Short notes1 from

M.H. Patel, Dynamics of Offshore Structures,

Butterworths (1989)

1 Basic fluid mechanics

The present notes do not want to substitute the study of the cited book, but to
give just a first glance of basic fluid mechanics which is needed to understand
the mathematics of linear wave theory and nonlinear wave theories. It is recalled
that in order to have a more general view on the basic fluid mechanics problem
the reader should consider to refer to chapter 3 of the book cited above.

1.1 Frame of reference

Since a fluid is treated as a continuous medium, a difficulty arises in identifying
a means or a frame of reference which can be used to describe the motion of all
the fluid within a flow field. Two frames of reference might be used: Lagrangian
and Eulerian. The first describes the behavior of individual fluid particles during
motion within a flow field as a function of their initial position and time. Thus,
the Cartesian coordinates (x, y, z) describing the position of a fluid particle can
be written as
x = F1 (a, b, c, t)
y = F2 (a, b, c, t) (1)
z = F3 (a, b, c, t)

where (a, b, c) are the coordinates of the particle’s initial position and t is
time. By differentiating these relations with respect to time velocity and accel-
eration are carried out.
On the contrary Eulerian frame of reference defines a flow field by identi-
fying fluid velocity components at discrete points within a volume of fluid and
observing the characteristic variation of the fluid as it flows past that point.
Thus, the velocity components are defined as

u = f1 (x, y, z, t)
v = f2 (x, y, z, t) (2)
w = f3 (x, y, z, t)

where (x, y, z) are the coordinates of a point fixed in space and t is time. The
Eulerian frame, compared to the Lagrangian one, identifies fixed points in space
and describes the velocity components of fluid particles as they pass the points.
The Eulerian scheme is the best for computing forces on offshore structures.
1 by Nicholas Fantuzzi, nicholas.fantuzzi@unibo.it (March 5, 2017)

1
1.2 Rotational and irrotational flow
Since a fluid readily shears under the action of an applied shear stress it is im-
portant to examine the precise nature of this shear action and to distinguish it
from other fluid motion such as rotation. The following figure depicts an irro-
tational and a rotational fluid, because the first exhibits distortion of the fluid
element without rotation, whereas figure on the right demonstrate a solid body
rotation without distortion. The irrotational flow condition is demonstrated in

Figure 1: Irrotational and rotational fluid.

detail in [Patel p.78], the demonstration is not reported because out of scope of
the present notes. In particular the following definitions of the vorticity along
the three Cartesian directions ξ, η, ζ are computed
∂w ∂v
ξ= −
∂y ∂z
∂u ∂w
η= − (3)
∂z ∂x
∂v ∂u
ζ= −
∂x ∂y
The irrotational flow condition states ξ = η = ζ = 0 so that
∂w ∂v
− =0
∂y ∂z
∂u ∂w
− =0 (4)
∂z ∂x
∂v ∂u
− =0
∂x ∂y
This conditions is very useful for gravity waves where the flow is taken as
irrotational because the fluid particle displacements are small if compared to
the length scales of the bulk of fluid. By considering a 2D problem the three
relations above reduce only to one
∂u ∂w
− =0 (5)
∂z ∂x
Please note that this is the same irrotational condition indicated in [Wilson
p.66 eq.(3.11a)].

2
1.3 Stream function and velocity potential
It is cumbersome to describe the motion of a fluid as a set of displacement
co-ordinates or velocity components within a flow field of large volume. A flow
field can be conveniently described in terms of mathematical functions in two- or
three-dimensions. Generally, the stream function ψ and the velocity potential
φ are used. The stream function ψ can describe a steady or un-steady flow
pattern in a two-dimensional plane. Consider the 2D field in figure

Figure 2: Graphical representation of stream function and velocity potential.

where stream lines of constant value ψ are shown. The fluid flows between
the stream lines, without crossing them. The stream function ψ represents the
volume rate or flux between the stream lines so if two points A and B are
considered on ψA and ψB

δψ = ψB − ψA = uδy − vδx (6)

But the same increment can be written as
∂ψ ∂ψ
δψ = δx + δy (7)
∂x ∂y
By comparison of the two equations above the definition of the stream func-
tion can be given
∂ψ ∂ψ
u= , v=− (8)
∂y ∂x
An alternate mathematical function, the velocity potential φ, can be also
used to define a flow field, so that
∂φ ∂φ ∂φ
u= , v= , w= (9)
∂x ∂y ∂z
Generally speaking lines of φ are called equipotentials and they are orthog-
onal to the constant ψ. Please note that φ is applied to irrotational flow only,

3
whereas the stream function ψ can be applied to both irrotational and rotational
flows.

1.4 Continuity and the Laplace equation

The principle of conservation of mass states that the rate of mass accumulation
within a control volume of fluid must equal the net mass flux into the volume
boundaries. This principle is used to derive the equation of continuity. The
detailed demonstration of such equation can be found in [Patel p.81] where the
equation of continuity for a compressible flow can be found
∂ρ ∂(ρu) ∂(ρv) ∂(ρw)
+ + + =0 (10)
∂t ∂x ∂y ∂z
If the flow is incompressible (ρ = 0) the continuity equation becomes
∂u ∂v ∂w
+ + =0 (11)
∂x ∂y ∂z
The Laplace equation can be found by substituting the velocity potential in
the equation above

∂2φ ∂2φ ∂2φ

+ 2 + 2 = 0 or ∇2 φ = 0 (12)
∂x2 ∂y ∂z
For a 2D problem the continuity equation becomes
∂u ∂w
+ =0 (13)
∂x ∂z
that can be found in [Wilson p.66 eq.(3.11b)].

1.5 The Navier-Stokes equations

The dynamic behavior of fluids is governed by the Navier-Stokes equations which
extend Newton’s second law of motion to a continuous fluid medium. An in-
finitesimal element is subjected to forces due to a pressure field and body forces
(e.g. gravity) together with direct and shear stresses induced by the motion of
the fluid. These forces are balanced by the product of the fluid density and the
direct as well as convective accelerations of the fluid. All of these physical effects
are incorporated in the Navier-Stokes equations which are described below for
a compressible flow. The demonstration can be found in [Patel p.82], the result
is reported below.
 2
∂ u ∂2u ∂2u

∂u ∂u ∂u ∂u X 1 ∂p
+u +v +w = − +ν + +
∂t ∂x ∂y ∂z ρ ρ ∂x ∂x2 ∂y 2 ∂z 2
 2
∂2v ∂ 2v

∂v ∂v ∂v ∂v Y 1 ∂p ∂ v
+u +v +w = − +ν + 2+ 2 (14)
∂t ∂x ∂y ∂z ρ ρ ∂y ∂x2 ∂y ∂z
 2
∂ w ∂ w ∂2w
2

∂w ∂w ∂w ∂w Z 1 ∂p
+u +v +w = − +ν + +
∂t ∂x ∂y ∂z ρ ρ ∂z ∂x2 ∂y 2 ∂z 2

4
 where X, Y , Z are the body forces and ν is the fluid viscosity. If the flow is
inviscid ν = 0 the Navier-Stokes equations reduce to the Euler equations which
take the form

∂u ∂u ∂u ∂u X 1 ∂p
+u +v +w = −
∂t ∂x ∂y ∂z ρ ρ ∂x
∂v ∂v ∂v ∂v Y 1 ∂p
+u +v +w = − (15)
∂t ∂x ∂y ∂z ρ ρ ∂y
∂w ∂w ∂w ∂w Z 1 ∂p
+u +v +w = −
∂t ∂x ∂y ∂z ρ ρ ∂z
The solution of the latter two equations is not generally possible due to
realistic flow fields because of the nonlinear convective acceleration terms on the
left-hand side of the equations. Numerical models must be employed for solving
such mathematical problems. The Navier-Stokes equations can be presented in
short matrix form as
Du
ρ = f − ∇p + ν∇2 u (16)
Dt
where u = [u, v, w]T , f = [X, Y, Z]T and D expresses the total derivative
operator
D ∂
= +u·∇ (17)
Dt ∂t
and
∂ ∂ ∂
∇2 = 2
+ 2+ 2 (18)
∂x ∂y ∂z
If the flow is inviscid (kinematic viscosity ν = 0) the result is known also as
Cauchy momentum equation (without the deviatoric stress tensor [see Wikipedia
”Derivation of Navier-Stokes equations”])
Du
ρ = f − ∇p (19)
Dt
D ∂
If a 2D problem is considered by neglecting nonlinear terms so that Dt = ∂t
T
and gravity forces are considered only f = [0, 0, ρg] the following equations can
be derived

∂u 1 ∂p
=−
∂t ρ ∂x
(20)
∂w 1 ∂p
= −g −
∂t ρ ∂z
These two equations are the momentum consideration equations in [Wilson
p.66 eqs.(3.11c), (3.11d)].

5
2 Gravity wave theories
Gravity waves are a large family of surface water waves ranging from wind gen-
erated waves in the oceans, to tidal action, water surface oscillation in harbors,
flood waves in rivers, tidal bores, hydraulic jumps, waves generated by seismic
disturbances, and so on. In the following wind generated waves will be analyzed.
Gravity waves exist at the interface of two fluids of different density such as air
and water. Winds blow over a water surface and generate waves which develop
with increasing wind duration and fetch. The fetch is the length of water surface
over which the wind blows. Gravity waves generally travel and can move out
of their area of formation, and when they run free of the winds that generated
them such waves are called a swell. The simplest wave theory of the whole
family of gravity waves is the linear wave theory which is a linearized version of
a more complex mathematical description. Due to the linearity, such theory is
able to describe irregular oceans as a superposition of waves of regular motions
with different periods and from different directions. The most difficult part in
the present modeling is the treatment of the free surface boundary condition.
This difficulty arises from the nonlinear convective terms in the Navier-Stokes
equations which occur in the form of velocities times the spatial gradients.
It should be recalled that wave data can be presented in two ways: design
wave approach and design to a statistical wave description. The design wave
approach is based on the selection of a certain wave height in a return period
of 50 or 100 years. This approach is realistic from the designing point of view
against structural failure. However, large waves do not permit to analyze fatigue
damage in the same design. In fact, for many offshore structures the wave height
is not the worst loading case. Waves of smaller heights and periods with wave
lengths that are a specific fraction of the structure dimension or with periods
that excite resonant motions can induce larger structural loadings than the
highest design wave. The design wave approach is used for both fixed and
floating offshore structures.
By using the statistical description the incident waves is expressed in terms
of a probability density function which describes the probability of occurrence
of waves with a specified height, period and direction. A statical description of
such waves does permit to design structures under fatigue loads. However, this
approach has a big drawback which relies on the linear superposition of wave
components. Therefore, this approach is prone to some error for high waves.
The requirement for linear superposition is valid for computing loads due to
small waves on large structures where linear wave theory applies. For structures
with members of small diameter, nonlinear drag forces are significant and the
extension to the statistical wave description should be done with caution.

2.1 Linear wave theory

This theory offers a powerful analytical solution for gravity waves. The following
assumptions should be evoked:
1. The water is of uniform density and constant water depth.

6
 2. Viscosity and surface tension are neglected.
3. Nonlinear velocity terms are neglected.
4. Irrotational wave motion is considered.
A 2D linear gravity wave has to satisfy the continuity condition, so for an
irrotational flow it takes the form

∂2φ ∂2φ
+ 2 =0 (21)
∂x2 ∂y
please note that this equation corresponds to
∂u ∂w
+ =0 (22)
∂x ∂z
which is [Wilson eq.(3.11b)]. Due to the difference between Patel’s and
Wilson’s books in considering the vertical axis (y for Patel and z for Wilson)
in the following the Patel’s equations are presented with reference to Wilson’s
notation in the x-z reference. So the continuity equations becomes

∂2φ ∂2φ
+ 2 =0 (23)
∂x2 ∂z
The dynamic equation of motion (or Bernoulli’s equation, which implies an
irrotational and inviscid flow) becomes
 2 2  2 2 !
p ∂φ 1 ∂ φ ∂ φ
= −gz − − + (24)
ρ ∂t 2 ∂x2 ∂z 2
The boundary condition of zero flow normal to the sea bed at z = −d is
 
∂φ
w = 0 at z = −d ⇔ =0 (25)
∂z z=−d
The boundary condition at free surface is difficult due to the unknown water
surface shape defined by function η = η(x, z, t). Such condition can be modeled
by considering a flow boundary condition to which the local velocity must be
tangential with nη normal to the surface so
 
∂φ
=0 (26)
∂nη z=0
This condition is also known as kinematic free surface boundary condition.
Another condition will be defined at free surface as constant pressure on the
free surface shape η. The nonlinearity of the above system of equations arises
from the fact that the free surface boundary condition is applied on an unknown
surface which is obtained only after the solution of the governing equations and
boundary conditions.
Water pressure at the sea bed can be obtained from equation (24)

7
    2  2 !
ps ∂φ 1 ∂φ ∂φ pt
+ gη + + + = (27)
ρ ∂t s 2 ∂x ∂z ρ
where s denotes the reference surface and pt is the total pressure. By differ-
entiating equation (27) with respect to time and by neglecting velocity squared
terms, it leads
 2 
1 ∂ps Dη ∂ φ
+g + =0 (28)
ρ ∂t Dt ∂t2 s
where
D ∂ ∂ ∂
= +u +w (29)
Dt ∂t ∂x ∂z
is the total derivative2 . The latter for waves of small height (mathematically
infinitesimal) becomes
 
Dη ∂η ∂φ
= = (30)
Dt ∂t ∂z z=0
∂η
Due to the fact that at z = 0, w = ∂t [Wilson eq.(3.12)]. Now if surface tension
is ignored, it leads

ps = pa
∂ps ∂pa ⇔ p = pa , at z = 0 (31)
= =0
∂t ∂t
Equation (31) can be found in [Wilson eq.(3.14)]. By using equations (31)
and (30), equation (28) becomes
   2 
∂φ ∂ φ
g + =0 (32)
∂z z=0 ∂t2 z=0
This is the linearized free surface boundary condition. Please note that the
linearization process makes the equation linear (by neglecting nonlinear terms)
but also fixes the free surface boundary. Therefore, the present theory can be
used for small to moderate wave heights.
Finally the linear wave problem can be stated as in the following, the gov-
erning equation is

∂2φ ∂2φ
+ 2 =0 (33)
∂x2 ∂z
and the correspondent boundary conditions are at the sea bed z = −d
2 This derivative has several names, for more details please see,
https://en.wikipedia.org/wiki/Material derivative

8
 
∂φ 
=0 (34)
∂z z=−d
and at the surface z = 0

∂ 2 φ 
 
∂φ 
g  + 2 =0 (35)
∂z z=0 ∂t z=0
The presented partial differential equation is of elliptic type with the solution
in the domain limited by horizontal planes z ∈ [−d, 0] and with no limit in the
x direction. The solution φ is of harmonic type as

φ = F (z) exp (−j (kx − ωt)) = F (z)e−j(kx−ωt) (36)

√
where j = −1 is the imaginary unit. First we need to compute the derivatives
of φ

∂2φ
= −F (z)k 2 e−j(kx−ωt)
∂x2 (37)
∂2φ d2 F (z) −j(kx−ωt)
= e
∂z 2 dz 2
Now these expression can be put into the governing equation (33) as follows

d2 F (z)
− k 2 F (z) = 0 (38)
dz 2
This differential equation has a solution in the form

F (z) = C1 ekz + C2 e−kz (39)

If the sea bed z = −d boundary condition (34) is applied
 
∂φ  ∂F (z) 
= = kC1 e−kd − kC2 ekd = 0 (40)
∂z z=−d ∂z z=−d
Therefore
1 C kd C −kd
kC1 e−kd = kC2 ekd = C → C1 = e ; C2 = e (41)
2 2k 2k
So F (z) takes the form

  
C kd kz C −kd −kz C 1 k(d+z) −k(d+z)
F (z) = e e + e e = e +e (42)
2k 2k k 2

Please note that the term between the square brackets is cosh(k(d + z)). Once
F (z) has been rewritten in harmonic form the solution can be carried out
C
φ= cosh(k(d + z)) e−j(kx−ωt) (43)
k

9
 where the constant C still needs to be evaluated. This can be done by
considering oscillatory variations due to the wave in equation (27). Since the
pressure above the wave surface is uniform and the nonlinear terms are neglected
as follows

✘ 2✘
 2  ✘ !
∂φ ✘✘∂φ
 
ps ∂φ 1 pt
✁ + gη + + ✘ ✘+ = ✁ (44)
✁ρ ∂t s ✘ ✘
2 ✘ ∂x ∂z ✁ρ
The surface elevation η can be written as
 
1 ∂φ
η=− (45)
g ∂t z=0
The derivative of φ with respect to time is
∂φ C
= jω cosh(k(d + z)) e−j(kx−ωt) (46)
∂t k
Finally the wave elevation η is
ωC
η=− cosh(kd) j e−j(kx−ωt) = A j e−j(kx−ωt) (47)
kg
where A is the wave amplitude.
ωC
A=− cosh(kd) (48)
kg
The definition of A can be used for finding the velocity potential (43) as
C Ag 1
=− (49)
k ω cosh(kd)
and then by substiuting into equation (43)

Ag cosh(k(d + z)) −j(kx−ωt)

φ=− e (50)
ω cosh(kd)
Now, the second boundary condition (free surface) (35) can be used by sub-
stituting the velocity potential φ in it. First, we calculate the derivatives

∂φ Ag k
=− sinh(k(d + z)) e−j(kx−ωt)
∂z ω cosh(kd)
(51)
∂2φ Ag cosh(k(d + z)) 2 −j(kx−ωt)
2
= ω e
∂t ω cosh(kd)

Thus, equation (35) takes the form


Ag k sinh(kd) −j(kx−ωt)

✟✯1
✟
Ag cosh(kd)
g − e + ✟ ω 2 e−j(kx−ωt) = 0 (52)
ω cosh(kd) ω ✟ ✟
cosh(kd)

10
After some simplifications the following relation is carried out

gk tanh(kd) = ω 2 (53)
which corresponds to the equation in [Wilson eq.(3.16)]. By taking the real
parts3 of equations (47) and (50) the following relationships are obtained

Ag cosh(k(d + z))
φ=− cos(kx − ωt) (54)
ω cosh(kd)

η = A sin(kx − ωt) (55)

The last equation represents the real wave elevation which can be found also
in [Wilson Table 3.1] with the only difference that the wave type is represented
by using a cosine function instead of a sine function. By taking the progressive
wave relationship (e.g. wave celerity)
ω λ
= fλ =
c= (56)
k T
By using equation (53) the following result can be carried out

ω2 gk tanh(kd) g 1/2
c2 = 2
= 2
→ c= tanh(kd) (57)
k k k
which is the same in [Wilson eq.(3.17) Table 3.1 8th row]. Equations (56)
and (53) can be combined as

ω 2 cosh(kd) cosh(kd)
g= and ω = kc → g = kc2 (58)
k sinh(kd) sinh(kd)
Therefore, the velocity potential (54) becomes

cosh(k(d + z))
φ = −Ac cos(kx − ωt) (59)
sinh(kd)

2.2 Wave period, wave length & celerity

In case of deep water waves (d → ∞), the hyperbolic terms of the equations of
the last section reduce to exponentials, so that
Ag kz
φ=− e cos(kx − ωt) or φ = −Ac ekz cos(kx − ωt) (60)
ω
Moreover, the wave frequency for deep water waves becomes

ω 2 = gk (61)
3 Remember that ejθ = cos θ + j sin θ

11
 At this point some simple relationships between wave period T , wave length
λ and celerity c can be defined for deep water waves.
r
gλ
c=
2π
s
2πλ (62)
T =
g
gT
c=
2π
For T = 10 s, λ = 156.1 m and c = 15.61 m/s, whereas T = 6 s, λ = 56.2
m and c = 9.37 m/s. Thus, long waves travel faster than short ones. So long
waves overtake short ones by generating a continually varying water surface.
The wave particle motions (ξ, η) can be carried out by integration as
∂φ
Z
ξ= dt = Aekz cos(kx − ωt)
∂x
(63)
∂φ
Z
η= dt = Aekz sin(kx − ωt)
∂z
The latter relations show that water particles in deep water waves move in
a circular path of radius A with periodicity T . The velocity components can be
carried out as
∂φ
u= = Aωekz sin(kx − ωt)
∂x (64)
∂φ
w= = −Aωekz cos(kx − ωt)
∂z
and the total velocity can be computed as
p
q = u2 + w2 = Aωekz (65)

We can be conclude that both particle displacements and velocities decrease

exponentially with depth. At one wave length below the surface (d = −λ)
the reducing ratio is e−2π = 0.00187. At half-wave length (d = −λ/2) the
reducing ratio is e−π = 0.0432. Therefore for engineering purposes at half-wave
length below the surface the effect of gravity waves is negligible. Note that wave
particles move in the direction of the wave propagation when they are in the
crest and they have opposite motion in the trough.
The pressure in waves p = p′ − pa , with pa atmospheric pressure, can be
computed by Bernoulli’s equation (24) (neglecting nonlinear terms) as

p′ + ρgz − ρgAekz sin(kx − ωt) = pa (66)

The pressure distribution below a linear gravity wave can be seen as a series
of constant surfaces with the pressure value for each surface increasing in the
depth below the surface.

12
 Figure 3: Wave particle orbits in deep and shallow waters.

Another important relation is the total energy of a gravity wave. The total
potential and kinetic energy of a gravity wave is derived below for the general
case of shallow water linear equations (the deep water case can be always de-
ducted by setting d → ∞). It is considered a linear wave of amplitude A (wave
height H = 2A), frequency ω and wave number k = 2π/λ in water of depth d.
The amount of energy bounded between two planes in a unit thickness of water
includes kinetic and potential energy. The kinetic energy comes from the wave
particle velocities, whereas potential energy derives from free surface change in
elevation within the Earth’s gravitational field. In terms of velocity potential
the total energy can be written as
ZZZ (  2  2  2 ! )
1 ∂φ ∂φ ∂φ
E=ρ + + + gz dx dy dz (67)
2 ∂x ∂y ∂z
It is recalled that the derivatives of the velocity potential (59) are the fluid
velocities in the wave field. By considering a 2D fluid flow and by integrating
along the water depth and for a whole wave length the total energy becomes

( 2 2 ) ( )
0 λ  Z ηZ λ
ρ ∂φ ∂φ
Z Z
E= + dx dz + ρg z dx dz (68)
2 −d 0 ∂x ∂z 0 0

Note that the potential energy (the term on the right in the equation above
is computed with reference to the still water level (SWL, z = 0). The derivatives
of the velocity potential (59) are

∂φ cosh(k(d + z))
= Ack sin(kx − ωt)
∂x sinh(kd)
(69)
∂φ sinh(k(d + z))
= −Ack cos(kx − ωt)
∂z sinh(kd)

13
Thus, the total energy takes the form

(
0 λ
ρA2 c2 k 2
Z Z
E= cosh2 (k(d + z)) sin2 (kx − ωt)+
2 sinh2 (kd) −d 0
) (70)
λ
1
Z
2 2 2
sinh (k(d + z)) cos (kx − ωt) dx dz + ρg η dx
2 0

By applying the double-angle formula for the hyperbolic functions4

(
0 λ
ρA2 c2 k 2 cosh(2k(d + z)) + 1
Z Z
E= sin2 (kx − ωt)+
2 sinh2 (kd) −d 0 2
) Z λ (71)
cosh(2k(d + z)) − 1 2 1
cos (kx − ωt) dx dz + ρg η 2 dx
2 2 0

The terms which depend on z can be collected and the other terms can be
collected too through the double-angle formula for trigonometric functions5 as

( )
0 λ
ρA2 c2 k 2 cosh(2k(d + z)) cos(2(kx − ωt))
Z Z
E= − dx dz+
2 sinh2 (kd) −d 0 2 2
Z λ (72)
1
ρg η 2 dx
2 0

By developing the integral along z and including the definition of wave ele-
vation η (55) it leads

( )
λ
ρA2 c2 k 2 sinh(2kd) d
Z
E= − cos(2(kx − ωt)) dx+
2 sinh2 (kd) 0 4k 2
(73)
A2 ρg λ 2
Z
sin (kx − ωt)dx
2 0

The final integration along x leads to

ρA2 c2 k 2 λ sinh(2kd) A2 ρgπ

E= + (74)
8k sinh2 (kd) 2k
Another form of the energy can be (using k = 2π/λ, c = ω/k and ω 2 =
gk tanh(kd))
4 Double-angle formula for hyperbolic functions: cosh(2θ) = −1 + 2 cosh2 θ = 1 + 2 sinh2 θ.

So that cosh2 θ = cosh(2θ)+1

2
and sinh2 θ = cosh(2θ)−1
2
5 Double-angle formula for trigonometric functions: cos(2θ) = cos2 θ − sin2 θ

14
 A2 ρgλ tanh(kd) sinh(2kd) A2 ρgλ
E= + (75)
8 sinh2 (kd) 4
By collecting all the terms6 in the expression above the following relation
can be applied
1
ρgA2 λ
E= (76)
2
Is is remarked that the total energy is independent on the water depth d.
Therefore, the energy per unit area of water surface is
E 1
= ρgA2
Ê = (77)
λ 2
Compare equation (77) with the one in [Wilson eq.(6.14) p.150].
The total amount of energy of a single wave can be computed by considering,
for example, amplitude A = 5 m and period T = 8 s. Thus, if deep water waves
are considered (see equations (62)), the wave length λ = gT 2 /(2π) ≃ 100 m
and the celerity c = gT /(2π) ≃ 12.5 m/s. By considering a sea water density
ρ = 1025 kg/m3 the total energy results to be E ≃ 125.7 kJ/m2 or 34.9 Wh/m2 .
One square kilometer (1 km2 = 106 m2 ) the total energy content is E = 125.7
GJ or 34.9 MWh. Since the energy content is travelling with c = 12.5 m/s,
the energy flux per unit length of wave front can be computed as E = 0.1257
MJ/m2 · 12.5 m/s = 1.57 MW/m.

6 Please note that sinh(2kd) = 2 sinh(kd) cosh(kd).

15