Massachusetts Institute of Technology
MIT Space Systems Lab


 




  About DMSE
Academics
Faculty
Research
News/Events
Home

TMT

 

TMT Project Report

MIT Space Systems Lab

 

Scott A. Uebelhart and Deborah J. Howell

July 6, 2005

 

Tradespace Analysis of the Thirty Meter Telescope (TMT)

The goals of this contract were to assist AURA (Association of Universities for Research in Astronomy) in the selection of several basic design variables as the first step in their design of a large, ground-based telescope with a 30-meter segmented aperture.  The MIT SSL was brought aboard due to our experience in producing integrated disturbance-to-performance models of space telescopes (which share many of the flexibility problems of such a large ground-based telescope), and our experience in trade study analyses.  This project was supported by AURA under contract C10413A.  The official title was Development of Methods and Software Tools for Analyzing Integrated Computer Models of Extremely Large Ground Based Telescopes with Dr. George Angeli as the Technical Monitor.  The MIT team consisted of Prof. David Miller, Prof. Olivier de Weck, Scott Uebelhart, Deborah Howell, Soon-Jo Chung, and Julien Lamamy.

Four design variables were identified as important in the conceptual design phase:  primary mirror focal ratio, final focal ratio, configuration of the secondary mirror (Cassegrain or Gregorian), and position of the elevation axis.  In order to examine the effect of changing these four structural design variables, it was necessary to build a parameterized model of a TMT-like telescope so that each variable could be changed and a new realization of the structure created.  Much of the early work involved using a combination of MATLAB, DOCS (an integrated modeling tool) and the finite element processor NASTRAN to create automated finite element representations of the telescope that could be produced and run for any set of design variable inputs.  The number of primary segment rafts, the height of the tower holding the secondary mirror, the truss structure supporting the primary mirror in addition to the aforementioned design variables were all made variable.  The overall architecture of this parameterized model was based on the Strawman “point-design” model provided by AURA.  An optical model, provided by AURA, was appended to this structural model in order to evaluate the optical performance of each instantiation of the telescope.  Figure 1 shows an example of two different realizations of the structure.

Figure 1 : Two realizations of the Thirty Meter Telescope automatic finite element model

Once the parametric model was assembled, the tradespace, consisting of the four critical design variables, was explored.  This provided insights into the system which are not immediately intuitive such as coupling between variables and areas of the tradespace warranting further investigation.  First, a static tradespace analysis was performed whereby the effect of a 650 elevation angle change on mirror distortion under gravity was investigated.  The elevation angle is a rotation about an axis perpendicular to the pointing direction.  Figure 2 is an example static analysis results plot comparing the change in image quality (between 00 and 650 elevation angle) vs. primary mirror (M1) focal ratio for Gregorian configuration with the elevation axis above the primary mirror.  The different colored lines represent different values of final focal ratio.

Figure 2:  TMT static analysis results

 

Next, a dynamic analysis was performed whereby sensor and actuator models, wind models (provided by AURA) and a simple closed-loop optical control system were appended to the parametric model in order to evaluate the telescope’s performance under normal wind-loading conditions. Much effort was required to model the actuators and sensor, and to ensure that the final closed loop system was stable.  Figure 3 is an example of the results of the dynamic trade space analysis relating two output variables, cost and mass.  The dynamic results were used to select the design variables which best impacted optical performance, mass and cost.  The final analysis showed the usefulness of parameterized models to AURA.

 

Figure 3: TMT dynamic trade space analysis results