A modified sliding block analysis, which allows the sliding resistance to be nonlinear visco‐elastoplastic, has been developed to predict deformations of embankments during earthquakes. Contrary to the assumption (of the original Newmark sliding block analysis) that a distinct slip surface forms, the centrifuge models were observed to deform along relatively wide shear bands during base shaking. The sliding resistance provided by a wide shear band is better characterized by visco‐elastoplastic than rigid‐plastic behavior. A detailed analysis concerning rational parameter selection is included. The modified analysis is used to predict the deformation of clay embankment centrifuge models during simulated earthquakes, with good agreement between experiment and theory. The centrifuge modeling technique is shown to be a useful tool for development, tuning, and verification of analytical models, especially in the subject of earthquake engineering where detailed field evidence is in acute shortage.

A unified, landform‐based probabilistic approach to geotechnical characterization and analysis is outlined for application to ground transportation routes. The approach can explicitly account for uncertainties, geotechnical variability and limitations in available geotechnical data/information. Using Bayesian probability concepts, landform identification and statistical characterization of landform soil property parameters, it can provide site‐specific estimates of soil property parameters and geotechnical behavior variables. The methodology is suited to situations where regional or local landform soil property parameters are or can be statistically well known, but site‐specific data may generally be either unavailable or sparse. Estimates can be based only on landform information or systematically updated and improved with site‐specific data to the extent they are available. Estimates include means, variances, and reliability/probability levels on the estimates, including the probability of any given estimate being above or below a given critical value.

The results of a laboratory investigation of the influences of dynamical driving methods and relative density on the behavior of piles in tension are presented. The study includes 31 bored piles, 12 driven piles, and 5 vibro‐driving pulling tests in dry sand. The sand is layered in a 750 mm ID vat. The testing program uses the same 35 mm OD model pile and the same 90 mm embedded depth. The different driving methods used are the high frequency vibro‐driving and driving methods. Some piles are also bored in sand. The experimental results show that the placement methods have a significant influence on the ultimate pulling resistance. A placement method coefficient is deduced from the tests.

A series of centrifugal model tests to investigate the bearing capacity of a strip footing on normally consolidated Kawasaki clay was performed. The proposed prototype is a normally consolidated clay foundation having undrained shear strength increasing with depth. To ensure that the proposed modeling was performed accordingly, pore pressure and settlement behaviors with time were recorded and compared with computed values obtained by solving the consolidation differential equation. As a further check, water content distribution of the model was obtained after loading test. All the observations gave positive evidences showing that correct modeling had been carried out successfully. The bearing capacities of the foundations were determined by plotting the load intensity settlement curve on double logarithm paper. The bearing capacities thus obtained were then compared with Davis‐Booker's plasticity solution and Nakase's slip circle solution. Adjustment was also made for the 90% consolidation. The results reveal that Davis‐Booker's solution (smooth) formed the lower limit of bearing capacity and that solutions for the bearing capacity of clay, having undrained shear strength increasing with depth, is relevant.

A method for numerically modeling the transient coupled heat and moisture flow in a multi‐dimensional soil domain subjected to time‐dependent heat and moisture generation is presented. The computer program, TRUCHAM, which is based on the Philip and DeVries model for flow in soils, utilizes an integrated finite difference procedure. The technique accommodates different types of linear and nonlinear boundary conditions and material properties. Validation of the program has been made with data from exact solutions, controlled laboratory experiments and field tests. This program will be useful in solving many heat and mass transport problems encountered in the field of underground power cables, heat storage and nuclear waste disposal programs.

A statistical model is proposed to estimate the cyclic shear strength of a saturated sand deposit. Large‐scale shaking table test results are used to develop the basic model since they reproduce the in situ test conditions more closely than other laboratory tests. Due to the unavailability of large‐scale shaking table test results for different soils and test conditions, cyclic triaxial test results are used to modify this basic model. The model is first developed to represent laboratory conditions. This laboratory relationship is then modified for in situ conditions. Factors affecting laboratory relationships that are considered in this study are methods of sample preparation, soil fabric, system compliance, mean grain size, multidirectionality of earthquake motions, and secondary factors. Factors affecting the in situ relationship are previous strain history, age, cementation, etc. It is observed that considerable uncertainty is expected in the estimation of the in situ cyclic shear strength. The uncertainty associated with the in situ relative density contributes significantly to the overall uncertainty. This type of uncertainty needs to be considered in any analysis where the cyclic shear strength of sand is an important parameter, e.g., liquefaction.

Quantitative estimates of the various soil-pile parameters on lateral pile head displacement, rotation and moment coefficients are presented in dimensionless terms through graphs. The results provide guidelines to predict the lateral response of free-head piles. The illustrative example highlights the utility of the results. These results can also be used to predict the responses of partially restrained head piles.

It is indisputable that nonuniformities of stresses and strains exist in soils tested not only in the simple shear apparatus (SSA) but in the triaxial apparatus as well. These nonuniformities alone cannot explain the differences observed from the results of soils tests in the SSA and the triaxial apparatus. Regardless of nonuniformities, the behavior of soils in the SSA and triaxial apparatus should differ from one another because they were designed to impose different stress conditions on the soil samples. This study shows that the often-ignored intermediate principal stresses in the SSA play a significant role in the response of soils tested under plane strain conditions. The stress path followed by sand in the two apparatus differ significantly when the intermediate principal stresses are taken into account.