For the purpose of this question I would like to assume there is no risk of life for the astronauts, or that the design would only be used for delivery of a non-ethically-sensitive payload. Looking at most of the designs for reentry lander like Soyuz or other space capsules it seems like the vast majority, if not all of them, exclusively use parachutes. I'm going to assume this is because of various reasons: parachutes are tried and true, placing lots of fuel near live humans can be bad, fuel tends to explode when overheated, etc...

I would like to know if any payloads have used something like a turbofan engine or other form of air-breathing engine to slow their descent/land. Would it be impractical compared to other tech if the payload was heavy enough to require more than parachutes?

  • $\begingroup$ related: aviation.stackexchange.com/questions/37525/… $\endgroup$ – user3528438 Aug 21 '18 at 21:07
  • 5
    $\begingroup$ What exactly are the engines going to do? For vertical landings, you'd need engines to get a thrust to weight ration bigger than one, but that didn't get burnt off during reentry. That is going to add a lot of weight and mechanical complexity. What benefit do we get? $\endgroup$ – zeta-band Aug 21 '18 at 21:24
  • 2
    $\begingroup$ @zeta-band Air breathing engines plus wings can get you a go-around capability that the US space shuttle lacked, with far less than 1:1 TWR. $\endgroup$ – Russell Borogove Aug 22 '18 at 0:14
  • $\begingroup$ @zeta-band Another advantage is to extend the range of accessible landing sites for given entry circumstances, i.e. direction of flight upon entry, entry flight path angle, location (lat-lon) of onset of aerodynamic deceleration, etc. This is why the Soviet engineers designed Buran with turbojet engines. $\endgroup$ – Tom Spilker Aug 22 '18 at 0:29
  • 1
    $\begingroup$ NF-104A, for very tentative values of "reentry" - ie, zoom/rocket climb to altitudes where neither the air-breathing engine nor aerodynamic controls were effective (necessitating an RCS), then descend into thicker air and restart the engine. $\endgroup$ – Chris Stratton Aug 22 '18 at 1:21

As @OrganicMarble alluded to, the Buran Soviet shuttle was designed with turbojet engines (see here; and here, under "The engines") to extend the range of possible landing locations given the re-entry circumstances. Test versions had those engines (the same engines used in the Su-27 fighter) installed, but those were never launched into space. The version that actually launched into space didn't have those engines installed.

So although there was a spacecraft design that used turbojet engines, that specific design never flew in space. No other detailed design (i.e., more than simple architectural drawings) used a turbojet or turbofan engine.

  • 2
    $\begingroup$ Do you know of any further reading on how the air-intakes/jets were expected to behave under the conditions of hypersonic reentry? Presumably they would have only actually run once the Buran had transitioned to a typical flight regime, but they'd obviously be exposed before that. I can ask a new question if the answer is quite complex (as I imagine it is)! $\endgroup$ – Jack Aug 22 '18 at 13:47
  • 3
    $\begingroup$ The intakes and exhausts were to be covered until they were needed. (Energiya-Buran; the Soviet Space Shuttle (by Hendrickx and Vis), page 327) $\endgroup$ – Hobbes Aug 22 '18 at 15:43
  • $\begingroup$ When I was writing my answer above I was thinking that maybe a ramjet would be better than a turbojet or turbofan. But then, thinking about where airbreather engines would need to be to survive re-entry without a lot of mass spent on a thermal protection system, I decided that the air flow at the intakes would be very turbulent. Ramjets don't particularly like turbulent flow at their intakes! Also they only operate while supersonic, so while they could provide gross air-supported trajectory control, they couldn't provide final-stage control or go-around capability. $\endgroup$ – Tom Spilker Aug 22 '18 at 18:27

I don't believe any vehicles equipped with air-breathing engines have flown to space and returned. Some test vehicles for Buran had jet engines installed, but they did not fly to space. In this picture of a Buran test vehicle, you can see that the jet engine mounts interfere with the reaction control jet nozzles, showing that this configuration could not be used on an orbital mission. Edit: Tom Spilker's link shows that the two strap-on jets were to allow the test article to take off; the two engines on either side of the vertical stabilizer are jets that were planned for the orbital vehicle. However, on Buran's only flight, these jets weren't installed.

enter image description here

"Could they be?" Maybe. Early Shuttle designs (edit: and plans for Buran) featured jet engines to fly to the runway after re-entry and for ferry flight. By the time of the Phase B designs of Shuttle, the jet engines were mostly gone, and didn't make it into the final design. Here's an example from Dennis R. Jenkins, "Space Shuttle", 1992 edition showing engines in over-wing-mounted pods on an early Orbiter concept.

enter image description here

There were several reasons for dropping them, I would say that it boiled down to the fact that the cost, complexity, and weight of a separate engine system wasn't worth it for the short time it would be used during the mission.

Edit 2: Chrysler's SERV (Single-stage Earth-orbital Reusable Vehicle) contender for the Shuttle program Phase A had maybe the largest number of jet engines ever proposed for a spacecraft - 28!

enter image description here

  • 2
    $\begingroup$ Is that 12 aerospike engines?! See this is actually what I was envisioning in my mind. I'm guessing that this design never actually went into production though? I did something VERY similar to this for re-entry in (space-gods punish me for thinking this means it's something viable in real-life) Kerbal Space Program and it worked well. No parachutes, just atmospheric engines and a very small amount of liquid fuel, no oxidzier at all. Nevermind-- I see the 28 tiny little landing engines now. $\endgroup$ – Magic Octopus Urn Aug 23 '18 at 16:16
  • 1
    $\begingroup$ Yes, you are right, never produced, just proposed. This design got zero traction but apparently could not be ruled out as unworkable. I think it would have been amazing myself but the mechanical complexity of all those doors and engines....a target-rich environment for failures. $\endgroup$ – Organic Marble Aug 23 '18 at 16:22
  • 1
    $\begingroup$ That article you linked talks about literally everything I asked-- I need to learn to finish reading before I ask stupid things that are answered in sources that are linked. Seems neato though. Wouldn't be a bad idea for the worlds first unassisted reusable SSTO, or at least it seems like it wouldn't be too bad of an idea now that tech has improved. Seems at the time it was universally denied due to technical restraints... $\endgroup$ – Magic Octopus Urn Aug 23 '18 at 16:37

Though it hasn't been flown yet, British company Reaction Engines has designed a Single-Stage to Orbit reusable spaceplane with rocket engines that utilize atmospheric oxygen for a substantial portion of the ascent, before switching to internal LOX tanks once above ~85,000 ft. They have already designed and performed some limited tests on the engine technology, but have yet to build or test the overall vehicle.

So not a turbofan or turbojet, and not fully developed tech, but we might see this in the not-too-distant future.

  • $\begingroup$ But unfortunately the ~30,000 ft. or 10 km is not a substantial portion of the height of about 400 km for a low orbit. Going to an orbit requires not only the energy for the height but even more energy to acheive the necessary speed of about 8 km/s. If you try to accelerate to 8 km/s within the atmosphere it would be very hot just like a reentry from orbit requiring a heat shield and much energy to compensate the drag of the atmosphere. $\endgroup$ – Uwe Aug 22 '18 at 10:55
  • 2
    $\begingroup$ @Uwe Interestingly, the SABRE engine linked is designed to go to 20% height/20% speed of orbit before going closed cycle: ~28km @ Mach ~5.1. Being able to remove 1.8km/s of non-vacuum dV in fuel could be very significant: The Saturn V first stage at MECO was at 68km @ Mach 8. It also represents 77% of the initial launch mass of a Saturn V. Watching a recent Falcon 9 launch, we see somewhat similar profile: First stage MECO is at 75km @ Mach 7.7 at around 83% of the total mass at launch. $\endgroup$ – TemporalWolf Aug 22 '18 at 21:26
  • 2
    $\begingroup$ @Uwe It's an interesting idea though, since those first 10km are the just about the hardest 10km of the launch. $\endgroup$ – Mast Aug 23 '18 at 6:29
  • $\begingroup$ -1, These would not be used to slow the vehicle $\endgroup$ – JCRM Aug 23 '18 at 7:14
  • 2
    $\begingroup$ @Uwe by that definition the ISS was not in LEO for the first decade or so, at 350km. I'm not sure where the "LEO starts at 400km" came from. And SABREs design goal is to switch to LOX at 28km, not 10km. $\endgroup$ – TemporalWolf Aug 24 '18 at 18:24

I'll look at the physical side why zu they are not used:

Air breathing engines (or rather an engine) consist of two parts:

  1. Fuel (usually not the same as the rocket engines - RP-1 is out)
  2. the Engines themselves

And here the tyrany of space flight kicks in: any mass you want to bring down, you first need to bring up. Which increases the dry weight in the Tsiolkovsky rocket formula... and thus reduces efficiency logarythmically.


The mass of a parachute is comparably tiny to a descent capsule, almost neglectable. An engine that can put out the same amount of work (force over time) during the descent (stored in fuel) as the parachute applies would increase the mass of the descent vessel massively: You add the dead weight of the engine and the weight of the fuel storing this work. Let's look at this in Formulae:$$W_{\text{Parachute}}=F_{\text{Lift}}\times T_{\text{Descent}} ; F_{\text{Lift}}\propto \frac {⌀_\text{Parachute}} m$$ $$W_{\text{Engine}}=\int_T F_{\text{Engine}}dt \propto \frac{dM}{dt} \times a$$

The crux here is needing to break the additional weight of the engine and the fuel (force ~ mass) inside the same time frame to keep the descent profile, which demands much higher accelerations, which again demands a more powerful (and heavier engine), driving the need for higher accelerations up. Note that this is for engines that are retroburning.


Now, that is much less of a problem for spacecraft other than capsules: Spaceshuttle and similar ideas do fly to their destination, using air lift over wings to redirect vertical to horizontal speed. Due to the lift of the ing and their much more shallow descent profile they CAN fly (and break) with a TWR<1. But again the tyrany of spaceflight kicks in! You need to haul the mass of the engines and their fuel (which, unless you use RP-1 isn't useable by your rocket engines) to space, and you won't bei able to use this again till you are in the atmosphere again. That is drastically reducing the payload (as it increases the dead weight, and Tsiolkovsky has a logarithmic factor of dead weight to start weight). For maximum efficiency, cargolifters are usually designed to best have every part functional as often and long as possible during the flight - so best is one type of fuel (RP-1 would work, but has not the best Power/weight ratio) and as few different engines as possible.


But What If parachutes are not enough? Well, then it is still better to use the next most efficient package to break the cargo - which is an SRB, as it has the least dead weight and the best Work/Mass ratio. Pretty much any capsule since Mercury uses a SRB retrorocket-set to break into its descent path from orbit. You get the best bang for your weight from them after all.

If steering is needed, this tips the balance for either using the RCS (the Space shuttle did this to get into descent) or rocket engines already in the design for orbital maneuvering, following the tyranny of spaceflight's dictate: minimize the dead weight.

  • $\begingroup$ You can't really blame the rocket equation on anything here. That is about the expenential expensiveness of achieving some Δv. Yeah, increased payload will also make the ascent more expensive, but only linearly – that's not specific for rockets; ordinary aircraft also need more fuel when fully loaded. Contrary to your argument, air-breathing engines would have the advantage over retrorockets that they're not subject to the rocket equation with the Δv you need for braking. That's also the benefit of parachutes, but those don't give you the nice controllability of a propelled landing. $\endgroup$ – leftaroundabout Aug 22 '18 at 10:02
  • $\begingroup$ ...I would agree however if you say that if you're going to bring air-breathing engines, you should make sure to also use them for ascend to avoid the rocket equation there where it's most painful. That's the idea behind Skylon. $\endgroup$ – leftaroundabout Aug 22 '18 at 10:07
  • 1
    $\begingroup$ @leftaroundabout You misunderstand what I try to say, and the rocket equation: you don't need a linear increased fuel, it is more: you don't just need adjust to carry the payload, you also need to carry the extra fuel, which again needs to be carried by a smaller increment - the Tsiolkovsky formula is logarythmic. Also, you miss the point: My argument is "work per weight", which is ginormous for parachutes (due to their low weight), very good for SRBs (Pretty much any descent vehicle uses them), Good for rocket engines (they benefit from gimbal). Turboprops are in comparison just too heavy. $\endgroup$ – Trish Aug 22 '18 at 14:31
  • $\begingroup$ $\Delta v = v_c\cdot \ln\tfrac{m_0}{m_f} \Longleftrightarrow m_0 = m_f\cdot \exp\left(\tfrac{\Delta v}{v_c}\right)$. The $m_f$ does not stand the exponential. $\endgroup$ – leftaroundabout Aug 22 '18 at 14:40
  • $\begingroup$ I'm afraid not. You argue that work/mass ratio (which basically amounts to specific impulse) is the crucial parameter, and parachutes > rockets > turbojets in that regard. But this is just not true: air-breathing engines have much higher specific impulse than any rocket engine (turbofan ca. 3000 s, the best rockets 450 s). Only when you compare thrust to mass ratios, rockets win, but thrust alone doesn't isn't sufficient for propulsive landing. The Soyuz approach is to first remove most velocity with parachutes and use light rockets only for the very last bit. $\endgroup$ – leftaroundabout Aug 22 '18 at 15:16

Problem 1: "Slow down on reentry"... dealing with many times the speed of sound. Turbojet and Turbo fan do not work well at many times the speed of sound, they normally slow down air to slower than speed of sound at combustion stage, so subsonic combustion can keep up with the air. (You are also dealing with vast amounts of energy that would likely destroy the turbines due to physical stresses at supersonic speeds)

Problem 2: Air gets very hot when compressed because of many times the speed of sound, add heat of combustion and materials in turbines can't take the heat.

Normally when talking many times speed of sound, scram jet or ram jets are used, scram jets are very bleeding edge and trying to use them to reverse thrust rather than add thrust would need a redesign and risk more destruction from heat higher than materials can handle.

Buron had engines to speed up rather than slow down, just like a normal commercial jet engine (commercial jets using engines as braking do it only when landing at slow speed for short periods of time). When Buron is down to speeds of a normal jet at end of reentry, the idea was engines could kick in and then Buron flew like a jet to airport to extend options for landing compared to gliding without power.

Normally adding drag... whether parachute, wings, variable wings, changing profile or shape of vehicle, etc makes much more sense than air breathing reverse thrust at mach 10 speeds... You still have issues with overheating, stresses, etc but much less than trying to do combustion and reverse thrust at same time. Once you are down to speeds similar to normal jet planes then normally can land like a jet plane if wanted or regular parachute otherwise (weight not a problem, just need bigger wings or parachute).

Yes in theory you could land like a helicopter rather than like a jet airplane, to allow landing on spot without long runway with an air breathing engine (a helicopter engine could be described as slowing down the helicopter), so far no one has seen need for that.

In case of Musk and friends "reusable rockets"... simpler and probably less weight to just use the rocket engine already there then try to do one of alternatives... and an airbreathing engine with enough thrust to do vertical landing would probably be one of poorer alternatives to gliding down with wings, parachute, etc.

In theory, you could have a very big plane that matches speed and grabs/docks with the reentry vehicle like a hawk grabbing a flying bird as its prey if you could master all the challenges like risk of them colliding. Possible your very big plane could have ram jets or even scram jets and grab the reentry rocket at mach 3+. So far "not worth it" to even seriously think about by guys like Musk making rockets, as too hard and expensive and adds to little advantage.

  • $\begingroup$ uhh... "you could land like a helicopter [...] so far no one has seen need for that" I'll just say Roton saw a need for it $\endgroup$ – JCRM Aug 23 '18 at 9:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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