Early Aircraft Fly-By-Wire Applications

By the 1950s, fully boosted flight controls were common, and the potential benefits of fly-by-wire were becoming increasingly apparent. Beginning during the Second World War and continuing postwar, fly-by­wire and power-by-wire flight control systems had been fielded in var­ious target drones and early guided missiles.[1114] However, most aircraft designers were reluctant to completely abandon mechanical linkages to
flight control surfaces in piloted aircraft, an attitude that would undergo an evolutionary change over the next two decades as a result of a broad range of NACA-NASA, Air Force, and foreign research efforts.

Beginning in 1952, the NACA Langley Aeronautical Laboratory began an effort oriented to exploring various aspects of fly-by-wire, including the use of a side stick controller.[1115] By 1954, flight-testing began with what was perhaps the first jet-powered fly-by-wire research aircraft, a modified former U. S. Navy Grumman F9F-2 Panther carrier-based jet fighter used as an NACA research aircraft. The primary objective of the NACA effort was to evaluate various automatic flight control systems, including those based on rate and normal acceleration feedback. Secondary objectives were to evaluate use of fly-by-wire with a side stick controller for pilot inputs. The existing F9F-2 hydraulic flight control system, with its mechan­ical linkages, was retained with the NACA designing an auxiliary flight control system based on a fly-by-wire analog concept. A small, 4-inch-tall side stick controller was mounted at the end of the right ejection seat arm­rest. The controller was pivoted at the bottom and was used for both lat­eral (roll) and longitudinal (pitch) control. Only 4 pounds of force were required for full stick deflection. The control friction normally present in a hydromechanical system was completely eliminated by the electrically powered system. Additionally, the aircraft’s fuel system was modified to enable fuel to be pumped aft to destabilize the aircraft by moving the cen­ter of gravity rearward. Another modification was the addition of a steel container mounted on the lower aft fuselage. This carried 250 pounds of lead shot to further destabilize the aircraft. In an emergency, the shot could be rapidly jettisoned to restabilize the aircraft. Fourteen pilots flew the modified F9F-2, including NACA test pilots William Alford[1116] and Donald

L. Mallick.[1117] [1118] Using only the side stick controller, the pilots conducted
takeoffs, stall approaches, acrobatics, and rapid precision maneuvers that included air-to-air target tracking, ground strafing runs, and pre­cision approaches and landings. The test pilots quickly became used to flying with the side stick and found it comfortable and natural to use.18

In mid-1956, after interviewing aircraft flight control experts from the Air Force Wright Air Development Center’s Flight Control Laboratory, Aviation Week magazine concluded:

The time may not be far away when the complex mechani­cal linkage between the pilot’s control stick and the airplane’s control surface (or booster valve system) is replaced with an electrical servo system. It has long been recognized that this"fly-by-wire” approach offered attractive possibilities for reducing weight and complexity. However, airplane designers and pilots have been reluctant to entrust such a vital function to electronics whose reliability record leaves much to be desired.[1119]

Even as the Aviation Week article was published, several noteworthy aircraft were under development that would incorporate various fly-by­wire approaches in their flight control systems. In 1956, the British Avro Vulcan B.2 bomber flew with a partial fly-by-wire system that operated in conjunction with hydraulically boosted, mechanically activated flight controls. The supersonic North American A-5 Vigilante Navy carrier – based attack bomber flew in 1958 with a pseudo-fly-by-wire flight control system. The Vigilante served the fleet for many years, but its highly com­plex design proved very difficult to maintain and operate in an aircraft carrier environment. By the mid-1960s, the General Dynamics F-111 was flying with triple-redundant, large-authority stability and command aug­mentation systems and fly-by-wire-controlled wing-mounted spoilers.[1120]

On the basic research side, the delta winged British Short S. C.1, first flown in 1957, was a very small, single-seat Vertical Take-Off and Landing
(VTOL) aircraft. It incorporated a triply redundant fly-by-wire flight con­trol system with a mechanical backup capability. The outputs from the three independent fly-by-wire channels were compared, and a failure in a single channel was overridden by the other two. A single channel failure was relayed to the pilot as a warning, enabling him to switch to the direct (mechanical) control system. The S. C. 1 had three flight control modes, as described below, with the first two only being selectable prior to takeoff.[1121]

• Full fly-by-wire mode with aerodynamic surfaces and noz­zles controlled electrically via three independent servo motors with triplex fail-safe operation in conjunction with three analog autostabilizer control systems.

• A hybrid mode in which the reaction nozzles were servo/ autostabilizer (fly-by-wire) controlled and the aerodynamic surfaces were linked directly to the pilot’s manual controls.

• A direct mode in which all controls were mechanically linked to the pilot control stick.

The S. C. 1 weighed about 8,000 pounds and was powered by four ver­tically mounted Rolls-Royce RB.108 lift engines, providing a total ver­tical thrust of 8,600 pounds. One RB.108 engine mounted horizontally in the rear fuselage provided thrust for forward flight. The lift engines were mounted vertically in side-by-side pairs in a central engine bay and could be swiveled to produce vectored thrust (up to 23 degrees for­ward for acceleration or -12 degrees for deceleration). Variable thrust nose, tail, and wingtip jet nozzles (powered by bleed air from the four lift engines) provided pitch, roll, and yaw control in hover and at low speeds during which the conventional aerodynamic controls were inef­fective. The S. C.1 made its first flight (a conventional takeoff and land­ing) on April 2, 1957. It demonstrated tethered vertical flight on May 26, 1958, and free vertical flight on October 25, 1958. The first transi­tion from vertical flight to conventional flight was made April 6, 1960.[1122]

During 10 years of flight-testing, the two S. C.1 aircraft made hun­dreds of flights and were flown by British, French, and NASA test pilots. A Royal Aircraft Establishment (RAE) report summarizing flight-test experience with the S. C. 1 noted: "Of the visiting pilots, those from NASA [Langley’s John P. "Jack” Reeder and Fred Drinkwater from Ames] flew the aircraft 6 or 7 times each. They were pilots of very wide experience, including flight in other VTOL aircraft and variable stability helicopters, which was of obvious assistance to them in assessing the S. C.1.”[1123] On October 2, 1963, while hovering at an altitude of 30 feet, a gyro input malfunction in the flight control system produced uncontrollable pitch and roll oscillations that caused the second S. C. 1 test aircraft (XG 905) to roll inverted and crash, killing Shorts test pilot J. R. Green. The air­craft was then rebuilt for additional flight-testing. The first S. C. 1 (XG 900) was used for VTOL research until 1971 and is now part of the Science Museum aircraft collection at South Kensington, London. The second S. C.1 (XG 905) is in the Flight Experience exhibit at the Ulster Folk and Transport Museum in Northern Ireland, near where the air­craft was originally built by Short Brothers.

The Canadian Avro CF-105 Arrow supersonic interceptor flew for the first time in 1958. Revolutionary in many ways, it featured a dual channel, three-axis fly-by-wire flight control system designed without any mechanical backup flight control capability. In the CF-105, the pilot’s control inputs were detected by pressure-sensitive transducers mounted in the pilot’s control column. Electrical signals were sent from the transducers to an electronic control servo that operated the valves in the hydraulic system to move the various flight control surfaces. The CF-105 also incorporated artificial feel and stability augmentation sys­tems.[1124] In a highly controversial decision, the Canadian government can­celed the Arrow program in 1959 after five aircraft had been built and flown. Although only about 50 flight test hours had been accumulated, the Arrow had reached Mach 2.0 at an altitude of 50,000 feet. During its development, NACA Langley Aeronautical Laboratory assisted the CF-105 design team in a number of areas, including aerodynamics, performance, stability, and control. After the program was terminated,

many Avro Canada engineers accepted jobs with NASA and British or American aircraft companies.[1125] Although it never entered production and details of its pioneering flight control system design were reportedly lit­tle known at the time, the CF-105 presaged later fly-by-wire applications.

NACA test data derived from the F9F-2 fly-by-wire experiment were used in development of the side stick controllers in the North American X-15 rocket research plane, with its adaptive flight control system.[1126] First flown in 1959, the X-15 eventually achieved a speed of Mach 6.7 and reached a peak altitude of 354,200 feet. One of the two side stick con­trollers in the X-15 cockpit (on the left console) operated the reaction thruster control system, critical to maintaining proper attitude control at high Mach numbers and extreme altitudes during descent back into the higher-density lower atmosphere. The other controller (on the right cockpit console) operated the conventional aerodynamic flight control surfaces. A CALSPAN NT-33 variable stability test aircraft equipped with a side stick controller and an NACA-operated North American F-107A (ex-USAF serial No. 55-5120), modified by NACA engineers with a side stick flight control system, were flown by X-15 test pilots during 1958— 1959 to gain side stick control experience prior to flying the X-15.[1127]

Interestingly, the British VC10 jet transport, which first flew in 1962, has a quad channel flight control system that transmits electrical signals directly from the pilot’s flight controls or the aircraft’s autopilot via elec­trical wiring to self-contained electrohydraulic Powered Flight Control Units (PFCUs) in the wings and tail of the aircraft, adjacent to the flight control surfaces. Each VC10 PFCU consists of an individual small self – contained hydraulic system with an electrical pump and small reservoir. The PFCUs move the control surfaces based on electrical signals pro­vided to the servo valves that are electrically connected to the cockpit flying controls.[1128] There are no mechanical linkages or hydraulic lines between the pilot and the PFCUs. The PFCUs drive the primary flight
control surfaces that consist of split rudders, ailerons, and elevators on separate electrical circuits. Thus, the VC10 has many of the attributes of fly-by-wire and power-by-wire flight control systems. It also features a backup capability that allows it to be flown using the hydraulically boosted variable incidence tail plane and differential spoilers that are operated via conventional mechanical linkages and separate hydraulic systems.[1129] The VC10K air refueling tanker was still in Royal Air Force (RAF) service as of 2009, and the latest Airbus airliner, the A380, uses the PFCU concept in its fly-by-wire flight control system.

The Anglo-French Concorde supersonic transport first flew in 1969 and was capable of transatlantic sustained supercruise speeds of Mach 2.0 at cruising altitudes well above 50,000 feet. In support of the Concorde development effort, a two-seat Avro 707C delta winged flight research aircraft was modified as a fly-by-wire technology testbed with a side stick controller. It flew 200 hours on fly-by-wire flight trails at the U. K. at Farnborough until September 1966.[1130] Concorde had a dual channel analog fly-by-wire flight control system with a backup mechanical capa­bility. The mechanical system served in a follower role unless problems developed with the fly-by-wire control elements of the system, in which case it was automatically connected. Pilot movements of the cockpit con­trols operated signal transducers that generated commands to the flight control system. These commands were processed by an analog electri­cal controller that included the aircraft autopilot. Mechanically operated servo valves were replaced by electrically controlled ones. Much as with the CF-105, artificial feel forces were electrically provided to the Concorde pilots based on information generated by the electronic controller.[1131]