Journal of Geotechnical and Geoenvironmental Engineering

An empirical, energy-based methodology for liquefaction hazard assessment and microzonation mapping is presented. The approach is probabilistic, considers the uncertainty in the liquefaction criterion, and is applicable to most earthquake-induced liquefaction analyses. The examples illustrated are for water-saturated sands at level ground. The energy of ground shaking is estimated from the Fourier amplitude spectra of the incident waves. The susceptible materials are characterized only by their corrected standard penetration test value $\overline{N}$ and overburden pressure σ₀. Illustrative microzonation maps of liquefaction opportunity are shown for the Los Angeles metropolitan area. Two types of maps are presented, one showing the average return period of liquefaction occurrence (for given $\overline{N}$ and σ₀), and another one showing distribution of $\overline{N}$ with equal probability to liquefy during 50 years exposure (for given σ₀). An advantage of the method is that the result is given directly in terms of in situ albeit simple soil characteristics rather than in terms of laboratory tests and peak acceleration. Possible applications of the computed opportunity maps are discussed.

Investigation programs are intended to reduce uncertainty in site conditions that impact design performance. However, existing theoretical tools for assessing uncertainty reduction, such as Bayes' theorem, have seen limited practical use because cumbersome numerical solutions are usually required. A first-order, second-moment Bayesian method (FSBM) that overcomes this limitation is described in this paper. FSBM is a practical, analytical approximation to Bayes' theorem that helps estimate the reduction in uncertainty achieved for a particular investigation outcome (inverse analysis) and the reduction expected to result for proposed programs (design or forward analysis). In this initial application, FSBM is formulated and applied to updating uncertainty in the geometry of a single subsurface feature for 1D search programs. Uncertainty in the existence of the feature is also addressed. FSBM provides solutions for example cases that are generally within 10% of solutions obtained by Monte Carlo simulation. The method is intended and shows promise for more general applications such as the design of 3D investigation programs or building probabilistic models of 3D subsurface features.

Unsaturated drainage layers (UDLs) have been demonstrated to greatly increase the lateral diversion capacity of capillary barriers. The inclusion of a UDL allows native soils suitable for vegetation growth to be used as the finer soil as lateral drainage properties of the layer no longer need to be considered. A comprehensive numerical study was conducted to investigate the influence of the interface slope and the UDL material on the system's ability to laterally divert downward moving moisture. A capillary barrier system with and without a UDL was simulated for 10 years using daily varying climatic data for three locations in the United States. Three different sands were simulated as the UDL and were modeled at slopes of 5, 10, and 20%. The numerical results confirm that the inclusion of an unsaturated drainage layer at the fine/coarse interface of a capillary barrier can provide significant improvements in the performance of the cover system by laterally draining water. This improvement in performance may allow the system to be successfully implemented in climates wetter than previously were thought suitable. The diversion length (the distance water is diverted laterally with no downward flow through the fine/coarse interface) of a capillary barrier with a UDL was found to be proportional to the slope of the fine/coarse interface. In addition, a relationship between lateral diversion lengths in a capillary barrier and the UDL material was developed and found to be dependent on the unsaturated flow characteristics of the UDL. These relationships allow the performance of a variety capillary barrier UDL designs to be calculated knowing the behavior of one system for a given location.

Model tests were performed at the Geotechnical Centrifuge Facility of Delft University of Technology, The Netherlands, to examine the mechanics of light nonaqueous phase liquid (LNAPL) movement in a partially saturated porous granular medium. The experiment simulated a 2D spill of LNAPL in an unsaturated sand prepared at two values of porosity. The duration of the centrifuge model tests corresponded to a prototype equivalent of 110 days. The choice of modeling a 2D flow together with the use of a transparent container enabled direct visual observation of the experiments. Scaling laws developed in connection with other centrifuge modeling studies were used to support the test results. Tests were conducted at two different centrifuge accelerations to verify, by means of the “modeling of models technique,” the similitude between the different experiments. The paper presents details of the experimental methodologies and the measuring techniques used to evaluate the final distribution of water and LNAPL content in the soils.

A laboratory testing program was conducted to identify and interpret the fundamental factors affecting the performance of a new microchip-laser based fluorescence sensor in soil and ground water. Investigations were performed using a versatile experimental apparatus designed to simulate the in situ interface between the laser induced fluorescence (LIF) sensor and contaminated media while providing complete control of test conditions. Attempts were made to isolate the effects of soil properties such as grain size, soil type, color, mineralogy, and organic content on in situ LIF observations. Test results indicate that soil has no measurable effect on the determination of pore fluid fluorescence lifetimes or the general form of pore fluid emission wavelength profiles. However, for a given contaminant concentration in the pore space of a soil with a narrow grain size distribution, decreases in soil grain size are accompanied by a decrease in the magnitude and variability of observed LIF signals. For soils containing a wide range of particle sizes, in-soil LIF observations are primarily influenced by pore space geometry in relation to the smallest particles present in the soil. These trends were found to be a primary function of the volume of pore fluid in a soil specimen that is in the direct path of laser excitation energy. Soil organic content and optical characteristics such as reflectivity also have potential influence on in-soil LIF observations, although on a secondary basis. After reviewing these experimental results, which indicate the relative impact of soil properties on fluorescence observations, a simple geometric model is presented that captures the primary effects of soil on pore fluid fluorescence observations.

A major cause for massive collapse of coal mines in the Illinois Basin is the softening of mine floors. Certain materials present in mine floors soften after coal extraction resulting in bearing capacity failure of coal pillars. The softening mechanisms considered in this study were (1) slaking/swelling due to moisture exposure; and (2) creep or strain softening due to sustained loads. The effect of floor slaking or swelling on floor stability can be fairly dramatic. In some cases, shortly after pooling of water on the mine floor, collapse occurs resulting in surface subsidence. In other cases, the time factor is greater. The effect of softening on the ultimate bearing capacity of the mine floor was assessed by modeling representative conditions using FEM and elastoplastic elements. In the models, the properties of softened mine floors were determined from results from subsurface exploration work in floor areas that were stable and others that had failed. The zone of softening was evaluated by stress-field analysis.

A unique configuration of horizontal sheet-like electrodes was used in the field at a site in Ohio that was underlain by silty clay glacial drift to induce electroosmotic flow and to characterize the effects of electroosmosis on soil properties (e.g., electrical conductivity and pH). The lower electrode was created at a depth of 2.2 m by filling a flat-lying hydraulic fracture with granular graphite, and the upper one was a metallic mesh placed at a depth of 0.4 m and covered with sand. The electrodes were attached to a DC power supply, creating an electrical gradient of 20–31 V/m and a current of 42–57 A within approximately 20 m³ of soil. Total energy applied was 5,500 kW⋅h during approximate 4 months of operation. Electroosmotic flow rates of 0.6–0.8 L/h were observed during tests lasting several weeks, although total flow rate (electroosmotic plus hydraulic) was strongly influenced by fluctuations of the ground-water table. The ratio of applied current to voltage decreased from 0.9 to 0.6 A/V and was mainly due to a decrease in electrical conductivity of the soil. A low pH front developed at the anode and migrated toward the cathode. The velocity of the pH front per unit voltage gradient was 0.014 (cm/day)/(V/m). This was 40 times slower than what has been reported from laboratory experiments using kaolinite as a medium. These results confirm the feasibility of using horizontal electrodes at shallow depths, but they also underscore some important differences between the geochemical effects observed during field tests in natural soils and those seen in laboratory tests using ideal materials.