How to get into space cheaply

As I sat with my colleagues at one of the outdoor bars at our hotel in Cancun last month, I had an idea for a method of getting into space cheaply.  Since this is a problem of enormous general interest and commercial importance, I was sure it must have been thought of before, but when I checked Wikipedia’s non-rocket spacelaunch page (highly recommended, by the way!), I didn’t see my idea listed.  So could it possibly be novel after all?

[By happy coincidence, it’s only a few days ago that the world’s first Sushi In Space video was released.  Why?  I couldn’t tell you.  But that’s where this image is taken from.]


As in the sushi example above, balloons have often been used to lift light objects to the edge of space.  They have even been used as a launch platform for light rockets — the combination is called a rockoon.  The problem with balloons is that as they climb higher, the pressure of the atmosphere decreases, so the gas density difference inside and outside the balloon decreases, and so does its lifting power.

My idea is a progressively evacuated rigid balloon (which we can conveniently abbreviate PERB).  I think it can be used to get easily up to — well, I don’t know what height, that would depend on material properties — but high enough to make it a reasonable launch platform for smaller rockets.  (Many of the other non-rocket spacelaunch techniques also do not aim to get into orbit as the space elevator does, but just to make a rocket’s task easier by starting from a point some way up through the atmosphere.)

The PERB is simply a large, rigid shell strong enough to withstand a pressure differential of, say, 0.1 atmospheres without imploding.  You use a vacuum pump to evacuate air until the pressure inside is down to 0.9 atmospheres.  If your PERB is sufficiently voluminous, it will now be floating: what’s needed of course is a volume such that one tenth the mass of contained air at one atmosphere is greater than the mass of the PERB itself: the shell, the pump mechanism and any payload.

The beauty of the PERB is this: if  you can make it take off at all, then you can take it higher and higher very easily: you simply keep evacuating the shell to a pressure that is 0.1 atmospheres lower than the surrounding air, and let it float higher and higher.  (I am neglecting some details, of course.  For example, you’d want to tether it to the ground, and the mass of the tether itself will grow linearly with altitude.  I am guessing this needn’t be a killer.)

What is the limit?  I suppose the PERB can keep rising while there is enough atmosphere to give a pressure differential capable of lifting it.  You’d end up by fully evacuating the shell, at a point where the atmospheric density is something less than a tenth of an atmosphere.  From there, you launch your rocket into orbit by conventional means.  (Notice that it’s pleasantly simple to bring your PERB back down to ground level: you merely bleed air slowly into the shell.)


If this can be made to work at all, it seems to have many important advantages.

First, the launch cost would seem to be negligible, on a par with running a vacuum cleaner: all the significant cost would be in manufacture of the light, rigid shell.  Presumably your tether would double as an electrical supply, so you don’t need to carry any fuel.

Second, simplicity.  The PERB has almost no moving parts — only the pump — and (fingers crossed) seems pretty much foolproof.

Third, safety: because everything happens so slowly, if things do go wrong there should be time to do something about it.  Provided the shell doesn’t fail catastrophically by buckling — which should be easy to avoid just by evacuating slowly enough to allow the PERB to climb — if a leak develops, it should only cause the PERB to descend slowly.

Fourth, reusability.  I don’t see any reason why a PERB should not be used many times.


The big one: do we have a material that the shell can be made from?  It needs to be strong, light, and impermeable to air (or nearly so — very slow leakage shouldn’t be a problem, as the vacuum pump will easily overcome it).  Here, I am lost.  I know almost nothing about material properties.  Anyone?  Carbon nanotubes?

The PERB would likely be very vulnerable to bad weather.  Wind is the obvious hazard — it could conceivably break the tether; it could throw the PERB upwards or downwards too fast for the pump to compensate for pressure changes, causing the shell the implode or explode; it could throw solid objects against the shell causing a local failure.  Perhaps even the pressure of rain could be enough to defeat the lifting power, if it acts over a wide enough area (although this only applies at low altitudes).

Would it need to operate at, or near, the equator?  I am not clear on why, exactly, but I seem to recall that the space elevator can only be built at the equator for reasons that elude me (and which the Wikipedia article is strangely silent about).

Have I missed any?


Obviously some calculations need to be done here.  The key one is this: can we make a shell voluminous and strong enough enough that it can sustain a pressure differential great enough to counteract its mass?  And how much additional mass can it carry while maintaining a healthy safety factor?  Does anyone out there have the relevant background to work through this with me?

Here is the easy part.  Consider a PERB with a shell of mass Ms and volume V, capable of withstanding a density differential of P atmospheres.  The mechanism (vacuum pump and associated stuff) has mass Mm and we want to lift a payload of mass Mp.  The density of air at one atmosphere is 1.2 kg/m^3, so the difference in mass between the enclosed partially evacuated volume and the equivalent volume of air is 1.2 P V.  That buoyancy has to overcome the mass of the shell, mechanism and payload, so we have lift-off when 1.2 P V > Ms + Mm + Mp.

The hard part is calculating the strength of a shell of mass Ms and volume V made from some specific material in some specific shape.  Is it capable of sustaining a pressure difference of P atmospheres, where P = (Ms + Mm +Mp)/1.2 V?  If so, we are good to go.  Does anyone have the background to comment on shell strength for various materials?


The big question: why has no-one done it before?  Surely that can’t be the case?  Does anyone out there know of any prior art?  Has the idea been proposed, shown to be stupid, and discarded?  (If so, it should at least be mentioned in the Wikipedia article.)  If not — is it stupid anyway, for a reason that I’ve missed?


25 responses to “How to get into space cheaply

  1. Interesting, but not at all the same thing. The point here is that progressive evacuation of the rigid shells allows us to maintain the same amount of lift as altitude varies, and therefore to reach very high altitudes where the atmosphere is very sparse.

  2. Problem is that altitude is only 10% of the energy to be in orbit. And even with the PERB, you’re still only getting 100 miles up (which is only half way) so you’ve only saved about 5% of the launch energy.

    And you need a tether, and you need to carry the rocket, and all-in-all, I don’t think you save enough. I haven’t got the time to do the sums now, but it doesn’t seem enough.

    Back-of-the-envelope: Orbital speed is about 8000 m/s, so KE is 32*10^6 J/kg. PE at 320 km is 32*10^5 J/kg.

  3. Do you need to generate that KE? Surely you start with nearly all of it by virtual of the fact that you’re rotating with the ground?

  4. Only at most at about 1500 km/hr. You need to get to 8 km/s, otherwise we’d nearly be in orbit simply by standing on the ground.

    V at ground level rotating with the Earth is about 4*10^7/(24 hrs) which is only 460 m/s. You still need another 7500 m/s to be in orbit.

  5. I don’t understand why you want to get up so high before launching your rocket. Surely if we can get to say 100 km (the Kármán line) then we’ve done plenty enough to hugely reduce the cost of a conventional rocket from there upwards? That’s only 200 pi = 628km additional distance to travel every 24h, which is 7.3 m/s.

  6. I know your figure of a tenth of an atmosphere was quasi-randomly pulled out of the air (heh), but a tenth of an atmosphere is only about 50,000-60,000 feet up, in the range of high-flying jet aircraft. Existing high-altitude balloons can go significantly higher than that.

    More generally, here is why your idea isn’t very good: you are effectively trying to make a balloon that saves weight by having no helium inside, but helium is already very light. The problem with helium balloons is not that helium weighs too much.

    More quantitatively: let’s assume for simplicity that air = diatomic nitrogen with molecular weight 28. Helium has molecular weight 4. By definition, when a helium balloon is at its maximum altitude, the total weight of the balloon equals the weight of the displaced air. One-seventh of the weight of the balloon is helium, and the other six-sevenths of the weight are the payload and skin of the balloon. For a high-altitude balloon, perhaps the skin is mylar a couple of mils thick.

    If you can have a rigid balloon with nothing but vacuum inside, it will be 14% lighter. I think the exponential scale height of the atmosphere is about 10 km, so that ought to take the balloon another 1500 meters or five thousand feet higher. But this assumes that you can have a rigid impermeable skin that is no heavier than mylar a couple of thousands of an inch thick.

  7. I don’t understand what you’re trying to say.

    To get to orbit you need vastly more energy than you get by drifting up on a balloon even to 100 kms.

    > Surely if we can get to say 100 km then we’ve done plenty
    > enough to hugely reduce the cost of a conventional rocket
    > from there upwards?

    No, you haven’t. Height isn’t the problem. Speed is the problem, and you have none.

    > That’s only 200 pi = 628km additional distance to travel
    > every 24h, which is 7.3 m/s.

    But you’re up high, in a balloon, not at 100km in orbit. You have effectively no rotational speed. Rotationally you’re only doing about 460 m/s. In LEO you need about 8000 m/s.

    Yes, you’ve saved yourself the effort of punching through some of the atmosphere, but that’s peanuts compared with the speed you have to get to to be in LEO. That’s why rockets launches are as much down range as altitudinal during the first phases of their trajectories.

  8. “here is why your idea isn’t very good: you are effectively trying to make a balloon that saves weight by having no helium inside, but helium is already very light. The problem with helium balloons is not that helium weighs too much […] If you can have a rigid balloon with nothing but vacuum inside, it will be 14% lighter. I think the exponential scale height of the atmosphere is about 10 km, so that ought to take the balloon another 1500 meters or five thousand feet higher.”

    Hmm. I have to admit that’s pretty darned convincing.

    Though vacuum is much cheaper than helium :-)

    Then perhaps the real question is why helium balloons aren’t more useful as a launch platform than they seem to be (i.e. not very)

  9. Chris Purcell

    A space elevator needs to be at the equator because it is actually in orbit, and just happens to touch the ground at one end. The only orbits which stay above a fixed point on the ground are, natch, equatorial. (All from recollection, IANARS, etc.)

  10. “A space elevator needs to be at the equator because it is actually in orbit.”

    Oh, of course. How sensationally stupid of me. Now I want to go back and edit the article to remove the evidence :-) (I won’t, though.)

    OK, so that won’t be an issue for the PERB.

  11. Colin is right too, with the following caveat. Punching through the atmosphere is a relatively minor part of the problem for the space launch vehicles we actually use. But, if there were no atmosphere, then schoolchildren could build multistage rockets to launch some sufficiently small payload into orbit, a ping pong ball or the head of a pin or something. The reason they can’t do that is that atmospheric drag does become a relatively more significant problem as you scale the rocket down.

    So, if you had the plan to use small multistage rockets to launch teeny-tiny payloads, then it might make good sense to launch from an aircraft or balloon.

    The Pegasus rocket, which carries smallish (though not teeny-tiny) payloads into space, actually *is* launched from an aircraft… but this is not so much to avoid atmospheric drag, as to avoid launch delays due to weather, and to skirt some of the bureaucracy involved in launching from the ground!

  12. Wouldn´t the tossing about (from the wind, even at 10% atmosphere, look at the camera footage of those helium balloons) of the balloon be the most problematic aspect of launching from it? By the time the rig gets to the desired height it might be pointing at nearly any direction and the launching platform to be completely chaotic in motion. Making the rocket also steer afterwards might nullify any benefit you get with this.

    Suppose you get to an altitude where a rocket might conceivably be launched into orbit through an orbital assist (by first moving further out of the atmosphere and then falling back, missing the Earth and then somehow climb to a stable orbit), on a randomly oriented and randomly moving platform getting it to go in the right direction will be very difficult. Not sure how much gyroscopes could help with this.

  13. Despite it not seeming to help launch rockets in the usual sense, I just like the idea.

    So, three orthogonal circular hoops. Or maybe a geodesic dome. And a skin that bulges in at the gaps between the struts. How do the tensile strength and area of the skin, and the compression strength and length of the struts, have to scale with the volume of the balloon? This is a capital-O problem but I’ll leave out al the O(…) notation as I write.

    There’s some optimal 2D catenary curve shape–how much you let the skin bulge inward in proportion to the linear dimension of the PERB. It’s constant with a given frame geometry and pressure difference…so with my capital-O license I ignore it.

    The area is V^(2/3). The length of struts and linear dimension of the skin pieces is V^(1/3). The force on a skin piece is like its area, but it’s distributed over the circumference, so V^(2/3) / V^(1/3) = V^(1/3) = the tension of the skin. The skin thickness has to be proportional to that. So the area of a skin piece times its density is V^(2/3) * V^(1/3) = V. The mass of the skin, for a given material, is proportional to the area of the PERB.

    (Hey wait, is that the holographic principle in reverse?)

    For the struts, the force on them is the force of the skin pieces (V^(2/3)) and this sets their thickness. Times their length (V^(1/3)) is V again. The mass of the struts, for a given material, is proportional to the area of the PERB.

    So, there’s neither a scaling advantage or disadvantage except that as the PERB gets bigger the payload and mechanism get proportionally smaller…but only to a limit where the total mass is that of the PERB itself. And the strength of the materials determines the pressure difference… and your final altitude is O(log(pressure difference)).

    For a helium balloon, the strength of the skin at the top has to hold up the weight of at least the skin at the bottom… it’s a different calculation but I’ll bet the mass of the skin at the limit of no payload is O(V) again. But… the constant is different…right?

  14. Aargh! I meant the mass of the skin, and the mass of the struts, is proportional to the volume! So that’s unlike the holographic principle, duh.

  15. Iain M. Banks mentions “vacuum balloons” in various Culture novels. Over the years I’ve (repeatedly, because I’m dumb and forget it from book to book) done the calculation that Richard Mason did above, to show that a vacuum balloon would only be 14% lighter than a helium balloon. That 14% is just not that huge a win — and I’m pretty sure that it would be swamped by whatever is necessary to make a rigid body that would withstand the inward pressure. (The big win with helium is that your gas envelope can be basically a bag — no rigid structure at all.)

    The delta-V argument (that most of the kinetic energy in a rocket launch goes into orbital [parrallel to the ground, not vertical] motion) is well-illustrated by — it went more or less straight up, above the 100km line, and then came more or less straight back down. Getting into an actual orbit is totally different and takes a lot more energy.

  16. The thought puzzle of getting into orbit has always been one of my favorites and I’ve thought of the baloon method too (although the PERB idea is definitely a new clever twist on the idea).

    But like has been mentioned, its the horizontal velocity that matters more than the vertical velocity. You only get vertical with a baloon.

    Think about what orbit really is. Its an object falling back towards the body because of gravity, but because its also moving horizontally by the time it falls its already moved far enough horizontally that its no closer to the body than it was when it started falling.

    Its something like walking. You’re always falling, with each step you fall forward slightly and then catch yourself from falling, over and over.

    The important part of walking is not that you can stand up, which can be possible even for a partially paralyzed person, but moving forward is the hard part.

    Of course this happens linearly and not in steps but the principle is similar. Draw a circle with a tangent line and then compare the ratio of the distance from a point on that line to the circle with the distance of that point to the tangent point.. its going to be high.. you should be able to intuitively visualize this.

    I think the real issue with getting into orbit is a matter of fuel. Its just hard to contain enough energy in a vehicle in order to propel it to the required velocity. This leaves 3 options: find fuels with a higher energy content, remove the fuel from the vehicle, or make the vehicle more efficient (require less fuel).

    The space elevator is a combination of the latter two.. efficiency (but very very slow, it would take days or weeks to ascend) and remote energy (electricity sent up the line or solar maybe). Space ship 1 & 2 is an attempt at option 1 and 3, efficiency and better fuels.

    Personally, I’m for nuclear rockets, which is solely option 1. We can do this now if it weren’t for the fear of nuclear power. We could pretty easily build a single stage to orbit rocket AND do a powered landing which eliminates all the issues with reentry heating. Maybe we’ll be able to make controlled fusion rocket engines eventually that would be safer…

  17. I have a thought on the tether… The linear increase in the weight of the tether would mean that in order to maintain your upward velocity, you’d need a matching decrease of pressure within the PERB. Also, there’s a phenomenon known as “voltage drop” which is caused by the resistance in electrical conductors, whereby, the longer your wire, the more power you lose. To overcome this, you use thicker (and heavier) wires. Perhaps, having your pump powered by the sun would offer a solution to the power problem. Of course, this would mean the balloon could only ascend when the sun is visible, but the mass could be greatly reduced.

  18. Solar power is a neat idea: needless to say, it becomes more effective as you rise higher and get more direct sunlight unfiltered by atmospheric effects. But sadly the whole idea is flawed, as other comments have pointed out.

  19. Interesting proposal, and even more interesting discussion here :-)
    Btw, since the magnitude of the force (i.e. lift) is proportional to the weight of the fluid that would otherwise occupy the column, i.e. the displaced fluid,
    how far up would such a vacuum-tank raise until it reaches equilibrium, i.e. the weight of the fixed volume empty baloon + vacuum pump is equal to the mass of the very thin air it displaces?
    Anybody who can do the math?

  20. Pingback: Answering 25 tough interview questions, part 2 | The Reinvigorated Programmer

  21. 1- With enough altitude, releasing an aerodynamic shape to gravitational gliding will develop speed horizontally.
    2- Glide fast enough to ignite a ramjet (or scramjet or better- something with no moving parts & light weight), and accelerate to what horizontal speed?
    3- Add rocket(s) to finish atmospheric escape if necessary. You’ll need them to maneuver in the airless void anyway…

    And so it goes…..

  22. Gabriel Theodoro

    How is the autor of this? How can I contact him/her?

  23. I am the author. By commenting, you have contacted me. Say on.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.