See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/338936731

Numerical Simulation of Flow Over Different Types of Airfoils

Conference Paper · January 2020

CITATION READS

1 1,139

3 authors:

Neslihan Aydın Mehmet Erman Çalışkan

Uludag University Uludag University
14 PUBLICATIONS   1 CITATION    8 PUBLICATIONS   19 CITATIONS   

SEE PROFILE SEE PROFILE

Irfan Karagoz
Uludag University
52 PUBLICATIONS   763 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Neslihan Aydın on 31 January 2020.

The user has requested enhancement of the downloaded file.

Numerical Simulation of Flow Over Different Types
of Airfoils
Neslihan Aydın*, Mehmet Erman Çalışkan*, İrfan Karagöz#
*
Department of Mechanical Engineering, Uludag University, Bursa, Turkey,16059
nslhngunes@uludag.edu.tr

#
Department of Mechanical Engineering, Uludag University, Bursa, Turkey,16059
karagoz@uludag.edu.tr
Abstract— Airfoils in various types are widely used in many normal airfoil design at the speed of 100 m/s [1]. Another
devices subjected to fluid flows such as aircraft, vehicles, experimental and numerical study was performed on the flow
turbines, etc. Therefore, analyzing the fluid flow around an around NACA 0015 airfoil design at angles of attack ranges
airfoil is one of the important subjects in fluid mechanics. In this from 2 to 18º. Numerical results were obtained by using
study, the conservation equations of two-dimensional
different turbulences models which are Spalart Allmaras and
compressible flow over standard aerofoils were solved by using
different numerical techniques. After a mesh independence k-epsilon. They found that the best results for lift and drag
study, applied mathematical model, numerical methods and coefficient were obtained 16º attack angle [2]. A numerical
obtained results are confirmed with experimental results given in simulation aimed to analyze the aerodynamic characteristics
the literature. Three different turbulence models, namely the k- of NACA0012 airfoil was carried out by a researcher who
, the Spalarat Almaras and Reynolds stress models were used in focused on the designing a airfoil with better aerodynamic
the solutions. The performances of turbulence models were performance. NACA 0012 and the turbulence flow structure
evaluated under the results obtained. The verified numerical were tested in detail with the Large Eddy Simulation Model
model was also applied to the flow over different types of blades, for different attack angles [3]. The flow behaviour over AG-
including a special airfoil design. Velocity and pressure fields
16 airfoil was studied and analysed numerically by researchers,
obtained around these airfoils were analysed, and their
aerodynamic performances in terms of the lift and drag
aerodynamic performance of AG-16 airfoil at different angles
coefficients were compared to each other at different angles of of attack ranging from 0 degree to 15 degree was obtained
attack. using ANSYS-FLUENT software. They found that the
maximum lift coefficient for high lift airfoil type AG-16
Keywords— renewable energy, computational fluid dynamics, recorded at 0.0116 and the drag coefficient magnitude was
equal to 0.0013 when the angle of attack reached the stall limit
airfoils, turbulence models, lift and drag coefficients. [4]. A numerical study was performed with k-ω shear stress
transport model (SST Model) for turbulence intensities of 1%
and 5%. NACA 0012 airfoil subjected to different flap angles
and Mach number. They calculated lift coefficients (CL), drag
I. INTRODUCTION coefficients (CD) and CL/CD ratio at different operating
Since energy is indispensable for human beings, research conditions and showed that with increasing Mach number (M)
works are underway on especially renewable energy sources CL increases but CD remains somewhat constant [5].
such as solar energy, wind energy, and hydroelectric energy. Matyushenko et al. investigated numerically and
One of the most widely used renewable energy types is wind experimentally airfoils with different 6shapes )
and thicknesses
energy in which energy is generally obtained by means of at high Reynolds numbers (Re≥10 and low turbulence
wind turbines. One of the parts to be considered in the design intensity (I≤0,1%) using two-dimensional Reynolds-Averaged
of the turbine is blade. The airflow around the turbine blades Navier-Stokes equations (RANS) closed by different
affects the amount of energy production. Therefore, the turbulence models. Results show that comparison with the
velocity of the airflow around an airfoil and profile shapes corresponding data for the γ-SST model; both models are
have been investigated by many researchers. unable to predict the dependence of the lift coefficient on the
Recently, flow phenomena passing through solids have angle of attack [6]. Holden et al. investigated the design of
been an important research subject due to aerodynamic effects wind turbine blade that inspired by maple seed in their study.
such as lifting and drag on solid bodies. Especially in airfoil As a result of the study biomimetic wind turbine airfoil maple
design, different shape airfoil profiles and their aerodynamic seed profile Cp power coefficient increased to a maximum
effects have been an important research subject. value of 0,59. They also reached C L value up to 0,8 for
Among the studies, bio-inspired wing designs are maximum Re=10000 [7].
noteworthy. Başak & Demirhan (2017) were inspired by the
fins of the humpback whales, and they developed tubercled
airfoil design. As a result they found that the tubercled airfoil II. TURBULENCE MODELS
increased efficiency by approximately 42.09 % compared to
 In order to examine the velocity and pressure distributions 𝜕 𝜕 𝜕𝜌 𝜕 𝜕𝑢𝑖 𝜕𝑢𝑗
(𝜌𝑢𝑖 ) + (𝜌𝑢𝑖 𝑢𝑗 ) = − + [µ ( + −
in a flow field, the conservation of mass and momentum 𝜕𝑡 𝜕𝑥𝑗 𝜕𝑥𝑖 𝜕𝑥𝑗 𝜕𝑥𝑗 𝜕𝑥𝑖
equations must be solved under the existing boundary 2 𝜕𝑢𝑖 𝜕
𝛿 )] + ̅̅̅̅̅̅
(−𝜌𝑢 ′ ′
𝑖 𝑢𝑗 ) (6)
conditions [8]. However, it is difficult and even impossible to 3 𝑖𝑗 𝜕𝑥𝑗 𝜕𝑥𝑗
define these conservation equations in complex geometries
and solve them analytically. Therefore, equations must be Here 𝑢𝑖 is velocity components and (−𝜌𝑢 ̅̅̅̅̅̅
′ ′
𝑖 𝑢𝑗 ) indicates
solved numerically. Reynolds turbulence stresses. Due to the chaotic nature of
According to Newton's law, the motion of a viscous fluid is turbulence there is no analytical method for calculating these
defined by the Navier-Stokes equations in differential form. values. Turbulence models are used to calculate the turbulence
Navier-Stokes equations formulize the differential form of stresses in the momentum equation. Turbulence models have
conservation of momentum [9]. Incompressible Navier-Stokes been developed to calculate these values.
equations in Cartesian coordinates are expressible as follows;
𝜕𝑢 𝜕𝑢 𝜕𝑢 𝜕𝑢 1 𝜕𝑃
𝑥 − 𝑑𝑖𝑟𝑒𝑐𝑡𝑖𝑜𝑛 = + 𝑢 + 𝑣 + 𝑤 = − + 𝑔𝑥 +
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝑃 𝜕𝑥
𝜕2 𝑢 𝜕2 𝑢 𝜕2 𝑢
𝑣 ( 2 + 2 + 2) (1)
𝜕𝑥 𝜕𝑦 𝜕𝑧

𝜕𝑣 𝜕𝑣 𝜕𝑣 𝜕𝑣 1 𝜕𝑃
𝑦 − 𝑑𝑖𝑟𝑒𝑐𝑡𝑖𝑜𝑛 = +𝑢 +𝑣 +𝑤 =− + 𝑔𝑦 +
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝑃 𝜕𝑦
𝜕2 𝑣 𝜕2 𝑣 𝜕2 𝑣
𝑣 ( + + ) (2)
𝜕𝑥 2 𝜕𝑦 2 𝜕𝑧 2
𝜕𝑤 𝜕𝑤 𝜕𝑤 𝜕𝑤 1 𝜕𝑃
𝑧 − 𝑑𝑖𝑟𝑒𝑐𝑡𝑖𝑜𝑛 = +𝑢 +𝑣 +𝑤 =− +
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝑃 𝜕𝑧
𝜕2 𝑤 𝜕2 𝑤 𝜕2 𝑤
𝑔𝑧 + 𝑣 ( + + ) (3)
𝜕𝑥 2 𝜕𝑦 2 𝜕𝑧 2

The continuity equation is based on the principle of Fig 1. Turbulence models used in ANSYS [10].
conservation of mass in control volume, and can be written in
differential form as follows; B. Spalart-Allmaras Model
This model is a one equation model for turbulent viscosity
𝜕𝜌 𝜕(𝜌𝑢) 𝜕(𝜌𝑣) 𝜕(𝜌𝑤)
+ + + =0 (4) and it solves just one transport equation for viscosity; Spalart-
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧
Allmaras is a low-cost RANS model solving a transport
By assuming constant density in equations (1), (2), (3) and equation for a modified eddy viscosity. In particular, it gives
(4); the unknown terms are pressure and velocity components good results in the flow around the wall in the boundary layer.
and in the x, y and z directions. By solving these four Turbomachinery has started to gain popularity in applications
equations, these four unknowns value can be obtained. On the [11,12].
other hand, the Navier-Stokes equations cannot be solved 𝜕 𝜕 1 𝜕 𝜕𝑣̂
(𝜌𝑣̂) + (𝜌𝑣̂𝑢𝑖 ) = 𝐺𝑣 + [ [{(µ + 𝜌𝑣̂) }+
analytically without any simplification. The flow problem is 𝜕𝑡 𝜕𝑥𝑗 𝜎𝑣̂ 𝜕𝑥𝑗 𝜕𝑥𝑗
defined within the numerical solution and the result is 𝜕𝑣̂ 2
obtained for each discrete point under specified boundary and 𝐶𝑏2 𝜌 ( )]] − 𝑌𝑣 (7)
𝜕𝑥𝑗
initial conditions. However, the Navier- Stokes equations need
to be modified in order to take into account turbulence and In equation (7); 𝑣̂ turbulence kinematic viscosity, 𝐺𝑣
adopt turbulence models. The Reynolds-Averaged Navier- turbulence production, 𝑌𝑣 turbulence destruction; 𝜎𝑣̂ and 𝐶𝑏2
Stokes (RANS) equations are obtained for this purpose. indicate constants.
A. Reynolds-Averaged Navier- Stokes Equations RANS
In these equations, flow properties are separated into two C. Standard k-ε Model
components: time-averaged and time-varying parts. The K-ε turbulence model is the most common model used in
Navier-Stokes equations are rearranged using this principle Computational Fluid Dynamics (CFD) to simulate mean flow
and Reynolds Averaged Navier-Stokes (RANS) equations are characteristics for turbulent flow conditions. It is a two
obtained. Time resolved RANS equations are called as Time- equation model which gives a general description of
Dependent Reynolds Averaged Navier- Stokes (URANS) turbulence by means of two transport equations [13].
equations. Continuity equation (5) and URANS equations (6) The standard k-ε turbulence model ; Launder and Spalding,
respectively can be written as: 1974 is used which is based on our best understanding of the
relevant processes, therefore minimizing unknowns and
𝜕𝜌 𝜕 presenting a set of equations which can be applied to a large
+ (𝜌𝑢𝑖 ) = 0 (5)
𝜕𝑡 𝜕𝑥𝑖 number of turbulent applications [14].
For turbulent kinetic energy (k);
 III. GEOMETRY AND MESH DESIGN PARAMETERS
𝜕(𝜌𝑘) 𝜕(𝜌𝑘𝑢𝑖 ) 𝜕 µ𝑡 𝜎𝑘
+ = [ ] + 2µ𝑡 𝐸𝑖𝑗 𝐸𝑖𝑗 − 𝜌𝜖 (8)
𝜕𝑡 𝜕𝑥𝑖 𝜕𝑥𝑗 𝜎𝑘 𝜕𝑥𝑗
A. Geometry Design
For dissipation (𝛆);
Naca 0015 airfoil profile that shown in Fig. 2. was used in the
̅̅̅̅̅̅̅̅
𝑢𝑖′ 𝑢𝑖′ analysis. This profile’s data files were taken from
ε=𝑣 (9) airfoiltools.com and geometry was designed on Solidworks
𝜕𝑥𝑘 𝜕𝑥𝑘
2018 as a 3d.

Where; ui represents velocity component in corresponding

direction, Eij represents component of rate of deformation, μij
represents eddy viscosity.
Fig. 2. Naca 0015 airfoil profile
D. Standard k- ω Model (WİLCOX Model)
The other common and simplest model is k-ω model. It For dimensions of Naca 0015 profile, Robert E. and friend’s
provides better modeling of the turbulent boundary layer than (1981) paper was used for reasonable compare. Dimension of
the standard k-ε model, however is more sensitive to the free- Naca 0015 airfoil that used in this paper were shown in Fig. 3.
stream turbulence levels [15]. Given the processes of
convection, diffusion and destruction or dissipation, the model
equation for ω is given below:
𝜕𝜔 𝜕𝜔 𝜕 𝜕𝜔
𝜌 + 𝜌𝑈𝑗 = ((𝜇 + 𝜎𝜇𝑡 ) ) + 𝑃𝜔 − 𝐷𝜔 (10)
𝜕𝑡 𝜕𝑥𝑗 𝜕𝑥𝑗 𝜕𝑥𝑗

The model attempts to predict turbulence by two partial

differential equations for two variables, k and ω, with the first
variable being the turbulence kinetic energy (k) while the
second (ω) is the specific rate of dissipation of the turbulence
kinetic energy (k).
𝜔 𝜕𝑈𝑖 𝜔
𝑃𝜔 = 𝛼 𝜏𝑖𝑗 = 𝛼 𝑃𝑘 (11)
𝑘 𝜕𝑥𝑗 𝑘
𝐷𝜔 = 𝛽𝜌𝜔2 (12)
Fig. 3. Dimensions of Naca 0015 airfoil profile (units are
mm)
E. Reynolds Stress Equation Model (RSM)
Reynolds stress equation model (RSM), also known as A computational flow domain must be covered this Naca 0015
second order or second moment closure model is the nearly airfoil profile. There are many flow domain types that used in
most complex classical turbulence model. Several numerical analysis.
shortcomings of k-ε turbulence model were observed when it In this study, C type geometry was used as a flow domain (Fig.
was attempted to predict flows with complex strain fields or 4.). Because C type mesh structure allows to create less mesh.
substantial body forces. Calculation time is longer than other Also C type domain is more convenient to create a real flow
turbulence models, but keep in sight the lack of isotropic area.
turbulence. The Reynolds averaged momentum equations for
the mean velocities are:
𝜕𝑈𝑖 𝜕(𝜌𝑢𝑖 𝑢𝑗) 𝜕 𝜕𝑈𝑖 𝜕𝑈𝑗 𝜕𝑝′′ ̅̅̅̅̅̅̅)
𝜕(𝜌𝑢 𝑖 𝑢𝑗
+ − [𝜇 ( + )] = − − +
𝜕𝑡 𝜕𝑥𝑗 𝜕𝑥𝑗 𝜕𝑥𝑡 𝜕𝑥𝑖 𝜕𝑥𝑖 𝜕𝑥𝑗
𝑆𝑀𝑖 (13)
Where p’’ is a modified pressure, 𝑆𝑀𝑖 is the sum of body
forces and the fluctuating Reynolds stress contribution
is −𝜌𝑢̅̅̅̅̅.
𝑗 𝑢𝑖
RSM needs more modelling. It is more difficult to
convergence. Strong streamline curves are suitable for
complex 3D streams with swirl and rotation.
 B. Mesh Generation

Mesh generation was made on HYPERMESH 13.0. In order

to ensure the desired mesh thickness around the airfoil,
different sized mesh structure was generated in different parts
of the flow domain geometry. This is done to keep the
magnitude of Y-plus as small as possible. The mesh structure
created was shown in Fig. 7. from different views.

Fig. 4. C type geometry that was designed.

Ansys fluent tutorials was used to determine the sizes of flow

areas. Ansys fluent tutorials lectures suggest 10x chord length
between air inlet and center of naca profile. Also it suggests
20x chord length between air outlet and center of naca airfoil
profile (Fig. 5.).

Fig. 5. Certain distances about the c type geometry

Flow domain was subdivided into a series of region. The

purpose of this, to create different sized mesh in different
regions. Thus, it was able to create finer meshes around the
airfoil. The regions on geometry are shown in Fig.6.
Fig. 7. Different views of mesh structure.

Also, fine mesh structure around the airfoil is seen in Fig. 8. in

detail. This view was created by middle section of flow
domain.

Fig. 6. Different regions on flow domain geometry.

 Fig. 10. Setup of the velocity inlet
Fig. 8. Section view of mesh structure.
Determined values for the reference values were shown in Fig.
The flow domain is formed by hexa mesh structure. This mesh 11. Area value from these values, was calculated by
structure’s number of elements is 9533632 and number of multiplying the chord length with the airfoil z length
nodes 10059476 (Fig. 9.). When creating this mesh structure, (0.152*0.91). The length value was taken as the chord length
2d mesh structure was created. 3d mesh structure was then of 0.152 m.
generated by these 2d mesh structure.

Fig. 9. Number of nodes and elements of mesh structure.

C. Set up of Analysis

For numerical analysis of Naca 0015 airfoil, ANSYS 16.2

and 17.1 fluent was used. During the analysis, the effect of
several parameters on the lift and drag coefficient of the this
airfoil was examined.
Fig. 11. Reference values
For this, certain setups have been made on the program. For
solution type, Pressure-Based and Steady Type have been
For the solution method, The Simple scheme was used. Also
used in the analysis.
other values were given in Fig. 12. in detail.
The inlet of air was adjusted as shown in the Fig. 5.
Magnitude of air inlet velocity, was determined to make the
Reynolds number of 40000. So magnitude of inlet velocity
was determined as a 4.71m/s.
The attack angles were also adjusted in this section to examine
the performance of the airfoil at certain attack angles (Fig.
10.).
 Fig. 14. Vectors of Cd components calculation of Cd

As shown in the Fig. 14., direction must be same as flow,

aligned with chord line or attack angle.

Fig. 12. Setup of solution method IV. RESULTS

In the analysis, firstly, the Cl and Cd coefficients of the Naca
For Cd (drag coefficient) and Cl (lift coefficient) calculation, 0015 airfoil were examined, with 7 ° attach angle and Re
certain angle values must be written on the x and y component number of 40000, by using different turbulence models. These
locations on monitors section in program. How these angle results were compared with the experimental data set in the
values were calculated was shown in the Fig. 13. and Fig. 14. study of Robert et al. (1981) and turbulence model which
gives the best results was determined (TABLE I).

TABLE II

COMPARISON OF CL AND CD COEFFICIENTS ACCORDING TO

TURBULENCE MODELS AT 7° ATTACK ANGLE AND 40000 RE
NUMBER

TURBULENCE MODEL Cl Cd
1 Spalart Almaras 0,48543 0,038633
2 k-ε (Realizable) 0,43315 0,053946
3 k-ω (SST) 0,46362 0,037405
4 k-ω (Standart) 0,46825 0,037019

Experimental Data 0,5730 0,0267

Fig. 13. Vectors of Cl components; calculation of Cl As can be seen from the TABLE III, Spalart Almaras model is
the most suitable model.
The results obtained from the Spalart Almaras model, with 7°
As shown in the Fig. 14., direction must be normal to flow or attack angle and 40000 Re number, were examined.
perpendicular to drag. So mainstream velocity vector must be
stated by 90°.
A. Scaled Residuals
C. Pressure Contours

Fig. 15. Scaled residuals of number 1 analysis in the TABLE

IV Fig. 17. Pressure contour of number 1 analysis in the TABLE
V
As can be seen from Fig.15. , convergence curves are
sufficiently reduced. The high pressure at the bottom inlet of the airfoil has
B. Yplus Values increased up to 13,75Pa. In the upper part of the airfoil,
because of the high velocity, the pressure has dropped to -
23,71Pa. When the pressure contours are examined, it is seen
that the pressure distribution around the airfoil develops
regularly.

D. Velocity Contours

Fig. 16. Y plus values of number 1 analysis in the TABLE I

Yplus value is often used to describe how coarse or fine a

mesh is for a particular flow pattern. In the airfoil analysis,
Yplus value should be examined on the airfoil. Usually this
Yplus value is required max value 1. Therefore, the mesh
thickness around the airfoil, which is shown in detail in the
Fig. 8, was reduced to 0.05 mm.
As can be seen from the Fig.16., the max y plus value on the
airfoil was obtained from the analysis is 1,405.

Fig. 18. Velocity contour of number 1 analysis in the TABLE

VI
As can be seen in the Fig.18, a wake region was formed in the
rear part where the air is separated from the airfoil. This wake
area was about 1 meter long. Air flowed slightly slower than
the inlet velocity in the wake region

E. Velocity Vectors

Fig. 20. CLand Cd values of Naca0015 Airfoil according to

attack angle with 40000 Re number.

TABLE II
CL AND CD VALUES OF NACA 0015 AIRFOIL ACCORDING TO
ATTACK ANGLE WITH RE NUMBER 40000

α Cl Cd Cl/Cd
5 0,35043 0,030808 11,37464
7 0,48543 0,038633 12,56516
10 0,67569 0,055131 12,25608
12 0,77991 0,069345 11,24681
15 0,85625 0,1055 8,116114
16 0,55288 0,18567 2,977756
18 0,54624 0,22305 2,448958

There is a stall angle where the coefficient of CL starts to

decrease while the angle of attack increases. As can be seen
Fig. 19. Velocity vector of number 1 analysis in the TABLE from the Fig.20, this stall angle is 15° for this analysis. Also,
VII the highest ratio of CL/CD was observed at 7°.

When air velocity vectors are considered, it is seen that air V. CONCLUSIONS
velocity increases up to 7,52 m/s in upper parts of airfoil after
entering air. In this paper, NACA0015 Airfoil are investigated with
different turbulence methods; k-𝜀, k-𝜔, k-𝜔 (SST) models.
F. Investigation of NACA 0015 AİRFOİL performance  The Spalart-Allmaras model provided CL and CD
according to attack angle values in the best agreement with experimental
After determining the most suitable turbulence model, CL and results. This model is generally recommended for
CD coefficients were examined according to attack angle with flow analysis over airfoils [2].
40000 Re number. The results obtained from these analyses  In addition, k- 𝜔 models have been shown to
between 5° and 18° attack angles are presented in the Fig.1 provide good results for flow analysis over the
and TABLE II. airfoil.
 CD value increased with attack angle but CL
value was initially increasing with attack angle
after a stagnation point these was called as a stall
angle or stall point CL decreased critically. In this
study stall angle was obtained as 15°.
 The performance of any airfoil is measured with
the CL/CD ratio. The highest ratio of CL/CD was
observed at 7°.
 The max y plus value on the airfoil was obtained
from the analysis is 1,405. Considering that the
 max yplus value should be 1, this result was
considered acceptable.
References
[1] H. Başak , H. Demirhan , Examination of Airfoil Profile Yield Inspired
by The Fins of Humpback Whale with CFD Analysis , Journal of Gazi
Engineering Sciences, 2017, 3(2):15-20.
[2] I. Şahin, A. Acir, Numerical and Experimental Investigation of Lift and
Drag Performances of NACA0015 Wind Turbine Airfoil, IJMMM
2015 Vol.3(1): 22-25 ISSN: 1793-8198.
[3] L.B. Streher, , “Large-eddy simulations of the flow around a
NACA0012 airfoil at different angles of attack”,Physics, Fluid
Dynamics thesis, Helmut-Schmidt-Universitat der Bundeswehr,
Hamburg, May 2017.
[4] A. H. Mutaib, A. AL-Khateeb, M. K. Khashan, F. Kamil,
Computational Study of Flow Characteristics Over High Lift Airfoil at
Various Angles of Attack, Journal of Mechanical Engineering
Research & Developments (JMERD) 42(1) (2019) 90-9.
[5] T. Ahmed, Md. T. Amin, S.M. R. Islam, S. Ahmed, Computational
Study of Flow Around a NACA 0012 Airfoil Flapped at Different Flap
Angles with Varying Mach Numbers, Global Journal of Researches in
Engineering General Engineering Vol.13(4)(2013):5-16.
[6] A. A. Matyushenko, E. V. Kotov, A. V. Garbaruk, Calculations of
Flow Around Airfoils Using Two Dimensional RANS: An Analysis of
the Reduction in Accuracy,St. Petersburg Polytechnical University
Journal: Physics and Mathematics Vol.3(4) (2017):15-21.
[7] J. Holden, T. M. Caley, M. G. Turner, Maple Seed Performance as a
Wind Turbine, Conference: AIAA SciTech 2015, At Kissimmee, FL,
Volume: AIAA 2015-1304. DOI: 10.2514/6.2015-1304.
[8] Y. A. Çengel, J. M. Cimbala, Solutions Manual for Fluid Mechanics:
Fundamentals and Applications Second Edition, McGraw-Hill, 2010.
[9] B. Apaçoğlu, “Cfd Analyses of Uncontrolled and Controlled Laminar
and Turbulent Flows Over A Circular Cylinder”, Mechanical
Engineering Master thesis, TOBB University of Economics and
Technology,Ankara 2010.
[10] ANSYS FLUENT 13.0 Lecture 06 Turbulence User Guide,2013.
[11] P.R.Spalart, S.R. Almaras, A One-Equation Turbulence Model for
Aerodynamic Flows. 1992. AIAA Paper 92-0439.
[12] H. B. Ekmekci, “Modification Of A Computational Fluid Dynamics
Model (Ansys-Fluent) For The Purpose Of River Flow And Sediment
Transport Modeling”, Master of Science in Civil Engineering Thesis,
The Graduate School of Engineering and Sciences of İzmir Institute of
Technology, İzmir, July, 2015.
[13] F.Kaya, İ. Karagöz, Investigation into the Suitability of Turbulence
Models in Swirling Flows, Uludag University Journal of The Faculty
of Engineering Vol 12(1) (2007):85-96.
[14] B.E. Launder, D.B. Spalding, The Numerical Computation of
Turbulent Flows. Computer Methods in Applied Mechanics and
Engineering, 3, (1974) 269-289.
[15] D.C. Wilcox, Turbulence Modeling for CFD. 2006, DCW Industries,
Inc.
[16] R. E Sheldahl, P. C. Klimas, Aerodynamic characteristics of seven
symmetrical airfoil sections through 180-degree angle of attack for use
in aerodynamic analysis of vertical axis wind turbines (No. SAND-80-
2114). Sandia National Labs., Albuquerque, NM (USA). (1981).

View publication stats