Relation between seepage force and velocity of sand particles during sand boiling
K. Fujisawa, A. Murakami, S. Nishimura and T. Shuku
ABSTRACT: Thus far, the focus of studies on seepage failure, known as sand boiling or the piping phenomenon, has been to determine the critical hydraulic gradient or the critical seepage flow velocity. However, the transport of soil after seepage failure also needs to be well investigated in order to estimate the damage to soil structures or the ground. The purpose of this study is to experimentally investigate the relationship between the seepage force and the velocity of the sand particles during sand boiling induced by upward and horizontal seepage flows. In the experiments, silica sand is used as the test material and the migration velocities of the seepage water and the sand particles are calculated from the measured amounts of their discharge. The test results reveal that the equilibrium of the forces exerted on the sand particles, i.e., gravity, buoyancy and fluid-particle interaction, can be successfully used to estimate the velocity of the sand particles subjected to upward seepage flow and that the seepage force needed for the horizontal transport of the sand tends to decrease as the velocity of the sand particles increases.
A density-and stress-dependent elasto-plastic model for sands subjected to monotonic undrained torsional shear loading
G. Chiaro, J. Koseki and L.I. Nalin De Silva
ABSTRACT: A density- and stress-dependent elasto-plastic model for saturated sands undergoing monotonic undrained torsional shear loading is presented in this paper. The model is developed under an extended general hyperbolic equation (GHE) approach, in which the void ratio and stress level dependence upon stress-strain response of sand is incorporated. Most importantly, a state-dependent stress-dilatancy relationship is introduced to account for the effect of density on the stress ratio. Such a stress-dilatancy relation is used for modeling the excess pore water pressure generation in undrained shear conditions as the mirror effect of volumetric change in drained shear conditions. In this paper, details of the model formulation and soil parameters calibration are described. By using the proposed model, numerical simulation of monotonic undrained torsional shear tests have been carried out on Toyoura sand. The model predictions show that undrained shear behavior, described in terms of stress-strain relationship and effective stress path for both loose and dense sands, can be modeled satisfactorily by using a single set of soil parameters.
1-G model test with digital image analysis for seismic behavior of earth dam
Y. Miyanaga, A. Kobayashi and A.Murakami
ABSTRACT: This paper proposes a new experimental method using small 1-G shaking table tests to investigate the seismic behavior of an earth dam. In this research, a digital image analysis was applied to observe strain distributions during excitation. White gauge points were placed on the surface of the model and the movements of these points were analyzed via successive digital image pictures. From the displacements of the gauge points, the distributions of shear strain and volumetric strain were examined as the functions of the acceleration response of the dam body. As a result of the volumetric strain distributions, it was found that vertical tension and compression zones exist in turn, and that tensile stress was observed at the top of the model. It was estimated that the crack at the crest of the earth dam, brought about by the earthquake, was attributed to the tensile stress.
X-ray CT imaging of 3-D bearing capacity mechanism for vertically loaded shallow foundations
D. Takano, J. Otani, M. Nakamura, and R. Mokwa
ABSTRACT: The three-dimensional bearing capacity of shallow foundations has not been fully investigated because of the difficulty in visualizing and understanding the complex three-dimensional failure mechanism. In this paper, a series of model loading tests were conducted on shallow foundations with square and rectangular shaped footings X-ray CT scanning techniques were utilized to nondestructively visualize and investigate the soil behavior during the loading tests. Based on the tomographic results for four different footing geometries, both surface deformations and below ground deformations along the failure surface cross section are described and three-dimensional failure mechanisms illustrated. A gradual change or progression of the bearing capacity failure pattern was observed and quantified as the footing aspect ratio changed from a square foundation to a rectangular shape that approached plane strain conditions. It is anticipated that the process and technique developed in this study could be beneficial in understanding and quantifying the mechanics associated with other soil-structure interaction problems, especially complex problems in which the failure mechanism is difficult or impossible to predict using conventional geotechnical engineering principles.
Modeling and bending test simulations of cement treated soil
K. Kaneda, T. Tanikawa and S. Onimaru
ABSTRACT: Cement treated soil, which is commonly utilized to prevent liquefaction and/or to increase the bearing capacity of soft clay foundations, is characterized by four basic properties: 1) its strength is greater than that of untreated soil, yet less than that of concrete, 2) it exhibits nonlinear behavior close to its peak strength, 3) softening occurs after its peak strength has been exceeded, and 4) the extension strength is found while the soil is not considered. In this study, the subloading surface model introduced by Hashiguchi was incorporated into a modified Drucker-Prager criterion, and undrained triaxial compression tests of cement treated soil were performed under constrained pressures of 0.1 and 0.4 MN/m2, after which bending tests and simulations were performed. The numerical results of these tests agreed well with the actual results of element wise and boundary condition testing.
Modelling viscous effects during and after construction in London Clay
S. D. Clarke and C. C. Hird
ABSTRACT: A new approach to allow the modelling of the viscous behaviour of clay soils has recently been developed (Clarke & Hird, 2012) based on the BRICK constitutive model (Simpson, 1992). In this approach viscous effects, such as creep and stress relaxation, and the effects of strain history on soil stiffness are modelled within a single conceptual framework. The developed model, SRD (strain-rate dependent) BRICK, has been incorporated in a finite element program, allowing boundary value problems to be analysed. In this paper two case histories involving London Clay, where viscous effects possibly had an influence on the observed displacements, are back-analysed. These cases are the Jubilee Line extension at St James’s Park (Standing et al., 1996, Nyren et al., 2001) and a deep basement at Horseferry Road (May, 1975, Chapman, 1999). The results of the numerical modelling show that, in each case, the SRD BRICK model is able to achieve closer agreement with the recorded displacements when compared with the unmodified BRICK model.
Modelling viscous effects during and after construction in London Clay
S. D. Clarke and C. C. Hird
ABSTRACT: A new approach to allow the modelling of the viscous behaviour of clay soils has recently been developed (Clarke & Hird, 2012) based on the BRICK constitutive model (Simpson, 1992). In this approach viscous effects, such as creep and stress relaxation, and the effects of strain history on soil stiffness are modelled within a single conceptual framework. The developed model, SRD (strain-rate dependent) BRICK, has been incorporated in a finite element program, allowing boundary value problems to be analysed. In this paper two case histories involving London Clay, where viscous effects possibly had an influence on the observed displacements, are back-analysed. These cases are the Jubilee Line extension at St James’s Park (Standing et al., 1996, Nyren et al., 2001) and a deep basement at Horseferry Road (May, 1975, Chapman, 1999). The results of the numerical modelling show that, in each case, the SRD BRICK model is able to achieve closer agreement with the recorded displacements when compared with the unmodified BRICK model.
