See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/262007507

Numerical simulation of interaction of a broken wave and a vertical

breakwater

Article  in  International Journal of Civil Engineering · March 2011

CITATIONS READS

13 188

2 authors:

Fatemeh Hajivalie Abbas Yeganeh-Bakhtiary

Ghent University Cardiff University
35 PUBLICATIONS   186 CITATIONS    98 PUBLICATIONS   1,093 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Performance Improvement of Oscillating Wave Surge Converters View project

Wind models View project

All content following this page was uploaded by Fatemeh Hajivalie on 31 July 2014.

The user has requested enhancement of the downloaded file.

 ISSN:1735-0522

A study of multilayer soil-fly ash

layered system under cyclic
loading 73
M. A. Khan, A. Usmani, S.S. Shah,
H. Abbas
Heat and contaminant transport
in unsaturated soil
H. Ghasemzadeh 90

INTERNATIONAL March

2 0 1 1
JOURNAL OF VOL.9

CIVIL ENGINEERING NO. 1

Dilation and particle breakage

effects on the shear strength of
calcareous sands based on
energy aspects
M. Hassanlourad, H. Salehzadeh,
H. Shahnazari

System dynamics approach for

construction risk analysis
F. Nasirzadeh, A. Afshar, M.
108

120
International Journal of Civil Engineering
Khanzadi

Fluid-structure interaction in
concrete cylindrical tanks under
harmonic excitations 132
K. Shahverdiani, A. R. Rahai, F.
Khoshnoudian

Assessment of conventional
nonlinear static procedures with
FEMA load distributions and
modal pushover analysis for 142
Iranian Society of Civil high-rise buildings
M. Poursha, F. Khoshnoudian, A.S.
Moghadam
Engineers
Iran University of Science
& Technology

Technical Note
Numerical simulation of the interaction of a broken wave and a
vertical breakwater
F. Hajivalie1, A. Yeganeh Bakhtiary2,*
Received: July 2009, Revised: August 2010, Accepted: September 2010

Abstract

In this paper, a two-dimensional Reynolds Averaged Navier-Stokes (RANS) model is developed to simulate the shoaling, breaking
and overtopping of a solitary wave over a vertical breakwater. Turbulence intensity is described by using a k- turbulence closure
model and the free surface configuration is tracked by Volume Of Fluid (VOF) technique. To validate the numerical model the
simulation results is compared with the Xie (1981) experimental data and a very good agreement between them is observed. The
results revealed that wave height and wave energy decrease considerably during the reflection from vertical wall, which
illustrates a considerable energy lost during the impaction and wave overtopping process. The turbulence production during the
broken wave interaction with vertical breakwater is very significant; consequently the vertical breakwater undergoes sever
turbulent and dynamic drag force.

Keywords: RANS model, shoaling, wave breaking, wave overtopping, k- turbulence closure model, VOF technique, solitary wave.

1. Introduction out that when the incident wave breaks, the broken wave
climbed up the face of breakwater and then fallen back;
Wave breaking in the nearshore zone forms a huge turbulent whereas little information about wave movement during the
bore propagating towards the coastal breakwaters and such process of breaking, impaction and reflection is presented.
violent flow may cause different types of breakwater failure. Sakakiyama and Liu [3] experimentally studied the free
Vertical breakwaters are preferred to rubble-mound surface displacements and velocity field in front of a caisson
breakwaters because the vertical breakwaters are more breakwater with wave-dissipating blocks, together with both
economic and take less time to construct. The vertical nonbreaking and breaking wave acting on a caisson
breakwater acts as a rigid structure and transmits most of the breakwater. They studied the generation of turbulence by
impinged broken wave forces to its foundation: it makes a wave-breaking and by the flow through the porous armor
vertical breakwater to become so sensitive to its foundation layer.
damages. Therefore, it is very important to study the broken The complexity involved in the wave interaction with vertical
wave interaction with a vertical breakwater. Most of the breakwaters has directed researchers to use the numerical
existing studies have been restricted to the interaction of simulation. Suh et al. [8] used a numerical model to compute
nonbreaking waves with the vertical breakwater. the reflection of regular and irregular waves from a partially
Xie [1] experimentally studied the interaction of standing perforated-wall caisson breakwater. Gislason et al. [10] studied
wave with a vertical breakwater and measured the the hydrodynamics of 2D laminar flow condition in front of a
distributions of maximum horizontal orbital velocities at the vertical breakwater. Hajivalie and Yeganeh-Bakhtiary [4]
wave node and antinode. Later on Gao and Inouchi [2] developed a numerical model based on Reynolds Averaged
investigated experimentally the broken wave impacts on a Navier Stokes (RANS) equations with a k- turbulence
vertical breakwater placed over a sloping bed. They pointed closuring model, to study the breakwater steepness effects on
the standing waves formation. The free surface configuration
* Corresponding Author:[email protected] was tracked by Volume Of Fluid (VOF) technique suggested
1 PhD Candidate, School of Civil Engineering, Iran University of by Hirt and Nichols [5]. Yeganeh-Bakhtiary et al. [6],
Science and Technology Narmak, Tehran 16844, IRAN employed this model to simulate the interaction process of the
2 Assistant Professor, Enviro-Hydroinformatics COE, School of
Civil Engineering, Iran University of Science and Technology wave overtopping with a vertical breakwater.
Narmak, Tehran 16844, IRAN On the other hand, solitary waves were usually utilized to

International Journal of Civil Engineering, Vol. 9, No. 1, March 2011 71

study the behavior of a broken wave approaching a shore. For
∂ε ∂ε ∂ε ∂ ⎡⎛ ν ⎞ ∂ε ⎤ ∂ ⎡⎛ ν ⎞ ∂ε ⎤ ε ε2
example Grilli et al. [7] numerically studied wave breakings + U + W = ⎢⎜⎜ν + t ⎟⎟ ⎥ + ⎢⎜⎜ν + t ⎟⎟ ⎥ + C1ε Pr − C2ε
process on different slopes by simulating the shoaling and ∂t ∂x ∂z ∂x ⎣⎝ σ ε ⎠ ∂x ⎦ ∂z ⎣⎝ σ ε ⎠ ∂z ⎦ k k
(5)
breaking of solitary waves on slopes from 1:100 to 1:8 using a ⎡ ⎧⎪⎛ ∂U ⎞ 2 ⎛ ∂W ⎞ 2 ⎫⎪ ⎛ ∂U ∂W ⎞ 2 ⎤
fully nonlinear wave model based on potential flow equations. pr = ν t ⎢2⎨⎜ ⎟ +⎜ ⎟ ⎬+⎜ + ⎟ ⎥ (6)
The wave characteristics at and beyond the breaking point is ⎣⎢ ⎪⎩⎝ ∂x ⎠ ⎝ ∂z ⎠ ⎪⎭ ⎝ ∂z ∂x ⎠ ⎦⎥
studied and a breaking criterion based on the nondimensional
k2
slope parameter is derived. Chang et al. [8] experimentally ν t = Cμ , Γ = ν +ν t (7)
studied the run up flow and related pressure of solitary waves ε
on a 1:20 beach; they measured swash flow velocity and where U and W are the mean velocity components of flow in
compared it with the existing analytical solution. They x (stream-wise) and z (vertical) direction, respectively; P is the
observed extra pressure exerted by the impact of swash flow. mean pressure; g is the acceleration of gravity;  is the fluid
Lin and Karunarathna [9], Ryu et al. [10] and Hsiao and Lin density;  and t are respectively the kinematic and kinetic
[11] studied the interaction of solitary waves with different eddy viscosity; k is the turbulence kinetic energy; pr is the
type of breakwaters and seawall. production of turbulence kinetic energy;  is the turbulence
Although there were several experimental and numerical dissipation rate. The model constants were set according to
studies about the interaction of nonbreaking wave and vertical Launder and Spalding [12] presented in Table 1.
breakwaters and also about breaking and overtopping of
solitary waves over coastal structures, there has been little 2.2. Initial and Boundary Condition
study about the interaction of a broken wave with a vertical The initial flow motion was specified with the zero mean
breakwater. Thus, in this study, we investigated numerically velocities and hydrostatic pressure. The initial conditions of
the ensuing hydrodynamic processes involved in the the turbulence field were set according to Lin [13] as follows:
interaction effect of a broken wave with a vertical breakwater.
For this purpose, an extended version of the pervious model 1 (8)
k = ut2 , ut = δ ci ͕ δ = 0.0025
of Hajivalie et al. [4] with introducing the initial and 2
boundary conditions corresponding to broken wave in front of
k2
vertical breakwaters were employed. Finite difference ε = Cμ , ν t = ζν ͕ ζ = 0.1 ͕ Cμ = 0.09 (9)
solutions to the governing equations were obtained, while the νt
free surface is tracked by using the VOF technique. Below, where ci is the wave celerity at the inlet boundary, C is the
Section 2 gave the mathematical framework for the model empirical coefficient suggested by Rodi [14],  and  are the
while Section 3 discusses the numerical model. In Section 4, model's constants.
the numerical results were presented and conclusions were Boundary conditions at the free surface, bottom, inlet and
drawn in Section 5. outlet were defined in the following. The free surface motion
was tracked by the VOF technique [5], which satisfies both the
2. Mathematical Formulation kinematic and dynamic free surface conditions, the
conservation of F or the volume fraction of fluid with the
In this section, the equations governing the flow and donor-acceptor algorithm and was expressed as follows:
turbulence fields are presented. Thereafter, the initial and
boundary conditions were discussed. The RANS equations, ∂F ∂F ∂F (10)
+U +W =0
which describe the mean flow field, in closure with a k- ∂t ∂x ∂z
turbulence model are applied to simulate the turbulent flow in The flow turbulence was not spread to free surface and
front of a vertical breakwater. therefore the vertical flux of k and  on the free surface should
reach to zero:
2.1. Governing equations
The governing equations consisted of the continuity, ∂k ∂ε (11)
=0 , =0
momentum and the k- equations were introduced in two- ∂n ∂n
dimensional coordinates as follow: here n is the normal vector. At the bottom, the wall function
was used to account for the rough bed condition. Based on the
∂U ∂W (1) numerical analysis, the gradient of the averaged velocity was
+ =0
∂x ∂z expressed as follows:
∂U ∂U ∂U 1 ∂P ∂ ⎛ ∂U ⎞ ∂ ⎛ ⎛ ∂U ∂W ⎞ ⎞ (2) ∂U u ∗ (12)
+U +W =− + ⎜ 2Γ ⎟ + ⎜ Γ ⎜ + ⎟ ⎟ =
∂t ∂x ∂z ρ ∂x ∂x ⎝ ∂x ⎠ ∂z ⎜⎝ ⎝ ∂z ∂x ⎠ ⎟⎠ ∂z κz

∂W ∂W ∂W 1 ∂P ∂ ⎧ ⎛ ∂U ∂ W ⎞ ⎫ ∂ ⎛ ∂ W ⎞ Table 4. The constants in k- turbulence model [12]

+U +W =− − g + ⎨Γ ⎜ + ⎟ ⎬ + ⎜ 2Γ ⎟ (3)
∂t ∂x ∂z ρ ∂z ∂x ⎩ ⎝ ∂z ∂x ⎠⎭ ∂z ⎝ ∂z ⎠
Cμ C1ε C2ε σk σε
∂k ∂k ∂k ∂ ⎡⎛ ν ⎞ ∂k ⎤ ∂ ⎡⎛ ν ⎞ ∂k ⎤
+ U + W = ⎢⎜⎜ν + t ⎟⎟ ⎥ + ⎢⎜⎜ν + t ⎟⎟ ⎥ + pr − ε (4) 0.09 1.44 1.92 1.0 1.3
∂t ∂x ∂z ∂x ⎣⎝ σ k ⎠ ∂x ⎦ ∂z ⎣⎝ σ k ⎠ ∂z ⎦

72 F. Hajivalie, A. Yeganeh Bakhtiary

 here z is the distance from the bed, u* is the shear velocity where Hs is the significant incident wave height,  is the
;с τ w ρ Ϳ͕ τ w
is the shear stress in the flow direction and  is the factor of incident wave angel (=1.0), s is the factor of
von-Karman constant (=0.41). If the above equation is permeability (=1.0) and Rc is the breakwater freeboard.
integrated, the logarithmic velocity distribution at the
boundary layer is obtained: 3. Numerical Model
U 1 ⎛ u z⎞
= ln⎜ E ∗ ⎟ (13) The governing equations were the two-dimensional
u∗ κ ⎝ ν ⎠ momentum and continuity, volume of fluid change function and
where E is a constant value (y9.0, [14]). Using the final k- equations. The governing equations were discretized in an
value of shear velocity, u* the boundary values of k and  at Eulerian staggered grid with rectangular cells. Scalar variables
the bed can be defined as such as pressure and density were calculated from the ordinary
u 3∗ nodes and the velocity components were calculated from the
ε= (14) staggered grid around the cells. Thus, the discretized continuity
⎡ ⎛ u z ⎞⎤
κy ⎢1 − exp⎜⎜ − ∗ ⎟⎟⎥ equation was converted to a pressure correction equation that
⎣ ⎝ 26ν t ⎠⎦ was defined on the scalar control volumes. Figure 1 illustrated
the staggered grid presented in the x-y plane of the Cartesian
⎡ ⎛ u ∗ z ⎞⎤
ν t = κ u ∗ y ⎢1 − exp⎜ − ⎟⎥ (15) coordinate system. The two-dimensional equations were
⎣ ⎝ 26ν ⎠⎦ discretized by finite-difference methods. In summary, the
u∗2 calculation procedure was as follows:
k= (16) • Using the momentum equations and the respective values of
Cd
time step n, the velocity components of time step n+1 were
In calculating the rate of turbulence dissipation a descending approximated.
function was used. Implementing this function lead to a nearly •Since the velocity obtained by the momentum equations
exact value of for the low turbulence condition, which is might not satisfy the continuity equation, the pressure in every
acceptable in case of the laminar sub-layer, while the cell should change so that the new velocities calculated from
turbulence production and its dissipation is only due to the bed the pressure variations satisfy the continuity equation. This
friction. At the inlet and outlet boundaries Neumann was done by iteration.
continuative boundary condition were assumed. •Using the k- equations, their values were calculated. Then,
k1, j = k2, j , ki , j = ki , j implementing the new values, the turbulent viscosity was
max −1 max − 2
(17) calculated and used in the next step.
ε1, j = ε 2, j , ε i , j = ε i , j
max −1 max − 2 •The value of the function F for each cell was defined at time
At the inlet boundary, the generating-absorbing boundary step n+1. Therefore the new position of the fluid was defined.
condition, introduced by Petit et al. [16] was implemented. By •(5) After calculating the time step size, dt, the next step of
imposing this condition, the incident waves can be generated calculations was performed at time t+dt,
while the reflected waves from the vertical breakwater were
absorbed simultaneously. This prohibits the collision
(intermixing) of the generated and reflected waves that
propagated in opposite direction of the interested zone:
∂Rr ∂R
+ Cr r = 0 (18)
∂t ∂x
Rr=Rt -Rin (19)
where Cr is the celerity of reflected wave, Rr, Rt and Rin are
the variable associated with the reflected, computed and the
theoretical wave values represents velocity, pressure, and free
surface displacement and Rin is the calculated by the Airy or
Stokes second order theory.
The outlet consisted of two different kinds of boundaries: (i) for
solid boundary of vertical breakwater, turbulence boundary
condition was defined; and (ii) over the breakwater, open boundary
condition was defined to allow wave overtopping as follows:
∂u ∂u (20)
+c = qo
∂t ∂x
here qo is the overtopping discharge, computed with Franco
and Franco) [17] overtopping formula:
⎛ ⎛ − 3Rc 1 ⎞ ⎞
qo = gH s × ⎜ 0.082 exp⎜ ⎟⎟
3
.
⎜ H γ γ ⎟⎟
(21) fig. 1. Grid geometry, control volumes of scalar and vector
⎜
⎝ ⎝ s β s ⎠⎠ quantities and location of variables in the x-y plane

International Journal of Civil Engineering, Vol. 9, No. 1, March 2011 73

3.1. Stability of the numerical scheme 4. Results and Discussion
Stability should be controlled mainly by the value of dt: its
value was consider in the following [15]: 4.1. Model validation
•The fluid was not allowed to travel more than one To assess the accuracy of the numerical model we should
computational cell in each time step: compare the numerical result with experimental data of
⎛ Δx Δy ⎞ interaction between waves and vertical breakwater. As it was
Δt ≤ Min⎜ , ⎟ (24) mentioned before to the best knowledge of the authors there is
⎜ Ux Uy ⎟
⎝ ⎠ no experimental data available on the study of interaction of
•Surface waves was not allowed to travel more than one cell breaking and/or broken waves and vertical breakwater, therefore
in each time step: Xie's [1] experimental data for nonbreaking wave has been
⎛ Δx ⎞ employed. In the experiments, nonbreaking waves were used to
Δt ≤ Min⎜ ⎟ (25) develop standing waves in front of a vertical breakwater. Fig. 2
⎜ ⎟
⎝ ghmax ⎠ depicts the experimental set up of vertical breakwater by Xie
where hmax is the maximum depth of flow. [1]. The experiment was conducted in a 38 m long, 0.8 m wide
•The relative variation of k and in a time step should be and 0.6 m deep wave flume. The water depth was equal to 0.45
significantly less than unity: m in the beginning of the flume and reached to 0.3 at the flat bed
near the breakwater with a 1:30 slope.
⎛k⎞
Δt ≤ Min⎜ ⎟ (26) Three different tests of Xie [1] experiments were simulated
⎝ε ⎠ and the characteristics of these tests was summarizes in Table
⎛ 1 k⎞ 2. In the numerical simulations, the computational domain has
Δt ≤ Min⎜⎜ ⎟⎟ (27) total length of 14.5 m and a height of 0.7 m, the length of
⎝ C2 ε ε ⎠ computational domain is shortened than the experimental
•The size of time step was adjusted to satisfy the diffusion flume to reduce the computation times (see Fig. 3). At the inlet
stability condition: boundary, Airy waves theory was employed to generate the
⎛ Δx 2 Δy 2 ⎞ incident waves into computational domain (see Chapter 2.2).
Δt ≤ Min⎜⎜ ⎟
⎝ 2ν e (Δx + Δy ) ⎠
2 2 ⎟ (28) Incident waves started to propagate into the numerical domain
at beginning of the simulation. The interferences of incident
here e =  + t waves impinged on a vertical breakwater and the reflected

Fig. 2. The sketch of physical model [1]

Fig. 3. Sketch of numerical domain

74 F. Hajivalie, A. Yeganeh Bakhtiary

 Table 4. The constants in k- turbulence model [12]

The incident wave Duration of numerical

H(m) T(s) d(m) L(m)
characteristics simulation (s)
Test No. 1 0.05 2.41 0.45-0.3 4.00 24.1

Test No. 2 0.065 1.53 0.45-0.3 2.40 15.3

Test No. 3 0.06 1.86 0.45-0.3 6.00 18.6

waves from it produce the standing waves in front of vertical the Hsiao and Lin [11] experimental data to validate the
breakwater. The flow velocities at nodes and antinodes then turbulence field of the model. Hsiao and Lin [11] experiments
can be observed and compared with that of the experimental were carried out in a wave flume, 22 X 0.5 X 0:75 m. Solitary
one. Xie [1] measured maximum horizontal velocity in two wave broken on a 1:20 sloping beach with 3.6 m length, the
sections, near the first node of standing waves and halfway of incident wave height and water depth were 0.07 m and 0.2 m,
node and antinode; the same velocities computed by numerical respectively. Fig. 5 shows both the result of the simulation
model were compared with the experiments in Fig. 4. As seen model on the turbulence energy and that presented by Hsiao
from the figure there is a very good agreement between the and Lin [11] at the breaking point. The solitary wave was
numerical results and experimental data. The figure also breaking as a plunging breaker with generating a considerable
indicated that the difference between the numerical results and turbulent energy at almost the same point. The maximum
experimental data increase with the increasing of H/d. turbulent energy generated from the simulation model was at
To best knowledge of the authors, the measuring turbulence the approximately same order of experiments and the trend of
parameters for broken waves in front of vertical breakwater has the generated turbulent kinetic energy is very similar in both
not been available yet. It is ineluctably used one of the cases of cases. The figure indicates that the model describes the
turbulence generation with a very satisfactory accuracy.

4.2. Wave field and transformation

To simulate the interaction of a broken wave and a vertical

breakwater the previous numerical domain has been employed.
To simulate the broken wave, a 0.21 m height solitary wave with
the wave celerity of 2.67 m/s was generated in the numerical
domain. The vertical breakwater was 0.65 m high and 1.0 m
wide. Fig. 6 shows the free surface configuration during the
numerical simulation. At the beginning the solitary wave

Fig. 4. Comparison of numerical model results and Xie's Fig. 5. Distribution of turbulent energy at breaking point; A)
experimental data, tests No. (1,2,3), continues line: numerical experimental data of Hsiao and Lin [11]
results, circles: Xie [1] experimental data and B) the simulation result

International Journal of Civil Engineering, Vol. 9, No. 1, March 2011 75

progressed into the numerical domain. As it reach over the slope, m/s in some instance. At this condition, the vertical breakwater
due the shoaling effect the wave height started to increase till it experienced a huge dynamic drag force. Whereas, Fig. 8-b
reached to a breaking height equal to 0.24 m (which fulfill the indicated that the breakwater foundation faced a high speed
McCowan breaking criteria), at this stage a plunging breaker was undertow flow during the reflection stage, which may
observed and the broken wave impact to the vertical breakwater. influence the breakwater stability.
A portion of flow body overtopped on the vertical breakwater
and the rest of flow body reflected from the breakwater. The 4.3. Turbulence Field
maximum observed overtopping discharge was about 0.14 (m3/s) Fig. 9 shows the quantity of k and  during the numerical
per length unit. The reflected wave height was 0.15 m, about simulation. As seen from the figure, before the broken wave
70% of the initial wave height that demonstrates noticeable reached to the bed slope, the maximum amount of k and  were
energy dissipation during the breaking and overtopping process. observed near the bottom boundary layer. However, during in
Fig. 7 shows the variation of horizontal and vertical the broken wave transform over sloped bed, the amount of
velocities. Fig.7-a indicated that wave crest has the maximum turbulence kinetic energy increases significantly in the vicinity
horizontal velocity, which increase with increasing of wave of free surface. On the other hand as the wave height started to
height during the shoaling effect. It was also indicated that the increase due the shoaling effect, turbulence energy and its
impinging velocity - with maximum of 1.0 m/s- was larger dissipation rate increased at the wave surface, while it reached
than reflecting velocity -with maximum of 0.5 m/s, which to its maximum at the breaking point. when the broken wave
shows that the energy lost happened during the wave impacted to the vertical breakwater, due the huge change in
impinging and overtopping the vertical breakwater.. Fig. 7-b velocity orientation and quantity because of wave reflection
shows that at the front of progressive wave, the vertical and overtopping, the turbulence intensity increased at the
velocity was upward at the front of the vertical wall and it was vicinity of the breakwater, consequently the breakwater crown
downward at the back side of wave crest. It was also indicated experiences a violent turbulent condition when the jet flow
that the vertical velocity increased during the shoaling effect overtopped the breakwater while lee side breakwater wall faced
and it was noticeably decreased in the reflected wave. the turbulent produced by wave reflection.
The particle velocity of broken wave near the breakwater at It was also observed that the order of maximum amount of k
impact and reflection stages from the vertical breakwater is and  at the breaking wave surface was about two orders more
shown in Fig. 8. The figure indicated that during the wave than the order of the maximum of these parameters at the
overtopping stage, a strong jet flow transfer over the crown of bottom boundary layer but it was almost the same order as the
vertical breakwater, namely the jet velocity was more than 2 overtopping jet flow.

Fig. 6. Snapshots of breaking wave development in front of vertical breakwater

76 F. Hajivalie, A. Yeganeh Bakhtiary

 (a) (b)

Fig. 7. Variation of a) U and b) V during shoaling, wave breaking and wave overtopping

Fig. 8. Snapshots of velocity vectors since wave impaction and overtopping over vertical breakwater

International Journal of Civil Engineering, Vol. 9, No. 1, March 2011 77

 Fig. 9. Variation of a) k and b)  during shoaling, wave breaking and wave overtopping

5. Conclusion experiment was performed similar to one of the cases reported

by Hsiao and Lin [11]. From this numerical study, the
A 2D numerical model based on RANS equation was been following conclusions were drawn:
developed to study the interaction between a broken wave with •Both wave height and velocity decrease considerably after
a vertical breakwater. The wave surface configuration was wave reflection, this indicates that wave energy has been
traced by the VOF method and a k- model was closured to dissipated during the impaction and overtopping process.
estimate the turbulence intensity. The model was not validated •Turbulence intensity at the vicinity of breakwater and in the
against the experimental data of broken waves in front of violent jet flow over the breakwater has the same order as
vertical breakwaters because the suitable data set has not been breaking point. It means that the interaction of broken wave
available currently. However, to demonstrate the model and the breakwater has a comparable role in turbulence
capacity the numerical model was validated using the Xie [1] production with breaking process.
experimental data for nonbreaking wave interaction with •Breakwater crown experiences significant turbulence
vertical breakwater and for the turbulence field, a numerical condition and dynamic drag forces.

78 F. Hajivalie, A. Yeganeh Bakhtiary

References slope. Coastal Engineering 56:1022-1034.
[9] Lin P, Karunarathna ASA. 2007. Numerical study of solitary
[1] Xie S. L. 1981. Scouring pattern in front of vertical breakwaters wave interaction with porous breakwaters. Journal of Waterway
and their influence on the stability of the foundation of the Port Coastal and Ocean Engineering-Asce 133:352-363.
breakwaters. Delft University of Technology, Delft, pp. 61. [10] Ryu YG, Chang KA, Mercier R. 2007. Runup and green water
[2] Gao X, Inouchi K. 1998. The characteristics of scouring and velocities due to breaking wave impinging and overtopping.
deposition in front of vertical breakwaters by broken clapotis. Experiments in Fluids 43:555-567.
Coastal Engineering Journal 40. [11] Hsiao SC, Lin TC. 2010. Tsunami-like solitary waves impinging
[3] Sakakiyama T, Liu PLF. 2001. Laboratory experiments for wave and overtopping an impermeable seawall: Experiment and
motions and turbulence flows in front of a breakwater. Coastal RANS modeling. Coastal Engineering, 57:1-18.
Engineering 44:117-139. [12] Launder BE, Spalding DB. 1971. The numerical computation of
[4] Hajivalie F, Yeganeh-Bakhtiary A. 2009. Numerical Study of turbulent flow. Journal of Computational Mechanic in
Breakwater Steepness Effect on the Hydrodynamics of Standing Application of Mechanic and Engineering 3.
Waves and Steady Streaming. Journal of Coastal Research SI [13] Lin P. 1998. Numerical modeling of breaking waves. Cornell
51:514-518. University.
[5] Hirt C, Nichols BD. 1981. Volume of fluid (VOF) method for [14] Rodi W. 1980. Turbulence model and their application in
the dynamics of free boundaries. Journal of Computational hydraulic-A state-of-the-art review. IAHR.
Physics 39. [15] Lemos CM. 1992. A simple numerical technique for turbulent
[6] Yeganeh-Bakhtiary A, Hajivalie F, Hashemi Javan A. 2010. flows with free surfaces. International Journal for Numerical
Steady Streaming and Flow Turbulence in front of Vertical Methods in Fluids 15:127-146.
Breakwater with Wave Overtopping. Applied Ocean Research: [16] Petit HAH, van Gent MRA, van Den Boscj P. 1994. Numerical
32; 91-102. simulation and validation of plunging breakers using a 2D
[7] Grilli ST, Svendsen IA, Subramanya R. 1997. Breaking Navier- Stokes model. 24th International Conference on Coastal
criterion and characteristics for solitary waves on slopes. Engineering, Kobe, Japan, pp. 511-524.
Journal of Waterway Port Coastal and Ocean Engineering-Asce [17] Franco C, Franco L. 1999. Overtopping formulas for caisson
123:102-112. breakwaters with nonbreaking 3D waves. Journal of Waterway
[8] Chang YH, Hwang KS, Hwung HH. 2009. Large-scale Port Coastal and Ocean Engineering-Asce 125:98-108.
laboratory measurements of solitary wave inundation on a 1:20

International Journal of Civil Engineering, Vol. 9, No. 1, March 2011 79

View publication stats