Optimum acceleration factors for iterative solutions of linear and non-linear systems.
Abstract
The objective of the research is the development and evaluation of iterative methods for solving large systems of linear and nonlinear algebraic equations, with emphasis on systems arising from the discretization by finite difference methods or by finite element methods of problems arising in computational fluid dynamics involving partial differential equations. Special attention is given to the determination of parameters to accelerate the convergence of iterative procedures, for solving the algebraic systems. Two types of approaches have been taken. One approach involves direct consideration of the algebraic system and includes optimization techniques, dual adaptive techniques, and hybrid techniques. As a tool for testing these and other techniques several software packages have been developed as part of the ITPACK project. The other approach involved selection of iteration parameters based on considering methods for solving related time-dependent problems using variable time steps. (KR)
Document Details
- Document Type
- Technical Report
- Publication Date
- May 15, 1988
- Accession Number
- ADA198408
Entities
People
- David M. Young Jr
Organizations
- University of Texas at Austin