The seismic stability of an embankment resting on saturated sand layers was studied by carrying out shaking table tests using large models. This paper is concerned with the failure mode of embankments during earthquakes ; the slope stability analysis method used here is based on the assumption of failure along a circular arc (Circular Arc Analysis). For each test a large model of an embankment on a saturated sand layer was placed in a shaking table box (8 m in length, 4 m in width and 2 m in height). The failure of the embankment was investigated by measuring the rapid rapid increase in residual settlement and lateral displacement in accordance with the increase in table acceleration. The pore pressure ratio was found to be greater than 0.5 when significant settlement and lateral displacement of the embankment took place, and the failure plane was determined by locating the breaks in the horizontal white sand layers (5 mm in thickness) in a cross-section of the embankment. The test results were in good agreement with the results calculated by a method of circular arc analysis which takes into account the effects of the seismic forces and the excess pore pressures. It was concluded that the stability of embankments on saturated sand layers during earthquakes can be evaluated by using this type of circular arc analysis method.

View full abstract

Horizontal plate anchors are widely used in civil engineering constructions to resist vertical pulling loads. The available theoretical model suggested by Vesic to estimate the breakout resistance of plate anchors is in significant error, particularly for dense sandy soils. A theoretical model assuming a curved surface of failure through the surrounding soil, to evaluate the ultimate breakout resistance of horizontal plate anchors, has been proposed. It successfully brings out the concept of characteristics relative depth beyond which the breakout factor reaches a constant final value. It is capable of predicting breakout factors for wide range of values of angle of shearing resistance of sand.

View full abstract

Saturated sandy soils have been observed to liquefy during earthquakes and in laboratory experiments. Following liquefaction the soil grains settle out, and the material solidifies from the base up. The history of pore pressure in a solidifying liquefied sand has been obtained in centrifuge experiments. It is found that the excess pore pressure lags the pore pressure decay which would be expected from the solidification of a mass of rigid grains. Consequently the accreting solid sand layer is consolidating with time during and following accretion. A mathematical description of this process is given by an adaptation of Gibson's (1958) theory of consolidation of a sedimenting clay layer. The resulting theory is compared with experimental centrifuge results, and found to give a satisfactory description of the observations.

View full abstract

During construction period of a filldam the soil medium is subjected to varied confining pressues by the placement of the additional fills due to step by step construction of the dam. Therefore, it is necessary to evaluate the influence of the confining pressure on permeability coefficients of the core material. This has been accomplished by developing a permeability testing apparatus, making use of a triaxial cell.A series of permeability tests were carried out on filldam core material and the corresponding results are presented in this paper. The test results showed that the confining pressure has a significant influence on the permeability coefficient of the compacted cohesive soil. Thus, it is very important that the influence of the confining pressure on the permeability coefficients should be considered by the construction engineers while they construct the embankments.

View full abstract

A computer simulation model has been developed to perform a probabilistic modeling of rockfill shear strength in the form of friction angle parameter for Khao Laem Dam in western Thailand. The modeling was carried out by using Monte Carlo techniques obeying the nearest-neighbor autoregressive space law. The domain was divided into a set of equisized blocks with a single permeability value assigned to each block. Porosity value for each block was obtained by cross-correlation with the permeability value using the bivariate normal density function. From the porosity values, Barton's Equivalent Roughness Method was applied to get the corresponding friction angle values of the rockfill material. A series of porosity realization gave a series of friction angle values of the rockfill. Using Simplified Bishop Method, the values of the factor of safety for both plane strain and three-dimensional slope stability analyses were obtained corresponding to each set of friction angle values in each realization. The probability of failure was computed from the mean and variance of the factor of safety and was based on a factor of safety of less than 1. The results show that based on three arbitrary failure surfaces, the dam is safe with large values of factor of safety and low values of probability of failure. The higher the values of both the autoregressive parameter and correlation factor, the higher the probability of failure using the two-dimensional analysis. The same trend was also obtained in the three-dimensional analysis except that the magnitudes of the probability of failure were higher.

View full abstract

Drained true triaxial tests on a normally consolidated clay under a constant mean principal stress condition are performed using a newly made apparatus in which two principal stresses are applied by rigid plates and the other principal stress by a chamber water pressure. After checking the accuracy of the tests, the deformation and strength characteristics of clay in three-dimensional stresses are experimentally discussed, and the test results are compared with the analytical results by the well-known Cam-clay model and those by the t_ij-clay model which has recently been developed with reference to the extended concept of "Spatially Mobilized Plane". Although the Cam-clay model does not explain uniquely the observed shear behavior of clay in three-dimensional stresses, the t_ij-clay model takes into account adequately the influence of the intermediate principal stress on the behavior of clay and can describe the deformation and strength characteristics of clay with unified values of soil parameters which are easily determined from a conventional triaxial test and others.

View full abstract

The deformation and strength characteristics of sand was studied by means of the torsional shear tests on hollow cylindrical samples with a height of 20 cm, an outer radius of 5 cm and an inner radius of 3 cm. It was found that in order to obtain accurate results it is essential both to very precisely measure the axial load, the torque and the effective confining pressure σ_c' and to correct measured stress values for the membrane forces. The methods for this purpose are described in detail. Each test was performed at a constant confining stress σ_c' between 0.3∼2.0 kgf/cm² (29.4∼196.2 kN/m²) and σ₃' at failure ranged between 0.08∼0.75 kgf/cm² (7.9∼73.7 kN/m²).It was found by measuring local strain distributions throughout the sample height that the deformation is rather uniform when shear distortion γ_at is less than approximately 5%. Correspondingly the pre-peak stress ratio-strain relationship is rather independent of σ_c' or σ₃'. It was also found that with increasing σ_c' a shear band or shear bands start to appear at a smaller γ_at and the degree of post-peak strain softening becomes more pronounced. The dependency of ψ on σ₃' in the torsional shear tests for the range of σ₃' in this study was found small as well as in the triaxial and plane strain compression tests. A tendency was obtained that the value of ψ is slightly influenced by the deformation pattern at failure.The pre-peak stress and strain relationship by the torsional shear test was found similar to that by the plane strain compression test under similar values of b=(σ₂'-σ₃')/(σ₁'-σ₃') and the angle of bedding plane σ to the σ₁'-direction. The value of ψ by the torsional shear test was found slightly larger than or similar to the value of ψ by the plane strain compression test under similar values of b and δ depending on the stress level at which the value of ψ is compared.

View full abstract

Common procedures for slope stability analysis are essentially two dimensional (plane strain) despite the fact that observed failure mechanisms in the field possess clear three dimensional characteristics. A review of the few existing analysis procedures which are applicable to three dimensional failure mechanisms revealed some serious shortcomings such as failure to degenerate to known solutions when boundary cases are considered or dependence on constants which are extremely difficult to estimate.Presented is a three dimensional failure mechanism which is obtained through a variational limiting equilibrium approach without arbitrary assumptions. This mechanism consists of the shape of the potential slip surface and the potential direction of slip at each point on this surface. The present work deals with homogeneous and symmetrical slopes which are constrained in the 3 rd direction ; e.g. a dam in a narrow valley or a trench of finite length.The analysis results are presented in a form of stability charts relating the geometrical parameters of the slope, soil properties, and the 3-D factor of safety. Qualitatively, the results show that 3-D end effects are most pronounced in slopes made of highly cohesive soil. For cohesionless soil, no end effects exist since the critical slip surface then coincides with the furface of the slope.

View full abstract

A pile-driving analysis procedure coupled with stress-wave monitoring, is developed for friction piles in clay or argillaceous rock. Emphasis of the analysis procedure is placed on treating an elastic-perfectly plastic constitutive law of the shaft interface, within the frame-work of one-dimensional wave propagation along with the method of characteristics. The analysis procedure thus developed is applicable to two types of pile-driving analysis : one is the inverse analysis for identifying the constitutive parameters of the shaft interface from a recorded stress waveform ; and the other is the drivability analysis, in terms of the already identified material parameters, for predicting the displacements or stresses in the pile caused by any single hammer blow. Both aspects of the proposed analysis procedure are examined against laboratory pile-driving tests on a "perfect" friction pile with no point resistance. In those tests, a steel pipe pile of 2.5 cm in outer diameter is quasi-statically jacked into a remarkably homogeneous mudstone, along a slightly undersized guide-hole, and then subjected to striking by a free-falling steel bar. It is then confirmed that there is good agreement between the calculated and measured pile-driving performances.

View full abstract

A test system is devised to minimize detrimental effects of membrane penetration on the undrained shear strength of granular soils. The system constituting an electropneumatic closed-loop can compensate for volume changes caused by membrane penetration in real time so that the constant volume of soil skeleton is maintained. It can readily be coupled to any conventional static or cyclic shear test apparatus.In order to evaluate effectiveness of the system, cyclic triaxial tests were performed on specimens 75 mm in diameter either with or without the system. Resulting pore pressure responses and liquefaction strength curves for specimens with and without membrane compliance are compared, and the effects of membrane penetration are discussed. The test results revealed that, while the liquefaction characteristics of a fine sand with a mean grain size of 0.17 mm were not affected significantly by membrane penetration, the cyclic shear stress ratios to cause liquefaction for a coarse sand with a mean grain size of 1 mm were overestimated by about 30% as a result of membrane penetration. These results are consistent with an analytical estimation by other investigators, indicating satisfactory performance of the system in liquefaction tests.

View full abstract

Frictional resistance between dry sand and mild steel is studied by laboratory experiments. The significance of possible influential factors is examined with the use of the experimental design method. The considered factors include surface roughness of steel, 50 percent diameter of sand, sand type, test type (simple shear and shear box), and uniformity coefficient of sand. Among the factors, the surface roughness of steel, 50 percent diameter of sand, and sand type are found to have significant influence on the coefficient of friction. In order to evaluate the angularity of sand particles, modified roundness of sand was measured. Normalized roughness is proposed to evaluate the relative roughness of sand-steel interface. Coefficient of friction at yield show good correlations with normalized roughness. Coefficient of friction is estimated as a function of modified roundness and normalized roughness. The estimation of coefficient of friction showed good agreement with measured values.

View full abstract

A simple method of measuring the volume change of a specimen by means of a low capacity differential pressure transducer (LC-DPT) has been used for laboratory tests. Several kinds of error in this method evaluated by calibrating this method with another automated volume change measuring method by means of an electronic balance are described. It was found that if water flows in the tube located between the burette and the LC-DPT the tube friction associated with the flow of water in the system can induce the largest amount of error among other kinds of error. It was also found that this kind of error described above can be kept to a very small value by using a proper connection method of joints and tube.

View full abstract