The Case for Mars (11 page)

I did not agree with Ben’s thinking. I found his robotic self-assembly scheme not credible. Moreover, with a requirement to launch 700 tonnes to low Earth orbit (LEO) for each flight, there would not be very many future missions launched to Mars, and the thirty-day surface stays would not allow sufficient time for much real exploration to occur. As far as I was concerned, we were not going to Mars to set a new altitude record; we were going there to explore and develop a planet. The possibility of a sustained presence on Mars required a large number of repeat missions, and the only way this could be done was if the mass, and therefore the cost of the mission, could be brought way down. The best way this could be done would be by manufacturing an entire mission’s return propellant on the surface of Mars. In fact, in 1989 I had done studies showing that if such a strategy were combined with the use of nuclear propulsion for the outbound leg of the mission, a single booster in the class of the Apollo-era Saturn V could launch an entire human Mars mission. Launched with a single booster, the whole system could be integrated on the ground at Cape Canaveral, and the issue of on-orbit assembl
y of interplanetary spacecraft would be moot. Moreover, by using locally produced propellant, the whole mission could be landed on Mars, with no liabilities left in Mars orbit, thereby enabling the kind of long surface stays I felt to be absolutely necessary if the program were to do anything useful. Direct launch with a single throw of a heavy-lift booster, use of nuclear propulsion on the outbound trajectory, and direct return from the planet’s surface using in-situ produced propellant—this was the way to go.

Enter David Baker. Baker was a very sharp engineer working at the time as the systems and design lead on Martin’s Space Transfer Vehicle (STV, a ferry for lunar missions) program. On STV, Baker was being driven out of his mind by arbitrary requirements levied by NASA. For example, the STV would have to be able to land on the Moon if any two engines failed. (The Apollo lunar lander had only
one
engine.) For reasons of thrust symmetry, this meant you had to have five engines, when one would do the job. Now you had too much thrust, which meant the engines had to be made to throttle back to 10 percent power, which they weren’t designed for, thus requiring a new costly engine development program. Furthermore, NASA required that the engines be reused. That meant dragging five heavy engines all the way to the Moon and back—adding enormously to the launch mass and cost of the mission—and then checking the engines out and refitting them in a multibillion dollar orbital facility. All this to do a job that could easily be done better with a single, off-the-shelf Pratt and Whitney RL-10 engine costing $2 million. On STV, Baker contributed what he could as a “team player,” but he confided to me at the time, “None of this makes any sense.”

Baker had participated in earlier Mars mission studies whose direction had been dictated by 90-Day Report type thinking, but it was clear that the logic behind them (or lack thereof) left him feeling distinctly uncomfortable. I hit him with my ideas; some he already agreed with, others, such as the central role of using locally produced return propellants in human Mars missions, I gradually was able to convince him to accept. On other points, however, he would not budge. In particular, he wouldn’t buy the idea of using nuclear propulsion as the basis for the first Mars missions. Its development would cost too much, he argued, and publehinceptance would be a problem. I didn’
t accept those arguments; on a sustained human Mars program the cost of developing a nuclear rocket engine would be paid back in reduced launch costs after only two or three missions, and if the public wanted a sustained Mars program, they would come to accept nuclear propulsion on that basis. But, Baker said, if you insist on using nuclear propulsion starting on the very first mission, you’ll delay the whole program, perhaps fatally.

That point struck home. I felt very strongly that a humans-to-Mars program had to be done on a rapid schedule. Fast schedules reduce program cost: cost equals people multiplied by time. Moreover, every year any major program has to go before Congress for continued funding where it faces risk of termination, often caused by deals or interpersonal frictions that have nothing to do with the program itself Every time a program goes before Congress for funds it is forced to play another game of Russian roulette. You can only expect to be lucky so many times.

In 1961, John F. Kennedy called for reaching the Moon by 1970. By 1968, administrations had changed, and even as the Apollo astronauts were landing on the Moon, President Richard Nixon was ripping the program to shreds. If Kennedy had called upon the nation to reach the Moon in twenty years as opposed to a decade, 1969 would have seen NASA in the final stages of the Mercury program, with the Moon still a long way off. The program would have been canceled, and reaching the Moon would still be considered an impossible dream today. If you want to get humans to Mars, you can’t do it in thirty years; you can’t do it in twenty years—ten years is the most you can hope for.

Nuclear propulsion, I conceded, might have to wait but the Mars mission couldn’t. By all means, use nuclear propulsion whenever it should materialize; it will increase mission payload capability and cut launch costs (by about a factor of two). But don’t delay the mission until you’ve got it. Go as soon as you can with what’s at hand. Improvements can come later. As Baker and I started spending a lot of time in conversation, debating many issues of vehicle and mission design, both technical and philosophical, we increasingly began to converge. We resolved to collaborate.

In many respects we were an unlikely pair for collaboration. I’m on the short side; Baker is extremely tall. I’m mercuric; he’s phlegmatic. I’m an optimist; he’s a pessimist. I’
m a romantic; he’s an existentialist. My favorite movie is
Casablanca;
his
Brazil.
My thought process moves in leaps and jumps; his in a steady tread. My credo agrees with that of Hegel: “Nothing great has ever been accomplished without passion.” When on one occasion I told that to Baker, he winced and walked out of the room. For Baker, passion and engineering do not mix. For Baker, apparently, it is sufficient to do excellent work and to live well. I want to change the world.

Nevertheless, collaborate we did, and for some time in 1990, extremely effectively. We had complementary strengths. I had a much better academic education in broad areas of math, science, and engineering; he had a lot more engineering experience and “knew the ropes.” I supplied creative drive, he supplied discipline. We never became close friends, but as a team, we worked.

As alluded to above, in 1989 I had shown in a number of papers that, if nuclear propulsion were available and in-situ produced propellants were used for Mars ascent and Earth return, a human Mars mission could be launched with a single booster in the Saturn V class. Baker had designed sin heavy lift vehicle for NASA. He called the vehicle “Shuttle Z” after Code Z, the NASA organization in charge of developing human space exploration plans at the time. Shuttle Z was basically a growth variant of NASA’s Shuttle C design, which replaced the orbiter on the Space Shuttle launch stack with an expendable cargo pod. Shuttle C could deliver about 70 tonnes to low Earth orbit (LEO). By adding a powerful hydrogen/oxygen upper stage inside an enlarged side-mounted cargo pod, Baker created Shuttle Z, and increased the LEO capability of the vehicle to about 130 tonnes, just 10 tonnes short of the Saturn V’s capability. Because all the key components of Shuttle Z were drawn from the Space Shuttle inventory, it would be possible to develop the vehicle quickly and inexpensively, a key requirement for a decade-long program.

So, we had our booster, but we didn’t have nuclear propulsion for either the outbound or homebound leg of the mission. Throwing our hardware to Mars without nuclear propulsion would now require two launches. In itself, this was not a show stopper, but it did make our mission architecture inelegant, at best. In our design, an Earth return vehicle (ERV) sat atop a habitat, which, in turn, sat a
top a partially filled Shuttle Z upper stage, which sat on another nearly filled upper stage. This stack was assembled in orbit via a rendezvous and docking maneuver, with the first three elements (ERV, habitat, and one partially filled upper stage) delivered by one Shuttle Z launch and the fourth (the second nearly filled upper stage) by another.

For a number of reasons this design was not very attractive. To begin with, the long stack was awkward, and whichever booster launched first would leave its payload in LEO for several months, during which time a significant amount of propellant in the upper stage would boil off. On arrival at Mars, the ERV/habitat payload would ride behind an aeroshell—a blunt, mushroom-shaped shield—and decelerate as it plowed through the Martian atmosphere. However, the combined ERV/habitat payload was so heavy that it was questionable whether even a folding aeroshell big enough for the job could be made to fit inside the Shuttle Z payload fairing. But there was an even bigger problem once at Mars.

When nuclear propulsion had been in the cards, I had designed a propulsion system that simply compressed and stored Martian carbon dioxide and then heated it with a nuclear reactor to produce a high-temperature vapor rocket exhaust. (About 95 percent of the Martian atmosphere is carbon dioxide, which liquefies under Martian temperatures when compressed to about 100 pounds per square inch.) Mechanically, such a propellant manufacturing system is very simple; basically all you need is a pump. In such a plan, it was reasonable to suggest that the astronauts could acquire their return propellant after they landed on Mars. Without nuclear propulsion, however, any propellant produced on Mars would have to be manufactured by some form of chemical synthesis. This would be considerably more complex than simply compressing and storing carbon dioxide. Undoubtedly, NASA would quite reasonably insist that all the propellant required for a return to Earth be produced before a crew headed off for Mars; otherwise the crew could find themselves stranded on the planet if fuel production failed.

In 1989, Jim French, an independent engineering consultant, had published an article in the
Journal of the British Interplanetary Society
with some of these considerations in mind. French sugge
sted landing a propellant manufacturing plant on Mars before a crew arrived. The plant would produce and stockpile fuel for the return leg of the crew’s journey. But this left the problem of landing the crew’s spaceship within a hose length of the propellant depot. This proved so problematical that French ended up conceding that the use of in-situ propellant would not be practical until after a manned base was well established on Mars, along with local infrastructure to provide backup against all sorts of contingencies.

So there matters stood: by dropping nuclear propulsion we gained the possibility of an accelerated schedule, but a host of problems appeared to come along with the bargain. The most intractable was the question of transferring prearrival-produced chemical propellant from storage to the ERV. Depend on a prelanded mobile robotic fuel truck? Too risky. Wrestling with this problem, I hit upon a novel architectural idea, which now seems so obvious.
Don’t send the crew out with their own return vehicle—send the return vehicle, with the propellant plant integrated into it, out first.
At one blow, this idea solved practically everything. The habitat and the ERV were by themselves each light enough that either could be launched
directly to Mars
by a single Shuttle Z. We would still require two launches, but now one Shuttle Z could launch the ERV and another could send the crew and their hab. Combined, the ERV/hab payload would have required a huge aerobrake that would have presented serious challenges to our aerobrake designers, but, flown singly, a manageable aerobrake for each could be made to fit inside the Shuttle Z fairing. To ensure our Mars crew would not be stranded, the ERV would fly one launch opportunity, or twenty-six months, prior to the launch of the astronauts. Thus all the propellant would be made before the crew ever left Earth, and, since the propellant plant was flown to Mars integrated with the ERV, there was no question about landing “within a hose length.” The plumbing that would deliver the Mars-manufactured propellant from the chemical synthesis unit into the ERV’s fuel tanks would be hardwired installed on Earth.

Best of all, at no point would the mission require on-orbit assembly or orbital rendezvous of any type. The only rendezvous required would be on the Martian surface, and that’s easy. During Apollo, we landed an Apollo crew within 200 meters of a
Surveyor
spacecraft that had landed on the Moon several years before, and we have much better avionics today. In an orbital rendezvous, if you miss by 10 meters, you’ve missed. But in a surface rendezvous, you can be off by 10 kilometers and it doesn’t matter—you can just walk or drive over. Moreover, as part of the habitat’s cargo, we included a pressurized ground rover with a one-way driving range of 1,000 kilometers; it would take
really bad
piloting to land further from the ERV than that. And say what you will about the NASA bureaucracy, the NASA astronaut corps contains pilots that are among the world’s best. Without question, the surface rendezvous would succeed.

Though unorthodox and apparently daring, sending the crew to Mars separate from the Earth return vehicle is actually much safer then landing the crew on Mars together with the vehicle that will deliver them back to Mars orbit. The reason is simple: if the ERV is sent first, the crew will know before they even leave Earth that they have a fully functional Mars ascent and Earth return system waiting for them on the Martian surface,
one that has already survived the trauma of landing.
In contrast, a crew that lands with their ascent system can only guess in what shape their Mars ascent vehicle will be after they’ve hit the surface. Moreover, in our plan the crew would launch to Mars in convoy with another ERV, which would land at a site within rover long-distance driving range. This second ERV would start propellant production for a second human mission to Mars, but in the event of an emergency, it could serve as akup vehicle for the crew of the first mission. In addition, the two ERVs on the Martian surface together with the crew’s own habitat module gives the crew a total of three redundant units capable of housing them and providing life support. As far as safety is concerned on Mars missions, it doesn’t get any better than that.