The mechanical behaviour of rock bridges is analysed on simple crack configurations. The results show that direct and induced tensile crack propagation occurs in either stable or unstable conditions depending On crack spacing and applied confinement stresses.
On analyse Ie comportement mecanique des ponts rocheux à partir de quelques simples configurations de fractures. Les resultats montrent que la rupture dans les ponts rocheux peut etre du type stable ou instable selon I'epaisseur des ponts rocheux et les contraintes de confinement appliquees.
Das mechanistic Verhalten von Gesteinsbruecken wurde anhand einfacher Rißverlaufe untersucht. Die Ergebnisse Eigen, day seine lurch Zugkrafte verursachte direkte und induzierte Verbreitung der Risse in enema stabile odder night stabile Muffled auftritt und von der Entfernung der Risse und von den Einsehlußkraften abhangt.
The presence of rock bridges in not fully persistent natural discontinuity sets is a significant factor affecting the stability of rock structures (e.g. rock slopes). The behaviour of rock bridges has been studied by several authors, both in itself and in relation to rock slope stability (see for instance: Jennings, 1970; Einstein et al., 1983; Stimpson, 1978), according to the traditional concepts of the resistance of materials. More recent studies have analysed these phenomena on the basis of fracture mechanics concepts and through the Displacement Discontinuity Method (DDM) (e.g., Sheen, 1993; Scavia, 1989, 1995). The approach adopted by Scavia, in particular, makes it possible to study the degree of stability of rock slopes on the basis of the propagation of natural discontinuities in the rock bridges located in the rock mass. For application purposes, the fundamental limitation of this approach is that, like all discontinue methods, it presupposes a knowledge of the location of the discontinuities in the rock mass. In the authors' opinion, this difficulty can be overcome only by resorting to a probabilistic approach (see for instance Einstein at air., 1983; Scavia et al. 1990), and by analysing several geometric configurations obtained from the statistical distributions of the geometrical characteristics of the discontinuities (orientation. spacing, persistence). However, when using the DDM to deal with a large number of complex geometric configurations, the calculation process proves exceedingly time-consuming. This paper illustrates a process, based on the method proposed by Scavia (1995), aimed at identifying an elementary fracture set which might be deemed to be representative of the phenomena taking place in more complex fracture sets. This approach would make it possible to study the behaviour of rock structures by conceiving them as the sum of several elementary fracture sets.
In relation to the boundary conditions of the system being examined, the analysis of crack propagation in rock bridges calls for the use of a numerical technique. Account taken of the states of stress present in rock Structures, this technique must be able to simulate the propagation of closed and open cracks in tensile and compressive stress fields. If we assume the behaviour of the material making up the rock bridge to be linear elastic, and that non linear effects are limited to the crack surfaces. it is advantageous to refer to the BEM technique of the DDM (Crouch and Star field, 1983) suitably modified in order to take into account the singular state of stress at the crack tips and the propagation process (see for instance, Sheen, 1993: Napier, 1995). In particular, following the method proposed by Scavia, by means of a special computation code (BEMCOM), each crack delimiting a rock bridge is subdivided into a certain number of elements (DD elements) characterized by a constant distribution of the displacement discontinuity and two square root elements located at the crack tips which make it possible to determine the stress intensity factors (Fig. I).