Geotechnical Frontiers 2017: Seismic Performance and Liquefaction

Site response analyses (SRA) provide a means to assess the seismic wave propagation phenomenon in shallow deposits and capture its influence on ground motions. One of the key model assumptions in SRA involves the characterization of the attenuation at the site. Typical damping models are developed by testing small-scale soil samples in the laboratory. These procedures can only characterize material damping and fail to capture other sources of attenuation as they occur in the field (e.g., scattering). The spectral decay parameter, kappa, is used in this study to define alternative damping models. These combined models of geotechnical and seismological attenuation descriptors provide larger estimates of damping than laboratory-based models. The most important practical application of this study is the definition of the portion of attenuation that is ignored when solely relying in dynamic laboratory testing. Insights on similarities between kappa and damping values used in geotechnical models are also provided.

Seventeen soil specimens were prepared in the laboratory as mixtures of commercially available montmorillonite or kaolinite with quartz and used to conduct cyclic simple shear tests. The soils were subjected to different amplitudes of strain-controlled sinusoidal loading at a frequency of 0.5 Hz. The results obtained were used to determine the relationship of the strain-dependent dynamic properties. Modulus reduction curves were developed and found to be dependent on both the plasticity index and the mineralogical composition of the soil specimens. Specifically, for two soils with the same plasticity index, but different mineralogical compositions, a lower reduction in the shear modulus with shear strain was observed in the soils containing montmorillonite as the clay mineral in comparison to the soils containing kaolinite as the clay mineral. Furthermore, at a given shear strain, the damping ratio in the soils with montmorillonite as the clay mineral was lower than the damping ratio for the soils with kaolinite as the clay mineral.

Case histories have demonstrated the disastrous consequences of cyclic mobility in the underlying fine-grained cohesive soils. Cyclic mobility is the reduction in shear strength resulting from an increase in the pore pressure in soils from dynamic loads. This increase in pore pressure will cause a reduction in the effective stresses in the soil mass. In this study, over twenty soil samples were subjected to sinusoidal cyclic loading in the cyclic simple shear apparatus at a frequency of 0.5 Hz. Immediately following the cyclic loading phase, the shear strength of the soils was determined. The pore pressure response of the fine-grained soils was back-calculated from changes in the normal stress in order to maintain constant volume in both phases of the testing. The results were used to determine the stress paths followed by the soils during the cyclic and post-cyclic loading phases. The behavior observed demonstrated that the post-cyclic shear strength of the soils could be calculated by estimating the pore pressure as a result of a cyclic loading. This pore pressure could be used to determine the effective stresses on the soil just following the cyclic loading and thus, compute the overconsolidation ratio for the soil. The SHANSEP principles could, then, be applied to determine the relationship between the undrained strength ratio and the overconsolidation ratio. Relationships between the multiplier and exponent of a power function representation of the variation between the undrained strength ratio and the overconsolidation ratio with the liquid limit and the plasticity index are proposed.

Soft clay deposits around the world pose several issues including low bearing support, high compressibility and an amplification of earthquake motions. As a result, ground improvement techniques have been implemented in order to improve the engineering properties of these soft soils. Furthermore, if these ground improvement techniques can help to reduce the amplification of seismic waves as they travel through soft soil deposits, the demands placed on infrastructure can also be reduced. In this study, the effectiveness of deep soil mixed panels on reducing the seismic amplification through a soft clay profile is examined with the use of an unidirectional shake table. The influence of replacement ratio of these panels is evaluated by preparing three models representing replacement of 0%, 10% and 20% of the plan area of the unimproved soil. Similar models were also prepared with compacted soil-cement panels and used for comparison. The amplification of the seismic waves was found to reduce as the replacement ratio of the deep soil mixed panels increased. Relationships between the replacement ratio and reduction in seismic amplification are presented in this paper.

The paper presents a centrifuge test of a level site consisting of granular soil deposits and an embedded structure subjected to various bi-axial base excitations. The test was conducted at RPI NEES centrifuge facility to assess the dynamic response characteristics of soil-structure interaction (SSI) under multidimensional conditions. Synthetic sinusoidal waves were used as base excitations to test dense model under biaxial shaking. A dense array of accelerometers was used to monitor the deposit response (beneath and outside of the structure, along with pore water pressure transducers. The observed acceleration and pore pressure are used along with non-parametric identification procedures to estimate the corresponding dynamic shear stress-strain histories. The measured results along with the obtained histories are used for two purposes. First, to shed light on the mechanisms of soil-structure interaction under biaxial shaking. Secondly, to show the difference in soil behavior at different locations beneath the footing when subjected to biaxial shaking. The later objective is evident when comparing soil stress-strain loops and dilation behavior at locations beneath the structure, and those measured in the free field.

The shear stress–shear strain behavior of soil is characterized by a hysteresis loop. The geometry of the loop over one cycle of loading can be used to establish a simple equation that relates secant shear modulus to shear damping. The equation applies to values of modulus and damping obtained at all values of shear strain. No other information is needed. Two simple hyperbolic expressions, with only two coefficients, accurately describe both the shear modulus degradation and damping increase with shear strain. The equations include reference strain to allow the results to be relatively independent of confining stress. They are obtained with an optimization program in a spreadsheet. Results from resonant column tests on sand, silt and clay soils demonstrate their efficacy for determining minimum and maximum values of damping.

Unsaturated soils are commonly encountered in natural soil deposit above the ground water table and in compacted soil structures. The use of compacted soils in earthquake prone areas is increasing. The accurate prediction of dynamic behavior of unsaturated soils is of practical importance to many geotechnical engineering problems, such as soil-structure interaction and liquefaction evaluation. The hydraulic hysteresis of soil water characteristic curves during wetting and drying cycles is not accounted for in most of the numerical analysis of dynamic behavior of unsaturated soils. To better understand the dynamic behavior of unsaturated soils, a fully coupled flow-deformation finite element code U_DYSAC2 with solid displacements, pore water pressure and pore air pressure as nodal variables is used to analyze an unsaturated embankment subjected to base shaking. A bounding surface elastoplastic model is utilized to describe the stress-strain behavior of unsaturated soils. The hydraulic hysteresis is also taken into account using a model developed based on the bounding surface plasticity theory. The U_DYSAC2 results show that the analysis with hydraulic hysteresis predicted smaller vertical displacements and larger suction change than the analysis without hysteresis.

The GMP (Green, Mitchell and Polito) pore pressure generation model uses the energy dissipated in a soil during cyclic loading to predict the residual pore pressure rise in a saturated soil subjected to cyclic loading. In addition to the shear stresses and shear strains applied to the soil deposit being modeled, the GMP model requires a single calibration parameter referred to as the “pseudo energy capacity” or “PEC”. The study reported in this paper examined the validity of using constant-volume cyclic simple shear tests to develop values of the PEC for input into the GMP model. The study performed consisted of using the PEC developed from a series of constant-volume cyclic simple shear tests performed on a loose sand to predict the residual pore pressures generated during a series of stress- and strain-controlled cyclic triaxial tests performed on the same sand prepared to the same density. It was found that once the pseudoenergy capacities from the constant-volume cyclic simple shear tests were scaled to account for the effects of specimen size, they did an excellent job of predicting the pore pressure generation that occurred during the cyclic triaxial tests.

Soil behavior is inherently nonlinear, including the strain-dependent modulus and damping values measured in the resonant column apparatus. This paper describes the frequency response behavior in the resonant column test using a theory that is valid for all frequencies including those that differ from the resonant frequency, and for strain amplitudes ranging from near zero to those approaching failure. The theory is based on the ASTM D4015-15 Standard for devices having a fixed specimen base, with torque and rotation measured at the top of the specimen. The theory accounts for damping in the apparatus. Results show that the frequency response at all shear strain amplitudes is significantly different from that predicted by linear viscoelastic theory, especially at small shear strains where both shear modulus and damping are most highly dependent on shear strain amplitude. Proof of this dependency at low strains is provided by plotting shear modulus and damping to natural scales rather than logarithmic scales as is traditionally done.

Shear wave velocity is an essential parameter to characterize the dynamic behavior of the soil in small-scale samples, centrifuge models as well as field deposits. In this paper, the bender elements system in the centrifuge lab at Rensselaer Polytechnic Institute is used to characterize the shear wave velocity and the small strain shear modulus of dry and saturated Ottawa sand. It was found that the shear wave velocity of saturated sand is smaller than that of dry sand having the same relative density. Also, the paper describes an analytical system identification (SI) method that can be used in the absence of bender elements to estimate shear wave velocity from acceleration recordings. The SI and bender elements were found to be in very good agreement with each other confirming the validity of both techniques, which are fully independent.

Unlike dynamic properties of municipal solid waste (MSW), the dynamic properties of engineered fill landfill covers at MSW and other landfill sites have not been extensively investigated. The two main reasons for the dearth of research on this topic are: (i) design engineers realize that modern landfill covers are relatively thin (on the order of 1.5 m, or less), and hence, their influence on the overall landfill response is assumed to be small; and (ii) there are readily available sets of dynamic soil properties of cohesive soils that could be assigned to landfill covers soils based upon the results of index testing. However, not all landfill covers are “thin,” and the readily available sets of dynamic soil properties may not be applicable for low-plasticity compacted and overconsolidated soils such as landfill covers. At many old landfills and Superfund sites, landfill cover thickness readily exceeds 5 m, and covers can be as thick as 10 to 15 m. These thicker covers may have significant influence on the overall response of landfills, and hence, careful evaluation of cover material properties is warranted. In this paper, results of in-situ nonlinear testing of landfill cover soils was used to develop modulus reduction curves. Shear strains in the cover soils ranging from 10^-4 % to 2×10^-2 % were induced by Vibroseis shakers. The material damping curves were estimated from modulus reduction curves by means of a nonlinear constitutive model, Masing rules, and engineering judgement. The modulus reduction and damping curves presented herein may be used for seismic design of landfill covers and other engineered fills constructed of soils of similar plasticity and with similar stiffness (i.e., shear wave velocity) and compaction characteristics.

In this study, normally consolidated clays were subjected to undrained cyclic simple shear followed by drainage. Then the shear strain amplitudes (g) and number of cycles (n) were changed and the effects of Atterberg’s limits on the pore water pressure accumulation and recompression were investigated. In conclusion, the pore water pressure accumulation and the settlement of saturated clays are significantly affected by the Atterberg’s limits. For Kaolin with low plasticity index (I_p), the pore water pressure rapidly increases and becomes stable after relatively small number of cycles n. Meanwhile for Tokyo Bay clay and Kitakyushu clay with higher I_p, the cyclic shear-induced pore water pressure is rather small leading to smaller post-cyclic settlement. The cyclic recompression index (C_dyn) of Kaolin is unchanged with n while those of Tokyo Bay clay and Kitakyushu clay significantly decrease with n. By using these results, the post-cyclic settlement of clays was sucessfully predicted.

The seismic performance of many nuclear power plant (NPP) structures depends on the cyclic shear stress-shear strain-volumetric strain behavior of dense, compacted coarse-grained soils used to support their foundations. However, when these deposits are thick, even relatively small volumetric strains can result in nontrivial settlements that can impact NPP structures. Here, we describe the results from dynamic centrifuge tests performed on thick layers (prototype thicknesses of 10.25 and 20.5 m) of saturated dense (relative density, Dr~95%), Ottawa sand. Models were excited using unidirectional (1D) and bidirectional (2D) broadband motions (Arias intensities ranged from 0.1 to 5 m/s). Centrifuge test results indicate minor differences between 1D and 2D response spectra. In contrast, 2D shaking in dense sands caused increases in porewater pressure (PWP) generation and volumetric strains (ε_v) of approximately 200% compared to 1D shaking. These increases in PWP and ε_v were considerably larger than observed by others for loose sands.

A series of cyclic simple shear tests were conducted on reconstituted specimens of clean, fine-grained Ottawa C109 sand at three relative densities. Small-amplitude drained cyclic preshearing strains of 0.02-0.3% were applied to the specimens for 30-1000 cycles. After drained cyclic preshearing, the specimens were subjected to higher amplitude cyclic shear strains under constant-volume conditions until liquefaction occurred. The test results are compared with constant-volume cyclic shear tests without preshearing to evaluate the influence that strain amplitude and the number of cycles of drained cyclic preshearing have on the liquefaction resistance of Ottawa C109 sand. Based on this comparison, a framework for understanding drained preshearing is proposed.

In order to monitor the ongoing seismicity in different parts of the active regions of India, 300 state of the art digital strong motion accelerographs are installed. The accelerograms from the recording stations are available in PESMOS website. The recording stations are classified based on the physical description of local geology and not based on in-situ field tests. Thus, there exist an ambiguity in the assessment of local site conditions for these recording stations. Unless the site class (SC) is known, recorded ground motions cannot be used with confidence. Several research efforts have been directed towards the use of horizontal to vertical spectral ratio (HVSR) technique as a tool to access the local site condition. There is consensus among the scientific community regarding the peak frequency (f_peak) obtained from the HVSR method to identify the soil fundamental frequency. In this study, theoretical HVSR curves for 12 recording stations located in the Tarai region of Uttarakhand are generated using a Matlab tool called Model HVSR to obtain the fpeak and peak amplification (A_peak). Further, these values of fpeak and Apeak for all the 12 recording stations are estimated using the HVSR based on strong ground motion records. Results obtained from both the above methods show 1:1 correspondence. The values of fpeak obtained for 12 stations range from 1 Hz to 5 Hz based on model HVSR and from 1.1 Hz to 3.3 Hz using the HVSR based on strong ground motion records. These findings are also matching with the existing field studies. Present analyses show a clear difference in the SCs suggested by PESMOS and the one obtained from the present study. Once correct information on SC is known, recorded ground motion at each station can be used with confidence.

Structures founded on dense, compacted sands are currently designed based on the shear stress–shear strain (γ_cs)–volumetric strain (ε_v) response of clean sands primarily obtained from unidirectional element tests. This may lead to underestimation of ε_v under multidirectional seismic loads and incorrect settlement estimation for sands with fines. This paper presents results from first of a kind strain-controlled unidirectional (sinusoidal) and bidirectional (circular, figure-8) cyclic simple shear tests. Clean Ottawa sand and a non-plastic Mississippi River silty sand were tested at modified Proctor relative compactions (RC) of 92 to 99%. At the same RC, presence of non-plastic fines decreased ε_v. Volumetric strain under bidirectional loading normalized by ε_v under equivalent unidirectional loading (comparable γ_cs, RC, overburden pressure) was found to be a function of the number of loading cycles, load path and γ_cs. A figure-8 loading path yielded higher settlement than an equivalent circular loading path.

Heap leach pads consist of layers of blasted and/or crushed ore placed over a liner containment system from which metallic substances are harvested via solution circulated into the heap. Also associated with this mining approach are waste dumps where the low-grade ore materials are deposited. When these pads and dumps are constructed in seismically active areas (such as the Andes Mountains in Peru), determination of the linear and nonlinear shear moduli of the materials used to construct the leach pads and waste dumps become essential in evaluating seismic slope stability of these systems. In this paper, the linear and nonlinear shear moduli of the two types of blasted mine materials determined from laboratory and in-situ testing are presented. Resonant column (RC) and cyclic triaxial (CTX) tests were conducted on reconstituted specimens at several mean effective stresses at shear strains ranging from the linear to highly nonlinear range. The two materials exhibit dynamic behavior generally consistent with that of gravels. Finally, in-situ shear wave velocity measurements from the final placement were conducted and the results are in good agreement with the shear wave velocities determined from RC testing.

The material properties and the geometry of near surface soil layers, known as local site conditions, can significantly change the input seismic motion compared to the simple case of homogeneous linear elastic half-space. Our recent studies have shown that the effects of topography coupled to site response can lead to ground motion aggravation larger than the superposition of site and geometry amplification. These soil-topography coupling effects arise from seismic waves trapped in the near surface soil layers, are amplified or deamplified as a consequence of stiffness contrast, and are further modified due to scattering caused by irregular interface and ground surface. In this study, we investigate the coupling effects for 2D idealized convex features through a systematic analysis. The resulting trends, which are presented in the form of dimensionless amplification factors, clearly demonstrate the nonlinear nature of coupling effects, which cannot be predicted by modifying simulations of topography effects on rock by 1D site amplification factors, a posteriori. We then quantify these coupling effects through 3D site-specific analyses at selected strong ground motion stations in California, which yield more realistic amplification patterns (using 1 arc-second DEM extracted surface topographies and measured Vs profile). The results of coupling effects provide a basis as to how it can be incorporated in the proposed design motion of seismic code provisions.

In practice, seismic site response analysis is often conducted with the entire soil profile being divided into a number of sub soil layers and site-specific parameters, such as shear wave velocity (V_S), damping and modulus reduction, are assumed to be constant within the extent of each layer. In this study, the effect of large contrast in V_S at shallow layer interfaces on the computed surface response is investigated considering the geologic and seismic settings for the Charleston, SC, area. The V_S profiles with sudden contrast at shallow depth were smoothened systematically using various hypothetical gradients to obtain derived Vs profiles. The computed responses from the original and the derived profiles were then compared. The hypothetical profiles produced lesser shear strains and greater surface accelerations as compared to the corresponding original profiles with significant V_S contrast. Therefore, such simplification of V_S profiles should be avoided in site response analysis, unless they represent actual geologic conditions.

When the subsurface profile at a given site cannot be simplified as horizontal layers or the depth to bedrock varies across the profile, validation of a typical 1-D soil column analysis is desirable in calculating the site-specific response spectrum. This point is illustrated through a detailed analysis of a multi-story building project site underlain by very soft marine clay deposits. The subsurface profile and the corresponding dynamic soil parameters are determined through field exploration, geophysical testing, and laboratory testing. Consequently, a 2-D dynamic finite element analysis is performed to provide insight into the effects of a sloping rock. The results indicate that the short period response across the sloping rock is substantially magnified in the 2-D analysis. The 1-D soil column underestimates the short period amplification factor of the spectral acceleration F_a while the 1-sec period amplification factor F_v generally increases with the thicker soil profile. The boundary effects of a 2-D analysis are also discussed.

This study quantifies uncertainty in Central and Eastern North America (CENA) Reference Rock (RR), Weathered Zone (WZ), and resulting earthquake ground surface site response predictions. The assumptions used for modeling the WZ have an effect on the predicted ground surface response. Random modeling of the RR and WZ is implemented using a simple Taylor Series expansion. Equivalent linear site response analysis is performed with a synthetically generated motion and includes the site attenuation parameter, kappa (κ), for modeling site damping. Probabilistic site response spectra and ratios of response spectra are presented. This study introduces a simple method for quantifying uncertainty in ground surface response while capturing RR and WZ variability.

A decomposition is used to express the mean squared deviation, quantifying the dissimilarities between time histories of input (or response) quantities of multiple replicas of a soil system centrifuge test, as a unique aggregate of three discrepancy measures associated with shape, phase and frequency-shift. The shape measure quantifies the deviations associated with dissimilarities in form and amplitude. The phase measure estimates the deviations associated with differences in phase angle. The frequency-shift measure quantifies the deviations associated with differences in frequency components. These measures are illustrated using simple synthetic motions and used to assess the discrepancies among six replicas of centrifuge input motion achieved at six different facilities. The conducted analysis shows that the proposed decomposition accurately quantifies the different types of discrepancies between time histories.

As the profession moves toward a performance-based approach for the evaluation and mitigation of liquefaction, increasingly the performance of the soil-foundation-structure system needs to be evaluated for a wide range of conditions. The influence of different soil and structural properties as well as intensity measures (IMs) need to be evaluated, before the system performance can be predicted in a probabilistic manner. Fully-coupled, 3D, nonlinear finite element analyses were performed to evaluate the integrated response of the soil-foundation-structure system on a layered liquefiable soil. After the numerical results were validated using centrifuge experiments, a parametric study followed to evaluate the influence of soil relative density, structure’s height/width ratio, fixed-base fundamental period, and foundation pressure as well as IMs on building’s permanent settlement and peak inter-story drift ratio. The results point to the importance of considering building’s inertial properties and interaction with soil when evaluating its settlement, rocking, and deflection on liquefiable ground.

In this paper, a classical CPT-based liquefaction model is integrated with geostatistical tools and random field models to evaluate and map liquefaction potentials over an extended area. Verification of random field-based liquefaction prediction is challenging and rarely addressed in current literature due to limitation of soil data and liquefaction observations. To this end, a three-dimensional synthetic digital soil field is generated through novel numerical models, providing extreme detailed soil properties and corresponding liquefaction hazard information quantified in terms of the liquefaction potential index (LPI). Different virtual field testing plans are designed to investigate the effect of data inference on the model performance. Three ranking criteria are used to evaluate model performance. Knowledge gained through the model verification process can be used to guide the inference of model parameters and understanding the performance of random field-based liquefaction hazard mapping.

Soil softening and soil liquefaction have long been a significant aspect of seismic risk, including at sites with gravelly soils. However, the performance of gravelly soils during and after earthquake events is still not well understood. A prototype large-size cyclic simple shear (CSS) device is used in this research to perform monotonic and cyclic shear tests of a gravel, a sand, and a gravelly sand mixture to evaluate the shear response of these materials. The importance of vertical strain control on constant volume behavior is also investigated for a uniform gravel material. Comparisons are made between monotonic, cyclic and post-cyclic shear response of a gravelly sand to show the relationship between these different loading conditions. The results of these tests showed that the same ultimate state line (friction angle) is attained for this material from both monotonic and post-cyclic loading.

This paper presents results of one-dimensional (1-D) liquefaction vulnerability index (LVI) analyses and two-dimensional (2-D) nonlinear deformation analyses (NDAs) of a potential lateral spreading site adjacent to a channelized segment of the meandering Çark River in Adapazari, Turkey, during the 1999 M7.5 Kocaeli earthquake. The site is underlain by fluvial deposits of clayey fine-grained sediments interlayered with potentially liquefiable sandy sediments. No lateral spreading damage was observed at the site after the Kocaeli earthquake despite an estimated peak ground acceleration (PGA) around 0.4 g. Current liquefaction triggering and lateral spreading analyses predict significant lateral displacements; results obtained with a multilinear regression model were presented by others and results obtained with LVI methods are presented herein. Additional LVI analyses were performed to consider the effects of transition and thin-layer corrections, different criteria for distinguishing between sand-like and clay-like soils, and site-specific fines content estimation. NDAs with stochastic realizations of interlayered sand and clay stratigraphy estimated from available CPT data were performed to assess the impact of accounting for 2-D site geometry, spatial variability in stratigraphy, and dynamic response. The realizations were produced using a transition probability geostatistical approach, and representative soil properties were selected by statistical examination of the properties estimated from the CPT data. Descriptions of the methods used for the LVIs and NDAs are included and the computed lateral displacements are compared to field observations. Lateral displacements from all LVI analyses are significant, while those from the NDAs are sufficiently small to be generally consistent with the absence of observable damage at the site. These comparisons illustrate that a realistic representation of the site and the dynamic loading conditions beyond what is accounted for by LVI methods can be important for evaluating potential lateral spreading in these types of interlayered sand and clay deposits.

A fully coupled numerical analysis using Biot’s theory with u-p formulation considering variable permeability during liquefaction is used in this study to simulate liquefaction induced lateral spreading phenomenon. A centrifuge test performed on liquefiable soil at Rensselaer Polytechnic Institute is simulated using the developed model. The numerical results are in good agreement with experimental observations. A number of numerical simulations under different geometric, site-specific, and ground acceleration parameters have been carried out in the course of this study. Based on the statistical analysis conducted on the results of 31 different models, a new relation is proposed for predicting the liquefaction-induced maximum lateral displacement for gently sloping grounds. The proposed relation can predict the maximum ground lateral displacement obtained by the rigorous finite element analysis with less than 10% error. The predicted lateral displacements are in a good consistency with the ground movement observations in 16 different centrifuge tests as well as more than 250 field observations. The proposed relation in this study shows a better match compared to other well-known relations.

An evaluation of liquefaction triggering and deformation analyses for both banks of the Napa River at the crossing of the Napa Valley Wine Train Bridge during the August 24, 2014 Magnitude 6.0 Napa Earthquake is presented. The subsurface stratigraphy, including differences between the point bar and cut bank sides of the river channel, are described using results of cone penetration test (CPT) soundings and borings with standard penetration test (SPT) data. Liquefaction triggering analyses are performed using a simplified procedure and equivalent linear site response analyses. The potential for lateral deformations and settlements are evaluated using one-dimensional liquefaction vulnerability indices and Newmark sliding block methods. The analysis results indicate that liquefaction effects would not be expected on the cut-bank side of the channel, whereas minor amounts of liquefaction-induced settlements would be expected on the point bar side. There were no visible or reported ground deformations, cracks, or differential settlements at or near the bridge abutments on either side of the channel. Possible reasons for the discrepancy between observed and predicted behavior on the point bar side of the channel are discussed.

Microbial induced calcite precipitation (MICP) is a biochemical reaction that takes place within the soil producing calcite cementation bonding soil grains together. In this study, MICP method was applied to both Nevada sand and Nevada sand containing 15% silt to assess their improvement in monotonic shear strength, cyclic resistance and also changes in permeability. The change in shear strength properties are evaluated by performing undrained monotonic and cyclic direct simple shear tests on both treated and untreated samples. Reduction in permeability as a result of calcite precipitation is also another objective of this research which is investigated by performing constant head tests on untreated and MICP treated samples. The results show improvement in shear response and reduction in excess pore water pressure in both treated samples. The cyclic resistance also increases by applying MICP. The permeability of soil reduces by applying MICP but the amount of reduction is low in comparison to the change in permeability when fines are added to the Nevada sand.

In this study, the effect of radio-frequency (RF) waves on the generation of excess pore-water pressure (EPWP), and hydraulic conductivity of coarse-grained soils is examined. Apparent viscosity of water can change due to the oscillation of dipole molecules within an alternating electric field. Altering the viscosity alters the hydraulic conductivity and can, hence, minimize the buildup of the EPWP. Constant-head permeability tests were performed to measure the hydraulic conductivity of two types of samples, glass beads and natural sand with/without RF waves. Impact tests were performed on a box made of Plexiglas to measure the EPWP of glass-bead samples. Different sets of tests were conducted to evaluate the effect of RF waves’ power and frequency. The results showed a positive correlation between the RF waves power and hydraulic-conductivity alteration whereas no change was observed in the generation of EPWP.

The results of a numerical study to predict the response of a Pleistocene uncemented sand deposit during an in situ liquefaction test with a mobile field shaker are presented in this paper. A plasticity model is used to represent cyclic stress-strain behavior of the sand. Input model variables are calibrated using results from field geotechnical testing and laboratory monotonic triaxial testing of intact specimens obtained by in situ freezing. The calibration required considerable adjustment of one primary input variable to account for increased liquefaction resistance due to aging processes. The simulations under amplitudes representative of the mobile shaker’s highest output predicted an initial tendency towards contraction and maximum excess pore pressure ratios (r_u) of < 20%, followed by a dilative response and subsequent reduction in pore pressure.

In this paper, the PM4Sand plasticity model developed by Boulanger and Ziotopoulou is investigated for predicting soil behavior under cyclic loading for sands containing substantial amounts of low plasticity fines. In this study, the model is compared with test data for sands with higher fines contents of low plasticity in the literature. Results from experimental studies on silty and clayey sands at different relative densities and confining pressures are used in the comparison. Simulated and measured values of number of cycles to liquefaction triggering and pore pressure response are compared. The results show that PM4Sand can predict the number of cycles to liquefaction and pore pressure response with reasonable accuracy for a variety of cyclic stress ratios for fines rich sands at different relative densities consolidated to a range of confining pressures. Limitations to the use of a sand plasticity model for soils with substantial fines are also discussed.

Effective mitigation of liquefaction requires a reliable evaluation of liquefaction triggering and its consequences in terms of excess pore pressures, accelerations, and displacements. However, reliable prediction of all these key response parameters remains challenging in the past. In this paper, the results of a centrifuge experiment modeling of a layered soil profile, including a liquefiable layer of Ottawa sand, are used to evaluate the predictive capabilities of two state-of-the-art constitutive models. The models were first calibrated using the same set of monotonic and cyclic triaxial tests and were then used to simulate the seismic performance of a layered soil deposit to a horizontal earthquake motion. This paper presents the systematic calibration process adopted for each constitutive model, followed by a comparison of the numerical results with centrifuge recordings. This effort aims to provide insight into the strengths and limitations of the adopted models.

Numerical predictions of the liquefaction experiments and analysis project (LEAP) centrifuge tests were revisited in order to evaluate the consistency of the numerical approach and the degree of agreement between numerical predictions and recordings for each facility. The calibration of the numerical model is examined against three Ottawa Sand laboratory data sets, and the validity of the original calibration is confirmed. The recordings of the centrifuge experiments for each facility are compared to the corresponding numerical model analysis and results are presented in terms of excess pore pressure time histories and spectral accelerations. The comparison shows that the numerical model, within the range of uncertainties involved, can satisfactorily predict the results of the experiment for most facilities, but experimental challenges of the experiments for few facilities result to an inconsistent agreement. Additional research is required to identify the underlying causes for the discrepancies but the project is deemed an overall success in yielding high-quality testing results, providing insights for future tests, and demonstrating the predictive capabilities of the numerical approach followed.

Porewater pressure (PWP) generation leads to soil softening and potential liquefaction of sandy soils during earthquakes, and can decrease ground motion amplitudes at high frequencies and increase the predominant period of shaking. This paper presents an evaluation of the generalized quadratic/hyperbolic constitutive model coupled with two PWP generation models, termed GQ/H+u, as implemented in DEEPSOIL V 6.1. The GQ/H+u model can represent large-strain shear strength and thus provides more realistic response at liquefiable sites where large strains are expected. First, sets of cyclic direct simple shear tests were used to evaluate the models. Comparisons of measured and computed stress-strain loops and excess PWP generation illustrate that the model reasonably captures measured cyclic stress-strain and PWP responses. The GQ/H+u model also can reasonably capture acceleration time histories and response spectra measured in centrifuge tests in loose and dense sands subjected to strong shaking.

Very limited number of computational studies has been presented for the analysis of the response of saturated granular soils to multidirectional shaking despite it being the realistic mode of loading that resembles an actual earthquake. Herein, we examine the capabilities of a recently developed coupled lattice Boltzmann method-discrete element method technique to model level and gently sloped soil deposits when subjected to bidirectional shaking. Results of conducted simulations show that bidirectional shaking may increase surface settlement of level deposits by about 30% compared to unilateral shaking. The depth along the deposit that experiences excess pore pressure ratio close to unity increases because of bidirectional shaking compared to unilateral shaking. In addition, bidirectional shaking increases the magnitude of lateral spreading and associated shear strains in sloping deposits.

The expansion of the Port of Long Beach-Pier S, a major U.S. port facility, has been completed through the past decades by placing loose man-made hydraulic fills. These fills and their subjacent natural estuarine and marine deposits, have shown to be susceptible to liquefaction. This case of study is located within a few miles of the Newport-Inglewood and the Palos Verdes faults, which represent a high seismic hazard to the port. In this paper, numerical analyses are performed using the UBC3D-PLM constitutive soil model to determine the onset of liquefaction and estimate free-field ground-induced settlements based on post-liquefaction excess pore water pressure dissipation. It is concluded that this model, even with certain limitations, is able to predict the onset of liquefaction capturing numerous features of dynamic soil behavior, and can be used as a first estimate of liquefaction-induced settlements arising from dissipation of excess pore water pressures.

The seismic performance of nuclear power plants (NPPs) constructed on compacted, dense sands depends on their cyclic response. NPPs founded on a thick deposit may experience nontrivial settlements due to small, but accumulated volumetric strains during an earthquake. Nonlinear, three-dimensional (3-D) numerical simulations can be utilized to assess the deformations of the soil-structure system. This work examines the performance of a class of distributed element constitutive models which has a parallel-series Iwan type distribution of elastic - perfectly plastic nested components. The performance evaluation is limited to deviatoric space since the volumetric component is yet not implemented. Soil-structure 3-D interaction and single element numerical simulations are compared to dynamic centrifuge and cyclic direct simple shear (cDSS) tests respectively. The simulations reasonably capture the measured shear behavior in small to medium shear strain ranges achieved in the centrifuge. However, the constitutive model deviates from measured response at the large shear strain levels achieved in cDSS.

This paper presents numerical liquefaction simulations of a hypothetical sand in cyclic direct simple shear tests for four different constitutive models. The four models are PM4Sand in FLAC, UBCSand in FLAC, Pressure Dependent Multi Yield 02 (PDMY02) in OpenSees, and the Manzari-Dafalias04 model in OpenSees. Parameters published by others for various sands were used for this work to avoid possible introduction of our bias into the calibration process. The material properties were determined by others for different uniformly graded sands, all with a relative density of approximately 50%. The simulations do not pertain to one sand or one set of laboratory data, so therefore, there is no right or wrong answer. Instead, the goal of this paper is compare a consistent set of results that show the implementations of each model behave as expected, and to illustrate basic differences in behavior of the different models. Cyclic strength curves (cyclic stress as a function of number of cycles) illustrate the behavior of the models over a range of cyclic stresses. Each model displays pore pressure build up, softening, and cyclic contraction-dilation cycles associated with cyclic mobility. Two of the models soften to a point, but then stabilize in a repeated hysteresis loop with no additional growth in the cyclic strain amplitude after some number of cycles.

Characterization of soil response under cyclic loading is one of the major challenges in evaluating liquefaction triggering. In this paper, we have performed numerical simulations to study dissipated energy and accumulated damage in soil structure at onset of liquefaction. For this purpose, at first, we validated Plasticity Model for Sands (PM4Sand) in capturing soil cyclic response with findings in experiments. Thereafter, the model was utilized to simulate soil behavior during uniform cyclic loading under controlled boundary conditions and stress paths. Simulations were performed on soils with different relative densities and under different confining pressures. The results of this study indicate that energy dissipation is directly related to PWP generation, and is independent of the amplitude, form and frequency of loading. Dissipated energy can be utilized as a versatile metric to characterize soil strength degradation and liquefaction triggering during cyclic loading.