Finally, about 10 A.M. on December 4 the crew managed to launch the first A-3, patriotically named “Deutschland.” For the first three seconds the rocket ascended vertically, then suddenly the parachute popped out of the side, trailed behind the still accelerating vehicle, and was incinerated. The rocket turned into the wind, and the engine shut off automatically when it tipped over too far. After about twenty seconds it crashed back onto the island only about 300 meters from the launch site, exploding violently on impact. According to Dornberger: “Eyewitness accounts were wildly contradictory. Everyone claimed to have seen something different. We decided to venture on a second launching.” When the second A-3 was sent on its way two days later, virtually the same thing happened, with the vehicle crashing only 5 meters offshore. Now that it was clear that the parachute had been deployed, it was only natural to blame the powder charge that pushed it out. The parachute was omitted for the third launch on December 8, and a signal flare was put in its place. The wind was stronger than on earlier attempts, and the rocket turned quickly into it, ejecting the flare after four seconds. Again the engine cut out automatically and the rocket crashed 2 kilometers out to sea. The last attempt on December 11 was almost identical.59
Those results were shocking and discouraging, but already during the many interminable delays on the Oie, Dornberger, von Braun, and the chief engineers threw themselves energetically into explaining the failures. Attention focused initially on the possibility of a static electricity buildup on the skin of the rocket, setting off the parachute charge. But ground tests conducted later in December indicated that that was definitely not the case. The fact that the A-3 tended to turn into the wind rather than stay on a vertical course also implied that the control system was too weak. The rocket appeared to be excessively stable; the fins had apparently moved the center of pressure so far back that the jet vanes lacked the power to fight back against the aerodynamic forces. The servomotors that moved the vanes also seemed to lack sufficient power, a consequence of the undeveloped state of this technology.60
While it was in fact true that the control system was too weak and the rocket too stable, those problems did not explain the ejection of the parachute or flare every time. Only after review of the launch films, repeated ground tests, and meetings with Kreiselgeräte did it become clear in January 1938 what had gone wrong. The Achilles heel of the Sg 33 guidance and control system was its inability to stop a rapid rolling of the A-3. For reasons of simplification, the stable platform had no ability to turn around the vertical axis and no roll gyro to sense whether the rocket was moving in that axis. If the vehicle rolled at a rate of more than six degrees per second, the forces acting on the platform gyros would quickly overwhelm their ability to compensate for the precession induced by the rolling. When one of the gyros hit the end of its allowed range of motion (30 degrees), it would lurch back, and the platform would tumble over, losing its ability to control the vehicle. The circuitry for letting out the parachute had been linked to the platform on the assumption that at the peak of the trajectory the rocket would turn over, upsetting the platform. But the fundamental flaw was that the control forces exerted by the jet vanes, on command of the rate gyro for the roll axis, were far too weak. Assymmetries in the fins, in conjunction with wind, would be enough to start a roll that overpowered the control system. In every case this had happened so fast that the platform toppled in the first three or four seconds.61
Because of a lack of experience, no one in Ordnance or Kreiselgeräte had seen this coming. Kreiselgeräte specialized in heavy naval systems, and the engineers in Army Ordnance had been completely dependent on the company and on Boykow’s original design. Thus von Braun came to rue his uncritical enthusiasm for the late inventor. It was clear that much more effort and resources had to be put into guidance and control and that competing companies had to be pulled into the program. Dornberger and his subordinates saw as well that it was a mistake after the A-2s to conclude that frequent launches were unnecessary. A new vehicle, called the A-5 (since the A-4 designation had already been assigned), would have to be built to test guidance systems systematically in the air, rather than only with the burning rocket on a test stand, as had been done at Kummersdorf. The excessive stability of the A-3 also confirmed earlier impressions that the wind tunnel testing at Aachen had been far too limited to give an adequate understanding of the forces acting on a flying rocket. Only the engine system of the A-3 had worked without a hitch. But the 25-ton-thrust A-4 engine was a huge step that required a massive infusion of resources and more systematic work. That at least had begun with Thiel’s transfer from Schumann’s research section and the construction of Peenemünde.62
The failure of the A-3s thus confirmed and strengthened the trend that had begun in 1936. If breakthroughs in the key technologies were needed to build something as revolutionary as a ballistic missile, the rocket program would have to spend much more money and build much more in-house expertise. But the A-3s were the epitome of what von Braun later called “successful failures” in the rocket business. So much had been learned from this experience that, given the highly favorable political and budget climate of the late 1930s, the technical obstacles could almost certainly be overcome. In the years between 1936 and 1941 the Army Ordnance group would do precisely that.
Chapter 3
Breakthrough in Key Technologies
Notwithstanding the important advances the Army group had made in the A-3 and rocket aircraft programs, the technological challenge of the A-4 remained gigantic. The engine would have to be seventeen times more powerful than the largest rocket motor so far constructed; the missile would have to fly at nearly five times the speed of sound when no Ordnance rocket had even approached the sound barrier; and the vehicle would have to be guided to targets nearly 300 kilometers away, when no liquid-fuel rocket built by the Germans had ever traveled more than a few thousand meters vertically. The A-3 failures only underlined how far away the engineers were from solving the guidance and control problem in particular. Yet by late 1941 Peenemünde had in its possession the technologies essential to the success of the A-4, and the first versions of that rocket were on the test stand.
The foundation for that remarkable technological achievement was Ordnance’s ability to mobilize money, manpower, and matériel for the ballistic missile project—something it was able to do because of the high priority placed on rocketry by the Army High Command. Access to resources alone, however, did not automatically lead to the dramatic breakthroughs necessary for the A-4. Under the leadership of Becker, Dornberger, and von Braun, the liquid-fuel program had to expand its engineering staff greatly, put innovative leaders at the head of critical projects, and gain control over additional research capability in universities and corporations. The research process itself had to be altered so that trial-and-error testing was replaced, where possible, with a more scientific and theoretical approach, although that became apparent only over time. The result, especially after the A-3 guidance failures, was to accelerate further the growth of the large government laboratory at the heart of Peenemünde-East. At the beginning of 1938 the facility had 411 employees. By September 1939 that number had tripled.1 Although no figures are available for late 1941, the number of people in development (as opposed to the new A-4 Production Plant) must have nearly tripled again to at least three thousand engineers, craftsmen, and office workers. With that vastly expanded staff came a corresponding increase in the facilities and materials available for research and testing.