb by equation (4).
Rgd = 2.2019. Rgd = 1.891.

1
1 2
2 4
4 6
8 6
8

10
10

Figure 5B. Optimum hull form by equation (5). Figure 6B. Optimum hull form by equation (6).
Rgd = 2.2559. Rgd = 1.7231.

1 2 1 2
4 4
6 6
8 8

10 10

Figure 7B. Optimum hull form by equation (7). Figure 8B. Optimum hull form by equation (8).
Rgd = 3.5675. Rgd = 2.4586.

114 Oceanic Engineering International

APPENDIX C: THE HULL SHAPES WITH THE LEAST WAVE RESISTANCE AT THE DIFFERENT FROUDE NUMBERS

1
2 1 2
4 4 6 8
6 8

10 10

Figure 1C. The ship hull with the least wave resistance Figure 2C. The ship hull with the least wave resistance
at Fn = 0.15. at Fn = 0.18.

1 2 4 6 8 1
2
4 6 8
10
10

Figure 3C. The ship hull with the least wave resistance Figure 4C. The ship hull with the least wave resistance
at Fn = 0.23. at Fn = 0.32.

Oceanic Engineering International 115