Jmse 09 01151 v2

Journal of
Marine Science
and Engineering
Article
CFD Prediction of Ship Seakeeping and Slamming Behaviors of
a Trimaran in Oblique Regular Waves
Xiyu Liao, Zhanyang Chen * , Hongbin Gui and Mengchao Du
Department of Ocean Engineering, Harbin Institute of Technology at Weihai, Weihai 264209, China;
[email protected] (X.L.); [email protected] (H.G.); [email protected] (M.D.)
* Correspondence: [email protected]; Tel.: +86-1350-6318-766
Abstract: The main hull encounters waves at first and causes waves to break, when trimarans are
subject to the slamming in head waves. At this moment, emergence phenomena of side hulls will not
occur. Thus, the slamming study of trimarans in oblique waves presents further practical significance.
In this study, a CFD method is used for trimaran seakeeping and slamming analysis. An overset
grid technique is adopted to simulate ship motions in waves. Firstly, to further verify the present
method, a series of verification and validation studies is conducted. Then, the motion responses and
slamming pressure with different control parameters, such as forward speed and ship heading angle,
are calculated and discussed. The comparative results indicate that the seakeeping and slamming
behaviors of trimarans differ significantly from those of conventional monohull ships. Finally, severe
bow slamming and green water in oblique waves are also observed and investigated, which should
be given enough attention during ship design and evaluation.
Keywords: CFD; seakeeping; slamming; trimaran; oblique waves

Citation: Liao, X.; Chen, Z.; Gui, H.;

Du, M. CFD Prediction of Ship
Seakeeping and Slamming Behaviors 1. Introduction
of a Trimaran in Oblique Regular Recently, high-performance ships have been attracting the attention of more and
Waves. J. Mar. Sci. Eng. 2021, 9, 1151.
more researchers. Trimarans with good maneuverability and high stability have wide
https://doi.org/10.3390/jmse9101151
development prospects [1]. When navigating in severe sea states, the phenomena of
emergence and water entry of the bow and stern of ships will occur frequently, which may
Academic Editors: Tahsin Tezdogan,
lead to slamming. Generally speaking, slamming is a transient process. The duration of
Jialong Jiao and Kostas Belibassakis
pressure generated by transient impact is very short, but the peak value is often extremely
large. The instantaneous impact of water entry will lead to local structural deformation
Received: 1 September 2021
Accepted: 11 October 2021
of the hull and even structural failure. The complex cross-deck structure makes it more
Published: 19 October 2021
difficult to study the vibration of trimarans [2].
At present, the related research on trimaran slamming mainly focuses on the simula-
Publisher’s Note: MDPI stays neutral
tion calculation and water-entry test by using the simplified hull section model. Nikfar-
with regard to jurisdictional claims in
jam [3] presents an experimental investigation on the pressure distribution on three wedge
published maps and institutional affil- sections with 15◦ , 20◦ and 30◦ deadrise during water-entry. The results give an appropriate
iations. approximation of the maximum pressures by the model resembling high-speed craft’s hull
sections, which can be used to estimate impact loads in different operational conditions.
Zong et al. [4] conducted a vertical water entry experiment of a 2-D trimaran section from
different drop heights, and flow fields were conducted using PIV technology for each con-
Copyright: © 2021 by the authors.
dition. It was found that the flow particle velocity at the bottom of the trimaran model and
Licensee MDPI, Basel, Switzerland.
the free surface between the main and side hulls was higher. By means of Flow-3D which is
This article is an open access article
a commercial computational fluid dynamics code, Ghadimi et al. [5] simulated seakeeping
distributed under the terms and of a wave-piercing trimaran in the presence of irregular waves via standard Bretschneider
conditions of the Creative Commons spectrum in sea state 5 in various seagoing modes. In the slamming analysis of trimaran,
Attribution (CC BY) license (https:// the relative vertical velocity of bow obtained from seakeeping analysis is used to study the
creativecommons.org/licenses/by/ problem of hull section entering the water in the most serious slamming mode. Wu et al. [6]
4.0/). reported the slamming loads on a typical trimaran section, based on both experimental
J. Mar. Sci. Eng. 2021, 9, 1151. https://doi.org/10.3390/jmse9101151 https://www.mdpi.com/journal/jmse

J. Mar. Sci. Eng. 2021, 9, 1151 2 of 23
measurements and numerical results. The convergence of the MPS method in numerical
simulation is verified by comparing the results of different particle configurations and
time-step sizes. The results show that the acceleration and slamming pressure simulated
by the MPS method are in good agreement with the experimental results. Sun et al. [7]
analyzed the simulation of different types of motions. It was found that the characteristics
of the dynamics during slamming strongly correlate to penetration depth, regardless of the
types of entry.
The above research mainly considers the slamming caused by trimaran longitudi-
nal/vertical motion in the head wave. In actual navigation, it is difficult for a ship to always
keep in the head wave. Since there is an angle between the hull and the wave propagation
in oblique waves, the transverse flow may cause the hull to roll, which will make the
side hulls periodically shake around the main hull. When the roll angle reaches a certain
value, the phenomena of emergence and water entry of the side hull will occur frequently.
Differing from the slamming caused by longitudinal/vertical motion, the slamming caused
by transverse motion is unique to a trimaran.
In addition, many researchers pay more attention to the hull vibration induced by
waves in all directions, but there are few studies on the impact of wave loads on trimaran
slamming based on the whole ship model [8,9]. Therefore, it is necessary to study the behav-
ior of the hull slamming that occurs in waves based on ship motion. Jiao and Huang [10,11]
simulated ship motions in both uni- and bi-directional regular waves based on the CFD
method. Moreover, severe bow slamming and green water under bi-directional waves
are also observed and investigated. Lin [12] numerically and experimentally investigated
the ship motions and slamming pressures of a 10,000. TEU container ship in regular
head waves. By calculating, they found the duration of the slamming, especially under
parametric roll conditions, mainly depends on the rolling period and encountering wave.
The hydrodynamic response analysis of trimaran has always been the authors’ re-
search focus. In the authors’ previous work [9], the nonlinear hydroelastic responses of the
trimaran are analyzed by means of the segmented model experiment. The present study
is the further expansion of previous studies. This study aims to conduct a comparative
study on the motion and slamming responses of trimaran advancing in oblique regular
waves by CFD simulations, which will give a better understanding of the influence of wave
direction on trimaran slamming performance. The main novelty of this study is disclosing
the motion characteristics leading to trimaran slamming as well as the relationship between
slamming position and slamming time in oblique regular waves. The conclusion drawn in
this study would guide the safety of ships navigating in waves.
2. Numerical Model Set-Up

The numerical calculations were applied to a trimaran model based on the CFD
technique. Details of the CFD model and numerical method are presented in this section.
2.1. Ship Hull and Simulation Conditions

Only the naked hull is involved in the seakeeping and slamming investigation.
The rudder, propeller or bilge keels are not appended to the model. The 3-D model of the
trimaran is built by using the CFD software package STAR-CCM+, as shown in Figure 1,
which gives a side and a top view of the model. The main dimensions of the ship model
are listed in Table 1, where VCG denotes vertical center of gravity, BL denotes baseline.
J.J.Mar.
Mar.Sci.
Sci.Eng.
Eng.2021,
2021,9,9,x1151
FOR PEER REVIEW 3 3ofof23
23
Overviewof
Figure1.1.Overview
Figure ofthe
thetrimaran.
trimaran.
Mainparticulars
Table1.1.Main
Table particularsof
ofthe
thetrimaran.
trimaran.
Item Item ValueValue

Length overall
Length overall LOA /m L OA/m 142 142
Breadth Breadth /m
B/m 27 27
D/m
DepthDepth /m 13 13
Draught T/m 6
Draught /m 6
Displacement ∆/ton 397
from BL ZG /m ∆/ton
VCG Displacement 7.85 397
Length ofVCG
a sidefrom
hull BL
LP /m /m 62 7.85
Breadth of a side
Length sideBPhull
of ahull /m /m 3.76 62
Depth of a side hull DP /m 8.26
Breadth of a side hull /m 3.76
Depth of a side hull /m 8.26
This study mainly focuses on investigating ship large amplitude motions and slam-
ming pressures
This in oblique
study mainly focusesregular waves. Typical
on investigating shipwave
large heading
amplitude angles are and
motions defined
slam-in
ming pressures in oblique regular waves. Typical wave heading angles are defined in Ta-is
Table 2. The port side of the trimaran is the wave-forward side, and the starboard side
the2.wave
ble backward
The port side ofside. The CFDissimulation
the trimaran conditions
the wave-forward in this
side, andstudy are listedside
the starboard in Table
is the3.
The wave
wave heading
backward in the
side. Therange
CFD of 0◦ –135◦ was
simulation selected in
conditions forthis
analysis
studyandarecomparison, so the
listed in Table 3.
selected forward speed and wave height are relatively gentle.
The wave heading in the range of 0°–135° was selected for analysis and comparison, so
the selected forward speed and wave height are relatively gentle.
Table 2. Definition of wave headings.
Table 2. Definition of wave headings.
Wave Headings Value
Wavewave
Head Headings 0◦Value
Head wave
Bow quarter wave 45◦ 0°
Beam wave
Bow quarter wave 90◦ 45°
Stern quartering wave 135◦
Beam wave 90°
Following wave 180◦
Stern quartering wave 135°
Following wave 180°
Table 3. CFD simulation conditions.
Table 3. CFD simulation conditions.
Wavelength/λ Wave Height/H Forward Speed/v Wave Heading/θ
Wavelength/ Wave Height/ Forward Speed/
6.5 m/s Wave 0◦ , Heading/
15◦ , 30◦ , 45◦ , 135◦
146 m 6m 6.5 m/s 5 m/s 0◦ , 45°,
0°, 15°, 30°, 45◦ , 135°
135◦
146 m 6m
5 m/s 0°, 45°, 135°
J. Mar. Sci. Eng. 2021, 9, x FOR PEER REVIEW 4 of 23
J. Mar. Sci. Eng. 2021, 9, 1151 4 of 23

In addition, according to the Equation (1) provided by STAR CCM +
In addition, [13],
UserGuide_14.04 according
the timeto
stepthe isEquation
calculated(1) provided
by using wave by STAR
period P andCCM +
number
UserGuide_14.04
of mesh n, [13], the time step is calculated by using wave period P and number
of mesh n,
In addition, according to the Equation (1) provided by STAR CCM + UserGuide_14.04 [13],
P
the time step τ is calculated by using waveτ period
= P P and number of mesh n, (1)
τ = 2.4n (1)
2.4
Pn
By calculation, ≈ 0.0503. Thus, inτ this
= study, the time step is 0.05 s. The wave pe-
(1)
riodBy calculation,
used in this paper ≈ 0.0503. Thus,the
is 200 times this 2.4n
in time study, the time
step, which stepthe
meets is 0.05 s. The wave
requirements pe-
recom-
riod used
mended in this paper
by ITTC [14,15].
By calculation, is 200 times the time step, which meets the requirements
τ ≈ 0.0503. Thus, in this study, the time step is 0.05 s. The wave period recom-
mended by ITTC [14,15].
used in this paper is 200 times the time step, which meets the requirements recommended
2.2.ITTC
by Numerical Scheme
[14,15].
2.2. Numerical Scheme
The CFD model is developed based on the Unsteady Reynolds-averaged Navier–
2.2. Numerical
The
Stokes (URANS) Scheme
CFD model is developed
method. based on
The turbulence the Unsteady
model Reynolds-averaged
of Realizable k-ε is used in this Navier–
study.
Stokes (URANS)
The CFD method.
model is The turbulence
developed based model
on the of Realizable
Unsteady
The Volume of Fluid (VOF) method is used to capture the free surface between air and k-ε is used
Reynolds-averaged in this study.
Navier–
The Volume
Stokes
water. Figureof2Fluid
(URANS) (VOF)the
method.
illustrates method
The is usedin
turbulence
free surface tothe
modelcapture the free
of Realizable
established CFDsurface
k-εmodel between
is used byin airstudy.
this
showing and
the
water.
water volume fraction profile around the hull surface. As shown in the scale bar, the the
The Figure
Volume of2 illustrates
Fluid (VOF) the free
method surface
is used in
to the established
capture the free CFD
surface model
between by showing
air and water.
vol-
water
Figure volume
ume fraction fraction
2 illustrates
of watertheprofile
isfree
used around
surface inthe
to reflect the hull
the surface.
established
status As
CFD
of the shown
model in by
computationaltheshowing
scale
cell. bar,
The the
the vol-of
water
value
ume
volumefraction
0.5 means of awater
fraction
that profileis used
around
computational to cell
reflect
the ishull the
filled status
surface.
with 50%of
Asthe computational
shown
water in the
and air,cell.
50%scale bar,The
which the value
volume
representsof
0.5 means
fraction that
of a
water computational
is used to cell
reflectis filled
the with
status 50%
of water
the and 50%
computational
the free surface. The value of 0 and 1 represents a computational cell is filled with air and air, which
cell. Therepresents
value of
the
0.5 free
meanssurface.
that
water, respectively.a The value
computational of 0 and
cell1isrepresents
filled with a computational
50% water and cell
50% is filled
air, whichwith air and
represents
the free
water, surface. The value of 0 and 1 represents a computational cell is filled with air and
respectively.
water, respectively.
Figure 2. Water volume fraction profile around the hull surface.

Dynamic Fluid
Dynamic Fluid Body
BodyInteraction
Interaction(DFBI) (DFBI)modulemodule is adopted
is adopted to simulate
to simulate trimaran
trimaranmo-
Dynamic Fluid BodyThe Interaction (DFBI) module is adopted to simulate trimaran mo-
motions in oblique waves. The release time is set to 0.3 s and the buffer time is sets.toIt
tions in oblique waves. release time is set to 0.3 s and the buffer time is set to 0.6
tions in oblique
is confirmed thatwaves. The releaseobtain
this module time isthe setship
to 0.3 s andresponses
the buffer time is set to 0.6 s. It
0.6 s. It is confirmed that thiscan module can obtainmotion
the ship motioninresponseswaves conveniently
in waves
isconveniently
confirmed that
and accuratelyand this module
[10,11,16].
accurately can
However, obtain
[10,11,16]. the ship
if sixHowever, motion
Degrees ofif Freedomresponses
six Degrees in
(6-DOF) waves conveniently
of the trimaran
of Freedom (6-DOF)are of
and
the accurately
released, the are
trimaran [10,11,16].
ship will not
released, However,
have
the if sixstability
course
ship will Degrees
not ofoblique
haveincourseFreedom (6-DOF)
waves.
stability inThus, of the
oblique 3-DOFtrimaran
of the
waves. are
tri-
Thus,
released,
3-DOF ofthe
maran that theship
include will
trimaran not
roll, have
pitch
that course
and
include heave stability
roll, are
pitch in oblique
released
and waves.
to ensure
heave thatThus,
are released the to3-DOF
trimaran
ensure ofdoes
the tri-
that not
the
maran
trimaranthat
deviate from include
doesthe roll,
notgiven
deviate pitch
route. and
from heave
Since are
the hull
the given released
is fixed
route. Sinceto ensure
inthe
thehull that
direction the
is fixed trimaran
ofinX the Y axes,not
does
anddirection the
of
deviate
X and Yfrom
incoming axes, the
thegiven
velocity can be
incomingroute. Since as
regarded
velocity the hull
the
can is fixedspeed.
trimaran
be regarded inasthe direction
As
the shown in
trimaran ofspeed.
X andAs
Figure Y according
3, axes, the
shown in
incoming velocity
to the velocity
Figure 3, according can be
decomposition regarded
to the velocity as
theorem, the trimaran
the incoming
decomposition speed. As
velocity
theorem, shown in Figure
v can be decomposed
the incoming 3, according
velocity v caninto be
to the velocity
(along decomposition
theinto
ship vlength) thetheorem, the incoming velocity vflowcanvelocity).
be decomposed into
decomposed l (alongand ship(transverse
length) andflow velocity).
vd (transverse
(along the ship length) and (transverse flow velocity).
Figure 3.
Figure 3. Decomposition of incoming
incoming flow
flow velocity.
velocity.
Figure 3. Decomposition of incoming flow velocity.
J. Mar. Sci. Eng. 2021, 9, 1151 5 of 23
2.3. Fluid Domain and Boundary Conditions

2.3. Fluid Domain and Boundary Conditions
An overset mesh scheme, which has been widely used by relevant scholars in simu-
lating An overset
complex meshwith
bodies scheme,
largewhich has been
amplitude widely
motions used by relevant
[10–12,16–21], scholars
is applied in simu-
to simulate
lating complex bodies with large amplitude motions [10–12,16–21],
the trimaran motions in waves. The computational fluid domain consists of a background is applied to simulate
the trimaran
region motionsregion,
and an overset in waves.
andThe computational
the overset meshesfluid domain
are used consists
for two of atobackground
regions exchange
region
data. and an overset region, and the overset meshes are used for two regions to ex-
change data.
In this paper, the computational domain is established with a dimension of −3.5 L <
In this paper, −2.3
x < 1.5 L in length, the computational
L < y < 2.1 L indomain
width isand
established
−1.6 L < with
z < 1 aLdimension
in height. of −3.5 L
Figure < x < 1.5 L
4 presents
the length, −
in general 2.3 Lof
view < ythe
< 2.1 L indomain
fluid width and and − L < z < 1 Lofinboundary
1.6definition
the height. Figure 4 presents
conditions the
corre-
general
sponding to view of the fluid domain and the definition of boundary conditions
θ = 45°. The coordinate origin is set at the same height level with the calm corresponding
to θ =
water 45◦ . The
surface. Thecoordinate origin
positive x-axis, is setand
y-axis at z-axis
the same
point height
to thelevel
shipwith
bow,the
portcalmsidewater
and
surface. The positive x-axis, y-axis and z-axis point to the ship bow, port side and the sky,
the sky, respectively. The velocity inlet boundary condition is used at the upstream of the
respectively. The velocity inlet boundary condition is used at the upstream of the numerical
numerical towing tank. The outlet boundary condition is set to pressure outlet. The ship
towing tank. The outlet boundary condition is set to pressure outlet. The ship surface is set
surface is set as the no-slip wall boundary condition. Moreover, the velocity boundary
as the no-slip wall boundary condition. Moreover, the velocity boundary method is used
method is used to avoid the wave reflection caused by the boundary. Due to the complex
to avoid the wave reflection caused by the boundary. Due to the complex wave heading
wave heading in this paper, the boundary conditions of the fluid domain will change with
in this paper, the boundary conditions of the fluid domain will change with the wave
the wave direction angle. The maximum simulation time is 80 s. The second-order con-
direction angle. The maximum simulation time is 80 s. The second-order convergence
vergence scheme is used in time and space. The CFL_I value is set to 0.5, and the CFL _U
scheme is used in time and space. The CFL_I value is set to 0.5, and the CFL _U value is set
value is set to 1.
to 1.
(a) (b)
Figure
Figure4.4.Computational
Computationaldomain
domainand
andboundary
boundaryconditions.
conditions.(a)(a)Side
Sideview;
view;(b)
(b)Front
Frontview.
view.
2.4.
2.4.Monitoring
MonitoringPoints
PointsforforSlamming
SlammingPressure
Pressure
InIn
order to to
order record
recordthetheslamming
slamming pressure in oblique
pressure in obliquewaves during
waves water
during entryentry
water of theof
bow, a series
the bow, of monitoring
a series of monitoringpointspoints
are distributed on the on
are distributed main
thehull,
mainside hulls
hull, and
side cross-
hulls and
decks of the trimaran,
cross-decks as shownas
of the trimaran, in shown
Figure 5.inP1-P8
Figurelocate on the locate
5. P1–P8 centerline of bow,
on the and L1-of
centerline
L3bow,
locate
andbow flare
L1–L3 area.bow
locate Moreover, to investigate
flare area. Moreover, to theinvestigate
asymmetric thepressure distribution
asymmetric pressure
ondistribution
two sides on of the
two sides of the bow in oblique waves, the monitoring points are on
bow in oblique waves, the monitoring points are arranged both
arranged
sides of bow
on both sidessymmetrically. L4-L13 L4–L13
of bow symmetrically. locate on the on
locate left-side hull and
the left-side hullleft
andcross-deck, as
left cross-deck,
shown in Figure
as shown 5b. (L
in Figure 5b.and R represent
(L and portport
R represent and and
starboard, respectively).
starboard, respectively).
J. Mar. Sci. Eng. 2021, 9, 1151 6 of 23
(a)
(b)
Figure
Figure5.5.Monitoring
Monitoringpoints
pointsofofslamming
slammingpressure.
pressure.(a)
(a)Main
Mainhull;
hull;(b)
(b)Side
Sidehulls.
hulls.
3.3.Verification
VerificationandandValidation
Validationof ofthe
theNumerical
NumericalApproach
Approach
InIn
this study,
this study,the
theprimary
primaryfocus
focusisisononship
shipmotions
motionsand andslamming
slammingpressure
pressureininoblique
oblique
waves.
waves.To To further verifythe
further verify thework
workconducted,
conducted, along
along with with the methodology
the methodology proposed
proposed above,
above,
the verification and validation studies are conducted in this section. The verificationThe
the verification and validation studies are conducted in this section. will
verification
focus on gridwilluncertainty
focus on grid uncertainty
analysis, and theanalysis,
validationand theinclude
will validation will include
the ability the
to maintain
ability to maintain
the wave elevationthe wave elevation
of oblique waves and of oblique wavesofand
ship motions pitchship
andmotions
heave. of pitch and
heave.
3.1. Mesh Generation
3.1. Mesh
Mesh Generation
generation is performed using the automatic meshing tool of STAR-CCM+.
According to Ref. [11],isapproximately
Mesh generation performed using 3.74the
million cells of
automatic unstructured
meshing tool oftrimmed
STAR-CCM+.hexahedral
Ac-
mesh which
cording to Ref.include 2.7 million cells
[11], approximately in the background
3.74 million region and
cells of unstructured 1.04 million
trimmed in the
hexahedral
overset
mesh region
which are created
include for numerical
2.7 million cells in simulation. The general
the background regionview
and of themillion
1.04 computational
in the
grids isregion
overset depicted
are in Figurefor
created 6. numerical simulation. The general view of the computa-
tional grids is depicted in Figure 6.
J. J.Mar.
Mar.Sci.
Sci.Eng.
Eng.2021,
Eng. 2021,9,9,x x1151
FOR
FORPEER
PEERREVIEW
REVIEW 7 7of
7ofof2323
(a)
(a) (b)
(b)
Figure
Figure6.6.General
Generalview
viewofofthe
thecomputational
computationalmesh.
computational mesh.(a)
mesh. (a)Refined
(a) Refinedmeshes
Refined meshesofof
meshes offlow
flowfield;
flow field;(b)
field; (b)Overview
Overviewofofthe
Overview therefined
refinedmeshes
meshes
around the hull.
around the hull.
The
TheCFD
The CFDmesh
CFD meshconvergence
mesh convergencetest
convergence testshould
test shouldbe
should beconducted
be conductedfirst.
conducted first.The
first. Themethod
The methodofof
method ofRichard-
Richard-
Richard-
son’s
son’s extrapolation
son’s extrapolation is used
extrapolation isisused to estimate
used toestimate
estimate the
thetheuncertainty
uncertainty
uncertainty and
andand numerical
numerical
numerical error
error [22,23].
[22,23].
error TheThe
[22,23]. con-
The
condition
dition forfor
condition forbeam
beambeamseasea
was
sea was selected
selected
was forfor
selected forthe
the theverification
verification study.
study.
verification TheThe
study. Thecondition
condition is forisis
condition for
forwave-
wavelength
wave-
length
λ/L =λ/L
length λ/L=and
1.0 =1.0 and
andforward
1.0forward speed
forward speed vv=m/s.
v = 6.5
speed =6.5
6.5m/s.
m/s.The
The Thegrid
grid gridrefinement
refinement factor
factor
refinement r risisdefined
r is defined
factor as: as:
defined as:
ℎhℎ
r=
== ℎ
coarse (2)
(2)
hℎf ine
where ℎ denotes the coarser grid size,hℎℎ denotes
size, denotes the finer grid size.
where hℎcoarse denotes
where denotesthe thecoarser
coarsergrid
gridsize,
f ine
denotesthethe finer
finer grid grid size.
size.
The
The factor = =√2 √
The factor r = 2 is used to divide into three different sizes of grids.Three
factor √2 isisused
used totodivide
divide into
into three
three different
different sizes
sizes ofof grids.
grids. Threegrids
Three grids
grids
named
named Grid
Grid 1,1,Grid
Grid 2 2and
and Grid
Grid 3 3are
aredefined
defined ininthis
thisstudy.
study. Grid
Grid
named Grid 1, Grid 2 and Grid 3 are defined in this study. Grid 1 is the fine mesh 1 1isisthe
thefine
finemesh
mesh grid
grid
(5.60
(5.60million
million
grid (5.60 cells), Grid
cells),cells),
million 2 (3.74
Grid 2Grid (3.74million
million
2 (3.74 cells) is the
cells) cells)
million medium
is theismedium mesh
the medium gird
mesh mesh and
gird and Grid
girdGrid 3 (2.18
and 3Grid(2.183
million
millioncells)
cells)isis
(2.18 million the
thecoarse
cells) is the mesh
coarse meshgrid.
coarse grid.
mesh The
The grid
grid
grid. Thedetails
gridof
details ofthe
thehull
details hullpart
of thepart are
hull shown
arepart areinshown
shown inFigure
Figurein
7.7.
Figure 7.
(a)
(a) (b)
(b) (c)
(c)
Figure
Figure7.7.
Figure 7.The
Thethree
The threemesh
three meshcases
mesh casesfor
cases forverification
for verificationstudy
verification study(a)
study (a)fine
(a) finegrid;
fine grid;(b)
grid; (b)medium
(b) mediumgrid;
medium grid;(c)
grid; (c)coarse
(c) coarsegrid.
coarse grid.
grid.
TheThepitch
The pitch
pitch motion
motion
motion peak peak
peak and and
and the
the peak
the peak
peak pressure
pressure
pressure atatM8,
at
M8, R4
M8,R4and
R4 L4
andandL4monitoring
L4 monitoring
monitoring points are
points
points are
selected
are selected
selected for uncertainty
for for uncertainty
uncertainty analysis. Analyzed
analysis.
analysis. Analyzed
Analyzed results for
results
results the
forforthe grid
thegrid refinement
gridrefinement studies
refinementstudies are
studies are
summarized
summarized
summarized ininTable
in Table
Table 4.4.It
4.ItisIt
isobvious
is obvious
obvious that
that the
therelative
that the error
relative
relative error between
error Grid
between
between Grid1Grid
1and
and1Gridand2Grid
Grid 2( ( ) 2)
isis 21
(esignificantly smaller than 32
) is significantly
significantly smaller thanthat
smaller than
that between Grid
that between
between Grid22and andGrid
Grid 2 and
Grid 33( Grid
( ).).It3Itseems ). Itthat
(eseems seems
that the
theresults
that the
results
ofofthe
thenumerical
results simulation
of the numerical
numerical simulation tend
tendtotoconverge
simulation tend to with
converge withthe
theincreased
converge with thenumber
increased numberofofcells.
increased cells.The
number Theofmax-
cells.
max-
imum
imum Grid
GridConvergence
The maximum Index
Grid Convergence
Convergence Index(GCI)(GCI) isis0.03975
Index (GCI)atis
0.03975 atmonitoring
0.03975
monitoring point
pointR4
at monitoring R4and the
point
and theminimum
R4 and the
minimum
was 0.007549
minimum wasat the pitch
0.007549 motion
at the peak.
pitch Hence,
motion peak.according
Hence, to Table
according
was 0.007549 at the pitch motion peak. Hence, according to Table 4, the numerical uncer- 4,to the
Tablenumerical
4, the uncer-
numerical
tainty
tainty ininthe
uncertainty thefine-grid
in
fine-grid solution
the fine-gridsolution for
forthe
solution peak
for
the the pressure
peak peak atatR4
pressure
pressure R4atmonitoring
R4 monitoring
monitoring points
points should
points should
should be
be
be reported
reported
reported as 3.975%.
asas3.975%.
3.975%. In order
InInorder
order to achieve
totoachieve
achieve a compromise
aacompromise
compromise between
betweenbetween computation
computation
computation efficiency
efficiency
efficiency and
and
and accuracy,
accuracy,
accuracy, Grid2Grid
Grid 2 is for
2isisused
used used
forthe for
the the subsequent
subsequent
subsequent simulation.
simulation.
simulation.
J. Mar. Sci. Eng. 2021, 9, 1151 8 of 23
Table 4. Grid uncertainty analysis summary.
Pitch Motion
Table 4. Grid uncertainty analysis Peak
summary. M8 R4 L4
Grid 1 2.988° 67,038 Pa 53,946.9 Pa 31,515.8 Pa
Pitch Motion Peak M8 R4 L4
Grid 2 3.029° 67,693 Pa 53,129.8 Pa 32,187.4 Pa
Grid31
Grid 2.988◦
2.895° 67,038Pa
68,790 Pa 53,946.9
51,923.7PaPa 31,515.8
30,472.9PaPa
Grid 2 3.029◦ 67,693 Pa 53,129.8 Pa 32,187.4 Pa
e 0.01372 0.00977 0.01514 0.02131
Grid 3 2.895◦ 68,790 Pa 51,923.7 Pa 30,472.9 Pa
e e21 0.04432
0.01372 0.01621
0.00977 0.02322
0.01514 0.05326
0.02131
GCIe32 0.00754
0.04432 0.01809
0.01621 0.03975
0.02322 0.01043
0.05326
GCI21
fine 0.00754 0.01809 0.03975 0.01043
To capture the more accurate behavior of the free surface, The meshes around the
free surface are refined. The mesh size in the Z direction is 1/16 of the wave height, and
To capture the more accurate behavior of the free surface, The meshes around the free
the mesh
surface are size in the The
refined. X and Y directions
mesh is 1/80
size in the of the wave
Z direction is 1/16length.
of theTowave
reflect the slamming
height, and the
pressure
mesh sizedistribution
in the X andand green water
Y directions moreofaccurately,
is 1/80 the wave the hull To
length. surface
reflectmeshes and the
the slamming
meshes near the pressure monitoring points are further refined. In
pressure distribution and green water more accurately, the hull surface meshes and the transition region
the
between the overset region and the background region, a similar mesh size
meshes near the pressure monitoring points are further refined. In the transition region is used to en-
sure the mesh
between data transmission
the overset region andbetween the two regions.
the background region, A boundary
a similar layer
mesh grid
size is of 5 cells
used to
is set on the hull wall surface where wall value y+ is between 30−60.
ensure the mesh data transmission between the two regions. A boundary layer grid of
5 cells is set on the hull wall surface where wall value y+ is between 30−60.
3.2. Numerical Towing Tank Validation
3.2. Numerical
The waves Towing Tank Validation
generated by the CFD method above are compared with the theoretical
valuesThetowaves
verify generated
the reliability of the
by the CFD numerical towingare
method above tank. Three wave
compared with probes in front
the theoretical
of the ship
values are adopted
to verify for theofmeasurement
the reliability the numericaloftowing
wave elevation.
tank. ThreeThe longitudinal
wave probes in position
front of
of probe
the 1 isadopted
ship are 36 m (approximately L/4) from
for the measurement of the
waveship’s forward
elevation. Theperpendicular. The trans-
longitudinal position of
probe 1 is 36 m between
verse distance probe 2L/4)
(approximately from
(or 3) andthe
theship’s
ship’sforward perpendicular.
centerline The transverse
is 30 m, as shown in Figure
distance
8. between probe 2 (or 3) and the ship’s centerline is 30 m, as shown in Figure 8.
Figure 8.
Figure 8. Simulated waves and wave
wave probe
probe arrangement.
arrangement.
The calm water

watersurface
surfaceisisset z =z6=m6 (wave
setasas m (wave amplitude
amplitude is 3form)numerical
is 3 m) for numerical
simu-
simulation validation.
lation validation. FigureFigure 9 shows
9 shows the time
the time histories
histories of encountered
of encountered wave wave elevation
elevation in onein
one period
period measured
measured at Probe
at Probe 1 in
1 in comparisonwith
comparison withthe
thetheoretical
theoretical value
value corresponding to to
five different wave headings. According to
five different wave headings. According to the Airy the Airy wave theory, the theoretical value
theory, the theoretical value can can
be
be obtained
obtained byby trigonometric
trigonometric function
function operations.
operations. AsAs it
it is
is illustrated
illustrated inin Figure
Figure 8,
8, all
all the
the
wave amplitudes are close to 3 m in these cases, and there is no significant wave attenuation.
tenuated to avoid the reflection of waves at the boundary. Due to the different damping
of peaks and troughs, irregular elliptical boundaries appear. Moreover, the setting of
damping at the outlet has little effect on the wave near the hull and will not cause the
attenuation of the wave near the hull. The maximum and minimum theoretical values of
J. Mar. Sci. Eng. 2021, 9, 1151 wave elevation are 3 m and 6 m. The error between the maximum value and theoretical 9 of 23
value in the figure is caused by the size difference between the grid in the overset region
and the background region, which may lead to a certain deformation of the wave surface
near the interface. The maximum error, which is in the head wave, reaches 2% of the wave
The maximum error is around 7.3%, which is caused by the diffraction and reflection of
height.
waves by the trimaran advancing in waves.
10 10
Result from Probe 1 Theoretical value Result from Probe 1 Theoretical value
8 8
Wave height (m)

Wave height (m)
6 6
4 4
2 2
0 2 4 6 8 10 0 2 4 6 8 10
Time (s) Time (s)
(a) (b)
10 10 10
Result from Probe 1 Theoretical value Result from Probe 1 Theoretical value Result from Probe 1 Theoretical value
8 8 8
Wave height (m)
Wave height (m)

Wave height (m)
6 6 6
4 4 4
2 2 2
0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10
Time (s) Time (s) Time (s)
(c) (d) (e)
Figure 9. Wave elevation (λ/L = 1.0). (a) θ = 0◦ ; (b) θ = 15◦ ; (c) θ = 30◦ ; (d) θ = 45◦ ; (e) θ = 135◦ .
Figure 9. Wave elevation (λ/L = 1.0). (a) = 0°; (b) = 15°; (c) = 30°; (d) = 45°; (e) = 135°.
The initialized wave profiles with the trimaran model corresponding to the five
different wave headings are obtained and shown in Figure 10. It can be seen that the wave
shows an elliptical shape at the outlet boundary. This is because that wave damping is set at
the outlet boundary to eliminate the reflected wave. Wave damping is used at the boundary
of the flow field to damp and attenuate the waves near the selected boundary, so as to
reduce the oscillation near the boundary. The damping introduces vertical resistance to
the vertical motion. The waves generated by the hull can be damped and attenuated to
avoid the reflection of waves at the boundary. Due to the different damping of peaks and
troughs, irregular elliptical boundaries appear. Moreover, the setting of damping at the
outlet has little effect on the wave near the hull and will not cause the attenuation of the
wave near the hull. The maximum and minimum theoretical values of wave elevation are
3 m and 6 m. The error between the maximum value and theoretical value in the figure is
caused by the size difference between the grid in the overset region and the background
region, which may lead to a certain deformation of the wave surface near the interface.
The maximum error, which is in the head wave, reaches 2% of the wave height.
J. Mar. Sci. Eng. 2021, 9, 1151
x FOR PEER REVIEW 10of
10 of 23
(a) (b)
(c) (d)
(e)
Figure 10. Initialized

Figurewave profile. (a)
10. Initialized wave=profile. = 0◦(c)
0°; (b) (a)=θ15°; 15◦(d)
; (b) θ==30°; 30◦ ; (e)
= 45°;
; (c) θ = (d) θ == 45 ◦ ; (e) θ = 135◦ .
135°.
3.3. Numerical Method Validation in Head Wave

3.3. Numerical Method Validation in Head Wave
Due to the lack of experimental measurement for trimaran motions in oblique waves,
Due toCFD
the present the lack of experimental
method for trimaranmeasurement for trimaran
motion prediction motions
is validated in oblique waves,
by comparing it with
the present CFD method for trimaran motion prediction
the existing experimental measurement of trimaran motions in head regularis validated by comparing
waves. The it
with the existing experimental measurement of trimaran motions in
dimensionless results are compared with the experimental measurement and the 3-D po-head regular waves.
The dimensionless
tential flow theoreticalresults areby
results compared with
Chen et al. [9], the experimental
as shown measurement
in Figure 11. The errorand the
quanti-
3-D potential flow theoretical results by Chen et al.
fication study for the head wave case is presented in Table 5. [9], as shown in Figure 11. The error
quantification
It is foundstudy
from for the head
Figure 11 thatwave case is presented
the overall tendenciesinofTable 5. calculation and 3-D
the CFD
potential flow theoretical results show good agreement with the experimental measure-
ment, but the motion results obtained by the 3-D potential flow theory are smaller than
those from the CFD method, which can be also found in Table 5. As shown in Table 5,
error 1 denotes the differences between CFD results and experimental data, and error 2
denotes the differences between potential flow theoretical results and experimental data.
Compared with the experimental measurement, the calculating accuracy of the two meth-
ods decreases as wave length range increases, but the error 2 caused by potential flow
theory becomes more obvious, especially at λ/L > 1.3. This is because that the unsteady
ship–wave interactions are limited to linear assumption and forward speed approxima-
J. Mar. Sci. Eng. 2021, 9, 1151 11 of 23
tion is applied in the potential flow theory, while the CFD model can consider the nonlin-
ear interaction between wave and ship. Therefore, the CFD numerical method can be used
in the subsequent simulation of ship seakeeping and slamming behaviors.
1.4
1.8
1.2 1.6
Dimensionless pitch (φa/kζa）
Dimensionless Heave (Za/ζa）

1.4
1.0
1.2
0.8
1.0
0.6 0.8
0.4 0.6
Present CFD method 0.4

0.2
Present CFD method
3D potential flow theory 3D potential flow theory
0.2
Experimental measurement Experimental measurement
0.0
0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 0.0
0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75
λ/L λ/L
(a) (b)
Figure11.
Figure 11.Pitch
Pitchand
andheave
heaveresponses
responsesininhead
headwave.
wave.(a)
(a)Pitch;
Pitch;(b)
(b)Heave.
Heave.
Table
Table5.
5. Motion
Motionerror
errorquantification
quantification study
study for
for head
head wave
wave case.
case.
PresentPresent
CFD Method 3D 3D
Potential Flow Theory Experimental Data Error1 Error2
/ CFD Potential Flow
λ/L Pitch Method
Heave PitchTheory Heave ExperimentalPitch Data Heave Pitch Error 1 Heave Error 2Heave
Pitch
0.8 0.266
Pitch 0.216
Heave 0.268
Pitch Heave0.225 Pitch0.269 Heave 0.219 Pitch
1.12% Heave 1.37% 0.37%
Pitch 2.74%
Heave
0.9 0.425 0.465 0.4 0.475 0.45 0.485 5.56% 4.12% 11.11% 2.06%
0.8 0.266 0.216 0.268 0.225 0.269 0.219 1.12% 1.37% 0.37% 2.74%
1.0
0.9 0.585
0.425 0.673
0.465 0.55
0.4 0.643
0.475 0.45 0.62 0.4850.702 5.56%5.65% 4.12% 4.17% 11.29% 2.06%
11.11% 8.33%
1.1
1.0 0.755
0.585 1.029
0.673 0.71
0.55 0.938
0.643 0.62 0.8 0.7021.12 5.65% 5.62% 4.17% 8.15% 11.25% 16.29%
11.29% 8.33%
1.1
1.3 0.755
1.005 1.029
1.585 0.71
0.96 0.938
1.525 0.8 1.05 1.121.645 5.62% 4.29% 8.15% 3.65% 11.25%
8.57% 16.29%
7.29%
1.3
1.5 1.161.005 1.585
1.557 0.96
1.08 1.5251.35 1.05 1.2 1.645 1.69 4.29%
3.33% 3.65%
7.87% 8.57% 7.29%
10.00% 20.12%
1.5 1.16 1.557 1.08 1.35 1.2 1.69 3.33% 7.87% 10.00% 20.12%
1.7
1.7 1.146
1.146 1.385
1.385 1.02
1.02 1.181.18 1.22 1.22 1.5 1.5 6.07%
6.07% 7.67%
7.67% 16.39%
16.39% 21.33%
21.33%
1.9
1.9 1.091.09 1.2425
1.2425 0.98
0.98 1.081.08 1.16 1.16 1.38 1.38 6.03% 6.03% 9.96% 9.96% 15.52%
15.52% 21.74%
21.74%
2.1
2.1 1.075 1.1475
1.075 1.1475 0.95
0.95 1.025
1.025 1.1351.135 1.27 1.27 5.29% 5.29% 9.65% 9.65% 16.30%
16.30% 19.29%
19.29%
2.3
2.3 1.061.06 1.085
1.085 0.927
0.927 0.950.95 1.1271.127 1.22 1.22 5.94% 5.94% 11.07%
11.07% 17.75%
17.75% 22.13%
22.13%
2.5 1.045 1.0425 0.923 0.925 1.121 1.16 6.78% 10.13% 17.66% 20.26%
2.5 1.045 1.0425 0.923 0.925 1.121 1.16 6.78% 10.13% 17.66% 20.26%
It is found
4. Trimaran from Figure
Motion 11 that
in Oblique the overall
Regular Wavestendencies of the CFD calculation and 3-D
potential flow theoretical results show good agreement with the experimental measure-
4.1. Analysis of Trimaran Motions in Oblique Regular Waves
ment, but the motion results obtained by the 3-D potential flow theory are smaller than
those Itfrom
is confirmed that the CFD
the CFD method, whichnumerical
can be alsomethod
foundadopted
in Tablein5.this
As paper
showncan in simulate
Table 5,
trimaran motion responses in head regular waves well in Section 3. Thus,
error 1 denotes the differences between CFD results and experimental data, and error 2 the time-domain
motionsthe
denotes of the trimaranbetween
differences in oblique regularflow
potential waves will be analyzed
theoretical results and in this section. For
experimental the
data.
conditionswith
Compared of calculation time step
the experimental 0.05 s, each simulation
measurement, condition
the calculating accuracytookofabout 2 days
the two meth-to
obtain the trimaran motion and slamming simulation. In addition, the technical
ods decreases as wave length range increases, but the error 2 caused by potential flow specifica-
tions ofbecomes
theory the computing hardware
more obvious, used to at
especially conduct
λ/L > the
1.3. numerical simulation
This is because in unsteady
that the this paper
are Intel Core i7-9700F CPU with 16 GB RAM.
ship–wave interactions are limited to linear assumption and forward speed approximation
is applied in the potential flow theory, while the CFD model can consider the nonlinear
4.1.1. Influence
interaction of Wave
between waveHeading
and ship. Therefore, the CFD numerical method can be used in
the subsequent
Figure 12 simulation of ship
shows the time seakeeping
histories and slamming
of motion behaviors.
in oblique regular waves. The results
indicated that the motion signals at any wave heading present sinusoidal characteristics.
4. Trimaran Motion in Oblique Regular Waves
4.1. Analysis of Trimaran Motions in Oblique Regular Waves
It is confirmed that the CFD numerical method adopted in this paper can simulate
trimaran motion responses in head regular waves well in Section 3. Thus, the time-
domain motions of the trimaran in oblique regular waves will be analyzed in this section.
For the conditions of calculation time step 0.05 s, each simulation condition took about
2 days to obtain the trimaran motion and slamming simulation. In addition, the technical
J. Mar. Sci. Eng. 2021, 9, 1151 12 of 23
specifications of the computing hardware used to conduct the numerical simulation in this
paper are Intel Core i7-9700F CPU with 16 GB RAM.
4.1.1. Influence of Wave Heading

Figure 12 shows the time histories of motion in oblique regular waves. The results
Due to
indicated thethat
different initialsignals
the motion locations of the
at any hull
wave relativepresent
heading to the wave under
sinusoidal the different
characteristics.
wave headings, a phase shift will occur. Moreover, the pitch responses
Due to the different initial locations of the hull relative to the wave under the different of trimaran in bow
quarter waves, which are not much affected by wave heading, are
wave headings, a phase shift will occur. Moreover, the pitch responses of trimaran in bow larger than those in
stern
quarterquartering waves.
waves, which areUnlike pitchaffected
not much motion,bythe heave
wave motionare
heading, haslarger
a close correlation
than with
those in stern
wave heading.
quartering Within
waves. Unlikethepitch
rangemotion,
of bowthe quarter
heavewaves,
motionthe
has heave motion
a close increases
correlation withsignif-
wave
icantly
heading. asWithin
the wave the heading
range of bowanglequarter
increases. Moreover,
waves, the heave since oblique
motion wavessignificantly
increases will cause
transverse
as the waveincoming
heading flow anglevelocity
increases.to Moreover,
the hull, the liquid
since levelwaves
oblique difference between
will cause the left
transverse
and right-side hulls of the trimaran will occur. The hull pressure at the wave-forward
incoming flow velocity to the hull, the liquid level difference between the left and right-side side
is significantly
hulls greater
of the trimaran than
will that The
occur. at the wave
hull backward
pressure at theside, as shown in
wave-forward Figure
side 13.
is significantly
greater than that at
In addition, the wave
there is an backward
obvious phase side, as shown in Figure
relationship between13. roll and heave mo-
tionsIninaddition, there is
all the wave an obvious
heading phase
cases. Therelationship
roll curves between
always rolldifferand heave
from themotions
heave
in all the wave heading cases. The roll curves always differ from
curves by 1/4 cycle, that is, when the roll motion peaks, the heave motion is the aver- the heave curves by
1/4 cycle, that is, when the roll motion peaks, the heave motion
age value, and when the heave motion peaks, the roll motion is the average value. is the average value, and
whenindicates
This the heavethat motion
therepeaks, the roll
is a strong motion relationship
coupling is the averagebetween
value. This roll indicates
and heavethat in
there is a strong coupling relationship between roll and heave in oblique waves. Moreover,
oblique waves. Moreover, due to the side hull of the trimaran, large-amplitude roll
due to the side hull of the trimaran, large-amplitude roll motion may lead to a serious
motion may lead to a serious uneven draft of the two side hulls. At this moment, the
uneven draft of the two side hulls. At this moment, the whole ship needs to sink to a
whole ship needs to sink to a certain depth to maintain the balance of overall buoy-
certain depth to maintain the balance of overall buoyancy and gravity, which will bring the
ancy and gravity,
large-amplitude which
heave will bring the large-amplitude heave motion.
motion.
4 0.8
15° Pitch Roll Heave
2 0.4
Pitch/Roll (°)
Heave (m)
0 0.0
-2 -0.4
-4 -0.8
20 25 30 35 40 45 50 55
Time (s)
(a)
6 0.6
30° Pitch Roll Heave
3 0.3
Pitch/Roll (°)
Heave (m)
0.0
0
-0.3
-3
-0.6
-6 -0.9
-9 -1.2
30 35 40 45 50
Time (s)
(b)
Figure 12. Cont.

J. Mar. Sci. Eng. 2021, 9, 1151 13 of 23
1.0
4 45° Pitch Roll Heave 1.0
0.4
Heave (m)
Pitch/Roll (°)
Heave (m)
Pitch/Roll (°)
0 -0.2
-0.2
-4 -0.8
-4 -0.8
-8 -1.4
-8 -1.4
-12 -2.0
30
-12 35 40 45 50 -2.0
30 35 Time40(s) 45 50
Time (s)
(c)
(c)
0.5
2 135° Pitch Roll Heave0.0
0.0
Pitch/Roll (°)
Heave (m)
-2
Pitch/Roll (°)
Heave (m)
-2 -0.5
-0.5
-6 -1.0
-6 -1.0
-10 -1.5
-10 -1.5
-14 -2.0
30
-14 35 40 45 50 -2.0
30 35 40 45 50
Time (s)
Time (s)
(d)
(d)
FigureFigure
12. Time
12.histories of motion
Time histories in oblique
of motion regularregular
in oblique waves.waves.
(a) =(a)
15°; 15◦ ; (b)
θ =(b) = 30°; 30◦ ; (c)
θ = (c) = 45°; (d)◦ ; (d)= θ135°.
θ = 45 = 135◦ .
Figure 12. Time histories of motion in oblique regular waves. (a) = 15°; (b) = 30°; (c) = 45°; (d) = 135°.
Figure 13. Liquid level difference between the left and right-side hulls of the trimaran in oblique
Figure 13. Liquid level difference between the left and right-side hulls of the trimaran in oblique
waves.
Figure 13. Liquid level difference between the left and right-side hulls of the trimaran in
waves.
oblique waves.
4.1.2. Influence of Forward Speeds
4.1.2.Influence
4.1.2. InfluenceofofForward
ForwardSpeedsSpeeds
To investigate the influence of forward speed on trimaran motions in oblique waves,
ToToinvestigate
investigate
the time-domain theinfluence
the
motion influence
responsesofofforward
atforward speed
θ = 45°speed
are onontrimaran
displayedtrimaran motions
motions
in Figure Itinoblique
14.in isoblique waves,
waves,
seen that the
the
the time-domain
time-domain motion
motion responses
responses at
at θ =
= 45°
45 ◦ are
are displayed
displayed in
inFigure
Figure 14.
14. It is
It seen
is that
seen the
that
periods of time-domain results are different at different speeds, which are consistent with
θ
periods
the periods of time-domain
of time-domain results are
results different
are at
different different
at differentspeeds, which
speeds, are
which
their encounter periods. The change of velocity hardly affects the pitch amplitude, while consistent
are with
consistent
theirtheir
with encounter periods.
encounter The The
periods. change of velocity
change hardly
of velocity affects
hardly the pitch
affects amplitude,
the pitch while
amplitude,
J. Mar. Sci. Eng. 2021, 9, x1151
FOR PEER REVIEW 1414of
of 23
the increase in speed will significantly increase the amplitudes of roll and heave. Moreo-
the
ver,increase
whilethethe in speed
increase
emergence in will
phenomenasignificantly
speed will increase
ofsignificantly
the whole the amplitudes
increase
left-side hullthe
can be of roll and
amplitudes
observed heave.
ofwhen
roll theMoreo-
and heave.
speed
ver, the
Moreover,
reaches emergence
6.5the phenomena
emergence
m/s. However,phenomenaof the whole left-side
of the whole
when the forward speed hull can
left-side be observed
hullonly
is 5 m/s, can be
bow when
observed the
when
emergence speed
the
phe-
reaches
speed 6.5
reachesm/s.
6.5 However,
m/s. when
However, the
when forward
the speed
forward is 5
speed m/s,
is 5 only
m/s, bow
only emergence
bow
nomena of left-side hull occurs, not the bottom of side hull. Thus, when advancing in the phe-
emergence
nomena
phenomena
oblique of left-side
waves, the hull
of left-side occurs,
hull
trimaran not not
occurs,
should the bottom of side
the bottom
speed down hull.hull.
of avoid
to side Thus, when
the Thus, advancing
when rollinand
advancing
large-amplitude the
in
oblique
the waves,
oblique
heave motions. the
waves, trimaran
the should
trimaran shouldspeed
speeddown
down totoavoid
avoid the
thelarge-amplitude
large-amplitude roll
roll and
and
heave
heave motions.
4 1.0
4 1.0
0.4
0 0.4
(°)(°)
(m)
-0.2
Pitch/Roll
(m)
-0.2
Heave
Pitch/Roll
-4
Heave
-4 -0.8
-0.8
-8
-8 -1.4
-1.4
-12 -2.0
-1230 35 40 45 50-2.0
30 35 40 45 50
Time (s)
6.5 m/s Pitch 6.5 m/s Roll
Time
6.5 m/s Heave
(s)
56.5 m/s
m/s Pitch
Pitch 56.5 m/s
m/s Roll
Roll 56.5 m/s
m/s Heave
Heave
5 m/s Pitch 5 m/s Roll 5 m/s Heave
Figure 14. Time histories motion corresponding to different forward speeds (θ = 45°).
Figure 14. Time
Time histories motion corresponding to different forward
forward speeds
speeds (θ
(θ== 45 ◦ ).
45°).
4.2.
4.2. Comparison
Comparison of of Trimaran
Trimaran Motions
Motions in in Heading and Oblique Waves
Waves
4.2. Comparison of Trimaran Motions in Heading
Heading and
and Oblique
Oblique Waves
For
For aaabetter
betterunderstanding
understandingof ofthethemotion
motion behavior
behavior of trimaran
of trimaran in oblique
in oblique waves, tri-
waves,
maran For better
motion understanding
responses inin of
heading theandmotion
obliquebehavior
waves of
are trimaran
compared in oblique
in this waves,
section. tri-
Dif-
trimaran
maran from motion
motion responses heading and oblique waves are compared in this section. Dif-
fering
fering from theresponses
the headwave
head wavein heading
case,the
case, the and
oblique
oblique
oblique waves
wave
wave willwill
are compared
leadlead
to the
in this section.
to transverse
the transverse incoming
incoming
Dif-
flow.
fering
flow. from the
This means head wave
thatside
one sidecase, the oblique wave will lead to the transverse incoming
This means
flow. This that one
means that oneof the of
side hull
of
thewould
hull would
the hull would
be subject
be subject
be subject
to a continuous
to a continuous
totime load, the
a continuous
load, the am-
amplitude
load, the am-
plitude of which changes periodically. Figure 15 presents the
of which changes periodically. Figure 15 presents the time histories of motions at different histories of motions at
plitude
differentof which
wave changes
heading periodically.
angles. It is Figure
obvious 15
thatpresents
the roll the
motiontime ishistories
affected of
by motions
the wave at
wave heading angles. It is obvious that the roll motion is affected by the wave direction,
different
direction, wave heading
which0is almost angles.
0° in It is obvious
the head that the roll motion is affected by the wave
which is almost ◦ in the head wave. With wave. With in
the increase thetheincrease in the wave
wave heading angle,heading
the roll
direction,
angle, which
the roll is almostincreases
amplitude 0° in the head wave. With the increase in the wave heading
amplitude increases rapidly, which israpidly,
similar which is similar
to the law of heaveto the law
motion. of
Theheave motion.
peak-to-peak
angle,
The the roll amplitude
peak-to-peak value increases rapidly, which is ◦similar toheading
the law45° of heavewhichmotion.
value of heave motion is of
2.3heave
m in themotion
waveisheading
2.3 m in 45the wave
case, which is aroundcase, 2.2 times that is
The peak-to-peak
around 2.2 times value
that of of
the heave
head motion
wave. is 2.3
Moreover, m in
in the wave
oblique heading
wave cases, 45° case,
there arewhich
a largeis
of the head wave. Moreover, in oblique wave cases, there are a large number of phenomena
around
number 2.2 times
of phenomenathat of the head
of side hullwave. Moreover,
emergence in
and be oblique
green wave
water, in cases,
which there are a large
of side hull emergence and green water, which will discussed detailwill be discussed
in the following
number
in of
detailns. phenomena of side
in the following section ns. hull emergence and green water, which will be discussed
section
in detail in the following section ns.
1.0
1.0
0.2
0.2
Heave(m)
Heave(m)
-0.6
-0.6
-1.4
-1.4
-2.2
-2.230 35 40 45 50
30 35 40 45 50
Time (s)
Time (s)
0° 15° 30° 45° 135°
0° 15° 30° 45° 135°
(a)
(a)
Figure 15. Cont.
J. Mar. Sci. Eng. 2021, 9, 1151 15 of 23
2
Pitch(°)
-2
-4
30 35 40 45 50
Time (s)
0° 15° 30° 45° 135°
(b)
2
-2
Roll(°)
-6
-10
-14
30 35 40 45 50
Time (s)
0° 15° 30° 45° 135°
(c)
Figure15.
Figure 15.Time
Timehistories
historiesofofmotion
motionatatdifferent
differentwave
waveheading
headingangles
angles(v(v==6.5
6.5m/s).
m/s). (a)
(a) Heave;
Heave; (b)
(b) Pitch;
Pitch; (c)
(c) Roll.
Roll.
5. Slamming
5. Slamming Analysis
Analysis
It can
It can bebeseen
seenfrom
fromthe above
the aboveanalysis thatthat
analysis there is a is
there continuous load caused
a continuous by trans-
load caused by
verse incoming flow in oblique waves, which leads to uneven pressure
transverse incoming flow in oblique waves, which leads to uneven pressure distribution distribution on the
leftthe
on and right-side
left hulls of
and right-side theof
hulls trimaran, and even
the trimaran, full emergence
and even and and
full emergence green water
green of one
water of
side hull. Traditional studies on slamming of trimarans focus on numerical
one side hull. Traditional studies on slamming of trimarans focus on numerical simulation simulation or
experiment of water entry slamming by using a wedge with simplified
or experiment of water entry slamming by using a wedge with simplified main hull and main hull and side
hulls.
side However,
hulls. However, this this
approach, which
approach, onlyonly
which considers a single
considers period,
a single cannot
period, consider
cannot the
consider
actual motion state. Therefore, STAR-CCM+ is used to generate continuous
the actual motion state. Therefore, STAR-CCM+ is used to generate continuous waves for waves for
slamming response
slamming response comparison
comparisonin inmulti-period
multi-periodin inthis
thisstudy.
study.
5.1. Slamming
5.1. Slamming Pressure
Pressure in
in Oblique
Oblique Waves
Waves
This section
section focuses
focuseson onthetheslamming
slammingcharacteristics
characteristicsof of different
different positions
positions of the
of the tri-
trimaran corresponding
maran corresponding toto the
the speedofof6.5
speed 6.5m/s.
m/s.AsAs shown
shown in in Figure
Figure 16,16, when
when advancing
advancing in
in oblique
oblique waves,
waves, as the
as the portport
sideside is wave-forward
is wave-forward side,side, the pressures
the pressures on the onport
the side
portofside
the
of the hull
main mainarehull are greater
greater thanon
than those those on the starboard
the starboard side.
side. With theWith the increasing
increasing wave
wave heading
heading angle, the uneven pressure distribution on both sides of the trimaran
angle, the uneven pressure distribution on both sides of the trimaran is gradually obvious. is gradually
obvious.
The waveThe wave
impact toimpact to hull from
hull transfers transfers
bowfrom bow centerline
centerline area to starboard
area to starboard area in bow area in
quar-
bow quarterwhile
ter waves, waves,
thewhile
impacttheisimpact
mainlyisaround
mainly the
around
sternthe
in stern in stern quartering
quartering waves. waves.
Due to the roll motion, the pressure distributions of the side hulls and cross-decks are
becoming very complex. Firstly, there is a gradual difference in the pressure between the
left and right-side hulls. With the increasing of wave heading angle, the pressure on the
right-side hull decreases and the pressure on the left-side hull increases. This is because
that the emergence time of the right-side hull becomes longer. Secondly, the time histories
of slamming pressure on both side hulls will have a phase difference as the roll motion
increases. The time difference between the two slamming peaks will increase with the
increase in wave heading angle. Finally, as for the cross-decks, the water entry on the
wave-forward side and splash caused by roll motion lead to a certain pressure. However,
J. Mar. Sci. Eng. 2021, 9, 1151
as the cross-decks are neither directly impacted by incoming flow nor immersed much
16 of 23
below the free surface, the pressures of the cross-decks are still much smaller than those
of other parts of the trimaran.
80
Pressure(kPa)
60
40
20
0
30 35 40 45 50
Time (s)
15°-P4 15°-P6 30°-P4 30°-P6
45°-P4 45°-P6 135°-P4 135°-P6
(a)
80
60
Pressure(kPa)
40
20
0
30 35 40 45 50
Time (s)
15°-R3 15°-L3 30°-R3 30°-L3
45°-R3 45°-L3 135°-R3 135°-L3
(b)
60
50
40
Pressure(kPa)
30
20
10
0
30 35 40 45 50
Time (s)
15°-R4 15°-L4 30°-R4 30°-L4
45°-R4 45°-L4 135°-R4 135°-L4
(c)
Figure 16. Cont.

J. Mar. Sci. Eng. 2021, 9, 1151 17 of 23
15
10
Pressure(kPa)
-5
-10
30 35 40 45 50
Time (s)
15°-R7 15°-L7 30°-R7 30°-L7
45°-R7 45°-L7 135°-R7 135°-L7
(d)
Figure 16. Time
Figure histories
16. Time of slamming
histories of slammingpressure
pressureatatthe
theimportant positions.(a)(a)Centerline
important positions. Centerline of the
of the bowbow of main
of main hull; hull; (b) Side
(b) Side
of main hull; hull;
of main (c) Side hull;hull;
(c) Side (d) (d)
Cross-deck.
Cross-deck.
Due to theofroll
5.2. Comparison motion, Pressure
Slamming the pressure distributions
in Heading of the side
and Oblique hulls and cross-decks are
Waves
becoming very complex. Firstly, there is a gradual difference in the pressure between the
This
left andsection focuses
right-side hulls. on thethe
With difference
increasingbetween
of wavethe trimaran
heading slamming
angle, characteristics
the pressure on the
in head and oblique waves. The time histories of slamming pressure corresponding
right-side hull decreases and the pressure on the left-side hull increases. This is because to
different
that thewave headings
emergence time are presented
of the right-sidein Figure
hull 17. longer. Secondly, the time histories
becomes
of Firstly,
slamming since there is
pressure onlittle
bothroll
sidemotion in head
hulls will have awave,
phasethe pressure
difference as distributions
the roll motionon the
increases. The time difference between the two slamming peaks
port and starboard sides are completely consistent. The impact caused by the will increase withincoming
the
flowincrease in wave
is mainly heading angle.
concentrated Finally,
near the as for of
centerline thebow,
cross-decks, the waternear
so the pressure entry
theoncenterline
the
wave-forward side and splash caused by roll motion lead to a certain pressure. However,
is greater than that on both sides. Moreover, since the side hulls are located at the stern of
as the cross-decks are neither directly impacted by incoming flow nor immersed much
the trimaran, the main hull encounters waves at first and causes waves to break in head
below the free surface, the pressures of the cross-decks are still much smaller than those of
wave.
otherWhen
parts the wave
of the transmits to the side hulls, the energy that has decayed is not high
trimaran.
enough to cause the side hulls and cross-decks to slam. Thus, the traditional slamming
5.2. Comparison
experiment of Slamming
for local Pressure on
model focuses in Heading
the head andwave
Oblique Waves
case, in which the weakening of the
main hull Thistosection
the waves
focusesalong
on the ship length between
difference in the actual navigation
the trimaran is ignored.
slamming characteristics
in Secondly,
head and oblique waves. of
the slamming Themaintimehullhistories of slamming
is caused pressure corresponding
by the longitudinal/vertical to
motion in
different wave headings are presented in Figure 17.
the head wave, while the emergence of one side hull caused by roll motion is the main
cause ofFirstly, since there
slamming is little roll
in oblique waves.motion in head
Taking thewave,
case the pressure
of 45° as andistributions
example, at onthe
the same
port and starboard sides are completely consistent. The impact caused by the incoming
incoming flow speed, the pressure of the side hull at the wave-forward side in the case of
flow is mainly concentrated near the centerline of bow, so the pressure near the centerline
45°isisgreater
1.57 times thaton
than that of both
the head
sides.wave.
Moreover, since the side hulls are located at the stern of
Finally, the slamming of
the trimaran, the main hull encounters side hulls in head
waves at firstwaves requires
and causes waves serious seainconditions;
to break head
however, the roll motion caused by gentle oblique wave conditions
wave. When the wave transmits to the side hulls, the energy that has decayed is not may lead to high
the slam-
ming phenomenon
enough to cause the of side
sidehulls
hulls. Compared
and cross-decks with the main
to slam. Thus,hull slamming,slamming
the traditional the slamming
experiment
of side hulls and for local modelflow
incoming focuses
willoncause
the head wave case, into
the cross-decks which
bearthe weakening
additional of the mo-
bending
main hull to the waves along ship length in the actual navigation
ments. This is why the trimaran cross-decks are more likely to be damaged in waves. is ignored.
Therefore,Secondly,
comparedthe slamming
to the head of main hull case
wave is caused
withbythe thesame
longitudinal/vertical motionsail in
sea state, trimarans
in the head wave, while the emergence of one side hull caused by roll motion is the main
oblique waves with greater potential safety hazards. ◦
cause of slamming in oblique waves. Taking the case of 45 as an example, at the same
incoming flow speed, the pressure of the side hull at the wave-forward side in the case of
45◦ is 1.57 times that of the head wave.
J.J.Mar.
Mar.Sci.
Sci. Eng. 2021, 9,
Eng. 2021, 9, 1151
x FOR PEER REVIEW 18 18
of of
2323
80
Pressure(kPa)
60
40
20
0
30 35 40 45 50
Time(s)
0° -R3 15°-R3 30°-R3 45°-R3 135°-R3
(a)
80
65
Pressure(kPa)
50
35
20
-10
30 35 40 45 50
Time(s)
0°-L3 15°-L3 30°-L3 45°-L3 135°-L3
(b)
60
50
Pressure(kPa)
40
30
20
10
0
30 35 40 45 50
Time(s)
0°-R4 15°-R4 30° -R4 45° -R4 135° -R4
(c)
40
Pressure(kPa)
25
10
-5
30 35 40 45 50
Time (s)
0°-L4 15°-L4 30°-L4 45°-L4 135°-L4
(d)
Figure 17.
Figure 17. Time
Time histories
histories of
of slamming
slamming pressure
pressure corresponding
correspondingto todifferent
differentwave
waveheadings.
headings.(a)
(a)Starboard
Starboardside ofof
side main hull;
main hull;
(b) Port side of main hull; (c) Right-side hull; (d) Left-side hull.
(b) Port side of main hull; (c) Right-side hull; (d) Left-side hull.
J. Mar. Sci. Eng. 2021, 9, 1151 19 of 23
Finally, the slamming of side hulls in head waves requires serious sea conditions; how-
5.3.
ever,Relationship between
the roll motion Trimaran
caused Motions
by gentle and Slamming
oblique Pressuremay
wave conditions in Oblique
lead toWaves
the slamming
The simulated
phenomenon of sideimpact
hulls. pressures
Comparedand motions
with the mainat the
hullforward
slamming,perpendicular
the slamming in oblique
of side
waves
hulls andare incoming
presentedflowin Figure 18. the cross-decks to bear additional bending moments.
will cause
This It
isiswhy
seenthe
that the slamming
trimaran pressures
cross-decks are generally
are more likely to beperiodical
damaged andinsteady
waves.during this
Therefore,
compared
period. Thetoslamming
the head wave case with
pressures the same
are closely sea state,
related trimarans
to trimaran sail in
heave obliqueand
motion, waves
the
with greater
peak time of potential
the two issafety hazards.
just the opposite. The peak time of the port and starboard of the
main hull is around the peak time of the bow centerline, but it does not coincide. There is
5.3.obvious
no Relationship between Trimaran
relationship between Motions
the peakand Slamming
time of side Pressure
hulls andin the
Oblique
roll Waves
motion, which
means Thethesimulated impact pressures
pressure distribution on theand motions
side hulls isatnot
theonly
forward perpendicular
related in oblique
to the roll motion, but
waves are presented in Figure
also affected by the heave motion. 18.
4 100
80
0
Pressure(kPa)
Pitch/Roll(°)
60
-4 40
20
-8
0
-12 -20
30 35 40 45 50
Pitch Roll Time(s)

P6 R3 L3 R7 L7
(a)
3 100
0 80
Pressure(kPa)
Pitch/Roll(°)
60
-3
40
-6
20
-9 0
-12 -20
30 35 40 45 50
Time(s)
Pitch Roll
M4 R3 L3 R7 L7
(b)
1.0 100
0.5 80
Pressure(kPa)
0.0 60
Heave(m)
-0.5 40
-1.0 20
-1.5 0
-2.0 -20
30 35 40 45 50
Heave Time (s)
M6 R3 L3 R7 L7
(c)
Figure 18. Cont.

0.4 80
J.
J. Mar.
Mar. Sci.
Sci. Eng.
Eng. 2021, 9, 1151 20 of
of 23
0.0 9, x FOR PEER REVIEW
2021, 20 23
Pressure (kPa)
60
-0.4
Heave (m)
-0.8 40
-1.2
0.4 8020
-1.6
0.0
Pressure (kPa)
-2.0 600
-0.4
Heave (m)
30 35 40 45 50
-0.8 40
Time (s)
-1.2 Heave
20
M6 R3 L3 R7 L7
-1.6
(d) 0
-2.0
Figure 18.30Time histories of motion
35and slamming pressure.
40(a) Roll, pitch and slamming
45 pressure (θ = 45°); (b)
50 Roll, pitch
and slamming pressure (θ = 135°); (c) Heave and slamming pressure (θ = 45°); (d) Heave and slamming pressure (θ = 135°).
Time (s)
Heave
M6 R3 In addition,
L3 R7withinL7the range of bow quarter waves, with the increasing of wave
direction angle, the amplitude (d) of rolling motion will increase. Taking the wave heading
Figure 18.
Figure 18. Time histories of
Time histories 45° case
of motion
motion andas
and an example,
slamming
slamming the(a)
pressure.
pressure. green
(a) Roll,water
Roll, pitch and
pitch phenomenon
and slamming pressure
slamming for trimaran
pressure (θ== 45
(θ ◦is
45°); depicted
(b)
); (b) in Figure
Roll, pitch
Roll, pitch
and slamming
and slamming pressure
pressure (θ 19.
(θ== 135
135°);When
(c) Heave
◦ ); (c) the
Heave and bow experiences
and slamming
slamming pressure a wave
pressure (θ crest,
(θ== 45
45°); the green
(d) Heave
◦ ); (d) Heave and water
and slammingphenomenon
slamming pressure
pressure (θ on the
(θ== 135 foredeck
135°).
◦ ).
will occur, as shown in Figure 19a. Then bow experiences a wave trough, the green water
In
It isaddition,
on the foredeck within
seen thatdisappears. the range
the slamming At this of boware
moment,
pressures quarter
the waves,
stern
generally enters with the
the wave
periodical andincreasing
crest
steady area of wave
and
during green
this
direction
water The
period. onangle,
the the amplitude
afterdeck
slamming occurs,of
pressures asrolling
are closelymotion
shown in Figure
related will
to19b.increase.heave
trimaran Taking the wave
motion, and theheadingpeak
45°
timecase as an
ofAlthough
the twoexample,
isthe the
justtransverse
the green
opposite. water
incoming
The peakphenomenon
flowtime velocity
of thefor trimaran
inport
the and
wave isheading
depicted
starboard in
theFigure
of45°and main 135°
19.
hullWhen
cases has
is around the bow
the experiences
almost the same
peak ofa wave
time level,the the
bow crest,
roll the greenbut
amplitude
centerline, water phenomenon
corresponding
it does to θ
not coincide. on=the 45° foredeck
There is islarger,
no
will occur,
obvious
which can as shown
relationship
be observed in Figure
between 19a.
the peak
in Figures Then bow
time of
above. experiences
Thisside hulls that
means a wave
and the roll trough,
motion,
trimaran the green water
whichsensitive
is more means
ontothe
the foredeck
pressure
motion caused disappears.
distribution
by wave onAtthethis
impact moment,
side
onhulls
the bow.isthe
not stern enters
onlythe
Since areathe
related oftowave crest
theforedeck
the roll area
motion, and
is muchbutgreen
also
larger
affected
water
than on by
that the
theof heave
afterdeck
the motion.
occurs,
afterdeck, as shown
when in Figureflow
the incoming 19b.velocity reaches a certain value, the
In addition,
Although
amount of green thewithin
water thethe
transverse
on range
incomingof bow
afterdeck flow quarter
caused by waves,
velocity sternin the with
wave
quartering theheading
increasing
waves of
be wave
45°and
will 135°
signifi-
direction
cases
cantly angle,
hasgreater
almostthanthe
theamplitude
sameonlevel,
that theof foredeck
rolling
the rollmotioncausedwill
amplitude increase.
by corresponding
bow quarter Taking tothe
waves. θ =wave
As heading
45°a is larger,
result, the
◦ case as an example, the green water phenomenon for trimaran is depicted in Figure 19.
45peak
which can be observed in Figures above. This means that the
value of roll motion caused by stern quartering waves is greater, which is different trimaran is more sensitive
When
tofrom
motion the bow
caused
monohull experiences
by wave
ships aiswave
andimpact shown crest,
on in the
the bow.
the green
Since
figures water
the
above. phenomenon
area on the
of the foredeck
Although the isforedeck
heave much will
larger
amplitudes
occur,
than
in the as
that shown
waveof the in Figure
afterdeck,
heading 45°and19a.
when Then
135° the bow
incoming
cases experiences a wave
flowdifference,
have little trough,
velocity reachesthe motion the green
a certain water
time value,
histories onin
the
the foredeck
amount
stern of green
quartering disappears.
water
wavesonare At
thethis moment,
afterdeck
below inthe
thosecaused bow stern
by enters
stern
quarter the wave
quartering
waves, which crest
wavesmeansarea and
willthat green
be signifi-
the draft
water
cantly on the afterdeck
greater
will increase than that
when theoccurs,
on theasforedeck
trimaran shown
sails ininstern
Figure
caused by19b.
bow quarter
quartering waves. waves. As a result, the
peak value of roll motion caused by stern quartering waves is greater, which is different
from monohull ships and is shown in the figures above. Although the heave amplitudes
in the wave heading 45°and 135° cases have little difference, the motion time histories in
stern quartering waves are below those in bow quarter waves, which means that the draft
will increase when the trimaran sails in stern quartering waves.
(a) (b)
Figure
Figure 19.19.Phenomena
Phenomenaofof green
green water
water corresponding
corresponding toto
θ =θ 45 ◦ . (a)
= 45°. (a)Green
Greenwater
waterononforedeck;
foredeck;(b)
(b)Green
Green water
water onon afterdeck.
afterdeck.
Although the transverse

As concluded from Sectionincoming
4, due flow
to thevelocity in the
side hulls wave
of the heading
trimaran, 45◦ and 135◦
large-amplitude
cases ◦
roll has almost
motion maythelead
sameto level, the rolluneven
the serious amplitude corresponding
draft of the two side to θhulls.
= 45 Taking
is larger,the
which
wave
(a)can be observed in figures above. This means that the trimaran (b) is more sensitive to motion
caused
Figure 19. Phenomena of green bycorresponding
water wave impacttoon θ =the bow.
45°. Sincewater
(a) Green the area of the foredeck
on foredeck; (b) Greeniswater
muchon larger than that
afterdeck.
of the afterdeck, when the incoming flow velocity reaches a certain value, the amount of
greenAswater on the afterdeck
concluded from Sectioncaused by stern
4, due to thequartering
side hullswaves
of thewill be significantly
trimaran, greater
large-amplitude
thanmotion
roll that onmay
the foredeck caused
lead to the by bow
serious quarter
uneven draftwaves.
of the As
twoa side
result, the peak
hulls. Taking value
theof roll
wave
J. Mar. Sci. Eng. 2021, 9, 1151 21 of 23
J. Mar. Sci. Eng. 2021, 9, x FOR PEERmotion

REVIEWcaused by stern quartering waves is greater, which is different from monohull21ships
of 23
and is shown in the figures above. Although the heave amplitudes in the wave heading
45◦ and 135◦ cases have little difference, the motion time histories in stern quartering waves
are below45°
heading those inas
case bowan quarter
example, waves,
whenwhich means
the rolling that is
angle theupdraft willthe
to 10°, increase when the
phenomenon of
trimaran sails in stern quartering waves.
full emergence of one side hull will occur, as shown in Figure 20.
As concluded
Therefore, whenfrom Section 4,
advancing indue
sterntoquartering
the side hulls of the
waves, thetrimaran, large-amplitude
trimaran always maintains
roll
the motion may lead roll
large-amplitude to the serious
angle anduneven
sinkingdraft of Thus,
range. the two side hulls.
compared toTaking
the bowthequarter
wave
heading
waves with45◦ the
casesame
as anangle,
example, when thesails
the trimaran rolling anglequartering
in stern
◦ , the phenomenon of
is up to 10waves with greater po-
full emergence
tential of one side hull will occur, as shown in Figure 20.
safety hazards.
(a) (b)
Figure20.
Figure 20.Full
Fullemergence
emergenceof
ofthe
theleft-side
left-sidehull.
hull.(a)
(a)View
Viewabove
abovethe
thefree
freesurface;
surface;(b)
(b)View
Viewbelow
belowthe
thefree
freesurface.
surface.
Therefore, when advancing in stern quartering waves, the trimaran always maintains
6. Conclusions
the large-amplitude roll angle
This work studies and sinking
the motion range.
slamming Thus, compared
characteristics to the bow
of a trimaran in quarter
oblique
waves with the same angle, the trimaran sails in stern quartering waves with greater
regular waves based on the CFD method. The slamming and green water characteristics
potential safety hazards.
of the trimaran in oblique waves are also analyzed. The following conclusions are drawn
6.according to this work:
Conclusions
(1) This
Sincework
the side hullsthe
studies aremotion
locatedslamming
at the stern of the trimaran,
characteristics of a the main hull
trimaran encounters
in oblique reg-
waves at first and causes waves to break in head waves.
ular waves based on the CFD method. The slamming and green water characteristics When wave transmits to side
of
hulls, the
the trimaran inenergy
obliquehas decayed
waves and analyzed.
are also the slamming The of the side conclusions
following hulls and cross-decks
are drawnis
difficult
according to occur.
to this work: Compared with the traditional slamming experiment in the head
wave, the slamming study of trimarans in oblique waves presents further practical
(1) Since the side hulls are located at the stern of the trimaran, the main hull encounters
significance;
waves at first and causes waves to break in head waves. When wave transmits to side
(2) When advancing in oblique waves, trimaran will have a serious roll motion that may
hulls, the energy has decayed and the slamming of the side hulls and cross-decks is
possess a coupling relationship with the heave motion. Due to the larger mass of
difficult to occur. Compared with the traditional slamming experiment in the head
green water
wave, on the afterdeck,
the slamming the roll peak
study of trimarans value caused
in oblique waves by stern quartering
presents waves
further practical
will be greater,
significance; which lead to the full emergence of one side hull and green water;
(3) When
(2) The oblique waves
advancing incan lead waves,
oblique to uneven pressure
trimaran willdistribution
have a serious on therollleft and right-side
motion that may
hulls of the trimaran. At the same incoming flow speed,
possess a coupling relationship with the heave motion. Due to the larger the pressure of side hullofat
mass
wave-forward side in the case of 45° is 1.57 times of that in head
green water on the afterdeck, the roll peak value caused by stern quartering waves wave;
(4) will
The bedifferences in slamming
greater, which characteristics
lead to the full emergence of the trimaran
of one side hullin heading
and green and oblique
water;
waves are caused by rolling motion and transverse incoming
(3) The oblique waves can lead to uneven pressure distribution on the left and right-side flow. When the trima-
hulls of the trimaran. At the same incoming flow speed, the pressure of side hullmo-
ran moves in a head wave, the slamming is caused by the longitudinal/vertical at
tion. Due to theside
wave-forward weakening
in the case of waves
of 45◦ isby1.57
thetimes
main ofhull, the
that inside
headhullswave;and cross-decks
needdifferences
(4) The high enough sea state tocharacteristics
in slamming encounter slamming. Meanwhile,
of the trimaran in the oblique
in heading waves
and oblique
with the
waves areequivalent sea state,
caused by rolling the slamming
motion and transverse of theincoming
side hulls caused
flow. When bythe
thetrimaran
joint in-
fluencein
moves ofaroll
headmotion
wave,and the heave motions
slamming is more
is caused byobvious. Moreover, the splash
the longitudinal/vertical and
motion.
partial water entry in oblique waves lead to a certain pressure
Due to the weakening of waves by the main hull, the side hulls and cross-decks need on the cross-decks;
(5) high
Whenenough
advancing in stern
sea state quartering
to encounter waves, the
slamming. trimaran always
Meanwhile, maintains
in the oblique the large-
waves with
amplitude roll angle and sinking range. Thus, compared to the bow quarter waves
with the same angle, the trimaran sails in stern quartering waves with greater poten-
tial safety hazards.
J. Mar. Sci. Eng. 2021, 9, 1151 22 of 23
the equivalent sea state, the slamming of the side hulls caused by the joint influence
of roll motion and heave motions is more obvious. Moreover, the splash and partial
water entry in oblique waves lead to a certain pressure on the cross-decks;
(5) When advancing in stern quartering waves, the trimaran always maintains the large-
amplitude roll angle and sinking range. Thus, compared to the bow quarter waves
with the same angle, the trimaran sails in stern quartering waves with greater potential
safety hazards.
It is noted that although a detailed validation study for head wave cases is provided
in this paper, an uncertainty quantification study for any one of the oblique wave cases
would improve confidence in the results. Thus, seakeeping and slamming experiments of
the trimaran in oblique waves are needed in the future for further validation.
Author Contributions: Conceptualization, X.L. and Z.C.; methodology, X.L.; investigation, Z.C.;
software, X.L.; validation, Z.C.; formal analysis, M.D.; data curation, M.D.; writing—original draft
preparation, X.L.; writing—review and editing, Z.C.; supervision, H.G.; project administration,
H.G.; funding acquisition, Z.C. All authors have read and agreed to the published version of the
manuscript.
Funding: This research was funded by the financial support from the Shandong Provincial Natural
Science Foundation, China, grant number ZR2020ME262 and the Open Fund of State Key Laboratory
of Coastal and Offshore Engineering, Dalian University of Technology, grant number LP2001.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: All data, models, or code generated or used during the study are
available from the corresponding author by request.
Conflicts of Interest: The authors declared that they have no conflict of interest to this work.
References
1. Andrews, D.J.; Zhang, J.W. Trimaran Ships. The Configuration for the Frigate of the Future. Nav. Eng. J. 1995, 107, 77–94.
[CrossRef]
2. Ren, H.L.; Chen, L.L.; Feng, G.Q.; Duan, F.; Yu, P.Y. Cross-Deck Slamming Load Calculation of Trimaran Based on Modified Plate
Slamming Theory. Appl. Mech. Mater. 2014, 477, 325–329. [CrossRef]
3. Nikfarjam, M.; Yaakob, O.B.; Seif, M.S.; Koto, J. Investigation of wedge water-entry under symmetric impact loads by experimental
tests. Lat. Am. J. Solids Struct. 2017, 14, 861–873. [CrossRef]
4. Zong, Z.; Sun, Y.; Jiang, Y.C.; Sun, T.Z.; Yu, Y.Q. Evolution of slamming load and flow field in water-entry process of trimaran
ship section. Ocean Eng. 2020, 205, 107319. [CrossRef]
5. Ghadimi, P.; Nazemian, A.; Sheikholeslami, M. Numerical simulation of the slamming phenomenon of a wave-piercing trimaran
in the presence of irregular waves under various seagoing modes. Proc. Inst. Mech. Eng. Part M J. Eng. Marit. Environ. 2019, 233,
1198–1211. [CrossRef]
6. Wu, J.X.; Sun, Z.; Jiang, Y.C.; Zhang, G.Y.; Sun, T.Z. Experimental and numerical study of slamming problem for a trimaran hull.
Ships Offshore Struct. 2021, 16, 46–53. [CrossRef]
7. Sun, Z.; Deng, Y.Z.; Zou, L.; Jiang, Y.C. Investigation of trimaran slamming under different conditions. Appl. Ocean Res. 2020, 104,
102316. [CrossRef]
8. Tang, H.Y.; Ren, H.L.; Li, H. Experimental Investigation of Wave-Induced Hydroelastic Vibrations of Trimaran in Oblique
Irregular Waves. Shock Vib. 2016, 2016, 8794560. [CrossRef]
9. Chen, Z.Y.; Gui, H.B.; Dong, P.S.; Yu, C.L. Numerical and experimental analysis of hydroelastic responses of a high-speed trimaran
in oblique irregular waves. Int. J. Nav. Archit. Ocean Eng. 2019, 11, 409–421. [CrossRef]
10. Jiao, J.L.; Huang, S.X.; Carlos, G.S. Numerical investigation of ship motions in cross waves using CFD. Ocean Eng. 2021, 223,
108711. [CrossRef]
11. Huang, S.X.; Jiao, J.L.; Chen, C.H. CFD prediction of ship seakeeping behavior in bi-directional cross wave compared with in
uni-directional regular wave. Appl. Ocean Res. 2021, 107, 102426. [CrossRef]
12. Lin, Y.; Ma, N.; Gu, X.C. Potential-flow and CFD investigations of bow-flare slamming on a container ship in regular heading
waves. Ocean Eng. 2021, 219, 108278. [CrossRef]
13. STAR CCM + UserGuide_14.04. Siemens PLM Software. 2019. Available online: https://www.plm.automation.siemens.com/
global/en/products/simcenter/STAR-CCM.html (accessed on 12 June 2021).
J. Mar. Sci. Eng. 2021, 9, 1151 23 of 23
14. ITTC. Uncertainty Analysis in CFD Verification and Validation, Methodology and Procedures; ITTC–Recommended Procedures and
Guidelines, ITTC: Tokyo, Japan, 2021.
15. Eça, L.; Hoekstra, M.A. Procedure for the Estimation of the Numerical Uncertainty of CFD Calculations Based on Grid Refinement
Studies. J. Comp. Phys. 2014, 262, 104–130. [CrossRef]
16. Jiao, J.L.; Huang, S.X.; Carlos, G.S. Viscous fluid-flexible structure interaction analysis on ship springing and whipping responses
in regular waves. J. Fluids Struct. 2021, 106, 103354. [CrossRef]
17. Field, P.L. Comparison of RANS and Potential Flow Force Computations for the ONR Tumblehome Hullform in Vertical Plane
Radiation and Diffraction Problems. Master’s Thesis, Virginia Polytechnic Institute and State University, Blacksburg, VA, USA,
2013.
18. Hashimoto, H.; Yoneda, S.; Omura, T.; Umeda, N.; Matsuda, A.; Stern, F.; Tahara, Y. CFD prediction of wave-induced forces on
ships running in irregular stern quartering seas. Ocean Eng. 2019, 188, 106277. [CrossRef]
19. Romanowski, A.; Tezdogan, T.; Turan, O. Development of a CFD methodology for the numerical simulation of irregular sea-states.
Ocean Eng. 2019, 192, 106530. [CrossRef]
20. Tezdogan, T.; Demirel, Y.K.; Kellett, P.; Khorasanchi, M.; Incecik, A.; Turan, O. Full-scale unsteady RANS CFD simulations of ship
behaviour and performance in head seas due to slow steaming. Ocean Eng. 2015, 97, 186–206. [CrossRef]
21. Song, S.S.; Kim, S.H.; Paik, K.J. Determination of linear and nonlinear roll damping coefficients of a ship section using CFD.
Brodogradnja 2019, 70, 17–33. [CrossRef]
22. Ismail, B.C. Procedure for Estimation and Reporting of Uncertainty Due to Discretization in CFD Applications. J. Fluids Eng.
2008, 130, 078001.
23. Stern, F.; Wilson, R.V.; Coleman, H.V.; Paterson, E.G. Comprehensive approach to verifcation and validation of CFD simulations-
Part 1: Methodology and procedures. J. Fluids Eng. Transcr. ASME 2001, 123, 793–802. [CrossRef]

Jmse 09 01151 v2

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Jmse 09 01151 v2

Uploaded by

Copyright:

Available Formats

Journal of

 Keywords: CFD; seakeeping; slamming; trimaran; oblique waves

Citation: Liao, X.; Chen, Z.; Gui, H.;

J. Mar. Sci. Eng. 2021, 9, 1151. https://doi.org/10.3390/jmse9101151 https://www.mdpi.com/journal/jmse

2. Numerical Model Set-Up

2.1. Ship Hull and Simulation Conditions

Item Item ValueValue

J. Mar. Sci. Eng. 2021, 9, 1151 4 of 23

Figure 2. Water volume fraction profile around the hull surface.

2.3. Fluid Domain and Boundary Conditions

J. Mar. Sci. Eng. 2021, 9, 1151 8 of 23

Table 4. Grid uncertainty analysis summary.

The calm water

Wave height (m)

Wave height (m)

(c) (d) (e)

Figure 10. Initialized

3.3. Numerical Method Validation in Head Wave

Dimensionless Heave (Za/ζa）

Present CFD method 0.4

4.1.1. Influence of Wave Heading

Figure 12. Cont.

Figure 16. Cont.

Pitch Roll Time(s)

Figure 18. Cont.

Although the transverse

J. Mar. Sci. Eng. 2021, 9, x FOR PEERmotion

You might also like

Keywords: CFD; seakeeping; slamming; trimaran; oblique waves