This paper presents a simple procedure to evaluate the stress-displacement behavior of clean rock joints sheared under a constant normal stiffness. The predictions of the model are compared with test results on joints with different geometries. The results show good agreement.
L'article expose un precede permettant de determiner la relation tension-deformation dans des fractures de roches cisaillees dans la condition de rigidite normale constante. Les resultats du modèle, compares aux donnees experimentales obtenues sur des fractures de geometries differentes, montrent une fiabilite acceptable du modèle.
Der Artikel stellt ein Verfahren vor zur Untersuchung des Verhaltnisses Spannung-Versetzung von Rissan in Fels unter Konstanter Normal spannung. Die Vorhersagen des Modells werden mit experimentellen Ergebnissen, die in Rissan mit verschiedener Geometrien erhalten wurden, verglichen. Die Ergebnisse zeigen eine ausreichende Übereinstimmang.
The prediction of the shear strength developed in a rock joint as it undergoes shear displacements, is a subject which has received increased attention in recent years. Applications cover the design of underground excavations in rock as well as the behavior of rock slopes reinforced with rock anchors. Initial studies covered the case of joints sheared at constant normal stress (Ladanyi & Archambault, 1970; Celestino & Goodman, 1979; Barton et al., 1985). Later, other authors have developed models for the shear behavior of clean rock joints sheared under a stiff normal loading (Leichnitz, 1985; Johnston & Lam, 1989; Saeb & Amadei, 1990; Archambault et al., 1990, Benjellount et al., 1990; Skinas et al. 1990). Most of these models are difficult to apply because they may require a lot of tests, they use many equations, or because they were obtained for some particular case This paper presents a simple analytical procedure to evaluate the strength-displacement behavior of a clean rock joint sheared under a normal stiffness condition. The model is based on simple parameters. For a variety of joint profiles, the comparison of experimental values with those calculated using the model shows an agreement which, for practical purposes, can be adequate.
The testing program was described in a previous paper (Van Sint Jan, 1990) where some typical results were presented. All the samples tested were 16 cm long and 5 cm wide. The geometry of the joint surfaces, shown in Figure 1, included triangular, trapezoidal and sinusoidal teeth and casts from two real joint surfaces. The samples were prepared using a mixture with plaster of Paris. The unconfined compressive strength of the intact material was 0,92 MPa and the angle of friction measured on flat surfaces was 36°, constant within the range of normal stresses used in the tests (0,006 to 0,10 MPa). Shear tests were run with initial values of the normal pressure,, in the range of 0,006 to 0,10 MPa, and with the stiffness of the normal loading, K, either 0 or 0,39 MPa/cm. In most tests the tangential displacement reached one wave length or at least 25 mm, and the direction of movement was then reversed.
First, Johnston & Lam's equations were used to back calculate the shear behavior of the triangular joints. It was found that, although the shape of the curves is similar, the computed values of shear strength, normal force and dilatancy are larger than the experimental ones. For other geometries, such as the casts of real joints RA & RB, it is not clear how to use Johnston & Lam's procedure, thus, the comparison cannot be made.
The analytical model developed herein is a simple extension of Barton's (1973) model, to which the influence of normal load stiffness and joint surface wearing have been added.