Outside Earths atmosphere is the vacuum of space and its immense power. If a cylindrical tube of sufficient strength and size were built tall enough to reach the edge of atmosphere would it not be possible to have a vehicle inside this tube with an air lock and means to open hatch or door to allow the vacuum of space to suck the vehicle up and away? If the angle were correct and certain adjustments to tube were done orbit could be obtained. If my theory is correct.

  • 6
    $\begingroup$ Vacuum does not suck by itself. It is the air pressure on the other side that propels stuff. There is nothing like that in your evacuated tube. $\endgroup$
    – jkavalik
    Commented Mar 3, 2017 at 11:41
  • 4
    $\begingroup$ It's not vacuum pulling, it's our atmosphere pushing when you see implosions like this one. Otherwise our atmosphere would just be sucked into space. So a vacuum does not have "power" itself. $\endgroup$
    – DarkDust
    Commented Mar 3, 2017 at 11:52
  • 2
    $\begingroup$ @jkavalik I think the idea is to open a door at the bottom of the evacuated tube and let the whoosh of the atmosphere push the payload. The question is poorly worded, and needs work. $\endgroup$
    – uhoh
    Commented Mar 3, 2017 at 12:55
  • 8
    $\begingroup$ To the downvoter: just because the idea displays a misunderstanding of vacuum doesn't mean it's an invalid question. You can't learn unless you ask, right? $\endgroup$
    – Bear
    Commented Mar 3, 2017 at 14:18
  • $\begingroup$ But we are not able to build towers so high to reach the edge of the atmosphere and the minimum height of an orbit. $\endgroup$
    – Uwe
    Commented Mar 3, 2017 at 23:07

7 Answers 7


It's important to remember that getting into orbit requires not just gaining altitude but achieving horizontal speed of about 7700 m/s, so your vehicle would need an upper stage for circularization.

At sea level air pressure against vacuum, a piston about 10 meters wide would develop the same force as the 9 Merlin engines of a Falcon 9 first stage. Unlike a rocket engine, a vacuum piston arrangement would have an absolute upper limit equal to the speed of sound, so the technique would only be usable for an initial launch boost.

The obvious problem is that we don't know how to build a structure anywhere big enough to take advantage of this. The world's tallest building is less than 1km in height; an evacuated launch tube that big could save maybe 25-30 seconds of propellant, which would a substantial win. You'd need a seal at the top that was strong enough to keep the air out but weak enough to let the rocket through without damaging it. Of course, you'd have to ignite your rocket engines from on top of that moving piston, and anything that went wrong along the way would be catastrophic.

  • 1
    $\begingroup$ You can also just release it in GEO, although that requires building up to GEO (tends of thousands of km) rather than LEO (hundreds). $\endgroup$
    – ikrase
    Commented Mar 19, 2022 at 6:25
  • $\begingroup$ Also, there seems to be a misconception about the vacuum in the tube... You'd have to evacuate this tube before every single launch which costs a lot of energy. $\endgroup$
    – TrySCE2AUX
    Commented Mar 22, 2022 at 6:57

The existing answer all note reasons why it doesn't work, but miss one.

The air inside the tube doesn't really behave different than the air outside the tube. Neither is sucked into space. That's because gravity pulls it down.

So if we had a vacuum tube, and we let air in on the bottom, it would quickly shoot up. And if there's a snug fitting payload, it would go up as well. But as the air molecules below the payload go up, they lose energy to gravity, and near 500 kilometer there's not much energy left.

On a practical note, air expanding into vacuum also cools down quickly. This will freeze out the water, so the bottom of your tube will soon be clogged by ice.


You would have to create a vacuum initially, but as the tube fills with air, the pressure inside becomes equal to the pressure outside, and the flow of air will stop.


Even if you could build a tube from ground level high enough to reach space, the bottom of the tube would not have a vacuum. The weight of the air in the tube is enough to hold it down against the pull from the vacuum above it (really from its own pressure pushing it up). How do we know this? The rest of the air in the atmosphere does not push itself out into space, and it has nothing holding it down but its own weight.


So there is the obvious problem of building such a tower (and having it not collapse or burst somewhere along the way). However your question says:

If a cylindrical tube of sufficient strength and size were built tall enough to reach the edge of atmosphere would it not be possible to have a vehicle inside this tube with an air lock and means to open hatch or door to allow the vacuum of space to suck the vehicle up and away? If the angle were correct and certain adjustments to tube were done orbit could be obtained.

So we're working off the premise that we have such a tower and those problems can be left for another question.

Now, just to check I've pictured this right I've drawn it out (in paint so it isn't great but hopefully assists a description). Space air cannon

So the ball is our ship with a seal around it so air cannot escape past the ship. It would have to be supported at the bottom before the air is added.

We then need to release the pressurised air. In principle yes, this pushes a payload upwards (assuming the tower is fit for coping with these pressures). If we don't use pressurised air then the ship won't get to the top, the heavier the ship the higher the pressure we need.

You would need to suck the air back out again afterwards though.

Just a little extra: I'm not sure of the details but it seems to me that, if a tower like this could be built, then using a mechanical lift would be much safer and use less energy.


Others have mentioned the speed limit of such a system: you can't get above the speed of sound (around 340 m/s), or 1/20 of orbital speed.

In addition, you need a really high tower. Even at 500 km altitude, there's still some atmosphere left, which would (slowly) fill your tube if you left the tube open at the top.

At 100 km altitude, air pressure is on the order of 10-7 bar, or 100 mg/m3. You should be able to seal the tube with a membrane and have your rocket pierce it when it reaches the end of the tube.

At 50 km altitude, air pressure is on the order of 10-3 bar, or 1 kg/m3. Making a cover that will hold against this pressure, yet be easy for the rocket to break through, is going to be difficult.

We have no materials strong enough to build a tube this high. The highest building in the world is only 1 km high, and that was easier to build than a structure that has to hold a vacuum. Buildings this high are limited by 2 things: the compressive strength of the building materials, and the compressive strength of the ground underneath. Build higher than a few km, and a cylindrical tower collapses under its own weight, so you have to build a pyramid-like structure to spread the load. You end up with a building that's not only 100 km high, but also 100 km wide.

Many of the challenges for this scheme are similar to those of a space gun, with the added problem of an even lower maximum exit speed.


To add a little bit to @RussellBorogove's definitive answer there may be disappointing speed limitations due fluid dynamical realities; friction and turbulent flow may severely limit the maximum speed that air can enter the tube. It seems mostly sub-sonic so it's possible basic fluid dynamics texts could be used to estimate this effect.

If someone can do this, please go for it and post separately!

There may also be some low frequency (sub-Hz) oscillations as air mass in different parts of this multi-kilometer long tube moving at different velocity start bumping into each other.

  • $\begingroup$ Sub-sonic ( about 0.3 km/s) speed is nothing compared with the necessary speed for an orbit ( about 8 km/s ). A factor of more than 25 is missing. $\endgroup$
    – Uwe
    Commented Mar 4, 2017 at 9:24
  • 1
    $\begingroup$ @Uwe your random can't! won't! doesn't! isn't! wrong! bad! comments are getting tiresome. Nothing wrong with an engineering exercise exploring something that ultimately won't work. Based on your comment I think you haven't carefully read the linked answer first, where the need for substantial lateral velocity was already nicely pointed out. I'm simply adding to the list of why this wouldn't work. If you aren't going to put in the time to read please refrain from the negative "you don't know what you are talking about" - type comments. Your positive contributions are so much more valued here! $\endgroup$
    – uhoh
    Commented Mar 4, 2017 at 12:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.