------- --- __ ...

AEDC·TR·80·35 ARCHIVE COpy

DO NOT LOAN
c:,\

Investigations of Free-Jet Test Requirements

and Techniques with Emphasis on the
Adaptable Jet Stretcher

R. J. Matz and E. M. Kraft

ARO Inc.

April 1981

Final Report for Period October 1, 1978 - September 30, 1979

TECrll~ICl'\L REPORTS
.FILE COP'Y'

Approved for public release; distribution unlimited.

PROPERTY OF U.S. AIR FORCE 3

AEDe TECHNICAL LIBRARY
" ,.

ARNOLD ENGINEERING DEVELOPMENT CENTER

ARNOLD AIR FORCE STATION, TENNESSEE
AIR FORCE SYSTEMS COMMAND
UNITED STATES AIR FORCE
 ..... ~ ;.:.:~_.. - ,.......·:"'"c­
t" '.'1.,

·'·;',,::T'·t;\-}/I-"':""~~~~:::-:-:~~~:'::-~~':""';"':""-7:-"':""':':;'':''''';;;';;;~:';;~+:~~~:-:-:-:-::~~=~::''"i~~~~~~===
..
·.·.,·r~;k.:.:b~~~~~~~_~~_~~~_~~Qi
•
,'<';X, ~;;,~,/:;.

,.

.:, ..
r.
\
~
'.
i.I. ."
I •
i
 UNCLASSIFIED
READ INSTRUCTIONS
REPORT DOCUMENTATION PAGE
I REPORT NUMBER r GOVT ACC ESSION NO. ]
BEFORE COMPLETING FORM
REC,PIEN'''S CATALOG NUMBER

AEDC-TR-80-35
4 TITLE r."d Subtitle) 5 TyPE OF REPORT & PERIOD COVERED

INVESTIGATIONS OF FREE-JET TEST REQUIRE- Final Report - Oct. 1,

MENTS AND TECHNIQUES WITH EMPHASIS ON 1978 - Sent. 30, 1979
6. PERFORMING ORG. REPORT NUMBER
THE ADAPTABLE JET STRETCHER
7. AUTHOR!,) B. CONTRACT OR GRANT NUMBER!,)

R.J. Matz and E.M. Kraft, ARO, Inc.

(a Sverdrup Corporation Company)

9 PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT. PROJECT. TASt<
AREA & WORt< UNIT NUMBERS
Arnold Engineering Development Center/DOT Program Element 65807F
Air Force Systems Command
Arnold AF Station, TN 37389
II CONTROLLING OFFICE NAME AND ADDRESS 12. REPORT DATE

Arnold Engineering Development Center/DOS April 1981

Air Force Systems Command I J. NUMBER OF PAGES

Arnold AF Station, TN 37389 73

14 MONITORING AGENCY NAME &: ADDRESS(1f dlfteumt from ControJ"n~ Olliee) 15. SECURITY CLASS. (of thi~ report)

UNCLASSIFIED
15. DECL ASSI FI C ATI ON DOWNGRADING
SCHEDULE
N/A
16 DISTRIBUTION STATEMENT (of thJs Report)

Approved for public release; distribution unlimited.

17 DISTRIBUTION STATEMENT (of the abstract entered In BloeM. 10, Jf different from Report)

lB SUPPL EMEN T ARy NOTES

Available in Defense Technical Information Center (DTIC).

19. KEY WORDS (Continue on reverse ~;de If n e c e e e erv lind Identify by block number)

inlets subsonic flow

test facilities
test methods
supersonic flow
flow fields
20 A 85 T R AC T (Continue on rever"e ~;de If nece.s.sary and IdentIty by block number)

Preliminary investigations of approaches that will permit

relatively large air-breathing engine/inlet/forebody systems to
be tested in free-jet test facilities were accomplished. Con-
clusions were based upon near field flow properties (Mach number
and flow angle) obtained from inviscid flow computations for
slender, axisymmetric bodies at zero incidence using current
AEDC flow quality design goals as an acceptance criterion.

FORM
DD 1 JAN 73 1473 EDITION OF 1 NOV 65 IS OBSOLETE

UNCLASSIFIED
 UNCLASSIFIED

20. ABSTRACT, Concluded.

Results obtained with subsonic and supersonic free-flight boundary
conditions indicate that shortened forebodies can be used to
reduce the overall length of test installations required for
vehicles equipped with aft-mounted inlets. Bodies immersed in
a supersonic free jet experience unacceptable flow distortions
as a result of (1) bow shock reflections from the constant pres-
sure free-jet boundary and (2) waves emanating from the nozzle
lip because of exit plume static pressure mismatch. The adaptable
jet stretcher can potentially eliminate these disturbances over a
range of test conditions. A rigorous mathematical proof of jet
stretcher convergence to the desired interference-free geometry
was developed for supersonic flow. Convergence was also demon-
strated by a computer experiment for a slender axisymmetric body
in an off-design supersonic jet stretcher. Near field flow
disturbances were reduced to an acceptable level after two re-
adjustments of the jet stretcher geometry.

I\.FSC
"fno:>ld I'I.FS T ......

UNCLASSIFIED
 AEDC-TR-80-35

PREFACE

The work reported herein was conducted by the Arnold Engineering Development
Center (AEDC) of the Air Force Systems Command (AFSC). The results of the research
were obtained by ARO, Inc., AEDC Group (a Sverdrup Corporation Company), operating
contractor for the engine test facilities at AEDC, AFSC, Arnold Air Force Station,
Tennessee, under Project Number E35A-OIA. The authors gratefully acknowledge the
assistance of W. J. Armstrong and G. W. Lewis, ARO, Inc., in the computational activities.
The Air Force project manager was M. L. Laster. The manuscript was released for
publication on May 28, 1980.
 A E DC-TR -80-35

CONTENTS

Page

1.0 INTRODUCTION . . . . . . . . . . . . . . . . . .. 5
2.0 ANALYSIS
2.1 Background 5
2.2 Free-Jet Test Requirements 9
2.3 Flow-Field Computations 10
2.4 Assessment of Jet Stretcher Feasibility and
Implementation Problems 17
3.0 CONCLUDING REMARKS 19
REFERENCES 21

ILLUSTRATIONS

1. Flow-Field Characteristics about a Supersonic Vehicle with Different

Boundary Conditions 23
...
2. Schematic of Near Field Streamline Patterns for a Cylinder in
Crossflow with Various Boundary Conditions 27
3. Integration of the Test Cell and Computational Domain 28
4. Iterative Scheme for the Adaptable Jet Stretcher Concept 29
5. Details of Circular Arc Tangent/Ogive Bodies Used in Computations 30
6. Boundary Conditions and Computer Codes Used in the
Flow-Field Computations 31
7. Flow Conditions Near a l-eal Tangent/Ogive Body with Subsonic
and Transonic Free-Flight Conditions 32
8. Flow Conditions Near a 2-cal Tangem /Ogive Body with Subsonic
and Transonic Free-Flight Conditions 34
9. Flow Conditions Near a 3-cal Tangent/Ogive Body with Subsonic
and Transonic Free-Flight Conditions 36
10. Flow Conditions Near a 2-cal Tangent/Ogive Body with Supersonic
Free-Flight Conditions 38
11. Flow Conditions Near a 3-cal Tangem /Ogive Body with Supersonic
Free-Flight Conditions 40
12. Comparison of Free-Jet Boundaries, Free-Flight Streamlines, and
, . Cylindrical Duct Extensions Considered in the Subsonic
Computations (3-caI Tangent/Ogive, Moo = 0.6) 42

3
A E DC-TR-80-35

Page

13. Flow Conditions Near a 3-cal Tangent/Ogive Body with Moo = 0.6
Free-Flight and Free-Jet Operation 43
14. Flow Conditions Near a 3-cal Tangent/Ogive Body with Moo = 0.6
Free-Flight and Ducted Operation 45
15. Comparison of Free-Jet and Free-Flight Flow Details (50-percent
Blockage 3-cal Tangent/Ogive) 48
16. Flow-Field Details Produced by a 3-cal Tangent/Ogive Body in a
Supersonic Free Jet _ 50
17. Comparison of Flow-Field Conditions Near a 3-cal Tangent/Ogive
Body with Supersonic Free-Jet and Free-Flight Conditions 52
18. Extraneous Waves Introduced by Pressure Mismatch at Free-Jet
Nozzle Exit 54
19. Streamline Trajectories at Various Supersonic Free-Flight Conditions
about a 3-cal Tangent/Ogive Body 57
20. Computed Flow Conditions Near a 3-cal Tangent/Ogive Body with
Iterative Jet Stretcher Adjustments _ 58
21. Characteristics Lines for Second Internallteration 62
22. Preliminary Thoughts on Possible Probe-Type Sensors for Perforated
Adaptable Supersonic Jet Stretchers 63

TABLE

I. Current Free-Jet Design Goals 64

APPENDIX

A. Convergence of the Adaptable Jet Stretcher in Supersonic Flow 65

NOMENCLATURE , .. " " 72

4
 A E DC-TR -80-35

1.0 INTRODUCTION

Tests of full-scale engine/inlet systems in ground test facilities are limited at the present
time because of the large air-processing systems required to achieve flow conditions
representative of flight. Specially contoured ducts or "jet stretchers" have been proposed to
reduce air supply requirements for engine/inlet tests in supersonic free-jet test ficilities.
However, fixed-geometry jet stretchers are unattractive because each test vehicle/test
condition combination theoretically requires a different jet stretcher. Incorporation of
variability - particularly in a feedback-controlled, online, adaptive mode - would greatly
improve the potential usefulness of jet stretchers.

The objective of this study was to determine the requirements for, the alternatives to,
and the feasibility of adaptable jet stretchers for both subsonic and supersonic free-jet
testing. Feasibility was to be established in a rigorous mathematical sense and from
representative flow-field computations for an axisymmetric body in free-flight, free-jet, and
ducted flow environments. Results were also to be used to obtain a preliminary assessment
of mechanical, instrurnentational, and computational requirements and to provide a basis
for outlining additional work that must be accomplished before adaptable jet stretchers can
become a practical test tool.

2.0 ANALYSIS

-. 2.1 BACKGROUND

Problems arising from propulsion system integration in high-performance aircraft and

missiles have plagued the aerospace industry for many years. Investigation of these problems
with full-scale hardware has been hampered by the limited size and/or performance
capabilities of ground test facilities. Some engine/inlet interaction tests have been conducted
with full-scale air-breathing missile systems in free-jet test facilities of the AEDC Engine
Test Facility (ETF). However, tests to date have been limited to relatively small vehicles
having inlets located no more than two or three body diameters (or calibers) aft of the nose
with simulated flight Mach numbers greater than two and angles of attack less than 10 deg.
Under these conditions, it has been possible to locate the forward portion of the vehicle,
including the inlet cowl, within the test rhombus of either planar, variable Mach number, or
axisymmetric, fixed Mach number free-jet nozzles.

AEDC has received inquiries about supersonic, low-altitude tests of ram rocket-powered
missiles with aft-mounted inlets (e.g., Fig. la) at angles of attack approaching 30 deg. A
large wind tunnel (Fig. lb) could provide the right Mach number, but the required

5
A E 0 C -T R -80 -3 5

temperature and pressure conditions are beyond the capabilities of existing facilities. A
smaller free-jet facility, capable of furnishing the desired temperature and pressure
conditions, could be considered if extraneous effects such as the intersection of the bow
shock with the constant pressure boundary (Fig. Ic) can be avoided or at least controlled to
an acceptable degree. One scheme that has been proposed (Ref. 1) to effectively increase the
length ~f the interference-free test rhombus is the so-called "jet stretcher" (Fig. Id). A jet
stretcher, which in reality is a "test rhombus stretcher," is a duct with an internal contour
that corresponds to the portion of a streamline surface downstream of the bow shock
generated by the vehicle being tested. The stretcher is positioned so that the flow within it is
isolated from the bow shock interaction with the free boundary. Although the lip of the jet
stretcher introduces a disturbance, careful design, fabrication, and alignment should
produce flow conditions approaching the inlet that very closely approximate free flight.

Some theoretical and experimental investigations of axisymmetric jet stretchers, designed

for use with axisymmetric test bodies, have been conducted (Refs. 2 through 5) to verify
feasibility of the concept and to identify operational problems and performance constraints.
Since the internal surface of the jet stretcher corresponds to a stream surface in the
corresponding free-flight flow field about the test body, there is only one jet stretcher with a
given inlet area that corresponds to a given set of test conditions (Moo, Re:', and attitude)
for that test body. Some assessment of jet stretcher off-design performance, in terms of free-
stream Mach number, axial location, and angular orientation with respect to the test body
(Refs. 3 and 4) and porosity of the jet stretcher surface (Refs. 4 and 5), have been
investigated as possible means of reducing cost and increasing flexibility of jet stretcher
installations. Test body surface pressures obtained with free-stream Mach numbers that
differed by 0.25 to 0.35 from design conditions were found to be acceptable. However,
results obtained with angular mismatch and limited porosity in the jet stretcher are
inconclusive based upon investigations to date. Furthermore, jet stretcher design and
application has been limited to maximum angle-of-attack testing of about 5 to 10 deg where
significant boundary-layer separation begins to occur (Ref. 6) on slender bodies. The
relatively limited performance capability of a costly fixed-geometry jet stretcher, together
with limitations of available analytical techniques required for design purposes, has
significantly delayed development of the concept.

Continued concern about the interpretation of data obtained with wind tunnel models -
particularly with transonic test conditions - has aroused interest in "self-correcting" or

"The internal jet stretcher contour is modified from a stream surface to account for boundary-
layer development.

6
 A E DC-TR -80-35

"adaptive wall" wind tunnels (Refs. 7 and 8). The basic problem, of course, is that the walls
of a wind tunnel impose a boundary condition on the flow about a test vehicle (Fig. 2) that is
not present in free flight. For example, a fixed, solid wall tunnel compresses the streamlines
about a model (Fig. 2b) and artificially accelerates the local flow relative to conditions
obtained in free flight (Fig. 2a) at the same initial Mach number condition. However, proper
modi fication of the tunnel boundary condition by recontouring a solid wall tunnel (Fig. 2c)
or by setting appropriate suction flows in a ventilated tunnel (Fig. 2d) will lead to streamline
patterns and flow conditions that are comparable to free flight. The problem is the
determination of the proper tunnel modification required, which is the objective of the
adapative wall test philosophy.

To understand how the adaptive wall concept might be applied to an adaptable jet
stretcher, consider the flow situation created by an aerodynamic configuration in supersonic
flight in a real fluid of unlimited extent (Fig. 3a). To simulate the flow field of Fig. 3 by a
computational technique is difficult with the present state-of-the-art because of the complex
geometry, strong viscous effects near the body, and shock-boundary layer interactions.
Note, however, that the severe restrictions on the computational methods have to do with
the region near the body. Computational methods can adequately simulate the conditions
far from the body.

In contrast, if a test cell is considered as an analog computer, a good simulation of the

flow near the body, which accurately accounts for the geometry and viscous interactions,
can be provided. Unfortunately, because of the finite extent of the test cell it is difficult to
simulate accurately the far field conditions (unless the test article is infinitesimally small).

Consequently, the ideal device to simulate the flow field of Fig. 3a would be a hybrid
device using an analog simulator (the test cell) for the near field and a digital simulator
(computational fluid dynamics) for the far field. The merging of the analog and digital
devices can best be described by examining Fig. 3b where the infinite fluid region is divided
into two parts, exterior (E) and interior (I) to an imaginary surface, S.

First, consider the flow in the exterior region. In the exterior region there are no
immersed bodies; hence, viscous effects are essentially insignificant. Consequently, inviscid
theories (such as the method of characteristics or, for small perturbations, linearized small­
disturbance theory) which are well within the realm of practical computational methods can
be applied in the exterior region. It is clear, then, at least in the inviscid approximation, the
flow throughout the exterior region could be determined by prescribing t he exist ing
streamline slopes at S, for that would amount to prescribing the shape of the stream surface
there. All other flow variables throughout the external region (and at S) are determined by
these prescribed boundary values and the conditions at infinity. Hence, given the streamline

7
A E DC-TR -80-35

slopes at 5, other variables such as the static pressure are uniquely determined by the strong
conditions at in finity.

From the fact that the flow in the exterior region is determined by boundary values of a
single variable at the surface, 5, together with the conditions at in finity, it follows that two
variables at 5 are adequate to define the conditions for unconfined flow. In other words,
specification of two flow quantities all over 5 overdetermines the flow problem in the
exterior region unless they have the required functional relationship with each other that
satisfies unconfined flow at infinity.

Turning now to the interior region, note that the flow is determined by the stream
parameters, the aerodynamic vehicle, and the values of the flow variables at 5, without any
approximations. Furthermore, if the interior region is replaced with a test cell with an
adaptable jet stretcher (Fig. 3c), then to simulate the conditions of flight in that test cell, it is
necessary and sufficient that conditions at 5 be the same as in unconfined flow (where 5 is
any convenient surface within the cell). In general, these conditions are not met for an
arbitrary jet stretcher configuration. If two flow quantities are measured at 5, however, it
could be ascertained by consideration of the boundary value problem in the region exterior
to 5 if the measured flow quantities are consistent with the unconfined flow conditions at
infinity. If they are not, then the jet stretcher geometry would have to be altered in some
fashion to achieve unconfined flow conditions at the test article.

A basic iterative scheme of measurement and calculation for modifying the jet stretcher
to achieve unconfined flow is shown in Fig. 4. For concreteness, the pressure distribution
(PI) and the flow angle relative to 5 (81) are selected as the two flow variables measured at 5.
First, a flow field is established in the cell, and the flow variables, PI and 8J, are measured at
the given control surface, 5. The exterior unconfined region is then evaluated by specifying
8 E = 8) (subscript E designating the exterior region) as the boundary value at 5. If the
distribution at 5 of P E determined from the exterior region calculations does not agree with
PJ, the flow is still constrained by the jet stretcher and the jet stretcher must be readjusted.
The iteration continues until PI and P E agree. The relaxation factor, k, is introduced to
accelerate convergence of the iterative process.

The objective of the present work was to determine (1) requirements for and feasibility
adaptable jet stretchers at AEDC, (2) compatibility with other free-jet hardware, and (3)
additional work required for development of the concept. The method of approach included
an assessment of currently envisioned free-jet test requirements at AEDC and an evaluation
of alternate test techniques. A rigorous mathematical proof that the adaptive wall concept
must converge to give free-flight conditions with supersonic flow was developed to

8
 AEDC-TR-80-35

complement the proof previously developed (Ref. 9) for subsonic flow. Computations of
inviscid perfect gas flow field conditions about simple axisymmetric shapes were used to
determine differences between free-flight, free-jet, and ducted flow conditions.
Computations were made for a test body and a mismatched jet stretcher to demonstrate
convergence to supersonic free-flight conditions and to determine relative flow conditions at
intermediate iterations. Implementation considerations and additional work required to
ensure that the adaptable jet stretcher is, in fact, a practical test concept are also discussed.

2.2 FREE-JET TEST REQUIREMENTS

Current free-jet test activities at AEDC are limited to the ETF Aerodynamic and
Propulsion Test Unit (APTU) where supersonic free-stream Mach number conditions in the
range from 2 to 5 are achieved with fixed Mach number axisymmetric nozzles. Full-scale
vehicles with cross-sectional areas (i.e., blockages) ranging from 5 to 50 percent of the nozzle
exit area have been considered for test at preselected angles of attack up to 30 deg. In some
proposed tests, the vehicle inlet is as much as 12 body diameters aft of the forebody nose.
The APTU supersonic free-jet test approach assumes that only flow conditions in the
theoretical nozzle test rhombus are acceptably representative of free-flight conditions. The
forward portion of the vehicle to be tested, from the nose to the inlet cowl, must therefore be
positioned within the test rhombus, and the APTU exhaust diffuser must be designed and
operated to ensure that the free-jet nozzle is flowing full at all conditions of interest.
Maximum angles of attack that can be achieved depend upon the free-jet nozzle design
Mach number and size and the relative size and configuration of the test vehicle. Free-jet
tests accomplished to date in APTU have been limited to maximum angles of attack of
about 10 deg. No tests have been accomplished in APTU with jet stretchers.

Subsonic and supersonic free-jet test capabilities to be incorporated in the

Aeropropulsion Systems Test Facility (ASTF) are currently under consideration. Presently
perceived design goals (Table I) include both steady-state and transient capabilities with
both subsonic and supersonic flow and pitch and yaw variations. Tests of full-scale gas
turbine engines, inlets, and any airframe surfaces that might affect flow approaching the
inlets will require a large test rhombus since blockages, even at zero attitude conditions, will
be in the 30- to 50-percent range. At maximum attitude angles, effective blockages will be
even greater, and the flow process will be further complicated by impingement of the nozzle
free-jet boundary on the test vehicle. Because of the limited extent of the free-jet nozzle test
rhombus, it may not be possible to include all of the forward airframe components that
might affect flow into the engine inlet. Under these conditions, an approximate shortened
forebody will be used in conjunction with flow-field measurements in the vicinity of the
initial inlet ramp. The forebody simulator and test conditions provided in ASTF will be

9
AEDC-TR-BO-35

selected to produce flow conditions at some measurement station that are comparable to
those obtained from preceding wind tunnel inlet model tests. Achieving the ASTF now
quality goals with all of the desired operational capability will be a formidable task.

2.3 FLOW-FIELD COMPUTATIONS

2.3.1 Approach

Computations of subsonic and supersonic flow fields around bodies representative of

missile and aircraft forebodies were made to determine differences between free-flight, free­
jet, and ducted flow conditions. Analysis was limited to axisymmetric bodies (Fig. 5) at zero
angle of attack without inlet through-flow representation. Six different analytical models
were employed since no single existing computer code has the demonstrated capability to
handle this wide range of boundary conditions. Free-flight or infinite stream conditions
(Fig. 6a) were obtained in the subsonic and transonic regimes with the South-Jameson finite­
difference, relaxation solution of the full potential equation for axisymmetric inviscid flow
(S-J, Ref. 10) while corresponding supersonic conditions were obtained with the Lockheed
(LMOC, Ref. II) and Armstrong three-dimensional (A3DMOC, Ref. 12) rotational
method-of-characteristics codes. LMOC was used for the supersonic free jet (Fig. 6b)
calculations and both LMOC and A3DMOC were used for the supersonic jet stretcher (Fig.
6c) evaluations. The time-dependent, finite-difference Cline nozzle and plume code (CNAP,
Ref. 13) was the only analysis available for the subsonic free-jet (Fig. 6d) evaluations. Both
CNAP and the Wehofer-Moger time-dependent, finite-difference analysis (W-M, Ref. 14)
were used in the ducted subsonic flow (Fig. 6e) evaluations. The Douglas-Neumann panel
method potential flow code (D-N, Ref. 15) was used to obtain subsonic infinite stream
results (Fig. 60 for comparison with CNAP and W-M results.

CNAP, because of its unique capability with the boundary conditions, had to be
employed in the subsonic free-jet computations. However, considerable effort was necessary
to achieve reasonably stable solutions representative of steady-state conditions. Various
combinations of time step size and artifical viscosity factors were attempted with no
apparent success within the 500 to 1,000 time steps (5 to 10 min CPU* time with 1,600 grid
points) that were generally adequate for typical CNAP nozzle and plume computations. As
a last resort, the nose of the forebody was modified on the premise that the differencing
scheme used in CNAP has accuracy limitations which, in combination with the slope
discontinuity at the forebody nose, might be aggravating the computational instabilities.
The nose fairing was gradually increased in extent until reasonably stable flow conditions

"Computer times indicated are for the central processor unit (CPU) of an 181"1 370/165 machine.

10
 AEDC-TR-80-35

were achieved. This required a faired sting (Fig. 5) with cross-sectional area equal to 10
percent of the body. With this geometry and 8,000 time steps (95 min CPU time) stable, free-
jet solutions were finally achieved" with a free-stream Mach number of 0.6. CNAP and
W-M (55 min CPU time for 600 iterations with 21 by 111 grids) were used to evaluate the
modified forebody in a cylindrical duct under subsonic conditions, and equivalent infinite
stream results were obtained with D-N** (13 sec CPU time for 228 nodal points).

Unique capabilities and limitations of the A3DMOC and LMOC codes were weighed in
the selection of the appropriate MOC analysis to use in a particular case. A3DMOC, because
of its streamline tracking capability, was used to define supersonic jet stretcher geometries.
The entropy "smearing" problem that occurs to some degree in all rotational MOC codes
was noted in comparisons of near field flow properties (local total pressure and Mach
number) obtained with A3DMOC and LMOC (CPU times = 11.3 and 1.3 min,
respectively). Because of the reordering process necessary for the more general three-
dimensional computations, entropy smearing is more severe in A3DMOC than in LMOC.
For this reason and because of its general applicability to infinite stream, free-jet, and
ducted boundary conditions, LMOC was used for the majority of the supersonic
computations. AEDC modifications to the basic LMOC code provided plotting and
interpolating capabilities and arbitrary pressure boundary input options that were both
useful and necessary for manual iterations between internal and external adaptive jet
stretcher conditions. Because of LMOC limitations, only body shapes and free-stream Mach
numbers corresponding to attached bow shock conditions were considered in this study.

-. All of the computations made during this investigation were preliminary in nature since
they are restricted to axisymmetric geometries and inviscid conditions. Flow fields associated
with bodies and jet stretchers at angle of attack will be three-dimensional in nature with
significant viscous effects near the surfaces which will, in extreme cases, lead to regions of
separated flow. Furthermore, an inlet located at some point along the forebody, (Fig. Ia)
will obviously alter the flow fields, as will the relative inlet capture/spillage conditions.
However, theoretical investigations of the full viscous, compressible, three-dimensional
problem this represents cannot be achieved until the appropriate algorithms are developed
for this complex combination of boundary conditions.

*Computational instabilities increased as free-stream Mach number approached unity, even with
the modified forebody.
**D-N, which includes a tangent gas approximation for compressibility effects, had to be used
instead of the more precise S-l (3.3 min CPU time tor a 97 by 97 grid) because S-l could not readily
handle the modified forebody geometry. However, flow field computations made with both S-l and
D-N for an unmodified 3-cal tangenr/ogive and with D-N for the modified forebody with a Mach
number of 0.6 yield substantially identical results within about four body radii from the unmodified
nose station.

II
AEDC-TR-80-35

2.3.2 Infinite Stream Results

Forebody flow-field conditions experienced in an infinite stream were of interest for

comparative purposes and for assessment of the potential for short forebody simulators.
The ASTF flow quality goals (Table I) were used as a criterion for determining adequacy of
the test concept to represent free-flight conditions.

Body surface pressure coefficients and local free-stream Mach number and flow angle
variations near the body surface where inlets would probably be located were determined for
various caliber circular arc tangent ogives with subsonic, transonic, and supersonic flow
conditions. The subsonic results (Figs. 7 through 9) indicate significant flow angle variations
in the portion of the field near the body shoulder and for a distance corresponding to I to 2
body radii downstream. However, Mach number profiles indicate relatively uniform
conditions even one body radius downstream of the tangent point on the one-caliber body
(Fig. 7b). Therefore, for vehicles with forebodies like the circular arc tangent ogive with
inlets located more than one body radius downstream of the shoulder, Mach number and
flow angle variations should be uniform to well within the ASTF design goals of ± 0.05 on
Mach number and ± I deg on flow angle if boundary-layer perturbations and/or separation
are insignificant. Replacement of a longer forebody with a shorter version, therefore,
appears to be a viable subsonic test approach at zero angle of attack if the test installation
does not introduce significant extraneous effects such as shock/boundary interactions at
supercritical flow conditions.

Infinite stream supersonic flow results (Figs. 10 and 11) indicate significant variations in
flow angle, particularly near the body shoulder, and some Mach number variations that
result from entropy gradients across the axisymmetric bow shock wave. However,
downstream of the shoulder, particularly with free-stream Mach numbers less than 3, flow
angle and Mach number variations are uniform to within the ± 0.6 deg and ± 0.05 Mach
number ASTF design goals. Therefore, use of a shortened forebody also appears to be a
viable option at supersonic test conditions.

2.3.3 Subsonic Free-Jet Results

The major concern in all wind tunnel and free-jet tests is that the test section or free-jet
boundary will impress or reflect extraneous flow conditions on the test body that are not
representative of free flight. This was investigated with a limited series of computations
involving the modified 3-cal tangent ogive in free jets corresponding to 10- and 50-percent
blockage conditions. Because of the computational stability problems with CNAP, only one
subcritical free-stream condition corresponding to 0.6 Mach number could be considered
during the time period of this study.

12
 AEDC-TR-80-35

The constant pressure free-jet boundaries corresponding to both the 10- and 50 percent
blockage configurations were found to very closely approximate the infinite stream 10- and
50-percent streamlines for the case considered (Fig. 12). The constant pressure free-jet
boundary alters body pressure distributions upstream of the shoulder region to some degree
(Fig. 13a) and apparently causes a more rapid return to free-stream conditions downstream
of the shoulder than occurs under infinite stream conditions. Local free-stream Mach
numbers and flow angles (Fig. 13b) are within ± 0.01 and ± 0.5 deg, respectively, of infinite
stream conditions even on the contoured portion of the modified forebody where the most
significant variations occur in body surface pressure coefficients. Although these results
indicate surprisingly good agreement between free-jet and infinite stream conditions even
with substantial blockages, care must be exercised to avoid sweeping conclusions about
relative subsonic free-jet flow conditions based upon this single set of idealized (modified
forebody with inviscid, subcritical, zero angle-of-attack conditions) results without
additional theoretical or experimental con firmation.

2.3.4 Subsonic Jet Stretcher Consideration

One of the planned work items of this study was to accomplish a series of iterative
computations to demonstrate that flow conditions about a body do, in fact, relax to free-
stream conditions with an adaptable subsonic jet stretcher. Cylindrical extensions (Fig. 12)
of the free-jet nozzle were selected as an initial geometry for the jet stretcher, and
computations were initiated with CNAP. Although the CNAP computational instability
problems precluded completion, the available results do offer some information for
consideration. For example, choking considerations limit subsonic Mach numbers upstream
of a 50-percent blockage model in a cylindrical duct to something less than 0.3. With this
constraint it might be concluded that representative flow conditions for a free-flight Mach
number of 0.6 could not be achieved. This would indeed be true if comparable high quality
(i.e., wind tunnel) flow conditions were required over the entire body. However, in the
proposed ASTF free-jet test philosophy, representative flow conditions will only be required
in the vicinity of the induction system inlet. With this concession, it should be possible to
select test conditions (i.e., overall pressure ratio and/or jet stretcher geometry) to produce
desired nominal levels and distributions of Mach number and flow angle at the induction
system inlet even though there are significant mismatches in other regions of the flow field.
This is illustrated most dramatically by the CNAP results for a modified 3-cal tangent ogive
in a cylindrical duct.

Because of the choking problem and the CNAP input requirements, the ducted subsonic
computations were made with the duct exit Mach number set at 0.6. Pressure coefficients
(Figs. 14a and b, based on M = 0.6 at X/Ro = 16) indicate the significant overall fore-to-
aft deviations from free flight that can be expected with significant blockages on both the

13
A E DC-TR -80-3 5

model and the outer boundary. However, with matched downstream conditions, the major
differences between free flight and ducted flow are forced to occur near the upstream end of
the test body. Even for the extreme case represented by the 50-percent blockage model in a
cylindrical duct, computed flow conditions at the shoulder (X/R a = 6) agree with free flight
to within 0.03 on Mach number and 0.5 deg on flow angle (Fig. 14c). Four body diameters
(X/R a = 8) downstream of the nose station, all computed results (free flight, ducted, and
free jet) were found to agree within ± 0.01 on local Mach number and ± 0.3 deg on local
flow angle. Since the radial distributions are similar in shape, it appears that the agreement
between ducted, free-jet, and free-flight results could be further improved if necessary with
only slight adjustments in boundary conditions. Although these results are probably
representative for any slender body of interest with subcritical flow, additional
investigations are required to determine if they are significantly altered when supercritical
conditions occur.

2.3.5 Supersonic Free-Jet Results

The principal cause for differences between supersonic free-flight and free-jet flow
conditions is the reflection of body-generated compression and rarefaction waves which
significantly alter the constant pressure free boundary shape from the corresponding free-
flight streamline (Fig. 15). The reflected rarefaction zone from the bow shock/free
boundary interaction (Fig. 16a) and subsequent reflections from the body and free boundary
(Fig. 16b) can significantly alter near field conditions (Fig. 17), particularly when high
blockage bodies are evaluated at low supersonic Mach numbers.

Any mismatch between free-jet boundary (i.e., test cell) and free-jet nozzle lip pressure
also introduces extraneous waves that can distort supersonic free-jet test results. An
expansion fan (Fig. 18a) or a compression wave (Fig. 18b) will emanate from the free jet
nozzle lip depending upon the relative magnitudes of the nozzle lip and jet boundary
pressures. These waves will impinge upon and reflect from the test body in a manner that
could-with a significant pressure mismatch-unacceptably distort the flow (Fig. 18c)
entering an aft-mounted inlet.

2.3.6 Supersonic Jet Stretcher Results

A properly designed jet stretcher can minimize the effect of bow shock interactions with
a free-jet boundary. Also, if the jet stretcher lip size and location are properly selected, the
flow field of interest can be isolated (Fig. l d) from extraneous waves produced by
mismatches between the nozzle lip and test cell pressures. However, a fixed-geometry jet
stretcher has obvious limitations as indicated by a comparison of free-flight streamline

14
 AE DC-T R -80-35

trajectories (Fig. 19). Even with a particular body and blockage combination and allowance
for differences in the bow shock location, free-stream Mach number alters the downstream
streamline shape that a jet stretcher must assume. Considering the additional requirements
for angle-of-attack testing with a variety of vehicle geometries, the need for variability,
preferably with rapid, online feedback control, is almost a necessity to ensure practicability
of the jet stretcher concept.

A major concern in the adaptable jet stretcher philosophy is that convergence to free-
flight conditions can indeed be obtained from some arbitrary initial condition in a
reasonable number of iterations. For this reason previous mathematical proofs of the
convergence of transonic adaptive wall wind tunnels to unconfined or infinite stream
conditions (Ref. 9) were extended to supersonic flow and are included in the Appendix .
. Convergence was further demonstrated in a series of computations where the interior flow
field conditions between a typical body and an off-design jet stretcher, which would
normally be obtained in a test installation, were computed with LMOC. The corresponding
exterior region computations, which would normally be made online in an actual adaptive
test mode situation, were also accomplished with LMOC.

A 3-cal tangent/ogive was evaluated in a Mach number 2.3 airstream with an off-design
jet stretcher configuration corresponding to an axially shifted, 50-percent blockage Mach
number 5.0 streamline. Body surface pressure coefficients (Fig. 20a) and near field flow
conditions (Figs. 20b and c) obtained with the shifted, mismatched jet stretcher indicate
significant deviations from infinite stream conditions, largely because of the 1.9-deg
mismatch (Fig. 2Od) at the jet stretcher lip.

From the initial mismatched flow condition, the jet stretcher contour was varied through a
series of exterior linterior region iterations until near field Mach numbers and flow angles
agreed with infinite stream results to within ± 0.05 on Mach number and ± 0.6 deg on flow
angle. Streamwise distributions of static pressure near the jet stretcher surface were used to
define the outer boundary conditions for the interior region computations. The resulting
outer boundary streamline contours, computed from the interior solutions, were then used
as a boundary condition in the exterior region computations. Typical LMOC CPU times
were 20 and 50 sec, respectively, for the interior and exterior computations. LMOC had to
be modified to accept streamwise variations of pressure as a function of axial position and to
accomplish spline fits which could be used in the interior computations.

15
A E DC-TR -80-35

The sequence of events in the iterations was as follows:

1. Interior flow-field conditions were obtained with the off-design jet stretcher
contour as the outer boundary to establish initial pressure distributions [P?(X)]
along the jet stretcher surface with Moo = 2.3.

2. An initial exterior region computation was made with a hypothetical

axisymmetric body, having the off-design jet stretcher contour, in an infinite
Moo = 2.3 stream to define surface pressure distributions [P~x)].

3. The initial exterior and interior pressure distributions were compared and a first
iteration internal pressure distribution [pI (X)] was established from

where k, the weighting factor, was arbitrarily set at 0.5.

4. A first iteration internal solution was accomplished with a varying pressure outer
boundary represented by pI(X) to determine the first iteration jet boundary
shape.

5. A first iteration exterior solution was accomplished with a hypothetical

axisymmetric shape corresponding to the first iteration jet boundary contour to
determine the corresponding surface pressure distribution.

6. pI<X) and Pl.,<X) were used to establish a second iteration pressure distribution
from

Pf (\) = k [rl (X~ + (1 - k) f:~ (X~

7. A second iteration internal solution was accomplished with P/<X) as the outer
boundary.

Changes in body surface pressure coefficients and near field Mach number, static
pressure, and flow angle for the two iterations are shown in Fig. 20. Although all parameters
indicate a general convergence to infinite stream conditions, there is still evidence (Figs. 20d
and 21) of some residual mismatched conditions near the jet stretcher lip that are being
reflected downstream after the second iteration. This mismatch is attributed in part to the
upstream boundary conditions used in the interior computations and could probably be
eliminated with a more careful selection of LMOC starting line point conditions. Secondary
ripples in the body surface pressure coefficients, particularly with the initial off-design jet

16
 AEDC-T R -80-35

stretcher geometry, are probably indicative of irregularities in the jet stretcher contour that
are produced by the combination of the number of points selected and the resulting spline
fits.

2.4 ASSESSMENT OF JET STRETCHER FEASIBILITY AND

IMPLEMENTATION PROBLEMS

2.4.1 General Remarks

Although actual evaluation of the adaptable jet stretcher concept was very limited and
confined to the relatively simple case represented by an axisymmetric body at zero angle-of­
attack in axisymmetric supersonic and subcritical, subsonic inviscid jets, the results obtained
do indicate feasibility of the approach. However, there is obviously a large step from the
limited feasibility demonstrated in this study to the provision of a practical system that can
be used in either APTU or ASTF. Further investigations, both theoretical and experimental,
are obviously required. Three-dimensional subsonic and supersonic computational
capability for general bodies embedded in planar or axisymmetric flow fields is required to
extend the feasibility studies to relevant ASTF and APTU test considerations. Some
experimental model studies will also be .equired to evaluate control systems,
instrumentation concepts, and external region computational algorithms. When an
.. appropriate computer code becomes available, relative effects of flow into the engine inlet
on jet stretcher aspects should also be considered, particularly with subcritical inlet
operation.

High angle-of-attack tests of large bodies that will penetrate a subsonic free-jet
boundary will create unique flow problems that may be difficult to correct with a jet
stretcher. After the planned FY 81/82 ASTF subsonic free-jet model tests in Propulsion
Research Cell (R-2A2) are accomplished, a better assessment can be made about the
potential role of adaptable jet stretchers in subsonic free-jet testing.

High angle-of-attack testing of slender vehicles with aft-mounted engine inlets could also
be difficult to handle with jet stretchers. Extension to angles of attack up to about 20 deg,
where separation occurs on the vehicle in a fairly simple, time-wise steady manner (Ref. 6),
may be possible. However, the asymmetric shedding of multiple vortex sheets observed (Ref.
6) with slender vehicles at higher angles of attack will pose even more formidable
implementation problems.

Incorporation of adaptable jet stretchers in APTU will require a high-response control

system and a fast external flow algorithm that are compatible with the limited blowdown

17
A E DC-TR -80-35

times. Incorporation into ASTF will depend to a large degree on compatibility with the
planar free-jet nozzle and associated free-jet spillage diffuser systems actually used.

2.4.2 Jet Stretcher Control

In actual tests, the two independent flow parameters required for the jet stretcher
adjustments must be measured along some reference surface (or surfaces) in the vicinity of
the jet stretcher surface. Ideal measurable flow parameters for this purpose would be (I)
simple to obtain, (2) noninterfering, and (3) highly responsive to jet stretcher adjustments.
In current subsonic/transonic adaptive wind tunnel wall studies, flow-field measurements
made depend upon the manner in which the effective wall shape is achieved. If a flexible wall
is employed (Ref. 8), wall static pressure distribution and contour geometry are the control
parameters. In a perforated or slotted tunnel (Ref. 7), where effective outer boundary shape
is changed by air injection or removal, static pressures and flow angles are usually measured
with static pressure pipes and hemisphere/cylinder probes at some reference surface near the
wall but outside the boundary-layer interaction region. Static pressure measurements from
two different reference surfaces are also being considered as an alternative. Comparable
measurements could be made in a subsonic jet stretcher.

Actually recontouring the wall of a jet stretcher required for three-dimensional

supersonic flow with temperatures greater than 200 to 300°F will be extremely difficult. If
effective shaping is accomplished with air injection or removal, control parameter
measurements are more of a problem. Optical techniques, such as laser velocimeters, would
pose formidable installation and operational problems. On the other hand, insertion of
probes into the supersonic airstream will introduce extraneous compression and expansion
waves. Calibrated cone probes or double wedge airfoil sections (Fig. 22) extending beyond
the interaction zone might be used for supersonic adjustable jet stretcher control parameter
measurement if the extraneous waves can be tolerated. In any case, the probes must be
carefully designed to be as small as possible and to extend no farther beyond the boundary-
layerlinbleed interaction zone than necessary. Careful consideration must also be given to
the number and placement of the probes to ensure that measurements are obtained at the
most critical regions with a total number of parameters that can be accommodated by the
available data acquisition and conditioning equipment.

2.4.3 Online Exterior Flow Computations

The present studies indicate that the method of characteristics is a reasonable algorithm
to use, at least for axisymmetric flow, in the online exterior flow computations for
supersonic flow. A new computer code must be formulated to take the specified measured

18
 AEDC-TR-SO-35

quantities, make the appropriate characteristics computations, assess the adequacy of

current conditions, and then provide an output that can be converted into an appropriate
signal to inbleed/outbleed control valves. The relatively high computational speeds achieved
with a general program like LMOC suggest that CPU time should not be a problem for a
tailored MOC code. However, the code should be constructed with time optimization in
mind since this may be crucial to satisfactory operation with a blowdown facility like APTU.

The subsonic/transonic free-flight results presented herein suggest that elements of the
potential flow and finite-difference algorithms incorporated in D-N and S-J could be
employed for external region computations with subsonic adaptable jet stretchers. The
potential flow option is particularly attractive because of its flexibility and speed; however,
it is limited to subsonic conditions on the reference surface. Algorithms being developed for
adaptive transonic wind tunnel applications should also be considered for extension to
subsonic jet stretchers. For example, the Prandtl-Glauert form of the linear small
disturbance equations and a finite-difference formulation of the transonic small disturbance
equation have been used (Ref. 16), respectively, with subsonic and transonic flow conditions
at the reference surface. Further investigations are required to determine advantages and
disadvantages of extending these and other adaptive wind tunnel external flow analyses to
subsonic/transonic jet stretchers.

3.0 CONCLllDING REMARKS

-.. Approaches that will permit relatively large airbreathing engine/inlet/forebody systems
to be free-jet tested in APTU and ASTF received preliminary investigation. Conclusions
were based upon near field flow properties (Mach number and flow angle) obtained from
in viscid flow computations for slender, axisymmetric bodies at zero incidence using ASTF
flow quality design goals as an acceptance criterion.

Results obtained with subsonic and supersonic free flight boundary conditions indicate
that shortened forebodies can be used to reduce the overall length of test installations
required for vehicles equipped with aft-mounted inlets.

Limited axisymmetric calculations for slender bodies immersed in free jets at subcritical
fiow conditions indicate streamlines and local steady-state flow properties that are
comparable to free-flight conditions to well within the ASTF flow quality goals. However,
this agreement is favorably biased in the calculations by an artificial upstream extension to
the forebody which was necessary to reduce computational instabilities. Although the
instability problems precluded quantitative verification that an adaptable jet stretcher could
reduce differences between subsonic free-jet and free-flight results, the ongoing transonic

19
AEDC-TR-80-35

wind tunnel adaptive wall studies and the limited results obtained on this study do tend to
support this premise.

Bodies immersed in a supersonic free jet experience unacceptable flow distortions as a

result of (1) bow shock reflections from the constant pressure free-jet boundary and (2)
waves emanating from the nozzle lip because of exit plume static pressure mismatch. The
adaptable jet stretcher can potentially eliminate these disturbances over a range of test
conditions. A rigorous mathematical proof of jet stretcher convergence to the desired
interference-free geometry was developed for supersonic flow to complement the previously
developed proof for subsonic conditions. Convergence was confirmed by a computer
experiment for a slender axisymmetric body in an off-design supersonic jet stretcher. Near
field flow disturbances were reduced to an acceptable level after two readjustments of the jet
stretcher geometry.

The method of characteristics appears to be an acceptable algorithm for external region

calculations with supersonic adaptable jet stretchers. Adaptation of linear small disturbance
and finite-difference formulations under development for adaptive transonic wind tunnels
should be possible for external flow computations in subsonic jet stretchers.

~o
 A E DC-TR-80-35

REFERENCES

.- 1. Himrnler, E. B., et al. "A Special Technique for Free-Jet Testing of Aft-Inlet
Systems." The Marquardt Corporation Report MP-1429, November 1966.

2. German, R. C. "Simulation of Supersonic Flow Over a Body of Revolution Using an

Axisymmetric Jet Stretcher." AEDC-TR-70-166 (AD875834), October 1970.

3. Barebo, R. L. and Matkins, E. H. "Simulation of Supersonic Flow Over a Scale Model

Missile with Aft-Mounted Inlets Using an Axisymmetric Jet Stretcher." AEDC­
TR-71-37 (AD51489l), April 1971.

4. Eppright , B. R. "Theoretical Effects of Porosity and Angle-of-Attack on Jet Stretcher

Performance." University of Texas, PhD Dissertation, May 1971.

5. Bauer, R. C. et al. "A Theoretical and Experimental Study of a Jet Stretcher System."
Journal of Spacecraft and Rockets, Vol. 10, No.6, June 1973, pp. 395-405.

6. Przirembel, C. E. A. and Shereda, D. E. "Aerodynamics of Slender Bodies at High

Angles of Attack." Journal of Spacecraft and Rockets, Vol. 16, No. I, January­
February 1979, pp. 10-14.

7. Sears, W. R. "Self Correcting Wind Tunnels." Aeronautical Journal, Vol. 78, No.
758-759, February-March 1974, pp. 80-89.

8. Goodyer, M. J. "The Self-Streamlining Wind Tunnel." NASA Langley Research

Center Report NASA TMX-72699, August 1975.

9. Lo, C. F. and Kraft, E. M. "Convergence of the Adaptive Wall Wind Tunnel."

American Institute of Aeronautics and Astronautics Journal, No. I, January 1978, pp.
67-72.

10. South, J. c., Jr. and Jameson, A. "Relaxation Solutions for Inviscid Axisymmetric
Transonic Flow over Blunt or Pointed Bodies." Proceedings of the First AIAA
Computational Fluid Dynamics Conference, Palm Springs, California, July 19-20,
1973.

II. Smith, S. D. and Ratliff, A. W. "Rocket Exhaust Plume Computer Program

Improvement, Vol. IV." Lockheed Missiles and Space Company Report LMSC/HREC
DI62220-1V-A, June 1971.

12. Armstrong, W. C. and Bauer, R. C. "Analysis of Three-Dimensional Inviscid

Supersonic Flow Between a Body and an Outer wall (with Application to a Jet Stretcher
System)." AEDC-TR-76-103 (AD-A029123), August 1976.

21
AEDC-TR-80-35

13. Cline, M. C. "NAP: A Computer Program for the Computation of Two-Dimensional,

Time-Dependent, lnviscid Nozzle Flow." Los Alamos Scientific Laboratory Report
LA-5984, January 1977.

14. Wehofer, S. and Moger, W. C. "Analysis and Computer Program for Evaluation of
Air breathing Propulsion Exhaust Nozzle Performance." AEDC-TR-73-29(AD760541),
May 1973.

15. Hess, J. S. and Smith, A. M. O. "Calculation of Potential Flow About Arbitrary

Bodies." Progress in Aeronautical Sciences, Vol. 8, Pergamon Press, New York, 1967.

16. Kraft, E. M. and Parker, R. L., Jr. "Experiments for the Reduction of Wind Tunnel
Wall Interference by Adaptive-Wall Technology." AEDC-TR-79-51 (AD-A076555),
October 1979.

22
 ,
. # , . . .

Bow Shock
Moo 2.35
~Streamlines

_----'L-

-----
---
------ -- - ----

N
W DB I I

5.6 DB. ·1

a. Free flight
Figure 1. Flow-field characteristics about a supersonic vehicle ~
m
with different boundary conditions. o
C')

~
JJ
Co
o
W
\11
A E DC-TR -80-35
Qi c
i;:,
c
c .<:;
.-<
... oc
;:,
0
.-< "'C
~
C
~
~ ....
.-<
i1J
C
c
::l
E-o

.0 -
Q)
;:,
.E!
LL
8
::;;:
24
 .I ) , .

Constant Pressure Boundary

Free-Jet Nozzle

M00 = 2.35

N
VI

Waves Compression Waves

c. Free jet
Figure 1. Continued.

»
m
o
o
.:...
Jl
00
o
W
<J1
 :t>
m
o
o
~
JJ
00
o
W
(Jl

Free-Jet Nozzle Jet Stretcher

M00

I·..)
DB I I
0'­

Jet Boundary

d. Free jet with jet stretcher

Figure 1. Concluded.

~
 AE DC-TR -80-35

___ LstreamJines (Typ)

Mco ---~-----==-
--=::~ =-=-
--...... ::===
_ ..........
':: : - ;:.- ~::::-=--=--~
------' --..._-­
- - ---

a. Free flight

LLIIIIIII~IIII'II~Streamlines Distorted
11 co z z: .=-:~.::- _-_ _ with Respect to Free­
- - --:. ':. - - - Fl ight Streaml ines

b. Fixed, solid wall tunnel

111111~11l..LL IIIII . .
___ --;:--~,-_ .:::::,.----Streamllnes Lde n t t c a I
::::::::::~=:::=-_ to Free-Flight Streamlines
- - - - ~ _ __ wi th Proper Shaping

c. Adjustable, solid wall tunnel

Distributed Suction

Mco
JJ
___
- -= -::::.
_-;::"=
ZI2 'U
== _ ,
_-,_- _ZZ '
- __ ~ ~ .
St r-e am l i n e s Identical to
Free-Flight Streamlines
-- *"" ..-- . . . . .
----_~--- with Proper Suction
----:. ~ =--=--=---- Distribution

d. Ventilated wall tunnel

Figure 2. Schematic of near field streamline patterns
for a cylinder in crossflow with various
boundary conditions.

27
AEDC·TR·BO·35

\.I > 1
..

a. Unconfined supersonic flow

Exterior
Region S-Control Surface

Interior
Region

b. Definition of interior and exterior regions

Jet Stretcher

c. Replacement of the interior region with the test cell

Figure 3. Integration of the test cell and computational domain.

28
 I ,I

Control Surface
Test Cell Computer
S

P = f(6
r)

~~-~---D
E

CJ
~f2asuredPressure A
(PI) and Flow
r
Angle A on S
r
~ ~----
3 '\ f\ v.. Stretcher

Adjust Jet
Stretcher
It.,)
1-0 P kP
r Ne w E
+ (l - K) P
l Ol d

Correct External
Aerodynamics
:t>
m
o
()

~
Figure 4. Iterative scheme for the adaptable jet stretcher concept. :D
00
o
W
U'I
 »
m
o
Cl
~
JJ
Co
o
W
R U1

Caliber _ L /2 R
o 0

0.316
-
o
[
-.
-...I
I >(
R

hi L.,...."': I t i " I ... X

Faired Forward Sting For Circular Arc Tangent Ogive:

Modification (For R
C R2 + L 2
Subsonic Jet and o 0
Ducted Computations) RC = 2 R
o
w With 0 < X < L
o o

R
B
= R0 - RC +
-0 (R 2 - L2 ) + 2 L X - X2
Coo

With X > L :
o

H = R (=1 for this study)

B o

Figure 5. Details of circular arc tangent/ogive bodies used

in computations.

\
 A E DC-TR-80-35

Code Used

~L
o
R
__-- Subsonic:
su_p_e_r.,son i c:

I< • X
S-J
t~g~oc

a. Infinite stream (~ free flight)

,R
N
1 ".,.. ~
R Percent Blockage
(l/R N) 2 x 100% LMOC

I -------~ • X
o
b. Supersonic free jet

A3DMOC
L~IOC

Percent Blockage
(1/R x 100%
CNAP
N)2

Subsonic free jet

Percent Blocka~e =
R (l/R
NE)2
x 100<;C
, -'I",(X =
CNAP
'til/II 1111/11 1111 I /I /II I I 1/ 1//111
W-M
R _
NE I=====:: 1
o
e. Subsonic (in duct)

D-N
For Comparison
with Subsonic Free
Jet and Ducted
Cases

1 .X
f)

f. Subsonic infinite stream

Figure 6. Boundary conditions and computer codes used in the
flow-field computations.

31
 »
m
o
o
.:,
JJ
-0.8 00
o
W
Sym Moo (Jl

1.0
-0.6 -0- 0.9
0.8
0. 0.6
U
-0.4
+-'
l::
Q)
.....
.....o
'H
'H -0.2
Q)

W
0
U
N
Q)
...
;:l
[J;
(f)
o
Q) 2 4 6 8 10 12 14 16
...
0. Axial Position, X/R
o

O.H~
0.4
I
L,

a. Body surface pressure coefficient

Figure 7. Flow conditions near a 1-cal tangent!ogive body with
subsonic and transonic free-flight conditions.

"
 AEDC-TR-80-35

~ X/Ro
2 (Body Shoulder)
- .. - •. - 3
.- ---- 4
-._.- 6
- - - 10

1.0

Wl
( \!I
0.8

0.6
\\ \ \I.
\
0.4
III
0.2
II:
'I
-,
,
-, \
\ \1

I
1.0

1.0
III I
,I'
0.8 MI ,

0.6
.:/
0.4
t,,
c 0.2
.....c
+-'
..... o L.----l_-.uu...-..L._......I
rJ) 0.8
o
Po.
(j)
:>
-,-4
+-'
cd
......
~
1.0

0.8
r:
,: I
I
I':/
,
0.6
I

0.4

0.2
I ./ I
, I

o L.......l.......L.L.._...J....__
o 0.4 0.8 1.2 1.6 -2 -1 o 1 2 3
Local Mach Number Flow Angle, deg
M 0.6
b. Local Mach number and flow angle
Figure 7. Concluded.

33
0.4

a. Body surface pressure coefficient

Figure 8. Flow conditions near a 2-cal tangent!ogive body with
subsonic and transonic free-flight conditions.
 A E DC-T R -80-35

XjR
S
~ __0

4 (Body Shoulder)
----- F
_._.- 8

1.0 ,
0.8 ,,
\

\
0.6

0.4
, \
\
\
0.2

o:l 0
(:C
........ M00
1.0
(:C

Qi
U
oj 1.0
'H
s,
::l
tr.I 0.8
>,
"0
0 0.6
o:l
E 0.4
0
s,
'-
c 0.2
....,
.+-'
..., 0
(IJ
0
M00 0.8
P,
Qi

.....:> 1.0
+-'
r:l
.......
Qi
0:: 0.8

0.6

0.4 ~
.,II
~
0.2

0
0.5 0.7 0.9 1.1 1.3 -0.3 0 0.3 0.6 0.9
Local Mach Number Flow An~le. dep,­
M00 = 0.6

b. Local Mach number and flow angle

Figure 8. Concluded.

35
 ~
m
o
()

~
JJ
Co
o
-0.4 W
r..n

u 0. -0 . 2
..,
c
.,..,Q)
.,..,o ()

'+-'
12 14 16
'H
Q) Axial Position, X/R
a o
u
Q) 0.2
H M00
,~

10\ ::I Sym

ffJ
ffJ
Q)
1.00
~
0.. 0.95
1).4
0.90
0.80
0.60
0.6

a. Body surface pressure coefficient

Figure 9. Flow conditions near a 3-cal tangent/ogive body with
subsonic and transonic free-flight conditions.
 AEDC-TR-BO-35

X/R o

n (Body Shoulder)
_._.- R

------ 12

1.0

\
0.8

0.6 "d
I'
\
0.4 IiI'
~ 0.2
II
a
M00 1.0

OJ
C,)
<"d
'H
1.0
,, ,I
~
(/J
0.8
:II,
-2o 0.6
III
:1
EO.4 I'
o
I-<
'H
0.2
II
<: ~
o
...,
''';
a
''';
(/J 11 00 0.8
o
0..

1.0

0.8

0.6

0.4

0.2

a L........J'--.....L_......II..-....L._...l.._...1
a 0.2 0.4 0.6 0.8 1.0 1.2 -0.3 a 0.3 0.6
Local Mach ~umber Flow Angle, deg

b. Local Mach number and flow angle

Figure 9. Concluded.

37
 :l>
m
o
o
~
::u
00
o
W
(Jl

-0.2

0.
U

-+-' o
c 6 8 10 12 14 16
.,.,
Q)

.,.,o Axial Position, X/R

o
'H
'H
Q) 0.2
o
u
w Q) M00
00
'"'
;:l

~
Q)
0.4
I
iII:::= 3.0
5.0
_____ 2.3
'"'
0.

0.6

a. Body surface pressure coefficient

Figure 10. Flow conditions near a 2-cal tangent/ogive body
with supersonic free-flight conditions.
 A E DC-TR-80-35

Sym XjR o
4 (Body Shoulder)
--­ 6
- ---­ 9
_._-- 12

1.0
~\
0.8 "
\; ~ -:
0.6
1/
,
0.4
.-;
0.2

CO 0
0::
<,
II 00 5.0
0::

Q)
o 1.0 ,, /
\~
oj

""'
I-;
;:J
Cfl
0.8
:>.
'0
0
CO
0.6
(\j
\~
E 0.4
0
I-;

""'c 0.2
0
......
...,
...... 0
i
(f; M00 3.0
c
0.
Q;
;:.
......
..., 1.0
oj
......
Q)
0:: 0.8

0.6

0.4

0.2

0
1.0 3.0 5.0 -1 0 1 2 3 4 5 6 7 8 9 10
Local Mach Flow AnJ;le, deg
Number
Mcc 2.3

.. b. Local Mach number and flow angle

Figure 10. Concluded.

39
 »
m
o
o
~
:Il
00
a
W
OJ

0. -0.2
U

+->
l::
Q.I
0,-<
o o
0,-< 6 8 10 12 14 16
'+-<
'+-<
Q.I Axial Position, X/R
o o
U
Q.I 0.2
~ r-.
o ;:l
CJJ
(fJ
Q.I
!-<
0.
0.4

a. Body surface pressure coefficient

Figure 11. Flow conditions near a 3-cal tangent/ogive body
with supersonic free--fIight conditions.
 A E DC-TR -80-35

, .
41
 ~
m
o
o
~
JJ
Co
o
Sym W
(J1

Free-Jet Boundary
4 Free-Flight Streamline Blockage,
Cylindrical Duct Extension percent
o 10
a::
~ 3
c
....o
....+->
CI) 2
o
c, 50
.-;

-+:>
t-J
....
c{j

-0
~ 1

o 2 4 6 8 10 12 14 16
Axial Position, X/R
o

Figure 12. Comparison of free-jet boundaries, free-flight streamlines,

and cylindrical duet extensions considered in the subsonic
computations (3·cal tangent/ogive, M~ = 0.6).

,
 l .

Infinite Stream
-0.4 *10-percent Blockage (In Jet)
*50-percent Blockage (In Jet)
0..
U
*c p Based on M00 0.6 at X 16
...., -0.2 (and Along Jet Boundary)
c
...,
Q)

...,o
"-'
"-' o
Q)
-+:- 0 8 10 12 14 lEi
V-l U
Q) Axial Position, X/R
:... o
;::l
rJJ
rJJ
Q)
0.2
:...
c,

0.4

a. Body surface pressure coefficients

Figure 13. Flow conditions near a 3-cal tangent/ogive body
with M~ = 0.6 free-flight and free-jet operation.
»
m
o
o
~
:0
00
o
W
\1l
A E DC-TR-80-35

Boundary Condition Blockage (percent)

- - - Infinite Stream
- - - - Free Jet 10
-'-'-'- Free Jet 50

1.0

0.8

0.6

0.4

M 0.2
I
cff 0 L.-_ _.....L .....l._ _......
~ ....

<, X/Ro 2
IJ:::

1.0 I

:.. I
::l
(/J
0.8
'0
>, \1
o 0.6
co
E \
....o 0.4
....
c 0.2
.....o
;..J
..... 0
en
o X/Ro 4
c,
Ij)

.....:>
;..J
C':l
1.0
......
Q
IJ::: 0.8 I
0.6 II
0,4
!I
~I
0.2
~I
0
0.3 0.4 0.5 0.6 0.7 0 5 10 15
Local Mach Number Flow Angle, deg
X/Ro 6 (Body Shoulder)

b. Forebody Mach number and flow angle variations in

free-flight and free-jet environments.
Figure 13. Concluded.

44
 '. • J'
.'

-0.2 Infinite Stream

o k:::: ~I I 7/ '<
/4
~
jf~=:: I I ! ! I
- I
8 10 12 14 16
/
Axial Position, X/R
U
0. / o
/ ,

-z' /
0.2
+-'
~ / ~ *50-percent
Q)
..... I -.. . . - Blockage
.....o " / , (In Duct)
""'
""'Q)
0
0.4
*10-pe~cent /
,
+:0­
VI
U
Q)
I-<
Dlockage
(In Duct) ,
/ *c p Based on M 00
0.6 at X 16
::l
rn
rn
Q)
I-<
0.6
, /
P.
/
0.81- .-­ - - / - M _ 0.28

1.0

a. Body surface pressure coefficients

Figure 14. Flow conditions near a 3-cal tangent/ogive body l>
m
with M~ = 0.6 free-flight and ducted operation. o
o
.:,
:0
00
o
W
Ul
 l>
m
D
o
~
:0
00
-0.2 o
W
(11

.,---- ...
0 -'-­ - -'"'£3­
0.
o I 8
...., / Axial Position,
I::
.....ill 0.2
i
,
.....c RUS/R o Boundary Condition
,/
..... -~
.....
ill
-_._­ 3.162 Infinite Stream

,/
0
U --0-­ 3.162 10-percent Blockage Duct
ill 0.4 --- - - 1.414 Infinite Stream
~
0\
s..,
::l
(IJ
/ ~~- 1. 414 50-percent Blockage Duct
/
(IJ
ill
s..,
0. 0.6
C Based on H00 0.6 at X 16
./ p

/
0.8
----~/.

b. Comparison of pressure coefficients on the duct wall

and along a comparable surface in free flight.
Figure 14. Continued.
 A E DC-TR-80-35

Boundary Condition Blockage, percent

Infinite Stream
~ Duct 10
1:2e--oo- Duc t 50
1.0

0.8

0.6

0.4

...... 0.2

0
tIl
~ X/H 2
<; o
co:
III
U 1.0 .­
cO
'­
I-.
:l 0.8 I-
CJJ
>,
"0 0.6 l-
c
c::l !
E 0.4 l-
...e
'H
1
0.2 l­ )
e
0
.....
....,
.....
a I I I

rn X/H 4
0 o
0­
Q)
>
...., 1.0
oj
....
Q) 0.8
~

0.6

0.4

0.2

0
0.3 0.4 n.5 0.6 0.7 0 5 10 15
Local Mach Number Flow Angle, de~

X/R = 6 (Body Shoulder)

o

c. Forebody Mach number and flow angle variations

Figure 14. Concluded.

47
 l>
m
o
o
~
:0
00
o
W
Ul

Bow
5 tr

_1
Shock
2.00
0::
0
.......
0::

I:::
1. 50
---­ : ' :

0
.,...;
...,
.,...;
u: 1. 00
0
0..
.....
~ cd
00 .,...; 0.50
'0
cd
a:
o K ' , I , I , , , I , I I" I

o 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 12.00 13.00 14.00
Axial Position, X/R
o

a. Free flight
Figure 15. Comparison of free-jet and free-flight flow details
(50-percent blockage 3-cal tangent/ogive.)
 / . I

Reflected Waves

2.50
Constant Pressure Boundary Streamline
0
0::
........ 2.00
0::

I::
....0 1. 50
....+-'
rJl
0
P- 1.00
.--i
~ til
\0 .,..;
"0
til 0.50
0::

0
0 2.00 4.00 6.00 B.OO 10.00 12.00 14.00
Axial Position, X/R o

b. Free jet
Figure 15. Concluded.

l>
m
o
o
~
::0
cb
o
W
<Jl
 5.00 J>
m
o
o
~
4.50 :D
Right Running 00
Characteristics o
Lines (Typ) W
Q1
4.00
Bow Shock
Free Boundary
3.50
0
0::
.......
0::
3.00
s::
0
•..1
+-'
'..1 2.50
{fJ
0
P­
H
VI <1l 2.00
0 '..1
-0
<1l
0::
1. 50

1. 00

0.50

0
0 2.00 4.00 6.00 8.00 10.00 12.00 14.00
Axial Position, X/R
o

a. 10·percent blockage body with M~ = 3.0

Figure 16. Flow-field details produced by a 3-cal tangent/ogive
body in a supersonic free jet.

\ ,
 3.00

2.50
0 Free Boundary
0::
........
0::
2.00
c Bow Shock
...,0
....,
..., 1. 50
if)
0
Po.
.--1

...,
oj 1. 00
'0
oj

U'I
er::
0.50

0
0 2.00 4.00 6.00 8.00 10.00 12.00 14.00
Axial Position, X/R
o

b. 50-percent blockage body with M~ = 2.0

Figure 16. Concluded.

l>
m
o
(')

~
Jl
00
o
W
111
A E DC ·TR ·80·35

Infinite Stream
- - - - - Free Jet

1.0

0.8

0.6

0.4

0.2

0
X/R o 6 (Body Shoulder)

II)
tl
....c<l 1.0
r..
;:l
eo 0.8
~
-e
& 0.6
e
o
r..
0.4
....
~ 0.2
..-<
:::: 0 ---l ---l ~

Ul
o
p.
9
II)

.....>
~ 1.0
c<l
..... I /
~
<II
0.8 / ./
./
I
0.6

0.4
/
/
---
--:­
--
"...
/

0.2

3.0 3.5 o 1 2 3 4 5 6
Local Mach Number Flow Angle, deg
X/R o = 12

a. 10-percent blockage, M~ = 3.0

Figure 17. Comparison of flow-field conditions near a 3-cal
tangent/ogive body with supersonic free-jet and
free-flight conditions.

52
 AEDC-TR-80-35

Infinite Stream
----- Free Jet

1.0

0.8
\
0.6 \
'\
0.4
\ -,
<,
~ <,
0.2 <,
...... <,
<,
,-...
0
~ X/Ho 6 (Body Shoulder)
~
<,
a::
III 1.0
o
t<$
"-<
I-. 0.8
:l
CIJ \
:>. 0.6
'0
0
(Il

E
0.4 \
0 \
I-.
"-< 0.2 -,
c \
0
.....
.., 0
..... X/Il o 9
IJJ
0
0..
III
.....> 1.0
+'
t<$
.-<
III
0.8
a::
0.6 I

0.4
/
/
/ " ....... \

- -­
I
0.2 /
I
0
1.5 2.0 2.5 -10 -8 -6 -4 -2 0 2 4 6
Local Mach Number Flow Angle, deg
x/:qo = 12

b. 50-percent blockage, M = 2.0

Figure 17. Concluded.

53
 »
m
o
()
10-percent Blockage 3-cal Tangent/Ogive ~
JJ
Moo = 3.0 (Uniform)
Co
p o
/B = 0.99 W
U1
NE
4.00
Expansion Fan
Nozzle Lip
3.50

3.00~
~
0
a:
<,
a:
.
.::
0
''';

.....,.;
f/J
VI 0
~ Jl,

..-<
oj
.,.;
1. 50

'"=:
oj

1.00

0.50

2.00 4.00 6.00 8.00 10.00 12.00 14.00

Axial Position, X/R
o

a. Underexpanded nozzle flow

Figure 18. Extraneous waves introduced by pressure mismatch
at free-jet nozzle exit.
 ,. ...

10-percent Blockage 3-cal Tangent/Ogive

M~ = 3.0 (Uniform)
P
~ = 1.01
P
NE
4.00

Compression Wave
Nozzle Lip
3.50

o
0::
.......
0::
.
c
o
·M
+->
.~ 2.00
V1 o
p..
V1
.-<
cd
·M
'0
cd
0::

o 2.00 4.00 - -- 8.00 10.00 12.00

Axial Position, X/R o

b. Overexpanded nozzle flow :t>

m
Figure 18. Continued. o
o
.:;
::0
00
o
iN
(J1
A E DC-TR -80-35

Infini t e Stream
_ . - . - Free Jet PJB/P 0.90
NE
_ . - . - Free Jet PJB/Pr-'E 0.99

4;3-0- Free Jet - PJB/PNE ~ 1.01

- - - Free Jet PJB/PNE I. 10

1.0

0.8 / /
// .r
:

0.6 /
0.4 .. / /
0.2 c.
0'--_ _......L...a..._ _.L..­_ _....J
X/R 6 (Body Shoulder)
o

1.0

0.8

0.6

0.4

c 0.2
.....o
;..>
..... o
u: 9
o
0..
Q)
;.
.....
;..> 1.0
~
.....
Q)
ll:: 0.8

0.6

0.4

0.2

o
2.5 3.0 3.5 4.0 -1 0 1 2 3 4 5 6
Local Mach Number Flow Angle, deg
X/R = 12
o

c. Flow-field conditions
Figure 18. Concluded­

56
 ..

M00
Sym
-
-0-13- 2.3 Streamlines
4 for Indicated
--- 3.0
Blockage, percent
----- 5.0

rr.
-....
rr.
0
3 Bow Shock
- .-. 10

for M00 = 2.3 3.0 5.0

.
c
0
.:;:::-. ::=r'.~._ . - - - --.:.: - - = 2'0
'rl

I
.j-l
'rl 2
rJl
VI
-.J
0
0.
~~.Q G £3 a-;;rcu'O'*qjZ·UC.·- 50
.-1 ;
ell
"rl
'0 1
ell
rr.

o
o 2 4 6 8 10 12 14 16
Axial Position, X/R a

Figure 19. Streamline trajectories at various supersonic free-flight conditions

about a 3-cal tangent/ogive body.
»
m
o
o
~
Jl
00
o
W
U1
 :l>
m
o
()

~
:D
00
a
W
(J'I

-0.4 !?~
Infinite Stream

c,
-0-0- With Shifted. Mismatched Jet Stretcher
u First Internal Iteration
. -0.2
.j..)
c Second Internal Iteration
IIJ
.....
.....U
'H
'H 0
IIJ 6 8 10 12 14 16
o
U
VI Axial Position. X/R
00 IIJ o
H
;:l
~ 0.2
IIJ
H
P.

0.4

a. Comparison of body pressure coefficients

Figure 20. Computed flow conditions near a 3-cal tangent/ogive
body with iterative jet stretcher adjustments.
 AEDC-TR-80-35

Infinite Stream
~ Shifted, Mismatched Jet Stretcher
-- ---- Second Internal I tera t ion

0.8

0.6

0.4

0.2

o
2 (Unaffected by Jet Stretcher)

0.8

0.6

0.4
E
,..o
.... 0.2

0.8

0.6

0.4

0.2

o
2.0 2. 2 2.4 60 70 80 90 100 110 120 o 2 4 6 8 10
3)
Local P/P T (x 10 Flow Angle, deg
Mach a>

Number

b. Local Mach number, static pressure and flow angle

upstream of shoulder
Figure 20. Continued.

59
AEDC-TR-80-35

Infinite Stream
~ With Shifted ~ismatched Jet Stretcher
_._._.- First Internal Iteration
- - - - - Second Internal Iteration
0.8

0.6

0.4

0.2

o
6

0.8

0.6

0.4

0.2
c
........,c o
....
w
o
0,

........,~ 0.8

~ 0.6
(1)
0::
0.4

0.2

o
2.2 2.4 2.6 60 70 80 90 100110 -2 -1 o 1 2 3 4
3)
Local P/PT~(x 10 Flow Angle, deg
Mach ~

Number
X/Ho = 12

c. Local Mach number, static pressure, and flow angle

downstream of shoulder
Figure 20. Continued.

60
 AEDC-TR-80-35

Ideal Moo = 2.3 50-percent Streamline

-1:3-13- Shifted Moo = 5.0 Jet Stretcher
10
First Internal Iteration
Second Internal Iteration

4 1
Jet
Stretcher
Lip
Location
2
j
Axial Position. .'-... /"~16
'\

-2

d. Comparison of flow angles at jet stretcher

Figure 20. Concluded.

61
 »
m
o
(")

~
JJ
00
o
W
(J1
3.00

2.50 Irregular Reflections

0
0::
........
0::
. 2.00
I::
.....0
+-l
..... 1.50
CIJ
0
0..
.-<
.....til 1.00
Q\
l--J -e
til
0::
Nonparallel Lines
0.50 Evidence of Mismatch

0
0 2.00 4.00 6.00 8.00 10.00 12.00 14.00
Axial Position, X/R o

Figure 21. Characteristics lines for second internal iteration.

\
 A E DC-TR -80-35

Jet Boundary
Double Wedge
To Test Article Support

Extraneous Waves
Cone Probe
" Flow Angle
Pressure)

a. Single reference plane

~Mach Lines
Adequate Separation
for Required Sensitivity

Pressure Taps

b. Two reference plane

Figure 22. Preliminary thoughts on possible probe-type sensors
for perforated adaptable supersonic jet stretchers.

63
AEDC·TR·80·35

Table 1. Current Free-Jet Design Goals

OPERATIONAL CAPABILITY

Mach Number Range 0.2-0.8+

1.5 - 2.5
+

Minimum Attitude Variation

Pitch -10 deg - 25 deg
Yaw ±10 deg

NOZZLE FLOW QUALITY

Parameter Subsonic Supersonic

-----­
Mach Number ±0.05 ±0.05
Variation

Flow Angle ±1 deg for Pitch ~ 10 deg

Variation ±0.6 deg

± 1 percent of Pitch Angle

for Pitch> 10 deg

64
 AEDC-TR-BO-35

APPENDIX A
CONVERGENCE OF THE ADAPTABLE JET STRETCHER IN SUPERSONIC FLOW

To investigate the interation procedure for the adaptable jet stretcher in supersonic flow,
an analytical simulation of a jet stretcher flow field is used to establish conditions for
convergence to unconfined flow for supersonic, axisymmetric flow with arbitrary initial
conditions. By this simulation, it then becomes possible to ascertain the rate of convergence
and establish criteria to reduce significantly the number of iterations required to achieve
conditions representative of unconfined flow. The analysis is performed by using a
simplified model of the flow within a jet stretcher to examine critically the fundamental
theoretical validity of the adaptable jet stretcher concept. It should be emphasized that the
analytical simulation presented here differs from the actual test process in the following
respect. The power of the actual experimental adaptable jet stretcher installation is that it
never requires the calculation of the interior flow (which is presumably complicated,
viscous, and shock infested) between the model and the jet stretcher. Only the flow exterior
to the jet stretcher (presumed to be inviscid and with only small disturbances) must be
determined computationally in the actual test installation.

To provide an analytical proof of convergence of the adaptable jet stretcher concept, an

axisymmetric model in supersonic flow will be used. Consequently, the measuring plane, S,
consists of a cylinder of radius, Rs. The flow field can be described by the linearized small
disturbance equation

2 1
~B rb x x ~dJ
r rr --dJ
r r
= 0 (A-I)

where B2 = M2-I, cP is the perturbation potential, x is the axial coordinate, and r the radial
coordinate. For a body of revolution with a radius distribution Rtx), the boundary
conditions for unconfined flow are

lim (r adJ )
r -s o rlr
R (x) d R (x)
dx
~ S "{x )
217
(A-2)

¢ bounded as x 2 + r2 -> 00
(A-3)

where S(x) = 211" R(x)2 is the cross-sectional area of the body. Equation (A-3) is equivalent to
requiring no incoming waves from infinity. The boundary value problem can be solved using
the Heaviside operational transform

p foc g{x)e- PX dx (A-4)

o

65
AEDC-TR-BO-35

with the inverse

g(x) = _1_._
2711
f L
g(p)ePX dp
P
(A-5)

where L is any contour from A - i 00 to A + i 00, A is some finite real constant such that R(p)
> A, and L passes to the right of all singularities of the integrand. The solution for Eqs.
(A-I), (A-2), and (A-3) in the transformed plane is

-s' (p) (A-6)

271

where K, is the modified Bessel function. The corresponding velocity component in the axial
direction at the radius R, is

(A-7)

Uoo(p,R s) is therefore the reference value for unconfined flow at the reference surface R s •
Equation (A-6) can be inverted to the physical plane to yield

d( (A-8)

Equation (A-8) is the classical solution for supersonic flow over a slender body of
revolution.

If the same axisymmetric body is placed in an axisymmetric jet stretcher, it is presumed

that at the reference surface S, interior to the jet stretcher, the pressure distribution is
known. In practice, two flow quantities are measured at S; hence, in a simulation of the
interior flow, it is consistent to assume knowledge of a flow variable. However, in the
simulation it is necessary to calculate the interior flow to find the second flow variable at S.
Again it is emphasized that in a practical application the calculation of the interior flow is
unnecessary. The boundary value problem for the interior region is shown in Fig. A-I. The
resulting solution for the internal flow in the transformed plane is

<b] (p, r )
-s' (p ) '0 (pBr) + [
Gp(P) 5' (p)
(A-9)
271 + 271

66
 AEDC-TR-80-35

- -8
2
':!J r
xx
+ ,.+,
,r
rr
+ r
1
4J r
r
o

.,.;;,......""---"'----"-"'----"--'--""---'-----01""---""--~----------1.---.... x
1 im (r ¢ r )
s (X)
2n
r -" 0 r

Figure A·1. Boundary value problem for the interior region.

67
A E DC·TR·80·35

where G(p) is the transformed known pressure distribution at the surface S, i.e.,

.
G (X) = --
()¢[ /
= "r (X,' R )
, (A-to)
ax r = Rs
S 2

where C, = (p - Poo)/qoo is the pressure coefficient.

Consequently, the normal velocity (i.e., the second flow variable) at the control surface
is

(A-II)
I) (pBR s )

'0 (pBR s )

The next step in the adaptable jet stretcher concept is to use one of the measured flow
variables at S to determine the unique functional relationship with the other flow variable
under conditions of unconfined flow at infinity. If VE (X, Rs) = VI (X, R s) is the flow
variable chosen at the reference surface S, then the exterior region boundary value problem
is that shown in Fig. A-2.

The corresponding solution for the exterior region, in transformed variables, is

VI(p. Rsl K o (pBd

(A-I2)
pBK) (pBR s)

Substitution of Eq. (A-II) yields

(A-B)
I) (a)
+
, (a)
o

where a = pBR s . Consequently, the exterior value of the axial velocity at the reference
surface is

K (a)
o
(A-I4)

68
 ... . t I
f J

cP
E
< = -- ------- X <; "

<,

-,
-13 It E
2
xx
+
F:
IT
+
r
4J
E I'
o -,
c \
>":
\
0\
"'t::!
\
\0
-o
'-'.l
\
PE
r
't, I Given \
r r
\
-------T- -- L
R
S

Figure A-2.
---*-
Boundary value problem for the exterior
unconfined flow regions.
.X
:Po
m
o
o
~
:0
00
o
W
111
A E DC- TR -80-35

The convergence of the iteration procedure will be demonstrated in the transformed

plane, since the functional relationships are in simple algebraic form in that plane. The
iteration is initiated with an arbitrary distribution of tiiO) = <yo) at r = Rs . Following the
iteration scheme shown in Fig. 4, the nth iterative value for til at R, can be obtained by using
Eqs. (A-ll) and (A-13) appropriately

li}n) (p , R
s)
= ca uJn-]) (p, RJ + (l - w) uln- I) (p, R s)
(A-IS)
nne;(o)+A (I-W)
( I - n)

where

II (a) K o (a)
n = (1- w) - w - - - - ­ (A-16)
lo(a)KI(a)

A
S'
- -
(p )
(V
[
- P
Ko (a) ( K Ko
(a) + - - - - -
11 (a) <a)1] (A-17)
217 K (a) I I (a)
I 0

and where w is a relaxation parameter introduced to accelerate convergence of the iterative

procedure. It should be noted that since p is a complex variable. then n, A, and ware in
general also complex variables. As long as /n/ < I is required, then it can be asserted that as
n - 00 , then On -0, hence
lim uI(n) (p , R
A
s) 1- n
n-> ""

S' (p)
(A-18)
217

by comparison with Eq. (A-7). Hence, convergence of the adaptable jet stretcher concept to
unconfined flow is established independent of the initial condition, a(o), for axisymmetric
supersonic flow. For /n/ < / I requires
II (a)
l-w-w--­ < 1
J (a)
o

or

o < co <
2 (A-19)
(a) [I K
o
(a)
1+-­
I (0) K I (a)
o

70
 A E DC-TR -80-35

within this range it may be possible to select a value of w such that convergence to
unconfined flow can be reduced to a finite number of iterations. From Eq. (A-15) it is
obvious that the effect of the initial configuration, (;(0), would be diminished the closer the
magnitude of n is to zero. If n = 0, a value for w can be determined, namely,

w ~ a J0 (a) I( 1 (a) (A-20)

To see the effect of this value of the relaxation factor on convergence of the method,
Eq. (A-20) is substituted into Eq. (A-15) for the case n = I, i.e., the first iteration,

u,U) (p ' s
H ) = A = a Jo (a ) I( , a S' (I')
( ) --:2-17- [
- P ---
,I(K I
(a) + _K_'o_(a_l_'_J(_a_)
, (a)
)J
n
(A-21)
-S' (1')

'217

The value of win Eq. (A-20) is an optimum relaxation factor in that it produces unconfined
flow in a single adjustment of the adaptable jet stretcher.

In principle then, knowing the two velocity components, u\o)(X, R s ) and v\o)(X, R), is
sufficient to determine directly the requirements for unconfined flow. Combining Eqs.

- (A-12), (A-I5), and (A-20), it can be shown that

-
-(,,)
U I \P, R)J
s ~ U I p, J>\ S )
- (o ) (j. (A-22)

Using the convolution theorem for Heaviside transforms, Eq. (A-22) can be inverted to the
physical plane to yield

where the H-I are the inverse Heaviside transforms

(A-24)
217i

(A-25)
217;

71
A E D C-TR -80-35

In principle, Eqs. (A-24) and (A-25) can be inverted either analytically or numerically.
Hence, Eq. (A-23) verifies that if u\o)(X, Rs) and v\O)(X, R s) are known from
measurements at the control surface, the conditions for unconfined flow can be determined
directly without recourse to mathematical modeling of the test article or jet stretcher
configuration.

ln a similar manner it is possible to show that

(A-26)

which can be inverted to the physical plane in the same manner as Eq. (A-22).

NOMENCLATURE

Pressure coefficient (C, = P-P oo/qoo)

D Diameter

k Weighting factor

Length of forebody to shoulder

M Mach number

P Static pressure

q Dynamic pressure

R Radius

Local body radius

Rc Radius of curvature

Re oo Free flight or reference Reynolds number

Aft body radius (nondimensionalizing parameter for all geometric details)

S Reference surface

x Axial position

e Flow angle (positive, away from body axis)

72
 A E DC-TR -80-35

SUBSCRIPTS
-- E Exterior region

Interior region

JB Jet boundary

NE Nozzle exit or duct

us Undisturbed upstream streamline

00 Free flight or reference conditions

SVPERSCRIPTS

0, 1,2 Computational iteration number

COMPUTER PROGRAM ABBREVIATIONS

A3DMOC Armstrong three-dimensional method of characteristics (Ref. 12)

CNAP Cline nozzle and plume inviscid finite di fference (Ref. 13)

D-N Douglas-Neumann panel method potential flow (Ref. 15)

LMOC Lockheed method of characteristics (Ref. 11)

S-J South-Jameson finite-difference, infinite stream, full potential relaxation

(Ref. 10)

W-M Wehofer-Moger inviscid, finite-difference, time-dependent conservation

(Ref. 14)

73