Next: Input File Up: LEDG: A Tool For Previous: Introduction

Method

LEDG operates on a 3-dimensional lattice of cubic cells separated by faces and edges. The lattice can have arbitrarily many divisions in the X, Y, and Z directions, but is scaled so as to lie in the unit cube. Based on user input, LEDG groups the cells into one or more classes (which correspond to bodies). Faces separating cells of the same class are discarded (face popping), as are edges all of whose neighboring cells are of the same class.

LEDG uses either a bounded domain or the torus domain. When using the bounded domain (the default), LEDG treats the external vertices, edges, and faces differently from the internal ones. The external elements are not eliminated by face popping, and all external elements can be designated as constrained to the unit cube.

When using the torus domain, LEDG does not create external elements because there are none. After face popping only faces separating cells of different classes remain.

LEDG writes as output an evolver data file with the appropriate constraint and domain information. Bodies are given volumes corresponding to the number of cells they contain.