Two-Grid Domain Decomposition Methods for the Coupled Stokes-Darcy System

TL;DR

This paper introduces two innovative Robin-type two-grid domain decomposition methods for the coupled Stokes-Darcy system, improving computational efficiency and stability through coarse and fine grid strategies, validated by numerical experiments.

Contribution

The paper presents novel two-grid domain decomposition algorithms specifically designed for the coupled Stokes-Darcy system, enhancing efficiency and stability over existing methods.

Findings

The proposed schemes are effective under small mesh sizes and realistic parameters.

Error estimates confirm the stability and convergence of the methods.

Numerical experiments demonstrate the efficiency and validation of the algorithms.

Abstract

In this paper, we propose two novel Robin-type domain decomposition methods based on the two-grid techniques for the coupled Stokes-Darcy system. Our schemes firstly adopt the existing Robin-type domain decomposition algorithm to obtain the coarse grid approximate solutions. Then two modified domain decomposition methods are further constructed on the fine grid by utilizing the framework of two-grid methods to enhance computational efficiency, via replacing some interface terms by the coarse grid information. The natural idea of using the two-grid frame to optimize the domain decomposition method inherits the best features of both methods and can overcome some of the domain decomposition deficits. The resulting schemes can be implemented easily using many existing mature solvers or codes in a flexible way, which are much effective under smaller mesh sizes or some realistic physical parameters. Moreover, several error estimates are carried out to show the stability and convergence of the schemes. Finally, three numerical experiments are performed and compared with the classical two-grid method, which verifies the validation and efficiency of the proposed algorithms.