Mission requirements for satellites are limitless. Among them, researchers share a desire for larger apertures on observational antennae and telescopes. The larger the antenna’s aperture is, the better its resolution and sensitivity are. The advantages brought by enlargement of the antenna’s diameter cannot be reproduced by any other way. However, even if we enlarge the antenna, there is still a problem if there are large shape deformations. Thus, observational researchers are now bringing their accuracy requirements for the antenna’s mirror plane to us, satellite structural engineers.
In case of a satellite using a large antenna
The permissible error of the mirror plane (theoretical error from parabola plane) is about 1/20 to 1/50 of the observation wavelength used. In the case of the large antenna (deployable-mesh, mirror-plane antenna) of the Engineering Test Satellite KIKU-8 (ETS-VIII) launched in 2006, the mirror-plane error was under 2.4mm in square mean value since the S-band (frequency 2 to 4GHz, wavelength 150 to 75mm) was used. The design requirements were extremely strict because it had deployable 17m x 19m antennas. For the antenna of the ASTRO-G satellite (planned for launch in 2012) which takes over the large antenna technology, we envisage its mirror accuracy to be under 0.4mm, one order of magnitude smaller than KIKU-8’s, since it will use one order of magnitude shorter wavelength, Ka-band (frequency 27 to 40GHz, wavelength 11.1 to 7.5mm).
Concerning actual errors in mirror accuracy, there are a number of factors apart from fabrication and assembly errors. The biggest factor is thermal-distortion error caused by the heat of solar light and the earth’s infrared/albedo. We must also keep in mind changes in dimension due to material degradation (change in elastic modulus must be considered as well) caused by the radiation environment in space.