ENGINEERING BEHAVIOUR OF ANISOTROPIC ROCKS

Abstract

The construction of tunnels, caverns and such other underground structures in anisotropic rock formations, such as phyllites, slates, schists and shales in the Himalayan regions have posed complexities to the geotechnical engineers. The stability analysis of such structures needs to take into account the influence of the structural anisotropy on the mechanical behaviour of rocks. The influence of the intrinsic anisotropy on strength and deformation behaviour of rocks is one of the basic data required for predicting rock performance. In many cases, a particular class of anisotropic rocks, called transversely isotropic rocks, is encountered. The common character of such rocks is the structural anisotropy due to the existence of regular foliation planes. The importance of accounting for the strength anisotropy of rock in different aspects of rock engineering is well known and has been amply demonstrated. Over the past several decades, many investigators have devoted considerable efforts to the study of rock anisotropy. However, the review of literature has revealed that only in very few studies, in a limited way, an anisotropic rock has been systematically investigated under different test modes/conditions over the entire range of schistosity orientation varying from 0° to 90° so as to comprehensively delineate anisotropic nature of its different properties. In spite of the undisputable effect of water on the strength of rocks, only limited data on its effect revealing distinct changes of strength and deformational properties; and on the degree of anisotropy in different strength and deformation characteristics of schistose rocks is found in the literature. The failure criteria for transversely isotropic rocks proposed by different investigators (e.g. McLamore and Gray, 1967; Hoek and Brown, 1980; Ramamurthy et al., 1988; Duveau and Shao, 1998 and Tien and Kuo, 2001) generally provide fairly accurate simulation of the experimental data. However, an exhaustive review of the failure criteria in vogue with regard to their basic assumptions, techniques and calibration procedures lead to the conclusion that all these approaches require a wide range of tests, determination of a number of parameters and/or a considerable amount of curve fitting. Therefore the strength criteria available for transverse isotropic rocks lack in wider practical utility. In view of the above, a comprehensive investigation of the engineering behaviour of three anisotropic rocks has been planned with the following specific objectives: 1. To study basic lithology, texture, structure and index properties of the anisotropic rocks in the laboratory; 2. To determine compressional wave velocity, axial and diametral point load strength indices and Brazilian strength at different orientations of schistosity; 3. To study strength behaviour of anisotropic rocks in unconfined and confined states at different orientations ofschistosity planes with respect to loading direction; 4. To comprehend deformation behaviour of anisotropic rocks more thoroughly by continuously monitoring axial and diametral strains and in particular the direction dependent behaviour of tangent modulus of elasticity in unconfined and confined states; 5. To investigate the effect of saturation on different strength and deformation properties of anisotropic rocks; and 6. To develop simple models to predict strength and modulus ofelasticity of transversely isotropic rocks. Three intrinsically anisotropic rocks; namely phyllite, slate and banded orthoquartzite have been selected objectively from the construction sites of two mega hydroelectric schemes located in the lesser Himalayan region. Thin section studies, Xray diffraction analyses and scanning electron microscopy were performed to find out the petrography and petro-fabric of the foliated rocks. A large number of cylindrical specimens (d=38 mm) were prepared at different orientations of schistosity (P=0°, 15°, 30°, 45°, 60°, 75° and 90°). Electrical resistance strain gauges were used to automatically record the axial and diametral strains during uniaxial and triaxial compression tests. The axial loading was continuously monitored through load cells or pressure transducer. The triaxial tests were conducted at confining pressures (03) of 5, 15, 30 and 60 MPa respectively. One half of the total tests have been conducted on saturated specimens. In all 252 wave velocity tests, 210 axial point load tests, 168 diametral point load tests, 210 Brazilian tests, 126 uniaxial compression tests and 256 axi-symmetric triaxial compression tests were conducted in the present study. The compressional wave velocity of saturated samples has been found higher than that of the corresponding dry samples at all orientationsof foliations. The saturation of specimens has also resulted decrease in wave velocity anisotropy. The test results have clearly demonstrated the influence of foliations on strength indices of the rocks such as axial and diametral point load strength and Brazilian strength. Saturation of specimens resulted in reduction of these indices whereas enhancement in anisotropy associated with them. All the three rocks are found to exhibit 'U-shape' variation of strength with orientation of foliations in uniaxial and triaxial compression tests. The curves have been characterized with two maxima of strength, one at (3=90° and another at p=0° whereas the minimum strength around P=30°. The strength is found to increase non-linearly with confining pressure. However, the rate of strength increase with confining pressure has been found maximum at p=30° and minimum at P=90°. As a consequence, anisotropy in strength diminishes with confining pressure. The reduction in strength anisotropy has also been indicated by the flattening of strength versus orientation curves at higher 111 confining pressures. Saturation of specimens resulted reduction in strength. The test results have indicated that relatively more reduction in strength occurred at p=0° as compared to that at p=90°. The variation of tangent modulus of elasticity with orientation has been observed more or less systematic, resembling U-shape variation ofstrength to a large extent. The highest moduli values were obtained at p=0° whereas the minimum ones usually at p=30°. Saturation significantly reduces modulus ofelasticity values. However its effect onanisotropy ratio of Etso has not been very prominent. The shear strength parameters (c and <|>) of the foliated rocks have also been observed depending on orientation of schistosity, confining pressure and the physical condition ofthe specimens (dry or saturated). The cohesion exhibited more systematic variation with p and increased with confining pressure. The analysis of test results indicated decrease in friction values with confining pressure which has been observed more in saturated specimens as compared to those observed in the corresponding dry specimens. The performance of failure criteria proposed by Hoek and Brown (1980), Ramamurthy et al. (1988) and Tien and Kuo (2001) have been evaluated onthe basis of experimental data on strength of anisotropic rocks tested in present study as well as those collected from the available literature. Such an analysis revealed that the three criteria generally provide fairly accurate prediction of strength; nevertheless they involve a number of parameters and hence require more experimental data for their determination and/or a considerable amount ofcurve fitting. Asemi-empirical single parameter parabolic strength criterion has been proposed in the present study. The criterion has been developed on the basis of Barton's hypothesis assuming that the rocks reach a critical state of strength enhancement for confining pressures greater than or equal to the uniaxial compressive strength of the IV intact rock material. It has been found capable of estimating the strength of transversely isotropic rocks over the entire range of stress (i.e. low to very high). The performance of the proposed criterion has been exhaustively examined by its application on 21 different anisotropic rocks (data collected from literature including the present study). The quality of strength predictions were found fairly well. At the moment, it has been presented in two-dimensional form and the effect of intermediate principal stress was ignored. In view of a judicious balance in its simplicity and accuracy of strength prediction, it is hoped that the proposed criterion will find a wider practical utility in assisting the practicing engineers and consultants to estimate the strength of rocks with one set of weakness planes without much indulgence in performing tests in the laboratory. In absence of an approach for prediction of modulus of elasticity, a simple model has been suggested. The empirical proposition has been formulated in terms of the uniaxial compressive strength (acp), loading orientation (P) and the confining pressure (03) for estimation of the Etso of foliated rocks. In order to evaluate the performance of the proposed model, the predicted values are compared with their corresponding experimental values obtained in the present study. The quantitative comparison reveals a fairly good agreement between the predicted and the experimental results.