The Northrop YF-23 ATF Prototype Air Vehicle was a 5th generation, twin-engined, single seat prototype jet fighter aircraft that employed low observable or stealth technology to reduce the chance of detection in a combat environment, both in the microwave radar and infra-red spectrums. It was designed, developed and built specifically in response to the USAF's request for an Advanced Tactical Fighter
to replace the F-15 Eagle
. It was a 'bare-bones' prototype, devoid of the sensors, avionics and full weapons capacity specified in the ATF requirement, which were tested and demonstrated separately in a different part of the programme. It contained only those avionics sufficient to carry out its function of demonstrating flight performance commensurate with Northrop's predictions for the EMD Proposal
. It was capable of flight at sustained supersonic speed without the use of afterburner. It was built and flown in 2 different versions, to accommodate the 2 contenders for the ATF engine contract. The engine bays of the aircraft were designed in such a way as to accommodate the 2 different engine types with maximum structural commonality.
The overall shape was carefully designed to cater as much as possible to speed and all-aspect stealth, while offering credible agility and high alpha flight without departure. It was a triolithic configuration with a diamond shaped trapezoidal wing, a 'V' or butterfly tail, and conformed tightly to the Area Rule in the shape and distribution of its volume. The planform consisted of completely straight creases or edges aligned in 7 directions (6 angular and 1 longitudinal) to minimise radar return spikes as much as possible. All edges were parallel to one of these 7 alignments. The primary angle of alignment was 40°. 90° corners and vertical surfaces were kept to a minimum. The overall shape employed a series of flat surfaces or facets smoothly blended or bevelled into 3 essential masses with flowing curves. The sides of the aircraft also conformed as closely as possible to 4 different angular alignments when viewed from the front or rear. The configuration could essentially be thought of as a large clipped diamond wing to which 3 volumetric masses have been attached. The 2 rear masses consist of the engine nacelles and air intakes and house the main undercarriage, while the front mass is the forebody or fuselage containing the weapons bay, cockpit, avionics, and front undercarriage. The fuselage was pinched in volume almost precisely at the point of maximum wing span. The leading and trailing edge angle of sweep of the flying surfaces was 40°. The slits between the control surfaces on the wings were vertical. The leading edge was radiused, while the trailing edge was sharp. The V-tail offered significant IR shielding of the rear jet exhaust. It was canted outwards at an angle of 40° dihedral. The leading edge of the V-tail commenced precisely where the trailing of the wing left off. The conformity to area ruling was probably the tightest of any aircraft ever.The overall shape would probably most closely resemble a stingray, and it is widely considered to be one of the most beautiful aircraft ever to fly. It is the only supersonic fighter to ever employ a V-tail, the only fighter in the world to employ a completely symmetrical diamond wing planform, and was quite possibly (in PAV-2 guise) the fastest aircraft in the world in relation to supercruise.
The primary structural materials were carbon/bismaleimide and titanium. The load of the main wing was distributed between 4 primary spars. The firewalls, engine support frames, wing spars and attachment frames were titanium alloy. The undercarriage was steel. The nose was made out of aluminium and was devoid of any radar or sensors. The canopy was made out of polycarbonate, and the exhaust trough tiles at the back were made from Titanium Lamilloy.
The heart of the aircraft was the Vehicle Management System, designed to monitor and accept feedback from all systems aboard the aircraft, and to provide necessary control input. It integrated all functions into a single organism allowing for a far greater level of dynamic response with geater efficiency than a conventional system. It allowed for the complete integration of normally disparate systems: it controlled flight control surface movement, engine operation, BLC door position, nosewheel steering, anti-skid of the wheels, and air data/flight test data to do with the flight surfaces. The core of the VMS was the bank of Vehicle Management Computers (VMCs): 4 linked computers feeding into a completely digital control system for quadruple redundancy.
The YF-23 was a highly coupled and highly unstable vehicle aerodynamically to allow for maximum agility in flight. It was artificially stablised by the VMS. The aircraft employed a series of aerodynamic control surfaces that were fully integrated into the VMS, meaning that any surface could be used for any control task, and manoevure was accomplished by moving all the surfaces in concert for a result rather than the traditional method of having a dedicated control surface for a particular task. For example, the leading edge flaps could be used differentially, to roll the aircraft in concert with other surfaces; in unison, to provide lift for landing and takeoff or high AoA manoevure, or to dump lift and brake the aircraft on landing rollout. A control surface could perform at least 3 functions at once: supplying manoevure control vectors, cancelling unwanted off-axis effects from other surfaces (eg yaw input from the v tails when pitch was required), and maintaining constant flying stability. As a consequence, the aircraft had no separate dedicated airbrake. Braking was provided by deflection of the trailing edge wing flaperons. The inner main flaperons could be deflected hard down while the outer secondary flaperons were deflected hard up, or vice versa. In terms of specific control input tasking: yaw control was controlled by the V-tails (the trailing edge wing flaps compensated for roll caused by the V-tail). Roll was controlled by differential deflection of the flaperons. Pitch was controlled by deflection of the V-tail. The VMS maintained zero side-slip unless commanded otherwise by the pilot.
To provide feedback to the VMS, the nose of the aircraft featured flush air pressure sensors with non-steathy pitot tubes for backup. AoA static ports were located top and bottom on the centreline of the nose. Differential pressure recorded by the ports provided angle of attack information. Sideslip static ports were located on the nose offset to the centreline. Differential pressure provided angle of side-slip. Combined pressure from both sides provided static pressure for airspeed and altitude computation. 2 L-shaped pitot tubes on the lower fuselage aft of the nose provided total pressure measurements for airspeed backup computation. An air data calibration check was commenced by the VMS on take off roll at approx 45 knots, comparing airspeed, wheel spin, and INS velocity.
The weapons bay was a single compartment situated behind the cockpit. It was finished in zinc chromate primer on PAV-1, gloss white on PAV-2, and contained plumbing visible on the inside walls. The bay was 160 inches long. The design was provisionally configured to carry 3 AIM-120A AMRAAMs, and 2 AIM-9 Sidewinders. Missiles were to be attached to a single flat 'pallet' that swung down and forward on hydraulic actuators to lie flush with the underside of the fuselage for weapons launch. The launchers for the AMRAAMs were angled outwards at 17° to facilitate positive ejection outwards & downwards. The missiles would have been attached to LAU-106 adapters and were aligned 4.5° nose downwards when deployed. Small white baffles swung down into the airstream at the same time the weapons bay doors opened, to stabilise the airflow around the missiles, ensuring a safe and clear separation on launch. The weapons bay doors were relatively large single piece units, and opened out to lie sloped at a similar angle to the sides of the air intakes, effectively shielding the intakes from missile exhaust on launch. On PAV-1, the doors were zinc chromate primer on the interior surfaces while the hinge booms were white. The doors had relatively smooth interior surfaces comensurate with weapons bay acoustic tests. They were equipped to potentially mount a rail launched Sidewinder each using a LAU-114 adaptor. On PAV-2, the doors where white and had exposed structural ribs on the interior instead of smooth surfaces. This aircraft was scheduled for tasks other than weapons bay acoustic measurement. The doors on PAV-1 could operate in 3 modes: fast launch, in which the doors opened in 2.1 seconds; standard launch in which the doors opened in 5 seconds; or slow mode where they opened in 10 seconds. This was the default mode for ground operations to ensure safety for ground crew. The prototypes were flown in a basic configuration; weapons integration was planned for a later stage had Northrop won the EMD phase contract. Weapons release demonstrations were simulated using computer modelling. Provision was made for a standard GE M61 20mm Gatling gun with 500 rounds in the starboard wing root. Access to the ammunition was from an access hatch located on the underside of the a/c. Again, the gun was not actually carried. The space reserved for ammo was occupied by test equipment.
Quoting the Pilot's Manual:
"Armament System page 1-123. The armament system described is the config after the weapons integration enhancement. Some hardware might not be installed and weapon switches described may not be functional prior to the enhancement. The internal weapons carriage system consists of the weapons bay insert, Advanced Technology Launcher(s) (ATL), the door launch assembly, the weapons bay doors and their drive mechanism. Each... door is mechanically linked to an airflow spoiler. The spoilers extend 6 inches below the mold line when deployed. The weapons bay insert supports the advanced technology launcher and a modified LAU-106 for captive carriage of an AIM-120A. The platform, suspended at an outward angle of 17°, imparts a 'down-and-out' trajectory to missiles ejected by the launcher... After the missile is launched, the launcher automatically retracts... The ATL ejects the AIM-120A by releasing the missile at the end of the launcher linkage arm extension stroke. The missile is aligned 4.5° nose down on release with a slight nose down momentum. The launcher also has provisions to rail launch AIM-9 missiles with the installation of a LAU-114 launcher. Structural provisions are included to mount a weapons bay boor launch assembly for AIM-9 carriage and release. The weapons bay doors have 3 opening and closing speeds: 2.1 sec (normal), 5 sec (med), and 10 sec (slow) for ground use."
The cockpit had a unique layout but did not include any new equipment specifically designed for the ATF programme. All hardware in the cockpit was off-the-shelf. The instrument panel used display units from the F-15E but was of a different overal arrangement. Although the panel layout of the main CRT's was similar to the F-15E, the panel was devoid of the warning light cluster found on the right hand side of the F-15E panel, and in place of the analogue gauge clusters found directly under the F-15E's main left and right CRT's, there were simple air conditioning vents. There was an absolute minimum of analogue instrumentation in the YF-23, just a standby attitude indicator and cockpit altimeter on the extreme right of the instrument panel. The panel included a unique and new longitudinal CRT-type diagnostic/warning display unit on the right. The HUD system was identical to the F-15E. The side console layout was unique to the aircraft, although the cockpit switchology was similar to the F-15E. The consoles featured test equipment not normally found on a production aircraft. The aircraft sported a standard ACES II seat finished in black rather than the usual grey. The seat was reclined back at 18°. Red brackets (for mounting the flight-monitoring video cameras on) were installed on the seat ejection rails, although PAV-2 had a slightly beefier bracket on the left ejection seat rail. Cabin pressure regulation equipment featured quite prominently immediately behind the seat, looking like 2 black R2-D2 droid units! The pilot was able to see over the nose at a depressed angle of 15°. The canopy frame was finished in zinc chromate primer except for upwardly visible areas of the frame inside the actual cockpit area, which were finished in matt black. Rather than having rows of locking hooks, it featured two very large shark tooth shaped hooks towards the rear. The canopy frame mounted two mirrors and a pair of grab handles. The cockpit sill featured prominent "experimental" labels on both sides. The joystick was centre-mounted. The throttle quadrant was borrowed from the F-15E also. In PAV-2, the strut of the throttle quadrant was left in zinc chromate primer. The Pilots manual diagram of the instrument panel layout shows 2 spin gauges: these were not installed on the actual aircraft because spin tests were not carried out. Had the aircraft gone on to EMD, spin tests would have been conducted.
In addition to the VMS, the aircraft had the following avionics and systems aboard: a Garret Auxiliary Power Unit (APU), which provided pressurised air for engine starts on the ground; a Litton OBOGS from an AV-8B Harrier; a KY-58 UHF Secure Voice Have Quick radio; UHF antennae, C-Band beacons, an L-Band beacon; and IFF, INS, TACAN, and an ECS. The Environmental Control System (ECS) provided conditioned air for cockpit pressurisation and air conditioning, windscreen defog, canopy seal, anti-g, fuel tank pressurisation, oxygen generation, and avionics cooling. It was powered by engine bleed air, APU, or ground cart as necessary. Cockpit pressurisation was automatic, ambient pressure was the equivalent of 8,000ft up to 23,000 ft altitude actual; thereafter 5 psi differential to outside pressure. At no stage did either aircraft carry either a radar or combat sensors. The ECS temperature bleed air scoop could remain open in flight. On PAV-1, the ECS was located in the cockpit immediately ahead of the rear bulkhead, below the canopy line. On PAV-2 it was located in the foward section of the weapons bay.
The engine bay and avionics bay access doors were zinc chromate primer on the interior surfaces. The engine bay doors hinged downwards and inwards. The avionics bay doors had quick release fasteners and were completely removeable. There was an avionics rack mounted just ahead of the cockpit that slid down vertically on dual rails underneath the aircraft for maintenance access. While key panels had appropriate alignment with one of the 7 axes, many panels on the aircraft were right-angled in shape.
The aircraft employed the standard tricycle undercarriage configuration. The main oleo legs were trailing-axle cantilevered units unique to the aircraft, not borrowed from the F-15 or F-18, as popular opinion would suggest, although F-18 main wheels were employed. The configuration of the undercarriage was very straight-forward and simple. The rear units retracted rearwards and straight up into the body without any swivel of the wheel (unlike the F-15 and F-18 which both swivel the wheels to lie flat underneath the fuselage), and the front unit rotated up and forwards into the front fuselage, again, without any wheel swivelling. The front oleo strut was a stripped-down F-15 unit with different forward struts and the addition of a trailing arm to combat shimmy tendencies. The shock absorber travel was reduced to cater to the height requirements of the YF-23 fuselage, which sits lower to the ground than an F-15. Both front and rear undercarriage units employed single doors respectively, in contrast again to the teen series fighters which have multiple doors per unit. The front gear door opened and retracted extremely rapidly, in a virtual snap movement. Landing gear retraction was 4.5 sec, extension was 7 sec, and emergency extension could be accomplished in around 25-35 sec. Nosewheel steering was 20° above 15 knots, up to 45° below 15.
The PAVs flew using engines equipped with nozzles that were configured for STOL operations according to the original ATF requirement. This capability was not actually tested due to the redundancy of the requirement. The nozzle developed for the YF-23 was a very simple 2-dimensional design with a single moveable paddle. The P&W YF119 employed a moveable exterior shroud to the paddle, whereas the GE YF120 had a fixed shroud. The paddles were known officially as a Single Expansion Ramp Nozzle or SERN. Fuel was fed from 11 fuel tanks distributed throughout the wings and fuselage: 1 in each wing, 1 above each intake, 4 surrounding the weapons bay, and the remaining tanks in the centre fuselage.