Completion of the first A-12 was delayed several times because the performance specifications it had to meet put Johnson and the Skunk Works in uncharted territory. The aircraft, over 101 feet long and weighing up to 62 tons fully loaded, had to fly at Mach 3.2, or 2,150 miles per hour—as fast as a rifle bullet—at a mid-range altitude of 91,000 feet. The A-12 was expected to be over four times faster than the U-2 and go almost three miles higher. Moreover, the aircraft had to have the smallest feasible RCS to minimize the risk of detection and shootdown. To meet all these demands would require developing new structural materials, tools, and fabrication techniques, and special fuels, fluids, lubricants, sealants, paints, plastics, electronics, cables, windshields, fittings, fixtures, and tires. Bissell recalled that the A-12 “practically spawned its own industrial base,” and the 2,400 or so machinists, mechanics, and fabricators working on the project could do their own milling and forging. No assembly line techniques could be employed; every aircraft was essentially handmade. At the peak of production in the mid-1960s, nearly 8,000 workers were delivering an A-12 or a variant each month.

Getting to that point was long, hard, and expensive. As a consequence of the technical difficulties, the delivery date of the first aircraft began sliding and costs started rising. Originally promised for May 1961, delivery was moved to August and the first flight was moved to December. A vexed Bissell—then preoccupied with the fallout from the Bay of Pigs debacle that he had overseen as head of CIA’s covert operations—wrote to Johnson that “[t]his news is extremely shocking on top of our previous slippage…. I trust this is the last of such disappointments short of a severe earthquake in Burbank."  But it was not. In July, Johnson wrote in his log that Lockheed was “[h]aving a horrible time building the first airplane ...everyone on edge...and we still have a long, long way to go." To reduce expenditures—already at $136 million by October 1961 and still climbing—project officials decided to reduce production to 10 aircraft, at a total cost of over $161 million, and assigned a top-level CIA aeronautical engineer to work at Lockheed to monitor the program.

 

Finding the Right Metal

The most formidable set of challenges Lockheed faced was dealing with the great heat produced by air friction at the high speed the A-12 would reach. Most of the aircraft’s skin would be subjected to temperatures between 500 and 600 degrees F., and over 1,000 degrees F. at some spots near the engines. No metal used in aircraft production up to then could withstand such heat. The metals that could stand up under the conditions were too heavy for this project.

After evaluating many materials, Lockheed chose to build over 90 percent of the A-12’s airframe out of a titanium alloy. It was almost as strong as stainless steel but weighed half as much and could handle the intense heat. (The rest was made of radar-absorbing composite materials.) Lockheed’s supplier of the alloy had trouble delivering material of the requisite quality at first—95 percent was rejected—and the base metal was scarce enough that some had to be obtained covertly from the country with the largest known reserves—the Soviet Union.

Titanium proved to be very difficult to work with, however. Its extreme hardness caused problems in machining and shaping the material. Drills broke and tools snapped, and new ones had to be devised. By the end of the program, drill bits could make 100 holes before resharpening. Titanium also was very sensitive to contaminants such as chlorine and cadmium. Pentel pens could not be used to draw on sheets of the metal because their chlorine-based ink left etch marks. Wing panels that were spot welded in the summer failed within six or seven weeks, but those made in the winter lasted indefinitely. The problem was traced to Burbank’s water, which was heavily chlorinated in the summer to prevent algae growth but not in the winter. Switching to distilled water to wash the panels after acid treatment prevented a recurrence. When bolt heads dropped off under high heat, Skunk Works troubleshooters found that cadmium-plated wrenches left enough residue to weaken the fittings. Hundreds of tool boxes had to be inspected to get rid of the now-useless implements.

Fuels, Lubricants, and Sealants

To operate at design speeds and temperatures, the A-12 required fuel, lubricants, hydraulic fluids, and sealants that had not been invented yet. The fuel tanks, holding almost 11,000 gallons, made up the largest proportion of the aircraft and would heat up to about 350 degrees F. At that temperature the most advanced fuel blends then in use would boil off or blow up. Instead, a special fuel, called JP-7, with a low vapor pressure and high flash point had to be developed; a lighted match would not ignite it. A liquid chemical that exploded on contact with air would start the engines. Through the use of heat exchangers and smart valves, the fuel would also act as an internal coolant.

Synthetic lubricants were formulated to work at the extreme temperature range between Mach 3.1 missions and subsonic refueling. They were practically solid at room temperature and had to be heated before each flight. A hydraulic fluid that would not vaporize at high speed but was still usable at low altitudes was eventually found.
No sealant for the fuel tanks was ever developed that was simultaneously impervious to chemical effects caused by the fuel, and elastic enough to expand and contract as the tanks heated and cooled and were subjected to large pressure changes. Consequently the A-12’s tanks leaked, a quirk that was not detected until the first aircraft was delivered to the test site and filled with fuel, setting off a reaction that broke down the sealants. A “leak rate” of between five and 60 drops per minute, depending on the source, was considered acceptable. When the A-12 was about to go off on a test or mission, it would receive only enough fuel to get airborne. It would then rendezvous with a KC-135, top off its tanks, and immediately climb to operating altitude, which caused the metal to expand and the leaks to stop.

 The Engines

The J58 turbojet engines that would enable the A‑12 to fly so high and fast were the most persistent problem. Designed in 1956 for a Navy aviation project that was canceled, the engines had to undergo major modifications to turn them into the most powerful air-breathing propulsion devices ever made. Just one J58 had to produce as much power as all four of the Queen Mary’s huge turbines—160,000 horsepower or over 32,000 pounds of thrust. To crank it up, two Buick (later, Chevrolet) racecar engines on a special cart were used. The unmuffled, big block engines put out over 600 horsepower and made a deafening roar. The J58s themselves put out an almost incredible din. Recalling his visit to the test site to watch a midnight takeoff, DCI Richard Helms wrote that “[t]he blast of flame that sent the black, insect-shaped projectile hurtling across the tarmac made me duck instinctively. It was if the Devil himself were blasting his way straight from Hell." 

As with so much else on the A-12, getting the engines to work at design specifications posed never-before-encountered troubles with fabrication, materials technology, and testing. Not the least of them was the superhot conditions. Maximum fuel temperatures reached 700 degrees F.; engine inlet temperatures climbed to over 800; lubricants ranged from 700 to 1,000; and turbine inlets reach 2,000 degrees F. and above. A Pratt & Whitney engineer later wrote that “I do not know of a single part, down to the last cotter key, that could be made from the same materials as used on previous engines." 

Pratt & Whitney’s continuing difficulties with the weight, performance, and delivery of the J58 forced delays in the completion of the first A-12. After meeting with the manufacturer in early January 1962, Johnson noted in his log that [t]heir troubles are desperate. It is almost unbelievable that they could have gotten this far with the engine without uncovering basic problems which have been normal in every jet engine I have ever worked with... Prospect of an early flight engine is dismal, and I feel our program is greatly jeopardized. 
To prevent further scheduling setbacks, Johnson and CIA officials already had decided to use the less powerful J75 in early flights. The airframe had to be slightly altered to accommodate the substitute engine, which could power the craft only up to Mach 1.6 and 50,000 feet. Despite enormous development costs of the J58, the engines were not ready until January 1963, and the A-12 did not reach Mach 3 speed until the following July—more than a year after the first test flight.

The design feature that ultimately made it possible for the J58s to generate the power needed to fly at planned speed was a pair of retractable, spike-shaped cones that protruded from the engine inlets. The “spikes,” as they were known, served as regulators that would decelerate, compress, and superheat incoming air, which was further squeezed and heated by the bypass engines before fuel was added. The supercharged air was then expanded through the turbine and fed into the afterburners. This gas-air mix combusted at 3,400 degrees F., just 200 degrees below the maximum temperature for burning hydrocarbon fuels. Without the spikes, the J-58s would have produced only about 20 percent of the power the A-12 needed. Ben Rich, Lockheed’s lead propulsion engineer, recalled that “developing this air-inlet control system was the most exhausting, difficult, and nerve-racking work of my professional life." Rich and his colleagues did much of the testing in wind tunnels at a NASA facility in northern California. They had to work mostly at night because the tests drained too much electricity from the local power grid during the day.

The main issue with the inlets was that the system’s mechanical controls did not respond quickly enough to shock-wave-induced variations in the incoming air flow to prevent engine “unstarts” that would cause violent buffeting and severe yawing, and sometimes smash the pilot’s head against the cockpit. The unstarts and the “popped shocks” occurred at speeds between Mach 2.5 and 2.9 while the aircraft was on an accelerated climb to design speed. After more than a year and a change in subcontractors, a new electronic control was developed that, along with some other modifications, took care of the problem at lower speeds, but ultimately the inlet system had to be redesigned. In the new configuration, the spike could be moved in or out as much as 26 inches at supersonic speeds to capture and contain the shock wave.

 
The Cockpit and Flight Suit

Providing for the pilots’ safety and comfort was difficult because the external temperatures would make the uninsulated cockpit feel like the inside of a moderately hot oven. To cut weight, Lockheed did not even try to insulate the aircraft’s interior; instead, it counted on the pilot’s suit to protect him. Pilots would have to wear a type of space suit with its own cooling, pressure control, oxygen supply, and other life support capabilities. Two Lockheed subcontractors, the David Clark Company and the Firewel Corporation, developed a full-pressure suit and oxygen supply system based on ones created for pilots of the X-15 rocket aircraft. The aluminized suit and breathing apparatus would protect the pilot from heat radiated from the 400 degree F. windshield and the effects of depressurization and extreme cold encountered during a high-altitude bail-out. The S‑901 suits were custom-made and each cost $30,000 in the mid-1960s.

To further protect the pilots, the cockpit had an air conditioning system. It was tested by putting a pilot inside what one engineer described as “a broiler big enough to roast an ox medium rare"  —a cylinder was cooled to 75 degrees F. while the outer skin was heated to about 600 degrees F. Additionally, if the pilot had to eject from the cockpit, his feet would be held against the seat with cables while it cleared the aircraft, and a stabilization parachute would keep him from spinning and rotating as he descending more than 12 miles in around seven minutes to approximately 15,000 feet, when the main parachute would deploy and separate the pilot from the seat.

The Photo Gear

Notwithstanding its innovations in aeronautical engineering, the A-12 was a photographic reconnaissance platform, so the whole OXCART program would have been pointless if worthwhile pictures could not be taken. Project managers decided to have three different camera systems developed to provide a range of photography, from high-ground-resolution stereo to very-high-resolution spotting data.

Perkin-Elmer was the primary manufacturer. Its stereo camera, called Type I, had a 5,000-foot film supply and produced pairs of photographs covering a 71-mile swath with a ground resolution of 12 inches. To meet severe design constraints on size, weight, thermal resistance, coverage, and resolution, Perkin-Elmer employed concepts never before used in camera systems. Perkin-Elmer’s camera was installed on all 29 A-12 missions and failed only once, halfway through a sortie.
In case Perkin-Elmer ran into production problems, Eastman Kodak was also asked to build a camera. Called Type II, it had an 8,400-foot film supply and produced stereo photographs covering a 60‑mile swath with 17-inch resolution. A third firm, Hycon, built an advanced version of the spotting camera used on the U-2. Hycon’s device, Type IV, had a 12,000-foot film supply and covered a 41‑mile‑wide swath with a resolution of eight inches.
The integrity of the double quartz camera window demanded special attention because optical distortion caused by the effect of great heat (550 degrees F.) on the outside of the window and a much lower temperature (150 degrees F.) on the inside could keep the cameras from taking usable photographs. Three years and $2 million later, the Corning Glass Works came up with a solution: the window was fused to its metal frame by a novel process using high frequency sound waves.
In addition to the film cameras, other collection devices were developed or planned for the A-12: an infrared camera, a side-looking radar, a gamma spectrometer, and a particulate sampler. None of these was used on an A-12 mission.

Defenses

Finally, although it was intended to fly too high and too fast to be detected or shot down, the A-12 was equipped with several electronic countermeasures (ECMs) to foil hostile air defenses. The ECMs would warn the pilot his aircraft had been “painted” by a radar or missile guidance signal, and then jam or confuse them.

 
In the Air: “A Wild Stallion”

At last, the first A-12, known as Article 121, was built and ground tested in Burbank during January and February 1962. Because the aircraft was too secret to fly to the test site and too large to carry on a cargo plane, it had to be trucked. During the night of 26 February, a specially designed trailer truck loaded with a huge crate (35 feet wide and 105 feet long) containing the disassembled aircraft’s fuselage left the Skunk Works for the two-day trip to the Nevada facility, escorted by the California and Nevada highway patrols and CIA security officers. The box was so wide that some road signs had to be removed, trees trimmed, and road banks leveled. The wings were shipped separately and attached on site.

The A-12’s first flight—unofficial and unannounced in keeping with a Lockheed tradition—took place on 25 April 1962 and almost caused the loss of the only OXCART aircraft built so far. Lockheed test pilot Lou Schalk flew the plane less than two miles, at an altitude of about 20 feet, because serious wobbling—Johnson described the movements as “lateral oscillations which were horrible to see”—caused by improper hookup of some navigational controls. Instead of circling around and landing, Schalk put it down in the lake bed beyond the end of the runway. When the A-12’s nose appeared out of a cloud of dust and dirt, Johnson’s angry voice erupted over the radio, “What in Hell, Lou?" 
The next day, Schalk tried again, this time with the landing gear down, just in case. The flight lasted about 40 minutes. The takeoff was perfect, but after the A-12 got to about 300 feet it started shedding all the “pie slice” fillets of titanium on the left side of the aircraft and one fillet on the right. (On later aircraft, those pieces were paired with triangular inserts made of radar-absorbing composite material.) Technicians spent four days finding and reattaching the pieces. Nonetheless, the flight pleased Johnson. “We showed that the first flight troubles were not caused by basic aircraft [in]stability." 

Hauled disassembled and in boxes to its Nevada test site, the A-12 posed a significant traffic hazard. Once, an oncoming bus grazed a crate.
Once the fillets were repaired, Article 121 was rolled out for its first official flight on 30 April, just under one year later than originally planned. A number of senior Air Force officers and CIA executives, including Deputy Director for Research Herbert Scoville and former project chief Bissell (who left the Agency in February 1962), witnessed the long-awaited event. Schalk again was the pilot. He took the aircraft up for 59 minutes and reached 30,000 feet and just under 400 mph; most of the flight was made at under 300 mph. He reported that the A-12 responded well and was extremely stable. Johnson said this was the smoothest official first flight of any aircraft he had designed or tested. On 4 May, with Schalk at the controls, Article 121 made its first supersonic flight, reaching Mach 1.1 at 40,000 feet. Problems were minimal. DCI John McCone, who had shown a keen interest in the OXCART program since becoming director in November 1961, sent Johnson a congratulatory telegram.

Now began the arduous and often discouraging task of bringing the aircraft—christened “Cygnus” after the swan constellation in the northern sky—up to operational performance requirements. Ben Rich later called the A-12 “a wild stallion of an airplane. Everything about it was daunting and hard to tame…so advanced and so awesome that it easily intimidated anyone who dared to come close."

By the end of 1962, four more aircraft had arrived at the test site, and two were engaged in flight testing. Article 122, which arrived in late June, initially was used principally for checking electronic and propulsion systems and RCS. Its first test flight  was in January 1963. During its delivery, a Greyhound bus traveling in the opposite direction grazed the 35-foot-wide crate carrying a portion of it. Project managers quickly authorized the payment of nearly $5,000 for damage to the bus so that no insurance or legal inquiry would take place and compromise the program. Articles 123 and 125, after arriving in August and December, respectively, were outfitted for operational use.

Article 124, a trainer version nicknamed the “Titanium Goose,” was delivered in November. It was fitted with the less powerful J75 engines, could only reach Mach 1.6 and 40,000 feet, and was the only A-12 that Kelly Johnson ever flew in. (The CIA’s first deputy director for science and technology, Albert “Bud” Wheelon, also took a ride in the trainer to demonstrate his confidence in the A-12. John McCone, the director of central intelligence, “roundly criticized” him for “risking my person” that way, Wheelon recalled.) 

The remaining 10 aircraft in the fleet would arrive at the test site by mid-1964. Of those, eight were designated for reconnaissance missions, and two would become the “mother ships” for the D-21 drones in the TAGBOARD project.