NEXT WEEK, the space shuttle Discovery is due to begin a 10-day mission to
service the Hubble Space Telescope. After seven years in space, Hubble is
beginning to wear out. Radiation has degraded some of the spacecraft鈥檚
electronics, while micrometeoroids have occasionally slammed into its mirror,
damaging the surface. The craft also suffers from the effects of extreme
temperature changes as it passes in and out of the Earth鈥檚 shadow.
During the servicing mission, astronauts will remove and replace some of
Hubble鈥檚 ageing cameras, sensors and data recorders, giving the spacecraft a new
lease of life. If all goes to plan, a revamped Hubble will again be peering into
the depths of the Universe by the end of February. Further servicing missions in
1999 and 2002 should keep the spacecraft operating until 2005.
But Hubble can鈥檛 last forever. So NASA and American aerospace firms are
designing a replacement known as the Next Generation Space Telescope. Last
August, three competing designs were unveiled. The NGST will be bigger and
better than Hubble but lighter and less costly to build and operate. It will sit
some 1.5 million kilometres from Earth, at a point in space where the Earth and
the Sun鈥檚 gravitational fields cancel out. And even though the NGST would be
closer to the Sun than Hubble, its lonely post would help make it easier to
control, since it would be at a constant temperature and relatively unaffected
by gravity. Here it will observe the cosmos 24 hours a day, looking for young
galaxies at the edge of the Universe, imaging the planets and moons of the Solar
System, and hunting for Earth-like planets orbiting Sun-like stars.
Advertisement
But winning the funds and developing the technology for a successor to Hubble
will not be easy. Although hugely successful, Hubble has faced criticism over
its equally enormous budget. The telescope cost over $1 billion to build
and NASA was forced to spend a further $800 million to make good the
defects in its mirror. By the time it finally expires, Hubble will have cost a
further $4 billion to operate. 鈥淭he Hubble is wonderful. It鈥檚 terrific,
but at a quarter of a billion a year [to operate], we must plan a replacement,鈥�
says NASA chief Dan Goldin, who is leading the effort to build 鈥渃heaper, better
and faster鈥� spacecraft.
Counting the pennies
Goldin has placed strict limits on what NASA can spend on a new telescope.
The NGST must cost no more than $500 million to build and $400
million to operate for a decade. The key to achieving this cost saving will be
the use of a new generation of lightweight materials and autonomous computers,
explains Goldin. Many of these technologies will be taken from top-secret spy
satellite programmes while others must be developed from scratch. Nevertheless,
despite the reduced costs, some scientists say a replacement would be too
expensive, and that Hubble should be patched up one last time in 2005 and left
to soldier on until it finally wears out.
Competing for the contract to build the NGST are teams from TRW Space and
Electronics and Lockheed Martin, two American aerospace companies, and a third
team from NASA鈥檚 Goddard Space Flight Center. At the heart of their designs are
ultra-lightweight mirrors up to 8 metres across. Hubble鈥檚 glass mirror is 2.4
metres across and weighs 826 kilograms鈥攐ne of the reasons the 11.6-tonne
telescope had to be put in orbit by the space shuttle, at huge cost. The NGST
will weigh only 2500 kilograms in total and will be launched by a much cheaper
expendable rocket.
The Lockheed team proposes using a Russian Proton rocket, which is wide
enough to carry a 6-metre mirror. The mirror planned by Lockheed consists of a
layer of fused silica so thin that the entire reflecting surface will weigh less
than 7 kilograms. They will be supported by a carbon composite shell that is
flexible and rests on some 2000 actuators which can extend and contract,
adjusting the mirror鈥檚 shape to produce precisely the required optical surface
after the launch. The mirror would be impossible to operate on Earth, because
gravity would distort the surface, so the team is hoping to carry out tests on a
shuttle flight.
The other designs use similar deformable optics, but with a difference. Both
Goddard and TRW want to launch their 8-metre designs using American Atlas II
rockets, which cannot carry payloads measuring more than 3 metres across. So
these teams have developed ways of folding up their mirrors. TRW鈥檚 mirror, for
example, consists of six hexagonal gold-coated sections arranged around a
central hexagon, like petals round a flower (see
Diagram). For the launch, the
mirror segments will be stacked on top of each other but, in orbit, each segment
will swivel into position to form a mirror with a surface area of around 40
square metres.
Designed for spying
The idea comes from a top-secret project called the High Accuracy Reflector
Development Program run by TRW to develop optics for spy satellites. TRW claims
that this kind of approach could make it possible to unfold 20-metre telescopes.
鈥淵ou鈥檝e got to admit that鈥檚 innovative, and it鈥檚 certainly going to be tough,鈥�
admits Bernard Seery, NGST study team manager for the rival Goddard team.
One of the NGST鈥檚 major goals will be to look at young galaxies and how they
form. These galaxies can be seen at the edge of the visible Universe, and light
from events that took place there billions of years ago is only reaching us now.
The young galaxies are moving away from us at a colossal speed and this
鈥渞ed-shifts鈥� their light in much the same way that the frequency of a train
whistle changes as it passes by. The visible light from these galaxies is so
heavily red-shifted that it reaches us as infrared radiation. For this reason,
the NGST will be optimised for infrared imaging鈥攁lthough it will also be
able to capture visible light.
But infrared imaging introduces another problem. The telescope鈥檚 infrared
cameras must be cooled to about 30 kelvin (鈭�240 掳C) even though the
spacecraft will be in the constant glare of the Sun. This kind of problem
requires a whole new generation of technologies capable of operating reliably at
such low temperatures 鈥淭here鈥檚 a whole nature of what I call cryogenic
engineering that has to be brought to space. It鈥檚 not black magic, just
engineering,鈥� says Peter Stockman, NGST project scientist at the Space Telescope
Science Institute, whose primary site is on the Johns Hopkins University campus
in Baltimore.
The first priority is to shield the telescope and its cameras from the Sun.
Both Goddard and TRW have opted for sunshades that will unfold to about the size
of a tennis court. These shields and most of the spacecraft鈥檚 electronics will
operate at around 300 kelvin (30 掳C) insulated from the telescope by a short
mast. Once in shadow, any excess heat should radiate into space.
Reducing costs on the ground will be just as important as reducing them in
space. So the NGST will carry intelligent computers that will automate many of
the routine tasks such as the downloading of data and position
maintenance鈥攖asks that must be coordinated from the ground with Hubble.
When up and running, the new telescope could be run by one person working a
40-hour week, a small fraction of the labour needed to keep Hubble running.
But some scientists say that the NGST is still too expensive. They suggest
instead that improved ground control facilities could enable Hubble to last
beyond its planned demise, and that technical improvements would allow it to be
operated and maintained at a reduced cost. A study last year by the Association
of Universities for Research and Astronomy, based in Washingotn DC, concludes
that Hubble could be operated in a 鈥渘o repair, no upgrade鈥� mode well into the
21st century.
The proposal would require one additional servicing mission in 2005 to
replace worn-out equipment, a software upgrade that will make ground operations
more efficient and a transfer of ground control from the Goddard Space Center to
the Space Telescope Science Institute. This could reduce Hubble鈥檚 running costs
to about $50 million a year, one-fifth their present level. Later this
year, NASA will decide whether or not this is a viable option.
Cash squeeze
Ultimately, cost will be the deciding factor, and with falling budgets there
is a very real chance that the NGST could fall by the wayside. 鈥淭he laws of
physics say it鈥檚 possible, though the laws of humanity might not,鈥� says John
Mather, study scientist for the NGST at the Goddard Space Flight Center. Goldin
is more optimistic. He says new technology will allow a cheap and reliable
replacement. 鈥淲e鈥檝e opened up the technology flood gates a little bit and
there鈥檚 been a gusher.鈥�
Either way, Hubble is eventually going to stop functioning, says Robert
Williams, director of the Space Telescope Science Institute. And although he
supports the idea of extending its life, he admits: 鈥淭his and that will break,
and it will go brain dead.鈥�
Just what will happen to Hubble when it finally stops working has yet to be
decided. 杏吧原创s say it cannot be left to drift since an uncontrolled dive
into the Earth鈥檚 atmosphere could be a hazard for people on the ground. They
have a more romantic vision. They say Hubble should end its mission as it
started鈥攊n the payload bay of the space shuttle. The final shuttle to
visit Hubble should bring the telescope back to Earth, so that it can be placed
in a museum to inspire future generations of astronomers. Only this, they say,
will be a fitting end to a great mission.