Falcon 1 Falcon 1 (dev)

Reference sources split the Falcon 1 into two successive configurations, and Gunter's Space Page calls the first of them "Falcon-1 (dev)": the one that flew the opening two attempts, with a Merlin 1A engine on the first stage and a Kestrel 1 on the second, both burning liquid oxygen and kerosene. The advertised capacity of this version was 420 kg to low Earth orbit, below the 470 kg the same source later attributes to the operational configuration. The material consulted documents neither the thrust nor the specific impulse of the Merlin 1A, so no engine performance figures are given here: the difference between the two configurations is described by what the sources state, not by what might be inferred.

The first flight slipped repeatedly because of technical problems with the new vehicle. A scheduling conflict with a Titan IV launch at Vandenberg added to the delays and ended up moving the operation to the Reagan Test Site on Kwajalein Atoll, in the Marshall Islands. The date was first set for 31 October 2005 and let go; it was rescheduled for 25 November and missed again. During the attempt of 19 December 2005 a faulty valve created a vacuum inside the first-stage fuel tank: the walls sank inward and the structure was damaged. The whole stage had to be replaced before returning to the pad.

Lift-off finally came on 24 March 2006 at 22:30 UTC from the SpaceX launch site on Omelek Island — already the morning of Saturday the 25th in local time — carrying the United States Air Force Academy's FalconSAT-2 as a payload contracted by DARPA, a small technology satellite of about 20 kg according to the Encyclopedia Astronautica entry. The flight ended in under a minute. A fuel line leak started a fire in the engine bay; the vehicle showed a noticeable rolling motion from lift-off, pitched over sharply at T+26 seconds and struck a dead reef about 250 feet from the launch point at T+41 seconds. FalconSAT-2 separated from the booster and fell on the island, with damage estimates ranging from slight to significant depending on the source.

The cause was narrowed down to a fuel line nut. SpaceX initially blamed improper tightening; a later DARPA review concluded that the nut had in fact been properly tightened — its locking wire was still in place — and that it had failed through corrosion. The company's answer was procedural rather than a redesign: replacing aluminium hardware with stainless steel and increasing the number of pre-lift-off computer checks by a factor of thirty.

The second flight was planned for January 2007 and slipped because of second-stage problems; September, November and December 2006 had been floated before that. The 19 March attempt was held for forty-five minutes by a data relay issue and then scrubbed one minute and two seconds before launch, when the safety computer read a few milliseconds of hardware delay as a transmission failure. On the 20th a communications problem forced another postponement. On 21 March 2007 the count stopped at 00:05 GMT in the very last second, with the engine already lit, and even so a second try was ordered for the same day: the rocket left the pad at 01:10 GMT carrying the DemoSat payload for DARPA and NASA.

The first stage performed well. The trouble came at separation: the interstage fairing atop the first stage bumped the second-stage engine bell as the nozzle cleared the interstage, because the first stage was rotating far faster than expected — around 2.5 degrees per second against a maximum expected rate of 0.5. Elon Musk stated that the bump did not appear to have damaged the niobium nozzle. The second stage did not complete its job and the flight ended without orbit, a failure Gunter's sums up as a "second stage failure".

SpaceX presented the second flight as a successful test: it claimed to have flight-proven more than 95 per cent of the Falcon 1's systems, and its stated objectives were to exercise responsive launch procedures and gather data. The subsequent diagnosis, vetted by third-party experts, pointed at liquid oxygen slosh in the second-stage tank, correctable with baffles and a change to the control logic, and at the Merlin shutdown transient, to be addressed by initiating cut-off at a much lower thrust level even at some cost to engine reusability.

Those two flights closed the development configuration. The third attempt already flew with the Merlin 1C and the Kestrel 2, the combination that would carry the Falcon 1 to orbit in September 2008.