SP-4220 Wingless Flight: The Lifting Body Story





[129] Costing between two and three million dollars and involving 60 NASA employees, the rocket-powered lifting-body programs for the M2-F2, M2-F3, and HL-10 were major undertakings for the Flight Research Center. However, this effort seems small in comparison with the several hundred million dollars being invested by the United States at that time, mostly through the Air Force, in lifting re-entry technology.

In the early 1960s, the Air Force funded several studies within the aerospace industry of winged-vehicle configurations, variable-geometry slender bodies, and high-volume lifting bodies. However, having less confidence in wingless designs, the Air Force committed several hundred million dollars to winged vehicles, most of this money channeled between 1960 and 1964 into two hardware programs, the manned Boeing Dyna-Soar X-20 and the unmanned McDonnell ASSET (Aerothermodynamic/Elastic Structural Systems Environmental Tests) programs.

In 1963, a major shift occurred within the Air Force regarding aerospace concepts, interest waning in the winged-vehicle concept as interest grew steadily in the concept of high-volume lifting bodies. By that year, we had had nearly a year of solid flight experience at the Flight Research Center with the M2-F1 lifting body, and I was spending most of my time developing and selling the supersonic lifting-body program to NASA management. Since November 1960, the Air Force had had the Martin Aircraft Company under contract for developing a full-scale flight-testing program of a lifting re-entry vehicle. By December 1963, Martin had selected the SV-5 configuration, following the results of wind-tunnel tests on various lifting re-entry designs.

A high-volume lifting body, the SV-5 was the brain child of Hans Multhopp, an aerodynamicist at the Martin Aircraft Company. The SV-5 quickly became the centerpiece of a new Air Force program known as START (Spacecraft Technology and Advanced Reentry Tests). Established in January 1964, START consisted of dual programs-the unpiloted PRIME (Precision Recovery Including Maneuvering Entry) and the piloted PILOT (Piloted Lowspeed Tests).

In early 1964, I visited the Martin Aircraft Company to gather information on the SV-5 and possibly gain some support from Martin and the Air Force in convincing NASA management to fund a supersonic lifting-body flight-test program. I met Hans Multhopp, introduced to me as Martin's chief scientist and the designer of the SV-5. A soft-spoken man with a heavy German accent, Multhopp seemed to be highly respected and admired by others in Martin engineering. After a conversation with him about the SV-5, I could understand why he was so highly respected, for his knowledge of aerodynamics and aircraft design was impressive.

A former aeronautical engineer, Multhopp had worked during World War II for the Focke-Wulf Flugzeugbau in Bremen, Germany, first as head of the aerodynamics [130] department and then as chief of the advanced design bureau. One of his projects at Focke-Wulf was designing, in conjunction with Kurt Tank, the Ta-183. Information on the Ta-183 design obtained by the Russians at the end of World War II greatly influenced the design of the Russian Mig-15 jet fighter. The Pulqui-II, a derivation of the Ta-183 design flown in Argentina after World War II, had been built by former Focke-Wulf employees who had fled to Argentina.

Whisked out of Germany at the end of World War II, Multhopp went to work for the British at Farnborough. There, he designed the swept-wing British Lightning fighter, using calculation techniques he had developed. After four years, however, the British found his arrogance intolerable and he was sacked. He then became the chief scientist for the company that eventually became the giant American aviation and space contractor, Martin Marietta.

Multhopp was able to convince Martin management as well as the Air Force that the SV-5 shape was superior to NASA's M2-F3 and HL-10 shapes on the basis of six features. First, the SV-5 was a maneuverable lifting body with no essential surface components that would be destroyed on re-entry from orbit. Second, the vehicle had a hypersonic lift-to-drag ratio of 1.2 or better, permitting a lateral range of 1,000 miles. This feature would enable a recall to any preselected site at least once a day as well as emergency recall to a suitable location from every orbit.

Third, the low-speed aerodynamics of the SV-5 were suitable for making a tangential landing without resort to automatic controls. Fourth, volumetric efficiency was as high as possible, the shape giving as much volume forward as possible for center-of-gravity control. The resulting configuration gave more room up front for the pilot and equipment. The center-of-gravity could then be positioned sufficiently forward to provide adequate vehicle control without resorting to an unstable vehicle with a negative static margin. Fifth, positive camber was included in the body, allowing trimmed lift conditions at lower angles of attack as well as a high subsonic lift-to-drag ratio of about 4.0. Sixth, in regards to pilot visibility, the SV-5 cockpit canopy design was superior to that of the M2-F3 and the HL-10.

My first meeting with Hans Multhopp at Martin in early 1964 also turned out to be my last. After that visit, he seemed simply to disappear from public view. Later, when the X-24A was being flown at Edwards Air Force Base as the final stage of the PILOT portion of the SV-5 program, I was surprised to learn that my Air Force colleagues at Edwards had never even heard of Hans Multhopp. At that time, there was still considerable resentment in this country about using German engineers in American aerospace projects. Consequently, it became the usual practice to keep German engineers at low profile. However, this was not always true. A good example of an exception to this practice was Wernher von Braun, who rose to high rank in NASA in full public view and made a significant contribution to our space program.

The PRIME unpiloted SV-5 program began in November 1964. The Space Systems Division of the Air Force Systems Command gave the Martin Aircraft Company a contract to design, fabricate, and test a maneuverable re-entry vehicle in order to demonstrate whether a lifting body could, in fact, be guided from a straight [131] course and then returned to that course. Martin had already been studying lifting re-entry vehicles for some time-the company had, after all, been in the Dyna-Soar competition-and had invested more than two million hours in lifting-body studies.

Martin Aircraft Company refined the SV-5 design into the SV-5D, an 880-pound aluminum vehicle with an ablative heat shield. The Air Force ordered four of the SV-5D aircraft, which it designated the X-23A. Between December 1966 and mid-April 1967, the Air Force launched three of these vehicles atop Atlas boosters that blasted them at 14,900 miles per hour over the Pacific Ocean Western Test Range toward Kwajalein. The three vehicles performed so well that the Air Force canceled the fourth launch to save money. The PRIME project demonstrated that a maneuvering lifting body could indeed successfully alter its flight path upon re-entry. These tests also conclusively confirmed that lifting bodies were maneuverable hypersonic re-entry configurations.


From SV-5P to X-24A


As an expansion of Martin's PRIME work, the Air Force and Martin derived PILOT-a proposed "low-speed" (Mach 2) research vehicle that the Air Force could test for its supersonic, transonic, and subsonic-to-landing behavior. Martin designated the vehicle the SV-5P.

Colonel Chuck Yeager, then commandant of the Edwards Test Pilot School, had been a fan of the lifting bodies since his flight in the M2-F1. At the time, he had told Paul Bikle that the first lifting body handled well and that he would like to have a few jet-powered versions to use for training future lifting-body pilots. After learning of what Yeager had said, Martin proposed the SV-5J, a low-speed lifting-body trainer powered by a small turbojet, for use by the Air Force's test pilot school at Edwards.

Nothing came of this proposal, although Martin built the shells for two such vehicles and even tried to entice Milt Thompson to fly the SV-5J when it was completed. NASA had no interest in the vehicle, and Thompson was committed to supporting the objectives of the NASA lifting-body program. Calculations showed that the vehicle, because of its high drag and low thrust, would not only have marginal climb performance but would actually be dangerous to fly. Nevertheless, Martin offered Milt Thompson $20,000 if, on his own time, he would simply get the vehicle airborne. Thompson offered to accept Martin's $20,000 if he could get it airborne by simply bouncing the SV-5J a few inches into the air by running it across a two-by-four on the runway. Martin didn't accept Thompson's "flight plan."

Meanwhile, the SV-5P development program was progressing smoothly. In May 1966, the Air Force gave Martin a contract for building one SV-5P. Martin began development under the direction of engineers Buz Hello and Lyman Josephs. About a year later on 11 July 1967, Martin rolled out the SV-5P at its plant in Baltimore, Maryland. The Air Force designated the vehicle the X-24A.


[132] X-24A Crew and Wind-Tunnel Testing


Selecting staff and crew for the X-24A lifting-body project coincided with the winding down of the X-15 program. The X-24A gained experienced flight planners and flight-test engineers from both the X-15 and M2-F2 programs, including NASA's Jack Kolf and the Air Force's Johnny Armstrong, Bob Hoey, Paul Kirsten, and David Richardson. Chief NASA flight planner for the X-15, Jack Kolf became project manager of the program under the direction of John McTigue.

Norm DeMar became operations engineer for the X-24A. His crew included crew chief Jim Hankins; mechanics Chet Bergner, Mel Cox, and James "Catfish" Gorden; inspector LeRoy Barto; avionics technician Ray Kellogg; instrumentation technicians Bill Basko and Jay Maag; and, from Martin Aircraft Company, electrical engineer Bob Moshier and hydraulics and mechanical-systems engineer Jack Riddle. Wen Painter and Sperry Rand's Ron Kotfilm worked on the vehicle's stability and augmentation system.

On 24 August 1967, the X-24A was delivered to Edwards Air Force Base. An experienced lifting-body pilot who had probed the instability boundaries of the M2-F2, Jerry Gentry was assigned as project pilot. Although Paul Bikle and Jerry Gentry were anxious to keep the X-24A on schedule, the vehicle did not fly for the better part of another two years. The vehicle was not released for program activity until 5 October 1967, when DeMar and his crew began preparing the X-24A for wind-tunnel tests at NASA Ames.

Although the X-24A left Edwards on 19 February 1968, wind-tunnel testing at NASA Ames did not begin until 27 February, the extra days at the wind tunnel used to prepare the vehicle with a removable coating to simulate the ablative roughness that would be encountered after the heat of re-entry. Roughness measurements from recovered PRIME vehicles were used in preparing this coating. Afterwards, the X-24A was wind-tunnel tested with two skin conditions, with a clean metal skin and with the rough surface stuck to the skin with an adhesive.

The rough surface seemed to cause a significant reduction in the lift-to-drag ratio for landing, a reduction that would, in turn, reduce the time available for correcting control inputs during actual landings on re-entry from space. These conclusions, along with other aspects of the wind-tunnel test data on the X-24A, were published a year later in a NASA report written by Jon S. Pyle and Lawrence C. Montoya, two engineers at the NASA Flight Research Center, entitled Effects of Roughness of Simulated Ablated Material on Low-Speed Performance Characteristics of a Lifting-Body Vehicle1. However, flight tests at Edwards were planned for the vehicle only with a clean metal skin.


[133] Problems and More Problems


After the X-24A returned to Edwards on 15 March 1968, DeMar and his crew began preparing it for flight. However, problems began to appear that would slow them down. First, since the cockpit instrument panels had not been designed at Martin to be removable for check-out and maintenance, DeMar and his crew had to spend two months installing connectors on all electrical and pressure fittings in the panel. Next, when the hydraulic control system was operated, the actuators started leaking, so they had to change all of the servo valves. During hangar tests of the control system, when runaway control-surface oscillations were put in, structural feedback resulted, eventually traced to its origin through the soft actuator structural mounts. This problem occurred because Martin engineers-in their zeal to avoid having to add weight to the vehicle nose for balance-had designed the X-24A to be very light in its aft end, where the actuators were supported.

Competition had sprung up earlier between the Martin designers of the X-24A and the Northrop designers of the M2-F2, M2-F3, and HL-10. The Martin designers knew that the Northrop designers had had to add either nose ballast or redundant structure in the noses of the Northrop-built lifting bodies to maintain center-of-gravity, and they vowed that they would not do the same in their design of the X-24A lifting body. They claimed that one of the assets of the X-24A shape was that it offered more volume forward for the pilot, allowing heavy equipment to be installed in the nose. However, the Martin designers had been so frugal in weight control that the structure and actuators in the aft end of the X-24A were of minimum size and thicknesses. In fact, the aft end of the X-24A was so light that 140 pounds of ballast had to be added to it to balance the vehicle for flight.

DeMar and his crew had to beef up the structure to eliminate the control- system dynamic feedback encountered in ground tests, and this process delayed the X-24A schedule substantially. According to DeMar, he was called into Bikle's office almost weekly during this time to explain to Bikle and Gentry what was causing the latest delay. Even more delays came about as a result of the new wave of caution and conservatism that had engulfed the Flight Research Center following the crash of the M2-F2 the year before. It had always been a tradition at the Center to have a Flight Review Board made up of engineers and technicians not involved in a project to recommend when a project's aircraft was ready for flight-testing. The Board formed to examine every detail of the X-24A, however, proved to be very picky. Extra tests on systems were needed to assure the Board that the vehicle was flight-worthy, further delaying the schedule.


X-24A Glide-Flights, 1969-1970


After the X-24A finally was declared ready for flight, Jerry Gentry was set to pilot the vehicle in its first glide-flight on 17 April 1969, nearly two years after its roll-out...



Drawing illustrating the concept of shuttlecock stability of the X-24A transonic configuration with extended control surfaces

Drawing illustrating the concept of shuttlecock stability of the X-24A transonic configuration with extended control surfaces (original drawing by Dale Reed, digital version by Dryden Graphics Office).


....by Martin Aircraft Company. Gentry's first glide-flight of the X-24A turned out to be almost as hair-raising and exciting as Milt Thompson's and Bruce Peterson's first glide-flights respectively of the M2-F2 and HL-10.

Milt Thompson had experienced a lateral-directional pilot-induced oscillation (PIO) in the M2-F2 at low angles of attack when he moved the manual rudder-aileron interconnect wheel the wrong way. Bruce Peterson had experienced pitch and roll oscillations in the HL-10, the result of flow separation on the outer vertical fins at high angles of attack. Because of this flow separation, the pitch and roll control in the HL-10 was ineffective. Jerry Gentry faced somewhat different problems on the first flight of the X-24A.

The X-24A was more automatic and complex than either the M2-F2 or HL-10. First of all, roll control on the X-24A could come from either the lower split flap or the upper split flap. Roll control could be shifted from the lower to the upper flap in either of two ways: by scheduling an automatic biasing (shuttlecock configuration) of the upper flap for transonic flight, or by the pilot pulling back on the stick, resulting in retraction of the lower flap and extension of the upper flap.

Eliminating the dependency on the pilot to set the rudder-aileron interconnect manually, the X-24A included an automatic system that changed the interconnect ratio as the angle of attack varied. For its first glide-flight, the X-24A was launched from the B-52 with its upper flap set at 21 degrees upward from the body's upper skin [135] to give the lowest drag during the subsonic glide to landing. This setting meant that all roll control during the flight would come from the lower split flap.

The first flight of any air-launched lifting body is unique. With the vehicle's very rapid rate of descent, the pilot has only about two minutes to evaluate actual flight characteristics and determine that no serious deficiencies exist that could compromise a safe landing. During that same two minutes, the pilot also has to perform enough maneuvers in the aircraft to allow lift-to-drag performance and longitudinal trim to be determined, information that later will be compared with wind-tunnel predictions so that the second flight can be approached with an even higher degree of confidence.

The launch of the X-24A from the B-52 into its first glide-flight was smooth. However, one minute into the flight, the automatic interconnect system failed, causing the interconnect to stick in one position. During the landing approach at two degrees angle of attack and 300 knots, Gentry experiences an uncomfortable lateral-directional "nibbling." He said that the sensation was similar to one he had experienced in the M2-F2 with a characteristic that developed into a severe lateral-directional PIO tendency with large bank-angle excursions. At approximately 1,800 feet above ground, to stop the roll oscillation in the X-24A, Gentry increased the angle of attack to between four and five degrees, decreased airspeed to 270 knots, and used the landing rockets, a successful flare landing without rockets requiring an airspeed of 300 knots.

Just before touchdown, the lower flaps were rate-limited, the maximum surface rate from the actuators being insufficient to follow the large commands from both the roll rate-damper system and the pilot, which were in phase. During the flare, Gentry considered the longitudinal control to be good. However, due to actuator rate-limiting, the rate damper could not be fully effective during periods of surface rate-limiting. The result was that the vehicle's roll-rate excursions reached 20 degrees per second.

Something obviously needed to be changed on the X-24A. Johnny Armstrong, Bob Hoey, the NASA engineers, and the Air Force engineers Captain Charles Archie, Paul Kirsten, Major John Rampy, Captain John Retelle, and Paul Richard-son analyzed the flight data and concluded that the problems with roll oscillation and elevon actuator rate-limiting were caused by the failure of the automatic interconnect system. The poor handling qualities of the X-24A during the final approach were primarily the result of the higher-than-planned rudder-to-aileron interconnect that occurred when the automatic system failed.

Once the interconnect system problem was corrected and with no other changes to the vehicle, Jerry Gentry piloted the X-24A on its second glide-flight. However, the same problem occurred, the lower flaps again becoming rate-limited on the final approach, even though the rudder-aileron system was working properly.

Before the third glide-flight of the X-24A, the program's engineers conducted a considerable investigation by simulator to define the changes needed to improve the vehicle's flying qualities on final approach. Subsequent changes made to the control system included modifying the lower-flap control horns to approximately twice the maximum surface rate, modifying the rudder-aileron interconnect schedule with angle [136] of attack, and increasing the control-stick force gradient and stick-damping in roll. More effective rate-damping gain settings in roll and yaw were defined. Although the X-24A's response to motion in turbulence could not be duplicated adequately in the fixed-based simulator, the X-24A engineering team concluded that the effect of turbulence significantly contributed to the control problem.

Bob Hoey recalls that the most significant cause of the oscillations on the X-24A's first and second glide-flights was "an error in the prediction of the yawing-moment-due-to-aileron for the lower flap. The error was apparently caused by flow interference around the sting in the wind tunnel when the flaps were closed to nearly zero deflection. The flight data showed that the derivative was of opposite sign than predicted. Although we suspected the problem, we didn't measure this correct value until after fl[igh]t 2, when the pilot did some aileron doublets."

In retrospect, Hoey concluded, "we were lucky on fl[igh]t 1." Not only had the interconnect stuck too high at 35 percent but even more proverse was the aileron derivative. These effects were additive, Hoey said. "Later analysis showed that Gentry was well into the predicted PIO region on that approach, and his decision to slow down and use the rockets was a good one!" 2

During Gentry's third glide-flight, he noticed considerable improvement due to the changes in the control system. However, he continued to be concerned about the vehicle's response in turbulence. Gentry did not begin to lose this concern until, during additional glide-flights, he became convinced that the motions he was sensing stemmed from "riding qualities" aggravated by turbulence rather than from any serious deficiency in handling qualities. The increased surface rates of the lower flaps, furthermore, prevented the reoccurrence of the earlier problem with rating-limiting. Nine more glides were made in the X-24A before the vehicle's first powered flight.


X-24A Powered Flights, 1970-1971


By combining much larger fuel tanks with a lighter-weight structure in the X-24A, Hans Multhopp and the other Martin designers theoretically achieved the potential for the X-24A to attain much higher speed and altitude than either the M2-F3 or the HL-10. All of the powered lifting bodies had the same type of rocket engine, the LR-11, with a maximum theoretical vacuum thrust of 8,480 pounds. In structure, the X-24A was nearly 200 pounds lighter than the HL-10 and 700 pounds lighter than the M2-F3. The X-24A also carried about 1,600 pounds more in fuel than did the HL-10 or M2-F3. Fuel-to-vehicle weight ratios for the three powered lifting bodies were 0.45 for the X-24A, 0.35 for the HL-10, and 0.33 for the M2-F3. The X-24A seemed to have the potential for breaking lifting-body speed and altitude records.



View of the X-24A showing eight retracted control surfaces on the aft end of the vehicle in its subsonic, low-drag configuration.

View of the X-24A showing eight retracted control surfaces on the aft end of the vehicle in its subsonic, low-drag configuration. (NASA photo E68 18769)


Bob Hoey, however, felt that the maximum speed of the X-24A would not be greater than what had been achieved already with the other powered lifting bodies. "The reference area of the X-24A was 162 sq[uare] f[ee]t," he explained, compared to 139 square feet for each of the other two powered lifting bodies, "so it was larger with more wetted area. The X-24A also required a larger wedge angle (more drag) for stability at transonic and supersonic speeds. This is a desirable feature while decelerating during an entry, but undesirable when trying to accelerate with a rocket."

Actual X-24A entry, Hoey continued, would use 50 degrees of upper flap and 10 degrees of outward flare on the rudder down to Mach 2, identical to the configuration of the PRIME vehicle that deployed a drogue chute at Mach 2. As speed decreased below Mach 2 in the X-24A, Hoey theorized, the upper flap and rudder bias would begin to program inwardly. "We used 40 degrees of upper flap and 0 rudder as our transonic/supersonic configuration," Hoey said, "a compromise in reduced shuttlecock stability in order to get lower drag and higher speed under power. Our simulation showed that we would only reach about Mach 1.7 for an optimum, full duration burn." 3

Historic accounts including Richard P. Hallion's On the Frontier and The Hypersonic Revolution have suggested the X-24A had few, if any, negative points. [138] However, the X-24A's high reputation rests on the fact that the vehicle was not allowed to be flown in what might have been very uncontrollable flight regimes. Hans Multhopp and his fellow designers at Martin had designed the X-24A exclusively as a re-entry vehicle. It had not been designed to perform well in other situations, including being launched from a B-52, climbing to altitude, and diving to achieve a high Mach speed during rocket burn. 4

The X-24A had very serious angle-of-attack control limitations at transonic speeds. If the pilot increased angle of attack above about 12 degrees, he risked losing roll control due to roll-reversal boundary. If the pilot continued increasing angle of attack in the X-24A to near 20 degrees, wind-tunnel tests and simulations predicted the vehicle would depart in yaw due to lack of directional stability. According to these predictions, at these high angles of attack, neutral longitudinal stability also would occur. The X-24A also had a low angle-of-attack limitation, experiencing roll-reversal and pitch-instability problems at angles of attack lower than four degrees.

Nevertheless, it can be said that the X-24A had no constraints in handling or stability for an optimum, maximum speed boost profile. "Although the stability boundaries were well defined by flight test," Bob Hoey said about the X-24A, this use of flight-testing being fairly traditional by the end of the X-15 and M2 programs, "they DID NOT constrain the optimized trajectory. . . . We had adequate margins on both sides of the boundaries to safely fly an optimized trajectory." 5

Flight research teams for the various lifting bodies always wanted their vehicle to surpass the speed and altitude records of earlier lifting bodies. The less restrictive control boundaries of the HL-10 allowed its pilots to be able to fly more optimum-powered trajectories on speed and altitude missions than were allowed for the X-24A. The X-24A team chose to see the HL-10's speed record as mainly a matter of luck, saying that on its speed mission the HL-10 climbed and accelerated at lower altitudes with a tailwind, then climbed slightly into a jet-stream headwind that increased airspeed and added about 0.2 of its speed record of Mach 1.86. Perhaps partly in jest, the HL-10 team replied that they had planned it that way and that perhaps the X-24A team ought to do the same.

Nevertheless, the wave of caution that engulfed the Flight Research Center following the M2-F2 crash affected flight-planning for several years for the lifting bodies still being flight-tested-the HL-10, the X-24A, and eventually the M3-F3. As a [139] result, much care was taken to avoid crossing any possible out-of-control boundaries. Carefully considered restraint characterized the planning of maximum speed and altitude missions for the X-24A. Flight safety was paramount. Program objectives would be met if the rocket-powered lifting bodies, including the X-24A, could be flown at supersonic speeds near or greater than Mach 1.5 in order to test re-entry glide performance.

As objectivity prevailed, the X-24A team decided not to try to set speed and altitude records for the lifting bodies. Describing the X-24A team effort, Bob Hoey said, "We tried twice to get to the expected burnout point of 1.7 (actually 1.68 on the Flight Request). Both flights resulted in engine malfunctions. The X-24B program had already been approved, so we decided that the benefit of another tenth in Mach number was not worth the added risk to the vehicle and crew." As a result, "we stopped the X-24A program without ever flying a speed profile to burnout." 6

Eighteen powered flights were made in the X-24A between mid-March 1970 and early June 1971. A typical X-24A powered flight lasted just under eight minutes, consisting of a two-and-a-half-minute rocket-powered flight followed by a five-minute glide to landing. The vehicle's speed envelope in Mach number was expanded in successive small steps separated at times by pauses for investigating problems affecting handling. Primary flight objectives were not met on the first five powered flights due to system failures following launch.

Flight planning and crew preparation for the powered flights took consider-ably more time than had been required for the glide flights. Not only was the basic flight plan more complex for powered flight, but a large number of possible deviations had to be planned and practiced in simulation. Each X-24A pilot usually spent over 20 hours in flight simulation in preparing for each flight. Furthermore, actual flight practice in the F-104 was also increased to include landing approaches to as many as five different runways. Each of the three X-24A pilots-NASA's John Manke and Air Force Majors Jerry Gentry and Cecil Powell-performed as many as 60 landing approaches during the two weeks prior to a flight.

Generally, the primary objective of each powered flight was to perform data maneuvers near the planned maximum Mach speed, and this required precise control of the profile. Consequently, data maneuvers were generally limited to the angle-of-attack range required for profile control. To prevent the possibility of large upsetting maneuvers that could compromise the profile, all data maneuvers were done with the stability augmentation system engaged. The capability for individually operating the LR-11 rocket engine's four chambers made it possible to select a reduced thrust level upon reaching the desired test conditions to provide additional data time at quasi-steady flight conditions.

On 19 March 1970, Jerry Gentry piloted the X-24A in its first powered flight, reaching well into the transonic region by achieving a speed of Mach 0.87. After we [140] analyzed the data from the first powered flight, there were two changes. First, the center-of-gravity in the X-24A was moved forward by removing 140 pounds of ballast from the tail. Second, to help reduce longitudinal control sensitivity, the upper flap was biased upward from 30 to 35 degrees above the aircraft's body surface-in effect, opening the shuttlecock.

On the flights that followed, pilots John Manke and Cecil Powell steadily expanded the X-24A's performance envelope. During these flights, to increase directional stability, the shuttlecock was increased, biasing the upper flap upward to 40 degrees above the body surface, and the rudders were moved outboard. To improve handling, we also increased yaw damper gain and the rudder-to-aileron interconnect ratio.

Exactly 23 years after Chuck Yeager's first supersonic flight, on 14 October 1970, Manke piloted the X-24A on its first flight beyond Mach 1, reaching Mach 1.19 at 67,900 feet. Less than two weeks later, Manke simulated a Space Shuttle approach and landing in the X-24A from an altitude of 71,400 feet. On 29 March 1971, Manke reached Mach 1.60 in the X-24A, its fastest research flight. However, the 28th and final research flight of the X-24A on 4 June 1971 was disappointing. Only two of the LR-11 engine's four chambers ignited, limiting the X-24A to subsonic speeds.


1971: X-24A Ready for Space


The only lifting-body configuration completely flight-tested from near-orbital speeds to subsonic landing was the PRIME/X-24A. The unpiloted PRIME vehicle demonstrated hypersonic maneuvering flight from Mach 24 to Mach 2.0, while the piloted X-24A demonstrated maneuvering flight from Mach 1.6 to landing. By 1971, the technology existed for initiating a rapid-turnaround, low-cost, low-risk program that could place a piloted lifting body into orbit, using a Titan II booster from the Gemini program. Had such a program come into being then, it would have resulted in the world's first lifting re-entry to horizontal landing a decade before the Shuttle Orbiter.

The last flight to the moon was to occur in December 1972, leaving two complete Saturn V-Apollo systems unused. One of these rocket-and-spacecraft systems would eventually be used in a joint American/Soviet space effort, the Apollo/Soyuz orbital linkup. However, in 1971, there were still no plans for using either of the two Saturn V systems. A Northrop lifting-body engineer came up with the idea of using the vehicles for launching two lifting bodies into orbit. I thought it was a great idea. So did the NASA lifting-body project manager at the time, John McTigue.

I prepared a briefing for Wernher von Braun, then in charge of the NASA Marshall Space Flight Center in Alabama, who was visiting Paul Bikle at the Flight Research Center. The briefing was about launching two lifting-body/Saturn missions, carrying the HL-10 in the same space where the Lunar Lander had fit. The HL-10 would be modified for space flight with a heat-protective ablative coating to protect its aluminum structure. In the first mission, the HL-10 would be flown unpiloted back to [141] earth from orbit. In the second mission, the HL-10 would be flown back by a pilot on board. Because of its maturity, the X-24A probably would have been a better choice than the HL-10.

What made the concept attractive was the proven safety of the Apollo command module that would be used by three astronauts, one of whom would be the lifting-body pilot, for the launch and climb to orbit. During the first mission, the lifting-body pilot would transfer from the Apollo capsule to the cockpit of the lifting body, conduct pre-re-entry systems checks in the lifting body, and then return to the Apollo capsule. The astronauts would then send the lifting body back to Earth unpiloted for a runway landing. Later, the astronauts would themselves return safely to Earth in the capsule via parachute.

The second mission would follow the successful completion of the first, only this time the lifting body would be flown by the astronaut/pilot back from orbit for a runway landing. If the in-orbit cockpit checks of the lifting body proved to be unacceptable, the astronaut pilot could then simply return to earth with the other two astronauts in the Apollo capsule, as done in the first mission.

In my presentation to von Braun, I used a large Saturn V-Apollo model that I had built from a commercially available plastic kit. I had substituted a model of the HL-10 for the Lunar Lander module and had even devised a model of an extraction arm for placing the lifting body in free orbit. There was enough room in the model for either the M2-F3 or the X-24A, had we chosen one of those vehicles for the mission. However, at the time, I had decided to use a scale model of the HL-10 to show the compatibility of the Saturn V-Apollo with existing lifting bodies.

Wernher von Braun thought it was a fantastic idea. He told Bikle he would prepare the rockets at NASA Marshall if Bikle would prepare a lifting body at the Flight Research Center by adding an ablative heat shield to protect the vehicle's aluminum structure from the heat of re-entry. Imagine how I felt at that moment, if you will. I was sitting in a room with two of my heroes, making plans for the first piloted lifting re-entry from space-many years before the Shuttle.

Of course, I was disappointed when Paul Bikle said "no" to the project, even though I could respect why he had made that decision. He felt my idea was good, but he also believed it was a project beyond his experience and interest. Space was beyond his realm, and he was interested only in aircraft. Paul Bikle and Wernher von Braun had each demonstrated the ability to work outside the bureaucratic process. Together, I had little doubt, they would have made the proposed project a success. And if they had, we might have been able to keep the momentum going in the lifting-body program-all the way to space.

Although we still had another five years of flight evaluation to come on the M2-F3 and X-24B lifting bodies, putting a piloted lifting body into orbit would have been a fitting conclusion to our first seven years of lifting-body flight research. However, voices of support for the Space Shuttle concept were already being heard, voices that all too soon became loud enough to drown out our vocalized advocacy for the lifting-body approach in space applications. Nonetheless, our efforts in the lifting-body program [142] had two very significant influences on the immediate future in terms of spacecraft. First, we established the concept of horizontal landing as feasible for spacecraft recovery. Second, we established the fact that landing unpowered spacecraft with gliding lift-to-drag ratios as low as 3.0 could be conducted safely and routinely.


1969: Shuttle Concept Emerges


It wasn't until 1969-after six years of lifting-body flight at the Flight Research Center-that NASA's top-rank decision-makers and planners decided to switch from parachute recoveries of piloted spacecraft to horizontal landings. Chief of engineering at NASA Johnson Space Center, Max Faget was one of the leading figures who, at the time, was still hanging on to the parachute concept in spacecraft recovery. In fact, it was in 1969, while he was promoting the "Big G" concept for building a big Gemini capsule that could carry 12 astronauts, that he became convinced that the concept of horizontal landing was good and immediately switched sides. Studies began at NASA Johnson Space Center on lifting bodies, delta-wing configurations, and a straight-wing vehicle with a conventional horizontal and vertical tail designed by Max Faget himself. Studies led by Gene Love at NASA Langley evaluated candidates for the Shuttle configuration.

Lifting bodies remained major contenders for the Shuttle configuration until two significant events took place in 1969. The first was the invention of the lightweight silicone tile. The second was the mandate by Congress that the Shuttle design satisfy Air Force as well as NASA requirements, including the Air Force's requirements for hypersonic lift-to-drag ratio and a full-access payload compartment about the size of a railroad boxcar.

The early ablator heat shields, developed for spacecraft such as the Apollo capsule, could be applied directly to lifting bodies with much less weight penalty than when applied to winged vehicles. However, with the invention of the lightweight silicone tile by Lockheed Space Systems (later improved by Howard Goldstein and his team at NASA Ames), winged vehicles constructed of such low-cost materials as aluminum could compete with the lifting bodies as candidates for space. Thin surfaces, such as those found on wings and tails, could be covered with the tiles, adding only minimum weight. Minimum use of the heavier newly-developed carbon-carbon tiles could also protect leading-edge high-heat areas of winged vehicles.

Even though NASA had been granted the responsibility for developing the Shuttle, Congress dictated to NASA that the Shuttle design also had to satisfy requirements of the Air Force, which called for a payload size and cross-range requirements roughly twice those of NASA. The typical hypersonic lift-to-drag ratios of the high-volume lifting bodies that we were flight-testing were between 1.2 and 1.5, which would have served any of the projected NASA missions for hauling people and cargo to and from orbit. However, the Air Force projected greater cross-range capability requiring hypersonic lift-to-drag ratios as high as 2.0, a requirement that made winged vehicles more attractive as Shuttle candidates.

[143] The payload requirement of the Air Force was about 50,000 pounds to low orbit, to be contained in a compartment roughly 15 by 60 feet, or about the size of a railroad boxcar. Easy access to this compartment also required the use of full-size doors that could be opened in space. This requirement narrowed down the potential spacecraft shape to what basically resembled a rectangular box with lifting surfaces (wings and tails) attached to it, plus a rounded nose on the front and rocket motors on the back. Two basic shapes evolved for final consideration: Max Faget's configuration with unswept wing and tail surfaces, and a delta-wing design with a vertical tail attached. Studies continued through 1972, when the delta-wing shape was selected for the Shuttle.


Phoenix Rising: From M2-F2 to M2-F3


In the spring of 1970, two powered lifting bodies were in the air over Edwards Air Force Base and a third would enter flight testing by early June. Very popular with the pilots after its modification, the HL-10 was flown more times than any other of the rocket-powered lifting bodies, its final flight occurring 17 July 1970. Since the spring of 1968, it had been flown 36 times by four pilots-10 times by John Manke, nine times each by Bill Dana and Jerry Gentry, and eight times by Pete Hoag. The X-24A was about halfway through its two-year flight-test program by the spring of 1970, having been flown 25 times-13 times by Jerry Gentry and 12 by John Manke. Only three more flights would be made in the X-24A, all three by Cecil Powell.


The M2-F1 next to Shuttle prototype, Enterprise, showing the comparative sizes of the two vehicles. The Space Shuttle with its delta wings was selected over a lifting-body shape for the first reusable launch vehicle, but later the X-33 employed a lifting-body configuration.

The M2-F1 next to Shuttle prototype, Enterprise, showing the comparative sizes of the two vehicles. The Space Shuttle with its delta wings was selected over a lifting-body shape for the first reusable launch vehicle, but later the X-33 employed a lifting-body configuration. (NASA photo EC81 16288)


[144] Like the mythic phoenix rising from its own ashes, the M2-F3 arose from the wrecked M2-F2 after a nearly three-year "inspection process" of the M2-F2 that had crashed on 10 May 1967. Working closely with the Northrop lifting-body crew in Los Angeles, John McTigue parlayed resources from the shops at the Flight Research Center and Northrop along with about $700,000 from NASA Headquarters for this "inspection process," and by 2 June 1970, an essentially new rocket-powered lifting body- the M2-F3- was ready for its first flight.

McTigue had kept costs down for the M2-F3 by means of several methods. For instance, he appropriated idle X-15 crews during the winter months when the X-15s could not fly because the normally dry lakebeds used for landing experimental aircraft were wet. He also had sheet-metal and machined structural parts made in NASA's shops to Northrop's drawings and specifications and then sent to Northrop's Hawthorne facility for assembly, a uniquely cooperative venture between a government agency and a contractor that involved a most cost-effective use of labor and facilities, keeping expenses to an absolute minimum. McTigue also had the full support of Paul Bikle, a man with a reputation for supporting thrifty approaches in flight research.

Working under the direction of McTigue, Meryl DeGeer had kept the original M2-F2 team intact and involved in the building of the M2-F3. The original M2-F2 crew chief, Bill LePage, and mechanics Jay King and Bill Szuwalski continued on with the M2-F3. Although the M2-F3 resembled the M2-F2 externally, several systems had been modified, relocated, or added. The four-chamber LR-11 rocket engine, for example, was turned on its side so the lower flap could be retracted without having to build a bulge into the shape of the M2-F3's lower flap. Further-more, heavier items were moved forward and lighter items were moved aft to help eliminate nose ballast used in controlling center-of-gravity.

Some people consider the M2-F3 the "purest" lifting-body configuration, for it had no horizontal projections or tail surfaces that could be considered small wings of some sort. The other lifting bodies had canted fins projecting into horizontal and vertical planes. By 1970, we became convinced that any engineering information that we could produce from M2-F3 flight tests would be very valuable to those designing future spacecraft. Consequently, we decided to use the M2-F3 for conducting control-system research.

The first lifting body, the lightweight M2-F1, had used a very basic mechanical control system of pushrods and cables moved solely by the pilot's muscles. There were no power systems such as hydraulics or electric actuators because the pilot didn't need them to fly the M2-F1. Only when the heavyweight lifting bodies came along-starting with the M2-F2 which, fully fueled, weighed nearly ten times as much as the M2-F1- were hydraulic controls necessary to help the pilot move the control systems against the high pressures caused by high-speed flight.

The high speeds of the heavyweight lifting bodies introduced another control problem common to all the heavyweights, the tendency for overshoot or oscillation when the pilot made a control input. Although this problem tended to manifest itself [145] in varying degrees, depending on the configuration, it arose from the high inertia (mass distribution) and low aerodynamic damping of these vehicles. To solve this problem, we added a rate-damping stability augmentation system (SAS) to all of the heavyweight lifting bodies.

Figuratively speaking, the SAS worked like a very fast secondary electronic or robot pilot that shared control with the slower human pilot. Either "pilot" could move the control surface independently. However, since many of our early stability augmentation systems were single-strand or nonredundant, we never allowed the "robot" to have more than 50 percent of the authority on the control system, not trusting it to have more control than the human pilot. We also limited the control authority mechanically to guarantee that the pilot would always have 50 percent or more control in case of electronic failure in the SAS.

The primary task of the SAS was to respond to rate gyros by telling the control surfaces to oppose angular rate movements. We called this process "rate damping" because it slowed or resisted motions of the lifting body. Often, the SAS would oppose the pilot's control inputs, telling the control surfaces to move in the opposite direction to slow down the vehicle motions commanded by the pilot. To keep the pilot and the SAS from this kind of conflict, we designed a special washout circuit for the SAS, allowing the pilot to make normal-but not high-rate-turns.

In transforming the M2-F2 into the M2-F3, we used the basic mechanical portion of the M2-F2's control system. However, we increased authority in the speed brake, modifying the rudders to allow 25 degrees of outboard deflection, and increased aileron deflection from 10 to 20 degrees. Yaw was controlled through the rudder pedals that deflected either of the two rudder surfaces on the outboard side of the two outer vertical fins.

The primary manual control system in the M2-F3 was an irreversible dual hydraulic system. Pitch was controlled by moving the center stick longitudinally, positioning the lower flap. Roll was controlled by moving the same stick laterally, differentially positioning the upper flaps.

By adding a center fin to the M2-F3, we gained true roll control with differential body flaps, no longer having the vehicle's nose moving in the opposite direction from adverse aileron yaw, as had happened with the M2-F2. In essence, we eliminated roll reversal. Even though simulation with the new wind-tunnel data told us that the rudder-aileron interconnect was not needed once the center fin was added to the M2-F3, we kept the manual interconnect control wheel in the cockpit in case we wanted to use it during the flight-test program.

Two vehicle configurations-subsonic and transonic- were used to provide adequate stability at transonic speeds and low drag (that is, an increased lift-to-drag ratio) for approach and landing. For shuttlecock stability at speeds higher than Mach 0.65, the upper body flaps were moved from the average position of 11.8 degrees to 20 degrees upward from the body surface. Outboard biasing of both rudders was used solely as a speed brake-not for transonic shuttlecock stability, as was the case for the HL-10 and X-24A.



Schematic showing how the lifting-body control systems differed. The M2-F3 was a test-bed for four different control systems including reaction controls (original drawing by Dale Reed, digital version by Dryden Graphics Office).

Schematic showing how the lifting-body control systems differed. The M2-F3 was a test-bed for four different control systems including reaction controls (original drawing by Dale Reed, digital version by Dryden Graphics Office). [Click here for a larger image]


[147] Three Experimental Control Systems Added


To add reliability and flexibility to the electronic portion of the control system in the M2-F3, we eventually replaced the original single-strand SAS of the M2-F2 with a triply-redundant Sperry electronic control system similar to the one used in the X-24A. We also added three new control systems to the M2-F3, supplemental to the basic system, using simple rate-damping controlled by the pilot's center stick. We used a second "sidearm" control stick to the cockpit for flying the M2-F3 with the three different experimental control systems. The pilot could always revert back to the basic center-stick control system by throwing a switch on the center stick or on the front panel. We planned to install these systems after the speed and altitude envelope for the M2-F3 had been expanded while using the basic center stick and SAS.

The first experimental control system for the M2-F3 was a rate command augmentation system (CAS) in the roll and pitch axis, which we hoped would improve pilot control by smoothing out the configuration's nonlinear longitudinal aerodynamic characteristics. Conceptually, the CAS differed from the SAS normally used in the lifting bodies. Instead of sharing control of the control surface with the pilot and being capable of independent operation, as was the case with the SAS, the CAS received instructions from the pilot's control stick and then used gyro and other information to actually fly the vehicle for the pilot. For instance, if the pilot wanted a certain pitch or roll rate, he would move the side-stick accordingly. After receiving the electrical signals from the pilot's side-stick, the CAS would fly the M2-F3, avoiding overshoots and oscillations and steadying the vehicle at the angular rate the pilot had indicated by stick position.

With lead and lag electronic networks, the CAS actually could do a better job than the pilot in flying a dynamically unstable aircraft. In 1970, however, we didn't trust electronics the way we do today. We gave the CAS only 50-percent authority, the pilot retaining 50-percent mechanical authority so the pilot could control the M2-F3 with the center stick if a failure occurred in the CAS. Today, high-speed aircraft routinely use command augmentation systems with 100- percent authority that are based on redundant digital computers.

We added another feature to the CAS for maintaining a pilot-indicated angle of attack. The CAS would maintain constant angle of attack when the cockpit hold switch was engaged if the pilot's side-stick was in centered position. The side-stick had a detent switch so that when it was out of center, angle-of-attack hold was disengaged and a rate dependent on stick position was commanded until a new angle of attack was reached. Centering the stick would engage angle-of-attack hold. When angle-of-attack hold wasn't desired, the pilot could turn it off with the cockpit switch and then only rate command was operative. Another switch on the side-stick provided a vernier so that angle of attack could be changed without taking the stick out of detent. The pilot could regain center-stick control with the SAS at any time by disengaging the CAS switch on the instrument panel or on the center stick.

[148] The second experimental system installed in the M2-F3, a reaction-control rocket system, offered potential weight reduction and simplified design for potential future lifting-body re-entry vehicles. Such a vehicle could be greatly simplified if the same rockets used for maneuvering in space could be used for control during landing. Four 90-pound-thrust hydrogen-peroxide rocket motors installed on the base of the M2-F3 were designed to be operated in pairs, providing either rolling or pitching moments for roll and pitch control.

The rockets could be operated only in two states-basic ON or OFF-with no capability for variable thrust. Effective rolling or pitching moments could be achieved only by pulsing the rockets' burn times to produce the desired impulse for changing the vehicle's motion. At first, a pilot operated the rocket system by using a spring-loaded toggle switch on the right console. Later, we replaced the toggle switch with a side-arm controller-obtained surplus from an old World War II formation stick-that enabled the pilot to use his right wrist rather than his fingers to operate the system to produce the necessary "beep-beep" or "bang-bang" motion.

A third experimental system installed in the M2-F3 was a CAS to control the reaction-control rocket system. The ON-OFF (or "bang-bang") scheme of controlling the rockets seemed crude and marginal, so the CAS was modified to control the rockets rather than the aerodynamic control surfaces.


June 1970: Bomb on the Ramp!


The flight-test program for the M2-F3 benefited from the experience gained in the M2-F2, HL-10, and X-24A flight-test programs. Meryl DeGeer served as operations engineer only through the first two glide flights of the M2-F3 and then was reassigned as operations engineer on the newly established YF-12A flight program at the Flight Research Center that involved three of the Lockheed Blackbirds, similar to the SR-71A reconnaissance aircraft. Herb Anderson, who had been operations engineer on the HL-10 through its last flight, took over as M2-F3 operations engineer.

Most of the time, preparations for the M2-F3 flight tests progressed smoothly, methodically, and safely. However, an extremely dangerous incident occurred in June 1970 as the M2-F3 was being prepared for a powered flight following four unpowered glides. While hanging under the B-52's wing, the M2-F3 was being fueled on the ramp. During the fueling operation, crew member Danny Garrabrant noticed liquid was spilling out of the liquid oxygen vent onto the ramp.

Normally during fueling, the liquid-oxygen tank and the water-alcohol fuel tanks in the M2-F3 and other lifting bodies were protected by a "quad valve," a dual-redundant check valve that keeps the fuel from flowing into the liquid-oxygen tank. However, both sides of the valve failed on this occasion, allowing the fuel and liquid oxygen to mix, something that had never happened with any of the other lifting bodies. The mixture in the tanks immediately froze due to the temperature of the liquid oxygen (-270 to -290 degrees Fahrenheit), creating a bomb. The slightest jar could set off a gigantic explosion on the ramp under the fully fueled B-52.



M2-F3 launched from B-52.

M2-F3 launched from B-52. (NASA photo EC71 2774)


At once, Garrabrant sounded the alarm to his crew chief, Bill LePage. Herb Anderson and LePage immediately alerted the Air Force. The area was evacuated. All flights at Edwards Air Force Base were canceled, including all supersonic over-flights, for the jar from a sonic boom could trigger the explosion.

Anderson and LePage then set out to defuse their bomb. Using padded tools and being extremely careful not to drop anything on the M2-F3's tanks, they eventually succeeded, but only after several very long hours of extreme danger to themselves and the aircraft.


Flight-Testing the M2-F3


Project pilot Bill Dana flew the M2-F3 on 19 of its 27 flight missions, including the first three of four glide flights for determining how its character-is-tics compared with those of its predecessor, the M2-F2. Even though he had not flown the M2-F2, Jerry Gentry piloted the M2-F3 on its final glide before its program of 23 powered-flight missions began. Two other pilots made seven of the M2-F3's powered flights, including four by John Manke and three by Cecil Powell.

After the end of the vehicle's flight-test program in late 1972, Bill Dana helped write a pilots' report on the flight characteristics of the M2-F3 that included not only his own observations but also those of John Manke, Cecil Powell, and Jerry Gentry. Published in 1975, this final NASA report on the vehicle's handling qualities entitled [150] Flight Evaluation of the M2-F3 Lifting Body Handling Qualities at Mach Numbers from 0.30 to 1.61 was written by Bob Kempel and Alex Sim as well as Bill Dana.7 This report was based on the pilot ratings for all flights and is he main source for the comments, details, and summarized results that follow.

Beginning 2 June 1965 and ending 16 December 1971, the first 13 of the 27 flight tests were made using only the vehicle's center-stick system, with in-flight maneuvers to evaluate control characteristics with the SAS on and off. Maximum Mach speed for these flights was 1.27. After the thirteenth flight, the M2-F3 was grounded for six months-until 25 July 1972-while the experimental control systems and side-stick were installed for evaluation during the final 14 flights. The last flight occurred on 21 December 1972, the M2-F3 during the course of its flight-test program achieving a maximum speed of Mach 1.61 and altitude of 71,500 feet.


Glide Flights and Landings


During the first half of its flights, in glide and at subsonic speeds, the M2-F3 flew very well with the SAS on. Adding the center fin had made a dramatic change in the configuration, transforming the "angry machine" of the original M2-F2 into the very controllable and pleasant-to-fly M2-F3. The pilots reported that control in both longitudinal and lateral-directional axes was excellent with the rate-damping system (SAS) on. While the M2-F3 proved it could also be flown during glides with the SAS turned off in all axes, vehicle response was very sensitive and the pilots had to exercise great care to keep from over-controlling in both longitudinal and roll axes. According to the pilots, without the SAS, the M2-F3's nose would "hunt" up and down and roll maneuvers were "jerky."

During landings from the glide flights, the M2-F3 demonstrated characteristics that distinguished it from the other lifting bodies. Of the three lifting-body shapes tested, the M2 possessed the lowest subsonic lift-to-drag ratio. This fact did not create traffic-pattern difficulties due to the careful planning that went into each flight to provide sufficient altitude for comfortable landing under both normal and emergency conditions.

The low lift-to-drag ratio, however, did require more of the pilot's attention on final approach and flare than had been needed with the HL-10. Flare speed varied from 260 to 320 knots, but 260 knots proved insufficient for holding the aircraft off the ground while "feeling for the runway." About 290 knots of preflare airspeed gave a reasonable float time. However, the faster the final approach, the more comfortable it was for the pilot. Flare altitude also had to be carefully monitored for the vehicle to come level just above the ground, varying between 600 feet for final approach at 260 knots to 100 feet for 320-knot approaches.

[151] Turbulence response in the M2-F3 resembled that of the HL-10 and X-24A. A side gust would cause a high-frequency roll oscillation that would damp out without pilot input, the type of response caused by the vehicle's excessively high amount of effective dihedral. At first, low-level turbulence would make the M2-F3 pilots apprehensive due to the unusual nature of the vehicle's response. As with the HL-10 and X-24A, however, their apprehension decreased as additional experience showed that the unusual response did not mean the vehicle was on the threshold of divergent lateral oscillation. Nevertheless, we chose not to fly the M2-F3 on days when we expected high turbulence in the atmosphere over Edwards.

Having made sixteen of the X-15 flights, including its last flight, Bill Dana tended to be disappointed with the M2-F3's speed brakes. Spoiled by the X-15's powerful speed brakes, he wasn't impressed with the lesser effectiveness of those on the M2-F3. Dana also did not like the vehicle's large nose-down pitching moment when the speed brakes were applied by outboard biasing of both rudders.

The flat upper deck of the M2-F3 challenged the pilots' visibility, requiring them to switch back and forth quickly between looking over the top side and looking down through the nose window at their feet. The biggest problem with visibility in the M2-F3 was visually judging altitude just before touchdown when the nose was at high angle. Historic accounts claim that fighter pilots during World War II adapted well when they had little or no forward visibility due to the long noses on that era's aircraft, compensating by using their peripheral vision. Using the nose window, especially during approaches to touchdown, the pilots of the M2-F3 adjusted just as successfully to limited forward visibility.


Rocket-Powered Flight


During the vehicle's first rocket-powered flight in 1971, Bill Dana achieved the transonic speed of Mach 0.81. However, indications appeared shortly after launch that the M2-F3 had longitudinal problems transonically. Angle of attack drifted nearly uncontrollably due to a decrease in pitch stability and changes in trim as the Mach number increased.

As speeds were gradually increased on each additional rocket-powered flight, the pilots discovered that the most longitudinal instability occurred near Mach 0.85, when they had difficulty controlling angle of attack. The center-of-gravity was moved forward with ballast added to the nose. Increasing the pitch damper gain, or sensitivity, to its maximum value also helped the pilot steady the vehicle. However, even with these changes, longitudinal stability (pitch control) was only marginally acceptable in the transonic speed range. Consequently, the longitudinal rate-damping system (SAS) was never turned off in this speed range.

In contrast, the pilots rated the roll control of the vehicle at transonic speeds as very good. Just as at subsonic speeds, the M2-F3 could be flown with the roll and yaw damping system (SAS) turned off. However, as it had been in glide flights, the vehicle was very sensitive to roll control, and the pilots had to exercise great care to avoid [152] over-controlling the M2-F3 at transonic speeds with the roll and yaw damping system turned off.

At speeds between Mach 1.0 and 1.6, longitudinal control with the rate-damping system turned on was considerably better than it had been in transonic flight. However, the longitudinal control still wasn't as good as it was at subsonic speeds. We decided, consequently, that the longitudinal rate-damping system (SAS) should not be turned off at supersonic speeds. On the other hand, at supersonic speeds, the pilots felt comfort-able about turning off the lateral-directional rate-damping system (SAS), for roll control was sensitive with this system operating and pilots had to be very cautious to avoid over-controlling in roll.

After the side-stick and experimental control systems were installed in 1973, the final 14 flights of the M2-F3 evaluated them. Generally, the pilots were disappointed in the Command Augmentation System (CAS). Bill Dana had hoped the CAS would improve the vehicle's handling characteristics at transonic speed during the rocket-burning phase. While the CAS did improve the longitudinal control in rate-command mode slightly, it was far from satisfactory. The pilots preferred not to use the angle-of-attack-hold mode, for it did not work well. Furthermore, the CAS did nothing to improve lateral control, already good using only the basic SAS. It seems we had cut costs too much in developing the CAS and had failed to optimize its potential.

The sidearm controller selected for use with the CAS proved to be too rudimentary. One spring in the side-stick provided both force gradient and breakout force. Adjusting one required great care to prevent varying the other. Changing either parameter required disassembling the stick, threatening the integrity of the assembly. We should have located or developed an electric sidearm controller with external and independently adjustable force gradients and breakouts.

The potential for improvement in the CAS was never fully achieved due to the poor physical characteristics of the side-stick plus the system's requirement that the pilot wear a pressure suit, which not only limited pilot mobility but also aggravated the negative effects of the side-stick. Nevertheless, the potential for the CAS was recognized. In spite of its drawbacks, the system was a welcomed addition to the M2-F3.

The ON-OFF, or "bang-bang," rocket reaction-control system was first tried in roll with poor results. Manual control of the rockets was too responsive, resulting in jerky flying. Longitudinal control was not even tried for fear of losing control of the M2-F3.

In the reaction-control system with CAS, the pilot's side-stick was a proportional control with the stick's position commanding an angular roll rate. Tested in-flight, the CAS responded to pilots' input command, firing the control rockets with pulses timed to give the desired results in changing or holding the vehicle's angular rate. The system worked beautifully without moving the aerodynamic control surfaces. Bill Dana rated the system as excellent. The system's quality reflected the level of achievement possible from applying experience with previous systems first developed at the Flight Research Center back in the days before the NACA became NASA, experience that was then applied on the rocket-boosted F-104 zoom aircraft and even later on the X-15 and Lunar Lander Research Vehicle.

[153] A refinement on this rocket-control system eliminated unwanted yaw moments when applying roll control. The system worked almost perfectly in this mode when rockets were needed only to change roll rates. In the longitudinal mode, however, excessive use of the rockets was needed when the vehicle got out of trim by adjusting the longitudinal aerodynamic flap. A further refinement of the system, had we had the time and money to do so, would have been to combine the longitudinal reaction-control rockets with the body's longitudinal flap in an automatic control system.

The M2-F3 flight-test program was almost over, and we were nearly out of money. So we took what we had learned from the M2-F3, wrote our technical reports, and left the potential for application of what we had learned in the hands of the designers of future spacecraft. 8




1 Jon S. Pyle and Lawrence C. Montoya, Effects of Roughness of Simulated Ablated Material on Low-Speed Performance Characteristics of a Lifting-Body Vehicle (Washington, D.C.: NASA TM S-1810, 1969).

2 Typed comments of Robert G. Hoey to Dale Reed in conjunction with his technical review of the original manuscript for "Wingless Flight," Sept. 1993.

3 Hoey, comments to Reed.

4 Richard P. Hallion and John L. Vitelli, "The Piloted Lifting Body Demonstrators: Supersonic Predecessors to Hypersonic and Lifting Reentry," Chapter II: "The Air Force and the Lifting Body Concept," pp. 893-945, esp. p. 922 of Hallion, ed., The Hypersonic Revolution: Eight Case Studies in the History of Hypersonic Technology , 2 vols. (Wright-Patterson Air Force Base, Ohio: Aeronautical Systems Division, 1987), Vol. II; Hallion, On the Frontier, p.164.

5 Hoey comments to Reed.

6 Hoey comments to Reed, underlining in the original.

7 Robert W. Kempel, William H. Dana, and Alex G. Sim, Flight Evaluation of the M2-F3 Lifting Body Handling Qualities at Mach Numbers from 0.30 to 1.61 (Washington, D.C.: NASA Technical Note D-8027, 1975).

8 E.g., Kempel, Dana, and Sim, Flight Evaluation of the M2-F3 and Alex G. Sim, Flight-Determined Stability and Control Characteristics of the M2-F3 Lifting Body Vehicle (Washington, D.C.: NASA Technical Note D-7511, 1973).