Dark Eagles: A History of the Top Secret U.S. Aircraft Read online

  Only for certain special cases can [radar cross-section] be calculated rigorously; for most targets [it] has to be inferred from the radar data… Only a rough estimate of the cross-section of such targets as aircraft or ships can be obtained by calculation. Even if one could carry through the calculation for the actual target (usually one has to be content with considering a simplified model) the comparison of calculated and observed cross-section would be extremely difficult because of the strong dependency of the cross-section on aspect.[342]

  By the mid-1950s, basic research was underway in the United States on understanding the sources of a plane's radar cross section. A team headed by Bill Bahret at the Wright Air Development Center did much of this work.

  A large anechoic chamber was built to test the radar return of different shapes.

  By the late 1950s, Bahret and his team felt they understood the sources of large echoes. Once they knew this, the obvious next step was to reduce the echoes. This would have two advantages in terms of electronic countermeasures: the amount of power needed to hide the plane's echo would be reduced, and, for a given jammer, the effectiveness would be increased. As yet, there was no intent to build a plane invisible to radar.

  A second part of this effort was development of radar-absorbing material (RAM). Since World War II, Dr. Rufus Wright and a team at the Naval Research Laboratory had been working on RAM. Together with Emerson and Cuming Incorporated, a plastics manufacturer, they had developed a practical RAM. The material was in the form of thin, tilelike sheets. It was pliable like rubber and could be cut and formed into any shape. The navy lost interest in the project, and Wright went to the air force.

  The air force was very interested — the RAM was both thin and strong and, therefore, could be attached to the skin of an airplane. After tests with scale models, it was decided to cover a T-33 jet trainer with the RAM. This was to verify the echo reduction predicted by the scale tests. The project was code-named "Passport Visa," although the white-painted T-33 was better known as "Bahret's White Elephant."

  The Passport Visa T-33 was completely covered with the RAM. This included the skin, wing tanks, and control surfaces. The plane was only an experiment, with no operational applications in mind. The air force test pilot selected for the project was Capt. Virgil "Gus" Grissom. (The following year he was selected as a member of the first group of astronauts; he would later die in the 1967 Apollo 1 launchpad fire.) Test flights began in late 1958. The results were mixed — many of the echo reductions were confirmed, but the T-33's flight characteristics were degraded by the added thickness of material. Grissom found the plane was hard to control; it slid in turns, overdived, and coming in for a landing it behaved like a roller coaster.[343]

  Clearly, a plane's radar cross section could not be reduced simply by covering it with RAM. It would have to be designed in. Despite all these efforts, there was no simple way to calculate the radar cross section of a plane. With the computers and theoretical models of the time, too many factors entered into the calculation for it to be a practical possibility.

  This meant designers would have to take a crude cut-and-try approach.

  When Kelly Johnson wanted to test the radar cross sections of the A-12 and D-21, he first used small models. Then full-scale mock-ups were built and tested. From this data, the final designs were developed. Still, it was not until the planes actually took flight that the true radar cross section could be determined.

  Such efforts could be made for Black airplanes. Reduced radar cross section had little impact on the design of operational aircraft. Until Vietnam.

  PROJECT HARVEY

  The air defenses of North Vietnam required a fundamental change in tactics. A typical Rolling Thunder strike was composed of sixteen F-105D bombers. The force needed to protect them was made up of eight EF-105F "Wild Weasels," which attacked SAM sites, and six F-4D escorts against MiGs. Even though each F-105D carried individual ECM pods, two EB-66 jamming aircraft would also accompany the strike force. The EB-66s, in turn, each required two F-4Ds as protection against MiGs. Thus, to protect sixteen bombers, a total of twenty jamming and support aircraft were needed since the support aircraft themselves needed protection.[344] The net result was that most of the available aircraft were diverted from attack missions to defensive roles.

  The revolution in air defense caused by SAMs would be underlined in the October 1973 Yom Kippur War. The Egyptian and Syrian armies that attacked Israel were equipped with the new SA-6 Gainful SAM. Mounted on a tanklike transporter, it could move with the frontline troops. The Israeli air force did not have the ECM pods needed to counter the SA-6 and suffered heavy initial losses. During a single strike against a Syrian SA-6 battery, six Israeli F-4Es were lost. The air defenses also prevented the Israeli air force from providing close air support to ground troops.[345]

  Although the Israelis overcame the early setbacks, the SA-6 was a clear warning. As long as U.S. countermeasures and tactics were specifically tailored to enemy radars and SAMs, they would be vulnerable to technological surprise. The Soviets were then in the process of deploying a new generation of SAMs. In the event of a war in Europe, NATO forces could suffer the same huge losses as the Israelis had. Many academics theorized the end of manned aircraft was at hand. Technical advances in radar design, such as the traveling wave tube and computers, had increased power and the ability to defeat ECM. Any new technological advances in ECM would be countered by improved radars.

  Others realized that a new set of assumptions was needed. Countermeasures had always been based on overpowering the radar. Even Black aircraft with reduced RCS — the A-12 and Model 147–154 drones — used ECM equipment for protection. The key was not more powerful ECM, but to make the RCS a primary design consideration. It would be eliminated, not simply reduced. With no echo, the radar would be blind. No radar would provide early warning as the aircraft approached; no radar would direct MiGs, antiaircraft guns, or SAMs. There would be no need for support aircraft. Air defenses would revert to the 1930s, against an enemy traveling at near supersonic speeds.

  The problem was the amount of RCS reduction needed. A tenfold reduction would only shorten the range at which a plane could be detected. A hundredfold RCS reduction would merely degrade the effectiveness of radar. It would take a thousandfold reduction of a plane's RCS to make it undetectable to radar.[346]

  Moreover, to be fully effective this reduction in RCS would have to be combined with other design features to reduce detectability. Just as the aircraft could not reflect any radar signals, it also could not emit any — no bombing radar or ECM transmissions. The infrared emissions from the engine would have to be hidden. The engine could not produce smoke. The airplane also would have to be quiet; the sound of a plane gives warning of its approach. The plane could not produce a contrail — this had been a major problem with the Model 147 drones. The final problem was visibility.

  Although true optical invisibility was not possible, efforts had to be made to reduce the distance at which the plane could be seen. One problem was "glints" from the canopy. A plane could be seen at a distance of five to ten miles; the reflection of the sun could be seen at a distance several times that.

  The effort to make this possible became known as "Project Harvey," after the invisible rabbit in the play and film of the same name.[347]

  In 1974, the Defense Advanced Research Projects Agency (DARPA) issued requests to five aircraft manufacturers to study the potential for developing aircraft based on a minimal RCS. They were to design a small, low-cost test aircraft to demonstrate the possibilities. It was called the "XST," for "experimental survivable testbed." The companies were General Dynamics, Northrop, McDonnell Douglas, Grumman, and Boeing.[348] All had recent experience with fighter design and manufacturing. Lockheed, which had not built a fighter since the F-104 program of the early 1960s, was not included.

  By early 1975, Ben Rich had learned of the program. He had been involved with the work Lockheed had done on the Dirty Bird U-2s,
the A-12, SR-71, and D-21, and knew it gave Lockheed the experience needed for the DARPA project. Rich obtained a letter from the CIA granting permission to discuss the reduced RCS work of the earlier projects. This was part of the request to DARPA for Lockheed to be included in the program. The effort was successful, and Lockheed joined the design competition.

  The keys to Lockheed's efforts were Lockheed mathematician Bill Schroeder and Skunk Works software engineer Denys Overholser. They produced the conceptual […] that allowed a stealth aircraft to be designed.

  Schroeder went back to the basic equations derived by Scottish physicist James Clerk Maxwell a century before. These described how electromagnetic energy was reflected by a surface. Maxwell's equations were revised at the turn of the century by German electromagnetic expert Arnold Johannes Sommerfeld. For simple shapes, such as a cone, sphere, or flat plate, these formulas could predict how radar signals would be reflected. In the early 1960s, a Soviet scientist named Pyotr Ufimtsev developed a simplified approach which concentrated on electromagnetic currents set up in the edges of more complex shapes, such as disks.

  The Maxwell, Sommerfeld, and Ufimtsev equations still could not predict the RCS for a complex shape like that of an airplane. Schroeder's conceptual breakthrough was to realize that the shape of an airplane could be reduced to a finite set of two-dimensional surfaces. This reduced the number of individual radar reflections that would have to be calculated to a manageable number. Rather than a surface made of smoothly curving surfaces, the whole airplane would be a collection of flat plates, which reflected the echo away from the radar. This system of flat, triangular panels became known as "faceting," because it resembled the shape of a diamond.

  Schroeder asked Overholser to develop a computer program that could predict the RCS of a faceted aircraft shape. It took only five weeks for the Echo I program to be completed. Now, with the faceting concept and the Echo I program, it would be possible to predict the RCS of an aircraft. Possible designs could be tested and refined in the computer. The way was clear to build a truly invisible aircraft.[349]

  The initial design was dubbed the "Hopeless Diamond." When Overholser presented a sketch of the design to Ben Rich on May 5, 1975, Rich did not quite grasp what had been achieved. Rich kept asking how big the radar return of a full-size aircraft would be — as large as a T-38, a Piper Cub, a condor, an eagle, an owl? Overholser gave him the unbelievable answer.

  "Ben, try as big as an eagle's eyeball."

  The Hopeless Diamond met with a frosty reception by Kelly Johnson, who was still working as a consultant to the Skunk Works. Having built some of the most graceful planes ever to take to the skies, he was not impressed with this alien design. Johnson's opinion was shared by many of the Skunk Work's senior engineers and aerodynamicists. They preferred a disk-shaped design — a real flying saucer.

  A disk had the ultimate in low radar cross section. The convexed surface of the disk would scatter the radar signals away from the source. The problem with a disk-shaped aircraft was control. The WS-606 project of the mid-1950s was to have relied on a large spinning fan to provide gyroscopic stability, while directional control was to be provided by thrusters on the rim of the disk, fed by a complex network of ducts. A disk also has poor aerodynamic qualities, such as high subsonic drag. How to make a flying saucer fly was the problem, and, as Rich later noted, "The Martians wouldn't tell us."

  Johnson thought the radar return from the Hopeless Diamond would be larger than that of the D-21. A ten-foot mock-up of the Hopeless Diamond was built. On September 14, 1975, it was tested against the original mock-up of the D-21. The Hopeless Diamond had a radar return one-one thousandth that of the D-21. This was exactly that predicted by the Echo I program.[350]

  By October 1975, the DARPA competition had been reduced to the Lockheed and Northrop designs. The Northrop XST was a pure delta wing with a faceted fuselage. The rear of the wing was swept forward, giving it the appearance of a broad arrowhead. The single, large intake was located above the cockpit. To mask the inlet from radar, a fine mesh screen was used. Two tilted fins shielded the engine exhaust.[351]

  The Lockheed design, in contrast, had sharply swept-back wings—72.5 degrees. The rear of the fuselage came to a point; with the swept-back wings, this gave it a W shape. The two intakes were placed on the sides of the aircraft and were covered with grills. This allowed a higher speed than the screen on the Northrop XST, which was not usable above Mach 0.65.

  Twin inward-canted fins shielded the exhaust. These were slotlike and were called "platypus" nozzles.

  The XST design philosophy was to have the lowest possible radar return from the front and bottom of the aircraft. As the plane would fly at high altitude, the top was not considered as important.[352]

  In December 1975, Lockheed and Northrop built one-third-scale models of their XST designs. These were shipped to the Gray Butte Microwave Measurement Range in New Mexico and mounted on poles for radar signature testing. A second series of tests was run in January 1976 after minor modifications had been made. This was followed by a full-scale RCS model, which was tested at the air force measurement range at White Sands, New Mexico.[353]

  The results were a breakthrough in aircraft design. During an early outdoor test, the radar could not detect the model. The radar operator thought it had fallen off the pole. Then a reflection was picked up — from a crow that had perched on it. At the White Sands tests, the reflection from the pole was many times brighter than the model. It was also discovered that the model had to be kept clean. Bird droppings increased the return by 50 percent. The series of measurements showed that the Lockheed design had one-tenth the radar return of the Northrop model.[354]

  In April 1976, Lockheed was named the winner. It was to build two XST aircraft for aerodynamic and RCS testing. The contract was for $32.6 million from DARPA and the air force. Lockheed had to add another $10.4 million of its own money. The latter represented a big gamble on Lockheed's part. The Skunk Works had spent much of the late 1960s and early 1970s in an unsuccessful effort to sell a series of fighter designs. At the same time, losses in the L-1011 airliner program had brought Lockheed to the edge of bankruptcy. Even with federal guaranteed loans, Lockheed was still near failure in 1975 and 1976. But the $10.4 million investment was to bring in several billion dollars.[355]

  STEALTH GOES BLACK

  To this point, Project Harvey was unclassified, and stealth was freely talked about. In June 1975, Defense Daily carried a report that the air force was developing a small stealth fighter.[356] In August 1976, Aviation Week and Space Technology carried a brief story that Lockheed had won the development contract for a stealth fighter demonstrator.[357] The 1977-78 edition of Jane's All the World's Aircraft carried a one-paragraph item that a "small" stealth fighter was being built by Lockheed and was expected to fly in 1977.[358] A June 1977 issue of Aviation Week and Space Technology revealed that the "Stealth Fighter Demonstrator" used J85 engines, that Kelly Johnson had acted as a consultant on the project, and that it would make its first flight in 1977.[359]

  Soon after work on the XST started, Jimmy Carter was elected president.

  The program attracted the attention of the defense undersecretary for research and engineering, William J. Perry. The results of the model RCS tests indicated that stealth had the prospect of a fundamental breakthrough.

  As a result, the XST became a Black airplane in early 1977. Control was transferred from the largely civilian-staffed DARPA to the Air Force Special Projects Office. The word "stealth" also disappeared; it could not be used in any public statement or in an unclassified context. The program was pushed, even as the defense budget underwent major cuts.[360]

  The program also received a new two-word code name. Unlike Aquatone, Oxcart, and Tagboard, it was a computer selected designation. Because it was an aircraft technology development project, the prefix "have" was given to the program. This new Dark Eagle became the "Have Blue."

  HAVE BLUE

 
Have Blue was the first airplane whose shape was determined by electrical engineering, rather than aerodynamics. Not surprisingly, it had the aerodynamics of a household appliance. The design was inherently unstable in all three axes — pitch (longitudinal stability), roll (lateral stability), and yaw (directional stability). Every aircraft ever built had curved wing surfaces. On the Have Blue, the wings were made of long, wedged-shaped flat plates, meeting at a sharp edge.

  The first Have Blue prototype would be used for aerodynamic and control tests. It had a long (and unstealthy) nose boom for the air-speed system.

  Because of the design's instability, it used a fly-by-wire control system, built for the F-16A, that was modified to make the Have Blue stable in all three axes. (The F-16 was unstable only in the pitch axis.) Stability was critical if the design was to be developed into an attack aircraft; an unstable aircraft cannot bomb accurately.

  The second Have Blue prototype would be used to demonstrate the design's stealth qualities. It had an operational air-speed system and lacked a drag chute. Development work was also done on improved RAM and better ways to apply it. The prototype would also test the practical details.

  Unlike an RCS model, a real airplane has landing gear doors, a canopy, a fuel-fill door, screws, and vents. Any of these could greatly affect the plane's RCS. On the second Have Blue, greater care would be taken to insure that all gaps were sealed.

  The Have Blue aircraft were 38 feet long and had a wingspan of 22.5 feet.

  This was 60 percent of the size of the planned production aircraft. They would have a top speed of Mach 0.8 and were powered by a pair of J85 engines.

  These lacked afterburners to reduce the infrared signature. There was no weapons bay and no inflight refueling equipment. Weight of the Have Blue was 12,500 pounds, and it was limited to a one-hour flight time.