Japan to build Space elevator

[Surlethe]

(b) No, because the beanstalk is rotating with the Earth, and so isn't cutting across the Earth's magnetic field lines.

If the Earth's magnetic field is not perfectly stationary with respect to its surface, the immense length of the beanstalk will amplify any time-dependent variations in the magnetic field. You're right that the beanstalk will not cut across the large part of the Earth's magnetic field, but current flowing in the elevator is something to take into consideration.

[starslayer]

That's why I was suggesting stacking the string.

Or more accurately -

Build Elevator Shaft #1 as tall and wide and long as you can build it with commerically avialable materials.

On the floor it ends, start Shaft #2 across from it. Build it the same until it's end, then start Shaft #3

Sure, you'd have to keep moving things from one elevator car to another every few thousand stories, but that's GOT to be cheaper, and more environmentally friendly, then space launches currently are.

This is impossible. You're still building a compressive structure, and these cannot reach nearly high enough for our purposes.

If I understand rightly, a beanstalk does not stand on the ground and reach to orbit.

It hangs from orbit, and reaches the ground.

Correct. Building a compression structure this large is simply impossible. It would destroy itself long before we reached LEO, much less GEO and beyond. The cable must be deployed from the top down.

I'd love to hear how they go about securing the bottom end to the ground. The atmospheric interference alone would be a nightmare.

It really isn't that hard. Due to several other considerations, such as storms and orbital debris, a movable platform such as an oil rig is necessary anyways. Even with a wind-blown cable, a mobile platform could move with it and successfully maintain anchor.

You're right that the beanstalk will not cut across the large part of the Earth's magnetic field, but current flowing in the elevator is something to take into consideration.

Not really. The thing is, the cable would be made of carbon nanotubes, which IIRC have very high resistance. Given this, and the fact that the voltage generated really isn't all that much, there isn't much current flowing, even though the elevator cuts across a very large portion of the field (this thing is something like 100,000 km long, and it sways unavoidably due to deployment), all things considered. Edwards calculated that the worst-case scenario for cable heating resulted in .0064 Watts per meter, and this is in the strongest regions of the field.

[dragon]

I'd love to hear how they go about securing the bottom end to the ground. The atmospheric interference alone would be a nightmare.

Several of the purposals suggest using a floating platform.

The primary system is a ribbon attached at one end to Earth on a floating platform located in the equatorial Pacific Ocean

link

[Koolaidkirby]

uhh... one question about this that I'm sure anyone here could answer easily (I'm not the greatest at physics), now you've said that it basically hangs a "cable" from space (which I'm guessing would be some sort of space station/solar array anchor at the top), and I could see how that would work if its in synchronous orbit with whatever its floating above, but once you were to use, say, elevators on the "cable" (which pulls down on the cable to pull itself up if I remember this right) what would keep the "top" of the elevator from being pulled down? would it just be centripetal force of whatever mass is at the top of the elevator? or would there need to be something else?

[Crayz9000]

Compared to the mass of the space elevator, even a very heavy car is insignificant. It's kind of like Spider-Man crawling up a skyscraper: he doesn't drag it down with him.

[Shroom Man 777]

So the thing on top of the Space Elevator would have to be... what, some sort of great super space station with cables and winches to drag the heavy car up? If it's able to keep itself in orbit and not get dragged down, it must have serious stuff like... boosters to keep itself from losing orbit or something, yes? The station in orbit would have to be a never before seen feat of engineering - something totally massive.

[Ford Prefect]

So the thing on top of the Space Elevator would have to be... what, some sort of great super space station with cables and winches to drag the heavy car up? If it's able to keep itself in orbit and not get dragged down, it must have serious stuff like... boosters to keep itself from losing orbit or something, yes? The station in orbit would have to be a never before seen feat of engineering - something totally massive.

The counterweight is probably going to be a spacestation, it might as well. However, it will not have the apparatus required to lift the lifts into orbit - the carriages climb the cable. Also, it will not need boosters (or indeed, to be especially massive, but I am not an expert), to stay in geostationary orbit. Think of the analogy of spinning a weight around on a length of string, which has been mentioned already.

[Steel]

So the thing on top of the Space Elevator would have to be... what, some sort of great super space station with cables and winches to drag the heavy car up? If it's able to keep itself in orbit and not get dragged down, it must have serious stuff like... boosters to keep itself from losing orbit or something, yes? The station in orbit would have to be a never before seen feat of engineering - something totally massive.

The counterweight is probably going to be a spacestation, it might as well. However, it will not have the apparatus required to lift the lifts into orbit - the carriages climb the cable. Also, it will not need boosters (or indeed, to be especially massive, but I am not an expert), to stay in geostationary orbit. Think of the analogy of spinning a weight around on a length of string, which has been mentioned already.

Also the point of a space elevator is that you could have a continuous series of cars heading up to the top and coming down at the same time. This means that if you pack whatever waste or mining your station produces into the cars going down then the mass going up can always match the mass going down and there will be no difference in the orbit of the station, and no power needed to lift the cars.

[Molyneux]

I have to admit that I'm very skeptical about their capabilities - didn't Japan claim about a year ago that they wanted to have a fully functional moonbase by 2025, and then quietly scrap it? - but if they really want to put the effort towards getting a skyhook in place, then hell, more power to them! It's about damned time.

[Sikon]

The article title could more accurately read: "Private Japanese Space Elevator Association imagines a space elevator project."

Japan isn't building a space elevator any more than the U.S. is building one. Some individuals in Japan who are part of private organization with little if any funding dream of such.

“Just like travelling abroad, anyone will be able to ride the elevator into space,” Shuichi Ono, chairman of the Japan Space Elevator Association, said. [...]

According to Yoshio Aoki, a professor of precision machinery engineering at Nihon University and a director of the Japan Space Elevator Association, the cable would need to be about four times stronger than what is currently the strongest carbon nanotube fibre, or about 180 times stronger than steel.

That is the only group really in the article.

There are informal private space elevator advocacy groups in other countries too. I've seen such for a number of years on the internet.

Now, if the article instead said JAXA was spending so-and-so billion Yen government funding on starting such a project in the 2009 fiscal year, such could be described as Japan doing it, but that's not going to happen.

-----------

Let's consider a space elevator as in the popular sci-fi imagination, a structure reaching up to the distance at which its end station can be stationary with respect to a portion of earth's surface, allowing cargo to travel up a relatively stationary cable. That corresponds to an astronomically long cable from a station 36000 kilometers above earth to its surface, as GEO altitude is 36000 km. (The total structure would be still bigger since a counterweight extending into orbit above GEO would be needed with so much mass being suspended from the GEO station).

I think a lot of people imagine a space elevator like a highway, thinking such would transfer large amounts of goods affordably to space. There's a lot of differences actually. Let's progress towards a more accurate analogy.

Imagine that thousands-of-miles-long highway was vertical, made from exotic lab material costing more than gold. Under the most optimistic assumptions allowing it to work at all, it could just barely support its own weight without breaking under the stress, while its payload carried had to be orders of magnitude less than its total mass.

So don't imagine something like a multi-billion-dollar highway with thousands to millions of vehicles driving on it at once after construction, amortizing the capital cost over a lot of transit mass per unit time. Imagine an expensive highway upon which only *one* vehicle was driving down the whole thousands of miles of road at a time. Or it could be more than one car if multiple miniature cars, but the mass would have to remain very tiny compared to the mass of the highway to keep the vertical structure from breaking under its weight.

That is a key weakness of a hypothetical space elevator. If, for example, a climber could climb up the cable at X m/s, it takes 417 days divided by X for the climber to reach the top 36 million meters above earth's surface.

How fast can it climb? It can't just climb a thin super-stressed tape at more than a limited velocity, not without excessive abrasion among other practical issues. (Anything 36 million meters in length lifted into space with current launch technology affordably has to be rather thin, to say the least, to not be too massive).

Let's consider a velocity likely on the order of somewhere between 10 and 100 m/s depending on assumptions. Then one is talking about between several days and weeks for the climber to reach the top. There can't be multiple climbers going up at once unless their mass each is even tinier compared to the mass of the exponentially tapered 36 million meter cable.

So the giant super-expensive space elevator amortizes its capital cost over a very limited number and mass of payloads per month and per year.

People sometimes talk about several dollars per kilogram or less for the cost of a space elevator delivering payload to space for the *electrical energy* cost which is totally different from the capital cost.

In contrast, a mass driver is electrically powered too, 40% to 90% efficiency likely comparable or better than that of power transmission to an elevator climber. However, the huge difference is that a mass driver can fire up to multiple projectiles per minute, thousands of projectiles per day, and up to millions over its total operation. Literally orders of magnitude more mass can be launched per unit time, as it takes seconds or less for the cargo to transverse the mass driver launcher, instead of spending days or weeks to climb an astronomically long capable. That leads to *vastly* lesser amortized capital cost, far more launched.

Meanwhile, instead of trying to setup in space a 36000-km construction made from lab materials costing more than their weight in gold, a mass driver requires only a few billion dollars of investment in conventional materials: electromagnets, steel structure such as within a concrete tunnel in a mountain, and so on. None of the stationary hardware is subject to extreme temperatures or to trying to withstand many-GPa stress, which helps make it a good engineering solution. Unlike a space elevator, it doesn't have to be lifted into space itself or to "fly" so to speak, which helps keep it comparably less expensive per unit mass.

Personally, I suspect space elevators originally became a popular idea in sci-fi because the idea of a civilization making structures from ground to geosynchronous orbit was so outrageously difficult as to be a super-impressive feat demonstrating their power, something cool and awesome, then the general public and other sci-fi authors confused sci-fi popularity with realism.

It's ironic that in a thread some time ago many people expressed an unwarranted degree of doubt about the practicality of artificial gravity by having structures tens to hundreds of meter in diameter rotating to provide between fractional-g and 1-g pseudogravity. That corresponds to a moderate number of MPa stress, easily within the limits of materials. But here one is talking about an imaginary structure withstanding between 1g and a fraction of g force from earth's gravity over a length of a 36000000 meters, a situation where the structural stress and strength requirements are orders of magnitude different and utterly incomparable.

**************************

One should understand what assumptions are involved in imagining such a space elevator to be able to work at all without the 36 million meter cable breaking under the stress of gravity over its length. Size matters.

In the real world, we don't tend to build structures more than 1 or 2 kilometers high in part because the structural stresses on the base increase, and it becomes less practical, although tapered towers tens of kilometers or more high are theoretically possible if aerospace materials were used albeit expensive. It does help for the space elevator to have not a structure under compression and subject to buckling but rather a cable under tension. However, there still are limits to the strength of materials, though.

If one took an ordinary untapered rope and made it as long as one could dangling in earth's gravity before it broke under its own weight, it couldn't be more than a moderate number of kilometers long. Now, if one took the exceptionally high strength material of Kelvar 49 with 3.6 GPa tensile strength and 1.44 g/cm^3 density, one could nominally have a cable up to 255 kilometers long before it broke under its own weight.

As a random example, a 1cm diameter cable of Kelvar 49 of 250 kilometers in length would weigh 113 kilograms per kilometer of length, so it would weigh ~ 28.3 metric tons. Distributed over the 1cm diameter or ~ 0.785 cm^2 cross-section at the top, the resulting stress would be 3500 MPa, nominally just barely within the strength limit of the material. Really, that would be unsafe without a safety factor, and the cable should more appropriately not be designed to be more than a fraction as long. But it works for the illustration. That 3600 MPa tensile strength of Kelvar 49 is actually pretty good; for some perspective, steel cables are a fraction as much strength to weight ratio.

If one took the highest performance commercial fiber in existence maintaining strength over macroscopic lengths, some carbon fiber reaches up to ~ 7 GPa tensile strength with 1.85 g/cm^3 strength. Technically, nominally, one could theoretically make a non-tapered cable of such able to withstand its own weight over up to 390 kilometers length in earth's gravity, neglecting some practical issues against the actual feasibility of such.

One can exceed the limit for a non-tapered cable by having a tapered cable. But that is not without tradeoffs. The more one tapers the cable, the longer its total length can become without breaking, but its required mass goes up exponentially.

For example, tapering the preceding imaginary carbon-fiber cable enough to allow 2000 km instead of 390 km length while supporting a payload on the end can technically be done. However, then the cable must mass more than an order of magnitude beyond the mass of its final hundreds-of-km length and beyond the mass of its payload with all the tapering. Try for 10000 km instead, and there would be orders of magnitude involved. Etc.

With existing engineering materials being exceptionally impractical for such a space elevator, some have hypothesized about what if levels of strength seen on the micro scale in the lab could be translated to the macro scale.

Some materials are really strong on the micro scale in a lab environment. For some random illustrations, the theoretical tensile strength of diamond is [url=http://www.lrsm.upenn.edu%20~vitek/papers/friak:03b.pdf]around[/url] 100 to 200 GPa, that of aluminum is 11-12 GPa, and even iron is 13+ GPa. For example, microscopic single-crystal iron has been measured as having 10+ GPa strength on some samples, while, in contrast, iron and steel produced on the macro scale ranges from a fraction of 1 GPa for most types up to around 2 GPa at best. There's a lot of complications and details beyond the scope of this discussion like brittleness and practicality.

But for decades there have been some materials in the lab able to made with super high strength on the micro scale. Carbon nanotubes have record strength for such. To use an example given earlier in this thread:

The tensile strengths of individual multiwalled carbon nanotubes (MWCNTs) were measured with a "nanostressing stage" located within a scanning electron microscope. The tensile-loading experiment was prepared and observed entirely within the microscope and was recorded on video. The MWCNTs broke in the outermost layer ("sword-in-sheath" failure), and the tensile strength of this layer ranged from 11 to 63 gigapascals for the set of 19 MWCNTs that were loaded. Analysis of the stress-strain curves for individual MWCNTs indicated that the Young's modulus E of the outermost layer varied from 270 to 950 gigapascals.

From here

Since I'm not sure all readers would regularly know this, the Young's Modulus is totally different from the tensile strength. To give an illustration, common structural steel can have a Young's Modulus of 200 GPa but a tensile strength of 0.4 GPa. The Young's Modulus is just a measure of stiffness.

The more relevant figure in the article is the outer layer of microscopic individual nanotubes ranged from 11 to 63 GPa tensile strength as measured with the electron microscope.

I can barely express what a giant leap it would be to assume from that the creation of a macroscopic cable with strength in that range.

(In fact, it is so difficult that, for joining microscopic nanotubes into a long cable while avoiding defects much reducing strength, often space elevator cables have been considered in the context of first developing molecular nanotechnology with self-replicating nanobots, a countless trillion-dollar-significance technology for other applications, way out of current development range).

However, neglecting that and neglecting a safety factor, nominally such strength relative to 1.3 to 1.4 g/cm^3 carbon nanotube density would theoretically allow an untapered cable of around 830 to 4800 kilometers length before breaking under its own weight in a 1g gravity field. Adjusting for how the cable can instead be tapered and how earth's gravity diminishes gradually means that such technically can be sufficient strength for a space elevator with 36000 km length, although its mass after all the tapering will be huge compared to the permissible payload.

More precisely than a single circular cable, it could be imagined as tapes to reduce its vulnerability to a space junk impact, for it presents a pretty significant target over thousands of km of length, and far smaller satellites have been hit by such. A safety factor, some protection against atomic oxygen erosion, and much more would contribute further to reducing actual obtainable tension to mass ratio obtainable.

Hopefully by now one can have an idea of how optimistic and questionable assumptions are involved in imagining such a space elevator being able to be built at all. Meanwhile, as previously discussed, it obtains a vastly lesser launch rate than a mass driver is capable of reaching. A mass driver gets orders of magnitude better potential without any advancement in material technology needed.

**************************

There is a common source of confusion about space elevators. In total contrast to the sci-fi idea of a space elevator stationary with respect to earth's surface and 36000+ km in length, sometimes a more realistic idea of a space elevator in orbit of only a few hundred kilometers length has been considered. Because of the way size matters exponentially, such a different concept is actually relatively practical. It could be made with existing materials, mass only a comparably moderate number of times more than its payload, and its mass could be amortized over a much greater number of payloads since they wouldn't have to take nearly as long to transverse perhaps a couple orders of magnitude less cable length.

That's what has been most seriously considered, like some R&D investigation with a little actual funding by NASA, although it's still not going to be developed in the immediate foreseeable future. Unfortunately such tends to be described with the same term of space elevator even though it is so drastically different from the popular perception.

Such would only be one part of an overall launch system, providing a benefit like a fraction of a km/s reduction in the 8 km/s needed for a rocket to reach orbit. With a few hundred km length, its bottom end travels in a low orbit, at the speed of the bulk of its mass which is a station in a lesser-velocity higher orbit above.

One would still be using rockets to reach orbit, but the rockets might have significantly higher payload fractions as there is more than linear gain, where each 1% reduction of delta v required can results in multiple percent increase in rocket payload, as suggested by the exponential nature of the rocket equation.

-----------

However, by far the lowest cost per unit mass sent to space can come from mass drivers.

What I personally dislike about the popularity of 36000-km space elevators is that this impractical concept gets popular attention instead of more realistic methods for space launch cost reduction. It helps the general public think such requires future technology and hypothetical advancements in materials technology indefinite decades or generations away, rather than being able to be done now. Its popularity seems to come largely from an incorrect popular perception that it is the only alternative to current expensive rockets.

[madd0ct0r]

Surely mass drivers accelerate payloads far too quickly for a passenger (or anything easily squishable) to survive?



Ok, thought about it a little more and the rate of acceleration would be dependant on the length of time you accelerate for. duh.

So by increasing the length of the mass driver 'barrel' you could have a slower acceleration.

Now, you'd want to minimse this where you can to
1) decrease the size of the structure - less land, less capital ect.
2) spend less time at high speeds (drag becomes exponential at these kinda speeds)

But you knew all that already.

The attraction of a space elevator is it should have much lower energy needs to run (while hugely more to set up)
It also looks damn cool.

[Singular Intellect]

I should point out that when I said 'securing the cable', I meant actually physically capturing it while it is being lowered into the atmosphere. Once it's actually secured, there shouldn't be too much difficulty in holding it in position.

I'm curious as to how they go about getting it to be in a relatively accurate position for acquisition in order to anchor it.

[starslayer]

Also the point of a space elevator is that you could have a continuous series of cars heading up to the top and coming down at the same time. This means that if you pack whatever waste or mining your station produces into the cars going down then the mass going up can always match the mass going down and there will be no difference in the orbit of the station, and no power needed to lift the cars.

No, it isn't. You can only have one car at a time per cable; a single elevator would only have one cable, and thus could only hold one car at a time. The point is to vastly reduce the cost of getting to orbit (and beyond), even if it takes much longer to get there.

The counterweight is probably going to be a spacestation, it might as well.

In most proposals, there is either no counterweight, or a very small one. The cable would actually extend far past GEO, so that the cable can be its own counterweight. This has the additional advantage that you can simply roll right off the end, which happens to provide easy access to many destinations in the Solar System (described in detail in the NIAC reports).

Sikon, the highest-strength steels can be as strong as 5 GPa, not that it makes any difference to steel's feasibility as a material for the beanstalk. And I have a question about mass drivers; how are you going to deal with the incredible amount of heat that will be generated in both launcher and payload, even at 90% efficiency? You're pumping all the energy the payload needs to reach orbit in seconds, and it's going to be hitting the atmosphere at that speed, thickest part first. There's going to be a tremendous amount of heating, far worse than that encountered by the Shuttle or Soyuz, especially since the payload will be going much faster on launch.

[Sikon]

Sikon, the highest-strength steels can be as strong as 5 GPa, not that it makes any difference to steel's feasibility as a material for the beanstalk. And I have a question about mass drivers; how are you going to deal with the incredible amount of heat that will be generated in both launcher and payload, even at 90% efficiency? You're pumping all the energy the payload needs to reach orbit in seconds, and it's going to be hitting the atmosphere at that speed, thickest part first. There's going to be a tremendous amount of heating, far worse than that encountered by the Shuttle or Soyuz, especially since the payload will be going much faster on launch.

Heat in the mass driver itself is distributed amongst many thousands of tons of its stationary hardware, but of course the one part that does reach very high temperature is the projectile's exterior during atmospheric passage. With an ablative heat shield like that used by spacecraft for reentry, such just needs to mass a few percent of total projectile mass. Example:

The first time this question was considered seriously in a quantitative way, to the best of my knowledge, was at the 1977 NASA Ames summer study. The theory of ablation in a dense atmosphere had received recent attention in connection with the outer planet probe program, and two members of the Ames team applied the resulting software to the problem of the Earth launcher: Chul Park and Stuart Bowen. They found, much to everybody's surprise, that an Earth-launched vehicle would not have to be prohibitively large to survive: a vehicle the size and shape of a telephone pole could be launched out of the Solar System with a loss of only about 3% of its mass, and 20% of its energy to the atmosphere.

From here.

Although not strictly required, a mountainside would tend to be used, putting the launcher exit above up to half of the atmosphere's mass.

Technically a mass driver could fire through a super-high tower with several plasma windows at the top keeping most air out for a near-vacuum inside (the closest thing in real life to a sci-fi forcefield) or even be suspended from giant high-altitude balloons with a high-strength cable trailing down from its capacitor banks to a ground power station, but those are merely optional and probably an extra expense avoided for the first mass driver. After all, "brute-force" punching through the atmosphere with a heat shield on the projectiles works as previously described. It would be noisy from the sonic booms as projectile after projectile went up like a meteor in reverse, but there are some areas without population centers nearby.

[Sikon]

Surely mass drivers accelerate payloads far too quickly for a passenger (or anything easily squishable) to survive?



Ok, thought about it a little more and the rate of acceleration would be dependant on the length of time you accelerate for. duh.

So by increasing the length of the mass driver 'barrel' you could have a slower acceleration.

Now, you'd want to minimse this where you can to
1) decrease the size of the structure - less land, less capital ect.
2) spend less time at high speeds (drag becomes exponential at these kinda speeds)

But you knew all that already.

Chances are actual colonization of space would initially begin just with a short mass-driver for cargo launch, several miles length within a mountainside tunnel, high acceleration, and high velocity.

One can't have large numbers of people emigrating to space until supporting infrastructure is set up, with that being the primary problem to deal with first. As a result, if proceeding beyond today's exploration and if trying to set up a space civilization, literally more than 99% of the mass needing to be launched initially is hardware rather than people. As a random example, a 1975 NASA proposal involved 16000 tons of nuclear power plants and mass drivers to be sent to the lunar surface to subsequently fire 10 million tons to a spacestation, yet the 300 people sent to the lunar base would have a combined body mass of around 30 tons.

Nominally that's an illustration of a ratio of 500 to 1 for needed equipment mass to passenger mass for setting up that space infrastructure, although the ratio would be a bit different with adjustment for the non-organic mass of passenger capsules and for periodic rotation of crewmembers in spartan early working environments.

A similar situation would apply with a near-earth-asteroid retrieval ship, for example, which would be made to minimize life support mass and requirements, a little analogous to an oceanic supertanker of today with a dozen crewmembers despite 200000 tons payload.

The main thing is to minimize the launch cost of the 99% or so of the mass that is cargo, while the fraction of 1% consisting of a human personnel can initially be just launched on rockets.

Current rockets are expensive with figures ranging greatly from $3000 / kg for the Proton to $57000 / kg for the Space Shuttle, in part a result of a basic production and operations cost of millions to billions of dollars being amortized over a handful of tons per year. For example, a launch of the Atlas D used about $0.04 million of kerosene and liquid oxygen, converted to today's dollars, while the rocket itself cost $27 million ... a launch cost exceeding fuel and energy expense by three orders of magnitude. Since the Atlas is not reusable, its cost is not amortized over multiple launches.

Since current rockets are sufficient for today's funded directives, a handful of astronauts being launched for exploration, nobody has put full funding into reusable rapid-turnout rocket development, but it has been repeatedly estimated at NASA and elsewhere that $400 / kg and less would be an obtainable goal. The topic is discussed more in a past thread including a number of examples here.

Launch 99% of total mass on a mass driver for a few dollars a kilogram, launch the 1% or so corresponding to people for hundreds of dollars per kilogram, and the overall average is a few dollars per kilogram.

Of course, in the long-term, once there was eventually massive space infrastructure and manufacturing with a mostly self-sufficient space civilization, then the portion of launch mass corresponding to humans could be far greater, as little additional hardware launch became needed and huge numbers of people began able to emigrate per year.

Then progression beyond rocket launch of passengers is appropriate.

If that point was reached, it would imply a lot of funding available, so a more expensive mass driver hundreds of kilometers long for passenger launch could then be built. At human-permissible acceleration, such can send larger manned projectiles to a few km/s with small inexpensive rocket engines doing a final bit of delta v and circularizing the orbit into LEO.

That could eventually make emigrating to space only a moderate number of times the expense of an intercontinental airline flight today. (Energy cost in itself isn't too much of an issue: For example, the 8 km/s of LEO is 6.4 gigajoules per 200-kg before inefficiencies as implied by KE = 0.5mv^2; for perspective, the 400 kilograms per passenger of fuel used by an airline on a long 747 flight is about 43 MJ/kg and thus 17 gigajoules).

The attraction of a space elevator is it should have much lower energy needs to run (while hugely more to set up)
It also looks damn cool.

The power required is relatively low spread out over days of climbing ... although I see that as primarily indirectly a disadvantage since having a single moderate-mass car on the cable taking days to transverse its length means less mass sent to space per year than firing thousands of projectiles a day from a coilgun running off one large nuclear power plant on the ground.

Meanwhile, naturally, except for any differences in efficiency, the energy requirement to get to a point like GEO in the end is the same however it is reached, as determined by the change in gravitational potential energy. An elevator climber may not seem like much energy consumption, but climbing up 36000000 meters adds up.

Whether a mass driver or a space elevator has more overall efficiency depends on assumptions. The space elevator has the inefficiencies from converting ground-based electricity to lasers or microwaves beamed to the climber, then conversion back to electricity to power the climber's motor. There would also tend to be energy usage for high-ISP electrically powered thrusters keeping the GEO station part of the space elevator in orbit, since each additional payload climbing up the cable slightly pulls it down with a loss of orbital momentum needing eventual compensation.

But energy considerations are quantitatively relatively small anyway compared to other aspects like capital expense versus launch capability and the comparative practicality of building in the first place. (For example, although individual microscopic carbon nanotubes can have strengths like the 11 to 63 GPa previously mentioned, whenever something is made on the macro scale like CNT / epoxy or CNT / thermoplastic composites joining together nanotubes, the overall measured strength is either a figure similar to the 5 GPa illustration here or otherwise vastly less than the theoretical ideal).

[Commander 598]

The "shoot it out full force" mass driver that we're talking about here is not the only sort of mass driver launch available, there's the one that does simply half the work of launching a craft by getting it up to a decent speed then letting the craft complete it under it's own power.

[sketerpot]

Sikon, how about circular mass drivers? Like this:



The idea is that you can save on construction costs by spinning packages around and around, accelerating them gradually over the course of a few hours. That might give them less throughput, but the demand for sending lots of material into space won't manifest until it's possible. This could get mass drivers started.

Pic taken from this article.

[Sea Skimmer]

The "shoot it out full force" mass driver that we're talking about here is not the only sort of mass driver launch available, there's the one that does simply half the work of launching a craft by getting it up to a decent speed then letting the craft complete it under it's own power.

That’s exactly what the Iraqi supergun was going to do to put a very small satellite into a very low orbit, use a rocket assisted shell. Still, you need the mass driver to shoot the projectile to a very high velocity for it to make any sense vs. simply air launching a rocket. That’s pretty much always going to rule out manned launches.

Sikon, how about circular mass drivers? Like this:

[img]http://i37.tinypic.com/2k4gi.jpg[img]

The idea is that you can save on construction costs by spinning packages around and around, accelerating them gradually over the course of a few hours. That might give them less throughput, but the demand for sending lots of material into space won't manifest until it's possible. This could get mass drivers started.

Pic taken from this article.

A circular mass driver would be an even more insane technological challenge then a straight one… this does not strike me as a very practical proposition nor one likely to save money vs. simply mass producing 500 small booster rockets at once and firing them off F-15s as needed. The claims of hundreds or thousands of launches per day are obviously pipe dreams and rather incompatible with taking hours to accelerate each projectile…

[Sikon]

Sikon, how about circular mass drivers? Like this:

[img]http://i37.tinypic.com/2k4gi.jpg[img]

The idea is that you can save on construction costs by spinning packages around and around, accelerating them gradually over the course of a few hours. That might give them less throughput, but the demand for sending lots of material into space won't manifest until it's possible. This could get mass drivers started.

Pic taken from this article.

A circular mass driver would be an even more insane technological challenge then a straight one… this does not strike me as a very practical proposition nor one likely to save money vs. simply mass producing 500 small booster rockets at once and firing them off F-15s as needed. The claims of hundreds or thousands of launches per day are obviously pipe dreams and rather incompatible with taking hours to accelerate each projectile…

There's no insane technological challenge. If you had a specific argument, it would be easier to point out the flaws. Anyway, this is not a topic on which to make no-math assumptions.

Here's one illustration of a mass driver and high launch rates, in this case up to on the order of a thousand projectiles a day:

Vehicle: Telephone Pole Shaped, Mass of 1,000 kg
Launch Velocity: 12.3 km/s
Velocity at Top of Atmosphere: 11 km/s (escape velocity)
Kinetic Energy at Launch: 76 x 109 joule
Ablation Loss, Carbon Shield: 3% of mass
Energy Loss: 20%
Acceleration: 1,000 gee
Launcher Length: 7.8 km
Launch Duration: 1.26 second
Average Force: 9.8 x 106 newton = 2.2 x 106 pound
Average Power: 60 x 106 kilowatts
Charging Time From 1,000 MW Power Plant: 1.5 minute

From here.

The rapid acceleration results in large capacitor banks needed, not an excessive problem though as such cost between a fraction of a billion dollars and some billions of dollars depending on factors including the projectile mass desired. (The preceding gave an estimate of $11 billion except it isn't too meaningful to neglect economies of scale in capacitor production and some options for reduced cost, although even the $11 billion figure could be acceptable).

The beauty of a mass driver like this is once built it can launch thousands of tons a week and up to literally millions of tons over total operation. That's more of the way to progress to space infrastructure than sending up a handful of tons per year.

[Sikon]

Minor edit: I didn't correct some of the notation when copying the quote, like 106 where the formatting was lost, and there should be 10^6 obviously.

[Kanastrous]

Anyway, this is not a topic on which to make no-math assumptions.

That's a good turn of phrase.

I'm stealing it.

[Sea Skimmer]

There's no insane technological challenge.

Your own damn article says it’s a huge challenge and that they want a two year study just to look into feasibility before they even do real work! If its’ not such a huge challenge and its such a great and cost effective idea then why didn’t they just issue a development contract? No one has built anything remotely like this which handles worthwhile payloads, let alone hundreds or thousands of them per day. Sure it works like a practical accelerator, but size matters, an atom and a fifty pound satellite that is neither solid nor equal in mass distribution are just totally different cans of worms. Meanwhile projects to build military railguns and coilguns have hit all sorts of walls and aren’t expected to enter service for another two decades. They only want 1/4th the velocity this idea demands too.

These arguments remind me of the argument used to sell the Space Shuttle. Hey look lets radically push technology and we’ll get someone that’s cheap and easy to use… and no it wont have horrendous operational problems or limitations or say, a helium leak that shuts it down for an entire year…

[Sikon]

There's no insane technological challenge.

Your own damn article says it’s a huge challenge and that they want a two year study just to look into feasibility before they even do real work!

An insane technological challenge would be one for which successful completion was unlikely even after development funds or one for which costs exceeded gain. Neither applies here. It is not crazy or foolish but worthwhile. Of course any project like this has development time starting with paper studies.

Among the huge differences between this and a space elevator, it doesn't depend on questionable hypothetical advancements in materials technology (somehow making 10-60 GPa strength microscopic nanotubes get combined strength towards the upper end of that range on the macro scale when everything like all the CNT composites made so far have defects and limitations resulting in lesser performance on the order of the 5 GPa example previously mentioned). One could put any number of billions of dollars R&D into such with no true guarantee of success in decades or ever. It might possibly require even self-replicating molecular nanotechnology way out of current development range to make a 36000000-meter cable with few enough microscopic defects, if such wasn't promptly messed up by the space environment anyway.

Yet for the mass driver, there are no new materials needed, no scientific breakthroughs. What uncertainties exist are rather limited in comparison. For example, to have capacitors storing X GJ of energy, one can't absolutely state the exact cost yet with current information, since there's not been such a big array built historically, but one can be certain that such can be built, a matter of combining enough smaller capacitors. Some approximate upper limits on estimated cost are possible.

If its’ not such a huge challenge and its such a great and cost effective idea then why didn’t they just issue a development contract?

It fits within a context of funding massive space infrastructure and development but not within current goals like sending some astronauts per year to the ISS and back, an application for which it is irrelevant.

There aren't any entities in the world spending a few billion dollars a year on space launch except for one, NASA (as, for example, the Russian program has far lesser funds). Some people at NASA have considered options for massive space colonization, as opposed to the current funded goal of smaller scale exploration, concluding such is possible, but funding is a totally different matter.

From there came the 1975 NASA study involving lunar mass drivers sending 10 million tons up for a giant spacestation, a proposal which Congress was not that interested in funding. Among other examples discussed in the thread long ago here, Sea Dragon proposal cost estimates were judged by NASA to allow order-of-magnitude reduction in launch expense per unit mass and the capability to put many thousands of tons into orbit over repeated launches, compared to the handful sent up today. But that capability was judged unneeded for current goals, with Congress accordingly having little interest in putting billions of dollars of funding into such.

These arguments remind me of the argument used to sell the Space Shuttle. Hey look lets radically push technology and we’ll get someone that’s cheap and easy to use… and no it wont have horrendous operational problems or limitations or say, a helium leak that shuts it down for an entire year…

Fundamentally, what most makes each Space Shuttle orbiter cost $25000 a kilogram of dry weight is performance relative to weight requirements. Its hardware costs more than its weight in gold ($1.7+ billion each, 68.6 metric tons). The turbopumps in its rocket engine pump a number of gallons a minute that would be comparatively cheap if one could use a massive industrial pump, but they cost orders of magnitude more primarily because they must be so lightweight. There have been large liquid-fueled rocket engines on test stands made vastly cheaper than the SSMEs, but obtaining their thrust to weight performance astronomically drives up cost. Etc.

In contrast, there is a different situation with a mass driver's capacitors, its power plant, its electromagnets, and so on. None of that has to fly, and none of it has any reason to cost multiple times its weight in gold. The concrete and steel of its tunnel won't cost that much. The power plant for a mass driver like the earlier illustration can be the same as existing nuclear power plants used for commercial electricity for cities today. The capacitor array is giant, but it doesn't have to cost a much different number of dollars per megajoule than large capacitors in other terrestrial applications. The electromagnets don't need more teslas strength than some made before in other applications. Etc.

No one has built anything remotely like this which handles worthwhile payloads, let alone hundreds or thousands of them per day.

Naturally mass drivers like those made at MIT by a small team with a shoestring budget have tended to be in the kilojoule range rather than the gigajoule range, since working with thousands of dollars rather than billions of dollars limits the energy storage, the projectile mass, the velocity, and everything else. So one sees payloads like a fraction of a kilogram sent to a fraction of a km/s if making a lab demonstration working with so comparatively limited resources.

However, for something working within understood principles, it is possible for engineers to figure out what is involved in making a larger version. For example, in regard to the previously mentioned NASA study with lunar mass drivers, which they discuss here, there were no uncertainties so great as for them to doubt at all whether such could be built.

Rather, they were able to make quantitative estimates like the power plant mass needed, the 3 tesla strength for the electromagnets in that design (not too high, obtained before in other applications), and so on. Sure, they allocated a R&D budget with a lot of funds spent over years, as natural in working out all the details of a complete engineering design, but there isn't uncertainty over whether it would work.

Meanwhile projects to build military railguns and coilguns have hit all sorts of walls and aren’t expected to enter service for another two decades. They only want 1/4th the velocity this idea demands too.

That's a rather different situation. When those have issues, it is not a matter of the basic principles not working at all but rather how performance relative to size and expense compares in applications currently already fulfilled by more conventional weapons.

Here one doesn't need a mass driver equivalent of thousands of conventional weapons like naval guns, tank cannons, and artillery. Those conventional weapons cost a fraction of a million dollars each while having competitively low mass and all the other requirements of mobile military systems.

Here one only needs to effectively build *one* multi-billion dollar "artillery piece" so to speak, massing many thousands of tons, not intended to compete with conventional military hardware. Tradeoffs like a cost literally orders of magnitude above a conventional artillery piece, the requirement to run off a massive stationary power plant sufficient to power a city, and so on become acceptable in this application when the absolute requirement is to send projectiles to sufficient velocity never obtainable by conventional cannons.

A random example is one October 1991 mass driver proposal from Sandia National Laboratories (part of the U.S. DOE), SAND91-1600:UC-706, Hypervelocity Gun Report: Electromagnetic Coilgun.

For a coilgun (mass driver) of 960 meters length, 0.72 meters diameter, angled upwards at 25 degrees built on a mountain slope, firing 1.82 metric ton projectiles at 6 km/s, their cost estimates were $0.41 billion research and development expense and $2.3 billion for the gun and launch facility, plus $0.05 billion operating cost per year, launching 4000 projectiles in the first 7 years of operation.

[Admiral Valdemar]

I have to say, I always found the idea of a mass driver, SSTO plane and Skyhook to be more practical in getting payloads up to LEO and beyond than your standard space elevator with the associated technical and political ramifications.

Sikon, have you read Charlie Stross' piece on space colonisation at all? I haven't the time to go in detail on it, but given the subject matter, I would love to hear your feedback on his arguments.

[Sikon]

Sikon, have you read Charlie Stross' piece on space colonisation at all? I haven't the time to go in detail on it, but given the subject matter, I would love to hear your feedback on his arguments.

The sci-fi author who wrote "The High Frontier, Redux" article?

His discussion is interesting to the degree it shows how some people can tend to think, but he has a number of misunderstandings and miscalculations.

He incorrectly estimates $350 / kg energy cost to GEO for a space elevator, based on a wikipedia reference link apparently assuming just 0.5% to 2% maximum energy beaming efficiency.

I'm not a fan of space elevators beyond at most a short version rather than the 36000-km type, but, nevertheless, that is just plain silly in context.

Maybe one could get that low efficiency if one really didn't care about wasting power with most of the beam sloppily missing the receiver or if one used an inefficient laser hitting solar cells not optimized for its spectra, but a multi-billion-dollar imaginary space elevator project would tend to have more sophisticated engineering. (Some proposals for sending power to the solar cells of current satellites with lasers may have low efficiency, but that's for a variety of reasons including the fact that one wouldn't even much care about the efficiency there when the satellite only needs to receive a few kilowatts).

If the astronomical difficulties in the space elevator's construction were surmounted, it would be comparatively not too hard to have vastly greater power beaming efficiency since microwave beams can be generated at high efficiency (around 50-90+% depending on the source), then converted back to electricity by a rectenna at around 80-90+% efficiency.

The actual energy cost of escaping earth's gravity with electricity such as $0.05/kw-hr is rather about $3 / kg before inefficiencies. (60 MJ/kg, $0.05 / 3.6 MJ). The inefficiencies aren't nearly quite that high.

The real problem with space elevators isn't their energy cost but rather everything else.

He talks about how it costs thousands of dollars per kilogram to send material to the moon, which naturally it does with current rockets costing literally orders of magnitude more than their fuel expense and with them having billions of dollars of development & operations expense amortized over a handful of tons launched.

He overestimates the expense of fuel & oxidizer. In contrast, to use a random example, converted to today's dollars, the Atlas used around $23000 of liquid oxygen and $13000 of kerosene to put up to a 1400 kilogram payload in LEO, ~ $26 / kg (further discussion). Again, even with the high inefficiencies of that rocket, it isn't the energy cost but other expenses which have been the primary factor, like the rocket cost $27 million versus its $0.04 million for propellant.

He talks about how it requires megatons of energy for a moderate-size interstellar craft to reach the Centauri system in a few decades. Of course it does, as quite familiar to anyone who has looked at proposals for nuclear pulse propulsion starships like myself. Such is indeed much of the reason that nuclear pulse propulsion would tend to be the most affordable method if sending large masses, as opposed to the vastly greater cost of electrically powered beams per megaton, at least in the foreseeable future short of a scenario like self-replicating factory ubertech.

While observing how the economics of current systems are unsuitable for space colonization, which is in itself true, he doesn't understand how such can be carried out with certain technologies, like mass driver launch of hardware, eventually combined electromagnetic / rocket launch of passengers, retrieval of near-earth asteroid material, etc.

Almost everything launched into orbit now costs more than its weight in gold, rather impractical for colonization, but there are ways to vastly reduce that expense while making use of extraterrestrial materials, just like the 1975 NASA study didn't imagine sending 10 million tons to orbit at the ~ $10000/kg rate of current rockets for nominally $100 trillion or more than the entire world's GDP but rather doing it for a comparatively tiny fraction of 1% of that expense with a lunar mass driver.