- Press Release
- August 8, 2022
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.
by Sir Arthur C. Clarke
THE PROBLEM OF MATERIALS
The very minimum requirement for a space elevator is, obviously, a cable strong enough to support its own weight when
hanging from geostationary orbit down to earth, 36000 km below. That is a very formidable challenge; luckily, things are
not quite as bad as they look because only the lowest portion of the cable has to withstand one full gee.
As we go upwards, gravity falls off according to Newton’s inverse square law. But the effective weight ofthe cable
diminishes even more rapidly, owing to the centrifugal force* on the rotating system. At geostationary altitude the
two balance and the net weight is zero; beyond that, weight appears to increase again — but away from the Earth.
* My brief apologies to purists for invoking this fictitious entity.
So our cable has no need to be strong enough to hang 36000 km under sea-level gravity; allowing for the effects just
mentioned, the figure turns out to be only one-seventh of this. In other words, if we could manufacture a cable with
sufficient strength to support 5000 km (actually, 4960) of its own length at one gee, it would be strong enough to span
the gap from geostationary orbit to Equator. Mathematically — though not physically — Jacob’s ladder need be only
5OOO km long to reach Heaven…. This figure of 5000km I would like to call ‘escape length’, for reasons which will
soon be obvious.
How close are we to achieving this with known materials? Not very. The best steel wire could manage only a miserable 5O km
or so of vertical suspension before it snapped under its own weight. The trouble with metals is that, though they are
strong, they are also heavy; we want something that is both strong and light. This suggests that we should look at the
modern synthetic and composite materials. Kevlar (Tm) 29, for example  could sustain a vertical length of 200 km
before snapping — impressive, but still totally inadequate compared with the 5000 needed.
This ‘breaking length’, also known as ‘rupture length’ or ‘characteristic length’, is the quantity which enables one to
judge whether any particular material is adequate for the job. However, it may come as a surprise to learn that a cable
can hang vertically for a distance many times greater than its breaking length!
This can be appreciated by a simple ‘thought experiment’. Consider a cable which is just strong enough to hang vertically
for a hundred kilometres. One more centimetre, and it will snap….
Now cut it in two. Obviously, the upper 50 km can support a length of 50 km — the identical lower half. So if we put the
two sections side by side, they can support a total length of 100 km. Therefore, we can now span a vertical distance of
150 km, using material with only 100 km breaking length.
Clearly, we can repeat the process indefinitely, bundling more and more cables together as we go upwards. I’m sure that by
now you’ve recognised an old friend — the ‘step’ principle, but in reverse. Step rockets get smaller as we go higher; step
cables get bigger.
I apologise if, for many of you, I’m labouring the obvious, but the point is of fundamental importance and the rocket
analogy so intriguing that I’d like to take it a little further.
We fossils from the pre-space age — the Early Paleoastronautic Era — must all remember the depressing calculations we
used to make, comparing rocket exhaust velocities with the 11.2 km 5′ of Earth escape velocity. The best propellants we
knew then — and they are still the best today! — could provide exhaust velocities only a quarter of escape velocity.
From this, some foolish critics argued that leaving the Earth by chemicalrocket was impossible even in theory.
The answer, of course, was the step or multi-stagerocket — buteven this didn’t convince some sceptics. Willy Ley 
records a debate between Oberth and a leading German engineer, who simply wouldn’t believe that rockets could be built with
a mass-ratio of twenty. For Saturn V, incidentally, the figure is about five hundred.
We escaped from earth using propellants whose exhaust velocitywas only a fraction of escape velocity, by paying the heavy
price demanded by multi stage rockets. An enormous initial mass wasrequired for a small final payload.
In the same way, we can achieve the 5000 km ‘escape length’, even with materials whose breaking length is a fraction of
this, bysteadily thickening the cable as we go upwards. Ideally, this should be done not in discrete steps, but by a
continuous taper.The cable should flare outwards with increasing altitude, its cross-section at any level being just
adequate to support theweight hanging below.
With a stepped, or tapered, cable it would be theoretically possible to construct the space elevator from any material,
however weak. You could build it of chewing gum, though the total mass required would probably be larger than that of the
entire universe. For the scheme to be practical we need materials with a breaking length a very substantial fraction of
escape length. Even Kevlar 29’s 200 km is a mere 25th of the 5000 km goal; touse that would be like fuelling the Apollo
mission with damp gunpowder, and would require the same sort of astronomical ratio.
So, just as we were once always seeking exotic propellents, wemust now search for super-strength materials. And, oddly
enough, we will find them in the same place on the periodic table.
Carbon crystals have now been produced in the laboratory withbreaking lengths of up to 3000 km — that is, more than half
of escape length. How happy the rocket engineers would be, ifthey had a propellant whose exhaust products emerged with
60% of escape velocity!
Whether this material can ever be produced in the megaton quantities needed is a question that only future technologies
can answer; Pearson  has made the interesting suggestion that thezero gravity and vacuum conditions of an orbiting
factory may assist their manufacture, while Sheffield  and I  havepointed out that essentially unlimited quantities
of carbon are available on many of the asteroids. Thus when space mining is infull swing, it will not be necessary to use
super-shuttles to lift vast quantities of building material up to geostationary orbit — a mission which, surprisingly, is
somewhat more difficult than escaping from Earth.
It is theoretically possible that materials stronger — indeed,vastly stronger — than graphite crystals can exist.
Sheffield has made the point that only the outer electrons of theatoms contribute, through their chemical bonds, to
the strength of a solid. The nucleus provides almost all the mass, but nothingelse; and in this case, mass is just what we
So if we want high-strength materials, we should look at elements with low atomic weights — which is why carbon (A.W.12)
is good and iron (A.W.56) isn’t. It follows, therefore, that the best material for building space elevators is — solid
hydrogen! Infact, Sheffield calculates that the breaking length of a solidhydrogen crystal is 9118 km — almost twice
By a curious coincidence, I have just received a press release from the National Science Foundation headed ‘New form of
hydrogen created as Scientists edge closer to creating metallic hydrogen'. It reports that, at a pressure of half a
million atmospheres, hydrogen has been converted into a dense crystalline solid at room temperature. The scientists
concerned go on to speculate that, with further research — and I quote — “hydrogen solids can be maintained for long
periods without containment”.
This is heady stuff, but I wonder what they mean by ‘long periods’. The report adds casually that ‘solid hydrogen is 25 to
35 times more explosive than TNT’. So even if we could make structures from solid hydrogen, they might add a new dimension
to the phrase ‘catastrophic failure’.
However, if you think that crystallitic hydrogen is a tricky building material, consider the next item on Dr. Sheffield’s
shopping list. The ultimate in theoretical strength could be obtained by getting rid of the useless dead mass of the
nucleus, and keeping only the bonding electrons. Such a material has indeed been created in the laboratory; it’s
‘positronium’ — the atom, for want of a better word, consisting of electron-positron pairs. Sheffield calculates that the
breaking length of a positronium cable would be a fantastic 16,700,000 km! Even in the enormous gravity field of Jupiter, a
space elevator need have no appreciable taper.
Positronium occurs in two varieties, both unfortunately rather unstable. Para-positronium decays into radiation in one-tenth
of a nanosecond — but orthopositronium lasts a thousand times longer, a whole tenth of a microsecond. So when you go
shopping for positronium, make sure that you buy the brand marked ‘Ortho’.
Sheffield wonders wistfully if we could stabilise positronium, and some even more exotic speculations are made by Moravec
. He suggests the possible existence of ‘monopole’ matter, andhybrid ‘electric/magnetic’ matter, which would give not
only enormous strength but superconductivity and other useful properties.
Coming back to earth — or at least to this century — it seems fair to conclude that a small cable could certainly be
established from geostationary orbit down to sea level, using materials that may be available in the near future. But that,
ofcourse would be only the first part of the problem — a mere demonstration of principle. To get from a simple cable to
a working elevator system might be even more difficult. I would now like to glance at some of the obstacles, and suggest a
few solutions; perhaps the following remarks may stimulate others better qualified to tackle them.
DEPLOYING THE CABLE
The space elevator may be regarded as a kind of bridge, and many bridges begin with the establishment of a light initial
cable — sometimes, indeed, no more than a string towed across a canyon by a kite. It seems likely that the space elevator
will start in the same way with the laying of a cable between geo stationary orbit and the point on the equator
This operation is not as simple as it sounds, because of the varying forces and velocities involved, not to mention the
matter of air resistance after atmospheric entry. But there are two existing technologies which may provide a few answers,
or at least hints at them.
The first is that of submarine cable laying, now considerably more than a century old. Perhaps one day we may see in space
something analogous to the triumphs and disasters of the Great Eastern, which laid the first successful transatlantic
telegraph cable — the Apollo Project of its age.
But a much closer parallel, both in time and sophistication, lies in the development of wire-guided missiles. These
lethal insects can spin out their metallic gossamer at several hundred kilometres an hour. They may provide the prototype
of the vehicle that lays a thread from stationary orbit down to earth.
Imagine a spool, or bobbin, carrying some 40000 km of filament, a few tenths of a millimetre thick at the outer layers,
and tapering down to a tenth of this at the core — the end that finally reaches Earth. Its mass would be a few tons, and
the problem would be to play it out evenly at an average velocity of a kilometre a second along the desired trajectory.
Moreover, an equivalent mass has to be sent outwards at the same time, to ensure that the system remains in balance at the
My friend Professor Ruppehas investigated  the dynamics of the mission, and concludes that it can be achieved with
modest mass-ratios. But the mechanical difficulties would obviously be formidable, and it may well turn out that material
of such tensile strength is too stiff to be wound on to a spool of reasonable radius.
Sheffield  has suggested an alternative method of installation which I find — to say the least — hair raisingly
implausible. He proposes constructing the entire space elevator system in orbit, and then launching it towards the earth,
grabbing the lower end when it reaches the equator! The atmospheric entry of a few megatons dead weight, which must
impact within metres of the aiming point, seems likely to generate a lot of opposition from the environmentalists. I call
it ‘harpooning the Earth’, and would prefer not to be near one of the Poles if it’s ever tried out.