The feasibility of using neural networks to model the complex relationship between the seismic and soil parameters, and the liquefaction potential has been investigated. Neural‐networks are information‐processing systems whose architectures essentially mimic the biological system of the brain. A simple back‐propagation neural‐network algorithm was used. The neural networks were trained using actual field records. The performance of the neural‐network models improved as more input variables are provided. The model consisting of eight input variables was the most successful. These variables are: the standard penetration test (SPT) value, the fines content, the mean grain size $D_{50}$, the equivalent dynamic shear stress ${\tau }_{\mathrm{av}}/\sigma {\prime }_0$, the total stress ${\sigma }_0$, the effective stress $\sigma {\prime }_0$, the earthquake magnitude $M$, and the maximum horizontal acceleration at ground surface. The most important input parameters have been identified as the SPT and fines content of the soil. Comparisons indicate that the neural‐network model is more reliable than the conventional dynamic stress method.

Medium‐ and large‐size model drilled shafts up to 174 mm in diameter were constructed in laboratory‐prepared cohesive soil deposits. The models were subjected to cyclic axial loading, and the resulting displacements and changes in capacity were determined, while the in‐situ stresses at the foundation/soil interface were measured. The applied loading was designed to simulate realistic windstorm events, which are an important source of cyclic loading for foundations. Shaft response was found to depend critically on the type of cyclic loading. One‐way uplift, in which the direction of loading is always upward, caused large permanent displacements at high load levels but only small displacements at moderate load levels, while capacity reductions were modest. Two‐way loading, in which the direction of loading reverses twice every cycle, caused minimal response at low load levels but a sudden degradation in displacement response at moderate load levels, with an associated substantial reduction in capacity. Critical levels of repeated load, below which shaft response remained stable, were established; and the mechanisms responsible for capacity reduction are discussed.

Setup and relaxation were encountered during testing and subsequent installation of piled foundations in a glacial deposit of clean sand. Displacement piles driven into this medium dense deposit show time‐dependent increases in load capacity of from 40–80% as determined by loading tests and the results of wave equation analyses made using input obtained from dynamic measurements. Evidence is cited to show that setup results from a loss of strength due to disruption of the structure of the sand during pile driving, followed by a recovery in strength as the soil structure heals at constant effective stress. Relaxation occurred in large groups of piles where the cumulative effects of driving piles into the ground brought the sand to a dense state, creating conditions where subsequent pile driving caused sands in the failure zone to dilate during shearing displacements. Relaxation is attributed to dissipation of the resulting negative pore pressures and relief of high lateral ground stresses within the dilated sand.

A variational limiting equilibrium approach that is an extension of Baker and Garber's analysis is presented. In this extension a pseudostatic seismic force is included to assess the stability of simple slopes. The normal stress distribution over the potential slip surface is determined mathematically so as to render the minimum factor of safety while satisfying explicitly all limiting equilibrium equations. The equation describing this distribution contains terms related to seismicity and it makes the problem statically determinate without a priori statical assumptions. A closed‐form solution is then obtained for a log spiral failure mechanism. Consequently, a parametric study of the effects of pseudostatic seismic force is conducted with ease. Furthermore, design charts for assessing the seismic stability of simple slopes are presented. For nonseismic cases these charts coincide with Taylor's chart. For the seismic cases, identical results could also be obtained by using a conventional log spiral analysis (though with much greater effort). The design charts, therefore, should be acceptable by a large group of readers.

The effectiveness of treatment methods for collapsible soils was evaluated under field conditions using six full‐scale load tests performed on 1.5 m square footings. Treatments included conventional methods such as prewetting with water and partial replacement with compacted fill along with more innovative procedures such as prewetting with a sodium silicate solution and dynamic compaction under dry and wet conditions. A load test was also performed on untreated soil for comparison purposes. Soil improvement was evaluated using double oedometer testing on “undisturbed” samples along with cone penetration tests and pressuremeter tests. Settlements were monitored as a function of moisture penetration and compared with predictions based on oedometer testing. Settlement of the footing on untreated soil exceeded 400 mm and differential settlement was substantial. Prewetting prior to loading did not significantly improve footing performance. Use of a compacted fill mat delayed the onset of settlement and reduced differential movement, but settlement eventually exceeded 100 mm. Although the sodium silicate and dynamic compaction methods were more expensive than conventional procedures, they prevented settlements from exceeding 25 mm.

Liquefaction has been a source of major damage during severe earthquakes. During the past two decades researchers have attempted to determine the parameters that could better define the liquefaction potential of a soil deposit. The use of the energy concept to define the liquefaction potential is validated through laboratory tests on sand specimens. Numerous tests were conducted at several confining pressures, relative densities, and strain amplitudes using a torsional shear, hollow cylinder device; and the energy per unit volume required for liquefaction was measured. Relationships were developed between the energy per unit volume and the various parameters that were varied during the tests. The dissipated‐energy‐per‐unit‐volume approach is very promising and can accommodate nonuniform loading present in earthquake spectra.

The dynamic response of rigid foundations supported on a homogeneously nonlinear elastoplastic half‐space is investigated in the context of nonlinear finite‐element (FE) analysis. The constitutive equation for the half‐space is described by a bounding surface plasticity model with a vanishing elastic region. This model is based on a total stress formulation in which the plastic modulus is determined from an interpolated exponential hardening rule. The incompressibility constraint is handled via the mean dilatational formulation of the B‐method. The computational problem associated with the disparity between the longitudinal and shear wave velocities is addressed in the context of radiation boundary condition. The relative significance on the foundation response of the more rapidly propagating P waves, as they reflect spuriously from the FE boundaries, is investigated. It is shown that the reflected P waves have no significant influence on the dynamic response of finite‐size foundations; nor do they have any significant effect on the response of strip foundations vibrating in the rocking mode. Furthermore, it is shown that nonlinear effects can have a significant influence on the dynamic response of vibrating foundations over a wider range of excitation frequencies than originally thought. The numerical simulations presented in this paper are performed using data on real soils determined from resonant column, cyclic triaxial, and other one‐dimensional “radial” loading tests.

The discrete‐element method has successfully been applied in the past to study constitutive behavior of granular materials. The method is applied in this paper to cohesive soils, where in addition to mechanical forces, there are physicochemical forces between particles, and there is significant bending of individual particles. Based on a recent study on the double‐layer repulsive force between inclined particles, the repulsive force is modeled approximately. The attractive force between particles is not considered in the present study. Dividing each particle into a number of interconnected discrete elements, bending of particles is simulated as part of the overall analysis. By employing suitable force‐displacement laws for the mechanical contacts, formation of new contacts, deletion of existing contacts, and slip between particles are modeled. The preliminary results presented here show the potential of the method for carrying out fundamental research into the stress‐strain behavior of cohesive soils.

An approximate elastic perfectly plastic solution in a closed form for axially loaded piles based on the $t‐z$ curve method, is presented. The closed‐form solution allows prediction in a very simple way of vertical pile movements and settlement at the pile head as well as the value of the axial load at a certain depth from the ground surface. Differently from elastic solutions, the method can take into account the nonlinear behavior of the load‐settlement curve. Because of the approximation made to derive simple expressions in a closed form, the method does not apply to situations where a significant portion of the working load is derived from end‐bearing capacity. However, when the analysis is restricted to design loads and the choice of soil parameters is appropriate, the solution presented can give reasonable results. This is confirmed by comparisons with data from full‐scale pile tests taken from the literature.

The availability of wick drains has rapidly led to their replacing conventional sand drains as a means of accelerating the consolidation of soft soils. Wick drains have an oblong shape, but the available theories of radial consolidation have been derived for drains having a circular shape. What is needed to apply these theories to the design of field installations of wick drains is a method of calculating the equivalent diameter of an oblong shape. Previous investigators have suggested several equations for computing the equivalent diameter of wick drains, but they yield different results. The equivalent diameter of oblong drains is determined with an analog field plotter. The results of all methods are reviewed and then compared to published experimental values of equivalent diameters of wick drains. These values were determined from laboratory consolidation tests. Values of equivalent drain diameters backfigured from field data are discussed with respect to the influence of smear.