International Journal of Geomechanics

Following major earthquakes that occurred in New Zealand (2010–2011) and Japan (2011), soil multiple liquefaction, or reliquefaction, regained major attention in the field of geotechnical earthquake engineering. Not only can liquefaction occur multiple times at the same site, but the devastation caused by reliquefaction is often more severe than that triggered by the first liquefaction. In this study, to address this issue and provide new insights into reliquefaction mechanisms, a series of cyclic simple shear tests was conducted with the use of a newly developed stacked-ring shear apparatus. In the multiliquefaction tests, subsequent liquefaction stages were applied to a single Toyoura sand specimen sheared at different levels of maximum shear strain double amplitude (${\gamma }_{DA\text{max}}$), from 2% to 10%. Tests results showed that: (1) the increase in soil density during the postliquefaction reconsolidation stages had only a minor effect on sand resistance against multiple liquefaction; (2) the extent of ${\gamma }_{DA\text{max}}$ significantly influenced sand resistance against multiple liquefaction; and (3) the major impact of ${\gamma }_{DA\text{max}}$ was a change in the soil fabric during multiple liquefaction, as confirmed by image analysis results.

In this study, the transient properties of dominant frequencies and their components during soil liquefaction were evaluated by the nonstationary Fourier spectra of the records at Kawagishi-cho during the Niigata-ken Chuetsu-oki earthquake in 1964, at Port Island during the Hyogoken-Nanbu earthquake in 1995, and at K-NET CHB024 during the 2011 off the Pacific Coast of Tohoku earthquake. The deterioration ratio of soil shear stiffness was calculated from the ratio of dominant frequencies at the beginning and the ending of the liquefaction process. The displacement profile was calculated through numerical integration of the acceleration record, and the shear strain was calculated from the relative underground displacement of Port Island from 0 to 83 m from ground level (GL-0m to GL-83m) to GL-83m. The deterioration ratio of shear stiffness was also evaluated by fitting the maximum shear strain to the G-$\gamma$ relation, and the deterioration ratio was evaluated by the nonstationary spectrum was confirmed by the coincidence of both evaluated values. The property of the nonstationary Fourier spectra is confirmed from the analytical result to artificial design waves, which provide important information for revision of the design wave. The response property of pulse waves was analyzed by the response envelope spectra for the Ricker wavelet. The properties of dominant components in the artificial seismic wave for seismic design were compared with those of the measured seismic records.

The majority of slope failures are triggered by excessive rainfall and the consequent increase in pore-water pressure within the slope. This paper presents the results of a computer code that quantifies earth pressure coefficients. This code is based on limit-equilibrium analyses and is used for the internal design of geosynthetic reinforced soil structures and to identify the critical failure mechanism. The critical failure mechanism is the largest value of out-of-balance force. For this purpose, the nonlinear programing (NLP) approach was used, and a NLP optimization model, TMAX, was developed. The model was used for failure mechanisms, assuming that the failure surfaces were bilinear. The influence of pore-water pressure on the potential failure surface was analyzed. The model was developed under basic principles. Optimally, the system is best suited for structures with varying geometries, different backfill unit weights, varying types of soil shear resistance, and different pore-water pressures. A numerical example was used to demonstrate the effect of pore-water pressure on the required strength of reinforcement and on the efficiency of the introduced optimization approach.

Generally, slope-stability evaluation involves two coupled tasks: locating the critical slip surface and calculating its corresponding factor of safety. Various assessment methods have emerged in the past few decades, and the question of whether equivalence exists between these methods has become a controversial issue. Unfortunately, research that explores the roots of the slope-stability problem from the perspective of computational complexity is quite limited. This paper addresses this long-neglected but very important problem by rigorously proving that the evaluation of slope stability is essentially a nondeterministic polynomial-time complete problem, which is computationally one of the most difficult types of problems. A significant achievement of this proof is the inference that equivalence between the limit equilibrium method and the strength reduction method does not exist. The methods used throughout these routine analyses are exactly approximate and will not replace one another completely. The case study verified the aforementioned conclusions.

This study examined the liquefaction-induced ground settlement of silty sand ground, in which many nonplastic fines are contained. Investigation determined that if fines were included in the specimen, the minimum void ratio shrank more under water-immersion than dry conditions. In addition, the cyclic minimum void ratio was examined, which was obtained by cyclic liquefaction and drainage testing conducted by assuming earthquake excitation. The results illustrated that the difference between the estimated dry and cyclic minimum void ratios uniquely correlated with fines content. A new density index, the normalized margin of void ratio, was proposed for postliquefaction volumetric strain based on examination of the minimum void ratio of soils containing nonplastic fines. The test results confirmed that maximum and accumulated shear strains are satisfactory indices of loading history. Furthermore, the normalized margin of void ratio performed well as a density index for estimating postliquefaction settlements, regardless of the type or content of fines.

This article presents an assessment of the effects of pore water pressure generation of the soil foundation on the seismic road embankment response. Numerical simulations were carried out to study the preloading technique as an improvement method for reducing the liquefaction potential and the induced settlements in a sandy soil profile. The analyses showed that the use of preloading reduces the induced settlements mostly because of the increase in lateral confinement in the superficial soil layers that results from an increase of the coefficient of lateral earth pressure at rest ($k_o$). The research also showed that the efficiency of the countermeasure method was limited to cases in which earthquakes produced a liquefaction zone lower than the depth of the overconsolidated soil.

Because pipelines traverse large geographical areas, they frequently must cross active faults when constructed in locations vulnerable to earthquakes. In this study, the authors performed three-dimensional (3D) finite-element analyses to investigate the behavior of buried pipe subject to strike-slip fault movement in dry sand and, more realistically, in partially saturated sand. The performance of the finite-element model was first validated by comparing the computed results with the data from the full-scale experiments at Cornell University. The analysis was then extended by varying the initial conditions of the sand (e.g., sand type, density, moisture content), pipe material, pipe burial depth, and pipeline–fault-rupture inclination to assess the effect of these parameters on the soil loads applied to the pipe and the corresponding deformations. On the basis of the simulation results, the authors propose a soil–structure interaction mechanism for pipelines crossing active faults. The authors also propose design recommendations for the mitigation of ground-deformation effects at buried pipeline crossings of strike-slip faults.

Nankai megathrust earthquakes are expected to occur with a return period of approximately 90–200 years along the fault of the Nankai Trough. Because they are interplate earthquakes, it is predicted that they will cause heavy liquefaction damage as a result of the long duration of their ground motion, especially around bay areas with soft deposits. In this study, a numerical analysis was performed for the ground in Osaka City with a hypothetical earthquake in the Nankai Trough region. In the numerical simulation of the dynamic behavior, a liquefaction analysis program was used. To quantify the degree of liquefaction at certain points, an index, named the liquefaction risk index (LRI), was defined and obtained by the weighted integration of the effective stress-decreasing ratio (ESDR) with respect to depth. Moreover, the LRI values were evaluated by data from the 1995 Hyogoken-Nanbu earthquake. It was found that the obtained LRI values were greater than 6.0 for the points at which heavy liquefaction damage had occurred. From the numerical results, the risk of liquefaction was evaluated with these LRI values.

The Tohoku coastal area in Japan suffered massive damage in the Great Tohoku Earthquake, in which a prolonged major earthquake was followed by a large tsunami. The damage mechanisms of coastal structures during earthquake–tsunami events have not been fully explained. Thus, this study elucidates the damage mechanism of breakwaters by focusing on the interactions among earthquake–tsunami events, caisson structures, and soil composed of rubble mounds and seabed components. Centrifuge model tests, finite-element analyses, and smoothed particle hydrodynamics simulations with tsunami–soil–structure interactions were performed. The simulated breakwater was destabilized by not only wave pressure, but also long-acting tsunami seepage flow and overflow into the rubble mound and the seabed. These processes resulted in scour and fluidization/liquefaction, which decreased the bearing capacity. Moreover, the liquefaction resulting from earthquake motion caused caisson subsidence and excess pore water pressure in the soil components before the tsunami occurred. These problems decrease the ability of breakwaters to provide protection against tsunamis.

A mixed FEM–crushable discrete-element method (DEM) nested scheme in the frame of second-order computational homogenization for granular materials was proposed. The particle breakage followed by the discontinuity and dissipative relative movements between each of two immediate neighboring particles were modeled at the mesoscale to perform both the downscaling and upscaling between the mixed FEM at the macroscopic continuum scale and the crushable DEM at the mesoscopic discrete particle assembly scale. To develop the crushable DEM for modeling the mesostructural evolution within representative volume elements (RVEs) assigned to integrating points of the macroscopic Cosserat continuum in the homogenization, two grain breakage models consisting of crushing criteria and the fracture mode for an individual crushable particle were developed. Not only the contact forces, but also contact moments exerted on each individual grain via the contacting points on the grain surface, were taken into account to set up the proposed crushing criteria. The stress measures, responsible for the particle breakage, involved both the average Cauchy stress tensor and the average couple stress tensor exerted on a crushable particle modeled as the Cosserat continuum. A fracture mode to specify how a crushable parent particle is replaced with a specific arrangement of postcrushing fragments was proposed and implemented. The mass conservation in the postcrushing replacement of fragments was ensured, whereas neither overlaps among the fragments nor overlaps with the fragments of the immediate neighboring particles of the crushable parent particle were introduced. The numerical results demonstrate the performance of the proposed mixed FEM–crushable DEM nested scheme in the frame of second-order computational homogenization for granular materials and the effects of the particle breakage on the failure behavior of the overall geostructure.

To protect coastal structures from the damage caused by the impact force of a tsunami, a new concept of using waste tires behind such structures is introduced in this paper. A physical model for tsunami impact force simulation was developed to evaluate the reduction effect of tsunami impact force by the tire structures. Model tests also were performed to evaluate the stiffness of tire structures. From an esthetic point of view, cultivation of suitable plants inside the tires is also proposed. Field tests on planting trees that can grow in saline soil conditions were performed to see whether such a structure can preserve the greenery of the area. Results show that the tsunami impact force could be reduced considerably by placing filled tires (with a suitable material) behind seawalls, and this technique can protect the structures from the tsunami impact force and the resulting scouring. The greening effect could be maintained by the appropriate selection of the shrubs and trees planted inside the tires, making it one of the most cost-effective methods for recycling waste tires.

The 2004 Niigata-ken Chuetsu Earthquake in Japan caused serious damage to embankments. Considerable damage occurred in mountainous regions, and the recorded rainfall before this earthquake exceeded the annual rainfall. In this study, numerical simulations were performed to reproduce the damage and to investigate the effect of rainfall on seismic behavior. In the simulations, a three-phase (soil, water, and air) coupled analysis was adopted to consider the behavior of unsaturated soil, and a conventional elastoplastic constitutive model was used for unsaturated soil. The deformation of the embankments was quantitatively reproduced by incorporating rainfall in the seepage analysis before the dynamic response analysis, and it was found that heavy rainfall significantly affected the dynamic behavior. Furthermore, parametric studies of countermeasures showed that the embankments would deform as a result of poor-quality materials even if sufficient drainage or impermeable roadbeds were constructed.

A series of laboratory model tests was performed to investigate the behavior of geosynthetic reinforced stiff clay foundation systems under circular loading. The footing consisted of a rigid circular steel plate with a diameter of 150 mm. Five different series of tests were performed in both homogeneous (clay or sand) and layered configurations. The tests used planar geogrid and three-dimensional geocell reinforcements. Test results indicate that both types of reinforcements substantially improve the performance of the stiff clay foundation bed. A maximum threefold improvement was observed in bearing pressure, depending on the reinforcement type. However, geocell was found to be the most advantageous soil reinforcement technique, giving maximum performance improvement.

This study investigated the response of rigid surface foundations to moment loading based on nonlinear (hyperbolic) soil response. The response of a rectangular foundation on the surface of the ground was modeled as a nonlinear Winkler model. The nonlinear Winkler-based shallow foundation model was evaluated for its dependence on eccentricity of load, subgrade properties, etc. The key subgrade properties on which the proposed nonlinear model depends are the initial tangent subgrade modulus and ultimate bearing stress of the soil. The model was validated by comparing the predicted response of stress versus normalized displacement based on relevant soil properties with the response from the published data of field footing load tests. A parametric study quantified the effects of different parameters on the stress-displacement and moment-rotation responses of the foundation. The proposed analysis captured the inherent nonlinear deformations of the soil and the coupled nature of displacements and rotations of foundations subjected to eccentric loading. The predicted displacements compared well with the measured values.

In this paper, the authors propose an analytical model for the prediction of the slope safety factor as a function of basic geometrical parameters (slope height H and slope angle β) and soil factors (bulk density γ, cohesion c, angle of internal friction φ, and pore water pressure coefficient r_u). Research was performed by applying the statistical technique of experimental design, for which the input data were provided by stability analyses of different homogeneous slopes with a circular slip surface using the Spencer limit equilibrium method. The proposed model represents a nonlinear equation of a simpler form and higher prediction accuracy than those of the existing mathematical expressions, with predominant linear effect of the individual factors and significant influence of the two-factor interactions. Linear terms in a derived equation indicate a positive effect of c or φ and a negative effect of H, β, γ, or r_u on slope stability. Because of two-factor interactions, the effect of c is highly dependent on H, β, γ, and φ, whereas the influence of φ is predetermined by the values of β and r_u.

The variation of stress–strain state induces changes in the permeability of clayey soil and then affects the properties of seepage and consolidation. In previous studies, variations of the permeability of soil in shear process were mainly neglected, and there were no reasonable descriptions of the permeability in such a condition. To study the variation of the hydraulic conductivity of clayey soil in shear process, new triaxial seepage equipment was designed by changing the conventional triaxial apparatus. Based on a series of tests under different confining pressures carried out by this equipment, the influences of the stress–strain state of clayey soil on hydraulic conductivity were investigated. Furthermore, a mathematical model for the hydraulic conductivity of clayey soil under the condition of large shear deformation was proposed. The comparison between the seepage test results and the calculated values of the proposed model indicates that the mathematical model can calibrate the effects of the void ratio and the mesostructure on the hydraulic conductivity of soil in shear process very well.

This paper presents a finite-element (FE) analysis of the effect of subsoil in a reinforced piled embankment and compares the findings with predictions from a theoretical method. It shows that the subsoil can carry as much as 75% arching embankment load for the present standard model (pile spacing of 2.5 m, geogrid stiffness of 6 MN/m, embankment height of 6 m, and subsoil compression index of 0.7). Also, the proportion is much larger for the lower value of pile spacing, geogrid stiffness, embankment height, and subsoil compression index. The maximum settlement of subsoil is found to be more sensitive to the pile spacing than to the geogrid stiffness. The theoretical method shows reasonable agreement with the FE results and can therefore be used to predict the effect of subsoil in reinforced piled embankments. The geogrid has a significant role in carrying the entire embankment load without the subsoil support. However, this effect may lead to intolerable geogrid strain due to relatively large settlement. Two case studies are assessed by the theoretical method. They show that stiff subsoil can carry a significant portion of the embankment load, whereas very soft subsoil support may result in intolerable strain on the geogrid. The effect of the subsoil, therefore, should be considered in the design.

This paper presents a new earth-retaining structure, the moso bamboo soil-nailed wall. Moso bamboo (including branches) is employed as soil nails and piles, and used to weave bamboo strip grids (replacing rebar grids) for sprayed-concrete surface courses. In this paper, the design and construction methods of this structure are elaborated using laboratory and field tests, and numerical simulations for two completed projects with respect to stability and deformation, respectively. In-situ tests show that the pulling capacity of bamboo nails (with branches) increased by ∼2.5–2.8 times in stiff soil or soft clay, respectively, compared with steel-pipe nails. For one wall in soft clay, numerical analysis was conducted, which verified the stability of the bamboo-reinforced system (compared with the soil-nailed wall constructed with steel-pipe nails, which failed). For another wall, the deformation and internal force of the structure were obtained using three-dimensional (3D) nonlinear numerical software. The prediction agrees well with the measured settlements, and has good accuracy against the measured horizontal displacements. The study indicates that the row of nails the lowest level receives the highest axial force (which requires the longest length). This trend is opposite to that in using conventional soil-nailed walls. The moso bamboo method renders an increase in the depth of sliding and thus the stability of excavation, which failed in the wall constructed with conventional steel-pipe soil nails.

The combined pile–raft foundation (CPRF) has been widely recognized as economic and rational foundation for high-rise buildings when subjected to vertical loading because of its effectiveness in load sharing by both raft and pile components. This results in smaller total and differential settlements with a reduced number of piles as compared with group piles. Until recently, the behavior of CPRF when subjected to lateral and real earthquake loading conditions in addition to vertical loads has not been well understood as a result of the complexities involved in the interaction of the pile, soil, and raft under such loading considerations. In the present study, an attempt has been made to investigate the behavior of CPRF with the use of centrifuge testing and a numerical model under pseudostatic and dynamic loading conditions carried out with finite-element software. After successful validation of the present CPRF model with both centrifuge and numerical model results, the same model was used for further study under El-Centro 1979, Loma Prieta 1989, Bhuj 2001 and Sikkim 2011 real earthquake pseudostatic loading and real acceleration–time history. The maximum displacement and bending moment were observed at the pile head, which is attributable to the rigid fixity of the piles with the raft. The crossover point in the bending moment profile was observed at shallow depths for all cases of loading. The occurrence of the resonance condition that yielded maximum horizontal displacement of the CPRF as a result of one of the input motions is also highlighted in the present study. The difference in the response of soil just below the raft of the CPRF and at the far field indicated that near-field and far-field ground motions were not in tandem. A case study on an existing CPRF of Messeturm Tower, Frankfurt am Main, Germany, is also presented. The existing CPRF of Messeturm Tower was modeled and analyzed under static and pseudostatic loading conditions. Obtained results show good correlation with the static field-measured results. Results in terms of bending moment in piles, total settlements, and normalized lateral displacement (u/D) are reported. The present findings are useful because they provide broader understanding of the response of the CPRF under pseudostatic and dynamic loading conditions.

Many factors influence the axial bearing capacity and performance of deep foundations embedded in sand deposits. Among these factors, the installation method is of great importance. The objective of this study was to investigate the construction effects on pile performance via physical modeling. The frustum confining vessel of Amirkabir University of Technology (FCV-AUT) was employed for this purpose. Six different installation methods were studied: jacking, drilling and grouting, driving, screwing, drilling and placing, and postgrouting. Up to 30 axial compressive and tensile load tests were carried out on different piles on Babolsar sand, from the northern coast of Iran, with relative densities of 45% to 50% within the FCV-AUT. Experimental results show that among different pile installation methods, jacked and precast-in-place piles had the greatest and lowest capacities, respectively. Moreover, the performance results of the different pile types in the FCV-AUT indicate that the axial bearing capacity of piles can be boosted by modification of the installation method. These modifications result in a reduction of the number and size of piles, and consequently result in savings in construction costs and time requirements.

The strength and dilatancy behaviors of modeled dense rockfill material were systematically investigated through a series of true triaxial compression tests at different minor principal stress and intermediate principal stress ratios. It was found that the intermediate principal stress ratio had great influence on the critical-state friction angle, peak-state friction angle, and maximum dilatancy of the modeled rockfill material. Both the critical-state and peak-state friction angles at a given confining pressure first increased and then decreased with an increase in the intermediate principal stress ratio. An increase in the minor principal stress led to a decrease in both the critical-state and peak-state friction angles at a given intermediate principal stress ratio. The maximum dilatancy decreased with an increase in the minor principal stress or the intermediate principal stress ratio. It was also found that the intermediate principal stress ratio significantly affected the relationship between the peak-state friction angle and maximum dilatancy. An adapted stress–dilatancy equation (incorporating a state index and a function of the intermediate principal stress ratio) could be used to capture the stress–dilatancy behaviors of modeled rockfill materials at different intermediate principal stress ratios.

This paper presents a numerical study on culverts under rigid pavements subjected to static traffic loading. The numerical study was based on the results of a comprehensive field study of a single-cell low-fill box culvert under three roadway sections (concrete pavement, concrete shoulder, and unsurfaced fill). The culvert in the field test was instrumented with displacement transducers and pressure cells to capture the deformations and pressures resulting from different combinations of wheel loads from a test truck pulling a low-boy trailer loaded with a backhoe. Deflections under the culvert roof and pressures on the culvert were measured during loading. Properties of soil and pavement layers were determined in the laboratory. A three-dimensional (3D) numerical model of the culvert was developed using a finite-difference program. The numerical model with material properties was verified with the field test results. A parametric study was conducted to investigate the influence of cement concrete pavement thickness, fill depth, and culvert span on the load distributions over the culvert under wheel loads. The intensity of the vertical pressure gradually decreased with an increase in the pavement thickness, fill depth, and culvert span. However, the effect of the span on vertical pressures decreased with a top slab designed to control excessive deflections. The comparison of the calculated vertical pressures by the numerical method and the AASHTO distribution methods demonstrated that the current AASHTO pressure distribution methods are overly conservative for the wheel load distribution on a low-fill box culvert under a rigid pavement.

Undrained cylindrical cavity-expansion solutions that were based on the modified Cam-clay model have been widely used to predict and interpret geotechnical problems, such as pile installation, and in pressuremeter tests in saturated clay. However, the yield and failure of soils around the cavity might not be properly modeled by the extended Mises criterion adopted in the modified Cam-clay model. To consider the three-dimensional mechanical properties of soils during cavity expansion, an analytical solution with the spatially mobilized plane criterion-based Cam-clay model is derived for the undrained cylindrical cavity-expansion problem by using the stress-transformed method. The presented solution is verified by two well-documented pile-installation tests and is compared with the modified Cam-clay model–based solutions in detail. The results show that the prediction from the presented solution is closer to the actual stress field induced by the undrained cylindrical cavity expansion. However, the radial stress and excess pore-water pressure are overestimated, whereas the plastic zone radius is underestimated by the conventional modified Cam-clay model–based solutions.

Large-diameter monopiles are widely used foundations for offshore wind turbines. In the present study, three-dimensional finite-element (FE) analyses are performed to estimate the static lateral load-carrying capacity of monopiles in dense sand subjected to eccentric loading. A modified Mohr–Coulomb (MMC) model that considers prepeak hardening, postpeak softening, and the effects of mean effective stress and relative density on stress–strain behavior of dense sand is adopted in the FE analysis. FE analyses are also performed with the Mohr–Coulomb (MC) model. The load–displacement behavior observed in model tests can be simulated better with the MMC model than with the MC model. On the basis of a parametric study for different length-to-diameter ratios of the pile, load–moment capacity interaction diagrams were developed for different degrees of rotation. A simplified model, based on the concept of lateral pressure distribution on the pile, is also proposed for the estimation of its capacity.

Two key issues in landslide forecasting are finding the critical slip surface and determining its parameters. Generally, these issues are two parts of one problem. To solve this problem, a new two-phase search method is proposed: a newly proposed inversion method to determine the critical slip surface parameters, using a new bionics algorithm (i.e., particle swarm optimization), and based on the measured slip surface and the critical slip surface-searching process. The proposed method is verified by using two typical numerical examples used by the Association for Computer Aided Design (Australia) and an engineering example of a highway landslide. The results indicate that when the search over a critical slip surface and the inversion of its parameters are performed simultaneously, the calculated factor of safety is in good agreement with that of the actual scenario, whereas the factor of safety computed on the basis of known parameters of a slide body will deviate from that of the physical scenario.

The variation of pullout capacity of vertical strip anchor plates with width was studied while considering the effect of mean normal stress on the peak friction angle in a granular soil domain. This analysis was performed by using lower-bound finite-element limit analysis in conjunction with linear programing. The study was conducted for a wide range of anchor widths (B) and embedment depths (H) for both loose and dense sands. It was determined that the pullout capacity factor (F_γ = P_u/γBH) increases with decreasing values of normalized anchor width (γB/σ_A), where P_u is the ultimate pullout load; γ is the unit weight of soil mass; and σ_A is the atmospheric pressure. The rate of change for increments of F_γ with γB/σ_A was found to be more pronounced when the anchor is embedded in loose sand and at greater embedment depths.

A reliability-based analysis of the bearing capacity of strip footings considering the spatial variability of undrained shear strength of clay was performed using the random FEM. The effect of anisotropic correlation of spatially varying undrained shear strength on the mean and standard deviation of bearing capacity was obtained using the random FEM. It was found that the anisotropic correlation of undrained shear strength significantly affects the bearing capacity. The mean bearing capacity was found to be minimal for a specific combination of horizontal and vertical scales of fluctuation. The coefficient of variation of the bearing capacity increased with increases in horizontal or vertical scales of fluctuation, or both. The factor of safety required to achieve a specified target failure probability for different scales of fluctuation and coefficients of variation of the undrained shear strength of clay is presented.