Ike’s gambit: The development and operations of the KH-7 and KH-8 spy satellites
The bang-bang OCV
GAMBIT was started at the same time as the Air Force Samos E-6 search satellite that was intended to replace the CIA’s CORONA. In summer 1960 Air Force Undersecretary Joseph Charyk forbid Lockheed from competing to build the E-6 spacecraft in order to spread the work around and expand the industrial base for manufacturing reconnaissance satellites. At the time Lockheed already had a virtual monopoly on the manufacture of Air Force satellites. Air Force officials felt that the company was overbooked and were also unhappy with its performance on the CORONA program, which had suffered a string of failures. From Charyk’s viewpoint it made sense to broaden the satellite industrial base by giving contracts to other aerospace companies.
It is highly likely that Lockheed was also forbidden from competing to build the GAMBIT spacecraft. The same company that won the Samos E-6 spacecraft contract, General Electric, also won the contract to build the GAMBIT spacecraft. General Electric manufactured the Orbital Control Vehicle, or OCV, for the GAMBIT program. CORONA used Lockheed’s Agena upper stage to provide power and stability in orbit. GAMBIT would still require an Agena to reach orbit, but it would discard it and rely upon the OCV for highly precise pointing and overall stability.
The OCV was a squat cylinder 60 inches in diameter, the same diameter as the Lockheed Agena upper stage that boosted it to orbit. It contained horizon sensors for accurately orienting the spacecraft, and a cold gas control system—sometimes called a “bang bang system” because it would fire bursts from its jets in quick pulses—to stabilize and point the spacecraft in its orbit. These systems were important because the camera’s field of view was so small that it might point in the wrong direction and miss its target. This was one of the driving factors behind Director of Central Intelligence John McCone’s support of the KH-6 LANYARD satellite. LANYARD was started a year after the GAMBIT, in December 1961. It utilized the camera system from the Samos E-5, which had suffered several spacecraft failures. The CIA had started LANYARD to serve as an interim system until GAMBIT became operational. But McCone also viewed the LANYARD as “insurance” in case GAMBIT experienced problems.
The primary factor that affected any reconnaissance spacecraft’s pointing capabilities was moving mass inside the vehicle. Any moving mass could cause the spacecraft to move in the opposite direction. The major source of movement in the spacecraft was the camera system, and there were several parts of the GAMBIT camera that moved. The image reflecting mirror pitched back and forth to provide stereo photographs by changing the angle that light entered the aperture. However, the biggest source of movement was the film spools: the supply spool in the rear of the spacecraft and the takeup spool in the nose that collected the exposed film. They would impart a pitching movement on the spacecraft as they started and stopped. GAMBIT’s designers reduced the effects of this by looping some of the unexposed film back and forth before it went to the platen. That way the film could be drawn through the camera without having to turn the heavy spools at the same time the camera was exposing film. But the Orbital Control Vehicle’s cold gas stabilization system had to quickly dampen any movement as a result of camera operation.
At some point early in the GAMBIT’s development program managers made an important decision concerning its reentry vehicle. They decided to use the same reentry vehicle developed by General Electric for the CORONA program. Publicly this was known as the Discoverer Satellite Recovery Vehicle, or SRV, after the Discoverer cover story developed for CORONA. Although the SRV was relatively small, it had the virtue of being proven.
This decision had drawbacks, however. The GAMBIT’s film was over three times as wide as CORONA’s, but had to be stuffed into the same amount of space. It is unknown if the limiting factor for how much film the GAMBIT could carry was weight or volume. Eventually the CORONA SRV would carry two spools of ultra-thin film each 16,000 feet (4,877 meters) long and weighing 160 pounds (73 kilograms) total. The GAMBIT SRV carried a maximum of only 3,000 feet (914 meters), roughly equivalent in weight to 9,000 feet (2,743 meters) of CORONA film, so the limiting factor was probably the volume of the wider film in the SRV rather than its weight.
Another limitation of the CORONA SRV was reentry accuracy. The vehicle did most of its slowing down in the upper atmosphere, following a shallow trajectory. As a result, its reentry “footprint” could be quite large, extending 30 nautical miles (56 kilometers) to either side of its ground track and up to 200 miles (370 kilometers) long. Because its footprint was so large, many more aircraft had to be spread out over a much larger area to retrieve it. In fact, at the time that program managers chose the conservative option of using the CORONA SRV, General Electric was already working on developing a larger and more accurate SRV for the Samos E-6 satellite to reduce this footprint. They were also exploring the possibility of developing a lifting body reentry vehicle that could land inside the United States carrying reconnaissance film.
Although the Discoverer SRV had limitations, the program managers eventually realized that they had made the right decision to use a proven design. CORONA capsules returned from orbit regularly, but other Air Force efforts to develop larger and more precise reentry vehicles failed miserably.
Technology transfer was not all one-way, however. GAMBIT’s designers developed a backup battery system called “Lifeboat” which insured de-orbit of the recovery vehicle in event of spacecraft power failure. Lifeboat was soon incorporated into the CORONA.
By September of 1961 the Secretary of Defense made several organizational changes to clarify the management of satellite reconnaissance projects. The Office of Missile and Satellite Systems was renamed the National Reconnaissance Office, or NRO. The Samos Program Office was renamed the Office of Special Projects, or OSP. The NRO was a secret agency and Joseph Charyk was named its first director. Within the NRO the Office of Special Projects’ secret designation was Program A. Program A was responsible for developing the GAMBIT and other satellites.
By March 1962 GAMBIT was taken over by Colonel William G. King, who was the most experienced Air Force officer in the satellite reconnaissance field. He had taken over the WS-117L reconnaissance satellite office in early 1956 and had served in that post until 1958 when he was transferred to run the trouble-plagued Snark cruise missile program.
In 1962 the NRO created the KH series of designations for certain reconnaissance satellites. KH stood for KEYHOLE, which was the existing code word for the security compartment covering satellite imagery. But the KH designation was only assigned to covert satellite programs, not to the pre-existing Samos camera systems that were already public. GAMBIT was assigned the designation KH-7.
GAMBIT also had another designation that was probably applied in late 1961. It was known as Air Force Program 206. This was an unclassified designation used in official paperwork, such as travel orders for Air Force personnel working on the project.
As GAMBIT progressed it suffered schedule delays and cost overruns, but their nature, severity, and cause remain unknown.
The folding Atlas
By early 1963 the GAMBIT program was approaching its first launch, scheduled for the summer. But in May 1963 an Atlas-Agena D launch vehicle was on the pad at Vandenberg Air Force Base undergoing tests. Atop the Atlas was a non-operational payload simulating a GAMBIT satellite. As space historian Joel Powell recently wrote, this test ended in an embarrassing accident.
During fueling a bubble developed in the ground system pumping liquid oxygen into the Atlas, knocking a valve out of alignment. Ground crews then had to manually drain the liquid oxygen tank. But the Atlas received its structural strength from internal pressure. While the oxygen was being drained the vehicle collapsed, crumpling like an empty soda can and causing the Agena and its payload to fold over and hit the pad nose first. A surge of either fuel or residual liquid oxygen also damaged the launch tower, which had to be repaired before another Atlas could be launched.
The event was similar to an early failure in the CORONA program before the first launch, where a vehicle on the pad suffered a catastrophic failure. That event had been labeled “CORONA Zero,” and it had served as a wake-up call to the program’s managers about the importance of carefully checking all of the systems before the vehicle ever reached the pad. In the case of the crumpling Atlas in May 1963, although the Atlas was a total loss, the Agena and its payload were apparently only mockups, not flight hardware, and the program recovered quickly from the accident.
A conservative start
The first GAMBIT mission was launched on July 12, 1963. Its Atlas-Agena lifted off its launch pad at Vandenberg and headed south. The Atlas performed properly and when it burned out it fell away. The Agena’s Bell rocket engine then fired and pushed the payload into polar orbit, at 102 miles (189 kilometers) altitude. The mission was designated 4001.
An engineer at The Aerospace Corporation had recommended that, during GAMBIT’s first flights, the Orbital Control Vehicle should remain attached to the Agena throughout the flight. This was a confidence-building decision because the Agena was proven whereas the OCV was not. However, it meant that photographs could only be taken of targets directly below the vehicle. After the photographic phase of the mission was completed the reentry vehicle separated and came down over the ocean northwest of Hawaii, where it was caught in mid-air by a C-130 aircraft. Its film was then transported to Eastman Kodak in Rochester, New York, where it was processed and copied and then sent to Washington for analysis.
After the reentry vehicle was jettisoned the engineering phase of the mission began. The OCV was separated from the Agena and put through a series of tests to determine its stability and other characteristics. Its performance during these tests is unknown, but it did not totally silence GAMBIT’s skeptics, particularly in the CIA.
On September 6, 1963 the Air Force launched the second GAMBIT spacecraft on mission 4002. Like its predecessor, this GAMBIT also kept the Agena attached throughout the photographic phase of the mission and then detached for engineering tests after the reentry capsule had returned to Earth. The mission also was successful.
Even after the second GAMBIT launch, Albert Wheelon, the CIA’s Director of Science and Technology, expressed skepticism about GAMBIT. “The major question mark in our minds at this point is that the uncertainties involved in establishing the location of a satellite in orbit, combined with the small swath width delivered by the G system, may make it extremely difficult for us to have adequate assurance of covering the targets for which high-resolution photography is required. It is possible, therefore, that neither G nor L[ANYARD] will meet our technical intelligence requirement, and that we may have to develop a system with greater swath width and less resolution than G but smaller swath width and greater resolution than L and [CORONA]. We may also find that we cannot achieve a useable system yielding GAMBIT’s ground resolution from satellite vehicles.”
The third GAMBIT mission, number 4003, was launched on October 25, 1963 and was also successful. The Agena again remained attached. The film was ejected after the photographic phase and the capsule recovered. The OCV was then put through various tests once the intelligence goals had been achieved.
GAMBIT operations and problems
GAMBIT mission 4004 was launched on December 18, and for the first time the OCV and its payload detached from the Agena to conduct the photographic phase of the mission. It was successful, and the capsule was recovered the next day.
Four successful GAMBIT missions in a row proved that the Air Force could run a black satellite program that would work, and the GAMBIT was a powerful new intelligence tool. But CIA officials still complained. As John McMahon, an official in the CIA’s Directorate of Science and Technology, noted in a May 1964 memo, “In 1963 there were four GAMBIT launches. Total target coverage numbered only 15.” The CIA wanted to see GAMBIT’s intelligence return increase substantially, and fast.
By 1964 the KH-7 GAMBIT quickly shifted into high gear. After only four missions in the second half of 1963, the Air Force launched ten GAMBITs in 1964, but not without incident. In May 1964 mission 4008 suffered problems when its Agena lost roll control during the boost phase. The OCV also suffered system problems, but the mission was still able to return some imagery.
But late in the year the GAMBIT program suffered from major problems. In October 1964 a GAMBIT mission failed to achieve orbit when its Agena malfunctioned during launch. Two weeks later mission 4013 achieved orbit, but for unknown reasons returned no film. Ellis Lapin did not remember the specific cause of the problem, but thought that it might have been with the command system. “In those days we used wire-recorders for storing commands, and we did have problems with them. Tape recorders were not yet in vogue,” he explained.
The next mission, 4014, was launched in early December, but suffered a battery failure. “That failure was an explosion!” Lapin exclaimed. Something in the batteries had failed catastrophically and it required much effort to find the cause and fix it.
In January 1965 Brockway McMillan, the Director of National Reconnaissance, decided to delay plans to improve the launch readiness of GAMBIT until after its reliability problems had been solved. Mission 4019, launched in June 1965, and mission 4020, launched in July 1965, both apparently failed to return imagery. Similarly, mission 4023 returned few images and 4034, launched in November 1966, returned no imagery. Many years later retired General Electric executive Hill Paige remembered that after the program had experienced a string of successful missions it suddenly suffered a number of malfunctions—possibly the slew of failures in 1964 and 1965. Paige explained that this had ultimately been traced to a change in the launch pad tower at Vandenberg Air Force Base. An additional structure had been added to the top of the tower and this had reflected acoustic energy back on the vehicle during launch, shaking loose components in the spacecraft.
Despite the problems, the KH-7 GAMBIT was dramatically improving the quality of American intelligence collection. But intelligence officials wanted better results, and they would get them with a major upgrade of the spacecraft and camera system.