Journal of Geotechnical and Geoenvironmental Engineering

The nature and scope of geotechnical engineering are such that the problems and projects we deal with are a never ending source of both challenge and excitement. In most cases we do very well in defining the problems and developing suitable solutions resulting in successful completed projects. Unfortunately, however, there remain far too many cases in which something goes wrong. Often a failure or other bad outcome comes as a complete surprise. However, in a significant number of cases perhaps the unexpected or surprise outcome might have been anticipated. Three illustrative case histories are reviewed and then examined to see if what went wrong might reasonably have been anticipated. The first involves a well-known and much studied stability failure along the composite double liner system of a hazardous waste landfill. The second is concerned with unusual soil types that were the cause of major difficulties during earthwork construction of a large embankment dam. The third is about a very large, slow-moving landslide that caused major distress to a roadway and impacted the safety of a large bridge. Several reasons why unexpected adverse outcomes may occur are stated and some possible means for reducing their frequency and severity in the future are proposed.

Geosynthetic-reinforced column-supported (GRCS) embankments have increasingly been used in the recent years for accelerated construction. Numerical analyses have been conducted to improve understanding and knowledge of this complicated embankment system. However, most studies so far have been focused on its short-term or long-term behavior by assuming an undrained or drained condition, which does not consider water flow in saturated soft soil (i.e., consolidation). As a result, very limited attention has been paid to a settlement-time relationship especially postconstruction settlement, which is critical to performance of pavements on embankments or connection between approach embankments and bridge abutments. To investigate the time-dependent behavior, coupled two-dimensional mechanical and hydraulic numerical modeling was conducted in this study to analyze a well-instrumented geotextile-reinforced deep mixed column-supported embankment in Hertsby, Finland. In the mechanical modeling, soils and DM columns were modeled as elastic-plastic materials and a geotextile layer was modeled using cable elements. In the hydraulic modeling, water flow was modeled to simulate generation and dissipation of excess pore water pressures during and after the construction of the embankment. The numerical results with or without modeling water flow were compared with the field data. In addition, parametric studies were conducted to further examine the effects of geosynthetic stiffness, column modulus, and average staged construction rate on the postconstruction settlement and the tension in the geosynthetic reinforcement.

An investigation was conducted to characterize and relate in situ soil stress-strain behavior to roller-measured soil stiffness. Continuous assessment of soil stiffness via roller vibration monitoring has the potential to significantly advance performance based quality assurance of earthwork. One vertically homogeneous and two layered test beds were carefully constructed with embedded sensors for the field testing program. Total normal stress and strain measurements at multiple depths reveal complex triaxial soil behavior during vibratory roller loading. Measured cyclic strain amplitudes were 15–25% of those measured during static roller passes due to viscoelasticity and curved drum/soil interaction. Low amplitude vibratory roller loading induces nonlinear in situ modulus behavior. Roller-measured stiffness and its dependence on excitation force is influenced by the stress-dependent modulus function of each soil, the varying drum/soil contact area, and by layer characteristics (modulus ratio, thickness) when layering is present. On vertically homogeneous clayey sand, roller-measured stiffness decreased with increasing excitation force, a behavior attributed to stress-dependent modulus reduction observed in situ. On the crushed rock over silt test bed, roller-measured stiffness increased with increasing excitation force despite the mild stress-dependent modulus reduction observed in the crushed rock. In this case, the stiffer crushed rock takes on a greater portion of the load, resulting in the increase in roller-measured stiffness.

In order to simulate the effect of drainage on soils adjacent to gravel drains that are installed as countermeasure against liquefaction, several series of cyclic triaxial tests were performed on saturated sands under partially drained conditions. The condition of partial drainage under cyclic loading was simulated in the laboratory using triaxial testing equipment installed with a drainage control valve to precisely regulate the volume of water being drained from test specimens. Effects of both drainage conditions and loading frequencies on cyclic response were incorporated through the coefficient of drainage effect, ${\alpha }^{*}$ . Experimental results showed that for sand exhibiting strain softening, the partially drained response was controlled by the critical effective stress ratio while for sand showing strain hardening behavior, the controlling factor was the phase transformation stress ratio. Moreover, test results indicated that the minimum liquefaction resistance under partially drained conditions can be used as a parameter to describe the liquefaction resistance of sands improved by the gravel drain method. From these results, a simplified procedure for designing gravel drains based on the factor of safety $\left(F_L\right)$ concept was proposed.

Current design guidelines for shallow foundations supporting building and bridge structures discourage footing rocking or sliding during seismic loading. Recent research indicates that footing rocking has the potential to reduce ductility demands on structures by dissipating earthquake energy at the footing-soil interface. Concerns over cyclic and permanent displacements of the foundation during rocking and sliding along with the dependence of foundation capacity on uncertain soil properties hinder the use of footing rocking in practice. This paper presents the findings of a series of centrifuge experiments conducted on shear wall-footing structures supported by dry dense to medium dense sand foundations that are subjected to lateral cyclic loading. Two key parameters, static vertical factor of safety $\left({\mathrm{FS}}_V\right)$ , and the applied normalized moment-to-shear ratio $(M∕(H\cdot L))$ at the footing-soil interface, along with other parameters, were varied systematically and the effects of these parameters on footing-soil system behavior are presented. As expected, the ratio of moment to the horizontal load affects the relative magnitude of rotational and sliding displacement of the footing. Results also show that, for a particular ${\mathrm{FS}}_V$ , footings with a large moment to shear ratio dissipate considerably more energy through rocking and suffer less permanent settlement than footings with a low moment to shear ratio. The ratio of actual footing area $\left(A\right)$ to the area required to support the vertical and shear loads $\left(A_c\right)$ , called the critical contact area ratio $(A∕A_c)$ , is used to correlate results from tests with different moment to shear ratio. It is found that footings with similar $A∕A_c$ display similar relationships between cyclic moment-rotation and cumulative settlement, irrespective of the moment-to-shear ratio. It is suggested that shallow foundations with a sufficiently large $A∕A_c$ suffer small permanent settlements and have a well defined moment capacity; hence they may be used as effective energy dissipation devices that limit loads transmitted to the superstructure.

This paper is a critical evaluation of the interpretation criteria of drilled shafts under axial compression loading. A wide variety of load test data are used for analysis, and these data are divided into drained and undrained databases. The interpretation criteria are examined from these load test results to establish a consistent compression interpretation criterion. Among these criteria, the range of each interpretation method presents approximately the same trend for both drained and undrained conditions. The statistical results show that the smaller the compression displacement, the higher the coefficient of variation. Moreover, the undrained load test results reveal less variability than the drained results. The load-displacement curve of a drained loading also demonstrates more ductility than that for undrained loading. Based on these analyses, the relative merits and interrelationships of these criteria are established, and specific design recommendations for the interpretation of compression drilled shaft load test, in terms of both capacity and displacement, are given.

Centrifuge tests to investigate the behavior during initial reservoir filling of a concrete faced rockfill dam (CFRD) with face slab stiffnesses that vary by a factor of about two are described. The two centrifuge models exhibited similar deformations at the crest and along the face slab, with crest settlements averaging $0.19H$ (%) and maximum face slab deformations averaging $0.88H$ (%). The centrifuge test results suggest that the face slab stiffness had little effect on deformations, at least for the range of stiffnesses examined here. A parametric study of transition (supporting) zone stiffness was performed using a numerical model calibrated using the centrifuge results. The numerical results indicated that face slab deformation is more influenced by transition zone stiffness than face slab stiffness, supporting the centrifuge results. Deformation measurements for 25 in-service CFRDs (including six Korean CFRDs—one of which was used as the basis for the centrifuge model dam) are presented and compared with the experimental and numerical results. The centrifuge experiments exhibited crest settlements similar to the Korean CFRDs; however, the centrifuge models exhibited considerably larger maximum face slab deflections. The larger values measured in the centrifuge tests likely resulted from some experimental limitations. These limitations, as well as suggestions for improving future centrifuge studies of CFRDs, are discussed.

The scope of this paper is to develop a new three-part wedge method to analyze translational failure when using an engineered berm to increase landfill space. The possible failure face may pass over the back slope of the berm or pass under the bottom of the berm. This depends on the geometric dimensions of the berm which in turn influences the amount of airspace gained. Each possible failure condition must be checked to make sure both the waste mass and berm remain stable. If the waste boundary is kept to be unchanged, ${\mathrm{FS}}_{\text{over}}$ will increase and ${\mathrm{FS}}_{\text{under}}$ will decrease with increase of the back slope of the berm. The critical failure condition may be changed from passing over to passing under the berm by changing the back slope of the berm. If the location of the front toe of the berm is fixed, both ${\mathrm{FS}}_{\text{over}}$ and ${\mathrm{FS}}_{\text{under}}$ increase with increased height of the berm, however, the critical failure condition will change from passing under the berm to passing over the berm.

Results from dynamic penetration tests are traditionally interpreted on the basis of empirical correlations, this being a frequent criticism to these tests. An alternative rational method of interpretation is proposed in this paper from which the energy delivered to the composition of rods is used to calculate a dynamic force that represents the reaction of the soil to the penetration of the sampler $\left(F_d\right)$ . Interpretation of soil properties both in sand and clay is based on this calculated dynamic force from which the internal friction angle and the undrained shear strength can be estimated. This is achieved from a simple combination of limit equilibrium analysis and cavity expansion theory. Case studies gathered from the Brazilian experience are reported in this paper to illustrate the applicability of the proposed approach.

Ground strains were estimated from vibrations measured during pile driving operations at a buried, prehistoric archeological site to monitor potential construction impacts. Subsurface characteristics of the site were investigated using multiple cone penetration test (CPT) soundings and the shear wave velocity profile was measured using the seismic CPT method. Embedded geophones and surface accelerometers were then used to measure ground vibrations during pile driving. Displacement gradients were estimated from the vibrations using the following three methods: (1) the difference between adjacent displacements divided by sensor spacing; (2) peak particle velocity divided by depth-dependent wave velocity (i.e., at the depth where the sensor was placed); and (3) peak particle velocity divided by frequency-dependent wave velocity from a measured dispersion curve. Methods (1) and (3) agreed well, while method (2) caused errors that depended on depth of embedment of the sensors and distance from pile driving. Errors in (2) were attributed to a mismatch between the depth-dependent wave velocity and the wave velocity on the frequency band that carried the largest velocity pulse through the dispersive soil profile. Ground strains were related to displacement gradients based on theoretical solutions of harmonic body waves and Rayleigh waves in dispersive elastic media. The peak estimated ground strains were smaller than the threshold volumetric shear strain, but a few centimeters of settlement were nevertheless observed at the site. The spatial extent of the settlement is characterized using attenuation rules fit to the vibration data, and by calibration with a settlement gauge. Ground cracking and vertical offsets that could potentially mask the archaeological history of the site were neither observed nor predicted from the observed vibration amplitudes. Estimated impact on archeological interpretation of artifacts in their stratigraphic context was likely insignificant except in the immediate region where the piles were driven. This insight will assist in future planning at sites with similar subsurface stratigraphy.

Time domain reflectometry (TDR) is performed as a complement to the monitoring methods in subsurface deformation in slope together with the global positioning system (GPS) to monitor ground deformation of high-altitude landslides in Li-shan. Four TDR cables were installed in drill holes near the monitoring stations in the landslide area. According to the recorded TDR waveforms, there were shear and tensile zones under the B-5, B-9, C-1, and C-2 stations. A comparison of the TDR waveforms with the monitored data and boring log revealed that the subsurface sliding occurred between layers of colluvium and strongly weathered slate. Three GPS receivers were installed to measure ground displacement in the landslide area. The results from the GPS were compared with the surface extensometers data on-site. The two initial baseline lengths were 451,188.10 and $\mathrm{908,212.4}\mathrm{mm}$ , respectively. The optimal data reduction achieved used a $3\mathrm{h}$ session with moving average for each hour’s GPS data. The standard deviation values of the GPS were 2.16 and $2.44\mathrm{mm}$ , respectively, on-site. The results of TDR and GPS measurements showed their applicability in the deformation monitoring of high-mountain landslides.

The kinematics of overhanging rock slopes and the mechanical constraints associated with this specific slope geometry were studied. Investigation of the problem began with a generalized rigid body analysis and was followed by a numerical discontinuous deformation analysis, both of which were performed in two dimensions. It was found that eccentric loading and hence the development of tensile stresses at the base of overhanging rock slopes control their stability. Global slope instability, which is typically manifested in a forward rotation failure mode, may ensue if a through-going vertical discontinuity, typically referred to as “tension crack,” transects the slope at the back. The transition from stable to unstable configurations depends on the distance between the tension crack and the toe of the slope. On the basis of the analysis, a simple threefold stability classification—stable, conditionally stable, and unstable—is proposed. In addition, geometrical guidelines, based on standard field mapping data, for the above stability classification are provided. Finally, the optimal reinforcement strategy for overhanging slopes is explored. The stability of overhanging slopes is determined by their eccentricity ratio, defined by the ratio between the base $\left(B\right)$ and top $\left(L\right)$ lengths: $e_r=B∕L$ . It was found that an overhanging slope with eccentricity ratio of $e_r<0.38$ is unstable and requires reinforcement. With an eccentricity ratio between $0.38<e_r<0.62$ , the slope is considered conditionally stable against toppling failure, and reinforcement should be considered if the geometry approaches the lower bound eccentricity of 0.38. A comparison of full and partial face reinforcement schemes showed that full face reinforcement is preferable. The findings of this study were demonstrated by using an illustrative case study in which the stability of a $34\mathrm{m}$ high overhanging slope in a highly discontinuous rock mass was studied and an optimal rock bolt reinforcement scheme was designed.

The effects of aging of high-density polyethylene (HDPE) geomembranes on the diffusion and partitioning of a group of volatile organic compounds (VOCs) are examined. Two different $1.5\mathrm{mm}$ thick HDPE geomembranes were aged in the laboratory at $85°\mathrm{C}$ by immersing in a synthetic leachate for up to $32\text{months}$ . The results of partitioning and diffusion tests performed at room temperature on both unaged and aged geomembranes using a dilute aqueous solution containing four VOCs commonly found in landfill leachates [benzene, toluene, ethylbenzene, and xylenes (BTEX)] are reported. The diffusion and partitioning coefficients decreased with increased aging. The calculated permeation coefficients decreased by 36–62% after aging the geomembrane for about $10–32\text{months}$ . This decrease in diffusion, partitioning, and permeation coefficients is related to the increase in geomembrane crystallinity during aging. A relationship between partitioning, diffusion, and permeation coefficients with the geomembrane crystallinity is established and could potentially be used to evaluate the migration of VOCs through HDPE geomembranes. Aging of HDPE geomembrane did not increase diffusive transport of organic contaminants.

Adequate knowledge of the in situ state of stress can be essential to the analysis of geotechnical systems. However, the measurement and prediction of $k_0$ remain difficult. In particular, limited attention has been given to the evolution of $k_0$ during the formation history of the soil and diagenetic processes such as mineral dissolution. Experimental and numerical results show that grain mass loss due to mineral dissolution produces a pronounced horizontal stress drop under zero lateral strain conditions; the state of stress may reach the active shear failure $k_a$ condition and internal shear planes may develop. While horizontal stress recovery often follows upon further dissolution, marked differences in fabric are observed between the pre and postdissolution soil structures.

The cyclic behavior of 9.5 mm (3/8 in.) minus curbside-collected crushed glass (CG) blended with dredged material (DM), classified as an organic silt by the Unified Soil Classification System, was evaluated using a cyclic triaxial testing program. Tests were performed on 100% CG and 100% DM specimens, and 20/80, 40/60, 60/40, and 80/20 CG-DM blends (dry CG content is reported first). The specimens were compacted to a dry unit weight equivalent to 95% of the maximum dry density based on ASTM D1557. For each material, a minimum of three specimens was tested at cyclic stress ratios of 0.20, 0.35, and 0.45. The DM used in this study exhibited significant plasticity, which would be expected to display cyclic softening behavior according to liquefaction susceptibility criteria proposed by Boulanger and Idriss in 2006. However, the high density of the material resulted in transitional behavior between cyclic mobility and cyclic softening. These findings suggest that as long as the CG, DM, and CG-DM blends are compacted, they should not be susceptible to strength loss or large strain under cyclic loading.

A soil when sheared ultimately reaches a steady-state condition at which it deforms at a constant shear stress, effective normal stress, and void ratio. Various systems in nature dynamically evolve similarly from some initial condition, to a final steady-state condition. Such systems have been studied using dynamical systems theory. This technical note uses this theory to model monotonic shear of soil as a dynamical system. The principle proposed is simple—the rates of change of the shear stress, effective normal stress, and void ratio are proportional to the applied values of the shear and effective normal stress with the proportionality values decaying with strain until ultimately these proportionality values become zero at the steady-state condition. It provides a well-formed qualitative principle that fits closely the stress-strain-void ratio curves of undrained shear tests on uncemented, resedimented clays at various over consolidated ratios (OCRs), and drained shear tests on sands and silts at various relative densities, for various stress paths including compression, extension from standard triaxial, and true-triaxial tests. For the undrained shear of resedimented clay, these proportionalities and their decay rates vary smoothly with OCR. For drained shear of sand and silt, the model parameters show orderly variation with relative density. Its value lies in that a well-formed qualitative principle derived from the steady-state condition provides an alternate approach to current complex elastoplastic models based on critical state theory.

The spatial distribution of hydraulic gradients within a soil sample under cyclic flow with different frequencies was measured and compared with calculated results. The permeability within the soil before and after cyclic flow was compared. This study shows that a lower ratio of consolidation coefficient of the soil to the angular velocity of cyclic flow will result in higher hydraulic gradients at the boundaries, and lower gradients within the soil. This study suggests that besides geometrical parameters of soil and geosynthetics, the consolidation coefficient of the soil and the frequency of cyclic flow need to be considered as well in the design of filters under cyclic flow.