Let us suppose we manage to mine water on the Moon and return lots of it to LEO. Would it be hard to design an upper stage that could refuel at that depot, and uses that fuel to return to the surface and be reused?

I was reading about ULA's Advanced Cryogenic Evolved Stage, and in a Defense News article Tony Bruno said

"We had the idea, well, why do you have to bring it back to Earth just to reuse it?" Bruno said "Why don't we just leave it in space?"

...Yeah, but... why not bring it back to Earth and reuse it?

Is it just a matter of how to have enough fuel up there so that its reasonable to propose refueling a bunch of specialized upper stages with enough fuel to propulsively decelerate and land? Are there other complications?

Maybe I should mention that I'm thinking about this in terms of long-range lunar development. Jkavalik mentioned in comments that a big complication is that things are launched to many different orbits and so it would be rare for an upper stage to be in range of a fuel depot (though they mentioned an architecture change where payloads are transferred to space tugs that complete the orbit insertion process at the depot could get around that). In the particular case I'm thinking of, the payloads are destined for one staging area before continuing on to the Moon, probably a space station.

  • $\begingroup$ But how to reentry an upper stage without the necessary heat shield? Some fuel is necessary to initiate a reentry from a LEO, but without a heatshield the stage would be destroyed. If the decceleration should be done using fuel only and not the drag by the atmosphere, you would need more fuel that would fit in the upper stage. The first stage was needed to enable the second stage to reach the orbit. To reverse the ascent into a descent without heatshield, the first stage would be needed for the lower part of the descent. Just imagine you would try to return the stage without an atmosphere. $\endgroup$ – Uwe Sep 23 '17 at 17:10
  • 1
    $\begingroup$ Delta-v of a filled upper stage without the payload would be quite a bit higher and might allow to cancel most of the orbital velocity. $\endgroup$ – jkavalik Sep 23 '17 at 17:20
  • 1
    $\begingroup$ @kimholder has propulsive reentry from orbit been seriously proposed for the Earth's atmosphere by propulsion engineers? Can you offer a link where we can read more about it? I think the actual question in your question is really "Could propulsive reentry to avoid the use of heat shields possible in the Earth's atmosphere" and not about mining water on the moon, so the title should contain "propulsive reentry" and "Earth's atmosphere" explicitly instead of waiting until the fifth paragraph to mention it explicitly. $\endgroup$ – uhoh Sep 24 '17 at 3:25
  • 1
    $\begingroup$ @kimholder OK would you mind then if I asked as a separate question; "Has the use of propulsion during reentry of Earth's atmosphere from orbit ever been seriously considered or the issues addressed?" I'm thinking that it be a separate question because I'm asking about prior engineering work and you're asking "what would be the difficulties of using lunar water to get back to Earth". $\endgroup$ – uhoh Sep 24 '17 at 4:29
  • 1
    $\begingroup$ @kimholder let's restrict discussion to the question itself. I'm recommending you improve the question by focusing on the role of propulsion in Earth's atmosphere reentry. Does having more than a tiny bit of left-over necessary for the empty spacectraft to lower its apoapsis have any utility in reentering Earth's atmosphere? Could having more fuel make 2nd stage reentry from orbital velocity easier? Refueling strategies for LOX/LH2 is a whole separate question. Asking them separately allows for people with different expertise to address each part even if they don't have an answer to the other. $\endgroup$ – uhoh Sep 25 '17 at 3:48

It's close. Plugging in the numbers for the Falcon 9 upper stage I get 11,300m/s of delta-v. Since 9000m/s will get you to LEO that's enough to get you back down with something to spare. However the upper stage doesn't have landing gear. Since you have 5000kg of payload capacity (brings the delta-v down to 9111m/s) you could land it. The weight of the legs and other landing equipment would come directly off the weight of the payload, though, on a 1:1 ratio.

Landing this will be a nightmare. The upper stage has only one engine and it doesn't throttle.

Looking for some numbers on the Falcon 9 I find the landing burn is 267 m/s but that includes 117 m/s of gravity loss. The upper stage will be landing very, very hot, though—if we figure the max payload weight of 5000kg the engine still is going to be putting out 103m/s^2 at exhaustion. The actual velocity to kill is only 150 m/s. Figure 1.5 seconds for the burn which adds another 15 m/s of gravity loss, so figure 1.6 seconds for the burn. The accuracy required is going to be incredible. If the engine ignites 1 millisecond early it's going to shut down while still 66 cm in the air above the pad and the rocket will have to fall the rest of the way. If it ignites 1ms late it hits much harder, my gut says the same as a fall from 7 meters up and it's late enough I don't want to spend the time to confirm this. Note that the engines do not start with anything like 1ms precision!!

(Note that I am using the terminal velocity of the first stage. Nobody cares about the terminal velocity of the second stage as it doesn't fall anyway so the number isn't to be found. I would assume it's a bit lower. This will change the duration of the landing burn but will not change the touchdown numbers.)

  • $\begingroup$ No source at hand but I believe M-Vac can throttle. $\endgroup$ – jkavalik Sep 25 '17 at 5:00
  • 3
    $\begingroup$ Throttling isn't the issue - the MVac is optimized for vacuum operation, and the exhaust will be grossly overexpanded at near-sea-level pressure. That will result in flow instability and, most likely, a pile of smoking scrap metal at the landing site. $\endgroup$ – John Bode Sep 25 '17 at 14:48
  • $\begingroup$ If its maneuverability was somewhat close to that of an F9 first stage, then the stage could initiate reentry at a point that would bring it in to, say, the Nevada Test Range, and it could deploy parachutes for the final descent. Would that be enough? $\endgroup$ – kim holder Sep 25 '17 at 16:06
  • $\begingroup$ The stage will be destroyed when hitting the ground after the final descent with parachutes. All manned capsules landed with parachutes after reentry had additional equipment to limit the shock when hitting the ground. There have been many parachuters that broke a leg by a bad landing. $\endgroup$ – Uwe Sep 25 '17 at 17:41
  • 1
    $\begingroup$ Well, even the original SpaceX Reusable Launch System video shows the second stage using some secondary propulsion system (Draco-derived?) for the final landing (but together with the forward-facing heat shield, a flip in the atmosphere and retractable nozzle it is probably not very realistic). Adding such secondary system would again lower the mass to orbit. $\endgroup$ – jkavalik Sep 25 '17 at 18:00

You need to keep in mind that every kilogram you add for a recovery system on the upper stage robs an equivalent kilogram from the payload.

With a refueled stage that can shed most of the 9000 m/s ΔV in the deorbit burn, we don't need a heat shield (or at least as robust a heat shield). We could probably get away with something like the inflatable heat shield NASA has been testing.

I'm not sure about an engine-first re-entry, though; the end of the MVac nozzle is a little thicker than a soda can (remember they were able to manually trim four feet off of it using a pair of tin snips for COTS-1), and I'm not sure it wouldn't be damaged from the buffeting as the atmosphere got thicker. Then again, I'm not sure a nose-first re-entry would protect it all that much, either (not an aerospace engineer, so feel free to ignore that).

You won't be able to use the MVac for propulsive landing — it's optimized for vacuum operation and would be ridiculously unstable at near-sea-level pressure. You'd need something like the SuperDraco for an actual landing.

A parachute drop coupled with some serious shock-absorbing legs and a quick Soyuz-style retro-fire from a Draco or SuperDraco might work, but that's a lot of payload mass to sacrifice. The on-orbit use case makes a lot more sense.

You're still robbing a non-trivial amount of payload mass to pay for those legs, parachutes, and thrusters.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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