> Putting something radioactive on the launch pad and having it detonate in the atmosphere would be terrible too (Which is why we don't send nuclear materials into the sun.)
That's not why we don't do it. I worked on a "Space Disposal of Nuclear Waste" study at Boeing, under contract to the DOE. The risk reduction (about two cancer deaths a year on a statistical basis) was simply not worth the extra cost (about double that of burying it underground). Also, the Sun is not the safest place to dispose of it. If your rocket fails and leaves it crossing the orbit of Venus or Mercury, they could send it back to Earth by accident. The lowest risk is to place it in an orbit halfway between Venus and Earth (0.85 AU), and that also takes much less delta-V than hitting the Sun.
The nuclear waste was assumed to be glassified into coke-can sized segments, then formed into 2-meter "waste balls" surrounded by 20 cm thick steel alloy, which in turn was surrounded by heat shield tiles. The worst case accident is no on the launch pad, which is merely a lot of fire. The worst case is the rocket failing just before reaching orbit, where the payload kinetic energy is twice that of the best rocket fuel. So the heat shield enabled surviving re-entry, and the thick steel shell enabled surviving a terminal-velocity ground impact. It was also a corrosion-resistant alloy, because most launch failures end up dropping the payload in the ocean. We assumed a 2% launch failure rate.
The waste balls were so damage-resistant, that the study manager would have been happy to take one home and put it in the basement to keep the house warm in the winter (they generate 2 kW from radioactive decay heat). House fires, natural gas explosions, earthquakes, none of those would do any damage to it.
> the basic food for plants without which plants would not grow and we would all starve to death is not a bad thing.
Too much food, or anything for that matter, is a bad thing. We humans need water, but too much of it is called drowning. In the case of Carbon Dioxide, too much is a poison for people. Occupational guidelines are not to exceed 0.5% for extended periods.
There is no competition. Cotton farming in Alabama, for example, where I used to live, is highly mechanized. They spray the fields by airplane with herbicides to kill the leaves. Then the harvesting machine chews up the remainder of the plants. Stems are denser than the light and fluffy cotton bolls, so they get separated by air, and the stems are discarded. When the harvester is full, it dumps the cotton into a baler, making bales larger than a tractor-trailer in size (cotton is light). The bales eventually go to the mill to be processed. Around three people can harvest hundreds of acres at at time.
I used it as a long-term investment. I accumulated them from 2011 to earlier this year at an average price of $76/btc. Now that it has gone over my price target of $600/btc I'm selling. Unless the current block size logjam is broken, I don't anticipate it will go up that much more in the near term, so I will keep selling until it's gone.
The block size is currently limited to 1 MB/10 minutes. Blocks are sets of transactions which are secured by a series of special hashes that are chained together (hence the name blockchain for the whole transaction record). 1 MB can fit a few thousand transactions, and ~250,000 transactions per day simply can't serve day-to-day uses like coffee at Starbucks. It's fine for my purpose, or things like international payments and remittances, where bitcoin is far cheaper than bank wires or services like Western Union. Unless the limit is changed, it will have to be "second layer" payment services to handle small daily transactions, and bitcoin itself used for settlement between such services. This is similar to how banks use clearing-houses to settle up payments between themselves, on behalf of their individual customers. In other words, your bank doesn't send money to Walmart each time an individual shops there. Instead, they batch together all the payments for the day from your bank to Walmart's bank, and in turn that bank forwards the right amount to Walmart's account. Since Walmart's bank is doing the same thing to pay employees and suppliers, the daily clearing is a single payment between the two banks for the net amount going in both directions.
Bitcoin only reached filling the 1 MB of transactions per block on a regular basis earlier this year. People are working on second layer solutions, but they are mostly not ready yet.
No. The high gain antenna (dish) is not pointable now that it's attached to the comet, and may have been damaged on landing. Ditto for the solar arrays.
No, the liquid Oxygen is delivered as liquid on trucks, and stored in large tanks at the pad as liquid. The Helium would be cooled to LOX temperature by virtue of being inside a tank full of the stuff. This would lower the pressure of the stored helium, allowing you to put more in the tank, but it's not cold enough to liquefy.
One possible failure mode is something preventing the Helium from cooling down, in which case it could overpressure the tank and it blows up. That could be a problem with getting the LOX into the tank, bubbles around the He tank, etc. Or it could be something simple like a flow valve fracturing. They have all the telemetry data, so I can only speculate.
Nah, that's too obvious. A sniper rifle fired by a CCAFS security person with money troubles is less conspicuous. Security staff have a reason to be on the base, even patrolling the launch pad area. On the other hand, there's a whole lot of nobody else around the launch pads, for safety reasons. So all he has to do is find a good spot, pop off a shot, then drive over to the launch pad like a concerned security guy would do when something goes boom.
Why money troubles? The people with a motive, like United Launch Alliance, could pay off someone for a whole lot less than what they stand to lose by SpaceX eating their business. Even a six month delay and a few customers moving payloads to "spread their risks" is worth a billion or so in revenue.
> although they feel the timeline is too ambitious.
Musk operates on the Martian calendar, so everything takes 1.88 times longer when you convert to Earth years. Once you correct for that, all his timelines work out pretty well.
Luxembourg is a forward-looking country. They invested in communications satellites in the 1980's, and now operate the largest commercial constellation of satellites. Recently, they started investing in asteroid mining, and they are also a SpaceX customer. I don't think Musk is so dumb he didn't know a big rocket could go other places than Mars. I think what's happened is he has a customer who is *interested* in going other places than Mars. And he needs lots of commercial customers to help pay for the big rocket he wants to build.
Look, he's developing the Raptor engine ( https://en.wikipedia.org/wiki/... ) Assuming he uses 9 of them in the first stage, like the Falcon 9 has, that's 20.7 MN of liftoff thrust. Liquid rocket T/W on liftoff is typicall 1.3:1, which gives 2100 tons liftoff mass. A good chemical rocket typically has 4% payload mass, so 84 tons payload to LEO. All of that follows directly from the engine size and how many you use.
If you can put that much mass into Low Earth Orbit, you can get variable amounts of payload to different higher orbits. This is obvious to anyone who has much experience with rocketry. User handbooks for different launch vehicles have graphs showing the payload as a function of mission velocity, and that velocity is set by where you are going and the trajectory you follow. Perhaps Musk is slow to realize this, because of his focus on colonizing Mars, but it's no surprise to people in the industry like me, and probably to a lot of the people working at SpaceX either. I can imagine the staff meeting at SpaceX:
Musk: You mean this giant rocket we're building can go other places than Mars? Staff in unison: No shit, Sherlock.
The one in the movie 2001 was buried. They found it because it generated an anomalous magnetic field, hence the name "Tycho Magnetic Anomaly One", or TMA-1
In this case, it was Powell's g-mail account. His login credentials may have been reused on some other site that had their data stolen. In that case it would not be a hack so much as a stolen password.
Sorry, but this is wrong-headed. We have to start working on it *now*, so that in 50 years we have the experience to build the space factories. It was 60 years from the Wright Flyer to the 747, but you can't skip all the steps in between.
But "fully automated self-replication" is both a limiting concept, and *hard*. There is no reason you can't make different machines than the ones you start with, or different sizes. So a "starter set" can be smaller and simpler than the final factory. All the complexity is in the stored computer files that tell it what to build. There is also no reason that it has to be 100% automated and make 100% of its own parts. Those are theoretical ideals like 100% efficiency. We can tolerate some manual labor and buying parts and materials from outside. The only real requirements are to be efficient enough to compete with conventional manufacturing, and have enough surplus production to pay for the things you can't make on your own.
We currently have $95,000 of income producing assets and a 3 acre R&D location being developed. We are an *open source project*, not a venture capital startup, so we don't have paid staff, at least not yet. People contribute their time and funds to the project, and we do the best we can with it. Our workshop won't house all the machines and tools we need to build our prototypes. For that we rely on a network of makerspaces, individually owned equipment, university labs, etc.
We don't have a timeline for reaching a complete prototype factory, we haven't even finished conceptual design. For the moment the work is fleshing out the concepts, and renovating the workshop space on our property, so when we do have hardware to test, we will be able to. Things like solar furnaces need sunlight, which is why we have 3 acres. You can't test them indoors.
I'm a retired aerospace engineer. I devote my spare time to this project because it interests me. Other people's motivations are their own, I can't speak for them.
> Until such a thing exists this is just fiction. As far as I can tell, at the moment all plans for building a self-bootstrapping automated factory on Earth, much less the actual operating factory, are fiction also.
Industrial automation is a thing, and has been for decades. We don't have to reinvent that part. What makes a seed factory different is the CAD/CAM files include making parts for more machines, besides the salable end products that any factory makes. Again, most of this has already been done, machine tool makers use their own machines that they made to produce more machine tools to sell to others. Robot makers use their own robots in their robot factories to make more robots. So our design problem is coming up with a growth path from a small and simple starter set (we presently have 8 elements in the starter set) to a complete factory capable of producing new starter sets.
And because people like you justifiably question the ability to do this, our project goal is to design and build prototypes to prove it can be done. We are not done yet, far from it. But you have to start somewhere, even if it takes 10 or 20 years to reach the goal. You will note that unlike Mars One (who I criticized myself, here on Slashdot) we don't give some fictional completion date. All we say is "here's the idea, here's what we've done so far, we are working on it".
Also space is full of high levels of radiation, and it will scare them off. We only exist here on the Earth's surface because of the ozone layer, depth of our atmosphere, and magnetic field.
Any aerospace engineer, but apparently no member of Congress, knows the right approach to get most of the cost reduction:
* Stop throwing away several kg of aerospace hardware @ ~$1000/kg every time you launch. *
There's roughly 160 MJ/kg of fuel energy in a good conventional rocket, which results in 1 kg of payload with 31 MJ of orbital energy, so around 20% energy efficiency at best, and often 10% in not so good designs. But propellant is *cheap*, around $1/kg. By far most of the cost is the hardware.
Outside of NASA, most of the new rocket development involves re-using the hardware. The Space Shuttle was intended to save money by reusing most of the hardware, but the program was poorly designed and run and didn't achieve cost savings. They did, however fly hardware multiple times.
A 6 km/s tip velocity skyhook is not an optimum design at present. When you do the actual numbers, it comes out closer to 3 km/s, and the remainder is supplied by a ~4 km/s single stage rocket. The totals are not the same because the faster skyhook is larger, and has a higher center of mass. In turn, that means slower orbit velocity. Also, 3 km/s is sufficient to reach high orbits from low orbit, and that's all you really need. Finally, skyhook mass is highly non-linear in tip velocity. Assuming carbon fiber with conservative safety margins, you can allow 150 g-km stress.
The 3 km/s skyhook has 460 km g-km of stress, and mass ratio of one side (center to tip) is exponential in stress ratio, so e^(460/150) = 21.5. Since it has two arms, the total mass ratio is 43:1. The 6 km/s skyhook has 4 times the stress - 1835 g-km. So the total mass ratio becomes 205,500:1, which is unworkably high.
A 4 km/s single stage rocket is *easy* compared to reaching all the way to orbit all by itself, and you can dramatically improve the design margins so it can make many flights. The rocket then bears the same relationship to the elevator as an airplane does to an airport.
I've spent 39 years doing space systems engineering, and know more than most about getting to orbit. I *will* talk about space elevators, having taught a class about them last year. But not space elevators "as we know it", to paraphrase Spock. The ground-to-60000 km single cable version that most illustrations show is unworkable, even with carbon nanotubes. A feasible version uses two rotating cables, one in low orbit, and the other in high orbit, with nothing between them but orbit mechanics. Their combined length is under 2000 km, and can be built with today's carbon fiber. You still need a way to get from the ground to half low-orbit energy, but even a chemical rocket can do that, easily, with good design margins, and a single stage.
But even that type of elevator isn't justified yet by traffic rates. A space elevator is "transportation infrastructure", like a bridge or airport. You don't build those for a few trips a month, and neither do you build a space elevator for a few trips a month. But a few trips a month is all the space traffic you can get today. So before anyone thinks seriously about elevators, we have to work through more conventional ways to reduce cost and grow the market.
Automated factories already exist, and have for many years. What's different about a seed factory is having a planned growth sequence from stored CAD/CAM type design files. They also make useful end-products from the same type of design files, like current automated production follows.
> What cheaper method of getting out of Earth's gravity well can be implemented now?
Enough people are working on that problem (as I used to do at Boeing's space systems division) that a few more people won't make a material difference. Changing how much stuff you need to launch, by bootstrapping production from a starter set, has a lot of leverage, and not many people are working on it. That's why I choose to spend my time on it. No, you can't order a seed factory kit from Amazon at the present time. Neither can you order an iPhone 8 or Galaxy Note 8 yet. They have to be designed and prototyped first, which is what I work on.
> How much cheaper is it than how it's done now?
The potential is to get 98% of the mass of future space products from space. The other 2% is hard to make items or rare elements for which delivering from Earth is easier. The savings are therefore 50x in reduced launch cost, minus the cost and launch mass of the starter equipment. Nobody knows the final cost, because the whole concept is too new. The *potential* is large enough to justify working on it, just like the *potential* of integrated circuits was in 1960. But you have to do the actual work to find out the results.
Typical mission times for Near Earth Asteroids in good orbits is 2-3 years. That's going out, grabbing dirt off the surface of an asteroid, and coming back. You can use Lunar gravity assist in both directions, which reduces the acceleration time on the electric propulsion. Current ion thrusters are too small for mining tugs. What you want are 200 kW plasma thrusters, like the VASIMR, and gang up 5 of them for 1 MW total power. That gives you 28.5 N @ 1 AU, sufficient to accelerate a loaded tug (1000 tons payload, 35 tons tug) at 2.38 m/s/day or 868 m/s/year. Most of the time is consumed on the return trip, since the tug is vastly heavier then. On the outbound leg the tug can achieve 68 m/s/day, and do all the required delta-V in a month or so. You would choose asteroids and orbit positions so as to minimize the return leg, and just accept a less efficient outbound leg.
If you think 1 MW is a lot of solar power, modern solar panels the size of the ones on the Space Station (400 square meters) can produce 165 kW each, so six of them in a hexagon around the tug core can do it. By comparison the Space Station has eight main panels.
The main asteroid belt is 1.1 to 2.3 AU from Earth ( https://upload.wikimedia.org/w... ). The "Near Earth" group are by definition within 0.3 AU, so that's the ones you start with. They don't have water as ice, it's too hot that near the Sun. What they have is hydrated minerals, which release the water when heated to 200-300C. An example of a hydrated mineral is kaolinite, which has the formula Al2Si2O5(OH)4. It is a major component of clay on Earth. The OH's are what get driven off in the form of water vapor.
Hauling ice from beyond the "frost line" (2.8 AU), where average temperatures are low enough for ice to be stable, is certainly a possibility, but not for the early years of space mining. It's just too far away. There's lots and lots of water beyond the frost line, because Oxygen is the 3rd most common element, after Hydrogen and Helium, and water is H2O.
You can reduce metals directly in a vacuum. The point of the reduction reaction in smelting is to remove the oxygen from the mineral oxides. Carbon can do it because the oxygen has a greater affinity for it than for the metal. But simple heating in a vacuum can break down molecules, and the oxygen pumped away. It's not as easy on Earth, but in space we usually have an abundance of vacuum to play with.
Slashdot Mirror
> Putting something radioactive on the launch pad and having it detonate in the atmosphere would be terrible too (Which is why we don't send nuclear materials into the sun.)
That's not why we don't do it. I worked on a "Space Disposal of Nuclear Waste" study at Boeing, under contract to the DOE. The risk reduction (about two cancer deaths a year on a statistical basis) was simply not worth the extra cost (about double that of burying it underground). Also, the Sun is not the safest place to dispose of it. If your rocket fails and leaves it crossing the orbit of Venus or Mercury, they could send it back to Earth by accident. The lowest risk is to place it in an orbit halfway between Venus and Earth (0.85 AU), and that also takes much less delta-V than hitting the Sun.
The nuclear waste was assumed to be glassified into coke-can sized segments, then formed into 2-meter "waste balls" surrounded by 20 cm thick steel alloy, which in turn was surrounded by heat shield tiles. The worst case accident is no on the launch pad, which is merely a lot of fire. The worst case is the rocket failing just before reaching orbit, where the payload kinetic energy is twice that of the best rocket fuel. So the heat shield enabled surviving re-entry, and the thick steel shell enabled surviving a terminal-velocity ground impact. It was also a corrosion-resistant alloy, because most launch failures end up dropping the payload in the ocean. We assumed a 2% launch failure rate.
The waste balls were so damage-resistant, that the study manager would have been happy to take one home and put it in the basement to keep the house warm in the winter (they generate 2 kW from radioactive decay heat). House fires, natural gas explosions, earthquakes, none of those would do any damage to it.
> the basic food for plants without which plants would not grow and we would all starve to death is not a bad thing.
Too much food, or anything for that matter, is a bad thing. We humans need water, but too much of it is called drowning. In the case of Carbon Dioxide, too much is a poison for people. Occupational guidelines are not to exceed 0.5% for extended periods.
There is no competition. Cotton farming in Alabama, for example, where I used to live, is highly mechanized. They spray the fields by airplane with herbicides to kill the leaves. Then the harvesting machine chews up the remainder of the plants. Stems are denser than the light and fluffy cotton bolls, so they get separated by air, and the stems are discarded. When the harvester is full, it dumps the cotton into a baler, making bales larger than a tractor-trailer in size (cotton is light). The bales eventually go to the mill to be processed. Around three people can harvest hundreds of acres at at time.
http://www.walkinginhighcotton...
I used it as a long-term investment. I accumulated them from 2011 to earlier this year at an average price of $76/btc. Now that it has gone over my price target of $600/btc I'm selling. Unless the current block size logjam is broken, I don't anticipate it will go up that much more in the near term, so I will keep selling until it's gone.
The block size is currently limited to 1 MB/10 minutes. Blocks are sets of transactions which are secured by a series of special hashes that are chained together (hence the name blockchain for the whole transaction record). 1 MB can fit a few thousand transactions, and ~250,000 transactions per day simply can't serve day-to-day uses like coffee at Starbucks. It's fine for my purpose, or things like international payments and remittances, where bitcoin is far cheaper than bank wires or services like Western Union. Unless the limit is changed, it will have to be "second layer" payment services to handle small daily transactions, and bitcoin itself used for settlement between such services. This is similar to how banks use clearing-houses to settle up payments between themselves, on behalf of their individual customers. In other words, your bank doesn't send money to Walmart each time an individual shops there. Instead, they batch together all the payments for the day from your bank to Walmart's bank, and in turn that bank forwards the right amount to Walmart's account. Since Walmart's bank is doing the same thing to pay employees and suppliers, the daily clearing is a single payment between the two banks for the net amount going in both directions.
Bitcoin only reached filling the 1 MB of transactions per block on a regular basis earlier this year. People are working on second layer solutions, but they are mostly not ready yet.
No. The high gain antenna (dish) is not pointable now that it's attached to the comet, and may have been damaged on landing. Ditto for the solar arrays.
If modding games might turn kids into hackers, imagine what *writing* games and apps can do.
No, the liquid Oxygen is delivered as liquid on trucks, and stored in large tanks at the pad as liquid. The Helium would be cooled to LOX temperature by virtue of being inside a tank full of the stuff. This would lower the pressure of the stored helium, allowing you to put more in the tank, but it's not cold enough to liquefy.
One possible failure mode is something preventing the Helium from cooling down, in which case it could overpressure the tank and it blows up. That could be a problem with getting the LOX into the tank, bubbles around the He tank, etc. Or it could be something simple like a flow valve fracturing. They have all the telemetry data, so I can only speculate.
Nah, that's too obvious. A sniper rifle fired by a CCAFS security person with money troubles is less conspicuous. Security staff have a reason to be on the base, even patrolling the launch pad area. On the other hand, there's a whole lot of nobody else around the launch pads, for safety reasons. So all he has to do is find a good spot, pop off a shot, then drive over to the launch pad like a concerned security guy would do when something goes boom.
Why money troubles? The people with a motive, like United Launch Alliance, could pay off someone for a whole lot less than what they stand to lose by SpaceX eating their business. Even a six month delay and a few customers moving payloads to "spread their risks" is worth a billion or so in revenue.
> although they feel the timeline is too ambitious.
Musk operates on the Martian calendar, so everything takes 1.88 times longer when you convert to Earth years. Once you correct for that, all his timelines work out pretty well.
Solar City is the largest US installer of solar panels.
Luxembourg is a forward-looking country. They invested in communications satellites in the 1980's, and now operate the largest commercial constellation of satellites. Recently, they started investing in asteroid mining, and they are also a SpaceX customer. I don't think Musk is so dumb he didn't know a big rocket could go other places than Mars. I think what's happened is he has a customer who is *interested* in going other places than Mars. And he needs lots of commercial customers to help pay for the big rocket he wants to build.
Look, he's developing the Raptor engine ( https://en.wikipedia.org/wiki/... ) Assuming he uses 9 of them in the first stage, like the Falcon 9 has, that's 20.7 MN of liftoff thrust. Liquid rocket T/W on liftoff is typicall 1.3:1, which gives 2100 tons liftoff mass. A good chemical rocket typically has 4% payload mass, so 84 tons payload to LEO. All of that follows directly from the engine size and how many you use.
If you can put that much mass into Low Earth Orbit, you can get variable amounts of payload to different higher orbits. This is obvious to anyone who has much experience with rocketry. User handbooks for different launch vehicles have graphs showing the payload as a function of mission velocity, and that velocity is set by where you are going and the trajectory you follow. Perhaps Musk is slow to realize this, because of his focus on colonizing Mars, but it's no surprise to people in the industry like me, and probably to a lot of the people working at SpaceX either. I can imagine the staff meeting at SpaceX:
Musk: You mean this giant rocket we're building can go other places than Mars?
Staff in unison: No shit, Sherlock.
The one in the movie 2001 was buried. They found it because it generated an anomalous magnetic field, hence the name "Tycho Magnetic Anomaly One", or TMA-1
In this case, it was Powell's g-mail account. His login credentials may have been reused on some other site that had their data stolen. In that case it would not be a hack so much as a stolen password.
Sorry, but this is wrong-headed. We have to start working on it *now*, so that in 50 years we have the experience to build the space factories. It was 60 years from the Wright Flyer to the 747, but you can't skip all the steps in between.
> we don't even have a Wikipedia page on it yet?
We have a WikiBook half written about it: https://en.wikibooks.org/wiki/...
There's a Wikipedia page on self replicating machines: https://en.wikipedia.org/wiki/...
But "fully automated self-replication" is both a limiting concept, and *hard*. There is no reason you can't make different machines than the ones you start with, or different sizes. So a "starter set" can be smaller and simpler than the final factory. All the complexity is in the stored computer files that tell it what to build. There is also no reason that it has to be 100% automated and make 100% of its own parts. Those are theoretical ideals like 100% efficiency. We can tolerate some manual labor and buying parts and materials from outside. The only real requirements are to be efficient enough to compete with conventional manufacturing, and have enough surplus production to pay for the things you can't make on your own.
We currently have $95,000 of income producing assets and a 3 acre R&D location being developed. We are an *open source project*, not a venture capital startup, so we don't have paid staff, at least not yet. People contribute their time and funds to the project, and we do the best we can with it. Our workshop won't house all the machines and tools we need to build our prototypes. For that we rely on a network of makerspaces, individually owned equipment, university labs, etc.
We don't have a timeline for reaching a complete prototype factory, we haven't even finished conceptual design. For the moment the work is fleshing out the concepts, and renovating the workshop space on our property, so when we do have hardware to test, we will be able to. Things like solar furnaces need sunlight, which is why we have 3 acres. You can't test them indoors.
I'm a retired aerospace engineer. I devote my spare time to this project because it interests me. Other people's motivations are their own, I can't speak for them.
> Until such a thing exists this is just fiction. As far as I can tell, at the moment all plans for building a self-bootstrapping automated factory on Earth, much less the actual operating factory, are fiction also.
Industrial automation is a thing, and has been for decades. We don't have to reinvent that part. What makes a seed factory different is the CAD/CAM files include making parts for more machines, besides the salable end products that any factory makes. Again, most of this has already been done, machine tool makers use their own machines that they made to produce more machine tools to sell to others. Robot makers use their own robots in their robot factories to make more robots. So our design problem is coming up with a growth path from a small and simple starter set (we presently have 8 elements in the starter set) to a complete factory capable of producing new starter sets.
And because people like you justifiably question the ability to do this, our project goal is to design and build prototypes to prove it can be done. We are not done yet, far from it. But you have to start somewhere, even if it takes 10 or 20 years to reach the goal. You will note that unlike Mars One (who I criticized myself, here on Slashdot) we don't give some fictional completion date. All we say is "here's the idea, here's what we've done so far, we are working on it".
Also space is full of high levels of radiation, and it will scare them off. We only exist here on the Earth's surface because of the ozone layer, depth of our atmosphere, and magnetic field.
Any aerospace engineer, but apparently no member of Congress, knows the right approach to get most of the cost reduction:
* Stop throwing away several kg of aerospace hardware @ ~$1000/kg every time you launch. *
There's roughly 160 MJ/kg of fuel energy in a good conventional rocket, which results in 1 kg of payload with 31 MJ of orbital energy, so around 20% energy efficiency at best, and often 10% in not so good designs. But propellant is *cheap*, around $1/kg. By far most of the cost is the hardware.
Outside of NASA, most of the new rocket development involves re-using the hardware. The Space Shuttle was intended to save money by reusing most of the hardware, but the program was poorly designed and run and didn't achieve cost savings. They did, however fly hardware multiple times.
A 6 km/s tip velocity skyhook is not an optimum design at present. When you do the actual numbers, it comes out closer to 3 km/s, and the remainder is supplied by a ~4 km/s single stage rocket. The totals are not the same because the faster skyhook is larger, and has a higher center of mass. In turn, that means slower orbit velocity. Also, 3 km/s is sufficient to reach high orbits from low orbit, and that's all you really need. Finally, skyhook mass is highly non-linear in tip velocity. Assuming carbon fiber with conservative safety margins, you can allow 150 g-km stress.
The 3 km/s skyhook has 460 km g-km of stress, and mass ratio of one side (center to tip) is exponential in stress ratio, so e^(460/150) = 21.5. Since it has two arms, the total mass ratio is 43:1. The 6 km/s skyhook has 4 times the stress - 1835 g-km. So the total mass ratio becomes 205,500:1, which is unworkably high.
A 4 km/s single stage rocket is *easy* compared to reaching all the way to orbit all by itself, and you can dramatically improve the design margins so it can make many flights. The rocket then bears the same relationship to the elevator as an airplane does to an airport.
I've spent 39 years doing space systems engineering, and know more than most about getting to orbit. I *will* talk about space elevators, having taught a class about them last year. But not space elevators "as we know it", to paraphrase Spock. The ground-to-60000 km single cable version that most illustrations show is unworkable, even with carbon nanotubes. A feasible version uses two rotating cables, one in low orbit, and the other in high orbit, with nothing between them but orbit mechanics. Their combined length is under 2000 km, and can be built with today's carbon fiber. You still need a way to get from the ground to half low-orbit energy, but even a chemical rocket can do that, easily, with good design margins, and a single stage.
But even that type of elevator isn't justified yet by traffic rates. A space elevator is "transportation infrastructure", like a bridge or airport. You don't build those for a few trips a month, and neither do you build a space elevator for a few trips a month. But a few trips a month is all the space traffic you can get today. So before anyone thinks seriously about elevators, we have to work through more conventional ways to reduce cost and grow the market.
Automated factories already exist, and have for many years. What's different about a seed factory is having a planned growth sequence from stored CAD/CAM type design files. They also make useful end-products from the same type of design files, like current automated production follows.
> What cheaper method of getting out of Earth's gravity well can be implemented now?
Enough people are working on that problem (as I used to do at Boeing's space systems division) that a few more people won't make a material difference. Changing how much stuff you need to launch, by bootstrapping production from a starter set, has a lot of leverage, and not many people are working on it. That's why I choose to spend my time on it. No, you can't order a seed factory kit from Amazon at the present time. Neither can you order an iPhone 8 or Galaxy Note 8 yet. They have to be designed and prototyped first, which is what I work on.
> How much cheaper is it than how it's done now?
The potential is to get 98% of the mass of future space products from space. The other 2% is hard to make items or rare elements for which delivering from Earth is easier. The savings are therefore 50x in reduced launch cost, minus the cost and launch mass of the starter equipment. Nobody knows the final cost, because the whole concept is too new. The *potential* is large enough to justify working on it, just like the *potential* of integrated circuits was in 1960. But you have to do the actual work to find out the results.
Typical mission times for Near Earth Asteroids in good orbits is 2-3 years. That's going out, grabbing dirt off the surface of an asteroid, and coming back. You can use Lunar gravity assist in both directions, which reduces the acceleration time on the electric propulsion. Current ion thrusters are too small for mining tugs. What you want are 200 kW plasma thrusters, like the VASIMR, and gang up 5 of them for 1 MW total power. That gives you 28.5 N @ 1 AU, sufficient to accelerate a loaded tug (1000 tons payload, 35 tons tug) at 2.38 m/s/day or 868 m/s/year. Most of the time is consumed on the return trip, since the tug is vastly heavier then. On the outbound leg the tug can achieve 68 m/s/day, and do all the required delta-V in a month or so. You would choose asteroids and orbit positions so as to minimize the return leg, and just accept a less efficient outbound leg.
If you think 1 MW is a lot of solar power, modern solar panels the size of the ones on the Space Station (400 square meters) can produce 165 kW each, so six of them in a hexagon around the tug core can do it. By comparison the Space Station has eight main panels.
The main asteroid belt is 1.1 to 2.3 AU from Earth ( https://upload.wikimedia.org/w... ). The "Near Earth" group are by definition within 0.3 AU, so that's the ones you start with. They don't have water as ice, it's too hot that near the Sun. What they have is hydrated minerals, which release the water when heated to 200-300C. An example of a hydrated mineral is kaolinite, which has the formula Al2Si2O5(OH)4. It is a major component of clay on Earth. The OH's are what get driven off in the form of water vapor.
Hauling ice from beyond the "frost line" (2.8 AU), where average temperatures are low enough for ice to be stable, is certainly a possibility, but not for the early years of space mining. It's just too far away. There's lots and lots of water beyond the frost line, because Oxygen is the 3rd most common element, after Hydrogen and Helium, and water is H2O.
You can reduce metals directly in a vacuum. The point of the reduction reaction in smelting is to remove the oxygen from the mineral oxides. Carbon can do it because the oxygen has a greater affinity for it than for the metal. But simple heating in a vacuum can break down molecules, and the oxygen pumped away. It's not as easy on Earth, but in space we usually have an abundance of vacuum to play with.