The case for smaller launch vehicles in human space exploration (part 1)
Mission design using smaller launch systems: an example
Economic arguments for smaller boosters are all well and good, but can programs of lunar or Mars exploration really be accomplished without a heavy-lift launch vehicle? Let’s take a look.
Consider Robert Zubrin’s ground-breaking Mars Direct mission architecture, which is comprehensively discussed in reference . The Mars Direct plan requires two spacecraft per complete mission. One, an Earth Return Vehicle (ERV), is launched to Mars unmanned, lands on the Martian surface, and produces return propellant from surface resources (using well understood chemical processes). The other is a crewed habitat module (Hab) which flies out approximately two years later on a slightly faster trajectory. The Hab lands at the same site as the ERV, and after approximately 500 days of surface exploration, the crew departs Mars in the fully-fuelled ERV for a direct flight back to Earth. Both vehicles in the Mars Direct mission design weigh in at between 25 and 30 tonnes on the Martian surface (for the purposes of this analysis, we’ll assume both are exactly 30), and therefore require about 20 extra tonnes of additional gear in the form of landers, heat shields, and propellant at the time of dispatch to Mars. Thus, for each vehicle, approximately 50 tonnes must be injected to the Red Planet every two years to support each mission. Now, because you would presumably want to fly more than just one mission, an extra ERV would have to be sent out with the Hab to support a following expedition, meaning that an average of approximately 50 tonnes per year (two spacecraft every two years) would have to be launched trans-Mars to support a continuing human presence on the surface.
Zubrin advocates doing this with heavy-lift launchers—specifically, by throwing each Mars-bound payload directly into interplanetary space with a single launch of a shuttle-derived HLLV. Neither multiple launches nor orbital assembly is required, making the entire plan quite elegant—and indeed, Mars Direct has powerfully influenced the ways in which astronautical engineers consider piloted expeditions to the Red Planet.
But let’s assume we don’t have a heavy-lift booster (because we don’t), and that we won’t get one anytime in the near future (which we might not). Zubrin’s architecture is still among the most simple, robust, and cost-effective ways yet proposed for human Mars missions (indeed, that’s its tagline), so we’ll use it as a basis for our mission plan, and see if it can be accomplished with existing or near-term launch systems. (This is the approach of the “Mars for Less” architecture discussed in reference .)
We’ll assume the existence of a launch system that can deliver 25 tonnes to low-Earth orbit (the approximate capacity of the U.S. Delta 4, Atlas 5, European Ariane 5, or NASA’s proposed Crew Launch Vehicle). Each spacecraft we deliver to the Martian surface will have to mass approximately 50 tonnes in LEO to begin with, so each will require two launches of our candidate boosters to deploy. If we assume the use of high-performance hydrogen/oxygen propulsion stages massing 25 tonnes each (22 tonnes propellant per unit), then we would need to use four stages to send our 50-tonne payload trans-Mars. So pencil that into the launch schedule: for each Mars-bound payload, a total of six 25-tonne “medium-lift” boosters would be needed: two for the spacecraft itself and four to deliver the rocket stages needed for trans-Mars injection (TMI).
The overall performance of this four-stage propulsion system is, in fact, much better than it needs to be; indeed, the trans-Mars injection assumed above is based on a slightly expedited trajectory with a flight time on the order of about five months—advantageous for crewed missions, and which has the important feature of being able to return a spacecraft to Earth after two years in the event that landing on Mars is impossible (known as a “free return” trajectory). However, unmanned spacecraft, such as the mission’s Earth Return Vehicle, don’t need to travel quite so fast. An eight-month, near-minimum energy trajectory would suffice for the ERV, and would also allow the unmanned spacecraft to carry a greater amount of payload (about five tonnes) than the crewed launch. Alternatively, the extra performance margin could instead be used to increase the size of the spacecraft’s launch window. That would be highly beneficial in the event of an assembly delay, the possibility of which is often leveraged against mission designs employing multiple launches.
But what about delays? Because six individual launches would be required per spacecraft in this plan, delays could certainly become a serious issue, since the high-energy propellants required for injecting each spacecraft to Mars may have to sit in low Earth orbit for some time, and would consequently begin to boil away in the harsh environment of space. Yet it’s been shown  that, depending on the specific propulsion stage design being used, a wait time of up to half a year in orbit can be acceptable, assuming that multi-layer insulation is used and that each stage is delivered over an equal interval of the wait time. Even if they aren’t, for a fixed amount of time spent in orbit prior to TMI (call it a “hard ceiling” of six months), launch delays could actually have the effect of increasing propulsion performance, since delayed stages would end up spending less time in space prior to use. (If this hard ceiling is violated, of course, the mission would probably have missed its launch window, and all bets would be off anyway. This is a problem that all space missions beyond low Earth orbit must face.)
Six months is a lot of margin, and a six-month assembly time and its corresponding propellant loss is actually factored into the performance assessment of our four-stage propulsion system. Even if assembly took longer than half a year, the only consequence would be losing the capability to fly on a free-return trajectory: the propulsion system would still be able to accomplish minimum-energy or better flights for up to nine months wait time in orbit .
How about the issue of assembly itself? Well, out of all the things required for our Mars mission to succeed (trans-Mars injection, aerobraking into Mars orbital capture, successful surface landing and rendezvous, and making rocket fuel from Martian air), orbital rendezvous and docking are the two things which today’s space programs actually have the most experience with—and besides, it’s not like we’re talking about building a space station here. The system of propulsion stages outlined above would be much easier to assemble than spacecraft like Mir or the International Space Station, since there would be no need to extend a complex life support system across the entire series of modules. With discrete, compartmentalized propulsion stages, little more than truss sections would be required to react and retransmit propulsion loads between them. Since only structural considerations would be primary, the entire assembly and integration process would be much simpler than for projects like the ISS.
These arguments, of course, are debatable, and many reasonable counterpoints exist. Regardless of what some advocates would like to believe, mating spacecraft hundreds of kilometers above terra firma is never truly simple, but the need for orbital assembly versus the capabilities of available launch technology defines the relevant trade space in this kind of mission design. The key issues in using smaller launch systems will always revolve around orbital assembly and wait time, and these issues are certainly not trivial. It’s simply a question of which set of problems (getting it together up there, or getting it up there in the first place) are harder to deal with, and for the reasons discussed thus far—and several to come—this author contends the latter to be more problematic. NASA does orbital rendezvous all the time, but it’s been over thirty years since that they had a flight-ready HLLV.