Geração de fraturas auto-similares em meios desordenados: técnicas do caminho crítico e do caminho mínimo

Abstract

In this work we propose two models for fracture generation in regular substrates. In the first model, we iteratively apply the concept of critical path to systematically determine the lower “conductivity” element in the connected spanning network. At each iteration, once these elements are identified as local “cracks ́ ́, they are permanently removed from the structure up to the point in which a macroscopic fracture can destroy the global network connectivity. This fracture is then topologically characterized as self-similar with fractal dimension Dp ≈ 1.21. In the second model, we employ the algorithm of Dijkstra to determine the minimal path in a random energy landscape and remove its highest energy element. As in the previous model, these elements are considered to be local “cracks ́ ́ till a subset of them can be identified as a macroscopic fracture. The average over many samples of fractures calculated for different system sizes reveals the presence of a self-similar structure with fractal dimension Df ≈ 1.21. The resemblance between the two exponents Dp e Df suggests that the two models belong to the same universality class.