Best propulsion method for a "low cost" SSTO?

[Exosphere]

One of the most important factors in the design of a rocket is the choice of its propulsion system. Although all surface-to-orbit rockets today use chemical engines, there have been some proposals to use LANTR (LOx-augmented NTR) or closed-cycle gas core NTR-powered launchers to more effectively launch payloads into orbit.

While these propulsion methods would certainly lower the required fuel fraction necessary for a rocket to achieve the delta-v required to reach orbit, there is one problem -- they use nuclear reactors, which tend to be maintenance intensive. Since maintenance is one of the biggest factors in launch vehicle cost, wouldn't it make more sense just to use a chemical rocket-powered SSTO? Obviously, chemical engines aren't exactly maintenance-free themselves, but would the omission of the nuclear reactor improve the maintainability of a SSTO enough to offset the greater cost of the increased fuel fraction? As an aside, I recognize that fuel itself is a relatively small factor in launch vehicle cost, but that the secondary effects of a high fuel fraction (a more fragile rocket which is more difficult to reuse and requires more inspection/maintenance in between launches if it is reusable -- i.e. the Space Shuttle) are more costly.

[Agamemnon]

what about sabre?

atm the best shot, i guess

[Exosphere]

what about sabre?

atm the best shot, i guess

I was thinking more along the lines of VTVL rockets, but as far as space planes go... this is all my opinion, and I could be wrong, but it would seem that space planes have significant disadvantages compared to rockets. The wings are less efficient than a simple tube for carrying fuel, the extra drag of wings/tail(s)/control surfaces would also increase the required delta-v to get into orbit, a complex winged body seems more difficult (and thus expensive) to design than a simpler rocket (especially because space planes are usually designed to take a more shallow trajectory, and thus spend more time in atmosphere, where aerodynamic stresses could create issues), the shallower trajectory seems less efficient than a rocket's (again, more delta-v), and the use of multiple kinds of engines (either on separate aircraft (SpaceShip One), separate locations on the same plane (i.e. a space plane with a set of turbojets and a set of rockets), or combined into a single "engine" (i.e. Skylon)) would seem to increase complexity, and thus costs. Also, with a multi-vehicle design (SpaceShip One), the use of two different vehicles would probably increase launch costs and require a larger crew to control both craft (again, this would lead to greater costs), and could increase the cost of infrastructure. Finally, most space plane designs I've seen have lower payloads to LEO than rockets (especially compared to beasts such as the Falcon Heavy), which would seem to prevent them from realizing economies of scale as effectively as heavy-lift vehicles.

Of course, this is just what seems to be the case from my point of view. To me, space planes would seem to be less efficient, more costly, and more complex than rockets. Of course, there are several rocket designers with much more aerospace knowledge than me looking at building space planes, so I'm not so sure I'm correct. Am I missing some crucial advantage space planes have (besides their coolness factor, of course )?

Edited April 16, 2014 by Exosphere

[shynung]

SSTO spaceplanes have an advantage over VTVL rockets in that their engines need not to carry the weight of the whole craft (that is taken care by the wings), eliminating one obstacle in rocketry: gravity drag. This enables the usage of engines with TWR of less than 1, but far more efficient. Plus, VTVL rockets need a horizontal component to their velocity to reach orbital speed (hence the gravity turn), otherwise they would fall back to Terra. Spaceplanes have this horizontal component in the first place, which reduces the delta-v needed for orbital insertion later on.

For engines, I think air-augmented rockets are suitable to the task. It's less efficient than something like SABRE, but much simpler in engineering.

Edited April 16, 2014 by shynung

[Armchair Rocket Scientist]

One of the most important factors in the design of a rocket is the choice of its propulsion system. Although all surface-to-orbit rockets today use chemical engines, there have been some proposals to use LANTR (LOx-augmented NTR) or closed-cycle gas core NTR-powered launchers to more effectively launch payloads into orbit.

While these propulsion methods would certainly lower the required fuel fraction necessary for a rocket to achieve the delta-v required to reach orbit, there is one problem -- they use nuclear reactors, which tend to be maintenance intensive. Since maintenance is one of the biggest factors in launch vehicle cost, wouldn't it make more sense just to use a chemical rocket-powered SSTO? Obviously, chemical engines aren't exactly maintenance-free themselves, but would the omission of the nuclear reactor improve the maintainability of a SSTO enough to offset the greater cost of the increased fuel fraction? As an aside, I recognize that fuel itself is a relatively small factor in launch vehicle cost, but that the secondary effects of a high fuel fraction (a more fragile rocket which is more difficult to reuse and requires more inspection/maintenance in between launches if it is reusable -- i.e. the Space Shuttle) are more costly.

A LANTR could probably be used for an SSTO, or at least with liquid-oxygen drop tanks, and get reusability with a payload fraction comparable to a chemical rocket, but this is heavily dependent on getting really high TWR engines. The problem is, as you said, maintenance and infrastructure will be the most important factor in a reusable launch vehicle. The thing is, currently getting permission to use an NTR in space would be extremely difficult. Using one in atmo, particularly one that had already been through the stresses of launch and reentry, would probably require the vehicle's operator to take the entire engine a part and inspect every piece with a microscope between flights, if it was allowed at all.

Gas-core reactors? I assume the proposal you're referring to is this: http://www.projectrho.com/public_html/rocket/surfaceorbit.php#libertyship

Yeah, a 10%+ payload fraction is really attractive, but it suffers from all the problems of a solid core NTR, with the additional complication that we can't actually build one with today's technology. It's inaccurate to compare it to chemical rockets; such a design would most likely be competing with space elevators, railguns, or beamed power.

A chemical-rocket-only SSTO is certainly possible; many first stages could in theory be turned into one. The problem is, the payload fractions would be awful, and the vehicle would still be expendable, so a multi-stage rocket ends up actually being cheaper.

A reusable rocket-only SSTO is either impossible or would have such a low payload fraction that despite its simplicity, air-breathing SSTOs and multi-stage reusable rockets would mop the floor with it. Consider the following: with an ISP of 465 (near the upper limit for hydrogen) and a dV of 10 km/s (that extra 500 m/s is for landing and orbital manuevers, assuming you use parachutes or wings and therefore only need a few tens of m/s for landing) needs to have a wet:dry mass ratio of 9. The space shuttle orbiter and external tank have a wet:dry ratio of only about 6.4, and that's with the fuel tank being thrown away. Using dense fuels would make the vehicle slightly less painful to heatshield, but your mass ratio requirements become higher: for an ISP of 340 s, about what you'd get for kerolox using high-performance engines, you would need a mass ratio of 20! The Falcon 9 v1.1 first stage barely beats that, and it's only designed to survive reentry from less than half orbital speed. Basically, it ain't happening.

[Nibb31]

One of the most important factors in the design of a rocket is the choice of its propulsion system. Although all surface-to-orbit rockets today use chemical engines, there have been some proposals to use LANTR (LOx-augmented NTR) or closed-cycle gas core NTR-powered launchers to more effectively launch payloads into orbit.

Do you have any pointers for SSTO proposals that used nuclear power? All of the half-serious SSTO proposals that I've seen have used chemical propulsion.

I don't think there is future for nuclear propulsion in atmospheric flight, due to all sorts of reasons, including safety, proliferation, and politics. I wouldn't rule them out for upper-stages, space-tugs or interplanetary applications though. That's what past NTR studies were going to use them for.

So any future reusable SSTO would most definitely use chemical propulsion. Chemical SSTO is actually pretty easy. The original Atlas-Mercury was effectively an SSTO. The Titan I first stage could also have reached orbit without any payload, but what use is that?

The problem with SSTO is that:

a - You want a payload.

b - You need a specific amount of energy to accelerate that payload from 0 to 27000 km/h.

c - You need a specific volume of chemical propellant to provide that energy.

d - You need a structure to safely carry that propellant and engines to convert the chemicals into usable thrust.

e (optional) - You need additional hardware to re-enter and land safely.

The hard part is to combine a+b+c+d+e into a workable vehicle. b and c are pretty much fixed by the laws of physics and volumetric efficiency, and there is only limited improvement available in energetic efficiency. Therefore, the only way to achieve SSTO is to either reduce either a or d or to go without e. However, a is your payload, and e is the only reason you are going with SSTO, because if your vehicle was disposable, you wouldn't care about the number of stages in the first place.

Therefore, your main work should concentrate on making d as light as possible. Unfortunately, lightweight structures and engines don't usually go together with the sort of operational reliability and serviceability that you would want out of a fast-turnaround reusable vehicle. For example, a reusable vehicle needs hatches, attachment points, connectors and structural interfaces to easily replace and service parts, which add weight and complexity. It also needs a more rigid structure to deal with both launch loads, re-entry loads and landing loads, which are all different forces that act on different parts of the structure. These additions can rapidly negate any gains that you might have made from using lighter materials.

That's why reusable SSTO is such a tough nut to crack.

Edited April 16, 2014 by Nibb31

[Rakaydos]

A Magnohydrodynamic thruster that uses bow shock plasma as reaction mass.

Edited April 16, 2014 by Rakaydos

[Exosphere]

Do you have any pointers for SSTO proposals that used nuclear power? All of the half-serious SSTO proposals that I've seen have used chemical propulsion.

I don't think there is future for nuclear propulsion in atmospheric flight, due to all sorts of reasons, including safety, proliferation, and politics. I wouldn't rule them out for upper-stages, space-tugs or interplanetary applications though. That's what past NTR studies were going to use them for.

So any future reusable SSTO would most definitely use chemical propulsion. Chemical SSTO is actually pretty easy. The original Atlas-Mercury was effectively an SSTO. The Titan I first stage could also have reached orbit without any payload, but what use is that?

The problem with SSTO is that:

a - You want a payload.

b - You need a specific amount of energy to accelerate that payload from 0 to 27000 km/h.

c - You need a specific volume of chemical propellant to provide that energy.

d - You need a structure to safely carry that propellant and engines to convert the chemicals into usable thrust.

e (optional) - You need additional hardware to re-enter and land safely.

The hard part is to combine a+b+c+d+e into a workable vehicle. b and c are pretty much fixed by the laws of physics and volumetric efficiency, and there is only limited improvement available in energetic efficiency. Therefore, the only way to achieve SSTO is to either reduce either a or d or to go without e. However, a is your payload, and e is the only reason you are going with SSTO, because if your vehicle was disposable, you wouldn't care about the number of stages in the first place.

Therefore, your main work should concentrate on making d as light as possible. Unfortunately, lightweight structures and engines don't usually go together with the sort of operational reliability and serviceability that you would want out of a fast-turnaround reusable vehicle. For example, a reusable vehicle needs hatches, attachment points, connectors and structural interfaces to easily replace and service parts, which add weight and complexity. It also needs a more rigid structure to deal with both launch loads, re-entry loads and landing loads, which are all different forces that act on different parts of the structure. These additions can rapidly negate any gains that you might have made from using lighter materials.

That's why reusable SSTO is such a tough nut to crack.

Since the main problem behind reusable SSTOs seems to be payload weight, if you were only using the design for a crew-transfer vehicle (say to transfer a crew to an orbital station, from which they would depart for somewhere else), would that make the design easier? The Dragon capsule holds seven people and weighs six metric tons, which would make it a lot easier to launch on a SSTO than a massive fifty metric ton payload that you'd launch using a Falcon Heavy.

[bsalis]

Anything fancy is not going to be cheap.

Cheap is simple with modest goals. Something that was posted around here a while ago was this...

http://en.wikipedia.org/wiki/Aquarius_(rocket)

Target launch cost was one million each.

[Nibb31]

Since the main problem behind reusable SSTOs seems to be payload weight, if you were only using the design for a crew-transfer vehicle (say to transfer a crew to an orbital station, from which they would depart for somewhere else), would that make the design easier? The Dragon capsule holds seven people and weighs six metric tons, which would make it a lot easier to launch on a SSTO than a massive fifty metric ton payload that you'd launch using a Falcon Heavy.

Sure. Now look at what it takes to put those 6 tons into orbit: a two-stage Falcon 9 that weighs 500 tons.

To make the Falcon 9 an SSTO design, you would have to remove the upper-stage, which means that the first stage has to burn longer, which means more propellant, more thrust, therefore more engines and more weight... In the end, your SSTO rocket would have to be significantly bigger than a Falcon 9, which is why SpaceX (and most other rocket companies) went for a two-stage design.

It also has to re-enter from orbit, and designing a heatshield for an entire rocket is a major challenge. You are going to bring that heatshield and some sort of landing capability, whether wings or propulsive landing, all the way to orbit. All that stuff is going to weigh much more than your 6 ton payload, which means that the rocket needs to be even bigger.

Edited April 16, 2014 by Nibb31

[SargeRho]

Depends entirely on what you want to put into orbit.

If you just want to put a small sattelite into orbit, a smallish rocket will do.

For medium lift, I'd say Aerospikes, see Venture Star/X-33

And for heavy lift, it doesn't make sense to build SSTOs.

[shynung]

Anything fancy is not going to be cheap.

Cheap is simple with modest goals. Something that was posted around here a while ago was this...

http://en.wikipedia.org/wiki/Aquarius_(rocket)

Target launch cost was one million each.

Yeah... I remember that one.

[Idobox]

Simple, cheap SSTO? ram accelerator.

Basically, you build a steel tube at an angle on top of a mountain, fill it with an explosive mix of gases, and use a biconic ship (two cones). The thing behaves like a ramjet, but it doesn't have to carry its own fuel. And as it accelerates, it goes through different mixes at different pressures, keeping it efficient, and eventually working as a scramjet. If you leave the ground at 6km/s, it doesn't take much to reach orbit.

Of course, flying at 6km/s in the lower atmosphere has a few drawbacks. Like massive drag, noise and compression heating.

[shynung]

Simple, cheap SSTO? ram accelerator.

Basically, you build a steel tube at an angle on top of a mountain, fill it with an explosive mix of gases, and use a biconic ship (two cones). The thing behaves like a ramjet, but it doesn't have to carry its own fuel. And as it accelerates, it goes through different mixes at different pressures, keeping it efficient, and eventually working as a scramjet. If you leave the ground at 6km/s, it doesn't take much to reach orbit.

Of course, flying at 6km/s in the lower atmosphere has a few drawbacks. Like massive drag, noise and compression heating.

That system requires non-standard infrastructure (launch tube, rather than launch pad/runway) built on hard-to-reach locale (mountain tops). The orbiter vehicle might be cheap, but the investments for the launch facility would eventually drive launch costs upward. Same goes for mass-driver launches.

[Idobox]

That system requires non-standard infrastructure (launch tube, rather than launch pad/runway) built on hard-to-reach locale (mountain tops). The orbiter vehicle might be cheap, but the investments for the launch facility would eventually drive launch costs upward. Same goes for mass-driver launches.

The infrastructure is basically a steel tube. It would probably cost less than a single SLS launch.

[shynung]

The infrastructure is basically a steel tube. It would probably cost less than a single SLS launch.

Of what length? Surely 1km would suffice?

Other than that, the gas mix needs to be maintained at a safe-but-usable level, needing another subsystem. Also, one needs sensors along the tube to monitor for faults, and massive struts to keep it from wobbling around and destabilizing the orbiter at launch. Also important is a VAB nearby, to avoid having to haul a fully-assembled-fueled-and-loaded orbiter up a mountain.

You could say that the end result would be lower single-launch cost per ton. However, needing a brand new facility using newly-invented technology would be very costly, and would only be considered by those whose launch far more frequently than others. It's like a train; while the end vehicle is highly efficient, significant infrastructures need to be laid down first.

[DaveofDefeat]

Meh, making your crafts reusable is the biggest thing that will reduce costs, not SSTO.

[shynung]

Meh, making your crafts reusable is the biggest thing that will reduce costs, not SSTO.

This was the mindset for the Space Shuttle, and that thing ended up being more expensive than simple rockets.

[Idobox]

Of what length? Surely 1km would suffice?

Other than that, the gas mix needs to be maintained at a safe-but-usable level, needing another subsystem. Also, one needs sensors along the tube to monitor for faults, and massive struts to keep it from wobbling around and destabilizing the orbiter at launch. Also important is a VAB nearby, to avoid having to haul a fully-assembled-fueled-and-loaded orbiter up a mountain.

You could say that the end result would be lower single-launch cost per ton. However, needing a brand new facility using newly-invented technology would be very costly, and would only be considered by those whose launch far more frequently than others. It's like a train; while the end vehicle is highly efficient, significant infrastructures need to be laid down first.

For satellites, you could do with something between 500 and 2000m long (electronics don't mind 1000g accelerations)

The gas would be mixed just before launch, and needs to stay stable for less than 4s, and since it's going to explode anyway, you just need a thick tube.

The vehicles would be 1 or 2m in diameter, about the same length, you could bring them on a truck.

The main difficulty is that you need different mixes in different sections, so you will need some form of door or valve to separate them, that open very quickly when the ship is near. You also need to accelerate the ship to Mach .5 before it works as a ramjet, so you have to add that too.

Whatever system you choose, you will need large capital investment to develop the technology for SSTO. The ram accelerator is not new technology, it's already used for artillery shells and high velocity impact research. Of course, building something significantly larger than what already exists is difficult, but in the end, it's a steel tube with doors and some low tech plumbing and an oversized bombshell.

[Red Iron Crown]

Mass production and economies of scale are what are needed to drive costs down. A program launching once a week is going to drive down the cost per kg to orbit, no matter whether it is a conventional disposable rocket, reusable multistaged craft, or SSTO. Just spreading the R&D and infrastructure costs over many flights will be significant.

[Nibb31]

Meh, making your crafts reusable is the biggest thing that will reduce costs, not SSTO.

You are assuming that the major cost factor in the space industry is the hardware. It isn't. It's the people. Reusable launchers don't necessarily require less people to operate. On the contrary, because you are reusing your rockets, you need to build less of them, which makes each unit more expensive. You might need less people at the factory, but you need more people to handle and refurbish the reused rockets. So any actual cost reduction is marginal.

Reusable only makes sense when launches are frequent enough to justify reusing the hardware. This means that there must be market demand for frequent launches, which there isn't. Even if you manage to slash launch prices by 50%, you might increase demand somewhat, but you won't double the number of customers.

Edited April 16, 2014 by Nibb31

[shynung]

The main difficulty is that you need different mixes in different sections, so you will need some form of door or valve to separate them, that open very quickly when the ship is near. You also need to accelerate the ship to Mach .5 before it works as a ramjet, so you have to add that too.

Whatever system you choose, you will need large capital investment to develop the technology for SSTO. The ram accelerator is not new technology, it's already used for artillery shells and high velocity impact research. Of course, building something significantly larger than what already exists is difficult, but in the end, it's a steel tube with doors and some low tech plumbing and an oversized bombshell.

This reminds me of Jules Verne's work about going to the Moon via space gun.

If the gases are to be confined to a single segment of the tube, be burned by the passing vehicle, then disappear, why use doors? Membranes could do the job; the vehicle will penetrate it as it goes from one mix to the next.

Also, the vehicle would need some fuel for orbital insertion. Combining a ramjet with a rocket engine gives you an air-augmented rocket, which works as a ramjet while still in the tube, and burns fuel in atmosphere to combat atmospheric drag. After that, do circularization burn, and you're done. Now, we have to somehow cram all those, plus fuel and payload, in a 1m-diameter vehicle short enough to be carried by a truck that can climb mountains. But I'll leave that to our ingenious engineers.

[Aghanim]

Chicken and the egg problem: You couldn't start decreasing launch costs without more demand, and demand is limited because of high launch costs. Now how do we solve this?

[Rakaydos]

Chicken and the egg problem: You couldn't start decreasing launch costs without more demand, and demand is limited because of high launch costs. Now how do we solve this?

The way SpaceX is doing it- Clustering smaller engines. Because you need more engines per launch, you start to reap the benifits of mass production with a lower total number of launches.

[cantab]

There's no reason to insist it be SSTO. What matters is that it's cheap. Even if you accept that the only way to make it cheap will be to make the vehicle reusable, multistage reusable designs are entirely possible.