GTCdigital

Full list of news

Practically perfect mirrors

29/10/2003

The 36 segments that will make up the Gran Telescopio CANARIAS' (GTC's) primary mirror have to be treated with extraordinary care, not only during their manufacture but also when they are polished.

In this issue you can find out how the mirror blanks are quality checked when they leave the factory and how they are subsequently polished. The whole process is designed to make sure that nothing comes between us and a faithful view of the Universe.

WHO MADE THEM?

The German company "Schott" was awarded the contract to build the 42 blanks (36 plus 6 spares) that will make up the GTC’s primary mirror - a giant jigsaw puzzle 11,310 metres in diameter. The segments are 1.9 metres in diameter, 8 centimetres thick, weigh 470 kilos and are made of Zerodur™, a vitreous ceramic material extraordinarily uniform in composition that undergoes virtually no change as temperatures fluctuate.

Grantecan selected Schott because of the German company’s leading position as a producer of optical glass products, vitreous ceramics like Zerodur™ and everyday materials like Ceran™, which is used in ceramic cooker hobs. Schott also produced the blanks used for the four large VLT telescopes in Chile, the two Keck telescopes in Hawaii, and the mirrors for three of the “Meteosat-2” weather satellites. The 42 GTC’s blanks are finished and being polished at the SAGEM’s factory.

TESTING

To make sure that the blanks will meet the required standards, the raw material has to be checked for any impurities that could detract from the observing quality of the mirrors.

The level of perfection required for the GTC’s primary mirror components is so demanding that only materials like Zerodur™, one of the most uniform and stable in the world, will do. That is why the blanks were so rigorously checked. As Javier Castro, the engineer in charge of Optics at the GTC Projetc Office, explains, “we looked for tiny imperfections in a near-perfect material.” The task is made even more difficult by the sheer size of the blanks. So how is this material, whose properties are so hard to benchmark, tested for quality?

Broadly speaking, we subjected the blanks to four tests: they must be the right size, have a minimal expansion coefficient, have very low internal tension and have as few internal defects as possible.

At the same time as making sure that the segments were the right size - to within one hundredth of a millimetre - the thermal expansion coefficient was also checked. This is the amount the material expands with temperature fluctuations, which must be as small as possible so that the mirrors do not deform and so distort the image. It was measured by a high-precision dilatometer, specially designed by Schott for Zerodur™ quality control. The dilatometer used samples of material left over from before the machining, which therefore came from the same casting and had the same thermal history as the GTC’s hexagonal segments themselves.

Grantecan, the public company responsible for building the telescope, has specified an expansion coefficient of 0.050 microns / metre. In other words, a metre of Zerodur™ may only expand by approximately 5 hundredths of a micron (1/20,000 millimeters) per degree of temperature fluctuation. Put differently, a metre of Zerodur™ would have to be heated by 200 degrees to expand as much as one micron. Expansion like this, which would be barely noticeable, is called “zero expansion”, and only a very few materials in the world can achieve it. A metre of normal steel, for example, has an expansion coefficient of around 15 microns per metre per degree, and other ceramic materials can only attain two microns per metre per degree. Tests confirm that the blanks have an expansion coefficient ten times less than that specified.

The internal tensions of the material were also tested - if they are too slack the segments can deform. The benefit of these tests is that the hexagons that will make up the primary mirror will keep their shape for a long time. They are so consistent that, if a hexagon broke in two, the shape of each half would change by less than one micron.

Similarly the internal defects in the material - such as tiny air bubbles - that are unavoidable in any product produced by casting, were also tested. These air bubbles must be just 2 millimetres or less within the critical volume (the outermost 10mm of blank adjacent to the mirror’s surface) and 5 millimetres in the non-critical volume.

At present, once the blanks has been manufactured, they are being polished. Responsibility for this falls to the French company SAGEM. That is another story, though - one that we will return to in a later issue.

Text: José Manuel Abad Liñán
Natalia R. Zelman

Full list of news