Schumacher Achim, Devaney Nicholas, Montoya Luzma
Instituto de Astrofisica de Canarias, GTC Project, La Laguna, Tenerife, Spain.
Appl Opt. 2002 Mar 1;41(7):1297-307. doi: 10.1364/ao.41.001297.
Future telescopes with diameters greater than 10 m, usually referred to as extremely large telescopes (ELTs), will employ segmented mirrors made up of hundreds or even thousands of segments, with tight constraints on the piston errors between individual segments. The 10-m Keck telescopes are routinely phased with the narrow-band phasing technique. This is a variation of the Shack-Hartmann wave-front sensor in which the signal is the correlation between individual subimages and simulated images. We have investigated the applicability of this technique to ELTs, and in the process we have developed what to our knowledge is a new algorithm in which each subimage provides on its own a piston-dependent value. We also discuss an alternative algorithm to resolve the lambda ambiguity that allows detection of problematic cases, and a modification of the singular-value-decomposition procedure used to phase the whole mirror, using weightings on individual measurement errors. By means of simulations we show that the modified technique shows improved performance and that it can work with sufficient precision on telescopes as large as 100 m.
未来直径大于10米的望远镜，通常被称为极大望远镜（ELT），将采用由数百甚至数千个镜片组成的拼接镜面，对各镜片之间的活塞误差有严格限制。10米口径的凯克望远镜通常采用窄带相位调整技术进行相位调整。这是夏克-哈特曼波前传感器的一种变体，其中信号是各个子图像与模拟图像之间的相关性。我们研究了该技术对极大望远镜的适用性，在此过程中，我们开发了一种据我们所知的新算法，其中每个子图像自身提供一个与活塞相关的值。我们还讨论了一种用于解决波长模糊性的替代算法，该算法能够检测出问题情况，以及对用于对整个镜面进行相位调整的奇异值分解程序的一种修改，即对各个测量误差进行加权。通过模拟，我们表明改进后的技术性能有所提升，并且它能够在直径达100米的望远镜上以足够的精度工作。
