Professional Documents
Culture Documents
Ocean Engineering
journal homepage: www.elsevier.com/locate/oceaneng
A R T I C L E I N F O A B S T R A C T
Keywords: The aim of this study is to present numerical procedures for hydrodynamic analysis of a heaving-buoy-type WEC
Wave energy converter connected to a fixed jacket structure that is currently being actively studied. In order to accurately assess the
Point absorber effect of the float on the fixed structure, the augmented formulation procedure used in multibody dynamics was
Augmented formulation
used to calculate the reaction force due to the float motion acting on the hinge point of the structure. The hy
Dynamic response
Reaction force
drodynamic coefficients and wave excitation forces acting on the floats were calculated using the three-
dimensional frequency domain solver WAMIT. The wave radiation forces produced by the movement of floats
were calculated using the Cummins’ equation in the time domain. Compared to the classical method of formula-
derivation technique, the augmented formulation, which is an automated formulation process, has an advantage
that it can be easily calculated even if the degree of freedom is increased. Thus, it can be easily extended to
analyze hydrodynamic performance of various multibody ocean structures. To conduct the time integration, the
implicit Runge-Kutta 4th-order method was applied to the classical method. A two-loop numerical integration
with the Newmark-beta explicit method was used in the augmented formulation method. The responses of the
float and reaction force acting on the hinge point were evaluated using the two numerical procedures. Finally,
the calculation results of two floats for a multi-buoy HPA WEC platform were analyzed to emphasize the
importance of reaction force analysis.
* Corresponding author.
E-mail addresses: sanghwan.heo@gmail.com (S. Heo), nwavetank@gmail.com, wckoo@inha.ac.kr (W. Koo).
https://doi.org/10.1016/j.oceaneng.2020.107070
Received 18 August 2019; Received in revised form 10 December 2019; Accepted 2 February 2020
Available online 21 February 2020
0029-8018/© 2020 Elsevier Ltd. All rights reserved.
S. Heo and W. Koo Ocean Engineering 200 (2020) 107070
and reaction force acting on the hinge point were evaluated using the
two numerical procedures.
Finally, the calculation results of two floats for multi-buoy HPA type
WEC platform were analyzed to emphasize the importance of reaction
force analysis.
2. Mathematical formulation
2
S. Heo and W. Koo Ocean Engineering 200 (2020) 107070
Fig. 2. Descriptions of a float connected to a hinge point by an arm. Left: in 2D. Right: in 3D coordinate system.
� �
and �FDiff WAMIT � are the magnitude of the Froude-Krylov and diffraction where m is the mass of the float. Vs is submerged volume of the float in
forces calculated using WAMIT, respectively. ψ FK and ψ Diff are the phase still water. Aw is the water plane area of the float. uz ðtÞ is the vertical
angles of the Froude-Krylov and diffraction forces, respectively. These displacement of the float from the water surface.
phase angles vary depending on the float location. ω denotes the inci
dent wave frequency. Subscript k denotes the direction of the variables. 2.2. Equation of motion for the floats
The float moves both horizontally and vertically because the move
ment of the float is constrained by the arm. Therefore, the wave radia In this study, two numerical procedures were presented to solve the
tion forces generated by motion in two directions should be considered. motion of the float and calculate the reaction force acting on the hinge
The wave radiation force can be expressed using the Cummins’ equation point. The first method is a classical method using Newton’s second law
and the force consists of the acceleration proportional term and the for motion. This method derives the equation of motion considering the
velocity proportional term. The latter term can be expressed as a forces acting on the structure. Chao et al. (2018) obtained the equation
convolution integral using the retardation function (impulse response of motion for a two-body articulated wave energy device with a complex
function or memory function) and the velocity of a float (Cummins, geometric shape using Newton’s second law. In this procedure, how
1962; Ogilvie, 1964). The wave radiation forces in the k direction can be ever, additional equations using dynamic equilibrium are required to
expressed as determine the constraint force. This method has the disadvantage that it
Z t is very difficult to derive an equation if the structure is slightly more
FRad;k ¼ ma∞ ;kk ⋅u€k Kkk ðτÞ⋅u_k ðt τÞdτ (4) complicated. The second method is to use an augmented formulation.
This method is difficult to build on the initial process, but it is possible to
0
where ma∞ ;kk is the added mass with infinite frequency. uk is the automate the analysis process. Also, this method can evaluate both the
displacement of the float and Kkk ðtÞ is the retardation function, which motion of the float and constraint force, simultaneously. To do this, it
can be written as requires relatively more computer memory and calculation time than
Z the classical method. However, increasing the number of floats is not a
Kkk ðtÞ ¼
2 ∞
Bkk ðωÞ⋅cosðωtÞdω (5) probelm because the absolute amount of computer memory required is
π 0 small. If only calculation time is considered, the classical method is a bit
more advantageous. However, as mentioned earlier, considering the
where Bkk ðωÞ is the radiation damping coefficient calculated by WAMIT. entire analysis process, the classical method is very inefficient because
Three methods can be used to compute the convolution integral in new equations must be derived in each case. Therefore, the augmented
Cummins’ equation: Direct integration, State space method, and Prony formulation method has the advantage of greatly reducing the prepa
method (Armesto et al., 2015). In this study, the direct integration ration time for analysis. The results calculated by both procedures were
method was applied. showed and verified with the reference in chapter 3. These procedures
The PTO system is connected to the WEC to extract the wave energy. should give almost the same results because both processes are
Among the many types of PTO damping forces, a linear PTO damping
moment, as shown in Eq. (6), was applied to the external moment in this
Table 1
study. The horizontal and vertical PTO damping forces can be calculated
Typical characteristics of the two methods.
easily using the trigonometric function.
Classical method Augmented formulation
MPTO ¼ BPTO ⋅θ_ (6) Type of equation of Differential equation Differential-Algebraic
motion equation
where BPTO is the damping coefficient of the PTO system. Derivation of equation Hard (Depending on the Easy (Automatically
The self-weight of the float can be considered as a constant external of motion structure, new equations calculated)
force. In still water, the buoyancy also acts as a constant external force should be derived)
Determination of Hard (Requires additional Easy (Automatically
on the submerged volume. On the other hand, the buoyancy changes constraint forces equations) calculated)
with time if there is a wave. In this study, a linear buoyant force was Process automation Hard (The equations Easy
applied to the floats. This means that the submerged volume changes depend on the structure)
linearly depending on the submerged shape of the float. The draft of the Second time derivative Unnecessary Necessary
of constraint
floats was satisfied by ballasting water. This means that the self-weight
equations
and buoyant force in still water cancel each other out. When a float is in Computer memory Small Relatively large
a sea water, the self-weight and linear buoyant force can be expressed as required
Applicability Low High (Required modules
FWeight þ FBuoyant ¼ mg þ ρw gðVs Aw uz ðtÞÞ ¼ ρw gAw uz ðtÞ (7) can be added to the
process)
3
S. Heo and W. Koo Ocean Engineering 200 (2020) 107070
interpreting the same physical phenomena. Table 1 lists the typical calculated using dynamic equilibrium. The arm was assumed to be a
characteristics of the two methods. The detailed procedures for calcu massless rigid arm, and the problem can be considered to be a simple
lating the response and reaction force using two methods are described pendulum problem. Fig. 3 presents the dynamic equilibrium of a simple
in the following sections. pendulum problem. In Fig. 3, an object is connected to the hinge point
with a massless rigid arm and rotates about the hinge point. The applied
2.2.1. Classical method forces acting on the system (left side of Fig. 3) and the inertia forces
According to Newton’s second law for rotation, the sum of the mo (right side of Fig. 3) should be equal. The inertia forces can be divided
ments acting on the hinge point is equal to the moment of inertia of the into two components depending on the direction of body rotation:
float multiplied by the angular acceleration. normal and tangential components. In particular, the normal directional
X component is the centripetal force that makes a body follow a curved
I€
θ¼ M y ¼ MFK þ MDiff þ MRad þ MPTO þ MWeight þ MBuoyant (8) path. These forces can be calculated using the angular velocity and ac
celeration as follows:
where I is the moment of inertia of the float. M y is the moment acting
on the float in the inverse y-axis. The inverse y-axis direction means the (12)
2
Fn ¼ mlθ_
same axis with the axis of rotation, as shown in Fig. 2. Several studies
have calculated the moments for the center of rotation directly and Ft ¼ ml€
θ (13)
substituted into Eq. (8) (Zurkinden et al., 2014; Flavia� et al., 2017). The
For the applied forces, the notation Q in Fig. 3 represents the external
translational forces were used to calculate the moments as the following
forces acting on the float, as shown in Eq. (1). The notation R in Fig. 3
equation:
denotes the tension in the arm, and it can be decomposed into horizontal
� ! ! ! ��
� i k j �
�
X ! ! � � ! ! � X X � !
I €θ ¼ M y ¼ ð!
r � Q Þ⋅ð j Þ ¼ � rx rz 0 �⋅ð j Þ ¼ ð j Þ⋅ rx ⋅ Fz rz ⋅ Fx ⋅ð j Þ
�X
� P � (9)
Fx Fz 0 �
� X X �
¼ rx ⋅ Fz r z ⋅ Fx
and vertical forces, Rx and Rz , respectively. These forces are from the
hinge point acting on the arm. According to Newton’s third law, the
The relationship between the rotation angle and the translational forces acting on the hinge point by the arm are equal in magnitude and
displacement of the float can be derived as the following equations: opposite to the direction of the tension in the arm. In summary, the
reaction forces acting on the hinge point can be obtained as follows:
ux ¼ l cos θ þ xH uz ¼ l sin θ þ zH
_
u_x ¼ lθsin θ _
u_z ¼ lθcos θ (10) FH;x ¼
2
Rx ¼ Qx þ ml θ_ cos θ þ €
θsin θ
�
(14)
2 2
u€x ¼ lθ_ cos θ l€θsin θ u€z ¼ lθ_ sin θ þ l€
θcos θ
�
(15)
2
where ðxH ; zH Þ denote the location of the hinge point. The equation of FH;z ¼ Rz ¼ Qz þ ml θ_ sin θ €
θcos θ
motion of the float can be obtained by substituting Eqs. (2)–(6) and (10) The rotation angle and angular velocity can be combined into a state
into Eq. (9) and rearranging both sides: vector to obtain the numerical solution as follows:
� � �
Iþma∞ ;zz ⋅r2x þma∞ ;xx ⋅r2z ⋅€ _
θþðBPTO ⋅θÞþf ρw gAw ⋅ðrz þzH Þ⋅rx g θðtÞ
�Z t Z t � f ðtÞ ¼ _ (16)
θðtÞ
þl⋅ Kzz ðτÞ⋅cosðθðt τÞÞ⋅θðt _ τÞdτ⋅rx _
Kxx ðτÞ⋅sinðθðt τÞÞ⋅θðt τÞdτ⋅rz
�
0
� �� �
0
� 2 The first time derivative of the above equation is a state-space
¼ rx ⋅ FFK;z þFDiff ;z rz ⋅ FFK;x þFDiff ;x þ rx ⋅rz ⋅ðma;zz ma;xx Þ ⋅θ_ equation of the system. Using this equation, the second-order differen
(11) tial equation can be replaced by the first-order differential equation. In
this study, Runge-Kutta 4th-order method was applied for numerical
The motion of the float is constrained by the arm with a hinge joint.
time integration of this equation. The state-space equation can be
The external forces acting on the float are transmitted to the hinge point
expressed using the following equation:
through the arm, and the forces due to the movement of the float also act
� � � �
on the hinge point. The constraint force acting on the float can be _
θðtÞ _
θðtÞ
f_ðtÞ ¼ € ¼ � _ 2 _
��
θðtÞ BðtÞ þ DðtÞ⋅fθðtÞg EðtÞ BPTO ⋅θðtÞ GðtÞ JðtÞ
(17)
where
� �
Hinge Hinge BðtÞ ¼ rx ⋅ FFK;z þ FDiff ;z rz ⋅ FFK;x þ FDiff ;x
positive
direction
= positive
direction
DðtÞ ¼ rx ⋅rz ⋅ðma;zz
�Z t
ma;xx Þ
EðtÞ ¼ l⋅ _
Kzz ðτÞ ⋅ cosðθðt τÞÞ⋅θðt τÞdτ ⋅ rx
0
Z t �
_
Kxx ðτÞ ⋅ sinðθðt τÞÞ⋅θðt τÞdτ ⋅ rz
Fig. 3. Dynamic equilibrium of a simple pendulum problem (R:force at the 0
hinge point acting on the arm, Q: external forces acting on the float).
4
S. Heo and W. Koo Ocean Engineering 200 (2020) 107070
GðtÞ ¼ ρw gAw ⋅ðrz þ zH Þ⋅rx the two-loop procedure is the iterative process of the explicit or implicit
integration schemes, and the inner loop is the Newton-Raphson iterative
JðtÞ ¼ I þ ma∞ ;zz ⋅r2x þ ma∞ ;xx ⋅r2z method for solving Eq. (19). In this study, the Newmark-beta explicit
method was employed for the outer loop of the two-loop procedure.
The classical method, as explained so far, has the drawback that the In planar motion, an unconstrained object has three degrees of
translational motion should be converted to rotational motion in order freedom (two translational motions and a single rotational motion). In
to calculate the rotation angle of the float. In addition, the dynamic the case of the hinge joint, two translational motions are constrained by
equilibrium of the system should be considered to calculate the reaction one arm. Therefore, the object has one degree of freedom, which is the
forces acting on the hinge point. In the case of a complex system, the rotational motion (θ) about the rotation axis. In other words, the motion
equation of motion would be much more difficult to describe. of the float can be described using only the angle of rotation, and the
translational displacements (xf ; zf ) are dependent on this angle. In this
2.2.2. Augmented formulation case, the rotation angle can be defined as an independent coordinate
As mentioned in the previous section, it is difficult to calculate the (!u i ), and the translational displacements can be defined as dependent
constraint forces directly when applying Newton’s second law. The coordinates (! u d ). The conditions for the hinge joint are that the point
augmented formulation, which is one of a multibody dynamics’ where the two objects are connected remains constant throughout the
formulation, can be used to overcome this problem. Using the entire motion. Fig. 4 presents two bodies, the fixed wall and a float with
augmented formulation, the rotational motion of the float and the re an arm, which are connected by a hinge joint. The constraint equation
action forces acting on the hinge point can be calculated at the same for the hinge joint can be expressed as follows:
time. The constrained system is described using the following differen
tial and algebraic equations: Cð!
u ; tÞ ¼ !
r Of !
r OH !
r Hf ¼ 0 (23)
T
ug þ ½Cu � fλg ¼ fQg
½M�f€ (18) where !r ij denotes the position vector from point i to point j. The point O
is the origin of the system coordinate and can be set to ð0; 0Þ for
Cð!
u ; tÞ ¼ 0 (19)
simplicity. The r vectors can be expressed as follows:
where ½M� is the mass matrix of the float, Cð! u ; tÞ is the constraint !
r OH ¼ ðxH ; zH Þ (24)
equations of the system at the displacement level, ½Cu � is the partial �
differential matrix of constraint equations, fλg is the Lagrange multiplier !
r Of ¼ xf ; zf (25)
vector. The first and second derivatives of Eq. (19) define the constraint � �
equations at the velocity and acceleration levels as follows: !
r Hf ¼ ½A� rf (26)
Cu u_ ¼ Ct (20)
where ½A� is the transformation matrix from the arm-float coordinate
system to the global coordinate system. The vector frf g is the position
Cu u€ ¼ Qc ¼ ððCu uÞ
_ u u_ þ 2Cut u_ þ Ctt Þ (21)
vector of the float in arm-float coordinate system as follows:
where Ct is the partial derivative of the constraint equation with respect �
cos θ sin θ
�
to the time, Qc is the vector related to the second time derivative of the ½A� ¼ (27)
sin θ cos θ
constraint equation. Eqs. (18) and (21) can be combined to form the
following augmented formulation (Shabana, 2013): � �
rf ¼ ½ l 0 �T (28)
� �� � � �
½M� ½Cu �T fug
€ fQg Substituting Eqs. (24)–(28) into Eq. (23), the constraint equations
¼ (22)
½Cu � 0 fλg fQc g can be expressed as
The above equations can be solved numerically to obtain the accel �
xf xH l cos θ
�
erations and Lagrange multipliers vectors. In the numerical process, the Cð!u ; tÞ ¼ ¼0 (29)
zf zH l sin θ
solutions of Eq. (22) should always be satisfied with Eqs. (19) and (20).
Satisfying Eqs. (19), (20) and (22) means that the response vectors of the The constraint Jacobian matrix can be expressed as
float satisfy the constraint equations of the system at the displacement, � �
1 0 l sin θ
velocity and acceleration levels. To solve these equations in the time Cu ðtÞ ¼ (30)
0 1 l cos θ
domain, two-loop numerical integration procedures were proposed
(Shabana and Hussein, 2009; Zhang and Zhang, 2016). The outer loop of The constraint forces acting on the float can be calculated by
multiplying Eq. (30) by the Lagrange multiplier vector. The constraint
forces acting on the hinge point can be obtained by changing the point of
Hinge action from the float to the hinge point. The translational constraint
z point xH , z H forces are the same, but the moment should be changed according to the
following equation:
� !�T �T
fFH g ¼ ½Cu �T fλg 0 0 ! r Hf � Q (31)
x
Augmented formulation, as explained thus far, has the disadvantage
rOH rHf that a second derivative with respect to the time of the constraint
equation is required. On the other hand, this method has the advantages
that it is easy to program, and the constraint forces are calculated
automatically at every time step.
xO , zO rOf xf , zf
O
Fig. 4. Description of the hinge joint in planar motion.
5
S. Heo and W. Koo Ocean Engineering 200 (2020) 107070
0.3
0.25
0.2
RAO [rad/m]
0.15
0.1
0.05
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Wave frequency [rad/s]
using the Cummins’ equation. The damping coefficient of the PTO sys
tem (BPTO ) of 3.52 � 106 N⋅m⋅s was applied for analysis.
Fig. 5. Hemispherical float constrained to a hinge point (Flavi�
a et al., 2017).
Fig. 6 shows the time history of the rotation angle of the float ac
cording to the calculation method. For the simulation, a wave height of
-0.6 2 m and wave frequency of 1.0 rad/s were applied. The ramp function
was applied for the first 30 s to prevent sudden changes. Regardless of
the calculation method, the rotation angle was similar over the entire
-0.7 time range. This means that the angular velocity and angular accelera
tion of the float are the same in both methods, and the reaction forces
acting on the hinge point will also be the same. Fig. 7 shows the time
-0.8
history of the horizontal and vertical reaction forces acting on the hinge
point. As expected, the results were similar in both calculation methods.
-0.9 To verify the validity of both methods, the Response Amplitude
Operator (RAO) for the rotational motion of the float were compared
with the reference (Flavi�a et al., 2017). Fig. 8 compares the pitch RAO of
-1
0 10 20 30 40 50 60 70 80 90 100 the float. The maximum error between the calculation results and the
Time [sec] reference was approximately 3% over the entire frequency range. The
trend and magnitude of the pitch RAO agreed well with the reference
Fig. 6. Time history of the rotation angle of the float (H ¼ 2 m, ω ¼ values. Therefore, the validity of the calculation methods used in this
1:0 rad=s).
study was verified.
In addition to the above results, Fig. 9 compares the spectrum of the
3. Numerical results and discussion horizontal and vertical reaction force magnitudes between the calcula
tion methods. The peak frequency of the reaction force spectra was the
3.1. Validation of calculation methods same with the peak frequency of the pitch RAO. This means that the
reaction force acting on the hinge point increased with increasing
Prior to the validation, the results of two procedures were compared rotational motion of the float.
to check whether they provided the same results. Fig. 5 presents a In Figs. 8 and 9, the results of both methods were similar over the
hemispherical float constrained to a hinge point in the two-dimensional given frequency range. This means that both procedures well analyzed
coordinate system (Flavia � et al., 2017). The float could only rotate about the responses of the float and its effects. In terms of set up efficiency for
the axis of rotation (inverse y-axis, as shown in Fig. 2). The diameter of numerical analysis, the classical method is very inefficient because it
the float was 6 m, and the water depth was 20 m. The direction of the requires the formula-derivation process every time the analysis target
incident wave was the x-axis. The hydrodynamic coefficients and wave changes. Therefore, building the numerical process using the augmented
exciting forces acting on the float were calculated using WAMIT in the formulation is more efficient in terms of efficiency and applicability.
frequency domain. The time domain motions of the float were described
Fig. 7. Time history of the reaction forces at the hinge point (H ¼ 2 m, ω ¼ 1:0 rad=s).
6
S. Heo and W. Koo Ocean Engineering 200 (2020) 107070
105 105
3 3
2.5 2.5
2 2
Force [N]
Force [N]
1.5 1.5
1 1
0.5 0.5
0 0
0 0.5 1 1.5 2 0 0.5 1 1.5 2
Wave frequency [rad/s] Wave frequency [rad/s]
Fig. 9. Comparison of the spectrum of the reaction force magnitude between calculation methods (H ¼ 2 m).
0.3
0.25
0.2
RAO [rad/m]
0.15
0.1
0.05
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Wave frequency [rad/s]
Fig. 11. Comparison of the pitch RAOs between floats 1 and 2.
7
S. Heo and W. Koo Ocean Engineering 200 (2020) 107070
106
6
-2
-4
-6
110 110.5 111 111.5 112 112.5 113 113.5 114 114.5 115 115.5
Time [sec]
Fig. 12. Inverse y-axis moment acting on hinge points 1 and 2 (H ¼ 2 m, ω ¼ 1:14 rad=s).
0.3 0.5
0.2
0.25
0.1
0 0
-0.1
-0.25
-0.2
-0.3 -0.5
110 111 112 113 114 115 110 111 112 113 114 115
Time [sec] Time [sec]
Fig. 13. Angular displacement and acceleration of the floats (H ¼ 2 m, ω ¼ 1:14 rad=s).
105 105
4 4
3 3
Force [N]
Force [N]
2 2
1 1
0 0
0 0.5 1 1.5 2 0 0.5 1 1.5 2
Wave frequency [rad/s] Wave frequency [rad/s]
Fig. 14. Comparison of the spectrum of reaction force magnitude between hinge points 1 and 2 (H ¼ 2 m).
the responses when the pitch RAO of float 2 was the largest. The mo angular acceleration were different from the ratio of the pitch RAO of
ments were slightly nonlinear during a period, and the magnitude of the the float. This is because the angular acceleration has a little nonline
range of moment at hinge point 2 was 15.2% greater than that of hinge arity, and the result of the time integration also has some nonlinearity.
point 1. Because the angular acceleration of the float is directly pro Fig. 14 compares the spectrum of horizontal and vertical reaction
portional to the moment, as shown in Eq. (8), it can be predicted that the force magnitudes between hinge points 1 and 2. In contrast to the trend
angular accelerations will have the same tendency as the moments. of pitch RAO, the reaction forces acting on hinge point 1 were much
Fig. 13 presents the rotation angle and angular acceleration of the greater than those acting on hinge point 2. This means that although the
floats during a single wave period. The angular acceleration showed movement of the float is relatively small, the force acting on the hinge
slight nonlinearity similar to the tendency of the moment, as expected. point can be large. In addition, the magnitudes of the vertical reaction
In addition, the magnitude of the range of angular acceleration at hinge force were greater than the horizontal reaction force in both floats. This
point 2 was 15.5% greater than that of hinge point 1, which is similar to is because the linear buoyancy was included in the vertical external
the ratio of the moment. The magnitude ratios of the moment and the force. These horizontal and vertical reaction forces should be considered
8
S. Heo and W. Koo Ocean Engineering 200 (2020) 107070
as external forces acting on the hinge point when analyzing the behavior publication of this paper.
of the WEC platform.
The above results showed that even if the structural properties of the Acknowledgements
floats were the same, the responses can be affected not only by the di
rection of forces, but also their arrangements. Especially, the tendency This work was supported by INHA UNIVERSITY Research Grant
and magnitude of the reaction forces acting on the hinge points would
not be easily predictable without performing the reaction force analysis. References
In other words, both the motion of the float and the reaction force acting
on the hinge point must be evaluated to satisfy the structural stability of Armesto, J.A., Guanche, R., Del Jesus, F., Iturrioz, A., Losada, I.J., 2015. Comparative
analysis of the methods to compute the radiation term in Cummins’ equation.
the HPA-type WEC platform. J. Ocean Eng. Mar. Energy 1 (4), 377–393. https://doi.org/10.1007/s40722-015-
0027-1.
4. Conclusions Babarit, A., Hals, J., Muliawan, M., Kurniawan, A., Moan, T., Krokstad, J., 2012.
Numerical benchmarking study of a selection of wave energy converters. Renew.
Energy 41, 44–63. https://doi.org/10.1016/j.renene.2011.10.002.
This paper presented two numerical procedures to evaluate both the Chao, Z., Yage, Y., Aiju, C., 2018. Hydrodynamics research of a two-body articulated
dynamic responses of the float and the reaction forces acting on the wave energy device. Ocean Eng. 148, 202–210. https://doi.org/10.1016/j.
oceaneng.2017.11.029.
hinge point for HPA-type WEC platform. One is the classical method Cl�ement, A., McCullen, P., Falc~ ao, A., Fiorentino, A., Gardner, F., Hammarlund, K.,
using Newton’s second law for rotation, and the other is the augmented Lemonis, G., Lewis, T., Nielsen, K., Petroncini, S., Pontes, M.T., Schild, P.,
formulation, which is one of multibody dynamics formulation. The Sj€
ostr€om, B., Sørensen, H.C., Thorpe, T., 2002. Wave energy in europe: current status
and perspectives. Renew. Sustain. Energy Rev. 6 (5), 405–431. https://doi.org/
characteristics of the two procedures were explained in detail. Two
10.1016/S1364-0321(02)00009-6.
procedures give almost the same results because both methods deal with Cummins, W.E., 1962. The impulse response function and ship motions. In:
the same physical phenomena. An important part of the proposed pro Hydromechanics Laboratory Research and Development Report of Department of the
cedures is the calculation of the reaction forces acting on the hinge Navy David Taylor Model Basin, vol. 9, pp. 101–109.
Drew, B., Plummer, A.R., Sahinkaya, M.N., 2009. A review of wave energy converter
points of the WEC platform. From this perspective, the calculation technology. Proc. IME J. Power Energy 223 (8), 887–902. https://doi.org/10.1243/
process was explained and the results were compared. 09576509JPE782.
In the case of the HPA-type WEC platform, multiple floats were European Ocean Energy Association, 2010. Oceans of Energy - European Ocean Energy
Roadmap 2010-2050. EU-OEA, Brussels, Belgium.
connected to the main structure by structural arms and hinges. Each Falc~
ao, A., 2010. Wave energy utilization: a review of the technologies. Renew. Sustain.
numerical model for a float was modeled as a hemispherical shape and Energy Rev. 14 (3), 899–918. https://doi.org/10.1016/j.rser.2009.11.003.
connected to the hinge point by a massless rigid arm. The frequency- Flavi�a, F.F., Babarit, A., Cl�
ement, A.H., 2017. On the numerical modeling and
optimization of a bottom-referenced heave-buoy array of wave energy converters.
dependent hydrodynamic characteristics and wave excitation forces Int. J. Mar. Energy 19, 1–15. https://doi.org/10.1016/j.ijome.2017.05.004.
acting on the floats were calculated using the three-dimensional fre Karimirad, M., 2014. Offshore Energy Structures: for Wind Power, Wave Energy and
quency domain BEM solver. The movement of floats was described using Hybrid Marine Platforms. Springer.
Kim, S.J., Koo, W., Min, E.H., Jang, H., Youn, D., Lee, B., 2016. Experimental study on
Cummins’ equation in the time domain. hydrodynamic performance and wave power takeoff for heaving wave energy
The validity of the calculation methods was verified by comparing converter. J. Ocean Eng. Technol. 30 (5), 361–366. https://doi.org/10.5574/
the pitch RAO of a single float with the reference paper. The calculation KSOE.2016.30.5.361.
Kim, S.S., Lee, J.C., Kang, D., Lee, S.S., 2019. Motion characteristics of a floating wave
results using both methods agreed very well with each other over the
energy converter with wave activating body type. Int. J. Naval Arch. Ocean Eng. 11
given range of frequency. This means that both procedures well (1), 244–255. https://doi.org/10.1016/j.ijnaoe.2018.04.005.
analyzed the responses of the float and its effects. However, in terms of Kramer, M., Marquis, L., Frigaard, P., 2011. Performance evaluation of the wavestar
efficiency of numerical analysis preparation stage, the classical method prototype. In: Proceedings of the 9th European Wave and Tidal Energy Conference
(Southampton, UK).
is very inefficient because it requires much more complicated process Lee, C.H., 1995. WAMIT Theory Manual. Report No. 95-2. Department of Ocean
according to the model of the platform changes. Therefore, building the Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts.
numerical process using the augmented formulation is more efficient in Newmark, N.M., 1959. A method of computation for structural dynamics. J. Eng. Mech.
Div. 85 (3), 67–94.
terms of efficiency and applicability. In the case of the two floats, the Ogilvie, T., 1964. Recent progress towards the understanding and prediction of ship
results of each float were different due to the direction of force and the motions. In: Fifth Symposium on Naval Hydrodynamics, vol. 1, pp. 2–5.
position of the floats. Especially, the tendency and magnitude of the Penalba, M., Giorgi, G., Ringwood, J.V., 2017. Mathematical modelling of wave energy
converters: a review of nonlinear approaches. Renew. Sustain. Energy Rev. 78,
reaction forces would not be easily predictable without conducting the 1188–1207. https://doi.org/10.1016/j.rser.2016.11.137.
reaction force analysis. This means that it is essential to confirm the Shabana, A.A., Hussein, B.A., 2009. A two-loop sparse matrix numerical integration
effects of float motion on the main structure when performing stability procedure for the solution of differential/algebraic equations: application to
multibody systems. J. Sound Vib. 327 (3–5), 557–563. https://doi.org/10.1016/j.
analysis of the HPA-type WEC platform. These effects can be considered jsv.2009.06.020.
when applying the methods presented in this study. In the future, the Shabana, A.A., 2013. Dynamics of Multibody Systems, fourth ed. Cambridge university
structural analysis of the HPA-type WEC platform will be carried out press.
Wang, L., Kolios, A., Cui, L., Sheng, Q., 2018. Flexible multibody dynamics modelling of
considering the various environmental conditions (such as oblique
point-absorber wave energy converters. Renew. Energy 127, 790–801. https://doi.
waves) and the flexible arm which is composed by beam elements. org/10.1016/j.renene.2018.05.029.
Zhang, L., Zhang, D., 2016. A two-loop procedure based on implicit Runge–Kutta method
Author contribution section for index-3 dae of constrained dynamic problems. Nonlinear Dynam. 85 (1),
263–280. https://doi.org/10.1007/s11071-016-2682-8.
Zurkinden, A.S., Ferri, F., Beatty, S., Kofoed, J.P., Kramer, M., 2014. Non-linear
Mr. Heo and Prof. Koo developed the numerical procedures and numerical modeling and experimental testing of a point absorber wave energy
carried out all numerical calculations and wrote the manuscript. converter. Ocean Eng. 78, 11–21. https://doi.org/10.1016/j.oceaneng.2013.12.009.