Billy Brown wrote:
> Anders Sandberg wrote:
> > Hmm, exactly what problems do you see? That they need to be *big* is
> > obvious, you need space for a lot of frozen people, livestock and
> > gametes, equipment for colonization (you have to be prepared for
> > almost anything; most likely a portable asteroid factory would be a
> > good idea), space for thawed people to work, protection against
> > relativistic shrapnel and of course things like propulsion, shuttles
> > and life support.
> The fundamental problem with any attempt at interstellar colonization using
> conventional technology lies in the fact that machines don't work forever.
> Complex machines need constant maintenance to avoid breakdowns, and
> sometimes they break down anyway. High-performance systems are by
> definition highly stressed, and consume spare parts the way an airplane
> burns fuel. SF stories commonly deal with this problem by pretending it
> doesn't exist (the "starship runs itself for 300 years while we hibernate"
> routine), but in the real world we'll have to face up to it.
High performance systems are only highly stressed when under constant operation at peak operating levels. I for one would not design such a ship that way.
> This simple fact creates a complex, multidimensional problem for anyone who
> wants to build starships without nanotech. You need to have the equipment,
> manpower, raw materials and knowledge to keep your ship running indefinitely
> (which is essentially equivalent to being able to build the entire ship from
> raw materials). At the same time you need to be able to achieve a
> significant fraction of light speed, which means that payload mass is very
> expensive. Oh, and the whole thing needs to be cheap enough that a
> pre-nanotech civilization can actually afford to build it.
Making devices and systems which are durable over a long term is merely a matter
a) fewer moving parts
b) having moving parts that do no rubbing, as much rolling as possible, and any contact must be minimal in area of surface contact and in speed, if at there is any contact at all (magnetic bearings are wonderful things). c) building devices with maximum performance tolerances far in excess of the normal operating levels
d) build them modular so that those units that fail can be easily replaced by robot systems.
One example of an ideal device is an ion or plasma engine. There are no moving parts besides the pumps (which could be made to be magnetohydrodynamic if need be), and are proven to be able to run on constantly, or intermittently, for periods of years so far. For a 300 year voyage ion systems are sufficient along with nukes for power supply and a hydrogen scoop to supply fuel.
> Without advanced automation, your crew doesn't even have the manpower to do
> routine maintenance on their ship. Without nearly-sentient AI, they can't
> possibly know enough to repair things when they break. Without nanotech,
> the ship needs to carry so many different kinds of manufacturing machinery
> that you end up with a flying city.
Space probes like Voyager were equipped with redundant systems and simple 70's era AI software to manage device failures.
Now, with nano tech it is obvious that all you really need to carry for cargo are enough bootstrap assemblers and cad libraries in several redundant copies. In early 21st century tech we are talking a pea-pod's worth of cargo. Then you've just got to worry about the cryo-storage systems for the colony crew, and the ship itself, along with survival supplies and equipment for the crew for about 6 months. If the nanites haven't built a civilization for the crew by 6 months, then they are dead anyways.