Download as pdf or txt
Download as pdf or txt
You are on page 1of 14

JOURNAL OF AIRCRAFT

Vol. 56, No. 2, March–April 2019

Effect of Side Gust on Performance of External Compression


Supersonic Inlet

Hussein K. Halwas∗ and Suresh Aggarwal†


University of Illinois at Chicago, Chicago, Illinois 60607
DOI: 10.2514/1.C035093
This Paper reports a computational investigation on the effects of side gust on the performance of a double-cone
external compression supersonic inlet at Mach 1.8. The supersonic inlet geometry comprises three zones: a supersonic
zone, a transit zone, and a subsonic zone. The supersonic zone is designed numerically using the Taylor–Maccoll
method, while the transit and subsonic zones are designed using a methodology from literature. A three-dimensional
structured mesh is generated by using the ICEM computational fluid dynamics software. Detailed three-dimensional
simulations are performed using the ANSYS Fluent 18.2 code. The grid independency tests are performed by using
local grid refinements and varying the number of grid points from 2.5 to 9.5 million. Further, the Fluent code is
validated by comparing predictions with those using the SUPIN and Wind-US codes. The effects of side gust on the
Downloaded by Cranfield University on January 23, 2024 | http://arc.aiaa.org | DOI: 10.2514/1.C035093

flowfield and oblique and normal shocks are analyzed by examining the Mach number and pressure distributions in
the three zones. The inlet performance is characterized in terms of total pressure loss, flow distortion, and mass flow
ratio parameters for a side gust of 56 m∕s and gust angles of 30, 60, and 90 deg. Results indicate that the inlet
performance is significantly affected by the side gust, especially at small gust angles.

Nomenclature weak normal shock at the throat, which is located inside the inlet,
A = area, m2 followed by a subsonic diffuser [4]. The type of inlet selected depends
M = Mach number on the cruise flight Mach number. For cruise Mach number 2.0 or
m
_ = mass flow rate, kg/s smaller, an external compression inlet is more suitable due to high-
Po = total pressure, Pa pressure recovery and stable operating characteristics [5–7]. As
p = static pressure Pa discussed by Oswatitsch [1], the large decrease in stagnation pressure
T = temperature, K across a normal shock may be reduced by decelerating the supersonic
V = velocity magnitude, m/s flows to a subsonic flow via one or more external oblique shocks,
α = side gust angle, deg followed by a weak normal shock at the throat. The duct in such inlets
may be of rectangular, cylindrical, or conical shape with a fixed or a
Subscripts variable geometry depending on the compression surface such as a
changeable angle two-dimensional (2D) ramp or conical spike that
avg = average can move inside that duct. The inlet flow comprises two main zones.
i = inlet The first is the supersonic zone, which consists of one or more
max = maximum oblique shocks followed by weak or strong normal shock depending
min = minimum on its location in the inlet.‡ The second is the subsonic zone that has a
y = condition immediately behind normal shock wave diverging section or a diffuser, in which the subsonic flow is further
∞ = freestream condition slowed down to a speed as required by the engine; see, for example,
Fig. 1 taken from Ref. [8]. In between the two zones, there is also an
intermediate or blending zone extending from the cowl lip to the
I. Introduction subsonic diffuser. The purpose of this zone is to make the flow after
the normal shock turn back in the axial direction as it enters the
S UPERSONIC inlets have been developed since World War II, in
parallel with the development of jet engines and supersonic
flights. While supersonic inlets have many different designs, most of
diffuser [9,10]. The throat must be designed to avoid the chocked
flow at all flight conditions [11].
them can be grouped into three types, namely, external compression, The basic function of a supersonic inlet is to provide the correct
internal compression, and a combination of both. In the first design, quantity and quality of airflow to the compressor. The correct
the supersonic compression occurs externally on a compression quantity of air needs to be delivered to the engine generally at about
surface (a wedge or a cone) ahead of a cowl lip, followed by a normal Mach 0.4 ∼ 0.6, with minimum flow distortions and as small a loss in
shock near the lip and further compression via a subsonic diffuser total pressure Po as possible [7,12]. Thus, the inlet delivers airflow to
[1,2]. In the second type, the supersonic compression occurs inside the engine at a static pressure greater than the ambient and thereby
the inlet (converging–diverging duct) followed by a weak normal contributes to the compression process, as shown in Fig. 2 from
shock at the throat and compression in a subsonic diffuser [3]. In the Ref. [13]. The total pressure loss is minimized by creating a series of
mixed type, the supersonic compression occurs externally upstream oblique shock waves in front of the supersonic inlet, followed by a
of the cowl lip and then internally downstream of it, followed by a weak normal shock at the throat in the cowl lip [14]. The location of
the normal shock is controlled by the backpressure at the compressor
Received 15 May 2018; revision received 13 August 2018; accepted for inlet. At the design backpressure, a weak normal shock is located at
publication 24 August 2018; published online 24 November 2018. Copyright the throat in the cowl lip. If the backpressure is too low, the normal
© 2018 by the American Institute of Aeronautics and Astronautics, Inc. shock moves downstream into the diffuser to maintain the same mass
All rights reserved. All requests for copying and permission to reprint flow rate and thus becomes stronger, resulting in greater total
should be submitted to CCC at www.copyright.com; employ the eISSN pressure loss. Moreover, the interaction of this stronger shock with
1533-3868 to initiate your request. See also AIAA Rights and Permissions the boundary layer causes flow distortions. On the other hand, if the
www.aiaa.org/randp. backpressure is too high, the normal shock moves upstream and
*Ph.D. Student, Department of Mechanical and Industrial Engineering,
842 West Taylor Street; hhalwa2@uic.edu.
† ‡
Professor, Department of Mechanical and Industrial Engineering, 842 For the design or cruise condition, there is a weak normal shock located at
West Taylor Street; ska@uic.edu. the cowl lip or the throat.
569
570 HALWAS AND AGGARWAL

Because of the high costs associated with experiments at supersonic


conditions, the computational fluid dynamics (CFD)-based tools
provide an attractive approach for analyzing the performance of
supersonic inlets and provide guidelines during the design process.
Connors and Meyer [22] presented design charts for Mach numbers up
to 4 for single-oblique, double-oblique, and conical-shock inlets and
also for isentropic two-dimensional and axisymmetric compression
surfaces. They used the Taylor–Maccoll method [23,24] to optimize
geometric angles for the single- and double-cone inlets in term of total
Fig. 1 Schematic of a double-cone external-compression supersonic
pressure recovery. Chen and Caughey [25] employed a relaxation
inlet. method to compute inviscid supersonic flow in an axisymmetric inlet
with a centerbody. Several researchers have employed analytical
methods to study flows in supersonic inlets. Bangent et al. [26]
examined analytically the effect of the Mach number in a supersonic
inlet, while Shimabukuro et al. [27] analyzed an inlet engine
combination for Mach 2.2.
Various CFD approaches have been employed based on solving
either Euler equations or Navier–Stokes equations. Biringen [28]
computed inviscid flowfield in a two-dimensional inlet using the
Euler equations. Bradley et al. [29] used an artificial density method
Downloaded by Cranfield University on January 23, 2024 | http://arc.aiaa.org | DOI: 10.2514/1.C035093

for solving inviscid equations for steady supersonic conical flows


and reported results for circular and elliptic cones at different Mach
numbers and angles of attack. Jain [30] numerically examined the
startup problem associated with the mixed compression supersonic
inlets by solving 2D compressible Euler equations and suggested
that the starting problem can be addressed by either increasing the
throat area (variable geometry) or increasing the freestream Mach
number (fixed geometry). Lavante and Thompkins [31] computed
supersonic diffuser flow by solving the Navier–Stokes equations.
Walters et al. [32] solved the compressible Navier–Stokes equations
in conservative form and analyzed flow in a supersonic inlet. Knight
Fig. 2 Supersonic gas turbine cycle. [33] also employed Navier–Stokes equations for simulating
flows in two-dimensional supersonic inlets. Geetha et al. [34] used
Lagrange multipliers for a maximum total pressure recovery
possibly outside the cowl. Again, this results in a stronger normal method to optimize the turning angles in a scramjet engine inlet.
shock, and the flow becomes completely subsonic inside the inlet. Several inlet design codes have also been developed to provide
Consequently, the mass flow rate decreases as some of the flow spills analysis of supersonic inlet performance based on analytical,
over the cowl lip, and the flow distortion increases, causing increased empirical, and low-fidelity computational methods. Examples
drag. Operation at an off-design Mach number leads to consequences include 1) SUPIN, developed to compute flows for external-
compression-type inlets for Mach 1.6–2.1 [35]; 2) the U.S. Navy
similar to those at off-design backpressure [15].
Several experimental investigations dealing with the performance of NIDA code [36]; and 3) Inlet Performance Analysis Code [37].
Wind-US CFD Code [38] is a high-fidelity CFD code developed at
supersonic inlets have been reported. Oswatitsch [1] reported a series
NASA for simulating the external and internal super-/subsonic
of pioneering analytical and experimental studies using axisymmetric
flows. In addition, the Fluent commercial code has been used to
external compression supersonic inlets for missiles. He considered
compute a viscous flowfield in an axisymmetric mixed compression
multistage spikes to minimize the total pressure loss. Brajnikoff [16]
supersonic inlet [39]. Different inlet flow operations (critical,
experimentally analyzed an annular inlet with a ramp for a range of
supercritical, and subcritical) were predicted correctly, as the cowl
Mach numbers and observed significant improvement in total pressure inner wall static pressure distribution was in good agreement with
recovery, as much as 80% of the loss associated with a normal shock the experimental data.
inlet with the same freestream Mach number. Ferri and Nucci [17] The objective of this Paper is to numerically investigate the effect
performed an experimental investigation of axisymmetric supersonic of side gust on the performance of a supersonic inlet. Three-
inlets and compared the performance of external and internal dimensional (3D) simulations are performed using the ANSYS
supersonic compression in terms of the external drag and total pressure Fluent 18.2 commercial code. The performance is analyzed in terms
recovery. In another study, they [2] demonstrated the suitability of of the total pressure recovery, mass flow rate, and flow distortion for
axisymmetric fixed geometry external compression inlets for different gust parameters. The geometric model considered is an
supersonic airplanes and missiles for a range of Mach numbers and axisymmetric double-cone supersonic inlet. The cone angles are
mass flow rates. Connors and Meyer [18] reported an experimental optimized on the basis of total pressure recovery by using the Taylor–
investigation on an axisymmetric external compression supersonic Maccoll method with the numerical code in QBasic language. The
inlet, focusing on the overall performance in terms of the total pressure transit, or blending, zone, which includes the cowl, inner surface, and
recovery, mass flow rate, and external drag. McGregor [19] examined subsonic zone, is designed by following Refs. [9,10].
the performance of a rectangular double ramp compression inlet and Gust is one of the most common and important weather
suggested that the concept of breaking an external shock system could phenomena for both subsonic and supersonic flights. It is particularly
be extended to any desired number of steps with a considerable total of common occurrence in the troposphere (6–10 km) and at higher
pressure recovery. Tindell [20] investigated the drag and stability altitudes, where a majority of airplanes fly [40]. Gusts may arise due
aspects for semiconical and two-dimensional inlets and observed that to several weather phenomena, including hurricanes, thunderstorms,
the former inlet type has better stability but higher drag. Soltani et al. high winds, tornadoes, etc. They have played an important role in
[21] reported an experimental study on the performance of a mixed many catastrophic airplane accidents during the last century. The
compression supersonic inlet in terms of the total pressure recovery, effects of gust on these airplanes are important from both structural
mass flow rate, and flow distortion at different Mach numbers with and and aerodynamic (performance) considerations. However, relatively
without boundary layer bleeding. The presence of a bleed system was few studies have focused on the latter [41], although the structural
found to improve the intake performance. aspects have been extensively investigated. Moreover, while the
HALWAS AND AGGARWAL 571

previous experimental and computational studies have examined presents the shock pressure recovery as a function of two semi-cone
various aspects of supersonic inlets, the effects of gust on the flow angle is half of the cone angle, at the design Mach number of
characteristics and inlet performance have not been examined. The M∞  1.8. The objective of this figure was to optimize the
present work is motivated by this consideration. The gust effects are supersonic section design, including the two cone angles and the
characterized in terms of gust speed and direction. A gust speed of cowl lip angle. The total pressure loss here includes only losses that
56 m∕s (126 mph) was used in the present study. This represents one- occur across the two oblique shocks and one normal shock at the lip.
half of the strongest gust speed of 408 km∕h, recorded officially Thus, it does not include losses in the internal contraction region,
during tropical cyclone Olivia in Australia on 10 April 1996 [42]. viscous losses, and those due to shock/boundary-layer interaction.
There have been several higher gust speeds; however, they have not As indicated in the figure, the highest total pressure recovery
been recorded officially. To analyze the effect of gust direction, (0.984) is achieved using the first and second semi-cone angles of
simulations were performed with three gust angles of 30, 60, δC1  20 deg and δC2  28 deg, measured from the horizontal
and 90 deg. centerline, respectively. The internal cowl lip surface was designed
This Paper is organized as follows. In the next section, we describe to have the initial incline parallel to the local flow direction.
the design of the inlet, including the supersonic, transit, and subsonic For the optimum performance at the design condition, both the
parts. Section III discusses the numerical method introducing the oblique shock waves should impinge at the cowl lip [4,5,10,11].
governing equations, solution methodology, grid generation, Thus, the intersection point of the two oblique shocks should be
boundary conditions, performance parameters, grid dependency, the cowl lip itself, which is also the location of the normal shock at
and optimum grid and code validation. Results and discussion are the design condition. It is also important to note that, while the
presented in Sec. IV, while the conclusions are provided in Sec. V. first shock is conical and the flowfield behind it is conical and
irrotational (isentropic), the second shock is curved and the
Downloaded by Cranfield University on January 23, 2024 | http://arc.aiaa.org | DOI: 10.2514/1.C035093

flowfield behind it is rotational (nonisentropic). Therefore, to have


II. Design of Inlet Geometry the intersection point of the two shocks at the cowl lip, we need to
The supersonic inlet considered in the present Paper comprises construct a curved shock at the second cone. A satisfactory
three sections: supersonic, transit, and subsonic. The supersonic approximate method, proposed in Ref. [44], is based on the
section consists of two cones. The geometric angles of these cones assumption of a constant flow deflection through each point of the
and their locations are designed by using the Taylor–Maccoll method shock. The deflection angle is again equal to the difference between
[23,24] and optimized to minimize the total pressure loss, calculated the second and first cone semi-angles. Therefore, using a stepwise
based on the shock losses without considering the viscous losses. In calculation procedure starting from the cone surface and using the
this method, the Euler equations (continuity and momentum) in oblique shock equations, the curved shock can be constructed. An
spherical coordinates are transformed to ordinary differential alternative procedure was proposed by Kennedy [45], who used the
equations, which are solved numerically by using a fourth-order conical flow theory in the irrotational region and the method of the
Range–Kutta method. The simulations were performed by using a characteristic in the rotational region to construct the curved shock
code written in the QBasic language. The shock system includes two and find the shocks intersection point.
oblique shocks and one normal shock. The use of the Taylor–Maccoll The transit section is the region where the supersonic flow converts
method involves the assumption that the flow is inviscid and conical. to subsonic flow via a normal shock. This section starts at the cowl lip
The conical flow implies that the flow properties are constant and forms the annular duct of the inlet along with the subsonic
along the conical compression surfaces originating at the cone diffuser. The interior cowl surface angle should be initially aligned
vertex [43]. The field Mach number of the first cone is considered to with the local flow direction in order to make the exterior shock at the
be the average Mach number in the region behind the first oblique cowl lip weaker [9,22]. The exterior cowl surface angle should be
shock. This is different from that used in Refs. [22,44], in which it is greater than the interior angle by 3–5 deg for structural considerations
taken as the average of the Mach numbers immediately behind the [4,5,9]. The last section of the inlet is the subsonic diffuser located
oblique shock and at the cone surface. The total pressure loss across downstream of the transit section, where the flow is subsonic. The
the second oblique shock is then calculated using the assumption design data for this part have been taken from a previous study [9].
that flow undergoes a deflection angle equal to the difference It connects the annulus cross section at the end of the transit zone with
between the second and the first cone half-angles. Similar to the the compressor face. The compressor face has a radius of 53.34 cm
procedure used for the first cone, the average Mach number of the and a coannular shape with a hub-to-tip ratio of 0.2. As stated in
second cone can be determined, and then the additional total Ref. [9], the hub has an elliptical cross-section with a radius
pressure loss at the normal shock can be computed. Figure 3 (semiminor axis) of 10.67 cm and a length (semimajor axis) of
15.24 cm. The cowl planer profile is designed based on the area
distribution through the subsonic diffuser. The planar profiles
surfaces are then extruded about the axis of symmetry to create the
surfaces of the subsonic duct. Figures 4 and 5 show the geometry of
the supersonic inlet design based on the procedure described
previously, while Table 1 lists the geometric data corresponding to the
design conditions.

III. Numerical Method


A. Governing Equations
The 3D flow simulations of the complete supersonic/subsonic flow
in the inlet are based on the compressible form of the Navier–Stokes
equations (mass, momentum, and energy equations), along with the
equation of state. The turbulence is modeled using Menter’s two-
equation tab (shear stress transport) k − ω turbulent model [46].
As discussed in the cited study, it employs the combination of the
k − ω mode in the near-wall region and k − ε model in the outer
region and connects them by blending functions. The model
represents a significant improvement over the standard two-equation
Fig. 3 Total pressure recovery for the double-cone inlet design for turbulence model, especially for the prediction of adverse pressure
M  1.8. gradient flows.
572 HALWAS AND AGGARWAL

Fig. 4 Geometry of the upper half of the supersonic inlet model (numbers are in centimeters).
Downloaded by Cranfield University on January 23, 2024 | http://arc.aiaa.org | DOI: 10.2514/1.C035093

2 3 2 3 2 3
ρ ρu ρv
6 7 6 7 6 7
6 ρu 7 6 ρu2  p 7 6 ρuv 7
6 7 6 7 6 7
6 7 6 7 6 2 7
6 ρv 7 6 ρuv 7 6 ρv  p 7
6 7 6 7 6 7
6 7 6 7 6 7
q6
6 ρw 7; E  6 ρuw 7;
7 6 7 F6
6 ρvw 7;
7
6 7 6 7 6 7
6 e 7 6 ue  p 7 6 ve  p 7
6 7 6 7 6 7
6 7 6 7 6 7
6 ρk 7 6 ρuk 7 6 ρvk 7
4 5 4 5 4 5
ρω ρuω ρvω
2 3
ρw
6 7
6 ρuw 7
6 7
6 7
6 ρvw 7
6 7
Fig. 5 The full geometry of the supersonic inlet model. 6 7
G6
6 ρw 2p 7
7 (2)
6 7
6 we  p 7
The compact form of 3D Navier–Stokes equations along with the 6 7
two-equation shear stress transport k − ω turbulence model can be 6 7
6 ρwk 7
written as 4 5
ρwω

∂q ∂E ∂F ∂G ∂Ev ∂Fv ∂Gv


      S (1)
∂t ∂x ∂y ∂z ∂x ∂y ∂z
2 3 2 3
where q is the conservative variable vector; E, F, and G are the 0 0
6 7 6 7
inviscid flux vectors; Ev , Fv , and Gv are viscous flux vectors in 6 τxx 7 6 τyx 7
6 7 6 7
the three special directions; and S represents the source terms in the 6 7 6 7
6 τ 7 6 τ 7
turbulence model. These vectors can be written as 6 xy 7 6 yy 7
6 7 6 7
6 τ 7 6 τ 7
Ev  6
6
xz 7;
7 Fv  6
6
yz 7;
7
Table 1 Inlet geometric data for the design conditions 6 uτxx  vτxy  wτxz 7 6 uτyx  vτyy  wτyz 7
6   7 6   7
6 7 6 7
Parameter Value 6 μ  μt ∂k 7 6 μ  μt ∂k 7
6 σ k ∂x 7 6 σ k ∂y 7
Frist cone semi-angle, deg 20 6   7 6   7
4 μt ∂ω
5 4 μt ∂ω
5
Second cone semi-angle, deg 28 μ  σω ∂x μ  σ ω ∂y
Length of the first cone, cm 37.89
Overall length of on-design supersonic part, cm 60.67 2 3
Interior surface cowl initial angle, deg 13 0
6 7
Exterior surface cowl initial angle, deg 17 6 τzx 7
6 7
Inlet radius, cm 53.15 6 7
Outlet radius (engine face radius), cm 53.34 6 τ 7
6 zy 7
Engine spinner radius (semiminor axis), cm 10.67 6 7
6 τ 7
Gv  6 7
Overall axial length of the engine spinner (major axis), cm 30.48 zz
Overall length of the supersonic inlet, cm 220.29 6 7 (3)
6 uτzx  vτzy  wτzz 7
On-design freestream Mach number 1.8 6   7
6 7
Air model Ideal gas 6 μt ∂k
μ  σk ∂z 7
Altitude (H), m 13716 6 7
6   7
Mass flow rate (m),
_ kg∕s 154.71 4 5
Back to freestream static pressure ratio (pout ∕p∞ ) 4.5 μ  σμωt ∂ω∂z
HALWAS AND AGGARWAL 573

p 
The effective stress tensor is given by k 500μ
φ2  max 2 ; (17)
   0.09ωy ρy2 ω
∂ui ∂uj 2 ∂u 2
τij  μt  − δij m − δij ρk (4)
∂xj ∂xi 3 ∂xm 3 where F1 and F2 are the blending functions and D
ω is the positive
portion of the cross-diffusion term
The source term vector is given by
α∞  F1 α∞;1  1 − F1 α∞;2 (18)
2 3
0
6 7 βi;1 k2
6 0 7
6 7 α∞;1  − p (19)
6 0 7 β∞ σ ω;1 β∞
6 7
S6 0 7 (5)
6 7
6 0 7
6 P − ρβ kω 7 βi;2 k2
4 k 5 α∞;2  − p (20)

β∞ σ ω;2 β∞
α ωk Pk − ρβω2

The coefficient α in the production of ω is


The turbulent kinetic energy production rate is given by
 
      α∞ αo  Ret ∕Rω
1 ∂ui ∂uj 2 2 ∂um 2 2 ∂u α (21)
Pk  μ t  − − ρk m α 1  Ret ∕Rω
Downloaded by Cranfield University on January 23, 2024 | http://arc.aiaa.org | DOI: 10.2514/1.C035093

(6)
2 ∂xj ∂xi 3 ∂xm 3 ∂xm
Rω  2.95α (22)
The turbulent viscosity μt is computed as
The model constants are listed in Table 2.
ρk 1
μt  (7)
ω max1∕α ; SF2 ∕α1 ω
B. Solution Methodology
Simulations have been performed using the ANSYS Fluent 18.2
where α is a turbulent viscosity damping coefficient, calculated as
code. After designing the inlet geometry as described in a preceding
  section, AutoCAD 2016 was used to draw the scale model, which was
αo  Ret ∕Rk
α  α∞ (8) then exported in a format readable by grid-generation software ICEM
1  Ret ∕Rk CFD. The grid generation is briefly described in the next section. In
the Fluent solver, the governing equations are discretized using a
ρk finite-volume approach. Thus, the computational domain is divided
Ret  (9) into discrete control volumes using a general curvilinear grid. The
μω code was executed using the parallel solver on a desktop workstation
with four CPUs. The postprocessing and plotting of results were done
βi using the ANSYS package. Note that the Fluent code is very sensitive
αo  (10) to the initial flowfield used to obtain the converged solution. To
3
obtain a converged solution for a new case, the flow needs to be
Turbulent Prandtl numbers σ k and σ ω for k and ω, respectively, are initialized close to the stagnant state. Thus, the assumed initial
solution corresponds to a uniform flowfield with a small velocity, and
1 pressure and temperature corresponding to the ambient at the
σk  (11) assumed altitude. The energy equation is not solved during the first
F1 ∕σ k;1  1 − F1 ∕σ k;2
few iterations, about 50 or so. In addition, the underrelaxation factors
conditions of 0.3 and 0.4 are used for solving the discretized
1 momentum and pressure equations, respectively. As the solution gets
σω  (12) stable after a few iterations, the energy equation is turned on. An
F1 ∕σ ω;1  1 − F1 ∕σ ω;2
underrelaxation factor of 1 is used for the energy equation, while the
underrelaxation factor for the pressure equation is changed to 0.7. For
F1  tanhφ41  (13) a new simulation case, a converged solution requires about 50,000
iterations based on values of residuals being 10−5 to 10−7 depending
  p   upon the flow variable.
k 500μ 4ρk
φ1  min max ; 2 ; (14)
0.09ωy ρy ω σ ω;2 D
ωy
2 C. Grid Generation
A three-dimensional structured multiblock mesh was generated for
  the flow domain using the ICEM CFD software, which is available as
1 1 ∂k ∂ω a part of the ANSYS Fluent package. The mesh was refined near the
D
ω  max 2ρ ; 10−10 (15)
σ ω;2 ω ∂kj ∂xj walls to capture the turbulent boundary layer so that the first grid
point was away from the wall at distance of y ≈ 1 as required by the
selected turbulent model. The mesh was also refined in the regions
F2  tanhφ22  (16) where the shocks were expected to exist, such as the first and second

Table 2 Turbulence model constants


α∞ 1.0 α∞ 0.52 αo 0.112 β∞ 0.09 βi 0.072
Rβ 8.0 Rk 6.0 Rω 2.95 σk 2.0 σω 2.0
σ k;1 1.176 σ ω;1 2.0 σ k;2 1.0 σ ω;2 1.168 α1 0.31
βi;1 0.075 βi;2 0.0828 —— —— —— —— —— ——
574 HALWAS AND AGGARWAL

cone zones, from the transit zone to the beginning of the subsonic Table 3 Comparison of
zone. The computational domain consists of ten blocks, with 6.5 ANSYS Fluent, Wind-US, and
million nodes. Results dealing with the grid dependency and SUPIN results at the design
condition (critical operation)
optimum mesh are discussed in Sec. III.F. Figure 6 shows the three-
dimensional computational domain and its vertical grid plane. Method MFR TPR
ANSYS fluent 0.9735 0.9569
CFD (Wind-US) 0.9704 0.9654
SUPIN 0.9800 0.9663

Table 4 Freestream conditions for the four side gust cases studied
Case V gust α, Effective Net flow Effective
study ,m∕s deg V ∞ , m∕s angle, deg M∞ Po∞ , Pa
1 0 0 508.84 0 1.800 112380.7
2 56 30 558.04 2.876 1.972 146612.2
3 56 60 536.84 5.162 1.906 132323.2
4 56 90 511.92 6.280 1.810 114172.8
Downloaded by Cranfield University on January 23, 2024 | http://arc.aiaa.org | DOI: 10.2514/1.C035093

D. Boundary Conditions
The far-field boundary condition is applied to the inlet section of
the supersonic inlet. The pressure and temperature at the far field are
calculated using the following equations:

101.325
p∞  (23)
10H∕19200

T ∞  288.16–0.0065H (24)

Thus, p∞  19558.77 Pa and T ∞  199.006 K correspond to


the atmospheric pressure and temperature, respectively, at an altitude
of 13,716 m. The static pressure for flow outside the inlet is specified
as the far-field static pressure, while that at the outlet is specified as
Fig. 6 Computational grid and domain. 4.5 times the far-field pressure in order to have the normal shock at the
Mass flow rate through outer boundary

0.958 0.35

0.956 0.30

0.954 0.25

0.952 0.20
TPR

0.950 0.15

0.948 0.10

0.946 0.05

0.944 0.00
2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0
Grid number (million) Grid number (million)
a) TPR grid test b) Mass flow rate through outer boundary
0.980

0.978

0.976
MFR

0.974

0.972

0.970

0.968
2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0
Grid number (million)
c) MFR at exit section
Fig. 7 Grid refinement results and optimum mesh.
HALWAS AND AGGARWAL 575
Downloaded by Cranfield University on January 23, 2024 | http://arc.aiaa.org | DOI: 10.2514/1.C035093

Fig. 8 Mach number contours for the flow in double-cone inlet for four side gust cases (α  0–90 deg).

Fig. 9 3D view of Mach number contours for the four side gust cases as shown in Fig. 8.
576 HALWAS AND AGGARWAL
Downloaded by Cranfield University on January 23, 2024 | http://arc.aiaa.org | DOI: 10.2514/1.C035093

Fig. 10 Static pressure contours along two representative streamlines for the flow in the double-cone inlet for four side gust cases (α  0–90 deg).

cowl lip. No-slip flow and adiabatic thermal conditions are specified weighted average. An inlet with good performance implies as high as
at all the inlet walls. possible values of TPR and MFR and as low as possible a value of FD.

E. Inlet Performance Parameters F. Grid-Dependency and Optimum Grid


The inlet performance is characterized in terms of the following To examine grid dependency and determine an optimum mesh,
parameters: simulations were performed for eight different cases using grid
1) Total pressure recovery (TPR) is the ratio of total pressure at the numbers of 2.5, 3.5, 4.5, 5.5, 6.5, 7.5, 8.5, and 9.5 million. A
inlet exit (compressor engine face) to the freestream total pressure. structured mesh was used in the entire domain. The computational
The total pressure loss is the sum of shock and viscous losses. The domain was divided into nine blocks, which were connected to each
total pressure at the exit is calculated by the area-weighted average, other via their faces. Any change in the number of grid points in
the axial or radial direction in a given any block will lead to changes
Poexit in the blocks with common faces. The local grid refinement was
TPR  (25)
Po∞ implemented in the regions of high gradients as the number of grids
was increased. Clearly, a coarse mesh yielded poor results, including
2) The mass flow ratio (MFR) is the ratio of the actual inlet mass incorrect shock locations, while a very highly refined mesh entailed
flow rate to the maximum captured mass flow rate. high computational cost. Results for different grids were compared
in terms of TPR, the mass flow rate through the outer boundary,§
m_i A and the MFR at the exit section. These three parameters are shown in
MFR   i (26)
m
_ ∞ A∞ Figs. 7a–7c, respectively, plotted with respect to the number of grids.
As indicated, computational results seem to have become nearly grid
3) The flow distortion parameter (FD) measures the flow independent for grid points of 6.5 million and higher. The mass
uniformity at the exit and is defined in terms of the total pressure flow rate through the outer boundary provides a good measure of the
variation at the exit (i.e., the compressor face), correct normal shock location, which would yield a value equal to
approximately 0, since no flow should go through the top section. As
Po max − Po min indicated in Fig. 7b, this value became 0 for 6.5 million grid points
FD  (27)
Poavg and higher. Note that for all the cases the MFR at the exit section was
kept nearly constant by adjusting the backpressure so as to bring the
Note that the average total pressure in Eq. (27) is the same as the
§
exit total pressure in Eq. (25), and both are calculated using the area- The outer boundary is shown in Fig. 6.
HALWAS AND AGGARWAL 577

normal shock to the cowl lip. Results (not shown) further indicated (≈1%) than those obtained with the Wind-US and SUPIN codes. The
that the oblique and normal shock over the external part and cowl lip difference can be attributed to the fact that Slater used a three-stage
were not captured precisely for cases with less than 6.5 million grids. cone in the external compression section, while a two-stage inlet is
Thus, in the present study, detailed simulations were performed with employed in the present study.
6.5 million nodes. In terms of the number of grid points in various
sections, there were 90,248 structured grid points in the symmetry
plane of the inlet, 444 grid points distributed axially between the first
IV. Results and Discussion
cone apex and the compressor face, and 130 grid points distributed
vertically from the centerline to the cowl lip. The grid points were Results are now presented from 3D simulations using the Fluent
clustered near the walls (such that y ≈ 1) and in regions where the code for four different gust cases as outlined in Table 4. The base case
shocks were expected to exist, which included the first and second corresponds to the case with no side gust, while the other three cases
cones and the transit zone that contained the cowl lip and throat. The correspond to the side gust speed of 126 mph and gust angles of
inlet symmetry plane was extruded by rotation about the central axis α  30, 60, and 90 deg.
using 72 layers with 5 deg intervals. Figures 8 and 9 present the Mach number contours for the four
cases. While these two figures show essentially the same results,
Fig. 9 provides a 3D perspective and shows Mach number
G. Code Validation distribution in the exterior part of the inlet. Results indicate that the
In the absence of experiment data, computational results obtained simulations capture all the important shock and flow characteristics
using the ANSYS Fluent code were compared with those reported by both outside and inside the inlet. As expected, for the base case, the
Slater [9], who performed simulations using the SUPIN and Wind- shock waves and flowfields in the upper and lower parts of the inlet
Downloaded by Cranfield University on January 23, 2024 | http://arc.aiaa.org | DOI: 10.2514/1.C035093

US codes for an external-compression supersonic inlet at a critical are symmetric with respect to the longitudinal axis. The major
operating condition with an inflow Mach number of 1.8 and zero features captured by the simulations include 1) two oblique shocks
angle of attack. Table 3 summarizes this comparison in terms of two impinging close to the cowl lip, 2) normal shock at the cowl lip,
performance parameters, namely, MFR and TPR, indicating good 3) expansion waves after the normal shock followed by a small
agreement between the results obtained with the three codes. The normal shock bringing the flow to subsonic, and 4) subsonic flow
computed TPR value using the ANSYS Fluent code is slightly lower including the separated boundary layer in the diffusor section. In

Fig. 11 Static pressure variation along the upper and lower streamlines, as shown in Fig. 10, for the four side gust cases.
578 HALWAS AND AGGARWAL

addition, contours indicate that the second oblique shock is curved flow details, especially concerning the effect of side gust. The
due to the conical flow over the compression surface. pressure contours for the base case again indicate the symmetry of
The Mach number contours for the three gust cases (Figs. 8b–8d) flowfields with respect to the longitudinal axis. As expected, the
clearly show that the shock wave and flow characteristics are strongly static pressure increases as the flow moves through the supersonic
affected by the presence of gust. The effect seems to be the strongest zone, which contains the oblique and normal shocks, and then
for the gust angle of 30 deg. As indicated in Table 4, the change in through the subsonic zone, where the subsonic diffuser works on
the freestream Mach number caused by gust is also the largest for increasing the pressure as the flow approaches the outlet section.
this case. An obvious effect of side gust is that the flow becomes More quantitative information about the pressure distribution is
significantly asymmetric with respect to the longitudinal axis. As provided in Fig. 11, which plots the pressure variation along the two
indicated, the oblique shock waves in the lower part become stronger streamlines mentioned earlier. For the base case (Fig. 11a), the
compared to those in the upper part. This results in a stronger normal pressure variations along the two streamlines are identical due to flow
shock in the upper part due to the higher Mach number there. symmetry. Here, the first two pressure jumps are due to the two
Consequently, the interaction of stronger normal shock with the oblique shock waves. The gradual pressure rise in between the two
boundary layer causes boundary-layer separation to occur earlier in oblique shocks is due to the conical (irrotational) flow in that region.
the upper part. However, the separated flow region is larger in the The third sharp jump occurs at the normal shock located at the cowl
lower part due to more pressure in upper part, which is discussed lip, followed by sudden drop and rise in pressure, indicating the
further in the next figures. Also, due to the side gust, the oblique presence of expansion waves and an additional normal shock located
shock waves get located upstream of the cowl lip, and as a result, the at the tail of the expansion wave (Fig. 9a); finally, the gradual pressure
normal shock is expelled outside the cowl, especially in the lower part rise occurs in the subsonic diffuser section.
of the inlet. The asymmetric shock structures cause considerable Figures 10b–10d along with Figs. 11b–11d provide further details
Downloaded by Cranfield University on January 23, 2024 | http://arc.aiaa.org | DOI: 10.2514/1.C035093

nonuniformity in the flowfield. These features are further discussed about the effect of side gust on the shock wave and flow
by examining pressure distribution for the various side gust cases. characteristics. The pressure contours and pressure variation along
Figure 10 presents the static pressure contours for the four the two streamlines clearly depict the asymmetry caused by the
simulated cases and complements the information provided by the presence of side gust. As discussed earlier in the context of Fig. 8, an
Mach number contours in Figs. 8 and 9. Figure 10 also shows two obvious effect of side gust is that the oblique shocks are stronger in
representative streamlines that pass through the shock waves and the lower part compared to those in the upper part. Moreover, the
diffuser section in the upper and lower parts and provide additional asymmetry in the oblique shock strength increases as the side gust

Fig. 12 Streamlines for the flow in the double-cone inlet for four side gust cases (α  0–90 deg).
HALWAS AND AGGARWAL 579
Downloaded by Cranfield University on January 23, 2024 | http://arc.aiaa.org | DOI: 10.2514/1.C035093

Fig. 13 Total pressure contours (normalized) at the outlet section of the double-cone inlet for four side gust cases (α  0–90 deg).

Fig. 14 Mach number contours at the outlet section of the double-cone inlet for four side gust cases (α  0–90 deg).
580 HALWAS AND AGGARWAL

55
angle is increased. This can be clearly seen in Figs. 11b–11d, which
show greater pressure peaks along the lower streamline as the flow 50
passes the two oblique shocks. This leads to a significantly stronger
normal shock in the upper part due to the higher Mach number there. 45
Another important effect of side gust is the series of expansion and 40
compression waves generated due to the interaction of normal shock

FD (%)
with the separated boundary-layer flow inside the duct, especially in 35
the upper part. These waves are clearly depicted in Figs. 10b–10d and 30
Figs. 11b–11d. The plots in these figures also indicate that this effect
of side gust is stronger at the side gust angle of 30 deg compared to 25
that at higher gust angles. Moreover, the flow asymmetry caused by
20
side gust can also be observed by the two streamlines in Figs. 11b–
11d. As the gust angle is increased, the upper streamline is located 15
closer to the duct wall compared to the lower streamline. Another
10
effect of side gust is the asymmetry in static pressure distribution in
the diffuser section, with the pressure being higher in the upper part 0 30 60 90
compared to that in the lower part. This generates radial flow in the Side gust angle (deg)
diffuser section, with some streamlines going from the upper part to Fig. 17 FD variation with the side gust angle.
the lower part, as shown in Fig. 12, which plots streamlines for the
four side gust cases. This also increases the size of the separation zone
Downloaded by Cranfield University on January 23, 2024 | http://arc.aiaa.org | DOI: 10.2514/1.C035093

in the lower part compared to that in the upper part. This asymmetry the outlet as a function of the side gust angle. As indicated, the TPR
in the size of the separated flow zone can be seen more clearly in values are the highest and lowest for α  0 and 30 deg, respectively.
Figs. 13 and 14, which are discussed next. Figure 16 plots the variation of the MFR with the side gust angle. It is
Figures 13 and 14 present the total pressure (normalized) and interesting to note that the MFR value for the base case is lower than
Mach number contours, respectively, at the outlet section for the four those for the two side gust cases with α  30 and 60 deg. This is due
side gust angles. For the base case, the symmetry in the flowfield is to the fact that for the base case the intersection point of the first and
quite evident in both the total pressure and Mach number contours. second shock waves is not at the cowl lip but upstream of it, which
As a consequence, the total pressure loss is the minimum for the base causes some flow to spill over the cowl, and therefore the MFR is
case, while it is the highest for the side gust case of α  30 deg. This less than 1. In contrast, for the gust angle of α  30 deg, the effective
is shown more clearly in Fig. 15, which plots the total pressure ratio at freestream Mach number becomes high enough to push the
intersection point downstream of the cowl lip, and, consequently,
the MFR reaches its maximum value for this case. For larger angle
1.00 gust cases (α  60 and 90 deg), the increase in the gust angle
increases the asymmetry in shocks structures, as discussed earlier. As
a consequence, the intersection points are generally outside the cowl
0.95 lip, which leads to a lower MFR, especially for the case with
α  90 deg. Note that for the last case (α  90 deg) the high side
0.90
gust angle not only increases the shocks asymmetry but also
decreases the freestream Mach number to a value that causes the
TPR

normal shock wave to move outside.


0.85 Figure 17 shows the variation in the FD with respect to the side gust
angle. As expected, the FD has the lowest value for the base case
(α  0 deg), indicating the highest flow uniformity, while it has the
0.80 highest value for the side gust case with α  30 deg and decreases
subsequently as the gust angle is increased. This is consistent with the
results discussed earlier regarding the asymmetry in shock wave
0.75 characteristics caused by the side gust. This leads to asymmetry and
0 30 60 90 nonuniformity in Mach number and pressure distribution inside the
Side gust angle (deg) duct. The asymmetry and flow nonuniformity are the highest for
Fig. 15 TPR variation with the side gust angle. α  30 deg, as discussed earlier.

V. Conclusions
1.00 The effect of side gust on the performance of a double-cone
external-compression supersonic inlet has been numerically
0.98 investigated. The inlet geometry has been designed and optimized
by using a combination of the Taylor–Maccoll method and existing
procedures. A 3D structured mesh with 6.5 million nodes has been
0.96 generated, and 3D simulations have been performed for four side gust
MFR

cases using the ANSYS Fluent 18.2 code. Results of the grid-
refinement study and code validation are presented. The inlet
0.94
performance has been characterized in terms of three parameters,
namely, the total pressure recovery, mass flow ratio, and flow
0.92 distortion. Important observations are as follows:
1) For all the side gust cases, simulations capture the important
shock and flow characteristics both outside and inside the inlet. For
0.90 the base case (without side gust), the shock waves and flowfields are
0 30 60 90 symmetric with respect to the longitudinal axis, and the results
Side gust angle (deg) indicate two oblique shocks originating from the compression
Fig. 16 MFR variation with the side gust angle. surface and impinging close to the cowl lip, a normal shock at the
HALWAS AND AGGARWAL 581

cowl lip, followed by expansion waves and a small normal shock, and [16] Brajnikoff, G. B., “Pressure Recovery at Supersonic Speeds through
finally subsonic flow in the diffusor section. Annular Duct Inlets Situated in a Region of Appreciable Boundary layer.
2) The shock wave and flow characteristics are strongly affected by II- Effect of an Oblique Shock Wave Immediately Ahead of the Inlet,”
the presence of gust. The effect seems to be the strongest for the gust NACA RM-A8F08, Aug. 1948.
[17] Ferri, A., and Nucci, L. M., “Theoretical and Experimental Analysis of
angle of 30 deg and progressively decreases as the gust angle is Low-Drag Supersonic Inlets Having a Circular Cross Section and a
increased. The major effect of side gust is that the shock structures Central Body at Mach Numbers of 3.3, 2.75, and 2.45,” NACA Rept.
and the external and internal flows become asymmetric with respect 1189, 1954.
to the longitudinal axis. Thus, the oblique shock waves in the lower [18] Connors, J. F., and Meyer, R. C., “Performance Characteristics of
part of the inlet are stronger compared to those in the upper part, and, Axisymmetric Two-Cones and Isentropic Nose Inlets at Mach Number
consequently, there is a stronger normal shock in the upper part 1.9,” NACA TN-3589, Jan. 1955.
compared to that in the lower part. It also leads to stronger shock/ [19] McGregor, I., “Some Theoretical Parameters Relevant to the
boundary-layer interaction, causing significant boundary-layer Performance of Rectangular Air Intakes with Double-Ramp
separation, secondary (radial) flow, and a separated flow region in Compression Surfaces at Supersonic Speeds,” Royal Aircraft
the diffuser section, especially in the lower half. Another effect of side Establishment TR RAE/TR-71232 ARC 33616, 1971.
gust is that the normal shock gets located upstream or downstream of [20] Tindell, R. H., “Inlet Drag and Stability Considerations for MO  2.00
Design,” Journal of Aircraft, Vol. 18, No. 11, 1981, pp. 943–950.
the cowl lip, depending on the side gust angle. doi:10.2514/3.57584
3) The supersonic inlet performance is adversely affected by the [21] Soltani, M. R., Younsi, J. S., and Daliri, A., “Performance Investigation
presence of side gust. There is significant reduction in total pressure of a Supersonic Air Intake in the Presence of Boundary Layer Suction,”
recovery (more than 20%) and a corresponding increase in the FD for Journal of Aerospace Engineering, Vol. 229, No. 8, 2015, pp. 1495–
the low side gust angle (α  30 deg), while the mass flow ratio 1509.
Downloaded by Cranfield University on January 23, 2024 | http://arc.aiaa.org | DOI: 10.2514/1.C035093

decreases the most for the high side gust angle (α  90 deg). [22] Connors, J. F., and Meyer, R. C., “Design Criteria for Axisymmetric and
4) Other possible effects include flow instabilities that could Two-Dimensional Supersonic Inlets and Exists,” NACA TN-3589,
potentially lead to buzzing inlet operation. The investigation of such Jan. 1956.
scenarios, however, requires transient flow simulations using a time- [23] Taylor, G. I., and Maccoll, J. W., “The Air Pressure on a Cone Moving at
accurate 3D compressible flow solver and will be considered in a High Speeds,” Proceedings of the Royal Society of London, Series A,
Vol. 139, No. 838, 1933, pp. 298–311.
future study.
doi:10.1098/rspa.1933.0018
[24] Maccoll, J. W., “The Conical Shock Wave Formed by a Cone Moving at
High Speed,” Proceedings of the Royal Society of London, Series A,
Acknowledgments Vol. 159, No. 898, 1937, pp. 459–472.
Hussein K. Halwas would like to gratefully acknowledge his doi:10.1098/rspa.1937.0083
scholarship sponsor, represented by Iraqi Ministry of Higher [25] Chen, L., and Caughey, D. A., “Calculation of Transonic Inlet
Education and Scientific Research via University of Babylon. Flowfields Using generalized Coordinates,” Journal of Aircraft, Vol. 17,
No. 3, March 1980, pp. 167–174.
doi:10.2514/3.57888
References [26] Bangent, L. H., Santman, D. M., Horie, G., and Miller, L. D., “Some
Effects of Cruise Speed and Engine Matching on Supersonic Inlet
[1] Oswatitsch, K., “Pressure Recovery for Missiles with Reaction Design,” Journal of Aircraft, Vol. 19, No. 1, Jan. 1982, pp. 58–64.
Propulsion at High Supersonic Speeds (The Efficiencies of Shock doi:10.2514/3.57356
Diffusers),” NACA TM-1140, 1947 (in German). [27] Shimabukuro, K. M., Welge, H. R., and Lee, A. C., “Inlet Design Studies
[2] Ferri, A., and Nucci, L. M., “Preliminary Investigation of a New Type of for a Mach 2.2 Advanced Supersonic Cruise Vehicle,” Journal of
Supersonic Inlets,” NACA Rept. 1104, 1951. Aircraft, Vol. 19, No. 7, July 1982, pp. 513–518.
[3] Kantrowitz, A., and Donaldson, C. D., “Preliminary Investigation of
doi:10.2514/3.57423
Supersonic Diffusers,” NACA ACR-L5D20, 1945.
[28] Biringen, S., “Numerical Simulation of Two-Dimensional Inlet
[4] Seddon, J., and Goldsmith, E. L., Intake Aerodynamics, AIAA
Flowfields,” Journal of Aircraft, Vol. 21, No. 4, April 1984,
Educational Series, 2nd ed., AIAA, Reston, VA, 1999.
pp. 244–249.
[5] Mahoney, J. J., Inlets for Supersonic Missiles, AIAA Educational Series,
doi:10.2514/3.48254
AIAA, Washington, D.C., 1990.
[29] Bradley, P. F., Dwoyer, D. L., South, J. C., Jr., and Keen, J. M.,
[6] Seddon, J., and Goldsmith, E. L., Intake Aerodynamics, AIAA
“Vectorized Schemes for Conical Potential Flow Using the Artificial
Educational Series, AIAA, New York, 1985.
Density Method,” AIAA Journal, Vol. 24, No. 1, Jan. 1986, pp. 13–20.
[7] Farokhi, S., Aircraft Propulsion, 2nd ed., Wiley, Hoboken, NJ, 2014.
[8] Halwas, H. K., “Design of an Axisymmetric Supersonic Air Intake,” M. doi:10.2514/3.9216
S. Thesis, Univ. of Babylon, Hillah Province, Iraq, Sept. 2003. [30] Jain, V., “Numerical Investigation of Supersonic Mixed-Compression
[9] Slater, J. W., “Design and Analysis Tool for External-Compression Inlet Using Euler Equations,” M.S. Thesis, Dept. of Aerospace
Supersonic Inlets,” 50th AIAA Aerospace Sciences Meeting including Engineering, Indian Inst. of Technology Kanpur, India, April 2000.
the New Horizons Forum and Aerospace Exposition, AIAA Paper 2012- [31] Lavante, E. V., and Thompkins, W. T., Jr., “An Implicit, Bidiagonal
0016, Jan. 2012. Numerical Method for Solving the Navier-Stokes Equations,” AIAA
[10] Faro, I. D. V., “Supersonic Inlets,” AGARD, NATO AGARDograph Journal, Vol. 21, No. 6, June 1983, pp. 828–833.
102, France, May 1965. doi:10.2514/3.8159
[11] Ryu, K. J., Lim, S., and Song, D. J., “A Computational Study of the [32] Walters, R. W., Dwoyer, D. L., and Hassan, H. A., “A Strongly Implicit
Effect of Angle of Attack on a Double-Cone Type Supersonic Inlet with Procedure for the Compressible Navier-Stokes Equations,” AIAA
a Bleeding System,” Computers and Fluids Journal, Vol. 50, No. 1, Journal, Vol. 24, No. 1, Jan. 1986, pp. 6–12.
2011, pp. 72–80. [33] Knight, D. D., “Calculation of High-Speed Inlet Flows Using the
doi:10.1016/j.compfluid.2011.06.019 Navier-Stokes Equations,” Journal of Aircraft, Vol. 18, No. 9,
[12] Archer, R. D., and Saarlas, M., An Introduction to Aerospace Feb. 1981, pp. 748–754.
Propulsion, Prentice–Hall, Upper Saddle River, NJ, 1996. doi:10.2514/3.57557
[13] Kerrebrock, J. L., Aircraft Engines and Gas Turbines, 2nd ed., [34] Geetha, J. J., Ganesh, A. T. K., and Panneerselvam, S., “Forebody
Massachusetts Inst. of Technology, Cambridge, MA, 1992. Design for Aerodynamic-Propulsion Requirements of an Airbreathing
[14] Lim, S., Koh, D. H., Kim, S. D., and Song, D. J., “A Design Optimization Hypersonic Research Vehicle,” National Conference on Air Breathing
Study of Diffuser Shape in a Supersonic Inlet,” D3-4, Asian Joint Engines and Aerospace Propulsion (NCABE), DRDL, Hyderabad,
Conference on Propulsion and Power (AJCPP), Gyeongju, Korea, India, Dec. 2000, pp. 386–398.
March 2008. [35] Slater, J. W., “External-Compression Supersonic Inlet Design Code,”
[15] Jung, S. Y., “Preliminary Design for Axisymmetric Supersonic Inlet NASA Technical Reports Server (NTRS) TR Technical Conference,
Using Conical Flow Solution and Optimization Technique,” Journal of Cleveland, OH, March 2011.
the Korean Society for Aeronautical Science, Vol. 34, No. 9, 2006, [36] Haas, M., Elmquist, R. A., and Sobel, D. R., “The NIDA Code: A New
pp. 11–19. Tool for Supersonic Inlet Design and Analysis,” Joint Army Navy NASA
doi:10.5139/JKSAS.2006.34.9.011 Air Force (JANNAF) Propulsion Meeting, Vol. 1, Nov. 1993.
582 HALWAS AND AGGARWAL

[37] Barnhart, P. J., “IPAC-Inlet Performance Analysis Code,” NASA CR- WMO, 7biz, Avenue de la Paix, Case Postale No. 2300, CH-1211
204130, July 1997. Geneva 2, Switzerland, Jan. 2010.
[38] Yoder, D. A., “Wind-US User’s Guid Version 4.0,” NASA TM-2016- [43] Sritharan, S. S., and Seebass, A. R., “Finite Area Method for Nonlinear
219145, Sept. 2016. Supersonic Conical Flows,” AIAA Journal, Vol. 22, No. 2, Feb. 1984,
[39] Thangadurai, G. R. S., Chandran, B. S. S., Babu, V., and Sundararajan, pp. 226–233.
T., “Numerical Simulation of Supersonic Mixed Compression doi:10.2514/3.8372
Axisymmetric Air-Intakes,” National Conference on Air Breathing [44] Moeckel, W. E., Connors, J. F., and Schroeder, A. H., “Investigation of
Engines and Aerospace Propulsion (NCABE), DRDL, Hyderabad, Shock Diffusers at Mach Number 1.85. II-Projecting Double-Shock
India, Dec. 2000, pp. 602–611. Cones,” NACA RM-E6L13, June 1947.
[40] Burnham, J., “Atmospheric Gusts, A Review of the Results of Some [45] Kennedy, E. C., “Calculation of the Flow Fields Around a
Recent Research at The Royal Aircraft Establishment,” Monthly Series of Bi-Conic Bodies of Revolution Using the Method of
Weather Review, Vol. 98, No. 10, Oct. 1970. Characteristics as Applied to Supersonic Rotational Flow,” OAL
[41] Barry, F. W., “Development of Atmospheric Gust Criteria for Ordinance Aerophysics Lab. OAL/CM 873, Daingerfield, TX,
Supersonic Inlet Design,” NASA Rept. CR-1143 72, Hamilton June 1956.
Standard Division, United Aircraft Corp., HSER-5195, Windsor Locks, [46] Menter, F. R., “Two-Equation Eddy-Viscosity Turbulence Models for
CT, Dec. 1968. Engineering Applications,” AIAA Journal, Vol. 32, No. 8, Aug. 1994,
[42] “Info note No. 58: World Record Wind Gust: 408 km∕h,” World pp. 1598–1605.
Meteorological Assoc., Specialized Agency of the United Nations, doi:10.2514/3.12149
Downloaded by Cranfield University on January 23, 2024 | http://arc.aiaa.org | DOI: 10.2514/1.C035093

You might also like