Geo-Congress 2019: Earthquake Engineering and Soil Dynamics

To address well-documented lifeline system vulnerability to ground deformation, a new generation of hazard-resilient pipeline products have been developed. These systems require evaluation for extreme loading conditions before they are accepted as viable options for design in hazard-prone regions. Presented is an experimental evaluation of a steel pipeline with special structural inclusions designed to accommodate significant ground displacement while providing continual water conveyance for post-disaster response. In addition to material characterization, axial tension/compression, and four-point bending tests, the evaluation includes a large-scale fault rupture experiment with an 8.5-in. (216-mm) diameter, 30-ft (8.8-m) long pipeline buried in sand with wave features positioned at either side of a right-lateral, strike-slip fault crossing. System response was captured by pre/post-test surveys as well as 116 instruments. The pipeline accommodated 24 in. (600 mm) of relative fault movement, corresponding to 15.2 in. (390 mm) of axial compression, without leakage/rupture nor significant loss of cross-sectional conveyance capacity.

Estimating the seismic bearing capacity of soil under shallow foundation is important for seismically active areas. This paper uses modified pseudo-dynamic method to estimate the dimensionless seismic bearing capacity factor N_γe for dry sand underlying a surface strip footing. Soil amplification is considered inherently in this method and seismic acceleration is found to have significant effect on N_γe. The present study shows that the value of N_γe considering k_h=0.1, k_v=0, decreases by 86% as compared to the static condition for ϕ =30^o. The vertical seismic acceleration coefficient also significantly affects the seismic bearing capacity factor. It was found that N_γe decreased by 84% when k_h increased from 0.1 to 0.2 for k_v=0 and it decreased by 95% when k_v=k_h for ϕ=30^o.With decrease in ϕ from 40^o to 30^o, N_γe decreased by 68% for k_h= k_v=0.1. The proposed method can be used for seismic design of shallow strip foundations in sand.

This paper presents a series of large scale liquefaction-induced lateral spreading shake table tests conducted at the University of California, San Diego (UCSD). A total of 7 different experiments are illustrated with varying heights, inclinations, soil profiles, pile material, and cross-sections. The available capabilities discussed are part of overall large-scale UCSD laboratories, one of the largest and most active earthquake engineering laboratories worldwide. Specifically, testing using 2 different shake tables is discussed, one being the largest outdoor shake table in the world. The testing models were densely instrumented with top of the line sensor arrays. Instrumentation includes strain gauges, pressure and excess pore-pressure sensors, accelerometers, and displacement pots. The method of construction is presented, and the stages clearly described. A sample of results is presented and conclusions are drawn.

Deformations and displacements caused by applied service loads and extreme events such as earthquake can have a detrimental impact on the accessibility of rail tracks and increase the maintenance cost of these infrastructure elements. For the ballasted tracks, which are the most common form of railway construction, these deformations can be reduced by using alternative aggregates in combination with ballast and/or sub-ballast particles. This paper discusses the application of tire derived aggregates (TDA), which are recycled from scrap tires, and lightweight expanded clay aggregates (LECA), which are produced in rotary kilns, in the sub-ballast layer. Application of these alternative aggregates has potential to reduce deformations and displacements of the embankment caused by seismic loads. Further, applications of TDA as waste materials, and LECA as sustainable products contribute to cleaner environment. The presented experimental investigation consists of a series of shake table tests involving three railway embankments, each supported on a different sub-ballast layer, including dense soil, dense soil with TDA, and LECA with TDA. Results indicated that using TDA and LECA in the sub-ballast layer reduces the total displacement and the acceleration of railway embankments, and thus improves the seismic performance of these structures.

Integral abutment bridges (IAB) consist of a continuous concrete deck integrated with abutments generally supported by a row of piles and they are cost-effective with less maintenance efforts. However, the abutment walls of the these bridges are subject to significant seismic loads during strong seismic shaking as opposed to the bridges with joints separating deck and abutments and there are limited guidelines available in the literature to evaluate the seismic loads on the abutments walls. The paper presents the results of the dynamic soil-structure interaction (SSI) analysis carried out for a typical two-span bridge with integral abutments using the computer program FLAC^2D. The bridge structure has been represented by a series of beam and pile elements that interact with the continuum medium through normal and shear soil springs. The slippage between abutment and soil has been accounted for by introducing interface elements in the model. The piles supporting the abutment walls are driven on to the bedrock and the middle pier is support by rock sockets extended into the bedrock. The bedrock is classified as “Class A” in accordance with the 2014 Canadian Highway Bridge Design Code (CHBDC). The seismic loads computed from the SSI analyses were compared with those estimated from the simplified analyses available from the literature.

The Mexico City Valley is located on very thick soft clay deposits that amplify the seismic vibrations. This amplification generates displacements at the soil surface larger than those register at the clay-bedrock interface. This paper discusses the problems associated with pipe failure during earthquakes in these conditions. Several buried pipelines conducting potable water failed during the earthquakes of September 19 of 1985 and 2017 in Mexico City, due to large soil displacements produced by the seismic wave propagation effects. The mechanisms of these pipe failures and methodologies to analyze the forces acting along the soil-pipeline interface are described, focusing on the response of the continuous and segmented pipelines to seismic wave-induced soil movements. Preventive and emergency measures to deal with such seismic effects are given. Finally conclusions and recommendations related the vulnerability of lifeline systems to earthquakes on this type of soil conditions are put forward.

This paper focuses on the seismic response of gravity walls with small to medium cohesive (i.e., dirty sand) backfills and compares the results with that of the cantilever walls using finite difference method. The verification of the analysis procedure was conducted using experimental work obtained from literature. The intent of the analysis is to investigate the effects of backfill cohesion up to 30 kPa on seismic earth pressures, the location of point of action, the coefficient of total seismic earth thrust, and the coefficient of incremental seismic earth thrust during the shaking. Two different earthquake (i.e., Loma Prieta and Kobe) ground motions are used in the analysis. The obtained results show that the coefficients of total and incremental seismic earth pressure are higher for Loma Prieta than Kobe in both types of walls. The geometry of the wall influences the earth pressures at upper portions of the wall.

This paper discusses the results of dynamic centrifuge tests performed on 9-story model structures with different yield capacities founded on a layered liquefiable deposit with and without a silt cap. The goal was to evaluate the interacting role of soil interlayering and structure’s yield capacity on the response of soil-foundation-structure systems. The first test evaluated the seismic performance of a 9-story, steel, moment-resisting frame structure (B₁) on a layered liquefiable deposit without a silt cap. The second test simulated simultaneously two models of a 9-story structure (B₁ and B₂) on a similar soil profile, but with a silt cap. Structure B₂ was designed to have a greater yield capacity compared to B₁, with other dynamic properties kept constant. The presence of a silt cap and resulting void redistribution did not notably change Structure B₁’s average settlement, but it amplified residual rotations, foundation, and roof accelerations, and deformation demands on the superstructure. The stronger Structure B₂ also experienced large rotations but less than B₁, because of smaller flexural deflections and P-Δ moments imposed about its columns and foundation. These results point to the importance of characterizing soil interlayering as well as structure’s strength and other dynamic properties when evaluating the consequences of liquefaction.

The Port of Wellington experienced liquefaction of reclaimed land in the 14 November 2016 M_w7.8 Kaikoura earthquake. The liquefaction caused global settlement of the sandy gravel and sandy fill deposits and lateral movement of the fills towards the sea. The reclamations at the port are comprised of different soil types (gravels and sands), different ages (40 to 100 years), and different methods of construction (end-tipped and hydraulic fills). Cone penetration tests were advanced successfully in gravelly fills using robust equipment. When combined with conventional soil exploratory borings, these data characterized well the liquefaction hazard at the port. The 30% to 60% of sand and silt in the well-graded gravelly fill largely determined its liquefaction resistance and seismic performance, suggesting the soil matrix is dominated by these finer particles. State-of-the-practice CPT-based liquefaction triggering and post-liquefaction reconsolidation settlement assessments identified loose zones of fill that likely liquefied, which produced widespread ejecta and caused damaging vertical and lateral ground movements. CPT-based procedures estimated well the vertical settlement of the gravel fills and the sand fills. Recommendations are made on how to best address the site investigation challenges at the gravelly soil areas of the port.

The present paper investigates the effect of soil-structure interaction (SSI) of the dynamic response of FRP confined reinforced concrete bridge column supported by isolated spread footings on loose Ganga sand. The FRP confined reinforced concrete bridge column with SSI base is efficiently modelled in finite element framework of Opensees by using the FRPconfinedConcretematerial model for the confined concrete and BNWF model for soil-foundation interface. It is observed that SSI significantly increases the peak base shear of the FRP confined bridge column by 30.5% but reduces the peak column displacement by 15.3%. The average energy dissipation due to foundation flexibility is estimated up to about 14.26%.

A seismic retrofit design was carried out for an existing 110-year old railway bridge crossing a wide river. The existing foundations that have performed acceptably under static conditions, are founded on potentially liquefiable soils. A retrofit strategy was developed using large diameter driven pipe piles to transfer the bridge loads to non-liquefiable strata. The retrofit design is required to withstand a 475-year return period earthquake. A suite of 15 spectrally matched ground motions were developed based on the contribution to the uniform hazard response spectra (UHRS). The equivalent linear ground response analysis and simplified liquefaction assessment showed that the sand layer liquefies during the earthquake and the foundation soils undergo lateral spreading. The preliminary design of the retrofit piles was carried out based on the soil-pile interaction to lateral loads using a beam configuration with nonlinear lateral load transfer curves in the software LPile. Nonlinear seismic deformation analyses were performed using the finite difference program FLAC and effective stress based constitutive soil model UBCSAND during the detailed design. The extent of liquefaction estimated using the simplified liquefaction assessment was validated with the FLAC deformation analysis. Based on the FLAC simulation, the overall riverbed flattens during the seismic event, and the retrofit pile heads undergo a horizontal displacement of up to 0.5 m. The results of the FLAC analysis verified the preliminary LPile predictions.

Seismic behavior of concrete faced rockfill dams (CFRDs) has not been sufficiently examined due to limited measured response in the field. This study presents results of numerical experiments (finite element analyses), which investigate effects of hysteretic damping and interface between rockfill and concrete face on the seismic behavior of CFRDs. The analyses use a recently developed, three-dimensional, distributed element plasticity soil model (I-soil) which allows flexible control over hysteretic behavior and can model both Masing and user-defined non-Masing type un/reloading. Numerical results are compared with centrifuge test data performed by others, including spectral response, and bending moment increments on the concrete face. The best estimates were obtained using non-Masing type un/reloading rules along with frictional interaction between concrete face and rockfill interface. These results suggest that the seismic response of CFRDs can be better simulated with proper characterization of hysteretic damping and interaction between different components of the dam body.

Significant improvement has been made with the advent of numerical tools and methods in the predicting capabilities of behavior and building performance when an earthquake strikes. Current building designs tend to produce safer structures, being resiliency the most important consideration to minimize indirect losses due to strong earthquakes. This resiliency is quantified in terms of recovery time; the time necessary to achieve a safe operational level for a structure after an earthquake. Performance based earthquake engineering is a probabilistic assessment of building risk and damage frequently used to quantify earthquake-induced losses of structural and non-structural elements. The main goal of this paper is to estimate the intensity measures and engineering demand parameters for a hypothetical tall building located in an urban environment, downtown Los Angeles, including the influence of foundation soils in the building response. Numerical simulations of a hypothetical tall building supported in a mat foundation are performed using the pressure-dependent multi-yield surface constitutive soil model coded in OpenSees. The building is subjected to three earthquakes levels. Based on the building responses and computed peak story horizontal accelerations, maximum inter-story drifts, and settlements, conclusions are drawn about the influence of soil-structure interaction in the response of tall buildings.

A shallow foundation, 3 m×3 m in plan and embedded 0.9 m, was dynamically loaded at the national geotechnical experiment site (NGES) on the campus of Texas A&M University (TAMU) via a rotating mass vibrator and the measured results were compared with results computed from cone models of the soil/foundation system. The measured responses of the footing indicated that a rigid body motion model for movement of the footing was appropriate. This rigid body model based upon a two-degree-of-freedom (2DOF), mass-spring-dashpot concept, employing dynamic impedance functions computed via a cone model, and using free-field shear wave velocity values measured at the site with surface wave tests, provided a good assessment of the dynamic response of the foundation under horizontal excitation.

To accurately describe the dynamic characteristics of bridges, it is important in some instances to take into consideration the flexibility and damping of the soil-foundation system. The ability to evaluate those properties in the field can serve as both a check for the design assumptions, and as assistance in the design of bridges with similar superstructure/substructure loading and soil conditions in the future. The goal of the presented study is to demonstrate the use of large-amplitude shaking as an effective tool in measuring actual response/behavior of bridges, and developing better understanding of the dynamic response of bridge systems. For that purpose, a large-amplitude shaking of a bridge in Hamilton Township, New Jersey, was carried out. The T-Rex, a mobile shaker from the Natural Hazards Engineering Research Infrastructure (NHERI) experimental facility at the University of Texas, Austin, was employed to shake the bridge. A large number of sensors, geophones, and accelerometers, were installed at various locations on the bridge deck, pier cap, and on the adjacent ground to capture the dynamic response of the bridge system. Furthermore, the results from field testing were used to calibrate a 3D finite element model of the bridge. The model was used to conduct a comparative analysis of the bridge response for the assumption of the bridge with fixed foundation conditions, and the bridge with the consideration of dynamic soil-structure interaction (DSSI) effects. The comparison with the field testing results demonstrate that the fixed foundation assumption model does not fully capture the behavior of the bridge, as opposed to the model with DSSI considerations.

A multi-phase liquefaction mitigation program in support of a vertical expansion of the Newby Island Sanitary Landfill (NISL) on the southern margins of the San Francisco Bay has been ongoing since 2012. The landfill, including the site perimeter levee, is founded in soft, estuarine soils, and the proposed waste fill thickness of 87 m (285 ft) is relatively high, as is the M7.1 design earthquake peak horizontal ground acceleration (0.41 g). Site-specific evaluations showed that the stability of the landfill-subgrade system is governed by a series of potentially liquefiable sand lenses underlying the site. In situ cement deep soil mixing (CDSM) through the levee was constructed to mitigate the expected excessive seismic displacements in underlying liquefiable layers. To date, design and construction of three mitigation phases have been completed along 1,000 m (3,280 ft) of levee. Slope stability and seismic deformation evaluations for mitigation design were performed using a state-of-practice de-coupled approach based on limit equilibrium and Newmark principles. This paper presents results of preliminary dynamic numerical analyses of the NISL mitigation using the two-dimensional finite difference-based computer program FLAC. Preliminary deformation results from the FLAC model were comparable to those from the conventional de-coupled approach, providing a reasonable degree of confidence in the numerical results. Given the versatility and more realistic physical behavior of the FLAC model relative to the de-coupled approach, this numerical approach may be considered for optimization of future NISL liquefaction mitigation phases or similar projects.

A three-dimensional (3D) finite element (FE) model has been established using the commercial program ABAQUS to simulate the performance of soil nailing systems in seismic conditions. Comparisons with experimental data have confirmed that the developed FE model can simulate the seismic behavior of a soil nail system properly. Based on this, the verified FE model has been used to study the effect of different nail orientations, slope angles, and backslope angles on the seismic deformations and failure mechanism of the soil nailing systems. It was found that with an increase in nail orientation, horizontal displacement of the soil nailed system decreased almost in all of the models. However, there was an optimum value for nail inclination after which horizontal displacements started to increase or stay constant with the increase in nail inclination. As backslope angle increased, the seismic resistance of soil nailing systems became more unstable. A comparison between the results for slope angles of 75° and 90° showed that a steep nailed slope had less seismic resistance. Besides, for a slope angle of 90° the failure mechanism was a combination of translational and rotational deformations while models with slope angle equal to 75° exhibited predominantly translational failures.

The results of seismic pile-soil-structure interaction (SSI) analysis of a large diameter tank supported on piles driven to a firm stratum are presented. The piles penetrate a sand layer susceptible to liquefaction. The layer was improved by driving steel inclusions into the ground. The liquefiable layer extended outside the footprint of the tank. SSI analyses were performed to demonstrate that liquefaction treatment was not required beyond the tank foundation footprint. Site response analysis was performed using SHAKE91 and the SSI analysis was performed using MTR/SASSI. Soil nonlinearities were accounted for using an equivalent linear model. The tank and fluid content were modeled using a lumped-mass stick model. The concrete base slab and insulation space between the steel and concrete base slabs were modeled using flat shell (plate) and spring elements, respectively. The piles were modeled using linear beam finite elements connected to soil at all pile nodes. The improved soil and the liquefaction zones were explicitly modeled using solid elements. For the liquefaction zone, the solid elements were assigned a degraded shear wave velocity value based on CPT data and residual strength relationships for sand. In addition to liquefaction effects the assumption that the bottom of the foundation slab is fully bonded to the soil was studied. It was found that the pile stresses for the unbonded case could be 50 percent higher than the bonded case.

Coastal structures are prone to multi-hazard loading; for instance, earthquake ground shaking, and subsequent tsunami loading. The instability of seabed sand caused by residual and momentary liquefaction, as well as large bed shear stresses, are primary concerns for engineers. The foregoing soil response is related to the soil’s pore water pressure response during and after earthquake shaking, i.e., before the tsunami attack. Our objective is to develop and validate a one-dimensional numerical model in OpenSees to simulate the excess pore water generation and dissipation during and immediately after earthquake loading. The model consists of four node elements and the pressure dependent multi-yield material model, and we validate the model through comparisons with LEAP-GWU-2015 centrifuge test data. In addition, we perform a convergence study to examine sensitivity to mesh size, time step, and convergence criteria. The results show that the numerical model captures the magnitude of the excess pore water pressure and its long-term asymptotic dissipation response well, and the permeability has a large impact on the excess pore water pressure response, as expected. Our work gives reference parameters for the Ottawa-F65 sand used in the LEAP-GWU-2015 centrifuge tests. The results indicate that our model can be adopted for future research on the liquefaction potential for coastal structures during complex, multi-hazard loading scenarios.

Many landslide masses are inhomogeneous and complicated. To a limited level, the failure mechanism, such as the initiation time and threshold displacement, derived from classical constitutive models and numerical methods could be further discussed to ascertain its applicability. In this paper, a validated dynamic finite element procedure is employed to analyze a coseismic landslide subjected to earthquake loading. The sliding mass and slope are expressed using quadrilateral elements of elastoplastic behavior that obey Mohr-Coulomb criteria with associative and nonassociative plastic flow rules. Based on the results, the flow rule does not have a major influence on discriminating the initiation time but have significant impact on the coseismic deformations during ground motion. Our analysis provides preliminary investigation focusing on the earthquake-induced landslide considering different plastic flow rules.

Microbially-induced calcite precipitation (MICP) has been increasingly investigated as a novel ground improvement technique for the mitigation of earthquake-induced soil liquefaction. Experimental results including those from triaxial, direct simple shear, and small centrifuge tests have demonstrated the ability of MICP to increase the resistance of loose sands to liquefaction triggering and improve the post-triggering response of these materials when compared to similar untreated soils. Despite significant advances in the development of this technology for field-scale applications, currently no quantitative frameworks exist, which can accurately predict the behavior of cemented sands in the field. In this paper, we present an evaluation of the ability of an existing stress-ratio controlled, critical state compatible, bounding surface plasticity model (PM4Sand) to capture the response of (bio-) cemented sands. The results of this study will guide the reformulation of the existing constitutive model to capture experimentally observed soil behaviors.

Hysteresis models following the extended Masing rules are commonly used to represent soil’s un/reloading behavior. However, at moderate to large strain levels of shaking, models based on the extended Masing rules overestimate hysteretic damping measured in laboratory tests. This study evaluates the impact of hysteretic damping at moderate to large strain levels on simulation of dynamic soil-underground structure-superstructure interaction. The simulations use a recently developed, three-dimensional, distributed element plasticity soil model (I-soil), which allows flexible control over hysteretic behavior and can model both extended Masing and user-defined non-Masing type un/reloading. Three-dimensional finite element simulations are compared with results obtained from centrifuge experiments on medium-dense, dry sand in terms of acceleration, surface settlement, and wall deformations. Non-Masing unloading/reloading rules provide a better estimation of spectral accelerations at intermediate period ranges and surface settlements. Both cases computed similar surface spectral response at short and long periods as well as wall deformations.

Centrifuge experiments and parallel nonlinear numerical simulations were conducted to investigate the seismic response of potentially inelastic, multi-degree-of-freedom structures on layered, liquefiable soil deposits and the effectiveness of mitigation with prefabricated vertical drains (PVDs). Fully-coupled, 3D, nonlinear numerical simulations of the soil-foundation-structure system in centrifuge were performed in the finite element platform OpenSEES. Experimental results showed that PVDs can effectively speed up pore pressure dissipation after shaking and reduce building’s settlement and tilt potential. PVD were, however, shown to increase roof accelerations, displacements, and drift ratios, which need to be considered in design. Although the numerical results showed some discrepancies in capturing the dissipation rate of excess pore pressure and foundation’s residual tilt, they generally captured its settlement, particularly for thinner liquefiable layers. Future improvements are needed in the numerical models to better capture the rate of dissipation, volumetric strains due to sedimentation, and the localized deformations responsible for foundation’s cumulative rotation over time.

Fine coal refuse (FCR) is the waste produced in mining process. FCR can be classified as low-plasticity to non-plastic sandy silt or silty sand. Although FCR consists of appreciable amount of fines content (30% to 60%), there is a high liquefaction potential for hydraulically deposited FCR in impoundments due to its loose and saturated structure as well as its low permeability. Therefore, dynamic properties, particularly liquefaction resistance, of FCR should be investigated. In this study, cyclic direct simple shear (DSS) tests were performed on reconstituted samples of FCR to determine its dynamic properties and liquefaction resistance. Slurry deposition method, which mimics the fabric and structure of hydraulically deposited soils, was used to prepare representative samples of FCR. In addition, the results of cyclic DSS tests were used to calibrate plasticity constitutive models developed to simulate liquefiable materials. PM4Sand and PM4Silt are plasticity models that are developed to obtain monotonic and cyclic undrained shear behavior of liquefiable materials, for sands and low plasticity silts or clays, respectively. It was of interest to find which model better represents the dynamic characteristics of FCR since the material is composed of both silt and sand and exhibited behavior that could be characterized as both liquefaction and cyclic-softening. Therefore, the PM4Sand and PM4Silt constitutive models were calibrated for FCR using the DSS experimental results in FLAC, a 2-dimensional finite-difference program. The simulation results were compared against the experimental results. This research revealed the applicability as well as limitations of the two constitutive liquefaction models in simulating the cyclic shear response of fine coal refuse.

Storage of large-diameter metallic casks containing spent fuel required the construction of concrete pads at an existing nuclear power generating facility. The subsurface investigation which included: piezocone and seismic piezocone penetration testing (CPTu, SCPTu), standard penetration testing (SPT), intact thin-walled tube sampling and laboratory testing, encountered zones of weak/compressible (soft) and/or liquefiable soils at depths greater than about 55 feet (16.8 m). Understanding the site geology, particularly the depositional environment, was a key factor in the overall assessment A 2.3-foot (0.7-meter) SPT rod drop in one borehole indicated a possible void. A seismic event at the site could trigger settlements at depths below 55 feet (16.8 m) that could propagate to the ground surface. Seismic settlements at depth were calculated for two different mechanisms: (a) compression of weak, soft materials, and (b) liquefaction-induced settlements. Settlements at depth, including a postulated collapse of a void, were propagated to the ground surface using an empirical soft-ground tunneling approach and a numerical model using Plaxis 2D. The results showed that both the empirical and numerical models predicted similar settlements at the ground surface, which were significantly smaller than those at depth, and within acceptable limits for the concrete pads.

There are a significant number of studies into the failure of pile-supported structures exposed to earthquakes, however, there remain a difficulties with the in situ examination of pile response and performance during seismic excitation. A flexible wall barrel, shaking table test method is suitable for investigating pile behavior during an earthquake. Cost, time, and difficulties in identifying soil properties accurately in physical models, in addition to the effects of test conditions, have led to the current research where the physical test is replaced by numerical simulation. Many researchers have experienced difficulties in the validation of numerical models and have found that there is a lack of available information in the literature. Thus, developing a practical approach will extend the soil-structure interaction (SSI) database and promote the validation opportunities for studies into pile performance during strong excitations. This study provides an insight into a set of SSI problems and proposes a procedure for calibration of the advanced SSI analysis. A framework is performed to simulate a shaking table test of a model pile-foundation superstructure on soft clay. A variety of model scaling relationships are used to develop an approach that allows observation of the inherent dynamic and non-linear nature of SSI behavior. The three-dimensional, non-linear dynamic response and elastoplastic analysis are included in the simulation. Through the development of finite element analysis (FEA) using ABAQUS software, fully non-linear unidirectional input excitations, which are amplified from the base to the top and are capable of including all of the possible degrees of freedom, are applied to the model. The inertial, kinematic, and damping interaction components of the response are also examined. The gap-slap mechanism between soil and pile is a significant aspect to the model. The results are validated using physical test results.

Subsurface fluid injection into sedimentary reservoirs is potentially responsible for the sharply rising number of seismic events in oil and gas production regions. To assess the effects of fluid injection activities on fault reactivation and slip, we construct a hydro-mechanical joint-enriched finite element model, comprised of a multi-layer sequence embedded with a favorably-oriented normal fault. We consider six injection scenarios to address the effects of anthropogenic parameters including injection rate and injection volume on the stability of seismogenic fault. Simulation results demonstrate that maximum excess pore pressure and stress perturbation are highly dependent on the injection strategy. The injection program with faster initial rate may induce early fault reactivation. Higher injection volume also could result in a higher potential of fault reactivation and slip. The incorporation of joint elements into the model allows the capture of progressive fault damage under hydro-mechanical couplings. This study highlights the importance of appropriately-designed injection well operations in minimizing the likelihood of induced seismic events.

Dam-foundation interaction under seismic loading remains a challenging topic. Many factors affect the dynamic response of a dam system during an earthquake. Two governing parameters include type of the foundation rock and layering of the supporting geological media. Finite element methods have been proven to be effective techniques for evaluating dam-foundation interactions. Previous finite element based studies have assessed the dynamic response of a given concrete gravity dam located on a rigid or a monolith viscoelastic half foundation. The current study presents a new finite element based numerical model developed to assess the dynamic response of an earth dam system supported by a multi-layered rock foundation during an earthquake. The Pine Flat gravity dam was adopted as a case study, and the Taft earthquake data (1952) was utilized to replicate the ground motion. The conventional Mohr-Coulomb plasticity model has been adopted for characterization of the constitutive response of the layered rock foundation system. Several case scenarios have been simulated to assess effects of rock type and layering on the dynamic response of the dam system. Various nodes have been selected on the dam crest as well as within the geological medium, in which displacements, stresses, deformations, and potential energy have been physically monitored. The most critical case scenario is identified based on which design recommendations are made.

Previous research shows that shaking history affects the liquefaction resistance of the soil. A strong shaking event could reduce the soil liquefaction resistance due to increased anisotropy of soil. However, a mild shaking with partial or no liquefaction could greatly increase the liquefaction resistance. In this paper, the effect of shaking history on liquefaction resistance of sand is studied using 1-g shake table testing. A uniform liquefiable sand deposit was air-pluviated and fully saturated in a large laminar shear box (L×W×H: 2.29 m×2.13 m×1.83 m). The sand deposit was subjected to four different shaking events. Piezometers were embedded at different depths to capture the seismic response of sand deposit. The excess pore pressure readings from four shaking events were compared to investigate the effect of shaking history on liquefaction resistance of the sand deposit. The major findings of this paper are: (1) along with the shaking, the sand deposit became denser; (2) the shaking event with an amplitude of 0.25 g tended to densify the sand and increase the liquefaction resistance; and (3) the strong shaking with an amplitude of 0.4 g tended to damage the soil structure and reduce the liquefaction resistance.

This study assesses the effects of spatial variability on liquefaction-induced settlements through a series of medium-size 1-g shake table experiments. Past research including stochastic numerical modeling and evaluation of field case histories provided some insight on the effects of spatial variability on liquefaction-induced settlements; however, there is lack of experimental data in the literature focusing on this topic. Therefore, the need for quantifying the effects of non-uniform liquefiable soil layers on liquefaction-induced differential settlements seems inevitable. A series of 1-g medium-size shake table experiments with different liquefiable soil layering configurations were conducted to evaluate the effects of non-uniform liquefiable soil layer on free-field and foundation settlements. According to the results obtained, the pattern of the free-field settlement illustrated direct relation to the spatial variability of liquefiable layer. Furthermore, this research explored the correlations between the extent of liquefaction-induced settlements and the volume of observed sediment ejecta. Finally, this study provides preliminary observations on the impact of sediment ejecta on liquefaction-induced settlement of foundations.

Both field and laboratory studies have shown that loose saturated sands containing nonplastic fines are highly susceptible to liquefaction. However, current liquefaction countermeasures for these soils are restricted due to their low permeability, environmental effects, and high costs. Microbial induced partial saturation (MIPS) via denitrification is a new emerging soil improvement technique that utilizes biochemical activities to generate biogenic gas and increase the liquefaction resistance of fully saturated soils. MIPS can offer a promising, quick, sustainable, and cost-effective method to effectively improve liquefaction resistance of silty sands. In this study, a direct simple shear (DSS) apparatus was used to experimentally investigate the response of induced partially saturated silty sands to cyclic loads. Water pressure sensor and (LVDT) were used to monitor excess pore water generation and deformation of treated specimens under strain-controlled cyclic tests. Results revealed that MIPS treated samples had significantly higher liquefaction resistance compared to untreated samples. Both generation of excess pore pressure and vertical deformation were considerably reduced as the degree of saturation decreased. From this study, it can be concluded that MIPS may potentially introduce a more effective technique than other techniques such as grouting and microbial induced carbonate precipitation where the permeability of soil limits the effectiveness.

Offshore soft clays can potentially undergo a loss in shear resistance when subjected to “non-standard” cyclic events such as storms and earthquakes. This behavior, termed cyclic softening, or cyclic degradation, is caused by the cyclic shear-induced generation of excess porewater pressure and a progressive de-structuration of the clay fabric. To better understand whether the reconstitution process affects the cyclic response of soft clays, six strain-controlled cyclic undrained triaxial tests were performed on a reconstituted Gulf of Mexico clay and compared to six similar tests performed on intact specimens loaded beyond the preconsolidation pressure. The single-amplitude unidirectional cyclic shear strain and overconsolidation ratios (OCRs) investigated were 1% and 2% and 1, 2, and 3 respectively. All cyclic strains were applied at a frequency of 0.5 Hz. The tests illustrated that the normalized cyclic shear stress response was nearly identical for the reconstituted and intact specimens. Reconstitution did not affect the undrained shear modulus during the first cycle normalized by the preconsolidation pressure. However, the reconstituted specimens generated excess porewater pressure more slowly possibly due to the absence of silt seams, that would otherwise aid in porewater pressure equalization. The reconstituted specimens exhibited a slower decrease in secant shear modulus with cycling; the values of modulus degradation parameter, t, of the reconstituted specimens were on average 25% smaller than those of the intact specimens. However, the authors conclude that alternatives to “t” may better capture shear modulus degradation.

An examination of dynamic responses of clean sand under high effective confining pressure (σ_v^′) is conducted with two centrifuge tests using laminar container. The centrifuge experiments simulate a 5 m saturated clean sand layer under high effective overburden pressure of 6 atm with the relative density of 40%. To achieve the targeted overburden pressure, a dry layer of lead shot with an appropriate thickness was deposited on the top of the clean sand. Viscous fluid was used for saturation to keep a constant prototype permeability. The centrifuge tests were subjected to 10 cycles sinusoidal seismic motions with a different prototype peak acceleration to study the effects of pre-shaking history. Acceleration, pore pressure build-up and dissipation, and shear wave velocity were recorded during and after shaking. The recorded accelerations were further analyzed using an established system identification (SI) technique to evaluate the cyclic stresses and strains induced in the soil deposit. Furthermore, the effects of pre-shaking history were analyzed in the aspects of excess pore pressure build-up, shear wave velocity changes, and soil deposit densification.

Saturated sand lenses are usually encountered inside saturated clay deposits. During an earthquake the sand lenses could liquefy. After liquefaction, sand volcanoes appear on the surface of the clay deposits. To date, there is not a clear understanding of the mechanisms involved with the movement of the liquefied sand lenses. The liquefaction of the sand lenses is the result of shear stresses induced in the sand by seismic shear (S) waves. Another seismic wave that has been overlooked in the movement of liquefied sand lenses is the compressive (P) wave. This study presents a laboratory experiment that uses a shaking table that vibrates horizontally a saturated clay mass containing a simulated liquefied sand lense. The experiments indicated that the liquefied sand lense moved as a result of the compressive stress induced in the clay by the P waves and the stress concentration at the top of the sand lense.

The performance of silty soils in Emilia Romagna, Italy, during the 2012 earthquake offers an interesting opportunity for understanding their response under cyclic loading and susceptibility to liquefaction. A comprehensive investigation based on both in situ tests and laboratory tests (undrained cyclic simple shear tests) was performed for evaluating the cyclic strength of these low plasticity silty soils and thus analyzing data in light of the screening criteria for liquefaction triggering susceptibility. The laboratory results indicate that the soils are susceptible to liquefaction, whereas some types of interpretation approaches based on in situ tests appear to overestimate the liquefaction resistance of soils when applied to sands containing non plastic or low plasticity fines, thus suggesting a revision of the existing methods of analysis.

Accurate estimation of dynamic pore water pressure of submerged soil is needed for seismic design of quay walls. Studies have shown complexities involved in estimation of dynamic pore water pressure which are mainly caused by soil/water interactions and excess pore water pressure. In this study, a series of 1-g shaking table tests was conducted to evaluate factors affecting dynamic pore water pressure of submerged soil. Comparison of the pore water pressure estimated using existing solutions with those measured in the model tests indicates that existing solutions fail to accurately predict dynamic pore water pressure for cases where the permeability of backfill soil is in medium range (between 10^-4 to 1 cm/s). The results indicate that the discrepancies between the measured and estimated values are likely caused by soil/water interaction, dilative/contractive behavior of backfill soil, and permeability of backfill soil, factors which are not considered in the existing solutions.

Direct simple shear (DSS) is one of the most widely used tests for investigating monotonic and cyclic properties of soil. Undrained condition in DSS is produced by keeping the specimen volume constant during shearing. Since the specimen is radially constrained, constant volume is achieved by maintaining its height constant. ASTM D6528 standard test method recommends limiting the vertical strain to ±0.05% for monotonic tests using passive or active height control. In this study, the performance of a newly developed active height control system is examined. The system maintains the height by almost simultaneously reacting to the readings of the vertical displacement sensor. As a result, it is capable of maintaining the vertical strains well within ±0.03% even when the excess pore pressure ratio is approaching 1 or the specimen is experiencing large shear strains (i.e. as high as 10%). Furthermore, the effects of vertical strain generated during cyclic and monotonic DSS tests due to inadequate constant volume are discussed. It is shown that the number of cycles required for liquefaction and monotonic shear strength can be significantly affected if the vertical strain is not kept at minimum.

This paper presents the influence of plasticity on liquefaction characteristics of a selected fine-grained soil. The effect of plasticity was evaluated in terms of pore water pressure generation and deformation characteristics of the fine-grained soils with cyclic loading. Stress-controlled cyclic triaxial tests were performed for this purpose. Test samples were prepared at comparable dry densities with varying plasticity index (PI). Sil-Co-Sil #40, a non-plastic commercial silt and EPK Kaolin clay were used as base materials. These materials were mixed in different proportions to obtain desired PI values. The PIs of the tested samples were varied over reasonably wide range [non-plastic (NP), 2.7, 3.5, 5.7, 9.8, and 14.5]. The samples were tested at an initial effective confining pressure of 34.5 kPa (5.0 psi) using a cyclic stress ratio (CSR) of 0.2, 0.3, and 0.4. The results indicate that with an increase in PI of the soil samples tested from NP to 14.5, liquefaction resistance of the soil in terms of deformation decreases initially and then increases with an increase of PI. Similar effect of plasticity on the liquefaction characteristics of soil was observed while evaluating in terms of pore pressure generation. The tests also revealed that, the rate of pore pressure generation first increases with an increase of PI from non-plastic to 3.5 and then decreases with increase of PI beyond 3.5.

Instability and static liquefaction of medium dense silty-sand, which has experienced liquefaction during the devastating 2001 Bhuj, India, earthquake has been explored in the current research. Unlike previous studies, where static liquefaction has been reported to be prevalent in very loose sand as well as in loose silty-sand specimens, the findings in the present article reveal its occurrence in medium-dense silty-sand. Interestingly, the soil exhibited static liquefaction even at large confining pressures amounting to 200 kPa and 300 kPa (besides 100 kPa) indicating its susceptibility to complete loss of strength at greater depths. Experimentally observed mechanical behavior has been compared with the response predicted by using a non-associative state dependent elasto-plastic material model that can capture strain softening behavior in granular media. Vanishing of second order work indicated the onset of instability and evolution of hardening modulus captured subsequent liquefaction in the experimental results as well as model predictions. Specimens exhibited runaway instability when undrained boundary condition was imposed at stress states lying in the region of potential instability. Microstructural instability (liquefaction) within the soil mass has also been examined using disturbed state concept (DSC).

The paper presents a series of centrifuge tests simulating a level site consisting of granular soil deposits subjected to various bi-axial and uniaxial base excitations. Synthetic sinusoidal waves were used as base excitations to test loose models under biaxial and uniaxial shaking. Dense arrays of accelerometers were used to monitor the deposit response along with pore water pressure transducers. Two uniaxial tests and one bi-axial shaking test were conducted on three similar soil models to study the impact of multidirectional shaking on the generation of soil liquefaction. The two uniaxial shaking tests consisted of: i) test with input energy content using the arias intensity is similar to that of the bi-axial input shaking; and ii) test with 10% increase in one of the components of the biaxial shake amplitude, as commonly done in practice for uniaxial simulation of multidirectional field shaking. 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 stress and strain histories are used for two purposes. First, to shed the light on the properties of liquefaction occurring through the stratum subjected to uniaxial shaking. Second, to show the difference in soil behavior when it is subjected to biaxial shaking.

Adelaide Industrial (AI) sand has been used extensively for land development and as a construction material in Adelaide, Australia. However, the mechanical properties of AI sand are not yet well understood. In this study, a series of undrained monotonic triaxial compression tests were conducted on AI sand to characterize its stress-strain behaviour density states. The test conditions included a range of void ratios, e₀, from 1.09 to 0.73 and a range of mean effective confining stress, $p_0^{\text{'}}$, from 50 to 400 kPa. It was observed that the effective stress path in the undrained condition exhibited flow and limited flow behaviour with the triggering of strain softening at the peak deviatoric stress, q. The onset of strain softening is the initiation of instability or static liquefaction, which is characterized by the stress ratio, η_IS= q/$p^{\text{'}}.$ It was observed that the stress-strain behaviour of AI sand under undrained conditions, particularly η_IS, was influenced by both e₀ and $p_0^{\text{'}}.$

During earthquakes, many structures fail due to the liquefaction of saturated sand or silt. Installation of the stone column is one of the most adopted ground improvement methods used to minimize the damage caused by soil liquefaction. During the installation of the stone column, the permeability of the soil at the vicinity of the column may reduce due to densification. The available analytical solutions to determine the performance of stone columns are developed based on the infinite permeability of the stone columns and without considering the stiffness of columns. Thus, in this paper, an approach has been presented to evaluate the generation and dissipation of excess pore water pressure considering smear effect, well resistance, and stiffness of stone column. The effects of various important parameters of smear zone and stone column on the mitigation of soil liquefaction are evaluated and demonstrated in the results. Obtained results demonstrate that the reduction of the permeability of smear zone increases the pore pressure ratio by around 20% and reduction of the permeability of stone column further increases the peak value of excess pore pressure ratio by 18% to 71%. It is also observed that due to increase in compressibility ratio, the peak value of excess pore pressure ratio is decreased.

Seismic hazard in Mexico City governed by site effects. The M8.1 1985 subduction zone earthquake, which caused significant damage and loss of thousands of lives at 350 km epicentral distance, has become the quintessential example of the role that site effects can play in modifying the amplitude, frequency, and duration of ground shaking; and in aggravating the catastrophic consequences of earthquakes. We here present observations and analyses of the M7.1 September 19, 2017, event that—while triggered by an intraplate rupture at approximately half the epicentral distance of the 1985 event relative to Mexico City—caused severe structural damage to a few tens of buildings located in a relatively narrow zone between the hill and lake zones of the basin, known as the transition zone. We show that the M 7.1 mainshock exposed the vulnerabilities of the pre-1985 building code in the transition zone; but more importantly highlighted the improvement of the 1987 building code revision in terms of the performance of modern high-rise buildings that suffered catastrophic consequences during the 1985 Michoácan earthquake sequence. We next analyze several records collected at stations in the basin over the past 20 years. We highlight the importance of three-dimensional heterogeneity of the basin sediments, the coupling between hydrological setting and site response and their evolution with time, and the energy interaction between the deep basin edge and the shallow clay layers. Results presented are the collective effort of the GEER teams that were deployed to perform post-earthquake reconnaissance in the affected regions of the epicentral area and in Mexico City after the M 7.1 September 19, 2017, earthquake sequence.

Over 31,000 cone penetration test (CPT) soundings collected in Christchurch, New Zealand, in the aftermath of the 2010–2011 Canterbury earthquake sequence (CES) are available on the New Zealand Geotechnical Database (NZGD). This incredible dataset has allowed for detailed comparisons between observed land performance during the CES and CPT-based retrospective predictions of liquefaction triggering on an unprecedented spatial scale. Results from these comparisons indicate a significant number of “false positive” CPT-based liquefaction triggering predictions in which severe liquefaction was predicted but no, or very minor, surface manifestations of liquefaction were observed. In August of 2015, an in situ site characterization study was initiated to further investigate 31 false positive liquefaction sites in Christchurch. The testing program included: (a) seismic CPT, (b) high-resolution compression, and shear wave velocity measurements made via direct-push crosshole testing, and (c) continuous soil sampling via sonic drilling. This paper presents in situ test results from one of these case history sites (Palinurus Road), at which moderate surficial evidence of liquefaction was observed over part of the site during the February 2011 Christchurch earthquake, but not over the remainder of the site. Liquefaction triggering analyses performed on 12 CPT soundings across the site were very similar, yielding liquefaction potential indices indicative of severe liquefaction across the entire site. Refined analyses including consideration for (1) site-specific fines content data, (2) partial saturation (as indicated by V_P), (3) coarse-to-fine-grained soil interlayering, (4) non-liquefying crust thickness, and (5) soil microstructure are performed to attempt to reconcile the over-prediction of liquefaction severity. The over-prediction of liquefaction severity for the 2010 Darfield earthquake was reconciled as result of these adjustments; however, the over-prediction of liquefaction severity for the 2011 Christchurch earthquake was not reconciled.

Sites gained on the sea through near-shore reclamation are vulnerable to liquefaction in the absence of soil improvement provisions. The need, extent, and degree of such provisions are function of the assessment of liquefaction susceptibility which depends on the in situ characteristics of the fill materials. Given the nature of the materials and the process of placement, spatial variability of the soil parameters is believed to be a significant source of uncertainty that will impact the estimated risk of liquefaction and the design of liquefaction-mitigation measures. In this paper, an array of 118 boreholes with standard penetration tests performed at 1.5 m depth intervals is utilized to characterize the spatial variability in the fill across a recently reclaimed site. Conditionally-simulated random fields were used to model the spatial variability of the N-values and the fines content at the site. The randomly simulated N-values and fines content were then used as input to an SPT-based liquefaction assessment framework to assess the risk of liquefaction of different zones within the site when subjected to different seismic loading scenarios. Results of the probabilistic analysis point to the importance of modeling spatial variability using realistic random fields that reflect the uncertainty in the soil properties of the site as portrayed in the site investigation data.

Many procedures commonly used in geotechnical engineering rely on borehole data as input. While these procedures may characterize the uncertainty around their predictions, they typically do not provide guidance for evaluating uncertainty arising from imprecision or inaccuracy in measuring input parameters nor do they account for spatial variability in soil properties. This paper summarizes a methodology for generating synthetic borehole data using random fields and Monte Carlo simulation. Guidance is provided regarding selection of parameters and study size. Such a procedure allows geotechnical engineering practitioners to account for uncertainty in measurements of subsurface conditions, and to investigate whether their conclusions are sensitive to this uncertainty. Incorporation of uncertainty around the soil profile’s geometry and its relevant parameters gives a more complete estimate of possible outcomes (e.g., liquefaction triggering, surficial manifestation of liquefaction, or foundation damage due to liquefaction). As an example, the proposed procedure is then used to generate synthetic borehole data for a case study in Adapazari, Turkey.

Typical practice for assessing liquefaction-induced deformations is to use a number of in situ tests. However, the actual soil profile is often more complex and the sedimentary architectures are not always uniform. Due to this variability, the liquefaction-induced ground deformations are difficult to assess with certainty. Recent studies have shown that presence of a continuous impermeable layer within a liquefiable permeable deposit could affect the formation of the water film beneath it. However, little effort has been directed to the effect of discontinuous distribution of impermeable layers on the liquefaction potential. Thus, there is a need to explore the effect of discontinuous impermeable layers on liquefaction surface manifestation. In this paper, simplified site models are built to study this effect through numerical modeling. The results indicate that the discontinuity of the capping clay layer affects the redistribution of the pore pressure, and thus the patterns and magnitude of liquefaction-induced deformations.

There have been significant advances in the prediction of liquefaction triggering since the late 1960s. However, research to date has not provided consensus on preferred practices to estimate the consequences of liquefaction and quantify the large uncertainties involved. The severity of liquefaction vividly displayed during the 1995 Kobe earthquake, and more recently in the 2010–2011 Canterbury earthquake sequence has reinforced the need to develop practical analysis and design guidelines that more rigorously address liquefaction effects on critical infrastructure. Current approaches to estimate lateral spreading include empirical models, integration of permanent shear strains, Newmark sliding block analyses, and nonlinear dynamic analyses. The objective of this study is to assess the suitability of the integration of maximum shear strains to calculate lateral spreading as measured in centrifuge tests. The agreement between measured and estimated lateral deformations is analyzed for various site and loading conditions. Insights on the strengths and limitations of the considered approach are discussed.

Liquefaction-induced free-field settlements can be attributed to several mechanisms, including volumetric reconsolidation strain, ground loss due to ejecta, and vertical components of lateral spreads and flow slides. In practice, the volumetric-strain settlement is typically predicted using one of several semi-empirical methods, but there is no consensus of which, if any, is most accurate. Moreover, no method has been proposed to predict the magnitude of settlement owed to ejecta, even though this component could be larger than any other. The lack of a complete and trusted model for predicting free-field settlement is due, in part, to the scarcity of case-history data needed to apportion and model its individual components. However, the 2010–2011 Canterbury, New Zealand, earthquake sequence (CES) resulted in a liquefaction dataset of unprecedented size and quality, presenting a unique opportunity to assess and advance liquefaction analytics. Accordingly, this study compiles and analyzes 1,013 high-quality settlement case-histories from the CES, such that the contributing components of settlement can be reasonably accounted for and quantified. Presented herein are preliminary findings of this effort, summarized as follows: (1) popular CPT-based models for predicting volumetric-strain settlement exhibit strong bias, such that smaller total settlements (<~6 cm) are underpredicted while larger total settlements (>~6 cm) are significantly overpredicted; (2) these poor predictions cannot be explained by the uncertainty of CPT proxies of relative density, which are central to predictions of volumetric strain; (3) introduction of a depth-weighting function lessens the prediction bias, showing promise for future study; and (4) prediction-performance is related to the observed volume of liquefaction ejecta, illustrating the need to account for and model this mechanism of settlement. Collectively, the findings suggest popular methods for predicting free-field settlement have significant room for improvement.