Discrete ABP estimate and convergence rates for linear elliptic equations in non-divergence form (Q1656374)
From MaRDI portal
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Discrete ABP estimate and convergence rates for linear elliptic equations in non-divergence form |
scientific article |
Statements
Discrete ABP estimate and convergence rates for linear elliptic equations in non-divergence form (English)
0 references
10 August 2018
0 references
The authors are concerned with a two-scale FEM designed to solve a Dirichlet boundary value problem attached to a linear non-divergence elliptic equation on a bounded but not necessarily convex domain. The introduced two-scale FEM is a compromise between the fine scale accuracy provided by the discrete Laplace operator and the monotonicity and stability achieved at the coarse scale by a linear integro-differential operator. This integral formulation is equivalent with the original problem. Essentially, they provide a novel proof to the discrete Alexandroff-Bakelman-Pucci (ABP) estimate, which is more geometric in nature, suitable for the FEM and which is of some intrinsic interest. With this ABP discrete estimate, the authors establish several rates of convergence depending on the solution and data regularity. Some numerical experiments are carried out in order to explore the qualities as well as the limitations of the proposed FEM.
0 references
linear elliptic PDEs
0 references
non-divergence form
0 references
piecewise linear finite elements
0 references
discrete maximum principle
0 references
discrete Alexandroff estimate
0 references
discrete Alexandroff-Bakelman-Pucci estimate
0 references
2-scale approximation
0 references
maximum-norm error estimate
0 references
0 references
0 references
0 references
0 references
0 references
0 references
0 references
0 references
0 references
0 references
0 references
0 references
0 references
0 references
0 references