This PDF file contains the front matter associated with SPIE Proceedings Volume 7281, including the Title Page, Copyright information, Table of Contents, Introduction, and the Conference Committee listing.

Optical lithography has driven on-going miniaturization in the microelectronics industry, thereby enabling continuation of 'Moore's Law'. To achieve this, lithographers have steadily reduced the wavelength of the illumination light used in the optical systems. However, as we transition from the visible spectrum, through ultra-violet, and now towards the soft x-ray wavelength regime, a host of new challenges are introduced. The majority of these challenges are related to material properties, as wavelength reduction significantly narrows the field of available materials that are both sufficiently transparent, as well as radiation resistant to the illumination light. We also are limited by the actual wavelengths that can be produced which deliver sufficient power to provide a production-worthy light source. In this paper, we will examine the history of wavelength transition in optical lithography, explaining the key material developments that enabled wavelengths such as 248nm to be highly successful, as well as explain the reasons wavelengths such as 157nm and 126nm were not adopted.

Initiated in 1968 by the first order of Max-Planck-Institute in Heidelberg successful history of ZERODUR^(R) continues now since 40 years. ZERODUR^(R) zero expansion glass ceramic from SCHOTT has been the material of choice in astronomy for decades, thanks to its special properties such as its extremely high thermal and mechanical stability. Today most of the major modern optical telescopes of the 4 m class and of the 8 m to 10 m class are equipped with ZERODUR^(R) . For the future several Extremely Large Telescope (ELT) projects are in development, which are designed with even larger primary mirrors ranging from 30 m to 42 m. Also here ZERODUR^(R) is under consideration. A historical review, prominent examples of astronomical projects with glass ceramic mirror substrates, and an outlook to the future is given in this paper.

In the case of multi-point supports, this paper studies the surface of large aperture thin-mirror in different support schemes. A finite element analysis model of thin circular aperture optical components is established. The response characteristics of the surface varies by the support units' size and the support units' pitch in the frequency domain have been analyzed. According to the results, the basic principles are concluded that different size and pitch of support units should be employed based on optical system.

A novel approach of co-phasing segmented mirrors using a Dispersed Rayleigh Interferometer (DRI) which is used for measuring the optical path difference between two parallel beams is proposed. Basic principle of DRI is introduced and configuration of co-phasing segmented primary mirrors for a Cassegrain telescope is presented. A fringe processing method called Two-Dimension Fringe Analyzing (TDFA) is proposed to analyze Dispersed Rayleigh Interference Pattern (DRIP) automatically. A bread board including a co-phasing error generator and a DRI is established to demonstration the approach proposed. Preliminary results using the DRI established show that, measurement range reaches 200 μm, repeatability is better than 2nm in whole range under condition of S/N=20 and can be accepted when S/N>8, and measurement error standard deviation in 1 μm range is 5.8nm (S/N=20). All of the data imply that DRI is a wide-range and high-accuracy method for co-phasing of segmented mirror, and can be used in the orbit for its compact configuration and no movement element.

In view of the primary mirror of large optical system, the influence of surface error on ratio of scattering loss (RSL) is analyzed. The mathematical model among the root-mean square (RMS) error, the correlation length (l) of autocorrelation function of optical elements and the RSL is established by assuming that the surface error is a stationary Gaussian random process with statistical optics theory and Harvey-Shack scatter theory, and the influence of surface error on RSL with simulation computation is analyzed. It is found that RSL increases almost in exponential form with the increasing of RMS and declines in exponential form with the increasing of the correlation length. The mathematical model agrees with the simulation results very well, which can be used to analyze the influence of RMS and the correlation length on the RSL and provide the support for the further finishing and figuring of the optical surface profile.

In terms of large-scale optical telescope, the design of multi-motor friction drive is obviously advantageous than that of single motor direct drive on the expense for manufacture of motors. However, to keep the high accuracy of tracking of multi-motor friction drive in certain velocity and acceleration, synchronized control for multi-motor and compensation to the mechanical resonance are needed. After designing appropriate multi-motor drive and synchronized compensation device, we overcame the interference among running motors, restricted the velocity difference in smooth running to a smaller range, and set a good foundation for the design of correcting parameter. Besides, to expand the closed loop bandwidth of the system, the control loop model has been identified, and the compensator based on the identified model effectively improved the influence of the mechanical resonance. The experimental results showed that for multi-motor friction drive of the 1.2-m large-scale Alt-Azimuth optical telescope, the proposed approach obtained high accuracy when running at the max velocity of 3 deg/s.

Ultra-thin mirror has a large diameter-thickness rate, so it will be more deformable because of the change of working conditions. We have performed an ultra thin mirror with 1.5mm glass membrane of 340mm diameter which is attached to a sample support structure through a set of precise screws. In this paper, some analyse and computing simulation for active ultra thin mirror surface by using finite element method (FEM) are introduced:(1) simulated an ultra thin mirror with a large ratio by software "patran";(2) With Zernike polynomials several low-order aberrations are got, which simulate the profile error after ultra thin mirror fabrication and the deformation because of change of working conditions; (3) After getting the influence function of single actuator, the optimum scheme of actuator array for an ultra thin active mirror with maximum corrective capability is achieved;(4)The optimal size of flange which attaches the actuator to glass membrane is determined;(5) The test for correcting the surface Figure of this sample mirror demonstrates the correctness of result of FEM.

Ultra-high precision measurement of the form is critical in deterministic machining of optical components. It is generally accomplished with an interferometer. However, a standard interferometer is incompetent for large convex aspherical surfaces or those with large relative apertures. Hence subaperture stitching interferometry was proposed to extend the vertical range of measurement and enhance the lateral resolution, based on the idea of "stitching pieces together". This paper presents a prototype design for subaperture stitching interferometer (SASI) for large optics. It applies to large flats (clear aperture<700mm), large concave spherical or aspherical surfaces (clear aperture<350mm, 1:2<relative aperture<1:1). Firstly, three designs are compared to choose an appropriate layout of the optical path for testing large optics. Then the multi-axis mechanical design is introduced in detail. The kinematical model of SASI is built according to the kinematics of open-loop robots. It helps to determine the stroke, resolution and accuracy of each axis. The advantages of the iterative subaperture stitching algorithm, e.g., no requirements of precise prior knowledge of subaperture alignment error, considerably facilitate the mechanical design.

The reflector is influenced by force and temperature more and more seriously with the increasing of reflector diameter. The error analysis and adjustment of reflector is important to satisfy the optical performance. In fact, the error of the reflector include figure error, position error and curvature radius error, which are often coupled and difficult to adjust. The method of the separation of the three type errors and way of adjustment are studied in this paper. The separation of the error is following the belows: first, the deformation of the nods on the reflector surface are analyzed and transfered to sag deformation; second, the position error is obtained by the least-square method and adjusted; third, the residual error is fit with the Zernike polynomial and the curvature radius error is acquired and adjusted, at last; the remained error is the high-frequency figure error. A reflector under the load of heat is simulated and analyzed in this paper with the method above. The results indicate that the error of piston and curvature radius are the major errors, so it is possible to adjust the error of piston and curvature radius alone to improve the quality of reflector in optical system, in order to simplify the mount system of the reflector.

This paper analyses the deforming and moving feature, when using stressed lap on both 1M F/1.4 convex hyperboloid; and 4M F/1.5 concave hyperboloid.

Large-aperture segmented primary mirror has been widely used in high-resolution space telescopes. After deployment, position error and surface error of each segmented mirror will lead inacceptable wavefront errors together. A critical problem is how to decouple position error and surface error from exit-pupil wavefront aberrations. We proposed a decoupling control strategy based on sensitivity matrix retrieval method. The sensitivity matrix of segmented mirror is calculated by simulation method and a new type of orthogonal polynomials is recommended to fit wavefront aberrations in non-circular area. First we give a simulation example to show how to decouple position errors in each degree of freedom if there are no surface errors. Then a further simulation reveals the decoupling control process for position and surface errors. The low order aberrations are corrected by position control actuators and high order aberrations are corrected by figure control actuators.

Ultra-thin mirror often is used in active optics and adaptive optics. To maintain its surface, activators are used. Many factors are important including how much area should the attaching area be. A simplified deformation model was established considering mirrors' practical using. It would deform linearly outside attaching area and keep no change within attaching area. In this model, aspherical error is a key factor. Using zernike polynomials, the relationship between attaching area and surface error introduce by activator is established. At last, finite element method is used to testify theoretical analysis. It proves to be reasonable that decrease the ratio of attaching area and activator's controlling area. To decrease stiff concentration, the ratio should be about 1/10.

In order to study more remote universe and the detailed structures of near stars, large-scale astronomical telescopes are very needed with the development of astronomy and astrophysics. In this trend, astronomical telescope becomes more and more huge, which leads its driving system to bear heavy nonlinear disturbances. The increased nonlinear disturbances especially caused by friction torque in the control system can easily bring tingle and stick-slip phenomena when the telescope tracks an object with an ultra-low velocity. However, conventional control approaches are difficult to realize high-precision controls and can decrease the quality of a telescope's observations. Therefore, it will be of significance in theory and in practice to develop an advanced new control method to restrain nonlinear disturbance and improve telescope's observation performance. Sliding mode approach has been applied in many other mechanical control systems since it is invariable to various disturbances. However, conventional sliding mode approach may cause dangerous high-frequency vibrations in corresponding control system, which may influence control performance or even lead the system unstable. To counteract the effect of above nuisance, a high-order sliding mode (HOSM) controller of third-order has been suggested in the large telescope's drive system through theoretic deduction and analysis. On account of that the HOSM approach needs all system states available, a sliding mode observer has then been designed in order to get the acceleration state of the drive system. Simulation results show that this approach can obtain high control precision and may satisfy the requirements of a telescope for a nicely ultra-low velocity.

The control system stability is very important to be taken into account whether in fast tracking control systems or in slow tracking control systems. In particular, position tracking error becomes a great rush as long as the velocity is approximate to zero, or when the gimbals reverse into another direction. This is mainly because the friction torque causes the control system instability. Particularly in the only speed tracking system, it is very clear that the system oscillates back and forth until the velocity turns into a great value. By Lyapunov stability theory, the closed-loop system with only the velocity feedback proves to be unstable in low velocity tracking control. The acceleration is directly equivalent to torque, so a control algorithm base acceleration feedback control is proposed and introduced into the regular feedback system to eliminate the effects of low velocity friction, which includes velocity loop and tracking loop. The angular acceleration of the gimbals is computed from high precision accelerometers, while the velocity is filtered from the high-precision encoder. Experiments show that the proposed method with acceleration feedback loop eliminates the low velocity friction for stability and achieves high-precision tracking control performances.

The deformation of large- aperture optical mirror is induced by gravity. It is very important to analyze the support structure and format in grounding and testing. By finite element software, a Φ620mm mirror is created and its theoretical analysis about three comparative testing programs of vertical support is proposed. Zernike polynomials are used to fit optical surface and separate corresponding aberrations. By reducing the aberrations with great impact on mirror, at the same time, RMS excel 0.025λ (λ=632.8nm) has been achieved. Experiments of three kinds of support verify that the entirely bound by strip is the best way in all supports. It affords a more scientific and rational support to large-diameter optical mirror reflection. Through research and analysis of support on large-diameter mirror by finite element, it provides a theory of design and guidance for the development and test of reflection mirror in future.

With the development of Earth and Space missions, the quest for higher spatial resolution for telescopes will turn into urgent need. Optical sparse-aperture system can be designed to obtain high resolution in astronomical object imaging with several independent small-aperture telescopes. However, the images from separated telescopes are not only overlapped but also have same phase. The analysis and discussion on system error which would affect the co-phasing is presented. The piston error, namely the relative longitudinal phase shift across the sub-aperture is an important error to affect the performance of the imaging system. In this paper, we demonstrate the effects of the piston error on the point spread function of the system. Then the simulation has been fulfilled with the specific system parameters. The direct imaging results are obtained. Due to the image restoration techniques, we can also produce higher resolution and better imaging quality result after the restoration. By using the correlation coefficient as the evaluation criterion of imaging, the results show that image quality is reducing with the increase of the piston error periodicity.

According to the topologic system of parallel supporting structure, the geometrical constraints are obtained. And the forward kinematic model with analytical expressions can be established by eliminating the unknown terms. It is shown that the analytical solutions of the forward kinematic model have 16 groups. In order to decrease the number of solutions and get exact position-stance of the parallel structure, Newton-Raphson method is used to search the best numerical solutions by the judgement of terminal value. The corresponding numerical simulation proves that the exact forward solution can be found rapidly by iterative steps. Moreover, aiming at improving numerical precision, some measures on the choice of initial value and iterative step have been put forward. The forward kinematic model provides the basis of the perfect control of parallel supporting structure.

Based on the analysis of mechanical structures and characteristics of several shafting reflecting mirror at home and abroad, the composition principle of FSM is introduced in detail, and a concrete designing scheme of shaftless FSM is proposed. Using resonant frequency as a starting point, according to elastic mechanics, a vibration system of FSM is established. The structure of elastic system of FSM, material, manufacturing and heat treatment are also introduced. The computer simulation of the structure is completed by using COSMOS software. At last, the corresponding experimental result of property of open-loop of FSM is also given.

Recent advancements in computational fluid dynamics (CFD) software now make it possible to perform a full three-dimensional turbulent air flow analysis without the need for a supercomputer. The presented study investigates the airflow effects over 2-meter TSU telescope using CFD method. The effects of air velocity, air pressure about statics and dynamics and K energy have been calculated by using multi-physical coupling theory and CFD methods. It is shown that 90° angle between main optics axis and horizontal line is the best observation direction in which the air velocity distribution, airflow pressure deformation and K energy are the least to other directions.

The four-bar linkage and stressed lap had an outstanding result on grinding and polishing axisymmetrical asphere^[1]. Its mechanical structure is succinct and direct, and control is also easy. In "Mathematical analysis about deformation and movement feature, when using stressed lap to off-axis segment" ^[2] (following using MADMF to represent it in this paper), author put forward two mechanical structures to achieve the axis of lap tilting on both X and Y directions, and overlap with normal line of off-axis segment's contacting point, both structures have a lower universal coupling and 3 pairs of hoisting joints. Based on adopting four-bar linkage to off-axis segment, now we put forward a new mechanical structure, and analyze the movement feature. On the other hand, we analyze the movement feature of four-bar linkage, and remind the attentive matters when using this structure to mirror.

Reflecting telescopes have been widely used for many outstanding advantages. But the field of a common reflecting system is too small to meet the needs nowadays. When a correcting lens is inserted at the rear of R-C system, the off-axis image quality is significantly improved. After optimizing with CODE V optical design software, a new design scheme of reflecting telescope with large field of view is achieved.

To shape aspheric surfaces on ultra-thin spherical mirrors can avoid problems of manufacture and measurement for large-aperture aspherical mirrors. In order to study the ability of shaping aspheric surfaces and find out some aspherical parameters that can describe the ability, the limit ability of shaping aspheric surfaces is investigated. Firstly, the relation between asphericity gradient and stress of shaping aspheric surfaces on ultra-thin mirrors is analyzed, and the asphericity gradient of spheres is determined to represent the limit ability. Secondly, based on an example of off-axis large-aperture ultra-thin mirrors, the spheres with the different asphericity gradient are worked out, and the figure errors and the maximal stresses for shaping aspheric surfaces are gotten by Finite Element Method (FEM). Thirdly, according to analysis results, the relation between maximal asphericity gradient of sphere and maximal stresses is created, and the relation between initial figure errors of spheres and maximal asphericity gradient is presented on ZERODUR material. Finally, the maximal stresses of other materials after deformation are solved by using Hook law and FEM. Above analysis results show that the material ZERODUR applied shaping aspherical surfaces is not broken under conditions of the asphericity gradient threshold 1.62e-5, the corresponding initial figure errors RMS threshold 0.49mm, and the P-V threshold 1.74mm, when the aspherical accuracy is 21.09nm. In addition, according to the maximal stresses of ZERODUR, the maximal stresses of other materials are estimated, and their limit ability of shaping aspheric surfaces also can be defined.

SiC is a preferential material for space large aperture mirror at present. Aperture of 1 meter sic mirror can meet most requirements of space optical systems. For larger aperture sic mirror, it is not economical and safe to make the monolithic body directly. The SiC mirror brazed assembly with several segments can solve these problems. In this paper, one Φ600 mm sic plane mirror as demonstration welded with 3 radial segments is roughly grinded to the required figure at first, then the mirror is finely grinded and polished by CCOS technology, the RMS of the figure error is less than 1/20λ (λ = 632.8nm) finally. The problems of the preferential removal near braze joint and material removing rate influenced by deformation of mirror surface are solved. Methods of reducing the residual stress of mirror are discussed. At last, some measurements and experiments are carried on, and the results conclude that the brazed SiC mirror can meet the requirement of large aperture optical systems.

It is found that space optical sensor is influenced negligibly by atmosphere turbulence when targeting the ground, which makes primary mirror figure deformation caused by gravity t variation and temperature become a part of main reasons for resolution degration. At present, a reference wavefront is used to detect the primary mirror figure deformation. In the choice of a proper reference light source, many aspects must be considered, such as environment, size, energy consumption, signal to noise ratio,etc.. The reference source method available can't satisfy the demands mentioned above. As a result, a new approach for detecting primary mirror figure deformation of space optical remote sensor is put forward, namely, the beacon is embedded in the system, and some sampling mirrors are located in the primary mirror. Then we can get light spot location variation of every sampling mirror when the primary mirror figure deformation occur, and figure out deformation distribution by modal reconstruction. From simulation, it is found that this approach can recover primary mirror figure deformation with RMS error less than λ/30, when the peak-valley value of figure deformation is around 1 λ .

In order to ascertain the effects of main errors of alt-alt photoelectric telescope on measurement accuracy, orientation error is analyzed. For analyzing optical, mechanical and electrical errors from measured target to telescope picture, we construct 7 coordinates such as: the earth's core coordinate, alt-alt coordinate, etc., carry through 21 times coordinate transformation, form measuring equation with 26 variables. And we utilize Monte Carlo method to calculate the orientation error, analyze the effects of errors upon pointing. The result indicates that the measuring error standard deviation of alt-alt telescope is minimum value about 10" at the zenith zone. And the deviation increases with the rotation angles of longitude and latitude axes. The model and the method not only can analyze and synthesize the main errors, but also provide a reference to the whole design.

Segmented deployable primary mirror is an effective way to increase the size of primary mirror to get higher resolution. We explore research on primary mirror with 8 quasi annulus sectors surrounding an octagon central segment, which is quite different to generally used structure with regular hexagon segments. The PSF of quasi annulus sector is deduced in this paper which hasn't been mentioned before. The expression of total primary mirror is described and diffraction effect on image quality with gaps and piston is discussed and simulated.

Phase diversity (PD), proposed by Gonsalves, is a kind of wavefront sensing technique based on measurement of two or more images of object. The optical system involved is relatively simple. It makes use of the methods of optimization and image processing, which can jointly estimate phase aberration as well as object itself simultaneously. The most significant characteristic of this technique is that it works well with extended scenes. Steepest descent method and conjugate gradient method both are preferable algorithms for nonlinear optimization. As a matter of fact, any one of two methods has some limitations. Steepest descent method is a local property and conjugate gradient method's convergence rate is slow. Combining two methods to develop a mixed algorithm, we can avoid entrapping into a local minimum and raise global convergence rate. Simulation results demonstrate that the hybrid algorithm has the features of quick convergence rate, comparatively large convergence range, which make the method of phase diversity remarkably robust and numerically efficient.

The performance of space-ground optical system (optics remote sensing system) depends on two aspects, one is the design, manufacture and assembly of system, the other is the influencing factors which occur during system operation, such as atmospheric turbulence, thermal effect, zero-gravity effect, and flight platform vibration and so on. With the demand for high resolution and multiple spectrum, the aperture of optics system becomes bigger and bigger. In order to overcome the limitation of mirror fabrication and launching, the segmented deployable primary mirror has been developed in large space optical system, which brings super large aperture and high resolution in theory, but also increases wavefront errors of system. So adaptive optics system should be used to detect and correct space environment and optical system errors online. In the paper, the influences of main error sources such as atmospheric turbulence, thermal affect, zero-gravity affect and flight platform vibration on imaging quality of remote sensing system are analyzed in theory, and the quantitative affecting results are put forward by computer simulation. Based on the conclusion of error analysis, the adaptive optical system scheme and working mode are presented, whose feasibility is proved by our adaptive optical emulation system.

High precision orbit predict data is not always available for large telescope system, which is designed for tracking space targets. The data from image procession is the only implementation to closed loop control. However, when target is going through thick cloud or low-contrast area, image procession might fail to offer reliable data to closed loop control. Large telescope system usually is built with high precision (arc seconds or several ten arc seconds) while with small field of view. A little time of failure from image procession will lead to escape of target. In this paper, a new way of real-time orbit prediction is presented to substitute traditional way. Real-time orbit prediction produces real-time prediction of high precision from real-time observation data which guides the driver module of large telescope system so that it will keep tracking even if temporary failure of image precession occurred and then high precision observation data is absent after acquisition. This theory has been validated by experiments in a large telescope system. The prediction keeps same order of observational data when failure of image precession occurred and then high precision observation data is absent for targets with orbit altitude 500km and 1000km, so that the system keepst tracking successfully. In this paper, the results of the experiments are analyzed, and the conclusion is that real-time orbit prediction is highly useful for large high precision telescope systems when tracking space target whose high precision orbit prediction is absent.

A new method to describe the position-stance of parallel supporting structure is proposed based on Rodrigues theory. Comparing with others methods, the kinematic model with Rodrigues parameters has the advantages including least computational parameters, no trigonometric function calculation and convenient real-time control. The model of the inverse kinematics is established and the inverse solutions of the position-stance are obtained by analyzing the topologic structure of the parallel supporting structure with 3-RPS limb. By analyzing the vectors of the manipulator, the velocity and acceleration models of moving platform, limbs and end-effector arw deduced. According to the designed kinematic track, it is convenient to control accurately the supporting structure by the inverse kinematic model of the mechanism.

The surface temperture of Open Large-Aperture Solar Telescope's primary mirror can vulnerablely be warmed too high by exposure to direct sunlight. If its temperature over the external environment is too high, it is likely to decline the mirror seeing. So generally keeping the surface temperture of primary mirror same as the environment's is very important by cooling technology researched. By using Ansys software to analyze a meniscus-shaped aluminum metal primary mirror with a diameter of 2.5m, and thickness of 4.45cm to gain the primary mirror's temperature distribution in order to verify the feasibility of methods of cooling. Furthermore form a more perfect cooling project. Thus it can provide a technical guidance for Large-Aperture primary mirror's development in the future.

The design and system performance of a far-infrared interferometer with which to test rough aspheric surfaces are presented. It is based on the optical configuration of classical Fizeau interferometer. A CO2 laser is used as a light source operating at a 10.6μm wavelength. A He-Ne laser is introduced into the interferometer to solve problems concerning with alignment. The measuring error(RMS) of the interferometer is less than λ/200 (λ=10.6μm), this accuracy is able to fulfill the need of grinding primary mirror.