In fact, Atmo is very hostile for classical rockets. It makes them too sizable, heavy & inefficient as well. To shift this situation needs to get another propellant, but we have fallen short here. So, one of the available ways is to deal w/ air-breathing engines (turbojets or turbofans). Yes, very disadvantage gets to be low thrust-to-weight ratio == an air-breathing engine never beats the rocket ones.. Never. However, there is possible to use 2.5-staged design of rocket.

  1. rocket boosters to initiate sustainable aerodynamic lift: 100-150m/s.
  2. winged core with turbojet engine == it takes rocket up to 30km altitude & up to 800 m per second velocity (yes, for the most share of flight profiles, there will be more humble speeds & altitudes).
  3. upper stage, attached to the back of the core, makes the rest of its way to an orbit.
  4. core gets its way back to the runway.

rocket boosters & core stage run from the start, it's very reason why to call it 2.5-staged design (not 3-staged one).

Horizontal take-off is preferable to avoid gravity losses.

Design of wings is most tricky part in this scheme == one of possible solution looks to use partially dropping ones.

  • $\begingroup$ related question, but not fully the same: actually, i'd like consider air-breathing engines for mid & heavy-lift rockets too. $\endgroup$
    – Z0dCHiY8
    Aug 18, 2019 at 2:42
  • $\begingroup$ Horizontal take off does not avoid gravity losses. You need some power only to hold height. $\endgroup$
    – Uwe
    Aug 18, 2019 at 9:43
  • $\begingroup$ Is it useful to use three stages to orbit including an air breathing stage when two stages will do? $\endgroup$
    – Uwe
    Aug 18, 2019 at 9:47
  • $\begingroup$ @Uwe not sure about your words for gravity losses. Gravity losses take place when vehicle needs TWR > 1 to take off, otherwise we deal w/ aerodynamic drag losses. about "two stages": three stages consume less fuel anyway + air-breathing stage takes on board only kerosene, so flight profile becomes more flexible + that stage can be reusable many times.. much many than rocket one. $\endgroup$
    – Z0dCHiY8
    Aug 18, 2019 at 20:08

1 Answer 1


This looks pretty close to Stratolaunch which has a claimed payload of a couple of hundred tons vs the several thousand kerb weight for space shuttle or Saturn V.

So a heavy lift would need to scale the already massive Stratolaunch by factor of 10-20 which is probably not impossible but certainly not trivial, and would certainly only be able to fly from a dry lake bed or other similar massive surface.

Horizontal launch also a whole collection of complications for the vehicle itself, structurally, for fueling and since it basically takes off with the something rivaling a nuclear payload onboard in terms of damage capability has some exciting safety questions. See the answers on the relevant tag.

Safety is important here since for a massive jet powered lift stage to be sensible it has to lift a lot of flights per year to earn the design and build costs back. If a weekly flight regime means something like this falling out of the sky every year it gets harder to get approvals, even if the $ cost is better.

Looking at Stratolaunch as an example a viable air launcher might be more like a 200 tonne small dumb rocket akin to an overgrown pegasus that is intended to lift 2-5 tons to orbit on a daily tempo taking advantage of the wider inclination flexibility of air launch to hopefully produce scale and mass production advantages against a BFR style 100 tons at a time on an small number of expensive but reusable systems.

Electron's work in recovering 'dumbish' first stages would also impact the maths of this sort of thing.

  • $\begingroup$ Stratolaunch is good example, but i'm not agreed about scaling == upper stage is about 150-300t. Actually, weight savings for upper stage relates to altitude and speed of 1st one. Air-breathing engine is quite capable to reach about 3000kmph & 20-30km of altitude. But very problem is wings == they need to have large area at subsonic speed & relatively small for supersonics. Would you see any solutions? Thanks. $\endgroup$
    – Z0dCHiY8
    Aug 18, 2019 at 20:23
  • $\begingroup$ Supersonic external carriage gets complicated quickly en.wikipedia.org/wiki/Lockheed_D-21 and the SR71 barely qualifies as re-usable given maintenance support required. And still only gets you an extra 800ms of the needed 9kms of DV for orbit. So not impossible but even harder than stratolaunch to make cost effective. If you want to look at air launch, the game Kerbal Space Program simulates of the complications enough to meaningfully explore the problem. $\endgroup$ Aug 19, 2019 at 9:49
  • $\begingroup$ 1kmps is too high, i more inclined to range 600-800mps. $\endgroup$
    – Z0dCHiY8
    Aug 20, 2019 at 0:05

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