ABSTRACT:

In this paper the authors combine three ingredients to analyze the response of deep tunnels in granite rock masses. First, some servo-controlled tests on granite rock are presented and interpreted to estimate dilation. Then, the authors recall a method to characterize rock masses behaving in a strain-softening manner to estimate the most relevant parameters of three rock masses. Finally, this behavior model is introduced in a numerical model to study the response of a tunnel. Results show that whereas variable dilation models affects significantly tunnel displacement response, post-failure strength models control the extent of the failure zones.

1. INTRODUCTION

The extent of rock failure and the displacements induced by the process of excavation around an underground opening are closely related to the rock mass behavior. Whereas at low depth joints usually control instability mechanisms, in deep excavations the compression induced failure processes of rocks and large deformations of the opening are the basic concern. With the increasing use of numerical models in rock engineering in I'' cent decades, excavation design has often come to rely on numerical studies. A visit to a number of publications on this topic reveals that in many cases simple behavior models, such as the elastic perfectly plastic one, are used to analyze rock mass behavior. More complex behavior models accounting for post-failure behavior such as strain-softening are seldom utilized. Finally, the ral of the dilation angle is rarely taken into account; and when it is considered, the approach tends to be extremely simplistic. Hoek & Brown [I] suggested that rock mass behavior may vary according to the quality of the rock mass. They proposed that the elastic brittle (EB) behavior properly represents the behavior of good to very good quality rock masses (OSI>70).

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