The Advent of Digital Flight Control Systems
Digital flight control systems were more nuanced still.[702] Analog computers calculate solutions simultaneously, thus producing an instantaneous output for any input. Digital computers, although more precise than analog, calculate solutions in sequence, thus introducing a time delay between the input and the output, often referred to as "transport delay.” Early digital computers were far too slow to function in a realtime, flight control feedback system and could not compute the required servo commands fast enough to control the aircraft motions. As digital computation become faster and faster, control system designers gave serious attention to using them in aircraft flight control systems. NASA Dryden undertook the modification and flight-testing of a Vought F-8C Crusader Navy fighter to incorporate a digital fly-by-wire (DFBW) control system, based on the Apollo Guidance Computer used in the Apollo space capsule. The F-8 DFBW’s first flight was in 1972, and the test program completed 248 DFBW flights before its retirement at the end of 1985.
It constituted a very bold and aggressive research program. The F-8 used redundant digital computers and was the first airplane relying solely on fly-by-wire technology for all of its flights. (Earlier FBW efforts, such as the AF F-4 Survivable Flight Control System, used a mechanical backup system for the first few flights.) NASA’s F-8 DFBW program
not only set the stage for future military and civil digital flight control systems and fly-by-wire concepts, it also established the precedent for the operational procedures and built-in-test (BIT) requirements for this family of flight control systems.[703] The ground-testing and general operating methods that were established by NASA DFRC in order to ensure safety of their F-8 DFBW airplane are still being used by most modern military and civilian airplanes.
After the completion of the basic digital FBW demonstration program, the F8 DFBW airplane was used for additional research testing, such as identifying the maximum allowable transport delay for a digital system to avoid pilot-induced oscillations. This is a key measurement in determining whether digital computations are fast enough to be used successfully in a control system. (The number turned out to be quite small, on the order of only 100 to 120 milliseconds.) The stimulus for this research was the PIO experienced by Shuttle pilot-astronaut Fred Haise during the fifth and last of the approach and landing tests flown at Edwards by the Space Shuttle orbiter Enterprise on October 26, 1977. Afterward, the Shuttle test team asked the DFBW test team if they could run in-flight simulations of the Shuttle using the F-8 DFBW testbed, to determine the effect of transport delays upon control response. During this follow-on research-testing phase, NASA Dryden Flight Research Center pilot John Manke experienced a dramatic, and very scary, landing. As he touched down, he added power to execute a "touch and go” to fly another landing pattern. But instead of climbing smoothly away, the F-8 began a series of violent pitching motions that Manke could not control. He disengaged the test system (which then reverted to a digital FBW version of the basic F-8 control system) just seconds before hitting the ground. The airplane returned to normal control, and the pilot landed safely. The culprit was an old set of control laws resident in the computer memory that had never been tested or removed. A momentary high pitch rate during the short ground roll had caused the airplane to automatically switch to these old control laws, which were later
The Ling-Temco-Vought A-7D DIGITAC testbed was the first U. S. Air Force airplane with a digital flight control system. USAF. |
determined to be unflyable.[704] This event further reinforced the need for extensive validation and verification tests of all software used in digital flight control systems, no matter how expensive or time-consuming.
In 1975, the Air Force began its own flight-testing of a digital flight control system, using a Ling-Temco-Vought A-7D Corsair II attack aircraft modified with a digital flight control system (dubbed DIGITAC,) to duplicate the handling qualities of the analog Command Augmentation System of the baseline A-7D aircraft. As well, testers intended to evaluate several multimode features.
The model-following system was enhanced to allow several models to be selected in flight. The objective was to determine if the pilot might desire a different model response during takeoff and landing, for example, than during air-to-air or air-to-ground gunnery maneuvers. The program was completed successfully in only 1 year of testing, primarily because the airplane was equipped with the standard A-7D mechanical backup system. The airplane used two digital computers that were continuously compared. If a disagreement occurred, the entire system would disengage, and the backup mechanical system was used to safely recover the airplane. The pilot also had a paddle switch on the stick that
immediately disconnected the digital system. This allowed software changes to be made quickly and safely and avoided most of the necessary, but time-consuming, preflight safety procedures that were associated with NASA’s F-8 DFBW program.[705]
One of the more challenging flight control system designs was associated with the Grumman X-29 research airplane. The X-29 was designed to demonstrate the advantages of a forward-swept wing (FSW), along with other new technologies.
The airplane would fly with an unusually large level of pitch instability. The F-16, while flying at subsonic speeds, had a negative static margin of about 6 percent. The X-29 static margin was 35 percent unstable. (In practical terms, this meant that the divergence time to double amplitude was about half a second, effectively meaning that the airplane would destroy itself if it went out of control before the pilot could even recognize the problem!) This level of instability required extremely fast control surface actuators and state-of-the-art computer software. The primary system was a triplex of digital computers, each of which was backed up by an analog computer. A failure of one digital channel did not prevent the remaining two digital computers from continuing to function. After two digital channel failures, the system reverted to the three allanalog computers, thus maintaining fail-op, fail-op, fail-safe capability.
After completing the limit-cycle and resonance ground tests mentioned earlier, plus a lengthy software validation and verification effort, the flight-testing began in 1984 at NASA’s Dryden Flight Research Center.[706] The control system handled the high level of instability quite well, and the test program on two airplanes was very successful, ending in 1992. Although the forward-swept wing concept has not been incorporated in any modern airplanes, the successful completion of the X-29 program further boosted the confidence in digital FBW control systems.[707]
In recent years, the digital FBW systems have become the norm in military aircraft. The later models of the F-15, F-16, and F/A-18 were
equipped with digital FBW flight control systems. The C-17 Globemaster III airlifter and F-117 Nighthawk stealth fighter performed their first flights with digital FBW systems. The Lockheed Martin F-22 Raptor and F-35 Lightning II Joint Strike Fighter exploit later digital FBW technology. Each has three digital computers and, for added safety, three of each critical component within its control systems. (Such "cross-strapping” of the various components allows FOFOFS redundancy.) There are dual-air data systems providing the various state variables that are backed up by an inertial system. The various "survivability” features first examined and demonstrated decades previously with the F-4 SFCS program (wire-routing, separate component locations, etc.) were also included in their basic design.