Uniformly convergent difference schemes for singularly perturbed parabolic diffusion-convection problems without turning points (Q1119365): Difference between revisions
From MaRDI portal
Latest revision as of 10:32, 30 July 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Uniformly convergent difference schemes for singularly perturbed parabolic diffusion-convection problems without turning points |
scientific article |
Statements
Uniformly convergent difference schemes for singularly perturbed parabolic diffusion-convection problems without turning points (English)
0 references
1989
0 references
The authors consider the initial-boundary value problem for the parabolic equation \[ \epsilon y_{xx}+a(x,t)y_ x-b(x,t)y-d(x,t)y_ t=f(x,t) \] on [0,1]\(\times [0,T]\) for \(0<\epsilon \leq 1\), smooth positive a, b, d and smooth f. After deriving bounds for the solution y and its derivatives they develop difference schemes on arbitrarily spaced rectangular grids by the method of exponential fitting in the space variable x (by the Petrov-Galerkin approach with appropriate test functions) and classical differencing in the time variable t, thereby using the parameter \(\theta\) of implicity. If everything is sufficiently smooth convergence at the grid points is of first order in both maximum mesh widths H and K, uniformly in \(\epsilon\). In the case \(\theta =1/2\) (Crank-Nicolson) they obtain, as expected, second order convergence with respect to K. Unless \(\theta =1\) (fully implicit scheme) a Courant- Friedrichs-Lewy-type condition is required. The authors also consider the corresponding stationary case, i.e. the corresponding twopoint boundary value problem. They conclude by displaying results of a numerical case study.
0 references
singularly perturbed parabolic diffusion-convection problems
0 references
without turning points
0 references
Crank-Nicolson method
0 references
difference schemes
0 references
method of exponential fitting
0 references
Petrov-Galerkin approach
0 references
convergence
0 references
numerical case study
0 references
0 references
0 references
0 references
0 references
0 references
0 references
0 references