The research reported is concerned with the load transfer behavior of rock‐socketed bored piles. Several large‐diameter instrumented piles were load tested to at least twice their working/design loads. The piles were observed to behave in an elastic manner when tested to normal working loads. Much of the applied loads was transferred through the pile shaft especially when the socket length was in excess of twice the pile diameter, while relatively little load was transferred to the pile base. The observed mobilized unit shaft friction values compared well with those reported by other researchers. Instrument readings continued to be monitored throughout the superstructure construction period. It has been found that static load tests carried out within a short term showed significantly different load distribution along the pile shaft compared with that in service. The differences in the load transfer behavior may be attributed to the creep between pile body and soil/rock, group interaction, and the presence of pile cap.

A new test to measure the low‐stress dilation behavior of cohesionless soils subjected to plane strain has been developed. Very large strains (30–35%) have been applied to sand specimens while approximately maintaining the parellelopiped shape of the specimen. A small version of the apparatus designed to test sands is described. However, it should be relatively easy to construct larger devices with the potential for testing rockfill‐like materials. A model approximating the effect of initial density on dilation behavior is presented that suggests a method for defining and measuring index void ratios that are intrinsically related to dilation behavior. The ultimate constant volumetric strain void ratio for low‐stress plane strain was found to vary with the initial void ratio of the specimen. The effect of fabric anisotropy on dilation behavior has been studied and modeled. A dilation test on sand‐cement shows that contact cohesion increases the rate of dilation of a loose sand to the rate expected for an uncemented dense sand.

Problems associated with isolation, measurement, and modeling the elastic behavior of particulate materials are discussed. Because of pervasive yielding, exclusively elastic behavior of particulate materials is restricted to infinitesimal increments of unloading. Three‐dimensional elastic constitutive equations are formulated in terms of two scalar functions representing effects of void ratio and stress history, a reference fabric matrix, a stress‐compliance matrix, and a Poisson's ratio matrix. The stress‐compliance matrix accounts for the discovery in 1979 that elastic shear stiffness is independent of stress normal to the plane of shear. Transformation of coordinates and the definition of reference fabric are carefully considered in the developed formulation.

This paper describes a boundary element analysis for analyzing the behavior of a group of end‐bearing piles subjected to negative friction. It is found that group effects lead to a reduction in downdrag force in a pile, as compared with a single isolated pile, and that inner piles in a group develop less downdrag force than the outer piles. Among the factors influencing the distribution of downdrag force in the group are the number of piles, the relative stiffness of the piles, the distribution of soil modulus with depth, and the presence of pile‐soil slip. A significantly larger soil surface settlement is required to mobilize full pile‐soil slip along all piles in the group than along a single pile. When applied to a published field test, the theory shows reasonable agreement with the measured behavior.

Probabilistic risk analysis is a useful technique in assessment of dam safety issues, and also has important applications in the early phases of dam design when design options are being compared and exploration strategies planned. At this stage, few data are usually available and judgment‐based decisions are most important. An application of probabilistic risk analysis to design of a dike in karst terrain is described. Assessments based on geologic information, observations, and subsurface investigations are expressed using a simple probability model. Options considered to reduce the likelihood or consequences of sinkhole‐induced failure include a secondary containment structure, a warning system, and additional exploration. These options are evaluated using an expected monetary value criterion. It is shown that simple probabilistic formulations consistent with preliminary levels of information reduced major sinkhole uncertainties into more readily evaluated components. These techniques identified the most cost‐effective design strategy for reducing the risk of sinkhole‐induced dike failure and provided an effective vehicle for communicating geotechnical judgments, risks, and alternatives to the project owner.

This paper develops an analysis for the behavior of a single pile or symmetrical pile group in sand under cyclic axial loading. The analysis allows consideration of the effects of degradation of skin friction, base resistance and soil modulus, the influence of loading rate, and the accumulation of permanent displacements. Laboratory data for skin friction degradation are presented and it is shown that the amount of degradation due to cyclic loading depends on the cyclic displacement, the number of cycles, the soil type, and the type of pile. Theoretical solutions for a hypothetical pile and a recent field test emphasize the importance of determining whether the degradation of skin friction depends on the absolute cyclic displacement or the cyclic displacement relative to the pile diameter.

The lateral deflections of sheetpile or diaphragm walls are often used to evaluate the performance of braced excavations. This paper examines the influence of a number of factors affecting the lateral deflections of strutted or anchored sheet‐pile walls in clay using the finite element method, such as the undrained shear strength, the width and depth of excavation, the wall stiffness, the penetration depth, and the depth to hard stratum beneath the bottom of the excavation. A simple procedure to estimate the lateral deflection is proposed. It has been assumed that the walls are flexible and the lateral deflections are governed by the plastic yielding of the soil below the bottom of excavation. The yielding is inversely proportional to the factor of safety with respect to basal heave. A nonlinear hyperbolic stress‐strain relationship has been used in the analysis. Deflections computed by this method compare favorably with field records and published results.

The relationship between settlements and lateral deflections of the walls of floating‐roof steel tanks, and methods that can be used to estimate lateral wall deflections, are the main focus of this paper. As the walls of floating‐roof steel tanks settle unevenly, the upper parts of the walls deflect inward or outward, producing what has been termed “ovalization” of the tank. Ovalization leads to problems with floating roofs. Inward deflections may cause the roof to bind. Outward deflections may cause loss of seal and resulting emission problems. The amount of inward or outward deflection of the walls of floating‐roof tanks can be estimated readily once the basic shape has been recognized and the deformation axes, called fold lines, have been located. A simple procedure for recognizing fold lines and for calculating lateral deflections of walls is explained. Comparisons with other methods of analysis and with field measurements show that the fold‐line method is accurate enough for practical purposes, and that it provides a clear understanding of tank deformations.

In the case of a mixed Spatial and Material description of Large-Deformation Consolidation, it is shown that if the geometry is updated through an expression for the velocity of solids, then the complete Lagrangian description is obtained.