In the earthquake‐resistant design of structures such as nuclear power plants and above‐ground tanks, consideration is given to the effect of base uplift on the response of these structures. This paper presents an iterative procedure for analyzing the rocking response of a rigid base allowed to uplift. The base is placed on an elastic half‐space and is subjected to a gravitational force and an overturning moment. The proposed method is applied first to a rectangular base under static loading conditions, and the results are compared with those of a conventional method. Then, the method is applied to the problem of dynamic loading conditions. The effects of lift‐off on the real and imaginary parts of the rocking compliance are examined. A large increase is produced in the real part. However, the imaginary part is only slightly increased by the effect of uplift. Further, the dynamic effect on the minimum contact area of the base is negligible.

Free‐field limit analysis of a homogeneous layer of dry soil excited by uniform horizontal and vertical acceleration components leads to simple equations describing various stages of inertial shear fluidization. Orientations of potential shear flow spread out from the initial slip planes developed at moderate acceleration ratios, less than 0.3, until the general fluidization state is reached, when a broad band of planes are mobilized throughout the layer. In these directions, the layer then behaves as a viscous fluid. For loose saturated soils, it is postulated that initial fluidization will initiate liquefaction, and therefore the free‐field solution for this neutral case may serve as a useful benchmark for estimating liquefaction trigger accelerations. All intermediate stages, both active and passive, from static conditions to general dynamic fluidization are expressed. The solution seems relatively insensitive to boundary conditions, since the free‐field lateral pressures derived do not differ significantly from the classic solution for dynamic active or passive pressures on sliding retaining walls. Finally, both initial and general fluidization of a dry sand layer are demonstrated by shaking table tests of a circular footing and a submerged buoyant box.

A probabilistic model for analysis of the pullout capacity for a tension pile installed in clay is formulated, based on conventional methods for prediction of the axial capacity of a pile, the pile forms a part of the foundation for an offshore tension leg platform. Static as well as cyclic pullout of the pile is considered, the latter being based on the strain accumulation method. The model accounts for uncertainties in the loading as well as in the soil strength and deformation properties. An example pile in clay is considered, and its reliability against static and cyclic pullout is computed, based on first‐order reliability methods. A comparison is carried out between the reliability solutions for static and cyclic pullout of the pile, respectively, and an investigation into the effect of modeling a flexible pile rather than a rigid pile is made. An interpretation of the most important uncertainty sources is given, and a sensitivity study is included.

Mechanistic pavement design procedures based on elastic layer theory require the specification of elastic moduli for each material in the pavement section. Repeated load tests yielding a resilient modulus are frequently used to characterize the soil subgrade. Due to difficulties associated with cyclic testing, approximate methods are often used for design estimates of resilient modulus. These approximations are often based only on shear strength measures and do not account for the dependence on the magnitude of cyclic deviator stress. A procedure is described to relate the soil‐index properties and the moduli obtained from uncon‐fined compression tests, to resilient modulus. Two statistical models are described and demonstrated for 11 soils from throughout the state of Tennessee. One model provides an estimation of the breakpoint resilient modulus, or the modulus at a deviator stress of 6 psi (41 kPa). The second model provides a general nonlinear relationship for the modulus of fine‐grained soils as a function of deviator stress. Both models are demonstrated for a range of soils and are shown to provide a good characterization of the response for the soils investigated. Similar relationships can be developed for other subgrade soils, and may prove useful to agencies that use deterministic pavement design procedures, but lack the capability for high‐production repeated‐load testing.

A new experimental apparatus has been developed to investigate the transient and long‐term pore‐water pressure response of saturated soils subject to compressional stress‐wave loading. The laboratory apparatus is capable of generating compressive shock pulses on the order of 35,000 kPa having submillisecond rise times to peak stress. The experimental investigation was conducted to examine and establish an understanding of compressionally induced liquefaction in saturated soils. Specimens of water‐saturated Monterey No. 0/30 sand were tested at various relative densities and effective stresses. Boundary conditions were for one‐dimensional, confined compressive loadings without drainage. Results indicate that it is possible to liquefy Monterey No. 0/30 sand under these conditions. Significant residual pore‐water pressure increases were observed even at high relative densities and high effective stresses. The data suggest that a transient strain exceeding approximately 0.005% may result in explosively induced residual pore‐water pressure increases. Several empirical models are presented for estimating liquefaction potential as a function of initial relative density, initial effective stress, and the magnitude of the applied compressive strain.

The liquefaction behavior of a clean, fine, uniform sand designated F‐70 Ottawa banding sand is examined. Specifically, an investigation into the effects of loading systems and sample preparation methods on the stress‐strain behavior is made. A comparison between the results obtained with a dead‐load apparatus and a hydraulic loading system on samples prepared by dry pluviation, moist tamping, and moist vibration indicates differences in the soil's response during undrained shear. The former loading system generates and maintains over a greater strain range lower residual deviator stresses and peak pore‐water pressures during steady‐state flow. Furthermore, the steady‐state velocities of the samples tested in the dead‐load apparatus are two‐three times higher than the velocities of those loaded hydraulically. These differences can be attributed to the dead‐load system's ability to provide a constant driving force fast enough to achieve a sample's maximum rate of strain once liquefaction commences. The relative position of the steady‐state curves is unaffected by the type of loading system. However, the method of sample preparation dictates their position. Samples prepared by dry pluviation exhibit residual effective minor principal stresses that are approximately one‐quarter to one‐half the value of samples prepared by either moist method.

The concrete membrane face rockfill dam, in spite of its increasing popularity with construction engineers, is perhaps one structure where the design is still largely empirical and is based on past experience rather than theory. Studies were carried out to understand the prototype behavior of this type of rockfill dam and to identify the various factors influencing the development of stresses and strains in the face slab and the rockfill embankment. Different types of elements are used to simulate the prototype behavior of various components of the dam in the sequential nonlinear analysis and some representative results are presented. The behavior of rockfill embankment on application of reservoir water load in case of face dam is compared with that for the case of central earth core dam. The effect of valley abutment slopes on the movement of rockfill and consequent development of stresses in the embankment in cross‐valley transverse direction is brought out. Creep deformations in the rockfill occurring after filling of reservoir are found to be a major factor governing the development of stresses and deformations in the upstream face membrane.

In this study, one‐dimensional consolidation of fine‐grained soils is considered. By taking into account the results of the infiltration theory, a new approach, which is applicable to soils in either saturated or unsaturated conditions, is given. After definition and review of the relevant quantities, the general differential equation of one‐dimensional consolidation is derived on the principle of equating the work done by the internal and external forces. The resulting equation is a nonlinear partial differential equation, from which the classical linear equation of consolidation can be obtained by assuming the coefficient of consolidation to be a constant. In order to verify the theoretical results of the proposed approach, an experimental study was also performed, and the findings were compared with those given by the classical theory. The sorptivity, which is a rather new definition in soil mechanics, suddenly increases when the primary consolidation ends, and this phenomenon provides an opportunity to determine the coefficient of consolidation otherwise computed by the log‐time method.

Limit equilibrium analysis of slope stability is comprised of two coupled problems: kinematical (i.e., find the critical slip surface) and statical (i.e., assure the existence of global equilibrium at the defined limit state). Determination of the general‐shaped critical slip surface can be attained by available optimization techniques where the minimal factor of safety is sought. However, since the problem is statically indeterminate, assumptions are necessary. The available general methods suggests the wide variety of possible statical assumptions. Consequently, there might be a question whether the critical results are indeed critical. The presented method, which is a generalization of Baker and Garber's approach, attempts to avoid statical assumptions by using a variational technique to minimize the safety factor. The variational analysis results in an ordinary differential equation that describes the normal stress distribution over a slip surface. Solving this equation numerically, and substituting the resulting stress into the global limiting equilibrium equations for the sliding body, one can determine the factor of safety corresponding to the user's specified slip surface. The general‐shaped slip surface is left to be varied, as done in other rigorous limit equilibrium methods, until a feasible surface that renders the absolute minimum factor of safety is located. To enable the application of the presented generalized method, a numerical scheme is proposed. The solutions to three example problems, involving complex slopes, soil profiles, and total and effective stress analyses, are presented in detail. These examples provide insight as to the method's performance.

Centrifuge model tests have been carried out under controlled laboratory conditions to simulate the installation of jack‐up spud can‐type foundations into beds of uniform clay having undrained shear strengths in the range of 12–40 kPa. Much of the previously published field data relating to removal of objects from the seabed covers sites with ill‐determined strength profiles but generally rather lower strengths. Results are presented that indicate that the underbase adhesion necessary to sustain substantial breakout resistance is not obtained until a compressive bearing pressure approaching that required for a substantial penetration, i.e., surface failure, has been previously applied. Once compressive failure is approached then breakout forces increase towards the compression capacity but rarely achieve the theoretical upper limit of equality between tension and compression under undrained conditions.