Finite‐Element Calculations on Alliant FX/80

Abstract

The finite‐element method has proven to be an invaluable tool for analysis and design of complex, high‐performance systems, such as those typically encountered in the aerospace or automotive industries. However, as the size of the finite‐element models of such systems increases, analysis computation time using conventional computers can become prohibitively high. Parallel processing computers provide the means to overcome these computation‐time limits, provided the algorithms used in the analysis can take advantage of multiple processors. The writers have examined several algorithms for linear and nonlinear static analysis, as well as dynamic finite‐element analysis. The performance of these algorithms on an Alliant FX/80 parallel supercomputer has been investigated. For single load‐case linear static analysis, the optimal solution algorithm is strongly problem dependent. For multiple load cases or nonlinear static analysis through a modified Newton‐Raphson method, decomposition algorithms are shown to have a decided advantage over element‐by‐element preconditioned conjugate gradient algorithms. For eigenvalue/eigenvector analysis, the subspace iteration algorithm with a parallel decomposition is shown to achieve a relatively high parallel efficiency.