ABSTRACT
In an effort to gain a better understanding of shale Joint behavior, normal-only and shear creep tests were performed on artificially prepared shale Joints. Pro- and post-test surfaces were examined for changes in morphology. This paper discusses both of these programs and the results derived from them.
INTRODUCTION
Changes in stress may occur as a result of construction, excavation or the release of residual ground stresses. When the stress regime is altered there is a redistribution of stress throughout the rock mass. The process of stress equilibration includes both 'instantaneous' and time-dependent components. In some instances the cumulative time-dependent strains may exceed the instantaneous elastic response of the rock, by several orders of magnitude. The possibility of this occurrence necessitates that creep movements be considered in the design of surface and underground structures constructed in or on rate-sensitive materials. The strains associated with the readjustment of stress may be manifested as creep along the Joints and/or creep of the intact material. Both types of responses have been observed for shaly materials (Quigley et al, 1978; Lee and Klym, 1978; Bowden, 1982). The total response of a rock mass depends on the relative contribution from each source. Joint creep, is the time-dependent relative displacement of opposing Joint surfaces, caused by a stress smaller than the shear strength of the joint. Our concern for the creep behavior of joints can be justified on the basis of their pervasiveness and the fact that the shear strength of a joint is generally less than that of intact rock. As such, joint creep can account for a significant proportion of the total deformation of a rock mass. Numerical methods are available which can model jointed rock masses, including the time-dependent behavior of both the intact material and joints (Curran and Crawford, 1980; Crawford and Curran, 1983). However, in order for these models to be realistic, we must know the stress-strain-time characteristics of the discontinuities. Furthermore, individual discontinuities which may decisively influence the properties and behavior of the rock mass, must be correctly represented geometrically and be accorded strength and deformation characteristics which reflect their behavior in the field. The accuracy of the stresses and displacements that a model yields is proportional to the quality of the numerical relationships which we use to approximate actual joint behavior. Our ability to correctly and consistently determine these relationships depends on the number and types of factors which influence a joint's behavior and our understanding of them. Shale joints are particularly difficult to model because their frictional strength is, among other things, time-dependent, and we lack a basic understanding of shale joint phenomena. This paper discusses the response of artificially prepared shale joints to normal only, and combined normal and shear loading. The implications of the observed behavior on design criteria are also examined. A large scale, high load capacity single surface bi-directional shear machine (sample size 200 x 300 mm) was designed and constructed at the University of Toronto for the purpose of studying the creep behavior of shale joints. Plate 1 shows the assembled machine and support equipment prior to testing.
