The most plausible reusable SSTO concepts are the Phoenix-E (“Improved Phoenix”) and associated Delta Clipper, and the Chrysler Aerospace SERV proposal. The Phoenix was a base-reentry with propulsive vertical landing using a single plug nozzle; DC had a nose-forward reentry to achieve the desired crossrange with a propulsive vertical landing and (baseline) used conventional de Laval nozzles. Both would have had only modest cargo capacity at best, and required extremely lightweight structure (titanium honeycomb for the Phoenix, advanced graphite composite for the DC).
The SERV, intended to meet the Space Transportation System criteria (except for crossrange), was a much larger vehicle capable of Shuttle-sized payloads (11 MT, although in certain configurations it was estimated to carry as much a 50-57 MT to certain orbits). It would have used an modular integrated base plug aerospike (essentially using the entire base as an aerospike surface) and had base-forward reentry with air breathing engines specifically for terminal descent and landing. Structure would have been standard aluminum composite honeycomb and ablative base end heat shield which would cool for both ascent and descent. The SERV proposal in particular was studied by the Aerospace Corporation which found no major problems with the study other than the requirement for the novel base plug aerospike. For crewed flights it would have carried a separate lifting body shuttle either on top or in the interior bay which would return separately in a horizonal landing flight mode. In practice, no one has yet flown an aerospike ascent engine on a space launch class vehicle and some of the physics of aerospkes (especially base plug spikes) is still not fully characterized, but it is demonstratably possible to build an SSTO in concept. Whether it would be genuinely robust enough for economic reuse, and whether it could be made to carry a useful amount of cargo is a question that can only be answered by further development.
Making a partially reusable two stage vehicle should be possible, but making it fiscally viable would mean reducing the amount of checkout and refurbishment activity, which itself means accepting more mass (in the first stage) or substantially lower performance (to reduce loads and environments). Various “Big Dumb Booster” concepts have been proposed by contractors, some with limited reusability, but one of the most promising was the Bob Truax designed Sea Dragon, a massive pressure fed rocket built to shipyard tolerances and launcheddirectly from the ocean using ballast erection; it just gets towed out to broad ocean area and fueled with no launch platform or support structure. The empty first stage would return downrange using ballute deceleration and landing in the ocean, and the robust combustion chamber required little more processing than washing out and applying corrosion inhibitor between flights. The basic premise was demonstrated through proof-of-concept testing with a smaller system (“Sea Bee”) but the need to launch the massive bulk cargo at rates of 12 to 20 flights a year never emerged, and the concept has languished. TRW studied this concept and concluded that it would have been fiscally and technically viable.
I have a personal “toy model” simulation of a two stage reusable/repurposable vehicle that I’ve played with over the last decade or so. While it isn’t intended to represent an actual launch vehicle configuration that I would propose, it is useful for testing different technologies and parameters suitable for low cost, high volume flight. The focus is largely on simplicity over redundancy and high performance, and my evaluation of baseline reliability suggests that this would give launch to orbit costs around $1000/kg with conventional propulsion systems, and <$500/kg with more advanced propulsion (base plug aerospike) using single stage low pressure/high volume propellant feed or augmented (autogeneous) pressure fed using mixed fuel propellants. One of the keys I’ve found is abandoning the standard long, narrow cylinder shape in favor of a squat shape with either base reentry or lifting body/ballute deceleration with a water landing or glide/skid landing profile (non-propulsive landing), which allows an improvement in mass ratio by geometry rather than through the use of exotic materials. The upper stage uses high performance vacuum thrust engine (generally a conventional de Laval with shock augmentation) and is intended for either repurposing (e.g. a “wet workshop” module for a space station) or in-situ recycling for use in constructing other structures once its propulsive mission is complete.
I’ve even gotten this system to fly to orbit–albeit with dramatically reduced cargo–using propane or DME instead of RP-1 or LH[SUB]2[/SUB] as the propellant, which suggests there is a lot of trade space to develop better and cleaner vehicles by trading performance for simplicity. The use of expendable tankage (drop tanks) or thrust augmentation (solid propellent boosters for rapid thrust development to minimize gravity drag at low altitude) scales pretty well with payload capacity, although consideration for the added complexity and reduced reliabliity has to be factored in. If you assume advanced propulsion systems (continuous wave detonation engines or full flow nuclear thermal upper stage) the performance can be dramatically improved or more renewable fuels like DME can be used, although those technologies have a considerable way to go before demonstrating a real proof-of-concept on the scale needed for this application.
The long and short is that we can do a lot better than the conventional two and three stage cylindrical space launch vehicles without having to invoke highly complex and likely unworkable concepts like the Skylon. However, any new concepts would require a signiifcant investment in not only basic development but flight testing at various scales to get the experience necessary to develop a production-grade launch system, and for the most part there just doesn’t currently exist the demand to develop these on a commercial basis, and the US (and other) military is fairly wedded to traditional concepts to a point of excluding anything that looks too “funny” out of a (rational) fear of failures during development and early deployment.
In my personal opinion, the opening up of the small satellite launch market offers a great opportunity to develop this kind of launch system technology for lower criticality payloads and figure out how to control logistical and NRE costs which can then be scaled up to heavy lift and crewed launch vehicles. However, there is a chicken and egg problem there as well; payloaders build small satellites but are waiting for a cheap launch system in order to sell services economically; the launch industry is waiting for a critical mass of small sat payloaders before investing in developing dedicated systems. The current rideshare offerings for smallsats are limiting what they can do, but they don’t individually have enough impetus (yet) to alter the landscape of launch vehicle developers or providers. I’m hoping that this will change in the next ten years or so and that some of the nascent smallsat vehicle developers like RocketLabs, Firefly, and Generation Orbit will develop fiscally viable and reliable launch vehicles. Unfortuantely, failures like the recent Super Strypi failure (rail-launched spin-stab solid propellant vehicle) tend to dampen the spirit of investors even though such failures are just a natural consequence of trying out new methods, processes, and technologies to see what works. If you’re never failing, you’re never really trying to push the envelope.
Stranger