一种系统构造三角形杂交边缘元空间的新方法

近年来发展起来的杂交边缘元是一种消除伪解的有效方法。本文提出了一种能系统构造三角形杂交边缘元空间的方法。作为例子具体给出了一次和二次杂交边缘元插值函数的显形表达式，并且用它们分别计算了空波导，条形介质波导和有耗块状介质填充波导的传播常数。计算表明这种方法不仅能消除伪解，而且具有较高的计算效率，两种杂交边缘元计算结果的比较表明高次杂交边缘元的计算精度和收敛速率较低次元有明显的改进。     

一种新的高次边缘元方法

讨论了一种新的高次边缘元方法，它不仅消除了伪解，并具有简单高效的特点．着重研究了二次边缘元的空间构造；给出了有关计算公式．对具有精确解的条形介质填充波导主模和高次模本征值的计算表明：二次边缘元方法的计算精度比一次边缘元提高约一个数量级．块状介质填充和衬底各向异性的微带线主模色散曲线的计算证实了该方法的精确性和广泛的适用性．     

导波结构三维不连续性问题的高次六面体边缘元分析

该文从全磁场矢量泛函出发,讨论了一种54参量六面体边缘元的空间构造。这种高次三维边缘元方法不但有效地消除了伪解,而且具有很高的计算精度。用这种方法对导波结构三维不连续性散射问题的分析,证实了它的有效性和可靠性。与12参量六面体边缘元计算结果的比较表明:本文方法具有更高的精度和计算效率,是一种求解三维不连续性问题的高效数值方法,有着推广应用的实际价值。     

涂层圆柱爬行波传播常数和模式阻抗的分析与计算

本文分析了介质涂层圆柱爬行波的电磁特性。找到了一种求解爬行波传播常数和模式阻抗的有效方法－微扰－迭代法，这种方法简单、快速，便于工程应用。最后给出公式并计算了一些实例。     

周期性结构的边缘元分析

本文用边缘元、模匹配以及多模网络相结合的方法分析周期性结构。计算了ＮＲＤ波导周期结构和介质镜象线周期结合两个具体实例。数值结果表明这种方法用于分析周期性结构不仅是有效的，而且上仍很高的计算精度和效率。     

三层介质光纤包层模色散方程及包层HE／EH模

利用解析法研究了Tsao和Erdogan关于3层介质光纤模型的包层模色散方程，对2个色散方程的等价性给出了证明。在研究长周期光纤光栅(LPFG)中，对传播常数的计算应选用Tsao给出的方程，在计算与场方程有关的量时应选用Erdogan的方程。计算了部分包层模的传播常数，给出了区分包层HE／EH模的方法。     

一种分析微带交指滤波器的新思路

本文提出一种直接用耦合多导体多介质结构传输特性计算微波滤波器的方法，它不但考虑了相邻导体，也考虑了所有非相邻导体间的耦合，因此计算精度大大提高。此外，它还可以计算不对称耦合，适应面宽。分区均匀介质填充多导体结构有多个不同传播常数的本征模。这与均匀填充多导体结构有很大差别，即不但要考虑本征阻抗，而且要考虑不同的传播常数。本文用矩阵电报方程求出它们的本征值与本征矢量，从而求得传输特性，最后用网络等效方法优化求得滤波器特性。本文还介绍了由不同宽度谐振腔组成的滤波器的设计     

8-18GHz介质板极化器的分析与设计

介质板极化器在天线设计中的应用非常广泛。从介质板波导形成圆极化波的原理出发,推导出差相移与波导传播常数的关系式。因此只需利用数值方法计算波导的传播常数,从而减小了仿真计算量。分析表明在818GHz超宽带范围内,方波导中插入介质板的圆极化轴比理论上只能达到2．03dB。文中提出介质板插入结合波导口径变形的方法使圆极化轴比小于1．53dB,驻波比小于1．23。对介质板波导的关键参数进行误差分析,得到加工误差对圆极化性能的影响小于2．5％。     

手征等离子本填充圆波导的传播特性

本文在柱坐标系中导出了手征等离子体波导内横向场量与纵向场量之间的关系，给出从向场量的波动方程，利用波导壁上的电场边界条件，建立了手征等离子体介质填充的圆波导传播常数的特征方程，计算得到了几种低次模式的传播常数随归一化波导半径及材料参数的一系列变化曲线，分析得出了若干有意义的结论。     

镜象和非镜象型喇叭NRD漏波天线的研究

为了克服一般无辐射介质波导漏波天线(简称 NRD 漏波天线)在 xoz 平面上有两个对称大瓣的缺点,也为了提高天线的增益,作者曾提出在天线口面上加喇叭的方法,称为喇叭 NRD 漏波天线。本文从理论上解出了当喇叭张角较小时(不大于50度)喇叭 NRD 漏波天线的复传播常数,从而可以设计天线。文中给出了根据理论结果计算的相位常,数β及衰减常数α随喇叭张角θ_0及介质条位置 d 的关系曲线。理论计算数据及实验结果符合良好。本文并提出了镜象型喇叭 NRD 漏波天线的方案,使天线的面积及体积大大减小,也解决了介质条的安装及固定的问题,将这种天线向实用化推进了一大步。同样也求出了镜象型喇叭 NRD 漏波天线的复传播常数,并得到了实验验证。     

An efficient high-order mixed-edge rectangular-element method forlossy anisotropic dielectric waveguides

A new efficient high-order mixed-edge rectangular element method is proposed for the analysis of lossy anisotropic dielectric waveguides. The space construction of the high-order mixed-edge rectangular element is investigated and the explicit form of the shape function is given. The high-order mixed-edge element yields higher accuracy and faster convergence than the lowest order mixed-edge rectangular elements without spurious solutions, and is more efficient compared to the high-order covariant projection element. The computations of the propagation constants in the rectangular waveguide and the slab loaded waveguide show that the accuracy of this high-order mixed-edge element is about one order higher than that of the lowest order one, and the nodes used in the calculation are only two-thirds as many as those used in the high-order covariant projection element having the same accuracy. The calculations of the dispersion curves for the dominant mode in the waveguide loaded with the lossy anisotropic dielectric block verify the accuracy and efficiency of the present method     

A comparison of formulations for the vector finite element analysisof waveguides

The principal formulations that have been proposed for finding the modes of waveguides by the finite element method are reviewed and compared. In each case, it is shown how Maxwell's equations may be reduced to matrix form using the method of weighted residuals. The formulations are compared from several points of view: their ability to handle spurious modes, lossy materials, and reentrant corners; the number of field components; and the properties of the matrices. Three benchmark problems are described and used to compare the formulations: a rectangular waveguide partially loaded with lossless dielectric; an air-filled, double-ridged waveguide; and a shielded image guide with either lossless or lossy dielectric     

A multipole analysis of a dielectric loaded coaxial rectangularwaveguide

A multipole analysis of a coaxial rectangular waveguide whose inner conductor is circular is made in order to determine the TE and TM modes of the system. The analysis is based on using multipole (dipole, quadrupole etc.) electric and magnetic current sources to generate field solutions in the waveguide. These solutions are used to match the electromagnetic boundary condition in a homogeneous coaxial rectangular waveguide and to determine the TE and TM eigenvalues of the waveguide system. Eigenvalue results are compared with results of the generalized spectral domain method and to eigenvalue results for a ridged waveguide. Propagation in a coaxial rectangular waveguide is also studied when the coaxial rectangular waveguide is loaded with lossy inhomogeneous dielectric material. A variational formula is used to relate the TEM, TE, and TM modes of an empty coaxial rectangular waveguide to the propagation in the loaded inhomogeneous dielectric waveguide     

Microwave transmission and reflection of stratified lossy dielectric segments partially filling waveguide

The microwave transmission and reflection is evaluated for stratified lossy dielectric segments partially filling the rectangular waveguide by the method which combines the multimode network theory with the rigorous mode matching procedure. As an example, we investigate in detail the microwave scattering properties of II-VI-epitaxial layer on a lossy dielectric substrate inserted in the rectangular waveguide. The experimental data verify the accuracy and the effectiveness of the present method. Extensive numerical results are presented to establish useful guidelines for the contactless microwave measurement of the conductivity of the epitaxial layer.     

The resonance method for evaluating the impedances of theequivalent network for dielectric posts in a rectangular waveguide

The resonance method is applied to evaluate the impedances of the equivalent network for dielectric posts symmetrically located about a reference plane in a rectangular waveguide. Explicit relationships for the impedances of the equivalent lattice network of the posts in a rectangular waveguide are given in terms of the resonant frequencies of the rectangular cavity constructed by short-circuiting the waveguide at equal distances from the reference plane. A perturbational analysis is carried out to determine the resonant frequencies for a rectangular cavity loaded with slightly lossy circular posts whose largest cross-sectional dimensions are small compared with the wavelength. Sample numerical results for a single loss-free dielectric post illustrating the application of the derived expressions are given     

Numerical analysis of H-plane waveguide junctions bycombination of finite and boundary elements

A numerical method is formulated for the analysis of H-plane waveguide junctions with arbitrary cross sections. The junctions are loaded with arbitrarily shaped dielectric or ferrite. The method is a combination of the finite-element method and the boundary-element method, and is applied to the regions with and without dielectric or ferrite, respectively. To show the validity and usefulness of the method, a lossy dielectric post and a ferrite slab in a rectangular waveguide are investigated in detail, and the computed results are compared with earlier theoretical and experimental results     

A Variational Theory for Wave Propagation in Inhomogeneous Dielectric Slab Loaded Waveguides

A novel numerical technique based on the variational formulation defined only in the slab is developed to study the loaded rectangular waveguide with an inhomogeneous dielectric slab. The variational equation for the boundary value problem is formulated and solved numerically, using the finite element method with piecewise quadratic trial functions. A comparison of this new technique with the conventional variational ones is presented. Various propagation characteristics, such as the phase constant, useful bandwidth, power handling capacity, and attenuation constants due to conductor and dielectric losses, are investigated for the waveguide centrally loaded with a slab of parabolic dielectric profile. The effects of changes in dielectric profiles are discussed by examining the results for the slabs with constant and parabolic profiles.     

Propagation in Dielectric Slab Loaded Rectangular Waveguide

Propagation in dielectric loaded rectangular waveguide is investigated theoretically for varying slab thickness and dielectric constant. The slabs are placed across the center of the waveguide in the E plane. This geometry is found to offer bandwidths in excess of double that of rectangular waveguide for dielectrics having dielectric constants of approximately 18. Power handling capacities which are double or triple that of standard waveguide are achievable using the dielectric loaded waveguide. In addition to the theory, design curves of bandwidth, guide wavelength, cutoff wavelength, impedance, power handling capacity, wall losses, and dielectric losses are presented and compared to experiment where possible.     

Application of a Variation-Iteration Method to Waveguides with Inhomogeneous Lossy Loads

An approximate technique for solving eigenvalue equations, the variation-iteration method,is commonly used in theoretics physics. A previous paper presented the application of this method to the scalar case of a dielectric slab loaded rectangular waveguide. This paper presents its extension to the complex vector case of a lossy dielectric insert loaded waveguide. Starting from an initial trial function, iterates are calculated in which the components relative to the unwanted eigenfunctions are eliminated. Both an upper and a lower bound for the unknown eigenvalues are available. Each iterate is the solution of a system of algebraic simultaneous equations. This system is solved by the successive overrelaxation method using an automatically computed optimal accelerating factor. An extrapolation technique further accelerates the convergence. This yields the attenuation and propagation coefficients for the dominant as well as several other modes, together with the electric and magnetic field configurations.