Re-usable space orbiters are not as easy as they look. Rather than go all-out for a single-stage spaceplane, an incremental approach might prove more practicable. Here is a quick overview of the main issues and a possible design solution.
Getting into orbit is hard work. The climb to space is bad enough, but it uses only a fraction of the energy needed for orbital flight. The real work lies in accelerating to far higher speed once the craft has reached space. In energy terms, getting to Earth orbit is halfway to anywhere in the Universe.
The first obstacle any spacecraft faces on takeoff is the thick pea-soup atmosphere at ground level. Unrestrained, way before the craft ever reaches orbital speed, air pressure on the nose builds up to enormous levels, temperatures soar and the nose shatters or melts. So speed has to start moderate and carefully increase with altitude until the ship reaches space.
In the upper atmosphere, air pressure is much reduced and hypersonic speeds are practicable. However this can pose its own challenges, especially in engine design. At some point, air-breathing engines must become ineffective and only rockets can take you properly into space.
The thermodynamics of getting up are bad enough. But, perhaps surprisingly, getting down again is even worse. Not only does the aerodynamic drag of the initial flight get turned into heat all over again, but this time the huge amount of kinetic energy built up to reach orbital speed must also be dissipated through heat. Even solid rocks, known as meteorites, melt and explode when subjected to this kind of treatment.
Orbital spaceflight can thus be divided into half-a-dozen main phases, each with its particular demands on the spacecraft:
A good design will work effectively and safely throughout, and especially efficiently once out in space.
There are some basic design principles which apply to any space orbiter.
The most important one is to try and avoid carrying into orbit anything that is needed only for the first stages of flight. Anything - anything at all - not needed in space is dead weight for the hardest part of the trip, the acceleration to orbit. And because this last phase is far the hardest work, any dead weight carried will cause an immense waste of fuel. Indeed, if you try to carry everything in a simple rocket and just add more fuel to lift it then the extra fuel tankage needs more power to lit it, getting you into a vicious circle where the rocket can never reach orbit no matter how huge you make it. You have to find something more efficient. One approach is the multi-stage vehicle, which discards its heavy launcher before reaching space. Another is to gain what help you can from the air around the lower part of the trajectory. These approaches can be combined.
Spacecraft experience powerful accelerations and hence strong gee forces. Strong acceleration can dramatically improve overall performance, but at the same time everything has to be strengthened to take the extra gees and that adds weight which reduces performance. The two effects must be balanced out. Then again, some payloads, especially humans, cannot tolerate very high gee forces. This can limit the technologies available to the designer. Most satellites and other payloads are carefully designed and reinforced for high-gee flight, while even the most super-fit and highly trained astronauts can manage rather less gee. Ordinary scientists and passengers are more delicate and need lower gee still, posing the greatest engineering challenge. A design approach which suits one kind of payload may be unworkable for another.
A key principle to driving down cost is re-usability. Throwing away a space launcher after use and building another one for the next launch is obviously wasteful. But the technology to make a re-usable stage is difficult and hence also both expensive and potentially unreliable. The hardest stage to re-use has in practice proved to be the one which takes the orbital rocket as far as space but then drops back.
As technologies evolve, some of the design compromises shift and what suited past generations may no longer be optimal. But that does not mean that the laws of physics change. The most practical answer, at any given level of technical advancement, is not necessarily either the cheapest or the most appealing one.
The traditional design approach to getting into orbit is the vertical-launch rocket. The horizontal-launch spaceplane has long been proposed but never yet realised for orbital flight. A third and seldom studied possibility is the rotary wing, which falls somewhere between the other two.
A rocket is lifted entirely by the blast from its engines, so it takes off vertically. If the engines are not powerful enough to thrust the rocket up hard enough, it will waste time hovering on its exhaust, looking spectacular and burning fuel at a horrendous rate but taking a long time to get anywhere. This is hugely wasteful. For maximum efficiency it needs to get up into orbit as fast as possible. Rocket launches are thus characterised by hard acceleration and high gee forces. Unmanned rockets tend to demand that the payload be tough enough to take such high forces. But a manned rocket has to compromise on the acceleration in order to keep from squashing its crew, making it more expensive. Re-entry is especially hard on the crew, because of the enormous deceleration needed to slow down their space capsule. Unless you are prepared to waste huge amounts of fuel and money by taking things more slowly, only super-fit trained astronauts can ride a rocket into space and back.
Even an unmanned, high-gee rocket's speed will be limited at lower altitudes by the intense air pressures of high-speed flight. Accelerate too fast and the nose of the rocket will melt or break up. This puts a practical limit on the acceleration achievable in the atmosphere by all rockets. Again, re-entry is especially hard to manage due to the incredibly high speed of the re-entry vehicle. Hard deceleration and white-hot temperatures must be carefully balanced to stop the poor thing from melting.
A simple, one-step rocket is next to impossible to send into orbit: the engineering challenge of getting a single-stage rocket up there is about as impossible as things can get and, even if it were eventually achieved, would be ludicrously wasteful and expensive. Multiple rocket stages - the step-rocket - are essential to make such a rocket practical. Some step-rockets have a separate, unpowered re-entry capsule to bring the astronauts back to earth. But the last powered stage is the orbiter. It is just a big fuel tank with an engine at the back and a small payload at the front and its job is the long haul of acceleration in space. Behind it, a middle stage is needed to lift it up into space where it can work efficiently. This stage will then be discarded so that the orbiter need not take it along. And behind the middle stage is the real heavy-lifter, a huge construction of fuel tanks and giant engines, capable of lifting the upper stages off the launch pad and sustaining them – not too fast now – through the thick pea-soup to a higher altitude where they can get up some real speed. It is so powerful that it must inevitably run out of fuel quite quickly, so it gets left behind even earlier, when the second stage takes over. And that is just as well, because it is also the heaviest of the lot. In fact it is still often not powerful enough to lift everything including its own launch weight and extra boosters get strapped on for initial lift-off.
In contrast to a rocket, a spaceplane has wings so it can take off horizontally like an aeroplane and fly up through the air using only a fraction of the fuel needed for a thirsty rocket. It may also use air-breathing engines for the early stages, saving the weight of oxidant tanks. But when it reaches the edge of space its wings can no longer lift and its engines can no longer breathe. A hypersonic jet plane might be able to coast into a brief sub-orbital hop using a zoom climb manoeuvre, but for the long haul to orbit it has no choice but to switch to rocket power.
A great advantage of the spaceplane is that its launch site is less critical. A rocket benefits from careful choice of site to maximise benefit from the Earth's rotation. However a spaceplane can fly economically to wherever suits it best before aiming for space. It is also inherently easier to make re-usable, needing little more than the undercarriage it took off on for it to land safely again.
But the spaceplane has two particular problems to overcome. Firstly, during its horizontal flight in the atmosphere the main fuel tanks are laid sideways, like coke cans that have fallen over. In this position they sag and collapse very easily and must be strengthened. A vertical rocket does not suffer this problem, as a thin-walled cylinder containing pressurised fuel naturally supports itself when stood up on end and needs little extra stiffening. The spaceplane must carry the extra weight of the stiffening all the way into orbit, with all the waste that involves. Extending chines along the fuselage and blending them with the wing roots can help and can also improve the wing characteristics, but it is not a full solution and must still carry a weight penalty. Secondly, as the plane approaches space, those wings and air-breathing engine systems become yet more dead weight. In practice a multi-stage composite plane is still currently the most viable option, though neither the Pegasus air-launched rocket nor the Virgin Galactic SpaceShip has yet made it to orbit.
The single-stage-to-orbit (SSO) spaceplane is in reality little more practical than the single-stage rocket. The weight penalty on the horizontal spaceplane is more severe than most aerospace engineers seem to realise and the SSO plane must also carry along its low-altitude wings and air-breathing engines. In truth it remains more in the realm of science fiction than realistic planning. Not that that stops technologists from dreaming.
One hope is for a hybrid air-breathing rocket engine, such as the Reaction Engines SABRE, which would save so much weight of engines and tankage that the winged, single-stage-to-orbit (SSO) spaceplane becomes possible. The B.Ae HOTOL proposal caused quite a stir in the 1980s.
But possible in engineering terms does not mean efficient or economic. Just think what those advanced engines could bring to an inherently more efficient multi-stage spaceplane. More conventionally-powered designs already exist which use a heavy-lifting mothership with relatively low-power but very high-efficiency bypass turbofan engines, to carry the spaceplane up above the worst of the air and position it optimally, so that it can zoom up into space on rocket power. But such flights, whether manned or unmanned, remain sub-orbital and full orbital flight from anything other than a rocket launch remains a way off. How much SABRE engines could help the second stage along remains to be seen. At the time, my own take on the idea was for a cut-down HOTOL launched from a mothership, shown here.
Re-entry for a spaceplane is in theory easier, as it has wings it can use as airbrakes and to land gently on a runway. But it still has to take those wings all the way into orbit in the first place, a big penalty as we have seen.
The Space Shuttle represented something of a halfway house between these two approaches, being in essence a step-rocket on the way up and a spaceplane on the way back down. It had some ingenious modifications to improve low-gee operation and give some re-usability for only moderate penalties. Firstly, the main engines were all installed in the orbiter and were used throughout every powered flight stage. There was some penalty to this. They were a bit too small for takeoff and extra-large boosters were fitted to provide extra thrust throughout the first phase of flight. But they were also far more powerful than necessary for the small orbiter craft they were fitted to and some were just dead weight while the craft was in space. The spacecraft took off vertically so its main fuel tank, fitted externally, did not have the self-supporting problem. But the fuel needed for the orbital burn did need to be supported once the shuttle orbiter had tilted over in space. It was a very original arrangement, resolving a tight compromise between the complex design specification for moderate gee and re-usable orbiter against the technologies of the day, and it is unlikely ever to be seen again.
Another unusual possibility might be to use a rotary wing, like a helicopter, for at least part of the flight. Such a vehicle has characteristics between those of a rocket or a spaceplane. Flying vertically, it does not suffer the need for strengthened fuel tanks that the horizontal spaceplane does, while its rotors give it near-spaceplane lifting efficiency and airbreathing capability. But it is not an efficient spaceplane: although once tested as an SSO technology, its poor lift at altitude left it demonstrably unsuited. Nor would it perform much better as a powered wing for a multi-stage vehicle. The operational ceiling of a helicopter is much lower than that of a conventional aeroplane, which limits any potential advantages even over a ground-launched rocket. Similarly, its efficiency and range are not large, which limits its ability to position the second stage optimally. All in all, as a launcher it offers too little advantage over either step-rocket or spaceplane to make up for the cost, complexity and mechanical weight.
Re-entry may prove a different story. A rotary wing has yet to be tried in this mode, but it offers some potential advantages. Upon re-entry the rotor would auto-rotate without any need for power and would be flung out into the correct position by centrifugal force. Most modern helicopters can use auto-rotation to land safely in the event of an engine-out emergency. It is light weight: a rotor is stiffened by its centrifugal force and can be made lighter than a rigid fixed wing, while an unpowered rotor system (as in the autogyro) avoids much of the weight penalty of a helicopter. It is true that a rotor is less aerodynamically efficient then a fixed wing, but for the job of slowing descent that does not matter.
But at orbital speeds, a rotor would be difficult to protect from the heat of re-entry. Sub-orbital use is probably the best we could expect from it, and then only if it can be folded away neatly during launch. Such a foldaway structure is more practical for a rotor than a fixed wing, because the rotor hinges carry no bending and do not need to be locked rigid during flight.
A nominal three-stage design might have a mothership or carrier jet aeroplane which takes off horizontally and is fitted with supporting superstructure for a two-stage step rocket that is laid horizontally on top of the fuselage. In this way, the mothership can support the horizontal fuel tankage and absorb the accompanying weight penalty. The second stage forms the rear part of the rocket and has a rotating nose hub with rotor blades folded back down along its sides. The third, stub-winged orbiter stage is mounted in front of it. At altitude the composite craft performs a zoom-climb manoeuvre and throttles back its engines. This points it steeply upwards, temporarily offloading the rocket supports, so that when separation takes place the rocket is now in the correct position to support itself.
Could this all take place at supersonic speeds? Possibly, but not without proving the concept in a subsonic generation first. Any supersonic separation would need to ensure that shock waves generated by one stage did not impact and destabilise another. But more mundanely, the engineering and cost gains for the second stage would not be particularly enormous, while the engineering challenges and cost of a supersonic carrier certainly would be.
On separation, the second-stage engines fire and the rocket continues into space where it separates again in the usual way. The second stage falls tail-down until it meets the atmosphere. Its rotors then swing out and stream up behind it and, because of their slight inherent twist, begin auto-rotating. As they gather speed they swing outwards again, greatly increasing the air resistance and acting as air brakes. Soon the rotors can fully support the craft and conventional cyclic pitch control can be used to steer it onto its landing pad. At the last moment they angle up even more, expending their last energy in slowing it for a gentle landing. The Space Shuttle flared its nose up and slowed its descent at the last moment in just the same way. If airbreathing rockets such as the SABRE become available, the second stage can be significantly shrunk and lightened.
The third, orbital stage has small fixed wings for use as a shield and airbrake on re-entry, just like the Space Shuttle. But, relieved of the Shuttle's extra engine and fuel management systems, the orbiter is smaller and cheaper. It could also exert even gentler gee force, allowing its wings to be even smaller, although this might need several passes through the upper atmosphere and back into space to shed enough speed and heat before it could begin its final descent.
Some of the manoeuvres necessary to this scenario would have been far too dangerous to consider in the past. For example during the steep zoom-climb the weight of the upper stages would add to their drag to create a significant pitch-up moment. This would suddenly end when the engines were cut, ready for the upper stages to be released. Traditionally such drastic pitch changes would have been extremely dangerous and difficult to handle. But modern digital control systems have overcome all that, and safe operation would not be a problem.
Could a rotor system be adopted for the orbiter too? Possibly, but there are some difficult problems to overcome. A rotor orbiter would look much the same as a rotor second stage and for re-entry would also fall tail-down until it met the atmosphere. But it would differ significantly. Physically it would need a broad heat shield to swing closed across the engine nozzle during descent. Such a shield presents a real design challenge because it must be large enough to provide sufficient deceleration while also being strong enough to support the magnified weight of the rocket under that deceleration, all without weighing much itself or obstructing the landing feet. Assuming that to be possible, it could perhaps be a clamshell split down the middle or a multi-petaled flower. "Ouch!" You might think, "Won't the hypersoinc plasma of re-entry eat through the gap?" Well, remember that the Space Shuttle had heat shielding tiles across its underside. On its landing approach, doors opened in the shielding to reveal undercarriage fitted with ... rubber tyres! That's right, if rubber tyres can be protected by opening panels, I don't see why a rocket engine can't be.
Heat-shielding a set of spinning rotors would be a lot more difficult, probably impossible. It would most likely be necessary to keep them folded away, tucked behind the main heat shield, until the worst of the speed had been dissipated. The craft might at first fall very obliquely, skimming the outer atmosphere to steadily convert its orbital speed into heat and stream it away, while also being gentle on the astronauts. Only once it had slowed to speeds comparable to the sub-orbital second stage would the rotor be deployed.
Another quirk for the astronauts would be a need for the seats to tilt ninety degrees relative to the takeoff position. The Space Shuttle did exactly that the other way, changing from horizontal to vertical, but the whole idea of a spaceplane - or space copter as here - is to be less stressful. And for a cargo lifter it might not be possible to bring back large payloads shaped to be launched horizontally, as they might not survive being tipped over to land vertically.
Overall a rotor orbiter would be a far more risky technology to develop and may well prove impractical. Certainly, the rotor concept should be proven in a sub-orbital design before attempting orbital re-entry.
Updated 19 Jan 2018