SP-4220 Wingless Flight: The Lifting Body Story






[111] The crash of the M2-F2 left us with no lifting bodies to fly for almost a year. When the M2-F2 crashed in early May 1967, the HL-10 had been a hangar queen for over four months, and it would remain grounded for another eleven months while its aerodynamic problems were fixed before its second flight. Bikle had grounded the M2-F1 permanently, the "flying bathtub" that had launched the lifting-body effort four years earlier now destined to be a museum artifact. Another lifting body was in the works, the Air Force Flight Dynamic Laboratory at Wright-Patterson Air Force Base having a contract with the Martin Aircraft Company of Middle River, Maryland, for designing and building a piloted lifting body originally designated the SV-5P and later known as the X-24A. However, it would be another two years before it was ready to fly.

Despite the setbacks in lifting-body flight testing, competition continued to flourish between the flight-test teams of the NASA/Air Force M2-F2 and the NASA HL-10. With the Air Force and three different NASA sites-Ames on the M2-F2 and Langley on the HL-10, each in conjunction with the Flight Research Center-actively involved on the M2-F2 and in flight operations for the HL-10, the dynamic energy of their interaction could have been destroyed within the multiple organizational channels through which it had to travel. It was amazing to watch these teams cut across NASA and Air Force channels and remain unified, their first allegiance being to their shared lifting-body project.


Rebirth of the M2-F2


The crashed M2-F2 was pathetic-looking, nearly no skin panels without dents or damage. Rather than scrapping the M2-F2, John McTigue had the vehicle sent to Northrop's plant in Hawthorne, California, where Northrop technicians put the battered vehicle in a jig to check alignment, having removed the external skin and portions of the secondary structure, and then removed and tested all systems and parts, an inspection process that took the next two months. Many parts such as valves and tanks were tested at the Flight Research Center's rocket shop. Mean-while, the M2-F2 team tackled the difficult problem of fixing the vehicle's control problems. Over the next 60 days, the NASA Ames team, led by Jack Bronson, gave high priority to wind-tunnel tests for finding that solution. Using a make-shift model of the M2-F2, they tried five different approaches to fixing the problem with elevon adverse yaw.

[112] First, they tried canting the elevon hinge lines so that side force directly on the elevons would give favorable yaw into a turn. This approach failed, because there were still more pressure effects on the vertical fins that offset any favorable pressure on the elevons.

Second, they tried an extra horizontal surface with two elevons attached between the right and left vertical tail tips, putting favorable pressures on the vertical tails that would reverse the yawing moments. This approach was abandoned due to its complexity and to structural problems.

Third, they tried converting the elevons to a bi-plane arrangement with standoffs supporting a second horizontal surface above each elevon so that the original elevons and standoff surface would move as a control unit. This approach was abandoned because it did not produce the favorable pressure gradients they had hoped it would.

Fourth, they tried extending the elevons aft of the body, away from the vertical fins. This approach succeeded in eliminating about half of the adverse yawing moments, although it also became apparent that pressure gradients were being affected upstream near the vertical tails from elevon deflection.

Finally, they tried installing a center fin that would act as a splitter-plate between the right and left elevons, producing side forces that would counter those of the outer vertical fins. For example, following a right roll command by the pilot, the original M2-F2's right elevon trailing edge moved upward. The pressure field on the upper right side of the body would increase due to this deflection, pushing down on the right side of the body. This increased pressure would also push on the inner side of the right vertical tail, pushing the tail to the right and the nose to the left, resulting in adverse yaw. With the center fin installed on the M2-F2, however, this pressure would also push against the right side of the center fin, opposing the adverse yaw effects from the pressure pushing to the right against the right vertical tail and, as a result, canceling the moments of adverse yaw.

Jack Bronson's team at NASA Ames ran wind-tunnel tests on center fins of various sizes. As expected, the larger ones produced more proverse (favorable) yaw than did smaller ones. Meryl DeGeer, the M2-F2 operations engineer at the Flight Research Center, was asked to provide a clearance drawing of the largest vertical fin that would fit under the B-52 pylon. As it turned out, the M2-F2/B-52 adapter could not be used if a center fin were installed on the M2-F2, for it had a large beam running down the center. However, DeGeer and the Northrop designers decided that the HL-10 adapter-with a slight modification- could be used for both vehicles since it had been built to accommodate the center fin on the HL-10. NASA Ames tested the fin shown in DeGeer's drawing, and it worked. The fin not only neutralized the adverse yaw effects but it also produced a small amount of proverse yaw beyond what was needed to cancel adverse yaw.

A conference called by Gary Layton was held at the NASA Flight Research Center, attended by team members from both NASA Ames and the Flight Research Center as well as the Air Force. Due to the wind-tunnel test results, the center fin was unanimously accepted by the attending team members as the way to fix the control [113] problems on the M2-F2. The NASA Ames team then gathered a more complete set of data on the new configuration. The team at the Flight Research Center analyzed the Ames data that showed the elevons to have a small amount of proverse yaw, modified the M2-F2 simulator, and calculated new root-locus characteristics.

Bob Kempel remembers making some root-locus calculations on the old and the new M2-F2 configurations at that time. He found the difference in controllability to be as extreme as the difference between night and day. The new configuration with the center fin had good roll control characteristics with no tendencies for problems in pilot-induced oscillation (PIO). Although Kempel was officially on the HL-10 team at the time, he had a vested interest in the M2-F2 from having done some analysis on it early in its development. Never happy with the lateral control-system design on the original M2-F2, he had aligned himself with the HL-10, which he originally considered the better of the two heavyweight lifting bodies. With the center fin added to the M2-F2, Kempel agreed that the vehicle could become a good flying machine.

As their main mathematical tools in analyzing all motions made by an aircraft during flight, stability and control engineers such as Bob Kempel use La Place transforms, differential equations, and linear algebra. Winged aircraft normally have such typical motions as roll, spiral, and Dutch roll modes. Lifting bodies, on the other hand, can have a unique motion called a coupled roll-spiral mode, which Kempel documented on the M2-F2 in September 1971 in a NASA report entitled, "Analysis of a Coupled Roll-Spiral-Mode, Pilot-Induced Oscillation Experienced With the M2-F2 Lifting Body."1 Kempel explains that the oscillatory coupled roll-spiral mode results from a combination of non-oscillatory roll and spiral modes. When poor roll controls such as the M2-F2 elevons are used, PIO problems result.

The control problems in piloting a lifting body are somewhat like the control problems experienced by a lumberjack in maintaining his balance during the sport of log-rolling, something I know a little bit about from growing up near the logging industry in Idaho. A log is similar to a lifting body in that both are very slippery in a roll, neither having anything like wings that work to resist the rolling motion in water, for the log, or in air currents, for the lifting body. A lumberjack wearing spiked boots has a pair of good controls on the log he's rolling. With constant attention, he can use his spiked boots to control the log's motion. Were the lumberjack wearing instead a pair of ordinary slick-soled shoes, however, he'd have only a pair of poor controls to use. Even with constant attention, he'll eventually lose control of the log he's rolling and, when a wave (analogous to a side gust on a lifting body with poor controls) hits the log, he's going to get very wet.

By 1967, we had flown two lifting-body configurations and were about to fly a third, the M2-F3, the rebuilt M2-F2 with the added center fin. The log-roller analogy....



Schematic showing M2-F1, M2-F2 and the M2-F3. (original drawing by Dale Reed, digital version by Dryden Graphics Office).

Schematic showing M2-F1, M2-F2 and the M2-F3. (original drawing by Dale Reed, digital version by Dryden Graphics Office). [Click here for a larger image]


....applies as well to the differences among the M2-F1, the M2-F2, and the M2-F3. The M2-F1 had the large "elephant ears," the external elevons, that provided good roll control, similar to the lumberjack wearing the spiked boots. The "elephant ears" also served as flat surfaces that slow down, or damp, rolling motions, similar to what would happen if the lumberjack nailed a board to the log. When we went from the M2-F1 to the M2-F2 configuration, we essentially deprived the lumberjack of his spiked boots and removed the board from the log, depriving him of the means for good roll control and damping. When we converted the M2-F2 to the M2-F3 configuration with the center fin, we essentially gave back to the lumberjack his pair of spiked boots, equipping him with the means for good roll control. However, our lumberjack would still have a slick log with no way to slow down (damp) the rolling motions minus the board nailed to the log. What the board nailed to the log provides the lumberjack, a stability augmentation system (SAS) on a lifting body provides the pilot, both helping to damp oscillations and other quick movements.


Birth of the M2-F3


Northrop was enthusiastic about wanting to rebuild the M2-F2 into the M2-F3, strengthening the resolve of the NASA teams to seek approval from NASA Headquarters for continuing the M2 program. NASA Headquarters was reluctant about authorizing more M2 flight tests, but project manager John McTigue was not one to give up easily. Eventually his tenacity succeeded in getting NASA's Office of Advanced Research and Technology to authorize Northrop in March 1968 to continue its [115] "inspection" of the wrecked lifting body. The Northrop team that had built the M2-F2 was still intact, soon to be transferred onto other Northrop projects, so this was the last opportunity we had to have the vehicle rebuilt at low-cost, using the best possible Northrop team for the job.

The lifting-body program was also fortunate to have the help of Fred DeMerritte to keep the effort going at NASA Headquarters. DeMerritte and McTigue had an unwritten agreement that they would proceed quietly at a steady pace until the M2-F3 was ready to fly. McTigue had Bikle to back him up at the Flight Research Center, but DeMerritte was on his own at NASA Headquarters. There was no official authorization for conducting an M2-F3 flight program; however, DeMerritte managed to find a way to continue sending money in incremental amounts to John McTigue to keep the "inspection" going until official approval was obtained.

Just how tense the situation was around DeMerritte at NASA Headquarters in regards to the M2-F2/M2-F3 project is suggested by a conversation that Meryl DeGeer recalls having with DeMerritte on a visit to the Flight Research Center. DeMerritte privately asked DeGeer how things were going on the project. DeGeer said that everything was going fine but that if DeMerritte would give them some more money, they could have the M2-F3 ready to fly all the sooner. DeMerritte asked DeGeer not to push him, for then he'd be forced to say no to the project. It was nearly ten months later-on 28 January 1969-that NASA Headquarters officially announced that the Agency would repair and modify the M2-F2, returning the vehicle to service as the M2-F3, a process that took three years and cost nearly $700,000.

Since there wasn't enough money to contract out all of the work, most of the installation of systems was done by the "Skunk Works" at the Flight Research Center, similar to what had been done with the M2-F1. However, McTigue was able to retain from Northrop five engineers and nearly a dozen technicians to work with the Flight Research Center in fabricating the M2-F3 from the remains of the M2-F2.

Northrop's Fred Erb coordinated the Northrop technical effort while Meryl DeGeer, as NASA's M2-F3 operations engineer, headed up the rebuilding project at the Flight Research Center. Special design problems and parts that had to be manufactured at the Northrop facility were handled through Erb. To keep costs down, as much of the rebuilding as possible was done in the FRC shops. Working from Northrop drawings, LaVern Kelly and Jerry Reedy built new vertical tails for the M2-F3 in the FRC sheet-metal shop, two sheet-metal workers from Northrop at times assisting the shop technicians. The FRC machine shop remanufactured broken parts, including the landing gear. Rocket, fuel-system, and plumbing parts were built in the Center's rocket shop. The FRC aircraft electrical shop put together and installed the vehicle's wiring bundles and electrical systems. Besides the new central fin, a number of internal improvements and other additions were made to the M2-F3. For example, heavy components were moved farther forward, avoiding the need for nose ballast, and small changes in the cockpit area improved visibility and access to the controls. For a cleaner installation, we also rotated the LR-11 rocket engine 90-degrees.

[116] NASA hoped that the new hydrogen-peroxide jet-reaction roll-control system installed on the M2-F3 might be used as well on future lifting-body spacecraft so the pilot could rely on a single control system from orbit to landing, rather than the multiplicity of systems used on such aircraft as the X-15. NASA planned to use the M2-F3 as a testbed for research on the lateral control problems of lifting bodies. If we could eliminate the elevons and rudders, replacing them with reaction rocket controls, we would need only one flap on the bottom of the vehicle for longitudinal trim.

According to Air Force pilot Jerry Gentry, the transformation of the M2-F2 into the M2-F3 changed "some-thing I really did not enjoy flying at all into something that was quite pleasant to fly." 2


HL-10 Returns to Flight


Meanwhile, after fifteen and a half months of wind-tunnel tests, simulation, control-system analysis, and modification of the outer tail fins, the HL-10 was returned to flight. Jerry Gentry flew the HL-10 for the second time on 15 March 1968, launched from 45,000 feet at Mach 0.65. From B-52 launch to touchdown, total flight time was approximately 4.4 minutes.

"I think the whole Center came out to watch this flight," recalls Joe Wilson. "People were standing on the roof, by the planes [on the ramp, and at the edge of the] lakebed. I haven't seen so many observers for a first flight since I've been here. The day was almost absolutely clear and you could see the contrails of the B-52 and [the] chase [planes]. . .two F-104s, one T-38 and the F5D. On [the] drop, everything was O.K., and for a short time you could follow the contrails. The contrails began to pop in and out [of sight], and then were gone from view." 3

The flight plan called for mild pitch and roll maneuvers to 15-degrees angle of attack to evaluate the possibility of control degradation of the sort experienced during the first flight. To assess potential flare characteristics, Gentry executed a simulated landing flare to 2G at altitude.

A camera had been installed on the tip of the vertical fin to provide in-flight photo-graphs of the right inboard tip-fin flap and right elevon. These surfaces had been "tufted" so that a qualitative assessment of the flow field could be made from the photo-graphs. "Tufting" involves taping the ends of short pieces of wool yarn, called "tufts," on suspected problem surfaces of an aircraft for assessing the quality of airflow. If the flow is attached, the tufts lie flat in the direction of the flow across the surface. If the flow is separated, the tufts dance and flutter randomly. Generally, the conclusions following the flight were that the airflow did not separate significantly and [117] consequently that there had been no degradation of control. (When the airflow over control surfaces separates significantly, the control is degraded because it operates aerodynamically.) However, some over-sensitivity in pitch control was observed.

In the debriefing room following the flight, Gentry said the vehicle felt solid. It had no problems in roll sensitivity. It had good longitudinal stability. He also said that, on turning to final approach, flare, and landing, the HL-10 was better than the F-104. He reported that he had put the gear down somewhere after 250 knots and had felt a sharp jolt as the nose gear touched down.

Bob Kempel, Wen Painter, and the rest of the team were as proud as peacocks following the second flight of the HL-10. When someone asked him what kind of problems had occurred on the flight, Kempel said there had been no problems at all, that the flight was a complete success from everyone's point of view. The sensitivity of the longitudinal stick, noted during the flight, was considered acceptable.

The dynamics of the HL-10 in flight proved to be as good as had been indicated by the simulator. After the second flight, Kempel said, the HL-10 attracted the attention of the pilots. "From this point on, all the pilots wanted their shot at flying the HL-10." 4

After pilots establish confidence in a new aircraft and have a little more time to evaluate things, they often change their opinions. The situation was no different with the HL-10. Although no major modifications were required, minor adjustments continued to be made to the HL-10 throughout the remainder of the program. The HL-10 had 35 more successful flights, piloted by NASA's Bill Dana and John Manke and the Air Force's Jerry Gentry and Pete Hoag.


F-104 Used in Pilot Training


During 1968, pilots were becoming very dependent on the ground-based simulator for developing flight procedures and becoming as familiar as possible with the flight characteristics of the lifting bodies. Actual flight experience in the lifting bodies could not be relied upon to provide adequate pilot training because the typical flights were short-five to six minutes for glides, 10 to 15 minutes for rocket flights-and weeks or even months separated flights. Furthermore, for the lifting-body pilots, the first launch off the B-52 hooks was like being thrown into deep water for the first time: you either swim or sink.

In 1957-58, a young research pilot at the Flight Research Center by the name of Neil Armstrong-who, as a NASA astronaut, would later become the first human being to walk on the moon-had conducted a series of flights tests on the NASA F-104 designed to simulate lifting-body flight experience. The technique [118] involved landing an F-104 "dirty," with power off and with flaps, landing gear, and speed brakes extended.5 The pilots found it exciting to fly the F-104 this way, but they had to be careful to avoid losing control of the aircraft. The pilots' choice for preparing for lifting-body flight as well as for flying chase on lifting-body flights was clearly the F-104, a reliable aircraft that had the pilots' full confidence.

The F-104 provided excellent training experience for pilots as preparation for lifting-body flights. The aircraft's high-speed landing gear and large-speed brakes could be used to duplicate lifting-body lift-to-drag characteristics. The aspect ratio of the F-104 was only about 2.46 with a low-speed, clean configuration at a maximum lift-to-drag ratio of approximately 5.7. With the engine at idle, gear and flaps down, and modulation of speed brakes, the lift-to-drag ratio could be made to simulate each of the lifting-body configurations. In this sort of power approach at 170 knots, the lift-to-drag ratio was approximately 2.9. Thus, the lift-to-drag-ratio envelope of the F-104 essentially blanketed the lift-to-drag-ratio values of all of the lifting bodies.

Chasing lifting bodies in the F-104, however, was not totally without risk, as experienced by NASA pilot Tom McMurtry. Chasing one lifting-body flight, McMurtry inadvertently entered an uncontrolled spin. This was serious because the F-104 was not known as an aircraft that could successfully recover from a spin.

The incident occurred at 35,000 feet and 210 knots airspeed with gear down, flaps at takeoff, speed brakes out, and power at idle while McMurtry was maneuvering to join up with the lifting body. Maneuvering into position, McMurtry rolled to 45 degrees of bank and sensed the aircraft starting to slice to the right while in heavy buffet with the nose pitched up. The F-104 went into a spin. One of the other chase pilots, Gary Krier, saw what was happening and radioed McMurtry, calling for full forward stick and full forward trim. The F-104 was in a flat uncontrolled spin directly over the Edwards maintenance and modification hangar, rotating to the right at about 40 to 50 degrees per second.

The aircraft made four or five full turns before McMurtry stopped the rotation by holding full left rudder, neutral aileron, and stick and pitch trim at full nose-down. Recovery from the spin seemed very abrupt, completed at approximately 180 knots and 18,000 feet. The engine did not flame out, and the only configuration change made during the spin was the retraction of the speed brakes. McMurtry held the nose down until the F-104 reached 300 knots and then pulled out at slightly over 4G, the bottom of the pull-out occurring at 15,000 feet.

After the lifting body landed successfully, McMurtry joined the other chase aircraft in the traditional fly-by. Later, during the post-flight debriefing, discussion of [119] the lifting-body mission seemed almost trivial in comparison with McMurtry's description of his experience in the F-104.


From Analog to Digital Computer Simulation


By 1968, flight simulation was becoming an essential part of flight research at the Flight Research Center.6 Even Paul Bikle, who had been somewhat skeptical of the early simulation work with the M2-F1, was beginning to recognize the importance of flight simulation in planning lifting-body flights. Over the three and half years of flight-testing the HL-10, three NASA simulation engineers-Don Bacon, Larry Caw, and Lowell Greenfield- were involved. Air Force Captains John Rampy and John Retelle were also involved with the HL-10 simulator and stability and control.

The HL-10 real-time simulator was primarily an engineering tool, not a pilot-training simulator per se. The simulator was fixed-base-that is, it had no cockpit motion. It had an instrument panel similar to that of the flight vehicle as well as a pilot's control stick and rudder pedals closely approximating those of the actual aircraft. No visual displays were available, all piloting tasks being accomplished by using the instruments. The instrument panel included indicators showing airspeed, altitude, angle of attack, normal acceleration, and control surface position. A three-axis indicator provided vehicle attitude and sideslip information.

Both engineers and pilots used the simulation extensively. Engineers used the simulator for final validation of control-system configuration. Control gearing selection was always difficult with the fixed base. The pilots wanted high sensitivity until they were airborne. Then, the simulation engineers had to decrease the gearing. Modern motion simulators of today have moving cockpits and give high fidelity to control gearing selection.

The simulation was used later to plan each research flight mission, specifying maneuvers and determining flight profiles including Mach numbers, altitudes, angles of attack, and ground track needed for mission objectives to be achieved. Emergency procedures were also practiced on the simulator, inducing various failure modes and selecting alternate landing sites. The pilots were relatively willing subjects once they knew they would be flying the actual mission, and the training paid large dividends. From this information, flight cards were assembled and distributed at crew briefings to all involved personnel, including chase and B-52 pilots, the mission controller, participating flight-research engineers, and NASA and Air Force managers. Coordination was critical to the success of each mission.

The pilots were unanimous in reporting that, once in flight, the events of the mission always seemed to progress more rapidly than they had in the simulator. As a [120] result, engineers and pilots experimented with speeding up the simulation's integration rates, or making the apparent time progress faster. They found that the events in actual flight seemed to occur at about the same rate as they had in the simulator once that simulation time was adjusted so that 40 simulator seconds was equal to about 60 "real" seconds. Only the final simulation planning sessions for a given flight were conducted in this way. In his book At the Edge of Space, Milt Thompson discussed how this difference between simulator seconds and seconds as perceived by pilots in actual flight was first discovered during the X-15 program, the first aircraft research program that made extensive use of simulation in flight planning and pilot training, and resolved by Jack Kolf who originated the concept of fast-time simulation, compressing simulator time to approximate time as it appeared in actual flight. 7

The first simulation of the HL-10 was done with the Pace 231R analog computers then in use at the Flight Research Center. The real capability of the analog computer was its ability to integrate differential equations. Because the equations of motion for the lifting bodies were differential equations-as are all equations of motion for aerospace vehicles-the simulation engineers mechanized them on available analog computers. During the early to mid-1960s, digital computers were primarily used for data reduction, not for real-time simulation. Analog computers were fast, having no problems with cycle time. However, they left much to be desired when it came to mechanizing highly nonlinear functions common to aerodynamic data. Simulation engineers at the Flight Research Center could generate these nonlinear functions on analog computers-but only with great difficulty, patience, perseverance, and a lot of time.

With the aerodynamic data for the modified HL-10, the simulation engineers wanted to mechanize the highest fidelity simulation possible, so they purchased a relatively high-speed digital computer to generate the nonlinear functions. They interfaced the digital and analog computers, using the analog system for the integrations, and moved into the world of hybrid computerization. This approach proved quite successful, allowing them to make fast, efficient changes to the aerodynamic database when they were needed.

Although the program engineers were not aware of it, the simulation engineers-Don Bacon, Larry Caw, and Lowell Greenfield- decided to experiment with moving all of the mathematical computations, including the integrations, to the digital computer. Afterwards, they gave a demonstration of an all-digital, real-time computer simulation. Program engineers Bob Kempel and Wen Painter couldn't tell the difference. Neither could the pilots Bill Dana, Jerry Gentry, Pete Hoag, and John Manke.

The HL-10 program thus achieved another milestone, having successfully made the transition from simulation by analog computer to real-time simulation by digital computer. Today, analog computers have nearly gone the way of the dinosaur. At the Dryden Flight Research Center in the 1990s, all flight simulation is done by using small, high-speed digital computers.


[121] Brown-Bagged Panic: Crashing the Simulator


After the second flight of the HL-10 in March 1968, Jerry Gentry and John Manke alternated as pilots of the vehicle during eight more glide flights in subsonic configuration before the HL-10 was fitted with the rocket engine for supersonic flight in transonic configuration. The aerodynamics became quite different in the transonic, or "shuttlecock," configuration with the rudders moved outboard and the elevon flaps moved upward. Now that the flight envelope of the HL-10 was expanding to supersonic speeds at higher elevations, everyone on the project was a little edgy, including the pilots.

A diligent research pilot, John Manke didn't believe in wasting time when it came to practicing on the simulator for upcoming flights. One day, to practice for his first supersonic flight with the HL-10, he showed up during the lunch hour, bringing his bagged lunch with him. No program engineers were still in the room, and Manke was left alone with the simulator once the simulation engineer left for lunch after loading a data set into the simulator. However, inadvertently, the simulation engineer had loaded the wrong data set-a demonstration set, not used for flight planning, that had directional stability set at zero.

Manke began simulated flight, unaware of the error. Achieving planned altitude for acceleration to supersonic speed, Manke pushed the nose over, toward zero angle of attack, and the vehicle became violently unstable in the lateral direction. The result? Manke "crashed" in the simulator.

To a simulation engineer, "crashing" in simulated flight may seem no big deal, for the engineer may be primarily conscious of the fact that simulated flight is not real flight, but to a pilot who uses a simulator as a pre-stage to actual flight, "crashing" in the simulator can be a major big deal. With no program engineers around at the time, Manke expressed his concerns at once to NASA management.

As a result, project engineers Bob Kempel, Berwin Kock, Gary Layton, and Wen Painter quickly found themselves in the "Bikle barrel," Bikle's wood-paneled executive office, trying to explain to Paul Bikle, Joe Weil, and several other members of the NASA management why they were trying to kill a perfectly good test pilot-a guy all the project engineers liked very much, even if he was from South Dakota.

Kempel recalls feeling a long way from the office's door as a means of escape from this very uncomfortable meeting, a formidable barrier of high-level managers standing between it and the HL-10 project engineers. Once the feeding frenzy had abated, it occurred to the project engineers that the wrong data set must have being used. They explained the problem and followed up with a demonstration in the simulation lab, showing that with the correct flight data set loaded into the simulator, no dynamic instability occurred.


[122] From Rocket Power to Supersonic


On 23 October 1968, Jerry Gentry attempted the first lifting-body powered flight in the HL-10. Unfortunately, the rocket failed shortly after launch. Propellant was jettisoned, and an emergency landing was made successfully on Rosamond Dry Lake located about 10 miles southwest of Rogers Dry Lake within the boundary of Edwards Air Force Base. A few weeks later, on 13 November, John Manke successfully flew the HL-10 for the first time in powered flight.

Five months later, on 17 April 1969, Jerry Gentry flew the X-24A for its first flight. After the B-52 had launched Gentry in the X-24A that day, it was mated with the HL-10 and then launched John Manke in the HL-10 for that vehicle's fifteenth flight. For the first and only time in lifting-body history, two flights in two different vehicles were launched the same day from one mothership.

It's traditional, following a maiden flight, to douse the pilot. After Gentry's first flight that day in the X-24A, during the party at the Edwards Officers' Club, someone decided the swimming pool could be used for Gentry's dousing. However, no one had noticed the pool was nearly empty. Fortunately, Gentry survived his shallow immersion with only a few cracked front teeth.

A few weeks later, on a beautiful spring day in the Mojave Desert, John Manke made the world's first supersonic lifting-body flight in the HL-10 on 9 May 1969. The flight plan for the first supersonic flight of the HL-10 called for launching approximately 30 miles northeast of Edwards AFB, igniting of three rocket chambers, rotating to a 20-degree angle of attack, maintaining that angle of attack until the pitch attitude was 40 degrees, and maintaining that pitch attitude until the vehicle reached 50,000 feet. At that altitude, according to the flight plan, Manke would pushover to a six-degree angle of attack and accelerate to Mach 1.08, afterwards changing angle of attack, turning off one rocket chamber, and maintaining a constant Mach number while gathering data. Landing was planned as a typical 360-degree approach with a landing on Runway 18.

Later, Manke reported that there had been no significant problems during the flight and that generally everything had gone really well. Indeed, the actual flight went almost entirely according to plan, quite different from the ground simulation preparations for this flight. On this historic seventeenth flight, the HL-10 actually rose to an altitude of 53,300 feet and achieved a speed of Mach 1.13, both slightly above the planning figures.

Some special engineering events preceded the first supersonic lifting-body flight. These included completely reviewing the wind-tunnel aerodynamic data and reassessing the predicted dynamic and vehicle controllability characteristics in transonic and supersonic flight regimes. Between Mach 0.9 and 1.0, the data indicated an area of low, and even slightly negative, directional stability at angles of attack of 25.5 degrees and above. Predictions and the simulator showed acceptable levels of longitudinal and lateral-directional dynamic stability at all angles of attack and Mach speeds. The engineering team also prepared a detailed technical briefing that was presented to the NASA and Air Force management teams.

[123] The HL-10, like all lifting bodies, had very high levels of effective dihedral. This characteristic-along with positive angles of attack and acceptable levels of directional stability-ensured lateral-directional dynamic stability almost everywhere in the flight envelope. Before the flight, the HL-10 team demonstrated to project pilot John Manke that the HL-10 would exhibit this dynamic stability even if the static directional stability was zero or slightly negative, provided that the angle of attack did not approach zero.

Bob Kempel recalls that the actual flight was probably not as exciting as the events leading up to it. From what Kempel remembers of the flight, it was relatively uneventful-except for the fact of going supersonic. Nevertheless, in his book On the Frontier, Richard Hallion calls this first supersonic flight "a major milestone in the entire lifting-body program," adding that "the HL-10 [later] became the fastest and highest-flying piloted lifting body ever built." 8


Faster and Higher


About nine months after Manke's first supersonic flight, during the 34th flight of the HL-10 on 18 February 1970, Air Force pilot Major Pete Hoag bested Manke's Mach 1.13, achieving Mach 1.86. Nine days later, on the 35th flight, NASA pilot Bill Dana took the HL-10 to an altitude of 90,303 feet.

Hoag's Mach 1.86 in the HL-10 was, indeed, the fastest speed achieved in any of the lifting bodies. From B-52 launch to touchdown, the flight lasted 6.3 minutes. Except for the Mach number exceeding the preflight prediction, the flight was fairly routine.

The HL-10 had been launched about 30 miles southwest of Edwards AFB, heading 059 degrees magnetic, at 47,000 feet. According to flight plan, all four rocket chambers were ignited immediately after launch. The vehicle was rotated to a 23-degree angle of attack until a pitch attitude of 55 degrees was attained, that pitch attitude held until the vehicle reached 58,000 feet, followed by a pushover to zero G-angle of attack near zero-maintained until the fuel was exhausted. Predicted preflight Mach speed had been 1.66 at 65,000 feet. However, Hoag achieved Mach 1.86 at 67,310 feet.

The fourth NASA research pilot to fly the HL-10, Bill Dana had flown the 199th and last flight of the X-15 in late October 1968, six months later making his first HL-10 glide flight on 25 April 1969. When he took the HL-10 to 90,303 feet on 27 February 1970, Dana not only flew the HL-10 higher than it had ever been flown before, he also set the record for the highest altitude achieved by any lifting body. From B-52 launch to touchdown, the flight lasted 6.9 minutes.

[124] Dana's flight to maximum altitude was launched under the same initial conditions as Hoag's nine days earlier, except that launch was executed 2,000 feet lower (45,000 feet) with pushover 9,000 feet higher (67,000 feet) to a seven-degree angle of attack held to Mach 1.15. Speed brakes were deployed at that altitude and speed, angle of attack then increasing to 15 degrees. According to the flight plan, maximum altitude was to have been reached at this point. What was achieved was Mach 1.314 and an altitude of 90,303 feet. The rest of the flight was fairly routine, except that touchdown was changed from Runway 18 to Runway 23 to avoid high crosswinds.


HL-10: Lift and Drag


For success, any aerospace vehicle must have adequate controllability. The modified HL-10 had very good control characteristics. Equally important to the HL-10's success in the lifting-body program was its ability to generate and control lift, plus its relatively high lift-to-drag ratio in its subsonic configuration. As measured in flight with the landing gear up, the HL-10's maximum lift-to-drag ratio was 3.6, so its best subsonic glidepath angle was approximately -16 degrees (below the horizontal reference).

The HL-10 and the M2-F2 can be compared in terms of their lift-to-drag characteristics, for although the two lifting bodies were considerably different configurations, their missions were similar. Maximum lift-to-drag ratio for the HL-10 was 14 percent higher than for the M2-F2. Although both vehicles had similar lift-curve slopes, the M2-F2 had a much lower angle of attack at a specific lift coefficient than the HL-10. Both vehicles initiated a 300-knot approach at a lift coefficient of approximately 0.15, resulting in a flight path angle of about -25 degrees for the M2-F2 and about -16 degrees for the HL-10 and a landing approach at pitch attitude of about -25 degrees (nose down) for the M2-F2 and about -8 degrees for the HL-10. (The approach flight path angle of commercial airliners in 1990, by comparison, was about -3 degrees.) Never a problem for the lifting-body pilots, the steep approaches for landing were always breath-taking to watch, especially the particularly steep descents of the M2-F2.

At about Mach 0.6, the lift-to-drag ratio of the HL-10 in transonic configuration was approximately 26 percent lower than it was in subsonic configuration. Since lowering the landing gear decreased the lift-to-drag ratio by about 25 percent, the common landing technique with the HL-10 involved flaring in the clean subsonic configuration, then lowering the landing gear in the final moments of flight.


"Dive Bomber" Landing Approaches


After its modification, the HL-10 was often rated by the pilots who flew it as the best flying lifting body in terms of turns and the "dive bomber" landing approaches typical of the lifting bodies. On a rating scale of 1 to 10, with 1 being the highest [125] rating, the average of ratings for the HL-10 was a 2. Each pilot was asked to evaluate various piloting tasks or maneuvers during each of his flights. Following a flight, the pilot then completed a questionnaire involving numerical evaluations as well as comments. Of the 419 numerical ratings given on flights, 43 percent were 2, with 98 percent of the pilot ratings being 4 or better. The best possible rating, a 1, figured on 3 percent of the ratings, while the worst rating received, a 6, showed up in only 0.7 percent of them.

Following the modification, the HL-10 presented no serious problems in piloting. Pilots found it relatively easy to fly, the HL-10 landings being no more difficult than making a similar power-off landing approach in an F-104. To some, the steep and unpowered landing approaches seemed to be mere sport, a daring maneuver of little or no advantage. Often, until they have been apprised of the benefits, spacecraft designers and engineers have failed to appreciate the advantages of these steep, unpowered approaches. Air Force pilot Jerry Gentry, in fact, advocated this type of approach even for the F-104 in normal operations. Pilots found the high-energy, steep, unpowered approach to be safer and more accurate than the recommended low-energy approach for the F-104 because it allowed gentler, more gradual changes in altitude.

What Gentry and other pilots found to be true in the HL-10 and F-104 had been known to be true for many years in terms of accuracy in the old dive bombers, where it was generally accepted that the steeper the dive angle, the greater the accuracy. The approach task in the HL-10 involved positioning the vehicle on a flight path or dive angle to intercept a preflare aim point on the ground, similar to the targeting task of the dive bomber. The difficulty of the HL-10's task was minimized by using a relatively steep approach of -10 to -25 degrees.

There was never a problem in the HL-10 of being short on energy, because the approaches generally were begun well before the peak of the lift-to-drag curve-that is, at high speeds and relatively low angles of attack. Energy was modulated while arriving on the desired flight path by slowing, accelerating, or remaining at the same speed and using the speed brakes to make needed changes in the flight path. Speed brakes are critically important on any aircraft landing with power off, for speed brakes can be used much like a throttle to vary the parameters of the landing pattern. What is more, speed brakes add only minimal weight to the vehicle and require no fuel. The small emergency landing rockets installed on the HL-10 were used only for experimental purposes and during the first flight of the vehicle. On all later flights, the speed brakes were consistently used, instead.

Later in the lifting-body program, many spot landings were attempted in the HL-10 because it was generally believed that unpowered landings on a conventional runway would one day be a requirement, as it is currently with the Space Shuttle. On those spot landing attempts, the average miss distance was less than 250 feet. This degree of accuracy in landing is a benefit of the high-energy, steep, unpowered approach typical of the HL-10 landing body.

[126] Higher speed in the landing approach also provided better controllability of the vehicle. The conventional aircraft landing approach with high power and low speed, by contrast, is much more demanding on a pilot. During the conventional low-speed approach, the aircraft is operating past the peak of the lift-to-drag-ratio curve-that is, at a relatively high angle of attack-where the vehicle's stability, controllability, and handling qualities are degraded and where engine failure can be catastrophic.

Although the pilots thought highly of the HL-10 for its excellent control in turns and during the steep landing approaches, most of the pilots did not like the visibility they had from inside the vehicle. Even though the pilot was located far forward in the HL-10, the canopy had no conventional canopy bulge. What is more, the rails at the lowest extent of the plexiglass canopy were relatively high, providing a sideward field of view depression angle of approximately 16 degrees to the right and somewhat less on the left, due to the canopy defrost duct. Pilots in the HL-10 were supplied routinely with a squirt bottle of water to use in case the flow from the defrost duct wasn't enough to handle the fog of condensation obstructing their view during critical moments of flight.

The plexiglass nose window provided excellent forward vision for navigation and maneuvering for touchdown. Unfortunately, the nose window was lens-shaped and, distorting distance like the wide-angle sideview mirrors on today's cars and trucks, gave the pilots the impression that they were higher off the ground than they really were. After one of his flights in the HL-10, John Manke reported that he had touched down before he wanted to, due to the distorted view out the nose window. Some pilots on their first flights in the HL-10 waited until they were critically close to the ground before they extended the landing gear. Only the accumulation of actual flight experience in the HL-10 alleviated this problem for the pilots.


Mysterious Upsets and Turbulence Response


As one might imagine, all of the lifting bodies possessed some unique aerodynamic characteristics. One of the most unusual is what is called "dihedral effect." On conventional winged aircraft, the "dihedral" is the acute angle between the intersecting planes of the wings, usually measured from a horizontal plane. The "dihedral effect" is essentially the aerodynamic effect produced by wing dihedral that is related to the tendency of a winged aircraft to fly "wings level." It is also the effect which produces a rolling tendency proportional to the angle of sideslip (side gusts). Even though lifting bodies don't have wings, they possess very large amounts of dihedral effect, which means that a very large amount of rolling tendency is generated for small amounts of sideslip, the primary reason why lifting bodies were flown with "feet on the floor"-that is, with pilots deliberately keeping their feet off the rudder pedals. Rudder would induce sideslip, and the lifting bodies would respond primarily with rolls.

[127] Each of the lifting bodies experienced flight through turbulence which caused pilot anxiety out of proportion to the involved "upsets," or uncommanded disturbances of unknown origin. These upsets were so different from upsets as experienced in conventional winged aircraft that the pilots frequently became disturbed when encountering any turbulence in a lifting body. Aerodynamically, the lifting bodies were significantly different from winged aircraft and one might expect them to respond quite differently to turbulence, but what we were experiencing was something entirely new and unknown.

The pilots could not agree on what particular sensations triggered their anxiety, but they said that they often felt on the verge of instability. Early in the lifting-body pro-gram, the pilots reported feeling that the vehicles were going to "uncork" on them. Once the pilots became convinced that there was no real instability and that the vehicle disturbances were caused by turbulence, they rode through the disturbances with little concern.

The gust response of an unwinged vehicle is considerably different from that of winged aircraft. In conventional aircraft, turbulence primarily affects the vertical, felt in the seat of the pants. In a lifting body, turbulence primarily affects the horizontal, producing small amounts of sideslip disturbance, resulting in a high-frequency rolling sensation. This was particularly true at lower elevations where turbulence could be most severe. Following the crash of the M2-F2 in May 1967, the pilots became even more sensitized to upsets close to the ground, the crash of the M2-F2 during landing linked to the rolling motions from such an upset that temporarily disoriented the pilot. In turbulence at low elevations, the pilots felt they might be experiencing some impending dynamic instability in the vehicle, even though the engineers assured them that they were not.

Mysterious upsets occurred at altitude as well, usually during the powered portion of a profile. The pilots found these upsets "spooky." The program engineers hypothesized that these upsets were caused by wind shears. Consequently, on one flight a movie camera was positioned on the ground directly beneath the planned ground track, since the LR-11 rocket motor always left a distinctive white trail of exhaust condensation, or contrail, in any and all atmosphere conditions. Just before launch, the upward-facing camera was turned on to record the launch, powered portion of the flight, and the pilot's radio transmissions. As the pilot flew the powered portion, he called out where the vehicle "felt squirrely" in the lateral direction. Later, playing the film showed that the vehicle had indeed encountered wind shears, as shown by the disturbed contrail, when the pilot had reported that the vehicle "felt squirrely."

Over time and with experience, the pilots came to accept that the turbulence response of the HL-10 was considerably different from that of conventional winged aircraft and that the upsets did not mean that they were on the threshold of dynamic instability. This was new territory in aerospace exploration, one in which the lifting-body pilots and engineers found themselves having to separate the real from the imagined.


[128] Experiments with Powered Landings


After its 35th flight, when all of the major program objectives had been met, the HL-10 was reconfigured for a powered approach and landing study conducted over two flights on 11 June and 17 July 1970. For the study, the LR-11 rocket engine was removed and three small hydrogen-peroxide rockets were installed. The objective was to study shallower glide angles during final approach. Ignited during approach, the rockets reduced the angle of approach from approximately 18 to 6 degrees. The 37th and final flight of the HL-10, piloted (like the 11 June flight) by Pete Hoag, was also the last of the powered approach flights in this study.

The overall results of the study were negative, powered landings having no advantage over unpowered ones for the lifting body. Indeed, shallower powered approaches in the lifting body provided none of the benefits normally obtained in winged aircraft from powered landings. Another conclusion from the study was that the normal approach technique for any space re-entry vehicle-even if equipped with airbreathing engines with go-around capability-should be to operate the vehicle as if it were unpowered, relying on the engines only if the approach were greatly in error. This conclusion proved to be of great influence later in the design of the Space Shuttle, especially the decision not to install landing engines on the Shuttle. Yet much credit for that decision should go to Milt Thompson, especially to his perseverance in campaigning vigorously for unpowered Shuttle landings.

When we total up the flight time for the HL-10 in its 37 flights between 1966 and 1970, we come up with 3 hours, 25 minutes, and 3 seconds. Was that enough time for us to prove the value of the lifting-body concept? We think so, especially every time we watch a Space Shuttle landing.




1 R. W. Kempel, "Analysis of a Coupled Roll-Spiral-Mode, Pilot-Induced Oscillation Experienced With the M2-F2 Lifting Body" (Washington, D.C.: NASA Technical Note D-6496, 1971).

2 The quotation also appears in Wilkinson, "Legacy of the Lifting Body," p. 61.

3 Personal diary of NASA Flight Research Center employee Ronald "Joe" Wilson, entry for March 15, 1968, copy available in the Dryden Flight Research Center History Office.

4 Kempel, Painter, and Thompson, Developing and Flight Testing the HL-10, p. 29.

5 Gene J. Matranga and Neil A. Armstrong, Approach and Landing Investigation at Lift-Drag Ratios of 2 to 4 Utilizing a Straight-Wing Fighter Airplane (NASA High-Speed Flight Station, Edwards, Calif.: NASA TM X-31, 1959).

6 Thompson, At the Edge of Space, pp. 70-71.

7 Ibid.

8 Hallion, On the Frontier, p. 162. See immediately below in the narrative for the details Hallion is summarizing here.