Energy 215 (2021) 118915

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

Impacts of Gurney flap and solidity on the aerodynamic performance

of vertical axis wind turbines in array configurations
Lulu Ni a, Weipao Miao a, Chun Li a, b, *, Qingsong Liu a
a
School of Energy and Power Engineering, University of Shanghai for Science and Technology, Shanghai, 200093, China
b
Shanghai Key Laboratory of Multiphase Flow and Heat Transfer in Power Engineering, Shanghai, 200093, China

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

Article history: To improve the overall performance of VAWT farm configurations, the impacts of Gurney flap (GF) and
Received 23 June 2020 the parameter of solidity on the performance of VAWTs in three-turbine arrays are investigated via
Received in revised form computational fluid dynamics (CFD) simulations after a validation of computational model using
15 September 2020
experimental data in the present study. The results indicate that with the tip speed ratio (TSR) of the
Accepted 21 September 2020
Available online 7 October 2020
upstream pairs of VAWTs in array staggered configurations increasing, the flow velocity through the gap
is higher and the upstream pairs of VAWTs achieve higher power output at high TSR compared to an
isolated turbine. Then, the addition of GF on the upstream pairs of VAWTs can achieve higher flow
Keywords:
CFD
velocity through the gap compared to the upstream clean coupled VAWTs, which is beneficial to the
Gurney flap power output of downstream rotor. The downstream turbine in the configurations which consists of
Solidity three GF-VAWTs achieves the best aerodynamic performance at TSR < 3.1, and at TSR ¼ 2.51, this
Aerodynamic performance downstream rotor enhances the maximum average torque by 36.5% and 24% in comparison with the
Array configurations isolated clean turbine and the isolated GF-turbine respectively. Moreover, the downstream VAWT in
VAWT array configurations with each solidity obtains higher power extraction than the isolated counterpart at
overall TSR. At TSR ¼ 2.27, the downstream rotor with a chord length of 123.5 mm achieves the incre-
ment of 23.1% in comparison with isolated counterpart.
© 2020 Elsevier Ltd. All rights reserved.

1. Introduction In contrast, the mutual interactions between VAWTs in array con-

figurations can result in positive effects and improve their perfor-
In recent years, the wind energy, as a type of clean, abundant, mance [8]. Additionally, VAWT via its intrinsic advantages such as
renewable energy source, has been the fastest-growing energy due easy installation and maintain, no yaw system, more appropriate
to the exhausted traditional fossil fuels and the increasing energy combining with buildings and so on, has recently received resur-
demands [1,2]. Wind turbines, which are the most common ma- gent interest from academic communities [9]. However, the power
chines to harvest wind energy, can be mainly categorized into extraction of VAWT is currently lower than that of HAWT due to
vertical axis wind turbines (VAWTs) and horizontal axis wind tur- their complex unsteady aerodynamic characteristics such as dy-
bines (HAWTs) according to the orientation of the rotational axis namic stall and blade wake interactions [10]. Therefore, a great
[3]. In the past, the industry and experts paid main attention to the number of experts conducted numerous researches such as flow
HAWTs for their higher efficiency and suitability to large-scale control techniques and wind farm arrangement on improving the
wind farms [4]. Whereas there are still existing various chal- power coefficient of VAWTs.
lenges with HAWTs currently such as noise pollution, ecological The mechanism of power enhancement of the closely placed
problems, and fatigue problems [5]. Moreover, the mutual in- VAWTs and the optimal layout of VAWT configurations attracted a
teractions between HAWTs in wind farm cause serious damage to great number of experts’ attention. Schatzle [11] investigated the
their performances so that the turbines should be spaced far apart mutual aerodynamic interference between two VAWTs using a
enough to avoid interferences, occupying large areas of land [6,7]. vortex/lifting line model in 1981. In 2004, Thomas [12] firstly
patented that a pair of placed closely VAWTs could achieve higher
aerodynamic efficiency by taking advantage of the interactions
* Corresponding Author between the adjacent turbines. Then, Duraisamy [13] numerically
E-mail address: [email protected] (C. Li).

https://doi.org/10.1016/j.energy.2020.118915
0360-5442/© 2020 Elsevier Ltd. All rights reserved.
L. Ni, W. Miao, C. Li et al. Energy 215 (2021) 118915

is a small plate perpendicular to the chord line of the leading edge

Nomenclature [22]. Even though the impacts of GF on the performance of single
airfoil or isolated VAWT have been extensively investigated by
VAWT Vertical axis wind turbine many experts, few studies examined the effects of this attractive
HAWT Horizontal axis wind turbine device on the VAWT farm configurations. For instance, Liebeck [23]
2D Two-dimensional investigated experimentally the effects of GF on the airfoil, pointing
GF Gurney flap out that the addition of GF could effectively increase the lift force of
CFD Computational fluid dynamics airfoils with a slight reduction of drag force. It was also reported
TSR Tip speed ratio that the height of this device should be remained below 2% of the
l Tip speed ratio chord length (c) of an airfoil to avoid the sharply increase of extra
c Chord length drag force. Later, many pieces of research confirmed the capacity of
q Azimuth angle GF on enhancing the lift force of airfoils [24,25]. The influences of
u Angular velocity GF size attracted many scholars’ attention. Li [26] conducted a wind
V∞ Incoming wind tunnel experiment for NACA 0018 airfoil with different height
n Number of blades Gurney flaps. The results revealed that the maximum lift coefficient
yþ Dimensionless wall distance was augmenting with an increased GF height from 0.5% chord to 3%
Cp Power coefficient chord whereas the maximum lift-to-drag ratio decreased. Li
Cm Torque coefficient conclusively proposed that the height of this device should remain
AOA Angle of attack within the boundary layer. Additionally, the authors of [27,28]
R Rotor radius agreed with this result that the size of GF should keep below the
D Diameter boundary layer to achieve the maximum lift-to-drag ratio. Zhu [29]
s Solidity discussed the performance of VAWT with GF and dimple-GF. The
URANS Unsteady Reynolds-averaged Navier Stokes numerical results revealed that the addition of GF and dimple-GF
could obtain a higher power coefficient at overall TSR compared
to the clean turbine. Bianchini [30] carried out a systematically
numerical analysis to study the effect of GF with different heights
investigated the performance of a pair of side-by-side VAWTs and and locations on the performance of airfoil static polars, a single
reported that both counter-rotating pairs and co-rotating pairs rotating airfoil VAWT and a three-blade turbine. It was observed
could obtain higher power output compared to an isolated rotor. that the GF could enhance significantly the aerodynamic perfor-
Later, the authors of [14,15] observed similar performance mance of airfoil, especially at lower TSR. The GF with 2% c of height
augmentation by investigating experimentally twin VAWTs. when installed on the inner side of airfoil achieved the highest
Brownstein [16] investigated the performances of downstream power extraction. Furthermore, more and more scholars designed
VAWTs positioned offset from the wake of upstream turbines via and studied novel GFs. Zhang [31] pointed out a novel plasma
laboratory and field measurements. The results showed that the Gurney flap with a plasma actuator and carried out simulations to
downstream rotors were observed an increased power output. investigate the effect of this novel device on NACA 0012 airfoil. Feng
However, the key flow mechanisms responsible for the improved [32] developed a virtual Gurney flap which consisted of dielectric
performance of twin VAWTs were not obvious. Then Zanforlin [17] barrier discharge plasma actuator. The research result indicated
studied the key flow mechanisms via numerical simulations. The that this device could effectively enhance the lift coefficient. Lee
results revealed that the change of flow direction influenced by the [33] designed a matrix-permutation perforated Gurney flap.
adjacent rotor and the contraction of the wake contributed to the Solidity is one important geometrical parameter of VAWTs [34].
enhancement power output. Moreover, the author pointed out that The effects of the parameter on the aerodynamic performance of
the overall performance of staggered pairs was a litter worse than isolated VAWT have been conducted by many experts [35e37]. For
that of twin side-by-side rotors. Later, Peng [18] conducted nu- instance, Li [38] studied the aerodynamic performance of straight-
merical simulations on optimizing the power performance of twin bladed VAWT with four different solidities corresponding to the
VAWTs by changing the parameters of the airfoil section, solidity different numbers of blades through wind tunnel experiments. The
ratio, pitch angle, rotational direction and turbine spacing. Posa [19] result indicated that the pressure difference sharply decreased as
investigated the wake features of coupled VAWTs by means of the solidity increased. Then, Lam [39] conducted a series of sys-
Large Eddy Simulation. It was reported that the blockage effects tematically wind tunnel tests to investigate the effect of solidity
between the coupled rotors rather than the vortex interaction ratios with four different turbulence intensities on the power
played an important role in increasing the momentum flux of the extraction of VAWTs. The results revealed that the higher solidity
gap as well as the downstream locations. Ahmadi-Baloutaki [20] turbine achieved more power output irrespective of flow regimes.
conducted a series of wind tunnel measurements to study the While because of turbulent flows, the medium solidity turbine is
aerodynamic performance of VAWTs in various array configura- more suitable than the low and high one. Later, Rezaeiha [40] car-
tions. The results indicated that a pair of counter-rotating VAWTs ried out a great number of simulations to analyzes the impact of the
achieved a slight enhancement in the power coefficient compared number of blades as well as solidity changed via blade chord length
to the isolated one. Furthermore, the turbine located downstream on the aerodynamic performance of VAWTs. It was observed that
of pairs of rotors obtained better aerodynamic performance and the the turbine with relatively-low solidity is optimal for moderate to
optimum arrangement distance of the three-rotor configuration high TSR corresponding to low wind speed urban environment.
was presented. Later, Andrew [21] studied the optimal layout of Roh [41] performed numerical research on the performance of
VAWT farm configurations via CFD simulations. The results indi- wind turbines with various solidities in the range of 0.5e0.8.
cated that even though all of the proposed wind farm designs Okeoghene [42] discussed the performance and flow field of VAWT
achieved a better power coefficient, the staggered configuration with two different solidities by changing the rotor chord length via
presented the highest array power output. experimental tests. The result showed that the turbine with higher
Gurney flap is one of the most effective passive flow control chord length performed a better overall power coefficient and the
devices to enhance the aerodynamic performance of airfoil, which stepper gradient curve of Cp-TSR. Taher [43] performed numerical
2
L. Ni, W. Miao, C. Li et al. Energy 215 (2021) 118915

simulations to investigate the influences of the number of the study is adopted from the literature [46]. The specific geometric
blades and blade chord length on the performance of straight model of clean-VAWT is illustrated in Fig. 1, where the q is the az-
Darrieus wind turbine. The results indicated that the power coef- imuth angle and the u is rotating angular velocity. The number of
ficient of wind turbine was increasing as the blade chord length blades (n) is three and the upper blade calls blade 1 corresponding
increased within a certain range after which the power output to the azimuth angle of 0 . The blade cross-section is the sym-
dramatically decreased. Zhu [44] and Sagharichi [45] numerical metrical NACA 0021 airfoil section with chord length of 0.0858 m.
investigated the effect of solidity on the performance of VAWT with Fig. 2a and 2b illustrate the geometric models of clean-airfoil and
a passive flow control device of Gurney flap as well as active flow airfoil with Gurney flap. The geometric models and mesh genera-
control device of variable pitch respectively. tions are performed by the software ANSYS® ICEM CFD™. Noted
As briefly reviewed, the vast majority of recent works were that the parameters of the Gurney flap refer to the literature [29].
aimed at either the performance of the clean VAWTs without any The rotor radius (R) is 0.515 m and the diameter of shaft is 1/6 of c.
flow control devices in array configurations, the isolated turbines The geometric parameters of the clean VAWT are collected in
with the passive flow control device of Gurney flap, or the isolated Table 1.
turbine with various solidities. However, few studies have studied The tip speed ratio can be used to reflect the operation condition
the effect of Gurney flap as well as the important parameter of of turbines, defined as:
solidity on the performance of VAWT farm configurations. There-
fore in this present study, to maximize the overall power output of l ¼ Ru=V∞ (1)
the VAWT farm configurations, the impacts of GF and solidity on
the performance and flow field of VAWT configurations are inves- where u is the angular velocity, V∞ is the incoming wind.
tigated respectively through comprehensive 2D CFD simulations. Solidity is a significant dimensionless parameter of VAWT,
Moreover, the effect of the TSR on the flow filed of configurations is calculated as
analyzed. The influence of GF on the isolated turbine and the per-
d ¼ Nc=2R (2)
formance of isolated turbines with different solidities are evaluated
and discussed. The torque coefficient Cm and power coefficient Cp are the most
important parameters to demonstrate the aerodynamic perfor-
mance of VAWT, defined as:
2. Computational methodology
. 
2.1. Geometries of VAWT Cm ¼ 2M rARV∞
2
(3)

. 
The straight-bladed VAWT defined clean-VAWT in the present
Cp ¼ 2P rAV∞
3
(4)

where M is the torque, N/m; r is the air density, kg/m3; A is rotor

swept area, m2; P is the power output, W.

2.2. Computational domain and grids

The schematic of the computational domain and boundary

conditions of the isolated clean-VAWT are shown in Fig. 3. The type

Table 1
Geometric parameters of VAWT.

Parameter Value

Airfoil profile NACA 0021

Number of blades [n] 3
Solidity [s] 0.25
Chord length [c] 85.8 mm
Radius [R] 515 mm
Diameter [D] 1030 mm
Wingspan [H] 1456.4 mm
Fig. 1. The aerodynamic model of clean VAWT.

Fig. 2. The geometrical models of airfoils.

3
L. Ni, W. Miao, C. Li et al. Energy 215 (2021) 118915

three diameters respectively. Furthermore, the computational

domain and region separations of the isolated turbine are also
applied to the wind turbine configurations.
Fig. 5 illustrates the mesh distributions of the isolated clean
turbine and the enlarged details around blades. It is noted that the
number of grids has been verified by the following section. Regions
G1, G3 and G4 are all composed by structured girds while region G3
is covered by unstructured grids. The girds around blades and shaft
surfaces are more refined in order to simulate the complex flow
phenomenon. Moreover, the height of the first-layer girds on
blades is set at 3.173  105 m to ensure the yþz1, so that the
laminar sub-layer could be resolved accurately. The growth rate of
mesh size around blades and shaft is set at 1e1.15 and the sizes of
girds on the interface of different adjacent regions remain
approximately consistent to make the simulations accuracy and fast
convergence. The mesh distribution of turbine configurations and
enlarged mesh details are illustrated in Figs. 6 and 7, respectively.
Fig. 3. Schematic of the computational domain of the isolated clean VAWT.
These mesh settings are also applied to the turbine configurations
and the turbine with GF.
of C is adopted for the topology structure of the computational
_
domain where the semicircle radius of AD is 80 R meanwhile the 2.3. Boundary condition and numerical algorithm
distances of AB and BC are 150 R and 160 R respectively. The size of
the computational domain in the present study is enough to avoid _
As shown in Fig. 3, the boundaries of AD, AB and DC are defined
the influences from the boundary of inlet and outlet, according to
as velocity inlet. The boundary of BC is set at pressure outlet with
the previous literatures of [29,47].
gauge pressure of 0 Pa. The velocity of incoming wind (V∞) is 9 m/s
The computational domain is partitioned into the outer sta-
with turbulent intensity of 1%. The turbulent viscosity ratio is 1 and
tionary region (G1) and rotating regions (i.e., G2, G3 and G4) as
the density of air is 1.225 kg/m3.
shown in Fig. 3. The stationary and rotating regions communicate
Incompressible unsteady Reynolds-averaged Navier Stokes
with each other by means of a sliding interface condition. In
(URANS) are performed in the present study by means of the CFD
addition, the rotating regions are divided into sub-regions G2, G3
software FLUENT®. The SIMPLEC scheme with second-order up-
and G4 in order to refine the mesh around the blade surfaces and
wind scheme is used for pressure-velocity coupling. The time
shaft. The regions G2, G3 and G4 rotate around the rotational axis
formulation is performed by second-order implicit scheme. The
with the identical angular velocity. And the regions G2, G3 and G4
aerodynamic parameters of VAWTs are acquired during a period of
are associated by the interface. The surfaces of the blades and shaft
stable rotation with a time step of 0.5 . Every time step is iterated
are set as no-slip walls. As regards the VAWT configurations, the
arrangement is illustrated in Fig. 4. Noted that the layout refers to
Ref. [20], where the authors reported that the wind turbine con-
figurations could obtain the maximum total power output when
the distance between the two upstream counter-rotating turbines
and spacing of the downstream turbine from the pair were one and

Fig. 4. Layout of VAWTs in array configurations. Fig. 5. Computational mesh distributions.

4
L. Ni, W. Miao, C. Li et al. Energy 215 (2021) 118915

dimensional (2D) computational simulations. It was reported that

SST model variants including SST k-u, SSTI and TSST performed
more reasonable agreement with experimental data and predicted
more accurately the power coefficient with an extensive range TSR.
Furthermore, the four-equation TSST turbulence model could pre-
dict more accurately the laminar-to-turbulent transition compared
with other two-equation SST models due to its ability to capture the
transition from laminar flow to turbulent flow on the blade surface.
In the present study, the TSST model is adopted to simulate all the
cases.

2.4. Independence study and validation

To demonstrate the reliability of CFD calculations adopted in the

present study, the time step, computational domain and number of
mesh nodes are verified. A gird sensitivity analysis is firstly inves-
tigated under the condition that the incoming wind is 9 m/s and the
TSR is 2.62. Based on the clean isolated VAWT, four types of mesh
number, which are 160 thousand, 333 thousand, 437 thousand and
662 thousand respectively, are conducted by changing the nodes
around the blades surfaces and every regions simultaneously. The
Fig. 6. Mesh distributions of VAWTs in array configurations. variation of instantaneous torque coefficient of single blade versus
azimuth angle is shown in Fig. 8.
In Fig. 8, it is apparent that the curves of 160 thousand and 333
thousand are largely different with the finest grids in a revolution.
Then, the curve of 437 thousand mostly overlaps with that of 662
thousand and the average torque coefficient of two types of grids in
a revolution is also almost identical. So that in view of both reli-
ability of the simulations and the computational resource limita-
tion, the type of 437 thousand grids is selected in the present study.
All of the grids adopted in the following cases are based on the
verified mesh number.
To demonstrate the reliability of the size of the computational
domain in the present study and ensure the complete development
of wake flow of turbine, the variation of streamwise flow velocity
along the y-axis direction at the point of x ¼ 140 R is illustrated in
Fig. 9. According to the analysis in Ref. [47], the turbine wake length
was decreasing with increasing the value of TSR. Thus, the vorticity
profile of the whole flow filed at TSR ¼ 1.62 is shown in Fig. 10.
In Fig. 9, the flow velocity along the y-axis direction at the point
of x ¼ 140 R is almost identical with the incoming wind of 9 m/s,
which indicates that the turbine wake has been terminated early
and the distance of outlet is appropriate. From Fig. 10, it is observed
Fig. 7. Mesh distributions of the airfoil with GF.
that the vorticity magnitude at inflow and outflow boundaries is 0.

by 25 times. The 16th stable revolution is adopted to analyze the

simulation results. The Least Squares Cell-Based is selected for
computing the gradients of spatial discretization. Furthermore, the
detailed information is collected in Table 2.
Turbulence modeling is significant in simulation [48]. Abdolra-
him [49] investigated previously various turbulence models to
select the most accurate model for Reynolds-averaged Navier
Stokes (URANS) simulations of VAWTs by the means of two-

Table 2
Details of the under-relaxation factors.

Factors Value

Pressure 0.2
Density 1
Body forces 1
Momentum 0.4
Specific dissipation rate 0.5
Intermittency 0.5
Turbulent Viscosity 0.5
Fig. 8. Grid independence verification.

5
L. Ni, W. Miao, C. Li et al. Energy 215 (2021) 118915

Fig. 11. Instantaneous torque coefficient using various azimuthal increments.

Fig. 9. Variation of flow velocity at x ¼ 140 R with different tip speed ratios.

Fig. 10. Vorticity contour profiles of the whole flow filed.

Fig. 12. Power coefficient comparison between computation results and experiment
values.

Thus, the size of the computational domain in the present study is

enough to avoid the influences from the boundaries of inlet, outlet,
top and bottom. model adopted in this paper could predict more accurately in
Fig. 11 reportes the instantaneous torque coefficient Cm of single laminar-to-turbulent transition on the blade surface [48]. In addi-
blade for the last revolution versus azimuth angle at TSR ¼ 2.51 tion, the aerodynamic parameters of VAWTs in present study are
which corresponds to relatively moderate tip speed ratio using acquired during a period of stable rotation with a time step of 0.5 .
different azimuthal increments. As reported by Ref. [47], the CFD results with a time step of 1 is
In Fig. 11, the time step size equivalent to the time required for a significantly higher than that with a time step of 0.5 . Despite the
0.5 blade rotation is chosen in order to consider the accuracy of values of Cp in the present study are slightly higher than the
simulations and limitation of computing resource. experimental data, since the CFD model includes neither the spokes
Then, in order to demonstrate the reliability and accuracy of the nor the mechanical friction loss, it is significantly accurate and
turbulence model and other numerical setups, the comparison reliable to simulate the following cases in the present study. Noted
between the simulation results and experimental data producted that the peak value of the experimental curve shows a large span
by the same rotor from the literature [46] is assessed in Fig. 12. with the next point so that its actual peak value may occur between
In Fig. 12, it is observed that the tendency of simulation results in them. Thus, the present study increases two values of TSR by 2.72
the present study, especially at low TSR, is mostly identical with and 2.82 to precisely demonstrate the curve power coefficient of
that of experimental data. When compared with the numerical this clean-VAWT. It is observed that the optimal tip speed ratio
results of Ref. [46], the results of the present simulation show more occurs at 2.72. Then the value of power coefficient of the turbine is
coincident with experimental data so that the turbulence model decreasing.
and numerical setting adopted by this study are more appropriate To validate the accuracy and reliability of the CFD model when
for simulating VAWT. Explained that in comparsion with the SST k- simulating the airfoil with GF, the simulations of arifoil NACA 0021
u turbulence model and enhanced wall treatment k-ε Realizable with GF heights of 2%c are performed. And the present results are
model adopted by Ref. [46], the four-equation TSST turbulence compared with the numerical data of the same airfoil obtained by

6
L. Ni, W. Miao, C. Li et al. Energy 215 (2021) 118915

Fig. 13. Schematic of the computational domain of the airfoil.

Fig. 15. Variation of flow velocity at x ¼ 0.55 R with different tip speed ratios.

distance of the radius of turbine.

As shown in Fig. 15, the streamwise flow velocity in the turbine
configurations can be divided into three regions: the flow acceler-
ating area through the gap between the upstream side-by-side
turbines, the low-velocity area below the wind turbine and the
lateral velocity recovery area away from the gap. It is observed that
the flow velocity through the gap is higher than the incoming wind
at all TSR. Furthermore, the value of flow velocity becomes higher
with the TSR increasing. Explained that the accelerating phenom-
enon between the coupled upstream turbines is essentially attrib-
uted to the blockage effects, making the flow velocity and kinetic
energy higher than the incoming wind which is beneficial for the
performance of upstream coupled machines and particularly
Fig. 14. The comparison between persent results and reference values.
downstream ones. Moreover, with the value of TSR increasing, the
blockage effect is stronger so that the accelerating phenomenon
Ref. [30]. Fig. 13 shows the schematic of the computational domain through the gap is more obvious, since that the dynamic solidity
(defined by Daniel [50] as dD ¼ 1  2pdl 1 , representing the combi-
and boundary conditions of the airfoil with GF. And the simulation
results are reported in Fig. 14. nation of geometric solidity and rotation rate) of one turbine be-
In Fig. 14, the tendency of simulation results, especially at low comes larger along with the higher turbine rotational speeds.
AOA, is mostly identical with that of Ref. [30]. A slight deviation is Additionally, at the location of y > 4.5 R, the flow velocity begins to
observed at high AOA which corresponds to the stall angle of attack. recover to the speed of the incoming wind, meaning that the in-
Most of all, the CFD models are able to correctly simulate the airfoil fluence of the wake of turbine along with the x-axis direction starts
with Gurney flap. Thus, the mesh distribution, boundary condition to disappear.
and numerical algorithm adopted by the present paper to simulate To better appreciate the effect of TSR of upstream turbines on
the VAWT with Gurney flap are reliable. the flow field, the velocity contour of the array configurations at
different TSR is shown in Fig. 16.
From Fig. 16, it is observed that the flow velocity through the gap
3. Result and discussion is higher than the incoming wind of 9 m/s. And the phenomenon of
flow acceleration through the gap is more obvious as the TSR of
3.1. The effect of TSR of upstream turbines on flow field upstream pairs increases meanwhile the flow velocity below the
rotor is lower, which is in correspondence with Fig. 15. Moreover,
The TSR of the upstream coupled VAWTs in array staggered the blockage effects between the upstream pairs and consequent
configurations has a significant influence on the whole flow filed as flow acceleration are responsible for the power extraction
well as the downstream turbines. As reported by Posa [19], the augmentation of the coupled VAWTs as revealed by Ref. [19],
blockage effects between coupled VAWTs were more obvious at particularly the performance improvement of the downstream
higher TSR. Then, we focus on the effects of various values of TSR on rotor. This assessment will be confirmed at the following sections.
the flow filed of turbine configurations. In this simulation, the TSR
of upstream pairs is changed from 1.42 to 3.28 in the condition of
remaining the rotating velocity of the downstream turbine fixed to 3.2. The performance of upstream coupled turbines in
avoid other interferences. Fig. 15 illustrates the variation of configurations
streamwise flow velocity fields at different TSR along the y-axis
direction at the point of x ¼ 0.55 R which is approximately the To assess the performance of the side-by-side upstream coupled
7
L. Ni, W. Miao, C. Li et al. Energy 215 (2021) 118915

Fig. 16. The wake velocity contours at different tip speed ratio.

turbines in staggered array configurations, their power coefficients

versus TSR are compared with that of the isolated turbine in the
condition that the TSR of the downstream isolated turbine is fixed
to avoid other interferences. The result is reported in Fig. 17.
In Fig. 17, the average power extraction of upstream pairs is
almost identical with that of the isolated turbine at TSR < 2.72.
While at TSR  2.82, with the TSR increasing, the performance of
each upstream turbine, especially VAWT1, is apparently higher than
that of the isolated one. Besides, at TSR ¼ 3.1, the VAWT1 obtains
the highest power coefficient of 0.388 among three curves and
achieves enhancement of 8.3% compared to the isolated VAWT.
Thus it is concluded that at higher TSR corresponding to the higher
flow velocity of this gap, each of the upstream pairs can obtain
higher performance than the isolated one. One reasonable mech-
anism of this improvement could contribute to the blockage effect
which significantly increases the flow velocity and kinetic energy
[19]. Another is due to the change of lateral velocity through the gap
influenced by the presence of the adjacent turbine which could
favorably generate more lift and torque [17]. Note that the
Fig. 17. Variation of power coefficient versus tip speed ratios.

8
L. Ni, W. Miao, C. Li et al. Energy 215 (2021) 118915

aerodynamic performance of airfoil. In this section, the effect of GF

on the isolated turbine is investigated. Fig. 18 illustrates the com-
parison of the power coefficient between clean VAWT and VAWT
with GF versus TSR.
In Fig. 18, compared to clean wind turbines, the VAWT with GF
improves significantly the wind power extraction at TSR < 2.62.
Particularly at low TSR corresponding to the large variation of angle
of attack (AOA), the enhancement of GF is more obvious. It reveals
that the addition of GF can effectively suppress the flow separation
which easily occurs at low TSR and the aerodynamic performance
and stall characteristics of the airfoil are significantly improved.
Then, at TSR ¼ 2.51, the turbine with GF achieves the maximum
power coefficient of 0.382, with an enhancement of 10.1%
compared to the clean turbine (0.347) at the same TSR. However,
the GF loses effectiveness even leads to a slight decline in the power
output at higher TSR. Bianchini [30] reported the similar phe-
nomenon. At 2.62 TSR 2.62  TSR  3.1, the performance of the
GF-turbine is similar to the clean turbine. Then at TSR ¼ 3.28, the GF
Fig. 18. Variation of power coefficient of VAWT with GF versus tip speed ratios. turbine provides a slight worsening of the power extraction. To in-
depth identify the mechanisms of GF, the instantaneous torque
coefficient trends of single blade of two types of VAWTs versus
distinction of power coefficient between VAWT1 and VAWT2 is azimuth angle are assessed, and the curves are shown in Fig. 19.
resulted from the rotating direction of the downstream turbine. In Fig. 19, the addition of GF installed on the outer side of the
blade (seen Fig. 2) achieves a higher torque coefficient in the up-
3.3. The effect of GF on isolated VAWT stream region in comparison with the clean turbine while leads to a
decline of the torque in the downstream region at all TSR. At
Gurney flap is an effective simple device that can increase the TSR ¼ 1.62 and 2.35, where the variation of angle of attack is wide

Fig. 19. The variation of instantaneous torque coefficient of VAWT with GF at different TSR.

9
L. Ni, W. Miao, C. Li et al. Energy 215 (2021) 118915

Fig. 20. The flow structure of streamline and pressure contour at different azimuth angle.

and the phenomenon of flow separation is serious, the addition of Similarly, the output decline in the downstream region is calculated
GF can effectively suppress this separation phenomenon and in- by the average of torque coefficient of the azimuth angle of
crease the torque of the upstream region with an increment of 139% 180 e360 . Thus, the aerodynamic performance over the whole
and 55.8% respectively compared to the clean one. Whereas the revolution is calculated by the average of torque coefficient of the
blade rotating in the downstream region, the pressure surface and azimuth angle of 0 e360 .
suction surface of the blade are interchanged, making the location Moving forward to TSR ¼ 2.72 and 3.1, where the variations of
of GF changed from pressure side to the suction side. So that the GF the AOA are mostly below the static stall angle and the flow is
is ineffective even brings the augmentation of drag force, leading to nearly attached on the blade surface, the effect of GF is no longer so
a sharply drop of torque at TSR ¼ 1.62 and 2.35. In conclusion, over much obvious. Even if the GF increases the torque coefficient by
the whole revolution, the GF still improves the aerodynamic per- 22.2% and 15.7% respectively within upwind half, it results in a
formance of VAWT at TSR ¼ 1.62 and 2.35, increasing the torque decline of the torque over the whole revolution with a drop of 1.6%
coefficient by 119.4% and 26% respectively. Noted that the perfor- and 4.1% respectively. This is because the GF brings largely extra
mance improvement in the upstream region is calculated by the augmentation of drag force within downwind half so that the tor-
average of torque coefficient of the azimuth angle of 0 e180 . que coefficient is sharply decreased by 91.1% and 177.3%
10
L. Ni, W. Miao, C. Li et al. Energy 215 (2021) 118915

Table 3 flow velocity in which VAWT3 bears is higher than the incoming
The types of configurations of VAWTs. wind of 9 m/s due to the blockage effects through the gap. While
Array VAWT1 VAWT2 VAWT3 the flow filed of wakes of upstream turbines is too complex to
1 Clean-VAWT Clean-VAWT Clean-VAWT
calculate the actual velocity VAWT3 bearing, as seen in Fig. 16.
2 GF-VAWT GF-VAWT Clean-VAWT Therefore, the curves of average torque within one stable revolu-
3 GF-VAWT GF-VAWT GF-VAWT tion of different VAWTs are compared to avoid the overestimation
of the power coefficient of VAWT3. Additionally, the TSR of VAWT3
is still calculated by the value of 9 m/s to promise the comparability
respectively. It can be reported from Fig. 18 that the GF can improve of downstream turbines in different groups.
the wind power extraction of the blade at low-moderate TSR, while From a perusal of Fig. 21, it is apparent from the comparison of
at high TSR, the GF leads to a sharply enhancement of extra drag VAWT3 in Array 1 and the isolated clean turbine that the down-
force resulting in a litter worse performance compared to clean stream turbine can achieve significantly higher power extraction at
one. all TSR. Thus, the phenomenon of flow acceleration through the gap
Fig. 20 shows the flow structure of streamline and pressure of upstream pairs is largely beneficial for the performance of the
contour for the clean VAWT and VAWT with GF at different azimuth downstream turbine. Moreover, the improvement is more obvious
angle. at high TSR. At TSR ¼ 2.72, the downstream turbine obtains the
In Fig. 20, when the azimuth angle is 0 , the pressure distribu- maximum average torque, with an enhancement of 20.3%
tion of the blades for the clean VAWT and VAWT with GF is the compared with the clean isolated one.
same. When at 60 of azimuth angle, the pressure value of blade To evaluate the effect of GF on the flow filed of wind farm and
pressure surface of the VAWT with GF is higher than that of clean downstream turbine, the curves of Array 1 and Array 2 from Fig. 21
VAWT which leading to the pressure difference on both sides and are compared. It is clearly observed that the VAWT3 in Array 2 leads
tangential force of the blade for VAWT with GF becomes larger. to a slight improvement at TSR < 2.35. And at 2.35 TSR 2.72, two
When the azimuth angle changes from 60 to 180 , the separation curves mostly coincide, meaning that the GF can slightly improve
vortex occurs at the suction surface of trailing edge of clean VAWT. the performance of downstream turbuine. Then at TSR3.1, the
While the presence of GF improves the pressure distribution and downstream turbine in Array 2 achieves higher torque than that in
thus the separation vortex and separation regions are shrunk. Array 1, with 13.5% enhancement of average torque at l ¼ 3.1 and
Therefore, it is concluded that the GF has a better effect of sup- 3.2% at l ¼ 3.28 respectively. It reveals that the GF installed on the
pression of separation vortex and improves the performance of upstream pairs has an obvious effect on improvement of the power
blade. output of downstream turbine at high TSR. Noted that the detailed
mechanisms of this phenomenon are reported in Fig. 22.
3.4. The effect of GF on VAWT configurations As regards the VAWT3 in Array 3, the turbine can take advantage
of blockage effects and power augmentation effect of GF, obtaining
In this section, the performance of the downstream turbine ar- the highest aerodynamic performance among all curves at l  2.82.
ranged in array configurations and the effect of GF on the flow filed Besides, at l ¼ 2.51, VAWT3 obtains the peak value of torque, with
of the wind farm are assessed. The layout of VAWT configurations is augmentation of 36.5% and 24% in comparison with the isolated
collected in Table 3. Then the simulations aimed at Table 3 are clean turbine and the isolated GF-turbine respectively. At l > 2.82,
carried out. Noted that in order to keep the maximum power the effect of GF is not so obvious that the performance of VAWT3 in
extraction of the whole wind farm as well as avoid the variation of Array 3 is a litter lower than that in Array 2. Additionally, the torque
upstream blockage effects, the TSR of upstream pairs remains the value of VAWT3 in Array 3 is similar to that in Array 1, since that the
optimal value fixed by 2.62, then the rotational speed of down- GF installed on upstream pairs is beneficial for the downstream
stream turbine is varied. turbine.
Fig. 21 illustrates the average torque within one stable revolu- To further understand the effect of GF on the flow fields in
tion of the whole turbine versus TSR. It is worth explaining that the configurations, Fig. 22 provides the velocity contour of flow fileds of
Array 1 and Array 2.
As illustrated in Fig. 22, the flow velocity through the gap be-
tween the upstream coupled turbines in Array 2 is obviously higher
than that of Array 1 at all TSR. Therefore, the acceleration effect of
upstream pairs with GF is better than that of the clean turbines.
This phenomenon leads to an increment of the flow velocity in
which the downstream turbine bearing so that the power output of
this rotor is increased, which is corresponding to the results of
Fig. 21. Additionally, the wake velocity of upstream pairs in Array 2
is much lower than that of Array 1, revealing that the installation of
GF makes the turbine capture more wind power.
Fig. 23 reports the instantaneous tangential force of the single
blade of different turbines versus azimuth angle at various TSR.
In Fig. 23, it is found that the VAWT3 in Array 3 achieve signif-
icant improvement of tangential force compared with other tur-
bines and the maximum value is higher than 14 N at all TSR, which
confirms the above conclusions. Furthermore, at TAR ¼ 3.1 and
3.28, the tangential force of the turbine in Array 2 is higher than
that of Array 1 in the upwind region, corresponding to the opti-
mizing effect of GF on the downstream turbine at relative high TSR.

Fig. 21. Variation of power coefficient of VAWTs in array configurations versus TSR.

11
L. Ni, W. Miao, C. Li et al. Energy 215 (2021) 118915

Fig. 22. The wake velocity contours at different tip speed ratio.

Fig. 23. Instantaneous tangential force of one blade in a revolution.

12
L. Ni, W. Miao, C. Li et al. Energy 215 (2021) 118915

Table 4 paired upstream turbines at optimal TSR to maintain the maximum

Solidity parameters of VAWTs. power output of the whole wind farm. Fig. 25 shows the variation of
Number of blade Chord length/mm Solidity average torque within a revolution of the turbines at different so-
3 85.8 0.25
lidities versus TSR. The reasons why choose the value of torque to
3 68.64 0.2 compare have been mentioned.
3 102.96 0.33 In Fig. 25, it is apparent that the VAWT3 with each solidity ob-
3 123.5 0.36 tains higher power extraction than the isolated counterpart at
4 85.8 0.33
overall TSR. It means that the upstream pairs of VAWTs with each
5 85.8 0.41
solidity can completely products the acceleration effect which is
beneficial for the performance of the downstream turbine.
In Fig. 25a, at TSR ¼ 1.42 and 1.62, the VAWT3 in array config-
3.5. The effect of solidity on VAWT configurations
urations obtains significant power output augmentation compared
to the isolated rotor, indicating that the self-starting characteristic
In this section, the influence of solidity on the performance of
of turbine arranged in array configurations is sharply improved.
isolated VAWT is investigated. It is worth explaining that the value
Furthermore, at higher TSR, the increment of average torque of
of solidity is varied from 0.2 to 0.41 by changing the blade chord
VAWT3 with comparison in the isolated rotor is also considerable.
length or the blade numbers. The detailed informations of wind
In Fig. 25b, the optimal TSR of VAWT3 is a little higher than the
turbines at different solidities are collected in Table 4. Fig. 24a il-
isolated rotor. At TSR ¼ 2.62, the VAWT3 obtains the highest torque,
lustrates the curves of power coefficient of turbines with various
with an increment of 25.3% compared to an isolated rotor. In
chord lengths of blades by changing when the blade number keeps
Fig. 25c, even though the performance of turbine arranged in array
three and Fig. 24b shows the power coefficient of turbines with
configurations is much better than the isolate counterpart, it is
various blade numbers when the chord length keeps at 85.8 mm.
worse than that of turbines at other solidities. Noted that this tur-
Form Fig. 24, it is apparent that the optimal TSR of different
bine performs better at relatively low TSR so that it is more suitable
turbines is lower with the value of solidity increasing which cor-
for operating at condition of low TSR. Then, in Fig. 25d, the optimal
responds to the revelations of Ref. [40], meaning that the turbine
TSR of VAWT3 is also changed to location of a litter high TSR. At
with high solidity bears lower cumulative load fatigue damage and
TSR ¼ 2.51, the VAWT3 obtains the maximum average torque of
products lower operation noise when achieving the maximum
5.04 Nm, with improvement of 25.4% compared to isolated rotor
power output. Furthermore, the self-starting characteristic of the
(4.02 Nm). Besides, the phenomenon of power extraction
turbine is also improved as the value of solidity increases. While at
improvement is more obvious at relatively moderately-high TSR.
relatively high TSR, the performance of high solidity turbine is
Finally, Fig. 25e still shows that the optimal TSR of VAWT3 is also
worse. Therefore, it is concluded that the high-solidity turbines
changed to location of a litter high TSR. At TSR ¼ 2.27, VAWT3
achieve higher power output at relatively low TSR while the low-
achieves the maximum average torque of 5.35 Nm, with incre-
solidity rotors are more suitable to relatively moderately-high
ment of 23.1% with comparision to isolated rotor (4.35 Nm). The
TSR. Additionally, the peak value of the power coefficient of tur-
improvement is more obvious at relatively moderately TSR.
bines with three blades is higher with chord length increasing.
While when the blade number increases from 3 to 5, the maximum
4. Conclusion
power output is worse. This implies that the performance of VAWT
can be enhanced by appropriately increasing the chord length of
In the present study, the impacts of Gurney flap and solidity on
blade, and the augmentation of blade worsens the performance
the performance of VAWTs in array configurations are investigated.
since more intense interaction among blades occurs.
The power coefficient, average torque of the whole turbine within a
To understand the effect of solidity in VAWT array configura-
revolution, the tangential force of the single blade are analyzed. The
tions, the aerodynamic performances of turbines at different so-
main conclusions in this study are summarized as follows:
lidities arranged in configurations are investigated. The parameters
of Table 4 are adopted for different solidity turbines and the layout
1) The TSR of the upstream pairs of VAWTs in array staggered
is illustrated in Fig. 4. Additionally, the simulations are conducted
configurations has a significant influence on the whole flow
by changing the TSR of the downstream turbine and remaining the
filed. The flow velocity through the gap between the adjacent

Fig. 24. Variation of power coefficient of VAWTs versus TSR.

13
L. Ni, W. Miao, C. Li et al. Energy 215 (2021) 118915

Fig. 25. Variation of average torque of turbine at different solidities versus TSR.

side-by-side turbines is higher than the incoming wind at all VAWTs, which is beneficial for the downstream rotor. Besides,
TSR. Furthermore, when the TSR increases, the accelerating the VAWT3 in array 3 which consists of three GF-VAWTs ach-
phenomenon through the gap is more obvious. ieves the best aerodynamic performance at TSR < 3.1. At l ¼ 2.51,
2) The upstream pairs of VAWTs achieve higher power output at VAWT3 enhances the maximum average torque by 36.5% and
high TSR compared to isolated turbine since that the upstream 24% in comparison with the isolated clean turbine and the iso-
turbines can benefit from the blockage effects. Additionally at lated GF-turbine respectively.
TSR ¼ 3.1, the VAWT1 obtains the highest power coefficient of 5) The isolated VAWT achieves the lower optimal TSR and better
0.388, bringing enhancement of 8.3% compared to the isolated self-starting characteristics with the value of solidity increasing.
VAWT. The performance of VAWT can be enhanced by appropriately
3) The addition of GF on the outer side of the blade can improve increasing the chord length of blade, and the augmentation of
significantly aerodynamic of VAWT at TSR  2.51. While at the blade worsens the performance.
higher TSR, even though the GF achieves better performance 6) The downstream VAWT in array configurations with each so-
within upwind half, it results in sharply augmentation of drag lidity obtains higher power extraction than the isolated coun-
force within downwind half, leading to similar or slightly worse terpart at overall TSR. Additionally, at TSR ¼ 2.27, downstream
performance within a whole revolution compared to clean rotor. VAWT with a chord length of 123.5 mm achieves the maximum
So that the GF loses effectiveness even leads to a decline of the average torque of 5.35 N m, with an increment of 23.1%
power output at relatively higher TSR. compared to isolated counterpart.
4) The addition of GF on the upstream pairs of VAWTs can achieve
higher flow velocity through the gap compared to paired clean In this paper, to improve the overall performance of VAWT farm

14
L. Ni, W. Miao, C. Li et al. Energy 215 (2021) 118915

configurations, the effects of Gurney flap and solidity on the per- performance of twin vertical axis wind turbines via numerical simulations.
Renewable Energy 2020;147:43e54.
formance of turbines in array configurations are studied. However,
[19] Posa A. Wake characterization of coupled configurations of vertical axis wind
the conditions of incoming flow are kept constant. Thus, further turbines using Large Eddy Simulation. International Journal of Heat and Fluid
work is required to explore the effects of various types of incoming Flow 2019;75:27e43.
wind which may contain various direction in array configurations, [20] Ahmadi-Baloutaki M, Carriveau R, Ting SK. A wind tunnel study on the
aerodynamic interaction of vertical axis wind turbines in array configurations.
turbulent intensity and magnitude of wind speed on the perfor- Renewable Energy 2016;96:904e13.
mance of VAWT farm configurations. In addition, the optimal [21] Barnes A, Hughes B. Determining the impact of VAWT farm configurations on
arrangement distances of VAWT configurations could relate to the power output. Renewable energy 2019;143:1111e20.
[22] Liu TS, Montefort J. Thin-airfoil theoretical interpretation for gurney flap lift
solidity of turbine. So the effect of solidity on the optimal enhancement. Journal of Aircraft 2007;44(2):667e71.
arrangement distances of VAWT farm configurations would be [23] Liebeck Robert H. Design of subsonic airfoils for high lift. Journal of Aircraft
investigated in the further works. 1978;15(9):547e61.
[24] Storms BL, Jang CS. Lift enhancement of an airfoil using a Gurney flap and
vortex generators. Journal of Aircraft 1994;31(3):542e7.
Declaration of competing interest [25] Wang JJ, Li YC, Choi KS. Gurney flap-Lift enhancement, mechanisms and ap-
plications. Progress in Aerospace 2008;44(1):22e47.
[26] Li Y, Wang J, Zhang P. Effects of gurney flaps on a NACA 0012 airfoil. Flow,
The authors declare that they have no known competing Turbul Combust 2002;68(1):27e39.
financial interests or personal relationships that could have [27] Giguere P, Lemay J, Dumas G. Gurney flap effects and scaling for low-speed
appeared to influence the work reported in this paper. airfoils. Applied Aerodynamics Conference 1995.
[28] Brown L, Filippone A. Aerofoil at low speeds with Gurney flaps. The Aero-
nautical Journal 2003;107(1075):539e46.
Acknowledgements [29] Zhu HT, Hao WX, Li C, et al. A critical study on passive flow control techniques
for straight-bladed vertical axis wind turbine. Energy 2018;165:12e25.
The authors gratefully acknowledge the financial support of [30] Bianchini A, Balduzzi F, Di Rosa D, et al. On the use of Gurney Flaps for the
aerodynamic performance augmentation of Darrieus wind turbines. Energy
National Natural Science Foundation of China (grand number: Conversion and Management 2019;184:402e15.
51976131 and 51676131), and this project also funded by the [31] Zhang PF, Liu AB, Wang JJ. Aerodynamic modification of a NACA 0012 airfoil
Shanghai University Science and Technology Innovation Action by trailing-edge plasma gurney flap. Aiaa Journal 2009;47(10):2467e74.
[32] Feng LH, Choi KS, Wang JJ. Flow control over an airfoil using virtual Gurney
Plan Local University Capacity Building Project (grand number: flaps. Journal of Fluid Mechanics 2015;767:595e626.
19060502200). [33] Lee T. Aerodynamic characteristics of airfoil with perforated gurney-type
flaps. Journal of Aircraft 2009;46(2):542e8.
[34] Cheng Z, Madsen HA, Gao Z, et al. Effect of the number of blades on the dy-
References namics of floating straight-bladed vertical axis wind turbines. Renewable
Energy 2017;101:1285e98.
[1] Peng HY, Lam HF, Liu HJ. Power performance assessment of H-rotor vertical [35] Fiedler A, Tullis S. Blade offset and pitch effects on a high solidity vertical Axis
axis wind turbines with different aspect ratios in turbulent flows via experi- wind turbine. Wind Engineering 2009;33(3):237e46.
ments. Energy 2019;173:121e32. [36] Howell R, Qin N, Edwards J, et al. Wind tunnel and numerical study of a small
[2] Wekesa DW, Wang C, Wei Y, et al. Influence of operating conditions on un- vertical axis wind turbine. Renewable Energy 2010;35(2):412e22.
steady wind performance of vertical axis wind turbines operating within a [37] Joo S, Choi H, Lee J. Aerodynamic characteristics of two-bladed H-Darrieus at
fluctuating free-stream: a numerical study. J Wind Eng Ind Aerod 2014;135: various solidities and rotating speeds[J]. Energy 2015;90(1):439e51.
76e89. [38] Li Q, Maeda T, Kamada Y, et al. Effect of solidity on aerodynamic forces around
[3] Acarer S. Peak lift-to-drag ratio enhancement of the DU12W262 airfoil by straight-bladed vertical axis wind turbine by wind tunnel experiments
passive flow control and its impact on horizontal and vertical axis wind tur- (depending on number of blades). Renewable Energy 2016;96(A):928e39.
bines. Energy 2020;201:117659. [39] Lam HF, Liu YM, Peng HY, et al. Assessment of solidity effect on the power
[4] Miao WP, Li C, Pavesi G, et al. Investigation of wake characteristics of a yawed performance of H-rotor vertical axis wind turbines in turbulent flows. Journal
HAWT and its impacts on the inline downstream wind turbine using unsteady of Renewable and Sustainable Energy 2018;10(2):023304.
CFD. J Wind Eng Ind Aerod 2017;168:60e71. [40] Rezaeiha A, Montazeri H, Blocken B. Towards optimal aerodynamic design of
[5] Tran TT, Kim DH. A CFD study of coupled aerodynamic-hydrodynamic loads vertical axis wind turbines: impact of solidity and number of blades. Energy
on a semisubmersible floating offshore wind turbine. Wind Energy 2018;165(B):1129e48.
2018;21(1):70e85. [41] Roh SC, Kang SH. Effects of a blade profile, the Reynolds number, and the
[6] Lignarolo LEM, Ragni D, Krishnaswami C, et al. Experimental analysis of the solidity on the performance of a straight bladed vertical axis wind turbine.
wake of a horizontal-axis wind-turbine model. Renewable energy 2014;70: Journal of Mechanical Science and Technology 2013;27(11):3299e307.
31e46. [42] Okeoghene, Eboibi, Louis, et al. Experimental investigation of the influence of
[7] Mckay P, Carriveau R, Ting D. Farm wide dynamics: the next critical wind solidity on the performance and flow field aerodynamics of vertical axis wind
energy frontier. Wind Eng; 35(4):397-418. turbines at low Reynolds numbers. Renewable Energy 2016;92:474e83.
[8] Dabiri JO. Potential order-of-magnitude enhancement of wind farm power [43] Abu-El-Yazied TG, Ali AM, Al-Ajmi MS. Effect of number of blades and blade
density via counter-rotating vertical-axis wind turbine arrays. J Renew Sustain chord length on the performance of Darrieus wind turbine. American Journay
Energy 2011;3:043104. of Mechanical Engineering and Automation 2015;2(1):16e25.
[9] Mohamed SO, Ibrahim A, Etman A, et al. Numerical investigation of Darrieus [44] Zhu HT, Hao WX, Li C, et al. Numerical study of effect of solidity on vertical
wind turbine with slotted airfoil blades. Energy Conversion and Management axis wind turbine with Gurney flap. Journal of Wind Engineering and Indus-
2020;5:100026. trial Aerodynamics 2019;86:17e31.
[10] Qasemi K, Leila N. Optimization of the power output of a vertical axis wind [45] Sagharichi A, Zamani M, Ghasemi A. Effect of solidity on the performance of
turbine augmented with a flat plate deflector. Energy 2020;202:117745. variable-pitch vertical axis wind turbine. Energy 2018;161:753e75.
[11] Schatzle PR, Klimas PC, Spahr HR. Aerodynamic interference between two [46] Castelli MR, Ardizzon G, Battisti L, et al. Modeling strategy and numerical
Darrieus wind turbines. Journal of Energy 1980;5(2):84e8. validation for a Darrieus vertical axis micro-wind turbine. Proceedings of the
[12] Thomas RN. Coupled vortex vertical axis wind turbine. US Patent No.6 ASME 2010 International Mechanical Engineering Congress and exposition.
2004;566:784. Vancouver: British Columbia; 2010. p. 409e18.
[13] Duraisamy K, Lakshminarayan V. Flow physics and performance of vertical [47] Rezaeiha A, Montazeri H, Blocken B. Towards accurate CFD simulations of
Axis wind turbine arrays. Aiaa Applied Aerodynamics Conference2014. vertical axis wind turbines at different tip speed ratios and solidities: guide-
[14] Lam HF, Peng HY. Measurements of the wake characteristics of co- and lines for azimuthal increment, domain size and convergence. Energy Con-
counter-rotating twin H-rotor vertical axis wind turbines. Energy 2017;131: version and Management 2018;156:301e16.
13e26. [48] Wekesa DW, Wang C, Wei Y, et al. Analytical and numerical investigation of
[15] Vergaerde A, Troyer TD, Standaert L, et al. Experimental validation of the unsteady wind for enhanced energy capture in a fluctuating free-stream.
power enhancement of a pair of vertical-axis wind turbines. Renewable En- Energy 2017;121:854e64.
ergy 2020;146:181e7. [49] Rezaeiha A, Montazeri H, Blocken B, et al. On the accuracy of turbulence
[16] Kinzel M, Dabiri JO. Performance enhancement of downstream vertical-axis models for CFD simulations of vertical axis wind turbines. Energy 2019;180:
wind turbines. Journal of Renewable and Sustainable Energy 2016;8:053306. 838e57.
[17] Zanforlin S, Nishino T. Fluid dynamic mechanisms of enhanced power gen- [50] Araya DB, Colonius T, Dabiri JO. Transition to bluff-body dynamics in the wake
eration by closely spaced vertical axis wind turbine. Renewable Energy of vertical-axis wind turbines. Journal of Fluid Mechanics 2017;813:346e81.
2016;12:1213e26.
[18] Peng HY, Han ZD, Liu HJ, et al. Assessment and optimization of the power

15