The description of the behavior of rock masses is usually based on the behavior of intact rock samples studied in the laboratory. One of the difficulties in describing the rock mass behavior is assigning the appropriate constitutive model. With the progress in discrete element software, this limitation may be overcome as the user is not required to prescribe a constitutive model for the rock mass. Instead the mirco-scale properties of the intact rock and joints are defined and the macro-scale response results from those properties and the geometry of the problem. In this paper a rock mass is simulated using a discrete element model and the results compared to the traditional model where the constitutive model is prescribed. The discrete element model gives very contrasting results compared to the traditional model.
1. INTRODUCTION
The international community has been investigating the suitability of deep geological disposal facilities for nuclear waste since the early 1970's. These investigations have mainly focused on issues related to geochemistry, geo-hydrogeology, geology and geomechanics. In the Scandinavian countries of Finland and Sweden, their nuclear waste programs have progressed rapidly and these countries are now in the site selection stage. As a result, their focus is on the site evaluation to ensure the reliability of the selected site for final waste disposal.
It must be realized that much of the data collected from site investigations will be obtained from boreholes, and field mapping. Since the nuclear waste repositories are deep seated, the extent of information on which the site evaluation will be based, is limited. However, the data collected from the site investigations must be adequate for a complete site description, which is built up of models related to geology, rock mechanics, thermal properties, transport properties, hydro-geology, hydro-geochemistry, and surface ecosystems. One of these models is the Rock Mechanical Descriptive Model, developed for any site in hard crystalline rock. This descriptive model involves the characterization of rock mass by means of both the empirical relationships and a theoretical approach based on numerical modeling [1].
During the site investigation phase, various scenarios will have to be evaluated that require knowledge of the rock mass response. Extensive work was carried out in early 1990's to investigate the formation and extent of the excavation disturbed zone around underground openings, i.e. the rock mass response at low confinement. However, the evaluations of most of the scenarios require knowledge of the confined rock mass response.
Dr. E. Hoek in his keynote address at the U.S. Rock Mechanics Symposium (1999) noted that quantifying the post-peak response of rock masses was the biggest challenge facing the rock mechanics community. While the behavior of intact rock samples has been studied extensively in the laboratory and using numerical simulations, little progress has been made pertaining to rock masses.
At present, the rock mass strength and deformation characteristics are evaluated using empirical methods such as the rock mass classification systems of Bieniawski's RMR [2] and Barton's Q-values [3].
