Thompson—FDTD方法的高精度差分格式

在微分－Ｔｈｏｍｐｓｏｎ变换结合时域有限差分（ＦＤＴＤ）技术计算复杂目标的电磁散射特性的方法中，差分格式的构造和选取与解的精度存在着密度关系。本文将提出一种新的高精度差分格式，其数值实现进一步证实了Ｔｈｏｍｐｓｏｎ－ＦＤＴＤ方法能精确模拟任意形状目标的电磁散射过程。     

Thompson变换－FDTD方法模拟不规则形体的电磁散射问题

三角形由于尖锐的角点，半环形由于既有凹面又有角点，它们都给传统的ＦＤＴＤ数值求解造成了一定的困难。这是因为要获得较高的精度，必须细分网格，从而增加内存需求和计算时间。同时三角形柱和半环柱对Ｔｈｏｍｐｓｏｎ－ＦＤＴＤ方法来说也是“坏”的形状，它使得Ｔｈｏｍｐｓｏｎ变换的精度有所下降。本文通过模拟这两种典型的不利形体的散射问题，进一步验证了Ｔｈｏｍｐｓｏｎ－ＦＤＴＤ法对散射体几何形状变化具有较强的适应能力     

Thompson变换—FDTD方法模拟不规则形体的电磁散射问题

三角形由于尖锐的角点，半环形由于既有凹面又有角点，它们都给传统的ＦＤＴ Ｄ数值求解造成了一定的困难。这是因为要获得较高的精度，必须细分网络，从而增加内存需求和计算时间。同时三角形柱和半环柱对Ｔｈｏｍｐｓｏｎ－ＦＤＴＤ方法来说也是“坏”的形状，它使得Ｔｈｏｍｐｓｏｎ交换的精度有所下降。本文通过模拟这两种典型的不利形体的散射问题，进一步验证了Ｔｈｏｍｐｓｏｎ－ＦＤＴＤ法对散射体几何形状变化具有较强的适     

铁路道碴瞬态电磁散射特性分析

引用离散随机媒质结构来模拟铁路道碴的分布情况，采用在ＦＤＴＤ法的差分网格中进行随机填充来实现铁路道碴的计算模拟模型，并求解其瞬态电磁散射场。通过对该散射场的处理，提取出道碴的非相关散射场，然后再对该非相关场进行分析，得出了可描述道碴分布状况的电磁参量。数值分析结果显示，该参量与道碴空隙率之间在一定范围内具有指数关系。这表明，用电磁方法来探测铁路道碴的空隙率是可能的。     

REVERSAL OF THE DOUBLE－STRANDED－RNA－INDUCED INHIBITION OF PROTEIN SYNTHESIS BY A CATALYTICALLY INACTIVE MUTANT OF THE PROTEIN

ＲＥＶＥＲＳＡＬＯＦＴＨＥＤＯＵＢＬＥ－ＳＴＲＡＮＤＥＤ－ＲＮＡ－ＩＮＤＵＣＥＤＩＮＨＩＢＩＴＩＯＮＯＦＰＲＯＴＥＩＮＳＹＮＴＨＥＳＩＳＢＹＡＣＡＴＡＬＹＴＩＣＡＬＬＹＩＮＡＣＴＩＶＥＭＵＴＡＮＴＯＦＴＨＥＰＲＯＴＥＩＮ　ＫＩＮＡＳＥＰＫＲＴｙｓｏｎ...     

ENHANCED NITRIC OXIDE(NO )-MEDIATED VASODILATIONIN LOW POTASSIUM MEDIUM

ＥＮＨＡＮＣＥＤＮＩＴＲＩＣＯＸＩＤＥ（ＮＯ）－ＭＥＤＩＡＴＥＤＶＡＳＯＤＩＬＡＴＩＯＮＩＮＬＯＷＰＯＴＡＳＳＩＵＭＭＥＤＩＵＭｗａｇｎＹｕ；ＯＳ．Ｓｔｅｉｎｓｌａｎｄ，ＳＨ．Ｎｅｌｓｏｎ，ＡＥ.Ｆｅｅｒｉｃｋ，ＷＥ．Ｊｏｈｎｓｔｏｎｎ（Ｄｅｐａｒｔ?..     

DROPLET DISPERSION AND COALESCENCE IN A POLYMER BLEND IN FLOW FIELDS

ＤＲＯＰＬＥＴＤＩＳＰＥＲＳＩＯＮＡＮＤＣＯＡＬＥＳＣＥＮＣＥＩＮＡＰＯＬＹＭＥＲＢＬＥＮＤＩＮＦＬＯＷＦＩＥＬＤＳＺｈａｎｇＨｏｎｇｂｉｎ（张洪斌）（ＰｏｌｙｍｅｒｉｃＭａｔｅｒｉａｌｓＲｅｓｅａｒｃｈＩｎｓｔｉｔｕｔｅ，ＳｈａｎｇｈａｉＪｉａｏｔ...     

EFFECT OF GRINDING TISSUE ON ISOLATE RECOVERY FROM RABBIT CORNEAS WITH FUNGAL KERATITIS

ＥＦＦＥＣＴＯＦＧＲＩＮＤＩＮＧＴＩＳＳＵＥＯＮＩＳＯＬＡＴＥＲＥＣＯＶＥＲＹＦＲＯＭＲＡＢＢＩＴＣＯＲＮＥＡＳＷＩＴＨＦＵＮＧＡＬＫＥＲＡＴＩＴＩＳＲｉｃｈａｒｄＲｏｂｉｎｓｏｎ，ＹａｎｇＲｏｎｇ；Ｗ．ＳｔｅｖｅｎＨｅａｄ，ＤｅｎｔｓＯ＇Ｄａｙ（Ｄ...     

PROPHYLAXIS OF INTERMITTENT DIAZEPAM IN CHILDREN WITH FEBRILE CONVULSIONS（FC）

ＰＲＯＰＨＹＬＡＸＩＳＯＦＩＮＴＥＲＭＩＴＴＥＮＴＤＩＡＺＥＰＡＭＩＮＣＨＩＬＤＲＥＮＷＩＴＨＦＥＢＲＩＬＥＣＯＮＶＵＬＳＩＯＮＳ（ＦＣ）（ＤｅｐａｒｔｍｅｎｔｏｆＰｅｄｉａｔｒｉｃｓ，ＳｅｃｏｎｄＡｆｆｉｌｉａｔｅｄＨｏｓｐｉｔａｌ，Ｘｉ＇ａｎ７１...     

STUDY ON THE PLATELET INTRACELLULAR FREE Ca~(2 ) IN NORMAL PREGNANCY

Ｒｅｃｅｎｔｌｙ,ｓｔｕｄｉｅｓｏｎｐａｔｈｏｇｅｎｏｆｐｒｅｇｎａｎｃｙ－ｉｎｄｕｃｅｄ－ｈｙｐｅｒｔｅｎｓｉｏｎ（ＰＩＨ）ｓｕｇｇｅｓｔｅｄｔｈａｔｅｌｅｖａｔｉｏｎｏｆＰＦＣａ２＋ｃｏｎｃｅｎｔｒａｔｉｏｎｗａｓａｎｉｍｐｏｒｔａｎｔｆａｃｔｏｒｔｏｉｎｄｕｃｅＰＩＨ．ＴｈｅｓｔｕｄｉｅｓａｌｓｏｆｏｕｎｄｔｈａｔｔｈｅＰＦＣａ２＋ｏｆＰＩＨｐａｔｉｅｎｔｓｗａｓｏｂｖｉ－ｏｕｓｌｙｈｉｇｈｅｒｔｈａｎｔｈａｔｏｆｎｏｒｍａｌｐｒｅｇｎａｎｃｙ．Ｂｕｔｔｈｅｍａｓｓｓａｍｐｌｅｓｔｕｄｙｏ…     

Modified QUICK schemes for 3D advection-diffusion equation of pollutants on unstructured grids

In the framework of finite volume method (FVM), two modified schemes of quadratic upstream interpolation for convective kinematics (QUICK), namely quasi-QUICK (Q-QUICK) and normal quasi-QUICK (NQ-QUICK), for improving the precision of convective flux approximation are verified in 3D unsteady advection-diffusion equation of pollutants on unstructured grids. The constructed auxiliary nodes for Q-QUICK or NQQUICK are composed of two neighboring nodes plus the next upwind node; the later node is generated from intersection of the line of current neighboring nodes and their corresponding interfaces. The numerical results show that Q-QUICK and NQ-QUICK overwhelm central differencing scheme (CDS) in computational accuracy and behave similar numerical stability to upwind difference scheme (UDS), hybrid differencing scheme (HDS) and power difference scheme (PDS) after applying the deferred correction method. Their corresponding CPU time is approximately equivalent to that of traditional difference schemes. In addition, their abilities for adapting high grid deformation are robust. It is so promising to apply the suggested schemes to simulate pollutant transportation on arbitrary 3D natural boundary in the hydraulic or environmental engineering.     

Analysis of numerical integration error for Bessel integral identity in fast multipole method for 2D Helmholtz equation

In 2D fast multipole method for scattering problems, square quadrature rule is used to discretize the Bessel integral identity for diagonal expansion of 2D Helmholtz kernel, and numerical integration error is introduced. Taking advantage of the relationship between Euler-Maclaurin formula and trapezoidal quadrature rule, and the relationship between trapezoidal and square quadrature rule, sharp computable bound with analytical form on the error of numerical integration of Bessel integral identity by square quadrature rule is derived in this paper. Numerical experiments are presented at the end to demonstrate the accuracy of the sharp computable bound on the numerical integration error.     

Scattering Characteristics of Metallic Plates in Radial Waveguide

Mode matching is a useful method to analyze the mode characteristics during the section change of a waveguide.By applying this method to the radial waveguide,and solving specific problems due to the radial transmission of electromagnetic wave,this paper presents a calculating method in radial waveguide based on mode matching.The scattering characteristics of metallic plates in the radial waveguide are calculated,and verified qualitatively using an electromagnetic simulator,which confirmed the validity of this method.     

随机粗糙面上电磁散射的高效迭代IEM计算

在复合目标电磁散射计算中,目标与粗糙面间的耦合近场计算问题是制约算法的主要瓶颈。该文提出一种适用于二维随机粗糙面上电磁散射场计算的迭代积分方程法(IEM)。与传统IEM法不同,迭代IEM法基于近场格林函数建立,考虑了粗糙面面元间的多次电磁互耦作用,散射场不能简化为积分形式的近似解。数值实验表明,与传统MoM法相比,迭代IEM法的内存需求节省了9倍,计算速度比矩量法(MoM)法提升了4.5倍以上,更能有效地计算粗糙面上的散射场。     

水下声散射问题的奇异积分研究

奇异积分和近奇异积分的数值计算是实施边界元法研究水下目标声散射特性的一个难点,全局法是一种可以将面积分转化为计算沿边界曲线的线积分,基于全局法的思想,推导得到了边界元法中奇异积分和近奇异积分的数值计算表达式．仿真结果表明通过奇异积分和近奇异积分的精确处理,提高了边界元法的数值计算精度;且由于得到的数值表达式是闭合的,提高了积分的计算速度;得到的数值表达式不仅可以用于水下声散射特性研究,在声辐射、电磁散射等研究中都可以得到满意的精度．     

解纯对流方程几种向后特征差分格式的比较

介绍了几种向后特征差分格式,用两个典型算例对这几种特征差分格式的计算精度进行了分析比较,计算结果表明,随着插值结点的增多,差分格式的计算精度也越来越高但计算效率却越来越低;具有尖峰剖面与具有缓平剖面的对流方程,其最适宜的特征差分格式各不相同。     

改进透射边界条件的新方法

在时域有限差分方法中，吸收边界条件的吸收特性直接影响到时域有限差分的结果。本文提出了两种改善透射边界条件的方法，多层反射损耗方法和超吸收方法。     

Steady state analysis of towed marine cables

Efficient numerical schemes were presented for the steady state solutions of towed marine cables. For most of towed systems, the steady state problem can be resolved into two-point boundary-value problem, or initial value problem in some special cases where the initial values are available directly. A new technique was proposed and attempted to solve the two-point boundary-value problem rather than the conventional shooting method due to its algorithm complexity and low efficiency. First, the boundary conditions are transformed into a set of nonlinear governing equations about the initial values, then bisection method is employed to solve these nonlinear equations with the aid of 4th order Runge-Kutta method. In common sense, non-uniform (sheared) current is assumed, which varies in magnitude and direction with depth. The schemes are validated through the DE Zoysa's example, then several numerical examples are also presented to illustrate the numerical schemes.