Florida Institute of Technology

Scholarship Repository @ Florida Tech

Theses and Dissertations

4-2017

A Physical and Computational Comparison of Floating

Breakwater Design Efficiencies for Habitat Restoration in the
Indian River Lagoon
Abigail Lynn Stehno

Follow this and additional works at: https://repository.fit.edu/etd

Part of the Ocean Engineering Commons

A Physical and Computational Comparison of Floating Breakwater Design
Efficiencies for Habitat Restoration in the Indian River Lagoon

by

Abigail Lynn Stehno

Bachelors of Science
Ocean Engineering
Florida Institute of Technology
2013

A thesis submitted to the College of Engineering at

Florida Institute of Technology
in partial fulfillment of the requirements
for the degree of

Master of Science
in
Ocean Engineering

Melbourne, Florida
April 2017
We the undersigned committee hereby approve the attached thesis, “A Physical and
Computational Comparison of Floating Breakwater Design Efficiencies for Habitat
Restoration in the Indian River Lagoon,” by Abigail Lynn Stehno

_________________________________________________
Dr. Robert J. Weaver
Primary Advisor
Associate Professor of Ocean Engineering
Ocean Engineering and Sciences

_________________________________________________
Dr. Ronnal Reichard
Committee Member
Professor of Engineering
Ocean Engineering and Sciences

_________________________________________________
Dr. Theodore Petersen
Committee Member
Assistant Professor of Communication
School of Arts and Communications

_________________________________________________
Dr. Stephen Wood
Department Head
Associate Professor of Ocean Engineering
Ocean Engineering and Sciences
 Abstract

Title: A Physical and Computational Comparison of Floating Breakwater Design

Efficiencies for Habitat Restoration in the Indian River Lagoon

Author: Abigail L. Stehno

Advisor: Dr. Robert J. Weaver

This study examines the design and implementation of floating breakwaters (FB) in

the Indian River Lagoon (IRL) to support and protect living shoreline restoration

projects from damaging wave climates. The design significant wave climate of the

IRL in a 0.6 m water depth had an incident significant wave height of 0.20 m with a

wave period of 1.6 s. Based on previous studies, it has been found that habitat

restoration will be successful if the wave height is reduced from 0.20 m to 0.10 m

with the use of a FB structure. FB structures attenuate wave energies and are

transportable, allowing for removal once the living shoreline restoration is

established.

The transmission coefficient is physically tested in a wave channel and numerically

simulated in OpenFOAM for a rectangular FB structure to determine the effects of

the draft parameter (dr/d) and the breadth parameter (B/L). The dimension
iii
parameters were compared to the transmission coefficient, which is a non-

dimensional value traditionally used to compare the transmitted wave height to the

incident wave height. Additionally, the FB motion response to wave interaction is

simulated using a static structure and a dampened, dynamic structure. The physical

testing was performed in the Florida Institute of Technology (FIT) wave channel at

1:1 scale using a rectangular, wooden FB structure. The OpenFOAM solver,

olaFoam, developed in 2015 by Pablo Higuera for wave generation and absorption,

was applied to a numerical replicate of the FB structure in the FIT wave channel for

ease of comparison and validation.

A wave-by-wave analysis was performed to determine the significant wave height,

as well as a singular wave analysis that is comparable between each data set. In the

numeric modeling, the static FB structures attenuated the wave energies more

effectively than the dynamic structures due to the lack of motion response. The

numerically modeled predicted a lower transmission coefficient than the physical

testing, due to assumptions during the computation and a uniform dampening

coefficient throughout the dynamic cases. The suggested FB structure dimensions

for wave attenuation in the IRL during living shoreline restoration are a draft of 0.4

m and a breadth of 1-1.25 m in a 0.6 m water depth.

iv
 Table of Contents

List of Figures ........................................................................................................ viii

List of Tables..............................................................................................................x

Chapter 1: Introduction ..............................................................................................1

1.1 Overview of the Problem .................................................................................1

1.2 Proposed Solution.............................................................................................4

1.3 Limitation of Study ..........................................................................................5

Chapter 2: Background ..............................................................................................6

2.1 Water Waves ....................................................................................................6

2.1.1 Wave Theory..............................................................................................6

2.1.2 Wave Prediction .........................................................................................9

2.2 Living Shorelines ...........................................................................................10

2.2.1 Living Shoreline Value and Restoration ..................................................10

2.2.2 Restoration Protection..............................................................................12

2.3 Floating Breakwater Performance ..................................................................14

2.3.2 Common Testing Parameters ...................................................................15

2.3.1 Scaling .....................................................................................................16

v
 2.4 Physical Testing- Wave Tank .........................................................................17

2.4.1 Testing Domain........................................................................................18

2.4.2 Structure Designs .....................................................................................19

2.5 Computational Fluid Dynamic Modeling – OpenFOAM ..............................20

2.5.1 Program Organization ..............................................................................21

2.5.2 Solver .......................................................................................................22

2.5.3 Mesh Motion ............................................................................................27

2.5.4 Parallel Processing ...................................................................................28

Chapter 3: Implementation.......................................................................................30

3.1. IRL Wave Parameters....................................................................................30

3.2 Design Requirements .....................................................................................31

3.3 Testing Floating Breakwater Performance .....................................................32

3.3.1 Parameters ................................................................................................32

3.3.2 Scaling .....................................................................................................34

3.4 Physical Testing – Wave Tank .......................................................................34

3.4.1 Tank Domain ...........................................................................................34

3.4.2 FB Structure .............................................................................................36

3.4.3 Wave Gauges ...........................................................................................37

vi
 3.5 Computational Fluid Dynamic Modeling- OpenFOAM ................................38

3.5.1 CFD Domain ............................................................................................38

3.5.2 CFD solution ............................................................................................42

3.6 Data Processing ..............................................................................................46

Chapter 4: Results and Discussion ...........................................................................47

4.1 Dynamic FB ...................................................................................................47

4.2 Static FB .........................................................................................................53

4.3 Static FB vs. Dynamic FB ..............................................................................56

Chapter 5: Conclusions and Recommendations .......................................................59

References ................................................................................................................62

Appendix A: Wave Theory ......................................................................................73

Appendix B: Floating Breakwater Literature Review .............................................78

Appendix C: Tables of Raw Results ........................................................................87

vii
 List of Figures

Figure 1: Map of the Indian River Lagoon (Snazzy Maps) .......................................2

Figure 2: Deep water and shallow water wave motion (Anthoni, 2000) ...................8

Figure 3: Example of wave reduction past a floating breakwater, where 𝐻𝑖 is the

incident wave and 𝐻𝑡 is the transmitted wave .........................................................15

Figure 4: Florida Institute of Technology Wave Channel .......................................35

Figure 5: Wave flap and generator in channel, Riprap for wave absorption ...........35

Figure 6: Gauge locations in the wave channel for numerical and physical testing 37

Figure 7: Location of gauges and structure in the physical wave channel...............38

Figure 8: Background mesh representing FIT wave channel...................................39

Figure 9: Channel with FB structure using topoSet for breakwater 5 .....................40

Figure 10: Breakwater 5 with two-phase domain ....................................................42

Figure 11: Wave passing through channel with FB structure ..................................44

Figure 12: Dynamic FB 4 under wave forcing ........................................................45

Figure 13: Breadth parameter for dynamic FB, significant wave height .................48

Figure 14: Draft parameter for dynamic FB, significant wave height .....................49

Figure 15: Breadth parameter for dynamic FB, second-wave height ......................50

Figure 16: Draft parameter for dynamic FB, second-wave height...........................51

Figure 17: Breadth parameter for static FB, significant wave height ......................53

Figure 18: Draft parameter for static FB, significant wave height ..........................54
viii
Figure 19: Transmission coefficient for a FB with respect to breadth effects .........57

Figure 20: Transmission coefficient for a FB with respect to draft effects .............58

Figure 21: Water wave parameters (Dean and Dalrymple, 1991) ...........................73

ix
 List of Tables

Table 1: Breadth parameter testing schedule ...........................................................33

Table 2: Draft parameter testing schedule ...............................................................33

Table 3: Boundary Conditions used for olaFoam ....................................................42

Table 4: RMSE for the Dynamic FB tests ...............................................................48

Table 5: RMSE for the Static FB tests .....................................................................55

Table 6: Draft parameter raw results ........................................................................87

Table 7: Breadth parameter raw results ...................................................................88

x
 List of Symbols

B Breadth of floating breakwater structure [B]

B/L Non-dimensionalized breadth parameter []
C Wave celerity [m/s]
C(t) Constant with respect to time []
Ct Transmission coefficient []
D Particle diameter of sediment [m]
Du/Dt Total acceleration of the fluid [m/s2]
d Water depth from Mean Sea Level [m]
dr Draft of floating breakwater structure [m]
dr/d Non-dimensionalized draft parameter []
E Energy per water surface area of a wave [N/m2]
FBuoyancy Buoyant force [N]
Fr Froude Number []
Fx Force in x direction [N]
g Gravity [9.81 m/s2]
H Wave height [m]
Hi Incident wave height [m]
Ht Transmitted wave height [m]
kx,y Wave number- 2π/L [1/m]
L Wavelength [m]
Lo Deep water wavelength [m]
l Linear dimension [m]
m Mass [kg]

xi
n Outward pointing unit normal vector on boundary surface [m]
P Pressure [N/m2]
Re Reynolds number []
S Closed surface boundary [m2]
SPiston Stroke length of a piston wave generator [m]
SFlap Stroke length of a flap wave generator [m]
sϕ Volume source sink of the general tensor property [° or rad]
T Wave period [s]
Tp Limiting wave period [s]
t Time [s]
u Velocity vector [u v w] [m/s]
Uc Correction velocity [m/s]
uzi Unadjusted wind velocity [m]
uzi=10m Wind velocity at 10 m above Mean Sea Level [m/s]
V Volume [m3]
v Fluid velocity mesh motion [m/s]
vs Velocity of the boundary surface [m/s]
X, x, y, z Position vectors [m]
zw Measured wind height above Mean Sea Level [m]
ϕ General tensor property []
∇ Gradient []
∇ ∙ 𝒖𝑐 𝛼(1 − 𝛼) Artificial compression term [m/s]
α Indicator phase function []
γϕ Coefficient of diffusion []
η Free surface water level [m]
ηR Reflected free surface water level [m]
κ Curvature of the interface [1/° or 1/rad]
xii
μ Fluid viscosity [N·s/m2]
π Pi [3.14]
ρ Water density [kg/m3]
ρs Sediment density [kg/m3]
σ Angular frequency of a wave- 2π/T [1/s]
τ Shear stress on sediment [N/m2]
φ Wave phase [° or rad]
ψ Shields parameter []

xiii
 Acknowledgment

I would like to send a huge thank you to my fellow Coastal Engineering group,

Peyman, Sitara, and Andrew, who answered all of my questions, taught me the tricks

of grad school, accompanied me on my 4 o’clock walk, and made sure the tea was

always brewing. Also, thank you to the undergraduate students, Leigh, Andrew,

Nathan, and Aaron, who helped me with testing. Of course, none of us would be here

without Dr. Weaver, who brought this great group of supportive students together

and inspired us every day to critically think as we develop as young engineers. Thank

you to the Department of Ocean Engineering and Sciences who helped fund this

endeavor and provided opportunities throughout my college education.

I would also like to thank my FIT rowing family who have encouraged me along this

journey to always persevere. The opportunity that FIT rowing gave me allowed me

to accomplish a Master’s degree, and for that I cannot express my gratitude enough.

Thank you to my parents, Joe and Nancy, who rooted for me the whole way. You

guys are always welcome to use me as an excuse to vacation to the warm Florida

beaches. And to my other “parents”, Erica and Kyle, thank you for everything- I

couldn’t ask for better or more reliable friends. Thank you to all of my friends who

never left me as I dove into academia. Last but not least, thank you Kevin for sticking

beside me, feeding me, and giving me back rubs while I stayed up late working.
xiv
 Dedication

I would not be here without my coach and good friend Adam Thorstad. You have

taught me so much more than how to row, in fact rowing was just a catalyst. You

have been an integral part of my graduate experience; first by providing me the

opportunity to continue my education, then by becoming my massive workout

partner, sassy lunch buddy, compassionate lifelong mentor… You never fail to see

my potential, never let me settle for less, and always encourage me to surpass self-

expectations. Thank you for always being there to listen and reassure me through any

and all hassles and road bumps the past few years. I am constantly inspired by how

tirelessly passionate you are with your work, family, and friends. Your values and

(slightly embarrassing) personality encompasses the definition of a reliable,

dedicated, world-class friend and someone I aspire to be like. Thank you for all that

you have given me and for teaching me how to grow as a person.

"We are made to persist... that's how we find out who we are." -Tobias Wolf

xv
 Chapter 1: Introduction

1.1 Overview of the Problem

Coastal lagoons worldwide have suffered under the stresses of human development.

In recent times, increasing population coupled with the desire to live near the coast

has resulted in the removal of the plants and animals that have traditionally been the

buffer between land and sea. While aesthetically pleasing, the removal of these

species creates an ecological gap because natural process do not continue.

Restoration efforts are key to ensuring a vibrant coastal community and economy

through the redevelopment of productive coastal habitats. The Indian River Lagoon

(IRL) is an estuary where living shorelines were removed and now are being restored.

The IRL estuary spans approximately one-third of Florida’s east coast. The 195 km

long and 2-4 km wide section of the intercoastal waterway is an average of 1-3 m

deep. The IRL is home to wildlife, such as manatees, dolphins, sea turtles, sport fish

(snook, redfish, tarpon, etc.), and birds. The increasingly popular estuary is a

resource for both residents and visitors who enjoy the diverse ecosystem for

recreation and fishing. The IRL supported a 200% increase in the number of

registered recreational boaters in the surrounding counties from 1978 to 1993 (Smith

et. al, 1990; Adams et. al, 1996). Water recreation in the IRL brought in over $2

billion to the surrounding counties in 2007 alone, making the IRL a valuable asset

for the state and local businesses (Hazen and Sawyer, 2007).
1
 Figure 1: Map of the Indian River Lagoon (Snazzy Maps)

The increased activity and human development in and around the IRL negatively

impacted the water quality and habitat diversity due to shoreline hardening, water

pollution, and an increase in recreational boating activity. Since 2011, ecological

events such as harmful algae blooms and fish kills indicate the decline in the IRL

health. Environmental impacts during shoreline development are often considered

insignificant as economic benefits increase; however, the IRL can no longer sustain

the environmental stress caused by human interactions, intensifying the local support

and response to restore the estuary to its previous state. Actions to reestablish a

healthy IRL ecosystem include reducing pollution and restoring living shorelines by

reviving native species and bringing the shores closer to a natural state.

2
Human activities, water clarity, wave action, and sediment movement are all factors

that influence the success of a living shoreline restoration project. Restoration sites

must be protected from negative human interactions. Human interaction may

unintentionally destroy restoration sites through generating waves or scarring plants

with a boat propeller. This type of damage could be reduced through education and

community awareness. Water clarity is affected by water quality and turbidity.

Phytoplankton blooms, nutrient loading, and waste-water and surface runoff

decrease water quality (Fonseca et. al, 1998; Morris et. al, 2015). Turbidity, created

from boat wakes and wave actions, suspends sediments and other benthic material,

such as macro-algae. The increased energy and impaired water quality and clarity

hinder the growth of plants which require light for photosynthesis, such as seagrass,

and plants that take time to establish in a high energy shoreline environment, such as

mangroves.

Currents and wave energy create movement through the water column, benefiting

seagrass and mangrove growth by circulating nutrients and preventing stagnant

water; however, high energy waves may also uproot or prevent vegetation from

rooting. Energetic wave climates cause many restoration projects to fail because the

waves erode sediment, bury restoration sites, or dislodge juvenile vegetation and

shellfish (Koch et al., 2006); therefore, a method must be devised to reduce the

incident wave energy in order to protect restoration sites from energetic wave

climates while maintaining water circulation.

3
1.2 Proposed Solution

The success rate of living shoreline restoration projects depend on the wave climate

and the type of species used. The proposed solution involves the installation of a

transportable floating breakwater (FB) seaward of living shoreline restoration

locations in order to reduce wave energy for 2-3 years as the living shoreline is

established (Knutson et. al, 1990). FB structures reduce wave energy through

dampening and reflection as well as encourage water motion through transmitted

underflow, preventing stagnant water.

A numerical and physical analysis of the effects of FB dimensions are used to

determine the design of a FB structure for use in the IRL for living shoreline

restorations. The design will allow the structure to attenuate the incident wave while

flow under the structure will provide the circulation required for healthy living

shoreline growth. The numerical simulation is completed in the Computational Fluid

Dynamics (CFD) program, OpenFOAM, for an analysis and comparison of multiple

FB designs prior to physical modeling. Physical testing in the Florida Institute of

Technology (FIT) wave channel is performed to compare with and validate the

numerical results. The physical model consists of two wooden floating box

structures, allowing for dimension testing using two breadths and a set of drafts

obtained through ballast.

4
1.3 Limitation of Study

The project limitations were considered when planning the design and testing phases.

Key limitations are listed below:

 Time limits the mooring designs available to test in CFD and physical models

 Assumptions in the numeric model will distort final results

 Weak coupling between the motion solver and two phase wave solver

introduces instabilities in the solution, restricting solver run time

 Available materials limit physical testing and prototype deployment

 Wave tank dimensions limit testing to a quasi-3D flow

5
 Chapter 2: Background

Wave theory is presented in order to develop an understanding of wave interactions

with living shorelines and coastal structures. A background of living shorelines and

FB structures expresses their importance in coastal processes. FB performance

depends on structure dimensions, which is tested using physical and numerical

models. Physical modeling provides a small scale representation of how a FB

structure will attenuate and interact with waves. The numerical modeling program,

OpenFOAM, is introduced for the modeling of coastal zones, specifically with wave-

structure interactions.

2.1 Water Waves

Engineers require an understanding of wave theory using wave prediction in order to

design effective structures. Waves must also be evaluated at the design location to

determine their impact on the shoreline.

2.1.1 Wave Theory

Water wave phenomena is a common occurrence in everyday life, whether it occurs

as capillary ripples in a morning cup of coffee, swells that drive surfers along coastal

beaches, or tsunamis that devastate coastal communities. Scientifically, water waves

are a balance between two fluids of different densities, water and air, generated by a

force disturbance; e.g. gravitational pull, earthquake, wind and boat wake (Knauss,

1997). Dean and Dalrymple (1991) describe water waves as “a manifestation of

6
forces acting on the fluid tending to deform it against the action of gravity and surface

tension, which together act to maintain a level fluid surface”.

The initializing disturbance forces acting on a body of water transfer energy to the

resultant wave. The energy is only nominally reduced due to effects of friction and

viscosity in the fluid as the wave propagates from the origin, allowing the wave to

travel a great distance (Dean and Dalrymple 2002). Additional disturbance forces

acting on a wave during the duration of propagation compounds with the initial force,

either increasing or decreasing the total energy in a wave. The total energy in a wave

is the primary factor which influences the consequential wave height, Equation 1,

1
𝐸= 𝜌𝑔𝐻 2 Equation 1
8

where E is energy per unit surface area of a wave, ρ is water density, g is gravity, and

H is the wave height. The wave energy in a propagating wave is only reduced as a

result of the wave interacting with an object, e.g. a coastline or change in underwater

bathymetry, or a fluid of a different density, e.g. a pocket of water with a different

temperature or salinity. The extent to which wave energy is reduced by an object

depends on the physical shape, density, and porosity properties, which result in either

a reflected wave, energy dissipation through viscosity variance between water and

the object, or a combination of the two. Additionally, when a wave propagates into

shallower water the wave face becomes steeper until eventually the wave is too steep

to support the crest, resulting in the crest collapsing over the wave face, dissipating

7
energy through turbulence. A comprehensive overview of wave theory is presented

in Appendix A.

Wave energy dissipation on a shoreline creates a shear force on the sediment, which

when great enough results in particle suspension and redistribution corresponding to

the horizontally driven, cross-shore orbital motion in shallow water, Figure 2. Water

motion in shallow water creates shear stress on the bottom sediments, which when

great enough will suspend the sediment particles.

Figure 2: Deep water and shallow water wave motion (Anthoni, 2000)

In a perfect system, the net amount of sediment remains the same; however, due to

currents and irregular wave forcing, sediment accretion or erosion occurs. Erosion is

viewed as problematic, because landowners do not want to lose their property.

Higher energy wave climates create a greater shear stress on bottom sediments than

lower energy waves, increasing sediment movement under the wave.

8
2.1.2 Wave Prediction
The wave climate for a particular location or domain captures the effects of wave

forcing on coastal structures and shorelines, and can be found through wave gauge

measurements or theoretical wave hindcasting. Gauge deployments and wave

hindcasting both have advantages and limitations, resulting in biased data and

possibly excluding important wave events.

Capturing a non-biased sample of wave gauge measurements is limited by the

duration of wave gauge deployment; lengthy deployments require power support and

maintenance, while shorter deployments create the possibility of missing a

significant wave event. Research organizations, such as NOAA and FIT, deploy and

operate long-term wind and wave gauges, which NOAA publishes in a

comprehensive database online. Third party sources provide a substantial amount of

data to evaluate a wave climate; however, care must be taken to ensure the wave

properties are adjusted to account for a variation of depth or wave direction between

the gauge location and the design location or domain.

Theoretical wave climates are derived from measured wind data and rely on wave

theory and assumptions to estimate wave properties. Wind gauges are more abundant

than wave gauges because the gauges are not exclusively in water, but also on land,

where the gauges can easily be powered and maintained. Nomograms in the Shore

Protection Manual (SPM) and the Coastal Engineering Manual (CEM) correlate a

wind-driven wave climate as a result of wind speeds over a water fetch and wind

9
duration, both of which limit the maximum growth of a wave in a fully developed

sea. Additional information on wind driven wave theory is found in both the SPM

and the CEM (USACE, 1984, 2002).

2.2 Living Shorelines

Living shoreline is a term used to describe soft engineering shoreline protection

methods designed to integrate with natural shoreline processes to reduce wave

energy impacts along shorelines and maintain ecosystems through the use of strategic

installation of native plants (e.g. seagrasses, marsh grass, and mangroves) and native

animals (e.g. oyster and mussel colonies). Living shorelines play a crucial role for a

productive ecosystem by reducing erosion through wave reduction and sediment

stabilization, providing habitat for animals, and filtering toxins from air and water.

2.2.1 Living Shoreline Value and Restoration

The organisms in a living shoreline each serve an individual function in order for the

system to successfully promote sustainable and productive shoreline protection.

Seagrass and marsh grass beds not only attenuate wave energy to reduce erosion, but

also stabilize bottom sediments, cycle water nutrients, provide shelter to juvenile fish

and wildlife, and serve as food for animals such as sea turtles and manatees (Adams

et al., 1996; Morrison et. al, 2003). The vascular root system of mangroves also

prevents erosion by reducing wave energy and stabilizing sediment, as well as

sequestering and storing carbon dioxide in their biomass. Mangrove forests provide

a habitat for 220 species of fish, 181 species of birds, 24 species of amphibians and

10
reptiles, and 18 species of mammals in the IRL (Adams et al., 1996). Oysters

stabilize sediment through wave attenuation, as well as filter 6.8 liters of water per

hour per oyster, thereby cleaning the surrounding ecosystem of harmful nutrient rich

and light attenuating plankton (Riisgard, 1988).

The living, soft-engineered shoreline protection method was popularized once hard-

engineering systems, i.e. gabions, bulkheads, seawalls, rock armor, and revetments,

were found unsustainable. Hard-engineered structures mainly rely on reflection or

wave dissipation to reduce the impact of wave energy on a shoreline. Natural habitats

were removed in order to build the hard-engineered structures, reducing ecosystem

productivity around the structure location as well as promoting scouring and erosion

on the seaward side of the structure, due to wave reflection and the lack of sediment

stabilizing species. Additionally, hard-engineered structures require maintenance to

prevent structural failure.

Hard-engineering projects tend to have a high initial cost; a 1.2 m high bulkhead

costs between $125-200 per linear foot (Galveston Bay Foundation, 2011) and

continual maintenance is required to prevent structural failure. The base cost of a

living shoreline depends on the design and functionality. The marsh grass Spartina

alterniflora costs between $2-3 per plug and mangroves cost $10 per gallon pot

(Galveston Bay Foundation, 2011). The only maintenance cost is to replace

vegetation that does not survive the initial planting.

11
The need to restore living shorelines in the IRL has become critical with the increase

in population surrounding the estuary; which contributes to an increase in pollution,

waterway usage, and coastal infrastructure. While hard engineered structures only

offer shore protection from erosion, living shorelines protect the shore, increase

water quality, and encourage a healthy and productive ecosystem. Biological events

in the IRL, such as fish kills, phytoplankton blooms, and dolphin and manatee

mortality events indicate the deteriorating water quality. Scarring from boat

propellers damage and destroy seagrass beds, which take 3.6 to 6.4 years to fully

recover if no additional scarring occurs (Yates et. al, 2015). Seagrasses require 20-

33% surface irradiance for photosynthesis (Choice et. al, 2014, Morris et. al, 2015);

however, phytoplankton blooms have reduced surface irradiance and led to mass

seagrass mortality in the IRL. Mangrove removal to make room for infrastructure or

to enhance the view along the coastlines has had negative environmental impacts

ranging from loss of habitat to increased wave energy along the shore.

2.2.2 Restoration Protection

Seagrasses and mangroves naturally reproduce by releasing seeds and propagules,

which wave energy disperses until the juvenile plant settles on the shore or bottom

sediment. With an increase of hard engineered structures in the IRL, the wave energy

has increased along the shorelines due to reflection and lack of natural wave

dissipation, reducing the likelihood of juvenile plant settlement. If the plants

successfully establish roots, energetic waves may uproot the plant or suspend local

bottom sediments; therefore, reducing surface irradiance and increasing the chance
12
of plant burial. Additionally, suspended sediment from an increase in wave energies

can settle on oyster substrates, creating dead zones where oysters are unable to grow.

Between 1943 and 2000, wave energies led to an increase in dead zones in the IRL

by 27.3% (Grizzle et. al, 2002).

Restoration of living shorelines provides an advantage to juvenile plants because the

seed and propagule dispersion is replaced by individual planting. Restored juvenile

seagrass and marsh grass is directly planted into the substrate, leaving the freshly

planted propagules exposed to the wave climate and reducing the chance of survival.

Restored mangrove propagules are also exposed to wave energy after planting. In an

effort to protect the juvenile mangroves against wave attack, propagules were

encased in a PVC pipe using the Riley encasement method (REM). Johnson and

Herren (2008) observed a 10% average survival rate of REM restored mangroves

planted from 1996 to 2006, concluding that the PVC encasements constricted the

mangrove’s root system and reduced the available sunlight. Wave energy reduction

seaward of juvenile living shorelines has yet to be accomplished without harming or

restricting growth of the restored vegetation.

Complete wave energy reduction would reduce water circulation around the living

shorelines; therefore, the system would benefit more from partial wave reduction.

Research has shown that a wave height of 0.1m provides adequate water circulation

without destructive wave forcing on the plant and sediment (Roland and Douglass,

13
2005). Living shoreline restoration requires 2-3 years before the system is durable

enough to handle complete exposure to wave forcing (Knutson, 1990).

Breakwaters reduce wave energy to prevent erosion and protect harbors and

coastlines. Breakwater designs depend on the structure life, use, location, sediment

quality, sediment type, wave climate, water depth, and cost. Engineers design

traditional bottom mounted breakwaters with armor stones or other large rocks,

which limit the use on unstable sediment or in deep water and makes the structure

difficult to relocate or remove. Bottom mounted breakwaters can also create erosion

on the leeward side of the structure from scouring due to wave–structure interaction.

The non-traditional floating breakwater (FB) provides wave reduction through

reflection and dissipation, and does not obstruct underflow, allowing water to be

transmitted shoreward of the structure. Additional transmission of the wave is created

through overtopping and/or structure motion. Floating breakwaters are used as an

alternative wave attenuation system in locations where traditional breakwaters are

inadequate. A literature review is provided in Appendix B.

2.3 Floating Breakwater Performance

The performance of a FB structure depends on the structure parameters, wave

climate, and design requirements. Ideally, structures are designed to effectively

attenuate the greatest amount of energy with the least amount of material costs. Initial

performance testing evaluates the effectiveness of FB structures at attenuating waves

prior to prototyping and full-scale deployment.

14
2.3.2 Common Testing Parameters
A breakwater’s effectiveness at attenuating wave energy depends on the structure

design, including the structure breadth, draft, mass, moment of inertia, freeboard, and

mooring tension. The mass and moment of inertias depend on the weight distribution

of the structure while floating. Freeboard depends on the structure design and can

simply be an extension of the structure’s seaward wall or a more complicated pattern,

such as a step, to reduce wave energies above the free surface level. The tension in

the mooring lines affects the structure movement and rotation, where greater tension

will increase the wave dampening. Under standard mooring practices the

arrangement of mooring lines on the FB does not affect wave attenuation (Sannasiraj,

et al., 1998).

Figure 3: Example of wave reduction past a floating breakwater, where 𝐻𝑖 is the

incident wave and 𝐻𝑡 is the transmitted wave

The transmission coefficient quantifies breakwater efficiency representing the ratio

between the transmitted wave height and incident wave height, defining the

15
percentage of the incident wave energy allowed to be transmitted past the structure,

Equation 2,

Ht
𝐶𝑡 = Equation 2
Hi

where Ct is the non-dimensionalized transmission coefficient, Hi is the incident wave

height, and Ht is the transmitted wave height. The transmission coefficient allows for

a comparison of the effectiveness between different breakwater designs by

evaluating how well each FB design reduces the incident wave height, Figure 3.

Non-dimensionalized breakwater parameters are used to compare the transmission

coefficient against the breakwater design while omitting the effects of the wave

climate. The draft is non-dimensionalized using the water depth [dr/d] and the

breadth is non-dimensionalized using the wavelength [B/L]. The efficiency of the FB

draft and breadth parameters compared to the transmission coefficient and the non-

dimensionalized values, allowing for the design of the most effective and economical

FB within the design parameters.

2.3.1 Scaling
Small scale testing provides substantive results, without the assumptions required in

CFD modeling and at costs much less than those associated with a full-scale

prototype. When testing wave-structure interactions, it has become commonplace to

test small-scale models in wave channels; however, the dimensional similarity of

anticipated wave climate and physical parameters needs to be maintained using

scaling coefficients in order to obtain valid results.

16
Geometric similarity is maintained when the geometric ratio between the model and

prototype are equivalent for the length dimensions. Dynamic similarity of viscous

forces is maintained using the Reynolds number, Equation 3, and dimensional

similarity of gravitational forces is maintained using the Froude number and is

typically used for free surface flows, Equation 4,

𝜌𝑢𝑙
𝑅𝑒 = Equation 3
𝜇
𝑢
𝐹𝑟 = Equation 4
√𝑔𝑙

where Re is the Reynolds number, l is the linear dimension, μ is the fluid viscosity,

and Fr is the Froude number.

It is not always possible to scale models to both geometric and dynamic similarity

without distortion; therefore, the significance of each maintained parameter in the

final result determines which similarity is applied for each case. The restrictions

caused by scaling may cause significant errors because the scaling coefficient with

the greatest influence on the model is the only similarity maintained.

2.4 Physical Testing- Wave Tank

Physical testing in a wave channel creates a small-scale, controlled testing

environment, simulating real-world wave climates and requiring less material than

prototypes. The small-scale testing increases the ability to investigate structure

modifications over a range wave parameters which are not always readily found in

nature. Physical testing also enables engineers to receive immediate data using state-
17
of-the-art instrumentation; e.g. pressure sensors and wave gauges, which may be too

expensive or cumbersome to deploy in the field.

Carr (1950) was the first to experimentally evaluate the performance of a FB

structure, testing the Phoenix breakwater, a caisson structure that was used during

the Invasion of Normandy in 1944. The forces on the mooring design were found to

have the greatest importance on structure behavior Patrick (1951) further developed

the assessment of FB structures through the comparison of the structure dimensions

to the transmission coefficients through experimental studies. Physical tests have

since incorporated a wide array of physical parameters and structure designs that

affect the FB effectiveness.

2.4.1 Testing Domain

Wave channels provide a quasi-3D physical simulation of structure interactions with

an incident wave. The small-scale system of a wave channel creates a wave climate

in which large-scale structures can be simulated prior to expensive prototyping,

manufacturing, and deployment. The domain of the wave channel limits the size of

the wave, water depth, structure, and replicated shorelines; therefore, scaling may be

required. If the model must be scaled care must be taken to ensure similarity of the

full-scale model. Wave characteristics are estimated by wave maker theory, where

the stroke height determines the wave height for a piston, Equation 5, or flap,

Equation 6,

18
 𝐻
( )=𝑘∗𝑑 Equation 5
𝑆𝑃𝑖𝑠𝑡𝑜𝑛

𝐻 𝑘∗𝑑
( )= Equation 6
𝑆𝐹𝑙𝑎𝑝 2

where SPiston is the stroke length of the piston wave generator and S Flap is the stroke

length of the flap wave generator (Dean and Dalrymple, 1991). The wave period is

limited by the capability of the wave generator and is an important parameter for

scaling the wavelength properly.

2.4.2 Structure Designs

The theoretical draft for a rectangular, non-moving FB structure’s effectiveness can

be derived from the velocity potential equation, Equation 7,

𝐻 𝑔 cosh(𝑘(ℎ + 𝑧))
𝜙(𝑥, 𝑧, 𝑡) = − sin(𝑘𝑥 − 𝜎𝑡) Equation 7
2𝜎 cosh(𝑘ℎ)

where ϕ is the progressive wave velocity potential, H is the wave height, g is gravity,

σ is the angular frequency, k is the wave number, h is water depth, z is depth below

free surface, x is the location of a particle in the wave, and t is the time associated

with the movement of the particle. The velocity potential is evaluated over the entire

water column and compared to the velocity potential evaluated over the distance

between the draft and water depth to determine the transmission coefficient, Equation

8, which can be rearranged to find the desired draft of the structure with respect to

the transmission coefficient, Equation 9,

19
 𝜙(−ℎ 𝑡𝑜 𝑑𝑟𝑎𝑓𝑡)
𝐶𝑡 =
𝜙(−ℎ 𝑡𝑜 0) Equation 8

𝑑𝑟𝑎𝑓𝑡 = ℎ ∗ asinh(𝐶𝑡 ) Equation 9

The theoretical draft does not take into account structure motion; therefore, can only

be used to evaluate static structures.

The structure width depends on the location of the study site, and varies based on

permitting. Multiple FB structures may be used in series to attenuate wave energy

along a shoreline. The gap between the structures creates diffraction of the incident

waves, where a gap width greater than 5 times the wavelength will create

independent diffraction patterns behind each structure (Johnson, 1952).

The structure breadth must be designed to limit the horizontal pitching motion under

wave forcing. The natural period of the structure must be much greater than the wave

period in order to maintain upright position and effectively attenuate wave energy

(Patrick, 1951). The natural period is determined from the breadth and the

metacentric height, which can be adjusted through repositioning of ballast weight.

2.5 Computational Fluid Dynamic Modeling – OpenFOAM

Computational fluid dynamic (CFD) modeling is used to simulate fluid flow using

theoretical equations, assumptions, and user input. Open source Field Operation and

Manipulation (OpenFOAM) is a free, open-source CFD program produced by

OpenCFD Ltd. and supported by the user community. Open-source programs

20
eliminate the initial program costs that are found with comparable closed-source

programs, such as Ansys Fluent and Flow-3D, and allows for the customization and

expansion of the numerical solvers. The source codes are written in C++ and the

program computes the numerical calculations in the open-source Linux operating

system. The only costs in OpenFOAM results from processing time and computer

memory.

2.5.1 Program Organization

OpenFOAM is comprised of program files which are organized to represent physical

objects and abstract entities, designating each code file to its associated operation i.e.

meshing, solving, or post-processing. Each code file is independent and modifiable,

creating a modular structure allowing for flexibility in the development of

customizable PDE solvers and their utilities. OpenFOAM is initially compiled with

wmake, similar to UNIXmake, creating a dependency list to check whether its source

files have been updated, and then selectively updating the outdated files.

OpenFOAM does not require recompilation after the initial compilation unless a case

file is modified or if switching between release and debug modes.

Each case file (.C) contains the unique algorithms required during solving for each

specific case. The case file is suspended during initial runtime to source the header

files (.H), which are dynamically linked to shared object (.so) files, containing

information for initialization and supporting equations for solving. The modular

21
structure of OpenFOAM supports implementation and modification of source files,

which can also be used over multiple solvers.

Data may be analyzed during the simulation or after, using post-processing. The

command line interface initializes post processing tools which produce similar

results as instrumentations; such as field and flow rate calculations, forces, minimum

and maximums, and probe measurements. The ParaView Graphical User Interface

(GUI) is a third party, open-source visualization tool kit which allows users to

visually process flow output data as well as calculated variables, such as velocity and

pressure, in the domain, sections of the domain, or along individual boundaries.

2.5.2 Solver
Pablo Higuera developed the solver olaFoam in 2015 to solve Reynolds-average

Navier-Stokes (RANS) equations for two phase incompressible flows using finite

volume discretization (FV) and the volume of fluid method (VOF) to generate and

absorb a two-phase wave flow and simulate fluid interaction with porous media

through volume averaging RANS equations.

General Solution

The olaFoam solver iteratively solves for velocity and pressure, Equation 10, which

is governed by the continuity equation, Equation 11, and the momentum

conservation equation, Equation 12, allowing for the solution of RANS flow with the

FV method using the pressure linked equations (PIMPLE) algorithm,

22
 ∇2 𝑃 = 𝑓(𝑢, ∇𝑃) Equation 10

∇ ∙𝒖=0 Equation 11

𝛿𝜌𝒖
+ ∇ ∙ (𝜌𝒖𝒖) − ∇ ∙ (𝜇𝑒𝑓𝑓 ∇𝒖) = −∇𝑃 − 𝑔𝑿∇𝜌 + ∇𝒖 ∙ ∇𝜇𝑒𝑓𝑓 + 𝜎𝜅∇𝛼 Equation 12
𝛿𝑡

where P is pressure, u is velocity, ρ is density, μeff is the effective dynamic viscosity

which accounts for turbulent effects, g is gravity, X is the position vector, κ is the

curvature of the interface, and α is the indicator VOF function.

An indicator phase function quantifies the VOF in each cell, creating a distinct value

between air and water where a value of 0 represents an air filled cell, 1 represents a

water filled cell, and between 0 and 1 represents a partially filled cell. A sharp

waterline interface, Equation 13, is derived from cell density and the advection

equation, while a multi-dimensional limiter for explicit solution (MULES) factor

limits the fluxes of the discretized divergence term to ensure the indicator phase value

remains between 0 and 1 during calculation,

𝑑𝛼 Equation 13
+ ∇ ∙ 𝒖𝛼 + ∇ ∙ 𝒖𝑐 𝛼(1 − 𝛼) = 0
𝑑𝑡

where ∇ ∙ 𝒖𝑐 𝛼(1 − 𝛼) is an artificial compression term.

The PIMPLE algorithm is a combination of the PISO (pressure implicit with splitting

of operators) algorithm and the SIMPLE (semi-implicit method for pressure linked

equations) algorithm. PISO uses one predictor step and two corrector steps to ensure

mass conservation, providing a stable and quick calculation of velocity and pressure,

23
while SIMPLE ensures convergence within one time step using a relaxation factor,

Equation 14,

𝜙 𝑛 = 𝜙 𝑛−1 + 𝛼𝜙 (𝜙 𝑛𝑒𝑤 − 𝜙 𝑛−1 ) Equation 14

where Φ is the relaxation factor, n is the time step, αΦ is the variable undergoing

relaxation. PIMPLE attempts to predict velocity explicitly based on previous

pressure calculations, while also ensuring convergence with the relaxation factor,

resulting in quick and stable solutions.

Wave Generation

Current operational OpenFOAM models attempt to achieve accurate and cost-

effective wave generation using boundary inlet flow conditions, the VOF method,

and relaxation zones. The GroovyBC is a boundary condition that generates a wave

flow at the inlet using a simple mathematical wave theory expression. Compatible

solvers do not incorporate a VOF free surface calculation coupled with the

GroovyBC; therefore, additional mesh refinement is required along the projected free

surface to reduce step disturbances. The waves2Foam solver passively generates and

absorbs waves in relaxation zones for a VOF solution. The olaFoam solver couples

the VOF method with the wave generation at the inlet boundary face, reducing

overall computational costs compared to previous methods.

Current wave theories implemented in olaFoam are for regular, irregular, solitary,

and wave maker waves. Due to the modular structure of OpenFOAM, additional

24
wave theories may be implemented. Stokes 1 theory produces a simple and analytical

regular wave through solution linearization using the following assumptions:

 The fluid is inviscid, incompressible, continuous, and homogeneous

 Coriolis forces and surface tensions are neglected

 Pressure at the free surface is uniform and constant

 Flow is irrotational

 Bottom domain is fixed and impervious

𝐻
 Relative wave height is small (i.e. 𝑑 ≪ 1)

Wave motion is computed in the PIMPLE algorithm to solve for the free surface,

Equation 15, as well as the x- and z- direction velocity components for a 2

dimensional case, Equation 16 and Equation 17, using input files defining the wave

height, period, phase, and direction. Additionally, the dispersion equation is

rearranged to solve for the wavelength, Equation 18.

𝐻
𝜂= cos(𝑘𝑥 𝑥 + 𝑘𝑦 𝑦 − 𝜎𝑡 + 𝜑) Equation 15
2

𝐻 cosh(𝑘(𝑑 + 𝑧))
𝑢= 𝜎 cos(𝑘𝑥 − 𝜎𝑡 + 𝜑) Equation 16
2 sinh(𝑘𝑑)

𝐻 sinh(𝑘(𝑑 + 𝑧))
𝑤= 𝜎 sin(𝑘𝑥 − 𝜎𝑡 + 𝜑) Equation 17
2 sinh(𝑘𝑑)

𝑔𝑇 2 2𝜋𝑑
𝐿= tanh ( ) Equation 18
2𝜋 𝐿

25
where T is the wave period, η is the free surface water level, k is the wave number

(2 π/L), σ is the angular frequency of a wave (2π/T), t is the time vector, x and y are

position vectors, and φ is the wave phase. Oblique waves may be created in 3D

domains, further information is found in the olaFoam manual.

Wave Absorption

Computational costs restrict the size of fluid domain for free surface wave models;

however, restricted domains typically generate reflection along the outlet boundary,

distorting the results and possibly enable seiching. Passive and active wave

absorption techniques attempt to reduce the amount of reflection on the outlet

boundary. Passive wave absorption in current OpenFOAM models is achieved

through the addition of relaxation zones, which represent a physical beach or porous

media. Active wave absorption in CFD models is achieved through a numerically

imposed velocity profile, inverse to the incident wave velocity, resulting in a zero-

sum velocity at the outlet boundary.

The olaFoam solver implements an active wave absorption boundary to reduce

computational costs. Incident waves are absorbed through: a 2-D absorption theory,

for which the angle of incidence is perpendicular to the outlet boundary, Equation

19; a Quasi 3-D absorption theory, which requires a known angle of incidence; or a

3-D absorption theory, which uses a digital filter to measure the directionality of the

incident wave to adjust the angle of the correction velocity profile,

26
 𝑔
𝑈𝑐 = −√ 𝜂𝑅 Equation 19
𝑑

where U is the correction velocity and η is the reflected free surface water level. The
c R

2-D absorption theory may be extended to a 3-D mesh when the angle of incidence

is known.

2.5.3 Mesh Motion

Dynamic solvers allow defined or calculated mesh motion, where defined mesh

motion is specified through user input and calculated mesh motion is derived from

the physics of the case. Updated cell volume and the volume face flux of each cell is

solved using a second order discretization of the governing equation over an arbitrary

moving control volume, Equation 20, with the addition of the space conservation

law, Equation 21,

𝑑
∫ 𝜌𝜙𝑑𝑉 + ∮ 𝜌𝑛 ∙ (𝒗 − 𝒗𝒔 )𝜙𝑑𝑆 − ∮ 𝜌𝛾𝜙 𝒏 ∙ ∇𝜙𝑑𝑆 = ∫ 𝑠𝜙 𝑑𝑉 Equation 20
𝑑𝑡 𝑉
𝑉 𝑆 𝑆

𝑑
∫ 𝑑𝑉 − ∫ 𝒏 ∙ 𝒗𝑠 𝑑𝑆 = 0 Equation 21
𝑑𝑡 𝑉 𝑆

where ρ is density, V is the moving volume, n is the outward pointing unit normal

vector on the boundary surface, v is the fluid velocity, vs is the velocity of the

boundary surface, S is the closed surface boundary, γϕ is the coefficient of diffusion,

and sϕ is the volume source sink of the general tensor property, ϕ (Jasak and Tukovic,

2010).

27
OpenFOAM’s dynamic mesh solver 6DoFRigidBodyMotion calculates mesh

movement for 6 degrees of freedom (6DoF) while maintaining the rigid body

structure within the mesh. The 6DoF solver coupled with olaDyMFoam, the dynamic

olaFoam solver, provides translational displacement from fluid forces and rotational

displacement from momentum forces, dependent on the calculated pressure and fluid

velocity. Motion is customized with the use of constraints (reducing the degrees of

freedom by eliminating motion along an axis, line, orientation, plane, or point) or

restraints (restricting motion through translational or rotational damping).

2.5.4 Parallel Processing

OpenFOAM has the ability to process solvers and utilities in serial processing

(calculating the solution over the entire domain on a single core) or parallel

processing (dividing the domain between multiple processors; where each processor

calculates the solution for its allocated domain).

Parallel processing requires a decomposition of the mesh geometry through the

creation of subdomains. A simple, hierarchical, manual, or algorithmic division of

the mesh into domains is processed with the OpenFOAM utility decomposePar. Each

subdomain is then assigned to separate processers. The simple decomposition

method separates the geometry into equal parts, according to Cartesian coordinate

directions. The hierarchical method separates the geometry similar to the simple

method, but the order of directions is specified. The manual decomposition method

requires the user to specify the allocation of each cell to each processor. The Scotch

28
and METIS algorithms attempt to minimize the number of processor boundaries

during the decomposition; therefore, requiring less computational time during the

solving stage.

An open-source messaging passing interface (OpenMPI) allows inter-processor

communication during the simulation. Spreading the workload over multiple

processors increases the speed of the computation; however, OpenFOAM requires a

significant amount of inter-processor communication to ensure accurate results.

Inter-processor communication creates a bottleneck during the solving process,

preventing improved system performance with the addition of processors (Keough,

2014).

29
 Chapter 3: Implementation

The design of a FB in the IRL for habitat restoration depends on the wave climate

and the effectiveness of the structure. Measured wind and wave data in the IRL was

used to determine the design wave climate. Numerical modeling in OpenFOAM was

completed for static and dynamic FB structures and further tested in a wave channel

for validation. The data was processed using a wave-by-wave analysis to determine

the transmission coefficient for each structure.

3.1. IRL Wave Parameters

The wave climate in the IRL is primarily driven by wind and boat wake. Previous

studies in the IRL present a wind-driven wave climate with a wave height of 0.3 m

at a 2.1 s wave period in a 1.8 m water depth, hindcasted from a 3.4 – 5.1 m/s winter

wind speed in Martin County, FL (Douchock and Landry, 2006). A four month study

conducted in Jensen Beach, FL presented a boat induced wave climate with a wave

height of 0.4 m at a 2-2.5 s wave period in a 2.0 m water depth (Shermet, 2012).

Nomograms in the SPM and CEM correlate the wind speed to wind duration and

water fetch, representing the significant wind induced wave climate (USACE, 1984,

2002). Wind gauges, such as the one at Trident Pier, Cocoa, FL, owned and

maintained by NOAA National Ocean Service since 2005, and one at Sebastian Inlet,

FL owned and maintained by FIT since 2013, captured a 6.33 m/s significant wind

30
climate in the IRL between gauge deployments and 2015, a time frame in which

tropical storm and hurricane wind forces were not present. Following the

methodology outlined in the SPM and CEM, a significant wave height of 0.2 m at a

1.6 s wave period represents the fetch-limited wave climate in the IRL over a fetch

of 4 km, the greatest distance between the East-West shores where the majority of

the shorelines lay.

3.2 Design Requirements

Living shoreline restoration requires protection from wave climates that are too

energetic to support juvenile living shorelines. In the IRL, reducing the significant

incident wave height of 0.2 m to 0.1 m will provide the water motion required for

living shoreline restoration without destructive wave energies. A FB will provide

temporary wave attenuation for the establishment of living shorelines during the first

2-3 years of the restoration phase.

The effectiveness of a FB depends on the breadth and draft dimensions; where an

increase in FB size is proportional to not only the effectiveness, but also material

costs. The performance of a FB structure is defined by the transmission coefficient,

which can be compared to the non-dimensionalized structure parameters breadth vs.

wave length, B/L, and draft vs. depth, dr/d. FB structures are ineffective in deep

waters because the length of the wave is much greater than the breadth of a practical

structure. Traditionally, FB structures are located in shallow water along coastlines.

The IRL design wave is considered to be a shallow water wave in a 0.061 m water

31
depth, thus a FB in truly shallow water will not be effective, as the wave will break

before reaching that depth. An intermediate water depth of 0.6 m is used for the

design depth. Surveys in the IRL show that a 0.6 m design depth is within 3 to 10 m

from the shoreline, while also seaward of wave breaking for the design wave.

3.3 Testing Floating Breakwater Performance

The design requirements developed the framework for a FB structure in the IRL for

shoreline protection. The parameters required are established from the IRL wave

climate and could be scaled to accommodate physical testing in the wave channel.

3.3.1 Parameters
Numerical and physical testing was done to compare the effects of the breadth and

draft of a FB structure in the IRL wave climate. Seven structure designs were tested

to determine the effects of the FB breadth and six structure designs were tested to

determine the effects of the FB draft. Additionally, cases for both static and dynamic

structures were tested to compare the effects of structure motion. The natural pitch

period of each structure is provided to establish the comparison between structure

motion and wave period.

The unscaled IRL wave climate has a wave height of 0.2 m, at a 2 s period in a 0.6

m water depth. The structures tested are shown in Table 1and Table 2, where the X

denotes a test was completed. Physical testing was limited by the material needed to

make multiple FB structures. The mooring configuration and line tension did not

change during physical testing. The dynamic numerical testing was limited due to a

32
weak coupling between the two-phase wave generation solver and the dynamic mesh

in OpenFOAM. The restraints and constraints in the numerical dynamic testing

remained constant during testing.

Table 1: Breadth parameter testing schedule

B/L B Tf Numerical OpenFOAM Physical Wave Channel

[] [m] [s] Static Dynamic Static Dynamic
0.0836 0.25 negligible X
0.1672 0.50 1.11 X X X
0.2341 0.70 0.87 X X
0.2843 0.85 0.83 X X
0.3345 1.00 0.81 X X X X
0.4181 1.25 0.79 X X
0.6689 2.00 0.78 X
Table 2: Draft parameter testing schedule

dr/d Dr Tf Numerical OpenFOAM Physical Wave Channel

[] [m] [s] Static Dynamic Static Dynamic
0.0000 0.0 negligible X X
0.1667 0.1 0.49 X
0.3333 0.2 0.68 X X X
0.5000 0.3 0.82 X X X X
0.6667 0.4 0.94 X X X X
0.8333 0.5 1.03 X X

where B/L is the non-dimensionalized breadth parameter, B is design breadth, Tf is

natural pitch period, dr/d is the non-dimensionalized draft parameter, and d is the

design draft.

33
3.3.2 Scaling
The FIT wave channel is limited by the height of the walls, length of the tank, and

wave maker capabilities. Fortunately, the IRL design water depth and wave climate

are able to be replicated in the wave channel and the channel is long enough to

generate a fully developed wave before structure interaction; therefore, a 1:1 scale

was used for the physical testing. The ability to replicate the design wave at the 1:1

scale during the wave channel reduces errors caused by scaling parameters, where

the Froude number would exclude viscosity effects in order to maintain similitude in

a free surface flow. The domain used for the numerical testing replicates the

dimensions of the physical wave channel in order for reproducible and comparable

testing between the methodologies.

3.4 Physical Testing – Wave Tank

The IRL wave parameters were achieved in the FIT wave channel with a 1:1 scaling.

A wave flap was used to generate the wave, while a riprap shoreline passively

absorbed the transmitted wave. The FB design is based on the velocity potential in

order to obtain the desired wave attenuation.

3.4.1 Tank Domain

Floating breakwater physical tests were performed in the FIT Surf Mechanics

Laboratory, Melbourne, FL. The wave channel total length is 9.085 m, the effective

length with the wave paddle installed is 8.5 m, the width is 0.5746 m, and the depth

is 0.914 m, Figure 4.

34
 Figure 4: Florida Institute of Technology Wave Channel

The wave generator in the FIT wave channel creates a controlled testing environment

using either a flap or piston paddle with a stroke length of 0.1143 m, 0.2159 m, or

0.3683 m. Furthermore, the wave climate may be easily changed through

modification of the motor generator frequency, allowing for a range of wave periods.

A riprap structure eliminates reflection at the opposite end of the tank, reducing the

influence of reflected waves. While wave generator theory may provide a framework

for the resultant wave height, in practice the wave height is affected through

variances between individual wave generators. In order to simulate waves in the IRL,

the stroke length for the flap paddle was set at 0.3683 m to generate the design wave,

Figure 5.

Figure 5: Wave flap and generator in channel, Riprap for wave absorption

35
3.4.2 FB Structure
The wave channel testing evaluated the effects of the structure’s draft and breadth.

Two 0.508 m x 0.508 m x 0.7 m FB box structures were made of plywood walls and

supported with a base and cross beams made of 2x4 wood, spanning the width and

breadth of each structure. A 2x4 beam was installed in each box 0.2 m above the base

as a reinforcement for bolts, which connected the structures together and were sealed

with rubber gasket and mounting putty. Additional 2x4 beams were installed across

the base to provide a support for the weight plate ballast.

The FB boxes could be tested individually or connected, effectively doubling the

breadth. The draft to range between 0 -0.6 m and was controlled using weight plates

for ballast. The structure spanned the width of the tank, eliminating diffraction by

creating a quasi-3D system. Theoretically, a draft of 0.3 m is required to obtain a

transmission coefficient of 0.5 in a 0.6 m water depth for a static structure. For

comparison, drafts of 0.2 m and 0.4 m were tested to further validate the numerical

model. The theoretical derivation of the required draft does not include the structure

motion response to wave forcing, thus is only appropriate for a static FB, which is

impractical in the field; therefore, the effects of the draft for a dynamic FB were

found empirically and validated during physical testing.

The FB structure was set 3.85 m from the face of the wave generator paddle and

moored to the plexi-glass wave channel floor using suction cups and 1.1 m taught

fishing wire. Additionally, a 0.2 m nylon string attached the FB gunwale to the top

36
edge of the tank, limiting roll and surge motion and simulating a mooring system.

The mooring system was constant for the dynamic testing to eliminate variable

mooring designs. Weight plates were placed at the bottom of the FB structure as

ballast to obtain the desired draft. The static FB was held in place with wooden and

composite shims, located between the plexi-glass and each corner of the structure.

3.4.3 Wave Gauges

The Sonic Wave Sensor XB from Ocean Sensor Systems, Inc., recorded the free

surface height of the water in the wave channel. The sonar sensors are able to

measure up to a 0.1 mm water level resolution with a frequency up to 32 Hz. Water

level output is viewed in the GUI and saved in a .csv file for post-processing.

Figure 6: Gauge locations in the wave channel for numerical and physical testing

Two sonar gauges were positioned in the channel to record the incident wave height

2.30 m from the wave maker and transmitted wave height 5.56 m from the wave

maker, Figure 6 and Figure 7. The location of the incident wave height gauge ensured

the wave had time to equilibrate before it was measured, but also far enough from

the structure to reduce the measurement of the reflected wave. The location of the

transmitted wave height gauge was placed after the transmitted wave re-developed

behind passing the structure.

37
 Figure 7: Location of gauges and structure in the physical wave channel

3.5 Computational Fluid Dynamic Modeling- OpenFOAM

OpenFOAM setup to test the FB structure includes mesh generation, FB structure,

and boundary addition. The numerical wave solver, olaFoam, requires an

understanding of the solving process using numerical schemes and solutions.

Additional variables, i.e. restraints and constraints, further improve solver

performance.

3.5.1 CFD Domain

Mesh generation establishes the domain for a CFD case. OpenFOAM requires two

meshing steps in order to model a FB structure: the blockMesh function to establish

the background mesh and either the snappyHexMesh function or the topoSet function

to add FB geometry.

38
Mesh Generation

The CFD mesh creates a domain through sets of points, faces, cells, and boundaries.

The blockMesh function creates a 3D, hexahedral block with specified vertices,

allowing discretization for mesh refinement to prepare for the solving of partial

differential equations. The complexity of the domain determines if the mesh is

unstructured, which requires an iterative approach to solving; therefore, requiring

less processing memory, or structured, which requires simultaneous solving for each

discretized cell volume, converging the solution in fewer time steps.

Figure 8: Background mesh representing FIT wave channel

An 8.5 m x 0.575 m x 0.9 m block domain replicates the dimensions of the FIT wave

channel, ensuring repeatability and consistency between physical and numerical

testing, where gravity acts in the –z direction and wave propagation occurs along the

x-axis, Figure 8. The block domain is discretized into a 40 x 20 x 40 cell count, where

the cell count in the z-direction is proportionally greater than those in the x and y

directions due to refinement along the two-phase surface for an accurate solution;

however, the cell count in the z direction would be greater for a solver that did not
39
use the VOF method at the surface interface. A cell gradient may be added along the

free surface boundary for complicated cases.

Floating Breakwater Implementation

The snappyHexMesh and the topoSet functions manipulate the background mesh to

generate a FB structure. The snappyHexMesh function castellates a mesh around a

surface geometry, created in a computer animated drawing (CAD) software, by

removing unused cells and iteratively refining the mesh around the object to generate

intricate geometries. Instant pressure instabilities occur on snapped boundaries due

to a weak coupling between mesh motion, VOF solvers, and 6DoF solvers, leaving

the snappyHexMesh function ineffective for certain cases (Higuera). The topoSet

function redefines a set of cells, patches, or faces; for example, a block of cells are

redefined as a boundary face to create a FB structure. The topoSet function reduces

the pressure instabilities on the structure boundary, but limits the complexity of the

FB structure design, Figure 9.

Figure 9: Channel with FB structure using topoSet for breakwater 5

40
Boundary Definition

Boundary patches establish a physical surface on each domain face, while boundary

conditions define the fluid behavior on each boundary face. Boundary patches and

conditions represent the physical function of the boundary, such as a wall,

atmosphere, inlet, or outlet, Table 3.

Wave generation for the olaFoam solver is performed along the inlet boundary and

is defined in the modifiable waveDict file. The waveAbsorption2DVelocity

boundary condition calculates the expected height of the wave along the outlet

boundary at each time step, then measures the actual height of the water on the

boundary; allowing the water to “flow out” of the boundary if the expected height is

less than the actual height, reducing reflection. The Direchlet fixedValue boundary

condition creates a no-slip velocity boundary along the walls, and the zeroGradient

boundary condition creates a zero normal gradient on the boundary. The atmosphere

boundary condition alternates between the zeroGradient condition, when there is

outflow at the boundary, and fixedValue, when there is inflow at the boundary. The

fixedFluxPressure boundary condition creates a zero gradient condition at walls

where body forces such as gravity and surface tension, are present. The

movingWallVelocity boundary condition allows the boundary to move according to

the motions, such as 6DoF, specified in the dynamicMeshDict file.

41
 Table 3: Boundary Conditions used for olaFoam

U p_rgh alpha.water

Dimension m/s lb/m·s

Inlet waveVelocity fixedFluxPressure waveAlpha

Outlet waveAbsorption2DVelocity fixedFluxPressure zeroGradient

Wall fixedValue fixedFluxPressure zeroGradient

atmosphere pressureInletOutletVelocity totalPressure inletOutlet

FB movingWallVelocity fixedFluxPressure zeroGradient

The air-water interface is created using the setFieldsDict which establishes the initial

two-phase domain. The user assigns a group of cells either 0 or 1, where 0 is air and

1 is water, creating each phase, Figure 10. The density and viscosity properties of

each fluid is defined in the transportProperties file.

Figure 10: Breakwater 5 with two-phase domain

3.5.2 CFD solution

CFD testing is cheaper, easier, and faster than physical modeling. The FB structures

and mesh domain dimensions were based on the wave channel tests to allow for

42
reproducibility. The CFD computation of the RANS equation requires numerical

schemes and solution techniques to quickly converge. The 6DoF solver calculates

mesh motion from the water forces on the moving structure.

Numerical Scheme and Solution Techniques

Numerical schemes are define how the solver interprets terms in the equations, such

as derivatives, during finite volume discretization. Each solver requires a specific

integration approach to successfully and quickly solve an accurate solution. olaFoam

solves for a transient, first order implicit, bounded flow; therefore, the Euler method

is used for the numerical time scheme. Gaussian integration linearly interpolates

values between cell centers to face centers for the gradient, divergence, and

Laplacian schemes. The Laplacian term is evaluated using Gaussian integration for

the surface normal gradient scheme using explicit non-orthogonal correction.

Numerical schemes are defined in the fvScheme dictionary.

The method of lower upper (LU) decomposition is used to solve variable matrices

and algorithms using linear solvers. Preconditioners and smoothers reduce the

iterations needed to attain convergence. Solution tolerances are defined for each

solver to reduce the error in the solution. The solution techniques are defined in the

fvSolution dictionary.

The system of linear equations for alpha and the velocity variables are solved using

a smoothing solver with a symmetric Gaussian Seidel smoother, which defines the

matrices with non-zero diagonals. The pressure corrector matrix is solved with a
43
preconditioned conjugate gradient (PCG) which requires a preconditioner,

geometric-algebraic multi-grid (GAMG), and is also aided with the simplified-

diagonal based incomplete-Cholesky (DIC), Gaussian Seidel smoother. The DIC

Gaussian Seidel smoother uses the previously calculated reciprocal of the diagonal

in addition with the Gaussian Seidel smoother, to reduce variations caused by the

DIC smoother for symmetric matrices.

The GAMG solver is used to solve the pressure and cell displacement matrices by

generating a coarse grid from the mesh using agglomeration to guess the solution

over an unrefined mesh then generating the solutions over a fine grid. The GAMG

solver is faster and cheaper because it reduces the number of cells per calculation.

The DIC smoother is used with the pressure matrix and the Gaussian Seidel smoother

is used with the cell displacement matrix.

Figure 11: Wave passing through channel with FB structure

Tolerances for each of the solvers, preconditioners, and smoothers are defined in the

fvSolution. Lower tolerances increase iterations, which will increase computational

44
time but reduce errors and improve convergence. The PIMPLE algorithm controls

determine amount of momentum-pressure loop iterations. Relaxation factors are

defined in the fvSolution dictionary to improve stability of the computation by

limiting the amount of change in a variable from one iteration to the next. The final

product of the olaFoam case results in a wave passing by a floating structure in a

channel, Figure 11.

Structure Motion

The six-degrees of motion for the FB structure was defined in the dynamicMeshDict.

The 6DoF solver calculated the effect of fluid forces on the structure and adjusted

the FB mesh according to physics. Structure motion was calculated about the center

of mass, which is initially defined by the user then recalculated at each time step

during the motion calculations. The center of mass must be symmetrical between

processor boundaries during parallel processing to reduce unwanted boundary forces

on the structure, which create instability.

Figure 12: Dynamic FB 4 under wave forcing

45
The narrow space between the channel walls and the FB structure created an increase

in pressure using an unrestrained, 6DoF structure. The motion in the y-direction is

insignificant in the wave channel; therefore, constraints were added to eliminate

sway, yaw, and roll. Additionally, a spherical angular damper restraint reduced

instabilities in pitch motion, acting similarly to a mooring system. The angular

damper restraint was not changed during the dynamic cases, simulating a

systematical mooring arrangement.

3.6 Data Processing

The FB structure’s ability to attenuate wave energy is compared with respect to

structure dimensions. Simulated gauge readings in OpenFOAM quantified the wave

height using probes in the channel, generating a vector output of the water elevation

at the gauge location for each time step.

The resultant free surface data was numerically analyzed using a wave-by-wave

analysis to detect the significant wave heights at each location. The transmission

coefficient was determined by comparing the significant transmitted wave to the

significant incident wave. The breadth is non-dimensionalized using B/L and the

draft is non- dimensionalized using dr/d and each are plotted against the transmission

coefficient. The dynamic motion and static FB structures were also compared and

plotted against each other.

46
 Chapter 4: Results and Discussion

The breadth parameter and the draft parameter are compared to the transmission

coefficient, as well as a comparison between the numerical and physical testing

methods. The parameters were deemed satisfactory if the structure attenuated 50%

of the wave energy, having a transmission coefficient of 0.5. Additionally, the effects

of structure motion were presented and compared between the static and dynamic

cases. Raw results are provided in Appendix C.

Comprehensive testing over the range of non-dimensionalized draft and breadth

parameters is limited due to the weak coupling between the olaDyMFoam solver and

6DoF solver in OpenFOAM. Due to the instability of the numerical solver, both the

significant and second full wave heights were used to determine the transmission

coefficient for the dynamic cases.

4.1 Dynamic FB

The effects of the breadth parameter for a dynamic FB are shown in Figure 13. The

Root Mean Square Error (RMSE) for the dynamic FB with respect to breadth was

0.0722. The effects of the draft parameter on a dynamic FB are shown in Figure 14,

with an RMSE of 0.0114. A comparison between errors from previous studies, Table

4, were statistically analyzed and compiled by Bieshuvel, (2013) under similar

parameters. The numerical model predicted greater the wave attenuation compared

47
to the physical model for both the breadth and draft parameter testing; however, the

limited data available may have resulted in lower error values.

Figure 13: Breadth parameter for dynamic FB, significant wave height

Table 4: RMSE for the Dynamic FB tests

RMSE
Stehno, 2017 0.0114-0.0722

Macagno vs Peña 0.1694

Wiegel vs Peña 0.1606

Ursell vs Peña 0.2110

Kriebel and Bollmann vs Peña 0.2476

Roul et al. vs Peña 0.1570

48
The physical modeling for both the draft and breadth parameters did not satisfactorily

attenuate the significant wave climate. The mooring system for the physical dynamic

structures remained unchanged between each case; however, mooring systems

influence structure stability under wave action and must be further evaluated to

determine the effects of mooring line tension under wave forcing.

Figure 14: Draft parameter for dynamic FB, significant wave height

The results for the numerical dynamic cases were limited due to a weak coupling in

the numeric model between the motion solver and the two-phase wave solver. This

weak coupling created instabilities in the mesh before convergence, which created a

large divergence within one time step and “exploding” the results. The numerical

instability occurred at different time steps between each simulation; therefore, two

cases allowed for the development of a wave climate dataset before becoming

49
unstable. The significant wave height was found from datasets with a minimum of 7

full waves after the two initial building waves. In order to further understand the

capabilities of the dynamic FB structure effectiveness, the second wave in each data

set was used to compare the structure parameters to the transmission coefficient. This

method does not take into account a full wave spectrum, but rather a singular wave

transmitted past the FB at a constant time, Figure 15 and Figure 16. The second-wave

method was also applied to the dynamic wave tank results in order for comparison.

Figure 15: Breadth parameter for dynamic FB, second-wave height

The second wave was not in a fully developed wave climate; therefore, the incident

wave is smaller than the significant wave height. From the physical tests, the

transmission coefficient under the second-wave analysis was reduced except for

when B/L=0.1672, where the transmitted wave height was increased, possibly due to

the wave-structure interaction for a natural pitch period which is 0.3 seconds larger
50
than when B/L=0.3345. The second-wave analysis also reduced the transmission

coefficient in the numerical model. This is caused by the wave not being fully

developed at the time of measurement, and FB starting at rest.

Figure 16: Draft parameter for dynamic FB, second-wave height

The second-wave analysis provided more data to understand the performance of the

dynamic FB structure parameters. The FB draft and breadth parameters performed

as expected, where the transmission coefficient decreased as the structure dimension

lengths increased. The RMSE for the breadth parameter was 0.0534, and for 0.0452

the draft parameter, where the numerical model predicts a more effective structure.

The second-wave breadth parameter performed satisfactory when B/L=0.3345,

suggesting that the larger natural pitch period reduced the motion of the structure,

attenuating waves more effectively. The numerical model supported this finding,

51
showing that wave attenuation increased as the structure breadth increased. The

results from the numerical model are less pronounced due to the time required for

the model to generate a full wave climate.

The physical testing for the second-wave draft parameter generated satisfactory

results when dr/d=0.5 and 0.6667. As the draft increased, the structure effectiveness

increased; however, further testing is required to develop the wave attenuation

properties throughout the entire water column. The numerical testing, which only

was able to produce one result under the significant wave analysis, further proved

that the effectiveness of the structure increases as the draft increases, where all three

usable tests were considered satisfactory. Similar to the breadth parameter, the

numerical model results are less pronounced due to the spin up duration of wave

generation between models, where the physical model takes less time to produce a

regular wave than the numerical model.

Use of the second wave height method of analysis allow for the generation of a

comparable set of data to determine the effectiveness of dynamic FB structures.

Further improvements must be made on the model to increase the simulation time by

reducing instabilities. Additional physical testing would provide insight for the

mooring tension required to dampen the structure to achieve satisfactory wave

transmission.

52
4.2 Static FB

The results of the breadth parameter for a static FB are shown in Figure 17. The

numerical model tests result in a lower transmission coefficient that the physical

model tests, with an RMSE of 0.2193. The wave tank results for the static B/L testing

were considered unsatisfactory due to the large amount of wave transmission. While

only two cases were tested, the results confirm that the increase in breadth will

increase wave attenuation. The numeric model for the static B/L testing is considered

satisfactory for B/L greater than 0.1672. Interestingly, the transmission coefficient

increases for B/L=0.67, suggesting limitations to a static structure as a wave

attenuator.

Figure 17: Breadth parameter for static FB, significant wave height

The effects of the draft parameter for a static FB are shown in Figure 18. The

numerical modeled structures resulted in lower transmission coefficients than the

53
physical testing with an RMSE of 0.2624. Previous studies have presented smaller

error using a different numerical solution Table 5. As expected, the transmission

coefficient decreased with an increase of the draft parameter, verifying that an

increase in structure draft will allow the structure to better attenuate the wave energy.

While the physical testing only produced one satisfactory draft parameter, the results

confirmed that the structure’s effectiveness increases with an increase in draft.

During physical testing, the structure performance was considered satisfactory when

the dr/d was 0.6667. The numerical modeling results are all considered satisfactory.

The structure with zero draft attenuated more wave energy than the structures with a

dr/d between 0.1667 and 0.5, possibly due to a numerical error.

Figure 18: Draft parameter for static FB, significant wave height

54
 Table 5: RMSE for the Static FB tests

RMSE
Stehno, 2017 0.2193-0.2624

Koutandos et al., 2005 * 0.1500

Gesraha, 2006 * 0.1500 (irregular waves)

The second-wave analysis was applied to the static cases to further determine the

practical application of this method. The results from the static testing were

thorough; therefore, the second-wave analysis was not necessary to obtain desired

results. The RMSE for the breadth parameter was 0.0559 and 0.0452 for the draft

parameter under the second wave analysis; however, the effects of the static second-

wave analysis proved that this method reduced the transmission coefficient. Due to

the sufficient amount of collected data, further evaluation of the static second-wave

analysis is unnecessary for static cases.

Discrepancy between the numerical and physical model for the static cases may be

due to assumptions during the numerical calculations in order to reduce run time. An

average RMSE of the significant wave parameter tests resulted in approximately a

24% error, suggesting that the numeric model predicts a lower transmission

coefficient for the performance of a static FB. Previous studies have shown a RMSE

of 0.1500 between a similar semi-empirical model and physical testing in regular

waves, where theory produces a lower the transmission coefficient than physical

55
testing (Koutandos et al., 2005). Wave theory estimated a structure with dr/d of 0.5

to have a transmission coefficient of 0.5; therefore, the numerical model attenuated

greater energy and the physical model attenuated less energy when compared to the

wave theory approximation.

The design of a static floating breakwater in a wave climate would require a strong

mooring system or piling in order to remain truly stationary. Further testing is

required to understand the capabilities of the structure breadth with respect to

wavelength for a static case, as well as additional mooring design consideration.

4.3 Static FB vs. Dynamic FB

The static and dynamic cases were compared to understand the effect of structure

motion on wave attenuation. It is known that dynamic structures are not as effective

at attenuating wave energies as a static structure due to the structure movement. The

degree that the dynamic structure will attenuate a wave is based on the mooring line

tension. Only one mooring line tension was tested during for each the physical and

computational models and further studies are required to understand the significance

of mooring line tension under this specific design wave climate.

The breadth effects for both the dynamic FB and the static FB are shown in Figure

19, where significant wave height analysis is used for comparison. The physical

testing resulted in similar transmission coefficients for B/L=0.1672, which was

unexpected. The remainder of the results of the breadth parameter were as expected,

where the static structure reduced wave energies greater than the dynamic structure.
56
The computational model suggested a greater difference between the static and

dynamic structures compared to the physical model. This may be due to the

differences in mooring configurations between the physical and computational

models. It is interesting to note that the computational results produced a similar

slope for both the static and dynamic structures when B/L= 0.5-0.667, suggesting

that the structure’s effectiveness is linearly dependent on the mooring tension.

Further studies must be done to test this hypothesis.

Figure 19: Transmission coefficient for a FB with respect to breadth effects

The effects of the draft parameter on the transmission coefficient for both the static

and dynamic structures are shown in Figure 20. The dynamic and static results

performed similar for both dynamic and static cases. The physical model further

suggests that the wave attenuation properties are linearly dependent on the mooring

tension, where the results for the static and dynamic cases produce a similarly sloped
57
line. The computational model did not produce enough results to compare the static

and dynamic structures with respect to the draft parameter; however, the one dynamic

result further proves that the dynamic structure is not as effective at attenuating

waves as the static structure.

Figure 20: Transmission coefficient for a FB with respect to draft effects

58
 Chapter 5: Conclusions and

Recommendations

Temporary floating breakwaters would permit water flow while reducing wave

energy during the initial phases of restoration, allowing plants, such as mangroves,

marsh grasses, and seagrasses, to establish a secure root system and become

productive additions to the ecosystem. Floating breakwater designs were compared

for their efficiencies at attenuating a design IRL wave height of 0.2 m at a 2 s wave

period in a 0.6 m water depth. Based on previous studies, the ideal transmitted wave

height for living shoreline restoration is 0.1 m; therefore, the FB structure design is

required to attenuate 50% of the wave energy, or have a transmission coefficient of

0.5 for successful living shoreline restoration protection in the IRL. While a

transmission coefficient of 0.5 will adequately protect living shorelines from the

design wave, the IRL experiences wave heights greater than 0.2 m due to wave events

and non-shore normal waves; therefore, a transmission coefficient less than 0.5

would better protect restoration sites.

The suggested FB design is based on computational and physical testing using

industry methods to non-dimensionalized structure parameters, designing the FB to

effectively reduce wave energy in the IRL. The suggested dimensions of a FB for the

design IRL wave climate has at least draft of 0.4 m and a minimum breadth of 1 m.

59
These suggested dimensions require an increase in mooring line tension in order to

further dampen the structure under wave actions, where a completely static structure

will produce a transmission coefficient of 0.4194. The available results for the

dynamic structures did not successfully attenuate the wave enough to adequately

protect living shoreline restorations. Realistically, the mooring system of the static

structure is difficult to design without requiring permanent reinforcement, such as

pilings. The dynamic structure requires less mooring design; however, does not

attenuate wave energy as well as the static designs. These results suggest that a

transmission coefficient of 0.5 is able to be achieved with these structure dimensions

if mooring tension is increased.

The proposed structure dimensions are based on the current testing parameters. The

results indicates that an increase in either breadth or draft will further increase the

transmission coefficient. Previous studies and the computational theory support this

suggestion. Material availability restricted the physical tests; therefore, it is

recommended that a greater range of structure dimensions should be tested prior to

deployment.

Differences between the physical and numerical studies result from the variability of

the FIT wave channel, the wave analysis method, and assumptions used during

numerical testing. Further development of olaFoam is required in order to generate

a regular, progressive wave coupled with a dynamic, floating structure in order to

determine a correlation between the numerical and physical tests. Additional

60
research must be done to address the variability of wave climates in the IRL,

including irregular waves, waves propagating in a non-shore normal direction, and

the effects of overtopping on the FB structure.

61
 References

Adams, Amy, et al. "The Indian River Lagoon Comprehensive Conservation and

Management Plan." (1996)Print.

Adee, B. "Floating Breakwater Performance." Coastal Engineering Proceedings

1.15 (1976)Web.

Anthoni, J. Floor. 2000. "Oceanography: Waves." Accessed March 5, 2012.

www.seafriends.org.nz/oceano/waves.htm.

Bayram, Atilla. "Experimental Study of a Sloping Float Breakwater." Ocean

Engineering 27 (2000): 445-453. Print.

Brebner, A., Ofuya, A. "Floating Breakwaters." Coastal Engineering Proceedings

1.11 (1968)Web.

Carr, John H. , Mobile breakwater studies, Calif. Inst. of Tech. , Hydro. Lab., Report

No. N-64. 2, 54 pp., Dec. 1950 (unpublished).

62
Carver, R. D. Floating Breakwater Wave-Attenuation Tests for East Bay Marina,

Olympia Harbor, Washington; Hydraulic Model Investigation. HL-79-13

Vol. Vicksburg, Miss: US Army Engineer Waterways Experiment Station,

1979. Print.

Carver, Robert D., et al. Sloping Float Breakwater Study, Oregon Inlet, North

Carolina : Coastal Model Investigation. U.S. Army Engineer Waterways

Experiment Station, 1987. Print.

Chen, Zhi-jie, et al. "Time-Domain Hydrodynamic Analysis of Pontoon-Plate

Floating Breakwater." Water Science and Engineering 5 (2012): 291-303.

Print.

Choice ZD, Frazer TK, Jacoby CA. 2014. Light requirements of seagrasses

determined from historical records of light attenuation along the Gulf coast

of peninsular Florida. Marine Pollution Bulletin 81:94–102.

Dean, Robert, and Robert Dalrymple. Water Wave Mechanics for Engineers &

Scientists. 2 Vol. 1991. Print. Advanced Series on Ocean Engineering.

Dean, Robert, and Robert Dalrymple. Coastal Processes with Engineering

Applications, 2002. Print.

63
Dong, G. H., et al. "Experiments on Wave Transmission Coefficients of Floating

Breakwaters." Ocean Engineering 35.8–9 (2008): 931-8. Web.

Duchock, John, and Kevin Landry. "Twin Rivers Shoreline Restoration: Getting the

most Out of Florida's Publically Accessible Shorelines." (2006)Print.

Ekedahl, Ghassan, Rafic Younes, and Pascal Lafon. "Parametrical and Motion

Analysis of a Moored Rectangular Floating Breakwater." Journal of Offshore

Mechanics and Arctic Engineering 131 (2009).

Fonseca, MF, WJ Kenworthy, and GW Thayer. "Guidelines for the Conservation and

Restoration of Seagrasses in the United States and Adjacent Waters." NOAA

Coastal Ocean Program Decision Analyses Series.12 (1998)Print.

Fousert, M. W. "Floating Breakwater: Theoretical Study of a Dynamic Wave

Attenuating System." Masters TU Delft; Faculty of Civil Engineering and

Geosciences, section Hydraulic Engineering, 2007. Print.

Fugazza M, Natale L., “Energy losses and floating breakwater response.” Journal of

Waterway Port Coastal and Ocean Engineering, vol. 114, pp 191-205. (1998)

Galveston Bay Foundation. "Living Shorelines: A Natural Approach to Erosion

Control." (2011)Print.

64
Hales, L. Z. Floating Breakwaters: State-of-the-Art Literature Review. 81-1 Vol.

Fort Belvoir, VA: US Army Corps of Engineers Coastal Engineering

Research Center, 1981. Print.

Hazen and Sawyer Environmental Engineers & Scientists. “Indian River Lagoon

Economic Assessment and Analysis Update.” Prepared in conjunction with

the St. Johns River Water Management District for the Indian River Lagoon

National Estuary Program. August 2008.

Higuera Caubilla, Pablo. "Application of Computational Fluid Dynamics to Wave

Action on Structures." Universidad de Cantabria, 2015. Print.

Higuera, Pablo. Floating Object Help. E-mail.

Isaacson, M., and R. Byres. "Floating Breakwater Response to Wave Action."

Proceedings of the 21st Coastal Engineering Conference 3 (1988): 2189--

2200. Print.

Jasak, H., and Tukovic, Z. “Dynamic Mesh Handeling in Openfoam Applied to

Fluid-Structure Interaction Simulation.” V European Conference on

Computational Fluid Dynamics. 2010.

65
Johnson LK, Herren LW. “Re-establishment of fringing mangrove habitat in the

Indian River Lagoon Shoreline Restoration Project”. Florida Department of

Environmental Protection Office of Coastal and Aquatic Managed Areas.

(2008)

Jones, D.B., “Transportable Breakwaters- A survey of Concepts”, Technical Report

R-727, U.S. Navy Civil Engineering Laboratory, Port Hueneme, Calif., May

1971.

Knauss, John A. Introduction to Physical Oceanography. 2nd ed., 1997. Print.

Koch, E.W., Sanford, L.P., Chen, S.-N., Shafer, D.J., Smith, J.M., 2006. Waves in

seagrass systems: review and technical recommendations. U.S. Army Corps

of Engineers, Washington, DC

Kofoed, J. P., et al. "Prototype Testing of the Wave Energy Converter Wave

Dragon." Proceedings of the 8th World Renewable Energy Congress

(2004)Print.

Kolahdoozan, Morteza, Omidreza Taheri, and Hadi Ghasemi. "Hydrodynamic

Performance Evaluation of Step Floating Breakwaters through Experiment

and Artificial Intelligence." Web.

66
Koutandos, E., T. Karambas, and C. Koutitas. "Floating Breakwater Response to

Waves Action using a Boussinesq Model Coupled with a 2DV Elliptic

Solver." Journal of Waterway, Port, Coastal, and Ocean Engineering 130.5

(2004): 243-55. Web.

Koutandos, E., and P. Prinos. Design Formulae for Wave Transmission Behind

Floating Breakwaters. Greece: Aristotle University of Thessaloniki, 2005a.

Print.

Koutandos, E., P. Prinos, and X. Gironella. "Floating Breakwaters Under Regular

and Irregular Wave Forcing: Reflection and Transmission Characteristics."

Journal of Hydraulic Research 43.2 (2005b): 174-88. Web.

Knutson, P. L., Allen, H. H., and Webb, J. W. (1990). "Guidelines for vegetative

erosion control on wave-impacted coastal dredged material sites," Technical

Report D-90-13, U. S. Army Engineer Waterways Experiment Station,

Vicksburg, MS.

Mani, J. "Design of Y‐Frame Floating Breakwater." Journal of Waterway, Port,

Coastal, and Ocean Engineering 117.2 (1991): 105-19. Web.

McCartney, B. "Floating Breakwater Design." Journal of Waterway, Port, Coastal,

and Ocean Engineering 111.2 (1985): 304-18. Web.

67
Morris, L. J., and C. A. Jacoby R.C. Chamberlain. Summary Report for the Northern

Indian River Lagoon in L. Yarbro and P.R. Carlson, Eds. Seagrass Integrated

Mapping and Monitoring Report no. 2. Fish and Wildlife Research Institute

Technical Report TR-17B, St. Petersburg, Florida.Print.

Morrison, G., W. Sargent, and K. Madley. 2003b. Florida seagrass manager’s toolkit.

Report to the Florida Department of Environmental Protection, Florida

Coastal Management Program pursuant to National Oceanic and

Atmospheric Administration Award No. NA17OZ2330. FMRI File Code

F2361-02. Florida Fish and Wildlife Conservation Commission, Florida

Marine Research Institute, St. Petersburg, Florida

Murali, K., and J. Mani. "Performance of Cage Floating Breakwater." Journal of

Waterway, Port, Coastal, and Ocean Engineering 123.4 (1997): 172-9. Web.

NOAA National Ocean Service. Station SIPF1- Sebastian Inlet State Park, FL.,

2013-2014. Print.

NOAA National Ocean Service. Station TRDF1-8721604-Trident Pier, FL., 2005-

2013. Print.

Ofuya, A. O. On Floating Breakwaters. CE-60 Vol. Kingston, Ontario, Canada:

Queen's University, 1968. Print.

68
Oliver, J. G., et al. Floating Breakwaters A Practical Guide for Design and

Construction. 85th ed. 13 Vol. Brussels, Belgium: PIANC, 1994. Print.

Patrick, D. A. California University Berkley Institute of Engineering Resarch,

"Model Study of Amphibious Breakwaters." 1951.Web. /z-wcorg/.

Peña, E., J. Ferreras, and F. Sanchez-Tembleque. "Experimental Study on Wave

Transmission Coefficient, Mooring Lines and Module Connector Forces with

Different Designs of Floating Breakwaters." Ocean Engineering 38.10

(2011): 1150-60. Web.

Richey, Eugene P. Floating Breakwater Field Experience, West Coast. Fort Belvoir,

VA:, 1982. Print.

Riisgard H.U. (1988). Efficiency of particle retention and filtration rate in 6 species

of Northeast American bivalves. Marine Ecology Progress Series. 45:217-

223.

Ripken, John. "An Experimental Study of Flexible Floating Breakwater." St. Anthony

Falls Hydraulic Laboratory 31 (1960)Print.

Roland, Rebecca M., and Scott L. Douglass. "Estimating Wave Tolerance of Spartina

Alterniflora in Coastal Alabama." Journal of Coastal Research (2005): 453-

63. Web.

69
Sannasiraj, S., et al. (1998). “Mooring forces and motion responses of pontoon-type

floating breakwaters.” Ocean Eng., 25(1), 27–48.

Schmitt, K., et al. "Site-Specific and Integrated Adaption to Climate Change in the

Coastal Mangrove Zone of Soc Trang Province, Viet Nam." Journal Coastal

Conservation 17 (2013): 545-558. Print.

Sheremet, Alex, Uriah Gravois, and Miao Tian. "Boat-Wake Statistics at Jensen

Beach, Florida." Journal of Waterway, Port, Coastal, and Ocean

Engineering 139.4 (2012): 286-94. Web.

Smith, Ernest R., and Nicholas C. Kraus. Laboratory Study on Macro-Features of

Wave Breaking Over Bars and Artificial Reefs. CERC-90-12 Vol., 1990.

Print.

"Snazzy Maps." Web. <https://snazzymaps.com/>.

Tadayon, M. H., et al. "Effect of Geometric Dimensions on the Transmission

Coefficient of Floating Breakwaters." International Journal of Civil and

Structural Engineering 1.4 (2010): 775-81. Print.

Tan, Mark Dexter. "Wave Interaction of the H-Type Floating Breakwater." Bachelor

of Engineering Universiti Technologi Petronas, 2013. Print.Tronoh, Perak

70
Teh H M, and Ismail H. "Hydraulic Characteristics of a Stepped-Slope Floating

Breakwater." 4th International Conference on Energy and Environment 2013

(2013) Print.

Teh H.M., and N. I. Mohammed. "Wave Interactions with a Floating Breakwater".

Humanities, Science and Engineering (CHUSER), 2012 IEEE Colloquium

on. Web.

U.S. Army Corps of Engineers, and Coastal Engineering Research Center (U.S.).

Shore Protection Manual. 1 Vol. Washington D.C.: Vicksburg, Miss: Dept.

of the Army, Waterways Experiment Station, Corps of Engineers, Coastal

Engineering Research Center;, 1984. Print.

U.S. Army Corps of Engineers. Coastal Engineering Manual. 1110-2-1100 Vol.

Washington, D.C.: U.S. Army Corps of Engineers, 2002. Print.

Wang, H. Y., and Z. C. Sun. "Experimental Study of a Porous Floating Breakwater."

Ocean Engineering 37 (2010): 520-527. Print.

Williams, A. N., and A. G. Abul-Azm. "Dual Pontoon Floating Breakwater." Ocean

Engineering 24.5 (1997): 465-478. Print.

71
Yates, K.K., H. Greening, and G. Morrison. 2011. Integrating science and resource

management in Tampa Bay, Florida: U.S. Geological Survey Circular 1348.

U.S. Geological Survey.

Yoon, J., and Y-S Cho. "Experimental Research on Effective Floating Breakwater."

Journal of Coastal Research. 64 (2011): 631-635. Print.

Zidan, AR, et al. "Wave Interaction with Single and Twin Pontoons." Web.

72
 Appendix A: Wave Theory

Figure 9 shows wave characteristics that describe the movement of the wave.

Figure 21: Water wave parameters (Dean and Dalrymple, 1991)

in which:

L= wave length, the distance between two successive wave crests

η= free surface height from waterline

H= wave height

a= wave amplitude

h= water depth, which will be referred to as d

Other notable wave characteristics include the wave crest, which is the highest point

of the wave; the wave trough, which is the lowest point of the wave; and the wave

period, which is the time it takes for a complete wave cycle to pass a specific point.
73
Theoretical waves are found using linear wave theory. Dean and Dalrymple (1991)

describe the steps to find useful linear wave equations. The continuity equation

(Equation 22) uses incompressible fluid and a non-divergent, inviscid flow

assumptions. The hydrostatic pressure equation (Equation 23) describes the pressure

acted on a point due to fluid density and depth. Buoyancy force is a positive result of

the hydrostatic pressure acting on an object’s surface (Equation 24). Newton’s

second law derives the conservation of momentum (Equation 25) by assuming steady

flow and independence of time. The conservation of momentum finds surfaces forces

acting on a particle and is used to derive the Euler equations (Equation 26). The

velocity potential (Equation 27) allows the assumption of irrotational flow.

Assumptions are able to be made from these equations to find the Bernoulli equation

(Equation 28).

𝑑𝑢 𝑑𝑣 𝑑𝑤
+ + =0 Equation 22
𝑑𝑥 𝑑𝑦 𝑑𝑧

where u, v, and w are velocities in the x, y, and z directions, respectively.

𝑃 = −𝜌𝑔𝑧 Equation 23

𝐹𝑏𝑢𝑜𝑦𝑎𝑛𝑐𝑦 = 𝜌𝑔𝑉 Equation 24

where P is pressure, ρ is density, g is gravity, z is water depth from the surface,

Fbuoyancy is the buoyant force on the floating object, and V is volume of an object in

fluid.

74
 𝐷𝑢
∑ 𝐹𝑥 = 𝑚 Equation 25
𝐷𝑡

where Fx is the force in the X direction, m is the mass of the fluid, and Du/Dt is the

total acceleration of the fluid.

𝐷𝑢 1 𝑑𝑝 𝐷𝑣 1 𝑑𝑝 𝐷𝑤 1 𝑑𝑝
=− , =− , = − −𝑔 Equation 26
𝐷𝑡 𝜌 𝑑𝑥 𝐷𝑡 𝜌 𝑑𝑦 𝐷𝑡 𝜌 𝑑𝑧

𝑑𝜙 𝑑𝜙 𝑑𝜙
𝑢=− , 𝑣=− , 𝑤=− Equation 27
𝑑𝑥 𝑑𝑦 𝑑𝑧

where φ is the line integral of the velocities in the x, y, and z directions.

𝒖2 𝑝
+ + 𝑔𝑧 + 𝜙 = 𝐶(𝑡) Equation 28
2 𝜌

where C(t) is a constant.

Boundary conditions (BC) are assumptions applied to the Bernoulli equation,

resulting in equations for wave motion. Assuming an irrotational and incompressible

fluid, a velocity potential that satisfies the mass continuity is found (Equation 29)

where ∇ is the vector differential operator. The divergence of the flow u in Equation

28 leads to the Laplace Equation (Equation 30). Assumptions are made, resulting in

boundary conditions that restrict the fluid domain. The kinematic BC (Equation 31)

constrains fluid motion. The bottom BC (Equation 32) defines the fluid motion at the

boundary at the bottom of the domain. The kinematic free surface BC (Equation 33)

restricts the fluid from moving outside of the free surface, η. Gravity and the pressure

balance bound the wave creating the dynamic free surface BC (Equation 34), where

75
C is a constant function of time, neglecting tension forces. The lateral BC (Equation

35) allows the assumption for waves to propagate in only the x direction where L is

the wavelength and T is the wave period.

∇∙𝒖=0 Equation 29

𝑑2 𝜙 𝑑2 𝜙 𝑑2𝜙
∇2 𝜙 = + + =0 Equation 30
𝑑𝑥 2 𝑑𝑦 2 𝑑𝑧 2

𝑑𝐹
𝒖 ∙ 𝒏 = − 𝑑𝑡 Equation 31
|∇𝐹|

𝑤 𝑑𝑑
=− Equation 32
𝑢 𝑑𝑥

𝑑𝜂 𝑑𝜂 𝑑𝜂
𝑤= +𝑢 + 𝑣 𝑓𝑜𝑟 𝑧 = 𝜂𝑥,𝑦,𝑡 Equation 33
𝑑𝑡 𝑑𝑥 𝑑𝑦

𝑑𝜙 1 2 𝑃𝜂
− + (𝑢 + 𝑤 2 ) + + 𝑔𝑧 = 𝐶(𝑡) Equation 34
𝑑𝑡 2 𝜌

𝜙(𝑥, 𝑡) = 𝜙(𝑥 + 𝐿, 𝑡) , 𝜙(𝑥, 𝑡) = 𝜙(𝑥, 𝑡 + 𝑇) Equation 35

These equations combine to create the dispersion equation (Equation 36).

𝜎 2 = 𝑔𝑘 tanh(𝑘𝑑) Equation 36

where σ is the angular frequency, 𝜎 = 2𝜋/𝑇 and k is the wave number, 𝑘 = 2𝜋/𝐿.

Simplifying the dispersion equation results in a deep water approximation for the

wavelength (Equation 37) where Lo is the deep water wavelength.

𝑔𝑇 2
𝐿0 = Equation 37
2𝜋

76
where Lo is the deep water wave length. Wave celerity (Equation 38) is the speed

that the wave is moving.

𝐿 𝐿0
𝐶= = tanh(𝑘ℎ) Equation 38
𝑇 𝑇

where C is the wave celerity. The free surface displacement (Equation 39) bounds

the water surface created by waves, where H is the wave height, x is the direction

along the x axis, and t is the time as the wave passes. Wave height is a function of

outside forces acting on the fluid. Fetch and duration determine the wave growth

from wind. Fetch is the distance wind is able to travel across open water and duration

is the amount of time wind is able to travel across open water. Fully developed wave

height growth is limited by either the fetch or duration.

𝐻
𝜂= cos(𝑘𝑥 − 𝜎𝑡) Equation 39
2

The ratio of the wavelength to water depth defines whether the wave is in deep,

shallow, or intermediate water. Deep water waves have a depth greater than half of

the wavelength. Shallow water waves have a depth is less than one twentieth of the

wavelength. Intermediate water waves are found between shallow and deep water,

Figure 2.

77
 Appendix B: Floating Breakwater

Literature Review

Floating breakwaters have been used for over two centuries with the first recorded

FB being built in 1811 at Plymouth Port, England (Teh and Mohammed, 2012). The

Bombardon, built in 1944 for the Invasion of Normandy, created an artificial harbor

and sparked an interest in FB. A comparison of FB structure designs has been done

by Jones (1971), Hales and Natale (1985), and McCartney (1985).

Rectangular

The rectangular shaped breakwater is the simplest and most studied FB. The design

depth and breadth depend on the wave climate. Increasing the freeboard on the FB

minimizes overtopping. This structure relies on reflection from the vertical wall and

dampening from structure motion to reduce wave height.

Lab, field and numerical studies have resulted in different design sizes for the

rectangular FB. The breadth and draft lengths directly relate to the transmission

coefficient. Patrick (1951) was the first to test FB in the lab by comparing B/L to the

transmission coefficient using three drafts. Patrick proved that an increase in the FB

draft resulted in an increase of efficiency. Brebner and Ofuya (1968) tested a

rectangular FB using different drafts and freeboards comparing the transmission

78
coefficient to dr/d. Brebner and Ofuya determined the size of the FB is proportional

to its dampening effects. Isaacson and Byres (1988) tested rectangular FB models

both numerically and in the lab relating the transmission coefficient and heave

response. These results were compared with prototypes deployed in British

Columbia. Fugazza and Natale (1988) numerically and experimentally compared

structure motion to the transmission coefficient of a rectangular FB.

Kofoed et al. (2004) developed the Wave Dragon, reflecting waves off walls into a

chamber located on the top of the box. The Wave Dragon was used to generate

energy using a turbine. Koutandos et al. (2004) compared the motion response of a

fixed and heave motion rectangular FB using a numerical model. The numerical

model was validated by an experimental lab study. Koutandos et al. analyzed the

structure’s motion response due to wave interaction. They also found that the

structure breadth and draft are most important design parameters for a FB. Koutandos

and Prinos (2005a) validated the experimental design from Koutandos (2004) using

a semi-emperical model to find the wave reflection and transmission. Koutandos et

al. (2005b) compare the efficiency of a heave motion rectangular FB, fixed

rectangular FB, box FB with vertical plates, and a double pontoon FB. The heave

motion FB relies on dissipation to reduce the wave height while the fixed FB uses

reflection. Dong et al. (2008) used the rectangular FB and found that strong currents

reduce the FB performance. Elchahal et al. (2009) numerically calculated the

effectiveness and motion response of a rectangular FB.

79
Wang and Sun (2010) studied a porous rectangular FB, finding that the porous

material dissipates wave energy more effectively than a reflective wall. Taydon et al.

(2010) compared the rectangular FB to a box FB with vertical plates and determined

that a FB with a larger breadth is more effective than one with a smaller breadth.

Taydon et al. compared mooring arrangements and determined that mooring

arrangement increases the structure’s effectiveness. Yoon and Cho (2011) physically

and numerically correlated an increase of efficiency with an increase of draft. Yoon

and Cho also set up an array to show that multiple FB consecutively arranged reduce

long period waves more effectively than a single structure. Zidan et al. (2012) found

an empirical equation for estimating the transmission coefficient proving that an

increase in breadth and draft of the structure will increase the structures effectiveness.

Chen et al. (2012) numerically related the transmission coefficient and structure

motion response of a rectangular FB. Chen et al. determined that increasing the

structure breadth increases its effectiveness. Teh and Mohammed (2012) compared

a rectangular FB efficiency to the H-shaped FB.

Box FB with Plates

Studies from the rectangular FB showed that an increase in draft will increase the

structure effectiveness. Increasing the draft of a rectangular FB would require an

increase of cost and material. Reducing the draft and attaching plates underneath will

increase the effective draft without increasing the amount of material needed. There

are three different designs of the box FB with plates. These designs use either

80
reflection or dissipation to reduce the incident wave height. The first design is a

rectangular FB with two thin, vertical plates that extend below the box on both the

incident and transmissive sides of the structure, increasing the draft of the FB. The

plate on the seaward side of the structure uses reflection to reduce the wave height,

while the plate on the shoreward side is attached for stability. The second design adds

horizontal hanging plates under the box FB increasing dissipation and reflection.

These plates hang below the box, attached by wire. Wave energy is reflected by the

box, while the horizontal plates create turbulence below the surface. Dong et al.

designed a board-net FB by attaching rows of nets extending below a floating board.

While the nets are not plates, they increase the draft of the structure, reducing wave

energy.

Koutandos (2005b) concluded that a box FB with vertical plates has the same

transmission coefficient as a rectangular FB of equal draft. Taydon (2010)

determined that the box FB with vertical plates was more effective than a rectangular

FB. Taydon showed that the effectiveness of the structure increases when the plate

draft increases. Chen et al. (2012) numerically tested the box FB with one and two

horizontal plates and determined the design was more effective than the rectangular

FB because of dampening effects. Chen et al. found that the box with two plates was

more effective than the box with one plate, but the rate of effectiveness decreases

when the number of horizontal plates increase. Dong (2008) determined the breadth

81
and rigidity of the board-net FB affected the wave attenuation and the addition of the

net increased effectiveness of the structure.

Double Pontoon

A double pontoon FB is comprised of two rectangular FB connected by a rigid board,

creating a catamaran-like structure. Taydon et al. (2010), Ziden et al.(2012), and

Chen et al. (2012) noticed that a larger breadth of a structure will increase the

effectiveness of the rectangular FB. The double pontoon FB increases the effective

breadth without increasing material and costs.

Brebner and Ofuya (1968) compared the double pontoon, rectangular and A-frame

FB. They found that larger structures are more effective. They also noted that the

double pontoon’s large radius of gyration increased its effectiveness. Williams and

Abul-Azm (1996) numerically calculated the reflection coefficient of the double

pontoon FB. The structure effectiveness depends on the draft, spacing of the

pontoons, and the mooring line stiffness. Koutandos et al. (2005b) determined that a

porous plate is just as effective as a non-porous plate. Dong et al. (2008) compared

the breadth to the structure effectiveness for the double pontoon FB. Pena et al.

(2011) studied the wave response to a double pontoon FB finding that structure

rigidity increases effectiveness. Zidan et al. (2012) determined the double pontoon

FB was most effective when the spacing width is equal to the pontoon width.

82
Alaska

The Alaska FB is similar to the double pontoon FB; however there are multiple cross-

beams between the two pontoons, creating a ladder shape, increasing rigidity. The

structure uses reflection and dissipation for wave attenuation. The Alaska breakwater

was designed by the Department of Public Works, State of Alaska, in 1973 as a

cheap, easy way to manufacture FBs. It was installed in along Alaska and northern

United States pacific coasts.

Carver (1979) investigated a prototype Alaska FB structure in Olympia Harbor,

Washington by comparing the structure breadth to the transmission coefficient.

Carver determined that the wave period affects the transmission coefficient. Richey

(1982) reviewed field models deployed in Alaska, Washington, and Oregon. He

described the installation and performance of each structure. Richey noticed that the

weakest point in the design was the connections on the structure.

Y-Frame

The Y-Frame FB was designed by Mani (1991). The addition of PVC pipe extensions

on a trapezoidal box reduces cost and the transmission coefficient through wave

reflection, without compromising the motion of the structure. The breadth, draft, and

PVC diameter determine the effectiveness of the structure. Murali and Mani (1997)

designed a cage FB by rigidly connecting two Y-Frame FB structures together.

83
H-Shaped

The H-shaped FB was developed by Teh and Mohammed (2012) to reduce the

material used in a rectangular FB. The cross section is shaped as an H. Teh and

Mohammed found that the H design is more effective than a rectangular FB. Tan

(2013) tested a large scale model of the design and determined that the structure

becomes more effective as the draft is increased.

Stepped

The stepped FB was developed by Teh and Ismail (2013) by adding steps on the

incident and transmitted sides of a rectangular FB, creating a greater surface area for

the wave to interact with. This interaction creates reflection and reduces overtopping.

An increase in the structure’s breadth resulted in better wave attenuation. Teh and

Ismail also arranged two and three consecutive rows of stepped FB to compare the

arrangements effectiveness with a single row. Kolahdoozan et al. (2015) compared

the effects of the step width finding that the width is more prominent when the

breadth of the structure, the water depth, and the wave height increase.

Flexible-Membrane

The flexible membrane FB was created by Ripken (1960) and the U.S. Rubber

company to dissipate waves through turbulence and dampening. The horizontal wave

trap is comprised of two thick, flexible rubber blankets, one floats on top of the water

and the other floats below the surface. The subsurface blanket has one way valve

holes arranged in a staggered pattern, creating turbulence. The vertical wave trap
84
extends vertically through the water column and acts as a flexible, rubber wall. The

vertical blankets also have one way check valves to create turbulence. The structure

floats on top of the waves and dampened the waves as they passed by. Blanket

thickness and breadth both act to increase the efficiency of the wave blanket (Ripken

1960).

Sloping

The sloping FB is a long rectangular box that is ballasted at one end to incline the

structure. The structure dimension faces either towards or away from the incident

wave and is moored on the seaward side for wave dampening.

Patrick (1951) compared the effect of ballast and mooring for the sloping FB. 60%

ballast was the most effective, however, 87% ballast was also efficient and

transported seaward during the testing, reducing wave forcings on the mooring.

Carver (1987) compared configurations and moorings for a sloping FB. Bayram

(2000) numerically and experimentally tested the sloping FB to compare different

inclinations and mooring line lengths for the structure. He determined that the

transmission coefficient increases as the wave period decreases.

A-Frame

The A-frame FB uses reflection and structure motion to reduce wave energy. The

structure is comprised of two floating cylinders attached together by a horizontal

rigid board and two subsurface rigid boards connected into a V shape. A third rigid

85
vertical board also extends beneath the structure to increase the draft. Ofuya (1968)

connected structure effectiveness with design, wave and location characteristics, and

the radius of gyration. Brebner and Ofuya (1968) modified the radius of gyration by

adding a subsurface set of cylinders

Literature Review Conclusions

This history of floating breakwaters is impressive; however, the optimal design type

and dimensions have not been thoroughly studied due to the variations in wave

climates and locations for deployment. Separate studies have established the

efficiency of a wide variety of FB structure types, generally compared to the

rectangular structure type. A comprehensive study on FB structure type comparison

is missing, which may be due to the large assortment of designs and the constant

addition to the present collection. Individual studies currently investigate particular

designs, which may be more effective than the rectangular FB but are generally more

complicated to build; therefore, the tried and true rectangular structure will continue

to be the industry standard for floating breakwaters.

86
 Appendix C: Tables of Raw Results

Table 6: Draft parameter raw results

CFD Tank
Empty Static Dynamic Empty Static Dynamic
dr Hi Ht Ct Ht Ct Hi Ht Ct Ht Ct
Significant wave height

0.1243 0.1117
0 0.0387 0.3117
92

0.1 0.0618 0.4970

0.2 0.0492 0.3960 0.0876 0.7847 0.1038 0.9294
0.3 0.0448 0.3605 0.0719 0.5786 0.0636 0.5693 0.0727 0.6509
0.4 0.0386 0.3103 0.0468 0.4194 0.0659 0.5900
0.5 0.0366 0.2945
0.1455 0.1050
0 0.0335 0.2304 0.0404 0.2776
Second-wave

0.1 0.0571 0.3926

0.2 0.0444 0.3053 0.0656 0.6248 0.0790 0.7524
0.3 0.0406 0.2788 0.0490 0.3365 0.0597 0.5686 0.0480 0.4571
0.4 0.0338 0.2324 0.0404 0.2776 0.0502 0.4781 0.0204 0.1943
0.5 0.0303 0.2082 0.0359 0.2464
 Table 7: Breadth parameter raw results

CFD Tank
Empty Static Dynamic Empty Static Dynamic
B Hi Ht Ct Ht Ct Hi Ht Ct Ht Ct
0.1243 0.1117
Significant wave height

0.25 0.0778 0.6263

0.5 0.0650 0.5230 0.0840 0.7521 0.0844 0.7557
0.7 0.0549 0.4419
0.85 0.0485 0.3904 0.0772 0.6214
1 0.0448 0.3602 0.0719 0.5787 0.0636 0.5693 0.0727 0.6509
1.25 0.0438 0.3527
93

2 0.0526 0.4235
0.1455 0.1050
0.25 0.0599 0.4119
Second-wave

0.5 0.0515 0.3542 0.0711 0.6771 0.0973 0.9267

0.7 0.0471 0.3233 0.0492 0.3381
0.85 0.0434 0.2981 0.0503 0.3456
1 0.0406 0.2788 0.0490 0.3365 0.0416 0.3962 0.0480 0.4571
1.25 0.0390 0.2680 0.0481 0.3305
2 0.0394 0.2706