Ocean Engineering 312 (2024) 119202

Contents lists available at ScienceDirect

Ocean Engineering
journal homepage: www.elsevier.com/locate/oceaneng

An innovative design for floating breakwater with Multi-objective genetic

optimal method
Puxiu Mao a,b, Cheng Chen a,b , Xinquan Chen a,b,*, Qi Zhang b, Yan Bao a,b,
Qi Yang a,b,c
a
State Key Laboratory of Ocean Engineering, Shanghai Jiao Tong University, Shanghai, 200240, China
b
School of Ocean and Civil Engineering, Shanghai Jiao Tong University, Shanghai, 200240, China
c
Shanghai Jiao Tong University Underwater Engineering Institute Co.Ltd., Shanghai, 200231, China

A R T I C L E I N F O A B S T R A C T

Keywords: The hydro-meteorological conditions in large port terminals are complex, severely hindering the safe berthing
Floating breakwater and operations of ships; in this case, floating breakwaters are suitable alternatives to traditional fixed facilities
Wave attenuation performance because of their advantages such as maneuverability, adaptability, minimal impact on marine environments, and
Parametric optimization
convenient construction and transportation. In this study, a floating breakwater with wing-plate was indicated,
Multi-objective genetic algorithm
and the influence of the structural dimensions on wave attenuation performance is explored based on potential
Computational fluid dynamics
flow theory. Additionally, a parameterized scaling optimization method based on a multi-objective genetic al­
gorithm (MOGA) was introduced and applied to the newly designed floating breakwater. Then a numerical
hydrodynamic model was constructed according to the optimization result, and computational fluid dynamics
(CFD) simulations were conducted using STAR-CCM+. Finally, wave dissipation experiments were performed
and validated the superiority wave attenuation performance of the innovative floating breakwater over the
traditional square-box floating breakwater.

1. Introduction increasingly recognized as a viable alternative.

In the last century, the superior performance of floating wave-
Breakwaters are common engineering structures in ports and coastal dissipation facilities has garnered global attention. The wooden
areas, primarily designed to mitigate incoming waves from the open sea, floating wave-dissipation facility constructed in Plymouth Harbor, UK,
maintain calm waters within harbors, protect other port structures, and in 1811 can be considered the world’s first floating breakwater. During
ensure the safe navigation, berthing, and loading or unloading opera­ World War II in 1944, Allied forces placed the famous Bombardon
tions of vessels(Dai et al., 2018). Traditional breakwaters are nearshore floating breakwater offshore along the Normandy coast, approximately
and fixed structures, consisting of rubble, concrete, etc., cast in place to 1.6 km from the shore. Japan’s first floating breakwater was installed in
create a designated width of wave-dissipating area near the shore. In Aomori Harbor in 1930, serving as a testbed for its wave resistance and
conditions of intense wave action, they reflect and break waves, dissi­ wave attenuation performance. In China, a reinforced concrete
pating wave energy and reducing coastal erosion. As permanent struc­ rectangular-section floating navigation breakwater, with a total length
tures, traditional breakwaters are immovable and non-reusable, of 134 m, was constructed in the Danjiangkou Reservoir in 1962.
requiring lengthy construction periods and being significantly influ­ The wave dissipation mechanisms of floating breakwaters primarily
enced in cost by water depth. Consequently, fixed breakwaters are no include wave energy reflection dissipation and wave energy loss dissi­
longer suitable for situations involving deep water. Floating breakwa­ pation(Ozeren et al., 2009). The wave reflection type relies on the
ters, a novel concept built upon traditional breakwater principles, have floating breakwater to reflect the energy of incident waves to attenuate
emerged to address the need for mobility, reusability, and applicability the energy of the incident waves, thereby reducing the wave height
to deep water conditions(Luan and Chen, 2021). Consisting of floating behind the breakwater and achieving wave dissipation. The energy loss
structures and anchoring systems, these floating breakwaters are type primarily relies on the interaction between floating structures and

* Corresponding author. State Key Laboratory of Ocean Engineering, Shanghai Jiao Tong University, Shanghai, 200240, China.
E-mail address: [email protected] (X. Chen).

https://doi.org/10.1016/j.oceaneng.2024.119202
Received 8 May 2024; Received in revised form 27 August 2024; Accepted 4 September 2024
Available online 12 September 2024
0029-8018/© 2024 Published by Elsevier Ltd.
P. Mao et al. Ocean Engineering 312 (2024) 119202

Fig. 1. Three main structural types of floating breakwaters.

waves, which results in collision, friction, and circumfluent motion that achieve the same wave dissipation effect with a reduced structure
disrupts the regular movement of water particles, converting it into width-to-wavelength ratio, thereby reducing costs. K. Murali et al.
chaotic turbulence to achieve wave dissipation(Kar et al., 2019). (Murali and Mani, 1997) proposed a cage-type floating breakwater
Currently, floating breakwaters designed based on these wave dissipa­ based on the Y-shaped floating breakwaters. This facility consists of two
tion mechanisms mainly include floating box type, floating cylinder Y-shaped floating breakwater connected together, and the results
type, and floating raft type, as shown in Fig. 1. showed that it has good wave dissipation effects. The study also pointed
In order to enhance the performance of floating breakwaters, out that the presence of water flow would reduce the wave dissipation
scholars from various countries have conducted extensive experimental effect, so attention should be paid to the arrangement of breakwaters to
research on the types of such facilities. VSundar et al. (2003) conducted minimize the impact of water flow. Shunichi Ikesue et al. (2002) pro­
experimental studies on the hydrodynamic performance of three posed a double-box floating breakwater structure to cope with
different diameters of floating cylinder breakwater in irregular waves. deep-water long-period waves. This structure consists of two parallel
They recorded the reflected waves in front of the breakwater and the floating boxes connected together, with horizontal fins arranged on the
transmitted waves behind it, and plotted graphs to validate the effec­ inner and outer sides of the floating boxes. The wave-dissipation per­
tiveness of the floating cylinder breakwater. Y.Ozeren et al. (2011) formance of the floating breakwater was studied through
proposed a multi-cylinder section type of floating breakwater based on two-dimensional wave tank tests, and the scale of the floating boxes and
the floating single-cylinder breakwater. They designed various fins as well as the spacing between the two boxes were optimized based
constraint models under different conditions, as a result, they suggested on the test results. Nobuhiro Matsunaga et al. (2002) proposed a
that the wave dissipation capacity of fixed breakwaters depends on the truss-type floating box breakwater, adding a truss structure on each side
relative depth of the structure’s draft. Bayram(Atilla, 2000) conducted of the conventional floating box breakwater. The truss structure breaks
experimental research on regular waves in shallow water for inclined incoming waves, and the floating box suppresses the propagation of
floating breakwaters, indicating that the wave attenuation performance small broken waves. Experimental studies showed that the wave dissi­
of floating breakwaters is related to wave height, wave period, and the pation effect is related to the ratio of truss length to floating box width
tilt angle of the breakwater. Loukogeorgaki et al. (2014) conducted and the ratio of floating box width to water depth. Mohamed R. Gesraha
physical experiments to study the impact of incident waves with (Gesraha, 2006) added two thin side plates to a conventional rectan­
different parameters on floating breakwaters under both regular and gular floating breakwater to form a cross-sectional shape resembling the
irregular wave conditions. The experiments revealed a strong correla­ Greek letter Π. Numerical calculations and experimental results showed
tion between the incident wave angle and wave height and the motion that this configuration experiences smaller wave loads and has a smaller
response of the floating breakwater under a small period; however, in heave damping coefficient, but larger other hydrodynamic coefficients,
the high-frequency range, the correlation between the incident wave resulting in better wave dissipation performance within the given wave
angle, wave height, and the motion response of the floating breakwater period range. Masashi Kashiwagi et al. (2007) proposed an improvement
was weaker. method by adding horizontal fins at different positions on the Π-shaped
J.S.Mani(Mani, 1991) proposed a Y-shaped structure with a row of breakwater and analyzed the wave-dissipation effect of the improved
cylinders added underneath a trapezoidal floating box, disrupting the model through two-dimensional potential flow theory calculations. D.H.
surrounding flow field through the cylindrical structure to dissipate Jung et al. (2006) proposed a floating breakwater made of polyethylene
wave energy. The study results indicated that this configuration can pipes, which is effective for waves with a period of 6 s and a height of 1

2
P. Mao et al. Ocean Engineering 312 (2024) 119202

Table 1
Preliminary dimensions for the floating break­
water section.
Parameter Value

Box width 12 m
Box height 8m
Box draft 3m
Wing width 3m
Wing length 2m
Wing angle 130◦

Fig. 2. Floating breakwater section.

m, but less effective for long-period regular waves or irregular waves. H.

Y.Wang et al.(Wang and Sun, 2010) proposed a porous breakwater
composed of a large number of rhombic blocks. It dissipates wave energy
mainly by absorbing wave energy rather than reflecting it, effectively
reducing tension on the mooring cables. Morteza et al.(Kolahdoozan
et al., 2014) proposed a special design of breakwater structure, incor­
porating two vertical plates beneath both sides of the breakwater, and It
is indicated that this structure enhances its wave dissipation perfor­
mance for increasing frontal area created by the bottom vertical plates.
Scholars worldwide have also conducted extensive research on
breakwaters consisting of double-layer or multiple-layer plates com­
bined with floating boxes(Liu et al., 2008a; Wang and Shen, 1999; Usha
and Gayathri, 2005; Yong-et al., 2010). This type of floating breakwater
is designed to disrupt the movement of water particles and break waves
Fig. 3. Hydrodynamic model for the new floating breakwater.
by ensuring that the spacing between adjacent horizontal plates is
smaller than the diameter of the trajectory of water particle movement,
2. The structural design and parametric performance analysis of
thereby achieving the purpose of dissipating wave energy.
innovative floating breakwater
The wave-dissipation performance of floating breakwaters is re­
flected by the transmission coefficient(Kamath et al., 2015); currently,
2.1. Structal design of innovative floating breakwater
the effectiveness of floating breakwaters is primarily evaluated through
model experiments and numerical simulations (Ji et al., 2018). The
This paper proposed a novel floating breakwater equipped with in­
feasibility of new breakwater designs or comparisons of the performance
clined side wings, as depicted in Fig. 2, which is categorized as a
of different breakwaters must be verified through experiments. E. Pena
pontoon-type floating wave attenuation structure. It comprises two
et al. (Peña et al., 2011) proposed three improved designs based on the
main components: the box section, serving as the primary structural
existing floating breakwater deployed in the Baiona Harbor and con­
element, characterized by its width, height, and draft during the design
ducted experimental research on the transmission coefficient and
phase. Functionally, the pontoon section constitutes the primary buoy­
module connection forces of four types of floating breakwaters under
ancy, providing essential conditions for wave propagation obstruction.
two-dimensional and three-dimensional wave conditions. The experi­
The appendage section consists of inclined side wings, with dimensions
mental results indicated that the twin-hull type of floating breakwater
including wing width, length, and angle. By installing these wings onto
could enhance wave-dissipation performance but also led to increased
the pontoon-type wave attenuation structure, the effective draft of the
loads at the module connections.
breakwater is increased, thereby augmenting the wave-blocking area.
Luca Martinelli et al. (2008) studied the effects of different ar­
Moreover, it enhances the vertical wave-blocking area, thereby
rangements of floating breakwaters under various wave directions on
improving the heave and roll performance of the box section(Liu et al.,
the transmission coefficient, mooring cable tension, and module
2008b). Drawing from existing engineering examples, preliminary di­
connection point loads through experiments. G.H. Dong et al. (2008)
mensions for the floating breakwater section were selected, as outlined
conducted experimental research on the transmission coefficients of
in Table 1. Based on the chosen primary dimensions, a hydrodynamic
three types of floating breakwaters—single-box, twin-box, and plate-­
model for the floating breakwater was established, as depicted in Fig. 3.
net—to find a suitable solution for deep-sea fisheries, the plate-net was
This would be considered the benchmark model in the following
recommended for its its simple structure, low cost, and effectiveness in
research, and analysis in the next part employed potential flow theory.
protecting fish cages in deep sea.
The above literature review shows that researchers investigated the
wave attenuation performance of floating breakwaters and designed
2.2. Wave attenuation performance analysis based on potential flow
different floating breakwaters with varied shapes and sizes. Still, most
theory
floating breakwaters are designed according to experience or directly
modified by one or several dimension parameters. In this study, a nu­
Floating breakwaters are typically characterized by large size and
merical design method based on the Multi-Objective Genetic algorithm
relatively slow motions, the application of potential flow theory for
(MOGA) is developed and conducted to determine an optimal structure
analyzing the motion of floating breakwaters in waves can be considered
of floating breakwaters that guarantees wave attenuation performance
(Han et al., 2022). The potential flow theory idealizes fluids as perfect,
and costs less. Finally, CFD simulations and experiments were conducted
assuming them to be homogeneous, incompressible, irrotational, and
to verify the optimal result obtained from MOGA.
devoid of viscosity. The fundamental cause of viscosity is frictional
forces within the fluid. For air and water, the two most significant fluids,
their viscosity coefficients are very small, making them thin fluids. In

3
P. Mao et al. Ocean Engineering 312 (2024) 119202

Fig. 4. Reference point locations.

most cases, the viscous forces during their motion are negligible relatively high (Zong and Fangze, 2014). Therefore, this study primarily
compared to inertial and gravitational forces, allowing viscosity to be investigates the wave attenuation performance of floating breakwaters
disregarded without causing significant errors(Gao et al., 2021). In the under longer wave periods (5~10s).
study of wave interaction with objects, the wave height is much smaller For regular incident waves, the wave field near the breakwater ex­
than the characteristic length of large-scale structures, minimizing the hibits periodic variations, with a cycle matching that of the incident
impact of viscosity while diffraction effects are pronounced(jun et al., wave. Selecting the maximum fluctuation at a specific location within
2024). Therefore, employing potential flow theory to analyze the one wave cycle determines the wave height at that location. By plotting
interaction between floating breakwaters and waves is a reasonable the wave height distribution based on varying positions, a wave height
simplification and approximation. cloud map can be generated to depict the wave conditions near the
To obtain the wave attenuation performance, reference points P1 to breakwater. Since the wave height represents the worst-case scenario at
P4 are established and depicted in Fig. 4. The wave heights at these a particular location, the transmission coefficient derived from the wave
reference points are calculated, and the ratio of the measured wave height contour map tends to be conservative, providing a reliable indi­
height to the incident wave height yields the transmission coefficient at cation of wave attenuation effectiveness. The floating breakwater under
each location. Because of the neglection of viscosity in potential flow waves with a height of 1 m and varying periods is depicted through a
theory, overall simulation results tend to be overestimated, so the wave height contour map, which is generated based on the maximum
minimum transmission coefficient among the four positions is utilized as wave surface observed at different positions within one wave cycle, as
the indicator of its wave attenuation effectiveness, which was later shown in Fig. 5. Overall, the new floating breakwater exhibits superior
validated through CFD techniques. An analysis is then conducted to wave attenuation performance. Under wave periods ranging from 5s to
assess the influence of various design parameters of the floating break­ 7s, the wave attenuation behind the floating breakwater is satisfactory,
water on its wave attenuation performance. with wave heights predominantly ranging from 0.5m to 0.7m in most
The basic principles of wave transmission and reflection were areas. However, for wave periods between 8s and 10s, the wave heights
initially explored during the early 20th century, because of the advent of behind the floating breakwater mostly align with the incident wave
computers which allowed for more complex calculations and simula­ heights, indicating poor wave attenuation performance.
tions, the terminology and standardized methods for measuring and For the square-box floating breakwater, the transmission coefficient
calculating the transmission coefficient were developed. Recent decades based on reference points is summarized in Table 2. Under wave periods
have seen innovations in materials and designs, such as floating break­ between 6s and 10s, the breakwater essentially loses wave attenuation
waters, and the transmission coefficient continues to be a critical metric effectiveness.
in the research(Kamath et al., 2015; Ji et al., 2018; Peña et al., 2011; For the new floating breakwater, the transmission coefficient at
Martinelli et al., 2008; Dong et al., 2008). reference points is presented in Table 3. Under wave periods 6s and 7s,
The wave transmission coefficient serves as a crucial metric for most areas showed transmission coefficients below 0.6, however, in
assessing the effectiveness of wave attenuation structures, and a lower longer wave periods (8s–10s), the wave attenuation effectiveness of the
transmission coefficient indicates better wave dissipation performance. breakwater diminished.
It is defined in simplified terms as follows: By comparing the wave attenuation of the two types of floating
breakwaters under different wave periods, it is observed that for wave
Ht
Kt = (1) periods of 6s and 7s, the floating breakwater demonstrated superior
Hi
wave attenuation performance, while the box-type floating breakwater
In the equation, Kt is the transmission coefficient, and Hi is the fails to attenuate waves effectively behind it. In cases where the wave
incident wave height, Ht is the transmitted wave height. period exceeds 8s, both types of floating breakwaters exhibit negligible
Based on experience and numerical simulation, floating breakwaters wave attenuation effects. For typical oceanic conditions, taking the
exhibit effective wave attenuation under short wave periods, trans­ South China Sea as an example, wave energy is primarily concentrated
mission coefficients in different sizes and wave conditions are usually in the period range of 6s–8s (L et al., 2012). From this perspective, the
under 0.5; besides, the target floating breakwater is planned to be uti­ wave attenuation performance of the novel floating breakwater is ex­
lized in the southern China Sea where the local wave heights are pected to surpass that of the conventional box-type breakwater under

4
P. Mao et al. Ocean Engineering 312 (2024) 119202

Fig. 5. Wave attenuation effect of floating breakwaters in different wave periods.

real sea conditions.

Table 2
Transmission coefficient of square-box floating breakwater.
2.3. Investigation of the influence of sectional dimensions on the wave
Period P1 P2 P3 P4 attenuation performance
6s 0.940 0.940 0.960 0.960
7s 1.070 1.080 1.060 1.070 For the proposed floating breakwater, the inclined wing-plates at the
8s 1.030 1.040 1.030 1.030
bottom serve three purposes: augmenting the vertical water-retaining
9s 1.020 1.020 1.020 1.020
10s 1.010 1.010 1.010 1.010 area of the floating breakwater, thus aiding in ameliorating its heave
motion characteristics, and increasing the effective draft of the floating
breakwater, thereby enhancing its efficacy in wave obstruction and
reflection. Concurrently, they also enlarge the inherent roll period of the
Table 3
floating breakwater, rendering it capable of delivering satisfactory wave
Transmission coefficient of the newly designed floating breakwater.
attenuation even under longer wave periods. Both the dimensions of the
period P1 P2 P3 P4 square-box main body and the auxiliary wing-plates can influence the
6s 0.820 0.580 0.660 0.590 overall wave attenuation effectiveness. In this section, based on the
7s 0.580 0.630 0.420 0.490 parametric calculation model for wave attenuation of the floating
8s 1.110 1.130 1.080 1.100
breakwater, the impact of box width, box draft, wing length, and wing
9s 1.060 1.070 1.050 1.060
10s 1.040 1.040 1.030 1.030 angle of the floating breakwater on its wave attenuation performance
was investigated.

5
P. Mao et al. Ocean Engineering 312 (2024) 119202

Fig. 6. Analysis of influence of the sectional dimensions on wave attenuation performance.

To analyze the impact of sectional dimensions on wave attenuation motion for the floating breakwater, enabling the wing-plate-equipped
performance, the controlled variables method was employed, wherein floating breakwater to absorb more wave energy when facing longer
only one scale of the benchmark model was modified in a scenario. wave periods, thereby exhibiting better wave attenuation performance
Subsequently, the wave attenuation performance of the altered model than the square-box floating breakwater.Furthermore, an analysis was
was tested under a wave height of 1 m. The parameter distribution and conducted on the influence of the sectional parameters of the new
calculated transmission coefficient are shown in Fig. 6. floating breakwater on its wave attenuation performance, which will be
It is evident that the lowest transmission coefficient for box width the theoretical support for the optimization method in the next chapter.
occurred at a wave period of 7s, and the increasing width positively Overall, all sectional parameters only affect wave attenuation perfor­
contributed to wave attenuation performance. mance obviously during a wave period lower than 8s and almost lose the
Under wave periods of 6s and 7s, floating breakwaters under effect of wave elimination if the wave period is higher than 9s. it was
different drafts experienced a local minimum in transmission coefficient. found that a larger box width and draft can significantly improve wave
The larger the draught, the larger the corresponding period of local attenuation performance during a wave period lower than 8s, for larger
minimum transmission coefficient. At the periods corresponding to the box width and draft increase interaction area between the structure and
local minimum transmission coefficient, there were significant differ­ incoming waves, leading to greater wave energy absorption and atten­
ences in transmission coefficients among floating breakwaters with uation. The impact of wing-plate parameters on wave attenuation is
different draughts. Those with better wave attenuation effects had reflected in the periods corresponding to minimal transmission co­
transmission coefficients below 0.4, while those with poorer effects were efficients; longer wing-plate lengths and smaller angles result in longer
around 0.6. wave periods associated with minimal transmission coefficients. In
There are variations in the wave periods corresponding to the local order to avoid the interference of viscosity, the CFD method was
minimum transmission coefficients for floating breakwaters with compared with the potential flow method in this paper; under a wave
different wing-plate lengths. Specifically, larger wing-plate lengths period of 5s, the transmission coefficient is 51%–59.2%, which is close
correspond to larger wave periods for the local minimum transmission to the transmission coefficient of 55% calculated based on the potential
coefficients. At these wave periods, there are some differences in flow theory.
transmission coefficients among floating breakwaters with different
wing-plate lengths, but overall, the wave attenuation effects are satis­ 3. The novel parameterized optimization method based on
factory, with transmission coefficients ranging from 0.2 to 0.4. multi-objective genetic algorithm
Under wave periods of 6s and 7s, the floating breakwaters with
different wing-plate angles exhibit a local minimum in transmission 3.1. Introduction of multi-objective genetic algorithm
coefficient. Overall, the transmission coefficients are below 0.5, with
smaller wing-plate angles corresponding to larger periods of local min­ The Multi-Objective Genetic Algorithm (MOGA) in goal-driven
imum transmission coefficient. However, under the periods between 8s optimization is a hybrid variant based on the popular Non-dominated
and 10s, all floating breakwaters with any size lost wave dissipation Sorted Genetic Algorithm-II (NSGA-II) incorporating the concept of
ability. controlled elitism, facilitating support for discrete or continuous input
This chapter introduced structure and analyzed the motion and wave parameters (design variables)(Somervell et al., 2017). The Pareto sort­
attenuation performance of a new floating breakwater, comparing it ing scheme is achieved through a rapid non-dominated sorting method,
with a square-box floating breakwater of the same size. It was observed which is an order of magnitude faster than traditional Pareto sorting
that the installation of wing-plates increased the inherent period of roll methods. Constraint handling employs the same non-dominance

6
P. Mao et al. Ocean Engineering 312 (2024) 119202

principle as the objectives. Thus, there is no need for penalty functions Table 4
or Lagrange multipliers. This also ensures that feasible solutions are Distribution of optimization parameters.
consistently ranked higher than infeasible solutions. Parameter Unit Maximum Minimum
The MOGA optimization method addresses the following form of
Box width (W) m 8 25
constrained nonlinear optimization problem: finding the minimum Box draft (D) m 2 5
value of objective functions, where there is more than one objective Wing length (WL) m 1.5 4
function, Wing angle (Wa) ◦
120 180

min f(x) = f(f1 (x), f2 (x), ⋅ ⋅ ⋅ , fk (x)) (2)

noted that a uniform distribution of individuals across the problem space
In the equation, x represents the design variables, and k denotes the
often maximizes the likelihood of achieving a globally optimal solution.
number of objective functions.
The quality of individuals in this paper is primarily indicated by the
The following constraints are satisfied:
effectiveness of wave attenuation which decides the chance of passing
hl (x) = 0, l = 1, 2, ⋅⋅⋅，L (3) on their represented genetic factors (parameters) to the next generation.
In the process of genetic evolution, there are two mechanisms: crossover
gm (x) ≤ 0，m = 1, 2, ⋅⋅⋅，M (4) and mutation. Crossover involves two individuals contributing half of
their genetic information each; if an individual provides half of the ge­
In the form, L and M represent the number of equality constraints and
netic information with random modification, this mechanism is called
inequality constraints, respectively.
mutation. Through these mechanisms, the newly generated population
Additionally, there are constraints on the design variable values, i.e.,
can preserve the characteristics of superior individuals from the original
limits on the design space boundaries:
population to the greatest extent possible, while also avoiding local
xL ≤ x ≤ xU (5) convergence.
Optimization based on MOGA has two convergence criteria. If the
In the equation, xL represents the lower bound of the design variable,
proportion of individuals in the Pareto optimal solution set reaches a
and xM represents the upper bound of the design variable.
predefined threshold or the objective functions of individuals in a
Pareto optimality is highly significant in multi-objective optimiza­
certain generation remain at a similar level during optimization,
tion especially when there are conflicting objectives and constraints,
convergence can be assumed. To avoid continuing indefinitely when
either partially or entirely. In such scenarios, a single optimal solution
convergence is not achieved, stopping criteria are required, usually
might not be found; instead, the best solution, often referred to as the
maximum number of genetic iterations is applied.
non-dominated set, comprises a set of solutions. Choosing any solution
from this set over another always involves reducing quality for at least
one objective or constraint while simultaneously improving at least one 3.2. MOGA optimization model for floating breakwater
other.
Formally, the illustration of Pareto optimality for a general optimi­ In this study, the optimization model is established as follows:
zation problem can be represented by the following equation:
For a multi-objective optimization problem, (1) Optimization objectives: Under specific wave period conditions,
there are two optimization objectives. Firstly, to achieve the best
X : {x ∈ Rn , g(x) ≤ 0, h(x) = 0, xL ≤ x ≤ xU } (6) wave attenuation performance, reflected numerically by mini­
All solutions that satisfy this condition are termed feasible solutions. mizing the reference point wave height. For this optimization, the
When there exists a set of design variables x∗ within the feasible reference point is still located 30 m behind the centerline of the
domain such that floating wave attenuator. Secondly, to minimize the weight of
steel materials or construction costs. Considering the structure as
fi (x∗ ) ≤ fi (x), ∀x ∈ X(i = 1, ⋅ ⋅ ⋅ , k) (7) plate-shell, the unit length weight is represented by the typical
Which demonstrates this set of variables is considered the optimal cross-sectional perimeter of the floating wave attenuator. A
solution when altering any of these variables fails to bring about further smaller cross-sectional perimeter implies less material usage and
improvement in any objective. lower costs.
However, in practical applications, attaining an optimal solution in (2) Optimization variables: The parameters of the new floating wave
multi-objective optimization problems is typically quite challenging. For attenuator section, including the width, height, draft, length,
a given set of variables, if there doesn’t exist any variable x that satisfies width, and angle. Since the width of the wing plates has minimal
impact on wave attenuation performance and is not involved in
fi (x) ≤ fi (x∗ ) (8) optimization, there are a total of four independent parameter
For at least one objective function, i ∈ {1, ⋅ ⋅ ⋅, k}, there exists x such variables involved in the optimization design.
that (3) Constraint conditions: For single-objective optimization,
constraint conditions primarily manifest in the range of variable
fi (x) < fi (x∗ ) (9) values and the interrelationships among parameters, as specified
The space of solutions x that meet the above conditions is referred to
∗ in Table 4.
as the Pareto optimal solution set, also known as non-dominated or non-
inferior solutions. In practical applications, typically based on the spe­ 3.3. MOGA optimization result
cific requirements of the problem and the preferences of decision-
makers, one or several solutions are selected from the Pareto optimal Using a multi-objective genetic algorithm, the optimization of the
set as the optimal solution for the analyzed problem. cross-section of the floating breakwater is conducted under the condi­
The genetic operations are conducted simultaneously across tion of a regular wave with periods from 6s to 9s. Throughout the
numerous individuals rather than individually, thus, genetic optimiza­ optimization process, a considerable number of design points are
tion algorithms initially require generating an initial population based generated, including the initial population as well as each subsequent
on parameter ranges. Through successive generations of genetic evolu­ population obtained through genetic variation. Based on the computa­
tion, the algorithm eventually converges to an optimal solution, and it is tional results of these design points, an analysis is performed on the
variations of two objective variables: the reference point wave height

7
P. Mao et al. Ocean Engineering 312 (2024) 119202

wave height exceeds the incident wave height. This phenomenon pri­
marily arises from the short wavelength of waves in shorter periods,
coupled with the limited length of the unit. The interference of different
wave components behind the floating breakwater unit is complex,
leading to significant variations in the wave height distribution map. At
a wave period of 5s, the central area near the floating breakwater unit
exhibits poor wave attenuation, while better wave attenuation is
observed in the distant and lateral regions, with transmission co­
efficients below 0.6. For wave periods ranging from 6s to 8s, the ma­
jority of the rear region behind the floating breakwater unit experiences
relatively low wave heights, with transmission coefficients below 0.6,
indicating effective wave attenuation.

4. Numerical analysis and experimental verification

4.1. The CFD model and simulation results

In this study, the computational fluid dynamics (CFD) solver Star

CCM + based on overset grid technology is employed for numerical
simulations. The solver underwent secondary development to incorpo­
Fig. 7. MOGA optimization iteration for T = 6 s. rate a six-degree-of-freedom motion solver based on dynamic deform­
able grids, a three-dimensional numerical wave generation and
absorption module, a detached eddy simulation (DES) solver module, a
Table 5
Cross section dimensions of the optimal floating breakwater. solver module for floating structure mooring systems, an overset grid
module, and a multi-body motion module. These developments resulted
Box width Box draft Box height Wing length Wing width Wing angle
in a CFD numerical solver suitable for analyzing complex geometric
13.9m 4.7m 7.05m 3.8m 2m 130◦ shapes of the floating breakwater structures and other hydrodynamic
problems in marine engineering.
In the simulation, the fluid domain of the two-phase flow includes an
and the sectional perimeter of the floating breakwater; for instance, the
upper air domain and a lower water domain, with the water domain
iterative changes in the population for T = 6s are depicted in Fig. 7.
further divided into three parts: the wave generation zone, working
Overall, the sectional perimeter mainly concentrates within the
zone, and wave absorbing zone, as depicted in Fig. 9. Here, H represents
40m–70m; In comparison, the wave height predominantly exceeds
the height of the main body of the floating breakwater structure, while λ
0.5m, and the levels of wave height and sectional perimeter in the new
denotes the wavelength of the incident wave. The height of the air
population noticeably decrease as the optimization progresses. From the
domain above the free surface is set to 4H, and the height of the water
distribution perspective, the positions of design points gradually shift
domain below the free surface is set to 8H. The lengths of the wave
towards the left and bottom sides; the sectional perimeters of design
generation zone and wave absorbing zone are both twice the wave­
points are all below 50m, and the wave heights mainly concentrate
length, while the working zone extends three wavelengths forward and
below 0.4m by the fifth generation. According to the figure, it can be
backward from the leading and trailing edges of the floating breakwater
observed that design points gradually gather on the right side of the
structure. The parameter of the numerical floating breakwater model is
Pareto front as the iterations proceed, so the optimal sectional param­
shown in Table 5.
eters can be computed by the MOGA method.
In numerical wave tanks, waves propagating to the boundaries
Executing similar iteration computation and analysis for situations
typically result in reflections, thereby interfering with the flow within
under wave periods from 7s to 9s, and Summarizing the variations in
the computational domain and affecting computational accuracy.
reference point wave height and cross-sectional perimeter for all design
Hence, it is essential to incorporate wave absorption functionality to
points throughout the optimization history. It is imperative to minimize
prevent wave reflections caused by boundary effects. In the three-
the wave height to ensure effective wave attenuation while also mini­
dimensional numerical wave tank developed by the teams, wave ab­
mizing the cross-sectional perimeter to control costs; therefore, based on
sorption is achieved by employing a sponge layer. This is primarily
the optimization history, design points should be selected from the left
accomplished by introducing a wave absorption source term fs into the
and bottom boundaries of the distribution area. It can be concluded that
momentum equation (10), which can be expressed as:
points on the boundaries are Pareto dominant to the interior design
⎧
points, they constitute the Pareto front, which represents the optimal ( )2
⎨ − ρα x − xs ( U − U )，in the wave absorbing zone
⎪
solution set. fs (x) =
s
Ls
ref
(10)
The final optimal solution is a design point under a wave period of T ⎪
⎩
0，out of wave absorbing zone
= 8s which is shown in Table 5. Based on the sectional parameters of
design point 4, the length of the floating breakwater unit was set at 50 m. In the equation, the wave absorption term becomes effective starting
The wave attenuation was computed for the floating breakwater unit from position xs; Ls represents the length of wave absorption persistence;
under the condition of vertical incidence of regular waves with a wave αs denotes the wave absorption viscosity coefficient. The relative ve­
height of 1m. The wave height distribution near the floating breakwater locity term Uref is utilized to ensure identical velocities at the inlet and
unit is depicted in Fig. 8. outlet, thereby ensuring the conservation of fluid mass during the
From the wave height distribution maps of the floating breakwater computational process.
under various wave periods, it can be observed that, in the wave periods The primary distinction between the regional wave-making method
ranging from 3s to 4s, significant fluctuations in wave height occur in and the velocity-inlet wave-making method is that the former not only
the rear region of the floating breakwater unit. There are areas with requires wave generation at the boundary but also necessitates the
effective wave attenuation, where the transmission coefficient is less modification of the flow field within specific regional boundaries. This is
than 0.3, as well as areas with unfavorable wave attenuation, where the achieved through the use of relaxation zones, ensuring no wave

8
P. Mao et al. Ocean Engineering 312 (2024) 119202

Fig. 8. Wave attenuation performance of Floating breakwater.

reflection at the external boundaries while also preventing wave re­ breakwaters to test the method mentioned above, and a fifth-order
flections within the computational domain from interfering with the Stokes wave was introduced as the incident wave, with a wave height
wave generation boundary, a characteristic not present in the boundary of 2m, a water depth of 30m, and a wave period of 9s, resulting in a
wave-making method. wavelength (λ) of 117.14 m.
Additionally, relaxation zones can be chosen to be locked (Frozen Figs. 10 and 11 illustrate the temporal variation of wave height from
Type) within a specific region of the computational domain, meaning the front and rear edges of the wing-plate floating breakwater, and it can
that the relaxation zone is bound to the grids within that range. As the be observed that there is a noticeable wave height reduction in the area
computational domain moves, the relaxation zone moves accordingly. behind the breakwater structure due to its wave attenuation effect.
This approach is suitable for numerical simulations of large-scale The transmission wave height was determined by averaging the
maneuvering vessels operating within waves. This is also the reason wave height at four positions relative to the rear edge of the floating
why this wave-making method was introduced into the solver in this breakwater: 0.25λ, 0.5λ, 1λ, and 2λ. The transmission coefficients for the
project. square-box floating breakwater and the optimal breakwater were
Conducting CFD simulations for square-box and wing-plate floating calculated as 0.833 and 0.506, so it can be demonstrated that, compared

9
P. Mao et al. Ocean Engineering 312 (2024) 119202

Fig. 9. Schematic of the computational domain for the floating breakwater.

Fig. 10. Wave heights at positions 0.05λ, 0.1λ, 0.25λ, and 0.5λ from the front edge of the wing-plate floating breakwater.

with the square-box floating breakwater, the wing-plate floating performance than square-box floating breakwaters, Additionally,
breakwater exhibited a more pronounced attenuation of incident waves comparing the CFD results with the potential flow theory results from
in the rear region. Chapter Two, it is revealed that the average difference between the two
More simulations utilized irregular incident waves adopted JONS­ is approximately 0.05, indicating a close agreement between two nu­
WAP spectrum SJ(ω), which has been widely recognized in engineering merical methods.
practice, and it is expressed as follows:
( ( )2 )
( ( )− 4 ) exp − 0.5 ω− ωp 4.2. Experiment setup and result
5 ( ) 5 ω σωp
SJ (ω) = Aγ H2 ω4 ω− 5 exp − γ (11)
16 S p 4 ωp The experiment was conducted at the State Key Laboratory of Ocean
Engineering at Shanghai Jiao Tong University. The ocean engineering
In the equation, Hs and ωp denote significant wave height and basin at the laboratory can simulate various marine environmental
spectral peak frequency, respectively; γ = 3, which is the peak conditions, including wind, waves, and currents, making it a fully
enhancement factor controlling the sharpness of the JONSWAP spec­ equipped facility in China. The main dimensions of the basin are 50m ×
trum peak. The spectral density function of the JONSWAP spectrum is 30m × 6m, and it is equipped with a high-power hydraulic wave
shown in Fig. 12: generator from Kempf & Remmers, Germany, capable of producing
More simulations introducing JONSWAP irregular waves under pe­ regular or irregular long-crested waves with a maximum wave height of
riods from 6s to 10s for the square-box(SB) and wing-plate(WP) floating 0.5m. A wave-absorbing beach on the opposite side of the wave gener­
breakwaters were executed, and transmission coefficients were ator to prevent wave reflection and dissipate wave energy is also
measured 30m, 60m, and 90m behind the breakwater. According to applied.
transmission coefficients shown in Table 6, it can be demonstrated that The model experiment was conducted at a scale of 1:20, adhering to
wing-plate floating breakwaters exhibit a superior wave dissipation this scale in both model fabrication and simulation of the marine

10
P. Mao et al. Ocean Engineering 312 (2024) 119202

Fig. 11. Waves heights at positions 0.25λ, 0.5λ, 1λ, and 2λ from the trailing edge of the innovative floating breakwater.

flexible arrangement of ballast blocks (lead blocks) to adjust the model’s

draft, among other parameters (see Fig. 13).
The ballast can be set up as Fig. 14 shown. By adjusting the position
and weight of the ballast, the mass and inertia radius of the model are
kept consistent with the design; meanwhile, the draft can also be
achieved.
Prior to the floating breakwater (at one measurement point, c0) and
after the floating breakwater, there were six measurement points: c1 and

Table 6
CFD scenarios setup and results.
Fig. 12. Spectral density function of JONSWAP spectrum. No. Wave Wave Model 30m 60m 90m Average
period height Kt Kt Kt Kt

environment. The experimental model parameters are shown in Table 7, 1 4 6 SB 0.284 0.228 0.174 0.229
and its definition can be referred to in Fig. 2. WP 0.229 0.175 0.127 0.177
2 6 6 SB 0.520 0.473 0.453 0.482
Totally ten scenarios were tested for two floating breakwater models: WP 0.514 0.440 0.430 0.461
the square-box (SB), and the wing-plate (WP). Models were primarily 3 8 6 SB 0.685 0.686 0.634 0.668
constructed of wood, with a fiberglass waterproof layer and a yellow WP 0.645 0.656 0.612 0.638
paint coating applied to the outer surface. Its structural design featured a 4 10 6 SB 0.780 0.746 0.718 0.748
WP 0.770 0.731 0.702 0.734
removable top cover and an internal hollow structure, allowing for the

Fig. 13. Model in the experiment.

11
P. Mao et al. Ocean Engineering 312 (2024) 119202

Table 7 and 0.181, respectively, compared to WP, indicating that adding wing-
Model size for SB and WP. plate can enhance the wave energy dissipation performance of the SB.
Parameter SB WP These results prove the wave dissipation function of wing-plate because
of the increase of effective draft and expansion of wave-blocking area.
Length 10m 10m
Box width 0.695m 0.695m
Box height 0.3525m 0.3525m 4.3. Comparison between experiment and numerical simulation
Wing width / 0.1m
Wing length / 0.19m
Wing angle / 130◦
In addition to the simulations described in Section 2, more simula­
tions were conducted under experimental conditions for wing-plate
floating breakwater, and its transmission coefficients are displayed in
Table 9. It can be concluded that the absolute difference between
experimental and CFD outcomes is below 0.1, indicating the rationality
and high precision of CFD simulations. Furthermore, comparing exper­
imental results with the potential flow theory, similar trends are
exhibited, so it can be considered that the parameterized analysis based
on potential flow theory is effective. Generally, the wave attenuation
performance decreases with the increase of wave period and wave
height; in the scenario with a wave period of 8s, the transmission co­
efficient is abnormally high because the wave condition is close to the
resonance frequency of the floating breakwater. Compared to the nu­
merical simulation in Chapter 2, the breakwater with optimized
sectional parameters still shows significant wave attenuation perfor­
mance during the wave period 8s–10s under higher wave height, so it
can be demonstrated that the parameterized optimization method
computed a remarkable result.

5. Conclusions and future work

Fig. 14. Configure the weight and draft of floating breakwater. In this study, a novel floating breakwater was proposed, and its wave
attenuation performance was validated through experiments and CFD
e1 at a distance of 1.5m from the centerline, c2 and e2 at 3m from the
centerline, and c3 and e3 at 4.5m from the centerline. Among these Table 8
points, c0 to c3 were located at the midsection of the floating breakwater Experiment results.
model, while e1 to e3 were positioned at the edges of the model. Fig. 15 No. Wave period Wave Height Model Hi（m） Ht（m） Kt
shows the distribution of measurement points.
1 6 2.5 WP 3.86 2.02 0.523
The experiment simulated an ocean with a water depth of 30m. The
SB 3.04 1.26 0.416
irregular wave spectrum adopted is the JONSWAP spectrum, and wave 2 7 3.5 WP 4.34 2.49 0.573
conditions and scenario results are shown in Table 8. SB 4.15 2.46 0.593
Fig. 16 depicts the comparison of the transmission coefficients be­ 3 8 5 WP 5.67 3.87 0.682
tween the SB and WP. It is evident from the figure that, except for sce­ SB 5.91 4.32 0.730
4 9 5 WP 5.88 3.86 0.657
nario 1, the transmission coefficients of the WP are lower than SB under SB 5.63 4.23 0.751
the remaining four conditions. Specifically, for the rest scenarios, 5 10 5 WP 6.17 4.09 0.663
transmission coefficients of the SB are increased by 0.036, 0.072, 0.142, SB 5.57 4.36 0.783

Fig. 15. Measurement points distribution.

12
P. Mao et al. Ocean Engineering 312 (2024) 119202

Fig. 16. Comparison of the transmission coefficients between the WP and SB.

periods, and the best result was figured when the wave period
Table 9 was 8 s.
Comparion between CFD and experiment results. (3) CFD simulations and experiments were conducted to compare the
No. Wave period/s Wave height/m Experiment CFD Difference wave attenuation performance of the floating breakwater with
1 6 2.5 0.523 0.534 0.011 wing plates and the box-type breakwater. Additionally, experi­
2 7 3.5 0.573 0.652 0.079 ments were conducted to investigate the influence of wing plate
3 8 5 0.682 0.614 − 0.067 perforation design on wave attenuation performance. The results
4 9 5 0.657 0.676 0.019 confirmed that equipping wing plates can enhance the wave
5 10 5 0.663 0.745 0.082
attenuation effectiveness of the squre-box floating breakwater.

simulations. The following conclusions were drawn from the above It should be noted that the optimal sectional parameter is not the best
work: solution in all conditions, but the design results ensure a good wave
attenuation performance in as many cases as possible. According to the
(1) A floating breakwater with a box-type main body and wing plates experiments and simulation, the optimal floating breakwater achieves
at the bottom was designed, and the influence of structural pa­ the best wave attenuation performance at wave period 6s, and the
rameters on wave attenuation performance was studied based on average transmission coefficient from 6s to 10s is 0.620.
potential flow theory. It was found that increasing the width and Although CFD simulation applied regular and irregular waves were
draft of the box body improved wave attenuation performance, executed, experiments were conducted on the wave dissipation perfor­
while the length and angle of the wing plates had a significant mance of the new floating breakwater in this study, experiments only
impact only when the wave period was relatively short. considered irregular waves based on the JONSWAP spectrum and lacked
(2) A parameterized method based on the Multi-Objective Genetic tests under regular wave conditions. According to the CFD simulation,
algorithm (MOGA) was developed to optimize the dimensions of the influence of wave parameters and structural parameters on wave
the wing-plate floating breakwater. The proper dimensions of the attenuation performance under regular waves will show a more obvious
new floating breakwater were obtained for different wave trend, so it is more suitable for regularity analysis. Therefore, in future

Fig. 17. Perforations on the wing-plates.

13
P. Mao et al. Ocean Engineering 312 (2024) 119202

work, it is possible to consider completing comparative experiments on Jun Li, Ai, Chen, Yong kun, Liu, Yong, Lin, Lu, Wang, Xin yu, 2024. Dissipative potential
flow models and energy evolution characteristics for curtain-wall caisson
floating breakwaters in regular waves.
breakwaters under wave action. Ocean Eng. 303, 117826.
Additionally, perforations on the wing-plates of the novel floating Jung, D.H., Kim, H.J., Kim, J.H., Park, H.I., 2006. A preliminary experiment study for
breakwater are also proposed, as shown in Fig. 17; this design enhances development of floater of floating breakwater. In: ISOPE International Ocean and
water passage, reduces the force acting on the entire breakwater, and Polar Engineering Conference (Pp. ISOPE-I). ISOPE.
Kamath, K., Hegde, A.V., Rao, K.B., 2015. Influence of relative draft on transmission
facilitates the movement of the breakwater body, thereby better atten­ characteristics of floating pipe breakwater with relative spacing of three. Aquatic
uating waves (H et al., 2018). Procedia 4, 206–213.
Through CFD and model experiments, the wave dissipation perfor­ Kar, P., Sahoo, T., Behera, H., 2019. Effect of Bragg scattering due to bottom undulation
on a floating dock. Wave Motion 121–138.
mance of the structure was tested. However, in CFD simulations, the Kashiwagi, M., Yamada, H., Yasunaga, M., Tsuji, T., 2007. Development of floating body
transmission coefficient of the perforated floating breakwater was only with high performance in wave reflection. Int. J. Offshore Polar Eng. 17 (1).
0.073 lower than that of the unperforated one. Model experiments Kolahdoozan, M., Alizadeh, M.J., Tahershamsi, A., Abdolali, A., 2014. Experimental
study of the performance of floating breakwaters with heave motion. Civ. Eng.
showed that under short wave periods, the transmission coefficient of Infrastruct. J. 47 (1), 59–70.
the perforated breakwater was about 0.1 lower, while under longer L, F., X, B., W, J., 2012. The study on selection of wave spectrum in the deep water zone
wave periods, there was little difference between the two. Thus, further of the South China sea. Proceedings of 2012 International Conference on Civil
Engineering and Information Technology (CEIT 2012), pp. 187–192.
research is needed to investigate the impact of wing opening configu­ Liu, Y., Li, Y.C., Teng, B., Dong, S., 2008a. Wave motion over a submerged breakwater
rations on the wave dissipation performance of floating breakwaters. with an upper horizontal porous plate and a lower horizontal solid plate. Ocean Eng.
35 (16), 1588–1596.
Liu, Yong, Li, Yu-cheng, Teng, Bin, Dong, Sheng, 2008b. Wave motion over a submerged
CRediT authorship contribution statement
breakwater with an upper horizontal porous plate and a lower horizontal solid plate.
Ocean Eng. 16, 1588–1596.
Puxiu Mao: Writing – review & editing, Writing – original draft, Loukogeorgaki, E., Yagci, O., Sedat Kabdasli, M., 2014. 3D Experimental investigation of
Software, Methodology, Formal analysis, Data curation. Cheng Chen: the structural response and the effectiveness of a moored floating breakwater with
flexibly connected modules. Coastal Eng. 91, 164–180.
Software, Methodology, Formal analysis. Xinquan Chen: Supervision, Luan, Y., Chen, B., 2021. Research progress of floating breakwater. Port Waterway Eng.
Project administration, Methodology, Funding acquisition, Conceptu­ (3), 64–69. https://doi.org/10.16233/j.cnki.issn1002-4972.20210226.034.
alization. Qi Zhang: Validation, Data curation. Yan Bao: Validation, Mani, J.S., 1991. Design of Y-frame floating breakwater. J. Waterw. Port, Coast. Ocean
Eng. 117 (2), 105–119.
Software, Data curation, Conceptualization. Qi Yang: Supervision, Martinelli, L., Ruol, P., Zanuttigh, B., 2008. Wave basin experiments on floating
Project administration. breakwaters with different layouts. Appl. Ocean Res. 30 (3), 199–207.
Matsunaga, N.L., Hashida, M., Uzaki, K.I., Uragami, Y., Kanzaki, T., 2002. Performance
of wave absorption by a steel floating breakwater with truss structure. In: ISOPE
Declaration of competing interest International Ocean and Polar Engineering Conference. ISOPE (pp. ISOPE-I).
Murali, K., Mani, J.S., 1997. Performance of cage floating breakwater. J. Waterw. Port,
The authors declare that they have no known competing financial Coast. Ocean Eng. 123 (4), 172–179.
Ozeren, Y., Wren, D.G., Alonso, C.V., 2009. Experimental and Numerical Investigations
interests or personal relationships that could have appeared to influence of Floating Breakwater Performance. US Department of Agriculture, Agricultural
the work reported in this paper. Research Service, National Sedimentation Laboratory, MS, USA.
Ozeren, Y., Wren, D.G., Altinakar, M., Work, P.A., 2011. Experimental investigation of
cylindrical floating breakwater performance with various mooring configurations.
References
J. Waterw. Port, Coast. Ocean Eng. (6), 300–309.
Peña, E., Ferreras, J., Sanchez-Tembleque, F., 2011. Experimental study on wave
Atilla, Bayram, 2000. Experimental study of a sloping float breakwater. Ocean Eng. (4), transmission coefficient, mooring lines and module connector forces with different
445–453. designs of floating breakwaters. Ocean Eng. 38 (10), 1150–1160.
Dai, Jian, Wang, Chien Ming, Utsunomiya, Tomoaki, Duan, Wenhui, 2018. Review of Somervell, Lamanto T., Thampi, Santosh G., Shashikala, A.P., 2017. A novel approach for
recent research and developments on floating breakwaters. Ocean Eng. 158, the optimal design of a vertical cellular breakwater based on multi-objective
132–151. optimization. Coast Eng. J. (4), 1750019, 1.
Dong, G.H., Zheng, Y.N., Li, Y.C., Teng, B., Guan, C.T., Lin, D.F., 2008. Experiments on Sundar, V., Sundaravadivelu, R., Purushotham, S., 2003. Hydrodynamic characteristics
wave transmission coefficients of floating breakwaters. Ocean Eng. 35 (8–9), of moored floating pipe breakwaters in random waves. Proc. IME M J. Eng. Marit.
931–938. Environ. (2), 95–110.
Gao, Huixing, Yang, Song, Fang, Qinghe, Suchao, Li, 2021. Wave forces on box-girder- Usha, R., Gayathri, T., 2005. Wave motion over a twin-plate breakwater. Ocean Eng. 32
type bridge deck located behind trench or breakwater. Ocean Eng. 237, 109618. (8–9), 1054–1072.
Gesraha, M.R., 2006. Analysis of Π shaped floating breakwater in oblique waves: I. Wang, K.H., Shen, Q., 1999. Wave motion over a group of submerged horizontal plates.
Impervious rigid wave boards. Appl. Ocean Res. 28 (5), 327–338. Int. J. Eng. Sci. 37 (6), 703–715.
H, W., Z, J., S, W., L, Y., 2018. Experiments and numerical simulation on wave absorbing Wang, H.Y., Sun, Z.C., 2010. Experimental study of a porous floating breakwater. Ocean
ability of two types of porous floating structures. Chinese J. Hydrodynamics (1), Eng. 37 (5–6), 520–527.
81–88. https://doi.org/10.16076/j.cnki.cjhd.2018.01.010. Yong-xue, W., Hua-yang, D., Chong, L., 2010. Experimental study of a pile-restrained
Han, M.M., Wang, C.M., 2022. Potential flow theory-based analytical and numerical floating breakwater constructed of pontoon and plates. China Ocean Eng. (1),
modelling of porous and perforated breakwaters: a review. Ocean Eng. 249, 110897. 183–190.
Ikesue, S., Tamura, K., Sugi, Y., Takaki, M., Kihara, K., Matsuura, M., 2002. Study on the Zong Fangze, 2014. Research on the Distribution and Changes of Waves and Wave
performance of a floating breakwater with two boxes. In: ISOPE International Ocean Energy in the South China Sea in the Past 20 Years. Ocean University of China
and Polar Engineering Conference. ISOPE (pp. ISOPE-I). master’s thesis. https://kns.cnki.net/kcms2/article/abstract?v=4HX-653LtvL4Lc
Ji, Chun-yan, Yang, Ke, Cheng, Yong, Yuan, Zhi-ming, 2018. Numerical and StApwTbiVh0YdRW7zCcykCQleFYAEUPTobL-iqJYHIvse_C58fAO5jlO6tp6EZFOy
experimental investigation of interactions between free-surface waves and A floating CWUwzBMJojoi4V_d1Ky7e9m5Ehp-leQ5w7C5DorAeUSYK4abk56ONFb
breakwater with cylindrical-dual/rectangular-single pontoon. China Ocean Eng. (4), aBgauL7u2cIDu2w-htOBRE2_Ec&uniplatform=NZKPT&language=CHS.
388–399.

14