An Empirical Prediction Model of the Performance Impacts of Material Tolerances in Frequency Selective Surfaces Using the Monte Carlo Method
Date of Award
Master of Science
Department of Electrical and Computer Engineering
Peter J. Collins, PhD
Standard computational tools predicting frequency selective surface (FSS) performance model periodic structures as an infinite-by-infinite array of perfectly placed elements with perfect, identical dimensions, and with dielectric layers of uniform thickness and material composition. These models do not address perturbations caused by manufacturing tolerances of elements and dielectric layers, nor do they address edge effects caused by finite-by-finite dimensional arrays. The Monte Carlo Method was used to determine the effects of random variations in element dimension, placement, dielectric thickness, and dielectric material on FSS performance. A full-factorial experimental design was applied, and eight hundred twenty-five unique finite arrays of elements were generated with dimensions that varied randomly within tolerances. Each array was analyzed using the AIM code (Adaptive Integral Method), and a sensitivity analysis was performed to determine the influence of design parameter tolerances on FSS performance. Dipole length was shown to have the most significant impact on FSS performance. Dielectric material constant, and the combination of dipole length and dipole placement in the "length" direction also had statistically significant impact on FSS performance. In addition, low tolerances for two design parameter combinations--dielectric thickness alone, and the combination of all five design parameters--produced significant rapid variation in output performance data.
DTIC Accession Number
Craig, Matthew D., "An Empirical Prediction Model of the Performance Impacts of Material Tolerances in Frequency Selective Surfaces Using the Monte Carlo Method" (1999). Theses and Dissertations. 5248.