Domain decomposition based iterative methods for nonlinear elliptic finite element problems
- Univ. of Colorado, Boulder, CO (United States)
The class of overlapping Schwarz algorithms has been extensively studied for linear elliptic finite element problems. In this presentation, the author considers the solution of systems of nonlinear algebraic equations arising from the finite element discretization of some nonlinear elliptic equations. Several overlapping Schwarz algorithms, including the additive and multiplicative versions, with inexact Newton acceleration will be discussed. The author shows that the convergence rate of the Newton`s method is independent of the mesh size used in the finite element discretization, and also independent of the number of subdomains into which the original domain in decomposed. Numerical examples will be presented.
- Research Organization:
- Front Range Scientific Computations, Inc., Boulder, CO (United States); USDOE, Washington, DC (United States); National Science Foundation, Washington, DC (United States)
- OSTI ID:
- 223834
- Report Number(s):
- CONF-9404305--Vol.1; ON: DE96005735
- Country of Publication:
- United States
- Language:
- English
Similar Records
Overlapping Schwarz for Nonlinear Problems. An Element Agglomeration Nonlinear Additive Schwarz Preconditioned Newton Method for Unstructured Finite Element Problems
Newton-Krylov-Schwarz algorithms for the 2D full potential equation