Fully reusable Falcon 9 with Raptor upper stage

[Robotbeat]

Just doing some basic calculations, here.

I think the current F9R, in expendable mode, can get:
3.*ln((400+130)/(16+130))+3.41*ln(130/(7+20)) = 9.2km/s

..about 20 tons to LEO expendably vs 13 tons on their list price. This is to a low orbit.


Increase the thrust slightly (5%) of the first stage, and get a 150 ton upper stage (vs 110ton now?) :

3.*ln((400+177)/(16+177))+3.7*ln(177/(8+27)) = 9.3km/s
...and you should get at least 27 tons to orbit expendably.

With reuse:

3.*ln((400+169)/(60+20+169))+3.7*ln(169/(8+6+13)) = 9.3km/s

(first stage goes from 16 tons dry to 20 tons dry to account for fins and legs, with 60 tons of propellant left at staging to allow a generous delta-v for RTLS, reentry, and landing... worst case, you do a barging instead)

And for the upper stage, I give 75% of the dry mass of the upper stage for recovery hardware and landing propellant, which I think is pretty generous as well. (rule of thumb for TPS systems is 10% of entry mass, on top of 20% for legs/thrusters and allowing about 600m/s delta-v (given a 240s average Isp for thrusters) for reentry and landing, which should be about twice as much as you should need in ideal circumstances.

...so 13 tons to LEO fully reusable (even with RTLS, most likely) is feasible, I think. And with a barging, it definitely should be possible.

[ChrisWilson68]

Thanks for doing the calculations!

Now for the interesting discussion: how will it land?

I think it's pretty obvious a Pica-X heat shield with some combination of grid fins, cold gas thrusters, and Draco thrusters can get it down to terminal velocity, which is probably a few hundred mph at most.

So then what?  Parachutes?  Landing legs?  Front or back side?  Extra landing engine for propulsive landing?  Or ultra-deep throttling of the Raptor engine?  Or some other recovery mechanism?

[Lars-J]

Thanks for doing the calculations!
So then what?  Parachutes?  Landing legs?  Front or back side?  Extra landing engine for propulsive landing?  Or ultra-deep throttling of the Raptor engine?  Or some other recovery mechanism?

10g Raptor landing burn at the last second!  (okay probably not)

[NovaSilisko]

I've long had this sort of design in mind for how a reusable stage would land. Basically recreating the butt end of a Dragon on the top of an upper stage, out to the diameter of the current flange of the fairing. Might have to coat the whole damned thing in some sort of ablator. The center of mass strikes me as something that would be problematic, as well, with the engine being so high during entry (possibly compensated for by the landing thrusters and propellant?).

Excuse the 20 minute hackjob. No idea how big a raptor using stage would actually be so I just kept it the same as is shown on the SpaceX website...

[Dante80]

This is close to the original concept SpaceX had for Falcon9 S2.



How would the PAF bolt on to the heatshield without compromising it? It has to also bear a lot of load.

[NovaSilisko]

How would the PAF bolt on to the heatshield without compromising it? It has to also bear a lot of load.

Perhaps the same structural path that the landing legs take. For the fairing, maybe the fairing halves could encapsulate the outward bulge of the shield entirely. Also, I can't remember, how does the trunk mount to the capsule for Dragon?

edit: Something like is attached for the fairing as an alternative

It would require a slightly longer fairing to afford the same payload volume. Mounting the payload to the shield becomes the main issue, it may require an expendable mounting structure that attaches to the shield (transferring the load into the landing legs perhaps, as mentioned) by some means which is jettisoned before reentry. It is the means of attachment I'm not sure of.

Of course for all we know they might come up with anintegrated upper stage and reentry vehicle with a payload bay - a wingless space shuttle with propellant tanks if you will. Something like the second attachment.

[adrianwyard]

...Mounting the payload to the shield becomes the main issue, it may require an expendable mounting structure that attaches to the shield (transferring the load into the landing legs perhaps, as mentioned) by some means which is jettisoned before reentry. It is the means of attachment I'm not sure of.

This doesn't sound too problematic. I'm picturing something like ET umbilical doors in the heat shield which launch open, exposing hard points on S2 to which the expendable payload mounting structure is attached. Once the structure is jettisoned, they swing shut.

[Darga]

Hypothetically, what would the performance of F9/FH be with a same length but 5.2m diameter second stage with an optimized raptor?

[ArbitraryConstant]

Thanks for looking at this, I've been thinking along these lines as well.

(first stage goes from 16 tons dry to 20 tons dry to account for fins and legs, with 60 tons of propellant left at staging to allow a generous delta-v for RTLS, reentry, and landing... worst case, you do a barging instead)

One thought I had was that the upper stage grows quite a bit, and the first stage is de-stretched to compensate. That's a bad optimization for an expendable rocket, but in SpaceX's case it would make RTLS much easier. It's nearly an air launched SSTO, the downrange motion to counteract would be even smaller.

In the case of Falcon Heavy the de-stretch with heavier upper stage makes staging earlier as well, which helps RTLS for higher energy missions.

[rst]

Now for the interesting discussion: how will it land?

I think it's pretty obvious a Pica-X heat shield with some combination of grid fins, cold gas thrusters, and Draco thrusters can get it down to terminal velocity, which is probably a few hundred mph at most.

So then what?  Parachutes?  Landing legs?  Front or back side?  Extra landing engine for propulsive landing?  Or ultra-deep throttling of the Raptor engine?  Or some other recovery mechanism?

One possibility is in-air recovery, as ULA is planning for the Vulcan engine compartment.  That would have the nice side benefit of saving weight on landing hardware (trading chutes and parafoils for legs, separate engines and fuel for them, etc) -- which matters more than on the first stage, as weight of recovery hardware on the second stage trades off directly against payload mass.

[gadgetmind]

Any guesses what such a second stage would weigh? It's going to be quite a big lump to catch!

[Stan-1967]

Now for the interesting discussion: how will it land?

I think it's pretty obvious a Pica-X heat shield with some combination of grid fins, cold gas thrusters, and Draco thrusters can get it down to terminal velocity, which is probably a few hundred mph at most.

So then what?  Parachutes?  Landing legs?  Front or back side?  Extra landing engine for propulsive landing?  Or ultra-deep throttling of the Raptor engine?  Or some other recovery mechanism?

One possibility is in-air recovery, as ULA is planning for the Vulcan engine compartment.  That would have the nice side benefit of saving weight on landing hardware (trading chutes and parafoils for legs, separate engines and fuel for them, etc) -- which matters more than on the first stage, as weight of recovery hardware on the second stage trades off directly against payload mass.

I had a similar discussion on the "Falcon Super Heavy" thread about mid-air recovery of smaller boosters weighing less than 9000kg empty.   I didn't see much need for using full F9 S1 boosters as the performance of F9 FT has increased the overall capability so much. 

http://forum.nasaspaceflight.com/index.php?topic=21867.msg1473011#msg1473011

With all the extra performance of the F9 FT, and now additional plans for a "Raptor" S2 engine, S2 margins will become even greater, & re-use is possibly enabled.

If S2 empty mass can be kept under 9000 kg's, ( or whatever baseline payload capacity of the recovery helo)  why would SpaceX not try to perform mid air capture of S2?   

SpaceX has already shown they "evolve" re-use schemes to try for the least complicated approach.   They started with parachute recovery, and then moved to several iterations of propulsive & aero techniques to get to the latest successful RTLS flight.

It seems logical to progress S2 reuse along similar lines:
1.  Have enough propellant margin for a targeted propulsive de-orbit
2.  Propulsive reduction of velocity prior to re-entry depending on margins
3.  Deploy & orient heatshield for re-entry
4   Survive re-entry

From there you can branch off into parachute landing to ocean surface, mid air capture, propulsive landing etc.   

[NovaSilisko]

I almost want to try working the numbers and refining the design for that little Falcon-Shuttle design just as an exercise. Of course, we don't know a thing about this new Raptor design, so that's a little tricky...

[adrianwyard]

I almost want to try working the numbers and refining the design for that little Falcon-Shuttle design just as an exercise. Of course, we don't know a thing about this new Raptor design, so that's a little tricky...

IIRC people have argued that the additional cross-range that comes from a more aerodynamic S2 eases the challenge of getting to a landing facility - or even enables it. And having more heat shield area for its mass might ease the return from GTO case?

[rst]

Any guesses what such a second stage would weigh? It's going to be quite a big lump to catch!

Spaceflight 101 estimates the dry weight of the current F9 second stage at about 4000 kg.  That compares to roughly 5500 kg for the RD-180 engine alone, which ULA was talking about recovering in midair before they re-applied the concept to Vulcan.  (Their paper on the RD-180 variant is here: http://www.ulalaunch.com/uploads/docs/Published_Papers/Evolution/PartialRocketReuseUsingMidAirRecovery20087874.pdf).

BTW, on dealing with heat shield attachment, a furled HIAD (NASA's inflatable heat shield) *might* be easier to fit on the same rocket front end as spacecraft mount hardware.  (Also, since they typically inflate to well over the spacecraft's launch diameter, they'd make it somewhat easier to keep that large MVac engine bell out of the reentry slipstream...)

[david1971]

...so 13 tons to LEO fully reusable (even with RTLS, most likely) is feasible, I think. And with a barging, it definitely should be possible.

So, with some Legos, could you stack a Dreamchaser on top? 

[NovaSilisko]

BTW, on dealing with heat shield attachment, a furled HIAD (NASA's inflatable heat shield) *might* be easier to fit on the same rocket front end as spacecraft mount hardware.  (Also, since they typically inflate to well over the spacecraft's launch diameter, they'd make it somewhat easier to keep that large MVac engine bell out of the reentry slipstream...)

Hmm. Something that fits underneath a truss structure that holds the payload, which is jettisoned before the shield is inflated. Legs would be a bit tough to get out through that, unless you flipped the stage round and landed bell-first again (the shield would act as a very Mary Poppins esque airbrake, actually)

[RocketmanUS]

Does all engines including RCS use the LCH4/LOX as their propellants.

I assume the advantage of using LCH4/LOX so the stage could be used for days instead of hours?

If the were to replace the Merlin for Raptor on F9/FH I would expect them to replace F9/FH with a reusable two stage launcher. Why keep a tri-core when a single core would have less risk of failure? That could possible be nine full sized Raptor on the 1st stage. Then use more of the engines on the BFR later on after they have been tested on the smaller launcher.

[Lars-J]

Does all engines including RCS use the LCH4/LOX as their propellants.

Not necessarily, but it can be done. Helium can be eliminated as well. (ULA is planning such upgrades for Centaur, which is hydrolox, but the same can be done for methalox)

It would be a good proof of concept for MCT development, where it in practice MUST be done.

[douglas100]

One possibility is in-air recovery, as ULA is planning for the Vulcan engine compartment.  That would have the nice side benefit of saving weight on landing hardware (trading chutes and parafoils for legs, separate engines and fuel for them, etc) -- which matters more than on the first stage, as weight of recovery hardware on the second stage trades off directly against payload mass.

Yes, meekgee suggested this on another thread. It's also got the advantage of operational flexibility that the stage can be recovered at lots of different locations and quickly brought back to the launch site. The helicopter recovery can always be offshore for safety reasons as well.

It's pretty difficult to get rapid RTLS for an upper stage without a reasonable cross range ability. This technique gets round that difficulty and also keeps the recovery equipment to a minimum on the stage itself.

I'm very much in favour of nose first entry. That means nothing needs to be changed at the engine end of the stage. An inflatable heat shield sounds promising but the technology is not mature yet and the trade against a conventional TPS isn't clear yet.