Proceedings of the ASME 2021

International Design Engineering Technical Conferences and

Computers and Information in Engineering Conference
IDETC/CIE 2021
August 17 – 20, 2021, Virtual Conference

DETC2021-69977

NUMERICAL ANALYSIS AND PARAMETER OPTIMIZATION OF A PORTABLE TWO-BODY

ATTENUATOR WAVE ENERGY CONVERTER

Joseph Capper, Jia Mi, Qiaofeng Li, Lei Zuo*

Department of Mechanical Engineering, Virginia Tech, Blacksburg, VA, 24061, USA (leizuo@vt.edu)

ABSTRACT and those percentages are expected to increase as cities grow.

Easily portable, small-sized ocean wave energy converters Globally, usable wave energy has been estimated at over 2 TW
(WECs) may be used in many situations where large-sized WEC [3]. In the United States alone, wave energy could provide 2,640
devices are not necessary or practical. Power maximization for TWh/yr, which is about 2/3 of the electricity annually consumed
small-sized WECs amplifies challenges that are not as difficult in the U.S. [4]. Today, there is enormous diversity in the different
with large-sized devices, especially tuning the device’s natural types of wave energy converters (WECs), with no convergence
frequency to match the wave frequency and achieve resonance. on a widely accepted design as in wind energy. The three main
In this study, power maximization is performed for a small-sized, categories of WECs include oscillating water column,
two-body attenuator WEC with a footprint constraint of about overtopping devices, and oscillating bodies.
1m. A thin, submerged tuning plate is added to each body to Oscillating body WECs contain at least one moving body
increase added mass without significantly increasing hydrostatic that is driven by the wave motion. The oscillating motion is
stiffness in order to reach resonance. Three different body cross- converted into useful power using a power takeoff (PTO)
section geometries are analyzed. Device power absorption is system, which is typically either a hydraulic system or electric
determined through time domain simulations using WEC-Sim generator [5]. Oscillating body WECs generally fall into three
with a simplified two-degree-of-freedom (2DOF) model and a classes: heaving, surging, or pitching. Heaving devices, or point
more realistic three-degree-of-freedom (3DOF) model. Different absorbers, use bodies that float on the surface of the waves and
drag coefficients are used for each geometry to explore the effect contain a translational PTO system that is activated as the ocean
of drag. A mooring stiffness study is performed with the 3DOF waves cause movement in the vertical direction. Surging WECs,
model to investigate the mooring impact. Based on the 2DOF either bottom hinged or floating, take advantage of the wave
and 3DOF power results, there is not a significant difference in motion in the horizontal direction. Pitching devices, on the other
power between the shapes if the same drag coefficient is used, hand, use relative rotation between two or more bodies to
but the elliptical shape has the highest power after assigning a produce power [6].
different approximate drag coefficient to each shape. The Pitching devices are further classified as either bottom-
mooring stiffness study shows that mooring stiffness can be hinged or floating, also known as attenuators. Attenuators
increased in order to increase relative motion between the two contain a series of sections linked with flexible joints that allow
bodies and consequently increase the power. the sections to pitch relative to each other. The device is moored
to the seafloor and placed along with the wave direction so that
Keywords: Wave Energy Converter (WEC), Attenuator, there is oscillation with the wave amplitude [7]. A number of
Geometry Optimization, Resonance Tuning, Power different attenuators have been developed and tested, but much
Maximization of the work has focused on large-scale power generation,
including the McCabe Wave Pump, 750-1000 kW Pelamis, 500-
1. INTRODUCTION 6000 kW Wave Star, 375 kW Salter Duck, and 1000 kW
Ocean wave energy shows major promise as a clean, Anaconda [8]. Analysis of large-scale attenuators and floating
abundant, and renewable source of energy with a strategic two-body devices has been performed in literature such as Ref.
location in close proximity to global population centers. [9], [10], [11], [12], [13], and [14].
Presently, around 60% of the world’s population lives in coastal Increasingly, there is also interest in wave energy
regions [1], including 40% living within 20 km of the coast [2], conversion at the small scale in addition to large scale

1 © 2021 by ASME
applications. There are numerous low-energy marine devices Wave Propagation
that could potentially be powered by ocean wave energy, such as Mean Water Surface
navigation signs, illumination, sensors, survival kits, and
electronics charging [15][16]. The U.S. Department of Energy Power Takeoff Hinge
(DOE) highlights the importance of small-scale ocean wave
energy in the Powering the Blue Economy report [16]. The DOE Mooring modeled
has sponsored development of small-sized wave energy devices Seabed
as a linear spring
through design competitions including the Waves to Water
competition focusing on wave-powered desalination and the FIGURE 1: CONCEPT DIAGRAM FOR TWO-BODY
Ocean Observing competition focusing on wave-powered ATTENUATOR
devices for monitoring ocean conditions [17][18].
Power maximization and natural frequency tuning for a
small-sized attenuator are more challenging than for large-sized
devices. Many wave energy converters target a wave period of
5-12 seconds based on real ocean conditions [9][12][13][14], but
it is difficult to achieve a resonant period high enough to be in
that range with small-sized devices. Large-sized devices can use
large overall dimensions and physical or added mass to adjust Folded
the natural frequency and increase the power [9][12][13][14].
However, using such large dimensions and mass is not possible Tuning Body 1 Body 2
with small-sized attenuators because of portability constraints. Plate 1
An innovative method for frequency tuning of small-sized
attenuators is presented in this study through use of a thin tuning
plate attached to each body. PTO Hinge
The objective of this study is to develop a method of power Tuning
maximization for a small-sized, two-body attenuator through Plate 2
resonance tuning and numerical simulation. The study first
optimizes the geometry of the wave energy converter with a
frequency domain method to match the natural frequency of the
system with that of the incoming wave. The hydrodynamic
coefficients are calculated with ANSYS AQWA. The study then
adopts time domain simulation, using open-source tool WEC- Unfolded
Sim to optimize the PTO damping coefficient and power
generation. The time domain simulation enables us to more FIGURE 2: POSSIBLE SYSTEM DISASSEMBLY FOR
realistically consider the influences of drag damping, various TRANSPORT
cross-section shapes, and mooring stiffness. Conclusions are
drawn in Section 4. 2.2 System Governing Equations
In order to tune the two-body device to achieve resonance
2. MATERIALS AND METHODS by matching the device natural frequency to the wave frequency,
2.1 Design and Working Principle the equations of motion of the system were first derived. The
In this study, the device under consideration is a small-sized, equations defined in this section (Section 2.2) were used for
easily portable, two-body attenuator with a power takeoff around dimensional optimization through natural frequency tuning
the hinge between the two bodies, as shown in Figure 1. The size while the equations and models shown in Section 2.4 were used
is such that the device could be easily transported in the back of to determine the time domain results. The two-degree-of-
a car or other vehicle and carried by one or two people to be freedom (2DOF) system simplifies the device so that each body
deployed in the ocean. The two main bodies could be has one DOF in rotation around the hinge point (fixed), resulting
disconnected and the thin plates folded to ease transportation, as in a 2DOF overall system, as illustrated in Figure 3. The 3DOF
shown in Figure 2. Without the tuning plate, the length of each system allows the hinge to move in the heave direction with a
main body is 0.5 m, so the overall length of the device without mooring stiffness, as shown in Figure 4.
the plates will be about 1 m.

2 © 2021 by ASME
 Mean water surface
Buoyancy center Gravity center 𝑅𝑚 = 𝑚𝑔𝑙𝑐𝑜𝑠𝜃 − 𝜌𝑔𝑉𝜃 𝑙𝑐𝑜𝑠𝜃
= 𝑚𝑔𝑙𝑐𝑜𝑠𝜃 − 𝜌𝑔(𝑉0 + 𝑘1 𝜃 + 𝑘2 𝜃 2 + ⋯ )𝑙𝑐𝑜𝑠𝜃 (4)
By taking the derivative of Equation 4 with respect to 𝜃, we
l
have Equation 5, i.e., the stiffness equals to the rate of the change
Submerged volume Hinge (fixed)
of moment 𝑅𝑚 at 𝜃 = 0. It is shown that the individual body
FIGURE 3: 2DOF MOTION SCHEMATIC OF THE stiffness is a function of the body’s rate of change in volume with
ATTENUATOR WEC
respect to 𝜃 , as shown in Equation 6. By calculating 𝑘1
numerically, for example with a computer-aided design
Mean water surface
Buoyancy center Gravity center Z software, we are able to determine the rotational hydrostatic
stiffness of the two bodies of an arbitrary shape.

𝜕𝑅𝑚
l −𝑘 = |𝜃=0
hinge 𝜕𝜃
Submerged volume
Mooring Stiffness = −𝑚𝑔𝑙𝑠𝑖𝑛𝜃 − 𝜌𝑔(𝑘1 + 2𝑘2 𝜃 + ⋯ )𝑙𝑐𝑜𝑠𝜃 + 𝜌𝑔(𝑉0 +
𝑘1 𝜃 + 𝑘2 𝜃 2 + ⋯ )𝑙𝑠𝑖𝑛𝜃|𝜃=0 = −𝜌𝑔𝑘1 𝑙 (5)
𝜕𝑉𝜃
FIGURE 4:
3DOF MOTION SCHEMATIC OF THE 𝑘1 = |
𝜕𝜃 𝜃=0
(6)
ATTENUATOR WEC
Equation 7 gives the undamped natural frequency for one of
Taking the sum of the moments for the system around the
the bodies. In the case of the attenuator, as a hydrodynamic
hinge point, equations are derived that are applicable to both the
system, the spring stiffness can be replaced by the body’s
2DOF and 3DOF systems. The overall equation of motion for
hydrostatic stiffness and the mass becomes the sum of the inertia
the system is given by Equation 1, where J is the rotational
due to the body’s dry material JM and inertia due to added mass
inertia matrix with respect to the PTO hinge, A(ω) is the added
when submerged JA. The natural period can be calculated using
rotational inertia with respect to the PTO at frequency ω, C(ω)
Equation 8.
is the PTO damping matrix, K is the hydrostatic stiffness matrix,
M is the excitation moment vector, and 𝜃, 𝜃̇ , and 𝜃̈ are the 𝜔𝑛 = √𝐽
𝑘
(7)
angular displacement, velocity, and acceleration respectively. 𝑀 +𝐽𝐴
The expanded matrices are shown in Equation 2. 𝑇𝑛 = 𝜔
2𝜋
(8)
𝑛

(𝐽 + 𝐴(𝜔))𝜃̈ + 𝐶(𝜔)𝜃̇ + 𝐾𝜃 = 𝑀 (1)

𝐽1 0 𝐴1 (𝜔) 0 2.3 Frequency Domain Natural Frequency Tuning
𝐽=[ ], 𝐴(𝜔) = [ ], In order to maximize power absorption and achieve
0 𝐽2 0 𝐴2 ( 𝜔 ) resonance, the natural frequency of each body was tuned to
𝑐 −𝑐 𝑘 0 match the wave frequency. The target wave period was 6 seconds
𝐶 (𝜔 ) = [ ], 𝐾 = [ ] (2)
−𝑐 𝑐 0 𝑘 based on common wave conditions on the US east coast [17].
The dry material inertia JM was obtained from the CAD model.
We determine the hydrostatic stiffness of the two bodies The boundary element method (BEM) solver ANSYS AQWA
with the following procedure. At hydrostatic equilibrium, the was used to perform a hydrodynamic diffraction solution in the
vertical summation of the forces gives Equation 3, where m is frequency domain to obtain JA and k in order to use Equation 7
the body mass, g is acceleration due to gravity, ρ is the water [19]. Because of the portability constraint, it was necessary to
density, and V0 is the submerged volume with zero flap rotation. keep the overall body dimensions approximately close to the
target overall system dimensions shown in Figure 5.
𝑚𝑔 = 𝜌𝑔𝑉0 (3)

By taking the sum of the moments for the individual body

around the hinge point, we obtain Equation 4 where Rm is the
hydrostatic moment on one body with respect to the PTO hinge
and Vθ is the submerged volume. Note that Vθ is expanded around
𝜃 = 0 using a Taylor series where 𝑘1 , 𝑘2 , … are the coefficients
of the Taylor series.

3 © 2021 by ASME
 After tuning the device using traditional geometry proved
inadequate, a new geometry feature was added to make tuning
0.5 m possible. By adding a thin, submerged flat plate to the existing
shape, it was possible to greatly increase the added mass without
significantly increasing the hydrostatic stiffness. The body
dimensions could be kept at the target values and the plate length
varied to tune the body to the wave frequency, as shown in
Figure 6. Using the tuning plate, the natural periods for the three
MWS different body shapes were tuned to approximately 6 s, as shown
0.2 m Draft in Table 2.
0.5 m 0.1 m

FIGURE 5: DESIRED OVERALL DIMENSION CONSTRAINTS

FOR TWO-BODY ATTENUATOR

Table 1(a) shows a sample of geometry parameters and

results for the baseline shape, a rectangular flat plate. As shown
in Table 1(b), it is difficult to achieve a natural period much MWS
larger than 1 second because the hydrostatic stiffness k remained
large relative to JM and JA. An intuitive modification is to FIGURE 6: TWO-BODY ATTENUATOR WITH THIN TUNING
increase the buoy length in order to achieve an increased added PLATE
mass. However, as shown in Case 7 in Table 1, JM and JA were
greatly increased from the initial case when increasing the buoy TABLE 2: TUNED NATURAL FREQUENCY FOR EACH SHAPE
length, but hydrostatic stiffness also increased significantly, Thin Body
preventing the natural frequency from matching the targeted Plate Density K (N- J_M J_A wn
wave frequency. Length (kg/ m (kg* (kg.m²/ (rad/ Tn
Shape (m) m^3) /rad) m^2) rad) s) (s)
TABLE 1(a): SAMPLE OF GEOMETRY VARIATION AND Rectangular 0.90 606 338 2.52 299 1.06 5.93
NATURAL FREQUENCY RESULTS FOR BASELINE SHAPE Elliptical 0.95 627 342 1.78 320 1.03 6.09
Variable Parameters Quadrilateral 0.95 675 346 2.29 321 1.04 6.06
Buoy Lever
Draft Thickness Width Length Length Three different body cross sections were used for the
Case # (m) (m) (m) (m) (m) analysis, as shown in Figure 7: rectangular, elliptical, and
1 (initial) 0.15 0.2 0.50 0.5 0.075 quadrilateral. The shapes were chosen based on commonly
2 0.10 0.2 0.50 0.5 0.075 studied cross sections found in literature such as Ref. [9] and
3 0.25 0.3 0.50 0.5 0.075 [10]. The overall body dimensions were kept the same as those
4 0.15 0.2 0.75 0.5 0.075 shown in Figure 5. A draft of 0.1 m was used for each body so
5 0.15 0.2 0.50 0.5 0.150 that the mean water surface was in the center of the cross section.
6 0.15 0.2 0.50 0.5 0.500 Reducing the draft would increase portability by reducing the
7 0.15 0.2 0.50 1.0 0.075 body weight because as Equation 3 shows, body weight is
8 0.45 0.5 0.50 0.5 0.075 proportional to displaced volume. However, for tapered shapes
like the elliptical and quadrilateral, reducing the draft would also
TABLE 1(b): SAMPLE OF GEOMETRY VARIATION AND reduce the submerged body width. Excitation force decreases as
NATURAL FREQUENCY RESULTS FOR BASELINE SHAPE submerged body width decreases because less of the body is
Results exposed to the incoming wave front, leading to a reduction in
K (N- J_M J_A wn Tn power absorption. Consequently, the study is a fairer comparison
Case # m/rad) (kg*m^2) (kg.m²/rad) (rad/s) (s) if the submerged widths are same for the three geometries.
1
308.36 1.11 6.27 6.64 0.97
(initial)
2 305.34 0.74 6.39 6.54 0.96
3 302.08 2.00 6.10 6.11 1.03
4 462.54 1.81 11.35 5.93 1.06
5 445.00 1.15 9.24 6.55 0.96
6 1456.47 6.61 32.11 6.13 1.02
7 2061.67 7.19 56.73 5.68 1.11
8 289.51 4.93 5.35 5.31 1.18

4 © 2021 by ASME
 wave spectrum. Average power was calculated from the
Rectangular Draft=0.1m instantaneous PTO power after the simulation reached steady
MWS state. A wave period of T=6 s and wave height of H=0.5 m was
Elliptical used for each case. Assuming that the two-body attenuator would
be deployed close to shore, a water depth of 5 m was used. In
order to study the effects of drag, four different drag coefficients
Quadrilateral of Cd=0.5, 1, 1.5, 2 were used in calculation 𝐹𝑣 . These four drag
coefficients were determined to be in the reasonable range for
FIGURE 7: BODY CROSS SECTIONS USED FOR ANALYSIS the WEC type and body shapes based on literature such as Ref.
[21] and [22].
2.4 Time Domain Power Absorption Simulation Rigid Body Blocks
Power absorption modelling was performed using WEC-
Sim, an open-source code created in MATLAB/Simulink by the
National Renewable Energy Lab (NREL) for wave energy
converter simulation [11]. A time-domain numerical model was
created in WEC-Sim to solve the system dynamics. Initially, the
simplified 2DOF model was used followed by the more realistic
3DOF model. In the WEC-Sim software, the dynamic response Rotational constraint Rotational PTO
is calculated by solving the equation of motion for each body restricts motion to captures power
around its center of gravity using Equation 9, where M is the rotation around a fixed from relative
mass matrix, 𝑋̈ is the (translational and rotational) acceleration hinge point motion
vector of the body, Fexc(t) is the wave excitation force and torque
vector, Frad(t) is the force and torque vector resulting from wave
radiation, Fpto(t) is the PTO force and torque vector, Fv(t) is the
damping force and torque vector, FB(t) is the net buoyancy FIGURE 8: SIMULINK MODEL OF THE 2DOF SYSTEM
restoring force and torque vector, and Fm(t) is the force and
torque vector resulting from mooring connection [20]. Equation To obtain the max power and optimal damping for each
9 was used to determine the time-domain results while the cross section and Cd, WEC-Sim’s multiple condition run (MCR)
equations in Section 2.2 were only used for dimensional feature was used to run a series of cases with different rotational
optimization. damping values. For each cross section, each Cd was simulated
with rotational damping from 0-500 Nsm/rad with a step of 10
Nsm/rad, or in other words, 50 cases for each Cd.
𝑀𝑋̈ = 𝐹𝑒𝑥𝑐 (𝑡) + 𝐹𝑟𝑎𝑑 (𝑡) + 𝐹𝑝𝑡𝑜 (𝑡) + Wave energy flux, Jp, is the mean power per wave front
𝐹𝑣 (𝑡) + 𝐹𝐵 (𝑡) + 𝐹𝑚 (𝑡) width available in the wave. It can be calculated using Equation
(9) 11, where ρ is the water density, g is acceleration due to gravity,
T is wave height, and H is wave period [23]. Capture width (CW)
Figure 8 shows the Simulink model of the 2-body attenuator is the amount of wave crest width that is completely captured and
for the 2DOF model. The two rigid bodies are constrained to absorbed by the WEC. It is calculated according to Equation 12
rotate around a fixed center point using a rotational constraint. by taking the ratio of average absorbed wave power Pavg to wave
Heaving motion is not included in the 2DOF model, as if the energy flux J. As a non-dimensional way to quantify the
center point is attached to a fixed structure. The rotational PTO hydrodynamic efficiency, the capture width ratio CWR was
captures power based on the relative rotational velocity between calculated using Equation 13 by dividing the capture width (CW)
the two bodies according to Equation 10 where c is the PTO by the device characteristic dimension L [24]. In the case of the
rotational damping and 𝜃̇𝑟𝑒𝑙 is the relative rotational velocity. two-body attenuator, the characteristic dimension is the device
PTO damping is assumed to be constant with respect to the submerged width of 0.5 m, which is perpendicular to the
relative rotational velocity between the two bodies. incoming wave front.

𝜌𝑔2 𝑇𝐻 2
𝑃 = 𝑐𝜃̇𝑟𝑒𝑙
2
(10) 𝐽𝑝 = (11)
64𝜋

In each 2DOF simulation case, there was a ramp time of 150 𝐶𝑊 = 𝑃𝑎𝑣𝑔 /𝐽𝑝 (12)
s and time step of 0.05 s. The solver ode4 was used with the fixed
time step. Regular wave simulations used a duration of 500 s 𝐶𝑊𝑅 = 𝐶𝑊/𝐿 (13)
while irregular wave simulations used a duration of 750 s to
provide additional simulation data for taking the average power After simulations were performed with the 2DOF model, a
because power fluctuates more under irregular waves due to the 3DOF model that accounts for the WEC heave motion was used.

5 © 2021 by ASME
The 3DOF model is shown in Figure 9. To allow the device to optimal damping, maximum power, and capture width ratio. Out
heave up and down, a non-hydro body was connected to the of all the shapes and drag coefficients, maximum power occurred
seabed using a translational constraint. The non-hydro body has with rotational damping of 80-160 Nsm/rad. The rectangular
no hydrodynamic interaction in the model and is only used to shape with Cd 0.5 had the largest max power of the different
enable the heave motion. Two rotational PTOs are connected shapes and drag coefficients: 37.33 W with CWR of 10.15% for
between the side bodies and non-hydro body, allowing the side regular waves and 11.03 W with 3.00% CWR for irregular
bodies to pitch around the non-hydro body. For each case, the waves. There is not a significant difference in power absorption
two PTO damping values were set equal. To simulate a single between the shapes for a given drag coefficient. The small
mooring line attached to the WEC center point, a mooring difference is logical because submerged body widths are equal
stiffness acting in the heave direction was added to the non-hydro for each shape, so for the same drag coefficient, excitation forces
body. Power absorption was calculated using Equation 10 based and power absorbed are expected to be similar.
on the relative velocity of the two side bodies.
Rigid Body Blocks TABLE 3: 2DOF POWER RESULTS SUMMARY
Rectangular
Rotational Regular Wave Irregular Wave
C_opt P_max C_opt P_max
PTOs allow Cd (Nsm/rad) (W) CWR (Nsm/rad) (W) CWR
rigid bodies 0.5 80 37.33 10.15% 80 11.03 3.00%
to pitch 1.0 110 30.28 8.23% 110 9.02 2.45%
1.5 130 26.07 7.09% 130 7.89 2.15%
2.0 150 23.16 6.29% 150 7.13 1.94%

Mooring stiffness acts Non-hydro Elliptical

in heave direction rigid body Regular Wave Irregular Wave
C_opt P_max C_opt P_max
Translational Cd (Nsm/rad) (W) CWR (Nsm/rad) (W) CWR
0.5 90 33.51 9.11% 120 10.15 2.76%
constraint allows 1.0 120 27.09 7.36% 150 8.32 2.26%
heave motion 1.5 140 23.40 6.36% 170 7.29 1.98%
2.0 160 20.82 5.66% 190 6.60 1.79%
FIGURE 9: SIMULINK MODEL OF THE 3DOF SYSTEM
Quadrilateral
Using the 3DOF model, two different studies were Regular Wave Irregular Wave
performed: a drag study and mooring stiffness study. The 3DOF C_opt P_max C_opt P_max
model used the same wave period, wave height, and water depth Cd (Nsm/rad) (W) CWR (Nsm/rad) (W) CWR
as the 2DOF model. The same ramp time and simulation duration 0.5 90 32.50 8.83% 120 10.22 2.78%
1.0 120 27.12 7.37% 150 8.47 2.30%
were also used. The drag study was performed with the same four 1.5 140 23.71 6.44% 170 7.45 2.02%
drag coefficients. For the drag study, a baseline value for the 2.0 160 21.20 5.76% 190 6.75 1.83%
mooring stiffness kmooring was used by setting kmooring equal to the
combined heave hydrostatic stiffness khydrostatic of the side bodies. Average Power vs. Rotational Damping (Rectangular Average Power vs. Rotational Damping (Rectangular

The hydrostatic stiffness was obtained from AQWA, leading to

Cross Section, Regular Wave, 2DOF) Cross Section, Irregular Wave, 2DOF)
40 12

a baseline mooring stiffness of 5277.2 N/m, the same for all three 35 Cd 0.5 Cd 1 Cd 1.5 Cd 2
10

cross section geometries. Because of the increased system

Average Power (W)
Average Power (W)

30
8
25

stiffness due to the mooring, a smaller time step of 0.025 s was 20 6

necessary. As with the 2DOF model, each drag coefficient had a

15 4
10

rotational damping range of 0-500 Nsm/rad with a time step of

2
5 Cd 0.5 Cd 1 Cd 1.5 Cd 2

10 Nsm/rad, leading to 50 cases for each drag coefficient. For

0 0
0 100 200 300 400 500 0 100 200 300 400 500

the mooring stiffness study, a series of different mooring

Rotational Damping (Nsm/rad) Rotational Damping (Nsm/rad)

(a) (b)
stiffness values with respect to the hydrostatic stiffness were
FIGURE 10: 2DOF POWER RESULTS, RECTANGULAR CROSS
used. Regular waves were used for the mooring stiffness study. SECTION
Like the drag study, rotational damping was varied from 0-500
Nsm/rad with a step of 10 Nsm/rad.

3. RESULTS AND DISCUSSION

3.1 Two DOF Results
Figures 10, 11, and 12 show the 2DOF regular and irregular
wave power absorption results under different drag coefficients
for the three shapes. Table 3 summarizes the 2DOF results for

6 © 2021 by ASME
 Average Power vs. Rotational Damping (Elliptical
Cross Section, Regular Wave, 2DOF)
Average Power vs. Rotational Damping (Elliptical
Cross Section, Irregular Wave, 2DOF) One possible comparison is to choose an approximate drag
40 12 coefficient for each shape out of the four Cd values used for the
study, such as Cd=1 for the elliptical shape, Cd=1.5 for the
35 Cd 0.5 Cd 1 Cd 1.5 Cd 2
10

Average Power (W)

Average Power (W)

30
25
8
quadrilateral, and Cd=2 for the rectangular [22]. Figure 15 shows
20
15
6
this comparison. Using approximate drag coefficients, the
elliptical shape would have the highest power absorption of
4
10

27.09 W with 7.36% CWR at 120 Nsm/rad optimal damping for

2
5 Cd 0.5 Cd 1 Cd 1.5 Cd 2
0
regular waves and 8.32 W with 2.26% CWR at 150 Nsm/rad
0
0 100 200 300 400 500 0 100 200 300 400 500

optimal damping for irregular waves.

Rotational Damping (Nsm/rad) Rotational Damping (Nsm/rad)

(a) (b)
FIGURE 11: 2DOF POWER RESULTS, ELLIPTICAL CROSS Average Power vs. Rotational Damping (3 Cross
SECTION Sections, Regular Wave, 2DOF)
30 Elliptical, Cd 1
Average Power vs. Rotational Damping Average Power vs. Rotational Damping
(Quadrilateral Cross Section, Regular Wave, 2DOF) (Quadrilateral Cross Section, Irregular Wave, 2DOF) 25 Quadrilateral, Cd 1.5

Average Power (W)

40 12 Rectangular, Cd 2
20
35 Cd 0.5 Cd 1 Cd 1.5 Cd 2
10
Average Power (W)
Average Power (W)

30 15
8
25
20 6 10
15
4 5
10
2
5 Cd 0.5 Cd 1 Cd 1.5 Cd 2 0
0 0 0 100 200 300 400 500
(a)
0 100 200 300 400 500 0 100 200 300 400 500
Rotational Damping (Nsm/rad)
Rotational Damping (Nsm/rad) Rotational Damping (Nsm/rad)
Average Power vs. Rotational Damping (3 Cross
(a) (b) Sections, Irregular Wave, 2DOF)
FIGURE 12: 2DOF POWER RESULTS, QUADRILATERAL 9
CROSS SECTION 8
7

Average Power (W)

When rotational damping becomes large, the two bodies 6

tend to rotate together as one, causing small relative motion and

5
4
therefore small power extracted by the PTO. Figures 13 and 14 3
Elliptical, Cd 1

illustrate this effect. The rotational velocity is out of phase for

Quadrilateral, Cd 1.5
2

smaller damping values such as 80 Nsm/rad and nearly in phase 1 Rectangular, Cd 2

for large damping values like 1000 Nsm/rad.

0
0 100 200 300 400 500

(b) Rotational Damping (Nsm/rad)

FIGURE 15: 2DOF POWER RESULTS COMPARISON WITH

APPROXIMATE DRAG COEFFICIENTS

3.2 Three DOF Drag Study Results

Figures 16, 17, and 18 show the 3DOF drag study results.
Table 4 summarizes the 3DOF results for optimal damping,
maximum power, and capture width ratio. The optimal damping
ranged from 160-380 Nsm/rad. Like the 2DOF study, the
FIGURE 13: ROTATIONAL VELOCITY, RECTANGULAR rectangular shape with Cd 0.5 had the highest max power out of
CROSS SECTION, 80 Nsm/rad DAMPING, 0.5 Cd, REGULAR the different shapes and drag coefficients: 33.04 W with 8.98%
WAVE
CWR for regular waves and 8.51 W with 2.31% CWR for
irregular waves. As with the 2DOF study, the difference in power
between the shapes is small for the same drag coefficient, which
is expected because the three shapes have the same submerged
body width.

TABLE 4: 3DOF DRAG STUDY POWER RESULTS SUMMARY

Rectangular
Regular Wave Irregular Wave
FIGURE 14: ROTATIONAL VELOCITY, RECTANGULAR C_opt C_opt
(Nsm/ P_max (Nsm/ P_max
CROSS SECTION, 1000 Nsm/rad DAMPING, 0.5 Cd, REGULAR
Cd rad) (W) CWR rad) (W) CWR
WAVE
0.5 160 33.04 8.98% 130 8.51 2.31%
1.0 200 25.61 6.96% 200 6.82 1.85%
In reality, tapered cross section geometries like the elliptical 1.5 230 21.48 5.84% 260 6.00 1.63%
and quadrilateral shapes would have lower drag coefficients. 2.0 260 18.81 5.11% 320 5.50 1.50%

7 © 2021 by ASME
 As with the 2DOF study, an approximate drag coefficient
Elliptical can be used to compare the 3DOF drag study results, as shown
Regular Wave Irregular Wave in Figure 19. Based on this comparison, the elliptical shape
C_opt C_opt would have the highest power absorption of 22.52 W with 6.12%
(Nsm/ P_max (Nsm/ P_max
Cd rad) (W) CWR rad) (W) CWR
CWR at 210 Nsm/rad optimal damping for regular waves and
0.5 180 28.65 7.79% 150 7.60 2.07% 6.18 W with 1.68% CWR at 230 Nsm/rad optimal damping for
1.0 210 22.52 6.12% 230 6.18 1.68% irregular waves. This comparison agrees with the 2DOF
1.5 250 19.00 5.16% 310 5.51 1.50% comparison with approximate drag coefficients, which also
2.0 280 16.68 4.53% 390 5.10 1.39% showed the elliptical shape to have the highest power.

Quadrilateral Average Power vs. Rotational Damping (3 Cross

Sections, Regular Wave, 3DOF, k_m=k_h)
Regular Wave Irregular Wave
25
C_opt C_opt
(Nsm/ P_max (Nsm/ P_max
20

Average Power (W)

Cd rad) (W) CWR rad) (W) CWR
0.5 180 28.59 7.77% 150 7.59 2.06% 15
1.0 220 22.46 6.10% 240 6.16 1.67%
1.5 260 18.94 5.15% 310 5.48 1.49% 10
Rectangular, Cd 2
2.0 290 16.62 4.52% 380 5.06 1.37%
5 Quadrilateral, Cd 1.5

Average Power vs. Rotational Damping (Rectangular Average Power vs. Rotational Damping (Rectangular Elliptical, Cd 1
Cross Section, Regular Wave, 3DOF, k_m=k_h) Cross Section, Irregular Wave, 3DOF, k_m=k_h) 0
35 9 0 100 200 300 400 500
30
8
7
(a) Rotational Damping (Nsm/rad)
Average Power (W)
Average Power (W)

25
6 Average Power vs. Rotational Damping (3 Cross
20 5
4
Sections, Irregular Wave, 3DOF, k_m=k_h)
15
3 7
10
2
5 1 Cd 0.5 Cd 1 Cd 1.5 Cd 2 6
Cd 0.5 Cd 1 Cd 1.5 Cd 2
0 0
Average Power (W)
0 100 200 300 400 500 0 100 200 300 400 500 5
Rotational Damping (Nsm/rad) Rotational Damping (Nsm/rad)

(a) (b)
4

FIGURE 16: 3DOF POWER RESULTS, RECTANGULAR CROSS 3

Rectangular, Cd 2

SECTION 2 Quadrilateral, Cd 1.5

1 Elliptical, Cd 1
Average Power vs. Rotational Damping (Elliptical Average Power vs. Rotational Damping (Elliptical
Cross Section, Irregular Wave, 3DOF, k_m=k_h) 0
Cross Section, Regular Wave, 3DOF, k_m=k_h)
0 100 200 300 400 500
35 9

30
8
7
(b) Rotational Damping (Nsm/rad)

FIGURE 18: 3DOF POWER RESULTS COMPARISON WITH

Average Power (W)
Average Power (W)

25
6

APPROXIMATE DRAG COEFFICIENTS

20 5

15 4
3
10
2
5
Cd 0.5 Cd 1 Cd 1.5 Cd 2 1 Cd 0.5 Cd 1 Cd 1.5 Cd 2 3.3 Three DOF Mooring Study Results
The 3DOF mooring stiffness results for the three different
0 0
0 100 200 300 400 500 0 100 200 300 400 500

shapes are shown in Figure 20. Table 5 summarizes the mooring

Rotational Damping (Nsm/rad) Rotational Damping (Nsm/rad)

(a) (b)
stiffness power results. For low mooring stiffness values like the
FIGURE 17: 3DOF POWER RESULTS, ELLIPTICAL CROSS
kmooring=khydrostatic*0.1 case, there is no significant difference in
SECTION
power between the three shapes. With higher mooring stiffnesses
Average Power vs. Rotational Damping (Quadrilateral Average Power vs. Rotational Damping (Quadrilateral such as kmooring=khydrostatic and larger, the elliptical shape had the
highest power absorption when using the approximate drag
Cross Section, Regular Wave, 3DOF, k_m=k_h) Cross Section, Irregular Wave, 3DOF, k_m=k_h)
35 9

30
8
7
coefficients. The mooring stiffness study demonstrated that
increased mooring stiffness can be used to increase relative
Average Power (W)
Average Power (W)

25 6

motion between the two bodies in order to increase the power.

20 5
4
15

10
3
2
The mooring stiffness study can be applied to different WEC
5
Cd 0.5 Cd 1 Cd 1.5 Cd 2 1
0
Cd 0.5 Cd 1 Cd 1.5 Cd 2
types, geometries, and sizes.
0
0 100 200 300 400 500 0 100 200 300 400 500
Rotational Damping (Nsm/rad) Rotational Damping (Nsm/rad)

(a) (b)
FIGURE 18: 3DOF POWER RESULTS, QUADRILATERAL
CROSS SECTION

8 © 2021 by ASME
TABLE 5: 3DOF MOORING STUDY RESULTS SUMMARY
Rectangular (Cd 2) 4. CONCLUSION
C_opt P_max
When tuning the natural frequency of the small-scale device
k_m (N/m) (Nsm/rad) (W) CWR
k_m=k_h*0.1 190 2.07 0.56%
to the wave frequency, a major challenge was the small material
k_m=k_h 260 18.81 5.11% and added inertias (or masses) relative to the hydrostatic
k_m=k_h*10 340 40.52 11.01% stiffness. To overcome this issue, we propose to use a thin,
km=k_h*100 360 44.72 12.15% submerged plate to increase added mass without greatly
increasing stiffness. Three typical cross section shapes of the
Elliptical (Cd 1) same overall dimension constraint were compared: rectangular,
C_opt P_max elliptical, and quadrilateral. When variation in drag between the
k_m (N/m) (Nsm/rad) (W) CWR three shapes was not considered, there was not a significant
k_m=k_h*0.1 180 1.97 0.54%
difference in power absorption between the shapes. If an
k_m=k_h 210 22.52 6.12%
k_m=k_h*10 270 49.07 13.34% approximate drag coefficient was assigned to each shape, then
km=k_h*100 280 54.22 14.74% the elliptical shape had the highest power. Using the approximate
drag coefficient comparison for the 2DOF model, the elliptical
Quadrilateral (Cd 1.5) shape had a max power of 27.09 W with 7.36% CWR for regular
C_opt P_max waves and a max power of 8.32 W with 2.26% CWR for irregular
k_m (N/m) (Nsm/rad) (W) CWR waves. Using the approximate drag coefficient comparison for
k_m=k_h*0.1 200 2.15 0.58% the 3DOF model, the elliptical cross section had a max power of
k_m=k_h 260 18.94 5.15%
22.52 W with 6.12% CWR for regular waves and 6.18 W with
k_m=k_h*10 340 40.64 11.05%
km=k_h*100 350 44.82 12.18% 1.68% CWR for irregular waves. The mooring stiffness study
showed that increasing mooring stiffness can be used to increase
Average Power vs. Rotational Damping (Rectangular power by increasing relative motion between the bodies. For low
Cross Section, Regular Wave, 3DOF, Cd 2) mooring stiffness values, there was not a significant difference
60
k_m=k_h*0.1 k_m=k_h
in the results between the different shapes, but the elliptical
50 k_m=k_h*10 k_m=k_h*100 shape had the highest power for larger mooring stiffness values
Average Power (W)

40 when using approximate drag coefficients.

30

20 ACKNOWLEDGEMENTS
10
The material in this paper is based on work supported by the
0
US Environmental Protection Agency under Grant No. EPA
0 100 200 300 400 500 83991001-0.
(a) Rotational Damping (Nsm/rad)
Average Power vs. Rotational Damping (Elliptical REFERENCES
Cross Section, Regular Wave, 3DOF, Cd 1)
[1] Mafi-Gholami, D., Jaafari, A., Zenner, E. K., Kamari, A.
N., & Bui, D. T. (2020). Vulnerability of coastal communities to
60

50
climate change: Thirty-year trend analysis and prospective
Average Power (W)

40
prediction for the coastal regions of the Persian Gulf and Gulf of
30 Oman. Science of the Total Environment, 741, 140305.
20 [2] Huckerby, John, Henry Jeffrey, and Brighid Jay. "An
10
k_m=k_h*0.1 k_m=k_h
International Vision for Ocean Energy (Phase 1)." (2011).
k_m=k_h*10 k_m=k_h*100
0 [3] Vosough, Amir. "Wave energy." International journal of
0 100 200 300 400 500
multidisciplinary sciences and engineering 2, no. 7 (2011): 60-
(b) Rotational Damping (Nsm/rad)
63.
Average Power vs. Rotational Damping (Quadrilateral
Cross Section, Regular Wave, 3DOF, Cd 1.5) [4] Jacobson, Paul T., George Hagerman, and George
60 Scott. Mapping and assessment of the United States ocean wave
50
k_m=k_h*0.1
k_m=k_h*10
k_m=k_h
k_m=k_h*100
energy resource. No. DOE/GO/18173-1. Electric Power
Research Institute, 2011.
Average Power (W)

[5] António, F. de O. "Modelling and control of oscillating-

40

30
body wave energy converters with hydraulic power take-off and
20 gas accumulator." Ocean engineering 34, no. 14-15 (2007):
10 2021-2032.
0 [6] Prakash, S. S., K. A. Mamun, F. R. Islam, Rishaal
0 100 200 300 400 500 Mudliar, Cherie Pau'u, Manasa Kolivuso, and Semi Cadralala.
(c) Rotational Damping (Nsm/rad)
"Wave energy converter: a review of wave energy conversion
FIGURE 20: 3DOF MOORING STIFFNESS STUDY RESULTS technology." In 2016 3rd Asia-Pacific World Congress on

9 © 2021 by ASME
Computer Science and Engineering (APWC on CSE), pp. 71-77. [21] Heras, Pilar, Sarah Thomas, Morten Kramer, and Jens
IEEE, 2016. Peter Kofoed. "Numerical and experimental modelling of a wave
[7] López, Iraide, Jon Andreu, Salvador Ceballos, Iñigo energy converter pitching in close proximity to a fixed
Martínez De Alegría, and Iñigo Kortabarria. "Review of wave structure." Journal of Marine Science and Engineering 7, no. 7
energy technologies and the necessary power- (2019): 218.
equipment." Renewable and sustainable energy reviews 27 [22] Hoerner, Sighard F. "Fluid-Dynamic Drag. Theoretical,
(2013): 413-434. experimental and statistical information." Copyright by: SF
[8] Alamian, Rezvan, Rouzbeh Shafaghat, S. Jalal Miri, Hoerner Fluid Dynamics, Vancouver, Printed in the USA, Card
Nima Yazdanshenas, and Mostafa Shakeri. "Evaluation of Number 64-19666 (1965).
technologies for harvesting wave energy in Caspian [23] Guillou, Nicolas. "Estimating wave energy flux from
Sea." Renewable and sustainable energy reviews 32 (2014): 468- significant wave height and peak period." Renewable
476. Energy 155 (2020): 1383-1393.
[9] Zheng, Si-Ming, Yong-Hao Zhang, Yong-Liang Zhang, [24] Babarit, Aurélien. "A database of capture width ratio of
and Wan-An Sheng. "Numerical study on the dynamics of a two- wave energy converters." Renewable Energy 80 (2015): 610-
raft wave energy conversion device." Journal of Fluids and 628.
Structures 58 (2015): 271-290.
[10] Liu, ChangHai, QingJun Yang, and Gang Bao.
"Performance investigation of a two-raft-type wave energy
converter with hydraulic power take-off unit." Applied Ocean
Research 62 (2017): 139-155.
[11] So, Ratanak, Carlos Michelen, Bret Bosma, Pukha
Lenee-Bluhm, and Ted KA Brekken. "Statistical analysis of a 1:
7 scale field test wave energy converter using WEC-sim." IEEE
Transactions on Sustainable Energy 8, no. 3 (2017): 1118-1126.
[12] Wang, L., and John V. Ringwood. "Geometric
optimization of a hinge-barge wave energy converter."
In Proceedings of the 13th European Wave and Tidal Energy
Conference, Naples, Italy, pp. 1-6. 2019.
[13] Li, Biao, Fangfang Sui, and Bingsong Yang. "An
efficient multi-factor geometry optimization based on motion
analysis and resonance response for hinged double-body floating
wave energy converter." Science Progress 103, no. 3 (2020):
0036850420950151.
[14] Li, Qiaofeng, Jia Mi, Xiaofan Li, Shuo Chen, Boxi
Jiang, and Lei Zuo. "A self-floating oscillating surge wave
energy converter." Energy (2021): 120668.
[15] Blažauskas, Nerijus, Aleksas Pašilis, and Audrius
Knolis. "Potential applications for small scale wave energy
installations." Renewable and Sustainable Energy Reviews 49
(2015): 297-305.
[16] LiVecchi, A., A. Copping, D. Jenne, A. Gorton, R.
Preus, G. Gill, R. Robichaud et al. "Powering the blue economy;
exploring opportunities for marine renewable energy in maritime
markets." US Department of Energy, Office of Energy Efficiency
and Renewable Energy. Washington, DC (2019): 207.
[17] 2021, “American-Made Challenges: Waves to Water.”
from https://www.herox.com/wavestowater/guidelines
[18] 2021, “American-Made Challenges: Ocean Observing
Prize.”from
https://americanmadechallenges.org/oceanobserving/
[19] 2020, “ANSYS AQWA 2020 R2 AQWA User’s
Manual.”
[20] WEC-Sim (Wave Energy Converter SIMulator).
(2020). National Renewable Energy Laboratory (NREL) and
Sandia National Laboratories.

10 © 2021 by ASME