Model tests and theoretical analysis were done to investigate the arching in embankments on soft ground supported by piles with cap beams and geo‐textiles. The cap‐beam‐and‐geotextile combination may alleviate the uneven surface settlements that sometimes occur in embankments supported by piles with individual square caps. In the model tests, sand was placed on simulated cap beams and soft ground. The loads on the cap beams and soft ground were recorded and compared with theoretical analysis based on equilibrium of semicylindrical sand arches. In a second series of model tests, a geotextile layer was laid on the cap beams and the soft ground prior to placement of sand. Equations and charts were developed to show the parameters affecting the tension in the geotextile and to assess the effect of geotextile on load transfer. Theoretical solution showed reasonable agreement with experimental results for cases with no geotextile but only partial agreement for cases with geotextile. The limitations of the model tests, the factors affecting the results, and alternative interpretations are discussed.

Along the western shores of the Arabian Gulf, saline soils, known as sebkhas, are widely distributed. These soils are associated with many geotechnical problems that emerge principally because of their high salt content and their susceptibility to strength loss and collapse upon saturation. This research program was carried out to assess the performance of sebkha subgrade in the laboratory using conventional and soil‐fabric‐aggregate (SFA) systems under static and dynamic loading conditions. Variables investigated included subbase thickness, level of stress, type of geotextile, and saturation effect. Results indicate that the use of geotextiles significantly enhances the inferior properties of sebkha subgrade, particularly when the SFA system is saturated. The contribution of geotextiles to the performance of SFA systems diminishes when the subbase thickness increases. Results of the index laboratory tests on the small‐scale model have been used by local municipalities for road constructed on sebkha subgrade in the eastern province of Saudi Arabia.

The undrained strength of saturated cohesive soils after cyclic loading decreases and increases depending on whether the soil is initially normally consolidated or overconsolidated and whether drainage is included during or after cyclic loading. The methods for predicting these changes due to both strength degradation and its recovery in saturated cohesive soils after cyclic loading are presented in the current paper. The experimental parameters included in the proposed equations are easily determined as a function of plasticity index. The applicability of the methods is shown by investigating the results from both cyclic triaxial and direct simple shear tests on various kinds of normally consolidated cohesive soils including highly sensitive clays and peat. It is suggested from this consideration that they are valid for the prediction of the undrained strength for cohesive soils after cyclic loading with and without drainage.

The factor of safety for surficial stability is highly dependent on the effective cohesion. This study shows that even when triaxial tests are performed at effective pressures as low as 13.8 kPa (2.0 psi), the results may still provide effective cohesion intercepts that are too high. The reason is that the compacted sandy clay behaves as an overconsolidated soil, with effective stress paths bending to the right with low to negative A‐values. As demonstrated by this case study of a surficial failure, the calculated factor of safety was 1.33 using the extrapolated value of effective cohesion of 5.3 kPa (110 psf). An unconfined submerged triaxial test resulted in a lower effective strength envelope, which provided a factor of safety of 0.97 for the surficial failure. There are two factors that can increase the surficial stability of a compacted clay slope: aging and root reinforcement. For the surficial failure, the effect of aging was minor and root reinforcement was not a factor in this case, because the roots had not penetrated to a significant depth at the time of failure.

A bearing‐capacity analysis of shallow footings and strip foundations subjected to axial, eccentric, and inclined loading is presented for noncohesive soils. The influences of load eccentricity and inclination, subsoil surface inclination, and depth and shape of foundation with regard to classical and new calculation methods are discussed. Comparisons of different formulas of bearing‐capacity factors, shape and depth factors, load and surface inclination factors, as well as values of these factors are presented. A methodology and analysis of model test results of both the writer and other authors are also presented and compared with different calculation methods, which show considerable underestimation of results with respect to classical calculation methods and reasonable agreement for new ones. The results of experiments for strip foundations and footings have been statistically evaluated and empirical formulas to determine $N_{\gamma }$ values proposed. The quantitative data presented can be useful for verification of new bearing‐capacity calculation methods and practical engineering design.

Soil and geosynthetic components are used together in lining and cover systems for waste‐containment facilities to protect the environment by acting as a hydraulic barrier. Since the performance of the barrier depends on the components remaining intact, accurate predictions of tensile loads in the components are important. Lining and cover‐system components must be designed to resist tensile loads anticipated. Four methods for determining tensile loads are examined in this paper. Two common design methods, called the limit equilibrium method (LEM) and the limit method (LM), that consider only force equilibrium to predict tensile loads are compared with two proposed methods that consider both force equilibrium and displacement compatibility. An example multiple‐layered cover system is used to compare predictions from the four methods. The comparison indicates that displacement compatibility is an important consideration and can influence predictions significantly for tensile load. Estimates of tensile load in the geosynthetics using the state of practice may be too great (as for the LM) or too small (as for the LEM), compared with more rigorous displacement compatibility methods.

This paper presents the results of a cover‐subsidence sinkhole evaluation. In this case, the sinkhole activity caused a slow, gradual settlement of the State Road 434/Harbour Isle Way intersection, in Longwood, Florida. The subsoil conditions were explored with numerous cone penetrometer tests and permanent piezometers. The study indicated that the settlement was caused by internal soil erosion and raveling. Piezometric elevation contour maps revealed a well‐defined depression that coincided with the observed surface settlement. The piezometer monitoring results suggest this may be an effective technique for economically identifying potential sinkhole locations.

A finite‐element model capable of computing permanent deformations along preexisting sliding surfaces in dams and embankments is used to evaluate the error in the current state‐of‐practice two‐step procedures that are based on the simplifying assumption that the computation of dynamic response and of the resulting sliding displacements can be decoupled. Both sinusoidal and real earthquake ground motions are used as excitation, and Coulomb's friction law governs sliding along any point of the rigid‐plastic tensionless interface. It is found that only when the dominant excitation frequencies lie in the proximity of the fundamental dam frequency will the permanent deformation from the decoupling assumption exceed appreciably the “single‐step” value. As a validation of the single‐step procedure, an analysis is performed and shown to successfully explain the nonsymmetric response of La Villita Dam recorded in the 1985 Mexico Earthquake.

Practicing foundation engineers tend to consider the foundation load as the point load when they want to consider its influence on the adjacent foundations. They believe that it will yield conservative results when a Boussinesq equation is considered. This derivation shows that such an assumption is groundless. The effect of uniform pressure on the adjacent foundations could be as high as 25% more than its effect when the pressure is lumped as a concentrated load at the center of the footing. The derivation converts the square footing into a shape that fits the polar coordinates. Such conversion facilitates the integration of the uniform pressure effect of one footing on the adjacent footings. Parameters and limits of integration are derived, then used in a practical example.