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]
To this point, Project Harvey was unclassified, and stealth was freely talked about. In June 1975,
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 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).
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.
To keep the development time short, as many existing components as possible were used. In addition to the F-16 fly-by-wire control system, the Have Blue aircraft used F-5 ejector seats, landing gear, and cockpit instruments. The J85-GE-4A engines were supplied by the navy from its T-2B trainer program. The Have Blue aircraft were built by hand, without permanent jigs
Two test pilots were selected to fly the Have Blue. Lockheed test pilot William Park would make the first flights. Park was so highly regarded at the Skunk Works that Ben Rich obtained a special exemption from the air force so he could be chief test pilot for the Have Blue. (He was not a test pilot school graduate, nor did he have an
