You are on page 1of 20

Applied Mathematical Modelling 83 (2020) 683–702

Contents lists available at ScienceDirect

Applied Mathematical Modelling


journal homepage: www.elsevier.com/locate/apm

Computational fluid dynamics simulation of aerodynamic


performance and flow separation by single element and
slatted airfoils under rainfall conditions
Hossein Fatahian a, Hesamoddin Salarian a,∗, Majid Eshagh Nimvari b,
Jahanfar Khaleghinia a
a
Department of Mechanical Engineering, Nour Branch, Islamic Azad University, Nour, Iran
b
Faculty of Engineering, Amol University of Special Modern Technologies, Amol, Iran

a r t i c l e i n f o a b s t r a c t

Article history: This study investigated the effects of rainfall on flow separation and the aerodynamic per-
Received 15 November 2019 formance of single element and slatted NACA 0012 airfoils by using a mathematical model
Revised 17 January 2020
developed with the commercial computational fluid dynamics solver ANSYS FLUENT 18.2.
Accepted 27 January 2020
A two-way momentum coupled Eulerian–Lagrangian multiphase approach was used to
Available online 31 January 2020
simulate the formation of the water film layer on the airfoil’s surface. According to the
Keywords: results, very low values of the lift-to-drag ratio at low angles of attack reflected severe
Aerodynamic performance degradation of the aerodynamic performance of the airfoil in the presence of water accu-
Airfoil mulated on its surface. The impact of rain droplets on the leading-edge slat surface led
Flow separation to less water accumulating on the main section of the airfoil. In particular, the maximum
Leading-edge slat water film mass concentrated on the airfoil surface decreased from 15 g to 1 g compared
Rainy conditions with the single element airfoil. Hence, the thickness of the water film layer was not suf-
ficiently large to significantly affect the aerodynamic coefficients of the slatted airfoil, es-
pecially the maximum lift coefficient, compared with the thicker water film layer on the
single element airfoil. In addition, the use of slats clearly enhanced the aerodynamic co-
efficients and increased the stall angle from 13° to 22° in dry conditions, and from 16° to
24° in rainy conditions. Slats also significantly decreased the boundary layer thickness and
delayed the separation at higher angles of attack.
© 2020 Elsevier Inc. All rights reserved.

1. Introduction

The effect of heavy rain on the aerodynamic efficiency is a major reason for many aircraft crashes involving both trans-
port and general aviation airplanes [1]. The effect of rainfall on the flight of aircraft was first investigated in a wind tunnel
test by Rhode [2] who found that exposure to heavy rain was not sufficient to force aircraft to descend to the ground.
Haines and Luers [3,4] investigated the impacts of heavy rain on aircraft and concluded that heavy rain caused roughen-
ing of the wing, which led to a decrease in the stall angle. Thompson and Marrochello [5] determined the position of the
onset of rivulet formation for a surface water flow over a wing with a NACA 4412 airfoil and compared the results with


Corresponding author.
E-mail address: h_salarian@iaunour.ac.ir (H. Salarian).

https://doi.org/10.1016/j.apm.2020.01.060
0307-904X/© 2020 Elsevier Inc. All rights reserved.
684 H. Fatahian, H. Salarian and M.E. Nimvari et al. / Applied Mathematical Modelling 83 (2020) 683–702

Nomenclature

μ Air dynamic viscosity


c Airfoil chord length
α Angle of attack
ρD Density of droplets
ρ Density of fluid phase
ρw Density of water
Dp Diameter of rain particles
Y+ Dimensionless wall distance
ε Dissipation rate
CD Drag coefficient
μT Eddy viscosity
gi Acceleration due to gravity
I Injection line
CL Lift coefficient
LWC Liquid water content
mP Mass of parcel
Sαq Mass source term
Fi Momentum source term
Re p Particle Reynolds number
P Pressure
R Rainfall rate
Re Reynolds number
UT Terminal velocity
t Time
u Velocity vector of air
up Velocity vector of raindrop particles
αq Volume fraction

experimental data. The effects of heavy rain on the aerodynamic efficiency of the NACA 0012 airfoil were also in-
vestigated by Ismail et al. [6] who demonstrated that heavy rainfall conditions resulted in a remarkable increase in
drag and a decrease in lift. Wu [7] studied the impact of raindrops on an airfoil surface under natural drop-laden
two-phase flow conditions and found that the negative effects of water were reflected by the loss of lift and in-
creased drag. Wu et al. [8] investigated the effects of heavy rain on the stability of a DHC-6 Twin Otter aircraft and
showed that heavy rain could degrade the flight mechanical performance of aircraft. Ismail et al. [9] investigated the
effects of heavy rain on the aerodynamic efficiency of the NACA 23015 airfoil cruise and landing configurations, and
the NACA 64210 high lift configuration. In the simulations conducted for NACA 23105, the increase in drag was lower
compared with the simulations conducted for the NACA 64210 high lift configuration airfoil. Wu et al. [10] studied
the aerodynamic performance of the DLR-F12 transport aircraft under heavy rain conditions and their results indicated
that heavy rain degraded the dynamic flight performance and more fuel was consumed to compensate for the loss of
lift.
Rainfall at a rate of 1800 mm/h reduces the lift by 30% and increases the drag value by 20%, which can influ-
ence flow separation, the stall angle, flight safety, and aircraft maneuverability [11,12]. Flow separation leads to large
energy losses, which negatively affect the aerodynamic performance due to the loss of lift and increased drag. Thus,
it is necessary to delay or manage the occurrence of flow separation [13–16]. Flow separation can be controlled us-
ing active or passive methods. Many studies have investigated aerodynamic flow separation control using both active
and passive flow control methods [17–20]. Leading-edge slats are used as a passive flow control technique to prevent
leading-edge separation at low speeds by injecting a high momentum fluid through the gap between the slat and main
airfoil.
Airplanes cannot avoid encountering rainfall during flights, particularly in the take-off and landing stages [21]. Rain-
fall conditions have adverse effects on aerodynamic performance and many severe aviation accidents have occurred in
rainfall conditions. To the best of our knowledge, previous studies have not considered the effects of rainfall on flow
separation and the aerodynamic performance of a slatted airfoil, as well as the performance degradation caused by
rainfall when using a single element airfoil and slatted airfoil. Thus, these effects were investigated in the present
study. In addition, the aerodynamic performance and stall angles were compared for a single element airfoil un-
der different conditions, i.e., dry and rainy conditions, as well as those for a slatted airfoil under the same two
conditions.
H. Fatahian, H. Salarian and M.E. Nimvari et al. / Applied Mathematical Modelling 83 (2020) 683–702 685

2. Computational procedure

2.1. Modeling rainfall conditions

Evaluating the rainfall rate is required for rainfall simulations. A rainfall rate of 100 mm/h or greater is assumed to
represent heavy rain [22]. For both numerical and experimental studies, the relationship between the rainfall rate R (mm/h)
and liquid water content LWC(g/m3 ) was given by Bezos et al. [23] for thunder-type storms or heavy rain (1) as follows.
LW C = 0.054 × R0.84 (1)
For light rain (2), this formula is written as follows.
LW C = 0.0889 × R0.84 (2)
In addition, the velocity of the rain particles falling on the surface of the airfoil must be computed. The uniform velocity
of rain particles before they contact the airfoil’s surface is called the terminal velocity (UT ). This velocity varies with the
diameter of the particles and can be written as Eq. 3 [24]:
  
DP 1.147
UT = 9.58 1 − exp − , (3)
1.77
where DP is the diameter of the rain particles.

2.2. Two-phase modeling

In the present study, a coupled Eulerian–Lagrangian approach was used to model the formation of a water layer on the
airfoil due to rain. The Eulerian–Lagrangian approach solves the continuum conservation equations for the continuous phase
and integrates the Lagrangian equations of motion for the discrete phase, which is defined as the discrete phase model
(DPM) approach [22]. The parcel approach was used for calculating the droplet trajectories. Several droplets were assumed
to be present in each parcel, with the same density, diameter, and velocity. These rain parcels were injected into the airflow
field from an injection line (I) with a length of 10 chords and located three chords from the leading edge of the airfoil,
with a terminal velocity equal to the air stream velocity and uniform droplet diameter of 1 mm. In each parcel, the number
of droplets was selected such that the desired rainfall rate was maintained. In total, 300 parcels were injected from the
injection line at every flow time step of 1 × 10–4 s.

2.3. Continuous phase

In the present study, in order to solve the flow field around the airfoil, the unsteady Reynolds averaged Navier–Stokes
equations (URANS) of mass and momentum were solved with the k–ɛ standard turbulence model. This model was recom-
mended in previous studies because of its high accuracy at modeling the turbulent flow over an airfoil under heavy rainfall
conditions via a two-phase flow approach [8,9,12,25].
The Eulerian volume of fluid (VOF) model was implemented to simulate the air that conveys the rain particles. The Geo-
Reconstruct scheme was used to track the interface between the phases in order to resolve the interface between the airflow
and water layer [26].
For the qth phase, the continuity equation (4) is defined as follows [27]:

(α ρ ) + ∇ (αq ρq uq ) = Sαq , (4)
∂t q q
where α q , ρ q , and uq are the volume fraction, density, and velocity of the qth phase, respectively, and Sαq denotes the mass
source term for the qth phase. The primary phase volume fraction was calculated under the limitation that the volume
fractions of all phases summed to unity in each control volume. The momentum equation of the VOF model is written as
Eqs. 5 and 6:
∂ρ ui ∂ ∂ P ∂ τ ji
+ ( ρ u j ui ) = − + + Fi (5)
∂t ∂xj ∂ xi ∂ x j
 
∂ ui ∂ u j
τ ji = μ + − ρ u j u i , (6)
∂ x j ∂ xi
where Fi represents the momentum source term, −ρ u j u i is the Reynolds stress, ρ and μ are the air density and dynamic
viscosity, respectively, and ui is the air velocity component.
The k–ε standard turbulence model equation is given as Eqs. 7 and 8 [28]:
 
∂ ( ρ k ) ∂ ( ρ kui ) ∂ μt ∂ k
+ = + 2μt Ei j Ei j − ρε (7)
∂t ∂ xi ∂ x j σk ∂ x j
686 H. Fatahian, H. Salarian and M.E. Nimvari et al. / Applied Mathematical Modelling 83 (2020) 683–702

 
∂ (ρε ) ∂ (ρε ui ) ∂ μt ∂ε ε ε2
+ = + C1ε 2μt Ei j Ei j − C2ε ρ , (8)
∂t ∂ xi ∂ x j σε ∂ x j k k
2
where Eij is the component of the deformation rate and μt = ρCμ kε is the eddy viscosity. These equations contain some
adjustable constants comprising σ k ,σ ε ,C1ε ,C2ε , and Cμ , and the values of these constants are defined as follows [28].
Cμ = 0.09 σk = 1.00 σε = 1.30 C1ε = 1.44 C2ε = 1.92
Two source terms are added to the continuity and momentum equations in the VOF model to couple the DPM and VOF
models. When each rain droplet strikes the airfoil’s surface or enters the water layer around it, the droplets are removed
from the computational domain and the effect is accounted for by adding an instantaneous source term to both the VOF
volume fraction equation for the liquid phase and the momentum equation of the mixture at the beginning of that time
step. These source terms are then set to zero at the end of the time step. The values of these source terms are determined
by the properties of the impacting particle in the DPM model.
Mass source term for the liquid phase (9):
mP
Sα = . (9)
V t
Momentum source term for the mixture (10):
mP uri
Fi = , (10)
V t
where uri is defined in Eq. 11:
uri = uPi − u f i , (11)
mP is the mass of a parcel, t is the time-step size for the continuous phase solver, uri is the droplet velocity relative to the
continuous fluid phase in the i-direction, and V is the volume of the computational cell.

2.4. Discrete phase

The phase was considered to be a discrete phase for the rainfall particles. Thus, the DPM was applied to model the rain
particles. In the DPM, the wall film is formed by injecting the rain particles into the continuous flow field. In this model, the
rain particles impinge upon the airfoil’s upper and lower surfaces, and a thin film is formed [29,30]. In addition, raindrops
are assumed to be non-evaporating, non-interacting, and non-deforming spheres, which are subject only to drag and gravity
forces [22]. Bilanin [31] justified the assumption of non-interacting droplets. Furthermore, a two-way momentum coupled
Eulerian–Lagrangian approach was implemented to consider the two-way exchanges of mass, momentum, and energy be-
tween the discrete and continuous phases. The Lagrangian equations of motion for the raindrops (12) are written as [27]:

duPi
ρ − ρ
= β (ui − uPi ) + gi
D
, (12)
dt ρD
where uPi and ui are the parcel and fluid phase velocities in the i-direction, respectively, β (ui − uPi ) is the drag force per
unit particle mass, gi is the acceleration due to gravity, and ρ D and ρ are the densities of the droplets and the fluid phase.
β is defined in Eq. 13:
18μ CDd Re p
β= , (13)
ρP D2P 24
where Re p is the particle Reynolds number and CDd is the droplet drag coefficient used in the spherical drag law [32]. The
drag coefficient CDd is given by Eq. 14:

K1 K2
CDd = + 2 + K3 , (14)
Re p Re p
where K1 , K2 , and K3 are parameters that depend on the raindrop impact energy. The particle Reynolds number (Re p ) is
written as follows (15).

ρ DP ui − u pi
Re p = (15)
μ
3. Computational grid

Fig. 1 shows a two-dimensional view of the computational domain for the NACA 0012 airfoil with a chord length of 1 m,
which is extended as a circular region with a fixed diameter. The inlet has a uniform velocity corresponding to a chord based
on a Reynolds number (Re) of 3.1 × 105 . The outlet is defined as a pressure outlet and the no-slip condition is assumed on
H. Fatahian, H. Salarian and M.E. Nimvari et al. / Applied Mathematical Modelling 83 (2020) 683–702 687

Fig. 1. Two-dimensional view of the computational domain.

Fig. 2. Domain extent independence test results.

the airfoil surface. The computational domain is sufficiently large to avoid external effects on the airfoil flow field. Three
different computational domains with diameters equal to 20, 25, and 30 chords from the leading edge of the airfoil were
generated to consider the pressure coefficient (CP ) for the domain extent independence test, as shown in Fig. 2. This test
was conducted at an angle of attack of 20° and Re = 3.1 × 105 for the single element airfoil under dry conditions. It was
concluded that the domain with a diameter of 25 chords for the airfoil was sufficient for the present study.
An injection line (I) with a length of 10 chords was selected to inject the rain particles into the computational domain.
Four different injection line positions located one, two, three, and four chords upstream of the leading edge of the airfoil
were selected to determine a sufficient distance for this line to represent the airfoil under rainy conditions. The mass distri-
bution of the water film over the airfoil for the single element airfoil under rainy conditions at α = 20° is presented in Fig. 3.
It was concluded that the injection line located three chords upstream of the leading edge of the airfoil was sufficiently far
to simulate rainy conditions.
688 H. Fatahian, H. Salarian and M.E. Nimvari et al. / Applied Mathematical Modelling 83 (2020) 683–702

Fig. 3. Position of injection line in the independence test.

Fig. 4. O-type structured grid.

A two-dimensional structured O-type grid was employed to discretize the computational domain, as shown in Fig. 4. A
view of the grid for the controlled case is presented in Fig. 5. The finest grids were generated between the leading-edge slat
and main section of the airfoil. As shown in Figs. 6 and 7, the value of Y+ was kept at less than 1 by considering that the
distance of the nearest node from the airfoil surface was 1 × 10−5 (m) [28]. Further details of the grid cells and distribution
of Y+ at α = 10° and Re = 3.1 × 105 are presented in Tables 1 and 2. Moreover, the grid independence study was conducted
by comparing the lift and drag coefficients with four different cell sizes for liquid water contents (LWCs) of 0 g/m3 and
30 g/m3 , as shown in Tables 3 and 4, respectively. After considering the grid independence study, it was found that the grid
with 460 0 0 cells was suitable for use in the present study.

4. Numerical method

In the present study, the URANS were solved using the commercial computational fluid dynamics (CFD) solver ANSYS
FLUENT 18.2 with the pressure-based finite volume method. The effects of turbulence were modeled using the k–ɛ two-
H. Fatahian, H. Salarian and M.E. Nimvari et al. / Applied Mathematical Modelling 83 (2020) 683–702 689

Fig. 5. View of the grid for a slatted airfoil.

Fig. 6. Distributions of Y+ over the NACA 0012 airfoil at angles of attack of 10° and 14° under dry conditions.

Table 1
Distribution of Y+ at α = 10° and Re = 3.1 × 105 , and details of the grid cells under dry conditions.

Grid Number of cells Growth factor Height of the first cell (m) Maximum value of Y+ Minimum value of Y+ Average value of Y+

1 18000 1.1 2 × 10–3 14.13 3.42 8.14


2 32000 1.1 1 × 10–4 6.14 0.83 2.91
3 46000 1.1 1 × 10–5 0.65 0.01 0.24
4 65000 1.1 3 × 10–6 0.61 0.01 0.22

Table 2
Distribution of Y+ at α = 10° and Re = 3.1 × 105 , and details of the grid cells under rainy condition.

Grid Number of cells Growth factor Height of the first cell (m) Maximum value of Y+ Minimum value of Y+ Average value of Y+

1 18000 1.1 2 × 10–3 14.88 3.91 8.72


2 32000 1.1 1 × 10–4 6.63 0.96 3.27
3 46000 1.1 1 × 10–5 0.76 0.06 0.34
4 65000 1.1 3 × 10–6 0.68 0.03 0.26
690 H. Fatahian, H. Salarian and M.E. Nimvari et al. / Applied Mathematical Modelling 83 (2020) 683–702

Fig. 7. Distributions of Y+ over the NACA 0012 airfoil at angles of attack of 10° and 14° under rainy conditions.

Table 3
Grid independence study results for LWC = 0 g/m3 .

Number of cells α = 10° α = 14°


CL CD CL CD

18000 0.922 0.095 1.081 0.146


32000 0.868 0.082 0.951 0.134
46000 0.839 0.070 0.928 0.125
65000 0.837 0.068 0.926 0.122

Table 4
Grid independence study results for LWC = 30 g/m3 .

Number of cells α = 10° α = 14°


CL CD CL CD

18,000 0.882 0.109 1.174 0.143


32,000 0.824 0.094 0.962 0.131
46,000 0.780 0.086 0.939 0.122
65,000 0.777 0.083 0.936 0.120

Table 5
Time step independence study results for a single element airfoil under
dry conditions at α = 10° and Re = 3.1 × 105 .

CL CD Time step (s)

0.807 0.089 5 × 10–3


0.822 0.081 2 × 10–3
0.839 0.070 1 × 10–4
0.841 0.068 1 × 10–5

equation turbulence model. The SIMPLE pressure–velocity coupling algorithm and upwind second-order method were im-
plemented to discretize the governing equations. The Geo-Reconstruct scheme was utilized to discretize the Eulerian VOF
model by resolving the interface between the airflow and water layer [28].
The conversion criterion was set equal to 10–6 . Furthermore, a time-step independence study was conducted (Table 5) to
obtain a Courant–Friedrichs–Lewy number lower than 1, where four different time steps were selected to compare the lift
and drag coefficients for a single element airfoil under dry conditions at α = 10°. The results showed that a time step of
1 × 10–4 was acceptable for the present calculations.
In the numerical procedure, the injection of rain particles was initiated at t = 1.85 s to ensure that the single-phase sim-
ulation reached a steady state before the injection of the rain particles commenced. All of the simulations were conducted
for 16 s to ensure that they all reached a quasi-steady state under rainy condition. In total, 300 parcels were injected from
H. Fatahian, H. Salarian and M.E. Nimvari et al. / Applied Mathematical Modelling 83 (2020) 683–702 691

Fig. 8. Time histories of lift and drag coefficients at α = 10° and α = 16°.

Fig. 9. Comparison of the computational results obtained in the present study with the numerical and experimental results reported by Amaral et al. [33].
692 H. Fatahian, H. Salarian and M.E. Nimvari et al. / Applied Mathematical Modelling 83 (2020) 683–702

the injected line at every flow time step of 1 × 10–4 s and the number of rain particles in each parcel was determined so
the required rainfall rate was maintained.
Fig. 8 presents the time histories of the lift and drag coefficients at two angles of attack (α = 10° and 16°) at
Re = 3.1 × 105 . The plots are shown for a single element airfoil under dry conditions. Clearly, the time histories of the
lift and drag coefficients exhibited a combination of low and high-frequency oscillations at both angles of attack. The final
lift and drag coefficients obtained from the numerical study comprised the mean values.

5. Validation

Unfortunately, no previous numerical or experimental investigations have considered the effects of rain on the compo-
nents of a multi-element lifting surface [21]. Hence, the numerical model of the MD30P30N airfoil as a multi-element airfoil
was validated under conditions without rain at an angle of attack of 16°, as shown in Fig. 9. The pressure coefficient (Cp ) de-
termined in the present study was compared with the numerical and experimental results obtained by Amaral et al. [33] at
a Reynolds number of 9.62 × 105 and angle of attack of 16°. Good agreement was obtained between the computational
results obtained in the present study and the numerical and experimental data reported by Amaral et al. [33].
In addition, the lift and drag coefficients computed in the CFD simulation of a single element airfoil were compared
with the experimental results reported by Hansman and Craig [34], and the numerical results given by Wu et al. [12], as
shown in Fig. 10. Rainy conditions with a LWC of 30 g/m3 and Reynolds number of 3.1 × 105 were selected in order to

Fig. 10. Comparison of the computational results obtained in the present study with the experimental results reported by Hansman and Craig [34] and the
numerical results given by Wu et al. [12].
H. Fatahian, H. Salarian and M.E. Nimvari et al. / Applied Mathematical Modelling 83 (2020) 683–702 693

Fig. 10. Continued

ensure consistency with the experimental results reported by Hansman and Craig [34]. Our numerical results were in good
agreement with the previously reported experimental and numerical results. Hence, the turbulent model and numerical
solving method exhibited reliable accuracy.

6. Results and discussion

6.1. Effects of rain on aerodynamic coefficients

Figs. 11 and 12 show the variations in the lift and drag coefficients for single element and slatted airfoils under both
dry and rainy conditions. As mentioned above, the rainfall intensity was 30 g/m3 and the Reynolds number was 3.1 × 105 .
Figs. 11 and 12 show that rainfall inevitably had a negative impact on the aerodynamic performance of the airfoil over a
wide range of angles of attack. Clearly, the lift coefficient decreased whereas the drag coefficient increased up to 60% at
low angles of attack due to heavy rain. The maximum decrease of 10% in the lift coefficient was obtained at angles of
attack of 2–13° under heavy rain conditions because of water film formation. However, the slat had a remarkable effect on
improving the aerodynamic performance of the airfoil under both dry and rainy conditions. In general, the slatted airfoil
produced higher lift and lower drag compared with the single element airfoil under both dry and rainy conditions. The
critical angles of attack where the airfoil achieved the maximum lift coefficients were 13° and 16° under dry and rainy
conditions, respectively. Thus, the stall angle increased from 13° under dry conditions to 16° under rainy conditions, which
was a direct consequence of a premature boundary-layer transition induced by the rain [34] causing the delay of stalling.
Another interesting observation was the effect of the slat on improving the stall angle. The airfoil exhibited a sharp
deterioration in its aerodynamic performance as the lift decreased and drag increased after stalling under dry and rain con-
694 H. Fatahian, H. Salarian and M.E. Nimvari et al. / Applied Mathematical Modelling 83 (2020) 683–702

Fig. 11. Effects of rain on the lift coefficient at different angles of attack.

Fig. 12. Effects of rain on the drag coefficient at different angles of attack.

Fig. 13. Effects of rain on the lift-to-drag ratio at different angles of attack.
H. Fatahian, H. Salarian and M.E. Nimvari et al. / Applied Mathematical Modelling 83 (2020) 683–702 695

Fig. 14. Streamlines around the single element airfoil colored with velocity magnitude contours (m/s) at an angle of attack of 20° for dry (left) and rainy
(right) conditions.

Fig. 15. Streamlines around the slatted airfoil colored with velocity magnitude contours (m/s) at an angle of attack of 20° for dry (left) and rainy (right)
conditions.

Fig. 16. Contours of the water film thickness (m) on the upper surface of the slatted airfoil at an angle of attack of 20° for dry (left) and rainy (right)
conditions.

ditions. The differences between the slatted and single element airfoils were clearly demonstrated under these conditions.
The slat resulted in a significant increase in the stall angle from 13° to 22° under dry conditions, and the stall angle in-
creased from 16° to 24° under rainy conditions. Under rainy conditions, the slat increased the maximum value of the lift
coefficient from 1.07 at α = 16° to 1.64 at α = 24°. These desirable changes in the lift and drag coefficients as well as the
stall angle indicate that the slatted airfoil obtained better aerodynamic performance than the single element airfoil.
696 H. Fatahian, H. Salarian and M.E. Nimvari et al. / Applied Mathematical Modelling 83 (2020) 683–702

The lift-to-drag ratios representing the aerodynamic performance of the airfoil are presented in Fig. 13 for dry and rainy
conditions. The sensitivity of the aerodynamic performance of the airfoil was computationally investigated under a rainfall
intensity of 30 g/m3 . It should be noted that the aerodynamic performance of the airfoil degraded significantly because of
water film formation, especially at lower angles of attack. The maximum degradation of the lift-to-drag ratio was about 56%
at an angle of attack of 2° due to rainfall. The very low values of the lift-to-drag ratios at low angles of attack reflected the
severe degradation of the aerodynamic performance of the airfoil in the presence of water accumulation on the airfoil. The
slatted airfoil produced a higher lift-to-drag ratio under both dry and rainy conditions, especially at high angles of attack.
After stalling heavily, the presence of the slat increased the value of the lift-to-drag ratio under dry and rainy conditions
compared with the single element airfoil. The increase in the lift-to-drag ratio was remarkable under dry conditions. The
slat enhanced the aerodynamic performance of the airfoil but rain resulted in greater degradation of the aerodynamic per-
formance of the slatted airfoil compared with the slatted airfoil under dry conditions. The maximum lift-to-drag ratio was
observed at α = 12°. At this angle, the lift-to-drag ratio decreased from 14.09 to 9.88 with a slatted airfoil in the presence
of rain. By contrast, rain reduced this ratio from 10.36 to 9.27 for the single element airfoil at α = 12°.

6.2. Effects of rain on the flow field around the airfoil

The streamlines generated around the single element and slatted airfoils at an angle of attack of 20° colored with ve-
locity magnitude contours (m/s) are presented in Figs. 14 and 15 for dry and rainy conditions. For the single element air-
foil, a large recirculation region expanded on the suction side of the airfoil, possibly because heavy stalling occurred at
this angle under dry and rainy conditions, thereby considerably reducing the lift coefficient and increasing the drag coef-
ficient. As expected, using the leading-edge slat had a remarkable effect on the control of flow separation over the airfoil.

Fig. 17. Boundary layer velocity profiles at four chordwise positions on the upper surface of the single element airfoil at an angle of attack of 2°.
H. Fatahian, H. Salarian and M.E. Nimvari et al. / Applied Mathematical Modelling 83 (2020) 683–702 697

Fig. 17. Continued

Fig. 18. Comparison of the boundary layer velocity profiles under dry and rainy conditions for the single element airfoil at a chord length of 0.75 for the
airfoil at an angle of attack of 20°.
698 H. Fatahian, H. Salarian and M.E. Nimvari et al. / Applied Mathematical Modelling 83 (2020) 683–702

Fig. 19. Comparison of the boundary layer velocity profiles under dry and rainy conditions for the slatted airfoil at a chord length of 0.75 for the airfoil at
an angle of attack of 20°.

Fig. 20. Comparison of the boundary layer velocity profiles under dry conditions for the single element and slatted airfoils at a chord length of 0.75 for
the airfoil at an angle of attack of 20°.

Figs. 14 and 15 demonstrate that using a leading-edge slat delayed the flow separation at higher angles of attack, and thus
the aerodynamic performance of the airfoil improved. This control method was effective under both dry and rainy condi-
tions.
Another important consideration is the effect of the gap between the main section of the airfoil and the leading-edge
slat. Under rainy conditions, by setting an appropriate distance between the two sections, the airflow could pass through
the gap and the water film layer did not clog the gap. As shown in Fig. 15, the airflow was accelerated over the main section
of the airfoil, which allowed the boundary layer to attach to the airfoil’s surface and the water film was stripped off the
airfoil due to the shear stress. This observation was more evident when considering the water film thickness, as shown in
Fig. 16. Fig. 16 demonstrates that the water film was clearly thicker on the upper surface of the single element airfoil. Using
the leading-edge slat positively decreased the thickness of the water film. All of these effects enhanced the aerodynamic
performance of the airfoil.

6.3. Effects of rain on the boundary layer velocity profile

Fig. 17 shows the velocity profiles for the boundary layer at four different chordwise positions on the upper surface of
the airfoil for an angle of attack of 2°, at which the maximum degradation of the lift-to-drag ratio was obtained due to
H. Fatahian, H. Salarian and M.E. Nimvari et al. / Applied Mathematical Modelling 83 (2020) 683–702 699

Fig. 21. Comparison of the boundary layer velocity profiles under rainy condition for the single element and slatted airfoils at a chord length of 0.75 for
the airfoil at an angle of attack of 20°.

Fig. 22. Comparison of the mass distributions of the water films over the single element airfoil and slatted airfoil under rainy conditions at an angle of
attack of 2°.

rainy conditions. Clearly, when x/c increased, the magnitude of the velocity decreased under both dry and rainy conditions.
Fig. 17 demonstrates that the splashback rain particles decelerated the boundary layer near the leading edge compared
with the boundary layer under dry conditions. The boundary layer appeared to recover downstream [21]. Decelerating the
boundary layer increased the drag and decreased the lift and lift-to-drag ratio, thereby resulting in premature stalling and
flow separation. Moreover, at a chord length of 0.75, the turbulent profiles generated a thicker boundary layer than the
laminar profiles (chord length = 0.25).
The boundary layer velocity profiles at a chord length of 0.75 for the single element and slatted airfoils under dry and
rainy conditions at an angle of attack of 20° are presented in Figs. 18–21. As shown in Fig. 18, a reverse flow occurred
in the boundary layer under both dry and rain conditions, which corresponded to the considerable instability of the flows
that separated from the leading edge of the single element airfoil. Considerable changes in the shape of the boundary layer
velocity profile were observed for the slatted airfoil under dry and rainy conditions (Figs. 19, 20 and 21). The slat decreased
the boundary layer thickness. It should be noted that these significant changes in the shape of the boundary layer profile
improved the aerodynamic performance of the airfoil and delayed the separation at higher angles of attack.
700 H. Fatahian, H. Salarian and M.E. Nimvari et al. / Applied Mathematical Modelling 83 (2020) 683–702

Fig. 23. Comparison of the mass distributions of the water films over the single element airfoil and slatted airfoil under rainy conditions at an angle of
attack of 20°.

Fig. 24. Comparison of the rain particle diameters on the surfaces of the single element airfoil and slatted airfoil under rainy condition at an angle of
attack of 20°.

6.4. Mass distribution of water film and diameter of rain particles on the airfoil’s surface

Figs. 22 and 23 compare the mass distributions of the water film over the airfoil for the single element and slatted airfoils
under rainy conditions at α = 2° and α = 20°. The effect of water film formation on the airfoil surface corresponding to
aerodynamic degradation was considered under heavy stalling conditions. More raindrops adhered to the surface of the
single element airfoil and they formed a thicker film than that on the slatted airfoil at both angles of attack, where the
film effectively roughened the airfoil’s surface. After the raindrops impacted on the leading-edge slat’s surface, they broke
up into very fine water droplets and less water concentrated on the main section of the airfoil. Using the leading-edge slat
reduced the maximum water film mass from 18 g to 2.5 g compared with the single element airfoil at α = 2°, at which
the maximum lift-to-drag ratio degradation was observed. In addition, using the leading-edge slat decreased the maximum
water film mass from 15 g to 1 g compared with the single element airfoil at α = 20°. This observation was more obvious
when the diameter of rain particles on the airfoil’s surface was considered. The diameter of rain particles on the airfoil’s
surface was calculated by simulating heavy rainfall conditions using a wall film model, as shown in Fig. 24. Using the slat
resulted in smaller rain particles adhering to the airfoil chord length compared with the single element airfoil. All of these
positive effects increased the lift coefficient and lift-to-drag ratio but decreased the drag coefficient, and effectively delayed
the flow separation at higher angles of attack.
H. Fatahian, H. Salarian and M.E. Nimvari et al. / Applied Mathematical Modelling 83 (2020) 683–702 701

7. Conclusions

The present study comprehensively investigated the effects of rain on flow separation and the aerodynamic characteris-
tics of single element and slatted NACA 0012 airfoils. A two-way momentum coupled Eulerian–Lagrangian two-phase flow
approach was used to simulate the raindrop trajectories in the airflow field. The results obtained in the present study were
validated based on comparisons with the experimental results reported by Hansman and Craig [34], and the numerical re-
sults given by Wu et al. [12]. The CFD computations yielded satisfactory results and they were in good agreement with the
previously reported experimental and numerical results.
The results demonstrated that rainfall inevitably had negative effects on the aerodynamic coefficients for both the single
element and slatted airfoils. Due to the presence of rain, the lift coefficient decreased by up to 10% whereas the drag coef-
ficient increased by up to 60% at low angles of attack for the single element airfoil. The slatted airfoil produced greater lift
and lower drag compared with the single element airfoil, especially at high angles of attack under both dry and rainy condi-
tions. The presence of the slat significantly increased the stall angle from 13° to 22° under dry conditions and increased the
stall angle from 16° to 24° under rainy condition. The slat decreased the boundary layer thickness and changed the shape
of the boundary layer profile to enhance the aerodynamic performance of the airfoil and the flow separation was delayed at
higher angles of attack. The slat improved the aerodynamic performance of the airfoil but rain induced greater aerodynamic
degradation on the slatted airfoil compared with the slatted airfoil under dry conditions.
The results obtained in the present study may be important for aircraft designers because they provide a better under-
standing of the aerodynamic performance of aircraft, and they could be useful for helping pilots to fly safely under severe
atmospheric conditions with heavy rain.

Declaration of Competing Interest

The authors have no conflicts of interest to declare.

References

[1] M. Ismail, C. Yihua, A. Bakar, Z. Wu, Aerodynamic efficiency study of 2D airfoils and 3D rectangular wing in heavy rain via two-phase flow approach,
Proc. Inst. Mech. Eng., Part G: J. Aerosp. Eng. 228 (7) (2014) 1141–1155.
[2] R.V. Rhode, Some Effects of Rainfall on Flight of Airplanes and on Instrument Indication, National Aeronautics and Space Admin Langley Research
Center Hampton, VA, 1941 No. NACA-TN-803.
[3] J.K. Luers, P.A. Haines, Experimental measurements of rain effects on aircraft aerodynamics, in: AIAA 21st Aerospace Sciences Meeting, 1983.
[4] J. Luers, P. Haines, Heavy rain influence on airplane accidents, J. Aircr. 20 (2) (1983) 187–191.
[5] B.E. Thompson, M.R. Marrochello, Rivulet formation in surface-water flow on an airfoil in rain, AIAA J. 37 (1) (1999) 45–49.
[6] M. Ismail, C. Yihua, A. Bakar, Z. Wu, Aerodynamic efficiency study of 2D airfoils and 3D rectangular wing in heavy rain via two-phase flow approach,
Proc. Inst. Mech. Eng., Part G: J. Aerosp. Eng. 228 (7) (2014) 1141–1155.
[7] Z. Wu, Drop “impact” on an airfoil surface, Adv. Colloid Interface Sci. 256 (2018) 23–47.
[8] Z. Wu, B. Lv, Y. Cao, Heavy rain effects on aircraft lateral/directional stability and control determined from numerical simulation data, Aerosp. Sci.
Technol. 80 (2018) 472–481.
[9] M. Ismail, Z. Wu, A. Bakar, S. Tariq, Aerodynamic characteristics of airfoil cruise landing and high lift configurations in simulated rain environment, J.
Aerosp. Eng. 28 (5) (2014) 04014131.
[10] Z. Wu, Y. Cao, Y. Yang, Direct CFD prediction of dynamic derivatives for a complete transport aircraft in the dry and heavy rain environment, Aeronaut.
J. 122 (1247) (2018) 1–20.
[11] P. Haines, J. Luers, Aerodynamic penalties of heavy rain on landing airplanes, J. Aircr. 20 (2) (1983) 111–119.
[12] Z. Wu, Y. Cao, Numerical simulation of flow over an airfoil in heavy rain via a two-way coupled Eulerian–Lagrangian approach, Int. J. Multiph. Flow
69 (2015) 81–92.
[13] E. Fatahian, A.L. Nichkoohi, H. Fatahian, Numerical study of the effect of suction at a compressible and high Reynolds number flow to control the flow
separation over NACA 2415 airfoil, Progress Comput. Fluid Dyn. Int. J. 19 (3) (2019) 170–179.
[14] E. Fatahian, A.L. Nichkoohi, H. Salarian, J. Khaleghinia, Comparative study of flow separation control using suction and blowing over an airfoil
with/without flap, Sādhanā 44 (11) (2019) 220.
[15] J. Wild, Mach-, Reynolds-and Sweep effects on active flow separation control effectivity on a 2-element airfoil wing, Active Flow and Combustion
Control 2014, Springer, Cham, 2015, pp. 87–100.
[16] A. Farhadi, E.G. Rad, H. Emdad, Aerodynamic multi-parameter optimization of NACA 0012 airfoil using suction/blowing jet technique, Arab. J. Sci. Eng.
42 (5) (2017) 1727–1735.
[17] M.S. Genç, Ü. Kaynak, H. Yapici, Performance of transition model for predicting low Re aerofoil flows without/with single and simultaneous blowing
and suction, Eur. J. Mech.-B/Fluids 30 (2) (2011) 218–235.
[18] M. Tadjfar, E. Asgari, Active flow control of dynamic stall by means of continuous jet flow at Reynolds number of 1× 106 , J. Fluids Eng. 140 (1) (2018)
011107.
[19] K. Yousefi, R. Saleh, P. Zahedi, Numerical study of blowing and suction slot geometry optimization on NACA 0012 airfoil, J. Mech. Sci. Technol. 28 (4)
(2014) 1297–1310.
[20] H. Fatahian, H. Salarian, M. Eshagh Nimvari, E. Fatahian, Numerical study of suction and blowing approaches to control flow over a compressor cascade
in turbulent flow regime, Int. J. Autom. Mech. Eng. 15 (2) (2018).
[21] Y. Cao, Z. Wu, Z. Xu, Effects of rainfall on aircraft aerodynamics, Progress Aerosp. Sci. 71 (2014) 85–127.
[22] Z. Wu, Y. Cao, S. Nie, Y. Yang, Effects of rain on vertical axis wind turbine performance, J. Wind Eng. Ind. Aerodyn. 170 (2017) 128–140.
[23] G.M. Bezos, R.E. Dunham, G.L. Gentry, W.E. Melson, Wind Tunnel Aerodynamic Characteristics of a Transport-type Airfoil in a Simulated Heavy Rain
Environment, NASA, 1992 Technical. Report. TP-3184.
[24] A.H. Markowitz, Raindrop size distribution expressions, J. Appl. Meteorol. 15 (9) (1976) 1029–1031.
[25] Z. Wu, Y. Cao, M. Ismail, Heavy rain effects on aircraft longitudinal stability and control determined from numerical simulation data, Proc. Inst. Mech.
Eng., Part G: J. Aerosp. Eng. 229 (10) (2015) 1824–1842.
[26] M. Cai, E. Abbasi, H. Arastoopour, Analysis of the performance of a wind-turbine airfoil under heavy-rain conditions using a multiphase computational
fluid dynamics approach, Ind. Eng. Chem. Res. 52 (9) (2012) 3266–3275.
702 H. Fatahian, H. Salarian and M.E. Nimvari et al. / Applied Mathematical Modelling 83 (2020) 683–702

[27] A.C. Cohan, H. Arastoopour, Numerical simulation and analysis of the effect of rain and surface property on wind-turbine airfoil performance, Int. J.
Multiph. Flow 81 (2016) 46–53.
[28] H.K. Versteeg, W. Malalasekera, An Introduction to Computational Fluid Dynamics: The Finite Volume Method, Pearson Education, 2007.
[29] Z. Han, Z. Xu, N. Trigui, Spray/wall interaction models for multidimensional engine simulation, Int. J. Engine Res. 1 (1) (20 0 0) 127–146.
[30] S.V. Apte, M. Gorokhovski, P. Moin, LES of atomizing spray with stochastic modeling of secondary breakup, Int. J. Multiph. Flow 29 (9) (2003)
1503–1522.
[31] A.J. Bilanin, Scaling laws for testing airfoils under heavy rainfall, J. Aircr. 24 (1) (1987) 31–37.
[32] S.A.J. Morsi, A.J. Alexander, An investigation of particle trajectories in two-phase flow systems, J. Fluid Mech. 55 (2) (1972) 193–208.
[33] F.R. Amaral, F.H.T. Himeno, C.C. Pagani Jr., M.A.F. Medeiros, Slat noise from an MD30P30N airfoil at extreme angles of attack, AIAA J. 56 (3) (2017)
964–978.
[34] R.J. Hansman, A.P. Craig, Low Reynolds number tests of NACA 64-210, NACA 0012, and Wortmann FX67-K170 airfoils in rain, J. Aircr. 24 (8) (1987)
559–566.

You might also like