Journal of Geotechnical and Geoenvironmental Engineering

This paper presents a simple four-parameter model that can represent the shear modulus factors and damping coefficients for a granular soil subjected to horizontal shear stresses imposed by vertically propagating shear waves. The input parameters are functions of the confining pressure and density and have been derived from a generalized effective stress soil model referred to as MIT-S1. The predicted shear moduli and damping factors are in excellent agreement with high quality resonant column test data on remolded sands and confining pressures ranging from 30 kPa to 1.8 MPa. The proposed model has been implemented in a frequency domain computer code. By simulating the variations in stiffness and damping with confining pressure, the proposed model provides a more realistic simulation of ground amplification that does not filter out high frequency components of the base excitation.

Relationships between strut loads, earth pressures, temperatures, and the measurements provided by strain gauges are presented in this paper. A braced excavation up to 20-m deep, 9–20 m wide, and >650-m long constructed in competent glacially derived sand, silt, and clay soils (including glacial till) provided a significant amount of data for analysis. The excavation was supported by soldier piles and lagging with pipe struts and was covered with decking during construction. A direct correlation between incremental changes in strut load and temperature was observed during the course of the project. The few existing relationships between strut load and temperature were reexamined and were found to produce back-calculated elastic modulus values that were either without comparison or inconsistent with data from field tests and published sources. The relationships derived as a result of this work are supported by limited case-history data from other published sources and are consistent with practical application of elastic deformation concepts and published soil modulus values.

Field observations indicate that failure in soft rock is often associated with a slip surface or shear band, where deformation is concentrated in a narrow zone; displacements occur with decreasing stress within the shear band, whereas outside the band the material appears to be intact. In examining the propagation of the shear band, it is useful to establish the relation between shear stress and slip displacement. This was accomplished within a laboratory setting with a plane-strain compression apparatus, developed to study localized failure under controlled conditions. Tests on a soft rock, a sandstone with a uniaxial compressive strength of 10 MPa and a modulus of 2 GPa, were conducted to estimate fracture energy G_IIC, a quantity used to evaluate energy dissipation of the failure process. G_IIC was found to decrease by a factor of 3 when considering the actual displacements, rather than assuming tangential displacement only, that is, no displacement normal to the shear band. The experiments showed that the shear band was not completely formed until after peak strength and that sliding along the band during softening was associated with compaction; residual behavior exhibited virtually no volume change. The shear strength at peak stress was nonlinearly related to the normal stress, but the shear strength at the residual state displayed a linear relationship. For normal stresses less than the uniaxial strength, those typical of civil engineering practice, the response can be described as cohesion softening, with friction remaining constant in going from the peak to the residual stress states.

This paper presents a practical method that can be used to evaluate the settlement of pile groups. This method is developed based on variational solutions for pile groups embedded in soil modeled using a load-transfer curve model and elastic half-space model. Using this practical method, the settlement of pile groups can be estimated quickly without recourse to complex computer programs based on numerical models, resulting in savings in time and cost. The solutions obtained using this approach are in good agreement with rigorous variational solutions and are also in reasonable agreement with field measurements.

This paper presents results from a study of the dynamic response of pile foundations in liquefying sand during seismic loading. The study included a series of dynamic centrifuge tests of pile-supported structures and the back-calculation of time histories for the lateral resistance p and relative displacement y between a pile and the free-field soil. Details of the centrifuge experiments and the procedures used to back-calculate p and y time histories are described. The back-calculated p-y time histories provide a concise representation of the experimental results and can be compared to the equivalent p-y behavior predicted by soil-pile interaction analysis methods. The observed p-y behavior provides insight into the mechanisms of soil-pile interaction in liquefying sand, showing characteristics that are consistent with the undrained cyclic loading behavior of saturated sand, including the effects of relative density, cyclic degradation, pore-pressure generation, prior displacement (strain) history, and phase transformation behavior.

The mechanical behavior of a compacted metastable-structured residual gneiss soil was experimentally investigated. Volume changes were investigated using both a conventional oedometer cell and a triaxial permeameter system where the stress state variables were independently controlled. Wetting stress paths were utilized to reflect field conditions associated with collapsing earth structures. The compacted specimens were consolidated under both isotropic and k₀-oedometer conditions. Measurements of total volume change and water content change were made at specified matric suction values following a wetting stress path under a given loading. The experimental data were analyzed to define volume change constitutive relationships for the metastable-structured soil. Best-fit modeling was used for predicting the volume change behavior of the compacted metastable-structured residual soil during wetting-induced collapse. The measured data are discussed from both phenomenological and microscopic viewpoints. Predictions of the best-fit models are compared to experimental results.

A new three-dimensional (3D) slope stability analysis method is developed based on two-directional moment equilibrium. This method calculates not only the safety factor but also the possible direction of sliding for semispherical and composite failure surfaces. As a result, the possible errors associated with assuming a plane of symmetry in 3D stability analyses are eliminated. Another advantage of the new method is to eliminate the tedious work on the coordinate transformation prior to the analysis. Two examples of symmetrical failure surfaces are used to verify the basic formulation in the present study. Three additional examples further demonstrate the applicability of the proposed method in analyzing 3D asymmetrical failure surfaces. An analysis on a slope, subject to asymmetrical excavation unloading and geological conditions, shows that using the method of one-directional moment equilibrium may give an overestimated safety factor of the slope.

This paper presents the results of 254 drained triaxial compression tests. The results are analyzed with the aim of obtaining a phenomenological description of secant Young's modulus variation according to different test conditions. Using empirical fitting equations, the influences of different factors such as axial strain, vertical consolidation stress, consolidation stress ratio, and stress history on the secant Young's modulus under axisymmetric (triaxial) loading conditions at small and intermediate axial strains are singled out and presented. The main purpose of this paper is to identify the factors affecting the nonlinear secant stiffness of sand and to quantify their relevance at different levels of strain.