Journal of Fluids and Structures 40 (2013) 201–213

Contents lists available at SciVerse ScienceDirect

Journal of Fluids and Structures

journal homepage: www.elsevier.com/locate/jfs

The wave excitation forces on a truncated vertical cylinder

in water of infinite depth
William Finnegan a,c, Martin Meere b, Jamie Goggins a,c,n
a
College of Engineering and Informatics, National University of Ireland, Galway, Ireland
b
School of Mathematics, Statistics and Applied Mathematics, National University of Ireland, Galway, Ireland
c
Ryan Institute for Environmental, Marine and Energy Research, National University of Ireland, Galway, Ireland

a r t i c l e i n f o abstract

Article history: When carrying out any numerical modelling it is useful to have an analytical approxima-
Received 22 July 2012 tion available to provide a check on the accuracy of the numerical results and to give
Accepted 22 April 2013 insight into the underlying physics of the system. The numerical modelling of wave
Available online 31 May 2013
energy converters is an efficient and inexpensive method of undertaking initial optimisa-
Keywords: tion and experimentation. Therefore, the main objective of this paper is to determine an
Analytical approximation analytical approximation for the wave excitation forces on a floating truncated vertical
Excitation force cylinder in water of infinite depth. The approximation is developed by solving appropriate
Infinite depth boundary value problems using the method of separation of variables. A graphical
Scattering problem
representation of the analytical approximation for the truncated vertical cylinder and
Floating truncated cylinder
the cylinder of infinite depth are presented and are compared to the results from a
Wave energy converters
computational fluid dynamics analysis, using a commercial boundary element package.
The presented analytical approximation and the computational fluid dynamics analysis
results were found to be in good agreement. Furthermore, the presented analytical
approximation was found to be in good agreement with independent experimental data.
& 2013 Elsevier Ltd. All rights reserved.

1. Introduction

One of the main stages in the design of wave energy converters (WECs) is the numerical modelling of a given converter.
In this paper, an analytical solution for the wave excitation forces on a floating truncated vertical cylinder in water of infinite
depth is provided. This solution provides a method for validating the results of numerical models of WECs, since it estimates
the forces on a cylinder representation of an arbitrary shaped axisymmetric WEC.
The solution of the scattering and radiation problem for floating bodies, in finitely or infinitely deep water, has being
explored for decades for various shapes of bodies. Ursell (1949) explored the forces on an infinitely long horizontal floating
cylinder in infinitely deep water using a polynomial set of stream functions to derive the analytical solution. Havelock (1955)
employed a similar technique to solve the radiation problem for a floating half-immersed sphere in infinitely deep water.
MacCamy and Fuchs (1954) derived the analytical solution of a bottom mounted cylinder which penetrates the water
surface in water of finite depth. Garrett (1971) formulated the solution for the scattering problem of oblique waves around a
circular dock in water of finite depth and presented numerical results detailing the vertical force, horizontal force and

n
Corresponding author at: College of Engineering and Informatics, National University of Ireland, Galway, Ireland. Tel.: +353 91492609;
fax: +353 91494507.
E-mail address: [email protected] (J. Goggins).

0889-9746/$ - see front matter & 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jfluidstructs.2013.04.007
202 W. Finnegan et al. / Journal of Fluids and Structures 40 (2013) 201–213

Nomenclature SB wetted surface

t time (s)
a radius of cylinder (m) T location of pitch
A amplitude of incident wave (m) v flow velocity (m s−1)
As cross-sectional area (m2) x horizontal coordinate (m)
b draft of cylinder (m) z vertical coordinate (m)
Cm inertia coefficient εm Neumann symbol
Cd drag coefficient θ polar coordinate rads
F force (N) ξ separation constant
Fc Fourier cosine transform ρ density (kg m3)
F1,ext surge excitation force (N) φ frequency domain velocity potential (m s−1)
F3,ext heave excitation force (N) φI incident wave velocity potential (m s−1)
G gravity (m s−2) φis interior scattering velocity potential (m s−1)
k0 wavenumber (m−1) φed exterior diffraction velocity potential (m s−1)
m integer φes exterior scattering velocity potential (m s−1)
nj j-component of the normal Φ time domain velocity potential (m s−1)
pm(ξ) coefficient χi interior condensed function
qm0 coefficient χe exterior condensed function
qm(ξ) coefficient ω wave angular frequency (s−1)
r radius (m)

torque. Leppington (1973) examined the radiation properties of partially immersed three-dimensional bodies. A short-wave
asymptotic limit was imposed in order to derive the velocity potential of the outgoing wave of a heaving and rolling circular
dock and a heaving hemisphere. Bai (1975) developed a numerical method of linearising the boundary value problem by
using a variational principle equivalent in order to determine the diffraction of oblique waves by a horizontal infinitely long
floating cylinder. This constructed variational form is employed by a finite element discretisation of the fluid domain and the
numerical results are presented. Black (1975) investigated the wave forces on bodies which are vertically axisymmetric
using an integral equation formulation in water of finite depth. Yeung (1981) presented a set of theoretical added mass and
damping coefficients for a floating cylinder in water of finite depth and, also, truncated the solution for the infinite depth
problem. Hsu and Wu (1997) developed a boundary element method to study the sway and heave motion of a 2-D floating
rectangular structure in water of finite depth and an analytical solution for the problem was also presented. Mansour et al.
(2002) developed an analytical and boundary integral method (BIM) solution for a bottom-mounted uniform vertical
cylinder with cosine-type radial perturbations which penetrates the water surface in water of finite depth and compared the
numerical results to that of a circular cylinder. Bhatta and Rahman (2003) used the method of separation of variables
technique, which is similar to that employed by Havelock (1955), to analyse the scattering and radiation problem for a
floating vertical cylinder in water of finite depth and presented the formulation for the surge, heave and pitch motion
solutions. Liu et al. (2012a, 2012b) developed an analytical solution using a matched eigenfunction expansion for the wave
scattering by a submerged porous plate with finite thickness in water of finite depth and a boundary element method
solution is also presented to confirm the analytical solution. Hassan and Bora (2012) employed the method of separation of
variables to derive the exciting forces on a pair of coaxial hollow cylinder and bottom mounted cylinder in water of infinite
depth and presented numerical results for a variety of radius to water depth ratios. Kang et al. (2012) proposed an analytical
model for analysing the annular flow induced vibration of a simply supported cylinder, while also taking into account the
effects of friction. Liu et al. (2012a, 2012b) used the multipole expansion method to obtain the analytical solution for the
diffraction and radiation problem for a submerged sphere in water of infinite depth and presented a set of numerical
solutions for a variety of submerged depths. Similarly, Chatjigeorgiou (2012) employed the multipole expansion method in
order to derive a solution for the hydrodynamic diffraction problem for a submerged oblate spheroid which is being excited
by regular waves in deep water. Mohapatra et al. (2013) explored the effects of compressive flow on the wave diffraction on
a 2-D floating elastic plate. The solution is derived for both the infinite and finite water depth cases using an integro-
differential equation method.
However, an analytical study of the wave scattering problem of a floating truncated cylinder in water of infinite depth has
not been previously attempted. Thus, in this paper, the method of separation of variables is employed to construct
approximate analytical expressions for the wave excitation forces, by considering the wave scattering problem, on a floating
truncated vertical cylinder in water of infinite depth. The results are compared with the output from a numerical
computational fluid dynamics (CFD) analysis that was undertaken using the boundary element method package, (ANSYS
AQWA, 2010). The presented analytical approximation is, also, compared to a curtailed version of the finite depth solution of
Bhatta and Rahman (2003), where the water depth was set to a value which is considered deep, 200a, in order to be
comparable to the infinite depth approximation presented here. Furthermore, the presented approximation is compared
with the experimental results given in Fonseca et al. (2011).
 W. Finnegan et al. / Journal of Fluids and Structures 40 (2013) 201–213 203

2. Methodology

Since the desired solution deals with the scattering problem, the wave excitation forces on a fixed truncated vertical
cylinder are considered with an incident wave of amplitude, A, and angular frequency, ω. A definition sketch depicting the
set-up for the truncated vertical cylinder case is shown in Fig. 1. The wave progresses in the positive x-direction with z¼ 0
corresponding to the still water level (SWL) and the positive z-direction being taken to point directly down into the water.
In this analysis, Linear (Airy) wave theory (Coastal Engineering Research Center, 1977) is used. Therefore, the following
assumptions are made in the derivation of the governing mathematical model:

 The water is both incompressible, as frequencies are low, and effectively inviscid.
 As the air has such a small density relative to the water, pressure change is negligible and, thus, is at constant pressure.
 Surface tension at the air–water interface is negligible.
 The water is at constant density and temperature.
 The Reynolds number for the flow is sufficiently small for the flow to remain laminar.
 The waves are progressive and only travel in one direction and the wave motion is irrotational.
 The incident waves are of small amplitude compared to their wave length.

Following Yeung (1981) and Bhatta and Rahman (2003) the fluid domain was divided into an interior region and an
exterior region. The interior region corresponds to the area underneath the cylinder and the exterior region constitutes the
remainder of the domain (Fig. 1). The problem is solved in the frequency domain. Therefore, the time domain velocity
potential, Φ, to be solved is transformed to the frequency domain, as follows:
Φðr; θ; z; tÞ ¼ Re½φðr; θ; zÞe−iωt  ð1Þ
where r is the radial distance from the z-axis, θ is the angle about the x-axis, i is the standard imaginary unit, ω is the wave
angular frequency of the wave, t is time and φ is the frequency domain velocity potential. Applied forces to the cylinder are
then calculated by integrating the relevant velocity potential over the wetted surface area of the cylinder, SB, using the
following equation (Linton and McIver, 2001):

F^ j ¼ iρω∬SB φnj dS ; ð2Þ

where ρ is water density, nj is the j-component of the normal and F^ j is defined implicitly by F j ¼ ReF^ j e−iωt , where F j is the
force in the j-direction. The equations and boundary conditions, in cylindrical coordinates, that need to be satisfied are:
Laplace's equation, the deep water condition, the free surface equation and the radiation condition, respectively (Linton and
McIver, 2001):
1 ∂φ ∂2 φ 1 ∂2 φ ∂2 φ
∇2 φ ¼ þ 2 þ 2 2 þ 2 ¼ 0; ð3Þ
r ∂r ∂r r ∂θ ∂z

j∇φj-0 as z-∞; ð4Þ

∂φ
ω2 φ þ g ¼ 0 on z ¼ 0; r≥a; ð5Þ
∂z
 
pffiffiffi ∂φd
lim r −ik0 φd ¼ 0; ð6Þ
r-∞ ∂r

Fig. 1. Definition sketch for the boundary value problem for a truncated vertical cylinder (Finnegan et al., 2011).
204 W. Finnegan et al. / Journal of Fluids and Structures 40 (2013) 201–213

where ∇ is the gradient, g is the acceleration due to gravity, a is the radius of the cylinder, φd is the diffraction velocity
potential and k0 ¼ω2/g is the wavenumber. Since the motion is irrotational and incompressible, Laplace's equation was
arrived at by substituting v ¼ ∇φ into ∇v ¼ 0, where v is the flow velocity. The solution being developed is for infinitely deep
water. Thus, the deep water condition defines the flow velocity near the sea bed. The free surface equation defines the
velocity potential at the free surface away from the floating body. The radiation condition provides a constraint on the form
of the velocity potential of the wave at distances from the body where the effect of the body on the wave has dissipated.
The velocity potential, φ, is a combination of the scattering velocity potential, φS , and the radiation velocity potential, φR ,
i.e. φ ¼ φS þ φR (Linton and McIver, 2001). Since it is the excitation forces that is being sought and the radiation velocity
potential determines the radiation forces on an oscillating body, only the scattering velocity potential needs to be derived.
The truncated vertical cylinder problem considers a vertical cylinder of radius, a, and of draft, b, with an incident wave of
amplitude, A, and angular frequency, ω, as depicted in Fig. 1. Since the scattering problem deals with the excitation force on a
fixed body, the following structural boundary conditions must be imposed:

∂φiS
¼ 0 on z ¼ 0; where z ¼ z−b; ð7Þ
∂z

∂φeS
¼ 0 at r ¼ a; 0 o z o b; ð8Þ
∂r
where φiS and φeS are the interior and exterior scattering velocity potentials, respectively (see Fig. 1). The unknown
coefficients are then determined by matching the interior and exterior scattering velocity potentials and their radial
derivatives along their common boundary, r ¼a.

2.1. Derivation of velocity potential for the interior domain

The method of separation of variables is used to formulate an expression for the velocity potential. Since the problem
deals with infinite depth, a Fourier sine/cosine transform is employed when dealing with the vertical or z-component.
For the interior region, in order to satisfy the Neumann boundary condition on z ¼ 0 (Eq. (7)), a Fourier cosine transform is
required. Thus, the Fourier cosine transform of the interior scattering velocity potential, φiS , is defined as follows:
rffiffiffi Z
2 ∞ i
F c ðφiS ðr; θ; zÞÞ ¼ φS ðr; θ; zÞcos ξz dz≡ χ i ðr; θ; ξÞ; ð9Þ
π 0

where 0 oξ o ∞ and F c is the Fourier cosine transform. The inverse transform is then:
rffiffiffi Z
i 2 ∞ i
φS ðr; θ; zÞ ¼ χ ðr; θ; ξÞcos ξz dξ: ð10Þ
π 0

The method of separation of variables is used to solve Laplace's equation (Eq. (3)) in order to formulate an expression for
φiS . Taking the Fourier cosine transform of each part in Eq. (3) gives:

1 ∂χ i ∂2 χ i 1 ∂2 χ i 2 i
þ 2 þ 2 2 −ξ χ ¼ 0: ð11Þ
r ∂r ∂r r ∂θ
Seeking separable variable solutions to Eq. (11) of the form χ i ðr; θÞ ¼ RðrÞPðθÞ yields the following pair of ordinary
differential equations:

r 2 R″ þ rR′−ðξ2 r 2 þ κÞR ¼ 0; ð12Þ

P″ þ κP ¼ 0; ð13Þ
2
where κ is a separation constant. Solving Eq. (13) and imposing 2π periodicity yields κ ¼m and
P m ðθÞ ¼ Am cos mθ for m ¼ 0; 1; 2; 3; … ð14Þ

where Am are arbitrary constants. Solving Eq. (12), and imposing boundedness as r-0, gives:
(
Bm0 r m if ξ ¼ 0
Rm ðrÞ ¼ ; ð15Þ
Bm Im ðξrÞ if ξ 40

where the Bm0 and Bm are arbitrary constants and Im is the modified first Bessel function of order m. The separable variable
solutions are thus:
8 
< p2m0 ar m cos mθ if ξ ¼ 0
i
χ m ðr; θ; ξÞ ¼ Rm ðrÞP m ðθÞ ¼ ðξrÞ ; ð16Þ
: pm ðξÞ IImmðξaÞ cos mθ if ξ 40
 W. Finnegan et al. / Journal of Fluids and Structures 40 (2013) 201–213 205

where pm0 ¼ am A0 Bm0 =2 and pm ðξÞ ¼ Am Bm Im ðξaÞ. Superposing these gives:

∞ pm0  r m Im ðξrÞ
χ im ðr; θ; ξÞ ¼ ∑ þ pm ðξÞ cos mθ; r oa; ð17Þ
m¼0 2 a Im ðξaÞ

and taking the inverse transform yields:

"rffiffiffi Z #
i
∞ 2 ∞ Im ðξrÞ
φS ðr; θ; zÞ ¼ ∑ pm ðξÞ cos ξz dξ cos mθ; r oa; ð18Þ
m¼0 π 0 Im ðξaÞ

where the choice pm0 ¼ 0 has been made to remove a Dirac delta function from φiS .

2.2. Derivation of velocity potential in the exterior domain

Next the exterior domain, r 4a, is considered. The exterior scattering velocity potential is the sum of the incident and
diffraction velocity potentials (i.e. φS ¼φI+φd). In this problem, a plane incident wave of amplitude, A, coming from x ¼ −∞
and propagating in the x-direction is considered, so that:
gA −k0 z ik0 rcos θ
φI ðr; θ; zÞ ¼ −i e e ; ð19Þ
ω
since x ¼ rcosθ. Kim (2008) has expanded this in the form:
gA −k0 z ∞
φI ðr; θ; zÞ ¼ − e ∑ εm imþ1 Jm ðk0 rÞcos mθ; ð20Þ
ω m¼0

where Jm is a Bessel function of the first kind of order m and εm is the Neumann symbol, defined by ε0 ¼ 1 and εm ¼2 for m≥1.
Similar to the interior domain, when dealing with infinite depth in the method of separation of variables a Fourier sine/
cosine transform is used. In order to satisfy the free surface equation (Eq. (5)), a combination of the Fourier sine and Fourier
cosine transform is required. Thus, the Fourier cosine transform of the exterior diffraction velocity potential, φed , is defined as
follows:
rffiffiffi Z
2 ∞ e  
Fðφed ðr; θ; zÞÞ ¼ φd ðr; θ; zÞ ξcos ξz−k0 sin ξz dz ¼ χ e ðr; θ; ξÞ; ð21Þ
π 0
where 0 o ξ o∞. In order to obtain the inverse Fourier transform, Havelock's expansion theorem (Chakrabarti, 2000) is used,
which states that if:
rffiffiffi Z
2 ∞ C ðξÞ
f ðzÞ ¼ C 0 e−Kz þ ½ξcos ξz−Ksin ξzdξ; 0 o ξ o∞; ð22Þ
π 0 ðξ2 þ K 2 Þ

Then
Z ∞
C 0 ¼ 2K f ðzÞe−Kz dz; ð23Þ
0

rffiffiffi Z
2 ∞
CðξÞ ¼ f ðzÞ½ξcos ξz−Ksin ξzdz; ð24Þ
π 0
where C0 and K are constants and f(z) and its derivatives are continuous and integrable in the range (0–∞). Therefore, the
inverse Fourier transform of Eq. (21) is given as:
rffiffiffi Z
e −k0 z 2 ∞ χ e ðr; θ; ξÞ  
φd ðr; θ; zÞ ¼ χ 0 ðr; θÞe þ ξcos ξz−k0 sin ξz dξ; ð25Þ
π 0 ξ 2 þ k0 2

where
Z ∞
χ 0 ðr; θÞ ¼ 2k0 φed ðr; θ; zÞe−k0 z dz: ð26Þ
0

Now χ e also satisfies Eq. (11) and the variables are separated as before to obtain Eq. (14). Insisting that the solution
remain bounded as r-∞, the following is then obtained:
Rm ðrÞ ¼ Am Km ðξrÞ; ð27Þ

where Km is the modified second Bessel function of order m. Superposing the separated solutions gives:

∞ Km ðξrÞ
χ e ðr; θ; ξÞ ¼ ∑ qm ðξÞ cos mθ; r 4a: ð28Þ
m¼0 Km ðξaÞ
206 W. Finnegan et al. / Journal of Fluids and Structures 40 (2013) 201–213

In Eq. (25), χ 0 ðr; θÞ corresponds to ξ¼−ik0 and so the corresponding radial dependence is given by (Linton and McIver,
2001):

Km ðξrÞ ¼ Km ð−ik0 rÞ ¼ 1=2πimþ1 Hð1Þ

m ðk0 rÞ; ð29Þ

where Hð1Þ
m is a Hankel function of the first kind of order m. Hence, the appropriate expansion for χ 0 ðr; θÞ is of the form:
" #
ð1Þ
∞ Hm ðk0 rÞ
χ 0 ðr; θÞ ¼ ∑ qm;0 ð1Þ cos mθ; r 4a: ð30Þ
m¼0 Hm ðk0 aÞ

Therefore, since the scattering velocity potential is the sum of the incident and diffraction velocity potentials
(i.e. φS ¼φI+φd) and incorporating −gAω−1εmim+1 into the φed ðr; θ; zÞ term in Eq. (25), the scattering velocity potential for
the exterior problem takes the form:
"( ) rffiffiffi Z #
gA ∞ Hð1Þ
m ðk0 rÞ 2 ∞ Km ðξrÞ qm ðξÞ  
φeS ðr; θ; zÞ ¼ − ∑ εm imþ1 Jm ðk0 rÞ þ qm;0 ð1Þ e−k0 z þ ξcos ξz−k 0 sin ξz dξ cos mθ:
ω m¼0 Hm ðk0 aÞ π 0 Km ðξaÞ ξ2 þ k0 2
ð31Þ

2.3. Determination of unknown coefficients

The unknown coefficients of pm(ξ) in Eq. (18), and qm,0 and qm(ξ) in Eq. (31), are found by matching the velocity potentials
and normal velocities across the boundary at r ¼a, and imposing the structural boundary condition. The conditions which
are to be satisfied at the boundary are:

∂φeS ðr; θ; zÞ
¼ 0 if 0 ≤z ≤b; ð32Þ
∂r r¼a

φeS ðr; θ; zÞ ¼ φiS ðr; θ; zÞ if b ≤z ≤∞; ð33Þ
r¼a r¼a

∂φeS ðr; θ; zÞ ∂φi ðr; θ; zÞ

¼ S if b ≤z ≤∞: ð34Þ
∂r r¼a ∂r r¼a

Leppington (1973) derives the velocity potential along the free-surface in the outer region, as the majority of the velocity
potential of significant magnitude is at the free-surface for high-frequency waves. Leppington (1973) then expands from this
velocity potential to derive the full field outer velocity potential. In dealing with the inner region, a rescaling of coordinates
is performed so a solid body (i.e. a dock) having a semi-infinite extent is being solved for. A similar technique is employed in
this paper in approximating the exterior scattering velocity potential, as the majority of this potential with significant
magnitude is close to the free-surface in the region 0 ozob. Since the present study is not restricted to high-frequency
waves, this assumption is deemed valid by ensuring that the draft to radius ratio remains greater than unity, i.e. b/a41,
throughout the analysis.
In order to derive an analytical approximation for the interior and exterior velocity potentials, the following procedure is
adopted. Closed form expressions for qm,0 and qm ðξÞ (and consequently φeS ) are constructed by taking b¼∞ in the boundary
condition, given in Eq. (32), which corresponds to calculating the exterior scattering velocity potential for a cylinder of
infinite draft. The matching condition, given in Eq. (33), is then imposed to determine pm(ξ) and φiS . Finally, the surge and
heave excitation forces are then found using the calculated forms for φeS and φiS , respectively. The justification for this
approach lies in the comparison with corresponding numerical CFD results. The parameter space for the problem under
consideration here is two dimensional and may be represented by the region 0ok0a, b/a o∞ in the (k0a, b/a) plane.
Extensive CFD work conducted for the current study indicates the analytical expressions for the surge and heave forces
derived are acceptable provided the non-dimensional parameters (k0a, b/a) lie in the regions 0 ok0a o∞ and 1ob/a o∞.
The procedure just described is now implemented. In the exterior region, the condition which is to be satisfied is given in
Eq. (32), where b¼∞, and is then applied to Eq. (31), yielding:
( ) rffiffiffi Z
ð1Þ
gA mþ1 0 Hm ðk0 aÞ −k0 z gA mþ1 2 ∞ ξKm 0
ðξaÞ qm ðξÞ  
− εm i k0 Jm ðk0 aÞ þ qm;0 ð1Þ e − εm i ξcos ξz−k0 sin ξz dξ ¼ 0; ð35Þ
ω Hm ðk0 aÞ ω π 0 Km ðξaÞ ðμ2 þ k0 2 Þ

which is then rearranged to give

rffiffiffi Z
Hð1Þ ðk0 aÞ −k0 z 2 ∞ ξKm 0 ðξaÞ qm ðξÞ   0
k0 qm;0 mð1Þ
e þ ξcos ξz−k0 sin ξz dξ ¼ −k0 Jm ðk0 aÞe−k0 z ; ð36Þ
Hm ðk0 aÞ π 0 K m ðξaÞ 2
ðμ þ k Þ
2
0
 W. Finnegan et al. / Journal of Fluids and Structures 40 (2013) 201–213 207

where the prime is the derivative. Havelock's expansion theorem (Chakrabarti, 2000) is then applied to Eq. (36) in order to
determine the unknown coefficients. Therefore,
Z ∞
Hð1Þ ðk0 aÞ 0 1 −2k0 z ∞
qm;0 m
ð1Þ
¼ −2k0 Jm ðk0 aÞe−k0 z e−k0 z dz ¼ −2k0 Jm
0
ðk0 aÞ ½e 0
0 ¼ −Jm ðk0 aÞ; ð37Þ
Hm ðk0 aÞ 0 2k0

and
rffiffiffi Z ∞ rffiffiffi " #
π 2 0 −k0 z
  2 Jm0
ðk0 aÞ k0 ξ
qm ðξÞ ¼ J ðk0 aÞe ξcos ξz−k0 sin ξz dz ¼ ξ −k 0 2 ¼ 0: ð38Þ
2 πðξ2 þ k2 Þ 0 m π ξ2 þ k2 ξ2 þ k
2
ξ þk
2
0 0 0 0

Next, using this result, taking Eqs. (18) and (31) and the condition in Eq. (33) yields:
rffiffiffi Z
2 ∞ gA  
pm ðξÞcos ξðz−bÞdξ ¼ − εm imþ1 Jm ðk0 aÞ þ qm;0 e−k0 z : ð39Þ
π b ω
Therefore, inverting the Fourier cosine transform gives:
rffiffiffi Z  
2 ∞ gA  
pm ðξÞ ¼ − εm imþ1 Jm ðk0 aÞ þ qm;0 e−k0 z cos ξðz−bÞdz
π b ω
rffiffiffi( )
gA 2 0 Hð1Þ ðk0 aÞ e−k0 b k0
¼− εm imþ1 Jm ðk0 aÞ−Jm ðk0 aÞ m : ð40Þ
ω π Hð1Þ
m ðk0 aÞ ξ þ k0
2 2

Consequently, the interior scattering velocity potential, given in Eq. (18), is written as
" ( )Z #
ð1Þ ∞ −k0 b
i gA 2 ∞ mþ1 0 Hm ðk0 aÞ e k0 Im ðξrÞ
φS ðr; θ; zÞ ¼ − ∑ εm i Jm ðk0 aÞ−Jm ðk0 aÞ ð1Þ cos ξz dξ cos mθ; ð41Þ
ω πm¼0 Hm ðk0 aÞ 0 ξ þ k0 Im ðξaÞ
2 2

and the exterior scattering velocity potential, given in Eq. (31), is written as
"( ) #
gA ∞ 0 Hð1Þ
m ðk0 rÞ
φeS ðr; θ; zÞ ¼ − ∑ εm imþ1 Jm ðk0 rÞ−Jm ðk0 aÞ ð1Þ e−k0 z cos mθ: ð42Þ
ω m¼0 Hm ðk0 aÞ

In this section, the scattering velocity potentials for the exterior region, r 4a, and interior region, r oa, were formulated
and the unknown coefficients were approximated by matching these velocity potentials at r ¼a. Therefore, for the truncated
cylinder case, the interior scattering velocity potential (Eq. (41)) and the exterior scattering velocity potential (Eq. (42)) are
now approximated.

3. Results

In this section, the wave excitation forces for both the truncated and infinite draft cylinder cases are derived using the
relevant velocity potentials and Eq. (2). The three motions from which the wave excitation forces are been calculated for the
truncated cylinder case are the surge, heave and pitch. Since, for the infinite draft cylinder case, there is no vertical, or heave
motion, only the surge excitation force is calculated. These solutions are presented for a range of frequencies and, for the
truncated cylinder case, for a range of draft to radius ratios also. In order to verify the accuracy of the analytical solution, the
wave excitation forces are compared to the output from a numerical computational fluid dynamics (CFD) analysis that was
undertaken using the boundary element method package, ANSYS AQWA (2010). Since the numerical CFD analysis uses a
finite depth solution, it is necessary to set the water depth in the model to 1000 m, in order to be comparable to the
presented infinite depth analytical approximation.

3.1. Wave excitation forces on a truncated cylinder

When calculating the surge, or horizontal, excitation force the only non-zero contribution is for m ¼1, as this gives the
R 2π
only non-zero value amongst the integral 0 cos mθ cos θ dθ, which arise in the force calculation. Furthermore, when
integrating the velocity potential over the surface area, the integration is performed only over the curved surface of the
cylinder and, hence, the exterior velocity potential at the surface boundary, r ¼a, given in Eq. (42), is used. Therefore, the
surge excitation force, F^ 1;ext , is given as:
Z 2π Z b
F^ 1;ext ¼ iρω∬SB φeS ðr; θ; zÞn1 dS ¼ −iρω φeS ða; θ; zÞcos θa dz dθ
0 0
( ) Z 2π
ð1Þ
ρgAa ∞ 0 Hm ðk0 aÞ
¼− ∑ εm i
m
Jm ðk0 aÞ−Jm ðk0 aÞ ð1Þ ð1−e−k0 b Þ cos mθ cos θ dθ
k0 m ¼ 0 Hm ðk0 aÞ 0
( )
2πiρgAa Hð1Þ ðk0 aÞ
¼− J1 ðk0 aÞ−J10 ðk0 aÞ 1ð1Þ ð1−e−k0 b Þ; ð43Þ
k0 H ðk0 aÞ
1
208 W. Finnegan et al. / Journal of Fluids and Structures 40 (2013) 201–213

where n1 ¼ −cos θ. Graphical representations of surge excitation forces with respect to incident wave frequencies for various
draft to radius ratios of truncated vertical cylinders are shown in Fig. 2. As can been seen, the presented analytical solution
and the numerical CFD solution are in good agreement even for quite modest draft to radius ratios. The correspondence
between the analytical and numerical CFD solutions improves as the draft to radius ratio increases as would be expected,
since the exterior velocity potential is derived assuming that the cylinder in question has infinite draft. The shaded region on
the graph indicates where all the possible choices of draft to radius ratio which will return a valid solution. The upper bound
is calculated where the draft is set to infinity and the lower bound is where the draft is set to zero, for a very thin plate, in
the presented approximation for the surge excitation force, given in Eq. (43). Furthermore, the phase angle of the excitation
forces as k0a varies is shown in Fig. 3, where the phase angle, β, is defined in the following equation:
^ cosðωt þ βÞ;
FðtÞ ¼ jFj ð44Þ
where F^ is the relevant excitation force. As can be seen from Fig. 3, there is good agreement between the analytical solutions
and the numerical CFD solutions for the phase angle over a range of frequency waves (i.e. k0a).
When calculating the heave, or vertical, excitation force from the velocity potential, the only non-zero value is for m ¼0
R 2π
since this gives the only non-vanishing term in the integral 0 cosmθ dθ. Furthermore, when integrating the velocity
potential over the surface area, the integration is performed only over the base of the cylinder and, hence, the interior
velocity potential, at z ¼ 0, given in Eq. (41), is used. Therefore, the heave excitation force, F^ 3;ext , is given as
Z 2π Z a
F^ 3;ext ¼ iρω∬SB φiS ðr; θ; zÞn3 dS ¼ iρω φiS ðr; θ; 0Þn3 r dr dθ
0 0
rffiffiffi Z 2π Z a Z ∞
2 ∞ Im ðξrÞ
¼ −iρω ∑ pm ðξÞ dξr drcos mθ dθ
πm¼0 0 0 0 Im ðξaÞ
rffiffiffi Z
2 ∞ I1 ðξaÞ
¼ −2πiρωa p0 ðξÞ dξ; ð45Þ
π 0 ξI0 ðξaÞ
where n3 ¼−1. Graphical representations of heave excitation forces with respect to incident wave frequencies for various
draft to radius ratios of truncated vertical cylinders are shown in Fig. 4 and the phase angle of the heave excitation forces is

2.5

2
|F1,ext|/πρgAa2

1.5

0.5

0
0 1 2 3
k0 a

Fig. 2. The normalised surge (or horizontal) excitation force, in the frequency domain, as a function of k0a, for various draft–radius ratios compared to the
corresponding numerical CFD analysis.

Fig. 3. The phase angle for surge and heave, in the frequency domain, as a function of k0a, compared to the corresponding numerical CFD analysis.
 W. Finnegan et al. / Journal of Fluids and Structures 40 (2013) 201–213 209

0.8

|F3,ext|/πρgAa2
0.6

0.4

0.2

0
0 1 2 3
k 0a

Fig. 4. The normalised heave (or vertical) excitation force, in the frequency domain, as a function of k0a, for various draft–radius ratios compared to the
corresponding numerical CFD analysis.

Fig. 5. The normalised pitch (or torque) excitation force, in the frequency domain, as a function of k0a, for various draft–radius ratios.

2
1.8
1.6
1.4
|F1,ext|/πρgAa2

1.2
1
0.8
0.6
0.4
0.2
0
0 1 2 3
k 0a

Fig. 6. The normalised surge (or horizontal) excitation force, in the frequency domain, as a function of k0a, for various draft–radius ratios compared
to Bhatta and Rahman (2003) and to the corresponding numerical CFD analysis.

shown in Fig. 3. Again, in Fig. 4, as with Fig. 2, the agreement between the presented analytical solution and the numerical
CFD solutions is good and improves as the draft to radius ratio increases. However, in Fig. 4, the upper bound for the
normalised heave is calculated where the draft is set to zero, for a very thin plate, and the lower bound is where the draft is
set to infinity in the presented approximation, given in Eq. (45). This is to be expected as the water particle velocity
210 W. Finnegan et al. / Journal of Fluids and Structures 40 (2013) 201–213

decreases exponentially from the still water level downwards and, therefore, the vertical force on the base will, in turn,
decrease as the draft to radius ratio increases.
The pitch, or torque, excitation force arises from the surge and heave forces on the wetted surface of the cylinder.
The pitch is taken about the axis which is transverse to the incident wave at the centre of the base, as shown by T in Fig. 1.
When calculating the pitch the only non-zero contribution, similar to surge, is from m ¼1. Therefore, the pitch excitation
force, F^ 5;ext , is given as

F^ 5;ext ¼ iρω∬SB φS ðr; θ; zÞn5 dS

Z 2π Z b Z 2π Z a
¼ iρω φeS ða; θ; zÞðz−bÞcos θa dz dθ−iρω φiS ðr; θ; 0Þr 2 cos θ dr dθ
0 0 0 0
( )
Hð1Þ
1 ðk0 aÞ 1−k0 b−e
−k0 b
¼ −2πiρgAa J1 ðk0 aÞ−J1 0 ðk0 aÞ
Hð1Þ
1 ðk0 aÞ k0
2

rffiffiffi Z
∞
2 I2 ðξaÞ
þπiρωa2 p1 ðξÞ dξ; ð46Þ
π 0 ξI1 ðξaÞ

where n5 ¼(z−b)n1−r cos θn3. Graphical representations of the analytical pitch excitation forces with respect to incident
wave frequencies for various draft–radius ratios of truncated vertical cylinders are shown in Fig. 5. The trends of Fig. 5 are
very similar to that of Fig. 2 as the magnitude of the pitch increases as the draft of the cylinder increases and the lower
bound is where the draft is zero, for a very thin plate. However, since it is pitch that is being calculated and that is
proportional to the draft an upper limit for the problem cannot be calculated. This is to be expected in view of the choice of
axis about which the pitch is taken. It is also noted that the phase angle of the pitch excitation forces is the same as that for
the surge excitation forces, as shown in Fig. 3, and this was also observed by Garrett (1971) for the finite depth case.
The analytical approximation presented in this paper is compared to a curtailed version of the finite depth solution of
Bhatta and Rahman (2003) in Figs. 6 and 7 for the normalised surge and heave excitation forces, respectively. In order to

0.8
|F3,ext|/πρgAa2

0.6

0.4

0.2

0
0 1 2 3
k 0a

Fig. 7. The normalised heave (or vertical) excitation force, in the frequency domain, as a function of k0a, for various draft–radius ratios compared to Bhatta
and Rahman (2003) and to the corresponding numerical CFD analysis.

Fig. 8. Comparison of the normalised heave and surge wave excitation forces with the experimental results of Fonseca (2011).
 W. Finnegan et al. / Journal of Fluids and Structures 40 (2013) 201–213 211

ensure the two sets of results are comparable, the water depth was set to a value which is considered deep, 200a, in the
finite depth solution of Bhatta and Rahman (2003). The results from the corresponding numerical CFD analysis are also
included in Figs. 6 and 7. It is clear to see from this comparison that the numerical model and analytical approximation yield
similar trends of normalised force varying with frequency waves (k0a). As the draft to radius ratio increases, the two sets of
results are found to be in very good agreement. Furthermore, it may be noted that the presented analytical approximation
converges quicker to the numerical CFD than the finite depth solution of Bhatta and Rahman (2003) in the case of the surge
wave excitation force calculation.
Using physical models, Fonseca et al. (2011) estimated the first order wave excitation forces on a truncated vertical
cylinder with a radius, a¼0.325 m, and a draft, b¼0.2 m. Since the experimental results are being compared to an infinite
depth approximation, only the regular wave observations for a water depth of 3 m are used in the comparison and are
displayed in Fig. 8. This independently observed experimental data is compared to the presented solution for a truncated
vertical cylinder with a draft to radius ratio, b/a ¼0.62. From Fig. 8, it can be seen that the presented analytical expressions
are in good agreement with the experimental data.

3.2. Wave excitation forces on a cylinder of infinite depth

The scattering velocity potential, φS , was derived in Section 2.3 for a cylinder of infinite draft yielding the solution given
in Eq. (42). On the other hand, for a bottom mounted cylinder in water of finite depth, d, and using a concurrent axis
orientation as that presented in this paper, MacCamy and Fuchs (1954) obtained the solution:
( )
gA cos hk0 ðd−zÞ ∞ mþ1 0 Hð1Þ ðk0 rÞ
φS ðr; θ; zÞ ¼ − ∑ εm i Jm ðk0 rÞ−Jm ðk0 aÞ m cos mθ; ð47Þ
ω cos hk0 d m ¼ 0 ð1Þ
Hm ðk0 rÞ

for a bottom mounted cylinder in water of finite depth, d. For the infinite depth case, the z-component of the equation is
replaced with e−k0 z . Thus, the solution presented in Eq. (42) corresponds with the infinite depth version of MacCamy and
Fuchs (1954) solution given in Eq. (47). The surge excitation force, in the frequency domain, F^ 1;ext , is then calculated as
follows:
Z 2π Z ∞
F^ 1;ext ¼ iρω∬SB φS ðr; θ; zÞn1 dS ¼ iρω φS ða; θ; zÞn1 a dz dθ
0 0
" #Z
ρgAa ∞ m 0 Hð1Þ ðk0 aÞ 2π
¼− ∑ εm i Jm ðk0 aÞ−Jm ðk0 aÞ m
ð1Þ
cos mθcos θ dθ
k0 m ¼ 0 Hm ðk0 aÞ 0
" #
2πiρgAa Hð1Þ ðk0 aÞ
¼− J1 ðk0 aÞ−J1 0 ðk0 aÞ 1ð1Þ ; ð48Þ
k0 H ðk0 aÞ
1
R 2π
where n1 ¼−cos θ and the only non-zero solution to 0 cos mθ cos θ dθ is when m¼ 1. Since the exterior velocity potential is
calculated for the exterior domain, the solution is the same as when b ¼∞ in Eq. (43). In Fig. 9, the presented solution is also
compared to a numerical CFD solution, performed using the BEM software ANSYS AQWA (2010). Since the draft of the
cylinder cannot be set to infinity in the numerical CFD solution, a draft to radius ratio, b/a ¼20, is used. This comparison of
the two solutions is found to be in good agreement except at low frequencies. This difference is as a result of the parameter,
b/a ¼20, in the numerical CFD solution and greater values for the draft to radius ratio would yield closer agreement. When
this parameter is applied to the analytical solution, i.e. b/a ¼20, the two solutions match up very well as can be seen in Fig. 9.

Fig. 9. Dependence of normalised surge excitation force on k0a and compared to the corresponding numerical CFD analysis.
212 W. Finnegan et al. / Journal of Fluids and Structures 40 (2013) 201–213

4. Discussion and conclusions

An analytical approximation to determine the wave excitation forces, by solving the scattering problem, on a floating
truncated vertical cylinder in water of infinite depth has been presented in this paper. The presented analytical
approximation provides a solution which is far easier to use and implement than the already available analytical solutions.
For example, the solution for the finite depth case provided by Bhatta and Rahman (2003) requires a considerable amount of
further work if numerical values for a particular case are needed. On the other hand, the formulation presented in this paper
yields an easy to implement analytical approximation for the surge, heave and pitch wave excitation forces on a truncated
cylinder. This was achieved by imposing an approximation in deriving the unknown coefficients in the exterior region.
Furthermore, the presented analytical approximation still retains a high level of accuracy in calculating the heave and surge
wave excitation forces, when the draft to radius ratio is greater than unity.
The method of separation of variables is employed in order to formulate the velocity potentials and the unknown
coefficients are derived by matching these potentials along their common boundaries. In order to create an analytical
solution, the exterior velocity potential is derived for the case of a vertical cylinder of infinite draft and the interior velocity
potential is calculated by matching across their common boundary. However, it should be noted that this assumption is a
good approximation when the draft to radius ratio is greater than or equal to unity (i.e. b/a≥1), which can been seen in
Figs. 2, 4 and 5, where the presented analytical solution and numerical CFD solution are in increasingly better agreement as
the draft to radius ratio (b/a) increases.
The relationships between wave excitation forces and wave frequency (k0a) are shown graphically in Figs. 2, 4 and 5 for a
truncated cylinder having various draft to radius ratios. It may be noted that the pitch excitation forces are also derived in
the study, along with the heave and surge excitation forces. These results are validated by comparing to a numerical CFD
solution, performed using the BEM software ANSYS AQWA (2010), and they are found to be in very good agreement.
A shaded region on the graph is also included indicating where a viable solution may be obtained. For surge and pitch
motions, the upper bound is calculated where the draft is infinite and the lower bound is where the draft is zero, for a very
thin plate. However, for heave motion, the upper bound of the viable solution region is calculated where the draft is zero, for
a very thin plate, and the lower bound is where the draft is infinite. This is to be expected as the water particle velocity
decreases exponentially from the still water level downwards and, therefore, the vertical force on the base will, in turn,
decrease as the draft to radius ratio increases. These three figures (Figs. 2, 4, and 5) and the associated approximations
provide an efficient and relatively quick solution to the problem for an engineer or professional investigating the wave
excitation forces on a floating structure in deep water.
The phase angles of the wave excitation forces, relative to the incident wave, are also presented graphically in Fig. 3 and
compared to the results from a numerical CFD analysis. The presented analytical results were found to be in good agreement
with the results from the numerical CFD analysis. Furthermore, when compared to the independently observed
experimental data given in Fonseca et al. (2011), the analytical solution for a truncated vertical cylinder shows good
agreement, as shown in Fig. 8.
In addition, the presented analytical approximation is compared to a curtailed version of the finite depth solution of
Bhatta and Rahman (2003), which is shown graphically for the normalised surge and heave excitation forces in Figs. 6 and 7,
respectively. It is clear from this comparison that similar trends occur with respect to the relation between the normalised
excitation forces and frequency waves (i.e. k0a) for the presented analytical approximation and a curtailed version of the
finite depth solution of Bhatta and Rahman (2003), as with the numerical CFD results. As the draft to radius ratio increases,
the two sets of results are found to be in very good agreement.
Furthermore, when deriving the exterior velocity potential it was found to match to the infinite depth version of the
MacCamy and Fuchs equations (MacCamy and Fuchs, 1954). In addition, the presented solution, for the infinite draft cylinder
case, is also compared to a numerical CFD solution and they are found to be in very good agreement, which can be seen in
Fig. 9.
The analytical solution provided in this paper provides an engineer with a very good approximation of the wave
excitation forces on a structure which may be represented by a truncated cylinder. Furthermore, if only the k0a value is
known, the maximum forces can easily be calculated from the analytical solution, or estimated from the shaded regions of
the graphs in Figs. 2, 4, and 5, to aid in the design stage of an offshore structure.

Acknowledgements

The first author would like to acknowledge the financial support from the National University of Ireland under the
College of Engineering & Informatics Postgraduate Fellowship. The authors would also like to express their gratitude to the
anonymous reviewers of this paper for their constructive comments.

References

ANSYS AQWA, 2010. Release, vol. 13. ANSYS-Inc., Pennsylvania, USA.

Bai, K.J., 1975. Diffraction of oblique waves by an infinite cylinder. Journal of Fluid Mechanics 68 (3), 513–535.
 W. Finnegan et al. / Journal of Fluids and Structures 40 (2013) 201–213 213

Bhatta, D.D., Rahman, M., 2003. On scattering and radiation problem for a cylinder in water of finite depth. International Journal of Engineering Science 41
(9), 931–967.
Black, J.L., 1975. Wave forces on vertical axisymmetric bodies. Journal of Fluid Mechanics 67 (2), 369–376.
Chakrabarti, A., 2000. On the solution of the problem of scattering of surface-water waves by the edge of an ice cover. Proceedings: Mathematical, Physical
and Engineering Sciences 456 (1997), 1087–1099.
Chatjigeorgiou, I.K., 2012. Hydrodynamic exciting forces on a submerged oblate spheroid in regular waves. Computers & Fluids 57, 151–162.
Coastal Engineering Research Center, 1977. Shore Protection Manual/US Army Coastal Engineering Research Center. Fort Belvoir, VA, Washington (Supt. of
Docs., US Govt. Print. Off).
Finnegan, W., Meere, M., Goggins, J., 2011. The wave excitation forces on a floating vertical cylinder in water of infinite depth, in: World Renewable Energy
Congress, 2011, Linkoping, Sweden.
Fonseca, N., Pessoa, J.o., Mavrakos, S., Le Boulluec, M., 2011. Experimental and numerical investigation of the slowly varying wave exciting drift forces on a
restrained body in bi-chromatic waves. Ocean Engineering 38 (17–18), 2000–2014.
Garrett, C.J.R., 1971. Wave forces on a circular dock. Journal of Fluid Mechanics 46 (1), 129–139.
Hassan, M., Bora, S.N., 2012. Exciting forces for a pair of coaxial hollow cylinder and bottom-mounted cylinder in water of finite depth. Ocean Engineering
50, 38–43.
Havelock, T., 1955. Waves due to a floating sphere making periodic heaving oscillations. Proceedings of the Royal Society of London. Series A, Mathematical
and Physical Sciences 231 (1184), 1–7.
Hsu, H.-H., Wu, Y.-C., 1997. The hydrodynamic coefficients for an oscillating rectangular structure on a free surface with sidewall. Ocean Engineering 24 (2),
177–199.
Kang, H.S., Mureithi, N.W., Pettigrew, M.J., 2012. Analytical solution for a vibrating simply-supported cylinder subjected to 2-D concentric annular flow,
considering friction. Journal of Fluids and Structures 35, 1–20.
Kim, C.H., 2008. Nonlinear Waves and Offshore Structures. Advanced Series on Ocean Engineering. World Scientific Publishing Co. Pte. Ltd., New Jersey, USA.
Leppington, F.G., 1973. On the radiation and scattering of short surface waves. Part 3. Journal of Fluid Mechanics 59 (1), 147–157.
Linton, C.M., McIver, P., 2001. Handbook of Mathematical Techniques for Wave/Structure Interactions. Chapman & Hall/CRC, Boca Raton, FL.
Liu, Y.-y., Teng, B., Cong, P.-w., Liu, C.-f., Gou, Y., 2012a. Analytical study of wave diffraction and radiation by a submerged sphere in infinite water depth.
Ocean Engineering 51, 129–141.
Liu, Y., Li, H.-J., Li, Y.-C., 2012b. A new analytical solution for wave scattering by a submerged horizontal porous plate with finite thickness. Ocean
Engineering 42, 83–92.
MacCamy, R.C., Fuchs, R.A., 1954. Wave Forces on Piles: A Diffraction Theory, in Tech. Memo No. 69. US Army Corps of Engineers, Beach Erosion Board,
Washington, D.C., USA.
Mansour, A.M., Williams, A.N., Wang, K.H., 2002. The diffraction of linear waves by a uniform vertical cylinder with cosine-type radial perturbations. Ocean
Engineering 29 (3), 239–259.
Mohapatra, S.C., Ghoshal, R., Sahoo, T., 2013. Effect of compression on wave diffraction by a floating elastic plate. Journal of Fluids and Structures 36,
124–135.
Ursell, F., 1949. On the heaving motion of a circular cylinder on the surface of a fluid. Quarterly Journal of Mechanics and Applied Mathematics 2 (2),
218–231.
Yeung, R.W., 1981. Added mass and damping of a vertical cylinder in finite-depth waters. Applied Ocean Research 3 (3), 119–133.