The Space Reviewin association with SpaceNews

Radian One spaceplane
Despite the failures of dozens of past efforts, companies like Radian Aerospace continue to pursue spaceplanes. (credit: Radian Aerospace)

Spaceplanes: why we need them, why they have failed, and how they can succeed

Bookmark and Share

“Rockets are terribly inefficient and expensive.” This admission can be found here in NASA’s own educational piece on the equation that governs rocket performance, also known as Tsiolkovsky's equation. But what is the alternative?

Perhaps the most frustrating aspect of the space age is our inability to create the obvious one: a successful spaceplane. This is a launch vehicle that takes off like an airplane, flies up to orbit, and returns to settle back on the runway. This feat is known as single stage to orbit (SSTO) and if you look this up in Wikipedia, here is what you find:

It is considered to be marginally possible to launch a single-stage-to-orbit chemically fueled spacecraft from Earth. The principal complicating factors for SSTO from Earth are: high orbital velocity of over 7,400 meters per second (27,000 km/h; 17,000 mph); the need to overcome Earth's gravity, especially in the early stages of flight; and flight within Earth's atmosphere, which limits speed in the early stages of flight due to drag, and influences engine performance.

Perhaps the most frustrating aspect of the space age is our inability to create the obvious one: a successful spaceplane.

“Marginally possible.” This gloomy observation has not inhibited attempts to overcome the challenge of creating a spaceship that can fly up and down like an aircraft. Indeed, if you go to Wikipedia you will find a list of some 60 spaceplane projects since 1945 that had this objective. None can have been said to have succeeded. A more recent review was given by Joe Scott.

Unsurprisingly to aerospace engineers, the villain sabotaging this dream is Tsiolokovsky’s Equation. Konstantin Tsiolokovsky was a Russian mathematician who, in 1896 demonstrated that Newton’s laws, when applied to rockets propelled with oxygen plus a reductant, results in a formula that limits all the non-propellant mass—the unfueled rocket and the payload—to a small fraction of the total weight. Specifically, when launching to low earth orbit at the minimum speed required of 7,400 meters per second, and not allowing for losses from atmospheric drag and from gravitational pull, the equation gives this table:

Propellant Specific Impulse (Isp) Maximum Non-Propellant Mass
Liquid Hydrogen plus LOX 450 18.7%
Kerosene plus LOX 330 10.1%
Solid Rocket Motor (SRM) 270 6.1%

In practical terms the ultimate problem is that, because space has no oxygen, rockets have to carry up about 2.4 tons of it for every ton of fuel they carry. Not surprisingly, given this constraint, rockets which reduce their mass to orbit by dropping off empty sections on the way up—staging—is the system that has been universally adopted to maximise the non-propellant mass (another Tsiolokovsky insight). A further drawback of SSTO’s is that they must haul along a pair of wings that are useful only for a very small part of the flight path, which further limits the thin weight margin available for the payload.

In the last decade spaceplanes have become even less competitive. The development of reusable rockets by SpaceX has brought payload costs to low Earth orbit (LEO) down from over $10,000 a kilogram to around $3,000. Under these circumstances, SSTO concepts might be expected to have a vanishingly small chance of being economically viable.

Yet still they come. Radian Aerospace, based in Renton, Washington, is planning to develop a delta-winged spaceplane about the size of a small commercial jet air transport. This will launch horizontally using a rocket-powered sled to allow the craft to conserve as much fuel as possible. Once aloft, three rocket engines put the spacecraft into orbit under a low-g ascent, followed by reentry and landing on a runway three kilometers long.

But Radian is right: spaceplanes are going to be essential if we are to continue to use satellite-based services. We are painting ourselves into a corner here, with ever-larger rockets carrying ever-larger numbers of satellites up, but with no means of servicing them in orbit.

Radian have raised $27.5 million of what they call “seed capital,” so presumably the final cost will be of the order of hundreds of millions of dollars. Given the price squeeze originating from SpaceX it is difficult to see a justification for this. With a two-ton payload and assuming a net revenue of $1,000 per kilogram, cash recovery will be $2 million a launch, so at least 50 launches will be needed just to break even.

But Radian is right: spaceplanes are going to be essential if we are to continue to use satellite-based services. We are painting ourselves into a corner here, with ever-larger rockets carrying ever-larger numbers of satellites up, but with no means of servicing them in orbit. They cannot be easily repaired or their positions adjusted, and they cannot be readily deorbited when they become obsolete.

The problem can be seen most clearly in the space debris challenge. In November 2022, the US Space Surveillance Network reported tracking 25,857 artificial objects in orbit above the Earth, of which 5,465 were operational satellites. However, those 20,000-odd other objects represent the tip of an iceberg; they are the space debris items that are big enough to be trackable. There are now, according to NASA, perhaps a hundred million orbiting objects with a diameter between 1 and 10 centimeters, and over 36,500 pieces with diameters greater than 10 centimeters.

NASA also has a good survey of the scores of debris capture proposals and the regulatory situation here. What is lacking is a vehicle that can go about deorbiting space junk by whatever means, come back to Earth for re-equipping and return back up to continue its work. A space plane.

Checkmate, or so it may seem.

The way forward lies again in Tsiolokovsky’s equation. The form that is of use here is:

ln (Initial Mass/Final Mass) = Delta V/(g0 * Isp)

So there are just two variables involved, the increase in velocity and the specific impulse. There is nothing to be done about the Isp once the choice of propellant and oxidizer has been made, but what about the effect of gravity and drag losses on the delta V?

Conceptually, if we are able to use air-breathing ramjets to take the spaceplane to the edge of space before handing over propulsion to a rocket motor, then not only will the ensuing rocket drag loss be small enough to be almost negligible, the initial velocity of this second stage could be more than Mach 5, countering the gravitational drag. The evidence for this speed comes from several sources, such as:

  1. The Boeing ramjet-powered ASALM missile demonstrated its hypersonic ability in 1980 when it reached Mach 5.5 (about 1,900 meters per second) at 12,000 meters after a fuel valve stuck open.
  2. In 1951 NACA (the NASA predecessor) launched a ramjet powered missile which reached an apogee of 159,000 feet (48.5 kilometers). This missile was launched at a 75-degree angle and ran out of fuel at 67,200 feet (20.5 kilometers) when it was passing Mach 2.92 (about 1,000 meters per second).

It is possible to gain a measure of gravitational drag from the trajectory of the air-launched Pegasus rocket, which was developed by Orbital ATK in 1990 and later built and launched by Northrop Grumman. Its Users Guide, issued in October 2015, gave operating data that showed that the zoom effect between the first and second stage gave an altitude gain of nearly 16 kilometers in return for a delta V loss of about 58 meters per second, a penalty of 3.6 meters per second per kilometer. This is happening at an altitude of over 70 kilometers, so this loss is almost solely from gravitational drag.

From these figures it appears that the gravitational penalty for a SRM being used in our spaceplane to lift it from about 50 kilometers up at about 1,000 meters per second to 200 kilometers at the minimum orbital speed of 7,400 meters per second would be, very roughly, 150 x 3.6 meters per second, or about 540 meters per second. So gravitational drag will require us to add this value to the required orbital speed, giving a total of 7,940 meters per second. Because of the uncertainties surrounding this value, we can round it up to 8,000 meters per second.

However, if there is a first-stage ramjet propulsion stage for our spaceplane, and it achieves Mach 5.5 at 70 kilometers, this would remove about 1,700 meters per second from this orbital speed target, reducing the delta V-to-orbit requirement of the spaceplane’s rocket stage to about 6,300 meters per second. If this work is undertaken using a simple SRM with an Isp of 270, then the available non-propellant mass of 6.1% shown in the table above increases to 10.8%, or rather better than a kerosene plus lox liquid-fueled rocket on the same basis.

What does this mean in terms of a spaceplane? It will be necessary to make a number of informed guesses on the non-propellant mass items at this stage; here they are:

Item Mass
Payload 0.5t
Spaceplane structure 1.5t*
Cold gas thruster fuel 1.0t**
Control Systems 0.5t
Total Non-Propellant Mass 3.5t

*This may seem light, but there is no undercarriage on this vehicle. It is launched and captured on a separate carriage on a track controlled by a linear induction motor. Additionally, the ramjets are expected to need to run for no more than about three minutes after launch, and so can be made of thin heat-resistant steel.
** For the extensive in-orbit movements required of an orbital service vehicle, perhaps nitrogen or possibly propane from left-over ramjet fuel.

If this non-propellant mass of 3.5 tons now represents 10.8% of the total at the point where the SRM takes over from the ramjets, then the SRM propellant mass would be 32.4 tons approximately, giving a launch total of about 35.9 tons. In addition, at launch there would be an extra two to two-and-a-half tons of propane as fuel for the ramjets and perhaps for in-orbit thruster use as well.

So, finally, a practical spaceplane. A bonus is that by being able to reach orbit with ramjets and a simple SRM it will have almost no moving parts. The concept is expanded upon in

Note: we are now moderating comments. There will be a delay in posting comments and no guarantee that all submitted comments will be posted.