ABSTRACT:

The deformation of brittle rock is studied using a mechanical model that is developed based on the analyses of the microscopic fracture processes within rock. The simulations demonstrate that the elastic parameters (E and v) for the rock matrix and for the fractured specimen can be measured during the unloading and reloading cycle; the axial splitting failure brings about more abrupt bursts than the shear faulting failure; and the size of specimen affects the post-peak behaviour.

RESUME:

La defonnation des roches fragiles est etudiee à l'aide d'un modèle mecanique qui est base sur l'analyse du processus de microfracture à l'interieur de la roche. Les simulations demontrent que les paramètres d'elasticitè (E et v) de la matrice de la roche et de l'echantillon fracture peuvent être mesures durant le cycle de decharge-recharge; la rupture axiale provoque plus d'explosions brutales que la rupture en cisaillement; et la taille de l'echantillon affecte le comportement post-pic.

ZUSAMMENFASSUNG:

Die Deformation von Spödgestein wurde untersucht anhand eines mechanischen Modells, das auf den Analysen von mikroskopischen Bruchvorgangen im Gestein basiert. Die Simulationen zeigen, daβ die elastischen Parameter (E und v) der Gesteinsmatrix und der gebrochenen Proben wahrend des Austragtaktes und der Frischladungsaufnahme gemessen werden können; das Versagen durch axiale Risse bewirkt ein abrupteres Bersten als das Versagen durch Scherbruch; und die Gröβe der Proben beeinfluβt das Post-Peak-Verhalten.

1
INTRODUCTION

In underground excavations, it is often observed that a great number of parallel cracks appear in mine pillars and walls (Coates, 1981). Obviously, the cracks are induced under compression. Sometimes, pillars fail completely due to axial splitting, but most of the time pillars fail in a combination of spalling and shear faulting. In this case, axial cracks are initiated and grow to some extent within a pillar under compression. Then, spalling occurs in the middle of the pillar and develops towards the centre. At the final stage, pillars fail due to the shear faulting along the conjugate planes crossing the centre of the pillar or due to crushing. Horii and Nemat-Nasser (1985, 1986) conducted a series of experiments using resin specimens containing a single flaw or a large number of flaws that were different in length to examine the fracture pattern of the specimens under compression. For the specimen containing a single flaw, wing cracks were initiated at the tips of the flaw under compression. Then, they tended to grow in the loading direction. For specimens containing a large number of flaws, it was reported" that they failed in the form of either axial splitting under uniaxial compression or shear faulting under biaxial compression. Wing cracks were first initiated at the tips of those large flaws and grew in the direction of the maximum compressive stress. The wing cracks continued to grow with an increase in the uniaxial compressive stress. Finally, they coalesced and formed axial splitting fractures through the specimen. However, the fracture pattern was different when confining pressure existed. Wing cracks were first initiated at the tips of large flaws, but their growth was soon suppressed by the confining pressure. Wing cracks were then initiated on a large number of small flaws and coalesced to form shear faults. It is concluded, based on the observations in the field and the laboratory experimental results, that axial splitting is the basic local fracture pattern for brittle rocks under compression. In the pre-peak stage, axial splitting predominates the fracture process, while in the post-peak stage shear faulting does. A constitutive model has been developed for brittle rocks based on the analyses of the fracture process within rock. The deformation of rock, in this model, is decomposed into four components which are caused by (i) the elastic response of rock matrix, (ii) the closure of open cracks, (iii) the growth of cracks, and (iv) the shear faulting, respectively. In the pre-peak stage only the first three are considered, but all four components are involved in the post peak stage.

This content is only available via PDF.
You can access this article if you purchase or spend a download.