applied

sciences
Article
Numerical Study on Airfoil Aerodynamics in Proximity to
Wavy Water Surface for Various Amplitudes
Xing’an Liu, Dongli Ma, Muqing Yang *, Yang Guo and Haode Hu

School of Aeronautical Science and Engineering, Beihang University, Beijing 100191, China;
[email protected] (X.L.); [email protected] (D.M.); [email protected] (Y.G.);
[email protected] (H.H.)
* Correspondence: [email protected]

Featured Application: This study is meaningful for the design and flight safety of the wing–in–
ground craft.

Abstract: Wing–in–ground crafts will face waves with different amplitudes when flying over the
ocean, and the high amplitude situations are especially lack of exploration. Hence, the aerodynamic
characteristics of the NACA 4412 airfoil in proximity to the wavy water surface for various wave
amplitudes are inspected in this paper. By solving Navier–Stokes equations, the lift coefficients of
the airfoil when the angle of attack ranges from 0◦ to 4◦ are obtained. The results show that the
fluctuation amplitudes of aerodynamic coefficients increase remarkably with successive increases
in the wave amplitude and might threaten flight safety. The flow fields at 0◦ with low and high
wave amplitudes are investigated. It is revealed that the upward movement of the water surface is



the critical factor for the change of aerodynamics, and the mechanism varies with different wave

 amplitudes. Comparison of the flow fields at 0◦ and 2◦ further indicates that the influence of high
Citation: Liu, X.; Ma, D.; Yang, M.; amplitude waves depends on the distance between the leading edge of the airfoil and the water
Guo, Y.; Hu, H. Numerical Study on surface. This study discovers the reasons for the different aerodynamic characteristics under various
Airfoil Aerodynamics in Proximity to wave amplitudes and angles of attack, and is of great value for the design of wing–in–ground crafts.
Wavy Water Surface for Various
Amplitudes. Appl. Sci. 2021, 11, 4215. Keywords: wing–in–ground craft; ground effect; wavy water surface; various wave amplitudes;
aerodynamic characteristics; flow mechanism
https://doi.org/10.3390/app11094215

Academic Editor: Artur Tyliszczak

1. Introduction
Received: 17 March 2021
Accepted: 3 May 2021 When the aircraft flies in close proximity to an underlying surface, its lift–to–drag ratio
Published: 6 May 2021 will increase due to the ground effect. Wing–in–ground (WIG) crafts are designed to utilize
the favorable influence of the ground effect for higher operating efficiency and have been
Publisher’s Note: MDPI stays neutral developed by many countries [1]. Compared to the conventional aircraft, the WIG craft has
with regard to jurisdictional claims in some unique features in the application environment and aerodynamic characteristics. The
published maps and institutional affil- flow physics is also more complicated. To fully exploit the benefit of the ground effect on
iations. the WIG craft, a considerable number of studies have been conducted such as the research
on two–dimensional airfoils in the static ground effect [2–5], airfoils in the dynamic ground
effect [6,7] and three–dimensional aircrafts or wings in proximity to the ground [8–11].
Although extensive researches on the ground effect have been carried out, most of
Copyright: © 2021 by the authors. them considered the underlying surface as a rigid flat ground. However, considering
Licensee MDPI, Basel, Switzerland. that the WIG craft is designed to fly at a fairly low altitude, the ocean is a more proper
This article is an open access article environment for application without the danger of encountering hills, trees, or other
distributed under the terms and obstacles on the land. Therefore, WIG crafts usually cruise on the water surface for
conditions of the Creative Commons transportation. The water surface is deformable and free, therefore it cannot be substituted
Attribution (CC BY) license (https:// absolutely by ground for research. When the WIG craft flies over the ocean surface, waves
creativecommons.org/licenses/by/ often occur and sometimes have high amplitudes. Therefore, it is necessary to explore the
4.0/).

Appl. Sci. 2021, 11, 4215. https://doi.org/10.3390/app11094215 https://www.mdpi.com/journal/applsci

Appl. Sci. 2021, 11, 4215 2 of 19

aerodynamics of WIG crafts in proximity to the wavy water surface, especially for various
wave amplitudes.
To eliminate the inconsistency between the actual flight situation and the theoretical
research, numerous studies have attempted to explore the impact of the free surface on
the ground effect. Among them, Barber [12] investigated the deformation of the water
surface under WIG crafts and showed that the deformation was negligible in the two–
dimensional situation. However, three–dimensional CFD results suggested that the water
surface deformation behind the wing could not be neglected, which was probably caused
by wingtip vortices instead of the pressure distribution beneath the wing. Zong [13]
and Liang [14] took use of the lifting line theory to reveal lifting mechanisms of two–
dimensional and three–dimensional WIG crafts. The results showed that the free surface
deformation significantly affects the WIG craft aerodynamics at a low Froude number,
especially for the low altitude. But when the Froude number is high enough, the water free–
surface can be regarded as a rigid wall. Bal [15] proposed a modified iterative boundary
element numerical method for two–dimensional airfoils and three–dimensional wings
moving steadily over the water free–surface. It was demonstrated that the free surface
generated Kelvin wave deformation due to the ground effect, which might cause substantial
effects on the WIG craft performance. Therefore, it is necessary to consider the free surface
effects during the design of the WIG craft.
Since waves over the ocean surface complicate the flow field around WIG crafts, more
recent attention has been focused on the ground effect over the wavy ground. Im and
Chang [16] investigated aerodynamics of a NACA 6409 airfoil over the wavy ground by
solving the Euler equations. Numerical simulations were conducted under various ride
heights, wave amplitudes and wavelengths, and the results showed that the wavelength
affects the slope of aerodynamic phase curves. By adopting the sliding mesh technology,
Qu et al. [17] conducted a series of numerical simulations for a three–dimensional WIG
craft flying over the wavy ground. Their work argued that although the aerodynamic
coefficients are periodic under the wavy ground circumstance, the average values of the
coefficients vary in the same manner as the flat ground case when the flight height varies. It
was also found that lower flight height leads to higher aerodynamic coefficient amplitudes
and strengthens the flow separation at high angles of attack. Gao et al. [18] examined a
transonic airfoil over the wavy ground and observed different shock behaviors. Results of
different Mach numbers and ground clearances showed that the lower surface shock can
either remain attached to the ground or detach from the ground surface, which depends
on the flow conditions. In addition to numerical simulations, other authors also conducted
experiments. Lee and Tremblay–Dionne [19] used surface pressure and particle image
velocimetry measurements to obtain the aerodynamics and the flow field of a NACA 0015
airfoil over the wavy ground. It revealed that the peak values of aerodynamic coefficients
were larger than those in the flat ground case especially when the ground distance is
small, and the variation trend of different aerodynamic coefficients was not consistent. In
2019, they further studied the effect of the trailing–edge flap under the same experimental
conditions and reported that the fluctuation of aerodynamic coefficients is weakened due
to the deflected flap [20].
It should be noted that the heaving motion exists in water waves and the wavy behav-
ior is unsteady compared with the rigid wavy ground. There were only a few researchers
focusing on the numerical studies on the ground effect over the wavy water surface. Liang
et al. established the nonlinear lifting theory with the discrete vortex method for two–
dimensional airfoils [21] and three–dimensional wings [22] near the wavy water surface.
The method was further extended to the research on the heaving airfoils over progressive
water waves [23] in 2014. Kaneko [24] numerically surveyed the aerodynamics of a heaving
airfoil over an air–water interface and highlighted the impact of the water surface on the lift
force. Zhi et al. [25] compared the ground effect over the wavy wall with the wavy water
surface and reported that the wavy water disturbance velocity results in the differences in
aerodynamics, especially in a small ground clearance. They demonstrated that the flow
Appl. Sci. 2021, 11, 4215 3 of 19

properties change due to the interaction between the water wave and the airflow under the
airfoil but did not expound the mechanism of such interaction. Hu et al. [26] discovered
that the fluctuation of the wing aerodynamic coefficients in the wavy water surface case is
much more extensive than that in the wavy ground case. Analysis of the flow field clarified
that the air compression caused by the vertical movement of the water surface is the critical
factor for the significant lift increase. There are also some studies focusing on the design of
new vertical taking off and landing WIG crafts, such as the work conducted by Mi [27],
which studied the ducted fan performance over the ground, static water and dynamic
waves respectively. He demonstrated that the wave with a large wave height affected the
aerodynamic performance greatly and the wave surface was impacted by the jet flow.
Despite that a great deal of work has been carried out for the exploration of the
ground effect, the research on the airfoils over the wavy water surface is very limited. Some
researchers have attempted to clarify the interaction mechanism between the airfoil and
the wavy water surface, but the impact of the wave amplitude on the aerodynamics of
the airfoil has not been discussed adequately. Consequently, it is essential to investigate
the aerodynamics of the airfoil moving over the wavy water surface with various wave
amplitudes. The aim of the paper is to present guidance for the design of WIG crafts which
may face different sea conditions.
The remainder of this paper is organized as follows: Firstly, the numerical calculation
model is introduced, and the accuracy is verified. Then, the aerodynamic characteristics of
the NACA 4412 airfoil in proximity to the wavy water surface are numerically calculated
for different wave amplitudes. The lift coefficients of different angles of attack and wave
amplitudes are discussed according to the result. Moreover, the interaction mechanism
between the airfoil and the wavy water surface is clarified by analysis of the flow field.

2. Numerical Method and Validation

In this research, the NACA4412 airfoil of the chord length c = 0.5 m is investigated.
The freestream velocity V∞ is 30 m/s, and the Reynolds number Re based on the chord
length is 3 × 105 . The lift coefficient of the airfoil in proximity to the wavy water surface is
the main concern, and it is calculated at four angles of attack: α = 0◦ , 1◦ , 2◦ and 4◦ .
The calculation domain is presented in Figure 1. The inlet, upper, and lower bound-
aries are positioned 20c away from the airfoil, and the outlet boundary is located at 25c
from the trailing edge of the airfoil. A pressure outlet boundary condition is set for the
upper and outlet boundaries. The inlet boundary and lower boundary are set as velocity
inlet condition and nonslip wall condition respectively. In addition, the airfoil surface is set
as nonslip wall condition. Figure 1 also gives the definition of the ride height h and wave
parameters. The ride height is defined as the distance from the trailing edge of the airfoil to
the still water level and set as 0.2c. The wavelength λ is 5c which refers to existing studies
and for each angle, the wave amplitude a varies from 0.05c to 0.1c. The numerical wave
is generated by using the volume of fluid (VOF) technique [28]. Based on the first–order
linear wave theory [29], the wave propagating along the x–axis is generated by applying
the inflow wave making method based on the stream function [30]. The wave surface
elevation and the velocity components are specified at the velocity inlet, and they can be
calculated as
η = a cos(kx − ωt) (1)
cosh k(η + d)
Vx = aω cos(kx − ωt) (2)
sinhkd
sinhk(η + d)
Vz = aω sin(kx − ωt) (3)
sinhkd
where k = wave number, with k = 2π/λ; ω = wave frequency; t = time after the wave crest
passes through the original point; and d = water depth.
 sinh k   d 
VVz aa sinh k   d sin
sinkx
kxtt
(3)
(3)
z sinh kd
sinh kd
Appl. Sci. 2021, 11, 4215 4 of 19
where
where kk == wave
wave number,
number, with
with kk == 2π/λ;
2π/λ; ωω == wave
wave frequency;
frequency; tt == time
time after
after the
the wave
wave crest
crest
passes through the original point; and d = water
passes through the original point; and d = water depth.depth.

zz
20c
20c
xx
Flow
Flow

20c 25c
25c
20c

Wave cc
Wave
αα
VVz hh
z Vx
ηη Vx
20c
20c aa
λλ

Figure
Figure 1. Sketch
1.Sketch
Figure1. of
Sketchof the
ofthe computational
thecomputational domain.
computationaldomain.
domain.

The
The structural grid
Thestructural grid method
grid methodisisisemployed
method employedfor
employed for
for numerical
numerical
numerical simulations,
simulations,
simulations, as
as shown
as in
shown
shown in Fig-
in
Fig-
ure
Figure2. When
2. When simulating
simulating the high–amplitude
the high–amplitude wave, refinement
wave, refinementof the
of
ure 2. When simulating the high–amplitude wave, refinement of the grids around the wa- grids
the around
grids the
around wa-
the
water
ter surface
tersurface
surface is required
isisrequired
required to to achieve
toachieve
achieve better
better
better computational
computational
computational accuracy.
accuracy.
accuracy. The
TheThe
gridgrid
grid sizessizes
sizes around
around
around the
the
the water
water surface
surface in in
the the wavelength
wavelength and and
wave wave
heightheight directions
directions are are
1/10a
water surface in the wavelength and wave height directions are 1/10a and 1/25a, respec- 1/10a
and and
1/25a, 1/25a,
respec-
respectively.
tively. ThereThere
tively. There are
are 405 are
405 405 nodes
nodes
nodes on
on theon
the the airfoil,
airfoil,
airfoil, and and
and the
the y+
y+the y+ value
value
value for
for thefor
the themesh
first
first first mesh
mesh pointpoint
point away
away
away
from from
the the
airfoil airfoil
is is approximately
approximately 1. The 1. The
total total
number number
of meshof mesh
cells
from the airfoil is approximately 1. The total number of mesh cells is 0.39 million. cells
is 0.39is 0.39 million.
million.

Figure 2.
Figure2. Grids
2.Grids around
Gridsaround the
aroundthe airfoil.
theairfoil.
airfoil.
Figure
The numerical
Thenumerical simulations
numericalsimulationssimulations are are conducted
areconducted
conducted by by using
byusing
usingthe the commercial
thecommercial
commercialsoftware software
softwareSTAR–STAR–
STAR–
The
CCM
CCM++©.
CCM + © . The
© . The unsteady
Theunsteady
unsteady flow flow is numerically
flowisisnumerically
numericallysimulated simulated
simulatedby by solving
bysolving
solvingthe the unsteady
theunsteady
unsteadyincom-incom-
incom-
pressible Reynolds
pressibleReynolds
pressible Reynoldsaveraged averaged Navier–Stokes
averagedNavier–Stokes
Navier–Stokesequations equations
equations(RANS) (RANS)
(RANS)with with
withthe the k–ω
thek–ω shear
k–ωshear stress
shearstress
stress
transport
transport(SST)
transport (SST) turbulence
(SST)turbulence
turbulencemodel.model.
model.The The transport
Thetransport
transportequationsequations
equationsofoftheof the model
themodel
modelare are
are

kkkkvvkkt tkk** ff**kk00kk00GGkk GGnlnl GGbb SSkk (4) (4)
(ρk) +t∇
∂
t · (ρkv) = ∇ · ((µ + σk µt )∇k ) − ρβ ∗ f β∗ (ωk − ω0 k0 ) + Gk + Gnl + Gb + Sk (4)
∂t


 vvt tkk  ff2 2 2002GG DD SS
∂  2 2 (5)
(5)
(ρω ) +t∇ t · ( ρωv ) = ∇ · (( µ + σ µ
ω t )∇ k ) − ρβ f β ω − ω 0 + G ω + D ω + S ω (5)
∂t
where
where ffβ*β*represents
representsthe the free–shear
thefree–shear modification
free–shearmodification
modificationfactor, factor,
factor,ffβfββrepresents
represents
representsthe the vortex–stretching
thevortex–stretching
vortex–stretching
where fβ* represents
modification
modification factor,
factor, kk0 and ω0 are the ambient turbulence values that counteract turbulence
and ω are the ambient turbulence values that counteract turbulence
modification factor, k0 and ω 0 are the ambient turbulence values that counteract turbulence
0 0
decay,
decay, GGkkrepresents
represents the
the turbulent
turbulent production,
production, GGnlnlrepresents
represents the
the “non–linear”
“non–linear” production,
production,
decay, Gk represents the turbulent production, Gnl represents the “non–linear” production,
Gb represents the buoyancy production, GGωωrepresents
representsthe thespecific dissipation production,
GGb brepresents
representsthe thebuoyancy
buoyancyproduction,
production,G ω represents the specificdissipation
specific dissipationproduction,
production,
Dω represents the cross–diffusion term, Sk and Sω are source terms. The adaptable constants
have the following values: β* = 0.09, β1 = 0.075, β2 = 0.0828, σk1 = 0.85, σk2 = 1.0, σω1 = 0.5,
σω2 = 0.856, and a1 = 0.31.
Appl. Sci. 2021, 11, 4215 5 of 19

Appl. Sci. 2021, 11, x FOR PEER REVIEW 5 of 19

A second–order implicit scheme is used for time discretization and a second–order up-
wind scheme is applied for space discretization. The time step ∆t for unsteady computation
Dω represents the cross–diffusion term, Sk and Sω are source terms. The adaptable con-
is 0.0001 s to satisfy Equation (6),
stants have the following values: β* = 0.09, β1 = 0.075, β2 = 0.0828, σk1 = 0.85, σk2 = 1.0, σω1 =
V∞ · ∆t
0.5, σω2 = 0.856, and a1 = 0.31. <1 (6)
∆x
A second–order implicit scheme is used for time discretization and a second–order
where ∆x is the minimum grid size. The SIMPLE algorithm is used for the segregated
upwind scheme is applied for space discretization. The time step ∆t for unsteady compu-
flow solver, with the under–relaxation factor set as 0.8 and 0.2 for the velocity solver
tation is 0.0001 s to satisfy Equation (6),
and the pressure solver respectively. The momentum residual and the periodicity of
the aerodynamic coefficient are selectedVas  the
t convergence standard. If the residuals
1 −6 (6)
of the momentum, turbulent kinetic energy, xand continuity are less than 1 × 10 , and
the lift coefficient changes periodically simultaneously, the convergence of the numerical
where ∆x is the minimum grid size. The SIMPLE algorithm is used for the segregated flow
calculation is considered to be reached.
solver, with the under–relaxation factor set as 0.8 and 0.2 for the velocity solver and the
To verify the accuracy of the computational result, a validation for computational fluid
pressure solver respectively. The momentum residual and the periodicity of the aerody-
dynamics (CFD) method is conducted. The aerodynamic characteristics of the NACA4412
namic coefficient are selected as the convergence standard. If the residuals of the momen-
airfoil under the ground effect are calculated using the CFD method, and then compared
tum, turbulent kinetic energy, and continuity are less than 1 × 10−6, and the lift coefficient
with the experimental data obtained from the existing research [3,6,12,31]. The angle of
changes periodically simultaneously, the convergence of the numerical calculation is con-
attack is 4◦ , the ground clearance is h/c = 0.15, and the Reynolds number is 3.0 × 105 . These
sidered to are
conditions be reached.
consistent with those in the experiment performed by Ahmed et al. [32]. The
To verify
calculation the are
results accuracy of the
presented in computational
Figure 3, and itresult,can bea observed
validationthat forthe
computational
calculation
fluid dynamics (CFD) method is conducted. The aerodynamic
results are consistent with the experiment results. At the same time, the grid independence characteristics of the
is verified by constructing two new grids with additional refinement. Table 1 presentsthen
NACA4412 airfoil under the ground effect are calculated using the CFD method, and the
compared with
calculation results theunder
experimental
differentdata
gridobtained
numbers.from Thethe existing
relative research
errors of the[3,6,12,31]. The
lift coefficient
angle
for theseof attack
meshes is are
4°, the ground
1.25%, 1.25%,clearance
and 1.25% is h/c = 0.15, and and
respectively, the Reynolds
the relative number
errorsisof3.0
the×
10 . These
drag
5 conditions
coefficient for theseare meshes
consistent
arewith
6.61%, those in the
4.86%, andexperiment performed
5.14% respectively. by Ahmed
It can be noted et
al. [32]. The calculation results are presented in Figure 3, and
that the maximum difference between the medium grid and the fine grid is 0.26%, which it can be observed that the
calculation
indicates thatresults
the medium are consistent
grid canwith the experiment
basically ensure the results.
simulationAt the same time,
accuracy. the grid
Therefore the
independence is verified by constructing two new grids with additional
medium grid is selected to perform the numerical calculations. In addition, the fluctuation refinement. Table
1 presents of
amplitude thethecalculation results
lift coefficient andunder
dragdifferent
coefficientgrid numbers.
is less than 1The × 10 −6 when
relative errors of the
the result
liftconvergent.
is coefficient for Hence thesethemeshes are 1.25%,
unphysical 1.25%, caused
oscillation and 1.25% respectively,
by numerical and the relative
dispersion can be
errors ofduring
ignored the drag thecoefficient
simulation. for these meshes are 6.61%, 4.86%, and 5.14% respectively. It
can be noted that
According tothe
themaximum
wave theory,difference betweenlinear
the first–order the medium
wave isgrid basedandon thethefine grid is
inviscid
0.26%, which
hypothesis. indicatesthe
However, that theSST
k–ω medium
turbulencegrid can
modelbasically
is usedensure
for CFD thesimulation.
simulationTherefore,
accuracy.
itTherefore
is necessary the tomedium
ensure grid
that is
theselected
numerical to perform
wave is the numericalwith
in agreement calculations. In addition,
the theoretical wave.
the
A wave fluctuation
with theamplitude
wavelengthofλthe = 5clift
and coefficient and drag
the amplitude coefficient
a = 0.12c is lessinthan
is simulated 1 × 10of−6
the range
when
the the result is domain.
computational convergent. Hence in
As shown theFigure
unphysical oscillation
4, the wave caused
generated bybythenumerical
VOF method dis-
persion can
matches wellbewithignored during thewave.
the theoretical simulation.

Figure 3.
Figure 3. Comparison
Comparisonof
ofthe
thecomputed
computedand
andexperimental
experimentalpressure
pressurecoefficient
coefficientfor
forNACA
NACA 4412
4412 air-
airfoil.
foil.
 Table 1.
Table 1. Grid
Grid independence
independence study.
study.

Mesh
Mesh Number of
Number of Cells
Cells yy++ CLL
C CDD
C
Appl. Sci. 2021, 11, 4215 Coarse
Coarse 0.13 million
0.13 million 22 0.86890
0.86890 0.01162
0.011626 of 19
Medium
Medium 0.21 million
0.21 million 11 0.86896
0.86896 0.01143
0.01143
Fine
Fine 0.31 million
0.31 million 0.5
0.5 0.86892
0.86892 0.01146
0.01146
Experiment
Experiment
Table 1. Grid independence study. –– –– 0.8582
0.8582 0.0109
0.0109

Mesh y + CLis CDinviscid

According
According theNumber
to the
to of Cellsthe
wave theory,
wave theory, the first–order
first–order linear wave
linear wave is based
based onon the
the inviscid
hypothesis.
hypothesis. However, the
Coarse However, the k–ω
k–ω
0.13 SST turbulence
SST
million turbulence model
model
2 is used
is used0.86890
for CFD
for CFD simulation.
simulation.
0.01162 There-
There-
fore,Medium
fore, itit is
is necessary
necessary to 0.21 million
to ensure
ensure that the
that the numerical 1wave is
numerical wave is in 0.86896 with
in agreement
agreement with the0.01143
the theoretical
theoretical
wave. Fine
wave. A wave
A wave with
with the 0.31 million λ
the wavelength
wavelength λ == 5c
5c and 0.5 amplitude
and the
the amplitude0.86892
aa == 0.12c
0.12c is 0.01146in
is simulated
simulated in the
the
Experiment – – 0.8582
range of
range of the
the computational
computational domain.
domain. AsAs shown
shown inin Figure
Figure 4,4, the
the wave generated0.0109
wave generated by the
by the VOF
VOF
method matches
method matches well
well with
with the
the theoretical
theoretical wave.
wave.

Figure4.4.
Figure
Figure 4.Comparison
Comparisonof
Comparison ofthe
of thecomputed
the computedand
computed andtheoretical
and theoreticalwave
theoretical waveelevation
wave elevationatat
elevation atttt===1.2
1.2s.s.
1.2 s.

InInthis
In thisstudy,
this study,the
study, thewavy
the wavyground
wavy groundeffects
ground effectswere
effects wereevaluated
were evaluatedas
evaluated asaaacontrast.
as contrast.The
contrast. TheNACA
The NACA0015
NACA 0015
0015
airfoilisis
airfoil
airfoil ischosen
chosento
chosen toverify
to verifythe
verify thenumerical
the numericalmethod
numerical methodfor
method forpredicting
for predictingaerodynamics
predicting aerodynamicsof
aerodynamics ofan
of anairfoil
an airfoil
airfoil
overthe
over
over thewavy
the wavyground.
wavy ground.The
ground. Thesimulation
The simulationisis
simulation isconducted
conductedto
conducted tocompare
to comparewith
compare withthe
with thewind
the windtunnel
wind tunneldata
tunnel data
data
from
fromLee
from Leeand
Lee andTremblay–Dionne
and Tremblay–Dionne[19].
Tremblay–Dionne [19].The
[19]. Thesimulation
The simulationconditions
simulation conditionsare
conditions areconsistent
are consistentwith
consistent withthe
with the
the
experiment, including
experiment, including
experiment, including the the chord
the chord length
chord length of
length of the
of the airfoil
the airfoil set
airfoil set as
set as 25.4
as 25.4mm,
25.4 mm, the
mm, the Reynolds
the Reynolds number
Reynolds number
number
◦ , and the ground clearance set as h/c = 0.05.
set
setas
set as1.61
as 1.61 10555,, the
1.61××× 10 the angle
angle of of attack
attack setset as
as 12
12°,
12°, and thethe ground
ground clearance
clearance setset as
as h/c
h/c == 0.05.
0.05.
Besides,
Besides,the
Besides, the wave
the wave parameters
wave parameters are are aa ==
= 0.05c and λλ
0.05c and
and c.c.Figure
λ ===c. Figure555shows
Figure showsthe
shows thecomparison
the comparison
comparison of
of
of pressure
pressure
pressure coefficient
coefficient
coefficient distributions
distributions
distributions under
under
under thethe
the wavy
wavywavy ground
ground
ground condition
condition
condition with
with
with those
those
those under
under
under the
the
the
freefree
free stream
stream
stream condition.
condition.
condition. It reveals
It reveals
It reveals thatthat
that the numerical
the numerical
the numerical calculation
calculation
calculation produces
produces
produces the similar
the similar
the similar pres-
pres-
pressure distribution
sure distribution
sure distribution to to
to the the experimental
the experimental
experimental data. data.
data.

Figure 5. Comparison of experimental and computed pressure coefficient distributions for NACA
0015 airfoil. EXP = experimental.

3. Results and Discussion

3.1. Lift Behavior
The aerodynamics of the airfoil flying over the wavy water surface is periodic. For
clearer analysis, a relative time t* is defined to describe the relative position between the
 Figure 5. Comparison of experimental and computed pressure coefficient distributions for NACA
0015 airfoil. EXP = experimental.

3. Results and Discussion

Appl. Sci. 2021, 11, 4215 7 of 19
3.1. Lift Behavior
The aerodynamics of the airfoil flying over the wavy water surface is periodic. For
clearer analysis, a relative time t* is defined to describe the relative position between the
airfoil and
airfoil and wave.
wave. Supposing that T
Supposing that represents aa wave
T represents wave period
period which
which is is the
the time
time of
of the
the airfoil
airfoil
past one wavelength, the relative time is defined as t* = t/T.
moving past one wavelength, the relative time is defined as t* = t/T. At the initial moment
moving At the initial moment
of one
of one period,
period,thethetrailing
trailingedge
edgeof of
thethe
airfoil is directly
airfoil above
is directly the wave
above crest. crest.
the wave For each
Forangle
each
of attack, the lift coefficients of various amplitudes are shown in Figure 6. For
angle of attack, the lift coefficients of various amplitudes are shown in Figure 6. For com- comparison,
the lift coefficients under the case of free stream and flat ground effect at the corresponding
parison, the lift coefficients under the case of free stream and flat ground effect at the cor-
angle are calculated as well.
responding angle are calculated as well.

(a) (b)

(c) (d)
Figure 6.
Figure 6. Change
Change inin lift
lift coefficients
coefficients in
in two
two periods
periods at
atdifferent
differentwave
waveamplitudes:
amplitudes:(a)
(a)αα==00°; (b) αα== 11°;
◦ ; (b) (c) αα == 22°;
◦ ; (c) (d) α
◦ ; (d) α== 4°.
4◦ .
free = free stream; ground = flat ground.
free = free stream; ground = flat ground.

As can be
As can be seen
seen from
from Figure
Figure 6, 6, the
the shape
shape ofof the
the lift
lift coefficient
coefficient curves
curves atat low
low wave
wave
amplitudes isis similar
similar to sinusoidal waves. With the wave amplitude increasing,fluc-
to sinusoidal waves. With the wave amplitude increasing, the the
tuation amplitude
fluctuation amplitudeof theoflift
thecoefficient curves
lift coefficient increases
curves remarkably
increases and the
remarkably and shape of the
the shape
curve
of the changes. It is alsoItrevealed
curve changes. that the trend
is also revealed of aerodynamics
that the varying with
trend of aerodynamics the wave
varying with
amplitude is inconsistent
the wave amplitude at different
is inconsistent angles. When
at different angles.α When
= 0°, the 0◦ , the
α =most significant change
most significant
occurs
change at the valley
occurs at the of the lift
valley coefficient
of the curve,curve,
lift coefficient with the
with minimum valuevalue
the minimum decreasing con-
decreasing
siderably as the wave amplitude increases. When the
considerably as the wave amplitude increases. When the wave amplitude is wave amplitude is larger than
0.0875c, the minimum value of the lift coefficient is even less than zero, and this will will
threaten flight
flight safety
safety seriously.
seriously. While
Whilethethefluctuation
fluctuationamplitude
amplitude at at
thethe
peak
peak increases as
increases
as well, the maximum value of the lift coefficient still decreases. For the case of α = 1◦ ,
the amplitudes of both the wave crest and the trough increase, but the minimum value
decreases more. As for the case of α = 2◦ and 4◦ , the maximum value of the lift coefficient
increases notably, and the wave trough value decreases rarely especially for α = 4◦ . It
can be concluded that as the angle of attack increases, the enhancing effect caused by the
amplitude increase transfers from the valley of the lift curve to the crest, in the range of
low attack angles.
 well, the maximum value of the lift coefficient still decreases. For the case of α = 1°, the
amplitudes of both the wave crest and the trough increase, but the minimum value de-
creases more. As for the case of α = 2° and 4°, the maximum value of the lift coefficient
Appl. Sci. 2021, 11, 4215
increases notably, and the wave trough value decreases rarely especially for α = 4°. It can
8 of 19
be concluded that as the angle of attack increases, the enhancing effect caused by the am-
plitude increase transfers from the valley of the lift curve to the crest, in the range of low
attack angles.
The time–averaged
time–averagedlift liftcoefficients
coefficientsof of
different angles
different are then
angles calculated
are then and presented
calculated and pre-
in Figure
sented in 7, and the
Figure result
7, and thematches with thewith
result matches aforementioned analysis.analysis.
the aforementioned For α = For 0◦ and
α = 10°◦,

the time–averaged
and 1°, the time–averaged lift coefficients decreases
lift coefficients as the wave
decreases as theamplitude
wave amplitudeincreases and those
increases and
of α = 2 ◦ and 4◦ change inversely. The significant increase in the lift coefficient at α = 4◦
those of α = 2° and 4° change inversely. The significant increase in the lift coefficient at α
when the airfoil
= 4° when moves
the airfoil in proximity
moves to theto
in proximity wavy water water
the wavy surfacesurface
has beenhasdiscovered
been discoveredin our
previous researchresearch
in our previous and explained in detail in
and explained [26]. The[26].
detail flowThemechanism is that theisairthat
flow mechanism under
the the
air
airfoil is
under thecompressed by the particles
airfoil is compressed by theonparticles
the water onsurface duesurface
the water to theirdue
vertical movement,
to their vertical
and the hydrodynamic
movement, force is then transmitted
and the hydrodynamic force is thentotransmitted
the airfoil. According to this
to the airfoil. theory, the
According to
vertical disturbance of the wavy water enhances as the amplitude
this theory, the vertical disturbance of the wavy water enhances as the amplitude in- increases, and hence the
time–averaged
creases, and hence lift coefficient rises. However,
the time–averaged the theory
lift coefficient cannot
rises. However,directly
theexplain
theory the lift
cannot
loss at α = 0 ◦ and 1◦ . Compared the dashed line with the dotted line in Figure 6a, the lift
directly explain the lift loss at α = 0° and 1°. Compared the dashed line with the dotted
coefficient
line in Figurein the
6a,ground effect case is
the lift coefficient inlower than that
the ground in the
effect casefree streamthan
is lower casethat 0◦ . It
at αin= the is
free
apparent that ◦
stream case at there
α = 0°.exists a negativethat
It is apparent ground
there effect
existsat α = 0 , and
a negative the increase
ground effect atofαthe
= 0°,wave
and
amplitude seems to enhance the effect. Therefore, the influence of
the increase of the wave amplitude seems to enhance the effect. Therefore, the influence the wavy water at low
angles
of of attack
the wavy waterneeds
at lowmore clarification.
angles of attack needs more clarification.

Figure 7. Variation in the

Variation in the time–averaged
time–averagedlift
liftcoefficients
coefficientswith
withwave
waveamplitudes
amplitudesforfor different
different angles
angles of
of attack.
attack. a/ca/c
= 0=means
0 means flatflat ground.
ground.

3.2. Flow
3.2. Flow Field
Field Analysis
Analysis andand Discussion
Discussion
Based on
Based on the
the foregoing
foregoing analysis, the trend
analysis, the trend ofof the
the lift
lift coefficient
coefficient curve
curve varying
varying with
with
the wave amplitude is similar for α = 0 ◦ and 1◦ . Since the lift loss phenomenon is more
the wave amplitude is similar for α = 0° and 1°. Since the lift loss phenomenon is more
prominent ◦ the aerodynamics of this angle is taken as an example. We first choose a
prominent at at 00°,, the aerodynamics of this angle is taken as an example. We first choose a
low wave
low wave amplitude
amplitude case where aa is
case where is 0.05c
0.05c for
for analysis.
analysis. In this case,
In this case, thethe amplitude
amplitude is
is limited
limited
compared with the ride height. For clearer clarification of the flow mechanism, the airfoil
compared with the ride height. For clearer clarification of the flow mechanism, the airfoil
is divided into upper and lower surfaces through the leading and trailing edge points and
is divided into upper and lower surfaces through the leading and trailing edge points and
the lift
the lift coefficients
coefficients of ofthese
thesetwo
twoparts
partsare definedasasCC
aredefined L,up and CL,low
L,up and C
. Figure 8 shows the lift
L,low. Figure 8 shows the lift
coefficient curves
coefficient curves in intwo
twoperiods.
periods.ItItcan
canbebe seen
seen thatthat
thethe fluctuation
fluctuation amplitudes
amplitudes of Cof CL,up
L,up and
and CL,low are similar in this case, and therefore the influence of the wavy water vertical
CL,low are similar in this case, and therefore the influence of the wavy water vertical move-
movement on the lower surface is limited.
ment on the lower surface is limited.
 x FOR PEER REVIEW
Appl. Sci. 2021, 11, 4215 99of
of 19
Appl. Sci. 2021, 11, x FOR PEER REVIEW 9 of 19

Figure 8. Lift coefficients of the upper and lower surfaces at a = 0.05c and α = 0°.
Figure8.8.Lift
Figure Liftcoefficients
coefficientsof
ofthe
theupper
upperand
andlower
lowersurfaces
surfacesatataa==0.05c
0.05cand
andαα==00°.
◦.
For detailed analysis, one period of the airfoil moving over the wavy water surface
Fordetailed
For
is divided detailed
into fouranalysis,
analysis, oneperiod
one
time points, period ofofthe
with the the airfoil
airfoil
state = moving
at t*moving t* over
0 and over= 1the the wavy
wavy
being water
thewater
same. surface
surface
Figure is9
is divided
divided
shows into
the into four
four time
velocity time points,
points,
contours at withwith the
the state
different state
moments at t*
at t* =
and= 0 and
0 and t* =
t* =10
Figure 1 being
1 being
showsthe the same.
thesame. Figure
Figure
pressure 99
coef-
showsdistributions
shows
ficient thevelocity
the velocitycontours
contours atatdifferent
on the airfoil different momentsand
moments
at corresponding and Figure10
Figure
moments. 10shows
V* shows
is thepressure
the
defined pressure
as coef-
coeffi-
the ratio of
ficient
cient
the distributions
distributions
local velocity toon on
the the
the airfoil
airfoil
stream atatcorresponding
corresponding
velocity. Further, the moments.
pressureV*
moments. V*isisdefined
defined
coefficient as
Cpas theratio
isthe ratioas
defined of
of
thelocal
the localvelocity
Equation velocityto
(7), tothethestream
streamvelocity.
velocity.Further,
Further,the thepressure
pressurecoefficient
coefficientCCp pisisdefined
definedasas
Equation
Equation(7),(7),
pp−p∞p
p 
1 p2 
CCp = 2p2
(7)
C p 1/2ρV ∞V  (7)
 1 2 V 2
where p is the local pressure, pp∞∞ and V
local pressure, V∞∞are
arethe
thepressure
pressure andvelocity
and velocity of
of the
the incoming
incoming flow
where p is the local pressure,
respectively, and
far ahead respectively, and ρρ isp and
is the
∞ V
the air are the
air density.
density.
∞ pressure and velocity of the incoming flow
far ahead respectively, and ρ is the air density.
-0.3 -0.1 0.1 0.3 0.5 0.7

-0.3 -0.1 0.1 0.3 0.5 0.7

-0.3 -0.1 0.1 0.3 0.5 0.7

-0.3 -0.1 0.1 0.3 0.5 0.7

y/c

y/c
y/c

y/c

-0.5 -0.2 0.1 0.4 0.7 1.0 1.3 1.6 -0.5 -0.2 0.1 0.4 0.7 1.0 1.3 1.6
-0.5 -0.2 0.1 0.4x/c 0.7 1.0 1.3 1.6 -0.5 -0.2 0.1 0.4x/c 0.7 1.0 1.3 1.6
x/c x/c
(a) (b)
(a) (b)
-0.3 -0.1 0.1 0.3 0.5 0.7

-0.3 -0.1 0.1 0.3 0.5 0.7

-0.1 0.1 0.3 0.5 0.7

-0.1 0.1 0.3 0.5 0.7

y/c

y/c
y/c

y/c

A B
C
A B
C

-0.5 -0.2 0.1 0.4 0.7 1.0 1.3 1.6 -0.5 -0.2 0.1 0.4 0.7 1.0 1.3 1.6
-0.3

-0.3

-0.5 -0.2 0.1 0.4x/c 0.7 1.0 1.3 1.6 -0.5 -0.2 0.1 0.4x/c 0.7 1.0 1.3 1.6
x/c x/c
(c) (d)
(c) ◦ : (a)
(d)
Figure9.9.Velocity
Figure Velocitycontours
contoursof
ofdifferent
differentmoments
momentsatataa==0.05c
0.05cand
andαα==00°: (a) t*
t* == 0.0; (b) t* = 0.25; (c) t* == 0.5;
0.5; (d)
(d) t*
t* ==0.75.
0.75.
Figure 9. Velocity contours of different moments at a = 0.05c and α = 0°: (a) t* = 0.0; (b) t* = 0.25; (c) t* = 0.5; (d) t* = 0.75.
Appl. Sci.
Appl. Sci. 2021, 11,
11, 4215 PEER REVIEW 10 of 19
19
Appl. Sci. 2021,
2021, 11, xx FOR
FOR PEER REVIEW 10
10 of
of 19

Figure
Figure 10.
10. Pressure
Pressure coefficient distributions on the airfoil
airfoilatat different moments at aa == 0.05c and α=
Figure 10. Pressure coefficientdistributions
coefficient distributionson
onthe
theairfoil atdifferent
different moments
moments at at 0.05c
a = 0.05c α =α0=◦ .
and
and
0°.
0°.
As shown in Figure 9, the leading edge of the airfoil at α = 0◦ is close to the water
surface.As
As shown
shown
Therefore in
in Figure
Figure 9,
the cambered9, the
the leading
leadingedge
leading edgeof
edge the
ofwill
theleadairfoil
airfoil at
at αα == 0°
to shrinkage 0°isisatclose
close
the flowto
to the
the water
water
channel
surface.
surface.
inlet underTherefore
Therefore
the lower the cambered
the surface.
camberedFor leading
leading
instance,edge
edgethe will
will flowlead to
leadchannel shrinkage
to shrinkagein Figure at the
at the flow
9d flow channel
channel
is similar to
inlet
ainlet
2–D under
under
Venturi the
the lower
lowerwhich
tube, surface.
surface. isFor
firstinstance,
For instance,
convergent the
the flow
fromchannel
flow channel in
in Figure
the inlet—Section Figure 9d 9dAis
isto similar
similar to
to aa 2–
the throat— 2–
DD Venturi
Venturi
Section tube,
tube,then
B, and which
which is
is first
divergentfirst convergent
convergent from
from the
to the outlet—Section the inlet—Section
inlet—Section
C. In this case, A to
Atheto the
the throat—Section
throat—Section
incoming flow will
B,
B, and
firstand then
then divergent
accelerate to a maximum
divergent to
to the
the outlet—Section
velocity at the throat
outlet—Section C.
C. InIn this
section
this case,and
case, the
thethenincoming
decelerate
incoming flow will
will first
flowgradually. first
accelerate
However,
accelerate under to a maximum
the circumstance
to a maximum velocity
velocityofat at the
thethewavythroatwater
throat section
surface,
section and
andthe then decelerate
vertical
then movement
decelerate gradually.
of the
gradually.
However,
wave
However, under
influences
under the the circumstance
the flow channel. of
circumstance Tothe
of wavy
wavy water
illustrate
the this
water surface,
surface, the
phenomenon, the vertical
the casemovement
vertical of the airfoil
movement of
of
the
flyingwave
the wave over influences
the rigidthe
influences the flow
wavy channel.
flowground
channel.with To illustrate
the same
To illustrate this
this phenomenon,
wave parameters
phenomenon, the case
theiscase of
conducted the airfoil
as a
of the airfoil
flying
flying over
comparison.over the the rigid
rigid wavy
Figure 11 shows
wavy ground
ground the with
with the
velocity the same
same wave
contours wave ofparameters
the airfoil moving
parameters is
is conducted
conductedover the as
as aawavy
com-
com-
parison.
water
parison. Figure
surface
Figureand 11 shows the
wavy ground
11 shows velocity
the velocity contours
at t*contours of
= 0.75. Compared the airfoil
of the airfoilwith moving
moving Figure over
over the
11b, wavy
thewhether water
wavy water the
surface
high
surface and
and wavy
velocity region
wavy ground
at theat
ground t*
t* == 0.75.
throat
at section
0.75. Compared
or the low
Compared with
with Figure
velocity 11b,
11b, whether
Figureregions near the
whether the
the high
high veloc-
leading edge
veloc-
ity
ity region
and trailing
region at
at the
the throat
edge section
in Figure
throat section 11aor the
ordo low
thenotlow velocity
expand
velocity toregions near
the underlying
regions near the the leading
leading edge
surface. It is and
edge and trail-
apparent trail-
ing
ing edge in Figure 11a do not expand to the underlying surface. It is apparent that
that edge
the in
actualFigureflow 11a do
channel not inexpand
the wavy to the underlying
water case is surface.
not the It
exact is apparent
shape that the
formed by
the
actual
the flow
lower channel
surface of in the
the wavy
airfoil water
and watercase is not
surface.
actual flow channel in the wavy water case is not the exact shape formed by the lower the The exact shape
reason is formed
that the by the
particles lower on
surface
the water
surface of
ofthe airfoil
surface
the airfoil and
moveand water
water surface.
vertically, The
leading
surface. The reason
to the air
reason is
is that
nearthe
that theparticles
the water surface
particles on
on the
thewater
water surface
moving with
surface
move
them. vertically,
The outline leading
of the to the
actual air
flownear the
channel water
is surface
drawn
move vertically, leading to the air near the water surface moving with them. The outline inmoving
Figure with
11a them.
and is The outline
compressed
of
of the
theactual
compared actual flow
with channel
flowFigure
channel 11b.is
is drawn in
in Figure
Figure 11a
The throat—Section
drawn 11a and is
B moves
and is compressed
rearward,
compressed compared
and thewith
compared width
with Figureof it
Figure
11b.
11b. The throat—Section B moves rearward, and the width of it is narrower. Therefore the
is The
narrower. throat—Section
Therefore B
the moves
high rearward,
velocity and
region the
is width
larger, of
and it is
the narrower.
maximum Therefore
value of the
high
high velocity
velocity region
increases.
velocity region is is larger,
larger, and and thethe maximum
maximum value value of of the
the velocity
velocity increases.
increases.
0.7
0.7

0.5 0.7
0.5 0.7

0.3 0.5
0.3 0.5

Actual flow
Actual flow
0.1 0.3
0.1 0.3

channel
channel
y/c
y/c

A
y/c
y/c

A B C
B
-0.1 0.1
-0.1 0.1

Water surface
-0.3 -0.1

-0.3 -0.1

Water surface Wavy ground

Wavy ground
-0.3

-0.3

-0.5 -0.2 0.1 0.4 0.7 1.0 1.3 1.6 -0.5 -0.2 0.1 0.4 0.7 1.0 1.3 1.6
-0.5 -0.2 0.1 0.4 0.7 1.0 1.3 1.6 -0.5 -0.2 0.1 0.4 0.7 1.0 1.3 1.6
x/c x/c
x/c x/c
(a)
(a) (b)
(b)
Figure
Figure 11.
Figure 11. Velocity
11. Velocity contours
Velocity contours around
contours around the
aroundthe airfoil
theairfoil at t*
att*
airfoilat == 0.75:
t*= 0.75: (a)
0.75: (a) wavy
wavy water
wavy water surface;
water surface; (b)
surface; (b) wavy
(b) wavy ground.
wavy ground.
ground.

During
During the
During theperiod
the periodfrom
period from
from t*
t* =t*
= 00=to
to0t*
t*to== 0.25,
t* = the
0.25, the wave
0.25, the trough
wave wave moves
moves towards
troughtrough moves the
the leading
towardstowards the
leading
edge,
edge, and the distance between the leading edge of the airfoil and the water surface in-
leading and the
edge, distance
and the between
distance the leading
between theedge of
leading the airfoil
edge of and
the the water
airfoil and surface
the water
in-
creases.
surface With
With the
the cross
creases.increases. With section
crossthe crossgetting
section sectionwiden,
getting gettingthe
widen, shrinkage
widen,
the at
at the
the shrinkage
shrinkage the inlet decreases
at the
inlet corre-
inlet decreases
decreases corre-
spondingly
spondingly and
and weakens
correspondingly and weakens
weakens the
the acceleration
the acceleration
acceleration effect of
of the
effect
effect airflow
of the
the in
in the
airflow
airflow flow
in the
the channel.
flow
flow channel.
channel. Due
DueDueto
to
the
to deceleration
the of
deceleration the
of airflow,
the airflow, thethe pressure
pressure ononthe
the lower
lower surface
surface of
of
the deceleration of the airflow, the pressure on the lower surface of the airfoil increases, the
the airfoil
airfoil increases,
increases,
 Appl. Sci. 2021, 11, x FOR PEER REVIEW 11 of 19
Appl. Sci. 2021, 11, 4215 11 of 19

and the corresponding pressure coefficient gets higher in the range from x/c = 0.1 to x/c =
0.5 the
and in Figure 10. The liftpressure
corresponding coefficient of the lower
coefficient getssurface
higher consequently
in the range increases.
from x/c = From
0.1 tot* =
0.25 to t* = 0.5, the distance between the leading edge of the
x/c = 0.5 in Figure 10. The lift coefficient of the lower surface consequently increases. From airfoil and the water surface
t*decreases
= 0.25 to t*and = 0.5,enhances
the distance the between
blockingthe effect,
leadingwhileedgethe shrinkage
of the airfoil andat the inlet does
the water surface not
change significantly.
decreases and enhancesThus, the maximum
the blocking velocity
effect, while of the airflow
the shrinkage at theunder
inlet the
doeslower surface
not change
declines at the
significantly. Thus, leading edge, and
the maximum a significant
velocity pressure
of the airflow undercoefficient
the lower increase
surfacecan be ob-
declines
served from the range of x/c = 0 to x/c = 0.1 in Figure 10. Therefore,
at the leading edge, and a significant pressure coefficient increase can be observed from the the lift coefficient still
increases. From t* = 0.5 to t* = 0.75, the wave crest moves towards
range of x/c = 0 to x/c = 0.1 in Figure 10. Therefore, the lift coefficient still increases. From the airfoil, and the dis-
t*tance
= 0.5 between
to t* = 0.75,thethe leading
waveedge
crest of the airfoil
moves towards andthetheairfoil,
water and surface decreases.
the distance Monitoring
between the
the movement
leading edge of the of the water
airfoil and surface,
the water we surface
can see decreases.
that there exists an upward
Monitoring velocity com-
the movement of
ponent
the waterfor particles
surface, we canon the
seewater
that theresurface,
existsasan shown
upwardin Figure
velocity 12.component
The movement of parti-
for particles
oncles
theonwater
the water
surface, surface forces in
as shown theFigure
air near 12.theThewater surface to
movement ofmove vertically
particles on theand watercon-
sequently
surface forcesenhances
the air nearthe shrinkage
the water of the flow
surface channel
to move further.and
vertically Owing to this phenomenon,
consequently enhances
theshrinkage
the acceleration of theof flow
the airstream enhances,
channel further. and to
Owing thethis
pressure of the lower
phenomenon, surface as well
the acceleration of
as airstream
the the lift coefficient
enhances,decreases. Besides,offrom
and the pressure t* = 0.25
the lower to t* as
surface = 0.75,
well the
as theupward actuation
lift coefficient
decreases.
by the water Besides,
surface fromandt*the= 0.25 to t* =effect
blocking 0.75,at thetheupward
inlet ofactuation by the water
the flow channel causesurface
more air
and the blocking
to flow above theeffect upper at surface.
the inletThe of the flow channel
velocity cause on
of the airflow more theair
upperto flow above
surface the
thereby
upper surface.
increases, with The anvelocity
extension of the airflow
of high on theregions
velocity upper surface
in Figure thereby
9c,d. increases,
Therefore,with an
a lower
extension
pressure of high velocity
distribution regions
occurs on thein Figure
upper9c,d.surfaceTherefore,
with more a lower pressure
suction. distribution
Finally, the wave
occurs
movesondownward
the upper surface with more
at the leading edge suction.
of the Finally,
airfoil, andthe wave moves
the width ofdownward
the throat at the
section
leading edge of the airfoil, and the width of the throat section
begins to increase in the period of t* = 0.75~1.0. Consequently, the shrinkage of the flow begins to increase in the
period
channel of t*decreases,
= 0.75~1.0.and Consequently,
the acceleration the shrinkage of the flow
effect weakens. Thechannel
stream decreases,
velocity underand the the
acceleration
lower surface effect weakens.
decreases, and Thethestream
pressure velocity
increases,under the lower
resulting surface
in an increasedecreases, and
in the lift co-
the pressure
efficient. increases,
The combined resulting
effect of in antheincrease
blocking ineffect
the liftweakening
coefficient.and Thethecombined
upwardeffectforceofby
the
theblocking effect weakening
wave movement decreasing and the upward
leads force of
to the decline bythetheairflow
wave movement
over the upper decreasing
surface.
leads to the decline of the airflow over the upper surface.
The velocity of the airflow then decreases, and the suction on the upper surface also The velocity of the airflow thende-
decreases,
creases. and the suction on the upper surface also decreases.

Figure12.
Figure 12.Vertical
Vertical velocity
velocity ofof the
the particles
particles onon the
the water
water surface
surface below
below the
the airfoil.
airfoil.

The
Thecasecaseofofthethesame
sameangleangleofofattack
attackand
anda ahigher amplitudea =
higheramplitude a =0.1c,
0.1c,isisdiscussed
discussedinin
the following. Likewise, the change
the following. Likewise, the change of CL,up of C and C
L,up and CL,low is presented in Figure
L,lowis presented in Figure 13. 13. Under
Under the
the situation
situation of the
of the highhigh amplitude,
amplitude, the fluctuation
the fluctuation of C
of CL,up L,upCand
and CL,lowenhances,
L,low both both enhances,
but it is
but
clearit is clear
that the that the dramatic
dramatic decreasedecrease of CL,low dominates
of CL,low dominates the lift
the lift change. Thechange. The liftof
lift coefficient
coefficient
the lowerof the lower
surface surface
reaches reaches theatmaximum
the maximum around t* at around
= 0.5 t* = 0.5
and then and then
reaches reaches
the minimum
the
at minimum at t* = 0.8.
t* = 0.8. Velocity Velocity
contours atcontours at fourmoments
four different different moments and the corresponding
and the corresponding pressure
pressure coefficient distributions are shown in Figures 14 and 15
coefficient distributions are shown in Figures 14 and 15 respectively. Comparing Figure respectively. Comparing
Figure
14a,d,14a,d,
we can weseecanthat
seethe
thathigh
the high velocity
velocity region region
on the onlower
the lower
surfacesurface
of theofairfoil
the airfoil
at t*at=0
t*is= larger
0 is larger than that at t* = 0.8. The maximum value of the velocity at
than that at t* = 0.8. The maximum value of the velocity at t* = 0 is also higher. t* = 0 is also higher.
However,
However,Figure Figure1515indicates
indicatesthat thatthe
thepressure
pressurecoefficient
coefficientofofthethecorresponding
correspondingregion regionatat
t*t*==0.8
0.8isis lower
lower than that at
than that at t*t*==0,0,and
andananunusual
unusual lowlow pressure
pressure peakpeak occurs
occurs at t*at=t*0.8
= from
0.8
from the range of x/c = 0.1 to x/c = 0.4 on the lower surface. The result is in contradiction
 Appl. Sci. 2021, 11, x FOR PEER REVIEW 12 of 19
Appl. Sci.
Appl. Sci. 2021, 11, x4215
2021, 11, FOR PEER REVIEW 1212ofof 19
19

the range of x/c = 0.1 to x/c = 0.4 on the lower surface. The result is in contradiction with
the
therange
with ofofx/c
the theory
theory the=ofVenturi
0.1
thetoVenturi
x/ctube,
= 0.4 on the
tube,
whichwhichlower surface.
suggests
suggests that athatThe result
a higher
higher is incorresponds
contradiction
velocity
velocity corresponds with
to lower to
lower
the pressure.
theory of the Moreover,
Venturi the
tube, maximum
which pressure
suggests that coefficient
a higher on the
velocity
pressure. Moreover, the maximum pressure coefficient on the airfoil surface usually oc-airfoil surface
corresponds tousually
lower
occurs
curs atatthe
pressure. theleading
leadingedge
Moreover, edge
the stagnation
maximum
stagnation point,and
pressure
point, andthe
thevalue
coefficientvalue is equal
on the
is equal tosurface
airfoilto 11 according
usually
according to the
to oc-
the
definition,
curs at the in the case
leading of single–phase
edge stagnation incompressible
point, and the airflow.
value is While
equal to for
1 the two–phase
according to the
definition, in the case of single–phase incompressible airflow. While for the two–phase
wave with in
definition, a highcaseamplitude, there exists a pressure peak where the maximum pressure
wave with a the of single–phase
high amplitude, incompressible
there exists a pressure peakairflow.
where While for the two–phase
the maximum pressure
coefficient
wave with exceeds
aexceeds 1 at t* = 0.8 there
high amplitude, on the lowerasurface, aspeak
shown in Figure 15. These abnormal
coefficient 1 at t* = 0.8 on theexists pressure
lower surface, as shown where the maximum
in Figure pressure
15. These abnormal
phenomena
coefficient reveal 1that
exceeds at t*under
= 0.8 onhigh
thewave
loweramplitude
surface, assituations,
shown in the airflow
Figure velocity
15. These change
abnormal
phenomena reveal that under high wave amplitude situations, the airflow velocity change
caused by the
phenomena flowthat
reveal channel
under variation
high is not
wave the only situations,
amplitude factor that theshould be considered
airflow velocity when
change
caused by the flow channel variation is not the only factor that should be considered when
analyzing
caused the
by the pressure distribution.
analyzing theflow channel
pressure variation is not the only factor that should be considered when
distribution.
analyzing the pressure distribution.

Figure 13.
Figure 13. Lift
Lift coefficients
coefficients of
of the
the upper
upper and
and lower
lowersurfaces
surfacesat
ataa== 0.1c
0.1c and
and αα == 00°.
◦.
Figure 13. Lift coefficients of the upper and lower surfaces at a = 0.1c and α = 0°.
-0.3 -0.1 0.1 0.3 0.5 0.7
-0.3 -0.1 0.1 0.3 0.5 0.7

-0.3 -0.1 0.1 0.3 0.5 0.7

-0.3 -0.1 0.1 0.3 0.5 0.7
y/c

y/c
y/c

y/c

-0.5 -0.2 0.1 0.4 0.7 1.0 1.3 1.6 -0.5 -0.2 0.1 0.4 0.7 1.0 1.3 1.6
-0.5 -0.2 0.1 0.4 x/c 0.7 1.0 1.3 1.6 -0.5 -0.2 0.1 0.4 x/c 0.7 1.0 1.3 1.6
x/c x/c
(a) (b)
(a) (b)
-0.3 -0.1 0.1 0.3 0.5 0.7

-0.3 -0.1 0.1 0.3 0.5 0.7

-0.3 -0.1 0.1 0.3 0.5 0.7

-0.3 -0.1 0.1 0.3 0.5 0.7

y/c

y/c
y/c

y/c

-0.5 -0.2 0.1 0.4 0.7 1.0 1.3 1.6 -0.5 -0.2 0.1 0.4 0.7 1.0 1.3 1.6
-0.5 -0.2 0.1 0.4 x/c 0.7 1.0 1.3 1.6 -0.5 -0.2 0.1 0.4 x/c 0.7 1.0 1.3 1.6
x/c x/c
(c) (d)
(c) (d)
Figure 14. Velocity contours of different moments at a = 0.1c and α = 0°: (a) t* = 0.0; (b) t* = 0.25; (c) t* = 0.5; (d) t* = 0.8.
Figure 14.
Figure 14. Velocity
Velocitycontours
contoursof
ofdifferent
differentmoments
momentsat
ataa== 0.1c
0.1c and
and αα== 00°: (a) t*
◦ : (a) t* =
= 0.0;
0.0; (b)
(b) t*
t* == 0.25;
0.25; (c)
(c) t*
t* ==0.5;
0.5;(d)
(d)t*t*==0.8.
0.8.
Appl. Sci. 2021, 11, 4215 13 of 19
Appl. Sci. 2021, 11, x FOR PEER REVIEW 13 of 19

Figure15.
Figure 15.Pressure
Pressurecoefficient
coefficientdistributions
distributionsononthe
theairfoil
airfoilatat different
different moments
moments at at
a =a 0.1c
= 0.1c and
and α =α 0=◦ .
0°.
When the wave amplitude is higher, the distance between the wave crest and airfoil
When
is closer. Thethe wave analysis
former amplitude is higher,
implies that the
the distance between theofwave
upward movement crest andon
the particles airfoil
the
is closer.
water surfaceThemight
formerdirectly
analysis implies
impact on that the upward
the pressure movement
distribution of the
on the particles
airfoil besidesonthethe
waterof
shape surface
the flow might directly
channel. This impact on the pressure
is instructive distribution
for the analysis of theonliftthe airfoilatbesides
change differentthe
shape of the
moments, flow channel.
especially the abruptThis isdecline
instructive
near for
t* =the0.8.analysis
Therefore,of thewe liftexhibit
changethe at different
vertical
moments,
velocity of especially
the particles theon abrupt
the water decline near below
surface t* = 0.8. theTherefore,
airfoil at we exhibitmoments
different the vertical in
velocity of the particles
Figure 16. At t* = 0, although the on the water surface under the airfoil starts to move downwardin
surface below the airfoil at different moments
Figure 16. to
according AtFigure
t* = 0, although the waterwhen
16, the influence surfacetheunder
wavethe airfoil
crest starts
passes thetolower
move downward
surface of
according
the airfoil on to Figure 16, thestill
the airflow influence
exists. when the wave
It is shown thatcrest
thepasses the lower
flow channel is surface
compressed of the
airfoil on theinairflow
considerably Figurestill
14a.exists.
Therefore,It is shown
a large that
highthe flow channel
velocity is compressed
region appears and the consider-
pressure
ably in Figure
coefficients 14a.region
in this Therefore, a large
are fairly low.high velocity
At the sameregion
time, the appears
peak ofand thethe pressure
wave locates coef-
at
ficients
the in this
trailing edge region
of theare fairlyand
airfoil, low.theAtparticles
the sameon time,
the the
waterpeak of thestill
surface wave movelocates at the
upward
in this area
trailing edge (see Figure
of the 16) and
airfoil, andthe compress
particlesthe onair. Thus, surface
the water there isstilla highmove pressure
upwardregion in this
near
area the
(seetrailing
Figureedge16) andin Figure
compress 15. During
the air. the
Thus, period
therefrom t* = 0pressure
is a high to t* = 0.5, the distance
region near the
between
trailing edgethe airfoil and the
in Figure 15. water
During surface increases,
the period fromand t* = the
0 toshrinkage
t* = 0.5, the of the flow channel
distance between
at entry
the decreases.
airfoil The wavy
and the water surface water surfaceand
increases, underthe the airfoil of
shrinkage mainly
the flow moves downward,
channel at entry
and the interaction
decreases. The wavy between
water itsurface
and the air weakens.
under the airfoil Therefore,
mainly moves the change of the velocity
downward, and the
and pressure
interaction is similar
between to the
it and thecase of lower Therefore,
air weakens. amplitude.the Inchange
the period
of the t* = 0.5~0.8,
ofvelocity the
and pres-
crest of the wave moves towards the leading edge, and the distance
sure is similar to the case of lower amplitude. In the period of t* = 0.5~0.8, the crest of the between the wave
and
wave themoves
airfoiltowards
decreases. the Figure
leading16 shows
edge, andthat
the the particles
distance on the
between thewater
wavesurface
and thehave a
airfoil
fairly high Figure
decreases. vertical16velocity
shows that towards the airfoil.
the particles on theAs water
the position
surfaceof the astagnation
have fairly highpointverti-
iscal
close to the
velocity wave crest,
towards the compression
the airfoil. As the position of the particles
of the on thepoint
stagnation water surface
is close leads
to the waveto
acrest,
high the
pressure peak at the leading edge of the lower surface.
compression of the particles on the water surface leads to a high pressure peak Therefore, in Figure 15,
we canleading
at the see the edgepressure
of the coefficient
lower surface. of theTherefore,
leading edge at t* =15,
in Figure 0.8weis can
largerseethan 1. This
the pressure
sharp pressure increase enhances the shrinkage of the flow
coefficient of the leading edge at t* = 0.8 is larger than 1. This sharp pressure increase channel entry, and thus, the
air flowing under the airfoil is blocked. Simultaneously, the
enhances the shrinkage of the flow channel entry, and thus, the air flowing under the air- air which have passed the
leading edge accelerates
foil is blocked. towards the
Simultaneously, thetrailing
air which edgehavebecausepassedof thethecompression
leading edge byaccelerates
the wavy
water
towards surface. A peakedge
the trailing of negative
becausepressure then comes into
of the compression by theexistence on thesurface.
wavy water lower surface,
A peak
and C L,low falls into the minimum value rapidly. Figure
of negative pressure then comes into existence on the lower surface, and CL,low falls 17 shows the pressure contourinto
around
the minimumthe airfoilvalueat t* = 0.8, where
rapidly. Figurethe high pressure
17 shows the pressure region at the leading
contour around the edge and the
airfoil at t*
following low pressure region can be clearly observed. Additionally, the obstruction at
= 0.8, where the high pressure region at the leading edge and the following low pressure
the flow channel under the airfoil forces more air to flow above the airfoil. The velocity
region can be clearly observed. Additionally, the obstruction at the flow channel under
of the airflow above the upper surface increases consequently, causing the low pressure
the airfoil forces more air to flow above the airfoil. The velocity of the airflow above the
peak at the leading edge of the upper surface and the increase of CL,up . In the period of
upper surface increases consequently, causing the low pressure peak at the leading edge
t* = 0.8~1, the wave crest moves past the leading edge stagnation point. The blocking at the
of the upper surface and the increase of CL,up. In the period of t* = 0.8~1, the wave crest
entry of the flow channel is relieved, and the airflow accelerates to flow into. The pressure
moves past the leading edge stagnation point. The blocking at the entry of the flow chan-
coefficient on the lower surface returns to the normal value which is caused by the high
nel is relieved, and the airflow accelerates to flow into. The pressure coefficient on the
 Appl. Sci.
Appl. Sci. 2021,
2021, 11,
11, xx FOR
FOR PEER
PEER REVIEW
REVIEW 14 of
14 of 19
19
Appl. Sci. 2021, 11, 4215 14 of 19

lower surface
lower surface returns
returns toto the
the normal
normal value
value which
which is
is caused
caused by
by the
the high
high velocity.
velocity. In
In addi-
addi-
velocity.
tion, theIn addition,
suck peak the
on suck
the peak
upper on the
surface upper surface
disappears as disappears
the airflow as the
over airflow
the upper
tion, the suck peak on the upper surface disappears as the airflow over the upper surfaceover the
surface
upper surface
decreases, and
decreases, decreases,
and CCL,up and C decreases
decreases sharply.
L,up decreases sharply.
L,up sharply.

Figure 16.
Figure 16. Vertical
Vertical velocity
velocity of the
the particles
particles on
on the water
water surface
surface below
below the airfoil
airfoil at
at different
different mo-
mo-
Figure 16. Vertical velocity ofof
the particles on thethe
water surface below the the
airfoil at different moments.
ments.
ments.
0.5
0.5
0.3
0.3
0.1
0.1
y/c
y/c
-0.1
-0.1
-0.3
-0.3

-0.3
-0.3 -0.1
-0.1 0.1
0.1 0.3
0.3 0.5
0.5 0.7
0.7 0.9
0.9 1.1
1.1 1.3
1.3
x/c
x/c
Figure17.
Figure
Figure 17.Pressure
17. Pressurecontour
Pressure contouraround
contour aroundthe
around theairfoil
the airfoilatat
airfoil att*t*
t*===0.8.
0.8.
0.8.

To
Toclarify
To clarifythe
clarify thedifference
the differencebetween
difference betweenthe
between thelow
the lowand
low andhigh
and highamplitudes
high amplitudesfor
amplitudes forthe
for thewavy
the wavywater
wavy water
water
surface,
surface, the
the velocity
velocity of the
of particles
the particleson the
on water
the surface
water surfacein a larger
surface, the velocity of the particles on the water surface in a larger region at t* = 0.8in a region
larger at
regiont* = 0.8
at is
t* given
= 0.8 is
is
in Figure
given in
given 18.
in Figure For
Figure 18. different
18. For wave
For different amplitudes,
different wave
wave amplitudes, the magnitude
amplitudes, the the magnitude of the
magnitude of velocity
of the component
the velocity
velocity compo-
compo-in
the
nent
nenthorizontal
in the
in directiondirection
the horizontal
horizontal is similar,
direction is while that
is similar,
similar, while
while in that
the
thatvertical
in the
in direction
the vertical
vertical is in approximate
direction
direction is in
is in approx-
approx-
proportion
imate to the wave
imate proportion
proportion to the
to amplitude.
the wave Consequently,
wave amplitude.
amplitude. the influence
Consequently,
Consequently, of the highof
the influence
the influence ofamplitude
the high
the wave
high ampli-
ampli-
on thewave
tude
tude air under
wave on thethe
on the airairfoil
air underenhances
under airfoilmarkedly
the airfoil
the enhances when
enhances markedly
markedly the wavewhenmoves
when the wave
the wave upward.
movesBesides,
moves upward.
upward.
the airfoil
Besides, locates
the airfoilbetween
locates the dashed
between lines
the in
dashed Figure
lines 18.
in For
Figure
Besides, the airfoil locates between the dashed lines in Figure 18. For the case where the the case
18. where
For the the
case amplitude
where the
a = 0.05c, the
isamplitude wave is
is aa == 0.05c, scarcely
0.05c, the
the wave
wave isaffected
is scarcelyby the
scarcely affected pressure
affected by by the distribution
the pressure beneath
pressure distribution the
distribution beneath airfoil.
beneath
amplitude is
Whereas
the airfoil.forWhereas
airfoil. the high for amplitude highwave,
the high it shows
amplitude wave, thatit the
shows velocity distributions
that the
the below the
velocity distributions
distributions
the Whereas for the amplitude wave, it shows that velocity
airfoil
beloware thedifferent
airfoil are from
are those at
different froma wavelength
those at ahead of theahead
at aa wavelength
wavelength airfoil.ofAtthethe leading
airfoil. edge
At the
the of
lead-
below the airfoil different from those ahead of the airfoil. At lead-
the airfoil,
ing edge
edge of the
of the high pressure
the airfoil,
airfoil, the region
the high on
high pressure the lower
pressure region
region on surface
on the
the loweris so powerful
lower surface
surface is that
is so it
so powerfulputs
powerful that off the
that it
it
ing
upward
puts offmovement
the upward of the wave andofaccelerates
movement the wave the
and wave in the horizontal
accelerates the wave direction.
in the Hence,
horizontal
puts off the upward movement of the wave and accelerates the wave in the horizontal
there existsHence,
direction. strong there
interaction
exists between the airfoilbetween
strong interaction
interaction and the wavy water
the airfoil
airfoil andsurface
the wavy whenwater
wavy the
direction. Hence, there exists strong between the and the water
wave amplitude
surface whenwhen the is higher,
the wave and
wave amplitude it is
amplitude isthe main reason
is higher,
higher, and for
and it the
it is unusual
is the
the main change
main reason
reason forof lift
for the coefficient.
the unusual
unusual
surface
change of lift coefficient.
change of lift coefficient.
Appl. Sci.
Appl. Sci. 2021, 11, x4215
2021, 11, FOR PEER REVIEW 1515ofof 19
19
Appl. Sci. 2021, 11, x FOR PEER REVIEW 15 of 19

(a)
(a) (b)
(b)
Figure 18. Velocity of particleson onthe
thewater
watersurface:
surface:(a)
(a)horizontal
horizontalvelocity;
velocity;
(b)(b) vertical velocity.
VxV=V
x = velocity component
Figure 18.Velocity
Figure18. Velocityofofparticles
particles on the water surface: (a) horizontal velocity; vertical
(b) velocity.
vertical velocity. velocity component
x = velocity in
component
in
thethe horizontal
horizontal direction;
direction; andand
V V=z = velocity component in the vertical direction.
velocity component in the vertical direction.
in the horizontal direction; and z Vz = velocity component in the vertical direction.

ItItItcan
canbe
can beconcluded
be concludedthat
concluded thatthe
that theflow
the flowmechanism
flow mechanismof
mechanism ofthe
of theairfoil
the airfoilmoving
airfoil movingover
moving overthe
over thewavy
the wavy
wavy
water
water surface with a higher amplitude is different from that with a lower amplitude.For
water surface
surface with
with aa higher
higher amplitude
amplitude is
is different
different from
from that
that with
with a lower amplitude.
amplitude. For
For
the
the high
high amplitude
amplitude wave
wave where
where the
the wave
wave peak
peak is
is close
close enough
enough
the high amplitude wave where the wave peak is close enough to the airfoil, the vertical to
to the
the airfoil,
airfoil, the
the vertical
vertical
movement
movementof
movement ofthe
of thewater
the watersurface
water surfacedirectly
surface directlyaffects
directly affectsthe
affects thepressure
the pressuredistributions
pressure distributionson
distributions onthe
on theairfoil
the airfoil
airfoil
apart
apart from
from the
the compression
compression effect
effect on
on the
the flow
flow channel.
channel. When
When
apart from the compression effect on the flow channel. When the wave crest moves below the
the wave
wave crest
crest moves
moves below
below
the
theleading
the leadingedge
leading edgestagnation,
edge stagnation,the
stagnation, thecompression
the compressionby
compression bybythethewave
the wavecreates
wave createsaaahigh
creates highpressure
high pressureregion
pressure region
region
and
and prevents air from flowing under the lower surface. Thus
and prevents air from flowing under the lower surface. Thus a large low pressureregion
prevents air from flowing under the lower surface. Thus aa large
large low
low pressure
pressure region
region
with
withlow
with lowspeed
low speedoccurs
speed occursatat
occurs atthe
thelower
the lower
lower surface
surface
surface andand
and leadsleads
leads totolift
to loss.
liftlift The
loss.
loss. velocity
TheThe contour
velocity
velocity and
contour
contour and
pressure
and pressure coefficient distribution
coefficient distributionof α =
of 1° 1◦ with
α =with the same
the samewave wave amplitude
amplitude are shown
are showninin
pressure coefficient distribution of α = 1° with the same wave amplitude are shown
in Figures
Figures 19 19 and
and 20. 20. Similar
Similar phenomena
phenomena can becan be found
found at α = 1°, = 1◦ , that
at αexcept exceptthe that
peak the peak
magni-
Figures 19 and 20. Similar phenomena can be found at α = 1°, except that the peak magni-
magnitude
tude of the pressure
high pressure at leadingthe leading edge and the following lowpressure pressuredeclines.
declines.
tudeofofthe thehigh
high pressureatatthe the leadingedge edge
◦
and
andthe thefollowing
followinglow low pressure
◦ declines.
This is
This is because that that the leading
leading edge at at α = 1 is is higher than than that at at α = 0 , andand the larger
larger
This isbecause
because thatthe the leadingedge edge atαα==1° 1° ishigher
higher thanthat that atαα==0°, 0°, andthe the larger
distance weakens
distance weakens the the blocking
blocking effect.
effect.
distance weakens the blocking effect.
-0.3 -0.1 0.1 0.3 0.5 0.7
-0.3 -0.1 0.1 0.3 0.5 0.7
y/c
y/c

-0.5 -0.2 0.1 0.4 0.7 1.0 1.3 1.6

-0.5 -0.2 0.1 0.4 0.7 1.0 1.3 1.6
x/c
x/c
Figure 19. Velocity contour of α = 1°◦ at a = 0.1c and t* = 0.8.
Figure19.
Figure 19.Velocity
Velocitycontour
contourofofαα==11°at
ataa== 0.1c
0.1c and
and t*
t* == 0.8.
Appl. Sci. 2021, 11, 4215 16 of 19
Appl.Sci.
Appl. Sci.2021,
2021,11,
11,xx FOR
FOR PEER
PEER REVIEW
REVIEW 16 of
16 of 19
19

Figure20.
Figure
Figure 20.Pressure
20. Pressurecoefficient
Pressure coefficient
coefficient distribution
distribution
distribution onon
on the
the
the airfoil
airfoil
airfoil of
ofof α =1=◦1°
α α= 1° ata a=
atat a ==0.1c
0.1cand
0.1c andt*t*
and t*===0.8.
0.8.
0.8.

After ◦ and 1◦ under the case of high amplitudes,

Afterclarifying
After clarifying
clarifying thethe
the unusual
unusual
unusual lift
liftloss
lift loss
loss atat
at
ααα ===00°
0°and
and1° 1°under
underthe thecasecaseof ofhigh
highamplitudes,
amplitudes,
the flow
the flow
the mechanism
flow mechanism
mechanism of of higher
of higher angles
higher angles
angles of of attack
of attack
attack is is explained
is explained briefly.
explained briefly. Taking
briefly. Taking the result the resultof
result ofαα
of
α===2°
2 ◦ and a = 0.1c as an example, the velocity contour at t* = 0.8 is shown in Figure 21,
2°and
and aa == 0.1c0.1c asas an
an example,
example, the the velocity
velocity contour
contour at at t*
t* == 0.8
0.8 isis shown
shown in in Figure
Figure 21, 21, the
the
the pressure
pressure
pressure coefficient
coefficient
coefficient distributionsof
distributions
distributions ofofdifferent
differentmoments
different momentsare
moments are shown
are shown
shown in in Figure
in Figure22
Figure 22and
22 andthe
and the
the
change
changeofof
change oflift
liftcoefficients
lift coefficientsisis
coefficients isshown
showninin
shown inFigure
Figure23.
Figure 23.Figure
23. Figure21
Figure 21shows
21 showsthat
shows thatthe
that thedistance
the distancebetween
distance between
between
the airfoil leading edge and the wave crest is larger than that at = 0 ◦ and 1◦ . Compared
the airfoil
the airfoil leading
leading edge edge and and thethe wave
wave crest crest is is larger
larger than
than thatthat atat αα == 0°
α 0°andand 1°.1°. Compared
Compared
with
withFigure
with Figures
Figures 14d 14dor and
14d Figure
and 19,19,19,thethe
the blocking
blocking
blocking effect
effect
effect disappears.
disappears.
disappears. Thus,
Thus,
Thus, thethe
the low
low low velocity
velocity
velocity area
areaareaex-
ex-
expands
pands to
pands to nearly
to nearly
nearly the the whole
the whole
whole lowerlower surface
lower surface
surface of of the
of the airfoil
the airfoil instead
airfoil instead
instead of of assembling
of assembling
assembling at at the
at the leading
the leading
leading
edge.
edge.The
edge. Thehigh
The highpressure
high pressureregion
pressure regionnear
region nearthe
near thestagnation
the stagnationpoint
stagnation pointand
point andthe
and thefollowing
the followinglow
following lowpressure
low pressure
pressure
region
region in
region in Figures
in Figures
Figures 15 15 and
15 and
and 20 20
20 dodo not
do not occur.
not occur. Instead,
occur. Instead,
Instead, aa higha high pressure
high pressure
pressure region region
region occurs occurs
occurs at at the
at the
the area
area
area near
near the
near the
the trailingtrailing
trailing edge edge
edge as as
as shownshown
shown in in
in Figure Figure
Figure 22, 22,
22, whichwhich
which is is
is the the result
the result
result of of
of the the
the air air compressed
air compressed
compressed by by
by
thethe
the upward
upward
upward movement
movement
movement of of the
of the
the particles
particles
particles on the
on on water
the the
water water surface.
surface.
surface. The high
The The pressure
high high
pressurepressure area
areaagrees
area agrees
agrees
with the
with with
the liftthe
lift lift increase
increase
increase of the
of of
the lowerthe surface
lower lower surface
surface in Figure
in Figurein 23,
Figure
23, which
which 23,starts
whichat
starts atstarts
t* == 0.4
t* 0.4 t* the
at as
as =the0.4 as
wave
wave
the wave
under the under
the airfoil the
airfoil beginsairfoil
begins to begins
to move to
move upward move
upward and upward
and ends and
ends at at t* ends at
t* == 11 as
as thet* =
the wave1 as the
wave crest wave
crest passes crest
passes the the
under
passes
airfoilthe airfoil completely.
completely. The comparison The comparison
comparison variousof
of various various
angles angles
further further corroborates
corroborates the conclusion the
conclusion
airfoil completely. The of angles further corroborates the
conclusion that the upward movement of the wavy water surface can directly affect the
that the
that the upward
upward movementmovement of of the
the wavy
wavy waterwater surface
surface can can directly
directly affect
affect thethe pressure
pressure dis- dis-
pressure distribution on the lower surface of the airfoil. Besides, the distance between
tributionon
tribution onthe thelower
lowersurface
surface of ofthetheairfoil.
airfoil.Besides,
Besides,the the distance
distancebetweenbetweenthe theleading
leadingedge edge
the leading edge and the wave crest is the critical factor which determines whether the
and the
and the wave
wave crest crest isis the
the critical
critical factor
factor which
which determines
determines whether whether the the influence
influence will will gen-
gen-
influence will generate a blocking effect or transmit extra lift.
erate aa blocking
erate blocking effect effect oror transmit
transmit extra extra lift.
lift.
0.7
0.7
0.5
0.5
0.1 0.30.3
y/c
y/c
0.1 -0.1
-0.1
-0.3
-0.3

-0.5
-0.5 -0.2
-0.2 0.1
0.1 0.4
0.4 0.7
0.7 1.0
1.0 1.3
1.3 1.6
1.6
x/c
x/c

Figure21.
Figure
Figure 21.Velocity
21. Velocity
Velocity contour
contour
contour of
ofof α ==2◦2°
α α= 2° ata aa===0.1c
atat 0.1cand
0.1c andt*t*
and t*===0.8.
0.8.
0.8.
Appl. Sci.
Appl. Sci. 2021, 11,
11, 4215 PEER REVIEW 17 of 19
19
Appl. Sci.2021,
2021, 11,xxFOR
FOR PEER REVIEW 17
17 of
of 19

Figure
Figure22.
22.Pressure
Pressurecoefficient
coefficientdistributions
distributionsononthe
theairfoil
airfoilat
atdifferent
differentmoments
momentsat
ataa==0.1c
0.1cand
andαα==
Figure 22. Pressure coefficient distributions on the airfoil at different moments at a = 0.1c and α = 2◦ .
2°.
2°.

Figure
Figure 23.
23. Lift
Liftcoefficients
coefficients of
ofthe
of theupper
the upperand
upper andlower
and lowersurfaces
lower surfacesat
surfaces ataaa===0.1c
at 0.1c and
and ααα===2°.◦.
22°.

4.
4. Conclusions
4.Conclusions
Conclusions
The aerodynamics
The aerodynamics and flow
flow fields ofof the NACA
NACA 4412 airfoil
airfoil moving over
over the wavywavy
The aerodynamics andand flow fields
fields of the
the NACA 44124412 airfoil moving
moving over the the wavy
water surface
water surface with
with various
various wave
wave amplitudes
amplitudes areare numerically
numericallyinvestigated
investigatedin inthis
thispaper.
paper.
water surface with various wave amplitudes are numerically investigated◦in ◦this paper.
The calculations
The calculations are conducted
conducted in the
the range of low angles of attack (α == 00°~4°),
~4 ), and for for
The calculations areare conducted inin the range of low angles of attack (α (α = 0°~4°), andand for
each
each angle, the wave amplitude varies from 0.05c to 0.1c. Based on the result of this study,
eachangle,
angle,the
thewave
waveamplitude
amplitudevaries
variesfrom
from0.05c
0.05cto
to0.1c.
0.1c.Based
Basedononthe
theresult
resultofofthis
thisstudy,
study,
the following
the following conclusions
conclusions can
can be
be drawn.
drawn.
the following conclusions can be drawn.
(1)
(1) Thewavy
(1) The
The wavy water
wavy water surface
water surface willcause
surface will causethe thelift
the liftcoefficient
lift coefficientofof
coefficient ofthe
the
theairfoil
airfoil
airfoiltoto fluctuate
to fluctuate
fluctuate period-
peri-
peri-
ically.
odically. The
The fluctuation
fluctuation amplitude
amplitude ofofthe
the lift
lift coefficient
coefficient curves
curves
odically. The fluctuation amplitude of the lift coefficient curves increases remarkably increases remarkably
remarkably
withsuccessive
with
with successiveincreases
successive increasesin
increases in thewave
inthe
the waveamplitude
wave amplitudeand
amplitude andsharp
and sharppeaks
sharp peaksoccur
peaks occurin
occur inthe
in thecurves.
the curves.
curves.
WIG
WIG crafts flying over the ocean with high waves will generate severe oscillation.
WIGcrafts
craftsflying
flyingoveroverthe
theocean
oceanwithwithhigh highwaves
waveswill willgenerate
generatesevere severeoscillation.
oscillation.
(2) During
(2) Duringthe theperiodical
periodicalchange
changeofof lift, a high wave amplitude will lead to sudden
(2) During the periodical ◦ and 1change
◦ , which of lift, a high wave amplitude will lead to sudden
lift, a high wave amplitude will lead to sudden lift lift
lift
loss loss at = 0 even causes negative lift in some cases and threats
lossatatαα==0° 0°andand1°,1°,which
whicheven evencauses
causesnegative
negativelift liftin insome
somecasescasesand andthreats
threatsflight
α
flight
flight security
security for WIG crafts. With the increase of the angle of attack,significant
the significant
security forfor WIG
WIG crafts.
crafts. With
With thethe increase
increase of of the
the angle
angle of of attack,
attack, the the significant de- de-
decrease
crease in lift will gradually turn to theincrease
increaseininlift. lift.Therefore,
Therefore,WIG WIGcrafts
crafts should
crease in lift will gradually turn to the increase in lift. Therefore, WIG crafts should
in lift will gradually turn to the should
avoid cruisingat
avoid atsmall
smallangles
angles ofattack.
attack.
avoidcruising
cruising
◦ and 4at ◦ , small anglesof of attack.
(3)
(3) At α = 2 the time–averaged lift coefficientof of the airfoil
airfoil in proximity
proximity to the the
(3) At
At αα == 2°
2°and
and 4°,4°, the
the time–averaged
time–averaged lift lift coefficient
coefficient of the the airfoil in in proximity to to the
wavy
wavy water surface is larger than that of the flat ground, and it increases as the wave
wavywaterwatersurface
surfaceisislarger
largerthanthanthat
thatof ofthe
theflat
flatground,
ground,and andititincreases
increasesas asthe
thewave
wave
amplitudeincreases.
amplitude increases.This This is consistent with the former research, which indicated
amplitude increases. This is consistent with the former research, which indicatedthat
is consistent with the former research, which indicated that
thatvertical
the the vertical movement
movement of wavyof wavy
water water
surface surface
would would compress
compress the theand
air air lead
and leadto
the
to avertical
significantmovement
increase. of wavy
While water
at α = surface
0 would compress the
◦ and 1◦ , the time–averaged lift coefficientair and lead to aa
significant
significant increase.
increase. While
While at
at αα == 0°0°and
and 1°, 1°,andthe time–averaged
the the
time–averaged lift
lift coefficient
coefficient de-
de-
decreases
creases as as
the the wave
wave amplitude
amplitude increases,
increases, and the lift liftcaused
loss loss caused
by the by the ground
ground effect
creases as the
effect enhances. wave amplitude increases, and the lift loss caused by the ground effect
enhances.
enhances.
Appl. Sci. 2021, 11, 4215 18 of 19

(4) For a lower wave amplitude at α = 0◦ and 1◦ , the wavy water surface affects the flow
field indirectly by driving the air to move vertically, thereby changing the shape of the
actual flow channel. Under this situation, the change in the lift coefficient is mainly
influenced by the contraction and expansion of the flow channel as the airfoil moves
over wavy water.
(5) For a higher wave amplitude at α = 0◦ and 1◦ , the pressure distributions on the airfoil
are directly impacted by the vertical movement of particles on the wavy water surface.
When the wave crest moves below the leading edge stagnation, the compression by
the wave creates a high pressure region and prevents air from flowing under the
lower surface. A low pressure region consequently occurs after the leading edge and
decreases the lift coefficient significantly. This phenomenon is caused by the close
distance between the leading edge stagnation and the wave crest, and disappears at
higher angles of attack.
(6) The upward movement of high amplitude waves can transmit hydrodynamic force to
the airfoil by compressing the air under the airfoil. This is the reason for the significant
lift increase at α = 2◦ and 4◦ . The distance between the leading edge and the wave
crest is the critical factor which determines whether the impact will be blocked and
results in the inconsistent lift change at different angles of attack.
This paper studies the influence of wave amplitudes on the aerodynamics of the
airfoil moving over the wavy water surface, providing a new understanding of the ground
effect in the case of wavy water surface. These findings are instructive for the design and
application of WIG aircrafts, especially under high grade sea conditions.

Author Contributions: Conceptualization, X.L., D.M. and H.H.; methodology, Y.G. and H.H.; soft-
ware, X.L.; validation, X.L. and H.H.; formal analysis, Y.G. and H.H.; investigation, X.L. and Y.G.;
resources, D.M.; data curation, X.L.; writing—original draft preparation, X.L.; writing—review and
editing, M.Y. and H.H.; visualization, X.L.; supervision, D.M. and M.Y.; project administration, D.M.
and M.Y. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: The data presented in this study are available on request from the
corresponding author.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Rozhdestvensky, K.V. Wing-in-ground effect vehicles. Prog. Aerosp. Sci. 2006, 42, 211–283. [CrossRef]
2. Hsiun, C.-M.; Chen, C.-K. Aerodynamic Characteristics of a Two-Dimensional Airfoil with Ground Effect. J. Aircr. 1996,
33, 386–392. [CrossRef]
3. Ockfen, A.E.; Matveev, K.I. Aerodynamic characteristics of NACA 4412 airfoil section with flap in extreme ground effect. Int. J.
Nav. Arch. Ocean Eng. 2009, 1, 1–12. [CrossRef]
4. Jung, J.H.; Yoon, H.S.; Chun, H.H.; Hung, P.A.; Elsamni, O.A. Mean flow characteristics of two-dimensional wings in ground
effect. Int. J. Nav. Arch. Ocean Eng. 2012, 4, 151–161. [CrossRef]
5. Qu, Q.; Wang, W.; Liu, P.; Agarwal, R.K. Airfoil Aerodynamics in Ground Effect for Wide Range of Angles of Attack. AIAA J.
2015, 53, 1048–1061. [CrossRef]
6. Qu, Q.; Jia, X.; Wang, W.; Liu, P.; Agarwal, R.K. Numerical study of the aerodynamics of a NACA 4412 airfoil in dynamic ground
effect. Aerosp. Sci. Technol. 2014, 38, 56–63. [CrossRef]
7. Qin, Y.; Liu, P.; Qu, Q.; Guo, H. Numerical study of aerodynamic forces and flow physics of a delta wing in dynamic ground
effect. Aerosp. Sci. Technol. 2016, 51, 203–221. [CrossRef]
8. Lange, R.H.; Moore, J.W. Large Wing-in-Ground Effect Transport Aircraft. J. Aircr. 1980, 17, 260–266. [CrossRef]
9. Zhang, X.; Zerihan, J. Off-Surface Aerodynamic Measurements of a Wing in Ground Effect. J. Aircr. 2003, 40, 4. [CrossRef]
10. Jamei, S.; Maimun, A.; Mansor, S.; Azwadi, N.; Priyanto, A. Numerical Investigation on Aerodynamic Characteristics of a
Compound Wing-in-Ground Effect. J. Aircr. 2012, 49, 1297–1305. [CrossRef]
Appl. Sci. 2021, 11, 4215 19 of 19

11. Jung, J.H.; Kim, M.J.; Yoon, H.S.; Hung, P.A.; Chun, H.H.; Park, D.W. Endplate effect on aerodynamic characteristics of
three-dimensional wings in close free surface proximity. Int. J. Nav. Arch. Ocean Eng. 2012, 4, 477–487. [CrossRef]
12. Barber, T.J. A Study of Water Surface Deformation Due to Tip Vortices of a Wing-in-Ground Effect. J. Ship Res. 2007, 51, 182–186.
[CrossRef]
13. Zong, Z.; Liang, H.; Zhou, L. Lifting line theory for wing-in-ground effect in proximity to a free surface. J. Eng. Math. 2012,
74, 143–158. [CrossRef]
14. Liang, H.; Zhou, L.; Zong, Z.; Sun, L. An analytical investigation of two-dimensional and three-dimensional biplanes operating in
the vicinity of a free surface. J. Mar. Sci. Tech. 2013, 18, 12–31. [CrossRef]
15. Bal, S. Free surface effects on 2-D airfoils and 3-D wings moving over water. Ocean Syst. Eng. 2016, 6, 245–264. [CrossRef]
16. Im, Y.; Chang, K. Unsteady Aerodynamics of a Wing-in-Ground-Effect Airfoil Flying over a Wavy Wall. J. Aircr. 2000, 37, 690–696.
[CrossRef]
17. Qu, Q.; Lu, Z.; Liu, P.; Agarwal, R.K. Numerical Study of Aerodynamics of a Wing-in-Ground-Effect Craft. J. Aircr. 2014,
51, 913–924. [CrossRef]
18. Gao, B.; Qu, Q.; Agarwal, R.K. Aerodynamics of a Transonic Airfoil above Wavy Ground. In Proceedings of the 2018 AIAA
Aerospace Sciences Meeting, Kissimmee, FL, USA, 8–12 January 2018. [CrossRef]
19. Lee, T.; Tremblay-Dionne, V. Experimental Investigation of the Aerodynamics and Flowfield of a NACA 0015 Airfoil over a Wavy
Ground. J. Fluid Eng. 2018, 140, 071202. [CrossRef]
20. Tremblay-Dionne, V.; Lee, T. Effect of Trailing-Edge Flap Deflection on a Symmetric Airfoil over a Wavy Ground. J. Fluid Eng.
2019, 141, 064501. [CrossRef]
21. Liang, H.; Zong, Z.; Zou, L. Nonlinear lifting theory for unsteady WIG in proximity to incident water waves. Part 1: Two-
dimension. Appl. Ocean Res. 2013, 43, 99–111. [CrossRef]
22. Liang, H.; Zong, Z.; Zou, L. Nonlinear lifting theory for unsteady WIG in proximity to incident water waves. Part 2: Three-
dimension. Appl. Ocean Res. 2013, 43, 88–98. [CrossRef]
23. Liang, H.; Wang, X.; Zou, L.; Zong, Z. Numerical study of two-dimensional heaving airfoils in ground effect. J. Fluid Struct. 2014,
48, 188–202. [CrossRef]
24. Kaneko, Y.; Omori, T.; Kajishima, T. Numerical study on the aerodynamics of an airfoil moving close to an air-water interface.
Trans. JSME 2016, 82, 16-00112. (In Japanese) [CrossRef]
25. Zhi, H.; Xiao, T.; Chen, J.; Wu, B.; Tong, M.; Zhu, Z. Numerical Analysis of Aerodynamics of a NACA4412 Airfoil above Wavy
Water Surface. In AIAA Aviation 2019 Forum; American Institute of Aeronautics and Astronautics: Reston, VA, USA, 2019.
[CrossRef]
26. Hu, H.; Ma, D.; Guo, Y.; Yang, M. Airfoil Aerodynamics in Proximity to Wavy Water Surface. J. Aerosp. Eng. 2021, 34, 04020119.
[CrossRef]
27. Mi, B. Numerical investigation on aerodynamic performance of a ducted fan under interferences from the ground, static water
and dynamic waves. Aerosp. Sci. Technol. 2020, 100, 105821. [CrossRef]
28. Hirt, C.W.; Nichols, B.D. Volume of fluid (VOF) method for the dynamics of free boundaries. J. Comput. Phys. 1981, 39, 201–225.
[CrossRef]
29. Sundar, V. Ocean Wave Mechnics: Applications in Marine Structures, 1st ed.; John Wiley & Sons, Ltd.: Chichester, UK, 2015; pp. 41–52.
[CrossRef]
30. Elhanafi, A.; Fleming, A.; Leong, Z.; Macfarlane, G. Effect of RANS-based Turbulence Models on Nonlinear Wave Generation in a
Two-Phase Numerical Wave Tank. Prog. Comput. Fluid Dyn. Int. J. 2017, 17, 141–158. [CrossRef]
31. Firooz, A.; Gadami, M. Turbulence Flow for NACA 4412 in Unbounded Flow and Ground Effect with Different Turbulence
Models and Two Ground Conditions: Fixed and Moving Ground Conditions. In Proc. Conf. on Boundary and Interior Layers;
University of Gottingen: Lower Saxony, Germany, 2006; Volume 98, pp. 161–164.
32. Ahmed, M.R.; Takasaki, T.; Kohama, Y. Aerodynamics of a NACA4412 Airfoil in Ground Effect. AIAA J. 2007, 45, 37–47.
[CrossRef]