On the Algorithm of Accelerated Convergence for Waveguide Discontinuities

刘发林,徐善驾
DOI: https://doi.org/10.3969/j.issn.0253-2778.2001.04.017
2001-01-01
Journal of University of Science and Technology of China
Abstract:The Mode Matching method is an effective and simple technique to deal with waveguide discontinuities. However, it suffers from slow convergence and a lot of modes ar e needed as basis functions to guarantee effective convergence. The main objective of this study i s to accelerate convergence by use of new basis functions and to propose an algo rithm of acceler ated convergence. Instead of the conventional sinusoidal basis functions, a group of new basis fun ctions are used to deal with waveguide discontinuity problems. Taking into consideration the electr omagnetic edge conditions in the proximity of metal wedges, the modified trigonometric function s are constructed as basis functions to expand the electromagnetic fields at the discontinuities. By imposing open and short condition on the longitudinally symmetric plane of the discontinuity and m atching the tangential fields, a functional expression of the input admittance can be obtain ed. Then the Ritz method is adopted to construct a group of linear equations. And the input admitt ance can be calculated by solving the equations. The waveguide bifurcation structure is used to verify the proposed algorithm. As a result, a few basis functions are enough to well descri be the behavior of the discontinuity. Very good numerical convergence is observed. Furthermore, the zero thickness waveguide inductive membrane is analyzed to show the sharp contrast of the accel eration effect. The sharper the discontinuity, the more obvious the acceleration performance. Examples are also discussed by using different new basis functions. It is found that multiple choi ces of the basis functions are possible within certain constraints. And the good performances of accelerated convergence are also observed. Unlike the Coupled Integral Equation Technique (CIET), the order of the linear equation is the same as the number of basis functions in the field ex pansion for one discontinuity and this keeps the equation at the lowest order to facilitate the numerical solution.
What problem does this paper attempt to address?