A numerical comparison of predictions by a rigorous 3-D limit equilibrium analysis and a finite element method is presented. The finite element method was modified so that it can predict a factor of safety having the same definition as in limit equilibrium. Thus, a meaningful basis for comparison was formed. Comparisons were limited to vertical cuts and included a pseudo-static seismic force component. Safety factors and their associated slip surfaces, predicted by the finite element and limit equilibrium analyses, exhibited very good agreement. This is surprising since the two analyses are fundamentally different. This good agreement adds credibility to the results of either analysis. This is particularly important in 3-D since detailed empirical information for analysis validation is extremely limited.

View full abstract

After complation of an undrained triaxial compression test to failure with a constant rate of axial strain, a clay specimen was continuously subjected to the same rate of axial strain but with a pore water supply from outside through top and bottom boundaries. The clay, then, exhibited a softening-like behaviour with swelling. A series of these triaxial experiments was next numerically simulated based on the soil-water coupled finite deformation computation using the original Cam-Clay constitutive model, in which a clear distinction was made between a triaxial soil specimen and soil elements in it. The simulation exhibited almost the same post-failure behaviour as that observed in experiment. Apparent softening behaviour of the clay specimen was demonstrated due to the behaviour of soil elements that traversed the critical state line. Annealable behaviour of clay was also shown possible with swelling after failure, in which the clay made a "comeback" from a current normally consolidated state to a former normally consolidated state. The stress path for which in a q : p' : v space was also illustrated.

View full abstract

Huge offshore gravity structures such as deep-sea oil platforms are often equipped with a skirt system underneath their bases. When such foundations of substantial size with deep skirts are to be placed on soft clay deposits, it becomes important to take account of the influence of a non-homogeneous strength profile in bearing capacity calculations. In this study, plasticity theory was used to evaluate the influence of linearly increasing undrained shear strength with depth on the bearing capacity of shallow foundations, under both plane strain and axisymmetric conditions. Based on the results of computations, a simple method is proposed to calculate bearing capacity using a set of charts, which introduces the effect of increasing strength consistently into the framework of conventional bearing capacity theory. A series of centrifuge experiments was also conducted for circular models founded on clay beds. With a limited number of test cases, the experimental results have proved the validity of the proposed method. Furthermore, the failure observed in the model clay beds was due to punching shear; the mechanism of which was confined to shallow depths to avoid shearing of the clay quickly strangthening with depth.

View full abstract

Embankment and cut slope collapses along railways sometimes occur under heavy rainfall caused by typhoon etc. and have influence on safe operation of the railroad trains. The risky locations along the railway must be identified in order to safely control operations 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, is a desirable development goal. This paper deals with a discriminant method between a surface and a deep collapse for cut slope collapses, and a critical rainfall prediction method for the deep cut slope collapse during heavy rainfall. The critical rainfall is given by a multiplying an accumulated rainfall by an hourly rainfall, which have the powers 0.4 and 0.2 respectively, using a multivariate analysis. The critical rainfall is then given by external variables such as geometrical and structural conditions for the cut slope, soil and rock, catchment and experiential rainfall conditions. This evaluation indicates that the accumulated rainfall has more influence than the hourly rainfall on the deep collapse of cut slopes. Since it was confirmed that the risk estimation standard proposed satisfied some observed values for deep collapses, it is considered that it can be used not only for planning the disaster mitigation procedures along railway right of way but also for predicting cut slope collapse due to concentrated rainfall.

View full abstract

In order to analyze the effect of vertical drains in plane strain condition, a new matching procedure based on well resistance is proposed. The advantage of this procedure is that it not only matches the average degree of horizontal consolidation in axisymmetric by plane strain condition, but also yields a more realistic excess pore pressure distribution in horizontal direction. The method was incorporated into finite element analysis, and it was demonstrated that half of the drain spacing could be modelled by one-column of elements. Then, an actual case history of an embankment on vertical drain improved subsoil was analyzed by the proposed method. It was shown that vertical drains have two effects on subsoils response : (1) increasing settlement rate, and (2) reducing lateral displacement. For the case analyzed, the vertical drain used was without a filter. For matching the surface settlement, a low discharge capacity of about 1.5 m³/year was found (fixed smear effect). The analysis results were compared with measured values, and it was found that discharge capacity may be reduced significantly with an increase of confining pressure on the sleeve (wall) of the drain.

View full abstract

The Spatially Mobilized Plane (SMP) for frictional materials is extended to a plane for frictional and cohesive materials, which is named "Extended Spatially Mobilized Plane (Extended SMP)", by introducing a parameter of "bonding stress σ₀". The "Extended SMP" includes the SMP applicable to frictional materials such as granular materials (σ₀=0) and the octahedral plane applicable to cohesive materials such as metals (σ₀→∞) at the two extremes. This corresponds to the fact that the two extremes of frictional and cohesive materials are granular materials and metals. The constitutive law is verified quantitatively using experimental data of triaxial compression, triaxial extension and true triaxial tests on cemented sands, which are selected as an intermediate material with both friction and cohesion. The experimental stress-strain relationship for the cemented sands under three-dimensional stress conditions are uniquely arranged on the "Extended SMP", and the strength of the cemented sands under three-dimensional stress conditions are well predicted by "Extended SMP" failure criterion.

View full abstract

Three-dimensional (3D) FEM analysis was carried out using elastoplastic models which are capable of describing various elastoplastic behavior modes for clay and sand particularly in three dimensional stress. A ground model (lowering basement) that simulates the excavation mechanism for a tunnel was utilized for the FEM calculation. Two-dimensional (2D) FEM analysis was also carried out using the same elastoplastic models in order to compare with the 3D analytical results. Through this comparison, it was concluded that 3D analysis is necessary for a problem like tunneling where the construction procedure is not two dimensional even though it may look like a plane strain problem at the completion stage. In addition, it was shown that the positive and negative dilatancy for soil materials plays an important role in both settlement and earth pressure analysis.

View full abstract

This paper describes a series of laboratory tests on single instrumented model piles embedded in calcareous sand undergoing lateral movement. Key parameters influencing the maximum bending moment in the pile for a constant soil density have been identified to be pile head fixity condition, the ratio of the depth of moving soil to the pile embedded length, and pile diameter and stiffness. Normalized expressions for maximum bending moment are also presented. The agreement between the experimental results and the theoretical predictions by an existing boundary element program is shown to be reasonably good.

View full abstract

An experimental investigation of the behavior of sand near failure was undertaken to study the shape, location and movement of the plastic yield surface in the hardening regime and in the softening regime near peak failure. Triaxial compression tests on sand specimens at four different relative densities were performed, and it was confirmed that the yield surface defined as a contour of constant plastic work as measured from the origin of stress captures the behavior of soil with good accuracy in both the hardening and in the softening regime. It was observed that preshearing to peak failure produced effects similar to overconsolidation observed in clays in the region of lower confining pressures. In this region, the yield surface was found to move out beyond the failure surface for normally consolidated sand, i.e. the sand became stronger. In the region of confining pressures higher than that employed during initial preshearing to peak failure, the yield surface was located inside the failure surface and the sand was found to harden. The observed patterns of yielding is captured with good accuracy by the yield criterion employed in an existing single hardening constitutive model.

View full abstract

In this paper, the time-dependent settlement characteristics of granular fill on saturated soft foundation soil were studied using a mechanical model for the foundation. The model idealizes the behaviour of granular fill and saturated soft foundation soil with the Pasternak shear layer and spring-dashpot system respectively. The parametric studies revealed the effect of various parameters on the settlement response of the model at different stages of consolidation of the soft soil.

View full abstract

Knowledge of the small strain stiffness of the heavily overconsolidated Gault Clay, which has up to 30% calcium carbonate content, is rather limited. This has resulted in some difficulties in the analysis and design of structures constructed in the Gault. In this Technical note, the small strain stiffness of the Gault Clay is examined based on results from triaxial tests with internal small strain measurements, published geophysical data, and values deduced from the back-analysis of full-scale field observations of an excavation at Lion Yard Cambridge, U.K. Comparisons of stiffness values have also been made between the Gault Clay and the non-carbonate heavily overconsolidated London Clay. The results of these examinations have led to the conclusion that the stiffness-strain characteristic of Gault Clay is highly non-linear and exhibits first yield at a threshold shear strain of about 10^-5, beyond which the stiffness deteriorates dramatically from an initially very high value. After modest straining the stiffness reduces to values comparable to those for London Clay. The Gault Clay behaves like a low plasticity clay at small strains but as a high plasticity clay at medium to large strains. This behaviour is probably due to the breakdown of the weakly cemented bonding caused by the calcium carbonate content.

View full abstract

The results of consistency tests on marine clays were compared to investigate the influence of remolding on the liquid and plastic limits of soil. In the tests, the duration of the remolding process, in which soil is mixed with water, varied widely. It was shown that the liquid limit first decreases with increased remolding time because of a shear strength reduction during the mixing procedure. Further remolding causes the limit to increase, probably due to particle crushing. It was also shown that the plastic limit decreases as the remolding time increases. The conventional method for testing soil consistency does not define details of sample preparation procedure such as remolding time. Different test results may therefore result even for the same sample. A clear definition of the sample preparation procedure is required in order to ensure accurate test results.

View full abstract

The basic idea used in limit equilibrium method, of analysing some possible states of stresses (forces) for trial slip surfaces, is further expanded by adding consideration of possible kinematics of chosen sliding bodies divided into rigid slices (wedges), so that the local factors of safety at the bases and interfaces of the slices (wedges) could be defined. This approach enables determination of the stability of slopes in brittle soil, whose shear strength is shear displacement (strain) dependent. The analyses of instabilities of two slopes, using this method, showed that the method was capable of predicting the progressive types of failures, which occurred when brittle soil was loaded non-uniformly.

View full abstract