Click on “Download PDF” for the PDF version or on the title for the HTML version.


If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Incorporation of Multigrid and Sparse Matrix Techniques into a 3D FE Based Powder Compaction Analysis Software

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan www.asabe.org

Citation:  Paper number  033096,  2003 ASAE Annual Meeting . (doi: 10.13031/2013.13857) @2003
Authors:   B. Mittal, L. T. Zikatanov, V.M. Puri, J. Xu
Keywords:   Finite element method; Mulltigrid method; Sparse matrices; Conjugate gradient; Dry powder compression

To simulate powder compression, a 3D finite element prediction tool (PSU-FEM v2.0 software) with a fast iterative solver based on the multigrid (MG) method was developed. Results obtained from the multigrid solver were compared with a conventional Gauss-Siedel (GS) iterative method. From the analysis, an indistinguishable match between the results was achieved. The results from the multigrid solver of the PSU-FEM v2.0 software were also compared with the results from ABAQUS. The results and the trends were in very good quantitative and qualitative agreement with one another. Sparse matrix techniques were also incorporated into the PSU-FEM v2.0 software. Sparse matrix techniques led to a considerable decrease in solution time and convergence was obtained in few seconds as opposed to few minutes. However, the solution times from the MG solver were slightly higher compared with the GS and conjugate gradient (CG) solvers. The reason for this increase in solution time for MG solver is that given the nature of MG algorithm, the program has to perform a larger number of operations compared to CG and GS solvers. This same effect was also observed for solvers based on dense matrices.

(Download PDF)    (Export to EndNotes)