The Department of Astronomy of Kyoto University and Kyoto University Graduate School of Science Affiliated Observatories are constructing an optical infrared telescope with a large aperture of 3.8 m based on new technology to monitor planets outside the solar system and explosions that occur when a star dies. The telescope will be the largest and the first in Japan to use segmented mirrors, an indispensable technology for huge next generation telescopes. The larger the diameter of the primary mirror, the more light the telescope collects and the better the spatial resolution. However, limitations in production and transportation mean it is difficult to set up a mirror larger than 10 m. We thus need to produce segmented mirrors, i.e., mirror segments 1-2 m in size that are rearranged on the telescope. Optical evaluation of the segmented mirrors and development of a mirror control device are the topics of my research.
Fig. 1 Optical infrared telescope with a 3.8 m aperture based on new technology being assembled next to Okayama Observatory. The largest telescope in Japan will be completed by spring 2017.
Dividing the primary mirror into segments contracts the dimensions of each segment to a size that is easy to handle. However, the developers are faced with two difficulties. The first is that existing measurement techniques cannot confirm whether each segment has been processed into the optimum shape or not. The primary mirror of an optical telescope needs to have a predetermined curve over its entire surface with accuracy in the tens of nanometers. In general, the curve is measured by using a laser interferometer. Sample light that has the same wavefront shape as the mirror to be prepared is reflected on the prepared mirror and is superimposed on reference light that has the ideal wavefront shape; shape differences are detected from the resultant interference fringes. However, a segmented mirror has an off-axis aspherical surface, and it is thus very difficult to prepare light that has the same wavefront shape. To solve this problem, we developed an interferometer that uses a Computer Generated Hologram (CGH). The hologram is a common optical element that produces a three-dimensional virtual image by reproducing the wave surface emitted from an object. It is possible to measure an arbitrary shape by using CGH, with an interferometer. We are improving the precision and workability of the interferometer by making use of the characteristics of CGH and also developing software for designing CGH that reproduces a target wavefront shape.
The second difficulty is the extreme precision required to correctly arrange the mirror segments on the frame of the telescope. The new telescope will consist of 6 fan-shaped mirror segments surrounded by 12 segments arranged along the circumference of the former like petals of a flower. This is the first attempt in the world at an optical infrared telescope, and the result depends on the establishment of control technology. The telescope will be built firm and rigid, but the frame structureto support the mirrors may become slightly displaced upon rotating the telescope towards the target astronomical body. The impact of using the telescope outdoors is not negligible either. Because the telescope will be exposed to the external environment, a system is required to detect any minute dislocation caused by temperature changes or the wind and to correct the mirror’s position. We are currently developing a system that measures the positions of the mirrors 200 times a second and correct the positions so that the dislocation between two adjacent mirrors never exceeds 30 nm. We are planning to install 72 displacement sensors and 54 actuators to the 18 mirror segments to correct the position and inclination of the mirrors. The displacement sensors are eddy-current sensors and each measures the level difference between two adjacent mirror segments. To adjust the position of a mirror by tens of nanometers, we will develop a level-type decelerator from which play is eliminated by using elastic deformation and combine it with a linear actuator. Development of a control algorithm is also essential for correcting the mirrors with minimum time spent.
Fig. 3 Principal mirror control system during experiment. The two plates on the top are dummies of fan-shaped mirror segments. Level difference and inclination between the two ‘mirrors’ are maintained at 0 by using he three actuators installed on the back side of the mirror.
For these projects, I’m currently working on seeking new knowledge to find the most precise sensors and actuators. The optimum arrangement of each device has already been determined by simulation, and the behaviors of the devices have been tested indoors. However, to detect a dislocation at the 10 nm level, the measurement error of the displacement sensor needs to account for the external environment. This will be one of the issues for introducing the sensor in the actual telescope which will be used in outdoors. We need better ideas not only on precision and stability but also on size, weight, and cost.
These studies also aim to develop technologies to be implemented in future telescopes and astronomical observation systems. Similar technologies will also be used in the international joint projects of the Thirty Meter Telescope, a giant next generation telescope also consisting of mirror segments, and the European Extremely Large Telescope. The highly precise measurement and control technologies will also serve as foundations for diverse R&D activities. I will widely publicize the knowledge and technological information acquired, cooperate with researchers and engineers involved in actual manufacturing, and mediate between them and natural sciences.