Finite element analysis on distributed memory architectures (Q1207813)
From MaRDI portal
scientific article
| Language | Label | Description | Also known as |
|---|---|---|---|
| English | Finite element analysis on distributed memory architectures |
scientific article |
Statements
Finite element analysis on distributed memory architectures (English)
0 references
16 May 1993
0 references
The finite element computations are to be performed on a transputer array. The domain of the problem is partitioned into a set of subdomains. The interior elements and the interface elements are distinguished for each subdomain. All the sets of interface elements are assigned to the master (driver) processor, while each interior set is assigned to a (slave) transputer. An iterative element-by-element preconditioned conjugate gradient algorithm is presented [see \textit{T. J. R. Hughes}, \textit{I. Levit} and \textit{J. Winget}, Element-by-element implicit algorithms for heat conduction, ASCE J. Eng. Mech. 109, 576-586 (1983)]. A two-dimensional Dirichlet problem for an elliptic equation in a rectangle is considered to test the method on a PC, IBM 80386 with eight transputers.
0 references
distributed memory architectures
0 references
parallel processing
0 references
domain decomposition
0 references
transputer array
0 references
preconditioned conjugate gradient algorithm
0 references
Dirichlet problem
0 references
elliptic equation
0 references
0 references
