Simplified Weak Galerkin and New Finite Difference Schemes for the Stokes Equation
Provides a more efficient and practical numerical scheme for solving the Stokes equation, relevant to computational fluid dynamics and engineering applications.
This work simplifies the weak Galerkin finite element method for the Stokes equation, eliminating interior degrees of freedom while preserving local mass conservation. The method yields new 5- and 7-point finite difference schemes on rectangular grids, with optimal error estimates and superconvergence of order O(h^{1.5}) for velocity and pressure at cell centers.
This article presents a simplified formulation for the weak Galerkin finite element method for the Stokes equation without using the degrees of freedom associated with the unknowns in the interior of each element as formulated in the original weak Galerkin algorithm. The simplified formulation preserves the important mass conservation property locally on each element and allows the use of general polygonal partitions. A particular application of the simplified weak Galerkin on rectangular partitions yields a new class of 5- and 7-point finite difference schemes for the Stokes equation. An explicit formula is presented for the computation of the element stiffness matrices on arbitrary polygonal elements. Error estimates of optimal order are established for the simplified weak Galerkin finite element method in the H^1 and L^2 norms. Furthermore, a superconvergence of order O(h^{1.5}) is established on rectangular partitions for the velocity approximation in the H^1 norm at cell centers, and a similar superconvergence is derived for the pressure approximation in the L^2 norm at cell centers. Some numerical results are reported to confirm the convergence and superconvergence theory.