SpaceX Falcon 9 Launch Profile

By boosting up and back after separation aren't you just putting more energy in that you'll have to counter with additional deceleration later?

Why not just accelerate back, which will use less energy and conserve fuel?

Or am I missing something only a rocket scientist would know?

  • 1
    $\begingroup$ The graphic definitely shows an upward burn (I've seen some older animations showing the stages perform burns in a more horizontal attitude, but maybe these went out of date.) I guess they need a little more vertical velocity to keep out of the atmosphere - but this will depend a great deal on the precise velocity, climb rate and distance they have to return... $\endgroup$
    – Andy
    Commented Nov 21, 2016 at 9:44

2 Answers 2


This graphic was made for one of the first landing attempts, and it shows an unusual scenario: a landing on the droneship where the ship is positioned much closer to the coast than on later missions.

For a normal droneship landing, the stage does two burns: reentry and landing. Its trajectory is roughly a parabola, and doesn't exhibit this extra peak.

For this attempt, they did 3 burns: boostback, reentry and landing. The boostback burn has to kill all forward velocity, then accelerate the rocket on a parabolic trajectory towards the landing site. They did this to test the boostback burn without having the stage land on Cape Canaveral (which was considered too risky for the early landing attempts).

During the launch, the rocket is powered through the entire trajectory. At stage separation, the stage moves at 5000 km/h.
For the return flight to the landing site, the trajectory is an unpowered ballistic trajectory. The boostback burn has to provide enough energy to 'throw' the stage all the way back to the landing site. There are 2 ways to do this:

  1. accelerate the stage to 5000 km/h. This takes a lot of energy.
  2. accelerate the stage to a lower speed, and provide some energy by boosting it to a higher altitude. The width of a ballistic trajectory is determined by the speed of the rocket and its altitude. This takes less energy.

The graphic is also usable for landings back at Cape Canaveral.

  • $\begingroup$ Surely the width of the parabola doesn't matter. Couldn't they just move the landing platform and fall more vertically? $\endgroup$
    – user69911
    Commented Nov 21, 2016 at 12:17
  • 1
    $\begingroup$ I've revised my answer. They could fall more vertically, that's the profile they used in later missions. But that profile is not useful to test the boostback burn. $\endgroup$
    – Hobbes
    Commented Nov 21, 2016 at 12:31
  • 2
    $\begingroup$ It would seem that the diagram found here: spacex.com/sites/spacex/files/16892430560_f87dff78c0_o_1.jpg is a more accurate representation of the usual scenario, at least until they feel happy to land on land. $\endgroup$
    – user69911
    Commented Nov 21, 2016 at 13:14
  • $\begingroup$ I don't believe we have reason to suspect they boosted upwards at all. At staging the rocket already has substantial momentum, including a strong upward component. Thrusting retrograde to just the horizontal component of the booster's trajectory will get the results seen (looping up and around). $\endgroup$
    – Saiboogu
    Commented Sep 14, 2018 at 14:50
  • $\begingroup$ You gain nothing by boosting to higher altitude vs accelerating horizontally to higher speed. Boosting to higher altitude is simply adding gravitational potential energy to the rocket (m * g * h), which is no cheaper than adding adding kinetic energy (1/2 * m * v^2) which would come from horizontal acceleration. My guess, if they in fact boost up to higher altitude and not jut horizontally to higher speed, is that it puts the rocket stage on a better course toward the landing pad/ship. $\endgroup$
    – user39728
    Commented Apr 15, 2021 at 17:25

The graphic doesn't make it very clear, but at the moment of separation, the rocket still has vertical speed.

So it does only boost back, and not up, to cancel the horizontal speed. But the rocket will nonetheless go up until it cancels its vertical speed before it goes down.

  • 3
    $\begingroup$ Then that's an extremely bad graphic. $\endgroup$ Commented Nov 21, 2016 at 9:38
  • 1
    $\begingroup$ If that's the case: Then indeed the text "pushes the booster up and back" is totally inaccurate. It's probably just poor paraphrasing by someone non-technical of technical input, but really someone technical should have reviewed the graphic before it was distributed. Sloppy. $\endgroup$
    – user69911
    Commented Nov 21, 2016 at 12:23
  • 2
    $\begingroup$ PS. If you're right then it's not just the fine print (which is admittedly difficult to read at certain resolutions) – the top graphic of the rocket is clearly visible and should have been a red flag causing this to be sent back to be edited. I'd like to think that you're right, @radex, but I'd prefer to hear from someone at SpaceX before I stick a tick on this. $\endgroup$
    – user69911
    Commented Nov 21, 2016 at 12:36
  • 1
    $\begingroup$ Indeed, I could be wrong about this, I will delete my post if this is confirmed. Hobbes explanation is reasonable for this case, although in the general case it wouldn't make sense to accelerate up (and yet it will go up due to its vertical velocity). $\endgroup$
    – radex
    Commented Nov 21, 2016 at 18:07
  • $\begingroup$ That would make a lot more sense. Boosting horizontally is necessary to return to the landing spot. But boosting up is wasteful, because you do not need the altitude (you have plenty of it to kill) and because you will have to waste fuel just to kill off that extra gravitational potential energy as it gets converted to kinetic energy. Seems much cheaper and easier to just boost backward horizontally. $\endgroup$
    – user39728
    Commented Apr 15, 2021 at 17:31

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.