On stability of two-layer explicit schemes (Q2714022)
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: On stability of two-layer explicit schemes |
scientific article; zbMATH DE number 1603285
| Language | Label | Description | Also known as |
|---|---|---|---|
| English | On stability of two-layer explicit schemes |
scientific article; zbMATH DE number 1603285 |
Statements
10 June 2001
0 references
compound explicit scheme
0 references
two-layer scheme
0 references
stability
0 references
convergence
0 references
0.62358975
0 references
0.61876094
0 references
On stability of two-layer explicit schemes (English)
0 references
The authors continue in studying compound explicit schemes [see \textit{Yu. M. Laevskij} and \textit{P. Banushkina}, Sib. J. 3, No.~2, 165-180 (2000; Zbl 0956.65075)] wherein the authors investigated the questions of convergence and stability of the schemes under consideration.NEWLINENEWLINENEWLINEThe aim of the article under review is to expose conditions for the schemes which guarantee stability with respect to the right-hand of the scheme.NEWLINENEWLINENEWLINEThe following two-layer canonical scheme is considered: NEWLINE\[NEWLINE B\frac{u^{n+1} - u^n}{\Delta t} + Au^n = \varphi^n, NEWLINE\]NEWLINE where \(A\: H\to H\), \(B\: H\to H\) are matrix operators in a Hilbert space~\(H\) of the form NEWLINE\[NEWLINE A = \left( \begin{matrix} A_{11} & A_{12}\\ A_{12}^T & A_{22} \end{matrix} \right), \quad B = \left( \begin{matrix} I_{1} & 0_{12}\\ Q^{-1}RA^T_{12} & Q^{-1} \end{matrix} \right). NEWLINE\]NEWLINE Here \(Q\), \(R\) are operator polynomials, \(\varphi^n = (\varphi^n_1, \varphi^n_2)^T \in H\), \(\varphi_1^n = f_1^n\), and \(\varphi^n_2 = \frac{1}{p} Q^{-1}(P^{p-1}f_2^n + \cdots + Pf_2^{n + (p-2)/p} + f_2^{n + (p-1)/p})\), \(P = I_2 - \tau A_{22}\).NEWLINENEWLINENEWLINEThe following theorem holds:NEWLINENEWLINENEWLINETheorem. Let \(\Delta t\|A_{11}\|_{(1)}\leq 1 - \varepsilon\), \(\varepsilon\in (0,1)\), \(\tau\|A_{22}\|_{(2)}\leq 1\). Then in the case of zero initial data the following inequality holds: NEWLINE\[NEWLINE (Au^m,u^m)\leq \frac{1}{\varepsilon}\sum_{n=0}^{m-1}\Delta t\left( \|f_1^n\|^2 + 4\max_{k=1,\dots, p}\|f_2^{n + (k-1)/p}\|^2_2\right).NEWLINE\]
0 references