One of the principles of the program discussed here is that freight is delivered to the Earth station in appropriate sized packets. More detailed studies can determine this size; here a figure of 20 tons will be used. The figure is reduced in an integrated program, since hardware is re-used. The number of flights per year can be varied, as well as the number of copies of the booster in operation. The program can be designed so that freight is delivered to the station. There might however be situations where some other orbit from Earth launch would be preferred; this could be accommodated.
Other principles which guide the design of the booster are safety, cost, and integration into the overall program. It seems certain that the best design is an unmanned two stage booster with oceanic recovery. The second stage docks at the station; then reenters and lands in the ocean. The first stage lands downrange in the ocean.
Such a booster can be built in a short time, indeed with some existing components. The takeoff weight might be on the order of 500 tons. There is a possible program where the takeoff weight is lower. Takeoff weight reduction would result in reduction in scale and cost of operations; and increased options for design of more exotic configurations such as a large object vehicle.
Lower takeoff weight might be achieved by using ramjets for a portion of the flight, reducing the amount of on-board oxidizer. This question should be given adequate consideration before beginning construction of a new booster. A seemingly viable configuration uses rockets to take off and reach start-up velocity for variable geometry subsonic ramjets, and the subsonic ramjets to the end of their velocity range. At this point the first stage separates. The second stage uses rockets to achieve orbit. This configuration might be called "2+1", giving the number of cycles in each of the two stages. Each cycle uses its own engines, on the principle that multi-cycle engines don't reduce the weight sufficiently to justify their added complexity. Tanks, pumps, etc., are shared for the two cycles of the first stage. Some new hardware such as switching valves is required. Various refinements should be investigated, including "blending" the cycles by gradually switching, and using the onboard oxidizer to supplement the atmospheric oxidizer.
At one time the author considered using second stage supersonic ramjets to the end of their velocity range. This seems less likely to be advantageous than first stage ramjets, especially if the second stage cutoff can't be pushed much past Mach 12. Mach 8 has been mentioned as a possible first stage cutoff velocity, and if this is achievable second stage ramjets would almost certainly be of little value. There has been research on uasing both types of ramjets. The 2+1 configuration seems to be more workable (and cost effective) than SSTO's, aerospace planes, etc., and it seems quite likely that the ramjet engineering to determine the takeoff weight would be routine.
To provide the capability for a variety of vehicles, a vehicle has a "multiplicity''; this is the number of first stage rocket engines, first stage ramjets, etc. The weight overhead of this approach should be entirely tolerable, especially in view of the reduced weight. The payload per unit is a parameter which should be carefully studied. A value of 5 tons makes 10 tons the smallest payload, with 40 tons a realistic payload.
The on-board oxidizer for both stages should undoubtedly be liquid oxygen. The fuel might be liquid hydrogen throughout. Higher density is said to be preferred in the first stage of conventional boosters, because the lower tank weight offsets the higher exhaust velocity. This might not apply to a "2+1" booster, because the ramjets (which replace a greater weight of oxidizer) greatly increase the dry weight.
The Earth launch vehicles considered here are all "2+1" variants. They can be refurbished in a hanger (or hangers) at Cape Canaveral, with a small number of launch pads.
The issue arises of what engines to use for the rockets. There is a great deal of research underway on rocket engines. Those required for a 2+1 are comparatively low thrust, and oxy-hydrogen throughout should be considered. In a recoverable ramjet booster operational cost and lifetime are considerations of equal standing with exhaust velocity. A higher exhaust velocity in the 3th cycle might result in further takeoff weight reduction. As usual, any mass increase for the exhaust velocity increase must result in an overall reduction in initial weight. Finally, increased exhaust velocity is desirable in rocket engines to be used in Mars crew flights; here reusability is not an issue.