Single Stage to Orbit

[Nibb31]

The current rocket is the equivalent of throwing your car away once you get to work and buying a new car. SSTO is the system you hopefully use when dealing with cars

Yeah, but you are confusing "reusable" and "SSTO". A 100% reusable spaceplane can be multiple stages. A reusable VTOL rocket can be multiple stages. An SSTO launcher can be expendable.

If your goal is to have reusable hardware, it doesn't necessarily involve SSTO. In fact, if you want any usable payload fraction, it's probably going to require multiple reusable stages.

[jwenting]

2-3 reusable stages are much more efficient, thank you very much.

[KASASpace]

Yes, this I considered as well, and maybe a ramjet first stage using solids to get faster than mach 0.5, so they are pretty efficient. And it would really be a plane, so, recoverable.

[K^2]

Using planes for first stage doesn't work as great as you'd think. But if you have the first stage do a vertical climb, and then glide to landing, that might work. Ramjet can be significantly lighter than a turbojet, which would give it sufficient TWR to pull that off.

So yeah, this might be the best possible setup. First stage is solid-boosted (sc)ram jet, followed by a small LH2/LOX rocket plane that actually makes orbit. Scrams can get up to Mach 12, at least, and will eat up a big chunk of gravity and drag losses. That means second stage only has to pull about 4km/s of dV, which is very reasonable with cryogenic LH2/LOX. Everything but the boosters is fully reusable. Boosters can be partially reusable, like Shuttle's SRBs. Or if it proves economically sensible, perhaps the boosters can be reuseable liquid fuel rockets that work similar to SpaceX's grasshopper concept.

Takeoff would be entirely vertical, like a conventional rocket, and the two stages would land like planes.

So this leaves two main questions. Should the second stage be a pure LH2/LOX? Or should it bring a more stable fuel for orbital operations, a la Shuttle Orbiter. Should the boosters be solid rockets non/partially reusable? Or fully reusable liquid rockets?

I would lean towards using a MMH/UDMH/Hydrazine to be burned with nitrogen tetroxide for orbital operations, which can also be used as monoprop for maneuvering thrusters. So, basically, same as Orbiter. And I think SRBs would make things easier, and in the long run, it's probably a cheaper option anyways. But I can be easily persuaded otherwise.

I think I'll do some estimates on how big each of the stages is going to be, and maybe get some idea of costs.

Edited December 21, 2013 by K^2

[KASASpace]

Hmm, I think we may have formulated a good idea, ramjet/solid first stage, and then a second stage to only get a few km/s! Both land like planes, so almost complete re-usability.

[Horn Brain]

This is a cool idea, not unlike some stuff that the Australians are doing right now. However, you guys are basically completely overlooking the fact that it's very difficult to make an experimental scramjet, let alone one with the variable-geometry intakes, combustion chambers, and nozzles that would be required to have it function at low speeds initially. You basically have to design a transition from a converging-diverging-converging-diverging flow pattern to a converging-diverging flow pattern, with each intermediate position of the internal geometry being a problem as difficult as designing an entirely new vehicle.

It is very difficult to design these vehicles because at the speeds they're operating at, burning the fuel only produces about 10% as much energy as there is in the incoming flow (due to kinetic energy from the high speed). This means that drag reduction is as, if not more, important than thrust maximization. Therefore, you're going to have to find some way to inject, mix, and burn the fuel with almost ideal efficiency over a very wide range of speeds, using little to no redundant hardware (multiple parts which do the same job at different speeds), while the geometry of the vehicle drastically changes. This is why most of these ideas end up using rockets to get a pure scramjet up to speeds where it can operate with constant geometry, to avoid multiplying the difficulty of the problem by a factor of thousands (multiple geometries which may require different schemes for fuel injection and ignition).

[K^2]

However, you guys are basically completely overlooking

I'm brushing most of it aside in the conversation, but I'm definitely not overlooking it. There are some serious challenges there. There is, however, one very important thing to note. The (sc)ram jet stage would run on Jet-A. It needs to be efficient enough to get the job done, but if it's not very fuel efficient in absolute terms, that's not a big loss. A scram jet, however advanced, is nowhere near complexity of regular, run of the mill bipropellant liquid rocket. We've been building the later for a bit longer, so we have a lot of research to fall on, but once you figure out a good design, a scram jet is way cheaper to build and maintain. This will make a huge difference in cost to launch even with a reusable craft. So if we have to put a bit more kerosene in it to make it work, it's worth it.

variable-geometry intakes, combustion chambers, and nozzles

Here is my thinking on that, partially inspired by the Blackbird engines. Lets start with the easy part. I would use aerospike concept for nozzle. At low speeds, you don't need divergent part of the nozzle, so this will work fine. And at high speeds, aerospike should do exactly what you need in terms of the outgoing flow.

On to the hard part. I think the best way to handle the fact that low speed ram jet will need a diverging combustion chamber is to do it the way SR-71 solves similar problem. Place the ram jet combustion chamber inside a toroidal scramjet. In ram jet mode, the bypass is locked off, taking air through the diverging chamber. Once sufficient speeds for scramjet operation are reached, the central chamber is locked and bypass is opened. Bypass will narrow as it goes around the chamber, and give you the right flow for scramjet operation. This does involve moving parts, some expensive materials, and will require inspection and replacement. It's still way simpler than a turbojet, though. So it should be fine.

This leaves a few problems. First, the ram jet's combustion chamber has to be designed for specific speed, and will not be efficient otherwise. I say, forget that. At low speeds, we can take a hit of being not terribly efficient. Design the chamber to provide sufficient thrust after boosters are done, and otherwise, to give best overall efficiency, and call it the best we can do. Trying to actually vary geometry of the combustion chamber is not going to be worth it, unless somebody comes up with some pure genius idea similar to aerospike for that. I've got nothing though, so I just hope this is good enough.

Same problem with scramjet, technically, and we need great range on that. The scram bypass will be activated around Mach 3-4, and we need it to last to Mach 12 or so. Fortunately, the biggest problem in getting a speed-variable scram jet is the nozzle. As I've said, I'm betting on aerospike to solve that problem. Geometry of the combustion zone is also important, however. It might not be worth the trouble, but shifting the region where fuel is injected might work to improve that. Overall, it's a place where a performance hit will have to be taken. If nothing else can be done, design for best performance at high speeds and thrust at low speeds.

Overall, though, we have already improved on bipropellant fuel dramatically by using a scramjet, so if we have to take some performance hits, that's just life. With that in mind, lets go to intakes. Again, I'd turn to existing solutions here. The overall shape of the intake would have to be just the best we can do across the board, but we can have a moving spike without making things too complex. It worked for SR-71 and it works for Skylon. Hopefully, that's enough for this (sc)ram jet as well.

The rest is lots of trial and error. Fortunately, cylindrical symmetry of this type of engine would allow for very precise computer modeling, so you wouldn't have to build expensive prototypes until you have a good general idea of what works and what does not. The final build will have some moving parts, and is going to have to be manufactured with very high standards of precision, but it's still going to be way easier to build lots of these than anything else we could use to take a ship to orbit.

As I've indicated earlier in the thread. My interest would be in taking a small orbiter, with capacity under 10T, to LEO with the best cost-per-kg we can manage. Discounting methods that involve megastructures, like launch loop and tethers, I really think this is the best option. It would require a lot of research and testing, but it should be very affordable once the technology is in use.

[Aghanim]

P.S. While I still think maglev to 300m/s would be too expensive to be worth it, I'd definitely look into pneumatic catapults, similar to these used on aircraft carriers. Nothing new to invent, hardware exists, and it can get you to a pretty good fraction of that starting speed.

Well that is kinda funny, as currently the navies are considering to convert their steam catapults to electromagnetic catapults, and it's a standard feature in new aircraft carrier : https://en.wikipedia.org/wiki/Electromagnetic_Aircraft_Launch_System

But I do agree that 300 m/s maglev is overkill, the air resistance at that speed is kinda significant, right?

And linac? That is a weird word choice, as it usually means a linear particle accelerator, and linear motor have low maintenance either

Edited December 21, 2013 by Aghanim

[KASASpace]

But I do agree that 300 m/s maglev is overkill, the air resistance at that speed is kinda significant, right?

Well, do you know any craft that go less than supersonic at sea level, hmm? I sure do. And it could be at somewhat high altitudes, like 10,000 feet.

[Aghanim]

Well, do you know any craft that go less than supersonic at sea level, hmm? I sure do. And it could be at somewhat high altitudes, like 10,000 feet.

Its experiencing transonic drag, wave drag and all of that weird stuff, which give rise to sound barrier

If you want less drag, just go supersonic directly

Again from Wikipedia:

Supersonic aerodynamics are simpler than subsonic because the airsheets at different points along the plane often can't affect each other. Supersonic jets and rocket vehicles require several times greater thrust to push through the extra drag experienced within the transonic region (around Mach 0.85–1.2). At these speeds aerospace engineers can gently guide air around the fuselage of the aircraft without producing new shock waves but any change in cross sectional area further down the vehicle leads to shock waves along the body. Designers use the Supersonic area rule and the Whitcomb area rule to minimize sudden changes in size.

Its either go slow or go fast, you couldn't go in the middle. So weird

BUT..... What I forgot to account for is there are engines that could push it though the sound barrier

Edited December 22, 2013 by Aghanim