Optical Tolerance Analysis for the STEAMR Instrument
Paper in proceeding, 2013
The optics of the STEAMR instrument is a complex system involving off-axis mirrors designed to achieve precise imaging of the 14 receiver channel beams from the far field to the corresponding feed horns. An initial optical design was generated by Swedish Space Corporation, which laid the framework for the subsequent IAP design that further developed the optical system to meet the mission performance requirements. Omnisys Instruments is now the prime contractor for the complete STEAMR instrument.
Although simulations of the optics presently show good results, little is known about the sensitivity to mechanical errors, i.e. surface deviations and misalignments of the reflectors. This work encompasses a tolerance analysis for the complete optics chain consisting of a 28 reflector focal plane array (FPA) and 6 reflector relay optics. With six degrees of freedom for each reflector, the scale of the required mechanical tolerancing analysis is significant. The goal of this work is therefore to identify critical locations within the optics architecture that have the largest influence on performance.
Being a multi-beam instrument, the optics requirements for STEAMR can be divided into two parts: pointing and beam quality. Pointing errors were analysed using the commercial software package ZEMAX, which offers built-in routines for performing Monte-Carlo simulations specifically for tolerancing problems. Beam quality, i.e. sidelobe levels, beam efficiency and polarization plane, were analysed using physical optics routines in GRASP. Simulations in both programs have been carried out using single element perturbation and Monte-Carlo simulations on the complete optics chain.
In the first iteration of the analysis, all reflector surfaces were assumed to be perfect. In later analyses, surface errors were also added. Special attention was given to the 1.6 m x 0.8 m carbon fiber main reflector, which is the most sensitive in terms of errors in shape.
By running the optical analysis in parallel and in close cooperation with the mechanical design, it has been possible to assume realistic errors for the different parts of the optics. Measurements of the feed horns done by IAP show excellent agreement with simulations, where sidelobe levels around -40 dB was predicted. Therefore, the beams of all feeds have been modelled as perfect Gaussians.