Performance of NASA Equation Solvers on Computational Mechanics Applications

@inproceedings{Storaasli1996PerformanceON,
  title={Performance of NASA Equation Solvers on Computational Mechanics Applications},
  author={Olaf O. Storaasli},
  year={1996}
}
This paper describes the performance of a new family of NASA-developed equation solvers used for large-scale (i.e. 551,705 equations) structural analysis. To minimize computer time and memory, the solvers are divided by application and matrix characteristics (sparse/dense, real/complex, symmetric/nonsymmetric, size: in-core/out of core) and exploit the hardware features of current and future computers. In this paper, the equation solvers, which are written in FORTRAN, and are therefore easily… 

Figures from this paper

Scalability Analysis of Linear Equation Solvers for Sparse Positive Definite Systems

A collection of linear equation solvers is being developed at ARL that the process simulation code accesses in order to automatically select the optimal solver for the given problem at runtime.

Parallel implementation of large-scale structural optimization

The effects of problem size on sequential and parallel versions of this algorithm are studied to highlight the changes in the subroutine timings that are caused by the increase in problem size.

Three-Dimensional Nacelle Aeroacoustics Code With Application to Impedance Education

The research recommends that research be directed toward development of a rapid sparse assembler and exploitation of the multiprocessor capability of the solver to further reduce CPU time.

Multifrontal Solver Combined with Graph Partitioners

For large-scale problems with irregular meshes such as the finite element meshes of aerospace structures, the verification was made that the developed multifrontal solver combined with an efficient graph partitioner and an appropriate mesh mapping scheme (weighted-edge mapping) shows very good performance.

A MULTIFRONTAL SOLVER COMBINED WITH MESH PARTITIONERS

The multifrontal solver combined with graph or mesh partitioner shows much better performance than a single frontal solver with spectral element reordering for large size of problems with irregular meshes.

Turbofan Noise Propagation and Radiation at High Frequencies

This report summarizes progress on NASA Glenn Research Center Grant NAG3-2718 to the University of Missouri at Rolla This grant was awarded on February 22, 2002 and this report covers the performance

On Multifunctional Collaborative Methods in Engineering Science

Based on the results, the integrated modeling approach using the finite element method for multi-fidelity discretization among the subdomains is identified as most robust.

References

SHOWING 1-5 OF 5 REFERENCES

Three parallel computation methods for structural vibration analysis

The Lanczos (1950), multisectioning, and subspace iteration sequential methods for vibration analysis presently used as bases for three parallel algorithms are noted, in the aftermath of three

A Fast Parallel Algorithm for Generation and Assembly of Finite Element Stiffness and Mass Matricesó

  • Proceedings of the 32nd AIAA/ASMEASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference
  • 1991