Without a doubt, NASA's new 737 was a one-of-a-kind airplane. In addition to a conventional forward flight deck, Boeing modified the airplane with a second, experimental cockpit in the forward part of the main cabin that contained the advanced SST avionics. The aft flight deck (AFD) was enclosed in a fullsize fiberglass duplicate of the 737's front cockpit exterior that left just enough room for a small passageway to one side. A second fiberglass cab was used to build a highfidelity fixedbase ground simulator that replicated the aft flight deck. The simulator, run by a highspeed Control Data Corporation CDC6400 mainframe computer, allowed experiments to be tested and developed through realtime simulation on the ground before they were put on the airplane.

A typical map display using Cathode Ray Tube Technology

NASA flight test programs usually divided an aircraft's cockpit into two parts and installed any experimental equipment on one side, leaving a safety pilot on the other. The TCV researchers wanted the ability to evaluate new transport technology in a realistic, twocrew environment, however, so the 737 was outfitted with the complete second cockpit. For the NASA research flights, the aft cockpit was equipped with four monochrome cathode ray tube (CRT) displays. The pilot and copilot positions had both a primary flight display (PFD), and a navigation, or map, CRT display in front them, installed above a Control and Display Unit (CDU), which was the pilot's interface with the navigation computer. [Ref 2-21]

In order for the pilots to be able to see the full displays, the aft flight deck was equipped with two individual handles that came out of the instrument panel instead of conventional center control yokes. Sidestick controllers, like those eventually installed in the Airbus A320 airliner, were ruled out because they were a more dramatic departure from conventional yokes, and the researchers wanted to keep the 737's cockpit at least somewhat familiar to airline pilots. Although they were referred to in technical papers as "Panel Mounted Controllers," the dual handles were dubbed "Brolly Handles" by a British engineer who worked on the project, because their shape reminded him of an umbrella handle, or "brolly," as it is sometimes called in England. The instrument panel of the aft flight deck also included electromechanical engine instruments and a Boeing Advanced Guidance and Control Panel (AGCS), which was used to select different levels of automatic or manual flight control.

Because the focus of the TCV research was flight operations under Instrument Flight Rules (IFR), the complete lack of visibility from the aft flight deck was not considered a problem. The FAA, on the other hand, wanted its pilots to be able to try the experimental SST equipment in a less severe environment, where they could still use outside visual cues. So the 737 was also wired to allow one set of the displays, one CDU and the AGCS Panel to be installed on the right side of the forward flight deck. Instead of brolly handles, however, the righthand control column of the 737 was simply shortened to keep the yoke from obscuring the displays, since the FAA wanted to keep the configuration as conventional as possible. This "FAA" configuration, however, was only used once, for baseline testing of the SST equipment at the FAA's technical center in Atlantic City, New Jersey, in the fall of 1974.[Ref 2-22]

The airplane was also equipped to use the SST equipment in a third, "split" configuration, which would have put one set of the displays in the righthand side of the forward flight deck for monitoring by a safety pilot. This configuration was never actually used, however.

In all the configurations, control inputs from the experimental systems were processed through the SST program's digital flight control computer, which then interfaced with the airplane's autopilot. Because the computer relied on the autopilot system to actually drive the control surfaces, the aft flight deck was restricted to half the control authority of the forward cockpit, which used a conventional 737 powered control system.

This arrangement allowed experiments to be conducted in the aft cockpit while safety pilots monitored all the operations from the forward flight deck. This was an extremely valuable capability, as it allowed new and unproven technologies to be tested in an actual flight environment while maintaining an acceptable level of safety. It would probably have been far too risky, for example, to include autolands in the testing of the MLS curved path approaches if the plane did not have safety pilots up front who could see outside, monitor the progress of the approach, and take over if necessary. And yet the actual completion of those autolands was one of the things that made the MLS demonstrations so effective.

Tightly packed electronics racks required to conduct research on the Boeing 737 aircraft. Each pallet had a row of three seats for researchers. Note the aft flight deck partially shown in the center background.

The safety pilots could take over control of the airplane simply by pushing one of two buttons or operating a trim system switch, and the pilots in the aft flight deck could give control back to the forward flight deck by pushing a disconnect button. Annunciation lights in both cockpits would light up with any change in command, and the pilots would also verbally notify each other of the switch over the airplane's intercom system. Status messages and requests from the aft flight deck were monitored by the safety pilots on a "Control and Command" panel on the top of the front instrument panel. Gear and speed brakes, for example, could not actually be controlled by the aft flight deck. When a pilot in the back put his gear handle down, a "Gear" light on the Command and Control panel would light up, telling the safety pilot to extend the gear. The Control and Command panel also had an emergency disconnect knob that would physically disconnect the aft flight deck interface with the aircraft controls, in case the electrical disconnect system failed.

Behind the aft flight deck were rows of pallets that housed all the computers that ran the experimental systems, as well as data gathering equipment. Behind each row of pallets was a row of seats, so researchers could monitor the data and operation of each element of the system as well as any experiment that was being conducted.

There were three major experimental subsystems installed in the plane in addition to the data and video collection equipment. One subsystem operated the actual flight controls of the airplane (aileron, elevator, rudder), controlling the airplane's physical movement. A second subsystem provided computerized navigation functions, which controlled the airplane's flight path. The third subsystem operated the electronic flight displays in the aft cockpit.

The original experimental equipment for the three systems consisted of triply redundant General Electric (GE) ICP 723 flight control computers, a General Electric 701 digital display computer, and a Litton C4000 navigation computer. The aircraft also incorporated a triply redundant Litton LTN51 Inertial Navigation System (INS). The General Electric computers were all prototypes, developed specifically for the SST program. The ICP 723 computers were actually not fully digital but something called "incremental word" computers; an intermediate step between analog and digital systems.[Ref 2-23]

The biggest problem with the GE equipment was that it was never intended to be operational in an airplane for a long period of time. It had been designed to operate for only one ninetyday test. Yet NASA ended up flying the display computers, for example, for 12 years. The maintenance headaches this caused were complicated further by the fact that the equipment was what the NASA technicians termed "brassboard," or only one step better than the crudely connected systems electronics researchers would initially test in a lab. There were no maintenance manuals and no replacement systems. If a problem developed, it sometimes took phone calls to six or more GE engineers who had helped design the system as well as scanning pages of blueprint drafts to figure out how it could be fixed. Then the problem had to be troubleshot down to the level of individual parts on the circuit cards, because there were not even any spare circuit cards.

Since there were no spares in a lab to use to troubleshoot problems, the crew in charge of the experimental systems had to use the airplane as their lab. During flight test periods, this meant that repairs generally had to be made at night, so the next day's experiments could still be flown. Amazingly, although some flights had to be cancelled, the late hours and the resourcefulness of the Langley personnel who worked on the equipment kept the airplane from ever missing a major research program.

The NASA 737 on the ramp during flight tests in 1992 as
lightning strikes in the background.

Fortunately, the equipment was upgraded substantially over the years. The first change was made in 1976, when the GE ICP 723 computers were replaced with triply redundant GE 703 wholeword computers, which were full digital systems. In 1983 the GE 703 computers were replaced with a single Norden 1170 flight control computer. The single string flight control system was considered acceptable because of the backup provided by the conventional controls in the forward flight deck. The Norden 1170 also replaced the Litton C4000, giving the airplane its first complete flight management computer.

The system architecture of the experimental system was also changed in 1983 to a Digital Autonomous Terminal Access Communications (DATAC) data bus, invented by Boeing engineer Hans Herzog. Instead of requiring dedicated buses and connections for each item researchers wanted to get into or out of the computer, the DATAC system was a "broadcast" bus. All the information was broadcast in sequence down a single twisted pair of wires, and any station that needed information could simply connect to the bus and collect whatever data it required. The DATAC system also used magnetic coupling instead of hard connections, which made adding or changing experimental equipment vastly easier.

In 1986, the original GE 701 display computer was finally replaced with a second Norden 1170, and new Sperry/Honeywell color displays were installed. There was some debate among Langley engineers as to what size display to buy, because the only "off the shelf" displays available were small, 5" x 7" "Bsize" CRTs. Sperry/Honeywell was in the process of developing bigger 8" diameter "Dsize" displays, but they were not fully tested yet. However, the Langley managers decided that since the bigger displays were going to be the wave of the future in transport airplanes, they would take the risk of ordering them for the 737. Eventually, the aft flight deck was equipped with eight of the color monitors.[Ref 2-24]

Two years later, the Norden 1170 flight management computer was replaced with a Digital Equipment Corporation (DEC) MicroVax II computer, which was faster, smaller, cheaper, easier to cool, and took less power to operate. At the same time, the brolly handles on the left side of the aft flight deck were replaced with a McFadden side stick controller.

In 1990, a second DEC MicroVax II computer took the place of the Norden 1170 display computer, and second side stick controller was installed. The most recent upgrade to the system was the installation of new computer cards in the MicroVaxes that equipped the computers with a new processor. The upgrade changed the computers into MicroVax IVs and doubled their speed.[Ref 2-25]

The almost continuous upgrades in experimental equipment were necessary to keep pace with the rapid developments in computer technology over the last 20 years. The equipment has not been the only aspect of the program to change, however. The organizational structure and the level of support the program has enjoyed have both varied widely over the years.

Cutaway model of the Boeing 737 used for flight research showing
it aft cockpit and the location of various research pallets.