carrier image

Quality Partitioned Meshing of Multi-Material Objects

Zhang, Qin, Deukhyun Cha, Chandrajit Bajaj

24th International Meshing Roundtable, Elsevier Ltd., October 12-14 2015


24th International Meshing Roundtable
Austin, TX
October 12-14,2014

Institute of Computational Engineering and Sciences, Computational Visualization Center, University of Texas, Austin, TX 78712

We present a simple but effective algorithm for generating topologically and geometrically consistent quality triangular surface meshing of compactly packed multiple heterogeneous domains in R3. By compact packing we imply that adjacent homogeneous domains or materials share some 0, 1, and/or 2 dimensional boundary. Such packed multiple material (or multi-material) solids arise naturally from classification/partitioning/segmentation of homogeneous domains in R3 into different sub-regions. The multi-materials may also represent separate functionally classified sections or just be multiple component copies tightly fused together as perhaps by layered manufacturing processes. The input to our algorithm is a geometric representation of the entire multimaterial solid, and a volumetric classification map identifying the individual materials. As output, each individual material region is represented by a triangulated 2-manifold boundary, with adjacent material regions having shared boundaries. Our algorithm has been implemented, and applied to different multi-material solids, and the results are additionally presented with quantitative analysis of detection and cure of non-manifold interfaces as well as spurious small components. These meshes are useful for combined boundary element analysis, however these simulation results are not presented.

Download Full Paper (PDF Format)

Contact author(s) or publisher for availability and copyright information on above referenced article