The present paper aims at developing a mathematical theory for prediction of the lateral soil displacement induced by seismic liquefaction. By using the knowledge obtained from shaking table tests together with the principle of the minimum potential energy at force equilibrium, the theory of maximum possible displacement was developed, and is described in detail. Boundary conditions as well as irregular topography were also considered using the principle of the minimum potential energy. Prediction of the increase of displacement with time was also attempted. This aim was achieved by using a scalar function of time, denoted by λ, that stands for the increase of displacement from zero towards the maximum possible one. An equation of motion in terms of λ was derived by using the theory of Lagrangean equation of motion together with the Hamiltonian principle. By combining the maximum possible displacement and λ, it is possible to predict the variation of displacement with time. The proposed method is characterized by its need for a limited number of input data.

View full abstract

The analysis and computation of the ultimate lateral capacity of rigid piles is usually based on a simplified soil pressure distribution along the pile length. Actual soil pressure distributions were measured in a rigid model pile along its length and across the diameter and it was found that the simplified theoretical assumptions of pressure distribution needs a modification. A method is suggested in this paper to predict soil pressure distribution and ultimate lateral capacity for rigid piles in cohesionless soils. Field and laboratory data from published literature are used to validate the proposed method.

View full abstract

This paper introduces a time-dependent radiation damping for the time-domain wave equation analyses of single piles. By deriving the radiation damping ratios of the linear viscoelastic soils along the pile from the dynamic impedance functions (Novak, 1974), time-dependent damping coefficients are obtained from the inverse Fourier Transform of these damping ratios and the static soil spring constants. Numerical examples are studied by solving the wave equations with the finite difference scheme. Applications of the modeling are presented and discussed with emphasis on identifying the primary factors that control the performance of piles. It is important to know that this solution, in contrast to the ordinary frequency-domain approach, can predict the softening load-displacement response of the piles under cyclic loads. The pile response under monotonic loads would be significantly influenced by the proposed damping parameters.

View full abstract

Undisturbed clay samples were taken from alluvial deposits at seven different urban coastal locations in Japan. At several of the locations samples were taken from different depths in order to sample different sedimentary environments. The samples were then triaxially tested in both undisturbed and remoulded states. A testing programme was devised to separately examine the influence of plasticity index, aging factors and overconsolidation ratio on the undrained cyclic strength of clays. The cyclic data for normally consolidated samples was used to define the cyclic strength of the clays as an intrinsic function of the plasticity index I_p. This was modified to take into account the initial vertical yield stress ratio and degree of aging, and then further modified for the effect of geological OCR. A further series of tests was carried out on samples which were anisotropically consolidated by being subjected to an initial drained shear stress for 24 hours. The cyclic strength of the anisotropically consolidated clays at low initial drained shear stress ratios increased with increasing plasticity. This reversed as the drained shear stress ratio increased. It was possible to define a simple function for the cyclic strength normalised by the isotropic cyclic strength against the initial drained shear stress ratio. This function was shown to be dependent on plasticity index I_p but independent of stress history.

View full abstract

Estimation of overall (or average) settlement and differential settlement are the two most critical issues in the design of moderate to large sized raft and piled foundations. Recent studies have shown that a central pile group can largely eliminate differential settlements, but it is still necessary to estimate the average settlement of the combined pile group and raft foundation. This paper presents a simple method of estimating the overall stiffness of piled rafts in a non-homogeneous soil with finite depth, based on the method of Clancy and Randolph (1996). Although the method enables the estimation of the piled raft's stiffness even when piles are installed beneath the full raft area, particular emphasis is placed on the situation where the pile support is limited to the central region of the raft. The method combines the so called 'equivalent pier method'(Poulos and Davis, 1980) and the 'flexibility matrix method'(Randolph, 1983). In the presented method, a piled raft is replaced by a 'capped pier'. Firstly, a simple method of estimating the overall stiffness of stubby piers in a homogeneous or a non-homogeneous soil is presented. The estimated stiffnesses are compared with those calculated by a finite difference approach (FLAC, Itasca Corporation, Version 3.22). The comparison shows that the simple method gives approximate overall stiffness of piers with a wide range of slenderness ratios (length/radius) and pile-soil stiffness ratios. Secondly, the applicability of the equivalent pier method to pile group analysis is examined for homogeneous soil conditions. The calculated stiffnesses are compared with those obtained by FLAC, and the so called 'hybrid'method (HyPR : Clancy, 1993) where the existence of individual piles is considered. The applicability of the 'capped pier'method to piled raft analysis is then examined. The results show the presented simple method gives satisfactorily accurate stiffness of pile groups and piled rafts without any long computations. Finally, a short example of the application is presented for a centrifuge model piled raft. It was found that the presented method gave close overall stiffness to those obtained by observations, and by alternative rigorous calculations.

View full abstract

The experimental results reported in this paper were obtained performing monotonic, cyclic and dynamic torsional shear tests on a stiff Italian clay in the pre-failure stress-strain range. The experimental clearly demonstrate that the small strain values of shear stiffness and damping ratio are significantly affected by strain rate : while the initial shear modulus G₀ continuously increases with strain rate, the equivalent viscous damping ratio D₀ initially decreases with strain rate, then it reaches a minimum constant value and thereafter it increases significantly. The experimental data also demonstrate that, if the strain rate is properly accounted for, a linear behaviour for the soil is no longer distinguishable, even at strain levels as low as 10^-4%. As a whole, these effects could have meaningful practical implications, thus requiring to be accounted for in modelling soil behaviour in those problems where the rate of loading varies significantly.

View full abstract

A mixture of sand and bentonite is seen as an effective component of artificial barriers for disposal facilities of radioactive wastes from nuclear power stations. Therefore, to design and construct these facilities, we must evaluate the swelling characteristics of sand-bentonite mixtures. For this purpose, various laboratory tests on the swelling pressure and swelling deformation of sand-bentonite mixtures were performed. To discuss the process of the swelling characteristics of these mixtures, this study also observed the swelling behavior of bentonite in the mixtures using a scanning electron microscope which can control the temperature and vapor pressure around samples. The following conclusions were obtained through this study. (i) The swelling characteristics of sand-bentonite mixtures is dependent on the dry density and bentonite content of mixtures. (ii) From observation with the scanning electron microscope, we found that the voids of mixtures are filled up by the volume increase when bentonite absorbs water. Furthermore, the swelling process of sand-bentonite mixtures was considered on the basis of the swelling behavior of montmorillonite, a swelling clay mineral. (iii) This study proposes the simplified evaluation of the swelling characteristics of sand-bentonite mixtures using the parameter "Swelling volumetric strain of montmorillonite". This evaluation can obtain the maximum swelling pressure and the maximum swelling strain of sand-bentonite mixtures at various dry densities and bentonite contents. Therefore, this simplified evaluation is available when the disposal facilities of nuclear wastes will be designed and constructed.

View full abstract

Results of an experimental and numerical investigation of shallow foundation displacements behind cantilever sheet pile wall excavations in sand, are presented. A series of model tests have been carried out, in which the effect of sand density, excavation depth, foundation load and the distance between the foundation and the excavation edge on foundation settlement and horizontal displacement were investigated. Based on the model test results and the results of a parametric finite element analysis, shallow foundation displacements have been related to : a) the lateral sheet pile wall movement, b) the ratio of the distance between foundation and wall to the excavation depth, and c) the safety factor against bearing capacity failure of the foundation.

View full abstract

In this study, in order to comprehend the progressive failure of an overconsolidated clay, triaxial compression tests under constant load application for the overconsolidated Kawasaki clay are carried out. These tests show, after a certain period of time, axial strain rate fluctuations with marked peaks until the specimen reaches outright failure. The experimental procedure is then numerically simulated using the soil-water coupled finite element technique. While the numerical simulation for the triaxial compression test under axi-symmetric conditions did not exhibit the same behavior as in the triaxial test (shear band formation and fluctuation of axial strain rate in the axial strain rate vs. axial strain plot), under plane-strain conditions it properly described the fluctuation of axial strain rate as observed in the experiments. Such fluctuations in the axial strain rate are closely related to the progression of the localized shear zones within the specimen. This behavior is attibuted to the "non-local" characteristics of saturated clays. Since the pore water supply to the localized shear zone can not always be sufficient due to Darcy's law which limits the gradient of global seepage head in the soil, the specimen exhibits varying strain rates. Tertiary mudstone pebbles can be idealized as heavily overconsolidated clay. The progressive failure of a heavily overconsolidated clay suggests that the behavior of crushed Tertiary mudstone is due to slaking at the contact points between the pebbles.

View full abstract

A method to estimate the load-settlement relation at the toe of a large diameter pile is presented. Input information is the load-settlement relation of the toe of a small diameter pile obtained by loading test. First, soil parameters and coefficient of earth pressure at rest are determined as a set of plural solutions by back analysis, using finite element method of elastic-plastic soil and genetic algorithm. Since the input information is the load-settlement relation alone, soil parameters and so on can not be determined uniquely. Using the sets, finite element analyses are carried out which calculate load-settlement relations at the toe of a large diameter pile. Since the sets are plural, plural relations are obtained which determine the possible region of the relation. If a loading test for the large diameter pile is carried out, the load-settlement relation at the toe of the pile will coincide with one of the relations in the region. The load test results of large and small diameter concrete blocks which were constructed beneath the pile toes are analyzed. Using the tests result from the small diameter block, the possible region of the relation of the large diameter block is estimated and it is confirmed that the test results of the large diameter block is within the estimated region and the method is effective.

View full abstract

A hollow cylinder torsional shear apparatus was modified and improved so that automatic stress-strain control and measurements were possible. Using this apparatus, a series of undrained plane strain shear tests was conducted on sand, and stress-strain behavior, such as rotation of principal stresses, shifting of the magnitude of intermediate principal stresses, and variation of normal stresses and strains, were closely examined. An attempt was also made to compare the plane strain shear behavior with the undrained shear behavior of sand in simple shear conditions.

View full abstract