Response Mitigation of Floating

Downloaded from http://asmedigitalcollection.asme.org/offshoremechanics/article-pdf/145/5/051203/7013652/omae_145_5_051203.pdf by University College Dublin user on 26 July 2023

Wen-Huai Tsao1
Department of Civil and Environmental
Engineering,
Louisiana State University,
Baton Rouge, LA 70803
e-mail: whtsao@lsu.edu
Ying-Chuan Chen
KBR,
Houston, TX 77002
e-mail: ycchen@utexas.edu
Christopher E. Kees
Department of Civil and Environmental
Engineering,
Louisiana State University,
Baton Rouge, LA 70803
e-mail: cekees@lsu.edu
Lance Manuel
Department of Civil, Architectural and
Environmental Engineering,
The University of Texas at Austin,
Austin, TX 78712
e-mail: lmanuel@mail.utexas.edu
Platform by Porous-Media-Tuned
Liquid Dampers
A porous-media-tuned liquid damper (PMTLD) can serve as an effective and economical
dynamic vibration absorber. Placing porous media within a water tank can improve the
capacity for energy dissipation and optimize the performance by varying its material prop-
erties. Two numerical models are adopted to simulate the sloshing problem in PMTLD and
the dynamics of a floating platform in waves. Besides, the effectiveness of response mitiga-
tion can be verified numerically. The first potential-based approach employs a mixed-type
boundary value problem (BVP) solver and a free-surface particle tracker. This approach
not only simulates the inviscid water wave but also includes the nonlinear damping of
the PMTLD via a quadratic Forchheimer term. Another equivalent mechanical model is
used to reduce the degree-of-freedom of the PMTLD system. The Newmark method is incor-
porated to solve the rigid-body dynamics. The second viscous approach uses the finite
element method (FEM) to spatially discretize the Navier–Stokes (NS) equations and
handles the free surface via the volume of fluid (VOF) and the level set (LS) equations.
The multiphase simulation is implemented by computational modeling toolkits, Proteus
and Chrono, for the fluid and solid phases, respectively. The correlations between potential
flow and two-phase NS models are presented. The PMTLD is designed by analogy with the
tuned mass damper (TMD). Numerical results show that the PMTLD can effectively reduce
the structure’s dynamic response in terms of vibration amplitude around resonance. Such
damping devices have great potential for offshore platforms and wind turbine design.
[DOI: 10.1115/1.4062292]

Keywords: porous media, tuned liquid damper, equivalent mechanical model,

computational fluid dynamics, fluid–structure interaction, floating platform, dynamics of
structures

1 Introduction production systems (FPS), and wind turbines, have special

dynamic characteristics, so different devices are applied. For the
In the new era of renewable energy, the need for economical and
ships, the rolling motion is the biggest concern because the
clean resources from the ocean grows rapidly. Various kinds of off-
damping from the ambient fluid is usually too small to constrain
shore platforms have been created and served as the work station for
its amplitude, especially when resonance occurs. A common solu-
petroleum extraction, wind turbines, wave energy converters, and
tion is to equip the bilge keels on the hull to increase the hydrody-
offshore fishery. In the ocean environment, they are subject to
namic resistance and damping from the vortex [1]. An anti-rolling
many kinds of loading such as wind, wave, current, ice, and earth-
tank (ART), which consists of two vertical ducts connected at the
quake. These disturbances will result in tremendous vibration that
bottom with a horizontal duct, distributed along the centerline of
harms not only the seakeeping performance but also the longevity,
the ship is another effective damper for rolling reduction. Such
even breaking the integrity. Due to the complexity of the floating
liquid damper is also called tuned liquid column damper (TLCD)
body dynamics, the coupling effect brings unpredictable motions
in structural engineering. Their active control laws have been imple-
to the entire system. A single type of vibration absorber does not
mented by installing propellers or pumps in the middle of the
suit all purposes. Therefore, vibration control technology for off-
bottom tank [2–4]. Floating stations, such as fixed platforms, com-
shore structures remains challenging nowadays.
pliant towers, moored tension-leg platforms, and floating produc-
Typical floating structures, such as liquefied natural gas (LNG)
tion systems, have been equipped with dampers to reduce
vessels, floating storage regasification units (FSRU), floating
vibrations. TMD is one of the earliest solutions. It is an auxiliary
mass attached to the main structure through springs and dashpot ele-
ments. It generates an out-of-phase control force against the external
1
Corresponding author. excitations and dampens the kinetic energy of the main structure.
Contributed by the Ocean, Offshore, and Arctic Engineering Division of ASME for Once it is tuned to the resonant frequency, its performance can be
publication in the JOURNAL OF OFFSHORE MECHANICS AND ARCTIC ENGINEERING.
Manuscript received July 5, 2022; final manuscript received March 25, 2023; published
maximized. Gattulli and Ghanem [5] applied an active mass
online May 19, 2023. Assoc. Editor: Zhen Gao. damper to reduce the vortex-induced vibration of offshore

Journal of Offshore Mechanics and Arctic Engineering OCTOBER 2023, Vol. 145 / 051203-1
Copyright © 2023 by ASME
structures. Kawano [6] investigated the optimal control scheme of
the semi-active TMD applied in a jack-up offshore platform. The
TLCD had been also applied to the tension-leg platform (TLP).
Its effectiveness for angular motion reduction had been tested
numerically and experimentally [7]. Due to the increasing
demand for wind energy in recent decades, a race toward the gigan-
tism of wind turbines has begun. Larger wind turbines are more
flexible but meanwhile, have to bear higher loads. Besides, the
rotating blades produce significant gyroscopic effects that interact
with the platform’s yaw and pitch. Other complicated phenomena
like wind turbulence and shear, tower shadow, and wake effect
from the neighboring turbines also occur. Many researchers took
advantage of aeronautics technologies to reduce the loads on
wind turbines, such as blade pitch control [8] and movable flaps
on the trailing edge [9,10]. In the area of structural engineering,
most researchers had applied TMD, tuned liquid dampers (TLD),
magnetorheological (MR) fluid damper, and other energy sinks
[11] on the tower, nacelle, or blade. Several TMD optimizations
have been proposed [12,13]. The multi-TMD system had been
developed to reduce the sway motion of the nacelle [14]. Colwell
and Basu [15] carried out the fatigue analysis and showed that
TLCD can reduce the bending moment of the tower and extend
the fatigue life. Tong et al. [16] designed a bidirectional TLD
system located on the barge to reduce the load on the tower base.
Chen et al. [17] used several spherical TLDs for better performance
against seismic excitations. Martynowicz [18,19] studied the con-
trollable MR damper, which is a shock absorber equipped with elec-
trical windings to generate a magnetic field and change the viscosity
of the magnetic fluid and to alleviate the bending of the tower.
Staino et al. [20] proposed a new system that consists of tendons
and actuators embedded inside the blade to control the edgewise
vibration. Among all the solutions, TMD is popular because it is
easy to add or modify from an existing system. TLD has similar
dynamic characteristics and can be more attractive since it requires
lower construction costs and maintenance. However, its lack of
inherent damping cripples its performance, and the nonlinear beha-
vior sometimes leads to mistuning easily [21]. Past researchers had
proposed some solutions to enhance its damping effect by adding
some internal obstructions, such as wall baffles [22] and screens
[23]. Although many researchers have successively assessed their
performance [24–26], the entire progress to the practical application
remains stagnant. In recent years, another new type of TLD had
been invented by installing porous media inside the water tank
[27,28]. The porous media can bring a higher capacity for energy
dissipation. The equivalent mechanical model was proposed to
design the porous-media-tuned liquid damper (PMTLD) by
analogy with TMD. Tsao and Chang [29] completed the supple-
mentary solutions of the linear model and carried out experiments
to verify the dynamic characteristics change. Tsao and Huang
[30] included the quadratic damping equation in the numerical
method and determined some essential parameters from resonant
experiments. Tsao et al. [31] obtained the nonlinear damping ratio
from the resonance experiments. However, the preliminary research
only focused on the characteristic analysis and discussed its perfor-
mance on a pendulum-type structure. More studies on offshore
engineering problems should be carried out.
In this paper, the two-dimensional fluid–structure interaction
(FSI) problem of a PMTLD installed on a floating platform under-
going wave excitations is considered. The coupling algorithm for
fluid and structural dynamics has to satisfy the kinematic continuity
(same displacement and velocity) and dynamic continuity (equilib-
rium of stresses) on the fluid–solid interface. The potential flow and
Navier–Stokes flow models are employed for hydrodynamic com-
putation. The former is solved by a mixed Eulerian–Lagrangian
(MEL) method, which involves a mixed-type boundary value
problem solver implemented by the boundary element method
(BEM) with linear element and the free-surface particle tracker
via the second-order Tayler series expansion. Grilli et al. [32]
studied the wave problem in a similar frame and successfully sim-
ulated the wave absorption behavior by setting an appropriate
051203-2 / Vol. 145, OCTOBER 2023
radiation condition. Grilli and Horrillo [33] verified the numerical
method through a shoaling experiment. Chen et al. [34] followed
a similar approach with a new BVP solver considered for sloshing
analysis. These studies did not incorporate rigid-body dynamics in
the wave field. To do so, the Newmark method with linear acceler-
ation assumption is adopted in this paper. Some common numerical
instability issues should be addressed. Sen [35] obtained the con-
verged solutions of fluid field and body dynamics iteratively. Van
Daalen [36] combined the governing equations of the body and
fluid to derive an additional integral equation, which yields a com-
patible solution simultaneously. His method requires a little extra
computational effort but avoids numerical iterations. Tanizawa
Downloaded from http://asmedigitalcollection.asme.org/offshoremechanics/article-pdf/145/5/051203/7013652/omae_145_5_051203.pdf by University College Dublin user on 26 July 2023
[37] provided a similar formula in the implicit algorithm.
Hermans [38] calculated the reflection and transmission coefficients
of a flexible platform in waves. Koo and Kim [39] used direct and
indirect mode-decomposition methods and found that nonlinearity
could make the second and third harmonic surge forces bigger
than the first harmonic component. Jung et al. [40] carried out
wave tank experiments to study the viscous effect on a freely
rolling floating body. He concluded the linear potential theory
underestimates the roll motion at lower frequencies because of
the lack of eddy-making damping. Kim et al. [41] studied the cou-
pling effects between the ship and the sloshing tank above it. The
nonlinearity that comes from the sloshing fluid could significantly
affect the ship’s stability. Guerber et al. [42] and Dombre et al.
[43] distinguished the difference between several numerical
schemes for simulating the free motion of a fully-submerged
body. Many other strategies, such as the mode-decomposition
method [44,45] and indirect method [46], have been also proposed.
In this paper, the strongly coupled algorithm is completed via iter-
ative steps. In addition to the MEL method, the linear equivalent
mechanical model is adopted for PMTLD. This mechanical model
retains the dynamic characteristics of an actual PMTLD but signifi-
cantly reduces the degree-of-freedom of the fluid computation.
Besides, it is useful to design the PMTLD by analogy with TMD.
The two-phase Navier–Stokes model is also applied in the paper.
Compared to the potential flow model, it includes more realistic
phenomena, such as viscous effect, vortex, and breaking waves.
Piperno et al. [47] and Farhat and Lesoinne [48] showed the accu-
racy of the explicit scheme for FSI problems in Navier–Stokes flow.
However, the explicit schemes tend to be unstable when the added
mass effect is strong. The semi-implicit scheme consists of an
implicit solver for the solid (usually the Neumann part affecting
the stresses) and an explicit solver for the other part (usually the
Dirichlet part affecting the fluid velocity) [49]. Matthies et al.
[50] applied the implicit formulation to solve the FSI problem in
the arbitrary Lagrangian–Eulerian (ALE) framework. Rodrigues
et al. [51] applied FEM for the mooring lines of a floating drilling
platform. Palm et al. [52] studied the motion of the wave energy
converters by combining his VOF-RANS solver and FEM.
Tezduyar [53] demonstrates the numerical results of some impor-
tant topics in ocean engineering, such as propellers, pipes, free-
surface flow, and sloshing phenomena. Numerical approaches and
their applications were addressed in textbooks [54]. In this paper,
FEM is employed to spatially discretize the Navier–Stokes equa-
tions. The VOF method and the level set equations [55,56] are
applied to solve the free surface motion. An immersed boundary
method (IBM) [57] in conjunction with Nitsche’s technique is intro-
duced to complete the fluid–solid coupling. For PMTLD, the linear
Darcy’s flow and quadratic Forchheimer flow regimes are consid-
ered for the extra resistance induced by the porous media. Molin
[58] solved the pore flow by using amplitude-dependent coefficients
that account for added mass and damping effects. Mackay et al. [59]
and Liang et al. [60] proposed boundary conditions that include a
linear or quadratic pressure–velocity relation for the porous
surface. In this paper, the fluid computation is implemented by
the scientific toolkit Proteus [61], and the motion of the solid
phase was simulated via Chrono’s multibody dynamics engine
[62]. Both Proteus and Chrono are open-source frameworks and
available in the public domain.
Transactions of the ASME
 Downloaded from http://asmedigitalcollection.asme.org/offshoremechanics/article-pdf/145/5/051203/7013652/omae_145_5_051203.pdf by University College Dublin user on 26 July 2023
Fig. 1 The configuration of the rectangular PMTLD attached on a floating plat-
form in wave field for potential flow model

This manuscript is organized as follows. Sections 2 and 3 provide where R is the location of a free-surface particle. For wave channel,
details of the numerical methods used in the potential flow and two- the dynamic boundary condition on the free surface Γf,c can be
phase Navier–Stokes flow models, respectively. Section 4 describes expressed as
the numerical techniques for wave generation and absorption.
Section 5 presents the design strategy for PMTLD. Section 6 presents ∂ϕ 1
+ |∇ϕ|2 + gη = 0 (7)
the performance of PMTLD and the correlation between numerical ∂t 2
results through several benchmark tests. The advantages of the where η is the wave elevation. For sloshing fluid, the dynamic
PMTLD on offshore structures are briefly concluded in Sec. 7. boundary condition on the free surface Γf,d can be expressed as
∂ϕ 1
+ |∇ϕ|2 + (ϕ − Ψ)[α1 + α2 |∇ϕ − ∇Ψ|] + gη = 0 (8)
2 Nonlinear Potential Flow ∂t 2
√#
2.1 Governing Equations and Boundary Conditions. The where α1 = γμ/κρ and α2 = γcF / κ are constants related to the
configuration of the rectangular PMTLD attached to a floating plat- nature of the porous media, Ψ is the velocity potential of the
form in the wave field is shown in Fig. 1. For the incompressible, porous media. On the wetted boundaries Γw,c, Γb,c, Γw,d, Γb,d, and
inviscid, and irrotational fluid, the continuity and momentum equa- Γp,c, the impermeable condition is taken as
tions of fluid can be expressed as
∂ϕ
= ẋ · n (9)
∇·u=0 (1) ∂n
! " where n denotes the normal vector pointing outward to the fluid
∂u
ρ + (u · ∇)u = −∇p + ∇pd + ρg (2) boundary Γ, and ẋ is the velocity of the node on the wetted bound-
∂t ary. For the channel, the absorption of a wave with celerity c can be
where ρ is the fluid density, u is the fluid velocity, p is the pressure, represented by the radiation condition on the outlet boundary Γr,c as
pd is the pressure drop due to the porous media, and g is the grav- [32,67]
itational acceleration. Note that in freshwater, ∇pd = 0, then the
∂ϕ 1 ∂ϕ
nonlinearity only comes from the convective term; when the =− (10)
porous media is involved, Darcy’s flow and Forchheimer flow ∂n c ∂t
regimes are considered; therefore, the gradient of pressure drop
can be expressed as [63]
2.2 Mixed Eulerian–Lagrangian Method. By applying
γμ γcF ρ Green’s second identity to Laplace’s equation, the boundary inte-
∇pd = − (u − v) − √# |u − v|(u − v) (3)
κ κ gral equation (BIE) for a fluid boundary Γ can be obtained as
$ % &
where v is the velocity of porous media, μ is the dynamic viscosity ∂G ∂ϕ
of fluid, γ and κ are the porosity and permeability of porous media, cp ϕp = ϕ−G dΓ (11)
Γ ∂n ∂n
respectively, and cF is a dimensionless constant determined by
experiments [64–66]. Note that if u is sufficiently small, the first where G is the fundamental solution and cp is the resulting flux. In
linear-damping term in the right-hand side of Eq. (3) dominates, this paper, the BIE is solved by BEM. Therefore, Eq. (11) can be
while the second quadratic function becomes significant as u rewritten in matrix form as
increases and the drag due to solid obstacles is comparable with
the surface drag due to friction. In potential flow, the fluid velocity A{ϕ} = B{ϕn } (12)
can be expressed by the gradient of the velocity potential ϕ as where A and B are the kernel matrices, {ϕ} and {ϕn} are the vectors
u = ∇ϕ (4) containing the potential and the normal velocity at all of the bound-
ary nodes, respectively. Equation (12) arises a mixed-type BVP
with ϕ given on the free surface and ϕn given on the wetted bound-
Therefore, the Laplace equation should be satisfied as aries. The free-surface particle is tracked via the second-order
Taylor series expansion during a small time-step Δt as
∇2 ϕ = 0 (5)
DR Δt2 D2 R
The kinematic boundary condition on the free surface Γf,c and Γf,d R(t + Δt) = R(t) + Δt + + O[Δt 3 ] (13)
Dt 2 Dt 2
can be expressed as
DR Dϕ Δt 2 D2 ϕ
=u (6) ϕ(t + Δt) = ϕ(t) + Δt + + O[Δt 3 ] (14)
Dt Dt 2 Dt 2

Journal of Offshore Mechanics and Arctic Engineering OCTOBER 2023, Vol. 145 / 051203-3
 Downloaded from http://asmedigitalcollection.asme.org/offshoremechanics/article-pdf/145/5/051203/7013652/omae_145_5_051203.pdf by University College Dublin user on 26 July 2023
Fig. 2 Numerical flowcharts for the FSI simulation by (a) potential flow model and (b) two-phase Navier–Stokes model

The first-order derivative in Eq. (13) is obtained by the kinematic Laplace problems of ϕ and ϕt are expressed in the same geometry
boundary condition, while the first-order derivative in Eq. (14) can and for the same time-step, A and B shown in Eq. (12) can be
be obtained by the Bernoulli equation for Γf,c and Γf,d as reused [32].
Dϕ 1
= −gη + |∇ϕ|2 (15)
Dt 2
2.3 Dynamics of a Floating Body and Fluid–Structure
Dϕ 1 Interaction Coupling. The floating platform has three
= −gη + |∇ϕ|2 − α1 (ϕ − ψ) − α2 |∇ϕ − ∇ψ|(ϕ − ψ) (16) degrees-of-freedom (namely free to surge, heave, and pitch). The
Dt 2
motion equation can be expressed as
where D/Dt denotes the material derivative and Dϕ/Dt can be
regarded as the mechanical energy density of the fluid. The Mb ẍG = F + Mb g (22)
second-order derivatives of R and ϕ can be obtained as
IG θ̈G = M (23)
D2 Ri ∂2 ϕ ∂2 ϕ
2
= + uj ; i = 1, 2 (17)
Dt ∂t∂xi ∂xj ∂xi where ẍG and θ̈G are the translational and angular accelerations of
the body center of mass, respectively, Mb and IG are the mass and
D2 ϕ Dui Dη moment of inertia of the body, respectively. The resultant force F
= ui −g (18) and moment M from PMTLD and waves can be expressed as
Dt 2 Dt Dt
A new unknown variable ϕt = ∂ϕ/∂t appears in Eqs. (17) and $ $
(18). Since ϕt also satisfies Laplace’s equation, another mixed-type F= pn dΓ + pn dΓ (24)
BVP for ϕt and ϕtn arises and needs to be solved. The Dirichlet con- Γ p,c Γw,d +Γb,d
dition on the free surface Γf,c and Γf,d is given as
Dϕ $ $
ϕt = − |∇ϕ|2 (19) M= p(r × n) dΓ + p(r × n) dΓ (25)
Dt
Γ p,c Γw,d +Γb,d
The Neumann condition on the wetted boundaries Γw,c, Γb,c, Γw,d,
Γb,d, and Γp,c is given as [43] where r is the position of a point on the body boundary with respect
Dẋ · n ∂n to the center of mass, and the pressure along the wetted boundary
ϕtn = − · ∇ϕ − ẋ · ∇∇ϕ · n (20) Γp,c, Γw,d, and Γb,d can be obtained by the Bernoulli equation
Dt ∂t % &
The Neumann condition on the absorption boundary Γr,c is given ∂ϕ 1
p = −ρ + |∇ϕ|2 + gη (26)
as ∂t 2

∂2 ϕ The Newmark method based on constant acceleration assumption

ϕtn = c (21) is employed to solve the motion equation as
∂s2
where ẍ is the acceleration of the node on the wetted boundary, and Fk+1
ẍG,k+1 = (27)
s denotes the unit tangential vector of the boundary. Since the Mb

051203-4 / Vol. 145, OCTOBER 2023 Transactions of the ASME

 ẋG,k+1 = ẋG,k + 0.5Δt(ẍG,k + ẍG,k+1 ) (28) at the heights Zn as shown in Fig. 4. The equivalent masses, stiffness
kn, damping ratios ξn, and location heights can be expressed as [29]
xG,k+1 = xG,k + Δt ẋG,k + 0.25Δt 2 (ẍG,k + ẍG,k+1 ) (29) 8γρB
mn = tanhan hd (32)
where k is the time-step. This scheme is unconditionally stable. a3n Ld
Besides, no sub-step is required. The resultant force used in the
Newmark scheme comes from the fluid field. It is not a priori 8γρgB
known because, to solve the fluid field, the body kinematics used kn = tanh2 an hd (33)
a2n Ld
in Eqs. (9) and (20) should be given but they depend on body
motion as '##########
γμ cothan hd
ẋ = (ẋG + θ̇G × r) · n (30) ξn = (34)
2κρ an g

Downloaded from http://asmedigitalcollection.asme.org/offshoremechanics/article-pdf/145/5/051203/7013652/omae_145_5_051203.pdf by University College Dublin user on 26 July 2023

ẍ = [ẍG + θ̈G × r + θ̇G × (θ̇G × r)] · n (31) 2 1
Zn = hd + − (35)
To ensure the continuity of fluid field and body motion, the iter- an sinhan hd an tanhan hd
ative algorithm illustrated in Algorithm 1 is used [43]. The flow-
chart is presented in Fig. 2(a). ) *
∞
( 8
mf = γρBLd hd 1 − 3 L2 h
tanhan hd (36)
a
n=0 n d d
Algorithm 1 Algorithm for wave–structure interactions in
potential flow ) *
% 2 & ( ∞
1 Ld hd L3d
do k > 0 Zf = γρB + − mn Zn (37)
Start the iterative step with the solution at kth time-step as an initial guess
mf 2 12 n=0
ε = abs(ẍG,k,i − ẍG,k,i−1 )
where an = (2n + 1)π/Ld, n = 0, 1, 2 … , and mw = γρLdBhd denote
do while ε > εtolerance the total water mass in the tank. The wave elevation on the right-
Solve ϕi and ϕn,i (the first BVP) side wall can be expressed in terms of the equivalent displacement
Solve ϕt,i and ϕtn,i (the second BVP) xn as
Calculate pressure along wetted boundaries
(∞
Calculate force and moment act on the floating body 4
Define predictor of force and moment: η= tanhan hd · xn (38)
a L
n=0 n d
F∗ =Fk,i + ω(F∗i−1 −Fk,i )

Solve rigid body motion (Newmark method)

Update free surface and body boundary Sk,i+1 and ϕk,i+1
end do
xk+1 = xk,i+1

ẋk+1 = ẋk,i+1

ẍk+1 = ẍk,i+1

ϕk+1 = ϕk,i+1

Sk+1 = Sk,i+1
end do

2.4 Equivalent Mechanical Model. The equivalent mechani- Fig. 4 The schematic diagram of the equivalent mechanical
cal model of a rectangular PMTLD shown in Fig. 3 consists of a model
fixed mass mf located at the heights Zf and infinite
mass-dashpot-spring systems (denoted by mn, cn, and kn) located

Fig. 5 The configuration of the rectangular PMTLD attached on

Fig. 3 The schematic diagram of the porous media in a rectan- a floating platform in wave field for two-phase Navier–Stokes
gular tank model

Journal of Offshore Mechanics and Arctic Engineering OCTOBER 2023, Vol. 145 / 051203-5
 As the equivalent model is applied, the PMTLD-platform system The Eikonal equation is added for the calculation of a signed dis-
can be regarded as a multi-degree-of-freedom system so its response tance function ψd as
is solved by the Newmark method instead.
|∇ψ d | = 1 and ψ d = 0 on Γf (49)
The volume fraction is linked to the signed distance function by
3 Two-Phase Navier–Stokes Flow
3.1 Governing Equations. The configuration of the FSI λΔψ ′ = H(ψ d + ψ ′ ) − α with ∇ψ ′ · n = 0 (50)
problem in the two-phase Navier–Stokes flow is shown in Fig. 5. ′
where λ is a parameter that penalizes the deviation of ψ from a
The entire fluid domain consists of two incompressible Newtonian
global constant.
phases (air and water) separated by a sharp material interface.
Across the interface, the density and viscosity are discontinuous
but velocity and pressure are continuous. Surface tension is 3.3 Fluid–Structure Interaction Coupling via Force–

Downloaded from http://asmedigitalcollection.asme.org/offshoremechanics/article-pdf/145/5/051203/7013652/omae_145_5_051203.pdf by University College Dublin user on 26 July 2023

neglected. The continuity and Navier–Stokes equations formulated Displacement Co-Simulation. For space discretization, the residual-
for each fluid phase are based variational multiscale (RBVMS) method [68] is applied to
Eqs. (42), (43), and (47)–(50) with linear triangular mesh. Some strat-
∇ · ui = 0 (39) egies for improving stability and accuracy are employed but not
! " addressed in this paper [69]. Nitsche’s technique and an IBM are
∂ui used to accommodate the solid motion [70], and then, the motion of
ρ + (ui · ∇)ui = −∇pi + ∇ · σ i + f (40) the solid phase was calculated via Chrono’s multibody dynamics
∂t
engine [62]. The position and velocity of the structure are input to
where σi is the stress tensor and f is the body force. i ∈ {w, a} where the fluid solver as the boundary conditions. The sequential process is
subscript w and a denote the water and air phases, respectively. The illustrated in Algorithm 2. The flowchart is presented in Fig. 2(b).
proper boundary conditions on the air-water interface are provided
based on the assumption of continuity of velocity and stress across
the interface Γf as Algorithm 2 Algorithm for wave–structure interactions in
Navier–Stokes model
pw = pa , uw = ua , σ w · n = σ a · n (41)
do k > 0
Discretize and solve Navier–Stokes equations
Advance the level set and volume fraction
3.2 Level Set Method. By using Eq. (41) and neglecting the Re-initialize the level set function (Eikonal equation)
loss of smoothness due to the jump discontinuities in density and Apply mass conservative correction
viscosity, the continuous global pressure and velocity fields can Update level set function and volume fraction
be introduced. Therefore, the single Navier–Stokes equations for Calculate pressure and force
the entire fluid domain can be obtained as Solve rigid body motion (Chrono)
end do
∇·u=0 (42)
! "
∂u
ρ + (u · ∇)u = −∇p + ∇ · (2με) + f (43) 4 Wave Generation and Wave Absorption
∂t
In our examples, the periodic waves are generated by the piston
where ε = 12 (∇u + ∇uT ). The density and viscosity of the two-fluid wavemaker. For the potential flow model, the estimation of the rela-
system are given by tion between the target wave height H0 and the demand stroke of the
wavemaker S is described as [71]
ρ = ρw H(ψ) + ρa [1 − H(ψ)] (44)
sinh2k0 hc + 2k0 hc
S= H0 (51)
μ = μw H(ψ) + μa [1 − H(ψ)] (45) 2(cosh2k0 hc − 1)
where H(ψ) is the Heaviside function expressed as where k0 is the wavenumber. Therefore, the harmonic displacement
 of the wavemaker is given by
 0, ψ <0
H(ψ) = 1/2, ψ = 0 (46) S
 xf = sin(ωf t) (52)
1, ψ >0 2
The wavemaker’s displacement changes the geometry of the fluid
Note that Eq. (42) equals Eq. (1) but it is used in Ωa and Ωw. In domain. Meanwhile, the first- and second-order time derivatives of
this paper, the level set method is used to distinguish the fluid sub- Eq. (52) (i.e., the velocity and acceleration of the piston wave-
domains. The free surface is defined by the level set function ψ = 0. maker) will be used in Eqs. (9) and (20). For the outlet of the
ψ > 0 is for the water phase and ψ < 0 for the air phase. The free channel, the radiation condition shown in Eq. (10) is applied.
surface evolution can be described by For the two-phase Navier–Stokes model, the relaxation zones are
∂ψ set on both sides of the wave channel as shown in Fig. 6 for wave
+ u · ∇ψ = 0 (47) generation and wave absorption. In the generation zone, this method
∂t
does not need to move the physical boundary of the wavemaker to
To avoid unacceptable mass conservation errors, the differential generate waves but gradually imposes the target wave kinematics,
form of the mass conservation equation yields the volume fraction including velocity and free-surface position. In the absorption
equation that can be written as zone, it does not use the radiation boundary condition but
imposes a zero velocity. The target velocity is given by adding a
∂α source term in the Navier–Stokes equations as [72]
+ ∇ · (αu) = 0 (48)
∂t % &
∂u
where α is the volume fraction. Equations (42)–(47) constitute a ρ + u · ∇u − g − ∇ · σ = αb αc (u − ut ) (53)
complete model of the two-phase flow at the continuous level. ∂t

051203-6 / Vol. 145, OCTOBER 2023 Transactions of the ASME

 Fig. 6 Illustration of relaxation zones and the scaled distance in the numerical
channel
Fig. 7 Design chart of tank dimensions and porous media prop-
erty for different mass ratios
where αb is a blending function, αc is a constant, and ut is the target
velocity. The blending function αb is given by
d 3.5
e −1
αb = (54)
e−1
where d is the normalized scaled distance from the boundary to the end
of the relaxation zone. d = 1 means the source term is fully imposed at
the boundary and d = 0 means the source term vanishes at the end of
the relaxation zone. The detail and overall performance of the wave
generation module are described in previous research [73].
5 Design Parameters of PMTLD
By considering the first mode of the equivalent mechanical
model, the PMTLD can be designed based on the analogy with
TMD [27]. Therefore, the tuning frequency ratio ft and optimal
damping ratio ξopt can be obtained by [74]
ω0 1 The free-vibration responses of the floating platform when θ0 =
ft = = (55)
ωs 1 + Rm
Fig. 8 The element and node distribution on the computational domains and boundaries of
the wave channel and floating body (not to scale): (a) potential flow model and (b) two-phase
Navier–Stokes model
Journal of Offshore Mechanics and Arctic Engineering
Downloaded from http://asmedigitalcollection.asme.org/offshoremechanics/article-pdf/145/5/051203/7013652/omae_145_5_051203.pdf by University College Dublin user on 26 July 2023
.###########
3Rm
ξopt = (56)
8(1 + Rm )
where ωs is the natural frequency of the main structure, and Rm =
m0
is the mass ratio. Note that Eqs. (55) and (56) are
mf + Mb
derived based on some simple assumptions, so the design parame-
ters may not be truly “optimal” in our case. However, they are still
close and useful. The optimal parameters for different configura-
tions of mass dampers can be referred to in textbooks [75]. Once
Rm is given by the designer, the natural frequency and the
damping ratio of PMTLD are known. The corresponding tank
dimensions and porous media properties can be determined. A
simple design chart is shown in Fig. 7.
6 Examples
6.1 Dynamic Characteristics of Floating Platform. Since
the natural frequency of the floating platform is a required parame-
ter for PMTLD design, the free-decay test should be carried out in
advance. For the unit-width two-dimensional channel, the length is
10 m and the water depth is 2 m. The rectangular platform has a
length = 1 m, height = 0.5 m, and density = 450 kg/m3, resulting
the mass = 225 kg and moment of inertia = 23.44 kg · m2. No
waves are applied to the flow field. The rectangular platform is
not equipped with PMTLD. Its center of gravity is placed at the
height of the still water surface and has an initial pitch angle θ0 =
5 deg and θ0 = 10 deg. The schematic diagram of element and
node distribution on the computational domains and boundaries
of the wave channel and floating body are shown in Fig. 8. In
BEM, 1714 two-node linear elements are distributed on the wave
boundary. In FEM, 375 k triangular cells are distributed in the
entire domain. Some convergence tests had been carried out in
advance but will not be shown in this paper.
5 deg by two numerical models are shown in Fig. 9. The
OCTOBER 2023, Vol. 145 / 051203-7
 Downloaded from http://asmedigitalcollection.asme.org/offshoremechanics/article-pdf/145/5/051203/7013652/omae_145_5_051203.pdf by University College Dublin user on 26 July 2023
Fig. 9 The displacement, velocity, and acceleration of a freely moving floating
platform when θ0 = 5 deg: (a) surge, (b) heave, and (c) pitch (solid line: potential
flow and dashed line: NS flow)

Fig. 10 The response spectra of free vibration of the floating platform when θ0 = 5 deg (solid line: potential flow and
dashed line: NS flow)

051203-8 / Vol. 145, OCTOBER 2023 Transactions of the ASME


Fig. 11 The displacement, velocity, and acceleration of a freely moving floating platform
when θ0 = 10 deg: (a) surge, (b) heave, and (c) pitch (solid line: potential flow and dashed
line: NS flow)
corresponding response spectra are shown in Fig. 10. For surge
motion, the drifts in both models are small, while the harmonics
can be observed in the time histories of velocity and acceleration.
The horizontal force comes from the pressure difference between
the left and right sides of the body, which is in sync with the
pitch motion. This phenomenon explains the surge and pitch
motions have the same oscillatory frequency of 0.6104 Hz as
shown in Fig. 10. For the heave motion, the solutions by potential
and NS flow models are quite similar because the added mass effect
is significant. Both models have taken the added mass effect into
account. The heaving frequency = 0.6714 Hz as shown in Fig. 10.
For the pitch motion, the amplitude decay can be observed in the
NS model. As it had been mentioned in past research [40], wave
damping is important to pitch motion. Therefore, the potential
model overestimates the angular amplitude if no artificial wave
damping is applied. However, they give the same pitching
Journal of Offshore Mechanics and Arctic Engineering
Downloaded from http://asmedigitalcollection.asme.org/offshoremechanics/article-pdf/145/5/051203/7013652/omae_145_5_051203.pdf by University College Dublin user on 26 July 2023
frequencies shown in Fig. 10. The free-vibration responses of the
floating platform when θ0 = 10 deg by two numerical models are
shown in Fig. 11. The corresponding response spectra are shown
in Fig. 12. A larger initial pitch does not significantly change the
dynamic characteristics of the floating body. However, the differ-
ence between the two numerical models may grow because their
different physical assumptions for the wave field will affect the
FSI simulation.
6.2 PMTLD on a Floating Platform in Waves. In this
example, Rm is given as 0.1. When PMTLD is designed based on
the pitch mode (ωs = 3.84 rad/s), the tuning frequency can be
obtained as 3.49 rad/s. Therefore, the PMTLD has a length of
0.62 m and a water depth of 0.049 m. The porosity of the porous
media = 0.96. The optimal damping ratio = 18.4%; resulting α1
OCTOBER 2023, Vol. 145 / 051203-9
 Downloaded from http://asmedigitalcollection.asme.org/offshoremechanics/article-pdf/145/5/051203/7013652/omae_145_5_051203.pdf by University College Dublin user on 26 July 2023
Fig. 12 The response spectra of free vibration of the floating platform when θ0 = 10 deg (solid line: potential flow and dashed
line: NS flow)

Fig. 13 The element and node distribution on the computational domains and boundaries of
PMTLD (not to scale): (a) potential flow model and (b) two-phase NS model

Fig. 14 The surge motions of the floating platform with and without PMTLD when β = 1.0
(solid line a: potential flow with PMTLD, dotted line a: mechanical model with PMTLD,
dashed line a: potential flow without PMTLD, solid line b: NS flow with PMTLD, dashed line
b: NS flow without PMTLD)

051203-10 / Vol. 145, OCTOBER 2023 Transactions of the ASME

 Downloaded from http://asmedigitalcollection.asme.org/offshoremechanics/article-pdf/145/5/051203/7013652/omae_145_5_051203.pdf by University College Dublin user on 26 July 2023
Fig. 15 The heave motions of the floating platform with and without PMTLD when β = 1.0
(solid line a: potential flow with PMTLD, dotted line a: mechanical model with PMTLD,
dashed line a: potential flow without PMTLD, solid line b: NS flow with PMTLD, and dashed
line b: NS flow without PMTLD)

Fig. 16 The pitch motions of the floating platform with and without PMTLD when β = 1.0 (solid
line a: potential flow with PMTLD, dotted line a: mechanical model with PMTLD, dashed line a:
potential flow without PMTLD, solid line b: NS flow with PMTLD, and dashed line b: NS flow
without PMTLD)

Fig. 17 The forces and moment act on the floating platform when β = 1.0 (solid line: total
resultant, dashed line: wave load, dotted line: PMTLD reaction)

Journal of Offshore Mechanics and Arctic Engineering OCTOBER 2023, Vol. 145 / 051203-11
and α2 are assumed to be a reasonable value of 1.28 s−1 and
5.0 m−1, respectively, to satisfy the optimal design criterion [30].
The masses of the tank and porous media are neglected. The har-
monic waves are generated to propagate along the x-direction, from
left to right, with wave height = 0.05 m. In the two-phase NS
model, the lengths of the wave generation and absorption zones are
2 m. The schematic diagrams of element and node distribution on
the computational domains and boundaries of PMTLD are shown
in Fig. 13. In BEM, 80 two-node linear elements are distributed on
the PMTLD boundary. In FEM, 1211 triangular cells are distributed
in the PMTLD domain. In the equivalent mechanical model, only the
first mode is considered. In our tests, the numerical simulation termi-

Downloaded from http://asmedigitalcollection.asme.org/offshoremechanics/article-pdf/145/5/051203/7013652/omae_145_5_051203.pdf by University College Dublin user on 26 July 2023

nates before high-order (spurious) waves reach the structure.
When the PMTLD is installed on the floating platform, it is called
a controlled body, otherwise, it is called an uncontrolled body. The
forcing frequency ratio β is defined as the ratio between the fre-
quency of the wavemaker’s movement and the natural pitching fre-
quency of the body. For the resonant case (β = 1.0), the surge,
heave, and pitch motions of the controlled and uncontrolled floating
platforms are shown in Figs. 14–16, respectively. For surge motion,
PMTLD can slow down the lateral drift but will not be able to
reduce the velocity and acceleration of the platform. Although the
lateral drift is small, the results of the NS model are quite different
from that of the potential model. For heave motion, the mean values
of displacement are different because the additional weight of the
PMTLD will change the static displacement. The oscillatory ampli-
tudes are still the same showing that PMTLD is not effective for
heaving reduction. Besides, since the mass of PMTLD is only
10% of the main structure, it does not change the dynamic charac-
teristics of the heave motion much. The extra dynamic effect due to
increasing load is minor so the associate evaluation is not urgent in
this paper. For pitch motion, a significant reduction can be
observed. In the potential flow model, the PMTLD can cut down
the pitching amplitude from 12 deg to 2.5 deg (80% of vibration),
while in the NS flow model, the pitching amplitude decreases
from 7.7 deg to 2.5 deg (67% of vibration). Considering the wave
damping was neglected by the potential model, the reduction ratio
of 67% is a more reasonable value for reference. In the controlled
case, the angular amplitude in the potential model is at the same
level as that in the NS model. This shows that damping due to
porous media is a more important resource for response mitigation.
In our example, the results from the equivalent mechanical model
and nonlinear potential flow model are very close. This implies
the first modal response and the linear damping term dominate.
Figure 17 shows the forces and moment act on the floating plat-
form obtained from the potential flow model. For the horizontal
force Fx, PMTLD produces a sloshing force that is out of phase
with the wave load. However, in our case, no oscillation is observed
because the floating platform does not subject to any constraints
such as mooring lines. Therefore, the effectiveness is uncertain.
For the vertical force Fy, the sloshing force is simply a dead load.
PMTLD does not reduce the heave motion as shown in Fig. 15.
For the overturning moment Mθ , although the PMTLD is not per-
fectly tuned, the sloshing moment can cancel out the wave-induced
moment. Therefore, as the PMTLD dampens the kinetic energy of
the floating platform, the pitch motion is greatly reduced.
Fig. 18 The frequency responses of the floating platform’s pitch
The frequency responses of the root-mean-square (RMS) of the
motion
pitch motion are shown in Fig. 18. In uncontrolled cases, resonance
occurs at β = 1. The magnitudes of potential and NS models are dif-
ferent due to wave viscosity. As PMTLD is installed, the peaks of
the response are shifted to β = 1.05. This implies the system’s
natural frequency has changed. The magnitudes of both models Table 1 The computational cost of three numerical models for a
are close because the damping from porous media dominates. The 10-s simulation
overall reduction of the RMS response by PMTLD is about 75%.
Equivalent model Potential flow NS flow
Although the PMTLD is designed based on a displacement crite-
rion, it is effective to reduce the velocity and acceleration. Coupling algorithm Iterative Iterative Sequential
Table 1 shows the computational cost of three numerical models Number of node 1714 1794 375 k
for a 10-s simulation. The computations were completed on the Number of core 16 16 48
HPC machine QB3 at Louisiana Optical Network Infrastructure CPU time (h) 2.7 3 32
(LONI). Parallel computing is applied in three numerical

051203-12 / Vol. 145, OCTOBER 2023 Transactions of the ASME

 Conflict of Interest
There are no conflicts of interest.

Data Availability Statement

The data sets generated and supporting the findings of this article
are obtainable from the corresponding author upon reasonable
request.

Downloaded from http://asmedigitalcollection.asme.org/offshoremechanics/article-pdf/145/5/051203/7013652/omae_145_5_051203.pdf by University College Dublin user on 26 July 2023

Nomenclature
c = celerity
d = normalized scaled distance
Fig. 19 Number of iterations required by the MEL method for h = depth
potential flow model k = time-step
p = pressure
f = body force
schemes. BEM in the potential model costs only 9% of the comput- g = gravitational acceleration
ing time by FEM in the two-phase NS model due to simpler spatial n = unit outward normal vector of fluid boundary
discretization and time integration. The equivalent mechanical r = position vector
model avoids part of the fluid computation so it costs the least. u = fluid velocity
The number of iterations required by the MEL method for the v = intrinsic velocity
potential model during a 10-s simulation is shown in Fig. 19. x = position vector of node on wetted boundary
The same solver is executed at least ten times in one time-step. B = width
If the sequential algorithm or strong-coupled equations are adopted, G = green function
the MEL method is expected to save more computational time. H = heaviside function
L = length
S = demand stroke of wavemaker
7 Conclusions A = Kernel matrix
B = Kernel matrix
This paper validates the effectiveness of the response mitigation
F = force
for the floating platform by PMTLD through the numerical FSI
M = moment
simulation completed by the potential flow and the two-phase
R = position vector of node on free surface
Navier–Stokes flow model.
cF = dimensionless constant
The NS model describes the pitching oscillation better because it
cp = flux
takes the damping of ambient waves into account. However, most
ft = tuning frequency ratio
energy is consumed by PMTLD, so the numerical results of the con-
k0 = wavenumber
trolled cases in both models are similar. The linear equivalent mechan-
kn = equivalent stiffness
ical model successfully predicts the interaction between PMTLD and a
mf = fixed mass
floating platform in our case. This implies the first modal response and
mn = equivalent mass
the linear damping term dominate. Compared to FEM, BEM saves tre-
mw = water mass
mendous computational costs. Higher efficiency can be further
pd = pressure drop
obtained if the iterative algorithm is avoided. Through this research,
xf = displacement of wavemaker
researchers can also understand the trade-offs between physical fidel-
xn = equivalent displacement
ity and computational effort for practical problems.
ut = target velocity
Porous-media-tuned liquid damper is effective for reducing pitch
xG = translational displacement
motion. At our resonance test, 67% of vibration can be reduced. The
H0 = wave height
overall reduction of RMS response can be up to 80%. However, it
IG = moment of inertia of floating body
does not produce a useful control force in the vertical direction so
Mb = mass of floating body
heaving control is uncertain. Although the horizontal control force
Rm = mass ratio
cancels out the wave loads, no obvious reduction in surge motion is
Zf = fixed height
observed since the horizontal drift is unconstrained. Therefore, the
Zn = equivalent height
surge reduction is not validated. The PMTLD can be applied to
α = volume fraction
moored floating structures and wind turbine platforms for further
α1 = linear damping coefficient
investigation. Detailed parametric studies on the tuning condition
α2 = nonlinear damping coefficient
and optimal damping of PMTLD will be investigated. Experimental
αb = blending function
verification for both numerical models will be carried out in the future.
αc = constant
Γ = boundary of fluid domain
Δt = time interval
Acknowledgment γ = porosity
This work was partially supported by internal research funding ε = absolute error
from the Center for Computation and Technology and the η = wave elevation
College of Engineering at Louisiana State University. Portions of θG = angular displacement
this research were conducted with high-performance computing κ = permeability
resources provided by Louisiana State University.2 λ = parameter
μ = dynamic viscosity of fluid
ξn = equivalent damping ratio
2
http://www.hpc.lsu.edu ξopt = optimal damping ratio

Journal of Offshore Mechanics and Arctic Engineering OCTOBER 2023, Vol. 145 / 051203-13
 ρ = fluid density
σi = stress tensor
ϕ = velocity potential of fluid
Ψ = velocity potential of porous media
ψ = level set function
ψd = signed distance
ωf = forcing frequency of wavemaker
ωs = natural frequency of the main structure
References
[1] Burger, W., and Corbet, A. G., 1966, “Chapter III—Anti-Rolling Devices in
General,” Ship Stabilizers: A Handbook for Merchant Navy Officers, Pergamon
Press, Oxford, UK, pp. 31–37.
[2] Chen, Y. H., and Ko, C. H., 2003, “Active Tuned Liquid Column Damper With
Propellers,” Earthquake Eng. Struct. Dyn., 32(10), pp. 1627–1638.
[3] Yalla, S. K., Kareem, A., and Kantor, J. C., 2001, “Semi-Active Tuned Liquid
Column Dampers for Vibration Control of Structures,” Eng. Struct., 23(11),
pp. 1469–1479.
[4] Marzouk, O. A., and Nayfeh, A. H., 2009, “Control of Ship Roll Using Passive
and Active Anti-Roll Tanks,” Ocean Eng., 36(9–10), pp. 661–671.
[5] Gattulli, V., and Ghanem, R., 1999, “Adaptive Control of Flow-Induced
Oscillations Including Vortex Effects,” Int. J. Non Linear Mech., 34(5),
pp. 853–868.
[6] Kawano, K., 1993, “Active Control Effects on Dynamic Response of Offshore
Structures,” Proceedings of the 3rd International Offshore and Polar
Engineering Conference, Singapore, June 6–11, pp. 594–598.
[7] Lee, H. H., Wong, S. H., and Lee, R. S., 2006, “Response Mitigation on the
Offshore Floating Platform System With Tuned Liquid Column Damper,”
Ocean Eng., 33(8–9), pp. 1118–1142.
[8] Bossanyi, E. A., 2000, “The Design of Closed Loop Controllers for Wind
Turbines,” Wind Energy, 3(3), pp. 149–163.
[9] Andersen, P. B., Henriksen, L., Gaunaa, M., Bak, C., and Buhl, T., 2010,
“Deformable Trailing Edge Flaps for Modern Megawatt Wind Turbine
Controllers Using Strain Gauge Sensors,” Wind Energy, 13(2–3), pp. 193–206.
[10] Castaignet, D., Barlas, T. K., Buhl, T., Poulsen, N. K., Wedel-Heinen, J. J.,
Olesen, N. A., Bak, C., and Kim, T., 2013, “Full-Scale Test of Trailing Edge
Flaps on a Vestas V27 Wind Turbine: Active Load Reduction and System
Identification,” Wind Energy, 17(4), pp. 549–564.
[11] Rahman, M., Ong, Z. C., Chong, W. T., Julai, S., and Khoo, S. Y., 2015,
“Performance Enhancement of Wind Turbine Systems With Vibration Control:
A Review,” Renewable Sustainable Energy Rev., 51, pp. 43–54.
[12] Stewart, G., and Lackner, M., 2013, “Offshore Wind Turbine Load Reduction
Employing Optimal Passive Tuned Mass Damping Systems,” IEEE Trans.
Control Syst. Technol., 21(4), pp. 1090–1104.
[13] Si, Y., Karimi, H. R., and Gao, H., 2014, “Modelling and Optimization of a
Passive Structural Control Design for a Spar-Type Floating Wind Turbine,”
Eng. Struct., 69, pp. 168–182.
[14] Dinh, V. N., and Basu, B., 2014, “Passive Control of Floating Offshore Wind
Turbine Nacelle and Spar Vibrations by Multiple Tuned Mass Dampers,”
Struct. Control Health Monit., 22(1), pp. 152–176.
[15] Colwell, S., and Basu, B., 2009, “Tuned Liquid Column Dampers in Offshore
Wind Turbines for Structural Control,” Eng. Struct., 31(2), pp. 358–368.
[16] Tong, X., Zhao, X., and Karcanias, A., 2018, “Passive Vibration Control of an
Offshore Floating Hydrostatic Wind Turbine Model,” Wind Energy, 21(9),
pp. 697–714.
[17] Chen, J., Zhan, G., and Zhao, Y., 2016, “Application of Spherical Tuned Liquid
Damper in Vibration Control of Wind Turbine due to Earthquake Excitations,”
Struct. Des. Tall Special Build., 25(10), pp. 431–443.
[18] Martynowicz, P., 2015, “Vibration Control of Wind Turbine Tower-Nacelle
Model With Magnetorheological Tuned Vibration Absorber,” J. Vib. Control,
23(20), pp. 3468–3489.
[19] Martynowicz, P., 2017, “Control of a Magnetorheological Tuned Vibration
Absorber for Wind Turbine Application Utilising the Refined Force Tracking
Algorithm,” J. Low Freq. Noise Vib. Act. Control, 36(4), pp. 339–353.
[20] Staino, A., Basu, B., and Nielsen, S., 2012, “Actuator Control of Edgewise
Vibrations in Wind Turbine Blades,” J. Sound Vib., 331(6), pp. 1233–1256.
[21] Chen, Y. H., Hwang, W. S., and Tsao, W. H., 2018, “Nonlinear Dynamic
Characteristics of Rectangular and Cylindrical TLD’s,” J. Eng. Mech., 144(9),
p. 06018004.
[22] Zahrai, S. M., Abbasi, S., Samali, B., and Vrcelj, Z., 2012, “Experimental
Investigation of Utilizing TLD With Baffles in a Scaled Down 5-Story
Benchmark Building,” J. Fluids Struct., 28, pp. 194–210.
[23] Warnitchai, P., and Pinkaew, T., 1998, “Modelling of Liquid Sloshing in
Rectangular Tanks With Flow-Dampening Devices,” Eng. Struct., 20(7),
pp. 293–600.
[24] Faltinsen, O. M., Firoozkoohi, R., and Timokha, A. N., 2011, “Analytical
Modeling of Liquid Sloshing in a Two-Dimensional Rectangular Tank With a
Slat Screen,” J. Eng. Math., 70(1–3), pp. 93–109.
[25] Hamelin, J. A., Love, J. S., Tait, M. J., and Wison, J. C., 2013, “Tuned Liquid
Dampers With a Keulegan-Carpenter Number-Dependent Screen Drag
Coefficient,” J. Fluids Struct., 43, pp. 271–286.
[26] Tait, M. J., 2008, “Modelling and Preliminary Design of a Structure-TLD
System,” Eng. Struct., 30(10), pp. 2644–2655.
051203-14 / Vol. 145, OCTOBER 2023
[27] Tsao, W. H., and Hwang, W. S., 2018, “Tuned Liquid Dampers With Porous
Media,” Ocean Eng., 167, pp. 55–64.
[28] Tsao, W. H., and Hwang, W. S., 2019, “Dynamic Characteristics of Liquid
Sloshing in Cylindrical Tanks Filled With Porous Media,” IOP Conf. Ser. Earth
Environ. Sci., 351, p. 012007.
[29] Tsao, W. H., and Chang, T. J., 2020, “Sloshing Phenomenon in Rectangular and
Cylindrical Tanks Filled With Porous Media: Supplementary Solution and
Impulsive-Excitation Experiment,” J. Eng. Mech., 146(12), p. 04020139.
[30] Tsao, W. H., and Huang, Y. L., 2021, “Sloshing Force in a Rectangular Tank
With Porous Media,” Results Eng., 11, p. 100250.
[31] Tsao, W. H., Huang, L. H., and Hwang, W. S., 2021, “An Equivalent Mechanical
Model With Nonlinear Damping for Sloshing Rectangular Tank With Porous
Media,” Ocean Eng., 242, p. 110145.
[32] Grilli, S. T., Skourup, J., and Svendsen, I. A., 1989, “An Efficient Boundary
Downloaded from http://asmedigitalcollection.asme.org/offshoremechanics/article-pdf/145/5/051203/7013652/omae_145_5_051203.pdf by University College Dublin user on 26 July 2023
Element Method for Nonlinear Water Waves,” Eng. Anal. Boundary Elem.,
6(2), pp. 97–107.
[33] Grilli, S. T., and Horrillo, J., 1997, “Numerical Generation and Absorption of
Fully Nonlinear Periodic Waves,” J. Eng. Mech., 123(10), pp. 1060–1069.
[34] Chen, Y. H., Hwang, W. S., and Tsao, W. H., 2017, “Nonlinear Sloshing Analysis
by Regularized Boundary Integral Method,” J. Eng. Mech., 143(8), p. 040170046.
[35] Sen, D., 1993, “Numerical Simulation of Motions of Two-Dimensional Floating
Bodies,” J. Ship Res., 37(4), pp. 307–330.
[36] Van Daalen, E. F. G., 1993, “Numerical and Theoretical Studies of Water Waves
and Floating Bodies,” Ph.D. thesis, Universiteit Twente, The Netherlands.
[37] Tanizawa, K., 1995, “A Nonlinear Simulation Method of 3D Body Motions in
Waves,” J. Soc. Nav. Archit. Jpn., 178, pp. 179–191.
[38] Hermans, A. J., 2000, “A Boundary Element Method for the Interaction of
Free-Surface Waves With a Very Large Floating Flexible Platform,” J. Fluids
Struct., 14(7), pp. 943–956.
[39] Koo, W., and Kim, M. H., 2004, “Freely Floating-Body Simulation by a 2D Fully
Nonlinear Numerical Wave Tank,” Ocean Eng., 31(16), pp. 2011–2046.
[40] Jung, K. H., Chang, K. A., and Jo, H. J., 2006, “Viscous Effect on the Roll Motion
of a Rectangular Structure,” J. Eng. Mech., 132(2), pp. 190–200.
[41] Kim, Y., Nam, B. W., Kim, D. W., and Kim, Y. S., 2007, “Study on Coupling
Effects of Ship Motion and Sloshing,” Ocean Eng., 34(16), pp. 2176–2187.
[42] Guerber, E., Benoit, M., Grilli, S. T., and Buvat, C., 2012, “A Fully Nonlinear
Implicit Model for Wave Interactions With Submerged Structures in Forced or
Free Motion,” Eng. Anal. Boundary Elem., 36(7), pp. 1151–1163.
[43] Dombre, E., Benoit, M., Violeau, D., Peyrard, C., and Grilli, S. T., 2015,
“Simulation of Floating Structure Dynamics in Waves by Implicit Coupling of
a Fully Non-Linear Potential Flow Model and a Rigid Body Motion
Approach,” J. Ocean Eng. Mar. Energy, 1(1), pp. 55–76.
[44] Vinje, T., and Brevig, P., 1981, “Nonlinear, Two-Dimensional Ship Motions,”
Technical Report of the Norwegian Institute of Technology, R-112.81.
[45] Belibassakis, K. A., 2008, “A Boundary Element Method for the Hydrodynamic
Analysis of Floating Bodies in Variable Bathymetry Regions,” Eng. Anal.
Boundary Elem., 32(10), pp. 796–810.
[46] Wu, G. X., and Eatock-Taylor, R., 1996, “Transient Motion of a Floating Body in
Steep Water Waves,” Proceedings of the 11th International Workshop on Water
Waves and Floating Bodies, Hamburg, Germany, Mar. 17–20.
[47] Piperno, S., Farhat, C., and Larrouturou, B., 1995, “Partitioned Procedures for the
Transient Solution of Coupled Aeroelastic Problems Part I: Model Problem,
Theory and Two-Dimensional Application,” Comput. Methods Appl. Mech.
Eng., 124(1–2), pp. 79–112.
[48] Farhat, C., and Lesoinne, M., 2000, “Two Efficient Staggered Algorithms for the
Serial and Parallel Solution of Three-Dimensional Nonlinear Transient
Aeroelastic Problems,” Comput. Methods Appl. Mech. Eng., 182(3–4),
pp. 499–515.
[49] Guermond, J. L., Minev, P., and Shen, J., 2006, “An Overview of Projection
Methods for Incompressible Flows,” Comput. Methods Appl. Mech. Eng.,
195(44–47), pp. 6011–6045.
[50] Matthies, H. G., Niekamp, R., and Steindorf, J., 2006, “Algorithms for Strong
Coupling Procedures,” Comput. Methods Appl. Mech. Eng., 195(17–18),
pp. 2028–2049.
[51] Rodrigues, M. V., Correa, F. N., and Jacob, B. P., 2007, “Implicit Domain
Decomposition Methods for Coupled Analysis of Offshore Platforms,”
Commun. Numer. Methods Eng., 23(6), pp. 599–621.
[52] Palm, J., Eskilsson, C., Paredes, G. M., and Bergdahl, L., 2016, “Coupled
Mooring Analysis of Floating Wave Energy Converters Using CFD:
Formulation and Validation,” Int. J. Mar. Energy, 16, pp. 83–99.
[53] Tezduyar, T. E., 2001, “Finite Element Methods for Flow Problems With Moving
Boundaries and Interfaces,” Comput. Meth. Eng., 8(2), pp. 83–130.
[54] Karimirad, M., Michailides, C., and Nematbakhsh, A., 2018, Offshore
Mechanics: Structural and Fluid Dynamics for Recent Applications, Wiley, UK.
[55] Osher, S., and Sethian, J. A., 1988, “Fronts Propagating With
Curvature-Dependent Speed: Algorithms Based on Hamilton-Jacobi
Formulations,” J. Comput. Phys., 79(1), pp. 12–49.
[56] Osher, S., and Fedkiw, R., 2003, Level Set Methods and Dynamic Implicit
Surfaces: Applied Mathematical Sciences, Vol. 153. Springer-Verlag, New York.
[57] Mittal, R., and Iaccarino, G., 2005, “Immersed Boundary Methods,” Annu. Rev.
Fluid Mech., 37(1), pp. 239–261.
[58] Molin, B., 2011, “Hydrodynamic Modeling of Perforated Structures,” Appl.
Ocean Res., 33(1), pp. 1–11.
[59] Mackay, E., Liang, H., and Johanning, L., 2021, “A BEM Model for Wave Forces
on Structures With Thin Porous Elements,” J. Fluids Struct., 102, p. 103246.
Transactions of the ASME
[60] Liang, H., Zheng, S., Magee, A., and Greaves, D., 2022, “Water Wave
Interactions With Perforated Elastic Disks: Quadratic Pressure Discharge
Condition,” Phys. Rev. Fluids, 7(5), p. 054802.
[61] “Proteus: Computational Methods and Simulation Toolkit,” http://proteustoolkit.org.
[62] Tasora, A., Serban, R., Mazhar, H., Pazouki, A., Melanz, D., Fleischmann, J.,
Taylor, M., Sugiyama, H., and Negrut, D., 2016, Chrono: An Open
Source Multi-Physics Dynamics Engine, High Performance Computing in
Science and Engineering—Lecture Notes in Computer Science, Springer,
New York.
[63] Nield, A. D., and Bejan, A., 2013, Convection in Porous Media, 4th ed., Springer,
New York.
[64] Beavers, G. S., and Joseph, D. D., 1967, “Boundary Conditions at a Naturally
Permeable Wall,” J. Fluid Mech., 30(1), pp. 197–207.
[65] Steiros, K., Kokmanian, K., Bempedelis, N., and Hultmark, M., 2020, “The Effect
of Porosity on the Drag of Cylinders,” J. Fluid Mech., 901, p. R2.
[66] Xu, Y. P., Liu, J. Y., Chai, X. W., and Shi, S. S., 2022, “Experimental Study on
Measurement of Drag Coefficient of Sandy Soil Nonlinear Vadose,” Adv. Civil
Eng., 2022, p. 8104842.
[67] Clément, A., 1996, “Coupling of Two Absorbing Boundary Conditions for 2D
Time-Domain Simulations of Free Surface Gravity Waves,” J. Comput. Phys.,
126(1), pp. 139–151.
Journal of Offshore Mechanics and Arctic Engineering
[68] Bazilevs, Y., Calo, V. M., Cottrel, J. A., Hughes, T. J. R., Reali, A., and Scovazzi,
G., 2007, “Variational Multiscale Residual-Based Turbulence Modeling for Large
Eddy Simulation of Incompressible Flows,” Comput. Methods Appl. Mech. Eng.,
197(1–4), pp. 173–201.
[69] Kees, C. E., Akkerman, I., Farthing, M. W., and Bazilevs, Y., 2011, “A
Conservative Level Set Method Suitable for Variable-Order Approximations
and Unstructured Meshes,” J. Comput. Phys., 230(12), pp. 4536–4558.
[70] Peskin, C. S., 1977, “Numerical Analysis of Blood Flow in the Heart,” J. Comput.
Phys., 25(3), pp. 220–252.
[71] Ursell, F., Dean, R. G., and Yu, Y. S., 1960, “Forced Small-Amplitude Water
Waves: A Comparison of Theory and Experiment,” J. Fluid Mech., 7(1),
pp. 33–52.
[72] Jacobsen, N. G., Fuhrman, D. R., and Fredsøe, J., 2012, “A Wave Generation
Toolbox for the Open-Source CFD Library: OpenFoam,” Int. J. Numer.
Methods Fluids, 70(9), pp. 1073–1088.
Downloaded from http://asmedigitalcollection.asme.org/offshoremechanics/article-pdf/145/5/051203/7013652/omae_145_5_051203.pdf by University College Dublin user on 26 July 2023
[73] Dimakopoulos, A. S., de Lataillade, T., and Kees, C. E., 2019, “Fast Random
Wave Generation in Numerical Tanks,” Eng. Comput. Mech., 172(1), pp. 1–11.
[74] Den Hartog, J. P., 1956, Mechanical Vibrations, 4th ed., McGraw-Hill,
New York.
[75] Soong, T. T., and Dargush, G. F., 1997, Passive Energy Dissipation Systems in
Structural Engineering, Wiley, New York.
OCTOBER 2023, Vol. 145 / 051203-15