C1natural neighbor interpolant for partial differential equations
β Scribed by N. Sukumar; B. Moran
- Publisher
- John Wiley and Sons
- Year
- 1999
- Tongue
- English
- Weight
- 327 KB
- Volume
- 15
- Category
- Article
- ISSN
- 0749-159X
No coin nor oath required. For personal study only.
β¦ Synopsis
Natural neighbor coordinates [20] are optimum weighted-average measures for an irregular arrangement of nodes in R n . [26] used the notion of BΓ©zier simplices in natural neighbor coordinates Ξ¦ to propose a C 1 interpolant. The C 1 interpolant has quadratic precision in β¦ β R 2 , and reduces to a cubic polynomial between adjacent nodes on the boundary ββ¦. We present the C 1 formulation and propose a computational methodology for its numerical implementation (Natural Element Method) for the solution of partial differential equations (PDEs). The approach involves the transformation of the original Bernstein basis functions B 3 i (Ξ¦) to new shape functions Ξ¨(Ξ¦), such that the shape functions Ο3I-2(Ξ¦), Ο3I-1(Ξ¦), and Ο3I (Ξ¦) for node I are directly associated with the three nodal degrees of freedom wI , ΞΈI x , and ΞΈI y , respectively. The C 1 shape functions interpolate to nodal function and nodal gradient values, which renders the interpolant amenable to application in a Galerkin scheme for the solution of fourth-order elliptic PDEs. Results for the biharmonic equation with Dirichlet boundary conditions are presented. The generalized eigenproblem is studied to establish the ellipticity of the discrete biharmonic operator, and consequently the stability of the numerical method.
π SIMILAR VOLUMES
Approximations where the derivatives are corrected so as to satisfy linear completeness on the derivatives are investigated. These techniques are used in particle methods and other mesh-free methods. The basic approximation is a Shepard interpolant which possesses only constant completeness. The der
Simple, mesh=grid free, explicit and implicit numerical schemes for the solution of linear advection-di usion problems is developed and validated herein. Unlike the mesh or grid-based methods, these schemes use well distributed quasi-random points and approximate the solution using global radial bas
A family of numerical methods which are L-stable, fourth-order accurate in space and time, and do not require the use of complex arithmetic is developed for solving second-order linear parabolic partial differential equations. In these methods, second-order spatial derivatives are approximated by fo
A Gauss-Galerkin finite-difference method is proposed for the numerical solution of a class of linear, singular parabolic partial differential equations in two space dimensions. The method generalizes a Gauss-Galerkin method previously used for treating similar singular parabolic partial differentia
Elliptic PDEs with variable coefficients in a domain with complex geometry occur in many ocean models. The parallelization of the elliptic solver by the Shur complement method is presented for the ice-ocean model BRIOS. The Schur complement method is usually employed as an iterative solver, but for