VSKYLABS Information Page
The performance of the VSKYLABS MARTIN MARIETTA X-24A model
for X-PLANE flight simulator is accurately based on
all relevant data that is specified in this report.
ANALYSIS OF THE
FLARE AND LANDING
CHARACTERISTICS OF THE
X-24A LIFTING BODY
THIS REPORT MAIN TOPICS:
TEST AND EVALUATION
UPPER FLAP APPROACHES
LANDING GEAR CHARACTERISTICS
LANDING ROCKET USAGE
GROUND HANDLING AND SURFACE WINDS
ENERGY MANAGEMENT DURING THE APPROACH
F-104A LOW LIFT/DRAG DATA
SIMULATION TEST RESAULT
THIS PAGE CONTAINS DETAILED INFORMATION TAKEN FROM MATERIAL APPROVED FOR
UNLIMITED PUBLIC RELEASE.
This technology document presents.an analysis of the approach,
flare and landing characteristics of the X-24A lifting reentry research
vehicle. The X-24A made 28 landings on Rogers Dry Lake at Edwards Air
Force Base between 17 April 1969 and 4 June 1971. This test program was
conducted jointly with the NASA Flight Research Center (FRC). References
1 through 8 are related documents which have been or will be published.
The author wishes to acknowledge the efforts of Mr. Richard R. Larson
of NASA FRC who gathered the touchdown accuracy and landing rollout data,
and Mr. Robert G. Hoey of the AFFTC who provided the F-104A idle power
performance data. Acknowledgement is also extended to Mr. John A. Manke
of NASA FRC and Major Cecil W, Powell, who gave piloting comments and
assistance in preparing this document.
The participation of AFFTC personnel in this program was authorized
by Project Directive 69-38. The assigned Program Structure was 690A.
Foreign announcement and dissemination by the Defense nncumentation
Center are net authorized because of technology restrictions of the U.S.
Export Control Acts as implemented by &FR 400-10.
This report presents the results of the X-24A lifting body flight
test program pertaining to the approach, flare and landing. The approach
and flare were flown with clean configurations (landing gear up), having
maximum lift-to-drag ratios (L/D)'s between 2.95 and 4.30. The landing
configuration (gear down) had a maximum L/D value of 2.65. The landing
approach pattern was a 180-degree, unpowered, left-hand, circling approach
followed by a short wings-level, high energy final approach at airspeeds
ranging from 267 to 318 KCAS, terminating with a flare to near level
flight near the ground, and a deceleration to landing. The landing gear
was extended near the completion of the flare in close proximity to the
ground. The landings were at airspeeds between 168 and 205 KCAS at sink
rates of less than 5 feet per second and within 2,000 feet of the intended
landing point. An effective speed brake function was generated on the
X-24A by using the flap bias feature. This was essential to the accomplishment
of accurate landings. Visibility and handling qualities of
the X-24^ in the final approach and landing configuration were excellent;
however, lateral upsets» in turbulence were disconcerting to the pilot
especially near the ground. The trim change due to landing gear deployment
was also somewhat objectionable to the pilots. Differential braking and
rudder/aileron deflection were sufficient for ground steering of the X-24A
after touchdown for conditions of zero crosswind, but were inadequate for
light to moderate crosswinds. This program obtained results similar to
those of earlier lifting body programs and of routine X-15 operations
indicating that the unpowered visual approach and landing was relatively
easy for the project test pilots after extensive F-104A low L/D practice
prior to the X-24A flights .
The X-24A project was the second flight test project to use the SV-5
lifting body reentry configuration. The first project, called PRIME
(Precision Recovery Including Maneuvering Entry) used three subscale
unmanned SV-5's whicfT were boosted to orbital speeds on Atlas boosters.
This program provided data, and a feasibility demonstration of the SV-5
configuration in the technical areas of aerodynamics, stability, control,
heat protection, and maneuverability covering the speed range from orbital
velocity to Mach 2.0.
The purpose of the second project, the X-24A project, called PILOT
(Piloted Low Speed Test), was to investigate maneuverable lifting body
flight from the low supersonic speed range to touchdown. One of the main
X-24A project objectives was to gather data on and to prove that the SV-5
configuration could be maneuvered to a safe horizontal unpowered landing
at a pre-selected landing site. Twanty-eight successful X-24A landings
accomplished this objective. This report analyzes the X-24A unpowered
approach and landing characteristics from approximately 20,000 feet mean
sea level (MSL) down through the flare, touchdown, and ground rollout.
Pertinent pilot comments are included within the discussion.
The X-24A vehicle (figures 1 and 2) was a blunt-nosed, wingless
lifting body vehicle. The aft end had a large base area and three vertical
fins. The outboard fins supported the two upper and the two lower
rudders. The upper rudders were used for directional trim and control.
The four rudder surfaces could be biased inboard or outboard for improved
handling characteristics at the subsonic or supersonic airspeeds, respectively.
On the aft end were located the two upper and the two lower flaps
which were used for pitch and roll control. When either lower flap reached
the fully closed position (zero degrees), pitch and roll inputs were transferred
through a clapper mechanism to the corresponding upper flap. Both
pairs of upper and lower flaps could be deflected symmetrically (biased)
independent of pilot control input. The rudder bias was programmed as a
function of upper flap bias position. The flap bias feature was used as
a speed brake. The cockpit contained typical aircraft instruments, stick
and rudder pedals, zero -zero ejection seat, and a jettisonable bubble
canopy. The landing gear system was a conventional tricycle arrangement
with a quick acting pneumatic deployment system. In addition to the primary
propulsion system (XLR-11 rocket engine) with 8,480 pounds of vacuum
thrust, the vehicle was equipped with two 500-pound thrust hydrogen peroxide
rocket engines which were available to the pilot to extend his glide during
the landing approach if required. Reference 1 contains a more extensive
description of the X-24A. The subsystems pertinent to the approach and
landing task are discussed in appendix I and in reference 2.
The X-24A was launched from an NB-52 mother ship at an altitude of
40,000 to 47,000 feet MSL. On 10 flights, the entire flight was a preplanned
glide flight from launch to touchdown. On 18 flights, the first
portion of each flight was powered, in which case the vehicle accelerated
to a preplanned Mach number and altitude using the primary propulsion
system. On the powered flights the vehicle entered a gliding flight
phase at rocket engine shutdown. After launch on both the glide and
powered flights, various stability and performance data gathering maneuvers
were executed. During this phase, the pilot's attention was primarily
directed to data maneuvers and he was relying heavily on the ground
controller for navigation information. These ground calls, based on real
time radar data in the control room, consisted of vehicle position relative
to the preplanned ground track and altitude profile. These ground
calls were to guide the pilot to a low key point. Low key is defined as
the point where the final 180 degree turn was initiated and was considered
the dividing point between data maneuvers and the landing maneuver.
From low key the pilot performed, without ground controller assistance,
an unpowered visual gliding approach to the preselected runway where an
unpowered visual flare and landing were completed.
The approach and flare were accomplished with the landing gear retracted.
The landing gear was not extended until flare completion since
the low gear-down lift-to-drag ratio (maximum L/D of 2.65) and the geardown
limit airspeed (300 KCAS) would have made visual flares and landings
quite marginal (reference 9) if the entire approach and landing had been
flown with the gear extended.
Various parameters (appendix I) were telemetered to the ground for
later examination to determine the vehicle characteristics and pilot
technique during the approach, flare and landing. Bank angle ($) and
radar space positioning data were used to determine the time and location
of the low key points. The bank angle data were used to determine the
altitude at which the pilot had rolled out on the final approach. The
point at which the pilot had started the flare was determined by examining
the angle of attack (a), longitudinal stick position, normal acceleration
(nz, ) and pitch angle (8) data. Gear extension and touchdown were determined
from the landing gear strut position data.
For most of the flights, the pilots were asked to perform landings
at a predetermined point on a designated runway with the stipulation that
flight safety should in no way be compromised to achieve this objective.
On four flights (17 to 20), the pilots were requested to maintain a desired
final approach airspeed and forego the accuracy landings. The
pilots were also requested not to use the flap bias feature as a speed
brake for the first seven flights.
The desired approach and landing configuration (flaps and rudder
bias settings, stability augmentation system (SAS) gains, etc.) were
selected during the planning phase of each flight using a fixed base
six-degree-of-freedom simulator. Before each flight the pilots practiced
unpowered approaches to the primary runway and to the alternate runways
using an inflight performance simulator. This simulator was an F-104
aircraft in high drag configurations which provided L/D's and airspeeds
similar to those of the X-24A (appendix V).
Tables I, II and III present a listing of the vehicle landing weights
and pilots, together with the conditions existing during the initial and
final approaches,at flare initiation, gear extension, flare completion,
and touchdown for each flight. These tabulations form the nucleus of
this report and will be referred to frequently in subsequent sections.
The X-24A flight test program had thrne test pilots (A, B and C
in table III) . Table IV presents a summary of the lifting body piloting
experience of each of these pilots.
Vehicle gross weights at touchdown ranged from a low of 6,157 pounds
to a high of 7,023 pounds. The average touchdown weight was 6,387 pounds,
which corresponded to a planform loading of 39.4 pounds per square foot.
All the landings reported herein were made on marked, 300-foot wide,,
runways on the hard surface of Rogers Dry Lake at Edwards Air Force Base,
California. All of the flights terminated on ehe planned runway except
for flight 3 in which an alternate runway was used due to an inadvertent
early launch. Figure 3 is an aerial drawing of Rogers Dry Lake showing
the marked lakebed runways. Nineteen landings were made to runway 18
in figure 3. Seven landings were performed to runway 33. One landing
was made to runway 15 and the one early launch flight landed on runway 17.
During the course of the X-24A flight test program there were three
distinct subsonic configuration modes used during the approaches and
1. Configuration Mode 1 - Flights 1 to 7, pitch and roll control with
lower flaps, upper flap bias set at -18 to -23 degrees depending on
2. Configuration Mode 2 (figure 4) - Flights 8 to 18 and 21 to 28,
pitch and roll control with lower flaps, upper flap bias at -13
degrees until gaar extension at which time control crossed over from
the lower to the upper flaps and the actual landing was performed
using the upper flaps (figure 5). This configuration was the standard
subsonic approach and landing configuration for the X-24A.
3. Configuration Mode 3 - Flights 19 and 20, lower flaps fully closed
(zero degrees), pitch and roll control with the upper flaps which
varied between -8 and -12 degrees during the approach.
These configuration modes will be referred to throughout this document.
TEST AND EVALUATION
APPROACH, FLARE AHO LANDING PHASES
The unpowered visual approach, flare and landing were divided into
four phases for analysis (figures 6 and 7). In actual practice, these
phases were blended together by the pilot in a slightly different manneL
on each individual flight. For this reason, caution must be xxxxx in
generalizing the statistical values included herein.
I. Initial Approach Phase - A 180-degree circling unpowered turn starting
at a low key point between 18,000 and 21,000 feet MSL.
II. Final Approach Phase - A constant airspeed, gliding approach on the
runway heading with bank angle less than +10 degrees.
III. Landing Flare Phase - A wings level, angle of attack increase at
elevated load factor producing a transition from the steep final
approach glide angle (Y„) to a shallow deceleration-'uo-touciidown
glide e.iigle. The landing gear was extended near tho end of the
IV. Deceleration Phase - Deceleration along e shallow glide angle to
the point of touchdown.
The basic pattern used during the X-24A program was a 180-degree,
unpowered, left-hand, circling approach beginning at a low key point.
The advantages of this pattern over other possible patterns for the X-24A
1. The variation of bank angle, airspeed and speed brake deflection
provided for a broad energy management capability.
2. A 270- or 360-degree overhead pattern (X-15 pattern) would have
devoted too much valuable flight time to the landing pattern and
thus compromised the amount of research data gathered per flight.
3. A straight-in approach would be more difficult from a pilot's
judegement standpoint. It would be more difficult to judge glide
angle from the distances and altitudes associated with this type
of approach. Also, a straight-in approach would have substantially
reduced any energy management capability through bank angle modulation.
There were two low key points: the preplanned low key and the
actual low key.The preplanned low key point was the turn initiation
point for the X-24A landing approach with no wind.It was initially
determined by the pilots during their training when using the inflight
performance simulator aircraft and the fixed base X-24A simulator.The
preplanned low key point was located 2.8 to 3.4 nautical miles to the side
of, and 1.6 to 2.4 nautical miles downwind from the end of the runway(figures 6 and 7).
There was an altitude associated with this point to
aid the pilot in assessing his energy.This altitude was 19,000 to 21,000
feet MSL for configuration mode 1 and 18,000 feet for configuration modes
2 and 3; however, there was a comfortable altitude window of at least
5,000 feet (+4,000, -1,000 feet) associated with the preplanned low key
point.This point was changed several times during the early X-24A
flights as the pilot re-examined his evaluation of the pattern and as
different approach configurations were tested.
The actual low key point was the point at which the pilot actually
initiated the turn on each X-24A flight.
Because of the type of research mission flown in which the pilot
was performing data maneuvers down to the low key point and in some
cases beyond it, verbal guidance from a ground controller was necessary
to the pilot. The help that the pilot received from the ground controller
was an assessment of how far he was above or below the preplanned altitude
profile or left or right of the preplanned ground track! and how far he
was from the preplanned low key point. Based on this information the
pilot could judge his flight conditions, terminate data maneuvers and reconfigure
from the transonic configuration2 to the subsonic configuration (either modes 1, 2 or 3) prior to reaching low key.
A continuous assessment of total energy was made by the pilot based on known winds and
ground controller guidance information.This information plus a visual
assessment was used to determine the actual low key point (see the section
titled Effects of Winds and Turbulence in the Approach Pattern and Flare).
Orce the turn was started, the pilot received no energy management advice
from the ground controller.The pilot used his own experience and judgement
to fly the pattern. The ground controller was available to remind
the pilot of any specific events to be performed during the approach.
The chase aircraft pilots observed their airspeeds and power settings at
various times in the pattern when they were stabilized alongside the
X-24A. These observations were useful later in improving the inflight
performance simulation of the X-24A.
The piloting task during the initial approach phase was to position
the vehicle on an imaginary glide slope intercepting a preflare aim point
on the ground and accelerate to the desired final approach (preflare) airspeed.
(An instrument landing system (ILS) was not used with the X-24A.)
The pilots generally planned to have excess energy approaching the low
key point. They modulated this energy during the initial approach to
arrive on the desired glide slope either by airspeed and/or bank angle
modulation or they remained at approximately the same airspeed and used
the flap bias feature for speed brakes (appendix 111). The pilots had
an intermediate altitude window of between 12,000 and 15,000 feet MSL
at the 90-degree point of the pattern, and they used this to aid them
with their energy management. Figure 9 shows the X-24A at approximately
the 90-dearee point. Also refer to figure 8 (frames t = 66 to 46).
For the initial approach phase, the pilots tried to arrive at the
low key point with an airspeed between 190 and 230 KIAS and then allowed
it to increase during the turn while striving to have at least 250 KIAS
at the 90-degree point and the desired final approach speed at rollout
on final. The pilots decreased the angle of attack as they rolled out on
final to maintain the final approach airspeed.
The desired final approach (figure 6) was a wings level, constantindicated-
airspeed, gliding approach of short duration lined up with the
runway. Refer to figure 8 (frames t = 44 to 30), The pilot was not
trying to rollout on final at any specific altitude, but rather was trying
to position the vehicle on an imaginary glide slope which intercepted
the preflare aim point3. The pilots were striving to establish a final
approach airspeed of 300 KIAS, which was more than adequate for a flare.
However, the airspeed was allowed to vary somewhat in order to establish
a glide angle which would intercept the desired aim point. This permitted
the pilot to arrive at a flare initiation point with sufficient energy
to accomplish the flare and with excess airspeed after flare to allow for
some, margin for error.
The transitional maneuver between the steep final approach glide
angle and the shallow deceleration glide angle was a wings level angle of
attack increase to an elevated load factor. Refer to figure 8 (frames t
= 28 to 10). The pertinent characteristics (altitude and airspeed lost
and the timing) of this flare maneuver were predicted prior to the first
flight through the use of the X-24A fixed base simulator and the inflight
performance simulator using F-104 aircraft.
During-the actual X-24A flights, the pilots stated that their primary
cue for flare initiation was altitude. However, they indicated
that the determination of flare initiation was also a combination of
pilot experience and the interrelation of several factors including sink
rate, airspeed, altitude, and their position relative to the desired aim
point. The pilots made corrections to the desired final approach glide
slope/aim point right up to what they considered their flare initiation
point. Therefore, it was often difficult to differentiate between these
corrections, which produced "g" excursions, and what the pilots considered
their actual flare maneuver although technically these glide slope corrections
usually resulted in a partial flare maneuver.
A flare initiation airspeed of 300 KIAS was suggested to the pilot;
however, simulator studies indicated that flares could be initiated at
airspeeds as low as 250 KIAS. On four flights this target airspeed was
For postflight data analysis, the flare initiation point was determined
by examination of the normal acceleration, angle of attack, pitch
angle and stick position data. The flare completion point was arbitrarily
considered to be the point at which the flare initiation sink rate had
been arrested to 20 feet per second as determined by the radar altimeter.
For the X-24A, this prell ere aim point was 1*1/4 miles short o( the intended touchdown point.
At or near flare completion, the pilot extended the landing gear
and the vehicle decelerated along a shallow glide angle (less than
three degrees) to the point of touchdown. Refer to figure 8 (frames t
= 8 to 0). The length of time of this phase was very important because
i'; allowed for some margin for error in the flare maneuver. It also gave
tie pilot time to recover from the gear transient and to adjust the sink
rate to a very low value prior to touchdown. The amount of time that the
pilot had during the deceleration phase was obviously a function of the
gear down performance and the touchdown airspeed, but more importantly,
was related to the energy at flare initiation (airspeed and glide angle)
and the flare technique used (high or low load factor).
The chase aircraft pilot performed several functions during this
time period: (1) advised the X-24A pilot of his distance above the
ground, (2) reminded him to extend the landing gear, (3) reminded him to
be aware of the gear transient and (4) confirmed that the gear was down.
Figure 10 presents the L/D versus equivalent airspeed for the three
landing gear up approach configuration modes and the ianding gear down
touchdown configuration. These data are the flight data as presented in
reference 4. All of the gear-up data were obtained from push-over/pull-up
maneuvers and the gear down data were obtained during the deceleration
between gear extension and touchdown. The data in figure 10 were corrected
to a planform loading of 39.3 pounds per square foot which corresponds to
a gross weight of 6,360 pounds. All of the X-2 4A approaches were made
at Mach numbers of 0.55 and less.
There are certain characteristics of these L/D curves whi ^h are important
to test results. The positive slope of this L/D curve (higher
airspeeds to the left of maximum L/D) was commonly referred to as the
"front side" and was the speed stable flight regime. When operating
on the front side, lowering the nose would increase the airspeed and
decrease the L/D, producing a steeper flightpath angle. Raising the nose
would reduce the airspeed and increase the L/D, producing a shallower
flightpath angle. The vehicle would also be speed stable while ..lying
along a particular flightpath angle, i.e., excess speed along a given
flightpath angle would produce excess drag resulting in a deceleration to
the desired speed. The degree of speed stability would increase with the
steepness of the L/D curve. Flying on the front side of the L/D curve
was a natural task for the pilot and all X-24A approaches were made in
The negative slope portion of the L/D curve (lower speeds to the.
right of maximum L/D) was designated the "back side" and was characterised
by speed instability. Lowering the nose would increase the airspeed
but (unlike the case on the front side) the L/D increased with increasing
airspeed and consequently would result in a shallower flightpath
angle. It would have been an unnatural task for the pilot to conduct
the energy management task on this portion of the L/D curve. No X-24A
approaches were made on the back side of the L/D curve.
Figure 11 illustrates the L/D during a typical X-24A landing flare.
The approach and flare initiation airspeed of approximately 300 knots
placed the approach conditions well down the speed stable front side of
the L/D curve. During the flare maneuver, the normal acceleration and
lift- coefficient (CL) increased and the flare was accomplished at L/D
values approaching the maximum. As the flare was completed and the normal
acceleration decreased, the CL again dropped down the front side of
the L/D curve. The pilot lowered the landing gear while the vehicle
decelerated and glided essentially parallel to the ground. The C^
gradually increased during the deceleration, and touchdown occurred near
the point of maximum L/D.
The X-24A was not built with independent surfaces for speed brake
usage.Figure 12 presents the effect of using the flap bias feature as
a speed brake.Speed brake was the combined effect of the upper flap bias
position, rudder bias position, and the trim lower flap deflection. Speed
brake is expressed in terms of the upper flap b:as position where 5Ug =
-13 degrees is considered zero speed brakes. This figure illustrates that
the use of the flap bias control system feature provided a very effective
speed brake function (refer to appendix III).
Initial Approach Phase
Figure 13 shows the spread of the patterns flown to runway 18 from
flights 8 to 20. The envelope of the pattern radar data is indicated as
the shaded area. This figure illustrates the wide flexibility available
with this type of pattern. In general, the pilots reached the low key
point with excess energy to compensate for uncertain wind conditions and
possible misjudgement of distance. The excess energy was then expended
through the use of maneuvering flight and/or speed brakes during the
Figures 14 and 15 illustrate the size of the energy footprint at
low key that evolved during the X-24A flight test program. This footprint
was 1.3 nautical miles square and 5,000 feet deep. The scatter in
the data was expected since the preplanned low key point was a non-windcorrected
point and the pilot, who was correcting for wind, was not necessarily
trying to reach this exact point.
A summary of the speed braKe usage is presented in figure 16. This
summary does not include flights 1 to 7, 19, and 20, in which the pilots
were asked not to use the speed brakes (appendix III) . On 79 percent of
the flights in which speed brakes were available, they were used at some
time during the pattern as indicated on figure 16. Most of the usage
was below 15,000 feet MSL at an airspeed between 240 to 280 KCAS which
indicated that the pilot had passed through the 90-degree point of the
pattern and was converging on the preflare aim point. This figure also
illustrates that when speed brakes were used the upper flap bias was
deflected from the normal -13 degrees position to an average position
of between -13 and -24 degrees. Figure 17 is configuration mode 2 with
speed brakes (-25 degrees of upper flap bias).
Figure 18 shows the effect of vehicle configuration on the landing
pattern. In both cases, the winds at altitude were low (average value
of 12 to 15 knots). The patterns started at the same altitude, and the
flight with the lower lift-drag ratio used a higher bank angle and started
laterally closer to the touchdown point since the turn rate is directlyproportional
to the L/D and bank angle, and the turn radius is inversely
proportional to the turn rate (appendix II).
Figure 19 illustrates the glide slope, airspeed, bank angle, and
sink rate encountered during a typical approach in configuration modes
1 and 2 (configuration mode 3 was very similar to mode 2).
Table I shows that the overall average bank angle used with configuration
mode 1 was 44 degrees, and for modes 2 and 3 it was 39 degrees. The
average altitude lost during the turn was 16,300 feet for configuration
mode 1 and 12,700 feet for configuration modes 2 and 3. The flight time
from the actual low key point to rollout on the final approach ranged
between 50 to 87 seconds. This time was a function of the bank angle
(figure 20), configuration and airspeed (figures 2 and 3, appendix III)
and certainly the wind.
Pilots commented that the 180-degree circling approach pattern was
a very comfortable pattern. It was easy to pick up the airspeed during
the turn. At rollout on final, he had to reduce the angle of attack to
pick up or hold the desired airspeed of 300 knots. Bank angles as high
as 60 degrees were used on occasion in the turns without undue concern.
Pilots commented that the pilot task when flying the initial approach in
the X-24A was like flying a similar approach in an F-104 aircraft.
The altitude at rollout on final varied between 1,910 and 6.510 feet
AGL with an average of 3,350 feet for configuration mode 1 and 4,420 feet
for configuration modes 2 and 3 (table II). The time on final, defined
as the time from the attainment of wings level flight (ij> <+10 degrees)
to flare initiation varied between 1.0 and 32.4 seconds depending on the
airspeed and pilot technique. The average time on final was 7.8 seconds
for configuration mode 1 and 16.6 seconds for modes 2 and 3. Although
some of these times were short, one must remember that there still remained
22 to 30 seconds of wings level flight in the flare and deceleration prior
The ranges of average final approach airspeeds, flightpath angles,
and L/D's without speed brakes are shown in table V along with the glide
slope at 4,500 feet MSL. An average airspeed can be computed, because
as illustrated in tables II and III, the airspeed varied little between
rollout-on-final and flare initiation.
The flightpath angle was a parameter associated with the air mass
in which the' vehicle was flying and was calculated from the L/D. The
glide slope was the angle of glide path that terminated on the ground
at a fixed point and was calculated from radar data. The piloting task
associated with final approach was primarily related to establishing
the vehicle on an imaginary glide slope terminating at the aim point.
Figure 21 illustrates that all of the final approaches were made
well down the speed-stable front side of the L/D curve. The L/D data
points in figure 21 were calculated using flight data from actual
approaches and the equations from appendix II. The lines on this figure
were obtained from pushover-pullup data maneuvers (reference 4).
The final approach airspeeds of the first few flights (configuration
rode 1) were generally less than for the balance of the program.
This was the result of low angle of attack handling qualities problems
rather than energy management considerations and is discussed in depth
in reference 5.
Two flights (numbers 17 and 18) were flown in configuration mode 2
using 270 knots as the approach airspeed in preparation for the 270 KIAS
upper flap approaches (configuration mode 3). The maximum approach L/D
(3.44) in the X-24A program occurred on one of these two flights. This
high value of L/D coupled with the low airspeed accounted for the pilot
comment that the approach was very flat and he had plenty of time on
final. The times on final for these two flights were 32.4 and 20.5 seconds,
respectively, as compared to an average of 16.3 seconds for the 300-knot
Upper Flap Approaches
The maximum subsonic performance potential of the X-24A configuration
was achieved with the lower flaps stowed while controlling the
vehicle with the upper flaps in pitch and roll (configuration mode 3).
Two landings were performed in this configuration to demonstrate the
approach and landing capability. Wind tunnel and flight data, reference
6, showed that roll effectiveness with the upper flaps was about twice
that of the lower flaps and the pitch effectiveness was increased by
a factor of 1.5. Because of the increased sensitivity and the resulting
potential longitudinal pilot-induced oscillation (PIG) problem when flying
at an airspeed of 300 KIAS with the upper flaps, a final approach
airspeed of 270 KIAS was selected.
The two upper flap approaches (configuration mode 3) were flown on
flights 19 and 20. These approaches were planned for a 270-KIAS final
approach with the pilot option to explore the pitch and roll control sensitivity
at higher airspeeds. The average approach airspeed for these
flights was 2 87 KIAS because the pilots did explore the pitch and roll
sensitivity at the higher- airspeed.
The mechanization of the flight control system was such that the
flap bias feature could not be used as a speed brake while controlling
the vehicle with the upper flaps. The lack of the speed brake capability
made it difficult for the pilots to adjust their aim and touchdown points
for accurate landings. The pilots felt that the added energy management
provided by the flap bias "speed brake" feature far overshadowed any performance
or handling gualities improvements observed during the upper
flap approaches, thus the mode 2 configuration was used for all subseguent
X-2 4A landings.
Flare and Landing
Figures 22 to 30 present, in terms of flare initiation airspeed, a
summary of various flare parameters, including time from flare initiation
to touchdown, gear extension airspeed, flare initiation altitude, and
airspeed lost during the flare. It was interesting to note that the time
to touchdown (figures 22 and 23) appeared to be independent of the airspeed
at flare initiation. This is further confirmed by examining figures
24 and 25 where it is shown that when the flare initiation airspeed was
high the gear extension airspeed and the touchdown airspeed were also high.
The time for flare was dependent upon the normal acceleration used during
the flare (figure 26) and this accounts for most of the scatter in the
data of figure 23. Figures 22 and 23 also show the time from gear extension
to touchdown. Note, however, that this was the time from pilot action (pulled gear handle)
to touchdown. To get the time from gear-downand-locked to touchdown, one must
subtract the gear extension time of 1.2 to 1.5 seconds.
Figures 27 and 28 show the flare initiation altitude and figures 29
and 30 are the airspeed lost during the flare where the end of flare was
defined as the point at which the sink rate was arrested to 20 feet per
second. These parameters generally increased with flare initiation airspeed,
as might be expected.
The pilots indicated that their primary cue for flare initiation
was altitude. They used an indicated altitude of 800 to 1,200 feet AGL,
which was 300 feet below the actual altitude of 1,100 to 1,500 feet AGL
due to the position error (lag error was insignificant). Most of the
data in figure 28 fell within this altitude band. However, the pilots
also indicated that the determination of their flare initiation point was
a combination of pilot experience and the interred ii ticn of several factors
including sink rate, airspeed, altitude and position relative to the desired
preflare aim point.
The flare timing and the flare altitude as determined from an X-24A
fixed base simulator study (appendix V) are compared with flight data in
figures 22, 23, 27, and 28. In general, the data compared well; however,
the simulator flare timing was'2 to 3 seconds shorter than the flight data.
The reason for this was that the predicted gear-down drag data were 10 to
20 percent higher than the flight data (reference 4).
Figure 31 summarizes the flare altitude versus the time from flare
initiation to touchdown. Also shown on this figure are some boundaries
(dashed lines) from reference 10. T'.J.s reference contains an initial
attempt at defining the minimum and maximum requirements for flare and
landing of a lifting reentry vehicle. Reference 10 divides the flare
timing requirements into two segments: (1) the minimum time from flare
initiation to flare completion and (2) the minimum and maximum float time.
However, the pilots, who have flown the X-24A and other low to medium L/D
lifting body vehicles, feel that it is more meaningful to consider the
total time from flare initiation to touchdown rather than considering the
time from flare initiation to flare completion and the float time separately.
Therefore, the two boundaries from reference 10 have been combined to give
the boundaries (dashed lines) presented in figure 31. These boundaries
are for class IV lifting reentry vehicles with level 2 flying qualities,
where level 2 was defined in reference 10 as "Flying qualities adequate
to accomplish the mission flight phase, but some increase in pilot workload
or degradation in mission effectiveness, or both, exists." Class IV was
defined as: "Light to medium weight, medium to high cross range based
on hypersonic (L/D)max and normal load factor."
The total time from flare initiation to touchdown was considered
because the primary piloting task associated with the flare is to adjust
the sink rate to an acceptable touchdown value, this does not necessarily
require any minimum level attitude float time as required by reference 10.
In the case of the X-24A, the sink rate did not reach a value of five
feet per second (essentially a level attitude) until three to seven
seconds prior to touchdown. This was of no concern to the X-24A pilots,
in fact, they commented that they were impressed with the way the vehicle
handled near the ground. The pilots were more concerned about having
adequate total time from flare initiation to touchdown and energy to accomplish
the tasks of arresting the sink rate, extending the landing gear,
recovering from the gear trim change, and adjusting the sink rate to an
acceptable value prior to touchdown. These factors established a minimum
time from flare initiation to touchdown. A maximum time was dictated by
the fact that it is more difficult to judge the touchdown point as the
flare time increases. The X-24A pilots who have flown in other lifting
body vehicles suggest that a total time from flare initiation to touchdown
of 20 seconds minimum to 30 seconds maximum is adequate with 24 seconds
as an optimum for a medium L/D lifting body vehicle. These minimum and
maximum values are considered independent of the flare altitude and are
plotted as the solid line boundaries in figure 31.
The distance traveled after flare initiation was determined to a
large extent by the energy at flare initiation (figure 32). That is to
say that the touchdown point was largely determined by the airspeed
and altitude at flare initiation. This distance was also a small inverse
function of the average flare normal acceleration and touchdown airspeed
and a weak function of flightpath angle at flare initiation. The winds
would certainly affect this distance also.
Figures 33 and 34 show that the true angle of attack at touchdown,
generally increased with decreasing touchdown airspeed, as expected.
In general the touchdown angles of attack varied from 12 to 16 degrees
in configuration mode 1 and from 9 to 15 degrees in configuration modes
2 and 3. Also shown for reference in these figures is the angle of
attack for 1 g trimmed flight as predicted by the wind tunnel tests. Further
discussion of the wind tunnel and flight gear-down data is contained
in the section titled Landing Gear Characteristics.
Figure 35 was the sink rate at touchdown as recorded by the Honeywell
radar altimeter (appendix I). (There were only a few flights in
which this quantity was valid. The unit usually did not operate, or the
data were not accurately readable.) This parameter appeared to be independent
of touchdown airspeed and was usually less than four feet per
second. Figure 35 also contains a touchdown limit boundary as defined
by the landing gear/carry-through structural limits and the tire limits.
All the flare parameters: flare initiation airspeed, gear extension
airspeed, touchdown airspeed, normal acceleration, and flightpath angle
are complexly interrelated as to their effect on the flare altitude, distance
traveled, flare timing, and airspeed lost during the flare. This
interrelation accounts for the data scatter of figures 22 to 34.
Although the flightpath angle at flare initiation varied between -15
and -25 degrees, depending on the configuration, the pilots commented that
the flares were quite comfortable to perform. The high airspeeds associated
with these steep approaches provided ample "g" capability for the f.\are
and sufficient float time after flare completion/gear extension to provide
for some margin for error. As the flare initiation airspeed decreased,
the determination of the flare initiation altitude became more critical
because less time remained after flare completion for corrections.
The pilots felt that the X-24A had a very mild flare compared to
that of the other lifting body vehicles, and it seemed to flare more
quickly especially in configuration modes 2 and 3. In these configurations
the pilots commented that the vehicle seemed to flare itself, there was
no conscious flare maneuver, but rather just a slight easing back on the
stick. One of the reasons for these comments can be seen by comparing
the lift curve slope (CL alpha ) of the three lifting body vehicles: the X-24A
had a CL of 0.024 per degree (reference 4) and the M2-F2 and HL-10 had
CL 'S of 0.019 and 0.021 per degree, respectively (references 11 and 12).
That is, the "g" available from a small a change in the X-24A was greater
than that for the other two lifting bodies.
In summary, a typical X-24A flare in configuration mode 1 without
the use of the landing rockets was initiated at an airspeed of 2 82 KCAS
and an altitude of 1,820 feet AGL. The airspeed lost was 54.6 knots and
the landing gear was extended at an airspeed of 224 KCAS at 6.8 seconds
before touchdown. Touchdown was at 175 KCAS and an angle of attack of
14.8 degrees,24 seconds after flare initiation. A typical X-24A flare in
configuration mode 2 was initiated at an airspeed of 301 KCAS and an altitude
of 1,350 feet AGL. The airspeed lost was 47 knots and the gear was
extended at 256 KCAS at 9.6 seconds before touchdown. Touchdown was at
186 KCAS and an angle of attack of 11.9 degrees, 24.8 seconds after flare
There were no appreciable differences between the flares in configuration
modes 2 and 3 other than that the time from flare initiation to
touchdown was 30 seconds for mode 3 versus 24,8 seconds for mode 2.
This was due partially to the fact that the average normal acceleration (table
VI) used during the flares in configuration mode 3 (flights 19 and 20)
was less than that used with mode 2. Also, the L/D for configuration mode
3 was one-quarter of an L/D greater than for mode 2.
The lower flare initiation airspeeds of configuration mode 1 were a
result of the low angle-of-attack handling qualities problem on those
flights. The lower flare initiation airspeed combined with the larger
airspeed loss during flare for configuration mode 1 resulted in lower
gear extension speeds than for configuration mode 2 landings. The same
comments hold true for the differences in average touchdown airspeeds for
configuration modes 1 and 2.
Ml of the X-24A landings were reviewed in an attempt to isolate a
repeatable pattern in the piloting technique being used. This information
will be useful in establishing the flare characteristics of future
lifting reentry vehicles. During the X-24A program the flare maneuvers
could be divided into two basic flare techniques:
1. A constant nz. or a constant-rate-of-change-in-angle-of-attack
2. A constant-angle-of-attack flare.
Referring to table VT, it is significant to note that the constant nZb
flare was generally associated with the lower L/D configurations of
mode 1. The constant a flare was generally associated with the higher
L/D configurations of modes 2 and 3.
Figure 36 is a time history of a constant nz- flare in configuration
mode 1. In this type of flare maneuver, the pilot continued to pull back
on the stick throughout the flare as the speed decreased, resulting in a
constant a. while nZfc( remained approximately constant. The average values
of n2, used during this type of X-24A flare were from 1.20 to 1.30 g's.
Figure 37 is a time history of a constant a flare in configuration
mode 2. For this type of flare irumeuver the pilot pulled rapidly to some
initial value of normal acceleration and then held the stick relatively
fixed (constant a) while allowing the normal acceleration to bleed off
as airspeed decreased. The values of the maximum normal acceleration
obtained during these X-24A flares were from 1.3 to 1.6 g's.
Figure 38 is a time history of a constant o flare in configuration
mode 3. The two flares performed with this configuration were both constant
o flares, but there were two distinct a steps.
It should be pointed out that the X-24A flares were not programmed
maneuvers. The pilots did what they intuitively felt was necessary to
break the descent and bring the vehicle to a safe horizontal landing.
The three X-24A pilots used the 3ame basic flare techniques; however,
as shown in figures 23, -25, 28, 30, and 34, pilot A started his
flares at a lower altitude and airspeed than the other pilots. Also,
he landed at slower speeds, hence, the touchdown angles of attack for his
flights were higher than the average, and the time from gear extension
to touchdown was longer.
Landing Gear Characteristics
At or near flare completion isink rate of 20 feet per second and
an altitude of 40 to 130 feet AGL, table III), the landing gear was
extended and the vehicle decelerated along a shallow glide path (less
than 3 degrees) to the touchdown point.
The X-24A had a somewhat severe nosedown trim change associated with
landing gear extension. A fast acting pneumatic landing gear extension
system was used to deploy the gear. The average deployment time from
pilot action until gear-down-and-locked was 1.2 seconds for the main
gear and 1.5 seconds for -t-he nose gear. The landing gear trim change
was produced by both an aerodynamic effect which produced a nose-down
moment and a forward shift in the center of gravity which also produced
a nose-down moment. Reference 7 shows that this center of gravity shift
with gear extension was about 2 inches (0.72 percent c).
Longitudinal trim data for the three configuration modes during the
approach, flare and landing are shown in figures 39 to 41. The lower
flap increment (figure 39) required to compensate for the landing gear
trim change was about 7 degrees or 2 inches of longitudinal stick. The
upper flap increment (figure 41) required to compensate for the landing
gear trim change was 5 degrees or 1.0, inch of longitudinal stick.
Figure 41 contains data from some tests on the X-24A in the full
scale wind tunnel at NASA Ames Research Center. The incremental trim
change due to the landing gear was accurately predicted by the full scale
wind tunnel although the slopes were not. The X-24A fixed base simulator
contained this trim change based on the full scale wind tunnel data and
the gear up/down center of gravity measurements. However, an instantaneous
trim change was used in the simulator, whereas flight experience indicated
that there was a delayed trim change. Although the pilots were continually
exposed to this trim change during their training, the first two
pilots were caught by surprise on their first flights. A time delay was
added to the simulator, and briefing of the third pilot by his predecessors
prepared him for the delayed trim change. The pilots felt that even
though they became accustomed to the gear trim change, and it was easily
compensated for, it was still an undesirable feature especially in close
proximity to the ground where the workload was high and other distractions
could easily occur. However, there were occasions when the pilot extended
the gear higher above the ground than usual and in these cases he
knowingly used the closedown trim change tc converge with the ground.
In an effort to minimize the aerodynandc effect, a minor linkage
modification was made to reposition the nose gear door in the extended
position. It is seen in figure 40 that only a small change was realized,
1 or 2 degrees less flap required. This modification is discussed in
The fixed upper flap settings and center of gravity location for the
approach and landing on flights 1 through 7 were selected such that the
lower flaps were used for pitch and roll control throughout the flight,
thus avoiding the crossover region4 (configuration mode 1, figure 39).
Due to a certain amount of hysteresis in the control system linlto-re, there was a deadband of 2 degrees
of control surface frovsl in conjunction with the crossover from the lower to the upper flops (reference 8).
Test maneuvers were performed at altitude during these flights both in
the crossover region and using only the upper flaps for control without
adverse effects .Starting with flight 8 the upper flap setting (configuration
mode 2 figure 40) was selected such that the landing gear trim change
would result in a rapid crossover from the lower to the upper flaps, thus
avoiding any extended time in the crossover region.The pilot was not
aware of any obvious change in pitch sensitivity after gear extension
for this mode, although the increased effectiveness of the upper flap is
apparent in figure 40 by comparing the slopes (Se^, 6e\j vs a) before and
after gear extension.The effect was probably masked by the rapid speed
loss which occurred after gear extension.Note also that a comparison of
the slopes before and after gear extension in figures 39 and 41 showed
no trend in pitch sensitivity that could be attributed to the landing
The gear-down data points shown in figures 39 through 41 encompassed
the time span from gear extension to the instant of touchdown. A comparison
of slopes of the gear-up/gear-down data showed no obvious trend in
pitch sensitivity that could be attributed to ground effect.This is
consistent with the.limited wind tunnel results of reference 13 in which
tests were accomplished with a landing gear and ground board.These
wind tunnel tests could not be compared directly since the landing gear
and gear doors were not the same as those on the vehicle.
The pilots were impressed with the good vehicle response close to
the ground with the landing gear extended. They commented that it was
easy to make small corrections in the vehicle attitude while decelerating
to the touchdown point. The exceptions to this were the few landings
when a strong ground crosswind existed (see the section titled Effect of
Winds and Turbulence in the Approach Pattern and Flare).
After the landing gear was extended the vehicle continued to decelerate
to the touchdown point. An interesting phenomenon brought about
by the gear trim change was the difference in deceleration before and
after gear extension for configuration modes 1 and 3. The nose-up trim
requirement at gear extension resulted in a reduction in trim drag and
a reduction in lift-due-to-the-tail for mode 1 (lower flap closed, reduced
base area). This, in combination with the angle-of-attack increase
required at gear extension to compensate for the reduction in lift-dueto-
the-tail resulted in the fact that the L/D (figure 42) and the deceleration
did not change appreciably at gear employment in configuration
mode 1. In configuration mode 3, the same trim requirement resulted in
an increase in trim drag and a reduction in lift-due-to-the-tail (upper
flap opened, increased base area). In this case, the L/D decreased with
the result that a signficant increase in deceleration occurred at gear
deployment for configuration mode 3. This trend is also apparent in the
longitudinal accelerometer values before and after landing gear deployment
(figures 36a and 38a ) and the rate of change of airspeed (figures
36b and 38b and table VII) .
The change in deceleration for configuration mode 2 (figure 37) was
between those of modes 1 and 3, but closer to that of mode 3. The landing
rockets, which were used during 3 of the 7 flights in configuration
mode 1, decreased the deceleration by about 3 knots per second. There
were no pilot comments associated with this difference in deceleration
between the different configurations.Apparently the pilot's attention
was directed to altitude control and sink rate after gear deployment and
he was not aware of, or concerned with, the dxfference in the rate of
A description of the landing rocket system is given in appendix I.
Landing Accuracy and Rollout Distance
This system was available for use at the pilot's discretion to decrease
the rate of sink and extend the gjlide by producing an increase in the
effective L/D. The burn time available varied from 27 seconds for glide
flights to approximately 10 seconds following powered flights using the
XLR-11 rocket engine.
The first pilot to fly the X-24A used the landing rockets during
the first three flights. He elected not to maintain the desired high
energy approach airspeed (300 KIAS) due to handling qualities problems.
He therefore used an approach airspeed o£ 260 to 280 KIAS while relying
on the landing rockets for additional energy during the flare and landing.
After accomplishing several modifications to the control system,
and as the pilot became more familiar with the vehicle response in light
turbulence, landings were performed without the use of the landing
rockets. Also, as the flights progressed and airspeed calibration data
were obtained, it became evident that the X-24A had a larger positive
position error than expected (11 knots at an airspeed of 300 KIAS and a
Mach number of 0.5, reference 4).
Table VIII compares average values of various flare parameters with
and without use of the landing rockets in configuration mode 1. The
average time that the glide was extended by using the landing rockets
was five seconds and would have been greater if the average touchdown
airspeeds with and without the use of the landing rockets had been the
Each pilot specified his intended touchdown point prior to most
flights. This point was usually one and a quarter statute miles beyond
the preflare aim point. Figure 43 shows the frequency distribution of
the longitudinal distance from the intended touchdown point for 19 X-24A
landings. This figure shows that 32 percent of these landings were made
within 250 feet of the intended touchdown point and that all landings
were within 2,000 feet. Although this is a small sample of landings,
these results are similar to the X-15 findings of reference 14 which
covered 94 landings, in which 20 percent were within 250 feet and 84
percent were within 2,000 feet of the intended touchdown point. Reference
15 contains the touchdown accuracy data for the other two lifting body
The data in figure 43 must be qualified by the fact that the pilots
were not always allowed to devote their entire attention to the landing
from low key to touchdown as would be necessary to make an accurate
touchdown. As a result seven flights (asterisked values in figure 44)
were excluded from the statistics. For flights 7, 8, 19, and 20, a new
approach configuration was being evaluated. For flights 17 and 18, the
pilot flew the landing pattern at a specified airspeed of 270 KIAS. On
flight 10 (the first powered flight) the pilot realized he was high in
energy, but expressed concern that if he used extensive «peed brakes he
might lose the chase aircraft; therefore, he decided to forego the
accuracy landing. The pilots found that an effective speed bi;ike was
an absolute necessity to consistently make accurate landings.
Pilot A stated that his tenth flight in the X-24A (flight .12) was
the first flight in which he was able to devote his entire attention to
the approach and landing with the results that his touchdown occurred 197
feet short of the intended touchdown point. Likewise for pilot B, his
eighth flight in the X-24A (flight 21) was the first one in which he
really concentrated on achieving a touchdown point (although there was
no touchdown marker), and for that flight, the touchdown occurred 3 85
feet beyond the intended point. It should also be pointed out that the
last eight flights of the X-24A program were flown to a new runway without
a marker at the intended touchdown point.
Planform Landing Effects in the Approach, Flare, and Landing
Figure 45 is a plot of landing rollout distance as a function of
touchdown airspeed and braking. The braking technique varied from
moderate to no braking with or without speed brakes. All.braking began
at approximately the same airspeed, 100 to 120 KIAS, and was defined in
terms of toe pressure. Light braking meant light toe pressure on the
brakes with periods of no braking, whereas moderate braking consisted of
moderate toe pressure on the brakes. Figure 45 shows an increased rollout
distance with increased touchdown airspeed, as expected.
The task after touchdown was primarily to maintain a desired ground
track (straight down the runway). The stopping distance and/or braking
effectiveness were of academic interest only.
The range of wing loadings experienced during the X-24A approach
and landings was very limited (38.0 to 43.4 pounds per square foot). The
heaviest weight during the approach/ landing occurred on flight 15 (there
were no pilot comments relating to the heavy weight condition) . In
analyzing the actual data it was difficult to differentiate between planform
loading effects, wind effects or off-nominal pattern? because the
pilot compensates for each in the same manner using airspeed, speed
brakes, and bank angle.
Effects of Wind and Turbulence in the Approach Pattern and Flare
The low key point that the ground controller was guiding the pilot
toward was not adjusted for winds. However, there was adequate excess
energy in the standard approach pattern to allow the pilot to visually
compensate for winds using airspeed, bank angle, and/or speed brakes.
Winds aloft were determined from early morning weather balloon data.
These data were made available to the pilot prior to cockpit entry and
were updated just before launch if later balloon data were nvailable.
Also, on the morning of each X-24A flight the pilot, flying in the inflight
performance simulator (F-104), had an opportunity to practice the
X-24A approach and thus experience the effects of wind to be expected
during the actual X-24A flight. Based on this information, the pilot
could adjust the pattern inside or outside of the non-wind-corrected low
key point before reaching the low key altitude. Crosswinds on final
approach and landing were compensated by a combination of crab and bank
The effect of wind and the compensation technique used during
X-24A approaches were similar to those experienced on standard fighter
aircraft with the exception of the turbulence and surface winds. The
response of the X-24A to atmospheric turbulence was somewhat different
than that of most winged vehicles because of .i ts unusually large dihedral
effect. Although all three axes were disturbed by the turbulence, the
pilot was more aware of the lateral response of the vehicle than either
the pitch or yaw. These unexpected lateral upsets were somewhat disconcerting
to the first pilot on his first few flights, especially during
the final approach at high airspeed and below 10,000 feet. Once the
pilot was able to associate these motions with atmospheric turbulence
rather than with a degradation in lateral stability, he was able to ride
through turbulence on subsequent flights without apprehension.
Even through the flare, no difficulty was experienced in compensating
for the winds right up through gear extension and touchdown within
the specified wind limits (see the Ground Handling and Surface Winds
section). The exceptions to this were the few flights in which the surface
crosswind was near the limit value, and gusts and turbulence were
heavier than usual; the pilot experienced several disconcerting lateral
upsets between landing gear extension and touchdown.
Ground Handling and Surface Winds
The initial X-24A configuration incorporated nosewheel steering
for ground handling. A series of taxi tests was conducted to evaluate
ground handling characteristics with and without nosewheel steering engaged.
Vehicle response with nosewheel steering engaged was too sensitive.
The use of differential braking for directional control was concluded
to be adequate for landings on a dry lakebed; consequently, the
nosewheel steering was deleted prior to the first flight of the vehicle.
The primary task after touchdown was to maintain a desired ground
track (straight down the landing runway) . For the landings ir. which no
surface crosswind existed, the desired ground track was maintained through
the use of differential braking and rudder/aileron deflection. The deceleration
technique during rollout varied from moderate to no braking,
arc! on six occasions by opening the upper flaps to full deflection for
maximum aerodynamic braking after touchdown. Pilot comments on ground
handling under conditions of little or no crosswind indicated no significant
vehicle deficiencies and satisfactory results of both brakes and
deflected control surfaces in maintaining the desired ground track.
The surface wind limit used during the X-24A program was an arbitrary
limit based on experience with earlier lifting body programs. This
limit for flights 2 through 6 and 22 was a maximum surface wind of 10
knots or a crosswind component of 5 knots. For the balance of the X-24A
flights, this limit was a maximum surface wind of 15 knots or a crosswind
component of 10 knots.
The surface wind direction and velocity as listed in table III were
obtained from a combination of instruments all located a mile or more
from the touchdown point. In some cases, the wind reading was not taken
until some ti'.me after touchdown. In any case the best available data
were used for the wind values in table III.
Two landings (flights 19 and 24) were accomplished with crosswind
components at or near the maximum allowable. The pilots rated the vehicle
ground handling characteristics as poor on these landings. The vehicle
was blown downwind and on flight 24 despite moderate-to -heavy differential
braking applied concurrently with maximum control surface deflection,
the vehicle came to a stop approximately 2,500 feet laterally displaced
from the touchdown point. The wind direction for this landing (figure
46) produced a surface headwind component. As the vehicle was blown
downwind, the crosswind component tended to increase until the wind was
almost perpendicular to the ground track. The pilot did not have adequate
control or inherent vehicle ground stability to prevent this heading
On flight 19 using the same control techniques, the lateral displacement
was 300 feet. The wind direction for this landing (figure 46)
was such that there was a surface tailwind component. In this case, as
the vehicle was blown downwind, the crosswind component decreased until
the wind was almost all tailwind. As the vehicle slowed, the pilot was
able to gain control and turn to the runway heading once again.
In addition to the poor handling characteristics, the riding qualities
during ground rollout in a crosswind were unpleasant. The vehicle
turned away from the wind and the downwind landing gear strut was compressed.
On flight 19 this compression (the difference between windward
and downwind strut positions) amounted to about 1.5 inches, and on flight
24 it was 2.2 inches.
Visibility During the Approach, Flare, and Landing
Due to vehicle geometry (narrow wheel base and high center of
gravity above the ground, figure 47) the strut compression was experienced
by the pilot as an uncomfortable "heeled over" attitude downwind.
In attempting to maintain ground track against the crosswind, the downwind
strut compression was even greater, aggravating the poor riding
qualities. This "heeled over" attitude was experienced on all of the
lifting body vehicles and the pilots felt that if they had tried to
steer into the wind the vehicle might have overturned. The use of braking
on the windward wheel was net satisfactory because that wheel was
lightly loaded and because of the potential tire blowout problem. The
pilots were reluctant to use the brakes until the vehicle had decelerated
to below 120 knots.
The canopy on the X-24A was an acrylic resin bubble canopy (figure
48) very similar to the canopies on many fighter aircraft. However, the
X-24A canopy/cockpit was.q.iite wide, which limited the over-the-side
The pilots felt taat the X-24A had superior visibility compared to
the other two lifting'body vehicles (M2-F2 and HL-10) and compared it
favorably with the visibility from an F-104. The forward visibility
and depth perception in the X-24A were outstanding, and the pilots never
lost sight of the runway even when landing at the slower speeds. A
pressure suit degraded the visibility slightly primarily because it
restricted the pilot's head movements.
Three pilots, after a suitable period of training, flew the X-24A
lifting body vehicle and found that the unpowered visual approach and
landing task was relatively easy after extensive F-104 low/medium L/D
practice. There were several factors found important when performing
an unpowered low/medium L/D approach and landing.
An effective method of energy management was necessary to
perform accurate landings. The use of the flap bias feature
as a speed brake proved to be quite effective and valuable
during the approach and landing phases. The landing accuracy
during the X-24A program was similar to that observed on
other lifting body vehicles.
The external visibility, and consequently the canopy configuration,
were important in the performance of low/medium
L/D landings. The X-24A visibility was better than that of
either of the two previous lifting bodies. The depth perception
during the flare and landing was excellent, and the
pilot never lost sight of the runway.
Good handling qualities were very important. Following
control system developments early in the test program
the handling qualities of the X-24A in the approach and
landing configuration were excellent. However, lateral
upsets due to turbulence were disconcerting to the pilot,
particularly in close proximity to the ground.
Any large trim changes during approach and landing increased
pilot workload. The trim change associated with biasing the
flaps for speed brakes was minor and easily controlled in the
X-24A. The landing gear trim change was large and somewhat
objectionable to the pilots although it was easily compensated
As with past lifting body programs, the use of an inflight
landing simulator was very valuable. Without this the pilot
experience level would probably have been significantly lower
and possibly inadequate.
An effective method of steering on the ground was necessary
to stay within the limits of the marked lakebed runway.
Without a crosswind, the X-24A had good ground handling and
the desired ground track could be maintained through the use
of differential braking and rudder/aileron deflections. With
a crosswind those methods were not adequate to maintain the
desired ground track.
The total time from flare initiation to touchdown was important
in accomplishing the piloting tasks associated with the
flare and landing.The pilots, combining their experience
in the X-24A and other low to medium L/D lifting body vehicles,
suggest that 20 to 30 seconds is adequate to arrest the sink
rate, extend the landing gear and adjust the sink rate to an
acceptable touchdown value.The X-24A flare data fell within
this time band. The energy at flare initiation was very important
in order to give the pilots this necessary time interval.
LANDING GEAR SYSTEM
The landing gear was a conventional tricycle arrangement with cantilevered
air-oil shock struts. The nose gear was similar to the nose gear
used on the North American T-39 . The vehicle originally had nosewheel
steering, but it was removed before the first flight as the result of
the taxi tests which indicated that the steering was too sensitive.
The nosewheels were free-swivelling through a +35 degrees steering angle.
Cams in the nosewheel strut provided centering"of the whyels when the
strut was fully extended. Shimmy was prevented by means of two corotating
wheels . The nose gear was operated by a pneumatic actuator with
an integral lock in the extended position, functioning both as an actuator
and a drag brace.
Each main gear had a single wheel and was similar to the main gear on
the T-38 and the F-5A. The X-24A main gear had a folding drag brace with
a pneumatic actuator connected at the folding joint of the drag brace.
The actuator had an internal lock to hold the drag brace in the extended
Both the nose and main gear retracted aft with the strut in the
fully extended position. The nose gear strut contained a floating piston
to separate the oil and the air in the strut. To ensure reliable gear
extension, a direct mechanical link was provided from the pilot's gear
handle to the gear actuation valve. Gas flow for gear extension was
supplied simultaneously from the emergency helium supply tank and from
the landing gear helium supply bottle. The main and nose gear were
automatically locked in the extended position by positive mechanical
locks. During flight the gear could only be extended; there was no
capability for inflight retraction. There were no cockpit indicators to
tell if the gear was down and locked; however, the chase aircraft pilot
verified that the gear was down.
During the full s-.ale wind tunnel tests at Ames Research Center,
some gear extension tests were run with a dynamic pressure of 108 pounds
per square foot. These tests indicated that the average time from when
the gear handle was pulled until the gear was down and locked was:
left main gear - 1.17 seconds
right main gear - 1.13 seconds
nose gear - 1.4 8 seconds
The full scale wind tunnel tests indicated that 75 percent of the
gear-down aerodynamic pitch trim change was due to the nose gear system
alone and 30 percent of this was due to the nose gear forward flap assembly
(figure 1). The original assembly was mounted so that the angle
relative to the vehicle horizontal reference line (waterline) was 90 degrees
in the gear-down position (figure 1)? however, prior to flight 9
a modification was completed in which this angle was reduced to 4 5 degrees
(figures 2 and 3). The planform area of this assembly was 1,75
This system was used by the pilot if required to stretch the glide
or reduce the rate of descent by an increase in the effective L/D. This
system consisted of two variable thrust Bell monopropellant rocket motors.
They were rated at 500 pounds thrust each and used hydrogen peroxide as
a propellant. These motors were originally developed for the Lunar Landing
Since the hydrogen peroxide was also used to run the gas turbine
for the turbopumps in the primary propulsion system, the burn time of
the landing rockets was a function of the burn time of the main engine.
The maximum landing rocket burn time was 27 seconds as confirmed by
flight test data. The burn time available after a complete burn of the
main rocket was approximately 10 seconds.
The radar altimeter was a Honeywell YG9000D1 electronic altimeter
set which operated on hue Doppler shift principle. The unit contained
outputs for altitude ratJ (+300 and +60 feet per second) in addition to
two altitude ranges (0 to 1,000 feet and 0 to 5,000 feet). The manufacturer's
stated altitude accuracy was +5 feet plus 3 percent of actual
altitude. The altitude accuracy was checked against a ground-based beacon
tracking radar and was found to be within the manufacturer's specifications
No accuracy was given for the altitude rate data; however, a comparison
between ground-based radar and the Honeywell unit revealed that
the radar altimeter had good accuracy at altitude. No comparison could
be made near runway level because of. ground reflection problems with
ground-based radar. The unit read zero on the ground unless there
happened to be a zero shift in telemetry data. The accuracy was a function
of the type of terrain over which the vehicle flew. The data over
the hard, flat lakebed runway looked good. The +60 foot per second output
was not available until flight 15.
A problem with this particular unit was reliability.The unit had
a 2,5 second per cycle search mode and should have acquired a trackable
signal within one cycle.However, this unit often took several cycles
to acquire a signal.This appeared to be a malfunction in this particular
unit and should not have affected its accuracy.The output was not
presented to the pilot and was used only for data purposes.
Table I presents a summary of the parameters used in this study
including their range, sampling rate, and estimated accuracy. The
accuracy values given in table 1 included the processing, sensor, onboard
pulse code modulation (PCM), power supply and the calibration
A standard NASA noseboom was used to measure the total and static
pressure and the angle of attack. The bank angle and pitch angle were
obtained from attitude gyros. The flap, rudder, stick, and landing gear
strut positions came from several control position transducers. The
normal and longitudinal accelerations were measured by sensitive accelerometers
mounted close to the vehicle's center of gravity. The landing
rocket chamber pressure was measured by a pressure transducer. All of
the data were telemetered to the ground station through a PCM data
acquisition system. The raw data were recorded on a magnetic tape at
their respective sampling rates of 20 or 50 samples per seconds (SPS).
Calibrations and corrections were applied to the raw data and the results
were listed in engineering units at 10 SPS.
A surveyor's wheel was used to measure the longitudinal distance
from the intended to the actual touchäown point and the rollout distance.
Ground-based beacon tracking radars were used for space positioning
data, h MPS-19 long-range radar was used to provide real time space
position data in the control room which were used among other things to
guide the pilot to the low key point. Data from an FPS-16 long-range
radar were recorded on magnetic tape for later examination and were used
for the plots of altitude, ground track, and glideslope in this report.
ENERGY MANAGEMENT DURING THE APPROACH
There were three methods available to the pilots for energy management
during the approach:
1. Airspeed modulation
2. Bank angle modulation
3. Flap bias as a speed brake
All of the approaches were made on the high-speed "front side" of
the L/D curve.It was instinctive for t.h& pilot to fly on the front
side of the L/D curve, because when he pushed over, the airspeed increased,
the L/D decreased, the flightpath angle became steeper, and the
range decreased.If he pulled up, the airspeed decreased, the L/D increased,
the flightpath angle became shallower, and the range increased.
However, if the pilot were to fly on the low-speed "back side" of the
L/D curve, any energy management would have been contrary to pilot
instinct because (unlike the front side) an increase in airspeed would
have increased the L/D and thus, the range.
The variation in L/D and flightpath angle due to airspeed is summarized
in figure 1 which was based on flight data reduced to flightpath
angle by means of equation (3) in appendix II. The interesting characteristic
of these curves was that as the L/D versus CL slope decreased
(refer to figure 10 of the body of this report), a small change in liftdrag
ratio (AL/D) produced a larger change in flightpath angle (A-.-1 . For
example, a AL/D of -0.5 represented a Ay of -3.3 degrees for cor.f.-••.• 'ration
mode 1 and the same AL/D represented a Ay of only -1.5 degree, for
configuration mode 3. This factor was of no major consequence to the
X-24A, and the pilots did not comment on it.
By increasing or decreasing the bank angle, the turn rate will increase
or decrease, and the turn radius will consequently decrease or
increase.The turn rate at an altitude of 15,000 feet.is shown in table
II in the main body of this report.This data came from grcund-based
radar tracking.By examining this data it was not readily apparent that
any correlation existed between turn rate, bank angle and L/D.In an
effort to show this correlation, the turn rate was calculated based on
theoretical equations as presented in appendix II.These equations used
L/D data based on flight data. Figures 2 and 3 show the theoretical
turn rate as well as the actual values for configuration moies 1 and 2.
On the X-24A there were no speed brakes per se; however, by biasing
the upper and lower flaps a very effective speed brake was realized.
Speed brake (SSB) was expressed in terms of the upper flap bias position
(6Ug).The zero speed brake position was considered to be the point at
which fillß = -13 degrees .Note that the lower flap position was dictated
by the trim requirements for a particular angle of attack; therefore,
drag effects must also be compared at the same angle of attack.
The rudder bias position, which was a function of upper flap bias
position, figure 4, also contributed to the speed brake effectiveness.
Figure 5 shows the drag increment due to speed brakes. Figure 6 shows
the flightpath angle and range loss as a function of speed brake position.
These figures concurred with the pilot comments that the use of
the flap bias feature as a speed brake was very effective. The trim
change associated with the use of speed brakes was easily compensated
for by the pilots (figure 7).
As a result of the unknowns involved when flying at several upper
flap positions, it was decided that no speed brakes would be used during
the first several X-24A flights. Starting with configuration mode 2,
the speed brakes were used. No speed brakes were available when flying
with the upper flaps (configuration mode 3) because any attempt to bias
the upper flaps would result in a crossover to the lower flaps for control;
this would result in loss of the flight objective to use the upper
flaps for control.
Figure 1 presents a summary of the L/D _, data obtained for the
In addition to the F-104A, the F-104B, C, and D models were available
for use as inflight performance simulators and chase aircraft. There
were no appreciable differences between these models that would affect
the basic unpowered L/D other than the idle thrust engine differences.
The T-38 aircraft had high drag characteristics which made it
suitable for use as a chase airplane, but not as an inflight performance
simulator. During the flare, the T-38 had too great an L/D to accurately
simulate that portion of the flight.
The F-5D aircraft was used both as an inflight performance simulator
and a chase airplane, but there was only one available and it was phased
out early in the X-24A program.
SIMULATION TEST RESULTS
FLIGHT SIMULATED LANDING PATTERNS
The value of inflight performance simulation of the unpowered landing
patterns of low to medium L/D research vehicles was established prior
to the first X-15 flight. This philosophy has continued throughout the
lifting body program. Before the first X-24A flight, extensive inflight
simulation of the predicted X-24A landing pattern was performed using
an F-104. The F-1Ö4 configuration which best matched the L/D values of
the X-24A during approach and landing was takeoff flaps, speed brakes
and landing gear extended. Thrust modulation was used to vary the effective
L/D and to adjust for the reduction in wing loading as fuel was
burned (figure 1,' .
Baseline data for several F-10 4 configurations are included in
appendix IV. Figure 2 is a comparison between an X-24A pattern and
several F-104 simulations of that pattern.
Each pilot made 50 to 60 low L/D landings in the F-10 4 during the
two-week period prior to each X-24A flight including at least five landings
the morning of: the actual X-24A flight.This allowed him to practice
normal and alternate approaches to the primary and alternate runways.
The use of such inilight performance simulation practice permitted a high
level of pilot proficiency to be maintained even though actual lifting
body flights were short and relatively infrequent.It also allowed efficient
utilization of the available data gathering time on any one X-24A
flight since the pilot was not overly preoccupied with the landing task.
FIXED BASE SIMULATION OF THE FLARE AND DECELERATION
Since before the first X-24A flight there has existed a six-degreeof-
freedom, fixed base, hybrid simulator at AFFTC. Prior to the first
flight and throughout the flight test program, the pilots used this
simulator to predict (among other things) the X-24A flare characteristics,
primarily the flare initiation altitude. With this information on
hand, they used the inflight performance simulator (F-104 in high drag
configuration) to simulate the complete X-24A landing pattern, including
After determining the flare techniques being used by the pilots, a
simulator study was made using these flare techniques to observe how well
the fixed base simulator matched the actual X-24A flare characteristics.
Configuration modes 1 and 2 were used for this study.The flare technique
used with configuration mode 1 was a constant 1.3 g rotation to a sink
rate of 20 feet per second(see the section title Flare Technique).
For configuration mode 2, the flare technique was to pull to an initial
normal acceleration of 1.5 g's and then hold the angle of attack constant
until reaching a sink rate of 20 feet per second. The gear extension and
touchdown airspeeds used in the simulator study are in table I. The
simulator flares with configuration mode 1 were done with and without
the use of lending rockets.
1. Flight Planning and Conduct of the X-24A Lifting Body Flight Test
Program, FTC-TD-71-10, Air Force Flight Test Center, Edwards AFB,
California, to be published.
2. X-24A Lifting Body Systems Operation and Performance, FTC-TD-71-13,
Air Force Flight Test Center, Edwards AFB, California, to be published.
3. Flight Measured X-24A Lifting Body Control Surface Hinge Moments
and Correlation with Wind Tunnel Results, NASA TN , to be published.
4. Ash, Larry G., Captain USAF, Flight Test and Wind Tunnel Performance
Characteristics of the X-24A Lifting Body, FTC-TD-71-8, Air Force
Flight Test Center, Edwards AFB, California, June 1972.
5. Handling Qualities of the X-24A Lifting Body, FTC-TD-71-11, Air
Force Flight Test Center, Edwards AFB, California, to be published.
6. Kirsten, Paul W., Wind Tunnel and Flight Test Stability and Control
Derivatives for the X-24A Lifting Body, FTC-TD-71-7, Air Force
Flight Test Center, Edwards AFB, California, April 1972.
7. Retelle, John P., Jr., Captain USAF, Measured Weight, Balance, and
Moments of Inertia of the X-24A Lifting Body, FTC-TD-71-6, Air Force
Flight Test Center, Edwards AFB, California, November 1971.
8. Measured Characteristics of the X-24A Lifting Body Flight Control
System, FTC-TD-71-12, Air Force Flight Test Center, Edwards AFB,
California, to be published.
9. Matranga, Gene J., and Armstrong, Neil A., Approach and Landing
Investigation at Lift-Drag Ratios of 2 to 4 Utilizing a Straight-
Wing Fighter Airplane, NASA TM X-31, August 1959.
10. DiPranco, Dante A., and Mitchell, John F., Preliminary Handling
Qualities Requirements for Lifting Re-Entry Vehicles During Terminal
Flight, AFFDL-TR-71-64, Air Force Flight Dynamics Laboratory, Wright-
Patterson AFB, Ohio, August 1971.
11. Pyle, Jon S., and Swanson, Robert H., Lift and Drag Characteristics
of the M2-F2 Lifting Body During Subsonic Gliding Flight, NASA TMX-
1431, August 1967.
12. Pyle, Jon S., Lift and Drag Characteristics of the HL-10 Lifting
Body During Subsonic Gliding Flight, NASA TN D-62G3, March 1971
13. Wind Tunnel Test of a 20% Scale Model of the Martin SV-5 Configuration
Ninth Series , University of Maryland wind Tunnel Report No.
424, College Park, Maryland, February 1965.
14. Wilson, Ronald J., Statistical Analysis of Landing Contact Conditions
of the X-15 Airplane, NASA TN D-3801, January 19 67.
15. Larson, Richard R., Statistical Analysis of Landing Contact Conditions
for Three Lifting Body Research Vehicles, NASA TN D-670 8,
16. Schofield, B. Lyle, Richardson, David F.; and Koag, Peter C.,
Major USAF, Terminal Area Energy Management, Approach and Lancing
Investigation for Maneuvering Reentry Vehicles Using F-111A and
NB-523 Aircraft, FTC-TD-70-2, Air Force Flight Test Center, Edwards
AFB, California, June 19 70.
17. Herrington, Russell M., Major USAF, et al., Flight Test Engineering
Handbook, AF-TR-6273, Air Force Flight Test Center, Edwa::ds AFB,
California, January 1966