Computational & Technology Resources
an online resource for computational,
engineering & technology publications
Civil-Comp Proceedings
ISSN 1759-3433
CCP: 101
PROCEEDINGS OF THE THIRD INTERNATIONAL CONFERENCE ON PARALLEL, DISTRIBUTED, GRID AND CLOUD COMPUTING FOR ENGINEERING
Edited by:
Paper 39

Finite Element Mesh Conversion in Parallel

P. Iványi

Department of System and Software Engineering, Pollack Mihály Faculty of Engineering and Information Technology, University of Pécs, Hungary

Full Bibliographic Reference for this paper
, "Finite Element Mesh Conversion in Parallel", in , (Editors), "Proceedings of the Third International Conference on Parallel, Distributed, Grid and Cloud Computing for Engineering", Civil-Comp Press, Stirlingshire, UK, Paper 39, 2013. doi:10.4203/ccp.101.39
Keywords: finite element mesh, conversion of ASCII files, regular expressions, parallel conversion, long-term storage strategy.

Summary
Three scientific trends has influenced the research that is discussed in this paper. The first trend is that throughout the past couple of decades many researchers and companies have developed different simulation software that uses mainly their own finite element mesh format. For examples, the following finite element mesh formats exits: OpenFOAM, CFX, FLUENT, GMSH, MEDIT, LUSAS, LS-DYNA, ANSYS, IDEAS, VTK and so on. Although these mesh formats can be very different, but they could be classified into a couple of classes and converted into a single mesh format as published in Reference [1].

The second trend is that the size of the finite element meshes has continuously increased in the past decades and meshes with even more than 50 million elements are commonly used. This paper discusses some methods to convert such large meshes in a parallel fashion.

The third trend is that the long term storage of simulations is becoming more and more important. For example, since 2011 applications for a grant from the National Science Foundation in the United States of America, have had to be accompanied by a data management plan. In a recent monograph [2] it was shown that the long term preservation of digital documents can be achieved by simulation or by a constant conversion process.

In the case of the simulation the full computing environment, including hardware and software, must be stored with the data. Since this is not always possible, virtualization [3] plays an increasingly important role. The second technique is the conversion of the data to `recent' formats that can be handled in the current computing environment. This paper discusses this second technique for finite element meshes.

The paper will present some preliminary results, that are very promising. In the case of very large meshes, for example with more than 200,000 points the proof of concept implementation of the algorithms produces super-linear speed-up. On the other hand the limitations of the parallel parsing algorithms are also discussed in this paper. Beside these preliminary results further experiments are required to test the algorithms and strategies, discussed in this paper, with different finite element meshes on different computer architectures.

References
1
P. Ivanyi, "Finite element mesh conversion based on regular expressions", Advances in Engineering Software, 51: 20-39, 2012.
2
U.M. Borghoff, P. Rödig, J. Scheffczyk, L. Schmitz, "Long-Term Preservation of Digital Documents: Principles and Practices", Springer, 2010, page 289.
3
M. Portnoy, "Virtualization Essentials", Sybex, 2012, page 304.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £40 +P&P)