12 thoughts on “DragonFly”

  1. Will there be any niches in the launch profile that will -not- covered by “Separate all the parts, and fly everything possible back home?”

    And can it be ‘amphibious’ without a significant weight penalty (just for the emergency landings?)

    1. As to your first question, I think the answer is no. If and when they get serious about recovering the upper stage, they’ll likely do some landing tests with it as well. While it may seem similar to what they’re doing now, the upper stage is much smaller than the first stage and also has a much larger engine nozzle. I’ll bet they’ll want to test that configuration. For one thing, they’ll need to be able to throttle the upper stage Merlin engine low enough to handle the lighter stage. In future launches, they’ll also try to do deorbit burns and see how well their heat shielding works. That’s a much harder problem than first stage recovery.

      As to the second question, I’m sure the Dragon will always retain the ability to come down in water if they have to, such as an abort during launch. If they use their SuperDraco thrusters to abort, there likely won’t be enough propellant to do a propulsive landing. However, looking at some of their proposed tests, that may be a faulty assumption on my part.

      1. They can’t use the Merlin 1D on the upper stage for powered landing. One issue, it would have to throttle very, very low. Dry mass on the upper stage is about 10,000 lbs, and it’d be nearly dry. Merlin’s thrust is about 148,000 at sea level, so you’d need to throttle, even for a hoverslam, down by more than 90%. (They currently have to deep throttle a single Merlin on the first stage to even get close to a 1-1 thrust ratio, and the first stage weighs vastly more)

        And then there’s the nozzle. Throttle a first stage Merlin down that much and you’d get instability due to over expansion. But, the second stage nozzle is enormous; it’s already very over-expanded. That’s fine in a vacuum, but in atmosphere, it’s not.

        The third problem is keeping the small amounts of LOX remaining viable after reentry heating. That would be hard.

        On the other hand, two super dracos could easily handle landing the stage. It looks to me like the easiest, and perhaps only practical, way to land a second stage is to put a Dragon heat shield on the nose, use the MErlin only for a deorbit burn, do a ballistic reentry (pulling about 8.9G from LEO, or 9.5G from GTO, which the stage can handle in that axis) and fire the dracos a few seconds short of the pad. Due to aerodynamic and CG issues, it’s widely said that the second stage would be unstable entering nose first (and an active RCS would use too much fuel mass controlling it) so every surface will need to be protected by a heat shield. That’s not true and it wouldn’t work anyway; if it goes unstable during reentry it’d break up due to G and airo forces. Fortunately, there’s a solution; spin the stage prior to reentry for stabilization. (I have a hunch that they’re already testing that aspect when they deorbit the second stages, which we fist saw them do on CRS-3).

        I’ll bet you’re right that they do a “grasshopper” version of the second stage, but I’m betting they’ll use ballast in place of the Merlin.

        I think you’re right on the contingency water landing ability; they’d have to have it in case of aborts. Heck, even Soyuz can, and has, landed in salt water (and the crew was unharmed). And BTW, if they use the Superdracos to abort, you’re absolutely right; they won’t have the fuel to do a propulsive landing.

        1. I like your idea of using SuperDraco thrusters for the second stage. That would eliminate the problems I outlined with trying to use the vacuum Merlin D while having commonality with the Dragon thrusters. It would require adding hypergolic propellant tanks and sufficient propellant to the second stage which would require extending it somewhat. From what I recall, they were talking about using a blunt nose on the upper stage for heat shielding, so putting the hypergolic tanks high in the stage might help with center of gravity issues. That stage already has attitude control but they’d need to expand it for longer operations and perhaps more powerful to handle the aerodynamic forces during reentry.

          As for a grasshopper version of the second stage, that seems a safe bet. Whether they use ballast in place of the Merlin or just use a non-flight rated engine they have laying around it up to them. They’re bound to have some leftovers from the development program, so you might as well use them. It’d be cheaper than ballast.

          1. I have to admit, the use of two Superdracos on the second stage isn’t my idea. SpaceX put out a video a few years ago showing the second stage touching down with some kind of thrusters and not the main engine, so I just assumed they meant superdracos (recently; I’d never heard of superdracos when they put out that vid). The SD are throttleable, and two would more than handle the dry mass of the first stage easily. I also had a huge clue that that’s what SpaceX intends; A while back I saw a rough schematic of the second stage, which had two small tanks in the blunt conical end (top end on the pad). I knew they weren’t the helium or nitrogen tanks, so my guess was they had to be monometyl hydrazine and Nitrogen Tetroxide (Draco fuel).

            The only actual semi-original idea (and really, not even that; a bullet gave me the idea, plus spin stabilization is very common and old tech) in my post was spin-stabilizing the stage to deal with it’s inherent instability during reentry and flying a ballistic reentry profile (both by necessity, and to achieve precise targeting on the touchdown area).

            BAsically, I was wondering how SpaceX could recover the second stage./ I know they think they can, and thus have a plan, so I thought it’d be fun to see if I could figure out how they intend to do it. I knew mass was a huge issue so whatever it is has to be low mass. The rest just popped into my head (proof that nature does indeed abhor a vacuum) and when I ran the numbers, it worked; total added mass to the upper stage is around 790kg, and requires very minimal changes. I floated the idea on a thread over at Nasaspaceflight.com, expecting it to be shot down (I was sure I’d missed something) but it wasn’t, BTW, one part I didn’t mention here; they’d probably need to deploy two small cantable steering vanes (about one square foot each) from the stage base after reentry. These would handle despin and stabilization. Oh, and foe the spinup, they’ll probably need two regular draco thrusters. My SWAG for the needed Rpm is 120 to avoid nutation. That’d handle keeping it stable during reentry (though I’m not certain on the RPM needed).

            Great point on using a non-flight Merlin for ballast; that’s far better than boilerplate, and probably a lot cheaper to do as well.

    2. I think they will probably retain the capability to do water landings. But actual ground landings enable must faster mission turnaround. You have seen how much of a hard time they have been having with sea recovery. When Apollo did it they allocated a large chunk of the US Navy just to search for one capsule. That is anti-economic.

      I see no problem in separating the parts. This way they can optimize the recovery system. A large part of the mass and parts will be in the first stage and that comes down a lot slower than an upper stage so it needs less heat shielding. The upper stage will need robust heat shielding and the current second stage engine has a nozzle optimized for high altitude. It may not work that well for recovery and I would not be surprised if they had to redesign the nozzle or use some other recovery system. Also as Larry J said there is a question of how much speed can you bleed off the upper stage by e.g. aerobraking.

    3. Oh yeah. To me what is surprising is why no one else tried to do this before SpaceX. I have read a lot of proposals of similar systems over the years and I was always dumbstruck about it. It seems some people in the 1960s and 1970s thought it was a ‘waste of time’ interim step and that we should just go straight for Shuttle. However on unknown systems like these jumping intermediate development steps is not always a good idea. Then there was the DC-X. I hope SpaceX proves such a system works and is economic so we can actually advance launch vehicle design.

      Between this and Raptor plenty of interesting developments coming out of SpaceX. They have been exhausting the design space quick enough.

      1. I think a lot of factors are to blame.

        A big one is how screwed up NASA manned spaceflight has been since the start. NASA space science has been an above average government operation, partly because there is such a diversity of projects and each project tends to be not too large (with few exceptions). But manned spaceflight and as a consequence rocket development was driven originally by the moonshot, which required serious political machinations from LBJ, which are still in place.

        Additionally, private dabblings in spaceflight were difficult during the Cold War, which hampered development and funneled all efforts into government procurement programs, which are the worst for development of novel vehicles.

        Another aspect is the myth of NASA’s omnipotence. NASA does some good work now and again but they make plenty of mistakes and oversights too.

        This plays into two other factors, the lack of perception of commercial viability in spaceflight and the demand for the “coolness factor”. Not many investors have dabbled in spaceflight startups, and those that did tended to demand “cool” vehicles. Rotary rocket being a prime example. But NASA suffers from the same problem. There usually hasn’t been enough political pull to get an open ended R&D program going, or even an X-vehicles program. So that just leaves procurement. And the trend there has been to attempt procure-development, often with disastrous results. Typically there’s this idea that settles in that they can just skip a tech generation and proceed to an advanced vehicle. Which they tried with Shuttle, and VentureStar, and are doing mostly the same thing again with SLS and Orion. Partly it’s a pride thing, because getting behind schedule or failing to achieve success with something novel is easier to write off.

        The remarkable thing, of course, is that everything SpaceX is doing now was right under NASA’s nose decades ago, especially with the DC-X. But NASA has never actually been seriously committed to low cost spaceflight, they have always been distracted by shiny futuristic ideas that never quite pan out.

        1. First, NASA’s mission is basic technology development, so this type of approach was just too simple. That also is why when they made the X-33 selection they pick the most technical risky option, the Lockheed entry, instead of the more practical DC-Y or Shuttle II entrants.

          Second, if human space flight ever becomes easy NASA is out of a job. Its hard for a NASA astronaut to be exceptional if there are thousands of folks flying into space, and more importantly, hard to justify the continual flow of pork into the NASA Centers that is needed for continual Congressional support.

          Third, NASA science only looks good because there is no market for private firms to want to mess with it. But from I have done consulting with start ups I know many of those missions could be done for a fraction of the price. Like Project Apollo folks don’t have any other reference so they don’t realize how much cheaper and more productive those missions could be.

Comments are closed.