Efficient Implementation of the Exact NumericalFar Field Boundary Condition for PoissonEquation on an Infinite Domain
نویسنده
چکیده
Computation of exterior fluid problems often involves solving the Poisson equation in an unbounded domain. To cut down computational cost, it is desirable to minimize the computational domain. Therefore an important numerical issue is how to specify correctly the boundary condition at a finite computational boundary. In this note, we present an efficient method for such a task. The basic idea behind the current work and many previous works (see, for example, [1–4]) exploits solutions to the Laplace equation outside the compact support of the source. Therefore, in theory, all these methods are equivalent. However, the current numerical implementation is simpler than that of previous methods. In particular, we solve the Poisson equation with a false Dirichlet boundary condition, from which the correct solution can be obtained. This avoids handling mixed boundary conditions in the far field such as those used by Keller and Givolli [1]. Moreover, our strategy is independent of the discretization scheme; thus the derivation is more transparent than derivations employing explicit differencing schemes, as in the work of Anderson and Reider [3]. Because our method is independent of the discretization scheme, it can be implemented straightforwardly using a variety of discretization schemes, including finite difference and finite element methods. The current method can also be more efficient in high-order schemes than other methods involving inverting matrices whose complexity depends on the order of finite differencing [4]. Finally,
منابع مشابه
Implementation of Sinc-Galerkin on Parabolic Inverse problem with unknown boundary condition
The determination of an unknown boundary condition, in a nonlinaer inverse diffusion problem is considered. For solving these ill-posed inverse problems, Galerkin method based on Sinc basis functions for space and time will be used. To solve the system of linear equation, a noise is imposed and Tikhonove regularization is applied. By using a sensor located at a point in the domain of $x$, say $...
متن کاملImplementation of D3Q19 Lattice Boltzmann Method with a Curved Wall Boundary Condition for Simulation of Practical Flow Problems
In this paper, implementation of an extended form of a no-slip wall boundary condition is presented for the three-dimensional (3-D) lattice Boltzmann method (LBM) for solving the incompressible fluid flows with complex geometries. The boundary condition is based on the off-lattice scheme with a polynomial interpolation which is used to reconstruct the curved or irregular wall boundary on the ne...
متن کاملA Boundary Elements and Particular Integrals Implementation for Thermoelastic Stress Analysis
A formulation and an implementation of two-dimensional Boundary Element Method (BEM) analysis for steady state, uncoupled thermoelastic problems is presented. This approach differs from other treatments of thermal loads in BEM analysis in which the domain integrals due to the thermal gradients are to be incorporated in the analysis via particular-integrals. Thus unlike Finite Elements or Field ...
متن کاملAn Algorithm based on Predicting the Interface in Phase Change Materials
Phase change materials are substances that absorb and release thermal energy during the process of melting and freezing. This characteristic makes phase change material (PCM) a favourite choice to integrate it in buildings. Stephan problem including melting and solidification in PMC materials is an practical problem in many engineering processes. The position of the moving boundary, its veloci...
متن کاملAn Efficient Numerical Method to Solve the Boundary Layer Flow of an Eyring-Powell Non-Newtonian Fluid
In this paper, the boundary layer flow of an Eyring-Powell non-Newtonian fluid over a linearly stretching sheet is solved using the combination of the quasilinearization method and the Fractional order of Rational Chebyshev function (FRC) collocation method on a semi-infinite domain. The quasilinearization method converts the equation into a sequence of linear equations then, using the FRC coll...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2001