Chapter 7

The F 117 Nighthawk Story (Part 2)

by Jay Miller

Have Blue

In August of 1975, Lockheed, along with two other contractors (Boeing and Northrop), was invited by DARPA to participate in a competitive effort to develop and test an aircraft known as the Experimental Stealth Testbed (XST) (some sources claim the acronym stood for Experimental Survivable Testbed). In April of 1976 Lockheed was chosen to proceed with detail design, devel­opment, and test of this new aircraft. The program was given the name Have Blue.

Preliminary design responsibility for the new aircraft at Lockheed was given to Dick Scherer, Warren Gilmour, and Leo Celniker. Program Manager was Norm Nelson and the principle engi­neers assigned were Ed Baldwin, Alan Brown, Dick Cantrell, Henry Combs, Bob Loschke, the afore-mentioned Denys Overholser, and Bill Taylor. Manufacturing was placed under the direc­tion of Bob Murphy.

Have Blue was a subsonic, single-place air­craft powered by two General Electric J85GE-4A engines. The aircraft was 47 feet 3 inches in length, 7 feet 6-1/4 inches high, and had a wingspan of 22 feet 6 inches with a resultant wing area of 386 sq. feet. The wing planform was modified delta with a sweep of 72.5°. There were no flaps, speedbrakes, or high lift devices incorporated in the design. The structure was aluminium alloy with steel and titanium utilized in the hot areas. The surface controls were elevons, located inboard on the wing trailing edges, and two all moveable fins at the wing root that were swept back some 35° (leading edge) and canted inboard some 30°. A side stick controller and conventional rudder pedals operated the control surfaces through a fly-by-wire command and stability aug­mentation system without mechanical back-up. Elevon nose-down pitch control was augmented by a large, two-position flap called the platypus, which was deflected downward automatically whenever 12° angle-of-attack was exceeded

The aircraft had a tricycle landing gear with main gear antiskid braking. Nose wheel steering was installed on the second aircraft. While the for­ward retraction insured reliable emergency gear extension, it also meant that takeoff and landing would always occur at the most aft center-of-grav­ity location.

The test gross weight of the aircraft ranged from approximately 9,200 to 12,500 pounds. Zero fuel weight was 8,950 pounds. Fuel weight was 3,500 pounds and all fuel was carried in fuselage and wing tanks.

The unorthodox Have Blue configuration was designed to provide a highly maneuverable fighter aircraft with Very Low Observable (VLO) characteristics. As a result, the external shape evolved from VLO and controllability characteris­tics. This resulted in a relaxed static stability (RSS) aircraft which required a quad-redundant, fly-by­-wire flight control system (FCS) to provide han­dling qualities throughout the flight envelope.

The restrictions imposed by VLO require­ments were unprecedented and demanded new approaches to preserve efficient propulsion sys­tem performance. Each inlet duct was equipped with a flat, RCS treated grid given a porosity sized for the cruise condition. The airflow was aug­mented at takeoff with blow-in doors mounted on the upper fuselage surface. There was concern that the inlet grids would create problems with engine performance, but these worries proved unfounded. The grids actually had a beneficial side effect in that they helped straighten the vortex disturbed inlet airflow from the highly swept wing leading edges, especially at high angles of attack.

The General Electric J85-GE-4A non-after-burning engines were obtained as Government Furnished Equipment (GFE) from United States Navy T-2B trainer stores. No engine modifications except coating of the spin­ners were made.

The exhaust system design was likewise dri­ven by VLO requirements. To prevent radar energy from penetrating to the turbine face, the tailpipe was transitioned from a round duct to a 17­to-1 flattened slot convergent nozzle. The trailing edge of each nozzle was terminated on a 54° scarf angle to correspond to the airframe aft closure. Vanes which were interposed and angled in the slot exit helped straighten the exhaust flow back to the longitudinal axis, although some thrust vector "toe-in" remained. Sufficient bypass air was passed over the tailpipe to cool the aft fuselage structure.

The Have Blue test program was clearly out­lined in the original Development and Demonstration test plan to include: radar cross section and wind tunnel model tests of the proto­type design; qualification and proof tests for vari­ous systems and subsystems; preflight testing of the assembled aircraft and systems; and finally, flight tests of the aircraft.

As previously stated, the basic configuration was developed utilizing the Echo 1 program. A one-third scale RCS model of the Have Blue con­figuration was tested during December 1975 at the Grey Butte Microwave Measurement facility and a small model in Lockheed Anechoic Chamber. A second series of one-third scale model tests were conducted at Grey Butte in January of 1976. These tests confirmed signifi­cant R improvements made with a few minor configuration changes.

On the basis of these tests, low and high speed wind tunnel models were fabricated. Only 1,920 hours were required to tailor and define aerodynamic and propulsion characteristics.

A full-scale RCS model was constructed and used at the Air Force Ratscat Backscatter Measurement Range at White Sands, New Mexico to further develop and validate the VLO design. Many detail problems were resolved dur­ing this stage, allowing manufacture of the two test aircraft to proceed rapidly.

The initial engine runs were accomplished on the first Have 8lueaircraft on November4 1977 at Lockheed's Burbank facility. In order to maintain security, the aircraft was parked between two semi­-tractor-trailers over which camouflage net had been installed. The runs were performed at night after the airport was closed.

The biggest problem experienced was a phone call from a local resident who wanted to know what was making all the noise. Following the engine runs, the aircraft was partially disassembled (the wings were removed) and readied for shipment. It was delivered to the test location, via C-5A, on November 1 1977. Since this was the first time a C-5A had operated from the Burbank airport, the morning traffic became substantially more congested as people strained to see this “aluminum overcast” that appeared to hover over the city during its depar­ture.

The aircraft was off-loaded at the Groom Lake, and reassembled. Since most of the sys­tems had been checked out in Burbank, only a few System Check Outs (SCOs) needed to be accom­plished prior to first flight, which would be con­ducted under the direction of Have Blue Flight Test Manager Dick Miller. Engine thrust runs were performed and a series of four low and high speed taxi tests were conducted. During the third taxi test a problem developed that would become a nui­sance throughout the program. In particular, over­heated brakes which caused the wheel fuse plugs to melt. Flight control system performance was carefully monitored during the taxi tests and some minor changes were made to the yaw gains. Successful drag chute operation was verified and, following the fourth taxi test, the aircraft was deemed ready to fly.

On December 1 1977, with Lockheed test pilot Bill Park at the controls, Have Blue lifted into the air. A new era in military aviation had just begun. Only twenty months had passed since contract award.

As previously noted, the primary objective of the test program was to demonstrate VLO tech­nology. Towards this end, Have Blue 1001 would demonstrate loads/flutter, performance handling qualities, and stability and control. Have Blue 1002 was designated as the RCS test vehicle.

Have Blue 1001 accomplished 36 flights over the next 5 months and successfully expanded the flight envelope sufficiently to allow the RCS testing to be performed. Unfortunately, on May 4 1978, the aircraft sustained major dam­age during an attempted landing at less than design landing speed and had to be abandoned in flight by Park.

Have Blue 1002

Have Blue 1002 joined the program during July 1978 and flew for the first time on July 20 with Air Force pilot Norman "Ken" Dyson at the controls. This aircraft differed from Have Blue 1001 in that it possessed a "real" airspeed system (no nose boom) and did not have a drag chute installed. It also incorporated nose wheel steering to improve ground handling, and was adorned with all the coatings and materials required to per­form its intended task. Following some airspeed calibration flights, the aircraft accomplished 52 flights during the following 12 months and com­pleted the low observable testing.

Have Blue's pitot static system consisted of three separate static sources on the upper and lower forebody surfaces and three total pressure probes; one at the nose tip and two on the wind­shield center post. Have Blue 1001 also had a flight test nose boom which included pitot-static pressure sources and angle-of-attack and sideslip vanes. A flight path accelerometer (FPA) also was included as part of the basic airspeed probe and was located inside the probe at the angle-of-attack vane position. The forebody static pressure posi­tion error, determined from flight tests, agreed very well with wind tunnel data. The gear down position error, however, was less than the wind tunnel results. It should be noted that the design concepts of this airplane severely limited the choices of static and total pressure locations and, as a result, the static pressure position errors were quite large but were consistent.

Engine out characteristics were unusual in that rolling and yawing moments were in the direction of the operative engine, and more control effort was required in the roll axis than the yaw axis.

Baseline inflight RCS testing was com­pleted during September of 1978. After several modifications required by the results of the baseline testing were completed, penetration testing against ground based radars and IR systems was begun. IR detections matched predictions well. When a T-38 participated in or duplicated the test profiles, it was tracked to the maximum range of the terminal.

The final phase of testing in a simulated integrated air defense environment was com­pleted during July of 1979. The aircraft demon- strated its VLO capabilities against ground and airborne systems during these tests. Its low acoustic signature was also verified. Measured RCS data correlated well with those measured at Ratscat. Unfortunately, Have Blue 1002 also was lost when it was written-off at Groom Lake on July 11, 1979 after having completed 52 flights.

The project management of the Have Blue program for both the government and Lockheed can be characterized as small, close knit, streamlined, and tailored to the specific needs of the program objectives.

The project was accomplished with daily verbal and secure electronic communications between Lockheed and government program managers supplemented by frequent visits and more formal program reviews. Technical, schedule and cost performance tracking and reporting, although in simplified form, was accurate, timely, and responsive to customer needs.

Evidence of good contingency planning activity on Have Blue included actions taken to recover from: the non-availability of significant GFAE items, workarounds and rapid recovery from a three month IAM strike, and reactivation of flight test operations at Groom Lake.

Contractor interfaces were limited, but very important and well managed on the Have Blue program. General Electric was a subcontractor for J85 engine installation and perfor­mance interface. Lear Siegler was the principle subcontractor responsible for manufacture and field support of the flight control system.

Lockheed's security program was also quite successful during the Have Blue program. It ini­tially started as a "white world" program with mini­mum security classification requirements. This security posture remained until early 1977 when the government realized that a major break­through had been achieved in VLO technology, with great potential to national defense. Subsequently, the program was placed under the "special access required" (SAR) security umbrella. From that point forward Lockheed main­tained total secrecy of the program.

Have Blue was made possible by a great deal of contractor/customer cooperation in suc­cessfully achieving technical breakthroughs. The ability to reduce high risk technical areas on an accelerated basis and in a cost effective manner was demonstrated by the fact that design, manu­facture, and test of this innovative aircraft was accomplished in three years. The VLO aircraft, including engine inlets and exhausts, featured faceted surfaces covered with RAM. Adequate stability and control of the basically unstable air­craft was provided by a quad-redundant fly-by­wire flight control system, with unique Have Blue features and some basic components adapted from the F-16 system. The radical design involved breakthroughs in virtually all aircraft design disci­plines.

Areas of improvement for follow-on pro­grams were defined by the flight test results. In par­ticular: door closure designs were modified during the flight test program to provide more positive closing forces; fin and rudder installations were improved by relocating at the centerline instead of outboard (heat input was then reduced and poten­tial gaps were eliminated); air data probes were placed further forward to reduce air data correction factors; the handling quality requirements in Mil-F­8785B were inadequate for the design of the FCS required for the Have Blue aircraft (Have Blue's fixed side stick control was not totally satisfactory for many piloting tasks; lack of stick motion, stick orientation relative to seat location, and the need for lateral constraints for the pilots were all identi­fied as areas requiring improvement); the large flat plate upper surface of the platypus was unevenly heated when engine power settings were changed and the resulting differential expansion caused distortions which warped the surface (this distortion combined with manufacturing toler­ances associated with the nozzles to make the thrust vector toe-in angles asymmetric; the side forces generated by the asymmetry were picked up by the lateral accelerometer used in the FCS for directional stability augmentation; the resulting commands to the fins caused the aircraft to fly "crabbed", thus requiring the pilot to retrim the air­craft directionally each time that the flight condition changed; the automatic yaw trim feature in the FCS was a partial solution, but the final solution involved elimination of the lateral accelerometer and substitution of a direct measurement of the side slip angle for directional stability augmenta­tion; and the aerodynamic stability and control parameters derived from flight test results with the modified maximum likelihood estimator (MMLE) technique were in good agreement with wind tun­nel predictions except for directional stability which was lower than predicted (aircraft with unorthodox nozzle configurations and low inherent directional stability require special wind tunnel testing tech­niques, including flowing nozzles and inlets).

The Have Blue program was a low cost demonstration of a radically new concept in VLO aircraft design. Have Blue program accomplishments included from a technical standpoint: lowest RCS aircraft in the world by several orders of mag­nitude; VLO infrared signature; VLO visual signa­tures; VLO acoustic signature; and confirmation of complex aerodynamics. From a schedule stand­point: 20 months from prototype contract award to first flight; and 88 test sorties. From a cost stand­point; $43 million total; $32.6 million Air Force/DARPA funding; and $10.4 million from Lockheed. In conclusion, it was determined VLO tactical and strategic aircraft could be designed, produced, and operated.

F-117 DEVELOPMENT AND FLIGHT TEST

During 1978 following an assessment of the prototype results, the Air Force moved ahead with a decision to develop a full-scale production ver­sion of the original Lockheed design under the program codename Senior Trend. By now, mis­sion objectives had been generated and it thus had become possible for Lockheed to narrow its design focus. The attributes of the aircraft's low RCS were ideally suited for ground support mis­sions, and as a result the production aircraft was designed around that very specific requirement.

The new aircraft was to be optimized to covertly penetrate dense threat environments and attack high value targets with pinpoint weapon accuracy. Heavy emphasis would be placed on making the aircraft totally autonomous, totally pas­sive, and as elusive a target as technologically pos­sible; it would not be dependent upon external communications of any kind in order to accomplish its mission. Heavy emphasis would be placed on maintaining an almost negligible RCS, lowering the infrared (IR) signature, reducing the noise (acoustical) signature, reducing visibility (via size and paint constraints), and reducing power plant visible emissions (exhaust particulates and con­trail generation).

The advantages of very low observables or stealth technology, once successfully demon­strated by Lockheed's Have Blue prototypes, quickly led to a full-scale engineering develop­ment contract award from the Air Force on November 16, 1978. The fixed price production contract was signed thirteen months later. It called for five full-scale development (FSD) and fifteen production models of a single-seat, subsonic attack aircraft to be officially designated F-1 17A. Under Program Manager Norm Nelson (working under the direction of the Skunk Works' Vice President and General Manager, Ben Rich), engi­neering on the new aircraft proceeded at a rapid pace, utilizing the data base that had been devel­oped under Have Blue. Principle engineers assigned to the new aircraft were Ed Baldwin, Alan Brown, Dick Cantrell, Bob Loschke, and Bill Taylor. They were all veterans of the earlier low-observables program. Bob Murphy again was tapped to direct manufacturing.

The resulting unusual shape of the F-117 is the end product of the low- observables goal set for the aircraft at the program's beginning. Not surprisingly, it provided the aerodynamic and stability and control engineers with a significant challenge. A major effort was instituted to minimize performance penalties and provide satisfactory flight characteristics. In the end, these efforts were quite successful.

The F-1 17A incorporates a variety of design features to significantly reduce aircraft signature. There are seven different types of observable signatures of concern: radar, infrared, visual, contrails, engine smoke, and electromagnetic emissions. The three signature characteristics providing the greatest potential for exploitation by threat systems are radar, infrared, and  electromagnetic emissions. The F-117 was designed to minimize these signatures. Techniques utilized include highly swept surfaces, radar absorbing structure and materials, gridded inlets, high-aspect-ratio two-dimensional  nozzles, internal weapons carriage, special antennas, and radio frequency techniques.

Since the F-1 17A was a departure from normal aerodynamic design, a considerable effort was made to reduce development risk by using several proven systems from existing aircraft. Examples include: the General Electric F404 engine used in the McDonnell Douglas F/A-18fighter; cockpit components from the General Dynamics (now Lockheed Martin) F-16 and the McDonnell Douglas F/A-18; navigation and attack systems; computer and electronics; off-the-shelf weapons; and a modified fly-by-wire F-16 control system.

All aircraft designs are a compromise in one form or another, with the primary mission objective dominating these characteristics. The primary mission of the F-117A is to penetrate enemy air space, destroy high value targets and survive. As a result, low observability became the dominant design factor. Instead of an aerodynamic shape optimized for high speed or long range, the shaped was faceted for purposes of lowering the radar cross section.

Once Dick Cantrell, the Skunk Works’ chief aerodynamicist recovered from the initial shock that Alan Brown the low-observables expert had given him, he set out to achieve the desired compromises and still have a flyable aircraft. This proved a sizable challenge.

Since low observability – or stealth – was the primary goal, it established the external configuration and in particular the sweep angles of the

wings and tail. One of the larger challenges was to provide as much sweep as possible and still have sufficient aspect ratio for the needed lift-over-drag (L/D) to achieve the required range. This was accomplished by carrying the wing as far aft as possible in order to increase the span. The trailing edge of the delta was notched-out both for low observables considerations as well as to reduce wetted area.

Another major challenge was to provide ade­quate control to achieve the desired maneuver­ability within a reasonable angle-of-attack range for an unstable aircraft in both pitch and yaw. Since horizontal tails were not to be used, large full-span elevons were provided for both pitch and roll control. These were sized to handle the pitch instability which resulted in more roll control power than was needed.

The highly-swept V-tail surfaces were another concession to low observability. The objective was to reduce the height and size yet still provide adequate control for the unstable yaw axis. This required all-movable surfaces.

Alternate means of directional control were investigated, such as split elevon tips, but the V ­tail was preferred to provide adequate control power and reduce drag.

The resulting control configuration was not conducive to low takeoff and landing speeds. The full-span elevons could not be drooped for landing without leading edge devices or another means of pitch control. The solution was to use a drag chute for landing and accept a longer takeoff roll. The brake system capacity was subsequently improved reducing reliance on the drag chute.

Another low observability design considera­tion was to provide very sharp leading edges. This is good for a supersonic airfoil, but not optimum for a subsonic aircraft.

First Aircraft Designed by Electrical Engineers,

Since this was the first aircraft designed by electrical engineers, it was not surprising that a number of aerodynamic "sins" were committed. In fact, when unaugmented, the F-1 17A exhibits just about every mode of unstable behavior possi­ble-longitudinal and directional instability, pitch­up, pitch-down, dihedral reversal, and various other cross-axis couplings. Because of these characteristics, there was no question about what kind of flight control system was required. Piloted simulation showed it had to be a full-time, fly-by-­wire command augmentation system. Any mechanical back-up flight control system would just add weight since pilot control was impossible without stability augmentation.

When the F-1 17A program started, the F-16 was just being introduced into squadron service and F/A-18 was just beginning flight test. Since there was already enough risk in the new Skunk Works program, it proved prudent to use developed, off-the-shelf components to the maximum extent possible not only to reduce risk but also to reduce costs even it it meant some weight penalty. It was decided that the technology developed for the F-16 fly-by-wire flight control system was the only one mature enough to below risk and also rel­atively low cost due to the volume of production. The F-16 actuators, the flight control computer chassis, and power supplies were modified slightly to adapt them to the F-117A. New control laws had to be developed for implementation in the flight control computer, new interfaces for the new air data sensors defined, and a different actu­ator failure detection and redundancy manage­ment scheme developed. The air data probes and pilot's control stick were developed especially for the F-117A.

The pitch axis was implemented as a g-com­mand system optimized for flight path control to support the ground attack mission. An angle-of attack (AoA) limiter was incorporated to prevent departures and the elevons were sized to provide the necessary pitch control power. Because the pitch axis control required large elevons, the avail­able roll control power was much more than needed and the flight control computer incorpo­rated authority limiters to keep the roll rate down to minimize structural loads. The four elevons are used for roll control, and roll rate feedback is used to improve the roll damping.

The directional axis control is of particular interest since the F-117A is directionally unstable over large parts of its operational envelope. Operating the weapon bay doors makes it more unstable so that two large all-moveable fins were required to provide the necessary control power. Because of its shape, the aircraft has very low side force, which means that use of a conventional lat­eral accelerometer feedback for directional stabil­ity augmentation was not practical. The air data probes measure

differential pressure and the flight control computer acting through the fins literally keep the nose aligned with the relative wind, yaw rate, and roll rate. These are fed back to give the desired levels of roll damping and the production of pitch rate and roll rate is fed back to cancel inertia coupling.

The directional axis also incorpo­rates an automatic yaw trim when the gear is up and the pilot is not using the rudder pedals. This feature greatly simplified emergency procedures following an engine failure at lift-off since the pilot only has to retract the gear and concentrate on maintaining AoA and bank angle at the desired values.

Series trim is implemented in all three axes; that is, the stick and pedals are at neutral when the aircraft is in trim. Also, the stick does not move when the autopilot is operating. The autopilot incorporates pitch and roll attitude hold and head­ing hold with control stick steering (CSS). The CSS disables the autopilot while the pilot maneu­vers to a new attitude or heading with the stick and then holds the new attitude when the stick is released. The autopilot also incorporates altitude hold, Mach hold, and automatic navigation.

The net result of the F-117A's flight control system development is an aircraft with compara­ble pitch and roll response to that of conventionally shaped contemporary fighter and attack aircraft within certain boundaries. It is very maneuverable and fully aerobatic.

The F-1 17A flight test program began with a series of flights in the Calspan NT-33A where the suitability of the flight control laws was checked and where the eflects of some aerodynamic varia­tions were investigated. One of the variations assumed that the directional stability was even worse than predicted so that the pilot could see the effects and get some experience.

The actual first flight of the F-117A, under the direction of program Flight Test Manager Dick Abrams, took place in broad daylight (for safety reasons) on June 18, 1981 with Skunk Works test pilot Harold “Hal” Farley, Jr. at the controls. Since the air data probes were of a new design and had exhibited some pitch yaw coupling during vibration testing, it was decided to balance the aircraft to a forward centre of gravity location and turn off the AoA (Angle of Attack) and beta measurements to the FLCC to prevent any possible coupling. Switches were incorporated in the cockpit for the pilot to activate the AoA and beta feedbacks after attaining a stabi­lized flight condition at 15,000 feet. Extenders were installed on those switches so they could quickly be turned on if needed.

Immediately after lift off, it became apparent that the directional stability was significantly worse than predicted and the beta feedback to the flight control computer was immediately switched on. Fortunately, the probes worked exactly as pre­dicted, the aircraft stiffened up directionally, and the rest of the flight was more or less routine. This experience once again showed the value of using in-flight simulation to investigate possible aerody­namic variations prior to first flights. As it turned out, both the directional stability and directional control power were less than predicted and the fins were increased in size by 50% to get back up to the original predicted levels of stability and con­trol power.

Stealthy air data sensors were a unique design challenge. Ideally, flush-mounted "invisi­ble" sensors such as pressure ports on the fore body could be used. Unfortunately, Skunk Works engineers could not find four independent loca­tions for AoA and beta (actually sixteen places on the aircraft) without all variables being functions of all others (AoA, beta, Mach, and q) which would require too many correction coefficients in the flight control system. As a result, four probes (for quadruple redundancy) are utilized.

These probes were made of special materi­als in an unconventional shape which required a lot of test flying to get good position error calibrations on AoA, beta, airspeed, and altitude. Initial designs cracked when the heat was turned on, calibrations shifted, and quality control problems in probe manufacture required testing and retest­ing until deficiencies were identified and elimi­nated.

High angle-of-attack flight test to verify the adequacy of the AoA limiter was approached with extreme caution and took a lot of test time to com plete. There were two reasons for this. The first was that all the wind tunnel tests showed that pitch-up would occur at high AoA. Free flight test­ing of unpowered models verified that pitch-up and deep stall were possible, but that there were no identifiable spin modes. The second reason was that the high AoA tests were done without a recovery chute. There was never any intention to deliberately depart the aircraft because the small model free flight tests showed that all departures eventually wound up in a deep stall. Since there was a possibility that the normal drag chute would effect a recovery from that condition, it was decided that the high AoA testing could go forward without the installation of a special recovery chute.

Flutter testing became a bigger effort than originally planned as a result of a chain of events that started with the aircraft's original design. The big surprise came when an Air Force test pilot was flying a stores compatibility test mission. While performing a sideslip maneuver, explosive flutter of the left fin occurred. The fin was almost com­pletely lost and the pilot brought the F-117A home with difficulty, and with considerably less direc­tional stability than before. As noted earlier, the original fin was 50% smaller than the final version due to directional stability considerations. The cor­rective action was to increase the fin area by extending the fin edges, but without changing the size of the fin box structure. The net result was a reduction of fin stiffness.

Prior to this incident, an extensive flight flut­ter clearance program had been conducted, the results of which indicated there were no aeroelastic problems of any significance. Analysis per­formed subsequent to the incident revealed the problem to be aflutter mode, known as the "hump" mode, that was considered at first to have a flat damping trend that was actually found to be poten­tially flutter critical. In fact, the aircraft had been cleared to and flown many times at the incident flight conditions before the fin flutter problem was encountered.

The subsequent investigation also revealed that this critical mode, along with others, was highly affected by rudder post bearing friction which had masked the criticality of the mode in early flutter testing. In order to test the absolute worst case (no friction) low friction roller bearings were installed in the test aircraft replacing (for test purposes only) the standard journal bearings on the rudder posts. The testing continued and veri­fied satisfactory stability levels of the fin modes out to the desired speeds.

However, the flutter test results did show a significant loss of damping in sideslips. This loss, when coupled with the masking effect of the bearing friction, led to the fin flutter incident. Since then, the fins have been replaced with new fins made of graphite-thermoplastics. These new fins are much stiffer and have demonstrated a very large flutter margin.

The requirements for structural flight tests were fairly typical of a normal aircraft. All of the standard maneuvers had to be performed out to the structural limits of the aircraft. Two problems arose. The first was due to a change in fuel transfer sequencing which reduced the inertia relief pro­vided by wing fuel. This stopped flight testing until a strengthening of the aft wing attach points was accomplished. The second problem occurred when the aft fuselage loads measured in flight test turned out to be significantly higher than predicted by analysis. After much speculation without a rea­sonable explanation, it was decided to install pres­sure taps on the rear of the aircraft and compare the pressure distribution results from flight test with those from the wind tunnel. The tunnel data matched the flight test results. This led the Skunk Works engineers to look for other sources of load­ing which turned out to be the effect of the unusual exhaust system. Trying to take the thrust and squeeze it out of a high aspect-ratio two-­dimensional nozzle resulted in some unusual down bending moments which had not been accounted for in the analysis.

Engine tests also were conducted due to the unorthodox design of the exhaust nozzles. Problems were known to occur when a circular thrust pattern is expelled from a round engine then turned and flattened to exhaust out of a long, flat, two-dimensional nozzle. Hot spots, loads, tailpipe distortion, etc. had to be overcome requiring many tests and engineering exercises before a suitable exhaust configuration was developed.

One other area of concern also was the dis­tortion produced by the inlet grids and the engines' ability to tolerate this. Analysis indicated the grids should be no problem, but a little skepticism remained. The distortion levels in fact proved to be less than expected and the grids actually acted as flow straighteners giving a consistent source of air to the engine throughout the entire AoA and beta range of the aircraft.

A great deal of icing tunnel work was con­ducted as a result of inlet icing concerns. This indi­cated that the inlet grids not only looked like a giant ice cube tray, but acted like one, as well. Later, in-flight icing tests using the normal buildup approach were undertaken. It was concluded that a special de-icing system would have to be devised. As a result, the aircraft was equipped with a special wiper system complete with alcohol dis­pensing capability.

With the F-117A's handling qualities so dependent on good air data, the criticality of pitot-­static probe deicing was obvious. Requirements forced upon the aircraft by the low-observables engineering group made the design of the probes and their deicing system a serious challenge. Many different designs were tunnel tested before an acceptable one was found. Airborne results proved the systems worked well. The only real modification required was the repositioning of the icing detector within the engine inlet.

Avionics

The development of the F-117A avionics systems continues to be an ever-evolving program of changes, upgrades and improvements. Early FSD testing brought the initial avionics architecture to an initial operations capability during October of 1983.

Full capability development after initial operational capability brought most avionics systems to maturity. The Weapon System Computer Subsystem (WSCS) upgrade brought needed computational capability improvements to the F-1 17A while the Offensive Combat Improvement Program (OCIP) brought additional pilot situational awareness and reduced workload to the night single-seat attack mission. Ongoing programs and planned future capabilities will continue the evolutionary process.

The F-1 17A avionics systems development followed the same principles as the main airframe program: minimize development risks by using off-the-shelf hardware where possible, modify existing equipment where feasible, and invent new systems only when absolutely required. In this regard, the program was highly successful; however, some of the off-the-shelf and modified systems provided inadequate performance until additional improvements were made.

The stealth requirement for avionics design was just as stringent as the basic airframe and required substantial integration of the airframe/avionics systems. Avionics system designs must be aware that electromagnetic emissions from the aircraft, such as radar, are just as vulnerable to detectability as any of the other observables.

Considerable effort went into the incorpora­tion of features to preclude any emission from the F-117A. In the aircraft’s stealthy mode, the F-117A is incapable of any emission which may cause detection; i.e., UHF, IFF, radar altimeter, TACAN, etc. The laser target designator is the only exception and considerable forethought went into the amount of time and conditions under which it may fire. As a result, the avionics design of the F-117A operates independently of active emission and relies completely on passive systems for navigation, target acquisition, and weapon delivery.

As a result of the need to minimize develop­ment risks and maximize the use of off-the-shelf equipment, the cockpit became a mix of then state of-the-art glass cockpit technology and Century series aircraft type switches, lights, and dials. Much of the equipment came from front line aircraft, such as the F/A-18, but the aircraft includes components from practically every aircraft built since the T-33. Examples include parts taken from the SR-71, the P-3, the C-130, the L-1011, the S-3, the F-104, the P-2, and many others.

The main cockpit layout is the now familiar arrangement of Multifunction Displays with a HUD and center sensor display. During F117A conceptual design, the F/A-18 was the only US fighter using a similar arrangement. Due to limitations within the computer system and for risk reduction, most of the warning lights, indicators, and aircraft systems switches are external to the avionics architecture and have no provisions for control from the displays.

Most of the cockpit systems are derived from the F/A-18. These include the early multifunction display indicators, the HUD (Head Up Display), fuel gauge, engine instruments, stick grip, and throttles. The sensor display is provided by Texas Instruments, and is derived from the Vietnam-era OV-OD andP-3C programs

The original avionics architecture was distributed real-time processing system which used three Delco M362F computers from the F16 inter­connected with a dual redundant MILSTD-1553 data bus. The computers interface with the dis­plays, controls, INS, autopilot, stores management system (SMS), and the sensor systems. The weapon delivery computer (WDC) was the system executive. Besides providing over-all control, the WDC serviced and updated the cockpit displays, performed the weapon delivery ballistics calcula­tions, interfaced to the various sensor systems, and controlled the data distribution. The navigation con­trol computer performed all navigation and control functions including the inertial measurement unit, the control display unit, navigation steering, flight director steering, position update, attitude heading reference system integration, and the TACAN and ILS interface. The third computer provided control and data processing for an additional sensor sys­tem and was used as a back-up computer if one of the other two should fail. A data transfer module interface unit was provided to load preflight mission data via a data transfer module from the mission data planning system.

The underlying operating principle of the avionics system is the cueing of the sensor to the target via precision navigation system thus providing updated target information for accurate weapons release.

Given suitably accurate information about the location of a target and given the excellent accuracy of the onboard inertial system, the com­puter system cues the infrared (IR) system to the target. The field of view of the IR system being small, requires not only that the position accuracy of the INS be very good but also that the target location data be very accurate so that the IR sys­tem can be pointed very accurately. This was the critical program issue for the avionics system. Would sufficiently good target information be avail­able and would system performance be good enough to be able to find a target at night looking only through a small IR window? And, of course, could weapons be delivered accurately enough to destroy the target? Given that the desired target is within the field of view of the IR system, i.e., the pilot can see it on the sensor display, the pilot refines the aiming, designates the target, and con­sents for weapon release, which occurs via the SMS at the appropriate time.

The infrared acquisition and detection sys­tem (IRADS) was built by Texas Instruments. This was an off-the-shelf single turret system that was adapted to a twin turret design due to the unique "in contour" mounting requirements of the F-117A. Stealth characteristics of the F-1 17A required that the unique exterior shape be maintained. Since there was a need to be able to see from just above the horizon to well behind the aircraft, a forward ­looking IR (FLIR) turret and a downward-looking IR (DLIR) were required. This need doubled the size of the servo controller unit and the video tracker unit. Due to the mounting arrangement in the aircraft, the DLIR is inverted relative to the FUR and thus required the video to be inverted electron­ically when displayed to the pilot. This led to some interesting calibration and alignment problems. But, any turret may still be mounted in either posi­tion.

The F-1 17A employs screens over the FUR and DLIR cavities to maintain its low observable signature. The original screens were to be etched metal units from a process not unlike printed circuit boards. These screens proved unable to take the acoustic environment of the cavities, in particular the DLIR cavity where the screen broke on its first flight. This breakage led to a vibration and acoustics investigation of the cavities and redesign of the screens. Both FUR and DLIF cavi­ties now feature acoustic shrouds to limit acoustic affects. The screens are redesigned woven wire units capable of handling the acoustic loads.

The weapon bays are each equipped with a trapeze for loading and raising the weapons. Early concerns for possible damage to the aircraft and bay doors from fin scheduled weapons, like the GBU-10, required that these weapons be dropped trapeze down. This was a major detectability prob­lem for the aircraft. Early weapon certification was performed in this configuration. Later efforts by the aerodynamics department indicated that ade­quate clearance could be maintained with the trapeze up. This reduced the exposure times by more than a factor of 5. However, some weapons did end up with small speed restrictions due to weapon bay dynamics or airflow disturbances in the near flow field.

The Delco M362F computers were long known to be just adequate for the computational tasks of the F-117A. At the time of their selection, the MIL-STD-1750A instruction set computers that were on the horizon were still too big a risk. The first several years of software development prior to first flight were just as involved in getting the OFP to fit and run in the computer as much as implementing capabilities.

During 1984 the weapon system computational subsystem (WSCS) upgrade program was started to replace the Delco M362F computers. The IBM Federal Systems AP-102 MILSTD-1750A computer was selected. This was a repackaged version of the same computer used in the Rockwell International Space Shuttle.

The architecture for the WSCS computer upgrade was similar to the WDC version with some improvements. Three AP-102 computers were used with each computer controlling a dual redundant MIL-STD-1553 bus for a total of three in the system. The onboard systems were divided between buses 1 and 2 with the third computer and bus held as spares for growth. A unique high speed bus was incorporated for direct communi­cations between the three AP-102s and the expanded data transfer module interface unit.

Weapons Capability

The aircraft was also enhanced by the deci­sion to expand the weapon release capability from the use of a single weapon bay per pass to the abil ity to use both weapon bays. This was a significant change as the weapon bay doors, actuators, hydraulics, SMP, and cockpit controls all required redesign and modification.

With the growth potential of the WSCS com­puters, the program was positioned to embark upon introducing a number of new capabilities. Among the first of these was the incorporation of a significant new weapon capability. The GBU-27 laser guided bomb (LGB) brought new levels of accuracy and target penetration to the guided weapon inventory.

The GBU-27 was the marriage of a modified GBU-24 low-level laser guided bomb seeker, sometimes known as Paveway 3, and the BLU 109 improved 2,000 pound warhead. Changes to the GBU-24 seeker included modified canards to fit inside the F-1 17A weapon bays and a firmware change for trajectory shaping .

The GBU-27 features two guidance modes, each optimized to achieve the best penetration angle for horizontally or vertically oriented targets. The trajectory for vertically developed targets is essentially the ballistic path. For bunkers, rooftops, or any other target of horizontal orientation, the GBU-27 flies a commanded pitch down after release to strike the target in as near a vertical atti­tude as possible. These modes, coupled with the penetrating warhead and excellent accuracy of the weapon, caused the GBU-27 to become the pri­mary F-1 17A weapon in Desert Storm. While the exact circular error probable (CEP) of the GBU-27 is not presently releasable, the video tape aired during Desert Storm depicts weapons consistently striking the center of the crosshairs.

The most recent major avionics development is the offensive combat improvement pro­gram (OCIP). Based on the WSCS computer sys tem, a number of new systems and capabilities were added. These were color cockpit displays, a digital tactical situation display or moving map, a 4D flight management system (FMS), a new data entry panel, a display processor, autopilot improvements for vertical flight path control, an auto-throttle system, and a pilot activated auto­matic recovery system.

The OCIP program provides no new capabili­ties for target acquisition and attack. What it does do, is provide the pilot greater situational awareness, reduces pilot workload by allowing the FMS system to fly complex profiles automatically, pro­vides automatic speed and time over target con­trol, and provides unusual attitude recovery upon pilot command.

For the future, a number of major new sys­tems are planned for the aircraft. A new IRADS system is now in production refit. Goals for this program are to double the acquisition range of the IRADS system and increase the range and life of the laser. To replace the aging, out of producion SPNGEANS inertial navigation system, a ring laser gyro (RLG) system will be installed. This system will be supplemented by the addition of a global positioning (GPS) unit.

Although it incorporated many new technologies, the F-117A, typical of a Skunk Works prod­uct, was developed in significantly less time and for less cost than a comparable conventional fighter. This was achieved within the tight security of a special access program using streamlined management methods. The Air Force's Aeronautical Systems Division and Skunk Works personnel worked in a non-adversarial, problem-solving atmosphere with a minimum number of people. In addition, the use of proven components from other aircraft reduced risk and gave the confi­dence to proceed concurrently with development and low rate production.

Original F-117A program costs can be bro­ken down as follows: Total development to date­$2 billion; Procurement-$4.265 billion (Total flyaway -$2.515 billion; Unit flyaway  -$42.6 million); Military construction -$295 million; Total program cost -$6.560 billion. These costs include all government furnished equipment, including engines.

Part 3 to follow in the Combat Diaries: Production and Operational Service of the F-117