Jr

L_

AIAA 2001-1809
Hyper-X Flight Engine Ground Testing
for X-43 Flight Risk Reduction
Lawrence D. Huebner, Kenneth E. Rock, Edward G. Ruf
and David W. Witte
NASA Langley Research Center
Hampton, VA
Earl H. Andrews, Jr.
Swales Aerospace
Hampton, VA

AIAA/NAL-NASDA-ISAS
10th International Space Planes and Hypersonic
Systems and Technologies Conference
April 24-27, 2001
Kyoto, Japan
For permission to copy or to republish, contact the American Institute of Aeronautics and
Astronautics, 1801 Alexander Bell Drive, Suite 500, Reston VA, 20191-4344.
I
t
,r

AIAA 2001-1809

HYPER-X FLIGHT ENGINE GROUND TESTING FOR X-43 FLIGHT RISK REDUCTION

Lawrence D. Huebner*, Kenneth E. Rock t, Edward G. Ruf*, and David W. Witte t

NASA Langley Research Center
Hampton, VA, USA

Earl H. Andrews §
Swales Aerospace
Hampton, VA, USA

ACp,cowl pressure coefficient increment due to cowl

Airframe-integrated scramjet engine testing has opening
been completed at Mach 7 flight conditions in the ACp,fuel pressure coefficient increment due to
NASA Langley 8-Foot High Temperature Tunnel as part fueling
of the NASA Hyper-X program. This test provided FMS Force Measurement System
FMU Flight Management Unit
engine performance and operability data, as well as
H enthalpy (BTU/Ib m)
design and database verification, for the Mach 7 flight
tests of the Hyper-X research vehicle (X-43), which will H2 gaseous hydrogen
provide the first-ever airframe-integrated scramjet data HIL Hardware in the Loop
in flight. The Hyper-X Flight Engine, a duplicate Mach 7 HXFE Hyper-X Flight Engine
X-43 scramjet engine, was mounted on an airframe INU Inertial Navigation Unit
structure that duplicated the entire three-dimensional LO2 liquid oxygen
propulsion flowpath from the vehicle leading edge to the M Mach number
vehicle trailing edge. This model was also tested to OFP Operational Flight Program
verify and validate the complete flight-like engine N2 gaseous nitrogen
NASA National Aeronautics and Space
system. This paper describes the subsystems that were
Administration
subjected to flight-like conditions and presents
supporting data. The results from this test help to PAI Propulsion-Airframe Integration
reduce risk for the Mach 7 flights of the X-43. PSC Propulsion Subsystem Control
P pressure (psia)
Ngmcnclature Q heat flux (BTU/ft2sec)
8-Ft. HTT NASA Langley 8-Foot High Temperature q dynamic pressure
Tunnel SCFM standard cubic feet per minute
AETB Alumina-Enhanced Thermal Barrier Sill 4 gaseous silane
AOA, angle of attack (degrees) SiO 2 silicon dioxide
BLA Boeing Lightweight Ablator T temperature (°R)
drag coefficient TPS Thermal Protection System
CD
lift coefficient VFS Vehicle Flowpath Simulator (used with the
CL
CM pitching moment coefficient HXFE)
VIL Vehicle in the Loop
* Aerospace Engineer, Hypersonic Airbreathing Propulsion Branch, Xspindle streamwise location with origin at the
Associate Fellow, AIAA rotation point of the horizontal wing
* Aerospace Engineer, Hypersonic Airbreathing Propulsion Branch, X-43 flight vehicle designation for the Hyper-X
Member, AIAA research vehicles
;t Aerospace Engineer, Hypersonic Airbreathing Propulsion Branch f_ angle of sideslip (degrees)
§ Aerospace Engineer, Assigned to the Hyper-X Program Office,
fuel equivalence ratio
Associate Fellow, AIAA
Subscripts
Copyright © 2001 by the American Institute of Aeronautics comb facility combustor condition
and Astronautics, Inc. No copyright is asserted in the United t total condition
States under Title 17, U. S. Code. The U. S. Government has a freestream condition
royalty-free license to exercise all rights under the copyright isolator condition downstream of the internal inlet
claimed herein for Governmental Purposes. All other rights and upstream of the combustor
are reserved by the copyright owner.

1
American Institute of Aeronautics and Astronautics
 Introduction variations; and, the effects of these differences must be
NASA's Hyper-X program will advance accounted for in design and analysis methods when
technologies for vehicles utilizing hypersonic air- using wind tunnel results as an integral part of the
breathing pr0puision from the laboratory to the flight engine and vehicle design. The development of the
environment by obtaining data on a hydrogen-fueled, Mach 7 X-43 engine flowpath and its integration with an
airframe-integrated, duaI-mode supersonic combustion airframe are described in greater detail in Reference 3.
ramjet (scramjet) propulsion system in flight) These Two of the aforementioned Mach 7 flowpath
data wilI provide the first flight validation of analytical development tests were performed in the 8-Foot High
and computational techniques and wind-tunnel test Temperature Tunnel (8-Ft. HTT) as part of an overall
techniques used to design and analyze this class of effort to understand the major differences between the
vehicle. The Hyper-X program is jointly conducted by preliminary flowpath development databases and the
NASA Langley Research Center and NASA Dryden X-43 flight database. 4 The second of these tests
Flight Research Center. The flight-test project phase of provided the first-ever wind tunnel test of a full-scale,
the program involves the fabrication and flight testing of airframe-integrated, scram jet-engine flowpath at
three unpiloted, autonomous Hyper-X research vehicles, simulated flight conditions. The engine, known as the
designated X-43. These vehicles are fabricated by a Hyper-X FIight Engine (HXFE), is the only fuIl-width
contractor team led by MicroCraft that includes The Mach 7 Hyper-X scramjet engine tested prior to the X-43
Boeing Company and GASL, Inc. 2 flights. The Vehicle Flowpath Simulator (VFS) was an
The first two X-43 flights will be conducted at airframe structure to which the HXFE was mounted.
Mach 7, with the third flight being conducted at Mach The VFS represented a geometrically accurate tip-to-tail
10. Each engine will be tested at a freestream dynamic X-43 propulsion flowpath, including forebody
pressure of approximately 1000 psf. The flights will compression surfaces and aftbody expansion that
commence with a captive carry of the Hyper-X launch replicates the entire undersurface of the X-43.
vehicle (consisting of the X_43 and a modified Pegasus The main objectives of the HXFE/VFS test were:
booster) below the wing of a NASA Dryden B-52 (1) to validate the Mach 7 propulsion database and
aircraft to a Mach number of about 0.6 and an altitude of compare the results to other ground tests and correlate
about 20,000 ft. The launch vehicle will then be the data with the X-43 flight data and (2) to provide
dropped from the B-52 and will ascend to the planned important component and systems verification and
flight condition at approximately 95,000 ft. for the Mach 7 validation prior to flight. Not only were engine
flights and 110,000 ft. for the Mach 10 flight where the operability and performance data acquired during
X-43 will separate from the booster. Once the X-43 has testing, but unique data were acquired to provide
established post-separation, unpowered, controlled realistic estimates of the aero-propulsive vehicle force
flight, the inlet cowl door will open to start the inlet, and and moment increments due to both opening the full-
five seconds of unfueled tare data will be acquired. width cowl door (prior to engine ignition) and the
Following this, the fuel sequence will commence, effects of combustion itself.
resulting in about eight seconds of hydrogen-fueled This paper focuses on the second of these two
powered vehicle operation. Once the fuel is depleted objectives. Following a brief discussion of the facility
and engine operation is complete, unfueled tare data will and the model, the components and subsystems that
again be acquired, aerodynamic parameter identification were verified and validated will be presented.
maneuvers will be performed, the inlet door will be Exercising and demonstrating these components and
closed, and the X-43 will fly a controlled, unpowered, subsystems in a flight-like environment inherently
deceleration trajectory until completion of the flight. reduces risk associated with the first flight. The paper
These 700- to 1000-mile autonomous flights of the X-43 will conclude with a presentation of some relevant
will be flown on a due-west heading in the Western Test vehicle data that support both the Mach 7 flights and
Range over thgPacifiC Ocean off the California coast. verify key issues that will be addressed in future
A series of five wind tunnel tests were performed as scramjet-powered vehicle development.
part of Mach 7 propulsion flowpath verification
involving three different engine models in three 8-Foot High Temperature Tunnel
different facilities. These facilities and engines allowed The NASA Langley 8-Ft. HTT 5 has be used to
an integrated test program to be conducted in order to conduct aerothermaI loads, aerothermo-structures, high-
isolate and measure the effects on engine operability and enthalpy aerodynamics, and airbreathing propulsion
performance caused by such details as geometric scale, research. A schematic representation of the 8-Ft. HTT
dynamic-pressure, and test-gas differences. These configured for airbreathing engine testing is shown in
differences exist due to test-technique and facility Figure i. The facility, capable of testing large

2
American Institute of Aeronautics and Astronautics
hypersonic airbreathingpropulsionsystems at flight HX-FE/VFS Model Description
The HXFE/VFS installation in the 8-Ft. HTT is
enthalpiesfromMach4 to Mach7, employs a force
measurement system(FMS)to acquirelongitudinal shown in Figure 2. The HXFE/VFS flowpath is inverted
aero-propulsiveloads(axialforce,normalforce,and from the flight orientation to facilitate fluid and
pitchingmoment)of thetestarticles.TheFMSis instrumentation interfaces and to mitigate strut
attachedto thefacilitymodelelevator carriage,
which interference on the propulsion flowpath caused by
inserts
themodelassembly intothetestsection
oncethe mounting the model to the facility. The model simulates
properflowconditions areestablishedandretracts the the complete X-43 propulsion flowpath, including
modelat run completion.This insertioncapability vehicle geometry that affects the flow entering the
allowsfor thetestingof largermodelsby reducing engine inlet or interacting with the nozzle exhaust
blockage,improving tunnelstarting,reducingtransient plume. The pedestal houses the fuel control system and
loadsonthemodel, andincreasing safety. model instrumentation and provides access for internal
cavity purging/cooling of the airframe structures and
_-_ Air storage water cooling of engine leading edges. Angle-of-attack
(AOA) spacers provided the ability to change model
UlU attitude between runs; the nominal angle of attack for
the first flight is two degrees. Most of the runs were at
/_\ mode, this orientation, but the model was also tested at zero
and four degrees angle of attack to further examine the
Diffuser
design space for which the X-43 pre-flight performance
FMS
database was generated. Special support brackets
provided the ability to vary sideslip angle at the model-
vent slack to-pedestal interface between zero and three degrees.
Configurational details of the HXFE/VFS model are
Figure 1. Schematic drawing of the 8-Ft. HTT for shown in Figure 3. Many of these details are included
aibreathing propulsion testing. as part of the verification and validation for flight and
will be discussed in subsequent sections.
Testing of the HXFE/VFS was performed for three
[ 147.75" -'
- i
dynamic pressures at, above, and below the nominal
Mach 7 X-43 flight condition. The nominal tunnel
entrance _
combustor conditions and resulting simulated air _
.5" Flow
Diffuser

freestream flow parameters are presented in Table 1. =! 143.5" _'l

The simulated Mach number, static pressure, and flight
simulation conditions are within four percent of those
expected for the nominal Mach 7 X-43 flight condition.
Reference 4 contains more information about these test
el I i[ I _.
conditions.

Table 1: Sinmlated Freestream Conditions for 8-Ft. HTT

\. I
IIXFE/VFS Testing and Comparison to Flight

Tunnel Tunnel Target Tunnel

ql I II
Simulation
Case Flight Flight High q_ Figure 2. HXFFJVFS installation in the 8-Ft. HTT.
Low q_ q_ Point

Pcomb (psig) 1000 1585 2000

Air Cowl door- Ventral fin .--_ (,,,-"----'_ 7
Tcomb (OR) 3250 3350 3350

M_

p_ (psia)
6.84
0.140
6.92
0.211
7.00
0.204
6.87
0.263
[11 I t 't 1 I

q_ (psf) 647 I000 1000 1230

Wing stubs _

T_ (OR) 434 423 408 434 (gap size 0 0 to 0 30").-,-'J

Ht(BTU_b m) 1064 1052 1052 1064

Figure 3. HXFFJVFS configurational details.

3
American Institute of Aeronautics and Astronautics
 used in flight to ensure that a fully turbulent flow enters
The HXFE (Figure 4) is a flight-spare, Mach 7 the engine inlet so as to improve inlet operability
Hyper-X scramjet engine that is dedicated to ground marginl 6 Three different configurations were tested: no
testing. Physical attachment of the HXFE to the VFS is trips, the preliminary Mach 7 trips, and the final Mach 7
identical to attachment of the engine in the X-43 via the flight trips. The difference between the two trip designs
six mounting lugs shown. A detailed description of the (Figure 5) is that the trailing edges of the flight trips are
HXFE/VFS model, instrumentation, and interfaces with truncated. Although the trips are not necessarily
the facility are found in Reference 4. required for testing in the 8-Ft. HTT (the freestream
turbulence level of the facility flow is much higher than
in the atmosphere), they were incorporated and their
effects were addressed for a more complete
representation of the flight vehicle. Results on the use
of these trips (Figure 6) show pitot pressure surveys at
the centerline of the model just upstream of the cowl
leading edge at the target flight condition. No
significant difference in the pitot profiles is observed
between the two types of boundary-layer trip designs.
Furthermore, there is only a small increase in pitot
pressure with no trips installed, indicating a small
decrease in boundary-layer height. However, the
Figure 4. Image of the Mach 7 Hyper-X Flight Engine. boundary layer was turbulent for all runs, as previously
mentioned.
Test Summary
Fourteen unfueled runs were made with the HXFE/VFS.
Six of these runs were used to characterize the inlet
flow field plane via pitot pressure and total temperature
rake survey data for the three angles of attack at the
target flight dynamic pressure, two dynamic pressures at
the target flight angle of attack, and the three boundary-
layer trip options (discussed in the next section) at the
target flight dynamic pressure and angle of attack. The
remaining eight unfueled runs addressed cowl-door
actuation and quantification of force and moment
increments for the cowl-opening event.
Forty fueled runs were performed to primarily
examine engine performance and operability. In Figure 5. Comparison of forebody boundary-layer trips.
addition to preprogrammed fueling-sequence runs,
engine control-law development runs incorporating 4_
3.0
closed-loop feedback were performed, in which the fuel AO
v
delivery schedule was adjusted based on real-time O
2.5
sensing of engine pressure data. The issues addressed .2 O
during these runs included thermal effects on the
2.0 0
boundary layer entering the engine, dynamic-pressure
effects, angle-of-attack effects, data repeatability, effects J 1.5
0

of boundary-layer trips, and sideslip effects. O z_ No trips

O Preliminary trips

I_ Flight trips GIx

_5 1.0
Subsystem Verifi¢0tion and Validation
The following subsections provide a description of
subsystems incorporated in the HXFE/VFS model that 0.5

are identical, or nearly identical to the corresponding

tllx
X-43 subsystems. 0.0

Boundary-Layer Trim Pltot Pressure

The first forebody compression surface on the VFS Figure 6. Effect of boundary-layer trips on centerline
(and X-43) contains a slot to accommodate a set of pitot pressure rake data just upstream of the cowl
leading edge, M= =6.92, q_ =1000 psf, (x=2°.
boundary-layer trip devices (see Fig. 3). These trips are

4
American Institute of Aeronautics and Astronautics
AgTB-12 TPS Tiles forty fueled runs, the structural integrity of the engine
For the first ten runs of the HXFE/VFS test, two and the condition of the internal surfaces were nearly

panels of Alumina-Enhanced Thermal Barrier (AETB- 12) the same as prior to testing. The only difference was
thermal protection system (TPS) tiles comprised the that a layer of silicon dioxide (a by-product of using
second and third forebody ramps (see Fig. 3). This is silane for ignition) was usually deposited in the engine
the same material that covers a majority of the X-43 that required manual removal between runs. Even
airframe• A photograph of the AETB-12 TPS tile as though this engine was required to be tested only once
tested is shown in Figure 7. The following results were (in flight), it handled the rigorous ground test series
realized during these runs: (1) the AETB-12 tiles without any problems.
withstood multiple exposures to the flight-like, high The engine contained both static and dynamic seals
aerothermal loads with no degradation, (2) pre-test to control the flow between parts. The static seals
flight-like tile repairs were successfully performed and consisted of ceramic braided rope and were designed to
have leak rates of less than 0.20 SCFM for each inch of
the repaired tiles also withstood flight-like aerothermal
loads, and (3) flight-like installation of and data seal material. The dynamic seals wiping between the
acquisition from pressure taps and thermocouples in the engine cowl door and sidewall were made of l/8-in.
AETB-12 tiles were verified• square braided ceramic carbon-fiber rope with a split
inconel tube that act as a linear spring to keep the seal
material in place. This design functioned well for two
unfueled and five fueled runs; however, the first time the
engine unstarted during a run at flight dynamic pressure,
the seal material on the port side was dislodged.
(Subsequently, a non-flight repair of both the port and
starboard seals was performed to allow testing to
continue.) The implications for flight risk are minimal
because the engine will only see one thermal cycle
during flight, and, additionally, there is logic built into
the propulsion subsystem control (see next main
section) to keep the engine from unstarting due to over-
fueling.

Cowl and Sidewall Leading-Edge Water Cooling

The HXFE utilized sidewall- and cowl-leading edge
water-cooling passages identical to those in the X-43
flight engines. This subsystem during flight cools these
regions of significant aerodynamic heating both during
ascent and scramjet operation. Excessive heating could
result in leading-edge deformation (altering the captured
flowfield) or thermal expansion (causing the moving
cowl door to seize). Water is delivered to three open-
loop cooling paths (see Figure 4) that route the coolant
overboard through four small holes on external surfaces
to the flowpath at the pressure and mass-flow rate
Figure 7. HXFE/VFS during run showing water jets and
AETB-12 TPS tiles (view looking aft from above). expected in flight. Figure 7 shows the top view of the
water jets exiting the HXFE during a tunnel run. No
F,a.gia.e,.Ialr,
g_ problems were encountered with the water-cooling
The HXFE, including body surface, sidewall, cowl, subsystem during ground testing, and flight-like levels
and fuel injectors was fabricated from Glidcop®, a of pressure and mass-flow rate were realized; thus,
copper alloy that is dispersion strengthened with verifying the design of this subsystem.
ultraflne aluminum oxide particles. This yields a metal
Cow[ Actuation and Inlet Startinu
that possesses high strength at elevated temperatures
For the X-43 flight tests, the engine will be closed
and retains high thermal conductivity. Some internal
surfaces were coated with a layer of zirconia TPS to during ascent to protect internal engine components;
therefore, a mechanical actuation system was designed
provide a thermal barrier for the Glidcop® in high-
heating areas. Following fourteen unfueled runs and to open the cowl door and establish flow in (start) the

5
American Institute of Aeronautics and Astronautics
inlet.Therewasaminorconcern ofinletstartingduring The effects of cowl actuation rate on engine inlet
thecowlopeningsequence because of boundary-layerstarting and flowfield establishment were investigated
effects,internalcontractionratio effects,andcowl- by actuating the cowl at flight rate (less than 0.5 seconds
openingspeed.Themechanical nature of theactuation from cowl door closed to open) and at 1/4-flight rate.
designwasfullyreplicated
forHXFE/VFS testing.This No problems were encountered with inlet starting in
design (shown in Figure 8) consists of a linear electro- either case, and no differences were seen in the inlet
mechanical actuator motor and controller that rotates a flow structure.
torque tube connected with cams to connecting links The last run performed during this test series
that are attached through the engine sidewalls to the addressed the issue of cowl-door heating during its
cowl door. During initial testing of the final Mach 7 ascent on the booster to the Mach 7, 95,000-ft altitude,
Hyper-X scramjet flowpath on a partial-width engine test point. The objective of the run was to approximate
prior to the HXFE/VFS test, a similar cowl-actuation the cumulative heat load during ascent and, among other
system was employed. Two slider blocks that are flush things, verify cowl-door actuation. Based on thermal
with the internal engine sidewalls moved with the analysis, subjecting the engine to about 25 seconds of
connecting links and allowed for their proper freedom of Mach 7 flow in the 8-Ft. HTT at a dynamic pressure of
movement. These sliders were made out of the same 1230 psf approximates the heating load that the engine
engine material to very small tolerances. Thermal will encounter during flight from launch-vehicle release
growth and the use of similar metal resulted in galling of to booster separation. The model was in the test flow for
the sliders, which were then redesigned with increased 26 seconds before the cowl door was commanded open.
clearances and coated with a thin layer of chrome. The cowl door actuated to the full-open position in 0.41
HXFE/VFS testing verified the successful redesign of seconds with nominal torque outputs from the cowl
the sliders, which had the potential of preventing the actuator controller and yielded proper inlet flow
cowl door from opening during flight operation. The establishment.
cowl door was actuated 355 times under no
aerodynamic load (primarily during engine internal Wing-Gap Heatin_
inspection and pre-run preparation) and 52 times under The aII-moving horizontal wings that provide pitch
Mach 7 aerothermodynamic loads (during testing). For and roll control for the X-43 are attached to spindles that
all cases, the cowl successfully actuated. This test penetrate the vehicle airframe and are positioned by
allowed for the assessment of the actuation mechanism electromechanical actuators. A small gap exists
and understanding of the specifications for the actuator between the wing root and the vertical face of the
settings (speed and torque levels) required for flight. airframe. Gap heating at hypersonic speeds has been
addressed for a variety of gap configurations, but no data
\, is readily available for longitudinal wing gaps like those
on X-43. A set of wing stubs (truncated at the mid-
lsometrlc view span) were fabricated and a set of thermocouples were
positioned on the airframe surface adjacent to the wing
root to acquire temperature time history data to study
the heating phenomenology for this type of gap. Three
body walls removed)

different gap sizes were studied: no gap, flight gap, and

twice flight gap; data were also acquired with no wing
stubs present as a baseline condition.
A comparison of the chordwise variation in the
f ,/--.Actuator /_C: rrq;: i ruabn: / fst rV_ilCut ree
ratio of heating rates along the wing chord with respect
to the no-wing results is shown in Figure 9. 7 The
preliminary data indicate that heating for the flight gap
is less than 1.2 times the heating with no wing, except
for just upstream of the spindle where the ratio increases
Cowl__j// ((_Cc_n necting lirik _
to 1.3. At twice the flight gap, the heating is generally
Side view the same for the first half of the chord, yet increases in
excess of two times the no-wing heating values are seen
Figure 8. HXFE cowl actuation system schematic.
upstream of the spindle. This verified the thermal
analysis applied to the X-43 structural design in this
area.

6
American Institute of Aeronautics and Astronautics
 2.5i over a wide spectrum of possible conditions and
scenarios without putting unnecessary cycles on vehicIe
systems hardware. Fluid systems performance
verification tests provided complete stand alone flight
fluid systems hardware verification of the integrated
Right gap ,L' fluid systems; however, these tests were not run closed-
1.5 k;J
- - t3- - - Twice flight gap / loop with the PSC flight software hosted on the FMU.
Qa
This was accomplished during Vehicle-in-the-Loop
(VIL) testing in which the complete OFP hosted on the
1
FMU interfaced with the complete fluid systems while a
nominal flight scenario was simulated on the simulation
0.5 bench. Hence, the entire system had been tested with the
exception of scramjet engine operation running in
Wing rootoutline _-- Spindle closed-loop with the propulsion control laws. The
0:;5 .... e .... HXFE/VFS test provided this final element with the
"20 -15i .... I ....
"10 t ....
"5 0I j kL,l,,,,r5 10

X_.d_, (in,) HXFE operation in a wind-tunnel simulated flight

Figure 9. Effect of wing gap on gap heating, M= =6.92, environment. In these tests flight-type fluid and cowl
q_ =1000 psf, o_=2°. actuation systems controlled by a wind-tunnel specific
version of the PSC flight software hosted on a bench
PropulsionSubsystem Control Development. control system provided a closed-loop system with
Verificatiorl and V_lkl_ation scramjet operation. The following subsections discuss
To achieve the Hyper-X flight test objectives of the specific requirements and objectives for these tests
accelerating the X-43 under controlled autonomous free in detail.
flight, closed-loop control of the engine and engine
systems is required. Propulsion Subsystem Control
(PSC) software and hardware were designed and tested VIL Tests

to provide engine-inlet cowl-door control and engine

ignitor and fuel flow-rate control. The HXFE/VFS test Ignilor/Fuel
in the 8-Ft. HTT provided several unique elements to the
Signal Conditioning
overall development, verification, and validation of the 8-Ft. H'Iq" High-Pressure
Engine Operation Side
PSC flight software and engine and fluid systems Te Stin_ and Instrumentation
iil;i
hardware. The overall pre-flight effort consisted of • Inflow,q_ass Capture Fluid Systems
• Cowl Door/Active Cooling/ PerformanCe
multiple modeling, simulation, and test elements Thermal Barriers
• EngineThermal/ verlTtcatlon
integrated to provide a comprehensive verification and Structural Design Tests
• Vehicle PAll
validation of the complete system. Figure 10 illustrates Forces and Moments

the process used to ensure that each hardware and

Figure 10. Propulsion subsystem control development,
software element was thoroughly tested. Although no
verification, and validation schematic.
one test captured the complete system and flight
environment, 'overlapping tests in the integrated
Hardware and Software Requirements
verification and validation plan ensured comprehensive
Control System. A bench control system running
testing, including all interfaces.
the Matlab® Simulink® environment used for flight
The PSC development and simulation environment
software development was used to generate a wind-
provided an isolated computer simulation used to
tunnel specific version of the flight control laws in
develop and perform unit testing on the PSC control
which the control loops were identical to those
laws; however during these tests, the control laws were
imbedded in the flight software, but the interfaces were
neither integrated with the complete flight software nor
modified for wind-tunnel application. A top-level
resident on the flight computer. PSC flight software
schematic representing these control loops and the
performance was verified as part of the complete
required sensed parameters is shown in Figure 11. The
Operational Flight Program (OFP) during Hardware-in-
ignitor and fuel schedule is established based on the
the-Loop (HIt) testing. During HIt testing the flight
sensed flight condition and commanded vehicle attitude
software was hosted on the Flight Management Unit
and is then adjusted based on engine flowpath pressure
(FMU) while all other vehicle systems and flight
conditions were simulated on the simulation bench. feedback and vehicle acceleration feedback. The ignitor
and fuel mass-flow rates are then established through
This approach allowed extensive testing of the software

7
American Institute of Aeronautics and Astronautics
closed-loop mass flow control as indicated in the figure. valves, actuators, and venturi flow meters, as well as the
All signal conditioning, sample rates, and control-loop ignitor/fuel mixing manifold that feeds the bodyside and
rates were matched to the flight systems. Figure 12 cowlside fuel-injectors via tubing of flight-like length.
shows the compact nature of the control system, utilized Flight fuel delivery pressures and flight-like components
in HXFE/VFS testing to simulate the X-43 PSC downstream of the control valves, including flight fuel
functions. injectors, were utilized, and flight-identical instrumenta-
Fn)m, h_e tion was used throughout the fluid systems.

.i_--hqpd,,/Io to bodysido to cowlside injectors

Ni_l eenNt_s

-Ignitor/fuel
From mixing
Ignitor manifold
_ _lsum, to eslufl venturi
flowmeter-
ventud
flowmeter

aoee[er on
Ignitor
control valve
dyna n"4cpreasure, and actuator-
and sng_e of altack
X-43 Math number, r-_w._._ -_._
a_or
control valve
and actuator
Figure I 1. Ignitor and fuel control block diagram and
schematic drawing.

Figure 13. Flight-like ignitor and fuel system details

(located in the pedestal just below the model).

Elowpath Sensors. Multiple flowpath pressure

sensors were used for feedback as part of the PSC for
engine-unstart prevention. The sensors and their
installation were identical to those in the X-43 as shown
in the photograph of the HXFE in Fig. 4.

Control Law Verification

Ignition an_ Transition to Hydrogen Fuel. X-43
packaging constraints resulted in a limited supply of
both ignitor (mixture of 20% molar silane in hydrogen)
and hydrogen fuel, which subsequently resulted in a
limited duration test window in which to design an
ignitor and fuel sequence to meet the test objectives.
The ignitor and fuel flow rate profiles required for
ignition, transition to hydrogen-only fueling, and to
achieve vehicle acceleration were developed and
verified during these tests. This was achieved while
Figure 12. PSC computer used during I-IXFE/VFS testing. meeting the additional requirements of acquiring
steady-state enginelvehicle data and preventing engine
Ignitor and Fuel System. In order to verify the unstart due to over-fueling. The ability to perform these
inner-loops controlling the ignitor and fuel mass flow tests on the flight hardware at simulated flight
rates, the fluid system plumbing (including fittings, line conditions provided significant risk reduction to the
lengths and diameters), control valves, venturi flow flight test.
meters, and sensors had to be accurately simulated. To Engine-Unstart Prevention. The Hyper-X flight
achieve this, flight-type valves with similar mechanical project was aided by concurrent engineering in many
and fluid dynamic characteristics were used. Figure 13 key areas. While the HXFE/VFS test provided the first
shows the pedestal internal cavity just below the model scramjet data on the full-width, full-length engine/
that contains the closely-coupled ignitor and fuel control vehicle propulsion flowpath, the propulsion system

8
American Institute of Aeronautics and Astronautics
hardware, instrumentation, and control system software implementing this logic were finalized and verified
architecture was already in place. The final flight during these tests and data were acquired for the basis of
control laws were developed and verified within this setting the TripLevell and TripLevel2 constants used in
context. One of the most important features of PSC, flight.
whose design was finalized and subsequently verified as
part of these tests, is to prevent engine unstart during the Vehicle Performance
flight experiment, but still achieve vehicle acceleration. Because of the integrated nature of scramjet-
The control logic used to achieve this is presented in powered configurations, the basic aerodynamic
characteristics of this vehicle are strongly coupled with
Figure 14.
the propulsion system effects. Prior to HXFE/VFS
testing, estimates of the longitudinal force and moment
Inlerl! _n data were determined to develop the vehicle
__ 1 Reduce H e flow rate;
performance, stability, and control characteristics for the
I _ _ hold constant for 0.75 sec X-43 flight control laws. 8 These estimates were
developed from a combination of aerodynamic wind-
I tunnel testing of the closed-cowl configuration and the
I IContir,
u.fol owingl
t t \ / Hyper-X methodology of using analytical and
computational results to estimate the cowl opening and
powered effects. 9 The HXFE/VFS was tested at the off-
nominal angles of attack of zero and four degrees to
complement the data taken at the nominal flight angle of
attack of two degrees. This allowed better examination
Hold H2 flow rate of the design space from which the pre-flight
cor',stant =1deple ed J
• I performance database was generated. Furthermore,
excursions of one and three degrees in sideslip angle
Figure 14. Fuel control block diagram for unstart avoidance were addressed. The concern was that, at hypersonic
and meeting program objective. speeds, even a minor perturbation in sideslip angle from
zero might have an adverse effect on vehicle
Once engine ignition has been achieved and the
performance. This is due to an expansion fan (created
engine and vehicle have stabilized for acquisition of on the windward sidewall leading edge) and a shock
steady data, fuel flow is increased and this logic wave (created on the leeward sidewall leading edge)
becomes active. If the engine flowpath pressures
both being processed by the inlet.
indicate that combustor-isolator interaction is occurring,
Figure 15 presents a comparison of lift, drag, and
and simultaneously exceeding a pre-set lower limit
pitching moment coefficients (with no horizontal tail
(TripLevell), then the pre-programmed fuel schedule is deflection) for the cowl closed, cowl-open unfueled, and
interrupted. The fuel flow is reduced and held constant cowl-open fueled conditions, Data shown are for
for 0.75 seconds, after which the PSC evaluates vehicle comparative purposes. The database estimates, shown
acceleration based on inputs from the inertial navigation by the solid symbols, were developed by applying the
unit (INU). If the pre-set acceleration goal has been
met, the fuel flow is held constant until a controlled computed cowl-opening and power-on increments to the
fuel is experimentally derived aerodynamic database for the
flow-rate ramp down is commanded as the cowl-closed configuration. The increments included in
depleted; however, if the pre-set acceleration goal has the figure are the differences in the force and moment
not been met, the level of combustor-isolator interaction values between the unfueled cowl-closed and cowl-open
is evaluated based on a pre-set higher limit (TripLevel2). conditions and between the unfueled and fueled (at d_=
Additionally, if this limit is exceeded the fuel flow is 1.2) cowl-open conditions. The force and moment
held constant until the controlled ramp down is reached. increments acquired from the 8-Ft. HTT FMS are
If the acceleration goal has not been met and combustor-
applied to the same cowl-closed aerodynamic wind
isolator interaction is lower than the higher limit, then tunnel data to create the HXFE/VFS data and are
the fuel ramp-up is resumed. The intent of this logic
depicted by the open symbols.
sequence is to maximize the probability of achieving the In general, there is very good agreement between
acceleration goal while concurrently guarding against the estimated and measured increments. Where the
engine unstart. The two-level interaction criteria comparisons differ the most (C D fueled at 0_=4°), the
balances the risk of inlet unstart against the importance
experimental results actually show improved
of attaining the acceleration goal. The control laws
performance (less drag, i.e., more thrust). These results

9
American Institute of Aeronautics and Astronautics
are significant because they build confidence in the Propulsion-Airframe Integration
Hyper-X pre-flight database methodology for this One of the more important, yet often neglected,
configuration at conditions near the target scramjet test issues that needs to be addressed in a hypersonic air-
point, as well as demonstrate that there is a significant breathing propulsion vehicle design effort is propulsion-
effect of forebody and aftbody pressurization that airfi'ame integration (PAD. For this class of vehicle, PAI
affects both the aero-propulsive loads and the pitching- addresses the integration of a dual-mode scramjet to a
moment values and, thus, the ability to trim the vehicle. slender-body airframe capable of hypersonic speeds, as
Since associated trim-drag penalties have a significant well as the aero-propulsive interactions of the highly-
impact on net vehicle thrust, the ability to accurately coupled internal and external flowfields. These
estimate (and measure) this component in powered interactions include inlet spillage, exhaust pluming and
flowpath analysis and testing is important. Also the resulting impact on lifting and control surfaces, and
included in the figure are the results for the two nonzero engine/airframe forces and moments.
sideslip runs at c_ = 2 °. Results indicate that there is The test of the HXFE/VFS provided a unique
little, if any, degradation in aero-propulsive performance opportunity to address PAI since the test article
at these conditions even though the inlet possesses an comprises the key elements needed to assess the
asymmetric pressure field. interactions between the scramjet (HXFE) and a
realistic airframe surface (VFS). The forces and
moments issue was discussed in the previous section,
1" from aerodynamic data, cowl closed Samples of the external flowfield and exhaust pluming
- "41'- - Hyper-X me_odotogy, cowl open, no fuel
- -,lb.-- Hyper-X rnethodoiooy, fueled, $=1 2
results follow.
E] HXFENFS, cowl open, unfueled
C _,O,J .. "" O HXFFJVFS, cowl Open, fueled, $,=1,2
_> HXFENFS, _=t deg
4 HXFENFS, _--3 deg
Schliercn Experimental Flowfield Results.
Figure 16 depicts composite images of the flow
patterns emanating from the flowpath surface of the
....... o- ...... HXFE/VFS during the three distinct stages in the
C. scramjet test. For all runs, the shock structure upstream
of the cowl leading edge was quite stable, and included
the bow shock, weak waves from the boundary-layer
0_
trips, and additional compression emanating at each of
Figure 15. X-43 force and moment data versus a, the two forebody ramps. These shocks are part of the
M_ =6.92, q_ =I000 psf. overall compression process required to obtain the

Figure 16. Comparison of composite schlieren images for distinct scramjet test events, Moo=6.92, q_ =1000 psf, o_=2°.

10
American Institute of Aeronautics and Astronautics
desired pressure levels entering the combustor. A strong wing placement was verified in that the exhaust plume,
shock off the cowl leading edge is seen above the engine fueling levels, and other engine operability factors did
(as seen in the figure) in the cowl-closed configuration, not significantly affect control surface effectiveness;
and there is indication of significant shear flows and a thus, validating the aero-propulsive database, which
separation zone that occupies the first third of the assumed no power-on effects on the control surfaces.
aftbody behind the cowl trailing edge. When the cowl
door is opened, the shock generated off the cowl leading
edge changes, and the shear layer downstream of the
cowl trailing edge is seen by the alignment of the
cooling water jets (dark region) above the aftbody
surface. During fueled operation, the plume caused by
the increased engine pressure levels due to combustion
has shifted the shear layer farther away from the aftbody
surface. Waves seen downstream of the cowl-trailing
edge result from the internal nozzle geometry.
SiO2 Deoosits and Oil-Flow Results
Figure 17 indicates the lateral extent of the plume
expansion on the aftbody, evidenced in the deposition of
silicon dioxide (SiO 2, a white particulate and by-product
of the reaction of the silane/hydrogen ignitor gas
combustion process). The plume is seen to extend
laterally outward beyond the cusp line that defines the
aftbody nozzle surface. Vortical flow exists with a
separation/reattachment region shown, as evidenced by
the lack of SiO 2 just outside, and nearly parallel to, the Figure 18. VFS aftbody oil-flow patterns following a
cusp lines. fueled run, M_ =6.92, q =1000 psf, cz=2°.

Aftbody Pressure Distributions

The effects of cowl opening and fueling are
observed in the aftbody pressure distributions (Figures
19 and 20). Pressure-difference contours (in coefficient
form) were created from the discrete surface-pressure
data (identified by the small circles on the figures) such
that the difference in the pressure is calculated in the
following manner:
ACp, cowl = (Cp, cowl-open - Cp, cowl-closed) (1)

aCp, fuel = (Cp, _>0 - Cp, _--0) (2)

Figure 19 presents ACp, cowl data for two different runs,

both at the nominal flight Mach number, angle of attack,
and dynamic pressure. The difference is that Fig. 19a is
Figure 17. SiO 2 deposits on VFS aftbody following a fueled at 13=0° and Fig. 19b is at 13=3°. For reference, the cowl
run, M_ =6.92, q_ =1000 psf, cz=2°. trailing edge is located at the left middle of each
contour, and the wing stubs and ventral fins are shown
Aftbody oil-flow patterns (Figure 18) also indicate for clarity. When the cowl opens, the aftbody pressures
the symmetric nature of the powered exhaust flow. The react to the processing of the flow through the engine in
flow is symmetrically aligned relative to the vehicle a very symmetric sense for 13=0° , but very
centerline, and lateral expansion is visible as the plume
asymmetrically for 13=3° , albeit at about the same inflow
expands into the aftbody nozzle (including just
pressure level. The shock and expansion waves
downstream of the cowl trailing edge). In addition, no
emanating from the cowl sidewall leading edges are still
SiO 2 deposits were observed on the wing-stub surfaces, evident in the aftbody flowfield. The slight increase in
and there were no discernible differences in the wing-
windward chine pressures near the ventral fin is also
stub, oil-flow patterns for unpowered and powered runs. observed. When the engine is operating at the flight-
This finding is important because it indicates that proper

11
American Institute of Aeronautics and Astronautics
designed hydrogen fuel level, ACp, fuel contours (Figure
20) indicate that the external nozzle flowfield is very
similar for both sideslip angles, although the slight 5Cp, fuel
increase in chine pressures near the ventral fins indicates
that measurable lateral expansion is occurring under
both powered conditions. Furthermore, differences in
the pressures on the chines is still evident for the 13=3°
run.

), cowl
2.2
2.0
1.8
1.6
1.4
1.2
a. 13=0°.
f,0
0.8 Windward
0.6
0,4
ACp, fuel
0.2
0.0
-02

a. 13=0°.

Windward
ACp, cowl
2.2
2.0
1.8
Leeward
1.6
1.4
1.2
b. 13=3°.
1.0
0.8
0.6 Figure 20. Aftbody pressure coefficient increase due to
0,4 fueling, M**=6.92, q =1000 psf, _=2 °.
0.2
0,0
-0.2

Engine Unstart/Restart Capability

During the combustion process in a dual-mode
Leeward scram jet, high pressure develops in the combustor that,
with increased fueling, back-pressures the isolator and
b. 13=3
°.
ultimately can "unstart" the inlet (analogous to
Figure 19. Aftbody pressure coefficient increase due to cowl compressor stall in a turbojet). Although engine
opening, M_ =6.92, q_ = 1000 psf, (x=2°. controls can be used to throttle the engine and prevent
unstart, it is imperative to demonstrate the ability to "re-
Non-Flight Verification start" the inlet and combustion process. The PSC
Two significant studies were performed during this software used in HXFE/VFS testing was sufficiently
test that, although not specifically impacting the X-43 robust to allow, with minimal effort, the ability to test
Mach 7 flights, are significant in the future development this capability. One successful run demonstrated this by
of hypersonic air-breathing propulsion vehicles. One of purposely causing the HXFE inlet to unstart (via excess
these is related to propulsion control and the other is fueling), followed by rapidly throttling-down the fuel,
related to vehicle candidate materials and their effect on restarting the inlet flow by actuating the cowl door
engine performance. closed then open, and subsequently re-igniting the
engine. The entire restart process was achieved in 1.76
seconds, and the data indicate that it could have been

12
American Institute of Aeronautics and Astronautics
,,.;.

achieved significantly faster. (Note that no effort was 1000-psf dynamic pressure (both extremes for this
made to optimize the process, only to demonstrate TPS), 22-second run was very good; however, there was
feasibility.) In addition to demonstrating engine restart, a modest amount of surface pitting, primarily of the
the analysis of the force and moment data acquired forebody panel which appeared to be a combination of
during the unstart/restart process provides valuable impact and crack bridging on the surface (as seen in the
insight into vehicle dynamics and is an important figures). Lap joints and a repair of simulated damage on
component of the ability to establish and maintain the aftbody panel survived intact after each run.
controlled hypersonic flight. Temperature measurements acquired on the TPS/
substrate interface showed no appreciable temperature
Ablative TPS Effects rise during the runs and a maximum soak-back
The HXFE/VFS model was also employed to verify temperature rise of 35 F° was obtained on the forebody
the use of a new lightweight ablative TPS developed by panel six minutes after the run. During the runs, the
The Boeing Company. For two tunnel runs, the final HXFE was fueled with an X-43 flight-like sequence that
forebody compression ramp (just before the HXFE cowl ramped hydrogen to the target flight fuel equivalence
leading edge) and the first half of the external nozzle ratio. This is the first time this type of ablator material
were replaced with TPS panels known as BLA-20 (see has been tested with an operating engine. The HXFE lit
Figures 21 and 22, top). When the BLA-20 is subjected well for both runs and maintained robust combustion
to heat, organic compounds within the TPS ablate, throughout. The engine performance did not appear to
leaving a charred tile-like silica surface that maintains be affected by the aNator out-gassing. The second run
its geometric characteristics. The panels were thermally was the first time that this monolithic BLA-20 TPS was
isolated from the surrounding copper external surfaces. re-tested and provided data to address reusability of the
The objectives of these runs were to determine if the TPS for non-expendable applications. The TPS panels
out-gassing from the ablator affected engine were returned to The Boeing Company for measurement
performance and operability and to determine of mass loss, char density, and amount of recession.
survivability of the TPS both ahead of and behind the
engine. Overall, TPS survivability during the Mach 7

Figure 22. Comparison of pre-and post-run

aftbody BLA-20 TPS.
Figure 21. Comparison of pre-and post-ran
forebody BLA-20 TPS.

13
American Institute of Aeronautics and Astronautics
 Summary staff to incorporate complex engine support hardware
This paper presents a description of, and supporting and software, acquire a significant quantity of quality
data on, the significance of the Hyper-X Flight Engine/ model data, and handle real-time changes in test plans.
Vehicle Flowpath Simulator tests performed in the The authors thank the 8-Ft. HTT staff for their
NASA Langley 8-Foot High Temperature Tunnel in dedication and hard work.
support of risk reduction for the Mach 7 flights of the
X-43. Following a brief description of the facility, References
model, and test summary, the engine and flowpath 1. Rausch, V. L.: McClinton, C. R.; and Crawford, J. L.:
subsystems that were verified and validated for flight "Hyper-X: Flight Validation of Hypersonic
were documented. These include the use of forebody Airbreathing Propulsion." ISABE 97-7024,
boundary-layer trips, cowl actuation, water cooling of September 1997.
engine leading edges, engine structural integrity, wing 2. Rausch, V. L.; McClinton, C. R.; and Sitz, J. R.:
gap heating, and propulsion subsystem control hardware "The Hyper-X Program Overview" ISABE 99-
and software. Following this, supporting data was 7213, September 1999.
presented which included vehicle performance and 3. Rock, K. E.; Voland, R. T.; Rogers, R. C.; and
propulsion-airframe integration. Finally, a description Huebner, L. D.: NASA's Hyper-X Scramjet Engine
of two efforts that constitute future scramjet-powered Ground Test Program." ISABE 99-7214,
vehicle capabilities were presented; namely, the ability September 1999.
to restart the engine following an unstart and the use of 4. Huebner, L. D.; Rock, K. E.; Witte, D. W.; Ruf, E.
ablative TPS tiles on the engine flowpath. G.; and Andrews, Jr., E. H.: "Hyper-X Engine
This test was unique in that it addressed many Testing in the NASA Langley 8-Foot High
issues for the first time: (1) full airframe-integrated, Temperature Tunnel." AIAA 2000-3605, July
scramjet-powered model test in a wind tunnel at flight 2000.
conditions, (2) dual-mode scram jet flight control testing 5. Hodge, J. S.; and Harvin, S. F.: "Test Capabilities
for an airframe-integrated Scramjet-powered vehicle, (3) and Recent Experiences in the NASA Langley 8-
yaw-effect testing of a scramjet, (4) scramjet testing Foot High Temperature Tunnel." AIAA 2000-
using ablative TPS tiles on the forebody and external 2646, June 2000.
nozzle, and (5) engine data to validate conventional 6. Berry, S. A.; Auslender, A. H., Dilley, A. D.; and
scramjet module test techniques. In addition, this test Calleja, J.F.: "Hypersonic Boundary-Layer Trip
program (1) provided propulsion performance and Development for Hyper-X" AIAA 2000-4012,
operability over a range of dynamic pressures and August 2000.
angles of attaCk, (2) Validated the scramjet test article 7. Daryabeigi, K.; and Huebner, L. D.: "Heating Rate
structure and functionality, (3) verified the powered Measurements on Hyper-X Wing Gap." Proposed
aerodynamic database originally developed using NASA Technical Memorandum, Draft completed
aerodynamic models, CFD, and other analytical December 2000.
methods, and (4) was useful in quantifying the 8. Engelund, W. C.; Holland, S. D.; Cockrell, Jr., C. E.;
performance impact of partial-width versus full-scale and Bittner, R. D.: "Propulsion System Airframe
testing. Integration Issues and Aerodynamic Database
Development for the Hyper-X Research Vehicle."
Acknowledgement ISABE 99-7215, September 1999.
This test demonstrated the outstanding capabilities 9. Cockrell, Jr., C. E.; Engelund, W. C.; Bittner, R. D.;
of the 8-Foot High Temperature Tunnel to support large- Dilley, A. D.; and Jentink, T. N.: "Integrated Aero-
scale testing of hypersonic airbreathing propulsion Propulsive CFD Methodology for the Hyper-X
systems. Facility operations were executed well; a Flight Experiment." AIAA 2000-4010, August
thorough characterization of the tunnel test section flow 2000.
field exists; and there was active participation by the

I4
American Institute of Aeronautics and Astronautics