Journal of Applied Fluid Mechanics, Vol. 11, No. 4, pp. 1125-1131, 2018.

Available online at www.jafmonline.net, ISSN 1735-3572, EISSN 1735-3645.

DOI: 10.18869/acadpub.jafm.73.247.27928

Comparative Flow Field Analysis of Boundary Layer

Diverter Intake and Diverterless Supersonic Intake
Configuration
I. Arif†, S. Salamat, M. Ahmed, F. Qureshi and S. Shah
Department of Aerospace Engineering, College of Aeronautical Engineering,
National University of Sciences and Technology, H-12, Islamabad, Pakistan

†Corresponding Author Email: [email protected]

(Received April 30, 2017; accepted April 4, 2018)

ABSTRACT

In this paper comparative flow field analysis of two intake configuration i.e. Boundary Layer Diverter Intake
and Diverterless Supersonic Intake is carried out based on dimensionless parameters under various flow
conditions. Numerical analysis of aircraft intake is a complex phenomenon which involves both external and
internal flow analysis. In this research, both external and internal flow characteristics of intake duct are
analyzed in detail. A comprehensive mesh scheme is devised and implemented to accurately capture the flow
behavior in external surrounding of intake duct and flow passing through the intake duct. The analysis is
carried out at different flow conditions to analyze the flow behavior in subsonic and supersonic regimes.
Engine design mass flow rate is used for accurate intake analysis and results are validated with available
literature. Boundary layer diversion and pressure recovery are examined for each intake configuration and
comparative analysis based on pressure recovery is carried out subsequently. The analysis reveals that at
subsonic and transonic regimes, Boundary Layer Diverter intake is much more effective than Diverter less
Supersonic Intake, however, in supersonic regime Diverter less Supersonic Intake is found be to more
effective. The research can further help in modifying/ improving the design of an existing intake
configuration for enhanced intake efficiency.

Keywords: Aerodynamics; Boundary layer Diverter intake; Diverterless supersonic intake; Pressure recovery.

NOMENCLATURE

α angle of attack PR pressure recovery

DSI Diverterless Supersonic Inlet density
BLD Boundary Layer Diverter P static pressure
one equation Spalart-Allmaras turbulence
CFD Computational Fluid Dynamics S-A
model
M# Mach number RANS Reynolds Averaged Navier-Stokes
non-dimensional length scale associated
v velocity y+
with turbulence model

1. INTRODUCTION able to deliver required engine flow rate,

minimize pressure loss, drag and distortion
For a fighter aircraft the design of intake is one of (Sudhakar and Ananthkrishnan 1996, Taskinoglu
the very important factors. Intake characteristics and Knight 2002). Over the years Diverterless
directly affects the overall performance of an supersonic intake has gained significant
aircraft (Goldsmith and Seddon 1993). The importance due to its simplicity, stealth
design of a supersonic aircraft intake is one of the characteristics and effectiveness (Kim 2009).
most challenging task due to complex flow Diverterless Supersonic Inlet (DSI) has an added
characteristics such as diffusion and distortion. advantage of delivering required mass flow rates
Also, the design of intake requires efficient as compared to clean intake. The parent theory
operation over a wide range of flow conditions behind the design of bump in DSI is based on
(Paul, Kuppa et al. 2011). The intake must be Wave rider concept (Goldsmith and Seddon
I. Arif et al. / JAFM, Vol. 11, No.4, pp. 1125-1131, 2018.

1993). In boundary layer diverter intake 2. GEOMETRY AND MODELING

configuration, diverter separates the inlet from
the fuselage and the boundary layer. This design For this research, one of the most challenging task
feature not only causes added intake weight and was to develop/modify CAD geometry of aircraft
drag, but maintenance requirements are also with intake configuration under study. The
increased (Mattingly 2002). This intake type is geometrical model of aircraft with Boundary Layer
successfully designed and tested on F-16 aircraft Diverter intake (Config 1) was already available at
(Frant and Kozakiewicz 2011). Both intake Numerical Analysis Lab of the Institute. However,
configurations have their own characteristics the acquired model had number of additional
based on the engine requirements and aircraft features which were not required for this research
required maneuverability as shown in Fig. 1. and were removed such as external stores, wing
attachments, landing gears, landing gear doors,
antennas, exhaust nozzle and weapons etc. Due to
symmetry of aircraft in longitudinal axis, only right
half of the model was used which resulted in
optimal utilization of computational resources and
time. For further simplicity, only forward fuselage
area with intake was used and rear fuselage area
including vertical tail, horizontal tail, ventral fin and
exhaust nozzle were removed. CAD model of
aircraft before and after simplification are shown in
Fig. 2.
Fig. 1. Boundary layer diverter intake (left) and
diverterless supersonic intake (right)

During CFD analysis of aircraft, intakes are often

simplified due to complexities involved in flow
properties and involvement of both external and
internal flow dynamics. However, this
simplification usually affects the computed results.
This limitation necessitates the integration of
internal flow with external flow during CFD
Fig. 2. CAD model before simplification (Left)
analysis (Tu, Yeoh et al. 2012).
and after simplification (Right)
Although both intake configurations are associated
with certain pros and cons, but it is important to 3. COMPUTATIONAL SETUP
assess their performance quantitatively at similar
conditions. However, not much literature can be Computational Fluid Dynamics is a branch of fluid
found in this area due to complexities involved in mechanics which use basic conservation laws for
numerical analysis of intakes and assessing their numerical simulation. ANSYS® FLUENT solver
performance with different configurations. Hence, works on the principle of control volume approach.
great potential in literature exists in this field of For numerical analysis in this study, Reynolds-
research. For this purpose, dimensionless parameter averaged Navier-Stokes (RANS) set of equations
such as pressure recovery is selected for are used to account for time dependent behavior of
comparative analysis of intake performance on flow. RANS helps in optimum utilization of
existing aircraft configuration. computational resources by averaging the flow
For intake analysis, generally, only the forward quantities over the entire range of turbulence scale.
fuselage area of aircraft is modeled, therefore, RANS equations also cater for Reynolds stressors
same strategy is adopted in this research as well. that form an important part of flow analysis. The
Numerical analysis is performed at different flow governing conservation equations are:
speeds with varying angle of attack (AoA), zero Conservation of Mass:
side slip angle and design mass flow rate.
Validation of the results is carried out with     u     v     w 
available literature on the subject (Mattingly     (1)
2002, Ibrahim 2008). In the later part of this t x y z
research, a comparative analysis is performed
with an aircraft having Diverterless Supersonic Conservation of Momentum:
Aircraft based on dimensionless parameters at
similar flow conditions (Hassan, Masud et al.  u u u u  p
 u v w 
2015). The proposed methodology in this  t x y z  x
research is aimed to contribute in the research (2)
gap that exists in current literature as far as   2u  2u  2u 
     Fx
comparative qualitative assessment of intake  x 2 y 2 z 2 
configurations are concerned.  

1126
I. Arif et al. / JAFM, Vol. 11, No.4, pp. 1125-1131, 2018.

 v intake cowl lip, intake duct, and fuselage cone etc.

v v v 
 u v  w   In order to ensure mesh consistency, same mesh
 t x y z  strategy was adopted for both the geometries
(3) (Hassan, Masud et al. 2015). Domain size was kept
p   2v 2v  2v 
      Fy 20 times the fuselage diameter to accurately model
y  x 2 y 2 z 2  flight conditions away and in near vicinity of
  aircraft without the influence of far field (Masud
and Akram 2011). For surface meshing, ‘all tri
 w w w w  mesh’ technique with Patch independent method is
 u v w 
 t x y z  used in this work. For volume meshing Robust
(4) Octree method was used. This method actually
p  2 w 2 w 2 w  generates the volume by making layers of the
      Fz surface mesh. Turbulent y+ values were kept at
z  x 2 y 2 z 2 
  optimum level for subsonic and supersonic speeds
(Tu, Yeoh et al. 2012). A size function was also
Conservation of Energy:
applied to gradually coarsen the mesh size away
from the aircraft as shown in Fig. 3.
 T T T T 
C p  u v w Φ
 t x y z 
  T    T    T 
 k  k  k
x  x  y  y  z  z 
(5)
where, Fig. 3. Volume mesh (left) and cut plane view of
volume mesh (right)
ρ is the fluid density; µ is the kinematic viscosity;
u,v,w are the component of velocity in Cartesian 3.2 Grid Independence
coordinates; p is the pressure term; Fx, Fy, Fz are
The accuracy of results in numerical analysis
the body force terms; T is temperature in Kelvins;
largely depends upon the grid structure and mesh
and k is the heat transfer coefficient.
size. A very fine mesh can produce good results but
The numerical approach employed in the current can be computationally expensive and hence it can
study is based on the study carried out by Hassan et result in much larger computational time to achieve
al. , so that validation of results can be carried out at the desired results. Therefore, a balance must be
the later stages of this research. For numerical maintained between the mesh size and accuracy of
analysis, modeled geometry of aircraft with intake results. For said purpose, a grid independence test
duct was imported in ANSYS ICEM CFD ® for was carried out in order to select an optimized mesh
mesh generation. Semi spherical domain was made size which produce accurate results and is
around the aircraft and symmetry plane was used in computationally suitable as well. Four different
longitudinal axis. Mesh consistency was given prime meshes were generated based on number of cells.
importance for both types of intakes / configurations All four meshes were generated with different
to conduct comparative analysis at a later stage in this surface mesh size and size function. Pressure
research. Unstructured meshing scheme was used for Recovery at M # 0.6 and AoA 00 for each mesh was
both geometries in this research. This step was evaluated for comparative analysis. A brief
followed by numerical simulations in Fluent® summary is also presented in following Table 1 and
software. Analysis was carried out at different flow results are shown in Fig. 4.
conditions to analyze the flow behavior and intake
rsonic1. Grid independence analysis
performance in subsonic and supersonic regime. Grid Cells
Simulations were carried out at zero side slip angle Grid 1 4.1 million
and various angle of attack. Flow characteristics such Grid 2 8.3 million
as boundary layer diversion and pressure recovery are Grid 3 10.2 million
examined for each configuration. For comparative Grid 4 12.4 million
analysis of both types of intakes pressure recovery
values were compared.
3.1 Grid Generation
Grid generation is one of the most important factors
in numerical simulation. A number of public
domain and commercial mesh generation softwares
are available. For this research, ANSYS ICEM
CFD® was used for meshing, as it is industry
standard and can produce high quality hexa and
hybrid meshes. A very fine mesh could be
computationally expensive (Liu, Pekkan et al. Fig. 4. Grid independence
2004). Hence, in this work, a crafted grid was used,
where in mesh was kept fine at critical areas such as From Grid independence study, it was observed that
values of Pressure Recovery were almost identical

1127
I. Arif et al. / JAFM, Vol. 11, No.4, pp. 1125-1131, 2018.

for Grid 3 (10.2 million) and Grid 4 (12.4 million), results depicted a sinusoidal behavior and did not
whereas for Grid 1 and Grid 2, the values of converge at a single value. Hence, its final value
Pressure Recovery varied significantly. Based on was obtained by averaging the values of last 500
these results, a Grid size of 10.2 million cells was
selected for accurate results and optimum
computational efficiency.
3.3 Boundary Conditions
Aircraft and intake geometries were created
separately. These were required to be imported in
ANSYS Fluent for numerical analysis. Firstly,
meshed aircraft geometry and domain were
imported in Fluent and subsequently intake duct,
having the same global coordinates, was also
imported using ‘append’ command. The common
face of aircraft and intake duct geometries was
defined as ‘interface’. ‘Pressure outlet’ condition Fig. 5. Turbulence model independence
was used at engine inlet plane to control the engine
mass flow rate. Aircraft surfaces were treated as iterations. Also, its residuals remained above 1x10-
wall with no slip condition. Flow conditions were 4, whereas all other models had their residuals well
controlled by changing M # and Angle of attack below 1x10-6 and depicted better convergence
(AoA) with Pressure Far Field boundary condition rates. Based on these results, SA turbulence model
at the domain surfaces. Symmetry plane was was selected for further analysis. Also, Spalart-
defined as Symmetry boundary condition. Air is Allmaras (S-A) model is specially designed for
used as ideal gas throughout the analysis. aerospace applications involving wall-bounded
flows, as was the case in this research (Spalart and
3.4 Analysis Strategy Allmaras 1992). This RANS model solves transport
Numerical analysis is based on Reynolds-Averaged equation for modified eddy viscosity, and is
Navier-Stokes (RANS) equations to predict the therefore, not only computation less intensive, but,
flow behavior with time dependency. Density based the in the modified form, eddy viscosity is easy to
solver was chosen for present research and explicit resolve near the wall. A single solution took an
technique was implemented in Fluent. Flow average time of three days to stabilize on a high-end
discretization was selected as ‘2nd order upwind’, work station (sixteen core CPU with 32 Gigabytes
whereas ‘1st order upwind’ scheme was used for of RAM).
turbulent viscosity (Hassan, Masud et al. 2015).
Default Relaxation parameters were modified to 4. RESULTS AND DISCUSSION
stabilize the iterative process. Courant number, for
density based explicit solver, was set at its default In the present study, simulations was carried at two
value of “1”. To ensure the stability of the solution, subsonic flight conditions (M # 0.6 and 0.8) and one
Aircraft and ‘intake duct’ were initially simulated supersonic flight condition (M # 1.5). Furthermore,
separately in Fluent. Once the individual solution at each M #, two AoA (0º and 4º) were analyzed.
were stabilized, two geometries (aircraft and ‘intake Therefore, a total of six simulations were
duct’) were integrated using ‘append’ feature in performed. All cases were performed at engine
Fluent, and a combined case was simulated and design mass flow rate for each condition and no
results were obtained after attaining pre-defined sideslip angles. Flow behavior was analyzed
convergence criterion. qualitatively and compared for both intake
configurations. Qualitative analysis reveals flow
3.5 Turbulence Model Independence entrainment pattern into the intake and thus better
Turbulence Model selection also affects the intake configuration can be decided. Later, intake
accuracy of numerical solution obtained from CFD performance is calculated in the form of pressure
significantly, and hence careful selection of same recovery and comparative analysis is carried out
cannot be overlooked. For the purpose of this study, between both Config 1 and Config 2 for detailed
three different turbulence models were selected study.
(SA. k-epsilon and k-omega) based on the fact that
the primary physical phenomenon to be captured is
4.1 Configuration 1- Boundary Layer
flow inside intake duct. SA is a single equation Diverter (BLD) Intake
turbulence model while k-epsilon and k-omega are F-16 aircraft intake has a significant advantage due
two equation turbulence models. Comparative to location (under belly) with splitter plate which
results for these models are presented in Fig. 5 provides shielding and avoids boundary layer
below: suction at different flight regimes as shown in Fig.
It was observed that except Standard k-epsilon, all 6. The intake shape and curvature to F-16 duct has a
other models (SA and K-omega) were consistent in direct effect on pressure recovery.
calculated Pressure Recovery. In addition to this, it Flow field characteristics of Config 1 are presented
was also observed that K-omega model in detail. To analyze the flow behavior inside the
convergence was also a matter of concern as its intake duct, planes were made inside the duct and

1128
I. Arif et al. / JAFM, Vol. 11, No.4, pp. 1125-1131, 2018.

total pressure contours are plotted on these plane of intake duct. It was observed that centerline
using post processing tools in Fluent. The total curvature had a direct effect on flow as compared to
pressure contours plots at M # 0.6, 0.8 and 1.5 are cross sectional dimensions. The characteristics of
shown in Fig. 7 below. boundary layer in the direction of duct was directly
affected by local pressure changes. At the sharp
curvature of duct, the kinetic energy of flow
becomes high and results in large centrifugal forces.
Hence, the major cause of secondary flow is due to
pressure gradients involved. Low pressure region is
visible at duct plane edges at M # 1.5. With increase
in angle of attack from 0 degree to 4 degrees, it can
be observed that high pressure region is slightly
Fig. 6. Velocity streamlines diffused. The high pressure region is evident at the
centre of plane and slightly diffuses towards the
edges. The fuselage area before intake duct causes
the growth in low total pressure area at lower side
of intake and causes the velocity to decrease. In all
cases, with an increase in AoA, the twin swirls
strength decreases which exhibit the optimum
design feature of fuselage as flow straightener.
The performance of aircraft intake is usually gauged
in terms of pressure recovery, which is the ratio of
total pressure at engine inlet to the free stream total
pressure. Intakes are designed maximize the
pressure recovery and great efforts have been made
to minimize the pressure loss due to friction, shock
Fig. 7a. Pressure contour at intake duct exit waves and shock boundary layer interaction
plane at M # 0.6, AoA=00 (Whitford 1987, Mattingly 2002). Since pressure
recovery is a dimensionless parameter, it was
feasible to perform the comparative analysis based
on this parameter. The values of total pressure were
extracted from Fluent software at required areas at
different flow conditions. Comparative analysis was
performed at three different M # and two AoAs at
engine mass flow rate. The calculated values of
pressure recovery for Config 1 are shown in Table 2
and Table 3.
Table 2 Pressure recovery at AoA 00
Pengine Pengine /
M# ∞ (Pa)
(Pa) ∞
0.6 129240 123888 0.960
Fig. 7b. Pressure contour at Intake duct exit
0.8 154453 150790 0.976
plane at M # 0.8, AoA=00
1.5 371967 349649 0.940

Table 3 Pressure recovery at AoA 40

Pengine /
M# ∞ (Pa) Pengine (Pa)
∞
0.6 129240 124281 0.961
0.8 154453 149482.6 0.967
1.5 371967 345431 0.928

From the calculated results, it can be observed that

at a particular AoA, as the velocity is increased in
subsonic regime pressure recovery is increased.
Therefore pressure recovery is highest at M # 0.8 at
same AoA. Pressure recovery tends to decrease at
Fig. 7c. Pressure contour at intake duct exit supersonic speed at all conditions. This is due to the
plane at M # 1.5, AoA=00 fact that the shock waves at the inlet in supersonic
condition causes additional pressure loss and hence
From the figures, it is evident that the flow is
it results in lower pressure recovery as compared
largely affected by the duct geometry. The low
(Goldsmith and Seddon 1993, Mattingly 2002).
boundary layer flow does not disperse quickly due
This phenomena is quite similar in fixed intakes
to static pressure gradients and results in vortex
(Ibrahim, Ng et al. 2011). It is also evident that the
formation. At M # 0.8 and 1.5, twin swirl flow is
pressure recovery of BLD (Config 1) is quite
prominent at outlet plane due to symmetric nature

1129
I. Arif et al. / JAFM, Vol. 11, No.4, pp. 1125-1131, 2018.

similar for both AoA at M # 0.6, however, at high recovery values for Config 2 (DSI) has been
subsonic and supersonic speeds pressure recovery at published by Hassan et al. (Hassan, Masud et al.
AOA 0o is slightly higher than that at AOA 4º. This 2015) and same are being used in this research.
is due to the fact that the high velocity core moves
towards the bottom half of the intake exit plane with
the increase in AoA. Also, with the increase in
AoA, the flow distortion and instability have
pronounced effect on total pressure (Saha, Singh et
al. 2007). The variation in pressure recovery at
varying M # for different AoA are shown in Fig. 8.
The average percentage difference in pressure
recovery at high subsonic and supersonic speeds at
AoA 00 and 40 is observed to be 1% only.

Fig. 10. Pressure recovery comparison of config

1 and 2 at AoA 00
From the above graphs, it is evident that both intake
configurations have almost same pressure recovery
at moderate subsonic speeds of M # 0.6. As the
speed increases, in subsonic regime, Config 1 gives
better pressure recovery than Config 2. This is due
to the fact that boundary layer diverter
configuration is quite effective in low subsonic
Fig. 8. Pressure recovery vs M # regime due to its shape and location. BLD design
prevents ingestion of boundary layer inside the
The results obtained from numerical analysis were intake and thus increases intake performance.
validated from available data in literature of same Although, DSI configuration also diverts boundary
aircraft configuration (Mattingly 2002, Ibrahim layer but its effectiveness is less as compared to
2008). Pressure recovery obtained from CFD BLD. As speed is further increased to supersonic
analysis and literature are shown in Fig. 9. speeds, the pressure recovery of both configurations
is reduced due to formation of shock waves
(Goldsmith and Seddon 1993). In supersonic
regime, DSI shows better performance than
boundary layer diverter intake in terms of pressure
recovery. The pressure downstream of shock wave
formed on the bump surface has positive pressure
gradient which is responsible for boundary layer
diverting away from aircraft intake. At supersonic
speed for Config 2, the low energy boundary layer
is diverted away from the intake behind the shock
wave. Also, at design mass flow rate the oblique
shock structure for DSI configuration is similar to
design point shock structure which is not observed
in BLD configuration. Hence, DSI configuration
Fig. 9. Pressure recovery comparison

From comparative analysis, it can be observed that

the variation in pressure recovery is quite similar
except at M # 0.6 of Ref (Ibrahim, 2008) where
pressure recovery is high. Also, the pressure recovery
calculated in this work is slightly higher than other
analysis. This is due to the fact that average design
mass flow rate was used for each condition from
available literature (Frant and Kozakiewicz 2011)
which may differ from the mass flow rate used in
other analysis. However, the overall trend line and
results are quite satisfactory with available literature.
4.2 Comparative Analysis with
Configuration 2 (DSI) Fig. 11. Pressure recovery comparison of config
1 and 2 at AoA 40
In the next step, a comparative analysis of BLD
intake configuration is made with DSI at similar has better pressure recovery characteristics than
flight conditions and design mass flow rate for each boundary layer diverter intake configuration at high
configuration as shown in Fig. 10. Pressure supersonic speeds at design mass flow rate. At AoA

1130
I. Arif et al. / JAFM, Vol. 11, No.4, pp. 1125-1131, 2018.

of 40, a trend similar to AoA 00 is observed, Composite CFD. AIAA.

however, there is slight difference in the breakeven
point of pressure recovery in both configurations as Ibrahim, I., E. Ng and K. Wong (2011). Flight
shown in Fig. 11. Config 2 pressure recovery equals Maneuverability Characteristics of the F-16
Config 1 pressure recovery at M # 0.9 as compared CFD and Correlation with its Intake Total
to M # 1.04 at AoA 00 (numerical simulation details Pressure Recovery and Distortion. Engineering
not presented in this research). This is due to the Applications of Computational Fluid Mechanics
fact that the variation in pressure recovery of Config 5(2): 223-234.
2 with change in AoA is slightly less than that of Ibrahim, I. H. (2008). Fluid flow studies of the F-5E
Config 1. Hence, DSI performance is less affected and F-16 inlet ducts.
by variation in pitching motion of aircraft and
maintains optimum mass flow rate for better engine Kim, S. D. (2009). Aerodynamic design of a
performance. supersonic inlet with a parametric bump.
Journal of Aircraft 46(1): 198-202.
5. CONCLUSION Liu, Y., K. Pekkan, S. C. Jones and A. P.
Yoganathan (2004). The effects of different
In this work flow field and performance analysis of mesh generation methods on computational
BLD intake configuration is carried out. The results fluid dynamic analysis and power loss
were validated with available literature and a assessment in total cavopulmonary connection.
comparative analysis was carried out with an Journal of Biomechanical Engineering 126(5):
aircraft with DSI configuration. The analysis was 594-603.
carried out at different M # and angle of attack to
Masud, J. and F. Akram (2011). Flow field and
analyze the flow behavior in subsonic and
performance analysis of an integrated
supersonic regimes. The results revealed that BLD
diverterless supersonic inlet. Aeronautical
intake configuration is more effective in subsonic
Journal 115(1170): 471.
regime as compared to DSI configuration, whereas
at supersonic speeds DSI configurations gave Mattingly, J. D. (2002). Aircraft engine design,
superior performance. However, it may be noted AIAA.
that the simulations were carried out discretely at
zero side slip angle and at a particular height. For Paul, A. R., K. Kuppa, M. S. Yadav and U. Dutta
complete comparison of the two configurations, (2011). Flow improvement in rectangular air
more simulations at different numerical setups are intake by submerged vortex generators. Journal
under study by the College of Aeronautical of Applied Fluid Mechanics 4(2): 77-86.
Engineering (CAE) research group. Saha, K., S. Singh and V. Seshadri (2007). Effect of
angle of attack on performance of twin intake
ACKNOWLEDGMENTS duct. 37th AIAA Fluid Dynamics Conference
and Exhibition.
The authors acknowledge the use of Numerical Spalart, P. R. and S. R. Allmaras (1992). A one
Analysis Lab (NAL) of College of Aeronautical equation turbulence model for aerodinamic
Engineering, Risalpur, Pakistan. flows. AIAA journal 94.

REFERENCES Sudhakar, K. and N. Ananthkrishnan (1996). Jump

phenomena in Y-shaped intake ducts. Journal of
Frant, M. and A. Kozakiewicz (2011). Construction aircraft 33(2): 438-439.
of an air intake system model for F-100-PW-
229 engine in F-16 aircraft for intake vortex Taskinoglu, E. S. and D. Knight (2002). Numerical
development analysis. Prace Instytutu analysis of submerged inlets. AIAA Paper 3147.
Lotnictwa. 39-49. Tu, J., G. H. Yeoh and C. Liu (2012).
Goldsmith, E. L. and J. Seddon (1993). Practical Computational fluid dynamics: a practical
intake aerodynamic design, Amer Inst of approach, Butterworth-Heinemann.
Aeronautics &. Whitford, R. (1987). Design for air combat, Janes
Hassan, S., J. Masud and O. Khan (2015). Intake Information Group.
and Airframe Characterization through

1131