The method of Sherard, et al., where soil saturation extract cation data are plotted on a percent sodium versus TDS (total dissolved solids in milliequivalents per liter) graph to classify clayey soils as dispersive or nondispersive, is applied to 223 soil samples from six different soil groups. The method exhibited poor overall agreement with physical test results. These data were subsequently evaluated using analysis of variance and discriminant analysis (a multivariate classification technique) which indicated that each soil group exhibits unique classification criteria based on pore‐water cation data. It appears that, for most soil groups in this study, pore‐water cation data do not provide adequate identification of dispersive and nondispersive clay soils. Reasons for the observed lack of agreement between pore‐water data and physical tests are reviewed and suggestions for further research are presented.

A stress hybrid finite element procedure is developed for nonlinear, elastic and elastic‐plastic, analysis of soil‐structure interaction including simulation of construction sequences. The procedure is compared with a number of closed‐form solutions and field problems, and provides satisfactory and improved evaluation of stresses and deformations as compared to the displacement procedure.

The dynamic behavior of cross anisotropic clays was often studied in a resonant column device which is limited to strains smaller then ${10}^{-3}\text{,}$ and to high frequencies. Here the experimental study is extended to large strains at low frequency; and an attempt is made to model the behavior covering the whole spectrum of strains from small to large. Axial stresses, torsional stresses and their combinations when applied to thin long hollow cylinders lead to strains that show the influence of tridimensional states of stress on the dynamic response. The modulus degradation pattern is noticeably affected by those combined stresses. The inclinations of principal stresses play an important part in the way a clay responds to cyclic loading. Relations are suggested to predict shear moduli and damping ratios. The Ramberg‐Osgood‐Masing hypothesis is shown to be deficient and reduction factors are proposed to bring its prediction in line with experimental data.

This investigation concerns the use of conventional free earth support method as applied to anchored bulkheads in sands and to the development of a simplified pressure diagram for the purposes of design. The development of a computational pressure diagram (CPD) and range of values for the coefficients necessary to use the CPD are summarized. The use of the CPD to simplify the design procedure is explained. The suggested design procedure is found to be simple to use and shows good agreement with conventional methods. The method results in a safe penetration depth, and reduced free earth support moment. The suggested design procedure will be very useful in many preliminary design phases.

Two series of K_o tests were carried out on sand using the triaxial apparatus and lateral strain indicator. Jaóky's formula for the coefficient of earth pressure at rest K_o^NC (normally consolidated soils) was affirmed for dense samples only. The initial hydrostatic consolidation of samples decreased the dispersion of test results but increased the measured K_o^NC- value. Consequently, the K_o-strain path must be followed right from the beginning of testing. A deformation theory of K_o^NC is developed using Rowe's stress-dilitancy relation and oedometrically measured total and reversible axial strain increments. For overconsolidated samples, a power-law dependence of K_o^OC on the overconsolidation ratio OCR was found valid for a limited range of OCR. For very high OCR values, K_o^OC decreases. Maximum K_o^OC value corresponds to the Rankine coefficient of passive earth pressure. Initial porosity of samples is not the only structural parameter governing the behavior of samples.

A new procedure is proposed to evaluate in situ relative density using SPT values, and to quantify the uncertainty associated with that prediction. The relationship is first developed under laboratory conditions, then corrected for in situ conditions. The relationship is compared with commonly used relationships, namely the Gibbs and Holtz, Bazaraa, and Waterways Experiment Station. The predictability of the proposed relationship is better than the others when the relationships are compared to field measurements where all the relevant information is available. However, the uncertainty associated with all the predictive models is considerable.

Both theoretical analysis and laboratory test data indicate that for normally consolidated soft clays the coefficient of lateral earth pressure at rest, K_o, gradually increases with time. Physical reasonining suggests that for all soils K_o should asymptotically approach unity. This reasoning implies that for overconsolidated soils with initial values of K_o greater than one, K_o should decrease with time. The time dependence of K_o has significant implications with respect to how the value of K_o measured in laboratory tests relates to in-situ values of K_o.

An arbitrary depth is assumed for the sheet pile. The lateral forces on both sides of the pile are equalized graphically which satisfies one equation of equilibrium. The moment equation is not satisfied which produces residual moment. Such residual moment is calculated for three different depths and plotted against the assumed depths. The right depth is obtained at zero residual moment from the graph. This method is much simpler than the conventional method and yields results which are just as accurate. It has even a greater appeal in the assumption of nonlinearity of soil lateral pressure, for which none of the known methods—Conventional Method, Blume Method, Free Earth Support Method and Fixed Earth Support Method—offer any solution.

Pore water in marine sediments contains dissolved salts. For marine clays and biogenous oozes which have high void ratios, the dissolved salts should be taken into account in determining soil void ratio, water content, dry density, specific gravity, porosity and degree of saturation. This paper presents phase relations for marine sediments and provides correct equations for various index properties. It also provides equations for different relations taking into account the necessary corrections for dissolved salts in marine soils.

The ratio of ultimate undrained residual strength to consolidation pressure for San Francisco Bay mud is a constant in the manner that the ratio of peak undrained strength to consolidation pressure is a constant. Laboratory and field vane shear tests were used to study the variation of peak and ultimate residual undrained shear strength with water content for normally consolidated Bay mud. The average ratio of ultimate undrained residual strength to peak undrained strength determined from the laboratory vane tests was 0.42, and did not vary with water content. Based on a constant ratio of peak undrained strength to consolidation pressure equal to 0.32, the ratio of ultimate undrained residual strength to consolidation pressure is equal to 0.13. The values from the field vane tests were somewhat different, but the relationship was similar. A constant ratio of ultimate undrained residual strength to consolidation pressure for a cohesive soil allows a simple prediction of the ultimate undrained residual strength over a wide range in consolidation pressure from only a few tests. The ultimate undrained residual strength of a cohesive soil can provide an important lower bound in assessing the strength loss that accompanies landslide movement.

The stresses created within an elastic model material (gelatin) by a laterally loaded, rigid, finite pile are studied using three-dimensional photoelasticity. Cases of both slip and no-slip conditions at the pile/gelatin interface are investigated. Horizontal normal stresses near the pile surface are determined for both cases and compared to determine the importance of shear stress transfer. These results are also compared to the results of an elastic half-space model which neglects shear stress.