Frequency selective surfaces with lumped and time-varying loads, variable surface impedance, and multiple screens

Current numerical methods for the spectral domain analysis of thin, planar periodic structures can be applied to a limited class of surface geometries. Since the frequency response of a periodic structure is dependent upon the geometry of the surface, a typical approach to changing the response is to modify the geometry. In the past this meant choosing the element shape, such as a square path or cross, and adjusting the dimensions to give the proper resonant frequencies. This thesis adds the additional control parameters of variable surface impedance, lumped loading, nonlinear lumped loads, and multiple layers which are capable of generating entirely new response curves. This thesis, therefore, expands the realizable frequency responses of periodic structures and the ability to design and produce desired response characteristics.This thesis first studies the effects of loading a periodic structure with variable surface impedance and lumped loads. By starting with the resistive boundary condition, this problem is first formulated such that a general surface impedance can be given as a function of the unit cell discretization. This procedure leads to the formulation of a surface impedance matrix which modifies the method of moments impedance matrix.By defining new basis functions in conjunction with the general surface impedance formulation, the addition of lumped loading could also be accounted for in the surface impedance matrix. For more general nonlinear loads, this thesis also shows a unique development combining the frequency domain data of a periodic structure with time-varying nonlinear loads. The formulation developed uses an equivalent circuit to simplify the problem of finding the unknown load currents.For handling multiple screen systems a method is developed which bypasses the need for the generation and storage of the scattering matrices previously used in their analysis. Unlike the scattering matrix methods whose accuracies are dependent on the number of modes stored in the matrix, it requires no a priori estimate of the number of dominant Floquet harmonics. It is also capable of solving systems with small separation distances.U of I OnlyETDs are only available to UIUC Users without author permissio