Pattern evaluation for in-plane displacement measurement of thin films

Abstract

Ultra-lightweight spacecraft incorporating “gossamer” structures are extremely compliant, which complicates control, design, and ground testing in full scale. One approach to model the behavior of a full-scale gossamer structure is to construct a small-scale model test article that can be used to verify a corresponding small-scale computer model. Once the predictions of the computer model have been verified by measurement of the physical test article, it can be scaled up to allow computation of the full-scale structure behavior. As model verification requires accurate deflection measurements at multiple points along the surface of the structure, a sensing system that provides full-field data without changing the dynamic response of the structure must be developed. Hence, an optical approach is taken. Since the thin films used in gossamer space structures are typically smooth and featureless, targets must be incorporated into the film surface to enable tracking of both in-plane and out-of-plane displacements. A krypton fluoride excimer laser system was used to etch 35 μm wide linear features approximately 0.1 μm into the surface metallization of both 50.8 μm polyester and 127 μm polyimide films. These optically diffuse surface features, designed mainly to investigate the precision of the laser etching method, were used as targets for ultra-close-range photogrammetry, the method chosen for displacement tracking. A force applied to the surface of the etched mirror (test article) produced in-plane and out-of-plane deformations that were resolved via ultra-close-range photogrammetry. To measure the in-plane tracking resolution, 1.5 and 3.0 mm circular dots were added (using ink) to the surface of the thin film, and some of these targets were tracked as the test article was translated on a precision linear stage. In-plane tracking resolution using ultra-close-range photogrammetry was related to the ground sample distance of the camera, which in this case was 51.25 μm pixel−1 (equal to the ratio of sample dimension to number of pixels in the field of view). Using a manual technique to identify features of the etched pattern for tracking, the mean tracking error was about 13 μm (σ=43 μm). Using an automated, subpixel marking technique to identify the 1.5 mm circular targets, the mean tracking error was 22 μm (σ=13 μm). Neither of these methods achieved the desired 10 μm tracking resolution.