International Journal of Geomechanics

In landslide analysis, reliable prediction of landslide’s influence zone is significantly difficult due to the inherent heterogeneity of soils and their spatially varying geological properties. In this paper, a probabilistic framework is proposed to evaluate the landslide hazard zoning through the prediction of their influence zone considering the effects of heterogeneous soil properties. The material point method is used to simulate the large deformations during landslide failure. The spatial variation in the shear strength parameters of soils at undrained conditions is modeled by random fields, which are discretized by the Cholesky matrix decomposition method to incorporate the effects of the soil spatial heterogeneity on the postfailure deformations. Furthermore, a practical landslide hazard zoning is conducted to quantitively evaluate the level of disaster for facilities or structures located in the vicinity of the slope by using the exceedance probability of influence distance and runout distance. Five categories based on different thresholds of exceedance probability are used to visualize the area potentially affected by the landslide. To demonstrate the capability of the stochastic assessment framework, a benchmark example of the heterogeneous clayey slope is simulated, and the associated hazard zone is predicted and discussed; it presents a practical procedure for improving the landslide influence zone prediction and provides new insights for hazard zoning.

This paper presents the results of a study on the evolution of particle shape in carbonate sands that underwent different degrees of particle breakage. Carbonate sand specimens with three different initial particle sizes were subjected to impact loadings with different input energies. The particle sizes and shapes of the tested sands were analyzed by dynamic image analysis both before and after each loading. An increase in the input energy resulted in increases in particle breakage, aspect ratio, sphericity, and roundness to steady-state values, whereas convexity was barely influenced in the tests. It was found that the Weibull distribution could be used to describe the cumulative distributions of the particle-shape parameters. A relative shape-variation index for quantitatively describing the change in particle shape was correlated with the relative breakage index, which quantitatively describes the extent of particle breakage. Based on microscopic images of the particles at different loading states, the dominant particle-breakage mode gradually transited from the initial fracture mode to the attrition and abrasion mode with an increase in input energy, leading to the production of much more rounded and spherical particles.

Soft-soil reinforcement with stone columns is a popular ground improvement method. Installation of columns modifies the surrounding soft soil properties; this phenomenon is termed “smear.” Although several previous investigations were carried out on smear zone characterization, appropriate field-based studies are limited. The authors performed field-based investigations to quantify the installation effects by collecting a series of undisturbed soil samples from a reinforced soft ground site with newly installed stone columns and conducting laboratory testing followed by nonlinear regression analysis. It is observed that the stone column installation induced significant alteration in the soft soil properties when the columns are closely spaced. An explicit fast Lagrange finite-difference modeling is carried out employing unit cell analogy, where the nonlinear variation in soft clay permeability and compressibility is simulated. The disturbed soft-soil parameters obtained are successfully utilized to conduct numerical modeling. To cross-check, a 2D finite-element modeling by PLAXIS has been carried out as well. It is observed that the results obtained from finite-difference model were in sufficient agreement with the measured field data, compared with the finite-element results. Brief descriptions of the investigation and relevant conclusions drawn are presented in the paper.

This paper aims to show, through numerical parameter analysis on a finite-element method with axisymmetry hypothesis, the long-term effect involving the creep and shrinkage of the concrete lining in deep tunnel problems. The excavation process is simulated by the deactivation and activation method. A viscoplastic constitutive model for the rock mass and an elastic and viscoelastic constitutive model for the lining are used in the software ansys. Perzyna’s viscoplastic constitutive model with an associated flow rule with von Mises yield surface is considered. The viscoelastic model is implemented in ansys using the customization feature. The shrinkage is given by the CEB-FIP MC90 formulation and the creep part is modeled through a Generalized Kelvin chain, according to the solidification theory of Bažant and Prasannan, whose parameters are adjusted with the creep function of the CEB-FIP MC90. Compared with an elastic lining, the viscoelastic line showed a difference (in average terms), in the equilibrium convergence at the end of tunnel construction between 6% and 37%. This value increases to 8% and 72% in the long-term convergence.

In this study, the effects of grain size and moisture content on the interface shear strength of sandy soils against a polypropylene biaxial geogrid were evaluated using a large-scale shearing device. Soil specimens with various maximum grain sizes (2.0–13.2 mm) and moisture contents (5%–11%) were sheared against the geogrid at normal stress values of 250–750 kPa. The interface shear strength was primarily influenced by the soil’s internal shear strength, the friction between the soil and the geosynthetic plane surface, and the resistance provided by the transverse ribs. The results indicated that the shear strength of the soil–geogrid interface decreased as the maximum grain size of the soil decreased. The lower interface shear strength was primarily a consequence of the reduced internal shear strength and friction. The internal shear strength contributed more than 50% of the shear strength of the soil–geogrid interface, whereas the contributions from the friction or transverse ribs were less than 30%. The transverse ribs provided effective resistance when the median of the particle size distribution was greater than the thickness of the transverse ribs. Increasing the soil moisture content reduced the interface shear strength, especially when the moisture content exceeded the optimal moisture content. Increasing the moisture content also negatively impacted the bearing resistance of the transverse ribs.

Road-related sinkholes are of great concerns nowadays, appearing frequently in media stories. In the undrained stability analysis, a dimensionless stability number that is known as Broms and Bennermark’s original stability number expresses the combined effect of surface surcharge pressure, self-weight of the soil, and the supporting pressure. This “critical” undrained stability number, N_c, is used in this paper to formulate three-dimensional (3D) upper-bound and lower-bound solutions in active (collapse) and passive (blowout) scenarios of sinkhole stability using the classic trapdoor problem. Design charts, tables, and equations were developed to cover a wide range of practical design parameters. Together with modern geophysics techniques for the detection of underground cavity as well as the findings of the associated failure patterns in this paper, the study should, therefore, be of value to practitioners in their daily decision making.

Strainburst is the most common type of rockburst in deep excavations, which sometime result in heavy losses. This research studies the feasibility of a conceptualized destressing method for strainburst control. In this method, the rock is destressed by cutting slots at the excavation boundary. First, a circular tunnel is modeled and the stress distribution around the tunnel is calculated. In a separate model, a slot is added to the tunnel wall. Results show that when the slot is added to the model, the rock at the tunnel wall destresses, and the stress concentration zone transfers to a distance farther away from the wall. Also, the analysis of the failure zone around the tunnel and the velocities of the failed element nodes show that the failure in the slotted tunnel is less violent compared with the tunnel without the slot. Secondly, a parametric study is conducted to find the optimum dimensions of the slot. Finally, the rockburst damage in Jinping II drainage tunnel is simulated and then the application of the proposed destressing method for strainburst prevention in this tunnel is studied. Comparison of the modeling results with the field rockburst observations of the drainage tunnel shows that the proposed destressing method can be considered as a potential tool for strainburst control in deep excavations.

Grouting control is one of the key measures to reduce segment floating, to ensure safety and efficiency during shield tunnel construction. In this study, optimization of backfill synchronous grouting material for shield tunnel antifloating in the water-rich full-face rock stratum in Nanchang Metro Line 4 is conducted based on laboratory experiments and in situ application. First, orthogonal tests are designed and conducted to investigate the influence of four main factors, that is, water–binder ratio, binder–sand ratio, cement–fly ash ratio, and bentonite–water ratio, on grouting material properties. Based on this, optimization is proposed using response surface methodology combined with laboratory optimization tests. According to the optimization, an admixture mass proportion (bentonite: 11.81%, hydroxyethyl methylcellulose: 0.42%, nanosilicon: 0.7%) is chosen based on the original grouting material constituents and then verified through both laboratory and field tests. The results show that the optimization can significantly increase the setting time, 1-d unconfined compressive strength, and water–land strength ratio, reducing almost 36% of the uplifting value of the tunnel segment when water-rich full-face rock stratum is encountered in this case.

The effectiveness evaluation of long-hole destress blasting in an underground hard coal mine was made via the seismic effect method and the seismic moment tensor inversion. The seismic effect method is based on the difference between seismic and explosive energies, indicating that additional processes appeared to change the stress state in the rock mass, however, the seismic effect does not consider the focal mechanism. Therefore, the seismic moment tensor inversion was applied. Tremors that were an effect of long-hole destress blasting during the extraction of coal seam No. 507 in one of the hard coal mines localized in the Polish part of the Upper Silesian Coal Basin were analyzed. The results of the preceding applied methods were compared. High values of seismic effect indicate stress relaxation and thus the occurrence of additional processes, leading to a new stress equilibrium state in the rock mass. It has been shown that in analyzed cases, achieving such a state corresponded to the presence of a nonexplosive mechanism, with a large share or domination of the double couple component in the full seismic moment tensor solution. The reverse slip mechanism was present in the foci of provoked tremors, which can be associated with the displacement of rock blocks under conditions of high horizontal stress. Convergent results of the methods used confirm that the appropriate design of blasting parameters in relation to specific geological and mining conditions make it possible to provoke additional geomechanical processes in the rock mass.

Disc cutters are a key component of a tunnel boring machine (TBM) for rock rolling and crushing, yet their high cost and complicated maintenance process generally increase the construction cost of a project significantly. Therefore, understanding the pattern of disc cutter wear helps to not only improve TBM tunneling efficiency but also reduce construction costs. In this study, the cutter wear data were collected from a length of 4,300 m of TBM tunneling from the No. 2 bid section of the TBM construction of the Yellow River Diversion Project in central Shanxi Province, China. First, the influence of the lithology along the tunnel on the cutter wear rate was analyzed and, in particular, the relationship of the rock elements and the mineral content with the cutter wear rate was further analyzed to investigate the influence of the surrounding rock type on the cutter wear rate. Second, the influence of the lithology on the cutter failure mode was examined. The results showed the following: The cutter wear rate was the highest (1.72 mm · m⁻¹) in the quartzite and the lowest (0.22 mm · m⁻¹) in the dolomite. The cutter wear rate increased linearly with the increase of the quartz content in the quartzite. A higher surrounding rock type led to a higher cutter wear rate. When the joints and fissures of the rock mass were developed, the cutters had a high probability of abnormal wear, which accounted for a proportion of 80% for the dolomite lithology. Finally, some measures were put forward to reduce the cutter wear for TBM excavation with different lithologies, which could be used as a reference for similar projects.

This study analyses the classical problem of the ultimate lateral pressure (ULP) of a circular pile in undrained clay considering the undrained strength reduction induced by pile installation. A modified Tresca model considering strain softening proposed by Einav and Randolph is used to describe the variation in soil strength. The pile installation effect is evaluated through the cylindrical cavity expansion model (CEM). An analytical solution for CEM considering the undrained strength reduction is derived by introducing the modified Tresca model. The analytical solution is validated by comparison with the theoretical prediction and the three-dimensional (3D) large deformation finite-element results through CEL in ABAQUS. Subsequently, the obtained analytical solution for the radially nonuniform distribution of undrained strength is read into Optum G2 to conduct a two-dimensional plane strain finite-element limit analysis (FELA). A series of parametric studies are performed to investigate the influence of strain softening on the ULP. A closed-form solution for ULP considering the pile installation effect is given through the regression analysis of FELA data. The present analysis first considers the pile installation effect on ULP and could provide a basis for constructing the p–y model of a laterally loaded pile considering the pile installation.

Freeze–thaw cycles induced deterioration on rocks begins with the damage in microstructures such as pores and pore throats. This paper was devoted to providing a fundamental understanding of the microscopic deterioration characteristics of anhydrite rock exposed to freeze–thaw attacks by using a series of laboratory tests. Meanwhile, the relation between the microstructure evolution and the mechanical resistance deterioration was revealed. Results indicated that with the increase of freeze–thaw cycles, the proportion of micropores (r < 0.1 μm) and the pore throat radius within 0–0.1 μm (PT-I) decrease exponentially, while the proportion of mesopores (0.1 μm < r < 1 μm), macropores (r > 1 μm), the pore throat radius within 0.1–4 μm (PT-II), and porosity increase exponentially. Meanwhile, the pore fractal dimension and pore throat fractal dimension decrease exponentially with the freeze–thaw cycles. As the freeze–thaw cycles increase, the surface roughness and the area of micropores increase. It is characterized that the expansion and dissolution of anhydrite minerals in the water plays a crucial role in the morphology evolution. With the increase of freeze–thaw cycles, the mass variation increases, the average uniaxial compressive strength (UCS) and elastic modulus decrease exponentially. Furthermore, the average UCS and elastic modulus decrease exponentially with the damage variable represented by the porosity and increase exponentially with the pore structure fractal dimension. Finally, it is revealed that the proportion of the PT-II plays the leading role in the mechanical resistance evolution of anhydrite rock under freeze–thaw cycles. The experimental results presented are promoting the comprehending of the microscopic deterioration of anhydrite rock projects in cold districts.

This paper presents a model for compression and wetting-induced collapse behavior of unsaturated soils using the disturbed state concept. The volume change and collapse behavior of unsaturated soils were described using their compression behaviors in two reference states, including unsaturated condition before collapse as an initially intact state and saturated condition after wetting-induced collapse as a fully adjusted state. Using a disturbance function derived from experimental results, the collapse behavior of the soil was related to its behaviors in two references states. A power form function was proposed to describe the volume change of the soil during the collapse stage. Using a piecewise continuous disturbance function, the compression and collapse behavior of unsaturated soil could be expressed in the full range of loading, including before-collapse, during-collapse, and after-collapse stages in the effective stress and net stress space. A series of laboratory tests were conducted on three different types of collapsible soils to verify the proposed model. Verification of the model with laboratory results and the data available in the literature showed the ability of the proposed model to describe the compression and the wetting-induced collapse behavior of unsaturated soils before, during, and after the wetting-induced collapse.

Piping is a major failure mechanism in both dams and dikes. In this work, we develop a transparent specimen tank that allows a camera to clearly record the piping formation process. In addition, this apparatus is able to maintain a constant hydraulic head and to apply K0 stress states. The eroded soil and the discharged water are measured using a collection system. We conduct a series of laboratory tests to investigate the evolution process and mechanisms of piping in gap-graded soils under a K0 stress state. Our results show that the piping process can be divided into three stages: seepage, pipe formation, and pipe wall erosion. No visible soil erosion and deformation is observed in the seepage stage. In the pipe formation stage, the flow channel is formed gradually from the position where water is injected into the exits. In the pipe wall erosion stage, grains are eroded mainly from the pipe wall. Moreover, erosion is more profound on the convex side than on the concave side of the pipe. From the sensitivity analysis, a large percentage of coarse grains can hinder the formation of the flow channel is observed. Finally, the transition mechanism from suffusion to piping based on different testing conditions is revealed.

Damping is an important dynamic soil property that has been well-studied on the continuum scale. However, particle-particle interactions that are responsible for damping in dry soil are relatively less studied due to the difficulties in accessing energy loss at particle contacts. This study attempts to evaluate contact energy dissipation numerically on the particle scale. A model based on the discrete element method (DEM) was first developed and validated to simulate the cyclic torsional shear test. A parametric study was then carried out to study the distribution of contact energy dissipation inside the soil specimen. The effects of shear strain level, confining pressure, and specimen void ratio were investigated. The results show that, as shear strain increases, contact sliding and energy dissipation first occur in the weak-force network and propagate towards the strong-force network. At the same shear strain level, an increase in confining pressure results in a decrease in the number of sliding contacts and a higher energy loss per contact. At the same confining pressure and shear strain level, an increase in specimen void ratio results in less total energy loss and a smaller percentage of sliding contacts in the specimen. It is also found that particle contacts at the top one-third of the specimen have a higher tendency to slide and higher energy loss per contact and, therefore, contribute to the majority of the total energy loss in the specimen.

This paper presents Cerchar scratch tests performed on sandstone samples with a constant thrust force of 70 N. The scratch test is reproduced using discrete element method (DEM) models. Simulated scratching force–(depth-) displacement curves evolve in the same way as those of experimental results. Numerical scratching force is also comparable with laboratory results. All these accordance indicate that the presented model can represent the experiments. Microscopic observation of the scratch indicates pullout of grains and the associated tensile failure dominates the scratching process. Particle detachments and microcrack development in numerical simulations also verified this tensile-dominated failure mode. Compared with reported rock cutting studies with fixed cutting depth mode, the cutting force shows a much smaller variation in this fixed thrust force mode. This finding indicates that the applied thrust force plays a significant role in the cutting process, rock failure, and the tool–rock interaction. Factors that may result in a much larger scratch depth in simulations are carefully discussed. A parametric study was conducted to investigate the effect of scratching velocity, amplitudes of indentation velocity, particle sizes, and applied thrust load on the scratch behavior.

The present study delves into the propagation phenomenon of shear waves in a layered structure comprised of reduced Cosserat stratum imperfectly bonded to a reduced Cosserat substrate having abrupt thickening at the interface. The dispersion equation of the shear waves has been deduced in algebraic form by employing Fourier transformation along with a perturbation method. Moreover, the impact of differently shaped abrupt thickening existing at the imperfect interface on the frequency of shear wave has been studied numerically and demonstrated through graphing. A comparative study is carried out through graphical illustrations to highlight the concealed behavior of shear waves in response to various affecting parameters, such as an abrupt thickening parameter and imperfect bonding parameter. The present study provides a theoretical framework in the study of rotational seismology, lithosphere–asthenosphere boundary, and acoustic metamaterials.

The investigation of fluid flow in fractured rocks is a key issue in underground engineering. Three sandstone specimens with different roughnesses were generated by Brazilian splitting, and the geometric morphology and aperture distribution of specimens were obtained by three-dimensional scanning technology. Fractal dimension was introduced to characterize the roughness of fractures. Seepage experiments in rough fractures were conducted to analyze the influence of fractal dimension and contact ratio on the nonlinear flow behavior subjected to different normal stresses, which proves that the Forchheimer equation can better describe the flow nonlinearity. A modified Bandis model was proposed to calculate the max normal compression displacement of fracture under normal stress. Subsequently, a new model was developed to forecast the nonlinear coefficient based on fractal dimension and contact ratio. Moreover, a semiempirical equation was established to describe the critical Reynolds number. The influence of contact areas on the seepage path was simulated by COMSOL.

The morphology of fragments produced by grain fragmentation is a cause for great concern in many areas such as rockfill dam and the mining industry. Therefore, there is a need to conduct a quantitative analysis of the morphology characteristics of fragments produced by rock grain crushing. So far, few studies have focused on the qualitative analysis of fragment morphology, and quantitative studies are rare. This study presents a quantitative investigation of the morphology characteristics of fragments produced by the crushing of two groups of rounded and angular rock grains. The intact grain morphology is acquired using a 3D scanner, and the grain fracture and fragmentation are simulated using the combined finite-discrete element method (FDEM). The mesh refinement study indicates that a finite-element mesh size that leads to the fracture process zone to be discretized into at least five or six elements is used to eliminate or minimize the mesh size effect. Regular and spherical grains tend to get smaller largest fragments and larger second-largest fragments compared with grains with lower sphericity and convexity because of different fracture modes. Small differences have been observed in the distributions of fragment shape descriptors for grains in two groups. However, no monotonic relationship has been found between the distribution of shape descriptors and the fragment size, except sphericity. The distributions of sphericity for different size groups clearly suggest that larger-sized fragments tend to be more angular, which is consistent with the existing results.

This paper presents micro–macro behaviors of unsaturated residual soils within high suctions. A series of unsaturated soil specimens was prepared using the saturated-saline-solution-based vapor equilibrium technique, considering different initial dry densities. The strength of these specimens was tested under high suction using the direct shear test. Micropore structures for residual soil tests were analyzed using mercury intrusion porosimetry and scanning electron microscopy. The macromechanical strength of residual soil under direct shearing and unconfined compressing conditions was improved by increasing the suctions and densities. And, the level of pore volume distribution measured in microscale on unsaturated residual soil was greatly dependent on matric suction and specimen density. Moreover, soil–water characteristic curves greatly decreased and trended to flat in the high-suction condition (more than 50 MPa).

This paper presents a numerical study of the mechanical properties of a binary mixture of similar prolate ellipsoids of various particle shapes and fines’ contents using the discrete element method. The ratio of the major axis of a small ellipsoid to that of a large ellipsoid is 0.2. Because all particles have the same shape, the particle shape of the binary mixture is defined unambiguously. The effect of fines content is examined with samples of the same particle shape. The effect of particle shape is examined by comparing the samples of different particle shapes. Very dense samples are generated and sheared under drained and constant volume conditions. The macroscopic and microscopic behaviors of these samples are studied at both peak and critical states. The mechanical behavior can be grouped into two or three regions for our binary mixture depending on the fines content. Although the critical state line varies with fines content, a unified critical state line is identified using the intergranular void ratio, which is a function of fines content and a parameter. The parameter is a constant for the samples of the same particle shape. This constant is different for the samples of different particle shapes. The finding explains why some studies defined the parameter as a function of fines. The intergranular void ratio does not relate well to the peak friction angle. The microscopic behavior is examined through contact normal, particle orientation, and stress partition. Although some microscopic parameters can identify the region where the sample belongs, we do not find any dominant microscopic parameter that relates to the macroscopic behavior of a binary mixture of various fines contents.

This paper presents a generic analytical solution for stresses and displacements developed around an arbitrarily shaped underground opening excavated in an elastic soil/rock mass with biaxial in situ stresses. A series of separable functions satisfying the governing biharmonic equation is employed as the basic stress function for the problem considered. With the basic stress function and the stress rotation equation, the normal stress and the shear stress on the opening boundary surface are expressed in terms of separable function series with unknown coefficients to make use of the boundary condition. Based on the variational theory, the unknown coefficients are determined by minimizing the difference between the real stress boundary condition and the assumed stress function on the opening surface. The excavation responses around different shapes of openings, including elliptical-shaped, ovaloid-shaped, horseshoe-shaped, and D-shaped openings, are investigated and compared with those from the finite-element simulations. The results show that both the stress field and the displacement field agree well with those from the corresponding numerical model, which verify the proposed solving technique. The proposed solution provides an easy, generic, and feasible approach to assess the stresses and displacements developed around an arbitrarily shaped opening, which is simpler and more straightforward than the complex theory–based solutions.

Swelling pressure (SP) is one of the most basic properties of bentonite and bentonite-based materials. To date, extensive models have been proposed for predicting SP. However, most of the existing models are aimed at predicting the final steady SP at saturation, ignoring the development process of SP during hydration. In this study, the SP was considered as a result of competition between the accumulated wedging pressure (AWP) and dissipated wedging pressure (DWP) within the specimen. By assuming that the AWP increases continuously with infiltration time following an exponential function and that the DWP first increases and then decreases with infiltration time obeying the Weibull distribution, a new empirical model for modeling the SP kinetics was established. The model contains only five parameters, which all have definite physical meanings and are susceptible to various factors including the mineral composition, dry density, specimen dimensions, infiltrated solution concentration, and temperature. The model was verified with three types (I, II, and III) of experimental SP kinetics on bentonite and bentonite-based soils. Results show that the model is capable of modeling the SP kinetics of Types I and II, while for SP curves of Type III, the simulated results deviate significantly from the experimental ones due to the model ignoring the second decrease of SP induced by montmorillonite dissolution in the high alkaline solutions. Furthermore, a four-step simplified procedure for predicting SP kinetics at different dry densities was put forward and verified by experimental results of MX-80 bentonite pellet/powder mixtures and compacted GMZ bentonite.

The stability analysis of a weakly inclined subgrade is a crucial topic of geotechnical engineering. This work established a novel method based on the Swedish method to solve the stability analysis of a weakly inclined subgrade. Centrifuge model tests were used to examine the failure mechanisms of a weakly inclined subgrade. Then, a series of two-dimensional (2D) and three-dimensional (3D) numerical simulations analyses were performed to find the influencing factor affecting the weakly inclined subgrade’s stability. According to the experimental and numerical simulation results, there were obvious differences in the instability failure mode between the weakly inclined and the weakly horizontal subgrades. In the novel method, the mechanism that caused the instability of the weakly inclined subgrade was studied by introducing influential force and additional force. While the influential force suggested the impact of the potential difference, the additional force indicated the gravity impact emanating from the lateral embankment. After contrasting the results in the engineering example, the proposed method’s viability was confirmed. The proposed method was found more suitable for performing the stability analysis of a weakly inclined subgrade than the traditional slice methods.

The displacement-based back-analysis is an effective approach to determine the values of soil parameters of a slope. However, the applicability of different methods of the displacement-based back-analysis varies. This paper divides the displacement-based back-analysis into the deterministic back-analysis under frequentist inference and the probabilistic back-analysis under Bayesian inference. A framework for the deterministic back-analysis is proposed using the maximum likelihood estimation, which is a typical method of frequentist inference. In the framework for the deterministic back-analysis, the dual annealing (DA) algorithm is applied to search for the globally optimal solution. A framework for the probabilistic back-analysis considering spatial variability is proposed based on the Bayesian theory, random field theory, and truncated Karhunen–Loève expansion (KLE), which overcome the “curse of dimensionality.” It is explained that this framework will also work in the probabilistic back-analysis based on the random variable model. In the two frameworks, metamodels are constructed by gradient boosting decision trees (GBDT). This algorithm perfectly fits the relationship between parameters and displacements of a slope, which replaces the time-consuming finite-element simulations. The adaptability of the two proposed frameworks is illustrated with a real case study of a highway slope. The case study also proves that the fitting accuracy of metamodels constructed by the GBDT algorithm is higher than those constructed by the neural network (NN) and the random forests (RF). Processes of frequentist inference and Bayesian inference show that the difference in their results originates from the different perceptions of the parameter space. However, the two different back-analysis results are not in competition but should be complementary. They both play an important role in slope parameter determination and slope stability analysis.

The particle shape plays an important role in controlling the behavior of granular material, thus needing to be considered in the formation and evolution of soil arching in the pile-supported embankment. Based on the three-dimensional discrete-element method simulation, the effect of the soil particle shape on the formation and evolution of soil arching in the pile-supported embankment is unraveled from the construction to the operation period by adopting spherical, oval, and tetrahedron particles. An additional simulation of spherical particles with the rolling resistance contact model is conducted to reveal the applicability of the indirect method to reproduce the behavior of the case with irregular particles. After sample preparation, four simulation procedures are applied for each case: differential settlement, static loading, cyclic loading, and final equilibrium. The results indicate that due to interlocking, smaller surface settlement occurs for the case with irregular particles at a given simulation state. Furthermore, the case with irregular particles tends to induce a more significant soil arching than the case with spherical particles, also showing a higher resistance to the degradation of soil arching under external load. Owing to various homogeneity and angularity, the two cases with irregular particles present different mechanical patterns under the external load. The reorientation and destruction of the contact force network are the microscale reasons for the formation and degradation of soil arching. As the rolling resistance method cannot fully reproduce the behavior of irregular particles for the soil arching, this method should be carefully validated and used in further simulation of the pile-supported embankment.

Gently inclined and inclined coal pillars are both subjected to the combined action of compression and shear load. Due to the existence of inclination angle, the fracture and instability behaviors of inclined pillars are different from those of near-horizontal pillars, and the difference is closely related to the load level of the pillar. Therefore, it is of great significance to study the influence of different inclination angles and loading rates on mechanical property and fracture behavior of the coal pillar for its strength design. In this paper, a new rock mechanical properties testing system was used with the combination of acoustic emission (AE) technology, and the changes of mechanical properties of coal specimens under different inclination angles and loading rates were obtained. The main conclusions are as follows: (1) the large of inclination angle, the greater of the peak shear stress. When the inclination angle is small (0°, 5°, 10°, and 15°), the peak stress increases first and then decreases with the increase of loading rate, when the inclination angle is large (20° and 25°), the peak stress gradually decreases. When the inclination angle is greater than 15°, the peak stress decreases with the increase of loading rate, and the decreasing amplitude gradually decreases. (3) The failure mode of the coal specimen is changed from tensile-splitting failure (0°–5°) and tension–shear composite failure (10°) to single shear failure (15°–25°). The loading rate has little effect on the failure mode of the specimen, but a significant effect on its failure degree. The crack initiation (CI) threshold and the crack damage (CD) threshold of coal specimens increase first and then decrease with the increase of inclination angle, and the ratios of CI and CD to peak strength are almost constant. With the increase of loading rate, the ratio of CI and CD to peak strength increases first and then decreases in the inclination angle range of 10°. The ratio of CI and CD to peak strength is basically independent of the loading rate. To comprehensively consider the influence of inclination angle and load level on pillar strength, a new parameter of rock strength under the influence of inclination angle and load level is suggested to be introduced into the empirical formula of pillar strength.

This study investigates the utilization of alkali-activated granular blast furnace slag (GBFS) and fly ash (FA) for stabilizing rammed earth materials and assessing their mechanical and durability properties. Twenty mixtures were designed using the central composite design (CCD) of response surface methodology (RSM). The clay content of the soil, NaOH molarity, and GBFS to the binder ratio were selected as independent variables. Density, water absorption, 28-day compressive strength, and compressive strength loss after 60 days of exposure to sulfuric acid were measured as responses. Also, the microstructure of the specimens and the effects of the sulfuric acid attack were investigated by SEM and XRD analysis. The results revealed that an increase in the clay content of the soil reduced the compressive strength of specimens. Also, higher content of FA than GBFS in binder led to more acid-resistant specimens for mixtures containing a low amount of kaolin clay. On the other hand, for specimens with a high amount of kaolin clay, a higher proportion of GBFS than FA in binder resulted in more acid-resistant specimens. In addition, an increase in NaOH molarity and GBFS content in the binder was accompanied by an improvement in compressive strength. SEM investigations showed more homogenous and compact reaction products for these mixtures. Finally, a numerical optimization method was implemented to obtain mixtures with the highest compressive strength and minimum compressive strength reduction after acid exposure.

Installation of granular columns is a cost-effective and versatile in situ technique to improve the shear strength, settlement, and drainage behavior of weak soils. It involves backfilling vertical boreholes in the ground with granular materials stiffer than the native soil, such as stone or compacted sand. However, the massive use and overexploitation of sand and natural aggregates have depleted their reserves in recent decades, causing damage to the environment, creating sand shortages, and skyrocketing their price. Hence, it is essential to develop a sustainable alternative to natural aggregates to construct granular columns. The ever-increasing stockpiles of waste glass could be a potential replacement for natural sand in several geotechnical construction applications, noting that both materials have a similar chemical composition. Using crushed waste glass (CWG) as an alternative to traditional natural and manufactured (quarried) sands in granular columns could offer a multipronged benefit by recycling nonbiodegradable waste (glass) and by conserving a depleting natural resource (sand). Using a large direct shear (LDS) machine, this study investigated the shear strength behavior of kaolin (to represent a typical weak soil) reinforced with a central granular column. Three different materials were separately used to backfill the column, including natural sand (NS), manufactured sand (MS), and CWG. The results revealed that the geocomposites containing the CWG column have the highest peak friction angle and relatively greater shear strength under high normal stresses, favoring the potential use of CWG as a green alternative to traditional sands in backfilling granular columns, ultimately supporting resource conservation, waste recycling, and the paradigm shift toward a circular economy.

This paper presents an investigation into the influence of the size of nongrouted anchor nails on the role of anchorage capacity in a flexible protected system. A lateral pullout test was used to examine the deformation, pullout capacity, and failure mode of anchor nails with different lengths and diameters. There could be three failure modes, namely rotation failure, bending failure, and tension failure. The theoretical models were established for different failure modes, and a corresponding solution for critical pullout force was derived. Moreover, the applicability of the theoretical models was verified by the consistency between the measured value and calculated value, which seems to show that the models could predict the displacement and the resistance of anchor nails under lateral force. With a fixed diameter, an increase in the length of the anchor nail contributes to a longer bending length, but the ratio of bending length to total length is consistent with the variation of the ultimate pullout capacity—both of which increase first and then decrease. On the contrary, length unchanged, the bending length and the ratio of bending length to total length decrease with a larger diameter, while the ultimate pullout capacity increases.

In optimizing the design of retaining structures and monitoring their health, it is important to determine the actual nonlinear slip-failure surface and the associated nonlinear lateral stress distribution. This paper presents a model developed for the nonlinear geometry of active and passive slip-failure surfaces in cohesionless soils and for determining their three main associated variables; that is, lateral earth pressure distribution, coefficient of lateral earth pressure, and location of the resultant lateral force. The variational limit-equilibrium method, as applied to a normally consolidated dry granular media, and the plane-strain critical-state friction angle at failure are used to develop the model. The model outputs the governing geometry of the slip-failure surface and its associated lateral stress distribution as a unique nonlinear function of the ultimate shearing resistance at failure. Mostly existing studies are complex unilateral approaches of the lateral stress or the slip-failure geometry as separated issues. In contrast, the present paper addresses a simple coupled solution at critical state that uses a disambiguated friction angle and avoids arbitrary input assumptions such as the geometry of the slip-failure surface.

Rocks exhibit different failure strengths under triaxial compression and extension. The understanding and quantification of their relationship provide theoretical and practical benefits. Conventional Mohr-type criteria cannot distinguish rock failure strengths under triaxial compression and extension. To overcome this limitation, we present an analytical model based on Mohr failure theory to describe the relationship between the rock failure strengths under triaxial compression and extension in a Mohr plane. Our analytical model indicates that rock failure criteria under both triaxial compression and extension can be expressed in power-law forms. The corresponding strength magnitude and Mohr envelope curvature parameters are analytically determined using measured rock properties from uniaxial tension, uniaxial compression, and equibiaxial compression tests. This model is validated using experimental data of five rock types: granite, marble, basalt, dolomite, and sandstone. Compared with four well-known failure criteria, the proposed model demonstrates the ability to predict the failure strengths of rocks under both triaxial compression and extension.