More than 3000 multi-anchor walls have been built in Japan over the last decade. The paper briefly reviews a total of eight instrumented wall sections that can be used to estimate anchor loads at the end of construction. Measured loads are compared to predicted values using equations found in current design guidelines. The comparison shows that the current Japanese and UK design methods to compute anchor loads are reasonably accurate for walls with frictional backfills provided K_a is calculated using the Rankine equation. For walls with cohesive-frictional backfills, current design methods over-predict anchor loads by as much as a factor of two. The eccentricity term in current UK and Hong Kong design methods is shown to not improve the accuracy of load predictions and it is recommended that this additional complexity be removed from these equations. A new load equation is proposed and constant coefficients are back-fitted to measured anchor load data. The new method is demonstrated to give quantitatively better predictions of anchor loads based on the statistics for load bias values computed as the ratio of measured to predicted anchor loads at the end of construction.

View full abstract

The evaluation of undrained shear strength of soils is necessary in determining the possibility of occurrence of flow deformation during earthquakes. The present study is aimed at examining the evaluation of undrained shear strength of silty sands from field with Swedish weight sounding tests and cone penetration tests. Based on the outcome of the previous studies on laboratory triaxial tests, the undrained shear strength ratio is defined as the undrained shear strength divided by the initial effective major principal stress. The undrained shear strength ratio is then formulated with respect to the relative density. The penetration resistances of Swedish weight sounding and cone penetration tests are then formulated with respect to the effective overburden stress and relative density, based on laboratory calibration chamber tests. By combining these formulations, the correlations of the undrained shear strength with Swedish penetration resistance and cone tip resistance are established. The range of values of penetration resistances indicative of soil layers susceptible to flow deformation is discussed. The correlations of the undrained shear strength with field penetration resistances thus derived are then examined from case history studies. Two case history studies are carried out with Swedish weight sounding tests at the sites of flow failures induced during the recent earthquakes. A series of case history studies are reexamined, which were carried out with Dutch cone penetration tests in the past studies.

View full abstract

The viscous properties of a wide variety of unbound granular materials (GMs) were evaluated by drained shear tests. The specimens were reconstituted ones that were loose or dense and air-dried or moist or saturated, of mostly natural sands and gravels, having different mean particle diameters, uniformity coefficients, fines contents, degrees of particle angularity and particle crushabilities. The tests were mostly triaxial compression (TC) tests and partly plane strain compression tests, both at fixed confining pressure, and direct shear tests at fixed normal pressure. The viscous properties of GMs were evaluated by stepwise changing the loading rate and performing sustained loading (SL) tests during otherwise monotonic loading (ML) at a constant loading rate. The viscous properties are characterised in terms of the rate-sensitivity coefficient (β), the viscosity type parameter (θ=β_r/β) and the decay parameter (r₁). Correlations among these parameters and effects of particle characteristics on these parameters are analysed. Creep strains are compared with residual strains by cyclic loading under otherwise the same TC conditions. As the particles become less angular, as the grading becomes more uniform and as the particles become less crushable, the viscosity type deviates more from the Isotach type (i.e., θ=1.0) changing toward the P & N type (i.e., θ<0) associated with a decrease in β and r₁, while creep strains by SL decreases and residual strains by many unload/reload cycles increases. It is shown that the loading rate effects observed in the experiments can be simulated well by the three-component model taking into account the effects of particle characteristics on the viscous property parameters.

View full abstract

Cost-effective design is the primary motivation for adopting the performance-based design method. This method, however, requires that deformations be reliably estimated. While soft rocks are known to be competent foundation materials for large-scale structures, the deformation characteristics of this material when subjected to large cyclic loadings still have to be understood. In this study, the strength and deformation characteristics of soft rocks under cyclic loadings were investigated by conducting cyclic triaxial tests on natural soft rock samples. The loading histories that were applied to these samples were uniform amplitude cyclic loadings with loading periods between 1 s and 9000 s. The tests revealed that the longer the loading period, the larger is the residual strain accumulated for a certain number of loading cycles. This dependency of residual strain accumulation on loading period appears to be an intrinsic material property which is irrespective of loading amplitude and water content. From this finding, it can be inferred that the prevailing practice of soft rock cyclic loading tests at 300 s and 9000 s of cyclic loading periods, which are much longer than that of earthquakes, can result in overestimated residual strains.

View full abstract

Data and information on the cyclic loading behaviour of soft rocks, especially behaviour under irregular cyclic loading, are very limited. This paper shows that the present procedure of estimating residual strain accumulation due to irregular cyclic loading using a fatigue model from uniform amplitude cyclic loading can result in underestimated residual strains. Such underestimation occurs because the present procedure fails to take into account the so-called P2S effect on the softening behaviour of soft rocks under cyclic loading. The parameter P2S is defined as the sum of the magnitudes of the increments in residual strains due to the previous two loading half-cycles. When P2S is large, a large residual strain increment can be expected. This paper also shows that taking the P2S effect into account can improve the simulation of residual strain accumulation due to irregular cyclic loading.

View full abstract

A series of monotonic and cyclic shear tests, as well as pullout tests, were conducted on gravel-geotextile interfaces using a large-scale apparatus, with development of a new special pullout test element. The macroscopic response of stress and displacement, as well as the movement and crushing process of soil particles, were observed and measured. The interface exhibited evident strain-softening and aeolotropic normal displacement, which were significantly influenced by normal stress. Shear strength decreased and normal displacement increased with increasing number of shear cycles. Shear deformation was composed of slippage at the contact surface and deformation of the soil constrained by the geotextile; and the thickness was estimated at 5-6 times the average soil grain size. There was significant evolution of physical state due to shear application, including soil particle crushing and soil compression, as well as damage to the geotextile. The pullout test underestimated shear stiffness of the interface due to significant deformation of the geotextile itself. Shear strength increased with increasing normal stress, described by a logarithmic equation, according to the pullout tests, rather than the linear relationship obtained using direct shear tests. Therefore, an appropriate test method should be selected with careful consideration of the site conditions.

View full abstract

This paper presents the role of fly ash on strength and microstructure development in blended cement stabilized silty clay. Its strength was examined by unconfined compression test and its microstructure (fabric and cementation bond) by a scanning electron microscope (SEM), mercury intrusion porosimetry (MIP), and thermal gravity (TG) analysis. The flocculation of clay particles due to the cation exchange process is controlled by cement content, regardless of fly ash content. It increases dry unit weight of the stabilized clay with insignificant change in liquid limit. This results in irrelevant difference in optimum water content (OWC) for the unstabilized and the stabilized clay since OWC of low swelling silty clay is mainly controlled by liquid limit. It is found from the microstructural and the strength test results that the reactivity of fly ash (pozzolanic reaction) is minimal, which is different from concrete technology. This is possibly due to less amount of Ca(OH)₂ to be consumed. The role of fly ash in cement stabilization is to disperse the large clay-cement clusters into smaller clusters. Consequently, the reactive surfaces to be interacted with water increase, and hence the cementitious products (inter-cluster cementation bond). To conclude, the strength development in the blended cement stabilized clay is controlled by cementitious products due to combined effect: hydration and dispersion. Cementitious products due to hydration are governed by cement content, while cementitious products due to dispersion by fly ash content and fineness. Water content of 1.2OWC and 10% replacement ratio are regarded as the effective mixing condition for the stabilization, exhibiting the highest cementitious products.

View full abstract

There have been many theories and interpretations of the compression process for soil. However due to the complexity of the compression process for a wide range of soil states, i.e., suspended, liquid, plastic, semisolid or solid states like rock, the interaction between the states or the transition from one state to another, there have often been misunderstandings and difficulties regarding the application of these theories. In this study, an ideal particle distribution in an ideal ground system is derived to find the equation of state, a general compression process for a wide variety of soil types and states. In other words, it is a general relationship between the void ratio and effective overburden pressure in terms of interparticle energy and is thus a “law of soil.” The principle is based on the law of energy distribution in statistical mechanics. The derived equations can be applied to describe the void ratio profile and the compression process of various soils. Case studies show that this theoretical approach agrees well with the experimental data obtained for several soils such as sediments, Mexican City clay, sand and others.

View full abstract

As a basic study for future lunar/planetary explorations and the in-situ resource utilization missions, bearing capacity characteristics of shallow footing systems in a low gravity environment were investigated. A series of model loading tests on a simulated lunar soil (lunar soil simulant) and Toyoura sand were conducted on an aircraft that flew in parabolic paths to generate partial gravity fields. As a result of the model tests, it became clear that bearing characteristics, including the coefficient of subgrade reaction and ultimate bearing capacity of the lunar soil simulant in a low gravity environment is hardly influenced by the gravity levels, while Toyoura sand exhibits a high dependence on gravity. From the observation of the failure mechanisms, it was found that the gravity dependence seems to correlate well with soil compressibility. To rationally explain the dependence of ultimate bearing capacity on gravity, theoretical evaluations were attempted in the framework of the upper bound method. The proposed calculation method not only makes it possible to correlate quantitatively the failure mode with dependence on gravity, but also may allow us to predict the ultimate bearing capacity in the lunar surface environment.

View full abstract

In this paper, an innovative procedure is developed for estimating the uncertainty of an empirical geotechnical model. Here, the Youd et al. (2001) method, a deterministic model for liquefaction triggering evaluation, is examined for its model uncertainty. The procedure for evaluating this model uncertainty involves two steps: 1) deriving a Bayesian mapping function based on a database of case histories, and 2) using the calibrated Bayesian mapping function as a reference to back-figure the uncertainty of the model. Details of the developed procedure within the framework of the first-order reliability method (FORM) are presented. Using FORM with the calibrated model uncertainty, the probability of liquefaction can be readily determined, and thus, the results presented in this paper extend the use of the Youd et al. (2001) method.

View full abstract