A geologic anomaly is any inclusion that is of a different material type or has different physical properties from that included or normally expected in a design profile. The performance of a geotechnical system could be significantly affected if the anomaly has undesirable properties. However, the presence of such an anomaly and its properties can never be ascertained. Methods for updating the anomaly location and size distributions and the probability of anomaly presence based on the engineer's experience and judgment, supplemented with the encounter history from exploratory borings performed at the site, are formulated. The effectiveness of using an extensive boring program in limiting the likelihood of anomaly presence and, if present, in limiting the anomaly size and location, is found to depend on the overall probability of anomaly detection by the exploration scheme, and prior estimates of anomaly properties and anomaly presence probability. The updated anomaly location and size distributions provide valuable input data for a reliability evaluation of the geotechnical system.

The problems of steady seepage into long double‐walled cofferdams in a pervious stratum of finite depth were presented by King and Cockroft in 1972 for the cases of no excavation and full excavation inside the cofferdam. In 1992, Banerjee and Muleshkov presented the details of the exact solutions of the complete problem, including the special cases for which solutions were developed by King and Cockroft in 1972. These solutions have been derived in terms of elliptic integrals of the first and third kind. The equations depicting the rate of seepage, the distribution of hydraulic heads on the various segments of the domain boundaries, and the exit gradients at the base of the excavation were used to prepare the design charts presented in this paper. An example problem illustrating how these design charts can be used to facilitate computation of the required quantities including seepage rates factors of safety against basal heave, and piping have also been included.

Tests were performed on a clayey soil from a site in Texas to define ranges of water content and dry unit weight at which compacted test specimens would have: (1) Low hydraulic conductivity; (2) minimal potential for shrinkage upon drying; and (3) adequate shear strength. Test specimens were compacted with three compactive energies over a range of water content. Low hydraulic conductivity could be achieved over a broad range of water content, but relatively wet specimens underwent large shrinkage upon drying. A range of water content near the optimum value measured with the highest compactive energy proved to be suitable in meeting the objective of low hydraulic conductivity and shrinkage potential. The dry unit weight had to be greater than 96–98% of the maximum value from modified compaction (ASTM method D1557) to meet hydraulic conductivity, shrinkage, and strength objectives. A similar approach is suggested for development of compaction criteria at other projects in which low‐hydraulic‐conductivity liners or covers must be constructed in relatively arid regions or other situations in which potential cracking caused by desiccation is of concern.

Despite modern advances in technology, civil engineering failures continue to occur. Most investigations identify the technical causes and financial liability. This paper looks at where and why the engineering process (planning, design, construction, and operation) breaks down. The data base is largely from the writer's unpublished failure investigations. Approximately 500 failures are evaluated. Of these, 58% originated in design, 38% in construction, and 4% in operation. Approximately half of the problems occurring during construction originated in design; the other half in construction. Three causes are identified: absence, ignorance, and rejection of current technology. Absence accounts for 12%, ignorance for 33%, and rejection for 55%. Of the total, 88% reflect human shortcomings that can be reduced by acknowledging professional limitations, continuing education, modifying design and construction systems, and resisting the unbalanced pressures that thwart good engineering.

A new method for reinforcing soils has been developed that involves the incorporation of multioriented geosynthetic inclusions within or upon a compacted or naturally deposited soil. The results of an extensive laboratory experimental study consisting of California bearing ratio (CBR), triaxial, and permeability tests are presented. These results indicate that substantial improvements in strength and stiffness of granular soils can be obtained using this method of reinforcement. A comparison is made of the effectiveness of multioriented elements and fibers. The potential uses for multioriented elements as lightweight, highly porous artificial soil are also discussed.

A large specimen of compacted clay (diameter = 298 mm; thickness = 914 mm) was subjected to freeze‐thaw in the field for 60 days. Afterward, the hydraulic conductivity was measured. The hydraulic conductivity of the entire specimen remained essentially unchanged, but increases in hydraulic conductivity of 1.5–2 orders of magnitude were observed above the freezing plane. The increase in hydraulic conductivity was highest at the top of the specimen and decreased with depth. Changes in hydraulic conductivity also occurred at depths 150 mm below the freezing plane, where desiccation occurred because of water redistribution, Numerous horizontal and vertical cracks formed in the soil mass. Dissection of the sample after permeation revealed that the cracks were laden with water. Cracking was greatest at the surface and became less frequent with depth. For depths greater than 150 mm below the freezing plane, cracking was absent. The frequency of cracks is consistent with principles of mechanistic models of soil freezing. The results of laboratory tests were used to predict the hydraulic conductivity of the large specimen. Tests were conducted on specimens subjected to various freeze‐thaw cycles, temperature gradients, and states of stress. It was found that the predicted hydraulic conductivities were lower than those measured on the large specimen, but they closely resembled the trend in hydraulic conductivity with depth.

This paper describes the results of an experimental investigation of capillary effects on the low strain shear modulus of sands in a partially saturated condition. It studies: (1) Void ratio; (2) confining pressure; (3) grain shape; and (4) grain size distribution on shear modulus. The results of resonant column tests show that capillary stresses significantly increase the shear modulus of partially saturated sands. These effects are more pronounced for soils with low void ratios and confining pressures. The optimum degree of saturation increases with increasing void ratio and a linear relationship exists between the maximum shear modulus ratio and void ratio. The slope of this relationship is not affected by confining pressure and grain size distribution, and instead depends on the soil grain shape. If the smallest grains of a sand are larger than the 400 sieve size, the grain size distribution does not affect the optimum degree of saturation, but only affects the maximum shear modulus in a partially saturated condition. The content of the minus 400 sieve size fraction can affect both the maximum shear modulus and the optimum degree of saturation for unsaturated sands. Soil grain shape also affects both the maximum shear modulus and the optimum degree of saturation.

In Southern California the conventional method for evaluating surficial stability has a bias towards cohesion. It suggests that surficial failures are more likely in slopes made of sandy and gravelly soils, than those composed of clays and silts. This contradicts actual observations. A study of the equations governing seepage and rainfall data in Southern California suggests that permeability plays an important role in the surficial stability of slopes. In homogeneous slopes this study finds that soils with a low permeability are more prone to develop the conditions leading to a surficial failure mechanism. Hence, clayey and silty soils slopes are more prone to develop surficial instability than slopes made of sandy or gravelly soils. This study also finds that soils with a permeability greater than a certain limiting value, klim, will not become saturated. It is found that the Unified Soil Classification System is useful in evaluating the likelihood that a surficial failure mechanism with seepage parallel to the face of the slope may occur.

Postearthquake deformation analyses of the lower San Fernando dam were conducted using an incremental finite‐element method. In the analyses, an undrained elastoplastic model was used to simulate the collapse of liquefied materials. The model is developed based on the critical‐state boundary‐surface theory, the concept of steady‐state strength, and the undrained behavior of liquefiable soils. A triggering condition in terms of a collapse surface was considered in this model. The hyperbolic strain‐softening relationship has been introduced to simulate the postpeak behavior of liquefied materials. The analyses have shown that a progressive failure under undrained conditions may explain the observed response of the lower San Fernando dam following the 1971 earthquake. Stress redistribution initiated by the strain softening of liquefied materials is the main reason for undrained flow failures of dams, slopes, and foundations and can occur in a short period ranging from a few seconds to a few minutes. The liquefied zone after stress redistribution may be much larger than the initial liquefied zone caused directly by an earthquake. Therefore, a postearthquake deformation analysis may be essential in liquefaction stability evaluations.

Currently, helical anchors are being extensively used in the construction of transmission tower foundations, pipelines, and braced excavations. To arrive at the uplift capacity of these anchors, it is necessary to assume certain type of failure surface and in a few cases, cylindrical failure surface has been suggested. However, in this type of formulations the effect of spacing of helical plates is to be given due consideration and in the present study, this aspect has been taken into account. Pullout tests are conducted on four type of model anchors fabricated with different number of helical plates and spacing ratios in soft marine clay. From the test results, it is found that continuous cylindrical failure surfaces are formed in the case of anchors with helical plates spaced at a spacing ratio of 1.5. But in the case of anchors with helical plates at higher spacing ratios, well‐defined cylindrical failure surfaces could not be observed and in these cases; empirical formulations are suggested to estimate the capacities. The suggested formulations are verified with the reported lab and field test results and are found to be in good agreement.

A mathematical model for a combined soil subgrade plus structurallement in contact with the subgrade (e.g., a mat foundation) is presented. The structural portion of the model is a conventional flexural element. The subgrade portion of the model is Pasternak's hypothesis, which is fundamentally more accurate than the commonly used Winkler hypothesis. The novel element developed in this note is not the individual model components, but the fact that the combined model can be visualized as a spring‐supported beam‐column under constant axial tension. The column tension and springs represent the subgrade effects. This beam‐column analogy is useful in practice because it allows more accurate modeling of soil‐structure interaction within the capabilities of existing structural analysis computer software.

Proper selection of bedding material before construction can prevent loss of money and time due to change orders and sometimes obviate the need for the common remedial practice of overexcavation and replacement with costly granular materials. The bearing capacity and intrusion characteristics of two different types of granular materials, gravel and volcanic clinker, commonly used in northern Arizona as granular bedding materials in pipeline construction, were studied under controlled laboratory conditions. It was determined that the void or area ratio influences the amount of bedding intrusion that occurs in low strength subgrades. Granular bedding materials, such as gravels, were generally found to intrude a low strength subgrade less than coarser bedding materials with relatively higher void or area ratios, such as volcanic clinker.

An analysis based on a simplified form of the boundary element method is used to analyze the interaction between a piled raft and a soil subjected to consolidation or expansive movements. The analysis allows for such factors as a nonhomogenous soil profile, pile‐soil slip, failure of the soil at the pile tip, and failure of the soil below raft elements. Via a relatively simple example, it is demonstrated that additional pile forces and movements are developed in the piled raft, as compared with the corresponding free‐standing pile group. In consolidating soils, raft‐soil contact can be lost and the load is then carried by the piles alone. In swelling soils, uplift pressures are developed on the raft, and tensile forces are induced in the piles. The use of piled rafts under these circumstances should generally be avoided.