Fast Poisson solvers on nonequispaced grids: multigrid and Fourier methods compared
Gisela Poeplau, Daniel Potts
Abstract
Gisela Poeplau, Daniel Potts
Abstract
The solution of partial differential equations on adaptively generated grids play an important role in scienti£c computation. In this paper we compare two Poisson solvers for data on nonequispaced mesh points. A new meshless Fourier method based on NFFT is constructed in R3. This algorithm is compared to the well-established multigrid method working on nonequidistant meshes. Our investigations are motivated especially by simulations of the behaviour of charged particles in accelerators.
OpenAlex reports 2 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The solution of partial differential equations on adaptively generated grids play an important role in scienti£c computation. In this paper we compare two Poisson solvers for data on nonequispaced mesh points. A new meshless Fourier method based on NFFT is constructed in R3. This algorithm is compared to the well-established multigrid method working on nonequidistant meshes. Our investigations are motivated especially by simulations of the behaviour of charged particles in accelerators.
Key concepts: Multigrid method, Polygon mesh, Partial differential equation, Computer science, Poisson's equation, Poisson distribution, Fourier transform, Computation