In the Shadow of Black

The D-21B and Project “Senior Bowl" 

“A” Flight and the 4200th Support Squadron

Beale AFB, California


D-21 Drone

B-52H with two Lockheed D-21B Drones

Of the various Skunk Works Mach 3.0 aircraft programs, the least known to reach the operational hardware stage was undoubtedly the D-21 un-manned strategic reconnaissance drone.  Developed for and operated by the Central Intelligence Agency and the Air Force under a veil of extreme secrecy not penetrated until long after the program had ceased to exist as a viable national reconnaissance asset, it was uncovered only by accident.  During early 1977, aviation enthusiasts unexpectedly found 17 D-21s in storage at Davis Monthan AFB's Military Aircraft Storage and Disposition Center.  Their discovery resulted in the first public disclosure concerning the precedent-setting D-21 and the beginning of an ongoing curiosity about its history and operational service life.

            Code named “Tagboard,” the D-21 was, in fact, an extension of the A-12 program. It was brought to life in response to the US Government's decision to discontinue over flights following the loss of Gary Powers in his U-2 on May 1, 1960. Consequently, in October 1962, Kelly Johnson made the first log entry in what was to become this enigmatic aircraft's long-hidden history: "Over the past several years, we have had a number of discussions on the feasibility of making a drone with the A-12 aircraft. I have steadily maintained that we should not do this, as it is a much too large and complicated machine. On several different occasions, we studied the use of a QF-104 air-launched from the A-12. It became obvious at an early date that the CIA was totally uninterested in this project, [but] others wanted to do it very much."

             Regardless, on October 10, 1962, the Skunk Works from the CIA received authorization for a drone study. Johnson noted the event as follows: "We have now configured it to allow the use of plastic blankets overall for the basic structure. In order to avoid the F-104 problem of a high central vertical tail, I put two on the tips and one in the middle. Besides the aerodynamic benefits of this configuration, they will provide the basis for a landing gear during the flight test operation."

            Lockheed had extensive background with the ramjet-powered X-7A-1, X-7A-2 and X-7A-3 test vehicle series and a close working relationship with Marquardt, a neighboring company in the San Fernando Valley. Together they determined that a highly modified Marquardt RJ43-MA-11 ramjet engine, formerly used on the retired USAF/Boeing Bomarc IM-99B air defense weapon, could power the unmanned drone. 

            Coupled with propulsion technology developed by Marquardt, it was not difficult for the team to execute a functional reconnaissance platform in a relatively short period of time. Importantly, the technology base generated by initial flight tests of the A-12 had given the engineering team (under Johnson) considerable confidence in the aerodynamic and low-observable precedent (i.e., reduced RCS) set by the chine delta. This configuration was a given by the time initial design options were studied for the D-21.

            On October 24, Johnson, Ben Rich and Rus Daniell met with representatives from Marquardt to discuss the use of a ramjet propulsion system on the new Lockheed drone. Johnson wrote: "It is obvious that we cannot use the Bomarc engine without change. We will just do the best we can to use its major parts." The new project was then assigned to Art Bradley under the supervision of Dick Boehme. A small team was assembled to handle engineering and a piece of the Skunk Works shop at Burbank was walled off specifically to accommodate the new drone activity.

            The D-21 engine is properly identified as an XRJ43-MA20S-4. While the Bomarc A used Marquardt RJ43-MA-3 ramjet engines, the Bomarc B used two externally mounted RJ43-MA-11s. Each MA-11 engine had a fixed geometry, Mach 2.35, isentropic spike inlet. The engine’s integral combustion chambers/exit nozzles were designed for relatively low cruise altitudes and therefore did not require high-expansion-ratio nozzles. The ignition system in each engine consisted of two pyrotechnic flares. There was no re-ignition capability. The MA-11's fuel control and flame holder combustion limits would not allow operation at high altitudes. 

            The externally mounted MA-11 engines on the Bomarc B used an all pneumatic fuel control system that maintained a constant Mach number of two selectable speeds as determined by pressure signals from the built-in inlet. These engines had the ability 
to function as an independent external power plant on any vehicle that could reach sufficient speed to allow efficient inlet operation. The MA-11 was developed in supersonic wind tunnels at Marquardt's Van Nuys, California test facility, flight tested on the Lockheed X-7A-3 at Holloman AFB, Alamogordo, New Mexico, and deployed operationally on the Bomarc B. The MA20S-4 engine employed in the D-21 used many MA-11 components but was modified to operate at lower pressures and higher temperatures. The S-4 was immersed in the body of the D-21 and had no inlet structures of its own because it utilized the D-21’s inlet system. The engine’s center body and main structure remained to house the fuel control, fuel pump, fuel injector nozzles and flame-holder assemblies. The flame-holder system was redesigned to allow for stable combustion at extreme high-altitude, high-temperature and low-pressure situations. Ignition was by TEB (Tri Ethyl Boron) to allow for re-ignition in the event of flameout. The combustion chamber/exit nozzle was redesigned to provide for the much greater 
expansion ratio required for high-altitude cruising. The design also incorporated an ejector system for engine structure cooling. Of interest is that until the advent of the D-21, no ramjet had ever powered any craft for more that a few minutes. The D-21’s XRJ-MA20S-4 would be powered up during its entire flight of over 1.5 hours. 

            The pneumatic fuel control computer was also modified to function at much lower pressures and higher temperatures. The input pressure sensors were redesigned to accept air pressures from the D-21 inlet and to provide continual, full power operation limited only by the D-21’s inlet conditions and mission parameters. The fuel flow schedule, pumps, controls and injector nozzles were all updated to permit accurate flow control and injection at the much lower air flow requirements of extremely high altitudes. 

            During early November 1962, while preliminary design work progressed with considerable rapidity (on what was then being referred to in the Skunk Works as the “Q-12”), miscellaneous subsystem problems, including those of the proposed optical sensor gear, surfaced to cause concerns about program progress. Johnson had favored the Hycon Company to win the contract to build the compact camera, but he was not the only decision maker. Per Johnson: "I am very impressed by their recent design with a fast moving slit shutter arrangement I saw at their factory recently."

            On November 5, Johnson wrote: "The drone is developing without much discussion between Headquarters and us. I think I know what they want, but no one has spelled it out. We will try to get six-inch ground resolution photographically, a range of 3,000 nautical miles, and a payload of 425 pounds for the camera. We will attempt to save the expensive elements of the aircraft by parachute recovery of the nose."

            The full-scale mockup was completed on December 7. Because it included the antiradar characteristics of the actual drone, it was sent to a test facility on December 10 for eleven days of pole-model testing to measure its radar cross-section. When it was returned to Burbank, it was modified to improve its RCS and concurrently used by Hycon to check-fit a mockup of their new camera.

            Propulsion system work began to accelerate during this time period as well. A Marquardt RJ43-MA-3 Bomarc engine was wind tunnel tested at simulated Q-12 operating conditions. Per Johnson: "We were all amazed, including Marquardt, that the engine could be shut off as long as 45 seconds and still restart, due to hot engine parts."

            On March 20, 1963, Lockheed was given an official CIA letter of contract for the D-21. (At this early date, the D-21 was still being referred to as the Q-12; that designation was later changed.) The contract gave Lockheed responsibility for the navigation systems, the ramjet and the airframe. 

           The program reached another milestone in September 1963, when Marquardt demonstrated a very good engine run to the highest temperatures required, plus they appeared to be proceeding well in the development of the MA20S-4 variant. 

The D-21 was the heart of this program and was designed as an extremely high-speed, high-altitude reconnaissance vehicle. In a sense, a much smaller, un-manned version of the A-12 but with similar capabilities. It was considered to be a “one-way aircraft,” meaning that each D-21 would make one flight only and then self destruct. While still in the design stages, two Lockheed A-12s were modified to carry the D-21 on a top-mounted dorsal pylon located on the rear centerline between the engines and the vertical stabilizers. A second cockpit was installed for the D-21 Launch Control Officer (LCO) in the area of the A-12’s “Q” bay. The “Q” bay was already pressurized and refrigerated, making the modification relatively easy. 

            By October, the Q-12's overall configuration had been finalized and the equally difficult task of defining the A-12 launch system and its configuration was nearing an end. By now, the somewhat unusual M-21 designation had been assigned to the dedicated and purpose-built carrier A-12s, of which there were to be two. “M” stood simply for "Mother." At the same time, it was decided to rename the Q-12 “D-21,” thus making it the "Daughter" aircraft. The numerals "1" and “2" of A-12 and Q-12, etc., were reversed so as not to confuse the “Mother/Daughter" combination with other "A-12" variants.

            On October 1, Johnson wrote: "I proposed to Boehme that we simplify some of the load problems by letting the D-21 float at a zero moment incidence when attached to the M-21. While this concentrated the loads at one point, it reduces most of them and particularly the effect on the M-21. I also made sure that we can jettison the aircraft without power."

            On December 31, 1963, Johnson noted that wind tunnel and paper studies were leading him to believe there would be launching difficulties with the M-21 Mother ship. According to Johnson: "Going through the fuselage shock wave is very hard. I am insisting on launching at full power, but there are problems regarding fuel to air ratio to the engine and engine blow-out in this condition." A month later he wrote: "Reviewed launch conditions again, and was very upset by recent tunnel tests which show we must make a pushover to launch. This is due to making the pylon too short. This was done for structural reasons, but got us into aerodynamic troubles which weren't recognized at once, although I suspected that we might encounter such troubles when it was done."

            By late May of 1964, Johnson's impression of the D-21 had improved: "We have launch problems, transonic drag deficiencies in the basic airplane, and equipment problems as usual. But we can now haul the thing through Mach 1.0, I believe, if we can get performance like A-12 #129. Launching must be done as an automatic pushover maneuver."

            Concurrent with the D-21 work, the purpose building of two M-21s to serve as launch platforms was also underway at Burbank. A single, dorsally mounted, low-drag pylon had been developed that was sufficiently strong to support the 11,000-pound D-21 at Mach 3.0. The pylon contained a primary support post with locking hook, a secondary lock, provisions for emergency pneumatic jettison of the D-21, and a refueling line that was used to cool the D-21 and top off the D-21's tanks prior to launch. D-21 separation was to be accomplished by flying the drone off the M-21 during a slight (approximately 0.9 g) pushover maneuver. The D-21 was not forcibly ejected from the M-21. 

            A fit-check using M-21 #134 and D-21 #501 was completed successfully on June 19, 1964 in Building 309/310 at the Skunk Works. Few problems surfaced during the mating and over the next several weeks, final assembly of the first carrier aircraft was completed. On August 12, it was delivered to Groom Lake to undergo initial flight-testing. At the same time, the initial  D-21 static tests were successfully concluded. Johnson wrote in his log: "Engine deliveries are in good shape but equipment, particularly cameras, is not so good. Trying to get out seven drones this year."

            On December 22, 1964, the first flight of the D-21/M-21 combination (now referred to as the “M/D-21”) was successfully completed from Groom Lake. Per Johnson: "Bill Park flew at the end of the day. It flew well and, in spite of having low-powered engines, went supersonic on the first flight." Interestingly, on this same day, the first SR-71A successfully completed its first flight from Lockheed's Palmdale, California facility.

            Regardless of CIA and Air Force support, interest in the project remained difficult to ascertain. Johnson nevertheless doggedly pursued successful execution. Johnson wrote: "We are aiming to launch one by my birthday – February 27, 1965." But Johnson's wish was not to be. His birthday came and went with little fanfare and no D-21 launch. "We have all kinds of troubles," he wrote. "Minneapolis-Honeywell came in with a terrible story on the Kollsman star tracker which they had purchased for the M-21 guidance system. It was a complete shambles from beginning to end."

            Further flight-testing of the mated M/D-21 continued in the meantime, but without a launch. During April of 1965, Johnson noted that one flight resulted in the loss of "… both elevons on the D-21 due to flutter. We are going to put on balance weights, and add control surface locks, etc."

            In July 1965, the first of only two Air Force officers assigned to Tagboard, Major Hal Rupard, was transferred from his B-58 assignment at Little Rock AFB to Area 51 for an as yet unknown mission. Major Jack Reed followed Rupard 30 days later. Both Rupard and Reed had extensive experience as Bomber/Navigators in the Air Force’s B-58A Hustler.

            At this point, the Tagboard program was unfortunately progressing rather slowly overall. By mid-May 1965, they had the M/D-21 up to Mach 2.6 but not up to launch speeds. They were having problems using the Hamilton Standard inlet control system in conjunction with the Pratt & Whitney YJ-58s. Between May and October 21, 1965, Johnson decided to put the newer 34,000-pound-thrust Pratt & Whitney J-58 Model K engines into #135 and changed from the Hamilton Standard inlet control system to one designed by Johnson’s crew.

            Tagboard needed the engines that were supposed to go to the SR, as engine technology always lagged behind airframe technology. What was needed was to come off the tanker and accelerate right on up to cruise altitude. With the YJ-58s, you could not do that. The M/D-21 would come off the tanker, go to max burner in a dive, and then start its pull-up. The net result was that it took away from range.

            A team from the test site received the okay from the highest authorities to proceed via a C-130 to Lockheed’s SR-71 Assembly Facility at Air Force Plant 42, Site 2, Palmdale, California. Their task was to secure the first three Model K J-58s – the only three Model Ks in the inventory. These were to be used to power the Tagboard up to launch speed. On March 5, 1966, they finally saw light at the end of the tunnel and actually launched a D-21 from the back of the Mother bird. It was a great success, at least in terms of the launch. Unfortunately, the D-21 was lost about 120 nm from the launch point, but the exercise demonstrated that Lockheed had developed a successful launch technique. Johnson was quoted as saying: “This was the most dangerous maneuver we have ever been involved in, in any airplane I have ever worked on.” In this ground breaking mission, Bill Park, the Skunk Works’ Chief Test Pilot, was at the helm and Keith Beswick was the LCO. With this system, the drone could be launched outside of the EW line. The INS that Honeywell was building would update the drone from a one of a kind, 60-star celestial catalog. The other projects (A-12 and YF-12) did not have that system, but the Tagboard program wanted it and Honeywell said they could buy it. 

            Tagboard was approaching the point where the program needed additional direction; so on order from the Pentagon, Reed and Rupard flew to St. Petersburg where Honeywell was developing the INS system for the Mother bird. The Honeywell team was making progress but they needed a few more test flights and the program did not have any spare airplanes. Johnson said: “Well, you can use mine” and agreed to loan them his unique twin-engine Jet Star. 

            Reed and Rupard were tasked to write a statement of work so they could complete their assignment and put a cost on it. They rolled up their sleeves and worked for about 72 hours straight to get it done, assisted by Honeywell’s Chief Engineer, Stan Mojole, and his staff at the St Petersburg facility. A couple weeks later (around Easter 1966), all of the items were successfully incorporated. They secured Johnson’s Jet Star again and flew four or five flights with the system to validate the changes. Another critical issue in launching the D-21 from the Mother bird was fuel circulation. It was absolutely necessary for cooling that the fuel be circulated through the D-21 and then back to the Mother airplane, just like a radiator. This was of primary importance because at 35 nm per minute, the M/D-21 would be at launch point within five minutes and there could be no turning around. There was a fail-safe check system, a checklist to ensure all systems were “Go,” and when there was continuity in all of the systems, the “Ready for Launch” light would turn on. 

            The D-21 just passed air until was turned it on, but acted like a blunt instrument until it was lit. The corrective for that was to put the tanker in closer. The engine thrust and the circulation of fuel were the major concerns because the D-21’s skin temperature became very hot, running up to 600° F. The systems that ADP feared were going to have problems (e.g., control with the INS) did not turn out to be problems after all. Honeywell had done an excellent job. Computer wise, it was leading-edge technology at that time.

            With the D-21’s Marquardt power on, the Mother bird had an additional thousand pounds of thrust.. It was a play on time as to when to come up to full launch power on the drone.

            The characteristics of the ramjet were such that it did not fire up until the Mother bird was at or above 60,000 feet. It could function only at high altitudes. According to Park, things became quite hectic when approaching the launch point. His famous admonition to the test engineers was: “Either launch it or shut it down. We’re going through the window!” Once the D-21 was cranked up and given the added thrust, it was at Mach 3.5 in seemingly no time at all.

            The drone’s Marquardt ramjet was lit approximately five minutes before launch, and that gave the LCO a 160 nm back down track. There was a load cell in the pylon that indicated if the D-21 had sufficient thrust, and a full power indicator light as well. By and large, the Marquardt ramjet engine worked like a charm. Jerry Miller had developed a practically flawless operation. Following the very successful first launch, Johnson (accompanied by Boehme) went to Point Mugu to personally witness the second launch of a D-21, this time #506. Once again, Johnson was delighted with the launch. The D-21 flew over 1,200 nm 
while holding course to within one half mile for the entire flight. The bird reached 90,000 feet and a speed of Mach 3.3. It came to an abrupt halt and fell out of the sky, however, when a hydraulic pump overheated and failed. The failure was due to running the pump un-pressurized several times during checkout. It was at this point in the program that Johnson first proposed substituting the B-52H as a new launch platform, with the D-21 being propelled to speed and altitude by a solid rocket booster. As usual, Johnson’s goal was to get the most benefit from the program at the lowest cost. 

            On June 16, 1966, Lockheed staged the third and most successful D-21 launch from an M-21. It flew almost 1,600 nm, making eight programmed turns in order to stay within a line of sight from the tracking ship. Everything went as planned except ejection of the package due to some electronic failure. To quote Johnson: “It was a very successful go.”

            On July 30, 1966, on the fourth and final launch from a Mother bird, D-21 #504 suffered an asymmetrical un-start and crashed back into the M-21. The regrettable collision resulted in the death of one of the crewmembers. Ray Torick, the LCO, successfully ejected from the stricken aircraft and survived the mid-air collision, but as a result of injuries sustained during bailout, Torick was unable to get into his one-man life raft and drowned in the Pacific Ocean. Park, the Lockheed Test Pilot at the controls of the M-21, survived and went on to become the Senior Test Pilot for Lockheed’s famed Skunk Works. Rupard later recalled: “It was on the 30th of July that Torick was launching, and this was the first time that we’d ever had the other Mother bird, #134, up and flying chase. They had been trying to do this ever since the first launch. Art (Pete) Peterson and Beswick were in the chase M-21 airplane, Park and Torick were in the launch airplane, and they were flying in formation at Mach 3.3. They had just launched the D-21 when it rolled sharply to the left and fell down on the wing. Beswick was getting all of this on film. When the drone hit the Mother bird, it pitched the nose up and caused the nose to break off at the 715 splice. Torick and Park were in that part of it. The launch airplane went out of view in a hurry because the speed went from ultrasonic to zip in nothing flat. This was on a Saturday and Reed and I were at the command post. The next thing we knew, Peterson came up on frequency. They had a code word for launching the rescue forces and he initiated that code word.” There was an investigation, of course; a very lengthy investigation. The Air Force was not sure if the program should be cancelled, but Johnson convinced them that that was the only way to go. As a result, the Blackbird program’s first fatality proved to be the death of Tagboard. If Tagboard had gone operational, the launch location would have been staged at the test site, Area 51. There had been some 
discussion to keep Peterson and Park as the pilots, and Reed and Rupard as the LCOs. That way, if anything happened, they could legitimately be described as an Air Force crew.

To see the Mpeg of in-flight collision go here:



            There was a convergence of technologies happening at that time. The SR was coming into being but was not yet operational. The A-12bird was doing well, and the satellites were improving at lightening speed in their capability and reliability. The Aerospace Corporation examined Tagboard’s operation with the idea of trying to come up with better, more sophisticated ways of testing the D-21. Talented people on the ground (men like Beswick, Torrick, Bill Sass and John Wallace) were Johnson’s “brain trust” and they provided him with data and advice on a multitude of engineering matters, though Johnson was always the decision maker. It was said that Johnson could process data like a super computer! In typical fashion, Johnson said: “I’ve got an idea. If it won’t go up, we’ll drop it down. Gravity always works.” Of course, the immediate question was: “Down from what?” Johnson’s answer: “The B-52 Stratofortress.”

            With CIA and senior Air Force staffers working under Johnson’s direction, plans were advanced for Lockheed to modify the D-21 to be launched from the more conventional, tried and proven Air Force/Boeing B-52H Stratofortress. The B-52 had already demonstrated its success at NASA, launching hundreds of craft up to and including the North American X-15A2. It was readily accepted as being potentially safer than the previous Tagboard scheme. This program eventually evolved into project “Senior Bowl” (initially designated as “A” Flight and later as the 4200th Support Squadron out of Beale AFB, California). The 4200th, although officially a squadron, was actually a wing level unit with direct and primary responsibility to Strategic Air Command (SAC) headquarters. Administrative functions were through the 14th Air Division at Beale, while operational functions were handled through SAC and other still-classified hierarchy.

            Configuring the drop of the drone was the easier part of the transition, as similar operations had already met with success. Of greater difficulty were the challenges of its achieving a speed of Mach 3 at 80,000 feet, and ensuring a safe separation of the 
booster without damage to the “Tag,” as it was referred to throughout the entire program. Much work also needed to be done to finesse the remote engine start-up and on all aspects of the D-21’s navigation system.

            Early in the Senior Bowl program, the two Tagboard Air Force officers (Majors Rupard and Reed) had to go to Beale to meet with senior staff. Discussion centered around knowledge that the B?52s would ultimately be reassigned and relocated to Beale AFB, as the Agency didn’t want them at the Ranch anymore. SAC also sent some people from headquarters out to Beale with a whole shopping list of things that they wanted to get done.

            Reed and Rupard arrived at Beale on a Sunday night and were ready to go to work the next day. According to their testimony: “These guys from SAC came traipsing in and they started laying out the agenda. They said, ‘Today we’ll do this and Tuesday we’ll do that and Wednesday….’ We looked at each other and told them, ‘Tell you what: By 2:00 this afternoon, we’re going to have all this done.’ The SAC representatives countered with, ‘We’ve already told our bosses that it will take all week.” Rupard’s response was: “Well, you guys can do what you want, but we’re going to get it done. Are you ready? Let’s go!” 
They missed their 2:00 goal, but by 4:00 everything had been worked out. Each person had been assigned a task and everybody knew what they were going to do, including who and when to call. They had it all laid out – that‘s the way ADP did things. There was no fooling around because there was no time for it. As usual, Rupard and Reed worked outside normal Air Force procedures. They forged ahead and took care of housekeeping details, like how many people were needed to launch the airplanes. Initially the thought was that Reed and Rupard would fly the missions as the LCOs, but the brass at the IG’s 
office said: “No, you two have to do the other things. You have to run the operation because you’re already familiar with it and you’ve been on the ground out there for quite a while now.”

            The first task on Reed and Rupard’s plate was to obtain the B-52s. The program needed the B-52’s H model, as they were the last to be manufactured. Major Rupard talked to General Leo Geary, Colonel Newburn and Colonel Hartley about sending selected B-52H aircraft to the depot for modification. General Geary went to the top, stating that two airplanes were required and the modifications on one of them had to begin immediately. It was still a quasi-black program, so he was able to go the depot at Kelly AFB, San Antonio, Texas and simply pick one off the line with the order: “This airplane is going to Palmdale.” Johnson redirected his forces and put I.J. MacNamore in charge of the modification program, working together with Bob Murphy. Murphy made room for the aircraft and the Air Force delivered the first Boeing B-52H (#61?0021) to the Lockheed 
Palmdale facility, Air Force Plant 42, Site 2, on December 12, 1966. As soon as the B-52H touched down, it was pulled into the hanger. The ADP team began work on it immediately.

            It was important for the Lockheed engineers to get up to speed quickly on the various B-52 related systems. The IG’s office had made a commitment that Lockheed would be ready when ADP called them to action. Reed said: “Well, that’s easy to do. Just send them out to the air base at Marquette. That’s where the training is going on. They’ve got the systems all set up out there.” He set forth his plan to higher headquarters and obtained their support. The Lockheed engineers were given GS-16 status, along with the proper GS-16 ID cards. Reed and Rupard briefed the K.I Sawyer Wing Commander on the GS-16s who were being sent for training and assured him that they would take charge of them.

            During the change over from Tagboard to Senior Bowl, all the remaining D-21s had to be moved to Burbank for modification. Johnson said it was absolutely necessary to modify all of the D-21s to “B” standards. Reed suggested using C-5s to bring them in. They had to do it at night, like they always did, to keep prying eyes from seeing whatever they were moving in or out of the Skunk Works. Fortuitously, the aircraft were painted black. It was decided to bring them in at 2:30 in the morning. According to Rupard, when he met with the airport manager, he thought to himself: “I may as well just unload on him here 
before he sits down,” so he jumped in with: “I just wanted to come over and let you know that we’re going to need to close down your terminal for a brief period of time at night.” The airport manager’s initial response was like he’d sat on a tack! He leaped into the air and got red-faced and demanded to know why. But when Rupard replied: “We have some cargo that we’ve got to offload at nighttime,” the airport manager immediately understood: “I’ve got the picture. That’s all I need to know.” The terminal was closed from 2:00 AM until Southwest Airlines’ first early-morning takeoff.


A few months after the initial delivery, the second B-52H (#60‑0036) arrived.  Both B-52H aircraft required extensive modifications in order to carry and launch the D-21B.  In its proposed configuration, the Senior Bowl B-52H/D-21B combination could conceivably cover any location anywhere in the world with its global reach capabilities.  Coupled with the D-21’s combat radius of over 3,000 nm, the enemy could run but not hide.

Senior Bowl called for the modified D-21 to be carried on an under-wing pylon, similar to that used for launching the X-15 from the NASA/Boeing NB-52B.  With the change in launch aircraft and the subsequent modifications to the D-21, the program warranted a new designation.  There had been no mock-up, model or working drawings of the D-21A (although some casual sketches may have been made), thus the program’s progression officially went straight from the D-21 to D-21B. 

            Major modifications to the two B-52H Senior Bowl aircraft included elimination of the ECM operator and tail gunner’s panels at the upper-rear crew station, and installation of identical launch control panels for the D-21Bs on the right and left pylons.  Two camera mounting stations were installed in the B-52H’s left and right forward wheel wells.  These mounts held a set of 35mm very high-speed cameras used to record the launch of the D-21B from the B-52H.  The cameras were aimed at the D-21B from different angles and with a variety of lenses in order to capture the D-21B as it dropped from the pylon.  The cameras had film magazines that held 1,000 feet of color film and ran at a speed of about 100 inches per second.  In addition to the fuselage mounted cameras, there was also a wide angle, downward looking, high speed camera mounted inside each pylon to film the D-21B as it dropped away from the aircraft.  The cameras would run for about ten seconds, though the reality was a virtually instantaneous action with only a few seconds passing until the drone was out of sight of the camera’s eye.  When played back on a normal 16 inches-per-second projector, the film would provide an exceptional slow-motion picture of the launch.  This documentation, viewed millisecond by millisecond, proved to be exceedingly useful in diagnosing any problems that occurred during the launch.  It was always a fascinating experience to watch the drone drop away and light off!  The addition of special pylons, telemetry gear and communications systems, together with the associated wiring and instrumentation, completed the modification of the aircraft. 

Once the modifications were done, the aircraft and crew headed to the Ranch to begin the training process.  The operational phase had commenced and once the training phase was finished, they moved again, this time to Beale.

A major concern, once reassigned to Beale, was the booster.  It was a 60 foot rocket motor that hung from the bottom of the drone.  There were problems with the booster’s explosive bolts.  One of the very volatile, 60 foot long rocket motors dropped off a D-21 onto the ramp at Beale late one night.  Lockheed took off all the explosive bolts and sent them to the lab at Wright Patterson.  A team examined them and when their evaluation was complete, Johnson demanded: “Hire an expert and get him to talk to the people at Wright Patterson so we can get on with this.  Get a fix!

There was an expert, a doctor with the Aerospace Corporation, and Rupard asked: “May I brief him?”  General Geary’s office responded: “No,” so when Rupard met the doctor, he simply said: “We’d like to retain you to go back to Wright Patterson to examine some explosive bolts that have been malfunctioning on our system.  Would you go back there with me and talk to them about this problem?”  Rupard gave him the non-classified, technical rundown on everything and the doctor agreed to participate, so they jumped on a Lockheed support airplane and flew back to Wright Patterson.

It turns out that Lockheed had gone to sole source contracting.  Nobody had had any problems with these bolts before, but this bunch suffered from hydrogen imbrittlement.  What that meant was that in the process of manufacturing the bolts, too much hydrogen had been introduced in process and the bolts had become imbrittled.  There was no way it could have been detected.  The doctor told Lockheed that he’d seen a B-36 sitting on the ramp when its landing gear simply collapsed.  The primary bolts inside had given way, a typical case of hydrogen imbrittlement.

By the time the day was over, they had come up with a solution.  The first thing they had to do was go from a ¾″ bolt to a ⅞″ bolt, and Lockheed sent someone to the plant to watch the process to make sure no additional hydrogen was added.  Lockheed thought that they had solved this problem 20 years before, but the doctor said: “Oh, no.  This happens all the time.  I’m also called in on launches of satellites for this particular problem.”

On September 28, 1967, D‑21 #501 was accidentally dropped from a B‑52H in fight as a result of what Johnson referred to at the time as: "… poor workmanship ... a stripped nut in the forward right attachment to the pylon." Johnson also noted that it was: “… very embarrassing."  The booster, in fact, fired after the inadvertent launch and was quite a sight from the ground!  During later investigations, they determined that it was hydrogen imbrittlement, not poor workmanship, that led to the loss.  This occurred about 60 miles north of the Ranch.

All D-21B vehicles dropped from “A”-Flight/4200th Support Squadron B-52Hs were launched from the starboard (right side) pylon only.  The port (left side) pylon station carried the spare D-21B and booster and was never used for operational launches.

The Marquardt RJ43-MA20S-4 ramjet engine powered both the D-21 and the B-52-launched D-21B.  When released from the pylon of the sub-sonic B-52H, the D-21B’s Lockheed designed rocket booster propelled it to above Mach 1.5, at which speed the ramjet would light.  The nose cone of the booster was fitted with a Marquardt B-4 supersonic ram air turbine to provide the electrical and hydraulic power necessary during the drop and boost phase.  It was Lockheed’s Missile and Space Division in Sunnyvale, California that cleverly developed this unique propulsion system.

After the D-21B had been installed on the B-52H pylon, the rocket booster would be mounted to the underside of the D-21B using the original connecting points where the D-21 had formerly been attached to the back of the M-21. 

The B-52H, with its D-21Bs and boosters, was then ready to take off for the required sortie.  At a precise, pre-planned point in time and geographic location, the LCO on board the B-52H would start the sequence of operations by first dropping the D-21B from the pylon.  Following separation from the pylon, the booster would ignite and then propel the D-21B to a speed in excess of Mach 3.0 at or above 75,000 feet.  After a burn time of about 90 seconds, the booster would separate from the D-21B via explosive bolts and the D-21B would begin its programmed solo sortie. 


In the summer and early autumn of 1967, SAC began assembling the nucleus of the Air Force unit that would prepare, launch, fly, recover, and maintain the D-21Bs being carried by the modified B-52H aircraft.  Assignments were levied throughout SAC for approximately 180 officers and airmen having the following skills:


1.     Those personnel who would fly or work on the B-52H only. 

2.     Those personnel who would work on the D-21B only. 

3.     Those few personnel who would be involved in both the B-52H and D-21B programs. 

4.     A small, specially trained component of Security Police which would initially be responsible only to “A” Flight (later they would support the 4200th Support Squadron). 

5.     Supply, administrative and other support personnel that would not be directly involved with either the B-52H or D-21B programs, but who would make important contributions to these projects. 


In addition to the Air Force personnel who would comprise the bulk of the staff, many Lockheed, vendor-support and advisory personnel would also be part of the team. 

            The requirements for all personnel were exceedingly stringent.  The Air Force was, in effect, hand-picking these individuals in order to put together the very best support for Senior Bowl.  In November 1967, for example, Glenn Chapman, a Staff Sergeant Avionics Specialist (and former U-2 Sensor Specialist) then stationed at K.I. Sawyer AFB, Michigan, received a new assignment.  When Chapman asked about where he was being sent and the nature of his next job, he was informed that the assignment was for “A” Flight at Beale AFB, California.  When Chapman further inquired about the numerical designation of his new outfit, he was told only that it was “A” Flight, 14th Strategic Aerospace Division (SAD) out of Beale AFB, California.  No other information was forthcoming.

When Chapman arrived at Beale for in-processing on January 3, 1968, he was intrigued to find out that the term “A” Flight was a rather mysterious designation.  The staff in the personnel, finance and other permanent party support operations at Beale had indeed heard of this new term, but they still had absolutely no idea what it represented.  The general consensus was that it was a new addition to the very prestigious 9th Strategic Wing that flew the SR-71 aircraft.  In reality, “A” Flight never played even an indirect role with the 9th or the SR-71 program.  As an “A” Flight designee, Chapman was moved to the head of the line of all personnel who were in-processing at that time, even passing up a Lt.  Colonel who became quite indignant that a SSgt had been served before him.  Evidently all base services had been informed that anyone processing into “A” Flight should be given precedence in all administrative procedures.  

“A” Flight appeared to consist of nothing more than a single-story wooden building directly across the street from the 9th Strategic Reconnaissance Wing Headquarters.  No one seemed to know anything except the acting First Sergeant and Chief Clerk, and he wasn’t talking.  A scenario soon developed wherein everyone in “A”  Flight would report to this building at 0800 each morning to drink coffee and “swap a few lies” for about half an hour.  Oddly, they would then be dismissed, told to go home, and directed to simply call in once in a while.  For about six weeks, this was the ongoing routine.  In one respect the unanticipated freedom was wonderful, yet in another sense it was emotionally agonizing, with some project members feeling abandoned, useless, or worse.  However, outside their knowledge, this time was being well used for the project to achieve final operational status and obtain all of the clearances needed – all critical steps that had to take place before anyone could be told the nature of their assignments. 

Around the middle of February 1968, Chapman was called in for his first meeting with Squadron Commander Colonel Arden B. Curfman (who, behind his back, became fondly known as “ABC”).   ABC’s number two man was Lt.  Colonel Baldwin, “A” Flight’s “chief spook.”   In reality, Baldwin was the Air Force’s liaison at the Skunk Works for the Senior Bowl project.   This day would prove to be the one and only time anyone on base would see the Lt.  Colonel in uniform, and all personnel were instructed to refer to him simply as “Mr.  Baldwin” from that point forward. 

            Once Chapman was cleared for Senior Bowl, “ABC” and the “chief spook” proceeded to give him the most intense security briefing he had ever received.  The encounter lasted about three hours with no one except Curfman, Baldwin, Chapman, and occasionally the Chief Clerk in attendance.  When something highly classified needed to be communicated, it was not done orally.  Instead, Baldwin or Curfman would write the critical information on a piece of paper and hand it to Chapman.  After reading it, Chapman would be asked if he understood the material.  With his affirmative answer, the paper was set afire in an ashtray, destroying it immediately.  At the conclusion of the briefing, the Chief Clerk gathered the ashes from the ashtray for further disposition.  Chapman’s impression was that he read more than he heard that day, because virtually everything they told him was highly top secret material.  Most important was the information they gave him concerning where they would be testing the D-21B, proper conduct whenever he or anyone else had to actually visit the Skunk Works or the test site, and the location where it was anticipated the program would be put into operational use. 

            Later that week, along with three other men in his career field, Chapman began attending “night school” in the same briefing room.  This training was conducted by a vendor’s technical representative and lasted about a week, with focus on the “payload” system and on the D-21B itself.  The first night, before they were too far into the session, someone jokingly asked the trainer if they were going to be working on something like the X-15.  With a very stoic look, the trainer informed the group that they were: “… not too far off track on that.”   Chapman would later learn that this trainer never joked around.  The trainer’s personality proved to be a challenge to work with, but he was regarded as a very knowledgeable, intelligent, top-notch teacher.   Following that one week of schooling, the team members never saw him again. 

A few individuals failed to pass clearance and were immediately re-assigned to other bases or units.  They left having received no information other than the project name, Senior Bowl, and the knowledge that they had been temporarily assigned to something called “A” Flight. 

During 1968, most of the cleared “A” Flight personnel were sent out to the test site each week where they learned the ins and outs of the D-21B, the modified B-52H, specifics about the specialized equipment, and the plan of how everyone fit into the program.  The team would meet at Beale AFB Base Operations every Monday morning at about 0600 and climb on board a civilian Fairchild F-27 turboprop aircraft for the flight out to Area 51.  While stationed at the area, the men were housed in mobile homes.  They were not allowed to leave the test site but were free to wander within its boundaries.  At 1730 hours each Friday, the team would again board the F-27 for the flight back to Beale.  Their weekends were free, but come Monday morning the workweek away from home would begin all over again. 

Jerry Miller, of Marquardt, trained a number of the "A" flight personnel on the testing, operation and maintenance of the ramjet, ramjet test-set trailer, APU and RAT.  When "graduation day" came, members of the Beale team expressed their gratitude in a teasing way by chasing down the civilians with a stencil and a can of spray paint.   Miller proudly wore his decorated shirt later at both the area and at Beale.  It still hangs in his closet. 

In late 1968, the unit gained operational status and the “A” Flight designation was changed to the 4200th Support Squadron.  With the change in designation, program members lost some of the priority treatment they had been enjoying, but they retained enough status (far more than the SR-71 personnel) to know that they were in quite an elite outfit.  While the team had been stationed at the test site, civilian contractors had been busy remodeling the nose dock at Beale (located near the current site of the fuel cell and phase hangars) as the new home for the 4200th Support Squadron.  In December 1969, the unit was permanently moved to the remodeled quarters at Beale.  The two Senior Bowl B-52Hs (#61-0021 and #60-0036) were parked at the farthest point of the northern end of the ramp near the alert facility. 

The 4200th had its own cadre of Security Police staff who, like the rest of the 4200th, had been fully cleared for Senior Bowl.  All team members (from E-2s to Colonels) were allowed exemption from the usual inspections, Maintenance Standardization and Evaluation Program (MSEP) inspections, base details and other irritants that everyone else at Beale had to contend with.  The 180 or so people who comprised the 4200th Support Squadron thus became a very close-knit group. 


One of the things that really helped the 4200th Support Squadron was the fact that the SR-71 was there.  It was also a program with Johnson’s signature on it – that was obvious by sight.  Since both groups were important, operating protocol between the two organizations went smoothly and the gap was bridged.  The 4200th was very enthused about getting their systems together and the job done.  They knew their program was going to be a short-fuse operation because satellites were rapidly coming into operation and prominence.

When you have a pure black program, you are not stifled by the general regulations that prevent you from stepping on anybody’s toes.  In black programs, it’s a given that you never let anyone think you’ve stepped on their toes; you just tread so lightly that they don’t even know that you’re there!

Prior to an operational launch, a 24 hour heads-up would typically be given.  The birds were just about ready all the time.  The crews had done enough training at the Ranch that everybody knew what they were supposed to do.  The TO&E that Reed and Rupard had recommended was about 80 people but SAC had more than doubled that amount, so they had plenty of staff.  That was just how SAC operated!

The number of birds ready for launch was usually four, two on each airplane.  “A” Flight worked to keep them ready to go on a semi-alert basis at all times, in preparation for the outside chance that something big would come up where you had to scurry everything you had.

Whatever the White House, Pentagon or the NRO wanted to do, they would be able to select the system that they thought was best for that particular mission.  They liked this system because there wasn’t a man involved in it.  They would give 24 hours notice but the launch time might be delayed, depending on what was needed.  Sometimes the airplanes would be readied and it would be two days before they would be released and given the final go.  Mission planning was just like the normal mission planning for any B-52.  They would always look at the launch area weather because you wanted to have about a 24 hour forecast available.  Launch altitude for the B-52 didn’t make much difference so long as you had a clear hole above through which to send it (e.g., without a jet stream in the way).  The craft could be punched off and up to altitude in nothing flat.

Typically, the launch from the BUFF (“Big Ugly Fat Fellow”) was behind the accepted EW line (the early warning line of 300 nm).  Of course, one could never discount the fact that there might be a submarine out in the water or something, but it would have to be a fluke for that to happen.  The Air Force, CIA and the NRO knew where all the radars were and they would plot the EW line so you could launch comfortably outside the boundary.  The B-52 is relatively slow, but at 500 miles the fighters would run out of gas so it was considered quite safe.  Any detection would be minimal.

The CIA has shared bits and pieces regarding the effect of D-21 overflights over the People’s Republic of China.  There was some rather spirited conversation between the Chinese that other sources had picked up.  Even though the Chinese were well trained, they didn’t have the foggiest idea as to what was going on!  They didn’t understand the drone and it gave them bad radar signatures.  They did have some A-bird and SR-71 signatures and they kept thinking that this was one of them: “No, it’s bigger; no, it’s smaller; no, it’s this; no, it’s that.  How did it just show up?”  Their confusion verified that our forces had done a good job with security, and that’s one of the trickiest things to accomplish.

One D-21 is believed to have been shot down (or crashed) in denied territory on the last operational mission.  Somebody made a tape of it, from outside the sphere of interest, which popped up in another program.  The CIA put that together and got possession of the tape, which was quite interesting, particularly when you had the Chinese version of it!  You’ve heard of the Chinese fire drill?  Well, that’s about what it sounded like!

It was amusing footage as they tried to figure out what it was.  With ego being what it is, they were undoubtedly sure that they knew everything we had – but this did not fit into any of their templates.  More than one person was heard on the tape, and in the confusion and excitement of their communications together, their voices became quite shrill! 

Ground troops never had any clue as to when a mission would be levied.  From their perspective, the missions appeared to be called spontaneously.  They could go weeks (and sometimes even months) with no specific duties other than extensive training.  In addition to military staff, about 10 to 20 percent of the 4200th’s Support Squadron was comprised of Skunk Works personnel and vendors’ technicians (otherwise known as “technical representatives”).  These team members assisted in training and helped prepare for each levied sortie.  They also provided all parts and hardware for the D-21B, plus other equipment.

The D-21B was constructed primarily of titanium, stainless steel fasteners and composite structures.  Composites were in their infancy at that time.  All wiring in the vehicle, especially in the equipment hatch, was Teflon-coated and gold-alloyed.  It required a whole new set of skills to learn to work with this wiring which, because of its makeup, was extremely brittle and easily broken.  A special type of solder was developed for use with the wiring, as were unique new soldering techniques.  Lockheed developed a new tool called a “solder sleeve” that was frequently used.  It was basically an in-line, stakon-type wire splice, together with a shrinkable, sheathed plastic sleeve equipped with special solder encircling the inside of the sleeve.  The technicians would strip the wire insulation using high-temperature heat strippers and then put the stripped wire ends inside the solder sleeve.  Next they would subject the sleeve to high temperature via a heat gun (something like a blow dryer except much hotter) until the solder would melt and the plastic sheathing would shrink, creating a perfectly spliced connection.

The stainless steel screws used with the titanium parts were very unusual.  They were mostly flat and resembled a Phillips head.  The slots in the head were offset, however, and looked vaguely like the old Nazi symbol, hence they became known as “swastika” heads.  The concept behind this configuration was to provide higher torque for tightening or loosening.  Special swastika-shaped apex drivers were developed, and an apex holder welded to a piece of 3/8” square stock about 24” long became standard equipment.

When tightening these screws, the apex driver would dig into the screw rather than slip out of the head (as a Phillips-type head might be prone to do).  It was the same when taking them out.  Due to the extremely high stresses encountered during flight, the screws in the hatch covers would become even tighter.  The swastika drivers made it a little easier to get them out because they would again dig into the head rather than slip out.  Of interest is that each screw was used only once; if it had to be removed, it was thrown away and replaced by a new one.  This happened to literally tens of thousands of screws during the life of the program.

If crew a member needed additional screws, the request had to be made in person.  When one new staffer was making such a request, he was asked how many screws he needed and replied: “Just a handful.”  To his surprise, he was ordered to go back and count out the exact number of screws required.  He returned with the figure of “30 or so,” and the parts man proceeded to count out exactly 30 screws and made him sign for them.  He later learned that each one of these swastika-head screws (about ½” long by 10-32) cost more than $5.00 apiece – undoubtedly a hefty profit for Lockheed!

Lockheed was covering new ground with both the D-21 and its support systems.  Every item, including all of the test fixtures, needed to be designed from scratch.  As an example, one very innovative and talented Lockheed engineer designed a test rig that looked like something out of “Star Wars,” with a myriad of buttons, dials and switches.  There wasn’t a piece of solid state electronics in the entire fixture; it was all relays.  The rig was about four times the size of a standard office desk and he had to stay with it all the time just to keep it running.  Others thought it was an absolute nightmare, but he made it work!

            During one particular crunch time, Skunk Works engineers had to get a D-21B out the door IMMEDIATELY.  They had already been working around the clock for several days in a small building behind Lockheed’s main production building #309/310.  An inspection team arrived and the lead engineer felt like the team was interfering with the work, so he ordered them out of the facility and locked the door behind them!  The D-21B successfully completed its checkout and was shipped on time.  Later, Johnson asked the lead engineer whey he had taken such action and the reply was: "Because they were in my way.  They weren’t adding to anything.  They were just pi**ing me off!”  Johnson’s calm response was simply: “Oh, okay,” and that was the end of it.

Once a mission was levied, the first task was to determine who would be going TDY to the Advance Party (ADVON) locations.  This usually meant that at least 30 percent of the D-21B staff and some of the B-52H people would be away from Beale while preparation for the mission was taking place.  There were three land based locations where the team would most likely be sent to recover the mission payload: Anderson AFB, Guam, Kadena AFB, Okinawa (as a backup to Anderson), or Hickam AFB, Hawaii.  An assortment of Navy ships also served as destination points.

1.      The first location (Anderson AFB, Guam) could receive and deploy the B-52H/D-21B aircraft as scheduled, allowing the B-52H/D-21B to depart or recover from its mission.  On all but one occasion, however, the missions commenced directly from and returned to Beale.  If weather conditions prevented JC-130 mission recoveries at Anderson, then Kadena AFB, Okinawa would be used as the alternate recovery site.  The JC-130s would land at the closest of the three bases to off-load their valuable cargo into a waiting KC-135 for its flight back to Beale.

2.      Once the D-21B had completed its mission, the hatch would be transported on a KC-135 to Beale AFB, California for sensor recovery operations.

3.      The third location, a “floating TDY” (usually a Liberty Ship or Navy destroyer), could perform secondary recovery operations of the hatch in the event the “Cat’s Whiskers” (JC-130B) was unable to snag the hatch in flight.

No site-permanent personnel or personnel aboard the ships were cleared for Senior Bowl.  The captain of the vessel would have been briefed by the ranking individual (usually an officer or high-ranking NCO) only on what he needed to know in order to get the ship where it needed to be.  The ranking individual at the ground-based site would be briefed in the same fashion. 


The hatch would be removed from the D-21B and taken to the Payload Shop.  Other technicians would start built-in testing (BIT) and in-flight checkout (IFCO) procedures on the D-21B with special test equipment in the hangar.

Inside the hatch were the main avionics packages made by Honeywell.  These included the Inertial Navigation System (INS), the Automatic Flight Control System (AFCS), and the Air Data Computer (ADC).  These were installed within the hatch so they could take advantage of the same cooling system used for the payload, as well as be recovered with the hatch at the end of the mission.  After the hatch was sealed, these systems were also tested via IFCO checks.

The autopilot was a combination of digital and analog technology, all housed in a one-cubic-foot box with three line-replaceable units (LRUs)).   One LRU controlled the roll and yaw, one controlled the pitch axis, and the other was a power supply for the in-flight checkout equipment.  Due to the extreme lack of space and high vibration environment on the aircraft, new construction techniques were developed.  The autopilot utilized what was called “welded cordwood construction” because the resistors, capacitors and other pieces of equipment stood on end, resembling cordwood.  The ends of these units were welded together and to the motherboard with a technique similar to a spot weld but without using solder.  Maintenance proved to be exceptionally difficult because the units needing work had to be drilled out, the Mylar separators that held the devices in place had to be patched, and then the units being replaced had to be re-welded into position.

The in-flight checkout equipment was a “first” for its time, using an A-to-D converter plus control circuitry.  It had to exercise the entire aircraft before each mission, both on the ground and while airborne.  Final checkout in Burbank, California involved taking the aircraft into a small hanger and firing it up using an MA-1 air source.  The air needed to be cooled in order to run the turbines and the procedure took many hours.  When the Burbank team was finished, the unit would be packed into a covered system that looked like an ordinary box and trucked to Palmdale, going from there to either Area 51 or Beale AFB.

One of the checkout tasks was to simulate an actual mission using the test set to ensure that the D-21B was operating properly.  Interestingly, this same testing procedure would already have been run at the Skunk Works facility, and the vehicle would not have been shipped had it not been in perfect operating order.  The great puzzle was that while in transit between the Skunk Works and Beale, “gremlins” seemed to creep in and cause the D-21B to go out of whack.  Sometimes the readjustment needed at Beale was minor, but at other times it was a significant problem.  In fact, not once did a D-21B arrive at Beale in working order!  (Mind you, there could be no fault levied against the Skunk Works because when they shipped it, it was confirmed to be in perfect condition.)  Following the on-site tune-up, all would appear to be well until the D-21B commenced its take-off roll from the B-52H.  Again, problems seemed to pop up out of nowhere.  Sometimes following the launch, for no apparent reason, the whole thing would go haywire!  The D-21B was reputed to be a very fragile and temperamental aircraft.

Of note, Johnson himself dispatched a D-21 “Tiger Team” to find out why the D-21 was having so many problems and failures.  The Tiger Team’s determination was that while the D-21 had a cutting-edge airframe and propulsion system, it utilized outdated 1940s-era electro-mechanical relays, technology and hardware that were not compatible with speeds in excess of Mach 3.3 at or above 90,000 feet.

While the skilled technicians were working their magic, the Senior Bowl team would begin their own system checks and prepare the payload for installation into the hatch.  Once installation was complete, they would perform their BIT and IFCO checks.  When those were done, the hatch would be delivered back to the D-21B where its BITs and IFCOs were still in progress and then the tests would be rerun.  After all of the connections had been made, the hatch would be installed in the equipment bay at the forward underside of the vehicle.

Connectors, pins and broken wires caused innumerable problems, as did the difficulty of having to work in such tight confines.  Very stiff, Kapton-coated wire was used because it was more compact and saved space.  Kapton was new technology at that time and the awkward wires tended to break while the connectors were being seated, or they would create a real challenge by moving the pins.  One of the primary connectors was a square Cannon plug with 100 pins.  It had to mate perfectly with the matching connector inside the equipment bay.   This joining proved to be extremely difficult because the pins on the connector were easily damaged and there was minimal room in which to work.  There were times when everyone would take turns attempting to get it right until all of a sudden it would go together, seemingly all by itself!  On one occasion, a connector was found to have been damaged during preparation for a mission.  It took more than a week for the new harness and connector to arrive from Lockheed.  The hatch had to be taken apart again, the harness replaced, more BITs and IFCOs performed, the hatch re-sealed, and the component finally re-installed back on the D-21B.  To make matters worse, work on the BITs and IFCOs on the D-21B had to be resumed from the point where it had previously been suspended.  So much additional work caused by one small broken pin!

Of note is that prior to putting on the waterproof cover, workers would apply RTV sealant (the old, red, flaky-when-dry stuff) on and around any areas of the hatch that weren’t secured.  After that, about two hundred screws needed to be installed to hold the cover down.  The screw holes in the hatch had been accurately drilled and tapped at Lockheed.  As the crew worked to attach the cover, however, it often seemed as though some of the holes had miraculously moved!  With the RTV sealant now blocking their visibility, it proved very difficult for the crew to tighten the cover down properly.  They persevered, however, and the people using the swastika screwdrivers usually ended up “red-handed” due to the extreme physical effort.  On top of that, the RTV sealant was miserable stuff to work with and difficult to remove from hands and fingers.  A spray can of WD-40 helped to get it off.

When the D-21B and the hatch had passed all of their BITs and IFCOs, they would be taken out to the B-52H under a canvas shroud, usually very early in the morning.  The completed D-21B, minus its booster, was then uploaded onto the right inboard pylon.  Once securely hung from the pylon, the booster (or as it was called due to its similar shape, the “cigar”) would then be brought out to be installed on the underside of the D-21B.

            The solid rocket booster was extremely volatile and very sensitive to static electricity.  Had any static been released while in close contact with the D-21B and/or the booster, it could have caused a low-amperage, high-voltage arc of direct current (something like lightening) that could have ignited the booster and sent it on its way.  This was never allowed to occur, however, as it would have caused total havoc and resulted in the loss of the D-21B, probably the B-52H as well, and endangered other aircraft, buildings, people, and whatever else might have been in the line of fire.

To prevent static electricity, everyone within 25 feet of the booster had to wear “leg stats.”  Pads were worn underneath the each crew member’s shoes that connected to a strap worn around the upper part of the lower leg (looking somewhat like a garter).  This contraption allowed the crew members to maintain conductivity to the ground at all times, so long as everyone remembered to keep one foot fully flat on the ground anytime they had to kneel around the booster.  Each and every member of the “A” Flight/4200th crew made sure that the leg stats were always used.  As an extra precaution, most of the ground crew wore them when the D-21B was being installed and the cigar wasn’t even at the site yet!

When it came time to launch Senior Bowl, nearly all personnel were busy at the squadron or at specific duty stations taking care of any last-minute hang-ups or problems to ensure that the scheduled take-off time would be met.  Timing was everything.  If the bomber wasn’t in the air on time, the mission would have to be aborted and rescheduled.  Although this did happen on a few occasions, most take-off times were successfully met.

Before and during each Senior Bowl test flight, the D-21B’s electronics were monitored closely via telemetry to the radio shack in the base hangar.  One day, while the B-52’s engines were spooling up, Frank Brink (the Honeywell technician who was responsible for the "health" of the AFCS and ADC) noticed a problem on the telemetry readout.  He boldly marched out to the flight line, held up his hands in front of the B-52 and announced: “You're not going."   When the Crew Chief argued: “Yes, we are," Brink firmly retorted: “No, you're not!"  Brink won the argument, possibly preventing the loss of the mission in flight which would have been tragic and very expensive to the program.  It was just another day in the life of Senior Bowl!

Once the B-52H was up on mission, it was totally out of the ground crew’s control.  It was the LCO’s job to get the D-21B to its drop point accurately and exactly on time.  The LCOs were able to perform IFCO checks while airborne and they also had some manual control of the D-21B if needed.

Any velocity or position errors would grow over time during the long flight and these had to be corrected prior to launch.  For example, en route to an operational launch point in the northern part of the Sea of Japan or the South China Sea, a complex set of procedures had to be followed to update the D-21’s inertial set in order for it to go its full mission distance.  The velocity input was pulled off the B-52H.  The BUFF would go through a series of U-type maneuvers, making three orthogonal (right-angle) turns in order to update the velocity and position parameters.  The stellar systems were crude by today’s standards.  The computer in the navigation system was comprised of magnetic drums programmed in their own assembly language.  It was a 24-bit machine, limited in memory, and yet it had to store the entire sequence of tests for in-flight checkout because it was the only computer on board.

After the final IFCO checks had been performed by the LCO and if everything was in order, and if the operation had not been recalled (which happened occasionally), the LCO would initiate the launch sequence and drop the D-21B/booster.  Once the D-21B/booster had been released from the pylon, very little control of it (other than some telemetry signals from the LCO on board the B-52H) was possible.

Instantaneously as the D-21B/booster dropped, automatic sequencing within the D-21B would kick in.  Approximately one to three seconds following the drop, the booster would ignite and the AFCS, INS, ADC and other systems inside the D-21B would follow their sequencing to start up the ramjet and propel the D-21B into proper trajectory.  The AFCS had a very simple pitch program, putting the D-21B into a steep climb and leveling off at the end of the booster’s 90 second burn.  Following that, the booster would be jettisoned, the programmed operations from the on-board computer would commence, and the D-21B would be on its way.  Now all that could be done (other than a few actions remaining under control of the LCO) was to wait and hope.  The most vital command the LCO could order was a signal for immediate destruction of the D-21B, available in the event anything went wrong or if conditions demanded abandoning the bird for safety and/or security reasons.

The mission track was programmed into the INS, which provided steering commands to the roll channel of the AFCS.  The ADC was pre-set to keep the D-21B flying at a fixed Mach number.

Once en route and on target, the payload/camera equipment would begin operating and continue until the computer shut it down.  The D-21B would then be vectored into its final return leg.  At an exact spot determined by the INS, the Marquardt engine would shut down and the vehicle would decelerate and enter a controlled descent to a lower altitude of about 60,000 feet.  At a pre-determined point, explosive bolts would fire and the hatch would be ejected.  The hatch-less vehicle would then tumble on its way until an explosive charge would destroy the entire D-21B, leaving behind only a meaningless residue of ashes, debris, and whatever else remained from the blast.

After egress, the hatch would drop to an altitude of about 15,000 feet or so when the drogue and main chute (attached to the inside of the hatch) would engage, trailing the hatch via cable a few hundred feet below the chute.  There were calculated markings on the cable that the “Cat’s Whiskers” would hopefully engage.  The hatch would then be taken into the aircraft and delivered to the closest location for recovery operations by ”A” Flight/4200th Support Squadron personnel.

If the JC-130B missed its target, the ship (location number three) would then attempt recovery operations performed by ”A” Flight/4200th Support Squadron personnel already on board.  In the case of a successful recovery, the ship would return to port and the camera pallet would be sent to the photo analysis center at Beale AFB, California.  Unfortunately, the Navy was never successful in recovering a hatch.  It was always either recovered by the JC-130B or lost completely.



     Chart compiled and edited by Leland Haynes, Webmaster, SR-71 Blackbirds. Copyright use may be requested here:

Serial Number

Launch Date

Launch Vehicle

Distance Flown (nm)


Functional Fit/Captive Flight:

19 JUN 64



First fit check in building #309/310; no launch.


22 DEC 64



First mated flight of M/D-21; no launch.

Tagboard Launches:

05 MAR 66



Crew – Park/Beswick.


27 APR 66



Crew – Park/Torick.


16 JUN 66



Crew – Park/Beswick.


30 JUL 66



Crew – Park/Torick.  M-21/D-21 mid‑air collision; Torick lost his life; aircraft and drone lost; Tagboard program canceled.


28 SEP 67



Drone fell off the B-52H pylon after leaving the test site because of poor workmanship in attempting to re-tap a stripped nut in the right forward attachment to the pylon.  This was a very embarrassing and costly failure.  The booster fired and according to witnesses, it was quite a sight indeed!


06 NOV 67



The first B-52 launch was unsuccessful.   The booster took it to altitude but it nosed over and dove in after 150 nm of flight.


02 DEC 67



Number 509 flew only 500 nm.  It was too low in altitude, slow on speed, and quit flying when it ran out of hydraulic fluid.


19 JAN 68



After a few minutes, #508 went out of control and was lost.


10 APR 68



An unsuccessful launch; the engine did not light.


16 JUN 68



A very good launch, reaching an altitude of over 90,000 feet.  The hatch and camera were recovered.  The engine blew out in turns but re-lighted in climb-back.


01 JUL 68



The engine did not light; vehicle nosed over and was lost.


28 AUG 68



Carried two birds from Area 51 to Kauai, Hawaii.  D-21 #516 was put into a perfect launch position, but the  Marquardt MA20S-4 engine did not light and the bird was lost.


15 DEC 68



Hatch and camera recovered; photos okay.


11 FEB 69



Loss thought to have been caused by water contamination in the autopilot.


10 MAY 69



Hatch and camera recovered; photos fair.


10 JUL 69



Drone flew the "Captain Hook" route extremely well.  The program had now met all of the design requirements and objectives to the point where the Air Force deemed the program successful and complete up to the operational phase.  Hatch and camera recovered; photos good.


09 NOV 69



First operational mission: Crashed in the former USSR.  Subsequent to this failure, Lockheed re-addressed the navigation system and changed the programming to enable the drone to miss one destination checkpoint but continue on to a following one.


20 FEB 70



Ran another "Captain Hook" mission with the new navigation programming, flying almost 3,000 nm.  The D-21 performed superbly, reaching an altitude of over 95,000 feet.  It met all of the checkpoints within 2-3 nm.  Hatch and camera recovered; photos good.


16 DEC 70



Second operational mission: Hatch lost due to parachute failure.


04 MAR 71



Third operational mission: launched bird on a critical mission.  It returned after a fine flight.  The parachute was damaged by the hatch on descent and fell slowly into the water.  The hatch floated and the Navy recovery ship arrived.  During the recovery attempt, the hatch was run over by the Navy ship; hatch was damaged and sank.  Another Navy ship found the D-21 afloat but was unable to get cables around it before it also sank.


20 MAR 71



Fourth operational mission: Drone shot down three-quarters of the way through the mission to over-fly the Chinese nuclear weapons test facility near Lop Nor, China.



It is believed that all four operational launches targeted the People’s Republic of China nuclear weapons test facility in remote west central China near Lop Nor.  It so happened that months prior to the very first launch, Air Force Lt.  Colonel Reed (who was responsible for payload recovery) was tasked to Pearl to brief the Commander of the Pacific Fleet (CINCPAC) as to their responsibilities.  This was in the event the camera pallet landed in the Pacific. The Navy Admiral’s response to Reed’s visit was: “Just what could the Air Force tell the Navy that it doesn’t already know?”  With that, Reed went on his way. Of the operational missions, the first D-21B (#517, launched on November 9, 1969) did not successfully institute the return maneuver to take it back to the recovery area.  Instead, the D-21B continued on a straight course and crashed somewhere in the wilderness of the former Soviet Union.  The cause for this loss was later simulated by Honeywell and was determined to have been an error buildup in computation of sign/co-sign routines in the navigation system.

Interestingly, after the fall of the Soviet Union, Ben R. Rich (then retired president of Lockheed’s Skunk Works) finally had an opportunity to tour Russia himself.  While in Moscow, the KGB presented Rich with a gift of what they thought were the remains of a stealth fighter that had crashed in their territory.  As it turned out, the wreckage was actually pieces and parts of the lost D-21B drone.

The second operational mission (launched on December 16, 1970) was flown by D-21B #523.  It completed its 2,648 nm trip only to have the payload lost at sea due to a partially failed parachute.

The third Senior Bowl operational mission was launched on March 4, 1971.  D-21B #526 flew the complete mission profile of 2,935 nm, the JC-130B missed the aerial recovery due to damaged parachute lines, and the payload landed safely in the Pacific.  Once in the water, it was the Navy’s task to recover the package.  On this particular mission, the Navy recovery ship failed to pick up the package on the first pass, so a Navy SEAL team was put into the water to assist in the recovery.  Once in the water, the SEALs could not cut through the parachute cables since the cords had been reinforced with stainless steel wire.  The Navy ship tried again, following its typical recovery procedure of approaching the pallet from the windward side and then drifting towards the package.  The “fly in the ointment” was that the parachute (still attached) was now acting like a sea anchor.  When the recovery ship came up next to the pallet, the ship ended up drifting right over it – in a sense, doing a “keel hole” of the package.  The net result of this misfortune was that the only recoverable camera images from an operational mission were sunk at sea.

D-21B #527 launched on March 20, 1971 and was the fourth and final mission.  Experts at the 4200th Support Squadron and at Lockheed’s Skunk Works concluded that #527 must have malfunctioned.  It was thought to have gone down near Lop Nor.

For each D-21B delivered from the Skunk Works to the 4200th, the initial program cost (from mission through destruction) was estimated at 5.5 million in 1970 dollars.  This price included the rocket, mission evaluation, operation of the B-52H, and all other program-incurred “per-mission” expenses.


The Lockheed D-21B and modified Boeing B-52H combination was the nucleus of Senior Bowl operations.  Senior Bowl and the”A” Flight/4200th program lasted only three years.  It was a very difficult operation, utilizing the best state-of-the-art technology of that day.  Only one thing had to go wrong to jeopardize the entire mission, and one problem tended to lead to others.  The operation was extremely expensive and that aspect alone undermined the project over time.  Overall, however, the successes were overwhelmed by less-than-stellar or totally unacceptable outcomes.  Johnson and his Skunk Works geniuses did everything possible to make the program a success, as did the members of "A" Flight/4200th.

The entire “A” Flight/4200th team was expertly trained and highly motivated, and each and every member of the squadron was fully capable of performing his or her duties properly.  “A”Flight/4200th became a close-knit, highly-effective team that happened to be assigned an extremely challenging project with obstacles that proved to be insurmountable for various reasons.  Three months prior to the termination of the project, many of the troops (especially those with overseas-imbalance skills or critical skills) received orders for Vietnam.

Lockheed’s challenge has been described as “… an overhead mission with airborne technology that just didn’t run long enough to make it.”  With better financing and more time to work the problems, the outcome of the project could have been markedly different.  The Skunk Works had technology for air-breathing vehicles, but they were attempting to do a satellite-type mission and the implementation was flawed.  They overcame many difficult hurdles (fuel sealants, aerodynamics, matching engines to nacelles, etc.), but when it came to electrical or electronic issues, the experience just wasn’t there.  Even though Johnson stressed the importance of the reliability of electronics, Lockheed’s success in that arena was lacking.

The story goes that Johnson had a long-standing mistrust of anything electrical.  Jim Eastham, Lockheed’s Chief Pilot on the YF-12, once said: “If Johnson could have invented a hydraulic radio, he would have.”  Early in the development of the Tagboard program, Johnson had a meeting with Honeywell’s D-21/Tagboard engineering team in his Burbank office.  He had some questions about the reliability of the autopilot and navigation systems, which the Honeywell group had presented to him in a box.  Suddenly he shocked everyone by throwing it on the floor, demanding: “Okay, can it take that?  It has to take that!”  His point was that it might meet its specs, but it also had to be invulnerable to abuse.

During its operational lifetime, Senior Bowl was one of the best-kept secrets in the military.  During that time, “A” Flight and the 4200th Support Squadron launched 16 D-21B vehicles from the B-52H, both in training and operational sorties.  Only a very few select personnel were cleared for the project.  Each team member had to be cleared to top secret, and in addition required a Special Access Required (SAR) clearance above their already-established top secret level.  This unique level of clearance was (and still is) classified information.  All Senior Bowl personnel had to be cleared to the same level, regardless of rank or grade.

When the program began, everyone involved had to sign all kinds of security paperwork (jokingly referred to as “burn before reading” material) promising, in effect, that they would never mention the term D-21B nor discuss the Senior Bowl program with anyone not equally cleared.

Senior Bowl lasted from January 1968 until its abrupt termination on July 23, 1971.   The D-21B program had come to an unceremonious end.  The drones were moved to the test site and from there they went to Norton.  It was a while before they moved them to DM.

Like the U-2, SR-71, F-117 and other formerly “black” operations, many more details about Senior Bowl may someday be made available to the public.   Until that time, this report summarizes most of the information available to date about this unique operation at the Skunk Works – a program they sometimes fondly refer to as their “twice-unsuccessful success.”


Authors: Jim Goodall and Jay Miller

Editor's Note:

 Thu, 18 Jul 2002 23:05 The following email and information was provided by Harlan L. Gurney, Col., USAF, (Ret); one of the D-21B hatch recovery test pilots. Although he was not  involved in the operational phase of Senior Bowl, his recollections are an important historical addition to this web page and is posted here to preserve that segment of Senior Bowl history.

The D-21B Payload Aerial Recovery System

"Project Quick Job"


Harlan L. Gurney

I was one of the recovery test pilots in the recovery developmental test unit at Edwards AFB (AF Satellite Control Facility, Operating Location #1) that was involved with proving that the "hatch" could be safely boarded on the C-130 after aerial recovery, and to perform development testing on the parachute for the "hatch" recovery under the name "Project Quick Job". I then was reassigned to the operational aerial recovery unit in Hawaii in 1965.

I flew in support of each of the four "Tagboard" missions, although I had never heard that name used and was the pilot of one of the two Hickam-based aerial recovery C-130 aircraft, staging from Edwards AFB, in the mission area off the California coast for the 30 July 1966 M-21/D-21 launch mishap. The two HC-130H aerial recovery aircraft on the mission were from the 6593rd Test Squadron, 6594th Recovery Control Group, Hickam AFB. The aircraft serial numbers were 64-14854 and 64-14858 flown by Major Ed Bayer (Ozzie 54) and myself (then a USAF Captain ) respectively. Because of that mishap, we returned to Hickam the following day, which involved my aircraft (Ozzie58) in an event whose aftermath has created a unique display in the new NRO visitor's center, which opened yesterday, 17 July 2002, in Chantilly VA..

Your article is generally excellent but errs greatly in its description of the C-130 aircraft and aerial recovery technique. At that time, the regular satellite recovery aircraft were JC-130B models. However, we acquired three specially modified new long-range HC-130H aircraft equipped with JC-130B type All American Engineering recovery gear, specifically to support the D-21 mission. Our aircraft did not have "whiskers", but rather used the same aerial recovery gear and technique as was employed for film capsule recoveries for the Corona Project (as displayed in the Corona exhibit at the NASM of the Smithsonian Institution in Washington D.C.) See Curtis Peebles book "The Corona Project" or several other books written about the aerial recovery of film packages from earth orbiting satellites. It WAS necessary, however, to develop a unique parachute configuration suitable for the weight of the O-21 "hatch".

       HC-130H  #526 recovering a Corona Project test parachute 

 (Photo courtesy Colonel Gurney)

Sun, 18 Aug 2002 21:38 While still at Edwards I participated in the early trials to figure out how we could rig/tether the "Hatch" so that it could be brought aboard safely. Its shape could result in a lift coefficient (potentially disastrous winching it aboard in the turbulence behind an open C-130 ramp. The task was to make sure that the "Hatch" did not flutter around & damage the stabilizer, control hydraulics, or fuselage as it was winched under the empennage and up to the ramp/dolly fairlead. By making repeated ground station pickups of the "Hatch" Captain Bob Counts and our enlisted recovery technicians came up with a rigging configuration that was stable enough for safe boarding. I went PCS to Hickam from Edwards in 1965, and participated in the Tagboard and early Senior Bowl tests, as previously mentioned. I will be seeing and talking with Don Hard, Bob Counts, and former enlisted rigger Chuck Dorigan at our Test Squadron reunion next month.

Editors Note: More to come...

Editors Note:  

In an email received from Roger Martin, He adds his observations of activities of the 4200th during his tour of service with the 744th Bomb Squadron at Beale AFB:

( Reprinted by Permission )

Fri, 5 Jul 2002 10:26

I was assigned to Beale in the 744th Bomb Squadron during the time the D-21 was there. I noticed one thing in particular I have to disagree with. We would taxi out on our missions, timed to cover for the missions flown by the Hs. I specifically recall seeing the "Drones," as all the signage around the base referred to them, on both wings of the B-52--not only on the right side. Furthermore, while the test aircraft might have been stripped of the gunner's position, I recall watching the 20mm being loaded into the gun. So, the operational birds must have had gun functionality, even if the test bird(s) didn't. The demise of the unit closely coincided chronologically with Nixon's trip to China, so I don't know what part, if any, our cozying up to the Chicoms had to do with the shutdown. At that time, we picked up some of the aircrews from the Hs. They were quite silent about what they had been doing, and we didn't pester them much. We did get from them that their houses had been bugged to ensure the crews didn't divulge anything to friends and family over drinks or whatever, and they even pointed out the wooden decoration on the wall where the bugs had been least that was what I was told by my neighbor and flying buddy.

Roger Martin

There is an Image Archive of the "Tagboard" and "Senior Bowl" operations. It consists of 50 photos and includes the construction of the D-21B pylon for the B-52; testing of the Marquardt Engine at Area 51; assembly construction of the first D-21 #501 and all known photos of the mated B-52H Stratofortress and the Lockheed D-21B drone. Here is the URL for the "Senior Bowl" image archive:

This article is copyrighted by James Goodall and Jay Miller. Copyrighted on April 19, 2001. All Rights Reserved. Article may not be reproduced in any format or stored on any retrieval system without permission of the authors. Revised September 25, 2002. Permission to post this article to the "SR-71 Blackbirds" website has been granted by the authors. 

Credit is given to Glenn Chapman, Jerry Miller ( Marquardt Technical Representative, D-21 ramjet engine) , Jim Walborn, Bill Fox and many others that asked to remain behind the scenes. Tagboard B-52 Artwork is from "Tony Landis Collection"; Photo of Lop Nor, China courtesy DOD Corona; Lockheed Martin Corporation (D-21 Photos). HC-130H photo courtesy Harlan Gurney. Web page HTML; D-21/Tagboard Summary table formatting and page editing by Leland Haynes, Webmaster, " SR-71 Blackbirds"

Take Me to the D-21B "Senior Bowl" Image Archive



 SR-71 Front Page   Links Page   Index Page   Recollections   2001 Reunion  
"SR-71 Blackbirds"  Web Site Navigator
First Created: April 15, 1996 - Last Revised: May 12, 2010
Copyright © 1996 Leland R. Haynes     Email:

Page #62 of the "SR-71 Blackbirds"