¿ìè¶ÌÊÓƵ

Riding on a column of flame, the world’s first reusable rocket ship
took off from White Sands Missile Range, New Mexico, on 18 August 1993.
Guided by a crew on the ground, the ship rose 30 metres – then stopped.
Hovering, the stubby, 12-metre-tall, cone-shaped craft tilted and began
to travel sideways before coming to a gentle halt, nose pointing skywards,
90 metres downrange. Kicking out its tiny landing legs, the Delta Clipper-Experimental
(DC-X) rocket ship then lowered itself onto a modest concrete pad, landing
tail-first under its own power, the way God and Robert Heinlein intended.

Less than four weeks after its initial flight, this remarkable vehicle
repeated its performance, reaching three times the altitude and travelling
twice the horizontal distance. Two weeks later the DC-X flew yet again,
rising 365 metres above the desert plain before traversing, hovering and
landing. Between flights, the vehicle required no maintenance other than
a coat of paint to cover scorch marks on the bodywork. ‘Except for the weariness
of the crew,’ says Jess Sponable, the US Air Force officer who oversaw the
programme, ‘we could have flown the craft again within 48 hours.’

The DC-X is a one-third scale model of a reusable rocket that could
one day make space travel as routine as air travel is today. By the end
of the millennium, a full-scale Delta Clipper, the DC-1, could be flying
commercial cargo, astronauts, even passengers into low Earth orbit. But
in November 1993 the project ran out of money. The launch crew was recalled
to the Californian headquarters of McDonnell Douglas, the aerospace company
that built the craft, and the mothballed DC-X now sits in a temporary fabric
hanger in the windswept plains of the White Sands range.

Uncertain future

The DC-X was originally a product of the Strategic Defense Initiative
Organization (SDIO), better known as the Star Wars project, where it was
conceived as a cheap and easy way of launching military satellites in space.
Star Wars has since been cancelled, and although Congress allocated $40
million for further research on the DC-X, the Pentagon has no use for the
craft and would rather spend the money elsewhere. In April, it released
only $5 million to continue tests until the end of the year. The ship’s
fate is now in the hands of politicians in Washington, who are awaiting
publication next week of a government review of space policy that will point
the way forward for space rocket design.

The DC-X project is a step on the road to a single-stage-to-orbit (SSTO)
rocket of the kind envisioned by the earliest science fiction writers.
They described spaceships that flew, landed and took off again, like aeroplanes.
Dreams of an SSTO craft died in the 1920s, when scientists began to study
the ‘rocket equation’ first proposed in 1897 by Konstantin Tsiolkovski,
a Russian schoolteacher and early rocket enthusiast. The equation determines
the maximum speed of a rocket, given its weight and the power of its engines.
To go into orbit, the vehicle must attain a speed of 9100 metres per second,
and to achieve this, the equation demands that approximately 90 per cent
of the spacecraft’s launch weight – the mass fraction – must be fuel.

But a 90 per cent mass fraction was not possible with rocket technology
developed in the 1950s and 1960s. Engines were too inefficient and though
their fuselages were made of aluminium alloy, which had the best strength-to-weight
characteristics of any material available at the time, they were still too
heavy. A single-stage rocket vehicle built then could have achieved a mass
fraction of no more than 86 per cent, enough to achieve a velocity of only
7400 metres per second. The only designs that could reach orbit were built
in stages that were jettisoned to lighten the load as fuel was used up.
Rocket scientists have since designed most rockets like ammunition – to
be used once, then thrown away.

In the 1980s, however, the advent of lighter, stronger materials revived
the possibility of an SSTO. The materials were developed for a US hypersonic
aircraft called the National AeroSpace Plane, which was intended to reach
orbit with a single stage. The project was cancelled in 1992 after researchers
failed to solve problems associated with flying at speeds of up to Mach
18 in the dense lower atmosphere. But it led to several innovations, including
two key materials. One is a graphite-epoxy composite that is 60 per cent
lighter than aluminium and almost as strong, which could be used to build
the internal framework of a rocket and tanks holding the liquid hydrogen
fuel. The other is an aluminium-lithium alloy, 38 per cent lighter than
pure aluminium, which could be used to construct liquid oxygen fuel tanks.
With these materials, a 90 per cent mass fraction becomes possible.

Losing weight

In addition to lightweight materials, the programme produced materials
that can withstand the very high temperatures associated with hyper-sonic
flight and with re-entry into the Earth’s atmosphere. These include composites
of carbon and silicon caride, and a reusable, flexible, cross-woven carbon
fibre ‘quilt’ developed at NASA’s Ames Research Center. Such materials are
easier to manufacture and maintain than the tiles used on the space shuttle,
each of which must be individually moulded to follow the curve of the craft’s
surface. The composite of carbon and silicon carbide would cover the nose
of the craft which experiences the highest temperatures. Most of the rest
of the vehicle would be covered with honeycomb steel and honeycomb titanium
panels which spread the heat evenly across the surface. The carbon fibre
quilt, bonded to the underside of these panels, would also provide further
protection to parts of the craft. Such heatproofing would help an SSTO rocket
cope with temperatures of up to 1700 °C during re-entry.

But thermal protection materials add weight. To keep it to a minimum,
designers can use computer models developed for ballistic missiles, which
simulate the air flow around the craft during flight and re-entry and calculate
the heat generated on the surface. Using this information they can match
the amount of thermal protection material at each point to the temperatures
experienced there.

Other advances also help to reduce weight. Since the 1960s, electronic
components have become smaller, lighter and more efficient. The DC-X is
the first rocket to have an on-board computer-controlled autopilot, a system
taken directly from the McDonnell Douglas MD-11 jet airliner. And the hydraulic
control system that moves the engine nozzles during flight is smaller and
lighter than 1960s designs because of more efficient, computer-controlled
pumps and actuators. Engineers estimate that incorporating all this technology
into the full-scale Delta Clipper rocket will reduce its unfuelled weight
from 55 000 kilograms to around 47 000 kilograms, which will be light enough
to meet the required 90 per cent mass fraction.

The DC-X project began in 1989, when the SDIO asked aerospace contractors
to prepare designs for a prototype reusable SSTO rocket. This was intended
to pave the way for a full-sized design capable of placing a payload of
over 9100 kilograms in orbit. About 80 per cent of US satellites fall within
this range. The full-sized rocket was to be controllable from the ground
or by an on-board crew. While in space, the craft had to be capable of manoeuvring
easily to allow it to shift orbits in mid-flight. The craft would have to
be fully reusable, with a turnaround between flights of no more than seven
days using a ground crew of 50 or less.

This lean operation would make the new rocket cheap to fly. The SDIO
stipulated that the launch cost should be no more than $5 million, equivalent
to $550 per kilogram of payload. Conventional American rockets require
several thousand ground crew and cost around $50 million per flight. The
space shuttle is even more expensive. It requires 9000 people for each launch
on top of the 20 000 who maintain the four-shuttle fleet. Supporting this
standing army of experts costs NASA between $500 million and $1000 million
per flight, or $10 000 to $20 000 for each kilogram of payload placed in
orbit. On this basis, NASA estimates that an SSTO spacecraft could save
the US government up to $6 billion each year.

Safety was also an important factor in the design spec. The SDIO insisted
that the vehicle should be able to return to its launch site if an engine
failed during an ascent, just as an aircraft can. This opens the possibility
of carrying fare-paying passengers, like today’s commercial airliners.

Several aerospace contractors put forward proposals and all agreed that
an SSTO was feasible. Rockwell International advocated a vertical takeoff
and horizontal landing system, similar to the space shuttle. Boeing proposed
a horizontal takeoff and horizontal landing version. And a team from General
Dynamics studied a vertical takeoff and tail-first re-entry and landing
system that resembled an Apollo capsule.

The SDIO, however, chose the proposal from McDonnell Douglas. Its rocket
takes off and lands vertically but re-enters the atmosphere nose first.
This allowed the company to draw on its experience with ballistic missiles,
which also re-enter the atmosphere in this orientation. The craft then manoeuvres
into a tail-first attitude and lands under its own power. This allows it
to be launched again without having to be hoisted upright.

Cheap and cheerful

McDonnell Douglas claimed that the prototype could be built quickly
and cheaply, and in the event the DC-X was designed and built in less than
20 months at a cost of $60 million – about the same price as two zero-gravity
space shuttle toilets. Rocket scientists are half serious when they say
that the DC-X’s takeoff weight of 10 000 kilograms is less than the paperwork
required for each shuttle launch.

The prototype makes the most of off-the-shelf technology. For example,
the four rocket engines are Pratt & Whitney RL-10s fuelled with liquid
hydrogen and liquid oxygen, a design that has been used on the upper stages
of Titan and Atlas rockets since the 1960s. RL-10s normally produce a constant
thrust, but for the SSTO, engineers fitted control valves to vary the thrust.

The size of the nozzle mouth on rocket engines is a crucial factor,
the ideal size depending on the pressure of the surrounding air, if any.
If the nozzle mouth is too wide, the expansion of the exhaust occurs before
its force can be directed by the nozzle skirts. If it is too narrow, the
exhaust expands after it has left the nozzle mouth, wasting valuable thrust.
RL-10s were originally designed with wide-mouthed nozzles for use in the
upper atmosphere and in space. But the DC-X has to fly in the lower atmosphere,
where air pressure is higher, so its engines are fitted with smaller nozzles
to boost their efficiency.

The engines for an operational SSTO will have to operate both in space
and near the ground. So on the full-size rocket, which will be known as
Delta Clipper One (DC-1), the engine nozzles will have adjustable skirts.
At least eight of these modified RL-10s will be needed to power each DC-l.

The DC-X fuselage is a lightweight graphite-epoxy ‘aeroshell’ that was
moulded by Scaled Composites, a Californian firm that used similar materials
to build Voyager, the aircraft that in 1986 became the first to fly around
the world on a single tank of fuel. Because the prototype is not intended
for space travel, the designers did not consider a 90 per cent mass fraction
vital. They chose aluminium for the internal framework of the vehicle and
an aluminium alloy for the fuel tanks. The unfuelled craft weighs 10 000
kilograms and has low mass fraction of about 53 per cent.

Flight and ground operations for the prototype are identical to those
planned for the DC-l. Before takeoff, the craft sits on four simple vertical
steel posts that keep it a few feet above the ground. After landing, it
is towed back to the flight area. The shape of the DC-X mimics that of the
full-sized craft, so aerodyna-mic forces experienced by the prototype should
be similar to those on the full-scale version.

Even the White Sands launch site is itself a simulation of sorts. With
a temporary, fabric-covered hangar, simple concrete pads for takeoff and
landing, and fuel supplied by tanker lorries, the site is similar to an
emergency launch site that would have to be set up if a future Delta Clipper
were forced to land away from a spaceport.

The control system is equally modest. The three-person crew who fly
and monitor the vehicle are based in a 10-metre-long trailer known as the
flight operations control centre (FOCC), which is parked 5 kilometres from
the launch pad. During and after the flight, on-board sensors relay information
to PC-based workstations at the FOCC, monitoring everything from fuel levels
and whether the landing gear has extended, to the attitude of the ship and
the temperature on board.

Swoop of death

Tests of the DC-X had only just started when the project ran out of
cash. The successful trials involved simple manoeuvres at up to 400 metres:
hovering, horizontal movement and vertical landing. The modified rocket
engines worked as planned, as did the computer software that controls the
craft.

The DC-X also had to cope with ground effects – the interaction of
the exhaust plume with the ground. This can throw up debris that can damage
the craft, create shock waves which reflect upwards and batter the engines,
or even cause engine exhaust to bounce off the ground and incinerate the
vehicle. To counter these effects, a set of angled metal plates on the
launch pad directs the exhaust away from the craft during takeoff; on landing,
the engines tilt to the side. For further protection, the engine nozzles
are slightly inset into a heat shield on the bottom of the craft – an arrangement
that the designers liken to the shielding of the human eye by its eye socket.
The plans worked out well, and there were no reported problems with ground
effects during the tests.

The recently released $5 million will pay for more ambitious tests.
Later this year, with the DC-X travelling faster and at higher altitudes,
the team intends to simulate an engine failure and carry out a controlled
landing with the three still working. Other flights will test the craft
in bad weather, hovering 1500 metres above the ground in whatever wind,
rain and dust storms southern New Mexico can offer.

The riskiest manoeuvre, however, will be a simulated return from orbit
in which the craft has to be turned onto its tail in midair, after plummeting
nose first from a height of 6000 metres, reaching speeds of up to Mach
0.42 on the way. The shape of the craft will provide some aerodynamic lift
while hydraulically activated flaps on one side create drag that will rotate
it. During this manoeuvre, dubbed ‘the swoop of death’ by the DC-X team,
the forces on the craft will test the design to its limits. After the manoeuvre
is complete, the engines will fire and slow the craft down almost to a hover,
before it finally lands.

The swoop of death creates a unique set of problems. As liquid fuel
sloshes around in the tanks, the flow to the engines could be interrupted,
with catastrophic results. To combat this, engineers have fitted the tanks
with baffles, which computer simulations indicate should ensure a steady
supply of fuel.

Before embarking on the full-scale rocket, McDonnell Douglas plans an
intermediate design called the DC-X2. It will be twice the size of the DC-X
but built of the lightweight materials intended for the DC-1. With a mass
fraction of 70 per cent the DC-X2 will not reach orbit, but it will be able
to carry a 1360-kilogram payload to any spot on Earth in under an hour.
The final design, the DC-1, will stand 38 metres tall, weigh 580 100 kilograms
fully fuelled and include the option of a two-person crew compartment.

But the future of the Delta Clipper remains uncertain. American politics
have changed dramatically since McDonnell Douglas started building the DC-X.
But the administration is still taking an interest in reusable rocket technologies
and may give the DC-X rocket to NASA for further testing.

NASA, however, is not keen on the Delta Clipper. It does not believe
it can be made safe enough to land reliably, and fears the public relations
consequences of a spectacular failure should anything go wrong. Nor does
it believe that vertical landing technology offers advantages over winged
SSTOs. NASA would like to go for a single-stage craft that takes off vertically
but lands much like the space shuttle today.

But SSTO supporters such as Daniel Graham, the retired Army general
who in 1989 persuaded Vice President Dan Quayle to fund the project, doubt
that NASA is the right organisation to develop the DC-X anyway. ‘It takes
a small, dedicated crew of experienced engineers, using the best tools at
their disposal, to do this job cheaply and quickly,’ he says. ‘The DC-X
programme has demonstrated this in flying a revolutionary, reusable rocket,
while NASA produces nothing but stacks of paper studies.’

He may be right. Draft versions of the White House review of space policy
suggest that NASA should build a small-scale prototype SSTO by 1997. This
will be four years after the first flight of the DC-X and the year in which
McDonnell Douglas had originally planned to fly a full-sized Delta Clipper
into orbit.

Arlan Andrews is a freelance writer of nonfiction and science fiction,
and technology manager at the Sandia National Laboratories.