A study of the interaction of the electromagnetic field with slotted cylindrical structures employing the Method of Regularization

The research program described in this thesis analyzes the interaction of the electromagnetic field with several classes of open cylindrical structures, by using the semi-analytical Method of Regularization (MoR). The dielectric cylinders considered in this thesis are partially shielded by conformal perfectly electric conducting (PEC) strips, where significant coupling and re-radiation of energy are created by the presence of apertures and sharp edges. The problems studied include the scattering problem of a single cylindrical lens reflector (CLR) illuminated by an obliquely incident plane wave, the scattering problem of a finite array of CLR with different characteristics when illuminated by a normal plane wave, the analysis of the scattering from and penetration through a multi-layered CLR and a multiconductor cylinder, as well as the transmission line problem involving a multi-conductor cable. Each of the problems studied is interesting from both a theoretical point of view and as an idealization of scattering and coupling mechanism in real devices of technological interest. -- When the structures are of moderate or large electrical size, standard numerical approaches to solving these mixed boundary valued problems (MBVP) often encounter difficulties of convergence and accuracy of computed solution. Therefore, the MoR - which transforms the ill-posed nature of the standard formulation of the problem to a well-conditioned second kind Fredholm matrix equation - is well-suited for the class of problems considered here. Numerical algorithms based upon the solution of the matrix equation, after truncation to a finite system of Ntr equations, converge with guaranteed and predictable accuracy, as Ntr -α. Because the computed solutions to these problems are rigorously accurate (in the sense of guaranteed convergence - theoretically and numerically), they provide benchmark solutions to problems of significant complexity against which solutions computed by more general purpose numerical codes (which although of wider applicability have less firm theoretical underpinnings) may be validated.