Ok, Space Elevators have fallen out of the limelight, sorry for them, but the concept is intriguing just the same, so here it goes.

One of the worst problem space elevators face is the fact that they are not elevators at all; they are structures, instead, that allow a "crawler" to climb them, with endless problems about how to feed energy to it, the need for motors, etc, etc.

The reason for this is that the cables need to be thicker (more resistant) near the geostationary orbit than elsewhere. A moving cable (a loop) would need to have a cross section instead.

There are, of course, many other structural problems, but I'm not trying to address them all with a single question.

Let us assume the cable itself can be made. It will have to be "tapered", which means it needs to be thicker near center of gravity (geosynchronous orbit) where everything is suspended.

This is standard S.E. design.

In this design you need a "crawler" autonomously climbing the cable and you need to give it enough power to arrive at GSO.

Doing without a crawler would mean having a true elevator, with moving cables, but a cable loop (chairlift-style) needs a cable of uniform diameter, thus it's necessary to break the distance into suitable legs.

What I propose is to:

  • subdivide the whole length into sections where the cable can reasonably be of constant section/strength.
  • make each "leg" of the trip with one loop of cable (the thinnest possible) between two wheels and have it run at constant speed; so the tract would be connected by two cables, one going up and one going down.
  • add a number of static cables (the same size as the moving one) to connect the wheel rig to the "upper station" to make up for increased cable strength needed.
  • have "crawlers" be simple cabins that will accelerate and hang on to the cable "going in the right direction".
  • add more cables for redundancy.
  • add space platforms at the wheel rigging to have nice platforms for scientific/touristic purposes (and to dampen Coriolis); to hold them in place it might be necessary to add some more static cables.

This rigging has multiple benefits:

  • you get "station platforms" (almost) for free.
  • all cables are the same section, so manufacturing is way simpler.
  • the structure can be built incrementally and enlarged at will.
  • multiple cables are less vulnerable to all kinds of incidents.
  • if (some of) the static cables are conductive you get free electricity.
  • motors to keep the moving cables running can have minimal power and rely on cable inertia for "crawler" acceleration.
  • different "legs" of the trip may be done at different speeds (e.g.: first, in atmosphere, leg should be slower than long-distance to geostationary).
  • crawler can be really simple and lightweight.

As requested I have added an image (sorry, I'm not really able to draw!): diagram of proposed elevator design

What is wrong with this scheme?
Why is this not taken into account (at least I haven't found references to it)?
Please cross-check my proposed design.

  • 1
    $\begingroup$ I've seen the update, you still have the same problem with weight and distribution. a chain experiences the same tension forces as a single long cable. You're not gaining anything by dividing it like that. Your are not reducing the tension loading, you are increasing it. $\endgroup$
    – John
    Commented Aug 22, 2017 at 18:29
  • $\begingroup$ @John: I cannot and I'm not interested in reducing the tension loading. That cannot be done and an accurately tethered cable is the theoretical best possible. Problem I'm trying to address is movement up and down, without worsening too much situation. It will be worse than theoretical, of course, but I would like to know if and why it will be dramatically worse, such not to consider it. $\endgroup$
    – ZioByte
    Commented Aug 22, 2017 at 18:38
  • 7
    $\begingroup$ All the things you done to try to FIX the problems will introduce much more difficult to solve problems, without helping the movement up and down whatsoever. It will be dramatically worse than other space elevator designs because of the tension loading (which for some reason you're not interested in) $\endgroup$
    – user41674
    Commented Aug 22, 2017 at 19:08
  • 1
    $\begingroup$ Do you want to know if you have constructed a working space elevator (see top answer for example) or do you want to know something else, e.g. if this solves one specific problem? I think so but I think the question is ambiguous and confusing and seems to ask a million things at once, but then specifically excluding what people answered is wrong with the design. You should maybe consider completely rewritting it in that case. The answers are all over the place debating all sorts of problems with space elevators and I believe it is caused by vagueness and lack of a clear aim $\endgroup$
    – Raditz_35
    Commented Aug 22, 2017 at 19:36
  • $\begingroup$ This is like asking why don't we just make highways and train tracks conveyor belts so cars and trains don't have to carry their own engines. Answer: in constrained and optimial situations (like a hyperloop) it's possible and feasible, but generally such systems are difficult to engineer, expensive to build, and these examples don't even cover the mind-boggling challenge of overcoming gravity and tension. $\endgroup$
    – Jimmy
    Commented Aug 22, 2017 at 21:55

8 Answers 8


I agree fully with all the points brought up in sdfgeoff's excellent answer, however, I want to add some points which work for your design, and which are not evaluated correctly in some of the other answers.

  1. Yes, your design could work from a statical view-point: Your intermediate platforms, pulleys, and motors will definitely add weight, how much depends on the specifics of your design. This will force you to increase the count of cables further up. But, as long as the extra weight is not too large in relation to the weight of the cable itself, this just adds to the costs of the design, it won't make the thing itself infeasible.

  2. There will be no danger of cables rubbing against each other: The rising cable will need to gain angular momentum, the descending cable will need to loose the same amount of angular momentum. Both will do so by not following a vertical path, rather the rising cable will bulge to the west, while the descending cable will bulge to the east. Thus, they will bulge away from each other, and away from the vertical static tethering.

  3. It would be very easy to provide power down from the space station to the motors by transferring ground via the static tether and the power via the rotating cable. Due to the bulging of the rotating cable, the two electrical poles will naturally be separated from each other.

    In a design that uses only static tethers, this won't be so easy: The opposing electrical charges will attract each other, so you will need to add some isolation between the cables carrying the opposing charges. With the rotating cable, space itself works as an isolation.

  4. Providing power up from the ground station would be next to impossible, unless you add an extra static cable on the last leg down to ground. That extra cable would increase the cost of the entire thing by something like 50%, because you are adding a third, otherwise useless cable. This adds extra weight, which requires 50% more cables further up to carry the load. Of course, you can try to make that extra cable thinner than the rotating cable, reducing that extra cost factor accordingly.

    However, adding that third cable would immediately allow you to directly transfer the power of a ground-based power plant up to space. It might be worth it.

  5. You have to think about how your cabins clamp on to the moving cable: You need to somehow accelerate your cabins to the cable speed without damaging the cable. Just clamping instantly would put way too much stress on the cable, using a slipping clutch would expose the cable to abrasive forces.

    A solution that could work would be to use some pulleys to connect the cabin with the cable, and to connect those pulleys to a small generator. The generator would provide electricity to the interior of the cabin during the ascent/descent by letting the cabin slowly slip downwards relative to the cable, and it could be used to make the acceleration at the ends of the trip graceful by changing how much power is extracted from it. This change of load on the generator could easily be effected by use of a small battery inside the cabin.

    The effect of this slipping for power generation would be, that the descent would be a bit faster than the ascent, but that shouldn't be any problem.

I'd say, as long as you manage to keep the extra weight of the intermediate platforms down to, say, 1% of the weight of the cables, your design would definitely work. It would be significantly more costly than building a static tether, but it would definitely solve some of the headaches connected to riser design.


The structural problems with space elevators are on a very different scale than what you are thinking of. At the lengths needed for a space elevator most materials won't be able to support their own weight. This is an engineering problem that can't be fixed, just by adding more cables. Each cable that is added would also need to support their own weight.

At this scale the mass of the cable dwarfs the mass of any climber or payload. Managing the energy and stress of moving this cable is a much harder engineering challenge than that of a "conventional" climber.

Your proposed design would require tripling the amount of cabling plus adding a bunch of wheels and motors. Furthermore all that extra weight adds no functionality to the design.

  • 9
    $\begingroup$ If a space elevator isn't geostationary it will destroy itself by crashing into mountains as the earth turns. $\endgroup$
    – sphennings
    Commented Aug 22, 2017 at 14:18
  • 7
    $\begingroup$ @ZioByte if your cable is tied to the Earth and to the geostationary station, you have the same design as every space elevator, only with more stuff (heavy, heavy stuff) in between. Kind of like if you pull a boulder on a rope, tied to a chain, tied to another rope, tied to another chain, you're still pulling the full weight of the boulder (plus the ropes/chains). If somehow you expect some of these intermediary stations to go with different rotative speed (which would solve nothing), your space elevator will wrap around the Earth very quickly. $\endgroup$
    – Keelhaul
    Commented Aug 22, 2017 at 15:02
  • 24
    $\begingroup$ @ZioByte The problem is simply in that assumption. Basically you've said "I'm assuming I have an infinitely strong, inelastic, non-fatiguing cable". That's how the space elevator thought experiment came about, but any practical implementation has to consider those physical properties of a cable. Your assumptions are not physically possible, which is why your implementation would not work. $\endgroup$
    – Graham
    Commented Aug 22, 2017 at 17:13
  • 8
    $\begingroup$ Adding more cables to a space elevator with current technology is akin to saying "We're going to take a loss, but we'll make up for it on volume." $\endgroup$
    – corsiKa
    Commented Aug 22, 2017 at 21:25
  • 5
    $\begingroup$ @immibis So what keeps it up then? The conventional space elevator cable extends past geostationary and has a counterweight on the end (for example a space station) in order to keep the whole thing up in the air. $\endgroup$
    – Tim B
    Commented Aug 23, 2017 at 9:43

Your question doesn't address the primary structural issues with space elevators. Your assumption seems to be that as the end of the cable gets farther away from the earth, the cable needs to be stronger. This is not, inherently, true.

First, we need to define what we mean by 'cable': Conventionally, a cable is a spool of intertwined smaller solid-metal cables that bind together to form a stronger, larger, cord. However once we're talking on the scale of a space elevator, the term 'cable' becomes much more broad. In that context, the cable is simply the mechanism that holds the end of the elevator (the "space station" so to speak) to the surface of the earth. A better word for it would be the "tether", rather than a cable, because it can be made in many different ways and with different structures, depending on the design. The tether can also include delicate things like power lines, plumbing, data cables, etc, as long as none are part of the load bearing structure.

No design for a space elevator (no sane design anyway) relies on a single large physical cable to act as the elevator tether. All designs use a series of cables, or trusses, or interconnected subsections like you are proposing.

From an engineering standpoint, here is the issue: whatever you use as a tether needs to be incredibly strong because it needs to withstand immense tensile force. Remember, in a space elevator the tether doesn't hold the space station up, in fact just the opposite: the space station is constantly trying to pull away from the earth, but is held in place by the tether. This keeps the tether pulled taught and keeps the tether from falling back to earth.

This means that we have two forces acting on the tether, both of which must (by definition) be exactly opposite to each other. On the planets surface, you have the force of gravity trying to pull the tether down. The longer the tether is (and it has to be long to reach geostationary orbit) the more massive it is, and therefore the more gravity has an effect on it. This has a tendency to pull the tether back down to the earth.

On the space side, in order to keep the tether from collapsing under its own weight, the space station needs to be pulling with an equal force to counter the weight of the tether. This is achieved through orbital momentum in sci-fi, but would require some form of acceleration in reality (like a really big rocket). Because the space station is in geostationary orbit, this gets very, very complicated if you want to both hold the tether up and maintain your orbit.

In summary, the end result of this is that you end up with two forces on the tether, each pulling in opposite directions. Firstly, you have the gravitational force pulling down on the tether. Secondly, you have the orbital force of the space station holding the tether up. This results in the tether being under enormous tension between these two forces. Reinforcing the cable with more material just makes it heavier, compounding the problem.

So no, your design would not fix the primary issues with a space elevator design.

Diagram included for clarity

enter image description here

There are two critical things to note in this diagram:

  • The total force on the tether is equal to 2*(G*m), not just G*m
  • As the mass of the tether m increases, the force on the tether also increases
  • 5
    $\begingroup$ There would need to be more cable out past geostationary orbit to a counterweight. $\endgroup$
    – sphennings
    Commented Aug 22, 2017 at 16:42
  • 6
    $\begingroup$ @origimbo it does make some sense to put the most massive structure (that's what sphennings means by “counterweight” at the end, but that must be well beyond geostationary distance to compensate for the heavy tether which is obviously for to most part closer than geostationary. — Anyway that's not necessarily the smartest layout: if you want to use the elevator to launch space missions, you can gain a lot of free Δv by deliberately extending it a lot beyond geostationary orbit, to act as a slingshot. $\endgroup$ Commented Aug 22, 2017 at 18:46
  • 4
    $\begingroup$ An object at geostationary-orbit altitude wouldn't exert any force on the cable; an object at on a higher altitude traveling tangent to the earth fast enough to remain geostationary ,however, would need to be pulled toward the earth to remain at that same altitude. $\endgroup$
    – supercat
    Commented Aug 22, 2017 at 20:38
  • 3
    $\begingroup$ You don't need a rocket on the station at the end of the tether. The station is slightly past geostationary orbit (or extends a counterweight past it) and as a result is being pulled away from the Earth by centripetal force. Climbers/launches would steal some of its orbital energy but the pull on the tether should allow it to regain that by tapping into the rotational energy of the Earth. $\endgroup$
    – Tim B
    Commented Aug 23, 2017 at 9:47
  • 1
    $\begingroup$ here this will let you calculate the tension. pdfs.semanticscholar.org/d402/… $\endgroup$
    – John
    Commented Aug 23, 2017 at 15:20

Failure Rates:
The more moving parts you have, the more likely things are to break. Why is this? Well, if you have a part that works 99.99% of the time, and have a thousand of them, then your final product works only 90% of the time. Adding a pulley and moving the cable increase the probability of catastrophic failure thousands of times (can the cable even bend? Will it fatigue?). For every bolt you add, you both increase the strength requirements for the cable and you add another potential failure point.....

That's just the way it is. In the case of a failure for a space elevator, your station slings off into space and you have to catch it a week or so later when it returns (yay - orbits), not to mention the expense of replacing the line. As a result, nearly everyone assumes we will see a single cable supporting the structure (very simple), and anything else will clamp on/around it (failure is non catastrophic for the station). This likely has a near-zero failure rate if your material is strong/durable enough.

As hinted at before, every gram of weight turns into tensile strength at geostationary orbit. Thus, we want everything to be as light as possible. Any station you add will require a bigger counterweight and stronger (or a higher count of) cables.

If you have some cable material that is far far stronger than required (which is far far far far stronger than currently possible), then engineers may start considering something as you describe. A weight anywhere other than geostationary increases the load on the cable, and we can't currently produce a material that will work for a space elevator, let along one strong enough to hang other equipment off.

What is the most expensive part of a space elevator? Probably either:

  • Getting the counterweight into space
  • The cost of the cable

If you have multiple cables (as per your diagram) or a revolving one, you've hugely increased the cost of the elevator. You've also made it harder to manufacture, so it will require longer to build - again driving the cost up.


So if you have handwavium'ly strong cable, your engineers are fantastic at designing fail-proof systems, and you have a high budget you can make it work.

Ways of powering a climber that I can think (and hence by no means exhaustive) of include:

  • 2
    $\begingroup$ I'm glad you mentioned the failure rates. I was going to write a (nowhere near as well written) answer, but now I don't have to haha $\endgroup$
    – user41674
    Commented Aug 22, 2017 at 19:18
  • $\begingroup$ Three 4/0 copper wires feeding the car, and done. Pump it up to probably 800 amps, at 20kv or whatever the car wants. That's 28MW. Earth station increases the voltage as the car climbs to make up for resistance losses, 4/0 is 16 ohms per 100km. Want the wire hot enough to shed energy by radiation in space, but not so hot it anneals. $\endgroup$ Commented Aug 24, 2017 at 7:43
  • $\begingroup$ At 800 amps, with 16 ohms per 100km, you're shedding 10MW per 100km. (P = I^2 / R). If this cable has 10cm circumference (~3cm diameter), it has a surface area of 100,000 square meters. When it reaches thermal equilibrium, it will be dissipating that 10MW. Assuming thermal loss is only through radiation, the cable will hover at -68 degrees Celsius. I think we need to up the current a bit more to stop the wire from becoming brittle. That 28MW probably station isn't going to be enough for more than a few hundred kilometers. $\endgroup$
    – sdfgeoff
    Commented Aug 24, 2017 at 11:46

If I understand your drawing correctly, there would be a loop of high tensile cable with a gear system to turn it, so you could latch stuff to the cable and it will be lifted via the cable?

Couple of problems here. For starters, you gotta move the cable. So now you have a massive power requirement at the ends of the loops. So instead of powered cars riding up and down, you have unpowered cars but gotta power the entire loop instead.

Second, you will need to keep the loops from twisting. These loops are gonna have to be miles long, so they will rub and twist on each other unless the distance between each side is substantial or you have spacers between them, which adds more cost/weight.

Third, the speed of the cable will have to be pretty high if loads are going to get anywhere in a reasonable amount of time, and there is no easy way to add cars to it without having a separate parallel track for a car to get up to speed and then hop on the cable. This is additional cost and weight.

Of note, there is a concept somewhat like this called the "space fountain" space fountain concept which uses rail gun to shoot a steady stream of balls up into a "floating" platform that is held aloft by the momentum of the balls and which redirects them back down in a stream to be recaptured and fired again. This creates a discontinuous "loop" that works very similarly to your concept except that it provides lift to the top of the tower (if the balls were fired inside a supported tunnel, allowing things to move on the outside) and doesn't need the long counterweight that a traditional tethered space elevator (and your chain loop concept) would require.

enter image description here Image borrowed from orionsarm.com

What you really need for your concept (other than highly advanced materials and power sources) is a chain that can support its own weight in a rigid manner but is flexible enough to be looped. So imagine that the chain, when running up into the sky, stacks on itself to form a rigid pole that is essentially pushing itself up, then it loops back down. If this is long enough to reach low earth orbit it will suffice as a space elevator and won't need a long counterweight to hold itself up. Neal Stephenson talks about stuff like this in Seveneves which has a lot of looped chain based launch and recovery systems.

  • $\begingroup$ @Will : ​ ​ ​ ​ ​ ​ ​ The starting sentence of the wikipedia article for that theorem suggests that it is about ​ "stationary equilibrium" , ​ i.e., the meaning one would expect from the word "equilibrium". ​ ​ ​ A space fountain would most definitely not be trying to keep the balls in such an equilibrium. ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ $\endgroup$
    – user3576
    Commented Aug 23, 2017 at 3:19
  • $\begingroup$ @Will : ​ ​ ​ A space fountain would not be trying to do anything like balancing them on top of each other - It would instead be playing a game of galactic catch. ​ (There would be an upwards force on the space station due to it quickly accelerating the balls from upward to downward.) ​ ​ ​ ​ ​ ​ ​ ​ $\endgroup$
    – user3576
    Commented Aug 23, 2017 at 5:08
  • $\begingroup$ @Will : ​ ​ ​ ​ ​ ​ ​ One wouldn't. ​ (For my previous comment, "them" is the balls.) ​ ​ ​ Nothing. ​ (I'd imagine that both stations are large enough that balls are caught despite the sideways drift that occurs on each of their trips between the stations.) ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ (... continued) ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ $\endgroup$
    – user3576
    Commented Aug 23, 2017 at 6:31
  • $\begingroup$ (continued ...) ​ ​ ​ I'd imagine there being at least 3 downward streams, and that if there's only one upward stream, then that stream is in the center. ​ (I well realize that more may be needed.) ​ For each downward stream, the station could adjust the speed at which it sends that stream's balls. ​ ​ ​ (... continued) ​ ​ ​ ​ ​ ​ ​ $\endgroup$
    – user3576
    Commented Aug 23, 2017 at 6:31
  • $\begingroup$ (continued ...) ​ ​ ​ Shear forces imposed by the system on the ground station would be taken up by Earth or the celestial body, since the ground station would be a standard structure on Earth or that body. ​ I'd imagine that shear forces on the space station would be handled as I speculated two sentences ago. ​ Otherwise, what points in the system do you mean? ​ (When in either station, the balls would probably be subjected to strong shear forces to move them horizontally between the streams. ​ ​ ​ ​ ​ ​ ​ $\endgroup$
    – user3576
    Commented Aug 23, 2017 at 6:31

I've not seen anyone here mention that current elevator technology limits an elevator rise to around 500 meters in one go - this is a significant problem for tall buildings like the Birj Khalifa, which is 800+ meters tall and has 2 elevator runs to get to the top (actually 3, since there's an elevator in the spire), as well as the longest continuous run elevator in the world (504 meters). There is a new installation going into the Saudi Tower that will be 660M, and there is some new materials research that promises to extend the run to over 1 km, but still - that would be 100 runs to get to space (100KM) and probably 250 to get to the station (I've seen a lot of proposed heights for the station in various elevator discussions).

My point is that we'd need much improved materials technology to make a scheme like this work (or for a space elevator to work). Given current materials. doesn't seem like a win.

  • $\begingroup$ You'd need much more than 250km. You'd need a cable all the way to GEO and back, at least 72000km, to go all the way. $\endgroup$
    – jwenting
    Commented Mar 26, 2019 at 9:55

Please look at www.isec.org for a growing BOK on space elevators.

You can't terminate a cable at geo. You need to overcome gravity with centripetal acceleration from an apex node mass. This will allow you to maintain tether tension and allow mass to traverse the tether. The geo sync orbit is only one exit/entry gate to the tether. LEO, moon and mars are also available.

Tether tensil strength is certainly an major issue, but there are others. If you are interested in volunteering to work on these issue, join the isec.org and get involved.


This idea is bad. Why? Because they are already dropping cables in favor of crawlers for REGULAR SKYSCRAPER ELEVATORS. Thyssen-Krupp has been running a test site for crawler-"elevators" for just about 2 years now, don't even have the certification for human testing yet and ALREADY have a stable stream of customers for this tech. The reason for that is that at high speeds you not only need long and strong cables, but due to the cables moving at very high speeds you also have to keep them from moving, making the whole elevator even more complex, AND cables can only support one cabin per slot at one time, while with crawlers you can send up multiple at a time, meaning the whole thing is way more efficient.

Here's a news article: https://www.thetimes.co.uk/article/magnets-to-lift-the-first-elevators-without-cables-tnvbnx7zg [I didn't read that one, but its the first I found on the subject in English, since most of it would be German]

  • $\begingroup$ You missed completely the point. In a "real life" elevator distributing power is not a problem as you have a whole, wired, building around you. OTOH a "space elevator" has no way to bring power , so current design relies on self power (poor choice due to extra weight) or laser beams to send power "wirelessly" (efficiency is still low, but already way better than rockets). $\endgroup$
    – ZioByte
    Commented Mar 27, 2019 at 10:25
  • $\begingroup$ Nah, I ignored power transportation, because power-over-laser is already good enough for this application. At least it would be far better than having cables. Also given the recent developments with 1-atom thick carbon "wires" its even reasonable to assume that plating the cable with some sort of "wire" for power is feasible as well by then $\endgroup$
    – Hobbamok
    Commented Mar 28, 2019 at 7:48

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .