A suction‐potential model based on the diffusion equation is proposed whereby the suction is predicted at all times and depths in terms of a parameter K and the initial values of suction ${\beta }_0$ The model shows that the suction ratio decays exponentially with time. The exponential decay rate is not uniform throughout the expansive soil, but it increases with decreasing distance to the water source. Analysis reveals that the prediction error will not exceed 25% even if K value increases by 100%, suggesting the validity of the model in providing an adequate means for suction determination. Furthermore, close agreement is found between laboratory observations and the theoretical behavior. However, the theoretical estimate of the average suction potential does not agree well with the field observation. The discrepancies may be attributed to the heterogeneities of the soil and to the simplified boundary conditions adopted in the theoretical model.

A one‐dimensional model of the sedimentation and desiccation of high‐clay‐content, subaerially deposited tailings is described. The approach is semi‐empirical and predicts average density and water content of a layer of tailings during sedimentation, first‐stage evaporation, and second‐stage evaporation, during which desiccation occurs. The method employs both geomechanical and soil physics principles and allows for varying evaporation, varying rainfall, and variation of initial height and solids content. Five tailings are analyzed, and model parameters are determined for each. Comparisons of recent experimental field data and the predictive model show good agreement. The laboratory equipment developed to provide parameters for analysis and methods used to obtain model parameters for other tailings are described. The model allows the effects of different initial layer depths, different initial water contents, and different available evaporation potentials to be compared with total drying time and resulting final dry density. Some examples of this are shown.

Design of tunnels constructed with the new Austrian tunneling method (NATM) is considered. With this method the tunnel is generally excavated using hand‐mining equipment and immediately supported with a continuous shell of shotcrete. The shotcrete acts as a preliminary support during the construction. A liner made of concrete is usually installed later to ensure the long‐term stability of the tunnel wall. The construction process of NATM tunnels is analyzed using the finite element method. The softening technique, currently used for the analysis of NATM tunnels, is improved and applied to a parametric study of ovoid transportation tunnels constructed in undrained clays. Based on this study, a simplified method is provided for the preliminary design of the NATM tunnel support. The method allows the estimation of moments and thrusts acting on the liner for a range of commonly encountered conditions. Design values are shown to be primarily dependent on soil deformation properties, initial stress conditions, and amounts of stress release in the ground prior to shotcreting.

Field experiments are conducted on two groups of piles supporting rigid foundations and subjected to harmonic loading. The objective is to investigate the ability of linear elastic theories of pile‐group modeling to predict the responsecurve characteristics including the resonant frequency and amplitude. The first pile group consists of six steel pipes of length 3.05 m and outside diameter of 0.102 m driven in a layer of cohesionless soil. The second group consists of six cast‐in‐place concrete piles 7 m in length and 0.32 m in diameter constructed in a clay site. The response is measured at different levels of excitation force amplitudes. Moderate nonlinear behavior is observed. The group stiffness and damping required to generate the theoretical response curves are established using an approximate method based on combining the plane strain soil reactions solution of the single pile with dynamic pile‐soil‐pile interaction factors. The theoretical results are also verified using the more rigorous direct analysis approach. Comparison with experiments shows that the linear theory gives a good estimate of the group stiffness but over estimates the group damping.

Complete weathering of parent rock has resulted in the formation of residual soils that cover most of the land area of Sri Lanka. The physiography of the island demands that new engineering structures be located on sloping ground, thus posing many problems. The residual soils are heterogeneous due to variable weathering of the jointed rock mass. The heterogeneity of residual soils is due to the influence of relict joints, presence of boulders, and variability of the soil matrix. The repercussion of this heterogeneity is that properties of small samples of residual soils are unrepresentative of the mass, thus rendering conventional methods of field testing, laboratory testing, and analysis procedures for slope stability unsuitable. Heretofore, there have been no testing and analytical models available to represent residual soils. This paper describes an alternative method to predict shear strength parameters of residual soils by a correlation factor obtained from a multistage triaxial test and conventional slope stability analysis. Three landslides were investigated and analyzed to obtain the field shear strength parameters using the back analysis technique. On comparison with laboratory strength parameters, obtained from multistage triaxial tests, a correlation factor was ascertained. This correlation factor can be used to modify laboratory shear‐strength parameters to obtain field shear‐strength values of residual soils for design purposes.

When an unsupported chunk of compacted clay is placed in water, it will generally completely slake apart. However, if the sample is compacted over optimum and after the sample is allowed to cure, it may not slake apart. For the unsupported compacted clay sample, the horizontal total stress is zero, and if the sample is sheared at a very slow rate to enable the dissipation of negative pore pressures, then the effective cohesion can be determined. The effective cohesion for compacted silty clay varied from 8 psf (0.4 kPa) for one day of curing to 45 psf (2.2 kPa) for 21 days of curing. Curing allows for the development of bonds, yielding reduced swelling and higher effective cohesion. The effective shear strength envelope for the compacted silty clay at high normal stresses can be described by the linear relationship defined by the angle $\Phi {}^{\prime }=28°$ and $c{}^{\prime }=265$ psf (12.7 kPa). But in the low normal effective stress range, the strength envelope is curved with a significantly lower effective cohesion intercept.

Studies on compressibility and shear strength aspects are the concern of many investigators concerned with partly saturated soils. In soil engineering connected with partly saturated soils, there are no approaches connecting soil states and stress conditions. The present investigation is essentially a step in this direction. A generalized state parameter, identified with regard to material states is shown to be related to the compressibility and shear strength. The involved parameters are simple and normally determined in routine investigations. The advantage of this approach is that changes in soil states due to external stress conditions and the associated changes in strength can be examined particularly when different types of soils are involved.

In a clay-water-electrolyte system, there exists a double-layer repulsive force between the clay particles. In a recent study, the influence of the nonparallel orientation and the finite size of the clay particles are systematically studied by solving the nonlinear governing equation by a finite element numerical technique. In this note, a set of empirical formulas are presented for the magnitude and location of the net repulsive force and shown to represent the numerical data reasonably well.