International Journal of Geomechanics

In the recent past, geotextile reinforcement with wraparound ends has been recommended to strengthen foundation soil to enhance the effectiveness of using the reinforcement. This study investigated the effects of reinforcement geometrical parameters—namely, land width occupied by reinforcement and lap length of the ends of reinforcement, which were not investigated in detail in the past—on the bearing capacity of the foundation soil. The soil considered was sandy in three different states of denseness, defined in terms of relative density as 50, 70, and 90%. The numerical model was developed using the typical elements in PLAXIS-3D software. The model was used to critically analyze the reinforcing mechanisms for improved bearing capacity caused by wraparound ends. The results showed that the bearing capacity increased by approximately 120% when reinforced with a geotextile reinforcement using the wraparound reinforcement technique, whereas this improvement was approximately 45% with the reinforcement only without wraparound ends. It was also found that the wraparound technique can be highly beneficial in a location of limited land width for foundation construction.

Plate anchors are embedded into the ocean floor to provide holding capacity for offshore structures. Anchor holding capacity is a function of both the anchor and soil properties. Although plate anchors have been widely studied experimentally and numerically, there is still no universally agreed-upon design approach, indicating that the problem physics remain elusive. In this work, discrete-element method (DEM) simulations were used to investigate the behavior of plate anchors during pullout in an effort to elucidate some of the microscale physical processes that influence overall system behavior. Macroscale assembly response was compared to published experimental results and empirical solutions. The influence of embedment ratio, anchor roughness, soil density, and anchor size on holding capacity was investigated, and system-scale results reasonably agreed with previously published work. Thus, observations of the simulated contact force network and particle velocity during uplift were used to provide insight into anchor failure mechanisms. Finally, the model was used to briefly explore the response of a cyclically loaded plate anchor embedded in a granular assembly.

Failure characteristics of rocks under polyaxial compression have been the subject of extensive laboratory testing, although a limited amount of literature is published on the numerical simulation of rock failure under such conditions. The current study was conducted by using a commercially available finite-difference program, FLAC^3D, to model the hard rock strength and fracturing under polyaxial compression. The Mogi-Coulomb criterion, where the relevant parameters can be obtained from conventional triaxial tests, was simulated in the software for realistic reflection of the effects of intermediate principal stress on hard rock fracture characteristics. According to the elastoplastic mechanics, the incremental iterative calculation format of the Mogi-Coulomb strain-softening failure criterion was implemented in the program, and the results were compiled into a dynamically linked library (DLL) based on the program interface of secondary development provided by FLAC^3D. The numerical results of the calculated peak strength were analyzed and compared with previous experimental data for verification of the effectiveness and accuracy of the simulation models. Laboratory tests with consideration of various intermediate principal stresses were also conducted to investigate the hard rock failure modes under polyaxial compression states. The comparison of experimental results with numerical results using both Mogi-Coulomb and Mohr-Coulomb strain-softening models further explained the validity of the former method. The results showed that proper numerical modeling can be considered for simulating hard rock fracturing under polyaxial compression states for complex stress conditions in future studies.

Aeolian loess deposits are among problematic collapsible soils whose substantial settlements due to the increase in moisture content, stress level, or a combination of both, frequently cause serious structural failures, financial losses, and casualties worldwide. One way to increase the strength and bearing capacity of collapsible soils is to chemically stabilize them with lime. The behavior of lime-stabilized collapsible soils has been investigated mainly by conducting conventional laboratory tests without any measurement or control of unsaturated state variables, such as matric suction. In this research, the hydromechanical behavior of a reconstituted lime-stabilized loessial soil is investigated by conducting unsaturated suction-controlled odometer tests. For this purpose, variable vertical loads have been imposed on the unsaturated lime-stabilized loessial soil specimens (with different amounts of lime content) under constant matric suctions. During all unsaturated odometer tests, unsaturated state variables (e.g., matric suction, degree of saturation, volume change of the tested specimens) have been simultaneously controlled and measured. To capture the hydraulic characteristics of the soil, along with odometer tests, soil-water retention curves (SWRCs) of specimens have been obtained for both wetting and drying paths by conducting filter paper tests. Furthermore, to investigate the effects of adding lime to the evolution of soil pore structures, scanning electron microscope (SEM) micrographs have been captured from the structure of the soil specimens. Results of laboratory tests showed that adding a small amount of lime to collapsible loessial soils not only reduces the collapse-induced volume changes to a large extent, but also can increase the yield strength of the soil. Moreover, it was revealed that further increase of lime content in loessial soil specimens beyond a specific value has no significant influence on the hydromechanical characteristics of the treated soil. In a companion paper, the results from unsaturated odometer tests on treated and untreated loessial soil reported in this paper have been analyzed in an unsaturated effective stress context, and an empirical model for explaining the load-collapse behavior of the soil is presented. Also by implementing the disturbed state concept (DSC), a coupled semiempirical hydromechanical model is developed to predict the disturbance level and calculate the soil strains under applying vertical stress in the K_o condition.

Assessing the hydromechanical behavior of collapsible soils is always a challenging topic in an unsaturated soil mechanics context. Although numerous efforts have been made thus far to establish frameworks for explaining the stress–strain behavior of collapsible soils, many of them are developed based on the results from conventional tests without controlling or measuring the unsaturated state variables. Furthermore, modeling the hydromechanical behavior of chemical-stabilized collapsible soils has been rarely studied. To establish a constitutive model for predicting the behavior of a lime-stabilized collapsible soil, the results obtained from filter paper and unsaturated odometer tests on a lime-treated loessial soil (which are presented in a companion paper) have been analyzed by considering the effective stress approach for unsaturated soils. An empirical model for explaining the load-collapse behavior of lime-stabilized tested soil is presented. Moreover, by implementing the disturbed state concept (DSC) method, a coupled semiempirical hydromechanical model is developed to predict the disturbance level and calculate the soil strain due to application of the vertical stress in K_o condition. Both presented models in this research are introduced as functions of two important governing state and material variables, namely matric suction and lime content. Results show excellent conformance between the laboratory test results and model predictions.

Numerical modeling of geomechanical phenomena and geo-engineering problems often involves complex issues related to several variables and corresponding coupling effects. Under certain circumstances, both soil and rock may experience a nonlinear material response caused by, for example, plastic, viscous, or damage behavior or even a nonlinear geometric response due to large deformations or displacements of the solid. Furthermore, the presence of one or more fluids (water, oil, gas, etc.) within the skeleton must be taken into account when evaluating the interaction between the different phases of the continuum body. A multiphase three-dimensional (3D) coupled model of finite strains, suitable for dealing with solid-displacement and fluid-diffusion problems, is described for assumed elastoplastic behavior of the solid phase. Particularly, a 3D mixed finite element was implemented to fulfill stability requirements of the adopted formulation, and a permeability tensor dependent on deformation is introduced. A consolidation scenario induced by silo filling was investigated, and the effects of the adoption of finite strains are discussed.

A three-dimensional (3D) discrete element method (DEM)-based numerical model is used to simulate the macromechanical response of sand strengthened using microbially induced carbonate precipitation (MICP) under undrained triaxial compression and inform the particle-scale mechanisms responsible for the behavior. The constant volume method is used to simulate saturated media. Although simulations using rigid boundaries are capable of representing the response of uncemented sands, virtual undrained triaxial tests on cemented sands require the use of flexible boundaries. Flexible membrane boundaries are created using particle facets (PFacets) as the building blocks. A methodology to implement virtual undrained triaxial compression using PFacet-based membrane boundaries is developed. The macroscale response of sands with varying degrees of cementation is adequately captured by this model. A cohesive bond strength, used to express the degree of cementation, is found to be well related to the shear-wave velocity through the soil sample. The model correctly predicts the occurrence of strain localization in cemented media, and the expected trends in shear band formation. The evolution of normal contact force distributions and coordination numbers as functions of both the cementation level and axial strain are also predicted.

Embankments over pile-reinforced soft soil (called piled embankments) are known as one of the interesting techniques for soft soil improvement. They have been broadly applied in infrastructure projects thanks to their advantages, such as the decrease in settlement, the reduction in construction time, and the reasonable cost. The shearing and the soil-arching mechanisms within embankments result in an increase of the stress acting on the pile head and a reduction of the soft soil pressure. They can reduce the total and differential settlements. Numerous studies were carried out to better understand the behavior of this technique. However, most of the research focused on static loadings, and few studies concerning the cyclic loading were conducted. The aim of this study was to investigate the behavior of a piled embankment subjected to different traffic cyclic loadings. A three-dimensional numerical modeling using the finite-element method (FEM) was conducted with the use of Abaqus. An advanced constitutive soil model based on the hypoplasticity concept was used for granular soil and compared with the conventional one (a linear elastic–perfectly plastic model with the Mohr-Coulomb failure criterion) for aspects of static and cyclic loading. A high number of load cycles applied to a piled embankment was also considered in the study. The numerical results show that the hypoplastic model is better than the linear elastic–perfectly plastic model to deal with the soil-arching decrease and cumulative settlements under cyclic loading. In addition, the influence of the number of traffic load cycles, the vehicle speed, and the embankment height on the arching effect and the cumulative settlements is presented.

This paper presents a large strain solution for the problem of cylindrical cavity expansion embedded in hardening–softening, non-associative Mohr-Coulomb material. The solution is obtained by transforming the problem from the original radial domain to the effective stress space in the hardening regime and to the effective plastic strain in the softening regime. The robustness of the solution is demonstrated by tracing the entire loading history of a pressurized cavity embedded in weak sandstones in large expansions and the calculation of the limit pressure beyond which the hole expands without further pressurization. The calculated radial stress profiles are completely smooth at the elastoplastic (EP) and the hardening–softening boundaries. The calculated limit pressure increases with increasing degree of pressure sensitivity, as expected for a stronger material. The limit pressure decreases dramatically as the rate of material softening increases. The elastic perfectly plastic limit-pressure solution, which ignores hardening, is quite close to the value determined from the full EP hardening with the same peak strength. The combination of non-associativity and material softening causes the breakdown of the solution before reaching the limit pressure due to emergence of bifurcation in the form of shear banding and surface instabilities.

The use of structural and nonstructural overlays has gained popularity in recent years due to financial and environmental constraints; however, the impact of overlay thickness and the type of applied tack coat on induced strains, where overlay interfaces with an existing surface layer, has not yet been extensively evaluated. In this study, the three-dimensional (3D) finite-element modeling technique was used to evaluate existing pavements of varying thicknesses with various overlay thicknesses, and different interface bonding strengths that were modeled in Abaqus. The results of this study verify that the impact of interface bonding strength on the overlay durability and structural characteristics of pavement changes drastically by changing the overlay thickness. For nonstructural overlays with a thickness of 12.7 mm (0.5 in.) or 25.4 mm (1 in.), interface bonding strength (tack coat) was found to play a prominent role in overlay durability, whereas overlay thickness was found to have a negligible impact. Increasing the thickness of the overlay from 25.4 mm (1 in.) to 50.8 mm (2 in.) significantly decreased the impact of interface bonding strength, whereas the influence of overlay thickness increased.

Current literature lacks an analytical method for estimating pile load distribution due to the statically indeterminate nature of pile groups. In practice, such estimations are accomplished by finite-element-based analyses or numerical interaction methods, which require computer implementations. A disadvantage of these implementations is that they extend the total duration time and cost of a project. In order to address this issue, a group pile load formula was developed based on 120 finite-element (FE) model configurations. The group pile load formula incorporates the dimensions of a rectangular raft, diameter, length and spacing of piles, vertical uniform pressure, soil moduli up to five layers, ultimate pile-soil and pile-tip resistance, and the elastic modulus of piles. The formula is capable of calculating pile loads at 25 different locations. Additionally, the portion of load carried by the piles as well as through the raft–soil contact can be estimated. Predictions from the FE model and the formula were verified with respect to the observed results from three case studies. The formula developed will help engineers to optimize their group pile configurations efficiently by providing an accurate group pile load estimation.

The grain size of aggregate particles is crucial to the mixture gradation of discrete-element (DE) models when realistic aggregate shapes are simulated. The objective of this study was to answer the question of how to determine the grain size of aggregates using DE models based on virtual sieving analysis. First, virtual sieving analysis models were developed with prolate ellipsoid, oblate ellipsoid, and cubic-shaped particles, and virtual sieving was performed under three vibration patterns, namely, vertical, horizontal, and hybrid vibration. The influence and efficiency of the vibration patterns were analyzed based on the results of the virtual sieving analysis. Then, the virtual sieving analysis was conducted with realistic aggregate shapes. By analyzing the test results, the shape sieving factor (${\mathit{S}}_{\mathit{sf}}$) was derived and was used to calculate the grain size of individual particles. For further validation, the grain size (${\mathit{G}}_{\mathit{s}}$) of selected aggregates was measured by lab manual measurement and virtual sieving analysis, separately. Then the test results were analyzed and compared. The main findings from this study include the following: (1) vibration patterns had significant impacts on the results of the virtual sieving analysis, and vertical vibration is recommended for virtual sieving analysis; (2) particle shapes had important impacts on the results of the virtual sieving analysis, and it was determined that aggregates with cubic shapes are relatively difficult to pass through the sieve meshes; (3) most particles can pass through smaller sieve apertures than their equivalent-volume spheres; (4) the approach to virtual sieving analysis developed in this study was validated by lab sieving tests, and the shape sieving factor (${\mathit{S}}_{\mathit{sf}}$) derived from the virtual sieving analysis can be used to generate DE models with more accurate gradation.

A drive is currently underway in the construction industry to promote the use of in situ soils (including clays) that are reinforced with natural fibers. This effort is primarily driven by sustainability considerations and concerns. Understanding of the interface resistance between natural fibers and clays is critical for reliably predicting the response of structures that are built on, or with, fiber-reinforced clays. The objective of this paper was to investigate the interface shear strength between natural hemp fibers and clay through interface direct shear and single fiber pullout tests. The parameters that were varied are the compaction water content and the drainage condition at the interface. The results indicated that the interface shear strength parameters are significantly affected by the test mechanism and drainage conditions. For compacted fiber-reinforced clay systems that are governed by short-term stability conditions, interface interaction coefficients may be best quantified using fast single fiber pullout tests. For systems that are governed by long-term stability, the drained interface friction coefficient could be estimated using small-scale direct shear or single fiber pullout tests. Hemp fibers are relatively efficient at mobilizing the shear strength of the clay, with interface interaction coefficients that could approach unity.

A new frequency-dependent analytical model of railway substructure stiffness is proposed for analyzing ballasted rail-track systems subjected to motion-induced dynamic loading. The proposed model incorporates the effects of all essential substructural components, such as sleepers, rail pads, ballast and subballast, and subgrade. It was found that incorporation of sleepers may induce large deflections in the rail-track system at the isolation frequency of the sleeper-substructure system. Furthermore, consideration of the flexibility of the sleeper beam (incorporated by using frequency-dependent stiffness of the substructure) may significantly influence the response of the railway-track system. It was also observed that the incorporation of damping into the system reduced the rail-beam deflections without altering the critical velocity. The reliability of the proposed model was assessed through comparison with an experimental study. Contrary to the case of contemporary models, the computational time for the dynamic analysis in this case was much smaller, thus making the proposed model well suited for engineering practice.

The mechanical response of granular material to the rotation of stress principal axes is an issue of both practical and theoretical importance in soil mechanics. This study used a two-dimensional (2D) discrete-element method (DEM) simulation to investigate the deformation of granular material under the continuous rotation of stress principal axes while maintaining fixed principal stress values. The results show that under such rotation, the deformation can exceed that caused by fixed principal stress axes cyclic biaxial compression with a maximum deviatoric stress ratio that equals the fixed stress ratio during rotation. The volumetric strain was found to be contractive overall while oscillating within each load cycle for specimens with significant fabric anisotropy. The initial fabric-anisotropy orientation was found to have little influence on the overall evolution of volumetric strain, although it was found to affect the development of shear strain. During the cycles of rotation of initial stress principal axes, the dilatancy of granular material, in the contractive direction, was found to be dominated by the evolving orientation of the major principal stress axis in reference to the initial major principal stress axis during anisotropic consolidation, and the influence of the initial fabric-anisotropy orientation was found to be trivial. As stress rotation continued, the influence of the relative orientation between stress and fabric became prominent, causing the material to dilate when the major principal stress axis rotated from the normal of the bedding plane to being perpendicular to it and causing contraction during the other half of the cycle.

Guadalquivir blue marl is a high-plasticity overconsolidated carbonate clay. This soil presents an elevated fragility and high susceptibility to moisture changes. These characteristics have caused many geotechnical accidents, such as the Aznalcollar dam failure, in Seville, Spain. A comprehensive test campaign was conducted to determine the physical and chemical properties of the blue marl. Analysis by scanning electron microscopy (SEM) and mercury intrusion porosimetry (MIP) allowed characterization of its internal structure, revealing clear differences between the macro- and microstructure. A novel one-dimensional (1D) model for predicting the volumetric deformation (under oedometric conditions) of the Guadalquivir blue marl with suction and vertical pressure changes was proposed. The model, based on data from shrink–swell tests, provided an acceptable estimation of the volumetric behavior of the soil with a relatively simple set of parameters. The results were experimentally verified by suction-controlled oedometer tests and showed an acceptable agreement with the data measured. The proposed model is valid for ranges of swelling, shrinkage or collapse.

Local defects, such as surround cavities and insufficient thickness of the tunnel lining due to imperfect construction, will lead to local discontinuity in the ground–lining interaction. This study investigated the mechanical and cracking performance of an existing tunnel lining due to local construction defects. Field observation was carried out to detect the geometry and position of the local construction defects, whereas numerical analyses were performed to examine how these defects affected the stress distribution and cracking performance of the concrete lining. The development of cracking direction and distribution pattern on the existing tunnel lining due to the defects of local cavity and local insufficient lining thickness was demonstrated by using the extended finite-element method (XFEM). Scenarios with different construction defects were studied to account for the cracking performance at the interior and exterior sides of the existing tunnel lining. The results demonstrated that the local construction defect could lead the lining to local arch behavior, squeezing outward to the defect and consequently causing damages firstly at the invisible exterior side of the tunnel lining. The presented study demonstrated an effective way for modeling tunnel lining cracking and appropriately evaluating the actual cracking performance of the tunnel lining subject to local construction defects.

U-section steel sheet piles are widely used as support structures in bridge foundation pit excavation. Because of the reduced modulus action (RMA) effect of U-section steel sheet piles, it is difficult to balance structural safety and construction cost in engineering. In this paper, a case study of a foundation pit supported by U-section steel sheet piles in the middle and lower reaches of the Yangtze River was carried out. A three-dimensional FEM of the support structure was established and the RMA parameter analysis was conducted. According to the measured data, the RMA value was identified. Based on the RMA value, the support structure of the subsequent foundation pit was optimized and the deformation of the pile head was monitored during the excavation process. The measured deformation was in good agreement with the predicted value, verifying the feasibility of the inferred RMA value. Finally, to solve the uncertainty problem of the U-section steel sheet pile RMA in general situations, an optimization strategy for the design of the U-section steel sheet pile support structure was proposed.

Zi-Zhi Tunnel is a shallow-buried tunnel, and its geological conditions are generally upper soft and lower hard. Therefore, the designed excavation methods are the center diaphragm method and the bench method. For the considerations of safety, cost, and schedule, excavation methods should be converted in advance. When determining the conversion timing of excavation methods, the different positions between the tunnel face and the upper-soft and lower-hard stratum interface need to be taken into consideration. The finite-difference software FLAC3D was used to analyze the rules of three-dimensional deformation and the variation of displacement direction angle of the surrounding rock based on the theory of displacement direction angle under different working conditions. The field-monitoring data of vault settlement before and after the conversion of excavation methods were also compared and analyzed to verify the rationality of the conversion of excavation methods. The results indicated that when the tunnel face passes through 1.5B distance (B is the tunnel width) of the stratum interface, the displacement direction angle of the surrounding rock tended to be stable, and the longitudinal stress concentration of the surrounding rock was decreased, which was conducive to utilizing the self-bearing capacity of the surrounding rock. Therefore, we suggest that the conversion of excavation methods should be carried out after the vault of the tunnel face passes through 1.5B distance of the stratum interface. Engineering experience verified the reliability of the conversion timing.

In pile-reinforced dams and bank slopes, the antislide effect of piles and drawdown of reservoirs are two aspects that could significantly affect the slope stability. However, existing studies have incorporated these two factors separately, albeit not in tandem. Moreover, stability assessment of these earth structures is usually performed ignoring the three-dimensional (3D) effect. To address these issues, the kinematic approach of limit analysis is adopted in this technical note for evaluating slope stability based on the 3D rotational failure mechanism. In addition, the coupled effects of pile reinforcement and water drawdown are considered. The analysis is performed for four types of drawdown cases. The results demonstrate that the optimal pile location undergoes significant change during the external drawdown process, while the effect of the declining water level on slope stability follows the similar pattern for varying pile locations.

The technical note presents analytical solutions to interaction factors of two unequal length piles (ULPs) under horizontal loads by using the fictitious pile method. The bending moments along the fictitious piles and the horizontal displacements of the pile heads/bottoms were obtained based on a simplified Fredholm integral equation of the second kind. The proposed analytical solutions were validated against the three-dimensional finite-element simulations considering different pile slenderness ratios. Parametrical studies were then conducted to show the influence of some key parameters on the pile-to-pile interaction factors for ULPs.

In this study, a simplified mechanical model is proposed based on the failure characteristics of a gypsum pillar–roof system obtained through in situ investigation. Then, a cusp catastrophe model for the gypsum pillar–roof system is established based on the given stress–strain relations of the gypsum rock. Using this model, the instability mechanism for the gypsum pillar–roof system is investigated. The results of the analysis indicate that the stress–strain relation of gypsum rock derived from damage mechanics theory can accurately describe its strain-softening behavior after the peak stress and that increases in the relative humidity around a pillar or both a pillar and the roof bed can increase the stiffness ratio of the support system and consequently result in a more stable support system. However, if the relative humidity around the roof bed increases, the stiffness ratio of the system decreases, increasing the possibility of catastrophic failure of the support system. The case study revealed that the influence of the relative humidity on the support system is significant, higher relative humidity in the mine’s atmosphere increases the probability of instability in the support system, and the influence of the relative humidity change around the pillar on system stability is greater than that around the roof bed.

Two kinds of asphalt binders and four kinds of aggregate gradations were selected. Cylinder specimens of six types of asphalt mixtures were molded by the superpave gyratory compactor (SGC) rotary compaction method, and semicircular bending (SCB) tests were conducted with different temperatures and notch depths. The effects of notch depth, gradation, asphalt properties, and temperature on the crack resistance of the asphalt mixture were analyzed based on essential fracture energy and fracture toughness. The analysis results show that the greater the notch depth leads to lower fracture energy and weaker crack resistance. The effect of gradation is slight when the nominal maximum particle size is relatively large, and smaller nominal maximum particle size leads to greater effect of gradation on crack resistance of the asphalt mixture. Asphalt properties have a positive influence, which means higher quality bitumen would give higher crack resistance. The effects of gradation and asphalt properties on crack resistance are limited by temperature.