Petrov - Galerkin Spectral Elements 1 A Petrov - Galerkin Spectral Element Technique for Heterogeneous
نویسنده
چکیده
A novel spectral element technique is examined in which the test functions are different from the approximating elements. Examples are given for a simple 1D Helmholtz equation, Burger's Equation with a small viscosity, and for Darcy's Equation with a discontinous hydraulic conductivity. 1. Polynomial Approximation. Spectral element approximations have been shown to be an eeective tool for the approximations of certain classes of partial diierential equationss9, 14]. The approach utilized by Patera combines a spectral multi-domain method with the variational formulation commonly used for nite element approximations 14]. The technique proceeds by integrating against \test" functions whose support is limited to either a single subdomain or two adjacent subdomains. Traditionally the test functions are chosen from the same class as the approximating \trial" functions, and an approximation is constructed such that within each subdomain the approximation is a linear combination of orthogonal polynomials. In this paper a new set of trial functions is examined. The new approach is similar in spirit to a mixed nite element method or a Petrov-Galerkin scheme. If the sequence of orthogonal polynomials is given as fp n (x)g, ?1 x 1, then the new trial functions are given by ~ 2 : Integration against these new trial functions is a simpler process than integration against those used in traditional spectral element approaches and the integration can be done with respect to a Jacobi weight function. This new approach is discussed in detail below and is illustrated for two model problems. In addition, a 2-D example for an approximation of Darcy's equation for porous media ow,
منابع مشابه
A Petrov - Galerkin Spectral Element Technique for Heterogeneous
A novel spectral element technique is examined in which the test functions are different from the approximating elements. Examples are given for a simple 1D Helmholtz equation, Burger's Equation with a small viscosity, and for Darcy's Equation with a discontinous hydraulic conductivity. 1. Polynomial Approximation. Spectral element approximations have been shown to be an eeective tool for the a...
متن کاملFinite Element Methods for Convection Diffusion Equation
This paper deals with the finite element solution of the convection diffusion equation in one and two dimensions. Two main techniques are adopted and compared. The first one includes Petrov-Galerkin based on Lagrangian tensor product elements in conjunction with streamlined upwinding. The second approach represents Bubnov/Petrov-Galerkin schemes based on a new group of exponential elements. It ...
متن کاملMeshless Local Petrov-Galerkin Method for Elasto-Static Analysis of Thick-Walled Isotropic Laminated Cylinders
In this paper, one of the simplest and most regular members of the family of the Meshless Local Petrov-Galerkin (MLPG) methods; namely MLPG5, is applied to analyze the thick-walled isotropic laminated cylinders under elasto-static pressure. A novel simple technique is proposed to eliminate a very important difficulty of the meshless methods to deal with material discontinuities regarding to the...
متن کاملThe Petrov-Galerkin Method and Chebyshev Multiwavelet Basis for Solving Integro-Differential Equations
Abstract: There are some methods for solving integro-differential equations. In this work, we solve the general-order Feredholm integro-differential equations. The Petrov-Galerkin method by considering Chebyshev multiwavelet basis is used. By using the orthonormality property of basis elements in discretizing the equation, we can reduce an equation to a linear system with small dimension. For ...
متن کاملExistence of solution and solving the integro-differential equations system by the multi-wavelet Petrov-Galerkin method
In this paper, we discuss about existence of solution for integro-differential system and then we solve it by using the Petrov-Galerkin method. In the Petrov-Galerkin method choosing the trial and test space is important, so we use Alpert multi-wavelet as basis functions for these spaces. Orthonormality is one of the properties of Alpert multi-wavelet which helps us to reduce computations in ...
متن کامل