TY - JOUR
AU - Li,, Bing
AB - Abstract Research on the mechanical behaviour of rock masses with multiple joints has become a popular topic and has practical applications in natural slope stability. This paper aims to clarify the influence of joint geometry, joint orientation and joint connectivity ratio on the mechanical behaviour of rock specimens containing two pre-existing joints. Triaxial compression tests were conducted under various confining pressures to simulate the variation in external conditions. An exponential criterion was used to describe the relationship between the axial stress and confining pressure. The experimental crack propagation was explored by varying the joint orientation, joint connectivity ratio and confining pressure. The structural plane with a greater angle of inclination controlled the failure of the rock sample. Two failure patterns were observed under the loading conditions: shear failure and mixed failure. The failure surface trajectory presented similar deviations with the increase in joint inclination angle, joint connectivity ratio and confining pressure, which also accelerates the transition from mixed failure to shear failure. The experimental results highlight the significance of elucidating the influence of structural planes in practical engineering to predict the stability of natural slopes. specific edge cracks, triaxial compression tests, two failure modes, failure trajectory deviation, mechanical properties 1. Introduction Due to the high geostress and high-altitude environment in the mountainous and hilly areas of Western China, rock collapse, landslides and mudslides are typical geological disasters. However, a fairly large number of hydropower projects are under construction in the area. For example, the Zipingpu Reservoir is located on the Longmenshan fault zone in Sichuan Province, China. Hence, the behaviour of jointed rock masses in this area is closely related to the stability of rock engineering, such as hydropower engineering, underground tunnel engineering and mining engineering (Li et al.2014). In particular, the process of crack propagation in brittle rocks has attracted researchers’ attention (Park & Bobet 2010; Morgan et al.2013; Xue et al.2014). Based on research conducted over the last few decades, the numbers, angles and connectivity ratios of cracks have caused dramatic changes in crack initiation, propagation and coalescence. Previous studies (Panda & Kulatilake 1999; Wasantha et al.2012) on rock specimens with single cracks primarily investigated the effect of joint angle and joint length. Huang & Huang (2014) found that the peak stress ratio decreased as the crack inclination increased. Jin et al. (2017) carried out a series of laboratory tests and numerical simulations to explore the effect of flaw inclination on the strain energy, input energy and dissipation energy of a sample, which exhibit generally increasing trends as the flaw angle increases. Le et al. (2018) investigated samples with single unfilled cracks by performing uniaxial compression tests. They discovered that the uniaxial compressive strength, deformation modulus and axial strain at peak strength varied with the crack inclination angle and reached a minimum when the crack inclination angle reached 60°. Regarding multiple sets of joints in a natural rock mass, the interaction among the cracks, including crack propagation and coalescence, requires more consideration. By studying gypsum specimens via laboratory tests, Shen (1995) considered the propagation of pre-existing joints mainly as tensile fractures, shear fractures and mixed mode fractures. Similar work was later extended by Wong & Chau (1998) through uniaxial compression tests on plaster specimens containing two flaws, and nine types of coalescence patterns were identified. Song & Chunsheng (2018) established a damage constitutive model with non-persistent joints, taking the flaw propagation length and the friction effect of flaw closure into account. Although previous studies on regular joint distribution (e.g. parallel flaws) in rock specimens have achieved significant progress, the problem caused by irregularly distributed fissures in a rock mass exists in many geotechnical projects. Sheng-Qi & Yan-Hua (2018) compared numerical simulations of rock-like specimens containing two pre-fabricated non-parallel flaws with equivalent laboratory experimental results. They found that the location of the tensile stress concentration was controlled by the flaw angle. Afolagboye et al. (2018) related the flaw configuration of non-parallel flaws to the crack initiation, crack path and coalescence behaviour to study the influence of geometrical factors. Feng et al. (2019) coupled static and dynamic strain rates applied to both intact and fissured specimens and discovered that the variation in dynamic strain rate dominated the mechanical properties and the flaw coalescence. In terms of experimental methods, numerous studies (Zhang & Wong 2012; Yang et al.2013; Yanlin et al.2016) conducted uniaxial tests to explore the mechanical properties of internal flaws, although rock masses in geotechnical projects are not consistently under a unidirectional stress state. Taking a jointed rocky slope, located in the Three Gorges area of the Yangtze River, China, as an example (see figure 1), due to long-term weathering, the fissures developed from the free face into the slope, which is in a three-dimensional stress state. In addition, it is critical to determine which joints will play a major role in the slope instability process. Figure 1. Open in new tabDownload slide A jointed rocky slope in the Three Gorges area. Figure 1. Open in new tabDownload slide A jointed rocky slope in the Three Gorges area. As a result, it is important to understand the propagation mechanism of cracks originating from the edges of rock slopes. However, there are few investigations focused on non-coplanar edge flaws, especially under triaxial loading conditions (Zhenghong et al.2019). There are still many challenges in determining the interrelationship among loading conditions, joint geometries and joint distribution. Thus, the present work designed six types of samples prepared with a rock-like material and conducted triaxial compression tests. Three different connectivity ratios and two angle combinations were employed to examine the defining characteristics of the resulting unilateral crack propagation and mechanical response. We investigated the influence of joint geometry on the variation in strength and deformation during the experimental tests. In addition, an exponential criterion was used to relate the axial stress and the lateral stress. The research results could be used as a reference for unstable slope analysis. 2. Experimental procedure 2.1. Sample preparation To simulate the weak weathered sandstone in the Three Gorges reservoir area, cement mortar was used as a rock-like material to mass-produce jointed samples. A comparison of the sandstone and cement mortar parameters is presented in Table 1. Table 1. Comparison of the sandstone and mortar material parameters. Cohesion (MPa) Internal friction angle (°) Uniaxial compression strength (MPa) Elasticity modulus (GPa) Poisson's ratio Unit weight (kN m−3) Sandstone 16.2 35 39.29 10.25 0.4 22 Cement mortar 10.7 31.5 32.65 9.24 0.38 20.6 Cohesion (MPa) Internal friction angle (°) Uniaxial compression strength (MPa) Elasticity modulus (GPa) Poisson's ratio Unit weight (kN m−3) Sandstone 16.2 35 39.29 10.25 0.4 22 Cement mortar 10.7 31.5 32.65 9.24 0.38 20.6 Open in new tab Table 1. Comparison of the sandstone and mortar material parameters. Cohesion (MPa) Internal friction angle (°) Uniaxial compression strength (MPa) Elasticity modulus (GPa) Poisson's ratio Unit weight (kN m−3) Sandstone 16.2 35 39.29 10.25 0.4 22 Cement mortar 10.7 31.5 32.65 9.24 0.38 20.6 Cohesion (MPa) Internal friction angle (°) Uniaxial compression strength (MPa) Elasticity modulus (GPa) Poisson's ratio Unit weight (kN m−3) Sandstone 16.2 35 39.29 10.25 0.4 22 Cement mortar 10.7 31.5 32.65 9.24 0.38 20.6 Open in new tab To ensure the isotropic properties of the samples as much as possible, the cement mortar was made of cement (#425 ordinary Portland cement), sand (medium fine sand) and water with a mass ratio of 1:1.73:0.4. Then, the cement mortar that had not yet solidified was poured into a cylindrical plastic mould to form the final samples. As shown in figure 2, two straight slots were established by the addition of thin steel sheets on the mould surface to create edge cracks. In addition, the moulds and the steel sheets were smeared with a release agent to extrude the samples smoothly after the initial solidification. During sample preparation, a small shaking table system was used to eliminate air bubbles before solidification. Then, the cement mortar samples were preserved in a curing room at 18°C (room temperature) and 99% relative humidity for 28 days. Because the opening width of the cracks was approximately 1 mm, the joints were not filled. Finally, the end faces of the specimens were polished to ensure flatness. All the cylindrical specimens employed in this article had equivalent heights of H ≈ 100 mm and diameters of R ≈ 50 mm. Figure 2. Open in new tabDownload slide The cylindrical plastic moulds and schematic diagram of a specimen. Figure 2. Open in new tabDownload slide The cylindrical plastic moulds and schematic diagram of a specimen. 2.2. Joint geometries The first set of triaxial compression tests was conducted to explore the influence of joint connectivity ratio on the mechanical behaviour of the specimens with non-consistent joints. Here, the crack inclination was invariant. The joint connectivity ratio is defined as the area ratio between the joint plane and the joint cross section in the specimen, as shown in figure 2. Meanwhile, k represents the joint connectivity ratio, three values of which were tested (0.30, 0.45 and 0.60). In addition, all the other joint geometric parameters were invariant. Although the initial joints are non-penetrative, their extensions will extend to the other side of the specimen. The virtual plane of extension of the joint is recognized as a hypothetical damage pathway. The other set of triaxial compressive tests was performed to examine the effect of joint orientation on the mechanical response of the specimens. In this case, the joint connectivity ratio was invariant. The edge starting points of the joints were predetermined in the specimen. There were two flaws: one with a fixed inclination (30°) and another with an unfixed inclination (i.e. 45° or 60°). The joint inclinations were measured from the horizontal section. Other than the joint orientation, all the other joint geometric parameters were invariant. Figure 3 shows the geometric features used to generate six different specimens. Figure 3. Open in new tabDownload slide The schematic diagrams of the joint locations and joint connectivity ratios. Figure 3. Open in new tabDownload slide The schematic diagrams of the joint locations and joint connectivity ratios. 2.3. Loading scheme In this testing programme, an automated testing apparatus was employed, as shown in figure 4. Data were recorded in real-time during the experiment by a computer-controlled system. Figure 4. Open in new tabDownload slide Test apparatus. Figure 4. Open in new tabDownload slide Test apparatus. The Institute of Rock and Soil Mechanics (affiliated with the Chinese Academy of Sciences) designed and manufactured the testing apparatus. This apparatus can apply loading and unloading conditions to rock samples and carry out different types of rock mechanics tests (conventional triaxial tests, uniaxial tests, Brazilian disc splitting tests, direct shear tests etc.). The PCI-2 AE screening system was used to identify the defective samples. To study the mechanical properties of rock masses with non-consistent flaws, especially under triaxial loading conditions, five levels of confining pressure (2, 4, 6, 8 and 10 MPa) and different types of stress path were adopted, as listed in Table 1. The adopted testing process for the cylindrical rock specimens is as follows: Initially, choose the ‘force-confining pressure’ mode and load the axial force and confining pressures simultaneously to the predetermined values at the rates of 0.2 kN s−1 and 0.1 MPa s−1, respectively. Then, keep the confining pressure invariant and load the axial force at the rate of 0.5 kN s−1 until failure. Replace the specimen and perform the same procedure until the entire experiment is finished. 3. Results and discussions 3.1. Stress–strain curve Figure 5 presents the triaxial compression stress–strain behaviours of all the specimens under various confining pressures. Based on the similarity to the classic stress–strain curve, these curves can be clearly divided into the crack compaction segment, elastic segment, plastic segment and residual stress segment. As the confining pressure increases, the sample transitions from brittle to ductile. The strain interval between the peak strength and the residual strength directly reflects the brittle–ductile transition. According to the variation in the peak strength, a higher confining pressure contributes to a greater ultimate compressive strength and residual stress. As joint connectivity ratios increase from 0.3 to 0.6, the strain interval from the end of the elastic segment to the start of the residual stress segment exhibits a decreasing trend in the axial stress. A similar tendency emerges with the transition in inclination from 45° to 60°. Meanwhile, the deformation required for the specimen to reach the peak strength gradually decreases. In other words, lower connectivity ratios retard the occurrence of failure. It can be concluded that higher k values and flaw angles enhance the resistance of the sample to compression, leading to the delay of the residual stress segment. The stress–strain curve does not present a very steep drop, indicating that the crack does not significantly promote the brittleness of the sample. Figure 5. Open in new tabDownload slide Stress-strain curve (k denotes the joint connectivity ratio). (a) 30–45° combination cracks, k = 0.3; (b) 30–60° combination cracks, k = 0.3; (c) 30–45° combination cracks, k = 0.45; (d) 30–60° combination cracks, k = 0.45; (e) 30–45° combination cracks, k = 0.6; and (f) 30–60° combination cracks, k = 0.6. Figure 5. Open in new tabDownload slide Stress-strain curve (k denotes the joint connectivity ratio). (a) 30–45° combination cracks, k = 0.3; (b) 30–60° combination cracks, k = 0.3; (c) 30–45° combination cracks, k = 0.45; (d) 30–60° combination cracks, k = 0.45; (e) 30–45° combination cracks, k = 0.6; and (f) 30–60° combination cracks, k = 0.6. 3.2. Strength characteristics Figure 6 exhibits the change in triaxial compressive strength due to the variation in joint geometries. According to the existing test conditions, the highest strength occurs at a confining pressure of 10 MPa. Based on the triaxial compression test data, the maximum strength at a higher confining pressure (10 MPa) is almost twice the minimum, which occurs at a lower confining pressure (2 MPa). As a result, the increase in axial pressure is less than the increase in confining pressure. In addition, the confining pressure induces higher fissure friction loading on the rock specimen, which results in the full extension of the cracks. Figure 6. Open in new tabDownload slide Triaxial compressive strength under different confining pressures and various connectivity ratios. (a) The 30–45° samples and (b) the 30–60° samples. Figure 6. Open in new tabDownload slide Triaxial compressive strength under different confining pressures and various connectivity ratios. (a) The 30–45° samples and (b) the 30–60° samples. Numerous criteria were proposed to study and predict the rock strength. The Mohr–Coulomb criterion represents a linear relationship between the normal stress |${{\rm{\sigma }}_n}$| and the shear stress |$\tau $| based on the assumption that shear failure occurs across a plane. $$\begin{equation}\left| \tau \right| = {\rm{\ c}} + {\sigma _n}{\rm{tan}}\varphi, \end{equation}$$ (1) where |${\rm{c}}$| and |$\varphi $| represent the cohesion and the internal friction angle of the material, respectively. The Hoek–Brown criterion matches the experimental data better than the Mohr–Coulomb criterion does, although the Hoek–Brown criterion is an empirical equation (Brown & Hoek 1980). The initial form of the Hoek–Brown criterion for intact rock is $$\begin{equation} {\sigma _1} = {\sigma _3}\ + {\sigma _c}\sqrt {m{\sigma _3}/{\sigma _c} + 1}, \end{equation}$$ (2) where |${\sigma _1}$| denotes the maximum principal stress and |${\sigma _3}$| denotes the minimum principal stress. The generalized Hoek–Brown criterion is (Hoek et al.1992) $$\begin{equation} {\sigma _1} = {\sigma _3}\ + {\sigma _c}{\left( {m{\sigma _3}/{\sigma _c} + s} \right)^n}, \end{equation}$$ (3) where |${\sigma _c}$| represents the uniaxial compressive strength of the intact rock, and m and s are constants that rely on the rock characteristics. The value of n depends on the extent of the rock fracturing and varies from 0.5 to 0.65. You (2009, 2010) explored an exponential criterion (see equation (4)), which precisely described the relationship between the strength and the confining pressure: $$\begin{equation} {\sigma _s} - \ {\sigma _3} = {Q_\infty }\ - \left( {{Q_\infty } - {Q_0}} \right)exp\left( { - \frac{{\left( {{k_0} - 1} \right){\sigma _3}}}{{{Q_\infty } - {Q_0}}}} \right), \end{equation}$$ (4) where |${Q_0}$| denotes the uniaxial compressive strength, |${Q_{\infty}}$| denotes the limitation of the stress deviator under the condition that confining pressure (⁠|${\sigma _3}$|⁠) approaches infinity and |${k_0}$| denotes the rate of the strength increase with respect to a confining pressure of zero. Based on equation (4), the parameters (see Table 2) of the exponential criterion were calculated by the experimental data. The correlation coefficients show that the equations provide a better fit to the measured data. Table 2. Parameters of equation (4) according to different categories of specimens. Specimen category |${Q_\infty}$| (MPa) |${Q_0}$| (MPa) |${k_0}$| Correlation coefficient 30–45° combination cracks, k = 0.3 97.05 27.97 3.51 0.99 30–45° combination cracks, k = 0.45 75.08 22.32 4.52 0.91 30–45° combination cracks, k = 0.6 56.22 23.23 5.51 0.99 30–60° combination cracks, k = 0.3 69.73 35.41 6.97 0.90 30–60° combination cracks, k = 0.45 62.57 30.08 7.91 0.96 30–60° combination cracks, k = 0.6 69.22 40.83 3.98 0.85 Specimen category |${Q_\infty}$| (MPa) |${Q_0}$| (MPa) |${k_0}$| Correlation coefficient 30–45° combination cracks, k = 0.3 97.05 27.97 3.51 0.99 30–45° combination cracks, k = 0.45 75.08 22.32 4.52 0.91 30–45° combination cracks, k = 0.6 56.22 23.23 5.51 0.99 30–60° combination cracks, k = 0.3 69.73 35.41 6.97 0.90 30–60° combination cracks, k = 0.45 62.57 30.08 7.91 0.96 30–60° combination cracks, k = 0.6 69.22 40.83 3.98 0.85 Open in new tab Table 2. Parameters of equation (4) according to different categories of specimens. Specimen category |${Q_\infty}$| (MPa) |${Q_0}$| (MPa) |${k_0}$| Correlation coefficient 30–45° combination cracks, k = 0.3 97.05 27.97 3.51 0.99 30–45° combination cracks, k = 0.45 75.08 22.32 4.52 0.91 30–45° combination cracks, k = 0.6 56.22 23.23 5.51 0.99 30–60° combination cracks, k = 0.3 69.73 35.41 6.97 0.90 30–60° combination cracks, k = 0.45 62.57 30.08 7.91 0.96 30–60° combination cracks, k = 0.6 69.22 40.83 3.98 0.85 Specimen category |${Q_\infty}$| (MPa) |${Q_0}$| (MPa) |${k_0}$| Correlation coefficient 30–45° combination cracks, k = 0.3 97.05 27.97 3.51 0.99 30–45° combination cracks, k = 0.45 75.08 22.32 4.52 0.91 30–45° combination cracks, k = 0.6 56.22 23.23 5.51 0.99 30–60° combination cracks, k = 0.3 69.73 35.41 6.97 0.90 30–60° combination cracks, k = 0.45 62.57 30.08 7.91 0.96 30–60° combination cracks, k = 0.6 69.22 40.83 3.98 0.85 Open in new tab 3.3. Deformation characteristics Figure 8 shows the change in axial strain at the peak strength in accordance with various specimens. Due to different joint geometries, the deformation characteristics are anisotropic. As the confining pressure increases, the axial strain increases in general. Notably, a rock mass is more sensitive to tensile failure than to compressive failure. A higher confining pressure lessens the lateral tension that causes a greater axial deformation. However, the variation in joint connectivity ratios produces no corresponding increment or decrement of strain. A similar anomaly is observed in the statistical results of the elastic modulus (see figure 7). The asynchronous phenomenon is induced by stress-strain hysteresis (Jie et al.2017), as investigated by a number of scholars (Cook & Hodgson 1965; Ray et al.1999; Jie et al.2014) in recent decades. For the fissured samples, the total strain (denoted |${\rm{\varepsilon }}$|⁠) includes the deformation of the intact material (denoted |${{\rm{\varepsilon }}_1}$|⁠) and deformation of the cracks (denoted |${{\rm{\varepsilon }}_c}$|⁠). Nevertheless, the responses of |${{\rm{\varepsilon }}_1}$| and |${{\rm{\varepsilon }}_c}$| are not necessarily consistent at the same stress state. Jie et al. (2017) indicated that the asynchronous loading rates of stress and strain at different stages result in the fluctuation of deformation parameters. When the stress rate exceeds the strain rate, the elastic modulus definitely exceeds the average elastic moduli. In the opposite case, the elastic modulus will be less than the average elastic moduli. In addition, the alteration of joint inclination from 45° to 60° insignificantly influences the deformation parameters based on the experimental results. Hence, the confining pressure is the most sensitive factor of deformation. Figure 7. Open in new tabDownload slide Elastic moduli of specimens under various confining pressures. (a) Specimens with 30° and 45° flaws and (b) specimens with 30° and 60° flaws. Figure 7. Open in new tabDownload slide Elastic moduli of specimens under various confining pressures. (a) Specimens with 30° and 45° flaws and (b) specimens with 30° and 60° flaws. 4. Crack propagation Figure 9 presents the crack propagation paths after sample failure. One specimen is generally separated into two major fragments as failure occurs. However, the fragment form varies and is closely related to the joint geometry and the experimental stress state. Accordingly, two types of failure modes were identified, as summarized below: one is shear sliding along the joint orientation and the other is a mixed mode that includes shear along the pre-existing flaw and tensile fracturing from the end of the joint within the specimen. 4.1. The influence of confining pressure The degree of fragmentation and the initiation of the secondary crack can be used to evaluate the effect of confining pressure. At a lower confining pressure (i.e. 2 MPa), a secondary crack initiates from the internal edge of the pre-existing fissure in most cases. Meanwhile, the fragmented edge is relatively irregular, especially close to the specimen edges. As the confining pressures increase, the major fragments gradually exhibit sharper edges and flatter sliding surfaces. These phenomena are induced by the variation in crack bearing capacity under the loading condition. A higher confining pressure can generate a larger normal stress and frictional force to resist tensile fracturing, while a lower confining pressure cannot supply sufficient resistance to confine the interparticle breakage. We assume that a rock mass inherently possesses properties of cohesion and friction, synchronously affecting a rock particle. When the cohesion force plays the dominant role in specimen capacity, the main damage is interparticle breakage. Thus, in this case, the edges of fragments are irregular. In contrast, shear sliding occurs when friction dominates. Particles are squeezed and ground, resulting in a relatively smooth plane. Hence, confining pressure is a key conversion factor between cohesion and friction. 4.2. The influence of flaw inclination Two flaws with different inclinations were created in each specimen. The damage developed at the joint plane with a higher angle, without exception. When samples contain a 45° flaw, the percentage of mixed failure is approximately 100%. Concerning samples with 60° flaws, 26.7% of the specimens present shear failure along the joint plane. These phenomena reveal two facts: (i) The occurrence of failure depends on the greater flaw inclination angle and normally produces one damage face. (ii) An increase in joint angle increases the proportion of specimens that undergo shear failure. Considering the assumption of cohesion and friction discussed in section 4.1, we divided the joint plane into two states, pre-existing and potential penetrating. In the pre-existing part, the major movement between fragments is shear slipping. However, the movement of the potential penetrating part may be shear slipping or tensile failure. According to figure 8, determination of the potential movement relies partially on the joint angles. From 45° to 60°, the joint length increases at the same connectivity ratio, indicating the extension of the shear slipping area and the decrease in the potential penetrating part. This trend accelerates the proportion of specimens that undergo friction sliding, lessening the effect of cohesion force. As a result, the probability of shear failure intensifies. Based on the testing results, the crack with a lower inclination angle has an insignificant effect on the overall mechanical properties due to the flaw size with respect to the sample size. Figure 8. Open in new tabDownload slide The axial strain at the peak strength. (a) Specimens with 30° and 45° flaws and (b) specimens with 30° and 60° flaws. Figure 8. Open in new tabDownload slide The axial strain at the peak strength. (a) Specimens with 30° and 45° flaws and (b) specimens with 30° and 60° flaws. Figure 9. Open in new tabDownload slide Open in new tabDownload slide Sample failure under different confining pressures (k denotes the joint connectivity ratio). (a) 30–45° combination cracks, k = 0.3, (b) 30–45° combination cracks, k = 0.45, (c) 30–45° combination cracks, k = 0.6, (d) 30–60° combination cracks, k = 0.3, (e) 30–60° combination cracks, k = 0.45, and (f) 30–60° combination cracks, k = 0.6. Figure 9. Open in new tabDownload slide Open in new tabDownload slide Sample failure under different confining pressures (k denotes the joint connectivity ratio). (a) 30–45° combination cracks, k = 0.3, (b) 30–45° combination cracks, k = 0.45, (c) 30–45° combination cracks, k = 0.6, (d) 30–60° combination cracks, k = 0.3, (e) 30–60° combination cracks, k = 0.45, and (f) 30–60° combination cracks, k = 0.6. 4.3. The influence of joint connectivity ratio (denoted by k) The variation in joint connectivity ratios from 0.3 to 0.6 directly alters the crack length. From another perspective, this change produces various joint types in the specimens. Notably, the failure trajectory consists of the pre-existing flaws and subsequent cracks. These two sections are not directly connected with each other. Apparently, the higher the connectivity ratio is, the flatter the failure plane. The subsequent crack originates at an internal point of the pre-existing flaw and penetrates the residual segment (relative to the segment penetrated by the pre-existing fissure). The increase in the joint connectivity ratio has a similar impact on the conversion of friction and cohesion. When k rises, the probability of shear failure increases, resulting in a reduced likelihood of tensile fractures. In contrast, considering the orientation of the subsequent crack when k is 0.3, the trajectory deviates gradually from the cylindrical specimen's longitudinal axis to its surface. However, the extent of fragmentation is associated with the confining pressure. At a lower confining pressure (i.e. 2 MPa), the increase in k results in a change in the fragment edge shape from irregular to regular. In regard to a higher confining pressure (i.e. 10 MPa), the edge integrity presents no significant alteration. In conclusion, the increase in confining pressure, joint orientation and connectivity ratio benefits the transformation from mixed failure to shear failure (see figure 10). Furthermore, the expansion of the joint ratio coupled with the enhancement of the frictional force definitely contributes to the frictional sliding against cohesion. Figure 10. Open in new tabDownload slide The influence of joint geometry and confining pressure on the failure trajectory (the dotted line represents the failure surface formed in the test). (a) The influence of confining pressure, (b) the influence of joint connectivity ratio, and (c) the influence of joint inclination. Figure 10. Open in new tabDownload slide The influence of joint geometry and confining pressure on the failure trajectory (the dotted line represents the failure surface formed in the test). (a) The influence of confining pressure, (b) the influence of joint connectivity ratio, and (c) the influence of joint inclination. 5. Conclusions In this paper, the influence of two flaws on the strength, deformation and failure modes of specimens was investigated. A rock-like material was introduced to prepare specimens with edge flaws with specific geometries. Triaxial compression tests under different confining pressures were carried out to study the mechanical behaviour and failure process of these specimens. The resultant stress–strain curves indicate that the increase in confining pressure promotes the ductile property and the peak strength of specimens. However, the deformation characteristics were affected by the stress–strain hysteresis, presenting a lack of gradation. The relationship between the axial stress and lateral stress can be precisely described by an exponential criterion. Two failure modes were observed during crack propagation: shear failure and mixed failure. The transformation between shear sliding and mixed failure was controlled by the ratio of the friction and cohesion properties. Moreover, the damage plane originates from the flaw at the higher angle of inclination. The joint geometries and loading conditions determine the process of crack propagation. Additionally, the deviation of the failure path was studied to illustrate the damage evolution. Acknowledgements This work was supported by the National Nature Science Foundation of China (Grant No. 51439003), Natural Science Foundation of Hubei Province (Grant No. 2015CFA140), Fundamental Research Funds for the Central Universities (Hohai University) (Grant No. 2016B43014), Open Fund of Key Laboratory of Geological Hazards on Three Gorges Reservoir Area (China Three Gorges University) (Grant No. 2015KDZ12) and Key Project of Scientific Research Plan of Hubei Provincial Department of Education (Grant No. D20181202). Conflict of interest statement: We declare that we have no financial and personal relationships with other people or organizations that can inappropriately influence our work, there is no professional or other personal interest of any nature or kind in any product, service and/or company that could be construed as influencing the position presented in, or the review of, the manuscript entitled, “The mechanical behaviour of rock specimens with specific edge crack distributions under triaxial loading conditions”. References Afolagboye L.O. , He J. , Wang S. , 2018 . Crack initiation and coalescence behavior of two non-parallel flaws , Geotechnical and Geological Engineering , 36 , 105 – 133 . Google Scholar Crossref Search ADS WorldCat Brown E.T. , Hoek E. , 1980 . Underground Excavations in Rock . CRC Press , Boca Raton, USA . Google Preview WorldCat COPAC Cook N.G.W. , Hodgson K. , 1965 . Some detailed stress‐strain curves for rock , Journal of Geophysical Research , 70 , 2883 – 2888 . Google Scholar Crossref Search ADS WorldCat Feng P. , Dai F. , Liu Y. , Xu N.W. , Du H.B. , 2019 . Coupled effects of static-dynamic strain rates on the mechanical and fracturing behaviors of rock-like specimens containing two unparallel fissures , Engineering Fracture Mechanics , 207 , 237 – 253 . Google Scholar Crossref Search ADS WorldCat Hoek E. , Wood D. , Shah S. , 1992 . A modified Hoek – Brown failure criterion for jointed rock masses . In Rock Characterization: ISRM Symposium, Eurock'92 , Chester, UK , 14–17 September 1992 ( pp. 209 – 214 ). Thomas Telford Publishing . Google Preview WorldCat COPAC Huang R.Q. , Huang D. , 2014 . Evolution of rock cracks under unloading condition , Rock Mechanics and Rock Engineering , 47 , 453 – 466 . Google Scholar Crossref Search ADS WorldCat Jie L. , Jian-Lin L. , Liang T. , Jing H. , Liu-Jun F. , Jun X. , 2014 . Study of prediction model for milliseconds deformation rate in triangular wave unloading segment of calcareous sandstone , Rock and Soil Mechanics , 35 , 63 – 70 . ( in Chinese ) WorldCat Jie L. , Lizhou Y. , JIanlin L. , Tengfei Y. , Zhenmin Y. , Fei W. , 2017 . Study on the asynchronous phenomenon of the stress rate and strain rate under cyclic loading and unloading tests , Chinese Journal of Rock Mechanics & Engineering , 36 , 29 – 42 . ( in Chinese ) WorldCat Jin J. , Cao P. , Chen Y. , Pu C. , Mao D. , Fan X. , 2017 . Influence of single flaw on the failure process and energy mechanics of rock-like material , Computers and Geotechnics , 86 , 150 – 162 . Google Scholar Crossref Search ADS WorldCat Le H. , Sun S. , Kulatilake P.H.S.W. , Wei J. , 2018 . Effect of grout on mechanical properties and cracking behavior of rock-like specimens containing a single flaw under uniaxial compression , International Journal of Geomechanics , 18 , 04018129 . Google Scholar Crossref Search ADS WorldCat Li X. , Cao W. , Zhou Z. , Zou Y. , 2014 . Influence of stress path on excavation unloading response , Tunnelling and Underground Space Technology , 42 , 237 – 246 . Google Scholar Crossref Search ADS WorldCat Morgan S.P. , Johnson C.A. , Einstein H.H. , 2013 . Cracking processes in Barre granite: fracture process zones and crack coalescence , International Journal of Fracture , 180 , 177 – 204 . Google Scholar Crossref Search ADS WorldCat Panda B.B. , Kulatilake P.H.S.W. , 1999 . Effect of joint geometry and transmissivity on jointed rock hydraulics , Journal of Engineering Mechanics , 125 , 41 – 50 . Google Scholar Crossref Search ADS WorldCat Park C.H. , Bobet A. , 2010 . Crack initiation, propagation and coalescence from frictional flaws in uniaxial compression , Engineering Fracture Mechanics , 77 , 2727 – 2748 . Google Scholar Crossref Search ADS WorldCat Ray S.K. , Sarkar M. , Singh T.N. , 1999 . Effect of cyclic loading and strain rate on the mechanical behaviour of sandstone , International Journal of Rock Mechanics and Mining Sciences , 36 , 543 – 549 . Google Scholar Crossref Search ADS WorldCat Shen B. , 1995 . The mechanism of fracture coalescence in compression – experimental study and numerical simulation , Engineering Fracture Mechanics , 51 , 73 – 85 . Google Scholar Crossref Search ADS WorldCat Sheng-Qi Y. , Yan-Hua H. , 2018 . Failure behaviour of rock-like materials containing two pre-existing unparallel flaws: an insight from particle flow modeling , European Journal of Environmental and Civil Engineering , 22 , 1 – 22 . Google Scholar Crossref Search ADS WorldCat Song C. , Chunsheng Q. , 2018 . Composite damage constitutive model of jointed rock mass considering crack propagation length and joint friction effect , Arabian Journal of Geosciences , 11 , 283 . Google Scholar Crossref Search ADS WorldCat Wasantha P.L.P. , Ranjith P.G. , Viete D.R. , Luo L. , 2012 . Influence of the geometry of partially-spanning joints on the uniaxial compressive strength of rock , International Journal of Rock Mechanics and Mining Sciences , 50 , 140 – 146 . Google Scholar Crossref Search ADS WorldCat Wong R.H.C. , Chau K.T. , 1998 . Crack coalescence in a rock-like material containing two cracks , International Journal of Rock Mechanics and Mining Sciences , 35 , 147 – 164 . Google Scholar Crossref Search ADS WorldCat Xue L. , Qin S. , Sun Q. , Wang Y. , Lee L.M. , Li W. , 2014 . A study on crack damage stress thresholds of different rock types based on uniaxial compression tests , Rock Mechanics and Rock Engineering , 47 , 1183 – 1195 . Google Scholar Crossref Search ADS WorldCat Yang L. , Jiang Y. , Li S. , Li B. , 2013 . Experimental and numerical research on 3D crack growth in rocklike material subjected to uniaxial tension , Journal of Geotechnical and Geoenvironmental Engineering , 139 , 1781 – 1788 . Google Scholar Crossref Search ADS WorldCat Yanlin Z. , Lianyang Z. , Weijun W. , Chengzhi P. , Wen W. , Jingzhou T. , 2016 . Cracking and stress–strain behavior of rock-like material containing two flaws under uniaxial compression , Rock Mechanics and Rock Engineering , 49 , 2665 – 2687 . Google Scholar Crossref Search ADS WorldCat You M. , 2009 . True-triaxial strength criteria for rock , International Journal of Rock Mechanics and Mining Sciences , 46 , 115 – 127 . Google Scholar Crossref Search ADS WorldCat You M. , 2010 . Mechanical characteristics of the exponential strength criterion under conventional triaxial stresses , International Journal of Rock Mechanics and Mining Sciences , 47 , 195 – 204 . Google Scholar Crossref Search ADS WorldCat Zhang X. , Wong L.N.Y. , 2012 . Cracking processes in rock-like material containing a single flaw under uniaxial compression: a numerical study based on parallel bonded-particle model approach , Rock Mechanics and Rock Engineering , 45 , 711 – 737 . WorldCat Zhenghong H. , Shouchun D. , Haibo L. , Hong Z. , 2019 . Edge crack growth of mortar plate specimens under uniaxial loading tests , Journal of Rock Mechanics and Geotechnical Engineering , 11 , 300 – 313 . Google Scholar Crossref Search ADS WorldCat © The Author(s) 2019. Published by Oxford University Press on behalf of the Sinopec Geophysical Research Institute. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.
TI - The mechanical behaviour of rock specimens with specific edge crack distributions under triaxial loading conditions
JF - Journal of Geophysics and Engineering
DO - 10.1093/jge/gxz059
DA - 2019-10-01
UR - https://www.deepdyve.com/lp/oxford-university-press/the-mechanical-behaviour-of-rock-specimens-with-specific-edge-crack-7HO30tDPts
SP - 962
VL - 16
IS - 5
DP - DeepDyve
ER -