01 HI

VOLUME 5 NOVEMBER-DECEMBER 1968 NUMBER 6

Design and Development of an Air Intake for a

Supersonic Transport Aircraft
I. H. RETTIE*
Downloaded by UNIVERSITY OF ADELAIDE -INTERNET on January 3, 2018 | http://arc.aiaa.org | DOI: 10.2514/3.43977

British Aircraft Corporation (Operating) Ltd., Bristol, England

AND

W. G. E. LEWIS!
National Gas Turbine Establishment, Pyestock, England

The reasons for choice and the characteristics of an external-compression intake geometry
for operation at Mach numbers up to and beyond 2.0 are described, and the problems of ap-
plication to SST aircraft are discussed. Mounted under the wing of the Concorde, this inlet is
divided to provide independent supplies of air to a pair of engines, an arrangement that intro-
duces particular problems in allowing for the wing flowfield and avoiding interaction between
the twin inlets. The precise definition of an intake geometry for a supersonic transport
should have regard for the over-all performance of the propulsion system. The choice of con-
trolling parameters and the design of the control system must give good performance and
engine handling in a wide range of off-design conditions without demanding excessive com-
plexity. The aerodynamic and other development tests required to make the appropriate
decisions are described in detail. The results underline the suitability of this basic geometry
in association with the other components of the propulsion system for SST operation.

Nomenclature = main duct recovery

= ramp bleed pressure recovery
AM = secondary-nozzle shroud entry area (Fig. 11)
Aj = primary-nozzle exit area (Fig. 11)
e = bypass sonic throat area/intake capture-area ratio
DC6Q = (P m i n - P)/q I. Introduction
P = mean total pressure at the compressor face
JPmin = lowest mean total pressure over any 60° segment of the

q
compressor face
= mean dynamic head at the compressor face
I N this paper we shall review the aerodynamic aspects of the
design and development the engine air intake for the Anglo-
French Concorde aircraft. The manufacture of the Concorde
M = Mach number is being undertaken jointly by Sud Aviation at their factory in
NL = engine L.P. compressor rpm Toulouse, France, and by the British Aircraft Corporation's
PS = secondary flow total pressure
Pj = primary jet total pressure Filton Division at Bristol, England; division of work be-
TI = freestream total pressure tween the two firms gives the British team responsibility for
52 = forward ramp angles (Fig. 3) the over-all design of the power plant. The joint organization
e! = main duct mass flow ratio of the project has, of course, enabled the French team to con-
€B = ramp bleed mass flow ratio tribute wherever it was advantageous; both British and
IJL = secondary flow coefficient = WsTsllz/WjTjllz French wind tunnels have been used extensively.
Ws = secondary mass flow Consider first the constraints imposed upon the intake
Wj = primary jet mass flow design by the aircraft as a whole. The general arrangement
TS = secondary air temperature drawing (Fig. 1) shows the mounting of the engines in twin-
Tj = primary jet air temperature
underwing nacelles. Each powerplant, including intake,
Presented as Paper 67-752 at the AIAA/RAeS/CASI 10th engine, nozzle and the associated control gear, is independent
Anglo-American Aeronautical Conference, Los Angeles, Calif., in all important aspects. Aerodynamic independence is
October 18-20, 1967; submitted November 10, 1967. The au- achieved by means of a projection forward of the center wall
thors thank the British Aircraft Corporation and the Ministry (splitter plate) separating the pair of inlets. Maximum in-
of Technology for permission to publish this paper. They are take efficiency is an economic requirement for SST aircraft;
also indebted to D. P. Morriss of BAG and M. C. Neale of
NGTE for valuable assistance during the preparation of the Table 1 shows the relationship between intake performance
paper. and payload.
* Chief Aerodynamicist, Filton Division; now with the A detailed discussion of the reasons for the under-wing semi-
Boeing Company, Seattle, Wash. span position for the engines is outside the scope of this paper.
f Head of Department, Engine Research Department. Suffice it to say that a suitable layout for the aircraft has been
513
 514 I. H. RETTIE AND W. G. E. LEWIS J. AIRCRAFT

Table 1 Effect of aerodynamic loss on payload

Reduction of
Aerodynamic loss Payload
1% Intake efficiency (subsonic) 1%
1% Intake efficiency (supersonic) 2|%
1% Aircraft drag (subsonic) 1%
1% Aircraft drag (supersonic) 4i%
1% Intake efficiency (takeoff) 5%

In parallel with work on the external-compression design,

the external/internal compression system was brought to a
high state of development; an early product of this research
was the "self-starting" internal-compression system which, if
properly exploited, eliminates one major difficulty of that
shock geometry. More recent work suggests the possibility
of eliminating the incipient instability always associated with
Downloaded by UNIVERSITY OF ADELAIDE -INTERNET on January 3, 2018 | http://arc.aiaa.org | DOI: 10.2514/3.43977

Fig. 1 Aircraft general arrangement.

the partial internal-compression system operating at its
critical point. Those avenues are worth following, but were
achieved and that it satisfies the following requirements: acknowledged to be incompatible with a time scale appropri-
1) adequate length for the complete powerplant, 2) adequate ate to a first flight in the spring of 1968.
separation of the intake and the wing leading edge to avoid The choice of external compression has been well justified
interference between the intake flow and the vortex develop- by the simplicity of intake control in flight. There is, for
ment on the upper surface of the wing, 3) feasible undercarriage example, no "unstart" mode; there is a generous margin be-
geometry with adequate ground clearance for lift-off and tween any operating condition and buzz; best performance
touchdown design cases. We do not wish to protract here the over a wide range of flight conditions may be obtained by
arguments concerning the choice between two-dimensional allowing the bleed flow to compensate naturally for changes in
and axisymmetric intakes. The European choice is obvious, engine air flow demand, without the intervention of any
and is justified at least in part by the way it has fitted neatly artificial control system; good engine-face velocity distri-
into the over-all aircraft design. butions are obtained with an elementary throat bleed geom-
The third major consideration in selecting the intake design etry. These virtues have allowed the use of a low-gain con-
is that of flexibility. The intake must provide good engine trol system which is actually dormant during the majority of
face-pressure distributions and high efficiencies for all likely supersonic flight. We hope that development of the system
ambient temperatures to be met in cruise, preferably without during prototype flying will lead to proposals for a manual
demanding an extremely complex high-response control standby system, eliminating much complex monitoring and
system. It must also provide adequate subsonic performance, duplicated equipment.
especially since the fuel for diversion cruise has to be carried It would be wrong to assume that this high degree of stabil-
all the way across the Atlantic. The ways in which the ity and simplicity has been obtained merely by sacrificing
intake design chosen for the Concorde meets these require- aerodynamic performance; in fact, the final geometry (Fig. 2)
ments are our main concern here. results from a detailed development program. Tests demon-
strate the value of isentropic compression in the ramp, not
II. Choice of Compression System and only for its effect on the theoretical shock recovery, but also
Intake Geometry for the smooth ride it gives the sidewall boundary layers
through supersonic compression. The aerodynamicist will
Confining attention now to the two-dimensional arrange- note that this cowl generates the " strong solution" oblique
ment, both the external/internal and all external compression shock, across the capture flow; for a practical low-drag design,
systems merit consideration for a Mach number of 2.2.l the limit of drag reduction is imposed by shock detachment at
At first sight, Mach 2.2 appears a rather awkward speed; it is the cowl lip. This design thus combines a high standard of
certainly too low to make a cast-iron case for partial internal internal recovery and flow quality with low external drag.
compression, but considerable fears were entertained at the The purists can (and do) argue whether this development of
outset regarding the cowl drag of a conventional external- the bleed slot and cowl-lip geometries has introduced a mea-
compression geometry. External compression was chosen sure of internal compression.
because of the inherent stability of the associated shock pat- Note that the bleed slot, whose aerodynamic operation is
tern and the simplicity of the associated bleed arrangements described fully by Leynaert in Ref. 2, is ideally situated for
and control systems. dealing with the boundary-layer growth on the front ramp
and also with the impingement of the cowl-lip shock on the
- HINGED upper duct surface. As pointed out by M. Leynaert, large
RAMPS
variations in the flow extracted by the slot can occur without
significantly affecting the external shock pattern of the intake.
The way this feature simplifies powerplant management will
be discussed later.
Other noteworthy features of the intake geometry are
the splitter plate, which provides aerodynamic independence
of the twin intakes; the hinged ramps, which vary their
position according to Mach number and engine setting; and
the dump door, which opens to spill excess intake air in cases
of engine throttling or shut down.
SPLITTER P L A T E It was an aircraft requirement to provide safe operation in
RAMP
BLEED SLOT -DUMP DOOR
upset maneuvers from cruise at Mach 2.2. Optimization
of intake performance has, however, been carried out at
Fig. 2 Intake geometry. Mach 2.0, approximately the speed at which best aircraft
 NOV.-DEC. 1968 AIR INTAKE FOR SUPERSONIC TRANSPORT 515

100
range is, at present, obtained. Modifications to provide
optimization for Mach 2.2 cruise are being studied. 90

III. Characteristics of the Isolated Intake

Uniform-entry "isolated" tests (divorced from wing airflow) 70
have laid the foundation for present Concorde intake design. PERCENTAGE
When the basic design was frozen, intakes were manufactured CRUISE M A S S 60
FLOW FOR
for testing; they included T^ scale models, a -3- scale twin BUZZ 50
intake for tests at Centre d'Essais des Propulseurs (Saclay,
France), and a full-scale version running in Cell 4 at the Na-
tional Gas Turbine Establishment. The importance of ENGINE

Reynolds number was fully appreciated early in the test pro- WINDMILLING

gram, and many of the smaller scale models were tested in

wind tunnels at stagnation pressures of 5 atm, thus very
nearly simulating full-scale conditions.
N O R M A L CONTROL.
MODE
Downloaded by UNIVERSITY OF ADELAIDE -INTERNET on January 3, 2018 | http://arc.aiaa.org | DOI: 10.2514/3.43977

Performance

Because of the sensitivity of aircraft economics to intake Fig. 4 Cruise margins to buzz.
performance, the experimental derivation of performance
characteristics demands great precision. Figure 3 shows
typical pressure recovery/mass flow characteristics at a test after, on the characteristic, bleed conditions are constant.
Mach number simulating the mean underwing conditions at Throughout the whole process the total capture flow is un-
Mach 2.0 (freestream). Mass flow measurements shown are changed, so that changes in the bleed are compensated by
based on the performance of standard nozzles with accurately equal and opposite changes in the engine flow. In the sub-
known discharge coefficients. Detailed traverses of the flow critical regime, the cowl-lip shock becomes detached; total
ahead of the intake are also necessary. The sensitivity is capture flow is reduced by the upstream movement of this
such that an error of 0.005 in the estimated capture-plane shock in accommodating the required forespillage. The
Mach number leads to an error of \% in the capture-mass flow flow in the throat is then completely subsonic.
ratio. The characteristics of Fig. 3 are drawn for three different
Test results show a well-defined critical point near the values of bleed control area, i.e., three values of the ratio of
point of maximum pressure recovery. Calculations of theo- bleed pressure to bleed-mass flow. Note that to a good ap-
retical shock recovery show that the "extra-to-shock" losses proximation, the level of bleed recovery uniquely defines the
are reduced to only 1% of the f reestream total pressure. intake operating condition relative to the critical point, inde-
Thus, pressure recoveries in the region of 0.95 are achieved. pendently of throat bleed mass flow. For peak internal
An excursion into the supercritical region causes the throat- performance at the aircraft design point it is necessary to size
shock structure to move downstream; the consequent reduc- the intake and bleed passages for operation at an elevated
tion in local static pressure simultaneously reduces the bleed bleed pressure, close to critical conditions.
flow. Note that the supercritical leg of the characteristic is
vertical, indicating that the main duct flow increases by a Stability
compensating amount during this process. A significant feature of the characteristics is the "mild in-
When the shock translation is complete, supersonic condi- stability/7 which occurs as a result of shock oscillation when
tions are fully established in the throat of the intake; there- the intake is operating slightly subcritically. The oscillation
is probably initiated according to the Ferri criterion. It
FREE S T R E A M M A C H NO. = 2.0 commences upon the intersection of a vortex sheet with the
R A M P ANGLE=9.6°
ONSET OF MILD- cowl lip; the vortex sheet emanates from the intersection
of the near-normal shock, upstream of the cowl lip, with the
ramp hinge shock. A quiescent region exists between this
INTAKE region of mild instability and the onset of buzz, the latter
PRESSURE 92
being precipitated by intersection of the first wedge shock and
the expelled shock.
The onset of mild instability could provoke surge if the
engine is sensitive to flow oscillation at the compressor face.
This aspect of engine sensitivity will remain obscure until
flight experience is obtained; it is considered prudent at the
present time to avoid intake operation in this condition. This
BLEED 5
clearly imposes an upper limit to the level of ramp bleed-
PRESSURE pressure recovery for steady-state operation. Operation at
RECOVERY
["• = C R I T I C A L POINT ramp bleed pressure recoveries consistent with near-critical
intake conditions provides an adequate margin to cope with
rapid transient variations in mass flow.
Only remote circumstances such as, for example, undetected
failure of both control system channels followed by failure of
BLEED FLOW the affected engine can produce buzz conditions. Neverthe-
less, it is a design requirement that the intake structure with-
stand buzz conditions for a reasonable time. Buzz pressures
in the intake and bleed ducts were recorded in a model
equipped with a rotating valve to simulate buzz, surge, and
0.7 0.8 0.9 0.95

I N T A K E CAPTURE M A S S FLOW R A T I O
hammer-shock conditions. Changes in ramp angle and dump
door position accommodate reductions in engine mass flow
Fig. 3 Typical intake characteristics. without incurring buzz (Fig. 4). With full deflection of each,
 516 I. H. RETTIE AND W. G. E. LEWIS J. AIRCRAFT

0.004 ——————— PRE-ENTRY DRAG

D E R I V E D FROM
MASS FLOW MEASUREMENT
AND STATIC PRESSURE A N A L Y S I S
PORT INNER
2.0M
Downloaded by UNIVERSITY OF ADELAIDE -INTERNET on January 3, 2018 | http://arc.aiaa.org | DOI: 10.2514/3.43977

Fig. 5 Engine-face distortions total-pressure contours—

9 10
view looking downstream.
RAMP ANGLE 6 0

Fig. 7 Pre-entry drag at a local Mach number of 1.9.

the engine mass flow can be safely reduced to 5% of the
maximum continuous value at Mach 2.0 (ISA+5°C). The
figure also shows the schedule of ramp and dump-door move- supersonic flow ahead of the cowl lip. Subsequently, the pre-
ments in engine throttling, that will ensure acceptable engine- entry drag rise steepens as the cowl-lip shock detaches and
face distributions. causes a region of partially subsonic flow. There are simul-
taneous changes in the cowl drag.
Engine-Face Pressure Distributions Tests were made with intakes featuring two identical sides,
both having leading edges lying on the straight line joining
A typical distribution of engine-face pressures in cruise
ramp tip to cowl lip. Other tests used models with the "split-
(Fig. 5) shows one of many distributions that were simulated
ter plate" that separates the two intakes forming each power-
at an early stage by gauzes in an engine-test cell. No single
plant nacelle. Tests with the "nonsplitter" arrangement and
parameter can uniquely define how an engine reacts to a given
large sidewall windows permitted the pre-entry drag to be
distortion pattern; experience in Great Britain with the DC60
estimated from the observed shock pattern (Schlieren analy-
parameter3 has shown a fairly high correlation with the effect
sis) which is deflected upstream of the inviscid pattern be-
of flow distortions on engine-surge margin. The distortions
cause of boundary-layer effects. The shock pattern, and
are very small indeed, and the test work carried out in Cell 4
hence the pre-entry drag, could also be deduced from measure-
at NGTE gives every reason to expect trouble-free engine
ments of the static pressures on the ramp surface and the
operation in flight. Figure 6 shows the variation of the dis-
total capture-mass flow ratio. The two methods were in
tortion parameter with throat bleed conditions. Note the
excellent agreement (Fig. 7).
sensitivity of compressor flow distortion to bleed conditions
The introduction of the splitter plate renders deductions
for extreme values of the bleed flow; both an upper and a
drawn from experimental observations of the shock geometry
lower limit to throat bleed flow must be defined in order to
less certain. The problem was therefore approached from a
ensure satisfactory operation.
theoretical standpoint; the spanwise effect of the splitter
plate in deflecting the ramp shock geometry was computed
Pre-Entry Drag
theoretically, and the consequent reduction in capture mass
The positioning of the ramp shocks ahead of the cowl-lip flow compared with the experimental measurement. The
leads to a mass-flow ratio of less than 1.0 at all flight condi- excellent agreement of computation and experiment (Fig. 8)
tions. The maximum value of this ratio depends, of course, lends confidence to the computed effect of the splitter plate on
upon the particular setting of the variable ramp: associated pre-entry drag. Before applying this pre-entry drag data to
with the adjustment of the capture flow are changes in the aircraft performance predictions, corrections accounting for
external flow conditions. The pre-entry drag rises as the lift effects and for the nonuniform wing flowfield must be
spill is increased, at first at a rate corresponding with purely applied.

DC 60

Y
0.82

0.02 0.04 0.06 0.08

0.82 0.86 0.90 0.94

C B BLEED FLOW (%) CAPTURE MASS FLOW RATIO

Fig. 6 Effect of bleed flow on engine-face distortion. Fig. 8 Effect of splitter plate on capture mass flow.
 NOV.-DEC. 1968 AIR INTAKE FOR SUPERSONIC TRANSPORT 517

Use of Large-Scale Intakes

Because of their size, tests of the full-scale twin intakes with

engines and intake-control systems were impossible. How-
ever, Cell 4 at the NGTE (Pyestock, England) was able
to accommodate a single Concorde intake with Olympus 593
engine, making possible full-scale investigation of intake
characteristics and the effects on engine handling.
In addition to the full-scale single-intake test, French
officials decided to test an approximately one-third scale
twin intake with General Electric J85/J2 engines, using the
R5 facility at Centre d'Essais des Propulseurs (Saclay, France).
This made possible the study of twin duct interaction effects, Fig. 10 Diverter geometry.
including those resulting from engine surge, albeit with engines
whose surge sensitivity and characteristics were different
from those of the Olympus engine. recovery. The reduction in diverter drag, however, more
These two installations are equipped with versatile auto- than offsets this for degrees of immersion which lead to at
matic intake control systems, so that it is possible to explore least the outer 30% in depth of the boundary layer being
Downloaded by UNIVERSITY OF ADELAIDE -INTERNET on January 3, 2018 | http://arc.aiaa.org | DOI: 10.2514/3.43977

the stability and response characteristics of the complete swallowed by the intake. A small increase in ramp bleed
controlled powerplant and to check the validity of an analog flow is sufficient to regain the main duct recovery of the non-
simulation of the dynamics of the system. Operating limits immersed case.
can be assessed, and effects of sidewash, ramp travel and The diverter planform was developed as a result of con-
dump-door opening can all be investigated at various engine siderable testing of both isolated diverter models and the
settings in both steady-state and transient conditions. complete aircraft model. Early diverter designs showed an
A gratifying aspect of isolated engine testing is the good "unstarted" diverter flow where forespillage from the diverter
agreement between different test configurations. Figure 9 was interfering with the intake shock pattern and causing a
shows the agreement between the characteristics obtained significant loss of intake capture. It is clearly essential to
from full-scale tests in Cell 4 at NGTE and from other tests position the apex downstream of the diverter inlet, and it was
at five atmospheres on a ^-scale model. The compressor- found necessary to reduce the included angle to not more
face flow distortions also agree; at a typical operating point, than 20° (Fig. 10).
a DC60 value of —0.09 obtained from a model test compares
with a figure of —0.10 in the full-scale test. Intake Characteristics

Figure 9 shows a comparison between the intake character-

IV. Complete Aircraft Testing istics measured on the complete aircraft model with those of
Wing Flowfield Effects isolated intakes. In general, measured recoveries agree very
well and the complete aircraft model shows an improved
To study the effects of the wing flowfield, a y^-scale model maximum capture and a much-reduced level of instability,
of the complete aircraft was constructed, complete with twin particularly in the region just below the critical point.
nacelles fully instrumented for intake testing. First, a full The splitter plate was designed to ensure independence of
exploration of the underwing flowfield was made without the shock patterns in front of each intake when one or other
nacelles to determine boundary-layer depths and profiles, was operating in subcritical conditions; complete aircraft
intake-face Mach numbers, and flow directions. To account testing has confirmed the design. In stable running, there is
for the sidewash near the underwing surface at the appropriate no interaction between the two intakes. When buzz condi-
cruise incidences, the intakes were toed-in on the wing of the tions are reached in one duct, a very small sympathetic dis-
aircraft (Fig. 1), giving a banana-like curvature to the diffuscr. turbance is observed in the other intake: the condition be-
The amount of toe-in is 4° for the outer intake and 2° for the comes progressively more important as the buzz regime is
inner. penetrated. In practice, this means that when buzz is fully
It was expected that when boundary-layer diverter drag established in one duct, oscillations of about 30% of the buzz
was accounted for, an over-all optimum performance would be amplitude will be experienced in the neighboring duct.
found with some degree of immersion of the first wedge of the Another form of interaction results from induced yaw fol-
intake in the wing boundary layer. Tests on both the iso- lowing an engine failure, causing sidewash at the face of all
lated intake and the complete aircraft model show a small intakes by very nearly the sideslip angle attained by the air-
loss in capture coupled with a drop in main duct pressure craft, and a consequent deterioration of the engine face pres-
sure distributions. A modification to the sidewall geometry
of the intake is being incorporated in the third and subsequent
aircraft, and will significantly improve the tolerance of the
intake to sidewash. Tests in Cell 4 at NGTE on the proto-
type geometry have shown that the current engine surge
margin is adequate for the expected amounts of sidewash.

V. Integrated Powerplant Concept

General Arrangement of the Powerplant

Each Concorde nacelle houses a twin powerplant, each

member of the pair being independent in all possible aspects.
The internal air system (Fig. 11) has two principal streams of
air, denoted as primary and secondary.
The secondary stream is derived from the intake throat
at the ramp bleed slot and passes through the engine bay, to
0.80 0.90 1.00
be exhausted in the dual stream secondary nozzle. This
Fig. 9 Capture mass flow ratio. arrangement combines considerable flexibility of powerplant
 518 I. H. RETTIE AND W. G. E. LEWIS J. AIR Oil AFT

SECONDARY FLOW
SHUT-OFF FLAPS INTAKE
CHARACTERISTICS

MATCHING
LINE
BUZZ
BOUNDARIES

RAMP
BLEED SLOT
BLEED
CONTROL

PRIMARY AIR FLOW

- SECONDARY AIR FLOW

Fig. 11 Powerplant layout.

Downloaded by UNIVERSITY OF ADELAIDE -INTERNET on January 3, 2018 | http://arc.aiaa.org | DOI: 10.2514/3.43977

operation with good over-all performance. A correctly

matched dual-stream nozzle provides high exhaust efficiency
for the mixed primary and secondary flows. The added ad-
vantage of being able to use this air supply for engine bay
ventilation and cooling makes the arrangement much superior
to any other. .70 .80 .90 .95
An important development problem was the reduction of MAIN DUCT FLOW R A T I O
the pressure losses in secondary flow ducts and engine bay to
an acceptable level. Internal flow tests on an accurate half- Fig. 13 Intake/engine matching.
scale model of the powerplant were carried out, and the ar-
rangement was progressively modified until an efficient sys- what ambient temperature should be assumed for the cruise-
tem was obtained. The greater part of the losses were arising point? Second, where on the intake-recovery-vs-bleed-flow
at the entry to the engine bay; in the bay itself, Mach num- characteristics should the engine operate? Third, what
bers were low and the drag of the engine and its accessories margins should be built in to provide good stability? The
was not important. specified design point for the intake is Mach 2.0 at an ambient
Recall that the intake characteristics were presented for temperature of ISA-f-5°C (as a mean temperature, this is
various values of bleed control area, this being the test con- exceeded in only about 15% of occasions in the North Atlantic
figuration (Fig. 3). The effective bleed control area (Fig. 11) route).
is determined by the choking effect of the expanding primary The presentation of intake characteristics given in Sec. Ill
jet in the dual stream nozzle. The detailed implications of showed that operation at near-critical conditions with high
this arrangement have been dealt with at length by Talbot duct recovery and an operating margin from the mild instabil-
and Furness in Ref. 3, and are only summarized here. It is ity would be ensured by choosing a unique upper bound for the
clear that it is the engine and exhaust nozzle which normally ramp bleed pressure. Now the geometrical parameter which
dictate how the intake and its associated bleed flow operate. is most significant in determining the bleed control area is,
At a given speed, specific range will vary with bleed flow; an for a fixed engine setting, the nozzle shroud entry area. Fig-
optimum value for any cruise condition can be determined ure 12 shows the variation of main duct recovery with bleed-
only after a comprehensive study of intake, engine, engine mass flow for constant bleed pressure, and also the variation of
bay, and nozzle characterisitics. nozzle pumping characteristics with shroud entry area.J
The relationship between matching intake and nozzle shroud
Choice of Intake Area areas can be obtained from this diagram and the variation of
net thrust obtained as a function of bleed flow. The result
Beyond a brief reference to the effect of flow distortions, of this calculation shows a fairly constant net thrust for values
nothing has so far been said about the factors governing the of the bleed flow between 3 and 6%. Beyond this figure, the
choice of intake size in respect to a given engine-flow char- increasing secondary momentum loss and deteriorating nozzle
acteristic. There are three questions to be answered. First, performance are not offset by improvements in intake effi-
ciency. This range of bleed flow is compatible with acceptable
0.96 engine face distortions (see Fig. 6). The arguments detailed
C O N S T A N T BLEED
PRESSURE RECOVERY — below concerning flexibility of operation over the likely range
0.95 of flight conditions led us to associate an ambient temperature
INTAKE
/ of ISA-f5°C with a high value of bleed flow. The intake and
RECOVERY
0.94 nozzle shroud areas were therefore chosen so that at the design
point the intake operates at maximum bleed pressure, with a
ZL (%) bleed flow of about 6%, and with a ramp angle at the minimum
0.93
^- BLEED FLOW
consistent with adequate stability.

Effect of Ambient Temperatures

The flow characteristics for a supersonic intake operating

SECONDARY ° 20 at a constant Mach number are such that its airflow varies
PRESSURE
inversely as the square root of the ambient temperature; thus
a rise in ambient temperature leads to a drop in airflow.

3 4 5 6 7 8 9 % Secondary flow in the nozzle is composed of the ramp-bleed

flow plus cabin-air heat exchanger and other cooling flows ex-
Fig. 12 Intake/nozzle matching. hausted in the engine bay.
 NOV.-DEC. 1968 AIR INTAKE FOR SUPERSONIC TRANSPORT

STANDARD FOR 90% OF T O T A L

D E V I A T I O N = 8°K MILES F L O W N
TEMPERATURE ISA+5
UNACCEPTABLE
STRUCTURAL
TEMPERATURE

-10 -5 ISA

TEMPERATURE

Fig. 15 World-wide frequency weighted temperature

distribution.

low ramp bleed flows; engine face distributions would then be

poor, and bay ventilation inadequate; the engine rpm must
1.2 1.4 1.6 1.8 2.0 2.2 2.4 therefore be reduced until a tolerable level of ramp bleed
Downloaded by UNIVERSITY OF ADELAIDE -INTERNET on January 3, 2018 | http://arc.aiaa.org | DOI: 10.2514/3.43977

FREE S T R E A M MACH NO. pressure ratio is obtained; 2) intermediate conditions, where

Fig. 14 Intake operating regimes.
the engine may be operated at maximum continuous setting
and the ramp is at the minimum angle consistent with ade-
quate stability; 3) warm conditions, where the ramp angle is
Similarly the engine will also respond to ambient temperature increased to limit ramp void pressure recovery; 4) very warm
changes, but in a manner which approximates the direct in- conditions, where a large ramp angle would lead to poor engine
verse of ambient temperature (assuming constant rotational face distributions; the ramp angle is limited and the dump
speed and intake efficiency). It is therefore possible to door opened to maintain ramp void pressure.
match the powerplant for optimum performance at only one Figure 14 shows the ambient temperature ranges defining
ambient temperature. Maladjustments can also arise from these regimes with the chosen intake area and the Olympus
differences in engine flow characteristics from mean brochure 593 engine; Fig. 15 shows the probability of enroute ambient
figures, and other sources. The manner by which differences temperatures, frequency-weighted over worldwide long-dis-
are adjusted is an important aspect of supersonic intake de- tance routes. Note that the band of temperature in which
sign. the control system is (in normal circumstances) dormant
The primary manner in which the present intake is adjusted covers a large percentage of aircraft-cruise miles. Movement
to match engine conditions is by variation of supersonic fore- of the ramp and dump door is effected automatically by the
spillage, effected by variation of the ramp angle. At ambient intake control system. Throttling of the engine in the very
temperatures above the design condition, the ramp angle is in- cold regime is manual.
creased in order to maintain near-critical conditions. At
temperatures below the design condition the intake is allowed Intake-Control System
to operate in the supercritical regime.
Intake adjustment is illustrated in Fig. 13, where the intake The maximum and minimum ramp bleed pressure recover-
characteristics are plotted against main duct flow so that the ies and ramp angles as functions of Mach number are shown
intake and engine conditions can be directly related. In the in Fig. 16. The intake control system senses bleed pressure
supercritical regime, note that the recovery/flow character-
istics are no longer vertical. Since the total of duct and
bleed flow which is constant, the powerplant air-flow systems
are self-compensating (at ambient temperatures below that
for minimum ramp angle) in that changes in engine demand
can occur without changing the total intake capture flow.
Studies show that it is more efficient to maintain maximum
continuous engine conditions and operate the intake supercri-
tically than to throttle the engine to maintain intake re-
covery. The throttling mode which is most favorable for
cold-climate operation is a reduction of rpm at constant tur- .6 1.8 2.0 2.2

bine entry temperature; on the Olympus engine, this involves FREE S T R E A M MACH NO.

a reduction of primary nozzle area leading to a loss of in- Fig. 16a Ramp void-pressure recovery schedule.
stalled nozzle performance.
Turning now to flight conditions in which the ambient
temperature is higher than that for minimum ram}) angle (the
ramp-spill regime), near-critical intake operation is achieved
with good intake pressure recovery over an extensive range of
temperature variation when the ramp is controlled so as to
maintain a constant ramp bleed pressure. The effective
bleed control area, which increases fairly rapidly with ambient
temperature in the supercritical regime, is almost constant in
the ramp spill regime.

Powerplant Management

Having chosen the capture area of the intake, it is possible

1.6 1.8 2.0 2.2 2.4
to identify the following four regimes of powerplant operation
FREE S T R E A M MACH NO.
according to the ambient temperature; 1) very cold condi-
tions, where the large engine air flow demand would lead to Fig. 16b Ramp schedules.
 520 I. H. RETTIE AND W. G. E. LEWIS J. AIRCRAFT

LOW PRESSURE
COMPRESSOR
R P M

Fig. 17 Intake control system flow diagram.

recovery; if we consider a flight condition where the intake is HIGH PRESSURE COMPRESSOR R P M
operating at the maximum value of this parameter, it is clear
that a rise in ambient temperature or throttling of the engine Fig. 19 Effect of compressor face-flow distortion on engine
will lead first to a rise in bleed pressure recovery and then to a surge.
Downloaded by UNIVERSITY OF ADELAIDE -INTERNET on January 3, 2018 | http://arc.aiaa.org | DOI: 10.2514/3.43977

controlled increase in ramp angle. Similarly, an increase in

flight Mach number will lead to an increase of ramp angle. fitted with a representative Concorde powerplant in a central
The attainment of maximum ramp angle switches control position under the fuselage.
to the dump door, and this arrangement copes automatically The difficulty here is the powerful effect of engine-face
with throttling back or shut-down of the engine. To cover distortions on engine-surge margin in the high N/T1/2 engine
engine failure, an input based upon rate of change of engine conditions when running at maximum take-off thrust (Fig.
rpm increases the rate of operation of the ramp and dump 19). The simple form of auxiliary inlet might incure risks
door. An additional ramp over travel is provided to cover of poor engine handling; therefore, a sophisticated aero-
emergency cases, such as fire when the engine-bay flow must be dynamic design incorporating a flap and a cascade was de-
shut off, and the requirement for forespillage is therefore in- veloped in the wind tunnel. As shown in the figure, it can
creased. provide extremely good engine-face pressure distributions.
For the prototype Concorde, an extremely versatile intake-
control system is fitted (Fig. 17). Atmospheric data is de- Subsonic Performance
rived from the aircraft air-data system; full provision is
made for altering the control laws and, if necessary, elaborat- The importance of the secondary-flow system in subsonic
ing them to take account of incidence or sideslip effects. It is flight must not be underestimated. Development of the
hoped that these laws may be generated aerodynamically internal geometry of the ramp bleed slot has enabled bleed-
when they are finally defined for production aircraft. pressure recoveries of around 95% to be obtained, ensuring a
good supply of secondary air to the nozzle and allowing diver-
VI. Low-Speed Problems sion cruise (at Mach 0.93) with the nozzle-ventilation doors
closed. Decreased reliance on these ventilation doors should
The emphasis in this paper has been on development prob- simplify the development program and improve the confi-
lems associated with intake characteristics at the cruise Mach dence which can be placed on aircraft performance estimates.
numbers; the intake geometry is basically designed for this However, the particular problem posed by the intake at high
condition, but many difficulties have been encountered in subsonic Mach numbers is that of spill drag. There is no
obtaining satisfactory performance in other flight conditions. special difficulty entailed in securing either good pressure
recovery or flow uniformity at the compressor face, but the
Takeoff drag problem arises on two counts: first, an intake sized for
supersonic performance has to operate at low capture mass
The intake entry area is dimensioned for cruise and means flow ratios during subsonic cruise and the pre-entry drag is
of increasing the mass flow at low speeds must be fitted. Two thus significant; second, the ability to develop compensating
alternative designs of auxiliary door have been developed suction forces on the cowl is severely limited by the need for
(Fig. 18) and the study of engine/intake compatibility has a sharp low-angle lip to minimize drag in supersonic flight.
been assisted by the Vulcan flying test bed, a Vulcan aircraft Figure 20 shows schematically the familiar arrangement of the
forces involved. These are fundamental facts of life which
the designer of an intake for any supersonic airplane has to
face.

TYPE 'A' AUXILIARY TYPE 'B' AUXILIARY

ENTRY ENTRY

D = PRE-ENTRY DRAG
S = Z F O R C E S ON C O W L LIP AND S I D E W A L L S

TYPE 'B'
ENTRY
OR IGNORING COS 9

0.10 V A R I A B L E RAMP INTAKE

MACH NUMBER
Fig. 20 Ramp intake in subsonic flight (the forces in-
Fig. 18 Compressor face-flow distortion at low speed. volved).
Downloaded by UNIVERSITY OF ADELAIDE -INTERNET on January 3, 2018 | http://arc.aiaa.org | DOI: 10.2514/3.43977

NOV.-DEC. 1968 AIR INTAKE FOR SUPERSONIC TRANSPORT 521

During the past two years, much progress has been made.
Not only can drag estimates be advanced with confidence, but
also the favorable effects stemming from such possibilities as
trimming the ramp position are well understood. Figure 21
shows a typical case of the gratifying agreement now achieved
Fig. 21 Drag of
between experimental drag levels and those predicted by the
ramp intakes at method of Ref. 5. The sharp dependence of drag on capture
high-subsonic Mach flow ratio is most evident.
number.

References
1
0.2 0.4 0.6 0.8 1.0 Neale, M. C., "Intake Design and Performance Around a
C A P T U R E M A S S FLOW R A T I O Mach Number of 2.0," AGARDograph, No. 103.
2
Leynaert, J., "Fonctionnement du peige a couche limite
The proper appreciation of the magnitude of these forces interne d'une prise d'air a compression supersonique externe,"
and an awareness of the methods by which they might be AGARDograph, No. 103.
3
minimized is important. Two years ago it was possible for Talbot, J. E. and Furness, B., "A Fully Integrated Propulsion
Mount to observe that the whole area was one of "the many System for a Supersonic Transport Aircraft," AGARDograph,
little technical pockets that has been skipped over in the No. 108.
4
transition from high-speed subsonic to supersonic design point Reid, C., "The Effect of Maldistributions of Inlet Airflow on
Surge Characteristics of Axial Flow Compressors," in prepara-
flight" (Ref. 6). Among the reasons for this situation might tion, Rolls Royce; also Lovesey, A. C., "Gas Turbine Develop-
be supposed both a general failure to appreciate the severity ment," Journal of Royal Aeronautic Society, Aug. 1964.
of the problem, and the fact that in the whole field of experi- 5
Pike, M. R., unpublished work, Rolls Royce.
mental techniques, the accurate determination of spill drag 6
Mount, J. S., "The Effect of Inlet Additive Drag on Aircraft
presents probably the greatest challenge of all. Performance," Journal of Aircraft, Vol. 2, No. o, Sept.-Oct., 1965