Waveguide mode solution using a hybrid edge element approach

A hybrid edge element approach for the computation of waveguide modes is presented. The electric field is decomposed into its transverse and longitudinal components, which are modeled in terms of two-dimensional edge elements and scalar nodal elements, respectively, thereby satisfying the Dirichlet boundary condition at the perfect electric conductor boundaries and dielectric interfaces. Failure to do so results in the generation of spurious modes. This approach allows for the modeling of a three-dimensional field quantity over a two-dimensional boundary, namely the waveguide cross section. Another approach, the method of moments, serves as an excellent means of verifying the results obtained through use of hybrid edge elements. A comparison of the results obtained from both techniques are presented, along with the associated field plots obtained from the hybrid edge element approach for several geometries.     

Fictitious domain finite element methods using cut elements: I. A stabilized Lagrange multiplier method

We propose a fictitious domain method where the mesh is cut by the boundary. The primal solution is computed only up to the boundary; the solution itself is defined also by nodes outside the domain, but the weak finite element form only involves those parts of the elements that are located inside the domain. The multipliers are defined as being element-wise constant on the whole (including the extension) of the cut elements in the mesh defining the primal variable. Inf–sup stability is obtained by penalizing the jump of the multiplier over element faces. We consider the case of a polygonal domain with possibly curved boundaries. The method has optimal convergence properties.     

MHD flow past a circular cylinder using the immersed boundary method

The immersed boundary method (IB hereafter) is an efficient numerical methodology for treating purely hydrodynamic flows in geometrically complicated flow-domains. Recently Grigoriadis et als. [1] proposed an extension of the IB method that accounts for electromagnetic effects near non-conducting boundaries in magnetohydrodynamic (MHD) flows. The proposed extension (hereafter called MIB method) integrates naturally within the original IB concept and is suitable for magnetohydrodynamic (MHD) simulations of liquid metal flows. It is based on the proper definition of an externally applied current density field in order to satisfy the Maxwell equations in the presence of arbitrarily-shaped, non-conducting immersed boundaries. The efficiency of the proposed method is achieved by fast direct solutions of the two poisson equations for the hydrodynamic pressure and the electrostatic potential.The purpose of the present study is to establish the performance of the new MIB method in challenging configurations for which sufficient details are available in the literature. For this purpose, we have considered the classical MHD problem of a conducting fluid that is exposed to an external magnetic field while flowing across a circular cylinder with electrically insulated boundaries. Two- and three-dimensional, steady and unsteady, flow regimes were examined for Reynolds numbers Re_d ranging up to 200 based on the cylinder’s diameter. The intensity of the external magnetic field, as characterized by the magnetic interaction parameter N, varied from N=0 for the purely hydrodynamic cases up to N=5 for the MHD cases. For each simulation, a sufficiently fine Cartesian computational mesh was selected to ensure adequate resolution of the thin boundary layers developing due to the magnetic field, the so called Hartmann and sidewall layers. Results for a wide range of flow and magnetic field strength parameters show that the MIB method is capable of accurately reproducing integral parameters, such as the lift and drag coefficients, as well as the geometrical details of the recirculation zones. The results of the present study suggest that the proposed MIB methodology provides a powerful numerical tool for accurate MHD simulations, and that it can extend the applicability of existing Cartesian flow solvers as well as the range of computable MHD flows. Moreover, the new MIB method has been used to carrry out a series of accurate simulations allowing the determination of asymptotic laws for the lift and drag coefficients and the extent of the recirculation length as a function of the amplitude of the magnetic field. These results are reported herein.     

Error estimates for fictitious domain/penalty/finite element methods

We obtain error estimates for the finite element solution of elliptic problems with Neumann boundary conditions for domains with curved boundaries using fictitious domain/penalty methods.     

A new procedure for evaluating the time domain boundary influence matrix of unbounded media

The finite element method can only deal with finite domains with well defined boundaries. For dynamic problems involving unbounded media, the boundaries of the finite model distort the real physical behaviour of the problem if they remain untreated. For many problems it is possible to formulate so-called silent boundary conditions which perfectly simulate the effect of the truncated unbounded medium. Unfortunately, most of these conditions are properly formulated in the frequency domain.

The present paper introduces a new procedure which employs these frequency-dependent boundary conditions to calculate the time domain influence matrix of the truncated unbounded medium. This matrix, which was introduced in a previous publication, is used to calculate the reflection-free response of the truncation boundary one time step ahead of the present time station. The known boundary response is then used as a prescribed condition for the finite model. A one-dimensional example with a frequency-dependent boundary condition is presented to examine the effectiveness of the new procedure. Two other silent boundary conditions formulated directly in the time domain are also examined.     

Implementation of PEC boundary condition for Laguerre‐RPIM meshless method with novel uniform nodal distribution

The implementation of perfect electric conductor (PEC) boundary condition for meshless radial point interpolation method (RPIM) with the weighted decaying Laguerre polynomial (WLP) is investigated. A novel uniform nodal placement is proposed and the number of H‐nodes reduces to about a half of that for the conventional scheme. There are only E‐nodes scattered on the PEC boundary. The E‐node and H‐node have the same scale of support domain nodes. The image theory is adopted to update the implicit electric equations on the PEC boundary. This scheme is also suitable for nonuniform node distribution. Numerical results have verified the accuracy of the proposed method.     

层次式直接边界元计算VLSI三维互连电容

文中将Ａｐｐｅｌ处理多体问题的层次式算法思想实现于直接边界元法,用以计算ＶＬＳＩ三维互连寄生电容。直接边界积分方程同时含有边界上的电势与法向电场强度,能比间接边界元法更方便地处理多介质及有限介质结构,直接边界元法的层次式计算涉及对三种边界（强加边界、自然边界与介质交界面）及两种积分核（１／ｒ与１／ｒ＾３）的处理,显著区别于基于间接边界元法、仅处理强加边界与一种分核的层次式算法。文中以边界元的层次划     

MHD flow in a rectangular duct with a perturbed boundary

The unsteady magnetohydrodynamic (MHD) flow of a viscous, incompressible and electrically conducting fluid in a rectangular duct with a perturbed boundary, is investigated. A small boundary perturbation $\epsilon$ is applied on the upper wall of the duct which is encountered in the visualization of the blood flow in constricted arteries. The MHD equations which are coupled in the velocity and the induced magnetic field are solved with no-slip velocity conditions and by taking the side walls as insulated and the Hartmann walls as perfectly conducting. Both the domain boundary element method (DBEM) and the dual reciprocity boundary element method (DRBEM) are used in spatial discretization with a backward finite difference scheme for the time integration. These MHD equations are decoupled first into two transient convection–diffusion equations, and then into two modified Helmholtz equations by using suitable transformations. Then, the DBEM or DRBEM is used to transform these equations into equivalent integral equations by employing the fundamental solution of either steady-state convection–diffusion or modified Helmholtz equations. The DBEM and DRBEM results are presented and compared by equi-velocity and current lines at steady-state for several values of Hartmann number and the boundary perturbation parameter.     

Hybrid Spectral Difference Methods for Elliptic Equations on Exterior Domains with the Discrete Radial Absorbing Boundary Condition

The hybrid spectral difference methods (HSD) for the Laplace and Helmholtz equations in exterior domains are proposed. We consider the fictitious domain method with the absorbing boundary conditions (ABCs). The HSD method is a finite difference version of the hybridized Galerkin method, and it consists of two types of finite difference approximations; the cell finite difference and the interface finite difference. The fictitious domain is composed of two subregions; the Cartesian grid region and the boundary layer region in which the radial grid is imposed. The boundary layer region with the radial grid makes it easy to implement the discrete radial ABC. The discrete radial ABC is a discrete version of the Bayliss–Gunzburger–Turkel ABC without pertaining any radial derivatives. Numerical experiments confirming efficiency of our numerical scheme are provided.     

Application of CPML to truncate the open boundaries of cylindrical waveguides in 2.5-dimensional problems

1 Introduction Very strong nonlinear interactions can occur between the electron beam and electro-magnetic wave in the generation of HPM sources, and it is very expensive and requires very long time to design the HPM devices experimentally, so the computer simulation is often and widely used to design the HPM devices. The most widely used numerical technique is the PIC (particle-in-cell) method[1]. Based on this method, some full elec-tromagnetic PIC codes have been developed in the Unit…     

求解应力强度因子的样条虚边界元交替法

为求解平面裂纹问题的应力强度因子,提出基于Muskhelishvili基本解和样条虚边界元法的样条虚边界元交替法.该方法将平面内带裂纹有限域问题分解成带裂纹无限域问题与不带裂纹有限域问题的叠加.带裂纹无限域问题利用Muskhelishvili基本解法直接得出,不带裂纹有限域问题采用样条虚边界元法求解.利用该方法对复合型中心裂纹方板和I型偏心裂纹矩形板进行分析.数值结果表明该方法精度高且适用性强.     

A variational approximation method for 2nd order elliptic eigenvalue problems in a composite structure with nonstandard boundary conditions

In this paper we deal with the finite element analysis of a class of eigenvalue problems (EVPs) in a composite structure in the plane, consisting of rectangular subdomains which enclose an intermediate region. Nonlocal boundary conditions (BCs) of Robin type are imposed on the inner boundaries, i.e. on the interfaces of the respective subdomains with the intermediate region. On the eventual interfaces between two subdomains we impose discontinuous transition conditions (TCs). Finally, we have classical local BCs at the outer boundaries. Such problems are related to some heat transfer problems e.g. in a horizontal cross section of a wall enclosing an air cave.     

Numerical solution of static and dynamic equations of cables

The mechanical equations of an extensible, perfectly flexible curvilinear material (cable) are formulated. The static problem can be solved either by a minimization technique or by an iterative finite difference method which also permits dealing with forces that are not derived from a potential. In the dynamic alternative the solution of the equations is obtained from finite difference discretization explicit in time - this can apply to different kinds of constitutive laws and to different kinds of boundary conditions. Application of these numerical methods to electric transmission lines subject to electromagnetic forces shows a good agreement between computation and experimental data.     

Local absorbent boundary condition for non-linear hyperbolic problems with unknown Riemann invariants

Generally, in problems where the Riemann invariants (RI) are known (e.g. the flow in a shallow rectangular channel, the isentropic gas flow equations), the imposition of non-reflective boundary conditions is straightforward. In problems where Riemann invariants are unknown (e.g. the flow in non-rectangular channels, the stratified 2D shallow water flows) it is possible to impose that kind of conditions analyzing the projection of the Jacobians of advective flux functions onto normal directions of fictitious surfaces or boundaries. In this paper a general methodology for developing absorbing boundary conditions for non-linear hyperbolic advective–diffusive equations with unknown Riemann invariants is presented. The advantage of the method is that it is very easy to implement in a finite element code and is based on computing the advective flux functions (and their Jacobian projections), and then, imposing non-linear constraints via Lagrange multipliers. The application of the dynamic absorbing boundary conditions to typical wave propagation problems with unknown Riemann invariants, like non-linear Saint-Venant system of conservation laws for non-rectangular and non-prismatic 1D channels and stratified 1D/2D shallow water equations, is presented. Also, the new absorbent/dynamic condition can handle automatically the change of Jacobians structure when the flow regime changes from subcritical to supercritical and viceversa, or when recirculating zones are present in regions near fictitious walls.     

A fast immersed interface method for solving Stokes flows on irregular domains

We present a fast immersed interface method for solving the steady Stokes flows involving the rigid boundaries. The immersed rigid boundary is represented by a set of Lagrangian control points. In order to enforce the prescribed velocity at the rigid boundary, singular forces at the rigid boundary are applied on the fluid. The forces are related to the jumps in pressure and the jumps in the derivatives of both pressure and velocity, and are approximated using the cubic splines. The strength of singular forces is determined by solving a small system of equations via the GMRES method. The Stokes equations are discretized using finite difference method with the incorporation of jump conditions on a staggered Cartesian grid and solved by the conjugate gradient Uzawa-type method. Numerical results demonstrate the accuracy and ability of the proposed method to simulate Stokes flows on irregular domains.     

A comparison of finite-difference and finite-element methods for calculating free edge stresses in composites

This study considers the accuracy of the finite difference method in the solution of linear elasticity problems that involve either a stress discontinuity or a stress singularity. Solutions to three elasticity problems are discussed in detail: a semi-infinite plane subjected to a uniform load over a portion of its boundary; a bimetallic plate under uniform tensile stress; and a long, midplane symmetric, fiber-reinforced laminate subjected to uniform axial strain. Finite difference solutions to the three problems are compared with finite element solutions to corresponding problems. For the first problem a comparison with the exact solution is also made. The finite difference formulations for the three problems are based on second order finite difference formulas that provide for variable spacings in two perpendicular directions. Forward and backward difference formulas are used near boundaries where their use eliminates the need for fictitious grid points. Moreover, forward and backward finite difference formulas are used to enforced continuity of interlaminar stress components for the third problem. The study shows that the finite difference method employed in this investigation provides solutions to the three elasticity problems considered that are as accurate as the corresponding finite element solutions. Furthermore, the finite difference method appears to give a solution for the laminate problem that characterizes the stress distributions near an interface corner in a more realistic manner than the finite element method.     

A semi-analytical method for piezocomposite structures with arbitrary interfaces

A general three-dimensional semi-analytical solution of piezocomposite media consisting of various domains of distinct electro-mechanical properties is proposed. The interfaces between the domains may be non-planar. The geometry and boundary conditions may be arbitrary, and the applied loading can be in the form of traction, displacement, voltage, or any combination of them. In this approach, the unknown displacement field u₁, u₂, u₃, and the electric potential Φ are taken as appropriate functions in each domain, such that the continuity of the displacement field and the electric potential are exactly satisfied at the interfaces. Also, the continuities of traction stresses and normal electric displacement across the interfaces are resolved with high accuracy. The homogeneous and non-homogeneous kinematic boundary conditions associated with the electric potential and the displacement field are incorporated at the boundaries exactly.     

A level set-based topology optimization incorporating arbitrary Lagrangian Eulerian method for wavelength filter using extraordinary optical transmission

This paper deals with structural optimization for designing periodic structures in a hole array wavelength filter. The hole array wavelength filter that consists of metallic thin film and dielectric enables to transmit narrow bandwidth light. It is known that transmission spectrum can be changed not only by the periodicity of hole array but also by the shape of holes. For optimizing the hole shape, the level set method is used in this study. In the ordinary level set method, the boundaries are implicitly expressed by the zero level set of a scalar function, called the level set function, within a fixed mesh. Therefore, the material interpolation becomes numerically awkward within the elements across the implicit zero level set because those elements inevitably take on intermediate material properties even if the boundary of zero level set are mathematically clear. As the result, the optimization is likely to yield wrong solution. Here, a new level set optimization method incorporating Arbitrary Lagrangian Eulerian method is proposed to eliminate intermediate values on the interfaces perfectly. As a result, the proposed method can successfully perform the structural optimization of hole shape without intermediate values.     

Stabilized interior penalty methods for the time-harmonic Maxwell equations

We propose stabilized interior penalty discontinuous Galerkin methods for the indefinite time-harmonic Maxwell system. The methods are based on a mixed formulation of the boundary value problem chosen to provide control on the divergence of the electric field. We prove optimal error estimates for the methods in the special case of smooth coefficients and perfectly conducting boundary using a duality approach.