Journal of Geotechnical and Geoenvironmental Engineering

The objective of this study was to evaluate the chemical compatibility of model soil-bentonite backfills containing multiswellable bentonite (MSB) relative to that of similar backfills containing untreated sodium (Na) bentonite or a commercially available, contaminant-resistant bentonite (SW101). Flexible-wall tests were conducted on consolidated backfill specimens ($\text{effective stress}=34.5 \text{kPa}$) containing clean sand and 4.5–5.7% bentonite (by dry weight) using tap water and calcium chloride (CaCl₂) solutions (10–1,000 mM) as the permeant liquids. Final values of hydraulic conductivity ($k$) and intrinsic permeability ($K$) to the CaCl₂ solutions were determined after achieving both short-term termination criteria as defined by ASTM D5084 and long-term termination criteria for chemical equilibrium between the influent and effluent. Specimens containing MSB exhibited the smallest increases in $k$ and $K$ upon permeation with a given CaCl₂ solution relative to specimens containing untreated Na bentonite or SW101. However, none of the specimens exhibited more than a fivefold increase in $k$ or $K$, regardless of CaCl₂ concentration or bentonite type. Final $k$ values for specimens permeated with a given CaCl₂ solution after permeation with tap water were similar to those for specimens of the same backfill permeated with only the CaCl₂ solution, indicating that the order of permeation had no significant effect on $k$. Also, final $k$ values for all specimens were within a factor of two of the $k$ measured after achieving the ASTM D5084 termination criteria. Thus, use of only the ASTM D5084 criteria would have been sufficient to obtain reasonable estimates of long-term hydraulic conductivity for the specimens in this study.

Pile setup in cohesive soils has been a known phenomenon for several decades. However, a systematic field investigation to provide the needed data to develop analytical procedures and integrate pile setup into the design method rarely exists. This paper summarizes a recently completed field investigation on five fully instrumented steel H-piles embedded in cohesive soils, while a companion paper discusses the development of the pile setup method. During the field investigation, detailed soil characterization, monitoring of soil total lateral stress and pore-water pressure, collection of pile dynamic restrike data as a function of time, and vertical static load tests were completed. Restrike measurements confirm that pile setup occurs at a logarithmic rate following the end of driving, and its development correlates well with the rate of dissipation of the measured pore-water pressure. Based on the field data collected, it was concluded that the skin friction component, not the end bearing, contributes predominantly to the setup, which can be accurately estimated for practical purposes using soil properties, such as coefficient of consolidation, undrained shear strength, and the standard penetration test $N$-value.

This paper establishes a methodology to quantify pile setup by using recent field test data that was presented in a companion paper for steel H-piles driven in cohesive soils. Existing methods found in literature for the same purpose either require restrikes of piles onsite or are developed for a specific soil type and seldom use easily quantifiable soil properties, despite their significant influence on pile setup. Following a critical evaluation of the existing methods, a new approach for estimating pile setup was developed using dynamic measurements and analyses in combination with measured soil properties, such as the horizontal coefficient of consolidation, undrained shear strength, and the standard penetration test $N$ value. Using pile setup information available in the literature, the proposed approach has shown that it provides good estimates for the setup of steel H-piles, as well as for other types and sizes of driven piles.

The paper presents a physically simplified method for computing displacements and structural forces on piles under conditions of lateral spreading triggered by the large seaward displacement of a harbor quay wall. The method avoids the empirical selection of stiffness-reduction factors and the associated use of $p\text{-}y$ curves that current state-of-the-art methods use. Instead, the three-dimensional (3D) highly nonlinear problem is approximated in two steps, both involving two-dimensional (2D) plane-strain analyses. The first step involves a vertical (representative) slice in which the pile group has been omitted and that, shaken at its base, gives the permanent deformation of the quay wall and of the liquefiable soil. It is an effective stress analysis. In the second step, a horizontal (representative) slice taken from the middle of the liquefiable zone is subjected to an outward quay wall displacement; the goal is to evaluate the reduction of the pile displacement over the free-field one and the ensuing pile group distress. The pile resistance to ground deformation depends heavily on the constraints imposed by the superstructure, as well on the exact stiffness of the soil layers. Thus, the interplay between soil piles-quay wall under soil flow conditions is captured in a physically meaningful way. The predictions compare well with results from two centrifuge tests.

The response of a laterally loaded pile is normally dominated by the limiting force per unit length profile (LFP, ${\mathit{p}}_{\mathit{u}}$) mobilized along the pile to the depth of the plastic zone (e.g., maximum slip depth, max ${\mathit{x}}_{\mathit{p}}$) and depends on pile-soil relative rigidity (via the subgrade modulus of elasticity $\mathit{k}$). The values of ${\mathit{p}}_{\mathit{u}}$ and the shear modulus ($\mathit{G}$, thus $\mathit{k}$) were deduced extensively against the measured response of 32 and 20 piles tested in situ in clay and sand, respectively. In this paper, simplified closed-form solutions are presented for the design of laterally loaded piles. The parameters of the ${\mathit{p}}_{\mathit{u}}$ and the $\mathit{G}$ obtained previously are utilized to examine the impact of loading eccentricity on piles and to assess the validity of existing ${\mathit{p}}_{\mathit{u}}$-profiles. Expressions are also explored and provided regarding determinations of the ${\mathit{p}}_{\mathit{u}}$-profiles, the modulus of the subgrade reaction $\mathit{k}$ (via $\mathit{G}$), the ranges of the plastic zone (max ${\mathit{x}}_{\mathit{p}}$), and the depth of elastic influence (via the critical pile length ${\mathit{L}}_{\text{cr}}$), along with the correlations between the shear modulus of soil $\mathit{G}$ and undrained shear strength (${\mathit{s}}_{\mathit{u}}$) and/or the standard penetration test blow count ($\mathit{N}$). It is noted that the popularly adopted LFPs (such as Matlock’s LFP, Reese’s LFP, and American Petroleum Institute code methods) are not sufficiently accurate for 60–85% of the 52 piles investigated herein, especially for those with a diameter >1.5 m.

The prediction of the shaft capacity of displacement piles in clay still relies wholly on empirical or semiempirical approaches. This paper examines the predictive abilities of five well-known methods against an extended version of an existing database of pile tests. It is seen that coefficients of variation (COVs) of the ratio of calculated to measured capacities ($Q_c/Q_m$) are significantly higher than expected. The paper then uses the database of shaft capacities to examine potential relationships between local shaft friction and the cone penetration test end resistance ($q_t$). Predictions are subsequently compared with capacities measured in a series of centrifuge tests and with the distributions of peak frictions observed in two well-documented field tests. Widely different formulations adopted by each of the existing empirical methods give broadly similar COV values for $Q_c/Q_m$ ratios, and it is concluded that the database of high-quality pile tests needs to be expanded significantly if the reliability of estimates of shaft capacity in clay is to be improved.

This paper develops a Bayesian approach for probabilistic site characterization (i.e., on both stratigraphy and soil properties) using cone penetration tests (CPTs). The available site information prior to the project (e.g., existing geological maps, geotechnical reports, and local experience) is used in the Bayesian approach as prior knowledge, and it is integrated systematically with results of CPTs that are performed deliberately for the project. The inherent spatial variability of soil is modeled explicitly by random field theory. The proposed approach contains two major components: a Bayesian model class selection method to identify the most probable number of statistically homogenous layers of soil and a Bayesian system identification method to estimate the most probable layer thicknesses and soil properties probabilistically. Equations are derived for the Bayesian approach, and the proposed approach is illustrated using a set of real CPT data obtained from a site in Netherlands. It has been shown that the proposed approach correctly identifies the number and thicknesses/boundaries of the statistically homogenous layers of soil and provides proper probabilistic characterization of soil properties. The Bayesian approach provides a means to identify the statistically homogenous layers progressively by gradually zooming into local differences with improved resolution, and it also contains a mechanism to determine when to stop such zooming. In addition, a sensitivity study is performed to explore the effect of prior knowledge.

Explicit empirical relationships for correcting liquefaction resistance of unsaturated soil below the groundwater table using compression wave velocity ($V_P$) and shear wave velocity ($V_S$) are derived and evaluated in this paper. The relationships are derived using published laboratory test data for four different sands and one silt material. The laboratory test data based on $V_P$ exhibit less scatter than the test data based on the ratio $V_P/V_S$. For this reason, the $V_P$-based relationship is recommended for correcting the liquefaction resistance of unsaturated soil. The influence of loading cycles and relative density on the recommended relationship is investigated. From a review of 40 field case history sites, it is found that in the critical layers at 20% of the sites, average $V_P$ is less than 1,400 m/s, indicating unsaturated conditions. Liquefaction resistances computed for eight field case histories with $V_P<\mathrm{1,200} \text{m/s}$ in the critical layers are corrected and plotted on penetration- and $V_S$-based liquefaction evaluation charts. The results support the use of the correction for unsaturated conditions below the groundwater table.

Maximum seismic shear stresses (${\mathit{\tau }}_{\text{max}}$) have been recognized as one of the important parameters in design practice. This study develops ground-motion parameters for ${\mathit{\tau }}_{\text{max}}$ and implements these in probabilistic seismic hazard analysis to provide the ${\mathit{\tau }}_{\text{max}}$ distribution of deep soil layers for design purposes. The application of improved ground-motion parameters for ${\mathit{\tau }}_{\text{max}}$ is demonstrated at the Oakland International Airport, where a thick Young Bay Mud deposit exists under the artificial fill. Model biases in the predictive equations of seismic shear-stress reduction coefficients (${\mathit{r}}_{\mathit{d}}$) are evaluated by comparison with the site response analyses performed with a wide range of input ground motions. Based on these results, we introduce improved ground-motion parameters for ${\mathit{\tau }}_{\text{max}}\left({\mathit{I}}_{\mathit{tau}}\right)$ as a linear combination of spectral accelerations, implemented in probabilistic seismic hazard analysis to calculate seismic hazard curves. Conditional mean spectra are calculated, given ${\mathit{I}}_{\mathit{tau}}$ at 10% in 50 years to illustrate the variations in frequency contents with depth compared with the uniform hazard spectra. Finally, ${\mathit{\tau }}_{\text{max}}$ is calculated with depth by using hazard values of ${\mathit{I}}_{\mathit{tau}}$ and compared with the peak-ground-acceleration-based and uniform-hazard-spectra-based calculations. Analysis results show that ${\mathit{\tau }}_{\text{max}}$ will be underestimated for deep soil layers by peak-ground-acceleration-based calculation if the median value of ${\mathit{r}}_{\mathit{d}}$ is used in design practice.

The objective of this study is to develop and apply reflection methods for guided ultrasonic waves used in evaluating the grout condition of rock bolts installed in a rock mass. Three types of rock bolts with different grouting ratios are prepared. First, piezoelectric disk elements are used as the source of guided ultrasonic waves to investigate grouted nonembedded rock bolts. Second, reflection methods with piezo disk elements and hammer impacts are used with rock bolts embedded in concrete columns. Third, a hammer impact with a center punch is applied for investigating rock bolts embedded in a rock mass in the field. The measured guided waves were analyzed using wavelet transforms. The peak magnitudes of the wavelet transform are used for group velocity calculations. Although piezo disk elements are sufficient as sources for nonembedded rock bolts and rock bolts installed in concrete columns, they do not provide sufficient energy in the field. However, a hammer impact with a center punch can generate guided ultrasonic waves with enough energy to evaluate the nongrouted ratio in rock bolts embedded in a rock mass. The group velocities of the guided ultrasonic waves increase with increasing nongrouted ratio. This study demonstrates that the suggested hammer impact method is effective for evaluating the nongrouted ratio of rock bolts in the field.

Structural reliability analysis is not widely used for the design and assessment of concrete dams. In this paper, the system reliability of a spillway structure consisting of two monoliths is calculated. Limit state functions are defined from the failure modes sliding in the concrete rock contact, sliding in the rock mass, and adjusted overturning. Random variables in the limit state functions are defined by stochastic distributions. These are defined based on site investigations and laboratory tests from samples taken at the dam. Simulations and information from the literature are used for the remaining variables. The safety index is calculated by the first-order reliability method for each failure mode and monolith, and the system reliability is approximated by direct integration of the bivariate normal distribution. The output is the safety index including associated sensitivity values at the single failure mode, monolith, and system levels. The results show that the system safety is governed by a persistent rock joint beneath one monolith. A system reliability analysis is found to be a useful tool in the dam risk management process as it can be used to calculate the probability of failure and to identify important failure modes and variables.

A probabilistic porous model is used to simulate the retention curves of different types of soils. The model is based on the same principles of porous networks models; however, it shows important advantages over them. It considers two different elements: the sites and the bonds. It is able to simulate the retention curves of a material when its pore size distribution is known. It can also be used to interpret more realistically the results of mercury intrusion porosimetry tests. In this paper, the model is used to obtain the pore size distribution of different soils using the retention curve as its data source. The comparison between numerical and experimental results for different soils shows that the model approximately reproduces the pore size distribution obtained from mercury intrusion porosimetry tests.

Understanding the factors that control the transport of soil particles detached by water is essential in predicting the contingency of internal and surface erosion of embankments during storms and floods. It is particularly useful to be able to evaluate the sensitivity of the available soils to erosion at an early stage of a project and the risk assessment of an existing earthwork. The presented study is based on laboratory tests with a hole erosion test apparatus and tries to characterize the internal erosion of soils. The test consists of eroding a drilled soil specimen by water flow with a constant pressure drop at the boundaries of the specimen while recording the flow rate and studying the hole enlargement. A new version of the apparatus with improved instruments and an interpretation method is proposed, estimating the erosion rate based on the turbidity of the outflow and independent of hydraulic charge. Several remolded kaolinite-sand mixtures are tested as reference soil textures, and the results are analyzed with the proposed and existing interpretation methods.

In past years, railroad transportation has been of growing interest because of its efficiency and advancement in railway technologies. However, many issues arise because of the variability in subsurface conditions along the sizeable lengths of track that exist. One very important issue is the need for significant upkeep and maintenance for railways passing over areas of poor soil conditions as a result of continuous deformation and a lack of stiffness from the foundation. One general solution for lack of substructure integrity has been confinement, applied through a variety of reinforcement types, including geocell. To investigate the effectiveness of geocell confinement on substructure integrity, a series of embankment model tests with different configurations of geocell placement (one layer and two layers of geocell) were constructed and loaded monotonically and cyclically for comparison with unreinforced, control tests. On the completion of these tests, the model embankments were simulated numerically using finite-element procedures. The results, which matched reasonably well, were then used as validation for a parametric study, observing the effects of less competent geocell material, gravel, and foundation conditions and their implications. The tests and numerical simulations demonstrate that geocell confinement effectively increased stiffness and strength of a gravel embankment while reducing vertical settlement and lateral spreading. Additionally, the parametric study shows that the use of geocell provides a composite mattressing effect that distributes subgrade stress more uniformly than without reinforcement, increasing bearing capacity and reducing settlement, especially on soft foundations. The results suggested that in some site conditions, use of geocell might be an economical alternative to frequent maintenance and/or lower train speeds.

Particle shape affects the mechanical behavior of soils, including packing density, stiffness, volume change during shear, and strength. Laboratory experiments conducted to study the mechanical response of sand mixtures made of round and angular grains show an increase in void ratio, small strain shear modulus ${\mathit{G}}_{\text{max}}$ (constant fabric), oedometric compressibility $C_C$ (fabric changes), and friction angle but a decrease in lateral stress coefficient ${\mathit{k}}_{\text{0}}$ as the mass fraction of angular particles increases. These results reflect variations in particle mobility and highlight the relative role of contact stiffness versus fabric changes.