Proceedings of 8th International Research Conference, KDU, Published November 2015
Aerodynamics Analysis of F-16 Aircraft
HAM Peiris, PVS Nirmal#, HMHHS Bandara, DM Mahindarathne,
SLMD Rangajeeva and RMPS Bandara
General Sir John Kotelawala Defence University, Ratmalana, Sri Lanka
#
[email protected]Abstract— The Lockheed Martin F-16 Fighting Falcon is operative 4th Generation multi-role fighter aircraft
the world’s most prolific fighter with more than 2000 in (Martin, 2015). The main goal of the designers of F-16
service. The primary objective of this research project is to was to create a simpler with greater manoeuvrable
evaluate the aerodynamics behaviour of the F-16 aircraft fighter aircraft of the time. Their concepts challenged by
by conducting Computational Fluid Dynamics (CFD) the shape and the way it should fly. Finally they
analysis. The CFD simulations have been done in both accomplished with the results of: a level blended-wing
subsonic and supersonic flight regimes. As a means of
body with extra lift and control, a fly-by-wire system that
validating the results, the CFD analysis has been done
kept the design stable, with better response time and
with two different turbulence models. A 1:1 CATIA solid
model of F-16 aircraft was used to generate increased its agility and an enhanced cockpit – including a
computational mesh and subsequent CFD simulations tilted back ejection seat, side-mounted throttle and
were performed mainly with Fluent ANSYS. During the control stick, head-up display and bubble canopy with
post processing phase of the CFD results, the improved pilot survivability, visibility and control.
aerodynamic characteristics of the F-16 have been
The Fighting Falcon is the cumulative results of F-16
predicted in terms of lift coefficient and drag coefficient
pilots’ combat experience and built on the primary
over angles of attack ranges from 0° to 40°. In comparing
CFD predictions between turbulence models, a minimal strengths of the original Fighting Falcon design. With the
variation of those dimensionless quantities was recorded. time passes introducing new technologies into the
At Mach number 0.6, formation of two large leading edge cockpit, avionics, sensors and weapons, the aircraft has
vortices which is the subsonic lift generation mechanism, become more reliable, more maintainable and more
were observed on the main wings. A complete analysis of supportable (Martin, 2015).
shock waves and expansion fans formation around the
aircraft was also performed at supersonic speed, by B. Computational Fluid Dynamics
examining the static pressure variation. Further to Computational Fluid Dynamics (CFD) is the process of
examine flow over main wings, the surface static pressure using computers to simulate realistic flows. In computer
variation and pressure coefficient variation at different form, the geometry of aircraft designs can be readily
span-wise locations have been also studied. The flow defined and modified. CFD deals with the aerodynamicist
physics revealed with CFD analysis are well aligned with a means of exploring variety of aircraft shapes than can
both subsonic and supersonic theories. The forecasted usually be achieved, in available time scales. For the
values for aerodynamic efficacy and dimensionless simulation the type of model should dependent on the
parameters are lower than expected. It has been found accuracy needed, the computer power accessible and the
this particular fact is directly related to computational time scale to accomplish the analysis. The central point of
limitations associated with CFD. The outcomes from this CFD problems is the Navier–Stokes equations which
piece of research not only provides better sight to fine includes continuity, momentum and energy.
details about fluid dynamics in relation to F-16, but also
made vital recommendations for future CFD analysis of F-
C. Turbulence Models
16 aircraft.
Turbulence models are generally classified according to
which governing equations they apply. Furthermore they
Keywords— CFD, F-16 Aircraft, Turbulence Models
are classified by the number of additional transport
equations which one must solve in order to compute the
I. INTRODUCTION model contributions (Clelik, 1999).
A. F16 Aircraft
One class of the turbulence models is the Reynolds-
With the technological advancement and 138 different
Averaged Navier-Stokes (RANS) model, and is used for
configurations, the famed F-16 is the world’s largely
69
Proceedings of 8th International Research Conference, KDU, Published November 2015
most production applications. Further this is categorized Table 1 - Mesh Features
according to their usage of wall functions, the number of
additional variables solved for and what these variables
Cells Faces Nodes Partitions Cell Face
represent. All of these models augment the Navier- Zones Zones
Stokes equations with an additional turbulent viscosity 792571 2415081 830701 1 1 8
term (Jurij SODJA, 2007). But they differ in how it is
computed. During the analysis process to resolve the
details of the turbulent fluctuations, the solutions are
focused on two turbulence models namely;
a. Standard k-epsilon (Ɛ) model
b. Shear Stress Transport (SST) k-omega (ω) model
II. METHODOLOGY
A. Solid Modeling
The solid model was created using CATIA V5R18. The
model was basically drawn with the aid of CATIA V5R16.
Initially the different schematics views of F-16 Aircraft
were downloaded from the web. Three Extrude surfaces
were crated while offsetting from XY, YZ and ZX planes.
Then the reference pictures were applied onto these
three surfaces. Sketches were drawn to each and every Figure 1 - Generated Mesh
cross section while relocating them into the correct
positions. The Body, then the wing and finally the tail
were created with using 3D curves and “Freeform”
surfaces. C. Cfd Simulation
The simulation of continuum was done in ANSYS Fluent
B. Mesh Generation 14. In this initially the meshing of the continuum was
Second process was to develop a mesh as referred in read and then checked. Once the software approves it,
Figure 1. In order to obtain valid and accurate results the scale was selected in mm as the model was created
creating a fine mesh is important. Mesh was created and meshed using the unit mm. Then the models,
using the software OpenFOAM with the use of materials and boundary conditions were set.
“snappyHexMesh” commanding. Before creating the
aircraft mesh the domain mesh was successfully created 1) turbulence Models:
with the use of command “blockMesh”. Modifications The turbulence models used for these simulations were
were created to define three refinement boxes which the the k-ε model and k-ω SST. The solver was based on
first was covering whole aircraft, the second was covering “density based” as the all of the simulations were
the tail part of aircraft, and the third was covering the compressible. Energy equations were selected in order to
both wings. But to accommodate a mesh with more solve the cases as all of the simulations were
accuracy 5 layers were developed near to the aircraft compressible.
surface. Another development had been made to create
another block after the model till the end of domain 2) Materials:
where the flow is having its most critical phenomena The working fluid in this simulation was ideal gas as the
such as turbulence, wakes and flow separation. boundary condition “Pressure Far field” was compactible
with it. It was considered the F-16 aircraft was flying at
The number of nodes points and cells are as referred in sea level conditions and the viscosity was solved using
the Table 1. Sutherland equations.
3) Boundary Conditions:
Pressure far-field boundary conditions were used in this
simulation to model a free-stream compressible Flow at
infinity, with free-stream Mach number and static
70
Proceedings of 8th International Research Conference, KDU, Published November 2015
conditions specified. The aircraft was given with “wall” Figure 3. The dominant aspects of this flow are two
boundary condition. vortex patterns that occur in the vicinity of the highly
swept leading edges. These vortex patterns are created
4) Solution: by the mechanism of the pressure on the bottom surface
The solver was semi-implicit method for pressure-linked of the wing is higher than the pressure on the top
equations (SIMPLE) algorithm. This algorithm is an surface. Thus, the flow on the bottom surface in the
iterative procedure for solving equations for velocity and locality tries to curl around the leading edge from the
pressure, for steady-state. The courant number is set to 2 bottom to the top.
and the under relaxation factors for momentum and
pressure are set as 0.5 and for the turbulent kinetic
energy, turbulent dissipation rate and turbulent viscosity
is set to 0.5. For the discretization the pressure was kept
as standard, while the other parameters Momentum,
Turbulent Kinetic Energy, Turbulent dissipation Rate and
Energy were retained as Second Order Upwind.
Monitoring the convergence during the solution was
dynamically checked by force coefficient values rather
than checking for the convergence through residuals. The
data were printed, reported and displayed in plots of lift,
drag, and moment coefficients, and residuals for the
solution variables. In the Force Monitors, the Force
vectors Lift and Drag had to define with relative to the
free stream direction.
Figure 2 - Static pressure distribution over wing
5) Reference Values: span
When calculating force coefficients the reference values
should be given to get the actual results. So the
Reference Values of area, length, pressure, density,
temperature and velocities were given as 27.87m2,
14.23 m, 101325 Pa, 1.225 kgm-3, 288.16 K, 204ms-1 for
subsonic (Mach 0.6) and 408ms-1 for supersonic (Mach
1.2) respectively.
7) Iterate:
The numbers of iterations were set to 10000 to be
performed in the Number of Iterations field. Then the
FLUENT began with the calculations starting at iteration
1, using the initial solution. Then the graphs were
plotted, printed and written (in separate data files). Figure 3 - Velocity distribution over wing span
When the observations in the graphs of Coefficient of Lift The leading edge vortices are strong and stable. The
Vs iterations and Coefficient of Drag Vs iterations were local static pressure of the vortices is considerably
converged, the simulations were stopped. dropped due to the high vorticity flow. The leading edge
vortices are figuratively creating a strong “suction” on the
III. RESULTS AND DISCUSSION top surface near the leading edges due to the pressure
A. Formation Of Wing Tip Vortices At M=0.6 difference. The suction effect of the leading edge vortices
At Mach number 0.6, formation of two large leading edge increase the normal force which will enhances the lift.
vortices which is the subsonic lift generation mechanism, For this reason, the lift coefficient for a delta wing
were observed on the main wings. exhibits an increase in CL for values of high angles of
The subsonic (M=0.6) flow pattern over the top of a delta attack at which conventional wing planforms would be
wing in F-16 aircraft at AOA=40° is as for the Figure 2 and stalled.
71
Proceedings of 8th International Research Conference, KDU, Published November 2015
Since delta -winged aircraft are performing in supersonic
speeds in their missions; the aircraft has to perform in
subsonic speeds in their landing and takeoff. So the low-
E
speed aerodynamics of delta wings for F-16 aircraft is C
C D
having a pronounced importance in lift generation B
mechanism at subsonic speeds (Anderson, 2013). A
B. Shock Waves And Expansion Fans Around
Aircraft At Mach=1.2
F
Figure 5- Pressure Distribution along axis
of symmetry
The upswing of Mach number due to the expansion fan
occurred in the upper surface of the aircraft body,
another shockwave was arisen. So this oblique shock
wave was formed in front of the vertical stabilizer on the
body with the body, symbolized with the location “E” in
Figure 5, signify a rise in static pressure while having a
reduction in static temperature and velocity as for
referenced figures respectively in order to retard the flow
Figure 4 - Velocity distribution along axis of
to obtain free stream conditions (Anderson, 2013).
symmetry
Starting from nose and along the X- axis in the upper
surface of the F-16 aircraft, three shock waves (as
denoted with the locations “A”, “B” and “C” in Figure 5)
were created with a clear variation of static pressure
reduction, velocity and static temperature rise as for the
Figure 5, Figure 4 and Figure 6 respectively. Again with
ability to flow over a convex angle with the aircraft
represented by location “D” as showed in the Figure 5 a
series of expansion waves were found with satisfying the
isentropic conditions. The properties of expansion waves
are in a contradictory way to the flow properties of shock
waves. The results can be interpreted as rise in velocity
magnitude and reductions in static temperature as well
as the static pressure as shown in the Figure 4, Figure 6
and Figure 5 respectively.
Figure 6 - Static temperature distribution along
axis of symmetry
72
Proceedings of 8th International Research Conference, KDU, Published November 2015
C. Lift Curve, Drag Curve, Drag Polar And Moment become more and more one time the lift generate will
Curve not sufficient to bare the self-weight of the aircraft and
1) Lift Curve: will stall. At this time the coefficient of drag will be at the
maximum.
3) Drag Polar
Drag Polar at M=0.6 for turbulence models of k-ε and k-ω
SST as for the Figure 11. Here no clear difference
between two graphs can be observed. The graph is
according to the Equation (1), where CD,0 the parasite
drag coefficient at zero lift. Value represents the
parasite drag and induced drag due to lift. And also “e” is
referred to the Oswald Efficiency Factor and “AR” is
aspect ratio.
= + (1)
Figure 8- Coefficient of Lift Vs AOA
When considering the graph referred in Figure 8, it could
identify that the flow over the upper surface of the airfoil
is still not separated and it is in attached with the airfoil
surface. Though the CFD gave a high AOA for the CL max
in practical the flight control system of the F-16 limits it
up to 25.5˚. It was because to reduce the high g force on
the fighter pilots at their manoeuvres this was achieved
by the Flight Control System installed in the F-16 aircraft.
2) Drag Curve:
Figure 10- Drag polar
A comparison of drag polar at M=0.6 and M=1.2 for
different turbulence models as for the Figure 11. With
the increasing Mach number a distinct variation can be
identified in the CD,0 value at zero lift. The drag polar
curves also have identical difference in different Mach
numbers.
4) Lift To Drag Ratio Versus Angle Of Attack
The lift to drag ratio is measure of aerodynamic efficiency
Figure 9- Coefficient of Drag Vs AOA where the F-16 requires minimum thrust to operate. The
maximum value for the at M = 1.2 is smaller than
When consider the normal shape of the graphs we could that of the M = 0.6. It could be due to the high drag
produced at high speeds. So the ultimate value for the
see same change in graph of Figure 9. The value at zero
angle of attack is little bit stationary and it is because for value is getting smaller and this maximum value is
lower angles of attack the flow will not separate much obtained at AOA= 13˚.The results obtained for the values
over the airfoil. But when the angle of attack increases of maximum aerodynamic efficiencies are for the Table 2.
the flow will slightly separate from the upper surface of
the airfoil. There the present of wakes are prominent
than in lower angle of attacks. When the separation
73
Proceedings of 8th International Research Conference, KDU, Published November 2015
Table 2-Maximum aerodynamic efficiency values supersonic flight regimes with two different turbulence
Turbulence Models Maximum aerodynamic efficiency models. During the post processing phase of the CFD
results, the aerodynamic characteristics of the F-16 have
M=1.2 M=0.6
been predicted in terms of lift coefficient and drag
k-ε 3.15 4.2
coefficient over angles of attack ranges from 0° to 40°. In
k-ω SST 3.2 4.35 comparing CFD predictions between turbulence models,
a minimal variation of those dimensionless quantities was
recorded. At Mach number 0.6, formation of two large
D. Flow Properties Over Different Span Wise leading edge vortices which is the subsonic lift generation
Locations On The Wing mechanism, were observed on the main wings. The
Velocity magnitude and static pressure variations over suction effect of the leading edge vortices enhances the
different span wise locations on the wing have been lift coefficient for a delta wing exhibits for values of high
observed. From the axis of symmetry, the span wise angles of attack at which conventional wing planforms
locations 2m, 3m, and 4m on the wing were taken.
would be stalled. A series of shock waves in the upper
surface as well as in the lower surface of the aircraft was
observed with the drastic changes in flow properties such
as rise in static pressure, reductions in flow velocity and
rise in temperature. While expansion fans formation
around the aircraft was also observed at supersonic
speed, by examining the contradictory flow property
changes to the shock waves such as rise in velocity,
reductions in static pressure and temperature variation.
Further the research was examined the flow over main
Figure 12- Velocity magnitude wings, the surface static pressure variation and pressure
coefficient variation at different span-wise locations. Due
to the formation of shock waves with the static pressure
rise, a higher adverse pressure gradient has been
observed. The flow physics revealed with CFD analysis
are well aligned with both subsonic and supersonic
theories. The forecasted values for aerodynamic efficacy
and dimensionless parameters are lower than expected.
It has been found this particular fact is directly related to
Figure 13- Pressure magnitude computational limitations associated with CFD.
With the distractions created with the whole aircraft REFFERENCES
body, the particular flow properties over the wing were Anderson, J. D., 2013. Fundamentals of Aerodynamics.
unable to recognize properly. But with the distance from 5th ed. New York: Mc Graw Hill.
the axis of symmetry as for the Figure 12 and Figure 13
the flow properties on the wing was clearly detected. Anderson, J. D., 2013. Introduction to Flight. 6th ed. New
Shock waves on the leading edge of the wing were York: Mc Graw-Hill.
determined with the reduction of velocity magnitude and
a rise of static pressure as highlighted in the Figure 12 Clelik, B. I., 1999. Introductory Turbulence Modeling,
and Figure 13. But comparing to the velocity variation, a Morgantown: West Virginia University.
slight variation of static pressure is observed.
Jurij SODJA, R. P., 2007. Turbulence Models in CFD, s.l.:
IV. COCLUSION s.n.
This research project was evaluated the aerodynamics Martin, L., 2015. F-16 Fighting Falcon. [Online]
behavior of the F-16 aircraft by conducting Available at: http://usfighter.tripod.com/f16.htm
Computational Fluid Dynamics (CFD) analysis while [Accessed 03 08 2014].
performing the CFD simulations in both subsonic and
74
