The Shape of Starships to Come

This may be a controversial statement, but any ship large enough to support a crew is too large to be a realistic option for interstellar travel.

Space is big, unfathomably big, and the problem of venturing between stars has occupied theorists for quite a long time, leading to some audacious proposals. To see how vast an undertaking a conventional starship would be, we just have to look at the numbers:

• Project Daedalus: This was a study for a two-stage, fusion-powered probe undertaken in the 1970s. It called for a 54,000-tonne behemoth of a craft with only 500 tons of scientific payload. Obtaining the helium-3 needed for one flight would have required mining the atmosphere of Jupiter for 20 years, and the flight time would have been 50 years, with no provisions for slowing down at the target star. https://en.wikipedia.org/wiki/Project_Daedalus.
• Project Longshot: Study from the 1980s for a fusion-pulse-powered starship, relatively reasonable except for the travel time. 396 tonnes and 100 years to reach Alpha Centauri. https://en.wikipedia.org/wiki/Project_Longshot.
• The Enzmann Starship: A generation ship concept from the 1960s, featuring a three million ton (!) ball of frozen deuterium as fuel. All this would have brought a crew of 200 up to 9% of the speed of light. Source: https://en.wikipedia.org/wiki/Enzmann_starshiphttps://en.wikipedia.org/wiki/Enzmann_starship.
• Valkyrie: This is actually a design that made it into The Killing Star, which I reviewed here. It was also the inspiration for the ISV Venture Star from Avatar, above. The numbers seem pretty good (10% c, 200-ton spacecraft) until you get to the fuel, which is fifty freaking tons of antimatter. Ouch. Short of strip-mining the surface of Mercury and turning it into a solar-powered factory (which people have proposed), it’s not really practical to make that much antimatter, let alone store it. Source: http://www.projectrho.com/public_html/rocket/slowerlight.php#id–Go_Fast–Starships–Valkyrie_Antimatter_Starship.
• I’m not even going to talk about the Frisbee Antimatter Starship, which  is 700 kilometers long with 71 billion tons of dry mass.

What do all these starships have in common? They’re big, with payloads in the dozens to hundreds of tons, and more importantly they use no more energy than what they carry aboard, in the form of fusion bombs, antimatter, or three-million-ton balls of deuterium. All these energy sources require mass in the form of handling and containment, not to mention the mass of the fuel itself; thus, more fuel and more propellant are needed, rapidly ballooning the size of the craft to something that can only be constructed in orbit, over a span of decades, with the GDP of an entire planet.

It gets worse. Supposing you do build your titanic-scale probe, or your self-sustaining generation ship, it will face a bevy of problems during its long voyage through interstellar space:

• Any ships traveling near the speed of light will be at risk of particle collisions, requiring very heavy shielding to withstand. This in turn demands more fuel and more propellant.
• Slowing down requires just as much energy as accelerating. If you want to sit still and explore for a bit, or set up a colony, your delta-v requirements are doubled.
• Generation ships are tricky. Really tricky. Making an enclosed biosphere last the length of a trip is a task not much easier than building the ship in the first place–Earth ecosystems only reached any degree of stability after millions of years of trial and error–and that’s not even getting into the social/cultural problems of bottling up a civilization for centuries on end.
• Unless you sent a probe in advance of your generation ship, requiring centuries more time and effort, there is the risk that the destination planet is not actually habitable, rendering the entire voyage a waste of resources and colonists’ lives.

This is the part where I present my own favorite idea as a workaround to all the problems I just listed. Enter the unmanned lightsail.

Build a miniaturized spacecraft, attach it to a sail thinner than paper, and shoot it with a gigantic laser to send it hurtling off into the cosmos. This has several advantages. First, there is no need to carry propellant. Second, there is no onboard engine—the energy source is a laser battery near Earth, where it is easily maintained and can be used for more than one mission. Third, the craft itself is tiny, certainly less than a hundred tons. Even a large lightsail could be assembled with one or two launches, not the thousands needed for several of the examples above, meaning that one could send a fleet of expendable probes to every star within range. Such a miniaturized vessel would also be breathtakingly fast;  we’re talking 0.5 c, not 0.05, enough for the folks back home to see the entire mission with time to spare.

Slowing down at the destination is one of the main technical challenges with this approach. But there are some solutions are out there, and they’re easier than they sound. Most of them involve using a magnetic field as a sort of space parachute, braking directly against particles in the interstellar medium, and I’ve also seen a proposal to use the galactic magnetic field to swing the ship in an arc, turning it back around the way it came so that the original laser can slow it down. I’ll discuss these methods more in a later post.

The real downside of a wafer-thin lightsail craft is this: you can’t fit a crew in a wafer. You also can’t fit a crew in a starship the size of a telephone box, or even the size of a house, because the mass requirements for supporting humans for years on end (and landing them on the surface, and protecting them from cosmic radiation, etc) are vast. Easier to send a miniaturized AI or something, don’t you think?

I think the most realistic starship proposal would only be around one hundred tons, perhaps much less. It would comprise an equipment module and a kilometers-wide lightsail. The equipment module would include a powerful computer, to make decisions in the absence of human crew; a bare minimum of scientific instruments; a small nuclear reactor; reaction control; and facilities for in-situ resource utilization, perhaps using self-replicating robots. Not a gram of mass would be wasted. Most things—additional cameras, re-entry shields, surface exploration vehicles, antennae for communicating with home, etc—would be fabricated on-site, out of local materials. An asteroid in the target star system would do quite nicely for metals, and the local star would provide an additional energy source.

One could bootstrap a colony this way, too—it only requires some fairly speculative technologies, which are still more reasonable than putting fragile humans aboard a million-ton starship. All one really needs to bring is information; the AI could assemble human DNA from local organic compounds, such as those found on comets, and then grow bodies from there. Rinse and repeat for crops, beneficial bacteria, and every other element of a functioning ecosystem.

Would this require unprecedented mastery of biological processes? Of course. Would it entail AI smart enough to engineer a colony from scratch, yet small enough to fit within a few tonnes? Absolutely. These technologies may require centuries to figure out, but they only need to be invented once. A hundred of these ships, each a small seed of science and industry, could be sent to a hundred different star systems, and at least some would survive the perils of interstellar space. It would only require a small up-front investment of research and infrastructure, and then we could colonize the galaxy at our leisure.

This, I think, is far better than the titanic, ponderous starships of old.

Agree with my conclusions? Disagree? If you have anything to share, please comment below, and be sure to follow Let’s Get Off This Rock Already for more weekly space content!