--- PAGE 1 --- fri{) _i \ ~===~===~~:: : : : ~ /.,:. ,.... . . THE CENTRAL INTELLIGENCE AGENCY AND -:.,:-. OVERHEAD (7 ; ! ..,: • ~ •.... R ECONNA ISSANCE The U-2 and OX C.\RT Programs, 1954 - 1974 t);J; :~•;/-;:~: _:; .. ·::/ ' ' .,. .\~:·· ..:__.. Gregory 'vV. Pedlow and :>~~::.< :;;~..:.,· Donald E. vVelzenbach !:·-~~--.~..; ' ::::..:· t> ·r:~.;.,- -:_;~::-:...~ f ; ":· .:..:.... :: • ·+·,'(/\,-_;,' ,:,. . ••• ·.i: •• · : ·: -~ .;'~: :\·· ~. • :•-: ... . •:· ! /· ·, '. ~ .• : .· . . [ },<~-~.~:_.:~ :· .~- ::· .~_:•;··.:: · ?;.\ ..- ~:.-:··· :}·.··.\ :· -:.: ,•.·-:· · ''{ :·: ·- ·.. .. ... ..• > ~ --- PAGE 2 --- Secret l. ~)\~~,-~, Skunk Works Design Staff speed of Mach 0.8 or 460 knots at altitude. Its initial maximum alti­ tude would be 70,600 feet and the lllltimate maximum altitude would be 73,100 feet. According to these early December 1954 specifica­ tions, the new plane would take off at 90 knots, land at 76 knots, and be able to glide 244 nautical miles from an altitude of 70,000 feet. After discussing the reconnaissance bay with James Baker, Johnson had worked out various equipment combinations that would not ex­ ceed the weight limit of 450 pounds. Johnson ended his report by promising the first test flight by 2 August l 955 and the completion of four aircraft by I December I955.': " Kell y Johnson. ·• A High-Altitude: Rcconnai~san(;C: Aircraft," 9 Dc:cc:mbc:r 195-i. Lockheed Contract Filc:s. OSA Records (S ). SeeFet --- PAGE 60 --- Secr&t PJOFOFU'd Chapter 2 47 In designing the U-2 aircraft, Kelly Johnson was confronted with two major problems-fuel capacity and weight. To achieve interconti­ nental range, the aircraft had to carry a large supply of fuel, yet, it also had to be light enough to attain the ultrahigh altitudes needed to be safe from interception. Although the final product resembled a typ­ ical jet aircraft, its construction was unlike any other US military air­ craft. One unusual design feature was the tail assembly, which-to save weight-was attached to the main body with just three tension bolts. This feature had been adapted from sailplane designs. The wings were also unique. Unlike conventional aircraft, whose main wing spar passes through the fuselage to give the wings continu­ ity and strength, the U-2 had two separate wing panels, which were attached to the fuselage sides with tension bolts (again, just as in sail­ planes). Because the wing spar did not pass through the fuselage, Johnson was able to locate the camera behind the pilot and ahead of the engine, thereby improving the aircraft's center of gravity and re­ ducing its weight. The wings were the most challenging design feature of the entire airp~ane. Their combination of high-aspect ratio and low-_drag ratio (in other words, the wings were long, narrow, and thin) made them unique in jet aircraft design. The wings were actually integral fuel tanks that carried almost all of the U-2's fuel supply. The fragility of the wings and tail section, which were only bolted to the fuselage, forced Kelly Johnson to look for a way to pro­ tect the aircraft from gusts of wind at altitudes below 35,000 feet, which otherwise might cause the aircraft to disintegrate. Johnson again borrowed from sailplane designs to devise a "gust control" mechanism that set the ailerons and horizontal stabilizers into a posi­ tion that kept the aircraft in a slightly nose-up attitude, thereby avoiding sudden stresses caused by wind gusts. Nevertheless, the U-2 remained a very fragile aircraft that required great skill and concen­ tration from its pilots. The final major design feature was the lightweight, bicycle-type landing gear. The entire structure-a single oleostrut with two light­ weight wheels toward the front of the aircraft and two small, solid-mount wheels under the tail-weighed only 208 pounds yet could withstand the force of touchdown for this 7-ton aircraft. Because both sets of wheels were located underneath the fuselage, the U-2 was also equipped with detachable pogos (long, curved sticks with two small wheels on them) on each wing to keep the wings level during takeoff. The pilot would drop the pogos immediately after takeoff so Seeret --- PAGE 61 --- Sec, et NOFORN Chapter 2 48 U-2 at testing site before attachment of wings and tail assembly that they could be recovered and reused. The aircraft landed on its front and back landing gear and then gradually tilted over onto one of 1 the wingtips, which were equ ipped with landing skids. ) THE DEVELOPMENT OF THE CAMERA SYSTEM By December 1954, Kelly Johnson was at work on drawings for the U-2's airframe and Pratt & Whitn•:!Y was already building the J57 jet " For the design ft:atures of the U-'.! in early 1955. see R. F. Boc:hrne. Summary Report: Reconnaissance Aircraft . Lockheed Aircraft Corpor.ition Report I04'.!0. 28 January 1955. pp. 7-9. OSA Records. job 74-B-6-l5. box I (S). --- PAGE 62 --- Sec. et NOFOAN Chapter 2 49 .. ·- . ·. • ,• .. .:;~~~~;.\ii ~1i: • ~:; ..:. ~ .::,, , ',t •... . , " •. h -;.: . .-:.:.~~- ...___ ... : ~·-'~•-· ,:..- .... : .~:~~·~-~~~~:~;\ ~f~ ' • .. :-'·· ~::....:,,.;_:,-:/~~ ~- :·.'.:~-:---·~,... engine, but no firm plans existed for the all-important cameras. U-2 landing gear and pogos Existing cameras were too bulky and lacked sufficient resolution to be used in high-altitude reconnaissance. The workhorses of World W'ilf II aerial photography had been the Fairchild K-19 and K-21 framing cameras with lenses of varying focal lengths from 24 to 40 inches. Late in the war, the trimetrogon K-17 .mapping-camera system came into use. This system consis~ed of lhree separate cameras which made three photographs simultaneously: a vertical, an oblique to the left, and an oblique to the right The major shortcomings of the trimetrogon system were the large amount of film required and the system's lack of sharp definition on the obliques. The standard aerial cameras available in the e'ilfly 1950s c:ould achieve resolutions of about 20 to 25 feet (7 to 8 meters) on a side when used at an altitude of 33,000 feet (10,000 meters), or about 25 lines per millimeter in current tenns of reference. Such resolution was considered adequate because aerial photography was then used pri­ marily to choose targets for strategic bombing, to assess bomb dam­ age after air raids, and to make maps and charts. Unfortunately, a camera with a resolution of only 20 to 25 feet at a height of 33,000 feet was too crude to be used at twice that altitude. lndeed, for intelli­ gence purposes a resolution of less than 10 feet was necessary to dis­ cern smaller targets in greater detail. This meant that any camera carried to altitudes above 68, 000 feet had to be almost four times as good as existing aerial cameras in order to achieve a resolution of less than 10 feet As a result, some scientists doubted that useful photogra­ phy could be obtained from altitudes higher than 40,000 feet.'" " Baker imervil!w (S). .Sea,et- --- PAGE 63 --- Seeret NOFORN Chapter 2 50 The first success in designing very-high-acuity lenses came in + ·~;~ ,·\ ·~~:f~: .i ~ ~~\ ~~ the mid- I940s, when James G. Baker of Harvard and Richard S. ;.;.,-::., . Perkin of the Perkin-Elmer (P-E) Company of Norwalk, Connecticut, collaborated on a design for an ex.perimental camera for the Army Air Force. They developed a 48-inch focal-length scanning camera that was mounted in a modified B-36 bomber. When tested over Fort Worth, Texas, at 34,000 feet, the new camera produced photographs in which two golf balls on a putting green could be distinguished (in reality. however, the "golf balls··· were 3 inches in diameter). These photographs demonstrated the high acuity of Baker's lens, but the camera weighed more than a ton and was much too large to be carried aloft in an aircraft as small as the U-2. Realizing that size and weiglht were the major restraining factors in developing a camera for the U--2. James Baker began working on a radically new system in October 1954, even before the CIA adopted the Lockheed proposal. Baker quickly recognized, however, that he would need almost a year to produce a working model of such a com­ plex camera. Since Kelly Johnson had promised to have a U-2 in the James G. Baker air within eight months, Baker needed to find an existing camera that could be used until the new came:ra was ready. After consulting with his friend and colleague Richard Perkin. Baker decided to adapt for the U-2 an Air Force camera krnown as the K-38. a 24-inch aerial framing camera built by the Hycon Manufacturing Company of Pasadena, California. Perkin suggested modifying several standard K-38 cameras in order to reduce their weight to the U-2's 450-pound payload limit. At the same time, Baker would ma:ke critical adj ustments to existing K-38 lenses to improve their acuiity. Baker was able to do this in a few weeks, so several modified K-38s, now known as A-1 cameras, were ready when the first "Angel" aircraft took to the air in 15 mid-1955. CIA awarded Hycon a contract for the modified K-38 cameras, and Hycon, in tum. subcontracte:d to Perkin-Elmer to provide new lenses and to make other modifications to the cameras in order to make them less bulky. In its tum, Perkin-Elmer subcontracted to Baker to rework the existing K-38 lenses and later design an im­ proved lens system. To keep his liens-designing efforts separate from " Ibid. See,et --- PAGE 64 --- Secret NOFORN Chapter 2 51 A-1 camera his research associate duties at Harvard and his service on go,vem­ ment advisory bodies, Baker established a small firm known as Spica, Incorporated, on 31 January 1955. The A-1 camera system consisted of two 24-inch K-38 framing cameras. One was mounted vertically and photographed a 17.2° swath beneath the aircraft onto a roll of 9.5- inch fi lm. The second K-38 was placed in a rocking mount so that it alternately photographed the left oblique and right oblique out to 36.5° onto separate rolls of 9.5-inch fi lm. The film supplies unwound in opposite directions in ordler to minimize their effect on the balance of the aircraft. Both cameras used standard Air Force 24-inch focal-length lenses adjusted for max­ imum acuity by Baker. The development of the special rocking mount by Perkin-Elmer's Dr. Roderic M. Scott was a major factor in neduc­ ing the size and weight of the A- I system, because the mount pro­ vided broad transverse coverage with a single lens, ending the need for two separate cameras. 16 •• OSA History. chap. I. annt!x 3, pp. 1-3 (TS Co1kword). SeGF&t --- PAGE 65 --- Settet N0FORl)I Chapter 2 52 A-2 camera U-2s equipped with the A- I camera system also carried a Perkin-Elmer tracking camera using 2.75-inch film and a 3-inch lens. This device made continuous horizon-to-horizon photographs of the terrain passing beneath the aircraft. Because the A - I system was: new, it also included a backup camera system. a K-17 6-inch three-camera trimetrogon unit using 9-inch film. W hi le the A -1 system was still being developed, James )Baker was already working on the next generation of lenses for high-alrtitude reconnaissance. B aker was a pioneer in using computers to synthesize optical systems. His software algorithms made it possible to model lens designs and determine in advance the effects that variatio,ns in lens curvatures. glass compounds. and lens spacings would have on rays of light passing through a lens. These .. ray-tracing" programs re­ quired extensive computations, and, for this he turned to the most modem computer available, an IBM CPC (card-programmed callcula­ 17 tor) installation at nearby Boston University. " Ibid.. chap. I. pp. 7-8 (TS Codeword). -Seeret --- PAGE 66 --- Seeret NOFORN Chapter 2 53 Baker's new lenses were used in a camera system known as the A-2, which returned to a trimetrogon arrangement because of prob­ lems with the A-1 system's rocking mount. The A-2 consisted of three separate K-38 framing cameras and 9.5-inch film magazines. One K-38 filmed the right oblique, another the vertical, and a third the left oblique. The A-2 system also included a 3-inch tracking camera. All A-2 cameras were equipped with the new 24-inch f/8.0 Baker-designed lenses. These were the first relatively large photo­ graphic objective lenses to employ several aspheric surfaces. James Baker personally ground these surfaces and made the final bench tests on each lens before releasing it to the Agency. These lenses were able to resolve 60 lines per millimeter, a 240-percent improvement over existing lenses. 18 Once Baker and Scott had redesigned the 24-inch lens for the K-38 devices, they turned their attention to Baker's new camera de­ sign, known as the B model. It was a totally new concept, a high-reso­ lution panoramic-type framing camera with a much longer 36-inch f/10.0 aspheric lens. The B camera was a very complex device that - used a single lens to obtain photography from one horizon to the other, thereby reducing weight by having two fewer lenses and shutter assemblies than the standard trimetrogon configuration. Because its lens was longer than those used in the A cameras. the B camera achieved even higher resolution-100 lines per millimeter. The B camera used an 18- by 18-inch format, which was achieved by focusing the image onto two counterrotating but overlap­ ping 9. 5-inch wide strips of film. Baker designed this camera so that one film supply was located forward, the other aft. Thus, as the film supplies unwound, they counterbalanced each other and did not dis­ turb the aircraft's center of gravity. The B camera had two modes of operation. In mode I, the camera used a single lens to make seven unique exposures from 73.5° on the far right and far left obliques to vertical photos beneath the air­ craft, effectively covering from horizon to horizon. Mode II narrowed the lateral coverage to 21 S on either side of vertical. This increased the available number of exposures and almost doubled the camera's " "Basic Configuration and Camera Data:· 24 January 1956. OSA Records (TS Codeword); OSA History, chap. 5. annex 44 (TS Codeword). Seeret --- PAGE 67 --- &ee1 et NOFOAN Chapter 2 54 B camera . operating time. Three of the seven B-camera frames provided stereo coverage. The complex B cameras were engineered by Hycon's chief designer, William McFadden.'" James Baker's idea for the ultimate high-altitude camera wa:s the C model that would have a 240-inch focal length. In December 1954, he made preliminary designs for folding the optical path using three mirrors, a prism, and an f/20.0 lens system. Before working ouIt the detai ls of this design, however, Baker flew to California in early January 1955 to consult with Kelly Johnson about the weight and space limitations of the U-2·s payload compartment. Despite eve~y ef­ fort to reduce the physical dimensions of the C camera, Baker ne,eded an additional six inches of payload space to accommodate the bigger lens. When he broached this subject to Johnson, the latter replied, " Six more inches? I'd sell my grandmother for six more inches!I" ~ •·• Ibid.: Bakl!r interview (S), '' Bakc:r intl!rvkw (S). --- PAGE 68 --- 6ee1 et NOFOfitM Chapter 2 55 Realizing that the 240-inch lens was both too large and too heavy for the camera bay, Baker scaled the lens down to a 200-inch f/16. 0 system. This was still too big. Further reductions followed, re­ sulting by July 1955 in a 120-inch fl 10.9 lens that met both the weight and space limitations. Later in the year, Baker decided to make the mirrors for the system out of a new, lightweight foamed silica mate­ rial developed by Pittsburgh-Coming Glass Company. This reduced the weight significantly, and he was able to scale up the lens to a 180-inch f/13.85 reflective system for a 13- by 13-inch format. In the past, the calculations for such a complex camera lens would have taken years to complete, but thanks to Baker's ray-tracing computer program, he was able to accomplish the task in just 16 days. When a C camera built by Hycon was flight-tested on 31 January l 957, project engineers discovered that its 180-inch focal length, which was five times longer than that of the B camera, made the camera very sensitive to aircraft vibration and led to great difficulty in aiming the C camera from altitudes above 68,000 feet. The engi­ neers, therefore, decided to shelve the camera. More than five years later, a redesigned C camera was employed during the Cuban Missile Crisis in 0ctober 1962, but the results were not very satisfactory. The failure of the C camera design was not a serious setback to the high-altitude reconnaissance program, because the B camera proved highly successful. Once initial difficulties with the film-trans­ port system were overcome, the B camera became the workhorse of high-altitude photography. An improved version known as the B-2 is still in use. Both of the earlier A-model cameras were phased out after September 1958. During the period when he was designing lenses for the CIA's overhead reconnaissance program, James Baker was also working on classified lens designs for the Air Force and unclassified designs for the Smithsonian Institution. To protect the security of Baker's work for the Agency, Herbert Miller of the Development Projects Staff told Baker to work on lenses for the U-2 in the open and not make any effort to classify the documents connected with the project. Miller be­ lieved that by not calling attention to the effort through the use of spe­ cial security measures, the project could be completed faster and still not be compromised. This "hiding in the open" strategy proved very successful? " Ibid. $ee,et --- PAGE 69 --- Sec, et NOFORN Chapter 2 56 In addition to the camera systems, the U-2 carried one other im­ portant item of optical equipment, a periscope. Designed by James Baker and built by Walter Baird of Baird Associates, the optical peri­ scope helped pilots recognize targets beneath the aircraft and also proved to be a valuable navigational aid. 22 PREPARATIONS FOR TESTING THE U-2 As work progressed in California on the airframe, in Connecticut on the engines, and in Boston on the camera system, the top officials of the Development Projects Staff flew to California and Nevada to search for a site where the aircraft could be tested safely and secretly. On 12 April 1955 Richard Bissell and Col. Osmund Ritland (the se­ nior Air Force officer on the project staff) flew over Nevada with Kelly Johnson in a small Beechcraft plane piloted by Lockheed's chief test pilot, Tony Le Vier. They spotted what appeared to be an air­ strip by a salt flat known as Groom Lake, near the northeast corner of the Atomic Energy Commission's (AEC) Nevada Proving Ground. After debating about landing on the old airstrip, LeVier set the plane down on the lakebed, and all four walked over to examine the strip. The facility had been used during World War II as an aerial gunnery range for Army Air Corps pilots. From the air the strip appeared to be paved, but on closer inspection it turned out to have originally been fashioned from compacted earth that had turned into ankle-deep dust after more than a decade of disuse. If Le Vier had attempted to land on the airstrip, the plane would probably have nosed over when the wheels sank into the loose soil, killing or injuring all of the key fig­ ures in the U-2 project. 23 Bissell and his colleagues all agreed that Groom Lake would make an ideal site for testing the U-2 and training its pilots. Upon re­ turning to Washington, Bissell discovered that Groom Lake was not part of the AEC proving ground. After consulting with Dulles, Bissell and Miller asked the Atomic Energy Commission to add the Groom Lake area to its real estate holdings in Nevada. AEC Chairman Adm. Lewis Strauss readily agreed, and President Eisenhower also ap­ proved the addition of this strip of wasteland, known by its map des­ ignation as Area 51, to the Nevada Test Site. The outlines of Area 51 " Information supplied by James Baker to Donald E. Welzenbach, 12 May 1986 (U). '·' OSA History, chap. 8, pp. 1-2 (TS Codeword); Miller, Lockheed U-2. pp. 19-20. Sea,at --- PAGE 70 --- ~ M t N6f0Ri\J Chapter 2 57 are shown on current unclassified maps as a small rectangular area Area 51, the Ranch adjoini ng the northeast comer of the much larger Nevada Test Site. To make the new facility in the middle of nowhere sound more attractive to his workers. Kelly Johnson called it the Paradise Ranch, which was soon shortened to the Ranch. !• Although the dry lakebed could have served as a landing strip. project managers decided that a paved runway was needed so that testing could also take place during the times when rainwater runoff from nearby mountains filled the lake (at such times the base acquired yet another unofficial name, Watertown Strip). By July 1955 the base was ready, and Agency, Air Force, and Lockheed personnel began moving in. '' OSA History, chap. 8. pp. 2-6 (TS Co,kword): John:;on, "Log for Projectx:·::?5-29 April I955: Clarence L. "Kelly" Johnso n with Maggi. " t,,lemorandum for lhc Joint S1udy Group from James Q. Rc:bcr. "Handling of Requirements for 1he U-2:· 15 August 1960. IC Staff records. job 33-T-123A. box I 0. " CHALICE (Gen.:rall" (TS Cod.:worJ>. --- PAGE 95 --- Gee, et NOFOAN Chapter 2 82 ARC gave the top priority target list to the Project Director, and tlie project staff's operations section then used the list to plarn the tlightpachs for U-2 missions. Although the requirements committee was not responsible for developing fl ight plans. it assisted the plan­ ners with detailed target information as required. When a flight plan was ready for submission to the President for approval, the committee drew up a decailed justi fication for the selection of the targets. This paper accompanied the flight plan.(J [n developing and pnontIz1ng lists of targets, the committee members had to take into account the varying needs and interesIts of their parent organizations. Thus, the CIA representatives generally emphasized strategic intelligence: ai rcraft and munitions factories, power-generating complexes. nuclear establishments. roads. bridlges. inland waterways. In contrast, the military services usually plac,ed a heavier emphasis on order-of-baule data. The Air Force. in panicular. had a strong interest in gathering intelligence on the location of Soviet and East European airfields and radars. Arthur C. Lundahl A lthough the committee members kept the interests of their ser­ vices or agencies in mind. their awareness of the vital nature of their mission kept the level of cooperation high. The group always attempted to reach a consensus before issuing its recommendations, although oc­ casionally this was not possible and one or more agencies would add a 5 dissent to the recommendation o f the committee as a whole." PREPARATIONS TO HANDLE THE PRODUCT OF U-2 MISSIONS On 13 December 1954, DCI Allen Dulles and his assistant. Richard Bissell, briefed Arthur C. Lundahl. the chief of CIA's Photo­ Intelligence Division (PID), on Project AQU ATONE. At !DC[ Dulles's direction, Lundahl immediately set in motion within his divi­ sion a compartmented effort, known as Project EQUINE, to plan for the exploication of overhead photography from the U-2 project. With only 13 members, the PID staff was too small co handle the expe,cted ~ lbi