_.

NAO_ EM A55Fi6

CONFIDENTIAL

TABLE OF COI_t_

._.

Page

!.

INTRODUCTION

II.

DEFINITIONS
AIR-INDUCTION SYST/M ...................
DMSION
OF FORCES ............
........
PEEFO_tNCE
PARAMETERS .................
_=_ESSUHE RECOVERY

.
- .

c ,
.

DRAG ...............
" ..........
MASS F[0W
.......................
FEELIMIN._RY CONSIDERATIONS
.................
AIRCRAFT HEQUIREMENTS
..................
AI EFRAME- INDUCTION- SYST_4 CGMBINATION
ENGINE- INDUCTION-SYST_
COMBINATION
..........
MATC_NG
........................
OPTIMIZATION

.
.
. o

,
.

.

FLOW UNIFORMITY AND STEADINESS .......

. .....
DETAIL CONSIDERATIONS
...................
INDUCTION
........................
PRESSURE RECOVERY AND Fz_)W UNIFORMITY
.........
Ducts
........................
Area ratio ....................
Skin fric ion losses ................
Flow sep_ation

.
,

Desi_
.......................
Subsonic Flight
...................
Lip design .....................
Angle of attack
..................
Inlet asjnmnetry ..................
Supersonic Flight
..................

_
:
_.
:_
_

III.

J'
-_
_.
_'_
_=
.
'
_i

._
i

IV.

:%

,
?

_
,,
_;
:_
i
:"_,
_

ii

........................

3
3
4
9
9
]_1.
12
13
13
14
i_
16
19
20
22
22
23
23
23
24
29
27
31
33

36
37

37

Supersonic compression ................

Limiting internal contraction
...........
Limiting iILlet Mach number .............
Boundarj-layer shock-wave interaction
......
Lip design .....................
Mass-flow variation
................
Angle of attack
..................
DRAG ..........................
Subsonic Flight
...................
Supersonic Flight
..................
Extern_l wave drag with no spillage
........
External profile ..................
Additive drag
...................
Change in external wave drag ............
Lip bluntness
...................
Net wave drag
...................

I'

_ l '_IIII

llll

NI

]ll

,l l

38
40
43
45
47
48
51
52
53
56
57
61
63
64
65
66

I _

_I

--'L_

,
4 :

CO_I_DE_.
__

|t

m e#

)1)

:,..

!i!
"

s_

mo

..:
m_@

. : NACA _4 A55F16
@meQ

TABLE OF C0_ENTS - Concluded

!
Page

Subsonic Flight ....................

Choked flow .....................
l>_ctrumble .....................
T_n-duct instability ................
Supersonic Flight ...................
Causes of unsteadiness ...............
Character of unsteadiness ..............
Preventicn of unsteadiness .............
__CE
.......................
AIRCRAFT-INDUCTION SYSTem4 ................
Effects of Inlet Location ...............
Subsonic flight ...................
Supersonic flight ..................
induced Effects of _g_- of A_ _ck
..........
Bodies ......................

wings.......................
Effects of Forebody Boundary Layer ..........
Boundary-Layer Removal ................
Suction ........................

"_
;

:
;

_:

Submerged inlets ..................

Combined
Effects
..................
Diversion
......................
Scoo_,lnc;_ementaldrag
...............
W_Lkes ........................
INDU_FION-SYST_4 AIRCRAFf ................
Drag .........................
_ikinfriction and separation ............
_ausonic drag rise .................
Wave drag
Lift and Pitching Moment
Wing leading-edge inlets ..............
Wing-root inlets ..............
Scoops .......................
Nacelles ......................
APPENDI_ A - SYMBOLS .......................
APPENDIX B
............................
REFERENCES ............................
BIBLIOGRAPHY ...........................
FIGURES ..............................

68
68
69
70
71
71
73
74
76

77
77
77
79
81
81

8_

_-

86
87
88

_
_,

92
93
8_
93
94
94
95
95
95
96
98
99
99
I00
i01
104
108
113
138
171

41,

k
_ i

"2_,_

C(Sr_IDENT_

NACA EM A99F16

NATIONAL

ADVISORY

';

C0MM_TTEE

RESEARCH

AERODYNAMIC

_"

FOR AERONAUTICS

MEMORANDUM

PRINCIPLES

F0h _

JET-ENGIneS INDUCT!0N
By Wallace

: i,

DESIGN

OF

SYSTEMS

F. Davis and Richard

_cherrer

I. INTRODUCTION

!
_
'_
_'

An air-induction
system conveys air from the atmosphere to the
engine of an aircraft.
Its purpose is to supply, under all flight conditions, the air needed for best operation of the engine wlth the ].east
disturbance to the external flow.
In other words, to avoid penalties in
engine size, weight, and fuel consumption: an induction system must supply
air at the maximum presto,re and with the least drag and adverse interference po'_sible. The flow to the engine must be sufficiently uniform
and steady to maintain engine performance
_md to avoid vlb_ation and
structural failure.
The significance
of tae air-induction
system in

,;
_

high-speed-aircraft
design has been well illustrated by SuJ_kin in reference i.
It is sho_m "_hat for fighter aircraft flying at Mach numbers le.ss
than about i.I, the: p_-essure losses throt_h a typical normal-shock
inlet

!_
_:.
_

cause a loss in engin,_ thrust that is equivalent to less than i0 percent

of the wing dr_._; '_hereas, st a Mach nu_ber of 1.6, these pressure losses
reduce the engine thrast force by an amount equal to the wing drag.

A sizable quantity of research has been dil'ected toward finding

solutions to the problems of air-inductlon
system_, particularly
In the
Mach number range from 0 to 2; but th_ results have not been consolidated
Into an organized group of design principles.
Kuchemann and Weber have
written a textbook on propulsion (ref. 2) and present some discussion of
air induction.
Ho_ever, further consollda_;lon of information i_ required,
particularly for supersonic aircraft.
It Is the purpose of this report
to assemble principles of Induction-system
design for flight to a Mach
number of e and to use existing data to show the consequences
of compro-

_
:_
_
:_
;:
_.

mlslng them.
an extensive

._
_i

bibliography
based on this search is appended to the present report.
The bibliography
lists "reports published since 1948 and thus extends the
blbllography
of reference J. The authors acknowledge
wlth gratitude the
assistance given by Mr. Emmet A. Mos._man, Mr. Forrest E. Gowen, and
Mr. Warren E. Anderson In csx'rylng out the literature ser_rch and in maklng

'_:
.:i_,
_

other

_,_

<i
i_

In order to accomplish thls task it _s

necessary to make
search of existing literature on alr-lnductlon
systems.
A

contributions

to thls re_

_;

J
I

"

_D_E_L_tAI_

o,

CA _

A55FI6

The design of an air-iaduction syst_m for an aircraft is greatly

influenced by the design of both the airframe and the englne_ and the
performance
of airframe and engine can be seriously affected by the
induction system.
Therefore, the problems of air induction must be con-

sidered from an ovcr=_ll vic_olnt_

az_d _. broad outli.ue must be selected
tc relate design principles.
In this zeport, the problems of alr-induction
systems are arranged according _o the following outline, and the principles
that ha.'e been established for their solution are presented under the
app_opliate problem headings.
A.

B.
_C_

Definitions are presented to describe the forces involved

the terminology used in air-induction-system
design.
The relationships of the induction
engina are discussed to indicate
_
siderations.

_ r

and

system to both airfr_ne and

the preliminary de:_ign con-

The detail design problems of ensuring high p_rformance

of an
isolated air-lnduction
system and then of maintalning this
performance when in com_blnation with ether alrcraft components
are discussed
1.

'

under

'.

two beadlngs:

Induction, that is, the pr_ss'_re-recovery,

uniformity, ar_d flo_-_eadiness
problems
supplying air to at. engine.

d_ag_ flowencountered in

ft. Interference_
or how other parts of an airframe
induction system and vice versa.
This arrangement

is illustrated

by the following

affect

the

chart:

Air-induction
systems
I
Definitions
Preliminary
Aircraft

tv

co.nsideration s
I
requir .e_ents

r_Airframe-i.,_ductionsystem combinatlun

_.

i
!
_
I

,
Engine-ind.,ction.system combinstion

'

I
L-tail

considerations
Interference

lTessure
Drag

Flow steadiness

._,

.....

recovery
and uniformity

;, , ,,-....,_ .............
,_ ,,,_.

Airframe-induction
system I
Inductlan-sy_tem
airframe |

_.

I
,

,_.

Y_
t

In order to discuss induction-system design over a wide range of

operating conditions, it is necessary to have a consistent termluoiogy.
The definitions that have been selected for use in this report have all
been used previously; and in the man_ instances where several terms have
been used by various investigators to indicate the same concept, the
choice made here is based upon considerations of consistency, popular
usage, and conven_ence.

AIR-INDUCTION SYg_EM

To define the major factors involved, consider the general arrangement of the following sketch:

Exlern@lsurfaces
InternatIwfoces
Streamline

---.-----

!
--

"_

/_
_o._

._ _

_;_ ,o.,._

Lip 3

Sketch (1)

" '"

The alr-inductlon system (E_ations 1 to 3) is a part of the propulsion

system (stations ! to 4) sac is defined to be that portion of am aircraft
whose purpose is to convey air fr_z the atmosphere to an engine. The
induction system includes any measures taken to compress or divide the
oncoming air stream that eventuall_ flows through the engine, such as the
ramp and boundary-layer bleed (stations I to 2) shown in the sketch.
The inlet is at station 2, and the inlet area is measured in a plane
tangent to the most upstream point Qf the lip and normal to the mean flow
direction in this plane at maximum mass flow and zero angle of attack.
If the entire cowl lip does noz lie in the inlet plane, the inlet area is
taken as the area outlined by the forwardmost points on the llps projected
onto the _nlet plene. For particularly distorted inlet sha_es, these
definitions are not al_ays applicable; in such cases, an area should be
_-

. --......

.......

;.........

,_.

................

......

--

/':

.-

I
/
/

:;

_(_?.DE_:

i : ": : !NACA EM A55F16

chose_
is the most representative in terms of _nduction-systemj
_erfor_
c_
Many specific definitlon_ of inlet area have been employed
_W-_BB-literature; two of -;_hese
_Ich are particularly useful are the
capture area; the axial pr _jec_ionel_the _n!et ar_._and c_mpresslonm,_face frontal area onto the plane of station l, and the minimum cross_ection area, station 2'. Each of these definitions is convenient in
certaln cases, and they are identical for sharp-llp normal-shock inlets.
The duct (stations 2' to 3) in the general case includes an area and
shape variation along i_s length, bends_ and a plenum chamber. The _uglne
intake is at station 3 and is considered to be upstream of all components
that are normally supplied with an engine and that are present when static
tests of the engine are made. It .isthus ahead of screens and swirl vanes.
The inlet lip and the fairing of external surfaces .intoother parts of the
aircraft are considered to he prcblems of the induction system.

Ii
I

Generally speaking, there are two characteristics used to identify

alr-induction systems; namely, tDe location of the inlet on an aircraft
and the method used to produce compression upstream r.fth_ inlet. For
example, induction systems are denoted by such terms as nose, side scoop,
_lug-root, conical-shock, or internal-_nntr__ctioni_l_..q:and these
expressions _t'_combined for more complete designations.

DIVISION GF FORCES

The division of forces between a propulsive unit and nther parts

of an aircraft must be carefully defined to ensure consistency. (See
ref. 4_ for example.) The air that flows through a Jet-propt_lsionsystem
is coml_,res_ed,
heated, an,_*'.ben
expanded to atmospheric pressure w_+_
the r_.ction from the ensuing accel_,_ationof th_ gases used to ow_.rcome
the cestr_.Inlngforces of pressure and friction and to accelerate the
ai_c_'aft. *l_ne
division of the component forces that are included in these
t'arustand drag forces is, to a large exteu_, arbitrary, but for practical
z_easonsspe:ific definitions must be selected. The engine designer, having
no knowledge of the airframes in which an engine might be installed,
defines engine thrust .aithquantities that are independent of installation conditions. _ne ter_ used to describe the propelling force of an
isolated engine is the "net thrust" which is the rate of change of total
momentum (pressure plus momentum flux) of the gases handled by the engine
from the free stream to the tail-pipe exit. The aircraft desi@uer defines
the force available to accelerate an aircraft, that is, the net propulsive
force, as the sum of _
the fo_-ces,friction aud pressure, in the flight
direction that act on all the surfaces of the aircraft (both internal and
external) that are exposed to the flow of air. In using engine information
to calculate this net propulsive force, the designez must be consistent
because it is assumed in the engine data that the pro_alsive system
receives air with free-stream momentum, but in an aircraft installation
this _s generally n_t so. A correction _n_t be made for the difference

i
i
!
:

:_

_ll

l 11 1 II

111

11_

II

II

I I

IIII11

III rl 11

Ill

lllllll

Ill

"

......

:---

I
_.

r*,

_ .

ILl

rl

r_,
/

.......

NACA RM A95F16

CONFIDEN_qAL

,_

between the free.-streamand inlet total momentt_ in order to obtain the

net propu_Isiveforce. The following discussion i_lustrates the considera-

"_

tions whZeh are involved.

_:_

The net thrust force of sn engine is defined as (see Appendix A for

definitions of symbols snd sketch (i) for the positions in_.icat_db_ the
numerical subscripts )

Fn m m4V4 - moVo + A4(P4 - Po)

_!_
_

It is assumed in this equation that the velocity and pressure distribution

at stations 0 axed4 are uniform and steady and that A4 is normal to the
flight direction. The net propulsive force of an aircraft is defined as

_<

_i
-

(1)

Fnp -

in

(P - Po)dA - L_l
i

ex

(P - Po)dA + DVex

(2)

Here, the pressure forces f(p - Po)dA and the viscous forces DV are
the components in the flight direction, and they are divided between
internal and exterual surfaces, Ain and Aex,, A force tending to accelerate in
the
flight direct__onis
thus the pressure
reaction
from
the
accelerated
gases of a considered
jet enginepositive;
causes a positive

_
_
_
_

difference and a resultant positive force on the internal surfaces Ain.

The internal surfaces include those of the _ir-induction system (that is_
from the _tagnation point on the lea_ing edge of the ramp and from the
stagnation point on the inlet lip to the engine intake, station 3, in

_ii
._"

sketch
(1))
thethose
engine
nozzle
the exit.
surfaces
Aexandare
inand
sketch
(1)passages
from the to
forebody
nose The
to external
station

1 and from the stagnation point on the lip to station 4.

_!
_'_'_.

The first bracketed term of equation (2) less the force on the ramp
is, according to the momentum theorem, equal to the rate of momentum

_,j_
_

change between the exit _._d the plane which includes the stagnation
points on the inlet lip (for a three-dimensional inlet)

in

where

,_

'I

AI

I
t

_
eO

O0

@@
@_

area in the plane through the entry section enc].osedby the

stagnation points of the internal flow _I the llp; this plane
is here assumed nn_mal to T.heflight direction, _nd flowinclination angles _e

Fr

assumed to be negligibly small

_|
a
I

sum of the pressure and friction forces in the flight direction

acting on the rE_p; it is a negative force.

To u._,ili_eFn in determining Fn_, the equation for the former can be

rewritten as the sum of the rates 5f momentum change of the gases handled
by the engine between the exit and station AI and from AI to the free
stream
Fn = m4V4 + A4(P4 - Po) - MI

+ _

- moVo

mL

(4)

From equation (3),

Fn = I_Ain(P - Po)dA - DVin1 + Fr + MI - moVO

so, substituting in equation (2)

Fnp = Fn - (.\Ey-n_oVc)- Fz -I'_

_Aex
:

or

(p . Po)dA + DVe

f-

J
I
According to the momen_,_ntheorem, the rate of change o_ _:.>mentum
through
the boundary about a d finite volume of fluld is lug, '_.._he
resultant
of _he pressure integr_ over the 1'tee-fluidsurface -."ithe forces acting
on the fluid due to soll '__urfaces. (This statem_.n
t " ,he theorem assumes
ste_y flow and _o shesa_ !_crce_on the free-_u_d ,"_"':--.'e.)
For the
streamtube between AI and t_._f:ee Btresm,

PIVITM +/AI (PI"

...., .......
:'?:,'",',.""'
:"' .:''",,.,',
'-,".... ' .'"

Po'dA" mV _/AT_po '_0)dA"

--

"

.....

FB " Fr

_l_11I

._:_

NACA I_4A99FI6

-_

"CONFIUA_._

_ ;'

_ .

"

or

_'

AI

_J_ " mV = o

_"

(P - P)dA

" FB " Fr

(6)

where
FB is the body force b_tween the nose and station i in sketch (i)
acting on the air which eventually flows through the engine,
if the airinduction system has a boundary-layer bleed, as in sketch (i), which pre-

_.__
_.
.%\_

ven_s the boundary layer frc_ the forebody from entering the inlet, MI
would not Lnclude any of the momentum decrement of this boundary layer,
_o F B should then represent only the pressure drag on the s_rip of
external body surface which is affected by the flow to the engine.
Sub-

_'i_

stitutlng

/'
'_'/:'

equation

(6) into equation

Fnp = Fn -

-_,_
;
_!
._!_.
._._
j_.."
.i.
_

S
"_/
._
+

(9) gives the final relat-onship

(p - Po)dA + DVe x

ex

.i_

- po)dA

- F

(7)
B!

In subsonic flight, when the flow is neither separated nor an3_here

supersonic, the determination
of net propulsive force is s_newhat s-_mplifi_d.
For such conditions, the flow outside the boundary layer can be
considered irrotational, and D'Alembert's theorem states tha_ for a body
about which the streamlines close, the component of the pressure integral
in the flight direction must be zero over a bounding streamtube from the
upstream station at which the flow is undist1_bed to the similar downstream station provided, in the case of a three-dimensional
body, that
it carries no llft. l.ssuming"for ease of explanation that the external
flow reaches ambient _ressure at station 4 and that sketch (1) is axially
sy_netric, it follo:_s that
AI

_:o,-

:.
..
._i
,

a
(p - Po)dA +

_:

(p - Po)dA -- 0
AI

_j"

._',,,..
,_,;

:_"
'

Restating

the terms of equation

Fn__ Fn -

(7) in smaller

(P " Po)_

components

" Po)a_ + _x

- Po)a_ "

I (p-_o)dA-

,;

%,:
DVBI

(the integral designated

B
the nose to station l) so

is the pressure

.....

force

on the forebody

from

";_
._

_._

CONFg_DENTL_JJ.
q

,_

*e

_teB

:
etQ

-_: o":NACA _-_!

A99F16

6_

6._

oo

Fn_ = F_ - m_ex+_B

where

DVB

is she friction

(8)

force on the forebody

3urface that affects

the flow to the engine.

In equation (8) DVB (and in eqlmtion (7)

case
for the
of rotational flow) is the corrective term

AI(p-po)dA-FB

req_uired by the _finitlon

of the component forces of Fnp.
The engine
net thrust is the rate. _f m_entum
change from the free stream to the
tm_.l-pipe exit (eq. (1)), hut part of this momentum change
DVB
cannot
be cha_-ged to the internal flow because it i_ accounted for in the
F.xternal flow as a part of

DVe x.

avoid the inclusion

of

DVB

twice

in Fnp, the _x.entum at the i_itlal _catlon of the internal flow must be
corrected to lccal conditions, which means that DVB
_ast be added into
the equation for Fnp because the true inlet momentum is less than that
as definea (_,Vo) and thus tends to increase
FT.p. In the event the
boundary layer from external _u_faces is removed from the engine flow
by a b_Jndary-layer bleed _uch as that of sketch (1), F u is not affected
by th_s loss in stream m_nentum,
Then

and the correction

I>_B is unnecessary.

Fnp : Fn - DVex

(9)

Taking boundary layer into an induction system does not

of course, result
in only an additive correction, for F n decreases because of the loss
in pressure at the engine face _nd the decrease in m4 and V4 which _t
be suffered by an engine _th
a limiting design temperature.
Eowevel-, if
DVB

increases

faster

Fnp

as boundary

than

Fn

decreases,

layer is taken

there can be an improvement

into the _nuduccion system.

in
',

Quick in

reference 9 shows that for a certain engine a decrease in specific fuel

consumption and an _crease
in available thrust can be produced by taking
boundary layer from a forebody into the engine at flight speeds less than
about 300 mpn. At greater speeds, the thrust decreased r{pidly relative
tc that of an engine takLng in no boundary la_er because .;fthe increasing
comp2 esscr inlet temperature and because of the loss in dynamic compression 8_head of the engine.
(See also ref. 2, p. 209.)
If the p__essur_ a_. station

4 is not equal to ambient pressure,

/ A I (P - P)dA +/ 4(p - Po)dA + /_

o
AI
4

L-_"
I

----

.....

"

"

"-

7L

"

"

(p - Po)dA = 0

then

@_.;_

NACA _4 Ag.SF16

.and

'"_,#

CONFIDE_IIUikL

Fn-o = F n +

(p - Po)dA-

DVex + DVB

(]0)

?
*q

,_
_:
._:
%
_
i_
_..

in other words, a correction must be made for the mc_entum chs_ige occurring
in the
jet
which
affects
the flow and thus
the forces,
as p=ev-lously
defined, which act on the system.
This correction is a pressure-drag
force which acts on the external surfaces.
(See ref. 6.) The fact that
sy_ne_ry is not a necessary condition for the preceding equations for
subsonic potential
flow has been demonstrated
in reference
7.
It can
also be seen fr_n the fact that if a closed body, which according _o
the
asst_l
can have
drag, is drag
added
to the
aystem,
the flow
s_etry conditions
is destroyed
and no
thepressure
total
pressure
must
still.
be zero if the flow remains irrotationa!.

Z_:

_':
_

_t_ON4ANC_

P/_tANN_RS

The basic terms used in describing the performance of air-induction

systems are pressure recovery, drag, and r_ss flow.
A description of each
of these concepts follows.

_
5;

PRESSURE

EECOV_ NY

"_'_'_
_:
_,

Several.
terms
have been used
to describe
r,he performance
of airinduction systems in regard to their effectiveness in providing an engine
with hig_,-pressure air. The total-pressure ratio
Pts/Pt o is the average

_
_

total pressure at the engine intake

Pt._ divided by the total pressure
available from flight.
(Methods of measurement
and the determination
of

_
_

the effective
Pts
in non_uiform flow _e discussed in Appendix B. )
This ratio is used when an air-induction system is being considered in

--_
_'_

relation to an engine_airframe
combination because it is directly related
_o the net thrust and _he fuel consumption.
K_chemann and Weber show
by a simplified analysis of turbojet engines in reference 2 (p. 197) that

AFn
Fn-'-_
= Fni"Fnl Fna =L(I-

,_

"_.
_'

A(Q/Fn)
(Ql'_U)i

(Q/Fn) i
(Q/Fn)a
(Q/Fn)i

J__lll&

Pt_h
P-_o/

= (i - L)

(ii)

(12)
Pt

i0

;
,i :_

_ ...........
C@__AL:
.....

_
ii
quql ii

,,.

:l

@oe

s_

"':
.: : _ACA EM Ag_FI6
c,_s

t9

where

" vt-qJ j
+

-- -'gZT1

I
l

_j

jet eificiency,

l + (vj/vo)
Po
--

pressure

ratio

across

the engine

exit nozzle

Ptn
a

actual

installation

ideal installation

fuel

with

induction-system

without

induction-system

losses
losses

consumption
i

.i

_,_
_ depends on engine design and flight conditions and is greater
than i. _ decrease in total-pressure
ratio reduces the engine net thrust
and increases the specific fuel consumption with a greater effect on the
thrust reduction.
This occurs because the net thrust decreases with both

the mass flow and the jet velocity while the fuel that can be burned
decreases only as the mass flow for a fixed turbine inlet temperature.
(See _!so refs. 8 and 9-)
Ram-recovery
ratio (pts-Po)/(Pto-Po)
is the ratio of differences
in total pressure as measured at the engine face _nd ambient static pres-,
sure Pts'Po
and the total pressure and static pressure in the undisturbed stream
Pto'PoThis parameter is t_eful because experience ha_,
demonstrated
it to be only a weak function of Mach number for welldesigned systems in subsonic flow at a fixed mass-flow ratio.
(See
ref. 10. ) Thus, the results of low-speed wind-tunnel
tests can be extrapolated to high subsonic Mach numbers (of the order of 0.9) for condi-_ions
in which the total-pressure
profile at the inlet in flight is simulated
in the tests.l
Conversion from ram-recovery ratio to total-pressure
ratio is accomplished by the formula:

'

ii
i_

ZSee reference ll for a discussion of equiv_lent mass-flow ratios

to be used in low-speed tests s_mulating hlgh-speed conditions.
The
equivalent mass-flow ratio is one which produces the same pressure rise
ahead of an inlet at low speed as occurs at high speed and thus is useful
in simulating conditions for configurations
which nave a boundary layer
growing on surfaces ahead of the inlet. _ .....
L

C(,_!D_._EI/_L

NACA I_4A99F16

Pts " Po
_.

pt s
P+.

Pt-_

_u

ll

7 - 1

Poo

+ _

M2

- i

+ i

(i

7.i2

2)7_ I

C_es
of this variation for
7 = l.& are presented in figure
out thls report
7
is assumed to be uqual to 1.4.)
The parameter
_
_._
_.
_=
__
,_'
-'_
._
'_
_:
_'-.

i-

(13)

. -!_

[(Pts-Pto)/q2]

has frequently

i.

(Through-

been used to describe

losses in duct systams.

As with ram-recovery ratio, tests of subsonic
difusers
with unseparated flow have shown little variation of this parameter with Mach number; but, also, it is not directly related to engine
performance.
With air-induction systems, q2 can be estimated for most
operating condit__ons _ithout resorting to detailed flow measurements st
the inlet.
At the _hlgh mass-flow ratios which occILr in take-off, the
major losses in press_re occur at the inlet lips, and it is a fair asst_ption that Pt2---Pts Then, q2
can be calculated from the measured massflow, A2, and ptS . However, at mass-flow ratios of the order of i, the
major lo._ses occur in the d_ct and Pt2---Pto under which conditions it is
more reasonable to calculate
q2 on the basis of Pto" If the parameter
is used, _he conditions for the determination of q2 must be specifically
stated to a_oid confusion.

_&
DRAG

!-_
_
2
_.

ratio

The drag coefficient of an air-induction _ystem is the dimensionless

of force in the flight direction caused by an air-induction system

t:
_'i
_"
',_

being added
to an of
airframe-engine
combination to area
the of
product
of the
a_mamlc
pressure
flight and a characteristic
the induction
sys._em. As indicated in the previo_,s discussion, it is necessary to be
consistent in defining drag; the bracketed term of equation (7), the n._t
drag
E.,_, can be regarded a_ the drag force which is consistent with _he
definition of net thrust
Fn usually used in computing net propulsive

'-#.
"f_:
-_-__-_

force
Fno.
_ne bracketed term of equation (7), in the general case,
includes _uch more than the drag force of the air-induction sysbem, for
the drag of basic body, wing, tail, etc., must, of course, be included
in t ha net propulsive force.
However, for the present discuss.lon, it
is assumed that only a scoop arrangement such as that of sketch (I) is

_
_"
"_

-,4C

being considered.
The force on the air-induction system is the pressure
and friction z'orc_s caused by adding the scoop to a basic body plus the
pressure _ntegral on the free surface of the engine-flow streamtube minu._

"-_

"

I;
!
12

":[ " CAm

:

the body
integral

t -

_t

esl

4e

forces acting on +.his stre&utube. 2

This difference of pressure
and body force has been ca/led the "scoop incremental drag."

(See refs. 7 and 12..) In the present devclopment,

the ramp was considered
-_t
of the air-induction
system, and the force on it does not appear in
the scoop incremental drag.
However, if a ramp (possibly because it is
a portion of a colony) is considered no5 a part of the internal system,
but to contribute an external force, then the portion of it affecting
the engine flow must be included in FB
of the scoop incremental c1_ag.
if the configuration
has a nose inlet and there is no forebody acting on
the er<g:_neflow,
.
tnen only the pressure integral from the inlet to the
free stretm is effective; this force has been called the "additive drag."
(See refs. 7, 12, and 13.)
The "external drag" of an air-induction
system is the sum of the pressure and viscous forces in the flight direction acting on the external surfaces of the air-induction
system.
Many
reports on inlets define "external drag" as the sum of external pressure,
friction, and scoo_-inc _mental
drag forces; to prevent confusion, this
sum is called "net drag" in the subsequent discussion.

2
;_
,_
_
_

MASS FLOW

The mass-flow ratio used to describe the flow through air-induction

systems is the mass of air that flows through an inlet din[dea by a
reference flow rate

ma
--

mref

A2

pVaA
(14)

_ArefpVdA

(A discussion of mass-flow measurements

i_ presented In Appendix B. )
Many choices of the reference can be _ade, each hav_,g Fome _dvantage
for particular conditions.
In thls _port,
t-_o reference rates are
usually used:
i.
_

'_

The mass-flow

ratio

ma/m o

is based on the reference

_O=PoVQA2

which can be readily determined.

In subsonic, incc_np;_essii,Z
_ flow, me/m o
reduces to inlet-velocity
ratio
Ve/V o which has often been used to
describe air-induction-system
performance,
This definition of mass-flow
ratio has the disadvantage
that in mlpersonic flight it can be greater

than 1 if the inlet is located in a compression field whereas a definition

based on capture area h_s a maximum possible value of 1 if local flow

il

mAs indicated perviously,

if a boundary-layer
bleed removes all the
boundary layer from the streamtube enteri_,g the inlet, the body viscous
force
DVB
is part of the exteroa_.._
must not be included in the
bod_ force acting on the enEA "
. ......

_
I

,,

I'"
I

I
, .

i
|

AS rl6
_;_

properties

_,

_"
_'
]_.;
;

' :., : :,,

13

evaluate

are used.
However, in the general case, mo
is easier co
=fA
DVdA,
and
in
subsonic
flow
both
ratios
can be greater
than m c
cl

than 1.

(See p. 4 for definition

of capture

area

'

Acl.)

/ 2. The mass-flow ratio

m2_/m2,*
is used for _he static condition
when Vo=O.
This ratio is based on the flow rate for choked flow at
station 2'. The m_ss flow, m2'*, is equal to p*V*A2,
where
p* and V*
are the density and velocity for flow at a M_ch number o_ 1 at the prescribed ambient pressure and temperature.
This ratio has been found to
correlate data well, and it indicates how near the flow quantity is to
the maximum possible.
As will be shown later, t is a criterion of the
excellence of lip design for low-speed flight.
For flight speeds other
than zero and for isentropic
are related by the equation

_.L_;

0.579

flow, the two definitions

m2--_.
A2 ( 1 + _
ma' Am'

of mass-flow

ratio

Mo a_a(7.1)

- i

Aa,

"

sharp lip inlet, from reference

14, is also shown in figure

2.

llI. I._WS_L_NARY CONSIDERATIONS

"! !il
_(
__

which

is plotted

in figure

2 for
AIRCRAFT

= 1.O.

The choking limit for a

REQUIRE_--_.,_
_TS

"_"
_

As discussed
in reference
basis
for the
choice of both
airframe 15,
and aircraft
engine. requirements
Since one of are
the the
considerations of airframe design is that of the induction system and since the
engine performs_ce is affected by the internal aerodynamic problems of
induction, the considerations of the air-induction system enter into
the preli_tlnary layout of aircraft; and they must be viewed from the
standpoint of the flight requirements.
Aircraft range and endurance,
for instance, are dictated by fuel consumption, which is affected by the
drag and pressure recovery of the induction system.
Similarly, take-off
distanoe, rate of cl_mb, maneuvering accelerations, etc._ depend upon
net propulsive force and hence on induction-system
drag and pressure
reoovery.
Aside from these performsnce requArements that vary w__th aircraft parpoae, there are other, less tangible, requirements that must be
taken into a_:ount in any design.
For example, safety, vulnerability,

_!_
_

and serviceability consideration=. _ffect engine location and thus the

type of air-lnduct_.on syste.mr: The emphasis on any psA_ticular requirement

'_
_:
__
i_'

-'

14

CO_D' .a_.__.

{ "'" :"{

NACA EM A95FI6

depends upon the l;_tended mission.

Thus, the /[esign of an alr-inductlon
system must be adapted by compromises to suit many requirements in
various degrees.

AIRFRAME-INDUCTION-SYSTEM

COMBINATION

To illustrate some of the problems encountered in fitting sxl Enduetion system to an airframe and to introduce some of the types of inlets
that have been developed for various engine locations, the progression
of design problems with increasing size of airplane is briefly discussed.
C_-rrent design practice for high-speed turbojet-powered
indicated by the following compilation:

Airplane

length Number

el_

Engine

diameter

F-86D
F-86F
F4D-I
FSU-1
FTU-1
F-IO0
F-8_E
XF-I04
XF-105
F-89
F4D-2
F-IOI
B-97
A3D-!
F-IOPA
X-3

14._
14
15
16
17
17
17
18
18
20
20..9
21.9
22
23
24
30

B-47

40

B-_2

44

of

x ettypeand I

engln_=s

i
1
1
1
2
1
1
1
1
2
1
2
2
2
1
2
6
8

aircraft

/_----_e

location

Fuselage nose scoop

!Fuselage
nose
i
Wing
root open
Fuselage
nose scoop
Fuselage side scoops
Fuselage opn nose
_h_selage open nose
Fuselage side scoops
Extended wing root
Fuselage side scoops
Extended wing root
Wing root
Nacelles, open nose
Nacelles, open nose
Fuselage side scoops
Fuselage side scoops
Nacellos, open nose
Nacelles, open nose

_
_
_

can be

Xe h
diameter

9.5
9.5
4.91
9
6
9
6

9.71
71
2
9
3
1.9
i._
I01
3.9
1.9
1.5

,"

iThese airplanes have two inlets for one engine, and the ratio of duct
length to engine diameter is for a reference diameter corresponding
to half the engine frontal area.

!
_

Airplane size relative to the engine is indicated by the ratio of fuselage

length to engine diameter.
For small airplanes with one englne_ in which
this ratio is less than 18, an inlet located in the fuselage nose or
underslung
Just behind the no_e has been used most frequently.
From the
induction-system
standpoint, such locations are desirable because the
problems associa*ed with boundary layer fl_wlng into the inlet are either
eliminated or minimized.
_e underslung inlet, in addition, maintains
_

": .....

," , , ..........

:..........

H H .....

II

II

-.....

_'

NACA _

_
__
_
_
_"
_.
_
_

A55F16

C0_FIP_._

....

l_

performauce at off-design positive angles of attack because the flo_ is

deflected into the inlet by the nose. A_ the ratio of fuselage-to-englne
size increases, or if nose _-olumeis required for equipment, 3coops f_ther
back _n the fuselage or wlng-root inlets s_-eused. From the induction
standpoint, an underslung scoop position is again desirable because of
the off-design angle-of-attack performance _d because the body bounds_y
layer is the thinnest on the windward side_ '.This
I,o_tion has, however_
been avoided because of the possibility of foreign-object damage to engines
during _um-up, taxilr_, or take-off,s The wing-root inlet has a possible
advantage over _coops in that the portion of the inlet perimeter adjacent
to the body can be relatively short, thereby reducing the proportion of
body boundary layer flowing into the i_let. _ku_the_._more,
with m_Itiple

engiues the ducts cau be short and the flow unimped_

_i

wing
aircraft,
thethe
wing-root
inset
a region
of large
induced preflow
angles,
both from
body _ua
wing is
atin
subsonic
spe_s,
so special
cautions must be taken to instu-eadequate perfor22nce at off-deslgn angles
of attack. For a high-_ing airplane, a design probl,m_of the wing-root
inlet at angle of attack i_ the thick bounda_ layer on the leeboardside
of the body.

_'
_,
_
_
,
.i_.
_
__:'
_
d_
....
_b

by beng_so For mid-

For aircraft of greater relative size (fu_elage-length-to-englnediameter ratio _ 22) there are several possible location_ with the choice
depending on many considerations. For engines clustered in the fuselage_
scoop inlets can be used; for engines in the wlng-root or buried in the
wing, wing-root, _Ing-leadlng-edge, or, for very large aircraft, underslung wing scoops are possibilities. However, nacelles with a simple
nose inlet have been used most frequently. Such _a_ramEementsare deslrable from the air-induction standpoint because the ducts are short and
straight and the problems of aircraft-induction-system interfe__nce ere
generally reduced.

_:
'_

_.-2_GINE-INDUCTION-SYS_EM
COMBI_ATION

_
,_
'_!
_
_,
",_
_
_
.:.
:
_'
_',
_!
,_

The performance of a propulsive system depends not only on the

individual characteristics of the air-induction system and of the engine,
SThe studies of references 16 and 17 indicate that the flow into _.n
airplane induction system can seldom llft damaging objects by itself. For
instance, an inlet whose center line is two inlet diameters above _he
gro_,l _d through which the flow velocity is 700 feet per second cannct
pick up sand particles larger than about 0.02 inch in _ameter u_less a
vortex forme between the inlet and the ground. However, such a vortex
can form under th_ proper conditions, and if the damaging objects on the
ground are restrained laterally, as they would be if lodged in a crack in

._

a
runway,
the vortex
will
suck
into
the the
engine;
or,
if objects
which
cs_
do damage
(see ref.
18)
arethem
thro_u
into
air by
some
other me_'.s,
the engine can easily draw them into the inlet. Foreign.-objectdau_e to
engines _s generally considered to be an operational problem, that is, one
of using screens, of polici_g._mPs ._ r_n:raysand of pro_r taxii_ proced_es_ rather than a fact_
t location and airframe design.

_
_!.
,_
._
_

!
,_
.'

ee

oe_

eo

_,_o

Qe

but also on the compatibility of these characteristics through the range

of flight conditions. This problem of compatibil._tyarises because ram,i.
._
_
_ :

jet or turbojet engines require a specific schedule of air flow to achieve

rated th_st through the flJg_htMach number and alt_tude ranges. The
flow through a nonadjustable inlet combined with an engine varies with
fl_gbt conditions and deviates from the optLmum conditions selected for
the critical design point. If the range of operating conditions is sufficiently wide, the air-lnduction system is complicated by adjustments
that must be pr_ided to maintain its performance near optimum.
The general problem of combining an alr-induction system with an
engire can be divided into three parts: (i) matching, (2) optimization,
(3) evaluation. Matching is the determination of the mutually compatible
operating point for an engine and air-induction system at each flight
condition; it consists simply of relating the engine flow requirements
to the air-induction-system characteristics by means of the continuity
equation to determine inlet area or mass-flow ratio for prescribed operating conditions. Optimization is the determination of the matching conditions for maximum net propulsive force or minimum specific fuel consu_ption. This can consist of the calculation of the optimum inlet area
or mass-flow ratio for fixed system8 or of the proper variation of inlet
dimensions for variable systems. The two problems, matching and optimization, are presented in same detail in the following discussion. Fv_luation is the comparison of sever_! possible propuls1_.c_j_ems on an
airframe to determine the best system for a certain mission. Evaluations
can _Ivolve many considerations in _i_tion to those of aerodynamics,
such as structure, _eigh_, mecLanical complexity, etc. However, by
restricting the propulsion-system variables to net propulsive force and
fuel consumption fo_ prescribed flight plans, many valusble results can
be obtained from an evaluation study. For example, Fradenburgh and
Kremzier in reference 19 describe an evaluation of the effects of w_:cious
propulsive systems on aircraft range. Another approach, which is similsr
_o that used by W_worth and Kelber in comparing Jet engines (ref. 20),
is to determine the allowable weights fo. the _Ast_D_lationof each of
several air-lnduction systems on an airframe having a prescribed range.
Such an evaluation provides the designer with the information necessary
to select possible mechanical arrangements_ These studies are part of
the general problem of power-plant-aircraft optimization discussed in
reference 15.

._
,

[
_

W_

MA,Ib_IKING

The problem of matching an air-induction system and an engine requires

knowledge of the performance characteristics of each, and the problem of
optimizing the design for a special airplsne requires knowledgc of the

._

__

, _m=
i

i
'

'

--L_

.]lk_l

___2

"

_,_v_l

J ....

JJ-

t'

'

,_ 11'

NACA P_ Ag_F16

COE_DENTIAL

....

1.7
7.

characteristics
through _ wide r__a%e of flight conditions. 4
These characteristics are determined by analysis and tests, but since in the preliminary stages the air-induction
system has not yet been designed, its
performance _n_st be assume_ from past experience or by determining what
performance
is necessary and then striving to dezign and develop an
_'r_em_nt
that will accomplish the goal.

,
_i
_
<
/

To illustrate

<
:
..:
_

a method

for matching

a turbojet

engine

and an air-

induction-system
combinstion, the variation of corrected weight flow of
air for an engine (Wac=Wa_8/8) as a function of Mach number and the variation of the pressure recovery of the air-induction
system with massflow ratio as shown in sketch (2) are assumed to be known.

?i,
Wae

._/

Pt--!
Pt o

n,h

Mo, a,_8,A2/A.
Mo

'-:.

roW'too

_.

Sketch

_
_'
%.
'_il
_.

(2)

For a complete analysis, this information must be available for each

parameter indicated on the sketch; that is, the flow variation must be

'_

known for the expected range of engine rotational speed n

and of flight
altitude
h.
The induction-system
variation must be known for the Mach
number
Mo, angle-of-attack
_, and angle-of-sideslip
_ ranges, and
posslb_y for a range of the ratio of inlet area to body fron+al area

:_>
'-

_2/AM, although in the usual case changes in this ratio are small and
their effec%s are negligible.
Transposing the continuity equation

'_
_,_
_
_"

(assuming uniform flow at all stations) into engine-inlet

terminology by
_See reference 21 for a discussion of engine performance psrameters;
reference 22 for an analysis of turbojet-engine-inlet
matching; references 8, 23, and 24 for relationships between engine and induction-system
performance
and methods of determining optimum performance
conditions;

;,:

and references
mi smatchin_.

.......

2_ and _

Illlll

for studies

II

IIII I

of the penalties
' ........

--

I .L_..]II

II I

associated

Illllll

.
!

_Ith

I I

""i"
,'

II

.,,.]IIsL_i_II

,f

i
!

defining

8-pts/2SL,

8o-=Pto/PSL,

mm_

and q-e- T_s/T-_L

peV_A._

s
_'_

Ao

=o poVoA21_;

gives

.:
{

w_J_P_
A28

m2
.7'+1

Pto - gPSLa_SL _

= 8_._
_

';

-N_

._

(16)

l+

when

i.4
32.] 7 ft/sec 2

PSL

0.002376slugs/ft

ASL

lll7 ft/sec

ii
;
:

'.

This relationship
can be represented graphically so that from the kno'.m
engine and air-inducticm
system characteristics the inlet area required
to match the engine at the _elected induction-system
conditions can be
readily determined as illustl-ated in fi_e
3. Thus, for a given flight
condition o_ Mach number and altitude (sketch (2)), a mass-flow ratio
is selected and the corresponding pressure ratio determined from the
alr-induction-sy_tem
performance data; _he corrected engine weight flow
is determined from the engine curve; and the p_oper inlet area is determined by the intersection of the corresponding horizontal and vertical
lines in the third quadrant of figure 3. This inlet area f'zrnlshes the
engine the pruper volume rate of flow at the chosen mass-flow ratio, but,
this is_ o: course, not necessarily the mass-flo_ ratio that produces
the msmAmum net propulsive force or the minimum fuel conscription.
A similar method

ditiO_S

can also be used to study matching

where the _ass-flow

ratio

m_/mo

at static

has no significance.

ccn-

Defining

"]I

_- ....,
:_

.......................
"_..........

_ '"'
' _"'_-"
...............
'-

"

-=

_L

NACA I]4A_GFI6

o,

|_$

'..

m sll

C,)N_ID_.NT!_'. : '
Ii

inlet Mach number M2'

2 were isentropicj5

19

as that which would exist if the flow to station

Wa_
,.
_'

J_

Pts

A25 Pto

89 4M2r

(17)

_ ,2_e
(I + 0._a
,

?
_i

This equation corresponds to equation (16) if ma/mo=l and M2

is substituted for Mo. With these changes, firo_re3 can be adapted to static

r
!_

conditions. Information on p_/pt o as a function of

converted to a function of M2' by tne_:relation

me/me*

"

can be

me*
i

ma = 1.728M._'(I+ O._e'2) "s

(18)

and this
used
to determine
variationthe
together
inlet area
wzth required
the }mownto
engine
matchch_cacteristlcs
the engine or the
cad be
penalties resulting from mismatching.
OPTIMIZATION

<

:
!
;

To determine _:heinlet areas far maximum net propulsive force &_er

a range of flight conditions, the net thl_istof the engine Fn, the correction to e_gine net thrust due to pressure losses upstream of the engine
AFn_n
(see ref. 24) and the net drag of the air-inductlon _ystem, as
2_Pts/Pts
_hown in the following sketch, must be known:

n,h

in'h'-l_''t
Mo

Mo

Mo,a.B.AsIAu
me/ mo

":,'

Sketch (3)

"'_

_A prime symbol is used here with R_ to lndlcate that the number

represents a fictitious condition and is used only f,.'m"
convenience. As
will be sho_ later, the flow through :Lnlet_with practice_ li_ shapes
is not isentropic at 't.a]_e-of_""...............

-.!j_
.._"
i_

A,':V

o-

:....

Then, for the conditions for which

Aa was calculated
the net propulsive force can be detelnnined as

to match

Fnp = Fn - _Fn - Dn

the engln_,

(39)

l_ne opti_r.m inlet _eas fcr a M_ch nu_er

range at a constant altitude,
en_-_nP__.p_.
_
_,nA.=<_l_
_+ titudc ar^__,,_=
b_ _ui've_ oi"net
propulsive force as a function of inlet area as shown in the following
sketch :

F._

/_z
Sketch

(4)

..

Such curves provide the information required in fin _ evaluation, that

is, the penalties in net propulsive force that would result from flight
_th a constant inlet area or any other deviation from the ideal _ariable-

)"
T_
"

area system

_;

that might be required

considerations.
other variables,
acceum.t.

by mecb2_ical;

structl_sl,

or flight

Of course, to optimize for a _rescribed _.Ission the

such as altitude and P_'Le of aitack_ _st
be taken into

I_
;'"
)_"
4

Maximum net propulsive force is _npoz'tP.;:t,

but it i_ not _Lwnyl the
cr_tlcal design consideration.
For instance, w_th lor_-rs_e
aircraft
the fuel consum_ption per pou_
of net thrust might be more importsnt.
The procedure for optlndzing this parameter is similar to that just
described; fuel flow rates corresponding to the ealcriated net prop_Isiv,_
forces are determined from engine perforr_Ance curves, 8_,_dthe ratio
Wf/F n is plotted as a function of inlet .%tea for the range of flight
conditi_s
to determine the optimums.
'l_e inlet _/'ta for minimum specific
4

NACA _4 Ag_-FI6

C0k_+fDDNT/AI] '' : :"

- -

. _

s
s

01o

_
i
bl

21

fuel consumption is, in general, different from that for 1_ximum net
propulsive force, but for a well-designed
air-induction
system the difference, which depends on the diffez'ence in the mass-flow ratio for maximum pressure recovery and for minimum net drag, is usually s_ull.
The
importance of this difference depends on the intended mission.

FLOW L_IFOR_TY

_._/
-i
*_*
_
._._

_
_i
_
_
_
_.
_
z_
_.
+
,_
_

_
*!_
_,
_

AND STF_DII,_S_

Another problem of the engine-Luduction_system

cc_nbination is the
uniformity and steadiness of the flow that the air-induction
system presents to the engine and the effects of irreg_laritles
on engine performsmce.
Irregularities
in pressure at the face of a compressor, parhicularly
an axial-flow compressor, can reduce engine perfo_nance
and cause vibration; pressure pulses or fluctuating
flow angles can cause st1_ctural
failure of compressor blades.
Tolerances in flow uniformity h2ve been
suggested by Greatrex (ref. 27), but steadiness tolerances have not been
established
(see ref. 28).
The indications are that these tolerances
depend upon individual engine design.
Conrad _nd Sobolewski (ref. 29)
found that flu'_ nonuniformity
that was once thought to be unacceptabl_
had no large effect on the engine which they tested; however, the tests
of reference 30 with a different engfne shewed large reductions in performance.
In the investigation
of flow steadiness reported in reference Bl, it was found that, although the induction system by itself produced unsteady flow, operation with a turbojet engine had _ large
attenuating effect.
Differences
between engines in response to flow nonuniformity
can
often be explained by the f _ct that a compressor with a large pressure
rise across the first stage has blades operating at high lift coefficients,
sn& izregularihies
in the entering flow readily cause stall.
A first
stage _ith smaller loading can reach local stalled conditions only if the
,_,n_eringflow is more irre_alar.
An induction system with flow nom_ifortuity sufflcien, to stall one or more blades leads to the phenomenon
called "rotating sts/_l" of the compressor with ensuing reduction in engine
performsmce
(thrt.st, allowable fuel consumption, and acceleration margin)
and large vibratory stresses in the blades.
(See, e.g., refs. 32, 33,
54, and 35.) Since the trend in the design of co_@ressors
for the engines
of supei'sonlc aircrsd't ._s toward larger flow r:_es and pressure ratios
and toward lighter specific weight, blades are being made longer and
thinner a with the result that the induction-s_stem
problems of flow
In this report, a di_:tinction is made between the problems of flow
stability and steadiness which has o_ten not been made in the past.
By
siability is meant the property of flow which enables it to return to _'_"
original steady condition after being disburbed; thus, a normal shock
wave is unstable In a convergi_
channel because it can exist in a steady
condition only upstream of the inlet or downstream of the throat.
By
steadiness is meant the qua_.t__.___e
f l_-: in regard to velocity or pressure fluctuations.
_ _
\

*_

".,iZ

,,. __ _

,...

22

C_NF_DE_A_

: -

NACA MM A55F16
4b

uniformity and steadiness are becoming more critical because of the greater
likelihood of rotating stall and of structural fail_._e. Even if a cornpresser is designed to avoid rotating stall, the effect of intak.e flow
r._c_
distortion is to _ve
the compressor surge line to higher eo_
....
wei_wht flows, and thus toward the operating line, with an ensuing decrease
in the operating range possible with the engine.
_iso, the results of
reference 35 indicate that nonuniformity of the flow from the induction
system can cause nonuniformity
in the temperature distribution at the

turbine entry with subsequent turbine failure.

With ram-jet engines,
adverse effects also result from irregalar flow from the air-induction
system.
Reference 36 reports large losses in combustion efficiency on
account of variations in velocity profile at a burner, and references 37
and 38 show that pressure pulsations must be avoided.
Flow uniformity is related to the problem of en_ue
location.
Such
factors as the induced effects of other __ircrsft cc_ponents and the length
and path of ducts must be considered in preliminary desiga _e produce an
air-induction system with uniform flow at the engine face.
S_eadlness
of the engine flow, particul_rly
in supersonic flight, is affected by the
operating mass-flow ratio of the induction system,
in general; u_teady
flow results from operation at low mas_-flo'_-ratios, and the associated
pulsations can be violent.
For safety, the flow must be steady from the
opera%ing speed to the windmilling speed of the engine; and a variable
inlet area or an air bypass
may be necessary to maintain high inlet
mass-flow ratios.
Considerations
of these problems in relation to inlet
design are discussed subseqaently.

IV. DETAIL

CONSIDERATIONS

INDUCTI ON

The purpose of this section is to discuss the pressure recovery:

drag, flo_- uniformity, and flow steadiness of air-induction systems without describing in any detail considerations of other aircraft components.
These latter factors are discussed later under the heading INTER_CE.
The flow inside ducts can be treated independently of the flight Mach
number, and this subject is presented first under the heading PRESSURE
RECOVERY AND FLOW UNIFORMITY.
In general, the problems of conducting
air to an engine are described at subsonic and supersonic speeds to a
Mach number of 2.

!
i
i

}
'i
_i
;

It should be mentioned at the outset that insufficient theoretical

and experimental information is available to predict accurately the performance of practical alr-lnduction systems through all the possible
combinations and ranges of the many pertinent variables.
Fox" all but the

simplest

purpose

cases,
here

refined

design

is to discuss

must depend upon test observations.

what is known of basic design

o-

,-_._

,.__.,:

_'

The

principles.

t ",,,

'_-,_,,.

"

"

"F
NACA }_4 AS_F16

C0_T_DF_TI._I

PRESS/HE

RECOVEHY

_,

23

AND i__/_ b2,_9._iTY

The design objective in regsa'd to pressure rec ( very is to provide

a passage by wh:.ch the air requir_-_/for best operatl)n can flow to an
engine with the least pumping power requir_.ment a5 zero flight speed and
by _rhich the c_pression
available from the kinetic _nergy of flight can
b_ u_ilJzed to ;he maximum extent.
'i_necompression
_f more air _.e; unit
of engine intak._ area permits more fuel to be b,_rn,i for the spree limiting
temperature wita a resulting increase in the s;ecific thrust for a smaller
specific fuel consumption.
In other "words, as _hown by "--quatJons (ll)
and (12), the total-pressure ratio must be high, for losses affcct _,_rust
in more than a l:i ratlo.
The problem of flow uniformity Is discussed
together with _ressure recovery in this section because the two problems
are closely allied

;
:'

:
:_
,\
],

Ducts

j-

:
_

There is no general method for designing the ducts of practical alrindu-.tion systems because the flow in the usual case is viscous, compress:ble, and three-dimensional.
A su,,nary of preseI:t knowledge of duct
flow is presented here to develop empirical design rulec.
L_e two primary ge_:etrlc factors which are of concern are the inlet-to-engine-face

_
J;

area ratio and the duct path.

The area ratio is determined by the selected
de._ign conditions, and the duct path, or the length and offsets, is deter-

,_
&
.&
-'
,_
_

mined by the aircraft configuration and the necessity for avoiding press_ce losses.
The aerodynamic factors of concern are the initial flow
distribution azd the conditions which cause _ressure losses and nonunzformity in the flow,
The problem is to determine from consideration of
these factors the shape of duct that produces the best ope_ati. _ conditions
for the engine with the least cost in weight and complexity t,_%ne airframe.

_'

Area ratio.- In regard to the area ratio between _hc i_l__t and the
engine face, by assuming uniform, adiabatic flow of a perfect gas and
using the continuity _quation, it can be shown that (assuming
Ae=A2,)

'

'_

7+1

:'

PtJPto 1% 1 +

!l.

-1

IIIIII

--

II

II

I'I

L,..

il

24

CO_IOE_IAL
t
_

".

',
":
."_ACA
RM AS_FI6
s
4. o

zr

eJo

ee

Thus, for a given area AS at the entwine face, the inlet area A2
increases as the total-pressure ratio, and engine intak_ Math nu_o,_r, but
it decreases with increasing mass-flow ratio.
Other factors being constant, Am
is a minimlJ_mat a flight Mach number of 1.0. For present-day
turbojet engines in flight from sea level into the stratosphere at Mach
numbers from 0 to 2.0, MS
is in the range from 0.4 to 0.6; 5hus the area
ratio for an efficient air-induction system is between 0.7 and 0.9; and,
for grea_er engine-intake M__ch n,_ers
_ich
c_ be expected
in the
future, the ratio is more nearly 1. In other w6rds, the chsnge in area
between inlet end engine face is relatively smsdl and short ducts can be
fused
without requiring large divergence of the flow.
However, in the
I case of a rs_-jet engine with the Mach number at the burner about 0.2,
( the area of the inlet must be about half of that for a turbojet engine,
\

and the duct problem

Skin-friction

_ ;
_ .,
_

is more difficult.

losses..- In regard

to the duct path,

consider first a

straight duct with no initial boundary layer.

The boundary layer in the
usual case is nearly all turbulent and the fl_
is subsonic; so, as long
as the walls are relatively smooth and the length is short enough so that
pipe flow does not develop (less t_n
about 20 inlet diameters, see
ref. Bg), the skin friction can be estimated _ith sufficient accuracy
frc_ the formula

"

= o.ozIr
(see, e. g., refs.

(n)

_0 and 4:1) where

= T/qs

'r

shearing

dynamic

wetted

Reynolds number based on average flow properties

duct length
Z

i,

force
pressure

area
in dut

and on

'l_nedecree se in skin-friction coelficient with Mach number (ref. 42) and

with positive pressure gradient (ref. 43) need not be taken into account
in most cases because the effect of the former is small and neglect of
the latter' produces a conservative estimate.
.{

Beeton in reference 44 assumes one-dimensional compressible flow

and no chan_e in skin-friction coefficient with duct length in calculating

the total-pressure
ducts with conical

I
/,

ratlo_ resulting from skin-friction los_s

i_ circular
divergence.
Two of the curves from this reforence

_f
:_

NACA RM A95F3 6

cgI_FIDENTIAI

_ ,'

29

2
2
_

are reproduced
in figure 4;7 similar curves can be calculated by the
method of reference 49. Beeton shows that for the seve_.e condition of
Aa/A2,=I.2, M2:=0.8, and (_/dsxCf/O.O03)=10
the total pressure ratio is
0.96.
Since the loss in total pressure in this case is nearly proportiona! to the duct length, it is evident that here a shcr_er duct Is
desirable and that losses due to skin friction can be sizable.
(Refs. 29
__ud _ sh_.:
- that the !ncr_mental loss ._ +,,_+
.... _
+_,_+
A_'___
per unit decrease in total-pressure
_atlo is in the range of 1.2 to 1.5
for the flighL conditions under discuasion.)
For long-range,
subsonic
aircraft, intern_3, skin-friction
losse_ must be mi_:mized, and duct length
req_.ires careful consideration.
If thi_ duct were on a supersonic airplane with a very efficient method of external compression
(Ma'-->i.0),
the high inlet velocity and the result_m_ duct losses would counteract
the near_y isentropic itLlet flow, for the total-pressure
ratic would be
reduced to 0.99 by the greater internal skin friction.
However, in the
usual case of a supersonic design in which the duct is shorter and external compression occurs through shock waves, skin friction is a small
portion of the tot; L loss.
The maid concern in duct design Is a shape
that avoids separation and meintains uniform flow. s

:[
_:

Flow separation.- T"ne problem of avoiding scparation depends upon

initial flow conditions and duct shape.
For high-speed aircraft with

efficient air-induction
systems, the i_!et Mach
subsonic ra_e,
for if the flow is uniform

&.
_2

Me '
J

-_
i

y=l.4,

Pts/Pt2,_

is in the high

+ _ -_____l
Me _

=pte,_avMs
and with

Pts AS

number

+ ____________i
y
M32/

1.0 and As/Aa,=I.2 , Ma,=I.0

(22)
when Ms=0.6;

or

TSince the variation of total-pressure

ratio _ith the parameter
_,/dsxCf/0.O03 is linear to the extent zequired by the accuracy of ductdesign considerations
from values of 2 to i0, the range of interest, only
curves for values of 4 and 8 have been reproduced.
Total-pressure
ratios
for other conditions can be obt&ined with sufficient accuracy by interpolation or extr-_polatlon.
SGreatrex in reference 27 suggests that the ratio of the maximum-toaverage engine ints/te velocity
VM/V
be used as a criterion for flow
uniformity, and the example_ presented indicate that this ratio should be
less than about 1.e for satisfactory
engine operation.
For fully developed
pipe flow with a i/7-power velocity profile, VM_=I.23
Since the ducts
of the air-inductlc1_ systems for aircraft are seldom, if ever, long enough
for pipe flow to develop, it is evident that skin friction by itself is
not sufficient, in the usual case, to cause serious nonuniformity,

'

II

-,_

.
o_

/
f

e6

.......

t
:

..

_
,,.

(_c
_ ,_

....
' ' "

;,%t

_w

: .:
:_ACA

esgb

A)_I6

)u

M2'=0.7 when M3--0.9, Such a high subsonic Mach number fJtthe in/et makes
the design of the upstream section of a duct critical because, assuming
one-dimensional flow

or assuming isentropic flow from the free stream to a local station _n

the duct entry, pt=Pto , and

_p
'

i,

!
;
:

dx

+ ---

MO
_'7-m

dx

For a g_ven local tot_l pressure, or flight altituAe and Mach number in
the second case, the bracketed _erm of these equations has a maximum
value at a local Mach number of 0.79and changes little from M=0.6 to
1.0. As a result, deceleration of flow in this range causes the most
severe positive pressure gradients per unit of Mach number chau_,e,and
the effect is aggravated by low-altitude flight at high Mach num,er.
Since deceleration is produced by an e_)anding channel in subsonic flow,
the inltialportion of a duct must diverge slowly to avoid presm_e
gradients which separate the boundary layer.
With mauy induction systems, b_adary D_yer fr_fS.ow over surfaces
upstream of the iDlet enters the duc_. In this case, the duct shape
depends critically on the initial boumds:y-layer conditions because the
pressure gradient that a boundazy layer c_n withstand wi_1out _eparatlon
decreases as the boundary-layer shape parameter H increases.9 The
shape parameter is increased when the boundary layer flows through adverse
pressure gradients and over rough surfaces.
_SH=5*/8=clisplacement
thickness/momentum thickness. This ratio is a
measure of the shape of the boundary-layer profile and is u_eful for
indicating incipient separation. Reference 47 shows that separatio_ do_s
not occur in inc_m2ressible, two-ddam.usio_alflow if H<I.8, and reference 48 slmdl_ly shows that the criteri_ is valid for conical-diffuser
flow.
"...........
tw"

% V,:,,

_ "_,

C01_I>ENT_,AL"

NACA I_MAS_F16

_.Together with area ratio_ length, initial Mash number, and

initial boundary layer, the internal co,,toursof ducts require careful
consideration. The factors to be considered in axially sy_netric straight
ducts are shown in sketch (5). .Many ducts also include some offset of
the center line from entrsmce to exit, transitions in cross-section shape,
and j?mctures between ducts. Since turbulent boundary-layer theory is not

"-

I.E._try
lenql_
Inltiol _e
5.t_imumslo_
2

'

Sketch (_)

'

yet sufficiently refined to provide, even for simple cases, a method by

which an optimum diffuser can be determined (see refs. 43 and 49), the
qualitative indications of many e_eriments _st be utilized in design.

,
%
i,

_
ii

F_rml slope

In regard to entry length, a section of nearly constant duct area is

_eces_'5 to p.cc,;_de for reattachment of the flow for flight conditions
in which seperation occurs in the inlet. The data of Seddon (ref. 28)
for zero flight,speed indicate t_t for normal lip shapes, an entry length
of possibly one inlet radius is desirable. For engine installations in
supersonic aircraft, the data of references _0 and 9l show that entr_
ler_rthsof six inlet radii provide a relatively wide range of mass-flow
ratios in which engine flow is st _ly. Alsc, the studic_ of shock-,wa_e
stability of Kantrowitz (ref. 52) show that a constant-area section is
desirable to prevent downstream pressure pulsations from forcing '_terminal normal shock wave out of an inlet. (These considerations are further
discusP,
ed in refs. 53 and 54.) Because of boundary-layer growth through
the entry length, the duct walls must diverge slightly to provide a constant effective area. Study of duct data in which the boundary-layer
displacement thickness was measured, such as references 48 az#l_5, indicates that an axially symme_rlc entry s_.ctionsho_Id diverge at a halfs_gle of from 0._ to 1. (This range of incremental divergence angle
also appears to be satisfactory for bo1_zdary-layercompensc_ion in the
initial, mmxi_a, and e_it slope region, w" en the boundary layer is not
separated, i.e., H< 1.8.)

"

/j

'

8
:11

le

so

el

e tew

In regard to initial slope, equation (23) indicates that to minlmi._e

adverse press:,_egradients at high inlet Mach numbers the slope shoui_ b_
small and the change of curvature should be contJmuous. The need for
"_'
limitations is indicated by Naumann, reference 96 and is illustrated uy
the data of reference_ 48, 99, and 97 .I These data show that the abrup_
expansion where a i0 or 12 conical diffuser is attached to a straight
pipe causes nonuniformity, appreciable losses in pressure recovery, and
some reduction in the maximum mass flow when the approach Mach number
exceeds 0.7 to 0.8.
In regard to maximum sloes, it determines the shortest ducb which
can be used without auxiliary methods of suppressing separation, such as
those of references 98 and 99- As the local Mach number decreases from
the throat along the length of a diffuser, the walls can diverge at an
increasing rate without an increase in local pressure grad_.ent. Thus,
a maximum slope exists which depends upon the initial Mach number and the
initial b0undary-layer profile. The available experimental evidence,
such as references 48, 95, 96, 97, and 60 through 62 for conical diffusers,
indicates that the maximum included divergence angle io in the range from
6 to 19 with the largest angle being used on.l:" with thin initial boundary
layers.
In regard to final slope, the theoretical studies of rcf_rences 47
and 6B and the ex_perimentsof references 48, 55, and 97 sh..wthat for
minimum-langth diffusers having As/An'_2.0 this slope should be less
than the max_mm slope to avoid separation when the initial boundary layer
is thin. All of these studies were made with conical diffusers; the fact
that the final slope should have been less than the maximum slope is
indicated by the measurements of the final profile which, at high values
of M_., had _ >> 1.8. If _he initial boundary layer is thick, the
maximum slope cannot be large; in fact, the two slopes become equal. The
da_a indicate that a 3 final divergence angl_ on a wall, or a 6 included
angle, should be used with bo_J.hthick and thln initial boundary layers.

,
i

These qualitative considerations indicate t_at for thick initial

boundar_ layers and high inltisl Mach numbers, a diffusing straight duct
should have a faired entry section and p con _al diffuser of included
angle no greater than 8 (6 included angle plus a m_ximum of 2 for
boundary-layer compensation). For other conditions, _air duct shapes
which satisfy these considerations can be conveniently expressed as
_The data on conical diffusers from these references were analyzed
to determine desirable duct shapes by selecting longitudinal pressure
distributions for which H_ 1.8, and then calc_lating new duct shapes
from one-dimensional relationships for this pressure distribution and
value_ of M_, approaching i. The resulting'calculated shapes all hs_e
small initial slopes because, as shown by equation (23), the Mach number
gradient (i.e., the slope of the wall) must decrease to maintain a
pressure
constant initial
gradient with increasing local Mach ru_'_r.

.:__"

NACA EM AS_F16

_:

'-

CONFIDFI,_IAL

29

expo_lential functions of tnc duct axial coordinate.

Tests ,_ere made of
a fs_uily of such diffusers with _ ratio of throat area to exit area of
1 to 2 and a variation of the ratio of duct length to t_moat diameter of
fr_n _ to 9. Tests were made with both _=parated and attached initial
boundary layers at mass-flow ratios up to t_ maximum, and the results
are reported in reference 64.
Dat_
from these and other tests are co_1.00
_.-Predictionsby methodof ref. 44

pared
tion ofin ansketch
attached
(6)
initial
for the
boundary
condilayer.
It is apparent that, for this

_
,____:
_'Conical0ref. 56 I
/
_'_8" Con_coI-O
5
_'-_96 L o f "_._
..,;t_1
olq +

_,
;

comparison, the ratio of initial

boundary-layer
thickness to throat

_
o
8

recovery

than does diffuser

shape.

_:

The measurements

i','
:_
:,'

is
on the
flowimportant
uniformity
and of
steadiness,
that
effect
shape _
for
the uniformity
ratio
VM/Vductvaried
from 1.12 to 1.29 for ducts differing
:n total-press,are ratio by only 0.02
in tests with a thin initial boundary
layer ((S/'r)2' =0_0014) and a hig_
initial Mach number (142' _- 0.89).

,_

Furthermore,
t_
_qual uniformity

"

of reference

ducts having
and pressure

64 show

_.
,

i i

Midstream
Moth number,O.,_

-._ B8

I
I
I
]----,
I
......
0
.004 D08
.Or2
_6
D20
Bou.d_y-my_r,_c_essrotM,(8/r)_o
Sketch (6)

1.00 __
-_

nearly
recovery

differed
by flow
a large
amount in
quality of
steadiness
at the
hi@h
inlet Math numbers.
The comparison
of pressure recovery predicted by
the method of reference _ with the

,_

13.5-O, ref. 64_'--_

---_
l 12"Oo,'icoIaref 48_'_

recoveryfor tfiinnest1
unseparated,__Y
__
u._15.5__6_

96 _
4_
_ 92 _
._

experimental measurements
of sketch
(6) _ .88
shows that the prediction is only
acc_.ate when the initial boundarylayer thickness is very small.
If it
_
cx .84----'
I
is
not small,
the effective
skinfriction
coefficient
is larger
than
that indicated by equation (2_I) and
experiments are necessary for accurate
loss predictions.
(The data for
_ketch (6), and also (7), were calc_iated according to the mass-derived
method; see Appendix B.
The magnitude

!
:

M._%_eom Moch I

;_umbe,',08
fl
.80 |
[----_--R,eference
64|

D2
04
06
08
JO
Om_oceme_ th_kr_s_ ratk%(_*/r)z,

of the difTerence between experiment

Sketch (7)
and theory depends upon which method
of data reduction is used; the
zlThe ducts of reference 64 are designated by numbers which indicete
the maximum slope in terms of incl_ed
angle and the length of ez:t;-ysection in terms
of inlet radius.
Thus_ 8 conical -0._ indicates a conical
divergence of 8 and an exponentz_J_ly faired entry section of 0._ inlet
radius

"_I

_
_

:"
_'

in length.

.......

IIIII II

L_

_ 11 IU

IIII _ &

,,_

-_r 7 ....

_._

3o

CO_ZDENT_A,L:, : .'': ':NAOA _ A99F16

:

ix ;

6
te

difference
shown in sketch (6) would be smaller
redaced by the mss-fl_w
weighting method.)

if the data had been

Sketch (7) shows the results of tests reported in reference 64 for

three ducts w_th separated initial boundary layers.
The data show that
an extended entry section increases the skin-frlctlon
losses when the
initial boundary la_fer is unseparated;
therefore, if separaticm in the
entering flo_r can be avoided, a tong entry is umdesirable.
However, with
_,
i_

initial separation which, as will be discussed later, c_n occur in lowspeed flight _t high mass-flow ratios or in high-speed flight at low massflow ratios, some entry length improves duct performance because it gives
the boundary layer an opportunity
to reattach.
The fact tb_t _he _ressure
recovery can be higher for the long duct with the separated boun;_j
layer
than with the _mseparated
profile indicates that reattacP_nent occurred
after relatively extensive separation and that the small skin-f_.qction
force in the region of separation reduc _d the over-all losses.
In regard
to flow uniformity,
the results of reference 64 show that for short ducts
the flov is more uniform if the initial boundary layer is attached rather
than separated.
For a given initial profile of the separated type, the
final uniformity is improved if the duct is made longer.

Reference 64 reports tests which were inteLRed to investigate to r _ne

extent the mantu'acturing tolerances required in duct constractlon.
Measurements were made with a duct having d_xTerent degrees of surface roughness, waviness, and leakage.
It was found that roughness caused by
scratching the surfaces with coarse _andpaper or by putting discrete steps
in the duct walls, as could occur _Ith Joints that are not flush, had no
effect on the diffused flow.
The maximum magnitude of the roughness was
about 0.7 the momentum thickness of the boundary layer _t the duct throat.
The maximum wackiness tested was similar to that which would occur because
of pressure loads in hlgh-.speed flight; circumferential
stiffeners were

!
!

i
:
:

assumed to be 0.6re,
apart, and the deflection was varied up to 19 tiues
the momentum thickness, or 1.9 times the boundary-layer
thickness, at the
duct throat.
For m_ss-flow ratios
toe'/me'* below 0.89, even the maximum
waviness tested had a negligible effect on the final flow.
At greater
mass-flow ratic_ 3 the maximum waviness reduced the pressure recovery,
uniformity,
_I
steadiness only slightly.
Leakage, as might occur through
Joints in duct wslls during high mass-flow operation in run-up on take-off,
was fou_
to have negligible effects when the leaks were in the lowvelocity region of a duct
However, leakage nea_. the duct inlet caused
separation with ensuing sizable pressure losses and flow nonunlformity,

The internal-flow
systems of most aircraft h_ve some offset between
the inlet
Grid the exit, transitions
in cross-section
_hape, and J_mctures

!_
i

wiLh other dacts, all of which can cause losses in pressure recover[.-.
The general pzoblem in the design of these elements Is the same as that of

%1
,k,,.......
......

_-,..,-m,m_

_'

_111

'

NACA RM A99FI6

_
_

CO_IDF_I_FI_L

_S
_
_
-i_i
_
_
_

a _ubsonic diffuser, tk_t is. the preven'ion of local separation and

reduction of skin friction. I_
One design f_ture
that has always been
beneficial
is the use of generous fillets to a" _id angled corner_.
(See
refs. 67 and 68.)
He_ever, since the factors which cause pressure losses
differ w_th each duct configuration,
it is diff__cv_lt to apply accurately
general design inform%tion.
The dat_ of references 28, 60, 61, 69. and
70 indicate the trends to be expected.
The m_nitude
of the total-presure
losses in
s-bends is demonstrated by the tests of reference 71.
Eelatively short ducts (l/rs = 4.0) with several inlet cross-section
_hapes
and a circular exit we_ze tested at a _ach number of 1.9.
The inlet had

_
{_

a wedge-shaped
externsl-co_pression
the auct was offset 1.9 exi_ radii,

_
_,
_
_
_]
;_'
_
_:

maximum total-pressure ratios measured with the ducts were of the order
of 6 percent less than those measured wi_h a straight duct.
Reducing
the mass-flow ratio decreased this difference to about 3 percent, a fact
which indicatez the dependence of duct losses on inlet Mach number.
Altho_=h the total-pressure
losses could be reduced by reducing mass-flow
ratio, the exit velocity distributions
show considerable nonuniformity
for _hese conditions.
Tests with offsets of one and two inlet radii
reported in refe-._nce 64 indicate similar results.
The center lines of

_....
_

these offsets
were s_ooth
curves
similar
to thin
thoseinitial
of the boundary
duct-wall layer,
contours.
At a mass-.flow
ratio
of 0.9
with a
the 1-radius offset reduced the total-pressure
ra_io 3 percept from that

_
_-_"

_
_.

of a straight duct, and the _-radii offset r-_aced it 6 percent.

The
steadiness and uniformity qualities of the flow deck-eased in a corresponding manner.
For example, with the thin initial bou_dary layer, the maximum mass-flow ratio fcr steady flow was about 0.9 for the straight duct
and 0.7 for the duct w_th the O-radii offset.
A fourfold increase in
the initial boundary-layer
th_._kness reduced the l'_tter mass-flow ratio
to o.A. It is apparent that deviating from the optimum aerodynamic design
of a duct can have serious consequences.

Subsonic

_,
_
f

surface and the exit certer line of

rs, from the dnlet center line.
The

Since

in subsonic

flow,

Flight

press'_re losses

snd nommiformity

result

from skin fr_ction, separation, and en+ering flow that is asymmetric with
respect to the inlet, the _nduction-system
design problems in subsonic

,_'
_
_

_The
design prin?_p_es for annular subsonic diffusers are like
those of diffusers without center bodies, but the annular type, having
more wetted area, has l_rger frictional pressure losses.
Ltudies of
annular diffusers are reported in references 65 and 66.

52

"_C"_II_lPE_f'r-AL
, ,

:
,._

t I

*, -

: ": ! :"iNACA P_4 A. _._

t,

el

el@

ol

flight e_e to provide conditions that avoid or minimize thesa factors.

Skin friction and inte_al
separation _l'e problems of duct design; the
problems of separation in the inlet and symmetry ea'e discussed in this
sect, on.

"

;
,_

To illustrate the conditions which lead to the principP_ separation

problem of __nlet d_sign in subsonic flight, sketch _v
,c:
shows a typical
cui-'e of the air requirements
of a turbojet engine in terms of the freestre-o_msrea of the engine-air streamtube
A o as a function of fligh*

12
,

See ievel
Sfrofosphere
Flighf schedule

I0

___

1.0

-----

.-

.4

.8"1.2
Machnumber,Mo

1.6

20
k

Sketch (8)

Math number.
It is here assumed that the airplane accelerates at sea
level to a Math number of O. 8, climbs at this Mach number to sltitude,

_"

-"-7....

f,.':,, "" " " : ......1--_

.....
"'_

71

.....
A _

A55F16-

a_nd then accelerates

_

CO.'_IDY:I_I__AL

from this cnlise

condition

33

to a Mach number

of 2o

The air requiremen_

is not only a "unction of Mo, but also of tot_!pres._vu_eratio and altitude, as shown, and of engine desi_--nand power
setting.
Since czuising flight is usually _ important desi_, condition,
the irlet area
A2 must be selected to produce efficient cruise performence, and this, for high-speed aircraft, is g_nerally at a relatively high
mass-flow ratio, above sbuut 0.8.
The chuiue of this mass-fLo_" ratio is
a comnromi_e he_ween reaD_irements for other flight conditions p_ud the
cor.riicging interests of the intern_l and exterr,al flows.
A low mass-flow
ra_io (me/mo=Ao/Ae
<<I), that is, a diverging streamtube ahead of the
inlet, is desirable to the intelnal flow because then most o.T the kinetic
compression upstream of thp engine, being in the external stream, is
isentropic if there is no inLerference w-i_h a boundsA-y layer; and, since
the ir_le1_velocity is low, internal s_in-friction
losses are minimized.
On the other hand, a mass-flow ratio greater ths/l 0.6, at least, is desirable to %he eyternal flow for two reasons:
(i) External compression
can
" thicken or separate the boundary layer on an upstream surface which is
in the interference field of the engine flow; (_) a d_verging streamtube
subjects the inlet lips to large flow angles which can result in an
increase in external drag because of wave drag due to local supersonic
flow or because of slain friction due to i_ncdlate boundary-layer
transition.
in any event, the sketch shows that choice of s_i inlet area for
the cz_dise conlition produces a.n i_let .much smaller than the area A O
of low flight speeds.
Conseo2aently, at low speeds the mass-flow ratio is
high and the flow converges toward the inlet (Ao/Ao >> 1.0) at large angles
wi_ich can cause internal separation, low total-pressure
ratios, and flow
nonuniformity
unless special precautions
_e taken.
If the eritic_l
design condition is flight at 8 Mach number of 2 rather than subsonic
cxn/ise, the situation at low flight speeds is worse unless the .nlet area
can be varied with speed.
The area that ":,akesin the required air is even
sma21er at this high speed, and also little fe_iring of the ].ip profile is
possible because it must be thin to minimize the wave drag of supersonic

__

-'

._
_

_.

fZight.
From this, it is evident that the principal problem of inlet design
in subsonic flow is to select a lip shape and a variation of mass-flow
ratio that avoids internal-flow
sepPIaticn at low speeds and detrimental
disturbances
in the external flow at high speeds.
Of course, there is
the l_mitablcn that the inlet area must rLot be chosen to be so small that
iu chokes el, a low flight
_=
_'esure losse_ and
which a Macb number of 1.0
in uniform flow are shown

speed, for then the flow to the er4_ine suffers

is nonuniform u.'%_uun_t_ady. _i_e condiZions in
cc_u be reached in an inlet with a sharp Izp
in figure 2.

-_

Lip design_.- The import,triceof lip shape "to pressure recovery in subsonic flight can be seen from the analysis or' Fradenb_rgh
and Wyatt
(_.'ef.14).
The extreme case of a tube having very thin walls was studied
by momentum methods, and the predicted variation of total-pressure
ratio

pt2/Pt

with mass-flow

ratio
..........
for _ariou_

flight

Mach numbers

is reproduced

Y
-

.......

I II II

-' -- ' -it

eat

e_

v,_
@

:'"

o4_a

40

eo_

OOe

@mD

N/1-//

i,

'

!//

720

1___

.2

.4

Sketch

_?

__1

LO

(9)
i

in sketch
,
'

)
i

'
,
J

-_

'

(9)

(Losses

in the duct behind

the inlet

can be added to these

total-pressure
ratios to determine the pres=ure at an engine face
pt 3.
At high mass-flow _'_tios when the lip is stalled the duct losses are
small relative to those due to flow separation at the lip and are seld_
known. ) If the inlet area is r o_ ",cted for the altitude, cx-aise condition
and information
similar to that o_ sketch (9) shows that the mass-flow

ratio
me/m2*
is about 0._ ._n _ake-off_ the total-pressure
ratio
Pt2/Pt
at the inlet is then less than 0.9.
Such pressure losses correspond to a

,-

i_- to 20-perc6nt loss in englue thrust which, of course, represents

serious limitation on _he acceleration
characteristics
of an airplane.
The flow nonuniformlty
which accompanies the total-pressure
losses can
even further limit engine operation.
If a smaller inlet area were chosen
to suit more closely the requirements
of supersonic or low-altitude
highspeed flight, the losses would be even greater.
On the other.
" hand, the
effects of increasing fli_t
speed are rapidly alleviating,
These large pressure losses at low speeds that result from a sharp
lip can be avoided by several methods.
A curved internal lip _)rofile
which the flow cam follo,_ prevents separation and the attendant nonuniformlty at high mass-flow r_tios, or, for a given lip profile, the losses
can be reduced by decreasing the mass-flow ratio either by increas'ng the
inlet area or by taking air in through another inlet.
Tests of llp
profiles on circular nose inlets at low speeds are reported in references

72 to 79.

Some of _,he rest_Its, in terms of

pts/Pto,

are presented

'_

_._

NAOA RM A_FI6

CO_'D__._.,.A.'_

_5

in figure _ and are compared with the prediction of Ptm/Pto

for the
thin lip of sketch (9). Duct losses have not been subtrp.cted from the
theoretical prediction because a wide variety of duct designs are ccmpsmed,
and, in most cases, duct losses by themselves wero not measured.
For the
cases in _hich smooth, nearly straight ducts were tested, the agreement
between
Pts/Pt and Pt2/Pt
is good at zero forward
the los_es for the conical-shock
_uleL from reference

speed.
However,
!_ are considerably

in this particular test.

The scatter of data at the maximum mass-flow
ratio is considerable, and a large part c_ _t is undoubtedly due to
inaccuracic_ in total-pressure
measurement.
Blackaby _ud Watson _ref. 72)
point out that near choking the flow through ducts is very unstesdy 2 and_
as mentioned in Appendix B, measurements
of pressure recovery by normal
methods under these conditions are not _eliable.
The data on the F-8_F
and F-lO0 airplanes are from full-scale tests.
The fact that they correlate with the data from model tests "ndicate that the effects of scale
are small.
Also, since the predictions
of the momentum analysis which
have no relation to scale agree so well with experiment, negligible scale
effects in regard to lid losses are to be expected.
The tests of reference 73 indicate that for a reasonable variation
of shape external lip profile has practically no effect on internal flowo
At zero flight speed, the data of reference 72 show that pressure recovery
is not highly sensitive to internal profile, for there was little difference
between ellipticel and circular shapes.
However, as shown in figure ,
internal lip profile is impnrtant at higher flight speeds, for the elliptical shapes are better than the circular ones.
At the flight Mach number
of this figure, 0.33, a sharp lip causes relatively large losses at high
mass-flow ratios, as at zero forward speed; but, in this case, the prediction

of

pt2/Pto

is greater

than the measurement

of

pts/Pto

by 1 to

2 percent, whereas at zero forward speed there was no difference between

theory and experiment for high mass-flow ratios.
The desirability
of the
elliptical profile is furthe_ _ substantiated by the recommendations
of
Pendley, Milillo, and Fleming (re.
_. 76).
An elliptical internal shape
was selected for this investigation from previous experience, and it was
found that the profile resulted in high total-pressure
ratios for a nose
inlet at zero angle of attack in the Mach nt_zber rs_uge from 0.6 to 1.1.
At these flight speeds, the mass-flow ratio of an induction-system-engine
combination rapidly decreases to values less than 1 (see sketch (8)), snd
the problem of internal separation from the lip disappears.
In fact, even
for a perfectly sharp lip, sketch (9) shows that internal pressure losses
resulting from li_ separation at the mass-flow ratios of interest (up to
0.9) are small at flight Mach numbers above about 0._.
Thus, at high
subsonic speeds, skin friction is the major source of pressure loss in
well-designed
systems.

'

ratio

..

Some tests have been made of qrhemes for reduci,_ the mass-flow
in low-speed flight to avoid lip separaticn.
These methods consist

_1_. ;,_,

36

COffF]_NTIA_',.
e

i :
,on

6_

-: : _ACA
0.0

f_MA55FI6

ot

oi" increasing the area t_o_h

_'hich air can flow into the induction
z_v'.
bern. In reference 77 a sharp-lip nose inlet was tested with a secondary
teoop having sharp llps that opened into the underside of the duct a short
distance behind the inlet.
At zero flight speed, it was found that the
variation

of

Pts/Pt

with

mt/mt*

(where

mt

is the mass-flow

through

the total area) was nearly identical no matter how much area (up to 68 percent of +_o+ _? +_ _o_n .....
_-_ _" j_ ".'as_,,_.,.._,_--^---"
_- _ in the _uxiiiary _ccop.
--Tnus,
the _rovement
in pressure recovery that can be expected with this method
is entirely the result of reducing the mass-flo_" ratio for a given engine
operating condition.
In reference 78 a supersonic conical-shock
inlet
with a sharp lip was tested with a translating cowl; that is, a short
length of cowl including the sharp leadir_ edge could be moved forward
exposing a gap with a rounded lip and increasing the minimum throat area.
Since the curve of total pressure ratio as a function of mass-flow ratio
rotrot* (mt is here based on the increased throat area) for the extended
cowl lles above that with the cowl retracted, it is evident that this
method not only increases the available inlet area, but it also improves
the quality of the flow.

Angle of attack.- The flow approaching

_a inlet can be asy_netrie
with respect to the induction system axis because of the changing attitude
of aircraft for various flight conditions, because of the induced flow
field of the aircraft, or because the inlet is distorted by configuration
requi.'ements.
The ultimate resttlt of such asymmetry is internal _eparatlon.
Data from tests of circular nose inlets at angle of attack and a flight
Math number of 0.24 (ref. 79) show that an inlet with blunt lips maintains
high total-pressure
ratios and uniform flow to greater angles of attack
than one with sharp lips.
Fox" example, at an angle of _ttack of 19 and
a mass-flo'_ ratio of 2.0, the inlet with an elliptical blunt lip attained
a tot-__l-prLz_ure ratio of 0.97 whe_-eas one with a sharp lit attained or_ly
0.90.
The corresponding deterioration
_n flow unifoz_!ty was a difference
between maximum and _tuimum total-pressure
ratios in the duet of 0.08
for the elliptical lip and 0.16 for the sharp lip.

At Mach numbers from 0.4 to 1.1, the .esults of references 23, 76,
and 80 show that even _th
shal-p lips pressure recovery is nearly insen,

sitive to attitude to angle of attack of about 8 to mass-flow ratios as

high as 0.9.
At higher mass-flow ratios this range of insensitivity
decreases.
The sharp-l._p inlet of reference _3 suffered greater losses
at high angles and mass-flow ratios than did t_e blunter lips of the tests;
at a Mach number of 0.9, an angle of attack of 12 , and a mass-flow ratio
of 0.9 the total-pressure
ratio was 0.92 whereas a blunter, but still
relatively thin lip, had a total-pressure
ratio of 0.94.
For these flight
conditions, the mass-_lo_-ratio
(m_/mo) at _ich
choking occurred _,_th

:
'
_

the sharp lip was 0.9 and that of the blunt

lip was 0.9_.

The sensitivity of an air-induction

system to _g]
only a function of llp _roflle, but it _s also affecued

'

.
.......

':_
i_o - :....... _._---_"_,,,_:e_"'_

"
' ,,,

_"| |_i'

._

_f sttauk _: not
, +L_ _ivergence

'

'

. i

". : ..'

%w

NACA RM A_gF16

'

%
_
_

_if
/

C0k?UDE_FIAL

,".

37

of the flow behind the inlet

In the _ests of reference 76 it was found
that an NACA 1-40-200 cowl was more sen._itiveto angle of attack and massflow ratio than a longer cowl, NACA 1J'O-;_O0,because the duct in the
shorter cowl expanded more rapidly, i_us, some lip bluntness and slow
divergence of the flow behind the inlet p_ovldes high pressure recovery
over a mifficient angle-of-attack range for most purposes. For a stir
greater range of insensitivity, the lower lip can be drooped and staggered
as suggested in reference 76 and tested in reference 81. In the latter
investigation, a blunt, staggered-lip inlet was tested eta Mach number
of 0.14, and it _intained high pressure recovery throughout the range
of the tests from inlet velocity ratios of 0.6 to 2.2 and angles of attack
from -5 to 12.

5_
_
_
_.
_/
_#,

_
_,
_
_
_
_
_

Inlet asy_netry.- An inlet that is distorted relative to the axis of

an air-induction system can .hacelarger pressure losses and greater flow
non,Jniformitythan an axially sy_netric inlet. For instance, Seddon and
Trebble in reference 82 report tests of a wing-root inlet at zero forward
speed. In comparing an _let swept back _2 with an unswept inlet, it
was found that the losses and flow nonuniformity were about twice those
of the unswept inlet. A"_eadditional losses were due to separation in

t
_

the outbo_axlcorner of the inlet which resulted from the fact that, for
this operating condition, the flow must turn through a large angle to
enter the duct, since it approaches nearly normal to the inlet plane.
Guide vanes ali__._with the duct axis in the outboard portion reduced the
flow nonuniformityj but increased the pressure losses. Slots in the inlet
lips similar to wing-leading-edge slots, but not swept, reduce_ both the
losses and nonun_formity because they increased the inlet area and bled
high-energy air _zntothe region of potential separation.

s
S

....
_-

_v

e
_:
_

An ._mportanteffect of inlet frontal shape is ahown by comparison of

the flow-6,istributionmeasurements of references 83, 84, and 8_ from tests
of _ing-root inlets at Mach numbers from 0.6 to 1.4. The results show

__-

that the uniformity

of the boundary
flow in the
portion
the inlet
which -was
unaffected
by the fuselage
layer
- theof
outboard
portion
"_as

_,_
_

greatly improved as the shape was changed from the acute 8_le
angular inlet to a semielliptical or semicircular inlet.

_
_.

of a tri-

Supersonic Flight

_'
-_
_
_
::
!_

-_
_

The considerations of pressure recovery in sopersonic flight are

more complex tk%u those at subsonic speeds because in supersonic com-

!'_

_%

pression of engine air the pressure losses and flow nommiformlty can be
caused by two additional factors, shock waves and shock-wave-boundary-layer
interaction. These factors become increasingly important as the local
Mach number at which they occur increase_ above 1. Moreover, the necessary
increase in thrust of air-consuming jet engines with speed depends upon
t_e increase in tc _al pressure

_
_
_
_

_"
_
_%

/
/

38

@0_LDEN_AL :

_ ".

_. c

..A_AEM
:

-I_v'_

m_

Pto = Po (I + 0"2No 2)s'5

ann density

Pto = po(!
Little of the available pressure and
to overcome the large dr_
forces of
margin of excess thrust at supersonic
thrust-available
and thrusl-required
small losses in total pressure cause
maximum-speed
performance.

mass flow can be lost if an engine is

supersonic flight.
In many cases, the
speeds is reia_ively small, and the
curves are slowly convergent.
Then,
large reductions :n acceler_.tion and
i

I
Supersonic c_pressionZS.
- Since the local Mach number at the intake
of present-day engine_ must be subsonic, the flow to the ermine of a supersonic aircraft must be decelerated
through a Mach number of i.
Ideally,

!
I_
I

this compressicn of the air can be accomplished

isentropically through c
reversed Laval nozzle with no external wave drag as indicated in sketch

l
_

(i0) ; practically,

i
I

shock-free

internal

intet'nol
compresslo,,
througa
M_@ waves

Normal
_

compmuio,

flow canno_

be attained
/

because

Extecr_coml_ssio,t_mug_
shock,ayes

,ntmnr:
co_sskm ttwou_shockwaves

external
endinternal
compression
Sketch (Io)
_

ISFerri
in reference 86 and Lukasie_Icz
in references 53 .and 87 discuss many of the principles involved in supersonic compression.
In this
report, these principles are mentioned only briefly, and the emphasis is
].imitations
the flight that
conditions
under
consideration-.
on presentlngforinformation
is useful
in design
and in point.tug out

'/

e,_

.
-

NACA A

F!6

39

the flow through such a channel is in a state of neutral equillbri_n. Any

disturbance which causes a loss in total pressure between the entrance
and the thrcat causes _ deczease in mass flov through the throat because
here the area and ve _city are fixed. Air _mst then accumulate because
more flows into the p_sage than can flow out, and a normal sh_.k w_ve
is fozunedwhich must move upstream, continually growing stronger3 until
it is expelled from the channel and spills the excess aic. The shock
wave caz,
not _.'e-enter
tilechannel unless the throat is opened sufficiently
to pass Ch_ full mass flow at the stagnation resstu_eexisting behind the
normal shock wave in the free stream. (For detailed discussions of these
phenomena see refs. 86 through C9.)

_
:r

'L
_
v-.

.:

It is, of course, not necessary to attempt supersonic compression

either in a closed charmel or isentropically. The flow can be decelerated
externally s_idthrough discrete shock waves as shown for several possible
arraz_Eementsin sketch (lO). The crudest method which entails the greatest
losses is to _ccept a normal shock wave at the free-stream Mach number.
-i
Sine_
= these normal shock losses can be redaced by decreasing the Mach
number at which they occur, hlgher tote1-pressure ratios can be attained
by placing an inlet in a region of substream velocity on an aircraft, a_
will be discus.Bedsubsequently under !NTERFE._CE, or by creating obl_que
shock waves to reduce the local Mach number but with less loss than that
of a single normal shock wave. For a given local Mach number ahead of an
air-inductlon system, the question arises as how best to uti_
_ize oblique
shock waves. Oswatitsch (ref. 90) has shown that the maximum tot_-- ..... :"
T
pressur_ ratio of a two-dimen_ional
I0
multishock system occurs when the
total-pr6ssure ratio across each
oblique shock wave is the same " For
3

__
.
:-"
"
_'
/
_
_'
#
_o
_
_

. total-pressure ratio with Mach number

for shock-wave compression (n oblique

_-_I_
_
.o_"

such
conditions,
the(11).
variation
waves
plus
normal
shock
wave)
is shown
intermina&
sketch
It
isof
apparent that the losses _rough a _ingle
normal shock wave rapidly become
intolerable above a Mach number of
about 1.6 and that large improvements
can be made by utilizing oblique

<
%
._
,
i

,
S

;_
_ .6
_ 4 ....
_._

_ k

k_
_

_2

_
Normal
_'
_.

sho_w_

\ \

,,_

hck waves'_"

_O

The variation of total-pressure

ratio with deflection angle for various
approach Mach numbers in two-dimensional

O0

I_

Z6
_.

_4

42
_'
'._. ,

flow is shown in figure 6 for a two-shock

Sketch (ii)
system (one oblique a_.da terminal normal shock wave) and in figure 7 for
a th_-ee-shocksystem. Figure 8 presents these variations for a two-_hock
i4Detailed information and design charts on shock wa_es can be
obtained from such references as 91 and 92.

_
.....
_.
_','J)!
__;_
....._._""_

mR"_ .'_.

---

II

llllllI

--

_"_

._

_",W_

l l__

I._

.......

._

i,_
;_'

Jl.JL .

.....

.- I

_,

J_lIIII I

_"'

';'

'

'-i

--

_-_-.-_-1

....

-"

_-_--

4o

:2:i
:

NACA RM A55F!6

system in conical flow _d

is ts_ken from reference 53 where it is assumed
that the normal shock wave occurs at the average of the Mach number behind
the conical shock wave and on the cone surface, (Ms+Mc)/2.
This assumption
is adequate for the Math nt_nher and cone-angle range of interest in the
flight conditions being considered in this report because the difference
between
Ms and M c is sls_ll, less than 0.01.
It is apparent from this
fact that the maxi_dm total-pressure
ratio attainable in two-dimensional
and conical flows is about the same.
Lukasie_cz
in _eference 53 shows
that 5his difference in total-pressure
ratio at Mach numbers less then 2.0
is less than 0.015.
The curves of figures 6, 7, 8_Id 8 show that totalpres**_ureratios near the maximum can be maintained for a relatively wide
rang_ ef flow deflection angles, an important fact because an eagle can be
selected which produces nearly maximum recovery at the high-speed condition
with little decrease from the maximum possible for a considerable range of
lower Mach numbers.
Also, the angle can be chosen so that a detached shock
wave occurs only at a low supersonic speed where the entropy rise through a
normal shock wave is small.
For example, at an upstream Math number of
1.8, the maximum total-pressure
ratio with a two-shock system is 0.945,
and the corresponding
flow deflection angle is 14 , for which the detachme/It Mach number is 1.57. If a lO O deflection angle were selected, only
O.O1 would be lost in total-pressure
ratio at the design Mach number, but
the shock-detachment
Mach number would be reduced from 1.57 to 1.37 and,
in this Mach number range, recovery would be improved several percent.
The total-pressure
ratios decrease beyond the maximums (the values plotted
in sketch (ll) for the two-dimensional
cases) because the Zosses through
the oblique waves exceed those through the normal wave until finally the
oblique wave detaches from the deflecting surface and only the pressure
recovery through a single normal shock wave is possible.
The high level
of total-pressure
recovery that can be attaine_ by conical-shock
compression has been verified at Mach numbers to 2.1 in references 13, 93, and 94.
In reference 94 a center body contoured for isentropic compression at a
Mach number of 1.85 produced a total-pressure
ratio of 0.967; with three
oblique shock waves, the total-pressure
ratio was 0.954; and with two,
it was 0.945.
In all cases, a uniform flow was measured after diffusior.
These values are very close to those obtained by adding the predicted shock
losses to the experimental
duct losses described prevlous]y.

Limiting internal contraction.- For internal-compression

systems
through shock waves, the problem of flow stability exists as in the
reversed Laval nozzle because of the two possible stable positions of the
normal shock wave, ahead of the inlet or downstream of the t.%roat. However,
at the expense of complication,
this disadvantage
can be overcome, and this
form of supersonic compression hes the advantage over external compression
of deflecting the flow toward the system axis rather than away from it. The
frontal area, external drag, and amount of turnLug in the duct can thereby
be reduced.
Thus, the optimum sa.rangement for any specific case requires
detailed evaluation.
The relation between contraction ratio, total-pressure
ratio, and Mach number is

'

._

' C-t

NACA RM A_F16

"_

A2'

Me

_._,
_;
.:_
-_i

CO__%L'

Ma "2 _

Ma 2

41

p _

Pt2 '

.9

for isentropic flow to a Mach number of

This relation is plotted in sketch (l,2a)
.e
1 at the throat. Also shown is the con- _I_ .7
traction ratio which permits isentropic
&
flow to a throat Mach number of I fr_n
"g
the total pressure existing behind a
_ '_
normal shock wave. This is the con"
traction ratio at which supersonic flow
_

"_.
'_
t-

..

\_
_
..... _
.
\ _,,,f.ee

\
_-_,N

can be established in a fixed internal-

_j:

.4

_"

contraction
inlet
at a given
flightTotalMach
number
and is
designated
_start"
pressure-ratio curves for two positions

j_
_.

of
the
normal
wave for
are
also
shown
for shock
the cases
where_start
the normal

_
,_
._.
:_:
._.

shock wave is at the throat and in the free

stream. It is, of course, possible for the
normal shock wave to be downstream of the
throat, in which case the pressure recovery
decreases _oward the lower curve in
sketch (12b). It is apparent that the

_._

starting contraction ratio for a Mach

,.o_,

(_,_/

that permissible at a low(r Math number.

_
_

Thus, if
aircraft
is to reach
a Mash
number
of an
2.0
and maintain
the totalnumber of 2.0, for inst_mce, is less than

_
%.
_i.

pressure
ratios ratio
(Pte,/PtO)_star
t orwith
higher_
the contraction
must decrease
. .e_-increasing flight speed above a Mach num- =_"_ ....

_._
_
.L_L
.:
_':"
i_,
_

ber of i. Also, it is apparent that above

a Mach number of about 1.8, the totalpressure losses with _start are unacceptably large, and it Js desirable to
decrease contraction ratio and incr_:ase
supersonic _ompression toward the isentropic value. If the throat area is
adjustable, this can be done as long as

_i
_i
_

_'
(',

Mach number

=m

J_

I_1

.Me_ with

\
_
..

.z
_

_4

LS

z.z

.,

z6

_o

Skc.tch(12a)

;__'_

._f.
es(___
\/_/_,
%_

_ k
\\

.e---_
_ .5_=--_

'

_
X

____/
_-

.4_

the flow at the throat is supersonic.

For a given contraction ratio the Mach
number at the the'oatcan be calculated
from equation (2_), sad the maximum
total-_ressure ratio possibl.e is _hat
of a normal shock wave occurring a_

.s

"21.0

1,4

2.2

I_

Ze

"_

-_
:_

M.

Pt2 _/Pt2=I"
_ ....,

Sketch (12%)

II

.....................

'

II

I1111111I_111 I]1

...........

....

""'

'/

42

coNfiDENCe.'.

.....
[ t
_

ce

.":

eta

eo

eo

.: :e-:NACA _94 A95FI6

:.-

However, if the flow at the throat is subsonic due either to a contraction

ratio that is too small or to the inlet being too larg_' for the engineair requirement,
a normal shock wave _Jaead of the inlet reduces the totalpressure ratio to t_at of the lowest curve showD in sketch (12b).
In
fact, this type of air-induction
s_stem is sensitive to flow changes, and
close control of bOth inlet-area and contraction ratio are necessary if it
is to operate with an engine through a wide range of flight conditions.
The pressure recovery car. decrcasc abruptly from the _ximum
possible with
small changes in either mass flow or angle of attack (see ref. 93).
An induction system in which both inlet and throat areas were adjustable to match engire-air requirements and provide maxim_n total-pressure
ratio with -"_*_ 91 contraction through two oblique shock wave,_ and a
terminal n
,ave has been reported by Scherrer and Gowen in reference 68. I
,_b found, as shown by the data poJmts in sketch (12), that
in this particular test a contraction ratio well below
_start
could be
reached, but there were no significant improvements
in corresponding
totalpressure _-atios. It was concluded that the increasing supersonic compression was counteracted by increasing losses in the duct and that greater
refgmement in duct design was required.

'
!
,
,

Other methods than adjustable passage walls have been invest_

,ed for
avoiding the flow-stability
proble_ of internal-contraction
inlets.
Evvard
and Blak_y (ref. 9_) tested an open-nose inlet in which the contracting
passage was perforated _o penmit the escape of excess flow between the
inlet ard the throat as the normal shock wave moved into the channel with
increasing flight Mach nu_iber or mass-flow ratio.
A high maxx_ram totalpressure ratio, 0.93, was measured at a Mach numher of 1.89, and the inlet
was found to be relatively sensitive to ma_-s flow bud not to angle of
attack.
It was estimated that 9 percent of the total ,_ass flow was lost
through tee perforations.
Further tests on this meth,xl of flow stabilization are presented in references 96, 97, and 98.
Al_,hollghhigh pres.o,
ure
recovery is attained with this type of inlet, it is accompanied by high
drag if the flow through the perforations
is vente_ to the external stream.
For example, the data of references 97 and 98 _how that the d_ag of perforated inlets is as much as 29 perce_t great__r than that of unperforated
types.
A similar method of providing flow stability when th_ terminal
normal shock wave is at the throat has been repo::ted by Nelce, reference 99.
Here, the cha_ael walls are vented, immediately ahead of the throat to a
chamber to permit the escape of excess mass flow when a d:sturbance tends
to force

the normal

shock wave upstream

into +:_e converging

---

<
,
.

passage.

Rectangular
scoop _'_lets w_._ side walls swept .back towara the body
as described in references _3, 100, and lO1 are able to n_in_ain supersonic
_!
_

flow to the throat of a contracting passage at reduced ma_s-flow ratios

and flight Mach numbers because air can escape laterall,,"_s the normal
shock wave moves down the channel.
_lowever, at low flight Mach numbers
the first oblique sho.-k wave from the compression s,_-face is forward on
the fuselage, and it interacts with the boundary layer causing both high
dr_g and poor pressure recovery.
These difficulties
ha_e been partially

"

_J

<

_CA _ A59FI,5

}_
_1
_.

circumvented by use of a leadi_,-edge flap on the compreasion surface.

(See ref. i01.)
Deflectlon of this flap toward the body reduced the
pressure rise across "_he oblique shock wave at a given Mach number, and
delayed boundary-layer
separation to lower Mach ntunbers.

i_
'_.
_

For the conical-shock

inlet, internsl contraction can be used to
produce additional supersonic compression, but at the expense of encountering the flow-stability
problem and additional duct losses
L_kaslewicz

C0_DE%_ZAL

43

derives in reference 53 th
contraction ratio
_start
that can be used
with conical-shock
inlets, based upon the assumption that the entrance

._i:
_'_

M_ch number is the average of that behind the shock wave and on the cone
mu_face.
This variation is presenc_,d in sketch (13). It is seen that for

:._
._

large cone angles the permissible

contraction is small.
Experiments
at
M o =1.85 (ref. 93) show that for an inlet _th
a straight lip (not cambered to meet the local flow), internal contraction reduces the optimum
cone ang)e for maximum

pressure

recovery

!
:

IDO

_]
_"

inlet
with25only
to about
as conical-shock
compared to 30compresfor _n

_._
,{
:_

in
maximum
recovery
small. e
sion,
(fig. _ossib]e
8).
However2
the is
differene
Only for SL_Lll cone angles where the

96
.92

_,
_:

oblique shock wave is not being fully

utilized can internal contraction

88

;_
_

produce
great
have beenany
made
at advantage.
a Mach numberTests
of 1.85

_
_i

contraction and a perforated lip to provide :"low stability.

(See ref. 94. )

'!
;
,,_

The results
with
conical-shock
indicateinlets
very high
havingmaximum
internal
total-pressure
ratio, 0.95, for this
arraugement.
Both drag and pressure-

_i____
.76

_
_'
_',

recovery
measurements
werea made
for a
conical-shock
inlet with
20 cone
and a perforated cowl at Mach numbers

.7210
.

_,
_
_

of 1.59, 1.79, and 1.99 in reference 96.

Sketch (13)
The resulgs indicated that even thougJa
high pressure recovery was obtained at zero _ngle o_" attack a r_:latively

_
!_
_.
_:
_'
._

large increase in external drag occurred rela'i;ive to simil_r unperfor_ted

inlets.
The pressure recovery was relatively insensitive to mass-flo_
change above the mass-flow ratio at which shock oscillation occurred,
With increasing angle of attack both the range of mass flows for steaay
operation and the pressure recovery decreased at all M_ch numbers, the
latter being a more pronounced decrease than with simil_r unperforated

/.

_,'

_nlets.

:'i!

,_

Limiting inlet Mach number.- Fc,r extern_l-cmmpresslon

systems there
is no problem of flow stability as, there is with internal-com_ressi_
systems.
There is, however, a li_Itation on how nesmly isentropic the

compression

can be,

or, in other

_ds,

__k
\ _
30"
--

_
*'.84

\_

r
20"

.80

1.4

on the number

18

2._
Mo

of oblique

____ 5.O
2.6

o-_

-_;_
,:_
_
"i_

shock

g-

44

'
' ;

_iD_.WA._'.
.:

" .'. _ :
.....

NACA RM A_-F3.6

waves which it is practical to use.

This limitation 8a.ises because the
larger the number of shock waves, the higher the subsonic inlet Mach
number and the greater the duct losses.
Hence, _otlmum supersonic ccmpressic, requires excellence in duct design.
The _o. lowing +_ble sho_'s the
local Mach number and total-pressure
ratio after the terminal normal
shock wave in a pattern arranged wlth
n
oblique shock waves to produce
the maximum supersonic compression at approach Math numbers of 1.5 and 2.0.
Subta-acted fk-om these total-pressure
ratios are the duct losses corresponmi_
to the inlet Mach number ss measured with a duct w_th very small losses
in reference 64. Thus, for these conditions, which are probably about the

"

Duct13._ --o, (o/r_2= o.oo143

Mo = 1.5
Mo = 2.0

! _t_p;_-_t_
_t._ _t_
pt_-_t_
p_
n"_-_o Pto_o
_
_ _t-'_
0 0.70 0.93

0.02

0.9l 0.98 0.72

1 _ 98i o_ _ 74 _
3

-94 1.00

.0_

.96

.90

-97

O.O1

o2
.03

0.71

.94

best that can be expected in the present state of practical design knowledge, little ca,. be g_tned by using more t'han one obllqae shock wave _.t
a Mach number of 1.5 or two oblique wave_ at a Mach nz_ber of 2.6.
If
a poorer duct is used, say the duct "_Ith a thick initial boundary layer
and a two-radli offset as described in reference 64, the following results
P_e obtained when it is csmblned with shock-cumpression
inlets:

Duct13.9o - _ [Offset= 2r2),(e/r)2= 0.0156

-_

= 1.5

--

_t__t__-_t_
pt_

}YO

2.0

_t_pt_-pt_

&

Me

Pto

Pto

0 _,17oi0.931 o.09
1
2

._i
._i
_94

.98
.99!

_.00

-!:
Pt o-

M_

Pt

0.84 0..58 0.72

.14
.16

.84
.33

.7%
.83

.90
.99

.17

.83 .90 .97

Pt o

Pto

0.06

0.66

.i0
.i3

.80
.82

.l_

.82

-_re, the advantages of high supersonic compression are further reduced.

a )_ch number of 1.9, a normal shock wave might as well be used, and
_7, a Mach number of 2.0, a slx_le oblique shock wave very nearly produces
maximum pressure r-_covel_. 0swatltseh esteblishes this point in reference 90 by considering the arrangement
of oblique shock waves which would
produce the maximum static pressure behind the terminal normal shock wave.
This would be the best initial condition for a poor duet installation.

"I

-f

i c is shown that oblique shock waves pz.oduc_ no improvement

to a Mach
ntunber of 1.6 and that a sngle oblique wa_e is sufficient to s Much number
of 2.0.

.At flight Mach numbers greater than 2.0, another limit appears on
the n_nber of oblique shock waves that can be us_=<lbeneficially.
As
pointed out by Lukasiewicz
/_ z'ef_renc_ o_ and Oonnors and Woo]-iett in
reference 102, supersonic flow can be t'_vned and compressed by deflecting
surfaces through such lar6e angles that a normal shock wave must form at
the streamline which turns throt,4_hthe max'nnum angle possible for attached
flow.
This normal shock w_ve occurs at Mach numbers above about 2.2 before

i_

essentially _sentropic compzessmon can be achieved; at lower Mach numbers,

nearly i6entropic comFression _s possible without the ,occurrence of a
normal shock _ave from this cause.

.-

Boundar_-la_er
shock-wave interactior .- _ohably
the most important
limitation on supersonic compression
is caused by the interaction of
shock waves with boun "duty layers.
For iustance, Seddon in the note
appended to reference I03 shows that for a side intake without boundarylayer removal and only a nornml shock wa_:e for supersonic compression,
the _obal-pressure
loss due to this interaction was greater than the sum
of the losses from all other sources _t Macb nt_bers between 1.0 and 1.4

=
_
_.

and was about equal to that across a nc.'mml shock wave at 1.7, where;
in general, normal-shock
losses are unacceptably
high.
These interfez_nce losses were due to turbt_lent mixing in the flow after separation and
to changes in :_kin friction and. shock losses from "..heir_lues
in uuseparated flow.

"

The boundary layer separates at relatively low local supersonic Much

mmnbers, about 1.29 and F#'eater, when a nora-real,
shock wmve _nteracts with
a t'_mbulent boundary layer; it separates &_ very low su-_}ersonicMach
numbers, locally about 1.1, when the interaction is with a laminar
boundary layer.
(See refs. lOB through lOT. ) Of course, if the profile
of the boundary layer has develope_ an inflection (H_ 1.8 in incompressible flow) before the interaction,
a less intense shock wave causes
separation.
The data of reference 106 show that for the range of flight
conditions of interest in t_Is report, the static pressure-rls
ratio
at seoaratJon i:_ not a strong function of Reynolds number If the flow
to the point of reattach_.ent is t:_b_lent.
However, if transition oce_n.s
between separation and :'eattachment, there is a Reynolds number dependence.
In alr-induction-systea
design or testing in conditions In which a laminar
boundary ].ayer in th_ engine-flow
atreamtube could exist, provision should
be made for causing transition upstream of shock waves.
The reaso_s are
th_.t a shock :_ave of pr_ctlcally
any strength can separate a laminar
layer and that _my saving in skin friction due to maintaining
a laminar
layer is negligtble.
Also, the Reynolds number dependence if the initial
boundary layer were not turbulent could produce unreliable test measure-

:
.

_:
:
;
:
;
.
._

men_s.
Separation is to be avoided noc only because of pressure losses
but al._o because of flow unsteadiness
and nonun_ormity.
However, small

46

....
,

;:

'

C0_F_$:_.
': @:-': ":
:
:"{:CA

:.:

amounts of separation with subsequent reattachment are not necessarily

serious, and information is required o_ the allowable tolerances for
regions of separated flo_:.

"

:
!

!
!

With air-inductloa systems, the shock waves that interact _th a

boundary layer can originate from a change in surface slope, from neighbcring surfaces, or frc_ the normal shock wave which terminates supersono.
compression. Bogdonoi'f ._udKepler (ref. 10_) indicate that for local Mach
numbers through 2.0_ a static-pressure-rise ratio of about 2 causes separation. Gadd, Nolder, and Regan (ref. 106) show a value of 1.7; Nussdorfer
(ref. 104) suggests a value of 1.89; Lu_asiewicz (ref. 52), Seddon
(ref. 103)3 and Dailey (ref. 108) suggest 1.8, the pressure ratio across
a normal shock wave occ:_ring at a M_ch number of 1.3; and the c.'iterion
of Nitzberg aud Crandall [ (Usep/u_mltial)
. .
z - 1/2] corresponds to a staticpressure-rise rctio of 1.7 (ref. 109). Such differences are due to the
method used to determine separation and to test conditions. Nussdorfer's
criterion of statlc-pressure-rise ratio of 1.9 was derived from a study of
air-induction-system data which included both :?iaQeand con:ca/ compression
surfaces. If this criterlou is used as the one appropriate +o present
design methods for the case where a uo._na:shock wa:e interacts _th a
turbulent boundary layer, the limitations on shock compression because
of separation are those superimposed on the curves of total-pressure ratio
as a function of flow deflection angle and Mach nl_mberpresented in figures 6, 7, and 8. If it is assumed that the degree of separation at the
boundary determined by Nussdorfer's criterion is sufficient to reduce
in_uction-s)stam performance, it is evident that in the Mach number range
up to 2.0 inlets must be designed for nearly the optimum shock confi_Aratlon. If a smaller de_lectlon angle is used, the termin,_lnormsl _hock
wave is intense snoagh to cause separation. This interaction undoubtedly
decreases oerformance in cases where the boundary layer just ahead of the
normal shock wave is on the verge of sep_ation and where the subsequent
flow is not given an opportunity to reattach. For instance, the skatch
_u f!gur_ 7 shows a condition where the pressure rise in the vicinity of
the oblique-shock reflection could be sufficient to cause local separation
or at least disturb the boundary layer sufficiently so that the terminal
normal shock wave would erasureseparstion. The limitations for avoiding
seperstlon in this case are more severe than indicated in this figure.
Comparison of figures 6, 7, and 8 shows that a strict requirement of
avoiding bow-shock wave detachment and separa%ion due to the termiual
normal shock wave thro_gh a range of flight Mach numbers makes systems
in wlqlchthe configuration can be varied necessary at Mach numbers above
about 1.6 in two-dimensional flow and above abou_ 2.0 in conical flow.
(Other reasons for variable systems and information on those that have
been tested will be discussed subsequently.)
Separation due to changes in surface slope snd to impinging shock
wa,_esfrom other surfaces can be alleviated by reducing the pressure
gradient by dist_'Ibutiz_ the disturbance over some length. In other words,
discrete shock wave_ are to be avoided. For instance, C_pman, Kuehn,

-_

NACA EM A55F16

and Larson in some as yet unpublished

results found that the turbulent
boundary layer can vith_t_,d a large pressure rise on a curved surface
where it has sufficient d-,stance in which to re-energize itse]_f. (See
also refs
43 and llO.)

'_

C_ENTIA_'

If boundary-layer
-_%
_,

separation

',"'

47

due to interaction

with shock -_aves

c_-_nnotbe avoided ____r.

Inductlon-svstem, design_ it can, of course, be prevented by removing or re-energizlng
the approaching boundary layer.
Investigations of such methods are reported in references lll through llS.
The

inveJtlgations of boundary-layer
removal near the minimum-area
station
Both porous suction and slots show that some improvement in pressure

_
_
,_

by

recovery at low mass-flow ratios _an be achieved.

More importp,ut, however,
is the improvement in flow uniformity and steadiness over a vide range of
mass-flow ratios.
Similar results are obtained with blo_Lug methods of
boundary-l_yer
control in which the point of discharge i_ upstream of the
minimum-area
station.
(See refs. ll3 and ll4.)

--

_-

_:_
_.
:_
_:

To summarize, separation can easily be caused by the interaction

of
shock raves vith a boun _dary layer.
To avoid separation, the boundsa-ylayer profile approac _hing the region of supersonic compression shc_uld have
no inflection; changes in surface slope and impinging dlst_rbances
shoula
be distributed
to re&uce the pressure gradient; the proper arrangement
of
shock _ves
should be used to k,_ep the interaction pressure ratio at the
terminal normal shock -cave belo-a that which _uld
produce separation; au__
the initial subsequent cc_presslon should be sn__l.
Thus_ the mass-flov
ratio should be 'high to minimize subsonic compression behind the, terminal
shock wave, and _ nearly straight entav section should be used in the duct
to minimize the pressure gre;-ient an_ to pelnnit reattaehment
if some
selaration does occur.
The boundary layer can be removed or re-energized
to avoid or reduce the interaction.

_
_
:r

LiP design.- In su_,ersonlc flight, the problems c_ llp deslgn are

different from those of subsonic flight, for the.re,is no possibility
of
external streamlines converging upon the inlet and causing separation of

_,
"
_

the internal flow.

The problems arc those of locating end shaping the lip
properly to maintain high pressure recovery and low net drag without
severely compr_,__si_g these qu_.ities in s_bsonlc fllghc.

.c

_
"_
:

--

?
,
_
_
:
,

Tests of open-nose inlets to determine the effects of lip profile in

supersonic f!l_t
ere reported in references 23 aud 116.
It was found
that curved internal surfaces that are satisfactory at subsonic speeds can
be used at supersonic speeds at least to a P_ch number of 1.7 v_thout any
sacrifice ,n total-pressure
ratio.
In fact, a llp described in reference 23 vith (r/R)_= 1.19 produces higher Imessure recovery than a sharp
llp at Mach numbers to 1.9, and, s_ shown _m figure _, this lip m_intains

high recovery

,_-

to relatively

large mass-flov

ratios

at subsonic

Sl_d_,.

With internal-co, traction inlets designed for th_ coatraction rat.4.o

star_ (see p. 40)_ the profile _f the c_tracting
passage
can as _ell
be
....
_

_
_

"

i
I

43

::

C9__'.

_{ACA_4 A99FI6

',, ..:':..:.......

a straight line as a theoretically more efficient contour because the

permissible contraction is small to a flight Mach number of 2.0. For such
inlets _i_
_ an adjustable throat to Increase the contraction while in
flight to values less than Sstart, a straight-line profile at the lip is
_lso sui_fic_entlyrefined in this M_ch n,mnberrange. The deflection angle
at the lip leading edge should, of course, not exceed the angle for shockwave detachment or for regular reflec#ion (see refs. 53 and 9_0). However,
as shown by the results in reference 94, and as discussed previously, it
should be a sufficiently large angle to _minlmlzethe effects of interaction between the boundary layer and the terminal normal shock wa,e. (_e
results of Wyatt and Hunczak, ref. _, further show that an extended entry
section permlts greater supersonic compression in this type of airinduction system, presumably because the separated boundary layer which
follows a relatively strong normal shock wave has an opportunity to
reattach. )
Lukasiewicz (ref. _3) in discussing conical-shock inlets wibh sharp
lips shows that neither lip position nor l_p incidence have, within reasonable design limits, great significance in affecting pressure recovery at
Mach numbers less than about 2.0. I_[pposition is not important because
the velocity gradients for reasonable positions in practical conical flow
fields are small. Lip incidence has little importance because even if the
shock wave from the lip is detached, it is of small intensity in a desi._n
having the relatively large cone angle necessary for maximum pressure
recovery.
Although lip design has been found to be_of secondary importance
in regard to pressure recovery for external-compression inlets, it is of
great importance in regard to drag, which will be discussed later.

Mass-flow variation.- Air-lnduction systems without an adjustable

inlet area or a bypass _st operate through a r_=ge of mass flow as flight
conditions change. The previous discussion Qf supersonic cQmpression has
been concerned primarily with considerations of maximum total-pressure
ratio at a single design condition, usually the "critical mass-flow ratio."
This term denotes the internal flow _nen there is no subsonic spillage
and the terminal normal sl - wave occurs at the minimum-_ea section;
that is, when the supersonic compression for the system is maximum. If
the transition to subsonic flow occurs downstream of the mlnim_u_section,
the mass-flow ratio is the same as at the critical coadition because there
i_ olso no subsonic spills_e, but the total-pressure ratio is less because
the terminal shock wave occurs at a higher local Mach number. Such operation is termed "supercritical" and t_e total-pressure ratio is determined
by the flight conditions and the requirements of flow continuity and of
the flow schedule of the engine. From equation (16)

" . ,

. "

NACA RM A55F16

CONFIDE._'

_ :'"

I_
9

85,_)_A2

Pt--2
=
Dto

(26)
7+I

Thus, for a specific mass-flow ratio, a reduction in inlet area proa:aees

a low pressure recovery for a given engine corrected air flow and f]__.ght
_ch number; or, for a.given inlet area and mass-flow ratio, corrected
air flows or flight Math numbers shove the design value also reduce the
total-pressure ratio. Systems are sometimes designed to operate at supercr!tlcal conditions in order to avoid flow unsteadiness which often occurs
at mass-flow ratios Just below critical, particularly at angle of attack
with systems hsvlng a large amount of st,
personic compression and no Luterference which alleviates angle-of-attack effects. (See, for instance,
refs. ll7 and 118. ) Wh_m the trans_tiom to subsonic flow is upstream of
_he inlet, the suberitlcal condltic_a,a normal shock wave occurs exter
. nally and flow is spilled behind it to reduce the mass-flow ratio frc_
the maximum. The possible total-pressure ratio at these reduced mass flows
can be calculated from the known shock pattern if the pressure rise through
the shock waves is not so great as to cause separation losses or to distort a boundary layer enough to change the shock pattern.
Experimental inves_Igatlons of isolated alr-inductlon systems through
the range of mass-flow ratios show, in general, that inlets vhich attain
very high total-pressure ratios at the critical condition are very sensitive to changes in operating flow condltion:.. That is, total-pressure
ratio is markedly reduced if operation is ve_j far subcritical, and, as
_-ithany inlet, recovery decreases rapidly in the supercri:ical range.
The data summarized by Lt_asiewicz (ref. 53) illustrate this fact. Thus,
an open-nose inle_ which _ccepts supersonic compression through a normal
shock wave does not, as sho:,min sketch (ll), attain a high total-pressure
ratio, but essentially the maximt_ total-presstu_eratio with uniform flow
at the compressor face is maintained throughout the subcrltical range.
The total-pressure ratio which has been measured in experiments is that
calculated for the normal-shock _ave minus the duet losses. An internalcontraction inlet suffers an abrupt total-pressure loss and operates as
a noznal-shock inlet as soon as the flow becomes subcritleal. Conlcalshock inlets designed with more than one oblique shock wave also have this
disadvantage of an abrupt decrease in total-pressure ratio at subcritlcal
mass-flow ratios, presumabl_ because the botmdary-layer profile approaches
that for separation in passing through the large _dverse pressure gradients
of the supersonic compression. However, conical-shock inlets with one
oblique sLock wave designed fo_-near-maximum-total-pressure ratio can
maintain a high level of pressure recovery well into the subcritlcal
range. Use of less thsn the optimum cone angle (included angles less than
about 50) produces a terminal nar_l shock wave of too great intensity

L a

II

I -_

,:

/
/

50

""
:"

.....

.......
(

_'_

.(It

....

FM

1-

ql'ot

which adveruely affects subcritical operation. The most disturbing difficulties at reduced mass-flow ratios are flow nonuniformity and unsteadiness which are caused by separatiou that can arise from a number of
sources. An extended subcritical range of mass-flow ratios _n which the
flow is steady can be obtained by cholce of _he proper shock _attern and
duct design or by bounds-_y-layerremoval.

Since fixed-area intakes _-anbe unsatisfactory at mass-flow ratios

other than that chosen as the design point_ systems must be considered
in which a constant, or nearly constantp mass-flow ratio maintains a high
level of over-all Induction-system performance through a wide rauge of
flight conditions. _lis can be accomplished by varying the inlet area;
or, for a fixed inlet area_ excess air can be bypassed to satisfy the engine
air requirements while operatin6 the inductic_ system ne_r its best design
point. By these methods the reduction in propulsive-system perfo_-mmlce
fr__-_
-_d_i_ivedrag_ reduced pressure recovery, or flow nonuniformity a_,d
unsteadiness can be avoided at the expense of weight and complication.
For aircraft which must fly at widely _ifferent conditions of power, altitude, and speed, such ec_lication is necessary, The best arrangement
for any particular aircraft requires detailed evaluation.
Perhaps the simplest variable ss_tems for matching the air requirements of an engine are.an auxiliary scoop (ref. 74) and a bypass (ref. 119).
With the feigner,the main inlet is matched in area for the high-speed
flight condition and an auxiliary scoop is opened for flYght at lower
Mach numbers. With a bypass between the inlet an_ the engine, the inJet
area is generally chosen fur the altitude cruise condition and is large
for flight a_ high speed or low altitude. The excess air is dumped overboard through the bypass. The analyses of references 7_ and 119 show
that these systems have various advantages and are superior to other systems
fur certain flight cond/tions. Experiments have demonstrated that at Mach
numbers up to 2.0 the dra_ of the bypcss can be small as long as the air
is ejected neaxly parallel to the local flow direction. (ref. 120).

_ ,

Another varia'ole system is a conical-shock inlet in which the center

_ody can be move_ fore and aft to repot!atethe mass-flow ratio. This is
the tranalatins-cone inlet (rafs. 121_ 122_ and 123). When the oblique
shock wave from the cone apex intersects the inlet lip_ the mass-flow
ratio is the maximum. When the co_e is moved fo_v-_rdrelative to the llp_
the mass-flow ratio is reduce_ by supersonic spill_e and the additive
drag i_ not as large e_ if the spillage were behind a _ormal sho_k wave
(_ee p. 6_). Gorton shows in reference 122 that such inl_s c_u be designed
for high pressure recovery at Mach numbers from 1.5 to 2.0. The effects
of various _esisn c_r_ises
which must he ms_e in the design of such
t_anslatlng_cone inlets are studic_iin reference 123. The perTormance
of three inlets each in c_Ination
with three tux,
bo_et eagines is c_pare_. The choice of inlet was foun_ to _epend upon the enszne air-flow
schedule an_ the flight co_itions aelec_
as critical. I_ reference 31
_est_ with an operating turbojet e_e
of a translating-cone inlet an_
j_"_
,_':_'_,
:._:,_.

.... ,,,,.;.-,_
. ,_,,,

"'111111 -- ' '.-_'7_',._'_

o_,
'" ' - " " 1""

"

'

"

. "

- _''''::
, ,,,,
,,j
., ,,ILIL_L_
...............
,....................
"'_
"."_",I ,_'_*_'_"::::'J',,_'_'_"_"':n',_,._,:_i
_:'_i_,_,,T,,_gR/_vz_crc"_
, "-
_-:-" _'" "" ._,'",'.:-.
, " " ".,_",____,."
" " ' .'. ":...

,,,

.'..... /

,""' ......

AS_F16

_O_q_

51

of a bypass system a_ flight Mash numbers of 0, 0.6, 1.7, and _.0 are
described. Both systems e_zLminatedflow spillage behind a normal shock
_ve, but the,net propulsive forces _re not determimed. Thi_ imvestigatlo_ was extended in reference 12_ to include automatic control of a system

ii

with a trausiating cone and a bypass camblned. By s_.nsingtotal pressure

at the cone tip and cowl lip and static pressure Just inside the inlet,
the oblique shock "_mvc could be maintained at the lip _
the terminal
shock wave could be positioned Just inside the cowl. The total-pressure
recovery varied from 0.92 to 0.88 as the Mach number was changed from
1.7 to 2.0 (see fig. 9).

Air-induction systems in which the deflection angle of the supersonic

compression surfaces can be varied to provide for emCee-inlet matching
through a range of flight conditions have boon tested ix__ wide variety
of arrangements. In reference 125 a precompression ramp followed by a
variable second _u_p was used to improve the performance of a t-din-scoop
installation wi_b fixed-area inlets. Precompression ramp angles of 3
and i0 were tested in combination with the variable second ramp; the
lar@er angle produced the better pressure recovery. However, nonuniformity in the total-pressure distribution at the diffuser exit of more than
5 percent existed for all the configurations tested. An und__rslungscoop
ha_ing a variable horizontal ramp or a variable ver_ical-_.dge compression
surface is described in reference 112. The total-pressure ratios attained
in tests at Mach numbers from i._ to i.$ are shown in figure 9- It i_
seen that these systems produce relatively high total-pressure ratios.
Further tests reported in reference 112 of an underslung scoop with
boundary-laye_ r_moval through p_m_u_ suction over the compression mtrfaces
show an increase in total-pressure ratio of as much as 5 percent with
nearly the same gain in net propulsive force.

""

:_

'_.

_.
o_
._

The problem of providing high values of net propulsive force for a

self-accelerating ram-Jet missile requires s_e fca_nof variable inlet
area, and t)-evariation mast be acc_nplished in a simple manner. A dropable cowling to provide, in _ffect, two inlets is reported in reference
126. A cowling was exudedto a double-cone inlet designed for Mo = 2._
so that the ccmbiuati(m was a normal-shock _let, and tests ware made at

_
._.
!
'_'
_

Mach numbers of 0.6_, 1.5, and 2.0. Substantial :unproveme_tsin net propulsive force over tl_t of the _ouble-cone inlet "wereobtained at these

Mach numbers,

i_

In._._.igations_ inlets having k_th variable inlet aI_ throat areas

are reported in references 68 and 127 and the pressure recovery characteristics are e_mparc_ with those of other inlets in figure 9.

.:,

Angle of attack.- As in subsonic flight, the flow approaehin_ a_

air-induction sy_t'emab supersonlc speeds can be at an an_.._e
to the system

'_
,_

axis because
of the attitude
A_ in _he case _f _s-flow
total-pressure
x_tios
are,

._

_ the aircraft
end because of induced effects.
variatimas_ inlets which attain very high
in general,
sensitive
to an_.e of attack.

_..'

......

, ....

---

_,._

!
elo

Lukasiewlcz (ref. _B) shows that an open-nose inlet with normal-shock

compression is not affected by e_le of attack up to _o; but the other
inlets, that is, the internal-conzractio_ sad conical-shock types, suffer
losses in maximum to%el-pressure ratios of from B to _ percen_ at engles
of attack of 5. (See refs. 93, 122, and 128.) At higher angle_ of attack
separatioh from the lower lip of symmetric open-nose inlets reduces the
pressure recovery until at angles of attack of the order of 20 at a Mach
number of 1.42, the maximum total pressure ratio decreases frQm 0.9_ to
0.85 (fig. 10). The reductions in pressure recovery are greater for
con_.cal-shocksad internal-contraction inlets.
"

,
:-

Several methods for maintaining the zero-angle-of-attack level of

pressure recovery with changing angle of attack ha_e been proposed. A
summary of test results is presented in figure lO. Beheim suggested a
pivoted cone in reference 129, and found that relative $o a fixed-cone
inlet, an increase in maximum pressure recovery, mass-flow ratio, sad flow
steadiness could be obtained at angle of attack. However, there was no
improvement in flow uniformity, and maximum pressure recovery occ_red at
a red_ced mass-flow ratio. A method is proposed in reference 130 lu which
an inlet _Jl_ha vertical-wedge compression eurface inside a conical cowl
was modifiel by perforating the wedge cenre.rbodY and cutting back the
lower ha2f of the cowl lip. Total-pressure recovery obtained with this
_nl_t, although lower than with comparable conical-shock i_lets, was
essentially constant with increasing angle of attack up to an angle of at
least l0, the limit of the tests. There was an iucrease in the subcritical mass-flow range for steady flow, and twin-duct instability was elim_hated by cross-ventilation through the perforations. Other methods for
maintaining the level of pressure recovery with chang__ugangle of attack
consist _ either c_-ntin_+he inlet plane (refs. 131 end 132) or adding
flow deflecting surfaces (refs. 26 and 133). _rangements for utilizing
interference from other aircraft components to keep the flow alined with
the system axis are discussed later under INTEEFEEENCE.

DRAG

The design objective in regard to drag is to mi_._ize disturbances

in the external flow; that is, to maintain as mlch laminar flow as possible,
to avoid separation, and to avoid shock waves or reduce their intensity.
Since the forces of skin friction occur on all external surfa-.esand arc
not limited to those of air-induction systems, no detailed discussions
of skin friction or of the allied problem of boundary-layer transition
are presentel in this report. References 41, 4_, 134, 13_, and 136 conrain design imformatio_ on these subjects.
!

In this secticm, o_y the drag -_fisolated air-induction systems is

considerod; that is, wing-ro_ inlets and types which include interference
drag forces are not discussed. In general, drag coefficients are based

,:
_h..

'

__

'lllnl

I.

II

,,_.,__._:;:;_::::..,

,::.l_

[l_..,,...:_J_._

,. .:.: ,: :..: _, ..,:' . .....

:.&:.- :.
. . ..,

;,_'
....

-C

_
._
_

on the maximum frontal area of the cowling or fuselage.

As describel
prevlou_ly, scoop incremental or additive drag should be computed _o the
stagnation point on the inlet lips; however, since the location of the

_
_
__
_

stagnation point is seldom known, these quantities are here cc_pated to

the plane tangent to the leading edge of %he lips.
As discussed in reference 23, such an assumption is conservative.
In order to have a reference
for the relative importance of _he _._ag components considered, the following table of representative
aircraft dimensions and total drag coeffici_ats has been compiled.

/
_

Xnle_ Ma_d=.=
f_selageA_

-_

Aircraft

area,

A2,

sq ft

_!

"_i_

sq ft

Wing
area,

26.40

F_D. _
F-94C

_._8

_3.00

.171

_-86_

_._

_.I#

.102 _.o

_._o
3.72

_._o
26.50

AM

3,

,q Zt

#.#_

F-84_
_F-9_,
_-_F

._,.

AM,

?-100

_,-_1
F3H-I

f.,cUt._1area_

Mdrng

19._i
_._

_o.o

_15.0
557.0
232.8

T_ansonic

s_"_lel_o=

0.169 376.0 0.070

_

r _Se

.m

I.i0

'0'.92 0.01_0 0.0_35

._
.88

.o_7_
.o630
.015
.0515

.045

.90
.76

.0100
.0_5

.06_

l_er"
Supersonic

.0373

= i.5o crimes
0.0_30 137

1_7
137

.0380

.o_

.8_

.016o

139

.057

.77
._0
.80

.01_0
.0100
.0140

_._O
1#1
i_

661.0
385.0

051
.o6_

,90

.0100
,o_o

OZO0

.047

.87

.0140

.07_0

_.20

33 6O
_..7o

.i_

FTU-I

3_._0

_3._

.137 _96.0

0_70
,o3_o

,_

1_3
I_

i_5

Thus, _u approximate figure for the ratio of maximum cowling or fuselage

cross-section area to wln_g area for present-day
aircraft is 0.i and the
supersonic drag coefficient at a M_ch number of i._ is about 0.04.
This
figure corresponds
to 0.400 based on maximum_ frontal area. Drag-coefficient
reductions of 0.00_ at supersonic speeds and 0.00_ at subsonic s_eeds due
to improvement_
in the air-induction
system represent 1.2_-percent reduct_ons In airplane drag.
Such increments in drag coefficient _'e probably
the limit of prelim_lary design accuracy and are the leas_ significant
figures worthy of consideration
in the fo_lowlmg discussion.

_'
_
_',
_
_

i_7
138

3_3._
_3.0
302.0

_-i0Z
_-_05

.:,

_
._i
:-_
-:_
!_'
""
_
'_

?,i,

4,*I_

_
-._i
m_
-_
!_

SubsonicFlight

..._

In subsonicflightbelow the M_ch number for drag divergence,-_he

main drag problem of air-inductionsystemsis to reduce skin frictionby

_
_'_I

delaying

_(:_,

bc_ludary-layer

transition

and by _Linimizing wette,_ are_.

D_ag

due to sep_rationic of little concerto

even for the re_tively sharp lips

,'j_

of supersonic aircraft because, as shown by thediscusslon

of sketch (8),
mass-flew ratios are near or above i and the a_ity
_f the
._xternal

_i
!'_'

flow relativeto the inletlips is small. For subs_ic aircraftin whica

,:,_._

._

i_ is desirable

_%

;i
._
'_

and low mass-flow ra_ios, exte_ml

separation can be avoided by use of
blunt lips.
At the high ar_gles of attack in land_g
and take-off operations, mass-flow ratios are greater than i_ so the engine-induced
flow

to minimize

internal

losses

by having

a large

_nlet area

,_,:
2_
_

94

: .."
'
C O_:;D_WZ_-T,_
_ _ '
" ,,"

NACA EM A95FI6

counteracts the tendency toward external sep:_rationon upper inlet lips.

Climb with Jet-powered aircraft ordinarily occurs at relatively higja
speeds where the mass-flow ratio can be less than l; but, because of the
_peed, the angle of attack of the airplane is not large. At high subsonic
speeds, low mass-flow ratios must be avoided if divergence of the engineair stream_ube ahead of the inlet and shock stall on the inlet lips is to
be prevented. Thus, since the external shape of an air-inauction system
can be considered independently of the _ _t shape (see ref. 2, p. 60),
the design problem in regal-dto subsonic drag is to select an external
contour that encloses the necessary induction system and maintains laminsx,
shock-free flow thzo'_ghthe required range of mass _low and angle of attack.

i
!

"

,:

The net drag of an air-induction system is entirely due to skin friction as long as the flow is unseparated and Irrotational outside of the
boundary layer, for, as shown pzeviousl), the pressure force in the drag
direction along the free surface of the englne-flow streamtube in equation (7) is offset by a pressure force on the cowling surface in the
thrust direction. The experimental _-esultso_ Biackaby and Watson
(ref. 72) show that for a wedge-shaped lip profile (7-1/2 w _ge angle)
there is no net pressure drag in low-speed flow at mass-flow ratios above
0.8; for blunter llps, lower mass-flow ratios (less than 0.6) were reached
without external separation that caused any appreciable loss in lip suction
force. Similarly, measurements to a Mach number of 1 show little change
in net drag with mass-flow ratios as low as 0.8 for sharp llps and to
less with blunt llps. (See refs. 76 and 146.) From these results, it is
apparent that no net pressure drag need be experienced at subsonic
speeds in the mass-flow-ratlo range of interest. However, for the thin
llps required for high-speed flight, a very localized lip suction force
to counteract additive drag is not conducive to laminar flow, for a small
region of very low pressure is followed by a rising pressure which causes
transition to turbulent flow in the boundary layer. From the criterion
of }'_cmluand Millikan (ref. 147) that laminar separation occurs in a
posltl_e pressure gradient when the local velocity is about 0.9 the maxlmum velocity and that laminar separation results in transition, it appears
from the pressure-distribution data of reference 146 that at flight Mach
numbers greater than O.8 with a sharp lip, mass-flow ratios greater than
0.9 ere necessary to prevent transition from occurrSng on the lip. For
the NACA 1-series irlets of reference 76, mass-flow ratios to as low as
0.8 with no serious adverse pressure _u-adlentseem possible in flight to
a M,eh number of 1.O, although the scatter of the data prevents a definite
cc_ucluslon. The pressure-dl_%rlbutlon data on NACA 1-series inlets at a
Mach number of 0.4 (rof. 80) indicate that for usual ratios of inlet to
maximum dismetert
no suction
pressure
peak with subsequent
transition
need occur to mass-flow
ratios
as low as about O.h at zero an61e of attack.
Similarly, the "class C" profiles of _ehemann and Weber (ref. 2) create
n_ _adverse pressu_ gradient until very low mass-flow ratios, less than
O._:_ are reached. These shapes thus can produce low drag in subsonic
flight; however, because of their blunt shapep they create hi_ wave drag
in sul_ersonieflight (see, e.g., the da_a Gf ref,,I_8). For aircraft tha_

_
_

i
I'_
!

I:

!
Z

NACA RM

"

55

_ :" _ ;_" _,

:..

fly supersonicalty, thinner llps must be used together __ch a relatively

high mass-flow ratio, greater than a:nout0.8, to have low external drag
through the speed range.
:

The NACA l=series profiles (ref. _O)and those described by K_chemann

and Weber (ref. 2) were designed accol'dlngto the _riterion of maximizing
the critical Mach number of lips, that is, the flight Mach number at which
sonic velocity first occurs on the profile. I_ was thou_hC that this
Mach number would indicate the beginnlr_Eof the transonic drag rise ana
thus should occ_. Fitas high a speed as possible. Th_ drag rise is well
predicted by critical Mach number for cowl shapes over which the pressure
distribution is nearly uniform (see ref. 2); how:_er, it is not predicted
by the critical Mach number as applied to local high-velocity regions z5
Since, from the skln-friction standpoint, shapes must be chosen that have
a nearly uniform distrJbation of pressure the critical Mach number is a
good indication of the drag-rise Mach z_mber for the shapes ,:f interest.
The NACA 1-series and the E:uchemannand Weber class C series can thus be
used with reliance placed on the predicted drag-rise M:.cb_mbe:'. For
high Mach numbers of drag divergence, the cowls mu_t b_ slender as shown
in sketch (14). The results of reference 148 show that at high mass-flow
IO0
..,.,
ratios, the details of lip shape for
_
slender cowls have little effect on
_
ACA
the magnitude of the external pressure _ .
_,_'" Eli t-50-200

""

drag to flight Macb numbers of-.

"

axial distribution of cross-sectlon

area, particularly when in combination with other aiz_]anc components,
as
be discussed
later. is the
Thewill
imprtant
cnsideration

_,
._
_.
:;

0_#__,
.e

i_ 7

"_"

_
--

//'

:
i

_ _"OI'C0"10050

_;////

E]

'

As shown by tests reported

_=0"
in references 150 and 151, the
.6,
Mach number for drag divergenc_ and
0
2
4
6
D
IO
the magnitude of the transonic drag
Moss-flow
ratio,m,/mo
rise for ducted bodies can be deterS_etch (14)
mined experimentally by tests of
eq_uivalcntbodies. That is, the solid b_ly equivalent to a ducted bod_
from the external-_ve-drag standpoint i_ the ducted body with the freestream area of the engine s_reamtube subtracted from the longitudinal
area distribution. At mass-flow ratios less than i, an equivalc_t body
thus has 2 b]nnt nose; nevertheless, the experiments indicate that the
ASThe unimportance of localized high-velocity regions on cowls is
s_mlogous to the observations of Nitzberg and Crandsll regarding airfoils
(see ref. 149). Here, it is shown that drag-rise M_ch n_mber can best be
predicted by appljing the Praadtl-Glsuert rule to the pressure coefficient
at the airfoil crest; in other words, superscmic flow _mt extend over a
considerable portion of the surface for the drag rise to be predicted
accuratel_ b_ the critical .Ma..Ch.,
number.
, .

"

I IIII

L ""

.......

7 __

III

...........

_:

:_.l_--iL-i

2;,_2,_._:w_

........

,_
,_"

_
;:_.

'_
::_
._._

:"

''

"'

T
JL_

c f

eq_k_val_it-bo_y -_ethod is a reliable

-,

indication

or duote_-bo_y

to mass-flo_" ra_icS as low as 0.7.

The accuracy of eb-_.smethod
for fair equiv:--lentbodies having high fi._.enessratiss.
-_'Er'leel" ar.r.

tingle

of

at, tack

af

_r-Nuet-ton

systems

drag r_s_ -

"

is greatest

or. external

drag Js generallynot a sp2_ouspvob!_, At the lowestma_s-flowratio

that -_oul_normally occur in high-_eed
flight,
off the _-_ler off 0.6, the
-:

p_ssure-distrlbut.i_
da_ o_ the _NACA l-_erles Nets
zhow that angles
to4 can be reached without a serious suctl_n presaure peak for cowls
that are not too _lender.
A _sl_Kler cowl, the i-_0-_00, for instance,
devel_
a suctiom presm_epeel[
at this angle wh_e_
the 1-50-i_0 does
not because of the thicker lip.

su, sc night
i.c

The follAr__ug dis_asslon of the drag off isolate_ alr-ir_luc_io_ ._st_ms

at supersonic speeds is arra_
according to the e_po_ents
"_h
m_ka up
the net drag _ sho_. in sketch (i_). He_e, t)_ical varlatloms of tae c_mpoments of the net dra_ coefficient _Ith m_ss-flow ratio f_r a given f_i_ht

?
+

'........

.:+_

JtJl

i :"

'

=_

+,*

57

1.

_ue external
wave (or pressure)
drag When the
with no spillage, as in sketch (16a).

syste_.--perates

2.

The _ressure force on the _ef!ect_d engine-flow stre_mtmbe, as'ln

sketches (16b) and (16c). (This is 6dditlve dra_.)

3-

The change in external wave du_

due to a reduc_-ionin mass-flow

_

from
the maximum,
as _n sketch (16b) or (16c). :('This.
is called
the cowl
suction force.)
II.+ Skin fricIlon (as mentioned on p. _2, thls.csmpon_at of the
drag is not discussed in this report).

(a)

(b)

No spillage

Spillage due to o.normol

shock wave

(c)
Spillage due man oblique
o.d a no_?mlshot..
_-wove

'

Sketch (16)
External wave a,_g with no spillage.- Several m_thods have been
developed for estimating the pressure distribution and wave drag of axially
symmetric ducted bodies at zero angle of attack with an _tt_.chedshock wave
on the llp. These are listed with pertinent references as follows:
Linearized methods
Brown and Parker
Lighthill
Ward
Jack
Moore
Ferrari
Bolton-Shaw and Zienklewicz
Perker

References
86,192 /
153,154
155,156
197
158
199,160
161
162

Second-order method
Van Dyke

:"

:<
_
;_
_+'+'

_._
163,164,165

,_

}
v,

,:,+
<L J

Higher-order

7--..-"_.

_-_'-'_'-Tmt'

,e

method

ov

References

Ferrl

86,166

In general, t_e greater accuracy of the more compl_cated methodz

is oh -'-'o_-__.t th._ ___pensc _ _eatar
labor _._making ca!calations.
_Iso,
since _ae simpler methods utilize more assumptions,
their range of applicability !J le_:. but is often sufficient fo_ design purposes.
In refer~
ence 1_7. the linearized met_od of characteristics
is compared with the
source-distribution
method of reference 152.
It was found that to
produce the same accuracy the l*_uearized method of characteristics
requires
much more cemput_,g time.
In comparing with the characteristics
method
of reference 166, this latter procedure was found to require by fax" the
great2st amount of effort, but the comparison sho_,_d thet for large flow
deflection
angles _t _;ne lip (15.5 ) the linearized methods underestimate
the pressure on the lip and hence the drag, in this case (}4o= 1.8) by
36 percent.
In terms of airplane drag, such an error would be equivalent
to roughly 1 _ercent.
Ferri compares calculations by the method of characteristics with those of the small-distur.ba_ce theory of reference 192 for
a cowl with a 3 lip angle at a Mach number of 1.5 and finds that the
approximate method underestimates
only slightly the pressures along the
cowl.
In fact, rotation need be tsken In.to account only when a s_zong
curved external shock wave occurs and the variation of entropy along the
shock wave is great.
Similar comparisons at a flight Mach number of 2
have been made between the methods of references 152 and 164 for a conica _
and a curved cowl.
The conical cowl Lad a 3 semiapex angle and the ratio
of inlet-to-max!mum
__rea was 0.676.
The curved cowling had a 12.9 iuitial
deflection
angle, an area ratio of 0.5, a length-to-dismeter
ratio
Z/dM
of 3.18 and a practical profile which is defined by the relation

x = 4.38(r

- l) + 15.Sl(r

- l) 2 + 77.07(r

- l) s + 1.73

The outer surface of this lip is parallel %0 the local flow direction
when the shock wave fro_, a 50 cone in%ersects the lip. 16
_%e results
of this comparison are sunm_rized in the following table:

iSLukasie_icz
in reference 53 _eesents &esign ir_forr_ation on the
flow direction in conical flow fields and on _he conditions for regular
reflection and shock-wave detachment.
It is shown that a llp incidence
angle can be _elected that is good for a wide ran_
of Mach numbers.
Also,
a conical-shock
inlet designed with a straight lip to provide internal
contraction cezmot have regular r_flection at Mach numbers up to 2.0 if
cone angles greater than 25 are u_,_d. In two-dlmensional
flow, attached
flow on a straight llp is not possible at ,a Mach number of 2.0 if the flow
deflection
angle is greater than 15 .

"
',

___-

. ,

.:

' :'

, .....

,,,.

'/2.

, !

NACA _

A55FI6

....

"

' "' ' _

59

IDrag coefficient,
Dwave

Method

,
qAM
Conical
C-._v_d
[ cowl
cowl

Pressure rela .ionship

o_.r Cp _-_
ist order

l +-7-

- ve2/]

-1

o,ol78 0.035

.... do ..................................

.OI7_L

.030

Do....
Do....

u2-

u
Cp = -2 Vo

v2 + -- _
Vo2
Vo 2

.Oi7h

.031

Cp = -2 u
Vo

v_
Vo2

.017I'
,

.02"9

.o187

.028

D_---- cp= -2 %

As in the previous comparisons, the first-order method underestimates the

pressure on the lip and the drag; the difference is small if the deflection
s_igleat the lip is small, but the error becomes sizable in terms of cowl
drag for large _uglesxv (in this case I_ percent when the complete pressurecoefficient relationship is used). In terms of airplane drag coefficient,
even this error at large deflection _._.cs is negligible. Van Dykc in
reference 165 sho_m t_.atfor cones _t Mach numbers less than 2 and cone
angles to 30, the second-order and exact theories give practlcslly identical results. In this reference, it is also show_, as indicated in the
table, that higher order terms sho,xldbe retained in the pressure relationship for calculations _nvolving three-dlmensional flow. From these comparisons and knowledge of the shapes tha_ are of practical interest, which
will be discussed subsequently, it Is concluded that since large lip angles
create large drag forces that must be avoided by the designer, the linearized methods are of sufficient accuracy for most design p_rpos_s.

<
f

Comparison of the quasi-cylindrical theory of Lignthill (ref. 193)

with experimental measurements of wave dreg is made in referenaes 146

,_

"

I71n applying the second-order theory to the curved cowling, it was

found that considerably more computation time was required than expected.
Reference 164 gives certain rules for selecting intervals for computation.

',,?
_
_,

Whereas about 6 intervals are sufficient for solid ogival bodies, the
curved cowling required ii intervals, which incr_.'ased
the labor of corn-

ii_i_
.,_

putati_n fourfold.

./_

___

"-

]1 i

ii

I ii

R iii

i i

I I

m l

I I

-;.'_

.....................

......

--

_,

.....

....... , f..

"_.

IlJll

F
{

c ,

_C'-

_K

*-

_a

and 167. It was found that in spite of the fact thst the mo4els were
not quasi-cyllndl'ical (the ratios of..inlet-<)-maxi_r_ area were 029 and
6.90, _
the corre_onding
initial lip angles were 11.8 and 7.3 ) the
agreement was satisfactory,
as indicated in the followin% table:

Model

Mo Externsl wave drag coefflcler

__/
Measured
Theoretical
---[

Aa/_4

= O. 29 i. _l I

:.821

A2/A
M = o.._o
l._.ll
1.82 i

0.1!9

O. 136

.099

.1o4

.049

.0_

.o4o

.04:

tI
Error
jor_diction '
percen_
14

12
2.5

The theory overestimates the drag coefficient

in spite of the fac_ that
it imderestlmates
the cowl pressures because too large a frontal area is
assumed for the initial portion of the cowl in these cases.
The experimental meam_ements
also substantiate
the followffng predictions:
!. The pressure at the cowl lip corresponds to that downst,'_l of a
two-dimens_onal
oblique shock wave created by the lip deflection a_gle.
2.
The pressures on the rear of the cowl approach asymptotics_Lly
the value for a cone with the same slope.
(This is true for 8__I
mass-flow ratios. )
:

3- An expansion about a discontinuity

_n _urface slope is a Prand:lMeyer expansion.
At reduced .mass-flow ratios, the Mach number ahead
of the corner is determined by the local static pressure end the
total pressure behind the normal shock 'wave.

"
',
,

"

"
:

I '

At a Mach number of i. 33, the theory predicts the pressure on the cowl
I:D as well as it does at higher Mach numbers, but at Mo = 1.17 the
experiments show that the pressure is overestimated.
At lower supersonic
Math numbers this tendency increases.
It is therefore conc!ud_l that the
lower limit at which tlm linearized theory should be applied is a Mach
number of about 1.2.

Warren ar_ Gunn in reference 168 have extrapolated Ward's firstorder theory for conical cowls to small values of the ratio of isis,-tomaximum area.
The effect is to l-educe the overestimation
of wave dr_g
shown in the previous table.
Their method csm be. slightly improved at
low values of A2/AM and Mo
by using exact values for the drag of cones
(A_/A M = O) and calculatioas from second-order theory to indicate more
closel_ the proper trend of the extrapolation.
Results from such a procedure are shown in figure ll. (Drag coefficient
is based on maximum frontal
area. )

External profile.- From considerations of strictly supersonic flight

"_ _h inlets having no spillage, the lin,_arizedtheories-have beer. used to
determine the optimum profile Of ax_a'Llysy_e_rlc bodies fr_ the stagnation _olnt to the posit$o_,of maximum diameter. Ward (ref, 169)coucluded that the profile is very n_srly a straight li_e, that is, a st_,_ight
conical taper. Jack (ref. i__7)calcu3ated the drag of several profiles
for a conical-shock inlet at a _ach number of 2.0 and fotu..,.
.m
_ _hat less drag
was produced by a conical taper than the c_ved, profile<s. Using more exact
method_ and. imposing certain restrictive conditions, Ferrari (ref. 160) and
Parker (ref. 162) have found that the optimum profile is corved. Similarly,
Walters (ref. 150) and Howell (ref. 17O) have applied the transonic-aresrule concept to the design of bodies with nose inlets an_ have fom_._that
the method suggests a curved profile and does produce low drag. The method
is to add the longit _6inal area distribution of a minimum-drag solid body
_nd the area of the ez_ine-air streamtube to obtain the area dictribution
of th_ m_imum-dr,_ ducted body. Not only did this method produce a lower
drag at full fl(,_t_,authe other bodies _aich were tested, but also it is '.
stated in refere,lce!70 that more cowl suction force is obtained at reduced
mass flow. Bowever, the improvement in this regard is of small magnitu_
in terms of airplane drag coefficient.

"

been
with
of 3
than
that

In order to compare these proposed optimum shapes, calculations have

made for Mach numbers of ! .4 snd 2.0 for practical nacelle _h_pes
ratios of inlet-to-maximum area of 0.16 and 0.36 and flaenes_ ratios
and 6. (As _howz oy the data of reference 76, fineness ratios less
S create large drag. Fineness ratios greater tha_ 6 are so slender
small differences in profile have a negligible effect.)

....... YzLnlmum-c1_ag coefficients based on maximum cowl area for two

optim_ cowl shapes
Mo
Shape

Conical
(Ref.
Parker
1621

Am/AM = 0.16

0.099

0.019

.056

.016

Mo = 2.0

1._
A_/AM = 0.36

A2/AM = 0.16

A_/AM " 0'36

0.032 ! 0.010

0.0_9

0.016

0.0_9

0.009

.0_8

.014

.0_3

.008

.031
............
I

.00_

To indicate the differences in shapes, the radii of three minimum-drag cowls

_i__"

R_

62

@@

.': .. :.:.

A 1'6

ere compared with the conical

cowl in sketch
(17).
This_ coa_r._isoz
sho_
that both the difference_ in drag and radius distribution are small for
these low-drag shapeej and it is concll_led,as in the case of opti_._xm

io

I.....I

.
.
,
Kormonprofde,re, 171
witharea rule_

"
.6 ..,,,,mmd_

''"

Porker,
ref. 162!
Me_1.414

-_.___'#

.-

B_P_

o.,

witharearul6
_'-" M4 power_o6/, rlf 171

......
Az/AII-.3S
tld_ 4.7!

.I

._

xA

.6

.7

.8

.9

1.0

':"

Sketch (17)
solid bodies (see ref6. 171 and 172), that there is little difference no
matter which shape near the optimum is selected. For most practical purposes the conical cowl is the optimum shape.

Warren and Gunn (ref. 168) have presented charts for the optlmum
angle of conical taper and the corresponding drag coefficient (including
skin friction) as functions of "Mch number, skin-frictlon coefficient,
and area ratio. For _ given area ratio, an optimum conical angle exists
because the hess the angle the smaller the wave drag 'butthe greater the
skin-fric_ ou drsg. Charts rerJultimgfr_ the altered calculations mentioned on _age 6J are shown in figure 12, and they show that for a given
area ratio and skin-friction c_efficient, an increase in Mach number
increases the optimum angle and decreases the drag coefficient. However,
the differences about the optAmum are small.
:

!
!

For high-performauce conical-shock inlets without internal supersonic

compression, it is not possible to use a straight conical taper of nearoptimum angle from the lip leading edge because insufficient lip thickness
is ava'2.ablein which to enclose the req_ Lred du_t area snd turn the flow
back to the system center line. It is therefore _,ecessaryto camber the
lip to meet the deflected streamline "a_d have a curved external surface.
The calcu]:_tionsof Ferri (ref. 13) indicate that it is better to expand
and turn the flow in the immediate vicinity of the lip than _o distribute
the e_ansion along the length of tLe cowl. The position of the lip leading

/."..-.'"'i'_;: ,,,'"-_:

.,-_ ':

, .

._

'{

..

edge is of little importance in regard to external drag; but, as discussed

subsequently, it is of great importance in regard to net drag because to
avo/_ the =_ge force that can result from additive drag the lip should
Just inter:ect the oblique shock wave from the cone apex.
Additive drag.- As described in the section on definitions (p. 12),
additive drFg represents the momentum difference in the engine-flow streamtube between the inlet and the free stream when no aircraft components,
other than those of the air-induction system, _nterfere wlth the streamtube The simplest example of additive drag is that of an open-nose inlet
at reduced mass-flow ratio; the additive drag is the pressure integral along
the diverging streamtube between tbe _xternal normal shock .waveand the
stagnation pc.intou _he inlet lip. This drag component can be calculated
by the formula derived by Sibulkin (ref. 17S ) which is plotted in figure 13
for drag coefficient and mass-flow ratio based ca c_2ture area. Comparison
with experimental meas'_r_ents (see refs. 146 aa_ 173) eubstautiates the
reliability of these predictions. Since the table on _age _B shows that
a rough value for the ratio of _let_to-wlng area is 0.01, the additive
drag coefficient can, as an example, represent 0.0020 in airplane drag
coefficient at a mass-flow ratio of 0.8 and a Mach number of i._. This
force, particularly at lower mass-flow ratios and higher Mach numbers,
therefore, can be an appreciable part of airple_e drag, and, for efficient
flight at supersonic speeds, the operating mass-flow ratio must be near i.

:,_
"_

'_'

For a conical-shock inlet or one utilizing a wedge-type ramp, the

pressures on a diverging streamtube ahead of the inlet (see sketches (16b)
and (16c))are, of course, affected b_ the shape of the precompression
surface, and the problem of predicting additive drag is more complicated
,,
than for a _mple open-nose inlet. Sibulkin (ref. 173) has studied the
conical-shock inlet with supersonic inlet flow and presents the charts
shown in figure 14 for the additive drag coefficient sx_dmass-flow ratio
based on ca )ture area. The variation of cowT.-positionangle _ (see

_,

:'
<

sketch (18)

with mass-flow ratio is also sh_m.

The charts show that,

'_' _,
:' :,

:
,:
;_
'-'_
'

'i
[',

_;
_
'_
"
.,_

....

SXetch(18)

oti_.r factors be_g constant, the additiye drag coefficient in,Teases with
cone angle, and, contrary to the normal-shock nose inlet, the additive
drag coeficient decreases _rlthincreasing Mach number. For conicalshock inlets in which the flow at the inlet is not supersc_ic (sketch (18)),
Sibulkin in the same reference has studied the effec%s of the center body

,;!_
ii_
,_
"

sad of the assumed pressure r_ccvery.

The rest.its show that the additive
drag coefficient for these co_-_fi_dons at given values of cone angle, m_ssflow ratio, and Nach n_nber can be either _-eater or les_ than. thaG of
normal-shock
inlet, depending upon the location of the llp relative to
the conical shock wave.
If the lip is close to the ob3 i_ue shock wave
at maximum mass fT.",.: (_;7. = '_) as shown in sketch (18a), t_.e additive drag
coefficient
is high because tn_ deflecte_ streamtabe i_ _ubj_cLed to the
pressure behind a normal shock weve occurriug at stream Mach number.
How..
ever, if the lip is far behind the couical :_hock wave (sketch (!8b)), for
a reduced mass-flow ratio the pressure' on the streamtube is not as great
"_ .

as
in the fo_er
because Sibulkin
of the weaker
normal
wave. _or
In comparing
predictions
with case
experiment,
has found
goodshock
agreement
this
form of spillage.
Wyatt (ref. 12) has compared the additive dr_.g coefficlents
resulting from reduced flow of the three possible types ,__._shown
in sketch (19).
Thus, from the standpoint of drag, it is evident that air

I"
_
I
1

o..........
!
--!
.
1

Moss41ow
r_io,m_/m.
should
points

I,
t

not be spilled from belnd

a normal
Sketch
(19) shock wave, and, as Sibulkin
out, for flight M_ch numbers below the design value (_ = _), it ie

li

desirable to increase the center body projection

(trans_ating-cone-lnlet,
p. _0) to maintain supersonic flow at the inlet.
_r
a two-dimensional
inlet with a precompression
ramp the additive drag can be calculated from
momentum _lationships
as has been .do_' for conical-_hock
inlets.
Change in axternal wave drag.- .When mass-flow ratio is reduced below
the maximu_,_va/ue, the pressures on cowls change because the inclination
of the flo_ _with respe.ct to _he ._i_ leadf.ug edge ,_hanges. Because of the

_:'"
_I

_m--,-'

I
I

........

. -".wF"

. [J"

_j

greater inclination of the Iccal _treamlines, the co_! pressuve_ decrease3

thereby creating an increme.ntalsuction force that is in the _:hrustdirection. As shown, for insts_ce, by Fradenburgh and Wyatt (ref. 14), at
subsonic speeds this lip suction force counterbalances the additive drag
if the flow _emalns irrotational. However, st supersonic spec@_, the
_.udthis balance of fo_.ce_
prcscncc of shock _aves causes rotational _ ..,,_
....
cannot be accomplished. Several investigators have presented ans_yse_ of
the change in cowl pressure forces withdecreasing mass-flow rat!:,.
Fraenkel (ref. 174) has studied the problem as applied to norm_l-_hock
inlets using momentum methods, but experiment sho_s that the pr_.,a]cti_s
underestimate the cowl suction force at mass-flow ra_f_ ebove about 0.6
even though the co_:istested had sharp lips. (See refs. 3_ and 167.)
The an_l_sls of Graham (ref. 17_), which includes an allowance for lip
thickness, agrees with that of Fraenkel for mass-flow ratios greater than
0.8. Grlggs and Goldsmith (ref. 1_6) use the analysis of Mo_,ck,_l
(ref. ?.76)
to predict some portion of the llp suction force, but sinc_ the 'xholeco_l
is not considered, this method also underestlma%es measure_ _uctloa forc.=_.
Figure i_ presents a compilation of experimental date and a comparison _ith the prediction cf Fraenkel. (Drag coefficient is based on inlet area,
and the increment of mass-flow ratio ._(m/_,-)is 0.3 correspondlng to a
change in mass-flow ratio from 1.O to 0.7. It is assumed th'.'%
the ",'ariation of drag coefficient is essentially linear ove_ this range ) The data
of references 146 _ud 167 represent pressure-distrlbution measurements and,
for the more slender cowl (AM/A2 ::-_.0),the predicted decrease in ava_lable cowl suction force with flight M_ch number is fairly well substantiated. For the larger cowl angle (AM/A2 = 4.1)_ however, much more tota?.
suction force m_ recovered_ the pressure measurements show ti_t _he suction pressures are less in magnitude +.hart
those on the thinner lip b_t
they act on a greater frontal area. This increased suction force at low
mass-flow ratios is _t the expense of greater drag at a mass-flow ratio
of i. The re_ining data represent the results of force-test measurements,
and they sho_"considerable scatter, as _ould be expected since Lhe accuracy
in determi_ this relatively small force component is not so _ood _z with
pressure measurements. These results tend to substantiate the conclusion
that blunt lips can recover more suction force tha_ sharp lips.

_,
-_
_
_i
._
_'
i_

_
_'_

_,_

LiD bluntn,_,ss.
- Much of the previous discussion on drag at sunersonic
speeds has been :oncerned with thin, sharp lips on whleh shock waves _ould
be attached at m_ximum mass flow. Ho_.;e,,er,
since such lip shapes cause
large total-pressure losses at the high mass-flow ratios encotu.teredin
low-speed flight, the penalty in drag at supersonic speeds result."._
from bluntness must be la_ovn_n order to resolve the necesssi-ycomp. x_se.
As pointed out by Graham (ref. 17_), it Is to be expected that the maximum
cowl _uctlon force attainable is limite& by lip bluntness; that is, for
_ given atio of inlet-to-maximum-cowl area, above some degree of bluntness,
high pressures on _ne large frontal area at the leading edge more than
counterbalance the incremental suction force caused by expansion of the
flo_ over the rela::_velysmall frontal area between the lip aud the
maximum cowl diameter'.
-'

:_

__._
_
._

_.
:"

'

"

: :

NACA _M Ag_FI6

Frnenkel has studied the problem Of lip bluntness when

i8_me/mo)M=l.O
(ref. 167) by assu_
that the drag of the profile is that of an isolated
lip ply
a $m//
cob_ponent due t9 the expansion behind the llp acting on
the downstream profile.
These asstm_tions
teml to limit the an83ysis to
.....
;-- -- chart
w_s ob:_in_d.
Comparison of these results with other _xper_ents
produces
no reliable correlation.
__he ex_per_nts
of reference 23 show that with
an inlet of Ae/AM = O.l_
and a lip of (r/R) a = 1.17 there is no more net
dura6 thsn with a sharp lip at rosins-flow_atlos abo)-e 0.8 at supersonic
speeds.
At the hlgh-mas_41ow
ratios of low-speed flight, this lip causes
about half as much loss in total-pressure
ratio as does a sharp lip
(fig..5).
_e
tests also show that the net drag changes little to angles
of attack _, 9 .
From the _isc_msslon of lip shape_ In regard to pressure recovel-j and
drag, it appears thzt a reasonable lip profile for supersonic aircraft
(flight to a Math number of 2.0) is elliptical on the internal surface
with (r/R) 2; 1.15 and a/b =-3.6 (see fig. 9) to provide acceptable pressure rec_ery
is low-speed flight.
_'-e profile is straight on the external
_urface with the amgle be_ean
the surface and the approaching flow d_rection about 3 for the leas_ wave _rag in supersonic flight.
For inlet
areas of 2 to _ square feet, the thickness behand the leading edge of such
a lip %x)uld be fr_m 1 to l-l/2 Inches.
Net wave drag!9.- The previous discussion of drag has been largely concerned with relatively idealized configuratlons.
For air-induction
systems
which are complicated by t_e necessity of many design compromises, accurate
predictions of net drag can be made only for quite restricted conmltions.

ZSBecause of the contraction between the lip leading edge and s_ati_
2', it would be expected from one-dimensional
consideration_
that (ma(/mO)M
would be greater than 1. The experimental evidence Of Fraenkel f pz relatively blunt lips indicates that compression due to contraction
is hardly
realized sm_ the maximum mass-flow ratio is very nearly 1. Mosmm_u s_d
Anderson (ref. 23) _ound that for less blunt lips nearly the full effect
of the ce_traction is attained.
This result is confirmed by recent work
of _i
and O_en (NACA RM LS_CI6).
_SThe experimental determln_tiom
of net wave drag by means of direct
force measurements and total-pressure
survcjs is a dlfficul+ procedure
because sevea-al very accurate measurements
must be made to obtain reliable
val_s.
It is possible to determine this force in supersonic flow frc_
schlieren or shadowEraph photographs
by calculation of the entropy rise
or moment-am change '_hrough the external b _w shock wave.
However, acc_rate
evaluations by this method slso require coasiderable
care_
Descriptions
and studies Of the method s.qe presented in refer__ences 178 thro___181.

NACA _A95FI6

C_!

_. _

67

For instance, as shown in sketch (20): the rises in net drag with decreasing
mass-flow ratio for the vertical-_dge inlet of reference 182 and the
inlet with a flow deflector of reference 26 are considerably different.

"

L6

"

'

"?' I

----Tl_xet_.ol

.:
"_

"_

----TI_It_

|2

_,

!
(open nose) Rat 173

(W'zbdescow;
IkCtl0nmlflmote)Ref.174

deflector

<3

(model

----<)----- e"Ver)icel
_

B) Ref. 26

sco_,_f. 162

:
2-

"

_"

.2

:_

.6
Mo_-f_ow

.8

ro:'o,

1.0

i.2

-'

m_lm m

Sketch (2O)
,,

These inlets are similar in that both had a wedge-type precampression

surface; the flow+deflection ax_le for the vertical-_dge inlet was 8
and that of the flow-deflector inlet was 6.9. However, the inlets were
otherwise entirely dlffer___5. At mass-flow ratios above 0.7, the drag rise
of the two differ by a factor of about 2. The estimations of Sibulkin
(ref. 173) and of Fraenkel (ref. 17_), which take no acc_t
of the precompression surfaces or of skin friction, apparently predict the dra_ of '
the flow-deflector inlet very well. However, acc_mt must be taken of
the preccmpressio_ surface to predict the drag af the vertical-wedge inlet.
Obviously, the theories cannot be relied upon to predict the drag at low

i
_.

._

+.t._
_
.._
_,

mass-flow ratios of such distorted inlet _hapes. However_ in normal

Operation, supersonic aircraft must avoid low mass-flow ratios because of
the large additive drag force (or, at least_ air should not be spilled
from behind _ normal shock wave). Fo_ mass-flow ratios of about 0.9
and greater"the incremental drag due to a reduced mass flov is not a large
force, and the significauce of the error in estimating it is correspond_n&ly
reduced. Thez_+'ore,the fo]lowing simple formula of }Taenkel (ref. 174)
for the net wave drag of open-nose bodies at zero angle of attack is
i--

--

it--

ii i

_
i_
_

_?

ii[i

,_
L

I .

_ LI ,+ .l'lll

I_

I
'

'I I
+

+ "

.....
_,"3,,

"
"

'_

:' '-:"+ ' " :'_/+;':

..............
;t['_."

"r'+ll '"

:_,_

'

',

'+",+ +'+9+_'

'

,'

",

',_l

possibly useful for estimating the drag of slender complicated configurations at high mass=flow ratlos.

01_

= CD,,.,,.
---_ +

ttoAM

p,-_o

= C'De_,, + .,_
_
,,,

m[

(27;

and

(Here Pz is the static pressure behind a normal shock wave. ) Thus,

according to this estimation, net wave drag is the sum of the externel
wave drag of the cowl with no spillsge _u_' the product of the relative
static pressure behind a normal shock _ave (Pl - Po) and the annular f_'ontal area of the diverging streamtube (A2 = Ao). _e expression is a linear
function of mass-flow ratio. Since there is little dCfference in the
slopes of curves of additive and net wave drag coefficients with massflow ratio at mass-flow ratios above 0.8 according to Fraenkel, cowl
suctio_ force is of no consequence in this range for sl_nder cowls. However, _ i_-a_icated
in figure 15, a sizable portion of the additive drag
can be co_teracted with blunt cowls and, if the high drag of these cowls
_ith no spillage is acceptable, cowl suction force should, in this case,
be taken into account.

FLOW STEADINESS

In the operation of air-induction systems, unsteady flows limit

propulsion-system performance for several reasons = duct rumble, that is,
noise and vibration from the system which disturb the pilot, fluctuations
"#nichcause structural fatigue, or fluctuations which affect engine
operation In the follo'rlngsection, flow steadiness is discussed as a
basic property of air-inductlon systems as _as pressure recovery, flow
uniformity) and drag previously. In this discussion, however, some conSlderation is given to interference from other aircraft components because
unsteadiness in the engine flow often arises on account of the boundary
layer from other surfaces.

Subsonic Flight

_m..toked flow.- In low-speed flight with a fixed-area _nlet designed

for hi_-__'_light
at altitude, the mass-flow ratio can be large

,J

#jr

_&

_'._ "LJ ',.

.

.
r::

::-

_,
: _'!:T:.

:-:_:_::

, _ , .'- . ,_:,_' .'

_,-_.L,._:,_._--_:.,

,'_ ,,,,
,,
_;L,_; _::::

,' .,7'.
:: -:T-'_ _

.
_

,r'j ,'_ ,_
_

,.

" "
'_

'

"," ,,.;J_olF-"w-_
, .......
' _
_

'

'

)_

NACA _

AS_F]6

,_ONF_D_TIA_' : '_' _ i'_ '_

69

enough to choke the inlet. Asld_ from the low total-pressure ratio and
nonuniformity associated with thl_ condlt_ou, it must be avoided because
of flow unsteadiness. The results of BiLckaby and Watson (ref. 72) show
that at zero forward speed with a sharp.-iil"
inlet, fluctuations as large
as 8 perceut of the ambient pressure occur at frequencies up to _bout
200 cycles per second at mass-flow ratios ma/ma* above about 0.6. Such
unsteadiness was reduced both by increasing either the flight Mach number
or the radius of the inlet lip. The results of Mil_llo (ref. 73) in tests
at zero forward speed indicate large nonuniformlty in _he diffused flow,
differences in local total-pressure ratio of as much as O.lO, for inlets
with rounded lips just prior to choking. Thus, both flow unsteadiness
and nonuniformlty are to be expected in opezation near choked conditions.

Duct rumble.- Several aircraft in flight at high subsonic speeds

have encountered duct rumble. So far _-_is known, operation has been
affected only by the nolce and vibration which are sufficient to disturb
the pilot so that the conditions under which they occur are :onsciously
avoided. The phenomenon has been reported only with air-induction systems
having side inlets and is apparently t_e result af interfel_nce with the
approaching boundary layer. The tests of Mathews (r_f. 183) on an underslung scoop for the cool.!ngair of the engine of a propeller-driven ai_-plane indicate that duct rumble w_s due to flow separation ahead of the
scoop. The separstion was apparently caused by external c_pression
resulting from a low inlet-velocity ratio. The rumble _;:seliminated by
increasing the inlet-velocity ratio through _ reduction of the inlet area
and by relieving the flow through the boundary-layer gutter by increasing
its depth. An air bypass which increased the inlet-velocity ratio was 81so
a successful means of avoiding the rumble. Similsrly, reference 184
reports duct rumble at inlet-velocity ratios less than 0.4 at flight Mach
numbers from 0.6_ to 0.92. _in-duct instability is su@gested as the
cause of the rumble; upstream separation at the low Inlet-velocity ratios
was probably the cause of the unsteady nature of the instability. Other
instances of duct rumble have been encountered, but descriptions of them
have not been published

Since available evidence indicates that duct rumble is generally

caused by boundary-layer interference, it can be avoided by removing the
boundary layer from the influence of the compression field or by reducing
the compression field through an increase in massoflow ratio. (Methods
of boundary-layer removal are discussed !_ter under INTER_,%INCE.) Duet
rumble is to be expected when the static-pressure gradient i_ the external
compression field is sufficient to separate a t_rbulent boundary layer.
In two-dimensional subsonic flow a rough design criterion reg_uxlingturbulent separation is that it csn occur in positive pressure gradients
when the local velocity is les._than two-thirds of the initial velocity
However, larger pressure rises have been observed with air-induction_ystem installations possibly because the flow was three-dimensional or
because the gradient was small. 9___e
_,uundary-layer surveys i_s,
ediately

'
:;_:
_
_
_
_!_

ahead of the inlets described in references 18_ sad 186 show that without

_i;_

/
.........................

.._

ace

ee

I@.CA]_4A99FI6

eo

_L

boundary-layer removal an approaching boundary layer thickens rapidly

and separates at inlet-veloclty ratios less than about 0.6. With some
bouudary-layer removal this rapid thickening occurs at _mlet-velocity
ratios less tha_ about 0.4. These figures can be used as rough indications
of ".Senduct _
"--amblc
might be expected.
Twln-duct instability.-Martin and Holzhaustr (ref. 187) have studied
the stability problem of the flow through ducts from symnetrical twin
intakes emptying at a juncture into a cc_mon chamber as shown in sketch (QI).
From the assumption that the static

_._
"_

b__m_._

._-\

ture (which is here called station 3)

pr._ssureJust
downstream
of the
juncis
uniform across
the common
duct,
it
of recovered
is
demonstrated
static
thatpressur_sfor
a variation
as shown
in the sketch the flow is unstable at

\
_

S
a
_--k_.._---J

--

inlet-velocity ratios oz"the syste._

less than %hat for maximum st,_t[cpressure recovery. That is. if the
two ducts laitiaS_lyoperate at the
joint inlet-velocity ratio corresponding to point s, a small disturbance
which causes an increase in inletvelocity ratio in one duct causes the
flow in that duct to increase to point

Vz/Vo
8ks%on (21)

a and that in the other duct to

decrease to point b. From the contlnuity relationship in inco_ressible
flow, it is evident that

(V
(v2/Vo)s
=

o)a
* (V

o)b

(29)

Thus, as a res_< of the continuity requirement snd the assumption of uniform static pressure at station 3, it is apparent from simple geometry
that operation below the inlet-velocity ratio for maximum recovery is
possible either at s or at a and b. However, if s is above the
maximum, operation is possible orly at the Joint inlet-velocity ratio.
For these events to occur it is necessary that the shape of the curve be
similar to that of the sketch; that is, the negative slope at ulgh inle%velocity ratios must be grea er in absolute magnitude +ban the l_-sitive
slope at low Inlet-velocity ratios. The assumption of m,ifcrm static
pressure has been found from experiments to be realistic, "Jndthe shale
of the curve has also been found to be typical of those of twin-scoops
into which boundary layer flows. If two nose inlets or scoops with complete boundary-layer removal were used, the slope of the curve would not
reverse; it would decrease from an inlet-velocity ratio of zero. Unstable
flow could then not occur. From _he sketch it c_ be seen that if the

.......................

__ _

_._

"

' "

,, ,,:"i,T--"'-h'-

........

'

J' i

joint inlet-veloclty ratio is suffiniently Emall, the point b would be

at an inlet-veloclty ratio of zero. A disturbance _n duct a that then
reduced the static pressure at _ ,_o_Id caus,_a reversal of the flow
through duct a - a phenomenon that has been obse_ed.
Since the statlc-pressure-recovery c1,_e does not have a sharp peak
in actual flow, unsteadiness can be expect_d if the point s is in the
region of zero slope because disturbances im either duct could cause c_e
and then the other to operate at the high and then the low Inlet-velocltyratio conditions. The magnitudes of the disturbances and the slopes
determine how close to the peak s would _ve to be for such unsteadiness
to occur. If s were below som_ limit, _e operation would be stable
at a and b.
Since all the conditions which lead to twln-duct instability and
unsteadiness in subsonic fllght can exist at supersonic speeds_ these
diffi_-tLlties
can also occur as demoastrated in reference 188, and s)mtams
should be designed to avoid them. A method of reducing twin-duct interaction in an air-lnductlon system far supersonic aircraft is reported 4_n
reference 130. _le wall between two ducts upstresm of the Junction was
perforated to equalize _he static pressure and enable crossflow to provide ,_isccusdamping.
i
Supe'.'s
onic Flight

Causes of unstesdiness.- Unsteady flow in air-induction s_tems

occurs more readily in supersonic than in subsonic flight essentially
because larger positive pressure gradients are encountel-_dwhich separate
the flow. Unsteadiness occurs either at subcritical _Jass-flowratios or
at the very low total-pressure raClos of operation fa_ in the suporcritical
regime. The design problem is to malut_In steady flow through a range
of mass-flow ratios sufficient to satisfy all engine operating conditions.

":

Unsteadiness has been observed to occur in a variety of situations

some of which are illustrated in sketch (22).

The flz-s_two exam_l._ are

_
,,t

(0)
,.
-- .....

(b)
Lineof velocitydiscontinuity
Seporofedflow

()

"

_:
_>

Sketch (22)

'*

._
:

"'

!
/
I

NACA l_,I A_SFI6

.......

those described _ Ferri &nd Nucci in reference 50. Here, the velocity
discontinuity downstream of the intersection of an oblique shock wave and
the terminal normal shock wave entezs the inlet as a result of the normal
shock wave moving forward due to a reduction in mass-flow ratio. Since
the total pressure and velocity are less in the streamtube on the outside
of the llne of dizcontinuity, subsonic compressioa tends to bring this air
to rest sooner than it does the high-_elocity streamtube next to the
center body. When the local _ch number behind the oblique bow shock
wave is near 1.O, as it should be to avoid significant shock-wave boundarylayer interaction, the velocity difference across the discontinuity is
large, and the velocity of the outside streamtube approaches zero in the
duct while that of the inside streamtube is still high. Unsteady flow
results when the llne of discont_uuity Just crosses the lip because a
large percentage growth in streamtube area of the low velocity stream
occurs while a uniform static pressure is maintained across the discontinuity. Even though the contraction of the high-velocity stream is small,
it is sufficient to choke the major portion of the flow because of the
high local velocity, and air must be spilled. Once this happens, the
pressure recovery decreases, which tends to draw the flow back to its
original position, choking again occurs, and the cycle repeats. This
explanation is obviously oversimplified because the effects of viscosity
are ignored; neither turbulent mixir_Eacross the llne of discontinuity
nor the presence of a boundary layer is considered. The experiments
which were reported with this explanation show that an entry section which
is sufficiently long to permit mixing to reduce the ve_'ocitydiscontinuity
provides an increased range of steady subcritical mass-flow ratios. "When
separation occurred on the central body as shown in sketch (22b) in these
tests, it was found that unsteadiness occurred as the mass-flow ratio was
reduced when the velocity discontinuity from behind the lambda shock
approached the lip from the inside. When separation was prevented by
boundary-layer removal, unsteadiness resulted only from the prior explanation. It was concluded fromthls study that unsteadiness can be ave: led
by positioning the external compression surface so that the _ine of velocity discontinuity cannot move across the lip for the range of flight conditions of interest so long as extensive separation on the compression
surface is also avoided.
The results of references 91 and 189 show the importance of separation, as illustrated in sketch (22c),as a source of unstesdlne_s and
indicate that factors other than lines of velocity discontinuity must be
considered. It is shown in reference 51 that a conical-shock diffuser
with a 25 semlcone an._leand a 6 equivalent conical subocnic diffuser
has a very small range of steady subcritical flo'_ e_en though the relation
of the lip to the oblique bcw shock wave is changed. The same inlet,
however, with _ length o" duct-entry section of 3._ hydraulic diameters
alway_ had a much wider steady range. Since there was separation on the
cone surface throughout the subscritical mass-flow range in these tests,
it is apparent that this and the duct shape can be dominant causes of
unsteadiness. When the duct did not have an entry length of small pressure

....

. .

', '.' , ,.... % , ;., _._,,_:'-._C, "'-

I"
I

.->%.;.

NACA _

CC_IDF_-T_;_LL _ _....

A_SFI6

73

gradient sufficient to pe1_nitthe boundary layer to reat_ach and recover

a profile that could withstand subsequent compression (H < 1.8), unsteadiness resulted. This conclusion is substantiated by the results of
references ll3 ard 190 in which unsteadiness was eliminated by forcing
a separated boundary layer to reattach by suction. D1so, the results of
reference 128 show that relatively small irreguls_ities in area distribution in the entry section of a duct in which the pressure gradient is
positive can have serious consequences in reducing _he range of steady
flow.e Additional data, on the flow unsteadiness in one scoop-type airinduction system, are reported in reference 191.

Character of unsteadiness.- The wind-t_m_el tests of reference 192

for an air-induction system without an engine showed flow unsteadiness
after diffusion with a frequency,of about 20 cycles per second and amplitudes as great as 30 percent of the local static pressure. The quantities
are, of course, dependent upon the particular desi_, and also upon engine
operating conditions. Reference 193, for instance, shows that for a
ram-jet engine the effects of approaching flow unsteadiness are attenuated
by an increase in the pressure drop across the flame holder and that sn
increase in engine total-temperature ratio can amplify the pressure fluctuations. With a turbojet engine controlllng the flow through a conicalshock inlet, Nettles and Leissler, reference 31, found that the engtue
steadied the flow through the inlet. Both the range of steady operation
snd the intensity of fluctuations were less with the engine operating than
with the flo',controlled by a choked exit plug. In fact, in the latter
case the fluctuations built up to a violent level in certain ranges of
unsteadiness; whereas with the engine controlling the flow, the inlet.
could be operated through the same range of mass-flow ratios without difficulty. Since, in general, flow unsteadiness from the air-induction
system causes reduced performance with the degree of permissible unsteadiness dependent upon the refinement of the engine, the requirement in
air-induction-system design is to provide steady flow to engines over the
needed range of flow conditions. Thus, the detailed nature of flow
unsteadiness is of interest only inscPar as it shows when serious unsteadiness is to be expected or what parameters are effective iu alleviating
alverse effects.

__

Several investigations of unsteady internal flows have been reported.

(See refs. 38, 194, 19_, and 196.) ,_T,._a_
theoretical and experimental study
of Trimpi, which analyzes the problem by considering traveling plane waves,
indicates that the frequency of the flow oscillation decreases as the duct
length increases. The frequency is also affected by mass-flow ratio,
increasing somewhat with decreasins mass ?iov. Probably the most important

_In the tests reported in referel._e123, the models used had small
irregularities in area distribution near the duct entry, but the razAge
of steady mass-flow ratios was large. The cause of this difference wa_
that in this latter case the pressure gradient through the duct entry
was slightly negative or zero.

....

_ ....

me

_J
%
%

nT

/
l

IIit
i

II

e _1

I 11

conclusions are those related to the orlgdn of the unsteadiness.

It was
found that the relation between the time rates of change of entering mass
flow, of boundary-layer
growth at the inlet s_atlon, and of the instantaneous vs_iue of entropy averaged across the inlet was the critical factor
causing unsteadiness.
Further, it w_s shown that, s ltho_h
_.aves caused
by changes in engine ':hrust can move the shock pattern to a position at
which unst_edlness might arlse_ the disturbance
which initiates unsteadiness originates near the entrance and need not be sufficient to choke the
flow.
The experiments of references 194 and 199 indicate that th_ ._a_lit_e
of uusteadiness
_s caused by a llne
of velocity
disccutinuity
crossing a lip
(sketch
(22a))
is less
than that
caused by separation
of centerbody boundary
layer
(sketch
(22e)),
Since _umerous
_nlet
configurations
were investigated
in references
19_ _ud 199,
it
is _ossible
that
this
result could have some gener_!It_.
Prevention of unsteadiness.The obvious method of avoiding flo_
unsteadiness
is to operate a propulsive system only at nmss-flow ratios
near or _!ightly above the critical with an inlet designed so that a line
of velocity
discontinuity
does nob cross the lip and so that serious
boundary-layer
shock-wave interactlcm
is avoided.
The fact that this can
be _ccompllshed with a fi::ed-area inlet for a relatively wide range of
Mach number variations has been demonstrated
in reference 90. However,
for operation through a wide range of Mach numbers, altitudes, and po_er
settings, one of the variable systems described previously would be
required to maintain nearly a constant mass-flow ratio.
Since this remedy
is accompanied by the addition of weight and complication,
other methods
of avoiding unsteadiness
can be more desirable.
From the 6iscussion of
the causes of flow unsteadiness,
it is apparent that the difficulty can
be delayed by reducing severe velocity discontinuities
and adverse prossure gradients in the entering flow.
However, if these must occur, the
effects can be minimized by giving the flow an opportunity
to re-establlsh
a more _miform high-energy profile that can withstand additional compression.
As shown by references 91, lll, 190, and 197, this can be accomplished by removing uoundary-layer
air or by providing sufficient distance
for turbulent mixing to re-energize
the flow.
The latter method has been
investigated by providing a long entry section of very nearly constant
cross-section
area.
The increase in the range of steady subcritical massflow ratios that can be accomplished by this method is shown in sketch (23)
which is reproduced from the data of reference Ol. For the models tested,
the flow was steady through the mass-flow range at a Mach number of 1.9.
Ho_-_v_r, there was an appreciable loss in maximum pressure recovery at
this Mach number as entry length was increased because of the high local
Math namber at the inlet and the associate_ increase in friction losses.

'

The previous discussion of steadiness has been concerned only with

conditions at zero angle of attack.
It is, of course, necessary to maintain steady flow for satisfactory
engine operation during maneuvers.
In
the tests of conical-shock
inlets of reference 90_ the steady range of
mass-flow ratios was small at zero angle of attack, and it was slightly

------.
r

i,

.., .... ,,.....

_-_ ,,

...........

- ....

....
" ........

._
"

'

"_

: - ......

_._.

:o......

._

NACA EM AS_F16

C_

:....

i.o

LO|__._

"- \x',

[ ""
,
'

79

I
2
3
Entry length,x/Oh

1
--a

Wetted porimeterof annulus

I
2
3
EnlTylength,x/db

Sketch (23)
greater at angles of attack up to 9 . A similar result was found in the
tests of reference 91 fo_"conlcal-shock inlets which had small steady
ranges at zero angle. Howe_er, when a long entry passage was a;Ldedto
provide a wide r_nge of steady operation at zero angle of attack, there
was an abrupt reduction in the steady range at angles of attack from 3
to _o. At higher angles thel-ewas little difference between _he inlets
with the long and short entry sections. A tilting cone on e conicalshock inlet to provide improved steadiness at large angles of atte:ckis
reported in reference 129. A_ an angle of attack of lO, "_iththe cone
at 0 angle of attack, steady flow was m_intained to a mass-flow ratio of
0.4; with the cone and cowl at lO angle of attack, the rdmimum steady
msss-flow ratio was 0.9. In reference 198 tests of conical-shock inlets
wlth booms protruding from the center bodies ere described. An increase
in angle of attack to lO reduced the range o_'steady uass-flow ratio_
by 29 percent. Interaction between ahock waves and the boundary layer
on the booms was the cause of this large decrease.

:
i
-

Other investigations have demolmtrated meth_udsof improving flow

steeliness to some extent. References 197 and 199 show small incr,_ses
in the steady mass-flow ratio range (0.06 in ref. L97) as a result of the
internsl contraction with a blunt lip. Reference_ 197 and 200 show that
remov_l of the boundary layer frc_ the center bo_;f of a conical-shock
inlet reduces unstaadiness, with the greater eff_ctiv@-nessoccuATing when
removal is upstream of the terminal normal shock wave. In fact, at an
angle of attack _f 0O an improvement of 0.16 i_ the range of steady massflow ratio was _ttalned (ref. 197), bu_ it dee,'easedwlth increasing angle
of attack. Altho,_h these and most of the previous referen-.esare concerned with conical-shock 4.1p*.?,,
t_

_ri__/les of design for providing

:_

_:
....

l
[

_,

'

If

_,

It
I

lel

It

ge

&

steady flov are the same for other t_rges. (See, e.g., refs. 188, 190,
19i, 199 through e0e. )

I_2E_[_NOE

The purpose of this section J_ to discuss the aerodynamic factors

other than those of the induction system its._!which affezt design; it
is entitled "INTEEFEBENCE" because the change,_in the forces due to combin_ng an air-induction system and other aircraft cum_onents are considered.
The _ectlon is divided into two principal parts:
1. The iaterference of aircraft flow fields wlth those of inductlon systems - the inluced effectG of body shape, angle of attack,
and the viscous effects of forebody bouudary layer.
2. The interference of'a_-induction-system flow fields with
other aircraft components - the effects of induction systems on
aircraft drag, llft, and pitching moment.
The type of factols involved are illustrated in sketch (24).

Here, the

Sketch(2_)
th_ performance of an undc_ w_ng nacelle is affe:c_,':_,
1.

Bow shock wave of th_ f_,_lage

2.

Velocity increment at inl_,tdue to fu _.Is;_e

pressure field

3.

Shock wave fro_ wing leading edge

. ._

' r.,,;

NACA _

'_;

-_
_

A55F16

CONT_DEI_.V_L
, _,

77

4.

Velocity increment at inlet due to _rlngpressure field

_.

Uniformity of the _low velocity at the inlet

"

__._.e
perfo._naneeof the other aircraft components is affected by
1. Interference of pressure field of engine streamtube "_iththe
wing and fuselage boundary layers and presgure fields

,!

2. Interference of pressure field of engine fairing with the wing

and fuselage
b(mnd_"y layers
and pressure
fields
Obviously, the problems of inte_erence

!_
o_
_
!_

._,.
_'

are complicated, end qusntit_tive

evaluation requires experimental studies of specific configtu-ations.

However, an induction system that must be placed in the flow field of
another object can either benefit or suffer fr_n the resulting interference,
and careful consideration must be given to the conditions of shape and
position in order to produce favorable effects. (See, e.g., ref. 203.)
AIRCRAFT-INDUCTION SY_
ST_4

Effects of Inlet Location

Subsonic flight.- From the
standpoint
recovery at
the
inlet, of
thepressure
best ]ongitudinal
nat_on region near the nose of a body

I'O0_--]----T
._
_
_ .96
_

Fuse._
in_I

I
m "0"_
I^_ i
I'_o
I _<2_4_
I

2.00

.
i
_,
,
:_.
_

position of an inlet is in the stag94

and
any external
compression result_ .92_
because
the localMachnumber
is low
--i
'59._
Ing from a mass-flow ratio less than
927 .3 .4
7
5
.6
.8 9
1.0
I is essentially Isentropic
As an
inlet is moved aft along the" body,
Free-s_a_lach number,
Mo
the amount of boundary layer flowing
through it increases with a resulting _ 1.3[---T_-_--'-_--]-'-1

This direct effect of low-energy

'_,
_.
:,:
_

large in .:ubsonicflight, but second_

i.I
boundary-layer
air is normally
not
_-_--_
reduction
in flow
total-pressure
ratio.
I"
r
____t
ary effects,
nonuniformlty
and

._,
..
'_
"

unsteadiness, can be very important

at mass-flo_,ratios of the order ,of
ratio c_ mo_,i_ an NACA submerged
inlet operating at a mass-flow ratio

"_
"_
7
:,_
_:

1.2---

I,O_---_r-_-_&_

:i]

i_

wlng-fuselage combination Is shown in

_ketch
(2_)along
together
th__1_l
..__
of
0.8 aft
the with
ftu,
elage
of a

"o ,o

,o

--.
_._

Sketch_.(25)
'_

i_
_

_'_
'r_

'7

!
/

78

"

......
"_

'

CZ!

NACA l_
:

Ag_F16

_sh number dis_rigution

along the fuselage.
These results
were taken
from the d_t._ of references
20_ and 205. At flight
y_n n,_be_
less
than 0.3, there is essentially no effect of _oving the inlet _'t. The
_-eater boundary-layer thickness at the rearward staLions becomes ImporTant at a Mach number of about 0..%,and at .'_chnumbers above about 0.7,
it becomes of _reat importance _t the mos% re,___
station Here, the
to_al-pressure ratio decreases rapidly at high subsonic Mach numbers
because of both the high local Mach number illustrated in sketch (2p-b)
and the thick boundary layer. The .mostrearward location is lu the pressure field of the wing, and at a flight _t_ch ntm_er of 0.9, th_ local Mach
n_er
at the iDlet is supersonic (M = 1.22). Thus, pressure flelds with
l_rge induced velocities should be avoided.

"

!.
!*
i

A method for estimating the velocities in two-dlmensional comblned

subsonic velocity fields is discussed in reference 206. Superpos_tion is
a_su_d to be valid and the resulting relationship is

+_where
flow.

).+...

(30)

_Vloca! denotes the induced velocity increment in incampressible

This method _f predicting maximum induced velocity has been corn-

i,

pared -_ith experiment for a wing-nacelle combination in reference 207.

Here, the method predicted maximum velocity ratios about 3 percent less
than those measured. To predict tu effects of compressibility, the
Frandtl-Glauert rule can be used Ica"two-dimensional flow.

_,-V/-o
)o_,_essiblo_l- .o_ -- Vo,'_c_re_s_ble
and in the three-dimensional case, the methods of Her:lot (zef. 208)
should be ,._ed.
In terms of pressure
coefficient
in three-dimensional
flow_

CPeumT-.essibl
e

CPimcmpressible

=i+

In(l - Mm)

_t)

2 + O.6138

where t/Z is one-half the body fineness ratio. Herriot points out that
in.Junctures_ such as those between a wing and nacelle, the flow is more

.
,,

- -

[
i

"'

NACA

A9_16

CO_[D_

79

nearly
Olauert

than
rule is a better approximation
three-dimensional,for this and
case.
thus

the

Prandtl-

t
over a
typical body
at a subsonic
and shows
a supersonic
Mach number.
If in
Supersonic
flight.Sketch (26)
a comparison
of flow properties
the subso/_ic case the boundary layer is neglected, the total-pressure

i
h

'J

iii

,_

(@) Free st_

kEochnumli_r,

Me* 0.70

"

'
J

tO

":

(b) Free Slll14m M_H:h .wmb4_', Me* 1.7_

Sketch

._

(26)

area and
distance
position
however,
ratio of

the ].ocal dynamic pressure change little downstream frc_ a short

behind the nose.
Thus, from these standpoints, longitudinal
of an inlet makes little diff-_rence.
In the supersonic case,
there
is
an
initial
loss
in
total-pressure
bow
any streamtube about the body is i, and the ratio
mass due
flow to
pertheunit

"

shock wave, in this case i percent, and there are subsequez,_ changes in
local flow properties which have important consequences
in regard to air-

._

induction-systems
performance,
As an example, consider the flow conditions
at x/1 = 0.0_ and at x/_ = 0.9 wnere the local Mach numbers are 1.38
and 1.79, respectively.
If no significant radial
change in Mach n_mber

_
._i'

through
an engine
streamtube
is assumed,
a normal
wave occurringratio
at
the
forward
location
would create
a _-percent
loss shock
in total-pressure
and the loss through the optim_h oblique-normal-shock-wave
combination

"

_t

point
field

..........
'_.-_

_!

wo_Id be i percent (see _ig. 6).

However, at the rearward staticm, the
normal-_.hock loss would be 17 percent and the two-shock loss wo_/id be
percent.
If there were no b_ly, that is, if the supersonic ccmpressiea
occ_n_red at the free-stream Math number: the normal-shock
lo_s would be
l_ percent _um the two-shock lo_s, 4 percent.
Similarly, from the stand-

of flow rate per unit area, or inlet size, location in a compression

is advantageous.
From the st_mdpoint of drag per unit area, a

'

_ _

'

'

_--_--

":

"_

:........
'..... "_,'_':--J"

"

l" 1

'

80

-"

NACA_4 A99FI6

compressiGm flel.d is detrimeutal because of the high dynamic pressure.

However, for the condltions illustrated
ia sketch
(_2), the greater flow
rate is the dominant factor_ and %,he focused position of the inlet can
be shown to have 7-percent-less
external _e
drag than the rearward
position due to its smaller size.
_hus, location can ha%'e important
effects on net propulsive force, and ".[; c_ be beneficial to place an
inlet in %_e compression field of other a_rcrsft components.
In regard to the effects of the r_is_L velocity field _to which an
__nlet is placed, Hasel in reference 209. has investigated
the problem
experimentsD_ly at a M_ch number of 2.0. Half-conical-shock
inlets were
tested o_ a flat plate and on bodies of revolution havi_g forebody fineness ratios of 4.0, 6._, and 7.5; the botal-press_re
ratio of an inlet on
the bodies was always less than that of the inlet on a flat plate.
_%en
all of the forebody bou/idary layer was removed, the maximum total-pressur-e
ratio attained with an inlet on a body of fineness ratio 4 was 0.08 less
than that with the _klet on a flat plate; this difference was 0.04 with
th flnemess ratio 7r9 forebody.
About half ef these differences
could
be attribut_-_dto the bow-shock waves and the local Mach numbers at the
inlet stations; the remainder was thought to be due to the differences
In the radial velocity field.
Thus, appreciable losses are to be expected
from this cause with forebodies of low fineness ratio.
Since the local Mach number at 8n inlet determines the magnitude of
the Pressure losses through the shock waves used for supersonic compression, the forebody shape should be selected to minimize this M _ch number
without, of c_)urse, creating any additions/ drag.
Considerations
which
are important are indicated in sketch (27).
(See refs. 172 and 21(: )
2.1
2.1

'.

ZO

'---

=
/ i/
J
1.7 I{/

|"

.....

- 1.7

.l

-"
i.,2.0

M,.2_

%,

t#d._o

i#d.li.O
i

(a)
i.5

-.-I

20

Jl

Jt

(hi
_0

tSO

x/l,

(a)

.2

6
x/l

Jllill! zil.......

(b)

.8

tO

41

NACA _4 A95F16

_CONFIDF_

.......

81

For forebodies
of low fneness ratio, a considerable redactlo_ in local
Mach munber can be achieved by using conical, or m!nlmum-drag
shapes
ra_he'_"than sn ogive if the inlet r.ust be locate_ upstre.am of x/_ = I0
For fo::'ebodles of high fineness ratio, the differences
are sma!ie'_. _"Tne
data of cefe_ence 172 show that for a fineness ratio of S.O, the K_m_n

forebo_y dx.ag than the cone and ogive at zero angle of attack at supersonic
Mach numbers up to 2.0.
However, these minlmum-@rag
nose shapes have
blunt tips, and, depending upon the size of the engiue stresmtube, the
loss in total presmme
through the locally int_.se bow shock wave counteracts the drag difference.
.Reference 211, for instance, reports that a
rel_tlvely
small amount of _0i# bluntness that had a negligible effect on
mirC1mum drag caused 1-percent losses in tot_3-)?ressure ratio and maximum
mass-flow ratio as compared to a pointed tiL_. Thus, any specific design
requires s_I/dy and evaluation of these factors.
Because an air inlet at
positions other than the nose intercepts b-l_ a sma31 part of the air co_pressed by the bodD,, the major consideration
in choice of body shade is
drag.
The desi_
problem is to find the optimum inlet location on a lowdrag body.

Tests of very blunt noses, in whi_. the nose-radius to body-radlus

ratio was near 1.O, are repoi _ed in references 211 and 212.. It was found
tb_t a 4-percent loss in totai_pressu_e
ratio was suffered at a flight
Mach number of 1.4 and a 6-percent loss at a Mach number of 1.7 due to
nose bluntness and to the large radia3, velocity gradients.
The minimumdrag coefficients,
as compared to those of bodies with more slender shapes,
were more than doubled
Because of the reduced total pressure and the
overexpansion
of the flow behind the juncture of the hemisphere sad the
subsequent body, there were also com.siderable losses in maximum mass-flow
ratio in both inve_tigations.

_'
-.

In the general case, forebodles are not axia]3.y symmetric as has been
assumed in this disct_slon,
the theoretical
study of reference 213 indicates that small reductions la dr_g can be produced by axial asymmetry,
and a similar conclusion has been reached as a result of the tests reported
in reference 21_.
It is possible that circumferential pressure gradients
and reduced local Mach numbers can be prod,uteri b._iasys_etric bodies that
are beneficial
to air-induction-.systam performance.
To date, no studies
of this kind have been made.

Induced

.
body,

Effects

of Angle

_
._

of Attack

"._

Bodie_.____s,._
In selecting the circumferential
position of an inlet on a
the induced effects of angle of attack are of primary concern.
The

:-.

': ----'

- .........

=7.',' ":'_,_,_ i.

"

"_-_?,.-,,-',,.i;,',
"_".......

'

_'._
-_

'_._',,_.i._,_-,_

"

82

....
.......

" '"

''<

NACA 1_4

flow phenomena that must be considered are illustrated in sketch (28).

_J

St_ A-A

a<3"

Potentialflow streamline,tel 215

___
tolol pressure

8_
Bottom, Oe

I
Side, 90 e

J
Top, 160"

bioclt humber distribution Idowi.5 to 2.0

Ref, 216
Stotion AoA

Con_tont

totol-

preosure contcurs
Rot 217', 218

Sketch (28)

It is seen that along the top and bottom of a body in potential flow, the
flow direction is nearly parallel to the body c_nter line (i.e., at the
_ngle of attack _ with respect to the flight dire< 'ion); whereas along
the body sldes the flow inclination Is greeter, being _
on a right circulsa-cylinder SimJlarly, the local _.ch number is greatest on the body
sides and is lea_t in the forward bottom location _: th, leeward side
of the body, bhe flow is affected by viscosity so that the boundary layer
accumulates in lobes and, at sufficiently h_gh angles of attack, this lowener_ air leaves the surface of the body as a vortex wake. These general
characteristics of tL_ flow occur at subsonic"as well as supersonic speeds.
Several investigatiems of alr-_-.ductionsystems in the flow fields
of inclined bodies have been made. (See refs. 199, 20_, 218, 219, and 220. )

i
1_
I

|
f

NACA EM A_SF_6

i
.

Typical results are shown in sketch (29) in which the maximum totalpressure ratios attained are plotted as functions of angle of attack.

',

"L_L! i

'

I
--

, I

--

M, I)9

7-1tl
h'O
_;

30
"

,
"_
"

h-375

I
8

12

inch

4
8
I_.
Angie of ottock, e, de@

h=.750inch
0

I2

Sketc_ (29)
Half-conical shock inlets were mounted on a slender, low-drag body at about
the maximum-d_lammterstation, and the height of the boundary-layer diverter
h was varied. The 0.37_-inch diverter height h was about equal _o the
undisturbed boundary-layer thicknes_ at the inlet station at zero angle
of attack. These results confirm the desirability of the bottom location

':
_i

in regard to pressure recovery. This would be expected from the reduced

_iscous effects and flow angularity relative to this inlet which was
alined with the bod_ axis. The emgle-oi"_ttack performance of inlets in
the side location can be improved by use of the flow-deflector principle
(see ref. 26) or by alining the inlet axis with the mean flow directions.
_See re1_. 27-1.
)

_,
_
:
_

'

"
:,

Kremzier and Campbell in reference 220 compare the net prolralsive

force of a body-propulsion-unit c_mbination with the inlet on the top or
bottom of the body. Because of a lower drag of the inlet in the top
position, the net l,
ropulsive force was slightly greater at a given angle
of attack. However, at the same llft coefficie,,t,the _ottom location
was superior becauz_ of a negative shift in the angle for zero lift and
an increase in lift-curve slope for this position. In reTerence 2_2 tests
are described of the top inlet of reference 220 with two Barge triangular
fences extending ahee_lo_ the inlet to _hield it from the lemward boundar>

':,_

layer.

The net propulsive f_--_-_.,,_b_,art-.--_ement

at moderate angles
_"_,_IdIII_'

_ -_
i_
"_
_

_,

84

C_,ID_L
-

_,
c

_
"

NACA EM A99F16

/_

of attack was greater than that of the bottom inlet. A final evaluation
would, of course, require study of the effects of such large vertical
surfaces on aircraft directional stability and other related factors.
Since the upwash about a body decreases as the square of the distance
from the body center line (refs. 21_ and __,,_^_ "adverseeffect of angle
of attack on pressure recovery of side inlets can be alleviated by moving
the air-intake outboard. Thus, a comparison of the data of references
218 and 224 shows that if a nacelle with a conical-shock inlet were used
rather than a half-conlcal shock scoop,on the body sides, the same maximum" total-pressure ratio could be maintained by the nacelle at twice the
angle of attack of the scoop-body combination when the nacelle was over
about 1-1/2 nacelle diameters from the body center line.
Wi___n_.Wh@n the Mach number normal to the leading edge of wings is
subsonic, the circulation accompanying lift creates an upwasn field ahead
of wln@s which increases the effective angle of attack of inlets in or
n,-arthe leading edge. At low mass-flow ratios this upwash is aggravated
by the diverging engine streamtube. Fortunately, turbojet-powered supersonic aircraft, which are quite subject to lip stall because of thln llps,
seldo_ encouater the condition of high lift coefficient and low mass-flow
ratio Eigh-speed maneuvers are made with full power and normal landings
are made _ta some power at mass-flow ratios greater than 1. For subsonic
aircraft designed with a relatively large inlet area, internal lip stall
in landing would be more likely if it were not for the thicker llps that
can be u_ed.
An investigation of leading-edge inlets in a straight wing at subsonic
speeds is reported in reference 229 in _,hichit is shown that the induced
upwash from the wins causes an abrupt decrease in total-pressure ratio for
an inlet not designed to account for the additional flow inclination. For
example, an inlet with relatively thick lips maintained a total-pressure
ratio of 0.99 to an angle of attack of 6 , at which angle the pressure
recovery rapidly decreased _o 0.92 at an angle of 8.5. This decrease ir_
total-pressure ratio was caused by internal-flow separation from the
lower lip. It was found that the separation could be delayed by csnti:_
the duct axis Ju_t behind the llps downward and also staggering the inlet
plane. Tests of a similar leading-edge inlet at subsonic speeds in a
swept wing are reported in reference o_2E. Here, it was found that a
seriouK spanwise flow occurs in the inlet at low mass-flow ratios when
the wing carries lift. At mass-flow ratios greater than 0.4 and angles
of attack less than about 4, the perfo_n_anceof the inlet in the swept
wing was nearly equal to that in the unswept wing. At greater angles,
however, the pressure recovery decreased rapidly due to separation of the
i_,te_-u_al
flow. I_ is probable that this separation could have been delayed
somewhat by canting the lower inlet lip downward as was done with the
inlet in the unswept wing. At angles of attack greater than 6 to 8 and
a_ mass-flow ratios less than 0.8, separation occurred downstream of the

outboard edge of the inlet on the external surface of this swept wing
and resulted in an increase in drag and a loss in lift.
for supersonic aircraft have
the
Itlets located in
wir_ leading edge
received little attention because of the transitions and bends needed t,,

lead_ug-edge inlets for application to spllt-_Ing ramjets at Mach numbers

above
2.0through
are reported
references
127, 128,
227, 228,
and 229. Probducu air
a thinin
wing
to a turbojet
engine.
Infestiga_ions
of wing
ably the most important factor in the interference of the aircraft on
this type of inlet at supersonic speeds is that for unswept leading edges
there is no upwash induced ahead of the inlet by the wing. Body upwash,
however t can be present at supersonic as well as subsonic speeds.

i
:
"
_
i
:
_

From tests of wlng-root inlets, in which both the induced effects of

wing and body increase t_e local flow angles, it has been found that a
high level of pressure recovery can be maintained to angles of attack of
at least 8 at subsonic speeds by employing relatively thick lips with
stagger and negative incidence. (See refs. 186, 230, and 84.) The investigation of wing-root inlets of reference 84 included presm_e-recovery
measurements at Mach numbers up to 1.3. A total-pressure ratio of 0.89
Jas attslned at a Mach n_-ber of 1.2_, and this pressure ratio ,_asmaintalned frc_ -2 to 8 angle of attack

"_
_i

The results of reference 231 show that good angle-of-attack performauce

can be attsined by placing the inlet of an underwlng scoop downstream of
a wing leading edge so that the local flow direction is along the inductionsystem axis. A comp:llationof all these results from tests at subsonic
speeds is shown in sketch (30) as the
change in total-pressure ratio as the
angle of attack increases from zero. _I_

In
this
sketch,
the
wlng-ro_t
inlet
thobe
for
maxlm_n
pressure
The
mass-flow
ratios
of therecovery.
data are
of reference
84 shows
pressure
recovery
with improvement
increasing in

._

t
_

past

_._.84

6,r-,_,mT,_ _. z_u [

/o"5_1
_-S_"
mMn;m_in_s.mt.P.3q
,_,

/I
._ L_,,_,LA
_'s,_,,_._zswi,_--"_
I
_
o
4
_
_
_
_ _,m..._

"

|
_.0"_'0
m
2o

by
body
c_ossflo
w.
In
Sketch
(_0)
terms_h_n
of absolute
total-pressure
ratio at angle of _t_ack, the wimg _*ooti_lets are inferior to isolate_
inlets or those with upstream flow-deflecti_ surfaces. Although most
of these tests we.reperformed at M_ch numbers less tha_ 0.7, the low-spee_
results have been transformed to com_itio_ at a Mach mtm_er of 0.7 to
obtain a consistent correlation. As m_ntioned previously, this _,"amsformation can be reliably acc_mnli_h.e_if it Is assumed that the mea_ared

_
"

, .

_I_
_''_,_I.00
_
._ s,_

angle of attack because at zero angle

the recovery is relatively low (0.96). i
Angle of attack increases the pros_
sure recovery because the inlets are
canted and because part of the
_pproachcng boundary layer is swept

_
i_

W_,_

"

_i_

!]_

:_.'_
)_
._._
,_,:,,,
_
_
i

86.

CGN_I_E_Ir_,

_"

' NACA _

A5_FI6

ram-recovery
ra_io is iz_Lependent of _ubsonic fli@ht
Mach number and this
measurement is conw_rted -,,o
total-pressure ratio by equatio_ (13).

Effects of Forebody Boundary Layer

As previous dis=ussion has often indicated, forebc_y boundary layer

flo'_inginto an air-inductior system ce_ reducc engine performance because
of losses in total pressure, un_:teadiness,and nonuniformity. A comparison of the maximum total-presstu'u-athos as s fanction of flight M_ch
number attained with a variety of araaogeme,%s in which entering boundary
layer was not _cmoved is shown in sketch (31). The boundary-layer effects
intakes
which encircle
where
are particularly
large bodies
with annular

______
_'

_tP'4P'!|__Ss
the local Mach number is high. Such
s_--- __--inlets receive all the boundary layer
I
R,f._
I,_,_-c,.,_.._
a_nu'_r
i,_
from the flow over the forebody (the
. |
..... -I_
wltll S.k.t_mo_

_ _I
.4:
!
20

.4

large), and thi_.layer is either

_,_
_,_
_._11 . ratio of retarded to free air is
I _3_
_I
thickene_ior separated by compression
m,/%.aa
_
from the high local Mach number. The
=,o*
results of the tests of reference 232
i
s
_
_e
_.o _-4 show that total-pressure ratios of
u.
annular inlets mounted c_ an ogival
Sketch (31)

body are about O.3 less than those of

a normal shock wa_ occurring at flight
Mach numbers from 1.4 to 2.0. Similarly, the results at transonic speeds
of the nearly annular intake of references 237 and 233 indicate a relatively low to_al-pressure ratio when compared to nose or scoop inlets.
A conical-shock inlet with a small cone an_le suffers from these same
difficulties, and, as shown in reference Ill, bounda_v-layer removal is
necessary to provide steady operation. However, by using a scoop which
encompasses only a small portion of the forebody and thus receives a small
l_roportionof boundary-layer air, high total-pressure ratios can be more
readily attained. Thus, the results of references 234, 23_, and 236 show
that scoops mou_ed Just under the body nose where the botmdary layer is
thin and the local Mach number is low attain high pressure recovery.
However_ with scoops located downstream of the body nose where the approaching boundary layer is t1_ickand the local Mach number is nearly equal _o
or greater than that of flight, large total-pressure losses occur unless
the boundary layer is removed.
Seddon, in reference 28, _s correlated wind tunnel and flight data
to show the decrease in pressure recovery resulting from taking forebody
Boundary layer _mto air-inducticm systems. Seddon (see also ref. 2) (or
relates dcta by means of the relationship

'_.-_-_

..

_. , ,,.

_--

NACA HM Ag_F16 "

CONF_Eh__;_

87

'5

"

q--/-Pt_"
Pto
[ I +J(Vo/V
1 )

"
where

CfI represents the internal skin-frlctlon losses in terms of Apt/q2 and

J accounts for pre-entry effects. Thus, k is an empirical constant
which includes the effects of inlet-velocity _atlo Vo/V2, and _ is a
correction to the skin-friction coefficient due to the previous history
of the boundary Sayer before it reaches the inlet (_ = Cfforebody,Cfduct),
J

_b is an efficiency factor to account for the amount of boundary-layer

removal; and S/Am is the ratio of forebody surface wetted by the flow
to the inlet divided by the inlet area. At reduced inlet-velocity ratios
and high speeds without complete boundary-layer removal, the bound_ylayer thickness ahead of the entry inc-'easesrapiaSy and, as a result, k.
_, and _b become functions of Vo/Ve and Mach number which must be
evaluated experimentally if accurate results are to ie obtained

"_

Boundary-Layer Removal
.

The design problem with a boundary-layer removal system is _o avoid

incurring any appreciable drag penalty while removing sufficient retarded
air to minimize pressure losses, unsteadiness, and nonuniformlty in the
engine streamtube. The boundary layer can be removed by providing suction
across a slot or a porous surface or by raising the inlet from the forebody surface so that +&e boundary layer flows beneath the inlet and is
diverted around the externsl surfaces of ti_educt fairing.21 In doing
this, it is necessary to minimize any addit._onaltotal-pressure losses
and interference with other parts of the flo_ field. The following discussion on removal systems is divided according to the method by which
forebody boundary layer is prevented from entering the air-induction
system - by suction or by diversion. These methods are similar in some
respects, Out a suction method is one in which a pressure difference is
provlded across some length of closed duct to draw off the boundary layer,
and a diversion method is one in which the flow is unrestrained in a lateral
direction. Under certain conditions, the effects of boundary layer can
be minimized by providing large-scale mixing with the engine flow, as is
the case with the NACA submerg_.dinlet. This method is also discussed.

'

:
';

"

2mSome tests have been made of diffusers in which energy is added to

the boundary layer by blowing air from a high-pressure source along the
forebody wall; the results are reported in references 113, ll4, and 238.
However, extensive development of this method as applied to air-induction
s,vstemsha_ not _a_tb,:e_:Pe_ormed.

88

CQ

NACA _M A59TI6

Suction.- An ev,_luation of a suction-removal

system on the basis of
aircraft range has been reported by Fradenburgh
an6. Kremzler in reference 19.
Tests were mexle with half-conlc_l
shock inlets with semicone
angles of 25 and 30 with various heights of boundary-layer
removal slot
at Ma_h numbers o_' 1.5, 1.8, and 2.0.
Because of the large drag force
contribused by tLis specific boundary-layer
removal system as noted in
both references 19 and 239, boundary-layer
removal produced essentially
no increase in maximum range in spite of the substantial
improvement in
pressure recovery.
Thus, careful consideratiou mus_ be given to the detail
design of removal systems to prevent energy losses and to achieve the
potential improvement in perfozmsmce.

m_

The data of reference 18,5 show that in subsonic fllght_ operation

of an alr-induction
system at Inlet-velocity
ratios less than 0.6 causes
rapid thickening of the foreoody boundary layer flowing into an i_let.
The tests of a boundary-layer
removal system that were included in this
investigation
show that the inlet-_veloclty ratio of the removal system
must.be greater than about 0.9 to maintain a net drag force less than that
for the coD_flguratlon without boundary-layer
removal.
The boundary-layer
8coop in this study was in the plane of the main inlet and was produced by
indenting the foxebody.
It was found that an indentation approach ax_Dle
of 7 caused unsteady flow.
An approach angle of 3 resulted in satisfactory operation; however, as discussed later in regard to submerged
inlets, such approach angles would cause unacceptable
losses in pressure
recovery at supersonic flight speeds.
In the tests reported in reference 240, a removal slot of depth equal
to about twice the local boundsx,:-layer thickness was located ahead of a
semicircular main inlet a distance of about 85 percent of the inlet radius.
Tests were made at low speed at inlet-veioclty
ratios greater than 0.6_
hence, the effects of removal on total-pressure
ratio were not large.
In
these tests it was found that the boundary layer on the sla_face between
_he boundary-layer
scoop and the main inlet grows rapidly at low inletvelocity ratios.
Thus, this length should be minimized.
A study of boundary-layer
removal at a Math number of 1.88 for a
h_df-con_cal-shock
inlet mounted on a flat plate is reported in reference 202.
Here, it was shown that the maxirman total-pressure
ratio
attainable
in the main duct decreased appreciably
as the amount of boundary
layer removed was decreased.
As the parameter
h/8
was reduced from 1.O
to 0 (h is the boundary-layer-scoop
height and
8 is the local undisturbed boundary-layer
thickness) the maximum total-pressure
ratio decreased
from 0.86 to 0.72.
In this case, the m_ss-flow ratio of the removsl scoop
was the maximum possible; at any value of h/8
below 1.0, reducticns in
scoo_ mass-flow ratio caused additional total-pressure
losses.
Also, with
this air-inductien
system the flow became unsteady when the engine massflow ratio was reduced below that for maximum total-pressure
ratio.

,,_ada

m.

."a
4

i
_

Tests at N._chnumbers from 1.3 to 1.8 of a suction-removal system

for a normal-shock inlet are described by Frazer and Anderson lu reference 190. It was found that boundary-layer removal produced an improvement in total-pressure ratio of from 0.06 to 0.08 through the Mach namber
range of the tests. The fact that this improvement was considerably less
tha!._h_t _t+_
with the ,half-conicalshock _let of rcfcrcnce _ _
probably due to _he difference in _he methods of exte_al ccmpression and
of duct design. The air-induction system of reference 202 ws_ more refined
in regard to supersonic compression but less refined in the duct. Thus,
with nearly complete boundary-layer removal, higher total--pressureratios
were possible but gith n_ boundary-layer removal greater duct losses would
be expected. Frazer and Anderson show that pressure recovery could be

..

shock wave occurring at the _ocal Mach number of each element of the flow
approaching the inlet and adding an allowance for the skln_friction loss
in the duct. This method of prediction is also recommended in reference 241. The tests showed that, if h/5 _ 1.0 and no additional _-ethod
of
boundary-layer
removal
is used, the
leading
of the
suction
fairly
well predicted
by integrating
the
local edge
pressure
recovery
ofscoup
a normal

'

_
:

the
- not i_L_t
aheadand
of the boundary-18&er
scoop
- occt_ron
if flow
must intervening
be upstreamsurface
of the main
nozm_l ShOCk wave
mu_t
unsteadiness is to be avoided. For mass-flow ratios greater than 0.9 at
Mach nambers from 1.3 to 1.8; it was founa _n this test that thesuction
scoop must be at least a distance of 0.4 of the inlet r_-llusupstream
of the main inlet. (The cross secticn of the main inlet was a semicircle)
The mass-flow ratio of the s',ctionscoop was maintained a_ the maximum value
in this investig tion, and.by measuring the total pressure in both the main
and the .....
ootm_y-layer
ducts the net propulsive force possible with the
system was evaluated. It was found that the maximum net propulsive force
occurred when the suction-scoop height was 0.7 of the undisturbed boundary-

from 96 _o lO0 percent of _hose produced by a normal-shock inlet not in

layer thickness and that the system could produce net propulsive forces
In suction-removal systems, the performance penalty for removing the
boundary layer appears as the pressure loss in the removal duct. This,
the
presence
forebody
bounda,.'y
together
with of
the
mass flow
in thelayer.
scoop, allows calculation of an effective drag of _he boundary-layer r_oval system. A s_m_naryof available
data for the presm_e recovery of suctioa-

'
i
_i
_

flight
number
indicated
in
remvalMach
ducts
shwsas
a large
decrease
with _ __.
sketch (32). (See also ref. 242.)
s _.4Q_

"

drag of a boundary-layer d_version

system, the depth of the bo_mdarylayer passage should be no greater
than that required to maintain satisfactory engine flow, and the speed
and direction of the diverted flow

_,.
_

l
_

l
Z_
C
o

.4

should change as little as posslble._

-- .........
-_..
o

I "t,.o "'

"

A5___

! --_

I . I

!', I.... _,

_'_
,q_f.190._P_--_
'

h IOm
J' --._-_"
,--"IJ
'S 8
ram.

_
_._._,[
'-ts
_,,.0,._,.0-_----_'o---_--_
8

_2

l
I
,6 _o

"

_8

Sketch (32_)

|
-"L-_.

................
I_111
I

-{k

_,

IIIIIIr
_'_l_'_
" --

_'t_ _

___

[l_l
r "_]I_/'i
_ l "',_ IIllll
lll
ll

!
!

....
!_
__
,

..... , ,,

._ __ .
- .

el

: , ,.: .::
-If

III|

, , -

NACA I_4

Thus, the main Ir_ietand duct shotLldbe designed to be insensitive to at

least small amounts of retarded air from a forebody; wetted area and
def!eetion ez_!es in the diverter must be sma3"._and the pasEa_e helght
d_st diverge both longitudinally and laterally to mlaimJ=e flow resistance.
The b_audery-layor suction scoop of reference 202, which was tested
"
w_th a half-coaic_/ shock inlet on a flat pla_e, w_s conver_ed to a
diverter system by re_vlm_ the scoop side walls to a point about one
inlet radius aft 0% the cowl lip az,dtaking no flow through the boundarylayer duct. It was fo_Id that maklmum tot_l-pressure ratios _rQm 0.0_
to 0.03 les_ than those of the suction system could be attained by sweeping
back the leading edge of the plate fo_ming the upper surface of the
_dverter, tJaatis, the boundary plate, as shown in sketch (33). This
plate was swel_tback along a line
Joining tae apex of the cone with the
main inlet ll_ _-atherthan the leading
edge of the plate beln normal to the
st._enmdirection at the cone apex.
It was concluded from these tests that
sensitivity to removsl-system massflow ratio can be reduced by sweeping
the leading edges of the boundary
plate so that the intensity of..
the
,_

80undary plo|edisturbance
created
by the
shock
from the edges
and the
extent
of wave
the
upstream .influencethrough the boundary
Sketch (33)
layer are reduc=d. Swept edges also
create a latereO pressure gradient
which tend_ to divert the boundary la._er. It -_asfound that extending
the boundeu--j-layer
passage downstresr_beyond the plane of the main inlet
reduced the angle through which the boundary layer w_s dlverted and prevenbed the b_.mdary layer"from being drawn /a._othe engine streamtube.
(See also ref. 243.) Tests of other inlets which u+ilize these design
pzincip!es are described in references 3.82,244, 2_, and 2_6.

The restGts of tests of a wedge diverter of about 60 included angle

beneath a ball-conical shock inlet mounted on a flat plate are presented
is refereuce 2_3. As would be e:_ected from the results of Goelzer Bad
Cortzight (ref. 262), this large a wedge angle turned the boundary layer
so abraptly that it spilled over the sw_.ptleading edges of the boundary
plat_;Bad flowed into the main inlet. In order to attain the _otalpressure ratios possible with a suction scoop, it was _.cssary to have
a diverter passage height 1.4 times the local undl_urbed boundary
thickness; thus, a high dr_g wo_/tdbe expected. In reference 247, a
series of wedge_ were tested in an arraaement _Imulatlng a dlverter passage. it was found that the included wedge angle must be less than 28
if the pres._ure_rag is to be small and that the a_.exof th_ wedge must
be about one I,as_.agc
heig_ downstream a_ the apex of the leading edges
of a s_ept boum_
plate in order to eltuimate the upstream influence of

".

|
P

NACA _

A_FI6

/d. _

91

the wedge on the engine flow. The photographs of the boundary-layez

flow of Pierey and Johnson (ref. 247), which were obtained by use of a
liquid-film technique, emphaslz_ the importance of minimizing the disturbances imposed upon the boundary layer in the region of an alr-Inductionsystem inlet. The nece=sity of a small wedge angle, e swept and thin
boundar# plate, and a wedge apex downstream of the splitter-plate apex
are all graphically illustrated.
The drag forces on wedge diverters in various types of installations
have been measured and ar_ reported in references 218 aud 248. With _
16 included-angle di_=rter_ the pressure drag was negligible, but the
viscous component of the drag was large. In fact, even though the fr_tal
area of the diverter was o_ly 3 percent of the total frontal area _ the
model of reference 248s to a flight Mach number of 2.0 and h/8 = 1.O Lhe
drag of the diverter was 23 percen_ of the total model dry, or, in other
words, the aiverter-plus-interfsrence drag coefficient based on the diverter
frontal srea was hlghj 0.9_. Improvements can be expected through reduction
of the viscous drag due to shock-boundary-layer interaction and turbulent
mixing in the vortex from sharp side edges. Not only should wetted area
and velocity changes be minimized, but also a high lateral velocity component over nearly square side edges should be avoid_d because a vortex
develops under such conditions and dissipates energy az drag. (A vortex
frc_ this cause is used to advantage with NACA submerged .1_letsat subsonic
flight speeds. )
The fact that a low-drag passage between an air-induction system and
a body can be attained is illtu_tratedby the investigation of Kre_zier
and Dryer (ref. 249) in which a c_,-cularnacelle _s tested,in contact w_'th
a circular f',selage. This confl_,_rationis shown together with a body
scoop diverter in sketch (34). By comparing the drag coefficients of the
configtu_ationsless the body drag on the basis of equal area, it was found
that the drag coefficient of the scoop-diverter combination was about twice
that ol _hc nacelle. Som_ of this difference is due to the fact that the

le

i_

at

#1_

|el

c,

models were not strictly compareble; however, the difference is so large

that the superiority of the nacelle installation is apparent. Similar
results were obtained by comparing a ramp-type scoop inlet and diverter
(ref. 248) with the nacelle. These co_parlsons and present knowledKe of
diverter design indicate that a low-drag diverter should be designed
according to the following principles:
1.

To reduce the upstream influence of the dlverter, the leadin_

edges of the boundea_ plate should be swept back, when this is
cunslstent with the inlet-shock confi_oration, and the diverter
apex should be at least one diverter height back of the boundaryplate apex.

2.

To reduce pressure _I friction drag and to minimize the l_teral

velocity component, the included angle of the dlverter wedge
should be about 20 .

B.

To prevent the fox,

nation of a strong vortex, the boundary-layer
passage slde edges sholtldhave large-radius fairlngs rather than
sharp corners.

._

As discussed previously, the distribution of boundary layer about a

body at angle of attack is not uniform and it accumulates on the leeward
side (sketch (28)). If an inlet is located in this position, the design
of the boundary-layer removal system _ist account for the local growth of
the boundary layer in angle-of-attack operation. (See, e.g , refs. 220
and 2b_.) If a large boundary-layer diverter is necessary to maintain
engine performance at high angles of attack, a drag penalty results st
low angles. As shown by the data of reference 199, this difficulty is
avoided at positive angles of attack by a bottom locatie_ of a side inlet.
Tests have been made of combined suction and diverter systems; that
is, a portion of the approaching boundary layer i_ drawn into a closed
duct, usually for cooling pur_,oses,and the remaining boundary layer is
diverted, (See refs. ll6 and 290.) With the sucti_l scoop at the apex
of the diverter wedge, the upstream influence of the divezter is reduced
by increasing the local flow rate and reducing the local deflection angles;
in other words, it allows lower dlverter wedge angles. If the auxiliary

-"

eystem requires low-energy air, the best point at which to locate an

auxiliary inlet in a diverter passage might not always be at the wedge
leadi_ edge. It is _pparent that the lowest energy air can best be
obtained at the exit of the diverter passage. It is possible that such

z
[

an installation would have less drag than one with a forward auxiliary
air intake because the dynamic pressure of the local flow is smaller.
,_

Submerged inlets.- Inlets which are submerged in the surfaces of

bodies and wings have all the boundary-layer-removal problems of scoops.
A number of variations of inlets of this type have been investigated and,
as with scoops_ high pressure recovez-jcan be attained at subsonic speeds

..... ,r

....

'

" ' ,',

"

" 11_I

,_,,, ,,,,,

....

--

NACA EM A59FI6

when the adverse effects of the approaching boundary layer are removed.
Investigat._onsof submerged inlets haviug curved or steep-angle approach
ramps with parallel sides are reported in references 291, 2_2, and 253.
In general, the total-pressure ratios attained were less than those of
_imilarly placed scoops. A submerged i_iet having a relatively small
ramp angle (about 7) and diverging ramp side walls has been found to be
coupare_le to scoops in regard to pressure recovery. (See refs. 294, 2_,
and 2_. ) The experimental investigation of reference 257 and the Gheoretical study of reference 298 provide an explanation of the relatively
high pressure recovery of this arrangement. Flow over the square corner
of the ramp side _'allscrestes a vortex which thins the boundary layer on
the ramp and sweeps the retarded air into the vQrtex co_'e. When the
vortex flows into the inlet at high mass-flow ratios, it represents a loss
in total-pressure ratio, but less of a loss than if _he boundary layer
were permitted to _ow normally; at low mass,-flo_ratios, the vortex is
discharged externally and represents an increase in drag. Tests at low
subsonic s_eeds, reference 28, have indicated that the drag of submerged
inlets can be greater than that of scoo;s. However, flight tests comparing
a submerged and a scoop installatlcm (ref _26) have shown that the former
has equal or slightly better performance. Appsrently, the merits of the
two depend upon the installatlon, and they can be equal in subsonic flight.
However, iuve_tigatiom at supersnnlc speeds, reference _9, has sho_a
that the expansion of the flow over the ramp leads to a high inlet Mach
number and large pressure losses at flight Math numbers greater than about.
1.2. Thus, the Submerged inlet is limited in application to subsonic
airplanes _s either a main or an auxiliary air intake. (For the latter
application, see refs. 2_i and 260.)

!
:

_,

: -

;_

CONFIDENk_AI _,

-_

Scoo_ incremental dra_.- As

discuss--ed
previously, scoop incre-

'

ence in the total fl_ght momentum

of the air in the engine stream-

_ +

-_

Combined Effects

initial
of an at
alr-inductlon
tube andstation
the mmmentum
the
_
system. It is_ therefore, an
Interference force resulting from
both _he pressure and skln-friction
arag forces on surfaces upstream of
"

93

Mo,,.O
P%/P'o

kk_ _

/I
"---- i

'

fl_,,_
m_m o

an induction system when no proSketch (39)

vision is made for removing
forebody boundary layer from the emgine'streemtube. Klein (ref. 7) has
calculated scoop incremental drag-c.oefficient
CDs = Ds/qoAm as a function
mass_flow ratio, flight Mach numl)er__
total-pr_ssure ratio bellmen free
stream and inlet. An ex__._ion
is shown in sketch (35).

_,,9,,i_m,_m_k _

' il

_i

_ ,, . L___ '_'---

_ , ', ,, _+i_ _ ,_v-?-_-''"_'_"'_,'_;

_.

!
/
!

-|
_
......

c
f
%

_t_

(|
t
_

It

me

u@

@ 4
ore

_,

NACA _

AS_F16

oe,

ee

:"

Thus, when the average inlet Mach n_uzberis subsonic in cupersonl flight,
the scoop incremental drag_force is large at low mass-flow ratios, particulexly i_ the forebody wave an-/skin-frictlon drag forces are small,
because the _ the Ioc_i pressure rise ahead of the inlet is large. (The
symbol Pt2 is the average total pressure at the inlet, and it includes
the total-pressure loss of any entering boundary-layer air which eventually
flows to the engine. ) With supersonic flow into the inlet, the scoop
incr_nental drag coefficient is negative because t]'?_pillage drag22 is
small (zero at m_xinnnnmass flow) end the forebody drag term FB of eOuaticm (7) is dominant.
For air-induction systems having thxs interference force, the net drag
consists of the sum of the external wave drag when the inlet operetes with
no spillage, the scoop Incramen+_l wave d_ag, the change in external wave
drag due to a reduction in mass flow from the maximu_ and skin friction.
".bus,%h" zc_)p incremental drag replaces the additive drag of systems
having no xorebody interference.
.Uakos.-The pressure recovery of am alr-lnduction system that takes
in air frQm the wake of an upstream body is, of course, reduced. The tests
at a Mach number of 2.0 of reference 224 in wb4 =h a nacelle was placed
behind the tip of a canard control surface illustrate f_e magnitude of
this effect. With the control surface deflected I0 , the maximum totalpressure ratio atta_mable was O.lO less than when the nacelle was moved
outboard away fr_n the influence of the tip vorter.

INDUCTION-SYSTEM AIRCIVA_T

The interference between an air-induction system and uther aircraft

components can affect any of the forces and _ents
which determine performance. For instance, drag can be increased if a nacelle is placed so
that a positive pressure gradient from it causes boundary-layer transition
)r separation on a neighboring surface; the lift of s wing with a leading_dge i_let can be a function of mass-flow ratio; tail loads can be affected
by a change in circulation distribution resulting from changing the wing
plan form to extend the duct of a wing-root inlet; side force and yawing
moment can result free shock or expansion interference from an outboard
nacelle with a vertical tail surface, and this interference could be
changed by power setting. It is the purpose of this section to discuss
these problems and principles regarding them which have resulted _rom
theoretical and experimental studies.

22Spillage drag is the pressure force an the external stres_lines

which are affected by the inlet mass-flow ratio. In this case, it is the
local additive dra_.
.,,.

i.

il
.. ,
-

;-.........

.
.

"
,.,

_,....
..
_,,_:,..,,_,_,._:
_ "A'.':'."
._
':
'*
_...._.__
.-"
' ....
.'_'
'_.9._%"
.
.

,.,

.,. '._...............
--..--.__.,._,w.,,_,,..,_._,._...._.._-._-_.___.-._,_,,...._._
-|

Drag
a

"

Skin friction and se_araticm_.-In reference 207 a series of wing

naccl!es were tested to demonstrate a method for maintaining long runs of
laminar flow over the combinations. By making +.heleading edges coincident and matching the pressure &ist_ibutions so as to maintain a negative
_adient to the position of ._'._,_ thlcknes_ of th_ wins, the minimum
drag coefficient was r._duced_o less than two-thirds that of conventional
wing-nacelle combinstf_n_ when the inlet-velocity ratio was greater than

0._.
'i_aetuft studies of reference 185 show that an inlet-velocity ratio
less than 0.6 with a scoop in the presence of forebody boundary layer not
only causes separation of the internal flow, but also causes the separabed
region to spread around the inlet and to affect the external flow.
Although inte-_ferencedrag was not measured, it is undoubtedly increased
by the turbulent _ixingo The flight tests reported in reference 256 =.how
the possible effect of such separati_u. Drag measurements were m_de _ith
a boundary-layer b_pass sealed, and with it discharging normal to the
external flow, it was found that at a flight Mach number of 0.8, discharge
of the boundary-layer normal to the air stream increased the airplane drag
coefficient 0.0015, or 7 percent.
s

.
i
;

At supersonic speeds the boundary layer on other aircraft components

can be affected by shock waves or the pressure field from propulsive
systems, and, the local pressure gradient_ caused by shaping a surface so
as to minimize wave drag can be sufficient to separate a turbulent boundary.
iay_r. Therefore, this form ol interaction also requires caref_l attention.
Shock-wave boundary-layer interaction has been discussed previously, but
the studies of Morokovin, Migotsky, Bailey, and Phinney (ref. 261) are
particularly pertinent here. This investigation of the interaction of a
plane oblique shock wave intersecting a circular cylinder across the axis
shows that if the incident shock wave is weak, the pressure rise across
the reflection is that predicted for two-dimensional flow, However, if
the shock wave is relatively strong (flow deflection angle of ll.2 in
this case) uhe over-all pressure rise is but half that predicted for a
flat plate. This difference is presumably the result of three-dimensional
relief and the resulting lateral pK'essuregradient. Because of the
decreased surface press_'e rise for a given shock wave, it appears that
more intense shock waves can be _ithstood without encountering separa__on
of turbulent boundary layers in three-dimensional rather than twodimensional flow.
Transonic dra_jrise.- In general, the addition of an alr-induction
system to the pressure field of another body alters the pressure distribution and _hus the transonic drag rise. The investigetions of refere_ces 225, 84, 85, 2_0, 231, _nd 262 sho_, however, that wing root or
_ing leading-edge i_]ets e_d _acelle_ operating at mass-flow ratios near i

-.--.....

____
,

_ L,,

_._
.....
_-

..........................
,_

_.""_'_"_,":'".':_'._L"
i

"'

,.

.,...,!

".
i

!
/
t

T I

:
_

I
.

I
il

II1
I

II

tl

can be desired
so that they do not decre_e
appreciably
the d_-rise
$_
number of a viz_-body
comb_ation.
(Methods of predicting
the dragrise Mach number have been discussed previously and are presented in
references 149 and 208.)
For supersonic aircraft, the drag-rise Mach number is an important
cru_e consideration; the magnitude of the rise and methods for minimizing
it are of essential importance in determining acceleration performance and
fuel eonmm_ption. Re "transonic area rule" pr__sentedin references 26j
and 264 states that for slender configurations, the transcmic rise in wave
drag is a function of _ne longitudinal distribution of cross-section area
and is independent of cross-sec_ion shape. Thus, an aircraft with the
least drag rise has the same distribution of cross-section area as a
minimmn-drag b(xlyof revoluti(_n_ Conversely, the magnitude of the increase
in wave drag at transonic speeds for complicated configurations can be predicted for flight at zero angle of attack from information on bodies of
revolution with the same cross-sectional-area distribution. It follows
fram this rule that for low drag rise the equivalent body of revolution
must be fair and slender, and these design requirements also result in
high drag-rise Mach nmnber.

In regard to interference of the air-induction system on the aircraft,

the transonic area rula is a design criterion for placing aud shaping
induction systems. For instance, the data preseated in reference 265 show
that the drag rise is the least and the drag-rise Mach number i_ the greatest when the addition of au air-induction syst-=mto a wing-body combination
causes no abrupt or large changes in the distribution of cross-section area.
This result is il31ustratedby sketch (36) which was reproduced from reference _6_. References 170,
(_ _F_
TYPES (_ _LLE.S
2(_, and 267 present more
experimental information
Ct'O
" F.A /_
concerning the interference

"
,

of air-induction systems
with aircraft at transonic
flight speeds.

,,_,,
_8

0241
|

_
___

_______
_

sonic area rule has been

Wave-dra_.- The tranextended for application at

_
I@DI_J_$-_#/'_

R. T. Jones in reference 268,

thn

._

extension
have- been examined
by Lumax in reference 269,
It is shown that, for slender
L_
0 8

_+_
LO

12

L4
M

L6

LO 2.0

Sk_-tch (36)

....................

"

" "
,

"

_ _ "--_ "_'"_';
"__";
*

. ,',

_--

.... .....
, ,,

, ,

aircraft, cross-section areas

can be taken in planes through
a point on the body axis
inclined at the flight Mach

'.

--

--

"

Illill

":':;

....

;,

.........

|1

_
,

.....

,",

II....

' _:'"' : ....

i
,., .,""'_

NACA _4 A95F16

CONFIDE_P_',

97
i

angle to obtain an equivalent area distribution. A sufficient number of

planes mast be chosen so that an accurate average oblique section area can
be computed. Then, from this equivalent area distribution, the wave drag
can be _.alculatedby slender-body theory. For configurations in which
the ares distribution is chosen so as to minimize the drag, the design is
near optimum o_ly for a small range of flight Mach numbers about the
design point. The experiments of_reference _68 .qubstantiatethe use of
this method as a design criterion. _ualysis of drag data for a _ide
variety of configurations indicates that predictions of drag are iraerror
by a maximum of about 20 percent with a mean error of about 7 percent.
As pointed out by Jones, these area rules are basically methods of wave
cancellation - the pressure drag of one component is canceled by proper
use of the pressure field from another component.

More d_tailed theoretical investigations of wlng-body combluations

in supersonic flow indicate how components can best be shaped and arranged
to provide wave c_ncellation. Daldw_n and Dickey in reference 270 demonstrate the importance of the moments of the area distribution at Mach
numbers above 1. Both experiment an_ theory show that t_e Mach number for
drag rise. is high and the subsequent drag rise is low if t}:elongitudinal
distribution of the moments of area is smooth and grsdu_l. (At low supersonic speeds moments greater thau the second are of negligible importance.)
Nacelles can be used to improve the moment distribution of wing-body combinations, and the data of reference 270 show that the h_.gh-speeddrag
characteristics of a wing-body-ns_elle combination can be less than those
of the corresponding wing-body co_blnation. The studies of references 271
and 272 indicate that rotational asymmetry of body cross sections in the
region of a wing Juncture provides greater wave cancellation than a symmetric indentation. Nielsen (ref. 2_[2)employs linearlzed theory to determine the change in shape of a circular cylinder required to cancel the
wave drag of wings. The method can be extended to the interference problems of nacelle-fuselage or to nacelle-wing-fuselage combinations as long
as the flow is quasl-cylindrical.
In reference 273, Friedman and Cohen consider the minimum wave drag
of two- and three-body combinations. It is shown by linearized theory
that the least dr_g _s _!oduced in supersonic flight when the "oodles are
close together az_dstaggered so t at the pressure fields interact to
produce a buoyant fo:'cein the flight direction. The general txends of
this analysis have been substantiated by the experiments of reference 249.
Here, the forces on both single and twin nacelles with normal-shock inlets
operating supercriticaily were measured in the presence of a body of revol_tion having a parabolic-arc radius distribution. The nacelles were
moved both axially _.d radially, and it was found that the theoretical
predictions were fairly accurate for forward locations, but for rearward
inboard locations there was considerable deviation from experiment. The
favorable interference effects at the rear inboard locat:'onswere equal
to or _'eater than those indicated by theory. In reference 274 a nacelle
with a conical-shook inlet operating subcritically at a flight Mach number
,--

_.m.--mm_Imb

!
,

of 2.0 was located at tvo positions in the pressure field of a complete

aircraft configuration.
A lO-perc_nt increase in drag coefficient rereadted
at zero angle of attack when the nacelle was moved from a forward to _>
rearward locstion.
This large increase in drag, which is opposite tc b_e
trend indicated in reference 249, _as attributed te the strong shock w_v_
from the inlet.
Lift and Pitching

_:_

Moment

The lift force of an air-inductlon

system consists of the lift component of the pressure forces on the external surfaces and of the reaction
from the force required to turn the engine streamtube f_om the flight
direction to that of the induction-system
axis.
Th_s force from t n_iing
the internal flow is carried on the lips, see reference 27_; and, as shown
in reference 1_2, in terms of the incremental lift coefficient based on
maximum body frontal area for a slender body, it is

2_CL = _

m_

(_

(33)

in radiant)

the corresponding
incremental pitchLug-mc_ent
coefficient
is, of course,
the product of this lift coefficient and the distance from the inlet to
the moment reference point d_vided by the moment reference length.
Pierpont and Braden in reference 234 compare this prediction with data
taken at subsonic speeds on a b_ly having an underslung scoop Just behind
the nose.
The results for a flight Mach number of 0.8 are shown in
sketch (37).
The effect of mass-flow ratio on the llft of the body-scoop

.3

.2

._

Theot'etical

//'

internal

Theor,ticoi

/j

internal

///
l / / Basic body

@J

12

(o) Effect of moss-flow ratio

(37)

Sketch

(b) Effect

......
,..........,_.--.,,,_,-_._
-.-............................
,. ....... ',..
..

.......

'

.-

. _.,.,.----,<,--,,,_-.,.,_,.,,.._,_il_:

!2

._...-.

....

. -,_

.. _ .....

- ,.

,.,.-...........

.-

-'"

_'__t_:_i_T.

'-'

- -, .....

@f basic body

-','!,---_-..........'--., ,' .__-;',r,_

_:_ _ :._._,.._,,-._-.-._
_.

" _,_..

....

: "
,,,
-_,,,t,__ _'_'-: .............

"_.*- -,

-.-,
',_ ,,"_'1

*..

L T_

c_foination is closely predicted (sketch (37a)), but there is some error

in predictir<gthe lift resulting from addition of the scoop with a massflow ratio of 1.0 to the basic body (sketch (37b)). This difference is
probably due to the fact that interference with the pressure field of the
basic body resulting from addition of the scoop was not taken into account.
The in..'remental
pitching moment of this configuration was not well predicted
appa_'_:n_ly
b_cquse the drag component of the moment contributed by the
asyn_etric scoop chax_ge_with mass flow and angle of attack and counteracted the moment due I;othe incremental lift. A comparison of experimental
and theoretical lift (_oefficientsat supersonic speed for a slender_ opennosed body of revolution is presented in reference 192. Here it was found
that up to an az_gleof attack of 4 the prediction _ed
with experiment

_
<
::
[
_

within 7 percent and there was comparable accuracy in the pitching-moment

comparison. The contribution of these effects to the lift and pitching
moment of a complete airplane is, in g_ersl, relatively small; for
instance, the increment_ lift-curve slope due to turning the engine flow
a_ a mass-flow ratio of i is only about i percent of that of a normal,airplane. Thus, in most cases great acc-ca_acy
in predicting mas_-flow effects
on llft and its mGm_n_s is not necessary.

_
%
_
],

'

In the following discussior, the interference of various air-irduction

systems n llft and pitching moment are presented. Forces and moments
in other planes are not discussed becaase, in general, they result from
the same phenomena.
Wing i_aalng-edge ini_t__.-Tests of wing leading-ed_Eeinlets in both
stcaight and swept wings with NACA 63-012 airfoil sections are described
in references 2_9 and 226. For the straight wing, the effect of internal
flow on both the lift- and pitching-moment-curve slopes was negligible.
There was a large effect of inlet-velocity ratio on maximum llft coefficient at very low flow rates, but for the range of usual interest, inletvelocity ratios above 0.8, the maximum lift coefficient of the basic wing
was maintained. With the swept wing, there was a large change in the

_
j.
<

flow at thc downstream corner of the inlet at lift coefficients above

0.6. The maximum section lift coefficient at 0.8 i_,let-veloci_yratio
wss i.i0 at the upstream corner but, at the other corner, it was 0.72.
R,,_ducing
the inlet-velocity ratio to 0 or increasing it to 1.6 changed
_Ae_e section lift coefficients by 0.i0 at most. It is thus apparent that
with a swept wing, flow through a leadlng_edge inlet can ser_ausly interfere with the liftir_ force and its distribution.

!
_

.
!

;"W-

Win6-root inlets_- A wing-root i,_letof triangular frontal shape was

tested on a 49o sw_ptback wID_-body combination as described in reference 186. The inlet lips were parallel to %he wing leading edge, and the
lip profiles were refined by ci_anginginclination end stagger so that for
mass-flow ratios from 0 to 1.9 internal flow had no effect on lift-curve
slope or maximum llft coefficient. Tests reported in reference 8_ at
higher speeds showed no effects at mass-flow ;;atiosfrom O.h *o._.,
_ up
to a flight Mach number of 1.2. In reference 89, the res,_itsof tests

--_

,
"

.n,

r_

....
.,
of a stmilar configtu_tion differing only in inlet frontal shape (semielliptical rather than triangalar) are presented. The effects of internal
flow on lift _:ereagain negligible, but here pitching moment was measured.
It was found that at Mach numbers above 0.R, the presence of the inlet
with mass-flow ratios of 0.4 or 0.8 increased the static longitudinal
stability of the wing-oody combination tested by 25 percent. _n this test,
the i_et had no e_fect on the l!ft coefficient at _hich the slope of the
moment curve reversed.
Tests of a wing-root inlet mounted on a swept wing with the inlet
planenormal to the flight direction are described in reference o76.
The inlet plane was ahead of the leading edge of the root chord of the
wing alone, and thus the installation of the air-inductlou system modified
the wing plan form_ Flight tests revealed a strong pitch-up aoove an angle
of attack of 8, and w_nd-tunnel tests showed this to be caused by an
abrupt change in downwash at the tail, du_ to a change in circulation
about the wing as the angle of attack was increased. The pitch-up wa_
eliminated by changing the section contour of the outbos_'d9ortion of the
wing leading edge and by adding fences both at _he inlet and outboard on
the wing. An ir_letwith the outboard radius about half the scoop depth,
and an inlet width-to-height ratio near 1.0 also eliminated the pitch-up.
It was concluded that the wing plan form and sharp side edge resulting
from the addition of the extended wing-root inlet was the cause of the
unexpected downwash varla'_ion. Tests of a somewhat simi]_u_ config_aration
for a supersonic airplane 8_e reported _n reference 144. In this case
there was no longitudinal instability for the condition of no flap deflection, probably because of the low position of the horizontal-tail surfaces
and the rounded side eages of the inlet.

Sc__s.- The effects of scoops on the lift of a complete airplane

are generally not large, just as the body llft is net a large percentage
of the total lift unless the bo_y diameter and wing spen are nearly equal.
Thus, top and bottom scoop locations would be expected to have sm_ll
effects on lift and moment, and the effect of sid? scoops would depend on
the width of the body-scoop combination relative to She body diameter or
wing span. These trends are illustrated for subsonic speeds by the data
of references 234 and 277. The effect of body plum-form changes due to
the addition of scoops on the lift increment due to viscous crossflo_z
effects can be estimated by the method of reference ?_16.

_e lift and moment effects of scoops mounted on bodies in tests at

Mach numbers cf 1.5, 1.8, and 2.0 are described in references 220 and 239.
In the former _nvestigatlon, scoop locations on the top and bottom of a
body were compared. The scoop had a ramp_type compression surface and
was operated supercritlcslly. At zero angle of attack, the bottom location included a small positive lift foro_, and the top location induced
an equal negative ferce" the shift in the s_gle for zero llft from that
oi_the body alone was plus and minus 2, respectively. This difference
was maintained to an angle of attack of 8o; at _reater angles, the bottom

" "

NACA _4 A55F16

!'

"

_,0:_F_W3_A$

'

101

i
i
I_
"_

Nacelles.- The investigation of references 207, 231, 278, and 279

were of wi_g-nacelle combinations in which the nacelle inlet was at tae
?eading edge of both straight and swept wings and the nace]les extended
behind the trailing edges. As would be expected from such plan foxms,
the Aift-c_,rveslope and the stability of the combinations (basei on
wing d tmensions) were greater than those of the "_ug alone. The effects
of internal flo_ on lift-curve slope and maximum lift coefficieaxtwere

I,,,:.

show that
_panwise
position of
ward.
Thethe
results
of ref_rence27

,,_

_,

'

I i_
1 _-

i t

such a strut-mounted nacelle can be

_
Wingondb_dy--_\
_!_
selected so as to increase the lift
._
coefficient at which the slope af the _ .5
f==_"l
piSching-moment curve of a sweptb&ck
a
....

Wing_
wing reverses. Here, movi_ the
nacelle from 0.5 to 0.6 of "_hewing
semiapan changed the flo':r
n_er the
wing to such an extent that loss of
lift at the ti_s was delayed. As
shown in sketch (38),at flight M_ch
n_bers below about 0.9 this effect
_as large.
_

_
0a

balanced by mounting the nacelle from _

semispon
_
ward
a
vertical
nacelle
strut
locations
and moving
can be
itcounterdown__ .7I Nocelle
st0_----_
,
_

For nacelles that extend aheed of a wang, the lift on the projecting
body is destabilizAng. The magr_.tude c_ this effect for some nacelles in
subsonic flow is reported it.reference 81.
Some of the nacelles of this
reference were located Just below the wing; this position resulted in an
increase in the an__e for zero lift above that for the "_ug alone because
of the hi_ induced velocities on the lower wi_g surface in the region of
the wing-nacelle Juncture. This effect also changed the span loading of
the wing. The nacelles described in thiz reference did not change the
maximum llft coefficient attainable with the wing alone_ bu_ the liftcurve slope was increased as much as lO percent. This large increase was
due to the nacelles being tested o_!y on a short wing panel; on a complete
wing the increase in lift-curve slope would De of the order of 4 percent.
In reference 278 it is shown
that the destabilizing effec_ of for- _

"

locatio_ caused s_ incre =ing lit't-curves?ope, whereas the slope remained

nearly constant for the top location. Thzro_L the angle-of-attack range
of l0, the slope of the pitching..memeatcuratef_r th_ bottom ]ncation was
constant whereas that for the top location increased. This means that the
center of pressure moved rearward for the bottom scoop and forward for
the top scoop. The results of reference 239 for an underslung scoop confirm these trends. A reduetlon in m_ss-flow r_tio in these latter tests
frc_ 1.0 to 0.7 had no appreciable effecT.,
on lift or moment. ,

small iu the tests in which llft was measured, (:'efs.207 and 231).

f_

Nscellest la.,5
semispon

.5_

AR-B.5
,
I/4Ch_d sweep,47'
Ts_e_r0tio
0.2
I
B
1O
i_
_A,
Sketch (38)

'" ,""

!
/

102

....
i

: <r ,-

i
q

_tl

' TE , :
lit

NACA

The nacelles
tested at subsonic speeds as re_r_ed
in reference
278
and 39 were also tested at Mach n,mmbersof 1.6 and 2.0 as described in
reference_ ?_) an_ 28O.. These nacelles w_re mounted in several positions
on and below the chord plane of the sweptback wing at various spanwise
loc=tl_ms. _.._aerodynamic characteristics were similar to those at subsc_ic speeds; _hat is, all the nacelles increased the lift-curve slope,
the necelles _m the wing root increased stability but those mounted oatboard decress,
ed it. The magnitudes of the effects depend upon the specif2c confJ6uration, but they seldom exceeded lO percent of the lift or
m_nen_ of r_ wing alone.

A theoretical study of the lift of bodies and combinations of bodies

is presented in reference 282. Slender-body theory was used to predict
the i_terferenee of a fuselage on an open-nose nacelle downstream of the
intersection of the nacelle _._ththe f_elage bow-shock wave. In the
region of this intersection, slender-body theory is not applicable, but
account was msde of this by assuming that the reflection is that cf twodimensional flow, and the results of reference 261 substantiate this
assumption for weak shock waves. The _redicti_as of this method were
compare_ with e_periment as deszribed in reference 249. Here, t_,_tswere
perf_orme4at Math numbers of 1.8 and 2.0 with a slender fuselage having
open-nose nacelles mounted above and below in the pitch plane. Normal
force was _easured with the nacelles in several axis_ and gertlcal positions. With the nacelles slmost in contact with the fuselage, the sum of
the normal forces of the component bodies was as much as 25 percent greater
than the total measured normal force at an sngle of attack of 4. At
higher angle of attack, the normal force decreased to half the sum of the
component forces. This llft interference is, of course, due to the bodies
being in crossflow wakes. _oe theory proposed does not include all of the
factors involved in cros_flow and, depending upon relative location, predicted normal-force interference wizh errors from 0 to f5 percent of the
measured values. With the nacelle axis over 2 fuselage diameters from the
faselage axis there was no normal-force interference with_1 the limits of
angle of attac_ (8O) and axial spacing investigated.

!
:
!

"

The lift- and pitching-moment characteristics at Mach numbers of 1.5,

1.8, and 2.0 cf a canard eomfi_ratlon having one of the nacelle arrangement_ of reference 249 are described by Obery and Erasnow (ref. 283).
The nacelle axes were located one fuselage diameter from the fuselage
axi_ and the nacelle inlets were at 70 percent of the fuselage length
behind the nose. Since the nacelles were nearly half as long as the fuselage, _hey extended a considerable distance downstream and contributed a
stabilizing moment to the fuselage. Because _f lift _nterferenee due to
crossflow, the lift of the combination could not be accurately predicted
by the theory of r_fere_ce 282. _'.&ts model was also tested with the
nacelles in t_he hor_zonta! z_ther tha_the vertical plane (ref. 284). They
were plac=_ i-i/2 fuselage diameters from the center line an_ the inlets
were at abc_t the mld-l_ngth station of the fuselage. The increase in
lift-curve slope due to adding the n_;el!es (Z5 _ercent) was about twice
t

,,

.'

..,;

4F

"

NACA _4 A99FI6"

C_E_TIK_.

103

as gre:.t as when they were added in the vertical pl_ne, and the effect on
stability was not as great.
The addition of nacelles to the basic aircraft confi&mration
of references 266 and 267, howev_r_ resulted in large
percentage changes in pitchi_4_ moment.

The r.%uge
vsrious incombinations
investigated
in
references
239: performances
283. end 284 of
arethecompared
reference 19.
It w_ _o..nd
I

slightly _reater range than the configuration

wlth nacelles in the horithat the configuration %-Ith scoops which had the least minimum drag ha_.
zontal plane which had the least drag due to llft.
This evaluation depends,
of course, on the specific conditions assumed in the study.
The interference of a nacelle having a conical-shock
inlet o_erating
subcritically at a flight Mach number of 2.0 with an aircraft configuretlon
is described in reference 274. A comparison with the nacelle in a forward
and an aft location shows a decrease in lift-ctuT-e slope from 0.026 to
0o021 per degree s_d an increase in angle for zero lift from 0.9 to 1.'9

due to moving the nacelle from a locati_

forward below the body to ome
rearw_
and over th_ _-_ng.
the aerodynamic
center.

There was a corresponding

Ames Aeronautical
Laborato_j
National Advisory Committee
Moffett Field, Calif.,

for Aeronautics
June 16, 1955

f_d

shift of

J
!

104
......

....
g

RM

( g_

APPENDIX

S_4BOLS

area

Ac

capture

CD

drag ccefficient

Cf

skin-friction

CL

lift coefficient

Cp

pressure

arag

Dn

net

area

coefficient

based

on wett_

area

coefficient

diameter
Fn

engine

Fnp

net propulsive

_Fn

difference

gravitational

ratio of boundary-layer-displacement
5*

' "

"

net

th_st

force

between

ideal and actual

net thrus_,

Fni-Fna

constant
thickness

to momentum

thickness,_h

altitude

height

of boundary-layer

diverSer

length
MI

total momentum

Mach n_m_er

::

mass flow
maximum

,. .},r

of the ermine

mass flow, mM_PoVoA c

b i

,_

streamtube

in the inlet plane

"

NACA EM Ag_FI6

CC_I_]_SENTI_.:.

105
i

-_

engine

rotational

_I_._

number

of oblique

shock waves

pressure x

fuel

dynamic

Reynolds

gas constant

local

wing

temperature z

local

u,v,v

local velocity
respectively

stream

velocity

Wa

weight

flow of air

Wae

coz-rected weight

Wf

weight

x,y,z

Cartesian coordinates
direction

angle

angle of sidesl_p

ratio of specific

heats

relative

pressure,

boundary-layer

consm_ption
pressure
n_mber

radius
area or wetted

velocity

e_a

of flow, U2=(V+u)2+vm+w 2
components

flow

in the

x,y, and z

directions_

of air

flow of fuel
with

positive

in the stream

of attack

Pts

absolute

thickness

bounda_-la_r
displacement
thickness
,.
ZWhen used without the subscript
t, the symbols, p, p, and T denote
static pressure_ static demsit_t _ an@ static temperature_ respectively.
,
5*

'

.'

._

^,

_.

(( [

_ '

,,

cs , ,, ''"

angle of flow deflection

Tts
e

relative

boundary-layer

cowl angle, the _le

between the free-stream direction
and a line connecting the inlet _nd cowl maximum
diameters

_I,_2_s

funct-_ons defined
respectively

absolute

temperattt_e, T'S_

mc_entu_, thickness

by equations

(BS), (BT), and (B8),

mass density 2
cc_e

semiapex

cowl-posltion
compression

local

angle
angle, the _ngle between the apex cf a presurface and the cowl lip (see sketch (18))

shear stres_

shock-_ave

_ugle

Ae'
area ratio,

-_-

Subscripts

O_ X_2_2 t_S_I

,
|

denote

staticr_

actual

or additive

forebody

body

cone surface

ex

external

friction

hydraulic

@j,,.

mSee footnote

_t .page

in the flow

_-

as shown in sketch

diameter

I0_.

.......

,_'

....

.....

(I)

_ ._t,Aii

Nil ...........

j,,.

(._,
.,_ h*

!&_CA Y_4 A95FI6

CCNI'_I,

_ _.' _ : "

refers to -the plane enclosed

the inlet l_ps

in

internal

'.

ideal

jet
.

I07

by the sta_lati_n

points

on

lip

maximum

_net

ramp

shock wave

scoop

total

viscous

wave

cr

c:-iticel

isen

denotes

isentropic

SL

denotes

standard

flow

sea-level

conditions

Superscripts

( )*

denotes

conditions

where

( )

average

or effective

M = 1.O

value

-_ ,,!

llLlOiOg,ci_

"C__........__

m ill

al

__--

iiii

.o

............

Im

"

"

"Ilium

....

......................

--

,I ....

'_'

ligllg_mDIml_t.,_

i
I

108

__

CO_ID_D_L
c tt

_ (. tee
_

NACA EM A55F16
eg

APPENDIX B

_TS

AND INTERPEETATION

fn tests of air-_nduction systems accurate measurements must be made

T 4
_ _.__
pressure, _
_^"_. ratc, and drag. _,o_only must each
measurement be made accurately, but also the m_thod of data interpretation
must be one which best suits design purposes. Some of the considerations
involved are discussed in Lhis appendix.

MEASUREMENT ACCURACY

The accuracy of determining effective total pressure _ts f2om

measurements with a rake of total-pressure tubes depends upon the precision of each measurement of Pts" Pitot tubes alined with a subsonic
stream indicate true total pressure at the tube center Iine only when
the flow is uniform and steady. The information o! reference 28_ shows
that there is little error in measurement if the "_u'be
is alined wlthii00 of the flew direction if the bore ee a tube with a hemispherical head
is greater than about 0.3 the extern_-_ diameter. The study of reference o86 shows that when a tube is _-.a transverse total-pressure gradient,
the effective center of total presF . '_ is displracedtowards the region of
higher velocity by a small amount.
.is correction is negligible in the
testing of well-designed air-induc
.a sys+ -_sbecause sizable transverse
pressure gradients do not exist in _ _-ge_ort!on_ of the flow and the
pitot tubes are normally smal_ re.lativ_to L},_
_rea. Since duct
/

flow can often by unsteady, measurements unde_ _s _ ::._ifficns

are not
at all reliable. In reference 2_7 it is show_ that in incompressibie
flow the reading of a total-pressure tube alined with the mean stream
velocity V is

'i

V2

p +0

_,

"

(u
TM +

(B1)

where u, v, and w are the components of the turbulent fluctuations.

Thus, in unsteady flow the readings of pitot tubes are always _-eater
than the true value_ and calculations of effective total pressure, inten_al
drag._or mass flow based on the indication can be considerably in error.
(See also refs. 288 and 289.) The importance of this source of _rror is
indicated by the tests of reference _
in which measurements were.made
in the turbulent flow behind orifice plates. It was found that the
measurement of mean total pressure decreased with distance behind lhe
plate_ a trend to be expected frem the decay of turbulence. Errors in
the measurement of flow quantity of l0 to 19 percent resulted from readings

7
7
_+

"

NACA FM A55F16

.......
Ill

with pitot
i

r+l+i

C_ENTIAL
il

++ll

l+

tubes in this turbulent

some method
of indicating
unsteady
ments
in air-inducticm
systems.

109

_.

flow.

"

It i_ therefore

necessary

flow be used _'ith pitot-tube

that

measureD

In making measur_ .ants with a rake of pitot tubes, the number of tubes
which can be conveniently used is occasionally limited.
Under such clrcumstances the spacing of the tubes to give the most accurate average can
be chosen according to the metnod oi" Gauss, references 291 and 29_.
Integration must be performed accordlD_ to the Gaussian formula, which
requires more comp_,t_tion than do the norm_l methods,

+
.'+

A rake of pit )t au_ static tubes is used in induction-_ystem

testing
when ares- or msss-flow-av+_raged total pressures and flo_" uniformity are
to be measured.
Because of the errors which can arise and because of the
importance of ta_ mass-fio_ _ measurement
in detern_ming accurately net
drag and opt_-perfolmance
conditlons, it is a;_'sable to callbrate
rake installations
w__th a standard orlflse meter.
As a result of these
complications,
total pressure and m_ss flow are often determined slmply
from measurements
of static pressure at L_:o stations of different area
in the duct
If steady, one-dimenslonal,
isentroplc flow of a perfect
gas is assumed between the measurement
stations

Pt2

+
where
"

the subscript

+++

L(p,./p+) ++/+" CA-I/A+) +

2 refers

to the throat

<,,++>

...!

or minimum

s_ution

Hence,

the total pressure

Pt_
can be determined from measurements
of static
pressure and area at local stations 1 and 2. From knowledge of the total
tempers,ture and pressure, the static pressure, and the cros_-seetion
area
at a station, mass flow can be calculated from the formula

_--

I-,
me = PtmAm \

Tt _

./

\Pta/

" t,p-"_'-J .j

(B3)

<

These formulas involve assumptioas _nich often are not met in tests of
air-inductlon
systems, and again check calibrations
and careful consideration of sources of error are necessary.
(See, e.g., refs. 28_, 288, and
293)
The +uncertainty _,see ref. 293) in mass-flow measurement
is given
by the relationship

'"

__ ....

""

-'

....-+.c_,++:._-=+.,+.,+.+;,.,.7,.,..
' '-""-" ' ....
. i

"

" '

....
Z .....

Iiii ........
_

and is a function of Much number at the measurement station because

......
[(P/Pt)
2/7 -2__2

(B_)

Plott_-ug _l as a funcbion of Mach number shows that large errors in

mass-flo_ determination result from errors in measurement of static and
total pressure if the throat Much number is less than about 0.7. Similarly, the uncertainty in total pressure is

11/2

whet c

7(A,IA2)_(p,Ip_)'l=[__.
(p,lp_)_/"]
_2 ......

"" _ _
[(AilA2)_(pl/p2)il'r- i][

(pllu_)

(B7)

12i7 (AI/A2)2-

and

rIA IA _2'-_- _IITK_AIA '_- _- _2i_

[(AllA=>=CP.I_=,
"i7-

6__ __ _217+

1][(Al/A=,=(,:l./._=>
:l'=/" - 1]

._,
!
_

}
J
(Bs)

'_us, the error in total pressure i3 a function of the ratios AI/A2,

Pl/P2, and the component uncertainties in the area and pressure measuremerits. The variation of _Pt/Pt ariaamlm as a function of throat Much

i-,

':
,
.

!
,
L-

%,

..:.

"' "_i.... "_

....... ";_'i- _" _>::".........

_ ....

" "

I ""

'....... ,,_' ."

,_ ...........................................

',';.."_._.
.... ".......

' ;7.

5F

,_

NACA _M Ag_F].6

n,_mber M2 is illustrated in sketch (39) for an assumed error in staticpressure and area-ratio measurements of 1/2 percent. The uncertainties
are directly
for
other values
proportional
of the assumed
to theerror
errors
caninbe
these
determined
rstios,byand
simply
uncert_',
multi-

_QN_I_)E_IS_L-

Ill

plying by the proper factor. It is evident that _ contraction in area of

abgut 0.7 with a near sonic value of
}42 produces relatively great accuthrough a wide range of mass flow,
it is necessary to employ a var_-

a D2
raCY.able
throat.ln
order to maintain accuracy

_ D4 ____iA2

= 2

Tio

Three methods of deter_ining

effective total pressure at diffuser
exits are in co.non use.

None are

D6....

following equations for incompressible, two-dimensional flow with

st

,o

Method of equation (B2) (the

"Mass-Derived Method" of referexact. They are described by the
ence 29_)

or

_\_

_'

_
<_ _--_
"_02
,

(Bg)

.4

_\j

.6

IO

T_ _e_.M,

P-V
= P + f(u)
2

Sketch (39)
Area-weighting method
Pt
_

or

=P+_

p/U2dA

(BI0)

p'_ = p + f(_)

.
"

Pt =P+

dA

or

(BII)

_t =P+ _/US
f(U_)

M_.ss-f!owweighting method_
..

1
,.q

I!
/

[- ,

ll2

,_

-C(_NF.T._N'_$.L

......

NACA HM Ag_FI6

Since none of these can be substantiated by rigorous proofs as giving the

true effective tot?3 pressure, the question of accuracy must be settled
by comparison with a more exact estimate.
Such a compa_'ison is presented
by Wyatt in reference 295 where the more exact estimate is made by determining an effective ootal pressure which satisfies the momentum and contlnulty relationships
which are iavolve,i in cs/.culating engine thamst.
For uniform flow, all methods agree, but for nonuniform flow, such as
those which occur because of separation, the methods do not agree.
_%e
method of equation (B2) is, in general, the least accurate; but it i-equires
the simplest instrumentation,
for the other methods require a pitot-tube
survey.
Data reduction by the nmss-flow weighting method requlre_, the
most effort.
The area-weighting
method is usually as accurate a_ the massflow method, and it produces a conservative value of total-pres._ure ratic
which the mass-flow weighting method does not.
However, in the calculation
of the internal thrust of a wind-tunnel model, a conservative valise of
total press'&re produces too low an indication of net drag.
Under conditions which are normsJ_Ij enco'autered in well-designed
air-indaction
systems;
that is, relatively uniform steady flow, one method is as accurate as
another, and selection can be made on the basis of convenience
However,
for non_miform flow such as exists in ducts w_th bends, c_re must be
exercised in evaluating data.

_,

..

IL

NACA EM Ag_FI6

CC_IPT/)F_TIAL
',

"

EEFFRENCES

i.

S1_l"
in, M. A. : Aerodyn_nic Considerat_ons for the Design af Submerged
larboJet Power Plant Installations. I;_ Symposium on Flight Propulslon, Cleveland, Ohio, Mar. 16, 19'51. Pub. by Central Ai_'
Doc1_ents Office, Wright-Patterson Air Force Base, Ds.yton,Ohio,
May 19'51.

2.

E_uchemmnn,D., and Weber, J.:

Hill Book Co., 19.53.

3.

Anon. : Bibliography of NACA and Other Reports on Air Inlets and

Internal Flows. NACA BM 8J0'5,1948.

4.

Cushimg_ W. R. : An _malysls of Thrust Estimation.

vol. XXVII_ no. 311_ Jan. 1995, PP. 9-13.

9.

Quick, A. W.: Turbojet Intake Configurations.

pp. 679-684.

6.

Jacobsson, Bengt: Definition and Measurement of Jet Engine T_u_st.

Jour. Roy. Aero. Soc., Apr. 19'51,pp. 226-243.

7.

Kleln_ Harold: The Calculation of the Scoop Drag for a General Configuration in a Supersonic Stream. Rep. No. SM-15744, Douglas
Aircraft Co._ Santa Monica, Apr. 1990.

8.

BraJnikoff, George B. : Method and Graphs for the Evaluatioo of

Air-Induction Systems. NACA TN 2697, 19'52. (NACA Rep ll41, 1953)

9.

Sanders, Newell D., and Palasics, John: Analysis of Effect_ of

Inlet Pressure Losses on Performance of Axial-Flow Type T_rboJet
Engine. NACA EM E8Jf'sb,19_8.

Aerodynamics of Propulsion. McGraw-

Aircraft Eng.,

luteravia, Nov. 1949.

lO.

Hausent Frederick H., Jr., and Mossman, Emmet A. : Effect of Pressure

Recovery on the Performance Qf a Jet-Propelled Airplane.
NACA TN 1699, 1948.

ll.

Pahne, H. 0. : Approxfmate Methods to Transform Air Inlet Losses

Measured in Low-Spead Wind Tunnel to Var)-IngFlight Conditions.
SAAB TN 2, SAAB Ai_-craftCompany, 19'51.

12.

Wyatt, DeMarquis D : Aerodynamic Farces Associated With Inlets of

Turbojet Inst_.d.lst'.orm.
Acro. Eng. Rev., Oct. 1951, PP. 20-23, 31.

13.

Ferri, Antonio, and Nucci; Louis M. : Preliminary Investigaticm of

a New Type of Supersonic Inlet. NACA TN 2286_ 19_l. (Formerly
NACA BM L6J31)

"

113

qpmpmp.,.. .......

IIIImNmmwl --<-

.......

I]'1,
"

'

_q_

II

'

,_r

,{

:,

i.

'

"r:l

"

.......

.l

l._l

II

,._[,

l;_._

_;l_,

II

lit

Fradenburgh, Evan A., an_ Wyatt, DeMarquis D. : Theoretical Perf_.nmanceCharacteristics of Sharp-Lip Inlets at Subsonic Speeds.
_ACA T_ 3O04, 1993.
I_.

Optimization of Power Plant and Airplane Performance.

Aero. Eng. Rev., June 1954, pp. 42-61.

16.

Klein, Harold: Small-Scale Tests ok Jet Engine Pebble Aspiration.

Rep. No. _4.-1_89, Douglas Aircraft Co., Santa Momica, 1953.

17.

Rodert_ Lewis A., and Garrett I Floyd B. : Ingestion of Foreign

Objects into _"_rbineEngines oy Vortices. NACA TN 3330, 1995.

18.

Li_e11_ Keith: Foreign ObJe_ h Da_e

Tests - J39 _bo
Jet Engines.
AF Tech. Note WCLP-_2-83, Wrlght Air Development Ce_at__rl
Ohio2 1992.

19.

Fradenb_rgh, Evan A. 2 and Eremzler, Emil J. : Performance Comparison of

Three C4_ma_l-_jl_e
Ram-Jet Missile Configurations at Mach Numbers
from 1.5 to 2.0. NACA _ E93Fil, 1993.

20.

Woodworth, L. R. t and Kelb_rj C. C. : The C__neralizedApproach to

the Selectiun o,P Propulsion Systems for Aircraft. IAS Preprint
No. 349, I_DI.

__l. Meyer 2 C. A._ sad Fau_htt H.F.:

forma_ce of Turbojet Engines.
_2.

A Symposimn.

A Metbo; of Presenting the PerIAS Preprint No. 29.3,19[/0.

Wyatt_ DeMarquis D. : An Analysis of Turbojet-Engine-Inlet Matching.

NASA _ 30!_., 1953.

23. Moss_an_ Emmet A. _ and Anderson_ Warren E. : The Ef2ect of Lip Shape
o_ a Nose-Inlet Inztal!atio_ at Maeh Numbers fr_ 0 to 1._ and
a Methc_lfor Optimizing Engine-lnlet Combinations. NACA _4 AGUE08,
195_.
24.

Kr_azie_, Emil J. : A Method for Evaluating the Effects of Drag and

l_!et Pressure Recovery on Propulsion-Sy-t_ Performance.
_ACA _ 3_.
19_.

2_. Schueller_ Carl F., and Esenwet-, Fred T. : Analytical and Experimental
luvestigation ._fInlet-Emgine Matching for T_rboJet-Powered Air/j

_*
_

craft at Mach Numbers up to '2.0. NACA _4 E91/_20_19_2.

2.

A_s_,
Warren E., and Scherrer, Richard: Investigation of a Flow
D_flector and an Auxiliary Scoop for Improving Off-Design Performance of Nose Inlets. NACA _4 A94_06, 19_.

27.

Gr_aZa_.x_F. B. : Air Intake Efficieucy. Jour. Roy. Aero. Soc.,

vo!. 98_ no. 929, Sept. 199_ PP. 639-6_8.

AS F 6
.....'

n5

28.

Seddon, J. : Air Intakes for Aircraft Gas Turbines.

Aero. Soc., Oct. 1952, pp. 747-781.

Jour. _nv.

29.

Conrad, E. William, and Sobolewski, Adam E.: Investigation of

Effects of Inl_--Air Velocity Distortion on Performance of Turbojet Engine. NAI RM EOOGll, 1950.

30. Walker, Lewis E., Conrad, E. William, and Prince, William R.: Effect
of Uneven Air-Flow Distribution to _he _
Inlets of an _xialFlow Turbojet E_.&ine NACA RM E52K06, 1953.
31.

Nettles, J. C., and Lelssler, L.A.:

Investigation of Adjustable
Supersonic Inlet in Combination With J-34 Fmgine Up _o Mach
Number 2.0. NACA RM EShHll, 1954.

32. Huppert, ___erle

C. : Prelimlnsry Investigation of Flow Fluctuations
During Surge and Blade Row Stall in Axial-Flow Compressors.
NACA _ E52E28, 1992.
33-

Graham, Robert W., and Prise, Vasily D. : Experimental and Theoretical Investigation of Rotating-Stall Characteristics of SingleStage Axial-Flow Compressor With Hub-Tip Ratio of 0.76.
NACA _ E53109, 1953.

34.

Huppert t Merle C., Johnson, Donald F., and Castilow, Eleanor C.:
Preliminary Investigation of Compressor Blade Vibration Excitad
by Rotating Stall. NACA _ E52J15, 1952.

"

35. Walker, Curtis L., Silvo, Joseph N., and Jansen, Emmert T.: Effect
of Unequal Air-Flow Distribution from _win Inlet Ducts on Performance of an Axial-Flow Turbojet Engine. NACA i_4E54EI3, 1954.
36.

Conrad, E. William, Schulze, Frederick W., and Usow, Karl H.:

Effect of Diffuser Design, Diffuser-Exlt Velocity Profile and
Fuel Distribution on Altitude Perfm_mance of Several Afterbuxner
Configurations. NACA _ E93A30, 1993.

37.

Connors, James F.: Effect of Ram-Jet Pressure Pulsations on Super- /

sonic Diffuser Performance. NACA PM EgOH22, 1950.

38.

Sterbenz, Willi_n H., and E_srd, John C. : Criterions for Prediction

and Control of Ram-Jet Flow Pulsations. NACA _ E51C27, 1991.

39.

Shapiro, Ascher H., and. Smith, R. Douglas.: Frlction Coefficients

in the Inlet Length of Smooth Round Tubes. NACA ._'1785, 1948.

40.

Dodge, Russell A., and Thompson, Milton J.:

McGraw-Hill Book Co. _ Inc., 1937.

j'

....
.
, ,-

Flui_ t_ech_icso
i

........ _ ............................

_._-_

/
n6

,,
.",.,"_,',
dO_D_,_ i'! ,."
.: _cA_ A55FI6

41.

Monaghan, R. J. : Some Remarks on the Choice and Pre_entatlon

Formulae for Turbulent Skin Friction in Compressible Flow.
R.A.E. _3_ No. Aero. 22469, British A.R.O., May 1953.

42.

Chapman, Dean R., and Kester, Robert H. : Turbulent Boundary-Layer

and Sk_u-Friction Measurements
in Axial Flow Along Cylinders at
Between
q_
_
%x^
_A1 3097,
_=I,
_'_ch _,.'nn,,Dcrs
_" "
v._
and 3.6.
._CA^ _"
_-

43.

Clauser, Francis H. : Turbulent Boundary Layers in Adverse

Gradients.
Jour. Aero. Sci., Feb. 1954, pp. 91-108.

_4.

Beeton, A. B. P. : Curves for the Theoretical

Skin Friction Loss
Air Intake Ducts.
R.A.E. TN No. Aero. 2035, British A.R.C.
Feb. 195o.

45.

Shapiro, A. H., H_wthorne, W. R.: and Edelman, G. M.:

The Mechanics
and Thermodynamics
of Steady One-Dimensional
G_s Flow With Tables
for Numerical Solutions.
M.I.T. Meteor Re_._ ..'_":)ec. i, 19_7.

46.

IuskJ_, H., and Klein, H. : High Speed Aerodynamic

Problems of Txrbo_^_ In_:tai3ations.
(Contributed by the Aviation and Gas Turbine
Power Divisions for presentation
at the Annual Meeting of the
A.S.M.Eo, New York, N. Y., Nov. 26-Dec. i, 1950, A.S.M.E. Paper
50-A-I02, 1950.
(Also available as Rep. No. _qM-13830, Douglas
Aircraft Cc._ Santa Mop, ca, Sept. I, 1990)

47.

_//%8.

Von Doenhoff; Albert E._ and T_uervin, Neal:

Determination
Relations for _he Behavior (,f Turbulent Boundary Layers.
Rep. 772, 15:%3. (Formerly I_ACA ACR 3G13 and WR L-382)

"

of

and

Pressure

in

of General
NACA

L._ttle,
B. H.,
Jr,
_.lqdWilbur,
Stafford W.:
High-Subsonic
formance
Charac:.eri._tics
and Boundary-Layer
Investigations Perof
a 12 10-Inch-inlet-Diameter
Conical Diffuser.
NACA Y_4 LSOCO2a,

_990.

49.

Fage, A., and Raymer, W. G. : Note on Empirical Relations for a

Turbulent Boundary Laye_
R. & M. No. 2_59, British A.R.C., 1948.

50.

Ferri, Antonlo_ _nd Nucci, Louis M. : The Origin of Aerodynamic

Ins _ability of Supersonic Inlets at Subcritlcal
Conditions.
NACA _4 L_0K30, 1951.

51.

Nettles, J. C. : The Effect of Initial Rate of Subsonic Diffusion

on the Stable Subcrltical Mass-Flow Range of a Conical-Shock
Diffuser.
NACA _4 E53_96, 1953.

92.

Kantrowitz, Arthur:
The Formation and Stability
Waves in Channel Flow.
NACA TN 12_3, 1947.

of Normal

Shock
"

.....
_,g'5

---....
..............

LJI

.....

:".

_.

,;-,_-c.-;_-;-

--

'"' ''_ '

[]

,"......
''

_ ' _-_.'II

'

'"

NACA _M A55FI6

CQt_LS__._AL ,,, .....

117

53.

Lukasiewicz, J. : Diffusers for Small Supersonic Mach Numbers;

Design Data. R.A.E. TN No. Aero. 1973, (SD 84), British A.R.C.,
1948.

54.

wyatt, DeMarquis D., and Hunczak, Henry R. : An investigation of

Convergent-Divergent Diffusers at Maeh Number 1.85. NACA I_4ESOK07_
1991. (Formerly NACA F_4E6K2_I)

55.

Copp, Martin R., and Klevatt, f'aulL.:

Investigatisn of High-Subsonlc /_

Performance Characteristics of a 12 21-1nch Conical Diffuser,

_/
!nclud_
the Effects 3f C_Dge ___Inlet-Boundary-Layer Thickness.
NACA _4 L9KiO, 1950.
56.

Naumann: Wirkungs_'ad yon Diffusoren bei hohen Unterschallgeschwlndigkelten. FB Nx 1705, Luftfahrtforschung (Berlln-Adlershof), 1942.

57.

Copp 3 Martin R.: Effects o_ Inlet Wall Contour on the Pressure

Recovery of a l0 lO-Inch-Inlet-Diameter Conical Diffuser.
NACA RM LSiElla, 1951.

58.

M moni, L. R.: Wide Angle Diffusers Employing Boundary Layer Control.

Rep. R-94960-12, Contract AF-SB(038)-elS06, Res. Dept., United
Aircraft Corp., Mar. 19, 1952.

59.

Taylor, H. D. : Summary Report on Vortex Generators.

United Aircrs__tCorp., 1950.

60.

Patterson: G. N. : Modern Diffuser Design.

no. 1i3, Sept. 1938, PP. 267-273.

61.

Henry, John R.: Design of Power-Plant Installations. PressureLoss Characteristics of Duct Components. NACA WR L-2(kS,June 19h/_.
(Formerly NACA ARR L_F26)

62.

Squire, H. B.: Experiments on Conical Diff1_ers. R & M No. 2751,

British A.R.C., Nov. 1950. (Formerly BAE _ No. Aero. 2216)

63.

Garuer, H. C.: Development of Turbulent Boundary Layers.

No. 7133, British A.R.C., Jure 1944.

64.

Scherrer, Richard, and Anderson, Warrer E. : Prel_min.%ryInvestigation of a Family of Diffusers Designed for Near-Sonic Inlet Velocities. NACA _ 3668, 1956.

65.

Nelson, William J. and Popp, Eileau G. : Performance Characteristics /

of Two 6 and 12_ Diffusers at High Flow Rates. NA_ _M LgH09, 1949.

66.

Jo_n%ston,I. H. : The Effect of Inlet Conditloms on the Flow in

Annular Diffuserz. Memo No. M.167, Nat. Gas _%_rb_meE_tab.,
_u. 1993. (Also available as M0_ C.P. No, 178, British A.R.C. )

Rep. 05280-9,

Aircraft Eng., vol X,

R.&M.
v.

"

-., .

/
!

118

,.......
,
, ,

_ ',C0NFTffEN_/AL.......
, , ,
, ,

NACA ._4A99FI6

6'[. Valerlno, Alfred S : Effects of Interual Corner Fillets on Pressure

Recovery - |v_as_
Flow Characteristics of Scoop-Type Conical Supersonic inlets. NACA EM E92JIO, 1992.
68.

Scherrer, Richard, and Gowen, Forrest E. : Prelimlnsu_ Experimental

Tmv_stlgatlon of a Varlable-Area, Variable Internal-Contractlon
Inlet at Mach Numbers Between 1.42 and 2.44. NACA RM A99F23. 1995.

69.

Weske, John R.: Experimental Investigation of Velocity Distributions

Downstream of a Single Duct Bend. NACA _ 1471_ 1948.

70.

Rogallo, F. M. : Internal-Flow Systems for Aircraft.

1941. (Formerly NACA TN 777. )

71.

Piercy, Thomas G., and Welnstein, Maynard I. : Preliminary Iuvesttgation at Mach Number 1.9 of Simulated Wing Root Inlets. NACA RM
I_4E94124, 199_

72.

Blackaby, James R., and Watson, Earl C. :

NACA Rep. 713,

An Experimental Investiga-

tion at T_._Speeds of the Effects of Lip Shape on the Drag and

Pressure Recovery of a Nose Inlet in a Body of Revolution.
NACA T_ 3170, 199)4o

"

73.

Milillo, Joseph R. : Some Internal Flow Charm'teristics of Supersonic Inlets at Zero Flight Speed. NACA RM L_4E39, 19_4.

74.

Scherrer:,Richard, Stroud, John F., _nd Swift, John T.: Preliminary

Investigation of a Variable-Area Auxiliary Air.Intake System a_
Mach Numbers From 0 to 1.3. NACA _M A93A13, 19_3.

i.

"

7.5. Bryan, Carroll R., and Fleming, Frank F. : Some Internal-Flow Characteristics of Several Axlsymmetrlcal NACA 1-Serles Nose Air Inlets
A_ Zero Flight Speed. NACA I_4L54E19a, 1994.

"

76.

Pendley, Robert E., Milillo, Joseph R., and Fleming, Frank F. : An

Investigation of Three NACA l-Sel-iesNose Inlets at Subsonic
and Transonic Speeds. NACA RM L9_23, 1993.

/__

BraJnikoff, George B. _ and Stroud, John F." Experimental Investigatiom of the Effect of Entrance Width-tc-Height Ratio on the Performance of an Auxiliary Scoop-_.,10e
I_iet at Me.ohNumbers _rcm
0 to 1.3. NACA EM A5_.28; !953.
78.

Cortright, Edgar M., Jr. : Prel_mlnsry Investigation of a Translating

Cowl Techniqu_.for Improving Take-t_'fPe:_formanceof a Sharp-Lip
Supersonic Diffuse_-. NACA I_4E51"24,1991.

--

.....
TT

....

L1
_

I_

'

'

IIIIIII
I

III

II IIIII

_'

L6F

NACA RM A55F16

CO.NFIIJENTi_L

79.

jJ

ll9

J,

Blackaby, James R. : Low-Speed Investigation

of the Effects of Angle
of Attack on fhe Pressure Recovery of a Circular Nose Inlet With
Several Lip Shapes.
NACA TN 339 h , 1955-

_
_,

80.

Baals, Donald D., Smith, Norman F., and Wright, John B. : The
Development
and Application
of High-C_itical-Speed
Nose l/fiets.
T A A
_._C_ Ecp. r.or, 1948.
(Formerly "^_^ ACR, LSF3Oa)

81.

Danneabe_g,
Robert E. : The Development
of Jet-Engine
a High-Spe_d Pother Design.
NACA RM ATDIO, 1947.

82.

Seddon, J., and Trebble, W. J. G. : Experiments

on the Flow Into
a Swept Leadlag-_ge
Intake at Ze::o Fc_rward Speed With Notes on the
Wider Uses of a Slotted Intake. R.& M. No. 2909, British A.R.C.,
1954.

83.

Trescot, Charle_ D., Jr., and Keith, Arvid L., Jr. : Investigation
at Transonic Speeds of the Aerodynamic
Characteristics
of a Se_iCircular Air Inlet in the Root of a 45 Sweptback Wing.
NACA }94 L95AOSa, 1999.

84.

Howell,

Robert

Transonic
Installed

R., and Kelth,

Arvid L. Jr. :

Nacelles

An Investigation

for

at

Speeds of the Aerodynamic

Characteristics
of an Air inlet
in the Root of a 45
Sweptback WLng.
NACA F_4 L52H_,Sa,

1952.
85-

He_ell, Robert R., and Trescot, Charles D., Jr. :

Transonlc Speeds of Aerodynamic
Characteristics
Air Inlet in the Root _f a 45 Sweptback Wing.
1953.

86.

Ferri, Antonio:
The MacMillan

87.

Lukasiewicz,
J. : Supersonic
A.R.C., 1946.
(Originally

88.

Kantrow_tz, A., and Donaldson, C. du P.:

Preliminary
I/Ivestigatlon
of Supersonic Diffusers.
NACA WR L-713, 1945.
(Formez-ly D_kCA

Elements of Aerodynamics
Co., New York, ]9'_9-

Investigation
at
of a :_emielllpbical
NACA i_4 L53J22a,

in Supers_lic

Flows.

._

Diffusers.
R. & M. No. 2901, British
R.A.E. GAll 8, June 1946)

ACRLSD O)

"

....
,_ , ,

89.

Clauser,
Rocket
ll2.

90.

0swatitsch, K. : Pressure Recovery for _lissiles with Reaction

Propuls_oa at High Supersonic Speed; (TToe EfflcieD_y of Shock
Diffusers).
NACA i_4 1140, 194 _I.

L_

F. H.:
Ramjet Diffusers at Supersonic
Soc.. vol. 24, no. 2, Mar.-Apr., 19_4,

/[
,

II12_

c __

II
,

_LJ
,

I-_

L.

,,. ,._,-r4._
_

_{

Spe_ds, Jour. Ame_.

pp. 79-84, 94, and

,"

I
,

II

I I
,

.// III ,,,.

120

,,_

,:c__

.......
......

NACA

RM

A55F16

91.

Moeckel, W. E., mud Connors, J.F.:

Charts for the Determination
of Supersonic Air Flow Against Inclined Plane._ and Axially Symmeb-lc
Cones.
NACA TN 1373, 1947.

92.

Staff of A_es Aeronautical

Laboratory:
Equations_
for Com.,ressible Flow.
NACA Rep. 1139 1953.

93.

Moeckel, W. E., Col_%oi's, J. F., and Schroed_ _, A. H.:

inv_Ligation
of Shock Diffusors at Maca Numbe_ 1.85.
I.
Projecting SingleShock Cones.
NACA EM E6):27, 1947.

94.

Moeckel, W. E., and Connors, J.F.:

Investigation
of Shock Diffusers
at Mach Number 1.89.
III. Multiple-Shock
and Curved-Contour
Projecting Cones.
NACA PM E7FI3, 1947.

99.

Evvard, John C., and Blaaey, John W. :

for Efficient Supersonic Diffusion.
(Formerly NACA RM E7C26. )

96.

Madden, Robert T., and Kremzier_ Emi2 J. : Force and Pressure Charateristics for a Series of Nose IDlets at Mach Numbers frcm 1.59 to
1.99. IV. Co_ical-Splke External-Internal
Compression Inlet
Utilizing Perforated Cowl.
NACA RM ESIB09, 1951.

97.

McLafferty,
George:
Development
of the M_lti-Unit Perforated Diffuser
for Operation at Mach Number 2.0.
Rep. R-9348h30, Res. Dept.,
United Aircraft Corp., Sept. 1952.

98.

Oblinger, Fred G_:

Report of Lift, Drag, and Pitching-Moment
Tests
on the 8.14 Percent Scale Grumman XSSm-N-6a Rigel Tactical .Missile
Model TM-8 e.t M_ch Number 2.0_.
Rep. 24-1-9, Eng. CTR., Univ. of
Southern Calif., Sept. 8, 1993.

99.

Tables,

i_

aud Charts

The Use of Perforated

NACA P_4 ESiB10, 1951.

Inlets

'

Neice, St_ford
E. : A Method for Stabilizing Shock Waw_._ in Channel
Flow by Means of a Surge Chamber.
_ACA TN 2694, 1953.

i00.

C,_menzo, Raymond J., and Mackley, Ernest A. : Preliminary Investigation

of a Rectangular
Supersonic Scoop Inlet with Swept Sides Designed
for Low Dr_E at a Mach Number of 2.7. NACA RM L52j02, 1952.

101.

Fexri, Antonio:
Some Recent Advances in the Design of Supersonic
Diffusers.
Supersonic Inlet Symposium, jan. 23, 195_.
R_p. No.
1692, Wr_nt
Aero. Div., Wood-Ridge,
N. J., pp. F-1 - F-18.

lO'_.

Connors_ J. F., and Woollett, R.R.:

Characteristics
Axlally-Sy_aetric
Isentropic Spkes for Nose Inlets

'

of FLow About
at Mach Number

3._. _ACA_ Eg_F08,

199_.

_ ";'_

' '<--'"T_-""-

"

..............

P_.[_

ilil

i II

"

I--

Ill

]111 ......

R_CA RM A55F16

C0_IDEI_ISLiL

121

i03.

Holder, D.W., Pearcey, H. H., and Gadd, G. E.:

Between Shock wav_:s and Boundary Layers, With
Effects of the Interaction
on the Performance
Intakes By J. Seddon.
No. 16,526 (Perf. 1199;
Performance S_o-Committee,
British A.R.C., Feb

104.

Nut,doffer, T. J. : Some Obsezcations

Separation on Supersonic Diffusers.

105.

Bogdonoff, S. M., and Kepler, C.E.:

Separation of a Supersonic
Turbulent Boundary Layer.
Rep. 249, Dept. of Aero. Eng.,
Princeton Univ. ; Js_i_ 1954.

106.

Gadd, G. E., Holder. D W., and Regan, J.D.:

The =3perimental
Investigation
ol the Interaction
Between Shock Waves and Boundary
Layers.
I_oc. of the Roy. Soc., Ser. A., vol. 226, Noo. 1954,
pp. 227-253.

107.

Gadd, G. E., and Holder, D.W.:

Further Remarks on interactions
Between Wholly Laminar or Wholly Turbulent Boundary Layers and
Shock Waves Strong Enough to Cause Separation.
Jour. Aero. Sci.,
vol. 21, no. 8, Aug. 1954, pp. 571-572.

108.

Dailey,
Univ.

109.

Nitzberg, Gerald E., and Crandall, Stewart:

Some Fundamental
Similarities Between Boundary-Layer
Flow at Transonic and Low
Speeds.
NACA TN 1623, 1948.

ii0.

Drougge, Georg :
Da% Experimental
Investigation
of the Influence of
Strong Adverse _Tessure Gradients on Turbulent Boundary Layers
at Supersonic Sp_:eds. Fly_teknlska Forsoksanstalten,
Stockholm.
Meddelande 46, 1952.

iii.

0bery, Leonard J.; Englert, Gerald W.: and Nussdorfer,

Theodore J.:
Pressure R_covery, Drag, 8m.d Subcritical
S_ability Characteristics
of Conical Supersonic Diffusers Wlth Boundary-Layer
Removal.
NACA RM ESIH29, 1952.

112.

Campbell, R_'"-rt C.:

Performance
Internal Bound_<_ _Layer Scoop.

113.

Piercy, Thomas G. : Prelimin.ary Investigation

of Some Internsl
Boundary-_yer
Control Systems on a Side inlet at Mach Number
2.96.
NACA RM E54K01, 1955.

The Interaction
a Note on the
of Supersonic
FM 2017),
2, 1954.

of Shock-Induccd
NACA RM ESIL26,

C. L. : Diffuser Instabiiitv
in Subcritical
of Southern Calif., Sept. 26, 1950.

of a Supersonic
NACA RM E54101,

_arbulent
1954.

Operation.

Rsmp Inlet Wlth _//

1954.

....

._I

il

ii

.......
I

ii I ii

i _I

i i i

NACA _4 A55FI6
_ i

Davis, Wallace F., and Brajnlkoff,

George B.:
Pressure Recovery
at Supersonic Speed_ T_rough Annular Duct Inlets Situated in a
Region of Appreciable
Boundary Layer.
I - Addition of Energy
to the Bou_ndary Layer.
NACA FM ASA13, 1948.

llS.

0bery, I,. S., and Cubbison, R. W. : Effectiveness

of Boundary Layer
Removal Near the Throat of Ramp-Type Side Inlets at Free-Stream
Mach Number 2.0.
NACA RM E54114, 1954.

ll6.

Davids, Joseph, and Wise, George A.:

Investigation
at Mach Numbers
1.5 and 1.7 of Twin-Duct Side Intake Sy=:tem w_th Two-Dimensional
6o Compression
Ramps Mounted on a Supersonic Airplane.
NACA
EM ESBKI9, 1953.

ll7.

Clark, D. B., s_d McCrea, J. W. : The Development

of an Inlet from
Drawing Board to Flight.
Supersonic Inlet Symposium, Jan. 23,
1933.
Rep. No. 1692, Wright Aero. Div., Wood-R_dge,
N. J.
pp. A-1 - A-18.

ll8.

Clark, D. B., and Carlson, P. G. : Inlet Aerodynamics

as Related co
Ram Jet .Uontrol. Supersonic Inlet Symposium, Jan. 23, 1953.
Rep. No. 1692, Wright Aero. Div._ Wood-Ridge,
N. J., pp. B-I - B-8.

,_/z ll9.

Allen, J. L., and Beke, Andrew:

Performance Comparison at Supersonic
Speeds of Inlets Spilling Excess Flow by Means of Bow Shock,
Co_crl
Shock or Bypass.
NACA RM E53Hll, 1953.

120.

A13en J. L., s_d Beke_ Andrew:

Force and Pressure Recovery Characteristics
at Supersonic Speeds of a Conzcal Spike Inlet with
Bypasses Discharging
in an Axial Direction.
NAC_ RM E52K!4, 1953.

121.

Gorton, Gerald C. : Investigation

of Translating-Spike
Supersonic
Inlet as Means of M_ss-Flow Control at Mach NtLmbers of 1.5, 1.8_
and 2.0. NACA RM E53G10, 1953.

122.

Gorton, Gerald C. : Investigation

at Supersonic Speeds of a Translating Spike Inlet Employing a Steep-Lip Cowl.
NACA 154E54G29, ].95_.

123.

Gorton, Gerald C., and Dryer,

of T_islatlng-Spike
i_lets
NACA RM E54J07, 1955.

124.

Leiss_er, L. Abbott, and Nettles, J. Cary:

Investigation
to Mach
Number 2.0 of Shock-Posi_icnlng
Controls for Variable-Geometry
Islet in Cc_mbination With _ JR4 Turbojet Engine.
NACA BM E54127,

125.

Stitt,
_onard
Double-Ramp

"

:C_N_EIA_

.i

ll4.

:
iI

........

'

Murry:
Comparison at Supersoni'.. Speeds
}L_ving Blunt- and Sharp-Lip Ccwls.

E., Inlets.
and Wise, NACA
George
A. : Investigation
Side
RM E54D20,
1954.

of Several

NACA RM A55FI6

126.

CONFIDE_f,/AL

123

Pennington, Donald, Rabb, Leonard, and Simpkinson,

formsa_ce of a Double-Zone
Inlet With sm.d Without
Design Mach Ntnnbers. NACA RM E54L27, 1955.

Scott H. : PerShroud at Below-

127.

Comenzo, Raymond J.: A Preliminary Investigstifm

of the Pressure
Recovery of Several Two-Dimensional
Supersonic Inlets at a Math
N'_b_r of 2.01. PL_C& P_ LSIIDIII,1951z.

128.

l.eissler_ L. Abbott, and Sterbentz, William H. : Investigation

of a Translating-Cone
Inlet at Mech Numbers From 1.9 to 2.0.

L-_

I
129.

Inves

Angle
of Attack.
NACA RMa E_3130,
of a Diffuser
Employing
Pivoted

ii
Ii
:--_
I!

130.

_
131.
.

]993.
Cone to Improve

Mach Number
Operation

NACA RM ES_E!4,

132.

I
1

Carter, Howard S., and Merlet, Charles F. : Preliminary Investigation of the Total-pre_;sure-Recovery
Characteristics
of a Symmetric

tion of an Asymmetric SWept Nose Inlet of Circular Projection

Co,mors, James F., and Woollett, Richard R.:
Preliminary
Investigaat a Mach Nunber of 3.89.
NACA RM E94026, 19,5%.
Beheim, Milton A. : A Preliminary
Investigation
at Mach Number 1.91
of an Inlet Configuration
Designed for Insensitivity
to Positive
A_Igle-of-Attack Operation.
NACA EM E53E20, 1993.

134.

Young, A. B. : The Calculation

of the Profile Drag of Aerofoils
and Bodies of Revolution at Suoersonic Speeds.
Rep. No. 73,

13_.

Cza_mecki, K. R., and Sinclair, A. R.: Factors Affecting

tion at Supersonic Speeds.
NACA _ L93118a, 1993.

136.

Hilton, John H., Jr., and Czarnecki, K. R. : An Exploratory

Investigation
of S?-in Friction and Transition on Three Bodies
of Revolutiol_ at a Mach Number of 1.61.
NACA TN 3193, 19_4.

137.

Purser, Paul E. ; Comparison of Wind-Tunnel

Rocket, and Flight
Drag Measurements
of Eight Airplane Configurations
at Mach
N1_mbers Between 0.7 and 1.6.
NACA HM L54F18, 1994.

College
:_

of Attack

133.

.;
:

at

1952.

and an Asymmetric Nose Inlet Over a Wide Range of Angle

_t Supersonic P_ach Numbers.
NACA RM L93J30, 1993.

1.91

Leissler, L. Abbott, and Hearth, Donald P. : Preliminary Investigation

of Effect of Angle of Attack on Pressure Recovery and Stability
C_haracteristics for a Vertical-Wedge-Nose
Inlet at Mach Nut bet
of 1.90.

I;
,_

NACA RM E94B23, 1994.

Beheim, MiJ.ton A. : A Preliminary

of Aeronautics,

Crar_field, EngLo_nd, Apr.

1993.
,k'snsi-

_ _:

;i

: '",
_
!

I
8

.
t
I

It

ill

_tte

_t

II

;_p

NACA HM

'

136.

Owens, Billy F., and Curtis, Thomas H.: Phase II Flight Tests
of the Lockheed F-94C Airplane, USAF No. 50-596. Tech. Rep.
52-14, Edwards Air Force Base, Aug., 1952.

13_.

Redd, Joseph W., Tr., and Stephens, Robert L. : Phase IV Performance

Flight Tests ef the F-86D Aircraft USAF b/N 50-459. Tech Rep.
53-26, Edwards Air Force Base, (1953)

3_.

Phillips, ,if red D., and Stephens, R. L. : Phase II Performance

and Stability Tests of YHF-84F Airplane.USAF No. 51-1828.
Tech. Rep. 52-32, Edwards Air Force Base, Jan. 16, 1953.

141.

Redd, Joseph W., Yeager, Charles E., and Everest, Frank K.: Perfarmance Flight Tests of the XF-92A Airplane, USAF S/N 46-682
with a J33-A-29 Power Plant. Tech. Rep. 53-11, Edwards Air Force
Base, Mar. 1953.

142.

Yaucey, Marion H., Jr., and Carson, James S.. Phase IV Performance
Flight Tests of the F-_gF &Ir_lnne USAF No. 52-4349. Tech. Rep.
54-10, Edwards Air Force Base, May 1954.

14g.

Wesesky, John L., and Stephens, Robert L.: Phase II Flight Tests
of the YF-102 Airplane S/N U._kFNo. 52-7995. Tech. Rep. 5;_-14,
Edwards Air Force Base, July 1954.

144.

Fertel, P. F-105: Areas, O_mensions; and Data on 1/22 Scale Model.

Rep. EAR-311, Republic Aviation Corp., Farmingdale, N. Y.,
Feb. B, 1954.
L

145.

Green. M., and Carter, Wayne: Model XF7U-1, FTU-1 Airplanes,

Summsry of High Mach Number Characteristics. Rsp_ 8198 Engr.
Dept., Chance Vo1_ht Aircraft, Dallas, Aug. 24, 1950.

II
"
:

146.

Grlggs, C. F., and Goldsmith, E. L.: Measurements of Spillage Drag

on a Pitot Type .Tutakeat Supersonic Speeds. R.A.E. TN No.
Aero. 2256, (British), lO53.

147.

yon Karman, Th., and Millikan, C.B.:

On the Theory of Laminar
Boundary Layers Involving Separation. NACA Bep. 504, 1934.

148.

Sears, R. I.. Merlet, C. Fo, and Putland, L. W.: Flight Determination of Drag of Norm_1.-SnockNose D&lets With Various Cowling

',

Profiles at Mach Numbers from 0.9 to 1.5.

149.

,_

|:-

NACA RM L5312_a, 1953.

Nitzberg, Gerald E., and Crandall, Stewart: A Study of Flow Changes

Associated With Airfoil Section Dr_g Rise at Supercri+ical Speeds.
NACA TN 1833, 1949.

'

--_

._

...==..--

,,_',

4
Lp_l_...+._

.
...._,,.%

....._

'J,,,

_ _

JJ

--

--

Ill lilt

lilt

Ill ___

: _:_,

. _

....

,--

J
.

NACA RM AS_F16

CONFIDENTIAL

125

150.

Waiters, Richard E. :
Sharp-Lipped Ducted

Application
of Transonic Area Rule to a
Nacelle.
NACA _4 L53JOb,
1954.

151.

Hall, James R. : Comparison of Free-Flight

Measurements
of the ZeroI ift Dra_ Rise of Six Airplane Configurations
and Their Equivalent
Bodies of Revolution at Transonlc Speeds.
NACA RM L53J21a, 1954o

152.

Brown, Clinton _., and P&rker, Herman H. : A Method for the Calculation of Ex'4er0sl Lift, Moment, and Pressure Drag of Slender OpeuNose Bodie_ on Revo]utiou at Supersonic Speeds.
NACA Rep. 808,
1945.
(Formerly NACA ACR LSL29)

153.

Lighthill, M. J. : Supersonic Flow Past Bodies

R. & M. No. 2003, British A.R.C., 1945.

154.

Lighthill, M. J. : Supersonic Flow Past Slender Bodies of Revolution

the Slope of Whose Meridian Sectiou is Discontinuous
Quart.
Jour. Mech. and Appl. Math., vol. I, pt. l, Mar. 1948, pp. 90-102.

of Revolution.

155.

Ward, G. N. : The Approximate

Exterior and Interior Flow Past a
Quasi-Cylindrlcal
Tube Moving at Supersonic Speeds.
Quart.
Jour. Mech. and Appl. Math., vol. I, pt 2, June 1948, pp. 225-245.

156.

Ward_ G. N. :
Jour. Mech.

i.

Supersonic Flow Past Slender Pointed Bodies o Quart.

and App1. Math., vol. II, pt. l, Mar. 1949, pp. 75-97.
E

157.

Jack, John R. : Theoretical Wave Drags and Pressure Distributionc

for Axially S3_mnetric 0pen-Nose Bodies.
NACA TN 2115, 1950.

1_8.

Moore, Franklin:
Linearized Supersonic Axially S_mmetric Flow
About Open-Nosed Bodies Obtaintm by Use of Stream Function.
NACA TN 2116, 1950.

159.

Ferrari, Carlo:
Determination
of the Exterior Contour of a Body
of Revolution Wi-h a Central Duct so as to Give Minimum Drag in
Supersonic Flow, With Various Perlmetral Conditions Imposed Upon
the Missile Geometry.
Rep. No. AF-814-A1 (Contract No. N6 ori-ll)
Cornell Aero, Lab., Inc., Buffalo, Mar. 1953.

160.

Ferrari, Carlo:
Determlnstion
of the External Contour of a Body of
Revolution with a Central Duct So As to Give Minimum Drag in
Supersonic Flow, with Various Perimetral Conditions imposed
Upon the Missile Geometry.
Part III- Numerical Applications.
Rep. AF-814-A-2, Cornell Aeronautical
Lab., Inc., Buffalo,
Nov. 1953.

161.

Bolton-Shaw,
B. W., and Zienkiewicz,
H. K.:
The Rap_d, Accurate
Prediction of Pressure on Non-Lifting
0glval Heads of Arbttrsmy
Shape at _upersonic Speeds.
British ARC CP 154, 1954.

"'J

/
J

=i!!
::
162.

.....

Parker_ Eer_m H. : _-_

Revolutix
Based _ Linearized

i:i
. .

Ducted and Bointed

Supersonic
Theory.

Bodies c_
NA_ TN 3189,

19:_.
_

163.

"VanDyke, Milton D.: Firstand Sec_d-0rder

Theory of _personie
Flow r_t Bodies of Revoiu_Ic_. Jour. Aero. Sei., vol. 18,
no. 3, Mar. 1951, pp. 161-179.

16_.

Ya,u Dyke, Milton D. : Practical Caleulatio_ of Second-Order S_rpersonie

_Vlow Past No_liftin_ Bodies of Revolution. NACA _ 27_I_tI_2.

165.

Van Dyke, Milton D. : A Study of Second-Crder Supersonic-Flow Theory.

NACA Rep. 1081, 1952. (Fo_erly NACA _ 220C)

166.

Ferri, Antonio: Application of the Method of Characteristics to

Supersonic Rotational Flow. NACA Rep. 8_I, 1946.

167.

Fraeakel, L. E. : The F_terior Drag of Some Pitot-Type Intakes at

_upersonic Speeds, _rt II. British R.A.E. Rep. No. _ero. 2_22,

i_i_

168.

Warren, C. H. E., and Gunn, R. E. W. : Estimation of Exterior Drag

of an Axially Symmetric Conical Nose Entry for Jet Engines at
Superscmic Speeds. R.A.E. _N No. Aero. 193_, S. D. 66, Jan. 1948.

169.

War_, G. N. : The Wave Idft and Dr_g Forces on a Propulsive Duct

(Athodyd) _vln_ at Supersonic Speeds. Quart. Jour. Mech. and
Appl. Math., vol. i, no_ 2, 19_8.

170.

Howell, Robert R. : A Method for Designing Lcw Drag Nose Inlet Bodies
for Operation at Moderate Supersonic Speeds. NACA RM Lg_IOla,

19_4.
171.

Jorgensen, Leland H. : Nose Shapes for Minimum Pressure Drag at

Supersonic Mach Numbers. Jour. Aero. Sci., Apr. 19_4, pp. _7,6-_79.

172.

Perkins, Edward W., and Jorgensen, Leland H. : Investigation of the

Drag of Various Axially Sy_netric Nose Shapes of Fineness Ratio 3
for Mach Numbers From 1.24 to 3.67. NACA _M A52H28, 19.52.

173.

Sibulkin, Mer_in: Theoretical and Experimental Investigation of

Additive Drag. NACA RM E_lB13, 19_l.

174.

Fraenkel, L. E. : The Exterior Drag of So_e Pitot-Type Intakes at

Supersonic Speeds, Part I. R.A.E. Rep. No. Aero. 2380, June 19_0.

175.

Graham, E. W.: Notes on the Drag of Scoops and Blunt Bodies.

SM-13747, Douglas Aircraft Co.,Inc., Santa Mc_ica, 1990.

.....
-2.::....---;_..__,*-_*_
_

t _ .......

:_

_ _ ,

Rep.

:_,*_
'_,.,,_,-,_:
:,-_:_;.
_-.,_:;_,._;_'___
_

, _*

_.

[?F

XACa_ A95F16

UOm_/D_kTIA_

of Detached Sh(>;k Waves Ahead of Plane

NACA _ i_21, 1949.
I

176.

Moeekel,

W. E. :

Approximate

Method

or Axially

for Predicting

Sym_etrlc

Form

and Loea$ion

and Pres?,ure Recovery of a Nose Inlet of Parabolic Profile

Math :!tumbersFrom 0.8 to 1.7.
NACA _4 LglE(Y2, 1951.
178.

FelTi. Ant_io:
Method for Evaluating from Shadow
Photegraphs the Pre._sure DraE in Two-Dimensional
Symmetrical Flow Phenomena With Detached Shock.

Having

0_an,
Lars:
An [_xperimenta! Method of Determining
the Drag of 8
Shock Wave with Application
to a Ducted Body.
The Flygtekniska
F6_sSksanstalten,
StocR/_olm Meddelande 51, jan. 1994.

181_.

Bra_ikoff,
George B., and Rogers, Arthur W. :
Four Nose Inlets as Measured at Maeh Numbers
NACA _4. A51C12, 1951.

182.

Esenwein, Fred T.:

Performance Characteristics
at Mach Numbers to
u/
2.00 of Various Types of Side Inlets Mounted on Fuselage of Propose
Supersonic Airplane.
IIINormal-Wedge
Inlet with Semicircular
Cowl.
NACA _4 E52H20, 1992.
J

183.

Matthews, Howard F. : Elimination

of Rl_mble from the Cooling
of a Single-_ine
i_ursuit Airplane.
NACA WE A-70, 1943.

184.

Richter, A.:
Flight Tes_ Investigation
of Engine _nlet Duct Rumble
on the Model XF4D-I for a Mach Number Range of 0.7 to 0.92.
Rep. No. DEV-1338, Douglas Aircraft Co., Inc., Santa Monica,
June 29, 1993.

185.

Smith, Norman F., and Baals_ Donald B. : Wind-Ttoz_el Investigation

of a High-Critlcal-Speed
Fuselage Scoop Including the Effects of
Bo.mdary Layer.
NACA WR L-733, 1945.
(Formerly NACA ACR LgB01a)

186.

Kelth,

Arvid L., Jr., and Sch!ff,

Jack:

Characteristics
of.
Between 1.4 and 2.0.

Low-Speed

187.

._

Ducts

Wind Tt_nel

Inves_i_tlcn
of a Triangular Swept Back Air Inlet
of a 45
Sweptback Wing.
NACA _4 LSOIO1, 1950_

_ _

Inlets

1949

180.

L_ -J .......

at

or Schlieren
or AxiallyNACA TN 1808,

179. Nuccl, Louis M. : _e Externs.1-Shock D;ag of Supersonic

Subsonic Entrance Flow.
w_CA _4 LSCGI4a, 3.990.
_

Bodies.

in the Root

Martin, Norman J., and Holzhauser, Curt A.:

Ana_.ysis of Factors
Influencing the Stability Characterlstic,_ of Symmetrical
TwinIntake-Air-Induction
Systems.
NACA TN 2049, 19_0.

'_

....

,--_,_--__I_K_.-,

", .... ,i_

_

IIII
_.

Ur.<_,-IIII
_

,
-_
....
.
_. L__ ,_,_4_'_.'_',_,'-_'_'T

_2 __

i+

!
!

'c '

128

_[

188.

_+_
;

'_ :
.c:

&

C_+_/DE_L:
:
:

tec
r

--

_:i

z:

BI_

,.

:
" :t
|
ire

:
:t
t

Ill

NACA _4 A99FI6

:.1

Mossman, Emmet A., Pfyl, Frank A., and Lazzeroni, Frank A.:
Experimental
Investigation
at Mach Numbers Frc_ 0 to 1.9 of
Trapezoidal
and Circular Side-Inlets for a Fighter-Ty; e Airplane.

NACA _ A99DmT,1959.

_//

189.

Nussdorfer,
Theodore J., 0bery, Leonard J., and Englere, Gerald W.:
Pressure Recovery, Drag, and Subcritlca! Stability Characteristics
tf Three Conical Supersonic Diffusers at Streem Mach Numbers fram
" 7 to 9.0.
NACA RM EglH27, 1992.

190.

Shock Alao._
Scoop Inlet
Boundary-Layer
NACA RMof A93D29,
Frazer,
C._ andWith
Anderson,
Warren E.Control.
: Performance
a Normal-

1953.
19]..

gbery, Leo_ard J., Cubbiscm; Robert W., and Mercer, T. G.:

Stabilizati_u Techniques for Ramp _Type Side Inlets at Supersonic Speeds.
NACA l@4 E99Afi6, 1959.

192.

Perchonok, E., W+lzox, Fred, and Fennington. Donald:

F2fect of Angle
of Attack and Exit Nozzle Design on the Performance of a 16-Inch
Ram Jet at Mach Numbers from 1.9 to 2.0.
NACA RM EglG26, 1993.

193.

Dangl_-; E. E., Cervenka, A. J., and Perchonok, Eugene:

Mechanically
Induced Si_usoidal Air-Flow Oscillations
of a Ram-Jet Engine.
NACA }_4 E94D01, 1994.

]94.

T_-impi, Robert L.: A Theory for Stability and Buzz

tude in Ram Jets and an Experimental
Investlga_lon
Effects.
NACA _4 L93G28, 1993.

199.
>+

'.
,

196.

Griggs, C. F., add Goldsmith, E. L.:

Shock
Centre-Body
Intakes at Supersonic Speeds.
24"17, Sept. 1992.

Effect of
on Operation

Pulsation AmpliIncluding Scale

Oscillations Ahead of
R._ .E. Rep. No. Aero.

Dailey, C. L. : Development
of Supersonic Ramjet Diffuser.'+, Sunmmry
Rel_rt.
Rep. 8-1, Univ. of Southenl Calif., Jan. lO, 1951.

_/197.

Trimpi, Robert L., and Cohen, Nathaniel B.: .Effect of Several

Modifications
to Center Body and Cowling on Subcritlcal
Performance
of a Supersonic Inlet at Mach Number of 2.02.
NACA RM L9_C16, 1999.
198.

Englert, Gerald W., and 0bery, Leonard J.: Evaluation of Five Conical
Center-Body
Supersonic Diffusers at Sever-i Angles of Attack.
NACA _4 ESIL04, 1992.

199.

Hasel, Lowell E., Lankford, John L., and Robins, A. W.:

Investigation
of a Half-Conical
_ .... T_!_t Mounted at Five Alternate Circthmferen-

"

tlal Locations Around a Circular Fuselage.

Pressure-Recovery
Results at a Mach Number of _-.01. NACA RM L93D30b, 1993.

....

+L
-

. ,-.

" '-P,_' .','r"_,8_;","-'v_'_++_"'_

,, ,v,

,k,,,,_+
.+

, ....

.'; +'r+.+
,

_
,,

+
....

++..+_<%,'_+."., ...+.++.,_+.,,,,_J"+;,+.
..
'. ,.,'.
+ . +:'

@_'m+'.";'+_,+"-:;
, _+.+'
. - ,+
'"

NACA RM A95FI6

CON_ID_TZ_M_

200.

Piercy, Thomas G. : Prel_inary

Investigation
of Some Internal
Boundary-Layer-Control
Systems on a Side Inlet at Mach Number
2.%,,
NACA _4, E94KOI, 1999.

201.

Mossman,

Emmet A., Lazzeroni,

Type Normal-Shock

llJJJ

Inlet.

Frank A.,
NACA RM

and Pfyl, Frank A.:

An

A59AI3, 1999.

202.

Goelzer, H. Fred, and Cortr_ght, Edgsr M., Jr. : Investigation

at
Mmch Number 1 ._q of a Conical Spike Diffusel Mounted as a Side
I_llet With Bounda_"j Layer Control.
NACA EM Eg!G06, 19_i.

203.

Ferri, Antonio, and Clarke, Joseph H.:

On the Use of Interference
Effects for Shock Dr_g Reduction at Supersonic Speeds.
PIBAL
Rep. 258, Polytechnic Inst. of Brooklyn, Dept. of Aero. Eng. and
Appl, Mech., July 1994.

204.

Hall, Charles F., and Frs.uk_ Joseph L. : Ram-Recovery

Characteristics
of NACA Submerged Inlets at High Subsonic Speeds.
NACA EM A8129,
1948.

209.

Hall, Charles F., and Barclay s F. Dorn:

An Experimental
Investigation of _CA
Submerged Inlets at High Subsonic Speeds.
I - Inlets
Forward of the Wing Leading Edge.
NACA RM ASBI6, 1948.

206.

Robinson, Russell G., and Wright Ray H. : Estimation of Critical

Speeds of Airfoils and Streamlined Bodies _ NACA WR L-781, 1940.
(Formerly NACA ACR L781)

207.

Allen, H. Julian, Frick, Charles W., and Erlckson, Myles D.:

Experimental
Investigation
of Several Low-Drag Wing-Nacelle
binations With Internal Flow.
NACA ACR 5Al_, 194_.

208.

Herriot, Jobm G. : The Linear Perturbation

Theory of Axially Sys_etric Compressible
Flow With Application
to the Effect of Compressibility on the Pressure Coefficient
at the Surface of a Body of
Revolution.
NACA PM A6HI9, 1947.

209.

Hasel, Lowell
on Circular

_I0.

Jorgensen, Leland H. : Correlation by the Hypersonic Similarity Rule

of Pressure Distributions
and Wave Dr8ge for Minlmum-Drag
Nose
Shapes at Zero Angle of Attack.
NACA BM A53FI2, 1993.

211.

Stroud, John F. : Experimental

Investigation
of the "Effect of Forebody Bluntness on the Pressure Recovery and Drag of a Twln-Scoop
inlet-Body Combination
at Mach Numbers of i._ and 1.7.
NACA
_4 ASIKI_, 1992.

"

129

JILL

E.:
The Performance of Conical Supersonic
Fuselages.
NACA _4 L93Ii4a, 1953.

J.

'li

Imlf

"

.......

_ ....

r i,' _.'

vi

Scoop

An
Com-

_,"

Inlets_ I_

iiillLJ_

__

!
/

130

.......

t
............
c
c
(
i

C__'7..a._

',
-.

: .'.

|t

":
':

,,': NAOA i_
,,
o !

Ag_FI6

-i

212.

Merlet, Charles F., and Carter, Howsx'd S. : Total-Pressul.e

of a Circular Underslung
Inlet With Taree Different Nose
at a Mach Number of 1.42.
NACA RM LSIK05, 1952.

Recovery
Shs_es

213.

Ferrl, Antonio_ Ness, N., and Kapllta, T.:

Supersonic Flow Over
_o:,cal Bodie6 Without ,_
S_etz- i. Jo'.:-..,.._u.^"-_
_.,_"_ Aug. _j,_'_

pp.563-571.
2_14.

Stoney, Wi]_liem E., Jr., and Putland, Leonard W. : Some Effects of

Body Cross-Sectional
Shape, Including a Sunken-Canopy
Design, on
Drag as Shown by Rocke_-Powered-Model
Tests at Mach Numbers from
0.8 to 1.5.
NACA _4 L52D07, 1992.

_15.

Beskln, L.:
Determination
o." Upwash Around a Body of Revolution
Supersonic Velocities.
Rep. CM-251, Applied Physics Lab.,
Johns Hopkins Univ., May 27, 19_.

)16.

Allen, H. Julian, and Perkins, Edward

Over Inclined Bodies of Revolution.

2i7.

Luidens, Roger W., and Simon, Paul C.:

Aerodynamic
Characteristics
of NACA ]_4-i0 Missile in S- by 6-Foot Supersonic Wind Tunnel at
Mach Numbers of 1.49 to 1.98.
I - Presentation
and Analysis of
Pressure Measurements
(Stabilizing Fins Removed).
NACA RM ESODI0,
1950.

./_18.
_/

at

W.:
Characteristics
of Flow
NACA 1_4 ASGL07, 1.951.

Valerino_ Alfred S., Pennington_ Donald B., and Vargo, Donald J.:
Effect of Circumferential
Location on Angle of Attack Performance
of T_in Half-Conical
Scoop-Type Inlets Mounted Sy_netrically
on
the RM-IO Body of Revolution.
NACA RM E53G09, 1993.

i
_ //_i_"_19"

Weinstein,

M.

I.-

NACA RM E92A22,
220.
_.

'

Kremzier,

Emll

Perfor_ance

of

Supersonic

Scoop

Inlets.

19_2.

J., and Campbell,

Robert

C.:

Angle-of-Attack

Super-

sonic Performance
of a Configuration Consisting
of a Ramp-Type
Scoop Inlet Located Either o_ Top or Bottom of a Body of Revolution.
NACA RM E54C09, 1954.
22-1.

Schaefer, Raymond F. : Some Design Considerations

of Half-Bound
Slde Inlets.
Supersonic Inlet Symposium Paper in Wright Aeronautical Rep. No. 1692, Wright Aero. Div., Curtiss-Wrlght
Corp.,
Wood-Ridge,
N. J., Sec. D, Jan. 2B,1953, pp. 29-39.

222.

Kremzier, _nil J., and Campbell, Robert C.: Effect of Fuselage

Fences on the Angle=of-Attack
Supersonic Performance
of a Top-

V /

Inlet-Fuselage

Configuration.

__,
....
[,

NACA RM E54J04,

1999.

-.,,_,,,,f.....

._.._.,,_

......

.....

Jl,

III

I1 IIIIIII

II

__ II

_11

NACA RM A95F16

Spahr, J. Richard, and D_ckey: Robert R. : Wind-Tunnel

Investisation
of the Vortex Wake and Downwash Field P,
ehind Triangular Wings
_
Wing-Body CombLnations
a_ Supersenlq $pe._ds. NACA R/M A53D10, 1993.

22h.

0bery, Leonard J. _ snd Erasnow, H. S. : Influence of a Canardo_Ipe

ConLx'ol Surface on the inzern_l and External Performance Characteristics of Nacelle-Mounted
Supersonic Diffusers (Conical Centerbody)
at a Rearward Body Station for a Mach Number of 2.0. NACA
RM E52F16, 1952.

225.

Dannenberg, Robert E. : A Design Study of Leading-Edge

bnswept Wings.
NACA TN 3126, 1954.

226.

Dannenberg, Robert E. : Low-Speed

Wing With Leading-Edge
Inlets

227.

Com_ors: James F., and Woollett, Richard R.:

Experimental
Investigation of a Two-Dimensional
Split-Wing Ram-Jet Inlet at Mach Number

of 3.85.

ii

133.

223.
'

_ J,

CONFIDK_TT_uL

Characteristics
NACA RM AglE29,

Inlets

for

of a 45 Swept
1951.

NACA r_,ES_FO4, 1952.

228.

Douglass, Wm. M. : Wing-Ram Jet Development

USCAL Report
Univ. of So. Calif.
Aero. Lab. Navy Research Project.
1948.

229.

Anon. : Evaluation
No. ZM-9136-001,

230.

Seddon, J., and Kettle, D. J.:

Low Speed Wind Tunnel Tests on the
Characteristics
of I/-_adingEdge Air Intakes in Swept Wings_
R.A.E. Rep. Aero. 2402, Nov. 1950.

231.

Dannenberg, Robert E., and Blackaby, james R. : An Experimental

Investigation
of a Jet-Engine Nacelle in Several Positions on a
37.25 Sweptback Wi_.
NACA RM AgOA13, 1950.
/

232.

Davis, Wallace F., Brajnikoff,

George B., Goldstein,
Spiegel, Joseph M. : An Experimental
Investigatlon
Speeds of Anntular Duct Inlets Situated in a Region
Boundary Layer.
NACA }_4 ATG15, 1947.

233.

Carter, Howard S., and Merlet, Charles F. : Flight Determination

of
the Pressure Recovery and Drag Characteristics
of a Twin Side-Inlet
Model at Transonic Speeds.
NACA I_4 L93E05, 1993.

234.

Pierpont; P. Kenneth, and Braden, John A.:

Investigation
at Transonic Speeds of a Forward-Located
Underslung Air Inlet on a Body
of Revolution.
NACA I_4 L52K17# 1993.

/
V"

3-9,
June l_,

of Supersonic Split Wing Rsm Jets.

Convair Rep.
Consolidated
Vultee Aircraft Corp. _ May 1949 .

David L., and

at Supersonic
of Appreciable

.//
v

132

C0_II_._IAL
.

I
l
l

.....
{ _
I I

: .. "': .":

II

I I

NACA EM A55FI6

:.

Boswinkle, Robert W., Jr., and Mitchell, Meade H., Jr. : Experimental Investigation
of the Internal-Flow
Characteristics
of Porward
Underslung Fuselage Scoops With Un.swept and Sweptback Entrances
at Mach Numbers of 1.41 to 1o96.
NACA _ L52A24, 1952.

236.

Merlet. Charles F.:

Pressure Recovery _nd Drag Characteristics
of
a Forward Located Circuiar Scoop Inlet as Determined From Flight
Tests for M_ch Numbers From 0.8 to ].6.
NACA RM L54B23, 195h.

237.

Nichols_ Mark R., ,_nd Goral, Edwin B.: A Low Speed Investigation
of a Fuselage-Side
Air Inlet for Use at Transonic Flight Speeds_
NACA TN 2684, 1952.

238.

Spiegel, Joseph M., Hofstetter, Robert U., and Kueb_l, Donald M :

Applications
of Auxiliary Air Injectors to Supersonic Wind Tunnels
NACA RM A5310_, 1953.

239.

Fradenburgh,
Evan A., and Campbell, Robert C.:
Characteristics
a Canard-Type Missile Configuration
With an Underslung Scoop
Inlet at Mach Numbe-:s From 1.5 to 2.0.
NACA BM. E52J22, 1953.

240.

Watson,

Earl

C. :

Some Low-Speed

Characteristics

System Having Scoop-'l_ype Inlets With

Control.
NACA RM A51FIS, 1951.

'.
_
i

II

II

235.

Provisions

of

of an Air-induction
for Boundary-Layer

241.

McLafferty,
George:
Theoretical
Pressure Recovery Through a Normal
Shock in a Duct With Initial Boundary Layer.
Jour. Aero. Sci.,
vol. 20, no. 3, Mar. 1953, pp. 169-174.

242.

Simon, Paul C. : Internal Performance

of a Series of Circular
Auxiliary-Air
Inlets Immersed in a Turbulent Bo_mdarv Layer Mach
Number Range: 1.5 to 2.0.
NACA HM E54L03, 1955.

243.

l_rcy,
Thomas G., and Johnson, Harry U. : A Comparison of Several
Systems of Boundary-Layer
Removal Ahad of a Typical Conical
Extrnal-Compression
Side Inlet at Mach Numbers of 1.88 and 2.93.
NACA RM E53FI6, 1953.

244.

Valerino, Alfred S_ : Performance

Characteristics
at Mach Numbers to
2.0 of Vsmious Types of Side Inlets Mounted on Fuselage of Proposed Supersonic Airplane.
I - Two-Dimensional
Compression-R_mp
Inlets With SemJcircular
Cowls.
NACA RM E52E02, 1952.

245.

Allen, J. L., and Simon, P. C.:

Performance
Characteristics
at Mach
Numbers to 2.0 of Various Types of Side Inlets Mounted on Fuselage
of Proposed. Supersonic Airplane.
II - l_lets Utilizing Half of
a Conical Spike
NACA RM E_2G08, 1952.

X/"

"!

:
i

i
,
i

V"
!
_/
[

[
,

,
,ll

, i

i i

, ]

F'

I t _J

'

NACA RMA55FI6

"

246.

Simon, Paul C. : Performance Characteristics

at Mach Numbers to 2.0
of Various Types of Side IrLlets Mounted on __h_
_ _age of Proposed
Supersonic Airplane.
IV - Rectangular-Cowl
inlets With Twot//
Dimensional
Compression Ramps.
NACA P_4E5H29, 1952.

247.

P__ercy, Thomas G_, and Johnson, Harry W. : Experimental

Investigation at Mach Numbers 1.88, 3.16, and 3.83 of Pressure Drag of
Wedge Diverters Simulating Boundary-Layer-Removal
Systems for
Side Inlets.
NACA RM E53L14b_ 1954.

248.

Campbell, Robert C., and Kremzier, Emil J. : Performance

of WedgeType Boun6_ry Layer Diverters for Side Inlets at Supersonic Speeds.
NACA EM E54C2-_, 1954.

249

Kremzier, Emil J., and Dryer, Murray:

Aerodynamic
Interference
Effects on Normal and Axial Force Coefficients
of Several EngineStrut-Body Configurations
at Mach _.Ibers of 1.8 and 2.0.
_C_
RM E52B21, 1952.

:'
_
_

C0_'iDENTIAL

133

250.

Obery, Leonard J., and Stitt, Leonard E. : Investigation

at Mach
Numbers of 1.5 and 1.7 of Twin-Duct Side Air-lntake System with
9 Compression Ramp Including Modifications
to Boundary-LayerRemoval Wedges and Effects of a Bypass System.
NACA RM E53HO4,
1953.

251.

Roberts, Howard E., and Langtry, B. D.:

_ne Influence of Design
Parameters on the Performance of Subsonic Air Inlets.
IAS Preprint
No. 260, Jan. 25, 19_0.

2_2.

Nichols, Mark R., and Pierpont, P. v_.eth:

Preliminary
Investibation of a Submerged Air Scoo], U'ilizing Boundary-Layer
SucSion
to Obtain Increased Pressure Recozery.
NACA P_I LSOAI3, 1950.

253. "'Pierpont, P. Kenneth, and Howell, Robert

tion of a Semisubmerged Air Scoop With
Suction.
NACA RM LSOHI5, 1951.

R.: Low-Speed Investigaand Without Boundary-Layer

254.

Braden, John A., and Pierpont, P. Kenneth:

Pressure and Force
Characteristics
at Transonic Speeds of a Submerged Divergent-Walled
Air Inlet on a Body of Revolution.
NACA RM L53C13, 1953.

255.

Frank, Joseph L., and Taylor, Robert A. : Comparison of Drag, Pressure Recovery and Surface Pressure of a Scoop-Type Inlet and an
NACA Submerged Inlet at Transonic Speeds.
NACA RM ASIH20a, 1951.

256

Rolls, L. Stewart:
A Flight Comparison of a Submerged Inlet and a
Scoop Inlet st Transonic Speeds.
N,_CA RM A53A06, 1953.

!J /.

-.

/"

_//

li:

134

257.

._!

CO_IDZh_FIkL

NACA EM A55F16

Frick, Charles W., Davis, Wallace F., Randall, Lau_,_sM.- and

Mossman, En_met_ : An Experimental Investigation of NACA
Submerged-D_ict_utrances.

_/'_

I I

'. ....

NACA ACR A5120, 1945.

258.

_acks, /_vin _., and Sprelter, Jo_m R.. _neoret_c_l Investigation

of Submerged Inlets at Low Speeds. NACA TN 2323, 1951.

259.

A_Iderson,Warren E., and Frazer, Alson C. : Investigation of an

NACA Submerged Inlet at Mach Numbers From 1.17 to 1.99. NACA
A52FI7, 1952.

260.

Mach Numbers to 2.0 of Two Auxiliary Inlets Immersed in Fuselage

Pennington; Donald B_, and Simon, Paul C. : Internal Performsaqceat
Boundary Layer. NACA RM E53L2_b, 1954.

261.

Morkovin, M. V., Migotsky, E., Bailey, H. E., and Phinney, R. E.:

Experiments on Interaction cf Shock Waves and Cylindrical Bodies
at Supersonic Speeds. Jour. Aero. Sci., vol. 19, no. 4, Apr. 1952,
pp. 237-248.

262.

Keith, Arvid L., Jr.: Tzansonic Wind-Tunnel Investigation of the

Effects of Body i_dentation on the Aerodynamic Characteristics of
a Semlelliptical Sweptback Wing-Root inlet Configuration. NACA
RM L54A29, 1954.

263.

Whitcomb, Richard T. : A Study of the Zero-Lift Drag-Rise Characteristics of Wing-Body Combinations Near the Speed of Sound. NACA
RM L52H0_, 1952.

264.

Hayes, Wallace D.: Linearized Supersonic Flow. Rep. AL-222,

North American Aviation, Inc., Los Angeles, Calif., June 18, 1947.

265.

Smith, Norman F., Bielat, Ralph P., and Guy, Lawrence D.: Drag of
External Stores and Nacelles at Transonic and Supersonic Speeds.
NACA RM L53123b, 1953.

266.

Conrad, 0. : 2.4-Transitions Between Wing and Power Unit. KS - The

Installation of Jet Propulsion Units, D. Kuchemann, e_., _VA
Monographs, A. Betz, ed., VG 238, British Ministry of Aircraft
Prod_ction, Volkenrode, Oct. 15, 1947.

!,

.!

267.

Hoffman, Sherwood, a_d Wolff, Austir L. : Transonic Flight Tests to

Determine Zero-Lift Drag and Pressure Recovery of Nacelles Located
at the Wing Root on a 45 Sweptback Wing and Body Configuration.
NACA RM L53H20, 1953.

268.

Jones, Robert T.: _leory of Wing-Body Drag at Supersonic Speeds.

NACA EM A53H18a, 1933.

_,
i
_
:

- II

]"

"

'"'_

, ,.....

_,.:'_ . ..,.

li

I_.,_
.-" .-..:..

III lii

__

--

_.
__

_SF

NACA EM A95FI6

269.

Lom_,
Harvard:
The Wave Drag of Arbitrary Configurabions
in
Linearized Flow as Determined by Areas and Forces in Oblique
Planes.
NACA RM A55A18, 1955.

270.

Baldwin, Bar_ett S., Jr., and Dickey, Robert R.: Application

.... _-_v_j
_,_cv_.y to Di-ag Eeductlun
_L Low Supersonic
Speeds,
NACA P44A54JI9, 1999.

271.

Lomax, Harvard, and Heaslet, Max. A. : A Special MeLhod for Finding

Body Distortions That Reduce the Wave Drag of Wing and Body Combinations at Supersonic Speeds.
NACA PM A55B16, 1995.

272.

Nielsen, Jack N. : General Theory of Wave Drag Reduction for Combinations Employing Quasi-Cylindrical
Bodies ;'i_h an Application
to Swept Wing and Body Combinations.
NACA F_ AS_B07,1995.

2_3.

Friedman, Morris D., and Cohen, Doris. : Arrangement

of FusiformBodies to Reduce the Wave Drag at Supersonic SpeeCs.
NACA
TN 3345, ].954.

274.

Spahr, J. Richard, and Robinson, Robert A. : Wind-Tunnel

Investigation at Math Numbers of 2.0 and 2.9 of Several Configurations
of
a Supersonic

Ram-Jet

Test Vehicle.

NACA RM A50C20,

135

of

1950.

275.

Luskin_ Harold, and Kl_in, Harold:

The L__fluence c2 Turbo-Jet
Airflow on the Aerodynamic
Design of Airplanes.
Douglas Aircraft
Co., Rep. S_-19111, Jm_e 3, 1955.

276.

Bollech, Thomas V., and Kelley, H. Neale:

Lo,_-Speed Longitudinal
Stability and L_tera! Control Cnaracteristics
of a 0.3-Scale
;_c_e] of a 40 Swept Wing Fighter. Type Airplane.
NACA l_ L54BI7,
1954.

277.

Scherer, A. : 2.3-Flllet Between ._uselage a_,d Power Unit.

Ks - The
Inst_llation
of Jet Propulsion Units, D. K_chenmmm, ed., AVA
Monog;aphs, A. Betz, ed., VG 239, British Niinistry of Aircraft
Production, Volkenrode, Oct. 15, 1947.

278.

Bielat, Ralph P., and Harrison, Daniel E. : A TTansonic Wind-Tuz_nel

Investigation
of the Effects of Nacelle Shape and Position on the
Aerodynamic
Characteristics
of Two 47 Sweptback Wir_-Body Combinations.
NACA RM L52G02, 1952.

_79.

r_0NFID_A[J.

Carmel, , 41vln M., end Flschettl, Thomas L.: A Trs/%sonic WindTunnel Investigation
of the Effects of Nacelles on the AerodyDamlc
Characteristics
of a Complete Model Configuration.
NACA ]_M L95F22a
1953.

!
,/

28_.

Hase_, Lowell E. sad Sevier,

John R., Jr. : Aerod_nsmLtc Characteristics
at _cmic
Speeds af s Series
of Wing-Body Combinatioas
t_aving Ca_
Wings Aith an As uect Ratio of 3-5 add a _aper Ratio
cf J.2. Effect at M = 1.60 of Nacelle
Shape and Position am the

with_

Swe_

w_w_. _

_ LSL_,

19_.

28].. Driver, Cornelius: Aerodynamic .3_arac_eristicsat Su_rs_c

Speeds
-_ a Series of W._-Body Camblnatloms Ha_Im_ Cambered Wi_6s With
an Aspect Ratio of 3.9 a_i a Taper Ratio of 0.2. Effect at
M = 2.0]. of Nacelle Shape and. Position on the Aero_c
Characteristics in Pitch of Two Wlng-Body Combinatians With _7 S_eptback Win6_. NACA I_4L_F03, 1992.

zSz. _tz,

Barry:_te

Bodies, _terbodies,

farCnAculation
of L_t of

and Combinations of Bodies.

NACA _N 2669,

_z.
283.

0_m_X, Leunaz_ J., and Krasnow, Howard S. : Pez'fo_e

Ch_aeterIs_Ics of Csnard-'l_ Missile With Vertically Mounted Nacelle
Engines _t Nnch _rs
1.9 to 2.0. NACA RM E92R08, 1952.

28_.

Krem_ier, Emil J., and Davids, Joseph: Pe_fmm_mce Characteristics

of Canard-Type M_.ssile With WinE-Mounted Nacelle
Engines at Math
Numbers i._ to 2.0. RACA RM ESeJ08, 1992.

280.

Pankhurst, R. C., and Holder, D. W. :

Pitman Pub. Co., Immdon, 1952.

286.

Young, A. D., and Msas, J. H. : The Behavior of a P'tot Tube in a

Transverse To%al-Pressu_e Gradient. R. & M. 177C, British A.R.C.,
1936.

287.

Golds%eln, S.: A Rote on the Measurement of Total Head and Static

Pressure in a Turbulent Stream. Proc. of the Roy. Soc., ser A,
_oi. I_5, no. A-886, 1936, PP. _70-_7D.

288.

Fage, A. : On the Static Pressure in Fullj--DevelopedTurbulent Flow.

Proc. of the Roy. Sot., ser. A, wol. i_9, July I, 1936, Pp. 976-

:"
C _
:

Wim_ Tunnel Technique.

_96.
t

289.

Persh, Jerome, and Bailey, Bruce M. : A Method far Estimatin_ the

Effect of _k_bulent Veloc_ty Fluctuations in the Boundary Layer
on Diffuser Total-Pres.zre-Loss Measurements. NAC_ _N 312_, 195_.

29C.

Nielsen, Jack b.: _ffect of Turtulence c_ Air-Flow Measurements

Behind Orifice Plates. NAC_qWR .'-27_,1943. (Formerly NACA ARR

_30)

i
|

_I_

Ag_F16.....

i_

291.

_ :

,_

"" : :" .;,

137

2 ':: ::!......

Kro]l, A. Edgar: Gauss's Formula in Chem_icalEn6ineering Calculati_Is. Chemical Engineer_,

vol 53, no+ 9, Sept. 19_6, pp. I02109.

292.

.Sc_rbrough,James @. : Numerieal Mathema._iealAnalysis.

Hopkins Press, 1930.

293.

Dean, Robert. Jr., ed.; Aerod_,ms_c _easurements.

Laboratory, M.I .T., 19_3.

294.

Wyatt, DeMarquis D.: Analysis of Errors Intro_lucedby Several

Methods of WeiEhlnE Ncmuniform Duct Flows. NACA _N 3_O0, 195_

The Johns

Gas _bine

.a

Th_ following
bibliogrsp_ has been arranged chronologically under
the headings of the table of contents. The n_bers of the references
that are pertinent to each subject are also listed. The majority of the
reports lu:luded in this blbl: graphy were published after June 1947,
because the NACA Bibliography of reference 3 lists previous reports.

II. I]EFINITIONS

See references 2 and 5 through

13.

Rudnick, Philip: Momentum Relations in Propulsive Ducts.

Sci., vol. 14, no. 9, Sept. 1947, pp. 9_O-9_.

Jour. Aero.

III. PHELIMINARY CONSIDERATIONS

See references

15, 16, 20, and 25.

Cleveland
Laboratory
Staff: PerformanceSystem.
and Ranges
Various Types
of Aircraft-Propulsion
NACA of
TNApplication
1349, 1947.of
Brewster, J. H. :

The Determination of the Optimum Airplane-Powerp3_nt

Combination. SAE Preprint No. 89, Dec. 1947.

,
i

A I

'

Lubarsky_ Bernard: Performance sad Load-Range Characteristics of TurboJet Engine in Transonic Speed Range. NACA _ 2088, 1950.
Sturdevant 2 C. R._ and Woodworth, L. R.: A Generalized Turbojet Weight t
Size, and Perfox_aanceStudy. Rep. R-166, Rand Corp._ Santa Monica,
Dec. 19, 1949.
Krebs, Richard P., and Wilcox, E. Clinton: Analysis of the Turbojet
Engine for Propulsion of Supersonic Bumbers. NACA _ ES_A21_ 1954.
Koutz_ Stanley L o_ sad Hensley_ Reec6 V. : Loitering and Range Performance
of Turbojet-Powered Aircraft Determined by Off-Design Engine Cycle
Analysis. NACA RM ESIK29, 1952.
/"

Gabriel_ David S., Krebs2 Richard P.3 Wilcox, E. Clinton, sad Koutz,
Stauley L. : Imalysls of the Turbojet Engine for Propulsion of
Supersonic Fighter Airplanes. NACA I_4E)2FIT, 1953.

....

....:.:
....

l
!

i
t

NACA RM A55FI6

._ _

AIRFR_

See references I, 2,
Blatz, William J.:
1948.

_EN_.

_:

139

-INDUCTION-SYSTEM COMBINATIDN

5, I0, and 15 to 18.

Air Inlets and Nacelles.

IAS Preprlnt 142, Mar. 19,

ENGIRE-INDUCTION-STS_EM COMBINATIONS
_IING

See references 6, 8 to i0, ]2, and 21 to 24.

Sanders, Newell D., and Behun, Michael: Generalization of TurbojetEngine Performance in Terms of F_mping Characteristics. NACA TN 1927,
1949.

OPTIMIZATION

See references 8, 19, an_ 24.

q

Blackaby, James R. : An Analytical Stady of the Comparative Performance

of Four Air-Ynduction Systems for Turbojet-Powered Airplanes Designed
to Operate at Mach Numbers up to 1.5. NACA RM A52CI_, 1952.

"

Watson, Earl C.: An Analytical Study of the Comparative Performance of

Six Air-lnduc_tion Systems for Turbojet-Powered Air_lanes Designed to
0perate at Mach Numbers up to 2.0. NACA RM A53H03, 1953.

J
_"
_..j

McLafferty, George: Simplified Methods for Comparing the Performance of

Supersonic Ramjet Diffusers. Rep. R-23996-1, Res. Dept., United Aircraft Corp., Mar. 25, 1953.

,
:_

Allen, J. L., and Beke, Andrew: Performance Comparison at Supersonic

Speeds of Inlets Spilling _cess Flow by Means of Bow Shock, Conical
Shock, or Bypass. NACA RM E53Hil, 1953-

i\

FLOW UNIFORMITY AND STEADINESS

See references 27 to 38.

/
I

!'__-

"' (L.. ,-.J.............

'"''_"/1

.....
--

,._..-,, ....

'-_""'-_"__I',:_._''
:_
_:.
___

",_<-_
' _+,'-'_','
_":_;_"_
'_-"
:"_
__:i_."_',-,
'
_

,'m..............................

_i

."

Pearce, R. B.- Causes and Control of _owerplant Surge.

vol. 92, no. 3, Jan. 16, 19_0, pp. al-eg.

Aviation Week,

Mark, Herman, and Zettle, Eugene V. : Effect of Air Distribution on

Radial Teu_eraturc Distribution in 0ne-Sixtl Sector of Annular TurboJet Combustor. NACA RM E9122, 1950.

IV.

D_fAYL CONSIDERATIONB
INDL_TION

FEESSURE RECOVERY AND FLOW UNIFORMITY

Ducts
Skin friction and separation

i
I
J

See references 39 to 44, 63, 109, 139, and 136.

i
i

Griffith, A. A. : Reducing Surface Friction Between a Solid and a Fluid.

Patent 978,763 (Br.), Aug. 2_, 1942, no. 11973.

"

Regenseheit, B. : Drag Reduction by Suction of the Boundary Layer Separated Behind Shock-Wave Formation at High Mach Numbers. NACA TM 1168,
1947.

!
I

!
I

Pierpont, P. Kenneth: Investigation of Suctlon-Slot Shapes for Controlling a Turbulent Boundary Ls_er. NACA TN 1292, 1947.
Dz?/den,Hugh L. : Recent Advances in the Mechanics of Botmdary L_yer
Flow. Vol. I, Advances in Applied Mechanics, sec. i, yon Mises, Richard_
and yon K_rm_u, Theodore, eds., Academic Press, Inc., N. Y., 1948,
pp. 2-)-1.0.
Oswatitsch, K._ and Wieghardt, K.: Theoretical Analysis of Stationary,
Potential Flows and Boundary Layers at High Speed. NACA TM 1189, 1948.
Lagerstrom, Facol Cole, Julian M. e and Trilling, Leon: On Viscous Effects
in Compressible Flow. Paper presented at meeting of Institute for Fluid
Mechanics and Heat Transfer, Pasadena, Cal., June 23, 1948.
Kay, J. M. : Experimental Investigation of Boundar.wLayer Flow Along a
Flat Plate with Uniform Suction. Aero. Com. no. ii,_76, British A.R.C.,
1948.
Dickinson, H. B. : Flight and Tunnel Test Research on Boundary-Layer
Control. Jour. Aero. Sci., vol. 16, no. 4, Apr. 1949, pp. 243-2_1.
m_
"
".....
_)Li-_

g2t
..-,.
,,,
- ._......
-_
....

.... ,_._.__ram,m,"
.. - .
........

"

"_'_-_
_,_,;:_...._.
"r__'_"_':
v.. ...

Schliehting, H. :
Laminar Flows.
TM 1218, 1949,

Lecture Series, "Boundary Layer Theory."

Pt. I NACA TM 1217, 1949.
I_. II - Turbulent Flows.
NACA

Van Driest, E. R..

Turbulent Boundary- Layer for Comp[cessible Fluids on
an Insulated Flat Plate.
Rep. AL-958, North American Aviation, Inc.,
Sept. 15, 1949.
Kay, J. M. : Turbulent Boaudary Layer Flow with Uniform
Motion.
Sub-Com. British A.R.C. 12,193, 1949.

Suction.

Fluid

yon Doenhoff, A. L., and Loftln, L. K., Jr. : Present Status of Research
on Boundary-Layer
Control.
Jour. Aero. Sci., vol. 16, no. 12, Dec.
1949, pp. 729-740, 760.
Van Driest, E. R. : Turbulent Boundary ]_yer for Compressible
Fluids on
a Flat Plate with Heat Transfer.
Rep. AL-997, North American Aviation,
Inc., Jan. 27, 1950.
Ferrari, Carlo:
Study of the Boundary L_Lyer at Supersonic Speeds
Turbulent Flow:
Case of Flow Along a _lat Plate.
Quart_ Appl.
vol. VIIi, no. l, pp. 33-57, Apr. 1950.

..

Davis, Don D., Jr., and Woods,

Reflections
of Oblique Waves
1990.

in
M_th.

George P.: Preliminary

Investigation
from a Porous Wall.
NACA RM LSOG19a,

of

Andersou, K_ G. : Preliminary
Investigation
of Boundary Layer Control at
High Subsonic Speeds.
Tech. Rep. No. 6186, U. S. Air Force, WrightI

Patterson

Air Force Base,

1950.

Tucker, Maurice:
Approximate
Turbulent Boundary-Layer
Development
in
Plane Compressible
Flow along Thermally Insulated Surfaces with Application to Supersonic-Tunnel
Contour Correction.
NACA TN 2045, 1950.
Young, A. D. : The Equations of Motion and Energy and the Velocity Profile
of a Turbulent Boundary Layer in a Compressible Fluid.
College of
Aeronautics,
Crsnfield, England, Rep. 42, Jan. 1951.
Van Driest, E. R. : Turbulent Boundary Layer in Compressible
Fluids
Jour. Aero. Sci., vol. 18, no. 3, Mar. 19_i, pp. 145-160, 216.
Young, A. D. : Boundary Layers and Skin Friction
Jour. R. Ae. Soc., l_y, 1951, pp. 285-30o-.

"

in High-Speed

Flow.

Rubesin, Morris W., Ma_ew,

Randall C., and Varga, Steven A. : An _alalytical and Experimental
Investigation
of the Skin Friction of the Turbulent
Boundary Layer on a Flat Plate at Supersonic Speeds.
NACA TN 2305,
1951

':

142

: "'":!: : NACARM A95FI6

*_ _ _:'_
_O_I_T_
c

_kleker, Maurice:
Approximate
Development
in Compressible

Jt

tl

_ll

aB

@al

eal

Calculation
of Turbulent Boundary-Layer
Flow.
NACA TN 2337, 1991.

Klebanoff,
F. S., and Diehl, Z. We:
Some Features of Artificially
Thickened Frilly Developed Turbulent Boundary Layers With Zero Pressure
Gradient.
NACA TN 2479, 1991.
Rubert, Kennedy F., and Persh, Jerome:
A Procedure for Calculating
the
Development
of Turbulent Bounda_-j Layers Under the Influence of Adverse
Presstre Gradients.
_iACA TN 2478, 1951.
Schubauer, G. B., and Klebanoff,
the Turbulent Boundary Layer.
NACA TN 2133)

P. S. : Investigation
NACA Rep. 1030, 1951.

of Sep_ratlon
(Supersedes

of

Tetervin, Neal, and Lin, Chla Chlao_

A General Integral Form of the
Boundary-Layer
Equation for Incempresslble
Flow With an Application
to the Calculation
of the Separation Point of Turbulent Boundary Layers.
NACA Rep. I0_6, 1951.
(Supersedes NACA TN 2158)
McCt__lou_h, George B., and Ga_bt'cci, Bruno J.:
Bou_Idary-Layer Measurements
on Several Porous M_terial_ with Suction Applied.
NACA RM A52D01b, 1952.
Anderson, Oiv_ R. : Investigation
of Perforation
Exhaust Flow and Its
Effect on _xternal Skin Friction.
Rep. R-Sj484-31 , Research Dept.,
Unized Aircraft Corp., Sept. 1952.
Lee, J. D. : The Influence of High Adverse Pressure Gradients on Boundary
Layers in Supersonic Flow.
Rep. 21, Institute of Aerophyslcs,
University of Toronto,
Oct.1992.
Anderson,
Speeds.

K. G. : Investigation
of Boundary-Layer
Control
Tech. Rep. 6344, pt. i, U. S. Air Force, WADC,

at High Subsonic
Jar. 19_i.

Design

See references

29, 44, 49, 48, 55 to 65, 69 to 71, _2, and 118.

Gray, S.: A Survey of Existing Information

on the Flow in Bent C_hannels
and the Losses Involved.
Power Jets Rep. R. 1104, Power Jets (Res. and
Dev.), Ltd, June 1345.
Seddon, J., and Spenc_, A.: Wind Tunnel Measurement5
of Internal Loss
on a Full-Scs_e Model of & Wing Root Intake Duet for a Proposed Jet
Propelled Fi@_ter.
British, R.A.E. TNNo.
Aero. 1757, Dee. 1945.

I[
!

Cohen, Herbert N. : Investigation of Intake Ducts for a High-Speed Subsonic Jet-Propelled Airplane. NACA RM L7C24a, 1947.

Turner, L. Richard, Addle, Albert N., and Zimmerman, Richard H.: Charts
for the _m_.lysisof 0ne-Dimensional Stead>-Compressible Flow. _ACA
TN 1419, 1948.
Shapiro, Ascher H., and Ha%_horne, W. R. : The Mechanics and Thermodynamics
of Steady One-Dimensional Gas Flow. Jour. Appl. Mech., vol. 14, no. 4,
Dec. 1947, pp. A317-A336.
Butcher, Marie A. : Compressible Flow 'l_blesfor Air.

NACA TN 1992_ 1948.

Gratzer, L. B., and Smith, E. H. : Boundary Layer Control for Wide Angle
Diffusers. Rep. 300 (ONR Contract N6ori-217, Task Order I, Project
No. NR-061-00$), Univ. Washington Aero. lab., Nov. 22, 1948.
Palme, Hans Olof: An Investigation of the Effect of Bo?mdary Layer Suctiou on the Air Resistance in Channel Elbows. KTH-Aero TN 2, Roy. Inst.
of Tech.# Flygtekniska Institutionen, Stockholm, Sweden, 1948.

Taylor, H.
fusers.
Research
Dec. 31,

D. : Application of Vortex Generator Mixing ;_rlnclpleto DifConcluding report, Air Force contract W33-038 ac-21825.
L-pt. R-19064-9, United Aircraft Corp., East Hartford, Conn.,
1948.

Schwartz, Ira R. : Investigations of an Annular Diffuser-Fan Combination

Handling RotatiI_ Flow. NACA RM L9B28, 1949.
Nelson, William J., and Popp, Eileen G. : PerfformanceCharacterisTics of
Two 6 and Two 12 Diffusers at High Flo_'Rates. NACA RM LgH09, 1949.
Hawthorne, William R. : Seconda?v Circulation in Fluid Flow.
Lab., M. IT., May 1990.

_i

Gas Turbine

Neumann, Ernest P., and Lustwerk, F. : High-Efficiency Supersonic Diffusers. M.I.T. Meteor Rep. 96, June 1990.

Persh,
The
of the Inlet
and Inlet-Boundaryt//
LayerJerome:
Thickness
onEffect
the Performance
of Mach
a 23Number
Conicsl-Diffuser
- TailPipe Combination. NACA RM L9K10, 19_0.
Squire, H. B., and Carter, P.: Further Experiments on Conical Diffusers.
Rep. No. 13,499, British A.R.C., Nov. 6, 1950.

'

Sibulkin, Merwin, and Koffel, William K. : Chart for Simplifying Calculations of Pressure Drop of a High-Speed Compressible Fluid under Simultaneous Action of Friction and Heat Transfer - Application to Combustion
Chamber Cooling Passages. NACA TN 2067, 1990.

j:

.Jb

144

.'

C tVF E_I
_"_"......

tTl

, , ,NACA RM A55F!6

Squire, H. B., and Winter, K. G.: The Secondary Flow in a Cascade of

Airfoils in a Nonuniform Stream. Jour. Aero. Sci., vol. 18, no. 4,
Apr. 1951, pp. 271-277.
Kronauer, Richard E. : Se_ondary flo_ in fluid dynamics,
Congress of Appl. Mech., pp. 747-756, 1951.

Proc. Natlon_l

W_rren, C. H. E., Dudley, R. E._ and Herbert, F. J.: A Theoretical and

Experimental Investigation of the Flow in a Duct of Varying Cross
Section, with Particular Application to the Design of Ducts for FreeFlight Gro'omd-La,
mched Model Tests. British A.R.C. Tech. Rep. 13726,
C.P. No. 60, 1951.
Wood, Charles C. : Preliminary I_estigation of the Effects of Rectan_
galar Vortex Generators on the Performance of a Short 1.9:l StraightWall Annular Diffuser. NACA RM LSIG09, 1951.

\,

,/

Fersh, Jerome: The Effect of Surface Roughness on the Performance of a

23 Conical Diffuser at Subsonic Mach Numbers. NACA RM L_IK09, 1951.

\/

Valentine, E. Floyd, and Carroll, Raymond B. : Effects of Several Arrangements of Rectangular Vortex Generators on the Static-Pressure Rise
Through a Short 2:1 Diffuser. NACA RM LSOL04, 1951.
i

V_!ent_ne, E. Floyd, and Carroll, Raymond B.: Effects of Some Primary

Variables of Rectangular Vortex Generators on the Static-Pressure Rise
Through a Short Diffuser. NACA RM L52B13, 1952.

!
,_

i
:

Stanitz, John D. : Design of Two-Dimensional Channels with Prescribed

Velocity Distributions Along the Channel W&lls. I - Relaxation Solutions. NACA TN 2593, 1952.

I
r
I

Kramer, James J., and Stanitz, John D.:

a 90 Elbow. NACA TN 2736, 1952.

Two-Dimensional Shear Flow in

Wood, Charles C., and Higginbotham, James T. : Flow Diffusion in a

Constant-Diameter Duct Downstream of an Abraptly Ter.inated Center
Body. NACA RM L53D23, 1953.

Valentine, E. Floyd, s_d Copp, Martin R.: Investigation to Determine

Effects of Rectangular Vortex Generators on the Static-Pressure Drop
Through a 90 Circular Elbow.
/Wood,
;_

__

NACA RM L53G08, 1953.

Charles C., and Higginbotham, James T. :

Performance Cl_racteristics

of a 24 Straight-(_ter-Wall Annular-Diffuser-Tailpipe Combination

Utilizing Rectangular Vortex Generators for Flow Control. NACA
RM L53HI7a, 1993.

_'

'

J_
,_

[111

i i

ii

. .

....

ii i

II III

ii

......

:,

_,

I i

II

i i

-L.J

i i i ill

I i

ii

IIII[

--

_.i

NACA P&_A55FI6

C0_'_IDENT_L "'

145

Wood, Charles C., and Higginbotham, James T. : The Influence of Vertex

Generators on the Performance of a Short 1.9:l Sti_ight-Wall A_nnular
Diffuser With a Whirling Inlet Flow. NACA RM L52LOla, 1953.
Reid, Elliott G. : Performance Characteristics of Plane-Wall TwoDimensional Diffusers. NACA TN 2888, 1953.

Fersh, Jerome, and Bailey, Bruce M. : Effect of Surface Roughness Over

the Downstream Region of a 23 Conical Diffuser.

NACA TN 3066_ 1954.

Farley, John M., and Welna, Henry J. : Investigation of Conical Subsonic

riffusers for Ram-Jet Engines. NACA RM E_3L15_ 1954.
Mallett, William E., and Harp, James L., Jr.: Performance Characteristics
of Several Short Annular Diffusers for Turbojet Engine Afterburners.
NACA RM E54B09, 1954.
:_

Henry, John R. : Aspects of Internal-Flow-System Design for Helicopter

Propulsive Units. NACA P_ L54F29, 1954.
Wood, Charles C., and Higginbotham, James T. : Effects of Diffuser and
Center-Body Length on Performance of Annular Diffusers with ConstantDiameter Outer Wshls and with Vortex-Generator Flow Controls. NACA
RM L54G21, 1954.
Henry, John E., and Wilbur, Stafford W. : Preliminary Investigation of
the Flow in an Annular-Diffuser-Tailpipe Combination with an Abrupt
Area Expansion and Suction, Injection, and Vortex-Generator Flow
Controls. NACA RM L53K30, 1954.
Persh, Jerome, and Bailey 3 Bruce M. : Effect of Various Arrangements of
Triangular Ledges on the Performance cf a 23 Conical Diffuser at Subsonic _lachNumbers. NACA TN 3123, 1954.

Subsonic Flight

See references 14, 28, 73, 78, 82, and 185.

_
@
_

Dennard, John S. : An Investigation of the Low-Speed Characteristics of

Two S1mrp-Edge Supersonic Inlets Designed fo_"Essentially External
13upersonicCompression. _&CA RM LTD03, 1947.
Baals, Dom.JLdD., Smith, Norman F., and W_ight, John B. : _"neDevelopment
an@ Application of High-Crltical-Speed N_se Inlets. NACA Rep 920,
io48.

--

II J.I I
il

i, i _ -_

ill11 i

__J_ _

I ttntllluu

I
i

II
ii

._
i

ii

146

'

'C_B_NT_

, !

_ e

f t

Connor,
F.,
and Widllmd,
J. : Flight
Pressure
Recovery.
EFTMR No. 1193,

: .', '": .'I_CA RM A55F16

Ig

Tests
of
Republic

ee

eo

eee

0
J_

F_E
Induction
System
Aviation
Corp.,
Mar.

20,

1951.
Becker,
John Vernon:
Wind-Tunnel
Openings oh a Streamline Body.
NACA ACR, _ov. 1940)

Investigation
of Air
Inlet
and Outlet
NACA Re;,. 1038, 1951.
(Su)ersedes

BraJnikoff,
George B., 9_ud Stroud, John F. : Exper_mlental investigation
of tP,_ Effect of Entrance Width-to-Height
Ratio on the Performance
of
an _uxiliary Scoop-Type Inlet at Mach Ntmlbers From 0 to 1o3. NACA
Ye4 A53E28, 1953.
Bryan, Carroll R., and Fleming, Frank F. : Some Internal-Flow
Characteristics of Several Axisymmetric
NACA 1-Series Nose Air Inlets at Zero
Flight Speed.
NACA RM L54E19a, 1954.

Supersonic
Su_rsonic

Flight
compression

See references 13, 53, 54, 87 to 97, 100, _02, iii, 116, 125, 127, 177,
181, 182, 189, 190, and 219.
Moeekel, W. E., Connors, J. F., _.
"b_,.e_:"
".,A. H.:
Investigation
Shock Diffusers at Mach Nu_?_:,'l.bS.
ii - Projecting Double-Shock
Ccnes.
NACA RM E6LI3, i94"i.
I

of

Staff of the Computing Section, Center of Analysis, Ur_ler !h?:ction of

Zden_k Kopal:
Tables of Supersonic Flow Around Coneb . _......
..... Rep. l,

M.I.T.,

1947.

LaVallee_ Stanley_ and BiL!m_n; Louis S.:

Pressure Recove.ly
a RamJet Diffuser with a Con!cai Nose and an Annular Inlet.
_ :'_r Rep.
UAC-20, United Aircraft C_rp., Apr. 1948.

.!
']

Schroeder, Albert H., and C_.'moz-s, James F.:

Preliminar
: ?estigation
of Effects of Combustion i_ P_nn Jet on Pe_fonnance
.:_
.... -sonic Dif-

i.
!_

_i

fuse2s.

II - Perforated
Marie A. :

Su_x_-rsonic Inlet.

_.

Burcher,

Compressible

_/

Neice, Mary M.:

Tables and Charts
Shocks.
NACA TN 1673, A948.

NACA R!'..;:

_o._",]%bie._ foc Ai:_

of Flow

:o,1948.

..ACA [.'

Pal_ame_e_s &cross

_;

1592, 1948.

!:

Oblique

L
I_

'

,,,,

,7'_
"

__-

,,,

, ,

....

NACA _MA S_F16

CO;_IDEIk_FIAL

1_7

McLafferty, G. H. : An Estimation of the Momentum Recovery of FSow Lost

4n Perforated Diffusers with Shrouds.
Rep. M-12133-8, Research Dept.,
United Aircr_ t Corp., East Hartford, Conn., Nov. 1949
McLafferty_

George

East Hartford,

i{.:
_Comn.,

A Stepwise Method for Designing Perforated Super_-_O-_,

F,es_a'uh D_95
UniLed Aircraft Corp.,
Nov. 17, 19_9.

Fraenkel, L. E., and Goldsmith, E. L.:

the Perfo_nnance of Conical Supersonic
British, 1949.

A Preliminary
Investigation
Diffusers.
R.A.E. TN Aero.

of
2000,

Schweige2, M_%rvin: Internal FIo_ C,haractezisties

of the MUV-2 Diffuser
9Q
at Supersonic Velocities.
Meteor Rep. UAC-_j,
Research Dept., United
Aircraft Corp., East Ha,_ford, Conn., Apr. 3.949.
Roberts.n, J. M., and Ross, D. : Water Tu_unel Diffuser Flow Studies;
Part II - Experimental Research
Pezm. State College, School of Eugr.,
Ordnance Research I_b., 1949.
Hunczak_ Hen_r R., and Kremzier, Emll J.: Characteristics
of Perforated
Diffusers at Free-Stream Mach Number 1.90.
NACA RM EgOB02 t 19_0.
McLafferty, George H. : Tests of Perforated Convergent-Divergent
Diffusers
for Multi-Unit Ramjet Application.
Rep. R-53133, Research Dept., United
Aircrai't Corp., East Hartford, Conn., June 1990.

_"

Nicks, O. W., and Pearce, R. B. : Supersonic Diffuser Research.

North
American Aviation, Inc., Los Angeles, CM616 (AL-1097) June "
E_,
_ 19_0.
Pearce,

R. B. :

Comparison of Supersouic-Inlet
Diffuser Tests at M c = 2.8.
American Aviation, Inc., Los Angeles, AL-I027,

(ProjectMX-_70) North
July l, 1950.

Weinstein, M_ynard I. : Investigation

of Perforated Convergent-Divergent
Diffusers _ith Initial Boundary Layer.
NACA RM ESOFI2, 1990.
Pearce, R. B. : Tests of Multishock
AL-1117, North American Aviation,

Diffusers.
Inc., Sept.

(Project MX-770)
21, 19_O.

Rep.

McLafferty, George, and Schweiger, Marvin:

Low-Loss, Perforated,
Convergent-Divergent
Diffusers for Multi-Unit Application.
(For prcsentation s _ symposit_n on aerodynamics
of ramjet supersonic inlets,
Wright-Patterson
Air Force Base, Oct. 3-4, ]9_0)
Dnited Aircraft Corp.,
East Hartford, Conn., Research Dept.

McI_fferty, George:
Tests of Unit and Multi-Unit Perforated Diffusers
at Math Numbers up to 3.0. Rep. R-93372-8, Res. Dept., United Aircraft
Corp., East Hartford, Co?re., Dee., 19_0.

/i

ll]

III

i i

iit[llljl[i

'

4
ge

ee

@9#
t

@_e

c
fl

McLefferty_
Geor_:
A Stu_
cf Perforation
Confi_narations
for Supersonic
b_IfTusers. Rep. R-53372-7, Research Dept. United Aircraft Corp. East

Hartford,Con,. Dec.1950.
P_e,P_ioi_ _.: =
Cuw].%n._
DiffUser.
Ru_,

P.:

Recen% RemAits Obtained with the Streamlined

Applie_ Ph'rsicsLab., Johns Hopkins _niv., 1950.

Tw_-Dimensionsd Symmetrical Inlets With External Compression.

R_A _ 1_79,1950.
_/_

@old_mith_
of S_
Centre
E. L., _Fraenkel,
Diffusers
L. E.,
at Supersonic
and Griggs,Speeds.
C. F.:

R.A.E.
The Performance
Bep. No.

A_z_.237Z_;_ae1950.
Esenwein, Fred T., 8_ Valerino, Alfred S. : Force and Pressure C_aracteristies for a Series of Nose Inlets at M_ch Numbers from 1.99 to 1.99.
I -Conical Spike All-External Compression Inlet with Subsonic Cowl Lip.
NAEA RM ESO_,
i_I.
Obery_ L_ J._ _u_ Englert, G. W. : For=e _
Pressure C'nar_cteristics
for a Series Of Nose Inlets at Math Numbers from 1.59 to 1.99. II Isentrupie-Spike All-External Cc_p:ession Inlet. NACA RM ._OJ_,

95_
Welnstein, N_Tuard I._ and Davlds_ Joseph: Force and Pressure Characteristics for a Series of Nose Inlets at Mach Numbers from 1.59 to 1.99.
III- Conical-Spike _ll-External-Conpression Inlet with Supersonic Cowl
Lip. _ACA _ E_C_T30,1951.
Baud,
L. Eu_.ne_ and Gould, Lawrence I. : Investigation of Three _5_es
of Supersonic Diffusex-sOver a Range of Mzch Numbers From 1.75 to 2.7_.
NACA RM E_X)8, 1951.
.

Sears, Richard I., and Merlet, C. F. : Flight Determination of the Drag

and Pressur_ Recovery of an NACA 1-_O-250 Nose Inlet at Macb Numbers
frum 0.9 to 1.8. NACA RM LgOL18, i991.
..

Fox, Jerome L. : Supersonic 9Anne! Investigation by Means of InclinedPlate Technique to Determine Performance of Sev_r_l No_e Inlets Over
M_ch Number Range of 1.72 to 2.18. NACA RM ESOKI_ 19_I.
PeaTJ_onok_ Eugene_ and Farley_
John H.:
Internal
Flow and Burning Characteristics
of 16-Inch Ram Jet Oper_tinE in a Free Jet at Mach Numbers
of 1-35 arK11.73. NAEA R_ E51C16, 1951.

Moeckel_ W. E. : l_._owSeparation Ahead of Blunt Bodies at Supersonic

Speed_.

NACA _

2_18, 1991.

McLafferty, George: Tests of Perforated Diffusers for Multi-Unit Ra_et

App!icatlon-Ph_se IV. Rep. R-93_16-II 2 Research Dept., United Airmra_
Corp. _ E_rtford, Conn., Se_t. 3951.

Moeckel t W. E. t en_ Evans t P. J'1 Jr.: Preliminary Investigation of Use

of Conical Flow Sel_/ation for Efficient Supersonic Diffusion. NAEA
RM E_IJ08, 1931.
Fraenkel, L. E. : Same Curves for Use in Calculations of the Performance
of Conical Centrebody Intakes at Supers_ic Speeds an_ at Full M_ss
Flow. R_.E. _ No. Aerb. 2135, British, Dec. 1951.

./-

Beasts/i t D., _
Turnert J. : The Effect of a Spike _otzuding in Front
of a Bluff Body at Supersonic Speeds. R.A.E. TN No. Aero. P137, E_itish,
Jan. 1952.
Jones, Jim J.:
Number 2.72.

Flow Separati,_ frc_ Rods Aheaa of Blunt Noses at Math

NACA RM _i,

1952.

Cortrightt Edgar M., Jr., and CcJa-_rst James F.: Survey of Some Preliminary Investigations of Supersouic Diffusers at _igh MAch Numbers.
NACA RM E52E20, 1952.
F_ir, W. A. : Experimento on Sel_L_ationof Bou_
Layers on Probes in
Front of Blunt-Nosed Bodies in a Supersonic Air Stream. Phil. Mag. t
.

ser. 7, vol. 43, no. 342, Jul_:1952, pp. 695-716.

//!

Connors, James F._ and Woollett, Richard R.: Performance Characteristics

of Several Types of Axially Symmetric Nose Inlets at M_h Number 3.85.
NACA Z_ E52ZlS, 1992.
Beke t Andrew, and Allen, J. L. : Force and Pressure-Recovery Characteristics of a Conical-Type Nose Inlet Operating at M_ch Numbers of 1.6
to 2.0 and at Angles of Attack of 9. NACA RM E92130, 1952.
Selna_ _._es, Bright, Loren G., and Schlaff t Bernar_ A. : Imvesti_ation
of _Ts_ and Pressure Reccvery of a Scoop Inlet in the Transonic S_
._ange. _ACA _ A92F27, 19_2.
Evvard_ John C., and _slen_ S_ephen H. : Three-Dimension_l Supersonic
Nozzles and Inlets of Arbitrary Exit Cross Section. NACA T_ 2688,
1952
Bernstein, Harry, and Haefeli_ RudolPh C.:
.

Investi_tion

of _Prc_ure

Recovery of a Single-Conical-Shock Nose Inlet at M_ch Number 9.4.

N_A RM ES_A/2, 1993.

"_

//

Haefeli, Rudolph C.t and Bernstein, Harry: _erforms_ce of Sepaz_tion

Nose Inlets at Mach Number 9.9. RAfiARM E53123, 1953-

_.

':..

| _ ',

/
,

I_ A55mU.6

Sears, Richard I. : Some Considerations

Concerning Inlets and Ducte_
Bodies at M_ch Numbers from 0.8 to 2.0.
NACA RM L5312_b, 1953.
Hearth, Donald P._ and Gorton 2 Gerald C. : Investigation
at Supersonic
Speeds of an Inlet Employing Conical Flow Separation from a Probe Ahes_
of a Blunt Body.
NACA RM E52KIS, 19_3.

h/

Lukasiewicz,
J. : Supersonic Ramjet Performance.
vol. XXV, no. 296, Oct. 1953_ PP. 298-306.

Aircraft

Engineering,

Wallace, Donald A. : Report of Tests on .09 Scale Scoop Duct Model.

Chance
Vought Research and Development
Program at M_ch Numbers 1.6 s 1.8, 2.0
and Transonic.
Rep. USCEC 24-i-8_ Engr. Center, Univ. of Southern
California, Dec. 24, 1953.

_,/-

Bernstein,
Harry,
and Humber
Haefeli,of Rudolph
C. : RMPerforn_Ance
of Isentropic
Nose Inlets
at M_ch
9.6.
HACA
E_B24_
1994.
Hunczak, Henry R. : Pressure Recovery and M_ss-Flow Performance
of Four
An_Llar Nose Inlets Ope=_ting in Hach Number Region of 3.1 and Reynolds
Number Range of Approximately
0.49xi0 e to 2.2CKI0 e. NACA RM E94A07,

195_

Connors, James F., and Woollett, Richard R.:

Force, Moment, and Pressure
Characteristics
of Several Annular Nose Inlets at _ch
Number 3.89.

NACARM E93J09,1,994.
_rsh,
B. W._
sonic Ra_et

and Sears, G.A.:

Introduction
Power Plants.
Jet Propulsion,

to the Analysis of Supervol. 24, no. 3, May-June

1954,_. 199-161.
0ffenlm,rtz, Edward:
An Experimental
Investigation
of Two-Dimensional t
Suporsonic Cascade-Type
Inlets at a _ch
Number of 3.11.
NACA RM L54E17,

195_.
Lean, G. H. : Report on the Flow Phenomena at Supersonic Speed in the
Neighbourhood
of the Entry of a Propulsive Duct.
R. and M. 2827,
British A.R.C., 1994.
(Also issued as:
A.R.C. Engine Aerodynamics
and Propellers, Sub-Comm. 11,868, 1949)

Limiting

See reference

internal

contraction

and limiting

inlet Mach

number

113.

,_

......

,..-:__

[ ..........................................

OF

Connors, J. F., sad Schrocder, A. H. : Preliminary

Investigation
of
Effects of Combustion in Ram Jet on Performance
of Supersonic Diffusers.
I - Shock Diffuser With Triple-Shock
Projecting Cone.
NACA RM ESFIS,

Schroeder, Albert H., sad Connors, JE_=.S F. : Preliminary

Investigation
of ETfects of Combustion in Ram Jet _ Performance
of Supersonic
Diffusers.
III - Normal-Shock Diffuser.
NACA RM ESJI8, 1948.
Neumsan, E. P., and Lustwerk, F. : Supersonic Diffusers for Wind
Jour. Appi. Mech., vol. 16, no. 2, June 1949, pp. 195-202.
Stoolman, Leol and Frsacis, Donald
With and Without C_mbustiom.
_
Sept. 18, 1990.

Tunnels.

L. : Supersonic Diffuser Perfo_msace

Preprint, Jet Prop. Lab., C.I.T.,

Comnors, James F., and Woollett, Richard R. :

at the Throat of a Two-Dimensional
Diffuser

Some Observations
of Flow
at a Maeh Number of 3.89.

NACARM E52104,1952.

Himka, _eodore:
Methods of Starting Scoop-Type Inlets.
Wright Aero.
Rep. No. 1692, sec. G, Supersonic Inlet Symposium, Curtiss-Wright
Corp.,
Wright Aero. Div., Wood-Ridge, New Jersey, Jan. 23, 1953, PP. 67-80.

Bouadar_-la_er

See references

shock-wave

interaction

103 to 107.

Fage, A., and Sargent, R. F. : Shock-Wave and Bou_'-Layer

Phence_.na
N_r
_ Flat Surface.
Proc. Roy. Soc. (l_om),
ser. A., vol. 190,
no. 1020, June 17, 1947, PP. 1-20.
Lees, L. : Remarks on the Interaction Between Shock Waves _
Bo_
Layer in Transonic and Supersonic Flow
Rep. 120, Aero. Engr. Dept.,
Princeton Univ., Nov. i, 1947.
Pearce, R. B. : Shock Waves on Surfaces
Rep. AL-399, North American Aviation,
Dec. 30, 1947.

Ackeret,
Shocks

m4 .

"

with Thick Boundary Layers.

Inc._ Los Angeles, Calif.,

J., Feldmsan, F., Rott, N.:

Investi_tions
of Ccm_ressioa
an_ Bounty
Layers in G_ses Moving at wigh Sl_ee_. NACA _

Weise, A.:
On the Selmla_tion of Flow Due
NACA TM 1152, 1947.

to Cxm_resslbillty

8hock.

1113,

i
i

152

..

e e

@ g

(io

Q@

' ' NACA BM A55F16

o

_|g
v

Zalovcik, John A., and Luke_ Ernest P.: Some Flight _asurements of
Pressure-Distribution and Boundary-Layer Cha_-_cteristicsin the Presence
of Shock. NACA I_ML8C22_ 1948.
Lukasiewlcz, J. : Conical Flow as a Result of Shock and Boundary Layer
Interaction on a Probe. R. and M. 2669, British A.R.C. 12,023, Sept.
1948.
0utmsn, Vernon, and Lambert, Arthur:
Sci., vol. 15, 1948, pp. 671-67]_.

Transonic Separation.

Jour. Aero.
_

Howarth_ L. : The Pro_cion

of Steady Disturbances in a Supersonic
Stream Bounded on One Side by a Parallel Subsonic Stream. Proc. Cambridge Phil. Soc., vol. 44, pt. 3, 1947, pp. 380-390.
Lees, Lester: Interaction Between the Iamin_r Bo_udaIS"Layer Over a
l_e
Surface and an Incident Oblique Shock Wave. Rep. 143, Aero.
Engr. Lab., Princeton Univ., Jan. 24, 1949.
Neuman, B. G. : The Re-Attachment of a Turbulent Boundary-Layer Behind
a Spoiler. Rep. A 64, Dept. of Supply and Development, Aero. Research
Lab., Melboulme, Australia, Oct. 1949.
Tsien, H. S., and Finston, M. : Interaction Between Parallel Streams of
Subsonic and Supersonic Velocities. Jour. Aero. Sci., vol. 16, no. 9,
Sept. 1949, pp. 515-528.
Barry, F. W., Shapiro, A. H., and Neumann, E. P. : The interaction of
Shock Waves with Boundary Layers on a Flat Surface. Meteor Rep. 52_
M.I.T., M_r. 1950.
Robinson, A. : Wave Reflection Near a Wall.
nautics, Cranfield, Eugland, May 1950.

i
I

Rep. 37, Co]lege of Aero-

Bardsley, 0., and M_ir, W. A. : The Interacton Between an Oblique Shock

Wave and a Turbulent Boundary Layer. Phil. Mag., set 7, vol. 42,
Jan. 1991, pp. 29-36.

Barry_ F. W., Shapiro, A. H., and Neumann, E. P. : The Interaction of

Shock Waves With Boundary i_ayerson a Flat Surface. Jour. Aero. Sci.,
vol. 18, no. k, Apr. 1991, pp. 229-238.
Stewartso_ K. : On the Interaction Between Shock Waves and Boundary
I_yers. Proc. Cambridge Phil. Soc., vol. 47. pt. 3, July 1951,
_945-5_3.

:
s
j
_L

Cooper, George E., and Bray, Richard S. : Schlieren Investigation of the

Wing Shock-Wave Boundsx_-Layer Interactxon in Flight. NACA RM ASIG09,
1991"

"

',

'

..........

,,._,

'" _._

111 J

. I

Ill

II

II

' _

"

NACA _4 A59FI6

GONFIDE_AL

193

Bogdonoff, S. M., and Solarski, A. H.: A Prelimlna_r Investigation of

a Shock Wave-Turbulent Boundary Layer Interaction. Rep. 184, Princeton
Univ., Aero. Engr. Lab., Nov. 30, 1991.

Crocco, L., and Lees, L. : A Mixing Theory for the Interaction Between
Dissi__tlve Flows and Nearly-Isentropic Streams. Rep. 187, l_nceton
Univ., Aero. Engr. Dept., Jan. 15, 1992.
Johannesen, N. H. : Experiments on Two-Dimenslonal Supersonic Flow in
Corners and Over Concave Surfaces. British A.R.C. Fluid Motiou SubComm., 14,607 - F.M. 1669, Jan. 29, 1992.
Gszld,G. E., and Holder, D. W. : The Interaction of an Oblique Shock
Wave with the Boundary Layer on a Flat Plate. Part I - Results for
M = 2. British A.R.C. Fluid Motion Sub-Comm., 14,8_8 - F.M. 1714,
Apr. 24, 1992.
Gadd, G. E. : On the Interaction with a Complete]_ Laminar Boundary L_ver
of a Shock Wave Generated in the Mainstream. British A.R.C. Fluid
Motion Sub-Comm., 15,100 - F.M. 1770, Aug. l, 1952.

"

Barry, F. W.: A Review of Exper_,,entalResults on Boundary Layer Shock Wave Interaction. (Project MX 770) Rep. AL-1599, North American
Aviation, Inc., Downey, Calif., Dec. 19, 1952.
Cope, W. F. : The Measurement of Skin Friction in a Turbulent Bound"
Layer at a Mach Number of 2.9, Including the Effect of a Shock Wave.
Proc_ Roy. Sot., ser. A, vol. 919, 1952, pp. 84-99.
Liepmann, H. W., Rosbko, A., and Dhawan, S. : On Reflection of Shock
Waves from Boundary Layers. NACA Rep. llO0, 1952. (Supersedes NACA

TN 2334)
Barrin, E. N. : A Flight Investigation on the Effect of Shape and Thickness of the Boundary Layer on the Pressure Distribution in the Presence
of Shock. NACA TN 2765, 1952.
Gadd, G. E. : A Semi-Emplrical Theory for Interactions Between Tumbulent
Boundary Layers and Shock Waves Strong Enough to Cause Sel_ration.
British A.R.C. Fluid Motion Sub-Comm., 15,543 - F.M. 1849, Jan lO, 1953.
Gadd, G. E., Holder, D. W., and Regan, J. D. : The Interaction of an
Oblique Shock Wave with the Boundax_ Layer on a Flat Plate. Part II Interim Note on the Results for M = 1.5, 2, 3 and 4. British A.R.C.
Fluid Motion Sub-Comm., 15,591 -F.M. 1855, Jan. 30, 1953.

Bogdonoff, S. M., Kepler, C. E., and Saulorenzo, E. : A Study of Shock

Wave Turbulent Boundary La_er Interaction at M = 3- Rep. 2'2_,
Princeton Univ., Aero. Engr. Dept._ July 19_3.
"

._mmmdmmmmmMmb,_

.-ab

.....

154
::

: .....

'.: ..: :

......

iil

Kepler, C E., and Bogdonoff, S. M. : Interaction of a Turbulent Boundary

Layer with _ Step at M = 3. Eep. 238, Prlnceto_.Univ., Forrestal
Research Center, Sept. l, 1953.
Gadzi,G. E. : Interactions Between Wholly Laminar or Wholly Turbulent
B_"
La_rs and Shock Wzve_ Strong _k__ough
to Cause Separation.
Jour. Aero. Sci., vol. 20, no. ll, Nov. 1953, PP- 729-790.
Karrln, E. N. : A Flight Investigation of Laminar and Turbulent Boundary
L_yers Passing Through Shock Waves at Full-Scale Reynolds Numbers.
N_CA TN 3056, 1953.
Lighthill, M. J. : On Boundary Layers and Upstream Influence; Part I A Comlmmison Between Subsonic and Supersonic Flows. Proc. Roy. Soc.,
vol. 917, 1953, PP. 3_J_-357
Lighthill, M. J. : On Boundary Layers and Upstream Influence, Part II,
Supersonic Flows with Separation. Proc Roy. Soc., vol. 9-17,1953,
PP. _78-_07.
Dry,gEe, Georg: An Experimental Investigation of the Influence of Strong
Adverse Pressure Gradients on Turbulent Boundary Layers at Supersonic
Speeds. Flygtekniska Forsoksanstalten, Stockholm.
Lange, Roy H. : Present Status of Information Rel_tive to the Prediction
of Shoak-Induced Boundary-Layer Separation. NACA TN 3069, 1994.
Lip design

See references 23, 132_ 197, and _99.

,.
Nichols, Mark R., and Pendley, Robert E. : Performance of Air Inlets at
,.
Transonic and Low Superscnic Speeds NACA }_4L52A07, 1992.
IIolzhauser,Curt A. : The Effect of Entrance Math Number and Lip Shape
on the Subsonic Characteristics of a Scoop-Type-Air-Induction System
for a Supersonic Airplane. NACA RM AglJ19a, 1992.
Dennard, John S., and Nelson, William J. : Preliminary Investigation of
the Effect of Inlet Asymmetry on the Performance of ConvergingDiverging Diffusers at Transonic Speeds. NACA _4 L92J20, 1992.

M_ss-flow variation

NACA _4 A55FI6

CONFIDEN"3T_AL

155

Hayes, Clyde: Preliminary Investi_tion of a Variable Mass-Flow Supersonic Nose Inlet. NACA RM L9JII, 1949.
Allen, J. L., and Beke, Andrew: Force and Pressure Recovery Characteris tics at Supersonic Speeds of a Conical Spike Inlet with a Bypass
Discharging from the Top or Bottom of the Diffuser in an Axial Direction.
_0A _M ESBA29, 1953.
Hinners, Artht_rH., Jr., and Lee, John B. : Preliminary Investigation of
the Total-Pressure-Recovery Characteristics of a 15 Semiangle MovableCone Variable-Geometry Ram-Jet Inlet at Free-Jet Math Numbers of 1.62,
2.00, 2.53, sad 3.05. NACA RM LO2KlO, 1953.

_-

Beke, Andrew, and Allen, J.L.:

Force and Pressure-Recovery Characteristics at Supersonic Speeds of a Conical Nose Inlet with Bypasses Discharging Outward from the Body Axis. NACA _4 E52L18a, 1953.
Angle of attack
See references 26, 129, 130, 131, 133, and 198.

Beke, Andrew, and Allen, J. L. : Force and Pressure-Recovery Characteristics of a Conical-Type Nose Inlet Operating at Mach Numbers of ].6
to 2.0 and at Angles of Attack to 9. NACA RM E52130, 1952.
Comenzo, Raymond J., and Mackley, Ernest A. : Effect of Yaw and Angle of
Attack on Pressure Recovery and Mass Flow Characteristics of a Rectangular Supersonic Scoop Inlet at a )_ch Number of 2.71. NACA _4 L54G22,
1954.
DRAG
Subsonic FIight
See references 7e, 76, and 80.
Kuchemann, D., and Weber, J., eds. : Je Power-Unit Ducts. Ministry of
Aircraft Production, Volkenrode, VC Sv (Rep. and Trans. 987), June i,
1948, AVA Monographs, A. Betz, gen. J.
Nichols, Mark R., and Kelth, Arvid L., Jr.: Investigation of a Systematic Groap of NACA 1-Series Cowlings With and Without Spinners. NACA
Rep. 950, 1949. (Supersedes NACA _4 LSAIs)
Pendley, Robert E., and Rob_mson, Harold L. : An Investigation of Several
NACA 1-Serles Nose Inlets With and Without Protruding Central Bodies
at High-Subsonic Mach Numbers and at a Mach Number of 1.2. NACA
RM LL23a, 1950.
Booker, John Vernon: Wind-Tunnel Investigation of Air Inlet and O_tlet
Openings on a _treamline.Body..,
.,_._
R.el__.:.
i_038, 1951.

/
!

.
t

.
LII

.
i

Ii

ie

NACA }94A_gF16

Selna, James, Bright, Loren G., and Schlaff, Bernard A.: Investigation
of Drag and Pressure Recovery of a Scoop Inlet in the Transonic Speed
Range. NACA RM A92F27, 1952.
Cole, Richard I.: Pressure Distributions on Bodies of Revolution at
Subsonic and Transonic Speeds. NACA RM L_2D30, 1952.
Pendley, Robert E., Mili12o, Joseph R., Fleming, Frank F., and Brym_,
Carroll R. : An Exper!mental Study of Five Annular Air Inlet Configurations at Subsonic 8_idTransonic Speeds. NACA RM L53F18a, 19_3.

Supersonic Flight
External wave drag with no spill_age

See references 1343 150 to 171, 17R, 180, and 213.

Lighthill, M. J. : Supersonic Flow Past Slender Pointed Bodies of Revolution at Yaw. Quart. Jour. Mech. and Appl. Math., vol. I, pt. l,
Y_r. 191L8,pp. 76-89.
Stone, A. H. : (_,Supersonic Flow past a Slightly Yawing Cone.
Math. a_d l_s., vel. XXVII, no. l, Apr. 1948, pp. 67-81.
Lighthill, M. J. : Supersonic Flow Past Bodies of Revolution.
2003, British A.R.C., Jan. 3, 1949.
i
!

Jour.

R. and M.

Munk, YzlxM., and Crown, J. Conrad: The Head Shock Wave. Proc. 7th
Int. Congr. Appl. Math., vol. 2, pt. 2, Sept. 1948, pp. 470-_84.
Broderick, J. B. : Supersonic Flow Round Pointed Bodies of Revolution.
Quart. Jour. Yech. and Appl. Math., vol. II, pt. l, Mar. 1949,
pp. 98-120.
Graham, Ernest W. : The Pressure on a Slender Body of Non-Uniform CrossSeetionsl Shape in Axial Sup=rsonic Flow. Rep. SM-13346K, Douglas
Aircraft Co., Inc., Santa Monlca, July 20, _949.

Moeckel, W. E. : Use of Characteristic Surfaces for Unsymmetrical Supersonic Flow Problems. NACA %_ 1849, 191'_9.
Basemann, Adolf: A Review of Analytical Methods for the Treatment of
Flow with Detached Shocks. NACA TN 1898, 1949.
i
!

Nucci, Louis M. : The External-Shock Drag of Supersonic Inlets Having

Subsonic Entran;e Flow. NACA RM L_0G14a, 1950.
....

,-,_"-_4, _

,..

I
i

_.j

"-

NACA EM A55F16

CO_,FIDF_NTIAL

157

Stoney, William E., Jr.: Pressure Distributions at Mach Numbers From

0.6 to 1.9 Measured in Free Flight on a Parabolic Body oi'Revolution
With Sharply Convergent After_ody. NACA }94LSIL03, 1952.
Rossow, Vernon J. : Applicability of the Hypersonic Similarity Rule to
Pressure Distributions Which I_elude the Effects of Rotation for Bodies
of Revolution at Zero Angle of Attack. NACA TN 2399, 1951.
Fraenkel, L. E.: The Theoretical Wave Drag of Some Bodies of Revolation.
R.A._. _,_..._._o.2420, May 1951.
Ringleb, F. 0.: Theory and Application of the _low Over a Cusp.
Princeton Univ., Dept. of Aero. Engr., Mar. 1952.

Rep. 192,

Ferrarl, Carlo: The Determination of the Projectile of Minimum WaveResistance. RTP Trans. ll80, British Minlstry of Aircraft Production,
Sept., 1939.
Fraenkel, L. E.: Supersonic Flow Past Slender Bodies of Elliptic CrossSection. R.A.E. Rep. Aero. 2466, British, May 1952.
Additive drag
See references 12, 146, 173, and 174.
Chan6e in external wave dra_

See references 23, 146, 153, 162, 167, 173 to 176, and 178.
Lighthill, M. J. : The Position of the Shock-Wave in Certain Aerodynamic
Problems. Quart. Jour. Mech. and Appl. Math., vol. I, pt. 3, Sept.
1948, pp. 309-318.

_
:'
t

Moeckel, W. E. : Experimental Investigation of Supersonic Flow With

Detached Shock Waves for Mach Numbers Between 1.8 and 2.9. NACA
RM ESOD05, 1950.
Keberle, Juergen W., Wood, George P., and Gooderum, Paul B.: Data on
Shape and Location of Detached Shock Waves on Cones and Spheres.
NACA TN 2000, 1950.
Moeckel, W. E. : Flow Sel_ration Ahead of a Blunt Axis/.laSymmetric Body
at Mach numbers 1.76 to 2.10. NACA RM E51125, 1951.

Lip bl_tnes _

See references 23, 167, and 17_.

.....

-..........

-.--, v / .......,,,.

158

I1

,1

4_

dt

.....

.............

G
_
t

lie

......

I_

A95F16

Net wave dra_

See references 23, 98, 100, i16, 120, 148, 174, and ]76 to 182.
Kinghorn, George F., and Disher, John H.: Free-Flight Investigation of
16-Inch-Diameter Supersonic Ram-Jet Unit. NACA _ ES&26, 1928.
Ferri, Antonio, and Nucci, Louis M.: Theoretical and Experimental Analysis of Low-Drag Supersonic Inlets Having a Circ_ularCross Section and
a Central Body at Mach Numbers of 3.30, 2.75, and 2.45. NACK RM L8H13,
1948.
Fendley, Robert E., and Smith, Norman F. : An Investigation of the Characteristics of Three NACA 1-Serles Nose Inlets at Subcritical and Supercritical Mach Numbers. NACA RM LSL06, 1949.
Esenwein, Fred T., and Valerino, Alfred S_ : Force and Pressuze Characteristics for a Series of Nose Inlets at Mach Numbers from 1.59 to 1.99.
I - Conical-Spike All-External Compression Inlet with a Subsonic Cowl
Lip. NACA RM E50J26, 1951.
Messing, Wesley E., and Aeker, Loren W. : Transonic Free-Flight Drag
Results of Full-Scale Models of 16-Inch-Diameter Ram-Jet _gines.
NACA _ Eg_Ulg, 1972.

i
t
!

!
!

Messing, Wesley E., and Raob_ Leo._rd: Transonic Free-Flight Investigation of the Total Drag aud of the Component Drags (Cowl Pressure,
Additive, Base, Friction, and Internal) Encountered by a 16-InchDiameter Ram-Jet Engine for Mach Numbers from 0.80 to 1.43. NACA

RM E52F02,
1992.
Allen, J. L., mad Beke, Andrew: Force and Pressure-Recovery Characteristics of a Conical-T_e Nose Inlet Operating at Mach Numbers of 1.6
to 2.0 and at Angles of Attack of 9 . NACA RM E52130, 1952.
Allen J. L., and Beke, Andrew: Force and Fressure Recovery Characteristics at Supersonic Speeds of a Conical Spike Inlet with a Bypass
Discharging from the Top or Bottom of the Diffuser in an Axial Direction.
NACA RM E53A29, 1953.

Connors, James F., and Woollett, Richard R. : Force, Moment, and Pressure
CharacSeristlcs of Several Annular Nose Inlets at M_ch Number 3.87.
NACA EM EgSJ09, 1954.

i
i
i

0bery, Leonard J., Stitt, Leonard E., and Wise, George A. : Evaluation
at Supersonic Speeds of Twin-Duct Side-Intake System with TwoDimensional Double-Shock Inlet_. NACA RM E51_C08,1954.
.,

ilF
CO!_$_EI_IAL

NACARM A55Fi6

159

Merchant, D. G. : A Collection of Dl_g Data for Wings and Bodies at

Supersonic Speeds. Tech. Note HSA __6, Aero. Res. Labs., Dept. of
Supply, Australia, Aug. 1954.
Merlet, Charles F., and Putlsmd, Leonard W. : Flight Determination of
the Drag of Conical-Shock N_se inlets with V_iolls Co.-]_ngShape a and
Axial Positions of the Center Body at Math Numbers from 0.8 to 2.0.
NACA RM L54G21a, i954.

FLOW STEADINESS
i

Subsonic Flight
Choked _

See references 72 and 73_

Duct rumble

See references 183 *o 186.

Twin-duct instability

See references 130 and 187.

Supersonic Flight
Causes of unsteadiness

See references 50, 51, ill, 189 to 191, and 199.

Character of tu.steadiness

See references 37, 38, 52; 99, 108, 130, and 191 5o 196.
/

Dailey, C. L.: Comments on the Subcritical Buzz Phenomenon Encountered

with Supersonic Ram Jet Diffusers. USCA_ 5-1-16, Aero. Lab._ Univ. of
Southern California, June i, 1948. (Navy Contract NOa(s)9242.)

,:. t

t , 1141
, it

_--

lw_i

in

-m_

' /

' ,f' ,

/
/

Co_uors, Jeme_ F., _

Schroe_r I Albert
H. : EXl_nt_en_ml
Investi_tlon
of Pressure Fiuatuations in 3.6-Xnch Ram Jet at Mach Number 1.92.

Orlin, Wo J., and Dunsworth, L. C. : A Criterion for Flow Instability in

9u_rsoale Diffuser v_n!et_. (._ode!._Y/_qM.-N-6
_
XE -_ ua 1 _ =_
5!4h

(coat_etswoa(s)9_o3
_

AF 33(038)-L_31)
mmua_t

Ai_t

Co.,

Van Nuys_ Calif.I Apr. 2_ 19_i.

Orlin W. James: The Flow Instability Problem in Su_ersonlc Ramjet Engines.
Rep. _20_, Marquardt Aircraft Co., Van Nuys, Calif._ Nov. 20, 1951.
TTimpi, Robert L. : An Analysis of Buzzing in Supersonic Ram Je_s by a
M_Kllfie_One-Dimensional Nonstations_y W_ve Theory. NACA RM L_2AI8,

1952.
Sterbentz, William H., and Davids, Joseph: Amplitude of Supersonic
Diffuser Flow Pulsations. NACA _ E_2Y2_ 19_2.
Dailey, C.L.:

Supersonic Diffuser Instability.

Ph.D. Thesis, C.I.T._

z_.
Prevention of unstead/ness

See references _0, 51, iii, 123_ 129_ 189; and 199.

_
"
"

Fisher, R. E. : Cc_trolilng the Subcritlcal Stability of Conical Shock

Inlets. Me_quardt Aircraft Co._ Van Nuys, Calif. (Presented at the
symposium on the Aerodynamics of Ramjet Supersonic Inlets_ WrightPatterson Air Force Base, Ohio)_ Oct. 3 and 4, 1990.
FOx, Jerome L.: Preliminax_ Investi_tion of Helmhol"_.Resona
Damping Pressure Fluctuations in 3.6-Inch Ram Jet at Nach Number 1.90.
NACA RM E51CO5, 1951.

AIRCRAFT-I_DUC_0_
SYSTEM
Effects of Inlet Location
Subsonic flight

_.e references 204 to 208.

Pendley, Robert E. t Robinson, Harold L., and Williams, Claude V.: An

Investigation of Three Transonic Fuselage Air Inlets at Mach Numbers
from 0.4 to 0.94 and at a Mi_chNumber of 1.19. NACA RM L_OH24, 1950.
Thompson, Jim Rogers: Measurements of the Drag and Pressure Distribution
on a Body of Revolution Throu@hout Transition From Subsonic to Supersonic Speeds. P_CA _._Lgu_27,l_O.
Holzhauser, Curt A. : An Experimental Investigation at Subsouic Speeds
of a Scoop-Type Air-Induction System for a Supersonic Airplane. NACA
RM ASIE24, 19_l.
Nichols, Mark R., and Rinkoski, Donald W. : A Low-Speed Investigation of
an AruralarTransonic Air Inlet. NACA TN 2685, 1952. (Supersedes

_CA _M L6JO_)
-Supersonic fli_t

See refe:mnces 172, 209 to 214, and 259.

Induced Effects of Angle of Attack

Bodies

See references 89_ 199, 209, 213, 215, 216, 218 to 220, 224, 232,

233,

and 237.
Allen_ H. Julian: Estimation of the Forces and Moments Acting on Incline_d
Bodies of Revolution of High Fineness Ratio. NACA RM A9126, 1949.
Allen, H. Julian: Pressure Distribution and Some Effects of Viscosity
on Slender Inclined Bodies of Revolution. NACA TN _4,
1950.
Anon. : Design and Aualysis of Three Supersonic Side Inlet Diffuser Models.
Rep. 1795, Wright Aero. Div., Wood-Ridge, N Jp Sept. 22, i953.

See references 127, 186, and 229 to _31.

.......................

_,,-._".'
,'_._"?'":-._:_
. "'_ ' ........
:"_"__'

'

..........

"_,_

, _,,

"_._"
....
_:_
...........
-'_"._:'_."_','-'-_'
_.':,,_v_"..,_

Effects of Forebody Boundary Layer

See references 28, 114, and 232 to 238.

Davis, Walls_e F.s and Goldstein; David L. : Experimental Investigation
at Supersonic Speeds of Twin-Scoop I/act It.letsof Equal Ai-ea. I - _ul
Inlet Enclosing 61._ Percent of the Yeximam Circumference of the Forebody. NACA I_(A7.r27,1948.
Davis, Wallace F., and Goldstein_ David L. : Experimental Investigation
at Supersonic Speeds of Twin-Scoo_ Duct Inlems of Equal Area. II Effects of Slots upon an Inlet Enclosing 61.5 Percent of the Maximum
Circumference of t_heForebody. NACA RM ASCll, 1948.
Davis, Walla_e F., and Edwards_ Sherman S. : Experimen_l Investisation
at Supersonic Speeds of Twin-Scoop Duct Inlets of Equal Area. III Inlet Enclosing 37.2 Percent of the Veximum Circumference of the Fore-

bo_v.HAOA_AS_o_, 1948.
BraJnikoff; George B. : Pressure Recovery &t Supersonic Speeds Through
_
Duct Inlets Situated in a Region of Appreciable Boundary Layer.
II - Effect of an Oblique Shock Wave I_ediately Ahead. of the Inlet.
NACA RM A(_'08,19_-8.
Davis, W"
_].laceF., Edwards, Sherman S., and BraJnikoff_ George B. :
Experimental Investigstion at Supersonic Speeds of Twin-Scoop Duct
Inlets of Equal Area. IV - Some Effects of Internal Duct Shape Upon
an Inlet Enclosing 37.2 Percent of the Forebody Circumference.
NACA _M A9A31, 1949.
".

Wittliff, Charles E., and Byrne, Robert W.: Prellmlna_ Investigation of

a Supersonic Scoop Inlet Derived from a Conical-Spike Nose Inlet.
NACA RM LSiGll, 1951.

Boundary-Layer Removal
Suction

See references 19, 123, 18_, 202, and 240 to 242.

Edwards, Sherman S. : Experimental Investigation at Supersonic Speeds of
Side Scoops Employing Boundary-Layer Suction. NACA RM A9129, 1949.
Piercy, Thomas G._ and Johnson, Harry W. : Investigation at Math Number
2.93 of Half of a Conical-Spike Diffuser Mounted as a Side Inlet with
Boundary-I_yer Control. NACA RM E92G23, 1952.

.,

!
;

_ ._,,,%, . _-.-,"

._:_._
i

._

_.,

..,,,,

'. _

_'_ _

..

,., . . ,,.,
:....
- ., ._,,_,,_,_*,
. ._ ..
I_'--

i-

,,

._..,,,"_',_,_,_._&_]_A',_'_,.._, .

.;.:
.....
c.:_,.,"_:
"

._

.-,

NACA EM A5_16

..

_CONFID__

, _._ :

163

Flax I A. H., Treanor, C. E., and Curtis, J. T.: Stability of Flow in

Air-Induction Systems for Boundary-Layer Suction. Tech. Re,. 53-189,
Wright Air Development Corp., May 1953.

Diversion

See references ll6, 182, 202, ?_18,and 243 to 250.

o

Kochendorfer, Fred D. : Investigation at a Mach Number of 1.90 of a

Diverter-Type _Boundary-LayerRemoval System for a Scoop Inlet. NACA
RM E53DO7, 1953.
Johnson_ Harry W., and Piercy, Thomas G. : Effect of Wedge-Type Bounda/yLayer Diverters on Performance of Ha_-Conlcal Side Inlets at Y_ch
Number 2.96. NACA RM E54E20, 1954.

Submer_ed inlets

See references 20_, 205, and 2.51tv 260.

Mossmau, Emmet A., and Randall, Lauros M. : An Experimental Investigation
of the Design Variables for NACA Submerged Duct Entrances. NACA
RM A7130, 1948.
Delany, Noel K. : An Investigation of Submerged Air Inlets on a 1/4-Scale
Model of a _/pics/ Fighter-Type Airplane. NACA RM A_O,
19k8.
Axelson, John A., and Taylor, Robert A. : Preliminary Investigation of
the Transonic Charac_ristlcs of an NACA Submerged Inlet. NACA
RM A5OC13, 1990.
Holzhauser, Curt A. : An Experimental Investigation at Large Scale of _A
N_CA _uhmerged Inte_e an_ Deflector Installation on the Resa_
Portion
of a Fuselage. NACA RM A50FI3, 19_0.
Selna, James: Frel_u_
Investi_tion of a Submerged Inlet and a Nose
Inlet in the Transonic Flight Range With Free-Fall Models. RACA
RM ASIBI4, 1951.

Seddon, J._ and Raney, D. J. : Low Speed Wind Tunnel _xlel Tests of Submerged Air Intakes on the Undersurface of a Delta Wing. R.A.E. Rep.
Aero. 2428, British s July 1951.

!
f

-_

:
c

to:

_
_

I
lit
iI a

ill

Taylor, Robert A. : Some Effects of Si_-Wall

and Fressure Recovery of am NACA Submerged
NACA RM Agll_3a, 1952.

Combined
Scoop

iI
s
ill

;
i

;It

i
i
t c

Modifications
on the Drag
Inlet at Transonic Speeds.

Effects

incremental

drag

7.

See reference

Wakes

See reference

224.

Wise, George A., and Dryer, Murray:

Surface on Flow Field in Vicinity
bers 1.8 and2.0.
NACAITME922_I3,

Influence of a Canard-Type
Control
of Sy_netrical Fuselage at Mach Num1952.

Dryer, Murray, and Beke_Andrew:

Performance Characteristics
of a NormalShock Side Inlet Located Downstream of a Canard Control Surface at Mach
Numbers

of 1.9 and 1.8.

NACA RM E52F09,

1992.

Fradenburgh,
Evan A., 0bery, Leonard J._ and Mello, John F. : Influence
of Fuselage and Canard-Type Control Surface on the Flow Field Adjacent
to a Rearwsa_Fuselage
Station at a Mach Number of 2.0 - Data Presentattcm.
NACAI_4EglK09,
1952.

INDUCTION-SYSTEM

AIRCRAI_

Drag
S_In friction

See references
Seddon,
,

J.:

and separation

185, 256, and 267.

Fuselage

and Air Intake

on a Model of a Single-Ermined
B.A.E. TNAero.
2051, British,
Brodel, Walter:
Theory
TM1267,
1950.

of Plane,

Drag Measurements

Jet Aircraft
May 1990.
Sy_netrical

at LowM_ch

With Exit

Intake

Number

at the Tail.

Diffusers.

NACA
6

r
........
V

Jll

%"

!
....

_CA

-_

: . :_j
"_

RM A95F16

z _

, ' J_ CO_._._.___

Transonic

See

references

2_2

to

169

drag rise

267.

K_chemann, D , ed. : The Installation

of Jet-Propulsion
Units.
Ministry
of Airc_t
Pr_^;"'_
....icn, VG 2., (Rcp. _
Trams 9_.),
_
Oct. I, _I,_
_.,
and VG 240 (Rep. and Trans. 940), Oct. 19, 1947, AVA Monographs,
A. Betz, gen. ed.
Pepper, William B., Jr., and Hoffm8_1_ Sherwood:
Tram.sonic Flight Tests
toCompare
the Zero-Lift Drag of Underslung and S_metrlcal
Nacelles
Varied Chordwise at 40 Percent Semispan o: a 49 Sweptback, Tapered
Wing.
NACA BM LgOG17a, 1990.
Pepper_ William B., Jr., and
Drags Determined by Flight
.Mounted Nacelles in Various
and Body Combination.
NACA

Hoffman, Sherwood:
Comparison of Zero-Lift
Tests at Transonic Speeds of Symmetrically
Spanwlse Positions on a 49 8weptback Wing
RM LglD06, 1991.

Hoffman, Sherwood:
Comparison of Zero-Lift Drag Determined by Flight
Tests at Transonic Speeds of Pylon, Underslung,
and Sy_netrlcally
Mounted Nacellee at 40 Percent Semispan of a 49 Sweptback Wing and
Body Combination.
NACA RM LglD26, 1991.
Pepper, William B.j Jr., and
Drags Determined by Flight
Moun_ed Nacelles in Various
49 Sweptback Wing and Body

Hoffm8_, Sherwood:
Comparison of Zero-Lift
Tests at Transonic Speeds of Symmetrically
Chordwise Positions at the Wing _p
of a
Combination.
NACA RM LglF13, 1991.

_offman, Sherwood:
Transonic Flight Tests to Compare the Zero-Lift Drags
of Underslung Nacelles Varied Spanwlse on a 49 Sweptback Wing and
Body Combination.
NACA BM L92DO4a, 1952.
Hoffman, Sherwood, and Pepper, William B., Jr. : Transonic Flight Tests
to Determine Zero-Lift Drag and Pressure Recovery of Nacelles Located
at the Wing Tips on a _9 Sweptbaek Wing and Body Combination.
NACA
RM LglK02, 1992.
Hoffman, Sherwood, and Mapp, Richard C., Jr. : Transonic Flight Tests
to Compare the Zero-Lift Drags of 49 Sweptback Wings of Aspect
Ratio 3.9.5and 6.0 With and Without Nacelles at the Wing Tips.
NACA
RM LglL27, 1992.
"

"

Hoffman, Sherwood, and Pepper, William B. t Jr. : The Effect of Nacelle

Comblnations
and Size on the Zero-Lift Drag of a 49 Sweptback Wing
and Body Configuration
as Determined by Free-Flight
Tests at Maeh
Numbers Between 0.8 and 1.3.
NACA RM L93E29, 1953.

!
\

.
i

I
i

.
I_
ilt

. .

g
_ fl

eeI

I
,1

. 0 0
el
e _

NACA RM A_gFI6

I r
,Ii

Hopko, Russell N., Piland, Robert 0., and Hall, James R. : Drag Measurements at Low Lift of a Four-Nacelle Airplane Configuration Havlng a
Longitudinal Distribution of Cross-Sectional Area Conducive to Low
Transonic Drag Rise. NACA EM L93E29, 1953.
Holda%_, George H." Com_rison of Theoretical and Experimeutal Zerc-Lift
Drag-Rise Characteristics of Wing-Body-Tail Combinations _ear the Speed
of Sm_d.
NACA RM A93H17, 1993.
Whitcomb, Richard T. : 2_cent Results Pertaining to the Application of the
"Area Rule." NACA BM L93il_, 1993.
Hall, James Rud_rd:
ComlP_rlsonof Free-Flight Measurements of the ZeroLift Drag Rise of S_x Airglane Configurations and Their Equivalent
Bodies of Revolution at Transonic Speeds. NACA RM L53X21a, 1994.
Donlan, Charles J. : An Assessment of the Airplane Drag Problem _t Transonic and Supersonic Speeds. NACA BM L94F16, 1994.

Wave drag
e

See references 261 and 267 to 272.

Leiss, Abra_mn: Flight Measrren_nts at _ch Numbers 1.1 to 1.9 of the
Zero-Lift Drag of a Twln-Engine Supersonic Ram-Jet Configuration_
NACA EM L92V24, 1922.
Whitcomb, Richard T., and Fischetti, Thomas L. : Develooment of a Supersonic Area R_le and an Application to the Design of a Wing-Body Combination Having High Lift-tc-Drag Ratios. NACA RM L93H_la, 1953.

Lift _nd Pitching Moment

Wing leading-edge inlets

See references

225 to 230.

Ruden, P. : Two-Dimenslonal Symmetrical Inlets with External Compression.

NACA TM 1279, 1990.
Br_del, Walter: _%eory of Plane, Sy_netricR1 Inlet Difih/sers. NACA
TM 1267, 1950.
Smith, Norman F. : Hi@h-Speed Investig_tlon of Low-Drag Wing Inlets.
NACA WR L-732, 194_. (Supersedes NACA ACR L4118)

i_

-_
_X,,_..J

....
I

nlllll

II J

III

inin

._.,

IJl,_

NAOA P,MA55FI6

CONFID_B_
J a J

167

Von Doenhoff, Albert E., and Horton, Elmer A. : Preliminary Investigation

in the NACA Low-Turbulence _manel of Low-Drag Airfoil Sections Suitable
for Admitting Air at the Leading Edge. _CA WR L-694, 1942.
Racisz, Stanley F. : Development of Wing Inlets.
(Supersedes NACA ACR L6B18)
Bartlett, Walter A., Jr., and Gor_l, Edwin B.:
of Wing Inlets for a Four-Engine Airplane.

NACA WR L-727, 1946.

Wind-Tunnel Investigation

NACA RM L6LII, 1947.

Perl, W., and Moses, H. E. : Velocity Distributions on Two-Dimenslonal

Wing-Duct Inlets by CoILformal_pping.
NACA Rep. 893, 1948.
Douglass, William M. : Wing-Ramjet Development. USCAL Rep. 3-9, Univ.
of Southern California, Aero. Lab., Navy Pesearch ProJ., June 15, 1948.

Wing root inlets

See references 84, 85, 186, 230, 274, and 275.

Scoops

See references 220, 234, axed276.

Nacelles

See references 81, 207, 211, 249, 272, 273, and 277 to 284.
Hansen, Frederick H., Jr., and Dannenberg, Robert E.: Effect of a Nacelle
on the Low-Speed Aerodynamic Characteristics of a Swept-Back Wing.
NACAIM A8EI2, 1948.
Welsh, Clement J., and Morrow, John D.: Effect of Wing-Tank Location on
the Drag and Trim of a Swept-WingModel As Measured in Flight at Transonic Speeds. NACARMLSOAI9,
1990.

Silvers, H. Norman, King, Thomas, J.1 Jr., and Pasteur, Thomas B., Jr.:
Investigation of the Effect of a Nacelle at Various Chordwise and
Vertical Positions on the Aerodynamic Characteristics at High Subsonic
Speeds of a 45 Wing With and Without a Fuselage. NACARMLOIII16, 1951.

168

......
i f

_ iI

"', dO.,f_._rD__
(

II

I i

", , ' , ....., , _ NACA KM A55F16

iii

iI

i _ iI

I l

Bielat_ Ralph P., Harrison, Daniel E., and Coppolino_ Domenlc A. : An

Investigation at Transonic Speeds of the Effects of Thickness Ratio _nd
of Thickened Root Sections on the Aerodynamic Characteristics of Win6_
with 47 Sweepback, Aspect Ratio 3.9, and Taper Ratio 0.2 in the Slotted
Test Section of the Langley 8-Foot High-Speed Tunnel. NACA EM LgllO4a,
1951.
Spree_mann,Kenneth P., and Alford, William J., Jr.: Investigation of Zhe
Effects of Geometric Change3 in an Undezwlng Pylon-Suspended ExternalStore Installation on the Aerod)r._ic Characteristics of a 45 Sweptback
Wing at High Subsonic Speeds. NA_,A_ L50LIP, 1991.
Jacobsen, Carl R. : Effects of Systematically Varying the Spanwlse and
Vertical Location of an External Store on the Aerodynamic Characteristics
of an Unswept Wing of Aspect Ratio 4 at Mach Numbers of i._i, 1.62, and
1.96. NACA EM L52FI3, 1992.
Kolesar, Charles E. : Transonic sad Supersonic Nacelle Invpstigatlon.
Rep. D-13089, Boeing Aircraft Co., Apr. 24, 1992.
Jacobsen, Carl R. : Effects of the Spanwlse, Chordwise, and Vertical
Location of an External Store on the Aerodynamic Characteristics of a
60 Delta Wing at Mach Numbers of 1.41, 1.62, and 1.96. NACA RM L52H29,

1952.
Silvers, H. Non,nan, and King, Thomas J., Jr. : A Small-Scale Investigation
of the Effect of Spanwlse and Chordwise Positioning of an Ogive-Cylinder
Underwlug Nacelle on the High-Speed Aerodynamic Characteristics of a
45 Sweptback Tapered-in-Thickness Wing of Aspect Ratio 6. NACA I_M

L.5_22,1992.
i
;

Jacobsen, Carl R.: Effects of the Spauwise, Chordwise, and Vertical

Location of an External Store on the Aerodynamic Characteristics of a
49 Sweptback Tapered Wing of Aspect Ratio 4 at Mach Numbers of 1.4i,
1.62, and 1.96. NACA _: L52J27, 1953.
Carmel_ Melvin M. : Transonic Wind-Tunnel Investigation of the Effects
of Aspect Ratio, Spanwise Variations in Section Thlckn ss Ratio, and
a Body Indentation on the Aerodynamic Characteristics of a 45 Sweptback Wing-Body Combinatlon. NACA _ L52L26b, 1953.

APPENDIX B

MEASUREMENTS AND INSTRUMENTATION

See references 289 to 294.

J

_. ............
_-_,,,_\_,_,......
: _,,_-,,_. , '_,. _

.....

,,_,.

__

:_,_

._

,_
_......_ _.;_@_._

;:::

,_
Q:

NACA _

A99F16

O.0_FI.OE_'-_AL

169

Smith, Norman F. : Numerical _;aluatlon of Mass-Flow Coefficient and

Associated Parameters from Wake S1_rveyEquations. NACA 'I_N
-_
q_o_,
__J.7...
T,

Holder, D. W., North, R. J., and Chinneck, A. : Experiments with Static

Tubes in a Supersonic Airstream. Parts I and II. R. & M. 2782,
British A.R.C., 1993.
Luidens, Roger W., and Madden, Robert T.: Interpretation of Bo_mdaryLayer Pressure-Rake Data in Flow with a Detached Shock. I%_CA_ ES0129a,

195o.
Kraushasr, Robert J. : Manometers in Pulsating Systems.
NYU-14, New York Univ., Project SQUID, Aug. 22, 1951.

Tech. Memo.

170

.........
.....
II

Ill

', .....
_' ....

CO__L
I

ll
I

z
I

II

li

II

Ill
I

III

Ill.,

i :"_

---- .......

L'l,_

llli

I I I

NACA _
I

I
I

AS_FI6

NACA F_4A_F16

CONFIDEPTI_!J

171

I.O _

-_

'

I.O p. - p_

.96

.68,

.68

56 ,52

_
.I

.2

.3
.4.5
.6
.7
Free-_.reomMochnumber,Mo

'
8

.9

ID

Figure i.- Variation of tctal-!_ressureratin with flight Mach number for

various ram-recovery ratios.

.._1 L_

|.,

//

<,
i

iiil

, :

3L._

: : ",, ., :': oo: ,," ,,:

Ghokinglimit for

'."

i
t

a shorp-Iip inlet, rof. 14----7

/
0.10
0.15

Me
020

il I X it(l/i

i /

= ,/
: /

,/,
/',S/
',/Z

2.4

jr

(130

'"

/i

__/._
o

.2

i_.
!
_

,.oo

,.--,.,o
A2

#I

_i

il

1.0

12

Mass-flow rotio, m_/m_,

Figure

2.- Variation of =ass-flow ratio based

with mass-_low ratio based or, choked

:. -- .-..,+.

_,...

., _

,'t

on free-stream
properties
inlet properties.

/
{

.,
.,
..
: : .,.
,


a
,...........
. ...
...
# : ::..: .......
: ':: :
kO0

_'

-_ _

""_

.98
.97

1o1 _

0.003
C...__. =4

tOOlI

_---_

....

.,

_ ,

_,

___

..q9

.45

Me
2

l ....

__.e
_

"

I"

p_.
.gt

.9'
LO

L2

1.4

1.6

18

2.0

2.2

i
2.6

2.4

Exit-to-inletarea ratio, As
A2,

(b) _'_'o.oo_ 8

,
,.

Figure 4.- Variatio_ of tatal-pressure ratio with area ratio for varimm
flow co_iti_s
in circular co_ical _iffusers.

._:

'

'

'

, .
....

I Ill

..

. ..,
:; ;'

I I IIIIIII

.- .
.. _.

Ill

. ;.
: .. ,...,..

. ,
,
. . ....

/1/

., .
.,
, '..,
|

_1

,..,:.
,_,, .. _.':" ,, . ,
. =,
._,
,
, ,'_', : ..

'..,,,.'.
,'_,'

-,..

_,,....._"
.,,:._,.,*_ ._..._,

!
/

/
/

178

.... : : ..: _.O_.I_.

f

ee

go

2;

";

7777

: .....:

,:

eo

NACA RM A_FI6
e

"'"..........

Minimum cof_.-, _-_iOJe(h_ '-_- ' L,y possible

separation at normal :_
_, _3)

r77

.2

.l

-,.

I
_,

Figure

8.-

12

To_l-pressure

16
20
Cone half angle, _

ratios

24
deg

for conical

28

32

two-shock

compression.

_: ,

i,,,

, .--

......

..,,_,
*,

,o---'--J'l

IJ

I__J

_ _-,

II

.
I_.ll

>

. ,..

"___

36

180

..........

_)NF_DEI_IAL. ......

.....
.98

me

NACA i_4A95F16

,.'',:

....

2"- "(9"--

o.o

- .9----_

_.___

L_

e= .82

.....

O.
.J.

_-

_
.78 ....

.74 _

.7"0n

Figure i0.- Su_y

M_
Descripiion
(9 1.42 Asymmetric normcI shock nose inlet
13 1.42 Symmetric normalshock nose inlet
<_ 1.50 Normal shock nose inlet with deflector

Ref.
151
131
26

A 1.9:

Half-conical inlet with deflector

133

I_ 1.91 Conicalinlet withpivotedcone

129

,d 1.91
,_) 1.90

129
1.50

Conicolinlet
Vertical-wedge nose inlet

8
12
16
Angleof attack, a,de,]

20

of data for the varistionof total-press,_re

ratio with
angle of attack.

NACA RMA99FI6

.30

C0RFIDEF/T_AL

181

II

182

ee

JO

C.O._.._.ZAL

,e

d.... e

o|

ge

:'_

geu '

es
.., '

,_ (

uw

mCA _ A55n6
I

o8

Az/AM =C

.04

.02
_/,,

Mo-,.2

Iv"

.001

002

.001

-_'--2 '!

.-

002

.002

004

003

o,_>_-

_-

.001

.002

003

.004

Friction coefficient, (3/

_,:o

'_'_

004

M.
:zo

o ,
S
_"

003

A_.O

Friction coefficient, Cf

04

00!

Friction coefficient, Gf

",_:,:_ _,=,,4

__ 2 _rz/_
_

.004

I_
'_.08

003

Mo=,.2

2"

Friction coefficient, Gf

_l<t.,o

"

___..._
zl

-_'-'
.4

I-_,

_t--'_,

I
:

i
Ct =.0025
0
1.5 20
2.5
30
Free-streamMach number,Mo

L
c, .0025
0
1.5 20
25
3.0
Free-streamMach number,Mo

Figure 12.-,Theoretical variation of minimum drag coefficient and optimum

cone half-angle for conical cowls.

.,.,,.

......

'"

;r

i ....

"'

'J'L,'L

:F

NACA RM A55FI6

....11

CO.'W"II_F,
NTZ'd

: *!

s s
As

183

_i;

&

_.e_.

',' //////

_.o

"

_.x

lllJllll_J

o
_

,_

,_

..
CO

J
-.

,It

OO'IUOFPlltOoo6o,ap o^!IlPPV
"-Ii

ii_

i_,"_ ..................

:'_"_ Idr__

..............

'"'_ _'t

_'-*_ ulna, ,_= ................... d

_1_

'

-2.0

'

AI
- I_ --

"

Sharp
Lip
Sharp
13 Sharp
<_Porobolicarc

-- 16 --

....

I_

4.t0
._x
2.00
2.00
12.50

167
Ref.
167
]46
177

S_rp

5.40

23

Sharp
NACA
1-49-30

4.20'
, 148

Blunt

4.20,

Blunt

--I.4-

-----

5.40

-,B

-.6

i
/

,,_S

--I.0

23

"

_1

4.20

......
[
i
I

..

'.

:
k

./

/,

,..... __--'

\
&13

13 S
1_ _

13
....

-.2

.8

IX)

12

1.4

dB -/_
/-""- ".._ ...L ._.. "

L6

L8

2.0

Ref. i74

2.2

:,
',,

2.4

Free-stream Moch number,Mo

Figure 15.- Summary of data for cowl suction force sz_ c_ison
theory.

wit_.

II_CA - IL.,au_q_
71eld,V_.
,,_
'

"

._#;_
......

-._-.._
"L-

_._..._,.
.....
L - __EL_,_I_.

, .,
..............................
__i

-. ,-__;- .'_.,,,J ....

....

_#_,7..
7. ........

.....

"'

, _ , - _ ::.,_,f,''_
L!_,'IIL]_ _
i

'