Professional Documents
Culture Documents
Supersonic Flow Over A Double Circular Airfoil
Supersonic Flow Over A Double Circular Airfoil
Contents
1.
Introduction ..................................................................................................................................... 3
1.1
Task .......................................................................................................................................... 3
1.2
1.3
Theory...................................................................................................................................... 4
1.3.1
Shocks .............................................................................................................................. 4
1.3.2
Expansion waves.............................................................................................................. 5
1.3.3
1.3.4
1.3.5
1.3.6
1.4
2.
3.
4.
5.
1.4.1
Gridgen ............................................................................................................................ 9
1.4.2
Star-CCM+........................................................................................................................ 9
Calculations ................................................................................................................................... 10
2.1
2.2
2.2.1
Area_1 to Area_2........................................................................................................... 12
2.2.2
Area_1 to Area_4........................................................................................................... 12
2.2.3
Area_2 to Area_3........................................................................................................... 13
2.2.4
Area_4 to Area_5........................................................................................................... 14
2.3
2.4
Simulations .................................................................................................................................... 16
3.1
Mesh Generation................................................................................................................... 16
3.2
3.3
Comparison ................................................................................................................................... 19
4.1
4.2
4.3
5.2
5.2.1
Area_1 to Area_2........................................................................................................... 23
5.2.2
Area_1 to Area_4........................................................................................................... 23
5.2.3
Area_2 to Area_3........................................................................................................... 24
5.2.4
Area_4 to Area_5........................................................................................................... 25
5.3
5.4
Simulations .................................................................................................................................... 27
6.1
Mesh Generation................................................................................................................... 27
6.2
6.3
Comparison ................................................................................................................................... 30
7.1
7.2
7.3
1. Introduction
1.1
Task
Supersonic flow over a double circular arc airfoil is to be analyzed. The shape, given in figure 1,
consists of two circles with radius of curvature Rlower and Rupper. The o espo di g thi k ess is
upper and lower. The cord length is L. The airfoil is flying at a Ma h u e M a d a gle of atta k .
Given data is:
Rlower
L
upper
lower
Minf.
Rupper
Questions:
1) Working first with analytical methods, determine:
a. Thickness of the airfoil,
b. Location and strength of leading edge shocks,
c. Lift coefficient,
d. Drag coefficient.
2) Carry out flow simulation of the flow field in task 1 using Star-CCM+
3) Compare analytical result with simulation.
4) Additional task: Replace the biconical airfoil with double edge airfoil and repeat the same
tasks.
3
1.2
With the given task, the numerical part and simulation part have to be completed separately and
after the results it has to be compared whether the commercial codes in Star-CCM+ is using working
turbulence models.
The numerical analysis part has to be started by understanding the given data. The theory about the
given data then have to be analyzed and calculated with right understanding of the formulas. The
sequent for the solving the task would be started by geometrical analysis of the airfoil. Thickness of
the airfoil could be found easily with geometry. After determining the thickness of the airfoil,
location of the leading edge shocks has to be determined. Strength of the shock could be determined
after the calculation of the pressure difference after the shock over entrance pressure of the shock.
Lift and drag coefficient could be calculated by only neglecting the expansion waves caused by the
biconical surface. As the area and pressure varies by the geometry, differential calculations are used
in order to calculate both drag and lift coefficient.
1.3
Theory
1.3.1
Shocks
As we are figuring out the biconical airfoils behavior, we have to understand the basics of the
shockwaves. A shock wave is a type of propagating disturbance. Like an ordinary wave, it carries
energy and can propagate through a medium through a field such as the electromagnetic field. Shock
waves are characterized by an abrupt, nearly discontinuous change in the characteristics of the
medium. Across a shock there is always an extremely rapid rise in pressure,
temperature and density of the flow. In supersonic flows, expansion is achieved through
an expansion fan. The flow has the capability to bypass the speed barrier as a result of the viscosity
of the air. Thus, results moving shocks and oblique shocks. An oblique shock represents a moving
shock with a deflection angle.
An oblique shock wave, unlike a normal shock, is
formed with respect to the upstream flow direction.
It will occur when a supersonic flow encounters a
corner that effectively turns the flow into itself and
compresses. The upstream streamlines are uniformly
deflected after the shock wave. The most common
way to produce an oblique shock wave is to place a
Figure 2 Oblique Shock Wave
wedge into supersonic, compressible flow. Similar to
a normal shock wave, the oblique shock wave
consists of a very thin region across which nearly discontinuous changes in the thermodynamic
properties of a gas occur. While the upstream and downstream flow directions are unchanged across
a normal shock, they are different for flow across an oblique shock wave. Oblique shocks results in
pressure, temperature and density increase which is used as an intake flow advantage, or
disadvantage of fast deformation [1].
The sharp edges must be used in the wedges in order to create an attached oblique shock wave. The
wedge angle determines the shock to be attached or detached shock.
1.3.2
Expansion waves
Thus as it will also be calculated, in this project the edges are sharp enough to generate attached
oblique shock waves. However, as the first oblique shock occurs, the flow will be expanded because
of the biconical geometry. Thus, this makes the use of the Prandtl Meyer Expansion rules.
The exhaust outtake will be slipped in two regions. The density will be different in upper and lower
parts of the airfoil due to the angle of attack not being zero and the not symmetrical geometry. The
different densities will result as a slip line at the exhaust, thus this separation of the flows will result
in another compression angle. This angle will be the same as the angle of attack and two different
oblique shocks will be formed at the edge of the rare-end. So the leading edge oblique shocks and
expansion waves will be expected in the form of figure 5.
Oblique
Shocks
Mint
4
Slip Line
Expansion
Waves
upper lower
1 2 3 4
As i the figu e is the ep ese ts the defle tio a gle of the o li ue sho k a es. This is calculated
by the NACA (National Advisory Committee by Aeronautics) tables which have been determined by
mainly analytical calculations and experiments.
6
1.3.3
Our ambient pressure is given in mega Pascal where 0.101325 MPa is equals to 1 atm. This is
assumed that the ambient pressure is the same with the zero elevation. Thus, the pressure ratio in
our NACA tables could be used as it is so there is no need to convert the ambient pressure.
Temperature is given in Kelvin, as it is 300K. Also the temperature is assumed to be in a hot summer
day where the airfoil is considered once more to be in zero elevation.
1.3.4
Mach number is given 2.4 in our task. This makes the flow to be supersonic flow. After 1.0 M, the
flow is considered to be supersonic. However as it is known that between 0.8M and 1.2M the flow is
transonic where the shocks are formed in the expansion waves regardless of the edges. The Mach
number is defined by,
Where:
M: Mach number
: Velocity of the source relative to the medium
c: Speed of the sound in the medium where it is calculated by c
As fluid is considered ideal gas and isentropic, we can calculate the k by;
.
Kappa in air medium is calculated as Cp= 3.5R and Cv=2.5R and universal gas constant to be R=0,287
kj/kgK, thus making the k= 1.4. The NACA tables are only helpful if kappa is 1.4. Therefore, NACA
tables could be used after confirming the kappa.
The angle of attack has been given as minus, furthermore this is considered as the flow is reaching
the airfoil from upwards, thus this results to be landing position of the airfoil. The minus angle of
attack makes the upper part of the airfoils flow go faster than the lower part. This makes the
pressure on the lower part lower than the upper part, making the lift coefficient minus as well. The
drag remains the same as positive value of the angle of attack.
1.3.5
Lift and drag coefficients could be separated by the direction of the calculation. The force which is
lifting due to the pressure difference of two sides will be calculated by the vertical forces and named
lift coefficient. The result would be dimensionless number based on the dependencies of shape,
inclination and flow conditions.
7
The drag coefficient is a dimensionless quantity that is used to quantify the drag or resistance of an
object in a fluid medium. Where a lower drag coefficient indicates the object will have less
aerodynamic drag. The drag coefficient is always associated with a particular surface area.
The drag coefficient of any object comprises the effects of the two basic contributors to
fluid drag: skin friction and form drag. The drag coefficient of a lifting airfoil also includes the effects
of lift-induced drag. The drag coefficient of a complete structure such as an aircraft also includes the
effects of interference drag.
Drag coefficient CD can be calculated by the insertion of the complex dependencies into a formula as:
Where Cd is the lift coefficient, t is the thickness of the airfoil, c is the length of the airfoil and M is
the Mach number.
Main calculation in our task would require a complex calculation of both lift and drag, since the
surface area A is biconical and does not have an equal density distribution. Thus, the equation for the
lift and drag would be different then the subsonic transonic double edge, lift and drag coefficients.
1.3.6
A shock wave propagates through an undisturbed gas at the supersonic speed vo > ao, where a0 is the
speed of sound in the undisturbed gas. The value of v0 increases with the strength of the shock wave,
the strength being defined here as (p1 p0)/p0. As the strength of the shock wave approaches zero,
the speed of propagation approaches a0. The speed of a shock wave with respect to the compressed
gas behind it is subsonic: V1 > a1, where a1 is the speed of sound in the compressed gas behind the
shock wave. As the oblique shock wave fits the description of the vo > ao since it is supersonic flow the
strength of an oblique shock is calculated by the following formula:
1.4
Simulation Programs
1.4.1
Gridgen
Gridgen is a complete meshing toolkit used to generate two or three-dimensional grids for complex
geometries in a production environment, often where CFD is mission critical. The software's origins
are in the demanding U.S. aerospace industry, where Gridgen continues to earn its reputation for
usability and high quality grids, both of which are vital for reliable simulations. Today Gridgen is used
in aerospace, automotive, power generation, chemical process and other industries for which CFD is
an integral part of the design process [5].
Where the simulations were simulated in Star-CCM+, the meshes which needed for it were designed
in Gridgen V16.04R2.
1.4.2
Star-CCM+
Star-CCM+ is one of the most practical, complete and easy-to-use engineering simulation program in
the market. The main usage of this program is mainly computational fluid dynamics (CFD). This
p og a s ai ad a tages a e;
CAD & PLM integration, fully integrated with most common design programs such as ProEngineer, SolidWorks and Gridgen.
Built-in meshing technology, polyhedral meshing, which also will be used later to see the
results.
Intuitive simulation user environment
Multi-disciplinary solutions
Engineering analysis, especially for fluid mechanics department [6].
CFD (Computational Fluid Dynamics) is an essential part in fluid mechanics.. It represents a vast area
of u e i al a alysis i the field of fluids flo phe o e a. Head ay i the field of CFD simulations
is strongly dependent on the development of computer-related technologies and on the
advancement of our understanding and solving ordinary and partial differential equations (ODE and
PDE). Since direct numerical solving of complex flows in real-like conditions requires an
overwhelming amount of computational power success in solving such problems is very much
dependent on the physical models applied. These can only be derived by having a comprehensive
understanding of physical phenomena that are dominant in certain conditions.
2. Calculations
2.1
upper
L/2
Rupper
Where;
10
2
Rlower
2
L/2
lower
Where;
(1.3)
2.2
Shockwave Calculations
Area_2
Area_3
Area_1
Area_5
Area_4
11
Due to the attack of angle and irregular upper and lower bodies, the airfoil will have different shocks
at different areas. The area_1 will be the initial conditions, where the area_2 and area_4 is where the
first oblique shocks happen. Area_3 and area_4 will have another couple of oblique shocks due to
the flow having different densities at the upper part and lower part, thus forming a separation line.
The separation line will have the same angle as the angle of attack.
2.2.1
Area_1 to Area_2
We ill all the o espo di g o p essio a gle 1 and the result compression will be:
Area_1
1
M1
Area_2
From this point on, the formulas in the NACA reports will be used.
Mn1
1 1
M2n
M1
1
2.2.2
Area_1 to Area_4
We ill all the o espo di g o p essio a gle 4 and the result compression will be:
Area_1
M1
12
4
Area_4
3rd Oblique Shock
From this point on, the formulas in the NACA reports will be used.
2.2.3
Area_2 to Area_3
The expansion due to the biconical airfoil is calculated by the following formula:
Where the biconical airfoil angular length from start to end is 2*1, thus this will result the maximum
angle which M2 will turn and that is:
The v2(M*2) is the maximum turning angle during the biconical expansion waves. From table.2 we
can check the corresponding angle results ate the M*2=2.79 ( ~ 2.8) and 2=21o. The pressure ratios
for M2=1.89 M*2=2.79 are also:
13
The oblique shock wave will happen just at the rear edge of the airfoil because of the separation line.
And the separation line has to continue form of the angle of attack with the same angle. Thus the
calculations are:
We will call the corresponding compressio a gle 3 and the result compression will be:
M2
Area_2
Area_3
M*2
2
1
From this point on, the formulas in the NACA reports will be used.
1
Separation Line
From table 2, the values for M=1.31:
2.2.4
Area_4 to Area_5
The expansion due to the biconical airfoil is calculated by the following formula:
Where the biconical airfoil angular length from start to end is 2*1, thus this will result the maximum
angle which M2 will turn and that is:
14
The v4(M*4) is the maximum turning angle during the biconical expansion waves. From table.2 we
can check the corresponding angle results ate the M*4=3.03 and 2=19,27o. The pressure ratios for
M4=2.03, M*4=3.03 are also:
The oblique shock wave will happen just at the rear edge of the airfoil because of the separation line.
And the separation line has to continue form of the angle of attack with the same angle. Thus the
calculations are:
We ill all the o espo di g o p essio a gle 6 and the result compression will be:
Separation Line
4
From this point on, the formulas in the NACA reports will be used.
Area_4
From table 2, the values for M=1.54:
2.3
M*4
Area_5
2.4
3. Simulations
3.1
Mesh Generation
The required mesh to use in the
simulations was generated with the
help of the program which was already
mentioned in the theory section.
Gridgen v16 has the option to
automatically generate the required
domain after generating connectors.
The main mesh is a giant circle with
the airfoil in the middle. The left side is
defined to be the inlet and the right
side is the outlet. The generated mesh
has the Cartesian coordinate system
where the origin is the beginning of
the airfoil.
16
3.2
As discussed at the theory section, the simulation was done in Star-CCM+ in order to compare the
results with the numerical and experimental results. The following properties were selected in the
options of Star-CCM+.
Simulation Properties
Chosen Variables
Gas-Air
Ideal Gas
Implicit Unsteady
K-Epsilon Turbulence
Models
Coupled Flow
Two-Dimensional
Realizable K-Epsilon Two-Layer
RANS
Stationary
Reference Values
Initial Conditions
3.3
17
18
4. Comparison
After calculation numerically and carrying out a flow simulation in Star-CCM+ the following results
occur. First of all, numerical calculations are mainly analytical assumptions. This results for a varying
error. In simulation, the calculations are done by the Reynolds stress models which are done by
Reynolds Averaging of Navier Strokes Equations. Especially very high Reynolds number and using
RANS modeling the errors are considerably high. The results of the simulation also depends on the
quality of the mesh and test variables. By the variables, initial conditions and flow properties are in a
crucial role.
4.1
As it can be seen from the simulation scene, the shock locations and angles are as the same as the
numerical calculations. To compare the results for the strength the pressures on the exact solutions
are gotten from the simulation results.
19
4.2
Lift coefficient:
4.3
Drag coefficient:
Final Remarks
The location and the strength of the shocks are roughly similar with an error of 2-3o in angle deviation
and 0,5 error in the strength. However, the main topic is that the location and the strength of the
shocks are at the trailing edges and there are several expansion waves at the surface of the airfoil
due to divergent of the flow. Strength of the shock waves are high at the intake and low at the rear
edge, with this the simulation results are confirming the numerical calculations.
The lift and drag coefficients have also are similar to the numerical calculations, where the error is
mainly lower than 0,5. This also proves that the angle of attack greatly affects the lift coefficient but
the sharp edges affect the drag coefficient. This proves that the supersonic flow required sharp
edged design at the airfoil.
Even with a very thin airfoil, the expansion waves are inevitable. This wave increase Mach number by
far and requires a lot more attention for the rear end oblique shocks, where those shocks would be
much more powerful due to the increased velocity. Even with this thin airfoil setup, the expansion
waves could carry the Mach number after the first shock, all the way to the initial Mach number.
And the last concerns is the oblique shock waves and the expansion waves are forced to crash at the
far out of the airfoils, increasing the turbulences, hence increase the disturbances at the separate
parts of the machine which uses this airfoil.
20
upper
L/2
Rupper
Where;
21
Rlower
2
L/2
lower
Where;
(1.3)
5.2
Shockwave Calculations
Area_2
Area_3
Area_1
Area_5
Area_4
22
Due to the attack of angle and irregular upper and lower bodies, the airfoil will have different shocks
at different areas. The area_1 will be the initial conditions, where the area_2 and area_4 is where the
first oblique shocks happen. Area_3 and area_4 will have another couple of oblique shocks due to
the flow having different densities at the upper part and lower part, thus forming a separation line.
The separation line will have the same angle as the angle of attack.
5.2.1
Area_1 to Area_2
We ill all the o espo di g o p essio a gle 1 and the result compression will be:
Area_1
1
M1
Area_2
From this point on, the formulas in the NACA reports will be used.
Mn1
1 1
M2n
M1
5.2.2
Area_1 to Area_4
We ill all the o espo di g o p essio a gle 4 and the result compression will be:
Area_1
M1
23
4
Area_4
3rd Oblique Shock
From this point on, the formulas in the NACA reports will be used.
5.2.3
Area_2 to Area_3
Where the double edge airfoil angular turn is 2*1, thus this will result the maximum angle which M2
will turn and that is:
2=23,28
24
The oblique shock wave will happen just at the rear edge of the airfoil because of the separation line.
And the separation line has to continue form of the angle of attack with the same angle. Thus the
calculations are:
We ill all the o espo di g o p essio a gle 3 and the result compression will be:
M2
Area_2
Area_3
M*2
2
1
From this point on, the formulas in the NACA reports will be used.
1
Separation Line
From table 2, the values for M=1.11:
5.2.4
Area_4 to Area_5
Where the double edge airfoil angular turn is 2*1, thus this will result the maximum angle which M2
will turn and that is:
From table.2 we can check the corresponding angle results ate the M*4=2.63 and
pressure ratios for M4=2.16, M*4=2.63 are also:
25
2=22,35.
The
The oblique shock wave will happen just at the rear edge of the airfoil because of the separation line.
And the separation line has to continue form of the angle of attack with the same angle. Thus the
calculations are:
We ill all the o espo di g o p essio a gle 6 and the result compression will be:
Separation Line
4
From this point on, the formulas in the NACA reports will be used.
Area_4
From table 2, the values for M=1.32:
5.3
26
M*4
Area_5
5.4
Lift coefficient:
6 Simulations
6.1
Mesh Generation
The required mesh to use in the simulations was generated with the help of the program which was
already mentioned in the theory section. Gridgen v16 has the option to automatically generate the
required domain after generating connectors. The main mesh is a giant circle with the airfoil in the
middle. The left side is defined to be the inlet and the right side is the outlet. The generated mesh
has the Cartesian coordinate system where the origin is the beginning of the airfoil.
27
6.2
As discussed at the theory section, the simulation was done in Star-CCM+ in order to compare the
results with the numerical and experimental results. The following properties were selected in the
options of Star-CCM+.
Simulation Properties
Chosen Variables
Gas-Air
Ideal Gas
Implicit Unsteady
K-Epsilon Turbulence
Models
Coupled Flow
Two-Dimensional
Realizable K-Epsilon Two-Layer
RANS
Stationary
Reference Values
Initial Conditions
6.3
28
29
7 Comparison
After calculation numerically and carrying out a flow simulation in Star-CCM+ the following results
occur. First of all, numerical calculations are mainly analytical assumptions. This results for a varying
error. In simulation, the calculations are done by the Reynolds stress models which are done by
Reynolds Averaging of Navier Strokes Equations. Especially very high Reynolds number and using
RANS modeling the errors are considerably high. The results of the simulation also depends on the
quality of the mesh and test variables. By the variables, initial conditions and flow properties are in a
crucial role.
7.1
5
LOWER EXPANSION WAVES
As it can be seen from the simulation scene, the shock locations and angles are as the same as the
numerical calculations. To compare the results for the strength the pressures on the exact solutions
are gotten from the simulation results.
7.2
Lift coefficient:
7.3
Drag coefficient:
The location and the strength of the shocks are roughly similar with an error of 2-3o in angle deviation
and 0,5 error in the strength. However, the main topic is that the location and the strength of the
shocks are at the trailing edges and there are several expansion waves at the surface of the airfoil
due to divergent of the flow. Strength of the shock waves is high at the intake and low at the rear
edge, with this the simulation results are confirming the numerical calculations.
The lift and drag coefficients have also are similar to the numerical calculations, where the error is
mainly lower than 0,5. This also proves that the angle of attack greatly affects the lift coefficient but
the sharp edges affect the drag coefficient. This proves that the supersonic flow required sharp
edged design at the airfoil.
Even with a very thin airfoil, the expansion waves are inevitable. This wave increase Mach number by
far and requires a lot more attention for the rear end oblique shocks, where those shocks would be
much more powerful due to the increased velocity. Even with this thin airfoil setup, the expansion
waves could carry the Mach number after the first shock, all the way to the initial Mach number.
And the last concerns is the oblique shock waves and the expansion waves are forced to crash at the
far out of the airfoils, increasing the turbulences, hence increase the disturbances at the separate
parts of the machine which uses this airfoil.
As seen from these results, the sharp edges cause less lift. Therefore it is wiser to use biconical
airfoils. However this type of airfoils has little less drag coefficient.
31
References:
[1] [Shapiro, Ascher H. The Dynamics and Thermodynamics of Compressible Fluid Flow, Volume 1.]
[2] Chinese Journal of Aeronautics Volume 26, Issue 3, June 2013, Pages 501513
[3] The Dreese Airfoil Primer Copyright 2001-2007 John Dreese
[4] Landau, L. D., and E. M. Lifshits. Mekhanika sploshnykh sred, 2nd ed. Moscow, 1953
[5][Poinwise Products; http://www.pointwise.com/archive/faq-V13.shtml]
[6][Milo a Pe i, Futu e De elop e ts of ta -CD and Star-CCM+ , p. -27]
Tabel of Figures:
Figure 15.Double biconical airfoil in supersonic flow
Figure 16 Oblique Shock Wave
Figure 17 Attached and detached oblique shock waves [2].
Figure 18 Prandtl - Meyer Expansion waves in smooth surfaces [2]
Figure 19 Biconical airfoil shock waves
Figure 20 Drag and Lift Coefficients [3]
Figure 21 Mesh in Gridgen v16
Figure 22 Residuals of the Simulation of the Biconical Airfoil
Figure 23 Mach Number
Figure 24 Lift Coefficient
Figure 25 Drag Coefficient
Figure 26 Oblique Shock Waves and Expansion Waves
Figure 13 Double Edge Airfoil Mesh in Gridgen v16
Figure 14 Double Edge Airfoil Residuals of the Simulation of the Biconical Airfoil
Figure 15 Double Edge Airfoil Mach Number
Figure 16 Double Edge Airfoil Lift Coefficient
Figure 27 Double Edge Airfoil Drag Coefficient
Figure 28 Double Edge Airfoil Oblique Shock Waves and Expansion Waves
32