The Thirty Meter Telescope (TMT) project, a partnership between ACURA, Caltech, and the University of California, is Currently developing a 30-meter diameter optical telescope. The primary mirror will be composed of 492 low expansion glass segments. Each segment is hexagonal, nominally measuring 1.44m across the corners. Because the TMT primary mirror is Curved (i.e. not flat) and segmented With uniform 2.5mm nominal gaps, the resulting hexagonal segment outlines cannot all be identical. All segmentation approaches Studied result in some combination of shape and size variations. These variations range from fractions of a millimeter to several millimeters. Segmentation schemes for the TMT primary mirror are described in some detail. Various segmentation approaches are considered, with the goal being to minimize various measures of shape variation between segments, thereby reducing overall design complexity and cost. Two radial scaling formulations are evaluated for their effectiveness at achieving these goals. Optimal tuning of these formulations and detailed statistics of the resulting segment shapes are provided. Finally, we present the rationale used for selecting the preferred segmentation approach for TMT.

This paper presents refinements to the design of the TMT primary mirror segment passive-support system that are effective in reducing gravity print-through and thermal distortion effects. First, a novel analytical method is presented for tuning the axial and lateral support systems in a manner that results in improved optical performance when subject to varying gravity fields. The method utilizes counterweights attached to the whiffletrees to cancel astigmatic and comatic errors normally resulting when the lateral support system resists transverse loads induced by gravity. Secondly, several central diaphragm designs are presented and analyzed to assess lateral-gravity and thermal distortion performance: 1) a simple flat diaphragm, 2) a stress-relieving diaphragm having a slotted outer rim and a circumferential convolution near the outside diameter, and 3) a flat diaphragm having a slotted outer rim. The latter design is chosen based on results from analytical studies which show it to have better overall optical performance in the presence of gravity and thermal environments.

This paper describes the studies performed to establish a baseline conceptual design of the Segment Support Assembly (SSA) for the Thirty Meter Telescope (TMT) primary mirror. The SSA uses a combination of mechanical whiffletrees for axial support, a central diaphragm for lateral support, and a whiffletree-based remote-controlled warping harness for surface figure corrections. Axial support whiffletrees are numerically optimized to minimize the resulting gravityinduced deformation. Although a classical central diaphragm solution was eventually adopted, several lateral support concepts are considered. Warping harness systems are analyzed and optimized for their effectiveness at correcting second and third order optical aberrations. Thermal deformations of the optical surface are systematically analyzed using finite element analysis. Worst-case performance of the complete system as a result of gravity loading and temperature variations is analyzed as a function of zenith angle using an integrated finite element model.

Because of the large difference in coefficient of thermal expansion (CTE) between the lead converters and the silicon detectors in the SuperGLAST trays, failure levels are being predicted for the silicon detectors. This CTE mismatch is only further compounded by the substantial thickness of the lead converters required for the SuperGLAST trays. Solutions ranged from building lead/carbon composites, which lower the thermal growth of the lead converters, to some how decoupling the converters from the bias sheet through compliant adhesives. By replacing the epoxy adhesives used in bonding the converter layer to the kapton bias sheet with more compliant adhesives, the CTE interaction between the converters and the adjacent bias sheet would be reduced; thus, lowering the stress state in the silicon detectors to an acceptable level.

A crude, system level dynamic model of SNAP is created based on the results of initial structural design studies for some of the key telescope structures. The model is used to evaluate coupled dynamics of the SNAP spacecraft with both fixed base and free-free boundary conditions to mimic launch and on-orbit situations. The model is then used to eva luate the magnitudes of on -orbit telescope deflections induced by residual imbalances in spinning reaction wheels. It is shown that, when reaction wheel speeds are below fundamental resonances, the induced jitter is negligible or well below requirements. However, it is also found that if reaction wheel speeds are allowed to overlap structural resonances, angular mirror motions exceed requirements unless very high damping levels are present.

Optimal placement of tuning masses, actuators and other peripherals on large space structures is a combinatorial optimization problem. This paper surveys several techniques for solving this problem. The genetic algorithm approach to the solution of the placement problem is described in detail. An example of minimizing the difference between the two lowest frequencies of a laboratory truss by adding tuning masses is used for demonstrating some of the advantages of genetic algorithms. The relative efficiencies of different codings are compared using the results of a large number of optimization runs.

The equations of motion of actively controlled structures are usually reduced to first order form, a procedure which obscures symmetries that are present in the second order form of these equations. The loss of symmetry requires the calculation of both left and right eigenvectors for obtaining derivatives of the stability eigenvalues of the system. The paper shows that, for some control laws, only right eigenvectors are required if derivatives are obtained from the second order form of the equations of motion. The paper also examines reduced basis approximations for the eigenvalues and their derivatives. It is shown that including Ritz vectors representing the effect of local actuation forces in the reduced basis improves the accuracy of the eigenvalues and eigenvalue derivatives. Two active truss examples are used for demonstrating the improved accuracy.

This notes summarizes performance predictions of the current designs for the HAM seismic isolation systems. Three configurations using the baseline Viton rubber spring and two damped metal spring designs (multi-layer coil spring and leaf spring) are considered. Vertical and horizontal transmissibilities are evaluated as well as residual mirror motions due to floor seismic noise in the horizontal and vertical directions.

Worst case BSC support beam bellows deflections in axial and shear directions are calculated by combining actuation ranges (coarse and fine), deflections of facility floor and vacuum chamber due to vacuum load imbalance, and manufacturing and positioning tolerances of the chamber and support structure. The shear and axial bellows deformations obtained from this calculation are large and define critical selection/design requirements for the support beam bellows.

Support structures for the secondary mirror assembly in SNAP require high stiffness, ultra-high dimensional stability, and minimum obstruction of the telescope apertu re. This note summarizes a conceptual design trade study that explores various structural concepts in terms of their relative stiffness, obscuration and diffraction merit, as well as technical complexity and risk. The effects of variations in two key and presently unsettled design parameters (mass of secondary mirror assembly, and primary to secondary mirror separation) are also explored for each option through systematic 2-dimensional trend analyses.

This note summarizes conceptual design calculations for various approaches to build "low-effective-CTE" lead sheets. The options examined are sandwiches of pure lead and graphite fiber composites, either co -cured or secondary bonded, bonded Invar-lead sandwiches, and lead matrix graphite fiber composites. Sizing calculations are summarized and the various options are compared in terms of final effective CTE, total thickness, RL, and mass, and manufacturing considerations.

This report summarizes the results from mechanical tests conducted by Senior Flexonics on a prototype BSC bellows. The tests were intended to provide data on spring rates, life, leak rates, and behavior under axial twist. The test report from Senior Flexonics is attached.