Update 08 April 2017: Elon Musk tweeted an indication that they're revisiting the possibility of second stage recovery, so this question is again relevant, and the details of the new solution may be different from the old.

Fairing is ~$5M, but that should be reusable this year. Am fairly confident we can reuse upper stage too by late next year to get to 100%.

SpaceX wants to make the Falcon 9 completely reusable. They're making definite, if slow, progress toward recovery of the first stage via powered landing, which requires several features:

  • Cold gas thrusters to turn the stage tail-first after second stage separation
  • Aerodynamic control surfaces at the top of the stage for trajectory control during the landing approach
  • Deep throttling a single main engine to get a low enough TWR for a controlled landing
  • Landing legs

The upper stage will be coming in a much faster trajectory, so they have to add a full heat shield at the front of the stage.

At first glance, the much shorter second stage would seem to need shorter legs because its center of gravity would be lower, but the upper stage version of the Merlin engine has a very long nozzle, which is radiatively cooled -- which means exposed, not shrouded, so the legs will have to mount quite high, and be fairly long (and thus heavy).

The biggest problem, however, would seem to be the engine throttling. The empty upper stage weighs something like a quarter of what the first stage does, but will be firing almost the same thrust level.

Are they going to try to throttle deeper?

Are they going to do the same hover-slam trick as the first stage, only on a much shorter timeframe?

Are they going to add 2 or 3 Kestrels alongside the Merlin on the second stage and land gently on those?

Parachutes? Airbags? Really big pogo sticks?

  • That last suggestion reminds me of some of my Minmus landers in KSP... – Sean Jun 15 at 2:52
up vote 16 down vote accepted

Old/Original Proposal for Second Stage Reuse

Old/Original Proposal for Second Stage Reuse

Second stage reuse for Falcon 9 was a planned feature that has been discontinued. In this old render(Shows a Falcon 9v1.0 with square engine block) you can see a reuse plan very similar to what you described. The second stage nozzle would retract into the tank structure, small legs would deploy, and smaller engines(kestrel, draco?) would perform the actual landing.

However, second stage reuse is hard. The first stage is a massive beast that does most of the heavy lifting into orbit. However, the second stage provides most of the speed. Once the second stage inserts Dragon into orbit, it would need to have enough fuel to deorbit/intersect the atmosphere then use a heavy heatshield to reduce its velocity on reentry. All of these things add weight.

Due to the nature of staging and the rocket equation, roughly every 4 pounds of dead weight on the first stage takes away 1 pound of payload. On the second stage it is a 1:1 conversion. All the extra weight for reuse(heat shield, legs, extra fuel, machinery for nozzle retraction) take away from the total payload mass.

SpaceX already has an inefficient upper stage relative to most launch providers. Their use of Kerolox on both stages simplifies design and tanking but means that the second stage Isp is much worse than a cryogenic Hydrogen stage. Now why did SpaceX change their mind? Elon Musk explained this at the MIT Aero/Astro Centennial. When asked about second stage reuse

The next generation vehicles after the Falcon architecture will be designed for full reusability. I don't expect the Falcon 9 to have a reusable upper stage, just because the - with a kerosene-based system, the specific impulse isn't really high enough to do that, and a lot of the missions we do for commercial satellite deployment are geostationary missions. So, we're really going very far out. These are high delta-velocity missions, so to try to get something back from that is really difficult. But, with the next generation of vehicles, which is going to be a sub-cooled methane/oxygen system where the propellants are cooled close to their freezing temperature to increase the density, we could definitely do full reusability - and that system is intended to be a fully reusable Mars transportation system. So, not merely to low Earth orbit but all the way to Mars and back, with full reusability. [Within 3 years?] Ha. I am an optimistic person, but - I think we could expect to see some test flights in the five or six year time frame. But, we're talking about a much bigger vehicle, and we're also going to be upgrading to a new generation - a harder engine cycle, which is a full-flow staged combustion. What we have right now is an open cycle engine. Right now, I'd say, engines are our weakest point at SpaceX, but they will become as strong as the structures and avionics in the next generation.

Falcon 9 won't be able to pull this off but the BFR/MCT will

Falcon 9 won't be able to pull this off but the BFR/MCT will

  • 3
    Could the F9 upper stage have a second use in space? Upper stages (with some extras I suppose) of Chang'e 2 and 5 went to Lagrange points and around the Moon and past an asteroid after having delivered the main mission. Or is it not designed for such lifetime and flexibility, and whatever interesting could be done from GEO anyway? – LocalFluff Jul 31 '15 at 8:48
  • Great answer. Looks like Wikipedia's F9 page is well behind the news. – Russell Borogove Jul 31 '15 at 13:46
  • @localFluff not without exstensive modifications. The second stage is battery limited, with just enough power to make it to GTO injection before it dies. SpaceX could add deployable solar panels or extra batteries but it adds weight. Also because the 2nd stage is Kerolox, there isn't that much delta-v left over from most mission profiles. It is better to have a 3rd stage as part of the payload instead. – T.J. Tarazevits Aug 1 '15 at 0:09
  • 1
    Actually deorbiting something from a low-periapsis orbit, especially with high apoapsis, is very easy and takes very little fuel. Now for making it survive the rentry... that's a real challenge. – SF. May 12 '16 at 21:35
  • I think your not taking into account SF. May's observation but your also not taking into account that the Falcon 9 Engines are to be replaced/upgraded to the D variant which gives them ... I think the terms was ... more than enough to do this. While your right about the ascent phase of this your not taking into consideration the "cost of things" IF one has the fuel to do a Supersonic retro burn then one does NOT need to do a high velocity descent through the atmosphere. I can slow the sped to ... a manageable level. – Enigma Maitreya Apr 8 '17 at 17:44

One possible thing which might help is to switch to methalox on the second stage to get additional performance to pay for grid fins and a heat shield. The subscale raptor engine they tested back in September 2016 has a thrust of 1MN, which is similar to the uprated thrust of the vacuum merlin 1D. The engine is also of a similar size.

The dimensions of the second stage can't change much but methane's lower density than kerosene is somewhat offset by the higher O/F ratio and LOX's high density.

  • Do we know what ratio the subscale Raptor uses? I can't imagine it would be a win without stretching the stage. – Russell Borogove Apr 12 '17 at 3:51
  • According to wikipedia it's 3.8. While digging that number out, I did come across this reddit post that does some math to suggest that with the same second stage volume methalox out performs kerolox for lighter payloads. Stretching would improve things more, but yeah falcon 9 is already too long. – Aethernaught Apr 12 '17 at 5:24
  • Nice find! I'm surprised at that result. – Russell Borogove Apr 12 '17 at 13:58

After they got launching seriously, SpaceX decided that second stage recovery would be too difficult, and still maintain a useful payload.

Musk recently mused in a tweet, that:

Really tempting to redesign upper stage for return too (Falcon Heavy has enough power), but prob best to stay focused on the Mars rocket

This indicates that the better ISP of the Raptor engine under development for Mars missions might make it possible.

  • Have you considered the D variant engines that are replacing the current engines, provides increased "payload" capabilities which can translate to the fuel to slow the 2nd stage down? – Enigma Maitreya Apr 8 '17 at 17:46
  • The D variants did not bring gains in any way useful for enabling S2 reuse. The limiting factor is payload capacity going uphill - reserving more mass for recovery will cut directly out of the capacity to orbit, and they would lose high paying customers by reducing that capability. Higher thrust engines did not improve that - a tank stretch or fuel change may provide the option, but they are not taking that path. – Saiboogu Nov 23 at 19:13

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

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