User:Christopher Thomas/ThinkStarship

The following brain-dump was posted to User talk:Prometheuspan on WikiBooks, as a contribution to the ThinkStarship project initiated by that user. --Christopher Thomas 04:41, 17 April 2006 (UTC)

Brain dump of material for the "thinkstarship" project

Here's a short brain-dump of items that may be useful for your "thinkstarship" project. I don't have a WikiBooks account, and don't plan to get one in the near future. If you need to contact me, it's best to do so on my user talk page at Wikipedia.

Thoughts:

You're going to need a lot of delta-v to colonize the outer solar system, or to colonize anywhere quickly.
You're going to need at least moderate thrust to colonize anywhere quickly.
Combining these gives a fairly arbitrary requirement of a delta-v of 100-300 km/sec and an acceleration of at least 0.01 m/s². This will let you get just about anywhere in a year or less (Neptune or Pluto in about 2 years).
Relaxing this window to 10-20 years gives you a delta-v of 10-30 km/sec and a thrust of 0.001 m/s², and allows gravity assists. Simpler engines, but much, much more complicated environment systems, so probably a no-go.
Drive options that we know how to build at present are iffy for these constraints:
- A pebble-bed fission reactor that you run hydrogen through as reaction mass gives you a high-thrust drive with an exhaust velocity in the range of 4-5 km/sec when run as hot as you possibly can. This and a 90% fuel fraction give you a delta-v of around 10 km/sec with a single-stage fully reusable craft. Thrust is relatively high (potentially enough for soft-landings on the moon or even Mars, if most of the rest of the mass is engine). Drawback: delta-v is bad.
- A fission-electric engine has as much delta-v as you want it to, but is very low-thrust, due to the power to weight ratio of the reactor. Pulling numbers straight out of my tail gives a thrust of 0.0001 N/kg for 100 km/sec delta-v. You can probably reach ten times this, but a hundred times this will be very difficult. Delta-v and thrust are inversely proportional for this kind of drive.
- I get about 0.001 N/kg for 100 km/sec delta-v with solar-electric in Earth's orbit. This number is on a bit firmer ground than the fission-electric one. Farther out, it's far less useful. It's marginal for a Mars trip, and to do that at all you'd have to reduce delta-v.
The "electric" end of an electric drive would probably be a Hall-effect thruster, as that avoids electrode erosion problems, and is high enough thrust that it's used for stationkeping for satellites, and is well-understood.

The habitat for the trip will contain dried food and medicine as things that _must_ be stored as disposables. Air and water are easily recycled if you're willing to spend a lot of power to do it, but for short trips (1-2 years) it might not be worth the added mass. The mass of air required will be at most about three times the dry weight of the food. The mass of water required is on the order of a couple of kg per person per day, tops.
The colony itself will need several heavy pieces of equipment:
- Power plant. Either solar or fission-based (you can use ship drive reactors for this, but they'd have to be pretty beefy). Solar is only really practical in the inner solar system, without exotic construction techniques.
- Universal smelter. This would be a combination chemical and electrochemical widget that breaks down rock/dirt into oxygen and metals/metalloids, and then separates out desired metals present in few-percent quantities (aluminum, iron, and titanium, mostly). This can be an open-cycle system if you have a source of water to get hydrogen from. A closed-cycle system would do its best to recycle hydrogen, but would still need to be topped up periodically.
- Really good machine shop. This must be capable of fabricating spare parts for everything in the machine shop.
- Airmaker. This is slightly easier to make if it can work in an open-loop cycle. You can get oxygen from the smelter, if on a waterless world, but getting it from water is easier. In a pinch, an air recycler that's periodically topped up will work.
- Watermaker. You'll need to recycle water on a hydrogen-poor world like the moon or an asteroid. You'd do this by distilling what you can out of waste, and then burning the residue in a hydrogen atmosphere and electrolyzing any resulting water. The hydrogen would then be reclaimed (unless you're on a reducing world like Titan). On Mars, you can condense water from the air in trace amounts, or mine it from the ice caps. On any given world, you'll have either lots of oxygen or lots of hydrogen for the taking, given power. The other will have to be conserved if you're not on a water world.
- Greenhouse. This will be Bloody Huge, but can mostly be built from on-site materials. Fertilizer will have to come from nonmetals extracted from your smelter, though.
Humans are an integral part of this system. We don't know how to build a completely automated manufacturing plant, and even if we did, humans would likely be much cheaper.
Some items, like medicines and integrated circuit chips, will have to be imported until the colony grows big enough to have enough of an industrial base to build a hundred-million-dollar facility to produce that type of thing. This also includes things like diamond-impregnated tool bits, really good scientific equipment, and so forth. You're going to need lots and lots of spares for this type of supplies.
Everything else can be bootstrapped with a sufficiently kickass machine shop. The very first things made will be spare equipment for the machine shop (ideally, several machine shops). Second things made will be extra smelters and airmakers and watermakers, so that you can shut these down for regular maintenance or have one fail on you without losing the colony.

Light things that you're going to have will be a kickass library, and a kickass entertainment system, both on the ship and colony-side. You need to be able to train people up to be experts in a diverse variety of subjects, and you're going to need to keep people happy for years on end. This wasn't possible during the Apollo era. It's easily possible now.
You'll want a high-bandwidth communications link to Earth to keep your library and video collection up to date, as well as sending back status and survey and scientific information. Big dishes are cheap, if you're already hauling a hundred or more tonnes of stuff for a colony.

I guestimate the bare minimum colony base being 100 tonnes and 10 people. Add another 100 tonnes of supplies, and you can keep them alive for a few years, covering travel and colony-establishment. Add another hundred tonnes of engine and fuel tank, and 3000 tonnes of reaction mass, and you have yourself a ship that might be buildable. It would cost about $100 billion to launch at current market prices. Minimum manned survey craft is about a tenth this for all numbers. Unmanned survey craft a hundredth this or less.

If you're building somewhere close by, like the moon, you can reduce costs in the short term by sending regular supply craft from Earth, but you'll still need bootstrapping equipment for the colony. Say about $20 billion to launch at current market prices. With a moonbase, colony craft cost goes down to about $30 billion, as you can supply your reaction mass from the moon for far less expense than from Earth.

You'll want chemical rocket shuttles to use as landing craft, ferries, and so forth. If delta-v is in the 2-5 km/sec range, you can do this quite easily with single-stage reusable methane+LOX craft. Both methane and LOX are very easy to make in-situ if you have carbon, hydrogen, and oxygen available, and will in fact show up at various stages in your airmaker and smelter, so you don't need extra equipment to produce them. The components used in these craft are well-understood and are for the most part already in use. These can probably even be built and maintained with colony equipment, though building them on Earth and transporting them (without fuel) is probably cheaper.

That about covers it. For further conversation, contact me at Wikipedia. --Christopher Thomas 04:34, 17 April 2006 (UTC)

First response to brain dump commentary

I'll keep this relatively short, as we could end up discussing this all day without accomplishing much, as we seem to have different views on a lot of things:

You seem to be uncertain about what thrust and delta-v mean. Thrust is the force your engines generate at any given instant. Divide it by craft mass to get acceleration. Delta-v is the total change in velocity that your craft can attain by spending all of its fuel (doesn't matter how quickly or slowly it does this). The delta-v to get between various parts of the solar system is well-known. It depends on what type of course you use (minimum-energy elliptical transfer, or a complicated path that uses slingshot maneuvers, or a very complicated path that attempts to use both liberation points _and_ gravity assists, which seems to be the type you're looking at).

Figure 100-300 km/sec delta-v to get anywhere in single-digit years, 10-30 km/sec to get anywhere in double-digit years using slingshot assists, and forget about the "interplanetary superhighway". That takes far too long to get manned craft anywhere, and doesn't save you that much delta-v.

You're putting way, way too much faith in automated robots. I'm a computer engineer. We have no clue how to build a completely automated factory. We hope we'll know how to do it in 50 years, but I'm pretty skeptical. The NASA proposal for an automated moonbase is mostly useful because you can use it to figure out what you'd have to do with humans instead of robots running the place. Don't be fooled into thinking it's something we can actually build without several very fundamental advances in computer science. I think it was Werner von Braun who said, "a human is the best computer we can put into a spacecraft, and the only one that can be mass-produced by unskilled labour". Humans are expensive to maintain, but they pay off big time in flexibility. If you're doing anything complicated, it'll be done by humans.

Any outer solar system colony will already be culturally isolated by the time delay for messages. You won't need to enforce bandwidth caps on people.

The minimum size for a self-sustaining colony is about 100 people at the time of establishment. This is due to the 50/500 rule for population genetic diversity (you need at least 50 people to avoid inbreeding in the short term, and at least 500 for mutations to feed new genes into the gene pool as fast as they're lost by random selection). I used 10 people as the minimum number of people needed to _build_ the colony, and to keep it running until you can ferry more colonists over.

You need far more greenhouse space than colonist space. Look at the ratio of farmland to city land on Earth to see why. You're not going to be able to carry this on your ship - it's much, much too heavy to have around when you can just pack enough military rations, granola bars, or what-have-you to keep people fed for a trip that takes less than 10 years. Food would mostly be low-moisture, as you can rehydrate it from the water supply, which gets recycled with at least 80% efficiency, and more if you're willing to pay for it.

This is also why your airmaker and watermaker are going to be mechanical, not biological. A biological system doing this, in addition to being suceptible to disease, weighs far too much. The closest we've gotten to doing this was a Russian system that tried to use algae grown on wet canvas to recycle air, and that experiment was done for a relatively _short_ time, on Earth, not in space. You're going to end up with mildew and other crud competing with the algae long before you reach your destination, and will need a lot of algae. The mechanical systems for this are very straightforward to build and maintain, and are lighter than the supplies they conserve for trips over a year or two. They just take power, which you'll have lots of from the drive.

"Reducing atmosphere" means an atmosphere rich in hydrogen, as opposed to oxygen. In a reducing atmosphere, you have mostly methane and ammonia and water. In an "oxidizing atmosphere", by contrast, you have free oxygen, free nitrogen, carbon dioxide, and water. Earth and Mars have oxidizing atmospheres. Titan has a reducing one. If I recall correctly Venus has an oxidizing atmosphere, but it's inhospitable enough to never be colonized (very high pressure, very high temperature, very corrosive due to sulphur dioxide forming sulphuric acid in combination with atmospheric water). Most airless bodies in the inner solar system give you the same chemicals as an oxidizing atmosphere (the moon is aluminum and silicon oxides, mostly, with a few percent iron, titanium, chromium, phosphorus, and so on). In the outer solar system, you get icy worlds, which are oxygen-rich but have hydrogen bound into water. They're perfect for colonization, because hydrogen is probably the thing you'll run out of most.

Don't bother with small asteroids and ring particles until much later. You want something big enough to have gravity, but small enough that you can launch from it cheaply. The moon is about as big as this gets. An asteroid like Ceres is about as small as this gets. The icy moons of Jupiter and Saturn are perfect. Mars is big enough that it's expensive to launch from, but it has all of the elements you need, and an atmosphere, so it's probably also a good colony site.

Launch from any world will probably be via chemical rockets, because they're the only really high thrust drive we have. That's why it's so expensive (Earth is _barely_ small enough that we can get off it with chemical rockets at all). Anything the size of the moon or smaller can be launched from with a reusable LOX/methane craft, which is simple enough to build and fuel with the equipment your colony ship will already have.

Exhaust from the fission drive won't be very radioactive. A small fraction of the hydrogen will get transmuted to deuterium, but deuterium isn't radioactive either. You'll get trace contamination from erosion of carbon and tungsten from your PBR fuel pellet casings, but this is _not_ desired, because it reduces the life of your engine. A nuclear engine isn't quite high thrust enough to launch from Earth, and isn't politically usable on Earth, but anywhere else you don't have to worry about it (as long as jetwash isn't pointed directly at your colony). Background levels of radiation from space will be much, much higher.

Fission drives have to be much more carefully maintained than chemical rockets, so you won't have many of them if you can avoid it. They'll be the big drives on interplanetary ships. All shuttles will be based on chemical rockets. Colonies will have fission plants, which were probably the drive reactors from the ships that founded the colonies, plus extra equipment for electricity generation if the drive was originally a NERVA-style hydrogen-through-PBR design.

Launching from an airless world will in the long run be done with a mass driver (think "maglev train run at several km/sec off an open-ended track"). This is expensive enough to build that it won't happen until your colony has a GDP in the range of at least $1 billion, but it's how you'd get large amounts of building material for a permanent space station colony or what-have-you. You can't do this from any world that has an atmosphere, even one as thin as Mars's.

Mining asteroids is a royal pain in the tail. If at all possible, mining will be done where there's gravity and an atmosphere. Planets are great big balls of metals and useful minerals, so I just don't see the advantage. The only exception is that you might break off a tiny chunk of an icy body and tow it to an established colony on the moon, to give you carbon, hydrogen, and nitrogen. It might be cheaper just to haul up refined chemicals from a colony elsewhere and ship them, though.

Almost all of the mass of an interplanetary ship will be mass that has to be thrown away: reaction mass used by the drive. Even for an electric drive, travel time requirements mean working with as low an exhaust velocity as you can, which means higher thrust but higher mass consumption. Take the mass of anything that doesn't have to be thrown away, and multiply it by 10 to get the reaction mass you need to carry. That's why the colony ships I specced had 300 tonnes dry weight but 3000 tonnes of fuel. For a high-thrust nuclear drive, the reaction mass has to be hydrogen. Otherwise you lose the exhaust velocity advantage over chemical rockets. For a Hall effect thruster, it can in principle be anything, so you can use oxygen or nitrogen or what-have-you from lunar rock. However, I don't think a nuclear-electric craft will get you to the outer solar system in a reasonable length of time if it's a colony ship. Colonizing incrementally doesn't help you much - you still have several AU between outer-system planets, so every step is a very big one.

Equipment breaks down. Complicated equipment breaks down faster. This is why I don't think automated colony ships will work (besides being incredibly complicated and ludicrously expensive), and why I don't think trips with a travel time above about 5 years are practical with humans on board (the systems maintaining the environment are as simple as we can make them, but still pretty complicated).

Cultural problems and mob-tyrrany among the crew are far less of a problem than simply keeping them from getting bored out of their skulls, or hating each other because they're all effectively sharing an apartment that they can't get out of for years on end. The key to this, besides carefully screening crew for conflicts before launching, is to give each person personal space they can retreat do and rearrange, and to give everyone the best library, computer, gaming system, and TV you can. These are relatively light, and let people entertain themselves when they're sick of the group. You'd also have a "living room" type of space with an entertainment centre as well, for group entertainment, as well as board games and what-have-you for social relaxation. There'll be a (compact, lightweight) gym, to reduce bone loss and similar problems, but most people don't entertain themselves physically that way, and a facility for swimming or sports is far, far too heavy to put on a spaceship.

Similarly, happily-married couples are a good idea. Otherwise you'll get sexually-driven tension between crew members either on the ship or once the colony is established.

To avoid health problems, the ship will consist of three big containers. The central one has the drive and fuel tank, which don't rotate. The other two are connected to the central hub by cables, and spin around the hub for gravity, counterweighting each other. One has the crew and everything they need on the trip. The other has everything they don't need on the trip but will need when they establish the colony. If you try to keep humans in zero gravity for years on end, they won't be much good to you when they arrive at the destination. The interplanetary drive doesn't have thrust high enough to produce noticeable gravity.

All of these points pretty much _have_ to be the case for a successful colonization project. Otherwise you end up needing a ship far more massive than you can build, or far more complicated than you can build, and probably far more expensive than you can build given either of those two, and one that takes long enough to reach its destination (any destination) that it's very unlikely to arrive with the crew alive and ship intact. I hope you see why I made each of the specifications I did above. --Christopher Thomas 22:50, 17 April 2006 (UTC)

Second response to brain dump commentary

Regarding space drives, you seem to have some odd ideas, and to be grouping several different ideas together. Specifics of the different drive options are:

Chemical rockets. The best delta-v you can get for reasonable fuel amounts is about 5 km/sec. Maximum acceleration is several gravities (say about 30 m/s^2). These are used for your shuttles only, and probably for launching equipment from Earth.

Nuclear-thermal. This involves making a pebble-bed reactor core (a bunch of marble-sized spheres of uranium clad with graphite and coated with tungsten), letting it heat up to about 2500 degrees C, and running hydrogen through it. The hydrogen heats up and sprays out the end of the rocket. Because hydrogen has a much lower molecular weight than the products of rocket fuel combustion, it's moving much faster at any given temperature. This gives you a delta-v of around 10-15 km/sec, at most, with a reasonable amount of fuel. Maximum acceleration is probably in the 0.01 to 0.1 gravity range (about 0.1 to 1 m/s^2). If you're willing to accept a travel time on the order of 5-10 years, you can use this kind of engine for your ships, and use gravity assists to get you where you're going. You don't have enough delta-v to get anywhere else (other than the moon) in a shorter period of time. Outer solar system could take more than 10 years.

Nuclear-electric or solar-electric. This involves using either a nuclear plant or a solar panel array as an energy source, and using an electric drive to move a working fluid (usually a heavy noble gas like xenon, for ease of handling and ionization, but it can be almost anything else you can turn into a gas). Options like ion drives and older-style plasma drives have problems with the electrodes eroding, but a Hall effect drive will work quite well. The downside is that thrust is very low. This is an unavoidable problem, because the thrust per unit power goes down inversely as your specific impulse (and so delta-v per weight of fuel) goes up. You'd need a power plant with a very high power to weight ratio (lightweight, high power) for this drive to be useful. It's probably still the only option for outer solar system colonization, but trips could take decades.

The "nuclear drives for small craft" you'd heard about were probably radiothermal electric generators powering electric drives. These aren't reactors; they're subcritical amounts of material that decay fast enough to get warm. Thermocouples draw power off of the heat gradient that results. They're fine for extremely light probes that are allowed to take decades to get anywhere, but aren't useful for small craft. A "reactor" is an assembly of nuclear fuel that catalyzes its own decay, producing much higher heat output for a given amount of weight. They're useful, as noted above, but have a certain minimum size (a few tons for anything shielded).

These drive designs are well-studied enough that we could build them right now, with off-the-shelf parts, if we wanted to. More exotic options you seem to have been made aware of are:

Nuclear pulse propulsion with bombs. This involves building a ship weighing hundreds of thousands of tonnes with a big shock-absorber under it, flinging nuclear warheads in front of the shock absorber, and setting them off. The design (Orion) was studied extensively in the 1950s and 1960s, but is outright impossible to implement for political reasons. You need to amass a _really huge_ stockpile of warheads, and the ship itself is a mobile bomb-thrower with immense destructive potential, so nobody will let anyone else build one, and the radioactive plume out the back prevents it from being used for ground-to-orbit work, which is what it's mostly useful for.

Nuclear pulse propulsion with inertial confinement fusion. This involves building a small ship with a magnetic field acting as a rocket nozzle, firing pellets of frozen deuterium into the field pinch, and zapping the pellets with lasers or particle beams to initiate fusion. The problem is that we've been trying to build inertial confinement fusion systems for decades, and still don't have one that produces a beneficial amount of power. The pellets are also extremely difficult to make and handle, due to symmetry requirements, and the laser installation is the size of a large building, not counting power plant for the lasers. So, this isn't happening any time soon.

Bussard ramscoops. You keep mentioning these, but you don't seem to realize where they were intended to function. Even in Bussard's original models, you have to be travelling extremely quickly for them to work (delta-v of hundreds of km/sec needed to boost to that speed before ignition). They were meant as interstellar craft (launched, expensively, at the speeds required, and then boosting to much higher speeds). The problem is that this type of drive doesn't actually work at all. First, you're trying to fuse the trace amounts of deuterium in the already-very-tenuous amounts of hydrogen in the interstellar medium. This means most of the hydrogen that passes through your drive is inert, causing problems but giving no benefit. Second, the magnetic field ends up deflecting far, far more hydrogen than it pinches to fusion density. This means that even if you built a ramscoop and launched it at the required speeds, instead of igniting and speeding up, it actually just plows into the interstellar medium and slows _down_ without igniting. The drive was popularized in fiction dating from the middle of the 20th century, but will never actually be possible in practice for this reason. Even if it did magically work, it wouldn't be useful for colony work, for the reason mentioned above (you have to be going fast enough to use it that you don't need it if you have the means to get that fast in the first place).

Additional points:

Regarding moving out in "waves", you don't seem to realize exactly how empty space is, and how far apart useful sites are. The moon is the only thing we can reach quickly from Earth. Getting anywhere else, at all, means changing radius by 0.5 AU going in to Venus (which we don't want to, for reasons mentioned previously), or 1 AU out to Mars. Large bodies in the solar system are more or less set up in a geometric progression, so getting from anywhere else to anywhere else involves travelling at least 2 AU and usually more. Any of these trips takes years or longer, depending on delta-v available (you probably don't have the delta-v to do a straight transfer orbit for objects in the inner solar system, so you have to use an Earth or Moon flyby after one complete orbit at Earth's distance from the sun before going anywhere). Any of these trips also takes large amounts of delta-v to do at all, as noted in my first post.

Think of the Earth/Moon system as a double planet, for this reason. It's one location, and travel from Earth orbit to Lunar orbit is easy (though surface launches from Earth are prohibitively expensive, and soft-landing on the moon or lifting off from it takes quite a bit of fuel).

It's not just human health that's affected by microgravity. Microgravity mucks up just about any industrial process developed on Earth (i.e., that we have significant experience with and understanding of). If you're building any kind of machinery, including your airmakers and watermakers and refinery/smelter, you want it to operate in gravity if at all possible, even if it's low gravity. This is why you'd colonize the moon to mine material for large projects, and colonize moons of Jupiter, and Mars, and only the really big asteroids before even thinking about doing anything with the small rocks in the belt. In the outer solar system, where water can exist as ice in vacuum, you might break off chunks of small icy bodies to tow back to refineries on moons, but you'd only do this if you didn't have material already available, and just about all of the outer solar system moons have ices already.

For pack psychology vs. tribal psychology, remember that you're only dealing with a small group for as long as it takes to establish the colony infrastructure. Each subsequent ship, or subsequent ferry run if it's one ship moving back and forth, brings around 30 more people, because you're not hauling colony infrastructure equipment. As long as the small group only has to last 5 years or so, you're fine. This is why a fast transport is desirable (though not necessarily _possible_).

You also seem to be overlooking exactly how nasty small-town politics gets. As long as the group is small enough that everyone knows most of the others, or a large fraction of the others, social structures freeze in place. The threshold for this seems to my uneducated guess to be around a thousand people. This also determines how unusual someone can be and still find like-minded people (if your hobby or mindset is 1 in 100, then you can find like minds in a 1000-person town, but not if you're at 0.1% or less). Large towns of 10,000 or more would be required to reduce these effects. Either way, you're not going to haul that many colonists over, so in practice you're starting with 50-100 people, and hoping they sort themselves out after a few generations of large families.

This is another reason why communication with Earth is a good thing. It gives anyone alienated on the colony a support network of like-minded individuals (for better or for worse).

Growing food by hydroponics has just as many problems as making a biological airmaker: it takes a lot of power (plants are inefficient), and is suceptible to infection and other maintenance problems. The weight in dried food lasts quite a while and can be stored for years. You'd only start growing things when you could get light for free (a greenhouse on an inner-system world). Outer-system colonies would either need large concentrator mirrors, or would need to spend a very large amount of power, or would just have supplies shipped in periodically. On a spaceship, it's not volume that's a problem (for things that don't need shielding), but weight of the apparatus, and amount of power needed (which means weight in the power plant).

You'd only use a maglev launch system on a body with an escape velocity of about 1-2 km/s or more. This means the moon. Mars (like Earth) would _benefit_ from one but can't build one because of the atmosphere. Smaller bodies don't have an incentive to build one, because chemical shuttles work fine. Very small bodies, with escape velocities on the range of 100-200 m/s, would use short launch tracks of the type found on aircraft carriers on Earth, but that's a very different type of system (much cheaper to build).

The only worlds where a space elevator makes sense are Earth and Mars, because they're rotating quickly, have atmospheres preventing magnetic launches, and are massive enough that chemical rockets are very expensive to use. They're very big, very expensive to make, and very annoying to maintain, so if other options are usable, other options would get used. In practice, Mars is the only place I'd expect to see one, because an elevator for Earth requires materials we don't know how to make yet.

A rotating base colony-side is an interesting idea, but you'd avoid it due to mechanical annoyance if possible. Any world with 0.1 gravities or more would keep humans adequately healthy, though they'd have trouble if they went down to Earth. Ceres is 1000 km wide. I'd be reluctant to colonize anything smaller than 500 km, due to the difficulties microgravity causes industry. Typical outer solar system moons are thousands of km wide.

You want rotating parts of the ship to be on as long a baseline as possible. This makes the rotation rate as low as possible. Quickly rotating ships cause motion sickness problems, and sometimes mechanical nastiness due to the coroilis effect. This is why I specced the crew compartment to be at the end of a long cable. You only need one other rotating module, as a counterweight. You don't want the drive module to move relative to the ship's centre of mass, so it can't rotate. This gives you three modules, per my previous outline.

We're going to have to agree to disagree on the "armies of robot workers" thing. They'd be far, far more complex than any system we've designed to date, and the expense and difficulty of designing a system goes up extremely quickly as complexity is increased. The chance of it actually working goes down very quickly as complexity is increased. This is why I'm not optimistic about the technology for fully autonomous, self-replicating factories and robots being developed within any useful planning horizon. You're correct about it radically changing how you'd implement projects, but there are lots of technologies that would have effects like that. None look like they're happening in the near future.

I hope this explains why I made some of the statements I did previously. --Christopher Thomas 130.63.92.157 23:33, 19 April 2006 (UTC)

Third response to brain dump commentary

First, sorry for the delay getting back to you. I have been, and am still, extremely busy.

Regarding your most recent responses in the ThinkStarship thread:

I'm feeling pretty sensitive to "odd" ideas lately. I have some good ideas and some mediocre ideas. All of my ideas are pretty well covered by science on the one hand and my willingness to allow process to be what it is, not what i assume on the other. Please forgive me if this sounds annoyed, but all i am doing here is trying to get the experts together to get them to solve the problems, not pretend that i allready have the answers.

The thing is, some of the design choices you've made assume that you already have the answers. I'll try to give a better explanation below.

This is all very interesting and i hope that you will consider making a presentation of this kind of information in the actual document.

Regarding me adding material to WikiBooks for this project, I'm actually not sure this is the best place to put it. You're pitching this as a textbook, but what it actually appears to be is a "design study", which is a different type of beast. Further, anyone with sufficient skill to be in a position to help actually implement a space colonization project will already have expertise in the areas you'd be teaching about in any textbook collection associated with ThinkStarship. I can see a WikiBooks project happening after the design study is done, to teach people about it, but what would best serve the interests of the project at present would be just improving existing texts on orbital mechanics, spacecraft design, existing proposed space habitat designs, ecology of closed systems, and so forth. This sidesteps the question of whether or not ThinkStarship belongs on WikiBooks by restricting contributions to things that definitely _do_ belong here.

If I find myself with large amounts of free time, I'll probably put an expanded version of my comments here on my own site, but that's not likely to happen any time soon.

I think that this (nuclear pulse propulsion) is the primary form of "nuclear powered" rocket which i had been exposed to, but it was described very differently from what you have here. The bombs were going off almost inside the engine.

Then it wasn't an Orion-style drive. You might be thinking of inertial confinement fusion drives, that cause a pellet to undergo fusion and explode, but I've already commented on them. We can't presently build them, and any technology for them in the planning horizon you're using would be too heavy to use for a space drive.

Thats not true, my one claim to reasonable knowledge is that i found out 20 years ago basically what you are saying (about Bussard Ramjets) and was so heartbroken than i tinkered the idea up a lot to make it work in any case. My ramscoops aren't used for much in most cases other than to get more speed once allready moving, and they are nuclear powered magnetic/plasma driven, which means that they double as a magnetic thrust system.

Again, what you are describing isn't a ramscoop. Any ramscoop that tries to collect hydrogen from the interstellar or even interplanetary medium has to be going fast enough that you don't need it for your colonization project, and _also_ gives a net loss for the reasons mentioned for Bussard drives.

There are "nuclear powered magnetic/plasma" drive designs, but I've already covered these under "nuclear-electric" drives. The closest to what you describe seems to be the VASIMR drive, but any electric drive will have similar performance characteristics. I like Hall effect thrusters because they can be bought off the shelf right now and are very well understood, but both VASIMR and Hall thrusters avoid the erosion problems ion and classic plasma drives have, and so would be suitable.

This isn't true, i do know how big space is, and, this is why i am talking about asteroid colonization, you see, a key component of the plan is to piggyback onto objects that are allready traveling around the solar system. Cruithnes closest pass is only abit farther out than the moon.

Then you aren't factoring in how long you have to wait for your orbits to synchronize with each other. A requirement of any realistic colonization scheme is that you get from point A to point B in a short enough time that your environment maintenance systems don't break down. I'd handwaved this at 5-10 years in my original estimates. Unmanned probes can last a lot longer because they aren't as complicated and don't have to perform continual recycling. Colonies can only get away with longer lifetimes after being founded because they have a large amount of matter and energy freely available, making partially open-loop systems possible and relaxing power constraints. Your ship can't afford this, which I explain in more detail below, and you'd be stuck on the asteroid for a very long time before it passed close to another object you were interested in.

I don't think trying to avoid the problem by colonizing these _small_ asteroids is feasible. You have the problem of trying to run industry in microgravity, as previously mentioned, as well as lacking hydrogen (in the inner solar system), and having a higher rate of mass loss from your colony (matter lost goes away, rather than staying on the body you've colonized). You can't just spin the colony habitat - industry will be spread out all over the place, especially for power generation.

i have contemplated these problems and still consider going after small rocks to be the best early tactic. I am not alone in this, a lot of other people have similar ideas.

I've heard lots of people propose this, but haven't heard any reasonably sound plans for _how_ to do it. Certainly the claims of massive economical paybacks from the ones I've heard aren't convincing (we're already sitting on top of a ball of metal with lithophillic elements pre-concentrated). We already know how to build colonies and industry on larger bodies, so I fail to see why we wouldn't have these as the first targets of colonization.

Again, I am not a physicist and i don't do algebra. I have on the other hand studied ecology sciences in pretty good depth, and I think i know what is possible. I'd be going for a garden as soon as the ships got into space because i know thats the best way to do it.

This is the difficulty that I mentioned a couple of times above: you're assuming that something is the best solution without working the math to see when it makes sense and when it doesn't.

I ran numbers last week for recycling of air, water, and food, and got the following. The key is that it's _not_ free - you at the very least have to add power plant weight to handle the recycling load, and for air and water, that's the dominant constraint. Breakeven of stored supplies vs. recycling happens when supplies weight exceeds the weight of the recycling equipment (assumed here to mostly be power plant weight unless otherwise noted).

The power plant is assumed to produce 100 W/kg. This is quite good for a closed-loop system; I was assuming 10 W/kg for my original posts. Cars get about 300 W/kg of engine-related system, but are open-loop. The higher the power to weight, the more maintenance it's going to need, and you can't do much maintenance on spaceship turbines sitting next to a live fission reactor, so 100 W/kg is probably _optimistic_.

Recycling oxygen takes about 20 MJ/kg (mostly the cost of electrolyzing the water produced as an intermediate step). Humans consume about 1 kg/day (based on figures at Breathing on Wikipedia), or about 1e-5 kg/second. Power required to recycle oxygen is therefore about 200 W per person. This corresponds to an added power plant weight of 2 kg. Breakeven happens after 2 days for our plant specifications, or 20 days for a more conservative plant, so air recycling is desirable. Reclamation of hydrogen from the methane produced isn't perfect, but it's close enough that this figure isn't thrown off much by the need to store extra water for hydrogen production.

Recycling water takes a bit more than 2 MJ/kg (mostly the "heat of vaporization" of converting it to steam to distill). Humans consume about 2 kg/day on average (based on figures at Drinking on Wikipedia), or about 2e-5 kg/second. This includes water in food, but most foods in storage will be dry, so it's a good estimate. Power required to recycle water is therefore about 40 W/person. This corresponds to an added power plant weight of 0.4 kg. Breakeven happens in hours for our plant specifications, or about 2 days for a more conservative plant, so water recycling is desirable.

Producing food requires between 0.1 and 0.3 hectares per person, based on the best yields for crops given in Wikipedia's articles on agriculture in the first world. Power needed to light that crop area is an average of about 200 W per square metre (averaging over the day/night cycle and angle change for illumination on Earth), or 200 kW per person under the most productive possible conditions. This requires about 2000 kg of added power plant per person at 100 W/kg. Dried food consumed by a human is about 100 kg/year (0.3 kg/day), so breakeven against storage happens in 20 years for our power plant specifications, or 200 years with a conservative power plant, even if we ignore the weight of the crops and the hydroponics trays. Even the shortest breakeven time estimate is much longer than the estimated length of the trip, so producing food on the ship isn't practical.

You'd grow food on the ship if you were travelling between the stars, because the trip time is much longer than your breakeven time. You'd also grow food at a colony, because you get most of the weight of the power plant from local materials instead of materials you have to haul with you, making the breakeven (for transported materials vs. transported dried foods) much earlier. You wouldn't use it on an interplanetary ship, because carrying dried food is lighter.

It's important to be able to calculate this type of tradeoff, as it tells you what you can do under any given set of conditions, and what conditions you'd need to do any given thing.

It (a rotating colony) solves more problems than it generates.

I think you're underestimating the engineering problems it generates. A ship can do it, but a ship doesn't have to touch anything, and only has to do it for 10 years. A colony has lots of nonrotating infrastructure (for mining and power generation), which suffers the microgravity problems noted above and previously, and which your rotating colony has to interface with.

It's not impossible, but I don't see why we'd willingly do it.

Muscles atrophy probably in less than .4 G.

I don't see why muscle atrophy is a problem. Colonists would only have to move around at 0.1 G. I'd worry more about malfunctions of the body's internal organs or development problems for children born at the colony, and the numbers I'd heard cited for that said 0.1 G was acceptable. This lets us colonize all of the large gas giant moons, though unfortunately not Pluto or Ceres. It also provides enough gravity that heavy industry won't have insurmountable problems.

I think that you are missing the point. I am a realist and a pragmatist. I know that there are serious problems to be solved. I am anxious for us to start working to solve them. If we don't have operational robots in time to lanch the first ships, then we will have other things to do wha they might have.

My point is that you should assume that you're _not_ going to have operational autonomous general-purpose robots by launch time. The NASA study's estimate for a self-maintaining autonomous general-purpose facility was that it would be complex enough to require 1e+11 bits to describe the design. A human being, on the other hand, has a design stored in about 1e+10 bits. So, please realize that you're proposing to design from scratch a system ten times more complicated than a human being, when we can't even design a self-replicating single celled organism from scratch right now. It'll _eventually_ be something we know how to do, but if you want colonization in full swing in 50 years, you'll need to work with things we know how to build _now_. Going from "design idea using known parts" to "fully functional spaceship ready to move colonists" takes multiple decades. Going from "first colony ship launched" to "multiple colonies established and mostly self-sufficient" also takes multiple decades. Taking any technology from "first experimental demonstration" to "something that can be reliably used" takes at least a decade for things that are fundamentally _simple_. A system as complex as the robots you propose would take vastly longer, for reasons mentioned in my previous post (design difficulty and debugging difficulty go up worse-than-linearly with complexity). Any book on systems engineering or software design can explain why this is the case in far more detail than I am doing here, if you don't want to take my word for it.

In summary, to be realistic and pragmatic, you have to have tools that let you assess what approaches work best for moving your colony ship and keeping it habitable, and you have to assume that you'll be working with technology that's either widely available now or that has fully working prototypes now if you want colonization to occur in your intended timeframe.

--Christopher Thomas 00:52, 9 May 2006 (UTC)