International Journal of Geomechanics

A new analytical method for investigating the stability of unsaturated soil slopes was developed by considering a 3D failure wedge under different steady-state infiltration and evaporation. In this light, the general limit equilibrium approach and concept of effective stress in unsaturated soils were implemented. An extensive parametric study was performed as well, to study the effect of slope geometry, unsaturated soil properties and/or state/material variables, position of groundwater level (GWL), and infiltration and evaporation flow rates on the stability of soil slopes against failure and landslide. In addition, a quantitative comparison was made between the safety factors obtained from the proposed unsaturated 3D analysis and those obtained from two-dimensional (2D) and three-dimensional (3D) classical slope stability analysis. It was revealed that the steady-state infiltration/evaporation has a significant influence on the stability of the silty and clayey soil slopes. In addition, the geometrical aspects of the slope and the hydromechanical parameters of the soil largely affect the safety factor of soil slopes against failure. Results show that the safety factors obtained from unsaturated 3D analyses for the silty and clayey slopes are remarkably different from those obtained from 2D and 3D classical calculations; however, negligible differences were found for the sandy slopes in that respect. In addition, a comparative study was performed on the impact of governing parameters on the safety factors of unsaturated slopes under infiltration and evaporation using an unsaturated 3D column method. It was revealed that infiltration has a pronounced effect on the safety factor compared with that for evaporation and that silty and clayey slopes are more affected by evaporation and infiltration than sandy slopes. In general, this study showed that applying unsaturated soil mechanics concepts to 3D stability analysis not only leads to more precise safety factors against the slope failure and landslide initiation but may also reduce the uncertain conservatism considerations in some practical problems.

Compensation grouting has been widely applied for the vertical deformation control of constructed facilities. However, few studies have been conducted on the horizontal displacement control of underground facilities using grouting. A field test of compensation grouting was conducted in a greenfield site to investigate the mechanism of horizontal displacement control. Both horizontal displacement and pore-water pressure of soil were monitored. A sophisticated finite-element model was developed and verified by field measurements. A comparison of various models showed that it was necessary to consider the nonuniform expansion of the conditioned zone for displacement prediction and the fissures for the simulation of pore pressure dissipation. The parametric study showed that the grouting-induced horizontal displacement of the silty clay could be predicted by a power model. The displacement decreased to 10% of the maximum value at a distance of 3 m, which could be defined as the effective controlling range of grouting. A hyperbolic model was proposed to calculate the efficiency of horizontal displacement. The efficiency of horizontal displacement reached 90% of its limit value when the grouting volume was approximately 2.3 m³, which could be regarded as an optimum grouting volume in practice.

During deep excavation in weak rock masses, the time-dependent deformation of the rock and longitudinal advancement process have significant influences on the tunnel performance. Accordingly, a combined support system consisting of rockbolts and linings is commonly used in such tunnels to ensure their stability. However, there are no well-established solutions for quickly and accurately evaluating the safety of those support systems. To address this problem, mechanical model of a circular tunnel with a rockbolt–lining combined system is analytically built in this study. Then mathematical formulas for the stresses and displacements around the tunnel are derived, considering the time-dependent behaviors of the rocks and tunnel face advancement effect. Furthermore, the analytical formula proposed in this study is compared with existing solutions and numerical results obtained using the finite-element software Abaqus, and a good agreement is achieved. Finally, a parametric investigation is conducted regarding the influences on the tunnel’s mechanical performance. The results show that the installation time (t₀) and distribution density of rockbolts (S_θS_z) and the installation time of the lining (t₁) have a significant influence on the deformation of the surrounding rock. Although the installation times of bolt and lining are positively related to the deformation suppression effect of the surrounding rock, they are negatively related to the stress of the bolt and lining. It is found that the closer the installation time of bolt and lining, the greater the stress of lining; otherwise, the greater the stress of bolts. Furthermore, the initial release coefficient (m) and the ratio of tunnel excavation rate to influence radius can influence the deformation rate of surrounding rock. Finally, the study proposes a new method for quickly evaluating the viscoelastic deformation of a tunnel.

Granular columns are widely used as a soil improvement solution adopted in the case of soft clays in an effort to increase their bearing capacity and stiffness. Granular columns, whether sand drains or gravel columns, are designed and constructed in groups and not as individual elements, given that there is a clear need to study the group effects and load sharing under various drainage conditions. Toward this need, a new fully instrumented triaxial test setup was developed and utilized to investigate the overall response of the composite. Particular emphasis was placed on monitoring pore-water pressures and contact stresses over the sand columns and the surrounding clay. The results from clay specimens that were reinforced at area replacement ratios of 17.1% and 30.4% indicate that the test setup is capable of quantifying the distribution of the stresses in the columns and the surrounding clay at different levels of axial strain. As such, previous information about the dependency of the stress concentration ratio on the area replacement ratio, axial strain, rate of loading, and drainage conditions could be inferred. It is anticipated that the insights gained and reported in this paper will facilitate the development of design methodologies for soft clays reinforced with sand column groups.

The mechanical response of natural clays can exhibit significant directional dependence due to loading history and layered soil structure. We report an experimental investigation of the stress–strain-strength response of undisturbed Shanghai clay subjected to undrained hollow cylinder shear that covers a wide range of principal stress rotation angles and intermediate principal stress ratios. The three-dimensional rotational hardening model proposed by Huang et al. ( 2011) is used to interpret experimental observations. Test data indicate that the undrained strength first decreases and then increases as the major principal stress direction rotates away from the soil deposition direction. This nonmonotonic trend primarily results from different shear-induced pore pressure responses and cannot be explained by loading history-induced anisotropy. These findings suggest the presence of mechanical anisotropy of the tested natural clays that is related to the laminated soil structure. The measured undrained strength, on the other hand, monotonically decreases as the intermediate principal stress ratio increases. This variation is mainly due to the dependence of strength parameters on shear modes. We present a simple approach to include anisotropy related to the laminated soil structure into the constitutive model. The comparison between measured and simulated results validates the new proposition.

An improved vacuum preloading method combined with a sand sandwich structure (SSS) was developed to facilitate the consolidation of dredged clay-slurry fill and original soft marine clay for land reclamation. In this method, a novel sandwich structure is formed as a sand layer introduced on to original soft marine clay prior to the placement of reclaimed, dredged clay-slurry. This assists in transmitting the vacuum pressure to both the reclaimed, dredged clay-slurry and the original soft marine clay, through prefabricated vertical drains, with little vacuum loss, and the sand layer thereby improving the efficiency of the vacuum preloading. Field pilot tests were carried out on improved ground both with and without using a SSS in order to demonstrate the capacity of the improved vacuum preloading method combined with the SSS. Field measurements were performed to evaluate the ground settlement, degree of vacuum loss, effective consolidation stress, shear strength, and bearing capacity. The test data showed that better consolidation results were achieved using the improved vacuum preloading method combined with the SSS.

In this work, the stability of a reinforced tunnel face is examined by combining the homogenization method with the kinematic approach of limit analysis. A three-dimensional (3D) heterogeneous and anisotropic failure criterion for reinforced soils is first proposed according to the homogenization method, in which the possible tensile failure and pullout failure of reinforcing members are considered. The proposed homogenization method is then incorporated into the 3D rotational failure mechanism to assess the safety factor of a tunnel face by using the strength reduction technique. The applicability of the proposed method is illustrated by comparing with existing results and numerical simulations. The influences of model parameters on the calculated safety factors are finally discussed.

Using the classical (linear) Mohr–Coulomb (M–C) failure criterion, the failure mechanism of slopes is commonly treated as a completely shear failure. However, the tension failure mechanism has also been commonly observed in landslides, especially for those covered by cemented soils geometrical. Considering only the shear failure would overestimate the tensile capacity of geomaterial, which can lead to an optimistic result. In this paper, a modified M–C failure criterion with zero or low tensile strength (tension cutoff) was introduced that can characterize the shear–tension failure feature of slopes well. Combined with the limit upper bound theory, the expressions of stability factor (N_s) for slopes were derived considering (1) only soil self-weight; and two external conditions, (2) surcharge load, and (3) seismic load. Further, a detailed parametric analysis was conducted. The results show that the slope stability was greatly influenced by the surcharge coefficient (q_t) and the horizontal seismic acceleration coefficient (k_h). The influence of the degree of tension cutoff (ζ) on the slope stability strongly depends on the values of slope angle (β) and internal friction angle (φ). The difference in N_s under two extreme cases (ζ = 0 and ζ = 1) was significant, and the difference was more pronounced with the introduction of surcharge and seismic load.

The inhomogeneity of soil is a ubiquitous problem for pile foundations due to, for example, pile installation (radial inhomogeneity) or the natural sedimentation of soil (vertical inhomogeneity). Most continuum theory solves soil equations by artificially setting the displacement at the inner soil core as zero, which results in the misestimation of soil shear strength. Although the additional mass model avoids the utilization of artificial boundaries, the internal deformation of soil is overlooked. In this paper, the modified additional mass model is proposed. The capacity of this model is verified through comparison with former studies. Through a comprehensive parametric study, it was found that (1) the inhomogeneity of soil has significant influence on pile stiffness and damping amplitudes, (2) the length and strength of the pile would affect the resonance frequency of the soil–pile system, (3) only the soil within the range of 0.3r₁ to the pile shaft has a visible effect on the dynamic response at the pile head.

Soil–steel structures are increasingly used as a solution for roadway and railway overpasses because of their relative low cost and ease of construction compared with conventional concrete and steel bridges. Recently, soil–steel arched bridges have been developed to achieve large spans of up to 32.4 m. Current design guidelines for these structures require comprehensive numerical modeling to specifically consider the soil–structure interaction and live loading. This study presents three-dimensional finite-element (FE) analysis for a field monitored full-scale corrugated steel culvert subjected to single-axle loading. The FE model predictions were within 15% of the measured responses. In addition, the behavior of a large-span arched steel culvert using the world's deepest corrugated steel profile was analyzed under the effect of axle loading at different soil cover heights. The straining actions of the steel structure decreased by 30%–35% as the height of soil cover increased from 1.0 to 1.4 m. Furthermore, the effects of the numerical modeling technique on the predicted performance and straining actions were evaluated, including geometric nonlinearity, soil behavior constitutive model, soil–steel interface condition, and construction stages for backfilling. The results demonstrated that incongruous numerical simulation may lead to differences in the results by up to 60% with the anticipated results causing significant impacts on the design and cost of soil–steel structures.

The moving water table would vary soil layering and may significantly affect the isolation efficiency of wave barriers. A two-dimensional finite-element model is then developed using COMSOL to study the influence of the moving water table on the isolation efficiency of an in-filled trench. The effects of soil hydraulic properties, fill material characteristics, and geometrical parameters of the trench are investigated in detail. Numerical analyses show that a low permeability of poroelastic soil, which shows low-frequency behavior, would significantly reduce the screening performance of barriers. A resonance in stiff barriers may occur as the impedance ratio increases, resulting in worse isolation efficiency. Placing soft trenches in vibration-weakening zones results in worse screening effectiveness, while it is the opposite for stiff barriers. A single stiff barrier can hardly achieve a significant screening performance at the critical water table (H_c) because of the resonance in the dry layer. However, mounting a horizontal plate above H_c would efficiently mitigate the adverse effect of the resonance and improve the isolation efficiency of barriers.

This study aims to predict the undrained seismic bearing capacity and failure mechanisms of a shallow strip footing that is placed on homogeneous and heterogeneous cohesive soils above a single unsupported rectangular void using finite-element limit analysis (FELA). Upper and lower bounds will be computed for a comprehensive range of geometries, horizontal earthquake accelerations, materials, void shapes, and size combinations. In addition, parameter studies will be carried out to quantify the effect of various factors on the bearing capacity of the footing to finally derive a design table that could be used to quickly predict the bearing capacity of the desired footing. The results show that according to the void condition, earthquake acceleration, and soil characteristics, there are four modes of failure: footing bearing failure without void failure; footing bearing failure with roof void failure; footing bearing failure with side; and roof void failure and void collapse without bearing failure. Finally, a sensitivity analysis will be carried out to show that the void depth and soil undrained shear strength are the most important parameters for the ultimate footing bearing capacity.

This paper presents an efficient and robust probabilistic approach to analyze the geosynthetic reinforced slopes (GRSs). The deterministic analysis will be performed that employs the rigorous (5N−1) formulation of the horizontal slice method (HSM), which is made efficient using the nonlinear constrained optimization (N = number of slices). The probabilistic analysis will be performed using a surrogated assisted Monte Carlo simulation (MCS). Collocation based stochastic response surface (SRS) will be employed to build the surrogate model using a third-order multidimensional polynomial chaos expansion (PCE). The random variables include the internal friction angle of soil (ϕ), soil unit weight (γ), and tensile strength of the reinforcement (T_u). Dependence between the random variables will be established using the Gaussian copula. A comparative analysis of the results with the First-Order Reliability Method (FORM) will be presented. The performance function will be evaluated 125 times using the SRS method in contrast to the direct MCS where it will be evaluated ≥50,000 times. This will reduce the computation time from approximately 2 days to approximately 20 min. In addition, the influence of correlation between the random variables will be highlighted comprehensively by adopting a wide range of correlation coefficients. This study concludes that the SRS method that incorporates an accurate deterministic model is a highly efficient and powerful approach to analyze GRSs probabilistically.

The strain transfer between fiber-optic cable and soil plays a critical role in the deformation characteristics of a cable–soil interface. Existing findings cannot provide a clear understanding of the effects of key influencing factors including horizontal confining pressures, anchorage property of cables, and saturation of soils on the strain transfer and shear characteristics at the interface. A group of pullout tests of cables in soil were conducted to examine the strain transfer efficiency using the optical frequency domain reflectometry (OFDR) technique. Two kinds of cables were pulled out from sandy soil and sand–gravel–clay mixtures under the confining pressures of 0–1.2 MPa. Typical strain-hardening behavior was observed for cables with confining pressure and anchorage, and the interface shear strength could not be evaluated within the strain measurement range. To address this problem, the so-called interface shear coefficient was adopted, and the key influential factors were discussed quantitatively. The interface shear coefficient keeps a linear relationship with the confining pressure, and that of the anchored cable in saturated backfill mixtures is 2–3 times than that of the unanchored cable. These findings will guide the methods of gaining reliable data for revealing the failure mechanism of geostructures via distributed strain sensing.

Expansive soils are susceptible to ground movements and are problematic to overlying geotechnical structures. There is a growing shift to use nontraditional sustainable and economical alternatives. To date, limited research is available on investigating the reaction mechanism of nontraditional soil stabilization; thus, this research aims to fill this gap in the knowledge. A series of experiments was conducted to assess the mechanical and hydraulic behavior of field expansive soil with the enzyme-based soil stabilizer. Moreover, the reaction mechanism of enzyme soil stabilization was studied using microstructural-driven analytical techniques. The results indicate substantial increase in California bearing ratio, unconfined compressive strength, and resilient modulus with a considerable reduction in permeability with the addition of enzymes into the expansive soil. The enzyme additive is adsorbed by the clay particles, which reflects a reduced affinity for water, an improved density, and a reduced ingress of water within the soil. Adding enzymes causes the expansive clay mineral lattices to relax, as the enzyme stabilizers penetrate the inner layers leading to expansion and subsequent moisture entrapment of the inner layers, emphasizing densification and a reduced affinity for water. The outcomes from the study are beneficial for the understanding of enzyme-based ground improvement to inhibit the expansive nature of reactive soils.

Reliability analysis of complex geotechnical engineering is time-consuming since its performance function is highly nonlinear and implicit. In this paper, an adaptive sequential sampling metamodeling-based method is proposed to deal with such problems. Gaussian process regression (GPR), utilized to approximate the real performance function, is constructed by the initial design of experiments (DOEs). Based on the geometric meaning of the most probable point (MPP) in the first-order reliability method (FORM), the potential MPP, which is a point infinitely close to the limit-state surface and with the minimum distance to the origin while considering a distance constraint, is searched and added to the DOE to refine the GPR. Then, the Monte Carlo simulation (MCS) is adopted to evaluate the failure probability by the refined GPR. The above two procedures are repeated until the stopping criterion is reached. Three examples, including one mathematical example and two geotechnical engineering problems, are analyzed. The results show the proposed method requires fewer performance function calls and is an efficient, accurate, and robust reliability analysis method.

Face stability is a critical issue in the stability analysis of tunnels, especially in soft soils and crushed rock masses. In this paper, considering two three-dimensional collapse mechanisms, upper-bound solutions are presented for determining the face pressure of deep and shallow tunnels excavated in rock masses. The Hoek–Brown failure criterion was considered for the rock mass surrounding the tunnel. The proposed upper-bound formulations were compared with three-dimensional finite-element simulations and also with centrifuge test results available in the literature. The obtained results showed that the face pressure decreases with increasing σ_ci, m_i, and geological strength index. Besides, increasing the tunnel diameter and the rock mass density led to an increase in the face pressure, where for larger tunnels, the effect of the density became more pronounced. Finally, some charts were presented to specify the range of applicability of the concepts of deep and shallow tunnels. The results obtained from these charts demonstrated that the reduction in the Hoek–Brown parameters of the rock mass results in spreading the collapse mechanism from the tunnel face to the ground surface.

A force-based limit equilibrium analysis is presented for the stability of a two-dimensional, three-wedge sliding mass of soil with vertical interfaces between the wedges. The analysis is conducted using five failure planes and can accommodate variable conditions for wedge geometry, pore pressure, shear strength parameters, reinforcement, applied loads, and pseudostatic seismic coefficients. A constant factor of safety is assumed for each failure plane and reinforcement element, although this assumption can be relaxed through the selection of strength parameters. The factor of safety is obtained analytically and requires solution for the roots of a polynomial equation. Verification checks show exact agreement with existing solutions for simplified conditions. A numerical example is provided to demonstrate the method and illustrate the importance of several parameters for stability of an earth slope. The analytical solutions take compact form, provide insight for the three-wedge method, and offer good capability to tailor conditions for applications that can be suitably characterized by wedge failure.

The mechanical performance of rock mass subjected to the coupled influences of the elevated temperature and in situ stresses has always been a hot topic in underground rock engineering projects. In this study, a thermal-stress-aperture coupled model was first developed and then incorporated into the particle flow code for the coupled thermomechanical analyses in the fractured rock mass. With thorough considerations of the aperture-dependent thermal and mesomechanical parameters for the fractured rock, the model performed more realistic thermoelastic responses of the fractured rock to the temperatures and confining pressures. Comparative studies between the numerical simulations and previous experimental results indicated that the proposed model was suitable for modeling the thermomechanical behaviors of the fractured rock. Then, a series of numerical compression simulations with heating temperatures of 20°C–600°C and confining pressures of 0–20 MPa were conducted to comprehensively explore the interplay of the temperature and confining pressure on mechanical properties of fractured rock specimens. Finally, the mechanisms that affect the rock thermomechanical properties were further revealed. The results indicated that the compressive strength and elastic modulus increase with an increase in confining pressure for each temperature scenario. The thermal strengthening behavior of rock extrapolated to about 400°C takes place in confined compression tests and is more pronounced at higher confining pressures. The evolutions of thermal properties, microcracks, and mesostructures are the most decisive factors that could induce the variations of rock properties under the coupled temperature and confining pressure treatment. For analyzing the mechanisms behind strengthening and weakening contribution to rock properties, the positive effect of the decrease in average fracture aperture, the dual effects of increased porosity and thermal-induced microcracks, and the negative effect of stress-induced microcracks should be comprehensively considered.

Efficient and accurate boulder detection technology is essential for shield tunneling safely under complex floodplain geological conditions. The traditional seismic reflection method is not suitable for complex geological conditions and small targets near the surface, the elastic wave reflection method based on zero-offset observation is proposed in this article. Forward is the basis of inversion, a high-order staggered grid finite difference algorithm based on least-squares optimization (LS-SGFD) is established, which has the advantages of higher accuracy and smaller dispersion. Three typical geological models are designed, and the elastic wave reflection characteristics of the method proposed in this paper and the traditional method are systematically analyzed. From the dual perspectives of field experiment and engineering application, the effectiveness of zero-offset observation for the detection of boulders is discussed, and the basic criteria for boulder identification are given. Research shows that this method has good results in detecting boulders on the surface, and it has certain reference significance for similar projects.

The three-dimensional discrete element method (DEM) is employed to investigate the macro- and microscopic mechanical behaviors of unsaturated silt subjected to one-dimensional (K₀) and constant stress ratio loadings. A DEM sample was prepared by installing a contact model considering rolling and twisting resistances and the van der Waals force. Capillary forces were installed at contacts to represent the unfilled water effect. Strain-controlled one-dimensional and constant stress ratio compression tests were then simulated on the specimens with different constant matric suctions and stress ratios (η = q/p). The discrete element simulation results reproduce the primary macroscopic mechanical behavior observed in laboratory tests. Shear failure occurs in specimens compressed under stress ratios exceeding the strength slope M, while volumetric yielding occurs in specimens compressed under very small stress ratios η ≪ M. The compression test under a constant stress ratio larger than M reaches the critical state characterized by not only a constant stress state and void ratio, but also a constant effective particle number, (mechanical) an average coordination number, and fabric anisotropies of contact normals and contact forces. The contribution of normal contact forces to the deviator stress predominates over tangential contact forces throughout the simulation. The capillary stress (capillary force-induced stress tensor) anisotropy decreases under isotropic compression but increases under anisotropic compression (η = 1.25). The mechanical average coordination number versus mean stress is independent of the stress ratio and the matric suction. Fabric evolutions indicate that the rearrangement of contact forces is much quicker than that of contact normals at the beginning of compression.

A full-scale vacuum preloading test was performed for an embankment construction on a Canadian structured clay deposit at Saint-Roch-de-l’Achingan. Accordingly, extensive instrumentation was used to monitor field behavior over eight months. A multidrain finite-element analysis was also carried out using a plane strain configuration with which the structured soft clay behavior was simulated using the S-CLAY1S constitutive model. The numerical analyses were then compared with the field measurements in terms of settlements, horizontal displacements of the clayey foundation, and pore pressure developments in the soft clay deposit. The results showed that compression curves obtained numerically at two different depths compared reasonably well with both field and laboratory curves. Effective stress paths at different points in the clay layer clearly demonstrated the benefit of vacuum preloading in comparison with traditional embankment construction (i.e., without vacuum preloading). In addition, the results suggested that the finite-element approach was consistent with laboratory and field behavior.

This study focuses on the stability of unsupported vertical cuts in unsaturated soil. For this, a series of numerical analyses were carried out to estimate the stability of unsupported vertical cuts excavated in an unsaturated glacial till considering three possible practical scenarios: (1) excavation time, (2) depth and location of a tension crack, and (3) rainfall infiltration into a tension crack. The results showed that the influence of excavation time and rainfall infiltration into a tension crack may not be significant if an unsupported vertical cut is made with a certain safety margin (i.e., safe height). A tension crack did not affect the overall safety of an unsupported vertical cut in a case where the location and depth of the tension crack are within 30% and 20% of safe height, respectively. The proposed approaches were then applied to the deep unsupported vertical cut made into a clay (9.75 m; Kwan 1971) for the validation. The failure in the vertical cut was initiated due to the development of a tension crack, which matched the numerical analyses results obtained extending both effective and total stress approaches considering the influence of matric suction.

The uniform arrangement of pile groups widely used in practice often causes obvious differential settlement of pile foundations. To decrease differential settlement of pile groups, individual piles nonuniformly arranged can be adopted. This paper presents a model test study on the bearing capacity of pile foundations under different pile arrangements. A versatile full-automatic loading system and data acquisition and analysis systems of the model test were developed. Using the model test setup, the response of a single pile subjected to axial load was clarified, and the reinforcing effect between an axially loaded pile and a load-free pile was then analyzed. Furthermore, model tests were carried out to assess the behavior of the pile groups composed of individual piles with variable length and diameter. For practical purposes, some suggestions on the nonuniformly arranged individual piles within pile groups were given to decrease differential settlement among each individual pile and make maximum use of the bearing capacity of each individual pile.

Geotechnical engineers deal with uncertainties in different projects. These uncertainties can be captured using nondeterministic approaches. Although during the last two decades numerous probabilistic and nonprobabilistic studies have been performed in a variety of fields in geotechnical engineering there is a limited amount of research in the literature on the applicability of different nondeterministic methods to evaluate statistical moments, reliability index (β), and most likely values (MLVs) of a system’s response during excavation projects. In this paper, the point estimate method (PEM), random set (RS) theory, and Monte Carlo simulation (MCS) methods with different mathematical procedures will be employed. This paper aims to present and compare these methods when estimating the mean (μ), standard deviation (σ), MLVs, probability of unsatisfactory performance (P_f), or β in the deformation and stability analysis of an instrumented deep excavation project in the presence of normal and nonnormal basic variables. In addition, due to the successful performance of the RS theory that was reported in previous researches, the code for the RS method was provided for further practical and research purposes. Finally, recommendations on the suitability of the methods to estimate the aforementioned statistics during excavation are provided.

Geotechnical structures often have position-dependent soil strength parameters as a consequence of stress history, deposition process, mineralogical composition, and weathering effect. However, most available theoretical models for estimating active earth pressures focus only on homogeneous cases. By a point by point technique, this work develops a friction-angle-dependent discretized failure mechanism to evaluate the active earth pressure on backfills in the framework of the limit-equilibrium slice method. The generated failure mechanism is composed of a series of horizontal elements. Compared with previous theoretical models, the benefits of the proposed method are twofold: (1) the friction-angle-dependent failure mechanism agrees well with the actual failure surface for heterogeneous backfills; (2) not only the multilayer frictional backfill but also the backfill with a position-dependent friction angle can be accounted for. The total active earth pressure coefficient, the earth pressure distributions, and the failure mechanisms are obtained, all showing excellent agreement with the available experimental data and the existing analytical solutions. For practical use, applications to backfills with a weathered zone and spatially variable frictional angles are then conducted. Finally, the proposed model is degraded into a two-layer case in which the active earth pressure coefficients under various mechanical and geometrical parameters are tabulated.

Surrogate models are helpful tools to enhance the efficiency for intensive computations of the factor of safety (FoS) in probabilistic slope stability evaluation. This study presents a multilayer perceptron (MLP)-based surrogate model combined with the Monte Carlo simulation (MCS) for system reliability analysis of earth slopes. The MLP-based surrogate model is constructed derived from the space-filling Latin hypercube sampling (LHS) for a global prediction of the FoS. Several factors affecting the performance of the MLP model are studied in detail, including the training algorithm, the generation method and size of samples, and the hyperparameters. Three examples with system effects are tested to verify the performance of the proposed method. The results show that the MLP-based MCS can achieve high accuracy and efficiency for the system failure probability assessment of soil slopes in different failure zones.

The convergence–confinement method (CCM), a widely used tunnel design method, is an analysis approach that can intuitively reflect the interaction between the surroundings and the support in tunnels. In this article, the CCM approach is used to determine the support performance of a steel–concrete composite support (SCCS) system, which is characterized by an improved stiffness of the initial support due to its unique arch type (π-type) and construction steps. In view of the tunneling process, the support characteristics of a conventional support system and the SCCS system are determined from the perspectives of single and combined support structures. A sensitivity analysis by considering the arch installation spacing, wall thickness, and concrete grade is performed. The results reveal that the stiffness and support pressure of the hollow π-type steel arch increase linearly with a decrease of the installation spacing and an increase of the wall thickness, while its support performance only slightly improves with an increase of the inner concrete grade after filling with concrete. Additionally, the support characteristic curves for different support systems are obtained, and the safety factor of the SCCS system is found to be greater than 3.48. The maximum support pressure of the SCCS system is found to increase by 1.25 MPa and its support performance increases by 4.57 times as compared to those of the conventional support system. Finally, the construction technology of the SCCS system is discussed, and some application suggestions are provided.