Imagining what a starship would be like

[jollyreaper]

We tend to be a bit prejudiced towards the present in scifi. Starships will be just like navy ships, only in space. They'll need the same number of crew doing the same sort of things. Red Dwarf made a joke of this in the intro, of course, having one of the crew outside scraping and painting the hull. Space barnacles? lol And there's also the bit of WTF from the new Star Trek trailer where we see a 19th century ironworker welding together the hull plates of the Enterprise, presumably launching from the Earth's surface!

As part of the biowank movement we started to see organic ships capable of self-repair. In contrast to the Death Star with a crew of hundreds of thousands but could have just as easily been millions given its size, we have another planet destroyer Lexx which really doesn't need any living crew at all, just one captain with a key and he's just there to tell the ship what its allowed to blow up. Every system on the ship is capable of self-repair and doesn't need any human interaction.

A recent book, Killing Star, got me thinking about this. There's a high level of robotic automation in the book. Human subs and spaceships are all heavily roboticized and are capable of self-replication. While the details are not gone into specifically, these robots seem to be proper von neumman machines, capable of replacing themselves as needed from feedstock onboard the ship. Damaged parts could be replaced on-site by these bots.

So, just for general discussion, what's everyone's thoughts on this sort of thing? I'd personally be curious as to the amount of self-repair a ship could perform. Could nano-assembly replicate the kinds of chemical bonds we can only now make with massive amounts of heat and pressure? Would the natural process end up being too slow? We can build a motorcycle in a lot less time than it takes to raise a horse and it travels faster and has no need of rest but you can feed a horse grass from your field and a herd of horses can make more horses whereas you'd need a factory to make more motorcycles. You can build a house with conventional means far more quickly than a tree can grow and the tree would be presumably using the same processes to draw building materials from the ground as your nanobots would.

[Zixinus]

I am skeptical that you can replicate everything or even repair everything. The more delicate is something, the longer it likely will take to make it (and more importantly, the more equipment is required to make it, which too have to repaired/maintained/created). However, it would make plentiful sense to make various self-repair functions.

For example, a repair droid like in Star Wars's R2-D2 makes plenty of sense: if designed so, a robot can more readily go out and repair things than it takes for an astronaut to adept to high-oxygen levels (to avoid the bends) and get suited up. It would also make sense for these robots to be equipped for as many tasks as possible.

As for creating stuff onboard, that really depends on the mission type. You design the spacecraft according to the mission. Some things may be self-evident to bring (duct tape, various other sealants, some wires, a multitool, etc) and others may rely on the mission (it may be more economical to bring a spare than to bring a machine to make a spare).

[sirocco]

There are also the Replicators in Stargate:SG1.

Specifically the bugs version which are made of tiny pieces (some bogus called kyron pathways) totally independent from one another. They can repair themselves as long as there are kyrons left and they could basically scavenge the remnants of another bug robot since it is made of the same components. There is even an episode where they joined to make a spaceship.

Theoretically, it's a good idea (extreme modularity). But actually, you could only use them for specific purposes since the more complex you need them to be, the more time they would need to assemble into a working entity.

Breach in hull: OK, just send a dozen of robots to fix it.
One drive shut down: Well, not that easy.

[Eleventh Century Remnant]

http://www.bisbos.com/rocketscience/spa ... dalus.html-

Prepare for a blinding glimpse of the obvious; the problem with science fiction is that it's quite frequently based on fictional science. Alternative version of the above- reality sucks.
Project Daedalus is as close as we've got to an actual design study, and the Wardens there are about as good as you're going to get. They are basically free roaming industrial robots massing about ten tons each.

The ship itself is basically a probe, and I can't help wondering if there was supposed to be a phase II in which the second stage would be used to decelerate at the target instead of being used for further acceleration; in that case, a 450 ton colonisation payload- not even enough to set up a village, we're looking at assisted panspermia here, really- could be carried.

It almost certainly is more mass efficient and reliable to send an extensively prepared and programmed technician robot or multiple thereof than to ship a human, and their life (and sanity) support needs.

I don't "believe" in nanotech- not the science fiction version of it, anyway; some fascinating things can be done with extreme small scale down to molecular manipulation- but whether they can be done practically, outside a clean room, in quantities sufficient to matter, without the cost and mass burden of essentially shipping a chip factory, that I'm a bit doubtful of.

If you have to worry about mass ratio and actual economic cost, it's also going to be one flight, one way, once. Any starship sent out isn't coming back until the colony it establishes reaches a sufficient technical base to refuel it.

[jollyreaper]

I am skeptical that you can replicate everything or even repair everything. The more delicate is something, the longer it likely will take to make it (and more importantly, the more equipment is required to make it, which too have to repaired/maintained/created). However, it would make plentiful sense to make various self-repair functions.

I'm thinking it would be a matter of just how complex the technology is in the setting. Our carriers have impressive machine shops and can create a wide variety of parts on the high seas but they're not building a spare reactor. They're also not scratch-building new F-18's. It takes a whole lot of land infrastructure for that. But I'm wondering just how much you could build if you do have atom-by-atom nano-scale precision assembly. But there would also be a lot of different stages between here and there.

Going back to the age of sail, there was actually quite a lot that the sailors could do for maintaining their ships far from home. Still, they weren't exactly going to be able to stop at a deserted island to pick up fresh sails and rigging.

[Simon_Jester]

The trouble with atom-by-atom is that by nature it makes precision difficult, and that it is going to be slow. For making heavy machinery, it leaves a lot to be desired.

[jollyreaper]

The trouble with atom-by-atom is that by nature it makes precision difficult, and that it is going to be slow. For making heavy machinery, it leaves a lot to be desired.

That's what I was wondering. So maybe the starship could grow a new engine if it had to -- say it was stranded in an uninhabited system -- but swapping in a fully assembled replacement would be so much quicker.

The argument against grey goo nanowank is that the nanites would likely be so delicate, they'd require clean rooms and possibly vacuum to operate successfully and stray cosmic rays would slaughter them. So while they made for great scifi baddies, they're about as realistic as zombies. Or if we look at it from a certain perspective, a fungus infection can look like grey goo, same with mold in a building and it can be annoying to deal with but annoying ain't the same thing as omfg apocalypse time. Perhaps robust nanites in the wild would be just another form of pest to deal with like ants and bathroom mold.

[sirocco]

I thought about using the robots to mine an asteroid and transform it into a spaceship.

You could basically just put people into stasis chambers in the deepest part of you flying rocks and they'll be sleeping there either until the life support has been installed by your robots or until they get to their destination.

Wouldn't that be a big time saver?

[jollyreaper]

I thought about using the robots to mine an asteroid and transform it into a spaceship.

You could basically just put people into stasis chambers in the deepest part of you flying rocks and they'll be sleeping there either until the life support has been installed by your robots or until they get to their destination.

Wouldn't that be a big time saver?

Could be. The only question becomes what are people needed for if the bots are so capable? Seems like you either go the Dune route or fully embrace transhumanism.

[sirocco]

The objective is clearly to colonize extrasolar planets with "actual" humans.

Having to wake up some of them during the flight could be because they have a expertise. Since so far machines were created to perform repetitive actions better and faster than us, it may be smart to have some people to cater the special cases.

There will also be scientists who wants to work on-site. Or it could just be because your stasis technology is not that good and you can't keep people petrified for so much time (= in-universe explanations).

[PainRack]

What about a mini-earth in Space?

Shouldn't it be that difficult to create a starship that large enough, say 3km by 3km large or something, strap on small engines and nav gear on board and then recreate a biosphere on board?

[Zixinus]

Shouldn't it be that difficult to create a starship that large enough, say 3km by 3km large or something, strap on small engines and nav gear on board and then recreate a biosphere on board?

Generation ships? Or more like strapping a few rockets on an O'Neill cylinder?

The problem is: more mass is more of a design mess. The heavier something is, the larger engines (and several times more propellant, thing several times bigger than your payload) you need, unless you have some space opera gear hyperdrive or something like that.

Making a mini-earth has the problem that then you need everything on that and that requires mass. The more mass you have, the more difficult it becomes to push anything anywhere. Hence why small probes are easier to send out in space than human-flown ships.

I'm thinking it would be a matter of just how complex the technology is in the setting. Our carriers have impressive machine shops and can create a wide variety of parts on the high seas but they're not building a spare reactor.

Of course, my thoughts are similar. Again, it depends on the complexity on the to-be-repaired part or replacement part itself: simple circuit boards and can be made or improvised (say, for a lamp). However, to actually make semiconductors is another matter entirely.

My guess is that it depends on what you might need.

But I'm wondering just how much you could build if you do have atom-by-atom nano-scale precision assembly.

I personally thing that you would only should use such tools for making things that requires atom-by-atom precision. If not, then it might be better to bring a less-precise but cheaper tool.

[Sriad]

I'm fond of saying nano-scale von Neumann machines are wank, but we already know micro-scale is possible because bacteria exist. We even know of extremophiles that can survive being frozen in hard vacuum for extended periods of time; there's every reason to believe that an "intelligently designed" bacteria-scale machine would be at least as hardy as the ruggedest natural bacteria.

They certainly wouldn't be as fast as the classic Gray Goo (or a traditional machine shop, for macro-scale work), but to set up a colony in a new star system it's hard to beat in terms of mass efficiency. A can of micro-assemblers, a ready supply of elements expected to be scarce, a power source, and a few petabytes of radiation-hardened data storage is all you need.

[Coyote]

Probably the biggest problem with all that is the notion of "one-type-fits-all" industrial processes. In other words, nano-construction will be the future because nano-construction is "so cool". Or that bio-construction will be the future because bio-construction is "so cool".

More likely, construction will be carried out in a variety of ways, in which some large pieces are cast-moulded, for example, and then bioelectric conduits are grown on it to connect a network of nano-made fracture sensors, etc.

In my sci-fi, I have an internal bio-synthetic organism that eats the crews waste and secretes water and oxygen. It is not a 100% cycling process, there is loss as the organism metabolizes stuff for itself, but it helps extend operating range. But this system exists in a ship that is framework-and-metal/ceramic with nanotech components, etc.

"One size fits all" is simply either a lack of imagination, or... not the point of the story. If a story makes a "one size fits all" process the centerpoint of their story and uses that to explain why their ships are superior to all others, then in my opinion it'd be lazy writing or an agenda. But if it's just a handwave to explain how the ship got there while focusing on things like characters and plot, then I'd just shrug and let the writer have it while concentrating on the real story.

[sirocco]

Hence the automated factory in my solution. The main purpose is to build the bots tailored to do specific tasks, repair them when possible or just recycle the outdated versions.

With the asteroid-based ship, you could start with crude robots to engineer the main sections then as time passes make them smaller and able to make more precise tasks.

[jollyreaper]

Probably the biggest problem with all that is the notion of "one-type-fits-all" industrial processes. In other words, nano-construction will be the future because nano-construction is "so cool". Or that bio-construction will be the future because bio-construction is "so cool".

More likely, construction will be carried out in a variety of ways, in which some large pieces are cast-moulded, for example, and then bioelectric conduits are grown on it to connect a network of nano-made fracture sensors, etc.

That's an interesting thought. I've seen some scifi discussion of self-assembling structures. In the case of a ship, maybe the larger structure isn't cast-molded but sculpted into place and then that structure is the seed for growing the nano-stuff on top of it.

By the time we build these starships, I'm thinking that there won't be a human hand touching any piece of it in the construction.

[loomer]

Probably the biggest problem with all that is the notion of "one-type-fits-all" industrial processes. In other words, nano-construction will be the future because nano-construction is "so cool". Or that bio-construction will be the future because bio-construction is "so cool".

That was something Star Trek got at least partially right, even if in the wrong way - their organic computer cores. Shame they handled it in a godawful way, but at least they made the effort of incorporating biotech in a way that's more realistic than the rest of its appearances in the show.

[aussiemuscle308]

but trek also did the opposite with the borg who's ships seem to repair themselves automagically and without fail. makes me wonder why they needed biological units at all.

[Coyote]

but trek also did the opposite with the borg who's ships seem to repair themselves automagically and without fail. makes me wonder why they needed biological units at all.

"Logical consistency" (or really, any "consistency" at all) were never hallmarks of Trek.

[someone_else]

It all depends on the tech at hand.

Assuming as-hard-as-possible tech, here is short list of assumptions to guide the wanking that will follow. Just bear with me.

-Engines are very limited (acceleration in the milligees range), meaning long-ish voyages (months to inner system, years to outer system) and a very limited course (once the course is set you can at best alter it to do an abort course, if your trajectory still allows it).
And even for this I'm assuming ballpark unrealistic fusion engines, like the ones in Comparison of Fusion-Antiproton Propulsion Systems for interplanetary travel

-Since even such crappy engines put out non-trivial amounts of light and IR in their exaust (10-100 Mw range, depending on craft mass), and they have to keep thrusting for more or less the whole voyage, detection isn't so hard. Anyone with an anti-asteroid telescope system is going to see at least the deceleration burn, so they will know your arrival at least a couple weeks before you get there.
Using "stealthier" engines like NERVAs, solar/nuclear-electric or chemical rockets makes the trip kinda long or outright unfeasible (for the outer system for sure).

-life support (regenerative or not) and supplies needed by humans are heavy and dedicated radiation protection is too massive to shield the whole habitat section unless it is tiny. This means humans are a VERY COSTLY cargo to ship around.

-Armoring is rare due to both armor's mass and level of overkill most missiles pack. Anti-missile systems are much lighter and much more effective.

-Going to another (very close) solar system is barely feasible, but only with external-powered propulsion, like titanic lasers shining on a gargantuan sail or accelerating swarms of tiny drone sailcrafts to relativistic speeds and aiming them to impact on a "simple" magnetic nozzle bolted on the actual spacecraft (when the craft is far out of range of the titanic lasers).


With these premises, I can tell you something that can help you distinguish a setting from the average SF stuff.

Spacecraft aren't a "one piece" affair. Each spacecraft is assembled from a semi-standardized structure of structural supports and fuel tanks, where engines, reactors and payload are then bolted. Every spacecraft will be very "destination-specific", and it may even drop tanks or entire stages as it goes. In general, is plain stupid to waste mass in carrying around useless structures.

You don't fix anything while en-route, you design crafts with redundancy. Any kind of serious mainteneance (anything more than patching fuel tanks or cables or pipes) requires too much mass in tools and machinery to perform (also too much skill, and humans aren't always on board or at lag-friendly distances), they may also require too much time to be performed. If the engine shuts down and the craft cannot reactivate it fast enough (either to resume course or to follow a planned abort course) everything heads for deep space.

Expensive and "frequent" mainteneance (performed by orbital facilities. Due to reliability concerns and the lenght of most voyages (multiple months for inner system, years for outer system), the entire engine assembly has to go through a costly mainteneance cycle at more or less any port it stops in. You cannot mantain it while en-route (both due to mass considerations and the fact that it is more or less constantly active)
If failure isn't an option, then mainteneance gets costly.

Crews will be very very very small. Also their quarters won't be that big. The point is that the 99% of any voyage can be very well handled by computers alone, and any human on board consumes tons of mass for food, rad-shielding and pressurized sections. For the same reason, there is very little reason to have anything pressurized.

The spacecraft's volume will be mostly fuel. That means the spacecraft's main feature are the fuel tanks. If you have good eye you can spot the other smaller things bolted on.

Military and civilian craft will be much more similar than with seagoing vessels.That is, the engines and most of the bulky stuff will be the same. Although maybe military will have more fuel tanks, to allow more evasion and/or abort courses.

Spacecraft should look differently depending on where they are supposed to operate. In general, anything working in the inner-system (mars to mercury) will be white-ish or tin-foiled (for lack of better term) to reflect the most of the sun's radiation. Also, radiators and somesuch will have bigger sizes to dump the sun's heat that still gest through. Fuel tanks with cryogenic fuels will also be more bulky or have an active refrigeration. The more close to the sun they can go, the more extreme such things become.
Outer-system craft can do well with smaller radiators and lighter fuel tanks, and of the white or tin-foil paint. But may need materials that can endure the murderously cold conditions.

Starcraft (craft meant to go in other star systems) follow different rules. Unless you use an external propulsion system (the titanic lasers mentioned above or a handwavium equivalent) to reach the intended cruising speed (a speed that allows you to get to Alpha Centauri in less than 50 years at least), no known reaction engine (not even antimatter ones) can carry enough fuel to both accelerate and decelerate (the main problem of Daedalus).
Also, such crafts must endure the bombardment of interstellar dust and crap, that impact on its nose at relativistic speeds.
If you can travel like this with decent timescales, your race is badass. Truly badass.
Still, your travel distances before the people aboard die of old age are pretty limited without FTL.


These limitations are easily handwaved because most people don't even know of them, but you can at least pay them some lip-service for the sake of making something different. Like Avatar's interstellar spacecraft (the first fictional spacecraft ever that I've seen using radiators).

Qick example:
Let's say we have bio-ships. By the first rule of thumb, they change the size depending on destination. So you can say that the Star Whale has an "enlargeable belly" to store enough fuel, or that it takes it from drop tanks made of bio-stuff.

Again, bioships meant to operate closer to the sun will be white-ish and have bigger "fins" (whatever that means), while the outer-system ones will have more "fat".

And lastly, since wear-and-tear of the engine will be significant, and since being modular allows quicker repairs, you can have bioships that are in fact "colonies" of multiple critters, like the Engine-creature, the pressurized-section-creature, the fuel-pump-creature, the radiator creatures that look like coral or algae, and so on and so forth.

I'm fond of saying nano-scale von Neumann machines are wank, but we already know micro-scale is possible because bacteria exist. We even know of extremophiles that can survive being frozen in hard vacuum for extended periods of time; there's every reason to believe that an "intelligently designed" bacteria-scale machine would be at least as hardy as the ruggedest natural bacteria.

You can add another problem to claim they are wank.
Mutation.

Anything so small with the ability of reproduce will eventually screw up due to radiation damage or chemical damage or plain chance (the main source of mutation are the screwups that the DNA-synthetizing enzymes do) and produce a malfunctioning offspring.
The vast majority will just sutdown, but if the numbers are great (as in most projects where time is an issue), then you may end up with new strains, and some of those may go out of control. And actively working against the goal.

Nothing like Gray Goo, of course, but an annoyance and a reliability problem nontheless.
The only solution is having separate machinery that builds nanomachines unable to reproduce.

[Sarevok]

About the detection. You can certainly detect many spaceship sized objects at interplanetary ranges but in all likelihood won't find them. Telescopes look at very narrow sectors in the sky. Tracking a known body and finding something out there in the darkness is totally different. There is a ton of stuff out there right near Earth etc that we simply miss. The overwhelming majority of NEOs, rocks larger than aircraft carriers, are believed to be still undiscovered.

[Sarevok]

Still, your travel distances before the people aboard die of old age are pretty limited without FTL.

The solution is genetic engineering to slow down or stop aging.

Cheaper and simpler than physics breaking FTL. Or ludicrous STL drives that can reach significant fractions of lightspeed.

[sirocco]

Still, your travel distances before the people aboard die of old age are pretty limited without FTL.

The solution is genetic engineering to slow down or stop aging.

Cheaper and simpler than physics breaking FTL. Or ludicrous STL drives that can reach significant fractions of lightspeed.

There's also the solution of using nanomachines to make human bodies more durable, so than you can freeze then reanimate them without any significant degradation.

[mdiinican]

I pretty much agree with someone_else

In a hard scifi environment, I expect spacecraft to remain rather spindly affairs, with all the pressurized sections being roughly cylindrical or spherical for strength. Staging will probably remain not uncommon. The price of material delivered to the far reaches of interplanetary space will quite likely outweigh the price of an engine and some empty fuel tanks, just like it has for sending things into earth orbit. Every Kg of spacecraft you bring to your destination that you aren't actually using is a kg less of cargo you could bring instead.

I don't think it's unlikely that larger spacecraft might have some onboard manufacturing ability. There are a number of rapid prototyping technologies that can produce metal parts in a variety of alloys, as well as many more that can print stuff like plastic. IIRC, the ISS is due to eventually receive such a machine that works by depositing metal in a fashion similar to the operation of a MIG welder. These devices can get quite small, and may be practical for replacing any number of smallish metal bits on a spacecraft at a net overall reduction of mass when compared to carrying spares of every small to medium sized metal or plastic part that is likely to break at some point. I do not think that the capability to manufacture the majority of electronics on site will be a feature of spacecraft in the foreseeable future. The production of integrated circuits, or future integrated photonics is a rather more involved process.

I'm no expert on nanotechnology, but I figure it will share some traits in common with other things that share the same general size and perform roughly similar tasks, specifically bacteria and fungi. They do seem to be able to adapt to a fairly wide range of environment, though they need a power source, probably in the form of an energy-rich chemical solution. They do seem quite adept at producing interesting chemicals like proteins or enzymes or oils or alcohols given the raw materials plus what they need to live, but I'm not sure how good they would be at building specific macroscale things. Specific breeds could probably be made to produce a certain pre-set shape through processes similiar to developmental biology, but I have doubts about how programmable that would be, let alone how accurate or fast, or how flexible the design parameters could be. Another solution could be to use a process similar to lithography and build things up in layers, but I don't know how much of an advantage nanotech would be at that over more conventional alternatives.

Spacecraft are not liable to have very effective metal armor, not only for mass reasons, but also because of the effects of cosmic radiation. High energy protons and alpha particles from the sun have a nasty habit of causing heavier elements to give off secondary radiation when hit. Steel or depleted uranium components to composite armor like in a tank are probably not worth it when you consider the increased X-ray flux the crew will be getting at all times. Aluminum seems to be about as heavy as you want to go. Lighter elements tend to make better shielding against particle radiation. Plastics and water make good radiation shielding because they are made up of a bunch of lighter elements. Impact shielding is probably best left to methods like whipple shields. The crewed segments of a spacecraft might be constructed as cylinders surrounded by tough fiber-reinforced polymer insulating material that also works as radiation shielding, and that would be surrounded by spaced armor to act as a whipple shield. The insulating layer would also act as the inner layer of the whipple shield and would absorb the vaporized cloud of material produced by the initial impact.

About the detection. You can certainly detect many spaceship sized objects at interplanetary ranges but in all likelihood won't find them. Telescopes look at very narrow sectors in the sky. Tracking a known body and finding something out there in the darkness is totally different. There is a ton of stuff out there right near Earth etc that we simply miss. The overwhelming majority of NEOs, rocks larger than aircraft carriers, are believed to be still undiscovered.

Anyone building off-planet colonies in a realistic setting is probably doing it at least in part because they are concerned about extinction level events happening to their planet, and will therefore probably have more government run dedicated asteroid finding and tracking telescopes than our (zero). Sure, we have a bunch of amateur astronomers finding new stuff all the time, but they have to look through our atmosphere and asteroids and such can be quite dim, especially the carbonaceous or very small ones. Plus there's the matter of having more telescopes up there for keeping track of the elevated levels of dangerous space junk that comes with increased space travel, as well as telescopes for tracking all the actual traffic, for various reasons like space traffic control.

There are only so many viable transfer orbits and launch windows for a given realistic acceleration and maximum change in velocity, so even in the vastness of interplanetary space there would probably be a surprisingly high chance of getting close to another ship. I don't think it would take that much extra software to get all those sorts of sensors to recognize things that are out of place. Any one telescope doesn't have much chance of finding any given thing, but it's just a matter of having enough to cover the whole sky. There's also the fact that there really isn't a whole lot of anything in the solar system above and below the ecliptic plane, so you can mostly ignore those parts of the sky. Sure, you could be sneaky and fly up or below the ecliptic, and try to get the drop on something from there, but that's going to use an awful lot of fuel, and if anyone sees you on your way out, that's going to raise some serious eyebrows.

[Decimator]

In a hard scifi environment, I expect spacecraft to remain rather spindly affairs, with all the pressurized sections being roughly cylindrical or spherical for strength. Staging will probably remain not uncommon.

Spindly is putting it mildly. This particular craft is indeed staged, as well.