NASA Technical Memorandum 100025

Flight Test Research at

NASA Ames Research Center:
A Test Pilot's Perspective
G. Warren Hall, Ames Research Center, Moffett Field, California

September 1987

National Aeronautics and

Space Administration
Ames Research Center
Moffett Field, California 94035
 FLIGHT RESEARCH AT NASA AMES: A TEST PILOT'S PERSPECTIVE

C. Warren Hall'
NASA Ames Research Center
Moffett Field, California

Abstract sustain a maximum sink rate of 12 ft/sec at the

design landing gross weight of 50,000 lb.
In 1976 NASA elected to assign responsibility
for each of the various flight regimes to individ- When the QSRA was delivered to NASA Ames
ual Research Centers. NASA Ames Research Center at (August, 1978), it was apparent that the USB
Moffett Field, California was designated lead cen- propulsive lift generated for the full-flap case
ter for vertical and short takeoff and landing, was below what had been predicted. In addition,
V/STOL research. This paper will discuss from the the pilots also commented about an asymmetry in the
test pilot's perspective the three most recent roll axis and what appeared to be an unsteady flow
flight research airplanes being flown at the phenomenon; substantial lateral trim was required
Center: the Quiet Short Haul Research Aircraft, and random vertical accelerations were experi-
the XV-15 Tilt Rotor Research Aircraft, and the enced. Past experience with the QSRA model in the
Rotor Systems Research Aircraft. 40- by 80-Foot Wind Tunnel at NASA Ames suggested
that the problem was the less-than-optimum place-
ment of the vortex generators. Tufting t n e wing
The Quiet Short Haul Research Aircraft behind the engines was impractical, therefore a
careful study of the soot patterns on the uplw
The Quiet Short Haul Research Aircraft (QSRA), surface of the wings was done and it provided the
Fig. 1, was designed to further research the con- primary clues for the proper placement of the vor-
cept of upper surface blowing (USB) as a means of tex generators. It is interesting that the differ-
developing a quiet short takeoff and landing (STOL) ence in approach speed between having no vortex
airplane using powered lift. Four high-bypass generators, to having one with a properly designed
turbo jet engines are mounted above the wing. For vortex generator pattern was 15 knots. Another
high lift, the flow is turned downward over smooth factor found to have a significant effect on the
curved flaps located at the trailing edge of the USB lift generated was the character of the flow
wing and directly behind each engine. For approach between the two engines which proved to be critical
and landing, these flaps are deflected 50"; the to good flow-turning performance.
resulting downward flow converts part of the engine
thrust directly to a lift force. An important At low altitude dith maximum USB flap deflec-
feature of the QSRA is to provide propulsive lift tion (66"), maximum thrust, and full aft elevator,
at a low noise level. The above-the-wing location the QSRA has a nose-down pitching moment at or near
of the engines provides a high degree of shielding 8" angle of attack. This anomaly was attributed to
of the engine exhaust noise, and noise suppression an elevator-stall phenomenon. The pitching moment
material is installed in the fan inlets, around the exceeds the capability of the availab e elevator
core engines, and in the exhaust ducts. The causing the nose-down attitude change. Recovery is
resulting noise footprint is estimated to be 1/7 achieved by either reducing thrust, USR flap
that of an equivalent jet transport. deflection, or increasing airspeed. Resetting the
elevator incidence angle 3" more negative provided
The QSRA was designed solely as a low-speed some increase in the maximum pitch controi avail-
research aircraft. The fuselage, empennage, and able but did not totally eliminate the elevator-
landing gear are essentially that of a de Havilland stall phenomenon.
C-8A Buffalo. Numerous structural modifications
were made by the Boeing Aircraft Company. The Almost all pilots who check out in the QSRA at
major modification consisted of reinforcing the one time or another experience a "wheel barrowing"
fuselage and mounting the newly designed wing con- tendency during a touch-and-go landing. Once on
taining the four engines, the USB flaps, two double the ground, if power is added before the USB flaps
slotted flaps, one outboard of each USB flap and are retracted, the lift generated is sufficient to
drooped/blown ailerons. While the entire focus of raise the main gear off the runway while the nose
the QSRA research was in the low-speed region, the gear remains thereon. During the performance
wing itself was representative of a wing capable of testing, it was shown that at a normal approach
an efficient cruise at Mach 0.74. The maximum speed of 65 knots, the QSRA could barely maintain
demonstrated takeoff gross weight is 57,000 lb. level flight with 50" USB flap and full thrust;
The airplane has a relatively high wing loading, however, at 30" of USB flap, a climb angle of
95 lb/ft2, a thrust-to-weight ratio of 0.5, and can nearly +6" could be maintained. The standard
go-around procedure became one of retracting the
-- __ USB flaps to 30" and then advancing the throttles
'Research Pilot. AIAA membet . to maximum thrust. Climb performance was greatly
increased and the "wheel barrowing" tendency
This paper is declared a work of the U.S.Government and
therefore is in the public domain.
considerah?y reduced.

1
 Flight testing any powered lift airplane is Normally, a conventional aircraft flies a
complicated by the fact that lift performance is carrier approach at constant angle of attack. As
heavily dependent on engine thrust (which varies indicated previously, attempts at flying constant
with altitude), temperature, and thrust coefficient angle of attack approaches in the QSRA resulted in
(which varies with velocity). It is always neces- excessive airspeed excursions and only marginal
sary to collect and process large quantities of flight-path control. Previous experience had also
data to construct meaningful performance plots. shown that flying a flat approach angle resulted in
Figure 2 (from item 3 in Bibliography) shows flight a tendency to float because of a strongly positive
path versus velocity for the QSRA in the landing ground effect. It was decided to fly the carrier
configuration at 48,000 lb with the USB flaps at approach as steeply as possible without exceeding
50". A normal approach is at 65 knots with a the QSRA's landing gear limit sink rate of
flight-path angle of -6 to -7.5". For a fan speed 12 ft/sec. A 4.5" aerodynamic flight-path angle,
of 70% rpm (approximately one half the thrust flown at 65-70 knots, with a pitch attitude between
available), the pitch attitude is just above a + l o and +3" allowed the airplane to fly a
level attitude (between +3" and +4") with an angle comfortable approach, cross the ramp with adequate
of attack near 10". Note that lines of constant safety margin and touch down without a flare, with
pitch attitude are nearly vertical at this point a nominal sink rate of just under 9 ft/sec.
and that lines of constant angle of attack have a
significant slope. This illustrates the need to The basic longitudinal rate command stability
fly the QSRA--and in general most powered lift augmentation system (SAS) was required for all
airplanes--at constant pitch attitude rather than approaches. The more important control require-
constant angle of attack as is more common for ment, however, was for direct lift control (DLC)
conventional airplanes. provided by the highly effective spoilers on the
wings. Spoilers were deployed to a nominal setting
Pilot experience indicated that any distrac- and operated both up and down from that position.
tion during the approach could result in signifi- The increased drag was offset by a higher power
cant deviations from the desired approach speed. setting which, incidently, placed the engines in a
An automatic speed-stabilization system was devel- more responsive power range. Without DLC, flight-
oped that "slaved" the USB flaps to a digital air- path control was severely degraded and occasionally
speed system. The system had an authority of 210" resulted in a longitudinal pilot-induced oscilla-
whenever the USB flaps were extended beyond 30". tion near touchdown. It was a relatively easy
The system worked well and reduced pilot workload piloting task to fly the desired glide path until
during the approach, but some problems were experi- about 100 ft, at which point the QSRA encountered
enced in ground effect. the ship's aerodynamic wake or "burble." The ini-
tial tendency was to go slightly high followed by a
Another piloting problem was discovered during somewhat abrupt settling as the airplane crossed
engine-out approaches. There was strong tendency the carrier's round down. After a few approaches,
toward a lateral (roll) pilot-induced oscilla- it was possible to anticipate the settling and
tion. It was determined that the r o l l trim point increase power at the appropriate time to continue
was very close to the point where spoiler activa- on the glide slope. The Navy was especially
tion (30" wheel deflection) occurred. The ultimate impressed with the wave-off capability of the
solution was to program the spoilers to operate as QSRA. The more difficult task was runway align-
a continuous and linear input along with the ailer- ment. Since the steep approach angles exceeded the
ons. This modification greatly improved the preci- capability of the ship's standard Fresnel lens, it
sion of the roll control for engine-out was necessary to use a portable Fresnel lens on the
approaches. Further, the ability to deploy the ship. This lens was located 250 ft aft of the
double slotted flaps asymmetrically allowed the normal lens and farther to the left of the carrier
pilot to almost completely trim the engine-out centerline than it had been during the field car-
rolling moment and fly with a nearly neutral wheel rier landing practice. The result was a tendency
position. 1
to land to the left of the prescribed centerline.
In the spring of 1979, a joint NAVY/NASA Takeoffs were performed from both the axial
flight program was undertaken to investigate the and angle decks. The USB flaps were set at 10" and
application of propulsive lift technology to the the double slotted flaps at 59". Power was set at
aircraft-carrier environment. 80% fan speed prior to brake release and maximum
power applied immediately after brake release.
The primary objectives of the program were to
Full aft column was applied after the first indica-
determine the best techniques for ooerating STOL
tion of airspeed. The airplane rotated comfortably
aircraft aboard aircraft carriers and to obtain
at 60 knots and lifted off at 70 knots. Takeoffs
design data for development of operational criteria were smooth and precise.
for future Navy use. Of particular interest was to
develop the best STOL landing technique; determine In a four-day period of sea trials, 25 low
the effects of ship's wake, ground effect and approaches, 37 touch-and-go landings, and 16 full-
motion on the approach; and general shipboard oper- stop landings were made on the USS Kitty Hawk
ation and handling of the airplane. (CV-631, Fig. 3 . It was concluded that USB

2
propulsive lift technology presented no unusual two modified T-53 engines located in the wirq tip
problems in the aircraft-carrier environment. nacelles. Either engine is capable of driving both
proprotors through a cross shaft located in the
It is commonly believed that high-performance wing which allows for an engine failure and still
propulsive lift airplanes are difficult to fly. retains power to both proprotors. Single engine
The QSRA is certainly an exception to this rule. capability in the intermediate nacelle angle mode
In general, it has a very natural feel and pilots has been demonstrated by simulated failures and two
readily adapt to its characteristics. Twenty-five actual engine failures.
"guest" pilots with a wide variety of experience
and backgrounds were invited to fly the QSRA. By The tilt-rotor concept is not a new one. The
the third flight, each pilot was able to fly a XV-3 (Fig. 5 ) was flown in the 1950s and has demon-
steep STOL approach with one engine inoperative and strated the feasibility o f converting from a heli-
the SAS turned off. No unusual piloting skills are copter to an airplane. While there were several
required. technical reasons that prevented further develop-
ment at that time, the real "show stopper" was a
The research capability of the QSRA is cur- low-frequency rotor-pylon instability that was
rently being used to investigate advanced elec- discovered during intensive wind tunnel testing.
tronic display and advanced augmentation systems Advances in engine performance, structures, and
for a powered-lift airplane under instrument condi- stability and control led to the development of the
tions in the terminal area. XV-15. The XV-15 first flew in the spring of 1977
and has been in continuous flight test since that
time .
The XV-15 Tilt Rotor
The first thing a pilot notices when being
A fixed-wing airplane is limited in its low introduced to the XV-15 is that on the ground the
speed by the stall. On the other hand, a helicop- airplane is sensitive to lateral control inputs,
ter is limited in its high speed by the retreating and during ground taxi there is a tendency for the
blade stall. The XV-15 tilt-rotor research air- airplane to lean into turns, thereby requiring a
plane, Fig. 4, is an attempt to combine the quali- small amount of lateral control to keep the wings
ties of a fixed-wing airplane and a helicopter to level. Movement of the nacelles provides a respon-
achieve a high-speed vertical takeoff and landing sive and natural means of accelerating and deceler-
(VTOL) airplane with the hover efficiency o f a ating during taxi, practically eliminating any
helicopter. The XV-15 tilt-rotor research airplane requirement for longitudinal cyclic inputs. A
was built by Bell Helicopter Textron under contract nacelle tilt of only 2-3" results in a comfortable
to NASA and the U.S. Army. 10-knot ground speed with neutral cyclic. With a
little differential braking the XV-15 is easily
In the helicopter mode, lift is provided by turned within its own radius.
two three-bladed, 25-ft diameter proprotors
attached to wing tip mounted engine nacelles Hovering the XV-15 out of ground effect is
rotated to the near vertical position. In this considered quite easy, similar to most tandem rotor
mode, the aircraft flies like a twin-rotor heli- helicooters. Only small cyclic inputs are required
copter and is controlled by cyclic and collective to maintain a precise hover over a preselected
control inputs to the proprotor. The collective spot. The only complaint ever voiced is the
provides simultaneous input to all six blades for requirement for small yaw inputs to keep the air-
vertical control; lateral stick produces roll con- plane lined up directionally. In ground effect,
trol through differential collective pitch changes hover is equally good until the airplane reaches
to each proprotor; longitudinal cyclic inputs com- 2 to 3 ft above the ground. Pilot workload is
mand cyclic pitch changes simultaneously to both increased considerably, particularly in the roll
proprotors for fore and aft translation; and the axis. There is also a small but noticeable "suck-
rudder pedals command differential cyclic pitch down" when landing vertically, but it is easily
inputs to provide yaw control. During helicopter controlled. If space is available, it helps to be
operation, all of the fixed-wing control surfaces moving slightly forward at low speed for the final
(conventional ailerons, rudders, and elevators) are touchdown. As in all helicopters, the question
operable; however, they are simply ineffective at arises as to what happens if you lose an engine in
low speeds. The pilot initiates conversion by hover and do not have single engine hover capabil-
pressing a spring loaded "coolie-hat" switch ity. To cope with an engine failure in the XV-15,
located on the collective or power lever. All it has been concluded that a high hover is desir-
intermediate nacelle angles are available to 95" able to allow the pilot to tilt the nacelles for-
for the helicopter mode and to O o for the airplane ward 5" to loo to accelerate for a run-on landing.
mode. During conversion from helicopter to air-
plane mode, the helicopter controls are mechan- A wide hover envelope has been demonstrated
ically phased out as the nacelles rotate forward. which includes sideward and rearward flight to
As the speed increases, the conventional airplane 25 knots. In sideward flight, there is a notice-
controls become effective. It is, of course, pos- able increase in vibration. In rearward flight,
sible to fly at intermediate nacelle angles or to there is an observable nose-down pitching moment
reconvert as desired. The proprotors are driven by caused by airflow over the horizontal tail. The

3
preferred method of translating rearward is to use frequency and intensity of the sound in the cockpit
the full aft nacelle setting of 95" and rotate the is considerably reduced. Outside noise is also
nacelles forward to stop. In general, there is reduced.
really nothing unusual about the XV-15 in the heli-
copter mode. It is a highly stable hover platform In the airplane mode, the collective or power
allowing precision hover with low pilot workload. lever only controls engine power. It is possible
A s with conventional tandem rotor helicopters, it to differentially trim the proprotor collective
is affected very little by wind direction. pitch angles to reduce any steady state sideslip
that may be developed from a power asymmetry. Most
From a piloting viewpoint, the most interest- pilots feel the roll sensitivity is too low and
ing feature is the conversion. As indicated pre- this makes the XV-15 feel too heavy. The airplane
viously, nacelle tilt is controlled by a "coolie mode also exhibits a turbulence response that is
hat" located on the collective control. The pilot unusual. A vertical gust is felt as a longitudinal
can "beep" the controiler or hold it down for a "chugging" which is caused by the transient changes
continuous conversion. Nacelle tilt rate is in blade angle of attack and can result in notice-
7.5"/sec, slowing to 1.S0/sec within 5" of eic,her able thrust changes. In moderate turbulence, this
end of the tilt mode. The XV-15 conversion corri- fore and aft motion is best described as
dor is quite wide, meaning the pilot does not have uncomfortable.
mush concern over whether he will exceed some par-
ticular velocity for a given nacelle setting. The In the airplane mode, i;he XV-15 exhibits very
conversion is quite natural and there is very docile and conventional stall characteristics. In
little longitudinal trim change and low pilot work- the clean airplane configuration, stall speed
load. Of course, the conversion process can be varies between 95 and 110 knots and is usually
stopped at any intermediate setting or a reconver- preceded by a mild buffett or shudder about 5 knots
sion accomplished from any nacelle tilt angle. The prior to stall. Standard airplane recovery tech-
longitudinal acceleration associated with changing niques are quite effective. Because of the high
tilt angle is quite noticeable and occasionally it drag of the proprotors, the XV-15 is capable of
is Roted as being mildly abrupt, depending on the quite high sink rates with the power set at idle.
pilot's technique during "beeping" of the tilt
The reconversion process is simple. Below
controller.
150 knots the flaps are placed to 20" and proprotor
The limits on the conversion corridor are rpm is "beeped" up to 98%. A power reduction
structural and are not due to handling qualities. results in a positive deceleration. The nacelles
The established conversion procedure is to acceler- are unlocked and can be moved continuously upward
ate to between 60 and 80 kngts with the nacelles at when the airspeed is reduced below 150 knots.
the 70" to 80" position while retracting the land- Flaps are lowered to 40" at 60' nacelle tilt. The
ing gear. A t airspeeds past 90 knots the nacelles airplane decelerates rapidly and a forward cyclic
are usually "beeped" continuously. The flaps are input is required to stop a nose-up pitching
raised from 40" to 20" at 60" nacelle tilt angle moment.
and fully retracted by 150 knots. There is a
While no autorotative landings have been
definite increase in cockpit noise as the nacelle
accomplished, steady state autorotations have been
tilt passes 30" and the proprotor reaches its
closest point to the cockpit. Since the pylon tilt entered from 60" to 90" tilt angles. With a stabi-
rate is automatically reduced to 1.5O/sec within 5" lized rpm of 921, sink rates of 3,200 ft/min were
of the stops, a continuous conversion can be accom- found to exist at 80 knots.
plished right down to the stops. As the nacelles Both STOL takeoffs and landings have been
approach 0" (airplane mode) there is a pronounced demonstrated. Considerable increase in performance
nose-down pitching moment which must be corrected is achieved for a very short rolling distance.
with a little aft cyclic. There is little change Both are easily performed and make the pilot look
in roll sensitivity or of the handling qualities in good.
general as tilt angle is varied. The conversion is
best described as a straightforward process which Because of the problems encountered with
has been comfortable for all pilots who have flown rotor-pylon-wing instability in the XV-3 "Converti-
the XV-15. plane," investigations of this problem have con-
tinued in the XV-15. Destabilizing parameters
Once the nacelles reach O o , they are "locked" include swashplate/pylon coupling, wing mode
down using a small toggle switch located just for- effects, increasing airspeed and proprotor rpm.
ward of the conversion switch on the pilot's col- The data from the XV-15 tests are being used to
lective control. This hydraulically forces the avoid similar problems in the V-22 Osprey.
nacelles against the downstops. Proprotor rpm is
then reduced to 86% to preserve the proprotor effi- Extensive hover tests have been conducted to
ciency at the high forward speeds. Although a determine wing download and hover efficiency. It
greater proprotor efficiency exists around 76%, the was dramatic to experience the effect of wing
higher rpm reduces the vibratory loads on the con- configuration (flap deflection) on hover perfor-
version spindle and engine coupling gearboxes. At mance. The results were predictable and obvious,
the 0" nacelle tilt angle and reduced rpm, the as you reduce the wing area exposed to the

4
proprotor downwash, you reduce the power required The RSRA was designed to allow the existing
to hover. With the flaps full up, it required rotor to be removed and new advanced rotors o f
maximum continuous power to hover at 15,000 lb different numbers and lengths of blades to be
gross weight. As the flaps were extended to 75", installed. The RSRA has a full set of rotary-wing
the same power setting resulted in climb rates as and conventional aircraft controls, both of which
high as 1200 ft/min. can be operated either mechanically or through a
fly-by-wire system. Provisions are made in both
Again, in support of the V-22 Osprey, the control systems to allow them to accommodate dif-
XV-15 was used to Evaluate the use of a three-axis ferent rotor configurations.
sidestick controller in a tilt-rotor aircraft.
Several pilots flew the XV-15 through all modes of Because of the increased number of control
flight and unanimously judged the system to be surfaces available in the compound configuration, a
suitable. They found it easier to establish and means of "control shariry" is incorporated in the
hold stabilized airspeed points in level flight, flight-control system. This is accomplished
and altitude and speed during maneuvers. Pari of through a control phasing unit, or CPU, centrally
this was due to the trim follow-up incorporated in located on the console between the two pilots. The
design. None of the pilots liked the yaw axis on CPU allows the pilot to select the proportion of
the sidestick. his control inputs that will be made by the fixed-
or rotary-wing control surfaces. This means the
The XV-15 das used by the military services to pilot can select full rotary-wing or full fixed-
determine the potential of the tilt-rotor concept wing control, or any combination of the two.
for various operational missions. These tests
included: detectability and survivability against The :ail section of the RSRA contains a lower
air defense threat systems; contour and nap-of-the- horizontal all-flying stabilator, an upper fixed
Earth flight; Navy shipboard evaluation; search and horizontal tail plane, two large aft-mounted drag
rescue; external load capability; simulated aerial brakes, a conventional rudder, and a helicopter
refueling from a KC-130 tanker; forward-weapons tail rotor.
delivery patterns, and wake-turbulence evaluacions.
The rotor transmission gear train and tail
All operational tests had positive results. rotor are driven by two General Electric T-58
It was concluded that the tilt-rotor could operate engines. Even though the main rotor was removed
safely in the terrain flight environment and that for the fixed-wing flight tests, it was necessary
its survivability was improved over that of conven- to operate both T-58 engines to drive the electric
tional airplanes and helicopters. The tilt rotor generators, hydraulic pumps, and tail rotor. The
operated well in the shipboard environment and power plants which allow the RSRA to be flown as a
integrated easily into the helicopLer regime. It fixed-wing airplane are two General Electric TF-34
was shown that personnel could easily operate high-bypass turbojet engines mounted on either side
beneath the tilt rotor and the downwash presented of the fuselage. In the compound helicopter mode,
no unusual problems for external loads or for these engines are derated to 6,250 lb thrust
hoisting survivors. The full potential of the each. For the fixed-wing flight test program,
tilt-rotor concept has yet to be explored; it will thrust was increased to 8,250 lb. The throttles
require new employment tactics and will introduce for the two TF-34 engines are two twist grips
totally new missions for military use. located on the collective controller.
The XV-15 has been fitted with new composite The RSRA has a 45-ft wing which is unique
proprotors made of high-strain graphite. The because it can vary its angle of incidence from 9"
blades can be fitted with three different blade leading edge down to 15" leading edge up. Wing
tips and three blade cuffs. All phases of heli- incidence changes are made through two large
copter and conversion flight will be investigated. hydraulic pistons attached to the leading edge of
the wing. Control of wing incidence is provided
Overall the XV-15 has proven to be an out- through a handle on the center console between the
standing research vehicle resulting in major two pilots. The forces and moments generated by
improvements in the field of vertical and short the main and tail rotors, auxiliary engines, and
takeoff aircraft. the wing are measured by a series of load cells
and/or the active isolation and balance system
which is an integral part of the aircraft
Rotor Systems Research Aircraft structure.

The Rotor Systems Research Aircraft (RSRA) is The RSRA incorporates a unique blade-severance
a unique research airplane designed t o flight test and crew-escape system. This pyrotechnically oper-
advanced helicopter rotor systems. Its principal ated system allows the crew to independently sever
flight test configuration is as a compound helicop- the rotor blades and/or eject from the aircraft.
ter, Fig. 6 (combination fixed- and rotary-wing Two Martin Baker ejection seats are installed to
aircraft). The RSRA has been flown as a helicop- provide the pilots with this capability.
ter, Fig. 7, and also as a fixed-wing airplane,
Fig. a. Since the compound configuration is too heavy
to hover (minimum speed 40 knots), the R S ~ Acan

5
also be flown as a pure helicopter. This configur- Movement of the variable incidence wing in
ation fills the low-speed performance gap and flight proved to be rather benign. Essentially,
allows complete investigation of a candidate rotor the fuselage rotates about the wing and the angle
system in the hover and low-speed range. The basic of attack remains constant. The wing-to-elevator
airframe was first flown as a helicopter to check interconnect practically negates the requirement
out those systems common to all three configura- for a cyclic trim change.
tions prior to adding the airplane parts.
The only noticeable effect on the handlin:
The helicopter configuration is relatively qualities with a change in wing incidence is a new
easy to fly and handles very much like a Sikorsky flight attitude and quite different power require-
H-3 series helicopter. The original SAS gains were ments for the auxiliary engines during turning
overly sensitive and resulted in a lateral pilot- flight. The higher wing-incidence angles require
induced oscillation on landing. Reducing the SAS considerably more power in a turn to maintain a
gains greatly improved the handling qualities. The constant speed.
extreme length of the helicopter requires caution
during rapid flare or quick stop maneuvers near the The greatest influence of wing incidence is
ground. where the maximum stresses occur during high-speed
flight. Much of the flight-test effort has been
During the buildup to the compound configura- devoted to mapping the structural loads as a
tion, the helicopter was flown with the horizontal function of wing incidence and collective set-
tail attached and then with the TF-34 engines ting. The limiting structural loads at high speed
mounted on the fuselage but without the wing. It occur either in the main rotor blades or in the
was not possible to autorotate in the helicopter upper horizontal stabilizer. At high wing-
mode with the compound configuration horizontal incidence angle, the horizontal stabilizer reaches
tail installed. The pilot found that he approached its endurance level around 180 knots. At low wing-
the left rudder pedal limit with only a 100 ft/min incidence angle, the main rotor blades reach their
rate of descent. endurance levels at about the same speed. A lower-
ing of the collective tends to reduce the rotor
The first unique research program took advan- loads. A push rod load indicator in the cockpit
tage of the rotor-system force measurement sys- provides excellent information to the pilot when
tem. Highly accurate vertical drag data were the rotor loads start to exceed their limits. To
obtained in hover and at low speed allowing deter- date, the structural loads on the aircraft have
mination of rotor downwash and fuselage interfer- been mapped at wing angles of O", 5", 7.5", and 10"
ence effects, thus achieving measurements never from 50 to 180 knots at several collective
before possible in actual flight. settings.
The helicopter configuration was used exten- The compound configuration is landed very much
sively during the buildup to flying the compound like a fixed-wing airplane. Approach speeds of
configuration. It was possible to make running 110 knots downwind, 100 knots on base leg, and
takeoffs and run-on landings with the helicopter, 90 knots on final, provide an adequate margin above
matching the compound speeds. wing stall. Maintaining a constant 20% collective
setting and these speeds allows the compound to
The takeoff technique for the compound is a touchdown in a slight tail-low attitude at 70 knots
compromise between that used for a tail dragger with only a slight flare.
fixed-wing aircraft and a rolling takeoff for a
heavyweight helicopter. The following technique A 7.5" wing incidence places the aircraft in a
was established: Establish the cyclic in the near 3-point attitude with jets at idle and 70-knot
center position using the cockpit control position touchdown target. There is no tendency to float or
indicators keeping the collective full down until enter any kind of pilot-induced oscillation (PIO)
the takeoff roll begins. Slowly advance the TF-34 in ground effect. If, in rare occasions, one mis-
engines to takeoff power as the aircraft rolls judges the location of the wheels and ends up
forward. As takeoff power is stabilized, collec- slightly high prior to touchdown, a slight downward
tive is increased slowly to arrive at 40% collec- pressure on the collective provides the flexibility
tive position and 70 knots simultaneously. to recover gracefully.
With the wing set at a 10' incidence, the The fixed-wing configuration of the RSRA was
aircraft simply flies off the runway in the 3-point primarily considered an emergency fly-home mode in
attitude with only minimum control input required the event it became necessary to sever an unstable
to maintain that attitude. Acceleration to rotor system in flight. While it had always been
90 knots occurs quite rapidly, and without changing planned to flight test the fixed-wing configura-
power. A slight aft cyclic movement results in a tion, the selection of the RSRA as the flight test
90-knot climb. Up-and-away flight is very comfor- bed for the X-wing rotor accelerated these tests.
table, much more so than with the pure helicop- The X-wing rotor will be the first completely "new"
ter. The compound is more stable and exhibits less rotor system to be flight tested on the RSRA.
vibration.
The compound configuration was used exten-
sively in the build-up to the fixed-wing flight

6
tests, It was used to investigate the lift and transport. Control sensitivity at 150 knots is low
stall characteristics of the wing. Near zero lift and the damping high in all axes. Control sensi-
was obtained on the RSRA rotor and the fixed-wing tivity increases rapidly with speed, becoming quite
angle of attack increased to stall. 1s the wing sensitive above 180 knots.
stalled, lift was rapidly transferred from the wing
to the autorotating rotor. A rapid increase in The tail rotor is the limiting factor on the
rotor rpm became the best way to determine when the maximum speed that c,>uldbe attained. It is neces-
wing stalled. The wing, fuselage, and empennage sary to keep the tail rotor tip speed below Mach 1 ;
area were extensively tuffed. Photo coverage this is accomplished by reducing tail rotor rpm to
showed that the stall progressed in a classic 943 and limiting the maximum speed of the aircraft
manner for a straight-winged aircraft. The root of to 250 knots. To further reduce structural loads
the wing stalled first, although there was little
on the tail rotor, above 200 knots, rudder pedal
or no noticeable stall warning that could be felt
inputs are prevented from reaching the tail rotor
in the cockpit. Increasing the collective setting by moving the yaw control phasing unit to the full
to the rotor was a rapid and positive way to fixed-wing position and restricting sideslip to
decrease the angle of attack on the wing. less than 7.5'.

The compound configuration was also used to Most of the up-and-away flying was devoted to
evaluate the transfer of pilot control inputs from the primary task of determining the control power
a combination rotary- and fixed-wing input to available for each control axis. Performance data
fixed-wing only inputs. Pilot control inputs were were obtained during the envelope expansion
incrementally washed out to the rotary wing by flights. An anomaly noted during the stability
advancing the control phasing unit lever to the tests was that the airplane exhibited a different
f u l l fixed-wing position. The airplane was slug- response for nose-up and nose-down inputs. Nose-
gish in this configuration but quite controllable. down inputs tended to appear uncoupled from the
roll axis while nose-up inputs always resulted in a
The fixed-wing RSRA embodies all the charac- roll to the left. It is believed that this is
teristics not to build into a tail dragger airplane caused by air flow interaction with the tail rotor.
if one wanted to reduce its tip-over and ground
looping tendencies. The high center of gravity, For the landing approach, the airplane flies
high-thrust line, high side-force area, and narrow nicely at a 140-knot approach speed and is easily
controlled down to 120 knots. The planned touch-
gear, when coupled with low-frequency directional
control, caused considerable concern regarding the down speed is near 115 knots. The airplane could
ground handling of the airplane. The primary rea- be landed in the 3-point wheel landing or tail-
son for keeping the tail rotor on the airplane was wheel first attitude. The determining factor is
to assist in ground handling and to reduce the how slow you get in the landing flare. The actual
engine-out speed of this multiengine airplane. The landing is less of a problem than the considerable
high-speed taxi tests indicated that, while not attention required for directional control once on
particularly good, the groLnd handling qualities the ground. The shutdown of the right TF-34,
were acceptable. It was decided, however, to limit required to reduce landing distance, aggravates the
the airplane to crosswinds of less than 15 knots at directional control problem. Total landing
the lower weight and lower center of gravity con- distances as low as 5,500 ft were demonstrated.
figuration, and to 10 knots at the higher weight Stopping distances of 2,500-3,000 ft are required.
and higher center of gravity. During the fixed-wing flight tests, two dif-
The taxi tests confirmed the 5" wing incidence ferent rotor hubs were flown, providing a unique
and 15" flap setting for the lighter weight config- flight-test opportunity. With the variable inci-
uration. Because of the rapid acceleration of the dence wing, aircraft pitch attitude could be
airplane, TF-34 power settings had been limited to changed while maintaining the same airspeed and
70% fan speed or less during the ground taxi altitude. This allowed a complete set of pure
tests. At the lighter weight, the thrust-to-weight rotor-hub-drag data to be obtained by comparing the
ratio is approximately 0.7. The tail wheel came drag of the RSRA with and without the main rotor
off the ground between 75 and 85 knots. installed.

A maximum TF-34 fan speed of 80% was estab- The fixed-wing flight test program provided
lished for the first flight takeoff. On takeoff invaluable data for the design and flight-test
roll, the copilot calls out fan speed as the pilot efforts of the RSRA/X-wing aircraft. The X-wing is
manipulates the two throttles while the airplane an inflight stoppable and restartable rotor that
accelerates down the runway, Difficulty in match- uses circulation control for both lift and con-
ing engine power compounds the directional control trol. The capability of the RSRA to provide inde-
problem. pendent control of both lift and drag, together
with its unique flight-control system, make it the
At liftoff there is a slight tendency to over- ideal test vehicle to research and demonstrate
control the airplane in pitch. Climb performance X-wing technology in flight. Flight test of the
is impressive with climb speed at 150 knots. X-wing rotor should commence in the Fall of 1987.
Up-and-away, the fixed-wing airplane is quite
stable, very similar to a medium-weight

7
 The RSRA provides the flight test community a tal Aircraft." The Society of Experimental Test
versatile research tool. It will play an important Pilots 1983 Report to the Aerospace Profession,
role in advancing the state-of-the-art in helicop- 27th Symposium proceedings, Beverly Hills, CA,
ter rotor systems design. Vol. 18, No. 4, 1983.
2Dugan, D. C., "The XV-15 Tilt Rotor Pro-
Bibliography gram." The Society of Experimental Test Pilots
1985 Report to the Aerospace Profession 29th Syrnpo-
I sium Proceedings, Beverly Hills, CA, ISSN
QSRA
I
I 80742-3705, 1985.
'Cochrane, J. A., "The Quiet Short-Haul
Research Aircraft," NASA CR 166440, 1983. 3Dugan, D. C . , Erhart, R. G., and Schroers,
I L. G., "The Tilt Rotor Research Aircraft." The
2Cochrane, J. A., Riddle, D. W., and Stevens, Society of Experimental Test Pilots 1980 Report to
V. C., "Quiet Short-Haul Research Aircraft: The the Aerospace Profession, 24th Symposium Proceed-
First Three Years of Flight Research," AIAA Paper ings, Beverly Hills, CA, Vol. 15, No. 2, 1980.
81-2625, AIAA/NASA Pnes VSTOL Conference, Palo
Alto, CA, Dec. 7-9, 1981. 'Green, D. L. , "Flying the V-22 Predecessor:
I The NASA/Army Bell XV-15." Rotor and Wing
3Martin, J. L., "The Quiet Short-Haul Research International, June 1985.
Aircraft (QSRA)." The Society of Experimental Test
Pilots 1979 Report to the Aerospace Profession, 5Magee, J. P., "The Tilt Rotor Research Air-
23rd Symposium Proceedings, Beverly Hills, CA, craft (XV-15) Program." NASA CR 166440, 1983.
Vol. 14, No. 4, 1979.
h h r o e r s , L. G. , "Dynamic Structural Aero-
'Martin, J. L., Strickland, P. B., "Shipboard elastic Stability Testing of the XV-15 Tilt Rotor
Trials of the Quiet Short-Haul Research Aircraft Research Aircraft .'I NASA TM 84293, Dec. 1982.
(QSRA)." The Society of Experimental Test Pilots
1980 Report to The Aerospace Profession, 24th Sym- -
RSRA
posium Proceedings, Beverly Hills, CA, Vol. 15,
'Hall, G. W., and Merrill, P. K., "Flight
No. 2, 1980.
Testing the Rotor Systems Research Aircraft
5Queen, S., and Cochran, J. A., "QSRA Joint (RSRA)." NASA TM 85852, Oct. 1983.
Navy/NASA Sea Trails," AIAA Paper 81-0152, AIAA
19th Aerospace Sciences Meeting, St. Louis, MO, 2Hall, G. W., and Morris, P. M., "Flight
Testing the Fixed Wing Configuration o f the Rotor
I Jan. 12-15, 1981.
Systems Research Aircraft." The Society of Experi-
6Watson, D. M., "Quiet Short-Haul Research mental Test Pilots 1984 Report to the Aerospace
Airplane (QSRA) Mode Select Panel Function Descrip- Profession 28th Symposium proceedings, Beverly
tion." NASA TM 84234, May 1982. Hills, CA, ISSN 80742-3705, 1985.
3Linden, A.m W., "RSRA/X-wing: A Status
XV-15
Report." Sikorsky Aircraft Division Memorandum,
'Carpenter, R. B., and Ball, J. C., "XV-15 June 1986.
Experience, Joint Operational Testing an Experimen-

8
 Fig. 1 The Quiet Short-Haul Research Aircraft (QSRA).

t
1
'.I
5

0 -
---
DLC ON
USB 50'
WT 48000 Ilrs
-
ALT. SEA LEVEL
i APPROX. X
FAN rpm

.r
I-I
2
I-

o -5-
U-J
. TNY-0.27 (70%)
\

-10 -

.-
50 60 70 80 90
EQUIVALENT AIRSPEED - kti

Fig. 2 Flightpath vs. velocity diagram: landing configuration.

Fig. 3 QSRA on approach to USS Kitty Hawk (CVA-63).

9
 ORIGINAI; P A a 1s
OF POOR,QU&.

.
u
I

AIRPLANE

TRANSITION
Fig. 6 RSRA Helicopter Configuration.

'.
HELICOPTER

Fig. 4 XV-15 Tilt Rotor Aircraft.

Fig. 7 RSRA Compound Configuration.

Fig. 5 XV-3 Convertiplane.

Fig. 8 RSRA Fixed Wing Configuration

10
 Report Documentation Page
1. Report No. 2. Government Accession No. 3. Recipient's Catalog No.

NASA TM-100025
.
4. Title and Subtitle 5. Report Date

September 1987
Flight Research at NASA Ames Research Center: 6. Performing Organization Code
A Test Pilot's Perspective

7. Authorls) -I.-Performing Organization Report No.

G. Warren Hall A-87323 - _ _ ....

10. Work Unit No.

9. Performing Organization Name and Address

999-31-05
11. Contract or Grant No.
Ames Research Center
Moffett Field, CA 94035-5000 ______-.__
13. Type of Report and Period Covered
12. Sponsoring Agency Name and Address
Technical Memorandum
National Aeronautics and Space Administration 14. Sponsoring Agency Code
Washington, DC 20546-0001
15. Supplementary Notes

Point of Contact: G. Warren Hall, Ames Research Center, M/S 211-17

Moffett Field, CA 94035 (415) 694-5279 or FTS 464-5279

16. Abstract
.In 1976 NASA elected to assign responsibility for each of the various
flight regimes to individual Research Centers. NASA Ames Research Center at
Moffett Field, California was designated lead center for vertical and short
takeoff and landing, V/STOL research. This paper will discuss from the test
pilot's perspective the three most recent flight research airplanes being
flown at the Center: the Quiet Short Haul Research Aircraft; the XV-15 Tilt
Rotor Research Aircraft; and the Rotor Systems Research Aircraft.

17. Key Words (Suggested by A u t h o r M 18. Distribution Statement

Flight test, Flight research, Test pilot, Unclassified-Unlimited
Quiet Short Haul Research Aircraft, XV-15
Tilt Rotor, Rotor Systems Research Air-
craft, NASA flight research, Vertical and
short takeoff and landina aircraft
19. Security Classif. (of this report) 20. Security Classif. (of this pagel
-
Subject Cateaorv
21. No. of pages
- 05 ______
22. Price

Unclassified Unclassified 12 A02