![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
More on space elevators
Following my earlier post, some of the comments prompted me to write a small epic on why I'm so sceptical.
Unfortunately, it wasn't small enough to post as a comment, as I discovered the hard way that LJ has a size limit on comments.
So, a follow-up post is called for.
This Nasa article quotes the required material strength for a Space Elevator cable as 62GPa.
The current state-of-the-art in high-tensile-strength steel has a strength of about 2GPa. Carbon fibre reinforced plastics are lighter than steel, but no stronger. (CFRP is popular for more subtle engineering reasons) More exotic materials like boron fibre composites might be able to do a bit better, but not much (and they cost a FORTUNE)
So, with common engineering materials, we're an order of magnitude short, even BEFORE we start hanging things like elevators or counterweights off it.
Various sources quote carbon nanotubes as having tensile strength "similar to diamond". Single-crystal Diamond has a theoretical tensile strength [210kb PDF] of about 90-225GPa (it's directional, depending on how the crystal lattice is aligned). However, as many have pointed out, diamond is too brittle to be used for general-purpose construction.
However, I've found no reports directly reporting such figures for CNTs. The Physical Properties of Carbon Nanotubes site at Michigan State University quotes 30GPa (with some margin of error), which is a bit different from the figures for diamond. Note also that the figure quoted by MSU is an actual measured value, rather than the theoretical 100GPa+ figures being lobbed about by many other sources.
To complicate things further, the "Nanotube fibres" being quoted are themselves composite materials, comprising about 60-80% CNT. There is no mention of what the other 40% is, but it's unlikely to contribute significantly to the material strength. From what I see, the nanotubes themselves are microscopic (for comparison, in Carbon fibre reinforced plastics, the actual fibres can be metres long) which means that the composite is probably going to be pretty amorphous.
A 60%, amorphous, composite is likely to have a strength well below the theoretical maximum of the CNTs. The Nasa Institute for Advanced Concepts (NIAC) report [512kb PDF] quotes 22GPa for one sample they obtained, which doesn't line up terribly well with the 62GPa target.
Once you have these individual fibres, the cable itself will then have to be a "rope" (or ribbon) of many fibres. Unless they're planning to make every one of the fibres run the entire 100,000km, this cable will be weaker than the individual CNT-composite fibres. The precise details will depend on how the fibres are joined to form the cable. Again, they don't go into much detail. The "joiner every 10cm" approach mentioned in the NIAC report doesn't sound that great to me - most engineering composites use a continuous matrix, to provide maximum adhesion between individual fibres.
Finally, these materials are still largely theoretical. There are a few newly-founded companies that will sell you raw nanotubes at $500 per gram, but CNT composite fibres still seem to be lab-manufactured stuff, not available in useful engineering quantities.
Additionally, I can see two major problems with using a meter-wide, 0.002mm thick, ribbon of cable like that:
In short, I'm a LOOOOOOOOOONG way from being convinced on the materials front.
I have other problems with the "thin cable" approach. Most of the studies done indicate that any cable needs to taper to keep the required material strength as low(!) as the 60GPa quoted above - so the midpoint (GEO) will be much thicker than the ends - maybe 20-30 times thicker. Anything climbing the cable is probably therefore going to need some sort of rail assembly to hang onto, rather than just grabbing the cable itself. Of course, as soon as you hang things on the cable, the whole contraption gets heavier, so the cable gets thicker. Also, anything hanging off the cable is going to exert considerable localised bending forces (unless it wraps right around - which is tricky if the cable is tapered by a factor of 20). Getting round this is much easier if the cable is substantial enough to resist such bending.
As soon as the cable gets thick, falling cable becomes a problem. Even if it's chopped up, there's still a lot of stuff coming down. A self-destruct may help, but that's more stuff to hang on the cable (making it heavier again) and brings up the possibility of polluting near-Earth space with thousands of bits of cable if you blow the whole thing.
For timescales: NASA's stopped laughing, as they should, because if this can be made to work, it will be a HUGE leap forward for space travel. CNT composites will happen (probably fairly soon) but they're going to be ferociously expensive, and may not be strong enough for an elevator anyway.
When (if) the materials are up to it, I can only compare to the amount of time it takes to build (say) suspension bridges, where they take months or years to assemble cables "only" 2-3km long from conventional materials. 100,000km would be a HUGE challenge. This thing would take at least a decade to build, probably longer, even once the "how" problem is solved. I'll be impressed if there's one finished in 50 years, and I agree with![[livejournal.com profile]](https://www.dreamwidth.org/img/external/lj-userinfo.gif)
zotz that 15-20 years seems fanciful.
This stuff is worth looking at, because the potential payoff is HUGE, but right now it looks like it won't be feasible for some time yet.
Unfortunately, it wasn't small enough to post as a comment, as I discovered the hard way that LJ has a size limit on comments.
So, a follow-up post is called for.
This Nasa article quotes the required material strength for a Space Elevator cable as 62GPa.
The current state-of-the-art in high-tensile-strength steel has a strength of about 2GPa. Carbon fibre reinforced plastics are lighter than steel, but no stronger. (CFRP is popular for more subtle engineering reasons) More exotic materials like boron fibre composites might be able to do a bit better, but not much (and they cost a FORTUNE)
So, with common engineering materials, we're an order of magnitude short, even BEFORE we start hanging things like elevators or counterweights off it.
Various sources quote carbon nanotubes as having tensile strength "similar to diamond". Single-crystal Diamond has a theoretical tensile strength [210kb PDF] of about 90-225GPa (it's directional, depending on how the crystal lattice is aligned). However, as many have pointed out, diamond is too brittle to be used for general-purpose construction.
However, I've found no reports directly reporting such figures for CNTs. The Physical Properties of Carbon Nanotubes site at Michigan State University quotes 30GPa (with some margin of error), which is a bit different from the figures for diamond. Note also that the figure quoted by MSU is an actual measured value, rather than the theoretical 100GPa+ figures being lobbed about by many other sources.
To complicate things further, the "Nanotube fibres" being quoted are themselves composite materials, comprising about 60-80% CNT. There is no mention of what the other 40% is, but it's unlikely to contribute significantly to the material strength. From what I see, the nanotubes themselves are microscopic (for comparison, in Carbon fibre reinforced plastics, the actual fibres can be metres long) which means that the composite is probably going to be pretty amorphous.
A 60%, amorphous, composite is likely to have a strength well below the theoretical maximum of the CNTs. The Nasa Institute for Advanced Concepts (NIAC) report [512kb PDF] quotes 22GPa for one sample they obtained, which doesn't line up terribly well with the 62GPa target.
Once you have these individual fibres, the cable itself will then have to be a "rope" (or ribbon) of many fibres. Unless they're planning to make every one of the fibres run the entire 100,000km, this cable will be weaker than the individual CNT-composite fibres. The precise details will depend on how the fibres are joined to form the cable. Again, they don't go into much detail. The "joiner every 10cm" approach mentioned in the NIAC report doesn't sound that great to me - most engineering composites use a continuous matrix, to provide maximum adhesion between individual fibres.
Finally, these materials are still largely theoretical. There are a few newly-founded companies that will sell you raw nanotubes at $500 per gram, but CNT composite fibres still seem to be lab-manufactured stuff, not available in useful engineering quantities.
Additionally, I can see two major problems with using a meter-wide, 0.002mm thick, ribbon of cable like that:
- The bottom 20km or so of the cable will have to worry about weather. That ribbon is likely to have some REALLY undesirable aerodynamic properties. The FAQ on the ISR site mentions how they will make it thicker at the bottom, but still keep it as a ribbon. They think that it'll stand anything short of a Cat5 hurricane, but I don't see any evidence that they've taken flutter into consideration - who's seen the film of the Tacoma Narrows bridge shaking itself to pieces in a modest 30-40mph wind? How well would the cable, under tension, cope with, say, an alternating combined torsional/lateral push, with a frequency in the range 0.5-20 Hz? Look at London's "wobbly" millennium bridge for another example of how easily things like this get overlooked.
- If the cable is that thin, tensile strength is not going to be the failure mode. It'll tear first. Ever tested the tensile strength of aluminium foil? Or paper? They're both pretty strong, under tension - but any actual structural failure is almost *always* due to tearing, and occurs at much lower stresses than their theoretical tensile strength. I should point out that aluminium cooking foil and paper are both considerably thicker than the proposed metre-wide, microns-thick cable.
In short, I'm a LOOOOOOOOOONG way from being convinced on the materials front.
I have other problems with the "thin cable" approach. Most of the studies done indicate that any cable needs to taper to keep the required material strength as low(!) as the 60GPa quoted above - so the midpoint (GEO) will be much thicker than the ends - maybe 20-30 times thicker. Anything climbing the cable is probably therefore going to need some sort of rail assembly to hang onto, rather than just grabbing the cable itself. Of course, as soon as you hang things on the cable, the whole contraption gets heavier, so the cable gets thicker. Also, anything hanging off the cable is going to exert considerable localised bending forces (unless it wraps right around - which is tricky if the cable is tapered by a factor of 20). Getting round this is much easier if the cable is substantial enough to resist such bending.
As soon as the cable gets thick, falling cable becomes a problem. Even if it's chopped up, there's still a lot of stuff coming down. A self-destruct may help, but that's more stuff to hang on the cable (making it heavier again) and brings up the possibility of polluting near-Earth space with thousands of bits of cable if you blow the whole thing.
For timescales: NASA's stopped laughing, as they should, because if this can be made to work, it will be a HUGE leap forward for space travel. CNT composites will happen (probably fairly soon) but they're going to be ferociously expensive, and may not be strong enough for an elevator anyway.
When (if) the materials are up to it, I can only compare to the amount of time it takes to build (say) suspension bridges, where they take months or years to assemble cables "only" 2-3km long from conventional materials. 100,000km would be a HUGE challenge. This thing would take at least a decade to build, probably longer, even once the "how" problem is solved. I'll be impressed if there's one finished in 50 years, and I agree with
zotz that 15-20 years seems fanciful.This stuff is worth looking at, because the potential payoff is HUGE, but right now it looks like it won't be feasible for some time yet.
no subject
no subject
Hey, if I wanted it kept secret, I wouldn't have posted it on the WWW.
no subject
no subject
Whoa, whoa, whoa! NASA has misled you with its talk of a minimum tensile strength. Because all that strength is required for the structure to support *itself*, lightness counts. The correct figure of merit is not tensile strength, but tensile strength divided by density.
Taking your figure of 2GPa for both materials, and taking the density of steel as 8000kg m-3, the relevant figure is 250 kilonewton metres per kilogram. For CFRP, assuming a density of 2000kg m-3, then the figure is 1 meganewton metre per kilogram, four times higher than steel.
(note to readers: a gigapascal is just another way of saying 10^9 newtons per square meter)
So where weight is a paramount consideration, CFRP is not preferred to steel for subtle engineering reasons at all, but for the very unsubtle reason that steel could not support its own weight, and CFRP can.
Another common figure of merit, used to give a rough intuitive idea of how much weight a material can bear, is to divide the previous figure by the Earth's gravity at sea level, 10 newtons per kilogram, to arrive by how much of its own length a cable could support before breaking. For steel, using your numbers above, that's 25km, and for CFRP it's 100km, again four times as much.
Even though the gravity against which an orbital cable would have to support itself is less than sea level, decreasing rapidly with height, it's clear that neither of these materials is quite up to it, but it's equally clear that CFRP is closer to it than steel.
In short, I'm a LOOOOOOOOOONG way from being convinced on the materials front.
I'm with you there.
(Materials Science and Technology, Brunel University 1995)
no subject
no subject
Density (and hence related stuff like bending/buckling resistance) is the "subtle engineering reason" (subtle, that is, compared to pure tensile strength) I was referring to, and I meant the context of general engineering. As you correctly point out, in this particular context, specific tensile strength becomes even more important.
In fact, IIRC, the 62GPa figure I'm throwing around above is for a material with a density broadly similar to CFRP, or maybe even lower, and assumes that the cable is tapered (thin at the bottom, thick in the middle (GEO), thin at the far end) to minimise loads. For something as heavy as steel, the number is really stupidbig.
With a space elevator, it all gets horribly complex, thanks to the interaction with centripetal acceleration and the counterweight cable (or mass). The gravity drop-off is, I think, only relevant in the sense that it defines how high the mid-point (where the cable is apparently "weightless") is - and even there, centripetal acceleration dominates once you get more than a few thousand km up.
For the purposes of argument, I picked the lowest possible number to illustrate how far we currently fall short, even against THAT target.
(Aeronautical Engineering, Manchester 1994)
no subject
Not to mention, if you put an explosive charge even only every 500m along the stuff below the mid point (since anything higher won't hit the ground or pollute low orbits) that's still about 70,000 charges. I wouldn't want to be the one in charge of coming up with a control mechanism that could guarantee no accidental detonations.
no subject
::wonders if she still has that e-ddress::
::wanders off to go scouring old e-mail accounts::
no subject
Thing is, the company I linked just sells the actual nanotubes. You need a pretty serious microscope to even see them, and in terms of structural material, it's just a pile of (very stiff/strong) microscopic sticks.
To do useful engineering things with them, they need to be turned into a bulk material with known mechanical properties. The most plausible first step to that is turning a bunch of tubes into a fibre, probably by forming a composite with some other material such as epoxy resin. There are some reports that a joint team from the University of Texas and Trinity College Dublin managed to do this in moderately useful quantities earlier this year, producing hundred-metre long threads.
However, threads do not make an aeroplane. The next obvious step would be to take those threads, weave them into a cloth, then use THAT as the basis for another composite material, much like current (x)-Fibre Reinforced Plastics. With the thread coming out of the labs a few hundred metres at a time, and the primary ingredient of the thread costing $500 per gram, making bulk material is likely to be slow and very expensive.
As soon as anyone does produce the stuff in sensible quantities, the aerospace industry will be all over it. Even if it's not strong enough for a space elevator (I doubt it will be) it's still likely to have mechanical properties that will have aerospace engineers foaming at the mouth. But for now, I can't see any evidence that anyone is producing engineering quantities of this stuff.
All that said, I am willing to stand corrected. Despite my scepticism, I'd love to be proved wrong here.
no subject
no subject