Search

Current simplified models for predicting liquefaction triggering and manifestation do not account for the mechanisms of liquefaction triggering and surface manifestation in a consistent and sufficient manner. Specifically, an artifact of the way simplified triggering curves have traditionally been developed is that they inherently include some factors that influence surface manifestations, particularly for medium-dense to dense soils. As a result, using the simplified triggering curves in conjunction with the manifestation models results in the double-counting, omission, or general obscuration of distinct factors that influence triggering and manifestation. Accordingly, the objective of this paper is to develop a framework for deriving a true simplified liquefaction triggering model consistent with a defined manifestation model, such that factors influential to triggering and manifestation are handled more rationally and consistently. Operating in conjunction with the LSNish manifestation model, the performance of the true triggering curve is compared with an existing, popular triggering curve using a set of 50 global case histories; the results are favorable for the proposed framework. Furthermore, the ability of the true liquefaction model to be used in conjunction with manifestation models that account for complex free-field soil profile stratigraphies or developed sites is a significant advantage over traditionally developed triggering models. [ABSTRACT FROM AUTHOR]

One of the fundamental decisions when performing one-dimensional (1D) site response analyses (SRA) involves the selection of the depth and dynamic properties of the elastic halfspace (EHS). This boundary condition assumes linear and homogenous material underlying the soil column for an infinite depth. This assumption implies that waves refracted into the EHS are fully absorbed, and as a result, energy from waves that are potentially reflected back toward the surface from deeper impedance contrasts in the actual geologic profile are not accounted for in the SRA. If a strong soil-rock seismic impedance contrast is present at the site of interest, the EHS boundary is typically set at that depth. However, the actual geologic profile below this impedance contrast may not be in accord with the assumed properties of the EHS, which can lead to systematic errors in the SRA. An analytical expression to quantify these errors is derived in this study, verified using an idealized three-layer profile, and compared to case studies of nine real sites in Charleston, South Carolina. Our results show that the presence of a single strong impedance contrast does not suffice as the sole condition to define the EHS boundary. Frequency-dependent errors in site amplification associated with the assumptions inherent to the EHS used in the SRA can be evaluated as a function of multiple impedance contrasts present in the profile. Smaller errors are associated with strong impedance contrasts at shallower layers and/or minimal impedance contrast among layer interfaces at depth. We also find that strong impedance contrasts located at great depths within deep soil deposits introduce nonnegligible errors to site response results. [ABSTRACT FROM AUTHOR]

This paper proposes a new approach for incorporating the positive attributes of the small-strain shear wave velocity (VS), stress-based simplified procedure and the cyclic strain procedure into penetration test, stress-based simplified liquefaction triggering models, with the objective of more fully accounting for the influence of intrinsic soil properties and soil state variables on liquefaction triggering. Current simplified liquefaction procedures are limited in their ability to capture the effects of intrinsic properties (grain size, mineralogy, grain shape, etc.) and the state properties (stress state, void ratio, fabric, etc.). To overcome these limitations, a new mechanistically based Kγ factor is proposed that can be incorporated in penetration test, stress-based simplified liquefaction triggering models in place of the currently used Kσ factor. However, Kγ is conceptually very different from Kσ. While most Kσ relationships have largely been empirically based and relate to the soil's cyclic resistance to liquefaction, Kγ is more mechanistically based and relates to the loading imposed on the soil. Specifically, Kγ is based on equating the shear strain induced in a given soil at given initial stress state and subjected to a given shear stress to the induced shear strain when the soil is confined at a reference initial stress state, all else being equal. Analyses show that Kγ is able to capture the liquefaction triggering behavior in both lab and field data in a wide range of soils and stress states. Numerically, Kγ and Kσ are similar for young, normally consolidated sandy soils when the factor of safety (FS) against liquefaction triggering is close to one, but may differ significantly for other scenarios and/or conditions. This has important implications for probabilistic-based analyses which consider a range of shaking intensities imposed on the soil, not just the case where FS=1. [ABSTRACT FROM AUTHOR]

The severity of surface manifestation of liquefaction is commonly used as a proxy for liquefaction damage potential. As a result, manifestation severity index (MSI) models are more commonly being used in conjunction with simplified stress-based triggering models to predict liquefaction damage potential. This paper assesses the limitations of three existing MSI models and a fourth MSI model that is developed herein. The different models have differing attributes that account for factors influencing the severity of surficial liquefaction manifestations, with the newly proposed model accounting more factors than the others. The efficacies of these MSI models are evaluated using well-documented liquefaction case histories from Canterbury, New Zealand, with the deposits primarily comprising clean to nonplastic silty sands. It is found that the MSI models that explicitly account for the contractive/dilative tendencies of soil did not perform as well as the models that do not account for this tendency, opposite of what would be expected based on the mechanics of liquefaction manifestation. The likely reason for this is the double-counting of the dilative tendencies of medium-dense to dense soils by these MSI models because the liquefaction triggering model, to some extent, inherently accounts for such effects. This implies that development of mechanistically more rigorous MSI models that are used in conjunction with simplified triggering models will not necessarily result in improved liquefaction damage potential predictions and may result in less accurate predictions. This provides the impetus for the development of a new framework that clearly and distinctly separates triggering and manifestation. [ABSTRACT FROM AUTHOR]

Cone penetrometer testing (CPT) is a frequently used soil characterization technique for liquefaction assessment; however, this technique has shortcomings in accurately characterizing very thin soil layers having thicknesses less than two to three times the diameter of the cone. In this study, the material point method (MPM) is used to generate numerical "measured" (or "blurred") CPT tip resistance (qm) in complex soil profiles. Results show that MPM is capable of accurately simulating qm in soil profiles with layers as thin as 20 mm, even when using basic constitutive models and simplified drainage conditions. It is further shown that MPM simulations are able to replicate the tendency of the CPT to smear the boundaries between very thin, interbedded soil layers in a way that obscures their true thickness and stiffness (typically referred to as thin-layer, transition-zone, or multiple thin-layer effects). While previous numerical studies of CPT have been performed in profiles composed of two or three layers, this study considers highly interlayered profiles with upwards of 27 soil layers. Difficulties of developing and implementing multiple thin-layer corrections are presented. It is shown that a numerical framework like MPM can generate a larger set of data for the development and validation of multiple thin-layer correction procedures with the aim of improving liquefaction severity predictions in complex soil profiles. [ABSTRACT FROM AUTHOR]

In deterministic evaluations, liquefaction triggering potential is assessed by comparing the computed factor-of-safety (FS) against liquefaction triggering to some minimal acceptable FS. While some guidelines are available for selecting the minimal acceptable FS , there is no standard value. Herein, Receiver Operating Characteristic (ROC) analyses are used to develop an approach for selecting the optimal minimal acceptable FS (i.e., optimal FS) for a project based on the relative costs of mispredictions. Utilizing different liquefaction triggering models and their associated case-history databases, relationships are established between the optimal FS and the ratio of the cost of a false-positive prediction to the cost of a false-negative prediction (i.e., cost ratio, CR). Also, by combining the FS data from different models, a "generic" FS-CR relationship is developed that "averages out" the degree of conservatism inherent to the individual triggering models. Similarly, relationships relating the optimal probability of liquefaction triggering (PL) to CR are developed for the probabilistic variants of the triggering models, as well as a generic PL-CR curve. [ABSTRACT FROM AUTHOR]

Data from 58 high-quality liquefaction case histories from the Darfield and Christchurch earthquakes are utilized to investigate the efficacy of current liquefaction aging correction procedures. Toward this end, liquefaction case histories are analyzed in which aging corrections are and are not applied, and the resulting predictions are compared to the actual liquefaction response of the deposits. An error-index is calculated to quantify the efficacy of aging corrections. While all the sites located in the Christchurch area are classified as Holocene, based on their geological age, their liquefaction response is influenced more by the geotechnical age of the soil deposits. Aging correction was determined to be beneficial for the liquefaction assessment of soils that experienced recurrent liquefaction (i.e., geotechnical young deposits). However, aging corrections were determined to exacerbate the liquefaction assessment of relatively old (greater than ∼62–580 years) soil deposits. [ABSTRACT FROM AUTHOR]

The number of equivalent cycles (neq) concept plays an important role in geotechnical earthquake engineering and underlies the accounting for ground-motion duration in simplified liquefaction evaluation procedures, whether explicitly or implicitly. In this regard, several neq correlations have been proposed over the years that were developed using a similar variant of the Palmgren-Miner (P-M) fatigue theory. The correlations presented herein were developed using an alternative implementation of the P-M theory that better accounts for the nonlinear response of the soil and for multidirectional shaking. The proposed correlations are for shallow crustal earthquakes in both active tectonic and stable continental regimes. Additionally, two forms of the correlations are presented, one being expressed as a function of peak ground acceleration (amax). This relation shows a strong negative correlation between amax and neq, implying that motions with high amplitudes have short durations and vice versa. This negative correlation is not accounted for in most previously proposed neq correlations, which could result in the erroneous weighting of the unlikely scenarios of high amplitude-longer duration and low amplitude-short duration motions in liquefaction hazard studies. [ABSTRACT FROM AUTHOR]

Since its inception in the early 1970s, the stress-based simplified procedure has become the standard of practice worldwide for evaluating liquefaction triggering potential. Central to this procedure is the stress reduction coefficient, rd, which allows the estimation of the seismically induced stresses at depth in a soil profile without the need to perform a numerical site response analysis. Proposed herein is a new rd relationship that was developed from equivalent-linear site response analyses performed on soil profiles representative of those in the liquefaction case history databases. The input motions used in the analyses are representative of those from shallow crustal earthquakes. Two variants of the rd relationship are presented that allow it to be used when the profile's shear-wave velocities are either known or unknown, with the former yielding values having less uncertainty. Additionally, calibration coefficients are provided for both active and stable continental tectonic regimes. In comparison with other rd relationships that are commonly used, the relationship proposed herein should yield values that have less bias and uncertainty. [ABSTRACT FROM AUTHOR]

Seismic compression is the accrual of volumetric strains in unsaturated soils caused by cyclic loading and has caused significant damages to buildings and other structures during earthquakes. To date, the available methods for predicting the severity of seismic compression have mainly been simplified procedures, in which a number of equivalent cycles are used to represent the duration of earthquake loading. Often, however, the number of equivalent cycles is computed inconsistently with the underlying mechanics of seismic compression. This paper proposes a non-simplified procedure for predicting the severity of seismic compression. The procedure is based on a modified version of the Richart-Newmark cumulative damage hypothesis, wherein volumetric strain is used as the damage metric. The proposed model was calibrated using data from 425 constant-amplitude sinusoidal strain-controlled cyclic simple shear tests performed on clean sand and validated using test data from samples subjected to variable-amplitude sinusoidal and earthquake loadings. In addition to predicting the severity of seismic compression, the proposed model can be used to compute number of equivalent shear-strain cycles for use in simplified models, consistent with the seismic compression phenomenon. In comparison with other proposed nonsimplified models for computing seismic compression, the proposed model gives good agreement with the measured seismic compression. [ABSTRACT FROM AUTHOR]

While the liquefaction potential index (LPI) has been used to characterize liquefaction hazards worldwide, calibration of LPI to observed liquefaction severity is limited, and the efficacy of the LPI framework and accuracy of derivative liquefaction hazard maps are thus uncertain. Herein, utilizing cone penetration test soundings from nearly 1,200 sites, in conjunction with field observations following the Darfield and Christchurch, New Zealand, earthquakes, this study evaluates the performance of LPI in predicting the occurrence and severity of surficial liquefaction manifestations. It was found that LPI is generally effective in predicting moderate-to-severe liquefaction manifestations, but its utility diminishes for predicting less severe manifestations. Additionally, it was found that LPI should be used with caution in locations susceptible to lateral spreading, because LPI may inconsistently predict its occurrence. A relationship between overpredictions of liquefaction severity and profiles having soils with high inferred fines-content was also investigated. It was determined that the LPI procedure might be improved if it accounted for the characteristics of soils in the crust and interbedded nonliquefiable layers, as well as the characteristics of the soils predicted to liquefy. Further research is needed in this regard. Finally, it was found that for the in situ conditions in Christchurch, the computed LPI values were relatively sensitive to estimates of groundwater depth because of the proximity of liquefiable strata to the ground surface. [ABSTRACT FROM AUTHOR]

This paper discusses the applicability of two simple models for predicting pore water pressure generation in nonplastic silty soil during cyclic loading. The first model was developed by Seed et al. in the 1970s and relates the pore pressure generated to the cycle ratio, which is the ratio of the number of applied cycles of loading to the number of cycles required to cause liquefaction. The second model is the Green-Mitchell-Polito model proposed by Green et al. in 2000, which relates pore pressure generation to the energy dissipated within the soil. Based upon the results of approximately 150 cyclic triaxial tests, the writers show that both models are applicable to silty soils. A nonlinear mixed effects model was used for regression analyses to develop correlations for the necessary calibration parameters. The results show that the trends in both α and pseudoenergy capacity calibration parameters for the Seed et al. and Green et al. pore pressure generation models, respectively, differ significantly for soils containing less than and greater than ∼35% fines, consistent with the limiting fines content concept. [ABSTRACT FROM AUTHOR]

As an alternative to a field-based liquefaction resistance approach, cyclic triaxial tests with bender elements were used to develop a new correlation between cyclic resistance ratio (CRR) and overburden stress-corrected shear-wave velocity (VS1) for two nonplastic silts obtained from Providence, Rhode Island. Samples of natural nonplastic silt were recovered by block sampling and from geotechnical borings/split-spoon sampling. The data show that the correlation is independent of the soils’ stress history as well as the method used to prepare the silt for cyclic testing. The laboratory results indicate that using the existing field-based CRR-VS1 correlations will significantly overestimate the cyclic resistance of the Providence silts. The strong dependency of the CRR-VS1 curves on soil type also suggests the necessity of developing silt-specific liquefaction resistance curves from laboratory cyclic tests performed on reconstituted samples. [ABSTRACT FROM AUTHOR]

The “equivalent number of cycles” concept is integral to cyclic liquefaction evaluations, whether applied directly in laboratory evaluations or via magnitude scaling factors in field evaluations. The premise of the concept is that the random motions of an earthquake can be represented by an equivalently damaging number of uniform stress cycles (neq), which facilitates laboratory testing and provides a convenient metric for comparing the duration of earthquake motions. The most widely used procedure for computing neq was developed by the late Professor H.B. Seed and colleagues in the late 1960s to early 1970s and is based on the Palmgren–Miner (P–M) cumulative damage hypothesis developed for metal fatigue evaluations. However, the original P–M hypothesis is intended for high cycle fatigue conditions and therefore applies to conditions where the strains are constrained to the elastic range of the material. By contrast, low cycle fatigue conditions, such as the case of soils subjected to strong ground shaking, are characterized by significant amounts of plastic strain. The implications of using a high cycle fatigue hypothesis to compute neq for evaluating liquefaction are examined herein, and an alternative approach to implementing the P–M fatigue hypothesis that more appropriately accounts for the nonlinear behavior of soil is proposed. The alternative implementation procedure equates energy dissipated in the soil subjected to the actual earthquake motions and neq uniform cycles. The results of a parametric study using the alternative implementation procedure show that neq varies as a function of earthquake magnitude, site-to-source distance, and depth in a soil profile. [ABSTRACT FROM AUTHOR]