Seismic design of structures that are long in plan, such as bridges and buried pipelines, requires information on the spatial variation of ground displacements along their lengths. This variation in ground motion results in relative displacement between two points on the ground surface and horizontal strain in the soil. An analysis of empirical data from the 1971 San Fernando Earthquake is presented herein that suggests that relative displacement and strains can be modeled by surface wave propagation. A general procedure for determining those relative displacements and strains for design is suggested and illustrated with an example.

Resonant column tests of five fine‐grained cohesionless soils showed the influence of capillary stresses on values of the low‐amplitude shear modulus, $G_0\text{.}$ This influence was found to be greatest for the soils having the smallest effective grain diameter, $D_{10}\text{,}$ and for the lowest external confining pressure, ${\sigma }_0^{\prime }\text{.}$ The maximum capillary influence developed at degrees of saturation between 5% and 20% for the soils tested. For the soil with the smallest $D_{10}$ $(=0.0024\mathrm{mm})$ the maximum $G_0$ was about two times $G_0\left(\mathrm{dry}\right)$ for ${\sigma }_0^{\prime }=14.2$ psi (98 kN/m). This latter pressure would correspond roughly to 20 ft (6.1 m) of overburden of incompletely saturated soils. Therefore, it would be expected that capillarity would affect in situ shear wave velocity measurements required for the dynamic design of shallow footings, runways, or other structures supported by soil near the ground surface. An empirical expression was developed which permits corrections of the value of $G_0\text{,}$ measured at a particular degree of saturation, to the degree of saturation expected in prototype conditions.

Saturation of sand specimens in laboratory investigations is necessary to provide reliable measurements of volume change and pore pressure. Conventional methods utilize back pressuring or back pressuring with supplemental use of carbon dioxide. If no carbon dioxide is used, back pressures can easily become undesirably large. On the other hand, the percolation of carbon dioxide through the sand to reduce back pressures can lead to unwanted reaction with soil components. To overcome these problems, a new procedure is proposed that utilizes equipment available in almost any geotechnical laboratory. A theoretical and experimental basis is provided as justification. In the new method, an initial vacuum is applied to the sand specimen in such a manner as not to elevate effective stresses. The soil may be dry, saturated, or partially saturated. After the vacuum is held for no more than five minutes, normal back‐pressuring techniques are used. It is found that sands can be effectively saturated at very low values of back pressure. The procedure is especially useful in normally troublesome materials such as cemented sands.

Oil storage tanks built on weak and compressible foundations have suffered catastrophic failures due to foundation instability. These experiences are reviewed and used to establish simple procedures for evaluating factors of safety against base shear failure and edge shear failure. Methods for averaging foundation shear strength values for these two modes of failure are described, which make it possible to apply simple bearing capacity formulas to conditions where the shear strength of the foundation varies with depth, and where the foundation contains or is overlain by layers of sand or gravel. Recommendations concerning minimum values of factor of safety against undrained foundation failure are provided. In cases where the first loading of the foundation is very slow, or where drainage is accelerated by sand drains, wick drains, or natural sand lenses, appreciable drainage and strength gain occurs during water load tests. Methods of rehabilitating tanks after failure of the foundation are reviewed.

A simplified approach for the analysis of the static axial response of pile groups is presented. The approach is based on idealization of the soil around pile shafts as a “Winkler Model for Pile Group” and a transfer matrix approach. The Winkler Model for Pile Group can produce a pile‐soil‐pile interaction and is formulated so that the model can be defined from the Winkler model for a single isolated pile. Excellent agreement is found between the group factors computed by the present approach and by other approaches based on a continuous soil model. Charts are prepared for the construction of the transfer matrix. Using these prepared charts, the flexibility matrix at the heads of the pile group can be computed easily by hand for any complex soil profile.

A simple thin‐layer element is developed and used in a finite element procedure for simulation of various modes of deformanon in dynamic soil‐structure interaction. The constitutive behavior of the interface is defined by decomposing it into its normal and shear components. The soil is modeled as an elastic‐plastic hardening material. The numerical procedure is used to predict behavior of a model structure tested in the field, and the influence of interface behavior on displacements, velocities, and accelerations is delineated.

Pullout capacities of driven H piles under axial tension loads were determined experimentally in stiff clays, dense sands, silts, and stratified soils. The test results are interpreted by 3 different procedures. Measured pullout capacities are compared to predicted pullout capacities based on semi‐empirical approaches. Predicted capacities vary greatly with subsurface conditions of site, effective failure surfaces of piles, pile lengths, and soil shear strength parameters. Earth pressure parameters and adhesion values obtained from the tension load tests are generally found to be consistent with the previously published data.

It is shown for simple shearing that the liquefaction and densification potentials of a cohesionless sample of sand do not necessarily decrease because of preliquefaction or because of large amplitude (drained) prestraining. In fact, in simple shearing, the resistance to liquefaction (undrained) or densification (drained) of a preliquefied sample actually increases, as a result of the concomitant densification, if the preliquefaction is terminated at zero residual shear strain, whereas this resistance is reduced considerably, if the preliquefaction is terminated at zero residual shear stress. The inherent anisotropy associated with dry pluviating and moist tamping sample preparation techniques is shown to affect the sample response to consolidation and to simple cyclic shearing. It is found that, within each cycle of simple shearing, the induced anisotropy is essentially wiped out in the neighborhood of the zero shear strain (but not at zero shear stress), and the anisotropy that remains at this state basically stems from the sample preparation technique. With the aid of the dilatancy‐fabric relation proposed by Nemat‐Nasser for simple shearing, the volumetric strain of the sample is estimated using a simple distribution function for illustration purposes, and the results are compared with the experimental data.

An analysis for investigating undrained creep of a viscoelastic‐plastic soil mass is briefly described. Any form of undrained creep response can be considered, while failure is incorporated through a Mohr‐Coulomb failure criterion. This analysis is applied to two published field studies of embankments on soft clay, the East Atchafalaya Levee Test Section 3 and the M.I.T. Test Embankment. In each case, it is demonstrated that the variation with time of vertical and horizontal movements in the underlying soil, can be reasonably well‐predicted.

Modulus of deformation, $E_{\mathrm{sa}}\text{,}$ of sandy soils is estimated from load tests of drilled piers assuming that only the active depth lying directly under the pier base undergoes compression. The active depth is defined as the depth of the layer in which the increment of stress produced by the transmitted load is greater than a certain given fraction of the effective geostatic stress. Instrumented load tests yield the conclusion that the modulus, $E_{\mathrm{sa}}\text{,}$ is equal to the secant modulus of deformation $E_0$ as obtained from triaxial tests. The proposed method makes it possible to predict the settlement of piers at limit load using laboratory moduli determined from samples extracted from the layer underlying the pier base. An example illustrates how the settlement of axially loaded drilled piers bearing on sand can be calculated by the proposed method using the linear elastic theory as outlined by Poulos. The variance of the estimated modulus is indicated. It appears that lengthy and costly load tests of drilled piers could be avoided if the proposed method is applied.

To control the earthquake‐induced pore pressures in earth dams with pervious gravelly shells, a method is presented to design chimney drains to be installed in the upstream submerged part of the embankment. It has been developed using the theoretical work of Seed on the determination of the pore pressure buildup in a material in which simultaneous generation and dissipation take place, the experimental work of Lee and Albaisa on the reconsolidation volume change from an induced excess pore pressure condition and Terzaghi's theory of consolidation. An example taken from an actual case is presented to explain the different steps as well as conclusions on the characteristics of the materials to which the proposed solution might be successfully applied.