A capillary model was developed to predict under what conditions piping versus plugging is likely to occur as a result of clay particle dispersion in soils. In the intermediate size range, i.e., particles too small per se to clog pore constrictions and too large to wash through unimpeded, the outcome depends strongly on the concentration of dispersed particles in the seepage stream. This concentration in turn depends upon the rate at which the particles are eroded from the pore walls. The capillary model was used to ascertain the relationship between key variables at the onset of piping. This relationship involved the initial permeability and porosity of the soil, the critical tractive stress, and the hydraulic gradient. The model was also used to predict under what circumstances clay particle concentrations in a seepage stream will buildup suddenly to very high values, a condition favorable to clogging or particle holdup. Results of the analysis showed that very large particle concentration buildups tend to occur at permeabilities of less than ${10}^{-9}{\mathrm{cm}}^2$ ${10}^{-4}\mathrm{cm}/s$ for water at $20°C)\text{.}$ This prediction is consistent with previous findings reported in the geotechnical literature.

Methods used to estimate strength and deformability properties of discontinuous rock can be categorized into: (1) The direct testing method; (2) the empirical correlation method; and (3) the analytical decomposition model method. In this paper, a numerical method that combines analytical decomposition modeling with statistical simulation is presented. The method consists of: (1) Probabilistic modeling of discontinuity geometry; (2) generation of discontinuities in rock blocks by Monte‐Carlo simulation; and (3) finite element analysis of simulated rock blocks. Strength and deformability properties of intact rock and discontinuities are needed for the finite element analysis. Results of finite element analysis can be used to obtain statistical properties of the rock mass. The method was used to study the mass properties of a shale that contains slickensides. At a block size of 30 in. X 30 in. (25 cm), the mean value of the ratio between shale mass strength and intact shale strength was found to be about 0.6. For the same block size the mean value of the ratio between shale mass modulus and intact shale modulus was found to be around 0.45. At the same block size, Poisson's ratio of shale mass was found to be about two times the Poisson's ratio of intact shale. For ten simulations, the coefficient of variations of all three mass parameters were found to be less than 0.21.

Tests have been performed on triaxial samples of natural and remolded plastic clay from Winnipeg, Canada, to examine the way natural samples conform with modern conceptual models of soil behavior based largely on remolded samples. The clay contains mixed illites and montmorillonites. In its field condition, it is layered, fissured, overconsolidated, and contains many inclusions. Natural samples were carefully trimmed from blocks of the clay: Remolded samples were reconsolidated from a slurry. All reconsolidation was onedimensional. Normal consolidation and critical state lines for the two sets of samples were parallel, with natural values at higher specific volumes than remolded samples. Post‐yield compressibilities and pore‐water pressures were somewhat higher in natural samples, but strengths and moduli were lower. Yield stresses were similar, but the natural yields were again at higher specific volumes. Geotechnical and geological evidence suggests that the natural clay is cemented. Many fundamental features of critical state soil mechanics were found in the natural samples. Differences were in detail and not in principle.

The following investigations are presented: (1) Comparison between the results of 2D nonlinear and 3D nonlinear dynamic finite element analyses of an earth dam subject to two very different input ground motions; and (2) comparison between measured and computed earthquake responses of the dam. The study is based on rigorous nonlinear hysteretic analyses utilizing a multi‐surface plasticity theory. The backbone shear stress‐strain curve is assumed hyperbolic and symmetrical about the origin. Detailed comparisons of induced stresses, strains, accelerations, and permanent deformations at various locations in the dam are presented. The effects of three‐dimensionality on the dynamic response, particularly on resulting permanent deformations, are assessed. The suitability of 2D analyses in determining the dynamic behavior of such structures is evaluated.

Based on data obtained from seventy‐nine large‐scale one‐dimensional laboratory tests, limiting conditions were established to define the transition zone between retention and elutriation of chemical grout injected into a cohesionless soil permeated by water. Various combinations of four grouts, four sands, and one gravel were included in the experimental program, and the interpretation is presented in terms of seepage velocity versus gel time relationships. The position and extent of the transition zone are strongly dependent on the viscosity and miscibility of the grout and to a lesser extent, on the effective grain size of the soil. These parameters are largely responsible for the dilution that occurs and that, in turn, lengthens the gel time and decreases the viscosity, thereby governing the distribution of the injected grout and the resulting permeability of the soil. Even if a water seal is established, the grout may be dissolved or eroded, or both, after a relatively short time. Measurements from a limited number of tests demonstrate that the strength is substantially higher in the vicinity of the injection point.

Probabilistic seismic risk analysis is a promising method for evaluating design options and establishing seismic design parameters. However, there have been few examples in the literature to guide practitioners in its use. This paper demonstrates the value of risk analysis for mine tailings dams and provides a case‐history application for a seismically active portion of Nevada. Risk analysis provided the basis for selecting among design options having varying liquefaction resistance, and for establishing input parameters for dynamic analysis. Ranges are presented for the quantity and cleanup cost of tailings released in seismic failures to aid in determining expected failure consequences. It is shown that for many tailings dams, accepted lifetime failure probabilities of a few percent may provide a reasonable basis for probabilistic determination of seismic design criteria.

From the considerations of the truncated diffuse double layer due to partial saturation, an equilibrium equation is written in terms of the soil state parameter viz. void ratio, e, void ratio at liquid limit, e=wG, degree of saturation, S and external applied stress, p. The type of fine grained soil is subdued by its liquid limit. The viability of the formulated approach has been experimentally verified. This approach, essentially eliminates the difficulty in the measurement of pore air pressure, u pore water pressure, u and computation of ν parameter.

A discrete element method is described for the analysis of single piles subjected to torsion. The soil is modelled using the modulus of subgrade reaction, which is expressed in terms of the soil shear modulus and the pile radius. Discrete element matrices are listed for the pile, and the soil that can be readily implemented into a computer program. The accuracy of the discrete element approach is verified by comparison with analytical and finite element solutions. The nonlinear response of the pile-soil system is computed, and compared with two reported torsional load tests in the field.