In the late 50's , a need for another Recon platform to augment the U-2 was recognized. After Gary Powers was shot down over Russia on 5-1-1960, President Eisenhower stopped all overflights, and the need for another capability became more urgent.
Designs for a supersonic aircraft were proposed by Lockheed and Convair. Lockheed won. Our design studies were numbered from A-1 to A-12. "A" was for Archangel. The contract was awarded in late 1959, and the first flight was on 4-26-1962, 30 months later. That accomplishment was even more amazing given the fact that almost everything on the aircraft had to be invented.
The design studies were done by a core group of 14 engineering specialists, including Clarence L. (Kelly) Johnson. All calculations were done by hand, with slide rules and Friden mechanical calculators. We had a small analog computer to run flight conditions for flight control system design. I produced 3 or 4 pages of hand calculations to set up one flight condition.
Once the contract was awarded, the real work began. The structure was to be almost completely titanium in order to withstand extreme heat at Mach 3.2 and still be strong and light enough to carry the fuel and payload for extended time at cruise Mach. This had never been done. Temperatures ranged from stagnation: 800 ºF, surface: 400 ºF, internal: 350 ºF, and engine nacelles: 1000 to 1200 ºF. Some of the things that had to be invented to operate in the high temperature environment are: Hydraulic Fluid, Fuel, Control Cables, Control Servos, Electrical Wiring, Tires, Fuel Tank Sealant, Tank Inerting System, Lubricating Oil/Grease, Many Electric and Electronic Systems and of course the entire structure. The list is endless.
The fuel had to be a low aromatic to avoid ignition in empty tanks during descent from high altitude. A Nitrogen tank inerting system was included, using LN2 and a converter. The fuel, PWA 523A - later JP 7, was so inert that a lighted match could be thrown in a puddle of fuel and it would just go out. This is a good thing, because we never did manage to completely seal the tanks, and fuel leaks on the ground were common.
The remarkable accomplishment of producing this aircraft can be attributed to Kelly Johnson's engineering and management genius. He gathered a group of very talented engineers and gave them the responsibility and authority to do a job. Every design engineer followed his item through its entire life cycle - procurement, vendor coordination, manufacturing, ground test, flight test and operational support. As a Flight Test Engineer, I knew and worked closely with all of the design engineers.
Kelly insisted on minimum staffing on the customer and contractor sides. Security was a primary requirement. Our security on the A-12 rivaled the Manhattan Project. On the Government side, the President, a couple of Congressmen, a few CIA people and a couple of USAF people were on board. The CIA was the customer, and the only CIA person we saw at Lockheed was Mr. Norm Nelson, who later joined Lockheed and became the chief of the Skunk Works in the early 80's.
In February of 1961, I was encouraged by a friend to talk to the Skunk Works Flight Test people about job opportunities. I was interviewed by Hank Stockham, whom I had met socially in Palmdale. We sat in my car in the parking lot off Hollywood Way because they could not let uncleared personnel inside the gate. Hank told me that my experience qualified me for a job in their organization. He could not tell me what I would be working on. It could have been a space ship or a submarine, but I suspected it was an airplane since they wanted me - an aircraft FTE with 9 years experience. I was hired, and my assignment was Senior Flight Test Engineer assigned to the first A-12 aircraft.
Design and manufacturing of the first aircraft, S/N 121, continued through 1961. It was transported by truck to the Nevada test site, Area 51, Groom Lake, in February of 1962.
There was a lot of work to do in those first weeks. Pratt & Whitney was behind schedule with the J-58 engine, so we adapted the nacelles to accept J-75's. When the aircraft was fueled for the first time, fortunately outside the hangar, it literally leaked like a sieve. The tanks had to be totally stripped and resealed. The tanks are integral - no bladders. The structure inside is incredibly complex and therefore difficult to seal. The sealant has to survive the extreme high temperature environment, and development of a new material was difficult. We needed to do engine runs, and the tanks were not ready. Kelly said "Stick a couple of F-80 drop tanks above the wheel wells and get on with the engine runs". We did. We also did the first taxi tests with the "Kelly Tanks" installed. Many other development problems were encountered and handled.
In this time period, as Senior FTE, I was not really in charge. Kelly ran the show. Next in line was the Chief of Flt Test (Larry Bohanan), then the Flt Test Group Engineer (Glenn Fulkerson), then me.
On April 25, 1962, Lou Schalk, the first pilot on the program, took 121 out for the " High Speed Taxi Test", the normal prelude to first flight. The intention was to get to takeoff speed, raise the nose wheel, then put it down and roll out on the runway. Whether by accident or intent ( Lou was a free spirit), the aircraft lifted off. The aircraft was not prepared for flight. The stability augmentation system was left off, by intent, and the fuel load was light and aft. When he got airborne, Lou had his hands full getting the bird under control. He put it down as soon as possible and rolled out on the lake.
We discussed what to call this operation - Flt 1 or a taxi test. Conclusion: it was a taxi test, and in the Flight Test group, we called it the "Broad Jump".
The first flight was on April 26, 1962. It was short, gear down and successful. We lost a few chine fillets due to a latch failure. It looked dramatic, but was not serious - secondary structure. Kelly pushed his designers to save weight (not unusual). If something broke, we beefed it up. There was no compromise on primary structure; it was generally redundant.
The flight test program progressed quite rapidly, finding and fixing problems. The J-58 engines came along eventually. First flights with the J-58 were with a J-75 in one nacelle and a J-58 in the other. Subsequent aircraft were delivered about every 3 months. The second article, S/N 122, was initially used for radar cross-section measurements. Instead of the usual mock-ups, we used an actual aircraft for RCS measurements. It was supported on a pole about 40 feet from the ground. 122 was thereafter known as the "Polecat".
A serious performance problem was encountered. The inlets were not handling the shock waves as intended. One of our thermo engineers plotted inlet area vs. longitudinal station. The plot revealed a step discontinuity around the middle of the spike. It was a design error not caught in wind tunnel tests. Kelly solved it by adding "mice" inside the inlet, which were fairings shaped to produce a smooth airflow in the duct. This fix stayed with all the Blackbirds - A-12, YF-12 and SR-71. A properly shaped and more efficient inlet was proposed for the F-12B, which was not built.
The inlet control system didn't work well. It was a pneumatic system provided by Hamilton Standard. We went to an electronic system designed by AirResearch and moved on.
The J-58 engine could not use a traditional jet engine ignition system because of the high temperature high altitude environment, so a chemical system was used. A fluid called Triethylborane (TEB) was injected in the burners. TEB is pyrophoric, which means that it ignites in the presence of oxygen. It was initially called the Pyrophoric Ignition System (P.I.S.). When someone noticed the initials printed, the name was changed to "Chemical Ignition System" (C.I.S.), which doesn't spell anything.
The engine inlet is a complex system, requiring cosiderable development effort. It is a symetrical inlet (axisymetric) with a translating spike to position the shock waves. The incoming air goes through a series of shock waves to reduce it to subsonic speed. The air is diverted overboard and internally in a manner to produce efficient engine performance. The J-58 engine is "high bypass ratio" or "bleed bypass", meaning that a lot of the compressed air is diverted around the core engine and into the afterburner, making the A/B operate as a ramjet. An analysis of pressures in the nacelle revealed that the forward thrust on the airframe is actually exerted in the inlet structure, the back of the spike, at cruise Mach. Strain gages on the engine mounts would show the engine as a drag item.
Working on the A-12, and subsequent programs, at Area 51 required tight security. The Lockheed people flew to the test site on Monday morning from Burbank. We had three Lockheed Constellation airliners to transport people and parts. Typically, we worked 5 or 6 days and then flew home. I remember several occasions when I stayed at the test site for two to three weeks without a break. We worked 50 to 80 hour weeks. We were trained to avoid identification with Lockheed, and we never discussed work with family or friends.
In the Fall of 1962, the union workers at Lockheed went on strike. All the shop and flight line guys at the test site were obligated to stay home. For about two weeks we kept the test program going, and flew several flights on 121. Managers, engineers and test pilots did the work. The pilots, Lou Schalk and Bill Park, did the pre and post flight inspections. We were in the process of expanding the flight envelope, moving higher into supersonic speeds. There was a need for a camera fairing to be mounted externally. We normally would have made it out of metal, but we had no metal workers. Lou Schalk and John Wallis knew how to work with wood, so we made a beautiful faring. It flew on many subsequent flights, until 121 was getting over Mach 2.0 and the wooden creation came back burned black. We also used silver duct tape to hold instrumentation wiring and other things to the aircraft exterior temporarily. We called it Mach 2 tape because it would burn off above that speed.
In June of 1964, Lou Schalk and John Wallis left the program. John took a sabbatical in Europe and Lou went to another job.
In September of 1965, John returned to Flight Test at Area 51 on the MD-21 program. We had four FTE's: Keith Beswick, Ray Torick, Bill Sass and John Wallis. Keith and Ray flew as Launch Control Officers (LCO). Larry Bohanan was the boss.
Meanwhile, the YF-12A and SR-71 programs were up and running, and the A-12 was going operational. As an aside, when the A-12 operational aircraft were delivered to Area 51, their home was a set of new hangars built south of the main base. It was a new area and no one knew what to call it. People said " I'm going to the south end" or "to the other place". John Wallis called it "Baja 51", and that part of the test base is known as "Baja" to this day.
We launched the first D-21on 3/5/1966. Launch was good. The D took a nose dive. The launch scenario involved a "Kamikazi" maneuver to get transonic, then starting the D engine at about 1.4 Mach and accelerating to 3.2 Mach. At this point, the M-D was a real tiger. With two J-58's and the ramjet burning, the pilot had to throttle back to Min A/B to avoid exceeding 3.2 Mach.
Before takeoff with a Blackbird, the pilot runs the engines up to full power, one at a time, to adjust engine trim. We did it one at a time because with both J-58's at full power, the brakes and chocks would not hold the aircraft. I noticed that when #135 ran up the left engine, a little moisture appeared on the runway. I let them take off anyway. This happened one more time and I said "stop!". I put the airplane in the hangar, got the fuel tanks drained and purged and I personally climbed into the fuel tank at the F.S. 715 joint to see what was there. I was one of very few engineers slim and athletic enough to climb into a fuel tank through the "boy hole". The Blackbird forward fuselage is connected to the rest of the aircraft with four large steel bolts. I checked these bolts and the one at 9 o'clock came out in my hand. It had broken. It was as I suspected. Fuselage bending on engine run-up had created a fuel leak. The bolt was defective.
The early launches were done with a pushover to negative "g" in order to give the drone a separation boost. With each successive launch ( there were only four) Kelly reduced the negative "g" target. The last one was done at one "g" - level flight. This concerned me at the time, but I was just a lowly flight test engineer.
On 30 July 1966, my airplane, #135, with D #504, performed a launch mission. Pilot was Bill Park and LCO was Ray Torick. The launch flight went as planned, including the one "g" condition. The D unstarted as it passed through the main shock wave of the M. This was normal, but it was just a transient event. In this case, the effect was strong enough to cause the D to pitch down and impact the mother ship. The combination broke up ( at Mach 3.2) and Bill and Ray ejected. Bill hit the ocean, dumped his chute, climbed in his raft and survived. Ray apparently had a broken arm, was not able to get into the raft, opened his faceplate and drowned. We were all devastated. I was particularly upset because I had disagreed with the one "g" launch condition. Kelly cancelled the program. For this launch, for the first time, the other M-21, S/N 134, flew chase with Art Peterson and Keith Beswick on board. Keith got some remarkable movies of the accident. This may have been the first time two aircraft flew formation at 3.2 Mach.
We immediately reinvented the program. Now we would launch the D-21 from an underwing pylon on a B-52. Everyone from the test site moved to Burbank. Keith moved on to other things. Bill and I stayed with the D program.
It was amazing to see how fast Kelly's organization could move. In just a few months, we had a new system concept and design. Bill and I helped develop the Launch Control Officer's station in the B-52. We wrote the preliminary test plans. Larry brought Jim Billo on board to manage the field operation. The three of us and our Air Force friends from the "A" program, Hal Rupard and Jack Reed, went to Castle AFB in Modesto for B-52 school. It was great fun. We were incognito, of course. Not identified as Lockheed, but as government GS-18 types, at least military General rank. The people at Castle were very helpful, to say the least.
The two BUFF's were modified at Palmdale. That's the B-52's. It stands for "Big Ugly Fat F-----r". Two LCO stations were installed in the aft-facing upper crew stations. It took a while to get used to the fact that the "Left LCO" station was on the right when going aft to sit down.
The first BUFF, 0021, was test flown May 30, 1967. In August of 1967, we flew with a D-21 mated to the B-52. This was only a year after the accident that ended the first program.
The test program continued with many development test flights, including an excursion to Orlando, Fla, to test the air conditioning system in high humidity conditions. We got the humidity, and the system iced up and quit, as we feared.
While the D-21 was mated to the B-52, system checks were done to verify the D-21's readiness for launch and free flight. The primary test system was an electronic automated system known as IFCO - In Flight Check-Out. Group 5 IFCO checked the control surface responses. This required the elevons and rudder to move, therefore putting aerodynamic loads on the D-21 and pylon while mated. I had problems with Group 5, so decided to run it early in the launch rehearsal test flight of 28 Sept 1967. Normally it would be run last in the test sequence, which would put us over Long Beach. While preparing for flight, the crew had great difficulty installing one of the three bolts that secured the D-21 and Booster to the B-52 Pylon. The crew manager and the engineering manager jointly decided to run a tap through the problem nut in the pylon. This, of course, removes material from the threads and weakens the connection. I had no knowledge of this action. I was preparing to fly the mission and was not consulted or informed. Shortly after takeoff, on climb-out, halfway from Groom Lake to Currant VOR, I ran Group 5 IFCO. After a few cycles of the D-21's control surfaces, it fell off the pylon. Separation caused automatic booster ignition after 5 seconds. The D (#501) and booster spiraled out of contol to the ground. Boom!
It is interesting to speculate what might have happened if I ran Group 5 in the normal sequence, when we would have been over Long Beach, on our way to the over-ocean launch point. The mounting bolt was compromised, but we would have been at higher altitude and lower Q. Would the bolt have pulled out of the weakened nut?
The USAF Security people were immediately dispatched to the crash site. Upon arrival they found Mom and Pop and their pick-up camper. The people were wandering around the site taking photos. Security got the folks clear of the debris, confiscated the camera, and at that point, the destruct package in the D-21 blew. It cooked off in the post-crash fire.
Shortly thereafter the Flight Test Manager was transferred and John Wallis was promoted to Project Manager and Senior LCO.
We continued with the development test program. The second BUFF (0036) came to the test site. We flew mated flights with both B-52's. We started flying with two D-21's on the B-52. The plan was to carry two mission - capable D's and launch the secondary if the primary failed IFCO.
We brought the Air Force people on board and trained the maintenance and flight crews to operate the system. The LCO flight manual was written from my in-flight notes, and I trained the USAF LCO's. We carried two birds on most flights. I was on the B-52 for every Flight Test launch, except one. On that occasion I went to Hawaii to fly on the tracking aircraft. I launched six birds and supervised the rest
On one occasion , the USAF Col. in charge of the program (Frank Hartley) said that we have to have a very good reliability record for the program to continue. I told him to forget it. This system is too complex and too new to expect such reliability.
We moved the program from Area 51 to Beale AFB in early 1969. We were the only test program on the base, so it went dormant for a while.
We continued development testing at Beale. All of the launches were made to the west of Kauai using the Barking Sands tracking facility. Lots of 12+ hour missions were flown.
After the successful mission on July 10, 1969, I left the program and joined the L-1011 Flight Test group. The operational phase of the D-21 program was not successful. It was terminated after 5 flights.
The 1011 program encountered a delay due to Rolls Royce engine problems. The Skunk Works called me back, which pleased me greatly.
In 1971, I was called to Burbank to replace the SR-71 Assistant Program Manager. This was a major promotion. I stayed in that position until 1979, when I joined the Stealth Fighter (F-117) program as the Flight Test field manager.
I was back at Groom Lake. We refurbished the old A-12 hangars and the rest of the base, which had been dormant since we moved the D-21 program to Beale in 1969.