22 thoughts on “The Key To Reusable Launch”

  1. Stratolaunch, which plans to drop a “Pegasus II” rocket stack with expendable components, is projecting a 6,100-kilogram payload capability. Because the 3STO system described here has fully-reusable components with attendant weight penalties, it will likely have considerably less payload capability than Stratolaunch for comparable launch sites, perhaps 3,000 kilograms. Detailed engineering analysis of specific point designs—designs which specify propellants used, engine performance, thermal protection material choices, etc.—is needed to ascertain actual achievable payload.

    While he’s trying to make a shortcut calculation of the performance penalty of reuse, he needs to factor in that the Pegasus II (like the Pegasus) is all solid fueled. Unless you’re planning on using a parachute, I don’t know how you’d recover its stages for reuse. Adding a parachute to the first rocket stage wouldn’t have nearly that severe a performance penalty but recovering the upper stage isn’t likely to be possible or worth the effort. SpaceX dropped out of the Stratolaunch project because it required too many changes to their Falcon rocket. If you really want to recover and reuse the stages, it seems that solid propellants aren’t the way to go but I don’t know who would develop your liquid fueled rocket.

  2. The point of the solid rocket boosters is to be cheap thrust.

    I’m basically still stuck on “So… why aren’t they -cheap-?”

    We’re nowhere near having the fuel costs dominate launch costs for the whole rocket … but it should at least be true for the solid rocket boosters. Right?

    Where, exactly, do they end up so excessively expensive for their job? And why would we even want to recover them? (Empty tube, now stressed.)

    I do really understand the value of reusables, but the idea of true mass-produced expendables -also- seems under-explored. Estes manages solid rockets where the cost isn’t dominated by the cost of the cardboard shell, and the smaller rocket-assists for jets aren’t mind-blowingly expensive.

    At what size do they start requiring a serious design effort?

    1. Well, the idea of cheap mass produced expendibles has been explored. You can do a web search for Big Dumb Boosters and find some good write-ups. In some ways, the Soyuz booster falls into this category. It’s the most successful booster ever made: more launches than anything else; used to launch civilian and military missions including humans. Back in the 1980s, the old Soviet Union was doing 60+ space launches each year and most were on the good, reliable and relatively inexpensive Soyuz booster (or it’s Molynia variant with an additional upper stage). They made minimal changes to the rocket once they got the bugs worked out. If one failed, they didn’t go into a years-long session of naval gazing about it. They just rolled out the next one from the assembly building and launched the backup payload. In one case I remember from around 1986 or early 1987, they had the replacement satellite on orbit two weeks after a launch failure. Hard to beat that kind of responsiveness. The cost of a Soyuz launch has gone up considerably in recent years in part because the launch rate is so much lower and there’s little reason to lower the price when they don’t have much competition. You want to launch a crew to the ISS? You’re going to have to use a Soyuz for the time being. Their recent launch of a new variant without the distinctive Stage 0 strap-ons was interesting.

    1. Might come down on a bass boat.

      My only concern about really cheap air launch of small payloads is that it might put a kink in the price structure for satellites, moving all the smaller payloads away from launchers that have the potential for manned capability.

  4. The idea in a nutshell: air launch with your landing site downrange so the first stage doesn’t have to boost back to the launch site.

    It’s no more terrible than other air launch concepts.

  5. You have this humongo subsonic turbofan craft to carry a full loaded rocket stack west of the base for an airdrop launch, so the first rocket stage (Stage 2 in that reckoning) can glide to home base for landing.

    How about ground launch from a site having an eastward/downrange recovery runway for the initial rocket stage. You then transport that rocket stage westward, empty of fuel back to the launch site?

    OK, somebody tell me what I am missing.

    1. The cost of maintaining two sites. The effect having those two sites fixed on the Earth has on your launch flexibility.

      Of course, you pay for this by having a big stupid airplane to maintain, and all the other disadvantages of air launch.

    2. Still, he has a very good point. Flying an empty stage is vastly easier than flying a full one. We used to fly empty Saturn V stages from the factory to Florida without even thinking much about it. If you had a stage that could reliably land on any commercial runway, you’d have the vast number of existing runways as potential landing sites, and all you’d have to do is just fly a plane there to pick them up and fly them back, unless you opted to ship them FedEx or UPS. But if your stage is reliable enough to fly over populated areas and land on runways, it’s also reliable enough to partially refuel and launch westwards. If the Soviets tried it, they’d probably design the stage to land in any random field and then send an off-road fuel truck to it.

  6. Even though I have worked on at least one reusable SSTO proposal (with Rand, in fact), I have become convinced that it is not possible with chemical propulsion, launched from the earth. The version Rand and I worked was one of only two such concepts ever devised that would come close (Roton being the other). But “close” for reusable SSTO is useless.

    The logistics problem involved in multistage launches are formidable unless you have inland flight corridors. That is, unless you know a trick. I recently discovered that someone has devised that trick, and I could kick myself for not having had the same epiphany.

    There is no logistics barrier to a two or three stage RLV. Three stages increases the operational cost some, but is indifferent to launch site location. The same applies to two stage, or even four stage rockets.

    The answer certainly is not in solid rockets, which (in 1980) cost $25/pound of motor weight. The recurring cost of a reusable solid decreases with increasing weight, but only to a point. It is still ~$40 per pound of motor weight in today’s dollars. Multiply that by 30 to get the minimum cost per pound in orbit. That is the propulsion limited part of the cost of a solid rocket to orbit, and it is intractable.

    Three rocket propelled stages, using the proprietary trick to which I allude, would make reusable vehicles a relative breeze…unless you start by setting the payload weight to 60,000 pounds in. 28.5 deg, 100 nmi orbit. Then no business case closes.

    1. I detailed some concepts for horizontally-oriented vertically-launched multi-stage rockets here and over at Selenian Boondocks, but lately I’m wondering what effect it would have if someone built engines whose performance compares favorably to the SSME, RD-0120, RD-180, and NK-33, but whose price was less than perhaps $100G, which I think is feasible with a radical change in the way heat is transfered and how the compressors operate. I think a lot of the architectural corners we’re in are the direct result of high engine prices, which also drives our reluctance to bend metal or experiment with actual flight hardware without two truckloads of documentation and analysis. If early piston engines for aircraft cost $20 to $60 million dollars, the golden age of aviation would have produced maybe one percent as many experimental aircraft as it did, especially if success or failure meant you didn’t get the engine back.

  7. Why not just fly the aircraft/1st stage East with cargo or passengers for the ground site, launch the orbital stage for boost to orbit and land the aircraft/1st stage right there at a new launch base with the passengers? Then repeat at each launch site around the world. Six to a dozen sites will put LEO launch sites well within range. Spacex could do this with the current system and avoid spending fuel to get back West. Just keep going East.

    In two hops I can get to India in 30-60 minutes (more or less I have not worked the numbers) counting time to get on the next flight in Africa.

    This can’t be the first time this was suggested?

    1. There is the complication that the operating sites need to be in mutually friendly territories, connected by great circle segments that approximate the orbits to which payload is delivered, and all need access to deliver propellant and (ideally) payload.

  9. Another advantage of air-launch that’s often overlooked is not having to use (and pay, and wait in line for) one of the established national coastal launch ranges. At smaller vehicle/payload sizes range cost can dominate the overall cost of the launch – it’s a major obstacle in the path of the still-mythical million-dollar standalone launcher (and why many such proposals use air-launch.) Even at larger sizes the direct and indirect costs of the established ranges make a significant difference.

    1. SpaceX tried to go to Kwaj to avoid the range nonsense at Vandenberg.

      They ended up at Vandenberg and the Cape.

  10. Answer me this.

    I know that the folks who are train nuts are nuttier than the aerospace nuts, but there are parallel discussions between the passenger train community and the passenger space community.

    Think of it, NASA is, “Amtraaak iiiinnn spaaaacce!” (repeated letters denote announcer reverb)

    Or Amtrak is NASA with trains. Or something.

    So, there are a group of people who say we “need” long distance trains with sleeping cars and diners, mainly the folks who enjoy subsidized leisurely rides in sleeping car compartments taking meals in the dining car. And there are those who say, “no, get rid of the long-distance trains — concentrate on “corridors”” (Like the Acela train in the Northeast that is still subsidized but it is popular with business passengers paying high fares covering a bigger chunk of their costs).

    The argument goes, “You think the LD (long distance in the argot) trains are so expensive, get rid of ’em and see if Amtrak even saves any money!”

    Aren’t we rid of the Shuttle right now? Is NASA actually saving any money right now? Is this experiment of actually closing down the Shuttle and seeing if NASA spends less money, or how much less money, a good read on Shuttle direct operating costs?

  12. I’m a bit confused over the air launch concept when it comes to liquid-fueled rockets.

    Okay, with a solid, getting it to ignite is a fairly straightforward process. However, that’s not the case with non-hypergolic liquids, as we’ve seen recently, such as with Space X being unable to relight its first Falcon 9 1.1 upper stage (it was only being done for test purposes – and is a great example of why you test when the opportunity arises). We’ve also seen them (and others) have post-ignition pad aborts.

    However, if you’re air launching a multi-engine liquid rocket, it seems to me that either you ignite the engines while it’s still attached to the carrier aircraft (which raises all sorts of other issues!) or, you drop then ignite. If the latter, aren’t you massively increasing mission risk over a pad launch? What I mean is, if you drop the rocket and have some sort of problem that prevents full ignition (and for ground launch would result in a pad abort), it seems to me it would be, shall we say, problematic?

    Am I missing something, or does this factor weigh against air-launching a liquid rocket?

    And as a second point, is air launching really any better from a recovery POV than, say, landing the first stage on a large cheap barge? Sure, unlike boostback, it’s not back at your launch site, but it’s still recovered at very low cost (and you don’t need a massive launch aircraft).

      1. [i]How is that different from a standard stage separation and start?[/i]

        I was thinking complexity. On the F9, for example, there’s one upper stage engine. On the 1st stage, there are 9. My guess (I’m not an expert) is that the more engines you have, the greater the chances for a problem on startup. With an F9, that means a pad abort, which isn’t a big deal. But, if it was being air launched and ran into a problem with ignition, it would be a far bigger deal.

        I do see your point though; air start is something that’s done all the time with staging, so this isn’t as big an issue as I’d first thought.

Comments are closed.