An analytical model has been developed to predict permanent stresses and strains in sands under cyclic loading for general drainage and boundary conditions. Here the model is combined with the finite element method to predict the permanent displacement of a storm barrier under combined tidal and wave loading. The results are compared with predictions from previous analyses, yielding a good correlation. The effects of drainage conditions and soil densification on permanent displacement are evaluated. The results demonstrate the importance of including in the pore pressure calculation the effects of both the tendency for volume change and the change in the average total stress.

Criteria for estimating the first two probabilistic moments of the normalized parameters of $p‐y$ curves are presented. These parameters are the ultimate soil resistance and the initial tangent value of the soil reaction which are both normalized with respect to the values predicted by an adequate mathematical expression. Due to the scarcity of data a sequential Bayesian formulation of statistical inference is adopted which deals with engineering judgment; small scale models of piles under lateral loads, as well as full scale specimens and indirect information is contained in load‐deflection curves measured at the heads of laterally loaded piles. When assessing the possible deviations of the normalized variables, attention is given to: uncertainty attached to the prediction formula of the $p‐y$ curves; influence of uncertainty in basic soil properties; and discrepancies among results from small scale models and field tests. Mean and variance values of the normalized parameters of $p‐y$ curves corresponding to soft and medium saturated clays are obtained for a particular case.

Laboratory geotechnical data on piston cores from 31 sites on the mid‐Atlantic Upper Continental Slope show the near‐surface sediments vary from normally consolidated to somewhat overconsolidated clayey silts and silty clays of low to high plasticity. They also exhibit normalized behavior and their index property correlations with the effective‐stress friction angle, the undrained strength ratio, and the compression index are reasonably consistent with existing knowledge. Disturbance effects were effectively minimized in the consolidation and strength parameter determinations by using stresses and stress histories in excess of those in situ. However, substantial disturbance effects in preconsolidation stress determinations were found. The preconsolidation stress values derived from the triaxial data are high, and those from the consolidation data are low, compared with the values obtained from the laboratory vane data. Because existing knowledge concerning disturbance effects suggests that in‐situ preconsolidation stress values should lie between those derived from the triaxial and laboratory vane data, the preconsolidation stress values obtained from the consolidation data appear to be appreciably smaller than in‐situ values.

Recently soil arching theory has been extended to the study of forces and stresses exerted by a yielding soil mass against discrete piles embedded in a slope and extending into a firm, non‐yielding base. Many uses of discrete pile retaining wall systems exist in current geotechnical practice. In each case, soil arching has been relied upon to some extent as a means of stabilization without concise information about the required spacing of the stabilization piles. This lack of information requires a design engineer to err on the conservative side, and to place the piles closer together than they need be. This study examines the role of soil arching in discrete pile retaining walls and provides an information base for rational design. The present research objective is to experimentally model soil arching action between piles, to determine the influence of and relationships between the essential parameters, and to compare predictions from several new arching theories to experimental model results. The results of the experimental and theoretical comparisons indicate that discrete piles embedded in a slope into a firm, non‐yielding base, can provide significant additional stability to a slope if conditions for soil arching are met. Files can be installed without significantly decreasing slope stability during construction. For retaining walls that rotate vertically about the base between “fixed” side supports, the data obtained by the modified Bransby and Smith numerical method closely approximates experimental results. The stress transfer ability of loose soil is greater than previously thought. A loose deposit does not preclude the possibility of mobilizing soil arching or side friction. In soldier pile and lagging systems, material costs, strengths, and sizes can be optimized to achieve maximum benefit from soil arching.

Cyclic triaxial tests were performed on a well‐graded railroad ballast with a maximum particle size of 38 mm (1.5 in.). Samples were compacted in a split mold to reproduceable density states and subjected to repeated applications of compressive axial stresses. The cumulative permanent axial strains that developed, for constant‐amplitude loadings, were found to be dependent upon the permanent strain after one cycle of loading, the loading magnitude, and the number of applied loading cycles. Staged tests were conducted in which a sequence of variable loads was applied to the same sample. For these variableamplitude tests, the final permanent strains were independent of the order of the applied loadings. A superposition method is presented to characterize the mixed loading conditions. The results for the variable‐amplitude tests indicated that the maximum load controlled the permanent strain development, with negligible additional strains developing after returning to a lower repeated stress level. Partial unloading cycles were also investigated. These partial unloading cycles caused very little additional cumulative inelastic strains.

An experimental investigation was conducted to determine the change in engineering properties of a collapsible loess in Bulgaria resulting from dynamic compaction. The results of in situ tests conducted before and after compaction are described. A 13,605-kg (15-ton) circular concrete tamper was dropped from a height varying between 5 and 10 m (16 and 33 ft) to produce impact. Density measurements taken before and after compaction and in situ tests using dynamic cone and dilatometer, indicate that the effective depth influenced by the compaction was about 4 m (13 ft). The results suggest that the potential for collapse has been significantly reduced by dynamic compaction.