Motors drive vision of galaxies far, far away

Thousands of actuator motors and sensors will enable the European Extremely Large Telescope to focus light generated light years away into a clear visual image.

“And the Lord said let there be light.” The famous words from Genesis may have passed the lips of engineers, scientists and businessmen in December 2014, when the Council of the European Space Organisation ESO approved the construction of the European Extremely Large Telescope, or E-ELT. The green light on the two-phase build opened up the market for components required to build this enormous, pan-European project, valued Euro1.083bn at 2012 prices.


Located in Cerro Armazones, deep in Chile’s Atacama Desert, the telescope’s scientific purpose is to search for exoplanets – planets outside our known universe – and the stellar composition of nearby galaxies.

An extremely large telescope needs extremely accurate moving parts. E-ELT’s five mirrors must align precisely to steer the beam toward a Nasmyth focal surface, where the final visible image is produced. The E-ELT will gather more light than all of the existing 8-10 metre class telescopes in the world combined, and 100 million times more light than the human eye.

The engineering required to convert light millions of miles away to a focused image is out of this world. The optical structures are mounted on an altitude-azimuth telescope main structure that uses the “rocking chair” concept, with two massive cradles for the elevation motions and two major azimuth tracks, which operate the vertical and horizontal axes. The structure weighs approximately 3,000 tonnes.


Active and adaptive optics

One of the great challenges of the E-ELT is to calibrate its tremendous size with the demands of both active optics and adaptive optics.

Big primary mirrors allow astronomers to capture more light and a perfectly shaped mirror surface is needed to avoid distortions; the effective combination of the two makes it possible to observe fainter objects.  But with mirrors over five metres diameter, the image quality decreases enormously as gravity pulls the mirrors out of shape – maintaining a perfect shape becomes harder as telescope mirrors become larger.

ESO engineer Raymond Wilson came up with a clever and “simple” idea called active optics. A thin and deformable primary mirror would be controlled by an active support system would apply the necessary force to correct for gravitationally-induced deformations as the telescope changes its orientation

Active optics require actuators, motors that move very accurately and can be controlled precisely: by pushing the mirror, they correct its shape and compensate for the distortion produced by gravity. As the telescope moves, this active system can maintain the correct shape of the mirror. The corrections applied by the actuators are calculated in real time thanks to a computer with an image analyser that detects even the smallest deviations from the ideal mirror shape.

The VLT Unit Telescopes are equipped with the best active optics system constructed to date

The discipline of active optics faces its biggest challenge yet with the 39-metre primary mirror of the E-ELT and its 798 individual segments. Each segment can be moved by a piston and tip-tilt mechanism, making this mosaic work as a single giant mirror by compensating for the effects of temperature fluctuations and gravity.

The E-ELT also has to deal with the distortion caused by the Earth’s atmosphere. ESO scientists use adaptive optics here. Precisely engineered, deformable mirrors controlled by computers can correct in real-time for the atmospheric distortion, making the images obtained almost as sharp as those taken in space. Adaptive optics allows the corrected optical system to observe finer details of much fainter astronomical objects than is otherwise possible from the ground.

The engineering challenge for ESO is to synchronise the whole arrangement, blending adaptive optics with the primary active optics, simultaneously.

Optomechanics – The main mirror units, M1 to M5

The design of the business end of the telescope, the primary mirror control strategy, takes into account wind disturbances, sensor and actuator noise – both edge sensor and the embedded position actuator sensors – changing gravity, thermal expansion and vibrations. Several control strategies were evaluated at ESO and by an external study with the University of Liege.

Engineers made a trade-off analysis between global, modal, local and the LTSI (Linear Time and Space Invariant) time baselines and decided that modal control forms the baseline. Within this baseline the segments are commanded in piston tip-tilt to reduce cross-coupling at the segment level.

Manoeuvring the mirror segments

The E-ELT primary mirror (M1) is a 39-metre diameter elliptic concave mirror, with a 69-metre radius of curvature. The M1 mirror is made of discrete optical elements: the primary mirror segments.

There is a wonderful understatement in the E-ELT Construction Proposal. “Any strategy that addresses 798 mirrors with almost 2,500 actuators and 5000 pairs of edge sensors needs to be investigated for robustness”. Stability and wind strength are key variables.


The mirror segments are axially supported on 27-point identical whiffletrees, structures that spread load equally via linkages. A lateral restraint is located in the centre of the segment using a membrane to allow limited motion in the direction orthogonal to the back surface.

Actuators, edge sensors and whiffletrees work together to manoeuvre the mirror segments precisely at the correct speed, within the physical constraints of the environment. A total of seven sectors, 931 segments, are procured. Coating strategies have dictated the design; having seven segments per family allows for a realistic operation scheme in relation to coating, where it is estimated that up to two segments will be recoated per day.

Three position actuators move the whiffletree, and consequently the segment, in piston and tip-tilt. The actuators are required to have sufficient stroke to reposition the segment to its nominal position compensating for the deflections of the underlying telescope structure.

In addition, the actuators need to provide the resolution and accuracy necessary for phasing the primary mirror in the presence of disturbances. These are dominated by the wind across the front surface of the primary mirror and possible vibrations arising from machinery either directly under the segments or transmitted through the structure to the primary segments.

The position actuator is based on a high-bandwidth voice-coil actuator in series with a gravity off-loading electromechanical stage. An encoder is used to close the local loop.

Opportunities for UK companies

While many of the contracts for the dome and main structure have been awarded or will have been awarded by the end of 2015, no contracts have yet been awarded the main mirror assemblies where the majority of actuators, drives and sensors are fitted. Tenders for position actuators for the M1 unit will be received in 2017

“We are encouraging all UK companies that qualify to manufacture these mechanical components to bid,” says Sandi Wilson, UK Industry Liaison Officer for ESO at the Science & Technology Facilities Council in Edinburgh. “We believe that defence especially, and aerospace, companies will have the necessary capabilities to bid for these contracts and because E-ELT is a European member organisation, the UK has a very fair shout.”

The E-ELT literature says that the ELT programme will invest some Eu820 million in industry. In line with ESO’s procurement policy, the bulk of this investment will be made in the member states and Chile. How much of the project is Britain likely to get? The head of contracts and procurements at ESO said: “There is not a hard percentage per Member State. However ESO monitors the return per Member State, taking into account all the procurements, and factors this in to encourage companies from Member States with less work.”