Showing papers on "Finite difference method published in 1996"

Computational methods for fluid dynamics

Abstract: Preface. Basic Concepts of Fluid Flow.- Introduction to Numerical Methods.- Finite Difference Methods.- Finite Volume Methods.- Solution of Linear Equation Systems.- Methods for Unsteady Problems.- Solution of the Navier-Stokes Equations.- Complex Geometries.- Turbulent Flows.- Compressible Flow.- Efficiency and Accuracy Improvement. Special Topics.- Appendeces.

A practical guide to pseudospectral methods

Abstract: 1. Introduction 2. Introduction to spectral methods via orthogonal functions 3. Introduction to PS methods via finite differences 4. Key properties of PS approximations 5. PS variations/enhancements 6. PS methods in polar and spherical geometries 7. Comparisons of computational cost - FD vs. PS methods 8. Some application areas for spectral methods Appendices.

Radiation effect on mixed convection along a vertical plate with uniform surface temperature

Abstract: This paper investigates the effect of radiation on the forced and free convection flow of an optically dense viscous incompressible fluid along a heated vertical flat plate with uniform free stream and uniform surface temperature with Rosseland diffusion approximation. With appropriate transformations, the boundary layer equations governing the flow are reduced to local nonsimilarity equations valid in the forced convection regime as well as in the free convection regime. A group of transformation is, also, introduced to reduce the boundary layer equations to a set of local nonsimilarity equations valid in both the forced and free convection regimes. Solutions of the governing equations are obtained by employing the implicit finite difference methods together with Keller box scheme and are expressed in terms of local shear stress and local rate of heat transfer for a range of values of the pertinent parameters.

H‐p clouds—an h‐p meshless method

Abstract: A new methodology to build discrete models of boundary-value problems is presented. The h-pcloud method is applicable to arbitrary domains and employs only a scattered set of nodes to build approximate solutions to BVPs. This new method uses radial basis functions of varying size of supports and with polynomialreproducing properties of arbitrary order. The approximating properties of the h-p cloud functions are investigated in this article and a several theorems concerning these properties are presented. Moving least squares interpolants are used to build a partition of unity on the domain of interest. These functions are then used to construct, at a very low cost, trial and test functions for Galerkin approximations. The method exhibits a very high rate of convergence and has a greater -exibility than traditional h-p finite element methods. Several numerical experiments in I-D and 2-D are also presented. @ 1996 John Wiley & Sons, Inc. In most large-scale numerical simulations of physical phenomena, a large percentage of the overall computational effort is expended on technical details connected with meshing. These details include, in particular, grid generation, mesh adaptation to domain geometry, element or cell connectivity, grid motion and separation to model fracture, fragmentation, free surfaces, etc. Moreover, in most computer-aided design work, the generation of an appropriate mesh constitutes, by far, the costliest portion of the computer-aided analysis of products and processes. These are among the reasons that interest in so-called meshless methods has grown rapidly in recent years. Most meshless methods require a scattered set of nodal points in the domain of interest. In these methods, there may be no fixed connectivities between the nodes, unlike the finite element or finite difference methods. This feature has significant implications in modeling some physical phenomena that are characterized by a continuous change in the geometry of the domain under analysis. The analysis of problems such as crack propagation, penetration, and large deformations, can, in principle, be greatly simplified by the use of meshless methods. A growing crack, for example, can be modeled by simply extending the free surfaces that correspond to the crack [ 11. The analysis of large deformation problems by, e.g., finite element methods, may require the continuous remeshing of the domain to avoid the breakdown of the calculation due to

Robust Numerical Methods for Singularly Perturbed Differential Equations: Convection-Diffusion-Reaction and Flow Problems

Abstract: Ordinary Differential Equations.- The Analytical Behaviour of Solutions.- Numerical Methods for Second-Order Boundary Value Problems.- Parabolic Initial-Boundary Value Problems in One Space Dimension.- Analytical Behaviour of Solutions.- Finite Difference Methods.- Finite Element Methods.- Two Adaptive Methods.- Elliptic and Parabolic Problems in Several Space Dimensions.- Analytical Behaviour of Solutions.- Finite Difference Methods.- Finite Element Methods.- Time-Dependent Problems.- The Incompressible Navier-Stokes Equations.- Existence and Uniqueness Results.- Upwind Finite Element Method.- Higher-Order Methods of Streamline Diffusion Type.- Local Projection Stabilization for Equal-Order Interpolation.- Local Projection Method for Inf-Sup Stable Elements.- Mass Conservation for Coupled Flow-Transport Problems.- Adaptive Error Control.

Numerical modeling of cellular/dendritic array growth: spacing and structure predictions

Abstract: A numerical model of cellular and dendritic growth has been developed that can predict cellular and dendritic spacings, undercoolings, and the transition between structures. Fully self-consistent solutions are produced for axisymmetric interface shapes. An important feature of the model is that the spacing selection mechanism has been treated. A small, stable range of spacings is predicted for both cells and dendrites, and these agree well with experiment at both low and high velocities. By suitable nondimensionalization, relatively simple analytic expressions can be used to fit the numerical results. These expressions provide an insight into the cellular and dendritic growth processes and are useful for comparing theory with experiment.

Optimized Compact Finite Difference Schemes with Maximum Resolution

Abstract: Direct numerical simulations and computational aeroacoustics require an accurate finite difference scheme that has a high order of truncation and high-resolution characteristics in the evaluation of spatial derivatives. Compact finite difference schemes are optimized to obtain maximum resolution characteristics in space for various spatial truncation orders. An analytic method with a systematic procedure to achieve maximum resolution characteristics is devised for multidiagonal schemes, based on the idea of the minimization of dispersive (phase) errors in the wave number domain, and these are applied to the analytic optimization of multidiagonal compact schemes. Actual performances of the optimized compact schemes with a variety of truncation orders are compared by means of numerical simulations of simple wave convections, and in this way the most effective compact schemes are found for tridiagonal and pentadiagonal cases, respectively. From these comparisons, the usefulness of an optimized high-order tridiagonal compact scheme that is more efficient than a pentadiagonal scheme is discussed. For the optimized high-order spatial schemes, the feasibility of using classical high-order Runge-Kutta time advancing methods is investigated.

Nonsmooth data error estimates for approximations of an evolution equation with a positive-type memory term

Abstract: We study the numerical approximation of an integro-differential equation which is intermediate between the heat and wave equations. The proposed discretization uses convolution quadrature based on the first- and second-order backward difference methods in time, and piecewise linear finite elements in space. Optimal-order error bounds in terms of the initial data and the inhomogeneity are shown for positive times, without assumptions of spatial regularity of the data.

GT-PML: generalized theory of perfectly matched layers and its application to the reflectionless truncation of finite-difference time-domain grids

Abstract: A new mathematical formulation is presented for the systematic development of perfectly matched layers from Maxwell's equations in properly constructed anisotropic media. The proposed formulation has an important advantage over the original Berenger's perfectly matched layer in that it can be implemented in the time domain without any splitting of the fields. The details of the numerical implementation of the proposed perfectly matched absorbers in the context of the finite-difference time-domain approximation of Maxwell's equations are given. Results from three-dimension (3-D) simulations are used to illustrate the effectiveness of the media constructed using the proposed approach as absorbers for numerical grid truncation.

Finite-difference time-domain algorithm for solving Maxwell's equations in rotationally symmetric geometries

Abstract: In this paper, an efficient finite-difference time-domain algorithm (FDTD) is presented for solving Maxwell's equations with rotationally symmetric geometries. The azimuthal symmetry enables us to employ a two-dimensional (2-D) difference lattice by projecting the three-dimensional (3-D) Yee-cell in cylindrical coordinates (r, /spl phi/, z) onto the r-z plane. Extensive numerical results have been derived for various cavity structures and these results have been compared with those available in the literature. Excellent agreement has been observed for all of the cases investigated.

Large amplitude liquid sloshing in seismically excited tanks

Abstract: An innovative method of analysis was developed to simulate the non-linear seismic finite-amplitude liquid sloshing in two-dimensional containers. In view of the irregular and time-varying liquid surface, the method employed a curvilinear mesh system to transform the non-linear sloshing problem from the physical domain with an irregular free-surface boundary into a computational domain in which rectangular grids can be analysed by the finite difference method. Non-linearities associated with both the unknown location of the free surface and the high-order differential terms were considered. The Crank-Nicolson time marching scheme was employed and the resulting finite difference algorithm is unconditionally stable and very lightly damped with respect to the temporal co-ordinate. In order to minimize numerical instability caused by the computational dispersion in spatially discretized surface wave, a second-order dissipation term was added to the system to filter out the spurious high-frequency waves. Sloshing effects and structural response were measured in terms of sloshing amplitude, base shear and overturning moment generated by the hydrodynamic pressure of the liquid exerted on the container walls. Simulation results of liquid sloshing induced by earthquake and harmonic base excitations were compared with those of the linear wave theory and the limitations of the latter in assessing the response of seismically excited liquids were addressed.

FDTD analysis of lossy, multiconductor transmission lines terminated in arbitrary loads

Abstract: A hybrid method is presented for incorporating general terminations into the solution of lossy multiconductor transmission lines (MTLs). The terminations are characterized by a state-variable formulation which allows a general characterization of dynamic as well as nonlinear elements in the termination networks. The method combines the second-order accuracy of the finite difference-time domain (FDTD) algorithm for the MTL with the absolutely stable, backward Euler discretization of the state-variable representations of the termination networks. A compact matrix formulation of the recursion relations at the interface between the MTL and the termination networks allows a straightforward coding of the algorithm. Skin effect losses of the line conductors as well as the effect of an incident field are easily incorporated into the algorithm. Several numerical examples are given which contain dynamic and nonlinear elements in the terminations. These examples demonstrate the validity of the method and show that the temporal and spatial step sizes can be maximized, thereby minimizing the computational burden.

The perfectly matched layer (PML) boundary condition for the beam propagation method

Abstract: The perfectly matched layer (PML) boundary condition for the Helmoltz equation is developed and applied to the finite-difference beam propagation method. Its effectiveness is verified by way of examples.

Finding apparent horizons in numerical relativity.

Abstract: This paper presents a detailed discussion of the ``Newton's method'' algorithm for finding apparent horizons in 3+1 numerical relativity. We describe a method for computing the Jacobian matrix of the finite differenced $H(h)$ function by symbolically differentiating the finite difference equations, giving the Jacobian elements directly in terms of the finite difference molecule coefficients used in computing $H(h)$. Assuming the finite differencing scheme commutes with linearization, we show how the Jacobian elements may be computed by first linearizing the continuum $H(h)$ equations, then finite differencing the linearized (continuum) equations. We find this symbolic differentiation method of computing the $H(h)$ Jacobian to be {\em much} more efficient than the usual numerical perturbation method, and also much easier to implement than is commonly thought. When solving the discrete $H(h) = 0$ equations, we find that Newton's method generally converges very rapidly. However, if the initial guess for the horizon position contains significant high-spatial-frequency error components, Newton's method has a small (poor) radius of convergence. This is {\em not} an artifact of insufficient resolution in the finite difference grid; rather, it appears to be caused by a strong nonlinearity in the continuum $H(h)$ function for high-spatial-frequency error components in $h$. Robust variants of Newton's method can boost the radius of convergence by O(1) factors, but the underlying nonlinearity remains, and appears to worsen rapidly with increasing initial-guess-error spatial frequency. Using 4th~order finite differencing, we find typical accuracies for computed horizon positions in the $10^{-5}$ range for $\Delta\theta = \frac{\pi/2}{50}$.

Study of discrete test filters and finite difference approximations for the dynamic subgrid‐scale stress model

Abstract: This paper investigates the effects of discrete test filters and finite‐difference approximations for large‐eddy simulations using the dynamic subgrid‐scale stress model. Discrete explicit test filters based on finite‐difference formulations have been constructed and the characteristics of their transfer function are studied. Several definitions of the scaling factor are investigated in the context of the discrete test filters. Two test filters, one based on a discrete representation of the top‐hat filter (A), and another based on a high‐order filtering operation (C) are evaluated in simulations of the turbulent channel flow at Reτ=180. It is found that filter A calculates a higher turbulent viscosity than filter C, which behaves more like a cutoff filter. For the same test filtering operation, the results are found to be sensitive to the ratio of the characteristic lengths of the test and grid filters. By testing two approximations to the convection terms based on second‐order central difference and a nonconservative fifth‐order upwind biased scheme, it is found that the dynamic procedure is receptive to the difference between the two approximations and adjusts the dynamic constant accordingly. It is found that the dynamic subgrid‐scale stress model is more compatible with the Harlow–Welch scheme than dissipative schemes such as the fifth‐order upwind biased approximation.

Finite-difference, time-domain analysis of lossy transmission lines

Abstract: An active and efficient method of including frequency-dependent conductor losses into the time-domain solution of the multiconductor transmission line equations is presented. It is shown that the usual A+B/spl radic/s representation of these frequency-dependent losses is not valid for some practical geometries. The reason for this the representation of the internal inductance the at lower frequencies. A computationally efficient method for improving this representation in the finite-difference time-domain (FDTD) solution method is given and is verified using the conventional time-domain to frequency-domain (TDFD) solution technique.

Finite difference analysis of 2-D photonic crystals

Abstract: In this paper, a finite difference method is developed to analyze the guided-wave properties of a class of two-dimensional photonic crystals (irregular dielectric rods). An efficient numerical scheme is developed to deal with the deterministic equations resulting from a set of finite difference equations for inhomogeneous periodic structures. Photonic band structures within an irreducible Brillouin zone are investigated for both in-plane and out-of-plane propagation. For out-of-plane propagation, the guided waves are hybrid modes; while for in-plane propagation, the guided waves are either TE or TM modes, and there exist photonic bandgaps within which wave propagation is prohibited. Photonic bandgap maps for squares, veins, and crosses are investigated to determine the effects of the filling factor, the dielectric contrast, and lattice constants, on the band-gap width and location. Possible applications of photonic bandgap materials are discussed.

Conservative modeling of 3-D electromagnetic fields, Part I: Properties and error analysis

Abstract: Conservation of electric current and magnetic flux can be explicitly enforced by modeling Maxwell’s equations on a staggered grid, where the different field components are sampled at points offset relative to each other. A staggered finite‐difference (SFD) approximation gives divergence‐free magnetic fields and electric currents ensuring good behavior at all periods. Comparisons of SFD solutions with 2-D quasi‐analytic solutions are very good (∼1% rms error). When a modeled region can be subdivided into uniform subdomains, comparison of analytic solutions and SFD approximations show that the greatest differences occur near the Nyquist wavenumbers; the SFD solutions do not attenuate in space as rapidly as the analytic solutions. The accuracy of a computed SFD solution can be estimated from its wavenumber content. For test cases the accuracy estimates are surprisingly close to the actual accuracies. Grid requirements for modeling short horizontal wavelength components of a solution seem more demanding than ...