An analytical measure was developed so that the permanent displacement/deformation caused by seismic subsoil liquefaction can be predicted. Since the effects of topography on displacement is important, a three-dimensional analysis was developed which can fully consider the topography. Moreover, the loss of resistance in the surface unliquefied crust undergoing a tensile stress was taken into account. The empirical interpolation of the lateral displacement along the vertical axis helped reduce the amount of computation. Consequently, a finite-element two-dimensional analysis which is equivalent to a three-dimensional calculation was developed. The displacement obtained by the proposed method is the maximum possible to minimize the potential energy of the subsoil. By comparing the calculated displacement with the observation, the predicted displacement was to some extent greater than the observation. Hence, a dynamic analysis in the time domain, which was also developed by the authors, is necessary for a better matching between prediction and observation. In contrast, the direction of displacement was correctly predicted by the present analysis.

View full abstract

A two-phase mixture model for predicting the stress-strain relationship of mixtures with different materials is proposed based on consideration of the stress distribution and strain energy in the mixture. Elastic moduli of mixtures with two different elastic materials are estimated by the proposed model and its validity is confirmed by comparing with the calculated results from other theories and the experimental results. In-situ cement-treated soils are regarded as a kind of two-phase mixture consisting of improved and unimproved parts, and the proposed model is applied to such in-situ treated soils. A level of mixing of the treated soils is evaluated using the improvement rate obtained from the improved area in a cross section of sampling specimen. Deformation modulus and unconfined compressive strength of the in-situ treated soil with different improvement rates are derived based on the two-phase mixture model. In order to confirm the validity of the proposed equations, laboratory model tests simulating the actual mixing process at site were conducted and the calculated results were compared with the test results. The proposed equation for estimating the in-situ strength of the treated soils was also verified by the results of the in-situ improvement tests.

View full abstract

This work is concerned with the mechanical properties of freshly-deposited cohesive sediments. It was previously established by the authors that the undrained shear strength in the ultra-low stress range is effective stress dependent and increases with bed age. This paper consists of an experimental program to quantify the strength gain with time, and to study the effect of stress history on the strength of the sediment bed. The experimental program was conducted on fully consolidated Boston Blue Clay beds of various ages, 3.5 to 11 cm in thickness deposited from concentrated slurries. The bed age varied between 2 and 23 days, and the undrained shear strength measured with the Automated Fall Cone Device varied between 10 and 120 Pa for effective stress values below 400 Pa. For a given water content, the strength exhibited a two-fold increase over the 21 day period attributed to thixotropy. This thixotropic strength gain was mainly a function of the existing effective stress with some dependence on water content. Although no measurable swelling occurred upon unloading (unloaded effective stress up to 190 Pa), the shear strength decreased substantially depending mainly on the magnitude of the unloaded and the remaining effective stresses, but also on the existing water content.

View full abstract

The paper analyses the mechanical behaviour of a silty sand, as it was measured in laboratory tests on samples prepared with different density and moisture contents. After a description of the physical characteristics of the material, the methods used to prepare compacted and slurry-consolidated soil samples are reported, followed by a description of the apparatus and experimental procedures adopted. The compressibility characteristics were analysed by isotropic tests at ordinary stress levels (p'<1 MPa) and by oedometer tests at stress levels up to 40 MPa. The shear behaviour was investigated by means of triaxial compression tests following conventional undrained or drained paths with variable directions ; also, the initial shear stiffness was measured by resonant column tests. The basic behaviour of the slurry-consolidated silty sand appeared consistent with the theoretical framework of Critical State Soil Mechanics ; this soil could therefore be used as a reference material to which the mechanical properties of the compacted samples were compared throughout the paper. Two volumetric state loci were taken as reference curves to investigate the behaviour of the compacted material : (i) the isotropic compression line of the remoulded soil, and (ii) the critical state line, which proved to be independent of the soil preparation procedure. Whatever the preparation technique, the mechanical behaviour of the silty sand under triaxial compression was found to be stress-path dependent from the very beginning of the stress-strain curve. The net effect of the compaction could be assessed by comparing the stress-strain behaviour of dynamically compacted soil to that of remoulded material brought to comparable density through a mechanical overconsolidation process. The stress-strain behaviour of the compacted silty sand proved to be more fragile and showed lower initial stiffness than the remoulded soil. It was therefore observed that, even though the compaction may be considerably beneficial to compression behaviour, it does not necessarily optimise the stress-strain response under shear loads.

View full abstract

To apply the general flow theory by Barker to the conventional in-situ hydraulic tests, the type curves for the constant head and constant rate tests for a general condition are derived with the semi-analytical method. The results of the real in-situ hydraulic tests are examined by the type curves. By arranging the results and considering the different test results, it is found that the hydraulic characteristics can be systematically inferred. As the second subject, whether or not the conventional continuous and discontinuous numerical models can represent the flow behavior such that is observed at the in-situ test is examined. The flow of a single borehole test simulated through the various numerical models is tested by the type curve. The effect of the anisotropy and heterogeneity of the model on the dimension and the hydraulic conductivity obtained by matching with the type curve is tested, and the relation of the variance of the dimension and hydraulic conductivity to the models is examined. By using the examination results, the proper available models for the site of interest here is discussed. The process presented in this paper will be very helpful in selecting the numerical model for a real problem because it is seen from the results that some models show a distinctive tendency in the dimension.

View full abstract

The mechanism of the damage to a gravity type quay wall caused by liquefaction of the backfill ground during an earthquake is made clear through a shaking table test and theoretical examination. A series of model shaking table tests was conducted focusing on the occurrence of liquefaction in the backfill ground. The movement of the caisson is found to be quite different depending on whether liquefaction occurred in the backfill ground or not. The fluctuating earth pressure on the caisson suppresses the movement of the caisson when liquefaction does not occur. On the other hand, sliding of the caisson is enhanced since the fluctuating component of earth pressure and the inertial force coincide in phase angle when liquefaction occurs in the backfill ground. When liquefaction occurs, observed earth pressure agrees with that evaluated by Westergaard's formula originally derived for the water pressure on the dam. The fluctuating earth pressure acting on the back wall of the caisson in the process to liquefaction was carefully observed in the model shaking tests. It was found that the amplitude of the earth pressure first decreased to a very small value because of the reduction of the stiffness of the backfill due to the excess pore water pressure generation, and then increased because the phase angle of the earth pressure changed 180 degrees. This indicates that stability criteria of the caisson should be developed not by the onset of the liquefaction but by the sudden phase change. This feature is demonstrated by the simplified mass-spring-dashpot model proposed by the authors.

View full abstract

Relationship between allophane content and index properties for volcanic cohesive soil were obtained by statistical treatment of about 120 sample's data. The main results are as follows. 1) Specific surface area, cumulative pore volume and non-free water content increases linearly with increasing allophane content. 2) Natural water content, liquid limit, plastic limit and ignition loss increases linearly with increasing allophane content. 3) Liquidity index decreases with increasing allophane content. 4) Volcanic cohesive soil lies below A-line in plasticity chart and fall more far from the line with increasing allophane content. The soil has its characteristics linear line in the chart. 5) Liquidity index increases linearly with increasing coefficient of flow β which is defined as W_n/W_l. Reasonable classification can be obtained by using the two factors.

View full abstract

Conventional methods for the design of slopes in the residual Hong Kong soils that require determination of soil/rock parameters, selection of a failure model and assignment of a piezometric surface (or similar) provide unsatisfactory estimates of safety factor owing to the uncertainty levels that prevail. An alternative approach is discussed which, while only approximate, uses measured Hong Kong slope performance to estimate probabilities of failure for a variety of conditions. Such probabilities should be useful in conjunction with consideration of the consequences of failure and the formal assessment of risk.

View full abstract

This paper presents the centrifuge modeling of a prototype pull-out test conducted on enlarged base foundation using undisturbed block samples so that prototype values could be possible to extrapolate from model results. Undisturbed block samples were directly procured in specially designed centrifuge model containers to avoid sample disturbance. Samples were also reconstituted in laboratory to compare the behavior of undisturbed and reconstituted models. The undisturbed models showed higher pull-out resistance than that of reconstituted models and suggested the preservation of cohesion and in-situ soil fabric in the undisturbed models. When the results of undisturbed model were compared with the prototype results, a successful prediction regarding pull-out resistance was observed. However, the centrifuge models over predicted displacements of the prototype test. The deviation may arise from the scale and particle size effect due to the lack of similarity in shear band thickness as the peak values of pull-out resistance occur at different displacements. Together with the previously published data, it may be inferred that there might exist a correction factor for displacement in order to match the centrifuge data with the field test data.

View full abstract

The effects of superimposing cyclic shear stress on the undrained behavior of saturated sand under monotonic loading are identified and discussed. This stress-controlled cyclic shear stress is intended to represent the small seismic shaking persisting after the main strong shaking of an earthquake. It is characterized by the cyclic stress ratio and the number of cycles applied per 1% of shear strain developed by the sand along the direction of the strain-controlled monotonic load. The experimental program includes tests performed using bidirectional simple shear and hollow cylindrical torsional shear types of apparatus. The only material employed in this study is standard Japanese Toyoura sand. In both apparatus, the complementary cyclic stress is applied through and independent of the monotonic load. One of the most remarkable effects is the reduction of the strength attained by saturated sand under monotonic loading application alone.

View full abstract

The objective of this paper is to develop a new vibro-compaction machine by investigating experimentally the effects of the frequency of an oscillator mounted on a tracked vehicle on the vibro-compaction of a high lifted decomposed granite sandy soil. By measuring the amount of depression of the ground surface, the dry density distribution in depth using a cone penetrometer, the normal earth pressure distribution in depth using a soil stress transducer, and the vertical and horizontal acceleration distribution in depth, the effects of the tracked vehicle on the increment of the soil compacting dry density in a deep stratum were considered theoretically for various kinds of frequency. It was observed that an oscillator frequency of 16 Hz with a constant amplitude of exciting force of 9.8 kN showed the maximum amount of depression of the ground surface and the maximum dry density distribution at deep stratum for the range of frequency from 16 Hz to 51 Hz. The ratio of dynamic shear stress to normal stress at 16 Hz proved to be large enough at deep stratum, so that an optimal alternative shear strain due to the impact force was developed on the whole range of the high lifted stratum. It was assumed that the increment in soil compacting density was the result of the effective dilatancy phenomenon.

View full abstract

This paper presents a method of analysis for piled raft foundations on layered soils. The method combines both the finite element method for analysis of rafts and the finite layer method for analysis of soil-pile groups. This means that rafts of any shape that are subjected to point or uniform loads or moments may be analysed. As well, the soil may consist of layers of different stiffnesses or have a stiffness that increases with depth. Deflections of the piles and raft as well as the forces in the piles and moments in the raft may be computed. Several case studies that include a centrifuge model and full scale field tests for piled raft foundations are examined in this paper. The comparisons between calculated values and measured ones show that the method can be used successfully for analysis of piled raft systems in layered soils.

View full abstract

Most of the correlations readily available in the literature are based on laboratory tests performed on clean, low compressibility, rounded quartz sands. These correlations are extended to predict the behavior of all types of sands regardless of their angularity and mineralogy. While the importance of both the engineering characteristics and mechanical properties of sands has been recognized for quite some time, very few comparative studies are available. In this study, model plate load tests, model standard penetration tests and direct shear tests were conducted on three angular compressible natural sands and one commercially available rounded sand of low compressibility. Based on the results of these tests, it is shown that at a given relative density, blow count in the model standard penetration test, failure load in the model plate load test, Terzaghi's bearing capacity factor, N_γ, and angle of shearing resistance are much higher for the angular compressible sands as compared with those for the rounded, low compressibility sand. Moreover, a sand with micaceous flakes was found to behave in a much different way as compared to quartz sands during the model tests. The effect of sand type on packing characteristics obtained by raining technique is also emphasized.

View full abstract

The mechanical behaviors of sands are notably influenced by their physical properties as well as by environmental conditions : stress and strain conditions, and drainage conditions. The physical properties of soils can be classified into primary properties and secondary properties ; primary properties are permanent properties of soil grains such as density, hardness, shape and size of grains, and secondary properties are non-permanent properties of soil structures such as void ratio, packing and moisture content of soils. The objective of this study is to reveal the mechanical behaviors of soils with different primary properties. The mechanical properties of sands were investigated extensively ; the mechanical behaviors of some 200 granular materials were observed in a series of conventional triaxial compression tests. The significant influence of the primary properties on the deformation-strength characteristics was clarified. From the test results, the relationships of deformation-strength properties of sands with index physical properties are discussed ; the significant influence of grain shape, grading and crushability on the mechanical behaviors are presented.

View full abstract

A theoretical formulation for the incorporation of constraint conditions imposed upon the displacement/velocity field of the soil-water coupled system is presented along the finite element discretization scheme. The"no-length change", "no-angle change"and"no-direction change"conditions are introduced as internal constraints to the displacement/velocity field through the Lagrange multiplier method. These multipliers are then interpreted as the internal constraining forces acting between the soil and the"constraint mechanism"system. The mathematical background of the method is first outlined in the INTROKUCTION to give a clear perspective using introductory finite element formulation. The method of analysis is illustrated through several examples of the simulation of load controlled undrained triaxial tests and displacement controlled plane strain undrained compression tests of the normally consolidated original Cam clay. In these tests the load application is made through a rigid pedestal which provides a"mechanism"for applying constraint conditions to the displacement/velocity field at the top of soil samples. The interaction between soil and pedestal is solved automatically and the reaction force from the pedestal to the soil and its distribution along the pedestal are obtained simply by computing the vector of the Lagrange multipliers. Thus it is found that broad applicability to the geotechnical engineering problems emerges from the present formulation.

View full abstract

The delayed compression and progressive failure of a saturated assembly of crushed mudstone pebbles that occur under constant load application are presented and concluded to be the results of the"slaking"of the mudstone pebbles. The mudstone itself is first idealized as a heavily overconsolidated clay, and the slaking is interpreted as softening behavior with swelling under shear stress application. Due to the slaking, overconsolidated mudstone pebbles tend to return from the initial stiff solid state to a previous, almost normally consolidated soft clayey state. The occurrence of excess shear stress between pebbles is considered to occur possibly due to the concentrated contact forces between them. To substantiate this, one-dimensional compression tests and triaxial compression tests are carried out on the assembly of crushed mudstone pebbles. Although this assembly is to be submerged in water, the individual mudstone pebble itself may initially be in a dry, unsaturated, or saturated condition. The main results obtained through experiment are as follows. (1) The one-dimensional compression tests show that states of the loose specimens in e∼σ_v space move down gradually towards the normal consolidation line of the clay that was remolded from the mudstones. To prevent long lasting large settlements, it is necessary for the assembly of mudstones to be compacted densely to get their overall void ratio below the normal consolidation line. (2) Even packed densely, the assembly still exhibits progressive failure with time under high shear stress application. (3) Both the delayed settlement and the progressive failure mentioned above are caused by the slaking of mudstone pebbles that could occur due to high contact force concentration between pebbles' boundaries. (4) When mudstone pebbles are initially in unsaturated or dry condition, slaking occurs at a faster rate compared with the saturated case.

View full abstract

Limit loads on footings (bearing capacity) can be conveniently analyzed using the kinematic approach of limit analysis. Nonsymmetrical loads on strip footings include the horizontal component of the load and the moment. They can be also represented as load inclination and eccentricity. For the analysis to be sensitive to nonsymmetrical loads, the collapse mechanisms also need to be nonsymmetrical, i. e., they need to allow for footing rotation and horizontal component of displacement. Based on such mechanisms, inclination coefficients for surface strip footings are derived. These coefficients fit the numerical limit analysis results better than the coefficients suggested earlier in the literature. It is also shown that for cohesive-frictional soils the method often used to account for the load eccentricity (where the footing width is reduced by twice-the-eccentricity) yields a bearing capacity lower than that obtained from the kinematic limit analysis. However, this method overestimates the bearing capacity for purely frictional soils, particularly when the surcharge load is small.

View full abstract

The first author derived a generalized one-dimensional consolidation theory in 1961 involving strain that takes into account the changes in m_v, K, consolidation pressure and the depth of the clay layer (finite strain), during the consolidation process, and the effect of the selfweight of clay. This theory, however, applies only to clay layers with uniform consolidation properties ; it has limited applicability to clay layers in the field, which are usually composed of different soils. The consolidation theory was thus amended to account for the continuous change in consolidation properties along the depth. The derivation of this consolidation theory is detailed, and two examples of application to thick alluvial clayey layers in the Osaka Port area together with comparisons with field observations are presented.

View full abstract

An elastoplastic deformation law for sands incorporating the state concept is formulated. These fundamental concepts for modelling and characterization of sand behaviour are unified in order to build a rational sand model that enables integral modelling over a wide range of density and normal stress states. A modified hyperbolic stress-strain relation and an energy based stress-dilatancy relation are used as basic relations that embody the essence of sand behaviour into the constitutive model. The stress-strain relation is established within the framework of the state concept and uses the state index I_s to quantify the effects of the relative initial state (initial density and normal stress state) on the stress-strain curve. These effects are reflected on the dilatancy of the sand by specifying that the principal parameter of the stress-dilatancy relation depends on the shear strain. The most distinctive features of the plasticity formulation are the assumptions for continuous yielding and the dependence of the plastic strain increment direction on the stress increment direction. The ability to model undrained sand behaviour for various relative densities with the same values of the material parameters is demonstrated.

View full abstract

A numerical procedure is presented to analyse consolidation problems involving anisotropic layered soils which contain incompressible as well as compressible pore fluid. The governing differential equations are reduced to ordinary differential equations by applying a Fourier transform along with a Laplace transform, and the solution is found in the finite element fashion without great computational efforts. The soil deposit is discretized by one-dimensional elements, and for each element, use is made of quadratic interpolating functions that implicitly enforce interelement continuity. A number of comparisons with existing solutions are shown in order to assess the accuracy of the present method. Finally, a parametric study is carried out to investigate the influence of elastic anisotropy on the consolidation processes caused by surface loading.

View full abstract

The prototype of a friction pile-box foundation, one of the supports of an urban bridge, was instrumented in order to learn more about the performance of this usual type of foundation in the difficult soft clayey soils of Mexico City. Load-transfer mechanisms that develop at pile-soil and slab-soil interfaces during construction stage and the beginning of bridge operation are discussed in this paper, taking into account the measured applied loads on friction piles, pressure distribution at the raft-soil interface, piezometric heads in the subsoil water, and the foundation movements.

View full abstract

The need to model the behaviour of piled foundations during serviceability state has required a lot of attention paid to analytical methods, in order to consider the interaction between piles and the sorrounding soil. Among the simplified approaches, the concept of"interaction factors"introduced by Poulos (1968) which takes into account the influence of loading on a pile on the settlement of an adjacent pile is relevant. This note aims to develop an analytical expression for the interaction factor, taking into account the soil non-linear response. The expression, relevant for floating rigid piles, is derived assuming a linear variation of shear modulus G with radial distance and finally is used to model the case record of a 5-pile group.

View full abstract

In this study the time-dependent behavior of geomembranes is modeled using elasto-viscoplasticity. For this purpose a particular time-rate rule, incorporating the consept of yield function from the traditional theory of plasticity, introduced. This time-rate rule is completely defined by three independent material parameters, which can be easily determined experimentally. The validity of the proposed approach is checked by comparison with experimental results of long-term tests on high-density polyethylene geomembranes.

View full abstract

A numerical model is proposed to simulate the seepage through unsaturated soil based on some mechanical and probabilistic considerations on the particle scale. In the model the probability density functions are introduced to estimate the state of voids in soil. The hydraulic properties of unsaturated soil ; i. e. the relations among the water content, pF-value and permeability coefficient are theoretically derived by using the model, and numerically calculated by using a personal computer. The numerical result is compared with a pF-test and unsaturated permeability test on Toyoura sand. In the proposed model the hydraulic properties of unsaturated soil can be derived by only the grain size distribution.

View full abstract

Hypo-elasticity does not generally imply the existence of an elastic strain energy potential. However, it appears that for a hypo-elastic model describing inherent and stress state-induced anisotropy of uncemented granular materials, elastic strain energy can be stress path-independent. As a consequence, elastic strain energy represents a potential uniquely defining the direction of the elastic strain vector along stress paths with a constant stress ratio, which is confirmed by experimental data for some sands.

View full abstract

This paper gives a description of a new"simplified"direct box shear test apparatus in both small and large sizes, and the test results obtained by the new apparatus. In the new apparatus, there is no upper shear box as used in the standard direct box shear test. A loading plate is directly placed on the sample in the lower shear box and pulled horizontally by a flexible rope or chain. By doing this, the effect of frictional forces between the sample and the internal surfaces of the upper shear box induced in the standard direct box shear test during dilatancy is naturally eliminated. The test using the new apparatus is similar to a frictional test on usual materials. A number of tests are performed on glass beads, Toyoura sand and crushed sand using the new small-sized apparatus and on three kinds of granular materials including rockfill materials (the maximum grain size : 150 mm) using the new large-sized apparatus. The test results from the new apparatus agree well with those from standard direct box shear tests in which the effect of frictional forces is reduced, and are near to those from triaxial compression tests. It takes only about 30 minutes to test even the rockfill materials with the new apparatus, including the time to set the specimen.

View full abstract

The relationship between matrix suction and volumetric moisture content for a compacted residual clayey silt soil was established by using the filter paper method to measure the suction. For specimens compacted to a common dry density, the relationship between the logarithm of suction and the volumetric moisture content was approximately linear over the suction range 4 to 8000 kPa. The effect of compaction moisture content, and associated soil fabric, on the relationship was small. Consistent data were obtained from pressure membrane tests on specimens subjected to wetting.

View full abstract