A modified Hoek–Brown failure criterion for anisotropic intact rock

The Hoek–Brown criterion parameters (σ_ci, m_i and s) are significantly influenced by the strength anisotropy of intact rock. In the present study, the criterion was modified by incorporating a new parameter (k_β) to account for the effect of strength anisotropy, thus being able to determine the strength of intact anisotropic rock under loading in different orientations of the plane of anisotropy. The range of the parameter (k_β) for the rocks tested has been analytically investigated by carrying out triaxial tests, in different orientations of the foliation plane. The proposed modification was studied for metamorphic rocks (gneiss, schist, marble), but could also be applied to other rock types exhibiting “inherent” anisotropy, e.g. sedimentary as well as igneous rocks. The proposed modified criterion is intended for use for prediction of strength of intact rock, but can also be extended to rock masses.     

Generalized crack initiation and crack damage stress thresholds of brittle rock masses near underground excavations

The rock mass failure process is characterized by several distinct deformation stages which include crack initiation, crack propagation and coalescence. It is important to know the stress levels associated with these deformation stages for engineering design and practice.Extensive theoretical, experimental and numerical studies on the failure process of intact rocks exist. It is generally understood that crack initiation starts at 0.3 to 0.5 times the peak uniaxial compressive stress. In confined conditions, the constant-deviatoric stress criterion was found to describe the crack initiation stress level.Here, generalized crack initiation and crack damage thresholds of rock masses are proposed. The crack initiation threshold is defined by σ₁−σ₃=A σ_cm and the crack damage threshold is defined by σ₁−σ₃=B σ_cm for jointed rock masses, where A and B are material constants and σ_cm is the uniaxial compressive strength of the rock masses. For a massive rock mass without joints, σ_cm is equal to σ_cd, the long-term uniaxial strength of intact rock. After examining data from intact rocks and jointed rock masses, it was found that for massive to moderately jointed rock masses, the material constants A and B are in the range of 0.4 to 0.5, 0.8 to 0.9, respectively, and for moderately to highly jointed rock masses, A and B are in the range of 0.5 to 0.6, 0.9 to 1.0, respectively. The generalized crack initiation and crack damage thresholds, when combined with simple linear elastic stress analysis, assist in assessing the rock mass integrity in low confinement conditions, greatly reducing the effort needed to obtain the required material constants for engineering design of underground excavations.     

Rock mass strength parameters mobilised in tunnels

Experience suggests that there is an enhancement in the strength of rock mass around the tunnels due to constraints in fracture propagation. In this paper, mobilised strength parameters and modulus of deformation have been deduced from back analysis of the field experience for the purpose of realistic non-linear stress analysis of arched underground openings in nearly dry rock masses, and modified correlations have been suggested. It is also inferred from field observations that sympathetic failure of rock mass may take place at all points almost simultaneously within the failure zone if deviatoric strain exceeds a critical value at any point. Block shear tests support Hoek and Brown's (1980) criterion, which is recommended for analysis of rock slopes and open cut mines.     

Bolt length requirement in underground openings

A parametric study has been carried out using the numerical analysis code FLAC^3D to obtain the influence of various shapes of underground openings on the maximum induced boundary stress. Five shapes—viz. circular, horseshoe, rectangular, elongated D-shape and elliptical—have been considered. For each shape, four tunnel depths and five horizontal in situ stress models have been taken for the study of induced boundary stresses.The values of maximum and minimum induced boundary stresses in the roof and wall have been obtained from the analyses. This data has subsequently been used to develop correlations to estimate the normalized maximum and minimum boundary stresses, which have been subsequently compared with the strength of the rock mass obtained from the Sheorey's non-linear failure criterion for three rock masses represented by three values of Bieniawski's RMR and three values of crushing strength of intact rock material. The values of minimum factor of safety at the roof and the wall have been collected from all the plots. Using these data sets, different correlations have been developed to estimate the minimum factor of safety (f_min) in the roof and wall.Since the bolt length should be normalized with the opening size, some more computer models have been run with varying tunnel width of 5 and 20 m besides the earlier 10 m size to obtain the correlations for estimating the bolt length. The depth of factor of safety contour of 1.5 from the opening periphery has been picked up from all these models and the correlations have been developed for estimating the roof and wall bolt length for the five shapes of underground openings. The correlations for bolt length show that in addition to the shape of underground openings and in situ stress, the bolt length also varies with the rock mass type. These correlations have been verified for field cases of elongated D-shape openings.     

Strength and failure modes of rock mass models with non-persistent joints

Most problems faced by the practicing rock engineer involve the evaluation of rock mass strength and deformability. The theoretical evaluation of the mechanical properties of fractured rock masses has no satisfactory answer because of the great number of variables involved. One of these variables, the influence of which over rock mass behavior is poorly documented, is the degree of fracture persistence. This paper presents the results of biaxial tests performed on physical models of rock with non-persistent joints. The failure modes and maximum strengths developed were found to depend on, among other variables, the geometry of the joint systems, the orientation of the principal stresses, and the ratio between intermediate stress and intact material compressive strength (σ₂/σ_c). Tests showed three basic failure modes: failure through a planar surface, stepped failure, and failure by rotation of new blocks. Planar failure and stepped failure are associated with high strength behavior, and small failure strains, whereas rotational failure is associated with a very low strength, ductile behavior, and large deformation.     

In situ strength criteria for tunnel design in highly-stressed rock masses

The selection of rock mass strength parameters and the approach to assessing the mechanical and thermal-mechanical rock mass behaviour around tunnels are important aspects of designing underground openings for a spent nuclear fuel waste repository. This paper demonstrates how strength criteria based on in situ observations can be used to design stable underground openings in rock masses subjected to adverse stress conditions. The findings are based upon work conducted at AECL’s Underground Research Laboratory (URL) regarding rock mass strength around tunnels in sparsely fractured granite.     

Estimation of mobilised cohesion around underground openings

Based on the field data obtained from 12 Indian and several NGI case histories, a semi-empirical relation has been proposed in this paper for determination of “mobilised” cohesion around underground openings. It has been found that mobilisation of cohesion takes place at higher values around underground openings than the valves previously suggested and the values obtained from block shear tests. A strength enhancement factor has been suggested which is to be multiplied with the cohesion parameter obtained from block shear tests for practical application. This apparent strength enhancement may be attributed to anisotropy in strength, statistical variation in strength and confining conditions around tunnels.     

2D and 3D finite element analysis of underground openings in an inhomogeneous rock mass

The finite element analysis of the underground openings excavated for Koyna hydroelectric project, Maharashtra, India, has been conducted. 2D and 3D models have been developed assuming that the rock mass obeys Drucker–Prager failure criterion. The computed deformations and the stress distribution, around these openings, have been compared with the in situ measurements. The study reveals that the 2D elasto-plastic analysis underestimates the deformations. On the other hand, the 3D elasto-plastic analysis yields results, which compare reasonably well with the in situ measurements. The effect of weak zones in the rock mass and creation of multiple cavities in the nonhomogeneous rock mass has also been considered in the analyses. Such a study is found to be very helpful for evaluating the stability of underground openings when extensive realistic input data is available for nonhomogeneous rock mass.     

A statistical evaluation of intact rock failure criteria constrained by polyaxial test data for five different rocks

In this study we examine seven different failure criteria by comparing them to published polyaxial test data (σ₁>σ₂>σ₃) for five different rock types at a variety of stress states. We employed a grid search algorithm to find the best set of parameters that describe failure for each criterion and the associated misfits. Overall, we found that the polyaxial criteria Modified Wiebols and Cook and Modified Lade achieved a good fit to most of the test data. This is especially true for rocks with a highly σ₂-dependent failure behavior (e.g. Dunham dolomite, Solenhofen limestone). However, for some rock types (e.g. Shirahama Sandstone, Yuubari shale), the intermediate stress hardly affects failure and the Mohr–Coulomb and Hoek and Brown criteria fit these test data equally well, or even better, than the more complicated polyaxial criteria. The values of C₀ yielded by the Inscribed and the Circumscribed Drucker–Prager criteria bounded the C₀ value obtained using the Mohr–Coulomb criterion as expected. In general, the Drucker–Prager failure criterion did not accurately indicate the value of σ₁ at failure. The value of the misfits achieved with the empirical 1967 and 1971 Mogi criteria were generally in between those obtained using the triaxial and the polyaxial criteria. The disadvantage of these failure criteria is that they cannot be related to strength parameters such as C₀. We also found that if only data from triaxial tests are available, it is possible to incorporate the influence of σ₂ on failure by using a polyaxial failure criterion. The results for two out of three rocks that could be analyzed in this way were encouraging.     

Non-dilatant deformation and failure mechanism in two Long Valley Caldera rocks under true triaxial compression

We conducted laboratory rock strength experiments in two ultra-fine-grained brittle rocks, hornfels and metapelite, which together are the major constituent of the Long Valley Caldera (California, USA) basement in the 2025–2996 m depth range. Both rocks are banded, and have very low porosity. Uniaxial compression tests at different orientations with respect to banding planes reveal that while the hornfels compressive strength is nearly isotropic, the metapelite possesses distinct anisotropy. Conventional triaxial tests in these rocks reveal that their respective strengths in a specific orientation increase approximately linearly with confining pressure. True triaxial compression experiments in specimens oriented at a consistent angle to banding, in which the magnitudes of the least (σ₃) and the intermediate (σ₂) principal stresses are different but kept constant during testing while the maximum principal stress is increased until failure, exhibit a behavior unlike that previously observed in other rocks under similar testing conditions. For a given magnitude of σ₃, compressive strength σ₁ does not vary significantly in both Long Valley rock types, regardless of the applied σ₂, suggesting little or no intermediate principal stress effect. Strains measured in all three principal directions during loading were used to obtain plots of σ₁ versus volumetric strain. These are consistently linear almost to the point of rock failure, suggesting no dilatancy. The phenomenon was corroborated by SEM inspection of failed specimens that showed no microcrack development prior to the emergence of one through-going shear failure plane steeply dipping in the σ₃ direction. The strong dependency of compressive strength on the intermediate principal stress in other crystalline rocks was found to be related to microcrack initiation upon dilatancy onset, which rises with increased σ₂ and retards the failure process. We infer that strength independence of σ₂ in the Long Valley rocks derives directly from their non-dilatant deformation.     

Influence of intermediate principal stress on rock fracturing and strength near excavation boundaries—Insight from numerical modeling

The influence of the intermediate principal stress on rock fracturing and strength near excavation boundaries is studied using a FEM/DEM combined numerical tool. A loading condition of σ₃=0 and σ₁≠0, and σ₂≠0 exists at the tunnel boundary, where σ₁, σ₂, and σ₃, are the maximum, intermediate, and minimum principal stress components, respectively. The numerical study is based on sample loading testing that follows this type of boundary stress condition. It is seen from the simulation results that the generation of tunnel surface parallel fractures and microcracks is attributed to material heterogeneity and the existence of relatively high intermediate principal stress (σ₂), as well as zero to low minimum principal stress (σ₃) confinement. A high intermediate principal stress confines the rock in such a way that microcracks and fractures can only be developed in the direction parallel to σ₁ and σ₂. Stress-induced fracturing and microcracking in this fashion can lead to onion-skin fractures, spalling, and slabbing in shallow ground near the opening and surface parallel microcracks further away from the opening, leading to anisotropic behavior of the rock. Hence, consideration of the effect of the intermediate principal stress on rock behavior should focus on the stress-induced anisotropic strength and deformation behavior of the rocks. It is also found that the intermediate principal stress has limited influence on the peak strength of the rock near the excavation boundary.     

A simplified three-dimensional extension of Hoek-Brown strength criterion

The Hoek-Brown(HB) strength criterion has been applied widely in a large number of projects around the world.However,this criterion ignores the intermediate principal stress σ_2.Many evidences have demonstrated that the rock strength is dependent on σ_2. Thus it is necessary to extend the HB criterion into a three-dimensional(3D) form.In this study,the effect of σ_2 on the strength of rocks is identified by reviewing the true triaxial tests of various rock types reported in the literature.A simple 3D strength criterion is developed.The modified criterion is verified by the true triaxial tests of 13 rock types.The results indicate that the modified criterion can achieve a good fit to most of rock types.It can represent a series of criteria as b varies.For comparisons,several existing 3D versions of the HB criterion are selected to predict the strengths of these rock types.It is indicated that the proposed criterion works better than other criteria.A substantial relationship between parameter b and the unconfined compressive strength is established,which guarantees that the proposed criterion can still work well even in the absence of true triaxial test data.     

Anisotropic strength and deformational behavior of Himalayan schists

Anisotropy, which is characteristic of metamorphic rocks such as schists, is due to a process of metamorphic differentiation. Preferred orientation of minerals like mica and chlorite in response to tectonic stresses makes schistose rocks foliated. As a result their engineering properties vary with the direction of loading.The influence of transverse anisotropy on strength and deformational responses of four schistose rocks obtained from the foundation of two underground powerhouse sites in the Himalayas has been critically examined. Specimens at different orientation (β) of the foliations varying from 0° to 90° with respect to the axial stress (σ₁) in the unconfined state and also in the confined states up to 100 MPa of confining pressure were tested to evaluate the applicability of the non-linear strength criterion for the prediction of triaxial compressive strength and modulus. Based on the analysis of large experimental results it has been possible to predict strength and modulus with minimum pre-evaluation experimental data, i.e. only with three uniaxial compressive strength tests at 0°, 30° and 90° and two triaxial compression tests conducted at convenient confining pressures at β=90°orientation. Predicted non-linear stress–strain curves, using predicted values of strength and modulus have been found to match well with the experimental stress–strain curves even at higher confining pressures.     

Application of Weibull's theory to estimating in situ maximum stress σH by hydrofracturing

Hydrofracturing is a widely used and established method for rock stress measurement and is especially valuable at great depths. In conventional hydrofracturing (Haimson, Int. J. Rock Mech. Min. Sci. Geomech. Abstr. 15 (1978) 167), dealing with an axi-parallel fracture, the horizontal minimum stress σ_h is obtained as the shut-in pressure and the maximum stress σ_H is calculated from the breakdown pressure or reopening pressure. It has been pointed out, however, that σ_H is not as reliable as σ_h. This paper therefore presents a new approach for estimating σ_H. In this approach the probabilistic aspects of tensile failure are considered as new sources of information, because the probability density of fracture direction may provide valuable information concerning the stress difference σ_H−σ_h. As the basic theory to describe the tensile failure of rock, we adopted the Weibull’s weakest link theory. The applicability of the theory is first verified via tensile tests on rock specimens of different shape and size, then the probabilistic approach is applied to hydrofracturing to give the probability function of breakdown and the probability density function for the fracture direction. The applicability of the proposed method is presented through numerical calculations and an example in which σ_H−σ_h is estimated from the probabilistic variability of the fracture direction.     

Extending the Q system's prediction of support in tunnels employing fuzzy logic and extra parameters

Rock mass classifications predict support measures according to expert rules by rating rock mass and taking into account the span of the opening. A similar procedure is adopted, in this work, and computerized using statistics and fuzzy logic. Fuzzy expert systems are trained with data of previously constructed underground openings. Using subtractive clustering the systems have the intelligence to pick up the relations between input and output and define the rules that represent the system's behavior automatically. These systems are found to predict support to be used more successfully than the Q system. With the introduction of extra input variables, which are important in numerical analysis, such as depth and intact rock strength, an extended fuzzy system is developed. This system is suggested for preliminary use as it is able to predict support even better.     

Tunnel roof deflection in blocky rock masses as a function of joint spacing and friction – A parametric study using discontinuous deformation analysis (DDA)

The stability of underground openings excavated in a blocky rock mass was studied using the discontinuous deformation analysis (DDA) method. The focus of the research was a kinematical analysis of the rock deformation as a function of joint spacing and friction. Two different opening geometries were studied: (1) span B = h_t; (2) B = 1.5h_t; where the opening height was h_t = 10 m for both configurations. Fifty individual simulations were performed for different values of joint spacing and friction angle. It was found that the extent of loosening above the excavation was predominantly controlled by the spacing of the joints, and only secondarily by the shear strength. The height of the loosening zone h_r was found to be dependent upon the ratio between joint spacing and excavation span S_j/B: (1) h_r < 0.56B for S_j/B  2/10; (2) stable arching within the rock mass for S_j/B  3/10. The results of this study provide explicit correlation between geometrical features of the rock mass, routinely collected during site investigation and excavation, and the expected extent of the loosening zone at the roof, which determines the required support.     

Shaft resistance of a pile embedded in rock

A rational calculation procedure is proposed for establishing the shaft resistance of a pile embedded in rock, based upon the Hoek and Brown failure model. The state of the art of the calculation of the pile shaft resistance is analysed. Nearly all the recommendations that have appeared in the technical literature, for calculating the ultimate shear strength of a shaft embedded in rock (τ_ult) propose that τ_ult=ασ_c^k(σ_c;τ_ulten MN/m²) where the coefficient α, considered as a constant dimensional value, ranges from 0.1 to 0.8, if the unconfined compressive strength (σ_c) is expressed in MN/m². In most cases, the exponent k is 0.5.A comparison is made between the results yielded and the different empirical theories that have been put forward with respect to this shaft resistance. It can generally be stated that the results obtained with this theory are reasonable for long and deeply socketed piles (high confining pressures) but the results are on the safe side in some cases where short piles (low confining pressures) are involved.This paper is a continuation of the works developed by the same authors with piles working at the tip, socketed in rock.     

An equivalent plasticity theory for jointed rock masses

A non-representative volume element (NRVE) approach to equivalent rock mass properties shows that the form of the elastic–plastic constitutive equations is the same for homogeneous material elements and multiple-material elements, subsequently homogenized. Thus the average stress and strain increments in an arbitrary jointed rock mass volume are related by {dσ}=([C*_ep]){dε} where σ is effective stress. The equivalent elastic-plastic properties matrix [C*_ep] is the sum of an equivalent elastic moduli matrix [C^*] and a plastic ‘correction’ matrix [C*_p, as usual. However, there are no equivalent plastic potentials Y^* or yield functions, failure criteria F^* or strengths. The equivalent elastic-plastic properties are constructed from the elastic moduli and strengths of the rock mass joints, the intact rock between and strain influence functions that relate local to overall average strains. Numerical examples that simulate laboratory-like tests on jointed rock cubes illustrate the approach.     

Quantitative three-dimensional description of a rough surface and parameter evolution with shearing

The choice of a general criterion to determine the shear strength of rough rock joints is a topic that has been investigated for many years. The major problem is how to measure and then to express the roughness with a number (e.g., joint roughness coefficient) or a mathematical expression in order to introduce the morphology of the joint into a shear strength criterion. In the present research a large number of surfaces have been digitised and reconstructed using a triangulation algorithm. This approach results in a discretisation of the joint surface into a finite number of triangles, whose geometric orientations have been calculated. Furthermore, during shear tests it was observed that the common characteristic among all the contact areas is that they are located in the steepest zones facing the shear direction. Based on this observations and using the triangulated surface data, it is possible to describe the variation of the potential contact area versus the apparent dip angle with the expression A_θ^*=A₀[(θ_max^*−θ^*)/θ_max^*]^C, where A₀ is the maximum possible contact area, θ_max^* is the maximum apparent dip angle in the shear direction, and C is a “roughness” parameter, calculated using a best-fit regression function, which characterises the distribution of the apparent dip angles over the surface. The close agreement between analytical curves and measured data therefore suggests the possibility of defining the influence of roughness on shear strength by the simple knowledge of A₀, C and θ_max^*. Based on the samples studied here, the values of these parameters capture the evolution of the surface during shearing. Moreover, they tend to be characteristic for specific rock types, indicating that it might be possible to determine ranges for each rock type based on laboratory measurements on representative samples.     

A quantitative comparison of strength criteria for hard rock masses

Knowledge of the rock mass strength is important for the design of all types of underground excavations. A frequently applied approach for estimation of the rock mass strength is through an empirical failure criterion, often in conjunction with rock mass classification/characterisation systems. This paper presents a review of existing methods to estimate the rock mass strength using empirical failure criteria and classification/characterisation systems—in this study, commonly denoted as estimation methods. A literature review of existing methods is presented, after which a set of methods were selected for further studies. The selected methods were used in three case studies, to investigate their robustness and quantitatively compare the advantages and disadvantages of each method. A Round Robin test was used in two of the cases. The case studies revealed that the N, Yudhbir-RMR₇₆, RMi, Q-, and Hoek–Brown-GSI methods, appeared to yield a reasonable agreement with the measured strengths. These methods are thus considered the best candidates for realistic strength estimation, provided that care is taken when choosing values for each of the included parameters in each method. This study has also clearly shown the limits of presently available strength estimation methods for rock masses and further work is required to develop more precise, practical, and easy-to-use methods for determining the rock mass strength. This should be based on the mechanical behaviour and characteristics of the rock mass, which implies that parameters that consider the strength of intact rock, block size and shape, joint strength, and physical scale, are required.