Statistical properties such as the mean value and the coefficient of variation of undrained shear strength and consolidation parameters of disturbed soils were investigated through laboratory testing. A series of unconfined compression tests and standard consolidation tests were performed on natural deposits of Kuwana clay. The values of consolidation yield stress, P_c, and compression index, C_c, become larger than the actual values for a range of the ratio of q_u values of disturbed samples to the undisturbed sample, q_u-ratio, being greater than 0.8, and vice versa for a range of q_u-ratio being smaller than 0.8. The effects of sample disturbance on coefficient of consolidation, c_v, coefficient of volume compressibility, m_v, and coefficient of permeability, k, are different for the range of consolidation pressure, p, up to p_c. Therefore, the q_u-ratio is affected only by these consolidation parameters, and is independent of I_p, P, aud q_u values. For the normally consolidated region, the c_v, m_v, and k values of remolded soil are about 70%, 40%, and 80% smaller than those of the undisturbed soil, respectively. The relationship between consolidation parameters and sample disturbance is given as a function of the q_u-ratio. A comprehensive procedure to obtain the statistical values of strength and consolidation parameters of undisturbed specimens from those of disturbed samples is presented.

View full abstract

A saturated sand deposit placed below a structure is subjected to nonuniform vertical stresses as shown in Fig. 4. It is known that there is an interesting experimental fact that a saturated sand deposit near a structure on the same horizontal plane liquefies more quickly than it below a structure. In order to clarify the liquefaction characteristics and the amount of liquefaction-induced settlement of a saturated sand deposit under nonuniform vertical stresses, a series of shaking table tests were performed on a simplified model (1), prepared in three Kjellman's type shear boxes. This model consisted of three saturated sand layers for which the effective vertical stresses σ'_vo were 10.4 kPa, 30.0 kPa and 49.6 kPa respectively. Similar tests were also carried out on four models (models (2) to (5)) consisting of two sand layers having different RVS values (= (σ'_vo)_l/(σ'_vo)_h), in which (σ'_vo)_l, (σ'_vo)_h represent lower and higher vertical stresses for the two respective sand layers. Since the values of RVS and k_h affect the permeability of and the transmission of pore water pressure between the sand layers, we found significant differences in liquefaction occurrence for such model sand layers. It was also found that the liquefaction-induced settlement of the model sand layer increases substantially in the range of smaller vertical stresses and remains at an almost constant value as vertical stresses increase.

View full abstract

The results of a series of plane strain model tests of unreinforced and reinforced sand slopes loaded with a 10 cm-wide strip footing, which was described in the companion paper in the previous issue, are analyzed by limit equilibrium methods. A statically rigorous method, namely Janbu's method, is modified in order to incorporate the reinforcement force acting on the potential failure surface and the inter-slice faces in the force equilibrium formulation. It is shown that the results of stability analysis using a composite failure surface can predict rather accurately the stability and the location of failure surface for both unreinforced and reinforced slopes, provided that all three factors that impact the mobilization of the friction angle φ along the failure surface are taken into account. These factors are the pressure level-dependency of φ, the strength anisotropy, and the progressive failure. In that case, the mobilized angle of friction along the failure surface is not constant. The degree of progressive failure increases as the effect of reinforcing increases. The comparison of some other simplified stability analysis methods currently used in the design of reinforced slopes with the modified Janbu's method showed that a less realistic assumption concerning force equilibrium leads to a more unrealistic result, and when the increase in the shear strength of soil from the tensile force in reinforcement is not taken into account, the stability could be largely under-estimated.

View full abstract

Field vane shear strength, s_u(v) is presented for seven Japanese soft marine clays with the plasticity index, I_p ranging from 20 to 120. The value of s_u(v) was compared with that obtained from unconfined compression tests which is a standard for determining the shear strength of cohesive soils for design use in Japan. Several researchers have reported that the vane shear strength normalized with the consolidation yield pressure, s_u(v)/p'_c is dependent on I_p. However, the s_u(v) /p'_c values obtained from the present study were generally constant with I_p' with values ranging from 0.25 to 0.35. Bjerrum's correction factor, μ is commonly used to determine the shear strength for stability analysis from the vane shear strength. The shear strength modified by this method, however, was found to be considerably conservative for Japanese marine clay, when compared with the unconfined compressive strength that has been proved to provide an appropriate shear strength for practical use in Japan.

View full abstract

Slope collapse for embankments and cut slopes on railways generally occur during heavy rainfall from typhoons etc. and have a strong influence on running safety of trains. Risky locations along the railway right of way must be identified to control railway operations based on the conditions and characteristics for each railway line. For this purpose, a practical damage estimation method, which accurately predicts the hazard of slope collapse and can be used simply by railway workers, has been developed. This paper deals with this method of risk estimation for a surface collapse of a cut slope during times of heavy rainfall using a multivariate analysis. The critical rainfall is obtained using a combination of an accumulated rainfall and an hourly rainfall, which have the powers 0.2 and 0.9 respectively. The critical rainfall is then determined from external variables such as geometrical and structural configuration of the cut slope, subsurface soil and rock conditions, catchment condition and experiential rainfall condition. It was determined that the hourly rainfall has more influence than the accumulated rainfall on the surface collapse of cut slopes. Since it was confirmed that this risk estimation standard agreed with some values observed for the surface collapses, it is suggested that the method can be used not only for planning the disaster prevention along railway right-of-way but also for predicting of the cut slope collapse due to heavy rainfall.

View full abstract

Based on a study of remolded young clay, this paper shows that the time effects observed at constant strain rate shearing, stress relaxation and creep processes in triaxial tests and the secondary compression for K₀-consolidation process are commonly simulated by the time dependency of dilatancy. The fundamental aspect of the proposed procedure is to solve the following two state equations as a simultaneous equation, without using the plastic flow rule. One is an equation for specifying a volumetric strain due to consolidation and dilatancy. The other is the energy equation used in the original Cam Clay model, which gives a plastic shear strain corresponding to a plastic volumetric strain. The proposed procedure duplicates fairly well the time effects in undrained triaxial test. For the process of K₀-consolidation, the equation for specifying a volumetric strain and a constraint for restricting lateral deformation are solved as a simultaneous equation, because the energy equation may not hold for this process. The procedure provides a settlement curve which is consistent with the curve observed in a standard oedometer test for the normally consolidated region, and clarifies the following properties. 1) The secondary compression in remolded young clay, is represented by the delayed occurrence of dilatancy. 2) In the standard oedometer test, the secondary compression takes place repeatedly at each step of loading, due to the mean effective stress which initially exists within the test specimen. 3) The secondary compression coefficient tends to decrease with increasing loading pressure, because the magnitude of dilatancy expected may decrease with the progress of step loading. 4) The secondary compression begins in the primary consolidation stage, depending on the value of the stress ratio. 5) The K₀-value estimated by the procedure agrees well with the value monitored by triaxial equipment.

View full abstract

A solution for the hydrodynamic clean up of contaminated land with flow reversal has been developed using analytical and numerical methods. This paper presents the results of physical modelling of the clean up process and compares the data with the theoretical predictions. A series of one dimensional laboratory tests are described in which clean water was flushed through a site initially contaminated with sodium chloride. The effects of flow reversal and soil permeability on the clean up process are investigated. A centrifuge test of hydrodynamic clean up using flow reversal is also described. Prototype stress and permeability conditions are modelled enabling the simulation of the clean up of a large scale homogeneous site. The theorectical models produced fairly accurate predictions of the varying contaminant concentrations in the model tests.

View full abstract

The stress-strain behavior of a granular material is highly complicated relative to that of a cohesive soil. Due to the frictional nature of the material, a plasticity framework based on non-associated flow rule is generally considered to be most suitable for modeling its constitutive behavior. By employing two surfaces, it is shown in this paper that a plasticity model based on an associated flow rule provides a satisfactory means of simulating, under a monotonic loading, the stress-strain behavior of dense and loose sands under triaxial undrained and drained loading conditions.

View full abstract

The main requirements for assessing the system reliability of slopes are outlined briefly and a comprehensive procedure is presented based on potential slip surfaces of arbitrary shape. The evaluation of system reliability of soil structures such as slopes is important because of the spatial variability of geotechnical parameters within any soil layer and also the fact that slopes are frequently associated with multi-layered soil deposits. The performance function for the stability of a slope, defined in terms of a 'factor of safety', is usually inexplicit and non-linear. A 'rigorous' method of slope stability analysis such as the Generalised Procedure of Slices may be used as the basis of a reliability analysis. A complete analysis should include the search for a critical slip surface and the calculation of reliability index or the probability of failure associated with it. System reliability bounds are evaluated by considering the probabilities of failure associated with a number of slip surfaces including the critical slip surface. It is shown that the upper bound of the system failure probability can be significantly higher than the failure probability associated with the critical slip surface. This finding is significant for slope design and for geotechnical risk assessment generally.

View full abstract