Yielding is an important feature of the stress‐strain behavior of lightly overconsolidated clays. This paper describes tests on 36‐mm triaxial samples of natural Mexico City clay. The specimens are carefully trimmed, consolidated in increments along various stress paths to determine the yield‐point characteristic curve of Mexico City clay. Once yielding is clearly defined, the specimens are subjected to an undrained compression at a constant strain rate. The clay exhibits marked changes in stiffness when it yields, and, thus, a well‐defined yield envelope in stress space is obtained. This yield curve, together with other curves from the literature, gave information for natural clays with friction angles in normally consolidated range between 17.5° and 45°, allowing a generalization of the concept. The coefficient of earth pressure at rest in the normally consolidated range is also discussed. Data obtained on Mexico City clay show that the relationship proposed elsewhere is also valid for clays having a high friction angle.

Previous work on the seismic behavior of retaining walls is briefly reviewed, and in particular, the Mononobe‐Okabe analysis for the limiting passive resistance is discussed. A test series is described that considers passive failure of a wall rotating about an 'anchor' at the top. Passive failure is induced by subjecting a small test wall to a roughly constant horizontal force, then shaking it with discrete pulses. Development of failure surfaces is observed, and forces, displacements, and acceleration distributions are recorded for each pulse. It is concluded that use of the sliding‐block model based on the Mononobe‐Okabe passive equations is justified for analysis and design, but that in practice, because of the large toe forces involved, migration of the active and passive resultants and reduced anchor resistance, a tied‐back wall would be more likely to suffer an anchorage failure with translation and rotation about the bottom than with rotation about the top.

Deformation, strength, and flow parameters were determined by inverse analyses on geotechnical data obtained from two full‐scale test embankments constructed on improved soft Bangkok clay. One site was improved with prefabricated vertical drains (PVD), and an adjacent site was improved with compacted granular piles (CGP). In the PVD site, the ratio of $c_h$ (field) to $c_h$ (lab) ranged from 1.5‐4, and the $c_h$ value was $4.90m^2/\mathrm{year}\text{.}$ The value of $k_h/k_s$ was about 10, with $d_s/d_w$ of 2.5. In the CGP site, the ratio of $c_v$ (field) to $c_v$ (lab) ranged from 1.3—12. The smeared zone was considered by reduction to CGP diameter by one‐fifth. The ratio of $E_u$ to $S_u$ was determined to be 150. The observed final settlements in the CGP site was 30‐35% lower than those in the PVD site. Using back‐analyzed soil parameters, the settlements of the test embankments were predicted well. Both ground improvement schemes are effective in improving the soft Bangkok clay.

An experimental estimation of the chloride diffusion coefficient and the corresponding tortuosity factor for a saturated, intact mudstone is described. Laboratory tests simulating one‐dimensional diffusive transport are performed by placing distilled water directly above Bison mudstone specimens having high chloride porewater concentrations. Chloride and other species naturally occurring in the pore water are then permitted to diffuse out of the sample and into the distilled water reservoir for a period of up to 34 days. At the end of the test, the sample is sectioned, and the chloride pore water concentration profile is measured. Fickian diffusion theory is then used to deduce the chloride diffusion coefficient. The diffusion coefficient at a temperature of 10° C ranges from 1.5 to $2.0\times {10}^{-6}{\mathrm{cm}}^2/s\text{,}$ from which a tortuosity factor ranging from 0.15 to 0.20 can be calculated. For comparison, an attempt is made to obtain the diffusion coefficient for bromide diffusing into the sample simultaneous with chloride diffusing out. It is found, however, that the concentration profile obtained for bromide cannot be fitted by the diffusion theory, due to interactions between bromide and other species naturally occurring in the rock sample.

Both laboratory and field pullout tests are conducted using steel grid reinforcements with cohesive‐frictional backfill soils. The laboratory pullout tests are performed using a large‐scale pullout apparatus designed especially for this study. The field pullout tests are performed on the dummy welded‐wire reinforcements embedded in a full‐scale reinforced test wall/embankment system that utilize three different locally available, low‐quality, cohesive‐frictional backfill soils, namely, clayey sand, lateritic soil, and weathered clay, in the three sections along its length. It is observed that the magnitudes of the mobilized field pullout resistances as well as the strains induced in the reinforcing elements are strongly influenced by the response of the wall/embankment system to the subsoil movements and the resulting arching effects due to the presence of the inextensible reinforcements. Meanwhile, the laboratory pullout test results generally seem to provide a conservative approximation of the field pullout resistances of the grid reinforcements in cohesive‐frictional backfills. Using the finite element method to model the laboratory pullout tests, the analytical results are observed to agree fairly well with the experimental results.

All existing closed‐form solutions for sheet‐pile analysis and design are based on the assumption of material homogeneity below the dredge line. Unfortunately, the composition of natural deposits is often quite complex, and it is common for a cohesive foundation soil to possess an undrained shear strength that varies continuously, or in a stepwise fashion, with depth. Thus, a closed‐form soution is derived to permit total stress (ƛ = 0) analyses of cantilever sheetpiling in such inhomogeneous foundation profiles. The governing equations associated with multiple foundation strata are found to be similar in form to those associated with a single foundation layer. A computer program is developed to solve the generalized equations for the required piling penetration depth and the maximum bending moment. The new analytical technique is verified by performing manual equilibrium checks on a variety of computer‐generated solutions. The analytical technique is also used to establish appropriate piling penetration depths for several inhomogeneous foundation profiles that have been described in prior geotechnical literature.

A compacted clay is subjected to five cycles of freezing and thawing at constant water content. The test specimens are then permeated in flexible‐wall permeameters at a relatively low effective stress of 35 kPa (5 psi) to determine the effect of freeze‐thaw on hydraulic conductivity. For soil compacted dry of optimum, hydraulic conductivity after freezing is two to six times larger than before freezing. For soil compacted wet of optimum, hydraulic conductivity increases approximately 100‐fold. Susceptibility to freeze‐thaw damage is found to be insensitive to compactive effort; soils compacted with modified compactive effort are no more resistant to freeze‐thaw damage than soils compacted with standard effort. Freezing changes the void ratio, structure, and fabric of the soil, which in turn produces increases in hydraulic conductivity. These findings are consistent with other information reported in the literature and indicate that compacted clay subjected to low effective confining stress should not be allowed to freeze if low hydraulic conductivity must be maintained.

Calculations of the bearing capacity of cohesive soil under embankments are presented for cases of strength increase with depth and a strong crust at the surface. The analysis makes it possible to account for the arbitrary horizontal component of the load generated by the horizontal thrust in the embankment. The proposed analysis allows the effective calculation of critical heights of embankments. A proposal including embankment reinforcement is also presented. Calculations confirm that the largest increase in the critical height of an embankment due to reinforcement is achieved when the strength of the foundation soil increases with depth. A strong crust at the surface of the foundation soil significantly reduces the benefit of reinforcement, and, for embankments with low height‐to‐width ratio, the reinforcement may not lead to an increase in the critical height (or increase in the safety factor for a given height) at all.