A Domain Decomposition Approach for Calculating the Graph Corresponding to a Fibrous Geometry

Conference paper

DOI: 10.1007/978-3-642-02677-5_1

Part of the Lecture Notes in Computational Science and Engineering book series (LNCSE, volume 70)
Cite this paper as:
Brandt A., Iliev O., Willems J. (2009) A Domain Decomposition Approach for Calculating the Graph Corresponding to a Fibrous Geometry. In: Bercovier M., Gander M.J., Kornhuber R., Widlund O. (eds) Domain Decomposition Methods in Science and Engineering XVIII. Lecture Notes in Computational Science and Engineering, vol 70. Springer, Berlin, Heidelberg

Abstract

The effective properties of composite materials/media are in strong demand in engineering, geoscience, and environmental studies to name just a few examples. In [2], we presented an efficient algorithm for computing an approximation of the effective thermal conductivity tensor for high contrast fibrous geometries. The essential idea of the approach is to take into consideration the network-like structure of a given fibrous geometry and to perform all calculations on the induced unstructured grid. More precisely, the intersections of fibers are considered as nodes and the connecting fibers between nodes are considered as edges of an undirected graph. The weight of each edge depends on the diameter and the conductivity of the respective fiber and the distance of the connected nodes. A comparison between the results produced by our algorithm and classical methods, which resolve the fibrous geometry using volumetric elements, yields evidence of its efficiency and reliability for a large class of problems from engineering and science.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer-Verlag Berlin Heidelberg 2009

Authors and Affiliations

  1. 1.Department of Applied Mathematics & Computer ScienceThe Weizmann Institute of ScienceRehovotIsrael
  2. 2.Fraunhofer Institut für Techno- und WirtschaftsmathematikKaiserslauternGermany

Personalised recommendations