I've looked for better explanation, and all I can get is:

Stage separation occurred as planned, but because residual fuel in the new Merlin 1C engine evaporated and provided transient thrust, the first stage recontacted the second stage, preventing successful completion of the mission

- https://en.wikipedia.org/wiki/List_of_Falcon_1_launches#Third_flight

However, why couldn't the second stage still get to orbit?

Did the 'recontact':

  • Damage the second stage so much that it failed to light or was unsafe to light?
  • Knock its attitude so far that it couldn't make corrections? What would prevent the second stage reorient itself?
  • $\begingroup$ In plain language, the stages collided and the upper stage spun out of control. $\endgroup$ Jan 14, 2016 at 4:07
  • $\begingroup$ Does the second stage not have attitude control through thrusters and gimbals? Do you have a source on it having spun out of control? $\endgroup$
    – Ehryk
    Jan 14, 2016 at 16:59

1 Answer 1


According to Elon Musk, the problem was in a recontact between the stages. From a SpaceRef's SpaceX Determines Cause of Falcon 1 Launch Failure article after the media teleconference, Elon Musk was quoted:

We have a definitive understanding of what went wrong on Flight 3. The problem was due to a design error not a production or quality assurance issue. The thrust transient was longer than it was for the prior flight. The previous flight had an ablatively cooled engine. Flight 3 had a regeneratively cooled engine. The gap between engine cut off and staging was 1.5 seconds - which was fine for the ablatively cooled engine on Flight 2. But on Flight 3, with the regeneratively cooled engine, there was some residual thrust after engine shut down and this caused the first stage to be pushed back toward the second stage after separation and there was a recontact between the stages.

Note that Musk talks about a recontact, not a bump. I.e. the separation failed with the inverse of it, a recontact of the two stages. Here's a video of the relevant part of the SpaceX Falcon 1 launch 3 from the Omelek Island, the Kwajalein Atoll:

Last couple of frames just before the video jumps to fairing separation show upper stage engine ignition while the first stage was again in contact with it, catching up with up till that point inertial upper stage due to the mentioned residual burn. Since the upper stage was not designed for such engine ignition (some launch vehicles do exactly that and while lower stage still burns so the upper stage doesn't require ullage motors, but no US ones as far as I'm aware, and those that are use an open mesh interstage skirt at sufficient separation from the lower stage to allow exhaust gas expansion), this could have only resulted in two scenarios:

  • Recontact wasn't symmetric and exhaust of the upper stage engine ignition with lateral thrust vector forced the upper stage into a violent spin, or
  • upper stage exhaust reflecting upwards into itself and creating a pressure spike, resulting in a hard start, followed by rapid unscheduled disassembly.

Final few seconds of the video during fairing separation suggest the first scenario was the final demise of the upper stage, from which the stage apparently couldn't recover and likely eventually disintegrated as it reentered denser lower altitude atmosphere. There's only that much of translational control that gimballed main engine and RCS thrusters can provide to the stage, which is at that point far below orbital speed, and soon starts falling back towards the surface.

  • $\begingroup$ Could you elaborate on what the difference between a 'bump' and 'recontact' is? Did the first stage 'hook on' to the second stage? Could they not have just tried to have the second stage fire some thrusters to get free first, or was it all pre-programmed? $\endgroup$
    – Ehryk
    Jan 14, 2016 at 4:37
  • $\begingroup$ @Ehryk It appears there was just enough residual burn in the first stage to catch up with separated upper stage and remain inertial with respect to it. I.e. collision wasn't hard enough for them to bounce off and make distance again, but I don't know if they actually reattached (doesn't look like it, there's apparent misalignment). Anyway, these sequences are hard-wired. There was detection of separation and upper stage ignition sequence was initiated from then on. There's no time to waste either, upper stage would at that point be losing forward velocity at ~ 9.4 m/s². $\endgroup$
    – TildalWave
    Jan 14, 2016 at 5:07
  • $\begingroup$ So, it blew itself up with it's own engine ignition, then? Are there any sources that say as such, or this is just what might have happened? It would only be very a short time before it accelerated away from the first stage, wouldn't it? Why wouldn't it move away from the first stage more rapidly than damage could be done? $\endgroup$
    – Ehryk
    Jan 14, 2016 at 16:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.