NASA Technical Memorandum 4722

Design and Development of

an F/A-18 Inlet Distortion
Rake: A Cost and Time
Saving Solution

Andrew J. Yuhas, Ronald J. Ray, Richard R. Burley,

William G. Steenken, Leon Lechtenberg, and Don Thornton

October 1995
NASA Technical Memorandum 4722

Design and Development of

an F/A-18 Inlet Distortion
Rake: A Cost and Time
Saving Solution

Andrew J. Yuhas
PRC Inc.
Edwards, California

Ronald J. Ray
Dryden Flight Research Center
Edwards, California

Richard R. Burley
NASA Lewis Research Center
Cleveland, Ohio

William G. Steenken, Leon Lechtenberg

and Don Thornton
General Electric Aircraft Engines
Evendale, Ohio

National Aeronautics and

Space Administration
Office of Management
Scientific and Technical
Information Program
1995
 DESIGN AND DEVELOPMENT OF AN F/A-18 INLET DISTORTION RAKE:
A COST AND TIME SAVING SOLUTION
Andrew J. Yuhas*
PRC Inc.
Edwards, California

Ronald J. Ray**
NASA Dryden Flight Research Center
Edwards, California

Richard R. Burley†
NASA Lewis Research Center
Cleveland, Ohio

William G. Steenken,†† Leon Lechtenberg‡, and Don Thornton‡‡

General Electric Aircraft Engines
Evendale, Ohio

ABSTRACT ARP Aerospace Recommended Practice

ac alternating current
An innovative inlet total-pressure-distortion mea-
surement rake has been designed and developed for the CFD computational fluid dynamics
F/A-18 A/B/C/D aircraft inlet. The design was con- DFRC Dryden Flight Research Center,
ceived by NASA and General Electric Aircraft Engines Edwards, California
personnel. This rake has been flight qualified and flown
in the F/A-18 High Alpha Research Vehicle at NASA F/A fighter–attack aircraft
Dryden Flight Research Center, Edwards, California. FOD foreign object damage
The eight-legged, one-piece, wagon wheel design of GE General Electric
the rake was developed at a reduced cost and offered
reduced installation time compared to traditional GEAE General Electric Aircraft Engines,
designs. The rake features 40 dual-measurement ports Evendale, Ohio
for low- and high-frequency pressure measurements g inertial force expressed in multiples of
with the high-frequency transducer mounted at the gravity
port. This high-frequency transducer offers direct
absolute pressure measurements from low to high fre- HARV High Alpha Research Vehicle
quencies of interest, thereby allowing the rake to be HATP High Alpha Technology Program
used during highly dynamic aircraft maneuvers. Out-
ID inside diameter
standing structural characteristics are inherent to the
design through its construction and use of lightweight LeRC Lewis Research Center, Cleveland,
materials. Ohio
M Mach number
NOMENCLATURE
NASA National Aeronautics and Space
Acronyms Administration
NASTRAN NASA Structural Analysis, a finite
AIP aerodynamic interface plane element modeling program
AIR Aerospace Information Report OD outside diameter
ALF aft looking forward PSI Pressure Systems Incorporated,
AOA angle of attack, deg Hampton, Virginia
PSIPP pounds force per square inch, peak to stabilized and highly dynamic maneuvers at high
peak angles of attack.
RTV room temperature vulcanizing Traditional inlet rake systems typically use eight-
legged, duct-mounted cantilevered designs, such as
SAE Society of Automotive Engineers those used on the developmental F/A-18A1 or inlet
X experimental aircraft guide vane leading-edge designs.2 These designs have
commonly required extensive modifications to the air-
Symbols craft or the engine. In particular, the cantilevered sys-
tem, with its individual rake leg designs and
σ standard deviation installations, can be costly, because of separate rake leg
development and testing. In addition, this system can
INTRODUCTION be time consuming because of increased structural
modifications of the aircraft and flight qualification
Designing, developing, and installing an inlet testing.
total-pressure-distortion rake can be expensive and Figure 2 shows an inlet rake which was developed by
time consuming. Figure 1 shows the F/A-18 High the NASA Lewis Research Center (LeRC), Cleveland,
Alpha Research Vehicle (HARV) being flown at NASA Ohio, and DFRC. This rake was based on an innovative
Dryden Flight Research Center (DFRC), Edwards, Cal- design concept conceived by NASA and General Elec-
ifornia. The HARV inlet research program goals tric Aircraft Engines (GEAE), (Evendale, Ohio). This
required a low-cost, short-installation-time, and low- approach connects all of the rake legs together at the
maintenance solution to meet the need for a total- hub to form a one-piece, wagon wheel design. This
pressure inlet-distortion measurement system. The design simplifies rake installation and aircraft modifi-
HARV inlet research program evaluates the inlet cation requirements, thus greatly reducing cost and
characteristics of high-performance aircraft during weight. The rake, designed and fabricated by GEAE,

EC 91 495-15
Figure 1. NASA F/A-18 HARV aircraft (preproduction aircraft number 6 modified with multiaxis thrust vectoring
paddles).

2
has passed the needed flight qualification requirements but were not limited to instrumentation setup, rake
and has been flown on the HARV. positioning, and probe configuration. The design also
This paper describes the design, fabrication, installa- had to allow accurate data measurements to be gath-
tion, and qualification testing of the HARV inlet rake ered to meet all of the HARV inlet research objectives.
system. Comparisons of cost and installation time This requirement primarily concerned instrumentation
between this design and a previous design are made. selection. In addition, following established industry
The paper also details the design requirements and guidelines wherever possible was desired. Consider-
pressure transducer selection. All stages of flight quali- ations of cost and installation time are always impor-
fication testing, from laboratory to flight test, are tant factors that ultimately constrained the design
described. requirements.
Use of tradenames or names of manufacturers in this
paper does not constitute an official endorsement of Commonality
such products or manufacturers, either expressed or
implied, by the National Aeronautics and Space Inlet instrumentation was required on the HARV to
Administration. correlate flight data with and verify test results from the
planned 9- by 15-ft wind tunnel tests scheduled at the
INLET RAKE SYSTEM REQUIREMENTS LeRC as part of the High Alpha Technology Program
(HATP) inlet research program. In that aspect, the
When considering an inlet rake design for use on the HARV instrumentation was developed to emulate the
F/A-18 HARV flight program, the basic requirements wind tunnel set, limited primarily by the cost and com-
were first established. The most important requirement plexity of modifying the full-scale vehicle.
was to provide as much commonality as practical with Having commonality between the current flight test
the planned HARV inlet wind tunnel test at the LeRC rake and previous F/A-18 flight test rakes was also
and with previous F/A-18 inlet testing. Commonality desirable. In the mid-1970’s, the U.S. Navy along with
considerations with past and present testing included industry partners, McDonnell Douglas Aerospace

EC 93 41084-7
Figure 2. NASA and GEAE inlet pressure distortion rake mounted in HARV right inlet.

3
(St. Louis, Missouri), Northrop Aircraft Division Engineers (SAE) in 1978 to ensure a consistent
(Newbury Park, California), and GEAE conducted an approach to the development of the inlet instrumenta-
inlet evaluation on the second preproduction F/A-18A. tion configuration and to provide a proven and consis-
The original inlet rake design consisted of eight inde- tent method of data analysis.3 The SAE document
pendent, cantilevered rakes, each spaced equiangularly, recommends that the instrumentation and analysis
having five measurement ports per rake located on the methods be agreed upon among the involved parties
centroids of five equal areas.1 Each rake was mounted and remain invariant throughout the propulsion system
to various structural members behind the inlet duct life cycle for all testing. This general approach was fol-
wall. Because of the forces and moments acting on lowed for the design and manufacture of the HARV
each cantilevered rake, special attention was given to inlet rake described here. In 1983, SAE issued a more
the structural design and buildup. To meet the complex comprehensive report, Inlet Total-Pressure-Distortion
inlet rake structural requirements, the bulkhead on air- Considerations for Gas-Turbine Engines, AIR 1419,
craft number 2 was specifically designed to accommo- that provides more detailed information.2
date the inlet rake-mounting requirements. The cost The SAE established the “aerodynamic interface
and time required for this modification was obviously plane” (AIP) as the location of the instrumentation
significant. Figure 3 shows the original rake installed in plane used to define inlet distortion and performance.
aircraft number 2. This original inlet rake design gener- In general, the guide recommended that the AIP should
ally followed the instrumentation guidelines estab- be located in a circular duct as close to the engine face
lished by industry. 2,3 as practical. The engine face is defined by the leading
edge of the most upstream engine strut, vane, or blade
Technical Guidelines row. To be consistent with past F/A-18 testing, the
HARV AIP was required to be 4 in. in front of the
An Aerospace Recommended Practice titled Gas engine bullet nose. The standard also established a typ-
Turbine Engine Inlet Flow Distortion Guidelines, ARP ical 40-port rake array for distribution of the total pres-
1420, was established by the Society of Automotive sure ports at the AIP. This array consists of eight

EC 91 220-2
Figure 3. Preproduction F/A-18A aircraft number 2 cantilevered inlet rake system.

4
equiangularly spaced rakes with five ports per rake transducer can either be absolute or differential (with
located at the centroids of equal areas. The original an accurate reference source). This response system
F/A-18 rake (fig. 3) followed this configuration and measures an accurate absolute pressure level. The
was clocked (rotated clockwise aft looking forward high-frequency response probe typically consists of a
(ALF)) approximately 9°. Clocking the rake is often miniature transducer mounted at the AIP. This response
required because of structure installation consider- system measures the time-dependent component of the
ations. The HARV inlet flight test program goal was to pressure but, normally, not an accurate absolute pres-
meet this arrangement. Because the original F/A-18A sure level.
rake was installed in the left inlet, the HARV rake, to The HARV research objectives required instrumenta-
be installed in the right inlet, was required to be tion to measure stabilized and dynamic maneuvers.
clocked counterclockwise ALF approximately 9°. This The typical instrumentation setup described in the pre-
configuration would make the installation equivalent vious paragraph would not be adequate to meet the
because of symmetry. demanding requirement of measuring inlet characteris-
tics during dynamic maneuvers without introducing a
HARV Inlet Research Objectives large amount of measurement uncertainty. The HARV
requires a system which would minimize the effects of
Another consideration necessary for the design of the two known drawbacks of the typical instrumentation
inlet rake was its intended use during the HARV inlet setup that affect the ability to measure an accurate pres-
research flight test. The primary objectives of the sure level during a dynamic maneuver: pneumatic lag
research were as follows: and thermal zero shift. Pneumatic lag describes the
1. Determine whether highly dynamic aircraft condition where the pressure signal at the AIP is
maneuvers result in a significant increase in inlet delayed in reference to time to the transducer at the end
pressure distortion levels compared to correspond- of the tubing and, therefore, affects low-frequency
ing steady-flight conditions. response accuracy. Thermal zero shift affects the abil-
ity of the low- and, especially, the high-frequency
2. Determine whether sources other than spatial time- response transducer to accurately measure the pressure
variant distortion lead to engine aerodynamic level at varying inlet temperature conditions. Thermal
instabilities during aircraft departures. zero shift describes the calibration shift of the zero
3. Assess predicted inlet distortion from computa- voltage condition experienced as a pressure-sensing
tional fluid dynamics (CFD) as compared to flight element of the transducer varies with temperature.
test measured levels. Thus, the requirement was to develop an instrumenta-
tion setup that would allow for accurate measurement
These objectives have the dual requirement for gath- of the pressure level and the time-dependent compo-
ering accurate inlet measurements during stabilized nent of the pressure during highly dynamic maneuvers.
and dynamic maneuvers, including aircraft departures. The requirements for pressure and temperature
The instrumentation setup for dynamic maneuvers ranges were determined by the flight conditions where
would require more attention than the setup for stabi- research testing would take place. The HARV research
lized maneuvers. occurred between an altitude from 15,000 to 40,000 ft
and below Mach 0.9 (fig. 4). The necessary instrumen-
Instrumentation tation pressure range was determined to be 2 to 16 psia.
The temperature was from 395 to 618 °R.
Inlet pressure distortion instrumentation typically has Other instrumentation considerations outlined by
3
the requirement for accurately measuring pressure lev- industry through the SAE include that “the frequency
els at high frequencies. This requirement is typically response characteristic of the probe and transducer
achieved by measuring the inlet characteristics with a combination should be determined with reference to
dual-probe configuration using low- and high- system accuracy requirements.” GEAE determined that
frequency response sensors. The low-frequency the highest frequency of interest for the F404-GE-400
response probe usually consists of pneumatic tubing engine was 105 Hz. NASA chose to increase the high-
routed through and beyond the rake where it is con- est frequency of interest to 250 Hz. Industry require-
nected to a highly accurate transducer. This pressure ments for the highest frequency of interest vary. NASA

5
required the higher range to allow the HARV inlet the freestream total pressure was 20.4 psia, and the hot
research database to be used by all interested industry day total temperature was 618 °R.
customers. The instrumentation accuracy requirement Requirements for the inlet rake design included addi-
needed to meet or exceed the original F/A-18A flight tional aerodynamic and structural considerations.
test. This testing called for the following system accu- Designing the rake legs and center hub with an aerody-
racy as a percent of reading (2σ): 3.2 percent at 2 psia, namic shape that minimized airflow disturbance and
1.3 percent at 5 psia, and 1 percent at 32 psia.1 with a blockage factor as small as possible was desir-
A final requirement stated for the instrumentation able. The HARV blockage goal was to be equal to or
was ease of maintainability. Transducers needed to be less than the previous F/A-18A inlet rake design,
accessible without engine removal. The ability to which had a flow blockage of less than 8 percent.1
replace the transducer without removing the engine Structural requirements included meeting the worse-
was also desirable. Additionally, the transducer or case pressure loads. These loads include an inlet ham-
probe configuration should provide the sensing element mershock overpressure of 20 psi maximum caused by
of the transducer with protection from foreign object an engine surge. The HARV structural load limits with
damage (FOD). the thrust vectoring vanes installed are 5.4 normal g
loads and 2.0 lateral g loads. Of particular concern was
Aerodynamic and Structural Requirements the requirement to meet dynamic structural require-
ments as outlined in DFRC document “Process Specifi-
cation 21-2, Environmental Testing Electronic and
Figure 4 shows the aerodynamic design flight enve-
Electromechanical Equipment.”* In particular, the rake
lope. This envelope coincides with the normal operat-
had to be designed stiff enough so that it did not exceed
ing envelope of the HARV aircraft and was chosen to
stress limits at its predominant structural frequencies
allow unrestricted flight with the inlet rake installed.
Inlet research test points were primarily focused at the *NASA Dryden internal document, “Process Specification 21-2,
low-speed portion of the envelope between Mach 0.3 Environmental Testing Electronic and Electromechanical Equip-
and Mach 0.4. The worse-case dynamic pressure con- ment.” Original released on May, 1968 with current updates until
dition was at a Mach 0.7 at sea level conditions where Apr. 1989.

50 x 103

40

30
Altitude,
ft
20
450 kts

10

0 .2 .4 .6 .8 1.0
Mach number 940086

Figure 4. HARV inlet rake design flight envelope.

6
within the airframe and engine operating range. NASA HARV Inlet Rake System and Flight Qualification
also wanted a number of structural materials consid- Testing sections.
ered for the rake including composite materials. Use of One significant advantage of the new rake design
composites could translate into a lighter, more aerody- includes its transportability to other aircraft which use
namic design along with structural tailoring of the rake F404-GE-400 engines, such as other F/A-18, X-29, and
arms as compared to traditional designs. X-31 aircraft. GEAE is currently under contract to
determine the feasibility of extending the current
Cost and Time Considerations design to supersonic flight conditions and to scale this
design to larger inlet diameters.
Two of the critical requirements that had to be met
for the HARV inlet rake design were low cost and min- HARV INLET RAKE SYSTEM
imal installation time. An evaluation of the rake used in
the original F/A-18 inlet compatibility program indi- This section describes the HARV Inlet Rake System.
cated costs in excess of $1.5 million and installation The design and development of the HARV inlet rake
time on the order of 1 year or more. Driving both these system are detailed in the Rake Description, Pressure
factors were the complexity of using eight cantilever Transducer Selection, and Rake Fabrication subsec-
rakes. Those designs had to be developed, tested, and tions. Placement of the rake into the HARV is
installed independently. Installation would require the described in the Rake Installation subsection. A sum-
aft portion of the inlet duct to be extensively rein- mary description of the rake system is described in the
forced. Thus, this evaluation quickly revealed that this Operating Principles subsection.
approach was not viable for the HARV project. A liter-
ature search of past rake designs did not provide a via- Rake Description
ble alternative approach.4,5
During an early design conception meeting, NASA The HARV rake is similar to a wagon wheel with the
and GEAE personnel conceived an alternative streamlined centerbody acting as the hub, the eight
approach in which the eight rake legs would be joined aerodynamic rake bodies being the spokes, and the
at the center of the inlet with a hub similar to that of a inlet duct being the rim. The load-bearing structure is a
wagon wheel. The rake would thus be one piece and welded Inconel 625* unit that joins the rake bodies and
would only have to consider shear loads (no moments) the central hub into a single piece that is supported by
at its attachment points. This alternative would greatly integral footpads and bolted to the aircraft inlet duct
simplify the structural installation requirements. Also, flange (fig. 5). Each of the eight rake legs contain five
it was envisioned that the rake would slip up the back ports located on the centroids of five equal areas of the
of the inlet ahead of the engine and attach to the main flow area. The ports are aligned within 2° of the antici-
bulkhead near this location. pated steady-flow streamlines. The innermost port is
Similar wagon wheel rake designs of the past differ the only one that had to be angled (5.5°) with respect to
from the NASA and GEAE approach. Previous designs the rake body. All others were already within 2° of the
typically used a modified engine bullet nose which flow angle.
acted as the hub. The NASA and GEAE concept would The central hub also contains an isolated metal
be self-supported with no physical contact with the damper ring potted in the polyurethane centerbody
engine. This design would minimize vibrational and (fig. 6). This configuration allows the damping material
force transfer from the engine. The NASA and GEAE to dissipate vibration energy more effectively than an
design would greatly simplify installation and aircraft all-metal body. The same polyurethane material also
modifications and significantly reduce aircraft forms the streamlining of both the centerbody and the
down time. GEAE has successfully designed, devel- trailing edges of the rake bodies. The rake bodies
oped and built one prototype and two flight-worthy (or spokes) of the structure are made by forming sheet
rakes for less than $500,000. One flight-worthy rake metal into the leading edge and sides of the airfoil
had an entire set of high-frequency response transduc-
ers included in the cost. The design details and specifi- *Inconel 625 is a registered trademark of Huntington Alloy Prod-
cations associated with the delivered rake along with ucts Division, International Nickel Company, Huntington Beach,
its flight qualification testing are described in the West Virginia.

7
8
Section B

Section A

D D

Metal damper ring

Engine bullet nose

C
Front view of rake Side view of rake
940096

(a) Front and side views.

Figure 5. Inlet distortion rake.
 Stagnation shield Lead-out tubes

Shrink tube
High-frequency K-seal
Epoxy filler
response transducer
Fuselage Engine
bulkhead front frame
Shim

Footpad
Low-frequency response
pneumatic tubing

Measurement port Mounting installation

Section A Section B

Lead-out tubes Metal rake body

Elastomer filler

Bolt holes

Footpad Rake body cross-section

Section C Section D 940173

(b) Sections A through D.

Figure 5. Concluded.

9
 Figure 6. Floating damper ring.

shape. The sheet metal is left open at the trailing edge. allows the sensors to read true pressure levels at yaw
This configuration allows the installation of the sensor angles from ±25° and pitch angles of 15° and –25°.
and leadout tubes. Positive angle is toward the engine centerline. Maxi-
Flexane 94* is the polyurethane elastomer used as mum flow blockage is less than 8 percent of flowpath at
the potting and flow contour material. This 2 part mix, the maximum thickness of the rake. This maximum
room temperature curing, pourable urethane bonds thickness occurs 1.5 in. behind the AIP.
well to metal, meets the 618 °R maximum temperature A composite material was not chosen for the rake
requirement for this application, and has good vibra- because of increased design costs. Instead, the selec-
tion-damping properties. The rake assembly weighs tion of the Inconel 625 frame with the bonded elas-
approximately 15 lb. tomer was an adequate compromise. The elastomer
The measurement plane of the sensors is located 4 in. allowed the overall weight to be reduced and the rake
in front of the bullet nose of the engine with the rakes struts to be aerodynamically shaped more easily than
oriented 45° from each other starting at 9° counter- an all-metal body. In addition, the elastomer acted as
clockwise off top center (aft looking forward (ALF)) an excellent damping material.
for the right-hand inlet. This orientation satisfies the
mounting accessibility and correspondence to previous Pressure Transducer Selection and Installation
test vehicle requirements. The rake sensors are shielded
total-pressure-measuring sensors consisting of an The NASA and GEAE team decided to use the same
impact-mounted high-frequency-response pressure dual-probe configuration with low- and high-frequency
transducer and a 1/16-in.-diameter, steady-state pres- response probes which had been used during the previ-
sure tube. The stagnation shield configuration was ous F/A-18A flight test for the HARV inlet rake. The
tested to show its ability to measure the true total input low-frequency response probe would use 1/16-in. out-
pressure at varying flow angles. This configuration side diameter (OD) tubing routed through the rake into
the engine bay. A differential transducer with a refer-
*
Flexane 94 is a registered trademark of Devcon Corporation, Dan- ence pressure was selected. The DFRC has had a con-
vers, Massachusetts. siderable amount of experience using ESP-3205BSL

10
differential transducers (Pressure Systems Incorporated increased maintainability requirements. The engine
(PSI), Hampton, Virginia). This transducer unit (fig. 7) bay pressure was finally chosen because of its steady
was thermally stabilized to increase accuracy by mini- rate of change with altitude and its ease of installation
mizing thermal zero drift. This stabilization was on the aircraft. This pressure is also relatively unaf-
accomplished by wrapping the transducer unit in a fected by angle of attack.
temperature-controlled thermal blanket. Another fea- The reference pressure range of the Sonix PS1019
ture of the transducer, used to increase accuracy, was transducer (0.4 to 19 psia) met the requirement of 2 to
its ability to perform in-flight calibrations. This in- 16 psia. The range of the differential pressure trans-
flight calibration allows for any calibration bias error to ducer unit was sized to provide the smallest range
be removed during postflight data processing. The about the reference pressure, while allowing for the
in-flight calibration is accomplished by applying the expected differential between the engine bay pressure
reference pressure to both sides of the differential and the highly distorted total pressures at the inlet rake.
transducer. A range of ±5 psid was selected. This minimization of
The right-hand ALF engine bay pressure was the differential pressure range allowed for increased
selected as the reference pressure for the resolution. Both the differential and reference pressure
ESP-3205BSL transducers. A Sonix PS1019 trans- transducers were readily accessible. The differential
ducer (Pressure System Incorporated, Hampton, Vir- transducer was located in the missile bay under the
ginia) was selected to measure the reference pressure. right wing ALF at the fuselage. The reference pressure
This pressure was chosen after careful consideration of transducer was located in the right-hand ALF landing
a number of possible reference sources. Other possibil- gear wheel well. The differential and reference pres-
ities included inlet duct throat wall static pressure, an sure transducer low-frequency response setups met the
inlet rake total-pressure port, and a pressurized tank. system accuracy requirements during stabilized aircraft
The inlet duct throat wall static pressure and inlet rake maneuvers. This setup should not be depended upon
total-pressure ports were eliminated because of their during highly dynamic aircraft maneuvers. Instead, the
unsteady pressure levels caused by inlet airflow varia- high-frequency response instrumentation setup should
tion. The pressurized tank was eliminated because of be used.

Figure 7. Installed HARV low-response differential transducer setup.

11
 Each high-frequency response probe used an of high-frequency pressure levels during highly
XCS-27L-093-20A temperature-compensated pres- dynamic maneuvers and meets the system accuracy
sure transducer (Kulite Semiconductor Products, Inc., requirements.
Leona, New Jersey). This transducer has a 0.093-in. The temperature-compensated pressure transducers
diameter by 0.50 in. length with an absolute pressure were installed in the inlet rake 0.125-in. OD carrier
range from 0 to 20 psia. The frequency response of this tubes. These tubes have a 0.099-in. ID counterbore to
transducer met the 250 Hz requirement. Forty of these receive the transducers. Installation was accomplished
transducers were mounted at the rake measurement by feeding the transducer electrical leads through the
ports in close proximity to the low-frequency response 0.125-in. tube from the sensor end, coating the back of
pneumatic tubing (fig. 8). Each high-frequency the transducer with room temperature vulcanizing
response transducer had a protective FOD screen over (RTV) silicone adhesive, and inserting the transducer
the sensing element. This transducer was selected into the counterbore. Next, the transducer was covered
because of its ability to minimize thermal zero drift with heat-shrinkable tubing, and the silicone adhesive
through passive temperature compensation. To further was allowed to cure. Then, the shrink tube was
increase the accuracy of the transducer measurement, a trimmed flush with the transducer face. The lead exit
series of pressure calibrations were performed over the was also sealed with the silicone adhesive and shrink-
entire required pressure and temperature range, up to able tubing. This arrangement provided a secure
20 psia and at 395, 425, 460, 535, and 610 °R. Along mounting for the transducers but still allowed removal
with the measured engine inlet temperature, these cali- and replacement while the rake was still in the aircraft
brations would allow for any remaining zero thermal in case of transducer failure.
drift to be removed during postflight data processing.
The differential transducer low-frequency response Rake Fabrication
measurements will be used to verify the pressure levels
of the temperature-compensated pressure transducer at Fabrication of the rake unit began with forming the
stabilized conditions. This high-frequency response sheet metal rake body airfoils, drilling them for the
transducer setup allows for the accurate measurement sensor tubes, and welding them to a machined hub. The

Figure 8. HARV inlet rake sensor setup with high-response, temperature-compensated pressure transducer
installed.

12
footpads were then machined, formed to the duct Rake Installation
radius, and welded to the rake bodies. After stress
relieving, the welds were inspected using the florescent Installation of the rake on the aircraft was accom-
penetrant inspection technique. (This nondestructive plished after the right-hand engine (ALF) had been
inspection technique detects cracks by applying a pen- removed. The minimal airframe modifications that
etrating fluorescent solvent to the material.) Sensor were required consisted of 1/4-in. bolt holes being
tubes and stagnation shields were then brazed in place drilled in the inlet duct flange aft of the rear bulkhead.
with a gold and nickel braze. The bend radii in the These 16 holes were located by trial positioning the
rake in the inlet duct, confirming that none of the bolt
leadout tubes were kept as large as possible to facilitate
holes would interfere with any bulkhead bracing, and
installation of the transducer leads later in the assembly
transfer-punching the hole locations through the rake-
(fig. 9). A damper ring was then installed around the
mounting footpads. Then, the rake assembly was
hub and rake struts. Welds and brazes were reinspected
removed, and the holes were drilled in the flange using
after the metal fabrication portion of the assembly con- a hand-held drill. Backup washers were then placed at
cluded. Strain gages were applied to the hub and rake each hole on the outside of the duct and expoxied in
bodies in the areas that would later be covered with place to provide a solid and flat surface for seating the
polyurethane elastomer (fig. 10). The rake bodies were nuts of the rake-mounting bolts (fig. 5). This seating
then filled and the trailing edges of the rakes were was done by inserting a nylon mandrel through the
formed with the elastomer. The center hub form was washer and rake-mounting holes in the duct to position
then cast in place completing the airfoil blending of the the washers on an aluminum-filled epoxy base that
rakes and centerbody. Figure 11 shows the completed filled the gap under the washer. These mandrels were
rake assembly. removed after the epoxy had cured overnight.

EC 92 5192-3
Figure 9. Rake leg instrumentation tube leadouts.

13
 S 3405-2
Figure 10. Stain-gaged hub of distortion rake.

S 825-3
Figure 11. Completed prototype distortion rake with polyurethane elastomer rake leg fairings and centerbody.

14
 Testing showed that this configuration could with- airframe, installation of the rake assembly, modifica-
stand over 200 in.-lb torque applied to the mounting tion to the K-seal, and installation of the seal ready for
bolts without deforming the washer, epoxy, or flange. lead routing was accomplished in less than 3 days.
The originally planned rework to the exit lip of the inlet
duct was eliminated to retain as much of the original Operating Principles
integrity of the flange as possible. To accomplish this
goal, the leadout tubes of the rakes were reformed to Single-piece construction of the HARV rake assem-
exit the duct just behind the lip, and a support shim was bly resulted in a redundant structure that allows each
made from polyurethane elastomer to support the tubes spoke of the wheel configuration to share the load of
under the K-seal. This seal around the inlet duct exit the structure through the hub and transfer it to the air-
provides the transition from the inlet duct to the engine frame duct as shear loads to the bolts in the duct flange.
front-frame lip. The K-seal was contoured to allow the These rakes are fastened at the footpad on the duct
rake tubes to pass freely under it (fig. 12). wall, allowing the duct flange to support the rake with-
The rake was placed in the duct. An even fit for each out inducing any bending load in the sheet metal wall
rake strut was achieved by placing shims between the of the duct.
rake metal footpad and the inlet duct wall. The rake Vibratory energy within the rake struts is damped as
assembly was then installed in its final position. Room the flex and twist motion of the open-backed rakes put
temperature vulcanizing silicone adhesive was placed the polyurethane elastomer in shear. The elastomer is
on each side of the fiberglass and polyurethane elas- bonded to the sheet metal. Energy at the hub from the
tomer footpads. The metal footpads were then bolted in rake struts is dissipated using the action of the damping
place, and the silicone adhesive was cured. This config- ring imbedded within the elastomer.
uration gave the assembly a firm, elastomer-damped
mounting. FLIGHT QUALIFICATION TESTING
The K-seal was then bolted in place on the aft duct
flange. The electrical leads for the pressure transducers Flight qualification testing consisted of three phases:
and the steady-state pressure tubes were routed to their laboratory, ground testing, and flight testing. The labo-
respective connector locations. The rework to the ratory phase determines the baseline structural and

Figure 12. Instrumentation tubes passing under K-seal at exit lip of HARV inlet.

15
vibrational characteristics of the rake. This phase con- addition, accelerometers were mounted on the inlet test
sists of rake modeling using NASA Structural Analysis fixture and on the vibration table. This instrumentation
(NASTRAN) along with Zonic ping (Zonic Corpora- would verify that there was no displacement amplifica-
tion, Milford, Ohio) and vibrational shake table tion caused by resonance of the test fixture. This instru-
(Unholtz-Dickie Corporation, Wallingford, Connecti- mentation was also used to verify the predicted
cut) testing on a prototype rake. Results of this phase damping characteristics of the rake structure.
are compared to the environmental conditions that the Strain gage excitation and signal conditioning were
rake will experience in operation. Then, baseline accomplished by a GEAE designed and fabricated con-
results from the laboratory tests are used for compari- stant current excitation, alternating current (ac) cou-
son with the results from the installed ground testing pled amplification system. Data were recorded on an
and flight testing phases. Each phase is described in the X-Y plotter to show overall strain levels as a function
following subsections. of frequency. The instrumented inlet test fixture and
rake assembly was mounted on a 17,500 lb capacity
Laboratory Testing vibration table.
The ping test uses an instrumented hammer for
Computer modeling of the rake structure using the applying a force to the rake structure and measures the
NASTRAN program was used to identify the possible resultant motion of the rake. This method identified
vibration modes and the approximate frequencies of four primary modes at which the rake might be driven.
those modes. The computer model identified 15 poten- These modes were axial, torsion, strut bending, and
tial vibrational modes between 0 to 7000 Hz. Most of second axial.
these modes could not be driven to significant levels Vibration table testing consisted of three procedures:
during subsequent laboratory, ground, or flight testing. axial-sinusoidal, transverse-sinusoidal, and transverse-
The laboratory testing used a prototype rake instru- random excitation. The axial- and transverse-sinusoidal
mented with strain gages. This rake was installed in an vibration test procedures required several steps. First,
inlet test fixture manufactured from a modified F/A-18 the rake was excited axially with respect to the engine
aircraft inlet. This test phase consisted of ping tests and centerline. Next, the rake was excited transversely
vibration table testing. Objectives of vibration testing using sinusoidal-driving forces over the frequency
with the instrumented prototype were to: from 10 to 2000 Hz at 1-g excitation to locate the reso-
1. Measure the vibration characteristics of the rake nant frequencies of the rake. Then, the axial and trans-
with axial and transverse excitation. verse stress distributions were obtained by dwelling at
the resonant points and increasing the input accelera-
2. Determine the rake stress distribution at an estab- tion levels until the established maximum stress level
lished maximum rake stress level. was reached on one of the strain gages. Subsequently,
3. Excite the rake randomly at the flight qualification the rake was exposed to transverse-random excitation
excitation level for a specified period without dam- (nonsinusoidal) over the frequency range of 15 to
aging the rake. 2000 Hz at 18.4-g excitation (flight qualification level).
Finally, the rake was vibrated at resonance at an excita-
4. Excite the rake sinusoidally at its resonant fre- tion level necessary to verify the rake endurance limit.
quency at a specified excitation level to empirically Rake maximum stress levels were established by tak-
define the rake endurance limit. ing a number of factors into consideration. These fac-
tors included the material properties of the rake
5. Select the critical strain gage locations for ground
structure, placement and number of strain gages
test monitoring based on the strain-gage-derived
installed, stress concentrations, and safety margin. The
stress distribution.
stress limits of 30,000 and 50,000 lb/in2 peak-to-peak
Twenty-eight dynamic strain gages were applied to (PSIPP), dependent on strain gage location, were used.
two (of eight) rake struts. That is, 14 each on 2 struts The axial-input frequency resonant sweep test
spaced 90° with respect to each other and circum- showed an axial mode at 147 Hz with 1-g excitation.
ferentially oriented in the inlet so that one of the The input acceleration level was then increased until
instrumented struts would be exposed to maxi- the maximum stress level of 30,000 PSIPP was
mum excitation during the transverse vibration test. In reached. The maximum strain gage outputs were

16
located on the leading edge and trailing edge of the shown that axial excitation has never been a problem
rake foil at the hub. A decrease in the axial mode fre- for an inlet rake installed directly in front of an engine.
quency to 129 Hz at 3.8-g excitation was observed. The The transverse-sinusoidal input sweep showed a res-
maximum response of 30,000 PSIPP at 3.8-g axial onant frequency of 123 Hz (a combination of rake axial
input meets the damping and structural design require- mode and fixture resonance) and a strut bending mode
ments of the rake. The lower resonance frequency of 240 Hz with the stress distribution remaining con-
(129 Hz) at the higher vibration level is a result of the stant with increasing input levels. A transverse excita-
soft mount interaction of the fixture and the tion of 7.8 g was reached at the maximum stress level
elastomer-damping configuration. The soft mount of of 30,000 PSIPP.
the test fixture is not representative of a full aircraft The transverse-random test provided a nonsinusoidal
inlet. This soft mount was improved by adding bags of input to test the hardware for interaction and structural
lead shot to the fixture. The frequency shift of the rake flaws caused by 18.4-g random input (flight qualifica-
and the lower 1/rev crossover on the fan Campbell dia- tion level) in the 15 to 2000 Hz frequency band. The
gram (fig. 13) at the higher excitation level (147 Hz) is test had to be performed for a minimum of 20 min to
not critical to the highly damped rake structure. meet the flight qualification requirements. This test was
Figure 13 shows the various vibration mode frequen- performed for 26 min. Stress level and distribution data
cies of the rake (axial, torsion, strut bending, and sec- from the nonsinusoidal excitation are not related to the
ond axial) and lines representing multiples of fan rotor transverse sinusoidal sweep data. During the random
speed frequencies (1/rev, 2/rev) are superimposed tests, strain gages located on the leading and trailing
across the entire fan rotor speed range. Any intersec- edges of the rake foil near the footpad provided the
tion allows for fan rotor speeds to be identified where highest relative output signals. Spectral data from the
rake resonance frequencies are likely to occur. For random excitation again indicated resonant frequencies
example, the rake axial mode resonance (147 Hz) is of 129 Hz (rake axial with fixture resonance) and
likely to occur at a fan rotor speed of approximately 240 Hz (strut bending). Based on the data from the
8500 rpm. This low fan rotor speed means less energy vibration table tests, 12 strain gages at the leading and
which can be imparted onto the rake. Experience has trailing edges of the airfoil at both the hub and the

Idle 4/rev Takeoff Fan rotor speeds where rake

800
resonance frequencies are
likely to occur
700 3/rev
625 Hz

600 Second axial

500
2/rev
Frequency,
Hz 400

300 262 Hz

200 1/rev Strut bending

162 Hz
147
8500 Torsion
100 123 Hz Axial
Axial with fixture resonance
0
4 5 6 7 8 9 10 11 12 13 14
rpm x 1000
940095

Figure 13. Campbell diagram showing rake vibration responses.

17
outer ends of the rake strut were selected to be moni- 50,000 PSIPP were satisfactory for use during ground
tored during the ground and flight test program. This and flight testing. Use of these limits depended on the
test resulted in no structural damage to the rake. location of the strain gage.
Empirical determination of the structural endurance
limit was attempted by exciting the rake at its resonant Ground Testing
frequency at the required 10-g, transverse-sinusoidal
level. The 10-g, transverse excitation showed a stress A ping test was performed on the inlet rake installed
level of 41,000 PSIPP. The 10-g input was to be held in the F/A-18 HARV while in the hangar. This test was
until the rake structure failed or reached 107 vibration used to verify the results gathered during laboratory
cycles. The duration would define the endurance limit. testing. Frequencies showed good agreement with the
The resonant frequency changed from 122 to 117 Hz past laboratory ping test results.
during the test. This change probably resulted from the The installed rake was ground tested on the F/A-18
heating effect on modulus of the elastomer potting HARV with the aircraft tied down. Twelve strain gages
material. The 10-g input level was extremely severe on were monitored while the right-hand engine was oper-
the inlet test fixture. The test was shut down after 2 hr ated through its full range with a slow acceleration
of dwell time at 10 g to avoid catastrophic damage to from idle power to full maximum afterburning and a
slow deceleration to idle power. This procedure
the test fixture. Empirical determination of the rake
allowed predominant frequencies to be identified over
endurance limit could not be achieved because the
the entire fan rotor speed range. Stress levels and spec-
high-damping characteristic of the rake made it signifi-
tral data were taken during the test.
cantly more durable than the inlet test fixture. The inlet
The spectra showed predominant frequencies of 153
test fixture could not tolerate excitation levels neces-
and 162 Hz corresponding to axial and torsional fre-
sary to damage the rake. The rake system showed no
quencies and a broad band about 250 Hz at a lower
signs of structural damage for the duration of the test
level corresponding to strut bending modes. These pre-
and was confirmed with a posttest X-ray and visual
dominant frequencies showed consistent agreement
inspection of the rake hardware.
with the laboratory baseline frequencies. Maximum
The results of the vibration shake table test have been
stress levels during the test were approximately 10 per-
compared to the NASTRAN computer model predic-
cent of limit. Again, the strain gages showing the great-
tions and to the ping test results. These dynamic analy-
est stresses were located at the leading and trailing
sis results are listed next. edges of the airfoil, at the hub, and at the outer end of
the rake strut.
NASTRAN Ping Vibration
Mode model, Hz test, Hz table, Hz Flight Testing
Axial 140 155 147
The four most active strain gages were located at the
Torsion 147 162 N/A leading and trailing edges of the airfoil, at the hub, and
Axial with N/A N/A 123–129 at the outer end of the rake strut, and were selected for
fixture monitoring during rake qualification flights. Two were
resonance monitored by telemetry in real time, and two were
recorded on the aircraft for postflight analysis.
Strut bending 250 236–262 240
Qualification runs for the rake consisted of flight
Second axial 594 625 N/A points to give maximum unsteady loads (an angle of
attack of 60° at an altitude of 20,000 ft), temperature
The rake vibration testing confirmed the analytical and pressure (Mach 0.7 on the deck), and combination
design calculations. The laboratory results from the of temperature, pressure, and unsteady loads (Mach 0.9
ping and vibration table tests formed the baseline pre- at an altitude of 18,000 ft) within the HARV flight
dominant frequencies for comparison to ground and envelope. The latter two points were at the limits of the
flight testing. A frequency shift limit of ±20 percent HARV flight envelope. The first point was flown to a
was determined to be reasonable. Shifts greater than high-angle-of-attack condition. The latter two points
this limit indicate either a damaged or failed structural were obtained at the maximum maneuver-loading
component. The stress limits of 30,000 and limit.

18
 The spectra of the rake vibrations remained consis- design flight envelope to be expanded to supersonic
tent with the ground and laboratory tests. The stress conditions and allow scaling to other aircraft inlets.
levels observed were 26 percent of limits during high- The cost and time saving solution to the inlet rake
angle-of-attack flight and less than 10 percent of limits design will allow the F/A-18 HARV to conduct
during maximum maneuver-loading flight. The maxi- important inlet research at flight conditions never
mum stress limits observed were 30 percent of limits before explored.
during aircraft takeoff. Based on the laboratory,
ground, and flight test results, the rake is now fully ACKNOWLEDGMENT
cleared for conducting flight research within the entire
HARV flight envelope with no restrictions. The DFRC, LeRC, and GEAE team members wish to
recognize posthumously the creative energy brought by
CONCLUDING REMARKS co-author Leon Lechtenberg to this project. It is unfor-
tunate that his tragic and untimely death did not allow
An improved cost and installation-time saving inlet him to see the fruit of his labor that resulted in success-
distortion pressure rake was successfully designed, ful ground and flight clearance of this novel rake
built, and validated for flight testing on the F/A-18 design.
HARV research aircraft. The innovative design consists
of a one-piece, wagon wheel approach that resulted in REFERENCES
ease of installation with minimal aircraft modifications.
1Amin, N.F. and D.J. Hollweger, “F/A-18A Inlet/Engine
Design advantages include lightweight, high strength,
low structural resonance, low flow blockage, and easy Compatibility Flight Test Results,” AIAA-81-1393,
transducer removal and replacement. A prototype of July, 1981.
the new rake was environmentally tested in a labora- 2Inlet Total-Pressure-Distortion Considerations for Gas-
tory where it passed all vibration structural require- Turbine Engines, Society of Automotive Engineers
ments. Ground test verified the expected frequency Aerospace Information Report, AIR 1419, May 1983.
response predicted from laboratory test. Flight qualifi- 3Gas Turbine Engine Inlet Flow Distortion Guidelines,
cation was completed and the rake is now cleared for
Society of Automotive Engineers Aerospace Recom-mended
flight testing on the HARV aircraft. All stress levels Practice, ARP 1420, Mar. 1978.
observed during ground and flight qualification were
4Holzman, Jon K. and Gordon A. Payne, Design and Flight
less then 30 percent of limits.
The simple design approach and creative use of light- Testing of a Nullable Compressor Face Rake, NASA TN
weight materials such as elastomers resulted in a supe- D-7162, Jan. 1973.
rior rake design at a significantly lower cost than previ- 5Farr, A.P. and G.A. Schumacher, “System For Evaluation
ous design approaches. The rake has promise of ease of of F-15 Inlet Dynamic Distortion,” Symposium on
transportability to other F404-GE-400 equipped air- Instrumentation for Airbreathing Propulsion, MCAIR
craft. Further design efforts are expected to allow the 72-043, Sept. 1972.

19
 Form Approved
REPORT DOCUMENTATION PAGE OMB No. 0704-0188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources,
gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this
collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson
Davis Highway, Suite 1204, Arlington, VA 22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project (0704-0188), Washington, DC 20503.
1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED
October 1995 Technical Memorandum
4. TITLE AND SUBTITLE 5. FUNDING NUMBERS

Design and Development of an F/A-18 Inlet Distortion Rake: A Cost and

Time Saving Solution
6. AUTHOR(S)
WU 505-68
Andrew J. Yuhas, Ronald J. Ray, Richard R. Burley, William G. Steenken,
Leon Lechtenberg, and Don Thornton
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION
REPORT NUMBER
NASA Dryden Flight Research Center
P.O. Box 273 H-2078
Edwards, California 93523-0273

9. SPONSORING/MONOTORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING

AGENCY REPORT NUMBER

National Aeronautics and Space Administration

Washington, DC 20546-0001 NASA TM-4722

11. SUPPLEMENTARY NOTES

Presented as AIAA 94-2132 at the 7th Biennial AIAA Flight Test Conference, Colorado Springs, Colorado, June 20–23, 1994. A.J. Yuhas, PRC Inc.,
Edwards, California; R.J. Ray, Dryden Flight Research Center, Edwards, California; R.R. Burley, NASA Lewis Research Center, Cleveland, Ohio;
W.G. Steenken, L. Lechtenberg, and D. Thornton, General Electric Aircraft Engines, Evendale, Ohio.

12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE

Unclassified—Unlimited
Subject Category 02

13. ABSTRACT (Maximum 200 words)

An innovative inlet total-pressure-distortion measurement rake has been designed and developed for the
F/A-18 A/B/C/D aircraft inlet. The design was conceived by NASA and General Electric Aircraft Engines
personnel. This rake has been flight qualified and flown in the F/A-18 High Alpha Research Vehicle at NASA
Dryden Flight Research Center, Edwards, California. The eight-legged, one-piece, wagon wheel design of the
rake was developed at a reduced cost and offered reduced installation time compared to traditional designs.
The rake features 40 dual-measurement ports for low- and high-frequency pressure measurements with the
high-frequency transducer mounted at the port. This high-frequency transducer offers direct absolute pressure
measurements from low to high frequencies of interest, thereby allowing the rake to be used during highly
dynamic aircraft maneuvers. Outstanding structural characteristics are inherent to the design through its
construction and use of lightweight materials.

14. SUBJECT TERMS 15. NUMBER OF PAGES

F-18 aircraft; Flow measurement; In-flight monitoring; Inlet flow; Pressure 21

measurement; Pressure sensors 16. PRICE CODE
A03
17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. LIMITATION OF ABSTRACT
OF REPORT OF THIS PAGE OF ABSTRACT
Unclassified Unclassified Unclassified Unlimited
NSN 7540-01-280-5500 Available from the NASA Center for AeroSpace Information, 800 Elkridge Landing Road, Standard Form 298 (Rev. 2-89)
Prescribed by ANSI Std. Z39-18
Linthicum Heights, MD 21090; (301)621-0390 298-102