Grand visions

The space race is on between Europe and the US to develop the next generation of super-telescopes, but the engineering challenges will be daunting.

The world has embarked on a new space race. But rather than competing to send men to the Moon, scientists and astronomers from Europe and the US are racing to develop the next generation of ground-based, super-sized telescopes that could chart the most distant reaches of our galaxy.

Last month, scientists from the European Southern Observatory (ESO) in Garching, Germany, launched the science case for the design and development of a telescope that will dwarf all others. ESO is an intergovernmental organisation designed to provide research facilities for European astronomers, and is pushing for funding bodies and governments to loosen the purse strings enough to fund the Extremely Large Telescope (ELT).

The power of a telescope is necessarily limited by the size of the mirror that collects light. The bigger the mirror, the fainter the objects in space it can detect. According to the group of more than 100 ESO astronomers and scientists, the science case for a huge telescope with a light-capturing mirror of 50–100m diameter is overwhelming.

The real scientific impact of a telescope of this size is unknown, but astronomers hope an ELT would give them the opportunity to scan space for Earth-like, ‘blue’ planets orbiting distant stars. Astronomers predict an ELT could not only detect Earth-like planets but also be used to analyse those planets’ atmospheres for signs of life. With a giant telescope, scientists also hope finally to unravel the mystery of the dark matter and dark energy that makes up more than 95 per cent of outer space.

For the past five years, the crown jewel of European astronomy has been the Very Large Telescope array (VLT), at the European observatory in Cerro Paranal, Chile. This consists of four telescopes which can be used singly or as a unit. Each telescope has a main mirror of 8.2m diameter and is housed in a compact, thermally controlled building that moves and rotates with the telescope.

Housing astronomical instruments such as high-resolution spectrometers and multi-object spectrographs, the VLT can make observations in a broad spectral region from deep ultraviolet to midinfrared.

Since September 2000, when all four telescopes began functioning, this flagship facility has been the largest optical telescope array in the world. But an ELT would dwarf the VLT in size, engineering and the exciting science it could make possible.

There are currently two main designs vying to be Europe’s ELT. The smaller of the two is the Euro50, brainchild of a consortium of institutes headed by Sweden’s LundUniversity. The design is for a 50m primary mirror using adaptive optics which would boast a resolution around 200 times better than existing telescopes. Euro50 would be sited on the island of La Palma in the Canaries.

But Euro50 is a mere sprat compared with ESO’s proposal for a massive 100m telescope provisionally named OWL (Overwhelmingly Large Telescope) — an ironic reference to the industry’s proclivity for rather obvious titles.

OWL is huge, both in terms of scale and the technology needed to make it work. To put it into context, the world’s largest single ground-based telescope is the Keck telescope in California which spans a mere 10m. OWL’s massive primary mirror, ten times the size of Keck, will be made from 3,048 identical hexagonal elements, each 1.6m in diameter. Astronomers have not yet decided where OWL would be located, but suggestions include the South Pole, which would have optimum atmospheric properties.

The OWL is the largest of the projected supertelescope designs.

Meanwhile, in the US, scientists and astronomers are working on the Thirty-Meter Telescope (TMT) which highlights a fundamental difference in tactics between the US and Europe. In scaling down the size of the telescope the US team, which includes researchers from the University of California, aims to get its telescope operational by 2015, some five to 10 years before a European ELT is likely to see ‘first light’.

It is based on much of the same technology as Euro50, and European researchers will be monitoring its progress with interest. The TMT’s less ambitious size will make it a good testing ground for the technology required in the much larger European telescopes.

Dr Philippe Dierickz, design study project manager at ESO, is co-ordinating the OWL project. He sees the relationship between the US and European projects as one of cooperation as well as competition. ‘The

American design essentially extrapolates existing telescope technology, while OWL has a radically different approach. The Americans may have an advantage in that their design is smaller but it is by no means certain they will be ready first,’ he said.

The reason for Dierickz’s assertion that the Europeans still have a chance of beating their American cousins to the punch lies in the telescope’s design.

One of the crucial differences between OWL on one hand, and the Euro50 and TMT on the other is the choice of primary mirror design. Traditional telescopes have used aspherical mirrors which offer a far superior image to spherical ones.

However, the OWL design is based on a large spherical primary mirror. This means identical spherical segments can be fitted in stages, allowing OWL to be used — even when it is only partcompleted — at 60m diameter.

‘We can still do meaningful science with a diameter of 50–60m; it can be competitively used even half-finished,’ said Dierickz. He hopes OWL can be functioning and producing scientific results by 2017, but admits it is difficult to judge when it will be ready.

But David Walker, head of the Optical Science Laboratory at UCL and technical director at Zeeko, a Leicestershire-based optics company involved in the Euro50 project, is less effusive on its prospects. Walker claimed OWL’s design is based on the assumption that it’s not possible, economically or technically, to create an aspherical mirror of the necessary size or quality for an ELT. In his opinion, the decision to plump for a large spherical mirror is a poor compromise. ‘OWL’s spherical shell, made from thousands of identical spherical tiles, would leave the telescope open to the optical defect known as spherical aberration,’ claimed Walker. ‘For the same reason car headlights are parabolic, the best shape for telescope mirrors is aspherical.’

To overcome the spherical aberration problem, OWL would need a complex optical system involving a secondary, flat 30m diameter mirror, a couple of 8m diameter severely aspherical components plus smaller 4m optics. Walker is sceptical.

‘Their view seems to be: “Let’s make the primary spherical because it’s big and difficult to manufacture and then we’ll fix the enormous defects by adding other components downstream”. Using more segmented mirrors than necessary, you have many more gaps between segments. This will compromise the telescope’s performance.’

Traditionally, telescopes have aspherical primary mirrors which are difficult to manufacture. As the grinding or polishing tool moves across an aspherical surface it may make full contact in the middle, but will not contact intimately at the edges, according to Walker. ‘Euro50 will have a 50m diameter, aspheric primary mirror and a smaller, secondary mirror.’ he said. ‘This will give it a clean near-infrared signature but will require the manufacture of these large, aspherical segments. We believe it is possible to make it aspherical, and the Americans have the same idea for their 30m telescope.’

Zeeko has been collaborating with CranfieldUniversity to develop machines capable of creating aspheric optics. Together they applied for a Research Council grant to establish a national centre for ultra-precision surfaces and received £4.2m for a pilot plant, based in North Wales, for manufacturing aspherical segments. When it is finished it will comprise a 1m Zeeko polishing machine, a 1m grinding machine from Cranfield and a 1m atomic plasma machine, also from Cranfield.

The Zeeko machine uses two precision-polishing techniques. The first utilises an inflated membrane polisher and the other a polishing jet which ejects a stream of polishing slurry to remove material from the surface. Using this machine, Zeeko aims to increase both processing speed and accuracy in creating high-precision aspheric components, essential if the plan to create a massive aspheric primary mirror is to become a reality.

And this polishing technology already has possible applications beyond telescope optics manufacture. It is currently developing a method for polishing prosthetic knee joints to make them last longer.

Weight is also a big issue when designing telescopes of this magnitude. Both Euro50 and OWL researchers are looking into the possibility of making mirror segments out of silicon carbide as an alternative to glass. For the same reason, carbon-fibre reinforced polymer (CFRP) is also being researched as a possible material for much of Euro50’s structural composition.

How the Euro50, could look, from the front and the side.

However, the technology most crucial to the success of an ELT is the further development of adaptive optics (AO) to correct the disturbance in light that occurs as it passes through the atmosphere before reaching the telescope.

According to Walker, this will be an enormous technical challenge. An ELT will need deformable optics on a whole new scale. AO clears up distorted images and provides images almost as good as those from telescopes in space.

As pockets of air move, photons in the light are deflected, causing a distorted, blurred image of the object on the telescope’s light-sensitive detector. Atmospheric disturbances like this are the reason stars appear to twinkle. Adaptive optics technology uses an advanced high-speed image analyser to measure image quality and send corrective signals at 100 times a second to a flexible mirror in front of the telescope. The mirror continuously changes its shape, effectively cancelling the distortion and returning a restored, uniform image. It is this mirror which will have to grow exponentially in relation to the size of the telescope — a major challenge for optics manufacturers.

Dierickz’s team at ESO is constantly investigating different techniques and materials for use in adaptive optics that will be able to cope with the challenges posed by an ELT. ‘The adaptive optics system is a real challenge,’ admitted Dierickz. ‘The corrector will have to be twice as big and with a movement nearly 100 times greater than current best correctors.’

One development in optics technology that has made an ELT viable is the advance in active optics for countering the effects of wind turbulence on a telescope’s structure. Active optics allows real-time adjustments of the telescope mirror and was first employed in the ESO 3.5m New Technology Telescope at La Silla, Chile, in 1989. In 1994 the Keck telescope was the first to use active optics in a segmented mirror.

Although the technique is proven, the real challenge will be to scale it up for an ELT’s mirror. The primary mirror is mounted on force actuators that function as motor-driven springs to adjust the mirror’s shape by applying force. The slightest error in a mirror of this size would ruin an image, so the electronics and mechanics will have to be extremely sophisticated.

As well as constantly adjusting the whole mirror to keep it focused, the thousands of individual segments in the primary mirror also need careful marshalling. To keep them positioned within a fraction of a wavelength they must be re-aligned several times a second. Because of the size of the projected ELTs, the segments will have to cope with large changes in air temperature, mounting imperfections and — critical on a huge telescope — the effect of wind on the frame.

In the OWL design, the segments are connected to position actuators that reposition the segments within a few nanometres. Edge sensors between the segments measure misalignments and feed the data to a computer which then calculates the re-adjustments for the actuators.

The UK Astronomy Technology Centre is leading the UK part of the ELT design study. Colin Cunningham, the centre’s director of technology development, said it is likely the European projects will have to converge at some point when a design for an ELT is finally agreed upon.

At the moment things are up in the air,’ he said. ‘An ELT is going to cost millions and at the moment we are still looking into raising that. I don’t know which design we’ll go for, but I know it will be the one that gives us the best science.’

According to Cunningham the number of technological and engineering problems that still exist, means that an ELT, however desirable, is still some way off. ‘To reach this performance affordably requires us to address many engineering and technological challenges. We must keep the technology independent of any one design.’