Rockets launched to reach orbit usually start from a stationary position, engines are started and the rocket slowly gain acceleration upwards, eventually reaching orbit.

What I'm thinking is that the rocket "wastes" a few seconds reaching a speed/acceleration that could be gained by other means. That would save a little bit of fuel, but most importantly, it would mean the rocket leaves altitude zero at already some speed, making it easier/faster/cheaper to reach the final orbital speed it needs.

I'm thinking for example of a hole in the ground, a few hundred meters deep. You start the engines, then accelerate the rocket up until it reaches a speed of say 40km/h when at ground level, from where the rocket just continues with acceleration but already at an advantage. 40km/h doesn't sound like much, but by looking at rocket launch videos, they do take their time to reach that.

I understand it would be expensive to build such a mechanism, but it's a one time thing. I also understand that the calculations would be different, but again it would be a "somewhat" one time thing: once you figure it out, it's way easier to redo.

Am I subestimating these costs and complexities? If not, what would be the reason not to do something like that?

Note: this is cross-posted from Physics because I can't close the same post there, but it is about to be closed by members. Check the comments there.

  • $\begingroup$ "I'm thinking for example of a hole in the ground, a few hundred meters deep.". How is that a gain? Now you need even more fuel because you start lower. $\endgroup$
    – Polygnome
    Commented May 18, 2020 at 20:11
  • 1
    $\begingroup$ @Polygnome I think the benefit is that you don't necessarily need to spend rocket fuel to get that initial velocity. It could be something like mag-lev system pointed upward. I'm not suggesting that that's the best option or that this is a generally great idea, however. $\endgroup$
    – aranedain
    Commented May 18, 2020 at 20:16
  • $\begingroup$ What is the use of initial 40 km/h when you need about 28,800 km/h for orbit? $\endgroup$
    – Uwe
    Commented May 18, 2020 at 20:29
  • $\begingroup$ aranedain is correct, that is what I'm trying to ask in my question. It's not about the hole, it's about using it to catapult the rocket up. Hence the title, "launch in movement". $\endgroup$
    – msb
    Commented May 18, 2020 at 22:06

2 Answers 2


There have been a number of different QAs on this site over the years discussing various methods of giving a launcher some initial advantage:

In all these cases, the advantage that can be gained through these schemes will be relatively small, for one reason or another, mainly due to the dense atmosphere at low altitude and the extremely high horizontal speeds needed for orbit. In exchange, a large amount of complexity is added to the system.

As mentioned in a few of the QAs I linked, space historian Henry Spencer sums it up this way:

Many novel launch schemes need some amount of help from rockets. What kills a lot of them is doing a tradeoff study of just enlarging the rocket part and getting rid of the non-rocket part. Surprisingly often, that works out to be better and cheaper.

In other words, rather than dig a big hole and set up a gas-piston launch system to provide 3% of your total velocity to orbit, it's generally easier to just make the rocket 3% more powerful.

  • 1
    $\begingroup$ This is a very useful answer and link compilation! I wish we had a way to make these types of efforts more prominent and accessible so that they would be easier to find and point to. In meta we have a community policy repository and the main meta has a real metaFAQ. I wonder if we could have a spaceFAQ as well. I'll see if I can write up a meta question about it. Somewhere I've made a list of questions about magnetic field on Mars for terraforming... $\endgroup$
    – uhoh
    Commented Apr 10, 2021 at 23:49

Launching a rocket from a hole will make the problem harder, not easier. You'll have the exact same acceleration/velocity profile, except your altitude will be a few hundred meters below where it would have been had you just launched from the surface. You've only succeeded in making your trip to orbit longer, which will require more fuel.

Where this might help is if you dug a deep hole, and launched the rocket from a catapult silo, only firing the engine after the rocket has left the catapult. This would help somewhat, since you are getting extra velocity from an external source, which means you do not have to carry or expend on-board fuel to get the initial acceleration. The problem is that this is a significant engineering challenge, as you now need a massive, powerful catapult in addition to the massive and powerful rocket you already have (which you barely got funding for in the first place). A rocket needs to be traveling at tens of thousands of miles per hour to reach orbit, so gaining even a couple of hundred of mph isn't really worth the extra complexity and cost of a system like this. In a future with very frequent rocket launches, a reusable system like this might make more sense, but it's simply not practical today.

I'll also mention the possibility of gaining horizontal momentum, which is actually the larger requirement of getting to orbit. It's not that you need to go up very fast, it's that you need to be moving sideways fast enough that the earth's curvature "falls away" at least as fast as you're falling toward it. One possibility for this is a rocket sled launch, where a spacecraft is accelerated horizontally over a long distance, and thrown off a ramp toward orbit. Again, this is a massive engineering challenge, and it actually quite wasteful due to air friction. Atmospheric drag is higher nearer to sea level, so you want to spend as little time as possible in the lower parts of the atmosphere. This is why rockets start their ascent moving vertically, in order to get out of the high-drag atmosphere region quickly, and only then turn parallel to the earth and spend most of their energy accelerating horizontally. A rocket sled ramp will be on the ground where drag is highest, so it's particularly inefficient to accelerate to high speeds when near the ground.


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