Discretization of the Poisson equation with non-smooth data and emphasis on non-convex domains (Q2821188)
From MaRDI portal
| This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use this page instead for the normal view: Discretization of the Poisson equation with non-smooth data and emphasis on non-convex domains |
scientific article; zbMATH DE number 6628234
| Language | Label | Description | Also known as |
|---|---|---|---|
| English | Discretization of the Poisson equation with non-smooth data and emphasis on non-convex domains |
scientific article; zbMATH DE number 6628234 |
Statements
Discretization of the Poisson equation with non-smooth data and emphasis on non-convex domains (English)
0 references
16 September 2016
0 references
Poisson equation
0 references
non-smooth data
0 references
polygonal domain
0 references
method of transposition
0 references
very weak formulation
0 references
finite element method
0 references
regularization
0 references
convergence
0 references
error estimate
0 references
numerical example
0 references
0 references
0 references
0 references
0 references
The paper is concerned with the Poisson equation \(- \Delta y = f\) on a bounded polygonal domain \(\Omega \subset \mathbb{R}^2\) with boundary conditions \(y=u\) on \(\partial \Omega\) if \(u \in L^2 \left( \partial \Omega \right)\) is not smooth enough. In this case it cannot be expected that there exists a solution \(y \in H^1 \left( \Omega \right)\). The authors discuss several approaches for solving the problem. First, they study the method of transposition that leads to a very weak formulation and discuss a choice of trial and test spaces. Furthermore, Berggren's approach is studied and it is shown that it is equivalent to the method of transposition for a particular choice of test functions. Finally, the data are regularized and the solution is computed as the standard solution of the regularized problem by the finite element method. The advantage of the regularization approach is its simplicity. Convergence of the scheme using piecewise linear functions is proved and error estimates are derived. Numerical examples show that the experimental orders of convergence correspond to the theoretical convergence rates for domains with various maximal interior angles.
0 references