Dynamic tailoring and tuning for space-based precision optical structures

Abstract

Next-generation space telescopes in NASA's Origins missions require use of advanced imaging techniques to achieve high optical performance with limited launch mass. Structurally-connected Michelson interferometers meet these demands, but pose specific challenges in the areas of system dynamics and controls, uncertainty management and testing. The telescope optics must meet stringent positional tolerances in the presence of environmental and on-board disturbances, resulting in heavy demands on structural dynamics and control. In addition, fully integrated system tests are cost-prohibitive due to the size and flexibility of the system coupled with the severe differences between the on-orbit and ground testing environments. As a result, the success of these missions relies heavily on the accuracy of the structural and control models used to predict system performance. In this thesis, dynamic tailoring and tuning are applied to the design of precision optical space structures to meet aggressive performance requirements in the presence of parametric model uncertainty. Tailoring refers to changes made to the system during the design, and tuning refers to adjustments on the physical hardware. Design optimizations aimed at improving both performance and robustness are considered for application to this problem. It is shown that when uncertainty is high and performance requirements are aggressive, existing robust design techniques do not always guarantee mission success. Therefore, dynamic tuning is considered to take advantage of the accuracy of hardware performance data to guide system adjustments to meet requirements.

(cont.) A range of hardware tuning techniques for practical implementation are presented, and a hybrid model updating and tuning methodology using isoperformance analysis is developed. It is shown that dynamic tuning can enhance the performance of a system designed under high levels of uncertainty. Therefore, robust design is extended to include tuning elements that allow for uncertainty compensation after the structure is built. The new methodology, Robust Performance Tailoring for Tuning creates a design that is both robust to uncertainty and has significant tuning authority to allow for hardware adjustments. The design methodology is particularly well-suited for high-performance, high-risk missions and improves existing levels of mission confidence in the absence of a fully integrated system test prior to launch. In the early stages of the mission the design is tailored for performance, robustness and tuning authority. The incorporation of carefully chosen tuning elements guarantees that, given an accurate uncertainty model, the physical structure is tunable so that system performance can be brought within requirements. It is shown that tailoring for tuning further extends the level of parametric uncertainty that can be tolerated at a given performance requirement beyond that of sequential tailoring and tuning, and is the only design methodology considered that is consistently successful for all simulated hardware realizations.