| dc.description.abstract | Space-based reflector antennas are susceptible to a variety of conditions that can result in surface deviation that affect the antenna’s performance. This is especially true for small-satellite antennas, where the reflector surface is generally fabricated from flexible membrane materials that are folded for stowage during launch. In this work, various surface deviations are generated and simulated in an effort to characterize how differing surface deformations affect antenna performance metrics (i.e. maximum gain, relative side lobe levels, and beam width). The study utilizes a new surface generation tool to generate surfaces with both random and systematic errors. These surfaces are subsequently used in a commercial Finite Element Analysis (FEA) software to simulate the antennas. From the FEA results, antennas with larger reflector surfaces are shown to be less sensitive to surface errors, especially when the errors are random with low correlation length. Antenna performance is shown to be generally more sensitive to surface errors with larger correlation lengths. The surface Root Mean Square (RMS) is often used by a variety of analysis techniques to predict reflector antenna performance. One such method is Ruze’s antenna tolerance theory. The FEA results are compared to Ruze’s relatively simple analysis technique, and Ruze’s method is shown to over predict gain losses, under predict side lobe level increases for surface errors with large correlation length, and is completely incapable of capturing any changes to half-power beam width. Finally, the FEA results illustrate that the metric of RMS alone is not sufficient to characterize expected changes in the radiation pattern. The results show that for the various surface error types and correlation lengths, the predicted values could be off by as much as 1 dB for gain loss, 1.5 dB for sidelobe levels, and 0.5o for RMS values as small as λ/70. | |
