ABSTRACT: We show experimental results undertaken for direct observation of microcrack initiation and propagation by using a newly developed experimental system, then homogenization theories are applied to analyze the stress distribution in the vicinity of pre-existing microcrack and stress-induced microcrack at grain contact.
RÉSUMÉ: Nous presentons les resultats des experiences entreprises pour I'observation directe de la formation et du developpement des microcrevasses à I'aide d'un nouveau systeme experimental. On applique ensuite les theories d'homogeneisation pour analyser la repartition des contraintes au voisinage des microcrevasses preexistantes et des microcrevasses provoqúees par des contraintes au contact des grains.
ZUSAMMENFASSUNG: Wir zeigen die experimentellen Ergebnisse, die durch die direkte Beobachtung der Haarriβinitiation und- propagation mit dem neulich entwickelten experimentellen System gefuehrt wurde. Und dann die Homogenisierubgstheorien werden angewendet, um die Streβdistribution in der Umgebung vom pra-existierenden Haarriβ und den durch den Streβ induzierten Haarriβ im Kornkontakt zu analysieren.
1 INTRODUCTION Numerous recent studies have shown that the physical properties of rocks are not only affected by the constituent minerals (Olsson 1974; Wong 1990) and their preferred orientation (Kern et al. 1985) but also by the microcracks (Bombolakis 1973; Splunt et al. 1974; Hadley 1976; Batzle et al. 1976; Richter et al. 1977; Kranz 1979a,b; Nernat-Nasser et al. 1982; Sammis et al. 1986; Yukutake 1989; Gottschalk et al. 1990; Ahrens et al. 1993). Microscopic studies of cracks in postloaded samples have been made by using a scanning electron microscopes (Friedman et al. 1970; Wong 1982; Mardon et al. 1990). Especially for damage propagation, Chelmsford granite specimens were loaded to various levels, unloaded, cut lengthwise into halves and then observed (Peng et al. 1972). These experimental studies have shown that macroscopic fractures grow from grain-scale microcracks, which are abundantly found in crystalline rocks. Whereas, these works have the drawback of having been performed only under zero stress conditions using thin or cut section after experiment, or under artificially fractured conditions. To better understand the fundamental problems of damage initiation and propagation in granite, we have observed the actual micro-damage behavior during the deformation of granite specimens. Experimental studies of damage process in intact granite specimens were carried out by uniaxial compressive stress of coarse-grained granite with newly developed experimental system. This experimental system allows us to observe continuously damaging process under loading without unloading in the same specimen. Recently damage mechanics and crack tensor theories have been developed for estimating micro-damage distribution in a continuum body. If it is possible to assume that the "continuum damage" starts and propagates from pre-existing mesoscale or macroscale cracks, these theories are clearly effective (Kawamoto, Ichikawa and Kyoya 1988). However, the effect of microscale structures for fracturing process of rock is not so simple. In this study, a homogenization theory (Sanchez-Palencia 1980) is applied to analyze the stress distribution in the vicinity of pre-existing microcrack and stress-induced microcrack at grain contact. The analyzed results are compared with the experimental ones. We conclude that microcracking in the contact of constituent minerals plays an important role in the deformation process and leads to shear fault of brittle material.
2 EXPERIMENTS The microcrack formation and subsequent damage propagation examined in this study were produced in a series of compression test by using coarse-grained granite under nominally dried condition at room temperature. Average stress velocity of was applied in this experiment.
2.1 Granite specimen All specimens were made from a block of coarse-grained granite from Geochang, Korea by arranging the cut-planes in the same direction. The specimens are consisted mainly of 36.5% quartz, 56.3% feldspar and 7.2% biotite (modal test results). The bulk density and apparent porosity of the granite are 2.58 g/cm and 0.83%, respectively. The dimension of specimens is shown in Figure 1. The parallel and perpendicular degree of the end planes of the specimen is about 4/1000 and all surfaces were polished with 1000 gr. emery powder to observe easily by a stereoscopic microscope.
2.2 Experimental procedures The design of the specimen assembly and experimental system, consisting of three subsystems a) loading system, b)datarecording system and c)observation system, are illustrated in Figure 2. The loading system is designed as Figure 3.