The Central Intelligence Agency was created in 1947 principally to provide US leaders with strategic warning of attack by the Soviet Union. The Agency’s main mission during its first decade and a half was to deploy its collection and analytic assets to detect and preempt a nuclear Pearl Harbor. No other intelligence question had greater implications for the national interests of the United States—and its very survival—than determining what kinds of strategic weapons, and how many of them, the Soviet Union had, and how it intended to use them. With the USSR proving to be an extremely hard target for traditional espionage operations, the United States had to turn to technical collection to peer beyond the Iron Curtain.
- “Untouchable” by artist Dru Blair.
In 1954, CIA retained the Lockheed Corporation to build the U-2 reconnaissance aircraft. Essentially a jet-powered glider, the U-2 could fly at the unprecedented height of 70,000 feet—beyond the range of Soviet fighters and missiles—and take detailed photographs of Soviet Bloc military facilities. The aircraft was ready for operations in June 1956. At the time, CIA project officers had estimated that the U-2 would be able to fly safely over the Soviet Union for two years at most before it became vulnerable to Soviet air defenses. The Soviets tracked the U‑2 from its first mission, however. The estimate had proven too optimistic, especially after initial efforts to mask the U-2’s radar image proved ineffective. A more radical solution was needed—an entirely
[Top of page]
Genesis of the A-12 Design—GUSTO
After a year of discussion with aviation companies beginning in the late summer of 1956, CIA focused its attention on building a jet that could fly at extremely high speeds and altitudes while incorporating state-of-the-art techniques in radar absorption or deflection. This effort was codenamed GUSTO. In the fall of 1957, U-2 project manager Richard Bissell established an advisory committee to help select a design for the U-2’s successor. Chaired by Polaroid chief executive Edwin Land, the committee met seven times between November 1957 and August 1959. Designers from several aircraft manufacturers and senior officials from the Navy and Air Force attended some of the meetings. The two most prominent firms involved in the process were Lockheed, which already was investigating designs for the U-2’s replacement, and Convair, which was building a supersonic bomber for the Air Force, the B-58 Hustler.
Clarence “Kelly” Johnson, said that “It makes no sense to just take this one or two steps ahead, because we’d be buying only a couple of years before the Russians would be able to nail us again….I want us to come up with an airplane that can rule the skies for a decade or more.” On 21 April 1958, Lockheed’s Advanced Development Projects component, jokingly nicknamed the “Skunk Works” after the backwoods moonshine still in the comic strip Li’l Abner and already responsible for so many cutting-edge aviation achievements, began designing an aircraft that would cruise at Mach 3.0 at altitudes above 90,000 feet. “The higher and faster we fly, the harder it will be to spot us, much less stop us,” Johnson asserted. On 23 July 1958, Johnson presented his concept to Land’s committee, which expressed interest in the approach. By September, the Skunk Works had studied various configurations called “Archangel-1,” “Archangel-2,” and so forth—a carryover from the original moniker of “Angel” given to the U-2 during its development. The nomenclature soon became simply “A-1,” “A-2,” etc.
That same month, the Land committee met to review all concepts proposed so far. It rejected two Lockheed designs—for a tailless subsonic aircraft and a new supersonic model, the A-2—and approved continuation of Convair’s work on a Mach 4 “parasite” aircraft called the FISH that would be launched from beneath a modified Hustler called the B-58B. In November, Lockheed submitted its design for the A-3, which was much smaller and lighter than the previous concepts. Upon receiving recommendations from the committee, President Dwight Eisenhower in December approved funding for further R&D on the U-2’s successor. Bissell asked Convair and Lockheed to turn in detailed proposals. Both firms did by the summer of 1959.
Lockheed worked on seven more designs in the ensuing months. None of them met all the requirements for speed, altitude, range, and radar cross section (RCS). Johnson decided not to sacrifice aerodynamic capability to achieve a smaller RCS.
[Top of page]
In March 1959, Lockheed developed a design for the A‑11. It would have a top speed of Mach 3.2, a range of 3,200 miles, and an altitude of 90,000 feet, and could be ready by January 1961. The A-11’s main drawback was that it would be more detectable than Convair’s much smaller FISH. That manned, ramjet-powered vehicle was designed to fly at Mach 4.2 at 90,000 feet with a range of 3,900 miles, and also could be ready by January 1961. Three substantial uncertainties beset the FISH, however: the unproven technology of ramjet engines; the unavailability of the B-58B that would fly fast enough to launch it; and the possibility that the B-58B could not reach the necessary speed, or that if it did, the FISH could not operate under post-launch conditions. The Air Force’s cancellation of the B-58B project in June 1959 took the FISH out of the running, but the Land committee also rejected the A-11 because its RCS was still too large. The competition continued.
[Top of page]
Convair and Lockheed completed new proposals in August 1959. Convair’s entry, known as the KINGFISH, was a ground-launched, single-pilot jet with two Pratt & Whitney J58 engines—the most powerful available—and a small RCS. Lockheed’s design, the A-12, also would use the J58 engines. It would reach Mach 3.2 at up to 97,600 feet and have a range of around 4,600 miles. To save weight, Johnson decided not to construct the aircraft out of steel. Because standard lightweight metals such as aluminum could not withstand the heat generated at Mach 3 speeds, Johnson chose a titanium alloy. The A-12’s design incorporated a continuously curving airframe, a forebody with tightly slanted edges called chines, engine housings (nacelles) located mid-wing, canted rudders, and nonmetallic parts to decrease the RCS. A cesium fuel additive would reduce the radar detectability of the afterburner plume.
[Top of page]
The two firms submitted their designs to a selection panel with members from the Department of Defense, the Air Force, and CIA on 20 August 1959. The A-12’s specifications were slightly better than the KINGFISH’s, and its projected cost was significantly less. Convair’s design had the smaller RCS, however, and CIA’s representatives initially favored it for that reason. The companies’ respective track records proved decisive. Convair’s work on the B-58 had been plagued with delays and cost overruns, whereas Lockheed had produced the U-2 on time and under budget. In addition, it had experience running a “black” project. On 28 August, Johnson wrote in his project log,
Saw Mr. Bissell alone. He told me that we had the project and that Convair is out of the picture. They [CIA] accept our conditions (1) of the basic arrangement of the A-12 and (2) that our method of doing business will be identical to that of the U-2. He agreed very firmly to this latter condition and said that unless it was done this way he wanted nothing to do with the project either.
Much of the eventual success of the OXCART program can be attributed to CIA and Lockheed following the best practices from the U-2 project that Johnson and Bissell tacitly referred to: complete trust between customer and contractor, individual responsibility and accountability, start-to-finish ownership of design, willingness to take risks, tolerance for failure, and streamlined bureaucracy with minimal staffing and paperwork.
On 29 August 1959, the selection panel voted for the A‑12 but required Lockheed to demonstrate by 1 January 1960 that it could reduce the aircraft’s RCS sufficiently. CIA awarded a four-month contract to Lockheed to proceed with antiradar studies, aerodynamic structural tests, and engineering designs. Project GUSTO was terminated, and “by a sort of inspired perversity,” an Agency officer later wrote, OXCART was selected from a random list of codenames to designate this R&D and all later work on the A-12.  The aircraft itself came to be called that as well. Funding for the four-month period was $4.5 million.
During tests over the trial period conducted along with contractor EG&G, Lockheed showed that its concept of shape, nonmetallic parts, and fuel additive would produce the needed reduction in RCS. In the course of this phase of radar testing and after, which required a full-scale, pylon-mounted mock-up, and further wind tunnel tests, the A-12 took on more of its distinctive cobra-like shape that allowed for better dispersion of radar pulses. To further reduce those reflections, the two canted rudders were fabricated from laminated nonmetallic materials—the first time such substances had been used for an important part of an aircraft’s structure. (Later on, the production aircraft would be painted with a radar-absorbent coating of ferrite particles in a plastic binder.)
- To test its radar cross section, a full-size model of the A-12 was placed in various positions on a pylon as radar readings were taken.
To Bissell’s great distress, however, the changes also reduced the aircraft’s performance below what he had told the president it could achieve. Johnson had to reduce the A-12’s weight by 1,000 pounds and increase its fuel load by 2,000 pounds so it could reach the target altitude of 91,000 feet. He noted in his project log: “We have no performance margins left; so this project, instead of being 10 times as hard as anything we have done, is 12 times as hard. This matches the design number, and is obviously right." 
These modifications worked. On 26 January 1960, Bissell notified Johnson that CIA was authorizing the design, construction, and testing of the new aircraft. Four days later the official word came, and the contract for 12 A-12s was signed on 11 February. Lockheed’s original price quotation was $96.6 million, but technical difficulties soon made that figure impossible to meet. CIA included a clause providing for periodic reevaluation of costs. That provision had to be invoked a number of times over the next five years as the A-12’s price rose rapidly. With U-2 overflights of the Soviet Union no longer possible after Francis Gary Powers’s aircraft was shot down on 1 May 1960, the reliability of the new CORONA satellite still undetermined, and no other aerial or space vehicle considered feasible for the mission, US leaders were willing to pay handsomely to collect vital intelligence on America’s principal Cold War adversary.
Picking the OXCART Drivers
The pilots who would fly the A-12 also had to satisfy a rigorous set of “design specifications.” The Air Force, with advice from Kelly Johnson and CIA, drew up the selection criteria. Pilots had to be currently qualified and proficient, with at least 2,000 total flight hours, 1,000 of them in the latest high-performance fighter jets; married, emotionally stable, and well motivated; between 25 and 40 years old; and under six feet tall and 175 pounds so they could fit in the A-12’s cramped cockpit. Extensive physical and psychological screening of Air Force personnel files produced 16 candidates. CIA put this group through an intensive security and medical review. The process was kept so secret that the candidates’ superiors did not know what their subordinates were doing. Those who survived the screening were approached to work for the Agency on a highly classified project involving a very advanced aircraft. By November 1961, only five had agreed.
A second search and screening produced six more. The 11 pilots selected to fly missions in the A-12 were Kenneth B. Collins, Ronald J. “Jack” Layton, Francis J. “Frank” Murray, Walter L. Ray, Russell J. Scott, William L. Skliar, Dennis B. Sullivan, Mele Vojvodich Jr., Alonzo J. “Lon” Walter, Jack W. Weeks, and David P. Young. These pilots—known as “drivers,” like their U-2 counterparts—were then “sheepdipped” from Air Force to CIA employment, with compensation and insurance arrangements similar to those provided for the U-2 pilots. Scott, Walter, and Young left the program before the A-12 became operational. Skliar was attached to the Air Force portion of OXCART that developed a supersonic fighter-interceptor, the YF-12A, and did not fly the A-12 operationally. In addition, Ray W. Schrecengost flew the two-seat trainer version of the A-12.
Lockheed test pilots were: Jim Eastham, Bob Gilliland, Darrell Greenamyer, Bill Park, Art Peterson, Lou Schalk, and Bill Weaver. After the A‑12 was decommissioned in 1968, the six surviving operational pilots—Collins, Layton, Murray, Skliar, Sullivan, and Vojvodich—returned to the Air Force. (Ray and Weeks died in A-12 crashes in 1967 and 1968.) Collins and Skliar flew that service’s version of the A-12, the SR-71 Blackbird, as test or instructor pilots, and Layton flew the YF‑12A.
- CIA’s A-12 “drivers” and managers: (l. to r.) Layton, Sullivan, Vojvodich, Barrett, Weeks, Collins, Ray, BGen Ledford, Skliar, Perkins, Holbury, Kelly, and squadron commander Col. Slater.
[Top of page]
Partnership with the Air Force
In addition to providing pilots and assisting with their processing, the Air Force filled several indispensable roles in supporting the OXCART program. The service dispatched to the test site more than a dozen aircraft that were used for training and proficiency flights, cargo transport, search and rescue, administrative travel, and chase flights. (Two of the chase planes and their pilots were lost during the program’s testing phase.) The A-12 consumed huge amounts of fuel—22,000 pounds per hour at cruising speed and altitude—and had to be refueled during its missions. Massive amounts of fuel had to be positioned at special tank farms at several air bases outside the contiguous United States: Eielson, Alaska; Thule, Greenland; Kadena, Okinawa; and Adana, Turkey. The Air Force’s 903rd Air Refueling Squadron at Beale Air Force Base was given KC‑135 tankers for the refueling operations, and the Air Force detailed most of the support personnel and facilities at Kadena Air Base for Operation BLACK SHIELD, the reconnaissance activity the A-12 would undertake in East Asia. Also, the North American Air Defense Command established procedures so that its radar stations would not report detections of high‑performance aircraft. Lastly, Air Force orders for the YF-12A and SR-71 variants of the A-12 helped lower development and procurement costs on the OXCART program.
- Patches made for OXCART crew members. CYGNUS was the name given the A-12 in testing. The 1129th SAS was the unit designation of the A-12 team assigned to Kadena Air Base in Okinawa.
[Top of page]
1. Rich and Janos, 193.
2. Johnson, “Archangel Log,” 7.
3. McIninch, 1.
4. Johnson, “Archangel Log,” 20.