A multi-splitting method to solve 2D parabolic reaction-diffusion singularly perturbed systems

In this paper we design and analyze a numerical method to solve a type of reaction-diffusion 2D parabolic singularly perturbed systems. The method combines the central finite difference scheme on an appropriate piecewise uniform mesh of Shishkin type to discretize in space, and the fractional implic...

ver descrição completa

Detalhes bibliográficos
Autores: Clavero, Carmelo, Jorge Ulecia, Juan Carlos
Formato: artículo
Estado:Versión aceptada para publicación
Fecha de publicación:2023
País:España
Recursos:Universidad Pública de Navarra
Repositorio:Academica-e. Repositorio Institucional de la Universidad Pública de Navarra
OAI Identifier:oai:academica-e.unavarra.es:2454/44599
Acesso em linha:https://hdl.handle.net/2454/44599
Access Level:acceso abierto
Palavra-chave:Coupled 2D parabolic systems
Fractional step methods
Piecewise uniform Shishkin meshes
Reaction-diffusion
Splitting by components
Uniformly convergent methods
Descrição
Resumo:In this paper we design and analyze a numerical method to solve a type of reaction-diffusion 2D parabolic singularly perturbed systems. The method combines the central finite difference scheme on an appropriate piecewise uniform mesh of Shishkin type to discretize in space, and the fractional implicit Euler method together with a splitting by directions and components of the reaction-diffusion operator to integrate in time. We prove that the method is uniformly convergent of first order in time and almost second order in space. The use of this time integration technique has the advantage that only tridiagonal linear systems must be solved to obtain the numerical solution at each time step; because of this, our method provides a remarkable reduction of computational cost, in comparison with other implicit methods which have been previously proposed for the same type of problems. Full details of the uniform convergence are given only for systems with two equations; nevertheless, our ideas can be easily extended to systems with an arbitrary number of equations as it is shown in the numerical experiences performed. The numerical results show in practice the qualities of our proposal.