The Space Elevator: ‘Thought Experiment’, or Key to the Universe?
Editor’s Note: The following paper was first published in 1981 in Advances in Earth
Oriented Applied Space Technologies. It is reprinted here with the authors permission.
Visit the new Space Elevator Reference
by Sir Arthur C. Clarke
Abstract — The space elevator (alias Sky Hook, Heavenly Ladder, Orbital Tower, or Cosmic Funicular) is a structure
linking a point on the equator to a satellite in the geostationary orbit directly above it. By providing a ‘vertical
railroad’ it would permit orders-of-magnitude reduction in the cost of space operations. The net energy requirements
would be almost zero, as in principle all the energy of returning payloads could be recaptured; indeed, by continuing
the structure beyond the geostationary point (necessary in any event for reasons of stability) payloads could be given
escape velocity merely by utilising the ‘sling’ effect of the Earth’s rotation.
The concept was first developed in detail by a Leningrad engineer, Yuri Artsutanov, in 1960 and later by several American
engineers quite unaware of Artsutanov’s work. All studies indicate that the idea, outrageous though it appears at first
sight, is theoretically feasible and that its practical realisation could follow from the mass-production of high-strength
materials now known as laboratory curiosities.
This paper is a semi-technical survey of the rapidly expanding literature of the subject, with some speculations about
ultimate developments. Whether or not the Space elevator can be actually built, it is of great interest as the only known
device which could replace the rocket as a means of escaping from the earth. If it is ever developed, it could make mass
space travel no more expensive than any other mode of transportation.
INTRODUCTION
What I want to talk about today is a space transportation system so outrageous that many of you may consider it not even
science-fiction, but pure fantasy. Perhaps it is; only the future will tell. Yet even if it is regarded as no more than a
‘thought-experiment’, it is one of the most fascinating and stimulating ideas in the history of astronautics.
This paper is essentially a survey; in the unlikely event that it contains anything original, it’s probably wrong. Your
complaints should be addressed to Director-General Roy Gibson, who is responsible for getting me here.
What’s in a name?
First of all, we have a severe problem in nomenclature. It is very difficult to talk about something, until people
have agreed on its name. In this case, we have an embarrassingly wide choice.
The Russian inventor used the charming ‘heavenly funicular’. American writers have contributed ‘orbital tower’,
‘anchored satellite’, ‘beanstalk’, ‘Jacob’s Ladder — and, of course, ‘Skyhook’. I prefer’space elevator’; it is
euphonious (at least in English) and exactly describes the subject.
Historical
As usual, it all began with Tsiolkovski — specifically, with his 1895 paper ‘Day-Dreams of Heaven and Earth'[1].
During his discussion of possible ways of escaping from the earth, he considered the building of a high tower, and
described what would happen as One ascended it. I quote:
‘On the tower, as one climbed higher and higher up it, gravity would decrease gradually; and if it were constructed on
the Earth’s equator and, therefore, rapidly rotated together with the earth, the gravitation would disappear not only
because of the distancefrom the centre of the planet, but also from the centrifugal force that is increasing
proportionately to that distance. The gravitational force drops. . . but the centrifugal force operating in the
reverse direction increases. On the earth the gravity is finally eliminated at the top of the tower, at an elevation of
5.5 radii of the earth (36000 km).
‘As one went up such a tower, gravity would decrease steadily, without changing direction; at a distance of 36000 km,
it would be completely annihilated, and then it would be again detected. . . but its direction would be reversed, so that
a person would have his head turned towards the earth….’
Tsiolkovski then calculates the height of similar towers on the Sun and planets, but his comments — at least as I read
the translation (‘it would be excessive to discuss how possible these towers would be on the planets’ he suggests that he
does not regard the concept as a serious practical proposition. And of course he is quite right: it would be impossible,
if I dare use such a risky word, to construct free-standing towers tens of thousands of kilometres high. If Tsiolkovski
failed to mention the alternative solution, it may be because he was concerned only with the first steps away from earth.
And the space-elevator is completely useless in the pioneering days of astronautics, unless you are lucky (?) enough to
live on a very small, rapidly spinning planet.
Nevertheless, it is interesting to find how high atower we could build, if we really tried. In early 1962 the Convair
Division of General Dynamics carried out a feasibility study, to see if very high towers would be of value for astronomy,
high altitude research, communications and rocket launching platforms [2]. It turns out that steel towers could be built
up to 6 km high, aluminium ones up to almost 10. Nature can do just as well; it would be cheaper to use Mount Everest.
However, we now have much better materials than steel and aluminium in the form of composites, which give both high
strength and low density. Calculations show that a tower built of graphite composite struts could reach the very
respectable height of 40 km, tapering from a 6 km-wide base. I dare not ask what it would cost; but it’s startling to
realise that, even with today’s technology, we could build a structure 100 times as high as the world’s tallest building.
But the geostationary orbit is a thousand times higher still, so we can forget about building up towards it. If we hope to
establish aphysical link between Earth and space, we have to proceed in the opposite direction — from orbit, down-wards.
That it might be useful to hang a long cable from a satellite must have occurred to a great many people. I myself toyed
with the idea in 1963 while preparing an essay on comstats for UNESCO, published next year in Astronautics [3]. At that
time, there was still considerable uncertainty about the effects of the time delay in satellite telephone circuits; some
thought that it might have proved intolerable in ordinary conversation.
Although we should be duly grateful to Nature for giving us the geostationary orbit, one can’t help wishing, for
INTELSAT’S sake, that it was a good deal closer. So I wondered if it would be possible to suspend a satellite repeater
10000 or more kilometres below the 36000 km altitude that the law of gravity, and the Earth’srotational speed, has
dictated.
Some desultory calculations soon convinced me that it couldn’t be done with existing materials, but as I wanted to leave
the option open I wrote cautiously: ‘As a much longer-term possibility, it might be mentioned that there are a number of
theoretical ways of achieving a low-altitude, twenty-four hour satellite; but they depend upon technical developments
unlikely to occur in this century. I leave their contemplation as “an exercise for the student”!’ In 1969, 6 years later,
Collar and Flower came to exactly the same conclusion in a J.B.I.S. paper ‘A (relatively) low altitude 24-h satellite’
[4]. To quote their summary:
‘The scheme for launching a twin-satellite system into a 24-h orbit with the inner satellite relatively close to the
earth’s surface is theoretically possible, although with the materials currently available no operational advantage
would result. New materials now being developed, however, if used to the limit of their strength, could result in a
system that considerably improved communication efficiency. Even with materials that are strong enough and light enough
many problems exist. Static and dynamic stability investigations would need to be made, and temperature effects allowed
for. In the design of the system, means of deployment and of minimising meteorite damage would in particular need careful
consideration.
‘The final conclusion is that while theoreticaly possible, the twin satellite system is impractical at the present time,
but will show ever increasing promise as new, strong, light materials are developed.
Incidentally, Collar and Flower did mention that it would be possible for the cable to reach all the way down to
the Earth’s surface, though they did not elaborate on this point, and were apparently unaware of earlier work in this
field. For it now appears that at least half a dozen people invented the space elevator quite independently of each
other, and doubtles more pioneers will emerge from time to time.
In the West, the group that got there first consisted of John Isaacs, Allyn Vine, Hugh Bradner and George Bachus, from
the Scripps Institute of Oceanography and the Woods Hole Oceanographic Institute. It is, perhaps, hardly surprising that
oceanographers should get involved in such a scheme, since they are about the only people who concern themselves with
very long hanging cables. Very long, that is, by ordinary standards; but in their 1966 letter to Science [5] Isaacs et al.
discussed a cable over three thousand times longer than one to the bottom of the Marianas Trench, a mere 11 km down.
Their brief but very comprehensive paper made the following points:
The cable would have to be tapered, and would have to be spun out in both directions simultaneously — that is, towards
the Earth and away from it, so that the structure was always balanced around the geostationary point. One would start
with the smallest possible cable — perhaps with a minimum diameter of only a few thousandths of a centimetre — and the
lower end would have to be guided down to earth by some kind of reaction device. Once the initial cable had been
established between stationary orbit and the point on the equator immediately below it, it could be used to establish a
stronger cable, until one of the required carrying capacity was attained. In principle, it would then be possible to
hoist payloads from earth into space by purely mechanical means.
Now, you will recall that, as one ascends Tsiolkovski’s hypothetical space tower, gravity decreases to zero at stationary
orbit — and its direction then reverses itself. In other words, though one would have to do work to get the payload up
to the geostationary position, once it had passed that point it would continue to travel on outwards, at an
increasing acceleration — falling upwards, in fact. Not only would it require no energy to move it away from earth — it
could generate energy, which could be used to lift other payloads! Of course, this energy comes from the rotation of the
earth, which would be slowed down in the process. I have not attempted to calculate how much mass one could shoot off
into space before the astronomers complained that their atomic clocks were running fast. It would certainly be a long
time before anyone else could notice the difference….
Isaacs et al. go on to say:
‘In addition to their use for launching materials into space, such installations could support laboratories for
observation of conditions in space at high altitudes; they could resupply energy or materials to satellites or
spacecraft, collect energy or materials from space and the high atmosphere, support very tallstructures on the earth’s
surface, and others. There isno immediate limit to the total mass that could be retained near the l-day orbit by such a
cable.’
Isaacs et al., discussed only briefly the obviouslyvital question of possible materials, listing amongst others quartz,
graphite and beryllium. The total mass, with the best material, of a cable strong enough towithstand 200 km h -1 wind
forces, turns out to be surprisingly low — only half a ton! Needless to say, its diameter at the earth and would be
extremely small — one five-hundredth of a centimetre. And before anyone starts to spin this particular thread, I
should point out that the material proposed is quite expensive. I don’t know what the market quotation would be for
half a ton of — diamond.
The first reaction to the Isaacs paper came some three months later [6] when an unlucky American scientist fell into a
neat dynamical trap. He was not the only one to do so, apparently; but James Shea was unfortunate enough to have his
letter published. The objection he raised was ingenious and so apparently convincing that he stated flately: ‘The system
is inefficient as well as mechanically unsound and theoretically impossible.’ (My italics.)
Shea’s paradox can best be appreciated as follows: Consider the payload to be sent up the cable, when it is resting on
the equator at the beginning of its journey. Obviously, because of the Earth’s spin, it’s actually moving eastwards at
about 1700 km h -1.
Now it is sent up the cable — how it is, is for the moment, unimportant — until it reaches the geostationary orbit,
36000 km above the Earth. It is still exactly above the point from which it started, but almost six times further from the
centre of the Earth. So to stay here, it must obviously move six times faster — about 11000 km h -1. How does
it acquire all this extra tangential velocity?
There’s no problem when you consider the analogous case of a fly crawling from the hub out along the spoke of a spinning
wheel. The wheel is a rigid structure, and automatically transmits its rotational velocity to the fly. But how can a
flexible cable extending out into space perform the same feat?
The explanation may be found by looking at one of mankind’s simplest, oldest and most cost-effective weapons — the sling.
I wonder if Goliath’s technical advisers told him not to worry about that kid with the ridiculous loop of cloth — it
couldn’t possibly transfer any kinetic energy to a pebble. If so, they forgot that the system contained a rigid
component — David’s strong right arm. So also with the space elevator. Its lower end is attached to the 6000 km radio of
the Earth — quite a lever.
Having easily refuted this criticism, Isaacs & Co. were now in for a shock. More than a year after their letter had
appeared, Science printed a lengthy note [7] from Vladimir Lvov, Moscow correspondent of the Novosti Press Agency,
pointing out that they had been anticipated by a half a decade. A Leningrad engineer, Yuri N. Artsutanov, had already
published an article in Pravda which not only laid down all the basic concepts of the space elevator, but developed them
in far greater detail.
This 1960 paper, which may turn out to be one of the most seminal in the history of astronautics, has the unassuming
title ‘Into the Cosmos by electric vehicle’*. Unfortunately, it has never been translated into English, nor have the
extensive calculations upon which it is obviously based yet been published. The summary that follows is therefore based
on Lvov’s letter.
* Artsutanov’s own translation is: ‘Into space with the help of an electric locomotive’ (See Acknowledgements).
Artsutanov’s initial minimum cable, constructed from materials which already exist but which have so far only been
produced in microscopic quantities, would be able to lift two tons, would have a diameter of about one millimetre at
the earth’s surface, and would have a total mass of about 900 tons. It would extend to a height of 50000 km — that is,
14000 km beyond geostationary altitude, the extra length providing the additional mass needed to keep the whole system
under tension. (The weight, as it were, on the end of the sling.)
But this is just a beginning. Artsutanov proposed to use the initial cable to multiply itself, in a sort of boot-strap
operation, until it was strengthened a thousand fold. Then, he calculated, it would be able to handle 500 tons an hour or
12000 tons a day. When you consider that this is roughly equivalent to one Shuttle flight every minute, you will
appreciate that Comrade Artsutanov is not thinking on quite the same scale as NASA. Yet if one extrapolates from
Lindbergh to the state of transatlantic air traffic 50 yr later, dare we say that he is over-optimistic? It is doubtless
a pure coincidence, but the system Artsutanov envisages could just about cope with the current daily increase in the world
population, allowing the usual 22 kg of baggage per emigrant….
Lvov uses two names to describe Artsutanov’s invention: a ‘cosmic lift’, and a ‘heavenly funicular’. But a funicular,
strictly speaking, is a device operated by a rope or cable — and we may be sure that the space elevator will not hoist
its payloads with the aid of moving cables tens of thousands of kilometres long.
One would have thought that this correspondence, in one of the world’s leading scientific journals, would have triggered
a large scale discussion. Not a bit of it; to the best of my knowledge, there was no reaction at all. This may be because
the Apollo project was then moving towards its climax — the first moon landing was less than two years away — and
everyone was hypnotised by big rockets, as well they might be.
But the idea must have been quietly circulating in the U.S.S.R. because it is illustrated in the handsome volume of
paintings by Leonov and Sokolov “The Stars are Awaiting Us” (1967). On p. 25 there is a painting entitled
‘Space Elevator’, showing an assembly of spheres — hovering, I am pleased to see, over Sri Lanka — from which a cable
stretches down to the earth. Part of the descriptive text reads as follows:
‘If a cable is lowered from the (24 h) satellite to the earth you will have a ready cable-road. An “Earth-Sputnik-Earth”
elevator for freight and passengers can then be built, and it will operate without any rocket propulsion.’
Rather surprisingly, there is no reference to the inventor. Sokolov’s original painting, incidentally, hasbeen acquired by
my insatiably acquisitive friend Fred Durant for the National Air and Space Museum, which by now must surely have the
world’s finest collection of space art (and space hardware).
The next major development was not for another eight years. Then Jerome Pearson of the Flight Dynamics Laboratory,
Wright-Patterson Air Force Base, invented the idea all over again and published the most comprehensive study yet in Acta
Astronautica [8]. His computer search of the literature had failed to turn up any prior references, and in view of the
indexing problem I’m not surprised. How would you look up such a subject? Pearson called it an ‘orbital tower’, and
presumably never thought of telling his computer to hunt for ‘sky-hook’, which might have located the Science
correspondence.
I speak with some feeling on this, because for 2 years I was solely responsible for indexing Physics Abstracts, and you’ll
find some very strange entries round 1950. But the problem is insoluble, unless you can do retrospective re-indexing. When
a new phenomenon is discovered, you may not even know how to classify it, let alone what to call it (after all, we’re
still stuck with X-rays, after almost a century. . . ).
Pearson’s 1975 paper was the most thorough study of the project yet published, and emphasised one of the space elevator’s
most important characteristics. Like a terrestrial elevator, it could be used in both directions. Returning payloads could
be brought back to Earth without the use of heat-shields and atmospheric braking. Not only would this reduce environmental
damage; it would mean that virtually all the energy of re-entry could be recovered, and not wasted as is the case today.
This characteristic makes the space elevator unique — at least, until someone invents anti-gravity. It is a conservative
system. If, as would probably be the case, electrical energy is used to lift payloads up the elevator, and the mass flow
is the same in both directions, incoming traffic could provide all the energy needed to power outgoing traffic. In
practice, of course, there could be the inevitable conversion and transmission losses, but they could be quite small.
In this and subsequent papers [9] Pearson was the first to go into the dynamics of the system, discussing the vibration
modes of the structure due to launch loads, gravity, tides etc. He decided that none of these, though important, would
cause any insuperable problems. Indeed, I have suggested elsewhere that they could even be used to advantage [10].
Pearson has also located at least three other independent originators of the concept [11], though none prior to
Artsutanov’s 1960 paper. From now on, at least, further re-invention is unnecessary; however, as we shall see later, novel
and often surprising extensions of the basic system are still appearing.
