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

  EXPOSED

  154P-4 was on a long-range test flight when a warning light came on at the Holloman control center. A control surface actuator had failed, and the drone was seconds from going out of control. Ground control triggered the parachute recovery to save the drone, but the troubles of 154P-4 were only starting. It was descending toward the Los Alamos complex, during the lunch hour. Thousands of people saw it coming down under a 100-foot-diameter parachute. Suddenly, there were bright flashes as doors on its underside were blown off, and the bags used to cushion the landing impact inflated.

  The drone missed a three-story building and landed on a road just inside the complex. A noontime jogger was starting his fourth lap when 154P-4 came down in front of him. The tip of the right wing slid under the guard-rail at the edge of the road.

  The 154P-4 was undamaged, but there was still worse to come. Although the landing site itself was secure, only a few feet away was the fence marking the boundary with public land. Just across a narrow canyon was a residential area, and word quickly spread of what had happened. A few Los Alamos employees realized the strange airplane was probably classified, and hurriedly covered it with tarps. But before 154P-4 could be hidden, the press arrived and were able to photograph it from the perimeter fence. By the time an air force-Ryan recovery crew arrived an hour after landing, the fence was lined with reporters and television crews. Others had climbed trees for a better look. The Model 154 Firefly had made a very public debut.

  The Albuquerque Journal carried the headline "Secret Something Falls to Earth." The article said that "the emergency descent by parachute of a super secret unmanned aircraft… ripped some security wraps off 'Firefly.'"

  Two photos of 154P-4 were published in the Los Alamos Monitor, despite air force requests that they be withheld.[293] The New York Times carried a small, two-sentence report on the incident on page 24.[294]

  The air force released a cover story that the Firefly was simply a "relatively high altitude test of an Air Force target drone" (the story originally developed for the "Q-2D" test flights in 1960). No one was fooled; it was clear the Firefly was a secret project. As the Albuquerque Journal said: "If Firefly is simply a high altitude target drone for testing missile systems, the reason behind the strict Firefly or drone aircraft security lid remains a mystery."[295]

  The failure was traced to use of low-temperature solder. When it got hot, the solder softened, the wire pulled free, and the secret was out.

  FINAL TEST FLIGHTS

  Following the accident, the 154 was grounded for several weeks while an investigation was conducted. When flights resumed, they were restricted to the White Sands area. A flight by P-5 in September was successful, and the range restrictions were n'rteci. During subsequent '154 flights CH-3 helicopter was placed on alert at Holloman Air Force Base. Should the 154 land outside the recovery zone, the crew would fly out and secure the drone before it could be further compromised. Two long-range flights were made by P-4 without problems. Another flight on November 21, 1969, almost resulted in more publicity.

  The ill-fortuned P-4 was flying over the Navajo reservation in northern Arizona when there was a circuit failure. The 154 went into an automatic recovery. The DC-130 crew saw a group of people around the drone. The plane buzzed them to warn them off. When the recovery crew arrived, they were pleased to discover the 154 had been secured by the tribal police. One of the people who found it was an ex-air force sergeant who realized the drone was a secret aircraft. He called the tribal police, saw that a perimeter guard was set up, gathered up the parachute, and took charge in a very professional manner.

  Three long-range navigation flights were conducted in early 1970, followed by a final series of eight tests at Edwards Air Force Base between August and December 1971. These flights reached altitudes of 81,000 feet.

  The sixth and seventh flights also involved simulated Soviet Fan Song B and E radars at the navy's China Lake facility. These were the radars used by the SA-2. It was found that the 154 was nearly impossible to detect. The drone's small radar cross section alone was enough to protect it. By December 1971, the problems with the 154 were finally solved. The drone program had achieved a capability rivaling that of the U-2.[296]

  But now it had no place to go.

  THE END OF THE FIREFLY

  The Model 154 had been designed for overflights of China. This required a low radar return to prevent detection, a very high altitude to avoid interception, and a precise navigation system to cover the target. With President Nixon's trip to China, this possibility ended. There was no interest in using the 154s over North Vietnam because of the success of the 147SC low-altitude drones. There were suggestions that the 154 be used over Cuba during the spring of 1972. Following the outbreak of the Yom Kippur War in October 1973, use of the 154 was again proposed. Once more, it was turned down. The drones were placed in storage, then scrapped.[297]

  Although the 154 had its share of problems, these did not cause its down-fall. The Firefly had been overtaken by events. Starting in 1969, the high-altitude drone mission had started to fade. The 147T program was cut short, while the number of 147H missions in 1970-71 was reduced to a mere h a n d f u l. Nixon's halting of Chinese overflights ended a mission that was already coming to a close. Without its primary mission of Chinese overflights, however, the Model 154 Firefly was left an orphan. It was not needed for flights over North Vietnam, while the other possible targets were covered by SR-71 overflights. Both the A-12 and the 147 drones had to wait for an opportunity to show what they could do. When given that chance, these Dark Eagles excelled. The Model 154 was never given that chance.

  These circumstances also affected another Black drone.

  CHAPTER 6

  The Last Blackbird

  The D-21 Tagboard

  Rid plans of doubts and uncertainties.

  Sun Tzu ca. 400 B.C.

  The prototype SR-71 made its initial test flight on December 22, 1964. Its takeoff and landing at Palmdale was a public, very "White" debut. On hand were a number of Lockheed dignitaries. Once the SR-71 landed, they boarded a transport and took off. They did not, however, fly west, toward Burbank. Instead, the plane headed east, to Groom Lake.

  Another of the Blackbird family was also to make its first, very Black, test flight that day.

  EARLY EFFORTS

  The next step in reconnaissance drone development was obvious — use the technology of the A-12 to build a very high-speed, high-altitude drone.

  With performance superior to that of the 147 drones, it would be much more likely to survive than the modified Firebees. The drone could also have a longer range than the Model 147. This meant they could be used to cover targets otherwise out of range.

  Following the loss of Powers's U-2 over the Soviet Union, there were several discussions about using the A-12 itself as a drone. Although Kelly Johnson had come to support the idea of drone reconnaissance, he opposed an A-12 drone, contending that the aircraft was too large and complex for such a conversion. Another possibility was to use the A-12 as a launch aircraft for an unmanned QF-104 reconnaissance drone. Several times the possibility was examined, but the CIA expressed no interest.

  Although the CIA turned down the idea, Johnson found an ally with Brig. Gen. Leo Geary, director of air force special projects. General Geary arranged $500,000 from the Black projects contingency fund to begin a drone study in October 1962.

  To speed up development of the drone, which was initially called the "Q-12," Lockheed planned to use a Marquardt ramjet engine from the Bomarc SAM.

  On October 24, in the midst of the Cuban Missile Crisis, Kelly Johnson, Ben Rich, and Rus Daniell met with Marquardt representatives. From these discussions, it was clear that the Bomarc ramjet would have to be modified for use on the Q-12.[298]

  A section of the Skunk Works shop in Burbank was walled off for use by the Q-12 team. The effort was kept isolated from the A-12, YF-12A, and SR-71 development. Just as the A-12 was more secret than the U-2, the Q-
12 was more closely held. It required special passes to enter the walled-off section, which was dubbed the "Berlin Wall West."[299]

  Unlike the Model 147 drones, the Q-12 would not be recovered intact.

  This was done for cost reasons. Trying to put in a recovery system would make the vehicle bigger. It would require one or more large parachutes to lower it to a soft landing. A system that would allow a runway landing would add complexity, take up space, and increase both the cost and weight.

  One major problem was simply slowing the drone down to a speed at which any recovery system could work.[300]

  Instead, Johnson studied the possibility of recovering the Q-12's nose section by parachute. This would include not only the entire camera, but the guidance system. These were more costly than the airframe itself.

  By December 7, 1962, the Q-12 mock-up was completed. On December 10, it was sent to a test site for eleven days of radar cross section (RCS) measurements. These showed it had the lowest radar cross section of any Lockheed design — a record that would be held for the next decade. The Q-12 was then returned to Lockheed for test fits of a mock-up of the camera. The drone would use a Hycon camera which used a rotating mirror to provide panoramic photos. The goal was six-inch resolution from 90,000 feet. Wind-tunnel testing of the design was also beginning.

  At the same time, engine tests of the Marquardt RJ43-MA-3 ramjet were conducted in a wind tunnel that simulated the Q-12's flight profile. One concern about using a ramjet was "blowouts." When the drone turned, it was feared that the airflow into the engine would be disrupted. The effect was like blowing out a candle. Both Lockheed and Marquardt were amazed by the results. The ramjet could be shut down for as long as forty-five seconds and still relight. The only ignition source was the hot engine parts.[301]

  When Johnson presented the results to the CIA in February 1963, he found them unenthusiastic. With the A-12 still far from Mach 3, the U-2 operations against China, and other secret air operations such as Air America, the CIA was overextended and unable to undertake another risky project. In contrast, Air Force Secretary Harold Brown was interested in the Q-12 as a possible nuclear-armed cruise missile, as well as a reconnaissance drone.[302]

  The air force interest seems to have moved the CIA to take action. On March 20, 1963, the CIA issued a contract to begin full-scale development. It assigned responsibility to Lockheed for the airframe, navigation system, and the ramjet. Funding and operational control was split between the CIA and the air force.

  DOUBTS ABOUT THE LAUNCH PROFILE

  Once the Q-12 got under way, it was clear a major problem in development was the aerodynamics of the Q-12/A-12 launch profile. The Mach 3 shock waves between the two aircraft could interact, creating high temperatures and stresses on the airframe. The aerodynamics of the separation also had to be determined. The Q-12 would have to pass through the shock wave formed by the A-12's forward fuselage. The Q-12 would have to separate cleanly, as there were only a few feet between its wingtips and the A-12's inward-canted fins. The center of gravity and center of lift of the combination would have to be very carefully controlled.

  Wind-tunnel tests were made using metal models of the Q-12 fixed to the back of an A-12. For tests of the separation, a more complex arrangement was needed. The Q-12 model was mounted on a movable arm. This would lift it off the back of the model A-12 and provide data on the aerodynamic effects and stresses on the combination. It was not possible to reproduce actual free flight in the wind tunnel, however.[303]

  Problems with the launch profile appeared early and continued into 1964.

  The wind-tunnel data and calculations indicated getting the drone through the shock wave formed by the launch aircraft's fuselage would prove difficult. The ramjet would be at full power, and there could be air-fuel mixture and engine unstart problems as it passed through the shock wave. The launch would also have to be done in a pushover maneuver. It was not until May 1964 that Johnson began to feel any confidence. Although there were still problems with the launch profile, the drag of the combination, and the equipment, Johnson felt they could "haul the thing" through Mach 1.

  Ironically, the actual design and production of the drones had gone much more smoothly. The slight changes in the drone design from the RCS and wind-tunnel testing had changed the size of the payload bay. This forced Hycon to redesign the camera. By August 6, 1963, this had been completed, with no loss of resolution caused by the change. By October 1963, the drone's design had been finalized.

  At the same time, the Q-12 underwent a name change. To separate it from the other A-12-based projects, it was renamed the "D-21." (The "12" was reversed to "21.") "Tagboard" was the project's code name. The A-12 launch aircraft were similarly renamed, becoming the "M-21." The M stood for "mother," while the D was for "daughter."[304]

  THE D-21 TAGBOARD

  The D-21A resembled an A-12 nacelle, wings, and vertical tail. The D-21A was 42.8 feet long, with a wingspan of 19.02 feet. The airframe was built of titanium. The leading edges of the wings, the control surfaces, and the inlet spike were all made from radar-absorbing plastic. The D-21's fueled weight was about 11,000 pounds. The reconnaissance equipment was carried in a 76-inch-long Q-bay just behind the intake. The Hycon HR-335 camera was mounted on the recoverable "hatch." This also carried the inertial navigation system, the automatic flight controls, and the command and telemetry electronics. These were the high-cost elements of the drone. The D-21 itself would self-destruct.[305]

  A Marquardt RJ43-MA-11 ramjet, a modified version of the engine originally designed for the Bomarc SAM, powered the D-21. This could propel it at a speed of Mach 3.35. The normal cruising speed was Mach 3.25 at an altitude of 80,000 to 95,000 feet. This was far above that of the 147 drones.

  The ramjet burned JP-7 fuel, the same as the A-12. The D-21's range was over 3,000 nautical miles.[306] The ramjet's burn time of two hours represented a major technical accomplishment — the Bomarc's ramjets had operated for only ten minutes.[307]

  The D-21 was carried on a pylon, which held it at a nose-up attitude.

  The pylon had latches to hold the D-21, a compressed air emergency jettisoning system, and a fuel line to transfer fuel to the D-21 from the M-21's own tank's. Two additional launch aircraft, TVL-21 Articles 134 and 135, were built. Article 134 would be used for captive test flights, while 135 would make the actual launches. As with the other A-12 derivatives, the M-21s were two-seat aircraft. The launch control officer (LCO) sat in the Q-bay.

  When the D-21/M-21 reached the launch point, the sequence would begin. The first step would be to blow off the D-21's inlet and exhaust covers.

  These would fragment and leave the ramjet ready to start. With the D-21/M-21 at the correct speed and altitude, the LCO would start the ramjet and the other systems of the D-21. This was controlled via a panel — green lights would indicate the status of each system. The LCO could watch the D-21 through a periscope on the instrument panel. The M-21 launch aircraft had a camera mounted in what was called a "hot pod" to film the separation.

  With the D-21's systems activated and running, and the launch aircraft at the correct point, the M-21 would begin a slight pushover, the LCO would push a final button, and the D-21 would come off the pylon.[308]

  The first D-21 was completed in the spring of 1964. As with the U-2s and A-12s before them, the D-21s were given Article numbers. The first was Article 501, with seven D-21s planned for completion by the end of the year. The early D-21s were in a natural metal finish with the outer half of the wings in black. This marked the radar-absorbing plastic material.

  After four more months of checkouts and static tests, the aircraft was shipped to Groom Lake and reassembled. Lockheed test pilot William Park was selected to make all the captive and launch tests. Everything was judged ready for the first captive flight. A bridge crane on the ceiling of the hangar was waiting to lift the D-21, swing it over, and lower it onto the M-21.

  TAGBOARD FLIGHT TESTS

  The first D-21/M
-21 captive flight was scheduled for December 22, 1964. This was also the date for the SR-71's first flight at Palmdale. The dual first flights would be a big plus for Lockheed. The SR-71 test flight was planned for 8:00 A.M., with the D-21/M-21 takeoff to follow at 12:00 P.M. This would allow time for Kelly Johnson to supervise the SR-71 flight, then fly out to Groom Lake. Minor problems pushed back the SR-71's flight until about noon, however. At Groom Lake, William Park and the rest of the D-21/M-21 team awaited their chance. Finally, at about 2:00 P.M., a Lockheed Jetstar landed, Johnson ran up the ladder, patted Park on the head, and told him to take off.

  As the D-21/M-21 flew above the snow-dusted hills, Park checked out its handling. Despite the added drag and weight of the D-21, and low-power engines in the M-21, Park found the combination's flight characteristics were the same as the A-12. For this first flight, the M-21 had only a partial fuel load. As with the SR-71 earlier in the day, the M-21/D-21 went supersonic. Due to the late takeoff, it was getting toward dark when the D-21/M-21 and its F-104 chase planes landed.[309]

  Because the combination represented a whole new design, the speed and altitude envelope had to be explored in small steps. Johnson had hoped to make the first D-21 free flight on his birthday, February 27, but it was not to be. A problem with control surface flutter appeared with the D-21. During one captive flight in April 1965, both the D-21's elevens were ripped off by flutter. This required a redesign, which involved adding balance weights and control surface locks. By May, the flutter problem had been corrected, and the combination had reached Mach 2.6.

  It was at this time that the LCOs for the test launches were selected. Ray Torick and Keith Beswick would alternate as launch officer on the missions.

  The launches were to be made over the ocean, west of Point Mugu, on the California coast.