Stupid question obviously. But did you ever had an idea which sounded so brilliant, but you know it is totally stupid? So, lets hear my idea:

Do you know how we launch jet fighters from navy ships? If not, have look at this video for an example. In nutshell:

  • attach to catapult
  • Jet fighter goes to full throttle
  • launch the jet fighter

The plus sides of this method obviously is, that we save amount of space needed for acceleration. But we must, we simply must also save some fuel in jet fighter, don’t we?

The stupid idea:

Imagine same device attached to a space rocket horizontally. The launch sequence would be:

  • attach rocket to catapult
  • rocket blasts off its engines
  • catapult helps rocket gain initial speed
  • To the space we go

So, I am not saying to build catapult good enough to "fire" rocket to the space. I am reasoning to save rocket's fuel, which means we can put less fuel to the rocket and replace the fuel by (usable) cargo.

Because we are not using such approach, I:

  1. Am total genius and deserve Nobel prize in year 2015
  2. Have something terribly wrong in my thought process

And I am so obviously stuck with this stupid idea (catapults, catapults everywhere) that I cannot find anything wrong with the idea. Can someone of you please kindly explain me, where did I go wrong?

Legal disclaimer: If I am not wrong with the idea, I claim all the copyright for such glorious way to get to the space and would like to thank my parents. Getting Nobel prize for Space Exploration (introduced just because of me). It would not be possible without them.

  • 3
    $\begingroup$ Related questions Can gunpowder get you to the moon? & Can magnets be used to launch spacecraft? $\endgroup$ Commented Sep 29, 2014 at 10:42
  • 4
    $\begingroup$ Non-trivial question even if the accepted answer punches holes in it. p.s. The answer fills in considerations a newb such as I take for granted such as the one about delayed acceleration in the upper atmosphere $\endgroup$
    – Everyone
    Commented Sep 29, 2014 at 16:23
  • 3
    $\begingroup$ While I agree with the answers, I find the sardonicism unfair. It may seem stupid, but if people with more expertise considered it seriously in the past, why not someone else? The space catapult idea, and the space elevator amongst others, have been examined enough amongst experts to be considered clearly non-viable, but are common ideas amongst amateurs. It's like someone is being made fun of for not understanding why packing a tube with explosives is the optimal way to go to space, when they're only familiar with gunpowder. See @Everyone's comment above as well. $\endgroup$
    – mechalynx
    Commented Sep 30, 2014 at 2:47
  • 1
    $\begingroup$ This plan has been considered and seems to be feasible on the moon: with no air resistance and much lower gravity, a large railgun constructed on the equator could easily be used to send craft towards Earth. $\endgroup$
    – pjc50
    Commented Sep 30, 2014 at 12:43
  • 1
    $\begingroup$ OP's comment on this: I have rough idea why the rockets are build. And I KNOW that if catapult would be any better than current approach, we would use it. But I totally got so deep into that idea that I thought the best way will be to ask it out loud, even when it will sound dumb. $\endgroup$ Commented Sep 30, 2014 at 13:30

10 Answers 10


I'm leaning toward option 2, that you "have something terribly wrong in [your] thought process". Here's at least some arguments why; I by no means claim this to be an exhaustive list:

Mach 3 (which is pretty fast for a fighter aircraft, toward the upper end of the currently attainable range) is right about 1 km/s (Google calls it 1020.87 m/s), and you don't need to go that fast originally, just fast enough for sustained flight so probably closer to Mach 0.5 at most as you leave the runway. Earth orbital speed is about 7 km/s (for comparison, translunar is 11 km/s).

The weight of an aircraft is trivial compared to practically any spacecraft. Work to accelerate grows exponentially with target speed and object mass -- either of which makes this a serious problem.

Acceleration kills; go much above 10G (which is what, 100 m/s^2?) and a human becomes a wet spot on the rear wall. The acceleration during a spacecraft launch and climb process is nontrivial, yet much easier to manage because it can be kept to a safe level while climbing out of the worst depths of Earth's gravity well. No need to jump to anywhere near even orbital velocity right off the launchpad.

What if after the vessel is catapulted to a considerable speed the main engine fails to light up? If a current-design spaceship main engine fails to light up on the launchpad, that's a serious annoyance and probably grounds for some grim looks toward a number of technicians, but nothing really worse than that. If you are already airborne and moving at a considerable speed, in a fully fueled spacecraft (keep in mind that rocket fuel is some really nasty stuff), and the main engine doesn't kick in on schedule for any reason whatsoever, then what do you do? Depending on the (ballistic) flightpath and spacecraft maneuverability, even ignition a few seconds late could spell some very serious trouble downhill.

The atmosphere is thickest close to ground, exactly where with your scheme you would be seeing a very high rate of acceleration. With more common rocket designs, acceleration close to ground is relatively trivial, and they only really start picking up speed some distance above ground. (Close to the ground, you are more concerned with getting off the ground than you are with building up speed. Even if you climb slowly for the first 10 km, you still have plenty of time to build up speed before you reach orbital altitudes.)

It's not like the idea hasn't been proposed. Jules Verne did something quite close to what you are proposing in 1865 in De la terre à la lune, beating you to it by about 150 years. It was seriously considered in 1903 and concluded that it wouldn't work even from a technological standpoint, although perhaps with modern materials it could be slightly more practical. That still doesn't solve the issue of the acceleration, however, which was touched on even in fiction works in the 1950s at least.

Bottom line: since it has been proposed and found a non-viable alternative, I would suggest that you do the math and tell us why it would work, addressing the concerns raised in this as well as the (currently only) other answer.

  • 3
    $\begingroup$ Terrible idea connected with laziness to do the math :) Ever happened to you that you had an idea knowing that such idea is bad, but because you are so deep in the idea itself, you cannot see why? :) Thanks a lot for thoughtful answer. $\endgroup$ Commented Sep 29, 2014 at 10:40
  • 1
    $\begingroup$ Most of these points address the problems with an actual catapult trying to throw something up into orbit. But like the F-18 starts its Engines while still on deck, the rocket on the catapult could also start its engine before liftoff. Imagine the Starting-POD was a giant elevator going really fast upwards, while the rockets starts its engines. Wouldn't the rocket get additional energy from the elevator/catapult and thus need less fuel? $\endgroup$
    – Falco
    Commented Sep 30, 2014 at 9:45
  • 1
    $\begingroup$ About the point about atmosphere being thick closest to the ground. We can have a look at drag equation. Trying to go at high velocities through a thick fluid is wasting fuel. $\endgroup$ Commented May 26, 2016 at 11:10
  • 2
    $\begingroup$ @BenCrowell: Answer assumes carrying fuel, for which the exponential term of accelerating your fuel dominates for chemical fuels to orbit. $\endgroup$
    – Joshua
    Commented Dec 12, 2016 at 21:55
  • 2
    $\begingroup$ While this is broadly correct, the line about humans "becoming a wet spot after 10 g's" is catchy but definitely wrong. It would probably black out the pilot and passengers (definitely A Bad Thing), but it's almost impossible to 'squish' a human into a puddle with acceleration alone. The wikipedia article on g-force has some quick reading on human limits, which depend heavily on orientation and duration. 10 g won't necessarily even black you out in short durations - the human record in formal testing is 46.2 g on a rocket sled. $\endgroup$
    – brichins
    Commented Jan 15, 2018 at 22:39

Mass of carrier-based F-18: 40,000kg Mass of Delta-II rocket: 230,000kg Mass of the Space Shuttle stack: 2,000,000kg

Even if we can knock off the fuel needed to get it moving we still have to have enough to accelerate the thing to 7km/s, and that's most of the fuel. You are going to need a stronger structure to handle the catapult forces (which adds weight which requires more force which needs a stronger structure.....) and a mighty big cable to apply that much force. I leave the math as an exercise for the reader.

Rockets can simply apply smooth, steady thrust until they are going fast enough, and some of the earlier rockets like the Centaur had pressurized bodies - the frame was only about 1-2 mm thick and could not hold itself up without internal pressure. No way you could hook it up to a launcher.

And to totally invalidate your idea, it's already been done in fiction. Part II of 2001: A Space Odyssey opens with a Pan-Am spaceplane being track-launched before the rockets lit up. Arthur C. Clarke beat you to it by 45 years.


You are not the first to think of this, of course. The idea of an electromagnetic catapult to push a payload to orbital speeds has been floated before... a couple of plans which spring to mind are the Launch Loop and StarTram.



The key issue is the sheer size of the whole assembly, which is generally measured in hundreds or thousands of kilometres with the launch end of the system needing to be many thousands of metres up.

The 'Gen 1.5' version of the StarTram sounds closest to your proposition, as it only provides 4km/s of accelleration to the payload, with the rest provided by onboard propulsion systems. The peak accelleration for a shorter version of the system (50-80km) could be under 15g, and so potentially surviveable for fit humans (if not particularly pleasant).


I'll try a little bit of Fermi estimation.

Mass Limit

Similar to rockets, there is an exponential equation for tethers. The Wikipedia article on it is actually quite solid in terms of the gritty technical details. Easily enough, we can plug in some velocity at the point of release and use the material characteristic velocity for M5 fiber. As far as I can tell, this is the most reasonably optimistic out of the practically available materials. Then the ratio of the mass of the tether to the mass of the rocket will be given by:

$$ \frac M m = \sqrt { \pi } V_r \mathrm{e}^{ {V_r}^2 } \mathrm{erf} ( {V_r} ) $$

Keeping with the spirit of Fermi estimation, let's image a Falcon 9 with a launchpad mass of 500,000 kg, which puts 10,000 kg into orbit. If we use an engineering factor of 3, then we can easily calculate the velocity at which the catapult assist system will weigh as much as the rocket on the launchpad. This is in the neighborhood of 3 km/s.

It's important to not go much further than this mass limitation. No company does more than a few launches per year. If per-mass the tether costs a similar order of magnitude as the rocket, then the mass ratio is the number of launches you'll need to amortize the cost. If the tether material is 10% the cost of the entire launch assist system, and you need to pay off investors within one year, then we're clearly settling on (mass of rocket) = (mass of tether). It doesn't matter much if you change your financial risk tolerance here, because the mass equation is so insanely exponential.

Needless to say, this could still work. If a rocket reduced its Delta V budget by 3 km/s, then it could practically double or triple its payload delivered to orbit. You would be tempted to think this makes sense. Actually, in space, it does make sense. The problem is that we're still on Earth.

Length Limit

Strange enough, the above mass ratio equation says nothing about the tether length. It could be long or it could be short. If the edge velocity is the same, it doesn't change the total tether mass. If you go faster, the centripetal acceleration is higher so the tether must be thicker - and this perfectly balances the effect of a shorter tether.

Thus, g-forces actually limit the length of the tether, and these are constrained by biology. Let's say that our limit is 3g. The equation is simply $3g=\frac{v^2}{r}$. If you plug in 3 km/s, you get a tether radius of 300 km. Oh my. This will not do.

The most radical version of the tether launch assist system consists of either high altitude balloons or aircraft. These are limited in altitude due to atmosphere density, typically to about 10 km in the case of large jets. Combining the limitation of 3g and 10 km, you find that the launch assist system can only aid to the tune of 0.5 km/s. So with this system, you can only improve the rocket's mass fraction by about 20%, but only at the cost of reducing the starting mass by the requirement that a jet or balloon can pick it up in the first place. More than likely, it hurts more than it helps!

Ways Around Length Limit?

Your idea was to reduce the propellant the rocket needs. You can see from my calculations that the limitation isn't the bulk mass of the tether. The problem is that the length need is so high that you don't have anywhere to attach it to. But your idea of attaching it is:

Imagine same device attached to a space rocket horizontally.

Sounds like the tether connects the rocket to an anchor (I assume a moving anchor) on the ground. But the problem is that the rocket must pull the tether up with it here. That increases the amount of propellant the rocket needs, which is already violating your objective.

In fact, this is why any serious proposals in this family involve launching hardened commodity payloads, not people, and not conventional rockets. By increasing the acceleration, you can keep the structure to within roughly a kilometer scale. But such a megastructure is a hard sell when it will only marginally increase the payload in the first place.

Alternatively, you could try a scheme to "attach" it to something not bound by terrestrial constraints. For instance, something already in orbit. This is the approach of various partial space elevators. These can reasonably be >300 km in radius, but then the mass constraint pops back up again since you must put it in orbit to begin with.


As others have said, the scale of both the construction and the forces involved place this project firmly in the realm of the impossible. However, that does not mean the idea hasn't been considered in a different form.

While not exactly a catapult, There has been a somewhat similar idea involving what's basically a particle accelerator for space cargo that can withstand the pressure. Imagine the Cern Large Hadron collider, but built into a mountain with an exit port aimed at a stable orbit entry point. The theory is that you get the craft up to speed on the ground floor and then launch it into orbit from there. While the G-forces are not viable for anything living, properly protected equipment that can withstand both the G-forces and the magnetic forces involved could be brought into outer space for a ludicrously low price compared to traditional rocket-aided launch methods (on the order of 10,000 times lower for the launch itself).


Also see the different sources on the Wikipedia articles.


While the answers are right lets look at it in a much simpler fashion: Nothing about a catapult reduces the acceleration the craft being tossed experiences. Watch a rocket launch--multiple gs for hundreds of miles as it flies off into space. If you want to keep the acceleration down to the same amount you'll need as much space--your catapult will have to be hundreds of miles long.



As rockets consume their fuel, they get lighter, so the thrust they produce is applied to a progressively reducing mass. The result is increasing acceleration and therefore g-force on the vehicle.

Two examples: Apollo Saturn and the Space Shuttle both limited their acceleration by shutting down or throttling back engines so as not to exceed about 3g. This is both for the benefit of the crew and because the structures couldn't take any more than that. The reason is weight.

Everything on a rocket has to be made as light as possible to minimize the fuel required to get it into space. That limits how strong things can be. Rocket structures are made only as strong as they need to be, hence they are as light as they can possibly be.

Now if you want to catapult the thing off a launch pad, you need to make things stronger to take all those extra g-forces (only needed for the duration of the catapult) and therefore everything gets heavier and therefore you need to carry more fuel to get the job done once you're past the end of the catapult stroke.

  • $\begingroup$ There were acceleration values above your limit of 3 g, see the ascent data of Apollo missions. $\endgroup$
    – Uwe
    Commented Dec 21, 2017 at 13:22

We do not use catapults because we don't want to destroy the payload before reaching the orbit. The immense acceleration necessary would destroy every payload build in the same way as for a rocket launch. If a very robust payload would survive the catapult, it would be destroyed by the immense heat load of orbital speed within the dense lower parts of the atmosphere.
A reentry is possible because the orbital speed is only within the very thin upper parts of the atmosphere.


I agree with Pavel, and also think this is probably a concept to look at: The catapult has to be engineered as a pre-burning rocket phase. Meaning, we could use the catapult to launch a rocket, capice? So, we may engineer the catapult to accelerate at the maximum rate a human body can stand (let's say, 5g), so that it reaches considerable velocity (say, 600km/h), way bellow the limit that would make things burn in the lower atmosphere.

In this visionary mind experiment, we could have a rocket ignition, where the rocket has initial velocity of 600km/h. The rocket would keep accelerating at 5g after the catapult line is over; using it's rockets (as currently). The nice thing is that the burning-rocket would start its trip at 600Km/h instead of 0km/h . I guess the main impact would be on the necessary amount of initial fuel (someone could estimate that, please?). You can cut of all the fuel necessary to reach 600km/h in a non-catapult launch. The whole ship would be lighter!

Considering our catapult works at acceleration = 5g = 5x10 m^2/s . Going from V_0 = zero , to V = 600km/h , demands 3.3seconds . Lets round it to 3.5s . Then, as next step we may think “who big should be the catapult line to allow this rocket launch?“ , And the answer is not bad at all.

Accelerating at a=5g , from X_0 = 0 and V_0 = 0

X = X_0 + V_0 . t + (a . t^2)/2 ; using t=3.5s


Let’s say we would need something bigger, since the rocket itself could be 100m high… We are talking about a 500m rock-catapult lined up to the sky. don’t think this sounds crazy to build. I don’t know how much would cost such catapult. But we have to take into account take the rocket would be much lighter , just because we need much less fuel to reach upper atmosphere .

Many mountains could meet the requirements for build this mind-catapult… (someone could try extra more calculations considering Everest type launch … be careful to keep velocities lower than low-atmosphere burn! )

About some drawbacks mentioned from previous comments...

Guys, a robe to stand 5g and catapult a light rocket is not a problem (is it?)... I mean, even if a robe to "do the job" is the biggest problem you can imagine, I think this could be engineered to work.... for example, we could look for currently robes-specifications for catapulting foo-fighters

About the "problem" of having rocket-motor failure during the catapultagge... guys, come on... If a rocket-engine explodes, it will be a tragic problem from both ground and catapult launch. If a rocket fail to burn during catapulting, I agree it will be hard to stop the rocket. Probably we may lose the rocket. But we already have technology to eject crew (or any valuable-payload) (see the Abort Pad tests from NASA or SpaceX)
So, "rocket failure" should not be seem as a main strong argument against catapulting rockets (my view)...

Also, about extra structures needed to stand the catapulting... You know... maybe, we would not need to go to far to address this kind of problem. Today's launchers already stand >5g acceleration ... So, concerning this point, the current rockets are fine. Also, lets think for a moment... Where would we connect the robe? Well, we can simply apply the force in the same mechanic structure that holds the currently burning rocket engine pull . I mean... the currently burning engines already apply force enough for a >5g acceleration. So, we may apply the robe-force in the same place.

Any extra structure needed for that can disconnect from the rocket at the catapult-end; no big extra-load have to be take up because of that.

We could start a dedicated forum just to bird-eyes on this topic, in a more-realistic point of view... looking for materials and specifications.

See you!

  • 4
    $\begingroup$ 600 km/h sounds nice, but it's only 2% of orbital speed. So your savings would be 2% of the total rocket mass, in exchange for replacing the launch tower with a structure 500 m tall and capable of accelerating several hundred tons to 600 km/h. $\endgroup$
    – Hobbes
    Commented May 29, 2015 at 14:49
  • $\begingroup$ what do you mean by "robe"? $\endgroup$ Commented Dec 19, 2017 at 13:37

You ever see rocket city rednecks? These redneck rocket scientists (Father and Son) from Huntsville Alabama (and former and current NASA employees) came up with this idea and made it work on a small scale by taking it up a notch by attaching the a catapult to a weather balloon. They launched the rocket from the catapult that was suspended from the balloon in the air. It worked to.

  • $\begingroup$ You may include references to enhance your answer. $\endgroup$
    – Manu H
    Commented Aug 16, 2015 at 9:35
  • $\begingroup$ Structural design considerations are a huge part of rocket body design. They are weak in side-to-side strength, but strong in vertical compressive forces strength. A rocket assembly that is fuelled and carrying a payload for travel in horizontal slopping to vertical flight would need to handle loads,shifting solid and liquid payloads and structural vibration harmonics in multiple different attitudes outside the current, almost 100%, vertical compressive stress requirements. nakka-rocketry.net/fusestru.html Ever wondered why shuttles were transported to pad standing @ 3mph? $\endgroup$
    – JohnP
    Commented May 18, 2018 at 14:51

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