Updated 29 April 2010

At 05.12 Pacific Coast Time on 9 August, 2005, the Space Shuttle orbiter Discovery rolled to a stop on the runway at Edwards Air Force Base in California. The first shuttle mission for thirty months had been completed safely. If Mission STS-114 had flown according to its original schedule, there would probably have been little media attention. Instead, people all over the world were watching, and breathing a collective sigh of relief.

The anxiety provoked by Discovery's homecoming was stimulated by the tragedy that overtook its predecessor, Columbia. The last hour of any shuttle mission represents an extreme challenge to pilot and engineer alike. This Entry is an attempt to explain that challenge.

The Fundamental Problem in Re-Entry

The phase of a spaceflight during which the craft leaves earth orbit and descends through the upper atmosphere is generally known as 're-entry'. In order to be in stable earth orbit in the first place, the craft must have attained and maintained a critical velocity. This orbital velocity is nearly 30 times the speed of sound - around a little under 8 kilometres per second. If the craft moves any more slowly than this, it will descend to a lower orbit under the influence of gravity. Because the craft will now encounter atmospheric resistance, it will lose energy and fall to earth.

In order to make a safe landing, a returning spacecraft has to lose nearly all of that orbital speed. The operation is basically a reversal of the launch phase, and this means that the returning craft must sink as much kinetic energy¹ as the propulsion systems generated between lift-off and orbit. Theoretically speaking, there are four fundamentally different methods of doing this:

• Powered Deceleration
• Energy Exchange
• Mass Shedding
• Energy Dissipation

An explanation of the shuttle's methods will be helped by a brief consideration of all four.

Powered Deceleration

This can be achieved by rocket thrust opposed to the direction of motion. The shuttle does this at the very beginning of its re-entry, to trim the speed and so initiate the descent from orbit. The equivalence of re-entry and launch energy means that it's not a practical proposition for the whole descent, however. Using this method would more than double the fuel load required at launch², and require the craft to carry half of it while in space.

Energy Exchange

This ingenious braking method requires the conversion of kinetic energy into potential energy, which can then be stored in some kind of separate device. It's ultimately the best solution of the four on grounds of sustainability: the energy is not wasted and can in principle be used in subsequent spaceflights. The best-known concept of this kind is the space elevator, a theoretically possible machine but one that's far beyond present-day engineering feasibility. Energy exchange is not an option for today's spacecraft, therefore, though it is a serious objective for far-future technological development.

Mass Shedding

This method is conceptually exemplified by a pilot ejecting from his damaged plane - most of the system's energy remains invested in the part that carries on and crashes. Though the idea works, the shuttle program's fundamental premise of orbiter re-usability precludes it. The only mass shed by the shuttle is in the launch phase, when the spent rocket-boosters and empty fuel tank fall back to the sea. If the vehicle was dispensable, the astronauts could transfer into a small escape vehicle which would then be jettisoned from the main craft. The re-entry phase of other space programs (such as Apollo) made use of mass shedding. The shuttle program, though, is bound by the principle that everything that enters earth orbit comes home again.

Energy Dissipation

This method differs from energy exchange in that the kinetic energy is progressively (and wastefully) converted to another form, such as heat, as the descent proceeds. It's the principal method for re-entry braking in all mankind's space programs to date, and for the space shuttle it's the only method, once the descent proper has begun. The amount of energy that must be dissipated is very large. Stopping a one hundred tonne craft from a speed of 13 km per second in eighty minutes requires nearly 2,000 megawatts of power³.

Ballistics Versus Flight

In an earlier period of manned spaceflight, the re-entry descent through the atmosphere was always ballistic - a plunge with a calculable but otherwise uncontrollable trajectory. The ballistic method remains the simplest and probably the safest way to return to earth from space. Its disadvantage is that it only works for a residual capsule of small size and constrained geometry. A great deal of equipment has to be ejected to strip down to the ballistic re-entry configuration. A great deal of money, of course, is being thrown away too.

For the purposes of this Entry Apollo's command module can be considered to be an example of a ballistic re-entry vehicle⁴. It weighed about six tonnes, and was basically a steel cone just large enough to accommodate three men. The heat shield accounted for nearly a quarter of the initial weight of the module. Unlike the shuttle's long-lived tiles, this shield (made from fibreglass and resin) was ablative⁵ in function.

Ultimately, all practical re-entry vehicles devised to date dissipate their kinetic energy by converting it to heat. The shuttle is no exception, but since it weighs about 100 tonnes, it has to lose nearly twenty times as much energy as did the Apollo command module. This means that the shuttle must also generate and dissipate twenty times as much heat. If it had to do that in the short time it would take to fall from orbit, it would be incinerated - even if constructed using the most advanced materials available today. A ballistic re-entry is therefore not practical for a reusable spacecraft of any size.

The only alternative is to descend at a lower rate than free-fall. An acceptable rate of descent is one at which the rate of heating will be reduced to a manageable level. For reasons already given, the orbiter cannot use powered rockets to slow its fall. This means that the space shuttle has to be able to fly (ie, to generate an upward force as a result of its forward movement that tends to keep it airborne. Such a force is termed 'lift'). Since the orbiter is unpowered, except for small thrusters used to orient and steer the vehicle, it is technically a glider. It doesn't look much like a glider, and in conventional terms it's actually a very poor one, being much too heavy and squat to generate significant low-speed lift. This is not a conventional duty, though. By the time that the orbiter's speed matches that of a sports glider in flight, it's already on the tarmac and rolling to a stop⁶.

Whereas conventional gliders do not approach the speed of sound, the shuttle orbiter exhibits aerodynamics which are not merely supersonic; they are hypersonic. The conditions experienced by the flight surfaces are complex and characteristic, but the upshot of the huge speed range over which flight occurs is that the design of the craft has to cope with several wholly different regimes. From atmospheric penetration at around Mach⁷ 25 or more, the plasma shroud persists down to about Mach 15, whereupon it is replaced by a less energetic though still chemically-reactive flow. This gives out at around Mach 7, and the shock wave dissipates at about Mach 3. Below this speed, manual flight becomes routine, although there is still the descent through the sound barrier to come.

An obvious consequence of the need for the orbiter to be able to glide is that it must have wings. They are not added lightly, however. As we are about to see, structures that stick out are undesirable during the first period of re-entry.

Pressure Waves and Plasma

It's usually assumed that the mechanism of heating in re-entry is by friction ie, viscous drag in the atmosphere. In fact this is the predominant mechanism only at lower altitudes, as air density increases. During the fastest and hottest part of the descent, less familiar physics is in play.

A re-entering vehicle develops a very energetic pressure wave at its leading surfaces. The energy density is sufficient to cause atmospheric molecules to dissociate⁸, and their component atoms to become ionized. The vehicle thus descends in a superheated shroud of incandescent plasma.

Plasma, known as the fourth state of matter, does not conform to the gas laws of conventional thermodynamics, although it does share one familiar property - a proportionality between pressure and temperature in a contained system. The formation of the pressure wave, therefore, also creates extreme temperatures. The plasma stream is electrostatically-charged too, and so it concentrates at acute surface contours, a behaviour related to the ground-level phenomenon of corona discharge (St Elmo's Fire). The resultant effect is particularly intense local heating at the airframe's leading edges.

This is almost certainly what inflicted catastrophic damage to Columbia. It's known that insulation foam became detached from the main tank, and impacted on the left wing soon after lift-off. The impact is thought to have breached the heat-shield, also creating a jagged protrusion about which the plasma flare could concentrate. If plasma heating generates a concavity in a leading edge, the vehicle is in serious trouble, because the localised high radius of curvature is maintained as melting proceeds, and the gash propagates. Debris evidence from Columbia suggests that such a plasma blowtorch action damaged components as far back as the wheel-well, destroying cabling and sensors linked to critical avionics en route.

This is also the reason that the removal of protruding tile-spacers and insulation fabric was carried out in the STS-114 spacewalks. Protuberances such as these are plasma concentrators. Though the local melting effect will usually be self-arresting by smoothing out the affected area, the risk of a gash forming and allowing propagation through the structure was evidently judged to be unacceptable.

The First Phase of Re-Entry

About seventy minutes before touchdown, the actions that commit the orbiter to re-entry are taken. Until this point, it's in stable orbit at an altitude of some 150 km and a velocity a little over 29,000 kph. The attitude of the craft is nose-first and belly-up (ie topside facing the earth). Though such an orbit can be sustained indefinitely, a speed trim of about 300 kph is sufficient to initiate descent.

Shedding around one percent of the momentum of a vehicle might seem like a gentle touch on the brakes, but it's a complex operation for the shuttle. The orbiter is not yet flying, because there is no substantial atmosphere for flight surfaces to bear on at this altitude. The only way to exert forces on the vehicle is by jet action. Small thrusters are sufficient to alter its attitude, but a reduction of speed requires the firing of relatively large rocket motors pointing forward, opposite to the sense of travel.

The shuttle has engines at the back, pointing backwards. This orientation is the obvious one to the layman - it fits our idea of the design of both a rocket and an aeroplane - and it's the right orientation for the launch phase. Once in orbit, the main engines are useless, since their fuel supply was the now-detached main tank. The engines used for the re-entry burn are alongside them, though, mounted in pods adjacent to the tail. It will now be disconcertingly apparent that, at the one time the orbiter needs these engines, they're aligned in entirely the wrong direction.

This is not an engineering oversight, but it is a compromise. Any practical engine that faced forwards in the hot phase of re-entry would have plasma-concentrating contours - and would thus melt catastrophically due to plasma erosion. The orbital manoeuvring engines thus have to be sheltered in the lee of the craft during the main part of the descent, which means they have to be shrouded in a recess at the back. It also means that the vehicle must be turned round prior to the deceleration burn, and then turned back for re-entry. Moreover, it means that the burn itself has to be applied while flying backwards.

Thrusters flip the orbiter right over about its pitch axis, so that its tail now points forward. This is not as alarming as it seems, because as yet there is no frictional shear on the airframe. The main engines are then fired for about twenty seconds. The thrusters fire again in a different configuration, rotating the craft on its yaw axis. It emerges from this sequence in the descent attitude, nose forward and slightly upward, and belly facing downward. This belly is covered in the ceramic tiles that give the orbiter its nickname, the 'flying brickyard'. The speed of the vehicle is now below the critical orbital velocity, and so it starts to come down.⁹

If anything goes wrong after this point, it has to be dealt with in the strict time-frame and trajectory of the planned descent. There is no longer a way to pull out¹⁰.

Entry Interface

At around thirty-five minutes before touchdown and at an altitude of around 120 km, the orbiter enters a discernable atmosphere. Though still extremely rarefied (containing little oxygen), there is now enough external matter to undergo ionisation, and the plasma flare begins to form. The linear distance to the landing site is about 8,000 km, so that this event invariably takes place over the Pacific¹¹. From this point, the angle of attack is critical, and is maintained at 40 degrees by automatic and continuous thruster trims. Any shallower, and the orbiter will experience excessive lift and overfly its destination. Any steeper, and it will burn up.

At an altitude of 85 km, the flight surfaces of the orbiter become usable. Under automatic control, a series of four S-bend turns is now performed, with the craft banking through 80 degrees at the fullest extent of the roll. The object is to lose speed more quickly.

It's now twenty minutes to touchdown, and the plasma shroud is at its most intense. This is the hottest phase of re-entry, and in terms of the integrity of the structure the most dangerous. The progressive disintegration of Columbia began at about this point, culminating in the vehicle yawing almost sideways about five minutes later, in which attitude the stresses on the airframe tore what was left of it apart.

Unlike the Apollo missions, the shuttle never normally endures a period of radio silence. The enclosure of the vehicle by plasma is never complete : there is always a window at the tail through which a signal can be beamed out. Nowadays there are satellites to receive it too, and so a continuous transmission returns to earth. Cockpit communication is a small part of this broadcast; by far the biggest volume of data is the telemetry stream from the avionics system¹².

Final Descent

Ten minutes before touchdown, the orbiter is still 40 km up, and its speed is still nearly 10,000 kph (Mach 8). The peak skin temperature, on the underside of the wings close to the leading edges, is around 1,600°C - hot enough to melt steel. Plasma flaring has ceased, and has now been replaced by reactive hypersonic flow. Drag on the craft increases greatly, and deceleration takes place more rapidly. The pilot will now be ready to take over control, and normally does so about seven minutes out, as the speed drops below 4,000 kph and into a stable supersonic regime. For the first time, the vehicle is pitched forward so that the nose is pointing downwards. Conventional flight surfaces now become effective, such as the air brakes which can be applied through the fantail configuration of the orbiter's rudder¹³.

If landing at Cape Canaveral, the orbiter will be over the Florida Panhandle and four minutes from touchdown as it drops below the speed of sound. It produces two sonic booms as it does so. At about 40 kilometres range from the runway, the pilot begins to adjust the heading. The descent is aggressive: ninety seconds from touchdown, the altitude is still around 4,000 m, and for the next minute the orbiter will lose height about twenty times quicker than would a commercial airliner. At six hundred metres and thirty seconds from landing, the pilot pulls up the nose once more and deploys the landing gear.

The speed at touchdown is 350 kph. The rear wheels come down first, and a parachute is released to slow the vehicle - it has no means through which to exert reverse thrust. The orbiter's very robust wheel-brakes are progressively applied¹⁴. As the speed falls to about 150 kph, the nose-wheel touches down and full braking is employed. The parachute is jettisoned just before the craft rolls to a complete stop, somewhat more than two kilometres from its point of touchdown.

The Right Stuff

The Space Shuttle is nearing the end of its service life, and the politics of our times mean that the prospects for its successor are not assured. America has not wholly recovered from the twin shocks of Challenger and Columbia. The myth of straightforward space travel has been exploded by these tragedies, and other disasters such as 9/11 and Hurricane Katrina have added to a nation's doubts about the space program's priority in terms of funds.

This Entry has attempted to explain the scale of the technical challenge implicit in just one part of a shuttle mission. It's legitimate to question the usefulness of the space program, but it's wrong to overlook (let alone to denigrate) the remarkable technology, enterprise and organisation that underpins it. Many people believe that this aspect of the program alone is sufficient to justify its advancement. It surely marks one of the most brilliant achievements in science and engineering yet attained by our species.

Perhaps it's time that the United States lost its odd penchant for oversimplifying its greatest triumphs. There is nothing ordinary about the people who engineer and pilot machines like this one. They are an elite, and they personify the immense contribution of their nation to human technology and civilization.

The orbiter is much more than a flying brickyard. Its homecoming, though fraught with danger, is anything but a reckless trust-to-luck. Engineering doesn't come more audacious than this. The stars are beckoning, and the shuttle marks one more step on Mankind's Glorious Road.

The X-37B Military Spaceplane

In a first of its kind the US military launched a prototype unpiloted spaceplane from Florida on 23 April 2010. The X-37B military vehicle is due to carry out the first autonomous re-entry and landing in the history of the US space programme.