随机表面散射光场的格林函数法与基尔霍夫近似的比较

从简谐光波满足的亥姆霍兹方程出发，将由格林定理得到的介质分界面上的积分方程转化为以表面上的光波及其导数为未知量的线性方程组，并对其进行数值求解，实现了光场的数值计算. 同时，由透射光场的格林函数积分得出了基尔霍夫近似下光场的表达式. 通过类比推导夫琅和费面上散斑场自相关函数的方法，提出了产生随机表面及其导数的傅里叶变换方法. 在此基础上，对采用基尔霍夫近似进行自仿射分形随机表面的散射光场数值计算的精确程度进行了研究. 发现在随机表面粗糙度比较小时，基尔霍夫近似的精度比较高；在粗糙度相同的情况下，表面的分形 关键词： 格林函数积分 基尔霍夫近似 自仿射分形随机表面     

三维频域格林函数的高效率计算方法

将高斯积分引入频域格林函数及其导数的数值计算,计算结果与已有文献进行对比,证明该方法在满足足够精度的基础上,使计算过程简化,减少分区,提高计算效率.     

自由表面波GreenOV函数及其导数的快速算法

快速而又精确地求得领域Ｇｒｅｎｎ函数值是应用频域Ｇｒｅｅｎ函数求解波物相互作用问题的关键。探讨了无限水深频域自由面Ｇｒｅｅｎ函数及其导数的快速算法问题，提出了一个实用有效的计算方法。     

基于有限元法的光子并矢格林函数重整化及其在自发辐射率和能级移动研究中的应用

任意微纳结构中量子点的自发辐射率和能级移动均可用并矢格林函数表达.当源点和场点在同一位置时,格林函数的实部是发散的.为解决这一发散问题,可采用重整化格林函数方法.本文提出一种计算重整化格林函数和散射格林函数的方法.该方法利用有限元,计算点电偶极子的辐射场,将其在量子点体积内做平均得到重整化的并矢格林函数,减去均匀空间中解析的重整化格林函数,得到重整化的散射格林函数.在均匀空间情况下,本方法所得数值结果与解析解一致.将该方法应用到银纳米球系统,以解析的散射格林函数作为参考,结果表明该方法能准确处理散射格林函数的重整化问题.将该方法应用到表面等离激元纳米腔中,发现有极大的自发辐射增强和能级移动,且该结果不依赖于量子点的体积.这些研究在光与物质相互作用领域具有积极的意义.     

��ά�߽�CIP��ֵ��������Ӧ��

利用函数及其高阶导数值构造五次插值函数近似网格单元内的真实解,改进数值求解双曲类偏微分方程的CIP数值算法。基于之前的一维高阶CIP数值算法思想,不同于利用时间分裂技术,发展了二维高阶CIP数值算法。改进后的算法具有五阶数值精度和显示格式的优点。     

高阶CIP数值方法及其在相关物理问题中的应用

利用函数的高阶空间导数值构建其高次插值,得到高阶CIP(Constrained Interpolation Profile)数值算法,并在此基础上模拟研究等离子体物理中著名的伏拉索夫-泊松(Vlasov-Poisson)方程相关物理问题.高阶CIP数值方法具有更高数值精度,从而可以在同等精度的情况下减少计算格点数,加速数值计算速度.     

三维圆柱几何格林函数节块法中子扩散计算

发展了中子扩散计算三维圆柱几何格林函数节块法。首先通过横向积分将中子扩散方程化为三个互相耦合的一维偏通量方程。对于径向偏通量方程，将径向扩散微分算符分解为平板几何的扩散微分算符和一个修正项之和，将修正项移到方程右端作为修正源项。这样，三个方程都化为平板几何的一维方程莆式，再借助平板几何第二类边界条件格林函数，对主几相应体源作积分，建立偏通量积分方程，对于修正源项，通过分部积分方法将偏通量导数项转化     

小孔衍射和近场散射数值计算的格林函数方法

从简谐光波满足的亥姆霍兹方程出发,将由格林定理得到的介质分界面上的积分方程转化为以表面上的光波及其导数为未知量的线性方程组,并对其进行数值求解,实现了光场的数值计算。然后将这一方法应用于亚波长尺度的小孔衍射的光波以及自仿射分形表面产生的随机光场及其在近场区域范围内的传播的计算。在随机表面产生的光场计算中．提出了类比推导夫琅禾费面上散斑场自相关函数的方法产生随机表面,以及计算其导数的傅里叶变换方法。对光场的计算结果表明,在近场范围内,光场随离开表面的距离的增加而迅速变化,其传播特性完全不同于光场在远场范围内的传播特性。     

静电磁场不规则区域问题的小波插值Galerkin算法

讨论了用小波插值Galerkin方法(WIGM)求解椭圆型偏微分方程,特别是求解区域不规则时的问题.在归纳出WIGM一般形式的基础上,推导出该方法在Sobolev空间范数下的误差界限为C2-m.提出了一种解决不规则区域中静电磁场场分析问题的数值算法,其中选用对称插值尺度函数为基函数,它的对称性及其与平均插值尺度函数的关系可以在一定程度上降低数值求解的计算量.最后通过计算实例说明该算法的有效性.     

有限时间Lyapunov指数的高精度计算新方法

针对目前有限时间Lyapunov指数(FTLE)计算方法准确度不高和无法获得边界值的问题,基于对偶数理论提出了一种新的高精度计算方法.首先描述了基于有限空间差分方法计算FTLE的缺点和问题:其次介绍了基于对偶数理论的高精度导数计算方法及其显著优点,并将动力系统的柯西一格林形变张量计算问题转化为对偶数空间中非线性微分方程数值求解问题;最后对单摆和非线性Duffing振子两个典型物理动力系统进行了数值实验.结果表明:基于对偶数理论的新方法能有效、方便和高精度地计算出有限时间Lyapunov指数场,并成功识别出所包含的拉格朗日相关结构.     

The radiation impedance of a rectangular piston

The radiation impedance of a rectangular piston is expressed as the Fourier transform of its impulse response, which is obtained from the recent work of Lindermann [1]. The analytical evaluation of the transform is performed and new integral expressions are presented for both the radiation resistance and reactance. The integrals are readily evaluated in terms of elementary functions at both the low and high frequency limits. The integrals are also expressed as series of Bessel functions which are valid for all frequencies and aspect ratios. Numerical results are presented to illustrate the behavior of the radiation resistance and reactance as a function of the aspect ratio of the piston and a normalized frequency parameter. Additional numerical results are then presented to illustrate the accuracy of the analytical expressions for the radiation resistance and reactance at low and high frequencies. Finally, numerical results are presented to illustrate the application and accuracy of using standard FFT algorithms to evaluate the radiation resistance and reactance directly from the impulse responses.     

An integral representation of the Green function for a linear array of acoustic point sources

We present a new algorithm for the evaluation of the quasi-periodic Green function for a linear array of acoustic point sources such as those arising in the analysis of line array loudspeakers. A variety of classical algorithms (based on spatial and spectral representations, Ewald transformation, etc.) have been implemented in the past to evaluate these acoustic fields. However as we show, these methods become unstable and/or impractically expensive as the frequency of use of the sources increases. Here we introduce a new numerical scheme that overcomes some of these limitations allowing for simulations at unprecedentedly large frequencies. The method is based on a new integral representation derived from the classic spatial form, and on suitable further manipulations of the relevant integrands to render the integrals amenable to efficient and accurate approximations through standard quadrature formulas. We include a variety of numerical results that demonstrate that our algorithm compares favorably with several classical methods both for points close to the line where the poles are located and at high-frequencies while remaining competitive with them in every other instance.     

水平层状各向异性介质中电磁场并矢Green函数的一种高效算法

利用高阶窗函数结合连分式展开等技术研究并建立一种水平层状各向异性介质中电磁场并矢Green函数的快速有效算法. 首先借助于高阶窗函数将构成并矢Green函数的Sommerfeld积分转化成广义快速下降路径上积分,并给出高阶窗函数Hankel变换的一种新的更高阶幂级数展开式以及严格的Lommel函数表达式,以满足在全空间上高精度计算并矢Green函数的要求. 在此基础上,用Bessel函数的零点将积分路径划分成一系列小区间并通过改进的自适应Gauss求积公式确定各个小区间上的积分值,然后引入连分式展开法对各个区间上的积分值求和,从而使整个积分的收敛效率得到大大提高. 最后通过数值结果验证本方法的有效性. 关键词： 高阶窗函数 连分式展开 并矢Green函数 层状各向异性介质     

Explicit time-domain approaches based on numerical Green’s functions computed by finite differences – The ExGA family

The present paper describes a new family of time stepping methods to integrate dynamic equations of motion. The scalar wave equation is considered here; however, the method can be applied to time-domain analyses of other hyperbolic (e.g., elastodynamics) or parabolic (e.g., transient diffusion) problems. The algorithms presented require the knowledge of the Green’s function of mechanical systems in nodal coordinates. The finite difference method is used here to compute numerically the problem Green’s function; however, any other numerical method can be employed, e.g., finite elements, finite volumes, etc. The Green’s matrix and its time derivative are computed explicitly through the range [0, Δt] with either the fourth-order Runge–Kutta algorithm or the central difference scheme. In order to improve the stability of the algorithm based on central differences, an additional matrix called step response is also calculated. The new methods become more stable and accurate when a sub-stepping procedure is adopted to obtain the Green’s and step response matrices and their time derivatives at the end of the time step. Three numerical examples are presented to illustrate the high precision of the present approach.     

Hilbert space inverse wave imaging in a planar multilayer environment

Most diffraction tomography (DT) algorithms use a homogeneous Green function (GF) regardless of the medium being imaged. This choice is usually motivated by practical considerations: analytic inversions in standard geometries (Cartesian, spherical, etc.) are significantly simplified by the use of a homogeneous GF, estimating a nonhomogeneous GF can be very difficult, as can incorporating a nonhomogeneous GF into standard DT algorithms. Devaney has circumvented these issues by developing a purely numerical DT inversion algorithm [A. J. Devaney and M. Dennison, Inverse Probl. 19, 855-870 (2003)] that is independent of measurement system geometry, number of frequencies used in the reconstruction, and GF. A planar multilayer GF has been developed for use in Devaney's "Hilbert space" algorithm and used in a proof-of-principle nondestructive evaluation (NDE) experiment to image noninvasively a flaw in an aluminum/copper planar multilayer medium using data collected from an ultrasonic measurement system. The data were collected in a multistatic method with no beamforming: all focusing through the multilayer was performed mathematically "after-the-fact," that is, after the data were collected.     

Error control in the perfectly matched layer based multilevel fast multipole algorithm

The development of efficient algorithms to analyze complex electromagnetic structures is of topical interest. Application of these algorithms in commercial solvers requires rigorous error controllability. In this paper we focus on the perfectly matched layer based multilevel fast multipole algorithm (PML-MLFMA), a dedicated technique constructed to efficiently analyze large planar structures. More specifically the crux of the algorithm, viz. the pertinent layered medium Green functions, is under investigation. Therefore, particular attention is paid to the plane wave decomposition for 2-D homogeneous space Green functions in very lossy media, as needed in the PML-MLFMA. The result of the investigations is twofold. First, upper bounds expressing the required number of samples in the plane wave decomposition as a function of a preset accuracy are rigorously derived. These formulas can be used in 2-D homogeneous (lossy) media MLFMAs. Second, a more heuristic approach to control the error of the PML-MLFMA’s Green functions is presented. The theory is verified by means of several numerical experiments.     

A meshless algorithm with the improved moving least square approximation for nonlinear improved Boussinesq equation

An improved moving least square meshless method is developed for the numerical solution of the nonlinear improved Boussinesq equation. After the approximation of temporal derivatives, nonlinear systems of discrete algebraic equations are established and are solved by an iterative algorithm. Convergence of the iterative algorithm is discussed. Shifted and scaled basis functions are incorporated into the method to guarantee convergence and stability of numerical results. Numerical examples are presented to demonstrate the high convergence rate and high computational accuracy of the method.     

Non-conformal domain decomposition method with second-order transmission conditions for time-harmonic electromagnetics

Non-overlapping domain decomposition (DD) methods provide efficient algorithms for solving time-harmonic Maxwell equations. It has been shown that the convergence of DD algorithms can be improved significantly by using high order transmission conditions. In this paper, we extend a newly developed second-order transmission condition (SOTC), which involves two second-order transverse derivatives, to facilitate fast convergence in the non-conformal DD algorithms. However, the non-conformal nature of the DD methods introduces an additional technical difficulty, which results in poor convergence in many real-life applications. To mitigate the difficulty, a corner-edge penalty method is proposed and implemented in conjunction with the SOTC to obtain truly robust solver performance. Numerical results verify the analysis and demonstrate the effectiveness of the proposed methods on a few model problems. Finally, drastically improved convergence, compared to the conventional Robin transmission condition, was observed for an electrically large problem of practical interest.